{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 5.0, "eval_steps": 500, "global_step": 18790, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.00026609898882384245, "grad_norm": 5.56785774230957, "learning_rate": 0.0, "loss": 0.4847, "step": 1 }, { "epoch": 0.0005321979776476849, "grad_norm": 5.652049541473389, "learning_rate": 5.319148936170212e-10, "loss": 0.4737, "step": 2 }, { "epoch": 0.0007982969664715274, "grad_norm": 5.8579864501953125, "learning_rate": 1.0638297872340424e-09, "loss": 0.4891, "step": 3 }, { "epoch": 0.0010643959552953698, "grad_norm": 5.88652229309082, "learning_rate": 1.5957446808510637e-09, "loss": 0.4833, "step": 4 }, { "epoch": 0.0013304949441192123, "grad_norm": 5.943056106567383, "learning_rate": 2.1276595744680848e-09, "loss": 0.478, "step": 5 }, { "epoch": 0.0015965939329430547, "grad_norm": 5.530908584594727, "learning_rate": 2.6595744680851063e-09, "loss": 0.4894, "step": 6 }, { "epoch": 0.0018626929217668972, "grad_norm": 5.745955467224121, "learning_rate": 3.1914893617021273e-09, "loss": 0.4908, "step": 7 }, { "epoch": 0.0021287919105907396, "grad_norm": 5.8021559715271, "learning_rate": 3.7234042553191484e-09, "loss": 0.4887, "step": 8 }, { "epoch": 0.002394890899414582, "grad_norm": 5.883655548095703, "learning_rate": 4.2553191489361695e-09, "loss": 0.4822, "step": 9 }, { "epoch": 0.0026609898882384245, "grad_norm": 5.554807186126709, "learning_rate": 4.787234042553191e-09, "loss": 0.4862, "step": 10 }, { "epoch": 0.002927088877062267, "grad_norm": 5.897463798522949, "learning_rate": 5.3191489361702125e-09, "loss": 0.5158, "step": 11 }, { "epoch": 0.0031931878658861094, "grad_norm": 5.717147350311279, "learning_rate": 5.851063829787234e-09, "loss": 0.4868, "step": 12 }, { "epoch": 0.003459286854709952, "grad_norm": 5.318289279937744, "learning_rate": 6.382978723404255e-09, "loss": 0.5073, "step": 13 }, { "epoch": 0.0037253858435337944, "grad_norm": 5.5764923095703125, "learning_rate": 6.914893617021276e-09, "loss": 0.5023, "step": 14 }, { "epoch": 0.003991484832357637, "grad_norm": 5.383755207061768, "learning_rate": 7.446808510638297e-09, "loss": 0.49, "step": 15 }, { "epoch": 0.004257583821181479, "grad_norm": 5.777388095855713, "learning_rate": 7.97872340425532e-09, "loss": 0.4932, "step": 16 }, { "epoch": 0.004523682810005322, "grad_norm": 5.648228645324707, "learning_rate": 8.510638297872339e-09, "loss": 0.4755, "step": 17 }, { "epoch": 0.004789781798829164, "grad_norm": 5.672698020935059, "learning_rate": 9.042553191489362e-09, "loss": 0.4905, "step": 18 }, { "epoch": 0.005055880787653007, "grad_norm": 5.815916061401367, "learning_rate": 9.574468085106382e-09, "loss": 0.4642, "step": 19 }, { "epoch": 0.005321979776476849, "grad_norm": 5.656456470489502, "learning_rate": 1.0106382978723404e-08, "loss": 0.5057, "step": 20 }, { "epoch": 0.0055880787653006915, "grad_norm": 5.719663619995117, "learning_rate": 1.0638297872340425e-08, "loss": 0.4937, "step": 21 }, { "epoch": 0.005854177754124534, "grad_norm": 5.814630031585693, "learning_rate": 1.1170212765957447e-08, "loss": 0.4732, "step": 22 }, { "epoch": 0.0061202767429483764, "grad_norm": 5.615434646606445, "learning_rate": 1.1702127659574468e-08, "loss": 0.4769, "step": 23 }, { "epoch": 0.006386375731772219, "grad_norm": 5.378471374511719, "learning_rate": 1.2234042553191488e-08, "loss": 0.5018, "step": 24 }, { "epoch": 0.006652474720596061, "grad_norm": 6.246951103210449, "learning_rate": 1.276595744680851e-08, "loss": 0.4972, "step": 25 }, { "epoch": 0.006918573709419904, "grad_norm": 5.779706001281738, "learning_rate": 1.3297872340425531e-08, "loss": 0.5052, "step": 26 }, { "epoch": 0.007184672698243746, "grad_norm": 5.759213924407959, "learning_rate": 1.3829787234042552e-08, "loss": 0.4807, "step": 27 }, { "epoch": 0.007450771687067589, "grad_norm": 5.642969131469727, "learning_rate": 1.4361702127659576e-08, "loss": 0.4701, "step": 28 }, { "epoch": 0.007716870675891432, "grad_norm": 5.513338088989258, "learning_rate": 1.4893617021276594e-08, "loss": 0.4887, "step": 29 }, { "epoch": 0.007982969664715274, "grad_norm": 5.688876152038574, "learning_rate": 1.5425531914893615e-08, "loss": 0.483, "step": 30 }, { "epoch": 0.008249068653539117, "grad_norm": 6.074993133544922, "learning_rate": 1.595744680851064e-08, "loss": 0.5118, "step": 31 }, { "epoch": 0.008515167642362959, "grad_norm": 5.802435398101807, "learning_rate": 1.6489361702127658e-08, "loss": 0.513, "step": 32 }, { "epoch": 0.008781266631186802, "grad_norm": 5.582295894622803, "learning_rate": 1.7021276595744678e-08, "loss": 0.4769, "step": 33 }, { "epoch": 0.009047365620010643, "grad_norm": 6.195054531097412, "learning_rate": 1.75531914893617e-08, "loss": 0.5112, "step": 34 }, { "epoch": 0.009313464608834487, "grad_norm": 5.786961555480957, "learning_rate": 1.8085106382978724e-08, "loss": 0.4701, "step": 35 }, { "epoch": 0.009579563597658328, "grad_norm": 5.5109333992004395, "learning_rate": 1.8617021276595744e-08, "loss": 0.4808, "step": 36 }, { "epoch": 0.009845662586482172, "grad_norm": 6.174668312072754, "learning_rate": 1.9148936170212764e-08, "loss": 0.5257, "step": 37 }, { "epoch": 0.010111761575306013, "grad_norm": 5.861639976501465, "learning_rate": 1.9680851063829787e-08, "loss": 0.4932, "step": 38 }, { "epoch": 0.010377860564129857, "grad_norm": 5.584140777587891, "learning_rate": 2.0212765957446807e-08, "loss": 0.5132, "step": 39 }, { "epoch": 0.010643959552953698, "grad_norm": 5.957421779632568, "learning_rate": 2.0744680851063827e-08, "loss": 0.5015, "step": 40 }, { "epoch": 0.010910058541777541, "grad_norm": 5.5026068687438965, "learning_rate": 2.127659574468085e-08, "loss": 0.4834, "step": 41 }, { "epoch": 0.011176157530601383, "grad_norm": 5.716446399688721, "learning_rate": 2.1808510638297873e-08, "loss": 0.5099, "step": 42 }, { "epoch": 0.011442256519425226, "grad_norm": 5.414938926696777, "learning_rate": 2.2340425531914893e-08, "loss": 0.5028, "step": 43 }, { "epoch": 0.011708355508249068, "grad_norm": 5.497702598571777, "learning_rate": 2.2872340425531913e-08, "loss": 0.4783, "step": 44 }, { "epoch": 0.011974454497072911, "grad_norm": 5.60296630859375, "learning_rate": 2.3404255319148936e-08, "loss": 0.4894, "step": 45 }, { "epoch": 0.012240553485896753, "grad_norm": 5.996121406555176, "learning_rate": 2.393617021276596e-08, "loss": 0.4873, "step": 46 }, { "epoch": 0.012506652474720596, "grad_norm": 5.792969703674316, "learning_rate": 2.4468085106382976e-08, "loss": 0.4921, "step": 47 }, { "epoch": 0.012772751463544438, "grad_norm": 5.7287821769714355, "learning_rate": 2.5e-08, "loss": 0.527, "step": 48 }, { "epoch": 0.013038850452368281, "grad_norm": 5.846191883087158, "learning_rate": 2.553191489361702e-08, "loss": 0.4918, "step": 49 }, { "epoch": 0.013304949441192123, "grad_norm": 5.831531047821045, "learning_rate": 2.6063829787234042e-08, "loss": 0.483, "step": 50 }, { "epoch": 0.013571048430015966, "grad_norm": 5.3118720054626465, "learning_rate": 2.6595744680851062e-08, "loss": 0.4698, "step": 51 }, { "epoch": 0.013837147418839808, "grad_norm": 5.804596424102783, "learning_rate": 2.712765957446808e-08, "loss": 0.4896, "step": 52 }, { "epoch": 0.014103246407663651, "grad_norm": 5.262639999389648, "learning_rate": 2.7659574468085105e-08, "loss": 0.4825, "step": 53 }, { "epoch": 0.014369345396487493, "grad_norm": 5.583193302154541, "learning_rate": 2.8191489361702125e-08, "loss": 0.4706, "step": 54 }, { "epoch": 0.014635444385311336, "grad_norm": 5.795779705047607, "learning_rate": 2.872340425531915e-08, "loss": 0.494, "step": 55 }, { "epoch": 0.014901543374135177, "grad_norm": 5.4368767738342285, "learning_rate": 2.925531914893617e-08, "loss": 0.4837, "step": 56 }, { "epoch": 0.01516764236295902, "grad_norm": 5.464730262756348, "learning_rate": 2.9787234042553187e-08, "loss": 0.504, "step": 57 }, { "epoch": 0.015433741351782864, "grad_norm": 6.122529029846191, "learning_rate": 3.031914893617021e-08, "loss": 0.5072, "step": 58 }, { "epoch": 0.015699840340606706, "grad_norm": 5.8557915687561035, "learning_rate": 3.085106382978723e-08, "loss": 0.4862, "step": 59 }, { "epoch": 0.015965939329430547, "grad_norm": 5.745038986206055, "learning_rate": 3.138297872340425e-08, "loss": 0.482, "step": 60 }, { "epoch": 0.016232038318254392, "grad_norm": 5.894466876983643, "learning_rate": 3.191489361702128e-08, "loss": 0.5011, "step": 61 }, { "epoch": 0.016498137307078234, "grad_norm": 6.063644886016846, "learning_rate": 3.2446808510638297e-08, "loss": 0.4986, "step": 62 }, { "epoch": 0.016764236295902075, "grad_norm": 5.401123523712158, "learning_rate": 3.2978723404255316e-08, "loss": 0.4944, "step": 63 }, { "epoch": 0.017030335284725917, "grad_norm": 5.864663124084473, "learning_rate": 3.351063829787234e-08, "loss": 0.5101, "step": 64 }, { "epoch": 0.017296434273549762, "grad_norm": 5.265594959259033, "learning_rate": 3.4042553191489356e-08, "loss": 0.4866, "step": 65 }, { "epoch": 0.017562533262373604, "grad_norm": 5.903389930725098, "learning_rate": 3.457446808510638e-08, "loss": 0.4832, "step": 66 }, { "epoch": 0.017828632251197445, "grad_norm": 5.6586384773254395, "learning_rate": 3.51063829787234e-08, "loss": 0.4822, "step": 67 }, { "epoch": 0.018094731240021287, "grad_norm": 5.437417984008789, "learning_rate": 3.563829787234042e-08, "loss": 0.4793, "step": 68 }, { "epoch": 0.018360830228845132, "grad_norm": 6.054892063140869, "learning_rate": 3.617021276595745e-08, "loss": 0.4926, "step": 69 }, { "epoch": 0.018626929217668974, "grad_norm": 5.528632640838623, "learning_rate": 3.670212765957447e-08, "loss": 0.4819, "step": 70 }, { "epoch": 0.018893028206492815, "grad_norm": 5.840620040893555, "learning_rate": 3.723404255319149e-08, "loss": 0.5038, "step": 71 }, { "epoch": 0.019159127195316657, "grad_norm": 5.364729881286621, "learning_rate": 3.776595744680851e-08, "loss": 0.5027, "step": 72 }, { "epoch": 0.019425226184140502, "grad_norm": 5.7290191650390625, "learning_rate": 3.829787234042553e-08, "loss": 0.4929, "step": 73 }, { "epoch": 0.019691325172964343, "grad_norm": 5.830262184143066, "learning_rate": 3.882978723404255e-08, "loss": 0.5234, "step": 74 }, { "epoch": 0.019957424161788185, "grad_norm": 5.639779090881348, "learning_rate": 3.9361702127659574e-08, "loss": 0.4788, "step": 75 }, { "epoch": 0.020223523150612027, "grad_norm": 5.674466609954834, "learning_rate": 3.9893617021276594e-08, "loss": 0.4963, "step": 76 }, { "epoch": 0.02048962213943587, "grad_norm": 5.487886428833008, "learning_rate": 4.0425531914893614e-08, "loss": 0.4751, "step": 77 }, { "epoch": 0.020755721128259713, "grad_norm": 5.2764668464660645, "learning_rate": 4.095744680851064e-08, "loss": 0.4766, "step": 78 }, { "epoch": 0.021021820117083555, "grad_norm": 5.847195625305176, "learning_rate": 4.1489361702127654e-08, "loss": 0.4995, "step": 79 }, { "epoch": 0.021287919105907396, "grad_norm": 5.382130146026611, "learning_rate": 4.202127659574468e-08, "loss": 0.4824, "step": 80 }, { "epoch": 0.02155401809473124, "grad_norm": 5.73451042175293, "learning_rate": 4.25531914893617e-08, "loss": 0.4852, "step": 81 }, { "epoch": 0.021820117083555083, "grad_norm": 5.49914026260376, "learning_rate": 4.308510638297872e-08, "loss": 0.4713, "step": 82 }, { "epoch": 0.022086216072378925, "grad_norm": 5.4570417404174805, "learning_rate": 4.3617021276595746e-08, "loss": 0.4821, "step": 83 }, { "epoch": 0.022352315061202766, "grad_norm": 5.8995819091796875, "learning_rate": 4.4148936170212766e-08, "loss": 0.5091, "step": 84 }, { "epoch": 0.02261841405002661, "grad_norm": 5.759040832519531, "learning_rate": 4.4680851063829786e-08, "loss": 0.5088, "step": 85 }, { "epoch": 0.022884513038850453, "grad_norm": 5.772883892059326, "learning_rate": 4.5212765957446806e-08, "loss": 0.4597, "step": 86 }, { "epoch": 0.023150612027674294, "grad_norm": 5.645866870880127, "learning_rate": 4.5744680851063826e-08, "loss": 0.4936, "step": 87 }, { "epoch": 0.023416711016498136, "grad_norm": 5.615866661071777, "learning_rate": 4.6276595744680846e-08, "loss": 0.488, "step": 88 }, { "epoch": 0.02368281000532198, "grad_norm": 5.663012981414795, "learning_rate": 4.680851063829787e-08, "loss": 0.5104, "step": 89 }, { "epoch": 0.023948908994145823, "grad_norm": 5.482131481170654, "learning_rate": 4.734042553191489e-08, "loss": 0.4663, "step": 90 }, { "epoch": 0.024215007982969664, "grad_norm": 6.095068454742432, "learning_rate": 4.787234042553192e-08, "loss": 0.4975, "step": 91 }, { "epoch": 0.024481106971793506, "grad_norm": 5.749577522277832, "learning_rate": 4.840425531914893e-08, "loss": 0.4869, "step": 92 }, { "epoch": 0.02474720596061735, "grad_norm": 5.4970526695251465, "learning_rate": 4.893617021276595e-08, "loss": 0.4707, "step": 93 }, { "epoch": 0.025013304949441192, "grad_norm": 5.193124771118164, "learning_rate": 4.946808510638298e-08, "loss": 0.4623, "step": 94 }, { "epoch": 0.025279403938265034, "grad_norm": 5.294166564941406, "learning_rate": 5e-08, "loss": 0.4748, "step": 95 }, { "epoch": 0.025545502927088876, "grad_norm": 5.070937633514404, "learning_rate": 5.0531914893617024e-08, "loss": 0.4764, "step": 96 }, { "epoch": 0.02581160191591272, "grad_norm": 5.092440128326416, "learning_rate": 5.106382978723404e-08, "loss": 0.4605, "step": 97 }, { "epoch": 0.026077700904736562, "grad_norm": 5.029958248138428, "learning_rate": 5.1595744680851064e-08, "loss": 0.4726, "step": 98 }, { "epoch": 0.026343799893560404, "grad_norm": 5.134003162384033, "learning_rate": 5.2127659574468084e-08, "loss": 0.4765, "step": 99 }, { "epoch": 0.026609898882384245, "grad_norm": 5.511857032775879, "learning_rate": 5.26595744680851e-08, "loss": 0.5119, "step": 100 }, { "epoch": 0.02687599787120809, "grad_norm": 5.38721227645874, "learning_rate": 5.3191489361702123e-08, "loss": 0.5024, "step": 101 }, { "epoch": 0.027142096860031932, "grad_norm": 5.185988903045654, "learning_rate": 5.372340425531915e-08, "loss": 0.4659, "step": 102 }, { "epoch": 0.027408195848855774, "grad_norm": 5.358252048492432, "learning_rate": 5.425531914893616e-08, "loss": 0.4881, "step": 103 }, { "epoch": 0.027674294837679615, "grad_norm": 5.692956447601318, "learning_rate": 5.478723404255319e-08, "loss": 0.511, "step": 104 }, { "epoch": 0.02794039382650346, "grad_norm": 5.9050140380859375, "learning_rate": 5.531914893617021e-08, "loss": 0.5092, "step": 105 }, { "epoch": 0.028206492815327302, "grad_norm": 5.316998481750488, "learning_rate": 5.585106382978723e-08, "loss": 0.4961, "step": 106 }, { "epoch": 0.028472591804151143, "grad_norm": 5.814479827880859, "learning_rate": 5.638297872340425e-08, "loss": 0.4945, "step": 107 }, { "epoch": 0.028738690792974985, "grad_norm": 5.388632774353027, "learning_rate": 5.6914893617021276e-08, "loss": 0.4898, "step": 108 }, { "epoch": 0.02900478978179883, "grad_norm": 5.336748123168945, "learning_rate": 5.74468085106383e-08, "loss": 0.4848, "step": 109 }, { "epoch": 0.02927088877062267, "grad_norm": 5.526834964752197, "learning_rate": 5.7978723404255315e-08, "loss": 0.4647, "step": 110 }, { "epoch": 0.029536987759446513, "grad_norm": 5.35842752456665, "learning_rate": 5.851063829787234e-08, "loss": 0.4855, "step": 111 }, { "epoch": 0.029803086748270355, "grad_norm": 5.422550201416016, "learning_rate": 5.904255319148936e-08, "loss": 0.4736, "step": 112 }, { "epoch": 0.0300691857370942, "grad_norm": 5.508878231048584, "learning_rate": 5.9574468085106375e-08, "loss": 0.4928, "step": 113 }, { "epoch": 0.03033528472591804, "grad_norm": 5.709466934204102, "learning_rate": 6.01063829787234e-08, "loss": 0.4884, "step": 114 }, { "epoch": 0.030601383714741883, "grad_norm": 5.807901382446289, "learning_rate": 6.063829787234042e-08, "loss": 0.4669, "step": 115 }, { "epoch": 0.030867482703565728, "grad_norm": 5.277142524719238, "learning_rate": 6.117021276595744e-08, "loss": 0.4604, "step": 116 }, { "epoch": 0.03113358169238957, "grad_norm": 5.423722743988037, "learning_rate": 6.170212765957446e-08, "loss": 0.4879, "step": 117 }, { "epoch": 0.03139968068121341, "grad_norm": 5.302748203277588, "learning_rate": 6.22340425531915e-08, "loss": 0.4833, "step": 118 }, { "epoch": 0.03166577967003725, "grad_norm": 4.894257545471191, "learning_rate": 6.27659574468085e-08, "loss": 0.4824, "step": 119 }, { "epoch": 0.031931878658861094, "grad_norm": 5.401922702789307, "learning_rate": 6.329787234042553e-08, "loss": 0.4526, "step": 120 }, { "epoch": 0.032197977647684936, "grad_norm": 4.700743198394775, "learning_rate": 6.382978723404255e-08, "loss": 0.4636, "step": 121 }, { "epoch": 0.032464076636508785, "grad_norm": 5.1749348640441895, "learning_rate": 6.436170212765957e-08, "loss": 0.455, "step": 122 }, { "epoch": 0.032730175625332626, "grad_norm": 5.179829120635986, "learning_rate": 6.489361702127659e-08, "loss": 0.4873, "step": 123 }, { "epoch": 0.03299627461415647, "grad_norm": 5.114917278289795, "learning_rate": 6.542553191489361e-08, "loss": 0.4899, "step": 124 }, { "epoch": 0.03326237360298031, "grad_norm": 5.368479251861572, "learning_rate": 6.595744680851063e-08, "loss": 0.4686, "step": 125 }, { "epoch": 0.03352847259180415, "grad_norm": 4.985300064086914, "learning_rate": 6.648936170212765e-08, "loss": 0.4516, "step": 126 }, { "epoch": 0.03379457158062799, "grad_norm": 4.392096996307373, "learning_rate": 6.702127659574469e-08, "loss": 0.4562, "step": 127 }, { "epoch": 0.034060670569451834, "grad_norm": 4.524296283721924, "learning_rate": 6.755319148936169e-08, "loss": 0.4736, "step": 128 }, { "epoch": 0.034326769558275676, "grad_norm": 5.27925443649292, "learning_rate": 6.808510638297871e-08, "loss": 0.5036, "step": 129 }, { "epoch": 0.034592868547099524, "grad_norm": 5.100612640380859, "learning_rate": 6.861702127659575e-08, "loss": 0.4755, "step": 130 }, { "epoch": 0.034858967535923366, "grad_norm": 4.839081287384033, "learning_rate": 6.914893617021277e-08, "loss": 0.4748, "step": 131 }, { "epoch": 0.03512506652474721, "grad_norm": 4.804378032684326, "learning_rate": 6.968085106382978e-08, "loss": 0.4664, "step": 132 }, { "epoch": 0.03539116551357105, "grad_norm": 4.6754302978515625, "learning_rate": 7.02127659574468e-08, "loss": 0.4621, "step": 133 }, { "epoch": 0.03565726450239489, "grad_norm": 4.660831928253174, "learning_rate": 7.074468085106384e-08, "loss": 0.4755, "step": 134 }, { "epoch": 0.03592336349121873, "grad_norm": 4.440398693084717, "learning_rate": 7.127659574468084e-08, "loss": 0.4628, "step": 135 }, { "epoch": 0.036189462480042574, "grad_norm": 4.642533302307129, "learning_rate": 7.180851063829786e-08, "loss": 0.4611, "step": 136 }, { "epoch": 0.036455561468866415, "grad_norm": 4.2580084800720215, "learning_rate": 7.23404255319149e-08, "loss": 0.4658, "step": 137 }, { "epoch": 0.036721660457690264, "grad_norm": 4.509471893310547, "learning_rate": 7.28723404255319e-08, "loss": 0.4844, "step": 138 }, { "epoch": 0.036987759446514105, "grad_norm": 4.735282897949219, "learning_rate": 7.340425531914894e-08, "loss": 0.445, "step": 139 }, { "epoch": 0.03725385843533795, "grad_norm": 4.526485443115234, "learning_rate": 7.393617021276596e-08, "loss": 0.4859, "step": 140 }, { "epoch": 0.03751995742416179, "grad_norm": 4.536887168884277, "learning_rate": 7.446808510638298e-08, "loss": 0.4593, "step": 141 }, { "epoch": 0.03778605641298563, "grad_norm": 4.623271942138672, "learning_rate": 7.5e-08, "loss": 0.4758, "step": 142 }, { "epoch": 0.03805215540180947, "grad_norm": 4.318850994110107, "learning_rate": 7.553191489361702e-08, "loss": 0.4703, "step": 143 }, { "epoch": 0.03831825439063331, "grad_norm": 4.493073463439941, "learning_rate": 7.606382978723404e-08, "loss": 0.4594, "step": 144 }, { "epoch": 0.038584353379457155, "grad_norm": 4.542541027069092, "learning_rate": 7.659574468085106e-08, "loss": 0.4682, "step": 145 }, { "epoch": 0.038850452368281004, "grad_norm": 4.697274208068848, "learning_rate": 7.712765957446809e-08, "loss": 0.4707, "step": 146 }, { "epoch": 0.039116551357104845, "grad_norm": 4.687363624572754, "learning_rate": 7.76595744680851e-08, "loss": 0.4441, "step": 147 }, { "epoch": 0.03938265034592869, "grad_norm": 4.677285194396973, "learning_rate": 7.819148936170213e-08, "loss": 0.4643, "step": 148 }, { "epoch": 0.03964874933475253, "grad_norm": 4.607725620269775, "learning_rate": 7.872340425531915e-08, "loss": 0.4523, "step": 149 }, { "epoch": 0.03991484832357637, "grad_norm": 4.617980480194092, "learning_rate": 7.925531914893616e-08, "loss": 0.4627, "step": 150 }, { "epoch": 0.04018094731240021, "grad_norm": 4.813063621520996, "learning_rate": 7.978723404255319e-08, "loss": 0.4861, "step": 151 }, { "epoch": 0.04044704630122405, "grad_norm": 4.6284003257751465, "learning_rate": 8.031914893617021e-08, "loss": 0.4494, "step": 152 }, { "epoch": 0.040713145290047895, "grad_norm": 4.398841857910156, "learning_rate": 8.085106382978723e-08, "loss": 0.4465, "step": 153 }, { "epoch": 0.04097924427887174, "grad_norm": 4.326231479644775, "learning_rate": 8.138297872340425e-08, "loss": 0.4788, "step": 154 }, { "epoch": 0.041245343267695585, "grad_norm": 4.405913829803467, "learning_rate": 8.191489361702128e-08, "loss": 0.4408, "step": 155 }, { "epoch": 0.041511442256519426, "grad_norm": 4.578609943389893, "learning_rate": 8.24468085106383e-08, "loss": 0.4609, "step": 156 }, { "epoch": 0.04177754124534327, "grad_norm": 4.26684045791626, "learning_rate": 8.297872340425531e-08, "loss": 0.4584, "step": 157 }, { "epoch": 0.04204364023416711, "grad_norm": 4.703631401062012, "learning_rate": 8.351063829787234e-08, "loss": 0.4519, "step": 158 }, { "epoch": 0.04230973922299095, "grad_norm": 4.429041862487793, "learning_rate": 8.404255319148936e-08, "loss": 0.4422, "step": 159 }, { "epoch": 0.04257583821181479, "grad_norm": 4.423661231994629, "learning_rate": 8.457446808510638e-08, "loss": 0.4428, "step": 160 }, { "epoch": 0.042841937200638634, "grad_norm": 4.679871559143066, "learning_rate": 8.51063829787234e-08, "loss": 0.4882, "step": 161 }, { "epoch": 0.04310803618946248, "grad_norm": 4.347014904022217, "learning_rate": 8.563829787234042e-08, "loss": 0.4534, "step": 162 }, { "epoch": 0.043374135178286324, "grad_norm": 4.132262706756592, "learning_rate": 8.617021276595744e-08, "loss": 0.4639, "step": 163 }, { "epoch": 0.043640234167110166, "grad_norm": 4.149474620819092, "learning_rate": 8.670212765957446e-08, "loss": 0.4534, "step": 164 }, { "epoch": 0.04390633315593401, "grad_norm": 4.290040969848633, "learning_rate": 8.723404255319149e-08, "loss": 0.4711, "step": 165 }, { "epoch": 0.04417243214475785, "grad_norm": 4.041350841522217, "learning_rate": 8.77659574468085e-08, "loss": 0.4749, "step": 166 }, { "epoch": 0.04443853113358169, "grad_norm": 4.3781585693359375, "learning_rate": 8.829787234042553e-08, "loss": 0.4546, "step": 167 }, { "epoch": 0.04470463012240553, "grad_norm": 4.407985210418701, "learning_rate": 8.882978723404255e-08, "loss": 0.4477, "step": 168 }, { "epoch": 0.044970729111229374, "grad_norm": 4.176435947418213, "learning_rate": 8.936170212765957e-08, "loss": 0.4528, "step": 169 }, { "epoch": 0.04523682810005322, "grad_norm": 3.96480131149292, "learning_rate": 8.989361702127659e-08, "loss": 0.4614, "step": 170 }, { "epoch": 0.045502927088877064, "grad_norm": 4.399297714233398, "learning_rate": 9.042553191489361e-08, "loss": 0.4406, "step": 171 }, { "epoch": 0.045769026077700906, "grad_norm": 4.4072370529174805, "learning_rate": 9.095744680851063e-08, "loss": 0.4511, "step": 172 }, { "epoch": 0.04603512506652475, "grad_norm": 4.043140888214111, "learning_rate": 9.148936170212765e-08, "loss": 0.4764, "step": 173 }, { "epoch": 0.04630122405534859, "grad_norm": 4.217087745666504, "learning_rate": 9.202127659574468e-08, "loss": 0.4714, "step": 174 }, { "epoch": 0.04656732304417243, "grad_norm": 4.227458953857422, "learning_rate": 9.255319148936169e-08, "loss": 0.4615, "step": 175 }, { "epoch": 0.04683342203299627, "grad_norm": 4.006089210510254, "learning_rate": 9.308510638297871e-08, "loss": 0.4636, "step": 176 }, { "epoch": 0.04709952102182012, "grad_norm": 4.155174255371094, "learning_rate": 9.361702127659574e-08, "loss": 0.4463, "step": 177 }, { "epoch": 0.04736562001064396, "grad_norm": 3.9782564640045166, "learning_rate": 9.414893617021276e-08, "loss": 0.4299, "step": 178 }, { "epoch": 0.047631718999467804, "grad_norm": 3.885749340057373, "learning_rate": 9.468085106382978e-08, "loss": 0.4328, "step": 179 }, { "epoch": 0.047897817988291645, "grad_norm": 3.8034908771514893, "learning_rate": 9.52127659574468e-08, "loss": 0.4355, "step": 180 }, { "epoch": 0.04816391697711549, "grad_norm": 3.417891502380371, "learning_rate": 9.574468085106384e-08, "loss": 0.4913, "step": 181 }, { "epoch": 0.04843001596593933, "grad_norm": 3.3468968868255615, "learning_rate": 9.627659574468084e-08, "loss": 0.4387, "step": 182 }, { "epoch": 0.04869611495476317, "grad_norm": 3.467853307723999, "learning_rate": 9.680851063829786e-08, "loss": 0.4442, "step": 183 }, { "epoch": 0.04896221394358701, "grad_norm": 2.971710681915283, "learning_rate": 9.73404255319149e-08, "loss": 0.4381, "step": 184 }, { "epoch": 0.04922831293241086, "grad_norm": 3.0466606616973877, "learning_rate": 9.78723404255319e-08, "loss": 0.4109, "step": 185 }, { "epoch": 0.0494944119212347, "grad_norm": 3.2025463581085205, "learning_rate": 9.840425531914894e-08, "loss": 0.4175, "step": 186 }, { "epoch": 0.04976051091005854, "grad_norm": 2.9562184810638428, "learning_rate": 9.893617021276596e-08, "loss": 0.4308, "step": 187 }, { "epoch": 0.050026609898882385, "grad_norm": 2.9961934089660645, "learning_rate": 9.946808510638298e-08, "loss": 0.4352, "step": 188 }, { "epoch": 0.050292708887706226, "grad_norm": 2.912290096282959, "learning_rate": 1e-07, "loss": 0.4358, "step": 189 }, { "epoch": 0.05055880787653007, "grad_norm": 2.9178082942962646, "learning_rate": 1.0053191489361702e-07, "loss": 0.443, "step": 190 }, { "epoch": 0.05082490686535391, "grad_norm": 2.890637159347534, "learning_rate": 1.0106382978723405e-07, "loss": 0.4261, "step": 191 }, { "epoch": 0.05109100585417775, "grad_norm": 2.961845874786377, "learning_rate": 1.0159574468085106e-07, "loss": 0.427, "step": 192 }, { "epoch": 0.0513571048430016, "grad_norm": 2.9956369400024414, "learning_rate": 1.0212765957446807e-07, "loss": 0.4155, "step": 193 }, { "epoch": 0.05162320383182544, "grad_norm": 2.8205199241638184, "learning_rate": 1.0265957446808511e-07, "loss": 0.4334, "step": 194 }, { "epoch": 0.05188930282064928, "grad_norm": 2.470646381378174, "learning_rate": 1.0319148936170213e-07, "loss": 0.4114, "step": 195 }, { "epoch": 0.052155401809473124, "grad_norm": 2.8768670558929443, "learning_rate": 1.0372340425531913e-07, "loss": 0.441, "step": 196 }, { "epoch": 0.052421500798296966, "grad_norm": 2.7444653511047363, "learning_rate": 1.0425531914893617e-07, "loss": 0.4341, "step": 197 }, { "epoch": 0.05268759978712081, "grad_norm": 2.5937438011169434, "learning_rate": 1.0478723404255319e-07, "loss": 0.4253, "step": 198 }, { "epoch": 0.05295369877594465, "grad_norm": 2.612548589706421, "learning_rate": 1.053191489361702e-07, "loss": 0.4205, "step": 199 }, { "epoch": 0.05321979776476849, "grad_norm": 2.8994944095611572, "learning_rate": 1.0585106382978724e-07, "loss": 0.4258, "step": 200 }, { "epoch": 0.05348589675359234, "grad_norm": 2.5991554260253906, "learning_rate": 1.0638297872340425e-07, "loss": 0.4376, "step": 201 }, { "epoch": 0.05375199574241618, "grad_norm": 2.7369093894958496, "learning_rate": 1.0691489361702127e-07, "loss": 0.3917, "step": 202 }, { "epoch": 0.05401809473124002, "grad_norm": 2.572204828262329, "learning_rate": 1.074468085106383e-07, "loss": 0.4312, "step": 203 }, { "epoch": 0.054284193720063864, "grad_norm": 2.5818517208099365, "learning_rate": 1.079787234042553e-07, "loss": 0.4059, "step": 204 }, { "epoch": 0.054550292708887706, "grad_norm": 2.5691511631011963, "learning_rate": 1.0851063829787233e-07, "loss": 0.4035, "step": 205 }, { "epoch": 0.05481639169771155, "grad_norm": 2.591534376144409, "learning_rate": 1.0904255319148936e-07, "loss": 0.4049, "step": 206 }, { "epoch": 0.05508249068653539, "grad_norm": 2.398501396179199, "learning_rate": 1.0957446808510638e-07, "loss": 0.4231, "step": 207 }, { "epoch": 0.05534858967535923, "grad_norm": 2.7952494621276855, "learning_rate": 1.1010638297872339e-07, "loss": 0.4211, "step": 208 }, { "epoch": 0.05561468866418308, "grad_norm": 2.580528497695923, "learning_rate": 1.1063829787234042e-07, "loss": 0.4227, "step": 209 }, { "epoch": 0.05588078765300692, "grad_norm": 2.6125402450561523, "learning_rate": 1.1117021276595744e-07, "loss": 0.4114, "step": 210 }, { "epoch": 0.05614688664183076, "grad_norm": 2.5587363243103027, "learning_rate": 1.1170212765957446e-07, "loss": 0.4235, "step": 211 }, { "epoch": 0.056412985630654604, "grad_norm": 2.495790481567383, "learning_rate": 1.1223404255319149e-07, "loss": 0.4018, "step": 212 }, { "epoch": 0.056679084619478445, "grad_norm": 2.70658802986145, "learning_rate": 1.127659574468085e-07, "loss": 0.426, "step": 213 }, { "epoch": 0.05694518360830229, "grad_norm": 2.4774675369262695, "learning_rate": 1.1329787234042553e-07, "loss": 0.4001, "step": 214 }, { "epoch": 0.05721128259712613, "grad_norm": 2.5876331329345703, "learning_rate": 1.1382978723404255e-07, "loss": 0.4267, "step": 215 }, { "epoch": 0.05747738158594997, "grad_norm": 2.4218578338623047, "learning_rate": 1.1436170212765957e-07, "loss": 0.4118, "step": 216 }, { "epoch": 0.05774348057477382, "grad_norm": 2.5609848499298096, "learning_rate": 1.148936170212766e-07, "loss": 0.4168, "step": 217 }, { "epoch": 0.05800957956359766, "grad_norm": 2.36650013923645, "learning_rate": 1.1542553191489361e-07, "loss": 0.402, "step": 218 }, { "epoch": 0.0582756785524215, "grad_norm": 2.3312745094299316, "learning_rate": 1.1595744680851063e-07, "loss": 0.4279, "step": 219 }, { "epoch": 0.05854177754124534, "grad_norm": 2.385835886001587, "learning_rate": 1.1648936170212766e-07, "loss": 0.4321, "step": 220 }, { "epoch": 0.058807876530069185, "grad_norm": 2.5901451110839844, "learning_rate": 1.1702127659574468e-07, "loss": 0.4256, "step": 221 }, { "epoch": 0.05907397551889303, "grad_norm": 2.5791547298431396, "learning_rate": 1.1755319148936169e-07, "loss": 0.4142, "step": 222 }, { "epoch": 0.05934007450771687, "grad_norm": 2.5849697589874268, "learning_rate": 1.1808510638297872e-07, "loss": 0.4088, "step": 223 }, { "epoch": 0.05960617349654071, "grad_norm": 2.260072946548462, "learning_rate": 1.1861702127659574e-07, "loss": 0.4046, "step": 224 }, { "epoch": 0.05987227248536456, "grad_norm": 2.346299409866333, "learning_rate": 1.1914893617021275e-07, "loss": 0.4355, "step": 225 }, { "epoch": 0.0601383714741884, "grad_norm": 2.267061948776245, "learning_rate": 1.196808510638298e-07, "loss": 0.4122, "step": 226 }, { "epoch": 0.06040447046301224, "grad_norm": 2.377370834350586, "learning_rate": 1.202127659574468e-07, "loss": 0.4091, "step": 227 }, { "epoch": 0.06067056945183608, "grad_norm": 2.20328688621521, "learning_rate": 1.207446808510638e-07, "loss": 0.4168, "step": 228 }, { "epoch": 0.060936668440659925, "grad_norm": 2.4589662551879883, "learning_rate": 1.2127659574468084e-07, "loss": 0.4029, "step": 229 }, { "epoch": 0.061202767429483766, "grad_norm": 2.2482635974884033, "learning_rate": 1.2180851063829788e-07, "loss": 0.4291, "step": 230 }, { "epoch": 0.06146886641830761, "grad_norm": 2.2334094047546387, "learning_rate": 1.2234042553191488e-07, "loss": 0.4305, "step": 231 }, { "epoch": 0.061734965407131456, "grad_norm": 2.3749518394470215, "learning_rate": 1.2287234042553191e-07, "loss": 0.4255, "step": 232 }, { "epoch": 0.0620010643959553, "grad_norm": 2.14760684967041, "learning_rate": 1.2340425531914892e-07, "loss": 0.4, "step": 233 }, { "epoch": 0.06226716338477914, "grad_norm": 2.1774039268493652, "learning_rate": 1.2393617021276595e-07, "loss": 0.4026, "step": 234 }, { "epoch": 0.06253326237360297, "grad_norm": 2.316197156906128, "learning_rate": 1.24468085106383e-07, "loss": 0.4037, "step": 235 }, { "epoch": 0.06279936136242682, "grad_norm": 2.302567481994629, "learning_rate": 1.25e-07, "loss": 0.4117, "step": 236 }, { "epoch": 0.06306546035125067, "grad_norm": 2.2606041431427, "learning_rate": 1.25531914893617e-07, "loss": 0.4125, "step": 237 }, { "epoch": 0.0633315593400745, "grad_norm": 2.177029609680176, "learning_rate": 1.2606382978723403e-07, "loss": 0.392, "step": 238 }, { "epoch": 0.06359765832889835, "grad_norm": 2.3104159832000732, "learning_rate": 1.2659574468085107e-07, "loss": 0.4333, "step": 239 }, { "epoch": 0.06386375731772219, "grad_norm": 2.082763433456421, "learning_rate": 1.2712765957446807e-07, "loss": 0.4008, "step": 240 }, { "epoch": 0.06412985630654604, "grad_norm": 2.2947115898132324, "learning_rate": 1.276595744680851e-07, "loss": 0.3981, "step": 241 }, { "epoch": 0.06439595529536987, "grad_norm": 2.180763006210327, "learning_rate": 1.2819148936170211e-07, "loss": 0.3784, "step": 242 }, { "epoch": 0.06466205428419372, "grad_norm": 2.145167350769043, "learning_rate": 1.2872340425531915e-07, "loss": 0.4057, "step": 243 }, { "epoch": 0.06492815327301757, "grad_norm": 2.2236247062683105, "learning_rate": 1.2925531914893618e-07, "loss": 0.4047, "step": 244 }, { "epoch": 0.0651942522618414, "grad_norm": 2.223033905029297, "learning_rate": 1.2978723404255319e-07, "loss": 0.3983, "step": 245 }, { "epoch": 0.06546035125066525, "grad_norm": 2.083285093307495, "learning_rate": 1.303191489361702e-07, "loss": 0.4177, "step": 246 }, { "epoch": 0.06572645023948909, "grad_norm": 2.0411107540130615, "learning_rate": 1.3085106382978723e-07, "loss": 0.3814, "step": 247 }, { "epoch": 0.06599254922831294, "grad_norm": 2.009188413619995, "learning_rate": 1.3138297872340426e-07, "loss": 0.3976, "step": 248 }, { "epoch": 0.06625864821713677, "grad_norm": 2.0564379692077637, "learning_rate": 1.3191489361702127e-07, "loss": 0.4049, "step": 249 }, { "epoch": 0.06652474720596062, "grad_norm": 2.028106927871704, "learning_rate": 1.324468085106383e-07, "loss": 0.3968, "step": 250 }, { "epoch": 0.06679084619478445, "grad_norm": 2.002408981323242, "learning_rate": 1.329787234042553e-07, "loss": 0.3844, "step": 251 }, { "epoch": 0.0670569451836083, "grad_norm": 1.9447612762451172, "learning_rate": 1.335106382978723e-07, "loss": 0.4056, "step": 252 }, { "epoch": 0.06732304417243215, "grad_norm": 2.045677900314331, "learning_rate": 1.3404255319148937e-07, "loss": 0.3842, "step": 253 }, { "epoch": 0.06758914316125599, "grad_norm": 1.9212242364883423, "learning_rate": 1.3457446808510638e-07, "loss": 0.3922, "step": 254 }, { "epoch": 0.06785524215007983, "grad_norm": 1.9638829231262207, "learning_rate": 1.3510638297872338e-07, "loss": 0.3866, "step": 255 }, { "epoch": 0.06812134113890367, "grad_norm": 1.8929238319396973, "learning_rate": 1.3563829787234042e-07, "loss": 0.3978, "step": 256 }, { "epoch": 0.06838744012772752, "grad_norm": 1.784192681312561, "learning_rate": 1.3617021276595742e-07, "loss": 0.3933, "step": 257 }, { "epoch": 0.06865353911655135, "grad_norm": 1.950165867805481, "learning_rate": 1.3670212765957446e-07, "loss": 0.4036, "step": 258 }, { "epoch": 0.0689196381053752, "grad_norm": 1.873347520828247, "learning_rate": 1.372340425531915e-07, "loss": 0.3823, "step": 259 }, { "epoch": 0.06918573709419905, "grad_norm": 1.9700030088424683, "learning_rate": 1.377659574468085e-07, "loss": 0.4129, "step": 260 }, { "epoch": 0.06945183608302288, "grad_norm": 1.8691076040267944, "learning_rate": 1.3829787234042553e-07, "loss": 0.3903, "step": 261 }, { "epoch": 0.06971793507184673, "grad_norm": 1.7817296981811523, "learning_rate": 1.3882978723404256e-07, "loss": 0.392, "step": 262 }, { "epoch": 0.06998403406067057, "grad_norm": 1.7374426126480103, "learning_rate": 1.3936170212765957e-07, "loss": 0.4041, "step": 263 }, { "epoch": 0.07025013304949441, "grad_norm": 1.8507325649261475, "learning_rate": 1.398936170212766e-07, "loss": 0.3791, "step": 264 }, { "epoch": 0.07051623203831825, "grad_norm": 1.6183115243911743, "learning_rate": 1.404255319148936e-07, "loss": 0.3734, "step": 265 }, { "epoch": 0.0707823310271421, "grad_norm": 1.8673781156539917, "learning_rate": 1.4095744680851062e-07, "loss": 0.363, "step": 266 }, { "epoch": 0.07104843001596593, "grad_norm": 1.7094836235046387, "learning_rate": 1.4148936170212768e-07, "loss": 0.3748, "step": 267 }, { "epoch": 0.07131452900478978, "grad_norm": 1.5976227521896362, "learning_rate": 1.4202127659574468e-07, "loss": 0.3737, "step": 268 }, { "epoch": 0.07158062799361363, "grad_norm": 1.6192394495010376, "learning_rate": 1.425531914893617e-07, "loss": 0.4085, "step": 269 }, { "epoch": 0.07184672698243746, "grad_norm": 1.6600031852722168, "learning_rate": 1.4308510638297872e-07, "loss": 0.3886, "step": 270 }, { "epoch": 0.07211282597126131, "grad_norm": 1.4354572296142578, "learning_rate": 1.4361702127659573e-07, "loss": 0.3931, "step": 271 }, { "epoch": 0.07237892496008515, "grad_norm": 1.314693808555603, "learning_rate": 1.4414893617021276e-07, "loss": 0.3716, "step": 272 }, { "epoch": 0.072645023948909, "grad_norm": 1.5646166801452637, "learning_rate": 1.446808510638298e-07, "loss": 0.38, "step": 273 }, { "epoch": 0.07291112293773283, "grad_norm": 1.386844515800476, "learning_rate": 1.452127659574468e-07, "loss": 0.4025, "step": 274 }, { "epoch": 0.07317722192655668, "grad_norm": 1.2908846139907837, "learning_rate": 1.457446808510638e-07, "loss": 0.3695, "step": 275 }, { "epoch": 0.07344332091538053, "grad_norm": 1.5481040477752686, "learning_rate": 1.4627659574468084e-07, "loss": 0.3801, "step": 276 }, { "epoch": 0.07370941990420436, "grad_norm": 1.4766933917999268, "learning_rate": 1.4680851063829787e-07, "loss": 0.4157, "step": 277 }, { "epoch": 0.07397551889302821, "grad_norm": 1.4659122228622437, "learning_rate": 1.4734042553191488e-07, "loss": 0.3877, "step": 278 }, { "epoch": 0.07424161788185205, "grad_norm": 1.397257685661316, "learning_rate": 1.4787234042553191e-07, "loss": 0.3859, "step": 279 }, { "epoch": 0.0745077168706759, "grad_norm": 1.2352077960968018, "learning_rate": 1.4840425531914892e-07, "loss": 0.4026, "step": 280 }, { "epoch": 0.07477381585949973, "grad_norm": 1.2116451263427734, "learning_rate": 1.4893617021276595e-07, "loss": 0.3971, "step": 281 }, { "epoch": 0.07503991484832358, "grad_norm": 1.2562094926834106, "learning_rate": 1.4946808510638299e-07, "loss": 0.342, "step": 282 }, { "epoch": 0.07530601383714741, "grad_norm": 1.214053988456726, "learning_rate": 1.5e-07, "loss": 0.3818, "step": 283 }, { "epoch": 0.07557211282597126, "grad_norm": 1.2413554191589355, "learning_rate": 1.50531914893617e-07, "loss": 0.3671, "step": 284 }, { "epoch": 0.07583821181479511, "grad_norm": 1.216336965560913, "learning_rate": 1.5106382978723403e-07, "loss": 0.3771, "step": 285 }, { "epoch": 0.07610431080361894, "grad_norm": 1.1015149354934692, "learning_rate": 1.5159574468085107e-07, "loss": 0.3732, "step": 286 }, { "epoch": 0.07637040979244279, "grad_norm": 1.1399321556091309, "learning_rate": 1.5212765957446807e-07, "loss": 0.3591, "step": 287 }, { "epoch": 0.07663650878126663, "grad_norm": 1.0781142711639404, "learning_rate": 1.526595744680851e-07, "loss": 0.3803, "step": 288 }, { "epoch": 0.07690260777009048, "grad_norm": 1.1547927856445312, "learning_rate": 1.531914893617021e-07, "loss": 0.3614, "step": 289 }, { "epoch": 0.07716870675891431, "grad_norm": 1.0399833917617798, "learning_rate": 1.5372340425531915e-07, "loss": 0.3752, "step": 290 }, { "epoch": 0.07743480574773816, "grad_norm": 1.1443514823913574, "learning_rate": 1.5425531914893618e-07, "loss": 0.3382, "step": 291 }, { "epoch": 0.07770090473656201, "grad_norm": 1.063236951828003, "learning_rate": 1.5478723404255319e-07, "loss": 0.3631, "step": 292 }, { "epoch": 0.07796700372538584, "grad_norm": 1.0415939092636108, "learning_rate": 1.553191489361702e-07, "loss": 0.3646, "step": 293 }, { "epoch": 0.07823310271420969, "grad_norm": 0.9528990983963013, "learning_rate": 1.5585106382978722e-07, "loss": 0.3534, "step": 294 }, { "epoch": 0.07849920170303352, "grad_norm": 0.897293746471405, "learning_rate": 1.5638297872340426e-07, "loss": 0.3608, "step": 295 }, { "epoch": 0.07876530069185737, "grad_norm": 1.074846625328064, "learning_rate": 1.5691489361702126e-07, "loss": 0.3728, "step": 296 }, { "epoch": 0.07903139968068121, "grad_norm": 1.0918340682983398, "learning_rate": 1.574468085106383e-07, "loss": 0.3899, "step": 297 }, { "epoch": 0.07929749866950506, "grad_norm": 0.9259338974952698, "learning_rate": 1.579787234042553e-07, "loss": 0.3393, "step": 298 }, { "epoch": 0.0795635976583289, "grad_norm": 0.9076675176620483, "learning_rate": 1.585106382978723e-07, "loss": 0.3557, "step": 299 }, { "epoch": 0.07982969664715274, "grad_norm": 0.9692551493644714, "learning_rate": 1.5904255319148937e-07, "loss": 0.3731, "step": 300 }, { "epoch": 0.08009579563597659, "grad_norm": 0.9950113892555237, "learning_rate": 1.5957446808510638e-07, "loss": 0.3696, "step": 301 }, { "epoch": 0.08036189462480042, "grad_norm": 0.898563802242279, "learning_rate": 1.6010638297872338e-07, "loss": 0.3687, "step": 302 }, { "epoch": 0.08062799361362427, "grad_norm": 0.8432802557945251, "learning_rate": 1.6063829787234042e-07, "loss": 0.3432, "step": 303 }, { "epoch": 0.0808940926024481, "grad_norm": 0.921385645866394, "learning_rate": 1.6117021276595742e-07, "loss": 0.3698, "step": 304 }, { "epoch": 0.08116019159127195, "grad_norm": 0.8625635504722595, "learning_rate": 1.6170212765957446e-07, "loss": 0.3835, "step": 305 }, { "epoch": 0.08142629058009579, "grad_norm": 0.7616870999336243, "learning_rate": 1.622340425531915e-07, "loss": 0.3558, "step": 306 }, { "epoch": 0.08169238956891964, "grad_norm": 0.9189148545265198, "learning_rate": 1.627659574468085e-07, "loss": 0.3685, "step": 307 }, { "epoch": 0.08195848855774349, "grad_norm": 0.7548439502716064, "learning_rate": 1.6329787234042553e-07, "loss": 0.342, "step": 308 }, { "epoch": 0.08222458754656732, "grad_norm": 0.8144955635070801, "learning_rate": 1.6382978723404256e-07, "loss": 0.3744, "step": 309 }, { "epoch": 0.08249068653539117, "grad_norm": 0.7611618638038635, "learning_rate": 1.6436170212765957e-07, "loss": 0.3691, "step": 310 }, { "epoch": 0.082756785524215, "grad_norm": 0.7842602729797363, "learning_rate": 1.648936170212766e-07, "loss": 0.3537, "step": 311 }, { "epoch": 0.08302288451303885, "grad_norm": 0.7914226651191711, "learning_rate": 1.654255319148936e-07, "loss": 0.3737, "step": 312 }, { "epoch": 0.08328898350186269, "grad_norm": 0.7656060457229614, "learning_rate": 1.6595744680851062e-07, "loss": 0.3486, "step": 313 }, { "epoch": 0.08355508249068654, "grad_norm": 0.843612790107727, "learning_rate": 1.6648936170212767e-07, "loss": 0.3882, "step": 314 }, { "epoch": 0.08382118147951038, "grad_norm": 0.7463722825050354, "learning_rate": 1.6702127659574468e-07, "loss": 0.3622, "step": 315 }, { "epoch": 0.08408728046833422, "grad_norm": 0.744846522808075, "learning_rate": 1.675531914893617e-07, "loss": 0.3532, "step": 316 }, { "epoch": 0.08435337945715807, "grad_norm": 0.7379047870635986, "learning_rate": 1.6808510638297872e-07, "loss": 0.3738, "step": 317 }, { "epoch": 0.0846194784459819, "grad_norm": 0.6934646964073181, "learning_rate": 1.6861702127659573e-07, "loss": 0.3387, "step": 318 }, { "epoch": 0.08488557743480575, "grad_norm": 0.7452108860015869, "learning_rate": 1.6914893617021276e-07, "loss": 0.3526, "step": 319 }, { "epoch": 0.08515167642362959, "grad_norm": 0.7395684123039246, "learning_rate": 1.696808510638298e-07, "loss": 0.3512, "step": 320 }, { "epoch": 0.08541777541245343, "grad_norm": 0.7528880834579468, "learning_rate": 1.702127659574468e-07, "loss": 0.367, "step": 321 }, { "epoch": 0.08568387440127727, "grad_norm": 0.7142563462257385, "learning_rate": 1.707446808510638e-07, "loss": 0.3582, "step": 322 }, { "epoch": 0.08594997339010112, "grad_norm": 0.6994447708129883, "learning_rate": 1.7127659574468084e-07, "loss": 0.3685, "step": 323 }, { "epoch": 0.08621607237892497, "grad_norm": 0.6892454028129578, "learning_rate": 1.7180851063829787e-07, "loss": 0.3707, "step": 324 }, { "epoch": 0.0864821713677488, "grad_norm": 0.7053375840187073, "learning_rate": 1.7234042553191488e-07, "loss": 0.3532, "step": 325 }, { "epoch": 0.08674827035657265, "grad_norm": 0.5954549312591553, "learning_rate": 1.728723404255319e-07, "loss": 0.3405, "step": 326 }, { "epoch": 0.08701436934539648, "grad_norm": 0.6254027485847473, "learning_rate": 1.7340425531914892e-07, "loss": 0.3419, "step": 327 }, { "epoch": 0.08728046833422033, "grad_norm": 0.6373633742332458, "learning_rate": 1.7393617021276595e-07, "loss": 0.3399, "step": 328 }, { "epoch": 0.08754656732304417, "grad_norm": 0.6781458854675293, "learning_rate": 1.7446808510638299e-07, "loss": 0.3642, "step": 329 }, { "epoch": 0.08781266631186802, "grad_norm": 0.6428430080413818, "learning_rate": 1.75e-07, "loss": 0.3758, "step": 330 }, { "epoch": 0.08807876530069186, "grad_norm": 0.5920748114585876, "learning_rate": 1.75531914893617e-07, "loss": 0.3741, "step": 331 }, { "epoch": 0.0883448642895157, "grad_norm": 0.6037255525588989, "learning_rate": 1.7606382978723403e-07, "loss": 0.3262, "step": 332 }, { "epoch": 0.08861096327833955, "grad_norm": 0.5863408446311951, "learning_rate": 1.7659574468085106e-07, "loss": 0.3573, "step": 333 }, { "epoch": 0.08887706226716338, "grad_norm": 0.6412938237190247, "learning_rate": 1.7712765957446807e-07, "loss": 0.3572, "step": 334 }, { "epoch": 0.08914316125598723, "grad_norm": 0.5765994191169739, "learning_rate": 1.776595744680851e-07, "loss": 0.3274, "step": 335 }, { "epoch": 0.08940926024481106, "grad_norm": 0.6256553530693054, "learning_rate": 1.781914893617021e-07, "loss": 0.3482, "step": 336 }, { "epoch": 0.08967535923363491, "grad_norm": 0.6009851098060608, "learning_rate": 1.7872340425531914e-07, "loss": 0.3448, "step": 337 }, { "epoch": 0.08994145822245875, "grad_norm": 0.6103844046592712, "learning_rate": 1.7925531914893618e-07, "loss": 0.3481, "step": 338 }, { "epoch": 0.0902075572112826, "grad_norm": 0.5961686372756958, "learning_rate": 1.7978723404255318e-07, "loss": 0.3468, "step": 339 }, { "epoch": 0.09047365620010644, "grad_norm": 0.6213310360908508, "learning_rate": 1.803191489361702e-07, "loss": 0.3566, "step": 340 }, { "epoch": 0.09073975518893028, "grad_norm": 0.6063845157623291, "learning_rate": 1.8085106382978722e-07, "loss": 0.3472, "step": 341 }, { "epoch": 0.09100585417775413, "grad_norm": 0.6037337779998779, "learning_rate": 1.8138297872340426e-07, "loss": 0.3493, "step": 342 }, { "epoch": 0.09127195316657796, "grad_norm": 0.5938726663589478, "learning_rate": 1.8191489361702126e-07, "loss": 0.3631, "step": 343 }, { "epoch": 0.09153805215540181, "grad_norm": 0.6577502489089966, "learning_rate": 1.824468085106383e-07, "loss": 0.3394, "step": 344 }, { "epoch": 0.09180415114422565, "grad_norm": 0.6387424468994141, "learning_rate": 1.829787234042553e-07, "loss": 0.3783, "step": 345 }, { "epoch": 0.0920702501330495, "grad_norm": 0.5808179974555969, "learning_rate": 1.835106382978723e-07, "loss": 0.3541, "step": 346 }, { "epoch": 0.09233634912187334, "grad_norm": 0.6135080456733704, "learning_rate": 1.8404255319148937e-07, "loss": 0.3421, "step": 347 }, { "epoch": 0.09260244811069718, "grad_norm": 0.5304134488105774, "learning_rate": 1.8457446808510638e-07, "loss": 0.3445, "step": 348 }, { "epoch": 0.09286854709952103, "grad_norm": 0.556472897529602, "learning_rate": 1.8510638297872338e-07, "loss": 0.3529, "step": 349 }, { "epoch": 0.09313464608834486, "grad_norm": 0.5499981641769409, "learning_rate": 1.8563829787234042e-07, "loss": 0.3162, "step": 350 }, { "epoch": 0.09340074507716871, "grad_norm": 0.5462653636932373, "learning_rate": 1.8617021276595742e-07, "loss": 0.3297, "step": 351 }, { "epoch": 0.09366684406599254, "grad_norm": 0.5661852955818176, "learning_rate": 1.8670212765957446e-07, "loss": 0.3385, "step": 352 }, { "epoch": 0.09393294305481639, "grad_norm": 0.5285981297492981, "learning_rate": 1.872340425531915e-07, "loss": 0.3331, "step": 353 }, { "epoch": 0.09419904204364024, "grad_norm": 0.5318543910980225, "learning_rate": 1.877659574468085e-07, "loss": 0.326, "step": 354 }, { "epoch": 0.09446514103246408, "grad_norm": 0.5723559856414795, "learning_rate": 1.8829787234042553e-07, "loss": 0.3513, "step": 355 }, { "epoch": 0.09473124002128792, "grad_norm": 0.5747937560081482, "learning_rate": 1.8882978723404256e-07, "loss": 0.3515, "step": 356 }, { "epoch": 0.09499733901011176, "grad_norm": 0.5772479176521301, "learning_rate": 1.8936170212765957e-07, "loss": 0.3519, "step": 357 }, { "epoch": 0.09526343799893561, "grad_norm": 0.5493420958518982, "learning_rate": 1.898936170212766e-07, "loss": 0.3561, "step": 358 }, { "epoch": 0.09552953698775944, "grad_norm": 0.5558992624282837, "learning_rate": 1.904255319148936e-07, "loss": 0.3754, "step": 359 }, { "epoch": 0.09579563597658329, "grad_norm": 0.5792911648750305, "learning_rate": 1.9095744680851061e-07, "loss": 0.3391, "step": 360 }, { "epoch": 0.09606173496540713, "grad_norm": 0.5545791983604431, "learning_rate": 1.9148936170212767e-07, "loss": 0.3608, "step": 361 }, { "epoch": 0.09632783395423097, "grad_norm": 0.4920159876346588, "learning_rate": 1.9202127659574468e-07, "loss": 0.3386, "step": 362 }, { "epoch": 0.09659393294305482, "grad_norm": 0.5308285355567932, "learning_rate": 1.9255319148936169e-07, "loss": 0.3679, "step": 363 }, { "epoch": 0.09686003193187866, "grad_norm": 0.5439931154251099, "learning_rate": 1.9308510638297872e-07, "loss": 0.3502, "step": 364 }, { "epoch": 0.0971261309207025, "grad_norm": 0.5523415803909302, "learning_rate": 1.9361702127659573e-07, "loss": 0.3386, "step": 365 }, { "epoch": 0.09739222990952634, "grad_norm": 0.5924479961395264, "learning_rate": 1.9414893617021276e-07, "loss": 0.3582, "step": 366 }, { "epoch": 0.09765832889835019, "grad_norm": 0.5037036538124084, "learning_rate": 1.946808510638298e-07, "loss": 0.333, "step": 367 }, { "epoch": 0.09792442788717402, "grad_norm": 0.530329704284668, "learning_rate": 1.952127659574468e-07, "loss": 0.3466, "step": 368 }, { "epoch": 0.09819052687599787, "grad_norm": 0.5366110801696777, "learning_rate": 1.957446808510638e-07, "loss": 0.3633, "step": 369 }, { "epoch": 0.09845662586482172, "grad_norm": 0.4949975907802582, "learning_rate": 1.9627659574468084e-07, "loss": 0.3353, "step": 370 }, { "epoch": 0.09872272485364555, "grad_norm": 0.5265131592750549, "learning_rate": 1.9680851063829787e-07, "loss": 0.3332, "step": 371 }, { "epoch": 0.0989888238424694, "grad_norm": 0.48459696769714355, "learning_rate": 1.9734042553191488e-07, "loss": 0.3249, "step": 372 }, { "epoch": 0.09925492283129324, "grad_norm": 0.5504510998725891, "learning_rate": 1.978723404255319e-07, "loss": 0.3339, "step": 373 }, { "epoch": 0.09952102182011709, "grad_norm": 0.4802427887916565, "learning_rate": 1.9840425531914892e-07, "loss": 0.3505, "step": 374 }, { "epoch": 0.09978712080894092, "grad_norm": 0.5174222588539124, "learning_rate": 1.9893617021276595e-07, "loss": 0.3489, "step": 375 }, { "epoch": 0.10005321979776477, "grad_norm": 0.4977235198020935, "learning_rate": 1.9946808510638298e-07, "loss": 0.3489, "step": 376 }, { "epoch": 0.1003193187865886, "grad_norm": 0.49840447306632996, "learning_rate": 2e-07, "loss": 0.3267, "step": 377 }, { "epoch": 0.10058541777541245, "grad_norm": 0.4957316517829895, "learning_rate": 1.999999985446298e-07, "loss": 0.3337, "step": 378 }, { "epoch": 0.1008515167642363, "grad_norm": 0.501146137714386, "learning_rate": 1.999999941785193e-07, "loss": 0.3563, "step": 379 }, { "epoch": 0.10111761575306014, "grad_norm": 0.4876519739627838, "learning_rate": 1.9999998690166866e-07, "loss": 0.3466, "step": 380 }, { "epoch": 0.10138371474188398, "grad_norm": 0.48373943567276, "learning_rate": 1.9999997671407803e-07, "loss": 0.3329, "step": 381 }, { "epoch": 0.10164981373070782, "grad_norm": 0.5134963393211365, "learning_rate": 1.999999636157477e-07, "loss": 0.3628, "step": 382 }, { "epoch": 0.10191591271953167, "grad_norm": 0.45375028252601624, "learning_rate": 1.999999476066781e-07, "loss": 0.3203, "step": 383 }, { "epoch": 0.1021820117083555, "grad_norm": 0.4884037971496582, "learning_rate": 1.9999992868686965e-07, "loss": 0.3485, "step": 384 }, { "epoch": 0.10244811069717935, "grad_norm": 0.48264002799987793, "learning_rate": 1.9999990685632296e-07, "loss": 0.3473, "step": 385 }, { "epoch": 0.1027142096860032, "grad_norm": 0.520371675491333, "learning_rate": 1.999998821150386e-07, "loss": 0.3155, "step": 386 }, { "epoch": 0.10298030867482703, "grad_norm": 0.4640774130821228, "learning_rate": 1.9999985446301732e-07, "loss": 0.3215, "step": 387 }, { "epoch": 0.10324640766365088, "grad_norm": 0.4909495711326599, "learning_rate": 1.999998239002599e-07, "loss": 0.3353, "step": 388 }, { "epoch": 0.10351250665247472, "grad_norm": 0.443756103515625, "learning_rate": 1.999997904267673e-07, "loss": 0.3249, "step": 389 }, { "epoch": 0.10377860564129857, "grad_norm": 0.5128698348999023, "learning_rate": 1.9999975404254043e-07, "loss": 0.3511, "step": 390 }, { "epoch": 0.1040447046301224, "grad_norm": 0.4570084512233734, "learning_rate": 1.9999971474758036e-07, "loss": 0.3337, "step": 391 }, { "epoch": 0.10431080361894625, "grad_norm": 0.4496302008628845, "learning_rate": 1.9999967254188826e-07, "loss": 0.3388, "step": 392 }, { "epoch": 0.10457690260777008, "grad_norm": 0.43966370820999146, "learning_rate": 1.999996274254653e-07, "loss": 0.3289, "step": 393 }, { "epoch": 0.10484300159659393, "grad_norm": 0.4866158962249756, "learning_rate": 1.9999957939831287e-07, "loss": 0.3267, "step": 394 }, { "epoch": 0.10510910058541778, "grad_norm": 0.4865550696849823, "learning_rate": 1.999995284604323e-07, "loss": 0.3377, "step": 395 }, { "epoch": 0.10537519957424162, "grad_norm": 0.44094061851501465, "learning_rate": 1.9999947461182512e-07, "loss": 0.3216, "step": 396 }, { "epoch": 0.10564129856306546, "grad_norm": 0.4549223780632019, "learning_rate": 1.999994178524929e-07, "loss": 0.3335, "step": 397 }, { "epoch": 0.1059073975518893, "grad_norm": 0.4520883560180664, "learning_rate": 1.9999935818243726e-07, "loss": 0.3348, "step": 398 }, { "epoch": 0.10617349654071315, "grad_norm": 0.443739116191864, "learning_rate": 1.9999929560165992e-07, "loss": 0.3282, "step": 399 }, { "epoch": 0.10643959552953698, "grad_norm": 0.4682822525501251, "learning_rate": 1.9999923011016273e-07, "loss": 0.3489, "step": 400 }, { "epoch": 0.10670569451836083, "grad_norm": 0.44944238662719727, "learning_rate": 1.9999916170794764e-07, "loss": 0.3258, "step": 401 }, { "epoch": 0.10697179350718468, "grad_norm": 0.45946770906448364, "learning_rate": 1.9999909039501657e-07, "loss": 0.3689, "step": 402 }, { "epoch": 0.10723789249600851, "grad_norm": 0.44474661350250244, "learning_rate": 1.9999901617137162e-07, "loss": 0.3259, "step": 403 }, { "epoch": 0.10750399148483236, "grad_norm": 0.44001534581184387, "learning_rate": 1.9999893903701498e-07, "loss": 0.3303, "step": 404 }, { "epoch": 0.1077700904736562, "grad_norm": 0.47219938039779663, "learning_rate": 1.9999885899194883e-07, "loss": 0.3365, "step": 405 }, { "epoch": 0.10803618946248005, "grad_norm": 0.43242013454437256, "learning_rate": 1.9999877603617556e-07, "loss": 0.3294, "step": 406 }, { "epoch": 0.10830228845130388, "grad_norm": 0.4372362494468689, "learning_rate": 1.9999869016969758e-07, "loss": 0.325, "step": 407 }, { "epoch": 0.10856838744012773, "grad_norm": 0.44651421904563904, "learning_rate": 1.9999860139251737e-07, "loss": 0.3125, "step": 408 }, { "epoch": 0.10883448642895158, "grad_norm": 0.4624120891094208, "learning_rate": 1.999985097046375e-07, "loss": 0.341, "step": 409 }, { "epoch": 0.10910058541777541, "grad_norm": 0.4394981861114502, "learning_rate": 1.9999841510606066e-07, "loss": 0.3511, "step": 410 }, { "epoch": 0.10936668440659926, "grad_norm": 0.45270657539367676, "learning_rate": 1.999983175967896e-07, "loss": 0.3139, "step": 411 }, { "epoch": 0.1096327833954231, "grad_norm": 0.4686649441719055, "learning_rate": 1.9999821717682714e-07, "loss": 0.3345, "step": 412 }, { "epoch": 0.10989888238424694, "grad_norm": 0.40657109022140503, "learning_rate": 1.9999811384617625e-07, "loss": 0.2961, "step": 413 }, { "epoch": 0.11016498137307078, "grad_norm": 0.43064409494400024, "learning_rate": 1.999980076048399e-07, "loss": 0.3543, "step": 414 }, { "epoch": 0.11043108036189463, "grad_norm": 0.4409036338329315, "learning_rate": 1.9999789845282118e-07, "loss": 0.3189, "step": 415 }, { "epoch": 0.11069717935071846, "grad_norm": 0.45631009340286255, "learning_rate": 1.999977863901233e-07, "loss": 0.3313, "step": 416 }, { "epoch": 0.11096327833954231, "grad_norm": 0.4856877028942108, "learning_rate": 1.999976714167495e-07, "loss": 0.3251, "step": 417 }, { "epoch": 0.11122937732836616, "grad_norm": 0.4362773597240448, "learning_rate": 1.9999755353270308e-07, "loss": 0.3186, "step": 418 }, { "epoch": 0.11149547631718999, "grad_norm": 0.5027356743812561, "learning_rate": 1.9999743273798756e-07, "loss": 0.3472, "step": 419 }, { "epoch": 0.11176157530601384, "grad_norm": 0.40985921025276184, "learning_rate": 1.999973090326064e-07, "loss": 0.3093, "step": 420 }, { "epoch": 0.11202767429483768, "grad_norm": 0.4581221342086792, "learning_rate": 1.9999718241656322e-07, "loss": 0.3319, "step": 421 }, { "epoch": 0.11229377328366152, "grad_norm": 0.42118072509765625, "learning_rate": 1.999970528898617e-07, "loss": 0.3297, "step": 422 }, { "epoch": 0.11255987227248536, "grad_norm": 0.4535369575023651, "learning_rate": 1.999969204525056e-07, "loss": 0.3305, "step": 423 }, { "epoch": 0.11282597126130921, "grad_norm": 0.4076696038246155, "learning_rate": 1.999967851044988e-07, "loss": 0.321, "step": 424 }, { "epoch": 0.11309207025013306, "grad_norm": 0.4212122857570648, "learning_rate": 1.999966468458452e-07, "loss": 0.3318, "step": 425 }, { "epoch": 0.11335816923895689, "grad_norm": 0.42996591329574585, "learning_rate": 1.999965056765489e-07, "loss": 0.3156, "step": 426 }, { "epoch": 0.11362426822778074, "grad_norm": 0.42239636182785034, "learning_rate": 1.999963615966139e-07, "loss": 0.338, "step": 427 }, { "epoch": 0.11389036721660457, "grad_norm": 0.46019092202186584, "learning_rate": 1.999962146060445e-07, "loss": 0.3171, "step": 428 }, { "epoch": 0.11415646620542842, "grad_norm": 0.42835378646850586, "learning_rate": 1.9999606470484492e-07, "loss": 0.3429, "step": 429 }, { "epoch": 0.11442256519425226, "grad_norm": 0.42286840081214905, "learning_rate": 1.9999591189301949e-07, "loss": 0.3194, "step": 430 }, { "epoch": 0.1146886641830761, "grad_norm": 0.425481915473938, "learning_rate": 1.9999575617057274e-07, "loss": 0.3472, "step": 431 }, { "epoch": 0.11495476317189994, "grad_norm": 0.42244356870651245, "learning_rate": 1.9999559753750917e-07, "loss": 0.3147, "step": 432 }, { "epoch": 0.11522086216072379, "grad_norm": 0.3974209725856781, "learning_rate": 1.9999543599383337e-07, "loss": 0.3187, "step": 433 }, { "epoch": 0.11548696114954764, "grad_norm": 0.3930739164352417, "learning_rate": 1.9999527153955008e-07, "loss": 0.3267, "step": 434 }, { "epoch": 0.11575306013837147, "grad_norm": 0.4599141776561737, "learning_rate": 1.9999510417466406e-07, "loss": 0.3166, "step": 435 }, { "epoch": 0.11601915912719532, "grad_norm": 0.4314284920692444, "learning_rate": 1.999949338991802e-07, "loss": 0.3391, "step": 436 }, { "epoch": 0.11628525811601916, "grad_norm": 0.42881184816360474, "learning_rate": 1.9999476071310345e-07, "loss": 0.3548, "step": 437 }, { "epoch": 0.116551357104843, "grad_norm": 0.41905850172042847, "learning_rate": 1.999945846164388e-07, "loss": 0.3168, "step": 438 }, { "epoch": 0.11681745609366684, "grad_norm": 0.43408069014549255, "learning_rate": 1.999944056091915e-07, "loss": 0.3328, "step": 439 }, { "epoch": 0.11708355508249069, "grad_norm": 0.4169093370437622, "learning_rate": 1.9999422369136666e-07, "loss": 0.3208, "step": 440 }, { "epoch": 0.11734965407131454, "grad_norm": 0.3791236877441406, "learning_rate": 1.999940388629696e-07, "loss": 0.3202, "step": 441 }, { "epoch": 0.11761575306013837, "grad_norm": 0.39635536074638367, "learning_rate": 1.9999385112400568e-07, "loss": 0.3051, "step": 442 }, { "epoch": 0.11788185204896222, "grad_norm": 0.3957427442073822, "learning_rate": 1.9999366047448038e-07, "loss": 0.3094, "step": 443 }, { "epoch": 0.11814795103778605, "grad_norm": 0.4238659143447876, "learning_rate": 1.9999346691439928e-07, "loss": 0.3238, "step": 444 }, { "epoch": 0.1184140500266099, "grad_norm": 0.41186800599098206, "learning_rate": 1.9999327044376798e-07, "loss": 0.3271, "step": 445 }, { "epoch": 0.11868014901543374, "grad_norm": 0.40869012475013733, "learning_rate": 1.9999307106259222e-07, "loss": 0.3397, "step": 446 }, { "epoch": 0.11894624800425758, "grad_norm": 0.39044609665870667, "learning_rate": 1.999928687708778e-07, "loss": 0.3119, "step": 447 }, { "epoch": 0.11921234699308142, "grad_norm": 0.3867211639881134, "learning_rate": 1.9999266356863056e-07, "loss": 0.3038, "step": 448 }, { "epoch": 0.11947844598190527, "grad_norm": 0.42814669013023376, "learning_rate": 1.9999245545585655e-07, "loss": 0.329, "step": 449 }, { "epoch": 0.11974454497072912, "grad_norm": 0.39507606625556946, "learning_rate": 1.9999224443256177e-07, "loss": 0.3167, "step": 450 }, { "epoch": 0.12001064395955295, "grad_norm": 0.40263044834136963, "learning_rate": 1.9999203049875236e-07, "loss": 0.3232, "step": 451 }, { "epoch": 0.1202767429483768, "grad_norm": 0.4032459557056427, "learning_rate": 1.9999181365443461e-07, "loss": 0.2992, "step": 452 }, { "epoch": 0.12054284193720063, "grad_norm": 0.3965512216091156, "learning_rate": 1.9999159389961478e-07, "loss": 0.3215, "step": 453 }, { "epoch": 0.12080894092602448, "grad_norm": 0.38925445079803467, "learning_rate": 1.9999137123429927e-07, "loss": 0.3177, "step": 454 }, { "epoch": 0.12107503991484832, "grad_norm": 0.377194881439209, "learning_rate": 1.999911456584946e-07, "loss": 0.3033, "step": 455 }, { "epoch": 0.12134113890367217, "grad_norm": 0.39850443601608276, "learning_rate": 1.9999091717220722e-07, "loss": 0.3075, "step": 456 }, { "epoch": 0.12160723789249601, "grad_norm": 0.378618448972702, "learning_rate": 1.9999068577544392e-07, "loss": 0.3315, "step": 457 }, { "epoch": 0.12187333688131985, "grad_norm": 0.3974233865737915, "learning_rate": 1.9999045146821139e-07, "loss": 0.3331, "step": 458 }, { "epoch": 0.1221394358701437, "grad_norm": 0.3771480917930603, "learning_rate": 1.9999021425051644e-07, "loss": 0.3023, "step": 459 }, { "epoch": 0.12240553485896753, "grad_norm": 0.41467341780662537, "learning_rate": 1.9998997412236594e-07, "loss": 0.3289, "step": 460 }, { "epoch": 0.12267163384779138, "grad_norm": 0.39634227752685547, "learning_rate": 1.9998973108376696e-07, "loss": 0.3273, "step": 461 }, { "epoch": 0.12293773283661522, "grad_norm": 0.4868484139442444, "learning_rate": 1.9998948513472648e-07, "loss": 0.3557, "step": 462 }, { "epoch": 0.12320383182543906, "grad_norm": 0.37179556488990784, "learning_rate": 1.9998923627525176e-07, "loss": 0.2966, "step": 463 }, { "epoch": 0.12346993081426291, "grad_norm": 0.3750380277633667, "learning_rate": 1.9998898450534997e-07, "loss": 0.3063, "step": 464 }, { "epoch": 0.12373602980308675, "grad_norm": 0.3998767137527466, "learning_rate": 1.9998872982502846e-07, "loss": 0.3219, "step": 465 }, { "epoch": 0.1240021287919106, "grad_norm": 0.3908625841140747, "learning_rate": 1.9998847223429463e-07, "loss": 0.3152, "step": 466 }, { "epoch": 0.12426822778073443, "grad_norm": 0.3863421082496643, "learning_rate": 1.9998821173315599e-07, "loss": 0.3295, "step": 467 }, { "epoch": 0.12453432676955828, "grad_norm": 0.37672367691993713, "learning_rate": 1.9998794832162013e-07, "loss": 0.3125, "step": 468 }, { "epoch": 0.12480042575838211, "grad_norm": 0.38969525694847107, "learning_rate": 1.999876819996947e-07, "loss": 0.3102, "step": 469 }, { "epoch": 0.12506652474720595, "grad_norm": 0.3804856836795807, "learning_rate": 1.999874127673875e-07, "loss": 0.3016, "step": 470 }, { "epoch": 0.1253326237360298, "grad_norm": 0.4209743142127991, "learning_rate": 1.999871406247063e-07, "loss": 0.3286, "step": 471 }, { "epoch": 0.12559872272485365, "grad_norm": 0.3889736533164978, "learning_rate": 1.9998686557165906e-07, "loss": 0.3063, "step": 472 }, { "epoch": 0.1258648217136775, "grad_norm": 0.3712884783744812, "learning_rate": 1.9998658760825376e-07, "loss": 0.3235, "step": 473 }, { "epoch": 0.12613092070250134, "grad_norm": 0.38985756039619446, "learning_rate": 1.9998630673449853e-07, "loss": 0.3159, "step": 474 }, { "epoch": 0.12639701969132516, "grad_norm": 0.38237613439559937, "learning_rate": 1.9998602295040151e-07, "loss": 0.3134, "step": 475 }, { "epoch": 0.126663118680149, "grad_norm": 0.3951902687549591, "learning_rate": 1.9998573625597097e-07, "loss": 0.3085, "step": 476 }, { "epoch": 0.12692921766897286, "grad_norm": 0.3757186233997345, "learning_rate": 1.9998544665121527e-07, "loss": 0.326, "step": 477 }, { "epoch": 0.1271953166577967, "grad_norm": 0.38078296184539795, "learning_rate": 1.9998515413614281e-07, "loss": 0.3278, "step": 478 }, { "epoch": 0.12746141564662053, "grad_norm": 0.4181426465511322, "learning_rate": 1.9998485871076214e-07, "loss": 0.3216, "step": 479 }, { "epoch": 0.12772751463544438, "grad_norm": 0.3689579665660858, "learning_rate": 1.9998456037508183e-07, "loss": 0.2918, "step": 480 }, { "epoch": 0.12799361362426823, "grad_norm": 0.380497545003891, "learning_rate": 1.999842591291106e-07, "loss": 0.3079, "step": 481 }, { "epoch": 0.12825971261309208, "grad_norm": 0.4146079421043396, "learning_rate": 1.9998395497285717e-07, "loss": 0.3319, "step": 482 }, { "epoch": 0.12852581160191592, "grad_norm": 0.4030975103378296, "learning_rate": 1.999836479063304e-07, "loss": 0.3335, "step": 483 }, { "epoch": 0.12879191059073974, "grad_norm": 0.379884272813797, "learning_rate": 1.9998333792953925e-07, "loss": 0.307, "step": 484 }, { "epoch": 0.1290580095795636, "grad_norm": 0.38547250628471375, "learning_rate": 1.9998302504249277e-07, "loss": 0.3235, "step": 485 }, { "epoch": 0.12932410856838744, "grad_norm": 0.4136222004890442, "learning_rate": 1.999827092452e-07, "loss": 0.3363, "step": 486 }, { "epoch": 0.1295902075572113, "grad_norm": 0.4050493538379669, "learning_rate": 1.9998239053767019e-07, "loss": 0.3169, "step": 487 }, { "epoch": 0.12985630654603514, "grad_norm": 0.3730846345424652, "learning_rate": 1.999820689199126e-07, "loss": 0.3132, "step": 488 }, { "epoch": 0.13012240553485896, "grad_norm": 0.4033719003200531, "learning_rate": 1.9998174439193654e-07, "loss": 0.3227, "step": 489 }, { "epoch": 0.1303885045236828, "grad_norm": 0.3855295181274414, "learning_rate": 1.9998141695375152e-07, "loss": 0.3134, "step": 490 }, { "epoch": 0.13065460351250666, "grad_norm": 0.37922441959381104, "learning_rate": 1.9998108660536708e-07, "loss": 0.315, "step": 491 }, { "epoch": 0.1309207025013305, "grad_norm": 0.41410887241363525, "learning_rate": 1.9998075334679278e-07, "loss": 0.319, "step": 492 }, { "epoch": 0.13118680149015433, "grad_norm": 0.38255903124809265, "learning_rate": 1.9998041717803836e-07, "loss": 0.303, "step": 493 }, { "epoch": 0.13145290047897817, "grad_norm": 0.40724897384643555, "learning_rate": 1.9998007809911356e-07, "loss": 0.3101, "step": 494 }, { "epoch": 0.13171899946780202, "grad_norm": 0.3919048309326172, "learning_rate": 1.9997973611002835e-07, "loss": 0.3069, "step": 495 }, { "epoch": 0.13198509845662587, "grad_norm": 0.38622069358825684, "learning_rate": 1.9997939121079255e-07, "loss": 0.2958, "step": 496 }, { "epoch": 0.13225119744544972, "grad_norm": 0.3913390040397644, "learning_rate": 1.9997904340141632e-07, "loss": 0.3092, "step": 497 }, { "epoch": 0.13251729643427354, "grad_norm": 0.39583930373191833, "learning_rate": 1.9997869268190969e-07, "loss": 0.3301, "step": 498 }, { "epoch": 0.1327833954230974, "grad_norm": 0.38390302658081055, "learning_rate": 1.999783390522829e-07, "loss": 0.3235, "step": 499 }, { "epoch": 0.13304949441192124, "grad_norm": 0.38240301609039307, "learning_rate": 1.9997798251254626e-07, "loss": 0.3214, "step": 500 }, { "epoch": 0.13331559340074509, "grad_norm": 0.37457510828971863, "learning_rate": 1.9997762306271016e-07, "loss": 0.3081, "step": 501 }, { "epoch": 0.1335816923895689, "grad_norm": 0.3623429238796234, "learning_rate": 1.9997726070278504e-07, "loss": 0.3078, "step": 502 }, { "epoch": 0.13384779137839276, "grad_norm": 0.38916876912117004, "learning_rate": 1.9997689543278143e-07, "loss": 0.3324, "step": 503 }, { "epoch": 0.1341138903672166, "grad_norm": 0.35488665103912354, "learning_rate": 1.9997652725271003e-07, "loss": 0.3088, "step": 504 }, { "epoch": 0.13437998935604045, "grad_norm": 0.3757742643356323, "learning_rate": 1.9997615616258146e-07, "loss": 0.3238, "step": 505 }, { "epoch": 0.1346460883448643, "grad_norm": 0.3769417703151703, "learning_rate": 1.9997578216240657e-07, "loss": 0.3192, "step": 506 }, { "epoch": 0.13491218733368812, "grad_norm": 0.3790467083454132, "learning_rate": 1.9997540525219625e-07, "loss": 0.3369, "step": 507 }, { "epoch": 0.13517828632251197, "grad_norm": 0.3851090669631958, "learning_rate": 1.9997502543196147e-07, "loss": 0.3211, "step": 508 }, { "epoch": 0.13544438531133582, "grad_norm": 0.37709033489227295, "learning_rate": 1.999746427017133e-07, "loss": 0.3206, "step": 509 }, { "epoch": 0.13571048430015967, "grad_norm": 0.35060253739356995, "learning_rate": 1.9997425706146282e-07, "loss": 0.2989, "step": 510 }, { "epoch": 0.1359765832889835, "grad_norm": 0.37992340326309204, "learning_rate": 1.9997386851122132e-07, "loss": 0.3056, "step": 511 }, { "epoch": 0.13624268227780734, "grad_norm": 0.3894604444503784, "learning_rate": 1.9997347705100012e-07, "loss": 0.326, "step": 512 }, { "epoch": 0.13650878126663119, "grad_norm": 0.3650719225406647, "learning_rate": 1.9997308268081053e-07, "loss": 0.3085, "step": 513 }, { "epoch": 0.13677488025545503, "grad_norm": 0.37843674421310425, "learning_rate": 1.9997268540066412e-07, "loss": 0.3195, "step": 514 }, { "epoch": 0.13704097924427888, "grad_norm": 0.36798229813575745, "learning_rate": 1.9997228521057238e-07, "loss": 0.3135, "step": 515 }, { "epoch": 0.1373070782331027, "grad_norm": 0.3623092770576477, "learning_rate": 1.9997188211054702e-07, "loss": 0.3023, "step": 516 }, { "epoch": 0.13757317722192655, "grad_norm": 0.37259581685066223, "learning_rate": 1.9997147610059976e-07, "loss": 0.3113, "step": 517 }, { "epoch": 0.1378392762107504, "grad_norm": 0.4057176411151886, "learning_rate": 1.9997106718074237e-07, "loss": 0.3297, "step": 518 }, { "epoch": 0.13810537519957425, "grad_norm": 0.3829900324344635, "learning_rate": 1.999706553509868e-07, "loss": 0.2816, "step": 519 }, { "epoch": 0.1383714741883981, "grad_norm": 0.3459978699684143, "learning_rate": 1.9997024061134503e-07, "loss": 0.3095, "step": 520 }, { "epoch": 0.13863757317722192, "grad_norm": 0.42888298630714417, "learning_rate": 1.9996982296182913e-07, "loss": 0.3098, "step": 521 }, { "epoch": 0.13890367216604577, "grad_norm": 0.3485850989818573, "learning_rate": 1.9996940240245123e-07, "loss": 0.3053, "step": 522 }, { "epoch": 0.13916977115486961, "grad_norm": 0.3592708110809326, "learning_rate": 1.999689789332236e-07, "loss": 0.313, "step": 523 }, { "epoch": 0.13943587014369346, "grad_norm": 0.3812033534049988, "learning_rate": 1.9996855255415857e-07, "loss": 0.3178, "step": 524 }, { "epoch": 0.13970196913251728, "grad_norm": 0.3658882677555084, "learning_rate": 1.9996812326526854e-07, "loss": 0.2998, "step": 525 }, { "epoch": 0.13996806812134113, "grad_norm": 0.41772517561912537, "learning_rate": 1.99967691066566e-07, "loss": 0.3276, "step": 526 }, { "epoch": 0.14023416711016498, "grad_norm": 0.3877013325691223, "learning_rate": 1.9996725595806354e-07, "loss": 0.3258, "step": 527 }, { "epoch": 0.14050026609898883, "grad_norm": 0.38283899426460266, "learning_rate": 1.9996681793977382e-07, "loss": 0.3165, "step": 528 }, { "epoch": 0.14076636508781268, "grad_norm": 0.35893622040748596, "learning_rate": 1.9996637701170957e-07, "loss": 0.3087, "step": 529 }, { "epoch": 0.1410324640766365, "grad_norm": 0.3515489995479584, "learning_rate": 1.999659331738837e-07, "loss": 0.3063, "step": 530 }, { "epoch": 0.14129856306546035, "grad_norm": 0.3778691589832306, "learning_rate": 1.9996548642630903e-07, "loss": 0.3217, "step": 531 }, { "epoch": 0.1415646620542842, "grad_norm": 0.3432246744632721, "learning_rate": 1.999650367689986e-07, "loss": 0.2896, "step": 532 }, { "epoch": 0.14183076104310804, "grad_norm": 0.36932554841041565, "learning_rate": 1.9996458420196553e-07, "loss": 0.3011, "step": 533 }, { "epoch": 0.14209686003193187, "grad_norm": 0.37214216589927673, "learning_rate": 1.9996412872522296e-07, "loss": 0.3168, "step": 534 }, { "epoch": 0.1423629590207557, "grad_norm": 0.33537518978118896, "learning_rate": 1.999636703387842e-07, "loss": 0.3068, "step": 535 }, { "epoch": 0.14262905800957956, "grad_norm": 0.3654571771621704, "learning_rate": 1.9996320904266247e-07, "loss": 0.3212, "step": 536 }, { "epoch": 0.1428951569984034, "grad_norm": 0.3715420067310333, "learning_rate": 1.9996274483687136e-07, "loss": 0.3052, "step": 537 }, { "epoch": 0.14316125598722726, "grad_norm": 0.3519796133041382, "learning_rate": 1.9996227772142423e-07, "loss": 0.3087, "step": 538 }, { "epoch": 0.14342735497605108, "grad_norm": 0.36168769001960754, "learning_rate": 1.9996180769633476e-07, "loss": 0.3108, "step": 539 }, { "epoch": 0.14369345396487493, "grad_norm": 0.37384557723999023, "learning_rate": 1.9996133476161666e-07, "loss": 0.3225, "step": 540 }, { "epoch": 0.14395955295369878, "grad_norm": 0.34027722477912903, "learning_rate": 1.999608589172836e-07, "loss": 0.3014, "step": 541 }, { "epoch": 0.14422565194252263, "grad_norm": 0.3785979151725769, "learning_rate": 1.999603801633495e-07, "loss": 0.316, "step": 542 }, { "epoch": 0.14449175093134647, "grad_norm": 0.34600138664245605, "learning_rate": 1.9995989849982826e-07, "loss": 0.3076, "step": 543 }, { "epoch": 0.1447578499201703, "grad_norm": 0.3479349613189697, "learning_rate": 1.9995941392673394e-07, "loss": 0.2917, "step": 544 }, { "epoch": 0.14502394890899414, "grad_norm": 0.3447350263595581, "learning_rate": 1.9995892644408063e-07, "loss": 0.302, "step": 545 }, { "epoch": 0.145290047897818, "grad_norm": 0.36319226026535034, "learning_rate": 1.999584360518825e-07, "loss": 0.324, "step": 546 }, { "epoch": 0.14555614688664184, "grad_norm": 0.39075613021850586, "learning_rate": 1.9995794275015384e-07, "loss": 0.3101, "step": 547 }, { "epoch": 0.14582224587546566, "grad_norm": 0.3504732847213745, "learning_rate": 1.9995744653890901e-07, "loss": 0.3015, "step": 548 }, { "epoch": 0.1460883448642895, "grad_norm": 0.35884565114974976, "learning_rate": 1.9995694741816242e-07, "loss": 0.2985, "step": 549 }, { "epoch": 0.14635444385311336, "grad_norm": 0.3430880904197693, "learning_rate": 1.9995644538792867e-07, "loss": 0.3069, "step": 550 }, { "epoch": 0.1466205428419372, "grad_norm": 0.35654449462890625, "learning_rate": 1.999559404482223e-07, "loss": 0.308, "step": 551 }, { "epoch": 0.14688664183076106, "grad_norm": 0.3501586318016052, "learning_rate": 1.9995543259905804e-07, "loss": 0.2994, "step": 552 }, { "epoch": 0.14715274081958488, "grad_norm": 0.3849630355834961, "learning_rate": 1.9995492184045068e-07, "loss": 0.3136, "step": 553 }, { "epoch": 0.14741883980840872, "grad_norm": 0.36192670464515686, "learning_rate": 1.9995440817241505e-07, "loss": 0.3081, "step": 554 }, { "epoch": 0.14768493879723257, "grad_norm": 0.33561861515045166, "learning_rate": 1.9995389159496615e-07, "loss": 0.2967, "step": 555 }, { "epoch": 0.14795103778605642, "grad_norm": 0.33408603072166443, "learning_rate": 1.99953372108119e-07, "loss": 0.3057, "step": 556 }, { "epoch": 0.14821713677488024, "grad_norm": 0.35246655344963074, "learning_rate": 1.999528497118887e-07, "loss": 0.2935, "step": 557 }, { "epoch": 0.1484832357637041, "grad_norm": 0.3531337082386017, "learning_rate": 1.999523244062905e-07, "loss": 0.3093, "step": 558 }, { "epoch": 0.14874933475252794, "grad_norm": 0.4592053294181824, "learning_rate": 1.999517961913396e-07, "loss": 0.32, "step": 559 }, { "epoch": 0.1490154337413518, "grad_norm": 0.3572847545146942, "learning_rate": 1.999512650670515e-07, "loss": 0.2999, "step": 560 }, { "epoch": 0.14928153273017564, "grad_norm": 0.3594224154949188, "learning_rate": 1.9995073103344154e-07, "loss": 0.3201, "step": 561 }, { "epoch": 0.14954763171899946, "grad_norm": 0.46932658553123474, "learning_rate": 1.9995019409052533e-07, "loss": 0.2987, "step": 562 }, { "epoch": 0.1498137307078233, "grad_norm": 0.3753923177719116, "learning_rate": 1.9994965423831852e-07, "loss": 0.302, "step": 563 }, { "epoch": 0.15007982969664715, "grad_norm": 0.36340415477752686, "learning_rate": 1.9994911147683674e-07, "loss": 0.2917, "step": 564 }, { "epoch": 0.150345928685471, "grad_norm": 0.3503718376159668, "learning_rate": 1.999485658060959e-07, "loss": 0.2957, "step": 565 }, { "epoch": 0.15061202767429482, "grad_norm": 0.37649738788604736, "learning_rate": 1.9994801722611179e-07, "loss": 0.3038, "step": 566 }, { "epoch": 0.15087812666311867, "grad_norm": 0.38778433203697205, "learning_rate": 1.999474657369004e-07, "loss": 0.3102, "step": 567 }, { "epoch": 0.15114422565194252, "grad_norm": 0.3864658772945404, "learning_rate": 1.9994691133847784e-07, "loss": 0.3128, "step": 568 }, { "epoch": 0.15141032464076637, "grad_norm": 0.3219042718410492, "learning_rate": 1.9994635403086013e-07, "loss": 0.2909, "step": 569 }, { "epoch": 0.15167642362959022, "grad_norm": 0.32933029532432556, "learning_rate": 1.9994579381406362e-07, "loss": 0.2984, "step": 570 }, { "epoch": 0.15194252261841404, "grad_norm": 0.32277387380599976, "learning_rate": 1.9994523068810455e-07, "loss": 0.302, "step": 571 }, { "epoch": 0.1522086216072379, "grad_norm": 0.39731577038764954, "learning_rate": 1.9994466465299928e-07, "loss": 0.3099, "step": 572 }, { "epoch": 0.15247472059606174, "grad_norm": 0.36871784925460815, "learning_rate": 1.9994409570876436e-07, "loss": 0.2974, "step": 573 }, { "epoch": 0.15274081958488558, "grad_norm": 0.32451486587524414, "learning_rate": 1.999435238554163e-07, "loss": 0.2952, "step": 574 }, { "epoch": 0.15300691857370943, "grad_norm": 0.3553796112537384, "learning_rate": 1.9994294909297177e-07, "loss": 0.2938, "step": 575 }, { "epoch": 0.15327301756253325, "grad_norm": 0.3637763559818268, "learning_rate": 1.9994237142144752e-07, "loss": 0.3091, "step": 576 }, { "epoch": 0.1535391165513571, "grad_norm": 0.3449094593524933, "learning_rate": 1.999417908408603e-07, "loss": 0.3014, "step": 577 }, { "epoch": 0.15380521554018095, "grad_norm": 0.33983510732650757, "learning_rate": 1.9994120735122707e-07, "loss": 0.3093, "step": 578 }, { "epoch": 0.1540713145290048, "grad_norm": 0.3420371413230896, "learning_rate": 1.9994062095256476e-07, "loss": 0.2919, "step": 579 }, { "epoch": 0.15433741351782862, "grad_norm": 0.34776628017425537, "learning_rate": 1.9994003164489048e-07, "loss": 0.3026, "step": 580 }, { "epoch": 0.15460351250665247, "grad_norm": 0.36293089389801025, "learning_rate": 1.9993943942822138e-07, "loss": 0.315, "step": 581 }, { "epoch": 0.15486961149547632, "grad_norm": 0.3600641191005707, "learning_rate": 1.999388443025747e-07, "loss": 0.3151, "step": 582 }, { "epoch": 0.15513571048430017, "grad_norm": 0.34246012568473816, "learning_rate": 1.999382462679677e-07, "loss": 0.3113, "step": 583 }, { "epoch": 0.15540180947312401, "grad_norm": 0.38469189405441284, "learning_rate": 1.999376453244179e-07, "loss": 0.3125, "step": 584 }, { "epoch": 0.15566790846194783, "grad_norm": 0.33684617280960083, "learning_rate": 1.9993704147194269e-07, "loss": 0.2833, "step": 585 }, { "epoch": 0.15593400745077168, "grad_norm": 0.3476209044456482, "learning_rate": 1.999364347105597e-07, "loss": 0.3192, "step": 586 }, { "epoch": 0.15620010643959553, "grad_norm": 0.37684714794158936, "learning_rate": 1.9993582504028655e-07, "loss": 0.319, "step": 587 }, { "epoch": 0.15646620542841938, "grad_norm": 0.32393765449523926, "learning_rate": 1.9993521246114104e-07, "loss": 0.2982, "step": 588 }, { "epoch": 0.1567323044172432, "grad_norm": 0.339459627866745, "learning_rate": 1.9993459697314097e-07, "loss": 0.298, "step": 589 }, { "epoch": 0.15699840340606705, "grad_norm": 0.3627125024795532, "learning_rate": 1.9993397857630425e-07, "loss": 0.3107, "step": 590 }, { "epoch": 0.1572645023948909, "grad_norm": 0.3398129940032959, "learning_rate": 1.9993335727064894e-07, "loss": 0.3038, "step": 591 }, { "epoch": 0.15753060138371475, "grad_norm": 0.34548085927963257, "learning_rate": 1.9993273305619302e-07, "loss": 0.279, "step": 592 }, { "epoch": 0.1577967003725386, "grad_norm": 0.3668566644191742, "learning_rate": 1.999321059329547e-07, "loss": 0.2988, "step": 593 }, { "epoch": 0.15806279936136242, "grad_norm": 0.3628426790237427, "learning_rate": 1.999314759009523e-07, "loss": 0.3152, "step": 594 }, { "epoch": 0.15832889835018626, "grad_norm": 0.408966600894928, "learning_rate": 1.999308429602041e-07, "loss": 0.2957, "step": 595 }, { "epoch": 0.1585949973390101, "grad_norm": 0.39604780077934265, "learning_rate": 1.999302071107285e-07, "loss": 0.293, "step": 596 }, { "epoch": 0.15886109632783396, "grad_norm": 0.3589252531528473, "learning_rate": 1.9992956835254405e-07, "loss": 0.3454, "step": 597 }, { "epoch": 0.1591271953166578, "grad_norm": 0.361908495426178, "learning_rate": 1.9992892668566932e-07, "loss": 0.3123, "step": 598 }, { "epoch": 0.15939329430548163, "grad_norm": 0.4665786325931549, "learning_rate": 1.99928282110123e-07, "loss": 0.3142, "step": 599 }, { "epoch": 0.15965939329430548, "grad_norm": 0.35615187883377075, "learning_rate": 1.9992763462592385e-07, "loss": 0.2909, "step": 600 }, { "epoch": 0.15992549228312933, "grad_norm": 0.3306821882724762, "learning_rate": 1.9992698423309074e-07, "loss": 0.2945, "step": 601 }, { "epoch": 0.16019159127195318, "grad_norm": 0.43744269013404846, "learning_rate": 1.9992633093164256e-07, "loss": 0.3075, "step": 602 }, { "epoch": 0.160457690260777, "grad_norm": 0.33926302194595337, "learning_rate": 1.9992567472159832e-07, "loss": 0.3047, "step": 603 }, { "epoch": 0.16072378924960085, "grad_norm": 0.3741895258426666, "learning_rate": 1.9992501560297717e-07, "loss": 0.3079, "step": 604 }, { "epoch": 0.1609898882384247, "grad_norm": 0.3886670470237732, "learning_rate": 1.9992435357579827e-07, "loss": 0.3127, "step": 605 }, { "epoch": 0.16125598722724854, "grad_norm": 0.32639384269714355, "learning_rate": 1.9992368864008087e-07, "loss": 0.288, "step": 606 }, { "epoch": 0.1615220862160724, "grad_norm": 0.37474602460861206, "learning_rate": 1.9992302079584437e-07, "loss": 0.2966, "step": 607 }, { "epoch": 0.1617881852048962, "grad_norm": 0.3544592559337616, "learning_rate": 1.9992235004310816e-07, "loss": 0.302, "step": 608 }, { "epoch": 0.16205428419372006, "grad_norm": 0.30966097116470337, "learning_rate": 1.999216763818918e-07, "loss": 0.2833, "step": 609 }, { "epoch": 0.1623203831825439, "grad_norm": 0.3804732859134674, "learning_rate": 1.9992099981221484e-07, "loss": 0.3016, "step": 610 }, { "epoch": 0.16258648217136776, "grad_norm": 0.3533763289451599, "learning_rate": 1.9992032033409706e-07, "loss": 0.295, "step": 611 }, { "epoch": 0.16285258116019158, "grad_norm": 0.3821411728858948, "learning_rate": 1.999196379475582e-07, "loss": 0.3264, "step": 612 }, { "epoch": 0.16311868014901543, "grad_norm": 0.38038915395736694, "learning_rate": 1.9991895265261812e-07, "loss": 0.2892, "step": 613 }, { "epoch": 0.16338477913783928, "grad_norm": 0.35747188329696655, "learning_rate": 1.9991826444929672e-07, "loss": 0.3086, "step": 614 }, { "epoch": 0.16365087812666312, "grad_norm": 0.3333292007446289, "learning_rate": 1.9991757333761411e-07, "loss": 0.3055, "step": 615 }, { "epoch": 0.16391697711548697, "grad_norm": 0.3871196508407593, "learning_rate": 1.9991687931759039e-07, "loss": 0.2882, "step": 616 }, { "epoch": 0.1641830761043108, "grad_norm": 0.35090675950050354, "learning_rate": 1.9991618238924573e-07, "loss": 0.2764, "step": 617 }, { "epoch": 0.16444917509313464, "grad_norm": 0.32575100660324097, "learning_rate": 1.9991548255260043e-07, "loss": 0.2867, "step": 618 }, { "epoch": 0.1647152740819585, "grad_norm": 0.35190728306770325, "learning_rate": 1.9991477980767484e-07, "loss": 0.3131, "step": 619 }, { "epoch": 0.16498137307078234, "grad_norm": 0.3180832862854004, "learning_rate": 1.9991407415448944e-07, "loss": 0.2933, "step": 620 }, { "epoch": 0.16524747205960616, "grad_norm": 0.33891767263412476, "learning_rate": 1.999133655930648e-07, "loss": 0.2915, "step": 621 }, { "epoch": 0.16551357104843, "grad_norm": 0.33159202337265015, "learning_rate": 1.9991265412342148e-07, "loss": 0.2917, "step": 622 }, { "epoch": 0.16577967003725386, "grad_norm": 0.3616255819797516, "learning_rate": 1.9991193974558023e-07, "loss": 0.2984, "step": 623 }, { "epoch": 0.1660457690260777, "grad_norm": 0.34811413288116455, "learning_rate": 1.9991122245956184e-07, "loss": 0.3143, "step": 624 }, { "epoch": 0.16631186801490155, "grad_norm": 0.3278140127658844, "learning_rate": 1.9991050226538716e-07, "loss": 0.2889, "step": 625 }, { "epoch": 0.16657796700372537, "grad_norm": 0.3525492548942566, "learning_rate": 1.9990977916307718e-07, "loss": 0.3015, "step": 626 }, { "epoch": 0.16684406599254922, "grad_norm": 0.32658451795578003, "learning_rate": 1.9990905315265297e-07, "loss": 0.2989, "step": 627 }, { "epoch": 0.16711016498137307, "grad_norm": 0.3136416971683502, "learning_rate": 1.999083242341356e-07, "loss": 0.2711, "step": 628 }, { "epoch": 0.16737626397019692, "grad_norm": 0.39282187819480896, "learning_rate": 1.9990759240754635e-07, "loss": 0.2857, "step": 629 }, { "epoch": 0.16764236295902077, "grad_norm": 0.33335936069488525, "learning_rate": 1.9990685767290648e-07, "loss": 0.3023, "step": 630 }, { "epoch": 0.1679084619478446, "grad_norm": 0.32040414214134216, "learning_rate": 1.9990612003023736e-07, "loss": 0.281, "step": 631 }, { "epoch": 0.16817456093666844, "grad_norm": 0.3410910964012146, "learning_rate": 1.9990537947956053e-07, "loss": 0.2881, "step": 632 }, { "epoch": 0.1684406599254923, "grad_norm": 0.3165854811668396, "learning_rate": 1.9990463602089749e-07, "loss": 0.2919, "step": 633 }, { "epoch": 0.16870675891431614, "grad_norm": 0.36237993836402893, "learning_rate": 1.9990388965426988e-07, "loss": 0.3318, "step": 634 }, { "epoch": 0.16897285790313996, "grad_norm": 0.3438180983066559, "learning_rate": 1.9990314037969945e-07, "loss": 0.3127, "step": 635 }, { "epoch": 0.1692389568919638, "grad_norm": 0.32420244812965393, "learning_rate": 1.9990238819720803e-07, "loss": 0.3032, "step": 636 }, { "epoch": 0.16950505588078765, "grad_norm": 0.30676713585853577, "learning_rate": 1.9990163310681745e-07, "loss": 0.2775, "step": 637 }, { "epoch": 0.1697711548696115, "grad_norm": 0.33915677666664124, "learning_rate": 1.9990087510854972e-07, "loss": 0.2825, "step": 638 }, { "epoch": 0.17003725385843535, "grad_norm": 0.3287609815597534, "learning_rate": 1.9990011420242692e-07, "loss": 0.2714, "step": 639 }, { "epoch": 0.17030335284725917, "grad_norm": 0.352018803358078, "learning_rate": 1.9989935038847116e-07, "loss": 0.284, "step": 640 }, { "epoch": 0.17056945183608302, "grad_norm": 0.3250834047794342, "learning_rate": 1.9989858366670475e-07, "loss": 0.2876, "step": 641 }, { "epoch": 0.17083555082490687, "grad_norm": 0.35086721181869507, "learning_rate": 1.9989781403714988e-07, "loss": 0.292, "step": 642 }, { "epoch": 0.17110164981373072, "grad_norm": 0.3593180179595947, "learning_rate": 1.9989704149982906e-07, "loss": 0.2764, "step": 643 }, { "epoch": 0.17136774880255454, "grad_norm": 0.3294963240623474, "learning_rate": 1.9989626605476477e-07, "loss": 0.2908, "step": 644 }, { "epoch": 0.17163384779137839, "grad_norm": 0.38134655356407166, "learning_rate": 1.9989548770197953e-07, "loss": 0.3142, "step": 645 }, { "epoch": 0.17189994678020223, "grad_norm": 0.36291778087615967, "learning_rate": 1.9989470644149602e-07, "loss": 0.2967, "step": 646 }, { "epoch": 0.17216604576902608, "grad_norm": 0.37372303009033203, "learning_rate": 1.9989392227333696e-07, "loss": 0.2963, "step": 647 }, { "epoch": 0.17243214475784993, "grad_norm": 0.3391888737678528, "learning_rate": 1.9989313519752521e-07, "loss": 0.2959, "step": 648 }, { "epoch": 0.17269824374667375, "grad_norm": 0.3316959738731384, "learning_rate": 1.9989234521408367e-07, "loss": 0.3077, "step": 649 }, { "epoch": 0.1729643427354976, "grad_norm": 0.33514153957366943, "learning_rate": 1.9989155232303533e-07, "loss": 0.3134, "step": 650 }, { "epoch": 0.17323044172432145, "grad_norm": 0.33506378531455994, "learning_rate": 1.9989075652440327e-07, "loss": 0.294, "step": 651 }, { "epoch": 0.1734965407131453, "grad_norm": 0.3328542113304138, "learning_rate": 1.9988995781821064e-07, "loss": 0.3016, "step": 652 }, { "epoch": 0.17376263970196915, "grad_norm": 0.33319059014320374, "learning_rate": 1.998891562044807e-07, "loss": 0.2919, "step": 653 }, { "epoch": 0.17402873869079297, "grad_norm": 0.3759792149066925, "learning_rate": 1.9988835168323679e-07, "loss": 0.2976, "step": 654 }, { "epoch": 0.17429483767961682, "grad_norm": 0.3295424282550812, "learning_rate": 1.998875442545023e-07, "loss": 0.3016, "step": 655 }, { "epoch": 0.17456093666844066, "grad_norm": 0.33964693546295166, "learning_rate": 1.998867339183008e-07, "loss": 0.3115, "step": 656 }, { "epoch": 0.1748270356572645, "grad_norm": 0.3272683024406433, "learning_rate": 1.9988592067465578e-07, "loss": 0.3003, "step": 657 }, { "epoch": 0.17509313464608833, "grad_norm": 0.3499435484409332, "learning_rate": 1.9988510452359098e-07, "loss": 0.3065, "step": 658 }, { "epoch": 0.17535923363491218, "grad_norm": 0.3277088403701782, "learning_rate": 1.9988428546513014e-07, "loss": 0.2847, "step": 659 }, { "epoch": 0.17562533262373603, "grad_norm": 0.3465122878551483, "learning_rate": 1.998834634992971e-07, "loss": 0.3073, "step": 660 }, { "epoch": 0.17589143161255988, "grad_norm": 0.3225915729999542, "learning_rate": 1.9988263862611577e-07, "loss": 0.2882, "step": 661 }, { "epoch": 0.17615753060138373, "grad_norm": 0.3085285723209381, "learning_rate": 1.9988181084561019e-07, "loss": 0.2956, "step": 662 }, { "epoch": 0.17642362959020755, "grad_norm": 0.36045512557029724, "learning_rate": 1.998809801578044e-07, "loss": 0.2886, "step": 663 }, { "epoch": 0.1766897285790314, "grad_norm": 0.2956853210926056, "learning_rate": 1.9988014656272262e-07, "loss": 0.2871, "step": 664 }, { "epoch": 0.17695582756785525, "grad_norm": 0.3068711757659912, "learning_rate": 1.9987931006038913e-07, "loss": 0.3047, "step": 665 }, { "epoch": 0.1772219265566791, "grad_norm": 0.31326228380203247, "learning_rate": 1.9987847065082827e-07, "loss": 0.2939, "step": 666 }, { "epoch": 0.17748802554550291, "grad_norm": 0.3164639472961426, "learning_rate": 1.9987762833406446e-07, "loss": 0.2847, "step": 667 }, { "epoch": 0.17775412453432676, "grad_norm": 0.324419766664505, "learning_rate": 1.998767831101222e-07, "loss": 0.2903, "step": 668 }, { "epoch": 0.1780202235231506, "grad_norm": 0.3260335922241211, "learning_rate": 1.9987593497902613e-07, "loss": 0.316, "step": 669 }, { "epoch": 0.17828632251197446, "grad_norm": 0.31951698660850525, "learning_rate": 1.9987508394080092e-07, "loss": 0.2924, "step": 670 }, { "epoch": 0.1785524215007983, "grad_norm": 0.337880939245224, "learning_rate": 1.998742299954713e-07, "loss": 0.302, "step": 671 }, { "epoch": 0.17881852048962213, "grad_norm": 0.3003269135951996, "learning_rate": 1.9987337314306218e-07, "loss": 0.2896, "step": 672 }, { "epoch": 0.17908461947844598, "grad_norm": 0.3476768434047699, "learning_rate": 1.9987251338359852e-07, "loss": 0.3217, "step": 673 }, { "epoch": 0.17935071846726983, "grad_norm": 0.31209632754325867, "learning_rate": 1.9987165071710525e-07, "loss": 0.2873, "step": 674 }, { "epoch": 0.17961681745609367, "grad_norm": 0.34188181161880493, "learning_rate": 1.998707851436076e-07, "loss": 0.3228, "step": 675 }, { "epoch": 0.1798829164449175, "grad_norm": 0.35385265946388245, "learning_rate": 1.998699166631307e-07, "loss": 0.2983, "step": 676 }, { "epoch": 0.18014901543374134, "grad_norm": 0.3128959834575653, "learning_rate": 1.998690452756998e-07, "loss": 0.2763, "step": 677 }, { "epoch": 0.1804151144225652, "grad_norm": 0.34531670808792114, "learning_rate": 1.9986817098134035e-07, "loss": 0.2952, "step": 678 }, { "epoch": 0.18068121341138904, "grad_norm": 0.35921189188957214, "learning_rate": 1.998672937800777e-07, "loss": 0.3041, "step": 679 }, { "epoch": 0.1809473124002129, "grad_norm": 0.3100993037223816, "learning_rate": 1.9986641367193746e-07, "loss": 0.2642, "step": 680 }, { "epoch": 0.1812134113890367, "grad_norm": 0.34045493602752686, "learning_rate": 1.9986553065694521e-07, "loss": 0.2948, "step": 681 }, { "epoch": 0.18147951037786056, "grad_norm": 0.34483009576797485, "learning_rate": 1.9986464473512666e-07, "loss": 0.3099, "step": 682 }, { "epoch": 0.1817456093666844, "grad_norm": 0.3131784200668335, "learning_rate": 1.9986375590650758e-07, "loss": 0.282, "step": 683 }, { "epoch": 0.18201170835550826, "grad_norm": 0.3155391216278076, "learning_rate": 1.998628641711139e-07, "loss": 0.2962, "step": 684 }, { "epoch": 0.1822778073443321, "grad_norm": 0.35545679926872253, "learning_rate": 1.998619695289715e-07, "loss": 0.2936, "step": 685 }, { "epoch": 0.18254390633315593, "grad_norm": 0.32207319140434265, "learning_rate": 1.9986107198010647e-07, "loss": 0.2933, "step": 686 }, { "epoch": 0.18281000532197977, "grad_norm": 0.34311428666114807, "learning_rate": 1.9986017152454493e-07, "loss": 0.304, "step": 687 }, { "epoch": 0.18307610431080362, "grad_norm": 0.3287527859210968, "learning_rate": 1.9985926816231307e-07, "loss": 0.2829, "step": 688 }, { "epoch": 0.18334220329962747, "grad_norm": 0.30524006485939026, "learning_rate": 1.998583618934372e-07, "loss": 0.2878, "step": 689 }, { "epoch": 0.1836083022884513, "grad_norm": 0.3178045451641083, "learning_rate": 1.9985745271794368e-07, "loss": 0.2994, "step": 690 }, { "epoch": 0.18387440127727514, "grad_norm": 0.34270137548446655, "learning_rate": 1.99856540635859e-07, "loss": 0.3104, "step": 691 }, { "epoch": 0.184140500266099, "grad_norm": 0.34845227003097534, "learning_rate": 1.998556256472097e-07, "loss": 0.2886, "step": 692 }, { "epoch": 0.18440659925492284, "grad_norm": 0.3514746427536011, "learning_rate": 1.998547077520224e-07, "loss": 0.3101, "step": 693 }, { "epoch": 0.18467269824374669, "grad_norm": 0.3283919394016266, "learning_rate": 1.9985378695032385e-07, "loss": 0.2861, "step": 694 }, { "epoch": 0.1849387972325705, "grad_norm": 0.3158937394618988, "learning_rate": 1.998528632421408e-07, "loss": 0.2804, "step": 695 }, { "epoch": 0.18520489622139436, "grad_norm": 0.3528836667537689, "learning_rate": 1.9985193662750018e-07, "loss": 0.2973, "step": 696 }, { "epoch": 0.1854709952102182, "grad_norm": 0.3275820016860962, "learning_rate": 1.9985100710642895e-07, "loss": 0.2948, "step": 697 }, { "epoch": 0.18573709419904205, "grad_norm": 0.326596736907959, "learning_rate": 1.9985007467895417e-07, "loss": 0.2838, "step": 698 }, { "epoch": 0.18600319318786587, "grad_norm": 0.31514862179756165, "learning_rate": 1.9984913934510293e-07, "loss": 0.2863, "step": 699 }, { "epoch": 0.18626929217668972, "grad_norm": 0.3428614139556885, "learning_rate": 1.9984820110490256e-07, "loss": 0.3228, "step": 700 }, { "epoch": 0.18653539116551357, "grad_norm": 0.3377818167209625, "learning_rate": 1.998472599583803e-07, "loss": 0.2828, "step": 701 }, { "epoch": 0.18680149015433742, "grad_norm": 0.337080717086792, "learning_rate": 1.9984631590556354e-07, "loss": 0.2961, "step": 702 }, { "epoch": 0.18706758914316127, "grad_norm": 0.3308890163898468, "learning_rate": 1.9984536894647976e-07, "loss": 0.3169, "step": 703 }, { "epoch": 0.1873336881319851, "grad_norm": 0.32823991775512695, "learning_rate": 1.9984441908115654e-07, "loss": 0.2968, "step": 704 }, { "epoch": 0.18759978712080894, "grad_norm": 0.3085198700428009, "learning_rate": 1.9984346630962153e-07, "loss": 0.2909, "step": 705 }, { "epoch": 0.18786588610963278, "grad_norm": 0.3315134644508362, "learning_rate": 1.9984251063190248e-07, "loss": 0.3088, "step": 706 }, { "epoch": 0.18813198509845663, "grad_norm": 0.3292728364467621, "learning_rate": 1.9984155204802713e-07, "loss": 0.2798, "step": 707 }, { "epoch": 0.18839808408728048, "grad_norm": 0.3329780101776123, "learning_rate": 1.9984059055802347e-07, "loss": 0.2943, "step": 708 }, { "epoch": 0.1886641830761043, "grad_norm": 0.3303263485431671, "learning_rate": 1.9983962616191944e-07, "loss": 0.3037, "step": 709 }, { "epoch": 0.18893028206492815, "grad_norm": 0.33177515864372253, "learning_rate": 1.9983865885974315e-07, "loss": 0.2877, "step": 710 }, { "epoch": 0.189196381053752, "grad_norm": 0.33010509610176086, "learning_rate": 1.998376886515227e-07, "loss": 0.2921, "step": 711 }, { "epoch": 0.18946248004257585, "grad_norm": 0.34474635124206543, "learning_rate": 1.9983671553728638e-07, "loss": 0.302, "step": 712 }, { "epoch": 0.18972857903139967, "grad_norm": 0.32528895139694214, "learning_rate": 1.998357395170625e-07, "loss": 0.2983, "step": 713 }, { "epoch": 0.18999467802022352, "grad_norm": 0.3015414774417877, "learning_rate": 1.9983476059087946e-07, "loss": 0.2629, "step": 714 }, { "epoch": 0.19026077700904737, "grad_norm": 0.31074970960617065, "learning_rate": 1.998337787587657e-07, "loss": 0.2996, "step": 715 }, { "epoch": 0.19052687599787121, "grad_norm": 0.33207112550735474, "learning_rate": 1.998327940207499e-07, "loss": 0.2722, "step": 716 }, { "epoch": 0.19079297498669506, "grad_norm": 0.29352062940597534, "learning_rate": 1.998318063768607e-07, "loss": 0.2841, "step": 717 }, { "epoch": 0.19105907397551888, "grad_norm": 0.30646127462387085, "learning_rate": 1.9983081582712683e-07, "loss": 0.2728, "step": 718 }, { "epoch": 0.19132517296434273, "grad_norm": 0.3078974783420563, "learning_rate": 1.9982982237157707e-07, "loss": 0.2979, "step": 719 }, { "epoch": 0.19159127195316658, "grad_norm": 0.3232637047767639, "learning_rate": 1.9982882601024043e-07, "loss": 0.2824, "step": 720 }, { "epoch": 0.19185737094199043, "grad_norm": 0.30567771196365356, "learning_rate": 1.9982782674314585e-07, "loss": 0.2722, "step": 721 }, { "epoch": 0.19212346993081425, "grad_norm": 0.44017863273620605, "learning_rate": 1.9982682457032243e-07, "loss": 0.2957, "step": 722 }, { "epoch": 0.1923895689196381, "grad_norm": 0.32784637808799744, "learning_rate": 1.9982581949179936e-07, "loss": 0.2835, "step": 723 }, { "epoch": 0.19265566790846195, "grad_norm": 0.3384537994861603, "learning_rate": 1.9982481150760586e-07, "loss": 0.3093, "step": 724 }, { "epoch": 0.1929217668972858, "grad_norm": 0.3103841245174408, "learning_rate": 1.9982380061777132e-07, "loss": 0.2853, "step": 725 }, { "epoch": 0.19318786588610964, "grad_norm": 0.3054032623767853, "learning_rate": 1.998227868223251e-07, "loss": 0.2843, "step": 726 }, { "epoch": 0.19345396487493347, "grad_norm": 0.3555256426334381, "learning_rate": 1.9982177012129675e-07, "loss": 0.3139, "step": 727 }, { "epoch": 0.1937200638637573, "grad_norm": 0.301517516374588, "learning_rate": 1.9982075051471584e-07, "loss": 0.2948, "step": 728 }, { "epoch": 0.19398616285258116, "grad_norm": 0.30242419242858887, "learning_rate": 1.998197280026121e-07, "loss": 0.2807, "step": 729 }, { "epoch": 0.194252261841405, "grad_norm": 0.321865975856781, "learning_rate": 1.9981870258501524e-07, "loss": 0.2899, "step": 730 }, { "epoch": 0.19451836083022883, "grad_norm": 0.3436114490032196, "learning_rate": 1.9981767426195516e-07, "loss": 0.2771, "step": 731 }, { "epoch": 0.19478445981905268, "grad_norm": 0.2850719094276428, "learning_rate": 1.9981664303346172e-07, "loss": 0.2692, "step": 732 }, { "epoch": 0.19505055880787653, "grad_norm": 0.32524216175079346, "learning_rate": 1.9981560889956498e-07, "loss": 0.3107, "step": 733 }, { "epoch": 0.19531665779670038, "grad_norm": 0.34355589747428894, "learning_rate": 1.9981457186029502e-07, "loss": 0.2808, "step": 734 }, { "epoch": 0.19558275678552423, "grad_norm": 0.3064979910850525, "learning_rate": 1.9981353191568204e-07, "loss": 0.2922, "step": 735 }, { "epoch": 0.19584885577434805, "grad_norm": 0.31645306944847107, "learning_rate": 1.9981248906575635e-07, "loss": 0.2922, "step": 736 }, { "epoch": 0.1961149547631719, "grad_norm": 0.3182639181613922, "learning_rate": 1.9981144331054823e-07, "loss": 0.2868, "step": 737 }, { "epoch": 0.19638105375199574, "grad_norm": 0.332988977432251, "learning_rate": 1.9981039465008814e-07, "loss": 0.2908, "step": 738 }, { "epoch": 0.1966471527408196, "grad_norm": 0.36458820104599, "learning_rate": 1.9980934308440663e-07, "loss": 0.2825, "step": 739 }, { "epoch": 0.19691325172964344, "grad_norm": 0.3046194314956665, "learning_rate": 1.9980828861353428e-07, "loss": 0.2869, "step": 740 }, { "epoch": 0.19717935071846726, "grad_norm": 0.3302396833896637, "learning_rate": 1.9980723123750182e-07, "loss": 0.3105, "step": 741 }, { "epoch": 0.1974454497072911, "grad_norm": 0.29297876358032227, "learning_rate": 1.9980617095634003e-07, "loss": 0.2845, "step": 742 }, { "epoch": 0.19771154869611496, "grad_norm": 0.30998361110687256, "learning_rate": 1.998051077700797e-07, "loss": 0.2726, "step": 743 }, { "epoch": 0.1979776476849388, "grad_norm": 0.3156895935535431, "learning_rate": 1.9980404167875185e-07, "loss": 0.3103, "step": 744 }, { "epoch": 0.19824374667376263, "grad_norm": 0.32854148745536804, "learning_rate": 1.998029726823875e-07, "loss": 0.2897, "step": 745 }, { "epoch": 0.19850984566258648, "grad_norm": 0.33719587326049805, "learning_rate": 1.998019007810177e-07, "loss": 0.3242, "step": 746 }, { "epoch": 0.19877594465141032, "grad_norm": 0.30602124333381653, "learning_rate": 1.9980082597467375e-07, "loss": 0.2995, "step": 747 }, { "epoch": 0.19904204364023417, "grad_norm": 0.32021623849868774, "learning_rate": 1.9979974826338687e-07, "loss": 0.2686, "step": 748 }, { "epoch": 0.19930814262905802, "grad_norm": 0.33711525797843933, "learning_rate": 1.9979866764718843e-07, "loss": 0.2973, "step": 749 }, { "epoch": 0.19957424161788184, "grad_norm": 0.3548951745033264, "learning_rate": 1.9979758412610992e-07, "loss": 0.3086, "step": 750 }, { "epoch": 0.1998403406067057, "grad_norm": 0.31098490953445435, "learning_rate": 1.9979649770018282e-07, "loss": 0.2721, "step": 751 }, { "epoch": 0.20010643959552954, "grad_norm": 0.32844462990760803, "learning_rate": 1.9979540836943884e-07, "loss": 0.3038, "step": 752 }, { "epoch": 0.2003725385843534, "grad_norm": 0.3143274486064911, "learning_rate": 1.997943161339096e-07, "loss": 0.2936, "step": 753 }, { "epoch": 0.2006386375731772, "grad_norm": 0.33358490467071533, "learning_rate": 1.9979322099362695e-07, "loss": 0.2919, "step": 754 }, { "epoch": 0.20090473656200106, "grad_norm": 0.31973689794540405, "learning_rate": 1.9979212294862278e-07, "loss": 0.2867, "step": 755 }, { "epoch": 0.2011708355508249, "grad_norm": 0.3094802796840668, "learning_rate": 1.99791021998929e-07, "loss": 0.3001, "step": 756 }, { "epoch": 0.20143693453964875, "grad_norm": 0.3676900565624237, "learning_rate": 1.9978991814457763e-07, "loss": 0.3277, "step": 757 }, { "epoch": 0.2017030335284726, "grad_norm": 0.3100546896457672, "learning_rate": 1.997888113856009e-07, "loss": 0.2869, "step": 758 }, { "epoch": 0.20196913251729642, "grad_norm": 0.28830695152282715, "learning_rate": 1.9978770172203094e-07, "loss": 0.2569, "step": 759 }, { "epoch": 0.20223523150612027, "grad_norm": 0.2977626621723175, "learning_rate": 1.9978658915390007e-07, "loss": 0.2882, "step": 760 }, { "epoch": 0.20250133049494412, "grad_norm": 0.3271127939224243, "learning_rate": 1.997854736812407e-07, "loss": 0.3053, "step": 761 }, { "epoch": 0.20276742948376797, "grad_norm": 0.2966686189174652, "learning_rate": 1.9978435530408528e-07, "loss": 0.2648, "step": 762 }, { "epoch": 0.20303352847259182, "grad_norm": 0.3126748502254486, "learning_rate": 1.9978323402246637e-07, "loss": 0.294, "step": 763 }, { "epoch": 0.20329962746141564, "grad_norm": 0.3322683274745941, "learning_rate": 1.997821098364166e-07, "loss": 0.2804, "step": 764 }, { "epoch": 0.2035657264502395, "grad_norm": 0.32447806000709534, "learning_rate": 1.9978098274596865e-07, "loss": 0.2979, "step": 765 }, { "epoch": 0.20383182543906334, "grad_norm": 0.37002599239349365, "learning_rate": 1.9977985275115543e-07, "loss": 0.3048, "step": 766 }, { "epoch": 0.20409792442788718, "grad_norm": 0.31793680787086487, "learning_rate": 1.9977871985200976e-07, "loss": 0.301, "step": 767 }, { "epoch": 0.204364023416711, "grad_norm": 0.31652018427848816, "learning_rate": 1.997775840485646e-07, "loss": 0.2994, "step": 768 }, { "epoch": 0.20463012240553485, "grad_norm": 0.3087935149669647, "learning_rate": 1.9977644534085307e-07, "loss": 0.2682, "step": 769 }, { "epoch": 0.2048962213943587, "grad_norm": 0.34600818157196045, "learning_rate": 1.9977530372890826e-07, "loss": 0.3001, "step": 770 }, { "epoch": 0.20516232038318255, "grad_norm": 0.34300094842910767, "learning_rate": 1.9977415921276345e-07, "loss": 0.3091, "step": 771 }, { "epoch": 0.2054284193720064, "grad_norm": 0.35659047961235046, "learning_rate": 1.9977301179245193e-07, "loss": 0.2972, "step": 772 }, { "epoch": 0.20569451836083022, "grad_norm": 0.3376677930355072, "learning_rate": 1.9977186146800706e-07, "loss": 0.2965, "step": 773 }, { "epoch": 0.20596061734965407, "grad_norm": 0.37442782521247864, "learning_rate": 1.9977070823946235e-07, "loss": 0.2827, "step": 774 }, { "epoch": 0.20622671633847792, "grad_norm": 0.3387735188007355, "learning_rate": 1.9976955210685142e-07, "loss": 0.2991, "step": 775 }, { "epoch": 0.20649281532730177, "grad_norm": 0.3196760416030884, "learning_rate": 1.9976839307020788e-07, "loss": 0.2923, "step": 776 }, { "epoch": 0.20675891431612559, "grad_norm": 0.33084169030189514, "learning_rate": 1.9976723112956542e-07, "loss": 0.296, "step": 777 }, { "epoch": 0.20702501330494943, "grad_norm": 0.44182881712913513, "learning_rate": 1.9976606628495794e-07, "loss": 0.314, "step": 778 }, { "epoch": 0.20729111229377328, "grad_norm": 0.3337228000164032, "learning_rate": 1.9976489853641926e-07, "loss": 0.3163, "step": 779 }, { "epoch": 0.20755721128259713, "grad_norm": 0.3250776529312134, "learning_rate": 1.997637278839835e-07, "loss": 0.2854, "step": 780 }, { "epoch": 0.20782331027142098, "grad_norm": 0.31461894512176514, "learning_rate": 1.9976255432768458e-07, "loss": 0.2863, "step": 781 }, { "epoch": 0.2080894092602448, "grad_norm": 0.3302108645439148, "learning_rate": 1.9976137786755677e-07, "loss": 0.3005, "step": 782 }, { "epoch": 0.20835550824906865, "grad_norm": 0.2957718074321747, "learning_rate": 1.9976019850363426e-07, "loss": 0.2726, "step": 783 }, { "epoch": 0.2086216072378925, "grad_norm": 0.30345389246940613, "learning_rate": 1.9975901623595142e-07, "loss": 0.2691, "step": 784 }, { "epoch": 0.20888770622671635, "grad_norm": 0.307449609041214, "learning_rate": 1.9975783106454262e-07, "loss": 0.2818, "step": 785 }, { "epoch": 0.20915380521554017, "grad_norm": 0.32760968804359436, "learning_rate": 1.997566429894424e-07, "loss": 0.2805, "step": 786 }, { "epoch": 0.20941990420436402, "grad_norm": 0.3362222909927368, "learning_rate": 1.9975545201068532e-07, "loss": 0.2804, "step": 787 }, { "epoch": 0.20968600319318786, "grad_norm": 0.33403870463371277, "learning_rate": 1.99754258128306e-07, "loss": 0.2921, "step": 788 }, { "epoch": 0.2099521021820117, "grad_norm": 0.29712870717048645, "learning_rate": 1.9975306134233926e-07, "loss": 0.3022, "step": 789 }, { "epoch": 0.21021820117083556, "grad_norm": 0.3138263523578644, "learning_rate": 1.997518616528199e-07, "loss": 0.283, "step": 790 }, { "epoch": 0.21048430015965938, "grad_norm": 0.3226606249809265, "learning_rate": 1.9975065905978288e-07, "loss": 0.2794, "step": 791 }, { "epoch": 0.21075039914848323, "grad_norm": 0.2927797734737396, "learning_rate": 1.9974945356326317e-07, "loss": 0.2671, "step": 792 }, { "epoch": 0.21101649813730708, "grad_norm": 0.3351336419582367, "learning_rate": 1.9974824516329581e-07, "loss": 0.2731, "step": 793 }, { "epoch": 0.21128259712613093, "grad_norm": 0.3199041783809662, "learning_rate": 1.997470338599161e-07, "loss": 0.2866, "step": 794 }, { "epoch": 0.21154869611495478, "grad_norm": 0.321880042552948, "learning_rate": 1.997458196531592e-07, "loss": 0.2901, "step": 795 }, { "epoch": 0.2118147951037786, "grad_norm": 0.30418074131011963, "learning_rate": 1.9974460254306046e-07, "loss": 0.2734, "step": 796 }, { "epoch": 0.21208089409260245, "grad_norm": 0.31306344270706177, "learning_rate": 1.9974338252965531e-07, "loss": 0.2936, "step": 797 }, { "epoch": 0.2123469930814263, "grad_norm": 0.2942332625389099, "learning_rate": 1.9974215961297933e-07, "loss": 0.2779, "step": 798 }, { "epoch": 0.21261309207025014, "grad_norm": 0.31527796387672424, "learning_rate": 1.9974093379306804e-07, "loss": 0.2909, "step": 799 }, { "epoch": 0.21287919105907396, "grad_norm": 0.30588120222091675, "learning_rate": 1.9973970506995714e-07, "loss": 0.285, "step": 800 }, { "epoch": 0.2131452900478978, "grad_norm": 0.3107222616672516, "learning_rate": 1.9973847344368238e-07, "loss": 0.271, "step": 801 }, { "epoch": 0.21341138903672166, "grad_norm": 0.32463356852531433, "learning_rate": 1.9973723891427966e-07, "loss": 0.2769, "step": 802 }, { "epoch": 0.2136774880255455, "grad_norm": 0.39869749546051025, "learning_rate": 1.997360014817848e-07, "loss": 0.292, "step": 803 }, { "epoch": 0.21394358701436936, "grad_norm": 0.324387788772583, "learning_rate": 1.9973476114623398e-07, "loss": 0.304, "step": 804 }, { "epoch": 0.21420968600319318, "grad_norm": 0.30837008357048035, "learning_rate": 1.9973351790766322e-07, "loss": 0.2745, "step": 805 }, { "epoch": 0.21447578499201703, "grad_norm": 0.2918148338794708, "learning_rate": 1.9973227176610867e-07, "loss": 0.2671, "step": 806 }, { "epoch": 0.21474188398084088, "grad_norm": 0.3468053340911865, "learning_rate": 1.9973102272160667e-07, "loss": 0.2924, "step": 807 }, { "epoch": 0.21500798296966472, "grad_norm": 0.30315619707107544, "learning_rate": 1.9972977077419353e-07, "loss": 0.2946, "step": 808 }, { "epoch": 0.21527408195848854, "grad_norm": 0.31611916422843933, "learning_rate": 1.9972851592390573e-07, "loss": 0.2916, "step": 809 }, { "epoch": 0.2155401809473124, "grad_norm": 0.32679444551467896, "learning_rate": 1.9972725817077974e-07, "loss": 0.2893, "step": 810 }, { "epoch": 0.21580627993613624, "grad_norm": 0.2986888885498047, "learning_rate": 1.9972599751485224e-07, "loss": 0.2832, "step": 811 }, { "epoch": 0.2160723789249601, "grad_norm": 0.3447900116443634, "learning_rate": 1.9972473395615986e-07, "loss": 0.2907, "step": 812 }, { "epoch": 0.21633847791378394, "grad_norm": 0.3204165995121002, "learning_rate": 1.9972346749473942e-07, "loss": 0.3112, "step": 813 }, { "epoch": 0.21660457690260776, "grad_norm": 0.33191150426864624, "learning_rate": 1.9972219813062775e-07, "loss": 0.2985, "step": 814 }, { "epoch": 0.2168706758914316, "grad_norm": 0.31121084094047546, "learning_rate": 1.9972092586386186e-07, "loss": 0.2795, "step": 815 }, { "epoch": 0.21713677488025546, "grad_norm": 0.32636696100234985, "learning_rate": 1.997196506944787e-07, "loss": 0.2751, "step": 816 }, { "epoch": 0.2174028738690793, "grad_norm": 0.3098302185535431, "learning_rate": 1.9971837262251547e-07, "loss": 0.2773, "step": 817 }, { "epoch": 0.21766897285790315, "grad_norm": 0.31729206442832947, "learning_rate": 1.997170916480093e-07, "loss": 0.31, "step": 818 }, { "epoch": 0.21793507184672697, "grad_norm": 0.3058532774448395, "learning_rate": 1.997158077709975e-07, "loss": 0.2622, "step": 819 }, { "epoch": 0.21820117083555082, "grad_norm": 0.2864273190498352, "learning_rate": 1.9971452099151746e-07, "loss": 0.2727, "step": 820 }, { "epoch": 0.21846726982437467, "grad_norm": 0.36218079924583435, "learning_rate": 1.997132313096066e-07, "loss": 0.3278, "step": 821 }, { "epoch": 0.21873336881319852, "grad_norm": 0.3010192811489105, "learning_rate": 1.9971193872530253e-07, "loss": 0.2708, "step": 822 }, { "epoch": 0.21899946780202234, "grad_norm": 0.304377019405365, "learning_rate": 1.9971064323864275e-07, "loss": 0.2856, "step": 823 }, { "epoch": 0.2192655667908462, "grad_norm": 0.29640907049179077, "learning_rate": 1.9970934484966508e-07, "loss": 0.2938, "step": 824 }, { "epoch": 0.21953166577967004, "grad_norm": 0.31503474712371826, "learning_rate": 1.997080435584073e-07, "loss": 0.2604, "step": 825 }, { "epoch": 0.2197977647684939, "grad_norm": 0.28949809074401855, "learning_rate": 1.9970673936490723e-07, "loss": 0.2938, "step": 826 }, { "epoch": 0.22006386375731773, "grad_norm": 0.3167324662208557, "learning_rate": 1.9970543226920288e-07, "loss": 0.2825, "step": 827 }, { "epoch": 0.22032996274614156, "grad_norm": 0.2995777428150177, "learning_rate": 1.9970412227133225e-07, "loss": 0.2692, "step": 828 }, { "epoch": 0.2205960617349654, "grad_norm": 0.31492483615875244, "learning_rate": 1.997028093713335e-07, "loss": 0.2901, "step": 829 }, { "epoch": 0.22086216072378925, "grad_norm": 0.3025466203689575, "learning_rate": 1.9970149356924488e-07, "loss": 0.2957, "step": 830 }, { "epoch": 0.2211282597126131, "grad_norm": 0.29510727524757385, "learning_rate": 1.9970017486510465e-07, "loss": 0.2636, "step": 831 }, { "epoch": 0.22139435870143692, "grad_norm": 0.33184880018234253, "learning_rate": 1.9969885325895117e-07, "loss": 0.2921, "step": 832 }, { "epoch": 0.22166045769026077, "grad_norm": 0.301895409822464, "learning_rate": 1.9969752875082295e-07, "loss": 0.2671, "step": 833 }, { "epoch": 0.22192655667908462, "grad_norm": 0.34588122367858887, "learning_rate": 1.9969620134075853e-07, "loss": 0.2882, "step": 834 }, { "epoch": 0.22219265566790847, "grad_norm": 0.3137549161911011, "learning_rate": 1.9969487102879655e-07, "loss": 0.2702, "step": 835 }, { "epoch": 0.22245875465673232, "grad_norm": 0.3115924298763275, "learning_rate": 1.9969353781497574e-07, "loss": 0.2926, "step": 836 }, { "epoch": 0.22272485364555614, "grad_norm": 0.3223358988761902, "learning_rate": 1.9969220169933486e-07, "loss": 0.2708, "step": 837 }, { "epoch": 0.22299095263437999, "grad_norm": 0.3251856565475464, "learning_rate": 1.996908626819129e-07, "loss": 0.292, "step": 838 }, { "epoch": 0.22325705162320383, "grad_norm": 0.31815585494041443, "learning_rate": 1.9968952076274868e-07, "loss": 0.2941, "step": 839 }, { "epoch": 0.22352315061202768, "grad_norm": 0.3330617845058441, "learning_rate": 1.9968817594188145e-07, "loss": 0.2755, "step": 840 }, { "epoch": 0.2237892496008515, "grad_norm": 0.28969940543174744, "learning_rate": 1.9968682821935018e-07, "loss": 0.2706, "step": 841 }, { "epoch": 0.22405534858967535, "grad_norm": 0.3243652284145355, "learning_rate": 1.9968547759519424e-07, "loss": 0.2782, "step": 842 }, { "epoch": 0.2243214475784992, "grad_norm": 0.2878596782684326, "learning_rate": 1.9968412406945284e-07, "loss": 0.2687, "step": 843 }, { "epoch": 0.22458754656732305, "grad_norm": 0.295503705739975, "learning_rate": 1.996827676421654e-07, "loss": 0.2752, "step": 844 }, { "epoch": 0.2248536455561469, "grad_norm": 0.29461443424224854, "learning_rate": 1.9968140831337146e-07, "loss": 0.2883, "step": 845 }, { "epoch": 0.22511974454497072, "grad_norm": 0.3151642084121704, "learning_rate": 1.996800460831105e-07, "loss": 0.2734, "step": 846 }, { "epoch": 0.22538584353379457, "grad_norm": 0.3069179058074951, "learning_rate": 1.9967868095142224e-07, "loss": 0.2898, "step": 847 }, { "epoch": 0.22565194252261842, "grad_norm": 0.3190673887729645, "learning_rate": 1.996773129183464e-07, "loss": 0.2743, "step": 848 }, { "epoch": 0.22591804151144226, "grad_norm": 0.2988879084587097, "learning_rate": 1.9967594198392278e-07, "loss": 0.2643, "step": 849 }, { "epoch": 0.2261841405002661, "grad_norm": 0.274616003036499, "learning_rate": 1.996745681481913e-07, "loss": 0.2682, "step": 850 }, { "epoch": 0.22645023948908993, "grad_norm": 0.31480634212493896, "learning_rate": 1.9967319141119193e-07, "loss": 0.2778, "step": 851 }, { "epoch": 0.22671633847791378, "grad_norm": 0.32994842529296875, "learning_rate": 1.9967181177296475e-07, "loss": 0.285, "step": 852 }, { "epoch": 0.22698243746673763, "grad_norm": 0.34222084283828735, "learning_rate": 1.9967042923354994e-07, "loss": 0.27, "step": 853 }, { "epoch": 0.22724853645556148, "grad_norm": 0.33492311835289, "learning_rate": 1.9966904379298773e-07, "loss": 0.2873, "step": 854 }, { "epoch": 0.2275146354443853, "grad_norm": 0.30997374653816223, "learning_rate": 1.9966765545131845e-07, "loss": 0.3027, "step": 855 }, { "epoch": 0.22778073443320915, "grad_norm": 0.3128674626350403, "learning_rate": 1.9966626420858249e-07, "loss": 0.2944, "step": 856 }, { "epoch": 0.228046833422033, "grad_norm": 0.31440436840057373, "learning_rate": 1.9966487006482035e-07, "loss": 0.2878, "step": 857 }, { "epoch": 0.22831293241085684, "grad_norm": 0.29743537306785583, "learning_rate": 1.9966347302007262e-07, "loss": 0.2753, "step": 858 }, { "epoch": 0.2285790313996807, "grad_norm": 0.3028337359428406, "learning_rate": 1.9966207307437998e-07, "loss": 0.2777, "step": 859 }, { "epoch": 0.22884513038850451, "grad_norm": 0.32293760776519775, "learning_rate": 1.9966067022778317e-07, "loss": 0.2848, "step": 860 }, { "epoch": 0.22911122937732836, "grad_norm": 0.311013400554657, "learning_rate": 1.99659264480323e-07, "loss": 0.3025, "step": 861 }, { "epoch": 0.2293773283661522, "grad_norm": 0.3195640742778778, "learning_rate": 1.996578558320404e-07, "loss": 0.2832, "step": 862 }, { "epoch": 0.22964342735497606, "grad_norm": 0.30470511317253113, "learning_rate": 1.996564442829764e-07, "loss": 0.2973, "step": 863 }, { "epoch": 0.22990952634379988, "grad_norm": 0.2820771634578705, "learning_rate": 1.9965502983317204e-07, "loss": 0.2663, "step": 864 }, { "epoch": 0.23017562533262373, "grad_norm": 0.29732900857925415, "learning_rate": 1.996536124826685e-07, "loss": 0.2894, "step": 865 }, { "epoch": 0.23044172432144758, "grad_norm": 0.28821679949760437, "learning_rate": 1.9965219223150708e-07, "loss": 0.2784, "step": 866 }, { "epoch": 0.23070782331027143, "grad_norm": 0.3504526615142822, "learning_rate": 1.9965076907972907e-07, "loss": 0.3029, "step": 867 }, { "epoch": 0.23097392229909527, "grad_norm": 0.29771295189857483, "learning_rate": 1.9964934302737594e-07, "loss": 0.2907, "step": 868 }, { "epoch": 0.2312400212879191, "grad_norm": 0.332999050617218, "learning_rate": 1.9964791407448913e-07, "loss": 0.2789, "step": 869 }, { "epoch": 0.23150612027674294, "grad_norm": 0.3053016662597656, "learning_rate": 1.9964648222111032e-07, "loss": 0.2891, "step": 870 }, { "epoch": 0.2317722192655668, "grad_norm": 0.29576051235198975, "learning_rate": 1.996450474672811e-07, "loss": 0.2777, "step": 871 }, { "epoch": 0.23203831825439064, "grad_norm": 0.29725390672683716, "learning_rate": 1.9964360981304328e-07, "loss": 0.2689, "step": 872 }, { "epoch": 0.2323044172432145, "grad_norm": 0.3104524314403534, "learning_rate": 1.9964216925843874e-07, "loss": 0.2746, "step": 873 }, { "epoch": 0.2325705162320383, "grad_norm": 0.27845460176467896, "learning_rate": 1.996407258035093e-07, "loss": 0.2612, "step": 874 }, { "epoch": 0.23283661522086216, "grad_norm": 0.30781543254852295, "learning_rate": 1.996392794482971e-07, "loss": 0.2647, "step": 875 }, { "epoch": 0.233102714209686, "grad_norm": 0.28733840584754944, "learning_rate": 1.996378301928442e-07, "loss": 0.2584, "step": 876 }, { "epoch": 0.23336881319850986, "grad_norm": 0.3965802490711212, "learning_rate": 1.9963637803719274e-07, "loss": 0.2854, "step": 877 }, { "epoch": 0.23363491218733368, "grad_norm": 0.34147852659225464, "learning_rate": 1.99634922981385e-07, "loss": 0.3028, "step": 878 }, { "epoch": 0.23390101117615753, "grad_norm": 0.3256434202194214, "learning_rate": 1.996334650254634e-07, "loss": 0.2894, "step": 879 }, { "epoch": 0.23416711016498137, "grad_norm": 0.2999359667301178, "learning_rate": 1.9963200416947031e-07, "loss": 0.2682, "step": 880 }, { "epoch": 0.23443320915380522, "grad_norm": 0.30298978090286255, "learning_rate": 1.9963054041344824e-07, "loss": 0.2755, "step": 881 }, { "epoch": 0.23469930814262907, "grad_norm": 0.30733346939086914, "learning_rate": 1.9962907375743986e-07, "loss": 0.2813, "step": 882 }, { "epoch": 0.2349654071314529, "grad_norm": 0.3224698305130005, "learning_rate": 1.9962760420148782e-07, "loss": 0.2961, "step": 883 }, { "epoch": 0.23523150612027674, "grad_norm": 0.28426748514175415, "learning_rate": 1.996261317456349e-07, "loss": 0.2728, "step": 884 }, { "epoch": 0.2354976051091006, "grad_norm": 0.3072284460067749, "learning_rate": 1.9962465638992395e-07, "loss": 0.2851, "step": 885 }, { "epoch": 0.23576370409792444, "grad_norm": 0.33544185757637024, "learning_rate": 1.9962317813439795e-07, "loss": 0.2781, "step": 886 }, { "epoch": 0.23602980308674826, "grad_norm": 0.28492075204849243, "learning_rate": 1.9962169697909988e-07, "loss": 0.2795, "step": 887 }, { "epoch": 0.2362959020755721, "grad_norm": 0.2812511920928955, "learning_rate": 1.996202129240729e-07, "loss": 0.2769, "step": 888 }, { "epoch": 0.23656200106439595, "grad_norm": 0.2856532633304596, "learning_rate": 1.9961872596936017e-07, "loss": 0.2595, "step": 889 }, { "epoch": 0.2368281000532198, "grad_norm": 0.2988225817680359, "learning_rate": 1.99617236115005e-07, "loss": 0.2768, "step": 890 }, { "epoch": 0.23709419904204365, "grad_norm": 0.4074201285839081, "learning_rate": 1.996157433610507e-07, "loss": 0.2843, "step": 891 }, { "epoch": 0.23736029803086747, "grad_norm": 0.3475519120693207, "learning_rate": 1.9961424770754078e-07, "loss": 0.2733, "step": 892 }, { "epoch": 0.23762639701969132, "grad_norm": 0.3183870017528534, "learning_rate": 1.9961274915451876e-07, "loss": 0.2769, "step": 893 }, { "epoch": 0.23789249600851517, "grad_norm": 0.30214571952819824, "learning_rate": 1.9961124770202826e-07, "loss": 0.2912, "step": 894 }, { "epoch": 0.23815859499733902, "grad_norm": 0.3169389069080353, "learning_rate": 1.99609743350113e-07, "loss": 0.2789, "step": 895 }, { "epoch": 0.23842469398616284, "grad_norm": 0.3194439113140106, "learning_rate": 1.996082360988167e-07, "loss": 0.295, "step": 896 }, { "epoch": 0.2386907929749867, "grad_norm": 0.3259661793708801, "learning_rate": 1.996067259481833e-07, "loss": 0.2904, "step": 897 }, { "epoch": 0.23895689196381054, "grad_norm": 0.3227596580982208, "learning_rate": 1.996052128982567e-07, "loss": 0.293, "step": 898 }, { "epoch": 0.23922299095263438, "grad_norm": 0.31091901659965515, "learning_rate": 1.9960369694908102e-07, "loss": 0.2995, "step": 899 }, { "epoch": 0.23948908994145823, "grad_norm": 0.3110003173351288, "learning_rate": 1.9960217810070033e-07, "loss": 0.2665, "step": 900 }, { "epoch": 0.23975518893028205, "grad_norm": 0.3030633330345154, "learning_rate": 1.9960065635315883e-07, "loss": 0.2862, "step": 901 }, { "epoch": 0.2400212879191059, "grad_norm": 0.327632337808609, "learning_rate": 1.9959913170650085e-07, "loss": 0.2764, "step": 902 }, { "epoch": 0.24028738690792975, "grad_norm": 0.3091456890106201, "learning_rate": 1.9959760416077075e-07, "loss": 0.2888, "step": 903 }, { "epoch": 0.2405534858967536, "grad_norm": 0.29968690872192383, "learning_rate": 1.99596073716013e-07, "loss": 0.2824, "step": 904 }, { "epoch": 0.24081958488557745, "grad_norm": 0.32698705792427063, "learning_rate": 1.9959454037227213e-07, "loss": 0.2925, "step": 905 }, { "epoch": 0.24108568387440127, "grad_norm": 0.28863799571990967, "learning_rate": 1.9959300412959278e-07, "loss": 0.2629, "step": 906 }, { "epoch": 0.24135178286322512, "grad_norm": 0.297036349773407, "learning_rate": 1.9959146498801972e-07, "loss": 0.297, "step": 907 }, { "epoch": 0.24161788185204897, "grad_norm": 0.3005645275115967, "learning_rate": 1.9958992294759763e-07, "loss": 0.2547, "step": 908 }, { "epoch": 0.24188398084087281, "grad_norm": 0.3083956241607666, "learning_rate": 1.9958837800837149e-07, "loss": 0.2586, "step": 909 }, { "epoch": 0.24215007982969664, "grad_norm": 0.2967228591442108, "learning_rate": 1.9958683017038625e-07, "loss": 0.2709, "step": 910 }, { "epoch": 0.24241617881852048, "grad_norm": 0.3181453049182892, "learning_rate": 1.9958527943368696e-07, "loss": 0.2789, "step": 911 }, { "epoch": 0.24268227780734433, "grad_norm": 0.34894660115242004, "learning_rate": 1.995837257983188e-07, "loss": 0.3127, "step": 912 }, { "epoch": 0.24294837679616818, "grad_norm": 0.3574464023113251, "learning_rate": 1.995821692643269e-07, "loss": 0.2894, "step": 913 }, { "epoch": 0.24321447578499203, "grad_norm": 0.32112088799476624, "learning_rate": 1.995806098317566e-07, "loss": 0.3192, "step": 914 }, { "epoch": 0.24348057477381585, "grad_norm": 0.3095446825027466, "learning_rate": 1.9957904750065336e-07, "loss": 0.2723, "step": 915 }, { "epoch": 0.2437466737626397, "grad_norm": 0.289343923330307, "learning_rate": 1.9957748227106255e-07, "loss": 0.2881, "step": 916 }, { "epoch": 0.24401277275146355, "grad_norm": 0.3307884633541107, "learning_rate": 1.9957591414302982e-07, "loss": 0.2627, "step": 917 }, { "epoch": 0.2442788717402874, "grad_norm": 0.37872588634490967, "learning_rate": 1.9957434311660074e-07, "loss": 0.2703, "step": 918 }, { "epoch": 0.24454497072911122, "grad_norm": 0.3100660741329193, "learning_rate": 1.9957276919182109e-07, "loss": 0.2961, "step": 919 }, { "epoch": 0.24481106971793506, "grad_norm": 0.3111734688282013, "learning_rate": 1.9957119236873665e-07, "loss": 0.2767, "step": 920 }, { "epoch": 0.2450771687067589, "grad_norm": 0.3168143033981323, "learning_rate": 1.9956961264739337e-07, "loss": 0.2581, "step": 921 }, { "epoch": 0.24534326769558276, "grad_norm": 0.3156871497631073, "learning_rate": 1.9956803002783716e-07, "loss": 0.272, "step": 922 }, { "epoch": 0.2456093666844066, "grad_norm": 0.3192758560180664, "learning_rate": 1.9956644451011416e-07, "loss": 0.2714, "step": 923 }, { "epoch": 0.24587546567323043, "grad_norm": 0.2806965112686157, "learning_rate": 1.9956485609427046e-07, "loss": 0.2604, "step": 924 }, { "epoch": 0.24614156466205428, "grad_norm": 0.341981440782547, "learning_rate": 1.9956326478035232e-07, "loss": 0.2556, "step": 925 }, { "epoch": 0.24640766365087813, "grad_norm": 0.32795077562332153, "learning_rate": 1.9956167056840606e-07, "loss": 0.2833, "step": 926 }, { "epoch": 0.24667376263970198, "grad_norm": 0.27008524537086487, "learning_rate": 1.9956007345847804e-07, "loss": 0.2473, "step": 927 }, { "epoch": 0.24693986162852583, "grad_norm": 0.3221510946750641, "learning_rate": 1.9955847345061482e-07, "loss": 0.2819, "step": 928 }, { "epoch": 0.24720596061734965, "grad_norm": 0.33021727204322815, "learning_rate": 1.9955687054486297e-07, "loss": 0.2868, "step": 929 }, { "epoch": 0.2474720596061735, "grad_norm": 0.32974427938461304, "learning_rate": 1.995552647412691e-07, "loss": 0.2833, "step": 930 }, { "epoch": 0.24773815859499734, "grad_norm": 0.3211110234260559, "learning_rate": 1.9955365603987994e-07, "loss": 0.3096, "step": 931 }, { "epoch": 0.2480042575838212, "grad_norm": 0.28819727897644043, "learning_rate": 1.9955204444074233e-07, "loss": 0.256, "step": 932 }, { "epoch": 0.248270356572645, "grad_norm": 0.31353944540023804, "learning_rate": 1.995504299439032e-07, "loss": 0.2952, "step": 933 }, { "epoch": 0.24853645556146886, "grad_norm": 0.3123358190059662, "learning_rate": 1.9954881254940957e-07, "loss": 0.3012, "step": 934 }, { "epoch": 0.2488025545502927, "grad_norm": 0.3076358437538147, "learning_rate": 1.9954719225730845e-07, "loss": 0.2749, "step": 935 }, { "epoch": 0.24906865353911656, "grad_norm": 0.2995394766330719, "learning_rate": 1.9954556906764705e-07, "loss": 0.2988, "step": 936 }, { "epoch": 0.2493347525279404, "grad_norm": 0.2967182397842407, "learning_rate": 1.995439429804726e-07, "loss": 0.2621, "step": 937 }, { "epoch": 0.24960085151676423, "grad_norm": 0.28438273072242737, "learning_rate": 1.9954231399583241e-07, "loss": 0.273, "step": 938 }, { "epoch": 0.24986695050558808, "grad_norm": 0.30002209544181824, "learning_rate": 1.9954068211377395e-07, "loss": 0.2762, "step": 939 }, { "epoch": 0.2501330494944119, "grad_norm": 0.351039856672287, "learning_rate": 1.9953904733434467e-07, "loss": 0.3032, "step": 940 }, { "epoch": 0.25039914848323575, "grad_norm": 0.3355731666088104, "learning_rate": 1.9953740965759217e-07, "loss": 0.3032, "step": 941 }, { "epoch": 0.2506652474720596, "grad_norm": 0.31410911679267883, "learning_rate": 1.9953576908356414e-07, "loss": 0.2581, "step": 942 }, { "epoch": 0.25093134646088344, "grad_norm": 0.29226547479629517, "learning_rate": 1.9953412561230828e-07, "loss": 0.2675, "step": 943 }, { "epoch": 0.2511974454497073, "grad_norm": 0.3090289235115051, "learning_rate": 1.9953247924387252e-07, "loss": 0.2805, "step": 944 }, { "epoch": 0.25146354443853114, "grad_norm": 0.2824963629245758, "learning_rate": 1.995308299783047e-07, "loss": 0.2638, "step": 945 }, { "epoch": 0.251729643427355, "grad_norm": 0.2825135886669159, "learning_rate": 1.995291778156528e-07, "loss": 0.294, "step": 946 }, { "epoch": 0.25199574241617884, "grad_norm": 0.37974801659584045, "learning_rate": 1.99527522755965e-07, "loss": 0.2786, "step": 947 }, { "epoch": 0.2522618414050027, "grad_norm": 0.3184596300125122, "learning_rate": 1.9952586479928942e-07, "loss": 0.2849, "step": 948 }, { "epoch": 0.2525279403938265, "grad_norm": 0.3012629449367523, "learning_rate": 1.9952420394567436e-07, "loss": 0.28, "step": 949 }, { "epoch": 0.2527940393826503, "grad_norm": 0.28781890869140625, "learning_rate": 1.9952254019516812e-07, "loss": 0.2638, "step": 950 }, { "epoch": 0.2530601383714742, "grad_norm": 0.3297119438648224, "learning_rate": 1.9952087354781913e-07, "loss": 0.2782, "step": 951 }, { "epoch": 0.253326237360298, "grad_norm": 0.33833402395248413, "learning_rate": 1.995192040036759e-07, "loss": 0.2674, "step": 952 }, { "epoch": 0.25359233634912187, "grad_norm": 0.28327205777168274, "learning_rate": 1.9951753156278708e-07, "loss": 0.2826, "step": 953 }, { "epoch": 0.2538584353379457, "grad_norm": 0.2768760919570923, "learning_rate": 1.995158562252013e-07, "loss": 0.2634, "step": 954 }, { "epoch": 0.25412453432676957, "grad_norm": 0.2904130220413208, "learning_rate": 1.9951417799096733e-07, "loss": 0.2808, "step": 955 }, { "epoch": 0.2543906333155934, "grad_norm": 0.2741510272026062, "learning_rate": 1.99512496860134e-07, "loss": 0.2542, "step": 956 }, { "epoch": 0.25465673230441727, "grad_norm": 0.27979692816734314, "learning_rate": 1.995108128327503e-07, "loss": 0.268, "step": 957 }, { "epoch": 0.25492283129324106, "grad_norm": 0.300013929605484, "learning_rate": 1.995091259088652e-07, "loss": 0.2668, "step": 958 }, { "epoch": 0.2551889302820649, "grad_norm": 0.2922171354293823, "learning_rate": 1.9950743608852781e-07, "loss": 0.2686, "step": 959 }, { "epoch": 0.25545502927088876, "grad_norm": 0.32805299758911133, "learning_rate": 1.9950574337178736e-07, "loss": 0.2841, "step": 960 }, { "epoch": 0.2557211282597126, "grad_norm": 0.3201104998588562, "learning_rate": 1.9950404775869307e-07, "loss": 0.2859, "step": 961 }, { "epoch": 0.25598722724853645, "grad_norm": 0.3104017972946167, "learning_rate": 1.995023492492943e-07, "loss": 0.2827, "step": 962 }, { "epoch": 0.2562533262373603, "grad_norm": 0.31730496883392334, "learning_rate": 1.9950064784364052e-07, "loss": 0.285, "step": 963 }, { "epoch": 0.25651942522618415, "grad_norm": 0.3043709099292755, "learning_rate": 1.9949894354178118e-07, "loss": 0.2691, "step": 964 }, { "epoch": 0.256785524215008, "grad_norm": 0.34025606513023376, "learning_rate": 1.9949723634376598e-07, "loss": 0.2816, "step": 965 }, { "epoch": 0.25705162320383185, "grad_norm": 0.3375955820083618, "learning_rate": 1.994955262496446e-07, "loss": 0.2875, "step": 966 }, { "epoch": 0.25731772219265564, "grad_norm": 0.30790024995803833, "learning_rate": 1.9949381325946672e-07, "loss": 0.2861, "step": 967 }, { "epoch": 0.2575838211814795, "grad_norm": 0.32889997959136963, "learning_rate": 1.9949209737328232e-07, "loss": 0.274, "step": 968 }, { "epoch": 0.25784992017030334, "grad_norm": 0.3220454156398773, "learning_rate": 1.9949037859114127e-07, "loss": 0.3017, "step": 969 }, { "epoch": 0.2581160191591272, "grad_norm": 0.3376891016960144, "learning_rate": 1.9948865691309362e-07, "loss": 0.2883, "step": 970 }, { "epoch": 0.25838211814795103, "grad_norm": 0.2818279564380646, "learning_rate": 1.9948693233918948e-07, "loss": 0.2599, "step": 971 }, { "epoch": 0.2586482171367749, "grad_norm": 0.2857404947280884, "learning_rate": 1.994852048694791e-07, "loss": 0.2916, "step": 972 }, { "epoch": 0.25891431612559873, "grad_norm": 0.30892443656921387, "learning_rate": 1.9948347450401267e-07, "loss": 0.2832, "step": 973 }, { "epoch": 0.2591804151144226, "grad_norm": 0.296493262052536, "learning_rate": 1.9948174124284064e-07, "loss": 0.2739, "step": 974 }, { "epoch": 0.25944651410324643, "grad_norm": 0.2979562282562256, "learning_rate": 1.994800050860134e-07, "loss": 0.2892, "step": 975 }, { "epoch": 0.2597126130920703, "grad_norm": 0.35542774200439453, "learning_rate": 1.9947826603358151e-07, "loss": 0.2781, "step": 976 }, { "epoch": 0.25997871208089407, "grad_norm": 0.2819724380970001, "learning_rate": 1.9947652408559562e-07, "loss": 0.2651, "step": 977 }, { "epoch": 0.2602448110697179, "grad_norm": 0.28762274980545044, "learning_rate": 1.9947477924210637e-07, "loss": 0.2569, "step": 978 }, { "epoch": 0.26051091005854177, "grad_norm": 0.360694944858551, "learning_rate": 1.9947303150316458e-07, "loss": 0.2785, "step": 979 }, { "epoch": 0.2607770090473656, "grad_norm": 0.31751421093940735, "learning_rate": 1.9947128086882113e-07, "loss": 0.2944, "step": 980 }, { "epoch": 0.26104310803618946, "grad_norm": 0.2937675714492798, "learning_rate": 1.9946952733912696e-07, "loss": 0.2707, "step": 981 }, { "epoch": 0.2613092070250133, "grad_norm": 0.315540611743927, "learning_rate": 1.994677709141331e-07, "loss": 0.2952, "step": 982 }, { "epoch": 0.26157530601383716, "grad_norm": 0.3015115559101105, "learning_rate": 1.9946601159389073e-07, "loss": 0.2937, "step": 983 }, { "epoch": 0.261841405002661, "grad_norm": 0.3056514263153076, "learning_rate": 1.9946424937845103e-07, "loss": 0.2724, "step": 984 }, { "epoch": 0.26210750399148486, "grad_norm": 0.40656307339668274, "learning_rate": 1.9946248426786527e-07, "loss": 0.2614, "step": 985 }, { "epoch": 0.26237360298030865, "grad_norm": 0.307090163230896, "learning_rate": 1.9946071626218485e-07, "loss": 0.2554, "step": 986 }, { "epoch": 0.2626397019691325, "grad_norm": 0.3502711355686188, "learning_rate": 1.9945894536146122e-07, "loss": 0.2862, "step": 987 }, { "epoch": 0.26290580095795635, "grad_norm": 0.3166431188583374, "learning_rate": 1.9945717156574593e-07, "loss": 0.2696, "step": 988 }, { "epoch": 0.2631718999467802, "grad_norm": 0.3282492458820343, "learning_rate": 1.9945539487509063e-07, "loss": 0.2922, "step": 989 }, { "epoch": 0.26343799893560405, "grad_norm": 0.3605620265007019, "learning_rate": 1.99453615289547e-07, "loss": 0.296, "step": 990 }, { "epoch": 0.2637040979244279, "grad_norm": 0.2942301630973816, "learning_rate": 1.9945183280916687e-07, "loss": 0.2903, "step": 991 }, { "epoch": 0.26397019691325174, "grad_norm": 0.2938957214355469, "learning_rate": 1.994500474340021e-07, "loss": 0.2739, "step": 992 }, { "epoch": 0.2642362959020756, "grad_norm": 0.2884865701198578, "learning_rate": 1.9944825916410468e-07, "loss": 0.2759, "step": 993 }, { "epoch": 0.26450239489089944, "grad_norm": 0.3065718114376068, "learning_rate": 1.9944646799952664e-07, "loss": 0.2901, "step": 994 }, { "epoch": 0.26476849387972323, "grad_norm": 0.2881271541118622, "learning_rate": 1.9944467394032013e-07, "loss": 0.28, "step": 995 }, { "epoch": 0.2650345928685471, "grad_norm": 0.3181980550289154, "learning_rate": 1.9944287698653736e-07, "loss": 0.2886, "step": 996 }, { "epoch": 0.26530069185737093, "grad_norm": 0.3135482370853424, "learning_rate": 1.9944107713823068e-07, "loss": 0.2856, "step": 997 }, { "epoch": 0.2655667908461948, "grad_norm": 0.294893741607666, "learning_rate": 1.994392743954524e-07, "loss": 0.2685, "step": 998 }, { "epoch": 0.2658328898350186, "grad_norm": 0.29189062118530273, "learning_rate": 1.9943746875825504e-07, "loss": 0.2693, "step": 999 }, { "epoch": 0.2660989888238425, "grad_norm": 0.31348296999931335, "learning_rate": 1.9943566022669116e-07, "loss": 0.29, "step": 1000 }, { "epoch": 0.2663650878126663, "grad_norm": 0.2917400002479553, "learning_rate": 1.9943384880081337e-07, "loss": 0.2773, "step": 1001 }, { "epoch": 0.26663118680149017, "grad_norm": 0.3098462224006653, "learning_rate": 1.9943203448067444e-07, "loss": 0.2945, "step": 1002 }, { "epoch": 0.266897285790314, "grad_norm": 0.3255231976509094, "learning_rate": 1.994302172663272e-07, "loss": 0.284, "step": 1003 }, { "epoch": 0.2671633847791378, "grad_norm": 0.30707794427871704, "learning_rate": 1.9942839715782443e-07, "loss": 0.2697, "step": 1004 }, { "epoch": 0.26742948376796166, "grad_norm": 0.31526681780815125, "learning_rate": 1.994265741552192e-07, "loss": 0.2771, "step": 1005 }, { "epoch": 0.2676955827567855, "grad_norm": 0.3041480481624603, "learning_rate": 1.9942474825856458e-07, "loss": 0.2809, "step": 1006 }, { "epoch": 0.26796168174560936, "grad_norm": 0.3579048812389374, "learning_rate": 1.9942291946791367e-07, "loss": 0.3039, "step": 1007 }, { "epoch": 0.2682277807344332, "grad_norm": 0.3048693835735321, "learning_rate": 1.9942108778331973e-07, "loss": 0.2682, "step": 1008 }, { "epoch": 0.26849387972325706, "grad_norm": 0.307282954454422, "learning_rate": 1.9941925320483608e-07, "loss": 0.2903, "step": 1009 }, { "epoch": 0.2687599787120809, "grad_norm": 0.3072441518306732, "learning_rate": 1.994174157325161e-07, "loss": 0.2964, "step": 1010 }, { "epoch": 0.26902607770090475, "grad_norm": 0.28381094336509705, "learning_rate": 1.9941557536641328e-07, "loss": 0.2697, "step": 1011 }, { "epoch": 0.2692921766897286, "grad_norm": 0.32375967502593994, "learning_rate": 1.994137321065812e-07, "loss": 0.2925, "step": 1012 }, { "epoch": 0.2695582756785524, "grad_norm": 0.31693872809410095, "learning_rate": 1.994118859530735e-07, "loss": 0.2872, "step": 1013 }, { "epoch": 0.26982437466737624, "grad_norm": 0.294234961271286, "learning_rate": 1.994100369059439e-07, "loss": 0.2527, "step": 1014 }, { "epoch": 0.2700904736562001, "grad_norm": 0.3165724575519562, "learning_rate": 1.9940818496524628e-07, "loss": 0.2809, "step": 1015 }, { "epoch": 0.27035657264502394, "grad_norm": 0.29818201065063477, "learning_rate": 1.9940633013103447e-07, "loss": 0.2723, "step": 1016 }, { "epoch": 0.2706226716338478, "grad_norm": 0.3613722026348114, "learning_rate": 1.994044724033625e-07, "loss": 0.3108, "step": 1017 }, { "epoch": 0.27088877062267164, "grad_norm": 0.3077780306339264, "learning_rate": 1.9940261178228445e-07, "loss": 0.277, "step": 1018 }, { "epoch": 0.2711548696114955, "grad_norm": 0.29751577973365784, "learning_rate": 1.994007482678545e-07, "loss": 0.2909, "step": 1019 }, { "epoch": 0.27142096860031933, "grad_norm": 0.28754886984825134, "learning_rate": 1.9939888186012682e-07, "loss": 0.281, "step": 1020 }, { "epoch": 0.2716870675891432, "grad_norm": 0.29376837611198425, "learning_rate": 1.9939701255915576e-07, "loss": 0.2811, "step": 1021 }, { "epoch": 0.271953166577967, "grad_norm": 0.28759995102882385, "learning_rate": 1.9939514036499579e-07, "loss": 0.2739, "step": 1022 }, { "epoch": 0.2722192655667908, "grad_norm": 0.28688567876815796, "learning_rate": 1.9939326527770134e-07, "loss": 0.2747, "step": 1023 }, { "epoch": 0.2724853645556147, "grad_norm": 0.28578755259513855, "learning_rate": 1.99391387297327e-07, "loss": 0.2942, "step": 1024 }, { "epoch": 0.2727514635444385, "grad_norm": 0.3038218319416046, "learning_rate": 1.9938950642392744e-07, "loss": 0.2683, "step": 1025 }, { "epoch": 0.27301756253326237, "grad_norm": 0.2786334156990051, "learning_rate": 1.9938762265755744e-07, "loss": 0.2703, "step": 1026 }, { "epoch": 0.2732836615220862, "grad_norm": 0.3173103630542755, "learning_rate": 1.993857359982718e-07, "loss": 0.2921, "step": 1027 }, { "epoch": 0.27354976051091007, "grad_norm": 0.28405511379241943, "learning_rate": 1.993838464461254e-07, "loss": 0.2663, "step": 1028 }, { "epoch": 0.2738158594997339, "grad_norm": 0.28702837228775024, "learning_rate": 1.993819540011733e-07, "loss": 0.2796, "step": 1029 }, { "epoch": 0.27408195848855776, "grad_norm": 0.28677117824554443, "learning_rate": 1.9938005866347056e-07, "loss": 0.2681, "step": 1030 }, { "epoch": 0.2743480574773816, "grad_norm": 0.3293585479259491, "learning_rate": 1.9937816043307235e-07, "loss": 0.2706, "step": 1031 }, { "epoch": 0.2746141564662054, "grad_norm": 0.29125192761421204, "learning_rate": 1.9937625931003392e-07, "loss": 0.2664, "step": 1032 }, { "epoch": 0.27488025545502925, "grad_norm": 0.28900566697120667, "learning_rate": 1.993743552944106e-07, "loss": 0.3006, "step": 1033 }, { "epoch": 0.2751463544438531, "grad_norm": 0.3148365318775177, "learning_rate": 1.9937244838625785e-07, "loss": 0.2646, "step": 1034 }, { "epoch": 0.27541245343267695, "grad_norm": 0.28129011392593384, "learning_rate": 1.9937053858563115e-07, "loss": 0.2555, "step": 1035 }, { "epoch": 0.2756785524215008, "grad_norm": 0.2987145781517029, "learning_rate": 1.9936862589258604e-07, "loss": 0.259, "step": 1036 }, { "epoch": 0.27594465141032465, "grad_norm": 0.33909398317337036, "learning_rate": 1.993667103071783e-07, "loss": 0.2718, "step": 1037 }, { "epoch": 0.2762107503991485, "grad_norm": 0.33802667260169983, "learning_rate": 1.9936479182946355e-07, "loss": 0.2974, "step": 1038 }, { "epoch": 0.27647684938797235, "grad_norm": 0.3195553123950958, "learning_rate": 1.9936287045949776e-07, "loss": 0.2448, "step": 1039 }, { "epoch": 0.2767429483767962, "grad_norm": 0.29681259393692017, "learning_rate": 1.9936094619733678e-07, "loss": 0.2775, "step": 1040 }, { "epoch": 0.27700904736562, "grad_norm": 0.3175794184207916, "learning_rate": 1.9935901904303668e-07, "loss": 0.2663, "step": 1041 }, { "epoch": 0.27727514635444384, "grad_norm": 0.29605767130851746, "learning_rate": 1.9935708899665346e-07, "loss": 0.2871, "step": 1042 }, { "epoch": 0.2775412453432677, "grad_norm": 0.29267916083335876, "learning_rate": 1.993551560582434e-07, "loss": 0.2689, "step": 1043 }, { "epoch": 0.27780734433209153, "grad_norm": 0.28622013330459595, "learning_rate": 1.9935322022786272e-07, "loss": 0.2683, "step": 1044 }, { "epoch": 0.2780734433209154, "grad_norm": 0.2984265387058258, "learning_rate": 1.9935128150556774e-07, "loss": 0.2656, "step": 1045 }, { "epoch": 0.27833954230973923, "grad_norm": 0.2877916991710663, "learning_rate": 1.9934933989141497e-07, "loss": 0.2508, "step": 1046 }, { "epoch": 0.2786056412985631, "grad_norm": 0.34900715947151184, "learning_rate": 1.9934739538546084e-07, "loss": 0.3089, "step": 1047 }, { "epoch": 0.2788717402873869, "grad_norm": 0.32757166028022766, "learning_rate": 1.9934544798776196e-07, "loss": 0.2574, "step": 1048 }, { "epoch": 0.2791378392762108, "grad_norm": 0.3199847638607025, "learning_rate": 1.9934349769837506e-07, "loss": 0.3, "step": 1049 }, { "epoch": 0.27940393826503457, "grad_norm": 0.30960509181022644, "learning_rate": 1.9934154451735687e-07, "loss": 0.2461, "step": 1050 }, { "epoch": 0.2796700372538584, "grad_norm": 0.29585134983062744, "learning_rate": 1.993395884447643e-07, "loss": 0.262, "step": 1051 }, { "epoch": 0.27993613624268227, "grad_norm": 0.3039669394493103, "learning_rate": 1.993376294806542e-07, "loss": 0.2908, "step": 1052 }, { "epoch": 0.2802022352315061, "grad_norm": 0.28250256180763245, "learning_rate": 1.9933566762508365e-07, "loss": 0.2645, "step": 1053 }, { "epoch": 0.28046833422032996, "grad_norm": 0.29528898000717163, "learning_rate": 1.9933370287810972e-07, "loss": 0.2865, "step": 1054 }, { "epoch": 0.2807344332091538, "grad_norm": 0.30788904428482056, "learning_rate": 1.9933173523978965e-07, "loss": 0.2972, "step": 1055 }, { "epoch": 0.28100053219797766, "grad_norm": 0.3150455057621002, "learning_rate": 1.9932976471018068e-07, "loss": 0.2899, "step": 1056 }, { "epoch": 0.2812666311868015, "grad_norm": 0.3013632893562317, "learning_rate": 1.993277912893401e-07, "loss": 0.2734, "step": 1057 }, { "epoch": 0.28153273017562536, "grad_norm": 0.3130698502063751, "learning_rate": 1.993258149773255e-07, "loss": 0.2608, "step": 1058 }, { "epoch": 0.28179882916444915, "grad_norm": 0.306586891412735, "learning_rate": 1.9932383577419428e-07, "loss": 0.2649, "step": 1059 }, { "epoch": 0.282064928153273, "grad_norm": 0.3278392255306244, "learning_rate": 1.9932185368000412e-07, "loss": 0.272, "step": 1060 }, { "epoch": 0.28233102714209685, "grad_norm": 0.27304255962371826, "learning_rate": 1.9931986869481264e-07, "loss": 0.2569, "step": 1061 }, { "epoch": 0.2825971261309207, "grad_norm": 0.2842930555343628, "learning_rate": 1.993178808186777e-07, "loss": 0.2693, "step": 1062 }, { "epoch": 0.28286322511974454, "grad_norm": 0.3092314600944519, "learning_rate": 1.9931589005165713e-07, "loss": 0.2528, "step": 1063 }, { "epoch": 0.2831293241085684, "grad_norm": 0.30035698413848877, "learning_rate": 1.9931389639380883e-07, "loss": 0.2902, "step": 1064 }, { "epoch": 0.28339542309739224, "grad_norm": 0.26341748237609863, "learning_rate": 1.9931189984519092e-07, "loss": 0.2584, "step": 1065 }, { "epoch": 0.2836615220862161, "grad_norm": 0.27306148409843445, "learning_rate": 1.9930990040586146e-07, "loss": 0.2448, "step": 1066 }, { "epoch": 0.28392762107503994, "grad_norm": 0.31244972348213196, "learning_rate": 1.9930789807587862e-07, "loss": 0.2945, "step": 1067 }, { "epoch": 0.28419372006386373, "grad_norm": 0.29776594042778015, "learning_rate": 1.9930589285530076e-07, "loss": 0.2419, "step": 1068 }, { "epoch": 0.2844598190526876, "grad_norm": 0.29308000206947327, "learning_rate": 1.9930388474418618e-07, "loss": 0.2839, "step": 1069 }, { "epoch": 0.2847259180415114, "grad_norm": 0.2948700785636902, "learning_rate": 1.9930187374259334e-07, "loss": 0.2807, "step": 1070 }, { "epoch": 0.2849920170303353, "grad_norm": 0.32598695158958435, "learning_rate": 1.9929985985058083e-07, "loss": 0.2688, "step": 1071 }, { "epoch": 0.2852581160191591, "grad_norm": 0.2880301773548126, "learning_rate": 1.992978430682072e-07, "loss": 0.2623, "step": 1072 }, { "epoch": 0.285524215007983, "grad_norm": 0.3100305497646332, "learning_rate": 1.992958233955312e-07, "loss": 0.2881, "step": 1073 }, { "epoch": 0.2857903139968068, "grad_norm": 0.2756178677082062, "learning_rate": 1.9929380083261158e-07, "loss": 0.2542, "step": 1074 }, { "epoch": 0.28605641298563067, "grad_norm": 0.31665581464767456, "learning_rate": 1.9929177537950722e-07, "loss": 0.2741, "step": 1075 }, { "epoch": 0.2863225119744545, "grad_norm": 0.27285268902778625, "learning_rate": 1.9928974703627714e-07, "loss": 0.2486, "step": 1076 }, { "epoch": 0.2865886109632783, "grad_norm": 0.2764560878276825, "learning_rate": 1.992877158029803e-07, "loss": 0.2631, "step": 1077 }, { "epoch": 0.28685470995210216, "grad_norm": 0.29098501801490784, "learning_rate": 1.9928568167967584e-07, "loss": 0.2687, "step": 1078 }, { "epoch": 0.287120808940926, "grad_norm": 0.32311946153640747, "learning_rate": 1.9928364466642298e-07, "loss": 0.2758, "step": 1079 }, { "epoch": 0.28738690792974986, "grad_norm": 0.2857383191585541, "learning_rate": 1.99281604763281e-07, "loss": 0.2646, "step": 1080 }, { "epoch": 0.2876530069185737, "grad_norm": 0.2958225607872009, "learning_rate": 1.9927956197030932e-07, "loss": 0.273, "step": 1081 }, { "epoch": 0.28791910590739755, "grad_norm": 0.2591635286808014, "learning_rate": 1.9927751628756735e-07, "loss": 0.2667, "step": 1082 }, { "epoch": 0.2881852048962214, "grad_norm": 0.2802017331123352, "learning_rate": 1.9927546771511465e-07, "loss": 0.2775, "step": 1083 }, { "epoch": 0.28845130388504525, "grad_norm": 0.2976182699203491, "learning_rate": 1.9927341625301087e-07, "loss": 0.2539, "step": 1084 }, { "epoch": 0.2887174028738691, "grad_norm": 0.27514562010765076, "learning_rate": 1.9927136190131569e-07, "loss": 0.2645, "step": 1085 }, { "epoch": 0.28898350186269295, "grad_norm": 0.328469455242157, "learning_rate": 1.992693046600889e-07, "loss": 0.2799, "step": 1086 }, { "epoch": 0.28924960085151674, "grad_norm": 0.3890608847141266, "learning_rate": 1.9926724452939042e-07, "loss": 0.2759, "step": 1087 }, { "epoch": 0.2895156998403406, "grad_norm": 0.2946302592754364, "learning_rate": 1.992651815092802e-07, "loss": 0.2719, "step": 1088 }, { "epoch": 0.28978179882916444, "grad_norm": 0.28717127442359924, "learning_rate": 1.992631155998183e-07, "loss": 0.26, "step": 1089 }, { "epoch": 0.2900478978179883, "grad_norm": 0.2997320890426636, "learning_rate": 1.9926104680106482e-07, "loss": 0.2628, "step": 1090 }, { "epoch": 0.29031399680681214, "grad_norm": 0.29811277985572815, "learning_rate": 1.9925897511308e-07, "loss": 0.286, "step": 1091 }, { "epoch": 0.290580095795636, "grad_norm": 0.30899927020072937, "learning_rate": 1.9925690053592413e-07, "loss": 0.2534, "step": 1092 }, { "epoch": 0.29084619478445983, "grad_norm": 0.26760053634643555, "learning_rate": 1.9925482306965762e-07, "loss": 0.2506, "step": 1093 }, { "epoch": 0.2911122937732837, "grad_norm": 0.3082845211029053, "learning_rate": 1.992527427143409e-07, "loss": 0.2631, "step": 1094 }, { "epoch": 0.29137839276210753, "grad_norm": 0.2648814916610718, "learning_rate": 1.9925065947003454e-07, "loss": 0.2594, "step": 1095 }, { "epoch": 0.2916444917509313, "grad_norm": 0.34188008308410645, "learning_rate": 1.9924857333679922e-07, "loss": 0.281, "step": 1096 }, { "epoch": 0.29191059073975517, "grad_norm": 0.31519022583961487, "learning_rate": 1.9924648431469563e-07, "loss": 0.2663, "step": 1097 }, { "epoch": 0.292176689728579, "grad_norm": 0.5684780478477478, "learning_rate": 1.9924439240378455e-07, "loss": 0.2872, "step": 1098 }, { "epoch": 0.29244278871740287, "grad_norm": 0.29836633801460266, "learning_rate": 1.992422976041269e-07, "loss": 0.2667, "step": 1099 }, { "epoch": 0.2927088877062267, "grad_norm": 0.2911379337310791, "learning_rate": 1.9924019991578363e-07, "loss": 0.2544, "step": 1100 }, { "epoch": 0.29297498669505057, "grad_norm": 0.297827810049057, "learning_rate": 1.9923809933881581e-07, "loss": 0.2752, "step": 1101 }, { "epoch": 0.2932410856838744, "grad_norm": 0.3438936173915863, "learning_rate": 1.9923599587328462e-07, "loss": 0.2744, "step": 1102 }, { "epoch": 0.29350718467269826, "grad_norm": 0.31253281235694885, "learning_rate": 1.9923388951925122e-07, "loss": 0.2777, "step": 1103 }, { "epoch": 0.2937732836615221, "grad_norm": 0.30058300495147705, "learning_rate": 1.9923178027677698e-07, "loss": 0.2861, "step": 1104 }, { "epoch": 0.2940393826503459, "grad_norm": 0.2865025997161865, "learning_rate": 1.9922966814592327e-07, "loss": 0.2548, "step": 1105 }, { "epoch": 0.29430548163916975, "grad_norm": 0.31931379437446594, "learning_rate": 1.9922755312675155e-07, "loss": 0.2821, "step": 1106 }, { "epoch": 0.2945715806279936, "grad_norm": 0.31600821018218994, "learning_rate": 1.992254352193234e-07, "loss": 0.2923, "step": 1107 }, { "epoch": 0.29483767961681745, "grad_norm": 0.30939120054244995, "learning_rate": 1.992233144237005e-07, "loss": 0.2744, "step": 1108 }, { "epoch": 0.2951037786056413, "grad_norm": 0.30900683999061584, "learning_rate": 1.992211907399445e-07, "loss": 0.2921, "step": 1109 }, { "epoch": 0.29536987759446515, "grad_norm": 0.31720730662345886, "learning_rate": 1.992190641681173e-07, "loss": 0.2868, "step": 1110 }, { "epoch": 0.295635976583289, "grad_norm": 0.30460214614868164, "learning_rate": 1.9921693470828075e-07, "loss": 0.2658, "step": 1111 }, { "epoch": 0.29590207557211284, "grad_norm": 0.2942042946815491, "learning_rate": 1.9921480236049685e-07, "loss": 0.275, "step": 1112 }, { "epoch": 0.2961681745609367, "grad_norm": 0.29478347301483154, "learning_rate": 1.9921266712482765e-07, "loss": 0.2521, "step": 1113 }, { "epoch": 0.2964342735497605, "grad_norm": 0.35272666811943054, "learning_rate": 1.9921052900133533e-07, "loss": 0.2848, "step": 1114 }, { "epoch": 0.29670037253858433, "grad_norm": 0.31996145844459534, "learning_rate": 1.9920838799008212e-07, "loss": 0.2718, "step": 1115 }, { "epoch": 0.2969664715274082, "grad_norm": 0.31161534786224365, "learning_rate": 1.9920624409113027e-07, "loss": 0.2876, "step": 1116 }, { "epoch": 0.29723257051623203, "grad_norm": 0.3211936950683594, "learning_rate": 1.9920409730454227e-07, "loss": 0.2836, "step": 1117 }, { "epoch": 0.2974986695050559, "grad_norm": 0.3102578818798065, "learning_rate": 1.992019476303806e-07, "loss": 0.2605, "step": 1118 }, { "epoch": 0.29776476849387973, "grad_norm": 0.2943914234638214, "learning_rate": 1.991997950687078e-07, "loss": 0.2628, "step": 1119 }, { "epoch": 0.2980308674827036, "grad_norm": 0.2965167760848999, "learning_rate": 1.9919763961958652e-07, "loss": 0.2616, "step": 1120 }, { "epoch": 0.2982969664715274, "grad_norm": 0.3206964433193207, "learning_rate": 1.9919548128307953e-07, "loss": 0.2553, "step": 1121 }, { "epoch": 0.2985630654603513, "grad_norm": 0.29444634914398193, "learning_rate": 1.9919332005924964e-07, "loss": 0.2781, "step": 1122 }, { "epoch": 0.29882916444917507, "grad_norm": 0.2881354093551636, "learning_rate": 1.9919115594815974e-07, "loss": 0.2748, "step": 1123 }, { "epoch": 0.2990952634379989, "grad_norm": 0.30248892307281494, "learning_rate": 1.9918898894987284e-07, "loss": 0.2369, "step": 1124 }, { "epoch": 0.29936136242682276, "grad_norm": 0.3363252878189087, "learning_rate": 1.9918681906445203e-07, "loss": 0.2841, "step": 1125 }, { "epoch": 0.2996274614156466, "grad_norm": 0.30614709854125977, "learning_rate": 1.991846462919604e-07, "loss": 0.2934, "step": 1126 }, { "epoch": 0.29989356040447046, "grad_norm": 0.2955614924430847, "learning_rate": 1.9918247063246132e-07, "loss": 0.2738, "step": 1127 }, { "epoch": 0.3001596593932943, "grad_norm": 0.296810507774353, "learning_rate": 1.99180292086018e-07, "loss": 0.2592, "step": 1128 }, { "epoch": 0.30042575838211816, "grad_norm": 0.2835467457771301, "learning_rate": 1.9917811065269393e-07, "loss": 0.2649, "step": 1129 }, { "epoch": 0.300691857370942, "grad_norm": 0.2943263649940491, "learning_rate": 1.9917592633255253e-07, "loss": 0.2787, "step": 1130 }, { "epoch": 0.30095795635976585, "grad_norm": 0.2830063998699188, "learning_rate": 1.9917373912565746e-07, "loss": 0.2566, "step": 1131 }, { "epoch": 0.30122405534858965, "grad_norm": 0.29817935824394226, "learning_rate": 1.991715490320723e-07, "loss": 0.2726, "step": 1132 }, { "epoch": 0.3014901543374135, "grad_norm": 0.325920045375824, "learning_rate": 1.9916935605186089e-07, "loss": 0.2787, "step": 1133 }, { "epoch": 0.30175625332623734, "grad_norm": 0.30787569284439087, "learning_rate": 1.9916716018508702e-07, "loss": 0.2793, "step": 1134 }, { "epoch": 0.3020223523150612, "grad_norm": 0.30992868542671204, "learning_rate": 1.9916496143181458e-07, "loss": 0.2608, "step": 1135 }, { "epoch": 0.30228845130388504, "grad_norm": 0.2894720435142517, "learning_rate": 1.991627597921076e-07, "loss": 0.2642, "step": 1136 }, { "epoch": 0.3025545502927089, "grad_norm": 0.2808007001876831, "learning_rate": 1.9916055526603012e-07, "loss": 0.2614, "step": 1137 }, { "epoch": 0.30282064928153274, "grad_norm": 0.2941719889640808, "learning_rate": 1.991583478536464e-07, "loss": 0.2467, "step": 1138 }, { "epoch": 0.3030867482703566, "grad_norm": 0.32708409428596497, "learning_rate": 1.991561375550206e-07, "loss": 0.2665, "step": 1139 }, { "epoch": 0.30335284725918044, "grad_norm": 0.30471205711364746, "learning_rate": 1.9915392437021708e-07, "loss": 0.267, "step": 1140 }, { "epoch": 0.3036189462480043, "grad_norm": 0.29920902848243713, "learning_rate": 1.991517082993003e-07, "loss": 0.2647, "step": 1141 }, { "epoch": 0.3038850452368281, "grad_norm": 0.3195290267467499, "learning_rate": 1.9914948934233475e-07, "loss": 0.2769, "step": 1142 }, { "epoch": 0.3041511442256519, "grad_norm": 0.31483015418052673, "learning_rate": 1.99147267499385e-07, "loss": 0.273, "step": 1143 }, { "epoch": 0.3044172432144758, "grad_norm": 0.29580849409103394, "learning_rate": 1.991450427705157e-07, "loss": 0.2698, "step": 1144 }, { "epoch": 0.3046833422032996, "grad_norm": 0.34444352984428406, "learning_rate": 1.9914281515579164e-07, "loss": 0.2756, "step": 1145 }, { "epoch": 0.30494944119212347, "grad_norm": 0.29666054248809814, "learning_rate": 1.9914058465527765e-07, "loss": 0.2703, "step": 1146 }, { "epoch": 0.3052155401809473, "grad_norm": 0.31825876235961914, "learning_rate": 1.9913835126903865e-07, "loss": 0.3012, "step": 1147 }, { "epoch": 0.30548163916977117, "grad_norm": 0.3006742298603058, "learning_rate": 1.9913611499713969e-07, "loss": 0.2813, "step": 1148 }, { "epoch": 0.305747738158595, "grad_norm": 0.301351934671402, "learning_rate": 1.9913387583964577e-07, "loss": 0.2751, "step": 1149 }, { "epoch": 0.30601383714741887, "grad_norm": 0.28217634558677673, "learning_rate": 1.9913163379662218e-07, "loss": 0.2688, "step": 1150 }, { "epoch": 0.30627993613624266, "grad_norm": 0.3633362948894501, "learning_rate": 1.9912938886813411e-07, "loss": 0.2768, "step": 1151 }, { "epoch": 0.3065460351250665, "grad_norm": 0.32729461789131165, "learning_rate": 1.9912714105424692e-07, "loss": 0.2719, "step": 1152 }, { "epoch": 0.30681213411389036, "grad_norm": 0.30901405215263367, "learning_rate": 1.99124890355026e-07, "loss": 0.2652, "step": 1153 }, { "epoch": 0.3070782331027142, "grad_norm": 0.2769629657268524, "learning_rate": 1.9912263677053693e-07, "loss": 0.2614, "step": 1154 }, { "epoch": 0.30734433209153805, "grad_norm": 0.313815712928772, "learning_rate": 1.9912038030084525e-07, "loss": 0.2743, "step": 1155 }, { "epoch": 0.3076104310803619, "grad_norm": 0.28058797121047974, "learning_rate": 1.9911812094601667e-07, "loss": 0.2748, "step": 1156 }, { "epoch": 0.30787653006918575, "grad_norm": 0.27975431084632874, "learning_rate": 1.9911585870611695e-07, "loss": 0.2697, "step": 1157 }, { "epoch": 0.3081426290580096, "grad_norm": 0.31252503395080566, "learning_rate": 1.9911359358121196e-07, "loss": 0.2846, "step": 1158 }, { "epoch": 0.30840872804683345, "grad_norm": 0.2932601869106293, "learning_rate": 1.9911132557136759e-07, "loss": 0.2714, "step": 1159 }, { "epoch": 0.30867482703565724, "grad_norm": 0.32817694544792175, "learning_rate": 1.9910905467664986e-07, "loss": 0.2781, "step": 1160 }, { "epoch": 0.3089409260244811, "grad_norm": 0.34197819232940674, "learning_rate": 1.9910678089712487e-07, "loss": 0.2836, "step": 1161 }, { "epoch": 0.30920702501330494, "grad_norm": 0.2826898396015167, "learning_rate": 1.9910450423285882e-07, "loss": 0.2715, "step": 1162 }, { "epoch": 0.3094731240021288, "grad_norm": 0.27493563294410706, "learning_rate": 1.99102224683918e-07, "loss": 0.2716, "step": 1163 }, { "epoch": 0.30973922299095263, "grad_norm": 0.32634782791137695, "learning_rate": 1.9909994225036876e-07, "loss": 0.2853, "step": 1164 }, { "epoch": 0.3100053219797765, "grad_norm": 0.2638537585735321, "learning_rate": 1.9909765693227744e-07, "loss": 0.2363, "step": 1165 }, { "epoch": 0.31027142096860033, "grad_norm": 0.29033946990966797, "learning_rate": 1.990953687297107e-07, "loss": 0.2816, "step": 1166 }, { "epoch": 0.3105375199574242, "grad_norm": 0.3032911717891693, "learning_rate": 1.9909307764273502e-07, "loss": 0.2643, "step": 1167 }, { "epoch": 0.31080361894624803, "grad_norm": 0.28708380460739136, "learning_rate": 1.9909078367141716e-07, "loss": 0.2573, "step": 1168 }, { "epoch": 0.3110697179350718, "grad_norm": 0.2714173197746277, "learning_rate": 1.990884868158239e-07, "loss": 0.2605, "step": 1169 }, { "epoch": 0.31133581692389567, "grad_norm": 0.2961517870426178, "learning_rate": 1.9908618707602203e-07, "loss": 0.2633, "step": 1170 }, { "epoch": 0.3116019159127195, "grad_norm": 0.3283216953277588, "learning_rate": 1.9908388445207854e-07, "loss": 0.31, "step": 1171 }, { "epoch": 0.31186801490154337, "grad_norm": 0.3031774163246155, "learning_rate": 1.9908157894406043e-07, "loss": 0.2895, "step": 1172 }, { "epoch": 0.3121341138903672, "grad_norm": 0.2841116786003113, "learning_rate": 1.9907927055203486e-07, "loss": 0.2446, "step": 1173 }, { "epoch": 0.31240021287919106, "grad_norm": 0.28542232513427734, "learning_rate": 1.9907695927606896e-07, "loss": 0.2723, "step": 1174 }, { "epoch": 0.3126663118680149, "grad_norm": 0.3225198984146118, "learning_rate": 1.9907464511623003e-07, "loss": 0.2933, "step": 1175 }, { "epoch": 0.31293241085683876, "grad_norm": 0.29563915729522705, "learning_rate": 1.990723280725854e-07, "loss": 0.2698, "step": 1176 }, { "epoch": 0.3131985098456626, "grad_norm": 0.2805863320827484, "learning_rate": 1.990700081452026e-07, "loss": 0.2759, "step": 1177 }, { "epoch": 0.3134646088344864, "grad_norm": 0.2910847067832947, "learning_rate": 1.9906768533414903e-07, "loss": 0.264, "step": 1178 }, { "epoch": 0.31373070782331025, "grad_norm": 0.2887880504131317, "learning_rate": 1.990653596394924e-07, "loss": 0.2442, "step": 1179 }, { "epoch": 0.3139968068121341, "grad_norm": 0.3599671423435211, "learning_rate": 1.9906303106130035e-07, "loss": 0.2894, "step": 1180 }, { "epoch": 0.31426290580095795, "grad_norm": 0.3045908212661743, "learning_rate": 1.990606995996407e-07, "loss": 0.2461, "step": 1181 }, { "epoch": 0.3145290047897818, "grad_norm": 0.3122466504573822, "learning_rate": 1.990583652545813e-07, "loss": 0.2757, "step": 1182 }, { "epoch": 0.31479510377860564, "grad_norm": 0.2993810176849365, "learning_rate": 1.9905602802619005e-07, "loss": 0.2739, "step": 1183 }, { "epoch": 0.3150612027674295, "grad_norm": 0.2866901159286499, "learning_rate": 1.9905368791453507e-07, "loss": 0.2609, "step": 1184 }, { "epoch": 0.31532730175625334, "grad_norm": 0.26441776752471924, "learning_rate": 1.990513449196844e-07, "loss": 0.2622, "step": 1185 }, { "epoch": 0.3155934007450772, "grad_norm": 0.2849477231502533, "learning_rate": 1.9904899904170626e-07, "loss": 0.2628, "step": 1186 }, { "epoch": 0.315859499733901, "grad_norm": 0.29688847064971924, "learning_rate": 1.9904665028066894e-07, "loss": 0.2692, "step": 1187 }, { "epoch": 0.31612559872272483, "grad_norm": 0.2919471859931946, "learning_rate": 1.990442986366408e-07, "loss": 0.2615, "step": 1188 }, { "epoch": 0.3163916977115487, "grad_norm": 0.290645569562912, "learning_rate": 1.9904194410969032e-07, "loss": 0.2716, "step": 1189 }, { "epoch": 0.31665779670037253, "grad_norm": 0.3086948096752167, "learning_rate": 1.9903958669988595e-07, "loss": 0.2641, "step": 1190 }, { "epoch": 0.3169238956891964, "grad_norm": 0.2968531847000122, "learning_rate": 1.990372264072964e-07, "loss": 0.2773, "step": 1191 }, { "epoch": 0.3171899946780202, "grad_norm": 0.29194512963294983, "learning_rate": 1.9903486323199035e-07, "loss": 0.2459, "step": 1192 }, { "epoch": 0.3174560936668441, "grad_norm": 0.364729106426239, "learning_rate": 1.9903249717403656e-07, "loss": 0.2629, "step": 1193 }, { "epoch": 0.3177221926556679, "grad_norm": 0.2935527265071869, "learning_rate": 1.9903012823350392e-07, "loss": 0.2589, "step": 1194 }, { "epoch": 0.31798829164449177, "grad_norm": 0.3020325303077698, "learning_rate": 1.9902775641046136e-07, "loss": 0.2802, "step": 1195 }, { "epoch": 0.3182543906333156, "grad_norm": 0.29460591077804565, "learning_rate": 1.9902538170497793e-07, "loss": 0.2693, "step": 1196 }, { "epoch": 0.3185204896221394, "grad_norm": 0.3018026053905487, "learning_rate": 1.9902300411712275e-07, "loss": 0.2688, "step": 1197 }, { "epoch": 0.31878658861096326, "grad_norm": 0.3038135766983032, "learning_rate": 1.9902062364696506e-07, "loss": 0.2546, "step": 1198 }, { "epoch": 0.3190526875997871, "grad_norm": 0.2693786323070526, "learning_rate": 1.990182402945741e-07, "loss": 0.2558, "step": 1199 }, { "epoch": 0.31931878658861096, "grad_norm": 0.29192808270454407, "learning_rate": 1.9901585406001929e-07, "loss": 0.2724, "step": 1200 }, { "epoch": 0.3195848855774348, "grad_norm": 0.32011502981185913, "learning_rate": 1.9901346494337003e-07, "loss": 0.261, "step": 1201 }, { "epoch": 0.31985098456625866, "grad_norm": 0.28110167384147644, "learning_rate": 1.9901107294469593e-07, "loss": 0.2556, "step": 1202 }, { "epoch": 0.3201170835550825, "grad_norm": 0.30480366945266724, "learning_rate": 1.990086780640665e-07, "loss": 0.257, "step": 1203 }, { "epoch": 0.32038318254390635, "grad_norm": 0.3821755051612854, "learning_rate": 1.990062803015516e-07, "loss": 0.2567, "step": 1204 }, { "epoch": 0.3206492815327302, "grad_norm": 0.3000302314758301, "learning_rate": 1.9900387965722092e-07, "loss": 0.2914, "step": 1205 }, { "epoch": 0.320915380521554, "grad_norm": 0.33792293071746826, "learning_rate": 1.9900147613114437e-07, "loss": 0.2683, "step": 1206 }, { "epoch": 0.32118147951037784, "grad_norm": 0.2847381830215454, "learning_rate": 1.989990697233919e-07, "loss": 0.248, "step": 1207 }, { "epoch": 0.3214475784992017, "grad_norm": 0.31089237332344055, "learning_rate": 1.9899666043403355e-07, "loss": 0.2881, "step": 1208 }, { "epoch": 0.32171367748802554, "grad_norm": 0.2682235836982727, "learning_rate": 1.9899424826313948e-07, "loss": 0.2566, "step": 1209 }, { "epoch": 0.3219797764768494, "grad_norm": 0.2877580225467682, "learning_rate": 1.9899183321077985e-07, "loss": 0.2538, "step": 1210 }, { "epoch": 0.32224587546567324, "grad_norm": 0.3094998300075531, "learning_rate": 1.9898941527702502e-07, "loss": 0.2854, "step": 1211 }, { "epoch": 0.3225119744544971, "grad_norm": 0.2850254476070404, "learning_rate": 1.9898699446194528e-07, "loss": 0.2832, "step": 1212 }, { "epoch": 0.32277807344332093, "grad_norm": 0.2969076633453369, "learning_rate": 1.9898457076561122e-07, "loss": 0.2593, "step": 1213 }, { "epoch": 0.3230441724321448, "grad_norm": 0.3105247914791107, "learning_rate": 1.9898214418809327e-07, "loss": 0.2682, "step": 1214 }, { "epoch": 0.3233102714209686, "grad_norm": 0.3207141160964966, "learning_rate": 1.9897971472946212e-07, "loss": 0.2721, "step": 1215 }, { "epoch": 0.3235763704097924, "grad_norm": 0.29309138655662537, "learning_rate": 1.9897728238978845e-07, "loss": 0.2546, "step": 1216 }, { "epoch": 0.3238424693986163, "grad_norm": 0.29271623492240906, "learning_rate": 1.9897484716914313e-07, "loss": 0.2752, "step": 1217 }, { "epoch": 0.3241085683874401, "grad_norm": 0.305633008480072, "learning_rate": 1.9897240906759695e-07, "loss": 0.2761, "step": 1218 }, { "epoch": 0.32437466737626397, "grad_norm": 0.2916848957538605, "learning_rate": 1.9896996808522094e-07, "loss": 0.2443, "step": 1219 }, { "epoch": 0.3246407663650878, "grad_norm": 0.32295501232147217, "learning_rate": 1.9896752422208612e-07, "loss": 0.2721, "step": 1220 }, { "epoch": 0.32490686535391167, "grad_norm": 0.298542320728302, "learning_rate": 1.9896507747826362e-07, "loss": 0.2547, "step": 1221 }, { "epoch": 0.3251729643427355, "grad_norm": 0.31275299191474915, "learning_rate": 1.9896262785382473e-07, "loss": 0.2637, "step": 1222 }, { "epoch": 0.32543906333155936, "grad_norm": 0.31040334701538086, "learning_rate": 1.9896017534884067e-07, "loss": 0.2715, "step": 1223 }, { "epoch": 0.32570516232038316, "grad_norm": 0.30839812755584717, "learning_rate": 1.989577199633828e-07, "loss": 0.2503, "step": 1224 }, { "epoch": 0.325971261309207, "grad_norm": 0.3227369785308838, "learning_rate": 1.9895526169752273e-07, "loss": 0.2634, "step": 1225 }, { "epoch": 0.32623736029803085, "grad_norm": 0.29595232009887695, "learning_rate": 1.989528005513319e-07, "loss": 0.2694, "step": 1226 }, { "epoch": 0.3265034592868547, "grad_norm": 0.290327787399292, "learning_rate": 1.9895033652488196e-07, "loss": 0.2612, "step": 1227 }, { "epoch": 0.32676955827567855, "grad_norm": 0.32687172293663025, "learning_rate": 1.9894786961824463e-07, "loss": 0.2596, "step": 1228 }, { "epoch": 0.3270356572645024, "grad_norm": 0.29903364181518555, "learning_rate": 1.9894539983149175e-07, "loss": 0.2718, "step": 1229 }, { "epoch": 0.32730175625332625, "grad_norm": 0.29432445764541626, "learning_rate": 1.989429271646952e-07, "loss": 0.2594, "step": 1230 }, { "epoch": 0.3275678552421501, "grad_norm": 0.298196405172348, "learning_rate": 1.9894045161792695e-07, "loss": 0.2784, "step": 1231 }, { "epoch": 0.32783395423097395, "grad_norm": 0.2901698648929596, "learning_rate": 1.98937973191259e-07, "loss": 0.2706, "step": 1232 }, { "epoch": 0.32810005321979774, "grad_norm": 0.29128244519233704, "learning_rate": 1.9893549188476356e-07, "loss": 0.2499, "step": 1233 }, { "epoch": 0.3283661522086216, "grad_norm": 0.2901170253753662, "learning_rate": 1.9893300769851286e-07, "loss": 0.256, "step": 1234 }, { "epoch": 0.32863225119744544, "grad_norm": 0.31553518772125244, "learning_rate": 1.989305206325792e-07, "loss": 0.2739, "step": 1235 }, { "epoch": 0.3288983501862693, "grad_norm": 0.26803073287010193, "learning_rate": 1.989280306870349e-07, "loss": 0.2497, "step": 1236 }, { "epoch": 0.32916444917509313, "grad_norm": 0.28624090552330017, "learning_rate": 1.9892553786195254e-07, "loss": 0.2641, "step": 1237 }, { "epoch": 0.329430548163917, "grad_norm": 0.3128233551979065, "learning_rate": 1.989230421574046e-07, "loss": 0.2806, "step": 1238 }, { "epoch": 0.32969664715274083, "grad_norm": 0.2964620292186737, "learning_rate": 1.9892054357346375e-07, "loss": 0.2732, "step": 1239 }, { "epoch": 0.3299627461415647, "grad_norm": 0.295514315366745, "learning_rate": 1.9891804211020272e-07, "loss": 0.2577, "step": 1240 }, { "epoch": 0.3302288451303885, "grad_norm": 0.2854217290878296, "learning_rate": 1.9891553776769436e-07, "loss": 0.2634, "step": 1241 }, { "epoch": 0.3304949441192123, "grad_norm": 0.3245949149131775, "learning_rate": 1.989130305460115e-07, "loss": 0.2498, "step": 1242 }, { "epoch": 0.33076104310803617, "grad_norm": 0.2758715748786926, "learning_rate": 1.9891052044522713e-07, "loss": 0.2522, "step": 1243 }, { "epoch": 0.33102714209686, "grad_norm": 0.3250790536403656, "learning_rate": 1.9890800746541437e-07, "loss": 0.2702, "step": 1244 }, { "epoch": 0.33129324108568386, "grad_norm": 0.3123113512992859, "learning_rate": 1.989054916066463e-07, "loss": 0.2571, "step": 1245 }, { "epoch": 0.3315593400745077, "grad_norm": 0.3091909885406494, "learning_rate": 1.9890297286899616e-07, "loss": 0.2653, "step": 1246 }, { "epoch": 0.33182543906333156, "grad_norm": 0.30652061104774475, "learning_rate": 1.989004512525373e-07, "loss": 0.2648, "step": 1247 }, { "epoch": 0.3320915380521554, "grad_norm": 0.2898062467575073, "learning_rate": 1.988979267573431e-07, "loss": 0.2973, "step": 1248 }, { "epoch": 0.33235763704097926, "grad_norm": 0.3442482352256775, "learning_rate": 1.9889539938348702e-07, "loss": 0.2395, "step": 1249 }, { "epoch": 0.3326237360298031, "grad_norm": 0.28880196809768677, "learning_rate": 1.9889286913104265e-07, "loss": 0.2621, "step": 1250 }, { "epoch": 0.33288983501862696, "grad_norm": 0.30702584981918335, "learning_rate": 1.9889033600008362e-07, "loss": 0.2666, "step": 1251 }, { "epoch": 0.33315593400745075, "grad_norm": 1.0089579820632935, "learning_rate": 1.9888779999068369e-07, "loss": 0.2476, "step": 1252 }, { "epoch": 0.3334220329962746, "grad_norm": 0.2944033443927765, "learning_rate": 1.988852611029167e-07, "loss": 0.2593, "step": 1253 }, { "epoch": 0.33368813198509845, "grad_norm": 0.30366581678390503, "learning_rate": 1.9888271933685643e-07, "loss": 0.2762, "step": 1254 }, { "epoch": 0.3339542309739223, "grad_norm": 0.3129632771015167, "learning_rate": 1.98880174692577e-07, "loss": 0.2598, "step": 1255 }, { "epoch": 0.33422032996274614, "grad_norm": 0.31143102049827576, "learning_rate": 1.9887762717015241e-07, "loss": 0.2674, "step": 1256 }, { "epoch": 0.33448642895157, "grad_norm": 0.38434189558029175, "learning_rate": 1.9887507676965685e-07, "loss": 0.2803, "step": 1257 }, { "epoch": 0.33475252794039384, "grad_norm": 0.3682107627391815, "learning_rate": 1.988725234911645e-07, "loss": 0.2616, "step": 1258 }, { "epoch": 0.3350186269292177, "grad_norm": 0.2745867669582367, "learning_rate": 1.9886996733474972e-07, "loss": 0.2432, "step": 1259 }, { "epoch": 0.33528472591804154, "grad_norm": 0.2886018455028534, "learning_rate": 1.9886740830048692e-07, "loss": 0.2444, "step": 1260 }, { "epoch": 0.33555082490686533, "grad_norm": 0.31012752652168274, "learning_rate": 1.9886484638845053e-07, "loss": 0.2832, "step": 1261 }, { "epoch": 0.3358169238956892, "grad_norm": 0.2996438145637512, "learning_rate": 1.9886228159871522e-07, "loss": 0.2668, "step": 1262 }, { "epoch": 0.336083022884513, "grad_norm": 0.3134702742099762, "learning_rate": 1.9885971393135554e-07, "loss": 0.2738, "step": 1263 }, { "epoch": 0.3363491218733369, "grad_norm": 0.3002587854862213, "learning_rate": 1.9885714338644627e-07, "loss": 0.2839, "step": 1264 }, { "epoch": 0.3366152208621607, "grad_norm": 0.30547258257865906, "learning_rate": 1.988545699640623e-07, "loss": 0.2662, "step": 1265 }, { "epoch": 0.3368813198509846, "grad_norm": 0.2858930230140686, "learning_rate": 1.988519936642784e-07, "loss": 0.2791, "step": 1266 }, { "epoch": 0.3371474188398084, "grad_norm": 0.373664528131485, "learning_rate": 1.9884941448716965e-07, "loss": 0.257, "step": 1267 }, { "epoch": 0.33741351782863227, "grad_norm": 0.29824766516685486, "learning_rate": 1.9884683243281113e-07, "loss": 0.2733, "step": 1268 }, { "epoch": 0.3376796168174561, "grad_norm": 0.302634060382843, "learning_rate": 1.9884424750127796e-07, "loss": 0.2728, "step": 1269 }, { "epoch": 0.3379457158062799, "grad_norm": 0.3252600431442261, "learning_rate": 1.988416596926454e-07, "loss": 0.2492, "step": 1270 }, { "epoch": 0.33821181479510376, "grad_norm": 0.2934764325618744, "learning_rate": 1.9883906900698872e-07, "loss": 0.2591, "step": 1271 }, { "epoch": 0.3384779137839276, "grad_norm": 0.31167367100715637, "learning_rate": 1.9883647544438342e-07, "loss": 0.2886, "step": 1272 }, { "epoch": 0.33874401277275146, "grad_norm": 0.31036853790283203, "learning_rate": 1.9883387900490493e-07, "loss": 0.2716, "step": 1273 }, { "epoch": 0.3390101117615753, "grad_norm": 0.33565768599510193, "learning_rate": 1.9883127968862886e-07, "loss": 0.2856, "step": 1274 }, { "epoch": 0.33927621075039915, "grad_norm": 0.297624796628952, "learning_rate": 1.9882867749563083e-07, "loss": 0.2687, "step": 1275 }, { "epoch": 0.339542309739223, "grad_norm": 0.2854304015636444, "learning_rate": 1.988260724259866e-07, "loss": 0.2693, "step": 1276 }, { "epoch": 0.33980840872804685, "grad_norm": 0.2996024191379547, "learning_rate": 1.9882346447977203e-07, "loss": 0.2518, "step": 1277 }, { "epoch": 0.3400745077168707, "grad_norm": 0.2928781807422638, "learning_rate": 1.98820853657063e-07, "loss": 0.2624, "step": 1278 }, { "epoch": 0.3403406067056945, "grad_norm": 0.2883554995059967, "learning_rate": 1.9881823995793544e-07, "loss": 0.297, "step": 1279 }, { "epoch": 0.34060670569451834, "grad_norm": 0.30571141839027405, "learning_rate": 1.9881562338246556e-07, "loss": 0.2643, "step": 1280 }, { "epoch": 0.3408728046833422, "grad_norm": 0.29648131132125854, "learning_rate": 1.9881300393072942e-07, "loss": 0.2708, "step": 1281 }, { "epoch": 0.34113890367216604, "grad_norm": 0.4846916198730469, "learning_rate": 1.9881038160280333e-07, "loss": 0.2699, "step": 1282 }, { "epoch": 0.3414050026609899, "grad_norm": 0.2942487299442291, "learning_rate": 1.9880775639876358e-07, "loss": 0.2627, "step": 1283 }, { "epoch": 0.34167110164981374, "grad_norm": 0.27640700340270996, "learning_rate": 1.988051283186866e-07, "loss": 0.2441, "step": 1284 }, { "epoch": 0.3419372006386376, "grad_norm": 0.2955611050128937, "learning_rate": 1.9880249736264886e-07, "loss": 0.2581, "step": 1285 }, { "epoch": 0.34220329962746143, "grad_norm": 0.29373684525489807, "learning_rate": 1.9879986353072698e-07, "loss": 0.2516, "step": 1286 }, { "epoch": 0.3424693986162853, "grad_norm": 0.29697489738464355, "learning_rate": 1.987972268229976e-07, "loss": 0.2774, "step": 1287 }, { "epoch": 0.3427354976051091, "grad_norm": 0.30728209018707275, "learning_rate": 1.9879458723953746e-07, "loss": 0.285, "step": 1288 }, { "epoch": 0.3430015965939329, "grad_norm": 0.3045642077922821, "learning_rate": 1.987919447804234e-07, "loss": 0.2574, "step": 1289 }, { "epoch": 0.34326769558275677, "grad_norm": 0.3287731111049652, "learning_rate": 1.9878929944573234e-07, "loss": 0.265, "step": 1290 }, { "epoch": 0.3435337945715806, "grad_norm": 0.29683879017829895, "learning_rate": 1.987866512355413e-07, "loss": 0.2593, "step": 1291 }, { "epoch": 0.34379989356040447, "grad_norm": 0.3102242648601532, "learning_rate": 1.9878400014992733e-07, "loss": 0.2766, "step": 1292 }, { "epoch": 0.3440659925492283, "grad_norm": 0.3037049174308777, "learning_rate": 1.9878134618896757e-07, "loss": 0.2614, "step": 1293 }, { "epoch": 0.34433209153805217, "grad_norm": 0.31161579489707947, "learning_rate": 1.9877868935273935e-07, "loss": 0.2817, "step": 1294 }, { "epoch": 0.344598190526876, "grad_norm": 0.2949821352958679, "learning_rate": 1.9877602964131994e-07, "loss": 0.2627, "step": 1295 }, { "epoch": 0.34486428951569986, "grad_norm": 0.30126452445983887, "learning_rate": 1.987733670547868e-07, "loss": 0.2659, "step": 1296 }, { "epoch": 0.34513038850452366, "grad_norm": 0.3150445222854614, "learning_rate": 1.9877070159321738e-07, "loss": 0.2888, "step": 1297 }, { "epoch": 0.3453964874933475, "grad_norm": 0.2655925154685974, "learning_rate": 1.9876803325668932e-07, "loss": 0.2351, "step": 1298 }, { "epoch": 0.34566258648217135, "grad_norm": 0.309830904006958, "learning_rate": 1.9876536204528022e-07, "loss": 0.2585, "step": 1299 }, { "epoch": 0.3459286854709952, "grad_norm": 0.29440629482269287, "learning_rate": 1.987626879590679e-07, "loss": 0.2677, "step": 1300 }, { "epoch": 0.34619478445981905, "grad_norm": 0.2792375385761261, "learning_rate": 1.9876001099813015e-07, "loss": 0.2476, "step": 1301 }, { "epoch": 0.3464608834486429, "grad_norm": 0.2841527462005615, "learning_rate": 1.9875733116254493e-07, "loss": 0.2548, "step": 1302 }, { "epoch": 0.34672698243746675, "grad_norm": 0.2859515845775604, "learning_rate": 1.9875464845239021e-07, "loss": 0.2606, "step": 1303 }, { "epoch": 0.3469930814262906, "grad_norm": 0.24959614872932434, "learning_rate": 1.987519628677441e-07, "loss": 0.2339, "step": 1304 }, { "epoch": 0.34725918041511444, "grad_norm": 0.3313121497631073, "learning_rate": 1.9874927440868473e-07, "loss": 0.2415, "step": 1305 }, { "epoch": 0.3475252794039383, "grad_norm": 0.29349181056022644, "learning_rate": 1.987465830752904e-07, "loss": 0.257, "step": 1306 }, { "epoch": 0.3477913783927621, "grad_norm": 0.31460776925086975, "learning_rate": 1.9874388886763942e-07, "loss": 0.2701, "step": 1307 }, { "epoch": 0.34805747738158593, "grad_norm": 0.2649373412132263, "learning_rate": 1.9874119178581022e-07, "loss": 0.2518, "step": 1308 }, { "epoch": 0.3483235763704098, "grad_norm": 0.2848915755748749, "learning_rate": 1.9873849182988135e-07, "loss": 0.2403, "step": 1309 }, { "epoch": 0.34858967535923363, "grad_norm": 0.3166133761405945, "learning_rate": 1.987357889999313e-07, "loss": 0.2785, "step": 1310 }, { "epoch": 0.3488557743480575, "grad_norm": 0.3101045787334442, "learning_rate": 1.9873308329603883e-07, "loss": 0.2634, "step": 1311 }, { "epoch": 0.3491218733368813, "grad_norm": 0.27821919322013855, "learning_rate": 1.9873037471828266e-07, "loss": 0.2446, "step": 1312 }, { "epoch": 0.3493879723257052, "grad_norm": 0.3133960962295532, "learning_rate": 1.987276632667416e-07, "loss": 0.2804, "step": 1313 }, { "epoch": 0.349654071314529, "grad_norm": 0.2745024263858795, "learning_rate": 1.9872494894149462e-07, "loss": 0.2502, "step": 1314 }, { "epoch": 0.3499201703033529, "grad_norm": 0.32801344990730286, "learning_rate": 1.9872223174262072e-07, "loss": 0.2641, "step": 1315 }, { "epoch": 0.35018626929217667, "grad_norm": 0.3041805922985077, "learning_rate": 1.98719511670199e-07, "loss": 0.2719, "step": 1316 }, { "epoch": 0.3504523682810005, "grad_norm": 0.31275662779808044, "learning_rate": 1.9871678872430858e-07, "loss": 0.2558, "step": 1317 }, { "epoch": 0.35071846726982436, "grad_norm": 0.33940261602401733, "learning_rate": 1.987140629050288e-07, "loss": 0.2756, "step": 1318 }, { "epoch": 0.3509845662586482, "grad_norm": 0.27915146946907043, "learning_rate": 1.987113342124389e-07, "loss": 0.2689, "step": 1319 }, { "epoch": 0.35125066524747206, "grad_norm": 0.2769768536090851, "learning_rate": 1.987086026466184e-07, "loss": 0.2456, "step": 1320 }, { "epoch": 0.3515167642362959, "grad_norm": 0.27881693840026855, "learning_rate": 1.987058682076468e-07, "loss": 0.264, "step": 1321 }, { "epoch": 0.35178286322511976, "grad_norm": 0.3465430736541748, "learning_rate": 1.9870313089560361e-07, "loss": 0.3126, "step": 1322 }, { "epoch": 0.3520489622139436, "grad_norm": 0.2900238633155823, "learning_rate": 1.9870039071056862e-07, "loss": 0.2734, "step": 1323 }, { "epoch": 0.35231506120276745, "grad_norm": 0.3190738260746002, "learning_rate": 1.9869764765262148e-07, "loss": 0.2635, "step": 1324 }, { "epoch": 0.35258116019159125, "grad_norm": 0.30275118350982666, "learning_rate": 1.986949017218421e-07, "loss": 0.2515, "step": 1325 }, { "epoch": 0.3528472591804151, "grad_norm": 0.29763585329055786, "learning_rate": 1.9869215291831036e-07, "loss": 0.2761, "step": 1326 }, { "epoch": 0.35311335816923894, "grad_norm": 0.29168441891670227, "learning_rate": 1.9868940124210633e-07, "loss": 0.2574, "step": 1327 }, { "epoch": 0.3533794571580628, "grad_norm": 0.34342166781425476, "learning_rate": 1.9868664669331007e-07, "loss": 0.2793, "step": 1328 }, { "epoch": 0.35364555614688664, "grad_norm": 0.31495150923728943, "learning_rate": 1.9868388927200176e-07, "loss": 0.27, "step": 1329 }, { "epoch": 0.3539116551357105, "grad_norm": 0.32158026099205017, "learning_rate": 1.986811289782617e-07, "loss": 0.248, "step": 1330 }, { "epoch": 0.35417775412453434, "grad_norm": 0.28808072209358215, "learning_rate": 1.9867836581217016e-07, "loss": 0.2556, "step": 1331 }, { "epoch": 0.3544438531133582, "grad_norm": 0.2969965636730194, "learning_rate": 1.9867559977380756e-07, "loss": 0.2687, "step": 1332 }, { "epoch": 0.35470995210218204, "grad_norm": 0.2974429726600647, "learning_rate": 1.9867283086325454e-07, "loss": 0.2471, "step": 1333 }, { "epoch": 0.35497605109100583, "grad_norm": 0.2829299569129944, "learning_rate": 1.9867005908059158e-07, "loss": 0.2476, "step": 1334 }, { "epoch": 0.3552421500798297, "grad_norm": 0.2919142246246338, "learning_rate": 1.986672844258994e-07, "loss": 0.2591, "step": 1335 }, { "epoch": 0.3555082490686535, "grad_norm": 0.29469722509384155, "learning_rate": 1.9866450689925874e-07, "loss": 0.2612, "step": 1336 }, { "epoch": 0.3557743480574774, "grad_norm": 0.27350619435310364, "learning_rate": 1.986617265007505e-07, "loss": 0.2801, "step": 1337 }, { "epoch": 0.3560404470463012, "grad_norm": 0.29081377387046814, "learning_rate": 1.9865894323045555e-07, "loss": 0.2611, "step": 1338 }, { "epoch": 0.35630654603512507, "grad_norm": 0.27749091386795044, "learning_rate": 1.9865615708845493e-07, "loss": 0.2444, "step": 1339 }, { "epoch": 0.3565726450239489, "grad_norm": 0.27815359830856323, "learning_rate": 1.9865336807482973e-07, "loss": 0.2577, "step": 1340 }, { "epoch": 0.35683874401277277, "grad_norm": 0.29782846570014954, "learning_rate": 1.9865057618966115e-07, "loss": 0.2545, "step": 1341 }, { "epoch": 0.3571048430015966, "grad_norm": 0.30682817101478577, "learning_rate": 1.9864778143303042e-07, "loss": 0.2708, "step": 1342 }, { "epoch": 0.3573709419904204, "grad_norm": 0.26860320568084717, "learning_rate": 1.9864498380501896e-07, "loss": 0.2572, "step": 1343 }, { "epoch": 0.35763704097924426, "grad_norm": 0.2923111319541931, "learning_rate": 1.986421833057081e-07, "loss": 0.2467, "step": 1344 }, { "epoch": 0.3579031399680681, "grad_norm": 0.2809356451034546, "learning_rate": 1.986393799351794e-07, "loss": 0.2579, "step": 1345 }, { "epoch": 0.35816923895689196, "grad_norm": 0.2860639691352844, "learning_rate": 1.986365736935145e-07, "loss": 0.2493, "step": 1346 }, { "epoch": 0.3584353379457158, "grad_norm": 0.28919047117233276, "learning_rate": 1.9863376458079503e-07, "loss": 0.2548, "step": 1347 }, { "epoch": 0.35870143693453965, "grad_norm": 0.3102097511291504, "learning_rate": 1.9863095259710276e-07, "loss": 0.238, "step": 1348 }, { "epoch": 0.3589675359233635, "grad_norm": 0.2938694357872009, "learning_rate": 1.9862813774251958e-07, "loss": 0.25, "step": 1349 }, { "epoch": 0.35923363491218735, "grad_norm": 0.28750869631767273, "learning_rate": 1.9862532001712737e-07, "loss": 0.2422, "step": 1350 }, { "epoch": 0.3594997339010112, "grad_norm": 0.30356645584106445, "learning_rate": 1.986224994210082e-07, "loss": 0.2593, "step": 1351 }, { "epoch": 0.359765832889835, "grad_norm": 0.30334025621414185, "learning_rate": 1.9861967595424413e-07, "loss": 0.2586, "step": 1352 }, { "epoch": 0.36003193187865884, "grad_norm": 0.27217772603034973, "learning_rate": 1.9861684961691733e-07, "loss": 0.247, "step": 1353 }, { "epoch": 0.3602980308674827, "grad_norm": 0.275260865688324, "learning_rate": 1.9861402040911014e-07, "loss": 0.2522, "step": 1354 }, { "epoch": 0.36056412985630654, "grad_norm": 0.2843242883682251, "learning_rate": 1.9861118833090483e-07, "loss": 0.2677, "step": 1355 }, { "epoch": 0.3608302288451304, "grad_norm": 0.27497240900993347, "learning_rate": 1.9860835338238388e-07, "loss": 0.2474, "step": 1356 }, { "epoch": 0.36109632783395423, "grad_norm": 0.3009383976459503, "learning_rate": 1.9860551556362982e-07, "loss": 0.2681, "step": 1357 }, { "epoch": 0.3613624268227781, "grad_norm": 0.30264949798583984, "learning_rate": 1.9860267487472518e-07, "loss": 0.2527, "step": 1358 }, { "epoch": 0.36162852581160193, "grad_norm": 0.3374485671520233, "learning_rate": 1.9859983131575272e-07, "loss": 0.2803, "step": 1359 }, { "epoch": 0.3618946248004258, "grad_norm": 0.29360365867614746, "learning_rate": 1.985969848867952e-07, "loss": 0.2726, "step": 1360 }, { "epoch": 0.36216072378924963, "grad_norm": 0.3084438443183899, "learning_rate": 1.9859413558793544e-07, "loss": 0.2585, "step": 1361 }, { "epoch": 0.3624268227780734, "grad_norm": 0.26633909344673157, "learning_rate": 1.9859128341925638e-07, "loss": 0.249, "step": 1362 }, { "epoch": 0.36269292176689727, "grad_norm": 0.29062619805336, "learning_rate": 1.9858842838084104e-07, "loss": 0.2553, "step": 1363 }, { "epoch": 0.3629590207557211, "grad_norm": 0.34487393498420715, "learning_rate": 1.9858557047277258e-07, "loss": 0.2803, "step": 1364 }, { "epoch": 0.36322511974454497, "grad_norm": 0.2635214626789093, "learning_rate": 1.985827096951341e-07, "loss": 0.2446, "step": 1365 }, { "epoch": 0.3634912187333688, "grad_norm": 0.28571829199790955, "learning_rate": 1.985798460480089e-07, "loss": 0.2569, "step": 1366 }, { "epoch": 0.36375731772219266, "grad_norm": 0.28734683990478516, "learning_rate": 1.9857697953148035e-07, "loss": 0.2629, "step": 1367 }, { "epoch": 0.3640234167110165, "grad_norm": 0.31436091661453247, "learning_rate": 1.985741101456319e-07, "loss": 0.2482, "step": 1368 }, { "epoch": 0.36428951569984036, "grad_norm": 0.27763473987579346, "learning_rate": 1.9857123789054704e-07, "loss": 0.2653, "step": 1369 }, { "epoch": 0.3645556146886642, "grad_norm": 0.2958422303199768, "learning_rate": 1.9856836276630938e-07, "loss": 0.26, "step": 1370 }, { "epoch": 0.364821713677488, "grad_norm": 0.30776476860046387, "learning_rate": 1.985654847730026e-07, "loss": 0.2795, "step": 1371 }, { "epoch": 0.36508781266631185, "grad_norm": 0.3090752065181732, "learning_rate": 1.985626039107105e-07, "loss": 0.2573, "step": 1372 }, { "epoch": 0.3653539116551357, "grad_norm": 0.3100123703479767, "learning_rate": 1.985597201795169e-07, "loss": 0.2739, "step": 1373 }, { "epoch": 0.36562001064395955, "grad_norm": 0.31271660327911377, "learning_rate": 1.9855683357950577e-07, "loss": 0.2718, "step": 1374 }, { "epoch": 0.3658861096327834, "grad_norm": 0.3003235459327698, "learning_rate": 1.985539441107611e-07, "loss": 0.2418, "step": 1375 }, { "epoch": 0.36615220862160724, "grad_norm": 0.3005968928337097, "learning_rate": 1.9855105177336698e-07, "loss": 0.2567, "step": 1376 }, { "epoch": 0.3664183076104311, "grad_norm": 0.29768332839012146, "learning_rate": 1.985481565674077e-07, "loss": 0.257, "step": 1377 }, { "epoch": 0.36668440659925494, "grad_norm": 0.282180517911911, "learning_rate": 1.985452584929674e-07, "loss": 0.2864, "step": 1378 }, { "epoch": 0.3669505055880788, "grad_norm": 0.29564085602760315, "learning_rate": 1.985423575501305e-07, "loss": 0.2605, "step": 1379 }, { "epoch": 0.3672166045769026, "grad_norm": 0.28112122416496277, "learning_rate": 1.9853945373898146e-07, "loss": 0.2451, "step": 1380 }, { "epoch": 0.36748270356572643, "grad_norm": 0.3329603672027588, "learning_rate": 1.9853654705960475e-07, "loss": 0.2547, "step": 1381 }, { "epoch": 0.3677488025545503, "grad_norm": 0.29903337359428406, "learning_rate": 1.9853363751208502e-07, "loss": 0.2621, "step": 1382 }, { "epoch": 0.36801490154337413, "grad_norm": 0.2990029752254486, "learning_rate": 1.9853072509650694e-07, "loss": 0.2666, "step": 1383 }, { "epoch": 0.368281000532198, "grad_norm": 0.2914527356624603, "learning_rate": 1.985278098129553e-07, "loss": 0.2625, "step": 1384 }, { "epoch": 0.3685470995210218, "grad_norm": 0.3061884045600891, "learning_rate": 1.9852489166151493e-07, "loss": 0.2675, "step": 1385 }, { "epoch": 0.3688131985098457, "grad_norm": 0.32935869693756104, "learning_rate": 1.9852197064227076e-07, "loss": 0.2634, "step": 1386 }, { "epoch": 0.3690792974986695, "grad_norm": 0.34197676181793213, "learning_rate": 1.9851904675530787e-07, "loss": 0.2559, "step": 1387 }, { "epoch": 0.36934539648749337, "grad_norm": 0.31530824303627014, "learning_rate": 1.985161200007113e-07, "loss": 0.2589, "step": 1388 }, { "epoch": 0.36961149547631716, "grad_norm": 0.2871042490005493, "learning_rate": 1.985131903785663e-07, "loss": 0.2601, "step": 1389 }, { "epoch": 0.369877594465141, "grad_norm": 0.32172298431396484, "learning_rate": 1.985102578889581e-07, "loss": 0.2647, "step": 1390 }, { "epoch": 0.37014369345396486, "grad_norm": 0.3145880103111267, "learning_rate": 1.9850732253197208e-07, "loss": 0.2642, "step": 1391 }, { "epoch": 0.3704097924427887, "grad_norm": 0.26316577196121216, "learning_rate": 1.9850438430769364e-07, "loss": 0.2579, "step": 1392 }, { "epoch": 0.37067589143161256, "grad_norm": 0.40606504678726196, "learning_rate": 1.9850144321620837e-07, "loss": 0.2627, "step": 1393 }, { "epoch": 0.3709419904204364, "grad_norm": 0.2903010845184326, "learning_rate": 1.9849849925760185e-07, "loss": 0.2731, "step": 1394 }, { "epoch": 0.37120808940926026, "grad_norm": 0.2825796604156494, "learning_rate": 1.9849555243195972e-07, "loss": 0.2492, "step": 1395 }, { "epoch": 0.3714741883980841, "grad_norm": 0.30010631680488586, "learning_rate": 1.9849260273936785e-07, "loss": 0.2459, "step": 1396 }, { "epoch": 0.37174028738690795, "grad_norm": 0.2865167558193207, "learning_rate": 1.9848965017991205e-07, "loss": 0.2517, "step": 1397 }, { "epoch": 0.37200638637573175, "grad_norm": 0.2723991572856903, "learning_rate": 1.9848669475367821e-07, "loss": 0.2212, "step": 1398 }, { "epoch": 0.3722724853645556, "grad_norm": 0.3089700937271118, "learning_rate": 1.9848373646075243e-07, "loss": 0.2716, "step": 1399 }, { "epoch": 0.37253858435337944, "grad_norm": 0.3279455602169037, "learning_rate": 1.984807753012208e-07, "loss": 0.297, "step": 1400 }, { "epoch": 0.3728046833422033, "grad_norm": 0.31739482283592224, "learning_rate": 1.984778112751695e-07, "loss": 0.2571, "step": 1401 }, { "epoch": 0.37307078233102714, "grad_norm": 0.29233667254447937, "learning_rate": 1.984748443826848e-07, "loss": 0.2531, "step": 1402 }, { "epoch": 0.373336881319851, "grad_norm": 0.2698242664337158, "learning_rate": 1.9847187462385308e-07, "loss": 0.2458, "step": 1403 }, { "epoch": 0.37360298030867484, "grad_norm": 0.27540841698646545, "learning_rate": 1.9846890199876075e-07, "loss": 0.2566, "step": 1404 }, { "epoch": 0.3738690792974987, "grad_norm": 0.3009425699710846, "learning_rate": 1.9846592650749439e-07, "loss": 0.268, "step": 1405 }, { "epoch": 0.37413517828632253, "grad_norm": 0.3102268576622009, "learning_rate": 1.9846294815014053e-07, "loss": 0.2579, "step": 1406 }, { "epoch": 0.3744012772751463, "grad_norm": 0.2895706593990326, "learning_rate": 1.9845996692678594e-07, "loss": 0.2751, "step": 1407 }, { "epoch": 0.3746673762639702, "grad_norm": 0.2998179793357849, "learning_rate": 1.9845698283751733e-07, "loss": 0.2656, "step": 1408 }, { "epoch": 0.374933475252794, "grad_norm": 0.2831709086894989, "learning_rate": 1.9845399588242163e-07, "loss": 0.2316, "step": 1409 }, { "epoch": 0.3751995742416179, "grad_norm": 0.3188527822494507, "learning_rate": 1.984510060615857e-07, "loss": 0.2746, "step": 1410 }, { "epoch": 0.3754656732304417, "grad_norm": 0.26600873470306396, "learning_rate": 1.9844801337509665e-07, "loss": 0.258, "step": 1411 }, { "epoch": 0.37573177221926557, "grad_norm": 0.31537967920303345, "learning_rate": 1.9844501782304152e-07, "loss": 0.2686, "step": 1412 }, { "epoch": 0.3759978712080894, "grad_norm": 0.2864602208137512, "learning_rate": 1.9844201940550756e-07, "loss": 0.2679, "step": 1413 }, { "epoch": 0.37626397019691327, "grad_norm": 0.28959548473358154, "learning_rate": 1.9843901812258198e-07, "loss": 0.257, "step": 1414 }, { "epoch": 0.3765300691857371, "grad_norm": 0.32704949378967285, "learning_rate": 1.984360139743522e-07, "loss": 0.286, "step": 1415 }, { "epoch": 0.37679616817456096, "grad_norm": 0.2867986559867859, "learning_rate": 1.9843300696090563e-07, "loss": 0.2656, "step": 1416 }, { "epoch": 0.37706226716338476, "grad_norm": 0.2805382311344147, "learning_rate": 1.984299970823298e-07, "loss": 0.2455, "step": 1417 }, { "epoch": 0.3773283661522086, "grad_norm": 0.29789999127388, "learning_rate": 1.9842698433871233e-07, "loss": 0.2594, "step": 1418 }, { "epoch": 0.37759446514103245, "grad_norm": 0.29171717166900635, "learning_rate": 1.9842396873014094e-07, "loss": 0.2551, "step": 1419 }, { "epoch": 0.3778605641298563, "grad_norm": 0.29242265224456787, "learning_rate": 1.9842095025670334e-07, "loss": 0.2696, "step": 1420 }, { "epoch": 0.37812666311868015, "grad_norm": 0.266382098197937, "learning_rate": 1.9841792891848743e-07, "loss": 0.2435, "step": 1421 }, { "epoch": 0.378392762107504, "grad_norm": 0.30149877071380615, "learning_rate": 1.9841490471558115e-07, "loss": 0.2544, "step": 1422 }, { "epoch": 0.37865886109632785, "grad_norm": 0.3131255805492401, "learning_rate": 1.9841187764807255e-07, "loss": 0.2499, "step": 1423 }, { "epoch": 0.3789249600851517, "grad_norm": 0.35154473781585693, "learning_rate": 1.9840884771604972e-07, "loss": 0.2817, "step": 1424 }, { "epoch": 0.37919105907397554, "grad_norm": 0.3074374794960022, "learning_rate": 1.984058149196008e-07, "loss": 0.2713, "step": 1425 }, { "epoch": 0.37945715806279934, "grad_norm": 0.2996111810207367, "learning_rate": 1.9840277925881415e-07, "loss": 0.2481, "step": 1426 }, { "epoch": 0.3797232570516232, "grad_norm": 0.2654796540737152, "learning_rate": 1.9839974073377806e-07, "loss": 0.246, "step": 1427 }, { "epoch": 0.37998935604044703, "grad_norm": 0.27267780900001526, "learning_rate": 1.9839669934458104e-07, "loss": 0.2453, "step": 1428 }, { "epoch": 0.3802554550292709, "grad_norm": 0.28506872057914734, "learning_rate": 1.983936550913116e-07, "loss": 0.2627, "step": 1429 }, { "epoch": 0.38052155401809473, "grad_norm": 0.31095418334007263, "learning_rate": 1.9839060797405832e-07, "loss": 0.2695, "step": 1430 }, { "epoch": 0.3807876530069186, "grad_norm": 0.2769632041454315, "learning_rate": 1.983875579929099e-07, "loss": 0.2452, "step": 1431 }, { "epoch": 0.38105375199574243, "grad_norm": 0.3028271496295929, "learning_rate": 1.9838450514795515e-07, "loss": 0.2604, "step": 1432 }, { "epoch": 0.3813198509845663, "grad_norm": 0.32752662897109985, "learning_rate": 1.983814494392829e-07, "loss": 0.2556, "step": 1433 }, { "epoch": 0.3815859499733901, "grad_norm": 0.29271912574768066, "learning_rate": 1.983783908669821e-07, "loss": 0.2502, "step": 1434 }, { "epoch": 0.3818520489622139, "grad_norm": 0.31069329380989075, "learning_rate": 1.9837532943114176e-07, "loss": 0.2871, "step": 1435 }, { "epoch": 0.38211814795103777, "grad_norm": 0.26715680956840515, "learning_rate": 1.9837226513185103e-07, "loss": 0.2462, "step": 1436 }, { "epoch": 0.3823842469398616, "grad_norm": 0.29126667976379395, "learning_rate": 1.9836919796919907e-07, "loss": 0.2542, "step": 1437 }, { "epoch": 0.38265034592868546, "grad_norm": 0.3137394189834595, "learning_rate": 1.9836612794327521e-07, "loss": 0.2804, "step": 1438 }, { "epoch": 0.3829164449175093, "grad_norm": 0.3022909164428711, "learning_rate": 1.9836305505416874e-07, "loss": 0.2534, "step": 1439 }, { "epoch": 0.38318254390633316, "grad_norm": 0.275412380695343, "learning_rate": 1.983599793019691e-07, "loss": 0.239, "step": 1440 }, { "epoch": 0.383448642895157, "grad_norm": 0.2670103907585144, "learning_rate": 1.9835690068676588e-07, "loss": 0.235, "step": 1441 }, { "epoch": 0.38371474188398086, "grad_norm": 0.3331789970397949, "learning_rate": 1.9835381920864868e-07, "loss": 0.2806, "step": 1442 }, { "epoch": 0.3839808408728047, "grad_norm": 0.26859694719314575, "learning_rate": 1.9835073486770716e-07, "loss": 0.2602, "step": 1443 }, { "epoch": 0.3842469398616285, "grad_norm": 0.3132483661174774, "learning_rate": 1.9834764766403112e-07, "loss": 0.2701, "step": 1444 }, { "epoch": 0.38451303885045235, "grad_norm": 0.5579981207847595, "learning_rate": 1.9834455759771042e-07, "loss": 0.2493, "step": 1445 }, { "epoch": 0.3847791378392762, "grad_norm": 0.2912994623184204, "learning_rate": 1.9834146466883498e-07, "loss": 0.2608, "step": 1446 }, { "epoch": 0.38504523682810005, "grad_norm": 0.30823591351509094, "learning_rate": 1.9833836887749487e-07, "loss": 0.2674, "step": 1447 }, { "epoch": 0.3853113358169239, "grad_norm": 0.3007138669490814, "learning_rate": 1.9833527022378016e-07, "loss": 0.237, "step": 1448 }, { "epoch": 0.38557743480574774, "grad_norm": 0.2754075229167938, "learning_rate": 1.9833216870778105e-07, "loss": 0.2494, "step": 1449 }, { "epoch": 0.3858435337945716, "grad_norm": 0.2617608606815338, "learning_rate": 1.9832906432958782e-07, "loss": 0.2482, "step": 1450 }, { "epoch": 0.38610963278339544, "grad_norm": 0.3362690508365631, "learning_rate": 1.9832595708929085e-07, "loss": 0.2522, "step": 1451 }, { "epoch": 0.3863757317722193, "grad_norm": 0.3467326760292053, "learning_rate": 1.9832284698698057e-07, "loss": 0.2766, "step": 1452 }, { "epoch": 0.3866418307610431, "grad_norm": 0.2901795506477356, "learning_rate": 1.983197340227475e-07, "loss": 0.2552, "step": 1453 }, { "epoch": 0.38690792974986693, "grad_norm": 0.29913464188575745, "learning_rate": 1.9831661819668227e-07, "loss": 0.2602, "step": 1454 }, { "epoch": 0.3871740287386908, "grad_norm": 0.3021884262561798, "learning_rate": 1.9831349950887555e-07, "loss": 0.2673, "step": 1455 }, { "epoch": 0.3874401277275146, "grad_norm": 0.25859227776527405, "learning_rate": 1.9831037795941815e-07, "loss": 0.2336, "step": 1456 }, { "epoch": 0.3877062267163385, "grad_norm": 0.3463759124279022, "learning_rate": 1.983072535484009e-07, "loss": 0.2683, "step": 1457 }, { "epoch": 0.3879723257051623, "grad_norm": 0.2938261926174164, "learning_rate": 1.9830412627591472e-07, "loss": 0.2527, "step": 1458 }, { "epoch": 0.3882384246939862, "grad_norm": 0.30912357568740845, "learning_rate": 1.9830099614205066e-07, "loss": 0.2639, "step": 1459 }, { "epoch": 0.38850452368281, "grad_norm": 0.28787243366241455, "learning_rate": 1.982978631468999e-07, "loss": 0.2588, "step": 1460 }, { "epoch": 0.38877062267163387, "grad_norm": 0.2697773575782776, "learning_rate": 1.982947272905535e-07, "loss": 0.2296, "step": 1461 }, { "epoch": 0.38903672166045766, "grad_norm": 0.266329824924469, "learning_rate": 1.9829158857310286e-07, "loss": 0.2382, "step": 1462 }, { "epoch": 0.3893028206492815, "grad_norm": 0.28173303604125977, "learning_rate": 1.982884469946393e-07, "loss": 0.2508, "step": 1463 }, { "epoch": 0.38956891963810536, "grad_norm": 0.284255713224411, "learning_rate": 1.9828530255525423e-07, "loss": 0.2531, "step": 1464 }, { "epoch": 0.3898350186269292, "grad_norm": 0.2631402909755707, "learning_rate": 1.9828215525503917e-07, "loss": 0.2485, "step": 1465 }, { "epoch": 0.39010111761575306, "grad_norm": 0.30266115069389343, "learning_rate": 1.982790050940858e-07, "loss": 0.2599, "step": 1466 }, { "epoch": 0.3903672166045769, "grad_norm": 0.2924025356769562, "learning_rate": 1.9827585207248573e-07, "loss": 0.2572, "step": 1467 }, { "epoch": 0.39063331559340075, "grad_norm": 0.29837700724601746, "learning_rate": 1.982726961903308e-07, "loss": 0.2547, "step": 1468 }, { "epoch": 0.3908994145822246, "grad_norm": 0.29360970854759216, "learning_rate": 1.9826953744771285e-07, "loss": 0.2501, "step": 1469 }, { "epoch": 0.39116551357104845, "grad_norm": 0.2690393030643463, "learning_rate": 1.9826637584472382e-07, "loss": 0.245, "step": 1470 }, { "epoch": 0.3914316125598723, "grad_norm": 0.31789666414260864, "learning_rate": 1.9826321138145573e-07, "loss": 0.2445, "step": 1471 }, { "epoch": 0.3916977115486961, "grad_norm": 0.277167409658432, "learning_rate": 1.982600440580007e-07, "loss": 0.2465, "step": 1472 }, { "epoch": 0.39196381053751994, "grad_norm": 0.3049260079860687, "learning_rate": 1.9825687387445092e-07, "loss": 0.2533, "step": 1473 }, { "epoch": 0.3922299095263438, "grad_norm": 0.2835114598274231, "learning_rate": 1.9825370083089866e-07, "loss": 0.2438, "step": 1474 }, { "epoch": 0.39249600851516764, "grad_norm": 0.28525158762931824, "learning_rate": 1.9825052492743627e-07, "loss": 0.258, "step": 1475 }, { "epoch": 0.3927621075039915, "grad_norm": 0.3046717643737793, "learning_rate": 1.982473461641562e-07, "loss": 0.2646, "step": 1476 }, { "epoch": 0.39302820649281534, "grad_norm": 0.27876707911491394, "learning_rate": 1.98244164541151e-07, "loss": 0.2149, "step": 1477 }, { "epoch": 0.3932943054816392, "grad_norm": 0.3006533682346344, "learning_rate": 1.9824098005851324e-07, "loss": 0.2737, "step": 1478 }, { "epoch": 0.39356040447046303, "grad_norm": 0.3014675974845886, "learning_rate": 1.9823779271633562e-07, "loss": 0.2909, "step": 1479 }, { "epoch": 0.3938265034592869, "grad_norm": 0.28590866923332214, "learning_rate": 1.9823460251471097e-07, "loss": 0.2579, "step": 1480 }, { "epoch": 0.3940926024481107, "grad_norm": 0.32159656286239624, "learning_rate": 1.9823140945373207e-07, "loss": 0.2703, "step": 1481 }, { "epoch": 0.3943587014369345, "grad_norm": 0.30437949299812317, "learning_rate": 1.982282135334919e-07, "loss": 0.2635, "step": 1482 }, { "epoch": 0.39462480042575837, "grad_norm": 0.2828548848628998, "learning_rate": 1.982250147540835e-07, "loss": 0.249, "step": 1483 }, { "epoch": 0.3948908994145822, "grad_norm": 0.30336442589759827, "learning_rate": 1.9822181311559993e-07, "loss": 0.2567, "step": 1484 }, { "epoch": 0.39515699840340607, "grad_norm": 0.24826665222644806, "learning_rate": 1.9821860861813443e-07, "loss": 0.2259, "step": 1485 }, { "epoch": 0.3954230973922299, "grad_norm": 0.2764692008495331, "learning_rate": 1.9821540126178027e-07, "loss": 0.2484, "step": 1486 }, { "epoch": 0.39568919638105376, "grad_norm": 0.29933789372444153, "learning_rate": 1.9821219104663076e-07, "loss": 0.2657, "step": 1487 }, { "epoch": 0.3959552953698776, "grad_norm": 0.2884780168533325, "learning_rate": 1.982089779727794e-07, "loss": 0.2591, "step": 1488 }, { "epoch": 0.39622139435870146, "grad_norm": 0.3179933726787567, "learning_rate": 1.9820576204031967e-07, "loss": 0.2426, "step": 1489 }, { "epoch": 0.39648749334752525, "grad_norm": 0.323962926864624, "learning_rate": 1.982025432493452e-07, "loss": 0.2608, "step": 1490 }, { "epoch": 0.3967535923363491, "grad_norm": 0.2859404683113098, "learning_rate": 1.981993215999497e-07, "loss": 0.2553, "step": 1491 }, { "epoch": 0.39701969132517295, "grad_norm": 0.30248725414276123, "learning_rate": 1.981960970922269e-07, "loss": 0.2818, "step": 1492 }, { "epoch": 0.3972857903139968, "grad_norm": 0.2886900007724762, "learning_rate": 1.9819286972627065e-07, "loss": 0.2475, "step": 1493 }, { "epoch": 0.39755188930282065, "grad_norm": 0.3203423023223877, "learning_rate": 1.9818963950217496e-07, "loss": 0.2594, "step": 1494 }, { "epoch": 0.3978179882916445, "grad_norm": 0.2929001450538635, "learning_rate": 1.9818640642003378e-07, "loss": 0.2649, "step": 1495 }, { "epoch": 0.39808408728046835, "grad_norm": 0.30495232343673706, "learning_rate": 1.9818317047994126e-07, "loss": 0.2431, "step": 1496 }, { "epoch": 0.3983501862692922, "grad_norm": 0.30973806977272034, "learning_rate": 1.981799316819916e-07, "loss": 0.2641, "step": 1497 }, { "epoch": 0.39861628525811604, "grad_norm": 0.3042203187942505, "learning_rate": 1.9817669002627904e-07, "loss": 0.2702, "step": 1498 }, { "epoch": 0.39888238424693984, "grad_norm": 0.2966604232788086, "learning_rate": 1.9817344551289791e-07, "loss": 0.2614, "step": 1499 }, { "epoch": 0.3991484832357637, "grad_norm": 0.31042420864105225, "learning_rate": 1.9817019814194272e-07, "loss": 0.2498, "step": 1500 }, { "epoch": 0.39941458222458753, "grad_norm": 0.2739895284175873, "learning_rate": 1.9816694791350796e-07, "loss": 0.2411, "step": 1501 }, { "epoch": 0.3996806812134114, "grad_norm": 0.3002561330795288, "learning_rate": 1.9816369482768823e-07, "loss": 0.2709, "step": 1502 }, { "epoch": 0.39994678020223523, "grad_norm": 0.2891824245452881, "learning_rate": 1.981604388845782e-07, "loss": 0.2599, "step": 1503 }, { "epoch": 0.4002128791910591, "grad_norm": 0.297430157661438, "learning_rate": 1.981571800842727e-07, "loss": 0.2626, "step": 1504 }, { "epoch": 0.4004789781798829, "grad_norm": 0.2935275733470917, "learning_rate": 1.9815391842686652e-07, "loss": 0.247, "step": 1505 }, { "epoch": 0.4007450771687068, "grad_norm": 0.3004930913448334, "learning_rate": 1.9815065391245465e-07, "loss": 0.2653, "step": 1506 }, { "epoch": 0.4010111761575306, "grad_norm": 0.3013322055339813, "learning_rate": 1.981473865411321e-07, "loss": 0.2698, "step": 1507 }, { "epoch": 0.4012772751463544, "grad_norm": 0.3185555040836334, "learning_rate": 1.9814411631299395e-07, "loss": 0.28, "step": 1508 }, { "epoch": 0.40154337413517827, "grad_norm": 0.3059003949165344, "learning_rate": 1.9814084322813537e-07, "loss": 0.285, "step": 1509 }, { "epoch": 0.4018094731240021, "grad_norm": 0.2848001718521118, "learning_rate": 1.981375672866517e-07, "loss": 0.2478, "step": 1510 }, { "epoch": 0.40207557211282596, "grad_norm": 0.2767997682094574, "learning_rate": 1.9813428848863825e-07, "loss": 0.2602, "step": 1511 }, { "epoch": 0.4023416711016498, "grad_norm": 0.294192373752594, "learning_rate": 1.9813100683419046e-07, "loss": 0.2732, "step": 1512 }, { "epoch": 0.40260777009047366, "grad_norm": 0.30638810992240906, "learning_rate": 1.9812772232340386e-07, "loss": 0.2613, "step": 1513 }, { "epoch": 0.4028738690792975, "grad_norm": 0.28194186091423035, "learning_rate": 1.9812443495637407e-07, "loss": 0.228, "step": 1514 }, { "epoch": 0.40313996806812136, "grad_norm": 0.2873266935348511, "learning_rate": 1.9812114473319675e-07, "loss": 0.2435, "step": 1515 }, { "epoch": 0.4034060670569452, "grad_norm": 0.35697200894355774, "learning_rate": 1.9811785165396766e-07, "loss": 0.2935, "step": 1516 }, { "epoch": 0.403672166045769, "grad_norm": 0.2857469320297241, "learning_rate": 1.9811455571878265e-07, "loss": 0.2363, "step": 1517 }, { "epoch": 0.40393826503459285, "grad_norm": 0.2943231761455536, "learning_rate": 1.981112569277377e-07, "loss": 0.2548, "step": 1518 }, { "epoch": 0.4042043640234167, "grad_norm": 0.2909819781780243, "learning_rate": 1.9810795528092884e-07, "loss": 0.263, "step": 1519 }, { "epoch": 0.40447046301224054, "grad_norm": 0.2938629686832428, "learning_rate": 1.9810465077845206e-07, "loss": 0.2826, "step": 1520 }, { "epoch": 0.4047365620010644, "grad_norm": 0.3108561635017395, "learning_rate": 1.981013434204037e-07, "loss": 0.2577, "step": 1521 }, { "epoch": 0.40500266098988824, "grad_norm": 0.2818923890590668, "learning_rate": 1.9809803320687988e-07, "loss": 0.2568, "step": 1522 }, { "epoch": 0.4052687599787121, "grad_norm": 0.3411194682121277, "learning_rate": 1.9809472013797705e-07, "loss": 0.2651, "step": 1523 }, { "epoch": 0.40553485896753594, "grad_norm": 0.30580008029937744, "learning_rate": 1.9809140421379165e-07, "loss": 0.2652, "step": 1524 }, { "epoch": 0.4058009579563598, "grad_norm": 0.4050566554069519, "learning_rate": 1.980880854344201e-07, "loss": 0.2658, "step": 1525 }, { "epoch": 0.40606705694518364, "grad_norm": 0.35592037439346313, "learning_rate": 1.9808476379995912e-07, "loss": 0.2902, "step": 1526 }, { "epoch": 0.40633315593400743, "grad_norm": 0.3064337372779846, "learning_rate": 1.9808143931050533e-07, "loss": 0.2368, "step": 1527 }, { "epoch": 0.4065992549228313, "grad_norm": 0.275077760219574, "learning_rate": 1.980781119661555e-07, "loss": 0.2411, "step": 1528 }, { "epoch": 0.4068653539116551, "grad_norm": 0.2886120676994324, "learning_rate": 1.980747817670065e-07, "loss": 0.2597, "step": 1529 }, { "epoch": 0.407131452900479, "grad_norm": 0.27874135971069336, "learning_rate": 1.9807144871315522e-07, "loss": 0.2443, "step": 1530 }, { "epoch": 0.4073975518893028, "grad_norm": 0.295693963766098, "learning_rate": 1.9806811280469872e-07, "loss": 0.2704, "step": 1531 }, { "epoch": 0.40766365087812667, "grad_norm": 0.29081740975379944, "learning_rate": 1.980647740417341e-07, "loss": 0.258, "step": 1532 }, { "epoch": 0.4079297498669505, "grad_norm": 0.3031485378742218, "learning_rate": 1.9806143242435852e-07, "loss": 0.2696, "step": 1533 }, { "epoch": 0.40819584885577437, "grad_norm": 0.2848692834377289, "learning_rate": 1.9805808795266924e-07, "loss": 0.2379, "step": 1534 }, { "epoch": 0.4084619478445982, "grad_norm": 0.31282901763916016, "learning_rate": 1.980547406267637e-07, "loss": 0.2481, "step": 1535 }, { "epoch": 0.408728046833422, "grad_norm": 0.2671714425086975, "learning_rate": 1.9805139044673917e-07, "loss": 0.2458, "step": 1536 }, { "epoch": 0.40899414582224586, "grad_norm": 0.3123738765716553, "learning_rate": 1.980480374126933e-07, "loss": 0.2682, "step": 1537 }, { "epoch": 0.4092602448110697, "grad_norm": 0.26353153586387634, "learning_rate": 1.980446815247236e-07, "loss": 0.2235, "step": 1538 }, { "epoch": 0.40952634379989356, "grad_norm": 0.2850915491580963, "learning_rate": 1.9804132278292785e-07, "loss": 0.2571, "step": 1539 }, { "epoch": 0.4097924427887174, "grad_norm": 0.288982093334198, "learning_rate": 1.980379611874037e-07, "loss": 0.247, "step": 1540 }, { "epoch": 0.41005854177754125, "grad_norm": 0.2928670644760132, "learning_rate": 1.9803459673824906e-07, "loss": 0.2588, "step": 1541 }, { "epoch": 0.4103246407663651, "grad_norm": 0.2941928207874298, "learning_rate": 1.980312294355619e-07, "loss": 0.28, "step": 1542 }, { "epoch": 0.41059073975518895, "grad_norm": 0.270418643951416, "learning_rate": 1.9802785927944014e-07, "loss": 0.2316, "step": 1543 }, { "epoch": 0.4108568387440128, "grad_norm": 0.2891368269920349, "learning_rate": 1.9802448626998195e-07, "loss": 0.2466, "step": 1544 }, { "epoch": 0.4111229377328366, "grad_norm": 0.3158245086669922, "learning_rate": 1.9802111040728544e-07, "loss": 0.2592, "step": 1545 }, { "epoch": 0.41138903672166044, "grad_norm": 0.2949732542037964, "learning_rate": 1.9801773169144895e-07, "loss": 0.2411, "step": 1546 }, { "epoch": 0.4116551357104843, "grad_norm": 0.30284789204597473, "learning_rate": 1.980143501225708e-07, "loss": 0.2572, "step": 1547 }, { "epoch": 0.41192123469930814, "grad_norm": 0.29257702827453613, "learning_rate": 1.9801096570074937e-07, "loss": 0.2714, "step": 1548 }, { "epoch": 0.412187333688132, "grad_norm": 0.3066018223762512, "learning_rate": 1.9800757842608323e-07, "loss": 0.2726, "step": 1549 }, { "epoch": 0.41245343267695583, "grad_norm": 0.29737499356269836, "learning_rate": 1.98004188298671e-07, "loss": 0.2717, "step": 1550 }, { "epoch": 0.4127195316657797, "grad_norm": 0.2840505540370941, "learning_rate": 1.9800079531861126e-07, "loss": 0.2464, "step": 1551 }, { "epoch": 0.41298563065460353, "grad_norm": 0.3191086947917938, "learning_rate": 1.9799739948600285e-07, "loss": 0.2692, "step": 1552 }, { "epoch": 0.4132517296434274, "grad_norm": 0.3145340085029602, "learning_rate": 1.979940008009446e-07, "loss": 0.2511, "step": 1553 }, { "epoch": 0.41351782863225117, "grad_norm": 0.2957528233528137, "learning_rate": 1.979905992635354e-07, "loss": 0.24, "step": 1554 }, { "epoch": 0.413783927621075, "grad_norm": 0.2864665389060974, "learning_rate": 1.9798719487387428e-07, "loss": 0.2637, "step": 1555 }, { "epoch": 0.41405002660989887, "grad_norm": 0.27270445227622986, "learning_rate": 1.9798378763206033e-07, "loss": 0.2243, "step": 1556 }, { "epoch": 0.4143161255987227, "grad_norm": 0.2909444570541382, "learning_rate": 1.979803775381928e-07, "loss": 0.239, "step": 1557 }, { "epoch": 0.41458222458754657, "grad_norm": 0.3595362901687622, "learning_rate": 1.979769645923708e-07, "loss": 0.2519, "step": 1558 }, { "epoch": 0.4148483235763704, "grad_norm": 0.31045427918434143, "learning_rate": 1.979735487946938e-07, "loss": 0.2474, "step": 1559 }, { "epoch": 0.41511442256519426, "grad_norm": 0.29609423875808716, "learning_rate": 1.979701301452612e-07, "loss": 0.251, "step": 1560 }, { "epoch": 0.4153805215540181, "grad_norm": 0.2748085558414459, "learning_rate": 1.9796670864417244e-07, "loss": 0.2567, "step": 1561 }, { "epoch": 0.41564662054284196, "grad_norm": 0.335812509059906, "learning_rate": 1.9796328429152718e-07, "loss": 0.2865, "step": 1562 }, { "epoch": 0.41591271953166575, "grad_norm": 0.28640541434288025, "learning_rate": 1.9795985708742507e-07, "loss": 0.2624, "step": 1563 }, { "epoch": 0.4161788185204896, "grad_norm": 0.29298219084739685, "learning_rate": 1.979564270319659e-07, "loss": 0.2319, "step": 1564 }, { "epoch": 0.41644491750931345, "grad_norm": 0.30489295721054077, "learning_rate": 1.9795299412524944e-07, "loss": 0.2693, "step": 1565 }, { "epoch": 0.4167110164981373, "grad_norm": 0.31018492579460144, "learning_rate": 1.979495583673757e-07, "loss": 0.2458, "step": 1566 }, { "epoch": 0.41697711548696115, "grad_norm": 0.32799074053764343, "learning_rate": 1.9794611975844458e-07, "loss": 0.2636, "step": 1567 }, { "epoch": 0.417243214475785, "grad_norm": 0.2823637127876282, "learning_rate": 1.9794267829855625e-07, "loss": 0.2347, "step": 1568 }, { "epoch": 0.41750931346460884, "grad_norm": 0.29205840826034546, "learning_rate": 1.9793923398781086e-07, "loss": 0.2368, "step": 1569 }, { "epoch": 0.4177754124534327, "grad_norm": 0.28538817167282104, "learning_rate": 1.979357868263087e-07, "loss": 0.2554, "step": 1570 }, { "epoch": 0.41804151144225654, "grad_norm": 0.30077651143074036, "learning_rate": 1.9793233681415002e-07, "loss": 0.2648, "step": 1571 }, { "epoch": 0.41830761043108033, "grad_norm": 0.26505082845687866, "learning_rate": 1.9792888395143534e-07, "loss": 0.2409, "step": 1572 }, { "epoch": 0.4185737094199042, "grad_norm": 0.29462823271751404, "learning_rate": 1.979254282382651e-07, "loss": 0.2722, "step": 1573 }, { "epoch": 0.41883980840872803, "grad_norm": 0.29138579964637756, "learning_rate": 1.979219696747399e-07, "loss": 0.2483, "step": 1574 }, { "epoch": 0.4191059073975519, "grad_norm": 0.28531405329704285, "learning_rate": 1.9791850826096042e-07, "loss": 0.2648, "step": 1575 }, { "epoch": 0.41937200638637573, "grad_norm": 0.30864718556404114, "learning_rate": 1.979150439970274e-07, "loss": 0.2421, "step": 1576 }, { "epoch": 0.4196381053751996, "grad_norm": 0.3083479404449463, "learning_rate": 1.9791157688304167e-07, "loss": 0.2761, "step": 1577 }, { "epoch": 0.4199042043640234, "grad_norm": 0.26847365498542786, "learning_rate": 1.979081069191042e-07, "loss": 0.2488, "step": 1578 }, { "epoch": 0.4201703033528473, "grad_norm": 0.3205814063549042, "learning_rate": 1.9790463410531591e-07, "loss": 0.2765, "step": 1579 }, { "epoch": 0.4204364023416711, "grad_norm": 0.30605387687683105, "learning_rate": 1.9790115844177796e-07, "loss": 0.2784, "step": 1580 }, { "epoch": 0.42070250133049497, "grad_norm": 0.31813445687294006, "learning_rate": 1.978976799285915e-07, "loss": 0.2557, "step": 1581 }, { "epoch": 0.42096860031931876, "grad_norm": 0.3060058653354645, "learning_rate": 1.9789419856585773e-07, "loss": 0.2506, "step": 1582 }, { "epoch": 0.4212346993081426, "grad_norm": 0.29067564010620117, "learning_rate": 1.97890714353678e-07, "loss": 0.2746, "step": 1583 }, { "epoch": 0.42150079829696646, "grad_norm": 0.2869027554988861, "learning_rate": 1.9788722729215382e-07, "loss": 0.2632, "step": 1584 }, { "epoch": 0.4217668972857903, "grad_norm": 0.31960630416870117, "learning_rate": 1.9788373738138656e-07, "loss": 0.2709, "step": 1585 }, { "epoch": 0.42203299627461416, "grad_norm": 0.28283795714378357, "learning_rate": 1.9788024462147787e-07, "loss": 0.2413, "step": 1586 }, { "epoch": 0.422299095263438, "grad_norm": 0.32076236605644226, "learning_rate": 1.978767490125294e-07, "loss": 0.2725, "step": 1587 }, { "epoch": 0.42256519425226186, "grad_norm": 0.3307415246963501, "learning_rate": 1.978732505546429e-07, "loss": 0.2416, "step": 1588 }, { "epoch": 0.4228312932410857, "grad_norm": 0.3366162180900574, "learning_rate": 1.9786974924792022e-07, "loss": 0.2751, "step": 1589 }, { "epoch": 0.42309739222990955, "grad_norm": 0.28896698355674744, "learning_rate": 1.9786624509246322e-07, "loss": 0.2632, "step": 1590 }, { "epoch": 0.42336349121873335, "grad_norm": 0.28335219621658325, "learning_rate": 1.9786273808837394e-07, "loss": 0.2371, "step": 1591 }, { "epoch": 0.4236295902075572, "grad_norm": 0.43624815344810486, "learning_rate": 1.9785922823575447e-07, "loss": 0.2593, "step": 1592 }, { "epoch": 0.42389568919638104, "grad_norm": 0.2580292522907257, "learning_rate": 1.9785571553470696e-07, "loss": 0.2345, "step": 1593 }, { "epoch": 0.4241617881852049, "grad_norm": 0.30233898758888245, "learning_rate": 1.9785219998533365e-07, "loss": 0.249, "step": 1594 }, { "epoch": 0.42442788717402874, "grad_norm": 0.47782760858535767, "learning_rate": 1.9784868158773683e-07, "loss": 0.2789, "step": 1595 }, { "epoch": 0.4246939861628526, "grad_norm": 0.27839767932891846, "learning_rate": 1.9784516034201902e-07, "loss": 0.2347, "step": 1596 }, { "epoch": 0.42496008515167644, "grad_norm": 0.2622659206390381, "learning_rate": 1.978416362482826e-07, "loss": 0.2277, "step": 1597 }, { "epoch": 0.4252261841405003, "grad_norm": 0.288584440946579, "learning_rate": 1.978381093066302e-07, "loss": 0.2433, "step": 1598 }, { "epoch": 0.42549228312932413, "grad_norm": 0.3004129230976105, "learning_rate": 1.9783457951716444e-07, "loss": 0.2523, "step": 1599 }, { "epoch": 0.4257583821181479, "grad_norm": 0.28653544187545776, "learning_rate": 1.9783104687998814e-07, "loss": 0.2633, "step": 1600 }, { "epoch": 0.4260244811069718, "grad_norm": 0.2731463611125946, "learning_rate": 1.9782751139520407e-07, "loss": 0.2307, "step": 1601 }, { "epoch": 0.4262905800957956, "grad_norm": 0.2986607551574707, "learning_rate": 1.9782397306291517e-07, "loss": 0.259, "step": 1602 }, { "epoch": 0.42655667908461947, "grad_norm": 0.3064688444137573, "learning_rate": 1.9782043188322437e-07, "loss": 0.2606, "step": 1603 }, { "epoch": 0.4268227780734433, "grad_norm": 0.3027794361114502, "learning_rate": 1.978168878562348e-07, "loss": 0.2518, "step": 1604 }, { "epoch": 0.42708887706226717, "grad_norm": 0.27379265427589417, "learning_rate": 1.978133409820496e-07, "loss": 0.2377, "step": 1605 }, { "epoch": 0.427354976051091, "grad_norm": 0.28446459770202637, "learning_rate": 1.9780979126077204e-07, "loss": 0.2375, "step": 1606 }, { "epoch": 0.42762107503991487, "grad_norm": 0.2853374183177948, "learning_rate": 1.978062386925054e-07, "loss": 0.2509, "step": 1607 }, { "epoch": 0.4278871740287387, "grad_norm": 0.26743707060813904, "learning_rate": 1.9780268327735308e-07, "loss": 0.241, "step": 1608 }, { "epoch": 0.4281532730175625, "grad_norm": 0.3092416524887085, "learning_rate": 1.9779912501541863e-07, "loss": 0.267, "step": 1609 }, { "epoch": 0.42841937200638636, "grad_norm": 0.3080144226551056, "learning_rate": 1.9779556390680556e-07, "loss": 0.2616, "step": 1610 }, { "epoch": 0.4286854709952102, "grad_norm": 0.3133147954940796, "learning_rate": 1.9779199995161756e-07, "loss": 0.2611, "step": 1611 }, { "epoch": 0.42895156998403405, "grad_norm": 0.28530168533325195, "learning_rate": 1.9778843314995835e-07, "loss": 0.2516, "step": 1612 }, { "epoch": 0.4292176689728579, "grad_norm": 0.2881544530391693, "learning_rate": 1.9778486350193177e-07, "loss": 0.2755, "step": 1613 }, { "epoch": 0.42948376796168175, "grad_norm": 0.3066023290157318, "learning_rate": 1.9778129100764168e-07, "loss": 0.2436, "step": 1614 }, { "epoch": 0.4297498669505056, "grad_norm": 0.3096688687801361, "learning_rate": 1.9777771566719211e-07, "loss": 0.2649, "step": 1615 }, { "epoch": 0.43001596593932945, "grad_norm": 0.26375383138656616, "learning_rate": 1.9777413748068712e-07, "loss": 0.2366, "step": 1616 }, { "epoch": 0.4302820649281533, "grad_norm": 0.28687286376953125, "learning_rate": 1.9777055644823084e-07, "loss": 0.2289, "step": 1617 }, { "epoch": 0.4305481639169771, "grad_norm": 0.2683524191379547, "learning_rate": 1.9776697256992753e-07, "loss": 0.227, "step": 1618 }, { "epoch": 0.43081426290580094, "grad_norm": 0.2827534079551697, "learning_rate": 1.977633858458815e-07, "loss": 0.2717, "step": 1619 }, { "epoch": 0.4310803618946248, "grad_norm": 0.2860388457775116, "learning_rate": 1.9775979627619716e-07, "loss": 0.2509, "step": 1620 }, { "epoch": 0.43134646088344863, "grad_norm": 0.2873339056968689, "learning_rate": 1.9775620386097898e-07, "loss": 0.2529, "step": 1621 }, { "epoch": 0.4316125598722725, "grad_norm": 0.3117932081222534, "learning_rate": 1.977526086003315e-07, "loss": 0.2429, "step": 1622 }, { "epoch": 0.43187865886109633, "grad_norm": 0.28225287795066833, "learning_rate": 1.9774901049435941e-07, "loss": 0.2598, "step": 1623 }, { "epoch": 0.4321447578499202, "grad_norm": 0.2803475856781006, "learning_rate": 1.9774540954316742e-07, "loss": 0.241, "step": 1624 }, { "epoch": 0.43241085683874403, "grad_norm": 0.29940348863601685, "learning_rate": 1.9774180574686037e-07, "loss": 0.2628, "step": 1625 }, { "epoch": 0.4326769558275679, "grad_norm": 0.2826913297176361, "learning_rate": 1.977381991055431e-07, "loss": 0.2541, "step": 1626 }, { "epoch": 0.43294305481639167, "grad_norm": 0.37016087770462036, "learning_rate": 1.9773458961932067e-07, "loss": 0.274, "step": 1627 }, { "epoch": 0.4332091538052155, "grad_norm": 0.31109920144081116, "learning_rate": 1.977309772882981e-07, "loss": 0.2442, "step": 1628 }, { "epoch": 0.43347525279403937, "grad_norm": 0.2851846516132355, "learning_rate": 1.9772736211258052e-07, "loss": 0.2683, "step": 1629 }, { "epoch": 0.4337413517828632, "grad_norm": 0.28317901492118835, "learning_rate": 1.977237440922732e-07, "loss": 0.2622, "step": 1630 }, { "epoch": 0.43400745077168706, "grad_norm": 0.3026304244995117, "learning_rate": 1.977201232274814e-07, "loss": 0.2606, "step": 1631 }, { "epoch": 0.4342735497605109, "grad_norm": 0.27774935960769653, "learning_rate": 1.9771649951831056e-07, "loss": 0.23, "step": 1632 }, { "epoch": 0.43453964874933476, "grad_norm": 0.2773955464363098, "learning_rate": 1.9771287296486614e-07, "loss": 0.2446, "step": 1633 }, { "epoch": 0.4348057477381586, "grad_norm": 0.305808424949646, "learning_rate": 1.9770924356725367e-07, "loss": 0.275, "step": 1634 }, { "epoch": 0.43507184672698246, "grad_norm": 0.2836032807826996, "learning_rate": 1.9770561132557885e-07, "loss": 0.2509, "step": 1635 }, { "epoch": 0.4353379457158063, "grad_norm": 0.3512755334377289, "learning_rate": 1.9770197623994736e-07, "loss": 0.2425, "step": 1636 }, { "epoch": 0.4356040447046301, "grad_norm": 0.2884102761745453, "learning_rate": 1.9769833831046498e-07, "loss": 0.2667, "step": 1637 }, { "epoch": 0.43587014369345395, "grad_norm": 0.306318461894989, "learning_rate": 1.976946975372377e-07, "loss": 0.2764, "step": 1638 }, { "epoch": 0.4361362426822778, "grad_norm": 0.29530954360961914, "learning_rate": 1.9769105392037142e-07, "loss": 0.263, "step": 1639 }, { "epoch": 0.43640234167110165, "grad_norm": 0.30547067523002625, "learning_rate": 1.976874074599722e-07, "loss": 0.2598, "step": 1640 }, { "epoch": 0.4366684406599255, "grad_norm": 0.26924023032188416, "learning_rate": 1.9768375815614622e-07, "loss": 0.2531, "step": 1641 }, { "epoch": 0.43693453964874934, "grad_norm": 0.3047896921634674, "learning_rate": 1.9768010600899964e-07, "loss": 0.2498, "step": 1642 }, { "epoch": 0.4372006386375732, "grad_norm": 0.2846266031265259, "learning_rate": 1.9767645101863883e-07, "loss": 0.2586, "step": 1643 }, { "epoch": 0.43746673762639704, "grad_norm": 0.2642972469329834, "learning_rate": 1.9767279318517013e-07, "loss": 0.2442, "step": 1644 }, { "epoch": 0.4377328366152209, "grad_norm": 0.3099304437637329, "learning_rate": 1.976691325087e-07, "loss": 0.2673, "step": 1645 }, { "epoch": 0.4379989356040447, "grad_norm": 0.312857449054718, "learning_rate": 1.9766546898933506e-07, "loss": 0.2456, "step": 1646 }, { "epoch": 0.43826503459286853, "grad_norm": 0.2898359000682831, "learning_rate": 1.976618026271819e-07, "loss": 0.2604, "step": 1647 }, { "epoch": 0.4385311335816924, "grad_norm": 0.318547785282135, "learning_rate": 1.9765813342234724e-07, "loss": 0.2617, "step": 1648 }, { "epoch": 0.4387972325705162, "grad_norm": 0.2792694568634033, "learning_rate": 1.9765446137493785e-07, "loss": 0.2563, "step": 1649 }, { "epoch": 0.4390633315593401, "grad_norm": 0.3281499147415161, "learning_rate": 1.976507864850607e-07, "loss": 0.2711, "step": 1650 }, { "epoch": 0.4393294305481639, "grad_norm": 0.292690247297287, "learning_rate": 1.9764710875282265e-07, "loss": 0.2606, "step": 1651 }, { "epoch": 0.4395955295369878, "grad_norm": 0.29226529598236084, "learning_rate": 1.976434281783308e-07, "loss": 0.2647, "step": 1652 }, { "epoch": 0.4398616285258116, "grad_norm": 0.2662850618362427, "learning_rate": 1.9763974476169233e-07, "loss": 0.227, "step": 1653 }, { "epoch": 0.44012772751463547, "grad_norm": 0.32746651768684387, "learning_rate": 1.976360585030144e-07, "loss": 0.2298, "step": 1654 }, { "epoch": 0.44039382650345926, "grad_norm": 0.27729538083076477, "learning_rate": 1.9763236940240428e-07, "loss": 0.251, "step": 1655 }, { "epoch": 0.4406599254922831, "grad_norm": 0.2850281596183777, "learning_rate": 1.976286774599694e-07, "loss": 0.258, "step": 1656 }, { "epoch": 0.44092602448110696, "grad_norm": 0.2832871675491333, "learning_rate": 1.976249826758172e-07, "loss": 0.2446, "step": 1657 }, { "epoch": 0.4411921234699308, "grad_norm": 0.26245221495628357, "learning_rate": 1.9762128505005527e-07, "loss": 0.2553, "step": 1658 }, { "epoch": 0.44145822245875466, "grad_norm": 0.29561683535575867, "learning_rate": 1.9761758458279122e-07, "loss": 0.2588, "step": 1659 }, { "epoch": 0.4417243214475785, "grad_norm": 0.29846808314323425, "learning_rate": 1.976138812741327e-07, "loss": 0.2697, "step": 1660 }, { "epoch": 0.44199042043640235, "grad_norm": 0.29665833711624146, "learning_rate": 1.9761017512418756e-07, "loss": 0.2374, "step": 1661 }, { "epoch": 0.4422565194252262, "grad_norm": 0.27814772725105286, "learning_rate": 1.9760646613306366e-07, "loss": 0.2497, "step": 1662 }, { "epoch": 0.44252261841405005, "grad_norm": 0.2792379856109619, "learning_rate": 1.9760275430086897e-07, "loss": 0.2523, "step": 1663 }, { "epoch": 0.44278871740287384, "grad_norm": 0.27199074625968933, "learning_rate": 1.9759903962771153e-07, "loss": 0.2352, "step": 1664 }, { "epoch": 0.4430548163916977, "grad_norm": 0.3007422685623169, "learning_rate": 1.9759532211369945e-07, "loss": 0.2306, "step": 1665 }, { "epoch": 0.44332091538052154, "grad_norm": 0.2958345413208008, "learning_rate": 1.9759160175894096e-07, "loss": 0.2285, "step": 1666 }, { "epoch": 0.4435870143693454, "grad_norm": 0.29437294602394104, "learning_rate": 1.975878785635443e-07, "loss": 0.2602, "step": 1667 }, { "epoch": 0.44385311335816924, "grad_norm": 0.2783452570438385, "learning_rate": 1.9758415252761793e-07, "loss": 0.2501, "step": 1668 }, { "epoch": 0.4441192123469931, "grad_norm": 0.29376864433288574, "learning_rate": 1.9758042365127024e-07, "loss": 0.2596, "step": 1669 }, { "epoch": 0.44438531133581693, "grad_norm": 0.41466224193573, "learning_rate": 1.9757669193460976e-07, "loss": 0.264, "step": 1670 }, { "epoch": 0.4446514103246408, "grad_norm": 0.3018956482410431, "learning_rate": 1.9757295737774516e-07, "loss": 0.2497, "step": 1671 }, { "epoch": 0.44491750931346463, "grad_norm": 0.2913684546947479, "learning_rate": 1.9756921998078508e-07, "loss": 0.237, "step": 1672 }, { "epoch": 0.4451836083022884, "grad_norm": 0.31225475668907166, "learning_rate": 1.9756547974383837e-07, "loss": 0.2822, "step": 1673 }, { "epoch": 0.4454497072911123, "grad_norm": 0.30463647842407227, "learning_rate": 1.9756173666701388e-07, "loss": 0.2606, "step": 1674 }, { "epoch": 0.4457158062799361, "grad_norm": 0.2832431197166443, "learning_rate": 1.9755799075042052e-07, "loss": 0.2507, "step": 1675 }, { "epoch": 0.44598190526875997, "grad_norm": 0.27695432305336, "learning_rate": 1.9755424199416736e-07, "loss": 0.2415, "step": 1676 }, { "epoch": 0.4462480042575838, "grad_norm": 0.3063591718673706, "learning_rate": 1.9755049039836354e-07, "loss": 0.2483, "step": 1677 }, { "epoch": 0.44651410324640767, "grad_norm": 0.2939196228981018, "learning_rate": 1.975467359631182e-07, "loss": 0.2652, "step": 1678 }, { "epoch": 0.4467802022352315, "grad_norm": 0.2705158293247223, "learning_rate": 1.975429786885407e-07, "loss": 0.2487, "step": 1679 }, { "epoch": 0.44704630122405536, "grad_norm": 0.30481934547424316, "learning_rate": 1.9753921857474034e-07, "loss": 0.3009, "step": 1680 }, { "epoch": 0.4473124002128792, "grad_norm": 0.29155200719833374, "learning_rate": 1.9753545562182656e-07, "loss": 0.2431, "step": 1681 }, { "epoch": 0.447578499201703, "grad_norm": 0.31933432817459106, "learning_rate": 1.9753168982990894e-07, "loss": 0.2554, "step": 1682 }, { "epoch": 0.44784459819052685, "grad_norm": 0.3134215176105499, "learning_rate": 1.9752792119909705e-07, "loss": 0.245, "step": 1683 }, { "epoch": 0.4481106971793507, "grad_norm": 0.3061843514442444, "learning_rate": 1.975241497295006e-07, "loss": 0.2818, "step": 1684 }, { "epoch": 0.44837679616817455, "grad_norm": 0.3416574001312256, "learning_rate": 1.975203754212294e-07, "loss": 0.2639, "step": 1685 }, { "epoch": 0.4486428951569984, "grad_norm": 0.2609492540359497, "learning_rate": 1.9751659827439323e-07, "loss": 0.2263, "step": 1686 }, { "epoch": 0.44890899414582225, "grad_norm": 0.27815642952919006, "learning_rate": 1.9751281828910214e-07, "loss": 0.2527, "step": 1687 }, { "epoch": 0.4491750931346461, "grad_norm": 0.3029038608074188, "learning_rate": 1.975090354654661e-07, "loss": 0.2427, "step": 1688 }, { "epoch": 0.44944119212346995, "grad_norm": 0.2586315870285034, "learning_rate": 1.9750524980359516e-07, "loss": 0.2442, "step": 1689 }, { "epoch": 0.4497072911122938, "grad_norm": 0.30097827315330505, "learning_rate": 1.975014613035996e-07, "loss": 0.2429, "step": 1690 }, { "epoch": 0.44997339010111764, "grad_norm": 0.3047308325767517, "learning_rate": 1.9749766996558966e-07, "loss": 0.2644, "step": 1691 }, { "epoch": 0.45023948908994144, "grad_norm": 0.26410746574401855, "learning_rate": 1.9749387578967572e-07, "loss": 0.2375, "step": 1692 }, { "epoch": 0.4505055880787653, "grad_norm": 0.3083888292312622, "learning_rate": 1.9749007877596818e-07, "loss": 0.2393, "step": 1693 }, { "epoch": 0.45077168706758913, "grad_norm": 0.28903982043266296, "learning_rate": 1.9748627892457757e-07, "loss": 0.2534, "step": 1694 }, { "epoch": 0.451037786056413, "grad_norm": 0.29290130734443665, "learning_rate": 1.974824762356145e-07, "loss": 0.2627, "step": 1695 }, { "epoch": 0.45130388504523683, "grad_norm": 0.2805551588535309, "learning_rate": 1.9747867070918965e-07, "loss": 0.2437, "step": 1696 }, { "epoch": 0.4515699840340607, "grad_norm": 0.2761564552783966, "learning_rate": 1.974748623454138e-07, "loss": 0.2428, "step": 1697 }, { "epoch": 0.4518360830228845, "grad_norm": 0.28916415572166443, "learning_rate": 1.9747105114439782e-07, "loss": 0.2589, "step": 1698 }, { "epoch": 0.4521021820117084, "grad_norm": 0.2847821116447449, "learning_rate": 1.974672371062526e-07, "loss": 0.2383, "step": 1699 }, { "epoch": 0.4523682810005322, "grad_norm": 0.30073684453964233, "learning_rate": 1.9746342023108918e-07, "loss": 0.2669, "step": 1700 }, { "epoch": 0.452634379989356, "grad_norm": 0.28180089592933655, "learning_rate": 1.9745960051901867e-07, "loss": 0.2655, "step": 1701 }, { "epoch": 0.45290047897817987, "grad_norm": 0.2790069580078125, "learning_rate": 1.9745577797015223e-07, "loss": 0.2262, "step": 1702 }, { "epoch": 0.4531665779670037, "grad_norm": 0.2769038677215576, "learning_rate": 1.9745195258460115e-07, "loss": 0.2461, "step": 1703 }, { "epoch": 0.45343267695582756, "grad_norm": 0.27939754724502563, "learning_rate": 1.9744812436247672e-07, "loss": 0.2574, "step": 1704 }, { "epoch": 0.4536987759446514, "grad_norm": 0.314997136592865, "learning_rate": 1.9744429330389047e-07, "loss": 0.2412, "step": 1705 }, { "epoch": 0.45396487493347526, "grad_norm": 0.28845494985580444, "learning_rate": 1.9744045940895383e-07, "loss": 0.2627, "step": 1706 }, { "epoch": 0.4542309739222991, "grad_norm": 0.28560832142829895, "learning_rate": 1.974366226777784e-07, "loss": 0.2438, "step": 1707 }, { "epoch": 0.45449707291112296, "grad_norm": 0.3129023611545563, "learning_rate": 1.9743278311047589e-07, "loss": 0.2737, "step": 1708 }, { "epoch": 0.4547631718999468, "grad_norm": 0.26389390230178833, "learning_rate": 1.9742894070715806e-07, "loss": 0.2522, "step": 1709 }, { "epoch": 0.4550292708887706, "grad_norm": 0.2818462550640106, "learning_rate": 1.9742509546793672e-07, "loss": 0.2537, "step": 1710 }, { "epoch": 0.45529536987759445, "grad_norm": 0.28399917483329773, "learning_rate": 1.9742124739292381e-07, "loss": 0.2483, "step": 1711 }, { "epoch": 0.4555614688664183, "grad_norm": 0.304943323135376, "learning_rate": 1.9741739648223135e-07, "loss": 0.2763, "step": 1712 }, { "epoch": 0.45582756785524214, "grad_norm": 0.28741568326950073, "learning_rate": 1.9741354273597142e-07, "loss": 0.2339, "step": 1713 }, { "epoch": 0.456093666844066, "grad_norm": 0.2628532648086548, "learning_rate": 1.974096861542562e-07, "loss": 0.2472, "step": 1714 }, { "epoch": 0.45635976583288984, "grad_norm": 0.29366740584373474, "learning_rate": 1.974058267371979e-07, "loss": 0.2634, "step": 1715 }, { "epoch": 0.4566258648217137, "grad_norm": 0.2877090871334076, "learning_rate": 1.9740196448490893e-07, "loss": 0.2668, "step": 1716 }, { "epoch": 0.45689196381053754, "grad_norm": 0.27738720178604126, "learning_rate": 1.9739809939750168e-07, "loss": 0.2477, "step": 1717 }, { "epoch": 0.4571580627993614, "grad_norm": 0.27032914757728577, "learning_rate": 1.9739423147508862e-07, "loss": 0.2439, "step": 1718 }, { "epoch": 0.4574241617881852, "grad_norm": 0.29093214869499207, "learning_rate": 1.973903607177824e-07, "loss": 0.2656, "step": 1719 }, { "epoch": 0.45769026077700903, "grad_norm": 0.2962317168712616, "learning_rate": 1.9738648712569562e-07, "loss": 0.2469, "step": 1720 }, { "epoch": 0.4579563597658329, "grad_norm": 0.30758804082870483, "learning_rate": 1.9738261069894107e-07, "loss": 0.2805, "step": 1721 }, { "epoch": 0.4582224587546567, "grad_norm": 0.2963006794452667, "learning_rate": 1.9737873143763158e-07, "loss": 0.2425, "step": 1722 }, { "epoch": 0.4584885577434806, "grad_norm": 0.2955700755119324, "learning_rate": 1.9737484934188003e-07, "loss": 0.2662, "step": 1723 }, { "epoch": 0.4587546567323044, "grad_norm": 0.3273976743221283, "learning_rate": 1.9737096441179947e-07, "loss": 0.2268, "step": 1724 }, { "epoch": 0.45902075572112827, "grad_norm": 0.3002691864967346, "learning_rate": 1.97367076647503e-07, "loss": 0.2778, "step": 1725 }, { "epoch": 0.4592868547099521, "grad_norm": 0.2869487702846527, "learning_rate": 1.9736318604910367e-07, "loss": 0.2218, "step": 1726 }, { "epoch": 0.45955295369877597, "grad_norm": 0.2967793941497803, "learning_rate": 1.9735929261671483e-07, "loss": 0.2417, "step": 1727 }, { "epoch": 0.45981905268759976, "grad_norm": 0.28839176893234253, "learning_rate": 1.9735539635044976e-07, "loss": 0.252, "step": 1728 }, { "epoch": 0.4600851516764236, "grad_norm": 0.27941399812698364, "learning_rate": 1.9735149725042188e-07, "loss": 0.2362, "step": 1729 }, { "epoch": 0.46035125066524746, "grad_norm": 0.2791213393211365, "learning_rate": 1.973475953167447e-07, "loss": 0.2502, "step": 1730 }, { "epoch": 0.4606173496540713, "grad_norm": 0.28858673572540283, "learning_rate": 1.9734369054953177e-07, "loss": 0.2409, "step": 1731 }, { "epoch": 0.46088344864289515, "grad_norm": 0.2879990339279175, "learning_rate": 1.9733978294889678e-07, "loss": 0.2492, "step": 1732 }, { "epoch": 0.461149547631719, "grad_norm": 0.270477831363678, "learning_rate": 1.9733587251495342e-07, "loss": 0.2525, "step": 1733 }, { "epoch": 0.46141564662054285, "grad_norm": 0.2727735638618469, "learning_rate": 1.9733195924781555e-07, "loss": 0.2245, "step": 1734 }, { "epoch": 0.4616817456093667, "grad_norm": 0.2863978147506714, "learning_rate": 1.9732804314759708e-07, "loss": 0.2495, "step": 1735 }, { "epoch": 0.46194784459819055, "grad_norm": 0.31570926308631897, "learning_rate": 1.9732412421441196e-07, "loss": 0.2631, "step": 1736 }, { "epoch": 0.46221394358701434, "grad_norm": 0.2867180109024048, "learning_rate": 1.973202024483743e-07, "loss": 0.2362, "step": 1737 }, { "epoch": 0.4624800425758382, "grad_norm": 0.2930296063423157, "learning_rate": 1.9731627784959823e-07, "loss": 0.2469, "step": 1738 }, { "epoch": 0.46274614156466204, "grad_norm": 0.3314518630504608, "learning_rate": 1.97312350418198e-07, "loss": 0.2645, "step": 1739 }, { "epoch": 0.4630122405534859, "grad_norm": 0.2991669178009033, "learning_rate": 1.9730842015428792e-07, "loss": 0.272, "step": 1740 }, { "epoch": 0.46327833954230974, "grad_norm": 0.28593748807907104, "learning_rate": 1.9730448705798237e-07, "loss": 0.2545, "step": 1741 }, { "epoch": 0.4635444385311336, "grad_norm": 0.2769896388053894, "learning_rate": 1.9730055112939586e-07, "loss": 0.2523, "step": 1742 }, { "epoch": 0.46381053751995743, "grad_norm": 0.2971626818180084, "learning_rate": 1.9729661236864294e-07, "loss": 0.2755, "step": 1743 }, { "epoch": 0.4640766365087813, "grad_norm": 0.2923210859298706, "learning_rate": 1.9729267077583828e-07, "loss": 0.2604, "step": 1744 }, { "epoch": 0.46434273549760513, "grad_norm": 0.3143354058265686, "learning_rate": 1.9728872635109659e-07, "loss": 0.2789, "step": 1745 }, { "epoch": 0.464608834486429, "grad_norm": 0.30051514506340027, "learning_rate": 1.9728477909453266e-07, "loss": 0.2503, "step": 1746 }, { "epoch": 0.46487493347525277, "grad_norm": 0.26156482100486755, "learning_rate": 1.9728082900626142e-07, "loss": 0.2347, "step": 1747 }, { "epoch": 0.4651410324640766, "grad_norm": 0.33429500460624695, "learning_rate": 1.9727687608639784e-07, "loss": 0.2532, "step": 1748 }, { "epoch": 0.46540713145290047, "grad_norm": 0.28683704137802124, "learning_rate": 1.9727292033505698e-07, "loss": 0.2358, "step": 1749 }, { "epoch": 0.4656732304417243, "grad_norm": 0.3008774220943451, "learning_rate": 1.9726896175235397e-07, "loss": 0.2566, "step": 1750 }, { "epoch": 0.46593932943054817, "grad_norm": 0.28856706619262695, "learning_rate": 1.9726500033840402e-07, "loss": 0.2451, "step": 1751 }, { "epoch": 0.466205428419372, "grad_norm": 0.2874751687049866, "learning_rate": 1.9726103609332244e-07, "loss": 0.2394, "step": 1752 }, { "epoch": 0.46647152740819586, "grad_norm": 0.2827128767967224, "learning_rate": 1.972570690172247e-07, "loss": 0.2474, "step": 1753 }, { "epoch": 0.4667376263970197, "grad_norm": 0.30700913071632385, "learning_rate": 1.9725309911022616e-07, "loss": 0.2587, "step": 1754 }, { "epoch": 0.46700372538584356, "grad_norm": 0.2858197093009949, "learning_rate": 1.9724912637244245e-07, "loss": 0.2492, "step": 1755 }, { "epoch": 0.46726982437466735, "grad_norm": 0.39199110865592957, "learning_rate": 1.9724515080398915e-07, "loss": 0.2723, "step": 1756 }, { "epoch": 0.4675359233634912, "grad_norm": 0.30383244156837463, "learning_rate": 1.9724117240498202e-07, "loss": 0.2662, "step": 1757 }, { "epoch": 0.46780202235231505, "grad_norm": 0.2836339473724365, "learning_rate": 1.9723719117553684e-07, "loss": 0.2617, "step": 1758 }, { "epoch": 0.4680681213411389, "grad_norm": 0.27625125646591187, "learning_rate": 1.9723320711576949e-07, "loss": 0.2353, "step": 1759 }, { "epoch": 0.46833422032996275, "grad_norm": 0.2932957708835602, "learning_rate": 1.9722922022579596e-07, "loss": 0.2556, "step": 1760 }, { "epoch": 0.4686003193187866, "grad_norm": 0.2627798318862915, "learning_rate": 1.9722523050573225e-07, "loss": 0.2569, "step": 1761 }, { "epoch": 0.46886641830761044, "grad_norm": 0.2647307813167572, "learning_rate": 1.9722123795569457e-07, "loss": 0.2334, "step": 1762 }, { "epoch": 0.4691325172964343, "grad_norm": 0.30235737562179565, "learning_rate": 1.9721724257579906e-07, "loss": 0.2679, "step": 1763 }, { "epoch": 0.46939861628525814, "grad_norm": 0.299212247133255, "learning_rate": 1.9721324436616204e-07, "loss": 0.2473, "step": 1764 }, { "epoch": 0.46966471527408193, "grad_norm": 0.39402157068252563, "learning_rate": 1.9720924332689988e-07, "loss": 0.236, "step": 1765 }, { "epoch": 0.4699308142629058, "grad_norm": 0.3137006163597107, "learning_rate": 1.972052394581291e-07, "loss": 0.2597, "step": 1766 }, { "epoch": 0.47019691325172963, "grad_norm": 0.3277299404144287, "learning_rate": 1.9720123275996613e-07, "loss": 0.2621, "step": 1767 }, { "epoch": 0.4704630122405535, "grad_norm": 0.31674325466156006, "learning_rate": 1.9719722323252767e-07, "loss": 0.2583, "step": 1768 }, { "epoch": 0.47072911122937733, "grad_norm": 0.30360615253448486, "learning_rate": 1.9719321087593038e-07, "loss": 0.244, "step": 1769 }, { "epoch": 0.4709952102182012, "grad_norm": 0.29012417793273926, "learning_rate": 1.9718919569029114e-07, "loss": 0.2525, "step": 1770 }, { "epoch": 0.471261309207025, "grad_norm": 0.2631171941757202, "learning_rate": 1.9718517767572674e-07, "loss": 0.2309, "step": 1771 }, { "epoch": 0.4715274081958489, "grad_norm": 0.2679053246974945, "learning_rate": 1.9718115683235416e-07, "loss": 0.2467, "step": 1772 }, { "epoch": 0.4717935071846727, "grad_norm": 0.2960651218891144, "learning_rate": 1.9717713316029042e-07, "loss": 0.2486, "step": 1773 }, { "epoch": 0.4720596061734965, "grad_norm": 0.295584499835968, "learning_rate": 1.9717310665965266e-07, "loss": 0.249, "step": 1774 }, { "epoch": 0.47232570516232036, "grad_norm": 0.28251415491104126, "learning_rate": 1.971690773305581e-07, "loss": 0.2381, "step": 1775 }, { "epoch": 0.4725918041511442, "grad_norm": 0.3267378807067871, "learning_rate": 1.9716504517312396e-07, "loss": 0.2625, "step": 1776 }, { "epoch": 0.47285790313996806, "grad_norm": 0.31738996505737305, "learning_rate": 1.9716101018746765e-07, "loss": 0.2751, "step": 1777 }, { "epoch": 0.4731240021287919, "grad_norm": 0.30158546566963196, "learning_rate": 1.971569723737066e-07, "loss": 0.2471, "step": 1778 }, { "epoch": 0.47339010111761576, "grad_norm": 0.4181515872478485, "learning_rate": 1.9715293173195837e-07, "loss": 0.2525, "step": 1779 }, { "epoch": 0.4736562001064396, "grad_norm": 0.29933926463127136, "learning_rate": 1.9714888826234053e-07, "loss": 0.2699, "step": 1780 }, { "epoch": 0.47392229909526346, "grad_norm": 0.29025205969810486, "learning_rate": 1.9714484196497084e-07, "loss": 0.2492, "step": 1781 }, { "epoch": 0.4741883980840873, "grad_norm": 0.2991805076599121, "learning_rate": 1.97140792839967e-07, "loss": 0.2504, "step": 1782 }, { "epoch": 0.4744544970729111, "grad_norm": 0.31432756781578064, "learning_rate": 1.9713674088744688e-07, "loss": 0.2687, "step": 1783 }, { "epoch": 0.47472059606173495, "grad_norm": 0.2782236337661743, "learning_rate": 1.9713268610752847e-07, "loss": 0.2603, "step": 1784 }, { "epoch": 0.4749866950505588, "grad_norm": 0.26010769605636597, "learning_rate": 1.9712862850032978e-07, "loss": 0.2368, "step": 1785 }, { "epoch": 0.47525279403938264, "grad_norm": 0.29912129044532776, "learning_rate": 1.9712456806596892e-07, "loss": 0.2412, "step": 1786 }, { "epoch": 0.4755188930282065, "grad_norm": 0.2865196466445923, "learning_rate": 1.9712050480456405e-07, "loss": 0.2531, "step": 1787 }, { "epoch": 0.47578499201703034, "grad_norm": 0.2905884385108948, "learning_rate": 1.9711643871623343e-07, "loss": 0.2212, "step": 1788 }, { "epoch": 0.4760510910058542, "grad_norm": 0.28577885031700134, "learning_rate": 1.9711236980109548e-07, "loss": 0.2521, "step": 1789 }, { "epoch": 0.47631718999467804, "grad_norm": 0.2784094512462616, "learning_rate": 1.9710829805926855e-07, "loss": 0.2652, "step": 1790 }, { "epoch": 0.4765832889835019, "grad_norm": 0.28579947352409363, "learning_rate": 1.9710422349087124e-07, "loss": 0.2537, "step": 1791 }, { "epoch": 0.4768493879723257, "grad_norm": 0.2898848056793213, "learning_rate": 1.9710014609602208e-07, "loss": 0.2535, "step": 1792 }, { "epoch": 0.4771154869611495, "grad_norm": 0.3285640478134155, "learning_rate": 1.9709606587483983e-07, "loss": 0.2529, "step": 1793 }, { "epoch": 0.4773815859499734, "grad_norm": 0.3102917969226837, "learning_rate": 1.9709198282744318e-07, "loss": 0.2675, "step": 1794 }, { "epoch": 0.4776476849387972, "grad_norm": 0.3072963058948517, "learning_rate": 1.97087896953951e-07, "loss": 0.2736, "step": 1795 }, { "epoch": 0.47791378392762107, "grad_norm": 0.3038388192653656, "learning_rate": 1.9708380825448225e-07, "loss": 0.2468, "step": 1796 }, { "epoch": 0.4781798829164449, "grad_norm": 0.29930514097213745, "learning_rate": 1.970797167291559e-07, "loss": 0.2526, "step": 1797 }, { "epoch": 0.47844598190526877, "grad_norm": 0.26621630787849426, "learning_rate": 1.9707562237809107e-07, "loss": 0.2359, "step": 1798 }, { "epoch": 0.4787120808940926, "grad_norm": 0.3031366467475891, "learning_rate": 1.9707152520140694e-07, "loss": 0.2594, "step": 1799 }, { "epoch": 0.47897817988291647, "grad_norm": 0.31284210085868835, "learning_rate": 1.9706742519922274e-07, "loss": 0.2578, "step": 1800 }, { "epoch": 0.4792442788717403, "grad_norm": 0.32212674617767334, "learning_rate": 1.9706332237165778e-07, "loss": 0.2614, "step": 1801 }, { "epoch": 0.4795103778605641, "grad_norm": 0.28349530696868896, "learning_rate": 1.9705921671883158e-07, "loss": 0.232, "step": 1802 }, { "epoch": 0.47977647684938796, "grad_norm": 0.2797132730484009, "learning_rate": 1.970551082408636e-07, "loss": 0.2492, "step": 1803 }, { "epoch": 0.4800425758382118, "grad_norm": 0.29298070073127747, "learning_rate": 1.970509969378734e-07, "loss": 0.2293, "step": 1804 }, { "epoch": 0.48030867482703565, "grad_norm": 0.30740606784820557, "learning_rate": 1.9704688280998068e-07, "loss": 0.2589, "step": 1805 }, { "epoch": 0.4805747738158595, "grad_norm": 0.2933560013771057, "learning_rate": 1.9704276585730517e-07, "loss": 0.254, "step": 1806 }, { "epoch": 0.48084087280468335, "grad_norm": 0.27843210101127625, "learning_rate": 1.9703864607996672e-07, "loss": 0.2481, "step": 1807 }, { "epoch": 0.4811069717935072, "grad_norm": 0.27680402994155884, "learning_rate": 1.9703452347808525e-07, "loss": 0.25, "step": 1808 }, { "epoch": 0.48137307078233105, "grad_norm": 0.27841416001319885, "learning_rate": 1.970303980517807e-07, "loss": 0.2483, "step": 1809 }, { "epoch": 0.4816391697711549, "grad_norm": 0.26605910062789917, "learning_rate": 1.9702626980117326e-07, "loss": 0.235, "step": 1810 }, { "epoch": 0.4819052687599787, "grad_norm": 0.27950161695480347, "learning_rate": 1.9702213872638303e-07, "loss": 0.2529, "step": 1811 }, { "epoch": 0.48217136774880254, "grad_norm": 0.28804805874824524, "learning_rate": 1.9701800482753024e-07, "loss": 0.2345, "step": 1812 }, { "epoch": 0.4824374667376264, "grad_norm": 0.29097801446914673, "learning_rate": 1.9701386810473524e-07, "loss": 0.2384, "step": 1813 }, { "epoch": 0.48270356572645023, "grad_norm": 0.2968047559261322, "learning_rate": 1.9700972855811843e-07, "loss": 0.2615, "step": 1814 }, { "epoch": 0.4829696647152741, "grad_norm": 0.2884863317012787, "learning_rate": 1.970055861878003e-07, "loss": 0.2496, "step": 1815 }, { "epoch": 0.48323576370409793, "grad_norm": 0.3174475133419037, "learning_rate": 1.970014409939014e-07, "loss": 0.2679, "step": 1816 }, { "epoch": 0.4835018626929218, "grad_norm": 0.2946696877479553, "learning_rate": 1.9699729297654249e-07, "loss": 0.2333, "step": 1817 }, { "epoch": 0.48376796168174563, "grad_norm": 0.26258060336112976, "learning_rate": 1.9699314213584416e-07, "loss": 0.2394, "step": 1818 }, { "epoch": 0.4840340606705695, "grad_norm": 0.3125440180301666, "learning_rate": 1.9698898847192735e-07, "loss": 0.2538, "step": 1819 }, { "epoch": 0.48430015965939327, "grad_norm": 0.3124016225337982, "learning_rate": 1.9698483198491287e-07, "loss": 0.2793, "step": 1820 }, { "epoch": 0.4845662586482171, "grad_norm": 0.27908286452293396, "learning_rate": 1.9698067267492177e-07, "loss": 0.2318, "step": 1821 }, { "epoch": 0.48483235763704097, "grad_norm": 0.2577545940876007, "learning_rate": 1.9697651054207512e-07, "loss": 0.2406, "step": 1822 }, { "epoch": 0.4850984566258648, "grad_norm": 0.29743507504463196, "learning_rate": 1.9697234558649404e-07, "loss": 0.2369, "step": 1823 }, { "epoch": 0.48536455561468866, "grad_norm": 0.30946969985961914, "learning_rate": 1.9696817780829975e-07, "loss": 0.266, "step": 1824 }, { "epoch": 0.4856306546035125, "grad_norm": 0.28544703125953674, "learning_rate": 1.9696400720761356e-07, "loss": 0.2495, "step": 1825 }, { "epoch": 0.48589675359233636, "grad_norm": 0.29841431975364685, "learning_rate": 1.9695983378455692e-07, "loss": 0.2798, "step": 1826 }, { "epoch": 0.4861628525811602, "grad_norm": 0.273394376039505, "learning_rate": 1.9695565753925125e-07, "loss": 0.2526, "step": 1827 }, { "epoch": 0.48642895156998406, "grad_norm": 0.30213475227355957, "learning_rate": 1.9695147847181814e-07, "loss": 0.2463, "step": 1828 }, { "epoch": 0.48669505055880785, "grad_norm": 0.27803292870521545, "learning_rate": 1.9694729658237923e-07, "loss": 0.2387, "step": 1829 }, { "epoch": 0.4869611495476317, "grad_norm": 0.3013347089290619, "learning_rate": 1.9694311187105623e-07, "loss": 0.2505, "step": 1830 }, { "epoch": 0.48722724853645555, "grad_norm": 0.3602839410305023, "learning_rate": 1.9693892433797094e-07, "loss": 0.257, "step": 1831 }, { "epoch": 0.4874933475252794, "grad_norm": 0.2993117868900299, "learning_rate": 1.969347339832453e-07, "loss": 0.2547, "step": 1832 }, { "epoch": 0.48775944651410325, "grad_norm": 0.28177449107170105, "learning_rate": 1.9693054080700122e-07, "loss": 0.2204, "step": 1833 }, { "epoch": 0.4880255455029271, "grad_norm": 0.31335949897766113, "learning_rate": 1.9692634480936076e-07, "loss": 0.2622, "step": 1834 }, { "epoch": 0.48829164449175094, "grad_norm": 0.285550981760025, "learning_rate": 1.9692214599044607e-07, "loss": 0.2344, "step": 1835 }, { "epoch": 0.4885577434805748, "grad_norm": 0.2945801913738251, "learning_rate": 1.969179443503794e-07, "loss": 0.2611, "step": 1836 }, { "epoch": 0.48882384246939864, "grad_norm": 0.3097513020038605, "learning_rate": 1.96913739889283e-07, "loss": 0.2479, "step": 1837 }, { "epoch": 0.48908994145822243, "grad_norm": 0.2807091176509857, "learning_rate": 1.9690953260727924e-07, "loss": 0.2564, "step": 1838 }, { "epoch": 0.4893560404470463, "grad_norm": 0.2621830999851227, "learning_rate": 1.9690532250449063e-07, "loss": 0.2223, "step": 1839 }, { "epoch": 0.48962213943587013, "grad_norm": 0.2723638713359833, "learning_rate": 1.969011095810397e-07, "loss": 0.2366, "step": 1840 }, { "epoch": 0.489888238424694, "grad_norm": 0.3155728876590729, "learning_rate": 1.9689689383704905e-07, "loss": 0.2427, "step": 1841 }, { "epoch": 0.4901543374135178, "grad_norm": 0.30182892084121704, "learning_rate": 1.9689267527264143e-07, "loss": 0.2728, "step": 1842 }, { "epoch": 0.4904204364023417, "grad_norm": 0.2966359853744507, "learning_rate": 1.9688845388793962e-07, "loss": 0.2727, "step": 1843 }, { "epoch": 0.4906865353911655, "grad_norm": 0.2875083088874817, "learning_rate": 1.968842296830665e-07, "loss": 0.274, "step": 1844 }, { "epoch": 0.49095263437998937, "grad_norm": 0.28583449125289917, "learning_rate": 1.9688000265814495e-07, "loss": 0.2363, "step": 1845 }, { "epoch": 0.4912187333688132, "grad_norm": 0.3008604645729065, "learning_rate": 1.968757728132981e-07, "loss": 0.2464, "step": 1846 }, { "epoch": 0.491484832357637, "grad_norm": 0.2930481433868408, "learning_rate": 1.9687154014864906e-07, "loss": 0.2523, "step": 1847 }, { "epoch": 0.49175093134646086, "grad_norm": 0.29458507895469666, "learning_rate": 1.9686730466432098e-07, "loss": 0.247, "step": 1848 }, { "epoch": 0.4920170303352847, "grad_norm": 0.29678815603256226, "learning_rate": 1.968630663604372e-07, "loss": 0.2349, "step": 1849 }, { "epoch": 0.49228312932410856, "grad_norm": 0.26221826672554016, "learning_rate": 1.9685882523712105e-07, "loss": 0.2545, "step": 1850 }, { "epoch": 0.4925492283129324, "grad_norm": 0.29348936676979065, "learning_rate": 1.9685458129449598e-07, "loss": 0.2574, "step": 1851 }, { "epoch": 0.49281532730175626, "grad_norm": 0.2804020047187805, "learning_rate": 1.968503345326855e-07, "loss": 0.2358, "step": 1852 }, { "epoch": 0.4930814262905801, "grad_norm": 0.280061274766922, "learning_rate": 1.968460849518133e-07, "loss": 0.241, "step": 1853 }, { "epoch": 0.49334752527940395, "grad_norm": 0.27473556995391846, "learning_rate": 1.9684183255200302e-07, "loss": 0.243, "step": 1854 }, { "epoch": 0.4936136242682278, "grad_norm": 0.30482086539268494, "learning_rate": 1.9683757733337844e-07, "loss": 0.2445, "step": 1855 }, { "epoch": 0.49387972325705165, "grad_norm": 0.2622372806072235, "learning_rate": 1.9683331929606342e-07, "loss": 0.233, "step": 1856 }, { "epoch": 0.49414582224587544, "grad_norm": 0.2620757222175598, "learning_rate": 1.968290584401819e-07, "loss": 0.2196, "step": 1857 }, { "epoch": 0.4944119212346993, "grad_norm": 0.2895831763744354, "learning_rate": 1.968247947658579e-07, "loss": 0.2471, "step": 1858 }, { "epoch": 0.49467802022352314, "grad_norm": 0.2874458432197571, "learning_rate": 1.968205282732155e-07, "loss": 0.2462, "step": 1859 }, { "epoch": 0.494944119212347, "grad_norm": 0.27740129828453064, "learning_rate": 1.9681625896237895e-07, "loss": 0.2356, "step": 1860 }, { "epoch": 0.49521021820117084, "grad_norm": 0.2792331874370575, "learning_rate": 1.9681198683347246e-07, "loss": 0.246, "step": 1861 }, { "epoch": 0.4954763171899947, "grad_norm": 0.2834020256996155, "learning_rate": 1.9680771188662042e-07, "loss": 0.2381, "step": 1862 }, { "epoch": 0.49574241617881853, "grad_norm": 0.28814077377319336, "learning_rate": 1.9680343412194723e-07, "loss": 0.2529, "step": 1863 }, { "epoch": 0.4960085151676424, "grad_norm": 0.3106045126914978, "learning_rate": 1.9679915353957744e-07, "loss": 0.2339, "step": 1864 }, { "epoch": 0.49627461415646623, "grad_norm": 0.2944796085357666, "learning_rate": 1.9679487013963562e-07, "loss": 0.2606, "step": 1865 }, { "epoch": 0.49654071314529, "grad_norm": 0.25429150462150574, "learning_rate": 1.9679058392224646e-07, "loss": 0.228, "step": 1866 }, { "epoch": 0.4968068121341139, "grad_norm": 0.2783243954181671, "learning_rate": 1.9678629488753468e-07, "loss": 0.2309, "step": 1867 }, { "epoch": 0.4970729111229377, "grad_norm": 0.2682492733001709, "learning_rate": 1.9678200303562523e-07, "loss": 0.2343, "step": 1868 }, { "epoch": 0.49733901011176157, "grad_norm": 0.44627952575683594, "learning_rate": 1.967777083666429e-07, "loss": 0.2528, "step": 1869 }, { "epoch": 0.4976051091005854, "grad_norm": 0.2769223749637604, "learning_rate": 1.9677341088071277e-07, "loss": 0.243, "step": 1870 }, { "epoch": 0.49787120808940927, "grad_norm": 0.26643678545951843, "learning_rate": 1.9676911057795994e-07, "loss": 0.2263, "step": 1871 }, { "epoch": 0.4981373070782331, "grad_norm": 0.28352925181388855, "learning_rate": 1.9676480745850954e-07, "loss": 0.2486, "step": 1872 }, { "epoch": 0.49840340606705696, "grad_norm": 0.2854593098163605, "learning_rate": 1.9676050152248686e-07, "loss": 0.2251, "step": 1873 }, { "epoch": 0.4986695050558808, "grad_norm": 0.3021955192089081, "learning_rate": 1.967561927700172e-07, "loss": 0.28, "step": 1874 }, { "epoch": 0.4989356040447046, "grad_norm": 0.27895593643188477, "learning_rate": 1.9675188120122602e-07, "loss": 0.252, "step": 1875 }, { "epoch": 0.49920170303352845, "grad_norm": 0.2954285144805908, "learning_rate": 1.9674756681623877e-07, "loss": 0.2342, "step": 1876 }, { "epoch": 0.4994678020223523, "grad_norm": 0.28577008843421936, "learning_rate": 1.9674324961518106e-07, "loss": 0.245, "step": 1877 }, { "epoch": 0.49973390101117615, "grad_norm": 0.25992217659950256, "learning_rate": 1.9673892959817854e-07, "loss": 0.2269, "step": 1878 }, { "epoch": 0.5, "grad_norm": 0.2741873562335968, "learning_rate": 1.9673460676535697e-07, "loss": 0.2474, "step": 1879 }, { "epoch": 0.5002660989888238, "grad_norm": 0.28275495767593384, "learning_rate": 1.9673028111684214e-07, "loss": 0.2529, "step": 1880 }, { "epoch": 0.5005321979776477, "grad_norm": 0.30935001373291016, "learning_rate": 1.9672595265276e-07, "loss": 0.2637, "step": 1881 }, { "epoch": 0.5007982969664715, "grad_norm": 0.26063817739486694, "learning_rate": 1.9672162137323653e-07, "loss": 0.2099, "step": 1882 }, { "epoch": 0.5010643959552954, "grad_norm": 0.2705231308937073, "learning_rate": 1.967172872783978e-07, "loss": 0.2302, "step": 1883 }, { "epoch": 0.5013304949441192, "grad_norm": 0.26790863275527954, "learning_rate": 1.9671295036836994e-07, "loss": 0.2346, "step": 1884 }, { "epoch": 0.5015965939329431, "grad_norm": 0.27735981345176697, "learning_rate": 1.9670861064327923e-07, "loss": 0.2405, "step": 1885 }, { "epoch": 0.5018626929217669, "grad_norm": 0.2974962592124939, "learning_rate": 1.9670426810325194e-07, "loss": 0.2546, "step": 1886 }, { "epoch": 0.5021287919105908, "grad_norm": 0.29203277826309204, "learning_rate": 1.9669992274841453e-07, "loss": 0.2493, "step": 1887 }, { "epoch": 0.5023948908994146, "grad_norm": 0.3072669804096222, "learning_rate": 1.9669557457889344e-07, "loss": 0.2637, "step": 1888 }, { "epoch": 0.5026609898882384, "grad_norm": 0.28899508714675903, "learning_rate": 1.9669122359481521e-07, "loss": 0.2498, "step": 1889 }, { "epoch": 0.5029270888770623, "grad_norm": 0.31536975502967834, "learning_rate": 1.9668686979630655e-07, "loss": 0.2515, "step": 1890 }, { "epoch": 0.5031931878658861, "grad_norm": 0.2878094017505646, "learning_rate": 1.9668251318349415e-07, "loss": 0.2379, "step": 1891 }, { "epoch": 0.50345928685471, "grad_norm": 0.28214386105537415, "learning_rate": 1.966781537565048e-07, "loss": 0.2369, "step": 1892 }, { "epoch": 0.5037253858435338, "grad_norm": 0.2995879054069519, "learning_rate": 1.9667379151546546e-07, "loss": 0.2465, "step": 1893 }, { "epoch": 0.5039914848323577, "grad_norm": 0.2682172954082489, "learning_rate": 1.9666942646050302e-07, "loss": 0.2352, "step": 1894 }, { "epoch": 0.5042575838211815, "grad_norm": 0.27848976850509644, "learning_rate": 1.966650585917446e-07, "loss": 0.2375, "step": 1895 }, { "epoch": 0.5045236828100054, "grad_norm": 0.2928190529346466, "learning_rate": 1.966606879093173e-07, "loss": 0.2499, "step": 1896 }, { "epoch": 0.5047897817988292, "grad_norm": 0.2755085825920105, "learning_rate": 1.9665631441334836e-07, "loss": 0.245, "step": 1897 }, { "epoch": 0.505055880787653, "grad_norm": 0.3004577159881592, "learning_rate": 1.9665193810396508e-07, "loss": 0.2403, "step": 1898 }, { "epoch": 0.5053219797764769, "grad_norm": 0.2648378312587738, "learning_rate": 1.966475589812948e-07, "loss": 0.2443, "step": 1899 }, { "epoch": 0.5055880787653007, "grad_norm": 0.29142218828201294, "learning_rate": 1.9664317704546508e-07, "loss": 0.2584, "step": 1900 }, { "epoch": 0.5058541777541246, "grad_norm": 0.28178179264068604, "learning_rate": 1.9663879229660336e-07, "loss": 0.2517, "step": 1901 }, { "epoch": 0.5061202767429483, "grad_norm": 0.2653263807296753, "learning_rate": 1.9663440473483734e-07, "loss": 0.2216, "step": 1902 }, { "epoch": 0.5063863757317723, "grad_norm": 0.27531003952026367, "learning_rate": 1.9663001436029468e-07, "loss": 0.2552, "step": 1903 }, { "epoch": 0.506652474720596, "grad_norm": 0.2769027352333069, "learning_rate": 1.966256211731032e-07, "loss": 0.2285, "step": 1904 }, { "epoch": 0.50691857370942, "grad_norm": 0.29251086711883545, "learning_rate": 1.9662122517339081e-07, "loss": 0.2453, "step": 1905 }, { "epoch": 0.5071846726982437, "grad_norm": 0.3066920042037964, "learning_rate": 1.9661682636128544e-07, "loss": 0.2576, "step": 1906 }, { "epoch": 0.5074507716870675, "grad_norm": 0.3060716986656189, "learning_rate": 1.9661242473691505e-07, "loss": 0.2486, "step": 1907 }, { "epoch": 0.5077168706758914, "grad_norm": 0.27361300587654114, "learning_rate": 1.9660802030040786e-07, "loss": 0.2432, "step": 1908 }, { "epoch": 0.5079829696647152, "grad_norm": 0.27024826407432556, "learning_rate": 1.9660361305189205e-07, "loss": 0.2266, "step": 1909 }, { "epoch": 0.5082490686535391, "grad_norm": 0.3287641704082489, "learning_rate": 1.965992029914959e-07, "loss": 0.2624, "step": 1910 }, { "epoch": 0.5085151676423629, "grad_norm": 0.29771688580513, "learning_rate": 1.9659479011934774e-07, "loss": 0.252, "step": 1911 }, { "epoch": 0.5087812666311868, "grad_norm": 0.3273282051086426, "learning_rate": 1.9659037443557607e-07, "loss": 0.2454, "step": 1912 }, { "epoch": 0.5090473656200106, "grad_norm": 0.27479153871536255, "learning_rate": 1.965859559403094e-07, "loss": 0.2349, "step": 1913 }, { "epoch": 0.5093134646088345, "grad_norm": 0.2952025532722473, "learning_rate": 1.9658153463367632e-07, "loss": 0.2316, "step": 1914 }, { "epoch": 0.5095795635976583, "grad_norm": 0.3067583739757538, "learning_rate": 1.9657711051580553e-07, "loss": 0.235, "step": 1915 }, { "epoch": 0.5098456625864821, "grad_norm": 0.2865108549594879, "learning_rate": 1.965726835868258e-07, "loss": 0.239, "step": 1916 }, { "epoch": 0.510111761575306, "grad_norm": 0.29666054248809814, "learning_rate": 1.9656825384686604e-07, "loss": 0.2493, "step": 1917 }, { "epoch": 0.5103778605641298, "grad_norm": 0.26597556471824646, "learning_rate": 1.965638212960551e-07, "loss": 0.2469, "step": 1918 }, { "epoch": 0.5106439595529537, "grad_norm": 0.2987353205680847, "learning_rate": 1.965593859345221e-07, "loss": 0.2387, "step": 1919 }, { "epoch": 0.5109100585417775, "grad_norm": 0.2977309823036194, "learning_rate": 1.9655494776239606e-07, "loss": 0.2394, "step": 1920 }, { "epoch": 0.5111761575306014, "grad_norm": 0.2989174723625183, "learning_rate": 1.965505067798062e-07, "loss": 0.2524, "step": 1921 }, { "epoch": 0.5114422565194252, "grad_norm": 0.2811409533023834, "learning_rate": 1.9654606298688177e-07, "loss": 0.2568, "step": 1922 }, { "epoch": 0.5117083555082491, "grad_norm": 0.2866869568824768, "learning_rate": 1.9654161638375214e-07, "loss": 0.2389, "step": 1923 }, { "epoch": 0.5119744544970729, "grad_norm": 0.2607676386833191, "learning_rate": 1.965371669705467e-07, "loss": 0.2247, "step": 1924 }, { "epoch": 0.5122405534858967, "grad_norm": 0.2873866856098175, "learning_rate": 1.96532714747395e-07, "loss": 0.249, "step": 1925 }, { "epoch": 0.5125066524747206, "grad_norm": 0.30090099573135376, "learning_rate": 1.9652825971442667e-07, "loss": 0.2318, "step": 1926 }, { "epoch": 0.5127727514635444, "grad_norm": 0.29104870557785034, "learning_rate": 1.9652380187177125e-07, "loss": 0.2519, "step": 1927 }, { "epoch": 0.5130388504523683, "grad_norm": 0.2875748872756958, "learning_rate": 1.9651934121955862e-07, "loss": 0.2497, "step": 1928 }, { "epoch": 0.5133049494411921, "grad_norm": 0.3251810669898987, "learning_rate": 1.9651487775791858e-07, "loss": 0.2619, "step": 1929 }, { "epoch": 0.513571048430016, "grad_norm": 0.28142496943473816, "learning_rate": 1.96510411486981e-07, "loss": 0.2348, "step": 1930 }, { "epoch": 0.5138371474188398, "grad_norm": 0.281585693359375, "learning_rate": 1.96505942406876e-07, "loss": 0.2177, "step": 1931 }, { "epoch": 0.5141032464076637, "grad_norm": 0.26117968559265137, "learning_rate": 1.9650147051773357e-07, "loss": 0.2409, "step": 1932 }, { "epoch": 0.5143693453964875, "grad_norm": 0.3152022361755371, "learning_rate": 1.9649699581968387e-07, "loss": 0.2494, "step": 1933 }, { "epoch": 0.5146354443853113, "grad_norm": 0.28105607628822327, "learning_rate": 1.964925183128572e-07, "loss": 0.2323, "step": 1934 }, { "epoch": 0.5149015433741352, "grad_norm": 0.32178449630737305, "learning_rate": 1.9648803799738384e-07, "loss": 0.2531, "step": 1935 }, { "epoch": 0.515167642362959, "grad_norm": 0.2700273394584656, "learning_rate": 1.9648355487339423e-07, "loss": 0.2297, "step": 1936 }, { "epoch": 0.5154337413517829, "grad_norm": 0.2707112729549408, "learning_rate": 1.9647906894101885e-07, "loss": 0.2354, "step": 1937 }, { "epoch": 0.5156998403406067, "grad_norm": 0.38228949904441833, "learning_rate": 1.964745802003883e-07, "loss": 0.2448, "step": 1938 }, { "epoch": 0.5159659393294306, "grad_norm": 0.2736443281173706, "learning_rate": 1.9647008865163318e-07, "loss": 0.2361, "step": 1939 }, { "epoch": 0.5162320383182544, "grad_norm": 0.26854366064071655, "learning_rate": 1.964655942948843e-07, "loss": 0.2356, "step": 1940 }, { "epoch": 0.5164981373070783, "grad_norm": 0.33138230443000793, "learning_rate": 1.964610971302724e-07, "loss": 0.2655, "step": 1941 }, { "epoch": 0.5167642362959021, "grad_norm": 0.2763615548610687, "learning_rate": 1.9645659715792846e-07, "loss": 0.2226, "step": 1942 }, { "epoch": 0.517030335284726, "grad_norm": 0.30065274238586426, "learning_rate": 1.964520943779834e-07, "loss": 0.2627, "step": 1943 }, { "epoch": 0.5172964342735498, "grad_norm": 0.2827964723110199, "learning_rate": 1.9644758879056833e-07, "loss": 0.2475, "step": 1944 }, { "epoch": 0.5175625332623736, "grad_norm": 0.2792786955833435, "learning_rate": 1.9644308039581435e-07, "loss": 0.2325, "step": 1945 }, { "epoch": 0.5178286322511975, "grad_norm": 0.3007681965827942, "learning_rate": 1.9643856919385273e-07, "loss": 0.2332, "step": 1946 }, { "epoch": 0.5180947312400213, "grad_norm": 0.27624276280403137, "learning_rate": 1.9643405518481474e-07, "loss": 0.228, "step": 1947 }, { "epoch": 0.5183608302288452, "grad_norm": 0.276042640209198, "learning_rate": 1.964295383688318e-07, "loss": 0.2391, "step": 1948 }, { "epoch": 0.518626929217669, "grad_norm": 0.2782824635505676, "learning_rate": 1.9642501874603536e-07, "loss": 0.2509, "step": 1949 }, { "epoch": 0.5188930282064929, "grad_norm": 0.5370171666145325, "learning_rate": 1.9642049631655705e-07, "loss": 0.2305, "step": 1950 }, { "epoch": 0.5191591271953166, "grad_norm": 0.3052118122577667, "learning_rate": 1.9641597108052838e-07, "loss": 0.2548, "step": 1951 }, { "epoch": 0.5194252261841406, "grad_norm": 0.29908889532089233, "learning_rate": 1.9641144303808116e-07, "loss": 0.2435, "step": 1952 }, { "epoch": 0.5196913251729643, "grad_norm": 0.296631783246994, "learning_rate": 1.9640691218934714e-07, "loss": 0.246, "step": 1953 }, { "epoch": 0.5199574241617881, "grad_norm": 0.28721433877944946, "learning_rate": 1.9640237853445826e-07, "loss": 0.2514, "step": 1954 }, { "epoch": 0.520223523150612, "grad_norm": 0.25547346472740173, "learning_rate": 1.963978420735464e-07, "loss": 0.2214, "step": 1955 }, { "epoch": 0.5204896221394358, "grad_norm": 0.279003769159317, "learning_rate": 1.963933028067437e-07, "loss": 0.2345, "step": 1956 }, { "epoch": 0.5207557211282597, "grad_norm": 0.3256964683532715, "learning_rate": 1.9638876073418222e-07, "loss": 0.2969, "step": 1957 }, { "epoch": 0.5210218201170835, "grad_norm": 0.29241570830345154, "learning_rate": 1.9638421585599422e-07, "loss": 0.2568, "step": 1958 }, { "epoch": 0.5212879191059074, "grad_norm": 0.2555227279663086, "learning_rate": 1.9637966817231191e-07, "loss": 0.213, "step": 1959 }, { "epoch": 0.5215540180947312, "grad_norm": 0.30663222074508667, "learning_rate": 1.9637511768326771e-07, "loss": 0.2464, "step": 1960 }, { "epoch": 0.5218201170835551, "grad_norm": 0.30518078804016113, "learning_rate": 1.963705643889941e-07, "loss": 0.2649, "step": 1961 }, { "epoch": 0.5220862160723789, "grad_norm": 0.25907984375953674, "learning_rate": 1.9636600828962354e-07, "loss": 0.2185, "step": 1962 }, { "epoch": 0.5223523150612027, "grad_norm": 0.27910929918289185, "learning_rate": 1.9636144938528872e-07, "loss": 0.2403, "step": 1963 }, { "epoch": 0.5226184140500266, "grad_norm": 0.2780304551124573, "learning_rate": 1.963568876761223e-07, "loss": 0.2406, "step": 1964 }, { "epoch": 0.5228845130388504, "grad_norm": 0.291877806186676, "learning_rate": 1.9635232316225707e-07, "loss": 0.2418, "step": 1965 }, { "epoch": 0.5231506120276743, "grad_norm": 0.3060474097728729, "learning_rate": 1.9634775584382589e-07, "loss": 0.2589, "step": 1966 }, { "epoch": 0.5234167110164981, "grad_norm": 0.284668892621994, "learning_rate": 1.9634318572096168e-07, "loss": 0.2438, "step": 1967 }, { "epoch": 0.523682810005322, "grad_norm": 0.30065977573394775, "learning_rate": 1.963386127937975e-07, "loss": 0.2429, "step": 1968 }, { "epoch": 0.5239489089941458, "grad_norm": 0.31503957509994507, "learning_rate": 1.9633403706246647e-07, "loss": 0.2628, "step": 1969 }, { "epoch": 0.5242150079829697, "grad_norm": 0.29017698764801025, "learning_rate": 1.9632945852710173e-07, "loss": 0.2607, "step": 1970 }, { "epoch": 0.5244811069717935, "grad_norm": 0.29530227184295654, "learning_rate": 1.9632487718783658e-07, "loss": 0.2595, "step": 1971 }, { "epoch": 0.5247472059606173, "grad_norm": 0.29525887966156006, "learning_rate": 1.9632029304480434e-07, "loss": 0.2428, "step": 1972 }, { "epoch": 0.5250133049494412, "grad_norm": 0.30850672721862793, "learning_rate": 1.9631570609813844e-07, "loss": 0.2623, "step": 1973 }, { "epoch": 0.525279403938265, "grad_norm": 0.32446354627609253, "learning_rate": 1.9631111634797248e-07, "loss": 0.2592, "step": 1974 }, { "epoch": 0.5255455029270889, "grad_norm": 0.30975642800331116, "learning_rate": 1.9630652379443996e-07, "loss": 0.2617, "step": 1975 }, { "epoch": 0.5258116019159127, "grad_norm": 0.3053792417049408, "learning_rate": 1.9630192843767456e-07, "loss": 0.2694, "step": 1976 }, { "epoch": 0.5260777009047366, "grad_norm": 0.29852524399757385, "learning_rate": 1.962973302778101e-07, "loss": 0.2584, "step": 1977 }, { "epoch": 0.5263437998935604, "grad_norm": 0.31411904096603394, "learning_rate": 1.962927293149804e-07, "loss": 0.2431, "step": 1978 }, { "epoch": 0.5266098988823843, "grad_norm": 0.2569565176963806, "learning_rate": 1.9628812554931936e-07, "loss": 0.2278, "step": 1979 }, { "epoch": 0.5268759978712081, "grad_norm": 0.2512952387332916, "learning_rate": 1.96283518980961e-07, "loss": 0.2165, "step": 1980 }, { "epoch": 0.5271420968600319, "grad_norm": 0.2813040614128113, "learning_rate": 1.962789096100394e-07, "loss": 0.2456, "step": 1981 }, { "epoch": 0.5274081958488558, "grad_norm": 0.27130499482154846, "learning_rate": 1.962742974366887e-07, "loss": 0.2229, "step": 1982 }, { "epoch": 0.5276742948376796, "grad_norm": 0.2849302887916565, "learning_rate": 1.9626968246104322e-07, "loss": 0.2542, "step": 1983 }, { "epoch": 0.5279403938265035, "grad_norm": 0.29068952798843384, "learning_rate": 1.962650646832372e-07, "loss": 0.2295, "step": 1984 }, { "epoch": 0.5282064928153273, "grad_norm": 0.29457318782806396, "learning_rate": 1.9626044410340513e-07, "loss": 0.2658, "step": 1985 }, { "epoch": 0.5284725918041512, "grad_norm": 0.33008822798728943, "learning_rate": 1.9625582072168146e-07, "loss": 0.2735, "step": 1986 }, { "epoch": 0.528738690792975, "grad_norm": 0.2586044371128082, "learning_rate": 1.9625119453820078e-07, "loss": 0.2163, "step": 1987 }, { "epoch": 0.5290047897817989, "grad_norm": 0.3074473738670349, "learning_rate": 1.9624656555309773e-07, "loss": 0.2205, "step": 1988 }, { "epoch": 0.5292708887706227, "grad_norm": 0.2877015769481659, "learning_rate": 1.9624193376650707e-07, "loss": 0.2186, "step": 1989 }, { "epoch": 0.5295369877594465, "grad_norm": 0.2628241181373596, "learning_rate": 1.9623729917856358e-07, "loss": 0.2189, "step": 1990 }, { "epoch": 0.5298030867482704, "grad_norm": 0.30865320563316345, "learning_rate": 1.962326617894022e-07, "loss": 0.2627, "step": 1991 }, { "epoch": 0.5300691857370942, "grad_norm": 0.2888731360435486, "learning_rate": 1.962280215991579e-07, "loss": 0.2485, "step": 1992 }, { "epoch": 0.5303352847259181, "grad_norm": 0.2829483449459076, "learning_rate": 1.9622337860796575e-07, "loss": 0.2423, "step": 1993 }, { "epoch": 0.5306013837147419, "grad_norm": 0.30054208636283875, "learning_rate": 1.9621873281596092e-07, "loss": 0.2627, "step": 1994 }, { "epoch": 0.5308674827035658, "grad_norm": 0.27993044257164, "learning_rate": 1.962140842232786e-07, "loss": 0.23, "step": 1995 }, { "epoch": 0.5311335816923896, "grad_norm": 0.30778738856315613, "learning_rate": 1.9620943283005405e-07, "loss": 0.2444, "step": 1996 }, { "epoch": 0.5313996806812135, "grad_norm": 0.2558809816837311, "learning_rate": 1.9620477863642275e-07, "loss": 0.2275, "step": 1997 }, { "epoch": 0.5316657796700373, "grad_norm": 0.27629274129867554, "learning_rate": 1.9620012164252016e-07, "loss": 0.2368, "step": 1998 }, { "epoch": 0.531931878658861, "grad_norm": 0.29737159609794617, "learning_rate": 1.961954618484818e-07, "loss": 0.2496, "step": 1999 }, { "epoch": 0.532197977647685, "grad_norm": 0.31817758083343506, "learning_rate": 1.961907992544433e-07, "loss": 0.2512, "step": 2000 }, { "epoch": 0.5324640766365087, "grad_norm": 0.2784583270549774, "learning_rate": 1.9618613386054038e-07, "loss": 0.2391, "step": 2001 }, { "epoch": 0.5327301756253326, "grad_norm": 0.27836549282073975, "learning_rate": 1.9618146566690884e-07, "loss": 0.2218, "step": 2002 }, { "epoch": 0.5329962746141564, "grad_norm": 0.30166399478912354, "learning_rate": 1.9617679467368463e-07, "loss": 0.2456, "step": 2003 }, { "epoch": 0.5332623736029803, "grad_norm": 0.2924644947052002, "learning_rate": 1.9617212088100358e-07, "loss": 0.2485, "step": 2004 }, { "epoch": 0.5335284725918041, "grad_norm": 0.28526201844215393, "learning_rate": 1.9616744428900183e-07, "loss": 0.2307, "step": 2005 }, { "epoch": 0.533794571580628, "grad_norm": 0.3175643980503082, "learning_rate": 1.9616276489781546e-07, "loss": 0.2622, "step": 2006 }, { "epoch": 0.5340606705694518, "grad_norm": 0.286317378282547, "learning_rate": 1.961580827075807e-07, "loss": 0.234, "step": 2007 }, { "epoch": 0.5343267695582756, "grad_norm": 0.31864580512046814, "learning_rate": 1.9615339771843382e-07, "loss": 0.2325, "step": 2008 }, { "epoch": 0.5345928685470995, "grad_norm": 0.3048500418663025, "learning_rate": 1.9614870993051122e-07, "loss": 0.2411, "step": 2009 }, { "epoch": 0.5348589675359233, "grad_norm": 0.3089664876461029, "learning_rate": 1.961440193439493e-07, "loss": 0.2532, "step": 2010 }, { "epoch": 0.5351250665247472, "grad_norm": 0.276081919670105, "learning_rate": 1.961393259588846e-07, "loss": 0.2364, "step": 2011 }, { "epoch": 0.535391165513571, "grad_norm": 0.326726496219635, "learning_rate": 1.9613462977545375e-07, "loss": 0.2843, "step": 2012 }, { "epoch": 0.5356572645023949, "grad_norm": 0.3005406856536865, "learning_rate": 1.9612993079379344e-07, "loss": 0.2447, "step": 2013 }, { "epoch": 0.5359233634912187, "grad_norm": 0.2916666269302368, "learning_rate": 1.9612522901404042e-07, "loss": 0.257, "step": 2014 }, { "epoch": 0.5361894624800426, "grad_norm": 0.3315850794315338, "learning_rate": 1.961205244363316e-07, "loss": 0.2631, "step": 2015 }, { "epoch": 0.5364555614688664, "grad_norm": 0.2895415425300598, "learning_rate": 1.961158170608039e-07, "loss": 0.2475, "step": 2016 }, { "epoch": 0.5367216604576902, "grad_norm": 0.2939874231815338, "learning_rate": 1.961111068875943e-07, "loss": 0.2411, "step": 2017 }, { "epoch": 0.5369877594465141, "grad_norm": 0.30199357867240906, "learning_rate": 1.9610639391683994e-07, "loss": 0.2404, "step": 2018 }, { "epoch": 0.5372538584353379, "grad_norm": 0.27535170316696167, "learning_rate": 1.96101678148678e-07, "loss": 0.2318, "step": 2019 }, { "epoch": 0.5375199574241618, "grad_norm": 0.29536837339401245, "learning_rate": 1.960969595832457e-07, "loss": 0.2428, "step": 2020 }, { "epoch": 0.5377860564129856, "grad_norm": 0.2862374782562256, "learning_rate": 1.9609223822068043e-07, "loss": 0.2306, "step": 2021 }, { "epoch": 0.5380521554018095, "grad_norm": 0.30821365118026733, "learning_rate": 1.9608751406111966e-07, "loss": 0.2647, "step": 2022 }, { "epoch": 0.5383182543906333, "grad_norm": 0.3174232244491577, "learning_rate": 1.9608278710470078e-07, "loss": 0.2509, "step": 2023 }, { "epoch": 0.5385843533794572, "grad_norm": 0.2719883322715759, "learning_rate": 1.9607805735156146e-07, "loss": 0.2408, "step": 2024 }, { "epoch": 0.538850452368281, "grad_norm": 0.24898143112659454, "learning_rate": 1.9607332480183937e-07, "loss": 0.2196, "step": 2025 }, { "epoch": 0.5391165513571048, "grad_norm": 0.2667362689971924, "learning_rate": 1.9606858945567225e-07, "loss": 0.2257, "step": 2026 }, { "epoch": 0.5393826503459287, "grad_norm": 0.25659722089767456, "learning_rate": 1.9606385131319791e-07, "loss": 0.2252, "step": 2027 }, { "epoch": 0.5396487493347525, "grad_norm": 0.2803433835506439, "learning_rate": 1.9605911037455431e-07, "loss": 0.2365, "step": 2028 }, { "epoch": 0.5399148483235764, "grad_norm": 0.2859717309474945, "learning_rate": 1.960543666398794e-07, "loss": 0.2508, "step": 2029 }, { "epoch": 0.5401809473124002, "grad_norm": 0.29092416167259216, "learning_rate": 1.9604962010931128e-07, "loss": 0.228, "step": 2030 }, { "epoch": 0.5404470463012241, "grad_norm": 0.2673957943916321, "learning_rate": 1.9604487078298814e-07, "loss": 0.2216, "step": 2031 }, { "epoch": 0.5407131452900479, "grad_norm": 0.3004654049873352, "learning_rate": 1.9604011866104815e-07, "loss": 0.2376, "step": 2032 }, { "epoch": 0.5409792442788718, "grad_norm": 0.26734158396720886, "learning_rate": 1.960353637436297e-07, "loss": 0.2508, "step": 2033 }, { "epoch": 0.5412453432676956, "grad_norm": 0.29728445410728455, "learning_rate": 1.9603060603087114e-07, "loss": 0.2492, "step": 2034 }, { "epoch": 0.5415114422565194, "grad_norm": 0.30483269691467285, "learning_rate": 1.96025845522911e-07, "loss": 0.2552, "step": 2035 }, { "epoch": 0.5417775412453433, "grad_norm": 0.28463345766067505, "learning_rate": 1.9602108221988783e-07, "loss": 0.2375, "step": 2036 }, { "epoch": 0.5420436402341671, "grad_norm": 0.3047102689743042, "learning_rate": 1.9601631612194022e-07, "loss": 0.232, "step": 2037 }, { "epoch": 0.542309739222991, "grad_norm": 0.2958019971847534, "learning_rate": 1.96011547229207e-07, "loss": 0.278, "step": 2038 }, { "epoch": 0.5425758382118148, "grad_norm": 0.2991846203804016, "learning_rate": 1.960067755418269e-07, "loss": 0.2446, "step": 2039 }, { "epoch": 0.5428419372006387, "grad_norm": 0.27655595541000366, "learning_rate": 1.960020010599389e-07, "loss": 0.2373, "step": 2040 }, { "epoch": 0.5431080361894625, "grad_norm": 0.29222387075424194, "learning_rate": 1.9599722378368186e-07, "loss": 0.2347, "step": 2041 }, { "epoch": 0.5433741351782864, "grad_norm": 0.3094541132450104, "learning_rate": 1.9599244371319493e-07, "loss": 0.2694, "step": 2042 }, { "epoch": 0.5436402341671102, "grad_norm": 0.4000730514526367, "learning_rate": 1.959876608486172e-07, "loss": 0.274, "step": 2043 }, { "epoch": 0.543906333155934, "grad_norm": 0.28304556012153625, "learning_rate": 1.9598287519008786e-07, "loss": 0.2365, "step": 2044 }, { "epoch": 0.5441724321447579, "grad_norm": 0.2956477403640747, "learning_rate": 1.9597808673774627e-07, "loss": 0.2394, "step": 2045 }, { "epoch": 0.5444385311335816, "grad_norm": 0.2714490294456482, "learning_rate": 1.9597329549173178e-07, "loss": 0.231, "step": 2046 }, { "epoch": 0.5447046301224056, "grad_norm": 0.30370745062828064, "learning_rate": 1.9596850145218382e-07, "loss": 0.2696, "step": 2047 }, { "epoch": 0.5449707291112293, "grad_norm": 0.30172958970069885, "learning_rate": 1.95963704619242e-07, "loss": 0.2879, "step": 2048 }, { "epoch": 0.5452368281000533, "grad_norm": 0.2652203142642975, "learning_rate": 1.9595890499304587e-07, "loss": 0.2251, "step": 2049 }, { "epoch": 0.545502927088877, "grad_norm": 0.27059364318847656, "learning_rate": 1.9595410257373518e-07, "loss": 0.2456, "step": 2050 }, { "epoch": 0.545769026077701, "grad_norm": 0.28521355986595154, "learning_rate": 1.9594929736144973e-07, "loss": 0.2313, "step": 2051 }, { "epoch": 0.5460351250665247, "grad_norm": 0.2576060891151428, "learning_rate": 1.9594448935632934e-07, "loss": 0.2226, "step": 2052 }, { "epoch": 0.5463012240553486, "grad_norm": 0.3059050142765045, "learning_rate": 1.9593967855851398e-07, "loss": 0.2463, "step": 2053 }, { "epoch": 0.5465673230441724, "grad_norm": 0.2858406901359558, "learning_rate": 1.959348649681437e-07, "loss": 0.2181, "step": 2054 }, { "epoch": 0.5468334220329962, "grad_norm": 0.2873798906803131, "learning_rate": 1.959300485853586e-07, "loss": 0.2541, "step": 2055 }, { "epoch": 0.5470995210218201, "grad_norm": 0.28238773345947266, "learning_rate": 1.9592522941029885e-07, "loss": 0.2318, "step": 2056 }, { "epoch": 0.5473656200106439, "grad_norm": 0.28857576847076416, "learning_rate": 1.9592040744310474e-07, "loss": 0.2435, "step": 2057 }, { "epoch": 0.5476317189994678, "grad_norm": 0.26685962080955505, "learning_rate": 1.9591558268391665e-07, "loss": 0.2184, "step": 2058 }, { "epoch": 0.5478978179882916, "grad_norm": 0.28497201204299927, "learning_rate": 1.9591075513287493e-07, "loss": 0.2384, "step": 2059 }, { "epoch": 0.5481639169771155, "grad_norm": 0.27328160405158997, "learning_rate": 1.959059247901202e-07, "loss": 0.2287, "step": 2060 }, { "epoch": 0.5484300159659393, "grad_norm": 0.28434860706329346, "learning_rate": 1.95901091655793e-07, "loss": 0.2625, "step": 2061 }, { "epoch": 0.5486961149547632, "grad_norm": 0.27451956272125244, "learning_rate": 1.9589625573003405e-07, "loss": 0.2452, "step": 2062 }, { "epoch": 0.548962213943587, "grad_norm": 0.2819451093673706, "learning_rate": 1.9589141701298408e-07, "loss": 0.2416, "step": 2063 }, { "epoch": 0.5492283129324108, "grad_norm": 0.2477824091911316, "learning_rate": 1.9588657550478392e-07, "loss": 0.2269, "step": 2064 }, { "epoch": 0.5494944119212347, "grad_norm": 0.2984274625778198, "learning_rate": 1.9588173120557454e-07, "loss": 0.2446, "step": 2065 }, { "epoch": 0.5497605109100585, "grad_norm": 0.2847837209701538, "learning_rate": 1.958768841154969e-07, "loss": 0.2396, "step": 2066 }, { "epoch": 0.5500266098988824, "grad_norm": 0.29552537202835083, "learning_rate": 1.958720342346921e-07, "loss": 0.255, "step": 2067 }, { "epoch": 0.5502927088877062, "grad_norm": 0.2678492069244385, "learning_rate": 1.9586718156330133e-07, "loss": 0.2512, "step": 2068 }, { "epoch": 0.5505588078765301, "grad_norm": 0.28662800788879395, "learning_rate": 1.9586232610146581e-07, "loss": 0.2496, "step": 2069 }, { "epoch": 0.5508249068653539, "grad_norm": 0.32259008288383484, "learning_rate": 1.958574678493269e-07, "loss": 0.2677, "step": 2070 }, { "epoch": 0.5510910058541778, "grad_norm": 0.2923247516155243, "learning_rate": 1.95852606807026e-07, "loss": 0.2604, "step": 2071 }, { "epoch": 0.5513571048430016, "grad_norm": 0.28666093945503235, "learning_rate": 1.9584774297470457e-07, "loss": 0.235, "step": 2072 }, { "epoch": 0.5516232038318254, "grad_norm": 0.3247170150279999, "learning_rate": 1.958428763525042e-07, "loss": 0.2599, "step": 2073 }, { "epoch": 0.5518893028206493, "grad_norm": 0.2916679382324219, "learning_rate": 1.9583800694056657e-07, "loss": 0.2482, "step": 2074 }, { "epoch": 0.5521554018094731, "grad_norm": 0.27938327193260193, "learning_rate": 1.958331347390334e-07, "loss": 0.2551, "step": 2075 }, { "epoch": 0.552421500798297, "grad_norm": 0.26298707723617554, "learning_rate": 1.958282597480465e-07, "loss": 0.235, "step": 2076 }, { "epoch": 0.5526875997871208, "grad_norm": 0.30948111414909363, "learning_rate": 1.9582338196774777e-07, "loss": 0.2254, "step": 2077 }, { "epoch": 0.5529536987759447, "grad_norm": 0.2727518379688263, "learning_rate": 1.958185013982792e-07, "loss": 0.2402, "step": 2078 }, { "epoch": 0.5532197977647685, "grad_norm": 0.29912054538726807, "learning_rate": 1.9581361803978283e-07, "loss": 0.2377, "step": 2079 }, { "epoch": 0.5534858967535924, "grad_norm": 0.2960131764411926, "learning_rate": 1.9580873189240083e-07, "loss": 0.2472, "step": 2080 }, { "epoch": 0.5537519957424162, "grad_norm": 0.2694871425628662, "learning_rate": 1.9580384295627543e-07, "loss": 0.2313, "step": 2081 }, { "epoch": 0.55401809473124, "grad_norm": 0.30815431475639343, "learning_rate": 1.9579895123154887e-07, "loss": 0.2419, "step": 2082 }, { "epoch": 0.5542841937200639, "grad_norm": 0.283701092004776, "learning_rate": 1.9579405671836362e-07, "loss": 0.2451, "step": 2083 }, { "epoch": 0.5545502927088877, "grad_norm": 0.289159893989563, "learning_rate": 1.957891594168621e-07, "loss": 0.2332, "step": 2084 }, { "epoch": 0.5548163916977116, "grad_norm": 0.29469034075737, "learning_rate": 1.9578425932718682e-07, "loss": 0.2441, "step": 2085 }, { "epoch": 0.5550824906865354, "grad_norm": 0.3281014859676361, "learning_rate": 1.9577935644948049e-07, "loss": 0.2609, "step": 2086 }, { "epoch": 0.5553485896753593, "grad_norm": 0.2809586226940155, "learning_rate": 1.957744507838858e-07, "loss": 0.2424, "step": 2087 }, { "epoch": 0.5556146886641831, "grad_norm": 0.28668341040611267, "learning_rate": 1.9576954233054547e-07, "loss": 0.243, "step": 2088 }, { "epoch": 0.555880787653007, "grad_norm": 0.27130210399627686, "learning_rate": 1.9576463108960247e-07, "loss": 0.2436, "step": 2089 }, { "epoch": 0.5561468866418308, "grad_norm": 0.3007389008998871, "learning_rate": 1.957597170611997e-07, "loss": 0.2368, "step": 2090 }, { "epoch": 0.5564129856306546, "grad_norm": 0.2771264612674713, "learning_rate": 1.957548002454802e-07, "loss": 0.2398, "step": 2091 }, { "epoch": 0.5566790846194785, "grad_norm": 0.28732985258102417, "learning_rate": 1.9574988064258707e-07, "loss": 0.2399, "step": 2092 }, { "epoch": 0.5569451836083023, "grad_norm": 0.2845172882080078, "learning_rate": 1.9574495825266354e-07, "loss": 0.2468, "step": 2093 }, { "epoch": 0.5572112825971262, "grad_norm": 0.2885955572128296, "learning_rate": 1.9574003307585288e-07, "loss": 0.2426, "step": 2094 }, { "epoch": 0.55747738158595, "grad_norm": 0.28148823976516724, "learning_rate": 1.9573510511229845e-07, "loss": 0.242, "step": 2095 }, { "epoch": 0.5577434805747739, "grad_norm": 0.2783282697200775, "learning_rate": 1.9573017436214368e-07, "loss": 0.2318, "step": 2096 }, { "epoch": 0.5580095795635976, "grad_norm": 0.31365975737571716, "learning_rate": 1.957252408255321e-07, "loss": 0.2431, "step": 2097 }, { "epoch": 0.5582756785524216, "grad_norm": 0.277357816696167, "learning_rate": 1.9572030450260727e-07, "loss": 0.2394, "step": 2098 }, { "epoch": 0.5585417775412453, "grad_norm": 0.28536471724510193, "learning_rate": 1.9571536539351296e-07, "loss": 0.2457, "step": 2099 }, { "epoch": 0.5588078765300691, "grad_norm": 0.28811705112457275, "learning_rate": 1.9571042349839286e-07, "loss": 0.2557, "step": 2100 }, { "epoch": 0.559073975518893, "grad_norm": 0.25487157702445984, "learning_rate": 1.9570547881739085e-07, "loss": 0.2177, "step": 2101 }, { "epoch": 0.5593400745077168, "grad_norm": 0.285563200712204, "learning_rate": 1.9570053135065087e-07, "loss": 0.24, "step": 2102 }, { "epoch": 0.5596061734965407, "grad_norm": 0.28201568126678467, "learning_rate": 1.9569558109831688e-07, "loss": 0.2288, "step": 2103 }, { "epoch": 0.5598722724853645, "grad_norm": 0.2921004593372345, "learning_rate": 1.9569062806053297e-07, "loss": 0.2579, "step": 2104 }, { "epoch": 0.5601383714741884, "grad_norm": 0.27400609850883484, "learning_rate": 1.9568567223744337e-07, "loss": 0.2226, "step": 2105 }, { "epoch": 0.5604044704630122, "grad_norm": 0.254003643989563, "learning_rate": 1.956807136291923e-07, "loss": 0.2182, "step": 2106 }, { "epoch": 0.5606705694518361, "grad_norm": 0.3038879334926605, "learning_rate": 1.9567575223592407e-07, "loss": 0.2487, "step": 2107 }, { "epoch": 0.5609366684406599, "grad_norm": 0.26415887475013733, "learning_rate": 1.956707880577831e-07, "loss": 0.2484, "step": 2108 }, { "epoch": 0.5612027674294837, "grad_norm": 0.2673589885234833, "learning_rate": 1.9566582109491395e-07, "loss": 0.2199, "step": 2109 }, { "epoch": 0.5614688664183076, "grad_norm": 0.277776300907135, "learning_rate": 1.9566085134746108e-07, "loss": 0.2214, "step": 2110 }, { "epoch": 0.5617349654071314, "grad_norm": 0.2846212685108185, "learning_rate": 1.9565587881556927e-07, "loss": 0.2433, "step": 2111 }, { "epoch": 0.5620010643959553, "grad_norm": 0.3077375292778015, "learning_rate": 1.9565090349938315e-07, "loss": 0.255, "step": 2112 }, { "epoch": 0.5622671633847791, "grad_norm": 0.2890722155570984, "learning_rate": 1.956459253990476e-07, "loss": 0.2237, "step": 2113 }, { "epoch": 0.562533262373603, "grad_norm": 0.26391321420669556, "learning_rate": 1.956409445147075e-07, "loss": 0.2237, "step": 2114 }, { "epoch": 0.5627993613624268, "grad_norm": 0.284339964389801, "learning_rate": 1.9563596084650782e-07, "loss": 0.2537, "step": 2115 }, { "epoch": 0.5630654603512507, "grad_norm": 0.27172115445137024, "learning_rate": 1.9563097439459366e-07, "loss": 0.2354, "step": 2116 }, { "epoch": 0.5633315593400745, "grad_norm": 0.2835240662097931, "learning_rate": 1.9562598515911014e-07, "loss": 0.2382, "step": 2117 }, { "epoch": 0.5635976583288983, "grad_norm": 0.2809179723262787, "learning_rate": 1.9562099314020246e-07, "loss": 0.2359, "step": 2118 }, { "epoch": 0.5638637573177222, "grad_norm": 0.2540295124053955, "learning_rate": 1.9561599833801595e-07, "loss": 0.2273, "step": 2119 }, { "epoch": 0.564129856306546, "grad_norm": 0.2761416733264923, "learning_rate": 1.9561100075269602e-07, "loss": 0.2374, "step": 2120 }, { "epoch": 0.5643959552953699, "grad_norm": 0.28456270694732666, "learning_rate": 1.9560600038438812e-07, "loss": 0.2416, "step": 2121 }, { "epoch": 0.5646620542841937, "grad_norm": 0.25007322430610657, "learning_rate": 1.9560099723323774e-07, "loss": 0.2084, "step": 2122 }, { "epoch": 0.5649281532730176, "grad_norm": 0.3046674132347107, "learning_rate": 1.9559599129939057e-07, "loss": 0.2667, "step": 2123 }, { "epoch": 0.5651942522618414, "grad_norm": 0.2798779010772705, "learning_rate": 1.9559098258299232e-07, "loss": 0.235, "step": 2124 }, { "epoch": 0.5654603512506653, "grad_norm": 0.29017573595046997, "learning_rate": 1.9558597108418874e-07, "loss": 0.2373, "step": 2125 }, { "epoch": 0.5657264502394891, "grad_norm": 0.29124152660369873, "learning_rate": 1.9558095680312573e-07, "loss": 0.25, "step": 2126 }, { "epoch": 0.5659925492283129, "grad_norm": 0.2732543349266052, "learning_rate": 1.955759397399493e-07, "loss": 0.2326, "step": 2127 }, { "epoch": 0.5662586482171368, "grad_norm": 0.2746926248073578, "learning_rate": 1.9557091989480536e-07, "loss": 0.2386, "step": 2128 }, { "epoch": 0.5665247472059606, "grad_norm": 0.2862117886543274, "learning_rate": 1.955658972678401e-07, "loss": 0.1979, "step": 2129 }, { "epoch": 0.5667908461947845, "grad_norm": 0.28457555174827576, "learning_rate": 1.9556087185919972e-07, "loss": 0.2259, "step": 2130 }, { "epoch": 0.5670569451836083, "grad_norm": 0.2676441967487335, "learning_rate": 1.9555584366903048e-07, "loss": 0.2241, "step": 2131 }, { "epoch": 0.5673230441724322, "grad_norm": 0.3089742958545685, "learning_rate": 1.9555081269747875e-07, "loss": 0.2435, "step": 2132 }, { "epoch": 0.567589143161256, "grad_norm": 0.258186012506485, "learning_rate": 1.9554577894469093e-07, "loss": 0.2184, "step": 2133 }, { "epoch": 0.5678552421500799, "grad_norm": 0.2863290309906006, "learning_rate": 1.9554074241081358e-07, "loss": 0.2393, "step": 2134 }, { "epoch": 0.5681213411389037, "grad_norm": 0.275651752948761, "learning_rate": 1.955357030959933e-07, "loss": 0.231, "step": 2135 }, { "epoch": 0.5683874401277275, "grad_norm": 0.29215213656425476, "learning_rate": 1.955306610003767e-07, "loss": 0.251, "step": 2136 }, { "epoch": 0.5686535391165514, "grad_norm": 0.2976193130016327, "learning_rate": 1.955256161241107e-07, "loss": 0.2647, "step": 2137 }, { "epoch": 0.5689196381053752, "grad_norm": 0.3061404824256897, "learning_rate": 1.9552056846734197e-07, "loss": 0.2364, "step": 2138 }, { "epoch": 0.5691857370941991, "grad_norm": 0.3141506612300873, "learning_rate": 1.9551551803021752e-07, "loss": 0.25, "step": 2139 }, { "epoch": 0.5694518360830229, "grad_norm": 0.2754477858543396, "learning_rate": 1.955104648128844e-07, "loss": 0.2327, "step": 2140 }, { "epoch": 0.5697179350718468, "grad_norm": 0.2924790382385254, "learning_rate": 1.9550540881548957e-07, "loss": 0.2478, "step": 2141 }, { "epoch": 0.5699840340606706, "grad_norm": 0.27438199520111084, "learning_rate": 1.9550035003818027e-07, "loss": 0.2498, "step": 2142 }, { "epoch": 0.5702501330494945, "grad_norm": 0.2725508213043213, "learning_rate": 1.9549528848110376e-07, "loss": 0.2133, "step": 2143 }, { "epoch": 0.5705162320383182, "grad_norm": 0.27490413188934326, "learning_rate": 1.9549022414440737e-07, "loss": 0.2418, "step": 2144 }, { "epoch": 0.570782331027142, "grad_norm": 0.2996659576892853, "learning_rate": 1.9548515702823848e-07, "loss": 0.2534, "step": 2145 }, { "epoch": 0.571048430015966, "grad_norm": 0.29201796650886536, "learning_rate": 1.9548008713274462e-07, "loss": 0.2372, "step": 2146 }, { "epoch": 0.5713145290047897, "grad_norm": 0.3352966904640198, "learning_rate": 1.954750144580733e-07, "loss": 0.2808, "step": 2147 }, { "epoch": 0.5715806279936136, "grad_norm": 0.30906787514686584, "learning_rate": 1.9546993900437223e-07, "loss": 0.2457, "step": 2148 }, { "epoch": 0.5718467269824374, "grad_norm": 0.27815455198287964, "learning_rate": 1.954648607717891e-07, "loss": 0.2239, "step": 2149 }, { "epoch": 0.5721128259712613, "grad_norm": 0.2735860049724579, "learning_rate": 1.9545977976047174e-07, "loss": 0.2317, "step": 2150 }, { "epoch": 0.5723789249600851, "grad_norm": 0.29302069544792175, "learning_rate": 1.9545469597056807e-07, "loss": 0.2443, "step": 2151 }, { "epoch": 0.572645023948909, "grad_norm": 0.2763000726699829, "learning_rate": 1.9544960940222604e-07, "loss": 0.236, "step": 2152 }, { "epoch": 0.5729111229377328, "grad_norm": 0.28253835439682007, "learning_rate": 1.954445200555937e-07, "loss": 0.2471, "step": 2153 }, { "epoch": 0.5731772219265566, "grad_norm": 0.3093554973602295, "learning_rate": 1.9543942793081922e-07, "loss": 0.2442, "step": 2154 }, { "epoch": 0.5734433209153805, "grad_norm": 0.27405667304992676, "learning_rate": 1.9543433302805078e-07, "loss": 0.2413, "step": 2155 }, { "epoch": 0.5737094199042043, "grad_norm": 0.29057198762893677, "learning_rate": 1.9542923534743672e-07, "loss": 0.2498, "step": 2156 }, { "epoch": 0.5739755188930282, "grad_norm": 0.28862109780311584, "learning_rate": 1.954241348891254e-07, "loss": 0.2486, "step": 2157 }, { "epoch": 0.574241617881852, "grad_norm": 0.29712608456611633, "learning_rate": 1.9541903165326526e-07, "loss": 0.2576, "step": 2158 }, { "epoch": 0.5745077168706759, "grad_norm": 0.286009818315506, "learning_rate": 1.9541392564000488e-07, "loss": 0.2526, "step": 2159 }, { "epoch": 0.5747738158594997, "grad_norm": 0.2809487283229828, "learning_rate": 1.9540881684949286e-07, "loss": 0.2077, "step": 2160 }, { "epoch": 0.5750399148483236, "grad_norm": 0.2693142294883728, "learning_rate": 1.954037052818779e-07, "loss": 0.2373, "step": 2161 }, { "epoch": 0.5753060138371474, "grad_norm": 0.30882951617240906, "learning_rate": 1.9539859093730878e-07, "loss": 0.2366, "step": 2162 }, { "epoch": 0.5755721128259713, "grad_norm": 0.29796040058135986, "learning_rate": 1.953934738159344e-07, "loss": 0.2563, "step": 2163 }, { "epoch": 0.5758382118147951, "grad_norm": 0.3170446753501892, "learning_rate": 1.9538835391790368e-07, "loss": 0.2413, "step": 2164 }, { "epoch": 0.5761043108036189, "grad_norm": 0.2723487317562103, "learning_rate": 1.9538323124336566e-07, "loss": 0.2377, "step": 2165 }, { "epoch": 0.5763704097924428, "grad_norm": 0.2853761613368988, "learning_rate": 1.9537810579246942e-07, "loss": 0.243, "step": 2166 }, { "epoch": 0.5766365087812666, "grad_norm": 0.28565558791160583, "learning_rate": 1.9537297756536417e-07, "loss": 0.2354, "step": 2167 }, { "epoch": 0.5769026077700905, "grad_norm": 0.29729464650154114, "learning_rate": 1.9536784656219917e-07, "loss": 0.2496, "step": 2168 }, { "epoch": 0.5771687067589143, "grad_norm": 0.2875474989414215, "learning_rate": 1.953627127831238e-07, "loss": 0.246, "step": 2169 }, { "epoch": 0.5774348057477382, "grad_norm": 0.28562602400779724, "learning_rate": 1.9535757622828744e-07, "loss": 0.2425, "step": 2170 }, { "epoch": 0.577700904736562, "grad_norm": 0.303981751203537, "learning_rate": 1.9535243689783963e-07, "loss": 0.2651, "step": 2171 }, { "epoch": 0.5779670037253859, "grad_norm": 0.2997278571128845, "learning_rate": 1.9534729479193e-07, "loss": 0.2272, "step": 2172 }, { "epoch": 0.5782331027142097, "grad_norm": 0.2690514624118805, "learning_rate": 1.9534214991070813e-07, "loss": 0.2425, "step": 2173 }, { "epoch": 0.5784992017030335, "grad_norm": 0.2809203863143921, "learning_rate": 1.9533700225432387e-07, "loss": 0.2325, "step": 2174 }, { "epoch": 0.5787653006918574, "grad_norm": 0.33938685059547424, "learning_rate": 1.95331851822927e-07, "loss": 0.2661, "step": 2175 }, { "epoch": 0.5790313996806812, "grad_norm": 0.2629055380821228, "learning_rate": 1.9532669861666746e-07, "loss": 0.2095, "step": 2176 }, { "epoch": 0.5792974986695051, "grad_norm": 0.2702270746231079, "learning_rate": 1.9532154263569525e-07, "loss": 0.2151, "step": 2177 }, { "epoch": 0.5795635976583289, "grad_norm": 0.3032037615776062, "learning_rate": 1.9531638388016042e-07, "loss": 0.2529, "step": 2178 }, { "epoch": 0.5798296966471528, "grad_norm": 0.2765478491783142, "learning_rate": 1.9531122235021314e-07, "loss": 0.2407, "step": 2179 }, { "epoch": 0.5800957956359766, "grad_norm": 0.2970028221607208, "learning_rate": 1.9530605804600362e-07, "loss": 0.242, "step": 2180 }, { "epoch": 0.5803618946248005, "grad_norm": 0.33526086807250977, "learning_rate": 1.9530089096768226e-07, "loss": 0.2616, "step": 2181 }, { "epoch": 0.5806279936136243, "grad_norm": 0.28528401255607605, "learning_rate": 1.9529572111539937e-07, "loss": 0.2438, "step": 2182 }, { "epoch": 0.5808940926024481, "grad_norm": 0.2756010591983795, "learning_rate": 1.952905484893055e-07, "loss": 0.2411, "step": 2183 }, { "epoch": 0.581160191591272, "grad_norm": 0.28013381361961365, "learning_rate": 1.9528537308955118e-07, "loss": 0.2494, "step": 2184 }, { "epoch": 0.5814262905800958, "grad_norm": 0.3576555550098419, "learning_rate": 1.9528019491628705e-07, "loss": 0.2177, "step": 2185 }, { "epoch": 0.5816923895689197, "grad_norm": 0.33198297023773193, "learning_rate": 1.952750139696638e-07, "loss": 0.2576, "step": 2186 }, { "epoch": 0.5819584885577435, "grad_norm": 0.2893308699131012, "learning_rate": 1.9526983024983233e-07, "loss": 0.2367, "step": 2187 }, { "epoch": 0.5822245875465674, "grad_norm": 0.32533225417137146, "learning_rate": 1.9526464375694345e-07, "loss": 0.2517, "step": 2188 }, { "epoch": 0.5824906865353912, "grad_norm": 0.2723202109336853, "learning_rate": 1.952594544911481e-07, "loss": 0.2305, "step": 2189 }, { "epoch": 0.5827567855242151, "grad_norm": 0.2846347987651825, "learning_rate": 1.9525426245259742e-07, "loss": 0.2711, "step": 2190 }, { "epoch": 0.5830228845130389, "grad_norm": 0.26139160990715027, "learning_rate": 1.952490676414425e-07, "loss": 0.216, "step": 2191 }, { "epoch": 0.5832889835018626, "grad_norm": 0.2626197934150696, "learning_rate": 1.9524387005783449e-07, "loss": 0.2439, "step": 2192 }, { "epoch": 0.5835550824906865, "grad_norm": 0.26752883195877075, "learning_rate": 1.9523866970192472e-07, "loss": 0.2364, "step": 2193 }, { "epoch": 0.5838211814795103, "grad_norm": 0.29541441798210144, "learning_rate": 1.952334665738646e-07, "loss": 0.2357, "step": 2194 }, { "epoch": 0.5840872804683342, "grad_norm": 0.3008088171482086, "learning_rate": 1.9522826067380552e-07, "loss": 0.2532, "step": 2195 }, { "epoch": 0.584353379457158, "grad_norm": 0.2796821892261505, "learning_rate": 1.9522305200189903e-07, "loss": 0.2512, "step": 2196 }, { "epoch": 0.5846194784459819, "grad_norm": 0.3046518862247467, "learning_rate": 1.952178405582967e-07, "loss": 0.2334, "step": 2197 }, { "epoch": 0.5848855774348057, "grad_norm": 0.30393487215042114, "learning_rate": 1.952126263431503e-07, "loss": 0.2509, "step": 2198 }, { "epoch": 0.5851516764236296, "grad_norm": 0.31609460711479187, "learning_rate": 1.9520740935661157e-07, "loss": 0.2413, "step": 2199 }, { "epoch": 0.5854177754124534, "grad_norm": 0.2862743139266968, "learning_rate": 1.9520218959883234e-07, "loss": 0.2403, "step": 2200 }, { "epoch": 0.5856838744012772, "grad_norm": 0.27812132239341736, "learning_rate": 1.9519696706996456e-07, "loss": 0.2542, "step": 2201 }, { "epoch": 0.5859499733901011, "grad_norm": 0.2909002900123596, "learning_rate": 1.9519174177016024e-07, "loss": 0.2451, "step": 2202 }, { "epoch": 0.5862160723789249, "grad_norm": 0.3056434988975525, "learning_rate": 1.9518651369957152e-07, "loss": 0.2494, "step": 2203 }, { "epoch": 0.5864821713677488, "grad_norm": 0.2984057068824768, "learning_rate": 1.951812828583505e-07, "loss": 0.2457, "step": 2204 }, { "epoch": 0.5867482703565726, "grad_norm": 0.27784091234207153, "learning_rate": 1.9517604924664948e-07, "loss": 0.243, "step": 2205 }, { "epoch": 0.5870143693453965, "grad_norm": 0.2782236635684967, "learning_rate": 1.951708128646208e-07, "loss": 0.2311, "step": 2206 }, { "epoch": 0.5872804683342203, "grad_norm": 0.2852286696434021, "learning_rate": 1.9516557371241683e-07, "loss": 0.2362, "step": 2207 }, { "epoch": 0.5875465673230442, "grad_norm": 0.2739194333553314, "learning_rate": 1.9516033179019015e-07, "loss": 0.2496, "step": 2208 }, { "epoch": 0.587812666311868, "grad_norm": 0.2883507013320923, "learning_rate": 1.9515508709809327e-07, "loss": 0.2208, "step": 2209 }, { "epoch": 0.5880787653006918, "grad_norm": 0.3106815218925476, "learning_rate": 1.9514983963627886e-07, "loss": 0.2496, "step": 2210 }, { "epoch": 0.5883448642895157, "grad_norm": 0.29294726252555847, "learning_rate": 1.9514458940489968e-07, "loss": 0.2424, "step": 2211 }, { "epoch": 0.5886109632783395, "grad_norm": 0.33273935317993164, "learning_rate": 1.9513933640410855e-07, "loss": 0.2561, "step": 2212 }, { "epoch": 0.5888770622671634, "grad_norm": 0.30223727226257324, "learning_rate": 1.9513408063405835e-07, "loss": 0.2438, "step": 2213 }, { "epoch": 0.5891431612559872, "grad_norm": 0.29001423716545105, "learning_rate": 1.9512882209490212e-07, "loss": 0.2593, "step": 2214 }, { "epoch": 0.5894092602448111, "grad_norm": 0.2856733798980713, "learning_rate": 1.9512356078679285e-07, "loss": 0.2359, "step": 2215 }, { "epoch": 0.5896753592336349, "grad_norm": 0.26548680663108826, "learning_rate": 1.9511829670988368e-07, "loss": 0.2377, "step": 2216 }, { "epoch": 0.5899414582224588, "grad_norm": 0.2730826437473297, "learning_rate": 1.9511302986432789e-07, "loss": 0.2266, "step": 2217 }, { "epoch": 0.5902075572112826, "grad_norm": 0.2940472662448883, "learning_rate": 1.9510776025027879e-07, "loss": 0.2619, "step": 2218 }, { "epoch": 0.5904736562001064, "grad_norm": 0.28386303782463074, "learning_rate": 1.951024878678897e-07, "loss": 0.2398, "step": 2219 }, { "epoch": 0.5907397551889303, "grad_norm": 0.2713245451450348, "learning_rate": 1.9509721271731413e-07, "loss": 0.2281, "step": 2220 }, { "epoch": 0.5910058541777541, "grad_norm": 0.29659461975097656, "learning_rate": 1.950919347987056e-07, "loss": 0.2458, "step": 2221 }, { "epoch": 0.591271953166578, "grad_norm": 0.27727997303009033, "learning_rate": 1.9508665411221775e-07, "loss": 0.2325, "step": 2222 }, { "epoch": 0.5915380521554018, "grad_norm": 0.2856600880622864, "learning_rate": 1.9508137065800433e-07, "loss": 0.2343, "step": 2223 }, { "epoch": 0.5918041511442257, "grad_norm": 0.29898273944854736, "learning_rate": 1.9507608443621906e-07, "loss": 0.2333, "step": 2224 }, { "epoch": 0.5920702501330495, "grad_norm": 0.29646703600883484, "learning_rate": 1.9507079544701583e-07, "loss": 0.2333, "step": 2225 }, { "epoch": 0.5923363491218734, "grad_norm": 0.26739534735679626, "learning_rate": 1.950655036905486e-07, "loss": 0.2271, "step": 2226 }, { "epoch": 0.5926024481106972, "grad_norm": 0.29336434602737427, "learning_rate": 1.9506020916697136e-07, "loss": 0.2483, "step": 2227 }, { "epoch": 0.592868547099521, "grad_norm": 0.28657659888267517, "learning_rate": 1.9505491187643827e-07, "loss": 0.2136, "step": 2228 }, { "epoch": 0.5931346460883449, "grad_norm": 0.2925117313861847, "learning_rate": 1.9504961181910353e-07, "loss": 0.2266, "step": 2229 }, { "epoch": 0.5934007450771687, "grad_norm": 0.27177101373672485, "learning_rate": 1.950443089951214e-07, "loss": 0.2333, "step": 2230 }, { "epoch": 0.5936668440659926, "grad_norm": 0.27691829204559326, "learning_rate": 1.9503900340464616e-07, "loss": 0.2472, "step": 2231 }, { "epoch": 0.5939329430548164, "grad_norm": 0.27180683612823486, "learning_rate": 1.9503369504783232e-07, "loss": 0.239, "step": 2232 }, { "epoch": 0.5941990420436403, "grad_norm": 0.2744719386100769, "learning_rate": 1.9502838392483436e-07, "loss": 0.2257, "step": 2233 }, { "epoch": 0.5944651410324641, "grad_norm": 0.284546434879303, "learning_rate": 1.9502307003580692e-07, "loss": 0.2237, "step": 2234 }, { "epoch": 0.594731240021288, "grad_norm": 0.2632821202278137, "learning_rate": 1.9501775338090459e-07, "loss": 0.2442, "step": 2235 }, { "epoch": 0.5949973390101118, "grad_norm": 0.28401800990104675, "learning_rate": 1.950124339602822e-07, "loss": 0.2428, "step": 2236 }, { "epoch": 0.5952634379989356, "grad_norm": 0.2884228527545929, "learning_rate": 1.9500711177409452e-07, "loss": 0.233, "step": 2237 }, { "epoch": 0.5955295369877595, "grad_norm": 0.26357805728912354, "learning_rate": 1.9500178682249653e-07, "loss": 0.2255, "step": 2238 }, { "epoch": 0.5957956359765832, "grad_norm": 0.2848595380783081, "learning_rate": 1.949964591056432e-07, "loss": 0.227, "step": 2239 }, { "epoch": 0.5960617349654072, "grad_norm": 0.2949604392051697, "learning_rate": 1.949911286236896e-07, "loss": 0.2615, "step": 2240 }, { "epoch": 0.596327833954231, "grad_norm": 0.27282771468162537, "learning_rate": 1.9498579537679087e-07, "loss": 0.2408, "step": 2241 }, { "epoch": 0.5965939329430548, "grad_norm": 0.33361396193504333, "learning_rate": 1.9498045936510226e-07, "loss": 0.2561, "step": 2242 }, { "epoch": 0.5968600319318786, "grad_norm": 0.2526870667934418, "learning_rate": 1.949751205887791e-07, "loss": 0.2218, "step": 2243 }, { "epoch": 0.5971261309207025, "grad_norm": 0.27579164505004883, "learning_rate": 1.949697790479768e-07, "loss": 0.2461, "step": 2244 }, { "epoch": 0.5973922299095263, "grad_norm": 0.2767309844493866, "learning_rate": 1.949644347428508e-07, "loss": 0.2467, "step": 2245 }, { "epoch": 0.5976583288983501, "grad_norm": 0.29284101724624634, "learning_rate": 1.949590876735567e-07, "loss": 0.2351, "step": 2246 }, { "epoch": 0.597924427887174, "grad_norm": 0.29525288939476013, "learning_rate": 1.9495373784025008e-07, "loss": 0.2462, "step": 2247 }, { "epoch": 0.5981905268759978, "grad_norm": 0.276821494102478, "learning_rate": 1.949483852430867e-07, "loss": 0.2301, "step": 2248 }, { "epoch": 0.5984566258648217, "grad_norm": 0.2761668860912323, "learning_rate": 1.9494302988222238e-07, "loss": 0.2324, "step": 2249 }, { "epoch": 0.5987227248536455, "grad_norm": 0.29405468702316284, "learning_rate": 1.9493767175781297e-07, "loss": 0.2461, "step": 2250 }, { "epoch": 0.5989888238424694, "grad_norm": 0.2715296149253845, "learning_rate": 1.9493231087001443e-07, "loss": 0.2317, "step": 2251 }, { "epoch": 0.5992549228312932, "grad_norm": 0.267278254032135, "learning_rate": 1.9492694721898282e-07, "loss": 0.2432, "step": 2252 }, { "epoch": 0.5995210218201171, "grad_norm": 0.2944287061691284, "learning_rate": 1.9492158080487425e-07, "loss": 0.2636, "step": 2253 }, { "epoch": 0.5997871208089409, "grad_norm": 0.2878116965293884, "learning_rate": 1.949162116278449e-07, "loss": 0.2404, "step": 2254 }, { "epoch": 0.6000532197977647, "grad_norm": 0.2930990755558014, "learning_rate": 1.949108396880511e-07, "loss": 0.2229, "step": 2255 }, { "epoch": 0.6003193187865886, "grad_norm": 0.28405094146728516, "learning_rate": 1.949054649856492e-07, "loss": 0.2284, "step": 2256 }, { "epoch": 0.6005854177754124, "grad_norm": 0.3715607225894928, "learning_rate": 1.9490008752079562e-07, "loss": 0.2199, "step": 2257 }, { "epoch": 0.6008515167642363, "grad_norm": 0.32977935671806335, "learning_rate": 1.948947072936469e-07, "loss": 0.2453, "step": 2258 }, { "epoch": 0.6011176157530601, "grad_norm": 0.2903447151184082, "learning_rate": 1.9488932430435961e-07, "loss": 0.2548, "step": 2259 }, { "epoch": 0.601383714741884, "grad_norm": 0.25944119691848755, "learning_rate": 1.9488393855309054e-07, "loss": 0.2245, "step": 2260 }, { "epoch": 0.6016498137307078, "grad_norm": 0.29054179787635803, "learning_rate": 1.948785500399963e-07, "loss": 0.2396, "step": 2261 }, { "epoch": 0.6019159127195317, "grad_norm": 0.2808254063129425, "learning_rate": 1.9487315876523388e-07, "loss": 0.2453, "step": 2262 }, { "epoch": 0.6021820117083555, "grad_norm": 0.2994300127029419, "learning_rate": 1.9486776472896014e-07, "loss": 0.2174, "step": 2263 }, { "epoch": 0.6024481106971793, "grad_norm": 0.3002674877643585, "learning_rate": 1.9486236793133207e-07, "loss": 0.2259, "step": 2264 }, { "epoch": 0.6027142096860032, "grad_norm": 0.3065643608570099, "learning_rate": 1.9485696837250677e-07, "loss": 0.2483, "step": 2265 }, { "epoch": 0.602980308674827, "grad_norm": 0.3012641966342926, "learning_rate": 1.9485156605264145e-07, "loss": 0.2304, "step": 2266 }, { "epoch": 0.6032464076636509, "grad_norm": 0.2829281985759735, "learning_rate": 1.9484616097189327e-07, "loss": 0.2452, "step": 2267 }, { "epoch": 0.6035125066524747, "grad_norm": 0.27706456184387207, "learning_rate": 1.9484075313041964e-07, "loss": 0.2281, "step": 2268 }, { "epoch": 0.6037786056412986, "grad_norm": 0.27711689472198486, "learning_rate": 1.9483534252837794e-07, "loss": 0.2513, "step": 2269 }, { "epoch": 0.6040447046301224, "grad_norm": 0.2745124399662018, "learning_rate": 1.9482992916592567e-07, "loss": 0.2183, "step": 2270 }, { "epoch": 0.6043108036189463, "grad_norm": 0.287507027387619, "learning_rate": 1.9482451304322037e-07, "loss": 0.2443, "step": 2271 }, { "epoch": 0.6045769026077701, "grad_norm": 0.27677640318870544, "learning_rate": 1.948190941604197e-07, "loss": 0.2294, "step": 2272 }, { "epoch": 0.604843001596594, "grad_norm": 0.27104082703590393, "learning_rate": 1.948136725176814e-07, "loss": 0.2371, "step": 2273 }, { "epoch": 0.6051091005854178, "grad_norm": 0.2854343354701996, "learning_rate": 1.9480824811516324e-07, "loss": 0.2299, "step": 2274 }, { "epoch": 0.6053751995742416, "grad_norm": 0.3534744083881378, "learning_rate": 1.948028209530232e-07, "loss": 0.2443, "step": 2275 }, { "epoch": 0.6056412985630655, "grad_norm": 0.2814856171607971, "learning_rate": 1.9479739103141916e-07, "loss": 0.2394, "step": 2276 }, { "epoch": 0.6059073975518893, "grad_norm": 0.28291216492652893, "learning_rate": 1.9479195835050922e-07, "loss": 0.2229, "step": 2277 }, { "epoch": 0.6061734965407132, "grad_norm": 0.2785315215587616, "learning_rate": 1.947865229104515e-07, "loss": 0.2246, "step": 2278 }, { "epoch": 0.606439595529537, "grad_norm": 0.2939268946647644, "learning_rate": 1.947810847114042e-07, "loss": 0.2346, "step": 2279 }, { "epoch": 0.6067056945183609, "grad_norm": 0.28588441014289856, "learning_rate": 1.947756437535256e-07, "loss": 0.2231, "step": 2280 }, { "epoch": 0.6069717935071847, "grad_norm": 0.26103487610816956, "learning_rate": 1.9477020003697413e-07, "loss": 0.2273, "step": 2281 }, { "epoch": 0.6072378924960086, "grad_norm": 0.2964467406272888, "learning_rate": 1.9476475356190817e-07, "loss": 0.2466, "step": 2282 }, { "epoch": 0.6075039914848324, "grad_norm": 0.29742443561553955, "learning_rate": 1.9475930432848633e-07, "loss": 0.2386, "step": 2283 }, { "epoch": 0.6077700904736562, "grad_norm": 0.2812829613685608, "learning_rate": 1.9475385233686716e-07, "loss": 0.2486, "step": 2284 }, { "epoch": 0.6080361894624801, "grad_norm": 0.26130804419517517, "learning_rate": 1.9474839758720938e-07, "loss": 0.2272, "step": 2285 }, { "epoch": 0.6083022884513039, "grad_norm": 0.3372001349925995, "learning_rate": 1.9474294007967177e-07, "loss": 0.2477, "step": 2286 }, { "epoch": 0.6085683874401278, "grad_norm": 0.29380205273628235, "learning_rate": 1.9473747981441312e-07, "loss": 0.2478, "step": 2287 }, { "epoch": 0.6088344864289515, "grad_norm": 0.2841030955314636, "learning_rate": 1.9473201679159245e-07, "loss": 0.25, "step": 2288 }, { "epoch": 0.6091005854177755, "grad_norm": 0.2891140878200531, "learning_rate": 1.9472655101136875e-07, "loss": 0.2227, "step": 2289 }, { "epoch": 0.6093666844065992, "grad_norm": 0.29285910725593567, "learning_rate": 1.9472108247390108e-07, "loss": 0.2562, "step": 2290 }, { "epoch": 0.6096327833954232, "grad_norm": 0.2705739140510559, "learning_rate": 1.9471561117934866e-07, "loss": 0.2232, "step": 2291 }, { "epoch": 0.6098988823842469, "grad_norm": 0.30570080876350403, "learning_rate": 1.947101371278707e-07, "loss": 0.2378, "step": 2292 }, { "epoch": 0.6101649813730707, "grad_norm": 0.2682815492153168, "learning_rate": 1.947046603196266e-07, "loss": 0.2227, "step": 2293 }, { "epoch": 0.6104310803618946, "grad_norm": 0.2790854275226593, "learning_rate": 1.946991807547757e-07, "loss": 0.2505, "step": 2294 }, { "epoch": 0.6106971793507184, "grad_norm": 0.26922672986984253, "learning_rate": 1.9469369843347758e-07, "loss": 0.2333, "step": 2295 }, { "epoch": 0.6109632783395423, "grad_norm": 0.29552313685417175, "learning_rate": 1.9468821335589173e-07, "loss": 0.2609, "step": 2296 }, { "epoch": 0.6112293773283661, "grad_norm": 0.2992557883262634, "learning_rate": 1.9468272552217787e-07, "loss": 0.223, "step": 2297 }, { "epoch": 0.61149547631719, "grad_norm": 0.28786343336105347, "learning_rate": 1.9467723493249568e-07, "loss": 0.2495, "step": 2298 }, { "epoch": 0.6117615753060138, "grad_norm": 0.2663189768791199, "learning_rate": 1.9467174158700503e-07, "loss": 0.2396, "step": 2299 }, { "epoch": 0.6120276742948377, "grad_norm": 0.3107140064239502, "learning_rate": 1.946662454858658e-07, "loss": 0.2546, "step": 2300 }, { "epoch": 0.6122937732836615, "grad_norm": 0.31708601117134094, "learning_rate": 1.9466074662923794e-07, "loss": 0.245, "step": 2301 }, { "epoch": 0.6125598722724853, "grad_norm": 0.2912907004356384, "learning_rate": 1.9465524501728154e-07, "loss": 0.2534, "step": 2302 }, { "epoch": 0.6128259712613092, "grad_norm": 0.2667296528816223, "learning_rate": 1.946497406501567e-07, "loss": 0.2414, "step": 2303 }, { "epoch": 0.613092070250133, "grad_norm": 0.2719575762748718, "learning_rate": 1.946442335280237e-07, "loss": 0.2414, "step": 2304 }, { "epoch": 0.6133581692389569, "grad_norm": 0.286725789308548, "learning_rate": 1.946387236510428e-07, "loss": 0.2369, "step": 2305 }, { "epoch": 0.6136242682277807, "grad_norm": 0.3047254979610443, "learning_rate": 1.946332110193744e-07, "loss": 0.2512, "step": 2306 }, { "epoch": 0.6138903672166046, "grad_norm": 0.29309776425361633, "learning_rate": 1.9462769563317888e-07, "loss": 0.2473, "step": 2307 }, { "epoch": 0.6141564662054284, "grad_norm": 0.29525285959243774, "learning_rate": 1.9462217749261688e-07, "loss": 0.2672, "step": 2308 }, { "epoch": 0.6144225651942523, "grad_norm": 0.2944317162036896, "learning_rate": 1.9461665659784898e-07, "loss": 0.2426, "step": 2309 }, { "epoch": 0.6146886641830761, "grad_norm": 0.2968243658542633, "learning_rate": 1.9461113294903585e-07, "loss": 0.2431, "step": 2310 }, { "epoch": 0.6149547631718999, "grad_norm": 0.3006109893321991, "learning_rate": 1.9460560654633836e-07, "loss": 0.2682, "step": 2311 }, { "epoch": 0.6152208621607238, "grad_norm": 0.3127291202545166, "learning_rate": 1.9460007738991724e-07, "loss": 0.2498, "step": 2312 }, { "epoch": 0.6154869611495476, "grad_norm": 0.2960973083972931, "learning_rate": 1.945945454799335e-07, "loss": 0.2476, "step": 2313 }, { "epoch": 0.6157530601383715, "grad_norm": 0.2651980519294739, "learning_rate": 1.9458901081654818e-07, "loss": 0.2248, "step": 2314 }, { "epoch": 0.6160191591271953, "grad_norm": 0.295604407787323, "learning_rate": 1.9458347339992233e-07, "loss": 0.2452, "step": 2315 }, { "epoch": 0.6162852581160192, "grad_norm": 0.2713682949542999, "learning_rate": 1.9457793323021715e-07, "loss": 0.2474, "step": 2316 }, { "epoch": 0.616551357104843, "grad_norm": 0.2732336223125458, "learning_rate": 1.9457239030759394e-07, "loss": 0.2219, "step": 2317 }, { "epoch": 0.6168174560936669, "grad_norm": 0.2896391451358795, "learning_rate": 1.9456684463221394e-07, "loss": 0.2553, "step": 2318 }, { "epoch": 0.6170835550824907, "grad_norm": 0.3168598413467407, "learning_rate": 1.9456129620423868e-07, "loss": 0.2504, "step": 2319 }, { "epoch": 0.6173496540713145, "grad_norm": 0.3126926124095917, "learning_rate": 1.9455574502382964e-07, "loss": 0.2413, "step": 2320 }, { "epoch": 0.6176157530601384, "grad_norm": 0.31231001019477844, "learning_rate": 1.9455019109114833e-07, "loss": 0.2591, "step": 2321 }, { "epoch": 0.6178818520489622, "grad_norm": 0.30024632811546326, "learning_rate": 1.9454463440635645e-07, "loss": 0.2371, "step": 2322 }, { "epoch": 0.6181479510377861, "grad_norm": 0.26726648211479187, "learning_rate": 1.9453907496961576e-07, "loss": 0.2242, "step": 2323 }, { "epoch": 0.6184140500266099, "grad_norm": 0.2722643315792084, "learning_rate": 1.9453351278108806e-07, "loss": 0.2426, "step": 2324 }, { "epoch": 0.6186801490154338, "grad_norm": 0.2712058424949646, "learning_rate": 1.9452794784093526e-07, "loss": 0.238, "step": 2325 }, { "epoch": 0.6189462480042576, "grad_norm": 0.2764441967010498, "learning_rate": 1.9452238014931934e-07, "loss": 0.2414, "step": 2326 }, { "epoch": 0.6192123469930815, "grad_norm": 0.28469130396842957, "learning_rate": 1.9451680970640236e-07, "loss": 0.2325, "step": 2327 }, { "epoch": 0.6194784459819053, "grad_norm": 0.30456605553627014, "learning_rate": 1.9451123651234646e-07, "loss": 0.2333, "step": 2328 }, { "epoch": 0.6197445449707291, "grad_norm": 0.36237189173698425, "learning_rate": 1.9450566056731383e-07, "loss": 0.277, "step": 2329 }, { "epoch": 0.620010643959553, "grad_norm": 0.2765537202358246, "learning_rate": 1.9450008187146683e-07, "loss": 0.2323, "step": 2330 }, { "epoch": 0.6202767429483768, "grad_norm": 0.25489869713783264, "learning_rate": 1.9449450042496782e-07, "loss": 0.204, "step": 2331 }, { "epoch": 0.6205428419372007, "grad_norm": 0.27451685070991516, "learning_rate": 1.9448891622797924e-07, "loss": 0.2164, "step": 2332 }, { "epoch": 0.6208089409260245, "grad_norm": 0.26994070410728455, "learning_rate": 1.9448332928066366e-07, "loss": 0.2279, "step": 2333 }, { "epoch": 0.6210750399148484, "grad_norm": 0.290690153837204, "learning_rate": 1.9447773958318368e-07, "loss": 0.2341, "step": 2334 }, { "epoch": 0.6213411389036722, "grad_norm": 0.2785457968711853, "learning_rate": 1.9447214713570204e-07, "loss": 0.233, "step": 2335 }, { "epoch": 0.6216072378924961, "grad_norm": 0.2942511737346649, "learning_rate": 1.9446655193838143e-07, "loss": 0.2458, "step": 2336 }, { "epoch": 0.6218733368813198, "grad_norm": 0.29151466488838196, "learning_rate": 1.9446095399138483e-07, "loss": 0.2468, "step": 2337 }, { "epoch": 0.6221394358701436, "grad_norm": 0.2785976231098175, "learning_rate": 1.9445535329487508e-07, "loss": 0.2387, "step": 2338 }, { "epoch": 0.6224055348589675, "grad_norm": 0.31034305691719055, "learning_rate": 1.9444974984901528e-07, "loss": 0.2429, "step": 2339 }, { "epoch": 0.6226716338477913, "grad_norm": 0.28065910935401917, "learning_rate": 1.944441436539685e-07, "loss": 0.2427, "step": 2340 }, { "epoch": 0.6229377328366152, "grad_norm": 0.27513235807418823, "learning_rate": 1.9443853470989788e-07, "loss": 0.2304, "step": 2341 }, { "epoch": 0.623203831825439, "grad_norm": 0.27951428294181824, "learning_rate": 1.9443292301696677e-07, "loss": 0.2179, "step": 2342 }, { "epoch": 0.6234699308142629, "grad_norm": 0.2843054234981537, "learning_rate": 1.9442730857533842e-07, "loss": 0.2516, "step": 2343 }, { "epoch": 0.6237360298030867, "grad_norm": 0.3029267489910126, "learning_rate": 1.9442169138517636e-07, "loss": 0.2623, "step": 2344 }, { "epoch": 0.6240021287919106, "grad_norm": 0.3028777837753296, "learning_rate": 1.9441607144664397e-07, "loss": 0.2271, "step": 2345 }, { "epoch": 0.6242682277807344, "grad_norm": 0.2822102904319763, "learning_rate": 1.9441044875990488e-07, "loss": 0.2328, "step": 2346 }, { "epoch": 0.6245343267695582, "grad_norm": 0.2591930329799652, "learning_rate": 1.944048233251228e-07, "loss": 0.2013, "step": 2347 }, { "epoch": 0.6248004257583821, "grad_norm": 0.2736670970916748, "learning_rate": 1.9439919514246142e-07, "loss": 0.2296, "step": 2348 }, { "epoch": 0.6250665247472059, "grad_norm": 0.2953239679336548, "learning_rate": 1.9439356421208455e-07, "loss": 0.2334, "step": 2349 }, { "epoch": 0.6253326237360298, "grad_norm": 0.29043060541152954, "learning_rate": 1.9438793053415617e-07, "loss": 0.2469, "step": 2350 }, { "epoch": 0.6255987227248536, "grad_norm": 0.2717513144016266, "learning_rate": 1.9438229410884014e-07, "loss": 0.2418, "step": 2351 }, { "epoch": 0.6258648217136775, "grad_norm": 0.2594004273414612, "learning_rate": 1.9437665493630062e-07, "loss": 0.2344, "step": 2352 }, { "epoch": 0.6261309207025013, "grad_norm": 0.306568443775177, "learning_rate": 1.9437101301670172e-07, "loss": 0.2649, "step": 2353 }, { "epoch": 0.6263970196913252, "grad_norm": 0.2798781991004944, "learning_rate": 1.943653683502077e-07, "loss": 0.247, "step": 2354 }, { "epoch": 0.626663118680149, "grad_norm": 0.26625385880470276, "learning_rate": 1.9435972093698278e-07, "loss": 0.2434, "step": 2355 }, { "epoch": 0.6269292176689728, "grad_norm": 0.28379422426223755, "learning_rate": 1.9435407077719139e-07, "loss": 0.271, "step": 2356 }, { "epoch": 0.6271953166577967, "grad_norm": 0.2813240885734558, "learning_rate": 1.9434841787099802e-07, "loss": 0.2374, "step": 2357 }, { "epoch": 0.6274614156466205, "grad_norm": 0.28330886363983154, "learning_rate": 1.9434276221856716e-07, "loss": 0.2345, "step": 2358 }, { "epoch": 0.6277275146354444, "grad_norm": 0.2923918068408966, "learning_rate": 1.9433710382006345e-07, "loss": 0.2394, "step": 2359 }, { "epoch": 0.6279936136242682, "grad_norm": 0.28561028838157654, "learning_rate": 1.943314426756516e-07, "loss": 0.2458, "step": 2360 }, { "epoch": 0.6282597126130921, "grad_norm": 0.28272899985313416, "learning_rate": 1.9432577878549635e-07, "loss": 0.2397, "step": 2361 }, { "epoch": 0.6285258116019159, "grad_norm": 0.34897756576538086, "learning_rate": 1.9432011214976262e-07, "loss": 0.2726, "step": 2362 }, { "epoch": 0.6287919105907398, "grad_norm": 0.2723844349384308, "learning_rate": 1.9431444276861532e-07, "loss": 0.2298, "step": 2363 }, { "epoch": 0.6290580095795636, "grad_norm": 0.2828366160392761, "learning_rate": 1.9430877064221946e-07, "loss": 0.2219, "step": 2364 }, { "epoch": 0.6293241085683874, "grad_norm": 0.27481338381767273, "learning_rate": 1.9430309577074018e-07, "loss": 0.2227, "step": 2365 }, { "epoch": 0.6295902075572113, "grad_norm": 0.2847767770290375, "learning_rate": 1.9429741815434263e-07, "loss": 0.2614, "step": 2366 }, { "epoch": 0.6298563065460351, "grad_norm": 0.29977285861968994, "learning_rate": 1.9429173779319207e-07, "loss": 0.2379, "step": 2367 }, { "epoch": 0.630122405534859, "grad_norm": 0.2721654176712036, "learning_rate": 1.9428605468745388e-07, "loss": 0.2411, "step": 2368 }, { "epoch": 0.6303885045236828, "grad_norm": 0.27812904119491577, "learning_rate": 1.9428036883729342e-07, "loss": 0.2433, "step": 2369 }, { "epoch": 0.6306546035125067, "grad_norm": 0.2653452455997467, "learning_rate": 1.942746802428762e-07, "loss": 0.2238, "step": 2370 }, { "epoch": 0.6309207025013305, "grad_norm": 0.2674837112426758, "learning_rate": 1.9426898890436788e-07, "loss": 0.2375, "step": 2371 }, { "epoch": 0.6311868014901544, "grad_norm": 0.2877222001552582, "learning_rate": 1.9426329482193402e-07, "loss": 0.2614, "step": 2372 }, { "epoch": 0.6314529004789782, "grad_norm": 0.2794824242591858, "learning_rate": 1.9425759799574042e-07, "loss": 0.2443, "step": 2373 }, { "epoch": 0.631718999467802, "grad_norm": 0.2820330262184143, "learning_rate": 1.9425189842595287e-07, "loss": 0.2391, "step": 2374 }, { "epoch": 0.6319850984566259, "grad_norm": 0.2713216543197632, "learning_rate": 1.9424619611273725e-07, "loss": 0.2379, "step": 2375 }, { "epoch": 0.6322511974454497, "grad_norm": 0.31044602394104004, "learning_rate": 1.9424049105625962e-07, "loss": 0.2513, "step": 2376 }, { "epoch": 0.6325172964342736, "grad_norm": 0.2851547598838806, "learning_rate": 1.9423478325668595e-07, "loss": 0.2428, "step": 2377 }, { "epoch": 0.6327833954230974, "grad_norm": 0.2584555447101593, "learning_rate": 1.9422907271418243e-07, "loss": 0.2173, "step": 2378 }, { "epoch": 0.6330494944119213, "grad_norm": 0.31202125549316406, "learning_rate": 1.9422335942891529e-07, "loss": 0.2411, "step": 2379 }, { "epoch": 0.6333155934007451, "grad_norm": 0.2824191153049469, "learning_rate": 1.9421764340105075e-07, "loss": 0.229, "step": 2380 }, { "epoch": 0.633581692389569, "grad_norm": 0.274284303188324, "learning_rate": 1.942119246307553e-07, "loss": 0.2283, "step": 2381 }, { "epoch": 0.6338477913783928, "grad_norm": 0.28543606400489807, "learning_rate": 1.9420620311819528e-07, "loss": 0.2485, "step": 2382 }, { "epoch": 0.6341138903672167, "grad_norm": 0.2745669484138489, "learning_rate": 1.9420047886353738e-07, "loss": 0.2394, "step": 2383 }, { "epoch": 0.6343799893560405, "grad_norm": 0.2881491780281067, "learning_rate": 1.9419475186694804e-07, "loss": 0.2342, "step": 2384 }, { "epoch": 0.6346460883448642, "grad_norm": 0.3087506592273712, "learning_rate": 1.941890221285941e-07, "loss": 0.2268, "step": 2385 }, { "epoch": 0.6349121873336881, "grad_norm": 0.2882654070854187, "learning_rate": 1.9418328964864227e-07, "loss": 0.2338, "step": 2386 }, { "epoch": 0.6351782863225119, "grad_norm": 0.2805696427822113, "learning_rate": 1.9417755442725944e-07, "loss": 0.2524, "step": 2387 }, { "epoch": 0.6354443853113358, "grad_norm": 0.2870869040489197, "learning_rate": 1.941718164646125e-07, "loss": 0.2437, "step": 2388 }, { "epoch": 0.6357104843001596, "grad_norm": 0.26999416947364807, "learning_rate": 1.9416607576086852e-07, "loss": 0.22, "step": 2389 }, { "epoch": 0.6359765832889835, "grad_norm": 0.2626747488975525, "learning_rate": 1.9416033231619457e-07, "loss": 0.23, "step": 2390 }, { "epoch": 0.6362426822778073, "grad_norm": 0.27663081884384155, "learning_rate": 1.9415458613075781e-07, "loss": 0.2476, "step": 2391 }, { "epoch": 0.6365087812666312, "grad_norm": 0.2889481782913208, "learning_rate": 1.9414883720472555e-07, "loss": 0.2352, "step": 2392 }, { "epoch": 0.636774880255455, "grad_norm": 0.2568296492099762, "learning_rate": 1.941430855382651e-07, "loss": 0.2167, "step": 2393 }, { "epoch": 0.6370409792442788, "grad_norm": 0.2984501123428345, "learning_rate": 1.9413733113154382e-07, "loss": 0.2333, "step": 2394 }, { "epoch": 0.6373070782331027, "grad_norm": 0.3116039037704468, "learning_rate": 1.941315739847293e-07, "loss": 0.2304, "step": 2395 }, { "epoch": 0.6375731772219265, "grad_norm": 0.30217355489730835, "learning_rate": 1.9412581409798907e-07, "loss": 0.2456, "step": 2396 }, { "epoch": 0.6378392762107504, "grad_norm": 0.2965656816959381, "learning_rate": 1.9412005147149077e-07, "loss": 0.2587, "step": 2397 }, { "epoch": 0.6381053751995742, "grad_norm": 0.2906595468521118, "learning_rate": 1.9411428610540215e-07, "loss": 0.2505, "step": 2398 }, { "epoch": 0.6383714741883981, "grad_norm": 0.28499794006347656, "learning_rate": 1.9410851799989103e-07, "loss": 0.2371, "step": 2399 }, { "epoch": 0.6386375731772219, "grad_norm": 0.27395570278167725, "learning_rate": 1.941027471551253e-07, "loss": 0.2323, "step": 2400 }, { "epoch": 0.6389036721660458, "grad_norm": 0.32307401299476624, "learning_rate": 1.9409697357127296e-07, "loss": 0.2116, "step": 2401 }, { "epoch": 0.6391697711548696, "grad_norm": 0.25918883085250854, "learning_rate": 1.94091197248502e-07, "loss": 0.2096, "step": 2402 }, { "epoch": 0.6394358701436934, "grad_norm": 0.3081633746623993, "learning_rate": 1.940854181869806e-07, "loss": 0.2467, "step": 2403 }, { "epoch": 0.6397019691325173, "grad_norm": 0.26273635029792786, "learning_rate": 1.9407963638687698e-07, "loss": 0.203, "step": 2404 }, { "epoch": 0.6399680681213411, "grad_norm": 0.2922816574573517, "learning_rate": 1.9407385184835944e-07, "loss": 0.2367, "step": 2405 }, { "epoch": 0.640234167110165, "grad_norm": 0.2883230149745941, "learning_rate": 1.9406806457159632e-07, "loss": 0.2515, "step": 2406 }, { "epoch": 0.6405002660989888, "grad_norm": 0.2780055105686188, "learning_rate": 1.9406227455675608e-07, "loss": 0.2379, "step": 2407 }, { "epoch": 0.6407663650878127, "grad_norm": 0.26683545112609863, "learning_rate": 1.9405648180400725e-07, "loss": 0.2317, "step": 2408 }, { "epoch": 0.6410324640766365, "grad_norm": 0.27594441175460815, "learning_rate": 1.9405068631351847e-07, "loss": 0.2399, "step": 2409 }, { "epoch": 0.6412985630654604, "grad_norm": 0.26844486594200134, "learning_rate": 1.9404488808545838e-07, "loss": 0.2377, "step": 2410 }, { "epoch": 0.6415646620542842, "grad_norm": 0.2867467403411865, "learning_rate": 1.940390871199958e-07, "loss": 0.2375, "step": 2411 }, { "epoch": 0.641830761043108, "grad_norm": 0.27414870262145996, "learning_rate": 1.9403328341729956e-07, "loss": 0.2184, "step": 2412 }, { "epoch": 0.6420968600319319, "grad_norm": 0.2721431255340576, "learning_rate": 1.940274769775386e-07, "loss": 0.2301, "step": 2413 }, { "epoch": 0.6423629590207557, "grad_norm": 0.2917526960372925, "learning_rate": 1.9402166780088193e-07, "loss": 0.2344, "step": 2414 }, { "epoch": 0.6426290580095796, "grad_norm": 0.2860489785671234, "learning_rate": 1.9401585588749863e-07, "loss": 0.2394, "step": 2415 }, { "epoch": 0.6428951569984034, "grad_norm": 0.2821359932422638, "learning_rate": 1.9401004123755787e-07, "loss": 0.2597, "step": 2416 }, { "epoch": 0.6431612559872273, "grad_norm": 0.3087080419063568, "learning_rate": 1.9400422385122892e-07, "loss": 0.2531, "step": 2417 }, { "epoch": 0.6434273549760511, "grad_norm": 0.3052869141101837, "learning_rate": 1.9399840372868108e-07, "loss": 0.2538, "step": 2418 }, { "epoch": 0.643693453964875, "grad_norm": 0.2861783802509308, "learning_rate": 1.9399258087008378e-07, "loss": 0.2599, "step": 2419 }, { "epoch": 0.6439595529536988, "grad_norm": 0.30051568150520325, "learning_rate": 1.9398675527560648e-07, "loss": 0.223, "step": 2420 }, { "epoch": 0.6442256519425226, "grad_norm": 0.29773133993148804, "learning_rate": 1.9398092694541878e-07, "loss": 0.2552, "step": 2421 }, { "epoch": 0.6444917509313465, "grad_norm": 0.3097688555717468, "learning_rate": 1.9397509587969032e-07, "loss": 0.2399, "step": 2422 }, { "epoch": 0.6447578499201703, "grad_norm": 0.2972260117530823, "learning_rate": 1.9396926207859085e-07, "loss": 0.223, "step": 2423 }, { "epoch": 0.6450239489089942, "grad_norm": 0.29610589146614075, "learning_rate": 1.9396342554229011e-07, "loss": 0.2414, "step": 2424 }, { "epoch": 0.645290047897818, "grad_norm": 0.2943271994590759, "learning_rate": 1.93957586270958e-07, "loss": 0.2506, "step": 2425 }, { "epoch": 0.6455561468866419, "grad_norm": 0.27931687235832214, "learning_rate": 1.9395174426476457e-07, "loss": 0.2307, "step": 2426 }, { "epoch": 0.6458222458754657, "grad_norm": 0.28306126594543457, "learning_rate": 1.939458995238798e-07, "loss": 0.2519, "step": 2427 }, { "epoch": 0.6460883448642896, "grad_norm": 0.29375576972961426, "learning_rate": 1.939400520484738e-07, "loss": 0.2539, "step": 2428 }, { "epoch": 0.6463544438531134, "grad_norm": 0.28186386823654175, "learning_rate": 1.9393420183871681e-07, "loss": 0.2173, "step": 2429 }, { "epoch": 0.6466205428419372, "grad_norm": 0.2636243999004364, "learning_rate": 1.939283488947791e-07, "loss": 0.2298, "step": 2430 }, { "epoch": 0.6468866418307611, "grad_norm": 0.28095391392707825, "learning_rate": 1.9392249321683103e-07, "loss": 0.2359, "step": 2431 }, { "epoch": 0.6471527408195848, "grad_norm": 0.3154973089694977, "learning_rate": 1.9391663480504306e-07, "loss": 0.2629, "step": 2432 }, { "epoch": 0.6474188398084088, "grad_norm": 0.29702404141426086, "learning_rate": 1.939107736595857e-07, "loss": 0.2513, "step": 2433 }, { "epoch": 0.6476849387972325, "grad_norm": 0.2797849178314209, "learning_rate": 1.9390490978062954e-07, "loss": 0.2367, "step": 2434 }, { "epoch": 0.6479510377860564, "grad_norm": 0.3114733397960663, "learning_rate": 1.938990431683453e-07, "loss": 0.2421, "step": 2435 }, { "epoch": 0.6482171367748802, "grad_norm": 0.2912171483039856, "learning_rate": 1.938931738229037e-07, "loss": 0.2204, "step": 2436 }, { "epoch": 0.6484832357637041, "grad_norm": 0.25488147139549255, "learning_rate": 1.9388730174447563e-07, "loss": 0.206, "step": 2437 }, { "epoch": 0.6487493347525279, "grad_norm": 0.2962315082550049, "learning_rate": 1.9388142693323197e-07, "loss": 0.2387, "step": 2438 }, { "epoch": 0.6490154337413517, "grad_norm": 0.38311633467674255, "learning_rate": 1.938755493893437e-07, "loss": 0.2519, "step": 2439 }, { "epoch": 0.6492815327301756, "grad_norm": 0.2761583924293518, "learning_rate": 1.9386966911298196e-07, "loss": 0.224, "step": 2440 }, { "epoch": 0.6495476317189994, "grad_norm": 0.3023720383644104, "learning_rate": 1.938637861043179e-07, "loss": 0.2576, "step": 2441 }, { "epoch": 0.6498137307078233, "grad_norm": 0.34970906376838684, "learning_rate": 1.9385790036352271e-07, "loss": 0.2316, "step": 2442 }, { "epoch": 0.6500798296966471, "grad_norm": 0.27399152517318726, "learning_rate": 1.9385201189076777e-07, "loss": 0.2611, "step": 2443 }, { "epoch": 0.650345928685471, "grad_norm": 0.30614498257637024, "learning_rate": 1.9384612068622444e-07, "loss": 0.2304, "step": 2444 }, { "epoch": 0.6506120276742948, "grad_norm": 0.2764569818973541, "learning_rate": 1.9384022675006419e-07, "loss": 0.2297, "step": 2445 }, { "epoch": 0.6508781266631187, "grad_norm": 0.26730310916900635, "learning_rate": 1.938343300824586e-07, "loss": 0.2016, "step": 2446 }, { "epoch": 0.6511442256519425, "grad_norm": 0.31798896193504333, "learning_rate": 1.938284306835793e-07, "loss": 0.2726, "step": 2447 }, { "epoch": 0.6514103246407663, "grad_norm": 0.26116663217544556, "learning_rate": 1.93822528553598e-07, "loss": 0.2303, "step": 2448 }, { "epoch": 0.6516764236295902, "grad_norm": 0.310499906539917, "learning_rate": 1.938166236926865e-07, "loss": 0.2399, "step": 2449 }, { "epoch": 0.651942522618414, "grad_norm": 0.28162136673927307, "learning_rate": 1.938107161010167e-07, "loss": 0.2351, "step": 2450 }, { "epoch": 0.6522086216072379, "grad_norm": 0.2852618992328644, "learning_rate": 1.9380480577876052e-07, "loss": 0.241, "step": 2451 }, { "epoch": 0.6524747205960617, "grad_norm": 0.3043206036090851, "learning_rate": 1.9379889272609e-07, "loss": 0.2369, "step": 2452 }, { "epoch": 0.6527408195848856, "grad_norm": 0.3182758092880249, "learning_rate": 1.9379297694317723e-07, "loss": 0.2576, "step": 2453 }, { "epoch": 0.6530069185737094, "grad_norm": 0.28672054409980774, "learning_rate": 1.9378705843019447e-07, "loss": 0.2231, "step": 2454 }, { "epoch": 0.6532730175625333, "grad_norm": 0.2760550081729889, "learning_rate": 1.9378113718731398e-07, "loss": 0.2246, "step": 2455 }, { "epoch": 0.6535391165513571, "grad_norm": 0.28620365262031555, "learning_rate": 1.9377521321470803e-07, "loss": 0.2417, "step": 2456 }, { "epoch": 0.6538052155401809, "grad_norm": 0.2795310616493225, "learning_rate": 1.9376928651254914e-07, "loss": 0.25, "step": 2457 }, { "epoch": 0.6540713145290048, "grad_norm": 0.31134486198425293, "learning_rate": 1.9376335708100977e-07, "loss": 0.219, "step": 2458 }, { "epoch": 0.6543374135178286, "grad_norm": 0.26554811000823975, "learning_rate": 1.9375742492026258e-07, "loss": 0.2075, "step": 2459 }, { "epoch": 0.6546035125066525, "grad_norm": 0.29869621992111206, "learning_rate": 1.937514900304801e-07, "loss": 0.2186, "step": 2460 }, { "epoch": 0.6548696114954763, "grad_norm": 0.2676819860935211, "learning_rate": 1.9374555241183526e-07, "loss": 0.2221, "step": 2461 }, { "epoch": 0.6551357104843002, "grad_norm": 0.2823232412338257, "learning_rate": 1.9373961206450075e-07, "loss": 0.2312, "step": 2462 }, { "epoch": 0.655401809473124, "grad_norm": 0.27801892161369324, "learning_rate": 1.9373366898864952e-07, "loss": 0.2603, "step": 2463 }, { "epoch": 0.6556679084619479, "grad_norm": 0.2647506296634674, "learning_rate": 1.9372772318445455e-07, "loss": 0.2316, "step": 2464 }, { "epoch": 0.6559340074507717, "grad_norm": 0.2917279899120331, "learning_rate": 1.9372177465208897e-07, "loss": 0.2484, "step": 2465 }, { "epoch": 0.6562001064395955, "grad_norm": 0.2862991988658905, "learning_rate": 1.9371582339172585e-07, "loss": 0.232, "step": 2466 }, { "epoch": 0.6564662054284194, "grad_norm": 0.2816179692745209, "learning_rate": 1.9370986940353844e-07, "loss": 0.2399, "step": 2467 }, { "epoch": 0.6567323044172432, "grad_norm": 0.28990599513053894, "learning_rate": 1.9370391268770004e-07, "loss": 0.2432, "step": 2468 }, { "epoch": 0.6569984034060671, "grad_norm": 0.3002142608165741, "learning_rate": 1.9369795324438405e-07, "loss": 0.2524, "step": 2469 }, { "epoch": 0.6572645023948909, "grad_norm": 0.27911505103111267, "learning_rate": 1.9369199107376392e-07, "loss": 0.2255, "step": 2470 }, { "epoch": 0.6575306013837148, "grad_norm": 0.2645341753959656, "learning_rate": 1.9368602617601323e-07, "loss": 0.2213, "step": 2471 }, { "epoch": 0.6577967003725386, "grad_norm": 0.2828204929828644, "learning_rate": 1.9368005855130553e-07, "loss": 0.2345, "step": 2472 }, { "epoch": 0.6580627993613625, "grad_norm": 0.28045883774757385, "learning_rate": 1.936740881998146e-07, "loss": 0.2437, "step": 2473 }, { "epoch": 0.6583288983501863, "grad_norm": 0.5234538912773132, "learning_rate": 1.9366811512171414e-07, "loss": 0.2543, "step": 2474 }, { "epoch": 0.6585949973390101, "grad_norm": 0.32378825545310974, "learning_rate": 1.936621393171781e-07, "loss": 0.2339, "step": 2475 }, { "epoch": 0.658861096327834, "grad_norm": 0.2778835594654083, "learning_rate": 1.9365616078638033e-07, "loss": 0.2339, "step": 2476 }, { "epoch": 0.6591271953166578, "grad_norm": 0.2889712154865265, "learning_rate": 1.936501795294949e-07, "loss": 0.2271, "step": 2477 }, { "epoch": 0.6593932943054817, "grad_norm": 0.3046238422393799, "learning_rate": 1.9364419554669595e-07, "loss": 0.2208, "step": 2478 }, { "epoch": 0.6596593932943055, "grad_norm": 0.26043790578842163, "learning_rate": 1.9363820883815757e-07, "loss": 0.2188, "step": 2479 }, { "epoch": 0.6599254922831294, "grad_norm": 0.29414811730384827, "learning_rate": 1.936322194040541e-07, "loss": 0.2389, "step": 2480 }, { "epoch": 0.6601915912719531, "grad_norm": 0.2986350655555725, "learning_rate": 1.9362622724455982e-07, "loss": 0.2318, "step": 2481 }, { "epoch": 0.660457690260777, "grad_norm": 0.2884815037250519, "learning_rate": 1.9362023235984914e-07, "loss": 0.2159, "step": 2482 }, { "epoch": 0.6607237892496008, "grad_norm": 0.3190803825855255, "learning_rate": 1.936142347500966e-07, "loss": 0.243, "step": 2483 }, { "epoch": 0.6609898882384246, "grad_norm": 0.2747652232646942, "learning_rate": 1.9360823441547675e-07, "loss": 0.2284, "step": 2484 }, { "epoch": 0.6612559872272485, "grad_norm": 0.30068475008010864, "learning_rate": 1.9360223135616423e-07, "loss": 0.2308, "step": 2485 }, { "epoch": 0.6615220862160723, "grad_norm": 0.2759617865085602, "learning_rate": 1.9359622557233381e-07, "loss": 0.2407, "step": 2486 }, { "epoch": 0.6617881852048962, "grad_norm": 0.2994861602783203, "learning_rate": 1.935902170641603e-07, "loss": 0.255, "step": 2487 }, { "epoch": 0.66205428419372, "grad_norm": 0.26931026577949524, "learning_rate": 1.9358420583181854e-07, "loss": 0.2143, "step": 2488 }, { "epoch": 0.6623203831825439, "grad_norm": 0.2713092267513275, "learning_rate": 1.9357819187548356e-07, "loss": 0.2179, "step": 2489 }, { "epoch": 0.6625864821713677, "grad_norm": 0.3182220458984375, "learning_rate": 1.9357217519533038e-07, "loss": 0.2555, "step": 2490 }, { "epoch": 0.6628525811601916, "grad_norm": 0.2985818088054657, "learning_rate": 1.9356615579153414e-07, "loss": 0.2203, "step": 2491 }, { "epoch": 0.6631186801490154, "grad_norm": 0.2866062521934509, "learning_rate": 1.9356013366427005e-07, "loss": 0.2445, "step": 2492 }, { "epoch": 0.6633847791378393, "grad_norm": 0.3125109374523163, "learning_rate": 1.9355410881371338e-07, "loss": 0.2518, "step": 2493 }, { "epoch": 0.6636508781266631, "grad_norm": 0.3050571382045746, "learning_rate": 1.9354808124003954e-07, "loss": 0.232, "step": 2494 }, { "epoch": 0.6639169771154869, "grad_norm": 0.29351088404655457, "learning_rate": 1.9354205094342392e-07, "loss": 0.2245, "step": 2495 }, { "epoch": 0.6641830761043108, "grad_norm": 0.28277313709259033, "learning_rate": 1.935360179240421e-07, "loss": 0.2379, "step": 2496 }, { "epoch": 0.6644491750931346, "grad_norm": 0.320171058177948, "learning_rate": 1.9352998218206963e-07, "loss": 0.244, "step": 2497 }, { "epoch": 0.6647152740819585, "grad_norm": 0.2809485197067261, "learning_rate": 1.935239437176823e-07, "loss": 0.2461, "step": 2498 }, { "epoch": 0.6649813730707823, "grad_norm": 0.28395876288414, "learning_rate": 1.9351790253105574e-07, "loss": 0.2417, "step": 2499 }, { "epoch": 0.6652474720596062, "grad_norm": 0.2835882008075714, "learning_rate": 1.9351185862236586e-07, "loss": 0.2371, "step": 2500 }, { "epoch": 0.66551357104843, "grad_norm": 0.25346434116363525, "learning_rate": 1.9350581199178854e-07, "loss": 0.21, "step": 2501 }, { "epoch": 0.6657796700372539, "grad_norm": 0.3043253719806671, "learning_rate": 1.9349976263949989e-07, "loss": 0.2348, "step": 2502 }, { "epoch": 0.6660457690260777, "grad_norm": 0.28094613552093506, "learning_rate": 1.934937105656759e-07, "loss": 0.2457, "step": 2503 }, { "epoch": 0.6663118680149015, "grad_norm": 0.2785954177379608, "learning_rate": 1.934876557704927e-07, "loss": 0.2211, "step": 2504 }, { "epoch": 0.6665779670037254, "grad_norm": 0.28992295265197754, "learning_rate": 1.9348159825412665e-07, "loss": 0.2367, "step": 2505 }, { "epoch": 0.6668440659925492, "grad_norm": 0.28820937871932983, "learning_rate": 1.9347553801675396e-07, "loss": 0.2448, "step": 2506 }, { "epoch": 0.6671101649813731, "grad_norm": 0.2825653553009033, "learning_rate": 1.9346947505855105e-07, "loss": 0.2419, "step": 2507 }, { "epoch": 0.6673762639701969, "grad_norm": 0.29642385244369507, "learning_rate": 1.9346340937969442e-07, "loss": 0.2412, "step": 2508 }, { "epoch": 0.6676423629590208, "grad_norm": 0.2733689844608307, "learning_rate": 1.9345734098036064e-07, "loss": 0.2341, "step": 2509 }, { "epoch": 0.6679084619478446, "grad_norm": 0.2693597376346588, "learning_rate": 1.9345126986072632e-07, "loss": 0.2122, "step": 2510 }, { "epoch": 0.6681745609366685, "grad_norm": 0.2967769503593445, "learning_rate": 1.934451960209682e-07, "loss": 0.2321, "step": 2511 }, { "epoch": 0.6684406599254923, "grad_norm": 0.27510106563568115, "learning_rate": 1.93439119461263e-07, "loss": 0.25, "step": 2512 }, { "epoch": 0.6687067589143161, "grad_norm": 0.3147059381008148, "learning_rate": 1.9343304018178767e-07, "loss": 0.2652, "step": 2513 }, { "epoch": 0.66897285790314, "grad_norm": 0.2987494170665741, "learning_rate": 1.9342695818271912e-07, "loss": 0.2423, "step": 2514 }, { "epoch": 0.6692389568919638, "grad_norm": 0.2650598883628845, "learning_rate": 1.9342087346423444e-07, "loss": 0.2189, "step": 2515 }, { "epoch": 0.6695050558807877, "grad_norm": 0.2923736870288849, "learning_rate": 1.9341478602651066e-07, "loss": 0.2395, "step": 2516 }, { "epoch": 0.6697711548696115, "grad_norm": 0.28261980414390564, "learning_rate": 1.9340869586972503e-07, "loss": 0.2383, "step": 2517 }, { "epoch": 0.6700372538584354, "grad_norm": 0.2848859429359436, "learning_rate": 1.9340260299405477e-07, "loss": 0.2246, "step": 2518 }, { "epoch": 0.6703033528472592, "grad_norm": 0.29600003361701965, "learning_rate": 1.9339650739967727e-07, "loss": 0.2331, "step": 2519 }, { "epoch": 0.6705694518360831, "grad_norm": 0.2747430205345154, "learning_rate": 1.9339040908676996e-07, "loss": 0.2257, "step": 2520 }, { "epoch": 0.6708355508249069, "grad_norm": 0.30784931778907776, "learning_rate": 1.933843080555103e-07, "loss": 0.2544, "step": 2521 }, { "epoch": 0.6711016498137307, "grad_norm": 0.2813202738761902, "learning_rate": 1.933782043060759e-07, "loss": 0.2404, "step": 2522 }, { "epoch": 0.6713677488025546, "grad_norm": 0.3316592574119568, "learning_rate": 1.9337209783864444e-07, "loss": 0.2346, "step": 2523 }, { "epoch": 0.6716338477913784, "grad_norm": 0.2708277404308319, "learning_rate": 1.9336598865339364e-07, "loss": 0.2508, "step": 2524 }, { "epoch": 0.6718999467802023, "grad_norm": 0.26598820090293884, "learning_rate": 1.9335987675050137e-07, "loss": 0.2311, "step": 2525 }, { "epoch": 0.672166045769026, "grad_norm": 0.29059138894081116, "learning_rate": 1.9335376213014546e-07, "loss": 0.2203, "step": 2526 }, { "epoch": 0.67243214475785, "grad_norm": 0.2714228332042694, "learning_rate": 1.9334764479250393e-07, "loss": 0.213, "step": 2527 }, { "epoch": 0.6726982437466738, "grad_norm": 0.29675278067588806, "learning_rate": 1.9334152473775482e-07, "loss": 0.2387, "step": 2528 }, { "epoch": 0.6729643427354977, "grad_norm": 0.30357691645622253, "learning_rate": 1.933354019660763e-07, "loss": 0.2522, "step": 2529 }, { "epoch": 0.6732304417243214, "grad_norm": 0.29799216985702515, "learning_rate": 1.9332927647764658e-07, "loss": 0.2343, "step": 2530 }, { "epoch": 0.6734965407131452, "grad_norm": 0.26863372325897217, "learning_rate": 1.9332314827264394e-07, "loss": 0.2293, "step": 2531 }, { "epoch": 0.6737626397019691, "grad_norm": 0.26939451694488525, "learning_rate": 1.9331701735124674e-07, "loss": 0.2269, "step": 2532 }, { "epoch": 0.6740287386907929, "grad_norm": 0.2610328495502472, "learning_rate": 1.9331088371363352e-07, "loss": 0.2097, "step": 2533 }, { "epoch": 0.6742948376796168, "grad_norm": 0.3124130666255951, "learning_rate": 1.933047473599827e-07, "loss": 0.2619, "step": 2534 }, { "epoch": 0.6745609366684406, "grad_norm": 0.3163851201534271, "learning_rate": 1.9329860829047298e-07, "loss": 0.2579, "step": 2535 }, { "epoch": 0.6748270356572645, "grad_norm": 0.26046669483184814, "learning_rate": 1.93292466505283e-07, "loss": 0.2394, "step": 2536 }, { "epoch": 0.6750931346460883, "grad_norm": 0.3041304349899292, "learning_rate": 1.9328632200459153e-07, "loss": 0.2496, "step": 2537 }, { "epoch": 0.6753592336349122, "grad_norm": 0.3399668335914612, "learning_rate": 1.932801747885775e-07, "loss": 0.2457, "step": 2538 }, { "epoch": 0.675625332623736, "grad_norm": 0.2897587716579437, "learning_rate": 1.9327402485741974e-07, "loss": 0.2253, "step": 2539 }, { "epoch": 0.6758914316125598, "grad_norm": 0.25702762603759766, "learning_rate": 1.932678722112973e-07, "loss": 0.2029, "step": 2540 }, { "epoch": 0.6761575306013837, "grad_norm": 0.2975272238254547, "learning_rate": 1.932617168503893e-07, "loss": 0.2481, "step": 2541 }, { "epoch": 0.6764236295902075, "grad_norm": 0.2833004891872406, "learning_rate": 1.9325555877487485e-07, "loss": 0.2323, "step": 2542 }, { "epoch": 0.6766897285790314, "grad_norm": 0.28807470202445984, "learning_rate": 1.9324939798493322e-07, "loss": 0.227, "step": 2543 }, { "epoch": 0.6769558275678552, "grad_norm": 0.2764955461025238, "learning_rate": 1.9324323448074372e-07, "loss": 0.2369, "step": 2544 }, { "epoch": 0.6772219265566791, "grad_norm": 0.3123075067996979, "learning_rate": 1.932370682624858e-07, "loss": 0.2368, "step": 2545 }, { "epoch": 0.6774880255455029, "grad_norm": 0.28514108061790466, "learning_rate": 1.9323089933033887e-07, "loss": 0.2278, "step": 2546 }, { "epoch": 0.6777541245343268, "grad_norm": 0.29371878504753113, "learning_rate": 1.9322472768448256e-07, "loss": 0.2443, "step": 2547 }, { "epoch": 0.6780202235231506, "grad_norm": 0.28686895966529846, "learning_rate": 1.9321855332509648e-07, "loss": 0.2236, "step": 2548 }, { "epoch": 0.6782863225119744, "grad_norm": 0.24385066330432892, "learning_rate": 1.9321237625236035e-07, "loss": 0.2129, "step": 2549 }, { "epoch": 0.6785524215007983, "grad_norm": 0.2889416813850403, "learning_rate": 1.9320619646645396e-07, "loss": 0.2588, "step": 2550 }, { "epoch": 0.6788185204896221, "grad_norm": 0.31343239545822144, "learning_rate": 1.9320001396755724e-07, "loss": 0.266, "step": 2551 }, { "epoch": 0.679084619478446, "grad_norm": 0.28492051362991333, "learning_rate": 1.9319382875585008e-07, "loss": 0.2184, "step": 2552 }, { "epoch": 0.6793507184672698, "grad_norm": 0.30366644263267517, "learning_rate": 1.9318764083151254e-07, "loss": 0.2379, "step": 2553 }, { "epoch": 0.6796168174560937, "grad_norm": 0.28692761063575745, "learning_rate": 1.9318145019472473e-07, "loss": 0.2251, "step": 2554 }, { "epoch": 0.6798829164449175, "grad_norm": 0.2769697606563568, "learning_rate": 1.9317525684566683e-07, "loss": 0.2451, "step": 2555 }, { "epoch": 0.6801490154337414, "grad_norm": 0.26835954189300537, "learning_rate": 1.9316906078451914e-07, "loss": 0.2276, "step": 2556 }, { "epoch": 0.6804151144225652, "grad_norm": 0.29075103998184204, "learning_rate": 1.9316286201146205e-07, "loss": 0.2396, "step": 2557 }, { "epoch": 0.680681213411389, "grad_norm": 0.2627943456172943, "learning_rate": 1.9315666052667587e-07, "loss": 0.2214, "step": 2558 }, { "epoch": 0.6809473124002129, "grad_norm": 0.2769909203052521, "learning_rate": 1.9315045633034123e-07, "loss": 0.2205, "step": 2559 }, { "epoch": 0.6812134113890367, "grad_norm": 0.3126809298992157, "learning_rate": 1.9314424942263866e-07, "loss": 0.234, "step": 2560 }, { "epoch": 0.6814795103778606, "grad_norm": 0.2925463616847992, "learning_rate": 1.931380398037488e-07, "loss": 0.2281, "step": 2561 }, { "epoch": 0.6817456093666844, "grad_norm": 0.3029822111129761, "learning_rate": 1.9313182747385248e-07, "loss": 0.2233, "step": 2562 }, { "epoch": 0.6820117083555083, "grad_norm": 0.28491124510765076, "learning_rate": 1.9312561243313046e-07, "loss": 0.2347, "step": 2563 }, { "epoch": 0.6822778073443321, "grad_norm": 0.2817009687423706, "learning_rate": 1.9311939468176366e-07, "loss": 0.2348, "step": 2564 }, { "epoch": 0.682543906333156, "grad_norm": 0.26688718795776367, "learning_rate": 1.9311317421993304e-07, "loss": 0.2209, "step": 2565 }, { "epoch": 0.6828100053219798, "grad_norm": 0.2562592923641205, "learning_rate": 1.931069510478197e-07, "loss": 0.2332, "step": 2566 }, { "epoch": 0.6830761043108036, "grad_norm": 0.2861890196800232, "learning_rate": 1.9310072516560477e-07, "loss": 0.2419, "step": 2567 }, { "epoch": 0.6833422032996275, "grad_norm": 0.30891382694244385, "learning_rate": 1.9309449657346946e-07, "loss": 0.2371, "step": 2568 }, { "epoch": 0.6836083022884513, "grad_norm": 0.2778546214103699, "learning_rate": 1.930882652715951e-07, "loss": 0.2181, "step": 2569 }, { "epoch": 0.6838744012772752, "grad_norm": 0.2946532070636749, "learning_rate": 1.93082031260163e-07, "loss": 0.236, "step": 2570 }, { "epoch": 0.684140500266099, "grad_norm": 0.31203705072402954, "learning_rate": 1.9307579453935468e-07, "loss": 0.2322, "step": 2571 }, { "epoch": 0.6844065992549229, "grad_norm": 0.27470508217811584, "learning_rate": 1.9306955510935163e-07, "loss": 0.2223, "step": 2572 }, { "epoch": 0.6846726982437467, "grad_norm": 0.29845985770225525, "learning_rate": 1.930633129703355e-07, "loss": 0.2326, "step": 2573 }, { "epoch": 0.6849387972325706, "grad_norm": 0.272264689207077, "learning_rate": 1.9305706812248794e-07, "loss": 0.2144, "step": 2574 }, { "epoch": 0.6852048962213944, "grad_norm": 0.2586918771266937, "learning_rate": 1.9305082056599079e-07, "loss": 0.2164, "step": 2575 }, { "epoch": 0.6854709952102181, "grad_norm": 0.2963729798793793, "learning_rate": 1.9304457030102582e-07, "loss": 0.2474, "step": 2576 }, { "epoch": 0.685737094199042, "grad_norm": 0.29444509744644165, "learning_rate": 1.93038317327775e-07, "loss": 0.247, "step": 2577 }, { "epoch": 0.6860031931878658, "grad_norm": 0.2963295876979828, "learning_rate": 1.9303206164642035e-07, "loss": 0.2542, "step": 2578 }, { "epoch": 0.6862692921766897, "grad_norm": 0.3004390001296997, "learning_rate": 1.9302580325714395e-07, "loss": 0.2554, "step": 2579 }, { "epoch": 0.6865353911655135, "grad_norm": 0.27662861347198486, "learning_rate": 1.9301954216012792e-07, "loss": 0.24, "step": 2580 }, { "epoch": 0.6868014901543374, "grad_norm": 0.28291556239128113, "learning_rate": 1.9301327835555456e-07, "loss": 0.2239, "step": 2581 }, { "epoch": 0.6870675891431612, "grad_norm": 0.35586950182914734, "learning_rate": 1.9300701184360619e-07, "loss": 0.2344, "step": 2582 }, { "epoch": 0.6873336881319851, "grad_norm": 0.2634173631668091, "learning_rate": 1.9300074262446518e-07, "loss": 0.2201, "step": 2583 }, { "epoch": 0.6875997871208089, "grad_norm": 0.2913946211338043, "learning_rate": 1.9299447069831402e-07, "loss": 0.2223, "step": 2584 }, { "epoch": 0.6878658861096327, "grad_norm": 0.3023858368396759, "learning_rate": 1.9298819606533528e-07, "loss": 0.2163, "step": 2585 }, { "epoch": 0.6881319850984566, "grad_norm": 0.30332276225090027, "learning_rate": 1.9298191872571162e-07, "loss": 0.2354, "step": 2586 }, { "epoch": 0.6883980840872804, "grad_norm": 0.29661861062049866, "learning_rate": 1.929756386796257e-07, "loss": 0.2256, "step": 2587 }, { "epoch": 0.6886641830761043, "grad_norm": 0.33829811215400696, "learning_rate": 1.9296935592726034e-07, "loss": 0.2278, "step": 2588 }, { "epoch": 0.6889302820649281, "grad_norm": 0.2746559679508209, "learning_rate": 1.9296307046879843e-07, "loss": 0.2314, "step": 2589 }, { "epoch": 0.689196381053752, "grad_norm": 0.2990932762622833, "learning_rate": 1.9295678230442293e-07, "loss": 0.2407, "step": 2590 }, { "epoch": 0.6894624800425758, "grad_norm": 0.3076329529285431, "learning_rate": 1.9295049143431683e-07, "loss": 0.2536, "step": 2591 }, { "epoch": 0.6897285790313997, "grad_norm": 0.31027525663375854, "learning_rate": 1.929441978586633e-07, "loss": 0.2541, "step": 2592 }, { "epoch": 0.6899946780202235, "grad_norm": 0.3035404682159424, "learning_rate": 1.9293790157764546e-07, "loss": 0.2565, "step": 2593 }, { "epoch": 0.6902607770090473, "grad_norm": 0.2856816351413727, "learning_rate": 1.9293160259144664e-07, "loss": 0.2312, "step": 2594 }, { "epoch": 0.6905268759978712, "grad_norm": 0.29525139927864075, "learning_rate": 1.929253009002502e-07, "loss": 0.2252, "step": 2595 }, { "epoch": 0.690792974986695, "grad_norm": 0.302531898021698, "learning_rate": 1.9291899650423948e-07, "loss": 0.2254, "step": 2596 }, { "epoch": 0.6910590739755189, "grad_norm": 0.278195858001709, "learning_rate": 1.9291268940359803e-07, "loss": 0.2213, "step": 2597 }, { "epoch": 0.6913251729643427, "grad_norm": 0.28615203499794006, "learning_rate": 1.9290637959850945e-07, "loss": 0.2266, "step": 2598 }, { "epoch": 0.6915912719531666, "grad_norm": 0.29012468457221985, "learning_rate": 1.929000670891574e-07, "loss": 0.233, "step": 2599 }, { "epoch": 0.6918573709419904, "grad_norm": 0.2837730944156647, "learning_rate": 1.928937518757256e-07, "loss": 0.2218, "step": 2600 }, { "epoch": 0.6921234699308143, "grad_norm": 0.28562960028648376, "learning_rate": 1.9288743395839786e-07, "loss": 0.2203, "step": 2601 }, { "epoch": 0.6923895689196381, "grad_norm": 0.26169121265411377, "learning_rate": 1.9288111333735811e-07, "loss": 0.223, "step": 2602 }, { "epoch": 0.692655667908462, "grad_norm": 0.2838384509086609, "learning_rate": 1.9287479001279032e-07, "loss": 0.2411, "step": 2603 }, { "epoch": 0.6929217668972858, "grad_norm": 0.30142489075660706, "learning_rate": 1.9286846398487855e-07, "loss": 0.2342, "step": 2604 }, { "epoch": 0.6931878658861096, "grad_norm": 0.2798221707344055, "learning_rate": 1.9286213525380692e-07, "loss": 0.2444, "step": 2605 }, { "epoch": 0.6934539648749335, "grad_norm": 0.2776472568511963, "learning_rate": 1.9285580381975964e-07, "loss": 0.224, "step": 2606 }, { "epoch": 0.6937200638637573, "grad_norm": 0.305195689201355, "learning_rate": 1.92849469682921e-07, "loss": 0.2497, "step": 2607 }, { "epoch": 0.6939861628525812, "grad_norm": 0.2793996036052704, "learning_rate": 1.9284313284347539e-07, "loss": 0.2153, "step": 2608 }, { "epoch": 0.694252261841405, "grad_norm": 0.3076910078525543, "learning_rate": 1.9283679330160724e-07, "loss": 0.2145, "step": 2609 }, { "epoch": 0.6945183608302289, "grad_norm": 0.26865535974502563, "learning_rate": 1.9283045105750108e-07, "loss": 0.2361, "step": 2610 }, { "epoch": 0.6947844598190527, "grad_norm": 0.256868839263916, "learning_rate": 1.9282410611134155e-07, "loss": 0.2281, "step": 2611 }, { "epoch": 0.6950505588078766, "grad_norm": 0.29786229133605957, "learning_rate": 1.928177584633133e-07, "loss": 0.2277, "step": 2612 }, { "epoch": 0.6953166577967004, "grad_norm": 0.3801967203617096, "learning_rate": 1.9281140811360108e-07, "loss": 0.2497, "step": 2613 }, { "epoch": 0.6955827567855242, "grad_norm": 0.29756712913513184, "learning_rate": 1.9280505506238977e-07, "loss": 0.2413, "step": 2614 }, { "epoch": 0.6958488557743481, "grad_norm": 0.28267475962638855, "learning_rate": 1.9279869930986426e-07, "loss": 0.2398, "step": 2615 }, { "epoch": 0.6961149547631719, "grad_norm": 0.2899039089679718, "learning_rate": 1.9279234085620956e-07, "loss": 0.2291, "step": 2616 }, { "epoch": 0.6963810537519958, "grad_norm": 0.2804468274116516, "learning_rate": 1.9278597970161075e-07, "loss": 0.2336, "step": 2617 }, { "epoch": 0.6966471527408196, "grad_norm": 0.29222482442855835, "learning_rate": 1.92779615846253e-07, "loss": 0.2233, "step": 2618 }, { "epoch": 0.6969132517296435, "grad_norm": 0.29536327719688416, "learning_rate": 1.9277324929032153e-07, "loss": 0.2207, "step": 2619 }, { "epoch": 0.6971793507184673, "grad_norm": 0.2995297610759735, "learning_rate": 1.9276688003400166e-07, "loss": 0.2542, "step": 2620 }, { "epoch": 0.6974454497072912, "grad_norm": 0.29771360754966736, "learning_rate": 1.9276050807747877e-07, "loss": 0.2414, "step": 2621 }, { "epoch": 0.697711548696115, "grad_norm": 0.27944427728652954, "learning_rate": 1.9275413342093833e-07, "loss": 0.241, "step": 2622 }, { "epoch": 0.6979776476849388, "grad_norm": 0.29280513525009155, "learning_rate": 1.9274775606456592e-07, "loss": 0.2353, "step": 2623 }, { "epoch": 0.6982437466737627, "grad_norm": 0.2873554825782776, "learning_rate": 1.9274137600854715e-07, "loss": 0.2527, "step": 2624 }, { "epoch": 0.6985098456625864, "grad_norm": 0.27029183506965637, "learning_rate": 1.927349932530677e-07, "loss": 0.2358, "step": 2625 }, { "epoch": 0.6987759446514104, "grad_norm": 0.2654604911804199, "learning_rate": 1.9272860779831342e-07, "loss": 0.2178, "step": 2626 }, { "epoch": 0.6990420436402341, "grad_norm": 0.3223857879638672, "learning_rate": 1.927222196444701e-07, "loss": 0.2472, "step": 2627 }, { "epoch": 0.699308142629058, "grad_norm": 0.3178917169570923, "learning_rate": 1.9271582879172374e-07, "loss": 0.2261, "step": 2628 }, { "epoch": 0.6995742416178818, "grad_norm": 0.2674219608306885, "learning_rate": 1.9270943524026029e-07, "loss": 0.2371, "step": 2629 }, { "epoch": 0.6998403406067057, "grad_norm": 0.3179113268852234, "learning_rate": 1.9270303899026593e-07, "loss": 0.2395, "step": 2630 }, { "epoch": 0.7001064395955295, "grad_norm": 0.2618655264377594, "learning_rate": 1.9269664004192682e-07, "loss": 0.2318, "step": 2631 }, { "epoch": 0.7003725385843533, "grad_norm": 0.33168333768844604, "learning_rate": 1.9269023839542918e-07, "loss": 0.2336, "step": 2632 }, { "epoch": 0.7006386375731772, "grad_norm": 0.2564088702201843, "learning_rate": 1.926838340509594e-07, "loss": 0.2201, "step": 2633 }, { "epoch": 0.700904736562001, "grad_norm": 0.2817218601703644, "learning_rate": 1.9267742700870383e-07, "loss": 0.2158, "step": 2634 }, { "epoch": 0.7011708355508249, "grad_norm": 0.2915916442871094, "learning_rate": 1.92671017268849e-07, "loss": 0.2263, "step": 2635 }, { "epoch": 0.7014369345396487, "grad_norm": 0.3065217137336731, "learning_rate": 1.9266460483158148e-07, "loss": 0.2431, "step": 2636 }, { "epoch": 0.7017030335284726, "grad_norm": 0.2901429235935211, "learning_rate": 1.9265818969708795e-07, "loss": 0.225, "step": 2637 }, { "epoch": 0.7019691325172964, "grad_norm": 0.24695824086666107, "learning_rate": 1.9265177186555506e-07, "loss": 0.2075, "step": 2638 }, { "epoch": 0.7022352315061203, "grad_norm": 0.3204401135444641, "learning_rate": 1.9264535133716966e-07, "loss": 0.2273, "step": 2639 }, { "epoch": 0.7025013304949441, "grad_norm": 0.27052441239356995, "learning_rate": 1.9263892811211864e-07, "loss": 0.2313, "step": 2640 }, { "epoch": 0.7027674294837679, "grad_norm": 0.301705539226532, "learning_rate": 1.9263250219058895e-07, "loss": 0.2143, "step": 2641 }, { "epoch": 0.7030335284725918, "grad_norm": 0.30490919947624207, "learning_rate": 1.9262607357276764e-07, "loss": 0.2654, "step": 2642 }, { "epoch": 0.7032996274614156, "grad_norm": 0.2756442725658417, "learning_rate": 1.9261964225884185e-07, "loss": 0.2331, "step": 2643 }, { "epoch": 0.7035657264502395, "grad_norm": 0.284857839345932, "learning_rate": 1.9261320824899875e-07, "loss": 0.2459, "step": 2644 }, { "epoch": 0.7038318254390633, "grad_norm": 0.27984511852264404, "learning_rate": 1.9260677154342562e-07, "loss": 0.2293, "step": 2645 }, { "epoch": 0.7040979244278872, "grad_norm": 0.2863099277019501, "learning_rate": 1.9260033214230982e-07, "loss": 0.2327, "step": 2646 }, { "epoch": 0.704364023416711, "grad_norm": 0.27033698558807373, "learning_rate": 1.9259389004583876e-07, "loss": 0.1975, "step": 2647 }, { "epoch": 0.7046301224055349, "grad_norm": 0.28092119097709656, "learning_rate": 1.9258744525420004e-07, "loss": 0.2225, "step": 2648 }, { "epoch": 0.7048962213943587, "grad_norm": 0.30478495359420776, "learning_rate": 1.9258099776758114e-07, "loss": 0.2277, "step": 2649 }, { "epoch": 0.7051623203831825, "grad_norm": 0.26379865407943726, "learning_rate": 1.925745475861698e-07, "loss": 0.2152, "step": 2650 }, { "epoch": 0.7054284193720064, "grad_norm": 0.29598692059516907, "learning_rate": 1.9256809471015375e-07, "loss": 0.2588, "step": 2651 }, { "epoch": 0.7056945183608302, "grad_norm": 0.3016447424888611, "learning_rate": 1.925616391397208e-07, "loss": 0.257, "step": 2652 }, { "epoch": 0.7059606173496541, "grad_norm": 0.2956690788269043, "learning_rate": 1.9255518087505887e-07, "loss": 0.2042, "step": 2653 }, { "epoch": 0.7062267163384779, "grad_norm": 0.29092010855674744, "learning_rate": 1.9254871991635595e-07, "loss": 0.2397, "step": 2654 }, { "epoch": 0.7064928153273018, "grad_norm": 0.284396767616272, "learning_rate": 1.9254225626380008e-07, "loss": 0.2426, "step": 2655 }, { "epoch": 0.7067589143161256, "grad_norm": 0.39840367436408997, "learning_rate": 1.925357899175794e-07, "loss": 0.2541, "step": 2656 }, { "epoch": 0.7070250133049495, "grad_norm": 0.30049487948417664, "learning_rate": 1.9252932087788222e-07, "loss": 0.2322, "step": 2657 }, { "epoch": 0.7072911122937733, "grad_norm": 0.2630608081817627, "learning_rate": 1.925228491448967e-07, "loss": 0.2187, "step": 2658 }, { "epoch": 0.7075572112825971, "grad_norm": 0.29643720388412476, "learning_rate": 1.9251637471881125e-07, "loss": 0.2337, "step": 2659 }, { "epoch": 0.707823310271421, "grad_norm": 0.3021014332771301, "learning_rate": 1.9250989759981438e-07, "loss": 0.2398, "step": 2660 }, { "epoch": 0.7080894092602448, "grad_norm": 0.2829304039478302, "learning_rate": 1.925034177880946e-07, "loss": 0.2103, "step": 2661 }, { "epoch": 0.7083555082490687, "grad_norm": 0.2895389199256897, "learning_rate": 1.924969352838405e-07, "loss": 0.2242, "step": 2662 }, { "epoch": 0.7086216072378925, "grad_norm": 0.26183971762657166, "learning_rate": 1.9249045008724076e-07, "loss": 0.2383, "step": 2663 }, { "epoch": 0.7088877062267164, "grad_norm": 0.28220289945602417, "learning_rate": 1.9248396219848418e-07, "loss": 0.2383, "step": 2664 }, { "epoch": 0.7091538052155402, "grad_norm": 0.3043958246707916, "learning_rate": 1.924774716177596e-07, "loss": 0.2391, "step": 2665 }, { "epoch": 0.7094199042043641, "grad_norm": 0.2752787172794342, "learning_rate": 1.9247097834525595e-07, "loss": 0.2152, "step": 2666 }, { "epoch": 0.7096860031931879, "grad_norm": 0.2836426794528961, "learning_rate": 1.9246448238116223e-07, "loss": 0.2393, "step": 2667 }, { "epoch": 0.7099521021820117, "grad_norm": 0.29638296365737915, "learning_rate": 1.9245798372566745e-07, "loss": 0.2263, "step": 2668 }, { "epoch": 0.7102182011708356, "grad_norm": 0.33681970834732056, "learning_rate": 1.9245148237896088e-07, "loss": 0.2557, "step": 2669 }, { "epoch": 0.7104843001596594, "grad_norm": 0.2706678807735443, "learning_rate": 1.9244497834123167e-07, "loss": 0.2431, "step": 2670 }, { "epoch": 0.7107503991484833, "grad_norm": 0.29411080479621887, "learning_rate": 1.924384716126692e-07, "loss": 0.2478, "step": 2671 }, { "epoch": 0.711016498137307, "grad_norm": 0.27800869941711426, "learning_rate": 1.9243196219346283e-07, "loss": 0.212, "step": 2672 }, { "epoch": 0.711282597126131, "grad_norm": 0.3019680678844452, "learning_rate": 1.9242545008380202e-07, "loss": 0.2374, "step": 2673 }, { "epoch": 0.7115486961149547, "grad_norm": 0.2690466642379761, "learning_rate": 1.9241893528387637e-07, "loss": 0.2191, "step": 2674 }, { "epoch": 0.7118147951037787, "grad_norm": 0.27753257751464844, "learning_rate": 1.9241241779387543e-07, "loss": 0.2394, "step": 2675 }, { "epoch": 0.7120808940926024, "grad_norm": 0.2876778841018677, "learning_rate": 1.9240589761398896e-07, "loss": 0.2363, "step": 2676 }, { "epoch": 0.7123469930814262, "grad_norm": 0.27430635690689087, "learning_rate": 1.923993747444068e-07, "loss": 0.2302, "step": 2677 }, { "epoch": 0.7126130920702501, "grad_norm": 0.3000454008579254, "learning_rate": 1.9239284918531871e-07, "loss": 0.2456, "step": 2678 }, { "epoch": 0.7128791910590739, "grad_norm": 0.29172924160957336, "learning_rate": 1.9238632093691467e-07, "loss": 0.2422, "step": 2679 }, { "epoch": 0.7131452900478978, "grad_norm": 0.28884416818618774, "learning_rate": 1.9237978999938472e-07, "loss": 0.2314, "step": 2680 }, { "epoch": 0.7134113890367216, "grad_norm": 0.31402766704559326, "learning_rate": 1.9237325637291893e-07, "loss": 0.2309, "step": 2681 }, { "epoch": 0.7136774880255455, "grad_norm": 0.29400622844696045, "learning_rate": 1.9236672005770754e-07, "loss": 0.261, "step": 2682 }, { "epoch": 0.7139435870143693, "grad_norm": 0.272036612033844, "learning_rate": 1.9236018105394073e-07, "loss": 0.2274, "step": 2683 }, { "epoch": 0.7142096860031932, "grad_norm": 0.279997855424881, "learning_rate": 1.9235363936180882e-07, "loss": 0.2353, "step": 2684 }, { "epoch": 0.714475784992017, "grad_norm": 0.28337332606315613, "learning_rate": 1.923470949815023e-07, "loss": 0.2436, "step": 2685 }, { "epoch": 0.7147418839808408, "grad_norm": 0.29314205050468445, "learning_rate": 1.9234054791321162e-07, "loss": 0.2413, "step": 2686 }, { "epoch": 0.7150079829696647, "grad_norm": 0.25981074571609497, "learning_rate": 1.9233399815712734e-07, "loss": 0.2258, "step": 2687 }, { "epoch": 0.7152740819584885, "grad_norm": 0.2951703667640686, "learning_rate": 1.9232744571344013e-07, "loss": 0.229, "step": 2688 }, { "epoch": 0.7155401809473124, "grad_norm": 0.2768385410308838, "learning_rate": 1.923208905823407e-07, "loss": 0.2364, "step": 2689 }, { "epoch": 0.7158062799361362, "grad_norm": 0.2738455832004547, "learning_rate": 1.9231433276401985e-07, "loss": 0.2108, "step": 2690 }, { "epoch": 0.7160723789249601, "grad_norm": 0.27417534589767456, "learning_rate": 1.9230777225866843e-07, "loss": 0.2247, "step": 2691 }, { "epoch": 0.7163384779137839, "grad_norm": 0.27566206455230713, "learning_rate": 1.923012090664775e-07, "loss": 0.236, "step": 2692 }, { "epoch": 0.7166045769026078, "grad_norm": 0.3179967999458313, "learning_rate": 1.9229464318763799e-07, "loss": 0.2343, "step": 2693 }, { "epoch": 0.7168706758914316, "grad_norm": 0.2705002725124359, "learning_rate": 1.9228807462234106e-07, "loss": 0.2231, "step": 2694 }, { "epoch": 0.7171367748802554, "grad_norm": 0.3013603687286377, "learning_rate": 1.922815033707779e-07, "loss": 0.2411, "step": 2695 }, { "epoch": 0.7174028738690793, "grad_norm": 0.2950381338596344, "learning_rate": 1.922749294331398e-07, "loss": 0.2386, "step": 2696 }, { "epoch": 0.7176689728579031, "grad_norm": 0.24496865272521973, "learning_rate": 1.9226835280961806e-07, "loss": 0.2025, "step": 2697 }, { "epoch": 0.717935071846727, "grad_norm": 0.28800249099731445, "learning_rate": 1.9226177350040417e-07, "loss": 0.2423, "step": 2698 }, { "epoch": 0.7182011708355508, "grad_norm": 0.29204821586608887, "learning_rate": 1.9225519150568963e-07, "loss": 0.2373, "step": 2699 }, { "epoch": 0.7184672698243747, "grad_norm": 0.2983543574810028, "learning_rate": 1.9224860682566595e-07, "loss": 0.2326, "step": 2700 }, { "epoch": 0.7187333688131985, "grad_norm": 0.2845573127269745, "learning_rate": 1.922420194605249e-07, "loss": 0.2255, "step": 2701 }, { "epoch": 0.7189994678020224, "grad_norm": 0.2784920334815979, "learning_rate": 1.9223542941045813e-07, "loss": 0.2327, "step": 2702 }, { "epoch": 0.7192655667908462, "grad_norm": 0.2585446536540985, "learning_rate": 1.9222883667565752e-07, "loss": 0.2145, "step": 2703 }, { "epoch": 0.71953166577967, "grad_norm": 0.2792356312274933, "learning_rate": 1.922222412563149e-07, "loss": 0.2434, "step": 2704 }, { "epoch": 0.7197977647684939, "grad_norm": 0.27997756004333496, "learning_rate": 1.9221564315262232e-07, "loss": 0.2382, "step": 2705 }, { "epoch": 0.7200638637573177, "grad_norm": 0.2648625075817108, "learning_rate": 1.9220904236477184e-07, "loss": 0.2066, "step": 2706 }, { "epoch": 0.7203299627461416, "grad_norm": 0.27884501218795776, "learning_rate": 1.9220243889295552e-07, "loss": 0.2403, "step": 2707 }, { "epoch": 0.7205960617349654, "grad_norm": 0.28990480303764343, "learning_rate": 1.921958327373656e-07, "loss": 0.2254, "step": 2708 }, { "epoch": 0.7208621607237893, "grad_norm": 0.2701345682144165, "learning_rate": 1.9218922389819436e-07, "loss": 0.2275, "step": 2709 }, { "epoch": 0.7211282597126131, "grad_norm": 0.30960267782211304, "learning_rate": 1.9218261237563422e-07, "loss": 0.2552, "step": 2710 }, { "epoch": 0.721394358701437, "grad_norm": 0.2975165545940399, "learning_rate": 1.9217599816987755e-07, "loss": 0.2405, "step": 2711 }, { "epoch": 0.7216604576902608, "grad_norm": 0.26247742772102356, "learning_rate": 1.9216938128111693e-07, "loss": 0.2131, "step": 2712 }, { "epoch": 0.7219265566790847, "grad_norm": 0.27590179443359375, "learning_rate": 1.9216276170954492e-07, "loss": 0.2301, "step": 2713 }, { "epoch": 0.7221926556679085, "grad_norm": 0.34263738989830017, "learning_rate": 1.9215613945535424e-07, "loss": 0.2401, "step": 2714 }, { "epoch": 0.7224587546567323, "grad_norm": 0.2971380352973938, "learning_rate": 1.9214951451873758e-07, "loss": 0.2324, "step": 2715 }, { "epoch": 0.7227248536455562, "grad_norm": 0.28447049856185913, "learning_rate": 1.9214288689988782e-07, "loss": 0.23, "step": 2716 }, { "epoch": 0.72299095263438, "grad_norm": 0.2668474614620209, "learning_rate": 1.921362565989979e-07, "loss": 0.2269, "step": 2717 }, { "epoch": 0.7232570516232039, "grad_norm": 0.276432067155838, "learning_rate": 1.9212962361626075e-07, "loss": 0.2437, "step": 2718 }, { "epoch": 0.7235231506120277, "grad_norm": 0.28585222363471985, "learning_rate": 1.9212298795186949e-07, "loss": 0.2291, "step": 2719 }, { "epoch": 0.7237892496008516, "grad_norm": 0.2700658142566681, "learning_rate": 1.9211634960601723e-07, "loss": 0.2195, "step": 2720 }, { "epoch": 0.7240553485896754, "grad_norm": 0.2515268623828888, "learning_rate": 1.9210970857889725e-07, "loss": 0.2173, "step": 2721 }, { "epoch": 0.7243214475784993, "grad_norm": 0.2697259783744812, "learning_rate": 1.9210306487070276e-07, "loss": 0.2154, "step": 2722 }, { "epoch": 0.724587546567323, "grad_norm": 0.31490859389305115, "learning_rate": 1.9209641848162723e-07, "loss": 0.249, "step": 2723 }, { "epoch": 0.7248536455561468, "grad_norm": 0.2923659384250641, "learning_rate": 1.9208976941186408e-07, "loss": 0.2213, "step": 2724 }, { "epoch": 0.7251197445449707, "grad_norm": 0.27620142698287964, "learning_rate": 1.9208311766160684e-07, "loss": 0.2079, "step": 2725 }, { "epoch": 0.7253858435337945, "grad_norm": 0.3424578309059143, "learning_rate": 1.9207646323104914e-07, "loss": 0.2592, "step": 2726 }, { "epoch": 0.7256519425226184, "grad_norm": 0.28456199169158936, "learning_rate": 1.9206980612038465e-07, "loss": 0.2314, "step": 2727 }, { "epoch": 0.7259180415114422, "grad_norm": 0.29267454147338867, "learning_rate": 1.920631463298072e-07, "loss": 0.2387, "step": 2728 }, { "epoch": 0.7261841405002661, "grad_norm": 0.30810025334358215, "learning_rate": 1.9205648385951058e-07, "loss": 0.2287, "step": 2729 }, { "epoch": 0.7264502394890899, "grad_norm": 0.28759706020355225, "learning_rate": 1.9204981870968876e-07, "loss": 0.234, "step": 2730 }, { "epoch": 0.7267163384779138, "grad_norm": 0.2650372087955475, "learning_rate": 1.9204315088053568e-07, "loss": 0.261, "step": 2731 }, { "epoch": 0.7269824374667376, "grad_norm": 0.26458021998405457, "learning_rate": 1.920364803722455e-07, "loss": 0.2236, "step": 2732 }, { "epoch": 0.7272485364555614, "grad_norm": 0.3023631274700165, "learning_rate": 1.920298071850123e-07, "loss": 0.234, "step": 2733 }, { "epoch": 0.7275146354443853, "grad_norm": 0.30176296830177307, "learning_rate": 1.920231313190304e-07, "loss": 0.2501, "step": 2734 }, { "epoch": 0.7277807344332091, "grad_norm": 0.27697479724884033, "learning_rate": 1.9201645277449407e-07, "loss": 0.2084, "step": 2735 }, { "epoch": 0.728046833422033, "grad_norm": 0.2968454957008362, "learning_rate": 1.9200977155159775e-07, "loss": 0.2235, "step": 2736 }, { "epoch": 0.7283129324108568, "grad_norm": 0.29466474056243896, "learning_rate": 1.9200308765053586e-07, "loss": 0.2398, "step": 2737 }, { "epoch": 0.7285790313996807, "grad_norm": 0.27621638774871826, "learning_rate": 1.9199640107150297e-07, "loss": 0.2344, "step": 2738 }, { "epoch": 0.7288451303885045, "grad_norm": 0.29448646306991577, "learning_rate": 1.9198971181469373e-07, "loss": 0.2403, "step": 2739 }, { "epoch": 0.7291112293773284, "grad_norm": 0.29173901677131653, "learning_rate": 1.9198301988030278e-07, "loss": 0.2483, "step": 2740 }, { "epoch": 0.7293773283661522, "grad_norm": 0.2768477201461792, "learning_rate": 1.9197632526852502e-07, "loss": 0.2049, "step": 2741 }, { "epoch": 0.729643427354976, "grad_norm": 0.28221169114112854, "learning_rate": 1.9196962797955518e-07, "loss": 0.2358, "step": 2742 }, { "epoch": 0.7299095263437999, "grad_norm": 0.26604530215263367, "learning_rate": 1.9196292801358827e-07, "loss": 0.2275, "step": 2743 }, { "epoch": 0.7301756253326237, "grad_norm": 0.277112752199173, "learning_rate": 1.9195622537081935e-07, "loss": 0.23, "step": 2744 }, { "epoch": 0.7304417243214476, "grad_norm": 0.29201459884643555, "learning_rate": 1.9194952005144346e-07, "loss": 0.2407, "step": 2745 }, { "epoch": 0.7307078233102714, "grad_norm": 0.2807181179523468, "learning_rate": 1.919428120556558e-07, "loss": 0.2392, "step": 2746 }, { "epoch": 0.7309739222990953, "grad_norm": 0.3197695314884186, "learning_rate": 1.9193610138365158e-07, "loss": 0.2328, "step": 2747 }, { "epoch": 0.7312400212879191, "grad_norm": 0.2867830991744995, "learning_rate": 1.9192938803562616e-07, "loss": 0.2303, "step": 2748 }, { "epoch": 0.731506120276743, "grad_norm": 0.30322325229644775, "learning_rate": 1.9192267201177494e-07, "loss": 0.2333, "step": 2749 }, { "epoch": 0.7317722192655668, "grad_norm": 0.27981120347976685, "learning_rate": 1.9191595331229342e-07, "loss": 0.248, "step": 2750 }, { "epoch": 0.7320383182543906, "grad_norm": 0.2695596516132355, "learning_rate": 1.9190923193737716e-07, "loss": 0.2222, "step": 2751 }, { "epoch": 0.7323044172432145, "grad_norm": 0.28719624876976013, "learning_rate": 1.919025078872218e-07, "loss": 0.2507, "step": 2752 }, { "epoch": 0.7325705162320383, "grad_norm": 0.2947598397731781, "learning_rate": 1.9189578116202304e-07, "loss": 0.2436, "step": 2753 }, { "epoch": 0.7328366152208622, "grad_norm": 0.31623420119285583, "learning_rate": 1.918890517619767e-07, "loss": 0.242, "step": 2754 }, { "epoch": 0.733102714209686, "grad_norm": 0.2708337604999542, "learning_rate": 1.918823196872787e-07, "loss": 0.2281, "step": 2755 }, { "epoch": 0.7333688131985099, "grad_norm": 0.2860287129878998, "learning_rate": 1.9187558493812487e-07, "loss": 0.2564, "step": 2756 }, { "epoch": 0.7336349121873337, "grad_norm": 0.25942596793174744, "learning_rate": 1.9186884751471139e-07, "loss": 0.2116, "step": 2757 }, { "epoch": 0.7339010111761576, "grad_norm": 0.28914111852645874, "learning_rate": 1.9186210741723423e-07, "loss": 0.2443, "step": 2758 }, { "epoch": 0.7341671101649814, "grad_norm": 0.28322646021842957, "learning_rate": 1.9185536464588966e-07, "loss": 0.2236, "step": 2759 }, { "epoch": 0.7344332091538052, "grad_norm": 0.2942054271697998, "learning_rate": 1.9184861920087395e-07, "loss": 0.2415, "step": 2760 }, { "epoch": 0.7346993081426291, "grad_norm": 0.2960052788257599, "learning_rate": 1.918418710823834e-07, "loss": 0.2548, "step": 2761 }, { "epoch": 0.7349654071314529, "grad_norm": 0.28785058856010437, "learning_rate": 1.9183512029061445e-07, "loss": 0.2253, "step": 2762 }, { "epoch": 0.7352315061202768, "grad_norm": 0.3049631714820862, "learning_rate": 1.9182836682576357e-07, "loss": 0.2312, "step": 2763 }, { "epoch": 0.7354976051091006, "grad_norm": 0.27551570534706116, "learning_rate": 1.918216106880274e-07, "loss": 0.2192, "step": 2764 }, { "epoch": 0.7357637040979245, "grad_norm": 0.28337544202804565, "learning_rate": 1.9181485187760254e-07, "loss": 0.2395, "step": 2765 }, { "epoch": 0.7360298030867483, "grad_norm": 0.2791391611099243, "learning_rate": 1.9180809039468572e-07, "loss": 0.2229, "step": 2766 }, { "epoch": 0.7362959020755722, "grad_norm": 0.2704903781414032, "learning_rate": 1.918013262394738e-07, "loss": 0.2262, "step": 2767 }, { "epoch": 0.736562001064396, "grad_norm": 0.26171550154685974, "learning_rate": 1.917945594121636e-07, "loss": 0.2047, "step": 2768 }, { "epoch": 0.7368281000532197, "grad_norm": 0.2848527431488037, "learning_rate": 1.9178778991295212e-07, "loss": 0.2379, "step": 2769 }, { "epoch": 0.7370941990420437, "grad_norm": 0.2673186659812927, "learning_rate": 1.9178101774203643e-07, "loss": 0.2225, "step": 2770 }, { "epoch": 0.7373602980308674, "grad_norm": 0.3004118800163269, "learning_rate": 1.917742428996136e-07, "loss": 0.2365, "step": 2771 }, { "epoch": 0.7376263970196913, "grad_norm": 0.28373250365257263, "learning_rate": 1.9176746538588086e-07, "loss": 0.2355, "step": 2772 }, { "epoch": 0.7378924960085151, "grad_norm": 0.292270302772522, "learning_rate": 1.9176068520103545e-07, "loss": 0.2252, "step": 2773 }, { "epoch": 0.738158594997339, "grad_norm": 0.2944265305995941, "learning_rate": 1.9175390234527477e-07, "loss": 0.2329, "step": 2774 }, { "epoch": 0.7384246939861628, "grad_norm": 0.2721596658229828, "learning_rate": 1.9174711681879623e-07, "loss": 0.2219, "step": 2775 }, { "epoch": 0.7386907929749867, "grad_norm": 0.280666708946228, "learning_rate": 1.9174032862179732e-07, "loss": 0.2325, "step": 2776 }, { "epoch": 0.7389568919638105, "grad_norm": 0.26432400941848755, "learning_rate": 1.9173353775447567e-07, "loss": 0.2274, "step": 2777 }, { "epoch": 0.7392229909526343, "grad_norm": 0.30072811245918274, "learning_rate": 1.917267442170289e-07, "loss": 0.2375, "step": 2778 }, { "epoch": 0.7394890899414582, "grad_norm": 0.29999837279319763, "learning_rate": 1.917199480096548e-07, "loss": 0.2302, "step": 2779 }, { "epoch": 0.739755188930282, "grad_norm": 0.34437817335128784, "learning_rate": 1.9171314913255113e-07, "loss": 0.2248, "step": 2780 }, { "epoch": 0.7400212879191059, "grad_norm": 0.28389430046081543, "learning_rate": 1.9170634758591582e-07, "loss": 0.235, "step": 2781 }, { "epoch": 0.7402873869079297, "grad_norm": 0.2698524594306946, "learning_rate": 1.9169954336994684e-07, "loss": 0.2118, "step": 2782 }, { "epoch": 0.7405534858967536, "grad_norm": 0.28663456439971924, "learning_rate": 1.9169273648484228e-07, "loss": 0.2415, "step": 2783 }, { "epoch": 0.7408195848855774, "grad_norm": 0.309449166059494, "learning_rate": 1.916859269308002e-07, "loss": 0.2489, "step": 2784 }, { "epoch": 0.7410856838744013, "grad_norm": 0.2944347560405731, "learning_rate": 1.9167911470801887e-07, "loss": 0.2378, "step": 2785 }, { "epoch": 0.7413517828632251, "grad_norm": 0.2725279629230499, "learning_rate": 1.9167229981669654e-07, "loss": 0.2241, "step": 2786 }, { "epoch": 0.7416178818520489, "grad_norm": 0.31694337725639343, "learning_rate": 1.9166548225703164e-07, "loss": 0.2395, "step": 2787 }, { "epoch": 0.7418839808408728, "grad_norm": 0.26473256945610046, "learning_rate": 1.916586620292225e-07, "loss": 0.2348, "step": 2788 }, { "epoch": 0.7421500798296966, "grad_norm": 0.2964455187320709, "learning_rate": 1.9165183913346772e-07, "loss": 0.2295, "step": 2789 }, { "epoch": 0.7424161788185205, "grad_norm": 0.29786762595176697, "learning_rate": 1.9164501356996586e-07, "loss": 0.2273, "step": 2790 }, { "epoch": 0.7426822778073443, "grad_norm": 0.2848692834377289, "learning_rate": 1.9163818533891564e-07, "loss": 0.2028, "step": 2791 }, { "epoch": 0.7429483767961682, "grad_norm": 0.282585084438324, "learning_rate": 1.9163135444051575e-07, "loss": 0.227, "step": 2792 }, { "epoch": 0.743214475784992, "grad_norm": 0.3153277039527893, "learning_rate": 1.9162452087496505e-07, "loss": 0.2388, "step": 2793 }, { "epoch": 0.7434805747738159, "grad_norm": 0.34190234541893005, "learning_rate": 1.9161768464246245e-07, "loss": 0.2503, "step": 2794 }, { "epoch": 0.7437466737626397, "grad_norm": 0.29295268654823303, "learning_rate": 1.9161084574320693e-07, "loss": 0.2405, "step": 2795 }, { "epoch": 0.7440127727514635, "grad_norm": 0.27723392844200134, "learning_rate": 1.916040041773976e-07, "loss": 0.223, "step": 2796 }, { "epoch": 0.7442788717402874, "grad_norm": 0.27671507000923157, "learning_rate": 1.915971599452335e-07, "loss": 0.223, "step": 2797 }, { "epoch": 0.7445449707291112, "grad_norm": 0.2884395718574524, "learning_rate": 1.9159031304691396e-07, "loss": 0.2264, "step": 2798 }, { "epoch": 0.7448110697179351, "grad_norm": 0.2864048182964325, "learning_rate": 1.9158346348263817e-07, "loss": 0.2252, "step": 2799 }, { "epoch": 0.7450771687067589, "grad_norm": 0.2914586365222931, "learning_rate": 1.9157661125260557e-07, "loss": 0.2448, "step": 2800 }, { "epoch": 0.7453432676955828, "grad_norm": 0.3223208785057068, "learning_rate": 1.9156975635701558e-07, "loss": 0.2445, "step": 2801 }, { "epoch": 0.7456093666844066, "grad_norm": 0.2676781713962555, "learning_rate": 1.9156289879606775e-07, "loss": 0.2275, "step": 2802 }, { "epoch": 0.7458754656732305, "grad_norm": 0.2686799168586731, "learning_rate": 1.9155603856996167e-07, "loss": 0.2204, "step": 2803 }, { "epoch": 0.7461415646620543, "grad_norm": 0.26033470034599304, "learning_rate": 1.9154917567889703e-07, "loss": 0.2307, "step": 2804 }, { "epoch": 0.7464076636508781, "grad_norm": 0.2826031744480133, "learning_rate": 1.9154231012307362e-07, "loss": 0.2348, "step": 2805 }, { "epoch": 0.746673762639702, "grad_norm": 0.2918631434440613, "learning_rate": 1.9153544190269123e-07, "loss": 0.2269, "step": 2806 }, { "epoch": 0.7469398616285258, "grad_norm": 0.2749294340610504, "learning_rate": 1.9152857101794977e-07, "loss": 0.2302, "step": 2807 }, { "epoch": 0.7472059606173497, "grad_norm": 0.3169825077056885, "learning_rate": 1.9152169746904928e-07, "loss": 0.2296, "step": 2808 }, { "epoch": 0.7474720596061735, "grad_norm": 0.2843203842639923, "learning_rate": 1.9151482125618982e-07, "loss": 0.2168, "step": 2809 }, { "epoch": 0.7477381585949974, "grad_norm": 0.3078986406326294, "learning_rate": 1.9150794237957153e-07, "loss": 0.2522, "step": 2810 }, { "epoch": 0.7480042575838212, "grad_norm": 0.26488858461380005, "learning_rate": 1.9150106083939464e-07, "loss": 0.2385, "step": 2811 }, { "epoch": 0.7482703565726451, "grad_norm": 0.28779321908950806, "learning_rate": 1.9149417663585944e-07, "loss": 0.2243, "step": 2812 }, { "epoch": 0.7485364555614689, "grad_norm": 0.2879316210746765, "learning_rate": 1.914872897691663e-07, "loss": 0.2182, "step": 2813 }, { "epoch": 0.7488025545502927, "grad_norm": 0.30143648386001587, "learning_rate": 1.9148040023951573e-07, "loss": 0.2227, "step": 2814 }, { "epoch": 0.7490686535391166, "grad_norm": 0.28156614303588867, "learning_rate": 1.9147350804710821e-07, "loss": 0.2353, "step": 2815 }, { "epoch": 0.7493347525279404, "grad_norm": 0.29937735199928284, "learning_rate": 1.914666131921444e-07, "loss": 0.237, "step": 2816 }, { "epoch": 0.7496008515167643, "grad_norm": 0.2899397611618042, "learning_rate": 1.9145971567482495e-07, "loss": 0.2468, "step": 2817 }, { "epoch": 0.749866950505588, "grad_norm": 0.2698718011379242, "learning_rate": 1.9145281549535067e-07, "loss": 0.221, "step": 2818 }, { "epoch": 0.750133049494412, "grad_norm": 0.29021942615509033, "learning_rate": 1.9144591265392238e-07, "loss": 0.2379, "step": 2819 }, { "epoch": 0.7503991484832357, "grad_norm": 0.2661910653114319, "learning_rate": 1.9143900715074102e-07, "loss": 0.235, "step": 2820 }, { "epoch": 0.7506652474720596, "grad_norm": 0.26275748014450073, "learning_rate": 1.9143209898600755e-07, "loss": 0.214, "step": 2821 }, { "epoch": 0.7509313464608834, "grad_norm": 0.30209556221961975, "learning_rate": 1.914251881599231e-07, "loss": 0.2208, "step": 2822 }, { "epoch": 0.7511974454497073, "grad_norm": 0.29287445545196533, "learning_rate": 1.914182746726888e-07, "loss": 0.2297, "step": 2823 }, { "epoch": 0.7514635444385311, "grad_norm": 0.27513715624809265, "learning_rate": 1.914113585245059e-07, "loss": 0.2173, "step": 2824 }, { "epoch": 0.7517296434273549, "grad_norm": 0.26709944009780884, "learning_rate": 1.9140443971557567e-07, "loss": 0.2312, "step": 2825 }, { "epoch": 0.7519957424161788, "grad_norm": 0.3087577223777771, "learning_rate": 1.9139751824609958e-07, "loss": 0.245, "step": 2826 }, { "epoch": 0.7522618414050026, "grad_norm": 0.27617549896240234, "learning_rate": 1.91390594116279e-07, "loss": 0.2268, "step": 2827 }, { "epoch": 0.7525279403938265, "grad_norm": 0.29694417119026184, "learning_rate": 1.9138366732631552e-07, "loss": 0.2362, "step": 2828 }, { "epoch": 0.7527940393826503, "grad_norm": 0.2813529372215271, "learning_rate": 1.913767378764108e-07, "loss": 0.2205, "step": 2829 }, { "epoch": 0.7530601383714742, "grad_norm": 0.2960980236530304, "learning_rate": 1.9136980576676647e-07, "loss": 0.2334, "step": 2830 }, { "epoch": 0.753326237360298, "grad_norm": 0.3122749626636505, "learning_rate": 1.9136287099758434e-07, "loss": 0.2477, "step": 2831 }, { "epoch": 0.7535923363491219, "grad_norm": 0.28840211033821106, "learning_rate": 1.9135593356906625e-07, "loss": 0.235, "step": 2832 }, { "epoch": 0.7538584353379457, "grad_norm": 0.28255048394203186, "learning_rate": 1.9134899348141414e-07, "loss": 0.2347, "step": 2833 }, { "epoch": 0.7541245343267695, "grad_norm": 0.28232017159461975, "learning_rate": 1.9134205073483e-07, "loss": 0.2235, "step": 2834 }, { "epoch": 0.7543906333155934, "grad_norm": 1.3723360300064087, "learning_rate": 1.9133510532951596e-07, "loss": 0.2333, "step": 2835 }, { "epoch": 0.7546567323044172, "grad_norm": 0.2983147203922272, "learning_rate": 1.9132815726567415e-07, "loss": 0.2466, "step": 2836 }, { "epoch": 0.7549228312932411, "grad_norm": 0.2598191797733307, "learning_rate": 1.913212065435068e-07, "loss": 0.2111, "step": 2837 }, { "epoch": 0.7551889302820649, "grad_norm": 0.27792951464653015, "learning_rate": 1.9131425316321624e-07, "loss": 0.2415, "step": 2838 }, { "epoch": 0.7554550292708888, "grad_norm": 0.3122498691082001, "learning_rate": 1.9130729712500483e-07, "loss": 0.2664, "step": 2839 }, { "epoch": 0.7557211282597126, "grad_norm": 0.276044100522995, "learning_rate": 1.9130033842907513e-07, "loss": 0.2381, "step": 2840 }, { "epoch": 0.7559872272485365, "grad_norm": 0.3042062222957611, "learning_rate": 1.9129337707562962e-07, "loss": 0.2415, "step": 2841 }, { "epoch": 0.7562533262373603, "grad_norm": 0.30898934602737427, "learning_rate": 1.9128641306487094e-07, "loss": 0.241, "step": 2842 }, { "epoch": 0.7565194252261841, "grad_norm": 0.2878851592540741, "learning_rate": 1.9127944639700183e-07, "loss": 0.2204, "step": 2843 }, { "epoch": 0.756785524215008, "grad_norm": 0.3075234889984131, "learning_rate": 1.9127247707222502e-07, "loss": 0.2249, "step": 2844 }, { "epoch": 0.7570516232038318, "grad_norm": 0.2691819369792938, "learning_rate": 1.9126550509074337e-07, "loss": 0.2274, "step": 2845 }, { "epoch": 0.7573177221926557, "grad_norm": 0.2787667214870453, "learning_rate": 1.9125853045275986e-07, "loss": 0.2384, "step": 2846 }, { "epoch": 0.7575838211814795, "grad_norm": 0.27524444460868835, "learning_rate": 1.9125155315847747e-07, "loss": 0.2283, "step": 2847 }, { "epoch": 0.7578499201703034, "grad_norm": 0.2618286907672882, "learning_rate": 1.912445732080993e-07, "loss": 0.2311, "step": 2848 }, { "epoch": 0.7581160191591272, "grad_norm": 0.27307647466659546, "learning_rate": 1.912375906018285e-07, "loss": 0.2228, "step": 2849 }, { "epoch": 0.7583821181479511, "grad_norm": 0.2860243022441864, "learning_rate": 1.9123060533986837e-07, "loss": 0.2407, "step": 2850 }, { "epoch": 0.7586482171367749, "grad_norm": 0.27053123712539673, "learning_rate": 1.9122361742242217e-07, "loss": 0.2235, "step": 2851 }, { "epoch": 0.7589143161255987, "grad_norm": 0.2823949158191681, "learning_rate": 1.9121662684969332e-07, "loss": 0.2362, "step": 2852 }, { "epoch": 0.7591804151144226, "grad_norm": 0.2705933749675751, "learning_rate": 1.912096336218853e-07, "loss": 0.2319, "step": 2853 }, { "epoch": 0.7594465141032464, "grad_norm": 0.3134825527667999, "learning_rate": 1.912026377392017e-07, "loss": 0.2566, "step": 2854 }, { "epoch": 0.7597126130920703, "grad_norm": 0.29184573888778687, "learning_rate": 1.9119563920184613e-07, "loss": 0.2226, "step": 2855 }, { "epoch": 0.7599787120808941, "grad_norm": 0.32582205533981323, "learning_rate": 1.9118863801002224e-07, "loss": 0.2379, "step": 2856 }, { "epoch": 0.760244811069718, "grad_norm": 0.28429925441741943, "learning_rate": 1.911816341639339e-07, "loss": 0.2504, "step": 2857 }, { "epoch": 0.7605109100585418, "grad_norm": 0.30502748489379883, "learning_rate": 1.9117462766378493e-07, "loss": 0.2617, "step": 2858 }, { "epoch": 0.7607770090473657, "grad_norm": 0.2694632411003113, "learning_rate": 1.911676185097793e-07, "loss": 0.2251, "step": 2859 }, { "epoch": 0.7610431080361895, "grad_norm": 0.28640443086624146, "learning_rate": 1.9116060670212096e-07, "loss": 0.233, "step": 2860 }, { "epoch": 0.7613092070250133, "grad_norm": 0.26661786437034607, "learning_rate": 1.9115359224101414e-07, "loss": 0.2115, "step": 2861 }, { "epoch": 0.7615753060138372, "grad_norm": 0.2809983789920807, "learning_rate": 1.9114657512666285e-07, "loss": 0.221, "step": 2862 }, { "epoch": 0.761841405002661, "grad_norm": 0.2729557156562805, "learning_rate": 1.9113955535927144e-07, "loss": 0.2339, "step": 2863 }, { "epoch": 0.7621075039914849, "grad_norm": 0.2903822362422943, "learning_rate": 1.9113253293904422e-07, "loss": 0.2473, "step": 2864 }, { "epoch": 0.7623736029803087, "grad_norm": 0.3653493821620941, "learning_rate": 1.911255078661856e-07, "loss": 0.2207, "step": 2865 }, { "epoch": 0.7626397019691326, "grad_norm": 0.3358150124549866, "learning_rate": 1.9111848014090004e-07, "loss": 0.2471, "step": 2866 }, { "epoch": 0.7629058009579563, "grad_norm": 0.2926815152168274, "learning_rate": 1.911114497633921e-07, "loss": 0.2298, "step": 2867 }, { "epoch": 0.7631718999467803, "grad_norm": 0.28209879994392395, "learning_rate": 1.9110441673386645e-07, "loss": 0.2458, "step": 2868 }, { "epoch": 0.763437998935604, "grad_norm": 0.299312949180603, "learning_rate": 1.9109738105252776e-07, "loss": 0.248, "step": 2869 }, { "epoch": 0.7637040979244278, "grad_norm": 0.32337135076522827, "learning_rate": 1.9109034271958085e-07, "loss": 0.2236, "step": 2870 }, { "epoch": 0.7639701969132517, "grad_norm": 0.2699280381202698, "learning_rate": 1.9108330173523055e-07, "loss": 0.2079, "step": 2871 }, { "epoch": 0.7642362959020755, "grad_norm": 0.2795225977897644, "learning_rate": 1.9107625809968187e-07, "loss": 0.22, "step": 2872 }, { "epoch": 0.7645023948908994, "grad_norm": 0.27705833315849304, "learning_rate": 1.9106921181313977e-07, "loss": 0.2338, "step": 2873 }, { "epoch": 0.7647684938797232, "grad_norm": 0.280964195728302, "learning_rate": 1.9106216287580938e-07, "loss": 0.2459, "step": 2874 }, { "epoch": 0.7650345928685471, "grad_norm": 0.26949408650398254, "learning_rate": 1.9105511128789588e-07, "loss": 0.2119, "step": 2875 }, { "epoch": 0.7653006918573709, "grad_norm": 0.2674539089202881, "learning_rate": 1.910480570496045e-07, "loss": 0.2255, "step": 2876 }, { "epoch": 0.7655667908461948, "grad_norm": 0.3430941104888916, "learning_rate": 1.9104100016114058e-07, "loss": 0.2447, "step": 2877 }, { "epoch": 0.7658328898350186, "grad_norm": 0.27175644040107727, "learning_rate": 1.9103394062270954e-07, "loss": 0.2277, "step": 2878 }, { "epoch": 0.7660989888238424, "grad_norm": 0.32351920008659363, "learning_rate": 1.9102687843451685e-07, "loss": 0.2462, "step": 2879 }, { "epoch": 0.7663650878126663, "grad_norm": 0.27251625061035156, "learning_rate": 1.910198135967681e-07, "loss": 0.2169, "step": 2880 }, { "epoch": 0.7666311868014901, "grad_norm": 0.2822225093841553, "learning_rate": 1.9101274610966887e-07, "loss": 0.2425, "step": 2881 }, { "epoch": 0.766897285790314, "grad_norm": 0.2913682460784912, "learning_rate": 1.9100567597342492e-07, "loss": 0.2386, "step": 2882 }, { "epoch": 0.7671633847791378, "grad_norm": 0.2596149444580078, "learning_rate": 1.9099860318824205e-07, "loss": 0.2172, "step": 2883 }, { "epoch": 0.7674294837679617, "grad_norm": 0.2898721396923065, "learning_rate": 1.909915277543261e-07, "loss": 0.2356, "step": 2884 }, { "epoch": 0.7676955827567855, "grad_norm": 0.29260632395744324, "learning_rate": 1.9098444967188305e-07, "loss": 0.2223, "step": 2885 }, { "epoch": 0.7679616817456094, "grad_norm": 0.2644880414009094, "learning_rate": 1.9097736894111892e-07, "loss": 0.2159, "step": 2886 }, { "epoch": 0.7682277807344332, "grad_norm": 0.2773342430591583, "learning_rate": 1.9097028556223977e-07, "loss": 0.2295, "step": 2887 }, { "epoch": 0.768493879723257, "grad_norm": 0.2559302747249603, "learning_rate": 1.9096319953545185e-07, "loss": 0.208, "step": 2888 }, { "epoch": 0.7687599787120809, "grad_norm": 0.28689131140708923, "learning_rate": 1.9095611086096133e-07, "loss": 0.2321, "step": 2889 }, { "epoch": 0.7690260777009047, "grad_norm": 0.285703182220459, "learning_rate": 1.909490195389746e-07, "loss": 0.2169, "step": 2890 }, { "epoch": 0.7692921766897286, "grad_norm": 0.2627536356449127, "learning_rate": 1.9094192556969806e-07, "loss": 0.2246, "step": 2891 }, { "epoch": 0.7695582756785524, "grad_norm": 0.27248749136924744, "learning_rate": 1.9093482895333822e-07, "loss": 0.2149, "step": 2892 }, { "epoch": 0.7698243746673763, "grad_norm": 0.2881661057472229, "learning_rate": 1.9092772969010161e-07, "loss": 0.2522, "step": 2893 }, { "epoch": 0.7700904736562001, "grad_norm": 0.28279852867126465, "learning_rate": 1.9092062778019486e-07, "loss": 0.2404, "step": 2894 }, { "epoch": 0.770356572645024, "grad_norm": 0.25382333993911743, "learning_rate": 1.9091352322382473e-07, "loss": 0.2191, "step": 2895 }, { "epoch": 0.7706226716338478, "grad_norm": 0.3239094316959381, "learning_rate": 1.9090641602119798e-07, "loss": 0.2262, "step": 2896 }, { "epoch": 0.7708887706226716, "grad_norm": 0.29062995314598083, "learning_rate": 1.9089930617252147e-07, "loss": 0.2365, "step": 2897 }, { "epoch": 0.7711548696114955, "grad_norm": 0.29201167821884155, "learning_rate": 1.9089219367800223e-07, "loss": 0.2572, "step": 2898 }, { "epoch": 0.7714209686003193, "grad_norm": 0.30207452178001404, "learning_rate": 1.9088507853784718e-07, "loss": 0.2258, "step": 2899 }, { "epoch": 0.7716870675891432, "grad_norm": 0.30263784527778625, "learning_rate": 1.9087796075226352e-07, "loss": 0.2581, "step": 2900 }, { "epoch": 0.771953166577967, "grad_norm": 0.26406240463256836, "learning_rate": 1.9087084032145834e-07, "loss": 0.2175, "step": 2901 }, { "epoch": 0.7722192655667909, "grad_norm": 0.27621304988861084, "learning_rate": 1.9086371724563896e-07, "loss": 0.238, "step": 2902 }, { "epoch": 0.7724853645556147, "grad_norm": 0.2886875867843628, "learning_rate": 1.908565915250127e-07, "loss": 0.2244, "step": 2903 }, { "epoch": 0.7727514635444386, "grad_norm": 0.2691095471382141, "learning_rate": 1.90849463159787e-07, "loss": 0.2314, "step": 2904 }, { "epoch": 0.7730175625332624, "grad_norm": 0.28644412755966187, "learning_rate": 1.9084233215016928e-07, "loss": 0.2464, "step": 2905 }, { "epoch": 0.7732836615220862, "grad_norm": 0.26512593030929565, "learning_rate": 1.9083519849636714e-07, "loss": 0.2192, "step": 2906 }, { "epoch": 0.7735497605109101, "grad_norm": 0.3061898350715637, "learning_rate": 1.9082806219858823e-07, "loss": 0.2261, "step": 2907 }, { "epoch": 0.7738158594997339, "grad_norm": 0.3120075762271881, "learning_rate": 1.9082092325704027e-07, "loss": 0.2264, "step": 2908 }, { "epoch": 0.7740819584885578, "grad_norm": 0.2784843444824219, "learning_rate": 1.9081378167193102e-07, "loss": 0.2315, "step": 2909 }, { "epoch": 0.7743480574773816, "grad_norm": 0.2720649540424347, "learning_rate": 1.908066374434684e-07, "loss": 0.2363, "step": 2910 }, { "epoch": 0.7746141564662055, "grad_norm": 0.29200905561447144, "learning_rate": 1.9079949057186036e-07, "loss": 0.2449, "step": 2911 }, { "epoch": 0.7748802554550293, "grad_norm": 0.31790977716445923, "learning_rate": 1.9079234105731488e-07, "loss": 0.2552, "step": 2912 }, { "epoch": 0.7751463544438532, "grad_norm": 0.27543193101882935, "learning_rate": 1.907851889000401e-07, "loss": 0.2248, "step": 2913 }, { "epoch": 0.775412453432677, "grad_norm": 0.27846938371658325, "learning_rate": 1.907780341002442e-07, "loss": 0.2248, "step": 2914 }, { "epoch": 0.7756785524215007, "grad_norm": 0.27983903884887695, "learning_rate": 1.9077087665813542e-07, "loss": 0.2346, "step": 2915 }, { "epoch": 0.7759446514103246, "grad_norm": 0.33943086862564087, "learning_rate": 1.907637165739221e-07, "loss": 0.2623, "step": 2916 }, { "epoch": 0.7762107503991484, "grad_norm": 0.3038118779659271, "learning_rate": 1.9075655384781265e-07, "loss": 0.2436, "step": 2917 }, { "epoch": 0.7764768493879723, "grad_norm": 0.286533921957016, "learning_rate": 1.9074938848001557e-07, "loss": 0.2288, "step": 2918 }, { "epoch": 0.7767429483767961, "grad_norm": 0.26458367705345154, "learning_rate": 1.9074222047073944e-07, "loss": 0.2302, "step": 2919 }, { "epoch": 0.77700904736562, "grad_norm": 0.27766942977905273, "learning_rate": 1.907350498201929e-07, "loss": 0.2227, "step": 2920 }, { "epoch": 0.7772751463544438, "grad_norm": 0.2935722768306732, "learning_rate": 1.907278765285846e-07, "loss": 0.2238, "step": 2921 }, { "epoch": 0.7775412453432677, "grad_norm": 0.4383290112018585, "learning_rate": 1.9072070059612342e-07, "loss": 0.2288, "step": 2922 }, { "epoch": 0.7778073443320915, "grad_norm": 0.2794995605945587, "learning_rate": 1.9071352202301818e-07, "loss": 0.2205, "step": 2923 }, { "epoch": 0.7780734433209153, "grad_norm": 0.2747131884098053, "learning_rate": 1.9070634080947787e-07, "loss": 0.2262, "step": 2924 }, { "epoch": 0.7783395423097392, "grad_norm": 0.29276812076568604, "learning_rate": 1.9069915695571147e-07, "loss": 0.2471, "step": 2925 }, { "epoch": 0.778605641298563, "grad_norm": 0.4150083363056183, "learning_rate": 1.9069197046192814e-07, "loss": 0.2263, "step": 2926 }, { "epoch": 0.7788717402873869, "grad_norm": 0.27970072627067566, "learning_rate": 1.9068478132833699e-07, "loss": 0.2105, "step": 2927 }, { "epoch": 0.7791378392762107, "grad_norm": 0.25780290365219116, "learning_rate": 1.9067758955514735e-07, "loss": 0.2251, "step": 2928 }, { "epoch": 0.7794039382650346, "grad_norm": 0.2878689467906952, "learning_rate": 1.906703951425685e-07, "loss": 0.2467, "step": 2929 }, { "epoch": 0.7796700372538584, "grad_norm": 0.2794301509857178, "learning_rate": 1.9066319809080987e-07, "loss": 0.2326, "step": 2930 }, { "epoch": 0.7799361362426823, "grad_norm": 0.2793249785900116, "learning_rate": 1.9065599840008098e-07, "loss": 0.2215, "step": 2931 }, { "epoch": 0.7802022352315061, "grad_norm": 0.2882988452911377, "learning_rate": 1.9064879607059133e-07, "loss": 0.2232, "step": 2932 }, { "epoch": 0.78046833422033, "grad_norm": 0.3833085000514984, "learning_rate": 1.9064159110255062e-07, "loss": 0.2356, "step": 2933 }, { "epoch": 0.7807344332091538, "grad_norm": 0.26905274391174316, "learning_rate": 1.906343834961685e-07, "loss": 0.2317, "step": 2934 }, { "epoch": 0.7810005321979776, "grad_norm": 0.28443148732185364, "learning_rate": 1.906271732516548e-07, "loss": 0.2201, "step": 2935 }, { "epoch": 0.7812666311868015, "grad_norm": 0.29415372014045715, "learning_rate": 1.9061996036921945e-07, "loss": 0.2296, "step": 2936 }, { "epoch": 0.7815327301756253, "grad_norm": 0.28009137511253357, "learning_rate": 1.9061274484907232e-07, "loss": 0.2182, "step": 2937 }, { "epoch": 0.7817988291644492, "grad_norm": 0.26739633083343506, "learning_rate": 1.9060552669142347e-07, "loss": 0.2144, "step": 2938 }, { "epoch": 0.782064928153273, "grad_norm": 0.30546507239341736, "learning_rate": 1.90598305896483e-07, "loss": 0.2174, "step": 2939 }, { "epoch": 0.7823310271420969, "grad_norm": 0.2987000346183777, "learning_rate": 1.9059108246446106e-07, "loss": 0.2363, "step": 2940 }, { "epoch": 0.7825971261309207, "grad_norm": 0.3112039864063263, "learning_rate": 1.9058385639556792e-07, "loss": 0.2446, "step": 2941 }, { "epoch": 0.7828632251197446, "grad_norm": 0.27212560176849365, "learning_rate": 1.9057662769001395e-07, "loss": 0.229, "step": 2942 }, { "epoch": 0.7831293241085684, "grad_norm": 0.299898624420166, "learning_rate": 1.905693963480095e-07, "loss": 0.2238, "step": 2943 }, { "epoch": 0.7833954230973922, "grad_norm": 0.2765262722969055, "learning_rate": 1.905621623697651e-07, "loss": 0.2327, "step": 2944 }, { "epoch": 0.7836615220862161, "grad_norm": 0.2783462107181549, "learning_rate": 1.905549257554913e-07, "loss": 0.2025, "step": 2945 }, { "epoch": 0.7839276210750399, "grad_norm": 0.3023929297924042, "learning_rate": 1.9054768650539872e-07, "loss": 0.226, "step": 2946 }, { "epoch": 0.7841937200638638, "grad_norm": 0.28083136677742004, "learning_rate": 1.905404446196981e-07, "loss": 0.2258, "step": 2947 }, { "epoch": 0.7844598190526876, "grad_norm": 0.29787495732307434, "learning_rate": 1.9053320009860023e-07, "loss": 0.2002, "step": 2948 }, { "epoch": 0.7847259180415115, "grad_norm": 0.2949028015136719, "learning_rate": 1.9052595294231595e-07, "loss": 0.2333, "step": 2949 }, { "epoch": 0.7849920170303353, "grad_norm": 0.2824547588825226, "learning_rate": 1.9051870315105622e-07, "loss": 0.2435, "step": 2950 }, { "epoch": 0.7852581160191592, "grad_norm": 0.27188560366630554, "learning_rate": 1.9051145072503211e-07, "loss": 0.2232, "step": 2951 }, { "epoch": 0.785524215007983, "grad_norm": 0.27251771092414856, "learning_rate": 1.9050419566445467e-07, "loss": 0.229, "step": 2952 }, { "epoch": 0.7857903139968068, "grad_norm": 0.2817675471305847, "learning_rate": 1.9049693796953506e-07, "loss": 0.2155, "step": 2953 }, { "epoch": 0.7860564129856307, "grad_norm": 0.29782751202583313, "learning_rate": 1.9048967764048457e-07, "loss": 0.2386, "step": 2954 }, { "epoch": 0.7863225119744545, "grad_norm": 0.3221728205680847, "learning_rate": 1.9048241467751454e-07, "loss": 0.2422, "step": 2955 }, { "epoch": 0.7865886109632784, "grad_norm": 0.2653117775917053, "learning_rate": 1.9047514908083633e-07, "loss": 0.2354, "step": 2956 }, { "epoch": 0.7868547099521022, "grad_norm": 0.27991074323654175, "learning_rate": 1.9046788085066145e-07, "loss": 0.2337, "step": 2957 }, { "epoch": 0.7871208089409261, "grad_norm": 0.2704961597919464, "learning_rate": 1.9046060998720148e-07, "loss": 0.2171, "step": 2958 }, { "epoch": 0.7873869079297499, "grad_norm": 0.29266396164894104, "learning_rate": 1.9045333649066802e-07, "loss": 0.2481, "step": 2959 }, { "epoch": 0.7876530069185738, "grad_norm": 0.2631945312023163, "learning_rate": 1.9044606036127278e-07, "loss": 0.2365, "step": 2960 }, { "epoch": 0.7879191059073976, "grad_norm": 0.29000556468963623, "learning_rate": 1.9043878159922758e-07, "loss": 0.2282, "step": 2961 }, { "epoch": 0.7881852048962213, "grad_norm": 0.2672199606895447, "learning_rate": 1.9043150020474426e-07, "loss": 0.2146, "step": 2962 }, { "epoch": 0.7884513038850453, "grad_norm": 0.3098756968975067, "learning_rate": 1.9042421617803476e-07, "loss": 0.2548, "step": 2963 }, { "epoch": 0.788717402873869, "grad_norm": 0.27773937582969666, "learning_rate": 1.9041692951931112e-07, "loss": 0.2187, "step": 2964 }, { "epoch": 0.788983501862693, "grad_norm": 0.27979427576065063, "learning_rate": 1.9040964022878544e-07, "loss": 0.2031, "step": 2965 }, { "epoch": 0.7892496008515167, "grad_norm": 0.2563423216342926, "learning_rate": 1.9040234830666985e-07, "loss": 0.2165, "step": 2966 }, { "epoch": 0.7895156998403406, "grad_norm": 0.27731186151504517, "learning_rate": 1.9039505375317665e-07, "loss": 0.2166, "step": 2967 }, { "epoch": 0.7897817988291644, "grad_norm": 0.2843925356864929, "learning_rate": 1.9038775656851812e-07, "loss": 0.224, "step": 2968 }, { "epoch": 0.7900478978179883, "grad_norm": 0.27406150102615356, "learning_rate": 1.9038045675290673e-07, "loss": 0.2093, "step": 2969 }, { "epoch": 0.7903139968068121, "grad_norm": 0.27104172110557556, "learning_rate": 1.903731543065549e-07, "loss": 0.2184, "step": 2970 }, { "epoch": 0.7905800957956359, "grad_norm": 0.2863135039806366, "learning_rate": 1.903658492296752e-07, "loss": 0.2284, "step": 2971 }, { "epoch": 0.7908461947844598, "grad_norm": 0.28999096155166626, "learning_rate": 1.9035854152248022e-07, "loss": 0.251, "step": 2972 }, { "epoch": 0.7911122937732836, "grad_norm": 0.29191234707832336, "learning_rate": 1.9035123118518274e-07, "loss": 0.234, "step": 2973 }, { "epoch": 0.7913783927621075, "grad_norm": 0.2714828848838806, "learning_rate": 1.9034391821799554e-07, "loss": 0.2236, "step": 2974 }, { "epoch": 0.7916444917509313, "grad_norm": 0.31179821491241455, "learning_rate": 1.9033660262113144e-07, "loss": 0.2441, "step": 2975 }, { "epoch": 0.7919105907397552, "grad_norm": 0.4208495020866394, "learning_rate": 1.903292843948034e-07, "loss": 0.231, "step": 2976 }, { "epoch": 0.792176689728579, "grad_norm": 0.2626569867134094, "learning_rate": 1.9032196353922438e-07, "loss": 0.2188, "step": 2977 }, { "epoch": 0.7924427887174029, "grad_norm": 0.26772788166999817, "learning_rate": 1.903146400546076e-07, "loss": 0.2053, "step": 2978 }, { "epoch": 0.7927088877062267, "grad_norm": 0.27947428822517395, "learning_rate": 1.9030731394116608e-07, "loss": 0.2264, "step": 2979 }, { "epoch": 0.7929749866950505, "grad_norm": 0.275646835565567, "learning_rate": 1.9029998519911315e-07, "loss": 0.2315, "step": 2980 }, { "epoch": 0.7932410856838744, "grad_norm": 0.2772839069366455, "learning_rate": 1.9029265382866213e-07, "loss": 0.2194, "step": 2981 }, { "epoch": 0.7935071846726982, "grad_norm": 0.31172528862953186, "learning_rate": 1.9028531983002636e-07, "loss": 0.2626, "step": 2982 }, { "epoch": 0.7937732836615221, "grad_norm": 0.2942202687263489, "learning_rate": 1.9027798320341938e-07, "loss": 0.231, "step": 2983 }, { "epoch": 0.7940393826503459, "grad_norm": 0.33172592520713806, "learning_rate": 1.9027064394905473e-07, "loss": 0.249, "step": 2984 }, { "epoch": 0.7943054816391698, "grad_norm": 0.30484646558761597, "learning_rate": 1.90263302067146e-07, "loss": 0.2339, "step": 2985 }, { "epoch": 0.7945715806279936, "grad_norm": 0.2814132869243622, "learning_rate": 1.9025595755790694e-07, "loss": 0.23, "step": 2986 }, { "epoch": 0.7948376796168175, "grad_norm": 0.2754591107368469, "learning_rate": 1.9024861042155126e-07, "loss": 0.2368, "step": 2987 }, { "epoch": 0.7951037786056413, "grad_norm": 0.2726380228996277, "learning_rate": 1.9024126065829286e-07, "loss": 0.2326, "step": 2988 }, { "epoch": 0.7953698775944651, "grad_norm": 0.26836496591567993, "learning_rate": 1.9023390826834572e-07, "loss": 0.2232, "step": 2989 }, { "epoch": 0.795635976583289, "grad_norm": 0.30658984184265137, "learning_rate": 1.9022655325192376e-07, "loss": 0.245, "step": 2990 }, { "epoch": 0.7959020755721128, "grad_norm": 0.2831437587738037, "learning_rate": 1.902191956092411e-07, "loss": 0.2354, "step": 2991 }, { "epoch": 0.7961681745609367, "grad_norm": 0.2631738483905792, "learning_rate": 1.9021183534051195e-07, "loss": 0.2196, "step": 2992 }, { "epoch": 0.7964342735497605, "grad_norm": 0.28961846232414246, "learning_rate": 1.9020447244595048e-07, "loss": 0.238, "step": 2993 }, { "epoch": 0.7967003725385844, "grad_norm": 0.3226294219493866, "learning_rate": 1.9019710692577102e-07, "loss": 0.223, "step": 2994 }, { "epoch": 0.7969664715274082, "grad_norm": 0.2639995515346527, "learning_rate": 1.9018973878018798e-07, "loss": 0.2178, "step": 2995 }, { "epoch": 0.7972325705162321, "grad_norm": 0.2823006212711334, "learning_rate": 1.9018236800941582e-07, "loss": 0.2346, "step": 2996 }, { "epoch": 0.7974986695050559, "grad_norm": 0.30969002842903137, "learning_rate": 1.9017499461366907e-07, "loss": 0.2415, "step": 2997 }, { "epoch": 0.7977647684938797, "grad_norm": 0.29178687930107117, "learning_rate": 1.901676185931624e-07, "loss": 0.2296, "step": 2998 }, { "epoch": 0.7980308674827036, "grad_norm": 0.2686311900615692, "learning_rate": 1.9016023994811044e-07, "loss": 0.2412, "step": 2999 }, { "epoch": 0.7982969664715274, "grad_norm": 0.2790650427341461, "learning_rate": 1.9015285867872801e-07, "loss": 0.2297, "step": 3000 }, { "epoch": 0.7985630654603513, "grad_norm": 0.27559566497802734, "learning_rate": 1.9014547478522994e-07, "loss": 0.22, "step": 3001 }, { "epoch": 0.7988291644491751, "grad_norm": 0.29122695326805115, "learning_rate": 1.9013808826783116e-07, "loss": 0.2233, "step": 3002 }, { "epoch": 0.799095263437999, "grad_norm": 0.3014811873435974, "learning_rate": 1.9013069912674667e-07, "loss": 0.2424, "step": 3003 }, { "epoch": 0.7993613624268228, "grad_norm": 0.2746402323246002, "learning_rate": 1.9012330736219154e-07, "loss": 0.2139, "step": 3004 }, { "epoch": 0.7996274614156467, "grad_norm": 0.28725364804267883, "learning_rate": 1.9011591297438094e-07, "loss": 0.223, "step": 3005 }, { "epoch": 0.7998935604044705, "grad_norm": 0.2989172339439392, "learning_rate": 1.901085159635301e-07, "loss": 0.2406, "step": 3006 }, { "epoch": 0.8001596593932943, "grad_norm": 0.29145291447639465, "learning_rate": 1.9010111632985434e-07, "loss": 0.2439, "step": 3007 }, { "epoch": 0.8004257583821182, "grad_norm": 0.2776554822921753, "learning_rate": 1.90093714073569e-07, "loss": 0.2155, "step": 3008 }, { "epoch": 0.800691857370942, "grad_norm": 0.27028393745422363, "learning_rate": 1.9008630919488958e-07, "loss": 0.2189, "step": 3009 }, { "epoch": 0.8009579563597659, "grad_norm": 0.30401232838630676, "learning_rate": 1.9007890169403161e-07, "loss": 0.2237, "step": 3010 }, { "epoch": 0.8012240553485896, "grad_norm": 0.26990529894828796, "learning_rate": 1.900714915712107e-07, "loss": 0.2497, "step": 3011 }, { "epoch": 0.8014901543374136, "grad_norm": 0.28480780124664307, "learning_rate": 1.9006407882664255e-07, "loss": 0.2313, "step": 3012 }, { "epoch": 0.8017562533262373, "grad_norm": 0.2802984416484833, "learning_rate": 1.900566634605429e-07, "loss": 0.2223, "step": 3013 }, { "epoch": 0.8020223523150612, "grad_norm": 0.2800939977169037, "learning_rate": 1.900492454731276e-07, "loss": 0.2261, "step": 3014 }, { "epoch": 0.802288451303885, "grad_norm": 0.2949811816215515, "learning_rate": 1.900418248646126e-07, "loss": 0.2223, "step": 3015 }, { "epoch": 0.8025545502927088, "grad_norm": 0.2732856869697571, "learning_rate": 1.9003440163521382e-07, "loss": 0.2345, "step": 3016 }, { "epoch": 0.8028206492815327, "grad_norm": 0.2485295832157135, "learning_rate": 1.9002697578514742e-07, "loss": 0.2179, "step": 3017 }, { "epoch": 0.8030867482703565, "grad_norm": 0.2900030016899109, "learning_rate": 1.900195473146295e-07, "loss": 0.2149, "step": 3018 }, { "epoch": 0.8033528472591804, "grad_norm": 0.2898365557193756, "learning_rate": 1.900121162238763e-07, "loss": 0.2489, "step": 3019 }, { "epoch": 0.8036189462480042, "grad_norm": 0.27188706398010254, "learning_rate": 1.900046825131041e-07, "loss": 0.2237, "step": 3020 }, { "epoch": 0.8038850452368281, "grad_norm": 0.3074406683444977, "learning_rate": 1.899972461825293e-07, "loss": 0.2185, "step": 3021 }, { "epoch": 0.8041511442256519, "grad_norm": 0.3004029393196106, "learning_rate": 1.8998980723236833e-07, "loss": 0.2506, "step": 3022 }, { "epoch": 0.8044172432144758, "grad_norm": 0.3014337718486786, "learning_rate": 1.899823656628377e-07, "loss": 0.2554, "step": 3023 }, { "epoch": 0.8046833422032996, "grad_norm": 0.27891847491264343, "learning_rate": 1.8997492147415407e-07, "loss": 0.2358, "step": 3024 }, { "epoch": 0.8049494411921234, "grad_norm": 0.2767167389392853, "learning_rate": 1.899674746665341e-07, "loss": 0.2276, "step": 3025 }, { "epoch": 0.8052155401809473, "grad_norm": 0.29235514998435974, "learning_rate": 1.8996002524019454e-07, "loss": 0.2341, "step": 3026 }, { "epoch": 0.8054816391697711, "grad_norm": 0.29622921347618103, "learning_rate": 1.8995257319535223e-07, "loss": 0.233, "step": 3027 }, { "epoch": 0.805747738158595, "grad_norm": 0.2915458083152771, "learning_rate": 1.8994511853222405e-07, "loss": 0.224, "step": 3028 }, { "epoch": 0.8060138371474188, "grad_norm": 0.2758224904537201, "learning_rate": 1.89937661251027e-07, "loss": 0.2275, "step": 3029 }, { "epoch": 0.8062799361362427, "grad_norm": 0.2636059820652008, "learning_rate": 1.8993020135197818e-07, "loss": 0.2103, "step": 3030 }, { "epoch": 0.8065460351250665, "grad_norm": 0.2893470227718353, "learning_rate": 1.8992273883529471e-07, "loss": 0.2122, "step": 3031 }, { "epoch": 0.8068121341138904, "grad_norm": 0.29117661714553833, "learning_rate": 1.8991527370119375e-07, "loss": 0.2415, "step": 3032 }, { "epoch": 0.8070782331027142, "grad_norm": 0.270681768655777, "learning_rate": 1.8990780594989267e-07, "loss": 0.2156, "step": 3033 }, { "epoch": 0.807344332091538, "grad_norm": 0.26387226581573486, "learning_rate": 1.8990033558160882e-07, "loss": 0.2083, "step": 3034 }, { "epoch": 0.8076104310803619, "grad_norm": 0.2738587558269501, "learning_rate": 1.8989286259655958e-07, "loss": 0.2142, "step": 3035 }, { "epoch": 0.8078765300691857, "grad_norm": 0.26886704564094543, "learning_rate": 1.8988538699496255e-07, "loss": 0.2012, "step": 3036 }, { "epoch": 0.8081426290580096, "grad_norm": 0.25992298126220703, "learning_rate": 1.8987790877703527e-07, "loss": 0.2065, "step": 3037 }, { "epoch": 0.8084087280468334, "grad_norm": 0.2946590781211853, "learning_rate": 1.898704279429954e-07, "loss": 0.2543, "step": 3038 }, { "epoch": 0.8086748270356573, "grad_norm": 0.2753261923789978, "learning_rate": 1.8986294449306078e-07, "loss": 0.2341, "step": 3039 }, { "epoch": 0.8089409260244811, "grad_norm": 0.27482807636260986, "learning_rate": 1.8985545842744914e-07, "loss": 0.2243, "step": 3040 }, { "epoch": 0.809207025013305, "grad_norm": 0.2856035828590393, "learning_rate": 1.8984796974637842e-07, "loss": 0.2382, "step": 3041 }, { "epoch": 0.8094731240021288, "grad_norm": 0.28031834959983826, "learning_rate": 1.898404784500666e-07, "loss": 0.221, "step": 3042 }, { "epoch": 0.8097392229909527, "grad_norm": 0.29949718713760376, "learning_rate": 1.898329845387317e-07, "loss": 0.2175, "step": 3043 }, { "epoch": 0.8100053219797765, "grad_norm": 0.30952921509742737, "learning_rate": 1.8982548801259188e-07, "loss": 0.2352, "step": 3044 }, { "epoch": 0.8102714209686003, "grad_norm": 0.2990545630455017, "learning_rate": 1.898179888718653e-07, "loss": 0.2512, "step": 3045 }, { "epoch": 0.8105375199574242, "grad_norm": 0.3050568997859955, "learning_rate": 1.8981048711677032e-07, "loss": 0.2413, "step": 3046 }, { "epoch": 0.810803618946248, "grad_norm": 0.2785244882106781, "learning_rate": 1.8980298274752525e-07, "loss": 0.2328, "step": 3047 }, { "epoch": 0.8110697179350719, "grad_norm": 0.3011764585971832, "learning_rate": 1.8979547576434852e-07, "loss": 0.2388, "step": 3048 }, { "epoch": 0.8113358169238957, "grad_norm": 0.29180237650871277, "learning_rate": 1.897879661674586e-07, "loss": 0.2452, "step": 3049 }, { "epoch": 0.8116019159127196, "grad_norm": 0.2793327867984772, "learning_rate": 1.8978045395707417e-07, "loss": 0.2456, "step": 3050 }, { "epoch": 0.8118680149015434, "grad_norm": 0.28374025225639343, "learning_rate": 1.8977293913341378e-07, "loss": 0.2328, "step": 3051 }, { "epoch": 0.8121341138903673, "grad_norm": 0.2929835617542267, "learning_rate": 1.8976542169669627e-07, "loss": 0.2453, "step": 3052 }, { "epoch": 0.8124002128791911, "grad_norm": 0.2737928330898285, "learning_rate": 1.897579016471404e-07, "loss": 0.2217, "step": 3053 }, { "epoch": 0.8126663118680149, "grad_norm": 0.28196826577186584, "learning_rate": 1.897503789849651e-07, "loss": 0.2288, "step": 3054 }, { "epoch": 0.8129324108568388, "grad_norm": 0.28299906849861145, "learning_rate": 1.8974285371038925e-07, "loss": 0.2253, "step": 3055 }, { "epoch": 0.8131985098456626, "grad_norm": 0.3009043037891388, "learning_rate": 1.8973532582363195e-07, "loss": 0.2258, "step": 3056 }, { "epoch": 0.8134646088344865, "grad_norm": 0.28196287155151367, "learning_rate": 1.8972779532491232e-07, "loss": 0.2488, "step": 3057 }, { "epoch": 0.8137307078233103, "grad_norm": 0.2973261773586273, "learning_rate": 1.8972026221444956e-07, "loss": 0.2386, "step": 3058 }, { "epoch": 0.8139968068121342, "grad_norm": 0.26691415905952454, "learning_rate": 1.897127264924629e-07, "loss": 0.2099, "step": 3059 }, { "epoch": 0.814262905800958, "grad_norm": 0.2807754874229431, "learning_rate": 1.8970518815917172e-07, "loss": 0.2237, "step": 3060 }, { "epoch": 0.8145290047897819, "grad_norm": 0.29036185145378113, "learning_rate": 1.8969764721479543e-07, "loss": 0.238, "step": 3061 }, { "epoch": 0.8147951037786056, "grad_norm": 0.2901204824447632, "learning_rate": 1.896901036595535e-07, "loss": 0.2383, "step": 3062 }, { "epoch": 0.8150612027674294, "grad_norm": 0.2667819559574127, "learning_rate": 1.8968255749366557e-07, "loss": 0.2242, "step": 3063 }, { "epoch": 0.8153273017562533, "grad_norm": 0.28478628396987915, "learning_rate": 1.8967500871735124e-07, "loss": 0.2334, "step": 3064 }, { "epoch": 0.8155934007450771, "grad_norm": 0.27955108880996704, "learning_rate": 1.8966745733083025e-07, "loss": 0.2185, "step": 3065 }, { "epoch": 0.815859499733901, "grad_norm": 0.284407377243042, "learning_rate": 1.8965990333432238e-07, "loss": 0.2322, "step": 3066 }, { "epoch": 0.8161255987227248, "grad_norm": 0.31097495555877686, "learning_rate": 1.8965234672804752e-07, "loss": 0.2355, "step": 3067 }, { "epoch": 0.8163916977115487, "grad_norm": 0.29278600215911865, "learning_rate": 1.8964478751222562e-07, "loss": 0.24, "step": 3068 }, { "epoch": 0.8166577967003725, "grad_norm": 0.29061421751976013, "learning_rate": 1.8963722568707677e-07, "loss": 0.2373, "step": 3069 }, { "epoch": 0.8169238956891964, "grad_norm": 0.2977801263332367, "learning_rate": 1.8962966125282096e-07, "loss": 0.2223, "step": 3070 }, { "epoch": 0.8171899946780202, "grad_norm": 0.29206958413124084, "learning_rate": 1.8962209420967845e-07, "loss": 0.2492, "step": 3071 }, { "epoch": 0.817456093666844, "grad_norm": 0.2701861262321472, "learning_rate": 1.896145245578695e-07, "loss": 0.2214, "step": 3072 }, { "epoch": 0.8177221926556679, "grad_norm": 0.2826937437057495, "learning_rate": 1.896069522976144e-07, "loss": 0.239, "step": 3073 }, { "epoch": 0.8179882916444917, "grad_norm": 0.2832473814487457, "learning_rate": 1.8959937742913357e-07, "loss": 0.2205, "step": 3074 }, { "epoch": 0.8182543906333156, "grad_norm": 0.2653403878211975, "learning_rate": 1.895917999526475e-07, "loss": 0.1991, "step": 3075 }, { "epoch": 0.8185204896221394, "grad_norm": 0.28189322352409363, "learning_rate": 1.8958421986837681e-07, "loss": 0.2119, "step": 3076 }, { "epoch": 0.8187865886109633, "grad_norm": 0.2971813976764679, "learning_rate": 1.8957663717654206e-07, "loss": 0.2666, "step": 3077 }, { "epoch": 0.8190526875997871, "grad_norm": 0.25124162435531616, "learning_rate": 1.8956905187736397e-07, "loss": 0.1959, "step": 3078 }, { "epoch": 0.819318786588611, "grad_norm": 0.29418492317199707, "learning_rate": 1.8956146397106334e-07, "loss": 0.2334, "step": 3079 }, { "epoch": 0.8195848855774348, "grad_norm": 0.25189125537872314, "learning_rate": 1.8955387345786106e-07, "loss": 0.1864, "step": 3080 }, { "epoch": 0.8198509845662586, "grad_norm": 0.2944772243499756, "learning_rate": 1.89546280337978e-07, "loss": 0.259, "step": 3081 }, { "epoch": 0.8201170835550825, "grad_norm": 0.2699503004550934, "learning_rate": 1.895386846116353e-07, "loss": 0.2077, "step": 3082 }, { "epoch": 0.8203831825439063, "grad_norm": 0.27492034435272217, "learning_rate": 1.8953108627905392e-07, "loss": 0.2178, "step": 3083 }, { "epoch": 0.8206492815327302, "grad_norm": 0.28537631034851074, "learning_rate": 1.895234853404551e-07, "loss": 0.2238, "step": 3084 }, { "epoch": 0.820915380521554, "grad_norm": 0.31654536724090576, "learning_rate": 1.8951588179606007e-07, "loss": 0.2429, "step": 3085 }, { "epoch": 0.8211814795103779, "grad_norm": 0.2814430594444275, "learning_rate": 1.8950827564609013e-07, "loss": 0.2108, "step": 3086 }, { "epoch": 0.8214475784992017, "grad_norm": 0.2738760709762573, "learning_rate": 1.8950066689076672e-07, "loss": 0.231, "step": 3087 }, { "epoch": 0.8217136774880256, "grad_norm": 0.2731776237487793, "learning_rate": 1.8949305553031126e-07, "loss": 0.206, "step": 3088 }, { "epoch": 0.8219797764768494, "grad_norm": 0.26797592639923096, "learning_rate": 1.8948544156494532e-07, "loss": 0.2378, "step": 3089 }, { "epoch": 0.8222458754656732, "grad_norm": 0.2639763057231903, "learning_rate": 1.8947782499489054e-07, "loss": 0.2076, "step": 3090 }, { "epoch": 0.8225119744544971, "grad_norm": 0.273502916097641, "learning_rate": 1.8947020582036858e-07, "loss": 0.2268, "step": 3091 }, { "epoch": 0.8227780734433209, "grad_norm": 0.2878728210926056, "learning_rate": 1.8946258404160126e-07, "loss": 0.2364, "step": 3092 }, { "epoch": 0.8230441724321448, "grad_norm": 0.2690306007862091, "learning_rate": 1.8945495965881038e-07, "loss": 0.2309, "step": 3093 }, { "epoch": 0.8233102714209686, "grad_norm": 0.28079596161842346, "learning_rate": 1.894473326722179e-07, "loss": 0.2304, "step": 3094 }, { "epoch": 0.8235763704097925, "grad_norm": 0.2852667272090912, "learning_rate": 1.8943970308204582e-07, "loss": 0.2314, "step": 3095 }, { "epoch": 0.8238424693986163, "grad_norm": 0.2738044261932373, "learning_rate": 1.8943207088851621e-07, "loss": 0.2011, "step": 3096 }, { "epoch": 0.8241085683874402, "grad_norm": 0.304749995470047, "learning_rate": 1.8942443609185123e-07, "loss": 0.2464, "step": 3097 }, { "epoch": 0.824374667376264, "grad_norm": 0.27342095971107483, "learning_rate": 1.8941679869227308e-07, "loss": 0.2252, "step": 3098 }, { "epoch": 0.8246407663650878, "grad_norm": 0.27744221687316895, "learning_rate": 1.8940915869000415e-07, "loss": 0.219, "step": 3099 }, { "epoch": 0.8249068653539117, "grad_norm": 0.29106324911117554, "learning_rate": 1.8940151608526669e-07, "loss": 0.2379, "step": 3100 }, { "epoch": 0.8251729643427355, "grad_norm": 0.2957057058811188, "learning_rate": 1.8939387087828324e-07, "loss": 0.2397, "step": 3101 }, { "epoch": 0.8254390633315594, "grad_norm": 0.2551276683807373, "learning_rate": 1.8938622306927635e-07, "loss": 0.2228, "step": 3102 }, { "epoch": 0.8257051623203832, "grad_norm": 0.29453423619270325, "learning_rate": 1.8937857265846854e-07, "loss": 0.2344, "step": 3103 }, { "epoch": 0.8259712613092071, "grad_norm": 0.2770724594593048, "learning_rate": 1.893709196460826e-07, "loss": 0.2205, "step": 3104 }, { "epoch": 0.8262373602980309, "grad_norm": 0.2886373698711395, "learning_rate": 1.8936326403234123e-07, "loss": 0.2236, "step": 3105 }, { "epoch": 0.8265034592868548, "grad_norm": 0.2585175037384033, "learning_rate": 1.8935560581746722e-07, "loss": 0.216, "step": 3106 }, { "epoch": 0.8267695582756786, "grad_norm": 0.27695393562316895, "learning_rate": 1.893479450016836e-07, "loss": 0.2216, "step": 3107 }, { "epoch": 0.8270356572645023, "grad_norm": 0.2905639410018921, "learning_rate": 1.893402815852133e-07, "loss": 0.2225, "step": 3108 }, { "epoch": 0.8273017562533262, "grad_norm": 0.3042020797729492, "learning_rate": 1.8933261556827933e-07, "loss": 0.2245, "step": 3109 }, { "epoch": 0.82756785524215, "grad_norm": 0.29376310110092163, "learning_rate": 1.8932494695110487e-07, "loss": 0.2131, "step": 3110 }, { "epoch": 0.827833954230974, "grad_norm": 0.2676329016685486, "learning_rate": 1.8931727573391315e-07, "loss": 0.2274, "step": 3111 }, { "epoch": 0.8281000532197977, "grad_norm": 0.2779788076877594, "learning_rate": 1.8930960191692743e-07, "loss": 0.2339, "step": 3112 }, { "epoch": 0.8283661522086216, "grad_norm": 0.3020465075969696, "learning_rate": 1.8930192550037108e-07, "loss": 0.2393, "step": 3113 }, { "epoch": 0.8286322511974454, "grad_norm": 0.2715036869049072, "learning_rate": 1.8929424648446758e-07, "loss": 0.2161, "step": 3114 }, { "epoch": 0.8288983501862693, "grad_norm": 0.2756093144416809, "learning_rate": 1.892865648694404e-07, "loss": 0.2082, "step": 3115 }, { "epoch": 0.8291644491750931, "grad_norm": 0.29170966148376465, "learning_rate": 1.8927888065551316e-07, "loss": 0.2179, "step": 3116 }, { "epoch": 0.8294305481639169, "grad_norm": 0.29789984226226807, "learning_rate": 1.892711938429095e-07, "loss": 0.2448, "step": 3117 }, { "epoch": 0.8296966471527408, "grad_norm": 0.2591983675956726, "learning_rate": 1.8926350443185318e-07, "loss": 0.2191, "step": 3118 }, { "epoch": 0.8299627461415646, "grad_norm": 0.29237550497055054, "learning_rate": 1.89255812422568e-07, "loss": 0.2393, "step": 3119 }, { "epoch": 0.8302288451303885, "grad_norm": 0.28335171937942505, "learning_rate": 1.8924811781527792e-07, "loss": 0.2449, "step": 3120 }, { "epoch": 0.8304949441192123, "grad_norm": 0.28775936365127563, "learning_rate": 1.8924042061020683e-07, "loss": 0.213, "step": 3121 }, { "epoch": 0.8307610431080362, "grad_norm": 0.29463350772857666, "learning_rate": 1.892327208075788e-07, "loss": 0.2203, "step": 3122 }, { "epoch": 0.83102714209686, "grad_norm": 0.2858157455921173, "learning_rate": 1.8922501840761794e-07, "loss": 0.2133, "step": 3123 }, { "epoch": 0.8312932410856839, "grad_norm": 0.2772967219352722, "learning_rate": 1.892173134105485e-07, "loss": 0.2293, "step": 3124 }, { "epoch": 0.8315593400745077, "grad_norm": 0.2794767916202545, "learning_rate": 1.8920960581659472e-07, "loss": 0.2416, "step": 3125 }, { "epoch": 0.8318254390633315, "grad_norm": 0.28154629468917847, "learning_rate": 1.8920189562598092e-07, "loss": 0.2349, "step": 3126 }, { "epoch": 0.8320915380521554, "grad_norm": 0.2763676643371582, "learning_rate": 1.8919418283893155e-07, "loss": 0.1992, "step": 3127 }, { "epoch": 0.8323576370409792, "grad_norm": 0.2912153899669647, "learning_rate": 1.8918646745567113e-07, "loss": 0.231, "step": 3128 }, { "epoch": 0.8326237360298031, "grad_norm": 0.2907324731349945, "learning_rate": 1.8917874947642418e-07, "loss": 0.2271, "step": 3129 }, { "epoch": 0.8328898350186269, "grad_norm": 0.29683157801628113, "learning_rate": 1.8917102890141542e-07, "loss": 0.2426, "step": 3130 }, { "epoch": 0.8331559340074508, "grad_norm": 0.30650389194488525, "learning_rate": 1.891633057308695e-07, "loss": 0.2409, "step": 3131 }, { "epoch": 0.8334220329962746, "grad_norm": 0.29296883940696716, "learning_rate": 1.8915557996501128e-07, "loss": 0.2206, "step": 3132 }, { "epoch": 0.8336881319850985, "grad_norm": 0.2947395443916321, "learning_rate": 1.891478516040656e-07, "loss": 0.2282, "step": 3133 }, { "epoch": 0.8339542309739223, "grad_norm": 0.2766647934913635, "learning_rate": 1.8914012064825747e-07, "loss": 0.2209, "step": 3134 }, { "epoch": 0.8342203299627461, "grad_norm": 0.2649965286254883, "learning_rate": 1.8913238709781185e-07, "loss": 0.224, "step": 3135 }, { "epoch": 0.83448642895157, "grad_norm": 0.2936069965362549, "learning_rate": 1.8912465095295386e-07, "loss": 0.2327, "step": 3136 }, { "epoch": 0.8347525279403938, "grad_norm": 0.30367058515548706, "learning_rate": 1.8911691221390873e-07, "loss": 0.2274, "step": 3137 }, { "epoch": 0.8350186269292177, "grad_norm": 0.287409245967865, "learning_rate": 1.8910917088090165e-07, "loss": 0.2078, "step": 3138 }, { "epoch": 0.8352847259180415, "grad_norm": 0.3187166452407837, "learning_rate": 1.8910142695415798e-07, "loss": 0.2416, "step": 3139 }, { "epoch": 0.8355508249068654, "grad_norm": 0.2744429111480713, "learning_rate": 1.8909368043390313e-07, "loss": 0.2313, "step": 3140 }, { "epoch": 0.8358169238956892, "grad_norm": 0.2840537130832672, "learning_rate": 1.8908593132036254e-07, "loss": 0.2313, "step": 3141 }, { "epoch": 0.8360830228845131, "grad_norm": 0.284262090921402, "learning_rate": 1.8907817961376186e-07, "loss": 0.2332, "step": 3142 }, { "epoch": 0.8363491218733369, "grad_norm": 0.27622291445732117, "learning_rate": 1.8907042531432659e-07, "loss": 0.2269, "step": 3143 }, { "epoch": 0.8366152208621607, "grad_norm": 0.2718479335308075, "learning_rate": 1.8906266842228255e-07, "loss": 0.2343, "step": 3144 }, { "epoch": 0.8368813198509846, "grad_norm": 0.2786617875099182, "learning_rate": 1.8905490893785548e-07, "loss": 0.2275, "step": 3145 }, { "epoch": 0.8371474188398084, "grad_norm": 0.2833912670612335, "learning_rate": 1.890471468612712e-07, "loss": 0.2335, "step": 3146 }, { "epoch": 0.8374135178286323, "grad_norm": 0.2814355492591858, "learning_rate": 1.890393821927557e-07, "loss": 0.226, "step": 3147 }, { "epoch": 0.8376796168174561, "grad_norm": 0.29113852977752686, "learning_rate": 1.8903161493253502e-07, "loss": 0.212, "step": 3148 }, { "epoch": 0.83794571580628, "grad_norm": 0.26660364866256714, "learning_rate": 1.8902384508083517e-07, "loss": 0.2197, "step": 3149 }, { "epoch": 0.8382118147951038, "grad_norm": 0.28210362792015076, "learning_rate": 1.890160726378823e-07, "loss": 0.224, "step": 3150 }, { "epoch": 0.8384779137839277, "grad_norm": 0.3026793897151947, "learning_rate": 1.8900829760390274e-07, "loss": 0.2278, "step": 3151 }, { "epoch": 0.8387440127727515, "grad_norm": 0.31562259793281555, "learning_rate": 1.890005199791227e-07, "loss": 0.2387, "step": 3152 }, { "epoch": 0.8390101117615754, "grad_norm": 0.2890923321247101, "learning_rate": 1.8899273976376864e-07, "loss": 0.2491, "step": 3153 }, { "epoch": 0.8392762107503992, "grad_norm": 0.2735275626182556, "learning_rate": 1.8898495695806697e-07, "loss": 0.2316, "step": 3154 }, { "epoch": 0.839542309739223, "grad_norm": 0.2878872752189636, "learning_rate": 1.8897717156224427e-07, "loss": 0.2346, "step": 3155 }, { "epoch": 0.8398084087280469, "grad_norm": 0.25244301557540894, "learning_rate": 1.8896938357652712e-07, "loss": 0.2015, "step": 3156 }, { "epoch": 0.8400745077168706, "grad_norm": 0.295608252286911, "learning_rate": 1.8896159300114222e-07, "loss": 0.2351, "step": 3157 }, { "epoch": 0.8403406067056945, "grad_norm": 0.2887301743030548, "learning_rate": 1.8895379983631634e-07, "loss": 0.2198, "step": 3158 }, { "epoch": 0.8406067056945183, "grad_norm": 0.3017563223838806, "learning_rate": 1.8894600408227632e-07, "loss": 0.2325, "step": 3159 }, { "epoch": 0.8408728046833422, "grad_norm": 0.27143800258636475, "learning_rate": 1.8893820573924906e-07, "loss": 0.213, "step": 3160 }, { "epoch": 0.841138903672166, "grad_norm": 0.2890055477619171, "learning_rate": 1.8893040480746157e-07, "loss": 0.2208, "step": 3161 }, { "epoch": 0.8414050026609899, "grad_norm": 0.2882709801197052, "learning_rate": 1.8892260128714087e-07, "loss": 0.2402, "step": 3162 }, { "epoch": 0.8416711016498137, "grad_norm": 0.29348522424697876, "learning_rate": 1.8891479517851417e-07, "loss": 0.2305, "step": 3163 }, { "epoch": 0.8419372006386375, "grad_norm": 0.27436119318008423, "learning_rate": 1.889069864818086e-07, "loss": 0.2153, "step": 3164 }, { "epoch": 0.8422032996274614, "grad_norm": 0.2988050878047943, "learning_rate": 1.8889917519725156e-07, "loss": 0.2542, "step": 3165 }, { "epoch": 0.8424693986162852, "grad_norm": 0.29488658905029297, "learning_rate": 1.8889136132507031e-07, "loss": 0.2441, "step": 3166 }, { "epoch": 0.8427354976051091, "grad_norm": 0.2767471671104431, "learning_rate": 1.8888354486549234e-07, "loss": 0.2301, "step": 3167 }, { "epoch": 0.8430015965939329, "grad_norm": 0.2762274742126465, "learning_rate": 1.8887572581874518e-07, "loss": 0.2037, "step": 3168 }, { "epoch": 0.8432676955827568, "grad_norm": 0.29003384709358215, "learning_rate": 1.8886790418505639e-07, "loss": 0.2357, "step": 3169 }, { "epoch": 0.8435337945715806, "grad_norm": 0.2801204323768616, "learning_rate": 1.8886007996465364e-07, "loss": 0.233, "step": 3170 }, { "epoch": 0.8437998935604045, "grad_norm": 0.2765153646469116, "learning_rate": 1.8885225315776472e-07, "loss": 0.2365, "step": 3171 }, { "epoch": 0.8440659925492283, "grad_norm": 0.29364991188049316, "learning_rate": 1.8884442376461738e-07, "loss": 0.2428, "step": 3172 }, { "epoch": 0.8443320915380521, "grad_norm": 0.298979789018631, "learning_rate": 1.8883659178543955e-07, "loss": 0.2376, "step": 3173 }, { "epoch": 0.844598190526876, "grad_norm": 0.2914038300514221, "learning_rate": 1.8882875722045922e-07, "loss": 0.2287, "step": 3174 }, { "epoch": 0.8448642895156998, "grad_norm": 0.2929956316947937, "learning_rate": 1.888209200699044e-07, "loss": 0.2352, "step": 3175 }, { "epoch": 0.8451303885045237, "grad_norm": 0.31282979249954224, "learning_rate": 1.888130803340032e-07, "loss": 0.2469, "step": 3176 }, { "epoch": 0.8453964874933475, "grad_norm": 0.25602400302886963, "learning_rate": 1.8880523801298387e-07, "loss": 0.2065, "step": 3177 }, { "epoch": 0.8456625864821714, "grad_norm": 0.28621163964271545, "learning_rate": 1.887973931070746e-07, "loss": 0.2173, "step": 3178 }, { "epoch": 0.8459286854709952, "grad_norm": 0.2804507315158844, "learning_rate": 1.887895456165038e-07, "loss": 0.2287, "step": 3179 }, { "epoch": 0.8461947844598191, "grad_norm": 0.2907073199748993, "learning_rate": 1.8878169554149986e-07, "loss": 0.2327, "step": 3180 }, { "epoch": 0.8464608834486429, "grad_norm": 0.27086135745048523, "learning_rate": 1.8877384288229128e-07, "loss": 0.2143, "step": 3181 }, { "epoch": 0.8467269824374667, "grad_norm": 0.2740055024623871, "learning_rate": 1.8876598763910662e-07, "loss": 0.2426, "step": 3182 }, { "epoch": 0.8469930814262906, "grad_norm": 0.28801867365837097, "learning_rate": 1.8875812981217457e-07, "loss": 0.2167, "step": 3183 }, { "epoch": 0.8472591804151144, "grad_norm": 0.2846015989780426, "learning_rate": 1.8875026940172381e-07, "loss": 0.2152, "step": 3184 }, { "epoch": 0.8475252794039383, "grad_norm": 0.32308804988861084, "learning_rate": 1.8874240640798313e-07, "loss": 0.2477, "step": 3185 }, { "epoch": 0.8477913783927621, "grad_norm": 0.2778627276420593, "learning_rate": 1.8873454083118144e-07, "loss": 0.2137, "step": 3186 }, { "epoch": 0.848057477381586, "grad_norm": 0.3199569880962372, "learning_rate": 1.8872667267154762e-07, "loss": 0.2438, "step": 3187 }, { "epoch": 0.8483235763704098, "grad_norm": 0.28284361958503723, "learning_rate": 1.8871880192931077e-07, "loss": 0.2337, "step": 3188 }, { "epoch": 0.8485896753592337, "grad_norm": 0.2995285987854004, "learning_rate": 1.8871092860469995e-07, "loss": 0.2249, "step": 3189 }, { "epoch": 0.8488557743480575, "grad_norm": 0.26365137100219727, "learning_rate": 1.8870305269794434e-07, "loss": 0.2288, "step": 3190 }, { "epoch": 0.8491218733368813, "grad_norm": 0.26145219802856445, "learning_rate": 1.8869517420927315e-07, "loss": 0.2215, "step": 3191 }, { "epoch": 0.8493879723257052, "grad_norm": 0.26123711466789246, "learning_rate": 1.8868729313891576e-07, "loss": 0.2041, "step": 3192 }, { "epoch": 0.849654071314529, "grad_norm": 0.29853737354278564, "learning_rate": 1.8867940948710154e-07, "loss": 0.2589, "step": 3193 }, { "epoch": 0.8499201703033529, "grad_norm": 0.2692680358886719, "learning_rate": 1.8867152325405996e-07, "loss": 0.2211, "step": 3194 }, { "epoch": 0.8501862692921767, "grad_norm": 0.292241632938385, "learning_rate": 1.886636344400206e-07, "loss": 0.2413, "step": 3195 }, { "epoch": 0.8504523682810006, "grad_norm": 0.26020047068595886, "learning_rate": 1.88655743045213e-07, "loss": 0.2276, "step": 3196 }, { "epoch": 0.8507184672698244, "grad_norm": 0.2753600776195526, "learning_rate": 1.8864784906986692e-07, "loss": 0.2064, "step": 3197 }, { "epoch": 0.8509845662586483, "grad_norm": 0.2871253192424774, "learning_rate": 1.8863995251421217e-07, "loss": 0.221, "step": 3198 }, { "epoch": 0.8512506652474721, "grad_norm": 0.2962046265602112, "learning_rate": 1.886320533784785e-07, "loss": 0.2157, "step": 3199 }, { "epoch": 0.8515167642362959, "grad_norm": 0.31069353222846985, "learning_rate": 1.8862415166289593e-07, "loss": 0.251, "step": 3200 }, { "epoch": 0.8517828632251198, "grad_norm": 0.28439024090766907, "learning_rate": 1.886162473676944e-07, "loss": 0.2241, "step": 3201 }, { "epoch": 0.8520489622139436, "grad_norm": 0.2761746942996979, "learning_rate": 1.88608340493104e-07, "loss": 0.2321, "step": 3202 }, { "epoch": 0.8523150612027675, "grad_norm": 0.2776230275630951, "learning_rate": 1.8860043103935485e-07, "loss": 0.2327, "step": 3203 }, { "epoch": 0.8525811601915912, "grad_norm": 0.281328409910202, "learning_rate": 1.8859251900667724e-07, "loss": 0.2196, "step": 3204 }, { "epoch": 0.8528472591804152, "grad_norm": 0.31411460041999817, "learning_rate": 1.885846043953014e-07, "loss": 0.2597, "step": 3205 }, { "epoch": 0.8531133581692389, "grad_norm": 0.2856152057647705, "learning_rate": 1.8857668720545775e-07, "loss": 0.2321, "step": 3206 }, { "epoch": 0.8533794571580628, "grad_norm": 0.29889291524887085, "learning_rate": 1.8856876743737674e-07, "loss": 0.2447, "step": 3207 }, { "epoch": 0.8536455561468866, "grad_norm": 0.31870001554489136, "learning_rate": 1.8856084509128885e-07, "loss": 0.2141, "step": 3208 }, { "epoch": 0.8539116551357104, "grad_norm": 0.2986108064651489, "learning_rate": 1.885529201674247e-07, "loss": 0.2293, "step": 3209 }, { "epoch": 0.8541777541245343, "grad_norm": 0.2810216248035431, "learning_rate": 1.88544992666015e-07, "loss": 0.2217, "step": 3210 }, { "epoch": 0.8544438531133581, "grad_norm": 0.2786942422389984, "learning_rate": 1.8853706258729046e-07, "loss": 0.2163, "step": 3211 }, { "epoch": 0.854709952102182, "grad_norm": 0.28820741176605225, "learning_rate": 1.885291299314819e-07, "loss": 0.221, "step": 3212 }, { "epoch": 0.8549760510910058, "grad_norm": 0.26702365279197693, "learning_rate": 1.8852119469882022e-07, "loss": 0.2134, "step": 3213 }, { "epoch": 0.8552421500798297, "grad_norm": 0.30044010281562805, "learning_rate": 1.8851325688953642e-07, "loss": 0.2485, "step": 3214 }, { "epoch": 0.8555082490686535, "grad_norm": 0.2778932750225067, "learning_rate": 1.8850531650386153e-07, "loss": 0.2132, "step": 3215 }, { "epoch": 0.8557743480574774, "grad_norm": 0.27920469641685486, "learning_rate": 1.8849737354202668e-07, "loss": 0.2312, "step": 3216 }, { "epoch": 0.8560404470463012, "grad_norm": 0.2728170156478882, "learning_rate": 1.884894280042631e-07, "loss": 0.2235, "step": 3217 }, { "epoch": 0.856306546035125, "grad_norm": 0.2738375961780548, "learning_rate": 1.8848147989080197e-07, "loss": 0.2145, "step": 3218 }, { "epoch": 0.8565726450239489, "grad_norm": 0.3123062252998352, "learning_rate": 1.8847352920187475e-07, "loss": 0.2341, "step": 3219 }, { "epoch": 0.8568387440127727, "grad_norm": 0.27660003304481506, "learning_rate": 1.8846557593771278e-07, "loss": 0.2183, "step": 3220 }, { "epoch": 0.8571048430015966, "grad_norm": 0.265994668006897, "learning_rate": 1.8845762009854763e-07, "loss": 0.2045, "step": 3221 }, { "epoch": 0.8573709419904204, "grad_norm": 0.2659996449947357, "learning_rate": 1.884496616846108e-07, "loss": 0.2337, "step": 3222 }, { "epoch": 0.8576370409792443, "grad_norm": 0.2905157804489136, "learning_rate": 1.8844170069613399e-07, "loss": 0.2198, "step": 3223 }, { "epoch": 0.8579031399680681, "grad_norm": 0.2898196280002594, "learning_rate": 1.884337371333489e-07, "loss": 0.2268, "step": 3224 }, { "epoch": 0.858169238956892, "grad_norm": 0.31357917189598083, "learning_rate": 1.8842577099648737e-07, "loss": 0.2217, "step": 3225 }, { "epoch": 0.8584353379457158, "grad_norm": 0.2820356488227844, "learning_rate": 1.884178022857812e-07, "loss": 0.2375, "step": 3226 }, { "epoch": 0.8587014369345396, "grad_norm": 0.28156745433807373, "learning_rate": 1.884098310014624e-07, "loss": 0.2121, "step": 3227 }, { "epoch": 0.8589675359233635, "grad_norm": 0.2779792845249176, "learning_rate": 1.8840185714376296e-07, "loss": 0.222, "step": 3228 }, { "epoch": 0.8592336349121873, "grad_norm": 0.2718130350112915, "learning_rate": 1.8839388071291502e-07, "loss": 0.2242, "step": 3229 }, { "epoch": 0.8594997339010112, "grad_norm": 0.27110302448272705, "learning_rate": 1.883859017091507e-07, "loss": 0.2105, "step": 3230 }, { "epoch": 0.859765832889835, "grad_norm": 0.2733863890171051, "learning_rate": 1.883779201327023e-07, "loss": 0.2172, "step": 3231 }, { "epoch": 0.8600319318786589, "grad_norm": 0.2783741056919098, "learning_rate": 1.8836993598380207e-07, "loss": 0.2289, "step": 3232 }, { "epoch": 0.8602980308674827, "grad_norm": 0.30344709753990173, "learning_rate": 1.883619492626825e-07, "loss": 0.2185, "step": 3233 }, { "epoch": 0.8605641298563066, "grad_norm": 0.29511773586273193, "learning_rate": 1.88353959969576e-07, "loss": 0.2363, "step": 3234 }, { "epoch": 0.8608302288451304, "grad_norm": 0.25973427295684814, "learning_rate": 1.8834596810471514e-07, "loss": 0.2072, "step": 3235 }, { "epoch": 0.8610963278339542, "grad_norm": 0.2911379635334015, "learning_rate": 1.883379736683325e-07, "loss": 0.2581, "step": 3236 }, { "epoch": 0.8613624268227781, "grad_norm": 0.2936031222343445, "learning_rate": 1.8832997666066086e-07, "loss": 0.2391, "step": 3237 }, { "epoch": 0.8616285258116019, "grad_norm": 0.2751483917236328, "learning_rate": 1.8832197708193292e-07, "loss": 0.2319, "step": 3238 }, { "epoch": 0.8618946248004258, "grad_norm": 0.29388198256492615, "learning_rate": 1.8831397493238156e-07, "loss": 0.2174, "step": 3239 }, { "epoch": 0.8621607237892496, "grad_norm": 0.2686220109462738, "learning_rate": 1.883059702122397e-07, "loss": 0.2098, "step": 3240 }, { "epoch": 0.8624268227780735, "grad_norm": 0.28901049494743347, "learning_rate": 1.882979629217403e-07, "loss": 0.231, "step": 3241 }, { "epoch": 0.8626929217668973, "grad_norm": 0.26836100220680237, "learning_rate": 1.882899530611165e-07, "loss": 0.2245, "step": 3242 }, { "epoch": 0.8629590207557212, "grad_norm": 0.2800503075122833, "learning_rate": 1.882819406306014e-07, "loss": 0.2291, "step": 3243 }, { "epoch": 0.863225119744545, "grad_norm": 0.28706562519073486, "learning_rate": 1.8827392563042825e-07, "loss": 0.2259, "step": 3244 }, { "epoch": 0.8634912187333688, "grad_norm": 0.31249552965164185, "learning_rate": 1.882659080608303e-07, "loss": 0.2315, "step": 3245 }, { "epoch": 0.8637573177221927, "grad_norm": 0.29238566756248474, "learning_rate": 1.8825788792204095e-07, "loss": 0.2263, "step": 3246 }, { "epoch": 0.8640234167110165, "grad_norm": 0.2870599031448364, "learning_rate": 1.8824986521429364e-07, "loss": 0.2345, "step": 3247 }, { "epoch": 0.8642895156998404, "grad_norm": 0.2831817865371704, "learning_rate": 1.882418399378219e-07, "loss": 0.24, "step": 3248 }, { "epoch": 0.8645556146886642, "grad_norm": 0.27864202857017517, "learning_rate": 1.8823381209285932e-07, "loss": 0.2193, "step": 3249 }, { "epoch": 0.8648217136774881, "grad_norm": 0.2914462983608246, "learning_rate": 1.8822578167963955e-07, "loss": 0.2271, "step": 3250 }, { "epoch": 0.8650878126663119, "grad_norm": 0.29039254784584045, "learning_rate": 1.8821774869839638e-07, "loss": 0.219, "step": 3251 }, { "epoch": 0.8653539116551358, "grad_norm": 0.30991700291633606, "learning_rate": 1.8820971314936357e-07, "loss": 0.2349, "step": 3252 }, { "epoch": 0.8656200106439595, "grad_norm": 0.26177138090133667, "learning_rate": 1.8820167503277506e-07, "loss": 0.212, "step": 3253 }, { "epoch": 0.8658861096327833, "grad_norm": 0.29031044244766235, "learning_rate": 1.881936343488648e-07, "loss": 0.2453, "step": 3254 }, { "epoch": 0.8661522086216072, "grad_norm": 0.29557105898857117, "learning_rate": 1.8818559109786684e-07, "loss": 0.2315, "step": 3255 }, { "epoch": 0.866418307610431, "grad_norm": 0.283514142036438, "learning_rate": 1.8817754528001527e-07, "loss": 0.2226, "step": 3256 }, { "epoch": 0.8666844065992549, "grad_norm": 0.25930461287498474, "learning_rate": 1.8816949689554435e-07, "loss": 0.2045, "step": 3257 }, { "epoch": 0.8669505055880787, "grad_norm": 0.275241494178772, "learning_rate": 1.8816144594468826e-07, "loss": 0.222, "step": 3258 }, { "epoch": 0.8672166045769026, "grad_norm": 0.2953382134437561, "learning_rate": 1.881533924276814e-07, "loss": 0.219, "step": 3259 }, { "epoch": 0.8674827035657264, "grad_norm": 0.30007338523864746, "learning_rate": 1.881453363447582e-07, "loss": 0.241, "step": 3260 }, { "epoch": 0.8677488025545503, "grad_norm": 0.3015924096107483, "learning_rate": 1.881372776961531e-07, "loss": 0.2366, "step": 3261 }, { "epoch": 0.8680149015433741, "grad_norm": 0.290946900844574, "learning_rate": 1.881292164821007e-07, "loss": 0.2505, "step": 3262 }, { "epoch": 0.868281000532198, "grad_norm": 0.2827312648296356, "learning_rate": 1.8812115270283564e-07, "loss": 0.2325, "step": 3263 }, { "epoch": 0.8685470995210218, "grad_norm": 0.29984602332115173, "learning_rate": 1.8811308635859265e-07, "loss": 0.2633, "step": 3264 }, { "epoch": 0.8688131985098456, "grad_norm": 0.27309373021125793, "learning_rate": 1.8810501744960645e-07, "loss": 0.2156, "step": 3265 }, { "epoch": 0.8690792974986695, "grad_norm": 0.2880174517631531, "learning_rate": 1.8809694597611199e-07, "loss": 0.2285, "step": 3266 }, { "epoch": 0.8693453964874933, "grad_norm": 0.2722773849964142, "learning_rate": 1.8808887193834415e-07, "loss": 0.2129, "step": 3267 }, { "epoch": 0.8696114954763172, "grad_norm": 0.2837284505367279, "learning_rate": 1.8808079533653797e-07, "loss": 0.2121, "step": 3268 }, { "epoch": 0.869877594465141, "grad_norm": 0.2695823311805725, "learning_rate": 1.8807271617092857e-07, "loss": 0.2352, "step": 3269 }, { "epoch": 0.8701436934539649, "grad_norm": 0.28019872307777405, "learning_rate": 1.8806463444175103e-07, "loss": 0.2131, "step": 3270 }, { "epoch": 0.8704097924427887, "grad_norm": 0.27152130007743835, "learning_rate": 1.8805655014924068e-07, "loss": 0.2263, "step": 3271 }, { "epoch": 0.8706758914316126, "grad_norm": 0.2571098208427429, "learning_rate": 1.880484632936328e-07, "loss": 0.2146, "step": 3272 }, { "epoch": 0.8709419904204364, "grad_norm": 0.27240410447120667, "learning_rate": 1.8804037387516274e-07, "loss": 0.2128, "step": 3273 }, { "epoch": 0.8712080894092602, "grad_norm": 0.3173064887523651, "learning_rate": 1.8803228189406603e-07, "loss": 0.2482, "step": 3274 }, { "epoch": 0.8714741883980841, "grad_norm": 0.3015425205230713, "learning_rate": 1.8802418735057813e-07, "loss": 0.2319, "step": 3275 }, { "epoch": 0.8717402873869079, "grad_norm": 0.30118000507354736, "learning_rate": 1.8801609024493468e-07, "loss": 0.2441, "step": 3276 }, { "epoch": 0.8720063863757318, "grad_norm": 0.2790831923484802, "learning_rate": 1.880079905773714e-07, "loss": 0.2362, "step": 3277 }, { "epoch": 0.8722724853645556, "grad_norm": 0.256647527217865, "learning_rate": 1.87999888348124e-07, "loss": 0.213, "step": 3278 }, { "epoch": 0.8725385843533795, "grad_norm": 0.28935661911964417, "learning_rate": 1.8799178355742835e-07, "loss": 0.238, "step": 3279 }, { "epoch": 0.8728046833422033, "grad_norm": 0.28700509667396545, "learning_rate": 1.8798367620552036e-07, "loss": 0.2222, "step": 3280 }, { "epoch": 0.8730707823310272, "grad_norm": 0.2727194130420685, "learning_rate": 1.87975566292636e-07, "loss": 0.2213, "step": 3281 }, { "epoch": 0.873336881319851, "grad_norm": 0.2822307348251343, "learning_rate": 1.8796745381901133e-07, "loss": 0.22, "step": 3282 }, { "epoch": 0.8736029803086748, "grad_norm": 0.26321712136268616, "learning_rate": 1.879593387848825e-07, "loss": 0.2138, "step": 3283 }, { "epoch": 0.8738690792974987, "grad_norm": 0.28316959738731384, "learning_rate": 1.879512211904857e-07, "loss": 0.2234, "step": 3284 }, { "epoch": 0.8741351782863225, "grad_norm": 0.25746557116508484, "learning_rate": 1.8794310103605716e-07, "loss": 0.1955, "step": 3285 }, { "epoch": 0.8744012772751464, "grad_norm": 0.26073241233825684, "learning_rate": 1.8793497832183334e-07, "loss": 0.209, "step": 3286 }, { "epoch": 0.8746673762639702, "grad_norm": 0.34492191672325134, "learning_rate": 1.879268530480506e-07, "loss": 0.244, "step": 3287 }, { "epoch": 0.8749334752527941, "grad_norm": 0.2972741723060608, "learning_rate": 1.879187252149455e-07, "loss": 0.246, "step": 3288 }, { "epoch": 0.8751995742416179, "grad_norm": 0.2894054055213928, "learning_rate": 1.8791059482275455e-07, "loss": 0.2594, "step": 3289 }, { "epoch": 0.8754656732304418, "grad_norm": 0.2674938142299652, "learning_rate": 1.8790246187171447e-07, "loss": 0.2104, "step": 3290 }, { "epoch": 0.8757317722192656, "grad_norm": 0.2947293519973755, "learning_rate": 1.8789432636206195e-07, "loss": 0.2179, "step": 3291 }, { "epoch": 0.8759978712080894, "grad_norm": 0.2848900854587555, "learning_rate": 1.878861882940338e-07, "loss": 0.2398, "step": 3292 }, { "epoch": 0.8762639701969133, "grad_norm": 0.2981826066970825, "learning_rate": 1.8787804766786692e-07, "loss": 0.1963, "step": 3293 }, { "epoch": 0.8765300691857371, "grad_norm": 0.2884417772293091, "learning_rate": 1.8786990448379822e-07, "loss": 0.2184, "step": 3294 }, { "epoch": 0.876796168174561, "grad_norm": 0.2871271073818207, "learning_rate": 1.8786175874206477e-07, "loss": 0.2021, "step": 3295 }, { "epoch": 0.8770622671633848, "grad_norm": 0.3044596314430237, "learning_rate": 1.8785361044290365e-07, "loss": 0.266, "step": 3296 }, { "epoch": 0.8773283661522087, "grad_norm": 0.26969537138938904, "learning_rate": 1.8784545958655205e-07, "loss": 0.2078, "step": 3297 }, { "epoch": 0.8775944651410325, "grad_norm": 0.2769289016723633, "learning_rate": 1.8783730617324718e-07, "loss": 0.2313, "step": 3298 }, { "epoch": 0.8778605641298564, "grad_norm": 0.34962841868400574, "learning_rate": 1.8782915020322644e-07, "loss": 0.2267, "step": 3299 }, { "epoch": 0.8781266631186802, "grad_norm": 0.26737868785858154, "learning_rate": 1.8782099167672714e-07, "loss": 0.2253, "step": 3300 }, { "epoch": 0.8783927621075039, "grad_norm": 0.30506932735443115, "learning_rate": 1.8781283059398684e-07, "loss": 0.2379, "step": 3301 }, { "epoch": 0.8786588610963278, "grad_norm": 0.28971773386001587, "learning_rate": 1.87804666955243e-07, "loss": 0.2455, "step": 3302 }, { "epoch": 0.8789249600851516, "grad_norm": 0.2726895213127136, "learning_rate": 1.877965007607333e-07, "loss": 0.2199, "step": 3303 }, { "epoch": 0.8791910590739755, "grad_norm": 0.2918933033943176, "learning_rate": 1.8778833201069543e-07, "loss": 0.2423, "step": 3304 }, { "epoch": 0.8794571580627993, "grad_norm": 0.2985858619213104, "learning_rate": 1.8778016070536714e-07, "loss": 0.2368, "step": 3305 }, { "epoch": 0.8797232570516232, "grad_norm": 0.3697602152824402, "learning_rate": 1.8777198684498633e-07, "loss": 0.2216, "step": 3306 }, { "epoch": 0.879989356040447, "grad_norm": 0.288770854473114, "learning_rate": 1.8776381042979086e-07, "loss": 0.2144, "step": 3307 }, { "epoch": 0.8802554550292709, "grad_norm": 0.28446418046951294, "learning_rate": 1.8775563146001874e-07, "loss": 0.2195, "step": 3308 }, { "epoch": 0.8805215540180947, "grad_norm": 0.267218679189682, "learning_rate": 1.8774744993590805e-07, "loss": 0.2048, "step": 3309 }, { "epoch": 0.8807876530069185, "grad_norm": 0.28927847743034363, "learning_rate": 1.8773926585769693e-07, "loss": 0.2012, "step": 3310 }, { "epoch": 0.8810537519957424, "grad_norm": 0.30930599570274353, "learning_rate": 1.8773107922562355e-07, "loss": 0.2312, "step": 3311 }, { "epoch": 0.8813198509845662, "grad_norm": 0.2794780731201172, "learning_rate": 1.8772289003992632e-07, "loss": 0.2241, "step": 3312 }, { "epoch": 0.8815859499733901, "grad_norm": 0.300687313079834, "learning_rate": 1.8771469830084347e-07, "loss": 0.2363, "step": 3313 }, { "epoch": 0.8818520489622139, "grad_norm": 0.29483914375305176, "learning_rate": 1.8770650400861353e-07, "loss": 0.2128, "step": 3314 }, { "epoch": 0.8821181479510378, "grad_norm": 0.2768961787223816, "learning_rate": 1.87698307163475e-07, "loss": 0.2247, "step": 3315 }, { "epoch": 0.8823842469398616, "grad_norm": 0.28287747502326965, "learning_rate": 1.8769010776566643e-07, "loss": 0.2248, "step": 3316 }, { "epoch": 0.8826503459286855, "grad_norm": 0.2825919985771179, "learning_rate": 1.8768190581542652e-07, "loss": 0.2221, "step": 3317 }, { "epoch": 0.8829164449175093, "grad_norm": 0.2903488278388977, "learning_rate": 1.87673701312994e-07, "loss": 0.2254, "step": 3318 }, { "epoch": 0.8831825439063331, "grad_norm": 0.2699963450431824, "learning_rate": 1.8766549425860765e-07, "loss": 0.2099, "step": 3319 }, { "epoch": 0.883448642895157, "grad_norm": 0.25893595814704895, "learning_rate": 1.876572846525064e-07, "loss": 0.1992, "step": 3320 }, { "epoch": 0.8837147418839808, "grad_norm": 0.2642000913619995, "learning_rate": 1.8764907249492917e-07, "loss": 0.2104, "step": 3321 }, { "epoch": 0.8839808408728047, "grad_norm": 1.0391249656677246, "learning_rate": 1.8764085778611504e-07, "loss": 0.2044, "step": 3322 }, { "epoch": 0.8842469398616285, "grad_norm": 0.3064766526222229, "learning_rate": 1.8763264052630312e-07, "loss": 0.2419, "step": 3323 }, { "epoch": 0.8845130388504524, "grad_norm": 0.2859775125980377, "learning_rate": 1.8762442071573254e-07, "loss": 0.2197, "step": 3324 }, { "epoch": 0.8847791378392762, "grad_norm": 0.28351259231567383, "learning_rate": 1.8761619835464262e-07, "loss": 0.2442, "step": 3325 }, { "epoch": 0.8850452368281001, "grad_norm": 0.26693156361579895, "learning_rate": 1.8760797344327262e-07, "loss": 0.2178, "step": 3326 }, { "epoch": 0.8853113358169239, "grad_norm": 0.3053874373435974, "learning_rate": 1.87599745981862e-07, "loss": 0.2392, "step": 3327 }, { "epoch": 0.8855774348057477, "grad_norm": 0.3375668525695801, "learning_rate": 1.8759151597065022e-07, "loss": 0.2518, "step": 3328 }, { "epoch": 0.8858435337945716, "grad_norm": 0.3115219473838806, "learning_rate": 1.8758328340987685e-07, "loss": 0.2227, "step": 3329 }, { "epoch": 0.8861096327833954, "grad_norm": 0.2883923351764679, "learning_rate": 1.875750482997815e-07, "loss": 0.2233, "step": 3330 }, { "epoch": 0.8863757317722193, "grad_norm": 0.25283047556877136, "learning_rate": 1.8756681064060387e-07, "loss": 0.2096, "step": 3331 }, { "epoch": 0.8866418307610431, "grad_norm": 0.2727082371711731, "learning_rate": 1.8755857043258377e-07, "loss": 0.2045, "step": 3332 }, { "epoch": 0.886907929749867, "grad_norm": 0.27778568863868713, "learning_rate": 1.8755032767596102e-07, "loss": 0.2364, "step": 3333 }, { "epoch": 0.8871740287386908, "grad_norm": 0.29071474075317383, "learning_rate": 1.8754208237097556e-07, "loss": 0.2295, "step": 3334 }, { "epoch": 0.8874401277275147, "grad_norm": 0.2836453914642334, "learning_rate": 1.8753383451786736e-07, "loss": 0.1994, "step": 3335 }, { "epoch": 0.8877062267163385, "grad_norm": 0.29189807176589966, "learning_rate": 1.8752558411687653e-07, "loss": 0.2384, "step": 3336 }, { "epoch": 0.8879723257051623, "grad_norm": 0.29169395565986633, "learning_rate": 1.875173311682432e-07, "loss": 0.2305, "step": 3337 }, { "epoch": 0.8882384246939862, "grad_norm": 0.26629018783569336, "learning_rate": 1.8750907567220762e-07, "loss": 0.2253, "step": 3338 }, { "epoch": 0.88850452368281, "grad_norm": 0.2707933485507965, "learning_rate": 1.8750081762901003e-07, "loss": 0.2174, "step": 3339 }, { "epoch": 0.8887706226716339, "grad_norm": 0.31372740864753723, "learning_rate": 1.8749255703889083e-07, "loss": 0.2515, "step": 3340 }, { "epoch": 0.8890367216604577, "grad_norm": 0.27020931243896484, "learning_rate": 1.874842939020905e-07, "loss": 0.2204, "step": 3341 }, { "epoch": 0.8893028206492816, "grad_norm": 0.3092188835144043, "learning_rate": 1.874760282188495e-07, "loss": 0.2371, "step": 3342 }, { "epoch": 0.8895689196381054, "grad_norm": 0.2985215485095978, "learning_rate": 1.8746775998940846e-07, "loss": 0.2302, "step": 3343 }, { "epoch": 0.8898350186269293, "grad_norm": 0.31427228450775146, "learning_rate": 1.8745948921400805e-07, "loss": 0.2403, "step": 3344 }, { "epoch": 0.8901011176157531, "grad_norm": 0.2779868245124817, "learning_rate": 1.8745121589288897e-07, "loss": 0.1983, "step": 3345 }, { "epoch": 0.8903672166045768, "grad_norm": 0.27027010917663574, "learning_rate": 1.8744294002629208e-07, "loss": 0.2136, "step": 3346 }, { "epoch": 0.8906333155934008, "grad_norm": 0.33173784613609314, "learning_rate": 1.874346616144582e-07, "loss": 0.2386, "step": 3347 }, { "epoch": 0.8908994145822245, "grad_norm": 0.30396413803100586, "learning_rate": 1.874263806576284e-07, "loss": 0.2331, "step": 3348 }, { "epoch": 0.8911655135710485, "grad_norm": 0.2852587103843689, "learning_rate": 1.874180971560436e-07, "loss": 0.2362, "step": 3349 }, { "epoch": 0.8914316125598722, "grad_norm": 0.30841144919395447, "learning_rate": 1.8740981110994502e-07, "loss": 0.2314, "step": 3350 }, { "epoch": 0.8916977115486961, "grad_norm": 0.26489099860191345, "learning_rate": 1.874015225195738e-07, "loss": 0.2326, "step": 3351 }, { "epoch": 0.8919638105375199, "grad_norm": 0.2857590913772583, "learning_rate": 1.8739323138517117e-07, "loss": 0.2055, "step": 3352 }, { "epoch": 0.8922299095263438, "grad_norm": 0.2896304130554199, "learning_rate": 1.873849377069785e-07, "loss": 0.2273, "step": 3353 }, { "epoch": 0.8924960085151676, "grad_norm": 0.2950892150402069, "learning_rate": 1.8737664148523716e-07, "loss": 0.2221, "step": 3354 }, { "epoch": 0.8927621075039914, "grad_norm": 0.28686797618865967, "learning_rate": 1.873683427201887e-07, "loss": 0.2267, "step": 3355 }, { "epoch": 0.8930282064928153, "grad_norm": 0.3286893665790558, "learning_rate": 1.873600414120746e-07, "loss": 0.2474, "step": 3356 }, { "epoch": 0.8932943054816391, "grad_norm": 0.3000025749206543, "learning_rate": 1.8735173756113653e-07, "loss": 0.2253, "step": 3357 }, { "epoch": 0.893560404470463, "grad_norm": 0.29399630427360535, "learning_rate": 1.8734343116761621e-07, "loss": 0.2371, "step": 3358 }, { "epoch": 0.8938265034592868, "grad_norm": 0.2829022705554962, "learning_rate": 1.8733512223175536e-07, "loss": 0.2394, "step": 3359 }, { "epoch": 0.8940926024481107, "grad_norm": 0.3141011595726013, "learning_rate": 1.873268107537959e-07, "loss": 0.2064, "step": 3360 }, { "epoch": 0.8943587014369345, "grad_norm": 0.28396013379096985, "learning_rate": 1.8731849673397968e-07, "loss": 0.2268, "step": 3361 }, { "epoch": 0.8946248004257584, "grad_norm": 0.2713243067264557, "learning_rate": 1.8731018017254878e-07, "loss": 0.2085, "step": 3362 }, { "epoch": 0.8948908994145822, "grad_norm": 0.2886504828929901, "learning_rate": 1.873018610697452e-07, "loss": 0.2131, "step": 3363 }, { "epoch": 0.895156998403406, "grad_norm": 0.2727283537387848, "learning_rate": 1.8729353942581114e-07, "loss": 0.2483, "step": 3364 }, { "epoch": 0.8954230973922299, "grad_norm": 0.31416845321655273, "learning_rate": 1.872852152409888e-07, "loss": 0.2447, "step": 3365 }, { "epoch": 0.8956891963810537, "grad_norm": 0.30064740777015686, "learning_rate": 1.8727688851552049e-07, "loss": 0.2123, "step": 3366 }, { "epoch": 0.8959552953698776, "grad_norm": 0.27905064821243286, "learning_rate": 1.8726855924964853e-07, "loss": 0.2183, "step": 3367 }, { "epoch": 0.8962213943587014, "grad_norm": 0.2814425230026245, "learning_rate": 1.8726022744361543e-07, "loss": 0.2314, "step": 3368 }, { "epoch": 0.8964874933475253, "grad_norm": 0.2744932770729065, "learning_rate": 1.872518930976637e-07, "loss": 0.2358, "step": 3369 }, { "epoch": 0.8967535923363491, "grad_norm": 0.2874317169189453, "learning_rate": 1.8724355621203584e-07, "loss": 0.227, "step": 3370 }, { "epoch": 0.897019691325173, "grad_norm": 0.27372753620147705, "learning_rate": 1.8723521678697462e-07, "loss": 0.2039, "step": 3371 }, { "epoch": 0.8972857903139968, "grad_norm": 0.2800424098968506, "learning_rate": 1.8722687482272276e-07, "loss": 0.2373, "step": 3372 }, { "epoch": 0.8975518893028207, "grad_norm": 0.30453288555145264, "learning_rate": 1.8721853031952306e-07, "loss": 0.2354, "step": 3373 }, { "epoch": 0.8978179882916445, "grad_norm": 0.29997798800468445, "learning_rate": 1.872101832776184e-07, "loss": 0.2241, "step": 3374 }, { "epoch": 0.8980840872804683, "grad_norm": 0.2832961082458496, "learning_rate": 1.872018336972517e-07, "loss": 0.2291, "step": 3375 }, { "epoch": 0.8983501862692922, "grad_norm": 0.2834516167640686, "learning_rate": 1.871934815786661e-07, "loss": 0.2193, "step": 3376 }, { "epoch": 0.898616285258116, "grad_norm": 0.4011639356613159, "learning_rate": 1.8718512692210463e-07, "loss": 0.2229, "step": 3377 }, { "epoch": 0.8988823842469399, "grad_norm": 0.28957146406173706, "learning_rate": 1.8717676972781046e-07, "loss": 0.2067, "step": 3378 }, { "epoch": 0.8991484832357637, "grad_norm": 0.2766495645046234, "learning_rate": 1.871684099960269e-07, "loss": 0.2372, "step": 3379 }, { "epoch": 0.8994145822245876, "grad_norm": 0.27725616097450256, "learning_rate": 1.8716004772699724e-07, "loss": 0.2173, "step": 3380 }, { "epoch": 0.8996806812134114, "grad_norm": 0.27606433629989624, "learning_rate": 1.871516829209649e-07, "loss": 0.2266, "step": 3381 }, { "epoch": 0.8999467802022353, "grad_norm": 0.2683306932449341, "learning_rate": 1.871433155781734e-07, "loss": 0.2309, "step": 3382 }, { "epoch": 0.9002128791910591, "grad_norm": 0.29293569922447205, "learning_rate": 1.8713494569886619e-07, "loss": 0.236, "step": 3383 }, { "epoch": 0.9004789781798829, "grad_norm": 0.2691097557544708, "learning_rate": 1.8712657328328698e-07, "loss": 0.2088, "step": 3384 }, { "epoch": 0.9007450771687068, "grad_norm": 0.3011213541030884, "learning_rate": 1.8711819833167947e-07, "loss": 0.2191, "step": 3385 }, { "epoch": 0.9010111761575306, "grad_norm": 0.2914896607398987, "learning_rate": 1.8710982084428734e-07, "loss": 0.22, "step": 3386 }, { "epoch": 0.9012772751463545, "grad_norm": 0.3054610788822174, "learning_rate": 1.8710144082135454e-07, "loss": 0.2356, "step": 3387 }, { "epoch": 0.9015433741351783, "grad_norm": 0.2911411225795746, "learning_rate": 1.8709305826312498e-07, "loss": 0.2304, "step": 3388 }, { "epoch": 0.9018094731240022, "grad_norm": 0.2835981249809265, "learning_rate": 1.870846731698426e-07, "loss": 0.2288, "step": 3389 }, { "epoch": 0.902075572112826, "grad_norm": 0.274730920791626, "learning_rate": 1.870762855417515e-07, "loss": 0.2235, "step": 3390 }, { "epoch": 0.9023416711016499, "grad_norm": 0.2742769122123718, "learning_rate": 1.870678953790958e-07, "loss": 0.2279, "step": 3391 }, { "epoch": 0.9026077700904737, "grad_norm": 0.2645376920700073, "learning_rate": 1.870595026821198e-07, "loss": 0.2078, "step": 3392 }, { "epoch": 0.9028738690792975, "grad_norm": 0.2925876975059509, "learning_rate": 1.8705110745106767e-07, "loss": 0.2189, "step": 3393 }, { "epoch": 0.9031399680681214, "grad_norm": 0.2903039753437042, "learning_rate": 1.8704270968618382e-07, "loss": 0.2411, "step": 3394 }, { "epoch": 0.9034060670569452, "grad_norm": 0.30449116230010986, "learning_rate": 1.8703430938771272e-07, "loss": 0.2359, "step": 3395 }, { "epoch": 0.903672166045769, "grad_norm": 0.2906315326690674, "learning_rate": 1.8702590655589886e-07, "loss": 0.22, "step": 3396 }, { "epoch": 0.9039382650345928, "grad_norm": 0.29030635952949524, "learning_rate": 1.8701750119098677e-07, "loss": 0.2323, "step": 3397 }, { "epoch": 0.9042043640234168, "grad_norm": 0.272966593503952, "learning_rate": 1.8700909329322123e-07, "loss": 0.2257, "step": 3398 }, { "epoch": 0.9044704630122405, "grad_norm": 0.27126315236091614, "learning_rate": 1.8700068286284687e-07, "loss": 0.2209, "step": 3399 }, { "epoch": 0.9047365620010644, "grad_norm": 0.26670703291893005, "learning_rate": 1.8699226990010852e-07, "loss": 0.2039, "step": 3400 }, { "epoch": 0.9050026609898882, "grad_norm": 0.27740827202796936, "learning_rate": 1.8698385440525108e-07, "loss": 0.2247, "step": 3401 }, { "epoch": 0.905268759978712, "grad_norm": 0.3737030327320099, "learning_rate": 1.8697543637851947e-07, "loss": 0.2057, "step": 3402 }, { "epoch": 0.9055348589675359, "grad_norm": 0.2787000834941864, "learning_rate": 1.8696701582015877e-07, "loss": 0.2092, "step": 3403 }, { "epoch": 0.9058009579563597, "grad_norm": 0.26720499992370605, "learning_rate": 1.8695859273041404e-07, "loss": 0.2001, "step": 3404 }, { "epoch": 0.9060670569451836, "grad_norm": 0.28908777236938477, "learning_rate": 1.8695016710953043e-07, "loss": 0.1915, "step": 3405 }, { "epoch": 0.9063331559340074, "grad_norm": 0.30255937576293945, "learning_rate": 1.869417389577532e-07, "loss": 0.2171, "step": 3406 }, { "epoch": 0.9065992549228313, "grad_norm": 0.27399125695228577, "learning_rate": 1.8693330827532777e-07, "loss": 0.2205, "step": 3407 }, { "epoch": 0.9068653539116551, "grad_norm": 0.26368197798728943, "learning_rate": 1.8692487506249942e-07, "loss": 0.2, "step": 3408 }, { "epoch": 0.907131452900479, "grad_norm": 0.2553706467151642, "learning_rate": 1.8691643931951365e-07, "loss": 0.2136, "step": 3409 }, { "epoch": 0.9073975518893028, "grad_norm": 0.3094226121902466, "learning_rate": 1.86908001046616e-07, "loss": 0.2307, "step": 3410 }, { "epoch": 0.9076636508781266, "grad_norm": 0.2718745768070221, "learning_rate": 1.868995602440521e-07, "loss": 0.2157, "step": 3411 }, { "epoch": 0.9079297498669505, "grad_norm": 0.27387335896492004, "learning_rate": 1.8689111691206763e-07, "loss": 0.2061, "step": 3412 }, { "epoch": 0.9081958488557743, "grad_norm": 0.2901625633239746, "learning_rate": 1.868826710509084e-07, "loss": 0.2396, "step": 3413 }, { "epoch": 0.9084619478445982, "grad_norm": 0.28412482142448425, "learning_rate": 1.8687422266082016e-07, "loss": 0.2032, "step": 3414 }, { "epoch": 0.908728046833422, "grad_norm": 0.28300946950912476, "learning_rate": 1.8686577174204884e-07, "loss": 0.2075, "step": 3415 }, { "epoch": 0.9089941458222459, "grad_norm": 0.26870620250701904, "learning_rate": 1.8685731829484047e-07, "loss": 0.2396, "step": 3416 }, { "epoch": 0.9092602448110697, "grad_norm": 0.2722399830818176, "learning_rate": 1.8684886231944112e-07, "loss": 0.2163, "step": 3417 }, { "epoch": 0.9095263437998936, "grad_norm": 0.2871464192867279, "learning_rate": 1.8684040381609687e-07, "loss": 0.2245, "step": 3418 }, { "epoch": 0.9097924427887174, "grad_norm": 0.31809884309768677, "learning_rate": 1.8683194278505395e-07, "loss": 0.2461, "step": 3419 }, { "epoch": 0.9100585417775412, "grad_norm": 0.2827569842338562, "learning_rate": 1.8682347922655862e-07, "loss": 0.2257, "step": 3420 }, { "epoch": 0.9103246407663651, "grad_norm": 0.298671692609787, "learning_rate": 1.8681501314085723e-07, "loss": 0.2284, "step": 3421 }, { "epoch": 0.9105907397551889, "grad_norm": 0.29584255814552307, "learning_rate": 1.8680654452819625e-07, "loss": 0.2391, "step": 3422 }, { "epoch": 0.9108568387440128, "grad_norm": 0.3005746006965637, "learning_rate": 1.8679807338882214e-07, "loss": 0.2314, "step": 3423 }, { "epoch": 0.9111229377328366, "grad_norm": 0.31240591406822205, "learning_rate": 1.867895997229815e-07, "loss": 0.2183, "step": 3424 }, { "epoch": 0.9113890367216605, "grad_norm": 0.29114004969596863, "learning_rate": 1.867811235309209e-07, "loss": 0.2367, "step": 3425 }, { "epoch": 0.9116551357104843, "grad_norm": 0.27875271439552307, "learning_rate": 1.867726448128872e-07, "loss": 0.2334, "step": 3426 }, { "epoch": 0.9119212346993082, "grad_norm": 0.3191165328025818, "learning_rate": 1.8676416356912705e-07, "loss": 0.2252, "step": 3427 }, { "epoch": 0.912187333688132, "grad_norm": 0.29646191000938416, "learning_rate": 1.867556797998874e-07, "loss": 0.2278, "step": 3428 }, { "epoch": 0.9124534326769558, "grad_norm": 0.29515698552131653, "learning_rate": 1.8674719350541517e-07, "loss": 0.2269, "step": 3429 }, { "epoch": 0.9127195316657797, "grad_norm": 0.2794330418109894, "learning_rate": 1.8673870468595738e-07, "loss": 0.2302, "step": 3430 }, { "epoch": 0.9129856306546035, "grad_norm": 0.3303185999393463, "learning_rate": 1.867302133417611e-07, "loss": 0.2367, "step": 3431 }, { "epoch": 0.9132517296434274, "grad_norm": 0.273701012134552, "learning_rate": 1.8672171947307352e-07, "loss": 0.228, "step": 3432 }, { "epoch": 0.9135178286322512, "grad_norm": 0.3063184320926666, "learning_rate": 1.8671322308014184e-07, "loss": 0.2223, "step": 3433 }, { "epoch": 0.9137839276210751, "grad_norm": 0.2935197055339813, "learning_rate": 1.8670472416321336e-07, "loss": 0.2184, "step": 3434 }, { "epoch": 0.9140500266098989, "grad_norm": 0.31147339940071106, "learning_rate": 1.8669622272253555e-07, "loss": 0.2409, "step": 3435 }, { "epoch": 0.9143161255987228, "grad_norm": 0.2737542986869812, "learning_rate": 1.8668771875835576e-07, "loss": 0.2093, "step": 3436 }, { "epoch": 0.9145822245875466, "grad_norm": 0.2942516803741455, "learning_rate": 1.8667921227092156e-07, "loss": 0.2411, "step": 3437 }, { "epoch": 0.9148483235763704, "grad_norm": 0.29414740204811096, "learning_rate": 1.8667070326048055e-07, "loss": 0.2303, "step": 3438 }, { "epoch": 0.9151144225651943, "grad_norm": 0.265081524848938, "learning_rate": 1.8666219172728042e-07, "loss": 0.2247, "step": 3439 }, { "epoch": 0.9153805215540181, "grad_norm": 0.27430441975593567, "learning_rate": 1.866536776715689e-07, "loss": 0.2245, "step": 3440 }, { "epoch": 0.915646620542842, "grad_norm": 0.3218090534210205, "learning_rate": 1.8664516109359382e-07, "loss": 0.2347, "step": 3441 }, { "epoch": 0.9159127195316658, "grad_norm": 0.28236451745033264, "learning_rate": 1.8663664199360305e-07, "loss": 0.2143, "step": 3442 }, { "epoch": 0.9161788185204897, "grad_norm": 0.28886502981185913, "learning_rate": 1.866281203718446e-07, "loss": 0.2218, "step": 3443 }, { "epoch": 0.9164449175093135, "grad_norm": 0.320919007062912, "learning_rate": 1.8661959622856652e-07, "loss": 0.2483, "step": 3444 }, { "epoch": 0.9167110164981374, "grad_norm": 0.3114418387413025, "learning_rate": 1.8661106956401683e-07, "loss": 0.2474, "step": 3445 }, { "epoch": 0.9169771154869611, "grad_norm": 0.3031730651855469, "learning_rate": 1.8660254037844388e-07, "loss": 0.2234, "step": 3446 }, { "epoch": 0.9172432144757849, "grad_norm": 0.26603102684020996, "learning_rate": 1.8659400867209575e-07, "loss": 0.2158, "step": 3447 }, { "epoch": 0.9175093134646088, "grad_norm": 0.28080567717552185, "learning_rate": 1.8658547444522092e-07, "loss": 0.2228, "step": 3448 }, { "epoch": 0.9177754124534326, "grad_norm": 0.285458505153656, "learning_rate": 1.8657693769806775e-07, "loss": 0.2351, "step": 3449 }, { "epoch": 0.9180415114422565, "grad_norm": 0.26340052485466003, "learning_rate": 1.865683984308847e-07, "loss": 0.2206, "step": 3450 }, { "epoch": 0.9183076104310803, "grad_norm": 0.2957259714603424, "learning_rate": 1.8655985664392033e-07, "loss": 0.2313, "step": 3451 }, { "epoch": 0.9185737094199042, "grad_norm": 0.2944881021976471, "learning_rate": 1.8655131233742328e-07, "loss": 0.224, "step": 3452 }, { "epoch": 0.918839808408728, "grad_norm": 0.30884894728660583, "learning_rate": 1.865427655116423e-07, "loss": 0.2343, "step": 3453 }, { "epoch": 0.9191059073975519, "grad_norm": 0.2709883749485016, "learning_rate": 1.865342161668261e-07, "loss": 0.2347, "step": 3454 }, { "epoch": 0.9193720063863757, "grad_norm": 0.264994740486145, "learning_rate": 1.8652566430322354e-07, "loss": 0.2019, "step": 3455 }, { "epoch": 0.9196381053751995, "grad_norm": 0.2829481363296509, "learning_rate": 1.8651710992108356e-07, "loss": 0.2264, "step": 3456 }, { "epoch": 0.9199042043640234, "grad_norm": 0.2652241885662079, "learning_rate": 1.8650855302065515e-07, "loss": 0.2277, "step": 3457 }, { "epoch": 0.9201703033528472, "grad_norm": 0.2787627577781677, "learning_rate": 1.864999936021874e-07, "loss": 0.2254, "step": 3458 }, { "epoch": 0.9204364023416711, "grad_norm": 0.3040546476840973, "learning_rate": 1.864914316659294e-07, "loss": 0.2299, "step": 3459 }, { "epoch": 0.9207025013304949, "grad_norm": 0.29195740818977356, "learning_rate": 1.864828672121304e-07, "loss": 0.2318, "step": 3460 }, { "epoch": 0.9209686003193188, "grad_norm": 0.27841490507125854, "learning_rate": 1.8647430024103972e-07, "loss": 0.2234, "step": 3461 }, { "epoch": 0.9212346993081426, "grad_norm": 0.28249454498291016, "learning_rate": 1.8646573075290665e-07, "loss": 0.2107, "step": 3462 }, { "epoch": 0.9215007982969665, "grad_norm": 0.2980071008205414, "learning_rate": 1.8645715874798068e-07, "loss": 0.2231, "step": 3463 }, { "epoch": 0.9217668972857903, "grad_norm": 0.28414273262023926, "learning_rate": 1.864485842265113e-07, "loss": 0.2312, "step": 3464 }, { "epoch": 0.9220329962746141, "grad_norm": 0.2679571807384491, "learning_rate": 1.864400071887481e-07, "loss": 0.2416, "step": 3465 }, { "epoch": 0.922299095263438, "grad_norm": 0.27602073550224304, "learning_rate": 1.8643142763494074e-07, "loss": 0.221, "step": 3466 }, { "epoch": 0.9225651942522618, "grad_norm": 0.3064541220664978, "learning_rate": 1.864228455653389e-07, "loss": 0.2256, "step": 3467 }, { "epoch": 0.9228312932410857, "grad_norm": 0.27126774191856384, "learning_rate": 1.8641426098019246e-07, "loss": 0.2207, "step": 3468 }, { "epoch": 0.9230973922299095, "grad_norm": 0.291849821805954, "learning_rate": 1.8640567387975122e-07, "loss": 0.2291, "step": 3469 }, { "epoch": 0.9233634912187334, "grad_norm": 0.2976821959018707, "learning_rate": 1.8639708426426519e-07, "loss": 0.2357, "step": 3470 }, { "epoch": 0.9236295902075572, "grad_norm": 0.26856979727745056, "learning_rate": 1.8638849213398436e-07, "loss": 0.2184, "step": 3471 }, { "epoch": 0.9238956891963811, "grad_norm": 0.2440556287765503, "learning_rate": 1.8637989748915883e-07, "loss": 0.1991, "step": 3472 }, { "epoch": 0.9241617881852049, "grad_norm": 0.2863495647907257, "learning_rate": 1.8637130033003876e-07, "loss": 0.2435, "step": 3473 }, { "epoch": 0.9244278871740287, "grad_norm": 0.28428372740745544, "learning_rate": 1.863627006568744e-07, "loss": 0.2205, "step": 3474 }, { "epoch": 0.9246939861628526, "grad_norm": 0.2914256155490875, "learning_rate": 1.8635409846991608e-07, "loss": 0.2259, "step": 3475 }, { "epoch": 0.9249600851516764, "grad_norm": 0.2665097415447235, "learning_rate": 1.8634549376941414e-07, "loss": 0.2212, "step": 3476 }, { "epoch": 0.9252261841405003, "grad_norm": 0.29845255613327026, "learning_rate": 1.863368865556191e-07, "loss": 0.2409, "step": 3477 }, { "epoch": 0.9254922831293241, "grad_norm": 0.2553098201751709, "learning_rate": 1.8632827682878142e-07, "loss": 0.1985, "step": 3478 }, { "epoch": 0.925758382118148, "grad_norm": 0.2834581732749939, "learning_rate": 1.8631966458915177e-07, "loss": 0.2259, "step": 3479 }, { "epoch": 0.9260244811069718, "grad_norm": 0.27806541323661804, "learning_rate": 1.8631104983698084e-07, "loss": 0.2287, "step": 3480 }, { "epoch": 0.9262905800957957, "grad_norm": 0.3144395649433136, "learning_rate": 1.863024325725193e-07, "loss": 0.2565, "step": 3481 }, { "epoch": 0.9265566790846195, "grad_norm": 0.28404805064201355, "learning_rate": 1.862938127960181e-07, "loss": 0.2105, "step": 3482 }, { "epoch": 0.9268227780734434, "grad_norm": 0.4255083501338959, "learning_rate": 1.8628519050772802e-07, "loss": 0.2197, "step": 3483 }, { "epoch": 0.9270888770622672, "grad_norm": 0.31812649965286255, "learning_rate": 1.862765657079001e-07, "loss": 0.2295, "step": 3484 }, { "epoch": 0.927354976051091, "grad_norm": 0.27641966938972473, "learning_rate": 1.8626793839678536e-07, "loss": 0.2405, "step": 3485 }, { "epoch": 0.9276210750399149, "grad_norm": 0.27880874276161194, "learning_rate": 1.862593085746349e-07, "loss": 0.2346, "step": 3486 }, { "epoch": 0.9278871740287387, "grad_norm": 0.3086058497428894, "learning_rate": 1.8625067624169999e-07, "loss": 0.2288, "step": 3487 }, { "epoch": 0.9281532730175626, "grad_norm": 0.27327752113342285, "learning_rate": 1.862420413982318e-07, "loss": 0.2214, "step": 3488 }, { "epoch": 0.9284193720063864, "grad_norm": 0.25179219245910645, "learning_rate": 1.8623340404448173e-07, "loss": 0.1964, "step": 3489 }, { "epoch": 0.9286854709952103, "grad_norm": 0.2682383954524994, "learning_rate": 1.862247641807012e-07, "loss": 0.2196, "step": 3490 }, { "epoch": 0.928951569984034, "grad_norm": 0.28134214878082275, "learning_rate": 1.862161218071416e-07, "loss": 0.2187, "step": 3491 }, { "epoch": 0.929217668972858, "grad_norm": 0.2975175678730011, "learning_rate": 1.862074769240546e-07, "loss": 0.2471, "step": 3492 }, { "epoch": 0.9294837679616818, "grad_norm": 0.33298638463020325, "learning_rate": 1.8619882953169177e-07, "loss": 0.2268, "step": 3493 }, { "epoch": 0.9297498669505055, "grad_norm": 0.29025956988334656, "learning_rate": 1.861901796303048e-07, "loss": 0.2276, "step": 3494 }, { "epoch": 0.9300159659393294, "grad_norm": 0.2975033223628998, "learning_rate": 1.8618152722014553e-07, "loss": 0.2331, "step": 3495 }, { "epoch": 0.9302820649281532, "grad_norm": 0.29018861055374146, "learning_rate": 1.8617287230146577e-07, "loss": 0.2238, "step": 3496 }, { "epoch": 0.9305481639169771, "grad_norm": 0.28366875648498535, "learning_rate": 1.861642148745174e-07, "loss": 0.2251, "step": 3497 }, { "epoch": 0.9308142629058009, "grad_norm": 0.30621224641799927, "learning_rate": 1.8615555493955252e-07, "loss": 0.2273, "step": 3498 }, { "epoch": 0.9310803618946248, "grad_norm": 0.2744543254375458, "learning_rate": 1.8614689249682308e-07, "loss": 0.2201, "step": 3499 }, { "epoch": 0.9313464608834486, "grad_norm": 0.27138811349868774, "learning_rate": 1.861382275465813e-07, "loss": 0.2202, "step": 3500 }, { "epoch": 0.9316125598722725, "grad_norm": 0.2985500395298004, "learning_rate": 1.8612956008907936e-07, "loss": 0.2253, "step": 3501 }, { "epoch": 0.9318786588610963, "grad_norm": 0.29933905601501465, "learning_rate": 1.8612089012456956e-07, "loss": 0.2317, "step": 3502 }, { "epoch": 0.9321447578499201, "grad_norm": 0.3014257848262787, "learning_rate": 1.8611221765330427e-07, "loss": 0.2322, "step": 3503 }, { "epoch": 0.932410856838744, "grad_norm": 0.27489951252937317, "learning_rate": 1.8610354267553592e-07, "loss": 0.2025, "step": 3504 }, { "epoch": 0.9326769558275678, "grad_norm": 0.27117544412612915, "learning_rate": 1.86094865191517e-07, "loss": 0.2234, "step": 3505 }, { "epoch": 0.9329430548163917, "grad_norm": 0.28484439849853516, "learning_rate": 1.860861852015001e-07, "loss": 0.2273, "step": 3506 }, { "epoch": 0.9332091538052155, "grad_norm": 0.300538569688797, "learning_rate": 1.8607750270573787e-07, "loss": 0.23, "step": 3507 }, { "epoch": 0.9334752527940394, "grad_norm": 0.2815816402435303, "learning_rate": 1.8606881770448301e-07, "loss": 0.2193, "step": 3508 }, { "epoch": 0.9337413517828632, "grad_norm": 0.26431742310523987, "learning_rate": 1.8606013019798837e-07, "loss": 0.2141, "step": 3509 }, { "epoch": 0.9340074507716871, "grad_norm": 0.2746959626674652, "learning_rate": 1.860514401865068e-07, "loss": 0.2359, "step": 3510 }, { "epoch": 0.9342735497605109, "grad_norm": 0.282091349363327, "learning_rate": 1.8604274767029121e-07, "loss": 0.2232, "step": 3511 }, { "epoch": 0.9345396487493347, "grad_norm": 0.32798483967781067, "learning_rate": 1.8603405264959466e-07, "loss": 0.2351, "step": 3512 }, { "epoch": 0.9348057477381586, "grad_norm": 0.2708345949649811, "learning_rate": 1.8602535512467022e-07, "loss": 0.2132, "step": 3513 }, { "epoch": 0.9350718467269824, "grad_norm": 0.29591435194015503, "learning_rate": 1.8601665509577103e-07, "loss": 0.2343, "step": 3514 }, { "epoch": 0.9353379457158063, "grad_norm": 0.2892140746116638, "learning_rate": 1.8600795256315038e-07, "loss": 0.229, "step": 3515 }, { "epoch": 0.9356040447046301, "grad_norm": 0.26984331011772156, "learning_rate": 1.8599924752706156e-07, "loss": 0.222, "step": 3516 }, { "epoch": 0.935870143693454, "grad_norm": 0.2781253457069397, "learning_rate": 1.859905399877579e-07, "loss": 0.2241, "step": 3517 }, { "epoch": 0.9361362426822778, "grad_norm": 0.29093030095100403, "learning_rate": 1.8598182994549292e-07, "loss": 0.2374, "step": 3518 }, { "epoch": 0.9364023416711017, "grad_norm": 0.27872201800346375, "learning_rate": 1.859731174005201e-07, "loss": 0.2154, "step": 3519 }, { "epoch": 0.9366684406599255, "grad_norm": 0.2768096625804901, "learning_rate": 1.8596440235309308e-07, "loss": 0.2182, "step": 3520 }, { "epoch": 0.9369345396487493, "grad_norm": 0.25964128971099854, "learning_rate": 1.859556848034655e-07, "loss": 0.2107, "step": 3521 }, { "epoch": 0.9372006386375732, "grad_norm": 0.3098597824573517, "learning_rate": 1.8594696475189113e-07, "loss": 0.2274, "step": 3522 }, { "epoch": 0.937466737626397, "grad_norm": 0.2981788218021393, "learning_rate": 1.8593824219862376e-07, "loss": 0.2261, "step": 3523 }, { "epoch": 0.9377328366152209, "grad_norm": 0.2864568829536438, "learning_rate": 1.859295171439173e-07, "loss": 0.2168, "step": 3524 }, { "epoch": 0.9379989356040447, "grad_norm": 0.2813008427619934, "learning_rate": 1.859207895880257e-07, "loss": 0.223, "step": 3525 }, { "epoch": 0.9382650345928686, "grad_norm": 0.28942328691482544, "learning_rate": 1.8591205953120303e-07, "loss": 0.2186, "step": 3526 }, { "epoch": 0.9385311335816924, "grad_norm": 0.26428577303886414, "learning_rate": 1.8590332697370337e-07, "loss": 0.2203, "step": 3527 }, { "epoch": 0.9387972325705163, "grad_norm": 0.27300143241882324, "learning_rate": 1.8589459191578093e-07, "loss": 0.2089, "step": 3528 }, { "epoch": 0.9390633315593401, "grad_norm": 0.31180498003959656, "learning_rate": 1.858858543576899e-07, "loss": 0.2114, "step": 3529 }, { "epoch": 0.9393294305481639, "grad_norm": 0.2757982015609741, "learning_rate": 1.8587711429968466e-07, "loss": 0.2276, "step": 3530 }, { "epoch": 0.9395955295369878, "grad_norm": 0.31060537695884705, "learning_rate": 1.8586837174201958e-07, "loss": 0.243, "step": 3531 }, { "epoch": 0.9398616285258116, "grad_norm": 0.29367905855178833, "learning_rate": 1.858596266849492e-07, "loss": 0.2265, "step": 3532 }, { "epoch": 0.9401277275146355, "grad_norm": 0.28254738450050354, "learning_rate": 1.85850879128728e-07, "loss": 0.2091, "step": 3533 }, { "epoch": 0.9403938265034593, "grad_norm": 0.2822703421115875, "learning_rate": 1.858421290736106e-07, "loss": 0.2319, "step": 3534 }, { "epoch": 0.9406599254922832, "grad_norm": 0.27840062975883484, "learning_rate": 1.8583337651985174e-07, "loss": 0.2166, "step": 3535 }, { "epoch": 0.940926024481107, "grad_norm": 0.2748757004737854, "learning_rate": 1.8582462146770612e-07, "loss": 0.2237, "step": 3536 }, { "epoch": 0.9411921234699309, "grad_norm": 0.29262861609458923, "learning_rate": 1.8581586391742862e-07, "loss": 0.2294, "step": 3537 }, { "epoch": 0.9414582224587547, "grad_norm": 0.2821332514286041, "learning_rate": 1.8580710386927417e-07, "loss": 0.2157, "step": 3538 }, { "epoch": 0.9417243214475784, "grad_norm": 0.3013744056224823, "learning_rate": 1.8579834132349769e-07, "loss": 0.2241, "step": 3539 }, { "epoch": 0.9419904204364024, "grad_norm": 0.2842072546482086, "learning_rate": 1.857895762803543e-07, "loss": 0.2204, "step": 3540 }, { "epoch": 0.9422565194252261, "grad_norm": 0.2885734438896179, "learning_rate": 1.8578080874009907e-07, "loss": 0.2423, "step": 3541 }, { "epoch": 0.94252261841405, "grad_norm": 0.25431689620018005, "learning_rate": 1.8577203870298722e-07, "loss": 0.181, "step": 3542 }, { "epoch": 0.9427887174028738, "grad_norm": 0.298210084438324, "learning_rate": 1.8576326616927403e-07, "loss": 0.2166, "step": 3543 }, { "epoch": 0.9430548163916977, "grad_norm": 0.30104976892471313, "learning_rate": 1.8575449113921486e-07, "loss": 0.2365, "step": 3544 }, { "epoch": 0.9433209153805215, "grad_norm": 0.28615134954452515, "learning_rate": 1.8574571361306508e-07, "loss": 0.231, "step": 3545 }, { "epoch": 0.9435870143693454, "grad_norm": 0.28669798374176025, "learning_rate": 1.8573693359108025e-07, "loss": 0.2212, "step": 3546 }, { "epoch": 0.9438531133581692, "grad_norm": 0.27506202459335327, "learning_rate": 1.857281510735159e-07, "loss": 0.2245, "step": 3547 }, { "epoch": 0.944119212346993, "grad_norm": 0.2804330885410309, "learning_rate": 1.8571936606062763e-07, "loss": 0.2067, "step": 3548 }, { "epoch": 0.9443853113358169, "grad_norm": 0.28793683648109436, "learning_rate": 1.8571057855267118e-07, "loss": 0.2316, "step": 3549 }, { "epoch": 0.9446514103246407, "grad_norm": 0.29123741388320923, "learning_rate": 1.8570178854990237e-07, "loss": 0.2356, "step": 3550 }, { "epoch": 0.9449175093134646, "grad_norm": 0.2725684642791748, "learning_rate": 1.85692996052577e-07, "loss": 0.204, "step": 3551 }, { "epoch": 0.9451836083022884, "grad_norm": 0.2840830683708191, "learning_rate": 1.85684201060951e-07, "loss": 0.2367, "step": 3552 }, { "epoch": 0.9454497072911123, "grad_norm": 0.2777065336704254, "learning_rate": 1.8567540357528043e-07, "loss": 0.2226, "step": 3553 }, { "epoch": 0.9457158062799361, "grad_norm": 0.29922184348106384, "learning_rate": 1.8566660359582125e-07, "loss": 0.2008, "step": 3554 }, { "epoch": 0.94598190526876, "grad_norm": 0.2989305257797241, "learning_rate": 1.856578011228297e-07, "loss": 0.2247, "step": 3555 }, { "epoch": 0.9462480042575838, "grad_norm": 0.26279324293136597, "learning_rate": 1.8564899615656198e-07, "loss": 0.2258, "step": 3556 }, { "epoch": 0.9465141032464076, "grad_norm": 0.34626996517181396, "learning_rate": 1.8564018869727435e-07, "loss": 0.231, "step": 3557 }, { "epoch": 0.9467802022352315, "grad_norm": 0.27027562260627747, "learning_rate": 1.856313787452232e-07, "loss": 0.2124, "step": 3558 }, { "epoch": 0.9470463012240553, "grad_norm": 0.3046046495437622, "learning_rate": 1.8562256630066497e-07, "loss": 0.2238, "step": 3559 }, { "epoch": 0.9473124002128792, "grad_norm": 0.2781563103199005, "learning_rate": 1.8561375136385614e-07, "loss": 0.2283, "step": 3560 }, { "epoch": 0.947578499201703, "grad_norm": 0.2801470458507538, "learning_rate": 1.856049339350533e-07, "loss": 0.2378, "step": 3561 }, { "epoch": 0.9478445981905269, "grad_norm": 0.2835552990436554, "learning_rate": 1.855961140145131e-07, "loss": 0.2303, "step": 3562 }, { "epoch": 0.9481106971793507, "grad_norm": 0.2892826795578003, "learning_rate": 1.8558729160249226e-07, "loss": 0.2337, "step": 3563 }, { "epoch": 0.9483767961681746, "grad_norm": 0.27552473545074463, "learning_rate": 1.8557846669924758e-07, "loss": 0.2174, "step": 3564 }, { "epoch": 0.9486428951569984, "grad_norm": 0.2953459918498993, "learning_rate": 1.8556963930503597e-07, "loss": 0.2302, "step": 3565 }, { "epoch": 0.9489089941458222, "grad_norm": 0.2807513475418091, "learning_rate": 1.8556080942011434e-07, "loss": 0.2178, "step": 3566 }, { "epoch": 0.9491750931346461, "grad_norm": 0.2769887447357178, "learning_rate": 1.8555197704473966e-07, "loss": 0.2084, "step": 3567 }, { "epoch": 0.9494411921234699, "grad_norm": 0.26185327768325806, "learning_rate": 1.8554314217916912e-07, "loss": 0.2118, "step": 3568 }, { "epoch": 0.9497072911122938, "grad_norm": 0.28087082505226135, "learning_rate": 1.8553430482365978e-07, "loss": 0.2242, "step": 3569 }, { "epoch": 0.9499733901011176, "grad_norm": 0.28913038969039917, "learning_rate": 1.855254649784689e-07, "loss": 0.1974, "step": 3570 }, { "epoch": 0.9502394890899415, "grad_norm": 0.27755823731422424, "learning_rate": 1.8551662264385384e-07, "loss": 0.2149, "step": 3571 }, { "epoch": 0.9505055880787653, "grad_norm": 0.2977510690689087, "learning_rate": 1.855077778200719e-07, "loss": 0.2194, "step": 3572 }, { "epoch": 0.9507716870675892, "grad_norm": 0.2935078740119934, "learning_rate": 1.854989305073806e-07, "loss": 0.238, "step": 3573 }, { "epoch": 0.951037786056413, "grad_norm": 0.28135836124420166, "learning_rate": 1.8549008070603743e-07, "loss": 0.2279, "step": 3574 }, { "epoch": 0.9513038850452368, "grad_norm": 0.28987017273902893, "learning_rate": 1.8548122841629996e-07, "loss": 0.2425, "step": 3575 }, { "epoch": 0.9515699840340607, "grad_norm": 0.28039535880088806, "learning_rate": 1.8547237363842587e-07, "loss": 0.2099, "step": 3576 }, { "epoch": 0.9518360830228845, "grad_norm": 0.27625513076782227, "learning_rate": 1.8546351637267294e-07, "loss": 0.2066, "step": 3577 }, { "epoch": 0.9521021820117084, "grad_norm": 0.2582367956638336, "learning_rate": 1.8545465661929894e-07, "loss": 0.1971, "step": 3578 }, { "epoch": 0.9523682810005322, "grad_norm": 0.27253562211990356, "learning_rate": 1.8544579437856176e-07, "loss": 0.2302, "step": 3579 }, { "epoch": 0.9526343799893561, "grad_norm": 0.2744094729423523, "learning_rate": 1.8543692965071935e-07, "loss": 0.2161, "step": 3580 }, { "epoch": 0.9529004789781799, "grad_norm": 0.2628912031650543, "learning_rate": 1.854280624360298e-07, "loss": 0.2192, "step": 3581 }, { "epoch": 0.9531665779670038, "grad_norm": 0.2906242907047272, "learning_rate": 1.8541919273475112e-07, "loss": 0.241, "step": 3582 }, { "epoch": 0.9534326769558276, "grad_norm": 0.32443997263908386, "learning_rate": 1.8541032054714155e-07, "loss": 0.2287, "step": 3583 }, { "epoch": 0.9536987759446514, "grad_norm": 0.2859141528606415, "learning_rate": 1.8540144587345932e-07, "loss": 0.216, "step": 3584 }, { "epoch": 0.9539648749334753, "grad_norm": 0.2985920011997223, "learning_rate": 1.8539256871396271e-07, "loss": 0.2205, "step": 3585 }, { "epoch": 0.954230973922299, "grad_norm": 0.2730385661125183, "learning_rate": 1.8538368906891018e-07, "loss": 0.2134, "step": 3586 }, { "epoch": 0.954497072911123, "grad_norm": 0.28078845143318176, "learning_rate": 1.8537480693856017e-07, "loss": 0.222, "step": 3587 }, { "epoch": 0.9547631718999467, "grad_norm": 0.28872814774513245, "learning_rate": 1.8536592232317116e-07, "loss": 0.2227, "step": 3588 }, { "epoch": 0.9550292708887707, "grad_norm": 0.335854709148407, "learning_rate": 1.8535703522300184e-07, "loss": 0.2475, "step": 3589 }, { "epoch": 0.9552953698775944, "grad_norm": 0.30195385217666626, "learning_rate": 1.8534814563831082e-07, "loss": 0.2316, "step": 3590 }, { "epoch": 0.9555614688664184, "grad_norm": 0.30548661947250366, "learning_rate": 1.853392535693569e-07, "loss": 0.213, "step": 3591 }, { "epoch": 0.9558275678552421, "grad_norm": 0.25178396701812744, "learning_rate": 1.8533035901639889e-07, "loss": 0.2076, "step": 3592 }, { "epoch": 0.956093666844066, "grad_norm": 0.2741701006889343, "learning_rate": 1.8532146197969567e-07, "loss": 0.2132, "step": 3593 }, { "epoch": 0.9563597658328898, "grad_norm": 0.2781822085380554, "learning_rate": 1.8531256245950623e-07, "loss": 0.2306, "step": 3594 }, { "epoch": 0.9566258648217136, "grad_norm": 0.284816712141037, "learning_rate": 1.8530366045608962e-07, "loss": 0.2224, "step": 3595 }, { "epoch": 0.9568919638105375, "grad_norm": 0.2847689688205719, "learning_rate": 1.8529475596970497e-07, "loss": 0.2054, "step": 3596 }, { "epoch": 0.9571580627993613, "grad_norm": 0.29636988043785095, "learning_rate": 1.8528584900061142e-07, "loss": 0.2336, "step": 3597 }, { "epoch": 0.9574241617881852, "grad_norm": 0.2857840359210968, "learning_rate": 1.852769395490682e-07, "loss": 0.2244, "step": 3598 }, { "epoch": 0.957690260777009, "grad_norm": 0.3095129430294037, "learning_rate": 1.8526802761533474e-07, "loss": 0.2571, "step": 3599 }, { "epoch": 0.9579563597658329, "grad_norm": 0.27400973439216614, "learning_rate": 1.852591131996704e-07, "loss": 0.2121, "step": 3600 }, { "epoch": 0.9582224587546567, "grad_norm": 0.26140275597572327, "learning_rate": 1.852501963023346e-07, "loss": 0.195, "step": 3601 }, { "epoch": 0.9584885577434806, "grad_norm": 0.2895886301994324, "learning_rate": 1.8524127692358697e-07, "loss": 0.2199, "step": 3602 }, { "epoch": 0.9587546567323044, "grad_norm": 0.2895684540271759, "learning_rate": 1.8523235506368708e-07, "loss": 0.2176, "step": 3603 }, { "epoch": 0.9590207557211282, "grad_norm": 0.25223928689956665, "learning_rate": 1.8522343072289467e-07, "loss": 0.2088, "step": 3604 }, { "epoch": 0.9592868547099521, "grad_norm": 0.290118008852005, "learning_rate": 1.8521450390146943e-07, "loss": 0.2241, "step": 3605 }, { "epoch": 0.9595529536987759, "grad_norm": 0.2608511745929718, "learning_rate": 1.8520557459967128e-07, "loss": 0.2162, "step": 3606 }, { "epoch": 0.9598190526875998, "grad_norm": 0.2528879642486572, "learning_rate": 1.8519664281776006e-07, "loss": 0.2138, "step": 3607 }, { "epoch": 0.9600851516764236, "grad_norm": 0.27401116490364075, "learning_rate": 1.8518770855599576e-07, "loss": 0.2182, "step": 3608 }, { "epoch": 0.9603512506652475, "grad_norm": 0.2868657410144806, "learning_rate": 1.851787718146385e-07, "loss": 0.2341, "step": 3609 }, { "epoch": 0.9606173496540713, "grad_norm": 0.2910914421081543, "learning_rate": 1.8516983259394833e-07, "loss": 0.2283, "step": 3610 }, { "epoch": 0.9608834486428952, "grad_norm": 0.29753443598747253, "learning_rate": 1.8516089089418547e-07, "loss": 0.2461, "step": 3611 }, { "epoch": 0.961149547631719, "grad_norm": 0.28265079855918884, "learning_rate": 1.8515194671561023e-07, "loss": 0.1965, "step": 3612 }, { "epoch": 0.9614156466205428, "grad_norm": 0.2908861041069031, "learning_rate": 1.8514300005848287e-07, "loss": 0.228, "step": 3613 }, { "epoch": 0.9616817456093667, "grad_norm": 0.28509747982025146, "learning_rate": 1.8513405092306386e-07, "loss": 0.2284, "step": 3614 }, { "epoch": 0.9619478445981905, "grad_norm": 0.2956513464450836, "learning_rate": 1.8512509930961368e-07, "loss": 0.216, "step": 3615 }, { "epoch": 0.9622139435870144, "grad_norm": 0.2673508822917938, "learning_rate": 1.8511614521839289e-07, "loss": 0.2115, "step": 3616 }, { "epoch": 0.9624800425758382, "grad_norm": 0.28515079617500305, "learning_rate": 1.851071886496621e-07, "loss": 0.2341, "step": 3617 }, { "epoch": 0.9627461415646621, "grad_norm": 0.287689208984375, "learning_rate": 1.8509822960368207e-07, "loss": 0.2427, "step": 3618 }, { "epoch": 0.9630122405534859, "grad_norm": 0.27071067690849304, "learning_rate": 1.850892680807135e-07, "loss": 0.2075, "step": 3619 }, { "epoch": 0.9632783395423098, "grad_norm": 0.2832478880882263, "learning_rate": 1.8508030408101725e-07, "loss": 0.2239, "step": 3620 }, { "epoch": 0.9635444385311336, "grad_norm": 0.2718465328216553, "learning_rate": 1.8507133760485427e-07, "loss": 0.2013, "step": 3621 }, { "epoch": 0.9638105375199574, "grad_norm": 0.2757318317890167, "learning_rate": 1.8506236865248558e-07, "loss": 0.208, "step": 3622 }, { "epoch": 0.9640766365087813, "grad_norm": 0.3019773066043854, "learning_rate": 1.8505339722417213e-07, "loss": 0.2387, "step": 3623 }, { "epoch": 0.9643427354976051, "grad_norm": 0.3612479865550995, "learning_rate": 1.8504442332017517e-07, "loss": 0.2403, "step": 3624 }, { "epoch": 0.964608834486429, "grad_norm": 0.4598715305328369, "learning_rate": 1.8503544694075586e-07, "loss": 0.2261, "step": 3625 }, { "epoch": 0.9648749334752528, "grad_norm": 0.28083229064941406, "learning_rate": 1.8502646808617545e-07, "loss": 0.2371, "step": 3626 }, { "epoch": 0.9651410324640767, "grad_norm": 0.2805500030517578, "learning_rate": 1.8501748675669539e-07, "loss": 0.2198, "step": 3627 }, { "epoch": 0.9654071314529005, "grad_norm": 0.25826296210289, "learning_rate": 1.8500850295257696e-07, "loss": 0.2015, "step": 3628 }, { "epoch": 0.9656732304417244, "grad_norm": 0.28306934237480164, "learning_rate": 1.8499951667408178e-07, "loss": 0.2406, "step": 3629 }, { "epoch": 0.9659393294305482, "grad_norm": 0.3272726237773895, "learning_rate": 1.8499052792147136e-07, "loss": 0.2304, "step": 3630 }, { "epoch": 0.966205428419372, "grad_norm": 0.27908894419670105, "learning_rate": 1.8498153669500733e-07, "loss": 0.2398, "step": 3631 }, { "epoch": 0.9664715274081959, "grad_norm": 0.26117202639579773, "learning_rate": 1.8497254299495145e-07, "loss": 0.2116, "step": 3632 }, { "epoch": 0.9667376263970197, "grad_norm": 0.27662432193756104, "learning_rate": 1.8496354682156543e-07, "loss": 0.2065, "step": 3633 }, { "epoch": 0.9670037253858436, "grad_norm": 0.27010872960090637, "learning_rate": 1.8495454817511118e-07, "loss": 0.2192, "step": 3634 }, { "epoch": 0.9672698243746674, "grad_norm": 0.2952907979488373, "learning_rate": 1.8494554705585064e-07, "loss": 0.2119, "step": 3635 }, { "epoch": 0.9675359233634913, "grad_norm": 0.2807266414165497, "learning_rate": 1.8493654346404578e-07, "loss": 0.2177, "step": 3636 }, { "epoch": 0.967802022352315, "grad_norm": 0.281482458114624, "learning_rate": 1.8492753739995865e-07, "loss": 0.2198, "step": 3637 }, { "epoch": 0.968068121341139, "grad_norm": 0.28435540199279785, "learning_rate": 1.8491852886385145e-07, "loss": 0.2082, "step": 3638 }, { "epoch": 0.9683342203299627, "grad_norm": 0.288505494594574, "learning_rate": 1.8490951785598635e-07, "loss": 0.2247, "step": 3639 }, { "epoch": 0.9686003193187865, "grad_norm": 0.29252979159355164, "learning_rate": 1.8490050437662563e-07, "loss": 0.2218, "step": 3640 }, { "epoch": 0.9688664183076104, "grad_norm": 0.2911521792411804, "learning_rate": 1.848914884260317e-07, "loss": 0.224, "step": 3641 }, { "epoch": 0.9691325172964342, "grad_norm": 0.3005727231502533, "learning_rate": 1.8488247000446692e-07, "loss": 0.2389, "step": 3642 }, { "epoch": 0.9693986162852581, "grad_norm": 0.2758016586303711, "learning_rate": 1.8487344911219389e-07, "loss": 0.2269, "step": 3643 }, { "epoch": 0.9696647152740819, "grad_norm": 0.30899256467819214, "learning_rate": 1.848644257494751e-07, "loss": 0.2424, "step": 3644 }, { "epoch": 0.9699308142629058, "grad_norm": 0.29200151562690735, "learning_rate": 1.8485539991657323e-07, "loss": 0.2464, "step": 3645 }, { "epoch": 0.9701969132517296, "grad_norm": 0.28294670581817627, "learning_rate": 1.8484637161375095e-07, "loss": 0.2254, "step": 3646 }, { "epoch": 0.9704630122405535, "grad_norm": 0.28058841824531555, "learning_rate": 1.8483734084127116e-07, "loss": 0.2242, "step": 3647 }, { "epoch": 0.9707291112293773, "grad_norm": 0.269767701625824, "learning_rate": 1.848283075993966e-07, "loss": 0.2125, "step": 3648 }, { "epoch": 0.9709952102182011, "grad_norm": 0.27598103880882263, "learning_rate": 1.848192718883903e-07, "loss": 0.2214, "step": 3649 }, { "epoch": 0.971261309207025, "grad_norm": 0.29554513096809387, "learning_rate": 1.848102337085152e-07, "loss": 0.2329, "step": 3650 }, { "epoch": 0.9715274081958488, "grad_norm": 0.28030169010162354, "learning_rate": 1.8480119306003444e-07, "loss": 0.2421, "step": 3651 }, { "epoch": 0.9717935071846727, "grad_norm": 0.29079070687294006, "learning_rate": 1.8479214994321112e-07, "loss": 0.2367, "step": 3652 }, { "epoch": 0.9720596061734965, "grad_norm": 0.2837863862514496, "learning_rate": 1.8478310435830844e-07, "loss": 0.2213, "step": 3653 }, { "epoch": 0.9723257051623204, "grad_norm": 0.31454554200172424, "learning_rate": 1.8477405630558977e-07, "loss": 0.2597, "step": 3654 }, { "epoch": 0.9725918041511442, "grad_norm": 0.3137568235397339, "learning_rate": 1.847650057853184e-07, "loss": 0.2314, "step": 3655 }, { "epoch": 0.9728579031399681, "grad_norm": 0.2773192822933197, "learning_rate": 1.8475595279775784e-07, "loss": 0.2173, "step": 3656 }, { "epoch": 0.9731240021287919, "grad_norm": 0.2911182940006256, "learning_rate": 1.8474689734317154e-07, "loss": 0.2351, "step": 3657 }, { "epoch": 0.9733901011176157, "grad_norm": 0.29393744468688965, "learning_rate": 1.8473783942182312e-07, "loss": 0.2465, "step": 3658 }, { "epoch": 0.9736562001064396, "grad_norm": 0.2816178798675537, "learning_rate": 1.847287790339762e-07, "loss": 0.2241, "step": 3659 }, { "epoch": 0.9739222990952634, "grad_norm": 0.2655209004878998, "learning_rate": 1.847197161798945e-07, "loss": 0.2207, "step": 3660 }, { "epoch": 0.9741883980840873, "grad_norm": 0.27318790555000305, "learning_rate": 1.8471065085984185e-07, "loss": 0.2139, "step": 3661 }, { "epoch": 0.9744544970729111, "grad_norm": 0.2753976881504059, "learning_rate": 1.8470158307408211e-07, "loss": 0.2129, "step": 3662 }, { "epoch": 0.974720596061735, "grad_norm": 0.2976611256599426, "learning_rate": 1.8469251282287922e-07, "loss": 0.2325, "step": 3663 }, { "epoch": 0.9749866950505588, "grad_norm": 0.29974809288978577, "learning_rate": 1.8468344010649716e-07, "loss": 0.2113, "step": 3664 }, { "epoch": 0.9752527940393827, "grad_norm": 0.2931012213230133, "learning_rate": 1.8467436492520006e-07, "loss": 0.2272, "step": 3665 }, { "epoch": 0.9755188930282065, "grad_norm": 0.25676214694976807, "learning_rate": 1.8466528727925202e-07, "loss": 0.205, "step": 3666 }, { "epoch": 0.9757849920170303, "grad_norm": 0.2805984616279602, "learning_rate": 1.8465620716891731e-07, "loss": 0.224, "step": 3667 }, { "epoch": 0.9760510910058542, "grad_norm": 0.28412479162216187, "learning_rate": 1.8464712459446027e-07, "loss": 0.2356, "step": 3668 }, { "epoch": 0.976317189994678, "grad_norm": 0.2877090275287628, "learning_rate": 1.8463803955614516e-07, "loss": 0.216, "step": 3669 }, { "epoch": 0.9765832889835019, "grad_norm": 0.2970064580440521, "learning_rate": 1.8462895205423646e-07, "loss": 0.247, "step": 3670 }, { "epoch": 0.9768493879723257, "grad_norm": 0.29096072912216187, "learning_rate": 1.8461986208899875e-07, "loss": 0.216, "step": 3671 }, { "epoch": 0.9771154869611496, "grad_norm": 0.31246036291122437, "learning_rate": 1.8461076966069652e-07, "loss": 0.2431, "step": 3672 }, { "epoch": 0.9773815859499734, "grad_norm": 0.2786523401737213, "learning_rate": 1.846016747695945e-07, "loss": 0.225, "step": 3673 }, { "epoch": 0.9776476849387973, "grad_norm": 0.2732962369918823, "learning_rate": 1.8459257741595744e-07, "loss": 0.2152, "step": 3674 }, { "epoch": 0.9779137839276211, "grad_norm": 0.2784125506877899, "learning_rate": 1.8458347760005004e-07, "loss": 0.2332, "step": 3675 }, { "epoch": 0.9781798829164449, "grad_norm": 0.27950987219810486, "learning_rate": 1.8457437532213725e-07, "loss": 0.2001, "step": 3676 }, { "epoch": 0.9784459819052688, "grad_norm": 0.2598441541194916, "learning_rate": 1.8456527058248398e-07, "loss": 0.2212, "step": 3677 }, { "epoch": 0.9787120808940926, "grad_norm": 0.2981094419956207, "learning_rate": 1.8455616338135523e-07, "loss": 0.1987, "step": 3678 }, { "epoch": 0.9789781798829165, "grad_norm": 0.29237762093544006, "learning_rate": 1.8454705371901614e-07, "loss": 0.2486, "step": 3679 }, { "epoch": 0.9792442788717403, "grad_norm": 0.2619573771953583, "learning_rate": 1.8453794159573183e-07, "loss": 0.216, "step": 3680 }, { "epoch": 0.9795103778605642, "grad_norm": 0.29574981331825256, "learning_rate": 1.8452882701176757e-07, "loss": 0.2401, "step": 3681 }, { "epoch": 0.979776476849388, "grad_norm": 0.30691200494766235, "learning_rate": 1.845197099673886e-07, "loss": 0.2252, "step": 3682 }, { "epoch": 0.9800425758382119, "grad_norm": 0.31516721844673157, "learning_rate": 1.8451059046286033e-07, "loss": 0.2354, "step": 3683 }, { "epoch": 0.9803086748270357, "grad_norm": 0.3089110255241394, "learning_rate": 1.845014684984482e-07, "loss": 0.2265, "step": 3684 }, { "epoch": 0.9805747738158594, "grad_norm": 0.278164267539978, "learning_rate": 1.8449234407441773e-07, "loss": 0.2185, "step": 3685 }, { "epoch": 0.9808408728046834, "grad_norm": 0.2729807198047638, "learning_rate": 1.8448321719103453e-07, "loss": 0.213, "step": 3686 }, { "epoch": 0.9811069717935071, "grad_norm": 0.2813704013824463, "learning_rate": 1.844740878485642e-07, "loss": 0.2181, "step": 3687 }, { "epoch": 0.981373070782331, "grad_norm": 0.28445926308631897, "learning_rate": 1.8446495604727254e-07, "loss": 0.2303, "step": 3688 }, { "epoch": 0.9816391697711548, "grad_norm": 0.3604835867881775, "learning_rate": 1.8445582178742529e-07, "loss": 0.2615, "step": 3689 }, { "epoch": 0.9819052687599787, "grad_norm": 0.28232088685035706, "learning_rate": 1.8444668506928839e-07, "loss": 0.2166, "step": 3690 }, { "epoch": 0.9821713677488025, "grad_norm": 0.28990840911865234, "learning_rate": 1.8443754589312773e-07, "loss": 0.2406, "step": 3691 }, { "epoch": 0.9824374667376264, "grad_norm": 0.27134621143341064, "learning_rate": 1.8442840425920934e-07, "loss": 0.2241, "step": 3692 }, { "epoch": 0.9827035657264502, "grad_norm": 0.27197620272636414, "learning_rate": 1.8441926016779931e-07, "loss": 0.2212, "step": 3693 }, { "epoch": 0.982969664715274, "grad_norm": 0.29662662744522095, "learning_rate": 1.8441011361916386e-07, "loss": 0.2221, "step": 3694 }, { "epoch": 0.9832357637040979, "grad_norm": 0.27461570501327515, "learning_rate": 1.8440096461356913e-07, "loss": 0.2043, "step": 3695 }, { "epoch": 0.9835018626929217, "grad_norm": 0.2843227982521057, "learning_rate": 1.8439181315128147e-07, "loss": 0.2277, "step": 3696 }, { "epoch": 0.9837679616817456, "grad_norm": 0.24887509644031525, "learning_rate": 1.8438265923256724e-07, "loss": 0.1883, "step": 3697 }, { "epoch": 0.9840340606705694, "grad_norm": 0.29796844720840454, "learning_rate": 1.8437350285769294e-07, "loss": 0.2331, "step": 3698 }, { "epoch": 0.9843001596593933, "grad_norm": 0.2751643657684326, "learning_rate": 1.84364344026925e-07, "loss": 0.2245, "step": 3699 }, { "epoch": 0.9845662586482171, "grad_norm": 0.27604350447654724, "learning_rate": 1.8435518274053005e-07, "loss": 0.2204, "step": 3700 }, { "epoch": 0.984832357637041, "grad_norm": 0.3051399886608124, "learning_rate": 1.843460189987748e-07, "loss": 0.2105, "step": 3701 }, { "epoch": 0.9850984566258648, "grad_norm": 0.280321329832077, "learning_rate": 1.843368528019259e-07, "loss": 0.2251, "step": 3702 }, { "epoch": 0.9853645556146887, "grad_norm": 0.28651905059814453, "learning_rate": 1.8432768415025022e-07, "loss": 0.2278, "step": 3703 }, { "epoch": 0.9856306546035125, "grad_norm": 0.33172285556793213, "learning_rate": 1.843185130440146e-07, "loss": 0.2424, "step": 3704 }, { "epoch": 0.9858967535923363, "grad_norm": 0.2745101749897003, "learning_rate": 1.8430933948348598e-07, "loss": 0.2228, "step": 3705 }, { "epoch": 0.9861628525811602, "grad_norm": 0.288189172744751, "learning_rate": 1.8430016346893143e-07, "loss": 0.2306, "step": 3706 }, { "epoch": 0.986428951569984, "grad_norm": 0.2771863341331482, "learning_rate": 1.8429098500061798e-07, "loss": 0.2287, "step": 3707 }, { "epoch": 0.9866950505588079, "grad_norm": 0.28026092052459717, "learning_rate": 1.8428180407881283e-07, "loss": 0.2425, "step": 3708 }, { "epoch": 0.9869611495476317, "grad_norm": 0.29391905665397644, "learning_rate": 1.8427262070378318e-07, "loss": 0.2173, "step": 3709 }, { "epoch": 0.9872272485364556, "grad_norm": 0.2503069043159485, "learning_rate": 1.8426343487579636e-07, "loss": 0.2003, "step": 3710 }, { "epoch": 0.9874933475252794, "grad_norm": 0.2566724717617035, "learning_rate": 1.8425424659511973e-07, "loss": 0.2045, "step": 3711 }, { "epoch": 0.9877594465141033, "grad_norm": 0.27371829748153687, "learning_rate": 1.8424505586202078e-07, "loss": 0.224, "step": 3712 }, { "epoch": 0.9880255455029271, "grad_norm": 0.2573302090167999, "learning_rate": 1.8423586267676698e-07, "loss": 0.2172, "step": 3713 }, { "epoch": 0.9882916444917509, "grad_norm": 0.26362043619155884, "learning_rate": 1.8422666703962588e-07, "loss": 0.2134, "step": 3714 }, { "epoch": 0.9885577434805748, "grad_norm": 0.34077706933021545, "learning_rate": 1.8421746895086525e-07, "loss": 0.2546, "step": 3715 }, { "epoch": 0.9888238424693986, "grad_norm": 0.27554741501808167, "learning_rate": 1.8420826841075274e-07, "loss": 0.2266, "step": 3716 }, { "epoch": 0.9890899414582225, "grad_norm": 0.30969998240470886, "learning_rate": 1.8419906541955622e-07, "loss": 0.2477, "step": 3717 }, { "epoch": 0.9893560404470463, "grad_norm": 0.2524327039718628, "learning_rate": 1.841898599775435e-07, "loss": 0.2064, "step": 3718 }, { "epoch": 0.9896221394358702, "grad_norm": 0.2849220931529999, "learning_rate": 1.8418065208498256e-07, "loss": 0.2337, "step": 3719 }, { "epoch": 0.989888238424694, "grad_norm": 0.28847038745880127, "learning_rate": 1.841714417421414e-07, "loss": 0.2227, "step": 3720 }, { "epoch": 0.9901543374135179, "grad_norm": 0.32501280307769775, "learning_rate": 1.8416222894928813e-07, "loss": 0.2035, "step": 3721 }, { "epoch": 0.9904204364023417, "grad_norm": 0.2741785943508148, "learning_rate": 1.841530137066909e-07, "loss": 0.2248, "step": 3722 }, { "epoch": 0.9906865353911655, "grad_norm": 0.3021405041217804, "learning_rate": 1.8414379601461795e-07, "loss": 0.2124, "step": 3723 }, { "epoch": 0.9909526343799894, "grad_norm": 0.36331552267074585, "learning_rate": 1.8413457587333755e-07, "loss": 0.2083, "step": 3724 }, { "epoch": 0.9912187333688132, "grad_norm": 0.29394569993019104, "learning_rate": 1.8412535328311812e-07, "loss": 0.2168, "step": 3725 }, { "epoch": 0.9914848323576371, "grad_norm": 0.28296321630477905, "learning_rate": 1.8411612824422807e-07, "loss": 0.2189, "step": 3726 }, { "epoch": 0.9917509313464609, "grad_norm": 0.2824600636959076, "learning_rate": 1.8410690075693594e-07, "loss": 0.2126, "step": 3727 }, { "epoch": 0.9920170303352848, "grad_norm": 0.29428043961524963, "learning_rate": 1.8409767082151032e-07, "loss": 0.2225, "step": 3728 }, { "epoch": 0.9922831293241086, "grad_norm": 0.30802664160728455, "learning_rate": 1.8408843843821983e-07, "loss": 0.2041, "step": 3729 }, { "epoch": 0.9925492283129325, "grad_norm": 0.2715626358985901, "learning_rate": 1.8407920360733328e-07, "loss": 0.219, "step": 3730 }, { "epoch": 0.9928153273017563, "grad_norm": 0.28029531240463257, "learning_rate": 1.8406996632911938e-07, "loss": 0.2074, "step": 3731 }, { "epoch": 0.99308142629058, "grad_norm": 0.31459128856658936, "learning_rate": 1.8406072660384707e-07, "loss": 0.2514, "step": 3732 }, { "epoch": 0.993347525279404, "grad_norm": 0.2789345383644104, "learning_rate": 1.8405148443178523e-07, "loss": 0.2251, "step": 3733 }, { "epoch": 0.9936136242682277, "grad_norm": 0.3028048276901245, "learning_rate": 1.8404223981320298e-07, "loss": 0.2407, "step": 3734 }, { "epoch": 0.9938797232570517, "grad_norm": 0.30859676003456116, "learning_rate": 1.840329927483693e-07, "loss": 0.2374, "step": 3735 }, { "epoch": 0.9941458222458754, "grad_norm": 0.2813534736633301, "learning_rate": 1.8402374323755338e-07, "loss": 0.263, "step": 3736 }, { "epoch": 0.9944119212346993, "grad_norm": 0.2729000151157379, "learning_rate": 1.840144912810245e-07, "loss": 0.2135, "step": 3737 }, { "epoch": 0.9946780202235231, "grad_norm": 0.27833807468414307, "learning_rate": 1.840052368790519e-07, "loss": 0.2205, "step": 3738 }, { "epoch": 0.994944119212347, "grad_norm": 0.2950703203678131, "learning_rate": 1.8399598003190496e-07, "loss": 0.2106, "step": 3739 }, { "epoch": 0.9952102182011708, "grad_norm": 0.30502963066101074, "learning_rate": 1.8398672073985314e-07, "loss": 0.2183, "step": 3740 }, { "epoch": 0.9954763171899946, "grad_norm": 0.2788505554199219, "learning_rate": 1.8397745900316598e-07, "loss": 0.2276, "step": 3741 }, { "epoch": 0.9957424161788185, "grad_norm": 0.28269532322883606, "learning_rate": 1.8396819482211299e-07, "loss": 0.2066, "step": 3742 }, { "epoch": 0.9960085151676423, "grad_norm": 0.2633528411388397, "learning_rate": 1.8395892819696387e-07, "loss": 0.2083, "step": 3743 }, { "epoch": 0.9962746141564662, "grad_norm": 0.29109954833984375, "learning_rate": 1.8394965912798836e-07, "loss": 0.2308, "step": 3744 }, { "epoch": 0.99654071314529, "grad_norm": 0.3154776692390442, "learning_rate": 1.8394038761545625e-07, "loss": 0.2511, "step": 3745 }, { "epoch": 0.9968068121341139, "grad_norm": 0.28085651993751526, "learning_rate": 1.839311136596374e-07, "loss": 0.2347, "step": 3746 }, { "epoch": 0.9970729111229377, "grad_norm": 0.2846187651157379, "learning_rate": 1.839218372608018e-07, "loss": 0.2336, "step": 3747 }, { "epoch": 0.9973390101117616, "grad_norm": 0.3113308846950531, "learning_rate": 1.8391255841921937e-07, "loss": 0.2388, "step": 3748 }, { "epoch": 0.9976051091005854, "grad_norm": 0.2684200704097748, "learning_rate": 1.8390327713516025e-07, "loss": 0.2189, "step": 3749 }, { "epoch": 0.9978712080894092, "grad_norm": 0.26691383123397827, "learning_rate": 1.838939934088946e-07, "loss": 0.2062, "step": 3750 }, { "epoch": 0.9981373070782331, "grad_norm": 0.26189449429512024, "learning_rate": 1.838847072406926e-07, "loss": 0.2167, "step": 3751 }, { "epoch": 0.9984034060670569, "grad_norm": 0.2974793016910553, "learning_rate": 1.838754186308246e-07, "loss": 0.2299, "step": 3752 }, { "epoch": 0.9986695050558808, "grad_norm": 0.29972580075263977, "learning_rate": 1.8386612757956096e-07, "loss": 0.2057, "step": 3753 }, { "epoch": 0.9989356040447046, "grad_norm": 0.31530100107192993, "learning_rate": 1.8385683408717206e-07, "loss": 0.2319, "step": 3754 }, { "epoch": 0.9992017030335285, "grad_norm": 0.2637577950954437, "learning_rate": 1.8384753815392845e-07, "loss": 0.2061, "step": 3755 }, { "epoch": 0.9994678020223523, "grad_norm": 0.2801492512226105, "learning_rate": 1.8383823978010072e-07, "loss": 0.2005, "step": 3756 }, { "epoch": 0.9997339010111762, "grad_norm": 0.2812526822090149, "learning_rate": 1.8382893896595956e-07, "loss": 0.2165, "step": 3757 }, { "epoch": 1.0, "grad_norm": 0.2945724427700043, "learning_rate": 1.8381963571177562e-07, "loss": 0.2114, "step": 3758 }, { "epoch": 1.000266098988824, "grad_norm": 0.27816447615623474, "learning_rate": 1.838103300178197e-07, "loss": 0.2154, "step": 3759 }, { "epoch": 1.0005321979776476, "grad_norm": 0.2934087812900543, "learning_rate": 1.838010218843627e-07, "loss": 0.2061, "step": 3760 }, { "epoch": 1.0007982969664715, "grad_norm": 0.27150747179985046, "learning_rate": 1.8379171131167553e-07, "loss": 0.2023, "step": 3761 }, { "epoch": 1.0010643959552954, "grad_norm": 0.288504421710968, "learning_rate": 1.8378239830002925e-07, "loss": 0.2073, "step": 3762 }, { "epoch": 1.0013304949441193, "grad_norm": 0.2711051404476166, "learning_rate": 1.8377308284969485e-07, "loss": 0.2159, "step": 3763 }, { "epoch": 1.001596593932943, "grad_norm": 0.2715613842010498, "learning_rate": 1.8376376496094352e-07, "loss": 0.2081, "step": 3764 }, { "epoch": 1.0018626929217669, "grad_norm": 0.2813146114349365, "learning_rate": 1.837544446340465e-07, "loss": 0.2271, "step": 3765 }, { "epoch": 1.0021287919105908, "grad_norm": 0.28987354040145874, "learning_rate": 1.8374512186927508e-07, "loss": 0.2257, "step": 3766 }, { "epoch": 1.0023948908994147, "grad_norm": 0.3100341260433197, "learning_rate": 1.8373579666690056e-07, "loss": 0.2263, "step": 3767 }, { "epoch": 1.0026609898882384, "grad_norm": 0.28580230474472046, "learning_rate": 1.8372646902719444e-07, "loss": 0.2159, "step": 3768 }, { "epoch": 1.0029270888770623, "grad_norm": 0.3368815779685974, "learning_rate": 1.837171389504282e-07, "loss": 0.2546, "step": 3769 }, { "epoch": 1.0031931878658862, "grad_norm": 0.2699655294418335, "learning_rate": 1.8370780643687342e-07, "loss": 0.2064, "step": 3770 }, { "epoch": 1.0034592868547099, "grad_norm": 0.27813175320625305, "learning_rate": 1.8369847148680174e-07, "loss": 0.2237, "step": 3771 }, { "epoch": 1.0037253858435338, "grad_norm": 0.31991463899612427, "learning_rate": 1.8368913410048485e-07, "loss": 0.2213, "step": 3772 }, { "epoch": 1.0039914848323577, "grad_norm": 0.28710228204727173, "learning_rate": 1.836797942781946e-07, "loss": 0.2148, "step": 3773 }, { "epoch": 1.0042575838211816, "grad_norm": 0.2599191963672638, "learning_rate": 1.836704520202028e-07, "loss": 0.203, "step": 3774 }, { "epoch": 1.0045236828100053, "grad_norm": 0.2867121398448944, "learning_rate": 1.8366110732678138e-07, "loss": 0.2418, "step": 3775 }, { "epoch": 1.0047897817988292, "grad_norm": 0.3214898407459259, "learning_rate": 1.8365176019820233e-07, "loss": 0.244, "step": 3776 }, { "epoch": 1.005055880787653, "grad_norm": 0.291041761636734, "learning_rate": 1.8364241063473777e-07, "loss": 0.2276, "step": 3777 }, { "epoch": 1.0053219797764767, "grad_norm": 0.3041697144508362, "learning_rate": 1.836330586366598e-07, "loss": 0.2304, "step": 3778 }, { "epoch": 1.0055880787653007, "grad_norm": 0.275178462266922, "learning_rate": 1.8362370420424068e-07, "loss": 0.2318, "step": 3779 }, { "epoch": 1.0058541777541246, "grad_norm": 0.28295984864234924, "learning_rate": 1.8361434733775262e-07, "loss": 0.1948, "step": 3780 }, { "epoch": 1.0061202767429485, "grad_norm": 0.29154837131500244, "learning_rate": 1.8360498803746802e-07, "loss": 0.2315, "step": 3781 }, { "epoch": 1.0063863757317721, "grad_norm": 0.2756801247596741, "learning_rate": 1.835956263036593e-07, "loss": 0.2345, "step": 3782 }, { "epoch": 1.006652474720596, "grad_norm": 0.3012678921222687, "learning_rate": 1.8358626213659892e-07, "loss": 0.2287, "step": 3783 }, { "epoch": 1.00691857370942, "grad_norm": 0.30031725764274597, "learning_rate": 1.8357689553655952e-07, "loss": 0.2308, "step": 3784 }, { "epoch": 1.0071846726982439, "grad_norm": 0.286491334438324, "learning_rate": 1.8356752650381368e-07, "loss": 0.2229, "step": 3785 }, { "epoch": 1.0074507716870675, "grad_norm": 0.3000575006008148, "learning_rate": 1.8355815503863411e-07, "loss": 0.2232, "step": 3786 }, { "epoch": 1.0077168706758914, "grad_norm": 0.28757020831108093, "learning_rate": 1.8354878114129364e-07, "loss": 0.2226, "step": 3787 }, { "epoch": 1.0079829696647153, "grad_norm": 0.2732967734336853, "learning_rate": 1.8353940481206504e-07, "loss": 0.2135, "step": 3788 }, { "epoch": 1.008249068653539, "grad_norm": 0.28570878505706787, "learning_rate": 1.8353002605122132e-07, "loss": 0.2193, "step": 3789 }, { "epoch": 1.008515167642363, "grad_norm": 0.29104629158973694, "learning_rate": 1.835206448590354e-07, "loss": 0.2422, "step": 3790 }, { "epoch": 1.0087812666311868, "grad_norm": 0.270077645778656, "learning_rate": 1.8351126123578036e-07, "loss": 0.2107, "step": 3791 }, { "epoch": 1.0090473656200107, "grad_norm": 0.2927228808403015, "learning_rate": 1.8350187518172935e-07, "loss": 0.2438, "step": 3792 }, { "epoch": 1.0093134646088344, "grad_norm": 0.3219139873981476, "learning_rate": 1.8349248669715558e-07, "loss": 0.2475, "step": 3793 }, { "epoch": 1.0095795635976583, "grad_norm": 0.2784225046634674, "learning_rate": 1.8348309578233229e-07, "loss": 0.2052, "step": 3794 }, { "epoch": 1.0098456625864822, "grad_norm": 0.2911497950553894, "learning_rate": 1.8347370243753283e-07, "loss": 0.2069, "step": 3795 }, { "epoch": 1.010111761575306, "grad_norm": 0.28238511085510254, "learning_rate": 1.8346430666303066e-07, "loss": 0.2218, "step": 3796 }, { "epoch": 1.0103778605641298, "grad_norm": 0.2746407389640808, "learning_rate": 1.8345490845909922e-07, "loss": 0.2325, "step": 3797 }, { "epoch": 1.0106439595529537, "grad_norm": 0.26413559913635254, "learning_rate": 1.8344550782601208e-07, "loss": 0.2134, "step": 3798 }, { "epoch": 1.0109100585417776, "grad_norm": 0.29786884784698486, "learning_rate": 1.8343610476404286e-07, "loss": 0.2484, "step": 3799 }, { "epoch": 1.0111761575306013, "grad_norm": 0.30471447110176086, "learning_rate": 1.8342669927346531e-07, "loss": 0.2274, "step": 3800 }, { "epoch": 1.0114422565194252, "grad_norm": 0.2901673913002014, "learning_rate": 1.8341729135455312e-07, "loss": 0.2239, "step": 3801 }, { "epoch": 1.0117083555082491, "grad_norm": 0.2902225852012634, "learning_rate": 1.8340788100758017e-07, "loss": 0.2271, "step": 3802 }, { "epoch": 1.011974454497073, "grad_norm": 0.30381831526756287, "learning_rate": 1.833984682328204e-07, "loss": 0.2396, "step": 3803 }, { "epoch": 1.0122405534858967, "grad_norm": 0.28997117280960083, "learning_rate": 1.8338905303054775e-07, "loss": 0.2381, "step": 3804 }, { "epoch": 1.0125066524747206, "grad_norm": 0.28884613513946533, "learning_rate": 1.8337963540103626e-07, "loss": 0.2164, "step": 3805 }, { "epoch": 1.0127727514635445, "grad_norm": 0.2576584219932556, "learning_rate": 1.8337021534456013e-07, "loss": 0.188, "step": 3806 }, { "epoch": 1.0130388504523682, "grad_norm": 0.28428325057029724, "learning_rate": 1.8336079286139342e-07, "loss": 0.2154, "step": 3807 }, { "epoch": 1.013304949441192, "grad_norm": 0.3076273798942566, "learning_rate": 1.8335136795181055e-07, "loss": 0.2167, "step": 3808 }, { "epoch": 1.013571048430016, "grad_norm": 0.29209256172180176, "learning_rate": 1.8334194061608576e-07, "loss": 0.2198, "step": 3809 }, { "epoch": 1.01383714741884, "grad_norm": 0.2724366784095764, "learning_rate": 1.8333251085449347e-07, "loss": 0.2055, "step": 3810 }, { "epoch": 1.0141032464076636, "grad_norm": 0.2790094316005707, "learning_rate": 1.8332307866730812e-07, "loss": 0.221, "step": 3811 }, { "epoch": 1.0143693453964875, "grad_norm": 0.28840968012809753, "learning_rate": 1.8331364405480433e-07, "loss": 0.2333, "step": 3812 }, { "epoch": 1.0146354443853114, "grad_norm": 0.26599717140197754, "learning_rate": 1.8330420701725668e-07, "loss": 0.2127, "step": 3813 }, { "epoch": 1.014901543374135, "grad_norm": 0.2933604419231415, "learning_rate": 1.8329476755493986e-07, "loss": 0.2544, "step": 3814 }, { "epoch": 1.015167642362959, "grad_norm": 0.28055235743522644, "learning_rate": 1.8328532566812863e-07, "loss": 0.2196, "step": 3815 }, { "epoch": 1.0154337413517829, "grad_norm": 0.2911219894886017, "learning_rate": 1.832758813570978e-07, "loss": 0.2363, "step": 3816 }, { "epoch": 1.0156998403406068, "grad_norm": 0.2787172794342041, "learning_rate": 1.8326643462212233e-07, "loss": 0.2015, "step": 3817 }, { "epoch": 1.0159659393294305, "grad_norm": 0.2844879925251007, "learning_rate": 1.8325698546347712e-07, "loss": 0.2217, "step": 3818 }, { "epoch": 1.0162320383182544, "grad_norm": 0.28175926208496094, "learning_rate": 1.8324753388143726e-07, "loss": 0.2314, "step": 3819 }, { "epoch": 1.0164981373070783, "grad_norm": 0.2670617699623108, "learning_rate": 1.8323807987627783e-07, "loss": 0.1962, "step": 3820 }, { "epoch": 1.0167642362959022, "grad_norm": 0.293792188167572, "learning_rate": 1.8322862344827402e-07, "loss": 0.2374, "step": 3821 }, { "epoch": 1.0170303352847259, "grad_norm": 0.26472124457359314, "learning_rate": 1.8321916459770102e-07, "loss": 0.2199, "step": 3822 }, { "epoch": 1.0172964342735498, "grad_norm": 0.26971179246902466, "learning_rate": 1.832097033248343e-07, "loss": 0.2059, "step": 3823 }, { "epoch": 1.0175625332623737, "grad_norm": 0.266078382730484, "learning_rate": 1.8320023962994912e-07, "loss": 0.2175, "step": 3824 }, { "epoch": 1.0178286322511974, "grad_norm": 0.28880441188812256, "learning_rate": 1.83190773513321e-07, "loss": 0.2218, "step": 3825 }, { "epoch": 1.0180947312400213, "grad_norm": 0.3039831221103668, "learning_rate": 1.831813049752255e-07, "loss": 0.2142, "step": 3826 }, { "epoch": 1.0183608302288452, "grad_norm": 0.32403868436813354, "learning_rate": 1.8317183401593814e-07, "loss": 0.2348, "step": 3827 }, { "epoch": 1.018626929217669, "grad_norm": 0.29238542914390564, "learning_rate": 1.831623606357347e-07, "loss": 0.2163, "step": 3828 }, { "epoch": 1.0188930282064927, "grad_norm": 0.25253745913505554, "learning_rate": 1.831528848348908e-07, "loss": 0.2102, "step": 3829 }, { "epoch": 1.0191591271953166, "grad_norm": 0.3093230724334717, "learning_rate": 1.831434066136824e-07, "loss": 0.2148, "step": 3830 }, { "epoch": 1.0194252261841406, "grad_norm": 0.2806376814842224, "learning_rate": 1.831339259723853e-07, "loss": 0.2033, "step": 3831 }, { "epoch": 1.0196913251729642, "grad_norm": 0.2637574374675751, "learning_rate": 1.8312444291127543e-07, "loss": 0.1959, "step": 3832 }, { "epoch": 1.0199574241617881, "grad_norm": 0.26753106713294983, "learning_rate": 1.8311495743062886e-07, "loss": 0.2242, "step": 3833 }, { "epoch": 1.020223523150612, "grad_norm": 0.2768945097923279, "learning_rate": 1.831054695307217e-07, "loss": 0.2041, "step": 3834 }, { "epoch": 1.020489622139436, "grad_norm": 0.29763197898864746, "learning_rate": 1.8309597921183012e-07, "loss": 0.2282, "step": 3835 }, { "epoch": 1.0207557211282596, "grad_norm": 0.2781069874763489, "learning_rate": 1.8308648647423031e-07, "loss": 0.2206, "step": 3836 }, { "epoch": 1.0210218201170835, "grad_norm": 0.2799726128578186, "learning_rate": 1.8307699131819866e-07, "loss": 0.2166, "step": 3837 }, { "epoch": 1.0212879191059074, "grad_norm": 0.27932870388031006, "learning_rate": 1.8306749374401145e-07, "loss": 0.2358, "step": 3838 }, { "epoch": 1.0215540180947313, "grad_norm": 0.31798243522644043, "learning_rate": 1.8305799375194519e-07, "loss": 0.2366, "step": 3839 }, { "epoch": 1.021820117083555, "grad_norm": 0.3104202449321747, "learning_rate": 1.830484913422764e-07, "loss": 0.2226, "step": 3840 }, { "epoch": 1.022086216072379, "grad_norm": 0.2854885458946228, "learning_rate": 1.8303898651528167e-07, "loss": 0.2134, "step": 3841 }, { "epoch": 1.0223523150612028, "grad_norm": 0.2844628393650055, "learning_rate": 1.8302947927123764e-07, "loss": 0.2089, "step": 3842 }, { "epoch": 1.0226184140500265, "grad_norm": 0.276161789894104, "learning_rate": 1.8301996961042104e-07, "loss": 0.2197, "step": 3843 }, { "epoch": 1.0228845130388504, "grad_norm": 0.28231915831565857, "learning_rate": 1.830104575331087e-07, "loss": 0.2135, "step": 3844 }, { "epoch": 1.0231506120276743, "grad_norm": 0.2650793492794037, "learning_rate": 1.8300094303957745e-07, "loss": 0.2024, "step": 3845 }, { "epoch": 1.0234167110164982, "grad_norm": 0.2870996296405792, "learning_rate": 1.829914261301043e-07, "loss": 0.2247, "step": 3846 }, { "epoch": 1.023682810005322, "grad_norm": 0.2840869128704071, "learning_rate": 1.829819068049662e-07, "loss": 0.206, "step": 3847 }, { "epoch": 1.0239489089941458, "grad_norm": 0.3050197958946228, "learning_rate": 1.8297238506444023e-07, "loss": 0.2343, "step": 3848 }, { "epoch": 1.0242150079829697, "grad_norm": 0.29638561606407166, "learning_rate": 1.8296286090880358e-07, "loss": 0.2272, "step": 3849 }, { "epoch": 1.0244811069717934, "grad_norm": 0.2876259386539459, "learning_rate": 1.829533343383335e-07, "loss": 0.234, "step": 3850 }, { "epoch": 1.0247472059606173, "grad_norm": 0.2912015914916992, "learning_rate": 1.829438053533072e-07, "loss": 0.2325, "step": 3851 }, { "epoch": 1.0250133049494412, "grad_norm": 0.27886703610420227, "learning_rate": 1.829342739540021e-07, "loss": 0.2003, "step": 3852 }, { "epoch": 1.025279403938265, "grad_norm": 0.28646060824394226, "learning_rate": 1.829247401406956e-07, "loss": 0.2271, "step": 3853 }, { "epoch": 1.0255455029270888, "grad_norm": 0.2855707108974457, "learning_rate": 1.8291520391366526e-07, "loss": 0.2088, "step": 3854 }, { "epoch": 1.0258116019159127, "grad_norm": 0.26935702562332153, "learning_rate": 1.8290566527318863e-07, "loss": 0.2175, "step": 3855 }, { "epoch": 1.0260777009047366, "grad_norm": 0.28688544034957886, "learning_rate": 1.828961242195433e-07, "loss": 0.212, "step": 3856 }, { "epoch": 1.0263437998935605, "grad_norm": 0.2808331549167633, "learning_rate": 1.8288658075300708e-07, "loss": 0.2124, "step": 3857 }, { "epoch": 1.0266098988823842, "grad_norm": 0.2824205458164215, "learning_rate": 1.8287703487385765e-07, "loss": 0.2349, "step": 3858 }, { "epoch": 1.026875997871208, "grad_norm": 0.2879166007041931, "learning_rate": 1.8286748658237298e-07, "loss": 0.2303, "step": 3859 }, { "epoch": 1.027142096860032, "grad_norm": 0.26450514793395996, "learning_rate": 1.8285793587883092e-07, "loss": 0.2104, "step": 3860 }, { "epoch": 1.0274081958488557, "grad_norm": 0.2826872766017914, "learning_rate": 1.828483827635095e-07, "loss": 0.2329, "step": 3861 }, { "epoch": 1.0276742948376796, "grad_norm": 0.3029928207397461, "learning_rate": 1.8283882723668673e-07, "loss": 0.2497, "step": 3862 }, { "epoch": 1.0279403938265035, "grad_norm": 0.25724083185195923, "learning_rate": 1.828292692986408e-07, "loss": 0.2065, "step": 3863 }, { "epoch": 1.0282064928153274, "grad_norm": 0.3079726994037628, "learning_rate": 1.8281970894964992e-07, "loss": 0.2486, "step": 3864 }, { "epoch": 1.028472591804151, "grad_norm": 0.2788484990596771, "learning_rate": 1.8281014618999233e-07, "loss": 0.2317, "step": 3865 }, { "epoch": 1.028738690792975, "grad_norm": 0.2672310769557953, "learning_rate": 1.8280058101994645e-07, "loss": 0.1949, "step": 3866 }, { "epoch": 1.0290047897817989, "grad_norm": 0.26648420095443726, "learning_rate": 1.827910134397906e-07, "loss": 0.2228, "step": 3867 }, { "epoch": 1.0292708887706228, "grad_norm": 0.3060806095600128, "learning_rate": 1.8278144344980331e-07, "loss": 0.2219, "step": 3868 }, { "epoch": 1.0295369877594465, "grad_norm": 0.2991519868373871, "learning_rate": 1.8277187105026317e-07, "loss": 0.2301, "step": 3869 }, { "epoch": 1.0298030867482704, "grad_norm": 0.26691487431526184, "learning_rate": 1.8276229624144877e-07, "loss": 0.2235, "step": 3870 }, { "epoch": 1.0300691857370943, "grad_norm": 0.3066096603870392, "learning_rate": 1.827527190236388e-07, "loss": 0.2388, "step": 3871 }, { "epoch": 1.030335284725918, "grad_norm": 0.3055647909641266, "learning_rate": 1.8274313939711206e-07, "loss": 0.2297, "step": 3872 }, { "epoch": 1.0306013837147419, "grad_norm": 0.2655828595161438, "learning_rate": 1.8273355736214736e-07, "loss": 0.2174, "step": 3873 }, { "epoch": 1.0308674827035658, "grad_norm": 0.2976057827472687, "learning_rate": 1.8272397291902362e-07, "loss": 0.2128, "step": 3874 }, { "epoch": 1.0311335816923897, "grad_norm": 0.30208370089530945, "learning_rate": 1.8271438606801984e-07, "loss": 0.2343, "step": 3875 }, { "epoch": 1.0313996806812133, "grad_norm": 0.30696141719818115, "learning_rate": 1.8270479680941504e-07, "loss": 0.2361, "step": 3876 }, { "epoch": 1.0316657796700373, "grad_norm": 0.2609519362449646, "learning_rate": 1.8269520514348832e-07, "loss": 0.1946, "step": 3877 }, { "epoch": 1.0319318786588612, "grad_norm": 0.28399795293807983, "learning_rate": 1.8268561107051892e-07, "loss": 0.2154, "step": 3878 }, { "epoch": 1.0321979776476848, "grad_norm": 0.2746931314468384, "learning_rate": 1.8267601459078606e-07, "loss": 0.2302, "step": 3879 }, { "epoch": 1.0324640766365087, "grad_norm": 0.27120140194892883, "learning_rate": 1.826664157045691e-07, "loss": 0.2068, "step": 3880 }, { "epoch": 1.0327301756253326, "grad_norm": 0.2668604850769043, "learning_rate": 1.8265681441214742e-07, "loss": 0.2015, "step": 3881 }, { "epoch": 1.0329962746141566, "grad_norm": 0.27099403738975525, "learning_rate": 1.826472107138005e-07, "loss": 0.1976, "step": 3882 }, { "epoch": 1.0332623736029802, "grad_norm": 0.27182385325431824, "learning_rate": 1.826376046098078e-07, "loss": 0.2218, "step": 3883 }, { "epoch": 1.0335284725918041, "grad_norm": 0.24993163347244263, "learning_rate": 1.8262799610044905e-07, "loss": 0.2043, "step": 3884 }, { "epoch": 1.033794571580628, "grad_norm": 0.3369397222995758, "learning_rate": 1.8261838518600388e-07, "loss": 0.2441, "step": 3885 }, { "epoch": 1.0340606705694517, "grad_norm": 0.2532680928707123, "learning_rate": 1.8260877186675204e-07, "loss": 0.2081, "step": 3886 }, { "epoch": 1.0343267695582756, "grad_norm": 0.3123417794704437, "learning_rate": 1.825991561429733e-07, "loss": 0.2393, "step": 3887 }, { "epoch": 1.0345928685470995, "grad_norm": 0.2674771547317505, "learning_rate": 1.825895380149476e-07, "loss": 0.2029, "step": 3888 }, { "epoch": 1.0348589675359234, "grad_norm": 0.2578444480895996, "learning_rate": 1.8257991748295492e-07, "loss": 0.1957, "step": 3889 }, { "epoch": 1.0351250665247471, "grad_norm": 0.3041520118713379, "learning_rate": 1.8257029454727525e-07, "loss": 0.2216, "step": 3890 }, { "epoch": 1.035391165513571, "grad_norm": 0.29374104738235474, "learning_rate": 1.825606692081887e-07, "loss": 0.2282, "step": 3891 }, { "epoch": 1.035657264502395, "grad_norm": 0.2573068141937256, "learning_rate": 1.825510414659754e-07, "loss": 0.2029, "step": 3892 }, { "epoch": 1.0359233634912188, "grad_norm": 0.2890547513961792, "learning_rate": 1.8254141132091564e-07, "loss": 0.2051, "step": 3893 }, { "epoch": 1.0361894624800425, "grad_norm": 0.33899012207984924, "learning_rate": 1.8253177877328973e-07, "loss": 0.2263, "step": 3894 }, { "epoch": 1.0364555614688664, "grad_norm": 0.29161328077316284, "learning_rate": 1.8252214382337803e-07, "loss": 0.2137, "step": 3895 }, { "epoch": 1.0367216604576903, "grad_norm": 0.24684350192546844, "learning_rate": 1.82512506471461e-07, "loss": 0.1972, "step": 3896 }, { "epoch": 1.036987759446514, "grad_norm": 0.27546149492263794, "learning_rate": 1.8250286671781915e-07, "loss": 0.2164, "step": 3897 }, { "epoch": 1.037253858435338, "grad_norm": 0.2619822323322296, "learning_rate": 1.8249322456273305e-07, "loss": 0.2059, "step": 3898 }, { "epoch": 1.0375199574241618, "grad_norm": 0.2802460491657257, "learning_rate": 1.8248358000648334e-07, "loss": 0.2339, "step": 3899 }, { "epoch": 1.0377860564129857, "grad_norm": 0.28383609652519226, "learning_rate": 1.824739330493508e-07, "loss": 0.2223, "step": 3900 }, { "epoch": 1.0380521554018094, "grad_norm": 0.26953181624412537, "learning_rate": 1.8246428369161625e-07, "loss": 0.2096, "step": 3901 }, { "epoch": 1.0383182543906333, "grad_norm": 0.32326143980026245, "learning_rate": 1.824546319335605e-07, "loss": 0.2154, "step": 3902 }, { "epoch": 1.0385843533794572, "grad_norm": 0.292358934879303, "learning_rate": 1.824449777754645e-07, "loss": 0.2073, "step": 3903 }, { "epoch": 1.038850452368281, "grad_norm": 0.27949872612953186, "learning_rate": 1.8243532121760925e-07, "loss": 0.2386, "step": 3904 }, { "epoch": 1.0391165513571048, "grad_norm": 0.27411729097366333, "learning_rate": 1.8242566226027587e-07, "loss": 0.2159, "step": 3905 }, { "epoch": 1.0393826503459287, "grad_norm": 0.3033146858215332, "learning_rate": 1.8241600090374547e-07, "loss": 0.2215, "step": 3906 }, { "epoch": 1.0396487493347526, "grad_norm": 0.28947457671165466, "learning_rate": 1.8240633714829925e-07, "loss": 0.2252, "step": 3907 }, { "epoch": 1.0399148483235763, "grad_norm": 0.2922924757003784, "learning_rate": 1.8239667099421854e-07, "loss": 0.2367, "step": 3908 }, { "epoch": 1.0401809473124002, "grad_norm": 0.26315224170684814, "learning_rate": 1.8238700244178466e-07, "loss": 0.2033, "step": 3909 }, { "epoch": 1.040447046301224, "grad_norm": 0.297373503446579, "learning_rate": 1.8237733149127909e-07, "loss": 0.2462, "step": 3910 }, { "epoch": 1.040713145290048, "grad_norm": 0.27339547872543335, "learning_rate": 1.8236765814298328e-07, "loss": 0.2096, "step": 3911 }, { "epoch": 1.0409792442788717, "grad_norm": 0.32204288244247437, "learning_rate": 1.823579823971788e-07, "loss": 0.2308, "step": 3912 }, { "epoch": 1.0412453432676956, "grad_norm": 0.2952735722064972, "learning_rate": 1.823483042541473e-07, "loss": 0.2276, "step": 3913 }, { "epoch": 1.0415114422565195, "grad_norm": 0.27479660511016846, "learning_rate": 1.8233862371417047e-07, "loss": 0.2082, "step": 3914 }, { "epoch": 1.0417775412453432, "grad_norm": 0.2721025049686432, "learning_rate": 1.8232894077753007e-07, "loss": 0.2244, "step": 3915 }, { "epoch": 1.042043640234167, "grad_norm": 0.2923494279384613, "learning_rate": 1.82319255444508e-07, "loss": 0.2255, "step": 3916 }, { "epoch": 1.042309739222991, "grad_norm": 0.2951546609401703, "learning_rate": 1.8230956771538612e-07, "loss": 0.2168, "step": 3917 }, { "epoch": 1.0425758382118149, "grad_norm": 0.29284748435020447, "learning_rate": 1.8229987759044646e-07, "loss": 0.2221, "step": 3918 }, { "epoch": 1.0428419372006386, "grad_norm": 0.28520745038986206, "learning_rate": 1.8229018506997105e-07, "loss": 0.2209, "step": 3919 }, { "epoch": 1.0431080361894625, "grad_norm": 0.2646285891532898, "learning_rate": 1.8228049015424202e-07, "loss": 0.2165, "step": 3920 }, { "epoch": 1.0433741351782864, "grad_norm": 0.3042415380477905, "learning_rate": 1.8227079284354155e-07, "loss": 0.2291, "step": 3921 }, { "epoch": 1.0436402341671103, "grad_norm": 0.28489720821380615, "learning_rate": 1.8226109313815193e-07, "loss": 0.2151, "step": 3922 }, { "epoch": 1.043906333155934, "grad_norm": 0.28365013003349304, "learning_rate": 1.8225139103835545e-07, "loss": 0.2126, "step": 3923 }, { "epoch": 1.0441724321447579, "grad_norm": 0.2864338159561157, "learning_rate": 1.8224168654443458e-07, "loss": 0.2165, "step": 3924 }, { "epoch": 1.0444385311335818, "grad_norm": 0.26674869656562805, "learning_rate": 1.8223197965667173e-07, "loss": 0.2217, "step": 3925 }, { "epoch": 1.0447046301224054, "grad_norm": 0.290524423122406, "learning_rate": 1.8222227037534944e-07, "loss": 0.2238, "step": 3926 }, { "epoch": 1.0449707291112293, "grad_norm": 0.28312182426452637, "learning_rate": 1.8221255870075041e-07, "loss": 0.2205, "step": 3927 }, { "epoch": 1.0452368281000533, "grad_norm": 0.27064284682273865, "learning_rate": 1.822028446331572e-07, "loss": 0.2091, "step": 3928 }, { "epoch": 1.0455029270888772, "grad_norm": 0.27713581919670105, "learning_rate": 1.8219312817285265e-07, "loss": 0.2092, "step": 3929 }, { "epoch": 1.0457690260777008, "grad_norm": 0.27955108880996704, "learning_rate": 1.8218340932011954e-07, "loss": 0.1927, "step": 3930 }, { "epoch": 1.0460351250665247, "grad_norm": 0.2646961510181427, "learning_rate": 1.8217368807524078e-07, "loss": 0.2076, "step": 3931 }, { "epoch": 1.0463012240553486, "grad_norm": 0.2863714098930359, "learning_rate": 1.821639644384993e-07, "loss": 0.2151, "step": 3932 }, { "epoch": 1.0465673230441723, "grad_norm": 0.3123281002044678, "learning_rate": 1.821542384101782e-07, "loss": 0.2224, "step": 3933 }, { "epoch": 1.0468334220329962, "grad_norm": 0.27204373478889465, "learning_rate": 1.8214450999056047e-07, "loss": 0.208, "step": 3934 }, { "epoch": 1.0470995210218201, "grad_norm": 0.3084838390350342, "learning_rate": 1.821347791799294e-07, "loss": 0.2224, "step": 3935 }, { "epoch": 1.047365620010644, "grad_norm": 0.2701469957828522, "learning_rate": 1.8212504597856811e-07, "loss": 0.202, "step": 3936 }, { "epoch": 1.0476317189994677, "grad_norm": 0.30724233388900757, "learning_rate": 1.8211531038676e-07, "loss": 0.2425, "step": 3937 }, { "epoch": 1.0478978179882916, "grad_norm": 0.26847121119499207, "learning_rate": 1.821055724047884e-07, "loss": 0.2124, "step": 3938 }, { "epoch": 1.0481639169771155, "grad_norm": 0.30466943979263306, "learning_rate": 1.820958320329368e-07, "loss": 0.2177, "step": 3939 }, { "epoch": 1.0484300159659394, "grad_norm": 0.2890094220638275, "learning_rate": 1.8208608927148868e-07, "loss": 0.208, "step": 3940 }, { "epoch": 1.0486961149547631, "grad_norm": 0.2877458930015564, "learning_rate": 1.8207634412072763e-07, "loss": 0.2234, "step": 3941 }, { "epoch": 1.048962213943587, "grad_norm": 0.2853900194168091, "learning_rate": 1.820665965809373e-07, "loss": 0.2282, "step": 3942 }, { "epoch": 1.049228312932411, "grad_norm": 0.26054689288139343, "learning_rate": 1.8205684665240148e-07, "loss": 0.2121, "step": 3943 }, { "epoch": 1.0494944119212346, "grad_norm": 0.27698609232902527, "learning_rate": 1.8204709433540383e-07, "loss": 0.1986, "step": 3944 }, { "epoch": 1.0497605109100585, "grad_norm": 0.25898927450180054, "learning_rate": 1.8203733963022834e-07, "loss": 0.1951, "step": 3945 }, { "epoch": 1.0500266098988824, "grad_norm": 0.2778167426586151, "learning_rate": 1.8202758253715887e-07, "loss": 0.2151, "step": 3946 }, { "epoch": 1.0502927088877063, "grad_norm": 0.3291415870189667, "learning_rate": 1.820178230564795e-07, "loss": 0.2232, "step": 3947 }, { "epoch": 1.05055880787653, "grad_norm": 0.29695209860801697, "learning_rate": 1.8200806118847426e-07, "loss": 0.2148, "step": 3948 }, { "epoch": 1.050824906865354, "grad_norm": 0.2599940598011017, "learning_rate": 1.8199829693342727e-07, "loss": 0.2126, "step": 3949 }, { "epoch": 1.0510910058541778, "grad_norm": 0.2946036458015442, "learning_rate": 1.8198853029162277e-07, "loss": 0.2192, "step": 3950 }, { "epoch": 1.0513571048430017, "grad_norm": 0.24824561178684235, "learning_rate": 1.81978761263345e-07, "loss": 0.1889, "step": 3951 }, { "epoch": 1.0516232038318254, "grad_norm": 0.2926669418811798, "learning_rate": 1.819689898488784e-07, "loss": 0.217, "step": 3952 }, { "epoch": 1.0518893028206493, "grad_norm": 0.30631399154663086, "learning_rate": 1.819592160485073e-07, "loss": 0.2158, "step": 3953 }, { "epoch": 1.0521554018094732, "grad_norm": 0.30573898553848267, "learning_rate": 1.8194943986251622e-07, "loss": 0.2423, "step": 3954 }, { "epoch": 1.0524215007982969, "grad_norm": 0.25552883744239807, "learning_rate": 1.8193966129118977e-07, "loss": 0.2001, "step": 3955 }, { "epoch": 1.0526875997871208, "grad_norm": 0.30721384286880493, "learning_rate": 1.819298803348125e-07, "loss": 0.2247, "step": 3956 }, { "epoch": 1.0529536987759447, "grad_norm": 0.27849316596984863, "learning_rate": 1.8192009699366915e-07, "loss": 0.2166, "step": 3957 }, { "epoch": 1.0532197977647686, "grad_norm": 0.2816278636455536, "learning_rate": 1.8191031126804446e-07, "loss": 0.2053, "step": 3958 }, { "epoch": 1.0534858967535923, "grad_norm": 0.29826077818870544, "learning_rate": 1.8190052315822332e-07, "loss": 0.2235, "step": 3959 }, { "epoch": 1.0537519957424162, "grad_norm": 0.2796823978424072, "learning_rate": 1.8189073266449058e-07, "loss": 0.2078, "step": 3960 }, { "epoch": 1.05401809473124, "grad_norm": 0.282967746257782, "learning_rate": 1.8188093978713126e-07, "loss": 0.2078, "step": 3961 }, { "epoch": 1.0542841937200638, "grad_norm": 0.28865137696266174, "learning_rate": 1.8187114452643036e-07, "loss": 0.2495, "step": 3962 }, { "epoch": 1.0545502927088877, "grad_norm": 0.2662108242511749, "learning_rate": 1.8186134688267302e-07, "loss": 0.2107, "step": 3963 }, { "epoch": 1.0548163916977116, "grad_norm": 0.2824023962020874, "learning_rate": 1.8185154685614443e-07, "loss": 0.2267, "step": 3964 }, { "epoch": 1.0550824906865355, "grad_norm": 0.2794705927371979, "learning_rate": 1.8184174444712987e-07, "loss": 0.2156, "step": 3965 }, { "epoch": 1.0553485896753592, "grad_norm": 0.30647560954093933, "learning_rate": 1.8183193965591457e-07, "loss": 0.2213, "step": 3966 }, { "epoch": 1.055614688664183, "grad_norm": 0.282723605632782, "learning_rate": 1.8182213248278403e-07, "loss": 0.2347, "step": 3967 }, { "epoch": 1.055880787653007, "grad_norm": 0.2870355546474457, "learning_rate": 1.8181232292802363e-07, "loss": 0.2071, "step": 3968 }, { "epoch": 1.0561468866418307, "grad_norm": 0.2969743013381958, "learning_rate": 1.8180251099191896e-07, "loss": 0.2235, "step": 3969 }, { "epoch": 1.0564129856306546, "grad_norm": 0.2910982668399811, "learning_rate": 1.8179269667475555e-07, "loss": 0.2189, "step": 3970 }, { "epoch": 1.0566790846194785, "grad_norm": 0.30864134430885315, "learning_rate": 1.8178287997681917e-07, "loss": 0.2386, "step": 3971 }, { "epoch": 1.0569451836083024, "grad_norm": 0.301664799451828, "learning_rate": 1.817730608983955e-07, "loss": 0.2237, "step": 3972 }, { "epoch": 1.057211282597126, "grad_norm": 0.27226099371910095, "learning_rate": 1.817632394397703e-07, "loss": 0.2129, "step": 3973 }, { "epoch": 1.05747738158595, "grad_norm": 0.29465702176094055, "learning_rate": 1.817534156012295e-07, "loss": 0.224, "step": 3974 }, { "epoch": 1.0577434805747739, "grad_norm": 0.2602803707122803, "learning_rate": 1.8174358938305904e-07, "loss": 0.2173, "step": 3975 }, { "epoch": 1.0580095795635978, "grad_norm": 0.3033916652202606, "learning_rate": 1.8173376078554495e-07, "loss": 0.2324, "step": 3976 }, { "epoch": 1.0582756785524214, "grad_norm": 0.28491389751434326, "learning_rate": 1.8172392980897335e-07, "loss": 0.2306, "step": 3977 }, { "epoch": 1.0585417775412453, "grad_norm": 0.32759061455726624, "learning_rate": 1.817140964536303e-07, "loss": 0.2367, "step": 3978 }, { "epoch": 1.0588078765300692, "grad_norm": 0.2648443579673767, "learning_rate": 1.817042607198021e-07, "loss": 0.2266, "step": 3979 }, { "epoch": 1.059073975518893, "grad_norm": 0.29471710324287415, "learning_rate": 1.8169442260777497e-07, "loss": 0.2243, "step": 3980 }, { "epoch": 1.0593400745077168, "grad_norm": 0.2940070629119873, "learning_rate": 1.8168458211783535e-07, "loss": 0.2215, "step": 3981 }, { "epoch": 1.0596061734965407, "grad_norm": 0.3114984333515167, "learning_rate": 1.8167473925026962e-07, "loss": 0.2229, "step": 3982 }, { "epoch": 1.0598722724853646, "grad_norm": 0.31122273206710815, "learning_rate": 1.816648940053643e-07, "loss": 0.2137, "step": 3983 }, { "epoch": 1.0601383714741883, "grad_norm": 0.27216556668281555, "learning_rate": 1.8165504638340595e-07, "loss": 0.2142, "step": 3984 }, { "epoch": 1.0604044704630122, "grad_norm": 0.3273068070411682, "learning_rate": 1.8164519638468124e-07, "loss": 0.2271, "step": 3985 }, { "epoch": 1.0606705694518361, "grad_norm": 0.26343703269958496, "learning_rate": 1.8163534400947682e-07, "loss": 0.2256, "step": 3986 }, { "epoch": 1.06093666844066, "grad_norm": 0.27404719591140747, "learning_rate": 1.8162548925807952e-07, "loss": 0.228, "step": 3987 }, { "epoch": 1.0612027674294837, "grad_norm": 0.2768858075141907, "learning_rate": 1.8161563213077616e-07, "loss": 0.2193, "step": 3988 }, { "epoch": 1.0614688664183076, "grad_norm": 0.2647273540496826, "learning_rate": 1.8160577262785365e-07, "loss": 0.1998, "step": 3989 }, { "epoch": 1.0617349654071315, "grad_norm": 0.27463993430137634, "learning_rate": 1.81595910749599e-07, "loss": 0.2121, "step": 3990 }, { "epoch": 1.0620010643959552, "grad_norm": 0.2520335912704468, "learning_rate": 1.8158604649629924e-07, "loss": 0.1914, "step": 3991 }, { "epoch": 1.0622671633847791, "grad_norm": 0.27082690596580505, "learning_rate": 1.8157617986824154e-07, "loss": 0.229, "step": 3992 }, { "epoch": 1.062533262373603, "grad_norm": 0.2775748074054718, "learning_rate": 1.81566310865713e-07, "loss": 0.2254, "step": 3993 }, { "epoch": 1.062799361362427, "grad_norm": 0.26986804604530334, "learning_rate": 1.8155643948900098e-07, "loss": 0.2204, "step": 3994 }, { "epoch": 1.0630654603512506, "grad_norm": 0.27089813351631165, "learning_rate": 1.8154656573839273e-07, "loss": 0.2173, "step": 3995 }, { "epoch": 1.0633315593400745, "grad_norm": 0.2932595908641815, "learning_rate": 1.8153668961417573e-07, "loss": 0.2045, "step": 3996 }, { "epoch": 1.0635976583288984, "grad_norm": 0.2917627990245819, "learning_rate": 1.8152681111663735e-07, "loss": 0.2198, "step": 3997 }, { "epoch": 1.063863757317722, "grad_norm": 0.3115244209766388, "learning_rate": 1.815169302460652e-07, "loss": 0.2233, "step": 3998 }, { "epoch": 1.064129856306546, "grad_norm": 0.28028514981269836, "learning_rate": 1.8150704700274687e-07, "loss": 0.1951, "step": 3999 }, { "epoch": 1.06439595529537, "grad_norm": 0.2848566174507141, "learning_rate": 1.8149716138697007e-07, "loss": 0.2214, "step": 4000 }, { "epoch": 1.0646620542841938, "grad_norm": 0.2814963161945343, "learning_rate": 1.8148727339902242e-07, "loss": 0.204, "step": 4001 }, { "epoch": 1.0649281532730175, "grad_norm": 0.29532092809677124, "learning_rate": 1.814773830391919e-07, "loss": 0.2358, "step": 4002 }, { "epoch": 1.0651942522618414, "grad_norm": 0.2971549332141876, "learning_rate": 1.814674903077663e-07, "loss": 0.2324, "step": 4003 }, { "epoch": 1.0654603512506653, "grad_norm": 0.31810274720191956, "learning_rate": 1.8145759520503356e-07, "loss": 0.234, "step": 4004 }, { "epoch": 1.065726450239489, "grad_norm": 0.2776760756969452, "learning_rate": 1.8144769773128175e-07, "loss": 0.2207, "step": 4005 }, { "epoch": 1.0659925492283129, "grad_norm": 0.3177986741065979, "learning_rate": 1.814377978867989e-07, "loss": 0.2153, "step": 4006 }, { "epoch": 1.0662586482171368, "grad_norm": 0.3040836453437805, "learning_rate": 1.8142789567187325e-07, "loss": 0.2294, "step": 4007 }, { "epoch": 1.0665247472059607, "grad_norm": 0.2676386535167694, "learning_rate": 1.8141799108679296e-07, "loss": 0.215, "step": 4008 }, { "epoch": 1.0667908461947844, "grad_norm": 0.2622308135032654, "learning_rate": 1.8140808413184634e-07, "loss": 0.1834, "step": 4009 }, { "epoch": 1.0670569451836083, "grad_norm": 0.28501495718955994, "learning_rate": 1.8139817480732178e-07, "loss": 0.2219, "step": 4010 }, { "epoch": 1.0673230441724322, "grad_norm": 0.2717069983482361, "learning_rate": 1.813882631135077e-07, "loss": 0.2197, "step": 4011 }, { "epoch": 1.067589143161256, "grad_norm": 0.3844045102596283, "learning_rate": 1.813783490506926e-07, "loss": 0.2283, "step": 4012 }, { "epoch": 1.0678552421500798, "grad_norm": 0.31515559554100037, "learning_rate": 1.8136843261916506e-07, "loss": 0.2317, "step": 4013 }, { "epoch": 1.0681213411389037, "grad_norm": 0.2938205301761627, "learning_rate": 1.813585138192137e-07, "loss": 0.2274, "step": 4014 }, { "epoch": 1.0683874401277276, "grad_norm": 0.2951050102710724, "learning_rate": 1.8134859265112724e-07, "loss": 0.2198, "step": 4015 }, { "epoch": 1.0686535391165513, "grad_norm": 0.3078429102897644, "learning_rate": 1.8133866911519452e-07, "loss": 0.2093, "step": 4016 }, { "epoch": 1.0689196381053752, "grad_norm": 0.29873305559158325, "learning_rate": 1.8132874321170428e-07, "loss": 0.228, "step": 4017 }, { "epoch": 1.069185737094199, "grad_norm": 0.26747944951057434, "learning_rate": 1.8131881494094553e-07, "loss": 0.2113, "step": 4018 }, { "epoch": 1.069451836083023, "grad_norm": 0.29404208064079285, "learning_rate": 1.813088843032072e-07, "loss": 0.22, "step": 4019 }, { "epoch": 1.0697179350718466, "grad_norm": 0.279108464717865, "learning_rate": 1.8129895129877837e-07, "loss": 0.2105, "step": 4020 }, { "epoch": 1.0699840340606706, "grad_norm": 0.277398943901062, "learning_rate": 1.8128901592794813e-07, "loss": 0.2051, "step": 4021 }, { "epoch": 1.0702501330494945, "grad_norm": 0.293218195438385, "learning_rate": 1.8127907819100575e-07, "loss": 0.2168, "step": 4022 }, { "epoch": 1.0705162320383184, "grad_norm": 0.28484970331192017, "learning_rate": 1.8126913808824043e-07, "loss": 0.2112, "step": 4023 }, { "epoch": 1.070782331027142, "grad_norm": 0.2670040428638458, "learning_rate": 1.812591956199415e-07, "loss": 0.2018, "step": 4024 }, { "epoch": 1.071048430015966, "grad_norm": 0.27911487221717834, "learning_rate": 1.8124925078639838e-07, "loss": 0.2214, "step": 4025 }, { "epoch": 1.0713145290047899, "grad_norm": 0.27769359946250916, "learning_rate": 1.8123930358790051e-07, "loss": 0.2039, "step": 4026 }, { "epoch": 1.0715806279936135, "grad_norm": 0.27658918499946594, "learning_rate": 1.812293540247375e-07, "loss": 0.2155, "step": 4027 }, { "epoch": 1.0718467269824374, "grad_norm": 0.29684171080589294, "learning_rate": 1.8121940209719886e-07, "loss": 0.2262, "step": 4028 }, { "epoch": 1.0721128259712613, "grad_norm": 0.28360626101493835, "learning_rate": 1.812094478055743e-07, "loss": 0.2443, "step": 4029 }, { "epoch": 1.0723789249600852, "grad_norm": 0.2579723298549652, "learning_rate": 1.8119949115015362e-07, "loss": 0.2109, "step": 4030 }, { "epoch": 1.072645023948909, "grad_norm": 0.2589172124862671, "learning_rate": 1.8118953213122655e-07, "loss": 0.2105, "step": 4031 }, { "epoch": 1.0729111229377328, "grad_norm": 0.2938130795955658, "learning_rate": 1.8117957074908302e-07, "loss": 0.225, "step": 4032 }, { "epoch": 1.0731772219265567, "grad_norm": 0.29001787304878235, "learning_rate": 1.8116960700401296e-07, "loss": 0.2265, "step": 4033 }, { "epoch": 1.0734433209153806, "grad_norm": 0.26490721106529236, "learning_rate": 1.811596408963064e-07, "loss": 0.2145, "step": 4034 }, { "epoch": 1.0737094199042043, "grad_norm": 0.2809385359287262, "learning_rate": 1.811496724262534e-07, "loss": 0.2174, "step": 4035 }, { "epoch": 1.0739755188930282, "grad_norm": 0.30110132694244385, "learning_rate": 1.811397015941442e-07, "loss": 0.2246, "step": 4036 }, { "epoch": 1.0742416178818521, "grad_norm": 0.26454368233680725, "learning_rate": 1.811297284002689e-07, "loss": 0.2068, "step": 4037 }, { "epoch": 1.0745077168706758, "grad_norm": 0.30363476276397705, "learning_rate": 1.8111975284491785e-07, "loss": 0.2123, "step": 4038 }, { "epoch": 1.0747738158594997, "grad_norm": 0.2580968141555786, "learning_rate": 1.8110977492838147e-07, "loss": 0.2093, "step": 4039 }, { "epoch": 1.0750399148483236, "grad_norm": 0.2935899794101715, "learning_rate": 1.810997946509501e-07, "loss": 0.2257, "step": 4040 }, { "epoch": 1.0753060138371473, "grad_norm": 0.30734193325042725, "learning_rate": 1.810898120129143e-07, "loss": 0.2253, "step": 4041 }, { "epoch": 1.0755721128259712, "grad_norm": 0.2868809998035431, "learning_rate": 1.8107982701456464e-07, "loss": 0.2225, "step": 4042 }, { "epoch": 1.075838211814795, "grad_norm": 0.28340429067611694, "learning_rate": 1.810698396561917e-07, "loss": 0.2262, "step": 4043 }, { "epoch": 1.076104310803619, "grad_norm": 0.27096882462501526, "learning_rate": 1.8105984993808625e-07, "loss": 0.2176, "step": 4044 }, { "epoch": 1.0763704097924427, "grad_norm": 0.29623982310295105, "learning_rate": 1.8104985786053903e-07, "loss": 0.2077, "step": 4045 }, { "epoch": 1.0766365087812666, "grad_norm": 0.2687319219112396, "learning_rate": 1.810398634238409e-07, "loss": 0.1964, "step": 4046 }, { "epoch": 1.0769026077700905, "grad_norm": 0.25436946749687195, "learning_rate": 1.8102986662828274e-07, "loss": 0.2072, "step": 4047 }, { "epoch": 1.0771687067589144, "grad_norm": 0.28678107261657715, "learning_rate": 1.8101986747415555e-07, "loss": 0.2154, "step": 4048 }, { "epoch": 1.077434805747738, "grad_norm": 0.27510878443717957, "learning_rate": 1.8100986596175042e-07, "loss": 0.2185, "step": 4049 }, { "epoch": 1.077700904736562, "grad_norm": 0.2664378583431244, "learning_rate": 1.8099986209135843e-07, "loss": 0.1936, "step": 4050 }, { "epoch": 1.077967003725386, "grad_norm": 0.28115659952163696, "learning_rate": 1.8098985586327077e-07, "loss": 0.2128, "step": 4051 }, { "epoch": 1.0782331027142096, "grad_norm": 0.27139151096343994, "learning_rate": 1.8097984727777863e-07, "loss": 0.191, "step": 4052 }, { "epoch": 1.0784992017030335, "grad_norm": 0.2958664000034332, "learning_rate": 1.8096983633517348e-07, "loss": 0.2355, "step": 4053 }, { "epoch": 1.0787653006918574, "grad_norm": 0.2876130938529968, "learning_rate": 1.8095982303574658e-07, "loss": 0.2252, "step": 4054 }, { "epoch": 1.0790313996806813, "grad_norm": 0.27714505791664124, "learning_rate": 1.8094980737978943e-07, "loss": 0.2186, "step": 4055 }, { "epoch": 1.079297498669505, "grad_norm": 0.2659110724925995, "learning_rate": 1.8093978936759362e-07, "loss": 0.2036, "step": 4056 }, { "epoch": 1.0795635976583289, "grad_norm": 0.29061946272850037, "learning_rate": 1.8092976899945067e-07, "loss": 0.2436, "step": 4057 }, { "epoch": 1.0798296966471528, "grad_norm": 0.24064823985099792, "learning_rate": 1.809197462756523e-07, "loss": 0.1916, "step": 4058 }, { "epoch": 1.0800957956359767, "grad_norm": 0.263480544090271, "learning_rate": 1.809097211964902e-07, "loss": 0.2033, "step": 4059 }, { "epoch": 1.0803618946248004, "grad_norm": 0.2949579358100891, "learning_rate": 1.8089969376225618e-07, "loss": 0.2309, "step": 4060 }, { "epoch": 1.0806279936136243, "grad_norm": 0.26769551634788513, "learning_rate": 1.8088966397324215e-07, "loss": 0.2018, "step": 4061 }, { "epoch": 1.0808940926024482, "grad_norm": 0.29946115612983704, "learning_rate": 1.8087963182974006e-07, "loss": 0.2164, "step": 4062 }, { "epoch": 1.0811601915912719, "grad_norm": 0.2778393030166626, "learning_rate": 1.8086959733204184e-07, "loss": 0.2105, "step": 4063 }, { "epoch": 1.0814262905800958, "grad_norm": 0.28379160165786743, "learning_rate": 1.8085956048043961e-07, "loss": 0.2302, "step": 4064 }, { "epoch": 1.0816923895689197, "grad_norm": 0.30809977650642395, "learning_rate": 1.8084952127522559e-07, "loss": 0.2283, "step": 4065 }, { "epoch": 1.0819584885577436, "grad_norm": 0.27503183484077454, "learning_rate": 1.808394797166919e-07, "loss": 0.2107, "step": 4066 }, { "epoch": 1.0822245875465673, "grad_norm": 0.2817705273628235, "learning_rate": 1.808294358051308e-07, "loss": 0.2078, "step": 4067 }, { "epoch": 1.0824906865353912, "grad_norm": 0.2835984528064728, "learning_rate": 1.8081938954083475e-07, "loss": 0.1897, "step": 4068 }, { "epoch": 1.082756785524215, "grad_norm": 0.26284900307655334, "learning_rate": 1.808093409240961e-07, "loss": 0.2111, "step": 4069 }, { "epoch": 1.083022884513039, "grad_norm": 0.26653367280960083, "learning_rate": 1.8079928995520736e-07, "loss": 0.212, "step": 4070 }, { "epoch": 1.0832889835018626, "grad_norm": 0.28825536370277405, "learning_rate": 1.8078923663446108e-07, "loss": 0.2333, "step": 4071 }, { "epoch": 1.0835550824906865, "grad_norm": 0.2824823558330536, "learning_rate": 1.807791809621499e-07, "loss": 0.2232, "step": 4072 }, { "epoch": 1.0838211814795105, "grad_norm": 0.2908635437488556, "learning_rate": 1.8076912293856647e-07, "loss": 0.223, "step": 4073 }, { "epoch": 1.0840872804683341, "grad_norm": 0.28816694021224976, "learning_rate": 1.8075906256400364e-07, "loss": 0.2104, "step": 4074 }, { "epoch": 1.084353379457158, "grad_norm": 0.27900782227516174, "learning_rate": 1.8074899983875412e-07, "loss": 0.2006, "step": 4075 }, { "epoch": 1.084619478445982, "grad_norm": 0.2878638207912445, "learning_rate": 1.8073893476311095e-07, "loss": 0.2354, "step": 4076 }, { "epoch": 1.0848855774348058, "grad_norm": 0.28352636098861694, "learning_rate": 1.8072886733736697e-07, "loss": 0.2324, "step": 4077 }, { "epoch": 1.0851516764236295, "grad_norm": 0.3023815453052521, "learning_rate": 1.807187975618153e-07, "loss": 0.2376, "step": 4078 }, { "epoch": 1.0854177754124534, "grad_norm": 0.28566521406173706, "learning_rate": 1.8070872543674902e-07, "loss": 0.2315, "step": 4079 }, { "epoch": 1.0856838744012773, "grad_norm": 0.30140045285224915, "learning_rate": 1.8069865096246127e-07, "loss": 0.2242, "step": 4080 }, { "epoch": 1.085949973390101, "grad_norm": 0.25700533390045166, "learning_rate": 1.8068857413924536e-07, "loss": 0.1791, "step": 4081 }, { "epoch": 1.086216072378925, "grad_norm": 0.27883344888687134, "learning_rate": 1.8067849496739455e-07, "loss": 0.2205, "step": 4082 }, { "epoch": 1.0864821713677488, "grad_norm": 0.28792673349380493, "learning_rate": 1.8066841344720222e-07, "loss": 0.2261, "step": 4083 }, { "epoch": 1.0867482703565727, "grad_norm": 0.26224449276924133, "learning_rate": 1.8065832957896182e-07, "loss": 0.1963, "step": 4084 }, { "epoch": 1.0870143693453964, "grad_norm": 0.2883129119873047, "learning_rate": 1.8064824336296688e-07, "loss": 0.2334, "step": 4085 }, { "epoch": 1.0872804683342203, "grad_norm": 0.26453593373298645, "learning_rate": 1.8063815479951102e-07, "loss": 0.189, "step": 4086 }, { "epoch": 1.0875465673230442, "grad_norm": 0.27832457423210144, "learning_rate": 1.806280638888878e-07, "loss": 0.2003, "step": 4087 }, { "epoch": 1.087812666311868, "grad_norm": 0.5086332559585571, "learning_rate": 1.80617970631391e-07, "loss": 0.2386, "step": 4088 }, { "epoch": 1.0880787653006918, "grad_norm": 0.27178192138671875, "learning_rate": 1.8060787502731437e-07, "loss": 0.2178, "step": 4089 }, { "epoch": 1.0883448642895157, "grad_norm": 0.2739775776863098, "learning_rate": 1.8059777707695186e-07, "loss": 0.2278, "step": 4090 }, { "epoch": 1.0886109632783396, "grad_norm": 0.26060715317726135, "learning_rate": 1.8058767678059729e-07, "loss": 0.2267, "step": 4091 }, { "epoch": 1.0888770622671633, "grad_norm": 0.294816255569458, "learning_rate": 1.8057757413854469e-07, "loss": 0.2227, "step": 4092 }, { "epoch": 1.0891431612559872, "grad_norm": 0.29082196950912476, "learning_rate": 1.805674691510881e-07, "loss": 0.2122, "step": 4093 }, { "epoch": 1.089409260244811, "grad_norm": 0.2859703600406647, "learning_rate": 1.8055736181852174e-07, "loss": 0.2255, "step": 4094 }, { "epoch": 1.089675359233635, "grad_norm": 0.27949732542037964, "learning_rate": 1.805472521411397e-07, "loss": 0.2166, "step": 4095 }, { "epoch": 1.0899414582224587, "grad_norm": 0.2887176275253296, "learning_rate": 1.8053714011923627e-07, "loss": 0.2281, "step": 4096 }, { "epoch": 1.0902075572112826, "grad_norm": 0.25762566924095154, "learning_rate": 1.8052702575310586e-07, "loss": 0.2011, "step": 4097 }, { "epoch": 1.0904736562001065, "grad_norm": 0.29680684208869934, "learning_rate": 1.805169090430428e-07, "loss": 0.2429, "step": 4098 }, { "epoch": 1.0907397551889302, "grad_norm": 0.27039122581481934, "learning_rate": 1.8050678998934154e-07, "loss": 0.2149, "step": 4099 }, { "epoch": 1.091005854177754, "grad_norm": 0.285681813955307, "learning_rate": 1.8049666859229667e-07, "loss": 0.201, "step": 4100 }, { "epoch": 1.091271953166578, "grad_norm": 0.3109450042247772, "learning_rate": 1.8048654485220282e-07, "loss": 0.2145, "step": 4101 }, { "epoch": 1.091538052155402, "grad_norm": 0.29803720116615295, "learning_rate": 1.804764187693546e-07, "loss": 0.2322, "step": 4102 }, { "epoch": 1.0918041511442256, "grad_norm": 0.2750544846057892, "learning_rate": 1.804662903440468e-07, "loss": 0.2053, "step": 4103 }, { "epoch": 1.0920702501330495, "grad_norm": 0.3084236681461334, "learning_rate": 1.804561595765742e-07, "loss": 0.2167, "step": 4104 }, { "epoch": 1.0923363491218734, "grad_norm": 0.3167498707771301, "learning_rate": 1.804460264672317e-07, "loss": 0.2138, "step": 4105 }, { "epoch": 1.0926024481106973, "grad_norm": 0.27780330181121826, "learning_rate": 1.8043589101631424e-07, "loss": 0.2032, "step": 4106 }, { "epoch": 1.092868547099521, "grad_norm": 0.29070186614990234, "learning_rate": 1.8042575322411685e-07, "loss": 0.2133, "step": 4107 }, { "epoch": 1.0931346460883449, "grad_norm": 0.27425315976142883, "learning_rate": 1.8041561309093462e-07, "loss": 0.2166, "step": 4108 }, { "epoch": 1.0934007450771688, "grad_norm": 0.2940536141395569, "learning_rate": 1.8040547061706267e-07, "loss": 0.2291, "step": 4109 }, { "epoch": 1.0936668440659925, "grad_norm": 0.31614571809768677, "learning_rate": 1.8039532580279626e-07, "loss": 0.2178, "step": 4110 }, { "epoch": 1.0939329430548164, "grad_norm": 0.28048503398895264, "learning_rate": 1.8038517864843064e-07, "loss": 0.2132, "step": 4111 }, { "epoch": 1.0941990420436403, "grad_norm": 0.2764403522014618, "learning_rate": 1.8037502915426122e-07, "loss": 0.1976, "step": 4112 }, { "epoch": 1.0944651410324642, "grad_norm": 0.27813684940338135, "learning_rate": 1.8036487732058335e-07, "loss": 0.2147, "step": 4113 }, { "epoch": 1.0947312400212879, "grad_norm": 0.29972895979881287, "learning_rate": 1.803547231476926e-07, "loss": 0.2268, "step": 4114 }, { "epoch": 1.0949973390101118, "grad_norm": 0.2699314057826996, "learning_rate": 1.803445666358845e-07, "loss": 0.2038, "step": 4115 }, { "epoch": 1.0952634379989357, "grad_norm": 0.28123801946640015, "learning_rate": 1.8033440778545463e-07, "loss": 0.214, "step": 4116 }, { "epoch": 1.0955295369877593, "grad_norm": 0.27777954936027527, "learning_rate": 1.8032424659669878e-07, "loss": 0.2148, "step": 4117 }, { "epoch": 1.0957956359765832, "grad_norm": 0.3062267601490021, "learning_rate": 1.8031408306991266e-07, "loss": 0.2177, "step": 4118 }, { "epoch": 1.0960617349654072, "grad_norm": 0.28390491008758545, "learning_rate": 1.8030391720539214e-07, "loss": 0.2416, "step": 4119 }, { "epoch": 1.096327833954231, "grad_norm": 0.3032456040382385, "learning_rate": 1.8029374900343305e-07, "loss": 0.2308, "step": 4120 }, { "epoch": 1.0965939329430547, "grad_norm": 0.27178817987442017, "learning_rate": 1.8028357846433142e-07, "loss": 0.1972, "step": 4121 }, { "epoch": 1.0968600319318786, "grad_norm": 0.2682192325592041, "learning_rate": 1.802734055883833e-07, "loss": 0.2109, "step": 4122 }, { "epoch": 1.0971261309207025, "grad_norm": 0.3017023801803589, "learning_rate": 1.8026323037588473e-07, "loss": 0.2209, "step": 4123 }, { "epoch": 1.0973922299095262, "grad_norm": 0.3129788339138031, "learning_rate": 1.8025305282713196e-07, "loss": 0.23, "step": 4124 }, { "epoch": 1.0976583288983501, "grad_norm": 0.28023210167884827, "learning_rate": 1.8024287294242119e-07, "loss": 0.2144, "step": 4125 }, { "epoch": 1.097924427887174, "grad_norm": 0.29289519786834717, "learning_rate": 1.8023269072204875e-07, "loss": 0.2253, "step": 4126 }, { "epoch": 1.098190526875998, "grad_norm": 0.2760787606239319, "learning_rate": 1.80222506166311e-07, "loss": 0.2338, "step": 4127 }, { "epoch": 1.0984566258648216, "grad_norm": 0.27832284569740295, "learning_rate": 1.8021231927550437e-07, "loss": 0.1941, "step": 4128 }, { "epoch": 1.0987227248536455, "grad_norm": 0.27847185730934143, "learning_rate": 1.8020213004992543e-07, "loss": 0.2249, "step": 4129 }, { "epoch": 1.0989888238424694, "grad_norm": 0.3029113709926605, "learning_rate": 1.801919384898707e-07, "loss": 0.2265, "step": 4130 }, { "epoch": 1.0992549228312933, "grad_norm": 0.2886956036090851, "learning_rate": 1.8018174459563684e-07, "loss": 0.2183, "step": 4131 }, { "epoch": 1.099521021820117, "grad_norm": 0.3039020895957947, "learning_rate": 1.8017154836752062e-07, "loss": 0.223, "step": 4132 }, { "epoch": 1.099787120808941, "grad_norm": 0.5423263907432556, "learning_rate": 1.8016134980581878e-07, "loss": 0.2461, "step": 4133 }, { "epoch": 1.1000532197977648, "grad_norm": 0.2792983949184418, "learning_rate": 1.8015114891082818e-07, "loss": 0.2044, "step": 4134 }, { "epoch": 1.1003193187865885, "grad_norm": 0.2982251048088074, "learning_rate": 1.8014094568284573e-07, "loss": 0.2328, "step": 4135 }, { "epoch": 1.1005854177754124, "grad_norm": 0.2911139726638794, "learning_rate": 1.8013074012216846e-07, "loss": 0.2166, "step": 4136 }, { "epoch": 1.1008515167642363, "grad_norm": 0.2624255418777466, "learning_rate": 1.8012053222909337e-07, "loss": 0.2102, "step": 4137 }, { "epoch": 1.1011176157530602, "grad_norm": 0.2956613004207611, "learning_rate": 1.8011032200391765e-07, "loss": 0.2257, "step": 4138 }, { "epoch": 1.101383714741884, "grad_norm": 0.28753772377967834, "learning_rate": 1.8010010944693845e-07, "loss": 0.2347, "step": 4139 }, { "epoch": 1.1016498137307078, "grad_norm": 0.27164405584335327, "learning_rate": 1.8008989455845304e-07, "loss": 0.2096, "step": 4140 }, { "epoch": 1.1019159127195317, "grad_norm": 0.28268754482269287, "learning_rate": 1.8007967733875878e-07, "loss": 0.2155, "step": 4141 }, { "epoch": 1.1021820117083556, "grad_norm": 0.26067087054252625, "learning_rate": 1.80069457788153e-07, "loss": 0.2086, "step": 4142 }, { "epoch": 1.1024481106971793, "grad_norm": 0.26854678988456726, "learning_rate": 1.800592359069332e-07, "loss": 0.2103, "step": 4143 }, { "epoch": 1.1027142096860032, "grad_norm": 0.2758118808269501, "learning_rate": 1.8004901169539694e-07, "loss": 0.2162, "step": 4144 }, { "epoch": 1.102980308674827, "grad_norm": 0.2862413823604584, "learning_rate": 1.8003878515384175e-07, "loss": 0.2161, "step": 4145 }, { "epoch": 1.1032464076636508, "grad_norm": 0.2687090337276459, "learning_rate": 1.800285562825654e-07, "loss": 0.2216, "step": 4146 }, { "epoch": 1.1035125066524747, "grad_norm": 0.28422847390174866, "learning_rate": 1.8001832508186555e-07, "loss": 0.2239, "step": 4147 }, { "epoch": 1.1037786056412986, "grad_norm": 0.2714386284351349, "learning_rate": 1.8000809155204003e-07, "loss": 0.2115, "step": 4148 }, { "epoch": 1.1040447046301225, "grad_norm": 0.28054502606391907, "learning_rate": 1.7999785569338667e-07, "loss": 0.2006, "step": 4149 }, { "epoch": 1.1043108036189462, "grad_norm": 0.30517473816871643, "learning_rate": 1.7998761750620348e-07, "loss": 0.2581, "step": 4150 }, { "epoch": 1.10457690260777, "grad_norm": 0.33683544397354126, "learning_rate": 1.799773769907884e-07, "loss": 0.2036, "step": 4151 }, { "epoch": 1.104843001596594, "grad_norm": 0.28571900725364685, "learning_rate": 1.7996713414743953e-07, "loss": 0.2238, "step": 4152 }, { "epoch": 1.105109100585418, "grad_norm": 0.3006099760532379, "learning_rate": 1.7995688897645507e-07, "loss": 0.2257, "step": 4153 }, { "epoch": 1.1053751995742416, "grad_norm": 0.26059430837631226, "learning_rate": 1.7994664147813312e-07, "loss": 0.2133, "step": 4154 }, { "epoch": 1.1056412985630655, "grad_norm": 0.2708281874656677, "learning_rate": 1.7993639165277202e-07, "loss": 0.2146, "step": 4155 }, { "epoch": 1.1059073975518894, "grad_norm": 0.2832929790019989, "learning_rate": 1.7992613950067016e-07, "loss": 0.2095, "step": 4156 }, { "epoch": 1.106173496540713, "grad_norm": 0.28166377544403076, "learning_rate": 1.799158850221259e-07, "loss": 0.2148, "step": 4157 }, { "epoch": 1.106439595529537, "grad_norm": 0.2599562704563141, "learning_rate": 1.799056282174377e-07, "loss": 0.1917, "step": 4158 }, { "epoch": 1.1067056945183609, "grad_norm": 0.30468663573265076, "learning_rate": 1.798953690869041e-07, "loss": 0.2453, "step": 4159 }, { "epoch": 1.1069717935071848, "grad_norm": 0.2804996371269226, "learning_rate": 1.798851076308238e-07, "loss": 0.2049, "step": 4160 }, { "epoch": 1.1072378924960085, "grad_norm": 0.27571824193000793, "learning_rate": 1.7987484384949543e-07, "loss": 0.2138, "step": 4161 }, { "epoch": 1.1075039914848324, "grad_norm": 0.27705642580986023, "learning_rate": 1.7986457774321776e-07, "loss": 0.2388, "step": 4162 }, { "epoch": 1.1077700904736563, "grad_norm": 0.29702121019363403, "learning_rate": 1.7985430931228959e-07, "loss": 0.2277, "step": 4163 }, { "epoch": 1.10803618946248, "grad_norm": 0.2948888838291168, "learning_rate": 1.798440385570098e-07, "loss": 0.2359, "step": 4164 }, { "epoch": 1.1083022884513039, "grad_norm": 0.29005616903305054, "learning_rate": 1.7983376547767736e-07, "loss": 0.2249, "step": 4165 }, { "epoch": 1.1085683874401278, "grad_norm": 0.30466240644454956, "learning_rate": 1.7982349007459132e-07, "loss": 0.2292, "step": 4166 }, { "epoch": 1.1088344864289517, "grad_norm": 0.27675947546958923, "learning_rate": 1.7981321234805073e-07, "loss": 0.2146, "step": 4167 }, { "epoch": 1.1091005854177753, "grad_norm": 0.29330796003341675, "learning_rate": 1.7980293229835476e-07, "loss": 0.1912, "step": 4168 }, { "epoch": 1.1093666844065992, "grad_norm": 0.3077986240386963, "learning_rate": 1.7979264992580264e-07, "loss": 0.2303, "step": 4169 }, { "epoch": 1.1096327833954232, "grad_norm": 0.3066020607948303, "learning_rate": 1.7978236523069365e-07, "loss": 0.2165, "step": 4170 }, { "epoch": 1.1098988823842468, "grad_norm": 0.2893686294555664, "learning_rate": 1.7977207821332715e-07, "loss": 0.2231, "step": 4171 }, { "epoch": 1.1101649813730707, "grad_norm": 0.25752612948417664, "learning_rate": 1.7976178887400263e-07, "loss": 0.1956, "step": 4172 }, { "epoch": 1.1104310803618946, "grad_norm": 0.2835879623889923, "learning_rate": 1.7975149721301948e-07, "loss": 0.2149, "step": 4173 }, { "epoch": 1.1106971793507185, "grad_norm": 0.27636849880218506, "learning_rate": 1.7974120323067732e-07, "loss": 0.2044, "step": 4174 }, { "epoch": 1.1109632783395422, "grad_norm": 0.28793084621429443, "learning_rate": 1.797309069272758e-07, "loss": 0.2165, "step": 4175 }, { "epoch": 1.1112293773283661, "grad_norm": 0.2766759395599365, "learning_rate": 1.797206083031146e-07, "loss": 0.2139, "step": 4176 }, { "epoch": 1.11149547631719, "grad_norm": 0.28614822030067444, "learning_rate": 1.7971030735849347e-07, "loss": 0.1869, "step": 4177 }, { "epoch": 1.111761575306014, "grad_norm": 0.27057817578315735, "learning_rate": 1.797000040937123e-07, "loss": 0.2032, "step": 4178 }, { "epoch": 1.1120276742948376, "grad_norm": 0.26930615305900574, "learning_rate": 1.796896985090709e-07, "loss": 0.2186, "step": 4179 }, { "epoch": 1.1122937732836615, "grad_norm": 0.29044535756111145, "learning_rate": 1.7967939060486934e-07, "loss": 0.2221, "step": 4180 }, { "epoch": 1.1125598722724854, "grad_norm": 0.28542062640190125, "learning_rate": 1.7966908038140755e-07, "loss": 0.2089, "step": 4181 }, { "epoch": 1.112825971261309, "grad_norm": 0.27748456597328186, "learning_rate": 1.7965876783898572e-07, "loss": 0.2053, "step": 4182 }, { "epoch": 1.113092070250133, "grad_norm": 0.26828616857528687, "learning_rate": 1.7964845297790398e-07, "loss": 0.2236, "step": 4183 }, { "epoch": 1.113358169238957, "grad_norm": 0.26839789748191833, "learning_rate": 1.796381357984626e-07, "loss": 0.216, "step": 4184 }, { "epoch": 1.1136242682277808, "grad_norm": 0.25240883231163025, "learning_rate": 1.7962781630096184e-07, "loss": 0.2074, "step": 4185 }, { "epoch": 1.1138903672166045, "grad_norm": 0.2527110278606415, "learning_rate": 1.7961749448570212e-07, "loss": 0.1761, "step": 4186 }, { "epoch": 1.1141564662054284, "grad_norm": 0.2729703187942505, "learning_rate": 1.7960717035298384e-07, "loss": 0.2051, "step": 4187 }, { "epoch": 1.1144225651942523, "grad_norm": 0.29761508107185364, "learning_rate": 1.795968439031075e-07, "loss": 0.2065, "step": 4188 }, { "epoch": 1.1146886641830762, "grad_norm": 0.2987222671508789, "learning_rate": 1.7958651513637378e-07, "loss": 0.2305, "step": 4189 }, { "epoch": 1.1149547631719, "grad_norm": 0.27437156438827515, "learning_rate": 1.795761840530832e-07, "loss": 0.2018, "step": 4190 }, { "epoch": 1.1152208621607238, "grad_norm": 0.2793165445327759, "learning_rate": 1.7956585065353652e-07, "loss": 0.2186, "step": 4191 }, { "epoch": 1.1154869611495477, "grad_norm": 0.2878154218196869, "learning_rate": 1.7955551493803455e-07, "loss": 0.2185, "step": 4192 }, { "epoch": 1.1157530601383714, "grad_norm": 0.26378685235977173, "learning_rate": 1.7954517690687806e-07, "loss": 0.2189, "step": 4193 }, { "epoch": 1.1160191591271953, "grad_norm": 0.308188796043396, "learning_rate": 1.7953483656036803e-07, "loss": 0.2291, "step": 4194 }, { "epoch": 1.1162852581160192, "grad_norm": 0.3174145221710205, "learning_rate": 1.7952449389880543e-07, "loss": 0.234, "step": 4195 }, { "epoch": 1.116551357104843, "grad_norm": 0.2900084853172302, "learning_rate": 1.795141489224913e-07, "loss": 0.2413, "step": 4196 }, { "epoch": 1.1168174560936668, "grad_norm": 0.29727408289909363, "learning_rate": 1.7950380163172674e-07, "loss": 0.2419, "step": 4197 }, { "epoch": 1.1170835550824907, "grad_norm": 0.27136245369911194, "learning_rate": 1.794934520268129e-07, "loss": 0.2121, "step": 4198 }, { "epoch": 1.1173496540713146, "grad_norm": 0.2742839753627777, "learning_rate": 1.7948310010805113e-07, "loss": 0.2166, "step": 4199 }, { "epoch": 1.1176157530601383, "grad_norm": 0.28692564368247986, "learning_rate": 1.7947274587574266e-07, "loss": 0.245, "step": 4200 }, { "epoch": 1.1178818520489622, "grad_norm": 0.2811488211154938, "learning_rate": 1.7946238933018893e-07, "loss": 0.2071, "step": 4201 }, { "epoch": 1.118147951037786, "grad_norm": 0.27644285559654236, "learning_rate": 1.7945203047169138e-07, "loss": 0.2157, "step": 4202 }, { "epoch": 1.11841405002661, "grad_norm": 0.25439226627349854, "learning_rate": 1.7944166930055147e-07, "loss": 0.205, "step": 4203 }, { "epoch": 1.1186801490154337, "grad_norm": 0.278461754322052, "learning_rate": 1.7943130581707089e-07, "loss": 0.2296, "step": 4204 }, { "epoch": 1.1189462480042576, "grad_norm": 0.3001963496208191, "learning_rate": 1.794209400215512e-07, "loss": 0.2288, "step": 4205 }, { "epoch": 1.1192123469930815, "grad_norm": 0.2949415445327759, "learning_rate": 1.7941057191429416e-07, "loss": 0.2275, "step": 4206 }, { "epoch": 1.1194784459819052, "grad_norm": 0.50233393907547, "learning_rate": 1.7940020149560156e-07, "loss": 0.2268, "step": 4207 }, { "epoch": 1.119744544970729, "grad_norm": 0.2772933542728424, "learning_rate": 1.7938982876577528e-07, "loss": 0.2122, "step": 4208 }, { "epoch": 1.120010643959553, "grad_norm": 0.27788928151130676, "learning_rate": 1.793794537251172e-07, "loss": 0.2242, "step": 4209 }, { "epoch": 1.1202767429483769, "grad_norm": 0.2993742525577545, "learning_rate": 1.7936907637392932e-07, "loss": 0.2459, "step": 4210 }, { "epoch": 1.1205428419372006, "grad_norm": 0.30072906613349915, "learning_rate": 1.7935869671251376e-07, "loss": 0.2234, "step": 4211 }, { "epoch": 1.1208089409260245, "grad_norm": 0.2527308166027069, "learning_rate": 1.7934831474117254e-07, "loss": 0.211, "step": 4212 }, { "epoch": 1.1210750399148484, "grad_norm": 0.2734987437725067, "learning_rate": 1.793379304602079e-07, "loss": 0.2193, "step": 4213 }, { "epoch": 1.1213411389036723, "grad_norm": 0.27610042691230774, "learning_rate": 1.7932754386992212e-07, "loss": 0.2268, "step": 4214 }, { "epoch": 1.121607237892496, "grad_norm": 0.2735287845134735, "learning_rate": 1.7931715497061754e-07, "loss": 0.1975, "step": 4215 }, { "epoch": 1.1218733368813198, "grad_norm": 0.31523022055625916, "learning_rate": 1.7930676376259646e-07, "loss": 0.2073, "step": 4216 }, { "epoch": 1.1221394358701438, "grad_norm": 0.2799062430858612, "learning_rate": 1.7929637024616144e-07, "loss": 0.2114, "step": 4217 }, { "epoch": 1.1224055348589674, "grad_norm": 0.26742151379585266, "learning_rate": 1.7928597442161496e-07, "loss": 0.209, "step": 4218 }, { "epoch": 1.1226716338477913, "grad_norm": 0.2720832824707031, "learning_rate": 1.7927557628925965e-07, "loss": 0.211, "step": 4219 }, { "epoch": 1.1229377328366152, "grad_norm": 0.2960984706878662, "learning_rate": 1.7926517584939814e-07, "loss": 0.2158, "step": 4220 }, { "epoch": 1.1232038318254391, "grad_norm": 0.29001569747924805, "learning_rate": 1.7925477310233314e-07, "loss": 0.2262, "step": 4221 }, { "epoch": 1.1234699308142628, "grad_norm": 0.2881331443786621, "learning_rate": 1.7924436804836754e-07, "loss": 0.222, "step": 4222 }, { "epoch": 1.1237360298030867, "grad_norm": 0.28161177039146423, "learning_rate": 1.792339606878041e-07, "loss": 0.2114, "step": 4223 }, { "epoch": 1.1240021287919106, "grad_norm": 0.2974108159542084, "learning_rate": 1.792235510209458e-07, "loss": 0.2208, "step": 4224 }, { "epoch": 1.1242682277807345, "grad_norm": 0.3207288384437561, "learning_rate": 1.7921313904809562e-07, "loss": 0.2252, "step": 4225 }, { "epoch": 1.1245343267695582, "grad_norm": 0.2979962229728699, "learning_rate": 1.7920272476955667e-07, "loss": 0.2233, "step": 4226 }, { "epoch": 1.1248004257583821, "grad_norm": 0.3104831278324127, "learning_rate": 1.7919230818563205e-07, "loss": 0.223, "step": 4227 }, { "epoch": 1.125066524747206, "grad_norm": 0.26662713289260864, "learning_rate": 1.7918188929662492e-07, "loss": 0.1986, "step": 4228 }, { "epoch": 1.1253326237360297, "grad_norm": 0.2892214357852936, "learning_rate": 1.791714681028386e-07, "loss": 0.2331, "step": 4229 }, { "epoch": 1.1255987227248536, "grad_norm": 0.31024131178855896, "learning_rate": 1.791610446045764e-07, "loss": 0.2213, "step": 4230 }, { "epoch": 1.1258648217136775, "grad_norm": 0.28466513752937317, "learning_rate": 1.7915061880214177e-07, "loss": 0.2121, "step": 4231 }, { "epoch": 1.1261309207025014, "grad_norm": 0.42186257243156433, "learning_rate": 1.7914019069583812e-07, "loss": 0.2112, "step": 4232 }, { "epoch": 1.126397019691325, "grad_norm": 0.28827551007270813, "learning_rate": 1.79129760285969e-07, "loss": 0.2142, "step": 4233 }, { "epoch": 1.126663118680149, "grad_norm": 0.2558707594871521, "learning_rate": 1.7911932757283803e-07, "loss": 0.196, "step": 4234 }, { "epoch": 1.126929217668973, "grad_norm": 0.27855396270751953, "learning_rate": 1.7910889255674885e-07, "loss": 0.221, "step": 4235 }, { "epoch": 1.1271953166577968, "grad_norm": 0.28147026896476746, "learning_rate": 1.790984552380052e-07, "loss": 0.2195, "step": 4236 }, { "epoch": 1.1274614156466205, "grad_norm": 0.29159271717071533, "learning_rate": 1.7908801561691093e-07, "loss": 0.22, "step": 4237 }, { "epoch": 1.1277275146354444, "grad_norm": 0.2802271544933319, "learning_rate": 1.7907757369376982e-07, "loss": 0.2345, "step": 4238 }, { "epoch": 1.1279936136242683, "grad_norm": 0.28818660974502563, "learning_rate": 1.7906712946888593e-07, "loss": 0.2177, "step": 4239 }, { "epoch": 1.128259712613092, "grad_norm": 0.2868058383464813, "learning_rate": 1.7905668294256317e-07, "loss": 0.2077, "step": 4240 }, { "epoch": 1.128525811601916, "grad_norm": 0.29130715131759644, "learning_rate": 1.7904623411510564e-07, "loss": 0.2036, "step": 4241 }, { "epoch": 1.1287919105907398, "grad_norm": 0.2712976634502411, "learning_rate": 1.7903578298681746e-07, "loss": 0.2019, "step": 4242 }, { "epoch": 1.1290580095795635, "grad_norm": 0.29227501153945923, "learning_rate": 1.7902532955800286e-07, "loss": 0.2136, "step": 4243 }, { "epoch": 1.1293241085683874, "grad_norm": 0.2889769673347473, "learning_rate": 1.790148738289661e-07, "loss": 0.227, "step": 4244 }, { "epoch": 1.1295902075572113, "grad_norm": 0.2942367196083069, "learning_rate": 1.7900441580001153e-07, "loss": 0.2331, "step": 4245 }, { "epoch": 1.1298563065460352, "grad_norm": 0.27252134680747986, "learning_rate": 1.7899395547144355e-07, "loss": 0.2099, "step": 4246 }, { "epoch": 1.1301224055348589, "grad_norm": 0.2907449007034302, "learning_rate": 1.7898349284356664e-07, "loss": 0.2312, "step": 4247 }, { "epoch": 1.1303885045236828, "grad_norm": 0.27338075637817383, "learning_rate": 1.789730279166853e-07, "loss": 0.1937, "step": 4248 }, { "epoch": 1.1306546035125067, "grad_norm": 0.28523531556129456, "learning_rate": 1.789625606911042e-07, "loss": 0.2036, "step": 4249 }, { "epoch": 1.1309207025013306, "grad_norm": 0.27781370282173157, "learning_rate": 1.7895209116712798e-07, "loss": 0.2099, "step": 4250 }, { "epoch": 1.1311868014901543, "grad_norm": 0.26867327094078064, "learning_rate": 1.7894161934506138e-07, "loss": 0.2018, "step": 4251 }, { "epoch": 1.1314529004789782, "grad_norm": 0.2807351052761078, "learning_rate": 1.789311452252092e-07, "loss": 0.2106, "step": 4252 }, { "epoch": 1.131718999467802, "grad_norm": 0.2614468038082123, "learning_rate": 1.7892066880787634e-07, "loss": 0.1938, "step": 4253 }, { "epoch": 1.1319850984566258, "grad_norm": 0.2816929519176483, "learning_rate": 1.789101900933677e-07, "loss": 0.2103, "step": 4254 }, { "epoch": 1.1322511974454497, "grad_norm": 0.2898329794406891, "learning_rate": 1.7889970908198834e-07, "loss": 0.2267, "step": 4255 }, { "epoch": 1.1325172964342736, "grad_norm": 0.28785088658332825, "learning_rate": 1.788892257740433e-07, "loss": 0.2192, "step": 4256 }, { "epoch": 1.1327833954230975, "grad_norm": 0.27529630064964294, "learning_rate": 1.7887874016983776e-07, "loss": 0.2073, "step": 4257 }, { "epoch": 1.1330494944119212, "grad_norm": 0.27371135354042053, "learning_rate": 1.7886825226967687e-07, "loss": 0.2136, "step": 4258 }, { "epoch": 1.133315593400745, "grad_norm": 0.281657874584198, "learning_rate": 1.7885776207386594e-07, "loss": 0.2015, "step": 4259 }, { "epoch": 1.133581692389569, "grad_norm": 0.28454646468162537, "learning_rate": 1.788472695827103e-07, "loss": 0.2055, "step": 4260 }, { "epoch": 1.1338477913783929, "grad_norm": 0.2908768355846405, "learning_rate": 1.7883677479651538e-07, "loss": 0.2076, "step": 4261 }, { "epoch": 1.1341138903672165, "grad_norm": 0.28612202405929565, "learning_rate": 1.7882627771558663e-07, "loss": 0.2305, "step": 4262 }, { "epoch": 1.1343799893560405, "grad_norm": 0.30339205265045166, "learning_rate": 1.7881577834022962e-07, "loss": 0.2268, "step": 4263 }, { "epoch": 1.1346460883448644, "grad_norm": 0.2808389365673065, "learning_rate": 1.7880527667074995e-07, "loss": 0.2414, "step": 4264 }, { "epoch": 1.134912187333688, "grad_norm": 0.2657119333744049, "learning_rate": 1.7879477270745327e-07, "loss": 0.208, "step": 4265 }, { "epoch": 1.135178286322512, "grad_norm": 0.27016398310661316, "learning_rate": 1.7878426645064538e-07, "loss": 0.2161, "step": 4266 }, { "epoch": 1.1354443853113358, "grad_norm": 0.30118781328201294, "learning_rate": 1.78773757900632e-07, "loss": 0.2344, "step": 4267 }, { "epoch": 1.1357104843001598, "grad_norm": 0.2672651708126068, "learning_rate": 1.7876324705771912e-07, "loss": 0.2242, "step": 4268 }, { "epoch": 1.1359765832889834, "grad_norm": 0.30242469906806946, "learning_rate": 1.7875273392221257e-07, "loss": 0.2374, "step": 4269 }, { "epoch": 1.1362426822778073, "grad_norm": 0.2796161472797394, "learning_rate": 1.7874221849441847e-07, "loss": 0.2091, "step": 4270 }, { "epoch": 1.1365087812666312, "grad_norm": 0.3256562650203705, "learning_rate": 1.7873170077464282e-07, "loss": 0.203, "step": 4271 }, { "epoch": 1.1367748802554551, "grad_norm": 0.25531160831451416, "learning_rate": 1.787211807631918e-07, "loss": 0.1828, "step": 4272 }, { "epoch": 1.1370409792442788, "grad_norm": 0.3011200726032257, "learning_rate": 1.7871065846037156e-07, "loss": 0.2205, "step": 4273 }, { "epoch": 1.1373070782331027, "grad_norm": 0.3147968649864197, "learning_rate": 1.7870013386648843e-07, "loss": 0.2399, "step": 4274 }, { "epoch": 1.1375731772219266, "grad_norm": 0.30187591910362244, "learning_rate": 1.786896069818488e-07, "loss": 0.2486, "step": 4275 }, { "epoch": 1.1378392762107503, "grad_norm": 0.2803264260292053, "learning_rate": 1.7867907780675896e-07, "loss": 0.2052, "step": 4276 }, { "epoch": 1.1381053751995742, "grad_norm": 0.27939361333847046, "learning_rate": 1.7866854634152548e-07, "loss": 0.201, "step": 4277 }, { "epoch": 1.1383714741883981, "grad_norm": 0.280068963766098, "learning_rate": 1.786580125864549e-07, "loss": 0.2047, "step": 4278 }, { "epoch": 1.1386375731772218, "grad_norm": 0.2728881239891052, "learning_rate": 1.7864747654185378e-07, "loss": 0.2161, "step": 4279 }, { "epoch": 1.1389036721660457, "grad_norm": 0.2640136778354645, "learning_rate": 1.7863693820802883e-07, "loss": 0.2036, "step": 4280 }, { "epoch": 1.1391697711548696, "grad_norm": 0.25355878472328186, "learning_rate": 1.7862639758528678e-07, "loss": 0.2116, "step": 4281 }, { "epoch": 1.1394358701436935, "grad_norm": 0.2785889506340027, "learning_rate": 1.7861585467393447e-07, "loss": 0.2184, "step": 4282 }, { "epoch": 1.1397019691325172, "grad_norm": 0.26513755321502686, "learning_rate": 1.7860530947427875e-07, "loss": 0.195, "step": 4283 }, { "epoch": 1.139968068121341, "grad_norm": 0.28585284948349, "learning_rate": 1.7859476198662654e-07, "loss": 0.2173, "step": 4284 }, { "epoch": 1.140234167110165, "grad_norm": 0.2591307759284973, "learning_rate": 1.785842122112849e-07, "loss": 0.1918, "step": 4285 }, { "epoch": 1.140500266098989, "grad_norm": 0.2675327658653259, "learning_rate": 1.7857366014856088e-07, "loss": 0.2047, "step": 4286 }, { "epoch": 1.1407663650878126, "grad_norm": 0.2986229956150055, "learning_rate": 1.7856310579876164e-07, "loss": 0.2129, "step": 4287 }, { "epoch": 1.1410324640766365, "grad_norm": 0.29150328040122986, "learning_rate": 1.7855254916219437e-07, "loss": 0.2177, "step": 4288 }, { "epoch": 1.1412985630654604, "grad_norm": 0.2912864685058594, "learning_rate": 1.7854199023916637e-07, "loss": 0.2249, "step": 4289 }, { "epoch": 1.141564662054284, "grad_norm": 0.32222887873649597, "learning_rate": 1.785314290299849e-07, "loss": 0.2152, "step": 4290 }, { "epoch": 1.141830761043108, "grad_norm": 0.3010924756526947, "learning_rate": 1.785208655349575e-07, "loss": 0.2058, "step": 4291 }, { "epoch": 1.142096860031932, "grad_norm": 0.29713040590286255, "learning_rate": 1.7851029975439155e-07, "loss": 0.2192, "step": 4292 }, { "epoch": 1.1423629590207558, "grad_norm": 0.2868671715259552, "learning_rate": 1.7849973168859463e-07, "loss": 0.2187, "step": 4293 }, { "epoch": 1.1426290580095795, "grad_norm": 0.27850615978240967, "learning_rate": 1.7848916133787438e-07, "loss": 0.2149, "step": 4294 }, { "epoch": 1.1428951569984034, "grad_norm": 0.2987509071826935, "learning_rate": 1.784785887025384e-07, "loss": 0.2239, "step": 4295 }, { "epoch": 1.1431612559872273, "grad_norm": 0.29541724920272827, "learning_rate": 1.7846801378289448e-07, "loss": 0.2153, "step": 4296 }, { "epoch": 1.1434273549760512, "grad_norm": 0.302338570356369, "learning_rate": 1.7845743657925043e-07, "loss": 0.241, "step": 4297 }, { "epoch": 1.1436934539648749, "grad_norm": 0.27467525005340576, "learning_rate": 1.784468570919141e-07, "loss": 0.2062, "step": 4298 }, { "epoch": 1.1439595529536988, "grad_norm": 0.27660372853279114, "learning_rate": 1.7843627532119343e-07, "loss": 0.2084, "step": 4299 }, { "epoch": 1.1442256519425227, "grad_norm": 0.2828136384487152, "learning_rate": 1.7842569126739647e-07, "loss": 0.2163, "step": 4300 }, { "epoch": 1.1444917509313464, "grad_norm": 0.2961643934249878, "learning_rate": 1.7841510493083125e-07, "loss": 0.2111, "step": 4301 }, { "epoch": 1.1447578499201703, "grad_norm": 0.29223164916038513, "learning_rate": 1.7840451631180596e-07, "loss": 0.2353, "step": 4302 }, { "epoch": 1.1450239489089942, "grad_norm": 0.2753903269767761, "learning_rate": 1.7839392541062876e-07, "loss": 0.1961, "step": 4303 }, { "epoch": 1.145290047897818, "grad_norm": 0.26520875096321106, "learning_rate": 1.7838333222760792e-07, "loss": 0.2145, "step": 4304 }, { "epoch": 1.1455561468866418, "grad_norm": 0.2681615650653839, "learning_rate": 1.7837273676305183e-07, "loss": 0.2258, "step": 4305 }, { "epoch": 1.1458222458754657, "grad_norm": 0.30150771141052246, "learning_rate": 1.7836213901726883e-07, "loss": 0.2385, "step": 4306 }, { "epoch": 1.1460883448642896, "grad_norm": 0.2666589915752411, "learning_rate": 1.7835153899056742e-07, "loss": 0.1974, "step": 4307 }, { "epoch": 1.1463544438531135, "grad_norm": 0.2848973870277405, "learning_rate": 1.783409366832562e-07, "loss": 0.212, "step": 4308 }, { "epoch": 1.1466205428419372, "grad_norm": 0.2801276445388794, "learning_rate": 1.783303320956437e-07, "loss": 0.2307, "step": 4309 }, { "epoch": 1.146886641830761, "grad_norm": 0.2727839946746826, "learning_rate": 1.783197252280386e-07, "loss": 0.2054, "step": 4310 }, { "epoch": 1.147152740819585, "grad_norm": 0.28177082538604736, "learning_rate": 1.7830911608074967e-07, "loss": 0.2236, "step": 4311 }, { "epoch": 1.1474188398084086, "grad_norm": 0.2975480556488037, "learning_rate": 1.782985046540857e-07, "loss": 0.2332, "step": 4312 }, { "epoch": 1.1476849387972325, "grad_norm": 0.2941471040248871, "learning_rate": 1.7828789094835552e-07, "loss": 0.2176, "step": 4313 }, { "epoch": 1.1479510377860564, "grad_norm": 0.295092910528183, "learning_rate": 1.7827727496386816e-07, "loss": 0.216, "step": 4314 }, { "epoch": 1.1482171367748801, "grad_norm": 0.30746400356292725, "learning_rate": 1.7826665670093256e-07, "loss": 0.2294, "step": 4315 }, { "epoch": 1.148483235763704, "grad_norm": 0.29651686549186707, "learning_rate": 1.7825603615985778e-07, "loss": 0.2342, "step": 4316 }, { "epoch": 1.148749334752528, "grad_norm": 0.29943689703941345, "learning_rate": 1.78245413340953e-07, "loss": 0.2188, "step": 4317 }, { "epoch": 1.1490154337413518, "grad_norm": 0.3179521858692169, "learning_rate": 1.7823478824452738e-07, "loss": 0.2244, "step": 4318 }, { "epoch": 1.1492815327301757, "grad_norm": 0.2859724164009094, "learning_rate": 1.7822416087089023e-07, "loss": 0.2205, "step": 4319 }, { "epoch": 1.1495476317189994, "grad_norm": 0.2805781364440918, "learning_rate": 1.7821353122035086e-07, "loss": 0.195, "step": 4320 }, { "epoch": 1.1498137307078233, "grad_norm": 0.2768506705760956, "learning_rate": 1.7820289929321867e-07, "loss": 0.2212, "step": 4321 }, { "epoch": 1.1500798296966472, "grad_norm": 0.2814990282058716, "learning_rate": 1.7819226508980311e-07, "loss": 0.2149, "step": 4322 }, { "epoch": 1.150345928685471, "grad_norm": 0.27347323298454285, "learning_rate": 1.7818162861041376e-07, "loss": 0.2107, "step": 4323 }, { "epoch": 1.1506120276742948, "grad_norm": 0.3056589365005493, "learning_rate": 1.7817098985536018e-07, "loss": 0.2228, "step": 4324 }, { "epoch": 1.1508781266631187, "grad_norm": 0.2741968631744385, "learning_rate": 1.781603488249521e-07, "loss": 0.2153, "step": 4325 }, { "epoch": 1.1511442256519424, "grad_norm": 0.29834917187690735, "learning_rate": 1.7814970551949916e-07, "loss": 0.2436, "step": 4326 }, { "epoch": 1.1514103246407663, "grad_norm": 0.30168095231056213, "learning_rate": 1.7813905993931118e-07, "loss": 0.223, "step": 4327 }, { "epoch": 1.1516764236295902, "grad_norm": 0.2592213749885559, "learning_rate": 1.7812841208469814e-07, "loss": 0.2001, "step": 4328 }, { "epoch": 1.1519425226184141, "grad_norm": 0.2732047140598297, "learning_rate": 1.7811776195596982e-07, "loss": 0.1914, "step": 4329 }, { "epoch": 1.1522086216072378, "grad_norm": 0.2910260856151581, "learning_rate": 1.7810710955343629e-07, "loss": 0.206, "step": 4330 }, { "epoch": 1.1524747205960617, "grad_norm": 0.2944347560405731, "learning_rate": 1.7809645487740764e-07, "loss": 0.2179, "step": 4331 }, { "epoch": 1.1527408195848856, "grad_norm": 0.27463141083717346, "learning_rate": 1.7808579792819393e-07, "loss": 0.1972, "step": 4332 }, { "epoch": 1.1530069185737095, "grad_norm": 0.2905805706977844, "learning_rate": 1.780751387061054e-07, "loss": 0.2254, "step": 4333 }, { "epoch": 1.1532730175625332, "grad_norm": 0.28476905822753906, "learning_rate": 1.7806447721145228e-07, "loss": 0.2432, "step": 4334 }, { "epoch": 1.153539116551357, "grad_norm": 0.28698471188545227, "learning_rate": 1.7805381344454494e-07, "loss": 0.2323, "step": 4335 }, { "epoch": 1.153805215540181, "grad_norm": 0.3207419216632843, "learning_rate": 1.7804314740569374e-07, "loss": 0.2289, "step": 4336 }, { "epoch": 1.1540713145290047, "grad_norm": 0.27849045395851135, "learning_rate": 1.7803247909520918e-07, "loss": 0.2315, "step": 4337 }, { "epoch": 1.1543374135178286, "grad_norm": 0.2691495418548584, "learning_rate": 1.7802180851340175e-07, "loss": 0.1943, "step": 4338 }, { "epoch": 1.1546035125066525, "grad_norm": 0.2765692174434662, "learning_rate": 1.7801113566058205e-07, "loss": 0.196, "step": 4339 }, { "epoch": 1.1548696114954764, "grad_norm": 0.30937671661376953, "learning_rate": 1.7800046053706074e-07, "loss": 0.244, "step": 4340 }, { "epoch": 1.1551357104843, "grad_norm": 0.28141576051712036, "learning_rate": 1.779897831431486e-07, "loss": 0.2317, "step": 4341 }, { "epoch": 1.155401809473124, "grad_norm": 0.28986749053001404, "learning_rate": 1.779791034791563e-07, "loss": 0.2065, "step": 4342 }, { "epoch": 1.155667908461948, "grad_norm": 0.2762499153614044, "learning_rate": 1.779684215453948e-07, "loss": 0.2288, "step": 4343 }, { "epoch": 1.1559340074507718, "grad_norm": 0.3060736656188965, "learning_rate": 1.77957737342175e-07, "loss": 0.2316, "step": 4344 }, { "epoch": 1.1562001064395955, "grad_norm": 0.30609244108200073, "learning_rate": 1.7794705086980788e-07, "loss": 0.2141, "step": 4345 }, { "epoch": 1.1564662054284194, "grad_norm": 0.27424994111061096, "learning_rate": 1.7793636212860449e-07, "loss": 0.2034, "step": 4346 }, { "epoch": 1.1567323044172433, "grad_norm": 0.2611115872859955, "learning_rate": 1.7792567111887596e-07, "loss": 0.2039, "step": 4347 }, { "epoch": 1.156998403406067, "grad_norm": 0.2884558439254761, "learning_rate": 1.7791497784093347e-07, "loss": 0.2271, "step": 4348 }, { "epoch": 1.1572645023948909, "grad_norm": 0.27404728531837463, "learning_rate": 1.779042822950883e-07, "loss": 0.2009, "step": 4349 }, { "epoch": 1.1575306013837148, "grad_norm": 0.28149402141571045, "learning_rate": 1.7789358448165172e-07, "loss": 0.1928, "step": 4350 }, { "epoch": 1.1577967003725387, "grad_norm": 0.3127945363521576, "learning_rate": 1.7788288440093514e-07, "loss": 0.2189, "step": 4351 }, { "epoch": 1.1580627993613624, "grad_norm": 0.28199517726898193, "learning_rate": 1.7787218205325004e-07, "loss": 0.2037, "step": 4352 }, { "epoch": 1.1583288983501863, "grad_norm": 0.30013880133628845, "learning_rate": 1.7786147743890793e-07, "loss": 0.2047, "step": 4353 }, { "epoch": 1.1585949973390102, "grad_norm": 0.28993088006973267, "learning_rate": 1.7785077055822034e-07, "loss": 0.2161, "step": 4354 }, { "epoch": 1.158861096327834, "grad_norm": 0.30789950489997864, "learning_rate": 1.7784006141149897e-07, "loss": 0.2253, "step": 4355 }, { "epoch": 1.1591271953166578, "grad_norm": 0.29658883810043335, "learning_rate": 1.7782934999905553e-07, "loss": 0.2156, "step": 4356 }, { "epoch": 1.1593932943054817, "grad_norm": 0.302455872297287, "learning_rate": 1.778186363212018e-07, "loss": 0.2212, "step": 4357 }, { "epoch": 1.1596593932943056, "grad_norm": 0.29798054695129395, "learning_rate": 1.7780792037824958e-07, "loss": 0.2336, "step": 4358 }, { "epoch": 1.1599254922831292, "grad_norm": 0.2852323055267334, "learning_rate": 1.7779720217051086e-07, "loss": 0.2218, "step": 4359 }, { "epoch": 1.1601915912719531, "grad_norm": 0.2667693495750427, "learning_rate": 1.777864816982976e-07, "loss": 0.2112, "step": 4360 }, { "epoch": 1.160457690260777, "grad_norm": 0.27372997999191284, "learning_rate": 1.7777575896192176e-07, "loss": 0.2188, "step": 4361 }, { "epoch": 1.1607237892496007, "grad_norm": 0.26295679807662964, "learning_rate": 1.7776503396169558e-07, "loss": 0.2123, "step": 4362 }, { "epoch": 1.1609898882384246, "grad_norm": 0.27997496724128723, "learning_rate": 1.7775430669793116e-07, "loss": 0.2159, "step": 4363 }, { "epoch": 1.1612559872272485, "grad_norm": 0.2611351013183594, "learning_rate": 1.7774357717094076e-07, "loss": 0.2111, "step": 4364 }, { "epoch": 1.1615220862160724, "grad_norm": 0.28657326102256775, "learning_rate": 1.7773284538103667e-07, "loss": 0.208, "step": 4365 }, { "epoch": 1.1617881852048961, "grad_norm": 0.2834019064903259, "learning_rate": 1.777221113285313e-07, "loss": 0.1983, "step": 4366 }, { "epoch": 1.16205428419372, "grad_norm": 0.2782065272331238, "learning_rate": 1.7771137501373708e-07, "loss": 0.2007, "step": 4367 }, { "epoch": 1.162320383182544, "grad_norm": 0.29673662781715393, "learning_rate": 1.777006364369665e-07, "loss": 0.2324, "step": 4368 }, { "epoch": 1.1625864821713678, "grad_norm": 0.26014837622642517, "learning_rate": 1.7768989559853213e-07, "loss": 0.2148, "step": 4369 }, { "epoch": 1.1628525811601915, "grad_norm": 0.28604722023010254, "learning_rate": 1.7767915249874664e-07, "loss": 0.1922, "step": 4370 }, { "epoch": 1.1631186801490154, "grad_norm": 0.3012373745441437, "learning_rate": 1.7766840713792269e-07, "loss": 0.2353, "step": 4371 }, { "epoch": 1.1633847791378393, "grad_norm": 0.30726000666618347, "learning_rate": 1.776576595163731e-07, "loss": 0.2323, "step": 4372 }, { "epoch": 1.163650878126663, "grad_norm": 0.27520862221717834, "learning_rate": 1.7764690963441065e-07, "loss": 0.2323, "step": 4373 }, { "epoch": 1.163916977115487, "grad_norm": 0.28179076313972473, "learning_rate": 1.7763615749234825e-07, "loss": 0.2023, "step": 4374 }, { "epoch": 1.1641830761043108, "grad_norm": 0.3012482523918152, "learning_rate": 1.7762540309049894e-07, "loss": 0.22, "step": 4375 }, { "epoch": 1.1644491750931347, "grad_norm": 0.2948943078517914, "learning_rate": 1.7761464642917567e-07, "loss": 0.2195, "step": 4376 }, { "epoch": 1.1647152740819584, "grad_norm": 0.2597139775753021, "learning_rate": 1.7760388750869159e-07, "loss": 0.2138, "step": 4377 }, { "epoch": 1.1649813730707823, "grad_norm": 0.3130913972854614, "learning_rate": 1.775931263293598e-07, "loss": 0.2343, "step": 4378 }, { "epoch": 1.1652474720596062, "grad_norm": 0.28124111890792847, "learning_rate": 1.775823628914936e-07, "loss": 0.2179, "step": 4379 }, { "epoch": 1.1655135710484301, "grad_norm": 0.29900482296943665, "learning_rate": 1.7757159719540623e-07, "loss": 0.2378, "step": 4380 }, { "epoch": 1.1657796700372538, "grad_norm": 0.30629798769950867, "learning_rate": 1.775608292414111e-07, "loss": 0.2337, "step": 4381 }, { "epoch": 1.1660457690260777, "grad_norm": 0.28731420636177063, "learning_rate": 1.7755005902982162e-07, "loss": 0.2179, "step": 4382 }, { "epoch": 1.1663118680149016, "grad_norm": 0.3201312720775604, "learning_rate": 1.775392865609513e-07, "loss": 0.2425, "step": 4383 }, { "epoch": 1.1665779670037253, "grad_norm": 0.3116626739501953, "learning_rate": 1.7752851183511363e-07, "loss": 0.2264, "step": 4384 }, { "epoch": 1.1668440659925492, "grad_norm": 0.2979801595211029, "learning_rate": 1.775177348526223e-07, "loss": 0.2175, "step": 4385 }, { "epoch": 1.167110164981373, "grad_norm": 0.263258159160614, "learning_rate": 1.77506955613791e-07, "loss": 0.2246, "step": 4386 }, { "epoch": 1.167376263970197, "grad_norm": 0.2782290577888489, "learning_rate": 1.7749617411893342e-07, "loss": 0.2113, "step": 4387 }, { "epoch": 1.1676423629590207, "grad_norm": 0.29181453585624695, "learning_rate": 1.7748539036836347e-07, "loss": 0.2297, "step": 4388 }, { "epoch": 1.1679084619478446, "grad_norm": 0.2666268050670624, "learning_rate": 1.7747460436239496e-07, "loss": 0.2098, "step": 4389 }, { "epoch": 1.1681745609366685, "grad_norm": 0.2539033889770508, "learning_rate": 1.7746381610134193e-07, "loss": 0.211, "step": 4390 }, { "epoch": 1.1684406599254924, "grad_norm": 0.2986508011817932, "learning_rate": 1.774530255855183e-07, "loss": 0.2376, "step": 4391 }, { "epoch": 1.168706758914316, "grad_norm": 0.3056860566139221, "learning_rate": 1.774422328152382e-07, "loss": 0.224, "step": 4392 }, { "epoch": 1.16897285790314, "grad_norm": 0.28056231141090393, "learning_rate": 1.774314377908158e-07, "loss": 0.1906, "step": 4393 }, { "epoch": 1.1692389568919639, "grad_norm": 0.27594348788261414, "learning_rate": 1.774206405125653e-07, "loss": 0.2205, "step": 4394 }, { "epoch": 1.1695050558807876, "grad_norm": 0.295815110206604, "learning_rate": 1.7740984098080097e-07, "loss": 0.2138, "step": 4395 }, { "epoch": 1.1697711548696115, "grad_norm": 0.2942505478858948, "learning_rate": 1.7739903919583714e-07, "loss": 0.2235, "step": 4396 }, { "epoch": 1.1700372538584354, "grad_norm": 0.26596906781196594, "learning_rate": 1.7738823515798827e-07, "loss": 0.2153, "step": 4397 }, { "epoch": 1.170303352847259, "grad_norm": 0.25753408670425415, "learning_rate": 1.773774288675688e-07, "loss": 0.2021, "step": 4398 }, { "epoch": 1.170569451836083, "grad_norm": 0.31155499815940857, "learning_rate": 1.7736662032489328e-07, "loss": 0.2234, "step": 4399 }, { "epoch": 1.1708355508249069, "grad_norm": 0.2796761095523834, "learning_rate": 1.7735580953027633e-07, "loss": 0.2083, "step": 4400 }, { "epoch": 1.1711016498137308, "grad_norm": 0.26078519225120544, "learning_rate": 1.773449964840326e-07, "loss": 0.1989, "step": 4401 }, { "epoch": 1.1713677488025545, "grad_norm": 0.27273067831993103, "learning_rate": 1.7733418118647688e-07, "loss": 0.196, "step": 4402 }, { "epoch": 1.1716338477913784, "grad_norm": 0.2811301052570343, "learning_rate": 1.7732336363792394e-07, "loss": 0.2179, "step": 4403 }, { "epoch": 1.1718999467802023, "grad_norm": 0.29876062273979187, "learning_rate": 1.7731254383868862e-07, "loss": 0.2095, "step": 4404 }, { "epoch": 1.1721660457690262, "grad_norm": 0.2765076756477356, "learning_rate": 1.773017217890859e-07, "loss": 0.2222, "step": 4405 }, { "epoch": 1.1724321447578498, "grad_norm": 0.2598368525505066, "learning_rate": 1.7729089748943077e-07, "loss": 0.1864, "step": 4406 }, { "epoch": 1.1726982437466738, "grad_norm": 0.295327365398407, "learning_rate": 1.772800709400383e-07, "loss": 0.2328, "step": 4407 }, { "epoch": 1.1729643427354977, "grad_norm": 0.25536176562309265, "learning_rate": 1.7726924214122363e-07, "loss": 0.1854, "step": 4408 }, { "epoch": 1.1732304417243213, "grad_norm": 0.24967584013938904, "learning_rate": 1.7725841109330194e-07, "loss": 0.1967, "step": 4409 }, { "epoch": 1.1734965407131452, "grad_norm": 0.2987481951713562, "learning_rate": 1.772475777965885e-07, "loss": 0.2378, "step": 4410 }, { "epoch": 1.1737626397019691, "grad_norm": 0.2892363369464874, "learning_rate": 1.7723674225139864e-07, "loss": 0.2199, "step": 4411 }, { "epoch": 1.174028738690793, "grad_norm": 0.27620917558670044, "learning_rate": 1.7722590445804778e-07, "loss": 0.2017, "step": 4412 }, { "epoch": 1.1742948376796167, "grad_norm": 0.3013019859790802, "learning_rate": 1.7721506441685134e-07, "loss": 0.2111, "step": 4413 }, { "epoch": 1.1745609366684406, "grad_norm": 0.2858198881149292, "learning_rate": 1.7720422212812486e-07, "loss": 0.2255, "step": 4414 }, { "epoch": 1.1748270356572645, "grad_norm": 0.2685917615890503, "learning_rate": 1.7719337759218393e-07, "loss": 0.2211, "step": 4415 }, { "epoch": 1.1750931346460884, "grad_norm": 0.29861533641815186, "learning_rate": 1.771825308093442e-07, "loss": 0.2445, "step": 4416 }, { "epoch": 1.1753592336349121, "grad_norm": 0.29212599992752075, "learning_rate": 1.7717168177992144e-07, "loss": 0.2389, "step": 4417 }, { "epoch": 1.175625332623736, "grad_norm": 0.3062972128391266, "learning_rate": 1.7716083050423135e-07, "loss": 0.227, "step": 4418 }, { "epoch": 1.17589143161256, "grad_norm": 0.27365642786026, "learning_rate": 1.7714997698258984e-07, "loss": 0.2126, "step": 4419 }, { "epoch": 1.1761575306013836, "grad_norm": 0.2671796679496765, "learning_rate": 1.7713912121531283e-07, "loss": 0.2043, "step": 4420 }, { "epoch": 1.1764236295902075, "grad_norm": 0.2913285791873932, "learning_rate": 1.771282632027163e-07, "loss": 0.2325, "step": 4421 }, { "epoch": 1.1766897285790314, "grad_norm": 0.24027098715305328, "learning_rate": 1.771174029451163e-07, "loss": 0.1892, "step": 4422 }, { "epoch": 1.1769558275678553, "grad_norm": 0.2683682441711426, "learning_rate": 1.771065404428289e-07, "loss": 0.2098, "step": 4423 }, { "epoch": 1.177221926556679, "grad_norm": 0.2786751389503479, "learning_rate": 1.7709567569617032e-07, "loss": 0.2148, "step": 4424 }, { "epoch": 1.177488025545503, "grad_norm": 0.2724061608314514, "learning_rate": 1.770848087054568e-07, "loss": 0.2004, "step": 4425 }, { "epoch": 1.1777541245343268, "grad_norm": 0.3013104498386383, "learning_rate": 1.7707393947100467e-07, "loss": 0.2085, "step": 4426 }, { "epoch": 1.1780202235231507, "grad_norm": 0.28874966502189636, "learning_rate": 1.7706306799313026e-07, "loss": 0.2147, "step": 4427 }, { "epoch": 1.1782863225119744, "grad_norm": 0.3156776428222656, "learning_rate": 1.7705219427215002e-07, "loss": 0.2314, "step": 4428 }, { "epoch": 1.1785524215007983, "grad_norm": 0.28101903200149536, "learning_rate": 1.770413183083805e-07, "loss": 0.2207, "step": 4429 }, { "epoch": 1.1788185204896222, "grad_norm": 0.2808304727077484, "learning_rate": 1.7703044010213822e-07, "loss": 0.2386, "step": 4430 }, { "epoch": 1.179084619478446, "grad_norm": 0.32368749380111694, "learning_rate": 1.7701955965373984e-07, "loss": 0.2213, "step": 4431 }, { "epoch": 1.1793507184672698, "grad_norm": 0.30447322130203247, "learning_rate": 1.7700867696350205e-07, "loss": 0.2139, "step": 4432 }, { "epoch": 1.1796168174560937, "grad_norm": 0.31567662954330444, "learning_rate": 1.7699779203174163e-07, "loss": 0.2257, "step": 4433 }, { "epoch": 1.1798829164449174, "grad_norm": 0.29335781931877136, "learning_rate": 1.769869048587754e-07, "loss": 0.2213, "step": 4434 }, { "epoch": 1.1801490154337413, "grad_norm": 0.2627798020839691, "learning_rate": 1.769760154449203e-07, "loss": 0.2146, "step": 4435 }, { "epoch": 1.1804151144225652, "grad_norm": 0.27451804280281067, "learning_rate": 1.7696512379049323e-07, "loss": 0.2051, "step": 4436 }, { "epoch": 1.180681213411389, "grad_norm": 0.2933090031147003, "learning_rate": 1.7695422989581124e-07, "loss": 0.2202, "step": 4437 }, { "epoch": 1.180947312400213, "grad_norm": 0.350758820772171, "learning_rate": 1.7694333376119141e-07, "loss": 0.2288, "step": 4438 }, { "epoch": 1.1812134113890367, "grad_norm": 0.2840454578399658, "learning_rate": 1.7693243538695096e-07, "loss": 0.2161, "step": 4439 }, { "epoch": 1.1814795103778606, "grad_norm": 0.2769070267677307, "learning_rate": 1.7692153477340702e-07, "loss": 0.2298, "step": 4440 }, { "epoch": 1.1817456093666845, "grad_norm": 0.3363916575908661, "learning_rate": 1.76910631920877e-07, "loss": 0.2265, "step": 4441 }, { "epoch": 1.1820117083555082, "grad_norm": 0.2888171374797821, "learning_rate": 1.7689972682967813e-07, "loss": 0.2332, "step": 4442 }, { "epoch": 1.182277807344332, "grad_norm": 0.2914973199367523, "learning_rate": 1.7688881950012787e-07, "loss": 0.2089, "step": 4443 }, { "epoch": 1.182543906333156, "grad_norm": 0.265433132648468, "learning_rate": 1.7687790993254372e-07, "loss": 0.2098, "step": 4444 }, { "epoch": 1.1828100053219797, "grad_norm": 0.28821712732315063, "learning_rate": 1.7686699812724323e-07, "loss": 0.2257, "step": 4445 }, { "epoch": 1.1830761043108036, "grad_norm": 0.2691488564014435, "learning_rate": 1.7685608408454402e-07, "loss": 0.2215, "step": 4446 }, { "epoch": 1.1833422032996275, "grad_norm": 0.2958214282989502, "learning_rate": 1.7684516780476374e-07, "loss": 0.2326, "step": 4447 }, { "epoch": 1.1836083022884514, "grad_norm": 0.27433887124061584, "learning_rate": 1.7683424928822018e-07, "loss": 0.2028, "step": 4448 }, { "epoch": 1.183874401277275, "grad_norm": 0.30762043595314026, "learning_rate": 1.7682332853523109e-07, "loss": 0.2494, "step": 4449 }, { "epoch": 1.184140500266099, "grad_norm": 0.29602673649787903, "learning_rate": 1.7681240554611438e-07, "loss": 0.2403, "step": 4450 }, { "epoch": 1.1844065992549229, "grad_norm": 0.3393596112728119, "learning_rate": 1.76801480321188e-07, "loss": 0.2255, "step": 4451 }, { "epoch": 1.1846726982437468, "grad_norm": 0.30269715189933777, "learning_rate": 1.7679055286076994e-07, "loss": 0.2183, "step": 4452 }, { "epoch": 1.1849387972325705, "grad_norm": 0.28054770827293396, "learning_rate": 1.7677962316517826e-07, "loss": 0.2122, "step": 4453 }, { "epoch": 1.1852048962213944, "grad_norm": 0.27228882908821106, "learning_rate": 1.767686912347311e-07, "loss": 0.2041, "step": 4454 }, { "epoch": 1.1854709952102183, "grad_norm": 0.2962953448295593, "learning_rate": 1.7675775706974667e-07, "loss": 0.2244, "step": 4455 }, { "epoch": 1.185737094199042, "grad_norm": 0.26012012362480164, "learning_rate": 1.7674682067054325e-07, "loss": 0.1968, "step": 4456 }, { "epoch": 1.1860031931878658, "grad_norm": 0.2714958190917969, "learning_rate": 1.7673588203743914e-07, "loss": 0.1944, "step": 4457 }, { "epoch": 1.1862692921766897, "grad_norm": 0.331500381231308, "learning_rate": 1.7672494117075275e-07, "loss": 0.2457, "step": 4458 }, { "epoch": 1.1865353911655137, "grad_norm": 0.28039228916168213, "learning_rate": 1.7671399807080253e-07, "loss": 0.2279, "step": 4459 }, { "epoch": 1.1868014901543373, "grad_norm": 0.3183606266975403, "learning_rate": 1.7670305273790701e-07, "loss": 0.1998, "step": 4460 }, { "epoch": 1.1870675891431612, "grad_norm": 0.2676646411418915, "learning_rate": 1.766921051723848e-07, "loss": 0.2109, "step": 4461 }, { "epoch": 1.1873336881319851, "grad_norm": 0.273491233587265, "learning_rate": 1.766811553745545e-07, "loss": 0.2244, "step": 4462 }, { "epoch": 1.187599787120809, "grad_norm": 0.28518763184547424, "learning_rate": 1.766702033447349e-07, "loss": 0.202, "step": 4463 }, { "epoch": 1.1878658861096327, "grad_norm": 0.28041136264801025, "learning_rate": 1.7665924908324475e-07, "loss": 0.2089, "step": 4464 }, { "epoch": 1.1881319850984566, "grad_norm": 0.28847700357437134, "learning_rate": 1.766482925904029e-07, "loss": 0.2038, "step": 4465 }, { "epoch": 1.1883980840872805, "grad_norm": 0.26951363682746887, "learning_rate": 1.7663733386652827e-07, "loss": 0.1917, "step": 4466 }, { "epoch": 1.1886641830761042, "grad_norm": 0.27823665738105774, "learning_rate": 1.7662637291193985e-07, "loss": 0.2106, "step": 4467 }, { "epoch": 1.1889302820649281, "grad_norm": 0.2892266809940338, "learning_rate": 1.7661540972695663e-07, "loss": 0.2281, "step": 4468 }, { "epoch": 1.189196381053752, "grad_norm": 0.2929736077785492, "learning_rate": 1.766044443118978e-07, "loss": 0.2066, "step": 4469 }, { "epoch": 1.189462480042576, "grad_norm": 0.28164198994636536, "learning_rate": 1.765934766670825e-07, "loss": 0.2122, "step": 4470 }, { "epoch": 1.1897285790313996, "grad_norm": 0.29240867495536804, "learning_rate": 1.7658250679282993e-07, "loss": 0.2305, "step": 4471 }, { "epoch": 1.1899946780202235, "grad_norm": 0.281865656375885, "learning_rate": 1.7657153468945947e-07, "loss": 0.2168, "step": 4472 }, { "epoch": 1.1902607770090474, "grad_norm": 0.2779538631439209, "learning_rate": 1.7656056035729044e-07, "loss": 0.1956, "step": 4473 }, { "epoch": 1.1905268759978713, "grad_norm": 0.27372410893440247, "learning_rate": 1.765495837966423e-07, "loss": 0.1991, "step": 4474 }, { "epoch": 1.190792974986695, "grad_norm": 0.3096638023853302, "learning_rate": 1.7653860500783449e-07, "loss": 0.2211, "step": 4475 }, { "epoch": 1.191059073975519, "grad_norm": 0.29072391986846924, "learning_rate": 1.7652762399118664e-07, "loss": 0.2032, "step": 4476 }, { "epoch": 1.1913251729643428, "grad_norm": 0.27662569284439087, "learning_rate": 1.7651664074701838e-07, "loss": 0.2082, "step": 4477 }, { "epoch": 1.1915912719531665, "grad_norm": 0.265217661857605, "learning_rate": 1.7650565527564935e-07, "loss": 0.1953, "step": 4478 }, { "epoch": 1.1918573709419904, "grad_norm": 0.2738807797431946, "learning_rate": 1.7649466757739936e-07, "loss": 0.2097, "step": 4479 }, { "epoch": 1.1921234699308143, "grad_norm": 0.29388847947120667, "learning_rate": 1.764836776525882e-07, "loss": 0.2247, "step": 4480 }, { "epoch": 1.192389568919638, "grad_norm": 0.2946786880493164, "learning_rate": 1.764726855015358e-07, "loss": 0.1991, "step": 4481 }, { "epoch": 1.192655667908462, "grad_norm": 0.2739054262638092, "learning_rate": 1.7646169112456204e-07, "loss": 0.2122, "step": 4482 }, { "epoch": 1.1929217668972858, "grad_norm": 0.26840710639953613, "learning_rate": 1.7645069452198702e-07, "loss": 0.2064, "step": 4483 }, { "epoch": 1.1931878658861097, "grad_norm": 0.2846994400024414, "learning_rate": 1.7643969569413076e-07, "loss": 0.2226, "step": 4484 }, { "epoch": 1.1934539648749334, "grad_norm": 0.26373007893562317, "learning_rate": 1.7642869464131345e-07, "loss": 0.1833, "step": 4485 }, { "epoch": 1.1937200638637573, "grad_norm": 0.292442262172699, "learning_rate": 1.7641769136385526e-07, "loss": 0.2381, "step": 4486 }, { "epoch": 1.1939861628525812, "grad_norm": 0.3044460415840149, "learning_rate": 1.764066858620765e-07, "loss": 0.207, "step": 4487 }, { "epoch": 1.194252261841405, "grad_norm": 0.2866731882095337, "learning_rate": 1.7639567813629754e-07, "loss": 0.2016, "step": 4488 }, { "epoch": 1.1945183608302288, "grad_norm": 0.2535119354724884, "learning_rate": 1.763846681868387e-07, "loss": 0.1985, "step": 4489 }, { "epoch": 1.1947844598190527, "grad_norm": 0.2977617084980011, "learning_rate": 1.7637365601402055e-07, "loss": 0.2256, "step": 4490 }, { "epoch": 1.1950505588078766, "grad_norm": 0.27343684434890747, "learning_rate": 1.7636264161816353e-07, "loss": 0.2169, "step": 4491 }, { "epoch": 1.1953166577967003, "grad_norm": 0.2965017557144165, "learning_rate": 1.7635162499958832e-07, "loss": 0.2129, "step": 4492 }, { "epoch": 1.1955827567855242, "grad_norm": 0.2924748957157135, "learning_rate": 1.763406061586155e-07, "loss": 0.2255, "step": 4493 }, { "epoch": 1.195848855774348, "grad_norm": 0.31577444076538086, "learning_rate": 1.763295850955659e-07, "loss": 0.2294, "step": 4494 }, { "epoch": 1.196114954763172, "grad_norm": 0.30639758706092834, "learning_rate": 1.7631856181076024e-07, "loss": 0.214, "step": 4495 }, { "epoch": 1.1963810537519957, "grad_norm": 0.27552011609077454, "learning_rate": 1.763075363045194e-07, "loss": 0.2274, "step": 4496 }, { "epoch": 1.1966471527408196, "grad_norm": 0.30731090903282166, "learning_rate": 1.7629650857716433e-07, "loss": 0.2373, "step": 4497 }, { "epoch": 1.1969132517296435, "grad_norm": 0.2643125057220459, "learning_rate": 1.76285478629016e-07, "loss": 0.2168, "step": 4498 }, { "epoch": 1.1971793507184674, "grad_norm": 0.37644821405410767, "learning_rate": 1.7627444646039545e-07, "loss": 0.2039, "step": 4499 }, { "epoch": 1.197445449707291, "grad_norm": 0.2778985798358917, "learning_rate": 1.7626341207162382e-07, "loss": 0.2127, "step": 4500 }, { "epoch": 1.197711548696115, "grad_norm": 0.3668821454048157, "learning_rate": 1.7625237546302224e-07, "loss": 0.222, "step": 4501 }, { "epoch": 1.1979776476849389, "grad_norm": 0.28634294867515564, "learning_rate": 1.7624133663491202e-07, "loss": 0.224, "step": 4502 }, { "epoch": 1.1982437466737625, "grad_norm": 0.2989642918109894, "learning_rate": 1.762302955876145e-07, "loss": 0.2286, "step": 4503 }, { "epoch": 1.1985098456625864, "grad_norm": 0.28693854808807373, "learning_rate": 1.7621925232145095e-07, "loss": 0.2089, "step": 4504 }, { "epoch": 1.1987759446514104, "grad_norm": 0.27581650018692017, "learning_rate": 1.7620820683674286e-07, "loss": 0.2244, "step": 4505 }, { "epoch": 1.1990420436402343, "grad_norm": 0.326558917760849, "learning_rate": 1.7619715913381177e-07, "loss": 0.2131, "step": 4506 }, { "epoch": 1.199308142629058, "grad_norm": 0.27252817153930664, "learning_rate": 1.761861092129792e-07, "loss": 0.2236, "step": 4507 }, { "epoch": 1.1995742416178818, "grad_norm": 0.3061832785606384, "learning_rate": 1.7617505707456682e-07, "loss": 0.2172, "step": 4508 }, { "epoch": 1.1998403406067057, "grad_norm": 0.2839350700378418, "learning_rate": 1.7616400271889631e-07, "loss": 0.2171, "step": 4509 }, { "epoch": 1.2001064395955297, "grad_norm": 0.2867465615272522, "learning_rate": 1.7615294614628944e-07, "loss": 0.209, "step": 4510 }, { "epoch": 1.2003725385843533, "grad_norm": 0.27183181047439575, "learning_rate": 1.7614188735706807e-07, "loss": 0.2147, "step": 4511 }, { "epoch": 1.2006386375731772, "grad_norm": 0.2961457669734955, "learning_rate": 1.7613082635155402e-07, "loss": 0.214, "step": 4512 }, { "epoch": 1.2009047365620011, "grad_norm": 0.28121641278266907, "learning_rate": 1.7611976313006927e-07, "loss": 0.2102, "step": 4513 }, { "epoch": 1.2011708355508248, "grad_norm": 0.3188759982585907, "learning_rate": 1.761086976929359e-07, "loss": 0.2319, "step": 4514 }, { "epoch": 1.2014369345396487, "grad_norm": 0.2841799557209015, "learning_rate": 1.7609763004047592e-07, "loss": 0.2349, "step": 4515 }, { "epoch": 1.2017030335284726, "grad_norm": 0.27809059619903564, "learning_rate": 1.7608656017301153e-07, "loss": 0.2314, "step": 4516 }, { "epoch": 1.2019691325172963, "grad_norm": 0.2962881624698639, "learning_rate": 1.7607548809086492e-07, "loss": 0.2132, "step": 4517 }, { "epoch": 1.2022352315061202, "grad_norm": 0.30089473724365234, "learning_rate": 1.760644137943584e-07, "loss": 0.2456, "step": 4518 }, { "epoch": 1.2025013304949441, "grad_norm": 0.2955964505672455, "learning_rate": 1.7605333728381427e-07, "loss": 0.2287, "step": 4519 }, { "epoch": 1.202767429483768, "grad_norm": 0.2830379903316498, "learning_rate": 1.7604225855955499e-07, "loss": 0.2135, "step": 4520 }, { "epoch": 1.203033528472592, "grad_norm": 0.2761784493923187, "learning_rate": 1.7603117762190298e-07, "loss": 0.1976, "step": 4521 }, { "epoch": 1.2032996274614156, "grad_norm": 0.28908100724220276, "learning_rate": 1.7602009447118082e-07, "loss": 0.2325, "step": 4522 }, { "epoch": 1.2035657264502395, "grad_norm": 0.28502267599105835, "learning_rate": 1.7600900910771107e-07, "loss": 0.2183, "step": 4523 }, { "epoch": 1.2038318254390634, "grad_norm": 0.3287811875343323, "learning_rate": 1.7599792153181646e-07, "loss": 0.2209, "step": 4524 }, { "epoch": 1.204097924427887, "grad_norm": 0.2675454616546631, "learning_rate": 1.7598683174381965e-07, "loss": 0.2016, "step": 4525 }, { "epoch": 1.204364023416711, "grad_norm": 0.27058231830596924, "learning_rate": 1.7597573974404347e-07, "loss": 0.1879, "step": 4526 }, { "epoch": 1.204630122405535, "grad_norm": 0.29801732301712036, "learning_rate": 1.7596464553281079e-07, "loss": 0.2111, "step": 4527 }, { "epoch": 1.2048962213943586, "grad_norm": 0.2959214448928833, "learning_rate": 1.7595354911044451e-07, "loss": 0.2157, "step": 4528 }, { "epoch": 1.2051623203831825, "grad_norm": 0.2930718958377838, "learning_rate": 1.7594245047726762e-07, "loss": 0.2035, "step": 4529 }, { "epoch": 1.2054284193720064, "grad_norm": 0.27235063910484314, "learning_rate": 1.7593134963360318e-07, "loss": 0.2107, "step": 4530 }, { "epoch": 1.2056945183608303, "grad_norm": 0.26967567205429077, "learning_rate": 1.759202465797743e-07, "loss": 0.2005, "step": 4531 }, { "epoch": 1.205960617349654, "grad_norm": 0.31085965037345886, "learning_rate": 1.759091413161042e-07, "loss": 0.2289, "step": 4532 }, { "epoch": 1.2062267163384779, "grad_norm": 0.30181431770324707, "learning_rate": 1.7589803384291607e-07, "loss": 0.2284, "step": 4533 }, { "epoch": 1.2064928153273018, "grad_norm": 0.2660664916038513, "learning_rate": 1.7588692416053324e-07, "loss": 0.2107, "step": 4534 }, { "epoch": 1.2067589143161257, "grad_norm": 0.2760239839553833, "learning_rate": 1.7587581226927907e-07, "loss": 0.207, "step": 4535 }, { "epoch": 1.2070250133049494, "grad_norm": 0.28346097469329834, "learning_rate": 1.7586469816947704e-07, "loss": 0.1873, "step": 4536 }, { "epoch": 1.2072911122937733, "grad_norm": 0.28130394220352173, "learning_rate": 1.7585358186145064e-07, "loss": 0.2167, "step": 4537 }, { "epoch": 1.2075572112825972, "grad_norm": 0.29324278235435486, "learning_rate": 1.7584246334552338e-07, "loss": 0.2103, "step": 4538 }, { "epoch": 1.2078233102714209, "grad_norm": 0.28357183933258057, "learning_rate": 1.7583134262201897e-07, "loss": 0.2065, "step": 4539 }, { "epoch": 1.2080894092602448, "grad_norm": 0.27816906571388245, "learning_rate": 1.7582021969126104e-07, "loss": 0.2225, "step": 4540 }, { "epoch": 1.2083555082490687, "grad_norm": 0.29924675822257996, "learning_rate": 1.758090945535734e-07, "loss": 0.1898, "step": 4541 }, { "epoch": 1.2086216072378926, "grad_norm": 0.3018435835838318, "learning_rate": 1.7579796720927988e-07, "loss": 0.2179, "step": 4542 }, { "epoch": 1.2088877062267163, "grad_norm": 0.2750212252140045, "learning_rate": 1.757868376587043e-07, "loss": 0.2191, "step": 4543 }, { "epoch": 1.2091538052155402, "grad_norm": 0.28011974692344666, "learning_rate": 1.7577570590217068e-07, "loss": 0.2176, "step": 4544 }, { "epoch": 1.209419904204364, "grad_norm": 0.31281059980392456, "learning_rate": 1.7576457194000298e-07, "loss": 0.2501, "step": 4545 }, { "epoch": 1.209686003193188, "grad_norm": 0.2685055136680603, "learning_rate": 1.7575343577252535e-07, "loss": 0.2201, "step": 4546 }, { "epoch": 1.2099521021820117, "grad_norm": 0.4867071509361267, "learning_rate": 1.7574229740006187e-07, "loss": 0.2301, "step": 4547 }, { "epoch": 1.2102182011708356, "grad_norm": 0.2722283601760864, "learning_rate": 1.757311568229368e-07, "loss": 0.1968, "step": 4548 }, { "epoch": 1.2104843001596595, "grad_norm": 0.28503352403640747, "learning_rate": 1.7572001404147435e-07, "loss": 0.2208, "step": 4549 }, { "epoch": 1.2107503991484831, "grad_norm": 0.2929539084434509, "learning_rate": 1.757088690559989e-07, "loss": 0.2266, "step": 4550 }, { "epoch": 1.211016498137307, "grad_norm": 0.2851516008377075, "learning_rate": 1.756977218668349e-07, "loss": 0.201, "step": 4551 }, { "epoch": 1.211282597126131, "grad_norm": 0.28942441940307617, "learning_rate": 1.7568657247430673e-07, "loss": 0.2295, "step": 4552 }, { "epoch": 1.2115486961149549, "grad_norm": 0.28378549218177795, "learning_rate": 1.7567542087873896e-07, "loss": 0.217, "step": 4553 }, { "epoch": 1.2118147951037785, "grad_norm": 0.3196301758289337, "learning_rate": 1.7566426708045613e-07, "loss": 0.2172, "step": 4554 }, { "epoch": 1.2120808940926024, "grad_norm": 0.3241918683052063, "learning_rate": 1.75653111079783e-07, "loss": 0.2115, "step": 4555 }, { "epoch": 1.2123469930814263, "grad_norm": 0.292700856924057, "learning_rate": 1.7564195287704421e-07, "loss": 0.2223, "step": 4556 }, { "epoch": 1.2126130920702503, "grad_norm": 0.27515077590942383, "learning_rate": 1.7563079247256464e-07, "loss": 0.2137, "step": 4557 }, { "epoch": 1.212879191059074, "grad_norm": 0.2568323612213135, "learning_rate": 1.75619629866669e-07, "loss": 0.2036, "step": 4558 }, { "epoch": 1.2131452900478978, "grad_norm": 0.27765896916389465, "learning_rate": 1.756084650596823e-07, "loss": 0.2025, "step": 4559 }, { "epoch": 1.2134113890367217, "grad_norm": 0.3548843264579773, "learning_rate": 1.7559729805192948e-07, "loss": 0.2233, "step": 4560 }, { "epoch": 1.2136774880255454, "grad_norm": 0.2715297341346741, "learning_rate": 1.7558612884373564e-07, "loss": 0.2148, "step": 4561 }, { "epoch": 1.2139435870143693, "grad_norm": 0.2828320264816284, "learning_rate": 1.7557495743542582e-07, "loss": 0.2134, "step": 4562 }, { "epoch": 1.2142096860031932, "grad_norm": 0.30179932713508606, "learning_rate": 1.7556378382732522e-07, "loss": 0.2089, "step": 4563 }, { "epoch": 1.214475784992017, "grad_norm": 0.27819111943244934, "learning_rate": 1.7555260801975907e-07, "loss": 0.2022, "step": 4564 }, { "epoch": 1.2147418839808408, "grad_norm": 0.2806703448295593, "learning_rate": 1.755414300130527e-07, "loss": 0.2056, "step": 4565 }, { "epoch": 1.2150079829696647, "grad_norm": 0.28678151965141296, "learning_rate": 1.7553024980753142e-07, "loss": 0.2039, "step": 4566 }, { "epoch": 1.2152740819584886, "grad_norm": 0.26724520325660706, "learning_rate": 1.7551906740352068e-07, "loss": 0.2056, "step": 4567 }, { "epoch": 1.2155401809473123, "grad_norm": 0.3085921108722687, "learning_rate": 1.7550788280134596e-07, "loss": 0.2185, "step": 4568 }, { "epoch": 1.2158062799361362, "grad_norm": 0.2934015393257141, "learning_rate": 1.7549669600133284e-07, "loss": 0.2123, "step": 4569 }, { "epoch": 1.2160723789249601, "grad_norm": 0.2830889821052551, "learning_rate": 1.754855070038069e-07, "loss": 0.2117, "step": 4570 }, { "epoch": 1.216338477913784, "grad_norm": 0.26137930154800415, "learning_rate": 1.754743158090939e-07, "loss": 0.1904, "step": 4571 }, { "epoch": 1.2166045769026077, "grad_norm": 0.27461349964141846, "learning_rate": 1.7546312241751951e-07, "loss": 0.2135, "step": 4572 }, { "epoch": 1.2168706758914316, "grad_norm": 0.300352543592453, "learning_rate": 1.7545192682940955e-07, "loss": 0.2235, "step": 4573 }, { "epoch": 1.2171367748802555, "grad_norm": 0.31960996985435486, "learning_rate": 1.754407290450899e-07, "loss": 0.2343, "step": 4574 }, { "epoch": 1.2174028738690792, "grad_norm": 0.272847056388855, "learning_rate": 1.7542952906488652e-07, "loss": 0.2204, "step": 4575 }, { "epoch": 1.217668972857903, "grad_norm": 0.27752089500427246, "learning_rate": 1.7541832688912544e-07, "loss": 0.221, "step": 4576 }, { "epoch": 1.217935071846727, "grad_norm": 0.2820768654346466, "learning_rate": 1.7540712251813264e-07, "loss": 0.2036, "step": 4577 }, { "epoch": 1.218201170835551, "grad_norm": 0.2946132719516754, "learning_rate": 1.7539591595223428e-07, "loss": 0.2312, "step": 4578 }, { "epoch": 1.2184672698243746, "grad_norm": 0.3100142776966095, "learning_rate": 1.753847071917566e-07, "loss": 0.2275, "step": 4579 }, { "epoch": 1.2187333688131985, "grad_norm": 0.27832579612731934, "learning_rate": 1.7537349623702583e-07, "loss": 0.201, "step": 4580 }, { "epoch": 1.2189994678020224, "grad_norm": 0.29645970463752747, "learning_rate": 1.753622830883683e-07, "loss": 0.2164, "step": 4581 }, { "epoch": 1.2192655667908463, "grad_norm": 0.3110384941101074, "learning_rate": 1.7535106774611035e-07, "loss": 0.2309, "step": 4582 }, { "epoch": 1.21953166577967, "grad_norm": 0.28744611144065857, "learning_rate": 1.753398502105785e-07, "loss": 0.2231, "step": 4583 }, { "epoch": 1.2197977647684939, "grad_norm": 0.28288140892982483, "learning_rate": 1.753286304820992e-07, "loss": 0.2115, "step": 4584 }, { "epoch": 1.2200638637573178, "grad_norm": 0.267362505197525, "learning_rate": 1.7531740856099908e-07, "loss": 0.2169, "step": 4585 }, { "epoch": 1.2203299627461415, "grad_norm": 0.2823949158191681, "learning_rate": 1.7530618444760474e-07, "loss": 0.2133, "step": 4586 }, { "epoch": 1.2205960617349654, "grad_norm": 0.32245415449142456, "learning_rate": 1.7529495814224292e-07, "loss": 0.2059, "step": 4587 }, { "epoch": 1.2208621607237893, "grad_norm": 0.2945835590362549, "learning_rate": 1.7528372964524034e-07, "loss": 0.2171, "step": 4588 }, { "epoch": 1.2211282597126132, "grad_norm": 0.27449002861976624, "learning_rate": 1.752724989569239e-07, "loss": 0.2171, "step": 4589 }, { "epoch": 1.2213943587014369, "grad_norm": 0.285645991563797, "learning_rate": 1.7526126607762041e-07, "loss": 0.2212, "step": 4590 }, { "epoch": 1.2216604576902608, "grad_norm": 0.3126104176044464, "learning_rate": 1.7525003100765693e-07, "loss": 0.2416, "step": 4591 }, { "epoch": 1.2219265566790847, "grad_norm": 0.2752547562122345, "learning_rate": 1.7523879374736042e-07, "loss": 0.211, "step": 4592 }, { "epoch": 1.2221926556679086, "grad_norm": 0.2915039360523224, "learning_rate": 1.7522755429705796e-07, "loss": 0.2146, "step": 4593 }, { "epoch": 1.2224587546567323, "grad_norm": 0.26583805680274963, "learning_rate": 1.752163126570767e-07, "loss": 0.2058, "step": 4594 }, { "epoch": 1.2227248536455562, "grad_norm": 0.29820841550827026, "learning_rate": 1.752050688277439e-07, "loss": 0.2122, "step": 4595 }, { "epoch": 1.22299095263438, "grad_norm": 0.30099308490753174, "learning_rate": 1.7519382280938682e-07, "loss": 0.2294, "step": 4596 }, { "epoch": 1.2232570516232038, "grad_norm": 0.30066022276878357, "learning_rate": 1.751825746023328e-07, "loss": 0.2203, "step": 4597 }, { "epoch": 1.2235231506120277, "grad_norm": 0.27404293417930603, "learning_rate": 1.7517132420690922e-07, "loss": 0.2005, "step": 4598 }, { "epoch": 1.2237892496008516, "grad_norm": 0.2934686541557312, "learning_rate": 1.7516007162344363e-07, "loss": 0.2351, "step": 4599 }, { "epoch": 1.2240553485896752, "grad_norm": 0.2804255485534668, "learning_rate": 1.7514881685226345e-07, "loss": 0.2161, "step": 4600 }, { "epoch": 1.2243214475784991, "grad_norm": 0.28324243426322937, "learning_rate": 1.7513755989369634e-07, "loss": 0.2046, "step": 4601 }, { "epoch": 1.224587546567323, "grad_norm": 0.26693636178970337, "learning_rate": 1.7512630074806996e-07, "loss": 0.2073, "step": 4602 }, { "epoch": 1.224853645556147, "grad_norm": 0.296314537525177, "learning_rate": 1.7511503941571207e-07, "loss": 0.222, "step": 4603 }, { "epoch": 1.2251197445449706, "grad_norm": 0.3167436122894287, "learning_rate": 1.7510377589695035e-07, "loss": 0.231, "step": 4604 }, { "epoch": 1.2253858435337945, "grad_norm": 0.2895781695842743, "learning_rate": 1.7509251019211275e-07, "loss": 0.232, "step": 4605 }, { "epoch": 1.2256519425226184, "grad_norm": 0.2574502229690552, "learning_rate": 1.7508124230152714e-07, "loss": 0.2123, "step": 4606 }, { "epoch": 1.2259180415114423, "grad_norm": 0.2722451686859131, "learning_rate": 1.7506997222552157e-07, "loss": 0.2069, "step": 4607 }, { "epoch": 1.226184140500266, "grad_norm": 0.3150976002216339, "learning_rate": 1.7505869996442396e-07, "loss": 0.2213, "step": 4608 }, { "epoch": 1.22645023948909, "grad_norm": 0.27592533826828003, "learning_rate": 1.7504742551856252e-07, "loss": 0.2332, "step": 4609 }, { "epoch": 1.2267163384779138, "grad_norm": 0.2943585515022278, "learning_rate": 1.7503614888826533e-07, "loss": 0.2186, "step": 4610 }, { "epoch": 1.2269824374667375, "grad_norm": 0.29225218296051025, "learning_rate": 1.7502487007386072e-07, "loss": 0.2366, "step": 4611 }, { "epoch": 1.2272485364555614, "grad_norm": 0.2682282328605652, "learning_rate": 1.7501358907567694e-07, "loss": 0.2003, "step": 4612 }, { "epoch": 1.2275146354443853, "grad_norm": 0.2721850872039795, "learning_rate": 1.7500230589404234e-07, "loss": 0.2333, "step": 4613 }, { "epoch": 1.2277807344332092, "grad_norm": 0.29311856627464294, "learning_rate": 1.7499102052928537e-07, "loss": 0.2316, "step": 4614 }, { "epoch": 1.228046833422033, "grad_norm": 0.28903692960739136, "learning_rate": 1.749797329817345e-07, "loss": 0.209, "step": 4615 }, { "epoch": 1.2283129324108568, "grad_norm": 0.2944420576095581, "learning_rate": 1.7496844325171826e-07, "loss": 0.2247, "step": 4616 }, { "epoch": 1.2285790313996807, "grad_norm": 0.29463115334510803, "learning_rate": 1.7495715133956533e-07, "loss": 0.1882, "step": 4617 }, { "epoch": 1.2288451303885046, "grad_norm": 0.29948070645332336, "learning_rate": 1.7494585724560434e-07, "loss": 0.2094, "step": 4618 }, { "epoch": 1.2291112293773283, "grad_norm": 0.2861321270465851, "learning_rate": 1.7493456097016403e-07, "loss": 0.2028, "step": 4619 }, { "epoch": 1.2293773283661522, "grad_norm": 0.2997349202632904, "learning_rate": 1.749232625135732e-07, "loss": 0.2237, "step": 4620 }, { "epoch": 1.2296434273549761, "grad_norm": 0.26977819204330444, "learning_rate": 1.7491196187616076e-07, "loss": 0.202, "step": 4621 }, { "epoch": 1.2299095263437998, "grad_norm": 0.2891002595424652, "learning_rate": 1.749006590582556e-07, "loss": 0.2255, "step": 4622 }, { "epoch": 1.2301756253326237, "grad_norm": 0.304424911737442, "learning_rate": 1.7488935406018676e-07, "loss": 0.2254, "step": 4623 }, { "epoch": 1.2304417243214476, "grad_norm": 0.3108762800693512, "learning_rate": 1.7487804688228325e-07, "loss": 0.2314, "step": 4624 }, { "epoch": 1.2307078233102715, "grad_norm": 0.2762466073036194, "learning_rate": 1.748667375248742e-07, "loss": 0.2034, "step": 4625 }, { "epoch": 1.2309739222990952, "grad_norm": 0.29143664240837097, "learning_rate": 1.7485542598828884e-07, "loss": 0.218, "step": 4626 }, { "epoch": 1.231240021287919, "grad_norm": 0.29993948340415955, "learning_rate": 1.7484411227285637e-07, "loss": 0.2131, "step": 4627 }, { "epoch": 1.231506120276743, "grad_norm": 0.2730221450328827, "learning_rate": 1.7483279637890615e-07, "loss": 0.2137, "step": 4628 }, { "epoch": 1.231772219265567, "grad_norm": 0.3253190815448761, "learning_rate": 1.748214783067675e-07, "loss": 0.2331, "step": 4629 }, { "epoch": 1.2320383182543906, "grad_norm": 0.28452056646347046, "learning_rate": 1.748101580567699e-07, "loss": 0.2274, "step": 4630 }, { "epoch": 1.2323044172432145, "grad_norm": 0.29006054997444153, "learning_rate": 1.7479883562924282e-07, "loss": 0.2111, "step": 4631 }, { "epoch": 1.2325705162320384, "grad_norm": 0.299509197473526, "learning_rate": 1.7478751102451588e-07, "loss": 0.2071, "step": 4632 }, { "epoch": 1.232836615220862, "grad_norm": 0.31306418776512146, "learning_rate": 1.7477618424291868e-07, "loss": 0.2251, "step": 4633 }, { "epoch": 1.233102714209686, "grad_norm": 0.283622145652771, "learning_rate": 1.7476485528478091e-07, "loss": 0.2095, "step": 4634 }, { "epoch": 1.2333688131985099, "grad_norm": 0.2626267671585083, "learning_rate": 1.747535241504323e-07, "loss": 0.203, "step": 4635 }, { "epoch": 1.2336349121873336, "grad_norm": 0.2990768253803253, "learning_rate": 1.747421908402027e-07, "loss": 0.2165, "step": 4636 }, { "epoch": 1.2339010111761575, "grad_norm": 0.2820998430252075, "learning_rate": 1.74730855354422e-07, "loss": 0.208, "step": 4637 }, { "epoch": 1.2341671101649814, "grad_norm": 0.30726584792137146, "learning_rate": 1.7471951769342015e-07, "loss": 0.2142, "step": 4638 }, { "epoch": 1.2344332091538053, "grad_norm": 0.2954208254814148, "learning_rate": 1.7470817785752714e-07, "loss": 0.1845, "step": 4639 }, { "epoch": 1.2346993081426292, "grad_norm": 0.2815299928188324, "learning_rate": 1.7469683584707305e-07, "loss": 0.2128, "step": 4640 }, { "epoch": 1.2349654071314529, "grad_norm": 0.30274125933647156, "learning_rate": 1.74685491662388e-07, "loss": 0.2166, "step": 4641 }, { "epoch": 1.2352315061202768, "grad_norm": 0.30864235758781433, "learning_rate": 1.7467414530380223e-07, "loss": 0.238, "step": 4642 }, { "epoch": 1.2354976051091007, "grad_norm": 0.29971423745155334, "learning_rate": 1.7466279677164596e-07, "loss": 0.2288, "step": 4643 }, { "epoch": 1.2357637040979244, "grad_norm": 0.29355111718177795, "learning_rate": 1.7465144606624954e-07, "loss": 0.2065, "step": 4644 }, { "epoch": 1.2360298030867483, "grad_norm": 0.29050859808921814, "learning_rate": 1.7464009318794338e-07, "loss": 0.2188, "step": 4645 }, { "epoch": 1.2362959020755722, "grad_norm": 0.31746038794517517, "learning_rate": 1.7462873813705787e-07, "loss": 0.2389, "step": 4646 }, { "epoch": 1.2365620010643958, "grad_norm": 0.29151594638824463, "learning_rate": 1.746173809139236e-07, "loss": 0.2164, "step": 4647 }, { "epoch": 1.2368281000532197, "grad_norm": 0.2940707206726074, "learning_rate": 1.7460602151887112e-07, "loss": 0.2089, "step": 4648 }, { "epoch": 1.2370941990420437, "grad_norm": 0.2734266221523285, "learning_rate": 1.7459465995223104e-07, "loss": 0.2005, "step": 4649 }, { "epoch": 1.2373602980308676, "grad_norm": 0.28965890407562256, "learning_rate": 1.745832962143341e-07, "loss": 0.221, "step": 4650 }, { "epoch": 1.2376263970196912, "grad_norm": 0.27792689204216003, "learning_rate": 1.7457193030551104e-07, "loss": 0.2182, "step": 4651 }, { "epoch": 1.2378924960085151, "grad_norm": 0.2874142825603485, "learning_rate": 1.7456056222609276e-07, "loss": 0.2347, "step": 4652 }, { "epoch": 1.238158594997339, "grad_norm": 0.28463226556777954, "learning_rate": 1.7454919197641004e-07, "loss": 0.2165, "step": 4653 }, { "epoch": 1.238424693986163, "grad_norm": 0.28457698225975037, "learning_rate": 1.7453781955679396e-07, "loss": 0.2277, "step": 4654 }, { "epoch": 1.2386907929749866, "grad_norm": 0.26635295152664185, "learning_rate": 1.7452644496757548e-07, "loss": 0.1947, "step": 4655 }, { "epoch": 1.2389568919638105, "grad_norm": 0.31975874304771423, "learning_rate": 1.7451506820908567e-07, "loss": 0.2282, "step": 4656 }, { "epoch": 1.2392229909526344, "grad_norm": 0.2649420499801636, "learning_rate": 1.7450368928165574e-07, "loss": 0.1963, "step": 4657 }, { "epoch": 1.2394890899414581, "grad_norm": 0.3053910434246063, "learning_rate": 1.7449230818561682e-07, "loss": 0.2093, "step": 4658 }, { "epoch": 1.239755188930282, "grad_norm": 0.29314613342285156, "learning_rate": 1.7448092492130024e-07, "loss": 0.2412, "step": 4659 }, { "epoch": 1.240021287919106, "grad_norm": 0.30008894205093384, "learning_rate": 1.7446953948903733e-07, "loss": 0.2311, "step": 4660 }, { "epoch": 1.2402873869079298, "grad_norm": 0.3040171265602112, "learning_rate": 1.7445815188915947e-07, "loss": 0.2276, "step": 4661 }, { "epoch": 1.2405534858967535, "grad_norm": 0.2759995758533478, "learning_rate": 1.7444676212199817e-07, "loss": 0.2104, "step": 4662 }, { "epoch": 1.2408195848855774, "grad_norm": 0.286447137594223, "learning_rate": 1.744353701878849e-07, "loss": 0.211, "step": 4663 }, { "epoch": 1.2410856838744013, "grad_norm": 0.31295767426490784, "learning_rate": 1.7442397608715127e-07, "loss": 0.2208, "step": 4664 }, { "epoch": 1.2413517828632252, "grad_norm": 0.2859722673892975, "learning_rate": 1.7441257982012893e-07, "loss": 0.2166, "step": 4665 }, { "epoch": 1.241617881852049, "grad_norm": 0.31282809376716614, "learning_rate": 1.7440118138714964e-07, "loss": 0.2364, "step": 4666 }, { "epoch": 1.2418839808408728, "grad_norm": 0.27585092186927795, "learning_rate": 1.743897807885451e-07, "loss": 0.2142, "step": 4667 }, { "epoch": 1.2421500798296967, "grad_norm": 0.28754135966300964, "learning_rate": 1.743783780246472e-07, "loss": 0.225, "step": 4668 }, { "epoch": 1.2424161788185204, "grad_norm": 0.275351345539093, "learning_rate": 1.7436697309578786e-07, "loss": 0.2188, "step": 4669 }, { "epoch": 1.2426822778073443, "grad_norm": 0.29228630661964417, "learning_rate": 1.7435556600229902e-07, "loss": 0.2266, "step": 4670 }, { "epoch": 1.2429483767961682, "grad_norm": 0.2871767282485962, "learning_rate": 1.7434415674451268e-07, "loss": 0.2037, "step": 4671 }, { "epoch": 1.2432144757849921, "grad_norm": 0.27329397201538086, "learning_rate": 1.7433274532276098e-07, "loss": 0.207, "step": 4672 }, { "epoch": 1.2434805747738158, "grad_norm": 0.2707242965698242, "learning_rate": 1.743213317373761e-07, "loss": 0.1966, "step": 4673 }, { "epoch": 1.2437466737626397, "grad_norm": 0.3014738857746124, "learning_rate": 1.743099159886902e-07, "loss": 0.2135, "step": 4674 }, { "epoch": 1.2440127727514636, "grad_norm": 0.28451353311538696, "learning_rate": 1.742984980770356e-07, "loss": 0.2151, "step": 4675 }, { "epoch": 1.2442788717402875, "grad_norm": 0.2656688988208771, "learning_rate": 1.7428707800274465e-07, "loss": 0.2148, "step": 4676 }, { "epoch": 1.2445449707291112, "grad_norm": 0.28519973158836365, "learning_rate": 1.742756557661497e-07, "loss": 0.2379, "step": 4677 }, { "epoch": 1.244811069717935, "grad_norm": 0.28219905495643616, "learning_rate": 1.742642313675833e-07, "loss": 0.2361, "step": 4678 }, { "epoch": 1.245077168706759, "grad_norm": 0.26490628719329834, "learning_rate": 1.7425280480737796e-07, "loss": 0.2104, "step": 4679 }, { "epoch": 1.2453432676955827, "grad_norm": 0.2965129315853119, "learning_rate": 1.742413760858662e-07, "loss": 0.2212, "step": 4680 }, { "epoch": 1.2456093666844066, "grad_norm": 0.2608945667743683, "learning_rate": 1.7422994520338082e-07, "loss": 0.1713, "step": 4681 }, { "epoch": 1.2458754656732305, "grad_norm": 0.28846877813339233, "learning_rate": 1.7421851216025446e-07, "loss": 0.225, "step": 4682 }, { "epoch": 1.2461415646620542, "grad_norm": 0.2612246572971344, "learning_rate": 1.742070769568199e-07, "loss": 0.1926, "step": 4683 }, { "epoch": 1.246407663650878, "grad_norm": 0.27126601338386536, "learning_rate": 1.7419563959341002e-07, "loss": 0.205, "step": 4684 }, { "epoch": 1.246673762639702, "grad_norm": 0.3164907991886139, "learning_rate": 1.741842000703577e-07, "loss": 0.2038, "step": 4685 }, { "epoch": 1.2469398616285259, "grad_norm": 0.2880229651927948, "learning_rate": 1.7417275838799594e-07, "loss": 0.1823, "step": 4686 }, { "epoch": 1.2472059606173496, "grad_norm": 0.28258660435676575, "learning_rate": 1.741613145466578e-07, "loss": 0.2148, "step": 4687 }, { "epoch": 1.2474720596061735, "grad_norm": 0.2809467017650604, "learning_rate": 1.7414986854667631e-07, "loss": 0.2176, "step": 4688 }, { "epoch": 1.2477381585949974, "grad_norm": 0.2679360508918762, "learning_rate": 1.741384203883847e-07, "loss": 0.2069, "step": 4689 }, { "epoch": 1.2480042575838213, "grad_norm": 0.37602055072784424, "learning_rate": 1.7412697007211615e-07, "loss": 0.2194, "step": 4690 }, { "epoch": 1.248270356572645, "grad_norm": 0.2987140119075775, "learning_rate": 1.7411551759820399e-07, "loss": 0.2037, "step": 4691 }, { "epoch": 1.2485364555614689, "grad_norm": 0.32342931628227234, "learning_rate": 1.7410406296698152e-07, "loss": 0.226, "step": 4692 }, { "epoch": 1.2488025545502928, "grad_norm": 0.2920950651168823, "learning_rate": 1.7409260617878223e-07, "loss": 0.2096, "step": 4693 }, { "epoch": 1.2490686535391164, "grad_norm": 0.28742295503616333, "learning_rate": 1.7408114723393953e-07, "loss": 0.2152, "step": 4694 }, { "epoch": 1.2493347525279404, "grad_norm": 0.2842698395252228, "learning_rate": 1.7406968613278702e-07, "loss": 0.2058, "step": 4695 }, { "epoch": 1.2496008515167643, "grad_norm": 0.27807655930519104, "learning_rate": 1.740582228756582e-07, "loss": 0.2074, "step": 4696 }, { "epoch": 1.2498669505055882, "grad_norm": 0.30721408128738403, "learning_rate": 1.7404675746288684e-07, "loss": 0.214, "step": 4697 }, { "epoch": 1.2501330494944118, "grad_norm": 0.2959946393966675, "learning_rate": 1.7403528989480664e-07, "loss": 0.2225, "step": 4698 }, { "epoch": 1.2503991484832357, "grad_norm": 0.3084445297718048, "learning_rate": 1.7402382017175137e-07, "loss": 0.2105, "step": 4699 }, { "epoch": 1.2506652474720596, "grad_norm": 0.2860351502895355, "learning_rate": 1.740123482940549e-07, "loss": 0.2187, "step": 4700 }, { "epoch": 1.2509313464608836, "grad_norm": 0.288257360458374, "learning_rate": 1.7400087426205113e-07, "loss": 0.2346, "step": 4701 }, { "epoch": 1.2511974454497072, "grad_norm": 0.28032049536705017, "learning_rate": 1.7398939807607404e-07, "loss": 0.2404, "step": 4702 }, { "epoch": 1.2514635444385311, "grad_norm": 0.2943994998931885, "learning_rate": 1.739779197364577e-07, "loss": 0.2101, "step": 4703 }, { "epoch": 1.251729643427355, "grad_norm": 0.27699288725852966, "learning_rate": 1.7396643924353623e-07, "loss": 0.1931, "step": 4704 }, { "epoch": 1.2519957424161787, "grad_norm": 0.3243695795536041, "learning_rate": 1.739549565976437e-07, "loss": 0.198, "step": 4705 }, { "epoch": 1.2522618414050026, "grad_norm": 0.3275669515132904, "learning_rate": 1.7394347179911447e-07, "loss": 0.2102, "step": 4706 }, { "epoch": 1.2525279403938265, "grad_norm": 0.28559410572052, "learning_rate": 1.7393198484828274e-07, "loss": 0.2051, "step": 4707 }, { "epoch": 1.2527940393826502, "grad_norm": 0.3084040582180023, "learning_rate": 1.739204957454829e-07, "loss": 0.186, "step": 4708 }, { "epoch": 1.2530601383714741, "grad_norm": 0.28258639574050903, "learning_rate": 1.7390900449104934e-07, "loss": 0.2249, "step": 4709 }, { "epoch": 1.253326237360298, "grad_norm": 0.2871868908405304, "learning_rate": 1.7389751108531659e-07, "loss": 0.219, "step": 4710 }, { "epoch": 1.253592336349122, "grad_norm": 0.3122214376926422, "learning_rate": 1.7388601552861915e-07, "loss": 0.221, "step": 4711 }, { "epoch": 1.2538584353379458, "grad_norm": 0.2943148910999298, "learning_rate": 1.7387451782129162e-07, "loss": 0.2213, "step": 4712 }, { "epoch": 1.2541245343267695, "grad_norm": 0.29733410477638245, "learning_rate": 1.7386301796366873e-07, "loss": 0.2243, "step": 4713 }, { "epoch": 1.2543906333155934, "grad_norm": 0.27014869451522827, "learning_rate": 1.7385151595608516e-07, "loss": 0.1896, "step": 4714 }, { "epoch": 1.2546567323044173, "grad_norm": 0.2830713987350464, "learning_rate": 1.738400117988757e-07, "loss": 0.2099, "step": 4715 }, { "epoch": 1.254922831293241, "grad_norm": 0.2964029610157013, "learning_rate": 1.7382850549237524e-07, "loss": 0.2069, "step": 4716 }, { "epoch": 1.255188930282065, "grad_norm": 0.3092051148414612, "learning_rate": 1.7381699703691866e-07, "loss": 0.2194, "step": 4717 }, { "epoch": 1.2554550292708888, "grad_norm": 0.2733067274093628, "learning_rate": 1.7380548643284096e-07, "loss": 0.1929, "step": 4718 }, { "epoch": 1.2557211282597125, "grad_norm": 0.27049097418785095, "learning_rate": 1.737939736804772e-07, "loss": 0.2038, "step": 4719 }, { "epoch": 1.2559872272485364, "grad_norm": 0.30175188183784485, "learning_rate": 1.7378245878016246e-07, "loss": 0.2174, "step": 4720 }, { "epoch": 1.2562533262373603, "grad_norm": 0.3298170566558838, "learning_rate": 1.737709417322319e-07, "loss": 0.2456, "step": 4721 }, { "epoch": 1.2565194252261842, "grad_norm": 0.28074026107788086, "learning_rate": 1.737594225370208e-07, "loss": 0.209, "step": 4722 }, { "epoch": 1.256785524215008, "grad_norm": 0.272885799407959, "learning_rate": 1.7374790119486442e-07, "loss": 0.2082, "step": 4723 }, { "epoch": 1.2570516232038318, "grad_norm": 0.29371780157089233, "learning_rate": 1.737363777060981e-07, "loss": 0.2209, "step": 4724 }, { "epoch": 1.2573177221926557, "grad_norm": 0.33006298542022705, "learning_rate": 1.7372485207105732e-07, "loss": 0.2405, "step": 4725 }, { "epoch": 1.2575838211814796, "grad_norm": 0.30352672934532166, "learning_rate": 1.7371332429007747e-07, "loss": 0.2294, "step": 4726 }, { "epoch": 1.2578499201703033, "grad_norm": 0.287276953458786, "learning_rate": 1.7370179436349418e-07, "loss": 0.2209, "step": 4727 }, { "epoch": 1.2581160191591272, "grad_norm": 0.29340583086013794, "learning_rate": 1.73690262291643e-07, "loss": 0.2186, "step": 4728 }, { "epoch": 1.258382118147951, "grad_norm": 0.29168257117271423, "learning_rate": 1.7367872807485963e-07, "loss": 0.2139, "step": 4729 }, { "epoch": 1.2586482171367748, "grad_norm": 0.28228089213371277, "learning_rate": 1.736671917134798e-07, "loss": 0.2091, "step": 4730 }, { "epoch": 1.2589143161255987, "grad_norm": 0.2893063426017761, "learning_rate": 1.736556532078393e-07, "loss": 0.2312, "step": 4731 }, { "epoch": 1.2591804151144226, "grad_norm": 0.28773802518844604, "learning_rate": 1.7364411255827396e-07, "loss": 0.2229, "step": 4732 }, { "epoch": 1.2594465141032465, "grad_norm": 0.277315229177475, "learning_rate": 1.736325697651197e-07, "loss": 0.2016, "step": 4733 }, { "epoch": 1.2597126130920704, "grad_norm": 0.28336042165756226, "learning_rate": 1.7362102482871258e-07, "loss": 0.2263, "step": 4734 }, { "epoch": 1.259978712080894, "grad_norm": 0.2820166051387787, "learning_rate": 1.736094777493885e-07, "loss": 0.2045, "step": 4735 }, { "epoch": 1.260244811069718, "grad_norm": 0.284782737493515, "learning_rate": 1.7359792852748373e-07, "loss": 0.2127, "step": 4736 }, { "epoch": 1.2605109100585419, "grad_norm": 0.28166770935058594, "learning_rate": 1.735863771633343e-07, "loss": 0.202, "step": 4737 }, { "epoch": 1.2607770090473656, "grad_norm": 0.371104896068573, "learning_rate": 1.7357482365727653e-07, "loss": 0.212, "step": 4738 }, { "epoch": 1.2610431080361895, "grad_norm": 0.27108943462371826, "learning_rate": 1.7356326800964664e-07, "loss": 0.2102, "step": 4739 }, { "epoch": 1.2613092070250134, "grad_norm": 0.2905745506286621, "learning_rate": 1.7355171022078107e-07, "loss": 0.2017, "step": 4740 }, { "epoch": 1.261575306013837, "grad_norm": 0.29594361782073975, "learning_rate": 1.735401502910162e-07, "loss": 0.2208, "step": 4741 }, { "epoch": 1.261841405002661, "grad_norm": 0.27501827478408813, "learning_rate": 1.7352858822068843e-07, "loss": 0.2098, "step": 4742 }, { "epoch": 1.2621075039914849, "grad_norm": 0.2964693307876587, "learning_rate": 1.7351702401013442e-07, "loss": 0.24, "step": 4743 }, { "epoch": 1.2623736029803085, "grad_norm": 0.2548207640647888, "learning_rate": 1.735054576596907e-07, "loss": 0.1857, "step": 4744 }, { "epoch": 1.2626397019691324, "grad_norm": 0.31368497014045715, "learning_rate": 1.73493889169694e-07, "loss": 0.2282, "step": 4745 }, { "epoch": 1.2629058009579563, "grad_norm": 0.28603652119636536, "learning_rate": 1.7348231854048096e-07, "loss": 0.2179, "step": 4746 }, { "epoch": 1.2631718999467803, "grad_norm": 0.29984813928604126, "learning_rate": 1.7347074577238847e-07, "loss": 0.2199, "step": 4747 }, { "epoch": 1.2634379989356042, "grad_norm": 0.2814076542854309, "learning_rate": 1.734591708657533e-07, "loss": 0.2035, "step": 4748 }, { "epoch": 1.2637040979244278, "grad_norm": 0.28661248087882996, "learning_rate": 1.7344759382091242e-07, "loss": 0.2183, "step": 4749 }, { "epoch": 1.2639701969132517, "grad_norm": 0.2649727165699005, "learning_rate": 1.7343601463820281e-07, "loss": 0.1942, "step": 4750 }, { "epoch": 1.2642362959020756, "grad_norm": 0.29669153690338135, "learning_rate": 1.7342443331796147e-07, "loss": 0.2331, "step": 4751 }, { "epoch": 1.2645023948908993, "grad_norm": 0.27855467796325684, "learning_rate": 1.7341284986052552e-07, "loss": 0.1956, "step": 4752 }, { "epoch": 1.2647684938797232, "grad_norm": 0.24942785501480103, "learning_rate": 1.734012642662321e-07, "loss": 0.1877, "step": 4753 }, { "epoch": 1.2650345928685471, "grad_norm": 0.28373080492019653, "learning_rate": 1.7338967653541848e-07, "loss": 0.2207, "step": 4754 }, { "epoch": 1.2653006918573708, "grad_norm": 0.2707921266555786, "learning_rate": 1.7337808666842194e-07, "loss": 0.2064, "step": 4755 }, { "epoch": 1.2655667908461947, "grad_norm": 0.29309338331222534, "learning_rate": 1.733664946655798e-07, "loss": 0.2212, "step": 4756 }, { "epoch": 1.2658328898350186, "grad_norm": 0.28914743661880493, "learning_rate": 1.7335490052722953e-07, "loss": 0.2119, "step": 4757 }, { "epoch": 1.2660989888238425, "grad_norm": 0.2736097574234009, "learning_rate": 1.7334330425370856e-07, "loss": 0.2154, "step": 4758 }, { "epoch": 1.2663650878126664, "grad_norm": 0.28999239206314087, "learning_rate": 1.7333170584535441e-07, "loss": 0.2097, "step": 4759 }, { "epoch": 1.2666311868014901, "grad_norm": 0.2861192226409912, "learning_rate": 1.7332010530250472e-07, "loss": 0.2019, "step": 4760 }, { "epoch": 1.266897285790314, "grad_norm": 0.31880316138267517, "learning_rate": 1.7330850262549714e-07, "loss": 0.239, "step": 4761 }, { "epoch": 1.267163384779138, "grad_norm": 0.28147247433662415, "learning_rate": 1.732968978146694e-07, "loss": 0.2165, "step": 4762 }, { "epoch": 1.2674294837679616, "grad_norm": 0.2717115879058838, "learning_rate": 1.7328529087035925e-07, "loss": 0.209, "step": 4763 }, { "epoch": 1.2676955827567855, "grad_norm": 0.29468584060668945, "learning_rate": 1.732736817929046e-07, "loss": 0.2257, "step": 4764 }, { "epoch": 1.2679616817456094, "grad_norm": 0.283414363861084, "learning_rate": 1.7326207058264333e-07, "loss": 0.2062, "step": 4765 }, { "epoch": 1.268227780734433, "grad_norm": 0.2940208911895752, "learning_rate": 1.7325045723991338e-07, "loss": 0.2283, "step": 4766 }, { "epoch": 1.268493879723257, "grad_norm": 0.30200478434562683, "learning_rate": 1.7323884176505282e-07, "loss": 0.2024, "step": 4767 }, { "epoch": 1.268759978712081, "grad_norm": 0.31947189569473267, "learning_rate": 1.7322722415839975e-07, "loss": 0.218, "step": 4768 }, { "epoch": 1.2690260777009048, "grad_norm": 0.2814148962497711, "learning_rate": 1.732156044202923e-07, "loss": 0.2074, "step": 4769 }, { "epoch": 1.2692921766897287, "grad_norm": 0.307630330324173, "learning_rate": 1.7320398255106874e-07, "loss": 0.2146, "step": 4770 }, { "epoch": 1.2695582756785524, "grad_norm": 0.30801495909690857, "learning_rate": 1.7319235855106728e-07, "loss": 0.2549, "step": 4771 }, { "epoch": 1.2698243746673763, "grad_norm": 0.27428263425827026, "learning_rate": 1.7318073242062635e-07, "loss": 0.2196, "step": 4772 }, { "epoch": 1.2700904736562002, "grad_norm": 0.2777843773365021, "learning_rate": 1.731691041600843e-07, "loss": 0.201, "step": 4773 }, { "epoch": 1.2703565726450239, "grad_norm": 0.3118271827697754, "learning_rate": 1.7315747376977956e-07, "loss": 0.214, "step": 4774 }, { "epoch": 1.2706226716338478, "grad_norm": 0.3090323805809021, "learning_rate": 1.7314584125005073e-07, "loss": 0.2141, "step": 4775 }, { "epoch": 1.2708887706226717, "grad_norm": 0.2893133759498596, "learning_rate": 1.731342066012364e-07, "loss": 0.2321, "step": 4776 }, { "epoch": 1.2711548696114954, "grad_norm": 0.2965771555900574, "learning_rate": 1.7312256982367523e-07, "loss": 0.2183, "step": 4777 }, { "epoch": 1.2714209686003193, "grad_norm": 0.27922144532203674, "learning_rate": 1.7311093091770586e-07, "loss": 0.2157, "step": 4778 }, { "epoch": 1.2716870675891432, "grad_norm": 0.2842504680156708, "learning_rate": 1.7309928988366715e-07, "loss": 0.2019, "step": 4779 }, { "epoch": 1.2719531665779669, "grad_norm": 0.26514607667922974, "learning_rate": 1.7308764672189792e-07, "loss": 0.1941, "step": 4780 }, { "epoch": 1.2722192655667908, "grad_norm": 0.3118060827255249, "learning_rate": 1.7307600143273707e-07, "loss": 0.2286, "step": 4781 }, { "epoch": 1.2724853645556147, "grad_norm": 0.4296962320804596, "learning_rate": 1.7306435401652353e-07, "loss": 0.1838, "step": 4782 }, { "epoch": 1.2727514635444386, "grad_norm": 0.29432639479637146, "learning_rate": 1.7305270447359636e-07, "loss": 0.1943, "step": 4783 }, { "epoch": 1.2730175625332625, "grad_norm": 0.2960834801197052, "learning_rate": 1.7304105280429465e-07, "loss": 0.2064, "step": 4784 }, { "epoch": 1.2732836615220862, "grad_norm": 0.2722938060760498, "learning_rate": 1.7302939900895753e-07, "loss": 0.2188, "step": 4785 }, { "epoch": 1.27354976051091, "grad_norm": 0.2843971252441406, "learning_rate": 1.7301774308792426e-07, "loss": 0.2242, "step": 4786 }, { "epoch": 1.273815859499734, "grad_norm": 0.27440935373306274, "learning_rate": 1.7300608504153405e-07, "loss": 0.2059, "step": 4787 }, { "epoch": 1.2740819584885577, "grad_norm": 0.30088356137275696, "learning_rate": 1.7299442487012626e-07, "loss": 0.2208, "step": 4788 }, { "epoch": 1.2743480574773816, "grad_norm": 0.28294509649276733, "learning_rate": 1.7298276257404028e-07, "loss": 0.2112, "step": 4789 }, { "epoch": 1.2746141564662055, "grad_norm": 0.2960965037345886, "learning_rate": 1.7297109815361562e-07, "loss": 0.222, "step": 4790 }, { "epoch": 1.2748802554550291, "grad_norm": 0.30115824937820435, "learning_rate": 1.7295943160919172e-07, "loss": 0.2335, "step": 4791 }, { "epoch": 1.275146354443853, "grad_norm": 0.2894706726074219, "learning_rate": 1.7294776294110822e-07, "loss": 0.2129, "step": 4792 }, { "epoch": 1.275412453432677, "grad_norm": 0.28726962208747864, "learning_rate": 1.7293609214970476e-07, "loss": 0.2245, "step": 4793 }, { "epoch": 1.2756785524215009, "grad_norm": 0.282061368227005, "learning_rate": 1.72924419235321e-07, "loss": 0.1969, "step": 4794 }, { "epoch": 1.2759446514103248, "grad_norm": 0.33073803782463074, "learning_rate": 1.7291274419829674e-07, "loss": 0.2034, "step": 4795 }, { "epoch": 1.2762107503991484, "grad_norm": 0.2949589788913727, "learning_rate": 1.7290106703897185e-07, "loss": 0.2162, "step": 4796 }, { "epoch": 1.2764768493879723, "grad_norm": 0.29937461018562317, "learning_rate": 1.728893877576862e-07, "loss": 0.2209, "step": 4797 }, { "epoch": 1.2767429483767962, "grad_norm": 0.29139742255210876, "learning_rate": 1.7287770635477968e-07, "loss": 0.233, "step": 4798 }, { "epoch": 1.27700904736562, "grad_norm": 0.3207593560218811, "learning_rate": 1.7286602283059236e-07, "loss": 0.213, "step": 4799 }, { "epoch": 1.2772751463544438, "grad_norm": 0.29964402318000793, "learning_rate": 1.7285433718546432e-07, "loss": 0.2277, "step": 4800 }, { "epoch": 1.2775412453432677, "grad_norm": 0.257506400346756, "learning_rate": 1.728426494197357e-07, "loss": 0.1925, "step": 4801 }, { "epoch": 1.2778073443320914, "grad_norm": 0.29180440306663513, "learning_rate": 1.7283095953374667e-07, "loss": 0.2322, "step": 4802 }, { "epoch": 1.2780734433209153, "grad_norm": 0.2642928957939148, "learning_rate": 1.7281926752783752e-07, "loss": 0.2025, "step": 4803 }, { "epoch": 1.2783395423097392, "grad_norm": 0.31318944692611694, "learning_rate": 1.7280757340234857e-07, "loss": 0.2071, "step": 4804 }, { "epoch": 1.2786056412985631, "grad_norm": 0.28746694326400757, "learning_rate": 1.7279587715762022e-07, "loss": 0.2176, "step": 4805 }, { "epoch": 1.278871740287387, "grad_norm": 0.2873067855834961, "learning_rate": 1.7278417879399286e-07, "loss": 0.2217, "step": 4806 }, { "epoch": 1.2791378392762107, "grad_norm": 0.2747286260128021, "learning_rate": 1.7277247831180706e-07, "loss": 0.1999, "step": 4807 }, { "epoch": 1.2794039382650346, "grad_norm": 0.33111029863357544, "learning_rate": 1.7276077571140336e-07, "loss": 0.2106, "step": 4808 }, { "epoch": 1.2796700372538585, "grad_norm": 0.26644447445869446, "learning_rate": 1.7274907099312238e-07, "loss": 0.208, "step": 4809 }, { "epoch": 1.2799361362426822, "grad_norm": 0.30968213081359863, "learning_rate": 1.7273736415730486e-07, "loss": 0.2249, "step": 4810 }, { "epoch": 1.2802022352315061, "grad_norm": 0.28478220105171204, "learning_rate": 1.7272565520429154e-07, "loss": 0.2157, "step": 4811 }, { "epoch": 1.28046833422033, "grad_norm": 0.2684909403324127, "learning_rate": 1.727139441344232e-07, "loss": 0.1944, "step": 4812 }, { "epoch": 1.2807344332091537, "grad_norm": 0.27143150568008423, "learning_rate": 1.7270223094804077e-07, "loss": 0.1918, "step": 4813 }, { "epoch": 1.2810005321979776, "grad_norm": 0.3018171787261963, "learning_rate": 1.7269051564548515e-07, "loss": 0.2309, "step": 4814 }, { "epoch": 1.2812666311868015, "grad_norm": 0.2984880805015564, "learning_rate": 1.7267879822709736e-07, "loss": 0.209, "step": 4815 }, { "epoch": 1.2815327301756254, "grad_norm": 0.3132937550544739, "learning_rate": 1.7266707869321847e-07, "loss": 0.2061, "step": 4816 }, { "epoch": 1.281798829164449, "grad_norm": 0.2933313548564911, "learning_rate": 1.7265535704418958e-07, "loss": 0.2081, "step": 4817 }, { "epoch": 1.282064928153273, "grad_norm": 0.2779262661933899, "learning_rate": 1.7264363328035193e-07, "loss": 0.2052, "step": 4818 }, { "epoch": 1.282331027142097, "grad_norm": 0.2753312587738037, "learning_rate": 1.7263190740204672e-07, "loss": 0.2067, "step": 4819 }, { "epoch": 1.2825971261309208, "grad_norm": 0.284081369638443, "learning_rate": 1.7262017940961524e-07, "loss": 0.2147, "step": 4820 }, { "epoch": 1.2828632251197445, "grad_norm": 0.2797785699367523, "learning_rate": 1.7260844930339894e-07, "loss": 0.2091, "step": 4821 }, { "epoch": 1.2831293241085684, "grad_norm": 0.251736581325531, "learning_rate": 1.7259671708373919e-07, "loss": 0.1912, "step": 4822 }, { "epoch": 1.2833954230973923, "grad_norm": 0.27247247099876404, "learning_rate": 1.7258498275097745e-07, "loss": 0.2126, "step": 4823 }, { "epoch": 1.283661522086216, "grad_norm": 0.2900727093219757, "learning_rate": 1.725732463054554e-07, "loss": 0.1962, "step": 4824 }, { "epoch": 1.2839276210750399, "grad_norm": 0.2891381084918976, "learning_rate": 1.7256150774751454e-07, "loss": 0.2214, "step": 4825 }, { "epoch": 1.2841937200638638, "grad_norm": 0.28944146633148193, "learning_rate": 1.7254976707749663e-07, "loss": 0.2178, "step": 4826 }, { "epoch": 1.2844598190526875, "grad_norm": 0.2749071419239044, "learning_rate": 1.7253802429574337e-07, "loss": 0.1946, "step": 4827 }, { "epoch": 1.2847259180415114, "grad_norm": 0.2823864221572876, "learning_rate": 1.7252627940259653e-07, "loss": 0.2024, "step": 4828 }, { "epoch": 1.2849920170303353, "grad_norm": 0.2696758210659027, "learning_rate": 1.7251453239839803e-07, "loss": 0.2012, "step": 4829 }, { "epoch": 1.2852581160191592, "grad_norm": 0.2586912214756012, "learning_rate": 1.7250278328348978e-07, "loss": 0.1996, "step": 4830 }, { "epoch": 1.285524215007983, "grad_norm": 0.2824300229549408, "learning_rate": 1.7249103205821376e-07, "loss": 0.2076, "step": 4831 }, { "epoch": 1.2857903139968068, "grad_norm": 0.2819826006889343, "learning_rate": 1.72479278722912e-07, "loss": 0.2093, "step": 4832 }, { "epoch": 1.2860564129856307, "grad_norm": 0.2870686650276184, "learning_rate": 1.7246752327792665e-07, "loss": 0.2247, "step": 4833 }, { "epoch": 1.2863225119744546, "grad_norm": 0.3154757022857666, "learning_rate": 1.7245576572359983e-07, "loss": 0.239, "step": 4834 }, { "epoch": 1.2865886109632783, "grad_norm": 0.3230084776878357, "learning_rate": 1.724440060602738e-07, "loss": 0.218, "step": 4835 }, { "epoch": 1.2868547099521022, "grad_norm": 0.2657181918621063, "learning_rate": 1.724322442882909e-07, "loss": 0.2053, "step": 4836 }, { "epoch": 1.287120808940926, "grad_norm": 0.28621944785118103, "learning_rate": 1.7242048040799338e-07, "loss": 0.2272, "step": 4837 }, { "epoch": 1.2873869079297497, "grad_norm": 0.30845293402671814, "learning_rate": 1.7240871441972376e-07, "loss": 0.2162, "step": 4838 }, { "epoch": 1.2876530069185737, "grad_norm": 0.29626625776290894, "learning_rate": 1.7239694632382442e-07, "loss": 0.2017, "step": 4839 }, { "epoch": 1.2879191059073976, "grad_norm": 0.28207656741142273, "learning_rate": 1.72385176120638e-07, "loss": 0.1958, "step": 4840 }, { "epoch": 1.2881852048962215, "grad_norm": 0.29482999444007874, "learning_rate": 1.72373403810507e-07, "loss": 0.2245, "step": 4841 }, { "epoch": 1.2884513038850454, "grad_norm": 0.269930899143219, "learning_rate": 1.7236162939377418e-07, "loss": 0.216, "step": 4842 }, { "epoch": 1.288717402873869, "grad_norm": 0.29415759444236755, "learning_rate": 1.7234985287078216e-07, "loss": 0.2141, "step": 4843 }, { "epoch": 1.288983501862693, "grad_norm": 0.3072662055492401, "learning_rate": 1.7233807424187383e-07, "loss": 0.2287, "step": 4844 }, { "epoch": 1.2892496008515169, "grad_norm": 0.2746933400630951, "learning_rate": 1.7232629350739196e-07, "loss": 0.192, "step": 4845 }, { "epoch": 1.2895156998403405, "grad_norm": 0.29810452461242676, "learning_rate": 1.7231451066767948e-07, "loss": 0.2093, "step": 4846 }, { "epoch": 1.2897817988291644, "grad_norm": 0.28556329011917114, "learning_rate": 1.7230272572307935e-07, "loss": 0.2193, "step": 4847 }, { "epoch": 1.2900478978179883, "grad_norm": 0.407229483127594, "learning_rate": 1.7229093867393462e-07, "loss": 0.2406, "step": 4848 }, { "epoch": 1.290313996806812, "grad_norm": 0.2930578887462616, "learning_rate": 1.7227914952058837e-07, "loss": 0.2268, "step": 4849 }, { "epoch": 1.290580095795636, "grad_norm": 0.2633035182952881, "learning_rate": 1.7226735826338372e-07, "loss": 0.1881, "step": 4850 }, { "epoch": 1.2908461947844598, "grad_norm": 0.302826464176178, "learning_rate": 1.7225556490266394e-07, "loss": 0.2379, "step": 4851 }, { "epoch": 1.2911122937732837, "grad_norm": 0.26855847239494324, "learning_rate": 1.7224376943877225e-07, "loss": 0.2055, "step": 4852 }, { "epoch": 1.2913783927621076, "grad_norm": 0.2951742112636566, "learning_rate": 1.7223197187205205e-07, "loss": 0.2108, "step": 4853 }, { "epoch": 1.2916444917509313, "grad_norm": 0.2822885811328888, "learning_rate": 1.722201722028467e-07, "loss": 0.2031, "step": 4854 }, { "epoch": 1.2919105907397552, "grad_norm": 0.28887948393821716, "learning_rate": 1.722083704314996e-07, "loss": 0.2017, "step": 4855 }, { "epoch": 1.2921766897285791, "grad_norm": 0.3067217469215393, "learning_rate": 1.7219656655835436e-07, "loss": 0.2187, "step": 4856 }, { "epoch": 1.2924427887174028, "grad_norm": 0.28915804624557495, "learning_rate": 1.7218476058375452e-07, "loss": 0.2097, "step": 4857 }, { "epoch": 1.2927088877062267, "grad_norm": 0.2706346809864044, "learning_rate": 1.7217295250804372e-07, "loss": 0.2055, "step": 4858 }, { "epoch": 1.2929749866950506, "grad_norm": 0.3054967224597931, "learning_rate": 1.7216114233156564e-07, "loss": 0.24, "step": 4859 }, { "epoch": 1.2932410856838743, "grad_norm": 0.2777320444583893, "learning_rate": 1.721493300546641e-07, "loss": 0.1952, "step": 4860 }, { "epoch": 1.2935071846726982, "grad_norm": 0.2704888880252838, "learning_rate": 1.7213751567768294e-07, "loss": 0.2053, "step": 4861 }, { "epoch": 1.2937732836615221, "grad_norm": 0.27481743693351746, "learning_rate": 1.7212569920096593e-07, "loss": 0.2061, "step": 4862 }, { "epoch": 1.2940393826503458, "grad_norm": 0.2900649607181549, "learning_rate": 1.7211388062485712e-07, "loss": 0.1955, "step": 4863 }, { "epoch": 1.2943054816391697, "grad_norm": 0.3209651708602905, "learning_rate": 1.7210205994970048e-07, "loss": 0.2022, "step": 4864 }, { "epoch": 1.2945715806279936, "grad_norm": 0.2823769152164459, "learning_rate": 1.720902371758401e-07, "loss": 0.205, "step": 4865 }, { "epoch": 1.2948376796168175, "grad_norm": 0.2883530259132385, "learning_rate": 1.7207841230362007e-07, "loss": 0.2168, "step": 4866 }, { "epoch": 1.2951037786056414, "grad_norm": 0.30646002292633057, "learning_rate": 1.7206658533338466e-07, "loss": 0.2141, "step": 4867 }, { "epoch": 1.295369877594465, "grad_norm": 0.2694377899169922, "learning_rate": 1.72054756265478e-07, "loss": 0.21, "step": 4868 }, { "epoch": 1.295635976583289, "grad_norm": 0.2943260967731476, "learning_rate": 1.7204292510024452e-07, "loss": 0.2109, "step": 4869 }, { "epoch": 1.295902075572113, "grad_norm": 0.2693694233894348, "learning_rate": 1.7203109183802854e-07, "loss": 0.1948, "step": 4870 }, { "epoch": 1.2961681745609366, "grad_norm": 0.2899742126464844, "learning_rate": 1.7201925647917448e-07, "loss": 0.1989, "step": 4871 }, { "epoch": 1.2964342735497605, "grad_norm": 0.3009961247444153, "learning_rate": 1.7200741902402688e-07, "loss": 0.2433, "step": 4872 }, { "epoch": 1.2967003725385844, "grad_norm": 0.26455649733543396, "learning_rate": 1.719955794729303e-07, "loss": 0.2045, "step": 4873 }, { "epoch": 1.296966471527408, "grad_norm": 0.3073127269744873, "learning_rate": 1.719837378262293e-07, "loss": 0.2209, "step": 4874 }, { "epoch": 1.297232570516232, "grad_norm": 0.26989424228668213, "learning_rate": 1.719718940842686e-07, "loss": 0.1976, "step": 4875 }, { "epoch": 1.2974986695050559, "grad_norm": 0.27815011143684387, "learning_rate": 1.7196004824739294e-07, "loss": 0.2264, "step": 4876 }, { "epoch": 1.2977647684938798, "grad_norm": 0.2877812087535858, "learning_rate": 1.7194820031594714e-07, "loss": 0.2026, "step": 4877 }, { "epoch": 1.2980308674827037, "grad_norm": 0.2800963521003723, "learning_rate": 1.7193635029027603e-07, "loss": 0.2216, "step": 4878 }, { "epoch": 1.2982969664715274, "grad_norm": 0.2794320583343506, "learning_rate": 1.7192449817072452e-07, "loss": 0.2085, "step": 4879 }, { "epoch": 1.2985630654603513, "grad_norm": 0.273375928401947, "learning_rate": 1.7191264395763762e-07, "loss": 0.1954, "step": 4880 }, { "epoch": 1.2988291644491752, "grad_norm": 0.31591981649398804, "learning_rate": 1.719007876513604e-07, "loss": 0.2121, "step": 4881 }, { "epoch": 1.2990952634379989, "grad_norm": 0.26478537917137146, "learning_rate": 1.7188892925223793e-07, "loss": 0.2085, "step": 4882 }, { "epoch": 1.2993613624268228, "grad_norm": 0.2901550829410553, "learning_rate": 1.718770687606154e-07, "loss": 0.2107, "step": 4883 }, { "epoch": 1.2996274614156467, "grad_norm": 0.3013346791267395, "learning_rate": 1.71865206176838e-07, "loss": 0.2215, "step": 4884 }, { "epoch": 1.2998935604044703, "grad_norm": 0.2680404484272003, "learning_rate": 1.7185334150125106e-07, "loss": 0.2044, "step": 4885 }, { "epoch": 1.3001596593932943, "grad_norm": 0.2737624943256378, "learning_rate": 1.718414747341999e-07, "loss": 0.2031, "step": 4886 }, { "epoch": 1.3004257583821182, "grad_norm": 0.2884068787097931, "learning_rate": 1.7182960587602998e-07, "loss": 0.2029, "step": 4887 }, { "epoch": 1.300691857370942, "grad_norm": 0.27236849069595337, "learning_rate": 1.7181773492708672e-07, "loss": 0.1972, "step": 4888 }, { "epoch": 1.300957956359766, "grad_norm": 0.28518393635749817, "learning_rate": 1.7180586188771562e-07, "loss": 0.1925, "step": 4889 }, { "epoch": 1.3012240553485896, "grad_norm": 0.27034661173820496, "learning_rate": 1.7179398675826237e-07, "loss": 0.2095, "step": 4890 }, { "epoch": 1.3014901543374136, "grad_norm": 0.28860852122306824, "learning_rate": 1.7178210953907257e-07, "loss": 0.224, "step": 4891 }, { "epoch": 1.3017562533262375, "grad_norm": 0.28900986909866333, "learning_rate": 1.7177023023049195e-07, "loss": 0.2146, "step": 4892 }, { "epoch": 1.3020223523150611, "grad_norm": 0.28704240918159485, "learning_rate": 1.7175834883286626e-07, "loss": 0.2056, "step": 4893 }, { "epoch": 1.302288451303885, "grad_norm": 0.270741730928421, "learning_rate": 1.7174646534654138e-07, "loss": 0.2273, "step": 4894 }, { "epoch": 1.302554550292709, "grad_norm": 0.28331783413887024, "learning_rate": 1.7173457977186314e-07, "loss": 0.2045, "step": 4895 }, { "epoch": 1.3028206492815326, "grad_norm": 0.28937333822250366, "learning_rate": 1.7172269210917757e-07, "loss": 0.2026, "step": 4896 }, { "epoch": 1.3030867482703565, "grad_norm": 0.2548298239707947, "learning_rate": 1.7171080235883068e-07, "loss": 0.1945, "step": 4897 }, { "epoch": 1.3033528472591804, "grad_norm": 0.28526803851127625, "learning_rate": 1.716989105211685e-07, "loss": 0.2162, "step": 4898 }, { "epoch": 1.3036189462480043, "grad_norm": 0.2843574285507202, "learning_rate": 1.7168701659653717e-07, "loss": 0.1953, "step": 4899 }, { "epoch": 1.303885045236828, "grad_norm": 0.26726287603378296, "learning_rate": 1.7167512058528298e-07, "loss": 0.2035, "step": 4900 }, { "epoch": 1.304151144225652, "grad_norm": 0.28229033946990967, "learning_rate": 1.716632224877521e-07, "loss": 0.2209, "step": 4901 }, { "epoch": 1.3044172432144758, "grad_norm": 0.2935076951980591, "learning_rate": 1.7165132230429088e-07, "loss": 0.1992, "step": 4902 }, { "epoch": 1.3046833422032997, "grad_norm": 0.26962241530418396, "learning_rate": 1.716394200352457e-07, "loss": 0.2141, "step": 4903 }, { "epoch": 1.3049494411921234, "grad_norm": 0.285873681306839, "learning_rate": 1.7162751568096305e-07, "loss": 0.2117, "step": 4904 }, { "epoch": 1.3052155401809473, "grad_norm": 0.2691250443458557, "learning_rate": 1.7161560924178937e-07, "loss": 0.213, "step": 4905 }, { "epoch": 1.3054816391697712, "grad_norm": 0.25742587447166443, "learning_rate": 1.7160370071807126e-07, "loss": 0.1955, "step": 4906 }, { "epoch": 1.305747738158595, "grad_norm": 0.254528671503067, "learning_rate": 1.7159179011015534e-07, "loss": 0.2057, "step": 4907 }, { "epoch": 1.3060138371474188, "grad_norm": 0.2711280584335327, "learning_rate": 1.715798774183883e-07, "loss": 0.2028, "step": 4908 }, { "epoch": 1.3062799361362427, "grad_norm": 0.28257864713668823, "learning_rate": 1.7156796264311687e-07, "loss": 0.1951, "step": 4909 }, { "epoch": 1.3065460351250664, "grad_norm": 0.2523208260536194, "learning_rate": 1.715560457846879e-07, "loss": 0.1923, "step": 4910 }, { "epoch": 1.3068121341138903, "grad_norm": 0.27162736654281616, "learning_rate": 1.715441268434482e-07, "loss": 0.1966, "step": 4911 }, { "epoch": 1.3070782331027142, "grad_norm": 0.29711082577705383, "learning_rate": 1.7153220581974475e-07, "loss": 0.2263, "step": 4912 }, { "epoch": 1.307344332091538, "grad_norm": 0.280815064907074, "learning_rate": 1.715202827139245e-07, "loss": 0.2009, "step": 4913 }, { "epoch": 1.307610431080362, "grad_norm": 0.31213659048080444, "learning_rate": 1.7150835752633455e-07, "loss": 0.2224, "step": 4914 }, { "epoch": 1.3078765300691857, "grad_norm": 0.278503954410553, "learning_rate": 1.7149643025732195e-07, "loss": 0.1993, "step": 4915 }, { "epoch": 1.3081426290580096, "grad_norm": 0.297965943813324, "learning_rate": 1.7148450090723392e-07, "loss": 0.223, "step": 4916 }, { "epoch": 1.3084087280468335, "grad_norm": 0.2930034399032593, "learning_rate": 1.714725694764177e-07, "loss": 0.2119, "step": 4917 }, { "epoch": 1.3086748270356572, "grad_norm": 0.28963106870651245, "learning_rate": 1.7146063596522052e-07, "loss": 0.2085, "step": 4918 }, { "epoch": 1.308940926024481, "grad_norm": 0.27547603845596313, "learning_rate": 1.7144870037398978e-07, "loss": 0.2118, "step": 4919 }, { "epoch": 1.309207025013305, "grad_norm": 0.27147820591926575, "learning_rate": 1.7143676270307288e-07, "loss": 0.2049, "step": 4920 }, { "epoch": 1.3094731240021287, "grad_norm": 0.2829743027687073, "learning_rate": 1.7142482295281733e-07, "loss": 0.2164, "step": 4921 }, { "epoch": 1.3097392229909526, "grad_norm": 0.2946065664291382, "learning_rate": 1.714128811235706e-07, "loss": 0.2177, "step": 4922 }, { "epoch": 1.3100053219797765, "grad_norm": 0.2872174084186554, "learning_rate": 1.7140093721568037e-07, "loss": 0.227, "step": 4923 }, { "epoch": 1.3102714209686004, "grad_norm": 0.2823341488838196, "learning_rate": 1.7138899122949423e-07, "loss": 0.2326, "step": 4924 }, { "epoch": 1.3105375199574243, "grad_norm": 0.31656941771507263, "learning_rate": 1.7137704316535987e-07, "loss": 0.2291, "step": 4925 }, { "epoch": 1.310803618946248, "grad_norm": 0.268819659948349, "learning_rate": 1.7136509302362517e-07, "loss": 0.2083, "step": 4926 }, { "epoch": 1.3110697179350719, "grad_norm": 0.27087265253067017, "learning_rate": 1.713531408046379e-07, "loss": 0.1815, "step": 4927 }, { "epoch": 1.3113358169238958, "grad_norm": 0.27654850482940674, "learning_rate": 1.7134118650874597e-07, "loss": 0.2093, "step": 4928 }, { "epoch": 1.3116019159127195, "grad_norm": 0.31968435645103455, "learning_rate": 1.713292301362973e-07, "loss": 0.2215, "step": 4929 }, { "epoch": 1.3118680149015434, "grad_norm": 0.28515318036079407, "learning_rate": 1.7131727168763996e-07, "loss": 0.2014, "step": 4930 }, { "epoch": 1.3121341138903673, "grad_norm": 0.31351426243782043, "learning_rate": 1.71305311163122e-07, "loss": 0.2026, "step": 4931 }, { "epoch": 1.312400212879191, "grad_norm": 0.2750460207462311, "learning_rate": 1.7129334856309164e-07, "loss": 0.2126, "step": 4932 }, { "epoch": 1.3126663118680149, "grad_norm": 0.28354543447494507, "learning_rate": 1.7128138388789694e-07, "loss": 0.1939, "step": 4933 }, { "epoch": 1.3129324108568388, "grad_norm": 0.2553695738315582, "learning_rate": 1.712694171378863e-07, "loss": 0.2013, "step": 4934 }, { "epoch": 1.3131985098456627, "grad_norm": 0.2784138321876526, "learning_rate": 1.7125744831340793e-07, "loss": 0.2273, "step": 4935 }, { "epoch": 1.3134646088344863, "grad_norm": 0.3016432523727417, "learning_rate": 1.7124547741481025e-07, "loss": 0.2152, "step": 4936 }, { "epoch": 1.3137307078233103, "grad_norm": 0.33178263902664185, "learning_rate": 1.7123350444244174e-07, "loss": 0.2364, "step": 4937 }, { "epoch": 1.3139968068121342, "grad_norm": 0.32259291410446167, "learning_rate": 1.7122152939665087e-07, "loss": 0.2268, "step": 4938 }, { "epoch": 1.314262905800958, "grad_norm": 0.37727564573287964, "learning_rate": 1.712095522777862e-07, "loss": 0.205, "step": 4939 }, { "epoch": 1.3145290047897817, "grad_norm": 0.2930578589439392, "learning_rate": 1.7119757308619636e-07, "loss": 0.1976, "step": 4940 }, { "epoch": 1.3147951037786056, "grad_norm": 0.27410489320755005, "learning_rate": 1.7118559182223003e-07, "loss": 0.2108, "step": 4941 }, { "epoch": 1.3150612027674295, "grad_norm": 0.28651800751686096, "learning_rate": 1.7117360848623593e-07, "loss": 0.2174, "step": 4942 }, { "epoch": 1.3153273017562532, "grad_norm": 0.2882501780986786, "learning_rate": 1.711616230785629e-07, "loss": 0.2171, "step": 4943 }, { "epoch": 1.3155934007450771, "grad_norm": 0.27719607949256897, "learning_rate": 1.711496355995598e-07, "loss": 0.207, "step": 4944 }, { "epoch": 1.315859499733901, "grad_norm": 0.28886815905570984, "learning_rate": 1.7113764604957553e-07, "loss": 0.205, "step": 4945 }, { "epoch": 1.3161255987227247, "grad_norm": 0.3212164342403412, "learning_rate": 1.7112565442895912e-07, "loss": 0.2288, "step": 4946 }, { "epoch": 1.3163916977115486, "grad_norm": 0.28023889660835266, "learning_rate": 1.7111366073805955e-07, "loss": 0.2054, "step": 4947 }, { "epoch": 1.3166577967003725, "grad_norm": 0.29642847180366516, "learning_rate": 1.7110166497722598e-07, "loss": 0.2299, "step": 4948 }, { "epoch": 1.3169238956891964, "grad_norm": 0.2665169835090637, "learning_rate": 1.7108966714680758e-07, "loss": 0.1928, "step": 4949 }, { "epoch": 1.3171899946780203, "grad_norm": 0.28620728850364685, "learning_rate": 1.710776672471535e-07, "loss": 0.2069, "step": 4950 }, { "epoch": 1.317456093666844, "grad_norm": 0.2631979286670685, "learning_rate": 1.710656652786131e-07, "loss": 0.2108, "step": 4951 }, { "epoch": 1.317722192655668, "grad_norm": 0.2777761220932007, "learning_rate": 1.710536612415357e-07, "loss": 0.2062, "step": 4952 }, { "epoch": 1.3179882916444918, "grad_norm": 0.27495938539505005, "learning_rate": 1.7104165513627074e-07, "loss": 0.2044, "step": 4953 }, { "epoch": 1.3182543906333155, "grad_norm": 0.2854897975921631, "learning_rate": 1.7102964696316763e-07, "loss": 0.2178, "step": 4954 }, { "epoch": 1.3185204896221394, "grad_norm": 0.25646185874938965, "learning_rate": 1.7101763672257591e-07, "loss": 0.1987, "step": 4955 }, { "epoch": 1.3187865886109633, "grad_norm": 0.30305933952331543, "learning_rate": 1.7100562441484524e-07, "loss": 0.2091, "step": 4956 }, { "epoch": 1.319052687599787, "grad_norm": 0.27997714281082153, "learning_rate": 1.7099361004032514e-07, "loss": 0.2071, "step": 4957 }, { "epoch": 1.319318786588611, "grad_norm": 0.28472912311553955, "learning_rate": 1.709815935993654e-07, "loss": 0.2047, "step": 4958 }, { "epoch": 1.3195848855774348, "grad_norm": 0.29020413756370544, "learning_rate": 1.709695750923158e-07, "loss": 0.2156, "step": 4959 }, { "epoch": 1.3198509845662587, "grad_norm": 0.26346659660339355, "learning_rate": 1.709575545195261e-07, "loss": 0.1978, "step": 4960 }, { "epoch": 1.3201170835550826, "grad_norm": 0.26088184118270874, "learning_rate": 1.7094553188134627e-07, "loss": 0.2049, "step": 4961 }, { "epoch": 1.3203831825439063, "grad_norm": 0.268825888633728, "learning_rate": 1.709335071781262e-07, "loss": 0.2125, "step": 4962 }, { "epoch": 1.3206492815327302, "grad_norm": 0.31156808137893677, "learning_rate": 1.709214804102159e-07, "loss": 0.2302, "step": 4963 }, { "epoch": 1.320915380521554, "grad_norm": 0.2812795341014862, "learning_rate": 1.7090945157796547e-07, "loss": 0.1917, "step": 4964 }, { "epoch": 1.3211814795103778, "grad_norm": 0.28688445687294006, "learning_rate": 1.7089742068172498e-07, "loss": 0.2337, "step": 4965 }, { "epoch": 1.3214475784992017, "grad_norm": 0.28209611773490906, "learning_rate": 1.7088538772184472e-07, "loss": 0.2278, "step": 4966 }, { "epoch": 1.3217136774880256, "grad_norm": 0.2932055592536926, "learning_rate": 1.708733526986748e-07, "loss": 0.2375, "step": 4967 }, { "epoch": 1.3219797764768493, "grad_norm": 0.287203311920166, "learning_rate": 1.7086131561256567e-07, "loss": 0.2368, "step": 4968 }, { "epoch": 1.3222458754656732, "grad_norm": 0.3005666434764862, "learning_rate": 1.708492764638676e-07, "loss": 0.2117, "step": 4969 }, { "epoch": 1.322511974454497, "grad_norm": 0.3014110028743744, "learning_rate": 1.7083723525293107e-07, "loss": 0.213, "step": 4970 }, { "epoch": 1.322778073443321, "grad_norm": 0.27846795320510864, "learning_rate": 1.708251919801065e-07, "loss": 0.2181, "step": 4971 }, { "epoch": 1.323044172432145, "grad_norm": 0.28683871030807495, "learning_rate": 1.7081314664574455e-07, "loss": 0.1992, "step": 4972 }, { "epoch": 1.3233102714209686, "grad_norm": 0.32347872853279114, "learning_rate": 1.7080109925019576e-07, "loss": 0.2241, "step": 4973 }, { "epoch": 1.3235763704097925, "grad_norm": 0.28267186880111694, "learning_rate": 1.7078904979381077e-07, "loss": 0.1966, "step": 4974 }, { "epoch": 1.3238424693986164, "grad_norm": 0.28827548027038574, "learning_rate": 1.7077699827694035e-07, "loss": 0.2133, "step": 4975 }, { "epoch": 1.32410856838744, "grad_norm": 0.2937965393066406, "learning_rate": 1.707649446999353e-07, "loss": 0.2159, "step": 4976 }, { "epoch": 1.324374667376264, "grad_norm": 0.2688154876232147, "learning_rate": 1.7075288906314643e-07, "loss": 0.2041, "step": 4977 }, { "epoch": 1.3246407663650879, "grad_norm": 0.29851439595222473, "learning_rate": 1.7074083136692464e-07, "loss": 0.2245, "step": 4978 }, { "epoch": 1.3249068653539116, "grad_norm": 0.37462958693504333, "learning_rate": 1.7072877161162098e-07, "loss": 0.1989, "step": 4979 }, { "epoch": 1.3251729643427355, "grad_norm": 0.30365777015686035, "learning_rate": 1.707167097975864e-07, "loss": 0.2098, "step": 4980 }, { "epoch": 1.3254390633315594, "grad_norm": 0.28536108136177063, "learning_rate": 1.7070464592517203e-07, "loss": 0.2192, "step": 4981 }, { "epoch": 1.325705162320383, "grad_norm": 0.27608248591423035, "learning_rate": 1.7069257999472896e-07, "loss": 0.2225, "step": 4982 }, { "epoch": 1.325971261309207, "grad_norm": 0.28699159622192383, "learning_rate": 1.7068051200660847e-07, "loss": 0.2146, "step": 4983 }, { "epoch": 1.3262373602980309, "grad_norm": 0.2917039692401886, "learning_rate": 1.7066844196116178e-07, "loss": 0.2228, "step": 4984 }, { "epoch": 1.3265034592868548, "grad_norm": 0.2924198806285858, "learning_rate": 1.7065636985874024e-07, "loss": 0.207, "step": 4985 }, { "epoch": 1.3267695582756787, "grad_norm": 0.2851904332637787, "learning_rate": 1.7064429569969523e-07, "loss": 0.2183, "step": 4986 }, { "epoch": 1.3270356572645023, "grad_norm": 0.3054778575897217, "learning_rate": 1.7063221948437822e-07, "loss": 0.2072, "step": 4987 }, { "epoch": 1.3273017562533262, "grad_norm": 0.29567983746528625, "learning_rate": 1.7062014121314068e-07, "loss": 0.1975, "step": 4988 }, { "epoch": 1.3275678552421502, "grad_norm": 0.2651594281196594, "learning_rate": 1.7060806088633418e-07, "loss": 0.1827, "step": 4989 }, { "epoch": 1.3278339542309738, "grad_norm": 0.26916003227233887, "learning_rate": 1.7059597850431037e-07, "loss": 0.201, "step": 4990 }, { "epoch": 1.3281000532197977, "grad_norm": 0.2964465320110321, "learning_rate": 1.705838940674209e-07, "loss": 0.214, "step": 4991 }, { "epoch": 1.3283661522086216, "grad_norm": 0.283949077129364, "learning_rate": 1.705718075760176e-07, "loss": 0.2106, "step": 4992 }, { "epoch": 1.3286322511974453, "grad_norm": 0.28226238489151, "learning_rate": 1.7055971903045218e-07, "loss": 0.217, "step": 4993 }, { "epoch": 1.3288983501862692, "grad_norm": 0.2634921967983246, "learning_rate": 1.7054762843107656e-07, "loss": 0.1834, "step": 4994 }, { "epoch": 1.3291644491750931, "grad_norm": 0.3233945965766907, "learning_rate": 1.7053553577824262e-07, "loss": 0.202, "step": 4995 }, { "epoch": 1.329430548163917, "grad_norm": 0.2768101096153259, "learning_rate": 1.705234410723024e-07, "loss": 0.2063, "step": 4996 }, { "epoch": 1.329696647152741, "grad_norm": 0.3243022859096527, "learning_rate": 1.7051134431360793e-07, "loss": 0.2403, "step": 4997 }, { "epoch": 1.3299627461415646, "grad_norm": 0.2708144187927246, "learning_rate": 1.7049924550251128e-07, "loss": 0.2045, "step": 4998 }, { "epoch": 1.3302288451303885, "grad_norm": 0.2558469772338867, "learning_rate": 1.704871446393647e-07, "loss": 0.1804, "step": 4999 }, { "epoch": 1.3304949441192124, "grad_norm": 0.275450736284256, "learning_rate": 1.7047504172452028e-07, "loss": 0.1964, "step": 5000 }, { "epoch": 1.3307610431080361, "grad_norm": 0.2828391492366791, "learning_rate": 1.7046293675833045e-07, "loss": 0.2023, "step": 5001 }, { "epoch": 1.33102714209686, "grad_norm": 0.2838476896286011, "learning_rate": 1.7045082974114746e-07, "loss": 0.2041, "step": 5002 }, { "epoch": 1.331293241085684, "grad_norm": 0.27828094363212585, "learning_rate": 1.7043872067332375e-07, "loss": 0.2062, "step": 5003 }, { "epoch": 1.3315593400745076, "grad_norm": 0.27256548404693604, "learning_rate": 1.7042660955521176e-07, "loss": 0.1974, "step": 5004 }, { "epoch": 1.3318254390633315, "grad_norm": 0.2727116644382477, "learning_rate": 1.70414496387164e-07, "loss": 0.201, "step": 5005 }, { "epoch": 1.3320915380521554, "grad_norm": 0.2947762608528137, "learning_rate": 1.7040238116953314e-07, "loss": 0.2271, "step": 5006 }, { "epoch": 1.3323576370409793, "grad_norm": 0.3194948732852936, "learning_rate": 1.7039026390267172e-07, "loss": 0.2308, "step": 5007 }, { "epoch": 1.3326237360298032, "grad_norm": 0.2469319999217987, "learning_rate": 1.703781445869325e-07, "loss": 0.1941, "step": 5008 }, { "epoch": 1.332889835018627, "grad_norm": 0.2735699415206909, "learning_rate": 1.7036602322266822e-07, "loss": 0.2122, "step": 5009 }, { "epoch": 1.3331559340074508, "grad_norm": 0.2786540389060974, "learning_rate": 1.7035389981023172e-07, "loss": 0.2114, "step": 5010 }, { "epoch": 1.3334220329962747, "grad_norm": 0.2765510082244873, "learning_rate": 1.7034177434997586e-07, "loss": 0.2047, "step": 5011 }, { "epoch": 1.3336881319850984, "grad_norm": 0.30432766675949097, "learning_rate": 1.7032964684225357e-07, "loss": 0.2288, "step": 5012 }, { "epoch": 1.3339542309739223, "grad_norm": 0.31925180554389954, "learning_rate": 1.7031751728741787e-07, "loss": 0.2236, "step": 5013 }, { "epoch": 1.3342203299627462, "grad_norm": 0.2690846621990204, "learning_rate": 1.7030538568582186e-07, "loss": 0.1796, "step": 5014 }, { "epoch": 1.3344864289515699, "grad_norm": 0.30122125148773193, "learning_rate": 1.7029325203781858e-07, "loss": 0.2125, "step": 5015 }, { "epoch": 1.3347525279403938, "grad_norm": 0.3021165728569031, "learning_rate": 1.7028111634376127e-07, "loss": 0.1989, "step": 5016 }, { "epoch": 1.3350186269292177, "grad_norm": 0.2986474335193634, "learning_rate": 1.7026897860400315e-07, "loss": 0.2155, "step": 5017 }, { "epoch": 1.3352847259180416, "grad_norm": 0.2719052731990814, "learning_rate": 1.7025683881889749e-07, "loss": 0.222, "step": 5018 }, { "epoch": 1.3355508249068653, "grad_norm": 0.28647929430007935, "learning_rate": 1.702446969887977e-07, "loss": 0.2064, "step": 5019 }, { "epoch": 1.3358169238956892, "grad_norm": 0.3028673827648163, "learning_rate": 1.7023255311405715e-07, "loss": 0.235, "step": 5020 }, { "epoch": 1.336083022884513, "grad_norm": 0.30811452865600586, "learning_rate": 1.7022040719502933e-07, "loss": 0.2293, "step": 5021 }, { "epoch": 1.336349121873337, "grad_norm": 0.2650791108608246, "learning_rate": 1.702082592320678e-07, "loss": 0.197, "step": 5022 }, { "epoch": 1.3366152208621607, "grad_norm": 0.28089869022369385, "learning_rate": 1.7019610922552614e-07, "loss": 0.2118, "step": 5023 }, { "epoch": 1.3368813198509846, "grad_norm": 0.28019726276397705, "learning_rate": 1.70183957175758e-07, "loss": 0.2233, "step": 5024 }, { "epoch": 1.3371474188398085, "grad_norm": 0.2886694073677063, "learning_rate": 1.701718030831171e-07, "loss": 0.2266, "step": 5025 }, { "epoch": 1.3374135178286322, "grad_norm": 0.26737117767333984, "learning_rate": 1.7015964694795716e-07, "loss": 0.195, "step": 5026 }, { "epoch": 1.337679616817456, "grad_norm": 0.27102047204971313, "learning_rate": 1.7014748877063213e-07, "loss": 0.2005, "step": 5027 }, { "epoch": 1.33794571580628, "grad_norm": 0.3000333607196808, "learning_rate": 1.7013532855149583e-07, "loss": 0.2209, "step": 5028 }, { "epoch": 1.3382118147951036, "grad_norm": 0.27626433968544006, "learning_rate": 1.7012316629090218e-07, "loss": 0.2012, "step": 5029 }, { "epoch": 1.3384779137839276, "grad_norm": 0.32389354705810547, "learning_rate": 1.7011100198920527e-07, "loss": 0.2241, "step": 5030 }, { "epoch": 1.3387440127727515, "grad_norm": 0.2854040861129761, "learning_rate": 1.700988356467591e-07, "loss": 0.2133, "step": 5031 }, { "epoch": 1.3390101117615754, "grad_norm": 0.33109474182128906, "learning_rate": 1.7008666726391785e-07, "loss": 0.2362, "step": 5032 }, { "epoch": 1.3392762107503993, "grad_norm": 0.2749245762825012, "learning_rate": 1.700744968410357e-07, "loss": 0.2036, "step": 5033 }, { "epoch": 1.339542309739223, "grad_norm": 0.2834044098854065, "learning_rate": 1.700623243784669e-07, "loss": 0.2022, "step": 5034 }, { "epoch": 1.3398084087280469, "grad_norm": 0.2686867415904999, "learning_rate": 1.7005014987656576e-07, "loss": 0.2011, "step": 5035 }, { "epoch": 1.3400745077168708, "grad_norm": 0.28695911169052124, "learning_rate": 1.700379733356866e-07, "loss": 0.2177, "step": 5036 }, { "epoch": 1.3403406067056944, "grad_norm": 0.30805322527885437, "learning_rate": 1.7002579475618394e-07, "loss": 0.2276, "step": 5037 }, { "epoch": 1.3406067056945183, "grad_norm": 0.2829694449901581, "learning_rate": 1.7001361413841218e-07, "loss": 0.2038, "step": 5038 }, { "epoch": 1.3408728046833422, "grad_norm": 0.2915895879268646, "learning_rate": 1.700014314827259e-07, "loss": 0.2236, "step": 5039 }, { "epoch": 1.341138903672166, "grad_norm": 0.2758098542690277, "learning_rate": 1.6998924678947973e-07, "loss": 0.1895, "step": 5040 }, { "epoch": 1.3414050026609898, "grad_norm": 0.2909296452999115, "learning_rate": 1.6997706005902826e-07, "loss": 0.1981, "step": 5041 }, { "epoch": 1.3416711016498137, "grad_norm": 0.323138564825058, "learning_rate": 1.6996487129172633e-07, "loss": 0.2212, "step": 5042 }, { "epoch": 1.3419372006386376, "grad_norm": 0.302769273519516, "learning_rate": 1.6995268048792863e-07, "loss": 0.2087, "step": 5043 }, { "epoch": 1.3422032996274615, "grad_norm": 0.29761236906051636, "learning_rate": 1.6994048764799004e-07, "loss": 0.2156, "step": 5044 }, { "epoch": 1.3424693986162852, "grad_norm": 0.27317649126052856, "learning_rate": 1.6992829277226544e-07, "loss": 0.214, "step": 5045 }, { "epoch": 1.3427354976051091, "grad_norm": 0.4099772572517395, "learning_rate": 1.6991609586110979e-07, "loss": 0.215, "step": 5046 }, { "epoch": 1.343001596593933, "grad_norm": 0.2694850265979767, "learning_rate": 1.6990389691487817e-07, "loss": 0.2096, "step": 5047 }, { "epoch": 1.3432676955827567, "grad_norm": 0.2914589047431946, "learning_rate": 1.6989169593392559e-07, "loss": 0.2238, "step": 5048 }, { "epoch": 1.3435337945715806, "grad_norm": 0.3027366101741791, "learning_rate": 1.698794929186072e-07, "loss": 0.2249, "step": 5049 }, { "epoch": 1.3437998935604045, "grad_norm": 0.2945494055747986, "learning_rate": 1.698672878692782e-07, "loss": 0.2172, "step": 5050 }, { "epoch": 1.3440659925492282, "grad_norm": 0.2960456609725952, "learning_rate": 1.6985508078629392e-07, "loss": 0.2052, "step": 5051 }, { "epoch": 1.344332091538052, "grad_norm": 0.2934248745441437, "learning_rate": 1.6984287167000958e-07, "loss": 0.22, "step": 5052 }, { "epoch": 1.344598190526876, "grad_norm": 0.268439382314682, "learning_rate": 1.6983066052078057e-07, "loss": 0.2135, "step": 5053 }, { "epoch": 1.3448642895157, "grad_norm": 0.2795253396034241, "learning_rate": 1.698184473389624e-07, "loss": 0.1992, "step": 5054 }, { "epoch": 1.3451303885045236, "grad_norm": 0.2857441306114197, "learning_rate": 1.6980623212491044e-07, "loss": 0.2199, "step": 5055 }, { "epoch": 1.3453964874933475, "grad_norm": 0.27518510818481445, "learning_rate": 1.6979401487898035e-07, "loss": 0.2054, "step": 5056 }, { "epoch": 1.3456625864821714, "grad_norm": 0.2844908535480499, "learning_rate": 1.697817956015277e-07, "loss": 0.1979, "step": 5057 }, { "epoch": 1.3459286854709953, "grad_norm": 0.28093618154525757, "learning_rate": 1.6976957429290816e-07, "loss": 0.2121, "step": 5058 }, { "epoch": 1.346194784459819, "grad_norm": 0.31523531675338745, "learning_rate": 1.697573509534775e-07, "loss": 0.217, "step": 5059 }, { "epoch": 1.346460883448643, "grad_norm": 0.28158602118492126, "learning_rate": 1.6974512558359145e-07, "loss": 0.2079, "step": 5060 }, { "epoch": 1.3467269824374668, "grad_norm": 0.2817215025424957, "learning_rate": 1.6973289818360588e-07, "loss": 0.2305, "step": 5061 }, { "epoch": 1.3469930814262905, "grad_norm": 0.2890990078449249, "learning_rate": 1.6972066875387672e-07, "loss": 0.2114, "step": 5062 }, { "epoch": 1.3472591804151144, "grad_norm": 0.30245548486709595, "learning_rate": 1.697084372947599e-07, "loss": 0.2303, "step": 5063 }, { "epoch": 1.3475252794039383, "grad_norm": 0.2862842082977295, "learning_rate": 1.6969620380661146e-07, "loss": 0.2014, "step": 5064 }, { "epoch": 1.347791378392762, "grad_norm": 0.2688910961151123, "learning_rate": 1.6968396828978754e-07, "loss": 0.2091, "step": 5065 }, { "epoch": 1.3480574773815859, "grad_norm": 0.2867794334888458, "learning_rate": 1.696717307446442e-07, "loss": 0.2016, "step": 5066 }, { "epoch": 1.3483235763704098, "grad_norm": 0.2603475749492645, "learning_rate": 1.6965949117153768e-07, "loss": 0.1858, "step": 5067 }, { "epoch": 1.3485896753592337, "grad_norm": 0.2863585352897644, "learning_rate": 1.6964724957082423e-07, "loss": 0.209, "step": 5068 }, { "epoch": 1.3488557743480576, "grad_norm": 0.2836364209651947, "learning_rate": 1.696350059428602e-07, "loss": 0.2053, "step": 5069 }, { "epoch": 1.3491218733368813, "grad_norm": 0.3030086159706116, "learning_rate": 1.69622760288002e-07, "loss": 0.2292, "step": 5070 }, { "epoch": 1.3493879723257052, "grad_norm": 0.4767168164253235, "learning_rate": 1.6961051260660595e-07, "loss": 0.2152, "step": 5071 }, { "epoch": 1.349654071314529, "grad_norm": 0.30820679664611816, "learning_rate": 1.6959826289902863e-07, "loss": 0.219, "step": 5072 }, { "epoch": 1.3499201703033528, "grad_norm": 0.3033353388309479, "learning_rate": 1.695860111656266e-07, "loss": 0.211, "step": 5073 }, { "epoch": 1.3501862692921767, "grad_norm": 0.27349236607551575, "learning_rate": 1.6957375740675647e-07, "loss": 0.2007, "step": 5074 }, { "epoch": 1.3504523682810006, "grad_norm": 0.28722503781318665, "learning_rate": 1.695615016227749e-07, "loss": 0.212, "step": 5075 }, { "epoch": 1.3507184672698243, "grad_norm": 0.2938523590564728, "learning_rate": 1.6954924381403864e-07, "loss": 0.222, "step": 5076 }, { "epoch": 1.3509845662586482, "grad_norm": 0.2579273581504822, "learning_rate": 1.6953698398090445e-07, "loss": 0.1735, "step": 5077 }, { "epoch": 1.351250665247472, "grad_norm": 0.27529704570770264, "learning_rate": 1.6952472212372925e-07, "loss": 0.1941, "step": 5078 }, { "epoch": 1.351516764236296, "grad_norm": 0.30431535840034485, "learning_rate": 1.6951245824286987e-07, "loss": 0.2148, "step": 5079 }, { "epoch": 1.3517828632251199, "grad_norm": 0.29531553387641907, "learning_rate": 1.6950019233868333e-07, "loss": 0.215, "step": 5080 }, { "epoch": 1.3520489622139436, "grad_norm": 0.30712950229644775, "learning_rate": 1.6948792441152665e-07, "loss": 0.2066, "step": 5081 }, { "epoch": 1.3523150612027675, "grad_norm": 0.2757423222064972, "learning_rate": 1.694756544617569e-07, "loss": 0.2044, "step": 5082 }, { "epoch": 1.3525811601915914, "grad_norm": 0.27971524000167847, "learning_rate": 1.6946338248973126e-07, "loss": 0.2106, "step": 5083 }, { "epoch": 1.352847259180415, "grad_norm": 0.26785051822662354, "learning_rate": 1.6945110849580692e-07, "loss": 0.1885, "step": 5084 }, { "epoch": 1.353113358169239, "grad_norm": 0.2950770854949951, "learning_rate": 1.6943883248034112e-07, "loss": 0.2174, "step": 5085 }, { "epoch": 1.3533794571580628, "grad_norm": 0.29438671469688416, "learning_rate": 1.694265544436912e-07, "loss": 0.232, "step": 5086 }, { "epoch": 1.3536455561468865, "grad_norm": 0.6660184264183044, "learning_rate": 1.6941427438621457e-07, "loss": 0.2084, "step": 5087 }, { "epoch": 1.3539116551357104, "grad_norm": 0.27316322922706604, "learning_rate": 1.694019923082686e-07, "loss": 0.2082, "step": 5088 }, { "epoch": 1.3541777541245343, "grad_norm": 0.28621765971183777, "learning_rate": 1.6938970821021088e-07, "loss": 0.2193, "step": 5089 }, { "epoch": 1.3544438531133582, "grad_norm": 0.30485910177230835, "learning_rate": 1.6937742209239892e-07, "loss": 0.2, "step": 5090 }, { "epoch": 1.3547099521021821, "grad_norm": 0.27571702003479004, "learning_rate": 1.693651339551903e-07, "loss": 0.2043, "step": 5091 }, { "epoch": 1.3549760510910058, "grad_norm": 0.27991631627082825, "learning_rate": 1.693528437989428e-07, "loss": 0.2224, "step": 5092 }, { "epoch": 1.3552421500798297, "grad_norm": 0.2859942615032196, "learning_rate": 1.6934055162401404e-07, "loss": 0.2003, "step": 5093 }, { "epoch": 1.3555082490686536, "grad_norm": 0.30638185143470764, "learning_rate": 1.6932825743076186e-07, "loss": 0.2183, "step": 5094 }, { "epoch": 1.3557743480574773, "grad_norm": 0.3115886449813843, "learning_rate": 1.6931596121954413e-07, "loss": 0.2015, "step": 5095 }, { "epoch": 1.3560404470463012, "grad_norm": 0.294769823551178, "learning_rate": 1.6930366299071876e-07, "loss": 0.2102, "step": 5096 }, { "epoch": 1.3563065460351251, "grad_norm": 0.2757790982723236, "learning_rate": 1.6929136274464368e-07, "loss": 0.1919, "step": 5097 }, { "epoch": 1.3565726450239488, "grad_norm": 0.27115294337272644, "learning_rate": 1.6927906048167695e-07, "loss": 0.2094, "step": 5098 }, { "epoch": 1.3568387440127727, "grad_norm": 0.29377248883247375, "learning_rate": 1.6926675620217664e-07, "loss": 0.1941, "step": 5099 }, { "epoch": 1.3571048430015966, "grad_norm": 0.2923159897327423, "learning_rate": 1.6925444990650096e-07, "loss": 0.2064, "step": 5100 }, { "epoch": 1.3573709419904203, "grad_norm": 0.32935455441474915, "learning_rate": 1.6924214159500802e-07, "loss": 0.2326, "step": 5101 }, { "epoch": 1.3576370409792442, "grad_norm": 0.27202826738357544, "learning_rate": 1.6922983126805613e-07, "loss": 0.2136, "step": 5102 }, { "epoch": 1.357903139968068, "grad_norm": 0.26712730526924133, "learning_rate": 1.6921751892600357e-07, "loss": 0.1973, "step": 5103 }, { "epoch": 1.358169238956892, "grad_norm": 0.3175046741962433, "learning_rate": 1.6920520456920882e-07, "loss": 0.2288, "step": 5104 }, { "epoch": 1.358435337945716, "grad_norm": 0.37795957922935486, "learning_rate": 1.691928881980302e-07, "loss": 0.1891, "step": 5105 }, { "epoch": 1.3587014369345396, "grad_norm": 0.28876838088035583, "learning_rate": 1.6918056981282633e-07, "loss": 0.2181, "step": 5106 }, { "epoch": 1.3589675359233635, "grad_norm": 0.3238409459590912, "learning_rate": 1.6916824941395562e-07, "loss": 0.2104, "step": 5107 }, { "epoch": 1.3592336349121874, "grad_norm": 0.29349270462989807, "learning_rate": 1.691559270017768e-07, "loss": 0.22, "step": 5108 }, { "epoch": 1.359499733901011, "grad_norm": 0.31510961055755615, "learning_rate": 1.6914360257664847e-07, "loss": 0.226, "step": 5109 }, { "epoch": 1.359765832889835, "grad_norm": 0.3114333152770996, "learning_rate": 1.6913127613892945e-07, "loss": 0.2224, "step": 5110 }, { "epoch": 1.360031931878659, "grad_norm": 0.2586032748222351, "learning_rate": 1.6911894768897847e-07, "loss": 0.1905, "step": 5111 }, { "epoch": 1.3602980308674826, "grad_norm": 0.29272595047950745, "learning_rate": 1.6910661722715438e-07, "loss": 0.2029, "step": 5112 }, { "epoch": 1.3605641298563065, "grad_norm": 0.282600998878479, "learning_rate": 1.690942847538161e-07, "loss": 0.2122, "step": 5113 }, { "epoch": 1.3608302288451304, "grad_norm": 0.28534257411956787, "learning_rate": 1.690819502693226e-07, "loss": 0.2128, "step": 5114 }, { "epoch": 1.3610963278339543, "grad_norm": 0.30775219202041626, "learning_rate": 1.6906961377403286e-07, "loss": 0.2405, "step": 5115 }, { "epoch": 1.3613624268227782, "grad_norm": 0.26791805028915405, "learning_rate": 1.6905727526830602e-07, "loss": 0.198, "step": 5116 }, { "epoch": 1.3616285258116019, "grad_norm": 0.31919828057289124, "learning_rate": 1.690449347525012e-07, "loss": 0.2229, "step": 5117 }, { "epoch": 1.3618946248004258, "grad_norm": 0.2556561231613159, "learning_rate": 1.6903259222697762e-07, "loss": 0.1857, "step": 5118 }, { "epoch": 1.3621607237892497, "grad_norm": 0.28509271144866943, "learning_rate": 1.690202476920945e-07, "loss": 0.2201, "step": 5119 }, { "epoch": 1.3624268227780734, "grad_norm": 0.30224311351776123, "learning_rate": 1.690079011482112e-07, "loss": 0.2303, "step": 5120 }, { "epoch": 1.3626929217668973, "grad_norm": 0.28812357783317566, "learning_rate": 1.6899555259568707e-07, "loss": 0.2119, "step": 5121 }, { "epoch": 1.3629590207557212, "grad_norm": 0.2827504277229309, "learning_rate": 1.6898320203488152e-07, "loss": 0.2164, "step": 5122 }, { "epoch": 1.3632251197445449, "grad_norm": 0.2778528332710266, "learning_rate": 1.689708494661541e-07, "loss": 0.2259, "step": 5123 }, { "epoch": 1.3634912187333688, "grad_norm": 0.3073732256889343, "learning_rate": 1.6895849488986432e-07, "loss": 0.2274, "step": 5124 }, { "epoch": 1.3637573177221927, "grad_norm": 0.3395345211029053, "learning_rate": 1.689461383063718e-07, "loss": 0.2295, "step": 5125 }, { "epoch": 1.3640234167110166, "grad_norm": 0.30107516050338745, "learning_rate": 1.6893377971603625e-07, "loss": 0.2125, "step": 5126 }, { "epoch": 1.3642895156998405, "grad_norm": 0.2888781428337097, "learning_rate": 1.6892141911921734e-07, "loss": 0.2286, "step": 5127 }, { "epoch": 1.3645556146886642, "grad_norm": 0.28378185629844666, "learning_rate": 1.6890905651627485e-07, "loss": 0.2143, "step": 5128 }, { "epoch": 1.364821713677488, "grad_norm": 0.29711490869522095, "learning_rate": 1.6889669190756864e-07, "loss": 0.2275, "step": 5129 }, { "epoch": 1.365087812666312, "grad_norm": 0.3681948482990265, "learning_rate": 1.6888432529345865e-07, "loss": 0.2068, "step": 5130 }, { "epoch": 1.3653539116551356, "grad_norm": 0.2674607038497925, "learning_rate": 1.6887195667430478e-07, "loss": 0.202, "step": 5131 }, { "epoch": 1.3656200106439595, "grad_norm": 0.2792015075683594, "learning_rate": 1.688595860504671e-07, "loss": 0.2218, "step": 5132 }, { "epoch": 1.3658861096327835, "grad_norm": 0.2843976616859436, "learning_rate": 1.6884721342230565e-07, "loss": 0.217, "step": 5133 }, { "epoch": 1.3661522086216071, "grad_norm": 0.3070540726184845, "learning_rate": 1.6883483879018056e-07, "loss": 0.2281, "step": 5134 }, { "epoch": 1.366418307610431, "grad_norm": 0.2988263666629791, "learning_rate": 1.6882246215445207e-07, "loss": 0.2084, "step": 5135 }, { "epoch": 1.366684406599255, "grad_norm": 0.28475505113601685, "learning_rate": 1.6881008351548037e-07, "loss": 0.1948, "step": 5136 }, { "epoch": 1.3669505055880788, "grad_norm": 0.2757579982280731, "learning_rate": 1.6879770287362582e-07, "loss": 0.1863, "step": 5137 }, { "epoch": 1.3672166045769025, "grad_norm": 0.29203736782073975, "learning_rate": 1.687853202292488e-07, "loss": 0.2203, "step": 5138 }, { "epoch": 1.3674827035657264, "grad_norm": 0.29042357206344604, "learning_rate": 1.6877293558270967e-07, "loss": 0.2005, "step": 5139 }, { "epoch": 1.3677488025545503, "grad_norm": 0.3198578357696533, "learning_rate": 1.6876054893436894e-07, "loss": 0.2289, "step": 5140 }, { "epoch": 1.3680149015433742, "grad_norm": 0.2963845431804657, "learning_rate": 1.687481602845872e-07, "loss": 0.1933, "step": 5141 }, { "epoch": 1.368281000532198, "grad_norm": 0.2838164269924164, "learning_rate": 1.68735769633725e-07, "loss": 0.2158, "step": 5142 }, { "epoch": 1.3685470995210218, "grad_norm": 0.29693248867988586, "learning_rate": 1.68723376982143e-07, "loss": 0.2068, "step": 5143 }, { "epoch": 1.3688131985098457, "grad_norm": 0.27872195839881897, "learning_rate": 1.6871098233020197e-07, "loss": 0.2062, "step": 5144 }, { "epoch": 1.3690792974986694, "grad_norm": 0.30120325088500977, "learning_rate": 1.6869858567826266e-07, "loss": 0.2287, "step": 5145 }, { "epoch": 1.3693453964874933, "grad_norm": 0.28673088550567627, "learning_rate": 1.6868618702668585e-07, "loss": 0.2172, "step": 5146 }, { "epoch": 1.3696114954763172, "grad_norm": 0.28591015934944153, "learning_rate": 1.6867378637583253e-07, "loss": 0.2186, "step": 5147 }, { "epoch": 1.369877594465141, "grad_norm": 0.2835567891597748, "learning_rate": 1.6866138372606353e-07, "loss": 0.1924, "step": 5148 }, { "epoch": 1.3701436934539648, "grad_norm": 0.272664338350296, "learning_rate": 1.6864897907773996e-07, "loss": 0.2099, "step": 5149 }, { "epoch": 1.3704097924427887, "grad_norm": 0.2843245565891266, "learning_rate": 1.6863657243122287e-07, "loss": 0.2067, "step": 5150 }, { "epoch": 1.3706758914316126, "grad_norm": 0.29939019680023193, "learning_rate": 1.6862416378687336e-07, "loss": 0.2139, "step": 5151 }, { "epoch": 1.3709419904204365, "grad_norm": 0.306393563747406, "learning_rate": 1.686117531450526e-07, "loss": 0.2131, "step": 5152 }, { "epoch": 1.3712080894092602, "grad_norm": 0.3030489981174469, "learning_rate": 1.685993405061219e-07, "loss": 0.2393, "step": 5153 }, { "epoch": 1.371474188398084, "grad_norm": 0.28830647468566895, "learning_rate": 1.6858692587044247e-07, "loss": 0.2076, "step": 5154 }, { "epoch": 1.371740287386908, "grad_norm": 0.2992657721042633, "learning_rate": 1.6857450923837572e-07, "loss": 0.2152, "step": 5155 }, { "epoch": 1.3720063863757317, "grad_norm": 0.3102528750896454, "learning_rate": 1.685620906102831e-07, "loss": 0.2173, "step": 5156 }, { "epoch": 1.3722724853645556, "grad_norm": 0.29686105251312256, "learning_rate": 1.6854966998652599e-07, "loss": 0.2177, "step": 5157 }, { "epoch": 1.3725385843533795, "grad_norm": 0.29016995429992676, "learning_rate": 1.68537247367466e-07, "loss": 0.2262, "step": 5158 }, { "epoch": 1.3728046833422032, "grad_norm": 0.2774236500263214, "learning_rate": 1.685248227534647e-07, "loss": 0.2101, "step": 5159 }, { "epoch": 1.373070782331027, "grad_norm": 0.30085235834121704, "learning_rate": 1.6851239614488372e-07, "loss": 0.2193, "step": 5160 }, { "epoch": 1.373336881319851, "grad_norm": 0.2783963084220886, "learning_rate": 1.684999675420848e-07, "loss": 0.2176, "step": 5161 }, { "epoch": 1.373602980308675, "grad_norm": 0.28488120436668396, "learning_rate": 1.6848753694542962e-07, "loss": 0.2135, "step": 5162 }, { "epoch": 1.3738690792974988, "grad_norm": 0.276201456785202, "learning_rate": 1.6847510435528012e-07, "loss": 0.2037, "step": 5163 }, { "epoch": 1.3741351782863225, "grad_norm": 0.3078688681125641, "learning_rate": 1.6846266977199814e-07, "loss": 0.2015, "step": 5164 }, { "epoch": 1.3744012772751464, "grad_norm": 0.3257770240306854, "learning_rate": 1.6845023319594556e-07, "loss": 0.2354, "step": 5165 }, { "epoch": 1.3746673762639703, "grad_norm": 0.29027414321899414, "learning_rate": 1.6843779462748445e-07, "loss": 0.2089, "step": 5166 }, { "epoch": 1.374933475252794, "grad_norm": 0.3116455674171448, "learning_rate": 1.6842535406697682e-07, "loss": 0.2254, "step": 5167 }, { "epoch": 1.3751995742416179, "grad_norm": 0.2798795700073242, "learning_rate": 1.684129115147848e-07, "loss": 0.2013, "step": 5168 }, { "epoch": 1.3754656732304418, "grad_norm": 0.29183244705200195, "learning_rate": 1.684004669712706e-07, "loss": 0.2093, "step": 5169 }, { "epoch": 1.3757317722192655, "grad_norm": 0.27142372727394104, "learning_rate": 1.6838802043679634e-07, "loss": 0.2028, "step": 5170 }, { "epoch": 1.3759978712080894, "grad_norm": 0.34399738907814026, "learning_rate": 1.683755719117244e-07, "loss": 0.2273, "step": 5171 }, { "epoch": 1.3762639701969133, "grad_norm": 0.3057014048099518, "learning_rate": 1.6836312139641708e-07, "loss": 0.2087, "step": 5172 }, { "epoch": 1.3765300691857372, "grad_norm": 0.2949291467666626, "learning_rate": 1.6835066889123683e-07, "loss": 0.2089, "step": 5173 }, { "epoch": 1.376796168174561, "grad_norm": 0.30886727571487427, "learning_rate": 1.6833821439654606e-07, "loss": 0.2189, "step": 5174 }, { "epoch": 1.3770622671633848, "grad_norm": 0.2721150517463684, "learning_rate": 1.6832575791270734e-07, "loss": 0.209, "step": 5175 }, { "epoch": 1.3773283661522087, "grad_norm": 0.2571910321712494, "learning_rate": 1.683132994400832e-07, "loss": 0.199, "step": 5176 }, { "epoch": 1.3775944651410326, "grad_norm": 0.2698059380054474, "learning_rate": 1.6830083897903627e-07, "loss": 0.2047, "step": 5177 }, { "epoch": 1.3778605641298562, "grad_norm": 0.2708868980407715, "learning_rate": 1.6828837652992927e-07, "loss": 0.2181, "step": 5178 }, { "epoch": 1.3781266631186802, "grad_norm": 0.29716166853904724, "learning_rate": 1.6827591209312494e-07, "loss": 0.211, "step": 5179 }, { "epoch": 1.378392762107504, "grad_norm": 0.2885902225971222, "learning_rate": 1.6826344566898604e-07, "loss": 0.2094, "step": 5180 }, { "epoch": 1.3786588610963277, "grad_norm": 0.2753218710422516, "learning_rate": 1.682509772578755e-07, "loss": 0.2066, "step": 5181 }, { "epoch": 1.3789249600851516, "grad_norm": 0.31153005361557007, "learning_rate": 1.6823850686015625e-07, "loss": 0.2186, "step": 5182 }, { "epoch": 1.3791910590739755, "grad_norm": 0.2682363986968994, "learning_rate": 1.6822603447619126e-07, "loss": 0.1996, "step": 5183 }, { "epoch": 1.3794571580627992, "grad_norm": 0.28223875164985657, "learning_rate": 1.6821356010634352e-07, "loss": 0.2141, "step": 5184 }, { "epoch": 1.3797232570516231, "grad_norm": 0.3008047640323639, "learning_rate": 1.6820108375097618e-07, "loss": 0.2289, "step": 5185 }, { "epoch": 1.379989356040447, "grad_norm": 0.432468056678772, "learning_rate": 1.6818860541045234e-07, "loss": 0.2263, "step": 5186 }, { "epoch": 1.380255455029271, "grad_norm": 0.27905455231666565, "learning_rate": 1.681761250851353e-07, "loss": 0.216, "step": 5187 }, { "epoch": 1.3805215540180948, "grad_norm": 0.27944064140319824, "learning_rate": 1.6816364277538825e-07, "loss": 0.2089, "step": 5188 }, { "epoch": 1.3807876530069185, "grad_norm": 0.2828046679496765, "learning_rate": 1.6815115848157452e-07, "loss": 0.2092, "step": 5189 }, { "epoch": 1.3810537519957424, "grad_norm": 0.2842451333999634, "learning_rate": 1.6813867220405756e-07, "loss": 0.2098, "step": 5190 }, { "epoch": 1.3813198509845663, "grad_norm": 0.27860745787620544, "learning_rate": 1.6812618394320076e-07, "loss": 0.2, "step": 5191 }, { "epoch": 1.38158594997339, "grad_norm": 0.3342801332473755, "learning_rate": 1.6811369369936762e-07, "loss": 0.2012, "step": 5192 }, { "epoch": 1.381852048962214, "grad_norm": 0.2914660573005676, "learning_rate": 1.6810120147292173e-07, "loss": 0.2079, "step": 5193 }, { "epoch": 1.3821181479510378, "grad_norm": 0.2859715521335602, "learning_rate": 1.6808870726422665e-07, "loss": 0.2168, "step": 5194 }, { "epoch": 1.3823842469398615, "grad_norm": 0.27692711353302, "learning_rate": 1.680762110736461e-07, "loss": 0.2015, "step": 5195 }, { "epoch": 1.3826503459286854, "grad_norm": 0.2861398756504059, "learning_rate": 1.6806371290154386e-07, "loss": 0.2044, "step": 5196 }, { "epoch": 1.3829164449175093, "grad_norm": 0.27125024795532227, "learning_rate": 1.6805121274828362e-07, "loss": 0.1986, "step": 5197 }, { "epoch": 1.3831825439063332, "grad_norm": 0.2860058546066284, "learning_rate": 1.6803871061422927e-07, "loss": 0.2035, "step": 5198 }, { "epoch": 1.3834486428951571, "grad_norm": 0.2672708034515381, "learning_rate": 1.6802620649974473e-07, "loss": 0.202, "step": 5199 }, { "epoch": 1.3837147418839808, "grad_norm": 0.27840733528137207, "learning_rate": 1.6801370040519394e-07, "loss": 0.2076, "step": 5200 }, { "epoch": 1.3839808408728047, "grad_norm": 0.2941058576107025, "learning_rate": 1.680011923309409e-07, "loss": 0.201, "step": 5201 }, { "epoch": 1.3842469398616286, "grad_norm": 0.307802677154541, "learning_rate": 1.6798868227734974e-07, "loss": 0.2207, "step": 5202 }, { "epoch": 1.3845130388504523, "grad_norm": 0.2881079316139221, "learning_rate": 1.6797617024478456e-07, "loss": 0.2128, "step": 5203 }, { "epoch": 1.3847791378392762, "grad_norm": 0.2989393472671509, "learning_rate": 1.6796365623360957e-07, "loss": 0.2226, "step": 5204 }, { "epoch": 1.3850452368281, "grad_norm": 0.2782730162143707, "learning_rate": 1.6795114024418904e-07, "loss": 0.2171, "step": 5205 }, { "epoch": 1.3853113358169238, "grad_norm": 0.30191975831985474, "learning_rate": 1.6793862227688722e-07, "loss": 0.2045, "step": 5206 }, { "epoch": 1.3855774348057477, "grad_norm": 0.26711368560791016, "learning_rate": 1.679261023320685e-07, "loss": 0.1794, "step": 5207 }, { "epoch": 1.3858435337945716, "grad_norm": 0.27203795313835144, "learning_rate": 1.6791358041009734e-07, "loss": 0.1895, "step": 5208 }, { "epoch": 1.3861096327833955, "grad_norm": 0.25958770513534546, "learning_rate": 1.6790105651133818e-07, "loss": 0.1918, "step": 5209 }, { "epoch": 1.3863757317722194, "grad_norm": 0.292611300945282, "learning_rate": 1.6788853063615554e-07, "loss": 0.2268, "step": 5210 }, { "epoch": 1.386641830761043, "grad_norm": 0.3119863271713257, "learning_rate": 1.6787600278491406e-07, "loss": 0.2233, "step": 5211 }, { "epoch": 1.386907929749867, "grad_norm": 0.3047299087047577, "learning_rate": 1.6786347295797837e-07, "loss": 0.226, "step": 5212 }, { "epoch": 1.387174028738691, "grad_norm": 0.2918750047683716, "learning_rate": 1.678509411557132e-07, "loss": 0.2086, "step": 5213 }, { "epoch": 1.3874401277275146, "grad_norm": 0.2949208915233612, "learning_rate": 1.6783840737848332e-07, "loss": 0.2021, "step": 5214 }, { "epoch": 1.3877062267163385, "grad_norm": 0.29217028617858887, "learning_rate": 1.6782587162665355e-07, "loss": 0.2253, "step": 5215 }, { "epoch": 1.3879723257051624, "grad_norm": 0.2743974030017853, "learning_rate": 1.678133339005887e-07, "loss": 0.2105, "step": 5216 }, { "epoch": 1.388238424693986, "grad_norm": 0.27721452713012695, "learning_rate": 1.6780079420065382e-07, "loss": 0.2032, "step": 5217 }, { "epoch": 1.38850452368281, "grad_norm": 0.2977055013179779, "learning_rate": 1.6778825252721388e-07, "loss": 0.2199, "step": 5218 }, { "epoch": 1.3887706226716339, "grad_norm": 0.2774563133716583, "learning_rate": 1.6777570888063388e-07, "loss": 0.2187, "step": 5219 }, { "epoch": 1.3890367216604576, "grad_norm": 0.28832486271858215, "learning_rate": 1.6776316326127899e-07, "loss": 0.21, "step": 5220 }, { "epoch": 1.3893028206492815, "grad_norm": 0.28935033082962036, "learning_rate": 1.6775061566951436e-07, "loss": 0.2173, "step": 5221 }, { "epoch": 1.3895689196381054, "grad_norm": 0.29071059823036194, "learning_rate": 1.6773806610570526e-07, "loss": 0.2247, "step": 5222 }, { "epoch": 1.3898350186269293, "grad_norm": 0.281875878572464, "learning_rate": 1.677255145702169e-07, "loss": 0.1867, "step": 5223 }, { "epoch": 1.3901011176157532, "grad_norm": 0.25884145498275757, "learning_rate": 1.6771296106341466e-07, "loss": 0.1962, "step": 5224 }, { "epoch": 1.3903672166045768, "grad_norm": 0.28385302424430847, "learning_rate": 1.6770040558566392e-07, "loss": 0.2001, "step": 5225 }, { "epoch": 1.3906333155934008, "grad_norm": 0.28892847895622253, "learning_rate": 1.6768784813733016e-07, "loss": 0.2158, "step": 5226 }, { "epoch": 1.3908994145822247, "grad_norm": 0.27495908737182617, "learning_rate": 1.6767528871877893e-07, "loss": 0.2081, "step": 5227 }, { "epoch": 1.3911655135710483, "grad_norm": 0.2888483703136444, "learning_rate": 1.6766272733037575e-07, "loss": 0.2259, "step": 5228 }, { "epoch": 1.3914316125598722, "grad_norm": 0.280105859041214, "learning_rate": 1.6765016397248623e-07, "loss": 0.2092, "step": 5229 }, { "epoch": 1.3916977115486961, "grad_norm": 0.2922811210155487, "learning_rate": 1.6763759864547612e-07, "loss": 0.203, "step": 5230 }, { "epoch": 1.3919638105375198, "grad_norm": 0.2679198980331421, "learning_rate": 1.6762503134971108e-07, "loss": 0.1929, "step": 5231 }, { "epoch": 1.3922299095263437, "grad_norm": 0.28185877203941345, "learning_rate": 1.6761246208555704e-07, "loss": 0.2111, "step": 5232 }, { "epoch": 1.3924960085151676, "grad_norm": 0.2876119017601013, "learning_rate": 1.6759989085337973e-07, "loss": 0.2201, "step": 5233 }, { "epoch": 1.3927621075039915, "grad_norm": 0.28526490926742554, "learning_rate": 1.6758731765354515e-07, "loss": 0.204, "step": 5234 }, { "epoch": 1.3930282064928154, "grad_norm": 0.28899434208869934, "learning_rate": 1.675747424864192e-07, "loss": 0.2037, "step": 5235 }, { "epoch": 1.3932943054816391, "grad_norm": 0.29526981711387634, "learning_rate": 1.67562165352368e-07, "loss": 0.205, "step": 5236 }, { "epoch": 1.393560404470463, "grad_norm": 0.29245519638061523, "learning_rate": 1.6754958625175756e-07, "loss": 0.1975, "step": 5237 }, { "epoch": 1.393826503459287, "grad_norm": 0.28766289353370667, "learning_rate": 1.6753700518495408e-07, "loss": 0.1939, "step": 5238 }, { "epoch": 1.3940926024481106, "grad_norm": 0.2746298909187317, "learning_rate": 1.6752442215232372e-07, "loss": 0.208, "step": 5239 }, { "epoch": 1.3943587014369345, "grad_norm": 0.2830902934074402, "learning_rate": 1.6751183715423275e-07, "loss": 0.2112, "step": 5240 }, { "epoch": 1.3946248004257584, "grad_norm": 0.29105713963508606, "learning_rate": 1.674992501910475e-07, "loss": 0.2049, "step": 5241 }, { "epoch": 1.394890899414582, "grad_norm": 0.3217046856880188, "learning_rate": 1.6748666126313434e-07, "loss": 0.2206, "step": 5242 }, { "epoch": 1.395156998403406, "grad_norm": 0.30733317136764526, "learning_rate": 1.674740703708597e-07, "loss": 0.2119, "step": 5243 }, { "epoch": 1.39542309739223, "grad_norm": 0.30629587173461914, "learning_rate": 1.6746147751459007e-07, "loss": 0.2332, "step": 5244 }, { "epoch": 1.3956891963810538, "grad_norm": 0.27474287152290344, "learning_rate": 1.6744888269469198e-07, "loss": 0.2106, "step": 5245 }, { "epoch": 1.3959552953698777, "grad_norm": 0.273944616317749, "learning_rate": 1.6743628591153204e-07, "loss": 0.1998, "step": 5246 }, { "epoch": 1.3962213943587014, "grad_norm": 0.30781039595603943, "learning_rate": 1.6742368716547692e-07, "loss": 0.2256, "step": 5247 }, { "epoch": 1.3964874933475253, "grad_norm": 0.2792094349861145, "learning_rate": 1.6741108645689334e-07, "loss": 0.2092, "step": 5248 }, { "epoch": 1.3967535923363492, "grad_norm": 0.2806214988231659, "learning_rate": 1.6739848378614804e-07, "loss": 0.2091, "step": 5249 }, { "epoch": 1.397019691325173, "grad_norm": 0.2724461555480957, "learning_rate": 1.6738587915360791e-07, "loss": 0.1994, "step": 5250 }, { "epoch": 1.3972857903139968, "grad_norm": 0.2943551242351532, "learning_rate": 1.6737327255963978e-07, "loss": 0.2166, "step": 5251 }, { "epoch": 1.3975518893028207, "grad_norm": 0.277313232421875, "learning_rate": 1.6736066400461062e-07, "loss": 0.1988, "step": 5252 }, { "epoch": 1.3978179882916444, "grad_norm": 0.2654152512550354, "learning_rate": 1.6734805348888746e-07, "loss": 0.2141, "step": 5253 }, { "epoch": 1.3980840872804683, "grad_norm": 0.30354228615760803, "learning_rate": 1.673354410128373e-07, "loss": 0.2033, "step": 5254 }, { "epoch": 1.3983501862692922, "grad_norm": 0.29635950922966003, "learning_rate": 1.673228265768273e-07, "loss": 0.205, "step": 5255 }, { "epoch": 1.398616285258116, "grad_norm": 0.27531978487968445, "learning_rate": 1.673102101812246e-07, "loss": 0.2088, "step": 5256 }, { "epoch": 1.3988823842469398, "grad_norm": 0.29124879837036133, "learning_rate": 1.6729759182639646e-07, "loss": 0.2017, "step": 5257 }, { "epoch": 1.3991484832357637, "grad_norm": 0.27946802973747253, "learning_rate": 1.6728497151271016e-07, "loss": 0.2277, "step": 5258 }, { "epoch": 1.3994145822245876, "grad_norm": 0.5170531272888184, "learning_rate": 1.6727234924053307e-07, "loss": 0.2131, "step": 5259 }, { "epoch": 1.3996806812134115, "grad_norm": 0.2981695830821991, "learning_rate": 1.6725972501023251e-07, "loss": 0.2064, "step": 5260 }, { "epoch": 1.3999467802022352, "grad_norm": 0.2854890823364258, "learning_rate": 1.6724709882217602e-07, "loss": 0.2022, "step": 5261 }, { "epoch": 1.400212879191059, "grad_norm": 0.2699056565761566, "learning_rate": 1.672344706767311e-07, "loss": 0.2116, "step": 5262 }, { "epoch": 1.400478978179883, "grad_norm": 0.2997119426727295, "learning_rate": 1.6722184057426532e-07, "loss": 0.2112, "step": 5263 }, { "epoch": 1.4007450771687067, "grad_norm": 0.2686768174171448, "learning_rate": 1.6720920851514629e-07, "loss": 0.1899, "step": 5264 }, { "epoch": 1.4010111761575306, "grad_norm": 0.29776039719581604, "learning_rate": 1.6719657449974168e-07, "loss": 0.2002, "step": 5265 }, { "epoch": 1.4012772751463545, "grad_norm": 0.37796011567115784, "learning_rate": 1.6718393852841926e-07, "loss": 0.2323, "step": 5266 }, { "epoch": 1.4015433741351782, "grad_norm": 0.2831801176071167, "learning_rate": 1.6717130060154688e-07, "loss": 0.2073, "step": 5267 }, { "epoch": 1.401809473124002, "grad_norm": 0.3026335537433624, "learning_rate": 1.6715866071949232e-07, "loss": 0.2137, "step": 5268 }, { "epoch": 1.402075572112826, "grad_norm": 0.31760483980178833, "learning_rate": 1.6714601888262352e-07, "loss": 0.2189, "step": 5269 }, { "epoch": 1.4023416711016499, "grad_norm": 0.2830696403980255, "learning_rate": 1.6713337509130846e-07, "loss": 0.1992, "step": 5270 }, { "epoch": 1.4026077700904738, "grad_norm": 0.33031103014945984, "learning_rate": 1.6712072934591515e-07, "loss": 0.2322, "step": 5271 }, { "epoch": 1.4028738690792975, "grad_norm": 0.29760515689849854, "learning_rate": 1.671080816468117e-07, "loss": 0.2246, "step": 5272 }, { "epoch": 1.4031399680681214, "grad_norm": 0.2944144606590271, "learning_rate": 1.6709543199436623e-07, "loss": 0.1986, "step": 5273 }, { "epoch": 1.4034060670569453, "grad_norm": 0.29317960143089294, "learning_rate": 1.6708278038894693e-07, "loss": 0.2225, "step": 5274 }, { "epoch": 1.403672166045769, "grad_norm": 0.2838421165943146, "learning_rate": 1.6707012683092207e-07, "loss": 0.2107, "step": 5275 }, { "epoch": 1.4039382650345928, "grad_norm": 0.2991918921470642, "learning_rate": 1.6705747132065997e-07, "loss": 0.231, "step": 5276 }, { "epoch": 1.4042043640234168, "grad_norm": 0.27279895544052124, "learning_rate": 1.6704481385852902e-07, "loss": 0.2189, "step": 5277 }, { "epoch": 1.4044704630122404, "grad_norm": 0.28226664662361145, "learning_rate": 1.670321544448976e-07, "loss": 0.2177, "step": 5278 }, { "epoch": 1.4047365620010643, "grad_norm": 0.29736649990081787, "learning_rate": 1.6701949308013422e-07, "loss": 0.2108, "step": 5279 }, { "epoch": 1.4050026609898882, "grad_norm": 0.31221088767051697, "learning_rate": 1.670068297646074e-07, "loss": 0.2153, "step": 5280 }, { "epoch": 1.4052687599787121, "grad_norm": 0.2935565412044525, "learning_rate": 1.6699416449868578e-07, "loss": 0.2102, "step": 5281 }, { "epoch": 1.405534858967536, "grad_norm": 0.2972743809223175, "learning_rate": 1.6698149728273794e-07, "loss": 0.2103, "step": 5282 }, { "epoch": 1.4058009579563597, "grad_norm": 0.2841205298900604, "learning_rate": 1.6696882811713264e-07, "loss": 0.2154, "step": 5283 }, { "epoch": 1.4060670569451836, "grad_norm": 0.3116636276245117, "learning_rate": 1.6695615700223865e-07, "loss": 0.2144, "step": 5284 }, { "epoch": 1.4063331559340075, "grad_norm": 0.27709466218948364, "learning_rate": 1.6694348393842475e-07, "loss": 0.2033, "step": 5285 }, { "epoch": 1.4065992549228312, "grad_norm": 0.29570192098617554, "learning_rate": 1.669308089260599e-07, "loss": 0.2, "step": 5286 }, { "epoch": 1.4068653539116551, "grad_norm": 0.3044373393058777, "learning_rate": 1.669181319655129e-07, "loss": 0.231, "step": 5287 }, { "epoch": 1.407131452900479, "grad_norm": 0.28513920307159424, "learning_rate": 1.6690545305715289e-07, "loss": 0.212, "step": 5288 }, { "epoch": 1.4073975518893027, "grad_norm": 0.2869722843170166, "learning_rate": 1.6689277220134884e-07, "loss": 0.2243, "step": 5289 }, { "epoch": 1.4076636508781266, "grad_norm": 0.2807918190956116, "learning_rate": 1.6688008939846992e-07, "loss": 0.2113, "step": 5290 }, { "epoch": 1.4079297498669505, "grad_norm": 0.2621678113937378, "learning_rate": 1.668674046488852e-07, "loss": 0.1972, "step": 5291 }, { "epoch": 1.4081958488557744, "grad_norm": 0.29771679639816284, "learning_rate": 1.6685471795296397e-07, "loss": 0.2094, "step": 5292 }, { "epoch": 1.4084619478445983, "grad_norm": 0.2908155620098114, "learning_rate": 1.668420293110755e-07, "loss": 0.2077, "step": 5293 }, { "epoch": 1.408728046833422, "grad_norm": 0.2975737750530243, "learning_rate": 1.6682933872358912e-07, "loss": 0.2269, "step": 5294 }, { "epoch": 1.408994145822246, "grad_norm": 0.2943432927131653, "learning_rate": 1.6681664619087415e-07, "loss": 0.2006, "step": 5295 }, { "epoch": 1.4092602448110698, "grad_norm": 0.28289011120796204, "learning_rate": 1.6680395171330015e-07, "loss": 0.2098, "step": 5296 }, { "epoch": 1.4095263437998935, "grad_norm": 0.2894948422908783, "learning_rate": 1.6679125529123656e-07, "loss": 0.2246, "step": 5297 }, { "epoch": 1.4097924427887174, "grad_norm": 0.2845093607902527, "learning_rate": 1.6677855692505293e-07, "loss": 0.2104, "step": 5298 }, { "epoch": 1.4100585417775413, "grad_norm": 0.2889135479927063, "learning_rate": 1.6676585661511893e-07, "loss": 0.1997, "step": 5299 }, { "epoch": 1.410324640766365, "grad_norm": 0.2854629158973694, "learning_rate": 1.6675315436180418e-07, "loss": 0.2176, "step": 5300 }, { "epoch": 1.410590739755189, "grad_norm": 0.30804356932640076, "learning_rate": 1.667404501654784e-07, "loss": 0.2099, "step": 5301 }, { "epoch": 1.4108568387440128, "grad_norm": 0.28064876794815063, "learning_rate": 1.6672774402651145e-07, "loss": 0.1921, "step": 5302 }, { "epoch": 1.4111229377328365, "grad_norm": 0.2918589115142822, "learning_rate": 1.6671503594527311e-07, "loss": 0.2119, "step": 5303 }, { "epoch": 1.4113890367216604, "grad_norm": 0.28730401396751404, "learning_rate": 1.6670232592213332e-07, "loss": 0.2118, "step": 5304 }, { "epoch": 1.4116551357104843, "grad_norm": 0.293662965297699, "learning_rate": 1.66689613957462e-07, "loss": 0.1983, "step": 5305 }, { "epoch": 1.4119212346993082, "grad_norm": 0.30369576811790466, "learning_rate": 1.6667690005162916e-07, "loss": 0.2228, "step": 5306 }, { "epoch": 1.412187333688132, "grad_norm": 0.288738489151001, "learning_rate": 1.666641842050049e-07, "loss": 0.2077, "step": 5307 }, { "epoch": 1.4124534326769558, "grad_norm": 0.2687024474143982, "learning_rate": 1.666514664179593e-07, "loss": 0.1978, "step": 5308 }, { "epoch": 1.4127195316657797, "grad_norm": 0.5304100513458252, "learning_rate": 1.6663874669086262e-07, "loss": 0.2476, "step": 5309 }, { "epoch": 1.4129856306546036, "grad_norm": 0.28702932596206665, "learning_rate": 1.66626025024085e-07, "loss": 0.2008, "step": 5310 }, { "epoch": 1.4132517296434273, "grad_norm": 0.29906466603279114, "learning_rate": 1.666133014179968e-07, "loss": 0.2099, "step": 5311 }, { "epoch": 1.4135178286322512, "grad_norm": 0.29467493295669556, "learning_rate": 1.6660057587296837e-07, "loss": 0.2086, "step": 5312 }, { "epoch": 1.413783927621075, "grad_norm": 0.2885640263557434, "learning_rate": 1.6658784838937013e-07, "loss": 0.2224, "step": 5313 }, { "epoch": 1.4140500266098988, "grad_norm": 0.3528583347797394, "learning_rate": 1.6657511896757248e-07, "loss": 0.2355, "step": 5314 }, { "epoch": 1.4143161255987227, "grad_norm": 0.2812310457229614, "learning_rate": 1.6656238760794597e-07, "loss": 0.1919, "step": 5315 }, { "epoch": 1.4145822245875466, "grad_norm": 0.28021401166915894, "learning_rate": 1.6654965431086118e-07, "loss": 0.2195, "step": 5316 }, { "epoch": 1.4148483235763705, "grad_norm": 0.2972033619880676, "learning_rate": 1.6653691907668878e-07, "loss": 0.2115, "step": 5317 }, { "epoch": 1.4151144225651944, "grad_norm": 0.28146791458129883, "learning_rate": 1.6652418190579942e-07, "loss": 0.2052, "step": 5318 }, { "epoch": 1.415380521554018, "grad_norm": 0.265450119972229, "learning_rate": 1.6651144279856383e-07, "loss": 0.191, "step": 5319 }, { "epoch": 1.415646620542842, "grad_norm": 0.3057537376880646, "learning_rate": 1.6649870175535287e-07, "loss": 0.2091, "step": 5320 }, { "epoch": 1.4159127195316659, "grad_norm": 0.28703761100769043, "learning_rate": 1.6648595877653734e-07, "loss": 0.2052, "step": 5321 }, { "epoch": 1.4161788185204895, "grad_norm": 0.3158222734928131, "learning_rate": 1.664732138624882e-07, "loss": 0.2199, "step": 5322 }, { "epoch": 1.4164449175093135, "grad_norm": 0.2752878665924072, "learning_rate": 1.6646046701357637e-07, "loss": 0.2268, "step": 5323 }, { "epoch": 1.4167110164981374, "grad_norm": 0.29177311062812805, "learning_rate": 1.664477182301729e-07, "loss": 0.2014, "step": 5324 }, { "epoch": 1.416977115486961, "grad_norm": 0.3084959089756012, "learning_rate": 1.664349675126489e-07, "loss": 0.2158, "step": 5325 }, { "epoch": 1.417243214475785, "grad_norm": 0.28971418738365173, "learning_rate": 1.664222148613755e-07, "loss": 0.2139, "step": 5326 }, { "epoch": 1.4175093134646088, "grad_norm": 0.3167347311973572, "learning_rate": 1.664094602767239e-07, "loss": 0.221, "step": 5327 }, { "epoch": 1.4177754124534327, "grad_norm": 0.2758405804634094, "learning_rate": 1.663967037590653e-07, "loss": 0.1951, "step": 5328 }, { "epoch": 1.4180415114422567, "grad_norm": 0.2606692314147949, "learning_rate": 1.6638394530877112e-07, "loss": 0.1938, "step": 5329 }, { "epoch": 1.4183076104310803, "grad_norm": 0.2847763001918793, "learning_rate": 1.6637118492621261e-07, "loss": 0.211, "step": 5330 }, { "epoch": 1.4185737094199042, "grad_norm": 0.25947117805480957, "learning_rate": 1.6635842261176122e-07, "loss": 0.183, "step": 5331 }, { "epoch": 1.4188398084087281, "grad_norm": 0.29528915882110596, "learning_rate": 1.6634565836578845e-07, "loss": 0.2073, "step": 5332 }, { "epoch": 1.4191059073975518, "grad_norm": 0.2716302275657654, "learning_rate": 1.6633289218866586e-07, "loss": 0.2015, "step": 5333 }, { "epoch": 1.4193720063863757, "grad_norm": 0.3010428249835968, "learning_rate": 1.66320124080765e-07, "loss": 0.2252, "step": 5334 }, { "epoch": 1.4196381053751996, "grad_norm": 0.3266543745994568, "learning_rate": 1.6630735404245752e-07, "loss": 0.2108, "step": 5335 }, { "epoch": 1.4199042043640233, "grad_norm": 0.27730393409729004, "learning_rate": 1.6629458207411516e-07, "loss": 0.2032, "step": 5336 }, { "epoch": 1.4201703033528472, "grad_norm": 0.29017874598503113, "learning_rate": 1.662818081761096e-07, "loss": 0.2103, "step": 5337 }, { "epoch": 1.4204364023416711, "grad_norm": 0.28360462188720703, "learning_rate": 1.6626903234881272e-07, "loss": 0.215, "step": 5338 }, { "epoch": 1.420702501330495, "grad_norm": 0.30843567848205566, "learning_rate": 1.6625625459259637e-07, "loss": 0.2167, "step": 5339 }, { "epoch": 1.4209686003193187, "grad_norm": 0.2427082657814026, "learning_rate": 1.6624347490783254e-07, "loss": 0.192, "step": 5340 }, { "epoch": 1.4212346993081426, "grad_norm": 0.2815459966659546, "learning_rate": 1.6623069329489306e-07, "loss": 0.2097, "step": 5341 }, { "epoch": 1.4215007982969665, "grad_norm": 0.29408079385757446, "learning_rate": 1.6621790975415012e-07, "loss": 0.2281, "step": 5342 }, { "epoch": 1.4217668972857904, "grad_norm": 0.296541303396225, "learning_rate": 1.6620512428597574e-07, "loss": 0.2157, "step": 5343 }, { "epoch": 1.422032996274614, "grad_norm": 0.2920411229133606, "learning_rate": 1.661923368907421e-07, "loss": 0.2167, "step": 5344 }, { "epoch": 1.422299095263438, "grad_norm": 0.2880805432796478, "learning_rate": 1.6617954756882142e-07, "loss": 0.2095, "step": 5345 }, { "epoch": 1.422565194252262, "grad_norm": 0.27450239658355713, "learning_rate": 1.6616675632058593e-07, "loss": 0.2029, "step": 5346 }, { "epoch": 1.4228312932410856, "grad_norm": 0.25795477628707886, "learning_rate": 1.6615396314640792e-07, "loss": 0.2101, "step": 5347 }, { "epoch": 1.4230973922299095, "grad_norm": 0.2448965609073639, "learning_rate": 1.6614116804665985e-07, "loss": 0.1744, "step": 5348 }, { "epoch": 1.4233634912187334, "grad_norm": 0.2972310483455658, "learning_rate": 1.661283710217141e-07, "loss": 0.208, "step": 5349 }, { "epoch": 1.423629590207557, "grad_norm": 0.27945446968078613, "learning_rate": 1.661155720719432e-07, "loss": 0.2209, "step": 5350 }, { "epoch": 1.423895689196381, "grad_norm": 0.29641857743263245, "learning_rate": 1.6610277119771965e-07, "loss": 0.215, "step": 5351 }, { "epoch": 1.424161788185205, "grad_norm": 0.2967408001422882, "learning_rate": 1.6608996839941605e-07, "loss": 0.2276, "step": 5352 }, { "epoch": 1.4244278871740288, "grad_norm": 0.3129050135612488, "learning_rate": 1.6607716367740506e-07, "loss": 0.2178, "step": 5353 }, { "epoch": 1.4246939861628527, "grad_norm": 0.29455068707466125, "learning_rate": 1.6606435703205945e-07, "loss": 0.2202, "step": 5354 }, { "epoch": 1.4249600851516764, "grad_norm": 0.30798593163490295, "learning_rate": 1.660515484637519e-07, "loss": 0.2313, "step": 5355 }, { "epoch": 1.4252261841405003, "grad_norm": 0.2842120826244354, "learning_rate": 1.660387379728553e-07, "loss": 0.2128, "step": 5356 }, { "epoch": 1.4254922831293242, "grad_norm": 0.696979284286499, "learning_rate": 1.6602592555974248e-07, "loss": 0.2113, "step": 5357 }, { "epoch": 1.4257583821181479, "grad_norm": 0.2628174424171448, "learning_rate": 1.6601311122478642e-07, "loss": 0.1871, "step": 5358 }, { "epoch": 1.4260244811069718, "grad_norm": 0.29515787959098816, "learning_rate": 1.660002949683601e-07, "loss": 0.2266, "step": 5359 }, { "epoch": 1.4262905800957957, "grad_norm": 0.25637319684028625, "learning_rate": 1.6598747679083657e-07, "loss": 0.1839, "step": 5360 }, { "epoch": 1.4265566790846194, "grad_norm": 0.3084665536880493, "learning_rate": 1.6597465669258888e-07, "loss": 0.2233, "step": 5361 }, { "epoch": 1.4268227780734433, "grad_norm": 0.28523021936416626, "learning_rate": 1.6596183467399025e-07, "loss": 0.2133, "step": 5362 }, { "epoch": 1.4270888770622672, "grad_norm": 0.308692991733551, "learning_rate": 1.6594901073541392e-07, "loss": 0.2193, "step": 5363 }, { "epoch": 1.427354976051091, "grad_norm": 0.3065047562122345, "learning_rate": 1.6593618487723308e-07, "loss": 0.2329, "step": 5364 }, { "epoch": 1.427621075039915, "grad_norm": 0.26541852951049805, "learning_rate": 1.6592335709982113e-07, "loss": 0.1955, "step": 5365 }, { "epoch": 1.4278871740287387, "grad_norm": 0.28159022331237793, "learning_rate": 1.659105274035514e-07, "loss": 0.2013, "step": 5366 }, { "epoch": 1.4281532730175626, "grad_norm": 0.2879503071308136, "learning_rate": 1.6589769578879737e-07, "loss": 0.2189, "step": 5367 }, { "epoch": 1.4284193720063865, "grad_norm": 0.28286877274513245, "learning_rate": 1.6588486225593247e-07, "loss": 0.213, "step": 5368 }, { "epoch": 1.4286854709952101, "grad_norm": 0.2764381766319275, "learning_rate": 1.6587202680533037e-07, "loss": 0.2046, "step": 5369 }, { "epoch": 1.428951569984034, "grad_norm": 0.2735787332057953, "learning_rate": 1.6585918943736456e-07, "loss": 0.1938, "step": 5370 }, { "epoch": 1.429217668972858, "grad_norm": 0.2702314853668213, "learning_rate": 1.6584635015240874e-07, "loss": 0.2051, "step": 5371 }, { "epoch": 1.4294837679616816, "grad_norm": 0.31649231910705566, "learning_rate": 1.6583350895083667e-07, "loss": 0.1903, "step": 5372 }, { "epoch": 1.4297498669505055, "grad_norm": 0.27915337681770325, "learning_rate": 1.6582066583302208e-07, "loss": 0.2073, "step": 5373 }, { "epoch": 1.4300159659393294, "grad_norm": 0.2750610113143921, "learning_rate": 1.658078207993388e-07, "loss": 0.1945, "step": 5374 }, { "epoch": 1.4302820649281534, "grad_norm": 0.2821894884109497, "learning_rate": 1.6579497385016071e-07, "loss": 0.2147, "step": 5375 }, { "epoch": 1.430548163916977, "grad_norm": 0.28322821855545044, "learning_rate": 1.657821249858618e-07, "loss": 0.2149, "step": 5376 }, { "epoch": 1.430814262905801, "grad_norm": 0.2992863953113556, "learning_rate": 1.65769274206816e-07, "loss": 0.2225, "step": 5377 }, { "epoch": 1.4310803618946248, "grad_norm": 0.2698978781700134, "learning_rate": 1.657564215133974e-07, "loss": 0.221, "step": 5378 }, { "epoch": 1.4313464608834487, "grad_norm": 0.278093159198761, "learning_rate": 1.6574356690598013e-07, "loss": 0.2278, "step": 5379 }, { "epoch": 1.4316125598722724, "grad_norm": 0.2948463261127472, "learning_rate": 1.657307103849383e-07, "loss": 0.2142, "step": 5380 }, { "epoch": 1.4318786588610963, "grad_norm": 0.2917446196079254, "learning_rate": 1.657178519506462e-07, "loss": 0.2278, "step": 5381 }, { "epoch": 1.4321447578499202, "grad_norm": 0.3375149071216583, "learning_rate": 1.6570499160347803e-07, "loss": 0.2055, "step": 5382 }, { "epoch": 1.432410856838744, "grad_norm": 0.3090900480747223, "learning_rate": 1.6569212934380815e-07, "loss": 0.226, "step": 5383 }, { "epoch": 1.4326769558275678, "grad_norm": 0.28708145022392273, "learning_rate": 1.6567926517201097e-07, "loss": 0.2038, "step": 5384 }, { "epoch": 1.4329430548163917, "grad_norm": 0.27911293506622314, "learning_rate": 1.656663990884609e-07, "loss": 0.2063, "step": 5385 }, { "epoch": 1.4332091538052154, "grad_norm": 0.2841685712337494, "learning_rate": 1.6565353109353248e-07, "loss": 0.2016, "step": 5386 }, { "epoch": 1.4334752527940393, "grad_norm": 0.30223989486694336, "learning_rate": 1.6564066118760023e-07, "loss": 0.2223, "step": 5387 }, { "epoch": 1.4337413517828632, "grad_norm": 0.2675776779651642, "learning_rate": 1.6562778937103874e-07, "loss": 0.2074, "step": 5388 }, { "epoch": 1.4340074507716871, "grad_norm": 0.33137670159339905, "learning_rate": 1.6561491564422271e-07, "loss": 0.2327, "step": 5389 }, { "epoch": 1.434273549760511, "grad_norm": 0.29342174530029297, "learning_rate": 1.6560204000752688e-07, "loss": 0.2192, "step": 5390 }, { "epoch": 1.4345396487493347, "grad_norm": 0.26800402998924255, "learning_rate": 1.65589162461326e-07, "loss": 0.2091, "step": 5391 }, { "epoch": 1.4348057477381586, "grad_norm": 0.2682872414588928, "learning_rate": 1.655762830059949e-07, "loss": 0.208, "step": 5392 }, { "epoch": 1.4350718467269825, "grad_norm": 0.2602349519729614, "learning_rate": 1.6556340164190842e-07, "loss": 0.2012, "step": 5393 }, { "epoch": 1.4353379457158062, "grad_norm": 0.29095184803009033, "learning_rate": 1.655505183694416e-07, "loss": 0.2065, "step": 5394 }, { "epoch": 1.43560404470463, "grad_norm": 0.2789914011955261, "learning_rate": 1.6553763318896937e-07, "loss": 0.2124, "step": 5395 }, { "epoch": 1.435870143693454, "grad_norm": 0.2715967893600464, "learning_rate": 1.6552474610086683e-07, "loss": 0.2009, "step": 5396 }, { "epoch": 1.4361362426822777, "grad_norm": 0.2817920446395874, "learning_rate": 1.65511857105509e-07, "loss": 0.2169, "step": 5397 }, { "epoch": 1.4364023416711016, "grad_norm": 0.29363158345222473, "learning_rate": 1.6549896620327118e-07, "loss": 0.2052, "step": 5398 }, { "epoch": 1.4366684406599255, "grad_norm": 0.28778237104415894, "learning_rate": 1.6548607339452852e-07, "loss": 0.2195, "step": 5399 }, { "epoch": 1.4369345396487494, "grad_norm": 0.3094518780708313, "learning_rate": 1.6547317867965625e-07, "loss": 0.2268, "step": 5400 }, { "epoch": 1.4372006386375733, "grad_norm": 0.2733575105667114, "learning_rate": 1.6546028205902977e-07, "loss": 0.2072, "step": 5401 }, { "epoch": 1.437466737626397, "grad_norm": 0.2922941744327545, "learning_rate": 1.6544738353302444e-07, "loss": 0.2008, "step": 5402 }, { "epoch": 1.4377328366152209, "grad_norm": 0.3208385407924652, "learning_rate": 1.6543448310201572e-07, "loss": 0.2183, "step": 5403 }, { "epoch": 1.4379989356040448, "grad_norm": 0.2796511948108673, "learning_rate": 1.6542158076637907e-07, "loss": 0.1944, "step": 5404 }, { "epoch": 1.4382650345928685, "grad_norm": 0.30702072381973267, "learning_rate": 1.6540867652649012e-07, "loss": 0.2062, "step": 5405 }, { "epoch": 1.4385311335816924, "grad_norm": 0.2963249683380127, "learning_rate": 1.653957703827244e-07, "loss": 0.2401, "step": 5406 }, { "epoch": 1.4387972325705163, "grad_norm": 0.29039251804351807, "learning_rate": 1.6538286233545757e-07, "loss": 0.2143, "step": 5407 }, { "epoch": 1.43906333155934, "grad_norm": 0.28437310457229614, "learning_rate": 1.6536995238506543e-07, "loss": 0.2124, "step": 5408 }, { "epoch": 1.4393294305481639, "grad_norm": 0.25848737359046936, "learning_rate": 1.653570405319237e-07, "loss": 0.2065, "step": 5409 }, { "epoch": 1.4395955295369878, "grad_norm": 0.2803008556365967, "learning_rate": 1.653441267764082e-07, "loss": 0.2025, "step": 5410 }, { "epoch": 1.4398616285258117, "grad_norm": 0.3085046708583832, "learning_rate": 1.6533121111889482e-07, "loss": 0.2076, "step": 5411 }, { "epoch": 1.4401277275146356, "grad_norm": 0.2618398666381836, "learning_rate": 1.6531829355975956e-07, "loss": 0.205, "step": 5412 }, { "epoch": 1.4403938265034593, "grad_norm": 0.30979016423225403, "learning_rate": 1.6530537409937834e-07, "loss": 0.2216, "step": 5413 }, { "epoch": 1.4406599254922832, "grad_norm": 0.29397064447402954, "learning_rate": 1.6529245273812724e-07, "loss": 0.2271, "step": 5414 }, { "epoch": 1.440926024481107, "grad_norm": 0.28515827655792236, "learning_rate": 1.6527952947638236e-07, "loss": 0.213, "step": 5415 }, { "epoch": 1.4411921234699308, "grad_norm": 0.301584392786026, "learning_rate": 1.6526660431451992e-07, "loss": 0.225, "step": 5416 }, { "epoch": 1.4414582224587547, "grad_norm": 0.28375130891799927, "learning_rate": 1.6525367725291607e-07, "loss": 0.1949, "step": 5417 }, { "epoch": 1.4417243214475786, "grad_norm": 0.2887692451477051, "learning_rate": 1.652407482919471e-07, "loss": 0.1998, "step": 5418 }, { "epoch": 1.4419904204364022, "grad_norm": 0.2662779688835144, "learning_rate": 1.6522781743198934e-07, "loss": 0.2085, "step": 5419 }, { "epoch": 1.4422565194252261, "grad_norm": 0.27253293991088867, "learning_rate": 1.652148846734192e-07, "loss": 0.2032, "step": 5420 }, { "epoch": 1.44252261841405, "grad_norm": 0.27142956852912903, "learning_rate": 1.6520195001661305e-07, "loss": 0.1924, "step": 5421 }, { "epoch": 1.4427887174028737, "grad_norm": 0.2866562008857727, "learning_rate": 1.6518901346194748e-07, "loss": 0.1979, "step": 5422 }, { "epoch": 1.4430548163916976, "grad_norm": 0.47674331068992615, "learning_rate": 1.6517607500979896e-07, "loss": 0.2111, "step": 5423 }, { "epoch": 1.4433209153805215, "grad_norm": 0.27835750579833984, "learning_rate": 1.6516313466054413e-07, "loss": 0.2065, "step": 5424 }, { "epoch": 1.4435870143693454, "grad_norm": 0.2933036983013153, "learning_rate": 1.6515019241455966e-07, "loss": 0.2163, "step": 5425 }, { "epoch": 1.4438531133581693, "grad_norm": 0.293502539396286, "learning_rate": 1.6513724827222223e-07, "loss": 0.217, "step": 5426 }, { "epoch": 1.444119212346993, "grad_norm": 0.275813490152359, "learning_rate": 1.6512430223390863e-07, "loss": 0.2044, "step": 5427 }, { "epoch": 1.444385311335817, "grad_norm": 0.26112863421440125, "learning_rate": 1.6511135429999569e-07, "loss": 0.193, "step": 5428 }, { "epoch": 1.4446514103246408, "grad_norm": 0.28099215030670166, "learning_rate": 1.6509840447086027e-07, "loss": 0.1965, "step": 5429 }, { "epoch": 1.4449175093134645, "grad_norm": 0.29460573196411133, "learning_rate": 1.6508545274687935e-07, "loss": 0.2242, "step": 5430 }, { "epoch": 1.4451836083022884, "grad_norm": 0.3088260591030121, "learning_rate": 1.6507249912842988e-07, "loss": 0.2391, "step": 5431 }, { "epoch": 1.4454497072911123, "grad_norm": 0.2926187217235565, "learning_rate": 1.6505954361588894e-07, "loss": 0.2119, "step": 5432 }, { "epoch": 1.445715806279936, "grad_norm": 0.30211523175239563, "learning_rate": 1.6504658620963358e-07, "loss": 0.2091, "step": 5433 }, { "epoch": 1.44598190526876, "grad_norm": 0.26746857166290283, "learning_rate": 1.65033626910041e-07, "loss": 0.1876, "step": 5434 }, { "epoch": 1.4462480042575838, "grad_norm": 0.28244295716285706, "learning_rate": 1.650206657174884e-07, "loss": 0.1931, "step": 5435 }, { "epoch": 1.4465141032464077, "grad_norm": 0.2928134500980377, "learning_rate": 1.6500770263235305e-07, "loss": 0.2177, "step": 5436 }, { "epoch": 1.4467802022352316, "grad_norm": 0.29668083786964417, "learning_rate": 1.6499473765501227e-07, "loss": 0.2082, "step": 5437 }, { "epoch": 1.4470463012240553, "grad_norm": 0.2822165787220001, "learning_rate": 1.6498177078584342e-07, "loss": 0.1945, "step": 5438 }, { "epoch": 1.4473124002128792, "grad_norm": 0.26843559741973877, "learning_rate": 1.6496880202522392e-07, "loss": 0.1982, "step": 5439 }, { "epoch": 1.4475784992017031, "grad_norm": 0.3057202994823456, "learning_rate": 1.6495583137353132e-07, "loss": 0.1984, "step": 5440 }, { "epoch": 1.4478445981905268, "grad_norm": 0.30381202697753906, "learning_rate": 1.6494285883114315e-07, "loss": 0.2082, "step": 5441 }, { "epoch": 1.4481106971793507, "grad_norm": 0.300847589969635, "learning_rate": 1.6492988439843692e-07, "loss": 0.2329, "step": 5442 }, { "epoch": 1.4483767961681746, "grad_norm": 0.3043680191040039, "learning_rate": 1.649169080757904e-07, "loss": 0.2073, "step": 5443 }, { "epoch": 1.4486428951569983, "grad_norm": 0.27926456928253174, "learning_rate": 1.649039298635812e-07, "loss": 0.2067, "step": 5444 }, { "epoch": 1.4489089941458222, "grad_norm": 0.29692941904067993, "learning_rate": 1.6489094976218713e-07, "loss": 0.2073, "step": 5445 }, { "epoch": 1.449175093134646, "grad_norm": 0.26133376359939575, "learning_rate": 1.6487796777198603e-07, "loss": 0.1924, "step": 5446 }, { "epoch": 1.44944119212347, "grad_norm": 0.28960078954696655, "learning_rate": 1.648649838933557e-07, "loss": 0.2296, "step": 5447 }, { "epoch": 1.449707291112294, "grad_norm": 0.3069096505641937, "learning_rate": 1.6485199812667414e-07, "loss": 0.2276, "step": 5448 }, { "epoch": 1.4499733901011176, "grad_norm": 0.2720983326435089, "learning_rate": 1.648390104723193e-07, "loss": 0.1969, "step": 5449 }, { "epoch": 1.4502394890899415, "grad_norm": 0.30615830421447754, "learning_rate": 1.648260209306692e-07, "loss": 0.2247, "step": 5450 }, { "epoch": 1.4505055880787654, "grad_norm": 0.27555611729621887, "learning_rate": 1.6481302950210197e-07, "loss": 0.1887, "step": 5451 }, { "epoch": 1.450771687067589, "grad_norm": 0.3400362432003021, "learning_rate": 1.6480003618699572e-07, "loss": 0.2104, "step": 5452 }, { "epoch": 1.451037786056413, "grad_norm": 0.24699819087982178, "learning_rate": 1.6478704098572868e-07, "loss": 0.1923, "step": 5453 }, { "epoch": 1.4513038850452369, "grad_norm": 0.4157356023788452, "learning_rate": 1.6477404389867905e-07, "loss": 0.2026, "step": 5454 }, { "epoch": 1.4515699840340606, "grad_norm": 0.34678274393081665, "learning_rate": 1.6476104492622524e-07, "loss": 0.2514, "step": 5455 }, { "epoch": 1.4518360830228845, "grad_norm": 0.29743435978889465, "learning_rate": 1.6474804406874555e-07, "loss": 0.2192, "step": 5456 }, { "epoch": 1.4521021820117084, "grad_norm": 0.27366960048675537, "learning_rate": 1.647350413266184e-07, "loss": 0.2028, "step": 5457 }, { "epoch": 1.4523682810005323, "grad_norm": 0.31441155076026917, "learning_rate": 1.6472203670022232e-07, "loss": 0.2124, "step": 5458 }, { "epoch": 1.452634379989356, "grad_norm": 0.2601028084754944, "learning_rate": 1.6470903018993576e-07, "loss": 0.2015, "step": 5459 }, { "epoch": 1.4529004789781799, "grad_norm": 0.2954716980457306, "learning_rate": 1.6469602179613736e-07, "loss": 0.2018, "step": 5460 }, { "epoch": 1.4531665779670038, "grad_norm": 0.28398022055625916, "learning_rate": 1.6468301151920574e-07, "loss": 0.2132, "step": 5461 }, { "epoch": 1.4534326769558277, "grad_norm": 0.2701387107372284, "learning_rate": 1.6466999935951962e-07, "loss": 0.1883, "step": 5462 }, { "epoch": 1.4536987759446514, "grad_norm": 0.2703888714313507, "learning_rate": 1.6465698531745772e-07, "loss": 0.2051, "step": 5463 }, { "epoch": 1.4539648749334753, "grad_norm": 0.2806493639945984, "learning_rate": 1.6464396939339884e-07, "loss": 0.1934, "step": 5464 }, { "epoch": 1.4542309739222992, "grad_norm": 0.2893238961696625, "learning_rate": 1.6463095158772188e-07, "loss": 0.2085, "step": 5465 }, { "epoch": 1.4544970729111228, "grad_norm": 0.2872348427772522, "learning_rate": 1.6461793190080574e-07, "loss": 0.2063, "step": 5466 }, { "epoch": 1.4547631718999467, "grad_norm": 0.26083099842071533, "learning_rate": 1.6460491033302937e-07, "loss": 0.1984, "step": 5467 }, { "epoch": 1.4550292708887707, "grad_norm": 0.27131161093711853, "learning_rate": 1.645918868847718e-07, "loss": 0.198, "step": 5468 }, { "epoch": 1.4552953698775943, "grad_norm": 0.3057652711868286, "learning_rate": 1.6457886155641208e-07, "loss": 0.2035, "step": 5469 }, { "epoch": 1.4555614688664182, "grad_norm": 0.3091725707054138, "learning_rate": 1.6456583434832943e-07, "loss": 0.2233, "step": 5470 }, { "epoch": 1.4558275678552421, "grad_norm": 0.2898697257041931, "learning_rate": 1.64552805260903e-07, "loss": 0.2254, "step": 5471 }, { "epoch": 1.456093666844066, "grad_norm": 0.29066088795661926, "learning_rate": 1.6453977429451197e-07, "loss": 0.228, "step": 5472 }, { "epoch": 1.45635976583289, "grad_norm": 0.2642955780029297, "learning_rate": 1.645267414495357e-07, "loss": 0.2048, "step": 5473 }, { "epoch": 1.4566258648217136, "grad_norm": 0.25655049085617065, "learning_rate": 1.6451370672635352e-07, "loss": 0.1928, "step": 5474 }, { "epoch": 1.4568919638105375, "grad_norm": 0.27842020988464355, "learning_rate": 1.6450067012534488e-07, "loss": 0.2096, "step": 5475 }, { "epoch": 1.4571580627993614, "grad_norm": 0.2984466850757599, "learning_rate": 1.644876316468892e-07, "loss": 0.2121, "step": 5476 }, { "epoch": 1.4574241617881851, "grad_norm": 0.29901042580604553, "learning_rate": 1.6447459129136596e-07, "loss": 0.215, "step": 5477 }, { "epoch": 1.457690260777009, "grad_norm": 0.2855205833911896, "learning_rate": 1.644615490591548e-07, "loss": 0.1937, "step": 5478 }, { "epoch": 1.457956359765833, "grad_norm": 0.2834131717681885, "learning_rate": 1.6444850495063531e-07, "loss": 0.2113, "step": 5479 }, { "epoch": 1.4582224587546566, "grad_norm": 0.2778991460800171, "learning_rate": 1.644354589661872e-07, "loss": 0.2, "step": 5480 }, { "epoch": 1.4584885577434805, "grad_norm": 0.2887609302997589, "learning_rate": 1.644224111061902e-07, "loss": 0.2019, "step": 5481 }, { "epoch": 1.4587546567323044, "grad_norm": 0.2643478214740753, "learning_rate": 1.6440936137102405e-07, "loss": 0.2094, "step": 5482 }, { "epoch": 1.4590207557211283, "grad_norm": 0.3099418580532074, "learning_rate": 1.6439630976106865e-07, "loss": 0.2113, "step": 5483 }, { "epoch": 1.4592868547099522, "grad_norm": 0.2913157343864441, "learning_rate": 1.6438325627670385e-07, "loss": 0.2254, "step": 5484 }, { "epoch": 1.459552953698776, "grad_norm": 0.30919191241264343, "learning_rate": 1.6437020091830968e-07, "loss": 0.228, "step": 5485 }, { "epoch": 1.4598190526875998, "grad_norm": 0.2786180377006531, "learning_rate": 1.6435714368626602e-07, "loss": 0.2086, "step": 5486 }, { "epoch": 1.4600851516764237, "grad_norm": 0.2961767911911011, "learning_rate": 1.6434408458095305e-07, "loss": 0.2209, "step": 5487 }, { "epoch": 1.4603512506652474, "grad_norm": 0.3364708721637726, "learning_rate": 1.6433102360275086e-07, "loss": 0.229, "step": 5488 }, { "epoch": 1.4606173496540713, "grad_norm": 0.2952219843864441, "learning_rate": 1.643179607520396e-07, "loss": 0.2206, "step": 5489 }, { "epoch": 1.4608834486428952, "grad_norm": 0.30043551325798035, "learning_rate": 1.6430489602919948e-07, "loss": 0.2296, "step": 5490 }, { "epoch": 1.461149547631719, "grad_norm": 0.26665884256362915, "learning_rate": 1.6429182943461082e-07, "loss": 0.1845, "step": 5491 }, { "epoch": 1.4614156466205428, "grad_norm": 0.2935403287410736, "learning_rate": 1.6427876096865392e-07, "loss": 0.209, "step": 5492 }, { "epoch": 1.4616817456093667, "grad_norm": 0.31302523612976074, "learning_rate": 1.6426569063170922e-07, "loss": 0.2279, "step": 5493 }, { "epoch": 1.4619478445981906, "grad_norm": 0.32135412096977234, "learning_rate": 1.642526184241571e-07, "loss": 0.2293, "step": 5494 }, { "epoch": 1.4622139435870143, "grad_norm": 0.28292611241340637, "learning_rate": 1.642395443463781e-07, "loss": 0.205, "step": 5495 }, { "epoch": 1.4624800425758382, "grad_norm": 0.2995160222053528, "learning_rate": 1.6422646839875276e-07, "loss": 0.1952, "step": 5496 }, { "epoch": 1.462746141564662, "grad_norm": 0.3162403106689453, "learning_rate": 1.6421339058166168e-07, "loss": 0.2134, "step": 5497 }, { "epoch": 1.463012240553486, "grad_norm": 0.28685879707336426, "learning_rate": 1.6420031089548555e-07, "loss": 0.2191, "step": 5498 }, { "epoch": 1.4632783395423097, "grad_norm": 0.28708159923553467, "learning_rate": 1.6418722934060502e-07, "loss": 0.2057, "step": 5499 }, { "epoch": 1.4635444385311336, "grad_norm": 0.28343862295150757, "learning_rate": 1.6417414591740095e-07, "loss": 0.194, "step": 5500 }, { "epoch": 1.4638105375199575, "grad_norm": 0.3033439815044403, "learning_rate": 1.641610606262541e-07, "loss": 0.209, "step": 5501 }, { "epoch": 1.4640766365087812, "grad_norm": 0.2589016854763031, "learning_rate": 1.6414797346754536e-07, "loss": 0.2008, "step": 5502 }, { "epoch": 1.464342735497605, "grad_norm": 0.289221853017807, "learning_rate": 1.641348844416557e-07, "loss": 0.2195, "step": 5503 }, { "epoch": 1.464608834486429, "grad_norm": 0.30511510372161865, "learning_rate": 1.6412179354896606e-07, "loss": 0.2151, "step": 5504 }, { "epoch": 1.4648749334752527, "grad_norm": 0.3146350383758545, "learning_rate": 1.6410870078985748e-07, "loss": 0.2135, "step": 5505 }, { "epoch": 1.4651410324640766, "grad_norm": 0.3076663911342621, "learning_rate": 1.640956061647111e-07, "loss": 0.2207, "step": 5506 }, { "epoch": 1.4654071314529005, "grad_norm": 0.3056347072124481, "learning_rate": 1.6408250967390804e-07, "loss": 0.23, "step": 5507 }, { "epoch": 1.4656732304417244, "grad_norm": 0.2644990086555481, "learning_rate": 1.6406941131782952e-07, "loss": 0.198, "step": 5508 }, { "epoch": 1.4659393294305483, "grad_norm": 0.27830806374549866, "learning_rate": 1.6405631109685677e-07, "loss": 0.2142, "step": 5509 }, { "epoch": 1.466205428419372, "grad_norm": 0.27521926164627075, "learning_rate": 1.6404320901137115e-07, "loss": 0.2012, "step": 5510 }, { "epoch": 1.4664715274081959, "grad_norm": 0.28867945075035095, "learning_rate": 1.64030105061754e-07, "loss": 0.2265, "step": 5511 }, { "epoch": 1.4667376263970198, "grad_norm": 0.2872210144996643, "learning_rate": 1.6401699924838672e-07, "loss": 0.2252, "step": 5512 }, { "epoch": 1.4670037253858434, "grad_norm": 0.2788911759853363, "learning_rate": 1.6400389157165086e-07, "loss": 0.1939, "step": 5513 }, { "epoch": 1.4672698243746674, "grad_norm": 0.26788029074668884, "learning_rate": 1.6399078203192785e-07, "loss": 0.1733, "step": 5514 }, { "epoch": 1.4675359233634913, "grad_norm": 0.32885292172431946, "learning_rate": 1.6397767062959933e-07, "loss": 0.2337, "step": 5515 }, { "epoch": 1.467802022352315, "grad_norm": 0.28035807609558105, "learning_rate": 1.6396455736504696e-07, "loss": 0.205, "step": 5516 }, { "epoch": 1.4680681213411388, "grad_norm": 0.2898057699203491, "learning_rate": 1.639514422386524e-07, "loss": 0.1976, "step": 5517 }, { "epoch": 1.4683342203299627, "grad_norm": 0.287455677986145, "learning_rate": 1.639383252507974e-07, "loss": 0.2052, "step": 5518 }, { "epoch": 1.4686003193187867, "grad_norm": 0.26801279187202454, "learning_rate": 1.6392520640186373e-07, "loss": 0.1901, "step": 5519 }, { "epoch": 1.4688664183076106, "grad_norm": 0.2756821811199188, "learning_rate": 1.6391208569223334e-07, "loss": 0.215, "step": 5520 }, { "epoch": 1.4691325172964342, "grad_norm": 0.27861282229423523, "learning_rate": 1.6389896312228805e-07, "loss": 0.1899, "step": 5521 }, { "epoch": 1.4693986162852581, "grad_norm": 0.27858853340148926, "learning_rate": 1.6388583869240987e-07, "loss": 0.1983, "step": 5522 }, { "epoch": 1.469664715274082, "grad_norm": 0.27611446380615234, "learning_rate": 1.6387271240298078e-07, "loss": 0.211, "step": 5523 }, { "epoch": 1.4699308142629057, "grad_norm": 0.26367759704589844, "learning_rate": 1.638595842543829e-07, "loss": 0.2117, "step": 5524 }, { "epoch": 1.4701969132517296, "grad_norm": 0.26187989115715027, "learning_rate": 1.6384645424699836e-07, "loss": 0.1988, "step": 5525 }, { "epoch": 1.4704630122405535, "grad_norm": 0.27679064869880676, "learning_rate": 1.6383332238120925e-07, "loss": 0.2156, "step": 5526 }, { "epoch": 1.4707291112293772, "grad_norm": 0.3116190433502197, "learning_rate": 1.638201886573979e-07, "loss": 0.2335, "step": 5527 }, { "epoch": 1.4709952102182011, "grad_norm": 0.2800236940383911, "learning_rate": 1.6380705307594655e-07, "loss": 0.1956, "step": 5528 }, { "epoch": 1.471261309207025, "grad_norm": 0.2981015145778656, "learning_rate": 1.6379391563723758e-07, "loss": 0.2007, "step": 5529 }, { "epoch": 1.471527408195849, "grad_norm": 0.2906750738620758, "learning_rate": 1.6378077634165337e-07, "loss": 0.2215, "step": 5530 }, { "epoch": 1.4717935071846728, "grad_norm": 0.2556305527687073, "learning_rate": 1.6376763518957634e-07, "loss": 0.1964, "step": 5531 }, { "epoch": 1.4720596061734965, "grad_norm": 0.31945541501045227, "learning_rate": 1.6375449218138904e-07, "loss": 0.2195, "step": 5532 }, { "epoch": 1.4723257051623204, "grad_norm": 0.3096890449523926, "learning_rate": 1.6374134731747398e-07, "loss": 0.2115, "step": 5533 }, { "epoch": 1.4725918041511443, "grad_norm": 0.2674734890460968, "learning_rate": 1.6372820059821385e-07, "loss": 0.1951, "step": 5534 }, { "epoch": 1.472857903139968, "grad_norm": 0.2881351411342621, "learning_rate": 1.6371505202399124e-07, "loss": 0.1968, "step": 5535 }, { "epoch": 1.473124002128792, "grad_norm": 0.29794541001319885, "learning_rate": 1.6370190159518888e-07, "loss": 0.2136, "step": 5536 }, { "epoch": 1.4733901011176158, "grad_norm": 0.28169283270835876, "learning_rate": 1.6368874931218958e-07, "loss": 0.2045, "step": 5537 }, { "epoch": 1.4736562001064395, "grad_norm": 0.32405030727386475, "learning_rate": 1.6367559517537616e-07, "loss": 0.2018, "step": 5538 }, { "epoch": 1.4739222990952634, "grad_norm": 0.3071470260620117, "learning_rate": 1.6366243918513152e-07, "loss": 0.2141, "step": 5539 }, { "epoch": 1.4741883980840873, "grad_norm": 0.2804633378982544, "learning_rate": 1.6364928134183854e-07, "loss": 0.2067, "step": 5540 }, { "epoch": 1.474454497072911, "grad_norm": 0.2923538386821747, "learning_rate": 1.6363612164588026e-07, "loss": 0.2037, "step": 5541 }, { "epoch": 1.474720596061735, "grad_norm": 0.285281777381897, "learning_rate": 1.636229600976397e-07, "loss": 0.2177, "step": 5542 }, { "epoch": 1.4749866950505588, "grad_norm": 0.31184571981430054, "learning_rate": 1.6360979669749997e-07, "loss": 0.2326, "step": 5543 }, { "epoch": 1.4752527940393827, "grad_norm": 0.28386157751083374, "learning_rate": 1.6359663144584423e-07, "loss": 0.2053, "step": 5544 }, { "epoch": 1.4755188930282066, "grad_norm": 0.27244269847869873, "learning_rate": 1.6358346434305565e-07, "loss": 0.1919, "step": 5545 }, { "epoch": 1.4757849920170303, "grad_norm": 0.30131909251213074, "learning_rate": 1.6357029538951754e-07, "loss": 0.197, "step": 5546 }, { "epoch": 1.4760510910058542, "grad_norm": 0.32559967041015625, "learning_rate": 1.6355712458561318e-07, "loss": 0.2359, "step": 5547 }, { "epoch": 1.476317189994678, "grad_norm": 0.26945021748542786, "learning_rate": 1.6354395193172598e-07, "loss": 0.1888, "step": 5548 }, { "epoch": 1.4765832889835018, "grad_norm": 0.27741843461990356, "learning_rate": 1.6353077742823926e-07, "loss": 0.1962, "step": 5549 }, { "epoch": 1.4768493879723257, "grad_norm": 0.31298768520355225, "learning_rate": 1.6351760107553658e-07, "loss": 0.1868, "step": 5550 }, { "epoch": 1.4771154869611496, "grad_norm": 0.2885637879371643, "learning_rate": 1.635044228740015e-07, "loss": 0.2066, "step": 5551 }, { "epoch": 1.4773815859499733, "grad_norm": 0.2561722695827484, "learning_rate": 1.6349124282401755e-07, "loss": 0.1979, "step": 5552 }, { "epoch": 1.4776476849387972, "grad_norm": 0.2677553594112396, "learning_rate": 1.6347806092596835e-07, "loss": 0.1973, "step": 5553 }, { "epoch": 1.477913783927621, "grad_norm": 0.30284181237220764, "learning_rate": 1.634648771802376e-07, "loss": 0.2012, "step": 5554 }, { "epoch": 1.478179882916445, "grad_norm": 0.2822309732437134, "learning_rate": 1.634516915872091e-07, "loss": 0.1928, "step": 5555 }, { "epoch": 1.4784459819052689, "grad_norm": 0.30250293016433716, "learning_rate": 1.634385041472666e-07, "loss": 0.1965, "step": 5556 }, { "epoch": 1.4787120808940926, "grad_norm": 0.2860839366912842, "learning_rate": 1.6342531486079395e-07, "loss": 0.2116, "step": 5557 }, { "epoch": 1.4789781798829165, "grad_norm": 0.3195882737636566, "learning_rate": 1.6341212372817506e-07, "loss": 0.2134, "step": 5558 }, { "epoch": 1.4792442788717404, "grad_norm": 0.2961353063583374, "learning_rate": 1.633989307497939e-07, "loss": 0.1981, "step": 5559 }, { "epoch": 1.479510377860564, "grad_norm": 0.2837188243865967, "learning_rate": 1.633857359260345e-07, "loss": 0.224, "step": 5560 }, { "epoch": 1.479776476849388, "grad_norm": 0.26180320978164673, "learning_rate": 1.633725392572809e-07, "loss": 0.1957, "step": 5561 }, { "epoch": 1.4800425758382119, "grad_norm": 0.2693772614002228, "learning_rate": 1.633593407439172e-07, "loss": 0.1931, "step": 5562 }, { "epoch": 1.4803086748270355, "grad_norm": 0.29176971316337585, "learning_rate": 1.6334614038632764e-07, "loss": 0.223, "step": 5563 }, { "epoch": 1.4805747738158594, "grad_norm": 0.27769017219543457, "learning_rate": 1.6333293818489637e-07, "loss": 0.1957, "step": 5564 }, { "epoch": 1.4808408728046834, "grad_norm": 0.2883263826370239, "learning_rate": 1.6331973414000774e-07, "loss": 0.1896, "step": 5565 }, { "epoch": 1.4811069717935073, "grad_norm": 0.29579395055770874, "learning_rate": 1.6330652825204606e-07, "loss": 0.1998, "step": 5566 }, { "epoch": 1.4813730707823312, "grad_norm": 0.30603262782096863, "learning_rate": 1.632933205213957e-07, "loss": 0.206, "step": 5567 }, { "epoch": 1.4816391697711548, "grad_norm": 0.289427250623703, "learning_rate": 1.6328011094844115e-07, "loss": 0.2185, "step": 5568 }, { "epoch": 1.4819052687599787, "grad_norm": 0.2988455891609192, "learning_rate": 1.6326689953356684e-07, "loss": 0.2047, "step": 5569 }, { "epoch": 1.4821713677488026, "grad_norm": 0.28020721673965454, "learning_rate": 1.6325368627715737e-07, "loss": 0.2144, "step": 5570 }, { "epoch": 1.4824374667376263, "grad_norm": 0.30502834916114807, "learning_rate": 1.6324047117959733e-07, "loss": 0.2169, "step": 5571 }, { "epoch": 1.4827035657264502, "grad_norm": 0.2843972444534302, "learning_rate": 1.6322725424127138e-07, "loss": 0.1917, "step": 5572 }, { "epoch": 1.4829696647152741, "grad_norm": 0.2936203181743622, "learning_rate": 1.632140354625642e-07, "loss": 0.1998, "step": 5573 }, { "epoch": 1.4832357637040978, "grad_norm": 0.2651664912700653, "learning_rate": 1.6320081484386062e-07, "loss": 0.2057, "step": 5574 }, { "epoch": 1.4835018626929217, "grad_norm": 0.28415000438690186, "learning_rate": 1.6318759238554538e-07, "loss": 0.203, "step": 5575 }, { "epoch": 1.4837679616817456, "grad_norm": 0.29172804951667786, "learning_rate": 1.6317436808800343e-07, "loss": 0.2258, "step": 5576 }, { "epoch": 1.4840340606705695, "grad_norm": 0.29697155952453613, "learning_rate": 1.6316114195161962e-07, "loss": 0.2118, "step": 5577 }, { "epoch": 1.4843001596593932, "grad_norm": 0.31589823961257935, "learning_rate": 1.63147913976779e-07, "loss": 0.2203, "step": 5578 }, { "epoch": 1.4845662586482171, "grad_norm": 0.29003459215164185, "learning_rate": 1.6313468416386652e-07, "loss": 0.2144, "step": 5579 }, { "epoch": 1.484832357637041, "grad_norm": 0.2781437039375305, "learning_rate": 1.6312145251326736e-07, "loss": 0.2152, "step": 5580 }, { "epoch": 1.485098456625865, "grad_norm": 0.2887597978115082, "learning_rate": 1.6310821902536658e-07, "loss": 0.2123, "step": 5581 }, { "epoch": 1.4853645556146886, "grad_norm": 0.27906519174575806, "learning_rate": 1.6309498370054938e-07, "loss": 0.1966, "step": 5582 }, { "epoch": 1.4856306546035125, "grad_norm": 0.30342862010002136, "learning_rate": 1.630817465392011e-07, "loss": 0.2248, "step": 5583 }, { "epoch": 1.4858967535923364, "grad_norm": 0.33088186383247375, "learning_rate": 1.630685075417069e-07, "loss": 0.2215, "step": 5584 }, { "epoch": 1.48616285258116, "grad_norm": 0.3116476535797119, "learning_rate": 1.6305526670845225e-07, "loss": 0.218, "step": 5585 }, { "epoch": 1.486428951569984, "grad_norm": 0.29354438185691833, "learning_rate": 1.6304202403982248e-07, "loss": 0.2155, "step": 5586 }, { "epoch": 1.486695050558808, "grad_norm": 0.2800597846508026, "learning_rate": 1.6302877953620308e-07, "loss": 0.2062, "step": 5587 }, { "epoch": 1.4869611495476316, "grad_norm": 0.31270793080329895, "learning_rate": 1.630155331979796e-07, "loss": 0.2081, "step": 5588 }, { "epoch": 1.4872272485364555, "grad_norm": 0.28152573108673096, "learning_rate": 1.6300228502553753e-07, "loss": 0.1916, "step": 5589 }, { "epoch": 1.4874933475252794, "grad_norm": 0.2841043174266815, "learning_rate": 1.629890350192625e-07, "loss": 0.2031, "step": 5590 }, { "epoch": 1.4877594465141033, "grad_norm": 0.28056514263153076, "learning_rate": 1.6297578317954025e-07, "loss": 0.1963, "step": 5591 }, { "epoch": 1.4880255455029272, "grad_norm": 0.27655863761901855, "learning_rate": 1.6296252950675645e-07, "loss": 0.2275, "step": 5592 }, { "epoch": 1.4882916444917509, "grad_norm": 0.27266767621040344, "learning_rate": 1.629492740012969e-07, "loss": 0.196, "step": 5593 }, { "epoch": 1.4885577434805748, "grad_norm": 0.2936006486415863, "learning_rate": 1.6293601666354743e-07, "loss": 0.1837, "step": 5594 }, { "epoch": 1.4888238424693987, "grad_norm": 0.275034099817276, "learning_rate": 1.6292275749389393e-07, "loss": 0.2027, "step": 5595 }, { "epoch": 1.4890899414582224, "grad_norm": 0.2806098759174347, "learning_rate": 1.6290949649272231e-07, "loss": 0.2175, "step": 5596 }, { "epoch": 1.4893560404470463, "grad_norm": 0.2592627704143524, "learning_rate": 1.6289623366041862e-07, "loss": 0.198, "step": 5597 }, { "epoch": 1.4896221394358702, "grad_norm": 0.37678173184394836, "learning_rate": 1.628829689973689e-07, "loss": 0.2, "step": 5598 }, { "epoch": 1.4898882384246939, "grad_norm": 0.29119905829429626, "learning_rate": 1.628697025039592e-07, "loss": 0.2102, "step": 5599 }, { "epoch": 1.4901543374135178, "grad_norm": 0.3086448609828949, "learning_rate": 1.628564341805757e-07, "loss": 0.2081, "step": 5600 }, { "epoch": 1.4904204364023417, "grad_norm": 0.26465845108032227, "learning_rate": 1.6284316402760458e-07, "loss": 0.1812, "step": 5601 }, { "epoch": 1.4906865353911656, "grad_norm": 0.3030782639980316, "learning_rate": 1.6282989204543216e-07, "loss": 0.2112, "step": 5602 }, { "epoch": 1.4909526343799895, "grad_norm": 0.27893784642219543, "learning_rate": 1.628166182344447e-07, "loss": 0.203, "step": 5603 }, { "epoch": 1.4912187333688132, "grad_norm": 0.2781222462654114, "learning_rate": 1.628033425950286e-07, "loss": 0.2138, "step": 5604 }, { "epoch": 1.491484832357637, "grad_norm": 0.3112451732158661, "learning_rate": 1.6279006512757024e-07, "loss": 0.2163, "step": 5605 }, { "epoch": 1.491750931346461, "grad_norm": 0.2825128734111786, "learning_rate": 1.6277678583245614e-07, "loss": 0.2008, "step": 5606 }, { "epoch": 1.4920170303352847, "grad_norm": 0.30972421169281006, "learning_rate": 1.6276350471007282e-07, "loss": 0.2167, "step": 5607 }, { "epoch": 1.4922831293241086, "grad_norm": 0.3065846264362335, "learning_rate": 1.627502217608068e-07, "loss": 0.2244, "step": 5608 }, { "epoch": 1.4925492283129325, "grad_norm": 0.26907864212989807, "learning_rate": 1.6273693698504478e-07, "loss": 0.1857, "step": 5609 }, { "epoch": 1.4928153273017561, "grad_norm": 0.3196026086807251, "learning_rate": 1.627236503831734e-07, "loss": 0.2148, "step": 5610 }, { "epoch": 1.49308142629058, "grad_norm": 0.27645251154899597, "learning_rate": 1.6271036195557944e-07, "loss": 0.2093, "step": 5611 }, { "epoch": 1.493347525279404, "grad_norm": 0.2745801508426666, "learning_rate": 1.6269707170264965e-07, "loss": 0.2071, "step": 5612 }, { "epoch": 1.4936136242682279, "grad_norm": 0.27137047052383423, "learning_rate": 1.6268377962477091e-07, "loss": 0.2007, "step": 5613 }, { "epoch": 1.4938797232570518, "grad_norm": 0.28684020042419434, "learning_rate": 1.6267048572233008e-07, "loss": 0.2048, "step": 5614 }, { "epoch": 1.4941458222458754, "grad_norm": 0.26741424202919006, "learning_rate": 1.6265718999571415e-07, "loss": 0.1953, "step": 5615 }, { "epoch": 1.4944119212346993, "grad_norm": 0.2760040760040283, "learning_rate": 1.6264389244531012e-07, "loss": 0.1997, "step": 5616 }, { "epoch": 1.4946780202235233, "grad_norm": 0.2739270329475403, "learning_rate": 1.6263059307150501e-07, "loss": 0.1979, "step": 5617 }, { "epoch": 1.494944119212347, "grad_norm": 0.2852710783481598, "learning_rate": 1.6261729187468598e-07, "loss": 0.216, "step": 5618 }, { "epoch": 1.4952102182011708, "grad_norm": 0.29475975036621094, "learning_rate": 1.6260398885524014e-07, "loss": 0.2004, "step": 5619 }, { "epoch": 1.4954763171899947, "grad_norm": 0.2826225459575653, "learning_rate": 1.6259068401355475e-07, "loss": 0.199, "step": 5620 }, { "epoch": 1.4957424161788184, "grad_norm": 0.3090161979198456, "learning_rate": 1.6257737735001707e-07, "loss": 0.2081, "step": 5621 }, { "epoch": 1.4960085151676423, "grad_norm": 0.3071516454219818, "learning_rate": 1.6256406886501438e-07, "loss": 0.2351, "step": 5622 }, { "epoch": 1.4962746141564662, "grad_norm": 0.32326585054397583, "learning_rate": 1.6255075855893413e-07, "loss": 0.2224, "step": 5623 }, { "epoch": 1.49654071314529, "grad_norm": 0.2670265734195709, "learning_rate": 1.6253744643216367e-07, "loss": 0.1952, "step": 5624 }, { "epoch": 1.4968068121341138, "grad_norm": 0.28395259380340576, "learning_rate": 1.6252413248509056e-07, "loss": 0.2139, "step": 5625 }, { "epoch": 1.4970729111229377, "grad_norm": 0.29124686121940613, "learning_rate": 1.6251081671810226e-07, "loss": 0.2075, "step": 5626 }, { "epoch": 1.4973390101117616, "grad_norm": 0.2864789068698883, "learning_rate": 1.6249749913158642e-07, "loss": 0.2192, "step": 5627 }, { "epoch": 1.4976051091005855, "grad_norm": 1.1173052787780762, "learning_rate": 1.6248417972593064e-07, "loss": 0.2068, "step": 5628 }, { "epoch": 1.4978712080894092, "grad_norm": 0.3333924114704132, "learning_rate": 1.6247085850152264e-07, "loss": 0.2193, "step": 5629 }, { "epoch": 1.4981373070782331, "grad_norm": 0.28914904594421387, "learning_rate": 1.6245753545875015e-07, "loss": 0.2215, "step": 5630 }, { "epoch": 1.498403406067057, "grad_norm": 0.30929088592529297, "learning_rate": 1.6244421059800096e-07, "loss": 0.2202, "step": 5631 }, { "epoch": 1.4986695050558807, "grad_norm": 0.27048125863075256, "learning_rate": 1.6243088391966296e-07, "loss": 0.198, "step": 5632 }, { "epoch": 1.4989356040447046, "grad_norm": 0.2715226709842682, "learning_rate": 1.62417555424124e-07, "loss": 0.1858, "step": 5633 }, { "epoch": 1.4992017030335285, "grad_norm": 0.2725767493247986, "learning_rate": 1.624042251117721e-07, "loss": 0.2049, "step": 5634 }, { "epoch": 1.4994678020223522, "grad_norm": 0.26206305623054504, "learning_rate": 1.6239089298299523e-07, "loss": 0.1971, "step": 5635 }, { "epoch": 1.499733901011176, "grad_norm": 0.2690902054309845, "learning_rate": 1.6237755903818145e-07, "loss": 0.2175, "step": 5636 }, { "epoch": 1.5, "grad_norm": 0.2827305197715759, "learning_rate": 1.623642232777189e-07, "loss": 0.2129, "step": 5637 }, { "epoch": 1.5002660989888237, "grad_norm": 0.28154540061950684, "learning_rate": 1.623508857019957e-07, "loss": 0.1982, "step": 5638 }, { "epoch": 1.5005321979776478, "grad_norm": 0.36740002036094666, "learning_rate": 1.6233754631140015e-07, "loss": 0.1861, "step": 5639 }, { "epoch": 1.5007982969664715, "grad_norm": 0.35334548354148865, "learning_rate": 1.6232420510632044e-07, "loss": 0.214, "step": 5640 }, { "epoch": 1.5010643959552954, "grad_norm": 0.2923491597175598, "learning_rate": 1.62310862087145e-07, "loss": 0.2128, "step": 5641 }, { "epoch": 1.5013304949441193, "grad_norm": 0.2775861620903015, "learning_rate": 1.622975172542621e-07, "loss": 0.2013, "step": 5642 }, { "epoch": 1.501596593932943, "grad_norm": 0.2628116309642792, "learning_rate": 1.622841706080603e-07, "loss": 0.1897, "step": 5643 }, { "epoch": 1.5018626929217669, "grad_norm": 0.2762923240661621, "learning_rate": 1.6227082214892795e-07, "loss": 0.1991, "step": 5644 }, { "epoch": 1.5021287919105908, "grad_norm": 0.2703504264354706, "learning_rate": 1.6225747187725366e-07, "loss": 0.1999, "step": 5645 }, { "epoch": 1.5023948908994145, "grad_norm": 0.27611804008483887, "learning_rate": 1.6224411979342602e-07, "loss": 0.2026, "step": 5646 }, { "epoch": 1.5026609898882384, "grad_norm": 0.3198128044605255, "learning_rate": 1.6223076589783366e-07, "loss": 0.238, "step": 5647 }, { "epoch": 1.5029270888770623, "grad_norm": 0.31407907605171204, "learning_rate": 1.622174101908653e-07, "loss": 0.2227, "step": 5648 }, { "epoch": 1.503193187865886, "grad_norm": 0.3019453287124634, "learning_rate": 1.622040526729097e-07, "loss": 0.22, "step": 5649 }, { "epoch": 1.50345928685471, "grad_norm": 0.26903215050697327, "learning_rate": 1.6219069334435558e-07, "loss": 0.189, "step": 5650 }, { "epoch": 1.5037253858435338, "grad_norm": 0.3063419759273529, "learning_rate": 1.6217733220559186e-07, "loss": 0.2363, "step": 5651 }, { "epoch": 1.5039914848323577, "grad_norm": 0.28857606649398804, "learning_rate": 1.621639692570075e-07, "loss": 0.1972, "step": 5652 }, { "epoch": 1.5042575838211816, "grad_norm": 0.28541943430900574, "learning_rate": 1.6215060449899134e-07, "loss": 0.2113, "step": 5653 }, { "epoch": 1.5045236828100053, "grad_norm": 0.27309203147888184, "learning_rate": 1.621372379319325e-07, "loss": 0.2086, "step": 5654 }, { "epoch": 1.5047897817988292, "grad_norm": 0.3131561279296875, "learning_rate": 1.6212386955621997e-07, "loss": 0.2164, "step": 5655 }, { "epoch": 1.505055880787653, "grad_norm": 0.31387171149253845, "learning_rate": 1.621104993722429e-07, "loss": 0.2235, "step": 5656 }, { "epoch": 1.5053219797764767, "grad_norm": 0.2985694706439972, "learning_rate": 1.6209712738039046e-07, "loss": 0.207, "step": 5657 }, { "epoch": 1.5055880787653007, "grad_norm": 0.2749446630477905, "learning_rate": 1.6208375358105188e-07, "loss": 0.2072, "step": 5658 }, { "epoch": 1.5058541777541246, "grad_norm": 0.25393596291542053, "learning_rate": 1.6207037797461648e-07, "loss": 0.209, "step": 5659 }, { "epoch": 1.5061202767429482, "grad_norm": 0.31419748067855835, "learning_rate": 1.6205700056147348e-07, "loss": 0.2088, "step": 5660 }, { "epoch": 1.5063863757317724, "grad_norm": 0.2797854542732239, "learning_rate": 1.6204362134201234e-07, "loss": 0.1823, "step": 5661 }, { "epoch": 1.506652474720596, "grad_norm": 0.48659148812294006, "learning_rate": 1.620302403166225e-07, "loss": 0.2381, "step": 5662 }, { "epoch": 1.50691857370942, "grad_norm": 0.27851197123527527, "learning_rate": 1.620168574856934e-07, "loss": 0.2019, "step": 5663 }, { "epoch": 1.5071846726982439, "grad_norm": 0.27706289291381836, "learning_rate": 1.620034728496146e-07, "loss": 0.215, "step": 5664 }, { "epoch": 1.5074507716870675, "grad_norm": 0.2970532774925232, "learning_rate": 1.6199008640877572e-07, "loss": 0.2166, "step": 5665 }, { "epoch": 1.5077168706758914, "grad_norm": 0.32069292664527893, "learning_rate": 1.6197669816356636e-07, "loss": 0.2299, "step": 5666 }, { "epoch": 1.5079829696647153, "grad_norm": 0.281236857175827, "learning_rate": 1.6196330811437625e-07, "loss": 0.2199, "step": 5667 }, { "epoch": 1.508249068653539, "grad_norm": 0.30379703640937805, "learning_rate": 1.6194991626159512e-07, "loss": 0.2148, "step": 5668 }, { "epoch": 1.508515167642363, "grad_norm": 0.285854309797287, "learning_rate": 1.6193652260561277e-07, "loss": 0.2175, "step": 5669 }, { "epoch": 1.5087812666311868, "grad_norm": 0.2746090888977051, "learning_rate": 1.6192312714681908e-07, "loss": 0.2129, "step": 5670 }, { "epoch": 1.5090473656200105, "grad_norm": 0.2944503128528595, "learning_rate": 1.6190972988560394e-07, "loss": 0.2153, "step": 5671 }, { "epoch": 1.5093134646088346, "grad_norm": 0.2963242530822754, "learning_rate": 1.6189633082235728e-07, "loss": 0.2074, "step": 5672 }, { "epoch": 1.5095795635976583, "grad_norm": 0.2697899341583252, "learning_rate": 1.6188292995746916e-07, "loss": 0.1838, "step": 5673 }, { "epoch": 1.509845662586482, "grad_norm": 0.29727083444595337, "learning_rate": 1.6186952729132966e-07, "loss": 0.2104, "step": 5674 }, { "epoch": 1.5101117615753061, "grad_norm": 0.29003724455833435, "learning_rate": 1.6185612282432882e-07, "loss": 0.1905, "step": 5675 }, { "epoch": 1.5103778605641298, "grad_norm": 0.2567099630832672, "learning_rate": 1.6184271655685685e-07, "loss": 0.1881, "step": 5676 }, { "epoch": 1.5106439595529537, "grad_norm": 0.29522889852523804, "learning_rate": 1.61829308489304e-07, "loss": 0.213, "step": 5677 }, { "epoch": 1.5109100585417776, "grad_norm": 0.27239447832107544, "learning_rate": 1.6181589862206052e-07, "loss": 0.1889, "step": 5678 }, { "epoch": 1.5111761575306013, "grad_norm": 0.2974476218223572, "learning_rate": 1.6180248695551672e-07, "loss": 0.2156, "step": 5679 }, { "epoch": 1.5114422565194252, "grad_norm": 0.2785502076148987, "learning_rate": 1.61789073490063e-07, "loss": 0.206, "step": 5680 }, { "epoch": 1.5117083555082491, "grad_norm": 0.2634420394897461, "learning_rate": 1.6177565822608978e-07, "loss": 0.1874, "step": 5681 }, { "epoch": 1.5119744544970728, "grad_norm": 0.27420613169670105, "learning_rate": 1.6176224116398753e-07, "loss": 0.2029, "step": 5682 }, { "epoch": 1.5122405534858967, "grad_norm": 0.30828744173049927, "learning_rate": 1.6174882230414686e-07, "loss": 0.2449, "step": 5683 }, { "epoch": 1.5125066524747206, "grad_norm": 0.263536661863327, "learning_rate": 1.6173540164695824e-07, "loss": 0.194, "step": 5684 }, { "epoch": 1.5127727514635443, "grad_norm": 0.25497785210609436, "learning_rate": 1.617219791928124e-07, "loss": 0.1772, "step": 5685 }, { "epoch": 1.5130388504523684, "grad_norm": 0.29975366592407227, "learning_rate": 1.6170855494210001e-07, "loss": 0.2109, "step": 5686 }, { "epoch": 1.513304949441192, "grad_norm": 0.3127073049545288, "learning_rate": 1.616951288952118e-07, "loss": 0.2274, "step": 5687 }, { "epoch": 1.513571048430016, "grad_norm": 0.2821493148803711, "learning_rate": 1.616817010525386e-07, "loss": 0.2114, "step": 5688 }, { "epoch": 1.51383714741884, "grad_norm": 0.26849624514579773, "learning_rate": 1.6166827141447119e-07, "loss": 0.1939, "step": 5689 }, { "epoch": 1.5141032464076636, "grad_norm": 0.2745075225830078, "learning_rate": 1.6165483998140056e-07, "loss": 0.2114, "step": 5690 }, { "epoch": 1.5143693453964875, "grad_norm": 0.3012760877609253, "learning_rate": 1.616414067537176e-07, "loss": 0.2045, "step": 5691 }, { "epoch": 1.5146354443853114, "grad_norm": 0.32255569100379944, "learning_rate": 1.6162797173181333e-07, "loss": 0.2261, "step": 5692 }, { "epoch": 1.514901543374135, "grad_norm": 0.28853699564933777, "learning_rate": 1.6161453491607885e-07, "loss": 0.2047, "step": 5693 }, { "epoch": 1.515167642362959, "grad_norm": 0.3017551600933075, "learning_rate": 1.6160109630690523e-07, "loss": 0.224, "step": 5694 }, { "epoch": 1.5154337413517829, "grad_norm": 0.265951544046402, "learning_rate": 1.6158765590468365e-07, "loss": 0.1896, "step": 5695 }, { "epoch": 1.5156998403406066, "grad_norm": 0.2568245828151703, "learning_rate": 1.615742137098053e-07, "loss": 0.188, "step": 5696 }, { "epoch": 1.5159659393294307, "grad_norm": 0.30466338992118835, "learning_rate": 1.615607697226615e-07, "loss": 0.2169, "step": 5697 }, { "epoch": 1.5162320383182544, "grad_norm": 0.3036189675331116, "learning_rate": 1.6154732394364348e-07, "loss": 0.2138, "step": 5698 }, { "epoch": 1.5164981373070783, "grad_norm": 0.30177041888237, "learning_rate": 1.6153387637314268e-07, "loss": 0.2074, "step": 5699 }, { "epoch": 1.5167642362959022, "grad_norm": 0.284921258687973, "learning_rate": 1.6152042701155053e-07, "loss": 0.2197, "step": 5700 }, { "epoch": 1.5170303352847259, "grad_norm": 0.3097892105579376, "learning_rate": 1.6150697585925847e-07, "loss": 0.2198, "step": 5701 }, { "epoch": 1.5172964342735498, "grad_norm": 0.2692396640777588, "learning_rate": 1.6149352291665807e-07, "loss": 0.2192, "step": 5702 }, { "epoch": 1.5175625332623737, "grad_norm": 0.3057979643344879, "learning_rate": 1.6148006818414086e-07, "loss": 0.2034, "step": 5703 }, { "epoch": 1.5178286322511974, "grad_norm": 0.3123381733894348, "learning_rate": 1.614666116620985e-07, "loss": 0.2176, "step": 5704 }, { "epoch": 1.5180947312400213, "grad_norm": 0.28651750087738037, "learning_rate": 1.614531533509227e-07, "loss": 0.2185, "step": 5705 }, { "epoch": 1.5183608302288452, "grad_norm": 0.28859931230545044, "learning_rate": 1.6143969325100513e-07, "loss": 0.2098, "step": 5706 }, { "epoch": 1.5186269292176688, "grad_norm": 0.2685161828994751, "learning_rate": 1.6142623136273762e-07, "loss": 0.2047, "step": 5707 }, { "epoch": 1.518893028206493, "grad_norm": 0.2832171320915222, "learning_rate": 1.6141276768651202e-07, "loss": 0.1942, "step": 5708 }, { "epoch": 1.5191591271953166, "grad_norm": 0.31592217087745667, "learning_rate": 1.6139930222272018e-07, "loss": 0.235, "step": 5709 }, { "epoch": 1.5194252261841406, "grad_norm": 0.2636141777038574, "learning_rate": 1.6138583497175412e-07, "loss": 0.1915, "step": 5710 }, { "epoch": 1.5196913251729645, "grad_norm": 0.29279980063438416, "learning_rate": 1.6137236593400576e-07, "loss": 0.2127, "step": 5711 }, { "epoch": 1.5199574241617881, "grad_norm": 0.2983967363834381, "learning_rate": 1.6135889510986719e-07, "loss": 0.2141, "step": 5712 }, { "epoch": 1.520223523150612, "grad_norm": 0.3128395974636078, "learning_rate": 1.6134542249973048e-07, "loss": 0.2183, "step": 5713 }, { "epoch": 1.520489622139436, "grad_norm": 0.36253872513771057, "learning_rate": 1.6133194810398783e-07, "loss": 0.2092, "step": 5714 }, { "epoch": 1.5207557211282596, "grad_norm": 0.26072385907173157, "learning_rate": 1.6131847192303142e-07, "loss": 0.1946, "step": 5715 }, { "epoch": 1.5210218201170835, "grad_norm": 0.2978670597076416, "learning_rate": 1.613049939572535e-07, "loss": 0.2173, "step": 5716 }, { "epoch": 1.5212879191059074, "grad_norm": 0.2947273552417755, "learning_rate": 1.6129151420704635e-07, "loss": 0.2254, "step": 5717 }, { "epoch": 1.5215540180947311, "grad_norm": 0.25760623812675476, "learning_rate": 1.612780326728024e-07, "loss": 0.2079, "step": 5718 }, { "epoch": 1.5218201170835552, "grad_norm": 0.3303851783275604, "learning_rate": 1.61264549354914e-07, "loss": 0.2097, "step": 5719 }, { "epoch": 1.522086216072379, "grad_norm": 0.27143093943595886, "learning_rate": 1.6125106425377365e-07, "loss": 0.1863, "step": 5720 }, { "epoch": 1.5223523150612026, "grad_norm": 0.28342026472091675, "learning_rate": 1.6123757736977388e-07, "loss": 0.2052, "step": 5721 }, { "epoch": 1.5226184140500267, "grad_norm": 0.2754710614681244, "learning_rate": 1.6122408870330722e-07, "loss": 0.1841, "step": 5722 }, { "epoch": 1.5228845130388504, "grad_norm": 0.2907271385192871, "learning_rate": 1.6121059825476628e-07, "loss": 0.2067, "step": 5723 }, { "epoch": 1.5231506120276743, "grad_norm": 0.2806306481361389, "learning_rate": 1.6119710602454378e-07, "loss": 0.2207, "step": 5724 }, { "epoch": 1.5234167110164982, "grad_norm": 0.279471218585968, "learning_rate": 1.6118361201303241e-07, "loss": 0.2008, "step": 5725 }, { "epoch": 1.523682810005322, "grad_norm": 0.28929489850997925, "learning_rate": 1.6117011622062498e-07, "loss": 0.2018, "step": 5726 }, { "epoch": 1.5239489089941458, "grad_norm": 0.2910545766353607, "learning_rate": 1.6115661864771425e-07, "loss": 0.2184, "step": 5727 }, { "epoch": 1.5242150079829697, "grad_norm": 0.2676233947277069, "learning_rate": 1.6114311929469318e-07, "loss": 0.1998, "step": 5728 }, { "epoch": 1.5244811069717934, "grad_norm": 0.2848034203052521, "learning_rate": 1.6112961816195466e-07, "loss": 0.2074, "step": 5729 }, { "epoch": 1.5247472059606173, "grad_norm": 0.3007916808128357, "learning_rate": 1.6111611524989165e-07, "loss": 0.2189, "step": 5730 }, { "epoch": 1.5250133049494412, "grad_norm": 0.29477623105049133, "learning_rate": 1.6110261055889724e-07, "loss": 0.2106, "step": 5731 }, { "epoch": 1.5252794039382649, "grad_norm": 0.30435818433761597, "learning_rate": 1.6108910408936446e-07, "loss": 0.226, "step": 5732 }, { "epoch": 1.525545502927089, "grad_norm": 0.3086960017681122, "learning_rate": 1.610755958416865e-07, "loss": 0.2177, "step": 5733 }, { "epoch": 1.5258116019159127, "grad_norm": 0.29090243577957153, "learning_rate": 1.6106208581625654e-07, "loss": 0.2204, "step": 5734 }, { "epoch": 1.5260777009047366, "grad_norm": 0.5432490706443787, "learning_rate": 1.6104857401346776e-07, "loss": 0.2395, "step": 5735 }, { "epoch": 1.5263437998935605, "grad_norm": 0.2825673818588257, "learning_rate": 1.6103506043371352e-07, "loss": 0.1922, "step": 5736 }, { "epoch": 1.5266098988823842, "grad_norm": 0.2810426354408264, "learning_rate": 1.6102154507738717e-07, "loss": 0.2154, "step": 5737 }, { "epoch": 1.526875997871208, "grad_norm": 0.2816968262195587, "learning_rate": 1.610080279448821e-07, "loss": 0.2014, "step": 5738 }, { "epoch": 1.527142096860032, "grad_norm": 0.3595654368400574, "learning_rate": 1.609945090365917e-07, "loss": 0.2127, "step": 5739 }, { "epoch": 1.5274081958488557, "grad_norm": 0.25950005650520325, "learning_rate": 1.6098098835290951e-07, "loss": 0.2032, "step": 5740 }, { "epoch": 1.5276742948376796, "grad_norm": 0.2918642461299896, "learning_rate": 1.609674658942291e-07, "loss": 0.2161, "step": 5741 }, { "epoch": 1.5279403938265035, "grad_norm": 0.2521235942840576, "learning_rate": 1.6095394166094407e-07, "loss": 0.1829, "step": 5742 }, { "epoch": 1.5282064928153272, "grad_norm": 0.2969810664653778, "learning_rate": 1.6094041565344805e-07, "loss": 0.1999, "step": 5743 }, { "epoch": 1.5284725918041513, "grad_norm": 0.2861544191837311, "learning_rate": 1.6092688787213475e-07, "loss": 0.2147, "step": 5744 }, { "epoch": 1.528738690792975, "grad_norm": 0.2959790825843811, "learning_rate": 1.6091335831739797e-07, "loss": 0.2014, "step": 5745 }, { "epoch": 1.5290047897817989, "grad_norm": 0.2723054885864258, "learning_rate": 1.6089982698963147e-07, "loss": 0.1915, "step": 5746 }, { "epoch": 1.5292708887706228, "grad_norm": 0.31366339325904846, "learning_rate": 1.6088629388922912e-07, "loss": 0.2151, "step": 5747 }, { "epoch": 1.5295369877594465, "grad_norm": 0.30122068524360657, "learning_rate": 1.6087275901658487e-07, "loss": 0.2172, "step": 5748 }, { "epoch": 1.5298030867482704, "grad_norm": 0.30793818831443787, "learning_rate": 1.6085922237209264e-07, "loss": 0.2105, "step": 5749 }, { "epoch": 1.5300691857370943, "grad_norm": 0.2832941710948944, "learning_rate": 1.6084568395614646e-07, "loss": 0.2085, "step": 5750 }, { "epoch": 1.530335284725918, "grad_norm": 0.2600316107273102, "learning_rate": 1.6083214376914042e-07, "loss": 0.1928, "step": 5751 }, { "epoch": 1.5306013837147419, "grad_norm": 0.3247237503528595, "learning_rate": 1.6081860181146862e-07, "loss": 0.2138, "step": 5752 }, { "epoch": 1.5308674827035658, "grad_norm": 0.2822093665599823, "learning_rate": 1.6080505808352522e-07, "loss": 0.1949, "step": 5753 }, { "epoch": 1.5311335816923894, "grad_norm": 0.2913303077220917, "learning_rate": 1.607915125857045e-07, "loss": 0.2264, "step": 5754 }, { "epoch": 1.5313996806812136, "grad_norm": 0.27146750688552856, "learning_rate": 1.6077796531840066e-07, "loss": 0.1911, "step": 5755 }, { "epoch": 1.5316657796700373, "grad_norm": 0.27703657746315, "learning_rate": 1.6076441628200805e-07, "loss": 0.2075, "step": 5756 }, { "epoch": 1.531931878658861, "grad_norm": 0.336503803730011, "learning_rate": 1.607508654769211e-07, "loss": 0.199, "step": 5757 }, { "epoch": 1.532197977647685, "grad_norm": 0.2676593065261841, "learning_rate": 1.6073731290353413e-07, "loss": 0.2026, "step": 5758 }, { "epoch": 1.5324640766365087, "grad_norm": 0.2743823528289795, "learning_rate": 1.6072375856224173e-07, "loss": 0.2005, "step": 5759 }, { "epoch": 1.5327301756253326, "grad_norm": 0.29389438033103943, "learning_rate": 1.6071020245343837e-07, "loss": 0.218, "step": 5760 }, { "epoch": 1.5329962746141566, "grad_norm": 0.3402005136013031, "learning_rate": 1.6069664457751866e-07, "loss": 0.2022, "step": 5761 }, { "epoch": 1.5332623736029802, "grad_norm": 0.3623143434524536, "learning_rate": 1.6068308493487722e-07, "loss": 0.202, "step": 5762 }, { "epoch": 1.5335284725918041, "grad_norm": 0.2971780300140381, "learning_rate": 1.6066952352590871e-07, "loss": 0.2091, "step": 5763 }, { "epoch": 1.533794571580628, "grad_norm": 0.265972763299942, "learning_rate": 1.6065596035100793e-07, "loss": 0.1938, "step": 5764 }, { "epoch": 1.5340606705694517, "grad_norm": 0.3042720556259155, "learning_rate": 1.6064239541056962e-07, "loss": 0.2105, "step": 5765 }, { "epoch": 1.5343267695582756, "grad_norm": 0.3126526176929474, "learning_rate": 1.6062882870498866e-07, "loss": 0.1993, "step": 5766 }, { "epoch": 1.5345928685470995, "grad_norm": 0.2968270182609558, "learning_rate": 1.606152602346599e-07, "loss": 0.2194, "step": 5767 }, { "epoch": 1.5348589675359232, "grad_norm": 0.3152472972869873, "learning_rate": 1.606016899999783e-07, "loss": 0.2219, "step": 5768 }, { "epoch": 1.5351250665247473, "grad_norm": 0.31129691004753113, "learning_rate": 1.6058811800133888e-07, "loss": 0.2156, "step": 5769 }, { "epoch": 1.535391165513571, "grad_norm": 0.3124995231628418, "learning_rate": 1.605745442391366e-07, "loss": 0.2119, "step": 5770 }, { "epoch": 1.535657264502395, "grad_norm": 0.28611502051353455, "learning_rate": 1.6056096871376665e-07, "loss": 0.2129, "step": 5771 }, { "epoch": 1.5359233634912188, "grad_norm": 0.317190945148468, "learning_rate": 1.6054739142562416e-07, "loss": 0.2098, "step": 5772 }, { "epoch": 1.5361894624800425, "grad_norm": 0.32301798462867737, "learning_rate": 1.605338123751043e-07, "loss": 0.2134, "step": 5773 }, { "epoch": 1.5364555614688664, "grad_norm": 0.30252784490585327, "learning_rate": 1.6052023156260236e-07, "loss": 0.1901, "step": 5774 }, { "epoch": 1.5367216604576903, "grad_norm": 0.2837463915348053, "learning_rate": 1.605066489885136e-07, "loss": 0.2077, "step": 5775 }, { "epoch": 1.536987759446514, "grad_norm": 0.262798935174942, "learning_rate": 1.6049306465323337e-07, "loss": 0.1988, "step": 5776 }, { "epoch": 1.537253858435338, "grad_norm": 0.2974953353404999, "learning_rate": 1.6047947855715713e-07, "loss": 0.2083, "step": 5777 }, { "epoch": 1.5375199574241618, "grad_norm": 0.29452595114707947, "learning_rate": 1.6046589070068032e-07, "loss": 0.2171, "step": 5778 }, { "epoch": 1.5377860564129855, "grad_norm": 0.28333377838134766, "learning_rate": 1.6045230108419835e-07, "loss": 0.203, "step": 5779 }, { "epoch": 1.5380521554018096, "grad_norm": 0.31193578243255615, "learning_rate": 1.6043870970810695e-07, "loss": 0.1996, "step": 5780 }, { "epoch": 1.5383182543906333, "grad_norm": 0.28742465376853943, "learning_rate": 1.604251165728016e-07, "loss": 0.216, "step": 5781 }, { "epoch": 1.5385843533794572, "grad_norm": 0.265179306268692, "learning_rate": 1.6041152167867798e-07, "loss": 0.198, "step": 5782 }, { "epoch": 1.538850452368281, "grad_norm": 0.27968311309814453, "learning_rate": 1.6039792502613186e-07, "loss": 0.2157, "step": 5783 }, { "epoch": 1.5391165513571048, "grad_norm": 0.27353596687316895, "learning_rate": 1.6038432661555891e-07, "loss": 0.2143, "step": 5784 }, { "epoch": 1.5393826503459287, "grad_norm": 0.25195953249931335, "learning_rate": 1.6037072644735504e-07, "loss": 0.1773, "step": 5785 }, { "epoch": 1.5396487493347526, "grad_norm": 0.28658655285835266, "learning_rate": 1.6035712452191606e-07, "loss": 0.2098, "step": 5786 }, { "epoch": 1.5399148483235763, "grad_norm": 0.28172817826271057, "learning_rate": 1.603435208396379e-07, "loss": 0.2204, "step": 5787 }, { "epoch": 1.5401809473124002, "grad_norm": 0.2943706810474396, "learning_rate": 1.6032991540091654e-07, "loss": 0.2284, "step": 5788 }, { "epoch": 1.540447046301224, "grad_norm": 0.30365660786628723, "learning_rate": 1.6031630820614796e-07, "loss": 0.1909, "step": 5789 }, { "epoch": 1.5407131452900478, "grad_norm": 0.3121468126773834, "learning_rate": 1.6030269925572827e-07, "loss": 0.2367, "step": 5790 }, { "epoch": 1.540979244278872, "grad_norm": 0.2802201509475708, "learning_rate": 1.602890885500536e-07, "loss": 0.1861, "step": 5791 }, { "epoch": 1.5412453432676956, "grad_norm": 0.2603971064090729, "learning_rate": 1.602754760895201e-07, "loss": 0.1974, "step": 5792 }, { "epoch": 1.5415114422565193, "grad_norm": 0.28629475831985474, "learning_rate": 1.60261861874524e-07, "loss": 0.2028, "step": 5793 }, { "epoch": 1.5417775412453434, "grad_norm": 0.27360203862190247, "learning_rate": 1.6024824590546153e-07, "loss": 0.1924, "step": 5794 }, { "epoch": 1.542043640234167, "grad_norm": 0.30818453431129456, "learning_rate": 1.6023462818272905e-07, "loss": 0.2122, "step": 5795 }, { "epoch": 1.542309739222991, "grad_norm": 0.30571362376213074, "learning_rate": 1.6022100870672302e-07, "loss": 0.2087, "step": 5796 }, { "epoch": 1.5425758382118149, "grad_norm": 0.2989807724952698, "learning_rate": 1.6020738747783974e-07, "loss": 0.2178, "step": 5797 }, { "epoch": 1.5428419372006386, "grad_norm": 0.2891039550304413, "learning_rate": 1.601937644964757e-07, "loss": 0.1979, "step": 5798 }, { "epoch": 1.5431080361894625, "grad_norm": 0.26862433552742004, "learning_rate": 1.601801397630275e-07, "loss": 0.2137, "step": 5799 }, { "epoch": 1.5433741351782864, "grad_norm": 0.3012891411781311, "learning_rate": 1.601665132778917e-07, "loss": 0.2097, "step": 5800 }, { "epoch": 1.54364023416711, "grad_norm": 0.3015258014202118, "learning_rate": 1.601528850414649e-07, "loss": 0.2081, "step": 5801 }, { "epoch": 1.543906333155934, "grad_norm": 0.2745363712310791, "learning_rate": 1.6013925505414383e-07, "loss": 0.2005, "step": 5802 }, { "epoch": 1.5441724321447579, "grad_norm": 0.27352744340896606, "learning_rate": 1.6012562331632517e-07, "loss": 0.1903, "step": 5803 }, { "epoch": 1.5444385311335815, "grad_norm": 0.28279608488082886, "learning_rate": 1.6011198982840575e-07, "loss": 0.2149, "step": 5804 }, { "epoch": 1.5447046301224057, "grad_norm": 0.27953752875328064, "learning_rate": 1.6009835459078237e-07, "loss": 0.2029, "step": 5805 }, { "epoch": 1.5449707291112293, "grad_norm": 0.282254159450531, "learning_rate": 1.600847176038519e-07, "loss": 0.1863, "step": 5806 }, { "epoch": 1.5452368281000533, "grad_norm": 0.30111947655677795, "learning_rate": 1.6007107886801136e-07, "loss": 0.2225, "step": 5807 }, { "epoch": 1.5455029270888772, "grad_norm": 0.29456251859664917, "learning_rate": 1.6005743838365766e-07, "loss": 0.1943, "step": 5808 }, { "epoch": 1.5457690260777008, "grad_norm": 0.28129035234451294, "learning_rate": 1.600437961511879e-07, "loss": 0.1965, "step": 5809 }, { "epoch": 1.5460351250665247, "grad_norm": 0.2845054864883423, "learning_rate": 1.6003015217099912e-07, "loss": 0.1984, "step": 5810 }, { "epoch": 1.5463012240553486, "grad_norm": 0.255307674407959, "learning_rate": 1.6001650644348846e-07, "loss": 0.1813, "step": 5811 }, { "epoch": 1.5465673230441723, "grad_norm": 0.28857603669166565, "learning_rate": 1.6000285896905315e-07, "loss": 0.2133, "step": 5812 }, { "epoch": 1.5468334220329962, "grad_norm": 0.2799028158187866, "learning_rate": 1.5998920974809037e-07, "loss": 0.1983, "step": 5813 }, { "epoch": 1.5470995210218201, "grad_norm": 0.2606425881385803, "learning_rate": 1.5997555878099748e-07, "loss": 0.1941, "step": 5814 }, { "epoch": 1.5473656200106438, "grad_norm": 0.2912461459636688, "learning_rate": 1.5996190606817182e-07, "loss": 0.2239, "step": 5815 }, { "epoch": 1.547631718999468, "grad_norm": 0.30541232228279114, "learning_rate": 1.5994825161001074e-07, "loss": 0.222, "step": 5816 }, { "epoch": 1.5478978179882916, "grad_norm": 0.2902706563472748, "learning_rate": 1.599345954069117e-07, "loss": 0.1979, "step": 5817 }, { "epoch": 1.5481639169771155, "grad_norm": 0.29633527994155884, "learning_rate": 1.5992093745927223e-07, "loss": 0.1988, "step": 5818 }, { "epoch": 1.5484300159659394, "grad_norm": 0.2950620949268341, "learning_rate": 1.5990727776748986e-07, "loss": 0.1988, "step": 5819 }, { "epoch": 1.5486961149547631, "grad_norm": 0.27776971459388733, "learning_rate": 1.5989361633196216e-07, "loss": 0.2061, "step": 5820 }, { "epoch": 1.548962213943587, "grad_norm": 0.29949724674224854, "learning_rate": 1.5987995315308682e-07, "loss": 0.2168, "step": 5821 }, { "epoch": 1.549228312932411, "grad_norm": 0.31633731722831726, "learning_rate": 1.5986628823126148e-07, "loss": 0.2242, "step": 5822 }, { "epoch": 1.5494944119212346, "grad_norm": 0.290423721075058, "learning_rate": 1.5985262156688397e-07, "loss": 0.2136, "step": 5823 }, { "epoch": 1.5497605109100585, "grad_norm": 0.31776997447013855, "learning_rate": 1.5983895316035204e-07, "loss": 0.2261, "step": 5824 }, { "epoch": 1.5500266098988824, "grad_norm": 0.2757991850376129, "learning_rate": 1.5982528301206356e-07, "loss": 0.1839, "step": 5825 }, { "epoch": 1.550292708887706, "grad_norm": 0.3060232698917389, "learning_rate": 1.5981161112241642e-07, "loss": 0.2279, "step": 5826 }, { "epoch": 1.5505588078765302, "grad_norm": 0.28664320707321167, "learning_rate": 1.5979793749180858e-07, "loss": 0.2269, "step": 5827 }, { "epoch": 1.550824906865354, "grad_norm": 0.2941875457763672, "learning_rate": 1.5978426212063808e-07, "loss": 0.22, "step": 5828 }, { "epoch": 1.5510910058541778, "grad_norm": 0.27559810876846313, "learning_rate": 1.5977058500930287e-07, "loss": 0.1991, "step": 5829 }, { "epoch": 1.5513571048430017, "grad_norm": 0.26894158124923706, "learning_rate": 1.5975690615820116e-07, "loss": 0.2026, "step": 5830 }, { "epoch": 1.5516232038318254, "grad_norm": 0.29770711064338684, "learning_rate": 1.5974322556773108e-07, "loss": 0.2047, "step": 5831 }, { "epoch": 1.5518893028206493, "grad_norm": 0.2931695282459259, "learning_rate": 1.597295432382908e-07, "loss": 0.2191, "step": 5832 }, { "epoch": 1.5521554018094732, "grad_norm": 0.2898569405078888, "learning_rate": 1.597158591702786e-07, "loss": 0.2199, "step": 5833 }, { "epoch": 1.5524215007982969, "grad_norm": 0.3099619746208191, "learning_rate": 1.5970217336409283e-07, "loss": 0.2288, "step": 5834 }, { "epoch": 1.5526875997871208, "grad_norm": 0.2877415716648102, "learning_rate": 1.5968848582013177e-07, "loss": 0.1983, "step": 5835 }, { "epoch": 1.5529536987759447, "grad_norm": 0.3327454924583435, "learning_rate": 1.5967479653879388e-07, "loss": 0.213, "step": 5836 }, { "epoch": 1.5532197977647684, "grad_norm": 0.26968643069267273, "learning_rate": 1.596611055204776e-07, "loss": 0.1997, "step": 5837 }, { "epoch": 1.5534858967535925, "grad_norm": 0.27970361709594727, "learning_rate": 1.5964741276558145e-07, "loss": 0.2092, "step": 5838 }, { "epoch": 1.5537519957424162, "grad_norm": 0.27953875064849854, "learning_rate": 1.59633718274504e-07, "loss": 0.2225, "step": 5839 }, { "epoch": 1.5540180947312399, "grad_norm": 0.29763659834861755, "learning_rate": 1.5962002204764383e-07, "loss": 0.1775, "step": 5840 }, { "epoch": 1.554284193720064, "grad_norm": 0.3148444890975952, "learning_rate": 1.5960632408539963e-07, "loss": 0.2202, "step": 5841 }, { "epoch": 1.5545502927088877, "grad_norm": 0.28080710768699646, "learning_rate": 1.5959262438817008e-07, "loss": 0.1905, "step": 5842 }, { "epoch": 1.5548163916977116, "grad_norm": 0.3047644793987274, "learning_rate": 1.59578922956354e-07, "loss": 0.2204, "step": 5843 }, { "epoch": 1.5550824906865355, "grad_norm": 0.2641970217227936, "learning_rate": 1.5956521979035015e-07, "loss": 0.2055, "step": 5844 }, { "epoch": 1.5553485896753592, "grad_norm": 0.3024272620677948, "learning_rate": 1.5955151489055743e-07, "loss": 0.2133, "step": 5845 }, { "epoch": 1.555614688664183, "grad_norm": 0.28939151763916016, "learning_rate": 1.595378082573747e-07, "loss": 0.1964, "step": 5846 }, { "epoch": 1.555880787653007, "grad_norm": 0.32283154129981995, "learning_rate": 1.5952409989120099e-07, "loss": 0.1979, "step": 5847 }, { "epoch": 1.5561468866418307, "grad_norm": 0.2898785173892975, "learning_rate": 1.595103897924353e-07, "loss": 0.2265, "step": 5848 }, { "epoch": 1.5564129856306546, "grad_norm": 0.31817105412483215, "learning_rate": 1.5949667796147664e-07, "loss": 0.2111, "step": 5849 }, { "epoch": 1.5566790846194785, "grad_norm": 0.29676902294158936, "learning_rate": 1.5948296439872414e-07, "loss": 0.2107, "step": 5850 }, { "epoch": 1.5569451836083021, "grad_norm": 0.2795538902282715, "learning_rate": 1.5946924910457708e-07, "loss": 0.1906, "step": 5851 }, { "epoch": 1.5572112825971263, "grad_norm": 0.2758229970932007, "learning_rate": 1.5945553207943453e-07, "loss": 0.2036, "step": 5852 }, { "epoch": 1.55747738158595, "grad_norm": 0.287146657705307, "learning_rate": 1.5944181332369583e-07, "loss": 0.2116, "step": 5853 }, { "epoch": 1.5577434805747739, "grad_norm": 0.3089042007923126, "learning_rate": 1.5942809283776027e-07, "loss": 0.2138, "step": 5854 }, { "epoch": 1.5580095795635978, "grad_norm": 0.2845820486545563, "learning_rate": 1.5941437062202727e-07, "loss": 0.2086, "step": 5855 }, { "epoch": 1.5582756785524214, "grad_norm": 0.28977301716804504, "learning_rate": 1.5940064667689616e-07, "loss": 0.2151, "step": 5856 }, { "epoch": 1.5585417775412453, "grad_norm": 0.27606186270713806, "learning_rate": 1.593869210027665e-07, "loss": 0.1887, "step": 5857 }, { "epoch": 1.5588078765300692, "grad_norm": 0.2763429284095764, "learning_rate": 1.593731936000377e-07, "loss": 0.2104, "step": 5858 }, { "epoch": 1.559073975518893, "grad_norm": 0.3067164421081543, "learning_rate": 1.5935946446910946e-07, "loss": 0.2111, "step": 5859 }, { "epoch": 1.5593400745077168, "grad_norm": 0.30253809690475464, "learning_rate": 1.5934573361038133e-07, "loss": 0.207, "step": 5860 }, { "epoch": 1.5596061734965407, "grad_norm": 0.2898516356945038, "learning_rate": 1.5933200102425298e-07, "loss": 0.2034, "step": 5861 }, { "epoch": 1.5598722724853644, "grad_norm": 0.30260756611824036, "learning_rate": 1.593182667111241e-07, "loss": 0.2045, "step": 5862 }, { "epoch": 1.5601383714741885, "grad_norm": 0.28148725628852844, "learning_rate": 1.5930453067139452e-07, "loss": 0.1872, "step": 5863 }, { "epoch": 1.5604044704630122, "grad_norm": 0.3306540846824646, "learning_rate": 1.5929079290546407e-07, "loss": 0.251, "step": 5864 }, { "epoch": 1.5606705694518361, "grad_norm": 0.33643078804016113, "learning_rate": 1.592770534137325e-07, "loss": 0.21, "step": 5865 }, { "epoch": 1.56093666844066, "grad_norm": 0.29277682304382324, "learning_rate": 1.5926331219659988e-07, "loss": 0.1917, "step": 5866 }, { "epoch": 1.5612027674294837, "grad_norm": 0.2869696617126465, "learning_rate": 1.592495692544661e-07, "loss": 0.2181, "step": 5867 }, { "epoch": 1.5614688664183076, "grad_norm": 0.30198460817337036, "learning_rate": 1.592358245877312e-07, "loss": 0.2172, "step": 5868 }, { "epoch": 1.5617349654071315, "grad_norm": 0.2928541600704193, "learning_rate": 1.5922207819679528e-07, "loss": 0.2127, "step": 5869 }, { "epoch": 1.5620010643959552, "grad_norm": 0.27944833040237427, "learning_rate": 1.592083300820584e-07, "loss": 0.2023, "step": 5870 }, { "epoch": 1.5622671633847791, "grad_norm": 0.2955019176006317, "learning_rate": 1.5919458024392074e-07, "loss": 0.2171, "step": 5871 }, { "epoch": 1.562533262373603, "grad_norm": 0.2822404205799103, "learning_rate": 1.5918082868278258e-07, "loss": 0.2119, "step": 5872 }, { "epoch": 1.5627993613624267, "grad_norm": 0.31356537342071533, "learning_rate": 1.5916707539904418e-07, "loss": 0.2123, "step": 5873 }, { "epoch": 1.5630654603512508, "grad_norm": 0.27995631098747253, "learning_rate": 1.591533203931058e-07, "loss": 0.2048, "step": 5874 }, { "epoch": 1.5633315593400745, "grad_norm": 0.2786542475223541, "learning_rate": 1.5913956366536784e-07, "loss": 0.2079, "step": 5875 }, { "epoch": 1.5635976583288982, "grad_norm": 0.3012198209762573, "learning_rate": 1.5912580521623078e-07, "loss": 0.2115, "step": 5876 }, { "epoch": 1.5638637573177223, "grad_norm": 0.29431429505348206, "learning_rate": 1.5911204504609504e-07, "loss": 0.1942, "step": 5877 }, { "epoch": 1.564129856306546, "grad_norm": 0.30777066946029663, "learning_rate": 1.5909828315536115e-07, "loss": 0.2087, "step": 5878 }, { "epoch": 1.56439595529537, "grad_norm": 0.2636222541332245, "learning_rate": 1.5908451954442969e-07, "loss": 0.204, "step": 5879 }, { "epoch": 1.5646620542841938, "grad_norm": 0.2994536757469177, "learning_rate": 1.5907075421370124e-07, "loss": 0.2375, "step": 5880 }, { "epoch": 1.5649281532730175, "grad_norm": 0.30115970969200134, "learning_rate": 1.5905698716357654e-07, "loss": 0.2178, "step": 5881 }, { "epoch": 1.5651942522618414, "grad_norm": 0.30298909544944763, "learning_rate": 1.5904321839445628e-07, "loss": 0.2303, "step": 5882 }, { "epoch": 1.5654603512506653, "grad_norm": 0.2899323105812073, "learning_rate": 1.5902944790674125e-07, "loss": 0.2002, "step": 5883 }, { "epoch": 1.565726450239489, "grad_norm": 0.28132835030555725, "learning_rate": 1.5901567570083224e-07, "loss": 0.2144, "step": 5884 }, { "epoch": 1.5659925492283129, "grad_norm": 0.2661817967891693, "learning_rate": 1.5900190177713015e-07, "loss": 0.1909, "step": 5885 }, { "epoch": 1.5662586482171368, "grad_norm": 0.2703098952770233, "learning_rate": 1.5898812613603587e-07, "loss": 0.1952, "step": 5886 }, { "epoch": 1.5665247472059605, "grad_norm": 0.32418709993362427, "learning_rate": 1.5897434877795047e-07, "loss": 0.2151, "step": 5887 }, { "epoch": 1.5667908461947846, "grad_norm": 0.276741623878479, "learning_rate": 1.5896056970327485e-07, "loss": 0.2023, "step": 5888 }, { "epoch": 1.5670569451836083, "grad_norm": 0.3031286299228668, "learning_rate": 1.5894678891241013e-07, "loss": 0.2191, "step": 5889 }, { "epoch": 1.5673230441724322, "grad_norm": 0.2823694348335266, "learning_rate": 1.5893300640575743e-07, "loss": 0.1988, "step": 5890 }, { "epoch": 1.567589143161256, "grad_norm": 0.26683419942855835, "learning_rate": 1.5891922218371799e-07, "loss": 0.1887, "step": 5891 }, { "epoch": 1.5678552421500798, "grad_norm": 0.29369688034057617, "learning_rate": 1.5890543624669292e-07, "loss": 0.2006, "step": 5892 }, { "epoch": 1.5681213411389037, "grad_norm": 0.30467408895492554, "learning_rate": 1.5889164859508356e-07, "loss": 0.2137, "step": 5893 }, { "epoch": 1.5683874401277276, "grad_norm": 0.3028952479362488, "learning_rate": 1.5887785922929126e-07, "loss": 0.2096, "step": 5894 }, { "epoch": 1.5686535391165513, "grad_norm": 0.27936041355133057, "learning_rate": 1.5886406814971728e-07, "loss": 0.2232, "step": 5895 }, { "epoch": 1.5689196381053752, "grad_norm": 0.27608567476272583, "learning_rate": 1.5885027535676317e-07, "loss": 0.2026, "step": 5896 }, { "epoch": 1.569185737094199, "grad_norm": 0.2830093801021576, "learning_rate": 1.5883648085083033e-07, "loss": 0.2159, "step": 5897 }, { "epoch": 1.5694518360830227, "grad_norm": 0.2881518602371216, "learning_rate": 1.588226846323203e-07, "loss": 0.2109, "step": 5898 }, { "epoch": 1.5697179350718469, "grad_norm": 0.27336642146110535, "learning_rate": 1.5880888670163465e-07, "loss": 0.1987, "step": 5899 }, { "epoch": 1.5699840340606706, "grad_norm": 0.35424506664276123, "learning_rate": 1.58795087059175e-07, "loss": 0.2345, "step": 5900 }, { "epoch": 1.5702501330494945, "grad_norm": 0.28773361444473267, "learning_rate": 1.5878128570534305e-07, "loss": 0.2113, "step": 5901 }, { "epoch": 1.5705162320383184, "grad_norm": 0.2759089767932892, "learning_rate": 1.5876748264054046e-07, "loss": 0.1951, "step": 5902 }, { "epoch": 1.570782331027142, "grad_norm": 0.31220102310180664, "learning_rate": 1.5875367786516907e-07, "loss": 0.196, "step": 5903 }, { "epoch": 1.571048430015966, "grad_norm": 0.3038322627544403, "learning_rate": 1.5873987137963062e-07, "loss": 0.2012, "step": 5904 }, { "epoch": 1.5713145290047899, "grad_norm": 0.28754448890686035, "learning_rate": 1.587260631843271e-07, "loss": 0.2013, "step": 5905 }, { "epoch": 1.5715806279936135, "grad_norm": 0.2852868437767029, "learning_rate": 1.587122532796603e-07, "loss": 0.2094, "step": 5906 }, { "epoch": 1.5718467269824374, "grad_norm": 0.3046714961528778, "learning_rate": 1.586984416660323e-07, "loss": 0.218, "step": 5907 }, { "epoch": 1.5721128259712613, "grad_norm": 0.3126576244831085, "learning_rate": 1.5868462834384505e-07, "loss": 0.2289, "step": 5908 }, { "epoch": 1.572378924960085, "grad_norm": 0.29258373379707336, "learning_rate": 1.5867081331350067e-07, "loss": 0.1992, "step": 5909 }, { "epoch": 1.5726450239489091, "grad_norm": 0.2883410155773163, "learning_rate": 1.586569965754012e-07, "loss": 0.2241, "step": 5910 }, { "epoch": 1.5729111229377328, "grad_norm": 0.29596906900405884, "learning_rate": 1.586431781299489e-07, "loss": 0.2114, "step": 5911 }, { "epoch": 1.5731772219265565, "grad_norm": 0.2828449606895447, "learning_rate": 1.5862935797754594e-07, "loss": 0.2106, "step": 5912 }, { "epoch": 1.5734433209153806, "grad_norm": 0.27832505106925964, "learning_rate": 1.5861553611859457e-07, "loss": 0.21, "step": 5913 }, { "epoch": 1.5737094199042043, "grad_norm": 0.32646456360816956, "learning_rate": 1.586017125534972e-07, "loss": 0.2021, "step": 5914 }, { "epoch": 1.5739755188930282, "grad_norm": 0.284185528755188, "learning_rate": 1.585878872826561e-07, "loss": 0.2132, "step": 5915 }, { "epoch": 1.5742416178818521, "grad_norm": 0.2763395607471466, "learning_rate": 1.585740603064737e-07, "loss": 0.1949, "step": 5916 }, { "epoch": 1.5745077168706758, "grad_norm": 0.2922089695930481, "learning_rate": 1.5856023162535252e-07, "loss": 0.2069, "step": 5917 }, { "epoch": 1.5747738158594997, "grad_norm": 0.30332934856414795, "learning_rate": 1.5854640123969505e-07, "loss": 0.2336, "step": 5918 }, { "epoch": 1.5750399148483236, "grad_norm": 0.27113577723503113, "learning_rate": 1.5853256914990385e-07, "loss": 0.1868, "step": 5919 }, { "epoch": 1.5753060138371473, "grad_norm": 0.28291288018226624, "learning_rate": 1.5851873535638153e-07, "loss": 0.1998, "step": 5920 }, { "epoch": 1.5755721128259714, "grad_norm": 0.2814441919326782, "learning_rate": 1.5850489985953074e-07, "loss": 0.2118, "step": 5921 }, { "epoch": 1.575838211814795, "grad_norm": 0.28604456782341003, "learning_rate": 1.5849106265975425e-07, "loss": 0.2081, "step": 5922 }, { "epoch": 1.5761043108036188, "grad_norm": 0.28535908460617065, "learning_rate": 1.5847722375745483e-07, "loss": 0.2209, "step": 5923 }, { "epoch": 1.576370409792443, "grad_norm": 0.28050217032432556, "learning_rate": 1.5846338315303524e-07, "loss": 0.2015, "step": 5924 }, { "epoch": 1.5766365087812666, "grad_norm": 0.26572009921073914, "learning_rate": 1.5844954084689834e-07, "loss": 0.1881, "step": 5925 }, { "epoch": 1.5769026077700905, "grad_norm": 0.27981460094451904, "learning_rate": 1.584356968394471e-07, "loss": 0.2134, "step": 5926 }, { "epoch": 1.5771687067589144, "grad_norm": 0.26385217905044556, "learning_rate": 1.584218511310844e-07, "loss": 0.1927, "step": 5927 }, { "epoch": 1.577434805747738, "grad_norm": 0.2787485122680664, "learning_rate": 1.5840800372221333e-07, "loss": 0.1969, "step": 5928 }, { "epoch": 1.577700904736562, "grad_norm": 0.2949921488761902, "learning_rate": 1.5839415461323696e-07, "loss": 0.2139, "step": 5929 }, { "epoch": 1.577967003725386, "grad_norm": 0.2715529203414917, "learning_rate": 1.5838030380455835e-07, "loss": 0.2109, "step": 5930 }, { "epoch": 1.5782331027142096, "grad_norm": 0.27428364753723145, "learning_rate": 1.5836645129658067e-07, "loss": 0.1981, "step": 5931 }, { "epoch": 1.5784992017030335, "grad_norm": 0.27905333042144775, "learning_rate": 1.5835259708970715e-07, "loss": 0.2135, "step": 5932 }, { "epoch": 1.5787653006918574, "grad_norm": 0.2853282392024994, "learning_rate": 1.5833874118434104e-07, "loss": 0.2198, "step": 5933 }, { "epoch": 1.579031399680681, "grad_norm": 0.3080211281776428, "learning_rate": 1.5832488358088562e-07, "loss": 0.2113, "step": 5934 }, { "epoch": 1.5792974986695052, "grad_norm": 0.28645846247673035, "learning_rate": 1.583110242797443e-07, "loss": 0.1942, "step": 5935 }, { "epoch": 1.5795635976583289, "grad_norm": 0.2808751165866852, "learning_rate": 1.5829716328132051e-07, "loss": 0.2145, "step": 5936 }, { "epoch": 1.5798296966471528, "grad_norm": 0.30422741174697876, "learning_rate": 1.5828330058601757e-07, "loss": 0.2233, "step": 5937 }, { "epoch": 1.5800957956359767, "grad_norm": 0.2725033164024353, "learning_rate": 1.5826943619423915e-07, "loss": 0.1976, "step": 5938 }, { "epoch": 1.5803618946248004, "grad_norm": 0.3121735155582428, "learning_rate": 1.582555701063887e-07, "loss": 0.2069, "step": 5939 }, { "epoch": 1.5806279936136243, "grad_norm": 0.2934766411781311, "learning_rate": 1.5824170232286987e-07, "loss": 0.2238, "step": 5940 }, { "epoch": 1.5808940926024482, "grad_norm": 0.33847829699516296, "learning_rate": 1.5822783284408636e-07, "loss": 0.1785, "step": 5941 }, { "epoch": 1.5811601915912719, "grad_norm": 0.2908798158168793, "learning_rate": 1.5821396167044178e-07, "loss": 0.215, "step": 5942 }, { "epoch": 1.5814262905800958, "grad_norm": 0.2791889011859894, "learning_rate": 1.582000888023399e-07, "loss": 0.2229, "step": 5943 }, { "epoch": 1.5816923895689197, "grad_norm": 0.2995116710662842, "learning_rate": 1.581862142401846e-07, "loss": 0.2017, "step": 5944 }, { "epoch": 1.5819584885577433, "grad_norm": 0.2800939977169037, "learning_rate": 1.5817233798437966e-07, "loss": 0.2132, "step": 5945 }, { "epoch": 1.5822245875465675, "grad_norm": 0.29649272561073303, "learning_rate": 1.58158460035329e-07, "loss": 0.2088, "step": 5946 }, { "epoch": 1.5824906865353912, "grad_norm": 0.3059077262878418, "learning_rate": 1.5814458039343662e-07, "loss": 0.2237, "step": 5947 }, { "epoch": 1.582756785524215, "grad_norm": 0.3077349066734314, "learning_rate": 1.581306990591064e-07, "loss": 0.2112, "step": 5948 }, { "epoch": 1.583022884513039, "grad_norm": 0.2965260446071625, "learning_rate": 1.5811681603274252e-07, "loss": 0.2064, "step": 5949 }, { "epoch": 1.5832889835018626, "grad_norm": 0.28068798780441284, "learning_rate": 1.58102931314749e-07, "loss": 0.2142, "step": 5950 }, { "epoch": 1.5835550824906865, "grad_norm": 0.2989455759525299, "learning_rate": 1.5808904490553003e-07, "loss": 0.223, "step": 5951 }, { "epoch": 1.5838211814795105, "grad_norm": 0.28792956471443176, "learning_rate": 1.5807515680548977e-07, "loss": 0.1975, "step": 5952 }, { "epoch": 1.5840872804683341, "grad_norm": 0.3066748380661011, "learning_rate": 1.580612670150325e-07, "loss": 0.1903, "step": 5953 }, { "epoch": 1.584353379457158, "grad_norm": 0.30794423818588257, "learning_rate": 1.580473755345625e-07, "loss": 0.2181, "step": 5954 }, { "epoch": 1.584619478445982, "grad_norm": 0.2762504518032074, "learning_rate": 1.5803348236448412e-07, "loss": 0.1939, "step": 5955 }, { "epoch": 1.5848855774348056, "grad_norm": 0.31419074535369873, "learning_rate": 1.5801958750520173e-07, "loss": 0.2074, "step": 5956 }, { "epoch": 1.5851516764236298, "grad_norm": 0.27885475754737854, "learning_rate": 1.5800569095711981e-07, "loss": 0.1954, "step": 5957 }, { "epoch": 1.5854177754124534, "grad_norm": 0.2763127386569977, "learning_rate": 1.5799179272064284e-07, "loss": 0.2001, "step": 5958 }, { "epoch": 1.5856838744012771, "grad_norm": 0.30999955534935, "learning_rate": 1.5797789279617536e-07, "loss": 0.2007, "step": 5959 }, { "epoch": 1.5859499733901012, "grad_norm": 0.27844592928886414, "learning_rate": 1.5796399118412195e-07, "loss": 0.2107, "step": 5960 }, { "epoch": 1.586216072378925, "grad_norm": 0.31082862615585327, "learning_rate": 1.5795008788488725e-07, "loss": 0.21, "step": 5961 }, { "epoch": 1.5864821713677488, "grad_norm": 0.30542606115341187, "learning_rate": 1.5793618289887599e-07, "loss": 0.2191, "step": 5962 }, { "epoch": 1.5867482703565727, "grad_norm": 0.2650080621242523, "learning_rate": 1.5792227622649282e-07, "loss": 0.1858, "step": 5963 }, { "epoch": 1.5870143693453964, "grad_norm": 0.2771238684654236, "learning_rate": 1.5790836786814264e-07, "loss": 0.2246, "step": 5964 }, { "epoch": 1.5872804683342203, "grad_norm": 0.27077868580818176, "learning_rate": 1.578944578242302e-07, "loss": 0.1881, "step": 5965 }, { "epoch": 1.5875465673230442, "grad_norm": 0.289894163608551, "learning_rate": 1.5788054609516042e-07, "loss": 0.2107, "step": 5966 }, { "epoch": 1.587812666311868, "grad_norm": 0.2974430024623871, "learning_rate": 1.5786663268133822e-07, "loss": 0.218, "step": 5967 }, { "epoch": 1.5880787653006918, "grad_norm": 0.32511433959007263, "learning_rate": 1.5785271758316861e-07, "loss": 0.1999, "step": 5968 }, { "epoch": 1.5883448642895157, "grad_norm": 0.28436657786369324, "learning_rate": 1.5783880080105659e-07, "loss": 0.221, "step": 5969 }, { "epoch": 1.5886109632783394, "grad_norm": 0.30559974908828735, "learning_rate": 1.5782488233540728e-07, "loss": 0.2018, "step": 5970 }, { "epoch": 1.5888770622671635, "grad_norm": 0.2909489870071411, "learning_rate": 1.5781096218662574e-07, "loss": 0.2019, "step": 5971 }, { "epoch": 1.5891431612559872, "grad_norm": 0.3021303415298462, "learning_rate": 1.577970403551172e-07, "loss": 0.2011, "step": 5972 }, { "epoch": 1.589409260244811, "grad_norm": 0.28066274523735046, "learning_rate": 1.5778311684128694e-07, "loss": 0.2012, "step": 5973 }, { "epoch": 1.589675359233635, "grad_norm": 0.27854689955711365, "learning_rate": 1.5776919164554013e-07, "loss": 0.2173, "step": 5974 }, { "epoch": 1.5899414582224587, "grad_norm": 0.30815449357032776, "learning_rate": 1.5775526476828218e-07, "loss": 0.2135, "step": 5975 }, { "epoch": 1.5902075572112826, "grad_norm": 0.2581796646118164, "learning_rate": 1.5774133620991839e-07, "loss": 0.2046, "step": 5976 }, { "epoch": 1.5904736562001065, "grad_norm": 0.2941906750202179, "learning_rate": 1.5772740597085427e-07, "loss": 0.2052, "step": 5977 }, { "epoch": 1.5907397551889302, "grad_norm": 0.30143284797668457, "learning_rate": 1.5771347405149524e-07, "loss": 0.2026, "step": 5978 }, { "epoch": 1.591005854177754, "grad_norm": 0.2713354527950287, "learning_rate": 1.5769954045224681e-07, "loss": 0.1993, "step": 5979 }, { "epoch": 1.591271953166578, "grad_norm": 0.3119807839393616, "learning_rate": 1.576856051735146e-07, "loss": 0.2198, "step": 5980 }, { "epoch": 1.5915380521554017, "grad_norm": 0.2590358555316925, "learning_rate": 1.5767166821570416e-07, "loss": 0.1978, "step": 5981 }, { "epoch": 1.5918041511442258, "grad_norm": 0.28830522298812866, "learning_rate": 1.5765772957922126e-07, "loss": 0.2128, "step": 5982 }, { "epoch": 1.5920702501330495, "grad_norm": 0.2868286967277527, "learning_rate": 1.5764378926447154e-07, "loss": 0.2057, "step": 5983 }, { "epoch": 1.5923363491218734, "grad_norm": 0.2796379029750824, "learning_rate": 1.5762984727186077e-07, "loss": 0.1907, "step": 5984 }, { "epoch": 1.5926024481106973, "grad_norm": 0.33319026231765747, "learning_rate": 1.5761590360179477e-07, "loss": 0.1874, "step": 5985 }, { "epoch": 1.592868547099521, "grad_norm": 0.3139808475971222, "learning_rate": 1.5760195825467943e-07, "loss": 0.2234, "step": 5986 }, { "epoch": 1.5931346460883449, "grad_norm": 0.29220131039619446, "learning_rate": 1.5758801123092064e-07, "loss": 0.2091, "step": 5987 }, { "epoch": 1.5934007450771688, "grad_norm": 0.3859647214412689, "learning_rate": 1.5757406253092437e-07, "loss": 0.2004, "step": 5988 }, { "epoch": 1.5936668440659925, "grad_norm": 0.29396894574165344, "learning_rate": 1.5756011215509663e-07, "loss": 0.2011, "step": 5989 }, { "epoch": 1.5939329430548164, "grad_norm": 0.28505054116249084, "learning_rate": 1.5754616010384348e-07, "loss": 0.2097, "step": 5990 }, { "epoch": 1.5941990420436403, "grad_norm": 0.2782762050628662, "learning_rate": 1.5753220637757104e-07, "loss": 0.2072, "step": 5991 }, { "epoch": 1.594465141032464, "grad_norm": 0.2726694345474243, "learning_rate": 1.5751825097668545e-07, "loss": 0.2107, "step": 5992 }, { "epoch": 1.594731240021288, "grad_norm": 0.2864813506603241, "learning_rate": 1.5750429390159292e-07, "loss": 0.22, "step": 5993 }, { "epoch": 1.5949973390101118, "grad_norm": 0.2955746352672577, "learning_rate": 1.574903351526997e-07, "loss": 0.2087, "step": 5994 }, { "epoch": 1.5952634379989354, "grad_norm": 0.26337966322898865, "learning_rate": 1.5747637473041208e-07, "loss": 0.1887, "step": 5995 }, { "epoch": 1.5955295369877596, "grad_norm": 0.3325507640838623, "learning_rate": 1.5746241263513646e-07, "loss": 0.2001, "step": 5996 }, { "epoch": 1.5957956359765832, "grad_norm": 0.2866577208042145, "learning_rate": 1.5744844886727916e-07, "loss": 0.207, "step": 5997 }, { "epoch": 1.5960617349654072, "grad_norm": 0.28022462129592896, "learning_rate": 1.574344834272467e-07, "loss": 0.2001, "step": 5998 }, { "epoch": 1.596327833954231, "grad_norm": 0.292490154504776, "learning_rate": 1.5742051631544556e-07, "loss": 0.2095, "step": 5999 }, { "epoch": 1.5965939329430547, "grad_norm": 0.28709661960601807, "learning_rate": 1.5740654753228228e-07, "loss": 0.2084, "step": 6000 }, { "epoch": 1.5968600319318786, "grad_norm": 0.298633337020874, "learning_rate": 1.5739257707816345e-07, "loss": 0.2007, "step": 6001 }, { "epoch": 1.5971261309207025, "grad_norm": 0.2963992953300476, "learning_rate": 1.5737860495349574e-07, "loss": 0.2091, "step": 6002 }, { "epoch": 1.5973922299095262, "grad_norm": 0.30024227499961853, "learning_rate": 1.573646311586858e-07, "loss": 0.2128, "step": 6003 }, { "epoch": 1.5976583288983501, "grad_norm": 0.3171631395816803, "learning_rate": 1.573506556941404e-07, "loss": 0.2084, "step": 6004 }, { "epoch": 1.597924427887174, "grad_norm": 0.28918200731277466, "learning_rate": 1.5733667856026634e-07, "loss": 0.214, "step": 6005 }, { "epoch": 1.5981905268759977, "grad_norm": 0.2824185788631439, "learning_rate": 1.5732269975747042e-07, "loss": 0.1934, "step": 6006 }, { "epoch": 1.5984566258648218, "grad_norm": 0.29738545417785645, "learning_rate": 1.5730871928615953e-07, "loss": 0.2286, "step": 6007 }, { "epoch": 1.5987227248536455, "grad_norm": 0.28022468090057373, "learning_rate": 1.5729473714674063e-07, "loss": 0.1998, "step": 6008 }, { "epoch": 1.5989888238424694, "grad_norm": 0.28334060311317444, "learning_rate": 1.5728075333962071e-07, "loss": 0.2122, "step": 6009 }, { "epoch": 1.5992549228312933, "grad_norm": 0.28269660472869873, "learning_rate": 1.5726676786520678e-07, "loss": 0.1862, "step": 6010 }, { "epoch": 1.599521021820117, "grad_norm": 0.37141022086143494, "learning_rate": 1.5725278072390594e-07, "loss": 0.2158, "step": 6011 }, { "epoch": 1.599787120808941, "grad_norm": 0.32468193769454956, "learning_rate": 1.5723879191612527e-07, "loss": 0.2284, "step": 6012 }, { "epoch": 1.6000532197977648, "grad_norm": 0.29610690474510193, "learning_rate": 1.5722480144227203e-07, "loss": 0.2076, "step": 6013 }, { "epoch": 1.6003193187865885, "grad_norm": 0.2922115921974182, "learning_rate": 1.572108093027534e-07, "loss": 0.2269, "step": 6014 }, { "epoch": 1.6005854177754124, "grad_norm": 0.29645630717277527, "learning_rate": 1.5719681549797667e-07, "loss": 0.2084, "step": 6015 }, { "epoch": 1.6008515167642363, "grad_norm": 0.28565043210983276, "learning_rate": 1.571828200283491e-07, "loss": 0.1961, "step": 6016 }, { "epoch": 1.60111761575306, "grad_norm": 0.28892260789871216, "learning_rate": 1.5716882289427815e-07, "loss": 0.2208, "step": 6017 }, { "epoch": 1.6013837147418841, "grad_norm": 0.27573254704475403, "learning_rate": 1.5715482409617122e-07, "loss": 0.1748, "step": 6018 }, { "epoch": 1.6016498137307078, "grad_norm": 0.2828502058982849, "learning_rate": 1.5714082363443573e-07, "loss": 0.2019, "step": 6019 }, { "epoch": 1.6019159127195317, "grad_norm": 0.30221039056777954, "learning_rate": 1.5712682150947921e-07, "loss": 0.2324, "step": 6020 }, { "epoch": 1.6021820117083556, "grad_norm": 0.28257834911346436, "learning_rate": 1.5711281772170927e-07, "loss": 0.2072, "step": 6021 }, { "epoch": 1.6024481106971793, "grad_norm": 0.29275065660476685, "learning_rate": 1.570988122715335e-07, "loss": 0.217, "step": 6022 }, { "epoch": 1.6027142096860032, "grad_norm": 0.30057165026664734, "learning_rate": 1.5708480515935958e-07, "loss": 0.2212, "step": 6023 }, { "epoch": 1.602980308674827, "grad_norm": 0.277609258890152, "learning_rate": 1.5707079638559514e-07, "loss": 0.2078, "step": 6024 }, { "epoch": 1.6032464076636508, "grad_norm": 0.31681904196739197, "learning_rate": 1.5705678595064807e-07, "loss": 0.2394, "step": 6025 }, { "epoch": 1.6035125066524747, "grad_norm": 0.28913334012031555, "learning_rate": 1.5704277385492609e-07, "loss": 0.2153, "step": 6026 }, { "epoch": 1.6037786056412986, "grad_norm": 0.2858072519302368, "learning_rate": 1.5702876009883704e-07, "loss": 0.2, "step": 6027 }, { "epoch": 1.6040447046301223, "grad_norm": 0.2776874005794525, "learning_rate": 1.5701474468278887e-07, "loss": 0.2094, "step": 6028 }, { "epoch": 1.6043108036189464, "grad_norm": 0.2734650671482086, "learning_rate": 1.5700072760718954e-07, "loss": 0.2053, "step": 6029 }, { "epoch": 1.60457690260777, "grad_norm": 0.2580137550830841, "learning_rate": 1.56986708872447e-07, "loss": 0.1731, "step": 6030 }, { "epoch": 1.604843001596594, "grad_norm": 0.2812061607837677, "learning_rate": 1.5697268847896936e-07, "loss": 0.1916, "step": 6031 }, { "epoch": 1.605109100585418, "grad_norm": 0.28080689907073975, "learning_rate": 1.5695866642716467e-07, "loss": 0.1992, "step": 6032 }, { "epoch": 1.6053751995742416, "grad_norm": 0.2670900225639343, "learning_rate": 1.569446427174411e-07, "loss": 0.1941, "step": 6033 }, { "epoch": 1.6056412985630655, "grad_norm": 0.27500757575035095, "learning_rate": 1.5693061735020684e-07, "loss": 0.2036, "step": 6034 }, { "epoch": 1.6059073975518894, "grad_norm": 0.28645145893096924, "learning_rate": 1.5691659032587011e-07, "loss": 0.2122, "step": 6035 }, { "epoch": 1.606173496540713, "grad_norm": 0.28301334381103516, "learning_rate": 1.569025616448392e-07, "loss": 0.1934, "step": 6036 }, { "epoch": 1.606439595529537, "grad_norm": 0.30148550868034363, "learning_rate": 1.568885313075225e-07, "loss": 0.2213, "step": 6037 }, { "epoch": 1.6067056945183609, "grad_norm": 0.30303990840911865, "learning_rate": 1.5687449931432836e-07, "loss": 0.2151, "step": 6038 }, { "epoch": 1.6069717935071846, "grad_norm": 0.30695658922195435, "learning_rate": 1.5686046566566523e-07, "loss": 0.2096, "step": 6039 }, { "epoch": 1.6072378924960087, "grad_norm": 0.2739880383014679, "learning_rate": 1.5684643036194154e-07, "loss": 0.1907, "step": 6040 }, { "epoch": 1.6075039914848324, "grad_norm": 0.27596545219421387, "learning_rate": 1.5683239340356592e-07, "loss": 0.2021, "step": 6041 }, { "epoch": 1.607770090473656, "grad_norm": 0.2844869792461395, "learning_rate": 1.5681835479094683e-07, "loss": 0.1961, "step": 6042 }, { "epoch": 1.6080361894624802, "grad_norm": 0.29774168133735657, "learning_rate": 1.56804314524493e-07, "loss": 0.2092, "step": 6043 }, { "epoch": 1.6083022884513039, "grad_norm": 0.28765782713890076, "learning_rate": 1.5679027260461304e-07, "loss": 0.2161, "step": 6044 }, { "epoch": 1.6085683874401278, "grad_norm": 0.2566297650337219, "learning_rate": 1.5677622903171573e-07, "loss": 0.1936, "step": 6045 }, { "epoch": 1.6088344864289517, "grad_norm": 0.3038610816001892, "learning_rate": 1.5676218380620978e-07, "loss": 0.212, "step": 6046 }, { "epoch": 1.6091005854177753, "grad_norm": 0.28695356845855713, "learning_rate": 1.5674813692850406e-07, "loss": 0.2167, "step": 6047 }, { "epoch": 1.6093666844065992, "grad_norm": 0.2879659831523895, "learning_rate": 1.5673408839900742e-07, "loss": 0.1859, "step": 6048 }, { "epoch": 1.6096327833954232, "grad_norm": 0.28545475006103516, "learning_rate": 1.5672003821812877e-07, "loss": 0.2156, "step": 6049 }, { "epoch": 1.6098988823842468, "grad_norm": 0.2688868045806885, "learning_rate": 1.5670598638627704e-07, "loss": 0.2, "step": 6050 }, { "epoch": 1.6101649813730707, "grad_norm": 0.30009350180625916, "learning_rate": 1.5669193290386137e-07, "loss": 0.1979, "step": 6051 }, { "epoch": 1.6104310803618946, "grad_norm": 0.30351555347442627, "learning_rate": 1.5667787777129068e-07, "loss": 0.2131, "step": 6052 }, { "epoch": 1.6106971793507183, "grad_norm": 0.3161619305610657, "learning_rate": 1.566638209889741e-07, "loss": 0.2303, "step": 6053 }, { "epoch": 1.6109632783395424, "grad_norm": 0.297300785779953, "learning_rate": 1.5664976255732087e-07, "loss": 0.2128, "step": 6054 }, { "epoch": 1.6112293773283661, "grad_norm": 0.2777020037174225, "learning_rate": 1.566357024767401e-07, "loss": 0.2033, "step": 6055 }, { "epoch": 1.61149547631719, "grad_norm": 0.28499504923820496, "learning_rate": 1.5662164074764112e-07, "loss": 0.2017, "step": 6056 }, { "epoch": 1.611761575306014, "grad_norm": 0.28569552302360535, "learning_rate": 1.5660757737043318e-07, "loss": 0.2029, "step": 6057 }, { "epoch": 1.6120276742948376, "grad_norm": 0.28360018134117126, "learning_rate": 1.5659351234552566e-07, "loss": 0.2113, "step": 6058 }, { "epoch": 1.6122937732836615, "grad_norm": 0.26846352219581604, "learning_rate": 1.565794456733279e-07, "loss": 0.196, "step": 6059 }, { "epoch": 1.6125598722724854, "grad_norm": 0.2888031601905823, "learning_rate": 1.5656537735424942e-07, "loss": 0.2104, "step": 6060 }, { "epoch": 1.612825971261309, "grad_norm": 0.3201012909412384, "learning_rate": 1.5655130738869967e-07, "loss": 0.2386, "step": 6061 }, { "epoch": 1.613092070250133, "grad_norm": 0.28371670842170715, "learning_rate": 1.5653723577708817e-07, "loss": 0.1828, "step": 6062 }, { "epoch": 1.613358169238957, "grad_norm": 0.25803765654563904, "learning_rate": 1.5652316251982455e-07, "loss": 0.1945, "step": 6063 }, { "epoch": 1.6136242682277806, "grad_norm": 0.28131213784217834, "learning_rate": 1.5650908761731842e-07, "loss": 0.2055, "step": 6064 }, { "epoch": 1.6138903672166047, "grad_norm": 0.3054853677749634, "learning_rate": 1.5649501106997952e-07, "loss": 0.2024, "step": 6065 }, { "epoch": 1.6141564662054284, "grad_norm": 0.31936249136924744, "learning_rate": 1.564809328782175e-07, "loss": 0.211, "step": 6066 }, { "epoch": 1.6144225651942523, "grad_norm": 0.2834995985031128, "learning_rate": 1.564668530424422e-07, "loss": 0.2019, "step": 6067 }, { "epoch": 1.6146886641830762, "grad_norm": 0.2704993486404419, "learning_rate": 1.5645277156306337e-07, "loss": 0.2014, "step": 6068 }, { "epoch": 1.6149547631719, "grad_norm": 0.29249584674835205, "learning_rate": 1.5643868844049096e-07, "loss": 0.2239, "step": 6069 }, { "epoch": 1.6152208621607238, "grad_norm": 0.3028660714626312, "learning_rate": 1.5642460367513488e-07, "loss": 0.2112, "step": 6070 }, { "epoch": 1.6154869611495477, "grad_norm": 0.2821190059185028, "learning_rate": 1.5641051726740505e-07, "loss": 0.2017, "step": 6071 }, { "epoch": 1.6157530601383714, "grad_norm": 0.2765220105648041, "learning_rate": 1.5639642921771162e-07, "loss": 0.1944, "step": 6072 }, { "epoch": 1.6160191591271953, "grad_norm": 0.2879282236099243, "learning_rate": 1.563823395264645e-07, "loss": 0.1971, "step": 6073 }, { "epoch": 1.6162852581160192, "grad_norm": 0.2607193887233734, "learning_rate": 1.5636824819407387e-07, "loss": 0.1924, "step": 6074 }, { "epoch": 1.6165513571048429, "grad_norm": 0.2929001748561859, "learning_rate": 1.5635415522094993e-07, "loss": 0.2031, "step": 6075 }, { "epoch": 1.616817456093667, "grad_norm": 0.27269884943962097, "learning_rate": 1.5634006060750284e-07, "loss": 0.1963, "step": 6076 }, { "epoch": 1.6170835550824907, "grad_norm": 0.29688915610313416, "learning_rate": 1.5632596435414288e-07, "loss": 0.2147, "step": 6077 }, { "epoch": 1.6173496540713144, "grad_norm": 0.3070475459098816, "learning_rate": 1.5631186646128032e-07, "loss": 0.2372, "step": 6078 }, { "epoch": 1.6176157530601385, "grad_norm": 0.294222354888916, "learning_rate": 1.5629776692932555e-07, "loss": 0.2078, "step": 6079 }, { "epoch": 1.6178818520489622, "grad_norm": 0.28876620531082153, "learning_rate": 1.5628366575868897e-07, "loss": 0.2077, "step": 6080 }, { "epoch": 1.618147951037786, "grad_norm": 0.3170943260192871, "learning_rate": 1.5626956294978102e-07, "loss": 0.2058, "step": 6081 }, { "epoch": 1.61841405002661, "grad_norm": 0.2706688344478607, "learning_rate": 1.562554585030122e-07, "loss": 0.2076, "step": 6082 }, { "epoch": 1.6186801490154337, "grad_norm": 0.288387656211853, "learning_rate": 1.56241352418793e-07, "loss": 0.2196, "step": 6083 }, { "epoch": 1.6189462480042576, "grad_norm": 0.29748785495758057, "learning_rate": 1.5622724469753412e-07, "loss": 0.2182, "step": 6084 }, { "epoch": 1.6192123469930815, "grad_norm": 0.6304944157600403, "learning_rate": 1.5621313533964611e-07, "loss": 0.2063, "step": 6085 }, { "epoch": 1.6194784459819052, "grad_norm": 0.2976074814796448, "learning_rate": 1.561990243455397e-07, "loss": 0.2115, "step": 6086 }, { "epoch": 1.619744544970729, "grad_norm": 0.3065132200717926, "learning_rate": 1.561849117156256e-07, "loss": 0.2139, "step": 6087 }, { "epoch": 1.620010643959553, "grad_norm": 0.30858683586120605, "learning_rate": 1.5617079745031462e-07, "loss": 0.2242, "step": 6088 }, { "epoch": 1.6202767429483766, "grad_norm": 0.27532702684402466, "learning_rate": 1.5615668155001753e-07, "loss": 0.1981, "step": 6089 }, { "epoch": 1.6205428419372008, "grad_norm": 0.2928226888179779, "learning_rate": 1.561425640151453e-07, "loss": 0.1957, "step": 6090 }, { "epoch": 1.6208089409260245, "grad_norm": 0.29187995195388794, "learning_rate": 1.5612844484610878e-07, "loss": 0.207, "step": 6091 }, { "epoch": 1.6210750399148484, "grad_norm": 0.30228155851364136, "learning_rate": 1.5611432404331896e-07, "loss": 0.2175, "step": 6092 }, { "epoch": 1.6213411389036723, "grad_norm": 0.27756476402282715, "learning_rate": 1.5610020160718688e-07, "loss": 0.1981, "step": 6093 }, { "epoch": 1.621607237892496, "grad_norm": 0.26346534490585327, "learning_rate": 1.560860775381236e-07, "loss": 0.1919, "step": 6094 }, { "epoch": 1.6218733368813198, "grad_norm": 0.2682260572910309, "learning_rate": 1.5607195183654022e-07, "loss": 0.2053, "step": 6095 }, { "epoch": 1.6221394358701438, "grad_norm": 0.3035465180873871, "learning_rate": 1.560578245028479e-07, "loss": 0.2265, "step": 6096 }, { "epoch": 1.6224055348589674, "grad_norm": 0.33525195717811584, "learning_rate": 1.5604369553745786e-07, "loss": 0.2347, "step": 6097 }, { "epoch": 1.6226716338477913, "grad_norm": 0.29960203170776367, "learning_rate": 1.5602956494078137e-07, "loss": 0.213, "step": 6098 }, { "epoch": 1.6229377328366152, "grad_norm": 0.3084997832775116, "learning_rate": 1.5601543271322974e-07, "loss": 0.2081, "step": 6099 }, { "epoch": 1.623203831825439, "grad_norm": 0.2838175296783447, "learning_rate": 1.5600129885521427e-07, "loss": 0.2023, "step": 6100 }, { "epoch": 1.623469930814263, "grad_norm": 0.30016177892684937, "learning_rate": 1.5598716336714645e-07, "loss": 0.1977, "step": 6101 }, { "epoch": 1.6237360298030867, "grad_norm": 0.27829986810684204, "learning_rate": 1.559730262494376e-07, "loss": 0.2004, "step": 6102 }, { "epoch": 1.6240021287919106, "grad_norm": 0.2766217887401581, "learning_rate": 1.5595888750249935e-07, "loss": 0.1998, "step": 6103 }, { "epoch": 1.6242682277807345, "grad_norm": 0.2728029191493988, "learning_rate": 1.5594474712674317e-07, "loss": 0.2008, "step": 6104 }, { "epoch": 1.6245343267695582, "grad_norm": 0.27016782760620117, "learning_rate": 1.5593060512258064e-07, "loss": 0.19, "step": 6105 }, { "epoch": 1.6248004257583821, "grad_norm": 0.284076452255249, "learning_rate": 1.559164614904234e-07, "loss": 0.2149, "step": 6106 }, { "epoch": 1.625066524747206, "grad_norm": 0.2906135320663452, "learning_rate": 1.559023162306832e-07, "loss": 0.2116, "step": 6107 }, { "epoch": 1.6253326237360297, "grad_norm": 0.27286645770072937, "learning_rate": 1.5588816934377171e-07, "loss": 0.2003, "step": 6108 }, { "epoch": 1.6255987227248536, "grad_norm": 0.2733291983604431, "learning_rate": 1.558740208301007e-07, "loss": 0.2053, "step": 6109 }, { "epoch": 1.6258648217136775, "grad_norm": 0.2752580940723419, "learning_rate": 1.5585987069008203e-07, "loss": 0.2077, "step": 6110 }, { "epoch": 1.6261309207025012, "grad_norm": 0.2639964818954468, "learning_rate": 1.5584571892412757e-07, "loss": 0.1935, "step": 6111 }, { "epoch": 1.6263970196913253, "grad_norm": 0.29846784472465515, "learning_rate": 1.5583156553264923e-07, "loss": 0.1992, "step": 6112 }, { "epoch": 1.626663118680149, "grad_norm": 0.2694871425628662, "learning_rate": 1.5581741051605898e-07, "loss": 0.1855, "step": 6113 }, { "epoch": 1.6269292176689727, "grad_norm": 0.26559826731681824, "learning_rate": 1.5580325387476886e-07, "loss": 0.2029, "step": 6114 }, { "epoch": 1.6271953166577968, "grad_norm": 0.28739646077156067, "learning_rate": 1.5578909560919088e-07, "loss": 0.208, "step": 6115 }, { "epoch": 1.6274614156466205, "grad_norm": 0.31013184785842896, "learning_rate": 1.557749357197372e-07, "loss": 0.2147, "step": 6116 }, { "epoch": 1.6277275146354444, "grad_norm": 0.28834769129753113, "learning_rate": 1.5576077420681994e-07, "loss": 0.2037, "step": 6117 }, { "epoch": 1.6279936136242683, "grad_norm": 0.27240198850631714, "learning_rate": 1.5574661107085136e-07, "loss": 0.2056, "step": 6118 }, { "epoch": 1.628259712613092, "grad_norm": 0.39644843339920044, "learning_rate": 1.5573244631224362e-07, "loss": 0.2068, "step": 6119 }, { "epoch": 1.628525811601916, "grad_norm": 0.30120983719825745, "learning_rate": 1.5571827993140913e-07, "loss": 0.2124, "step": 6120 }, { "epoch": 1.6287919105907398, "grad_norm": 0.2747046947479248, "learning_rate": 1.5570411192876015e-07, "loss": 0.1965, "step": 6121 }, { "epoch": 1.6290580095795635, "grad_norm": 0.27230966091156006, "learning_rate": 1.5568994230470913e-07, "loss": 0.1828, "step": 6122 }, { "epoch": 1.6293241085683874, "grad_norm": 0.2927454710006714, "learning_rate": 1.5567577105966848e-07, "loss": 0.2083, "step": 6123 }, { "epoch": 1.6295902075572113, "grad_norm": 0.3051891624927521, "learning_rate": 1.556615981940507e-07, "loss": 0.2182, "step": 6124 }, { "epoch": 1.629856306546035, "grad_norm": 0.3046305179595947, "learning_rate": 1.556474237082683e-07, "loss": 0.2022, "step": 6125 }, { "epoch": 1.630122405534859, "grad_norm": 0.29581695795059204, "learning_rate": 1.556332476027339e-07, "loss": 0.227, "step": 6126 }, { "epoch": 1.6303885045236828, "grad_norm": 0.46942657232284546, "learning_rate": 1.556190698778601e-07, "loss": 0.21, "step": 6127 }, { "epoch": 1.6306546035125067, "grad_norm": 0.2905982434749603, "learning_rate": 1.556048905340596e-07, "loss": 0.2068, "step": 6128 }, { "epoch": 1.6309207025013306, "grad_norm": 0.3066267967224121, "learning_rate": 1.555907095717451e-07, "loss": 0.2269, "step": 6129 }, { "epoch": 1.6311868014901543, "grad_norm": 0.2940993905067444, "learning_rate": 1.555765269913294e-07, "loss": 0.2056, "step": 6130 }, { "epoch": 1.6314529004789782, "grad_norm": 0.26895871758461, "learning_rate": 1.5556234279322528e-07, "loss": 0.2083, "step": 6131 }, { "epoch": 1.631718999467802, "grad_norm": 0.28306758403778076, "learning_rate": 1.5554815697784564e-07, "loss": 0.1982, "step": 6132 }, { "epoch": 1.6319850984566258, "grad_norm": 0.29495105147361755, "learning_rate": 1.5553396954560337e-07, "loss": 0.201, "step": 6133 }, { "epoch": 1.6322511974454497, "grad_norm": 0.2947186529636383, "learning_rate": 1.5551978049691146e-07, "loss": 0.2004, "step": 6134 }, { "epoch": 1.6325172964342736, "grad_norm": 0.2883055806159973, "learning_rate": 1.555055898321829e-07, "loss": 0.2157, "step": 6135 }, { "epoch": 1.6327833954230973, "grad_norm": 0.2908290922641754, "learning_rate": 1.5549139755183073e-07, "loss": 0.1959, "step": 6136 }, { "epoch": 1.6330494944119214, "grad_norm": 0.3007413148880005, "learning_rate": 1.5547720365626806e-07, "loss": 0.2069, "step": 6137 }, { "epoch": 1.633315593400745, "grad_norm": 0.3079388439655304, "learning_rate": 1.55463008145908e-07, "loss": 0.2021, "step": 6138 }, { "epoch": 1.633581692389569, "grad_norm": 0.3036147952079773, "learning_rate": 1.5544881102116381e-07, "loss": 0.2153, "step": 6139 }, { "epoch": 1.6338477913783929, "grad_norm": 0.29079440236091614, "learning_rate": 1.554346122824487e-07, "loss": 0.2167, "step": 6140 }, { "epoch": 1.6341138903672165, "grad_norm": 0.3534318804740906, "learning_rate": 1.55420411930176e-07, "loss": 0.1961, "step": 6141 }, { "epoch": 1.6343799893560405, "grad_norm": 0.28707343339920044, "learning_rate": 1.5540620996475896e-07, "loss": 0.1979, "step": 6142 }, { "epoch": 1.6346460883448644, "grad_norm": 0.2902940511703491, "learning_rate": 1.5539200638661102e-07, "loss": 0.2008, "step": 6143 }, { "epoch": 1.634912187333688, "grad_norm": 0.26197439432144165, "learning_rate": 1.5537780119614562e-07, "loss": 0.1975, "step": 6144 }, { "epoch": 1.635178286322512, "grad_norm": 0.3008262813091278, "learning_rate": 1.5536359439377616e-07, "loss": 0.2283, "step": 6145 }, { "epoch": 1.6354443853113358, "grad_norm": 0.28568440675735474, "learning_rate": 1.5534938597991626e-07, "loss": 0.1982, "step": 6146 }, { "epoch": 1.6357104843001595, "grad_norm": 0.25286218523979187, "learning_rate": 1.5533517595497944e-07, "loss": 0.1835, "step": 6147 }, { "epoch": 1.6359765832889837, "grad_norm": 0.30814188718795776, "learning_rate": 1.553209643193793e-07, "loss": 0.2253, "step": 6148 }, { "epoch": 1.6362426822778073, "grad_norm": 0.29614177346229553, "learning_rate": 1.5530675107352958e-07, "loss": 0.2154, "step": 6149 }, { "epoch": 1.6365087812666312, "grad_norm": 0.3029862642288208, "learning_rate": 1.552925362178439e-07, "loss": 0.2124, "step": 6150 }, { "epoch": 1.6367748802554551, "grad_norm": 0.3091874122619629, "learning_rate": 1.5527831975273607e-07, "loss": 0.24, "step": 6151 }, { "epoch": 1.6370409792442788, "grad_norm": 0.276212602853775, "learning_rate": 1.5526410167861988e-07, "loss": 0.1997, "step": 6152 }, { "epoch": 1.6373070782331027, "grad_norm": 0.29101303219795227, "learning_rate": 1.552498819959092e-07, "loss": 0.1812, "step": 6153 }, { "epoch": 1.6375731772219266, "grad_norm": 0.27228695154190063, "learning_rate": 1.5523566070501785e-07, "loss": 0.1961, "step": 6154 }, { "epoch": 1.6378392762107503, "grad_norm": 0.2938539683818817, "learning_rate": 1.552214378063599e-07, "loss": 0.1999, "step": 6155 }, { "epoch": 1.6381053751995742, "grad_norm": 0.30476170778274536, "learning_rate": 1.5520721330034923e-07, "loss": 0.2115, "step": 6156 }, { "epoch": 1.6383714741883981, "grad_norm": 0.2673094570636749, "learning_rate": 1.5519298718739994e-07, "loss": 0.1926, "step": 6157 }, { "epoch": 1.6386375731772218, "grad_norm": 0.28363436460494995, "learning_rate": 1.5517875946792614e-07, "loss": 0.2022, "step": 6158 }, { "epoch": 1.638903672166046, "grad_norm": 0.28951844573020935, "learning_rate": 1.5516453014234186e-07, "loss": 0.2158, "step": 6159 }, { "epoch": 1.6391697711548696, "grad_norm": 0.30791231989860535, "learning_rate": 1.5515029921106138e-07, "loss": 0.2272, "step": 6160 }, { "epoch": 1.6394358701436933, "grad_norm": 0.7136533260345459, "learning_rate": 1.5513606667449884e-07, "loss": 0.2111, "step": 6161 }, { "epoch": 1.6397019691325174, "grad_norm": 0.28663912415504456, "learning_rate": 1.5512183253306862e-07, "loss": 0.2091, "step": 6162 }, { "epoch": 1.639968068121341, "grad_norm": 0.28921225666999817, "learning_rate": 1.5510759678718494e-07, "loss": 0.2191, "step": 6163 }, { "epoch": 1.640234167110165, "grad_norm": 0.2969372868537903, "learning_rate": 1.5509335943726223e-07, "loss": 0.2126, "step": 6164 }, { "epoch": 1.640500266098989, "grad_norm": 0.3075863718986511, "learning_rate": 1.5507912048371483e-07, "loss": 0.2089, "step": 6165 }, { "epoch": 1.6407663650878126, "grad_norm": 0.30386218428611755, "learning_rate": 1.5506487992695726e-07, "loss": 0.2052, "step": 6166 }, { "epoch": 1.6410324640766365, "grad_norm": 0.28862303495407104, "learning_rate": 1.5505063776740403e-07, "loss": 0.1987, "step": 6167 }, { "epoch": 1.6412985630654604, "grad_norm": 0.2703950107097626, "learning_rate": 1.5503639400546967e-07, "loss": 0.2017, "step": 6168 }, { "epoch": 1.641564662054284, "grad_norm": 0.28900593519210815, "learning_rate": 1.5502214864156878e-07, "loss": 0.2136, "step": 6169 }, { "epoch": 1.641830761043108, "grad_norm": 0.265975683927536, "learning_rate": 1.55007901676116e-07, "loss": 0.1941, "step": 6170 }, { "epoch": 1.642096860031932, "grad_norm": 0.2937730550765991, "learning_rate": 1.5499365310952602e-07, "loss": 0.1917, "step": 6171 }, { "epoch": 1.6423629590207556, "grad_norm": 0.2850506603717804, "learning_rate": 1.5497940294221362e-07, "loss": 0.2227, "step": 6172 }, { "epoch": 1.6426290580095797, "grad_norm": 0.26806920766830444, "learning_rate": 1.5496515117459353e-07, "loss": 0.1985, "step": 6173 }, { "epoch": 1.6428951569984034, "grad_norm": 0.30385878682136536, "learning_rate": 1.549508978070806e-07, "loss": 0.22, "step": 6174 }, { "epoch": 1.6431612559872273, "grad_norm": 0.2566565275192261, "learning_rate": 1.5493664284008972e-07, "loss": 0.1929, "step": 6175 }, { "epoch": 1.6434273549760512, "grad_norm": 0.29976487159729004, "learning_rate": 1.5492238627403582e-07, "loss": 0.2038, "step": 6176 }, { "epoch": 1.6436934539648749, "grad_norm": 0.29387155175209045, "learning_rate": 1.5490812810933384e-07, "loss": 0.2032, "step": 6177 }, { "epoch": 1.6439595529536988, "grad_norm": 0.27727192640304565, "learning_rate": 1.5489386834639882e-07, "loss": 0.213, "step": 6178 }, { "epoch": 1.6442256519425227, "grad_norm": 0.2962554693222046, "learning_rate": 1.548796069856458e-07, "loss": 0.2301, "step": 6179 }, { "epoch": 1.6444917509313464, "grad_norm": 0.31305477023124695, "learning_rate": 1.5486534402748999e-07, "loss": 0.2112, "step": 6180 }, { "epoch": 1.6447578499201703, "grad_norm": 0.29801198840141296, "learning_rate": 1.5485107947234642e-07, "loss": 0.1979, "step": 6181 }, { "epoch": 1.6450239489089942, "grad_norm": 0.2960793077945709, "learning_rate": 1.5483681332063033e-07, "loss": 0.202, "step": 6182 }, { "epoch": 1.6452900478978179, "grad_norm": 0.31877201795578003, "learning_rate": 1.5482254557275703e-07, "loss": 0.2137, "step": 6183 }, { "epoch": 1.645556146886642, "grad_norm": 0.27966219186782837, "learning_rate": 1.5480827622914175e-07, "loss": 0.2041, "step": 6184 }, { "epoch": 1.6458222458754657, "grad_norm": 0.2768496870994568, "learning_rate": 1.5479400529019984e-07, "loss": 0.2206, "step": 6185 }, { "epoch": 1.6460883448642896, "grad_norm": 0.2720966935157776, "learning_rate": 1.5477973275634674e-07, "loss": 0.1934, "step": 6186 }, { "epoch": 1.6463544438531135, "grad_norm": 0.27967435121536255, "learning_rate": 1.5476545862799784e-07, "loss": 0.2077, "step": 6187 }, { "epoch": 1.6466205428419372, "grad_norm": 0.2869836688041687, "learning_rate": 1.5475118290556862e-07, "loss": 0.1977, "step": 6188 }, { "epoch": 1.646886641830761, "grad_norm": 0.28760865330696106, "learning_rate": 1.5473690558947465e-07, "loss": 0.2135, "step": 6189 }, { "epoch": 1.647152740819585, "grad_norm": 0.2500157058238983, "learning_rate": 1.5472262668013148e-07, "loss": 0.1766, "step": 6190 }, { "epoch": 1.6474188398084086, "grad_norm": 0.38387221097946167, "learning_rate": 1.547083461779547e-07, "loss": 0.1934, "step": 6191 }, { "epoch": 1.6476849387972325, "grad_norm": 0.3084285855293274, "learning_rate": 1.5469406408336005e-07, "loss": 0.213, "step": 6192 }, { "epoch": 1.6479510377860564, "grad_norm": 0.3071926534175873, "learning_rate": 1.546797803967632e-07, "loss": 0.209, "step": 6193 }, { "epoch": 1.6482171367748801, "grad_norm": 0.293050080537796, "learning_rate": 1.546654951185799e-07, "loss": 0.2168, "step": 6194 }, { "epoch": 1.6484832357637043, "grad_norm": 0.28646767139434814, "learning_rate": 1.5465120824922598e-07, "loss": 0.2069, "step": 6195 }, { "epoch": 1.648749334752528, "grad_norm": 0.2830461263656616, "learning_rate": 1.5463691978911732e-07, "loss": 0.206, "step": 6196 }, { "epoch": 1.6490154337413516, "grad_norm": 0.31947991251945496, "learning_rate": 1.5462262973866972e-07, "loss": 0.2233, "step": 6197 }, { "epoch": 1.6492815327301757, "grad_norm": 0.29599493741989136, "learning_rate": 1.5460833809829924e-07, "loss": 0.1926, "step": 6198 }, { "epoch": 1.6495476317189994, "grad_norm": 0.2623303532600403, "learning_rate": 1.5459404486842184e-07, "loss": 0.1877, "step": 6199 }, { "epoch": 1.6498137307078233, "grad_norm": 0.27080678939819336, "learning_rate": 1.5457975004945353e-07, "loss": 0.2092, "step": 6200 }, { "epoch": 1.6500798296966472, "grad_norm": 0.29280611872673035, "learning_rate": 1.545654536418104e-07, "loss": 0.2055, "step": 6201 }, { "epoch": 1.650345928685471, "grad_norm": 0.46401724219322205, "learning_rate": 1.5455115564590863e-07, "loss": 0.2392, "step": 6202 }, { "epoch": 1.6506120276742948, "grad_norm": 0.2781248092651367, "learning_rate": 1.5453685606216431e-07, "loss": 0.2022, "step": 6203 }, { "epoch": 1.6508781266631187, "grad_norm": 0.27615827322006226, "learning_rate": 1.5452255489099376e-07, "loss": 0.2013, "step": 6204 }, { "epoch": 1.6511442256519424, "grad_norm": 0.30944517254829407, "learning_rate": 1.5450825213281316e-07, "loss": 0.2179, "step": 6205 }, { "epoch": 1.6514103246407663, "grad_norm": 0.29686641693115234, "learning_rate": 1.5449394778803888e-07, "loss": 0.1897, "step": 6206 }, { "epoch": 1.6516764236295902, "grad_norm": 0.2934468686580658, "learning_rate": 1.544796418570873e-07, "loss": 0.2181, "step": 6207 }, { "epoch": 1.651942522618414, "grad_norm": 0.3047161400318146, "learning_rate": 1.5446533434037478e-07, "loss": 0.2181, "step": 6208 }, { "epoch": 1.652208621607238, "grad_norm": 0.2958424687385559, "learning_rate": 1.544510252383178e-07, "loss": 0.2066, "step": 6209 }, { "epoch": 1.6524747205960617, "grad_norm": 0.35622933506965637, "learning_rate": 1.5443671455133284e-07, "loss": 0.1978, "step": 6210 }, { "epoch": 1.6527408195848856, "grad_norm": 0.2891838550567627, "learning_rate": 1.5442240227983648e-07, "loss": 0.1991, "step": 6211 }, { "epoch": 1.6530069185737095, "grad_norm": 0.3089326024055481, "learning_rate": 1.544080884242453e-07, "loss": 0.2099, "step": 6212 }, { "epoch": 1.6532730175625332, "grad_norm": 0.2673177719116211, "learning_rate": 1.543937729849759e-07, "loss": 0.1947, "step": 6213 }, { "epoch": 1.653539116551357, "grad_norm": 0.28367266058921814, "learning_rate": 1.5437945596244504e-07, "loss": 0.206, "step": 6214 }, { "epoch": 1.653805215540181, "grad_norm": 0.2797791659832001, "learning_rate": 1.543651373570694e-07, "loss": 0.2062, "step": 6215 }, { "epoch": 1.6540713145290047, "grad_norm": 0.287940114736557, "learning_rate": 1.5435081716926575e-07, "loss": 0.2035, "step": 6216 }, { "epoch": 1.6543374135178286, "grad_norm": 0.2897278368473053, "learning_rate": 1.5433649539945098e-07, "loss": 0.2184, "step": 6217 }, { "epoch": 1.6546035125066525, "grad_norm": 0.32247892022132874, "learning_rate": 1.5432217204804187e-07, "loss": 0.1855, "step": 6218 }, { "epoch": 1.6548696114954762, "grad_norm": 0.2914626896381378, "learning_rate": 1.543078471154554e-07, "loss": 0.2021, "step": 6219 }, { "epoch": 1.6551357104843003, "grad_norm": 0.2587285339832306, "learning_rate": 1.542935206021085e-07, "loss": 0.1861, "step": 6220 }, { "epoch": 1.655401809473124, "grad_norm": 0.2846858501434326, "learning_rate": 1.542791925084182e-07, "loss": 0.1963, "step": 6221 }, { "epoch": 1.655667908461948, "grad_norm": 0.29616573452949524, "learning_rate": 1.5426486283480151e-07, "loss": 0.1885, "step": 6222 }, { "epoch": 1.6559340074507718, "grad_norm": 0.3002754747867584, "learning_rate": 1.542505315816756e-07, "loss": 0.1996, "step": 6223 }, { "epoch": 1.6562001064395955, "grad_norm": 0.28506892919540405, "learning_rate": 1.5423619874945754e-07, "loss": 0.1808, "step": 6224 }, { "epoch": 1.6564662054284194, "grad_norm": 0.29907095432281494, "learning_rate": 1.5422186433856457e-07, "loss": 0.2122, "step": 6225 }, { "epoch": 1.6567323044172433, "grad_norm": 0.2594096064567566, "learning_rate": 1.5420752834941394e-07, "loss": 0.1886, "step": 6226 }, { "epoch": 1.656998403406067, "grad_norm": 0.2774294316768646, "learning_rate": 1.5419319078242285e-07, "loss": 0.1823, "step": 6227 }, { "epoch": 1.6572645023948909, "grad_norm": 0.2715829908847809, "learning_rate": 1.5417885163800873e-07, "loss": 0.1952, "step": 6228 }, { "epoch": 1.6575306013837148, "grad_norm": 0.29153844714164734, "learning_rate": 1.541645109165889e-07, "loss": 0.2102, "step": 6229 }, { "epoch": 1.6577967003725385, "grad_norm": 0.28365975618362427, "learning_rate": 1.541501686185808e-07, "loss": 0.1971, "step": 6230 }, { "epoch": 1.6580627993613626, "grad_norm": 0.29098227620124817, "learning_rate": 1.5413582474440188e-07, "loss": 0.1887, "step": 6231 }, { "epoch": 1.6583288983501863, "grad_norm": 0.32413527369499207, "learning_rate": 1.5412147929446968e-07, "loss": 0.2152, "step": 6232 }, { "epoch": 1.65859499733901, "grad_norm": 0.2855300307273865, "learning_rate": 1.5410713226920173e-07, "loss": 0.2057, "step": 6233 }, { "epoch": 1.658861096327834, "grad_norm": 0.23928828537464142, "learning_rate": 1.5409278366901567e-07, "loss": 0.1657, "step": 6234 }, { "epoch": 1.6591271953166578, "grad_norm": 0.2803589105606079, "learning_rate": 1.540784334943291e-07, "loss": 0.1995, "step": 6235 }, { "epoch": 1.6593932943054817, "grad_norm": 0.2897261381149292, "learning_rate": 1.5406408174555975e-07, "loss": 0.1996, "step": 6236 }, { "epoch": 1.6596593932943056, "grad_norm": 0.2894885540008545, "learning_rate": 1.5404972842312536e-07, "loss": 0.2035, "step": 6237 }, { "epoch": 1.6599254922831292, "grad_norm": 0.27717018127441406, "learning_rate": 1.5403537352744373e-07, "loss": 0.2055, "step": 6238 }, { "epoch": 1.6601915912719531, "grad_norm": 0.28912004828453064, "learning_rate": 1.5402101705893267e-07, "loss": 0.1961, "step": 6239 }, { "epoch": 1.660457690260777, "grad_norm": 0.27555206418037415, "learning_rate": 1.5400665901801006e-07, "loss": 0.1845, "step": 6240 }, { "epoch": 1.6607237892496007, "grad_norm": 0.28375914692878723, "learning_rate": 1.5399229940509382e-07, "loss": 0.2173, "step": 6241 }, { "epoch": 1.6609898882384246, "grad_norm": 0.25923028588294983, "learning_rate": 1.5397793822060194e-07, "loss": 0.1876, "step": 6242 }, { "epoch": 1.6612559872272485, "grad_norm": 0.27226462960243225, "learning_rate": 1.5396357546495243e-07, "loss": 0.2093, "step": 6243 }, { "epoch": 1.6615220862160722, "grad_norm": 0.29012227058410645, "learning_rate": 1.539492111385634e-07, "loss": 0.2175, "step": 6244 }, { "epoch": 1.6617881852048964, "grad_norm": 0.2824667990207672, "learning_rate": 1.5393484524185284e-07, "loss": 0.1943, "step": 6245 }, { "epoch": 1.66205428419372, "grad_norm": 0.2611207365989685, "learning_rate": 1.5392047777523902e-07, "loss": 0.1781, "step": 6246 }, { "epoch": 1.662320383182544, "grad_norm": 0.2817305624485016, "learning_rate": 1.5390610873914008e-07, "loss": 0.2047, "step": 6247 }, { "epoch": 1.6625864821713678, "grad_norm": 0.2740084230899811, "learning_rate": 1.538917381339743e-07, "loss": 0.1868, "step": 6248 }, { "epoch": 1.6628525811601915, "grad_norm": 0.27718591690063477, "learning_rate": 1.5387736596015994e-07, "loss": 0.2044, "step": 6249 }, { "epoch": 1.6631186801490154, "grad_norm": 0.2836194932460785, "learning_rate": 1.5386299221811533e-07, "loss": 0.2038, "step": 6250 }, { "epoch": 1.6633847791378393, "grad_norm": 0.26773571968078613, "learning_rate": 1.538486169082589e-07, "loss": 0.2093, "step": 6251 }, { "epoch": 1.663650878126663, "grad_norm": 0.27671706676483154, "learning_rate": 1.53834240031009e-07, "loss": 0.2196, "step": 6252 }, { "epoch": 1.663916977115487, "grad_norm": 0.2699076533317566, "learning_rate": 1.538198615867842e-07, "loss": 0.1815, "step": 6253 }, { "epoch": 1.6641830761043108, "grad_norm": 0.29985836148262024, "learning_rate": 1.5380548157600294e-07, "loss": 0.2177, "step": 6254 }, { "epoch": 1.6644491750931345, "grad_norm": 0.28600767254829407, "learning_rate": 1.5379109999908385e-07, "loss": 0.2113, "step": 6255 }, { "epoch": 1.6647152740819586, "grad_norm": 0.28585782647132874, "learning_rate": 1.5377671685644549e-07, "loss": 0.2209, "step": 6256 }, { "epoch": 1.6649813730707823, "grad_norm": 0.2802051901817322, "learning_rate": 1.5376233214850656e-07, "loss": 0.1997, "step": 6257 }, { "epoch": 1.6652474720596062, "grad_norm": 0.2820582091808319, "learning_rate": 1.5374794587568568e-07, "loss": 0.2063, "step": 6258 }, { "epoch": 1.6655135710484301, "grad_norm": 0.2887101471424103, "learning_rate": 1.537335580384017e-07, "loss": 0.1971, "step": 6259 }, { "epoch": 1.6657796700372538, "grad_norm": 0.2949484586715698, "learning_rate": 1.537191686370733e-07, "loss": 0.2, "step": 6260 }, { "epoch": 1.6660457690260777, "grad_norm": 0.29478374123573303, "learning_rate": 1.5370477767211942e-07, "loss": 0.2183, "step": 6261 }, { "epoch": 1.6663118680149016, "grad_norm": 0.2798481285572052, "learning_rate": 1.536903851439589e-07, "loss": 0.2129, "step": 6262 }, { "epoch": 1.6665779670037253, "grad_norm": 0.30066409707069397, "learning_rate": 1.5367599105301068e-07, "loss": 0.1951, "step": 6263 }, { "epoch": 1.6668440659925492, "grad_norm": 0.26049235463142395, "learning_rate": 1.5366159539969373e-07, "loss": 0.196, "step": 6264 }, { "epoch": 1.667110164981373, "grad_norm": 0.26635995507240295, "learning_rate": 1.5364719818442705e-07, "loss": 0.1927, "step": 6265 }, { "epoch": 1.6673762639701968, "grad_norm": 0.2970098853111267, "learning_rate": 1.5363279940762976e-07, "loss": 0.2063, "step": 6266 }, { "epoch": 1.667642362959021, "grad_norm": 0.2835790514945984, "learning_rate": 1.5361839906972093e-07, "loss": 0.216, "step": 6267 }, { "epoch": 1.6679084619478446, "grad_norm": 0.2939968407154083, "learning_rate": 1.5360399717111968e-07, "loss": 0.1964, "step": 6268 }, { "epoch": 1.6681745609366685, "grad_norm": 0.29127347469329834, "learning_rate": 1.535895937122453e-07, "loss": 0.2094, "step": 6269 }, { "epoch": 1.6684406599254924, "grad_norm": 0.3039625287055969, "learning_rate": 1.5357518869351695e-07, "loss": 0.2164, "step": 6270 }, { "epoch": 1.668706758914316, "grad_norm": 0.29083916544914246, "learning_rate": 1.5356078211535401e-07, "loss": 0.2093, "step": 6271 }, { "epoch": 1.66897285790314, "grad_norm": 0.30392026901245117, "learning_rate": 1.5354637397817577e-07, "loss": 0.2268, "step": 6272 }, { "epoch": 1.6692389568919639, "grad_norm": 0.28041911125183105, "learning_rate": 1.5353196428240156e-07, "loss": 0.193, "step": 6273 }, { "epoch": 1.6695050558807876, "grad_norm": 0.3128564953804016, "learning_rate": 1.535175530284509e-07, "loss": 0.2121, "step": 6274 }, { "epoch": 1.6697711548696115, "grad_norm": 0.27865085005760193, "learning_rate": 1.535031402167432e-07, "loss": 0.1841, "step": 6275 }, { "epoch": 1.6700372538584354, "grad_norm": 0.2856203019618988, "learning_rate": 1.5348872584769804e-07, "loss": 0.2128, "step": 6276 }, { "epoch": 1.670303352847259, "grad_norm": 0.2905958294868469, "learning_rate": 1.5347430992173498e-07, "loss": 0.2079, "step": 6277 }, { "epoch": 1.6705694518360832, "grad_norm": 0.27746379375457764, "learning_rate": 1.5345989243927354e-07, "loss": 0.1969, "step": 6278 }, { "epoch": 1.6708355508249069, "grad_norm": 0.2955522835254669, "learning_rate": 1.534454734007335e-07, "loss": 0.2215, "step": 6279 }, { "epoch": 1.6711016498137305, "grad_norm": 0.27491533756256104, "learning_rate": 1.5343105280653445e-07, "loss": 0.1939, "step": 6280 }, { "epoch": 1.6713677488025547, "grad_norm": 0.2954773008823395, "learning_rate": 1.534166306570962e-07, "loss": 0.2177, "step": 6281 }, { "epoch": 1.6716338477913784, "grad_norm": 0.30321037769317627, "learning_rate": 1.5340220695283855e-07, "loss": 0.2039, "step": 6282 }, { "epoch": 1.6718999467802023, "grad_norm": 0.2886231541633606, "learning_rate": 1.5338778169418127e-07, "loss": 0.2028, "step": 6283 }, { "epoch": 1.6721660457690262, "grad_norm": 0.264677494764328, "learning_rate": 1.533733548815443e-07, "loss": 0.2001, "step": 6284 }, { "epoch": 1.6724321447578498, "grad_norm": 0.27735888957977295, "learning_rate": 1.5335892651534756e-07, "loss": 0.1994, "step": 6285 }, { "epoch": 1.6726982437466738, "grad_norm": 0.2900986075401306, "learning_rate": 1.53344496596011e-07, "loss": 0.2089, "step": 6286 }, { "epoch": 1.6729643427354977, "grad_norm": 0.2580729126930237, "learning_rate": 1.5333006512395468e-07, "loss": 0.1899, "step": 6287 }, { "epoch": 1.6732304417243213, "grad_norm": 0.2733091413974762, "learning_rate": 1.5331563209959862e-07, "loss": 0.1973, "step": 6288 }, { "epoch": 1.6734965407131452, "grad_norm": 0.2923431992530823, "learning_rate": 1.5330119752336295e-07, "loss": 0.1965, "step": 6289 }, { "epoch": 1.6737626397019691, "grad_norm": 0.2504792809486389, "learning_rate": 1.5328676139566777e-07, "loss": 0.1793, "step": 6290 }, { "epoch": 1.6740287386907928, "grad_norm": 0.28888604044914246, "learning_rate": 1.5327232371693338e-07, "loss": 0.2164, "step": 6291 }, { "epoch": 1.674294837679617, "grad_norm": 0.2794518768787384, "learning_rate": 1.5325788448757994e-07, "loss": 0.186, "step": 6292 }, { "epoch": 1.6745609366684406, "grad_norm": 0.291931688785553, "learning_rate": 1.532434437080278e-07, "loss": 0.2256, "step": 6293 }, { "epoch": 1.6748270356572645, "grad_norm": 0.258868545293808, "learning_rate": 1.532290013786972e-07, "loss": 0.1893, "step": 6294 }, { "epoch": 1.6750931346460884, "grad_norm": 0.33123543858528137, "learning_rate": 1.532145575000086e-07, "loss": 0.214, "step": 6295 }, { "epoch": 1.6753592336349121, "grad_norm": 0.2791343331336975, "learning_rate": 1.532001120723824e-07, "loss": 0.2001, "step": 6296 }, { "epoch": 1.675625332623736, "grad_norm": 0.28030019998550415, "learning_rate": 1.531856650962391e-07, "loss": 0.1971, "step": 6297 }, { "epoch": 1.67589143161256, "grad_norm": 0.2898203730583191, "learning_rate": 1.5317121657199918e-07, "loss": 0.2089, "step": 6298 }, { "epoch": 1.6761575306013836, "grad_norm": 0.29523810744285583, "learning_rate": 1.5315676650008317e-07, "loss": 0.196, "step": 6299 }, { "epoch": 1.6764236295902075, "grad_norm": 0.30040180683135986, "learning_rate": 1.531423148809117e-07, "loss": 0.2052, "step": 6300 }, { "epoch": 1.6766897285790314, "grad_norm": 0.30420467257499695, "learning_rate": 1.531278617149055e-07, "loss": 0.2059, "step": 6301 }, { "epoch": 1.676955827567855, "grad_norm": 0.2910194993019104, "learning_rate": 1.5311340700248514e-07, "loss": 0.1924, "step": 6302 }, { "epoch": 1.6772219265566792, "grad_norm": 0.29674726724624634, "learning_rate": 1.5309895074407137e-07, "loss": 0.2002, "step": 6303 }, { "epoch": 1.677488025545503, "grad_norm": 0.28211772441864014, "learning_rate": 1.5308449294008506e-07, "loss": 0.198, "step": 6304 }, { "epoch": 1.6777541245343268, "grad_norm": 0.2754468619823456, "learning_rate": 1.5307003359094698e-07, "loss": 0.201, "step": 6305 }, { "epoch": 1.6780202235231507, "grad_norm": 0.2967934012413025, "learning_rate": 1.5305557269707805e-07, "loss": 0.2032, "step": 6306 }, { "epoch": 1.6782863225119744, "grad_norm": 0.26321887969970703, "learning_rate": 1.5304111025889912e-07, "loss": 0.2018, "step": 6307 }, { "epoch": 1.6785524215007983, "grad_norm": 0.29269078373908997, "learning_rate": 1.5302664627683122e-07, "loss": 0.2122, "step": 6308 }, { "epoch": 1.6788185204896222, "grad_norm": 0.2804378271102905, "learning_rate": 1.5301218075129529e-07, "loss": 0.1996, "step": 6309 }, { "epoch": 1.679084619478446, "grad_norm": 0.2866182029247284, "learning_rate": 1.5299771368271244e-07, "loss": 0.1798, "step": 6310 }, { "epoch": 1.6793507184672698, "grad_norm": 0.27627086639404297, "learning_rate": 1.5298324507150377e-07, "loss": 0.1745, "step": 6311 }, { "epoch": 1.6796168174560937, "grad_norm": 0.2648032009601593, "learning_rate": 1.5296877491809038e-07, "loss": 0.1958, "step": 6312 }, { "epoch": 1.6798829164449174, "grad_norm": 0.27921751141548157, "learning_rate": 1.5295430322289349e-07, "loss": 0.1943, "step": 6313 }, { "epoch": 1.6801490154337415, "grad_norm": 0.29279598593711853, "learning_rate": 1.5293982998633435e-07, "loss": 0.2133, "step": 6314 }, { "epoch": 1.6804151144225652, "grad_norm": 0.2674993872642517, "learning_rate": 1.529253552088342e-07, "loss": 0.1996, "step": 6315 }, { "epoch": 1.6806812134113889, "grad_norm": 0.2927189767360687, "learning_rate": 1.5291087889081435e-07, "loss": 0.2187, "step": 6316 }, { "epoch": 1.680947312400213, "grad_norm": 0.28567224740982056, "learning_rate": 1.5289640103269623e-07, "loss": 0.2085, "step": 6317 }, { "epoch": 1.6812134113890367, "grad_norm": 0.2621636986732483, "learning_rate": 1.5288192163490123e-07, "loss": 0.1757, "step": 6318 }, { "epoch": 1.6814795103778606, "grad_norm": 0.28246021270751953, "learning_rate": 1.5286744069785077e-07, "loss": 0.1936, "step": 6319 }, { "epoch": 1.6817456093666845, "grad_norm": 0.29139244556427, "learning_rate": 1.5285295822196642e-07, "loss": 0.2135, "step": 6320 }, { "epoch": 1.6820117083555082, "grad_norm": 0.2747904658317566, "learning_rate": 1.5283847420766964e-07, "loss": 0.2094, "step": 6321 }, { "epoch": 1.682277807344332, "grad_norm": 0.28192904591560364, "learning_rate": 1.528239886553821e-07, "loss": 0.2112, "step": 6322 }, { "epoch": 1.682543906333156, "grad_norm": 0.2625228762626648, "learning_rate": 1.5280950156552538e-07, "loss": 0.1842, "step": 6323 }, { "epoch": 1.6828100053219797, "grad_norm": 0.2933684289455414, "learning_rate": 1.527950129385212e-07, "loss": 0.2108, "step": 6324 }, { "epoch": 1.6830761043108036, "grad_norm": 0.29076337814331055, "learning_rate": 1.527805227747913e-07, "loss": 0.1925, "step": 6325 }, { "epoch": 1.6833422032996275, "grad_norm": 0.2728210985660553, "learning_rate": 1.527660310747574e-07, "loss": 0.1933, "step": 6326 }, { "epoch": 1.6836083022884512, "grad_norm": 0.2620333731174469, "learning_rate": 1.5275153783884135e-07, "loss": 0.1943, "step": 6327 }, { "epoch": 1.6838744012772753, "grad_norm": 0.2864636480808258, "learning_rate": 1.52737043067465e-07, "loss": 0.203, "step": 6328 }, { "epoch": 1.684140500266099, "grad_norm": 0.27415549755096436, "learning_rate": 1.5272254676105024e-07, "loss": 0.1943, "step": 6329 }, { "epoch": 1.6844065992549229, "grad_norm": 0.28306499123573303, "learning_rate": 1.5270804892001909e-07, "loss": 0.2004, "step": 6330 }, { "epoch": 1.6846726982437468, "grad_norm": 0.330413818359375, "learning_rate": 1.5269354954479342e-07, "loss": 0.229, "step": 6331 }, { "epoch": 1.6849387972325705, "grad_norm": 0.27392423152923584, "learning_rate": 1.5267904863579535e-07, "loss": 0.1966, "step": 6332 }, { "epoch": 1.6852048962213944, "grad_norm": 0.3098813593387604, "learning_rate": 1.5266454619344696e-07, "loss": 0.2131, "step": 6333 }, { "epoch": 1.6854709952102183, "grad_norm": 0.27575457096099854, "learning_rate": 1.526500422181704e-07, "loss": 0.1945, "step": 6334 }, { "epoch": 1.685737094199042, "grad_norm": 0.2579006850719452, "learning_rate": 1.526355367103878e-07, "loss": 0.1847, "step": 6335 }, { "epoch": 1.6860031931878658, "grad_norm": 0.29753565788269043, "learning_rate": 1.526210296705214e-07, "loss": 0.2117, "step": 6336 }, { "epoch": 1.6862692921766897, "grad_norm": 0.29398006200790405, "learning_rate": 1.5260652109899348e-07, "loss": 0.2046, "step": 6337 }, { "epoch": 1.6865353911655134, "grad_norm": 0.25368639826774597, "learning_rate": 1.5259201099622626e-07, "loss": 0.1883, "step": 6338 }, { "epoch": 1.6868014901543376, "grad_norm": 0.297648161649704, "learning_rate": 1.5257749936264221e-07, "loss": 0.2032, "step": 6339 }, { "epoch": 1.6870675891431612, "grad_norm": 0.2742213308811188, "learning_rate": 1.5256298619866365e-07, "loss": 0.1793, "step": 6340 }, { "epoch": 1.6873336881319851, "grad_norm": 0.31481432914733887, "learning_rate": 1.5254847150471304e-07, "loss": 0.2142, "step": 6341 }, { "epoch": 1.687599787120809, "grad_norm": 0.27068665623664856, "learning_rate": 1.5253395528121282e-07, "loss": 0.2, "step": 6342 }, { "epoch": 1.6878658861096327, "grad_norm": 0.28715258836746216, "learning_rate": 1.5251943752858566e-07, "loss": 0.2107, "step": 6343 }, { "epoch": 1.6881319850984566, "grad_norm": 0.2802050709724426, "learning_rate": 1.5250491824725397e-07, "loss": 0.1971, "step": 6344 }, { "epoch": 1.6883980840872805, "grad_norm": 0.26574385166168213, "learning_rate": 1.5249039743764045e-07, "loss": 0.2005, "step": 6345 }, { "epoch": 1.6886641830761042, "grad_norm": 0.26184096932411194, "learning_rate": 1.5247587510016773e-07, "loss": 0.1954, "step": 6346 }, { "epoch": 1.6889302820649281, "grad_norm": 0.2891428470611572, "learning_rate": 1.5246135123525856e-07, "loss": 0.2283, "step": 6347 }, { "epoch": 1.689196381053752, "grad_norm": 0.2964155972003937, "learning_rate": 1.524468258433357e-07, "loss": 0.2234, "step": 6348 }, { "epoch": 1.6894624800425757, "grad_norm": 0.28658342361450195, "learning_rate": 1.524322989248219e-07, "loss": 0.2006, "step": 6349 }, { "epoch": 1.6897285790313998, "grad_norm": 0.2972939908504486, "learning_rate": 1.5241777048013999e-07, "loss": 0.2008, "step": 6350 }, { "epoch": 1.6899946780202235, "grad_norm": 0.3095504641532898, "learning_rate": 1.5240324050971286e-07, "loss": 0.2191, "step": 6351 }, { "epoch": 1.6902607770090472, "grad_norm": 0.30128657817840576, "learning_rate": 1.5238870901396356e-07, "loss": 0.1943, "step": 6352 }, { "epoch": 1.6905268759978713, "grad_norm": 0.3077504336833954, "learning_rate": 1.5237417599331489e-07, "loss": 0.2179, "step": 6353 }, { "epoch": 1.690792974986695, "grad_norm": 0.2958630323410034, "learning_rate": 1.5235964144818993e-07, "loss": 0.1998, "step": 6354 }, { "epoch": 1.691059073975519, "grad_norm": 0.35477113723754883, "learning_rate": 1.5234510537901182e-07, "loss": 0.2188, "step": 6355 }, { "epoch": 1.6913251729643428, "grad_norm": 0.27585890889167786, "learning_rate": 1.5233056778620355e-07, "loss": 0.1822, "step": 6356 }, { "epoch": 1.6915912719531665, "grad_norm": 0.27012255787849426, "learning_rate": 1.5231602867018837e-07, "loss": 0.196, "step": 6357 }, { "epoch": 1.6918573709419904, "grad_norm": 0.27907419204711914, "learning_rate": 1.5230148803138943e-07, "loss": 0.221, "step": 6358 }, { "epoch": 1.6921234699308143, "grad_norm": 0.3251585364341736, "learning_rate": 1.5228694587022996e-07, "loss": 0.1931, "step": 6359 }, { "epoch": 1.692389568919638, "grad_norm": 0.2669861614704132, "learning_rate": 1.5227240218713325e-07, "loss": 0.1811, "step": 6360 }, { "epoch": 1.6926556679084621, "grad_norm": 0.2916394770145416, "learning_rate": 1.5225785698252264e-07, "loss": 0.2107, "step": 6361 }, { "epoch": 1.6929217668972858, "grad_norm": 0.28829970955848694, "learning_rate": 1.5224331025682148e-07, "loss": 0.1931, "step": 6362 }, { "epoch": 1.6931878658861095, "grad_norm": 0.27668628096580505, "learning_rate": 1.5222876201045325e-07, "loss": 0.1892, "step": 6363 }, { "epoch": 1.6934539648749336, "grad_norm": 0.3005102276802063, "learning_rate": 1.5221421224384132e-07, "loss": 0.2056, "step": 6364 }, { "epoch": 1.6937200638637573, "grad_norm": 0.27955392003059387, "learning_rate": 1.5219966095740927e-07, "loss": 0.2079, "step": 6365 }, { "epoch": 1.6939861628525812, "grad_norm": 0.285739541053772, "learning_rate": 1.5218510815158063e-07, "loss": 0.2046, "step": 6366 }, { "epoch": 1.694252261841405, "grad_norm": 0.2818797528743744, "learning_rate": 1.5217055382677897e-07, "loss": 0.1882, "step": 6367 }, { "epoch": 1.6945183608302288, "grad_norm": 0.28640538454055786, "learning_rate": 1.5215599798342798e-07, "loss": 0.2113, "step": 6368 }, { "epoch": 1.6947844598190527, "grad_norm": 0.29921847581863403, "learning_rate": 1.521414406219513e-07, "loss": 0.2224, "step": 6369 }, { "epoch": 1.6950505588078766, "grad_norm": 0.2865184545516968, "learning_rate": 1.5212688174277266e-07, "loss": 0.2038, "step": 6370 }, { "epoch": 1.6953166577967003, "grad_norm": 0.2906835079193115, "learning_rate": 1.5211232134631583e-07, "loss": 0.2055, "step": 6371 }, { "epoch": 1.6955827567855242, "grad_norm": 0.2979803681373596, "learning_rate": 1.5209775943300463e-07, "loss": 0.2094, "step": 6372 }, { "epoch": 1.695848855774348, "grad_norm": 0.29876190423965454, "learning_rate": 1.5208319600326295e-07, "loss": 0.232, "step": 6373 }, { "epoch": 1.6961149547631718, "grad_norm": 0.2795311212539673, "learning_rate": 1.5206863105751464e-07, "loss": 0.2079, "step": 6374 }, { "epoch": 1.6963810537519959, "grad_norm": 0.2723202407360077, "learning_rate": 1.5205406459618368e-07, "loss": 0.1925, "step": 6375 }, { "epoch": 1.6966471527408196, "grad_norm": 0.27598267793655396, "learning_rate": 1.5203949661969407e-07, "loss": 0.1922, "step": 6376 }, { "epoch": 1.6969132517296435, "grad_norm": 0.31795117259025574, "learning_rate": 1.5202492712846984e-07, "loss": 0.2253, "step": 6377 }, { "epoch": 1.6971793507184674, "grad_norm": 0.30316275358200073, "learning_rate": 1.5201035612293504e-07, "loss": 0.2166, "step": 6378 }, { "epoch": 1.697445449707291, "grad_norm": 0.2670302391052246, "learning_rate": 1.5199578360351387e-07, "loss": 0.2057, "step": 6379 }, { "epoch": 1.697711548696115, "grad_norm": 0.2666575312614441, "learning_rate": 1.5198120957063035e-07, "loss": 0.1927, "step": 6380 }, { "epoch": 1.6979776476849389, "grad_norm": 0.302743524312973, "learning_rate": 1.5196663402470887e-07, "loss": 0.2301, "step": 6381 }, { "epoch": 1.6982437466737625, "grad_norm": 0.26904308795928955, "learning_rate": 1.5195205696617356e-07, "loss": 0.1854, "step": 6382 }, { "epoch": 1.6985098456625864, "grad_norm": 0.33362382650375366, "learning_rate": 1.5193747839544875e-07, "loss": 0.2283, "step": 6383 }, { "epoch": 1.6987759446514104, "grad_norm": 0.2948561906814575, "learning_rate": 1.5192289831295883e-07, "loss": 0.1921, "step": 6384 }, { "epoch": 1.699042043640234, "grad_norm": 0.2947750389575958, "learning_rate": 1.5190831671912816e-07, "loss": 0.2038, "step": 6385 }, { "epoch": 1.6993081426290582, "grad_norm": 0.2883302867412567, "learning_rate": 1.5189373361438117e-07, "loss": 0.2061, "step": 6386 }, { "epoch": 1.6995742416178818, "grad_norm": 0.26934999227523804, "learning_rate": 1.5187914899914232e-07, "loss": 0.1827, "step": 6387 }, { "epoch": 1.6998403406067057, "grad_norm": 0.3028572201728821, "learning_rate": 1.5186456287383616e-07, "loss": 0.2144, "step": 6388 }, { "epoch": 1.7001064395955297, "grad_norm": 0.2962227463722229, "learning_rate": 1.5184997523888723e-07, "loss": 0.1875, "step": 6389 }, { "epoch": 1.7003725385843533, "grad_norm": 0.26871150732040405, "learning_rate": 1.5183538609472013e-07, "loss": 0.2046, "step": 6390 }, { "epoch": 1.7006386375731772, "grad_norm": 0.2780589461326599, "learning_rate": 1.5182079544175954e-07, "loss": 0.1876, "step": 6391 }, { "epoch": 1.7009047365620011, "grad_norm": 0.27257344126701355, "learning_rate": 1.5180620328043013e-07, "loss": 0.2025, "step": 6392 }, { "epoch": 1.7011708355508248, "grad_norm": 0.28285887837409973, "learning_rate": 1.5179160961115672e-07, "loss": 0.2071, "step": 6393 }, { "epoch": 1.7014369345396487, "grad_norm": 0.26050427556037903, "learning_rate": 1.5177701443436397e-07, "loss": 0.1844, "step": 6394 }, { "epoch": 1.7017030335284726, "grad_norm": 0.26120880246162415, "learning_rate": 1.5176241775047679e-07, "loss": 0.1741, "step": 6395 }, { "epoch": 1.7019691325172963, "grad_norm": 0.279204785823822, "learning_rate": 1.5174781955992e-07, "loss": 0.1899, "step": 6396 }, { "epoch": 1.7022352315061204, "grad_norm": 0.2886045277118683, "learning_rate": 1.517332198631186e-07, "loss": 0.2175, "step": 6397 }, { "epoch": 1.7025013304949441, "grad_norm": 0.30280470848083496, "learning_rate": 1.5171861866049747e-07, "loss": 0.2049, "step": 6398 }, { "epoch": 1.7027674294837678, "grad_norm": 0.2876613736152649, "learning_rate": 1.5170401595248164e-07, "loss": 0.2026, "step": 6399 }, { "epoch": 1.703033528472592, "grad_norm": 0.29313695430755615, "learning_rate": 1.5168941173949614e-07, "loss": 0.2107, "step": 6400 }, { "epoch": 1.7032996274614156, "grad_norm": 0.28531092405319214, "learning_rate": 1.5167480602196605e-07, "loss": 0.1844, "step": 6401 }, { "epoch": 1.7035657264502395, "grad_norm": 0.2769777774810791, "learning_rate": 1.516601988003166e-07, "loss": 0.1892, "step": 6402 }, { "epoch": 1.7038318254390634, "grad_norm": 0.2883667051792145, "learning_rate": 1.5164559007497282e-07, "loss": 0.2065, "step": 6403 }, { "epoch": 1.704097924427887, "grad_norm": 0.272419273853302, "learning_rate": 1.5163097984636005e-07, "loss": 0.2073, "step": 6404 }, { "epoch": 1.704364023416711, "grad_norm": 0.28610846400260925, "learning_rate": 1.516163681149035e-07, "loss": 0.2078, "step": 6405 }, { "epoch": 1.704630122405535, "grad_norm": 0.3009265065193176, "learning_rate": 1.5160175488102853e-07, "loss": 0.2247, "step": 6406 }, { "epoch": 1.7048962213943586, "grad_norm": 0.274374783039093, "learning_rate": 1.515871401451604e-07, "loss": 0.2027, "step": 6407 }, { "epoch": 1.7051623203831825, "grad_norm": 0.3079632818698883, "learning_rate": 1.5157252390772463e-07, "loss": 0.2275, "step": 6408 }, { "epoch": 1.7054284193720064, "grad_norm": 0.3160359263420105, "learning_rate": 1.5155790616914654e-07, "loss": 0.2063, "step": 6409 }, { "epoch": 1.70569451836083, "grad_norm": 0.26735514402389526, "learning_rate": 1.515432869298517e-07, "loss": 0.2015, "step": 6410 }, { "epoch": 1.7059606173496542, "grad_norm": 0.29067811369895935, "learning_rate": 1.5152866619026564e-07, "loss": 0.1938, "step": 6411 }, { "epoch": 1.7062267163384779, "grad_norm": 0.28360089659690857, "learning_rate": 1.5151404395081385e-07, "loss": 0.2001, "step": 6412 }, { "epoch": 1.7064928153273018, "grad_norm": 0.26445090770721436, "learning_rate": 1.5149942021192197e-07, "loss": 0.1931, "step": 6413 }, { "epoch": 1.7067589143161257, "grad_norm": 0.27684950828552246, "learning_rate": 1.514847949740157e-07, "loss": 0.1923, "step": 6414 }, { "epoch": 1.7070250133049494, "grad_norm": 0.31518977880477905, "learning_rate": 1.514701682375208e-07, "loss": 0.2456, "step": 6415 }, { "epoch": 1.7072911122937733, "grad_norm": 0.2862834334373474, "learning_rate": 1.514555400028629e-07, "loss": 0.1887, "step": 6416 }, { "epoch": 1.7075572112825972, "grad_norm": 0.28902024030685425, "learning_rate": 1.514409102704678e-07, "loss": 0.2131, "step": 6417 }, { "epoch": 1.7078233102714209, "grad_norm": 0.33900952339172363, "learning_rate": 1.514262790407614e-07, "loss": 0.2315, "step": 6418 }, { "epoch": 1.7080894092602448, "grad_norm": 0.2834320068359375, "learning_rate": 1.5141164631416952e-07, "loss": 0.1929, "step": 6419 }, { "epoch": 1.7083555082490687, "grad_norm": 0.25913211703300476, "learning_rate": 1.5139701209111815e-07, "loss": 0.1757, "step": 6420 }, { "epoch": 1.7086216072378924, "grad_norm": 0.3632933795452118, "learning_rate": 1.5138237637203316e-07, "loss": 0.1973, "step": 6421 }, { "epoch": 1.7088877062267165, "grad_norm": 0.2878328263759613, "learning_rate": 1.5136773915734065e-07, "loss": 0.2035, "step": 6422 }, { "epoch": 1.7091538052155402, "grad_norm": 0.2917737066745758, "learning_rate": 1.5135310044746658e-07, "loss": 0.2105, "step": 6423 }, { "epoch": 1.709419904204364, "grad_norm": 0.30162450671195984, "learning_rate": 1.5133846024283715e-07, "loss": 0.1994, "step": 6424 }, { "epoch": 1.709686003193188, "grad_norm": 0.2961471378803253, "learning_rate": 1.513238185438784e-07, "loss": 0.2152, "step": 6425 }, { "epoch": 1.7099521021820117, "grad_norm": 0.2788825035095215, "learning_rate": 1.5130917535101654e-07, "loss": 0.1939, "step": 6426 }, { "epoch": 1.7102182011708356, "grad_norm": 0.2962992489337921, "learning_rate": 1.5129453066467784e-07, "loss": 0.215, "step": 6427 }, { "epoch": 1.7104843001596595, "grad_norm": 0.28464341163635254, "learning_rate": 1.5127988448528854e-07, "loss": 0.191, "step": 6428 }, { "epoch": 1.7107503991484831, "grad_norm": 0.2783406376838684, "learning_rate": 1.5126523681327493e-07, "loss": 0.2197, "step": 6429 }, { "epoch": 1.711016498137307, "grad_norm": 0.27365174889564514, "learning_rate": 1.512505876490634e-07, "loss": 0.2082, "step": 6430 }, { "epoch": 1.711282597126131, "grad_norm": 0.2775052785873413, "learning_rate": 1.5123593699308032e-07, "loss": 0.2062, "step": 6431 }, { "epoch": 1.7115486961149546, "grad_norm": 0.29408133029937744, "learning_rate": 1.5122128484575213e-07, "loss": 0.1968, "step": 6432 }, { "epoch": 1.7118147951037788, "grad_norm": 0.2890850603580475, "learning_rate": 1.5120663120750541e-07, "loss": 0.1934, "step": 6433 }, { "epoch": 1.7120808940926024, "grad_norm": 0.29496949911117554, "learning_rate": 1.5119197607876656e-07, "loss": 0.1947, "step": 6434 }, { "epoch": 1.7123469930814261, "grad_norm": 0.3027768135070801, "learning_rate": 1.5117731945996218e-07, "loss": 0.2148, "step": 6435 }, { "epoch": 1.7126130920702503, "grad_norm": 0.26463061571121216, "learning_rate": 1.5116266135151894e-07, "loss": 0.2003, "step": 6436 }, { "epoch": 1.712879191059074, "grad_norm": 0.2927600145339966, "learning_rate": 1.5114800175386345e-07, "loss": 0.1968, "step": 6437 }, { "epoch": 1.7131452900478978, "grad_norm": 0.2770858407020569, "learning_rate": 1.5113334066742244e-07, "loss": 0.1934, "step": 6438 }, { "epoch": 1.7134113890367217, "grad_norm": 0.2841123640537262, "learning_rate": 1.5111867809262267e-07, "loss": 0.1855, "step": 6439 }, { "epoch": 1.7136774880255454, "grad_norm": 0.2737947702407837, "learning_rate": 1.5110401402989092e-07, "loss": 0.2042, "step": 6440 }, { "epoch": 1.7139435870143693, "grad_norm": 0.2814136743545532, "learning_rate": 1.5108934847965398e-07, "loss": 0.2172, "step": 6441 }, { "epoch": 1.7142096860031932, "grad_norm": 0.2814842760562897, "learning_rate": 1.510746814423388e-07, "loss": 0.2016, "step": 6442 }, { "epoch": 1.714475784992017, "grad_norm": 0.28658539056777954, "learning_rate": 1.510600129183722e-07, "loss": 0.2061, "step": 6443 }, { "epoch": 1.7147418839808408, "grad_norm": 0.27794769406318665, "learning_rate": 1.5104534290818122e-07, "loss": 0.1982, "step": 6444 }, { "epoch": 1.7150079829696647, "grad_norm": 0.32097044587135315, "learning_rate": 1.5103067141219286e-07, "loss": 0.201, "step": 6445 }, { "epoch": 1.7152740819584884, "grad_norm": 0.32027938961982727, "learning_rate": 1.5101599843083412e-07, "loss": 0.2124, "step": 6446 }, { "epoch": 1.7155401809473125, "grad_norm": 0.29344460368156433, "learning_rate": 1.5100132396453216e-07, "loss": 0.2169, "step": 6447 }, { "epoch": 1.7158062799361362, "grad_norm": 0.2796841859817505, "learning_rate": 1.5098664801371407e-07, "loss": 0.2046, "step": 6448 }, { "epoch": 1.7160723789249601, "grad_norm": 0.2780798375606537, "learning_rate": 1.5097197057880705e-07, "loss": 0.2121, "step": 6449 }, { "epoch": 1.716338477913784, "grad_norm": 0.28106003999710083, "learning_rate": 1.5095729166023832e-07, "loss": 0.205, "step": 6450 }, { "epoch": 1.7166045769026077, "grad_norm": 0.3582358658313751, "learning_rate": 1.5094261125843515e-07, "loss": 0.2154, "step": 6451 }, { "epoch": 1.7168706758914316, "grad_norm": 0.2872956693172455, "learning_rate": 1.509279293738248e-07, "loss": 0.2136, "step": 6452 }, { "epoch": 1.7171367748802555, "grad_norm": 0.28542360663414, "learning_rate": 1.509132460068347e-07, "loss": 0.2079, "step": 6453 }, { "epoch": 1.7174028738690792, "grad_norm": 0.30202803015708923, "learning_rate": 1.5089856115789216e-07, "loss": 0.211, "step": 6454 }, { "epoch": 1.717668972857903, "grad_norm": 0.30814534425735474, "learning_rate": 1.5088387482742467e-07, "loss": 0.224, "step": 6455 }, { "epoch": 1.717935071846727, "grad_norm": 0.32314303517341614, "learning_rate": 1.5086918701585973e-07, "loss": 0.2387, "step": 6456 }, { "epoch": 1.7182011708355507, "grad_norm": 0.3138416111469269, "learning_rate": 1.5085449772362484e-07, "loss": 0.2148, "step": 6457 }, { "epoch": 1.7184672698243748, "grad_norm": 0.27978137135505676, "learning_rate": 1.5083980695114754e-07, "loss": 0.1879, "step": 6458 }, { "epoch": 1.7187333688131985, "grad_norm": 0.27477723360061646, "learning_rate": 1.5082511469885546e-07, "loss": 0.1939, "step": 6459 }, { "epoch": 1.7189994678020224, "grad_norm": 0.27073416113853455, "learning_rate": 1.5081042096717626e-07, "loss": 0.2056, "step": 6460 }, { "epoch": 1.7192655667908463, "grad_norm": 0.2830875813961029, "learning_rate": 1.5079572575653767e-07, "loss": 0.1977, "step": 6461 }, { "epoch": 1.71953166577967, "grad_norm": 0.30301326513290405, "learning_rate": 1.5078102906736737e-07, "loss": 0.2116, "step": 6462 }, { "epoch": 1.7197977647684939, "grad_norm": 0.30270296335220337, "learning_rate": 1.5076633090009315e-07, "loss": 0.2279, "step": 6463 }, { "epoch": 1.7200638637573178, "grad_norm": 0.3092145621776581, "learning_rate": 1.5075163125514282e-07, "loss": 0.1991, "step": 6464 }, { "epoch": 1.7203299627461415, "grad_norm": 0.298095166683197, "learning_rate": 1.5073693013294433e-07, "loss": 0.2076, "step": 6465 }, { "epoch": 1.7205960617349654, "grad_norm": 0.3189013600349426, "learning_rate": 1.5072222753392554e-07, "loss": 0.2108, "step": 6466 }, { "epoch": 1.7208621607237893, "grad_norm": 0.3010607063770294, "learning_rate": 1.5070752345851438e-07, "loss": 0.2084, "step": 6467 }, { "epoch": 1.721128259712613, "grad_norm": 0.3103037476539612, "learning_rate": 1.506928179071389e-07, "loss": 0.2329, "step": 6468 }, { "epoch": 1.721394358701437, "grad_norm": 0.2842949628829956, "learning_rate": 1.506781108802271e-07, "loss": 0.1983, "step": 6469 }, { "epoch": 1.7216604576902608, "grad_norm": 0.3077099025249481, "learning_rate": 1.506634023782071e-07, "loss": 0.2088, "step": 6470 }, { "epoch": 1.7219265566790847, "grad_norm": 0.29086270928382874, "learning_rate": 1.5064869240150697e-07, "loss": 0.209, "step": 6471 }, { "epoch": 1.7221926556679086, "grad_norm": 0.28219643235206604, "learning_rate": 1.5063398095055494e-07, "loss": 0.2084, "step": 6472 }, { "epoch": 1.7224587546567323, "grad_norm": 0.25992444157600403, "learning_rate": 1.5061926802577916e-07, "loss": 0.1798, "step": 6473 }, { "epoch": 1.7227248536455562, "grad_norm": 0.28146010637283325, "learning_rate": 1.5060455362760793e-07, "loss": 0.1912, "step": 6474 }, { "epoch": 1.72299095263438, "grad_norm": 0.2832871377468109, "learning_rate": 1.5058983775646955e-07, "loss": 0.2086, "step": 6475 }, { "epoch": 1.7232570516232038, "grad_norm": 0.34989410638809204, "learning_rate": 1.5057512041279234e-07, "loss": 0.2193, "step": 6476 }, { "epoch": 1.7235231506120277, "grad_norm": 0.29800906777381897, "learning_rate": 1.505604015970047e-07, "loss": 0.205, "step": 6477 }, { "epoch": 1.7237892496008516, "grad_norm": 0.26921412348747253, "learning_rate": 1.5054568130953502e-07, "loss": 0.2038, "step": 6478 }, { "epoch": 1.7240553485896752, "grad_norm": 0.27608299255371094, "learning_rate": 1.5053095955081184e-07, "loss": 0.2018, "step": 6479 }, { "epoch": 1.7243214475784994, "grad_norm": 0.2817678451538086, "learning_rate": 1.5051623632126361e-07, "loss": 0.2241, "step": 6480 }, { "epoch": 1.724587546567323, "grad_norm": 0.2941746115684509, "learning_rate": 1.505015116213189e-07, "loss": 0.1912, "step": 6481 }, { "epoch": 1.7248536455561467, "grad_norm": 0.2881355881690979, "learning_rate": 1.504867854514063e-07, "loss": 0.211, "step": 6482 }, { "epoch": 1.7251197445449709, "grad_norm": 0.35189181566238403, "learning_rate": 1.504720578119545e-07, "loss": 0.2146, "step": 6483 }, { "epoch": 1.7253858435337945, "grad_norm": 0.2880881428718567, "learning_rate": 1.5045732870339213e-07, "loss": 0.2161, "step": 6484 }, { "epoch": 1.7256519425226184, "grad_norm": 0.3046955466270447, "learning_rate": 1.5044259812614792e-07, "loss": 0.1941, "step": 6485 }, { "epoch": 1.7259180415114423, "grad_norm": 0.2892932593822479, "learning_rate": 1.5042786608065066e-07, "loss": 0.216, "step": 6486 }, { "epoch": 1.726184140500266, "grad_norm": 0.299685001373291, "learning_rate": 1.5041313256732916e-07, "loss": 0.1994, "step": 6487 }, { "epoch": 1.72645023948909, "grad_norm": 0.28676512837409973, "learning_rate": 1.5039839758661225e-07, "loss": 0.2004, "step": 6488 }, { "epoch": 1.7267163384779138, "grad_norm": 0.25693538784980774, "learning_rate": 1.5038366113892885e-07, "loss": 0.2156, "step": 6489 }, { "epoch": 1.7269824374667375, "grad_norm": 0.2995724678039551, "learning_rate": 1.5036892322470794e-07, "loss": 0.2194, "step": 6490 }, { "epoch": 1.7272485364555614, "grad_norm": 0.2879326641559601, "learning_rate": 1.503541838443784e-07, "loss": 0.1887, "step": 6491 }, { "epoch": 1.7275146354443853, "grad_norm": 0.26492777466773987, "learning_rate": 1.5033944299836933e-07, "loss": 0.1934, "step": 6492 }, { "epoch": 1.727780734433209, "grad_norm": 0.28539201617240906, "learning_rate": 1.5032470068710978e-07, "loss": 0.198, "step": 6493 }, { "epoch": 1.7280468334220331, "grad_norm": 0.31883543729782104, "learning_rate": 1.5030995691102883e-07, "loss": 0.2249, "step": 6494 }, { "epoch": 1.7283129324108568, "grad_norm": 0.2912880778312683, "learning_rate": 1.502952116705557e-07, "loss": 0.1958, "step": 6495 }, { "epoch": 1.7285790313996807, "grad_norm": 0.33516019582748413, "learning_rate": 1.5028046496611951e-07, "loss": 0.2291, "step": 6496 }, { "epoch": 1.7288451303885046, "grad_norm": 0.30942144989967346, "learning_rate": 1.502657167981496e-07, "loss": 0.2315, "step": 6497 }, { "epoch": 1.7291112293773283, "grad_norm": 0.30611518025398254, "learning_rate": 1.5025096716707516e-07, "loss": 0.2036, "step": 6498 }, { "epoch": 1.7293773283661522, "grad_norm": 0.29061010479927063, "learning_rate": 1.5023621607332554e-07, "loss": 0.2268, "step": 6499 }, { "epoch": 1.7296434273549761, "grad_norm": 0.28329238295555115, "learning_rate": 1.502214635173301e-07, "loss": 0.1923, "step": 6500 }, { "epoch": 1.7299095263437998, "grad_norm": 0.4057712256908417, "learning_rate": 1.5020670949951827e-07, "loss": 0.1779, "step": 6501 }, { "epoch": 1.7301756253326237, "grad_norm": 0.2902078628540039, "learning_rate": 1.5019195402031946e-07, "loss": 0.1914, "step": 6502 }, { "epoch": 1.7304417243214476, "grad_norm": 0.27758318185806274, "learning_rate": 1.501771970801632e-07, "loss": 0.2088, "step": 6503 }, { "epoch": 1.7307078233102713, "grad_norm": 0.2901419401168823, "learning_rate": 1.50162438679479e-07, "loss": 0.2003, "step": 6504 }, { "epoch": 1.7309739222990954, "grad_norm": 0.27263855934143066, "learning_rate": 1.5014767881869646e-07, "loss": 0.2098, "step": 6505 }, { "epoch": 1.731240021287919, "grad_norm": 0.2706345319747925, "learning_rate": 1.5013291749824527e-07, "loss": 0.1716, "step": 6506 }, { "epoch": 1.731506120276743, "grad_norm": 0.2774123549461365, "learning_rate": 1.5011815471855493e-07, "loss": 0.1985, "step": 6507 }, { "epoch": 1.731772219265567, "grad_norm": 0.2613252103328705, "learning_rate": 1.5010339048005528e-07, "loss": 0.1859, "step": 6508 }, { "epoch": 1.7320383182543906, "grad_norm": 0.2819478213787079, "learning_rate": 1.5008862478317598e-07, "loss": 0.1936, "step": 6509 }, { "epoch": 1.7323044172432145, "grad_norm": 0.2724685072898865, "learning_rate": 1.5007385762834694e-07, "loss": 0.2027, "step": 6510 }, { "epoch": 1.7325705162320384, "grad_norm": 0.26761698722839355, "learning_rate": 1.500590890159979e-07, "loss": 0.2022, "step": 6511 }, { "epoch": 1.732836615220862, "grad_norm": 0.2760135233402252, "learning_rate": 1.5004431894655874e-07, "loss": 0.2066, "step": 6512 }, { "epoch": 1.733102714209686, "grad_norm": 0.29489555954933167, "learning_rate": 1.5002954742045937e-07, "loss": 0.2017, "step": 6513 }, { "epoch": 1.7333688131985099, "grad_norm": 0.2945531904697418, "learning_rate": 1.500147744381298e-07, "loss": 0.2004, "step": 6514 }, { "epoch": 1.7336349121873336, "grad_norm": 0.32394924759864807, "learning_rate": 1.5e-07, "loss": 0.215, "step": 6515 }, { "epoch": 1.7339010111761577, "grad_norm": 0.28714513778686523, "learning_rate": 1.4998522410650004e-07, "loss": 0.2089, "step": 6516 }, { "epoch": 1.7341671101649814, "grad_norm": 0.2688346803188324, "learning_rate": 1.4997044675805996e-07, "loss": 0.1905, "step": 6517 }, { "epoch": 1.734433209153805, "grad_norm": 0.2709796726703644, "learning_rate": 1.4995566795510994e-07, "loss": 0.1867, "step": 6518 }, { "epoch": 1.7346993081426292, "grad_norm": 0.27630311250686646, "learning_rate": 1.4994088769808016e-07, "loss": 0.2153, "step": 6519 }, { "epoch": 1.7349654071314529, "grad_norm": 0.30977168679237366, "learning_rate": 1.4992610598740076e-07, "loss": 0.2218, "step": 6520 }, { "epoch": 1.7352315061202768, "grad_norm": 0.2789197266101837, "learning_rate": 1.4991132282350208e-07, "loss": 0.207, "step": 6521 }, { "epoch": 1.7354976051091007, "grad_norm": 0.2959396541118622, "learning_rate": 1.4989653820681437e-07, "loss": 0.2271, "step": 6522 }, { "epoch": 1.7357637040979244, "grad_norm": 0.2692979574203491, "learning_rate": 1.49881752137768e-07, "loss": 0.1895, "step": 6523 }, { "epoch": 1.7360298030867483, "grad_norm": 0.3049774467945099, "learning_rate": 1.4986696461679335e-07, "loss": 0.2151, "step": 6524 }, { "epoch": 1.7362959020755722, "grad_norm": 0.28791433572769165, "learning_rate": 1.4985217564432082e-07, "loss": 0.2055, "step": 6525 }, { "epoch": 1.7365620010643958, "grad_norm": 0.2832852900028229, "learning_rate": 1.498373852207809e-07, "loss": 0.2104, "step": 6526 }, { "epoch": 1.7368281000532197, "grad_norm": 0.2828238010406494, "learning_rate": 1.498225933466041e-07, "loss": 0.1979, "step": 6527 }, { "epoch": 1.7370941990420437, "grad_norm": 0.3575706481933594, "learning_rate": 1.4980780002222099e-07, "loss": 0.2215, "step": 6528 }, { "epoch": 1.7373602980308673, "grad_norm": 0.2698723077774048, "learning_rate": 1.497930052480621e-07, "loss": 0.1869, "step": 6529 }, { "epoch": 1.7376263970196915, "grad_norm": 0.35851922631263733, "learning_rate": 1.4977820902455816e-07, "loss": 0.2413, "step": 6530 }, { "epoch": 1.7378924960085151, "grad_norm": 0.2902946472167969, "learning_rate": 1.4976341135213978e-07, "loss": 0.2146, "step": 6531 }, { "epoch": 1.738158594997339, "grad_norm": 0.2843674421310425, "learning_rate": 1.4974861223123771e-07, "loss": 0.1922, "step": 6532 }, { "epoch": 1.738424693986163, "grad_norm": 0.2892574667930603, "learning_rate": 1.497338116622827e-07, "loss": 0.2038, "step": 6533 }, { "epoch": 1.7386907929749866, "grad_norm": 0.28133586049079895, "learning_rate": 1.4971900964570558e-07, "loss": 0.1851, "step": 6534 }, { "epoch": 1.7389568919638105, "grad_norm": 0.28860464692115784, "learning_rate": 1.4970420618193718e-07, "loss": 0.21, "step": 6535 }, { "epoch": 1.7392229909526344, "grad_norm": 0.3027787506580353, "learning_rate": 1.496894012714084e-07, "loss": 0.2135, "step": 6536 }, { "epoch": 1.7394890899414581, "grad_norm": 0.3270544111728668, "learning_rate": 1.4967459491455013e-07, "loss": 0.2237, "step": 6537 }, { "epoch": 1.739755188930282, "grad_norm": 0.27210134267807007, "learning_rate": 1.4965978711179342e-07, "loss": 0.1984, "step": 6538 }, { "epoch": 1.740021287919106, "grad_norm": 0.29207345843315125, "learning_rate": 1.4964497786356924e-07, "loss": 0.2188, "step": 6539 }, { "epoch": 1.7402873869079296, "grad_norm": 0.2711358070373535, "learning_rate": 1.4963016717030865e-07, "loss": 0.1978, "step": 6540 }, { "epoch": 1.7405534858967537, "grad_norm": 0.3201755881309509, "learning_rate": 1.4961535503244276e-07, "loss": 0.2062, "step": 6541 }, { "epoch": 1.7408195848855774, "grad_norm": 0.28298574686050415, "learning_rate": 1.4960054145040272e-07, "loss": 0.1893, "step": 6542 }, { "epoch": 1.7410856838744013, "grad_norm": 0.28133565187454224, "learning_rate": 1.495857264246197e-07, "loss": 0.2057, "step": 6543 }, { "epoch": 1.7413517828632252, "grad_norm": 0.3035803735256195, "learning_rate": 1.495709099555249e-07, "loss": 0.2174, "step": 6544 }, { "epoch": 1.741617881852049, "grad_norm": 0.2783547639846802, "learning_rate": 1.4955609204354965e-07, "loss": 0.1894, "step": 6545 }, { "epoch": 1.7418839808408728, "grad_norm": 0.29911887645721436, "learning_rate": 1.4954127268912524e-07, "loss": 0.2122, "step": 6546 }, { "epoch": 1.7421500798296967, "grad_norm": 0.3045242726802826, "learning_rate": 1.49526451892683e-07, "loss": 0.2235, "step": 6547 }, { "epoch": 1.7424161788185204, "grad_norm": 0.28074100613594055, "learning_rate": 1.4951162965465432e-07, "loss": 0.1918, "step": 6548 }, { "epoch": 1.7426822778073443, "grad_norm": 0.3244452476501465, "learning_rate": 1.4949680597547068e-07, "loss": 0.2024, "step": 6549 }, { "epoch": 1.7429483767961682, "grad_norm": 0.296183317899704, "learning_rate": 1.4948198085556354e-07, "loss": 0.2129, "step": 6550 }, { "epoch": 1.743214475784992, "grad_norm": 0.28647834062576294, "learning_rate": 1.494671542953644e-07, "loss": 0.2089, "step": 6551 }, { "epoch": 1.743480574773816, "grad_norm": 0.2811059355735779, "learning_rate": 1.4945232629530485e-07, "loss": 0.2047, "step": 6552 }, { "epoch": 1.7437466737626397, "grad_norm": 0.27377718687057495, "learning_rate": 1.494374968558165e-07, "loss": 0.1896, "step": 6553 }, { "epoch": 1.7440127727514634, "grad_norm": 0.279441773891449, "learning_rate": 1.4942266597733094e-07, "loss": 0.2029, "step": 6554 }, { "epoch": 1.7442788717402875, "grad_norm": 0.3074260652065277, "learning_rate": 1.4940783366027993e-07, "loss": 0.2471, "step": 6555 }, { "epoch": 1.7445449707291112, "grad_norm": 0.29202109575271606, "learning_rate": 1.4939299990509517e-07, "loss": 0.2135, "step": 6556 }, { "epoch": 1.744811069717935, "grad_norm": 0.26220542192459106, "learning_rate": 1.493781647122084e-07, "loss": 0.2108, "step": 6557 }, { "epoch": 1.745077168706759, "grad_norm": 0.28199297189712524, "learning_rate": 1.4936332808205148e-07, "loss": 0.2007, "step": 6558 }, { "epoch": 1.7453432676955827, "grad_norm": 0.2922610938549042, "learning_rate": 1.4934849001505625e-07, "loss": 0.1806, "step": 6559 }, { "epoch": 1.7456093666844066, "grad_norm": 0.2862916588783264, "learning_rate": 1.493336505116546e-07, "loss": 0.2089, "step": 6560 }, { "epoch": 1.7458754656732305, "grad_norm": 0.32193705439567566, "learning_rate": 1.493188095722785e-07, "loss": 0.2373, "step": 6561 }, { "epoch": 1.7461415646620542, "grad_norm": 0.27770838141441345, "learning_rate": 1.493039671973599e-07, "loss": 0.191, "step": 6562 }, { "epoch": 1.746407663650878, "grad_norm": 0.28929322957992554, "learning_rate": 1.4928912338733083e-07, "loss": 0.1999, "step": 6563 }, { "epoch": 1.746673762639702, "grad_norm": 0.3079686164855957, "learning_rate": 1.4927427814262337e-07, "loss": 0.2092, "step": 6564 }, { "epoch": 1.7469398616285257, "grad_norm": 0.2892700433731079, "learning_rate": 1.4925943146366957e-07, "loss": 0.2059, "step": 6565 }, { "epoch": 1.7472059606173498, "grad_norm": 0.2960789203643799, "learning_rate": 1.4924458335090164e-07, "loss": 0.2021, "step": 6566 }, { "epoch": 1.7474720596061735, "grad_norm": 0.2953714430332184, "learning_rate": 1.4922973380475177e-07, "loss": 0.1926, "step": 6567 }, { "epoch": 1.7477381585949974, "grad_norm": 0.2867305874824524, "learning_rate": 1.4921488282565217e-07, "loss": 0.2235, "step": 6568 }, { "epoch": 1.7480042575838213, "grad_norm": 0.29572248458862305, "learning_rate": 1.492000304140351e-07, "loss": 0.1912, "step": 6569 }, { "epoch": 1.748270356572645, "grad_norm": 0.31151697039604187, "learning_rate": 1.491851765703329e-07, "loss": 0.2056, "step": 6570 }, { "epoch": 1.7485364555614689, "grad_norm": 0.29098445177078247, "learning_rate": 1.4917032129497793e-07, "loss": 0.2083, "step": 6571 }, { "epoch": 1.7488025545502928, "grad_norm": 0.2726256847381592, "learning_rate": 1.4915546458840253e-07, "loss": 0.1919, "step": 6572 }, { "epoch": 1.7490686535391164, "grad_norm": 0.29703718423843384, "learning_rate": 1.4914060645103925e-07, "loss": 0.2243, "step": 6573 }, { "epoch": 1.7493347525279404, "grad_norm": 0.280627965927124, "learning_rate": 1.4912574688332046e-07, "loss": 0.1906, "step": 6574 }, { "epoch": 1.7496008515167643, "grad_norm": 0.29779937863349915, "learning_rate": 1.4911088588567874e-07, "loss": 0.2103, "step": 6575 }, { "epoch": 1.749866950505588, "grad_norm": 0.2895907461643219, "learning_rate": 1.4909602345854666e-07, "loss": 0.2053, "step": 6576 }, { "epoch": 1.750133049494412, "grad_norm": 0.3069079518318176, "learning_rate": 1.4908115960235682e-07, "loss": 0.2218, "step": 6577 }, { "epoch": 1.7503991484832357, "grad_norm": 0.3071562349796295, "learning_rate": 1.4906629431754185e-07, "loss": 0.2012, "step": 6578 }, { "epoch": 1.7506652474720596, "grad_norm": 0.2972204089164734, "learning_rate": 1.4905142760453445e-07, "loss": 0.2188, "step": 6579 }, { "epoch": 1.7509313464608836, "grad_norm": 0.2898423969745636, "learning_rate": 1.4903655946376735e-07, "loss": 0.2063, "step": 6580 }, { "epoch": 1.7511974454497072, "grad_norm": 0.3264787495136261, "learning_rate": 1.4902168989567334e-07, "loss": 0.207, "step": 6581 }, { "epoch": 1.7514635444385311, "grad_norm": 0.30254924297332764, "learning_rate": 1.490068189006852e-07, "loss": 0.2048, "step": 6582 }, { "epoch": 1.751729643427355, "grad_norm": 0.28440558910369873, "learning_rate": 1.4899194647923584e-07, "loss": 0.1965, "step": 6583 }, { "epoch": 1.7519957424161787, "grad_norm": 0.2868591248989105, "learning_rate": 1.489770726317581e-07, "loss": 0.2105, "step": 6584 }, { "epoch": 1.7522618414050026, "grad_norm": 0.26206305623054504, "learning_rate": 1.4896219735868495e-07, "loss": 0.2035, "step": 6585 }, { "epoch": 1.7525279403938265, "grad_norm": 0.28385841846466064, "learning_rate": 1.4894732066044934e-07, "loss": 0.1994, "step": 6586 }, { "epoch": 1.7527940393826502, "grad_norm": 0.3023940622806549, "learning_rate": 1.4893244253748435e-07, "loss": 0.2264, "step": 6587 }, { "epoch": 1.7530601383714743, "grad_norm": 0.28157296776771545, "learning_rate": 1.48917562990223e-07, "loss": 0.2111, "step": 6588 }, { "epoch": 1.753326237360298, "grad_norm": 0.28226912021636963, "learning_rate": 1.489026820190984e-07, "loss": 0.1919, "step": 6589 }, { "epoch": 1.753592336349122, "grad_norm": 0.2930077910423279, "learning_rate": 1.488877996245437e-07, "loss": 0.2107, "step": 6590 }, { "epoch": 1.7538584353379458, "grad_norm": 0.29273441433906555, "learning_rate": 1.4887291580699208e-07, "loss": 0.1997, "step": 6591 }, { "epoch": 1.7541245343267695, "grad_norm": 0.2693907618522644, "learning_rate": 1.4885803056687682e-07, "loss": 0.1903, "step": 6592 }, { "epoch": 1.7543906333155934, "grad_norm": 0.28784918785095215, "learning_rate": 1.488431439046311e-07, "loss": 0.2139, "step": 6593 }, { "epoch": 1.7546567323044173, "grad_norm": 0.26715749502182007, "learning_rate": 1.488282558206883e-07, "loss": 0.1983, "step": 6594 }, { "epoch": 1.754922831293241, "grad_norm": 0.26793184876441956, "learning_rate": 1.4881336631548176e-07, "loss": 0.1927, "step": 6595 }, { "epoch": 1.755188930282065, "grad_norm": 0.2788327932357788, "learning_rate": 1.4879847538944485e-07, "loss": 0.1868, "step": 6596 }, { "epoch": 1.7554550292708888, "grad_norm": 0.2894516885280609, "learning_rate": 1.4878358304301105e-07, "loss": 0.2075, "step": 6597 }, { "epoch": 1.7557211282597125, "grad_norm": 0.29258859157562256, "learning_rate": 1.4876868927661379e-07, "loss": 0.2075, "step": 6598 }, { "epoch": 1.7559872272485366, "grad_norm": 0.32713252305984497, "learning_rate": 1.4875379409068663e-07, "loss": 0.1995, "step": 6599 }, { "epoch": 1.7562533262373603, "grad_norm": 0.2956000864505768, "learning_rate": 1.4873889748566309e-07, "loss": 0.2297, "step": 6600 }, { "epoch": 1.756519425226184, "grad_norm": 0.27366653084754944, "learning_rate": 1.4872399946197682e-07, "loss": 0.1986, "step": 6601 }, { "epoch": 1.756785524215008, "grad_norm": 0.276906281709671, "learning_rate": 1.4870910002006145e-07, "loss": 0.2043, "step": 6602 }, { "epoch": 1.7570516232038318, "grad_norm": 0.2762184739112854, "learning_rate": 1.4869419916035062e-07, "loss": 0.2028, "step": 6603 }, { "epoch": 1.7573177221926557, "grad_norm": 0.2808970808982849, "learning_rate": 1.4867929688327806e-07, "loss": 0.1838, "step": 6604 }, { "epoch": 1.7575838211814796, "grad_norm": 0.28396451473236084, "learning_rate": 1.4866439318927762e-07, "loss": 0.1937, "step": 6605 }, { "epoch": 1.7578499201703033, "grad_norm": 0.28399598598480225, "learning_rate": 1.48649488078783e-07, "loss": 0.1955, "step": 6606 }, { "epoch": 1.7581160191591272, "grad_norm": 0.26591044664382935, "learning_rate": 1.4863458155222813e-07, "loss": 0.1833, "step": 6607 }, { "epoch": 1.758382118147951, "grad_norm": 0.27375859022140503, "learning_rate": 1.4861967361004686e-07, "loss": 0.2101, "step": 6608 }, { "epoch": 1.7586482171367748, "grad_norm": 0.292851984500885, "learning_rate": 1.4860476425267312e-07, "loss": 0.1947, "step": 6609 }, { "epoch": 1.7589143161255987, "grad_norm": 0.2966088056564331, "learning_rate": 1.485898534805409e-07, "loss": 0.2093, "step": 6610 }, { "epoch": 1.7591804151144226, "grad_norm": 0.32555851340293884, "learning_rate": 1.4857494129408424e-07, "loss": 0.2174, "step": 6611 }, { "epoch": 1.7594465141032463, "grad_norm": 0.298292338848114, "learning_rate": 1.4856002769373716e-07, "loss": 0.2099, "step": 6612 }, { "epoch": 1.7597126130920704, "grad_norm": 0.2875332832336426, "learning_rate": 1.485451126799337e-07, "loss": 0.1911, "step": 6613 }, { "epoch": 1.759978712080894, "grad_norm": 0.2712528109550476, "learning_rate": 1.4853019625310813e-07, "loss": 0.1903, "step": 6614 }, { "epoch": 1.760244811069718, "grad_norm": 0.32487139105796814, "learning_rate": 1.485152784136945e-07, "loss": 0.2056, "step": 6615 }, { "epoch": 1.7605109100585419, "grad_norm": 0.2720268666744232, "learning_rate": 1.4850035916212707e-07, "loss": 0.1919, "step": 6616 }, { "epoch": 1.7607770090473656, "grad_norm": 0.28797629475593567, "learning_rate": 1.4848543849884015e-07, "loss": 0.2159, "step": 6617 }, { "epoch": 1.7610431080361895, "grad_norm": 0.28384050726890564, "learning_rate": 1.48470516424268e-07, "loss": 0.2037, "step": 6618 }, { "epoch": 1.7613092070250134, "grad_norm": 0.28962400555610657, "learning_rate": 1.4845559293884498e-07, "loss": 0.2079, "step": 6619 }, { "epoch": 1.761575306013837, "grad_norm": 0.28767111897468567, "learning_rate": 1.4844066804300544e-07, "loss": 0.2126, "step": 6620 }, { "epoch": 1.761841405002661, "grad_norm": 0.2965261936187744, "learning_rate": 1.4842574173718383e-07, "loss": 0.213, "step": 6621 }, { "epoch": 1.7621075039914849, "grad_norm": 0.2613345980644226, "learning_rate": 1.4841081402181465e-07, "loss": 0.1795, "step": 6622 }, { "epoch": 1.7623736029803085, "grad_norm": 0.7940399646759033, "learning_rate": 1.4839588489733236e-07, "loss": 0.2024, "step": 6623 }, { "epoch": 1.7626397019691327, "grad_norm": 0.2904520332813263, "learning_rate": 1.483809543641715e-07, "loss": 0.1919, "step": 6624 }, { "epoch": 1.7629058009579563, "grad_norm": 0.2754562199115753, "learning_rate": 1.4836602242276668e-07, "loss": 0.206, "step": 6625 }, { "epoch": 1.7631718999467803, "grad_norm": 0.28297004103660583, "learning_rate": 1.483510890735525e-07, "loss": 0.1905, "step": 6626 }, { "epoch": 1.7634379989356042, "grad_norm": 0.29531151056289673, "learning_rate": 1.4833615431696371e-07, "loss": 0.2086, "step": 6627 }, { "epoch": 1.7637040979244278, "grad_norm": 0.3105422854423523, "learning_rate": 1.4832121815343498e-07, "loss": 0.2307, "step": 6628 }, { "epoch": 1.7639701969132517, "grad_norm": 0.2946527302265167, "learning_rate": 1.48306280583401e-07, "loss": 0.2081, "step": 6629 }, { "epoch": 1.7642362959020756, "grad_norm": 0.266647070646286, "learning_rate": 1.4829134160729666e-07, "loss": 0.198, "step": 6630 }, { "epoch": 1.7645023948908993, "grad_norm": 0.3059976398944855, "learning_rate": 1.4827640122555672e-07, "loss": 0.2143, "step": 6631 }, { "epoch": 1.7647684938797232, "grad_norm": 0.28805452585220337, "learning_rate": 1.482614594386161e-07, "loss": 0.2004, "step": 6632 }, { "epoch": 1.7650345928685471, "grad_norm": 0.2747862935066223, "learning_rate": 1.4824651624690977e-07, "loss": 0.1938, "step": 6633 }, { "epoch": 1.7653006918573708, "grad_norm": 0.2727499306201935, "learning_rate": 1.4823157165087254e-07, "loss": 0.1988, "step": 6634 }, { "epoch": 1.765566790846195, "grad_norm": 0.2831343710422516, "learning_rate": 1.482166256509395e-07, "loss": 0.198, "step": 6635 }, { "epoch": 1.7658328898350186, "grad_norm": 0.3089476525783539, "learning_rate": 1.482016782475457e-07, "loss": 0.2208, "step": 6636 }, { "epoch": 1.7660989888238423, "grad_norm": 0.295501172542572, "learning_rate": 1.4818672944112623e-07, "loss": 0.2159, "step": 6637 }, { "epoch": 1.7663650878126664, "grad_norm": 0.30256080627441406, "learning_rate": 1.4817177923211612e-07, "loss": 0.1909, "step": 6638 }, { "epoch": 1.7666311868014901, "grad_norm": 0.2918214201927185, "learning_rate": 1.4815682762095062e-07, "loss": 0.1934, "step": 6639 }, { "epoch": 1.766897285790314, "grad_norm": 0.2820045053958893, "learning_rate": 1.481418746080649e-07, "loss": 0.1937, "step": 6640 }, { "epoch": 1.767163384779138, "grad_norm": 0.27864569425582886, "learning_rate": 1.4812692019389422e-07, "loss": 0.2048, "step": 6641 }, { "epoch": 1.7674294837679616, "grad_norm": 0.28270819783210754, "learning_rate": 1.4811196437887387e-07, "loss": 0.1998, "step": 6642 }, { "epoch": 1.7676955827567855, "grad_norm": 0.282062828540802, "learning_rate": 1.4809700716343914e-07, "loss": 0.1934, "step": 6643 }, { "epoch": 1.7679616817456094, "grad_norm": 0.30844178795814514, "learning_rate": 1.4808204854802542e-07, "loss": 0.216, "step": 6644 }, { "epoch": 1.768227780734433, "grad_norm": 0.282703697681427, "learning_rate": 1.4806708853306812e-07, "loss": 0.1792, "step": 6645 }, { "epoch": 1.768493879723257, "grad_norm": 0.26759570837020874, "learning_rate": 1.4805212711900267e-07, "loss": 0.1822, "step": 6646 }, { "epoch": 1.768759978712081, "grad_norm": 0.25411805510520935, "learning_rate": 1.4803716430626454e-07, "loss": 0.1769, "step": 6647 }, { "epoch": 1.7690260777009046, "grad_norm": 0.34187960624694824, "learning_rate": 1.4802220009528929e-07, "loss": 0.1933, "step": 6648 }, { "epoch": 1.7692921766897287, "grad_norm": 0.2999921441078186, "learning_rate": 1.480072344865125e-07, "loss": 0.2207, "step": 6649 }, { "epoch": 1.7695582756785524, "grad_norm": 0.3112598955631256, "learning_rate": 1.4799226748036978e-07, "loss": 0.1972, "step": 6650 }, { "epoch": 1.7698243746673763, "grad_norm": 0.3010234236717224, "learning_rate": 1.4797729907729673e-07, "loss": 0.2094, "step": 6651 }, { "epoch": 1.7700904736562002, "grad_norm": 0.27038589119911194, "learning_rate": 1.4796232927772907e-07, "loss": 0.1929, "step": 6652 }, { "epoch": 1.7703565726450239, "grad_norm": 0.35593870282173157, "learning_rate": 1.4794735808210258e-07, "loss": 0.2143, "step": 6653 }, { "epoch": 1.7706226716338478, "grad_norm": 0.3307666480541229, "learning_rate": 1.4793238549085297e-07, "loss": 0.2226, "step": 6654 }, { "epoch": 1.7708887706226717, "grad_norm": 0.28993386030197144, "learning_rate": 1.4791741150441608e-07, "loss": 0.2059, "step": 6655 }, { "epoch": 1.7711548696114954, "grad_norm": 0.2917409837245941, "learning_rate": 1.4790243612322772e-07, "loss": 0.2009, "step": 6656 }, { "epoch": 1.7714209686003193, "grad_norm": 0.2900799810886383, "learning_rate": 1.4788745934772386e-07, "loss": 0.1785, "step": 6657 }, { "epoch": 1.7716870675891432, "grad_norm": 0.2952491343021393, "learning_rate": 1.4787248117834037e-07, "loss": 0.2151, "step": 6658 }, { "epoch": 1.7719531665779669, "grad_norm": 0.31260937452316284, "learning_rate": 1.478575016155133e-07, "loss": 0.2376, "step": 6659 }, { "epoch": 1.772219265566791, "grad_norm": 0.2883649468421936, "learning_rate": 1.4784252065967857e-07, "loss": 0.2047, "step": 6660 }, { "epoch": 1.7724853645556147, "grad_norm": 0.33560821413993835, "learning_rate": 1.478275383112723e-07, "loss": 0.2106, "step": 6661 }, { "epoch": 1.7727514635444386, "grad_norm": 0.2780805230140686, "learning_rate": 1.4781255457073059e-07, "loss": 0.2094, "step": 6662 }, { "epoch": 1.7730175625332625, "grad_norm": 0.27703720331192017, "learning_rate": 1.4779756943848956e-07, "loss": 0.1885, "step": 6663 }, { "epoch": 1.7732836615220862, "grad_norm": 0.27010855078697205, "learning_rate": 1.4778258291498538e-07, "loss": 0.1703, "step": 6664 }, { "epoch": 1.77354976051091, "grad_norm": 0.2945290803909302, "learning_rate": 1.4776759500065424e-07, "loss": 0.2132, "step": 6665 }, { "epoch": 1.773815859499734, "grad_norm": 0.28866907954216003, "learning_rate": 1.477526056959325e-07, "loss": 0.1984, "step": 6666 }, { "epoch": 1.7740819584885577, "grad_norm": 0.3027404248714447, "learning_rate": 1.4773761500125634e-07, "loss": 0.2119, "step": 6667 }, { "epoch": 1.7743480574773816, "grad_norm": 0.2901100516319275, "learning_rate": 1.477226229170622e-07, "loss": 0.1973, "step": 6668 }, { "epoch": 1.7746141564662055, "grad_norm": 0.27083513140678406, "learning_rate": 1.4770762944378641e-07, "loss": 0.1981, "step": 6669 }, { "epoch": 1.7748802554550291, "grad_norm": 0.31054723262786865, "learning_rate": 1.4769263458186539e-07, "loss": 0.2153, "step": 6670 }, { "epoch": 1.7751463544438533, "grad_norm": 0.27468302845954895, "learning_rate": 1.476776383317356e-07, "loss": 0.2004, "step": 6671 }, { "epoch": 1.775412453432677, "grad_norm": 0.26774224638938904, "learning_rate": 1.4766264069383357e-07, "loss": 0.2031, "step": 6672 }, { "epoch": 1.7756785524215006, "grad_norm": 0.32089680433273315, "learning_rate": 1.476476416685958e-07, "loss": 0.2455, "step": 6673 }, { "epoch": 1.7759446514103248, "grad_norm": 0.28937897086143494, "learning_rate": 1.4763264125645892e-07, "loss": 0.2047, "step": 6674 }, { "epoch": 1.7762107503991484, "grad_norm": 0.2828442454338074, "learning_rate": 1.476176394578595e-07, "loss": 0.1994, "step": 6675 }, { "epoch": 1.7764768493879723, "grad_norm": 0.29525983333587646, "learning_rate": 1.4760263627323428e-07, "loss": 0.219, "step": 6676 }, { "epoch": 1.7767429483767962, "grad_norm": 0.3004536032676697, "learning_rate": 1.475876317030199e-07, "loss": 0.1884, "step": 6677 }, { "epoch": 1.77700904736562, "grad_norm": 0.3235546052455902, "learning_rate": 1.475726257476531e-07, "loss": 0.2282, "step": 6678 }, { "epoch": 1.7772751463544438, "grad_norm": 0.3182786703109741, "learning_rate": 1.475576184075707e-07, "loss": 0.227, "step": 6679 }, { "epoch": 1.7775412453432677, "grad_norm": 0.2733442187309265, "learning_rate": 1.4754260968320952e-07, "loss": 0.1885, "step": 6680 }, { "epoch": 1.7778073443320914, "grad_norm": 0.30743902921676636, "learning_rate": 1.475275995750064e-07, "loss": 0.2048, "step": 6681 }, { "epoch": 1.7780734433209153, "grad_norm": 0.281473308801651, "learning_rate": 1.4751258808339827e-07, "loss": 0.2082, "step": 6682 }, { "epoch": 1.7783395423097392, "grad_norm": 0.30773916840553284, "learning_rate": 1.4749757520882206e-07, "loss": 0.2264, "step": 6683 }, { "epoch": 1.778605641298563, "grad_norm": 0.27334538102149963, "learning_rate": 1.4748256095171477e-07, "loss": 0.194, "step": 6684 }, { "epoch": 1.778871740287387, "grad_norm": 0.2823817729949951, "learning_rate": 1.474675453125134e-07, "loss": 0.1997, "step": 6685 }, { "epoch": 1.7791378392762107, "grad_norm": 0.2714548408985138, "learning_rate": 1.4745252829165507e-07, "loss": 0.188, "step": 6686 }, { "epoch": 1.7794039382650346, "grad_norm": 0.3567245602607727, "learning_rate": 1.474375098895768e-07, "loss": 0.1933, "step": 6687 }, { "epoch": 1.7796700372538585, "grad_norm": 0.27606022357940674, "learning_rate": 1.4742249010671581e-07, "loss": 0.1828, "step": 6688 }, { "epoch": 1.7799361362426822, "grad_norm": 0.3066200315952301, "learning_rate": 1.4740746894350928e-07, "loss": 0.2006, "step": 6689 }, { "epoch": 1.7802022352315061, "grad_norm": 0.3128347396850586, "learning_rate": 1.4739244640039438e-07, "loss": 0.222, "step": 6690 }, { "epoch": 1.78046833422033, "grad_norm": 0.2799991965293884, "learning_rate": 1.4737742247780849e-07, "loss": 0.2115, "step": 6691 }, { "epoch": 1.7807344332091537, "grad_norm": 0.2896055281162262, "learning_rate": 1.4736239717618879e-07, "loss": 0.1844, "step": 6692 }, { "epoch": 1.7810005321979776, "grad_norm": 0.2992366850376129, "learning_rate": 1.4734737049597273e-07, "loss": 0.1979, "step": 6693 }, { "epoch": 1.7812666311868015, "grad_norm": 0.3043414354324341, "learning_rate": 1.473323424375976e-07, "loss": 0.2035, "step": 6694 }, { "epoch": 1.7815327301756252, "grad_norm": 0.2878936231136322, "learning_rate": 1.473173130015009e-07, "loss": 0.1823, "step": 6695 }, { "epoch": 1.7817988291644493, "grad_norm": 0.2728201746940613, "learning_rate": 1.4730228218812006e-07, "loss": 0.187, "step": 6696 }, { "epoch": 1.782064928153273, "grad_norm": 0.2792714238166809, "learning_rate": 1.4728724999789263e-07, "loss": 0.2038, "step": 6697 }, { "epoch": 1.782331027142097, "grad_norm": 0.27630874514579773, "learning_rate": 1.4727221643125612e-07, "loss": 0.1912, "step": 6698 }, { "epoch": 1.7825971261309208, "grad_norm": 0.25891679525375366, "learning_rate": 1.4725718148864812e-07, "loss": 0.1905, "step": 6699 }, { "epoch": 1.7828632251197445, "grad_norm": 0.28256070613861084, "learning_rate": 1.472421451705063e-07, "loss": 0.1982, "step": 6700 }, { "epoch": 1.7831293241085684, "grad_norm": 0.28361350297927856, "learning_rate": 1.4722710747726827e-07, "loss": 0.2101, "step": 6701 }, { "epoch": 1.7833954230973923, "grad_norm": 0.28058141469955444, "learning_rate": 1.4721206840937178e-07, "loss": 0.1974, "step": 6702 }, { "epoch": 1.783661522086216, "grad_norm": 0.39471861720085144, "learning_rate": 1.4719702796725454e-07, "loss": 0.2113, "step": 6703 }, { "epoch": 1.7839276210750399, "grad_norm": 0.2783827781677246, "learning_rate": 1.471819861513544e-07, "loss": 0.1913, "step": 6704 }, { "epoch": 1.7841937200638638, "grad_norm": 0.26834550499916077, "learning_rate": 1.4716694296210913e-07, "loss": 0.1908, "step": 6705 }, { "epoch": 1.7844598190526875, "grad_norm": 0.2846907377243042, "learning_rate": 1.4715189839995662e-07, "loss": 0.2099, "step": 6706 }, { "epoch": 1.7847259180415116, "grad_norm": 0.2890234589576721, "learning_rate": 1.4713685246533475e-07, "loss": 0.204, "step": 6707 }, { "epoch": 1.7849920170303353, "grad_norm": 0.2560493052005768, "learning_rate": 1.471218051586815e-07, "loss": 0.1911, "step": 6708 }, { "epoch": 1.7852581160191592, "grad_norm": 0.2978205680847168, "learning_rate": 1.471067564804349e-07, "loss": 0.2147, "step": 6709 }, { "epoch": 1.785524215007983, "grad_norm": 0.3050074875354767, "learning_rate": 1.470917064310329e-07, "loss": 0.2249, "step": 6710 }, { "epoch": 1.7857903139968068, "grad_norm": 0.2592598795890808, "learning_rate": 1.4707665501091358e-07, "loss": 0.1774, "step": 6711 }, { "epoch": 1.7860564129856307, "grad_norm": 0.3013363182544708, "learning_rate": 1.4706160222051508e-07, "loss": 0.2076, "step": 6712 }, { "epoch": 1.7863225119744546, "grad_norm": 0.2901267409324646, "learning_rate": 1.4704654806027556e-07, "loss": 0.2005, "step": 6713 }, { "epoch": 1.7865886109632783, "grad_norm": 0.2967197000980377, "learning_rate": 1.4703149253063316e-07, "loss": 0.1988, "step": 6714 }, { "epoch": 1.7868547099521022, "grad_norm": 0.2926912009716034, "learning_rate": 1.4701643563202611e-07, "loss": 0.2081, "step": 6715 }, { "epoch": 1.787120808940926, "grad_norm": 0.28135761618614197, "learning_rate": 1.4700137736489274e-07, "loss": 0.1983, "step": 6716 }, { "epoch": 1.7873869079297497, "grad_norm": 0.3087460696697235, "learning_rate": 1.4698631772967127e-07, "loss": 0.2027, "step": 6717 }, { "epoch": 1.7876530069185739, "grad_norm": 0.2813622057437897, "learning_rate": 1.4697125672680012e-07, "loss": 0.195, "step": 6718 }, { "epoch": 1.7879191059073976, "grad_norm": 0.2849574685096741, "learning_rate": 1.4695619435671763e-07, "loss": 0.204, "step": 6719 }, { "epoch": 1.7881852048962212, "grad_norm": 0.2746783196926117, "learning_rate": 1.4694113061986224e-07, "loss": 0.1735, "step": 6720 }, { "epoch": 1.7884513038850454, "grad_norm": 0.3096171021461487, "learning_rate": 1.4692606551667242e-07, "loss": 0.223, "step": 6721 }, { "epoch": 1.788717402873869, "grad_norm": 0.26070213317871094, "learning_rate": 1.4691099904758666e-07, "loss": 0.2019, "step": 6722 }, { "epoch": 1.788983501862693, "grad_norm": 0.3194710910320282, "learning_rate": 1.4689593121304356e-07, "loss": 0.2001, "step": 6723 }, { "epoch": 1.7892496008515169, "grad_norm": 0.2962004244327545, "learning_rate": 1.4688086201348165e-07, "loss": 0.1859, "step": 6724 }, { "epoch": 1.7895156998403405, "grad_norm": 0.2864803075790405, "learning_rate": 1.4686579144933957e-07, "loss": 0.1967, "step": 6725 }, { "epoch": 1.7897817988291644, "grad_norm": 0.26567211747169495, "learning_rate": 1.4685071952105595e-07, "loss": 0.1785, "step": 6726 }, { "epoch": 1.7900478978179883, "grad_norm": 0.31071558594703674, "learning_rate": 1.468356462290696e-07, "loss": 0.1989, "step": 6727 }, { "epoch": 1.790313996806812, "grad_norm": 0.33177170157432556, "learning_rate": 1.4682057157381916e-07, "loss": 0.2185, "step": 6728 }, { "epoch": 1.790580095795636, "grad_norm": 0.30125492811203003, "learning_rate": 1.4680549555574344e-07, "loss": 0.2229, "step": 6729 }, { "epoch": 1.7908461947844598, "grad_norm": 0.2639627754688263, "learning_rate": 1.4679041817528128e-07, "loss": 0.1914, "step": 6730 }, { "epoch": 1.7911122937732835, "grad_norm": 0.29589059948921204, "learning_rate": 1.4677533943287156e-07, "loss": 0.2054, "step": 6731 }, { "epoch": 1.7913783927621076, "grad_norm": 0.268775999546051, "learning_rate": 1.4676025932895312e-07, "loss": 0.1978, "step": 6732 }, { "epoch": 1.7916444917509313, "grad_norm": 0.30578774213790894, "learning_rate": 1.4674517786396497e-07, "loss": 0.215, "step": 6733 }, { "epoch": 1.7919105907397552, "grad_norm": 0.29756295680999756, "learning_rate": 1.4673009503834607e-07, "loss": 0.1996, "step": 6734 }, { "epoch": 1.7921766897285791, "grad_norm": 0.3104115426540375, "learning_rate": 1.4671501085253542e-07, "loss": 0.2141, "step": 6735 }, { "epoch": 1.7924427887174028, "grad_norm": 0.2866760790348053, "learning_rate": 1.4669992530697213e-07, "loss": 0.2193, "step": 6736 }, { "epoch": 1.7927088877062267, "grad_norm": 0.2954728603363037, "learning_rate": 1.4668483840209524e-07, "loss": 0.2104, "step": 6737 }, { "epoch": 1.7929749866950506, "grad_norm": 0.28920045495033264, "learning_rate": 1.466697501383439e-07, "loss": 0.1977, "step": 6738 }, { "epoch": 1.7932410856838743, "grad_norm": 0.29738011956214905, "learning_rate": 1.4665466051615734e-07, "loss": 0.2003, "step": 6739 }, { "epoch": 1.7935071846726982, "grad_norm": 0.3001498878002167, "learning_rate": 1.4663956953597473e-07, "loss": 0.225, "step": 6740 }, { "epoch": 1.7937732836615221, "grad_norm": 0.27956002950668335, "learning_rate": 1.466244771982354e-07, "loss": 0.1818, "step": 6741 }, { "epoch": 1.7940393826503458, "grad_norm": 0.29268544912338257, "learning_rate": 1.4660938350337855e-07, "loss": 0.203, "step": 6742 }, { "epoch": 1.79430548163917, "grad_norm": 0.2754947543144226, "learning_rate": 1.4659428845184357e-07, "loss": 0.2001, "step": 6743 }, { "epoch": 1.7945715806279936, "grad_norm": 0.30053168535232544, "learning_rate": 1.4657919204406987e-07, "loss": 0.2216, "step": 6744 }, { "epoch": 1.7948376796168175, "grad_norm": 0.25940725207328796, "learning_rate": 1.465640942804968e-07, "loss": 0.1791, "step": 6745 }, { "epoch": 1.7951037786056414, "grad_norm": 0.288790225982666, "learning_rate": 1.4654899516156384e-07, "loss": 0.1993, "step": 6746 }, { "epoch": 1.795369877594465, "grad_norm": 0.29209649562835693, "learning_rate": 1.4653389468771052e-07, "loss": 0.2052, "step": 6747 }, { "epoch": 1.795635976583289, "grad_norm": 0.3103453516960144, "learning_rate": 1.465187928593763e-07, "loss": 0.2137, "step": 6748 }, { "epoch": 1.795902075572113, "grad_norm": 0.3403991460800171, "learning_rate": 1.4650368967700083e-07, "loss": 0.2081, "step": 6749 }, { "epoch": 1.7961681745609366, "grad_norm": 0.3113594353199005, "learning_rate": 1.4648858514102375e-07, "loss": 0.2024, "step": 6750 }, { "epoch": 1.7964342735497605, "grad_norm": 0.28730881214141846, "learning_rate": 1.464734792518846e-07, "loss": 0.2084, "step": 6751 }, { "epoch": 1.7967003725385844, "grad_norm": 0.3161880373954773, "learning_rate": 1.4645837201002315e-07, "loss": 0.2164, "step": 6752 }, { "epoch": 1.796966471527408, "grad_norm": 0.2676730453968048, "learning_rate": 1.4644326341587913e-07, "loss": 0.1903, "step": 6753 }, { "epoch": 1.7972325705162322, "grad_norm": 0.3212451934814453, "learning_rate": 1.4642815346989231e-07, "loss": 0.207, "step": 6754 }, { "epoch": 1.7974986695050559, "grad_norm": 0.3005869686603546, "learning_rate": 1.464130421725025e-07, "loss": 0.2061, "step": 6755 }, { "epoch": 1.7977647684938796, "grad_norm": 0.30073821544647217, "learning_rate": 1.463979295241495e-07, "loss": 0.2065, "step": 6756 }, { "epoch": 1.7980308674827037, "grad_norm": 0.3074568212032318, "learning_rate": 1.4638281552527328e-07, "loss": 0.2099, "step": 6757 }, { "epoch": 1.7982969664715274, "grad_norm": 0.2797574996948242, "learning_rate": 1.463677001763137e-07, "loss": 0.1947, "step": 6758 }, { "epoch": 1.7985630654603513, "grad_norm": 0.2793128192424774, "learning_rate": 1.463525834777108e-07, "loss": 0.2056, "step": 6759 }, { "epoch": 1.7988291644491752, "grad_norm": 0.2927803099155426, "learning_rate": 1.463374654299045e-07, "loss": 0.2076, "step": 6760 }, { "epoch": 1.7990952634379989, "grad_norm": 0.3025144636631012, "learning_rate": 1.4632234603333492e-07, "loss": 0.2154, "step": 6761 }, { "epoch": 1.7993613624268228, "grad_norm": 0.29382479190826416, "learning_rate": 1.4630722528844212e-07, "loss": 0.2239, "step": 6762 }, { "epoch": 1.7996274614156467, "grad_norm": 0.3076229989528656, "learning_rate": 1.4629210319566625e-07, "loss": 0.2319, "step": 6763 }, { "epoch": 1.7998935604044703, "grad_norm": 0.283760130405426, "learning_rate": 1.4627697975544741e-07, "loss": 0.1973, "step": 6764 }, { "epoch": 1.8001596593932943, "grad_norm": 0.3120848834514618, "learning_rate": 1.462618549682259e-07, "loss": 0.2207, "step": 6765 }, { "epoch": 1.8004257583821182, "grad_norm": 0.3177171051502228, "learning_rate": 1.4624672883444187e-07, "loss": 0.2281, "step": 6766 }, { "epoch": 1.8006918573709418, "grad_norm": 0.2950533926486969, "learning_rate": 1.4623160135453566e-07, "loss": 0.2101, "step": 6767 }, { "epoch": 1.800957956359766, "grad_norm": 0.2864872217178345, "learning_rate": 1.4621647252894755e-07, "loss": 0.2008, "step": 6768 }, { "epoch": 1.8012240553485896, "grad_norm": 0.26990777254104614, "learning_rate": 1.4620134235811793e-07, "loss": 0.1936, "step": 6769 }, { "epoch": 1.8014901543374136, "grad_norm": 0.2963883876800537, "learning_rate": 1.4618621084248722e-07, "loss": 0.1956, "step": 6770 }, { "epoch": 1.8017562533262375, "grad_norm": 0.29009491205215454, "learning_rate": 1.4617107798249578e-07, "loss": 0.1921, "step": 6771 }, { "epoch": 1.8020223523150611, "grad_norm": 0.29621872305870056, "learning_rate": 1.4615594377858422e-07, "loss": 0.2127, "step": 6772 }, { "epoch": 1.802288451303885, "grad_norm": 0.29499703645706177, "learning_rate": 1.4614080823119293e-07, "loss": 0.2033, "step": 6773 }, { "epoch": 1.802554550292709, "grad_norm": 0.2974400520324707, "learning_rate": 1.461256713407625e-07, "loss": 0.2173, "step": 6774 }, { "epoch": 1.8028206492815326, "grad_norm": 0.2796463370323181, "learning_rate": 1.461105331077336e-07, "loss": 0.2018, "step": 6775 }, { "epoch": 1.8030867482703565, "grad_norm": 0.28261059522628784, "learning_rate": 1.4609539353254677e-07, "loss": 0.2099, "step": 6776 }, { "epoch": 1.8033528472591804, "grad_norm": 0.3055698871612549, "learning_rate": 1.4608025261564272e-07, "loss": 0.2161, "step": 6777 }, { "epoch": 1.8036189462480041, "grad_norm": 0.3165585398674011, "learning_rate": 1.4606511035746215e-07, "loss": 0.2113, "step": 6778 }, { "epoch": 1.8038850452368282, "grad_norm": 0.29239651560783386, "learning_rate": 1.4604996675844582e-07, "loss": 0.1956, "step": 6779 }, { "epoch": 1.804151144225652, "grad_norm": 0.27118390798568726, "learning_rate": 1.4603482181903455e-07, "loss": 0.1886, "step": 6780 }, { "epoch": 1.8044172432144758, "grad_norm": 0.30613529682159424, "learning_rate": 1.4601967553966917e-07, "loss": 0.2148, "step": 6781 }, { "epoch": 1.8046833422032997, "grad_norm": 0.27548184990882874, "learning_rate": 1.4600452792079048e-07, "loss": 0.2061, "step": 6782 }, { "epoch": 1.8049494411921234, "grad_norm": 0.2871055603027344, "learning_rate": 1.4598937896283942e-07, "loss": 0.2078, "step": 6783 }, { "epoch": 1.8052155401809473, "grad_norm": 0.3047076463699341, "learning_rate": 1.45974228666257e-07, "loss": 0.2127, "step": 6784 }, { "epoch": 1.8054816391697712, "grad_norm": 0.27659082412719727, "learning_rate": 1.4595907703148411e-07, "loss": 0.191, "step": 6785 }, { "epoch": 1.805747738158595, "grad_norm": 0.2844659090042114, "learning_rate": 1.4594392405896185e-07, "loss": 0.2048, "step": 6786 }, { "epoch": 1.8060138371474188, "grad_norm": 0.29509711265563965, "learning_rate": 1.4592876974913123e-07, "loss": 0.2135, "step": 6787 }, { "epoch": 1.8062799361362427, "grad_norm": 0.2651783525943756, "learning_rate": 1.4591361410243338e-07, "loss": 0.1963, "step": 6788 }, { "epoch": 1.8065460351250664, "grad_norm": 0.3006054759025574, "learning_rate": 1.4589845711930943e-07, "loss": 0.2077, "step": 6789 }, { "epoch": 1.8068121341138905, "grad_norm": 0.27320364117622375, "learning_rate": 1.458832988002006e-07, "loss": 0.1987, "step": 6790 }, { "epoch": 1.8070782331027142, "grad_norm": 0.2785598635673523, "learning_rate": 1.4586813914554805e-07, "loss": 0.1923, "step": 6791 }, { "epoch": 1.8073443320915379, "grad_norm": 0.28314462304115295, "learning_rate": 1.4585297815579308e-07, "loss": 0.1913, "step": 6792 }, { "epoch": 1.807610431080362, "grad_norm": 0.2508185803890228, "learning_rate": 1.4583781583137695e-07, "loss": 0.1737, "step": 6793 }, { "epoch": 1.8078765300691857, "grad_norm": 0.28746914863586426, "learning_rate": 1.4582265217274104e-07, "loss": 0.2078, "step": 6794 }, { "epoch": 1.8081426290580096, "grad_norm": 0.29618653655052185, "learning_rate": 1.4580748718032667e-07, "loss": 0.205, "step": 6795 }, { "epoch": 1.8084087280468335, "grad_norm": 0.30313676595687866, "learning_rate": 1.4579232085457533e-07, "loss": 0.2172, "step": 6796 }, { "epoch": 1.8086748270356572, "grad_norm": 0.2940548360347748, "learning_rate": 1.457771531959284e-07, "loss": 0.2119, "step": 6797 }, { "epoch": 1.808940926024481, "grad_norm": 0.3272869884967804, "learning_rate": 1.4576198420482742e-07, "loss": 0.19, "step": 6798 }, { "epoch": 1.809207025013305, "grad_norm": 0.286878764629364, "learning_rate": 1.457468138817139e-07, "loss": 0.2204, "step": 6799 }, { "epoch": 1.8094731240021287, "grad_norm": 0.2866973876953125, "learning_rate": 1.457316422270294e-07, "loss": 0.2056, "step": 6800 }, { "epoch": 1.8097392229909528, "grad_norm": 0.30119842290878296, "learning_rate": 1.457164692412155e-07, "loss": 0.2121, "step": 6801 }, { "epoch": 1.8100053219797765, "grad_norm": 0.2616921067237854, "learning_rate": 1.457012949247139e-07, "loss": 0.186, "step": 6802 }, { "epoch": 1.8102714209686002, "grad_norm": 0.3263673186302185, "learning_rate": 1.4568611927796628e-07, "loss": 0.2425, "step": 6803 }, { "epoch": 1.8105375199574243, "grad_norm": 0.2812100946903229, "learning_rate": 1.4567094230141438e-07, "loss": 0.2302, "step": 6804 }, { "epoch": 1.810803618946248, "grad_norm": 0.2853187322616577, "learning_rate": 1.456557639954999e-07, "loss": 0.1929, "step": 6805 }, { "epoch": 1.8110697179350719, "grad_norm": 0.2873852252960205, "learning_rate": 1.4564058436066467e-07, "loss": 0.2088, "step": 6806 }, { "epoch": 1.8113358169238958, "grad_norm": 0.29452699422836304, "learning_rate": 1.4562540339735056e-07, "loss": 0.1852, "step": 6807 }, { "epoch": 1.8116019159127195, "grad_norm": 0.2853315472602844, "learning_rate": 1.456102211059994e-07, "loss": 0.207, "step": 6808 }, { "epoch": 1.8118680149015434, "grad_norm": 0.2868593633174896, "learning_rate": 1.455950374870531e-07, "loss": 0.1913, "step": 6809 }, { "epoch": 1.8121341138903673, "grad_norm": 0.30469396710395813, "learning_rate": 1.4557985254095368e-07, "loss": 0.2088, "step": 6810 }, { "epoch": 1.812400212879191, "grad_norm": 0.2862718403339386, "learning_rate": 1.455646662681431e-07, "loss": 0.2037, "step": 6811 }, { "epoch": 1.8126663118680149, "grad_norm": 0.29411906003952026, "learning_rate": 1.4554947866906337e-07, "loss": 0.2294, "step": 6812 }, { "epoch": 1.8129324108568388, "grad_norm": 0.29770490527153015, "learning_rate": 1.4553428974415661e-07, "loss": 0.1941, "step": 6813 }, { "epoch": 1.8131985098456624, "grad_norm": 0.27961182594299316, "learning_rate": 1.4551909949386493e-07, "loss": 0.1951, "step": 6814 }, { "epoch": 1.8134646088344866, "grad_norm": 0.29779931902885437, "learning_rate": 1.4550390791863043e-07, "loss": 0.2151, "step": 6815 }, { "epoch": 1.8137307078233103, "grad_norm": 0.2850624918937683, "learning_rate": 1.454887150188953e-07, "loss": 0.1975, "step": 6816 }, { "epoch": 1.8139968068121342, "grad_norm": 0.28168150782585144, "learning_rate": 1.454735207951018e-07, "loss": 0.1846, "step": 6817 }, { "epoch": 1.814262905800958, "grad_norm": 0.3002559244632721, "learning_rate": 1.4545832524769217e-07, "loss": 0.2321, "step": 6818 }, { "epoch": 1.8145290047897817, "grad_norm": 0.26543205976486206, "learning_rate": 1.4544312837710873e-07, "loss": 0.1884, "step": 6819 }, { "epoch": 1.8147951037786056, "grad_norm": 0.2746518850326538, "learning_rate": 1.454279301837938e-07, "loss": 0.1906, "step": 6820 }, { "epoch": 1.8150612027674295, "grad_norm": 0.299463152885437, "learning_rate": 1.454127306681898e-07, "loss": 0.2124, "step": 6821 }, { "epoch": 1.8153273017562532, "grad_norm": 0.2954133152961731, "learning_rate": 1.453975298307391e-07, "loss": 0.2044, "step": 6822 }, { "epoch": 1.8155934007450771, "grad_norm": 0.3900117576122284, "learning_rate": 1.453823276718842e-07, "loss": 0.2105, "step": 6823 }, { "epoch": 1.815859499733901, "grad_norm": 0.2824835777282715, "learning_rate": 1.4536712419206759e-07, "loss": 0.2036, "step": 6824 }, { "epoch": 1.8161255987227247, "grad_norm": 0.33367207646369934, "learning_rate": 1.4535191939173176e-07, "loss": 0.2032, "step": 6825 }, { "epoch": 1.8163916977115488, "grad_norm": 0.26977017521858215, "learning_rate": 1.453367132713193e-07, "loss": 0.1881, "step": 6826 }, { "epoch": 1.8166577967003725, "grad_norm": 0.2715953290462494, "learning_rate": 1.453215058312729e-07, "loss": 0.1955, "step": 6827 }, { "epoch": 1.8169238956891964, "grad_norm": 0.28824445605278015, "learning_rate": 1.4530629707203508e-07, "loss": 0.1902, "step": 6828 }, { "epoch": 1.8171899946780203, "grad_norm": 0.30093884468078613, "learning_rate": 1.4529108699404859e-07, "loss": 0.2198, "step": 6829 }, { "epoch": 1.817456093666844, "grad_norm": 0.26912921667099, "learning_rate": 1.4527587559775616e-07, "loss": 0.2036, "step": 6830 }, { "epoch": 1.817722192655668, "grad_norm": 0.29054704308509827, "learning_rate": 1.4526066288360056e-07, "loss": 0.223, "step": 6831 }, { "epoch": 1.8179882916444918, "grad_norm": 0.27229002118110657, "learning_rate": 1.4524544885202458e-07, "loss": 0.1932, "step": 6832 }, { "epoch": 1.8182543906333155, "grad_norm": 0.2968658208847046, "learning_rate": 1.4523023350347105e-07, "loss": 0.2043, "step": 6833 }, { "epoch": 1.8185204896221394, "grad_norm": 0.30479514598846436, "learning_rate": 1.4521501683838286e-07, "loss": 0.2276, "step": 6834 }, { "epoch": 1.8187865886109633, "grad_norm": 0.2856389880180359, "learning_rate": 1.4519979885720297e-07, "loss": 0.2007, "step": 6835 }, { "epoch": 1.819052687599787, "grad_norm": 0.2749333381652832, "learning_rate": 1.4518457956037428e-07, "loss": 0.202, "step": 6836 }, { "epoch": 1.8193187865886111, "grad_norm": 0.27103716135025024, "learning_rate": 1.4516935894833978e-07, "loss": 0.2195, "step": 6837 }, { "epoch": 1.8195848855774348, "grad_norm": 0.28973984718322754, "learning_rate": 1.451541370215425e-07, "loss": 0.219, "step": 6838 }, { "epoch": 1.8198509845662585, "grad_norm": 0.27622565627098083, "learning_rate": 1.4513891378042557e-07, "loss": 0.2157, "step": 6839 }, { "epoch": 1.8201170835550826, "grad_norm": 0.28525158762931824, "learning_rate": 1.4512368922543208e-07, "loss": 0.214, "step": 6840 }, { "epoch": 1.8203831825439063, "grad_norm": 0.29558926820755005, "learning_rate": 1.4510846335700513e-07, "loss": 0.2139, "step": 6841 }, { "epoch": 1.8206492815327302, "grad_norm": 0.28041964769363403, "learning_rate": 1.4509323617558794e-07, "loss": 0.1983, "step": 6842 }, { "epoch": 1.820915380521554, "grad_norm": 0.26480886340141296, "learning_rate": 1.4507800768162371e-07, "loss": 0.1832, "step": 6843 }, { "epoch": 1.8211814795103778, "grad_norm": 0.27637961506843567, "learning_rate": 1.4506277787555571e-07, "loss": 0.1968, "step": 6844 }, { "epoch": 1.8214475784992017, "grad_norm": 0.28292277455329895, "learning_rate": 1.4504754675782729e-07, "loss": 0.1862, "step": 6845 }, { "epoch": 1.8217136774880256, "grad_norm": 0.2902821898460388, "learning_rate": 1.4503231432888177e-07, "loss": 0.1873, "step": 6846 }, { "epoch": 1.8219797764768493, "grad_norm": 0.2948130965232849, "learning_rate": 1.4501708058916246e-07, "loss": 0.2175, "step": 6847 }, { "epoch": 1.8222458754656732, "grad_norm": 0.29724863171577454, "learning_rate": 1.4500184553911282e-07, "loss": 0.1961, "step": 6848 }, { "epoch": 1.822511974454497, "grad_norm": 0.29333987832069397, "learning_rate": 1.449866091791763e-07, "loss": 0.1943, "step": 6849 }, { "epoch": 1.8227780734433208, "grad_norm": 0.3011998236179352, "learning_rate": 1.4497137150979642e-07, "loss": 0.2196, "step": 6850 }, { "epoch": 1.823044172432145, "grad_norm": 0.27223607897758484, "learning_rate": 1.4495613253141666e-07, "loss": 0.1763, "step": 6851 }, { "epoch": 1.8233102714209686, "grad_norm": 0.2958657145500183, "learning_rate": 1.449408922444806e-07, "loss": 0.2216, "step": 6852 }, { "epoch": 1.8235763704097925, "grad_norm": 0.30742985010147095, "learning_rate": 1.4492565064943187e-07, "loss": 0.2156, "step": 6853 }, { "epoch": 1.8238424693986164, "grad_norm": 0.2908937931060791, "learning_rate": 1.4491040774671414e-07, "loss": 0.202, "step": 6854 }, { "epoch": 1.82410856838744, "grad_norm": 0.3431652784347534, "learning_rate": 1.4489516353677098e-07, "loss": 0.2138, "step": 6855 }, { "epoch": 1.824374667376264, "grad_norm": 0.2794860601425171, "learning_rate": 1.4487991802004623e-07, "loss": 0.1812, "step": 6856 }, { "epoch": 1.8246407663650879, "grad_norm": 0.30042892694473267, "learning_rate": 1.4486467119698357e-07, "loss": 0.2145, "step": 6857 }, { "epoch": 1.8249068653539116, "grad_norm": 0.3047049939632416, "learning_rate": 1.4484942306802683e-07, "loss": 0.2188, "step": 6858 }, { "epoch": 1.8251729643427355, "grad_norm": 0.26920342445373535, "learning_rate": 1.4483417363361982e-07, "loss": 0.1901, "step": 6859 }, { "epoch": 1.8254390633315594, "grad_norm": 0.2963099777698517, "learning_rate": 1.4481892289420644e-07, "loss": 0.2164, "step": 6860 }, { "epoch": 1.825705162320383, "grad_norm": 0.2761651277542114, "learning_rate": 1.4480367085023057e-07, "loss": 0.2222, "step": 6861 }, { "epoch": 1.8259712613092072, "grad_norm": 0.2850273847579956, "learning_rate": 1.4478841750213618e-07, "loss": 0.2093, "step": 6862 }, { "epoch": 1.8262373602980309, "grad_norm": 0.2876419723033905, "learning_rate": 1.4477316285036727e-07, "loss": 0.1853, "step": 6863 }, { "epoch": 1.8265034592868548, "grad_norm": 0.29253295063972473, "learning_rate": 1.447579068953678e-07, "loss": 0.202, "step": 6864 }, { "epoch": 1.8267695582756787, "grad_norm": 0.2960905432701111, "learning_rate": 1.4474264963758188e-07, "loss": 0.2025, "step": 6865 }, { "epoch": 1.8270356572645023, "grad_norm": 0.30383360385894775, "learning_rate": 1.4472739107745366e-07, "loss": 0.2091, "step": 6866 }, { "epoch": 1.8273017562533262, "grad_norm": 0.29150986671447754, "learning_rate": 1.447121312154272e-07, "loss": 0.192, "step": 6867 }, { "epoch": 1.8275678552421502, "grad_norm": 0.28551700711250305, "learning_rate": 1.4469687005194667e-07, "loss": 0.2009, "step": 6868 }, { "epoch": 1.8278339542309738, "grad_norm": 0.2777537703514099, "learning_rate": 1.446816075874563e-07, "loss": 0.1918, "step": 6869 }, { "epoch": 1.8281000532197977, "grad_norm": 0.25103265047073364, "learning_rate": 1.4466634382240034e-07, "loss": 0.1992, "step": 6870 }, { "epoch": 1.8283661522086216, "grad_norm": 0.28717175126075745, "learning_rate": 1.4465107875722308e-07, "loss": 0.2085, "step": 6871 }, { "epoch": 1.8286322511974453, "grad_norm": 0.3179858326911926, "learning_rate": 1.446358123923689e-07, "loss": 0.2439, "step": 6872 }, { "epoch": 1.8288983501862695, "grad_norm": 0.29288026690483093, "learning_rate": 1.446205447282821e-07, "loss": 0.2018, "step": 6873 }, { "epoch": 1.8291644491750931, "grad_norm": 0.2423373907804489, "learning_rate": 1.446052757654071e-07, "loss": 0.176, "step": 6874 }, { "epoch": 1.8294305481639168, "grad_norm": 0.30064770579338074, "learning_rate": 1.4459000550418833e-07, "loss": 0.2179, "step": 6875 }, { "epoch": 1.829696647152741, "grad_norm": 0.27191153168678284, "learning_rate": 1.4457473394507028e-07, "loss": 0.1932, "step": 6876 }, { "epoch": 1.8299627461415646, "grad_norm": 0.27949875593185425, "learning_rate": 1.4455946108849747e-07, "loss": 0.2051, "step": 6877 }, { "epoch": 1.8302288451303885, "grad_norm": 0.27204498648643494, "learning_rate": 1.4454418693491443e-07, "loss": 0.2061, "step": 6878 }, { "epoch": 1.8304949441192124, "grad_norm": 0.2617051899433136, "learning_rate": 1.4452891148476575e-07, "loss": 0.1833, "step": 6879 }, { "epoch": 1.8307610431080361, "grad_norm": 0.2849035859107971, "learning_rate": 1.445136347384961e-07, "loss": 0.197, "step": 6880 }, { "epoch": 1.83102714209686, "grad_norm": 0.2783347964286804, "learning_rate": 1.4449835669655013e-07, "loss": 0.1998, "step": 6881 }, { "epoch": 1.831293241085684, "grad_norm": 0.29332101345062256, "learning_rate": 1.4448307735937252e-07, "loss": 0.2097, "step": 6882 }, { "epoch": 1.8315593400745076, "grad_norm": 0.3097696900367737, "learning_rate": 1.4446779672740804e-07, "loss": 0.219, "step": 6883 }, { "epoch": 1.8318254390633315, "grad_norm": 0.27640897035598755, "learning_rate": 1.4445251480110143e-07, "loss": 0.1873, "step": 6884 }, { "epoch": 1.8320915380521554, "grad_norm": 0.2960611879825592, "learning_rate": 1.4443723158089758e-07, "loss": 0.2092, "step": 6885 }, { "epoch": 1.832357637040979, "grad_norm": 0.2978258728981018, "learning_rate": 1.4442194706724125e-07, "loss": 0.2266, "step": 6886 }, { "epoch": 1.8326237360298032, "grad_norm": 0.28840744495391846, "learning_rate": 1.4440666126057742e-07, "loss": 0.2208, "step": 6887 }, { "epoch": 1.832889835018627, "grad_norm": 0.2904018461704254, "learning_rate": 1.4439137416135095e-07, "loss": 0.2252, "step": 6888 }, { "epoch": 1.8331559340074508, "grad_norm": 0.27006855607032776, "learning_rate": 1.4437608577000683e-07, "loss": 0.2056, "step": 6889 }, { "epoch": 1.8334220329962747, "grad_norm": 0.30295291543006897, "learning_rate": 1.4436079608699016e-07, "loss": 0.2253, "step": 6890 }, { "epoch": 1.8336881319850984, "grad_norm": 0.2876354455947876, "learning_rate": 1.443455051127458e-07, "loss": 0.1986, "step": 6891 }, { "epoch": 1.8339542309739223, "grad_norm": 0.272317111492157, "learning_rate": 1.4433021284771896e-07, "loss": 0.2001, "step": 6892 }, { "epoch": 1.8342203299627462, "grad_norm": 0.27053236961364746, "learning_rate": 1.4431491929235472e-07, "loss": 0.1841, "step": 6893 }, { "epoch": 1.8344864289515699, "grad_norm": 0.2956564426422119, "learning_rate": 1.4429962444709827e-07, "loss": 0.2135, "step": 6894 }, { "epoch": 1.8347525279403938, "grad_norm": 0.2924056947231293, "learning_rate": 1.4428432831239475e-07, "loss": 0.2154, "step": 6895 }, { "epoch": 1.8350186269292177, "grad_norm": 0.27750933170318604, "learning_rate": 1.442690308886894e-07, "loss": 0.1971, "step": 6896 }, { "epoch": 1.8352847259180414, "grad_norm": 0.2821102440357208, "learning_rate": 1.4425373217642754e-07, "loss": 0.2003, "step": 6897 }, { "epoch": 1.8355508249068655, "grad_norm": 0.3066467344760895, "learning_rate": 1.4423843217605442e-07, "loss": 0.2046, "step": 6898 }, { "epoch": 1.8358169238956892, "grad_norm": 0.2647549510002136, "learning_rate": 1.4422313088801542e-07, "loss": 0.1814, "step": 6899 }, { "epoch": 1.836083022884513, "grad_norm": 0.32884278893470764, "learning_rate": 1.442078283127559e-07, "loss": 0.2231, "step": 6900 }, { "epoch": 1.836349121873337, "grad_norm": 0.28343358635902405, "learning_rate": 1.4419252445072126e-07, "loss": 0.2127, "step": 6901 }, { "epoch": 1.8366152208621607, "grad_norm": 0.30831050872802734, "learning_rate": 1.44177219302357e-07, "loss": 0.1984, "step": 6902 }, { "epoch": 1.8368813198509846, "grad_norm": 0.2868627905845642, "learning_rate": 1.441619128681086e-07, "loss": 0.2077, "step": 6903 }, { "epoch": 1.8371474188398085, "grad_norm": 0.27846941351890564, "learning_rate": 1.4414660514842156e-07, "loss": 0.1972, "step": 6904 }, { "epoch": 1.8374135178286322, "grad_norm": 0.29835766553878784, "learning_rate": 1.4413129614374146e-07, "loss": 0.1897, "step": 6905 }, { "epoch": 1.837679616817456, "grad_norm": 0.2808195948600769, "learning_rate": 1.4411598585451393e-07, "loss": 0.206, "step": 6906 }, { "epoch": 1.83794571580628, "grad_norm": 0.26818767189979553, "learning_rate": 1.441006742811846e-07, "loss": 0.1879, "step": 6907 }, { "epoch": 1.8382118147951036, "grad_norm": 0.30307191610336304, "learning_rate": 1.4408536142419915e-07, "loss": 0.2351, "step": 6908 }, { "epoch": 1.8384779137839278, "grad_norm": 0.32369959354400635, "learning_rate": 1.4407004728400332e-07, "loss": 0.2106, "step": 6909 }, { "epoch": 1.8387440127727515, "grad_norm": 0.2652294933795929, "learning_rate": 1.440547318610428e-07, "loss": 0.1824, "step": 6910 }, { "epoch": 1.8390101117615754, "grad_norm": 0.2723790109157562, "learning_rate": 1.4403941515576344e-07, "loss": 0.2045, "step": 6911 }, { "epoch": 1.8392762107503993, "grad_norm": 0.27731090784072876, "learning_rate": 1.4402409716861104e-07, "loss": 0.1994, "step": 6912 }, { "epoch": 1.839542309739223, "grad_norm": 0.31668227910995483, "learning_rate": 1.4400877790003147e-07, "loss": 0.2116, "step": 6913 }, { "epoch": 1.8398084087280469, "grad_norm": 0.32993754744529724, "learning_rate": 1.4399345735047066e-07, "loss": 0.2206, "step": 6914 }, { "epoch": 1.8400745077168708, "grad_norm": 0.29410019516944885, "learning_rate": 1.4397813552037454e-07, "loss": 0.2036, "step": 6915 }, { "epoch": 1.8403406067056944, "grad_norm": 0.2894151210784912, "learning_rate": 1.4396281241018908e-07, "loss": 0.2128, "step": 6916 }, { "epoch": 1.8406067056945183, "grad_norm": 0.3126409649848938, "learning_rate": 1.439474880203603e-07, "loss": 0.2174, "step": 6917 }, { "epoch": 1.8408728046833422, "grad_norm": 0.272018700838089, "learning_rate": 1.4393216235133426e-07, "loss": 0.1909, "step": 6918 }, { "epoch": 1.841138903672166, "grad_norm": 0.2886989712715149, "learning_rate": 1.43916835403557e-07, "loss": 0.2059, "step": 6919 }, { "epoch": 1.84140500266099, "grad_norm": 0.2901458144187927, "learning_rate": 1.4390150717747474e-07, "loss": 0.2097, "step": 6920 }, { "epoch": 1.8416711016498137, "grad_norm": 0.29293978214263916, "learning_rate": 1.4388617767353358e-07, "loss": 0.2082, "step": 6921 }, { "epoch": 1.8419372006386374, "grad_norm": 0.28938642144203186, "learning_rate": 1.4387084689217973e-07, "loss": 0.1962, "step": 6922 }, { "epoch": 1.8422032996274615, "grad_norm": 0.3218437731266022, "learning_rate": 1.438555148338594e-07, "loss": 0.2352, "step": 6923 }, { "epoch": 1.8424693986162852, "grad_norm": 0.28342553973197937, "learning_rate": 1.4384018149901897e-07, "loss": 0.2002, "step": 6924 }, { "epoch": 1.8427354976051091, "grad_norm": 0.28452733159065247, "learning_rate": 1.4382484688810465e-07, "loss": 0.2291, "step": 6925 }, { "epoch": 1.843001596593933, "grad_norm": 0.30808836221694946, "learning_rate": 1.4380951100156284e-07, "loss": 0.2131, "step": 6926 }, { "epoch": 1.8432676955827567, "grad_norm": 0.2664114534854889, "learning_rate": 1.4379417383983993e-07, "loss": 0.1883, "step": 6927 }, { "epoch": 1.8435337945715806, "grad_norm": 0.2635791301727295, "learning_rate": 1.437788354033823e-07, "loss": 0.1788, "step": 6928 }, { "epoch": 1.8437998935604045, "grad_norm": 0.2946169078350067, "learning_rate": 1.4376349569263645e-07, "loss": 0.2066, "step": 6929 }, { "epoch": 1.8440659925492282, "grad_norm": 0.29182693362236023, "learning_rate": 1.437481547080489e-07, "loss": 0.2169, "step": 6930 }, { "epoch": 1.844332091538052, "grad_norm": 0.28100043535232544, "learning_rate": 1.4373281245006613e-07, "loss": 0.2187, "step": 6931 }, { "epoch": 1.844598190526876, "grad_norm": 0.31950676441192627, "learning_rate": 1.4371746891913471e-07, "loss": 0.2148, "step": 6932 }, { "epoch": 1.8448642895156997, "grad_norm": 0.3012053370475769, "learning_rate": 1.4370212411570134e-07, "loss": 0.2093, "step": 6933 }, { "epoch": 1.8451303885045238, "grad_norm": 0.2950417399406433, "learning_rate": 1.4368677804021257e-07, "loss": 0.2052, "step": 6934 }, { "epoch": 1.8453964874933475, "grad_norm": 0.34505707025527954, "learning_rate": 1.4367143069311514e-07, "loss": 0.2046, "step": 6935 }, { "epoch": 1.8456625864821714, "grad_norm": 0.2540322542190552, "learning_rate": 1.4365608207485572e-07, "loss": 0.1756, "step": 6936 }, { "epoch": 1.8459286854709953, "grad_norm": 0.300921767950058, "learning_rate": 1.4364073218588116e-07, "loss": 0.2303, "step": 6937 }, { "epoch": 1.846194784459819, "grad_norm": 0.2719704508781433, "learning_rate": 1.4362538102663818e-07, "loss": 0.1893, "step": 6938 }, { "epoch": 1.846460883448643, "grad_norm": 0.2692101299762726, "learning_rate": 1.4361002859757361e-07, "loss": 0.2085, "step": 6939 }, { "epoch": 1.8467269824374668, "grad_norm": 0.28402453660964966, "learning_rate": 1.4359467489913436e-07, "loss": 0.2034, "step": 6940 }, { "epoch": 1.8469930814262905, "grad_norm": 0.2567652463912964, "learning_rate": 1.4357931993176728e-07, "loss": 0.1914, "step": 6941 }, { "epoch": 1.8472591804151144, "grad_norm": 0.31407180428504944, "learning_rate": 1.4356396369591938e-07, "loss": 0.1986, "step": 6942 }, { "epoch": 1.8475252794039383, "grad_norm": 0.27136415243148804, "learning_rate": 1.435486061920376e-07, "loss": 0.2007, "step": 6943 }, { "epoch": 1.847791378392762, "grad_norm": 0.3010987937450409, "learning_rate": 1.4353324742056897e-07, "loss": 0.2098, "step": 6944 }, { "epoch": 1.848057477381586, "grad_norm": 0.29047414660453796, "learning_rate": 1.4351788738196055e-07, "loss": 0.211, "step": 6945 }, { "epoch": 1.8483235763704098, "grad_norm": 0.2682918906211853, "learning_rate": 1.435025260766594e-07, "loss": 0.1846, "step": 6946 }, { "epoch": 1.8485896753592337, "grad_norm": 0.2758488059043884, "learning_rate": 1.434871635051127e-07, "loss": 0.2059, "step": 6947 }, { "epoch": 1.8488557743480576, "grad_norm": 0.2894424796104431, "learning_rate": 1.434717996677676e-07, "loss": 0.1872, "step": 6948 }, { "epoch": 1.8491218733368813, "grad_norm": 0.3077656328678131, "learning_rate": 1.4345643456507124e-07, "loss": 0.2085, "step": 6949 }, { "epoch": 1.8493879723257052, "grad_norm": 0.3063506484031677, "learning_rate": 1.4344106819747088e-07, "loss": 0.188, "step": 6950 }, { "epoch": 1.849654071314529, "grad_norm": 0.2848341464996338, "learning_rate": 1.4342570056541384e-07, "loss": 0.2042, "step": 6951 }, { "epoch": 1.8499201703033528, "grad_norm": 0.3079037368297577, "learning_rate": 1.4341033166934745e-07, "loss": 0.23, "step": 6952 }, { "epoch": 1.8501862692921767, "grad_norm": 0.28647565841674805, "learning_rate": 1.43394961509719e-07, "loss": 0.2007, "step": 6953 }, { "epoch": 1.8504523682810006, "grad_norm": 0.26077941060066223, "learning_rate": 1.4337959008697586e-07, "loss": 0.1863, "step": 6954 }, { "epoch": 1.8507184672698243, "grad_norm": 0.33528995513916016, "learning_rate": 1.4336421740156551e-07, "loss": 0.219, "step": 6955 }, { "epoch": 1.8509845662586484, "grad_norm": 0.31442710757255554, "learning_rate": 1.433488434539354e-07, "loss": 0.1975, "step": 6956 }, { "epoch": 1.851250665247472, "grad_norm": 0.2746850550174713, "learning_rate": 1.4333346824453302e-07, "loss": 0.1849, "step": 6957 }, { "epoch": 1.8515167642362957, "grad_norm": 0.2901971638202667, "learning_rate": 1.4331809177380586e-07, "loss": 0.1929, "step": 6958 }, { "epoch": 1.8517828632251199, "grad_norm": 0.30731844902038574, "learning_rate": 1.4330271404220153e-07, "loss": 0.2228, "step": 6959 }, { "epoch": 1.8520489622139436, "grad_norm": 0.3402206003665924, "learning_rate": 1.4328733505016765e-07, "loss": 0.2203, "step": 6960 }, { "epoch": 1.8523150612027675, "grad_norm": 0.2808026373386383, "learning_rate": 1.432719547981518e-07, "loss": 0.1991, "step": 6961 }, { "epoch": 1.8525811601915914, "grad_norm": 0.28883180022239685, "learning_rate": 1.4325657328660173e-07, "loss": 0.2063, "step": 6962 }, { "epoch": 1.852847259180415, "grad_norm": 0.2915298640727997, "learning_rate": 1.4324119051596512e-07, "loss": 0.2162, "step": 6963 }, { "epoch": 1.853113358169239, "grad_norm": 0.28902217745780945, "learning_rate": 1.4322580648668973e-07, "loss": 0.2159, "step": 6964 }, { "epoch": 1.8533794571580628, "grad_norm": 0.30927184224128723, "learning_rate": 1.4321042119922336e-07, "loss": 0.2172, "step": 6965 }, { "epoch": 1.8536455561468865, "grad_norm": 0.274293452501297, "learning_rate": 1.431950346540138e-07, "loss": 0.2021, "step": 6966 }, { "epoch": 1.8539116551357104, "grad_norm": 0.2728838622570038, "learning_rate": 1.4317964685150896e-07, "loss": 0.1961, "step": 6967 }, { "epoch": 1.8541777541245343, "grad_norm": 0.30076533555984497, "learning_rate": 1.431642577921567e-07, "loss": 0.2128, "step": 6968 }, { "epoch": 1.854443853113358, "grad_norm": 0.27995043992996216, "learning_rate": 1.4314886747640497e-07, "loss": 0.1857, "step": 6969 }, { "epoch": 1.8547099521021821, "grad_norm": 0.31418630480766296, "learning_rate": 1.4313347590470175e-07, "loss": 0.1996, "step": 6970 }, { "epoch": 1.8549760510910058, "grad_norm": 0.28267213702201843, "learning_rate": 1.4311808307749505e-07, "loss": 0.2027, "step": 6971 }, { "epoch": 1.8552421500798297, "grad_norm": 0.2873254716396332, "learning_rate": 1.431026889952329e-07, "loss": 0.206, "step": 6972 }, { "epoch": 1.8555082490686536, "grad_norm": 0.28712740540504456, "learning_rate": 1.4308729365836337e-07, "loss": 0.1973, "step": 6973 }, { "epoch": 1.8557743480574773, "grad_norm": 0.27747252583503723, "learning_rate": 1.430718970673346e-07, "loss": 0.2023, "step": 6974 }, { "epoch": 1.8560404470463012, "grad_norm": 0.27060139179229736, "learning_rate": 1.4305649922259474e-07, "loss": 0.1944, "step": 6975 }, { "epoch": 1.8563065460351251, "grad_norm": 0.2873412072658539, "learning_rate": 1.4304110012459202e-07, "loss": 0.2042, "step": 6976 }, { "epoch": 1.8565726450239488, "grad_norm": 0.2809629440307617, "learning_rate": 1.430256997737746e-07, "loss": 0.2142, "step": 6977 }, { "epoch": 1.8568387440127727, "grad_norm": 0.30419328808784485, "learning_rate": 1.4301029817059077e-07, "loss": 0.1987, "step": 6978 }, { "epoch": 1.8571048430015966, "grad_norm": 0.28039824962615967, "learning_rate": 1.4299489531548886e-07, "loss": 0.2029, "step": 6979 }, { "epoch": 1.8573709419904203, "grad_norm": 0.3042882978916168, "learning_rate": 1.4297949120891716e-07, "loss": 0.1989, "step": 6980 }, { "epoch": 1.8576370409792444, "grad_norm": 0.4383663833141327, "learning_rate": 1.429640858513241e-07, "loss": 0.2151, "step": 6981 }, { "epoch": 1.857903139968068, "grad_norm": 0.27912941575050354, "learning_rate": 1.42948679243158e-07, "loss": 0.1857, "step": 6982 }, { "epoch": 1.858169238956892, "grad_norm": 0.2630625367164612, "learning_rate": 1.429332713848674e-07, "loss": 0.198, "step": 6983 }, { "epoch": 1.858435337945716, "grad_norm": 0.2924792170524597, "learning_rate": 1.4291786227690072e-07, "loss": 0.2133, "step": 6984 }, { "epoch": 1.8587014369345396, "grad_norm": 0.30433782935142517, "learning_rate": 1.4290245191970653e-07, "loss": 0.212, "step": 6985 }, { "epoch": 1.8589675359233635, "grad_norm": 0.32596734166145325, "learning_rate": 1.4288704031373334e-07, "loss": 0.2418, "step": 6986 }, { "epoch": 1.8592336349121874, "grad_norm": 0.2901002764701843, "learning_rate": 1.4287162745942975e-07, "loss": 0.2245, "step": 6987 }, { "epoch": 1.859499733901011, "grad_norm": 0.3079976737499237, "learning_rate": 1.4285621335724445e-07, "loss": 0.2026, "step": 6988 }, { "epoch": 1.859765832889835, "grad_norm": 0.26880159974098206, "learning_rate": 1.4284079800762606e-07, "loss": 0.1973, "step": 6989 }, { "epoch": 1.860031931878659, "grad_norm": 0.3099972903728485, "learning_rate": 1.4282538141102323e-07, "loss": 0.1975, "step": 6990 }, { "epoch": 1.8602980308674826, "grad_norm": 0.3184162974357605, "learning_rate": 1.4280996356788473e-07, "loss": 0.2135, "step": 6991 }, { "epoch": 1.8605641298563067, "grad_norm": 0.3026188015937805, "learning_rate": 1.4279454447865934e-07, "loss": 0.2138, "step": 6992 }, { "epoch": 1.8608302288451304, "grad_norm": 0.2773163318634033, "learning_rate": 1.4277912414379592e-07, "loss": 0.1954, "step": 6993 }, { "epoch": 1.861096327833954, "grad_norm": 0.27684441208839417, "learning_rate": 1.4276370256374325e-07, "loss": 0.1914, "step": 6994 }, { "epoch": 1.8613624268227782, "grad_norm": 0.2986638844013214, "learning_rate": 1.4274827973895025e-07, "loss": 0.1983, "step": 6995 }, { "epoch": 1.8616285258116019, "grad_norm": 0.26772454380989075, "learning_rate": 1.427328556698658e-07, "loss": 0.1796, "step": 6996 }, { "epoch": 1.8618946248004258, "grad_norm": 0.294193834066391, "learning_rate": 1.427174303569389e-07, "loss": 0.2046, "step": 6997 }, { "epoch": 1.8621607237892497, "grad_norm": 0.277969628572464, "learning_rate": 1.427020038006185e-07, "loss": 0.2111, "step": 6998 }, { "epoch": 1.8624268227780734, "grad_norm": 0.2805419862270355, "learning_rate": 1.4268657600135365e-07, "loss": 0.2048, "step": 6999 }, { "epoch": 1.8626929217668973, "grad_norm": 0.30437391996383667, "learning_rate": 1.426711469595934e-07, "loss": 0.1865, "step": 7000 }, { "epoch": 1.8629590207557212, "grad_norm": 0.2729874849319458, "learning_rate": 1.4265571667578687e-07, "loss": 0.173, "step": 7001 }, { "epoch": 1.8632251197445449, "grad_norm": 0.8593712449073792, "learning_rate": 1.4264028515038314e-07, "loss": 0.1863, "step": 7002 }, { "epoch": 1.8634912187333688, "grad_norm": 0.278646320104599, "learning_rate": 1.4262485238383149e-07, "loss": 0.2008, "step": 7003 }, { "epoch": 1.8637573177221927, "grad_norm": 0.30354541540145874, "learning_rate": 1.42609418376581e-07, "loss": 0.2005, "step": 7004 }, { "epoch": 1.8640234167110163, "grad_norm": 0.285498708486557, "learning_rate": 1.42593983129081e-07, "loss": 0.2006, "step": 7005 }, { "epoch": 1.8642895156998405, "grad_norm": 0.3026873767375946, "learning_rate": 1.4257854664178076e-07, "loss": 0.2004, "step": 7006 }, { "epoch": 1.8645556146886642, "grad_norm": 0.2797206938266754, "learning_rate": 1.4256310891512958e-07, "loss": 0.1894, "step": 7007 }, { "epoch": 1.864821713677488, "grad_norm": 0.2705998718738556, "learning_rate": 1.425476699495768e-07, "loss": 0.188, "step": 7008 }, { "epoch": 1.865087812666312, "grad_norm": 0.29736706614494324, "learning_rate": 1.4253222974557183e-07, "loss": 0.2249, "step": 7009 }, { "epoch": 1.8653539116551356, "grad_norm": 0.3009137511253357, "learning_rate": 1.4251678830356407e-07, "loss": 0.2292, "step": 7010 }, { "epoch": 1.8656200106439595, "grad_norm": 0.3127601742744446, "learning_rate": 1.42501345624003e-07, "loss": 0.2072, "step": 7011 }, { "epoch": 1.8658861096327835, "grad_norm": 0.26031294465065, "learning_rate": 1.4248590170733814e-07, "loss": 0.1784, "step": 7012 }, { "epoch": 1.8661522086216071, "grad_norm": 0.26032471656799316, "learning_rate": 1.4247045655401894e-07, "loss": 0.1936, "step": 7013 }, { "epoch": 1.866418307610431, "grad_norm": 0.3059631288051605, "learning_rate": 1.4245501016449507e-07, "loss": 0.2242, "step": 7014 }, { "epoch": 1.866684406599255, "grad_norm": 0.27863726019859314, "learning_rate": 1.4243956253921606e-07, "loss": 0.1873, "step": 7015 }, { "epoch": 1.8669505055880786, "grad_norm": 0.2766591012477875, "learning_rate": 1.4242411367863162e-07, "loss": 0.1992, "step": 7016 }, { "epoch": 1.8672166045769027, "grad_norm": 0.27527642250061035, "learning_rate": 1.4240866358319135e-07, "loss": 0.1773, "step": 7017 }, { "epoch": 1.8674827035657264, "grad_norm": 0.280728280544281, "learning_rate": 1.4239321225334498e-07, "loss": 0.1931, "step": 7018 }, { "epoch": 1.8677488025545503, "grad_norm": 0.29902204871177673, "learning_rate": 1.4237775968954232e-07, "loss": 0.2066, "step": 7019 }, { "epoch": 1.8680149015433742, "grad_norm": 0.2875461280345917, "learning_rate": 1.4236230589223307e-07, "loss": 0.2069, "step": 7020 }, { "epoch": 1.868281000532198, "grad_norm": 0.3150806427001953, "learning_rate": 1.4234685086186712e-07, "loss": 0.2166, "step": 7021 }, { "epoch": 1.8685470995210218, "grad_norm": 0.26575151085853577, "learning_rate": 1.4233139459889428e-07, "loss": 0.1877, "step": 7022 }, { "epoch": 1.8688131985098457, "grad_norm": 0.283218115568161, "learning_rate": 1.4231593710376445e-07, "loss": 0.1992, "step": 7023 }, { "epoch": 1.8690792974986694, "grad_norm": 0.293899267911911, "learning_rate": 1.4230047837692753e-07, "loss": 0.2174, "step": 7024 }, { "epoch": 1.8693453964874933, "grad_norm": 0.27899858355522156, "learning_rate": 1.422850184188336e-07, "loss": 0.1904, "step": 7025 }, { "epoch": 1.8696114954763172, "grad_norm": 0.30921825766563416, "learning_rate": 1.422695572299325e-07, "loss": 0.1779, "step": 7026 }, { "epoch": 1.869877594465141, "grad_norm": 0.27724361419677734, "learning_rate": 1.4225409481067437e-07, "loss": 0.1969, "step": 7027 }, { "epoch": 1.870143693453965, "grad_norm": 0.28648683428764343, "learning_rate": 1.4223863116150927e-07, "loss": 0.2041, "step": 7028 }, { "epoch": 1.8704097924427887, "grad_norm": 0.3299691081047058, "learning_rate": 1.4222316628288723e-07, "loss": 0.2125, "step": 7029 }, { "epoch": 1.8706758914316126, "grad_norm": 0.3184742331504822, "learning_rate": 1.4220770017525852e-07, "loss": 0.2085, "step": 7030 }, { "epoch": 1.8709419904204365, "grad_norm": 0.3049008250236511, "learning_rate": 1.4219223283907322e-07, "loss": 0.2316, "step": 7031 }, { "epoch": 1.8712080894092602, "grad_norm": 0.3176620304584503, "learning_rate": 1.4217676427478155e-07, "loss": 0.1993, "step": 7032 }, { "epoch": 1.871474188398084, "grad_norm": 0.27740660309791565, "learning_rate": 1.421612944828338e-07, "loss": 0.2124, "step": 7033 }, { "epoch": 1.871740287386908, "grad_norm": 0.29658010601997375, "learning_rate": 1.4214582346368026e-07, "loss": 0.2096, "step": 7034 }, { "epoch": 1.8720063863757317, "grad_norm": 0.31170886754989624, "learning_rate": 1.421303512177712e-07, "loss": 0.2203, "step": 7035 }, { "epoch": 1.8722724853645556, "grad_norm": 0.2867373824119568, "learning_rate": 1.4211487774555703e-07, "loss": 0.1971, "step": 7036 }, { "epoch": 1.8725385843533795, "grad_norm": 0.2760554552078247, "learning_rate": 1.420994030474881e-07, "loss": 0.1948, "step": 7037 }, { "epoch": 1.8728046833422032, "grad_norm": 0.2971016764640808, "learning_rate": 1.4208392712401488e-07, "loss": 0.2027, "step": 7038 }, { "epoch": 1.8730707823310273, "grad_norm": 0.28968545794487, "learning_rate": 1.4206844997558779e-07, "loss": 0.2057, "step": 7039 }, { "epoch": 1.873336881319851, "grad_norm": 0.27217426896095276, "learning_rate": 1.4205297160265736e-07, "loss": 0.1874, "step": 7040 }, { "epoch": 1.8736029803086747, "grad_norm": 0.274387389421463, "learning_rate": 1.420374920056741e-07, "loss": 0.189, "step": 7041 }, { "epoch": 1.8738690792974988, "grad_norm": 0.30991530418395996, "learning_rate": 1.420220111850886e-07, "loss": 0.2308, "step": 7042 }, { "epoch": 1.8741351782863225, "grad_norm": 0.3187463879585266, "learning_rate": 1.420065291413515e-07, "loss": 0.2224, "step": 7043 }, { "epoch": 1.8744012772751464, "grad_norm": 0.2941374182701111, "learning_rate": 1.4199104587491338e-07, "loss": 0.1927, "step": 7044 }, { "epoch": 1.8746673762639703, "grad_norm": 0.29025885462760925, "learning_rate": 1.4197556138622492e-07, "loss": 0.2088, "step": 7045 }, { "epoch": 1.874933475252794, "grad_norm": 0.30834493041038513, "learning_rate": 1.4196007567573685e-07, "loss": 0.2061, "step": 7046 }, { "epoch": 1.8751995742416179, "grad_norm": 0.25787025690078735, "learning_rate": 1.4194458874389997e-07, "loss": 0.1899, "step": 7047 }, { "epoch": 1.8754656732304418, "grad_norm": 0.31268855929374695, "learning_rate": 1.41929100591165e-07, "loss": 0.2225, "step": 7048 }, { "epoch": 1.8757317722192655, "grad_norm": 0.2741401791572571, "learning_rate": 1.4191361121798282e-07, "loss": 0.216, "step": 7049 }, { "epoch": 1.8759978712080894, "grad_norm": 0.28784576058387756, "learning_rate": 1.418981206248042e-07, "loss": 0.1905, "step": 7050 }, { "epoch": 1.8762639701969133, "grad_norm": 0.25868093967437744, "learning_rate": 1.4188262881208008e-07, "loss": 0.2063, "step": 7051 }, { "epoch": 1.876530069185737, "grad_norm": 0.28432518243789673, "learning_rate": 1.418671357802614e-07, "loss": 0.1918, "step": 7052 }, { "epoch": 1.876796168174561, "grad_norm": 0.2907182276248932, "learning_rate": 1.4185164152979908e-07, "loss": 0.2155, "step": 7053 }, { "epoch": 1.8770622671633848, "grad_norm": 0.2867300808429718, "learning_rate": 1.4183614606114413e-07, "loss": 0.1982, "step": 7054 }, { "epoch": 1.8773283661522087, "grad_norm": 0.28865814208984375, "learning_rate": 1.4182064937474763e-07, "loss": 0.1802, "step": 7055 }, { "epoch": 1.8775944651410326, "grad_norm": 0.29515543580055237, "learning_rate": 1.4180515147106058e-07, "loss": 0.1955, "step": 7056 }, { "epoch": 1.8778605641298562, "grad_norm": 0.3094286322593689, "learning_rate": 1.4178965235053414e-07, "loss": 0.2288, "step": 7057 }, { "epoch": 1.8781266631186802, "grad_norm": 0.29239097237586975, "learning_rate": 1.4177415201361945e-07, "loss": 0.1937, "step": 7058 }, { "epoch": 1.878392762107504, "grad_norm": 0.27749720215797424, "learning_rate": 1.4175865046076765e-07, "loss": 0.2098, "step": 7059 }, { "epoch": 1.8786588610963277, "grad_norm": 0.2845059931278229, "learning_rate": 1.4174314769242993e-07, "loss": 0.2024, "step": 7060 }, { "epoch": 1.8789249600851516, "grad_norm": 0.2741910219192505, "learning_rate": 1.417276437090576e-07, "loss": 0.1893, "step": 7061 }, { "epoch": 1.8791910590739755, "grad_norm": 0.2639797031879425, "learning_rate": 1.4171213851110187e-07, "loss": 0.1831, "step": 7062 }, { "epoch": 1.8794571580627992, "grad_norm": 0.28204742074012756, "learning_rate": 1.416966320990141e-07, "loss": 0.1775, "step": 7063 }, { "epoch": 1.8797232570516234, "grad_norm": 0.29773637652397156, "learning_rate": 1.4168112447324563e-07, "loss": 0.2167, "step": 7064 }, { "epoch": 1.879989356040447, "grad_norm": 0.3016526997089386, "learning_rate": 1.4166561563424787e-07, "loss": 0.2235, "step": 7065 }, { "epoch": 1.880255455029271, "grad_norm": 0.2916228771209717, "learning_rate": 1.4165010558247222e-07, "loss": 0.2013, "step": 7066 }, { "epoch": 1.8805215540180948, "grad_norm": 0.26826199889183044, "learning_rate": 1.4163459431837014e-07, "loss": 0.1915, "step": 7067 }, { "epoch": 1.8807876530069185, "grad_norm": 0.3253234028816223, "learning_rate": 1.416190818423931e-07, "loss": 0.2266, "step": 7068 }, { "epoch": 1.8810537519957424, "grad_norm": 0.28649386763572693, "learning_rate": 1.4160356815499268e-07, "loss": 0.1975, "step": 7069 }, { "epoch": 1.8813198509845663, "grad_norm": 0.2860060930252075, "learning_rate": 1.415880532566204e-07, "loss": 0.2042, "step": 7070 }, { "epoch": 1.88158594997339, "grad_norm": 0.49899229407310486, "learning_rate": 1.4157253714772787e-07, "loss": 0.2191, "step": 7071 }, { "epoch": 1.881852048962214, "grad_norm": 0.29894527792930603, "learning_rate": 1.4155701982876673e-07, "loss": 0.2132, "step": 7072 }, { "epoch": 1.8821181479510378, "grad_norm": 0.2870078384876251, "learning_rate": 1.4154150130018864e-07, "loss": 0.201, "step": 7073 }, { "epoch": 1.8823842469398615, "grad_norm": 0.2736658751964569, "learning_rate": 1.415259815624453e-07, "loss": 0.1972, "step": 7074 }, { "epoch": 1.8826503459286856, "grad_norm": 0.29819735884666443, "learning_rate": 1.415104606159885e-07, "loss": 0.2084, "step": 7075 }, { "epoch": 1.8829164449175093, "grad_norm": 0.2777526378631592, "learning_rate": 1.4149493846126992e-07, "loss": 0.1909, "step": 7076 }, { "epoch": 1.883182543906333, "grad_norm": 0.29183319211006165, "learning_rate": 1.4147941509874145e-07, "loss": 0.218, "step": 7077 }, { "epoch": 1.8834486428951571, "grad_norm": 0.3104752004146576, "learning_rate": 1.4146389052885487e-07, "loss": 0.2014, "step": 7078 }, { "epoch": 1.8837147418839808, "grad_norm": 0.3074817955493927, "learning_rate": 1.4144836475206216e-07, "loss": 0.2321, "step": 7079 }, { "epoch": 1.8839808408728047, "grad_norm": 0.29862838983535767, "learning_rate": 1.4143283776881516e-07, "loss": 0.2101, "step": 7080 }, { "epoch": 1.8842469398616286, "grad_norm": 0.28412875533103943, "learning_rate": 1.414173095795658e-07, "loss": 0.1973, "step": 7081 }, { "epoch": 1.8845130388504523, "grad_norm": 0.2988251745700836, "learning_rate": 1.4140178018476608e-07, "loss": 0.2096, "step": 7082 }, { "epoch": 1.8847791378392762, "grad_norm": 0.30517154932022095, "learning_rate": 1.4138624958486807e-07, "loss": 0.1861, "step": 7083 }, { "epoch": 1.8850452368281, "grad_norm": 0.2760249674320221, "learning_rate": 1.4137071778032382e-07, "loss": 0.2222, "step": 7084 }, { "epoch": 1.8853113358169238, "grad_norm": 0.2950367033481598, "learning_rate": 1.4135518477158536e-07, "loss": 0.2068, "step": 7085 }, { "epoch": 1.8855774348057477, "grad_norm": 0.2874002456665039, "learning_rate": 1.4133965055910482e-07, "loss": 0.1952, "step": 7086 }, { "epoch": 1.8858435337945716, "grad_norm": 0.29495853185653687, "learning_rate": 1.4132411514333443e-07, "loss": 0.1949, "step": 7087 }, { "epoch": 1.8861096327833953, "grad_norm": 0.46178898215293884, "learning_rate": 1.4130857852472634e-07, "loss": 0.2176, "step": 7088 }, { "epoch": 1.8863757317722194, "grad_norm": 0.3064477741718292, "learning_rate": 1.412930407037328e-07, "loss": 0.2025, "step": 7089 }, { "epoch": 1.886641830761043, "grad_norm": 0.2701607942581177, "learning_rate": 1.4127750168080603e-07, "loss": 0.1957, "step": 7090 }, { "epoch": 1.886907929749867, "grad_norm": 0.30030983686447144, "learning_rate": 1.4126196145639837e-07, "loss": 0.1878, "step": 7091 }, { "epoch": 1.887174028738691, "grad_norm": 0.26624664664268494, "learning_rate": 1.4124642003096213e-07, "loss": 0.1938, "step": 7092 }, { "epoch": 1.8874401277275146, "grad_norm": 0.26498594880104065, "learning_rate": 1.4123087740494976e-07, "loss": 0.1984, "step": 7093 }, { "epoch": 1.8877062267163385, "grad_norm": 0.2720329761505127, "learning_rate": 1.4121533357881353e-07, "loss": 0.2022, "step": 7094 }, { "epoch": 1.8879723257051624, "grad_norm": 0.30741825699806213, "learning_rate": 1.41199788553006e-07, "loss": 0.2153, "step": 7095 }, { "epoch": 1.888238424693986, "grad_norm": 0.29292571544647217, "learning_rate": 1.4118424232797955e-07, "loss": 0.2019, "step": 7096 }, { "epoch": 1.88850452368281, "grad_norm": 0.2901279330253601, "learning_rate": 1.411686949041868e-07, "loss": 0.2102, "step": 7097 }, { "epoch": 1.8887706226716339, "grad_norm": 0.27817559242248535, "learning_rate": 1.411531462820802e-07, "loss": 0.2056, "step": 7098 }, { "epoch": 1.8890367216604576, "grad_norm": 0.31113341450691223, "learning_rate": 1.411375964621124e-07, "loss": 0.2123, "step": 7099 }, { "epoch": 1.8893028206492817, "grad_norm": 0.3045447766780853, "learning_rate": 1.4112204544473596e-07, "loss": 0.2147, "step": 7100 }, { "epoch": 1.8895689196381054, "grad_norm": 0.26508957147598267, "learning_rate": 1.4110649323040353e-07, "loss": 0.1868, "step": 7101 }, { "epoch": 1.8898350186269293, "grad_norm": 0.2912362515926361, "learning_rate": 1.4109093981956784e-07, "loss": 0.2047, "step": 7102 }, { "epoch": 1.8901011176157532, "grad_norm": 0.28816425800323486, "learning_rate": 1.4107538521268159e-07, "loss": 0.2208, "step": 7103 }, { "epoch": 1.8903672166045768, "grad_norm": 0.3456589877605438, "learning_rate": 1.410598294101975e-07, "loss": 0.2186, "step": 7104 }, { "epoch": 1.8906333155934008, "grad_norm": 0.2674504220485687, "learning_rate": 1.410442724125684e-07, "loss": 0.1852, "step": 7105 }, { "epoch": 1.8908994145822247, "grad_norm": 0.30098944902420044, "learning_rate": 1.410287142202471e-07, "loss": 0.2174, "step": 7106 }, { "epoch": 1.8911655135710483, "grad_norm": 0.29076093435287476, "learning_rate": 1.4101315483368646e-07, "loss": 0.2099, "step": 7107 }, { "epoch": 1.8914316125598722, "grad_norm": 0.29455405473709106, "learning_rate": 1.4099759425333933e-07, "loss": 0.2249, "step": 7108 }, { "epoch": 1.8916977115486961, "grad_norm": 0.29279831051826477, "learning_rate": 1.4098203247965874e-07, "loss": 0.203, "step": 7109 }, { "epoch": 1.8919638105375198, "grad_norm": 0.2986944913864136, "learning_rate": 1.409664695130976e-07, "loss": 0.2016, "step": 7110 }, { "epoch": 1.892229909526344, "grad_norm": 0.32278549671173096, "learning_rate": 1.4095090535410888e-07, "loss": 0.2275, "step": 7111 }, { "epoch": 1.8924960085151676, "grad_norm": 0.3080020844936371, "learning_rate": 1.4093534000314562e-07, "loss": 0.214, "step": 7112 }, { "epoch": 1.8927621075039913, "grad_norm": 0.27261272072792053, "learning_rate": 1.409197734606609e-07, "loss": 0.1958, "step": 7113 }, { "epoch": 1.8930282064928154, "grad_norm": 0.2762836217880249, "learning_rate": 1.409042057271078e-07, "loss": 0.1976, "step": 7114 }, { "epoch": 1.8932943054816391, "grad_norm": 0.32024767994880676, "learning_rate": 1.4088863680293952e-07, "loss": 0.2105, "step": 7115 }, { "epoch": 1.893560404470463, "grad_norm": 0.2744636535644531, "learning_rate": 1.408730666886092e-07, "loss": 0.1968, "step": 7116 }, { "epoch": 1.893826503459287, "grad_norm": 0.296711266040802, "learning_rate": 1.4085749538457e-07, "loss": 0.1959, "step": 7117 }, { "epoch": 1.8940926024481106, "grad_norm": 0.28611868619918823, "learning_rate": 1.408419228912752e-07, "loss": 0.1995, "step": 7118 }, { "epoch": 1.8943587014369345, "grad_norm": 0.28474530577659607, "learning_rate": 1.4082634920917808e-07, "loss": 0.1994, "step": 7119 }, { "epoch": 1.8946248004257584, "grad_norm": 0.2833541929721832, "learning_rate": 1.4081077433873192e-07, "loss": 0.1911, "step": 7120 }, { "epoch": 1.894890899414582, "grad_norm": 0.2618454098701477, "learning_rate": 1.407951982803901e-07, "loss": 0.1713, "step": 7121 }, { "epoch": 1.895156998403406, "grad_norm": 0.280443400144577, "learning_rate": 1.4077962103460598e-07, "loss": 0.1935, "step": 7122 }, { "epoch": 1.89542309739223, "grad_norm": 0.28894397616386414, "learning_rate": 1.4076404260183297e-07, "loss": 0.1846, "step": 7123 }, { "epoch": 1.8956891963810536, "grad_norm": 0.3186306059360504, "learning_rate": 1.4074846298252452e-07, "loss": 0.2081, "step": 7124 }, { "epoch": 1.8959552953698777, "grad_norm": 0.2776063084602356, "learning_rate": 1.4073288217713414e-07, "loss": 0.2074, "step": 7125 }, { "epoch": 1.8962213943587014, "grad_norm": 0.29680192470550537, "learning_rate": 1.407173001861153e-07, "loss": 0.2216, "step": 7126 }, { "epoch": 1.8964874933475253, "grad_norm": 0.3100031912326813, "learning_rate": 1.4070171700992158e-07, "loss": 0.1852, "step": 7127 }, { "epoch": 1.8967535923363492, "grad_norm": 0.28383827209472656, "learning_rate": 1.4068613264900654e-07, "loss": 0.1929, "step": 7128 }, { "epoch": 1.897019691325173, "grad_norm": 0.30340614914894104, "learning_rate": 1.4067054710382385e-07, "loss": 0.2173, "step": 7129 }, { "epoch": 1.8972857903139968, "grad_norm": 0.2799181342124939, "learning_rate": 1.4065496037482709e-07, "loss": 0.1962, "step": 7130 }, { "epoch": 1.8975518893028207, "grad_norm": 0.2938293516635895, "learning_rate": 1.4063937246247003e-07, "loss": 0.1933, "step": 7131 }, { "epoch": 1.8978179882916444, "grad_norm": 0.27898749709129333, "learning_rate": 1.4062378336720634e-07, "loss": 0.2039, "step": 7132 }, { "epoch": 1.8980840872804683, "grad_norm": 0.2969246208667755, "learning_rate": 1.4060819308948982e-07, "loss": 0.2012, "step": 7133 }, { "epoch": 1.8983501862692922, "grad_norm": 0.2817286252975464, "learning_rate": 1.405926016297742e-07, "loss": 0.1974, "step": 7134 }, { "epoch": 1.8986162852581159, "grad_norm": 0.3047667443752289, "learning_rate": 1.405770089885134e-07, "loss": 0.2068, "step": 7135 }, { "epoch": 1.89888238424694, "grad_norm": 0.27790164947509766, "learning_rate": 1.4056141516616116e-07, "loss": 0.1851, "step": 7136 }, { "epoch": 1.8991484832357637, "grad_norm": 0.2812650799751282, "learning_rate": 1.4054582016317148e-07, "loss": 0.2027, "step": 7137 }, { "epoch": 1.8994145822245876, "grad_norm": 0.2869095206260681, "learning_rate": 1.4053022397999824e-07, "loss": 0.1808, "step": 7138 }, { "epoch": 1.8996806812134115, "grad_norm": 0.25309333205223083, "learning_rate": 1.405146266170954e-07, "loss": 0.169, "step": 7139 }, { "epoch": 1.8999467802022352, "grad_norm": 0.35812628269195557, "learning_rate": 1.4049902807491704e-07, "loss": 0.2032, "step": 7140 }, { "epoch": 1.900212879191059, "grad_norm": 0.2825220823287964, "learning_rate": 1.4048342835391704e-07, "loss": 0.2112, "step": 7141 }, { "epoch": 1.900478978179883, "grad_norm": 0.29469797015190125, "learning_rate": 1.4046782745454958e-07, "loss": 0.2127, "step": 7142 }, { "epoch": 1.9007450771687067, "grad_norm": 0.3010563254356384, "learning_rate": 1.4045222537726878e-07, "loss": 0.2143, "step": 7143 }, { "epoch": 1.9010111761575306, "grad_norm": 0.31587645411491394, "learning_rate": 1.404366221225287e-07, "loss": 0.2234, "step": 7144 }, { "epoch": 1.9012772751463545, "grad_norm": 0.30025166273117065, "learning_rate": 1.4042101769078355e-07, "loss": 0.2047, "step": 7145 }, { "epoch": 1.9015433741351782, "grad_norm": 0.2745387852191925, "learning_rate": 1.404054120824875e-07, "loss": 0.1874, "step": 7146 }, { "epoch": 1.9018094731240023, "grad_norm": 0.2909812033176422, "learning_rate": 1.4038980529809486e-07, "loss": 0.203, "step": 7147 }, { "epoch": 1.902075572112826, "grad_norm": 0.26656094193458557, "learning_rate": 1.403741973380598e-07, "loss": 0.1868, "step": 7148 }, { "epoch": 1.9023416711016499, "grad_norm": 0.2753961980342865, "learning_rate": 1.4035858820283674e-07, "loss": 0.1955, "step": 7149 }, { "epoch": 1.9026077700904738, "grad_norm": 0.3071448504924774, "learning_rate": 1.4034297789287995e-07, "loss": 0.2095, "step": 7150 }, { "epoch": 1.9028738690792975, "grad_norm": 0.31986871361732483, "learning_rate": 1.4032736640864383e-07, "loss": 0.204, "step": 7151 }, { "epoch": 1.9031399680681214, "grad_norm": 0.28341153264045715, "learning_rate": 1.4031175375058278e-07, "loss": 0.206, "step": 7152 }, { "epoch": 1.9034060670569453, "grad_norm": 0.2794446349143982, "learning_rate": 1.4029613991915124e-07, "loss": 0.1868, "step": 7153 }, { "epoch": 1.903672166045769, "grad_norm": 0.3057924211025238, "learning_rate": 1.402805249148037e-07, "loss": 0.219, "step": 7154 }, { "epoch": 1.9039382650345928, "grad_norm": 0.283577561378479, "learning_rate": 1.4026490873799466e-07, "loss": 0.2054, "step": 7155 }, { "epoch": 1.9042043640234168, "grad_norm": 0.312452107667923, "learning_rate": 1.4024929138917868e-07, "loss": 0.2128, "step": 7156 }, { "epoch": 1.9044704630122404, "grad_norm": 0.3097096085548401, "learning_rate": 1.402336728688103e-07, "loss": 0.204, "step": 7157 }, { "epoch": 1.9047365620010646, "grad_norm": 0.2773188054561615, "learning_rate": 1.4021805317734421e-07, "loss": 0.2165, "step": 7158 }, { "epoch": 1.9050026609898882, "grad_norm": 0.2867109179496765, "learning_rate": 1.4020243231523502e-07, "loss": 0.1876, "step": 7159 }, { "epoch": 1.905268759978712, "grad_norm": 0.275371789932251, "learning_rate": 1.401868102829374e-07, "loss": 0.1947, "step": 7160 }, { "epoch": 1.905534858967536, "grad_norm": 0.2675423324108124, "learning_rate": 1.4017118708090608e-07, "loss": 0.1832, "step": 7161 }, { "epoch": 1.9058009579563597, "grad_norm": 0.28326866030693054, "learning_rate": 1.4015556270959582e-07, "loss": 0.1933, "step": 7162 }, { "epoch": 1.9060670569451836, "grad_norm": 0.3155865967273712, "learning_rate": 1.4013993716946135e-07, "loss": 0.2011, "step": 7163 }, { "epoch": 1.9063331559340075, "grad_norm": 0.3269105553627014, "learning_rate": 1.4012431046095756e-07, "loss": 0.2015, "step": 7164 }, { "epoch": 1.9065992549228312, "grad_norm": 0.30855780839920044, "learning_rate": 1.4010868258453928e-07, "loss": 0.2146, "step": 7165 }, { "epoch": 1.9068653539116551, "grad_norm": 0.29531431198120117, "learning_rate": 1.4009305354066136e-07, "loss": 0.2095, "step": 7166 }, { "epoch": 1.907131452900479, "grad_norm": 0.2998502254486084, "learning_rate": 1.400774233297788e-07, "loss": 0.1965, "step": 7167 }, { "epoch": 1.9073975518893027, "grad_norm": 0.28542348742485046, "learning_rate": 1.4006179195234644e-07, "loss": 0.2, "step": 7168 }, { "epoch": 1.9076636508781266, "grad_norm": 0.29285478591918945, "learning_rate": 1.4004615940881937e-07, "loss": 0.2173, "step": 7169 }, { "epoch": 1.9079297498669505, "grad_norm": 0.27229762077331543, "learning_rate": 1.4003052569965262e-07, "loss": 0.199, "step": 7170 }, { "epoch": 1.9081958488557742, "grad_norm": 0.335868239402771, "learning_rate": 1.4001489082530115e-07, "loss": 0.2026, "step": 7171 }, { "epoch": 1.9084619478445983, "grad_norm": 0.29297542572021484, "learning_rate": 1.3999925478622014e-07, "loss": 0.2144, "step": 7172 }, { "epoch": 1.908728046833422, "grad_norm": 0.27872011065483093, "learning_rate": 1.3998361758286466e-07, "loss": 0.1984, "step": 7173 }, { "epoch": 1.908994145822246, "grad_norm": 0.30856049060821533, "learning_rate": 1.399679792156899e-07, "loss": 0.2075, "step": 7174 }, { "epoch": 1.9092602448110698, "grad_norm": 0.2769159972667694, "learning_rate": 1.3995233968515102e-07, "loss": 0.1989, "step": 7175 }, { "epoch": 1.9095263437998935, "grad_norm": 0.29233574867248535, "learning_rate": 1.399366989917033e-07, "loss": 0.2131, "step": 7176 }, { "epoch": 1.9097924427887174, "grad_norm": 0.30340254306793213, "learning_rate": 1.399210571358019e-07, "loss": 0.2234, "step": 7177 }, { "epoch": 1.9100585417775413, "grad_norm": 0.3185959756374359, "learning_rate": 1.3990541411790225e-07, "loss": 0.2199, "step": 7178 }, { "epoch": 1.910324640766365, "grad_norm": 0.2816523015499115, "learning_rate": 1.398897699384596e-07, "loss": 0.1903, "step": 7179 }, { "epoch": 1.910590739755189, "grad_norm": 0.3006073832511902, "learning_rate": 1.398741245979293e-07, "loss": 0.1941, "step": 7180 }, { "epoch": 1.9108568387440128, "grad_norm": 0.2949252724647522, "learning_rate": 1.3985847809676679e-07, "loss": 0.2037, "step": 7181 }, { "epoch": 1.9111229377328365, "grad_norm": 0.2937493324279785, "learning_rate": 1.3984283043542743e-07, "loss": 0.2017, "step": 7182 }, { "epoch": 1.9113890367216606, "grad_norm": 0.3771069645881653, "learning_rate": 1.3982718161436678e-07, "loss": 0.2134, "step": 7183 }, { "epoch": 1.9116551357104843, "grad_norm": 0.28254449367523193, "learning_rate": 1.3981153163404024e-07, "loss": 0.1875, "step": 7184 }, { "epoch": 1.9119212346993082, "grad_norm": 0.2752287685871124, "learning_rate": 1.397958804949034e-07, "loss": 0.1857, "step": 7185 }, { "epoch": 1.912187333688132, "grad_norm": 0.3118472695350647, "learning_rate": 1.397802281974118e-07, "loss": 0.1801, "step": 7186 }, { "epoch": 1.9124534326769558, "grad_norm": 0.3112482726573944, "learning_rate": 1.3976457474202106e-07, "loss": 0.2332, "step": 7187 }, { "epoch": 1.9127195316657797, "grad_norm": 0.31234490871429443, "learning_rate": 1.397489201291868e-07, "loss": 0.2082, "step": 7188 }, { "epoch": 1.9129856306546036, "grad_norm": 0.2766404151916504, "learning_rate": 1.3973326435936466e-07, "loss": 0.2026, "step": 7189 }, { "epoch": 1.9132517296434273, "grad_norm": 0.30171746015548706, "learning_rate": 1.3971760743301037e-07, "loss": 0.1945, "step": 7190 }, { "epoch": 1.9135178286322512, "grad_norm": 0.29771101474761963, "learning_rate": 1.3970194935057967e-07, "loss": 0.2084, "step": 7191 }, { "epoch": 1.913783927621075, "grad_norm": 0.3477087914943695, "learning_rate": 1.396862901125283e-07, "loss": 0.2137, "step": 7192 }, { "epoch": 1.9140500266098988, "grad_norm": 0.2952139973640442, "learning_rate": 1.3967062971931206e-07, "loss": 0.2175, "step": 7193 }, { "epoch": 1.9143161255987229, "grad_norm": 0.29523780941963196, "learning_rate": 1.396549681713868e-07, "loss": 0.202, "step": 7194 }, { "epoch": 1.9145822245875466, "grad_norm": 0.28382566571235657, "learning_rate": 1.3963930546920836e-07, "loss": 0.1884, "step": 7195 }, { "epoch": 1.9148483235763702, "grad_norm": 0.27989456057548523, "learning_rate": 1.396236416132327e-07, "loss": 0.2061, "step": 7196 }, { "epoch": 1.9151144225651944, "grad_norm": 0.3121044635772705, "learning_rate": 1.3960797660391568e-07, "loss": 0.2347, "step": 7197 }, { "epoch": 1.915380521554018, "grad_norm": 0.30002540349960327, "learning_rate": 1.3959231044171332e-07, "loss": 0.2177, "step": 7198 }, { "epoch": 1.915646620542842, "grad_norm": 0.3146650195121765, "learning_rate": 1.395766431270816e-07, "loss": 0.2164, "step": 7199 }, { "epoch": 1.9159127195316659, "grad_norm": 0.2980334162712097, "learning_rate": 1.3956097466047653e-07, "loss": 0.2213, "step": 7200 }, { "epoch": 1.9161788185204895, "grad_norm": 0.27247023582458496, "learning_rate": 1.3954530504235423e-07, "loss": 0.1984, "step": 7201 }, { "epoch": 1.9164449175093135, "grad_norm": 0.3051202893257141, "learning_rate": 1.3952963427317083e-07, "loss": 0.1992, "step": 7202 }, { "epoch": 1.9167110164981374, "grad_norm": 0.29935309290885925, "learning_rate": 1.3951396235338234e-07, "loss": 0.1854, "step": 7203 }, { "epoch": 1.916977115486961, "grad_norm": 0.2959674894809723, "learning_rate": 1.3949828928344503e-07, "loss": 0.2172, "step": 7204 }, { "epoch": 1.917243214475785, "grad_norm": 0.2620013654232025, "learning_rate": 1.3948261506381508e-07, "loss": 0.1708, "step": 7205 }, { "epoch": 1.9175093134646088, "grad_norm": 0.2777532637119293, "learning_rate": 1.3946693969494872e-07, "loss": 0.1992, "step": 7206 }, { "epoch": 1.9177754124534325, "grad_norm": 0.2984269857406616, "learning_rate": 1.394512631773022e-07, "loss": 0.2054, "step": 7207 }, { "epoch": 1.9180415114422567, "grad_norm": 0.2872675955295563, "learning_rate": 1.3943558551133186e-07, "loss": 0.2158, "step": 7208 }, { "epoch": 1.9183076104310803, "grad_norm": 0.29910072684288025, "learning_rate": 1.3941990669749398e-07, "loss": 0.2091, "step": 7209 }, { "epoch": 1.9185737094199042, "grad_norm": 0.3069804608821869, "learning_rate": 1.3940422673624503e-07, "loss": 0.2081, "step": 7210 }, { "epoch": 1.9188398084087281, "grad_norm": 0.2668302357196808, "learning_rate": 1.3938854562804134e-07, "loss": 0.1775, "step": 7211 }, { "epoch": 1.9191059073975518, "grad_norm": 0.2785673141479492, "learning_rate": 1.3937286337333935e-07, "loss": 0.1794, "step": 7212 }, { "epoch": 1.9193720063863757, "grad_norm": 0.28900134563446045, "learning_rate": 1.3935717997259553e-07, "loss": 0.1997, "step": 7213 }, { "epoch": 1.9196381053751996, "grad_norm": 0.2906067967414856, "learning_rate": 1.393414954262664e-07, "loss": 0.21, "step": 7214 }, { "epoch": 1.9199042043640233, "grad_norm": 0.26693806052207947, "learning_rate": 1.3932580973480848e-07, "loss": 0.1853, "step": 7215 }, { "epoch": 1.9201703033528472, "grad_norm": 0.2887839674949646, "learning_rate": 1.3931012289867833e-07, "loss": 0.2061, "step": 7216 }, { "epoch": 1.9204364023416711, "grad_norm": 0.3060206174850464, "learning_rate": 1.392944349183326e-07, "loss": 0.1908, "step": 7217 }, { "epoch": 1.9207025013304948, "grad_norm": 0.29906946420669556, "learning_rate": 1.3927874579422787e-07, "loss": 0.1997, "step": 7218 }, { "epoch": 1.920968600319319, "grad_norm": 0.2615182101726532, "learning_rate": 1.3926305552682086e-07, "loss": 0.1979, "step": 7219 }, { "epoch": 1.9212346993081426, "grad_norm": 0.28093719482421875, "learning_rate": 1.3924736411656825e-07, "loss": 0.2102, "step": 7220 }, { "epoch": 1.9215007982969665, "grad_norm": 0.30487456917762756, "learning_rate": 1.3923167156392675e-07, "loss": 0.2082, "step": 7221 }, { "epoch": 1.9217668972857904, "grad_norm": 0.2921591103076935, "learning_rate": 1.392159778693532e-07, "loss": 0.2041, "step": 7222 }, { "epoch": 1.922032996274614, "grad_norm": 0.2686072885990143, "learning_rate": 1.392002830333043e-07, "loss": 0.1934, "step": 7223 }, { "epoch": 1.922299095263438, "grad_norm": 0.2656094431877136, "learning_rate": 1.39184587056237e-07, "loss": 0.1729, "step": 7224 }, { "epoch": 1.922565194252262, "grad_norm": 0.3066990077495575, "learning_rate": 1.3916888993860808e-07, "loss": 0.2319, "step": 7225 }, { "epoch": 1.9228312932410856, "grad_norm": 0.288495272397995, "learning_rate": 1.3915319168087447e-07, "loss": 0.1971, "step": 7226 }, { "epoch": 1.9230973922299095, "grad_norm": 0.29492923617362976, "learning_rate": 1.391374922834931e-07, "loss": 0.202, "step": 7227 }, { "epoch": 1.9233634912187334, "grad_norm": 0.3160020112991333, "learning_rate": 1.39121791746921e-07, "loss": 0.2149, "step": 7228 }, { "epoch": 1.923629590207557, "grad_norm": 0.28882288932800293, "learning_rate": 1.3910609007161506e-07, "loss": 0.1942, "step": 7229 }, { "epoch": 1.9238956891963812, "grad_norm": 0.3000263273715973, "learning_rate": 1.3909038725803243e-07, "loss": 0.2079, "step": 7230 }, { "epoch": 1.924161788185205, "grad_norm": 0.2719027101993561, "learning_rate": 1.390746833066301e-07, "loss": 0.182, "step": 7231 }, { "epoch": 1.9244278871740286, "grad_norm": 0.285496324300766, "learning_rate": 1.3905897821786519e-07, "loss": 0.1955, "step": 7232 }, { "epoch": 1.9246939861628527, "grad_norm": 0.2742283344268799, "learning_rate": 1.3904327199219485e-07, "loss": 0.1912, "step": 7233 }, { "epoch": 1.9249600851516764, "grad_norm": 0.3084053695201874, "learning_rate": 1.3902756463007623e-07, "loss": 0.2025, "step": 7234 }, { "epoch": 1.9252261841405003, "grad_norm": 0.2841232717037201, "learning_rate": 1.3901185613196652e-07, "loss": 0.2061, "step": 7235 }, { "epoch": 1.9254922831293242, "grad_norm": 0.2715793251991272, "learning_rate": 1.3899614649832302e-07, "loss": 0.1927, "step": 7236 }, { "epoch": 1.9257583821181479, "grad_norm": 0.2760571241378784, "learning_rate": 1.389804357296029e-07, "loss": 0.2002, "step": 7237 }, { "epoch": 1.9260244811069718, "grad_norm": 0.2990305721759796, "learning_rate": 1.3896472382626354e-07, "loss": 0.1969, "step": 7238 }, { "epoch": 1.9262905800957957, "grad_norm": 0.31838440895080566, "learning_rate": 1.3894901078876223e-07, "loss": 0.2128, "step": 7239 }, { "epoch": 1.9265566790846194, "grad_norm": 0.30255740880966187, "learning_rate": 1.389332966175563e-07, "loss": 0.2022, "step": 7240 }, { "epoch": 1.9268227780734435, "grad_norm": 0.2862800359725952, "learning_rate": 1.3891758131310328e-07, "loss": 0.2062, "step": 7241 }, { "epoch": 1.9270888770622672, "grad_norm": 0.2829427719116211, "learning_rate": 1.3890186487586046e-07, "loss": 0.1993, "step": 7242 }, { "epoch": 1.9273549760510909, "grad_norm": 0.2965478003025055, "learning_rate": 1.388861473062854e-07, "loss": 0.209, "step": 7243 }, { "epoch": 1.927621075039915, "grad_norm": 0.3050122559070587, "learning_rate": 1.388704286048355e-07, "loss": 0.1959, "step": 7244 }, { "epoch": 1.9278871740287387, "grad_norm": 0.2931605279445648, "learning_rate": 1.388547087719684e-07, "loss": 0.2163, "step": 7245 }, { "epoch": 1.9281532730175626, "grad_norm": 0.28218168020248413, "learning_rate": 1.3883898780814158e-07, "loss": 0.2063, "step": 7246 }, { "epoch": 1.9284193720063865, "grad_norm": 0.3193168044090271, "learning_rate": 1.388232657138127e-07, "loss": 0.2131, "step": 7247 }, { "epoch": 1.9286854709952101, "grad_norm": 0.2940095365047455, "learning_rate": 1.3880754248943934e-07, "loss": 0.184, "step": 7248 }, { "epoch": 1.928951569984034, "grad_norm": 0.2681525945663452, "learning_rate": 1.3879181813547918e-07, "loss": 0.189, "step": 7249 }, { "epoch": 1.929217668972858, "grad_norm": 0.3013707995414734, "learning_rate": 1.387760926523899e-07, "loss": 0.2081, "step": 7250 }, { "epoch": 1.9294837679616816, "grad_norm": 0.2878863513469696, "learning_rate": 1.387603660406293e-07, "loss": 0.1982, "step": 7251 }, { "epoch": 1.9297498669505055, "grad_norm": 0.2978900969028473, "learning_rate": 1.38744638300655e-07, "loss": 0.191, "step": 7252 }, { "epoch": 1.9300159659393294, "grad_norm": 0.2602960467338562, "learning_rate": 1.3872890943292496e-07, "loss": 0.1812, "step": 7253 }, { "epoch": 1.9302820649281531, "grad_norm": 0.2694632112979889, "learning_rate": 1.387131794378969e-07, "loss": 0.1882, "step": 7254 }, { "epoch": 1.9305481639169773, "grad_norm": 0.3235064446926117, "learning_rate": 1.3869744831602868e-07, "loss": 0.2214, "step": 7255 }, { "epoch": 1.930814262905801, "grad_norm": 0.3115925192832947, "learning_rate": 1.3868171606777825e-07, "loss": 0.2291, "step": 7256 }, { "epoch": 1.9310803618946248, "grad_norm": 0.2823883295059204, "learning_rate": 1.3866598269360348e-07, "loss": 0.1983, "step": 7257 }, { "epoch": 1.9313464608834487, "grad_norm": 0.3021731674671173, "learning_rate": 1.3865024819396235e-07, "loss": 0.1723, "step": 7258 }, { "epoch": 1.9316125598722724, "grad_norm": 0.3144538104534149, "learning_rate": 1.3863451256931284e-07, "loss": 0.2361, "step": 7259 }, { "epoch": 1.9318786588610963, "grad_norm": 0.29666435718536377, "learning_rate": 1.3861877582011303e-07, "loss": 0.1828, "step": 7260 }, { "epoch": 1.9321447578499202, "grad_norm": 0.2892526388168335, "learning_rate": 1.3860303794682086e-07, "loss": 0.2001, "step": 7261 }, { "epoch": 1.932410856838744, "grad_norm": 0.3058989346027374, "learning_rate": 1.3858729894989455e-07, "loss": 0.2265, "step": 7262 }, { "epoch": 1.9326769558275678, "grad_norm": 0.27382737398147583, "learning_rate": 1.385715588297921e-07, "loss": 0.2039, "step": 7263 }, { "epoch": 1.9329430548163917, "grad_norm": 0.29211530089378357, "learning_rate": 1.3855581758697177e-07, "loss": 0.206, "step": 7264 }, { "epoch": 1.9332091538052154, "grad_norm": 0.3016088306903839, "learning_rate": 1.3854007522189172e-07, "loss": 0.2116, "step": 7265 }, { "epoch": 1.9334752527940395, "grad_norm": 0.3018273413181305, "learning_rate": 1.3852433173501007e-07, "loss": 0.2158, "step": 7266 }, { "epoch": 1.9337413517828632, "grad_norm": 0.30378052592277527, "learning_rate": 1.3850858712678519e-07, "loss": 0.2084, "step": 7267 }, { "epoch": 1.9340074507716871, "grad_norm": 0.29327404499053955, "learning_rate": 1.3849284139767533e-07, "loss": 0.2206, "step": 7268 }, { "epoch": 1.934273549760511, "grad_norm": 0.281403511762619, "learning_rate": 1.3847709454813882e-07, "loss": 0.1996, "step": 7269 }, { "epoch": 1.9345396487493347, "grad_norm": 0.3253389596939087, "learning_rate": 1.3846134657863396e-07, "loss": 0.2036, "step": 7270 }, { "epoch": 1.9348057477381586, "grad_norm": 0.29265692830085754, "learning_rate": 1.3844559748961918e-07, "loss": 0.2166, "step": 7271 }, { "epoch": 1.9350718467269825, "grad_norm": 0.4443837106227875, "learning_rate": 1.3842984728155288e-07, "loss": 0.2179, "step": 7272 }, { "epoch": 1.9353379457158062, "grad_norm": 0.2859960198402405, "learning_rate": 1.384140959548935e-07, "loss": 0.2075, "step": 7273 }, { "epoch": 1.93560404470463, "grad_norm": 0.2767214775085449, "learning_rate": 1.3839834351009954e-07, "loss": 0.1826, "step": 7274 }, { "epoch": 1.935870143693454, "grad_norm": 0.29306209087371826, "learning_rate": 1.383825899476295e-07, "loss": 0.2039, "step": 7275 }, { "epoch": 1.9361362426822777, "grad_norm": 0.3003107011318207, "learning_rate": 1.383668352679419e-07, "loss": 0.2064, "step": 7276 }, { "epoch": 1.9364023416711018, "grad_norm": 0.39397668838500977, "learning_rate": 1.3835107947149535e-07, "loss": 0.2025, "step": 7277 }, { "epoch": 1.9366684406599255, "grad_norm": 0.27652662992477417, "learning_rate": 1.383353225587485e-07, "loss": 0.1954, "step": 7278 }, { "epoch": 1.9369345396487492, "grad_norm": 0.25935351848602295, "learning_rate": 1.383195645301599e-07, "loss": 0.1758, "step": 7279 }, { "epoch": 1.9372006386375733, "grad_norm": 0.2882143557071686, "learning_rate": 1.383038053861883e-07, "loss": 0.212, "step": 7280 }, { "epoch": 1.937466737626397, "grad_norm": 0.2828713059425354, "learning_rate": 1.3828804512729235e-07, "loss": 0.1965, "step": 7281 }, { "epoch": 1.9377328366152209, "grad_norm": 0.30092763900756836, "learning_rate": 1.3827228375393082e-07, "loss": 0.2005, "step": 7282 }, { "epoch": 1.9379989356040448, "grad_norm": 0.27593499422073364, "learning_rate": 1.3825652126656254e-07, "loss": 0.1831, "step": 7283 }, { "epoch": 1.9382650345928685, "grad_norm": 0.2684282064437866, "learning_rate": 1.382407576656462e-07, "loss": 0.1972, "step": 7284 }, { "epoch": 1.9385311335816924, "grad_norm": 0.3008694052696228, "learning_rate": 1.3822499295164073e-07, "loss": 0.2165, "step": 7285 }, { "epoch": 1.9387972325705163, "grad_norm": 0.29941484332084656, "learning_rate": 1.382092271250049e-07, "loss": 0.2129, "step": 7286 }, { "epoch": 1.93906333155934, "grad_norm": 0.2892858684062958, "learning_rate": 1.3819346018619777e-07, "loss": 0.2198, "step": 7287 }, { "epoch": 1.9393294305481639, "grad_norm": 0.3285612463951111, "learning_rate": 1.3817769213567814e-07, "loss": 0.2151, "step": 7288 }, { "epoch": 1.9395955295369878, "grad_norm": 0.2812759280204773, "learning_rate": 1.38161922973905e-07, "loss": 0.2195, "step": 7289 }, { "epoch": 1.9398616285258115, "grad_norm": 0.2794170677661896, "learning_rate": 1.3814615270133737e-07, "loss": 0.1933, "step": 7290 }, { "epoch": 1.9401277275146356, "grad_norm": 0.29640164971351624, "learning_rate": 1.381303813184343e-07, "loss": 0.2118, "step": 7291 }, { "epoch": 1.9403938265034593, "grad_norm": 0.3056979179382324, "learning_rate": 1.3811460882565484e-07, "loss": 0.2063, "step": 7292 }, { "epoch": 1.9406599254922832, "grad_norm": 0.29699674248695374, "learning_rate": 1.380988352234581e-07, "loss": 0.2173, "step": 7293 }, { "epoch": 1.940926024481107, "grad_norm": 0.29265594482421875, "learning_rate": 1.3808306051230314e-07, "loss": 0.19, "step": 7294 }, { "epoch": 1.9411921234699308, "grad_norm": 0.31581589579582214, "learning_rate": 1.380672846926492e-07, "loss": 0.2148, "step": 7295 }, { "epoch": 1.9414582224587547, "grad_norm": 0.29011106491088867, "learning_rate": 1.3805150776495544e-07, "loss": 0.2038, "step": 7296 }, { "epoch": 1.9417243214475786, "grad_norm": 0.2864242196083069, "learning_rate": 1.3803572972968106e-07, "loss": 0.1794, "step": 7297 }, { "epoch": 1.9419904204364022, "grad_norm": 0.2890528738498688, "learning_rate": 1.3801995058728536e-07, "loss": 0.2035, "step": 7298 }, { "epoch": 1.9422565194252261, "grad_norm": 0.3053484857082367, "learning_rate": 1.3800417033822764e-07, "loss": 0.2195, "step": 7299 }, { "epoch": 1.94252261841405, "grad_norm": 0.2981395423412323, "learning_rate": 1.3798838898296716e-07, "loss": 0.2176, "step": 7300 }, { "epoch": 1.9427887174028737, "grad_norm": 0.30432331562042236, "learning_rate": 1.3797260652196337e-07, "loss": 0.2183, "step": 7301 }, { "epoch": 1.9430548163916979, "grad_norm": 0.30638280510902405, "learning_rate": 1.3795682295567555e-07, "loss": 0.2238, "step": 7302 }, { "epoch": 1.9433209153805215, "grad_norm": 0.25429511070251465, "learning_rate": 1.3794103828456322e-07, "loss": 0.188, "step": 7303 }, { "epoch": 1.9435870143693454, "grad_norm": 0.2935379445552826, "learning_rate": 1.379252525090857e-07, "loss": 0.1984, "step": 7304 }, { "epoch": 1.9438531133581693, "grad_norm": 0.30199000239372253, "learning_rate": 1.3790946562970262e-07, "loss": 0.1942, "step": 7305 }, { "epoch": 1.944119212346993, "grad_norm": 0.28559014201164246, "learning_rate": 1.3789367764687342e-07, "loss": 0.2086, "step": 7306 }, { "epoch": 1.944385311335817, "grad_norm": 0.2904413938522339, "learning_rate": 1.378778885610576e-07, "loss": 0.2077, "step": 7307 }, { "epoch": 1.9446514103246408, "grad_norm": 0.293154239654541, "learning_rate": 1.3786209837271483e-07, "loss": 0.2111, "step": 7308 }, { "epoch": 1.9449175093134645, "grad_norm": 0.36444419622421265, "learning_rate": 1.3784630708230466e-07, "loss": 0.2092, "step": 7309 }, { "epoch": 1.9451836083022884, "grad_norm": 0.32042860984802246, "learning_rate": 1.378305146902868e-07, "loss": 0.2201, "step": 7310 }, { "epoch": 1.9454497072911123, "grad_norm": 0.28168007731437683, "learning_rate": 1.3781472119712086e-07, "loss": 0.1951, "step": 7311 }, { "epoch": 1.945715806279936, "grad_norm": 0.29269424080848694, "learning_rate": 1.3779892660326655e-07, "loss": 0.1861, "step": 7312 }, { "epoch": 1.9459819052687601, "grad_norm": 0.2674993574619293, "learning_rate": 1.3778313090918368e-07, "loss": 0.1911, "step": 7313 }, { "epoch": 1.9462480042575838, "grad_norm": 0.3278270363807678, "learning_rate": 1.3776733411533194e-07, "loss": 0.2128, "step": 7314 }, { "epoch": 1.9465141032464075, "grad_norm": 0.2792365252971649, "learning_rate": 1.377515362221712e-07, "loss": 0.1899, "step": 7315 }, { "epoch": 1.9467802022352316, "grad_norm": 0.2882739007472992, "learning_rate": 1.377357372301612e-07, "loss": 0.1962, "step": 7316 }, { "epoch": 1.9470463012240553, "grad_norm": 0.3013015687465668, "learning_rate": 1.377199371397619e-07, "loss": 0.214, "step": 7317 }, { "epoch": 1.9473124002128792, "grad_norm": 0.28927892446517944, "learning_rate": 1.3770413595143319e-07, "loss": 0.197, "step": 7318 }, { "epoch": 1.9475784992017031, "grad_norm": 0.3619077801704407, "learning_rate": 1.3768833366563494e-07, "loss": 0.1886, "step": 7319 }, { "epoch": 1.9478445981905268, "grad_norm": 0.2779255211353302, "learning_rate": 1.3767253028282717e-07, "loss": 0.2048, "step": 7320 }, { "epoch": 1.9481106971793507, "grad_norm": 0.30206573009490967, "learning_rate": 1.3765672580346987e-07, "loss": 0.2116, "step": 7321 }, { "epoch": 1.9483767961681746, "grad_norm": 0.2785831093788147, "learning_rate": 1.3764092022802305e-07, "loss": 0.1863, "step": 7322 }, { "epoch": 1.9486428951569983, "grad_norm": 0.2878943383693695, "learning_rate": 1.3762511355694677e-07, "loss": 0.1923, "step": 7323 }, { "epoch": 1.9489089941458222, "grad_norm": 0.27170246839523315, "learning_rate": 1.3760930579070115e-07, "loss": 0.1769, "step": 7324 }, { "epoch": 1.949175093134646, "grad_norm": 0.3133644163608551, "learning_rate": 1.3759349692974628e-07, "loss": 0.2101, "step": 7325 }, { "epoch": 1.9494411921234698, "grad_norm": 0.30141687393188477, "learning_rate": 1.375776869745423e-07, "loss": 0.2095, "step": 7326 }, { "epoch": 1.949707291112294, "grad_norm": 0.2801762521266937, "learning_rate": 1.3756187592554945e-07, "loss": 0.1927, "step": 7327 }, { "epoch": 1.9499733901011176, "grad_norm": 0.27938196063041687, "learning_rate": 1.375460637832279e-07, "loss": 0.1814, "step": 7328 }, { "epoch": 1.9502394890899415, "grad_norm": 0.2731154263019562, "learning_rate": 1.3753025054803794e-07, "loss": 0.2111, "step": 7329 }, { "epoch": 1.9505055880787654, "grad_norm": 0.2840169668197632, "learning_rate": 1.375144362204398e-07, "loss": 0.1772, "step": 7330 }, { "epoch": 1.950771687067589, "grad_norm": 0.28440359234809875, "learning_rate": 1.3749862080089385e-07, "loss": 0.1877, "step": 7331 }, { "epoch": 1.951037786056413, "grad_norm": 0.3109281659126282, "learning_rate": 1.3748280428986044e-07, "loss": 0.2064, "step": 7332 }, { "epoch": 1.9513038850452369, "grad_norm": 0.27910521626472473, "learning_rate": 1.3746698668779988e-07, "loss": 0.1841, "step": 7333 }, { "epoch": 1.9515699840340606, "grad_norm": 0.2794899642467499, "learning_rate": 1.3745116799517266e-07, "loss": 0.1956, "step": 7334 }, { "epoch": 1.9518360830228845, "grad_norm": 0.32277950644493103, "learning_rate": 1.3743534821243912e-07, "loss": 0.2356, "step": 7335 }, { "epoch": 1.9521021820117084, "grad_norm": 0.30003243684768677, "learning_rate": 1.3741952734005983e-07, "loss": 0.1988, "step": 7336 }, { "epoch": 1.952368281000532, "grad_norm": 0.30401182174682617, "learning_rate": 1.3740370537849526e-07, "loss": 0.1985, "step": 7337 }, { "epoch": 1.9526343799893562, "grad_norm": 0.29101499915122986, "learning_rate": 1.3738788232820593e-07, "loss": 0.195, "step": 7338 }, { "epoch": 1.9529004789781799, "grad_norm": 0.2858649492263794, "learning_rate": 1.3737205818965242e-07, "loss": 0.2047, "step": 7339 }, { "epoch": 1.9531665779670038, "grad_norm": 0.30798572301864624, "learning_rate": 1.3735623296329534e-07, "loss": 0.2075, "step": 7340 }, { "epoch": 1.9534326769558277, "grad_norm": 0.3096494972705841, "learning_rate": 1.3734040664959534e-07, "loss": 0.2172, "step": 7341 }, { "epoch": 1.9536987759446514, "grad_norm": 0.2896106541156769, "learning_rate": 1.3732457924901302e-07, "loss": 0.2059, "step": 7342 }, { "epoch": 1.9539648749334753, "grad_norm": 0.2956511378288269, "learning_rate": 1.3730875076200913e-07, "loss": 0.2057, "step": 7343 }, { "epoch": 1.9542309739222992, "grad_norm": 0.27752944827079773, "learning_rate": 1.3729292118904438e-07, "loss": 0.1826, "step": 7344 }, { "epoch": 1.9544970729111228, "grad_norm": 0.26883092522621155, "learning_rate": 1.372770905305795e-07, "loss": 0.1799, "step": 7345 }, { "epoch": 1.9547631718999467, "grad_norm": 0.30135291814804077, "learning_rate": 1.3726125878707533e-07, "loss": 0.2203, "step": 7346 }, { "epoch": 1.9550292708887707, "grad_norm": 0.2727035880088806, "learning_rate": 1.3724542595899264e-07, "loss": 0.1879, "step": 7347 }, { "epoch": 1.9552953698775943, "grad_norm": 0.27661019563674927, "learning_rate": 1.3722959204679232e-07, "loss": 0.2158, "step": 7348 }, { "epoch": 1.9555614688664185, "grad_norm": 0.30072906613349915, "learning_rate": 1.3721375705093523e-07, "loss": 0.2261, "step": 7349 }, { "epoch": 1.9558275678552421, "grad_norm": 0.2723449766635895, "learning_rate": 1.371979209718823e-07, "loss": 0.2037, "step": 7350 }, { "epoch": 1.956093666844066, "grad_norm": 0.30340078473091125, "learning_rate": 1.3718208381009448e-07, "loss": 0.2076, "step": 7351 }, { "epoch": 1.95635976583289, "grad_norm": 0.2718595862388611, "learning_rate": 1.3716624556603275e-07, "loss": 0.1963, "step": 7352 }, { "epoch": 1.9566258648217136, "grad_norm": 0.2762238681316376, "learning_rate": 1.371504062401581e-07, "loss": 0.2011, "step": 7353 }, { "epoch": 1.9568919638105375, "grad_norm": 0.30411863327026367, "learning_rate": 1.371345658329316e-07, "loss": 0.2095, "step": 7354 }, { "epoch": 1.9571580627993614, "grad_norm": 0.2750348448753357, "learning_rate": 1.371187243448143e-07, "loss": 0.2067, "step": 7355 }, { "epoch": 1.9574241617881851, "grad_norm": 0.28891196846961975, "learning_rate": 1.371028817762673e-07, "loss": 0.2022, "step": 7356 }, { "epoch": 1.957690260777009, "grad_norm": 0.3138081431388855, "learning_rate": 1.3708703812775175e-07, "loss": 0.2137, "step": 7357 }, { "epoch": 1.957956359765833, "grad_norm": 0.31062009930610657, "learning_rate": 1.3707119339972883e-07, "loss": 0.1881, "step": 7358 }, { "epoch": 1.9582224587546566, "grad_norm": 0.2941916584968567, "learning_rate": 1.370553475926597e-07, "loss": 0.2068, "step": 7359 }, { "epoch": 1.9584885577434807, "grad_norm": 0.2735537886619568, "learning_rate": 1.3703950070700564e-07, "loss": 0.19, "step": 7360 }, { "epoch": 1.9587546567323044, "grad_norm": 0.30234354734420776, "learning_rate": 1.370236527432279e-07, "loss": 0.1953, "step": 7361 }, { "epoch": 1.959020755721128, "grad_norm": 0.27101704478263855, "learning_rate": 1.370078037017877e-07, "loss": 0.1797, "step": 7362 }, { "epoch": 1.9592868547099522, "grad_norm": 0.30833807587623596, "learning_rate": 1.3699195358314648e-07, "loss": 0.2266, "step": 7363 }, { "epoch": 1.959552953698776, "grad_norm": 0.2982029318809509, "learning_rate": 1.3697610238776554e-07, "loss": 0.2075, "step": 7364 }, { "epoch": 1.9598190526875998, "grad_norm": 0.310389906167984, "learning_rate": 1.3696025011610626e-07, "loss": 0.2015, "step": 7365 }, { "epoch": 1.9600851516764237, "grad_norm": 0.29480868577957153, "learning_rate": 1.3694439676863003e-07, "loss": 0.2171, "step": 7366 }, { "epoch": 1.9603512506652474, "grad_norm": 0.27815011143684387, "learning_rate": 1.3692854234579837e-07, "loss": 0.2014, "step": 7367 }, { "epoch": 1.9606173496540713, "grad_norm": 0.27201202511787415, "learning_rate": 1.369126868480727e-07, "loss": 0.1826, "step": 7368 }, { "epoch": 1.9608834486428952, "grad_norm": 0.30063512921333313, "learning_rate": 1.3689683027591458e-07, "loss": 0.1897, "step": 7369 }, { "epoch": 1.961149547631719, "grad_norm": 0.31552600860595703, "learning_rate": 1.3688097262978554e-07, "loss": 0.2281, "step": 7370 }, { "epoch": 1.9614156466205428, "grad_norm": 0.31322047114372253, "learning_rate": 1.3686511391014713e-07, "loss": 0.2295, "step": 7371 }, { "epoch": 1.9616817456093667, "grad_norm": 0.29200059175491333, "learning_rate": 1.3684925411746095e-07, "loss": 0.1942, "step": 7372 }, { "epoch": 1.9619478445981904, "grad_norm": 0.31760773062705994, "learning_rate": 1.3683339325218872e-07, "loss": 0.212, "step": 7373 }, { "epoch": 1.9622139435870145, "grad_norm": 0.31599435210227966, "learning_rate": 1.36817531314792e-07, "loss": 0.2161, "step": 7374 }, { "epoch": 1.9624800425758382, "grad_norm": 0.3094426393508911, "learning_rate": 1.3680166830573257e-07, "loss": 0.2221, "step": 7375 }, { "epoch": 1.962746141564662, "grad_norm": 0.2652767598628998, "learning_rate": 1.3678580422547212e-07, "loss": 0.1765, "step": 7376 }, { "epoch": 1.963012240553486, "grad_norm": 0.3285217583179474, "learning_rate": 1.367699390744724e-07, "loss": 0.2137, "step": 7377 }, { "epoch": 1.9632783395423097, "grad_norm": 0.2662503123283386, "learning_rate": 1.3675407285319528e-07, "loss": 0.2014, "step": 7378 }, { "epoch": 1.9635444385311336, "grad_norm": 0.27952685952186584, "learning_rate": 1.3673820556210246e-07, "loss": 0.1919, "step": 7379 }, { "epoch": 1.9638105375199575, "grad_norm": 0.28404462337493896, "learning_rate": 1.3672233720165592e-07, "loss": 0.2017, "step": 7380 }, { "epoch": 1.9640766365087812, "grad_norm": 0.2627456486225128, "learning_rate": 1.3670646777231747e-07, "loss": 0.1904, "step": 7381 }, { "epoch": 1.964342735497605, "grad_norm": 0.3118263781070709, "learning_rate": 1.3669059727454906e-07, "loss": 0.2167, "step": 7382 }, { "epoch": 1.964608834486429, "grad_norm": 0.27737030386924744, "learning_rate": 1.366747257088126e-07, "loss": 0.1928, "step": 7383 }, { "epoch": 1.9648749334752527, "grad_norm": 0.32777008414268494, "learning_rate": 1.3665885307557016e-07, "loss": 0.2284, "step": 7384 }, { "epoch": 1.9651410324640768, "grad_norm": 0.29942581057548523, "learning_rate": 1.3664297937528363e-07, "loss": 0.2162, "step": 7385 }, { "epoch": 1.9654071314529005, "grad_norm": 0.3162686228752136, "learning_rate": 1.3662710460841514e-07, "loss": 0.206, "step": 7386 }, { "epoch": 1.9656732304417244, "grad_norm": 0.27770137786865234, "learning_rate": 1.3661122877542675e-07, "loss": 0.2024, "step": 7387 }, { "epoch": 1.9659393294305483, "grad_norm": 0.2703913748264313, "learning_rate": 1.3659535187678052e-07, "loss": 0.1869, "step": 7388 }, { "epoch": 1.966205428419372, "grad_norm": 0.28614386916160583, "learning_rate": 1.365794739129386e-07, "loss": 0.1989, "step": 7389 }, { "epoch": 1.9664715274081959, "grad_norm": 0.30369991064071655, "learning_rate": 1.365635948843632e-07, "loss": 0.2105, "step": 7390 }, { "epoch": 1.9667376263970198, "grad_norm": 0.29048413038253784, "learning_rate": 1.365477147915165e-07, "loss": 0.1935, "step": 7391 }, { "epoch": 1.9670037253858434, "grad_norm": 0.29251599311828613, "learning_rate": 1.3653183363486075e-07, "loss": 0.196, "step": 7392 }, { "epoch": 1.9672698243746674, "grad_norm": 0.277534544467926, "learning_rate": 1.3651595141485811e-07, "loss": 0.2076, "step": 7393 }, { "epoch": 1.9675359233634913, "grad_norm": 0.28361135721206665, "learning_rate": 1.3650006813197098e-07, "loss": 0.1971, "step": 7394 }, { "epoch": 1.967802022352315, "grad_norm": 0.29486241936683655, "learning_rate": 1.3648418378666162e-07, "loss": 0.2122, "step": 7395 }, { "epoch": 1.968068121341139, "grad_norm": 0.26865917444229126, "learning_rate": 1.3646829837939243e-07, "loss": 0.1967, "step": 7396 }, { "epoch": 1.9683342203299627, "grad_norm": 0.2853677272796631, "learning_rate": 1.364524119106257e-07, "loss": 0.1884, "step": 7397 }, { "epoch": 1.9686003193187864, "grad_norm": 0.2883319854736328, "learning_rate": 1.3643652438082394e-07, "loss": 0.2004, "step": 7398 }, { "epoch": 1.9688664183076106, "grad_norm": 0.2782043218612671, "learning_rate": 1.3642063579044956e-07, "loss": 0.2072, "step": 7399 }, { "epoch": 1.9691325172964342, "grad_norm": 0.2799913287162781, "learning_rate": 1.3640474613996506e-07, "loss": 0.198, "step": 7400 }, { "epoch": 1.9693986162852581, "grad_norm": 0.2908676862716675, "learning_rate": 1.363888554298329e-07, "loss": 0.1886, "step": 7401 }, { "epoch": 1.969664715274082, "grad_norm": 0.2851880192756653, "learning_rate": 1.3637296366051563e-07, "loss": 0.2054, "step": 7402 }, { "epoch": 1.9699308142629057, "grad_norm": 0.296050488948822, "learning_rate": 1.3635707083247586e-07, "loss": 0.221, "step": 7403 }, { "epoch": 1.9701969132517296, "grad_norm": 0.29025277495384216, "learning_rate": 1.3634117694617614e-07, "loss": 0.2027, "step": 7404 }, { "epoch": 1.9704630122405535, "grad_norm": 0.27161988615989685, "learning_rate": 1.3632528200207914e-07, "loss": 0.187, "step": 7405 }, { "epoch": 1.9707291112293772, "grad_norm": 0.26618868112564087, "learning_rate": 1.3630938600064745e-07, "loss": 0.1879, "step": 7406 }, { "epoch": 1.9709952102182011, "grad_norm": 0.29380467534065247, "learning_rate": 1.3629348894234386e-07, "loss": 0.2039, "step": 7407 }, { "epoch": 1.971261309207025, "grad_norm": 0.2930307686328888, "learning_rate": 1.36277590827631e-07, "loss": 0.2138, "step": 7408 }, { "epoch": 1.9715274081958487, "grad_norm": 0.31611502170562744, "learning_rate": 1.362616916569717e-07, "loss": 0.2057, "step": 7409 }, { "epoch": 1.9717935071846728, "grad_norm": 0.26095902919769287, "learning_rate": 1.362457914308287e-07, "loss": 0.1749, "step": 7410 }, { "epoch": 1.9720596061734965, "grad_norm": 0.299780935049057, "learning_rate": 1.362298901496648e-07, "loss": 0.2048, "step": 7411 }, { "epoch": 1.9723257051623204, "grad_norm": 0.285311222076416, "learning_rate": 1.3621398781394287e-07, "loss": 0.2027, "step": 7412 }, { "epoch": 1.9725918041511443, "grad_norm": 0.2808194160461426, "learning_rate": 1.3619808442412578e-07, "loss": 0.1863, "step": 7413 }, { "epoch": 1.972857903139968, "grad_norm": 0.2638983130455017, "learning_rate": 1.361821799806765e-07, "loss": 0.1997, "step": 7414 }, { "epoch": 1.973124002128792, "grad_norm": 0.27254918217658997, "learning_rate": 1.3616627448405785e-07, "loss": 0.1848, "step": 7415 }, { "epoch": 1.9733901011176158, "grad_norm": 0.2921788990497589, "learning_rate": 1.3615036793473288e-07, "loss": 0.1982, "step": 7416 }, { "epoch": 1.9736562001064395, "grad_norm": 0.28991928696632385, "learning_rate": 1.3613446033316455e-07, "loss": 0.202, "step": 7417 }, { "epoch": 1.9739222990952634, "grad_norm": 0.30020883679389954, "learning_rate": 1.3611855167981593e-07, "loss": 0.2148, "step": 7418 }, { "epoch": 1.9741883980840873, "grad_norm": 0.2737704813480377, "learning_rate": 1.3610264197515e-07, "loss": 0.1874, "step": 7419 }, { "epoch": 1.974454497072911, "grad_norm": 0.2749815285205841, "learning_rate": 1.3608673121962992e-07, "loss": 0.1916, "step": 7420 }, { "epoch": 1.9747205960617351, "grad_norm": 0.2957476079463959, "learning_rate": 1.360708194137188e-07, "loss": 0.1995, "step": 7421 }, { "epoch": 1.9749866950505588, "grad_norm": 0.2949954569339752, "learning_rate": 1.3605490655787976e-07, "loss": 0.202, "step": 7422 }, { "epoch": 1.9752527940393827, "grad_norm": 0.28421550989151, "learning_rate": 1.3603899265257606e-07, "loss": 0.1748, "step": 7423 }, { "epoch": 1.9755188930282066, "grad_norm": 0.29481932520866394, "learning_rate": 1.3602307769827082e-07, "loss": 0.1894, "step": 7424 }, { "epoch": 1.9757849920170303, "grad_norm": 0.364475816488266, "learning_rate": 1.3600716169542735e-07, "loss": 0.2282, "step": 7425 }, { "epoch": 1.9760510910058542, "grad_norm": 0.2801762521266937, "learning_rate": 1.3599124464450885e-07, "loss": 0.2087, "step": 7426 }, { "epoch": 1.976317189994678, "grad_norm": 0.2919997572898865, "learning_rate": 1.3597532654597871e-07, "loss": 0.2125, "step": 7427 }, { "epoch": 1.9765832889835018, "grad_norm": 0.28943267464637756, "learning_rate": 1.359594074003002e-07, "loss": 0.1936, "step": 7428 }, { "epoch": 1.9768493879723257, "grad_norm": 0.385952353477478, "learning_rate": 1.3594348720793672e-07, "loss": 0.1987, "step": 7429 }, { "epoch": 1.9771154869611496, "grad_norm": 0.2759706676006317, "learning_rate": 1.3592756596935163e-07, "loss": 0.1896, "step": 7430 }, { "epoch": 1.9773815859499733, "grad_norm": 0.2662688195705414, "learning_rate": 1.359116436850084e-07, "loss": 0.1991, "step": 7431 }, { "epoch": 1.9776476849387974, "grad_norm": 0.27716606855392456, "learning_rate": 1.3589572035537048e-07, "loss": 0.1995, "step": 7432 }, { "epoch": 1.977913783927621, "grad_norm": 0.27923986315727234, "learning_rate": 1.358797959809013e-07, "loss": 0.1914, "step": 7433 }, { "epoch": 1.9781798829164448, "grad_norm": 0.28797969222068787, "learning_rate": 1.3586387056206444e-07, "loss": 0.2175, "step": 7434 }, { "epoch": 1.9784459819052689, "grad_norm": 0.28746888041496277, "learning_rate": 1.3584794409932345e-07, "loss": 0.2052, "step": 7435 }, { "epoch": 1.9787120808940926, "grad_norm": 0.27767688035964966, "learning_rate": 1.358320165931419e-07, "loss": 0.1861, "step": 7436 }, { "epoch": 1.9789781798829165, "grad_norm": 0.3084064722061157, "learning_rate": 1.3581608804398336e-07, "loss": 0.1955, "step": 7437 }, { "epoch": 1.9792442788717404, "grad_norm": 0.2991097569465637, "learning_rate": 1.3580015845231144e-07, "loss": 0.1993, "step": 7438 }, { "epoch": 1.979510377860564, "grad_norm": 0.2991158664226532, "learning_rate": 1.3578422781858992e-07, "loss": 0.1934, "step": 7439 }, { "epoch": 1.979776476849388, "grad_norm": 0.3040710687637329, "learning_rate": 1.3576829614328242e-07, "loss": 0.2057, "step": 7440 }, { "epoch": 1.9800425758382119, "grad_norm": 0.2738858759403229, "learning_rate": 1.357523634268527e-07, "loss": 0.2025, "step": 7441 }, { "epoch": 1.9803086748270355, "grad_norm": 0.2809737026691437, "learning_rate": 1.357364296697645e-07, "loss": 0.1991, "step": 7442 }, { "epoch": 1.9805747738158594, "grad_norm": 0.30480554699897766, "learning_rate": 1.3572049487248162e-07, "loss": 0.2198, "step": 7443 }, { "epoch": 1.9808408728046834, "grad_norm": 0.27238035202026367, "learning_rate": 1.3570455903546791e-07, "loss": 0.2026, "step": 7444 }, { "epoch": 1.981106971793507, "grad_norm": 0.3116972744464874, "learning_rate": 1.3568862215918716e-07, "loss": 0.2112, "step": 7445 }, { "epoch": 1.9813730707823312, "grad_norm": 0.3050766885280609, "learning_rate": 1.3567268424410335e-07, "loss": 0.2105, "step": 7446 }, { "epoch": 1.9816391697711548, "grad_norm": 0.2828212380409241, "learning_rate": 1.3565674529068025e-07, "loss": 0.2013, "step": 7447 }, { "epoch": 1.9819052687599787, "grad_norm": 0.3137172758579254, "learning_rate": 1.356408052993819e-07, "loss": 0.2291, "step": 7448 }, { "epoch": 1.9821713677488026, "grad_norm": 0.329468697309494, "learning_rate": 1.3562486427067225e-07, "loss": 0.2059, "step": 7449 }, { "epoch": 1.9824374667376263, "grad_norm": 0.2838151156902313, "learning_rate": 1.356089222050153e-07, "loss": 0.1963, "step": 7450 }, { "epoch": 1.9827035657264502, "grad_norm": 0.2749573588371277, "learning_rate": 1.3559297910287506e-07, "loss": 0.2091, "step": 7451 }, { "epoch": 1.9829696647152741, "grad_norm": 0.2789728045463562, "learning_rate": 1.3557703496471564e-07, "loss": 0.205, "step": 7452 }, { "epoch": 1.9832357637040978, "grad_norm": 0.32095038890838623, "learning_rate": 1.355610897910011e-07, "loss": 0.2006, "step": 7453 }, { "epoch": 1.9835018626929217, "grad_norm": 0.28833186626434326, "learning_rate": 1.3554514358219558e-07, "loss": 0.2021, "step": 7454 }, { "epoch": 1.9837679616817456, "grad_norm": 0.2812528610229492, "learning_rate": 1.355291963387632e-07, "loss": 0.1838, "step": 7455 }, { "epoch": 1.9840340606705693, "grad_norm": 0.3063814043998718, "learning_rate": 1.355132480611682e-07, "loss": 0.2012, "step": 7456 }, { "epoch": 1.9843001596593934, "grad_norm": 0.28242209553718567, "learning_rate": 1.3549729874987473e-07, "loss": 0.1885, "step": 7457 }, { "epoch": 1.9845662586482171, "grad_norm": 0.3165386915206909, "learning_rate": 1.3548134840534707e-07, "loss": 0.2385, "step": 7458 }, { "epoch": 1.984832357637041, "grad_norm": 0.2992924451828003, "learning_rate": 1.3546539702804948e-07, "loss": 0.2058, "step": 7459 }, { "epoch": 1.985098456625865, "grad_norm": 0.2991969883441925, "learning_rate": 1.3544944461844626e-07, "loss": 0.196, "step": 7460 }, { "epoch": 1.9853645556146886, "grad_norm": 0.3043181896209717, "learning_rate": 1.3543349117700175e-07, "loss": 0.2237, "step": 7461 }, { "epoch": 1.9856306546035125, "grad_norm": 0.3213086426258087, "learning_rate": 1.3541753670418032e-07, "loss": 0.1968, "step": 7462 }, { "epoch": 1.9858967535923364, "grad_norm": 0.2739744186401367, "learning_rate": 1.3540158120044638e-07, "loss": 0.1721, "step": 7463 }, { "epoch": 1.98616285258116, "grad_norm": 0.282737672328949, "learning_rate": 1.3538562466626432e-07, "loss": 0.2032, "step": 7464 }, { "epoch": 1.986428951569984, "grad_norm": 0.28376099467277527, "learning_rate": 1.3536966710209858e-07, "loss": 0.2015, "step": 7465 }, { "epoch": 1.986695050558808, "grad_norm": 0.31686732172966003, "learning_rate": 1.353537085084137e-07, "loss": 0.1955, "step": 7466 }, { "epoch": 1.9869611495476316, "grad_norm": 0.2620345652103424, "learning_rate": 1.3533774888567412e-07, "loss": 0.1658, "step": 7467 }, { "epoch": 1.9872272485364557, "grad_norm": 0.29491689801216125, "learning_rate": 1.3532178823434447e-07, "loss": 0.1996, "step": 7468 }, { "epoch": 1.9874933475252794, "grad_norm": 0.26762768626213074, "learning_rate": 1.3530582655488926e-07, "loss": 0.1826, "step": 7469 }, { "epoch": 1.9877594465141033, "grad_norm": 0.2851411998271942, "learning_rate": 1.352898638477731e-07, "loss": 0.2011, "step": 7470 }, { "epoch": 1.9880255455029272, "grad_norm": 0.3062319755554199, "learning_rate": 1.3527390011346062e-07, "loss": 0.2032, "step": 7471 }, { "epoch": 1.9882916444917509, "grad_norm": 0.2991567552089691, "learning_rate": 1.352579353524165e-07, "loss": 0.2197, "step": 7472 }, { "epoch": 1.9885577434805748, "grad_norm": 0.28094711899757385, "learning_rate": 1.3524196956510545e-07, "loss": 0.2008, "step": 7473 }, { "epoch": 1.9888238424693987, "grad_norm": 0.3008536100387573, "learning_rate": 1.3522600275199217e-07, "loss": 0.2021, "step": 7474 }, { "epoch": 1.9890899414582224, "grad_norm": 0.32817214727401733, "learning_rate": 1.352100349135414e-07, "loss": 0.2137, "step": 7475 }, { "epoch": 1.9893560404470463, "grad_norm": 0.27068203687667847, "learning_rate": 1.3519406605021795e-07, "loss": 0.1957, "step": 7476 }, { "epoch": 1.9896221394358702, "grad_norm": 0.2642461955547333, "learning_rate": 1.3517809616248662e-07, "loss": 0.1914, "step": 7477 }, { "epoch": 1.9898882384246939, "grad_norm": 0.28697341680526733, "learning_rate": 1.351621252508122e-07, "loss": 0.2075, "step": 7478 }, { "epoch": 1.990154337413518, "grad_norm": 0.29903969168663025, "learning_rate": 1.3514615331565965e-07, "loss": 0.191, "step": 7479 }, { "epoch": 1.9904204364023417, "grad_norm": 0.29241639375686646, "learning_rate": 1.3513018035749384e-07, "loss": 0.2147, "step": 7480 }, { "epoch": 1.9906865353911654, "grad_norm": 0.2801326811313629, "learning_rate": 1.3511420637677966e-07, "loss": 0.2023, "step": 7481 }, { "epoch": 1.9909526343799895, "grad_norm": 0.28272947669029236, "learning_rate": 1.3509823137398215e-07, "loss": 0.1988, "step": 7482 }, { "epoch": 1.9912187333688132, "grad_norm": 0.3130277097225189, "learning_rate": 1.3508225534956623e-07, "loss": 0.219, "step": 7483 }, { "epoch": 1.991484832357637, "grad_norm": 0.2685030698776245, "learning_rate": 1.3506627830399695e-07, "loss": 0.1839, "step": 7484 }, { "epoch": 1.991750931346461, "grad_norm": 0.28347474336624146, "learning_rate": 1.3505030023773937e-07, "loss": 0.2123, "step": 7485 }, { "epoch": 1.9920170303352847, "grad_norm": 0.295502632856369, "learning_rate": 1.3503432115125855e-07, "loss": 0.2186, "step": 7486 }, { "epoch": 1.9922831293241086, "grad_norm": 0.2717786133289337, "learning_rate": 1.350183410450196e-07, "loss": 0.1767, "step": 7487 }, { "epoch": 1.9925492283129325, "grad_norm": 0.28242477774620056, "learning_rate": 1.350023599194877e-07, "loss": 0.2003, "step": 7488 }, { "epoch": 1.9928153273017561, "grad_norm": 0.3003600537776947, "learning_rate": 1.3498637777512793e-07, "loss": 0.2106, "step": 7489 }, { "epoch": 1.99308142629058, "grad_norm": 0.2948787212371826, "learning_rate": 1.3497039461240559e-07, "loss": 0.1908, "step": 7490 }, { "epoch": 1.993347525279404, "grad_norm": 0.31520238518714905, "learning_rate": 1.3495441043178583e-07, "loss": 0.2023, "step": 7491 }, { "epoch": 1.9936136242682276, "grad_norm": 0.27248644828796387, "learning_rate": 1.3493842523373394e-07, "loss": 0.1744, "step": 7492 }, { "epoch": 1.9938797232570518, "grad_norm": 0.2750701606273651, "learning_rate": 1.3492243901871522e-07, "loss": 0.1987, "step": 7493 }, { "epoch": 1.9941458222458754, "grad_norm": 0.2922378480434418, "learning_rate": 1.34906451787195e-07, "loss": 0.1952, "step": 7494 }, { "epoch": 1.9944119212346993, "grad_norm": 0.310983806848526, "learning_rate": 1.3489046353963857e-07, "loss": 0.2186, "step": 7495 }, { "epoch": 1.9946780202235233, "grad_norm": 0.2872074544429779, "learning_rate": 1.3487447427651136e-07, "loss": 0.1934, "step": 7496 }, { "epoch": 1.994944119212347, "grad_norm": 0.27586522698402405, "learning_rate": 1.3485848399827876e-07, "loss": 0.1954, "step": 7497 }, { "epoch": 1.9952102182011708, "grad_norm": 0.2938779592514038, "learning_rate": 1.3484249270540615e-07, "loss": 0.2059, "step": 7498 }, { "epoch": 1.9954763171899947, "grad_norm": 0.2751739025115967, "learning_rate": 1.3482650039835907e-07, "loss": 0.1858, "step": 7499 }, { "epoch": 1.9957424161788184, "grad_norm": 0.2638588845729828, "learning_rate": 1.3481050707760303e-07, "loss": 0.1915, "step": 7500 }, { "epoch": 1.9960085151676423, "grad_norm": 0.298049658536911, "learning_rate": 1.3479451274360346e-07, "loss": 0.2186, "step": 7501 }, { "epoch": 1.9962746141564662, "grad_norm": 0.28828805685043335, "learning_rate": 1.34778517396826e-07, "loss": 0.1766, "step": 7502 }, { "epoch": 1.99654071314529, "grad_norm": 0.31986722350120544, "learning_rate": 1.3476252103773617e-07, "loss": 0.2347, "step": 7503 }, { "epoch": 1.996806812134114, "grad_norm": 0.3915792405605316, "learning_rate": 1.3474652366679966e-07, "loss": 0.2167, "step": 7504 }, { "epoch": 1.9970729111229377, "grad_norm": 0.3021872937679291, "learning_rate": 1.34730525284482e-07, "loss": 0.2203, "step": 7505 }, { "epoch": 1.9973390101117616, "grad_norm": 0.3248370587825775, "learning_rate": 1.34714525891249e-07, "loss": 0.2237, "step": 7506 }, { "epoch": 1.9976051091005855, "grad_norm": 0.310994416475296, "learning_rate": 1.3469852548756625e-07, "loss": 0.2137, "step": 7507 }, { "epoch": 1.9978712080894092, "grad_norm": 0.2925804853439331, "learning_rate": 1.3468252407389948e-07, "loss": 0.194, "step": 7508 }, { "epoch": 1.9981373070782331, "grad_norm": 0.285359650850296, "learning_rate": 1.3466652165071454e-07, "loss": 0.2131, "step": 7509 }, { "epoch": 1.998403406067057, "grad_norm": 0.28169646859169006, "learning_rate": 1.3465051821847714e-07, "loss": 0.1882, "step": 7510 }, { "epoch": 1.9986695050558807, "grad_norm": 0.2785094678401947, "learning_rate": 1.3463451377765313e-07, "loss": 0.197, "step": 7511 }, { "epoch": 1.9989356040447046, "grad_norm": 0.2999937832355499, "learning_rate": 1.3461850832870835e-07, "loss": 0.2049, "step": 7512 }, { "epoch": 1.9992017030335285, "grad_norm": 0.2886216640472412, "learning_rate": 1.346025018721087e-07, "loss": 0.1968, "step": 7513 }, { "epoch": 1.9994678020223522, "grad_norm": 0.288033664226532, "learning_rate": 1.3458649440832004e-07, "loss": 0.2242, "step": 7514 }, { "epoch": 1.9997339010111763, "grad_norm": 0.3073088228702545, "learning_rate": 1.3457048593780834e-07, "loss": 0.2014, "step": 7515 }, { "epoch": 2.0, "grad_norm": 0.27889201045036316, "learning_rate": 1.3455447646103956e-07, "loss": 0.1909, "step": 7516 }, { "epoch": 2.0002660989888237, "grad_norm": 0.30836570262908936, "learning_rate": 1.3453846597847965e-07, "loss": 0.2113, "step": 7517 }, { "epoch": 2.000532197977648, "grad_norm": 0.32168784737586975, "learning_rate": 1.3452245449059473e-07, "loss": 0.2187, "step": 7518 }, { "epoch": 2.0007982969664715, "grad_norm": 0.30898502469062805, "learning_rate": 1.3450644199785078e-07, "loss": 0.1925, "step": 7519 }, { "epoch": 2.001064395955295, "grad_norm": 0.3157370090484619, "learning_rate": 1.344904285007139e-07, "loss": 0.2064, "step": 7520 }, { "epoch": 2.0013304949441193, "grad_norm": 0.28102096915245056, "learning_rate": 1.3447441399965016e-07, "loss": 0.2067, "step": 7521 }, { "epoch": 2.001596593932943, "grad_norm": 0.31640058755874634, "learning_rate": 1.344583984951258e-07, "loss": 0.2104, "step": 7522 }, { "epoch": 2.001862692921767, "grad_norm": 0.27478140592575073, "learning_rate": 1.344423819876069e-07, "loss": 0.215, "step": 7523 }, { "epoch": 2.002128791910591, "grad_norm": 0.27503257989883423, "learning_rate": 1.3442636447755965e-07, "loss": 0.1989, "step": 7524 }, { "epoch": 2.0023948908994145, "grad_norm": 0.29455801844596863, "learning_rate": 1.3441034596545038e-07, "loss": 0.222, "step": 7525 }, { "epoch": 2.0026609898882386, "grad_norm": 0.2734989821910858, "learning_rate": 1.3439432645174527e-07, "loss": 0.1872, "step": 7526 }, { "epoch": 2.0029270888770623, "grad_norm": 0.28645211458206177, "learning_rate": 1.3437830593691062e-07, "loss": 0.2014, "step": 7527 }, { "epoch": 2.003193187865886, "grad_norm": 0.3033584952354431, "learning_rate": 1.3436228442141273e-07, "loss": 0.1816, "step": 7528 }, { "epoch": 2.00345928685471, "grad_norm": 0.31870952248573303, "learning_rate": 1.3434626190571798e-07, "loss": 0.2232, "step": 7529 }, { "epoch": 2.0037253858435338, "grad_norm": 0.27028897404670715, "learning_rate": 1.343302383902927e-07, "loss": 0.1883, "step": 7530 }, { "epoch": 2.0039914848323575, "grad_norm": 0.2897983193397522, "learning_rate": 1.3431421387560337e-07, "loss": 0.1919, "step": 7531 }, { "epoch": 2.0042575838211816, "grad_norm": 0.2913813591003418, "learning_rate": 1.342981883621163e-07, "loss": 0.2246, "step": 7532 }, { "epoch": 2.0045236828100053, "grad_norm": 0.2703035771846771, "learning_rate": 1.3428216185029806e-07, "loss": 0.2, "step": 7533 }, { "epoch": 2.0047897817988294, "grad_norm": 0.28361228108406067, "learning_rate": 1.3426613434061509e-07, "loss": 0.1952, "step": 7534 }, { "epoch": 2.005055880787653, "grad_norm": 0.29696667194366455, "learning_rate": 1.342501058335339e-07, "loss": 0.207, "step": 7535 }, { "epoch": 2.0053219797764767, "grad_norm": 0.3077656030654907, "learning_rate": 1.3423407632952112e-07, "loss": 0.2046, "step": 7536 }, { "epoch": 2.005588078765301, "grad_norm": 0.29086577892303467, "learning_rate": 1.3421804582904323e-07, "loss": 0.2089, "step": 7537 }, { "epoch": 2.0058541777541246, "grad_norm": 0.3059023320674896, "learning_rate": 1.342020143325669e-07, "loss": 0.2176, "step": 7538 }, { "epoch": 2.0061202767429482, "grad_norm": 0.3026469945907593, "learning_rate": 1.3418598184055867e-07, "loss": 0.1922, "step": 7539 }, { "epoch": 2.0063863757317724, "grad_norm": 0.3002711236476898, "learning_rate": 1.3416994835348534e-07, "loss": 0.2146, "step": 7540 }, { "epoch": 2.006652474720596, "grad_norm": 0.2934146821498871, "learning_rate": 1.341539138718135e-07, "loss": 0.1945, "step": 7541 }, { "epoch": 2.0069185737094197, "grad_norm": 0.28192412853240967, "learning_rate": 1.341378783960099e-07, "loss": 0.1833, "step": 7542 }, { "epoch": 2.007184672698244, "grad_norm": 0.28012633323669434, "learning_rate": 1.341218419265413e-07, "loss": 0.2032, "step": 7543 }, { "epoch": 2.0074507716870675, "grad_norm": 0.2804059386253357, "learning_rate": 1.341058044638745e-07, "loss": 0.193, "step": 7544 }, { "epoch": 2.007716870675891, "grad_norm": 0.2803230583667755, "learning_rate": 1.340897660084763e-07, "loss": 0.1977, "step": 7545 }, { "epoch": 2.0079829696647153, "grad_norm": 0.3030903935432434, "learning_rate": 1.3407372656081347e-07, "loss": 0.2038, "step": 7546 }, { "epoch": 2.008249068653539, "grad_norm": 0.31412649154663086, "learning_rate": 1.34057686121353e-07, "loss": 0.2086, "step": 7547 }, { "epoch": 2.008515167642363, "grad_norm": 0.2857455313205719, "learning_rate": 1.3404164469056167e-07, "loss": 0.1838, "step": 7548 }, { "epoch": 2.008781266631187, "grad_norm": 0.2990646958351135, "learning_rate": 1.3402560226890646e-07, "loss": 0.1967, "step": 7549 }, { "epoch": 2.0090473656200105, "grad_norm": 0.2983144521713257, "learning_rate": 1.3400955885685433e-07, "loss": 0.2097, "step": 7550 }, { "epoch": 2.0093134646088346, "grad_norm": 0.2974988520145416, "learning_rate": 1.3399351445487222e-07, "loss": 0.2117, "step": 7551 }, { "epoch": 2.0095795635976583, "grad_norm": 0.2969135046005249, "learning_rate": 1.3397746906342717e-07, "loss": 0.1859, "step": 7552 }, { "epoch": 2.009845662586482, "grad_norm": 0.27036252617836, "learning_rate": 1.3396142268298624e-07, "loss": 0.1968, "step": 7553 }, { "epoch": 2.010111761575306, "grad_norm": 0.29473528265953064, "learning_rate": 1.3394537531401646e-07, "loss": 0.1994, "step": 7554 }, { "epoch": 2.01037786056413, "grad_norm": 0.3081591725349426, "learning_rate": 1.3392932695698494e-07, "loss": 0.1947, "step": 7555 }, { "epoch": 2.0106439595529535, "grad_norm": 0.27708905935287476, "learning_rate": 1.339132776123588e-07, "loss": 0.193, "step": 7556 }, { "epoch": 2.0109100585417776, "grad_norm": 0.27865317463874817, "learning_rate": 1.3389722728060523e-07, "loss": 0.197, "step": 7557 }, { "epoch": 2.0111761575306013, "grad_norm": 0.2786633372306824, "learning_rate": 1.338811759621914e-07, "loss": 0.1918, "step": 7558 }, { "epoch": 2.0114422565194254, "grad_norm": 0.3183973431587219, "learning_rate": 1.3386512365758446e-07, "loss": 0.2173, "step": 7559 }, { "epoch": 2.011708355508249, "grad_norm": 0.3486466407775879, "learning_rate": 1.3384907036725171e-07, "loss": 0.2034, "step": 7560 }, { "epoch": 2.011974454497073, "grad_norm": 0.3020719885826111, "learning_rate": 1.3383301609166044e-07, "loss": 0.2076, "step": 7561 }, { "epoch": 2.012240553485897, "grad_norm": 0.2904253900051117, "learning_rate": 1.338169608312779e-07, "loss": 0.2075, "step": 7562 }, { "epoch": 2.0125066524747206, "grad_norm": 0.3455037474632263, "learning_rate": 1.3380090458657144e-07, "loss": 0.2217, "step": 7563 }, { "epoch": 2.0127727514635443, "grad_norm": 0.2918993830680847, "learning_rate": 1.337848473580084e-07, "loss": 0.1849, "step": 7564 }, { "epoch": 2.0130388504523684, "grad_norm": 0.3254282772541046, "learning_rate": 1.337687891460562e-07, "loss": 0.2265, "step": 7565 }, { "epoch": 2.013304949441192, "grad_norm": 0.2864461839199066, "learning_rate": 1.337527299511822e-07, "loss": 0.2133, "step": 7566 }, { "epoch": 2.0135710484300158, "grad_norm": 0.26216959953308105, "learning_rate": 1.337366697738539e-07, "loss": 0.1937, "step": 7567 }, { "epoch": 2.01383714741884, "grad_norm": 0.2875281572341919, "learning_rate": 1.3372060861453873e-07, "loss": 0.2086, "step": 7568 }, { "epoch": 2.0141032464076636, "grad_norm": 0.2899344563484192, "learning_rate": 1.3370454647370417e-07, "loss": 0.2114, "step": 7569 }, { "epoch": 2.0143693453964877, "grad_norm": 0.28724396228790283, "learning_rate": 1.3368848335181779e-07, "loss": 0.2112, "step": 7570 }, { "epoch": 2.0146354443853114, "grad_norm": 0.3032990097999573, "learning_rate": 1.3367241924934714e-07, "loss": 0.2023, "step": 7571 }, { "epoch": 2.014901543374135, "grad_norm": 0.301383912563324, "learning_rate": 1.336563541667598e-07, "loss": 0.2067, "step": 7572 }, { "epoch": 2.015167642362959, "grad_norm": 0.2921728491783142, "learning_rate": 1.3364028810452335e-07, "loss": 0.1779, "step": 7573 }, { "epoch": 2.015433741351783, "grad_norm": 0.298110693693161, "learning_rate": 1.3362422106310548e-07, "loss": 0.198, "step": 7574 }, { "epoch": 2.0156998403406066, "grad_norm": 0.2848002016544342, "learning_rate": 1.336081530429738e-07, "loss": 0.1773, "step": 7575 }, { "epoch": 2.0159659393294307, "grad_norm": 0.3090902864933014, "learning_rate": 1.335920840445961e-07, "loss": 0.219, "step": 7576 }, { "epoch": 2.0162320383182544, "grad_norm": 0.32038313150405884, "learning_rate": 1.3357601406844005e-07, "loss": 0.233, "step": 7577 }, { "epoch": 2.016498137307078, "grad_norm": 0.272945761680603, "learning_rate": 1.335599431149734e-07, "loss": 0.1775, "step": 7578 }, { "epoch": 2.016764236295902, "grad_norm": 0.2890634536743164, "learning_rate": 1.3354387118466396e-07, "loss": 0.1846, "step": 7579 }, { "epoch": 2.017030335284726, "grad_norm": 0.2783821225166321, "learning_rate": 1.335277982779795e-07, "loss": 0.2088, "step": 7580 }, { "epoch": 2.0172964342735495, "grad_norm": 0.28100213408470154, "learning_rate": 1.335117243953879e-07, "loss": 0.2098, "step": 7581 }, { "epoch": 2.0175625332623737, "grad_norm": 0.29585492610931396, "learning_rate": 1.33495649537357e-07, "loss": 0.2033, "step": 7582 }, { "epoch": 2.0178286322511974, "grad_norm": 0.278510719537735, "learning_rate": 1.334795737043547e-07, "loss": 0.1997, "step": 7583 }, { "epoch": 2.0180947312400215, "grad_norm": 0.2944156527519226, "learning_rate": 1.3346349689684897e-07, "loss": 0.216, "step": 7584 }, { "epoch": 2.018360830228845, "grad_norm": 0.5774492025375366, "learning_rate": 1.3344741911530773e-07, "loss": 0.1976, "step": 7585 }, { "epoch": 2.018626929217669, "grad_norm": 0.2813994884490967, "learning_rate": 1.3343134036019895e-07, "loss": 0.1944, "step": 7586 }, { "epoch": 2.018893028206493, "grad_norm": 0.284906268119812, "learning_rate": 1.3341526063199066e-07, "loss": 0.2007, "step": 7587 }, { "epoch": 2.0191591271953166, "grad_norm": 0.27731233835220337, "learning_rate": 1.3339917993115092e-07, "loss": 0.2012, "step": 7588 }, { "epoch": 2.0194252261841403, "grad_norm": 0.2823810875415802, "learning_rate": 1.3338309825814773e-07, "loss": 0.1918, "step": 7589 }, { "epoch": 2.0196913251729645, "grad_norm": 0.2976570725440979, "learning_rate": 1.3336701561344925e-07, "loss": 0.187, "step": 7590 }, { "epoch": 2.019957424161788, "grad_norm": 0.2920086979866028, "learning_rate": 1.333509319975236e-07, "loss": 0.1887, "step": 7591 }, { "epoch": 2.020223523150612, "grad_norm": 0.283745139837265, "learning_rate": 1.3333484741083888e-07, "loss": 0.1997, "step": 7592 }, { "epoch": 2.020489622139436, "grad_norm": 0.2738461196422577, "learning_rate": 1.333187618538633e-07, "loss": 0.181, "step": 7593 }, { "epoch": 2.0207557211282596, "grad_norm": 0.26561737060546875, "learning_rate": 1.333026753270651e-07, "loss": 0.1967, "step": 7594 }, { "epoch": 2.0210218201170838, "grad_norm": 0.29707157611846924, "learning_rate": 1.3328658783091249e-07, "loss": 0.2167, "step": 7595 }, { "epoch": 2.0212879191059074, "grad_norm": 0.2991909086704254, "learning_rate": 1.3327049936587372e-07, "loss": 0.1942, "step": 7596 }, { "epoch": 2.021554018094731, "grad_norm": 0.27434709668159485, "learning_rate": 1.3325440993241714e-07, "loss": 0.1981, "step": 7597 }, { "epoch": 2.0218201170835552, "grad_norm": 0.293325275182724, "learning_rate": 1.3323831953101097e-07, "loss": 0.1901, "step": 7598 }, { "epoch": 2.022086216072379, "grad_norm": 0.30257707834243774, "learning_rate": 1.3322222816212367e-07, "loss": 0.1934, "step": 7599 }, { "epoch": 2.0223523150612026, "grad_norm": 0.3023972511291504, "learning_rate": 1.3320613582622354e-07, "loss": 0.2117, "step": 7600 }, { "epoch": 2.0226184140500267, "grad_norm": 0.30782613158226013, "learning_rate": 1.33190042523779e-07, "loss": 0.2073, "step": 7601 }, { "epoch": 2.0228845130388504, "grad_norm": 0.27418050169944763, "learning_rate": 1.331739482552585e-07, "loss": 0.2001, "step": 7602 }, { "epoch": 2.023150612027674, "grad_norm": 0.29605212807655334, "learning_rate": 1.3315785302113054e-07, "loss": 0.1962, "step": 7603 }, { "epoch": 2.0234167110164982, "grad_norm": 0.30340906977653503, "learning_rate": 1.3314175682186358e-07, "loss": 0.2022, "step": 7604 }, { "epoch": 2.023682810005322, "grad_norm": 0.30316632986068726, "learning_rate": 1.331256596579261e-07, "loss": 0.202, "step": 7605 }, { "epoch": 2.023948908994146, "grad_norm": 0.2732735276222229, "learning_rate": 1.3310956152978668e-07, "loss": 0.1759, "step": 7606 }, { "epoch": 2.0242150079829697, "grad_norm": 0.3009876608848572, "learning_rate": 1.330934624379139e-07, "loss": 0.2003, "step": 7607 }, { "epoch": 2.0244811069717934, "grad_norm": 0.27969861030578613, "learning_rate": 1.3307736238277637e-07, "loss": 0.1985, "step": 7608 }, { "epoch": 2.0247472059606175, "grad_norm": 0.27469658851623535, "learning_rate": 1.3306126136484268e-07, "loss": 0.21, "step": 7609 }, { "epoch": 2.025013304949441, "grad_norm": 0.2980394661426544, "learning_rate": 1.3304515938458152e-07, "loss": 0.2042, "step": 7610 }, { "epoch": 2.025279403938265, "grad_norm": 0.29014185070991516, "learning_rate": 1.3302905644246157e-07, "loss": 0.1933, "step": 7611 }, { "epoch": 2.025545502927089, "grad_norm": 0.29767122864723206, "learning_rate": 1.3301295253895154e-07, "loss": 0.1925, "step": 7612 }, { "epoch": 2.0258116019159127, "grad_norm": 0.2989773452281952, "learning_rate": 1.329968476745202e-07, "loss": 0.215, "step": 7613 }, { "epoch": 2.0260777009047364, "grad_norm": 0.2816234529018402, "learning_rate": 1.3298074184963628e-07, "loss": 0.1969, "step": 7614 }, { "epoch": 2.0263437998935605, "grad_norm": 0.28999024629592896, "learning_rate": 1.329646350647686e-07, "loss": 0.2033, "step": 7615 }, { "epoch": 2.026609898882384, "grad_norm": 0.2909587025642395, "learning_rate": 1.32948527320386e-07, "loss": 0.2053, "step": 7616 }, { "epoch": 2.0268759978712083, "grad_norm": 0.2914329469203949, "learning_rate": 1.3293241861695733e-07, "loss": 0.2053, "step": 7617 }, { "epoch": 2.027142096860032, "grad_norm": 0.26458823680877686, "learning_rate": 1.3291630895495143e-07, "loss": 0.1829, "step": 7618 }, { "epoch": 2.0274081958488557, "grad_norm": 0.3154368996620178, "learning_rate": 1.3290019833483724e-07, "loss": 0.1978, "step": 7619 }, { "epoch": 2.02767429483768, "grad_norm": 0.28553980588912964, "learning_rate": 1.3288408675708375e-07, "loss": 0.1969, "step": 7620 }, { "epoch": 2.0279403938265035, "grad_norm": 0.27974432706832886, "learning_rate": 1.328679742221598e-07, "loss": 0.1779, "step": 7621 }, { "epoch": 2.028206492815327, "grad_norm": 0.3009468615055084, "learning_rate": 1.3285186073053455e-07, "loss": 0.2033, "step": 7622 }, { "epoch": 2.0284725918041513, "grad_norm": 0.2751823365688324, "learning_rate": 1.3283574628267686e-07, "loss": 0.189, "step": 7623 }, { "epoch": 2.028738690792975, "grad_norm": 0.29051584005355835, "learning_rate": 1.3281963087905588e-07, "loss": 0.1957, "step": 7624 }, { "epoch": 2.0290047897817987, "grad_norm": 0.28392964601516724, "learning_rate": 1.3280351452014066e-07, "loss": 0.1946, "step": 7625 }, { "epoch": 2.029270888770623, "grad_norm": 0.3029007017612457, "learning_rate": 1.327873972064003e-07, "loss": 0.1991, "step": 7626 }, { "epoch": 2.0295369877594465, "grad_norm": 0.2726505696773529, "learning_rate": 1.3277127893830394e-07, "loss": 0.1805, "step": 7627 }, { "epoch": 2.02980308674827, "grad_norm": 0.2937769889831543, "learning_rate": 1.3275515971632078e-07, "loss": 0.1962, "step": 7628 }, { "epoch": 2.0300691857370943, "grad_norm": 0.3225167691707611, "learning_rate": 1.3273903954091992e-07, "loss": 0.1983, "step": 7629 }, { "epoch": 2.030335284725918, "grad_norm": 0.30871644616127014, "learning_rate": 1.3272291841257062e-07, "loss": 0.2274, "step": 7630 }, { "epoch": 2.030601383714742, "grad_norm": 0.3252830505371094, "learning_rate": 1.3270679633174217e-07, "loss": 0.205, "step": 7631 }, { "epoch": 2.0308674827035658, "grad_norm": 0.283449649810791, "learning_rate": 1.3269067329890378e-07, "loss": 0.2263, "step": 7632 }, { "epoch": 2.0311335816923894, "grad_norm": 0.27627426385879517, "learning_rate": 1.3267454931452474e-07, "loss": 0.1763, "step": 7633 }, { "epoch": 2.0313996806812136, "grad_norm": 0.31735262274742126, "learning_rate": 1.3265842437907445e-07, "loss": 0.2334, "step": 7634 }, { "epoch": 2.0316657796700373, "grad_norm": 0.2804131507873535, "learning_rate": 1.3264229849302222e-07, "loss": 0.1948, "step": 7635 }, { "epoch": 2.031931878658861, "grad_norm": 0.2791973352432251, "learning_rate": 1.3262617165683741e-07, "loss": 0.1892, "step": 7636 }, { "epoch": 2.032197977647685, "grad_norm": 0.2743712365627289, "learning_rate": 1.326100438709895e-07, "loss": 0.2003, "step": 7637 }, { "epoch": 2.0324640766365087, "grad_norm": 0.28663837909698486, "learning_rate": 1.3259391513594787e-07, "loss": 0.1682, "step": 7638 }, { "epoch": 2.0327301756253324, "grad_norm": 0.3149174153804779, "learning_rate": 1.3257778545218197e-07, "loss": 0.224, "step": 7639 }, { "epoch": 2.0329962746141566, "grad_norm": 0.2976604402065277, "learning_rate": 1.3256165482016134e-07, "loss": 0.1951, "step": 7640 }, { "epoch": 2.0332623736029802, "grad_norm": 0.28659024834632874, "learning_rate": 1.325455232403555e-07, "loss": 0.1968, "step": 7641 }, { "epoch": 2.0335284725918044, "grad_norm": 0.3043486177921295, "learning_rate": 1.3252939071323395e-07, "loss": 0.216, "step": 7642 }, { "epoch": 2.033794571580628, "grad_norm": 0.29674646258354187, "learning_rate": 1.325132572392663e-07, "loss": 0.1963, "step": 7643 }, { "epoch": 2.0340606705694517, "grad_norm": 0.27142134308815, "learning_rate": 1.324971228189222e-07, "loss": 0.1897, "step": 7644 }, { "epoch": 2.034326769558276, "grad_norm": 0.28554391860961914, "learning_rate": 1.3248098745267118e-07, "loss": 0.2054, "step": 7645 }, { "epoch": 2.0345928685470995, "grad_norm": 0.2799741327762604, "learning_rate": 1.3246485114098295e-07, "loss": 0.2071, "step": 7646 }, { "epoch": 2.034858967535923, "grad_norm": 0.2863723337650299, "learning_rate": 1.3244871388432721e-07, "loss": 0.19, "step": 7647 }, { "epoch": 2.0351250665247473, "grad_norm": 0.28160935640335083, "learning_rate": 1.3243257568317368e-07, "loss": 0.1984, "step": 7648 }, { "epoch": 2.035391165513571, "grad_norm": 0.2766771912574768, "learning_rate": 1.324164365379921e-07, "loss": 0.1868, "step": 7649 }, { "epoch": 2.0356572645023947, "grad_norm": 0.2689671218395233, "learning_rate": 1.3240029644925218e-07, "loss": 0.1834, "step": 7650 }, { "epoch": 2.035923363491219, "grad_norm": 0.28347107768058777, "learning_rate": 1.3238415541742375e-07, "loss": 0.1985, "step": 7651 }, { "epoch": 2.0361894624800425, "grad_norm": 0.28421831130981445, "learning_rate": 1.3236801344297666e-07, "loss": 0.2023, "step": 7652 }, { "epoch": 2.036455561468866, "grad_norm": 0.2756589353084564, "learning_rate": 1.3235187052638076e-07, "loss": 0.2064, "step": 7653 }, { "epoch": 2.0367216604576903, "grad_norm": 0.2747892439365387, "learning_rate": 1.3233572666810585e-07, "loss": 0.1907, "step": 7654 }, { "epoch": 2.036987759446514, "grad_norm": 0.28470495343208313, "learning_rate": 1.3231958186862196e-07, "loss": 0.2028, "step": 7655 }, { "epoch": 2.037253858435338, "grad_norm": 0.26769521832466125, "learning_rate": 1.323034361283989e-07, "loss": 0.208, "step": 7656 }, { "epoch": 2.037519957424162, "grad_norm": 0.29179906845092773, "learning_rate": 1.3228728944790675e-07, "loss": 0.2065, "step": 7657 }, { "epoch": 2.0377860564129855, "grad_norm": 0.2741677463054657, "learning_rate": 1.322711418276154e-07, "loss": 0.2117, "step": 7658 }, { "epoch": 2.0380521554018096, "grad_norm": 0.29316991567611694, "learning_rate": 1.3225499326799494e-07, "loss": 0.207, "step": 7659 }, { "epoch": 2.0383182543906333, "grad_norm": 0.2664419114589691, "learning_rate": 1.3223884376951536e-07, "loss": 0.1947, "step": 7660 }, { "epoch": 2.038584353379457, "grad_norm": 0.27677100896835327, "learning_rate": 1.322226933326467e-07, "loss": 0.1906, "step": 7661 }, { "epoch": 2.038850452368281, "grad_norm": 0.3131655752658844, "learning_rate": 1.3220654195785915e-07, "loss": 0.2071, "step": 7662 }, { "epoch": 2.039116551357105, "grad_norm": 0.28273898363113403, "learning_rate": 1.3219038964562276e-07, "loss": 0.1926, "step": 7663 }, { "epoch": 2.0393826503459285, "grad_norm": 0.3047980070114136, "learning_rate": 1.3217423639640773e-07, "loss": 0.2082, "step": 7664 }, { "epoch": 2.0396487493347526, "grad_norm": 0.2828019857406616, "learning_rate": 1.3215808221068422e-07, "loss": 0.2087, "step": 7665 }, { "epoch": 2.0399148483235763, "grad_norm": 0.29661279916763306, "learning_rate": 1.321419270889224e-07, "loss": 0.2106, "step": 7666 }, { "epoch": 2.0401809473124004, "grad_norm": 0.2856602072715759, "learning_rate": 1.3212577103159257e-07, "loss": 0.2105, "step": 7667 }, { "epoch": 2.040447046301224, "grad_norm": 0.2872137129306793, "learning_rate": 1.3210961403916494e-07, "loss": 0.1854, "step": 7668 }, { "epoch": 2.0407131452900478, "grad_norm": 0.2880684733390808, "learning_rate": 1.3209345611210984e-07, "loss": 0.1938, "step": 7669 }, { "epoch": 2.040979244278872, "grad_norm": 0.3157080113887787, "learning_rate": 1.3207729725089756e-07, "loss": 0.2062, "step": 7670 }, { "epoch": 2.0412453432676956, "grad_norm": 0.29222559928894043, "learning_rate": 1.3206113745599844e-07, "loss": 0.2098, "step": 7671 }, { "epoch": 2.0415114422565193, "grad_norm": 0.34474611282348633, "learning_rate": 1.3204497672788284e-07, "loss": 0.2216, "step": 7672 }, { "epoch": 2.0417775412453434, "grad_norm": 0.3234858810901642, "learning_rate": 1.320288150670212e-07, "loss": 0.2235, "step": 7673 }, { "epoch": 2.042043640234167, "grad_norm": 0.30523884296417236, "learning_rate": 1.3201265247388387e-07, "loss": 0.2078, "step": 7674 }, { "epoch": 2.0423097392229907, "grad_norm": 0.2869611382484436, "learning_rate": 1.3199648894894137e-07, "loss": 0.1938, "step": 7675 }, { "epoch": 2.042575838211815, "grad_norm": 0.2890915870666504, "learning_rate": 1.3198032449266417e-07, "loss": 0.208, "step": 7676 }, { "epoch": 2.0428419372006386, "grad_norm": 0.29002121090888977, "learning_rate": 1.3196415910552275e-07, "loss": 0.1932, "step": 7677 }, { "epoch": 2.0431080361894627, "grad_norm": 0.30021992325782776, "learning_rate": 1.3194799278798764e-07, "loss": 0.2045, "step": 7678 }, { "epoch": 2.0433741351782864, "grad_norm": 0.28627851605415344, "learning_rate": 1.3193182554052944e-07, "loss": 0.1886, "step": 7679 }, { "epoch": 2.04364023416711, "grad_norm": 0.4462837874889374, "learning_rate": 1.319156573636187e-07, "loss": 0.2353, "step": 7680 }, { "epoch": 2.043906333155934, "grad_norm": 0.30526986718177795, "learning_rate": 1.3189948825772603e-07, "loss": 0.2212, "step": 7681 }, { "epoch": 2.044172432144758, "grad_norm": 0.30313795804977417, "learning_rate": 1.3188331822332207e-07, "loss": 0.2189, "step": 7682 }, { "epoch": 2.0444385311335815, "grad_norm": 0.3035599887371063, "learning_rate": 1.3186714726087752e-07, "loss": 0.2003, "step": 7683 }, { "epoch": 2.0447046301224057, "grad_norm": 0.2828495502471924, "learning_rate": 1.3185097537086305e-07, "loss": 0.1796, "step": 7684 }, { "epoch": 2.0449707291112293, "grad_norm": 0.2857897877693176, "learning_rate": 1.318348025537494e-07, "loss": 0.2027, "step": 7685 }, { "epoch": 2.045236828100053, "grad_norm": 0.2908227741718292, "learning_rate": 1.3181862881000728e-07, "loss": 0.1896, "step": 7686 }, { "epoch": 2.045502927088877, "grad_norm": 0.29865318536758423, "learning_rate": 1.3180245414010752e-07, "loss": 0.2167, "step": 7687 }, { "epoch": 2.045769026077701, "grad_norm": 0.2853696346282959, "learning_rate": 1.3178627854452086e-07, "loss": 0.2066, "step": 7688 }, { "epoch": 2.046035125066525, "grad_norm": 0.3052525818347931, "learning_rate": 1.3177010202371821e-07, "loss": 0.2047, "step": 7689 }, { "epoch": 2.0463012240553486, "grad_norm": 0.2963128685951233, "learning_rate": 1.3175392457817036e-07, "loss": 0.2104, "step": 7690 }, { "epoch": 2.0465673230441723, "grad_norm": 0.2950897514820099, "learning_rate": 1.317377462083482e-07, "loss": 0.2101, "step": 7691 }, { "epoch": 2.0468334220329965, "grad_norm": 0.33552512526512146, "learning_rate": 1.3172156691472267e-07, "loss": 0.2133, "step": 7692 }, { "epoch": 2.04709952102182, "grad_norm": 0.29496654868125916, "learning_rate": 1.3170538669776467e-07, "loss": 0.1905, "step": 7693 }, { "epoch": 2.047365620010644, "grad_norm": 0.28386378288269043, "learning_rate": 1.3168920555794522e-07, "loss": 0.1995, "step": 7694 }, { "epoch": 2.047631718999468, "grad_norm": 0.26757335662841797, "learning_rate": 1.3167302349573524e-07, "loss": 0.177, "step": 7695 }, { "epoch": 2.0478978179882916, "grad_norm": 0.27558499574661255, "learning_rate": 1.316568405116058e-07, "loss": 0.1832, "step": 7696 }, { "epoch": 2.0481639169771153, "grad_norm": 0.30874910950660706, "learning_rate": 1.316406566060279e-07, "loss": 0.2002, "step": 7697 }, { "epoch": 2.0484300159659394, "grad_norm": 0.3158039152622223, "learning_rate": 1.316244717794727e-07, "loss": 0.2285, "step": 7698 }, { "epoch": 2.048696114954763, "grad_norm": 0.28360694646835327, "learning_rate": 1.3160828603241121e-07, "loss": 0.203, "step": 7699 }, { "epoch": 2.048962213943587, "grad_norm": 0.31709200143814087, "learning_rate": 1.315920993653146e-07, "loss": 0.2199, "step": 7700 }, { "epoch": 2.049228312932411, "grad_norm": 0.2621070444583893, "learning_rate": 1.31575911778654e-07, "loss": 0.192, "step": 7701 }, { "epoch": 2.0494944119212346, "grad_norm": 0.3069678544998169, "learning_rate": 1.3155972327290056e-07, "loss": 0.2156, "step": 7702 }, { "epoch": 2.0497605109100587, "grad_norm": 0.31303566694259644, "learning_rate": 1.3154353384852555e-07, "loss": 0.2015, "step": 7703 }, { "epoch": 2.0500266098988824, "grad_norm": 0.27939504384994507, "learning_rate": 1.3152734350600016e-07, "loss": 0.2164, "step": 7704 }, { "epoch": 2.050292708887706, "grad_norm": 0.2964824438095093, "learning_rate": 1.3151115224579566e-07, "loss": 0.1993, "step": 7705 }, { "epoch": 2.05055880787653, "grad_norm": 0.3361903727054596, "learning_rate": 1.3149496006838331e-07, "loss": 0.2099, "step": 7706 }, { "epoch": 2.050824906865354, "grad_norm": 0.27544254064559937, "learning_rate": 1.314787669742345e-07, "loss": 0.2014, "step": 7707 }, { "epoch": 2.0510910058541776, "grad_norm": 0.2948909103870392, "learning_rate": 1.314625729638205e-07, "loss": 0.1943, "step": 7708 }, { "epoch": 2.0513571048430017, "grad_norm": 0.2716939151287079, "learning_rate": 1.314463780376127e-07, "loss": 0.1976, "step": 7709 }, { "epoch": 2.0516232038318254, "grad_norm": 0.2808322012424469, "learning_rate": 1.3143018219608247e-07, "loss": 0.2064, "step": 7710 }, { "epoch": 2.051889302820649, "grad_norm": 0.30515581369400024, "learning_rate": 1.3141398543970125e-07, "loss": 0.2074, "step": 7711 }, { "epoch": 2.052155401809473, "grad_norm": 0.28984498977661133, "learning_rate": 1.3139778776894047e-07, "loss": 0.2055, "step": 7712 }, { "epoch": 2.052421500798297, "grad_norm": 0.31109151244163513, "learning_rate": 1.3138158918427162e-07, "loss": 0.2231, "step": 7713 }, { "epoch": 2.052687599787121, "grad_norm": 0.28084203600883484, "learning_rate": 1.313653896861662e-07, "loss": 0.1854, "step": 7714 }, { "epoch": 2.0529536987759447, "grad_norm": 0.3056643307209015, "learning_rate": 1.313491892750957e-07, "loss": 0.1944, "step": 7715 }, { "epoch": 2.0532197977647684, "grad_norm": 0.26183515787124634, "learning_rate": 1.3133298795153173e-07, "loss": 0.1727, "step": 7716 }, { "epoch": 2.0534858967535925, "grad_norm": 0.28789353370666504, "learning_rate": 1.313167857159458e-07, "loss": 0.201, "step": 7717 }, { "epoch": 2.053751995742416, "grad_norm": 0.29003003239631653, "learning_rate": 1.3130058256880956e-07, "loss": 0.1706, "step": 7718 }, { "epoch": 2.05401809473124, "grad_norm": 0.27345454692840576, "learning_rate": 1.3128437851059466e-07, "loss": 0.1916, "step": 7719 }, { "epoch": 2.054284193720064, "grad_norm": 0.2729375660419464, "learning_rate": 1.312681735417727e-07, "loss": 0.1829, "step": 7720 }, { "epoch": 2.0545502927088877, "grad_norm": 0.3223869800567627, "learning_rate": 1.3125196766281543e-07, "loss": 0.2329, "step": 7721 }, { "epoch": 2.0548163916977114, "grad_norm": 0.2941427230834961, "learning_rate": 1.312357608741945e-07, "loss": 0.2026, "step": 7722 }, { "epoch": 2.0550824906865355, "grad_norm": 0.3096591830253601, "learning_rate": 1.3121955317638167e-07, "loss": 0.2141, "step": 7723 }, { "epoch": 2.055348589675359, "grad_norm": 0.292936772108078, "learning_rate": 1.3120334456984869e-07, "loss": 0.2034, "step": 7724 }, { "epoch": 2.0556146886641833, "grad_norm": 0.31198397278785706, "learning_rate": 1.3118713505506742e-07, "loss": 0.2058, "step": 7725 }, { "epoch": 2.055880787653007, "grad_norm": 0.2765111029148102, "learning_rate": 1.311709246325096e-07, "loss": 0.1965, "step": 7726 }, { "epoch": 2.0561468866418307, "grad_norm": 0.2886355221271515, "learning_rate": 1.311547133026471e-07, "loss": 0.2159, "step": 7727 }, { "epoch": 2.0564129856306548, "grad_norm": 0.2967188060283661, "learning_rate": 1.3113850106595177e-07, "loss": 0.1972, "step": 7728 }, { "epoch": 2.0566790846194785, "grad_norm": 0.31474900245666504, "learning_rate": 1.3112228792289556e-07, "loss": 0.2241, "step": 7729 }, { "epoch": 2.056945183608302, "grad_norm": 0.29284146428108215, "learning_rate": 1.3110607387395031e-07, "loss": 0.1984, "step": 7730 }, { "epoch": 2.0572112825971263, "grad_norm": 0.3234371542930603, "learning_rate": 1.3108985891958805e-07, "loss": 0.2156, "step": 7731 }, { "epoch": 2.05747738158595, "grad_norm": 0.2707163691520691, "learning_rate": 1.3107364306028072e-07, "loss": 0.2051, "step": 7732 }, { "epoch": 2.0577434805747736, "grad_norm": 0.3049677014350891, "learning_rate": 1.310574262965003e-07, "loss": 0.2109, "step": 7733 }, { "epoch": 2.0580095795635978, "grad_norm": 0.29448264837265015, "learning_rate": 1.3104120862871886e-07, "loss": 0.2092, "step": 7734 }, { "epoch": 2.0582756785524214, "grad_norm": 0.30835530161857605, "learning_rate": 1.3102499005740844e-07, "loss": 0.2191, "step": 7735 }, { "epoch": 2.0585417775412456, "grad_norm": 0.306130975484848, "learning_rate": 1.310087705830411e-07, "loss": 0.1955, "step": 7736 }, { "epoch": 2.0588078765300692, "grad_norm": 0.29189467430114746, "learning_rate": 1.3099255020608894e-07, "loss": 0.1925, "step": 7737 }, { "epoch": 2.059073975518893, "grad_norm": 0.28483134508132935, "learning_rate": 1.3097632892702414e-07, "loss": 0.2178, "step": 7738 }, { "epoch": 2.059340074507717, "grad_norm": 0.29915469884872437, "learning_rate": 1.3096010674631887e-07, "loss": 0.207, "step": 7739 }, { "epoch": 2.0596061734965407, "grad_norm": 0.29057347774505615, "learning_rate": 1.3094388366444524e-07, "loss": 0.2082, "step": 7740 }, { "epoch": 2.0598722724853644, "grad_norm": 0.29643508791923523, "learning_rate": 1.3092765968187553e-07, "loss": 0.2117, "step": 7741 }, { "epoch": 2.0601383714741885, "grad_norm": 0.276438444852829, "learning_rate": 1.309114347990819e-07, "loss": 0.1874, "step": 7742 }, { "epoch": 2.0604044704630122, "grad_norm": 0.3105838894844055, "learning_rate": 1.3089520901653668e-07, "loss": 0.2046, "step": 7743 }, { "epoch": 2.060670569451836, "grad_norm": 0.29346850514411926, "learning_rate": 1.3087898233471217e-07, "loss": 0.1947, "step": 7744 }, { "epoch": 2.06093666844066, "grad_norm": 0.3015304505825043, "learning_rate": 1.3086275475408064e-07, "loss": 0.2008, "step": 7745 }, { "epoch": 2.0612027674294837, "grad_norm": 0.28753674030303955, "learning_rate": 1.3084652627511446e-07, "loss": 0.2119, "step": 7746 }, { "epoch": 2.0614688664183074, "grad_norm": 0.2934446334838867, "learning_rate": 1.30830296898286e-07, "loss": 0.2126, "step": 7747 }, { "epoch": 2.0617349654071315, "grad_norm": 0.266886830329895, "learning_rate": 1.3081406662406761e-07, "loss": 0.184, "step": 7748 }, { "epoch": 2.062001064395955, "grad_norm": 0.32266363501548767, "learning_rate": 1.307978354529318e-07, "loss": 0.1988, "step": 7749 }, { "epoch": 2.0622671633847793, "grad_norm": 0.30212175846099854, "learning_rate": 1.3078160338535097e-07, "loss": 0.2095, "step": 7750 }, { "epoch": 2.062533262373603, "grad_norm": 0.30672889947891235, "learning_rate": 1.307653704217975e-07, "loss": 0.2105, "step": 7751 }, { "epoch": 2.0627993613624267, "grad_norm": 0.30386629700660706, "learning_rate": 1.3074913656274407e-07, "loss": 0.2224, "step": 7752 }, { "epoch": 2.063065460351251, "grad_norm": 0.3115798830986023, "learning_rate": 1.307329018086631e-07, "loss": 0.2205, "step": 7753 }, { "epoch": 2.0633315593400745, "grad_norm": 0.2835429906845093, "learning_rate": 1.3071666616002713e-07, "loss": 0.195, "step": 7754 }, { "epoch": 2.063597658328898, "grad_norm": 0.28811541199684143, "learning_rate": 1.3070042961730876e-07, "loss": 0.2012, "step": 7755 }, { "epoch": 2.0638637573177223, "grad_norm": 0.29885852336883545, "learning_rate": 1.3068419218098062e-07, "loss": 0.1962, "step": 7756 }, { "epoch": 2.064129856306546, "grad_norm": 0.30470845103263855, "learning_rate": 1.306679538515153e-07, "loss": 0.182, "step": 7757 }, { "epoch": 2.0643959552953697, "grad_norm": 0.36877819895744324, "learning_rate": 1.3065171462938547e-07, "loss": 0.2195, "step": 7758 }, { "epoch": 2.064662054284194, "grad_norm": 0.31014811992645264, "learning_rate": 1.3063547451506382e-07, "loss": 0.2013, "step": 7759 }, { "epoch": 2.0649281532730175, "grad_norm": 0.33872851729393005, "learning_rate": 1.3061923350902306e-07, "loss": 0.1977, "step": 7760 }, { "epoch": 2.0651942522618416, "grad_norm": 0.2960881292819977, "learning_rate": 1.306029916117359e-07, "loss": 0.1969, "step": 7761 }, { "epoch": 2.0654603512506653, "grad_norm": 0.27401813864707947, "learning_rate": 1.3058674882367513e-07, "loss": 0.1921, "step": 7762 }, { "epoch": 2.065726450239489, "grad_norm": 0.295401006937027, "learning_rate": 1.3057050514531352e-07, "loss": 0.2084, "step": 7763 }, { "epoch": 2.065992549228313, "grad_norm": 0.4106268286705017, "learning_rate": 1.3055426057712384e-07, "loss": 0.1778, "step": 7764 }, { "epoch": 2.066258648217137, "grad_norm": 0.2794632613658905, "learning_rate": 1.30538015119579e-07, "loss": 0.1802, "step": 7765 }, { "epoch": 2.0665247472059605, "grad_norm": 0.2900896966457367, "learning_rate": 1.3052176877315186e-07, "loss": 0.2046, "step": 7766 }, { "epoch": 2.0667908461947846, "grad_norm": 0.28616541624069214, "learning_rate": 1.3050552153831525e-07, "loss": 0.1886, "step": 7767 }, { "epoch": 2.0670569451836083, "grad_norm": 0.2835589349269867, "learning_rate": 1.304892734155421e-07, "loss": 0.1974, "step": 7768 }, { "epoch": 2.067323044172432, "grad_norm": 0.29583391547203064, "learning_rate": 1.3047302440530536e-07, "loss": 0.2066, "step": 7769 }, { "epoch": 2.067589143161256, "grad_norm": 0.28936752676963806, "learning_rate": 1.3045677450807804e-07, "loss": 0.2035, "step": 7770 }, { "epoch": 2.0678552421500798, "grad_norm": 0.2965959310531616, "learning_rate": 1.304405237243331e-07, "loss": 0.1955, "step": 7771 }, { "epoch": 2.0681213411389034, "grad_norm": 0.2816261649131775, "learning_rate": 1.304242720545435e-07, "loss": 0.1957, "step": 7772 }, { "epoch": 2.0683874401277276, "grad_norm": 0.2841563820838928, "learning_rate": 1.3040801949918238e-07, "loss": 0.208, "step": 7773 }, { "epoch": 2.0686535391165513, "grad_norm": 0.2775135636329651, "learning_rate": 1.3039176605872272e-07, "loss": 0.1969, "step": 7774 }, { "epoch": 2.0689196381053754, "grad_norm": 0.2939140796661377, "learning_rate": 1.303755117336377e-07, "loss": 0.2011, "step": 7775 }, { "epoch": 2.069185737094199, "grad_norm": 0.30333104729652405, "learning_rate": 1.303592565244004e-07, "loss": 0.1888, "step": 7776 }, { "epoch": 2.0694518360830227, "grad_norm": 0.2631702721118927, "learning_rate": 1.3034300043148397e-07, "loss": 0.1881, "step": 7777 }, { "epoch": 2.069717935071847, "grad_norm": 0.27052202820777893, "learning_rate": 1.3032674345536157e-07, "loss": 0.1946, "step": 7778 }, { "epoch": 2.0699840340606706, "grad_norm": 0.2853175699710846, "learning_rate": 1.303104855965064e-07, "loss": 0.1914, "step": 7779 }, { "epoch": 2.0702501330494942, "grad_norm": 0.31648802757263184, "learning_rate": 1.3029422685539175e-07, "loss": 0.1894, "step": 7780 }, { "epoch": 2.0705162320383184, "grad_norm": 0.3161999583244324, "learning_rate": 1.3027796723249078e-07, "loss": 0.2074, "step": 7781 }, { "epoch": 2.070782331027142, "grad_norm": 0.2779003381729126, "learning_rate": 1.3026170672827678e-07, "loss": 0.2013, "step": 7782 }, { "epoch": 2.0710484300159657, "grad_norm": 0.2810528576374054, "learning_rate": 1.3024544534322308e-07, "loss": 0.1941, "step": 7783 }, { "epoch": 2.07131452900479, "grad_norm": 0.31843116879463196, "learning_rate": 1.3022918307780302e-07, "loss": 0.1922, "step": 7784 }, { "epoch": 2.0715806279936135, "grad_norm": 0.31464052200317383, "learning_rate": 1.3021291993248992e-07, "loss": 0.2046, "step": 7785 }, { "epoch": 2.0718467269824377, "grad_norm": 0.31180569529533386, "learning_rate": 1.3019665590775714e-07, "loss": 0.2181, "step": 7786 }, { "epoch": 2.0721128259712613, "grad_norm": 0.2835647761821747, "learning_rate": 1.3018039100407813e-07, "loss": 0.2074, "step": 7787 }, { "epoch": 2.072378924960085, "grad_norm": 0.2649685740470886, "learning_rate": 1.3016412522192632e-07, "loss": 0.1902, "step": 7788 }, { "epoch": 2.072645023948909, "grad_norm": 0.29342401027679443, "learning_rate": 1.3014785856177514e-07, "loss": 0.2107, "step": 7789 }, { "epoch": 2.072911122937733, "grad_norm": 0.28035902976989746, "learning_rate": 1.3013159102409805e-07, "loss": 0.1901, "step": 7790 }, { "epoch": 2.0731772219265565, "grad_norm": 0.27468815445899963, "learning_rate": 1.301153226093686e-07, "loss": 0.2181, "step": 7791 }, { "epoch": 2.0734433209153806, "grad_norm": 0.29591453075408936, "learning_rate": 1.300990533180603e-07, "loss": 0.2073, "step": 7792 }, { "epoch": 2.0737094199042043, "grad_norm": 0.2862091660499573, "learning_rate": 1.3008278315064674e-07, "loss": 0.1971, "step": 7793 }, { "epoch": 2.073975518893028, "grad_norm": 0.29863280057907104, "learning_rate": 1.3006651210760143e-07, "loss": 0.1908, "step": 7794 }, { "epoch": 2.074241617881852, "grad_norm": 0.2820762097835541, "learning_rate": 1.30050240189398e-07, "loss": 0.2052, "step": 7795 }, { "epoch": 2.074507716870676, "grad_norm": 0.27986940741539, "learning_rate": 1.3003396739651012e-07, "loss": 0.2157, "step": 7796 }, { "epoch": 2.0747738158595, "grad_norm": 0.31404218077659607, "learning_rate": 1.3001769372941146e-07, "loss": 0.2142, "step": 7797 }, { "epoch": 2.0750399148483236, "grad_norm": 0.31937268376350403, "learning_rate": 1.3000141918857564e-07, "loss": 0.2238, "step": 7798 }, { "epoch": 2.0753060138371473, "grad_norm": 0.287935733795166, "learning_rate": 1.2998514377447645e-07, "loss": 0.2171, "step": 7799 }, { "epoch": 2.0755721128259714, "grad_norm": 0.2691234350204468, "learning_rate": 1.2996886748758754e-07, "loss": 0.2123, "step": 7800 }, { "epoch": 2.075838211814795, "grad_norm": 0.3021089434623718, "learning_rate": 1.2995259032838276e-07, "loss": 0.1946, "step": 7801 }, { "epoch": 2.076104310803619, "grad_norm": 0.27903613448143005, "learning_rate": 1.2993631229733583e-07, "loss": 0.2061, "step": 7802 }, { "epoch": 2.076370409792443, "grad_norm": 0.3074961304664612, "learning_rate": 1.2992003339492054e-07, "loss": 0.1891, "step": 7803 }, { "epoch": 2.0766365087812666, "grad_norm": 1.6589686870574951, "learning_rate": 1.2990375362161077e-07, "loss": 0.226, "step": 7804 }, { "epoch": 2.0769026077700903, "grad_norm": 0.273584246635437, "learning_rate": 1.298874729778804e-07, "loss": 0.2022, "step": 7805 }, { "epoch": 2.0771687067589144, "grad_norm": 0.2794111967086792, "learning_rate": 1.2987119146420327e-07, "loss": 0.1999, "step": 7806 }, { "epoch": 2.077434805747738, "grad_norm": 0.2931794822216034, "learning_rate": 1.2985490908105333e-07, "loss": 0.2053, "step": 7807 }, { "epoch": 2.077700904736562, "grad_norm": 0.2995983958244324, "learning_rate": 1.298386258289045e-07, "loss": 0.2165, "step": 7808 }, { "epoch": 2.077967003725386, "grad_norm": 0.3102326989173889, "learning_rate": 1.2982234170823075e-07, "loss": 0.2198, "step": 7809 }, { "epoch": 2.0782331027142096, "grad_norm": 0.2685156762599945, "learning_rate": 1.2980605671950608e-07, "loss": 0.1931, "step": 7810 }, { "epoch": 2.0784992017030337, "grad_norm": 0.3139615058898926, "learning_rate": 1.2978977086320449e-07, "loss": 0.2106, "step": 7811 }, { "epoch": 2.0787653006918574, "grad_norm": 0.2852191627025604, "learning_rate": 1.297734841398e-07, "loss": 0.1895, "step": 7812 }, { "epoch": 2.079031399680681, "grad_norm": 0.3040030002593994, "learning_rate": 1.2975719654976672e-07, "loss": 0.217, "step": 7813 }, { "epoch": 2.079297498669505, "grad_norm": 0.2891275882720947, "learning_rate": 1.2974090809357867e-07, "loss": 0.2022, "step": 7814 }, { "epoch": 2.079563597658329, "grad_norm": 0.27947598695755005, "learning_rate": 1.2972461877171e-07, "loss": 0.1895, "step": 7815 }, { "epoch": 2.0798296966471526, "grad_norm": 0.272573858499527, "learning_rate": 1.297083285846349e-07, "loss": 0.2023, "step": 7816 }, { "epoch": 2.0800957956359767, "grad_norm": 0.2687355577945709, "learning_rate": 1.296920375328275e-07, "loss": 0.1992, "step": 7817 }, { "epoch": 2.0803618946248004, "grad_norm": 0.3155519962310791, "learning_rate": 1.2967574561676194e-07, "loss": 0.2244, "step": 7818 }, { "epoch": 2.080627993613624, "grad_norm": 0.28485366702079773, "learning_rate": 1.2965945283691247e-07, "loss": 0.1871, "step": 7819 }, { "epoch": 2.080894092602448, "grad_norm": 0.2887071669101715, "learning_rate": 1.296431591937534e-07, "loss": 0.1938, "step": 7820 }, { "epoch": 2.081160191591272, "grad_norm": 0.28721484541893005, "learning_rate": 1.2962686468775887e-07, "loss": 0.1964, "step": 7821 }, { "epoch": 2.081426290580096, "grad_norm": 0.2973549962043762, "learning_rate": 1.296105693194033e-07, "loss": 0.2136, "step": 7822 }, { "epoch": 2.0816923895689197, "grad_norm": 0.3143801987171173, "learning_rate": 1.2959427308916088e-07, "loss": 0.2073, "step": 7823 }, { "epoch": 2.0819584885577433, "grad_norm": 0.2853979766368866, "learning_rate": 1.2957797599750603e-07, "loss": 0.209, "step": 7824 }, { "epoch": 2.0822245875465675, "grad_norm": 0.3024251163005829, "learning_rate": 1.295616780449131e-07, "loss": 0.2044, "step": 7825 }, { "epoch": 2.082490686535391, "grad_norm": 0.3325158655643463, "learning_rate": 1.295453792318565e-07, "loss": 0.2083, "step": 7826 }, { "epoch": 2.082756785524215, "grad_norm": 0.29016223549842834, "learning_rate": 1.2952907955881057e-07, "loss": 0.1858, "step": 7827 }, { "epoch": 2.083022884513039, "grad_norm": 0.3034279942512512, "learning_rate": 1.2951277902624984e-07, "loss": 0.2202, "step": 7828 }, { "epoch": 2.0832889835018626, "grad_norm": 0.2793925106525421, "learning_rate": 1.2949647763464877e-07, "loss": 0.1931, "step": 7829 }, { "epoch": 2.0835550824906863, "grad_norm": 0.29277127981185913, "learning_rate": 1.2948017538448177e-07, "loss": 0.1947, "step": 7830 }, { "epoch": 2.0838211814795105, "grad_norm": 0.2989635765552521, "learning_rate": 1.2946387227622344e-07, "loss": 0.2062, "step": 7831 }, { "epoch": 2.084087280468334, "grad_norm": 0.27613896131515503, "learning_rate": 1.294475683103483e-07, "loss": 0.1997, "step": 7832 }, { "epoch": 2.0843533794571583, "grad_norm": 0.29818034172058105, "learning_rate": 1.2943126348733087e-07, "loss": 0.2124, "step": 7833 }, { "epoch": 2.084619478445982, "grad_norm": 0.3159143924713135, "learning_rate": 1.294149578076458e-07, "loss": 0.2169, "step": 7834 }, { "epoch": 2.0848855774348056, "grad_norm": 0.33364832401275635, "learning_rate": 1.293986512717677e-07, "loss": 0.2248, "step": 7835 }, { "epoch": 2.0851516764236298, "grad_norm": 0.32193565368652344, "learning_rate": 1.2938234388017114e-07, "loss": 0.2097, "step": 7836 }, { "epoch": 2.0854177754124534, "grad_norm": 0.2863803207874298, "learning_rate": 1.293660356333309e-07, "loss": 0.2051, "step": 7837 }, { "epoch": 2.085683874401277, "grad_norm": 0.2860652208328247, "learning_rate": 1.2934972653172159e-07, "loss": 0.2072, "step": 7838 }, { "epoch": 2.0859499733901012, "grad_norm": 0.28808164596557617, "learning_rate": 1.2933341657581793e-07, "loss": 0.1938, "step": 7839 }, { "epoch": 2.086216072378925, "grad_norm": 0.30458176136016846, "learning_rate": 1.2931710576609466e-07, "loss": 0.2002, "step": 7840 }, { "epoch": 2.0864821713677486, "grad_norm": 0.26410263776779175, "learning_rate": 1.2930079410302662e-07, "loss": 0.1931, "step": 7841 }, { "epoch": 2.0867482703565727, "grad_norm": 0.2652830481529236, "learning_rate": 1.2928448158708852e-07, "loss": 0.1821, "step": 7842 }, { "epoch": 2.0870143693453964, "grad_norm": 0.30240365862846375, "learning_rate": 1.292681682187552e-07, "loss": 0.215, "step": 7843 }, { "epoch": 2.0872804683342205, "grad_norm": 0.2739143371582031, "learning_rate": 1.2925185399850146e-07, "loss": 0.183, "step": 7844 }, { "epoch": 2.0875465673230442, "grad_norm": 0.29485419392585754, "learning_rate": 1.2923553892680225e-07, "loss": 0.1939, "step": 7845 }, { "epoch": 2.087812666311868, "grad_norm": 0.2868785858154297, "learning_rate": 1.2921922300413238e-07, "loss": 0.1942, "step": 7846 }, { "epoch": 2.088078765300692, "grad_norm": 0.32161745429039, "learning_rate": 1.2920290623096683e-07, "loss": 0.2205, "step": 7847 }, { "epoch": 2.0883448642895157, "grad_norm": 0.30230259895324707, "learning_rate": 1.2918658860778044e-07, "loss": 0.1934, "step": 7848 }, { "epoch": 2.0886109632783394, "grad_norm": 0.3125637471675873, "learning_rate": 1.291702701350483e-07, "loss": 0.2097, "step": 7849 }, { "epoch": 2.0888770622671635, "grad_norm": 0.2796918451786041, "learning_rate": 1.291539508132453e-07, "loss": 0.1977, "step": 7850 }, { "epoch": 2.089143161255987, "grad_norm": 0.2806087136268616, "learning_rate": 1.291376306428465e-07, "loss": 0.1802, "step": 7851 }, { "epoch": 2.089409260244811, "grad_norm": 0.2922167479991913, "learning_rate": 1.2912130962432693e-07, "loss": 0.2071, "step": 7852 }, { "epoch": 2.089675359233635, "grad_norm": 0.2888966500759125, "learning_rate": 1.2910498775816166e-07, "loss": 0.1918, "step": 7853 }, { "epoch": 2.0899414582224587, "grad_norm": 0.27026093006134033, "learning_rate": 1.2908866504482575e-07, "loss": 0.1792, "step": 7854 }, { "epoch": 2.090207557211283, "grad_norm": 0.3182189464569092, "learning_rate": 1.2907234148479433e-07, "loss": 0.1923, "step": 7855 }, { "epoch": 2.0904736562001065, "grad_norm": 0.29991644620895386, "learning_rate": 1.2905601707854256e-07, "loss": 0.1868, "step": 7856 }, { "epoch": 2.09073975518893, "grad_norm": 0.2972901463508606, "learning_rate": 1.2903969182654554e-07, "loss": 0.2021, "step": 7857 }, { "epoch": 2.0910058541777543, "grad_norm": 0.3035969138145447, "learning_rate": 1.290233657292785e-07, "loss": 0.196, "step": 7858 }, { "epoch": 2.091271953166578, "grad_norm": 0.2869938313961029, "learning_rate": 1.2900703878721666e-07, "loss": 0.1922, "step": 7859 }, { "epoch": 2.0915380521554017, "grad_norm": 0.2841357886791229, "learning_rate": 1.289907110008352e-07, "loss": 0.2066, "step": 7860 }, { "epoch": 2.091804151144226, "grad_norm": 0.2948966920375824, "learning_rate": 1.2897438237060948e-07, "loss": 0.2208, "step": 7861 }, { "epoch": 2.0920702501330495, "grad_norm": 0.3229195773601532, "learning_rate": 1.2895805289701468e-07, "loss": 0.1823, "step": 7862 }, { "epoch": 2.092336349121873, "grad_norm": 0.28894874453544617, "learning_rate": 1.2894172258052617e-07, "loss": 0.2004, "step": 7863 }, { "epoch": 2.0926024481106973, "grad_norm": 0.29719069600105286, "learning_rate": 1.289253914216192e-07, "loss": 0.2227, "step": 7864 }, { "epoch": 2.092868547099521, "grad_norm": 0.2826484143733978, "learning_rate": 1.2890905942076926e-07, "loss": 0.1952, "step": 7865 }, { "epoch": 2.0931346460883447, "grad_norm": 0.30066707730293274, "learning_rate": 1.2889272657845164e-07, "loss": 0.2024, "step": 7866 }, { "epoch": 2.093400745077169, "grad_norm": 0.2767598330974579, "learning_rate": 1.2887639289514176e-07, "loss": 0.1965, "step": 7867 }, { "epoch": 2.0936668440659925, "grad_norm": 0.28746792674064636, "learning_rate": 1.2886005837131505e-07, "loss": 0.2051, "step": 7868 }, { "epoch": 2.0939329430548166, "grad_norm": 0.2836889326572418, "learning_rate": 1.28843723007447e-07, "loss": 0.1972, "step": 7869 }, { "epoch": 2.0941990420436403, "grad_norm": 0.33851009607315063, "learning_rate": 1.2882738680401306e-07, "loss": 0.2087, "step": 7870 }, { "epoch": 2.094465141032464, "grad_norm": 0.3001492917537689, "learning_rate": 1.2881104976148873e-07, "loss": 0.2064, "step": 7871 }, { "epoch": 2.094731240021288, "grad_norm": 0.3192121982574463, "learning_rate": 1.2879471188034955e-07, "loss": 0.2039, "step": 7872 }, { "epoch": 2.0949973390101118, "grad_norm": 0.3113660216331482, "learning_rate": 1.2877837316107106e-07, "loss": 0.2008, "step": 7873 }, { "epoch": 2.0952634379989354, "grad_norm": 0.2973521053791046, "learning_rate": 1.2876203360412887e-07, "loss": 0.1832, "step": 7874 }, { "epoch": 2.0955295369877596, "grad_norm": 0.2862984538078308, "learning_rate": 1.2874569320999858e-07, "loss": 0.1932, "step": 7875 }, { "epoch": 2.0957956359765832, "grad_norm": 0.33876869082450867, "learning_rate": 1.2872935197915578e-07, "loss": 0.2316, "step": 7876 }, { "epoch": 2.096061734965407, "grad_norm": 0.2787758409976959, "learning_rate": 1.2871300991207613e-07, "loss": 0.1846, "step": 7877 }, { "epoch": 2.096327833954231, "grad_norm": 0.2781781852245331, "learning_rate": 1.286966670092353e-07, "loss": 0.1897, "step": 7878 }, { "epoch": 2.0965939329430547, "grad_norm": 0.27182766795158386, "learning_rate": 1.2868032327110903e-07, "loss": 0.1901, "step": 7879 }, { "epoch": 2.096860031931879, "grad_norm": 0.311836838722229, "learning_rate": 1.2866397869817301e-07, "loss": 0.209, "step": 7880 }, { "epoch": 2.0971261309207025, "grad_norm": 0.29063767194747925, "learning_rate": 1.2864763329090298e-07, "loss": 0.2082, "step": 7881 }, { "epoch": 2.0973922299095262, "grad_norm": 0.29857975244522095, "learning_rate": 1.2863128704977474e-07, "loss": 0.202, "step": 7882 }, { "epoch": 2.0976583288983504, "grad_norm": 0.31641870737075806, "learning_rate": 1.2861493997526408e-07, "loss": 0.2042, "step": 7883 }, { "epoch": 2.097924427887174, "grad_norm": 0.2761826515197754, "learning_rate": 1.285985920678468e-07, "loss": 0.2032, "step": 7884 }, { "epoch": 2.0981905268759977, "grad_norm": 0.27382606267929077, "learning_rate": 1.2858224332799878e-07, "loss": 0.201, "step": 7885 }, { "epoch": 2.098456625864822, "grad_norm": 0.2922814190387726, "learning_rate": 1.2856589375619586e-07, "loss": 0.2083, "step": 7886 }, { "epoch": 2.0987227248536455, "grad_norm": 0.38669490814208984, "learning_rate": 1.2854954335291395e-07, "loss": 0.183, "step": 7887 }, { "epoch": 2.098988823842469, "grad_norm": 0.28528308868408203, "learning_rate": 1.2853319211862896e-07, "loss": 0.2102, "step": 7888 }, { "epoch": 2.0992549228312933, "grad_norm": 0.2611343562602997, "learning_rate": 1.2851684005381683e-07, "loss": 0.1828, "step": 7889 }, { "epoch": 2.099521021820117, "grad_norm": 0.2766665518283844, "learning_rate": 1.2850048715895353e-07, "loss": 0.2048, "step": 7890 }, { "epoch": 2.0997871208089407, "grad_norm": 0.2809724509716034, "learning_rate": 1.2848413343451502e-07, "loss": 0.1956, "step": 7891 }, { "epoch": 2.100053219797765, "grad_norm": 0.27756017446517944, "learning_rate": 1.284677788809774e-07, "loss": 0.186, "step": 7892 }, { "epoch": 2.1003193187865885, "grad_norm": 0.282586932182312, "learning_rate": 1.2845142349881663e-07, "loss": 0.1943, "step": 7893 }, { "epoch": 2.1005854177754126, "grad_norm": 0.2803298532962799, "learning_rate": 1.284350672885088e-07, "loss": 0.1859, "step": 7894 }, { "epoch": 2.1008515167642363, "grad_norm": 0.2952745258808136, "learning_rate": 1.2841871025052994e-07, "loss": 0.1998, "step": 7895 }, { "epoch": 2.10111761575306, "grad_norm": 0.3432902991771698, "learning_rate": 1.2840235238535626e-07, "loss": 0.1767, "step": 7896 }, { "epoch": 2.101383714741884, "grad_norm": 0.28263401985168457, "learning_rate": 1.2838599369346386e-07, "loss": 0.202, "step": 7897 }, { "epoch": 2.101649813730708, "grad_norm": 0.2712894678115845, "learning_rate": 1.2836963417532886e-07, "loss": 0.1912, "step": 7898 }, { "epoch": 2.1019159127195315, "grad_norm": 0.28770384192466736, "learning_rate": 1.2835327383142747e-07, "loss": 0.1989, "step": 7899 }, { "epoch": 2.1021820117083556, "grad_norm": 0.2753477394580841, "learning_rate": 1.2833691266223592e-07, "loss": 0.1982, "step": 7900 }, { "epoch": 2.1024481106971793, "grad_norm": 0.28644850850105286, "learning_rate": 1.2832055066823036e-07, "loss": 0.2066, "step": 7901 }, { "epoch": 2.1027142096860034, "grad_norm": 0.27388131618499756, "learning_rate": 1.2830418784988718e-07, "loss": 0.1968, "step": 7902 }, { "epoch": 2.102980308674827, "grad_norm": 0.28776735067367554, "learning_rate": 1.282878242076826e-07, "loss": 0.1925, "step": 7903 }, { "epoch": 2.103246407663651, "grad_norm": 0.2960626482963562, "learning_rate": 1.2827145974209284e-07, "loss": 0.2007, "step": 7904 }, { "epoch": 2.103512506652475, "grad_norm": 0.29624155163764954, "learning_rate": 1.2825509445359427e-07, "loss": 0.1939, "step": 7905 }, { "epoch": 2.1037786056412986, "grad_norm": 0.27953866124153137, "learning_rate": 1.2823872834266335e-07, "loss": 0.1841, "step": 7906 }, { "epoch": 2.1040447046301223, "grad_norm": 0.2853677570819855, "learning_rate": 1.282223614097763e-07, "loss": 0.1736, "step": 7907 }, { "epoch": 2.1043108036189464, "grad_norm": 0.35609549283981323, "learning_rate": 1.2820599365540961e-07, "loss": 0.2051, "step": 7908 }, { "epoch": 2.10457690260777, "grad_norm": 0.29583829641342163, "learning_rate": 1.2818962508003966e-07, "loss": 0.2096, "step": 7909 }, { "epoch": 2.1048430015965938, "grad_norm": 0.297038197517395, "learning_rate": 1.2817325568414297e-07, "loss": 0.2036, "step": 7910 }, { "epoch": 2.105109100585418, "grad_norm": 0.2874082922935486, "learning_rate": 1.281568854681959e-07, "loss": 0.1916, "step": 7911 }, { "epoch": 2.1053751995742416, "grad_norm": 0.2805434465408325, "learning_rate": 1.2814051443267504e-07, "loss": 0.1799, "step": 7912 }, { "epoch": 2.1056412985630653, "grad_norm": 0.27634313702583313, "learning_rate": 1.2812414257805687e-07, "loss": 0.1825, "step": 7913 }, { "epoch": 2.1059073975518894, "grad_norm": 0.28852590918540955, "learning_rate": 1.2810776990481794e-07, "loss": 0.2223, "step": 7914 }, { "epoch": 2.106173496540713, "grad_norm": 0.2943896949291229, "learning_rate": 1.280913964134348e-07, "loss": 0.2069, "step": 7915 }, { "epoch": 2.106439595529537, "grad_norm": 0.2811870276927948, "learning_rate": 1.2807502210438404e-07, "loss": 0.2111, "step": 7916 }, { "epoch": 2.106705694518361, "grad_norm": 0.30628645420074463, "learning_rate": 1.2805864697814227e-07, "loss": 0.2111, "step": 7917 }, { "epoch": 2.1069717935071846, "grad_norm": 0.30956852436065674, "learning_rate": 1.2804227103518618e-07, "loss": 0.2191, "step": 7918 }, { "epoch": 2.1072378924960087, "grad_norm": 0.2782313823699951, "learning_rate": 1.2802589427599234e-07, "loss": 0.1852, "step": 7919 }, { "epoch": 2.1075039914848324, "grad_norm": 0.29185983538627625, "learning_rate": 1.280095167010375e-07, "loss": 0.193, "step": 7920 }, { "epoch": 2.107770090473656, "grad_norm": 0.28291887044906616, "learning_rate": 1.2799313831079838e-07, "loss": 0.1833, "step": 7921 }, { "epoch": 2.10803618946248, "grad_norm": 0.3018980622291565, "learning_rate": 1.2797675910575167e-07, "loss": 0.209, "step": 7922 }, { "epoch": 2.108302288451304, "grad_norm": 0.29075509309768677, "learning_rate": 1.279603790863742e-07, "loss": 0.1895, "step": 7923 }, { "epoch": 2.1085683874401275, "grad_norm": 0.280234158039093, "learning_rate": 1.2794399825314264e-07, "loss": 0.1911, "step": 7924 }, { "epoch": 2.1088344864289517, "grad_norm": 0.2798197865486145, "learning_rate": 1.279276166065338e-07, "loss": 0.1791, "step": 7925 }, { "epoch": 2.1091005854177753, "grad_norm": 0.2865909934043884, "learning_rate": 1.2791123414702463e-07, "loss": 0.1976, "step": 7926 }, { "epoch": 2.1093666844065995, "grad_norm": 0.26274845004081726, "learning_rate": 1.2789485087509187e-07, "loss": 0.1787, "step": 7927 }, { "epoch": 2.109632783395423, "grad_norm": 0.32423657178878784, "learning_rate": 1.2787846679121242e-07, "loss": 0.2098, "step": 7928 }, { "epoch": 2.109898882384247, "grad_norm": 0.3083861768245697, "learning_rate": 1.2786208189586322e-07, "loss": 0.2222, "step": 7929 }, { "epoch": 2.110164981373071, "grad_norm": 0.2867703139781952, "learning_rate": 1.2784569618952113e-07, "loss": 0.2103, "step": 7930 }, { "epoch": 2.1104310803618946, "grad_norm": 0.2820870876312256, "learning_rate": 1.2782930967266315e-07, "loss": 0.1914, "step": 7931 }, { "epoch": 2.1106971793507183, "grad_norm": 0.26677918434143066, "learning_rate": 1.2781292234576622e-07, "loss": 0.1801, "step": 7932 }, { "epoch": 2.1109632783395424, "grad_norm": 0.2799242436885834, "learning_rate": 1.2779653420930731e-07, "loss": 0.1851, "step": 7933 }, { "epoch": 2.111229377328366, "grad_norm": 0.2874021530151367, "learning_rate": 1.277801452637635e-07, "loss": 0.209, "step": 7934 }, { "epoch": 2.11149547631719, "grad_norm": 0.28226014971733093, "learning_rate": 1.2776375550961175e-07, "loss": 0.197, "step": 7935 }, { "epoch": 2.111761575306014, "grad_norm": 0.31384289264678955, "learning_rate": 1.277473649473292e-07, "loss": 0.219, "step": 7936 }, { "epoch": 2.1120276742948376, "grad_norm": 0.26577168703079224, "learning_rate": 1.2773097357739287e-07, "loss": 0.1779, "step": 7937 }, { "epoch": 2.1122937732836613, "grad_norm": 0.2805081009864807, "learning_rate": 1.2771458140027994e-07, "loss": 0.2038, "step": 7938 }, { "epoch": 2.1125598722724854, "grad_norm": 0.2986027002334595, "learning_rate": 1.276981884164675e-07, "loss": 0.2127, "step": 7939 }, { "epoch": 2.112825971261309, "grad_norm": 0.2634809613227844, "learning_rate": 1.2768179462643268e-07, "loss": 0.1828, "step": 7940 }, { "epoch": 2.1130920702501332, "grad_norm": 0.25375452637672424, "learning_rate": 1.276654000306527e-07, "loss": 0.1672, "step": 7941 }, { "epoch": 2.113358169238957, "grad_norm": 0.2827334403991699, "learning_rate": 1.2764900462960477e-07, "loss": 0.1904, "step": 7942 }, { "epoch": 2.1136242682277806, "grad_norm": 0.2979433536529541, "learning_rate": 1.276326084237661e-07, "loss": 0.2031, "step": 7943 }, { "epoch": 2.1138903672166047, "grad_norm": 0.2663734257221222, "learning_rate": 1.27616211413614e-07, "loss": 0.1916, "step": 7944 }, { "epoch": 2.1141564662054284, "grad_norm": 0.3043944835662842, "learning_rate": 1.275998135996256e-07, "loss": 0.2143, "step": 7945 }, { "epoch": 2.114422565194252, "grad_norm": 0.2816953659057617, "learning_rate": 1.2758341498227833e-07, "loss": 0.1921, "step": 7946 }, { "epoch": 2.114688664183076, "grad_norm": 0.2672521471977234, "learning_rate": 1.275670155620495e-07, "loss": 0.188, "step": 7947 }, { "epoch": 2.1149547631719, "grad_norm": 0.27957138419151306, "learning_rate": 1.2755061533941638e-07, "loss": 0.1984, "step": 7948 }, { "epoch": 2.1152208621607236, "grad_norm": 0.3082587718963623, "learning_rate": 1.2753421431485638e-07, "loss": 0.202, "step": 7949 }, { "epoch": 2.1154869611495477, "grad_norm": 0.28561949729919434, "learning_rate": 1.275178124888469e-07, "loss": 0.2001, "step": 7950 }, { "epoch": 2.1157530601383714, "grad_norm": 0.29784539341926575, "learning_rate": 1.2750140986186537e-07, "loss": 0.2081, "step": 7951 }, { "epoch": 2.1160191591271955, "grad_norm": 0.2631798982620239, "learning_rate": 1.2748500643438918e-07, "loss": 0.1867, "step": 7952 }, { "epoch": 2.116285258116019, "grad_norm": 0.2909848690032959, "learning_rate": 1.2746860220689582e-07, "loss": 0.1915, "step": 7953 }, { "epoch": 2.116551357104843, "grad_norm": 0.2699466645717621, "learning_rate": 1.274521971798628e-07, "loss": 0.1874, "step": 7954 }, { "epoch": 2.116817456093667, "grad_norm": 0.32675015926361084, "learning_rate": 1.2743579135376759e-07, "loss": 0.2232, "step": 7955 }, { "epoch": 2.1170835550824907, "grad_norm": 0.2952888309955597, "learning_rate": 1.274193847290877e-07, "loss": 0.1895, "step": 7956 }, { "epoch": 2.1173496540713144, "grad_norm": 0.2899588644504547, "learning_rate": 1.2740297730630073e-07, "loss": 0.1888, "step": 7957 }, { "epoch": 2.1176157530601385, "grad_norm": 0.3011763393878937, "learning_rate": 1.2738656908588424e-07, "loss": 0.1933, "step": 7958 }, { "epoch": 2.117881852048962, "grad_norm": 0.28312402963638306, "learning_rate": 1.2737016006831583e-07, "loss": 0.1975, "step": 7959 }, { "epoch": 2.118147951037786, "grad_norm": 0.29674816131591797, "learning_rate": 1.2735375025407315e-07, "loss": 0.2018, "step": 7960 }, { "epoch": 2.11841405002661, "grad_norm": 0.31934231519699097, "learning_rate": 1.273373396436338e-07, "loss": 0.1991, "step": 7961 }, { "epoch": 2.1186801490154337, "grad_norm": 0.29530513286590576, "learning_rate": 1.2732092823747546e-07, "loss": 0.2004, "step": 7962 }, { "epoch": 2.118946248004258, "grad_norm": 0.2806771695613861, "learning_rate": 1.273045160360759e-07, "loss": 0.1993, "step": 7963 }, { "epoch": 2.1192123469930815, "grad_norm": 0.2741929292678833, "learning_rate": 1.2728810303991272e-07, "loss": 0.1998, "step": 7964 }, { "epoch": 2.119478445981905, "grad_norm": 0.2817156910896301, "learning_rate": 1.272716892494637e-07, "loss": 0.1948, "step": 7965 }, { "epoch": 2.1197445449707293, "grad_norm": 0.2808346748352051, "learning_rate": 1.2725527466520666e-07, "loss": 0.1988, "step": 7966 }, { "epoch": 2.120010643959553, "grad_norm": 0.27715685963630676, "learning_rate": 1.2723885928761934e-07, "loss": 0.1983, "step": 7967 }, { "epoch": 2.1202767429483766, "grad_norm": 0.36156073212623596, "learning_rate": 1.272224431171795e-07, "loss": 0.1993, "step": 7968 }, { "epoch": 2.1205428419372008, "grad_norm": 0.3713735342025757, "learning_rate": 1.2720602615436508e-07, "loss": 0.2064, "step": 7969 }, { "epoch": 2.1208089409260245, "grad_norm": 0.26496410369873047, "learning_rate": 1.2718960839965386e-07, "loss": 0.1854, "step": 7970 }, { "epoch": 2.121075039914848, "grad_norm": 0.31481990218162537, "learning_rate": 1.2717318985352373e-07, "loss": 0.1987, "step": 7971 }, { "epoch": 2.1213411389036723, "grad_norm": 0.28699159622192383, "learning_rate": 1.2715677051645258e-07, "loss": 0.1973, "step": 7972 }, { "epoch": 2.121607237892496, "grad_norm": 0.2832048833370209, "learning_rate": 1.271403503889184e-07, "loss": 0.1949, "step": 7973 }, { "epoch": 2.12187333688132, "grad_norm": 0.3188285827636719, "learning_rate": 1.2712392947139906e-07, "loss": 0.2095, "step": 7974 }, { "epoch": 2.1221394358701438, "grad_norm": 0.2919365167617798, "learning_rate": 1.2710750776437255e-07, "loss": 0.2032, "step": 7975 }, { "epoch": 2.1224055348589674, "grad_norm": 0.2863491475582123, "learning_rate": 1.2709108526831688e-07, "loss": 0.1808, "step": 7976 }, { "epoch": 2.1226716338477916, "grad_norm": 0.2788983881473541, "learning_rate": 1.2707466198371006e-07, "loss": 0.1908, "step": 7977 }, { "epoch": 2.1229377328366152, "grad_norm": 0.27127519249916077, "learning_rate": 1.2705823791103014e-07, "loss": 0.1911, "step": 7978 }, { "epoch": 2.123203831825439, "grad_norm": 0.30626243352890015, "learning_rate": 1.2704181305075516e-07, "loss": 0.2033, "step": 7979 }, { "epoch": 2.123469930814263, "grad_norm": 0.28805428743362427, "learning_rate": 1.2702538740336322e-07, "loss": 0.2045, "step": 7980 }, { "epoch": 2.1237360298030867, "grad_norm": 0.28988364338874817, "learning_rate": 1.270089609693324e-07, "loss": 0.1873, "step": 7981 }, { "epoch": 2.1240021287919104, "grad_norm": 0.2664649188518524, "learning_rate": 1.2699253374914088e-07, "loss": 0.1945, "step": 7982 }, { "epoch": 2.1242682277807345, "grad_norm": 0.30772775411605835, "learning_rate": 1.2697610574326677e-07, "loss": 0.199, "step": 7983 }, { "epoch": 2.1245343267695582, "grad_norm": 0.29507115483283997, "learning_rate": 1.269596769521883e-07, "loss": 0.2061, "step": 7984 }, { "epoch": 2.124800425758382, "grad_norm": 0.28441494703292847, "learning_rate": 1.269432473763836e-07, "loss": 0.1941, "step": 7985 }, { "epoch": 2.125066524747206, "grad_norm": 0.2922639548778534, "learning_rate": 1.2692681701633093e-07, "loss": 0.1929, "step": 7986 }, { "epoch": 2.1253326237360297, "grad_norm": 0.29442620277404785, "learning_rate": 1.2691038587250856e-07, "loss": 0.213, "step": 7987 }, { "epoch": 2.125598722724854, "grad_norm": 0.2908995747566223, "learning_rate": 1.2689395394539471e-07, "loss": 0.1968, "step": 7988 }, { "epoch": 2.1258648217136775, "grad_norm": 0.27199748158454895, "learning_rate": 1.2687752123546768e-07, "loss": 0.1887, "step": 7989 }, { "epoch": 2.126130920702501, "grad_norm": 0.2855135202407837, "learning_rate": 1.2686108774320578e-07, "loss": 0.1934, "step": 7990 }, { "epoch": 2.1263970196913253, "grad_norm": 0.3021095097064972, "learning_rate": 1.2684465346908738e-07, "loss": 0.2107, "step": 7991 }, { "epoch": 2.126663118680149, "grad_norm": 0.28649112582206726, "learning_rate": 1.2682821841359083e-07, "loss": 0.1971, "step": 7992 }, { "epoch": 2.1269292176689727, "grad_norm": 0.3402320444583893, "learning_rate": 1.2681178257719452e-07, "loss": 0.2012, "step": 7993 }, { "epoch": 2.127195316657797, "grad_norm": 0.32667362689971924, "learning_rate": 1.2679534596037683e-07, "loss": 0.2008, "step": 7994 }, { "epoch": 2.1274614156466205, "grad_norm": 0.28463053703308105, "learning_rate": 1.2677890856361613e-07, "loss": 0.1987, "step": 7995 }, { "epoch": 2.127727514635444, "grad_norm": 0.28606876730918884, "learning_rate": 1.2676247038739099e-07, "loss": 0.1925, "step": 7996 }, { "epoch": 2.1279936136242683, "grad_norm": 0.30814579129219055, "learning_rate": 1.2674603143217982e-07, "loss": 0.213, "step": 7997 }, { "epoch": 2.128259712613092, "grad_norm": 0.28365346789360046, "learning_rate": 1.267295916984611e-07, "loss": 0.1959, "step": 7998 }, { "epoch": 2.128525811601916, "grad_norm": 0.2923797369003296, "learning_rate": 1.2671315118671335e-07, "loss": 0.2006, "step": 7999 }, { "epoch": 2.12879191059074, "grad_norm": 0.3046281039714813, "learning_rate": 1.2669670989741517e-07, "loss": 0.2003, "step": 8000 }, { "epoch": 2.1290580095795635, "grad_norm": 0.3081190288066864, "learning_rate": 1.2668026783104505e-07, "loss": 0.2055, "step": 8001 }, { "epoch": 2.1293241085683876, "grad_norm": 0.2958044409751892, "learning_rate": 1.2666382498808164e-07, "loss": 0.2226, "step": 8002 }, { "epoch": 2.1295902075572113, "grad_norm": 0.46574267745018005, "learning_rate": 1.2664738136900349e-07, "loss": 0.2144, "step": 8003 }, { "epoch": 2.129856306546035, "grad_norm": 0.4096170961856842, "learning_rate": 1.266309369742893e-07, "loss": 0.1762, "step": 8004 }, { "epoch": 2.130122405534859, "grad_norm": 0.29173338413238525, "learning_rate": 1.2661449180441764e-07, "loss": 0.2016, "step": 8005 }, { "epoch": 2.130388504523683, "grad_norm": 0.28111353516578674, "learning_rate": 1.2659804585986725e-07, "loss": 0.2015, "step": 8006 }, { "epoch": 2.1306546035125065, "grad_norm": 0.29463642835617065, "learning_rate": 1.2658159914111676e-07, "loss": 0.2052, "step": 8007 }, { "epoch": 2.1309207025013306, "grad_norm": 0.30318015813827515, "learning_rate": 1.2656515164864498e-07, "loss": 0.1999, "step": 8008 }, { "epoch": 2.1311868014901543, "grad_norm": 0.28176429867744446, "learning_rate": 1.265487033829306e-07, "loss": 0.1917, "step": 8009 }, { "epoch": 2.131452900478978, "grad_norm": 0.268218457698822, "learning_rate": 1.265322543444524e-07, "loss": 0.1875, "step": 8010 }, { "epoch": 2.131718999467802, "grad_norm": 0.3109705448150635, "learning_rate": 1.2651580453368917e-07, "loss": 0.229, "step": 8011 }, { "epoch": 2.1319850984566258, "grad_norm": 0.2885734438896179, "learning_rate": 1.264993539511197e-07, "loss": 0.2067, "step": 8012 }, { "epoch": 2.13225119744545, "grad_norm": 0.27401232719421387, "learning_rate": 1.2648290259722286e-07, "loss": 0.1782, "step": 8013 }, { "epoch": 2.1325172964342736, "grad_norm": 0.2888568937778473, "learning_rate": 1.264664504724775e-07, "loss": 0.1884, "step": 8014 }, { "epoch": 2.1327833954230973, "grad_norm": 0.3116743564605713, "learning_rate": 1.2644999757736247e-07, "loss": 0.2283, "step": 8015 }, { "epoch": 2.1330494944119214, "grad_norm": 0.2547343671321869, "learning_rate": 1.2643354391235667e-07, "loss": 0.1727, "step": 8016 }, { "epoch": 2.133315593400745, "grad_norm": 0.29864856600761414, "learning_rate": 1.2641708947793907e-07, "loss": 0.2058, "step": 8017 }, { "epoch": 2.1335816923895687, "grad_norm": 0.32022830843925476, "learning_rate": 1.2640063427458856e-07, "loss": 0.205, "step": 8018 }, { "epoch": 2.133847791378393, "grad_norm": 0.3065949082374573, "learning_rate": 1.2638417830278417e-07, "loss": 0.196, "step": 8019 }, { "epoch": 2.1341138903672165, "grad_norm": 0.32659009099006653, "learning_rate": 1.2636772156300483e-07, "loss": 0.2242, "step": 8020 }, { "epoch": 2.1343799893560407, "grad_norm": 0.28175055980682373, "learning_rate": 1.2635126405572958e-07, "loss": 0.1757, "step": 8021 }, { "epoch": 2.1346460883448644, "grad_norm": 0.26327747106552124, "learning_rate": 1.2633480578143745e-07, "loss": 0.1693, "step": 8022 }, { "epoch": 2.134912187333688, "grad_norm": 0.2881781756877899, "learning_rate": 1.2631834674060748e-07, "loss": 0.1978, "step": 8023 }, { "epoch": 2.135178286322512, "grad_norm": 0.2802857756614685, "learning_rate": 1.2630188693371883e-07, "loss": 0.1737, "step": 8024 }, { "epoch": 2.135444385311336, "grad_norm": 0.2826301157474518, "learning_rate": 1.2628542636125053e-07, "loss": 0.2062, "step": 8025 }, { "epoch": 2.1357104843001595, "grad_norm": 0.3156769573688507, "learning_rate": 1.2626896502368167e-07, "loss": 0.2052, "step": 8026 }, { "epoch": 2.1359765832889837, "grad_norm": 0.31413593888282776, "learning_rate": 1.262525029214915e-07, "loss": 0.2115, "step": 8027 }, { "epoch": 2.1362426822778073, "grad_norm": 0.288545161485672, "learning_rate": 1.262360400551591e-07, "loss": 0.2032, "step": 8028 }, { "epoch": 2.136508781266631, "grad_norm": 0.2785353660583496, "learning_rate": 1.2621957642516369e-07, "loss": 0.2038, "step": 8029 }, { "epoch": 2.136774880255455, "grad_norm": 0.3036711513996124, "learning_rate": 1.262031120319845e-07, "loss": 0.2007, "step": 8030 }, { "epoch": 2.137040979244279, "grad_norm": 0.2800135314464569, "learning_rate": 1.2618664687610077e-07, "loss": 0.185, "step": 8031 }, { "epoch": 2.1373070782331025, "grad_norm": 0.30494168400764465, "learning_rate": 1.261701809579917e-07, "loss": 0.2093, "step": 8032 }, { "epoch": 2.1375731772219266, "grad_norm": 0.29652130603790283, "learning_rate": 1.261537142781367e-07, "loss": 0.1971, "step": 8033 }, { "epoch": 2.1378392762107503, "grad_norm": 0.31277695298194885, "learning_rate": 1.261372468370149e-07, "loss": 0.2214, "step": 8034 }, { "epoch": 2.1381053751995744, "grad_norm": 0.3123784363269806, "learning_rate": 1.2612077863510574e-07, "loss": 0.2057, "step": 8035 }, { "epoch": 2.138371474188398, "grad_norm": 0.29133346676826477, "learning_rate": 1.2610430967288852e-07, "loss": 0.1917, "step": 8036 }, { "epoch": 2.138637573177222, "grad_norm": 0.30446797609329224, "learning_rate": 1.2608783995084267e-07, "loss": 0.2018, "step": 8037 }, { "epoch": 2.138903672166046, "grad_norm": 0.303252637386322, "learning_rate": 1.2607136946944748e-07, "loss": 0.2144, "step": 8038 }, { "epoch": 2.1391697711548696, "grad_norm": 0.29595866799354553, "learning_rate": 1.2605489822918242e-07, "loss": 0.195, "step": 8039 }, { "epoch": 2.1394358701436933, "grad_norm": 0.29746338725090027, "learning_rate": 1.2603842623052697e-07, "loss": 0.186, "step": 8040 }, { "epoch": 2.1397019691325174, "grad_norm": 0.3142116665840149, "learning_rate": 1.260219534739605e-07, "loss": 0.2131, "step": 8041 }, { "epoch": 2.139968068121341, "grad_norm": 0.3098134696483612, "learning_rate": 1.2600547995996257e-07, "loss": 0.2136, "step": 8042 }, { "epoch": 2.140234167110165, "grad_norm": 0.30696165561676025, "learning_rate": 1.259890056890126e-07, "loss": 0.2268, "step": 8043 }, { "epoch": 2.140500266098989, "grad_norm": 0.31039267778396606, "learning_rate": 1.2597253066159015e-07, "loss": 0.2149, "step": 8044 }, { "epoch": 2.1407663650878126, "grad_norm": 0.2758871614933014, "learning_rate": 1.259560548781748e-07, "loss": 0.1891, "step": 8045 }, { "epoch": 2.1410324640766367, "grad_norm": 0.2832775115966797, "learning_rate": 1.259395783392461e-07, "loss": 0.1857, "step": 8046 }, { "epoch": 2.1412985630654604, "grad_norm": 0.28078389167785645, "learning_rate": 1.2592310104528358e-07, "loss": 0.1855, "step": 8047 }, { "epoch": 2.141564662054284, "grad_norm": 0.3258446156978607, "learning_rate": 1.259066229967669e-07, "loss": 0.24, "step": 8048 }, { "epoch": 2.141830761043108, "grad_norm": 0.2816947400569916, "learning_rate": 1.2589014419417573e-07, "loss": 0.1924, "step": 8049 }, { "epoch": 2.142096860031932, "grad_norm": 0.2877315282821655, "learning_rate": 1.2587366463798964e-07, "loss": 0.1909, "step": 8050 }, { "epoch": 2.1423629590207556, "grad_norm": 0.2673547565937042, "learning_rate": 1.2585718432868843e-07, "loss": 0.1788, "step": 8051 }, { "epoch": 2.1426290580095797, "grad_norm": 0.2781239449977875, "learning_rate": 1.2584070326675162e-07, "loss": 0.2121, "step": 8052 }, { "epoch": 2.1428951569984034, "grad_norm": 0.31926214694976807, "learning_rate": 1.258242214526591e-07, "loss": 0.2272, "step": 8053 }, { "epoch": 2.143161255987227, "grad_norm": 0.3002726435661316, "learning_rate": 1.2580773888689055e-07, "loss": 0.228, "step": 8054 }, { "epoch": 2.143427354976051, "grad_norm": 0.2862679660320282, "learning_rate": 1.2579125556992568e-07, "loss": 0.1898, "step": 8055 }, { "epoch": 2.143693453964875, "grad_norm": 0.28225281834602356, "learning_rate": 1.2577477150224436e-07, "loss": 0.2022, "step": 8056 }, { "epoch": 2.1439595529536986, "grad_norm": 0.29069894552230835, "learning_rate": 1.2575828668432635e-07, "loss": 0.2201, "step": 8057 }, { "epoch": 2.1442256519425227, "grad_norm": 0.2812928557395935, "learning_rate": 1.257418011166515e-07, "loss": 0.2048, "step": 8058 }, { "epoch": 2.1444917509313464, "grad_norm": 0.2676779329776764, "learning_rate": 1.2572531479969968e-07, "loss": 0.1653, "step": 8059 }, { "epoch": 2.1447578499201705, "grad_norm": 0.28746330738067627, "learning_rate": 1.2570882773395073e-07, "loss": 0.2008, "step": 8060 }, { "epoch": 2.145023948908994, "grad_norm": 0.2941049635410309, "learning_rate": 1.2569233991988455e-07, "loss": 0.2171, "step": 8061 }, { "epoch": 2.145290047897818, "grad_norm": 0.30642542243003845, "learning_rate": 1.2567585135798104e-07, "loss": 0.2148, "step": 8062 }, { "epoch": 2.145556146886642, "grad_norm": 0.28381508588790894, "learning_rate": 1.256593620487202e-07, "loss": 0.1957, "step": 8063 }, { "epoch": 2.1458222458754657, "grad_norm": 0.2677234709262848, "learning_rate": 1.2564287199258194e-07, "loss": 0.1713, "step": 8064 }, { "epoch": 2.1460883448642893, "grad_norm": 0.2936767041683197, "learning_rate": 1.2562638119004624e-07, "loss": 0.1892, "step": 8065 }, { "epoch": 2.1463544438531135, "grad_norm": 0.31432467699050903, "learning_rate": 1.2560988964159317e-07, "loss": 0.2083, "step": 8066 }, { "epoch": 2.146620542841937, "grad_norm": 0.2975905239582062, "learning_rate": 1.2559339734770267e-07, "loss": 0.1999, "step": 8067 }, { "epoch": 2.1468866418307613, "grad_norm": 0.27319592237472534, "learning_rate": 1.2557690430885482e-07, "loss": 0.1886, "step": 8068 }, { "epoch": 2.147152740819585, "grad_norm": 0.30966830253601074, "learning_rate": 1.2556041052552976e-07, "loss": 0.1936, "step": 8069 }, { "epoch": 2.1474188398084086, "grad_norm": 0.299010694026947, "learning_rate": 1.2554391599820743e-07, "loss": 0.2074, "step": 8070 }, { "epoch": 2.1476849387972328, "grad_norm": 0.2933781147003174, "learning_rate": 1.2552742072736807e-07, "loss": 0.2089, "step": 8071 }, { "epoch": 2.1479510377860564, "grad_norm": 0.28110945224761963, "learning_rate": 1.2551092471349175e-07, "loss": 0.2048, "step": 8072 }, { "epoch": 2.14821713677488, "grad_norm": 0.2985401153564453, "learning_rate": 1.254944279570587e-07, "loss": 0.2121, "step": 8073 }, { "epoch": 2.1484832357637043, "grad_norm": 0.28837233781814575, "learning_rate": 1.25477930458549e-07, "loss": 0.1811, "step": 8074 }, { "epoch": 2.148749334752528, "grad_norm": 0.28224554657936096, "learning_rate": 1.2546143221844293e-07, "loss": 0.1942, "step": 8075 }, { "epoch": 2.1490154337413516, "grad_norm": 0.277322381734848, "learning_rate": 1.2544493323722067e-07, "loss": 0.1902, "step": 8076 }, { "epoch": 2.1492815327301757, "grad_norm": 0.3131134510040283, "learning_rate": 1.2542843351536247e-07, "loss": 0.2065, "step": 8077 }, { "epoch": 2.1495476317189994, "grad_norm": 0.29708388447761536, "learning_rate": 1.2541193305334858e-07, "loss": 0.2028, "step": 8078 }, { "epoch": 2.149813730707823, "grad_norm": 0.275677889585495, "learning_rate": 1.253954318516593e-07, "loss": 0.2039, "step": 8079 }, { "epoch": 2.1500798296966472, "grad_norm": 0.27366575598716736, "learning_rate": 1.2537892991077494e-07, "loss": 0.1873, "step": 8080 }, { "epoch": 2.150345928685471, "grad_norm": 0.29484322667121887, "learning_rate": 1.2536242723117583e-07, "loss": 0.2044, "step": 8081 }, { "epoch": 2.1506120276742946, "grad_norm": 0.24953730404376984, "learning_rate": 1.2534592381334235e-07, "loss": 0.1809, "step": 8082 }, { "epoch": 2.1508781266631187, "grad_norm": 0.2760589122772217, "learning_rate": 1.253294196577548e-07, "loss": 0.2008, "step": 8083 }, { "epoch": 2.1511442256519424, "grad_norm": 0.2865604758262634, "learning_rate": 1.253129147648936e-07, "loss": 0.2093, "step": 8084 }, { "epoch": 2.1514103246407665, "grad_norm": 0.2796821892261505, "learning_rate": 1.2529640913523923e-07, "loss": 0.1968, "step": 8085 }, { "epoch": 2.15167642362959, "grad_norm": 0.324941486120224, "learning_rate": 1.25279902769272e-07, "loss": 0.2147, "step": 8086 }, { "epoch": 2.151942522618414, "grad_norm": 0.28283849358558655, "learning_rate": 1.252633956674725e-07, "loss": 0.2029, "step": 8087 }, { "epoch": 2.152208621607238, "grad_norm": 0.29942673444747925, "learning_rate": 1.2524688783032112e-07, "loss": 0.2018, "step": 8088 }, { "epoch": 2.1524747205960617, "grad_norm": 0.29223158955574036, "learning_rate": 1.2523037925829837e-07, "loss": 0.1891, "step": 8089 }, { "epoch": 2.1527408195848854, "grad_norm": 0.31817367672920227, "learning_rate": 1.252138699518848e-07, "loss": 0.1939, "step": 8090 }, { "epoch": 2.1530069185737095, "grad_norm": 0.29536697268486023, "learning_rate": 1.2519735991156096e-07, "loss": 0.1949, "step": 8091 }, { "epoch": 2.153273017562533, "grad_norm": 0.28439998626708984, "learning_rate": 1.2518084913780739e-07, "loss": 0.2059, "step": 8092 }, { "epoch": 2.1535391165513573, "grad_norm": 0.2906438410282135, "learning_rate": 1.251643376311047e-07, "loss": 0.1863, "step": 8093 }, { "epoch": 2.153805215540181, "grad_norm": 0.289908766746521, "learning_rate": 1.2514782539193345e-07, "loss": 0.1772, "step": 8094 }, { "epoch": 2.1540713145290047, "grad_norm": 0.28286561369895935, "learning_rate": 1.2513131242077431e-07, "loss": 0.1866, "step": 8095 }, { "epoch": 2.154337413517829, "grad_norm": 0.2842279374599457, "learning_rate": 1.2511479871810792e-07, "loss": 0.1931, "step": 8096 }, { "epoch": 2.1546035125066525, "grad_norm": 0.28655654191970825, "learning_rate": 1.2509828428441495e-07, "loss": 0.2067, "step": 8097 }, { "epoch": 2.154869611495476, "grad_norm": 0.286273330450058, "learning_rate": 1.250817691201761e-07, "loss": 0.2168, "step": 8098 }, { "epoch": 2.1551357104843003, "grad_norm": 0.28212106227874756, "learning_rate": 1.2506525322587204e-07, "loss": 0.2118, "step": 8099 }, { "epoch": 2.155401809473124, "grad_norm": 0.273650199174881, "learning_rate": 1.2504873660198358e-07, "loss": 0.1788, "step": 8100 }, { "epoch": 2.1556679084619477, "grad_norm": 0.2844366431236267, "learning_rate": 1.2503221924899143e-07, "loss": 0.1883, "step": 8101 }, { "epoch": 2.155934007450772, "grad_norm": 0.2841322124004364, "learning_rate": 1.2501570116737634e-07, "loss": 0.1941, "step": 8102 }, { "epoch": 2.1562001064395955, "grad_norm": 0.3065018057823181, "learning_rate": 1.2499918235761918e-07, "loss": 0.1868, "step": 8103 }, { "epoch": 2.156466205428419, "grad_norm": 0.3149533271789551, "learning_rate": 1.2498266282020068e-07, "loss": 0.2202, "step": 8104 }, { "epoch": 2.1567323044172433, "grad_norm": 0.3014468848705292, "learning_rate": 1.249661425556018e-07, "loss": 0.204, "step": 8105 }, { "epoch": 2.156998403406067, "grad_norm": 0.2787107229232788, "learning_rate": 1.2494962156430332e-07, "loss": 0.209, "step": 8106 }, { "epoch": 2.157264502394891, "grad_norm": 0.30257171392440796, "learning_rate": 1.2493309984678613e-07, "loss": 0.1954, "step": 8107 }, { "epoch": 2.1575306013837148, "grad_norm": 0.28164607286453247, "learning_rate": 1.249165774035311e-07, "loss": 0.2026, "step": 8108 }, { "epoch": 2.1577967003725385, "grad_norm": 0.2987675070762634, "learning_rate": 1.2490005423501927e-07, "loss": 0.2136, "step": 8109 }, { "epoch": 2.1580627993613626, "grad_norm": 0.2916068732738495, "learning_rate": 1.2488353034173146e-07, "loss": 0.2054, "step": 8110 }, { "epoch": 2.1583288983501863, "grad_norm": 0.28196561336517334, "learning_rate": 1.2486700572414868e-07, "loss": 0.2097, "step": 8111 }, { "epoch": 2.15859499733901, "grad_norm": 0.319845974445343, "learning_rate": 1.2485048038275196e-07, "loss": 0.2192, "step": 8112 }, { "epoch": 2.158861096327834, "grad_norm": 0.2980666756629944, "learning_rate": 1.2483395431802226e-07, "loss": 0.2047, "step": 8113 }, { "epoch": 2.1591271953166578, "grad_norm": 0.30785414576530457, "learning_rate": 1.2481742753044066e-07, "loss": 0.2041, "step": 8114 }, { "epoch": 2.1593932943054814, "grad_norm": 0.2767099142074585, "learning_rate": 1.2480090002048815e-07, "loss": 0.1864, "step": 8115 }, { "epoch": 2.1596593932943056, "grad_norm": 0.32403403520584106, "learning_rate": 1.2478437178864586e-07, "loss": 0.2021, "step": 8116 }, { "epoch": 2.1599254922831292, "grad_norm": 0.3014116585254669, "learning_rate": 1.2476784283539485e-07, "loss": 0.2075, "step": 8117 }, { "epoch": 2.1601915912719534, "grad_norm": 0.2882249653339386, "learning_rate": 1.2475131316121626e-07, "loss": 0.2058, "step": 8118 }, { "epoch": 2.160457690260777, "grad_norm": 0.3089424967765808, "learning_rate": 1.247347827665912e-07, "loss": 0.2021, "step": 8119 }, { "epoch": 2.1607237892496007, "grad_norm": 0.2729775607585907, "learning_rate": 1.2471825165200083e-07, "loss": 0.1862, "step": 8120 }, { "epoch": 2.160989888238425, "grad_norm": 0.29815319180488586, "learning_rate": 1.2470171981792631e-07, "loss": 0.2163, "step": 8121 }, { "epoch": 2.1612559872272485, "grad_norm": 0.28269898891448975, "learning_rate": 1.2468518726484888e-07, "loss": 0.196, "step": 8122 }, { "epoch": 2.1615220862160722, "grad_norm": 0.3066907227039337, "learning_rate": 1.246686539932498e-07, "loss": 0.2159, "step": 8123 }, { "epoch": 2.1617881852048964, "grad_norm": 0.2796708345413208, "learning_rate": 1.246521200036102e-07, "loss": 0.2094, "step": 8124 }, { "epoch": 2.16205428419372, "grad_norm": 0.30287835001945496, "learning_rate": 1.2463558529641138e-07, "loss": 0.191, "step": 8125 }, { "epoch": 2.1623203831825437, "grad_norm": 0.3798708915710449, "learning_rate": 1.2461904987213467e-07, "loss": 0.2172, "step": 8126 }, { "epoch": 2.162586482171368, "grad_norm": 0.3118168115615845, "learning_rate": 1.2460251373126133e-07, "loss": 0.2079, "step": 8127 }, { "epoch": 2.1628525811601915, "grad_norm": 0.3073289692401886, "learning_rate": 1.245859768742727e-07, "loss": 0.2246, "step": 8128 }, { "epoch": 2.163118680149015, "grad_norm": 0.28745684027671814, "learning_rate": 1.2456943930165012e-07, "loss": 0.1766, "step": 8129 }, { "epoch": 2.1633847791378393, "grad_norm": 0.2998977303504944, "learning_rate": 1.2455290101387496e-07, "loss": 0.2, "step": 8130 }, { "epoch": 2.163650878126663, "grad_norm": 0.29207825660705566, "learning_rate": 1.2453636201142858e-07, "loss": 0.1894, "step": 8131 }, { "epoch": 2.163916977115487, "grad_norm": 0.30633851885795593, "learning_rate": 1.2451982229479242e-07, "loss": 0.2049, "step": 8132 }, { "epoch": 2.164183076104311, "grad_norm": 0.2935566306114197, "learning_rate": 1.2450328186444792e-07, "loss": 0.1994, "step": 8133 }, { "epoch": 2.1644491750931345, "grad_norm": 0.28030604124069214, "learning_rate": 1.2448674072087645e-07, "loss": 0.1926, "step": 8134 }, { "epoch": 2.1647152740819586, "grad_norm": 0.3138507902622223, "learning_rate": 1.2447019886455957e-07, "loss": 0.2055, "step": 8135 }, { "epoch": 2.1649813730707823, "grad_norm": 0.31858858466148376, "learning_rate": 1.2445365629597878e-07, "loss": 0.2036, "step": 8136 }, { "epoch": 2.165247472059606, "grad_norm": 0.30141621828079224, "learning_rate": 1.2443711301561553e-07, "loss": 0.2145, "step": 8137 }, { "epoch": 2.16551357104843, "grad_norm": 0.27684637904167175, "learning_rate": 1.2442056902395134e-07, "loss": 0.1897, "step": 8138 }, { "epoch": 2.165779670037254, "grad_norm": 0.2904568314552307, "learning_rate": 1.2440402432146778e-07, "loss": 0.1866, "step": 8139 }, { "epoch": 2.166045769026078, "grad_norm": 0.2974269390106201, "learning_rate": 1.2438747890864647e-07, "loss": 0.1963, "step": 8140 }, { "epoch": 2.1663118680149016, "grad_norm": 0.2940210700035095, "learning_rate": 1.2437093278596898e-07, "loss": 0.192, "step": 8141 }, { "epoch": 2.1665779670037253, "grad_norm": 0.3004995882511139, "learning_rate": 1.243543859539169e-07, "loss": 0.2058, "step": 8142 }, { "epoch": 2.1668440659925494, "grad_norm": 0.27289336919784546, "learning_rate": 1.2433783841297185e-07, "loss": 0.174, "step": 8143 }, { "epoch": 2.167110164981373, "grad_norm": 0.27052661776542664, "learning_rate": 1.2432129016361554e-07, "loss": 0.1991, "step": 8144 }, { "epoch": 2.167376263970197, "grad_norm": 0.309156209230423, "learning_rate": 1.2430474120632964e-07, "loss": 0.1957, "step": 8145 }, { "epoch": 2.167642362959021, "grad_norm": 0.27830415964126587, "learning_rate": 1.2428819154159582e-07, "loss": 0.2093, "step": 8146 }, { "epoch": 2.1679084619478446, "grad_norm": 0.2655031979084015, "learning_rate": 1.2427164116989581e-07, "loss": 0.1865, "step": 8147 }, { "epoch": 2.1681745609366683, "grad_norm": 0.30695781111717224, "learning_rate": 1.2425509009171136e-07, "loss": 0.2064, "step": 8148 }, { "epoch": 2.1684406599254924, "grad_norm": 0.2880595624446869, "learning_rate": 1.2423853830752418e-07, "loss": 0.2005, "step": 8149 }, { "epoch": 2.168706758914316, "grad_norm": 0.2822563946247101, "learning_rate": 1.2422198581781612e-07, "loss": 0.202, "step": 8150 }, { "epoch": 2.1689728579031398, "grad_norm": 0.3096115291118622, "learning_rate": 1.242054326230689e-07, "loss": 0.1969, "step": 8151 }, { "epoch": 2.169238956891964, "grad_norm": 0.2973141372203827, "learning_rate": 1.241888787237644e-07, "loss": 0.2239, "step": 8152 }, { "epoch": 2.1695050558807876, "grad_norm": 0.2819913923740387, "learning_rate": 1.2417232412038447e-07, "loss": 0.1797, "step": 8153 }, { "epoch": 2.1697711548696117, "grad_norm": 0.3010112941265106, "learning_rate": 1.2415576881341092e-07, "loss": 0.1983, "step": 8154 }, { "epoch": 2.1700372538584354, "grad_norm": 0.29785817861557007, "learning_rate": 1.2413921280332568e-07, "loss": 0.2073, "step": 8155 }, { "epoch": 2.170303352847259, "grad_norm": 0.29129883646965027, "learning_rate": 1.2412265609061064e-07, "loss": 0.2018, "step": 8156 }, { "epoch": 2.170569451836083, "grad_norm": 0.27976539731025696, "learning_rate": 1.2410609867574772e-07, "loss": 0.199, "step": 8157 }, { "epoch": 2.170835550824907, "grad_norm": 0.2737976014614105, "learning_rate": 1.2408954055921882e-07, "loss": 0.1898, "step": 8158 }, { "epoch": 2.1711016498137305, "grad_norm": 0.29449358582496643, "learning_rate": 1.2407298174150597e-07, "loss": 0.1907, "step": 8159 }, { "epoch": 2.1713677488025547, "grad_norm": 0.3146668076515198, "learning_rate": 1.2405642222309112e-07, "loss": 0.1915, "step": 8160 }, { "epoch": 2.1716338477913784, "grad_norm": 0.28945082426071167, "learning_rate": 1.2403986200445626e-07, "loss": 0.2068, "step": 8161 }, { "epoch": 2.171899946780202, "grad_norm": 0.2830842435359955, "learning_rate": 1.2402330108608346e-07, "loss": 0.1779, "step": 8162 }, { "epoch": 2.172166045769026, "grad_norm": 0.29528486728668213, "learning_rate": 1.2400673946845473e-07, "loss": 0.1985, "step": 8163 }, { "epoch": 2.17243214475785, "grad_norm": 0.3074290454387665, "learning_rate": 1.2399017715205218e-07, "loss": 0.2094, "step": 8164 }, { "epoch": 2.172698243746674, "grad_norm": 0.2815669775009155, "learning_rate": 1.2397361413735783e-07, "loss": 0.2031, "step": 8165 }, { "epoch": 2.1729643427354977, "grad_norm": 0.2957061231136322, "learning_rate": 1.239570504248538e-07, "loss": 0.196, "step": 8166 }, { "epoch": 2.1732304417243213, "grad_norm": 0.26130175590515137, "learning_rate": 1.239404860150223e-07, "loss": 0.187, "step": 8167 }, { "epoch": 2.1734965407131455, "grad_norm": 0.29797568917274475, "learning_rate": 1.239239209083454e-07, "loss": 0.1855, "step": 8168 }, { "epoch": 2.173762639701969, "grad_norm": 0.27362921833992004, "learning_rate": 1.2390735510530524e-07, "loss": 0.1758, "step": 8169 }, { "epoch": 2.174028738690793, "grad_norm": 0.3017716705799103, "learning_rate": 1.2389078860638404e-07, "loss": 0.2071, "step": 8170 }, { "epoch": 2.174294837679617, "grad_norm": 0.27513641119003296, "learning_rate": 1.2387422141206403e-07, "loss": 0.1965, "step": 8171 }, { "epoch": 2.1745609366684406, "grad_norm": 0.29003968834877014, "learning_rate": 1.2385765352282742e-07, "loss": 0.19, "step": 8172 }, { "epoch": 2.1748270356572643, "grad_norm": 0.2867961525917053, "learning_rate": 1.238410849391565e-07, "loss": 0.182, "step": 8173 }, { "epoch": 2.1750931346460884, "grad_norm": 0.2754557430744171, "learning_rate": 1.2382451566153345e-07, "loss": 0.1851, "step": 8174 }, { "epoch": 2.175359233634912, "grad_norm": 0.3095216453075409, "learning_rate": 1.2380794569044065e-07, "loss": 0.2116, "step": 8175 }, { "epoch": 2.175625332623736, "grad_norm": 0.3002975881099701, "learning_rate": 1.2379137502636034e-07, "loss": 0.2186, "step": 8176 }, { "epoch": 2.17589143161256, "grad_norm": 0.27188411355018616, "learning_rate": 1.2377480366977487e-07, "loss": 0.1853, "step": 8177 }, { "epoch": 2.1761575306013836, "grad_norm": 0.31293779611587524, "learning_rate": 1.2375823162116664e-07, "loss": 0.2034, "step": 8178 }, { "epoch": 2.1764236295902077, "grad_norm": 0.2853552997112274, "learning_rate": 1.2374165888101795e-07, "loss": 0.2027, "step": 8179 }, { "epoch": 2.1766897285790314, "grad_norm": 0.29788780212402344, "learning_rate": 1.237250854498112e-07, "loss": 0.2005, "step": 8180 }, { "epoch": 2.176955827567855, "grad_norm": 0.28522083163261414, "learning_rate": 1.2370851132802882e-07, "loss": 0.1929, "step": 8181 }, { "epoch": 2.1772219265566792, "grad_norm": 0.2827768921852112, "learning_rate": 1.2369193651615326e-07, "loss": 0.1935, "step": 8182 }, { "epoch": 2.177488025545503, "grad_norm": 0.27455225586891174, "learning_rate": 1.2367536101466692e-07, "loss": 0.1866, "step": 8183 }, { "epoch": 2.1777541245343266, "grad_norm": 0.300339937210083, "learning_rate": 1.2365878482405227e-07, "loss": 0.2, "step": 8184 }, { "epoch": 2.1780202235231507, "grad_norm": 0.2919304072856903, "learning_rate": 1.2364220794479185e-07, "loss": 0.2074, "step": 8185 }, { "epoch": 2.1782863225119744, "grad_norm": 0.27092593908309937, "learning_rate": 1.2362563037736817e-07, "loss": 0.1884, "step": 8186 }, { "epoch": 2.1785524215007985, "grad_norm": 0.2841029465198517, "learning_rate": 1.2360905212226368e-07, "loss": 0.188, "step": 8187 }, { "epoch": 2.178818520489622, "grad_norm": 0.2872014045715332, "learning_rate": 1.2359247317996105e-07, "loss": 0.2026, "step": 8188 }, { "epoch": 2.179084619478446, "grad_norm": 0.27920177578926086, "learning_rate": 1.2357589355094274e-07, "loss": 0.1852, "step": 8189 }, { "epoch": 2.17935071846727, "grad_norm": 0.27535346150398254, "learning_rate": 1.2355931323569136e-07, "loss": 0.1901, "step": 8190 }, { "epoch": 2.1796168174560937, "grad_norm": 0.268191933631897, "learning_rate": 1.2354273223468958e-07, "loss": 0.198, "step": 8191 }, { "epoch": 2.1798829164449174, "grad_norm": 0.3815017640590668, "learning_rate": 1.2352615054841998e-07, "loss": 0.2007, "step": 8192 }, { "epoch": 2.1801490154337415, "grad_norm": 0.31535592675209045, "learning_rate": 1.2350956817736523e-07, "loss": 0.201, "step": 8193 }, { "epoch": 2.180415114422565, "grad_norm": 0.2968709468841553, "learning_rate": 1.2349298512200795e-07, "loss": 0.2075, "step": 8194 }, { "epoch": 2.180681213411389, "grad_norm": 0.2858129143714905, "learning_rate": 1.2347640138283093e-07, "loss": 0.1912, "step": 8195 }, { "epoch": 2.180947312400213, "grad_norm": 0.28210702538490295, "learning_rate": 1.234598169603168e-07, "loss": 0.1843, "step": 8196 }, { "epoch": 2.1812134113890367, "grad_norm": 0.28705894947052, "learning_rate": 1.234432318549483e-07, "loss": 0.1915, "step": 8197 }, { "epoch": 2.1814795103778604, "grad_norm": 0.2602635324001312, "learning_rate": 1.234266460672082e-07, "loss": 0.1771, "step": 8198 }, { "epoch": 2.1817456093666845, "grad_norm": 0.2886003255844116, "learning_rate": 1.2341005959757928e-07, "loss": 0.1912, "step": 8199 }, { "epoch": 2.182011708355508, "grad_norm": 0.29485654830932617, "learning_rate": 1.233934724465443e-07, "loss": 0.1921, "step": 8200 }, { "epoch": 2.1822778073443323, "grad_norm": 0.28958308696746826, "learning_rate": 1.2337688461458607e-07, "loss": 0.1929, "step": 8201 }, { "epoch": 2.182543906333156, "grad_norm": 0.287277489900589, "learning_rate": 1.233602961021874e-07, "loss": 0.1857, "step": 8202 }, { "epoch": 2.1828100053219797, "grad_norm": 0.28568312525749207, "learning_rate": 1.233437069098312e-07, "loss": 0.2013, "step": 8203 }, { "epoch": 2.183076104310804, "grad_norm": 0.2822065055370331, "learning_rate": 1.2332711703800028e-07, "loss": 0.1937, "step": 8204 }, { "epoch": 2.1833422032996275, "grad_norm": 0.30604881048202515, "learning_rate": 1.2331052648717758e-07, "loss": 0.1943, "step": 8205 }, { "epoch": 2.183608302288451, "grad_norm": 0.26948246359825134, "learning_rate": 1.2329393525784596e-07, "loss": 0.1905, "step": 8206 }, { "epoch": 2.1838744012772753, "grad_norm": 0.2813415825366974, "learning_rate": 1.2327734335048836e-07, "loss": 0.1979, "step": 8207 }, { "epoch": 2.184140500266099, "grad_norm": 0.30214112997055054, "learning_rate": 1.2326075076558774e-07, "loss": 0.2019, "step": 8208 }, { "epoch": 2.1844065992549226, "grad_norm": 0.34471595287323, "learning_rate": 1.2324415750362708e-07, "loss": 0.1955, "step": 8209 }, { "epoch": 2.1846726982437468, "grad_norm": 0.262853741645813, "learning_rate": 1.2322756356508932e-07, "loss": 0.1796, "step": 8210 }, { "epoch": 2.1849387972325705, "grad_norm": 0.2816280722618103, "learning_rate": 1.2321096895045752e-07, "loss": 0.1873, "step": 8211 }, { "epoch": 2.1852048962213946, "grad_norm": 0.31065598130226135, "learning_rate": 1.2319437366021463e-07, "loss": 0.1849, "step": 8212 }, { "epoch": 2.1854709952102183, "grad_norm": 0.26714619994163513, "learning_rate": 1.2317777769484382e-07, "loss": 0.2019, "step": 8213 }, { "epoch": 2.185737094199042, "grad_norm": 0.2854897379875183, "learning_rate": 1.2316118105482801e-07, "loss": 0.1998, "step": 8214 }, { "epoch": 2.186003193187866, "grad_norm": 0.32462412118911743, "learning_rate": 1.231445837406504e-07, "loss": 0.2052, "step": 8215 }, { "epoch": 2.1862692921766897, "grad_norm": 0.2876609265804291, "learning_rate": 1.2312798575279403e-07, "loss": 0.2052, "step": 8216 }, { "epoch": 2.1865353911655134, "grad_norm": 0.27670004963874817, "learning_rate": 1.2311138709174208e-07, "loss": 0.1972, "step": 8217 }, { "epoch": 2.1868014901543376, "grad_norm": 0.26892927289009094, "learning_rate": 1.2309478775797765e-07, "loss": 0.1868, "step": 8218 }, { "epoch": 2.1870675891431612, "grad_norm": 0.2866199314594269, "learning_rate": 1.230781877519839e-07, "loss": 0.1954, "step": 8219 }, { "epoch": 2.187333688131985, "grad_norm": 0.27192986011505127, "learning_rate": 1.23061587074244e-07, "loss": 0.1956, "step": 8220 }, { "epoch": 2.187599787120809, "grad_norm": 0.2972622811794281, "learning_rate": 1.2304498572524123e-07, "loss": 0.2231, "step": 8221 }, { "epoch": 2.1878658861096327, "grad_norm": 0.29522427916526794, "learning_rate": 1.2302838370545874e-07, "loss": 0.2088, "step": 8222 }, { "epoch": 2.1881319850984564, "grad_norm": 0.3093058168888092, "learning_rate": 1.2301178101537978e-07, "loss": 0.1963, "step": 8223 }, { "epoch": 2.1883980840872805, "grad_norm": 0.269378662109375, "learning_rate": 1.2299517765548765e-07, "loss": 0.198, "step": 8224 }, { "epoch": 2.188664183076104, "grad_norm": 0.2878420352935791, "learning_rate": 1.2297857362626558e-07, "loss": 0.1938, "step": 8225 }, { "epoch": 2.1889302820649283, "grad_norm": 0.29364240169525146, "learning_rate": 1.229619689281969e-07, "loss": 0.212, "step": 8226 }, { "epoch": 2.189196381053752, "grad_norm": 0.29744720458984375, "learning_rate": 1.2294536356176494e-07, "loss": 0.2103, "step": 8227 }, { "epoch": 2.1894624800425757, "grad_norm": 0.27907583117485046, "learning_rate": 1.2292875752745302e-07, "loss": 0.1856, "step": 8228 }, { "epoch": 2.1897285790314, "grad_norm": 0.2823983430862427, "learning_rate": 1.2291215082574452e-07, "loss": 0.182, "step": 8229 }, { "epoch": 2.1899946780202235, "grad_norm": 0.2612954378128052, "learning_rate": 1.2289554345712274e-07, "loss": 0.1732, "step": 8230 }, { "epoch": 2.190260777009047, "grad_norm": 0.31495407223701477, "learning_rate": 1.228789354220712e-07, "loss": 0.1998, "step": 8231 }, { "epoch": 2.1905268759978713, "grad_norm": 0.3124603033065796, "learning_rate": 1.2286232672107322e-07, "loss": 0.2136, "step": 8232 }, { "epoch": 2.190792974986695, "grad_norm": 0.290004700422287, "learning_rate": 1.2284571735461227e-07, "loss": 0.1919, "step": 8233 }, { "epoch": 2.1910590739755187, "grad_norm": 0.2895418405532837, "learning_rate": 1.228291073231718e-07, "loss": 0.1867, "step": 8234 }, { "epoch": 2.191325172964343, "grad_norm": 0.30092138051986694, "learning_rate": 1.228124966272353e-07, "loss": 0.2241, "step": 8235 }, { "epoch": 2.1915912719531665, "grad_norm": 0.28630614280700684, "learning_rate": 1.227958852672863e-07, "loss": 0.183, "step": 8236 }, { "epoch": 2.1918573709419906, "grad_norm": 0.3255375325679779, "learning_rate": 1.2277927324380822e-07, "loss": 0.1793, "step": 8237 }, { "epoch": 2.1921234699308143, "grad_norm": 0.3134017586708069, "learning_rate": 1.2276266055728466e-07, "loss": 0.2099, "step": 8238 }, { "epoch": 2.192389568919638, "grad_norm": 0.31203752756118774, "learning_rate": 1.2274604720819915e-07, "loss": 0.2016, "step": 8239 }, { "epoch": 2.192655667908462, "grad_norm": 0.29031428694725037, "learning_rate": 1.2272943319703526e-07, "loss": 0.2141, "step": 8240 }, { "epoch": 2.192921766897286, "grad_norm": 0.2794274091720581, "learning_rate": 1.227128185242766e-07, "loss": 0.1788, "step": 8241 }, { "epoch": 2.1931878658861095, "grad_norm": 0.3031397759914398, "learning_rate": 1.2269620319040673e-07, "loss": 0.1886, "step": 8242 }, { "epoch": 2.1934539648749336, "grad_norm": 0.28877967596054077, "learning_rate": 1.2267958719590935e-07, "loss": 0.1901, "step": 8243 }, { "epoch": 2.1937200638637573, "grad_norm": 0.2953895926475525, "learning_rate": 1.2266297054126804e-07, "loss": 0.2116, "step": 8244 }, { "epoch": 2.193986162852581, "grad_norm": 0.3053476810455322, "learning_rate": 1.2264635322696654e-07, "loss": 0.2061, "step": 8245 }, { "epoch": 2.194252261841405, "grad_norm": 0.3022424280643463, "learning_rate": 1.2262973525348848e-07, "loss": 0.2211, "step": 8246 }, { "epoch": 2.1945183608302288, "grad_norm": 0.5834170579910278, "learning_rate": 1.2261311662131758e-07, "loss": 0.2068, "step": 8247 }, { "epoch": 2.1947844598190525, "grad_norm": 0.2836010456085205, "learning_rate": 1.2259649733093757e-07, "loss": 0.2054, "step": 8248 }, { "epoch": 2.1950505588078766, "grad_norm": 0.2909964621067047, "learning_rate": 1.225798773828322e-07, "loss": 0.1782, "step": 8249 }, { "epoch": 2.1953166577967003, "grad_norm": 0.2857836186885834, "learning_rate": 1.2256325677748525e-07, "loss": 0.2017, "step": 8250 }, { "epoch": 2.1955827567855244, "grad_norm": 0.2910758852958679, "learning_rate": 1.2254663551538044e-07, "loss": 0.1876, "step": 8251 }, { "epoch": 2.195848855774348, "grad_norm": 0.2915637791156769, "learning_rate": 1.2253001359700163e-07, "loss": 0.1888, "step": 8252 }, { "epoch": 2.1961149547631718, "grad_norm": 0.28412342071533203, "learning_rate": 1.225133910228326e-07, "loss": 0.1765, "step": 8253 }, { "epoch": 2.196381053751996, "grad_norm": 0.3004191815853119, "learning_rate": 1.2249676779335728e-07, "loss": 0.2026, "step": 8254 }, { "epoch": 2.1966471527408196, "grad_norm": 0.29754894971847534, "learning_rate": 1.2248014390905938e-07, "loss": 0.1913, "step": 8255 }, { "epoch": 2.1969132517296432, "grad_norm": 0.2909834682941437, "learning_rate": 1.224635193704229e-07, "loss": 0.1907, "step": 8256 }, { "epoch": 2.1971793507184674, "grad_norm": 0.2792259752750397, "learning_rate": 1.2244689417793168e-07, "loss": 0.1846, "step": 8257 }, { "epoch": 2.197445449707291, "grad_norm": 0.2866796553134918, "learning_rate": 1.224302683320697e-07, "loss": 0.1865, "step": 8258 }, { "epoch": 2.197711548696115, "grad_norm": 0.27518990635871887, "learning_rate": 1.224136418333208e-07, "loss": 0.1821, "step": 8259 }, { "epoch": 2.197977647684939, "grad_norm": 0.340356707572937, "learning_rate": 1.22397014682169e-07, "loss": 0.2254, "step": 8260 }, { "epoch": 2.1982437466737625, "grad_norm": 0.2939477562904358, "learning_rate": 1.2238038687909826e-07, "loss": 0.2124, "step": 8261 }, { "epoch": 2.1985098456625867, "grad_norm": 0.31536537408828735, "learning_rate": 1.2236375842459254e-07, "loss": 0.2158, "step": 8262 }, { "epoch": 2.1987759446514104, "grad_norm": 0.2769858241081238, "learning_rate": 1.2234712931913594e-07, "loss": 0.1813, "step": 8263 }, { "epoch": 2.199042043640234, "grad_norm": 0.26390063762664795, "learning_rate": 1.2233049956321237e-07, "loss": 0.1703, "step": 8264 }, { "epoch": 2.199308142629058, "grad_norm": 0.2834869921207428, "learning_rate": 1.2231386915730595e-07, "loss": 0.1939, "step": 8265 }, { "epoch": 2.199574241617882, "grad_norm": 0.35679593682289124, "learning_rate": 1.2229723810190073e-07, "loss": 0.222, "step": 8266 }, { "epoch": 2.1998403406067055, "grad_norm": 0.2738540768623352, "learning_rate": 1.2228060639748083e-07, "loss": 0.1776, "step": 8267 }, { "epoch": 2.2001064395955297, "grad_norm": 0.2946462631225586, "learning_rate": 1.2226397404453032e-07, "loss": 0.2143, "step": 8268 }, { "epoch": 2.2003725385843533, "grad_norm": 0.341745525598526, "learning_rate": 1.2224734104353335e-07, "loss": 0.2202, "step": 8269 }, { "epoch": 2.200638637573177, "grad_norm": 0.28148534893989563, "learning_rate": 1.2223070739497403e-07, "loss": 0.1945, "step": 8270 }, { "epoch": 2.200904736562001, "grad_norm": 0.2745470702648163, "learning_rate": 1.2221407309933652e-07, "loss": 0.1858, "step": 8271 }, { "epoch": 2.201170835550825, "grad_norm": 0.3101697564125061, "learning_rate": 1.2219743815710507e-07, "loss": 0.1977, "step": 8272 }, { "epoch": 2.201436934539649, "grad_norm": 0.307311087846756, "learning_rate": 1.2218080256876377e-07, "loss": 0.2002, "step": 8273 }, { "epoch": 2.2017030335284726, "grad_norm": 0.26463738083839417, "learning_rate": 1.2216416633479692e-07, "loss": 0.1881, "step": 8274 }, { "epoch": 2.2019691325172963, "grad_norm": 0.2981972098350525, "learning_rate": 1.2214752945568875e-07, "loss": 0.2026, "step": 8275 }, { "epoch": 2.2022352315061204, "grad_norm": 0.2604946494102478, "learning_rate": 1.2213089193192346e-07, "loss": 0.1848, "step": 8276 }, { "epoch": 2.202501330494944, "grad_norm": 0.36316749453544617, "learning_rate": 1.2211425376398544e-07, "loss": 0.2104, "step": 8277 }, { "epoch": 2.202767429483768, "grad_norm": 0.28322237730026245, "learning_rate": 1.2209761495235886e-07, "loss": 0.1866, "step": 8278 }, { "epoch": 2.203033528472592, "grad_norm": 0.28301453590393066, "learning_rate": 1.220809754975281e-07, "loss": 0.1881, "step": 8279 }, { "epoch": 2.2032996274614156, "grad_norm": 0.2754552364349365, "learning_rate": 1.2206433539997748e-07, "loss": 0.2006, "step": 8280 }, { "epoch": 2.2035657264502393, "grad_norm": 0.28464972972869873, "learning_rate": 1.2204769466019136e-07, "loss": 0.2018, "step": 8281 }, { "epoch": 2.2038318254390634, "grad_norm": 0.2813706696033478, "learning_rate": 1.2203105327865406e-07, "loss": 0.1954, "step": 8282 }, { "epoch": 2.204097924427887, "grad_norm": 0.3071606457233429, "learning_rate": 1.2201441125585e-07, "loss": 0.2113, "step": 8283 }, { "epoch": 2.2043640234167112, "grad_norm": 0.2751637399196625, "learning_rate": 1.2199776859226363e-07, "loss": 0.1781, "step": 8284 }, { "epoch": 2.204630122405535, "grad_norm": 0.2792101204395294, "learning_rate": 1.219811252883793e-07, "loss": 0.197, "step": 8285 }, { "epoch": 2.2048962213943586, "grad_norm": 0.29104214906692505, "learning_rate": 1.2196448134468152e-07, "loss": 0.191, "step": 8286 }, { "epoch": 2.2051623203831827, "grad_norm": 0.2820243537425995, "learning_rate": 1.219478367616547e-07, "loss": 0.1782, "step": 8287 }, { "epoch": 2.2054284193720064, "grad_norm": 0.25724464654922485, "learning_rate": 1.219311915397833e-07, "loss": 0.176, "step": 8288 }, { "epoch": 2.20569451836083, "grad_norm": 0.28372395038604736, "learning_rate": 1.219145456795519e-07, "loss": 0.207, "step": 8289 }, { "epoch": 2.205960617349654, "grad_norm": 0.317050963640213, "learning_rate": 1.21897899181445e-07, "loss": 0.2047, "step": 8290 }, { "epoch": 2.206226716338478, "grad_norm": 0.29005369544029236, "learning_rate": 1.2188125204594706e-07, "loss": 0.2082, "step": 8291 }, { "epoch": 2.2064928153273016, "grad_norm": 0.301969438791275, "learning_rate": 1.218646042735427e-07, "loss": 0.1979, "step": 8292 }, { "epoch": 2.2067589143161257, "grad_norm": 0.27435302734375, "learning_rate": 1.2184795586471647e-07, "loss": 0.1998, "step": 8293 }, { "epoch": 2.2070250133049494, "grad_norm": 0.2824346125125885, "learning_rate": 1.2183130681995299e-07, "loss": 0.1815, "step": 8294 }, { "epoch": 2.207291112293773, "grad_norm": 0.2747620642185211, "learning_rate": 1.2181465713973686e-07, "loss": 0.1976, "step": 8295 }, { "epoch": 2.207557211282597, "grad_norm": 0.32646656036376953, "learning_rate": 1.217980068245527e-07, "loss": 0.2209, "step": 8296 }, { "epoch": 2.207823310271421, "grad_norm": 0.2965348958969116, "learning_rate": 1.2178135587488515e-07, "loss": 0.1881, "step": 8297 }, { "epoch": 2.208089409260245, "grad_norm": 0.2866843342781067, "learning_rate": 1.217647042912189e-07, "loss": 0.1995, "step": 8298 }, { "epoch": 2.2083555082490687, "grad_norm": 0.30771610140800476, "learning_rate": 1.217480520740386e-07, "loss": 0.2195, "step": 8299 }, { "epoch": 2.2086216072378924, "grad_norm": 0.2876879572868347, "learning_rate": 1.21731399223829e-07, "loss": 0.1953, "step": 8300 }, { "epoch": 2.2088877062267165, "grad_norm": 0.28140661120414734, "learning_rate": 1.2171474574107477e-07, "loss": 0.1706, "step": 8301 }, { "epoch": 2.20915380521554, "grad_norm": 0.27885404229164124, "learning_rate": 1.2169809162626066e-07, "loss": 0.1838, "step": 8302 }, { "epoch": 2.209419904204364, "grad_norm": 0.27517181634902954, "learning_rate": 1.2168143687987144e-07, "loss": 0.1819, "step": 8303 }, { "epoch": 2.209686003193188, "grad_norm": 0.2696007192134857, "learning_rate": 1.2166478150239194e-07, "loss": 0.1834, "step": 8304 }, { "epoch": 2.2099521021820117, "grad_norm": 0.3000546991825104, "learning_rate": 1.2164812549430686e-07, "loss": 0.206, "step": 8305 }, { "epoch": 2.210218201170836, "grad_norm": 0.2853257656097412, "learning_rate": 1.2163146885610106e-07, "loss": 0.1976, "step": 8306 }, { "epoch": 2.2104843001596595, "grad_norm": 0.2929277718067169, "learning_rate": 1.2161481158825938e-07, "loss": 0.2107, "step": 8307 }, { "epoch": 2.210750399148483, "grad_norm": 0.286323219537735, "learning_rate": 1.2159815369126666e-07, "loss": 0.1989, "step": 8308 }, { "epoch": 2.2110164981373073, "grad_norm": 0.2868177890777588, "learning_rate": 1.2158149516560773e-07, "loss": 0.1814, "step": 8309 }, { "epoch": 2.211282597126131, "grad_norm": 0.29011833667755127, "learning_rate": 1.2156483601176754e-07, "loss": 0.1967, "step": 8310 }, { "epoch": 2.2115486961149546, "grad_norm": 0.3324914872646332, "learning_rate": 1.2154817623023097e-07, "loss": 0.2164, "step": 8311 }, { "epoch": 2.2118147951037788, "grad_norm": 0.2751784920692444, "learning_rate": 1.215315158214829e-07, "loss": 0.1836, "step": 8312 }, { "epoch": 2.2120808940926024, "grad_norm": 0.291699081659317, "learning_rate": 1.215148547860084e-07, "loss": 0.1946, "step": 8313 }, { "epoch": 2.212346993081426, "grad_norm": 0.30247703194618225, "learning_rate": 1.2149819312429227e-07, "loss": 0.1995, "step": 8314 }, { "epoch": 2.2126130920702503, "grad_norm": 0.2675580680370331, "learning_rate": 1.2148153083681955e-07, "loss": 0.2023, "step": 8315 }, { "epoch": 2.212879191059074, "grad_norm": 0.28835898637771606, "learning_rate": 1.2146486792407526e-07, "loss": 0.2094, "step": 8316 }, { "epoch": 2.2131452900478976, "grad_norm": 0.29868799448013306, "learning_rate": 1.214482043865444e-07, "loss": 0.2146, "step": 8317 }, { "epoch": 2.2134113890367217, "grad_norm": 0.2990188002586365, "learning_rate": 1.2143154022471201e-07, "loss": 0.216, "step": 8318 }, { "epoch": 2.2136774880255454, "grad_norm": 0.2860904037952423, "learning_rate": 1.2141487543906313e-07, "loss": 0.2002, "step": 8319 }, { "epoch": 2.2139435870143696, "grad_norm": 0.28190067410469055, "learning_rate": 1.213982100300828e-07, "loss": 0.1878, "step": 8320 }, { "epoch": 2.2142096860031932, "grad_norm": 0.2842577397823334, "learning_rate": 1.2138154399825615e-07, "loss": 0.1805, "step": 8321 }, { "epoch": 2.214475784992017, "grad_norm": 0.2863568961620331, "learning_rate": 1.2136487734406828e-07, "loss": 0.2143, "step": 8322 }, { "epoch": 2.214741883980841, "grad_norm": 0.3033764064311981, "learning_rate": 1.213482100680043e-07, "loss": 0.1748, "step": 8323 }, { "epoch": 2.2150079829696647, "grad_norm": 0.2811336815357208, "learning_rate": 1.2133154217054934e-07, "loss": 0.1784, "step": 8324 }, { "epoch": 2.2152740819584884, "grad_norm": 0.27289092540740967, "learning_rate": 1.213148736521886e-07, "loss": 0.1903, "step": 8325 }, { "epoch": 2.2155401809473125, "grad_norm": 0.37446725368499756, "learning_rate": 1.2129820451340721e-07, "loss": 0.2054, "step": 8326 }, { "epoch": 2.215806279936136, "grad_norm": 0.2767349183559418, "learning_rate": 1.212815347546904e-07, "loss": 0.1788, "step": 8327 }, { "epoch": 2.21607237892496, "grad_norm": 0.34850040078163147, "learning_rate": 1.2126486437652338e-07, "loss": 0.1957, "step": 8328 }, { "epoch": 2.216338477913784, "grad_norm": 0.28863176703453064, "learning_rate": 1.2124819337939136e-07, "loss": 0.1999, "step": 8329 }, { "epoch": 2.2166045769026077, "grad_norm": 0.2916102409362793, "learning_rate": 1.212315217637796e-07, "loss": 0.1911, "step": 8330 }, { "epoch": 2.216870675891432, "grad_norm": 0.26582470536231995, "learning_rate": 1.2121484953017338e-07, "loss": 0.1743, "step": 8331 }, { "epoch": 2.2171367748802555, "grad_norm": 0.29662346839904785, "learning_rate": 1.2119817667905795e-07, "loss": 0.1865, "step": 8332 }, { "epoch": 2.217402873869079, "grad_norm": 0.28826582431793213, "learning_rate": 1.2118150321091865e-07, "loss": 0.2028, "step": 8333 }, { "epoch": 2.2176689728579033, "grad_norm": 0.28199535608291626, "learning_rate": 1.2116482912624079e-07, "loss": 0.1981, "step": 8334 }, { "epoch": 2.217935071846727, "grad_norm": 0.289010614156723, "learning_rate": 1.2114815442550972e-07, "loss": 0.1992, "step": 8335 }, { "epoch": 2.2182011708355507, "grad_norm": 0.2638569176197052, "learning_rate": 1.2113147910921076e-07, "loss": 0.1931, "step": 8336 }, { "epoch": 2.218467269824375, "grad_norm": 0.30053985118865967, "learning_rate": 1.211148031778293e-07, "loss": 0.193, "step": 8337 }, { "epoch": 2.2187333688131985, "grad_norm": 0.30186712741851807, "learning_rate": 1.210981266318508e-07, "loss": 0.1843, "step": 8338 }, { "epoch": 2.218999467802022, "grad_norm": 0.3158533573150635, "learning_rate": 1.2108144947176057e-07, "loss": 0.2096, "step": 8339 }, { "epoch": 2.2192655667908463, "grad_norm": 0.2859300374984741, "learning_rate": 1.210647716980441e-07, "loss": 0.1966, "step": 8340 }, { "epoch": 2.21953166577967, "grad_norm": 0.27115052938461304, "learning_rate": 1.210480933111868e-07, "loss": 0.2009, "step": 8341 }, { "epoch": 2.2197977647684937, "grad_norm": 0.287317156791687, "learning_rate": 1.2103141431167418e-07, "loss": 0.1901, "step": 8342 }, { "epoch": 2.220063863757318, "grad_norm": 0.28102993965148926, "learning_rate": 1.2101473469999167e-07, "loss": 0.1822, "step": 8343 }, { "epoch": 2.2203299627461415, "grad_norm": 0.2736363112926483, "learning_rate": 1.2099805447662482e-07, "loss": 0.1917, "step": 8344 }, { "epoch": 2.2205960617349656, "grad_norm": 0.27124112844467163, "learning_rate": 1.2098137364205915e-07, "loss": 0.1902, "step": 8345 }, { "epoch": 2.2208621607237893, "grad_norm": 0.2546786963939667, "learning_rate": 1.2096469219678012e-07, "loss": 0.1783, "step": 8346 }, { "epoch": 2.221128259712613, "grad_norm": 0.30829179286956787, "learning_rate": 1.2094801014127333e-07, "loss": 0.2051, "step": 8347 }, { "epoch": 2.221394358701437, "grad_norm": 0.2745251953601837, "learning_rate": 1.209313274760244e-07, "loss": 0.1989, "step": 8348 }, { "epoch": 2.2216604576902608, "grad_norm": 0.2766488492488861, "learning_rate": 1.2091464420151885e-07, "loss": 0.18, "step": 8349 }, { "epoch": 2.2219265566790845, "grad_norm": 0.284777969121933, "learning_rate": 1.2089796031824234e-07, "loss": 0.1796, "step": 8350 }, { "epoch": 2.2221926556679086, "grad_norm": 0.2785731554031372, "learning_rate": 1.2088127582668044e-07, "loss": 0.1883, "step": 8351 }, { "epoch": 2.2224587546567323, "grad_norm": 0.29547980427742004, "learning_rate": 1.208645907273188e-07, "loss": 0.1947, "step": 8352 }, { "epoch": 2.222724853645556, "grad_norm": 0.27760395407676697, "learning_rate": 1.2084790502064314e-07, "loss": 0.194, "step": 8353 }, { "epoch": 2.22299095263438, "grad_norm": 0.287855863571167, "learning_rate": 1.2083121870713905e-07, "loss": 0.2094, "step": 8354 }, { "epoch": 2.2232570516232038, "grad_norm": 0.2771185338497162, "learning_rate": 1.208145317872923e-07, "loss": 0.1814, "step": 8355 }, { "epoch": 2.223523150612028, "grad_norm": 0.3058951497077942, "learning_rate": 1.2079784426158856e-07, "loss": 0.2048, "step": 8356 }, { "epoch": 2.2237892496008516, "grad_norm": 0.28281018137931824, "learning_rate": 1.2078115613051357e-07, "loss": 0.1865, "step": 8357 }, { "epoch": 2.2240553485896752, "grad_norm": 0.2638280987739563, "learning_rate": 1.2076446739455312e-07, "loss": 0.1864, "step": 8358 }, { "epoch": 2.2243214475784994, "grad_norm": 0.308829665184021, "learning_rate": 1.207477780541929e-07, "loss": 0.208, "step": 8359 }, { "epoch": 2.224587546567323, "grad_norm": 0.28738534450531006, "learning_rate": 1.2073108810991878e-07, "loss": 0.1951, "step": 8360 }, { "epoch": 2.2248536455561467, "grad_norm": 0.31076350808143616, "learning_rate": 1.2071439756221647e-07, "loss": 0.2006, "step": 8361 }, { "epoch": 2.225119744544971, "grad_norm": 0.2914457321166992, "learning_rate": 1.2069770641157186e-07, "loss": 0.1995, "step": 8362 }, { "epoch": 2.2253858435337945, "grad_norm": 0.27936452627182007, "learning_rate": 1.2068101465847073e-07, "loss": 0.1901, "step": 8363 }, { "epoch": 2.225651942522618, "grad_norm": 0.31311988830566406, "learning_rate": 1.2066432230339895e-07, "loss": 0.2037, "step": 8364 }, { "epoch": 2.2259180415114423, "grad_norm": 0.2648821175098419, "learning_rate": 1.2064762934684243e-07, "loss": 0.1841, "step": 8365 }, { "epoch": 2.226184140500266, "grad_norm": 0.2684688866138458, "learning_rate": 1.20630935789287e-07, "loss": 0.179, "step": 8366 }, { "epoch": 2.2264502394890897, "grad_norm": 0.27526625990867615, "learning_rate": 1.2061424163121864e-07, "loss": 0.1911, "step": 8367 }, { "epoch": 2.226716338477914, "grad_norm": 0.30361393094062805, "learning_rate": 1.205975468731232e-07, "loss": 0.1973, "step": 8368 }, { "epoch": 2.2269824374667375, "grad_norm": 0.3047080934047699, "learning_rate": 1.2058085151548668e-07, "loss": 0.218, "step": 8369 }, { "epoch": 2.2272485364555616, "grad_norm": 0.2951933443546295, "learning_rate": 1.20564155558795e-07, "loss": 0.1961, "step": 8370 }, { "epoch": 2.2275146354443853, "grad_norm": 0.29618045687675476, "learning_rate": 1.205474590035341e-07, "loss": 0.211, "step": 8371 }, { "epoch": 2.227780734433209, "grad_norm": 0.29758235812187195, "learning_rate": 1.205307618501901e-07, "loss": 0.1814, "step": 8372 }, { "epoch": 2.228046833422033, "grad_norm": 0.30392125248908997, "learning_rate": 1.2051406409924887e-07, "loss": 0.2023, "step": 8373 }, { "epoch": 2.228312932410857, "grad_norm": 0.28529736399650574, "learning_rate": 1.2049736575119651e-07, "loss": 0.1934, "step": 8374 }, { "epoch": 2.2285790313996805, "grad_norm": 0.2897436320781708, "learning_rate": 1.2048066680651908e-07, "loss": 0.2042, "step": 8375 }, { "epoch": 2.2288451303885046, "grad_norm": 0.3237864375114441, "learning_rate": 1.204639672657026e-07, "loss": 0.2021, "step": 8376 }, { "epoch": 2.2291112293773283, "grad_norm": 0.28115198016166687, "learning_rate": 1.2044726712923315e-07, "loss": 0.1876, "step": 8377 }, { "epoch": 2.2293773283661524, "grad_norm": 0.2957967221736908, "learning_rate": 1.2043056639759686e-07, "loss": 0.1992, "step": 8378 }, { "epoch": 2.229643427354976, "grad_norm": 0.30582359433174133, "learning_rate": 1.2041386507127983e-07, "loss": 0.2011, "step": 8379 }, { "epoch": 2.2299095263438, "grad_norm": 0.2976044714450836, "learning_rate": 1.2039716315076818e-07, "loss": 0.2111, "step": 8380 }, { "epoch": 2.230175625332624, "grad_norm": 0.27438631653785706, "learning_rate": 1.203804606365481e-07, "loss": 0.1742, "step": 8381 }, { "epoch": 2.2304417243214476, "grad_norm": 0.2690243124961853, "learning_rate": 1.203637575291057e-07, "loss": 0.1788, "step": 8382 }, { "epoch": 2.2307078233102713, "grad_norm": 0.2923828661441803, "learning_rate": 1.2034705382892721e-07, "loss": 0.2064, "step": 8383 }, { "epoch": 2.2309739222990954, "grad_norm": 0.3116092383861542, "learning_rate": 1.203303495364988e-07, "loss": 0.2078, "step": 8384 }, { "epoch": 2.231240021287919, "grad_norm": 0.28421562910079956, "learning_rate": 1.2031364465230672e-07, "loss": 0.1985, "step": 8385 }, { "epoch": 2.2315061202767428, "grad_norm": 0.3089025020599365, "learning_rate": 1.2029693917683713e-07, "loss": 0.2005, "step": 8386 }, { "epoch": 2.231772219265567, "grad_norm": 0.3057638704776764, "learning_rate": 1.202802331105764e-07, "loss": 0.2147, "step": 8387 }, { "epoch": 2.2320383182543906, "grad_norm": 0.278959184885025, "learning_rate": 1.2026352645401072e-07, "loss": 0.1915, "step": 8388 }, { "epoch": 2.2323044172432143, "grad_norm": 0.28708741068840027, "learning_rate": 1.2024681920762642e-07, "loss": 0.1984, "step": 8389 }, { "epoch": 2.2325705162320384, "grad_norm": 0.2962665259838104, "learning_rate": 1.2023011137190976e-07, "loss": 0.2011, "step": 8390 }, { "epoch": 2.232836615220862, "grad_norm": 0.3097464442253113, "learning_rate": 1.202134029473471e-07, "loss": 0.209, "step": 8391 }, { "epoch": 2.233102714209686, "grad_norm": 0.29420313239097595, "learning_rate": 1.2019669393442478e-07, "loss": 0.2048, "step": 8392 }, { "epoch": 2.23336881319851, "grad_norm": 0.26107287406921387, "learning_rate": 1.201799843336291e-07, "loss": 0.1853, "step": 8393 }, { "epoch": 2.2336349121873336, "grad_norm": 0.36329910159111023, "learning_rate": 1.201632741454465e-07, "loss": 0.2218, "step": 8394 }, { "epoch": 2.2339010111761577, "grad_norm": 0.2907710373401642, "learning_rate": 1.2014656337036333e-07, "loss": 0.1887, "step": 8395 }, { "epoch": 2.2341671101649814, "grad_norm": 0.30950209498405457, "learning_rate": 1.20129852008866e-07, "loss": 0.1953, "step": 8396 }, { "epoch": 2.234433209153805, "grad_norm": 0.31727540493011475, "learning_rate": 1.2011314006144095e-07, "loss": 0.2183, "step": 8397 }, { "epoch": 2.234699308142629, "grad_norm": 0.2931196391582489, "learning_rate": 1.2009642752857464e-07, "loss": 0.1971, "step": 8398 }, { "epoch": 2.234965407131453, "grad_norm": 0.2714839279651642, "learning_rate": 1.2007971441075348e-07, "loss": 0.1932, "step": 8399 }, { "epoch": 2.2352315061202765, "grad_norm": 0.2543438971042633, "learning_rate": 1.2006300070846398e-07, "loss": 0.1774, "step": 8400 }, { "epoch": 2.2354976051091007, "grad_norm": 0.3091367185115814, "learning_rate": 1.2004628642219263e-07, "loss": 0.2171, "step": 8401 }, { "epoch": 2.2357637040979244, "grad_norm": 0.31077665090560913, "learning_rate": 1.2002957155242592e-07, "loss": 0.2187, "step": 8402 }, { "epoch": 2.2360298030867485, "grad_norm": 0.27965059876441956, "learning_rate": 1.2001285609965042e-07, "loss": 0.2136, "step": 8403 }, { "epoch": 2.236295902075572, "grad_norm": 0.289031982421875, "learning_rate": 1.199961400643526e-07, "loss": 0.1812, "step": 8404 }, { "epoch": 2.236562001064396, "grad_norm": 0.3239278793334961, "learning_rate": 1.1997942344701906e-07, "loss": 0.2183, "step": 8405 }, { "epoch": 2.23682810005322, "grad_norm": 0.27047988772392273, "learning_rate": 1.1996270624813642e-07, "loss": 0.1707, "step": 8406 }, { "epoch": 2.2370941990420437, "grad_norm": 0.27018141746520996, "learning_rate": 1.1994598846819118e-07, "loss": 0.1935, "step": 8407 }, { "epoch": 2.2373602980308673, "grad_norm": 0.3091118633747101, "learning_rate": 1.1992927010767005e-07, "loss": 0.2039, "step": 8408 }, { "epoch": 2.2376263970196915, "grad_norm": 0.2839283049106598, "learning_rate": 1.1991255116705958e-07, "loss": 0.1947, "step": 8409 }, { "epoch": 2.237892496008515, "grad_norm": 0.29171305894851685, "learning_rate": 1.198958316468465e-07, "loss": 0.1997, "step": 8410 }, { "epoch": 2.238158594997339, "grad_norm": 0.28741785883903503, "learning_rate": 1.1987911154751735e-07, "loss": 0.1936, "step": 8411 }, { "epoch": 2.238424693986163, "grad_norm": 0.3083969056606293, "learning_rate": 1.1986239086955894e-07, "loss": 0.2089, "step": 8412 }, { "epoch": 2.2386907929749866, "grad_norm": 0.313868910074234, "learning_rate": 1.1984566961345788e-07, "loss": 0.1976, "step": 8413 }, { "epoch": 2.2389568919638103, "grad_norm": 0.32019752264022827, "learning_rate": 1.198289477797009e-07, "loss": 0.1916, "step": 8414 }, { "epoch": 2.2392229909526344, "grad_norm": 0.31336531043052673, "learning_rate": 1.198122253687747e-07, "loss": 0.2057, "step": 8415 }, { "epoch": 2.239489089941458, "grad_norm": 0.2713046669960022, "learning_rate": 1.197955023811661e-07, "loss": 0.1917, "step": 8416 }, { "epoch": 2.2397551889302822, "grad_norm": 0.29640284180641174, "learning_rate": 1.1977877881736185e-07, "loss": 0.1949, "step": 8417 }, { "epoch": 2.240021287919106, "grad_norm": 0.2916387617588043, "learning_rate": 1.1976205467784867e-07, "loss": 0.2109, "step": 8418 }, { "epoch": 2.2402873869079296, "grad_norm": 0.2773827314376831, "learning_rate": 1.197453299631134e-07, "loss": 0.2094, "step": 8419 }, { "epoch": 2.2405534858967537, "grad_norm": 0.30495309829711914, "learning_rate": 1.1972860467364286e-07, "loss": 0.2023, "step": 8420 }, { "epoch": 2.2408195848855774, "grad_norm": 0.2798457443714142, "learning_rate": 1.1971187880992384e-07, "loss": 0.1955, "step": 8421 }, { "epoch": 2.241085683874401, "grad_norm": 0.5691960453987122, "learning_rate": 1.196951523724432e-07, "loss": 0.1909, "step": 8422 }, { "epoch": 2.2413517828632252, "grad_norm": 0.28045159578323364, "learning_rate": 1.1967842536168783e-07, "loss": 0.2141, "step": 8423 }, { "epoch": 2.241617881852049, "grad_norm": 0.2907545268535614, "learning_rate": 1.196616977781446e-07, "loss": 0.1806, "step": 8424 }, { "epoch": 2.241883980840873, "grad_norm": 0.26808127760887146, "learning_rate": 1.1964496962230037e-07, "loss": 0.1779, "step": 8425 }, { "epoch": 2.2421500798296967, "grad_norm": 0.285724401473999, "learning_rate": 1.1962824089464212e-07, "loss": 0.2033, "step": 8426 }, { "epoch": 2.2424161788185204, "grad_norm": 0.296172171831131, "learning_rate": 1.196115115956567e-07, "loss": 0.218, "step": 8427 }, { "epoch": 2.2426822778073445, "grad_norm": 0.29087093472480774, "learning_rate": 1.1959478172583111e-07, "loss": 0.1928, "step": 8428 }, { "epoch": 2.242948376796168, "grad_norm": 0.2732618749141693, "learning_rate": 1.195780512856523e-07, "loss": 0.1928, "step": 8429 }, { "epoch": 2.243214475784992, "grad_norm": 0.3278380334377289, "learning_rate": 1.1956132027560728e-07, "loss": 0.2178, "step": 8430 }, { "epoch": 2.243480574773816, "grad_norm": 0.30016669631004333, "learning_rate": 1.1954458869618301e-07, "loss": 0.1931, "step": 8431 }, { "epoch": 2.2437466737626397, "grad_norm": 0.2830526828765869, "learning_rate": 1.195278565478665e-07, "loss": 0.1848, "step": 8432 }, { "epoch": 2.2440127727514634, "grad_norm": 0.29657265543937683, "learning_rate": 1.1951112383114477e-07, "loss": 0.2048, "step": 8433 }, { "epoch": 2.2442788717402875, "grad_norm": 0.2797603905200958, "learning_rate": 1.194943905465049e-07, "loss": 0.1881, "step": 8434 }, { "epoch": 2.244544970729111, "grad_norm": 0.31402328610420227, "learning_rate": 1.1947765669443396e-07, "loss": 0.1996, "step": 8435 }, { "epoch": 2.244811069717935, "grad_norm": 0.2844025790691376, "learning_rate": 1.1946092227541898e-07, "loss": 0.2011, "step": 8436 }, { "epoch": 2.245077168706759, "grad_norm": 0.3576471507549286, "learning_rate": 1.1944418728994706e-07, "loss": 0.2048, "step": 8437 }, { "epoch": 2.2453432676955827, "grad_norm": 0.3039526045322418, "learning_rate": 1.1942745173850536e-07, "loss": 0.2002, "step": 8438 }, { "epoch": 2.245609366684407, "grad_norm": 0.29668647050857544, "learning_rate": 1.1941071562158103e-07, "loss": 0.2083, "step": 8439 }, { "epoch": 2.2458754656732305, "grad_norm": 0.2899618446826935, "learning_rate": 1.193939789396611e-07, "loss": 0.2074, "step": 8440 }, { "epoch": 2.246141564662054, "grad_norm": 0.2863354980945587, "learning_rate": 1.1937724169323283e-07, "loss": 0.2071, "step": 8441 }, { "epoch": 2.2464076636508783, "grad_norm": 0.27981072664260864, "learning_rate": 1.1936050388278335e-07, "loss": 0.1964, "step": 8442 }, { "epoch": 2.246673762639702, "grad_norm": 0.2671913206577301, "learning_rate": 1.193437655087999e-07, "loss": 0.1875, "step": 8443 }, { "epoch": 2.2469398616285257, "grad_norm": 0.3091306984424591, "learning_rate": 1.1932702657176964e-07, "loss": 0.2121, "step": 8444 }, { "epoch": 2.24720596061735, "grad_norm": 0.28559941053390503, "learning_rate": 1.1931028707217982e-07, "loss": 0.1828, "step": 8445 }, { "epoch": 2.2474720596061735, "grad_norm": 0.2999703586101532, "learning_rate": 1.1929354701051767e-07, "loss": 0.2, "step": 8446 }, { "epoch": 2.247738158594997, "grad_norm": 0.2898949384689331, "learning_rate": 1.1927680638727048e-07, "loss": 0.1976, "step": 8447 }, { "epoch": 2.2480042575838213, "grad_norm": 0.34501221776008606, "learning_rate": 1.1926006520292553e-07, "loss": 0.219, "step": 8448 }, { "epoch": 2.248270356572645, "grad_norm": 0.3318520784378052, "learning_rate": 1.1924332345797005e-07, "loss": 0.2104, "step": 8449 }, { "epoch": 2.248536455561469, "grad_norm": 0.2882716953754425, "learning_rate": 1.1922658115289138e-07, "loss": 0.1929, "step": 8450 }, { "epoch": 2.2488025545502928, "grad_norm": 0.2683003842830658, "learning_rate": 1.192098382881769e-07, "loss": 0.1888, "step": 8451 }, { "epoch": 2.2490686535391164, "grad_norm": 0.29329606890678406, "learning_rate": 1.1919309486431389e-07, "loss": 0.1906, "step": 8452 }, { "epoch": 2.2493347525279406, "grad_norm": 0.3100540041923523, "learning_rate": 1.1917635088178972e-07, "loss": 0.2042, "step": 8453 }, { "epoch": 2.2496008515167643, "grad_norm": 0.3216390907764435, "learning_rate": 1.1915960634109174e-07, "loss": 0.211, "step": 8454 }, { "epoch": 2.249866950505588, "grad_norm": 0.2952669560909271, "learning_rate": 1.191428612427074e-07, "loss": 0.1875, "step": 8455 }, { "epoch": 2.250133049494412, "grad_norm": 0.30614033341407776, "learning_rate": 1.1912611558712405e-07, "loss": 0.2071, "step": 8456 }, { "epoch": 2.2503991484832357, "grad_norm": 0.33527424931526184, "learning_rate": 1.1910936937482917e-07, "loss": 0.1819, "step": 8457 }, { "epoch": 2.2506652474720594, "grad_norm": 0.2734050452709198, "learning_rate": 1.1909262260631013e-07, "loss": 0.2081, "step": 8458 }, { "epoch": 2.2509313464608836, "grad_norm": 0.29640668630599976, "learning_rate": 1.1907587528205442e-07, "loss": 0.208, "step": 8459 }, { "epoch": 2.2511974454497072, "grad_norm": 0.2899770736694336, "learning_rate": 1.1905912740254951e-07, "loss": 0.2047, "step": 8460 }, { "epoch": 2.251463544438531, "grad_norm": 0.29141631722450256, "learning_rate": 1.190423789682829e-07, "loss": 0.2008, "step": 8461 }, { "epoch": 2.251729643427355, "grad_norm": 0.30214381217956543, "learning_rate": 1.190256299797421e-07, "loss": 0.1874, "step": 8462 }, { "epoch": 2.2519957424161787, "grad_norm": 0.30021893978118896, "learning_rate": 1.190088804374146e-07, "loss": 0.2078, "step": 8463 }, { "epoch": 2.252261841405003, "grad_norm": 0.2814265489578247, "learning_rate": 1.1899213034178792e-07, "loss": 0.2048, "step": 8464 }, { "epoch": 2.2525279403938265, "grad_norm": 0.2888423204421997, "learning_rate": 1.1897537969334965e-07, "loss": 0.1982, "step": 8465 }, { "epoch": 2.25279403938265, "grad_norm": 0.29711249470710754, "learning_rate": 1.1895862849258738e-07, "loss": 0.2082, "step": 8466 }, { "epoch": 2.2530601383714743, "grad_norm": 0.26683157682418823, "learning_rate": 1.1894187673998861e-07, "loss": 0.181, "step": 8467 }, { "epoch": 2.253326237360298, "grad_norm": 0.2943970859050751, "learning_rate": 1.1892512443604101e-07, "loss": 0.1954, "step": 8468 }, { "epoch": 2.2535923363491217, "grad_norm": 0.28533077239990234, "learning_rate": 1.1890837158123217e-07, "loss": 0.191, "step": 8469 }, { "epoch": 2.253858435337946, "grad_norm": 0.292529433965683, "learning_rate": 1.1889161817604973e-07, "loss": 0.2057, "step": 8470 }, { "epoch": 2.2541245343267695, "grad_norm": 0.2878079116344452, "learning_rate": 1.1887486422098136e-07, "loss": 0.1943, "step": 8471 }, { "epoch": 2.2543906333155936, "grad_norm": 0.31173810362815857, "learning_rate": 1.1885810971651472e-07, "loss": 0.2022, "step": 8472 }, { "epoch": 2.2546567323044173, "grad_norm": 0.2970435619354248, "learning_rate": 1.1884135466313741e-07, "loss": 0.2014, "step": 8473 }, { "epoch": 2.254922831293241, "grad_norm": 0.29598480463027954, "learning_rate": 1.1882459906133721e-07, "loss": 0.2007, "step": 8474 }, { "epoch": 2.255188930282065, "grad_norm": 0.28236985206604004, "learning_rate": 1.1880784291160182e-07, "loss": 0.2121, "step": 8475 }, { "epoch": 2.255455029270889, "grad_norm": 0.2952515482902527, "learning_rate": 1.1879108621441895e-07, "loss": 0.2073, "step": 8476 }, { "epoch": 2.2557211282597125, "grad_norm": 0.28288334608078003, "learning_rate": 1.1877432897027635e-07, "loss": 0.2041, "step": 8477 }, { "epoch": 2.2559872272485366, "grad_norm": 0.28741660714149475, "learning_rate": 1.1875757117966177e-07, "loss": 0.2096, "step": 8478 }, { "epoch": 2.2562533262373603, "grad_norm": 0.3043084442615509, "learning_rate": 1.18740812843063e-07, "loss": 0.1985, "step": 8479 }, { "epoch": 2.256519425226184, "grad_norm": 0.2916516065597534, "learning_rate": 1.1872405396096783e-07, "loss": 0.2019, "step": 8480 }, { "epoch": 2.256785524215008, "grad_norm": 0.29489845037460327, "learning_rate": 1.1870729453386407e-07, "loss": 0.1887, "step": 8481 }, { "epoch": 2.257051623203832, "grad_norm": 0.29452380537986755, "learning_rate": 1.1869053456223956e-07, "loss": 0.207, "step": 8482 }, { "epoch": 2.2573177221926555, "grad_norm": 0.27738288044929504, "learning_rate": 1.186737740465821e-07, "loss": 0.1967, "step": 8483 }, { "epoch": 2.2575838211814796, "grad_norm": 0.29723507165908813, "learning_rate": 1.1865701298737956e-07, "loss": 0.2005, "step": 8484 }, { "epoch": 2.2578499201703033, "grad_norm": 0.3386036455631256, "learning_rate": 1.1864025138511981e-07, "loss": 0.2071, "step": 8485 }, { "epoch": 2.258116019159127, "grad_norm": 0.3000890016555786, "learning_rate": 1.1862348924029073e-07, "loss": 0.2039, "step": 8486 }, { "epoch": 2.258382118147951, "grad_norm": 0.28963714838027954, "learning_rate": 1.1860672655338024e-07, "loss": 0.1817, "step": 8487 }, { "epoch": 2.2586482171367748, "grad_norm": 0.314532607793808, "learning_rate": 1.1858996332487627e-07, "loss": 0.1997, "step": 8488 }, { "epoch": 2.258914316125599, "grad_norm": 0.29361411929130554, "learning_rate": 1.1857319955526675e-07, "loss": 0.2079, "step": 8489 }, { "epoch": 2.2591804151144226, "grad_norm": 0.28035101294517517, "learning_rate": 1.1855643524503957e-07, "loss": 0.1946, "step": 8490 }, { "epoch": 2.2594465141032463, "grad_norm": 0.27423015236854553, "learning_rate": 1.1853967039468278e-07, "loss": 0.1751, "step": 8491 }, { "epoch": 2.2597126130920704, "grad_norm": 0.28817200660705566, "learning_rate": 1.1852290500468431e-07, "loss": 0.208, "step": 8492 }, { "epoch": 2.259978712080894, "grad_norm": 0.2686794698238373, "learning_rate": 1.1850613907553217e-07, "loss": 0.1777, "step": 8493 }, { "epoch": 2.2602448110697178, "grad_norm": 0.32932049036026, "learning_rate": 1.1848937260771437e-07, "loss": 0.2181, "step": 8494 }, { "epoch": 2.260510910058542, "grad_norm": 0.32361099123954773, "learning_rate": 1.1847260560171895e-07, "loss": 0.2012, "step": 8495 }, { "epoch": 2.2607770090473656, "grad_norm": 0.2779484987258911, "learning_rate": 1.1845583805803391e-07, "loss": 0.1861, "step": 8496 }, { "epoch": 2.2610431080361897, "grad_norm": 0.3043864071369171, "learning_rate": 1.1843906997714736e-07, "loss": 0.1927, "step": 8497 }, { "epoch": 2.2613092070250134, "grad_norm": 0.3150641620159149, "learning_rate": 1.1842230135954739e-07, "loss": 0.1997, "step": 8498 }, { "epoch": 2.261575306013837, "grad_norm": 0.28806793689727783, "learning_rate": 1.1840553220572202e-07, "loss": 0.2069, "step": 8499 }, { "epoch": 2.261841405002661, "grad_norm": 0.30316483974456787, "learning_rate": 1.1838876251615942e-07, "loss": 0.2033, "step": 8500 }, { "epoch": 2.262107503991485, "grad_norm": 0.30798593163490295, "learning_rate": 1.1837199229134767e-07, "loss": 0.2079, "step": 8501 }, { "epoch": 2.2623736029803085, "grad_norm": 0.2898472845554352, "learning_rate": 1.1835522153177494e-07, "loss": 0.1886, "step": 8502 }, { "epoch": 2.2626397019691327, "grad_norm": 0.32150253653526306, "learning_rate": 1.1833845023792941e-07, "loss": 0.2247, "step": 8503 }, { "epoch": 2.2629058009579563, "grad_norm": 0.2936309576034546, "learning_rate": 1.1832167841029916e-07, "loss": 0.2005, "step": 8504 }, { "epoch": 2.26317189994678, "grad_norm": 0.2644898593425751, "learning_rate": 1.1830490604937241e-07, "loss": 0.1709, "step": 8505 }, { "epoch": 2.263437998935604, "grad_norm": 0.3042216897010803, "learning_rate": 1.1828813315563742e-07, "loss": 0.2066, "step": 8506 }, { "epoch": 2.263704097924428, "grad_norm": 0.293872594833374, "learning_rate": 1.1827135972958233e-07, "loss": 0.2039, "step": 8507 }, { "epoch": 2.2639701969132515, "grad_norm": 0.2928999662399292, "learning_rate": 1.182545857716954e-07, "loss": 0.1931, "step": 8508 }, { "epoch": 2.2642362959020756, "grad_norm": 0.2936229109764099, "learning_rate": 1.1823781128246489e-07, "loss": 0.1832, "step": 8509 }, { "epoch": 2.2645023948908993, "grad_norm": 0.29181626439094543, "learning_rate": 1.1822103626237903e-07, "loss": 0.1908, "step": 8510 }, { "epoch": 2.264768493879723, "grad_norm": 0.31621336936950684, "learning_rate": 1.1820426071192613e-07, "loss": 0.193, "step": 8511 }, { "epoch": 2.265034592868547, "grad_norm": 0.32399433851242065, "learning_rate": 1.1818748463159445e-07, "loss": 0.236, "step": 8512 }, { "epoch": 2.265300691857371, "grad_norm": 0.2675657570362091, "learning_rate": 1.1817070802187234e-07, "loss": 0.1803, "step": 8513 }, { "epoch": 2.265566790846195, "grad_norm": 0.32494693994522095, "learning_rate": 1.1815393088324808e-07, "loss": 0.2255, "step": 8514 }, { "epoch": 2.2658328898350186, "grad_norm": 0.35646775364875793, "learning_rate": 1.1813715321621002e-07, "loss": 0.2174, "step": 8515 }, { "epoch": 2.2660989888238423, "grad_norm": 0.28334352374076843, "learning_rate": 1.1812037502124656e-07, "loss": 0.1914, "step": 8516 }, { "epoch": 2.2663650878126664, "grad_norm": 0.2854917347431183, "learning_rate": 1.1810359629884601e-07, "loss": 0.1971, "step": 8517 }, { "epoch": 2.26663118680149, "grad_norm": 0.2882292568683624, "learning_rate": 1.1808681704949677e-07, "loss": 0.2005, "step": 8518 }, { "epoch": 2.2668972857903142, "grad_norm": 0.28375551104545593, "learning_rate": 1.1807003727368722e-07, "loss": 0.1979, "step": 8519 }, { "epoch": 2.267163384779138, "grad_norm": 0.321076363325119, "learning_rate": 1.1805325697190585e-07, "loss": 0.2039, "step": 8520 }, { "epoch": 2.2674294837679616, "grad_norm": 0.27950742840766907, "learning_rate": 1.1803647614464105e-07, "loss": 0.2005, "step": 8521 }, { "epoch": 2.2676955827567857, "grad_norm": 0.30668511986732483, "learning_rate": 1.1801969479238123e-07, "loss": 0.194, "step": 8522 }, { "epoch": 2.2679616817456094, "grad_norm": 0.2965618371963501, "learning_rate": 1.180029129156149e-07, "loss": 0.1922, "step": 8523 }, { "epoch": 2.268227780734433, "grad_norm": 0.28905534744262695, "learning_rate": 1.1798613051483051e-07, "loss": 0.1964, "step": 8524 }, { "epoch": 2.2684938797232572, "grad_norm": 0.3245444893836975, "learning_rate": 1.1796934759051658e-07, "loss": 0.2246, "step": 8525 }, { "epoch": 2.268759978712081, "grad_norm": 0.273759663105011, "learning_rate": 1.1795256414316157e-07, "loss": 0.1802, "step": 8526 }, { "epoch": 2.2690260777009046, "grad_norm": 0.25721609592437744, "learning_rate": 1.1793578017325405e-07, "loss": 0.1752, "step": 8527 }, { "epoch": 2.2692921766897287, "grad_norm": 0.32944369316101074, "learning_rate": 1.1791899568128252e-07, "loss": 0.2113, "step": 8528 }, { "epoch": 2.2695582756785524, "grad_norm": 0.2832483947277069, "learning_rate": 1.1790221066773555e-07, "loss": 0.1845, "step": 8529 }, { "epoch": 2.269824374667376, "grad_norm": 0.30884358286857605, "learning_rate": 1.1788542513310176e-07, "loss": 0.2069, "step": 8530 }, { "epoch": 2.2700904736562, "grad_norm": 0.2804204523563385, "learning_rate": 1.1786863907786964e-07, "loss": 0.1956, "step": 8531 }, { "epoch": 2.270356572645024, "grad_norm": 0.28680193424224854, "learning_rate": 1.1785185250252785e-07, "loss": 0.1889, "step": 8532 }, { "epoch": 2.2706226716338476, "grad_norm": 0.32694360613822937, "learning_rate": 1.1783506540756496e-07, "loss": 0.211, "step": 8533 }, { "epoch": 2.2708887706226717, "grad_norm": 0.27614548802375793, "learning_rate": 1.1781827779346967e-07, "loss": 0.2001, "step": 8534 }, { "epoch": 2.2711548696114954, "grad_norm": 0.2821823060512543, "learning_rate": 1.1780148966073054e-07, "loss": 0.2028, "step": 8535 }, { "epoch": 2.2714209686003195, "grad_norm": 0.2764173448085785, "learning_rate": 1.1778470100983627e-07, "loss": 0.1872, "step": 8536 }, { "epoch": 2.271687067589143, "grad_norm": 0.2669466733932495, "learning_rate": 1.1776791184127553e-07, "loss": 0.1789, "step": 8537 }, { "epoch": 2.271953166577967, "grad_norm": 0.29324212670326233, "learning_rate": 1.1775112215553699e-07, "loss": 0.1959, "step": 8538 }, { "epoch": 2.272219265566791, "grad_norm": 0.279328316450119, "learning_rate": 1.1773433195310942e-07, "loss": 0.1964, "step": 8539 }, { "epoch": 2.2724853645556147, "grad_norm": 0.2861197590827942, "learning_rate": 1.1771754123448148e-07, "loss": 0.1825, "step": 8540 }, { "epoch": 2.2727514635444384, "grad_norm": 0.2831718325614929, "learning_rate": 1.177007500001419e-07, "loss": 0.1738, "step": 8541 }, { "epoch": 2.2730175625332625, "grad_norm": 0.2821536064147949, "learning_rate": 1.1768395825057947e-07, "loss": 0.1877, "step": 8542 }, { "epoch": 2.273283661522086, "grad_norm": 0.2820669114589691, "learning_rate": 1.1766716598628291e-07, "loss": 0.192, "step": 8543 }, { "epoch": 2.2735497605109103, "grad_norm": 0.27927520871162415, "learning_rate": 1.1765037320774105e-07, "loss": 0.1934, "step": 8544 }, { "epoch": 2.273815859499734, "grad_norm": 0.30275654792785645, "learning_rate": 1.1763357991544262e-07, "loss": 0.2046, "step": 8545 }, { "epoch": 2.2740819584885577, "grad_norm": 0.2780642509460449, "learning_rate": 1.1761678610987648e-07, "loss": 0.1789, "step": 8546 }, { "epoch": 2.274348057477382, "grad_norm": 0.310773640871048, "learning_rate": 1.1759999179153143e-07, "loss": 0.2005, "step": 8547 }, { "epoch": 2.2746141564662055, "grad_norm": 0.27174821496009827, "learning_rate": 1.1758319696089634e-07, "loss": 0.1803, "step": 8548 }, { "epoch": 2.274880255455029, "grad_norm": 0.2842903733253479, "learning_rate": 1.1756640161846002e-07, "loss": 0.1983, "step": 8549 }, { "epoch": 2.2751463544438533, "grad_norm": 0.3012930452823639, "learning_rate": 1.1754960576471135e-07, "loss": 0.2044, "step": 8550 }, { "epoch": 2.275412453432677, "grad_norm": 0.2694490849971771, "learning_rate": 1.1753280940013923e-07, "loss": 0.1827, "step": 8551 }, { "epoch": 2.2756785524215006, "grad_norm": 0.30968913435935974, "learning_rate": 1.1751601252523258e-07, "loss": 0.2111, "step": 8552 }, { "epoch": 2.2759446514103248, "grad_norm": 0.2778250277042389, "learning_rate": 1.1749921514048024e-07, "loss": 0.1834, "step": 8553 }, { "epoch": 2.2762107503991484, "grad_norm": 0.26939263939857483, "learning_rate": 1.1748241724637124e-07, "loss": 0.1864, "step": 8554 }, { "epoch": 2.276476849387972, "grad_norm": 0.28293564915657043, "learning_rate": 1.1746561884339443e-07, "loss": 0.2003, "step": 8555 }, { "epoch": 2.2767429483767962, "grad_norm": 0.3012126386165619, "learning_rate": 1.1744881993203879e-07, "loss": 0.2066, "step": 8556 }, { "epoch": 2.27700904736562, "grad_norm": 0.27312058210372925, "learning_rate": 1.1743202051279334e-07, "loss": 0.1911, "step": 8557 }, { "epoch": 2.2772751463544436, "grad_norm": 0.4362618923187256, "learning_rate": 1.1741522058614704e-07, "loss": 0.1913, "step": 8558 }, { "epoch": 2.2775412453432677, "grad_norm": 0.28325653076171875, "learning_rate": 1.1739842015258885e-07, "loss": 0.1917, "step": 8559 }, { "epoch": 2.2778073443320914, "grad_norm": 0.3412744402885437, "learning_rate": 1.1738161921260783e-07, "loss": 0.2194, "step": 8560 }, { "epoch": 2.2780734433209155, "grad_norm": 0.2768169641494751, "learning_rate": 1.1736481776669305e-07, "loss": 0.1703, "step": 8561 }, { "epoch": 2.2783395423097392, "grad_norm": 0.30376091599464417, "learning_rate": 1.1734801581533347e-07, "loss": 0.2065, "step": 8562 }, { "epoch": 2.278605641298563, "grad_norm": 0.2978210747241974, "learning_rate": 1.1733121335901821e-07, "loss": 0.197, "step": 8563 }, { "epoch": 2.278871740287387, "grad_norm": 0.2894818186759949, "learning_rate": 1.173144103982363e-07, "loss": 0.2049, "step": 8564 }, { "epoch": 2.2791378392762107, "grad_norm": 0.30888447165489197, "learning_rate": 1.1729760693347688e-07, "loss": 0.1975, "step": 8565 }, { "epoch": 2.2794039382650344, "grad_norm": 0.2979387938976288, "learning_rate": 1.1728080296522906e-07, "loss": 0.1971, "step": 8566 }, { "epoch": 2.2796700372538585, "grad_norm": 0.31650403141975403, "learning_rate": 1.172639984939819e-07, "loss": 0.1961, "step": 8567 }, { "epoch": 2.279936136242682, "grad_norm": 0.3047340214252472, "learning_rate": 1.1724719352022458e-07, "loss": 0.1949, "step": 8568 }, { "epoch": 2.2802022352315063, "grad_norm": 0.273120254278183, "learning_rate": 1.1723038804444623e-07, "loss": 0.1943, "step": 8569 }, { "epoch": 2.28046833422033, "grad_norm": 0.30802708864212036, "learning_rate": 1.1721358206713606e-07, "loss": 0.1953, "step": 8570 }, { "epoch": 2.2807344332091537, "grad_norm": 0.2850021719932556, "learning_rate": 1.1719677558878315e-07, "loss": 0.1951, "step": 8571 }, { "epoch": 2.281000532197978, "grad_norm": 0.29029107093811035, "learning_rate": 1.171799686098768e-07, "loss": 0.1928, "step": 8572 }, { "epoch": 2.2812666311868015, "grad_norm": 0.3039025664329529, "learning_rate": 1.1716316113090617e-07, "loss": 0.2011, "step": 8573 }, { "epoch": 2.281532730175625, "grad_norm": 0.30799195170402527, "learning_rate": 1.1714635315236045e-07, "loss": 0.2107, "step": 8574 }, { "epoch": 2.2817988291644493, "grad_norm": 0.2866242527961731, "learning_rate": 1.1712954467472895e-07, "loss": 0.2007, "step": 8575 }, { "epoch": 2.282064928153273, "grad_norm": 0.28598374128341675, "learning_rate": 1.1711273569850087e-07, "loss": 0.1814, "step": 8576 }, { "epoch": 2.2823310271420967, "grad_norm": 0.3187403082847595, "learning_rate": 1.1709592622416547e-07, "loss": 0.2064, "step": 8577 }, { "epoch": 2.282597126130921, "grad_norm": 0.2838538587093353, "learning_rate": 1.1707911625221205e-07, "loss": 0.191, "step": 8578 }, { "epoch": 2.2828632251197445, "grad_norm": 0.29595473408699036, "learning_rate": 1.1706230578312992e-07, "loss": 0.1992, "step": 8579 }, { "epoch": 2.283129324108568, "grad_norm": 0.2948421537876129, "learning_rate": 1.1704549481740835e-07, "loss": 0.1901, "step": 8580 }, { "epoch": 2.2833954230973923, "grad_norm": 0.3005967140197754, "learning_rate": 1.1702868335553669e-07, "loss": 0.196, "step": 8581 }, { "epoch": 2.283661522086216, "grad_norm": 0.3124294877052307, "learning_rate": 1.1701187139800426e-07, "loss": 0.2177, "step": 8582 }, { "epoch": 2.28392762107504, "grad_norm": 0.2915455400943756, "learning_rate": 1.1699505894530044e-07, "loss": 0.1978, "step": 8583 }, { "epoch": 2.284193720063864, "grad_norm": 0.3100857436656952, "learning_rate": 1.1697824599791459e-07, "loss": 0.211, "step": 8584 }, { "epoch": 2.2844598190526875, "grad_norm": 0.300032377243042, "learning_rate": 1.1696143255633606e-07, "loss": 0.198, "step": 8585 }, { "epoch": 2.2847259180415116, "grad_norm": 0.2643956243991852, "learning_rate": 1.1694461862105427e-07, "loss": 0.1965, "step": 8586 }, { "epoch": 2.2849920170303353, "grad_norm": 0.2821665406227112, "learning_rate": 1.1692780419255864e-07, "loss": 0.203, "step": 8587 }, { "epoch": 2.285258116019159, "grad_norm": 0.28382545709609985, "learning_rate": 1.169109892713386e-07, "loss": 0.1849, "step": 8588 }, { "epoch": 2.285524215007983, "grad_norm": 0.26803478598594666, "learning_rate": 1.1689417385788354e-07, "loss": 0.192, "step": 8589 }, { "epoch": 2.2857903139968068, "grad_norm": 0.29685690999031067, "learning_rate": 1.1687735795268294e-07, "loss": 0.208, "step": 8590 }, { "epoch": 2.286056412985631, "grad_norm": 0.3003022074699402, "learning_rate": 1.1686054155622629e-07, "loss": 0.2298, "step": 8591 }, { "epoch": 2.2863225119744546, "grad_norm": 0.28216755390167236, "learning_rate": 1.1684372466900305e-07, "loss": 0.1903, "step": 8592 }, { "epoch": 2.2865886109632783, "grad_norm": 0.31392040848731995, "learning_rate": 1.1682690729150273e-07, "loss": 0.2067, "step": 8593 }, { "epoch": 2.2868547099521024, "grad_norm": 0.2776358425617218, "learning_rate": 1.1681008942421482e-07, "loss": 0.2025, "step": 8594 }, { "epoch": 2.287120808940926, "grad_norm": 0.2964027225971222, "learning_rate": 1.1679327106762887e-07, "loss": 0.2004, "step": 8595 }, { "epoch": 2.2873869079297497, "grad_norm": 0.2756046652793884, "learning_rate": 1.1677645222223435e-07, "loss": 0.2109, "step": 8596 }, { "epoch": 2.287653006918574, "grad_norm": 0.2858763635158539, "learning_rate": 1.1675963288852094e-07, "loss": 0.1957, "step": 8597 }, { "epoch": 2.2879191059073976, "grad_norm": 0.3150143623352051, "learning_rate": 1.1674281306697808e-07, "loss": 0.2121, "step": 8598 }, { "epoch": 2.2881852048962212, "grad_norm": 0.285810649394989, "learning_rate": 1.1672599275809541e-07, "loss": 0.1896, "step": 8599 }, { "epoch": 2.2884513038850454, "grad_norm": 0.3031412363052368, "learning_rate": 1.1670917196236252e-07, "loss": 0.1988, "step": 8600 }, { "epoch": 2.288717402873869, "grad_norm": 0.2883535623550415, "learning_rate": 1.1669235068026902e-07, "loss": 0.2097, "step": 8601 }, { "epoch": 2.2889835018626927, "grad_norm": 0.3208674490451813, "learning_rate": 1.1667552891230454e-07, "loss": 0.1936, "step": 8602 }, { "epoch": 2.289249600851517, "grad_norm": 0.286836713552475, "learning_rate": 1.1665870665895872e-07, "loss": 0.1815, "step": 8603 }, { "epoch": 2.2895156998403405, "grad_norm": 0.259570837020874, "learning_rate": 1.166418839207212e-07, "loss": 0.1827, "step": 8604 }, { "epoch": 2.289781798829164, "grad_norm": 0.315440833568573, "learning_rate": 1.1662506069808164e-07, "loss": 0.2207, "step": 8605 }, { "epoch": 2.2900478978179883, "grad_norm": 0.26790106296539307, "learning_rate": 1.1660823699152973e-07, "loss": 0.1774, "step": 8606 }, { "epoch": 2.290313996806812, "grad_norm": 0.2881409823894501, "learning_rate": 1.1659141280155518e-07, "loss": 0.1913, "step": 8607 }, { "epoch": 2.290580095795636, "grad_norm": 0.28309252858161926, "learning_rate": 1.1657458812864764e-07, "loss": 0.1891, "step": 8608 }, { "epoch": 2.29084619478446, "grad_norm": 0.3067916929721832, "learning_rate": 1.1655776297329692e-07, "loss": 0.2094, "step": 8609 }, { "epoch": 2.2911122937732835, "grad_norm": 0.300963431596756, "learning_rate": 1.1654093733599269e-07, "loss": 0.1998, "step": 8610 }, { "epoch": 2.2913783927621076, "grad_norm": 0.3067825138568878, "learning_rate": 1.1652411121722472e-07, "loss": 0.202, "step": 8611 }, { "epoch": 2.2916444917509313, "grad_norm": 0.28642937541007996, "learning_rate": 1.1650728461748278e-07, "loss": 0.1986, "step": 8612 }, { "epoch": 2.291910590739755, "grad_norm": 0.2643429636955261, "learning_rate": 1.1649045753725664e-07, "loss": 0.1992, "step": 8613 }, { "epoch": 2.292176689728579, "grad_norm": 0.2901060879230499, "learning_rate": 1.1647362997703614e-07, "loss": 0.1897, "step": 8614 }, { "epoch": 2.292442788717403, "grad_norm": 0.3392128348350525, "learning_rate": 1.1645680193731102e-07, "loss": 0.2212, "step": 8615 }, { "epoch": 2.292708887706227, "grad_norm": 0.2970559000968933, "learning_rate": 1.1643997341857111e-07, "loss": 0.2006, "step": 8616 }, { "epoch": 2.2929749866950506, "grad_norm": 0.28770554065704346, "learning_rate": 1.1642314442130625e-07, "loss": 0.1855, "step": 8617 }, { "epoch": 2.2932410856838743, "grad_norm": 0.2817991375923157, "learning_rate": 1.164063149460063e-07, "loss": 0.1871, "step": 8618 }, { "epoch": 2.2935071846726984, "grad_norm": 0.2896691858768463, "learning_rate": 1.1638948499316115e-07, "loss": 0.2041, "step": 8619 }, { "epoch": 2.293773283661522, "grad_norm": 0.2882559597492218, "learning_rate": 1.1637265456326064e-07, "loss": 0.1879, "step": 8620 }, { "epoch": 2.294039382650346, "grad_norm": 0.28485700488090515, "learning_rate": 1.1635582365679466e-07, "loss": 0.1887, "step": 8621 }, { "epoch": 2.29430548163917, "grad_norm": 0.3228631317615509, "learning_rate": 1.163389922742531e-07, "loss": 0.2028, "step": 8622 }, { "epoch": 2.2945715806279936, "grad_norm": 0.2962712049484253, "learning_rate": 1.1632216041612592e-07, "loss": 0.2043, "step": 8623 }, { "epoch": 2.2948376796168173, "grad_norm": 0.2936691343784332, "learning_rate": 1.1630532808290303e-07, "loss": 0.1892, "step": 8624 }, { "epoch": 2.2951037786056414, "grad_norm": 0.2739031910896301, "learning_rate": 1.1628849527507441e-07, "loss": 0.1955, "step": 8625 }, { "epoch": 2.295369877594465, "grad_norm": 0.2988605797290802, "learning_rate": 1.1627166199312994e-07, "loss": 0.2041, "step": 8626 }, { "epoch": 2.2956359765832888, "grad_norm": 0.32525622844696045, "learning_rate": 1.1625482823755963e-07, "loss": 0.2076, "step": 8627 }, { "epoch": 2.295902075572113, "grad_norm": 0.3107789158821106, "learning_rate": 1.162379940088535e-07, "loss": 0.2172, "step": 8628 }, { "epoch": 2.2961681745609366, "grad_norm": 0.28214454650878906, "learning_rate": 1.1622115930750153e-07, "loss": 0.1943, "step": 8629 }, { "epoch": 2.2964342735497603, "grad_norm": 0.27952665090560913, "learning_rate": 1.162043241339937e-07, "loss": 0.1914, "step": 8630 }, { "epoch": 2.2967003725385844, "grad_norm": 0.278083860874176, "learning_rate": 1.161874884888201e-07, "loss": 0.192, "step": 8631 }, { "epoch": 2.296966471527408, "grad_norm": 0.28848057985305786, "learning_rate": 1.1617065237247072e-07, "loss": 0.2001, "step": 8632 }, { "epoch": 2.297232570516232, "grad_norm": 0.3000424802303314, "learning_rate": 1.1615381578543567e-07, "loss": 0.2081, "step": 8633 }, { "epoch": 2.297498669505056, "grad_norm": 0.2745084762573242, "learning_rate": 1.1613697872820493e-07, "loss": 0.1789, "step": 8634 }, { "epoch": 2.2977647684938796, "grad_norm": 0.29588010907173157, "learning_rate": 1.1612014120126869e-07, "loss": 0.1965, "step": 8635 }, { "epoch": 2.2980308674827037, "grad_norm": 0.3060033321380615, "learning_rate": 1.1610330320511696e-07, "loss": 0.1983, "step": 8636 }, { "epoch": 2.2982969664715274, "grad_norm": 0.28746920824050903, "learning_rate": 1.1608646474023991e-07, "loss": 0.1998, "step": 8637 }, { "epoch": 2.2985630654603515, "grad_norm": 0.2937203347682953, "learning_rate": 1.1606962580712765e-07, "loss": 0.1913, "step": 8638 }, { "epoch": 2.298829164449175, "grad_norm": 0.28655487298965454, "learning_rate": 1.1605278640627028e-07, "loss": 0.1927, "step": 8639 }, { "epoch": 2.299095263437999, "grad_norm": 0.29510030150413513, "learning_rate": 1.1603594653815796e-07, "loss": 0.1931, "step": 8640 }, { "epoch": 2.299361362426823, "grad_norm": 0.3146284222602844, "learning_rate": 1.1601910620328091e-07, "loss": 0.1943, "step": 8641 }, { "epoch": 2.2996274614156467, "grad_norm": 0.30109086632728577, "learning_rate": 1.1600226540212923e-07, "loss": 0.1972, "step": 8642 }, { "epoch": 2.2998935604044703, "grad_norm": 0.285927414894104, "learning_rate": 1.1598542413519321e-07, "loss": 0.1971, "step": 8643 }, { "epoch": 2.3001596593932945, "grad_norm": 0.29532986879348755, "learning_rate": 1.1596858240296297e-07, "loss": 0.1985, "step": 8644 }, { "epoch": 2.300425758382118, "grad_norm": 0.3000183701515198, "learning_rate": 1.1595174020592877e-07, "loss": 0.2038, "step": 8645 }, { "epoch": 2.300691857370942, "grad_norm": 0.2953904867172241, "learning_rate": 1.1593489754458081e-07, "loss": 0.1959, "step": 8646 }, { "epoch": 2.300957956359766, "grad_norm": 0.27436143159866333, "learning_rate": 1.1591805441940939e-07, "loss": 0.1829, "step": 8647 }, { "epoch": 2.3012240553485896, "grad_norm": 0.2900640070438385, "learning_rate": 1.1590121083090472e-07, "loss": 0.1909, "step": 8648 }, { "epoch": 2.3014901543374133, "grad_norm": 0.2833309769630432, "learning_rate": 1.1588436677955707e-07, "loss": 0.1958, "step": 8649 }, { "epoch": 2.3017562533262375, "grad_norm": 0.3243449330329895, "learning_rate": 1.1586752226585674e-07, "loss": 0.2102, "step": 8650 }, { "epoch": 2.302022352315061, "grad_norm": 0.2823745608329773, "learning_rate": 1.1585067729029405e-07, "loss": 0.1934, "step": 8651 }, { "epoch": 2.302288451303885, "grad_norm": 0.27892112731933594, "learning_rate": 1.1583383185335933e-07, "loss": 0.2007, "step": 8652 }, { "epoch": 2.302554550292709, "grad_norm": 0.29862523078918457, "learning_rate": 1.1581698595554285e-07, "loss": 0.2028, "step": 8653 }, { "epoch": 2.3028206492815326, "grad_norm": 0.27394723892211914, "learning_rate": 1.1580013959733499e-07, "loss": 0.1795, "step": 8654 }, { "epoch": 2.3030867482703568, "grad_norm": 0.3512401580810547, "learning_rate": 1.1578329277922609e-07, "loss": 0.2164, "step": 8655 }, { "epoch": 2.3033528472591804, "grad_norm": 0.2749541401863098, "learning_rate": 1.1576644550170651e-07, "loss": 0.1909, "step": 8656 }, { "epoch": 2.303618946248004, "grad_norm": 0.26819872856140137, "learning_rate": 1.1574959776526665e-07, "loss": 0.1631, "step": 8657 }, { "epoch": 2.3038850452368282, "grad_norm": 0.2921152412891388, "learning_rate": 1.1573274957039687e-07, "loss": 0.1838, "step": 8658 }, { "epoch": 2.304151144225652, "grad_norm": 0.2866242825984955, "learning_rate": 1.1571590091758761e-07, "loss": 0.2022, "step": 8659 }, { "epoch": 2.3044172432144756, "grad_norm": 0.2808285653591156, "learning_rate": 1.1569905180732927e-07, "loss": 0.1912, "step": 8660 }, { "epoch": 2.3046833422032997, "grad_norm": 0.2812801003456116, "learning_rate": 1.1568220224011232e-07, "loss": 0.193, "step": 8661 }, { "epoch": 2.3049494411921234, "grad_norm": 0.2993382513523102, "learning_rate": 1.1566535221642717e-07, "loss": 0.1905, "step": 8662 }, { "epoch": 2.3052155401809475, "grad_norm": 0.3023096024990082, "learning_rate": 1.1564850173676429e-07, "loss": 0.1959, "step": 8663 }, { "epoch": 2.3054816391697712, "grad_norm": 0.31723305583000183, "learning_rate": 1.1563165080161414e-07, "loss": 0.2121, "step": 8664 }, { "epoch": 2.305747738158595, "grad_norm": 0.29326945543289185, "learning_rate": 1.1561479941146724e-07, "loss": 0.1823, "step": 8665 }, { "epoch": 2.306013837147419, "grad_norm": 0.2832202613353729, "learning_rate": 1.1559794756681409e-07, "loss": 0.1898, "step": 8666 }, { "epoch": 2.3062799361362427, "grad_norm": 0.2816363275051117, "learning_rate": 1.1558109526814515e-07, "loss": 0.186, "step": 8667 }, { "epoch": 2.3065460351250664, "grad_norm": 0.28935450315475464, "learning_rate": 1.15564242515951e-07, "loss": 0.1789, "step": 8668 }, { "epoch": 2.3068121341138905, "grad_norm": 0.2988150715827942, "learning_rate": 1.1554738931072214e-07, "loss": 0.1764, "step": 8669 }, { "epoch": 2.307078233102714, "grad_norm": 0.29670441150665283, "learning_rate": 1.1553053565294919e-07, "loss": 0.1938, "step": 8670 }, { "epoch": 2.307344332091538, "grad_norm": 0.2782287299633026, "learning_rate": 1.1551368154312264e-07, "loss": 0.1944, "step": 8671 }, { "epoch": 2.307610431080362, "grad_norm": 0.2976895272731781, "learning_rate": 1.1549682698173311e-07, "loss": 0.2152, "step": 8672 }, { "epoch": 2.3078765300691857, "grad_norm": 0.28927117586135864, "learning_rate": 1.1547997196927117e-07, "loss": 0.1868, "step": 8673 }, { "epoch": 2.3081426290580094, "grad_norm": 0.3091030418872833, "learning_rate": 1.1546311650622747e-07, "loss": 0.2121, "step": 8674 }, { "epoch": 2.3084087280468335, "grad_norm": 0.3048909902572632, "learning_rate": 1.1544626059309259e-07, "loss": 0.2043, "step": 8675 }, { "epoch": 2.308674827035657, "grad_norm": 0.2743563652038574, "learning_rate": 1.1542940423035714e-07, "loss": 0.1999, "step": 8676 }, { "epoch": 2.308940926024481, "grad_norm": 0.28196603059768677, "learning_rate": 1.1541254741851183e-07, "loss": 0.1927, "step": 8677 }, { "epoch": 2.309207025013305, "grad_norm": 0.3317275643348694, "learning_rate": 1.1539569015804724e-07, "loss": 0.2326, "step": 8678 }, { "epoch": 2.3094731240021287, "grad_norm": 0.2793770432472229, "learning_rate": 1.1537883244945415e-07, "loss": 0.1981, "step": 8679 }, { "epoch": 2.309739222990953, "grad_norm": 0.2911434471607208, "learning_rate": 1.1536197429322312e-07, "loss": 0.1909, "step": 8680 }, { "epoch": 2.3100053219797765, "grad_norm": 0.27706214785575867, "learning_rate": 1.1534511568984493e-07, "loss": 0.1739, "step": 8681 }, { "epoch": 2.3102714209686, "grad_norm": 0.2790438234806061, "learning_rate": 1.1532825663981026e-07, "loss": 0.2064, "step": 8682 }, { "epoch": 2.3105375199574243, "grad_norm": 0.2758624851703644, "learning_rate": 1.1531139714360985e-07, "loss": 0.1979, "step": 8683 }, { "epoch": 2.310803618946248, "grad_norm": 0.30068594217300415, "learning_rate": 1.1529453720173439e-07, "loss": 0.2007, "step": 8684 }, { "epoch": 2.3110697179350717, "grad_norm": 0.5453477501869202, "learning_rate": 1.1527767681467471e-07, "loss": 0.1998, "step": 8685 }, { "epoch": 2.311335816923896, "grad_norm": 0.2839440703392029, "learning_rate": 1.152608159829215e-07, "loss": 0.1761, "step": 8686 }, { "epoch": 2.3116019159127195, "grad_norm": 0.2822588086128235, "learning_rate": 1.1524395470696556e-07, "loss": 0.1996, "step": 8687 }, { "epoch": 2.3118680149015436, "grad_norm": 0.2720421850681305, "learning_rate": 1.1522709298729769e-07, "loss": 0.1737, "step": 8688 }, { "epoch": 2.3121341138903673, "grad_norm": 0.3055179715156555, "learning_rate": 1.1521023082440866e-07, "loss": 0.205, "step": 8689 }, { "epoch": 2.312400212879191, "grad_norm": 0.27228739857673645, "learning_rate": 1.1519336821878931e-07, "loss": 0.1734, "step": 8690 }, { "epoch": 2.312666311868015, "grad_norm": 0.2717420160770416, "learning_rate": 1.1517650517093047e-07, "loss": 0.1863, "step": 8691 }, { "epoch": 2.3129324108568388, "grad_norm": 0.2988523840904236, "learning_rate": 1.1515964168132297e-07, "loss": 0.2044, "step": 8692 }, { "epoch": 2.3131985098456624, "grad_norm": 0.28928112983703613, "learning_rate": 1.1514277775045766e-07, "loss": 0.1834, "step": 8693 }, { "epoch": 2.3134646088344866, "grad_norm": 0.2896319627761841, "learning_rate": 1.151259133788254e-07, "loss": 0.1907, "step": 8694 }, { "epoch": 2.3137307078233103, "grad_norm": 0.32050856947898865, "learning_rate": 1.151090485669171e-07, "loss": 0.1991, "step": 8695 }, { "epoch": 2.313996806812134, "grad_norm": 0.3013485372066498, "learning_rate": 1.1509218331522361e-07, "loss": 0.2063, "step": 8696 }, { "epoch": 2.314262905800958, "grad_norm": 0.28921666741371155, "learning_rate": 1.1507531762423586e-07, "loss": 0.1998, "step": 8697 }, { "epoch": 2.3145290047897817, "grad_norm": 0.3900254964828491, "learning_rate": 1.1505845149444473e-07, "loss": 0.2391, "step": 8698 }, { "epoch": 2.3147951037786054, "grad_norm": 0.28918495774269104, "learning_rate": 1.1504158492634119e-07, "loss": 0.193, "step": 8699 }, { "epoch": 2.3150612027674295, "grad_norm": 0.2952665388584137, "learning_rate": 1.1502471792041617e-07, "loss": 0.1832, "step": 8700 }, { "epoch": 2.3153273017562532, "grad_norm": 0.3086968660354614, "learning_rate": 1.1500785047716064e-07, "loss": 0.1975, "step": 8701 }, { "epoch": 2.3155934007450774, "grad_norm": 0.2984521687030792, "learning_rate": 1.1499098259706552e-07, "loss": 0.175, "step": 8702 }, { "epoch": 2.315859499733901, "grad_norm": 0.25386813282966614, "learning_rate": 1.1497411428062183e-07, "loss": 0.1717, "step": 8703 }, { "epoch": 2.3161255987227247, "grad_norm": 0.28336405754089355, "learning_rate": 1.1495724552832055e-07, "loss": 0.1953, "step": 8704 }, { "epoch": 2.316391697711549, "grad_norm": 0.303072988986969, "learning_rate": 1.149403763406527e-07, "loss": 0.2113, "step": 8705 }, { "epoch": 2.3166577967003725, "grad_norm": 0.2782662808895111, "learning_rate": 1.149235067181093e-07, "loss": 0.2061, "step": 8706 }, { "epoch": 2.316923895689196, "grad_norm": 0.28393417596817017, "learning_rate": 1.1490663666118135e-07, "loss": 0.1919, "step": 8707 }, { "epoch": 2.3171899946780203, "grad_norm": 0.48566722869873047, "learning_rate": 1.148897661703599e-07, "loss": 0.2138, "step": 8708 }, { "epoch": 2.317456093666844, "grad_norm": 0.321981281042099, "learning_rate": 1.1487289524613601e-07, "loss": 0.2154, "step": 8709 }, { "epoch": 2.317722192655668, "grad_norm": 0.27895388007164, "learning_rate": 1.1485602388900079e-07, "loss": 0.1877, "step": 8710 }, { "epoch": 2.317988291644492, "grad_norm": 0.28928831219673157, "learning_rate": 1.1483915209944528e-07, "loss": 0.207, "step": 8711 }, { "epoch": 2.3182543906333155, "grad_norm": 0.3154509365558624, "learning_rate": 1.1482227987796056e-07, "loss": 0.1947, "step": 8712 }, { "epoch": 2.3185204896221396, "grad_norm": 0.302689790725708, "learning_rate": 1.1480540722503778e-07, "loss": 0.1911, "step": 8713 }, { "epoch": 2.3187865886109633, "grad_norm": 0.27077409625053406, "learning_rate": 1.1478853414116802e-07, "loss": 0.1875, "step": 8714 }, { "epoch": 2.319052687599787, "grad_norm": 0.3047930598258972, "learning_rate": 1.1477166062684245e-07, "loss": 0.2025, "step": 8715 }, { "epoch": 2.319318786588611, "grad_norm": 0.30148205161094666, "learning_rate": 1.147547866825522e-07, "loss": 0.2034, "step": 8716 }, { "epoch": 2.319584885577435, "grad_norm": 0.26581746339797974, "learning_rate": 1.1473791230878839e-07, "loss": 0.1794, "step": 8717 }, { "epoch": 2.3198509845662585, "grad_norm": 0.27486881613731384, "learning_rate": 1.1472103750604225e-07, "loss": 0.1903, "step": 8718 }, { "epoch": 2.3201170835550826, "grad_norm": 0.26004624366760254, "learning_rate": 1.1470416227480492e-07, "loss": 0.1769, "step": 8719 }, { "epoch": 2.3203831825439063, "grad_norm": 0.3015524446964264, "learning_rate": 1.146872866155676e-07, "loss": 0.1966, "step": 8720 }, { "epoch": 2.32064928153273, "grad_norm": 0.2873896658420563, "learning_rate": 1.1467041052882149e-07, "loss": 0.1979, "step": 8721 }, { "epoch": 2.320915380521554, "grad_norm": 0.2631438374519348, "learning_rate": 1.1465353401505782e-07, "loss": 0.1771, "step": 8722 }, { "epoch": 2.321181479510378, "grad_norm": 0.27101191878318787, "learning_rate": 1.1463665707476785e-07, "loss": 0.1827, "step": 8723 }, { "epoch": 2.3214475784992015, "grad_norm": 0.3077120780944824, "learning_rate": 1.1461977970844276e-07, "loss": 0.2024, "step": 8724 }, { "epoch": 2.3217136774880256, "grad_norm": 0.28035253286361694, "learning_rate": 1.1460290191657387e-07, "loss": 0.1901, "step": 8725 }, { "epoch": 2.3219797764768493, "grad_norm": 0.3077261447906494, "learning_rate": 1.1458602369965243e-07, "loss": 0.1989, "step": 8726 }, { "epoch": 2.3222458754656734, "grad_norm": 0.3088200092315674, "learning_rate": 1.145691450581697e-07, "loss": 0.2234, "step": 8727 }, { "epoch": 2.322511974454497, "grad_norm": 0.2856018841266632, "learning_rate": 1.1455226599261699e-07, "loss": 0.2007, "step": 8728 }, { "epoch": 2.3227780734433208, "grad_norm": 0.2688208818435669, "learning_rate": 1.1453538650348558e-07, "loss": 0.1902, "step": 8729 }, { "epoch": 2.323044172432145, "grad_norm": 0.3026716113090515, "learning_rate": 1.1451850659126683e-07, "loss": 0.2016, "step": 8730 }, { "epoch": 2.3233102714209686, "grad_norm": 0.29670968651771545, "learning_rate": 1.1450162625645203e-07, "loss": 0.1915, "step": 8731 }, { "epoch": 2.3235763704097923, "grad_norm": 0.2767029106616974, "learning_rate": 1.1448474549953256e-07, "loss": 0.1898, "step": 8732 }, { "epoch": 2.3238424693986164, "grad_norm": 0.28166624903678894, "learning_rate": 1.1446786432099976e-07, "loss": 0.185, "step": 8733 }, { "epoch": 2.32410856838744, "grad_norm": 0.2778620421886444, "learning_rate": 1.14450982721345e-07, "loss": 0.1797, "step": 8734 }, { "epoch": 2.324374667376264, "grad_norm": 0.29563483595848083, "learning_rate": 1.1443410070105961e-07, "loss": 0.1823, "step": 8735 }, { "epoch": 2.324640766365088, "grad_norm": 0.28938814997673035, "learning_rate": 1.1441721826063509e-07, "loss": 0.1959, "step": 8736 }, { "epoch": 2.3249068653539116, "grad_norm": 0.28198570013046265, "learning_rate": 1.1440033540056276e-07, "loss": 0.2069, "step": 8737 }, { "epoch": 2.3251729643427357, "grad_norm": 0.2781689167022705, "learning_rate": 1.1438345212133404e-07, "loss": 0.1845, "step": 8738 }, { "epoch": 2.3254390633315594, "grad_norm": 0.2837231755256653, "learning_rate": 1.1436656842344036e-07, "loss": 0.1971, "step": 8739 }, { "epoch": 2.325705162320383, "grad_norm": 0.30321571230888367, "learning_rate": 1.143496843073732e-07, "loss": 0.2224, "step": 8740 }, { "epoch": 2.325971261309207, "grad_norm": 0.2888127267360687, "learning_rate": 1.1433279977362398e-07, "loss": 0.1996, "step": 8741 }, { "epoch": 2.326237360298031, "grad_norm": 0.2900117039680481, "learning_rate": 1.1431591482268419e-07, "loss": 0.1937, "step": 8742 }, { "epoch": 2.3265034592868545, "grad_norm": 0.31253671646118164, "learning_rate": 1.1429902945504529e-07, "loss": 0.2065, "step": 8743 }, { "epoch": 2.3267695582756787, "grad_norm": 0.2995840311050415, "learning_rate": 1.1428214367119875e-07, "loss": 0.1807, "step": 8744 }, { "epoch": 2.3270356572645023, "grad_norm": 0.29666590690612793, "learning_rate": 1.1426525747163611e-07, "loss": 0.2064, "step": 8745 }, { "epoch": 2.327301756253326, "grad_norm": 0.2981971800327301, "learning_rate": 1.1424837085684884e-07, "loss": 0.206, "step": 8746 }, { "epoch": 2.32756785524215, "grad_norm": 0.365426629781723, "learning_rate": 1.1423148382732852e-07, "loss": 0.1944, "step": 8747 }, { "epoch": 2.327833954230974, "grad_norm": 0.2950236201286316, "learning_rate": 1.1421459638356665e-07, "loss": 0.2243, "step": 8748 }, { "epoch": 2.3281000532197975, "grad_norm": 0.3074079155921936, "learning_rate": 1.1419770852605478e-07, "loss": 0.208, "step": 8749 }, { "epoch": 2.3283661522086216, "grad_norm": 0.30641379952430725, "learning_rate": 1.1418082025528446e-07, "loss": 0.2123, "step": 8750 }, { "epoch": 2.3286322511974453, "grad_norm": 0.29092395305633545, "learning_rate": 1.1416393157174731e-07, "loss": 0.1872, "step": 8751 }, { "epoch": 2.3288983501862695, "grad_norm": 0.29643064737319946, "learning_rate": 1.1414704247593487e-07, "loss": 0.1848, "step": 8752 }, { "epoch": 2.329164449175093, "grad_norm": 0.28309187293052673, "learning_rate": 1.1413015296833877e-07, "loss": 0.1819, "step": 8753 }, { "epoch": 2.329430548163917, "grad_norm": 0.2736297845840454, "learning_rate": 1.1411326304945059e-07, "loss": 0.1732, "step": 8754 }, { "epoch": 2.329696647152741, "grad_norm": 0.2703704833984375, "learning_rate": 1.14096372719762e-07, "loss": 0.1964, "step": 8755 }, { "epoch": 2.3299627461415646, "grad_norm": 0.272452175617218, "learning_rate": 1.1407948197976456e-07, "loss": 0.1842, "step": 8756 }, { "epoch": 2.3302288451303887, "grad_norm": 0.29769232869148254, "learning_rate": 1.1406259082995e-07, "loss": 0.2033, "step": 8757 }, { "epoch": 2.3304949441192124, "grad_norm": 0.3372887372970581, "learning_rate": 1.1404569927080989e-07, "loss": 0.2105, "step": 8758 }, { "epoch": 2.330761043108036, "grad_norm": 0.2965072691440582, "learning_rate": 1.1402880730283597e-07, "loss": 0.2007, "step": 8759 }, { "epoch": 2.3310271420968602, "grad_norm": 0.2899589538574219, "learning_rate": 1.1401191492651989e-07, "loss": 0.1933, "step": 8760 }, { "epoch": 2.331293241085684, "grad_norm": 0.269206166267395, "learning_rate": 1.1399502214235335e-07, "loss": 0.1745, "step": 8761 }, { "epoch": 2.3315593400745076, "grad_norm": 0.30674588680267334, "learning_rate": 1.1397812895082804e-07, "loss": 0.1938, "step": 8762 }, { "epoch": 2.3318254390633317, "grad_norm": 0.314999520778656, "learning_rate": 1.139612353524357e-07, "loss": 0.2145, "step": 8763 }, { "epoch": 2.3320915380521554, "grad_norm": 0.3177576959133148, "learning_rate": 1.1394434134766803e-07, "loss": 0.217, "step": 8764 }, { "epoch": 2.332357637040979, "grad_norm": 0.26257869601249695, "learning_rate": 1.1392744693701682e-07, "loss": 0.1743, "step": 8765 }, { "epoch": 2.332623736029803, "grad_norm": 0.30235496163368225, "learning_rate": 1.1391055212097378e-07, "loss": 0.2203, "step": 8766 }, { "epoch": 2.332889835018627, "grad_norm": 0.2821119427680969, "learning_rate": 1.138936569000307e-07, "loss": 0.193, "step": 8767 }, { "epoch": 2.3331559340074506, "grad_norm": 0.2884232997894287, "learning_rate": 1.1387676127467931e-07, "loss": 0.1906, "step": 8768 }, { "epoch": 2.3334220329962747, "grad_norm": 0.2775042951107025, "learning_rate": 1.1385986524541147e-07, "loss": 0.1761, "step": 8769 }, { "epoch": 2.3336881319850984, "grad_norm": 0.27318453788757324, "learning_rate": 1.1384296881271891e-07, "loss": 0.1841, "step": 8770 }, { "epoch": 2.333954230973922, "grad_norm": 0.29232165217399597, "learning_rate": 1.1382607197709346e-07, "loss": 0.1903, "step": 8771 }, { "epoch": 2.334220329962746, "grad_norm": 0.30296212434768677, "learning_rate": 1.1380917473902697e-07, "loss": 0.1918, "step": 8772 }, { "epoch": 2.33448642895157, "grad_norm": 0.27705204486846924, "learning_rate": 1.1379227709901127e-07, "loss": 0.1871, "step": 8773 }, { "epoch": 2.334752527940394, "grad_norm": 0.2981630861759186, "learning_rate": 1.1377537905753818e-07, "loss": 0.2026, "step": 8774 }, { "epoch": 2.3350186269292177, "grad_norm": 0.3044903874397278, "learning_rate": 1.1375848061509959e-07, "loss": 0.215, "step": 8775 }, { "epoch": 2.3352847259180414, "grad_norm": 0.2777198255062103, "learning_rate": 1.1374158177218736e-07, "loss": 0.1959, "step": 8776 }, { "epoch": 2.3355508249068655, "grad_norm": 0.3330056965351105, "learning_rate": 1.1372468252929333e-07, "loss": 0.2155, "step": 8777 }, { "epoch": 2.335816923895689, "grad_norm": 0.30170297622680664, "learning_rate": 1.1370778288690945e-07, "loss": 0.2051, "step": 8778 }, { "epoch": 2.336083022884513, "grad_norm": 0.27829858660697937, "learning_rate": 1.1369088284552759e-07, "loss": 0.1969, "step": 8779 }, { "epoch": 2.336349121873337, "grad_norm": 0.28222420811653137, "learning_rate": 1.1367398240563967e-07, "loss": 0.2012, "step": 8780 }, { "epoch": 2.3366152208621607, "grad_norm": 0.31180045008659363, "learning_rate": 1.1365708156773764e-07, "loss": 0.208, "step": 8781 }, { "epoch": 2.336881319850985, "grad_norm": 0.28062495589256287, "learning_rate": 1.1364018033231341e-07, "loss": 0.1997, "step": 8782 }, { "epoch": 2.3371474188398085, "grad_norm": 0.3002614676952362, "learning_rate": 1.1362327869985899e-07, "loss": 0.2083, "step": 8783 }, { "epoch": 2.337413517828632, "grad_norm": 0.29218798875808716, "learning_rate": 1.1360637667086626e-07, "loss": 0.1752, "step": 8784 }, { "epoch": 2.3376796168174563, "grad_norm": 0.2653525471687317, "learning_rate": 1.1358947424582724e-07, "loss": 0.1819, "step": 8785 }, { "epoch": 2.33794571580628, "grad_norm": 0.28625601530075073, "learning_rate": 1.1357257142523394e-07, "loss": 0.1909, "step": 8786 }, { "epoch": 2.3382118147951036, "grad_norm": 0.29370126128196716, "learning_rate": 1.1355566820957827e-07, "loss": 0.193, "step": 8787 }, { "epoch": 2.3384779137839278, "grad_norm": 0.2948078513145447, "learning_rate": 1.1353876459935235e-07, "loss": 0.2055, "step": 8788 }, { "epoch": 2.3387440127727515, "grad_norm": 0.3067842423915863, "learning_rate": 1.1352186059504811e-07, "loss": 0.1872, "step": 8789 }, { "epoch": 2.339010111761575, "grad_norm": 0.2890761196613312, "learning_rate": 1.1350495619715763e-07, "loss": 0.1945, "step": 8790 }, { "epoch": 2.3392762107503993, "grad_norm": 0.2936643362045288, "learning_rate": 1.1348805140617292e-07, "loss": 0.1905, "step": 8791 }, { "epoch": 2.339542309739223, "grad_norm": 0.3149295747280121, "learning_rate": 1.1347114622258611e-07, "loss": 0.1982, "step": 8792 }, { "epoch": 2.3398084087280466, "grad_norm": 0.29733559489250183, "learning_rate": 1.1345424064688915e-07, "loss": 0.1822, "step": 8793 }, { "epoch": 2.3400745077168708, "grad_norm": 0.29259511828422546, "learning_rate": 1.1343733467957421e-07, "loss": 0.1988, "step": 8794 }, { "epoch": 2.3403406067056944, "grad_norm": 0.2803628742694855, "learning_rate": 1.1342042832113335e-07, "loss": 0.1831, "step": 8795 }, { "epoch": 2.340606705694518, "grad_norm": 0.30949917435646057, "learning_rate": 1.1340352157205865e-07, "loss": 0.2167, "step": 8796 }, { "epoch": 2.3408728046833422, "grad_norm": 0.29053565859794617, "learning_rate": 1.1338661443284228e-07, "loss": 0.1904, "step": 8797 }, { "epoch": 2.341138903672166, "grad_norm": 0.2956649661064148, "learning_rate": 1.1336970690397627e-07, "loss": 0.2054, "step": 8798 }, { "epoch": 2.34140500266099, "grad_norm": 0.2962166965007782, "learning_rate": 1.1335279898595283e-07, "loss": 0.2015, "step": 8799 }, { "epoch": 2.3416711016498137, "grad_norm": 0.2984072268009186, "learning_rate": 1.1333589067926406e-07, "loss": 0.18, "step": 8800 }, { "epoch": 2.3419372006386374, "grad_norm": 0.2844351828098297, "learning_rate": 1.1331898198440218e-07, "loss": 0.1895, "step": 8801 }, { "epoch": 2.3422032996274615, "grad_norm": 0.2865920662879944, "learning_rate": 1.1330207290185927e-07, "loss": 0.1943, "step": 8802 }, { "epoch": 2.3424693986162852, "grad_norm": 0.30546239018440247, "learning_rate": 1.1328516343212758e-07, "loss": 0.2146, "step": 8803 }, { "epoch": 2.342735497605109, "grad_norm": 0.2875330150127411, "learning_rate": 1.1326825357569927e-07, "loss": 0.2021, "step": 8804 }, { "epoch": 2.343001596593933, "grad_norm": 0.28230980038642883, "learning_rate": 1.1325134333306656e-07, "loss": 0.1906, "step": 8805 }, { "epoch": 2.3432676955827567, "grad_norm": 0.37318453192710876, "learning_rate": 1.1323443270472166e-07, "loss": 0.1991, "step": 8806 }, { "epoch": 2.343533794571581, "grad_norm": 0.28426340222358704, "learning_rate": 1.132175216911568e-07, "loss": 0.1809, "step": 8807 }, { "epoch": 2.3437998935604045, "grad_norm": 0.2956603467464447, "learning_rate": 1.1320061029286417e-07, "loss": 0.1882, "step": 8808 }, { "epoch": 2.344065992549228, "grad_norm": 0.27289214730262756, "learning_rate": 1.1318369851033602e-07, "loss": 0.1722, "step": 8809 }, { "epoch": 2.3443320915380523, "grad_norm": 0.281353235244751, "learning_rate": 1.1316678634406469e-07, "loss": 0.1768, "step": 8810 }, { "epoch": 2.344598190526876, "grad_norm": 0.2665773034095764, "learning_rate": 1.1314987379454239e-07, "loss": 0.1802, "step": 8811 }, { "epoch": 2.3448642895156997, "grad_norm": 0.2954171895980835, "learning_rate": 1.1313296086226138e-07, "loss": 0.205, "step": 8812 }, { "epoch": 2.345130388504524, "grad_norm": 0.3119159936904907, "learning_rate": 1.13116047547714e-07, "loss": 0.227, "step": 8813 }, { "epoch": 2.3453964874933475, "grad_norm": 0.29060235619544983, "learning_rate": 1.1309913385139254e-07, "loss": 0.1875, "step": 8814 }, { "epoch": 2.345662586482171, "grad_norm": 0.3117450773715973, "learning_rate": 1.1308221977378929e-07, "loss": 0.1974, "step": 8815 }, { "epoch": 2.3459286854709953, "grad_norm": 0.286467969417572, "learning_rate": 1.1306530531539658e-07, "loss": 0.2001, "step": 8816 }, { "epoch": 2.346194784459819, "grad_norm": 0.2879638075828552, "learning_rate": 1.1304839047670677e-07, "loss": 0.2153, "step": 8817 }, { "epoch": 2.3464608834486427, "grad_norm": 0.4266204535961151, "learning_rate": 1.1303147525821219e-07, "loss": 0.1873, "step": 8818 }, { "epoch": 2.346726982437467, "grad_norm": 0.30143609642982483, "learning_rate": 1.1301455966040522e-07, "loss": 0.2144, "step": 8819 }, { "epoch": 2.3469930814262905, "grad_norm": 0.3218693733215332, "learning_rate": 1.129976436837782e-07, "loss": 0.2266, "step": 8820 }, { "epoch": 2.3472591804151146, "grad_norm": 0.2826073169708252, "learning_rate": 1.1298072732882351e-07, "loss": 0.1787, "step": 8821 }, { "epoch": 2.3475252794039383, "grad_norm": 0.28689807653427124, "learning_rate": 1.1296381059603354e-07, "loss": 0.1831, "step": 8822 }, { "epoch": 2.347791378392762, "grad_norm": 0.28256702423095703, "learning_rate": 1.1294689348590074e-07, "loss": 0.183, "step": 8823 }, { "epoch": 2.348057477381586, "grad_norm": 0.2732437551021576, "learning_rate": 1.1292997599891745e-07, "loss": 0.1788, "step": 8824 }, { "epoch": 2.34832357637041, "grad_norm": 0.26842615008354187, "learning_rate": 1.1291305813557615e-07, "loss": 0.1854, "step": 8825 }, { "epoch": 2.3485896753592335, "grad_norm": 0.27527958154678345, "learning_rate": 1.1289613989636924e-07, "loss": 0.1819, "step": 8826 }, { "epoch": 2.3488557743480576, "grad_norm": 0.2960624098777771, "learning_rate": 1.1287922128178922e-07, "loss": 0.1814, "step": 8827 }, { "epoch": 2.3491218733368813, "grad_norm": 0.2988038659095764, "learning_rate": 1.1286230229232849e-07, "loss": 0.1871, "step": 8828 }, { "epoch": 2.3493879723257054, "grad_norm": 0.27443283796310425, "learning_rate": 1.1284538292847952e-07, "loss": 0.1862, "step": 8829 }, { "epoch": 2.349654071314529, "grad_norm": 0.2860839366912842, "learning_rate": 1.1282846319073481e-07, "loss": 0.2002, "step": 8830 }, { "epoch": 2.3499201703033528, "grad_norm": 0.2978864908218384, "learning_rate": 1.1281154307958685e-07, "loss": 0.2018, "step": 8831 }, { "epoch": 2.350186269292177, "grad_norm": 0.2869648337364197, "learning_rate": 1.1279462259552814e-07, "loss": 0.1946, "step": 8832 }, { "epoch": 2.3504523682810006, "grad_norm": 0.2809300422668457, "learning_rate": 1.127777017390512e-07, "loss": 0.179, "step": 8833 }, { "epoch": 2.3507184672698243, "grad_norm": 0.28817516565322876, "learning_rate": 1.127607805106485e-07, "loss": 0.1955, "step": 8834 }, { "epoch": 2.3509845662586484, "grad_norm": 0.3266841471195221, "learning_rate": 1.1274385891081264e-07, "loss": 0.2012, "step": 8835 }, { "epoch": 2.351250665247472, "grad_norm": 0.30589863657951355, "learning_rate": 1.1272693694003613e-07, "loss": 0.1958, "step": 8836 }, { "epoch": 2.3515167642362957, "grad_norm": 0.2990671694278717, "learning_rate": 1.1271001459881154e-07, "loss": 0.208, "step": 8837 }, { "epoch": 2.35178286322512, "grad_norm": 0.27853530645370483, "learning_rate": 1.1269309188763144e-07, "loss": 0.1884, "step": 8838 }, { "epoch": 2.3520489622139436, "grad_norm": 0.2907341420650482, "learning_rate": 1.1267616880698836e-07, "loss": 0.1889, "step": 8839 }, { "epoch": 2.3523150612027672, "grad_norm": 0.30134716629981995, "learning_rate": 1.1265924535737492e-07, "loss": 0.207, "step": 8840 }, { "epoch": 2.3525811601915914, "grad_norm": 0.3042822480201721, "learning_rate": 1.1264232153928375e-07, "loss": 0.2101, "step": 8841 }, { "epoch": 2.352847259180415, "grad_norm": 0.29148006439208984, "learning_rate": 1.126253973532074e-07, "loss": 0.1859, "step": 8842 }, { "epoch": 2.3531133581692387, "grad_norm": 0.28952401876449585, "learning_rate": 1.1260847279963849e-07, "loss": 0.2091, "step": 8843 }, { "epoch": 2.353379457158063, "grad_norm": 0.2719436585903168, "learning_rate": 1.1259154787906972e-07, "loss": 0.1894, "step": 8844 }, { "epoch": 2.3536455561468865, "grad_norm": 0.2738366723060608, "learning_rate": 1.1257462259199366e-07, "loss": 0.1979, "step": 8845 }, { "epoch": 2.3539116551357107, "grad_norm": 0.27851128578186035, "learning_rate": 1.12557696938903e-07, "loss": 0.1757, "step": 8846 }, { "epoch": 2.3541777541245343, "grad_norm": 0.2870139181613922, "learning_rate": 1.1254077092029038e-07, "loss": 0.2051, "step": 8847 }, { "epoch": 2.354443853113358, "grad_norm": 0.3043462038040161, "learning_rate": 1.125238445366485e-07, "loss": 0.1985, "step": 8848 }, { "epoch": 2.354709952102182, "grad_norm": 0.2812007963657379, "learning_rate": 1.1250691778847e-07, "loss": 0.1895, "step": 8849 }, { "epoch": 2.354976051091006, "grad_norm": 0.29421523213386536, "learning_rate": 1.1248999067624763e-07, "loss": 0.2128, "step": 8850 }, { "epoch": 2.3552421500798295, "grad_norm": 0.33770594000816345, "learning_rate": 1.12473063200474e-07, "loss": 0.2018, "step": 8851 }, { "epoch": 2.3555082490686536, "grad_norm": 0.3165925443172455, "learning_rate": 1.1245613536164193e-07, "loss": 0.2098, "step": 8852 }, { "epoch": 2.3557743480574773, "grad_norm": 0.2919086813926697, "learning_rate": 1.1243920716024409e-07, "loss": 0.1919, "step": 8853 }, { "epoch": 2.3560404470463014, "grad_norm": 0.2798537313938141, "learning_rate": 1.1242227859677323e-07, "loss": 0.188, "step": 8854 }, { "epoch": 2.356306546035125, "grad_norm": 0.3353654444217682, "learning_rate": 1.1240534967172208e-07, "loss": 0.21, "step": 8855 }, { "epoch": 2.356572645023949, "grad_norm": 0.2760600447654724, "learning_rate": 1.1238842038558343e-07, "loss": 0.1904, "step": 8856 }, { "epoch": 2.356838744012773, "grad_norm": 0.29251959919929504, "learning_rate": 1.1237149073885002e-07, "loss": 0.2079, "step": 8857 }, { "epoch": 2.3571048430015966, "grad_norm": 0.3279587924480438, "learning_rate": 1.1235456073201466e-07, "loss": 0.2094, "step": 8858 }, { "epoch": 2.3573709419904203, "grad_norm": 0.3007766306400299, "learning_rate": 1.1233763036557011e-07, "loss": 0.2086, "step": 8859 }, { "epoch": 2.3576370409792444, "grad_norm": 0.2946563959121704, "learning_rate": 1.1232069964000915e-07, "loss": 0.1796, "step": 8860 }, { "epoch": 2.357903139968068, "grad_norm": 0.7145461440086365, "learning_rate": 1.1230376855582462e-07, "loss": 0.2031, "step": 8861 }, { "epoch": 2.358169238956892, "grad_norm": 0.30071181058883667, "learning_rate": 1.1228683711350932e-07, "loss": 0.194, "step": 8862 }, { "epoch": 2.358435337945716, "grad_norm": 0.281819611787796, "learning_rate": 1.1226990531355611e-07, "loss": 0.198, "step": 8863 }, { "epoch": 2.3587014369345396, "grad_norm": 0.2924704849720001, "learning_rate": 1.1225297315645783e-07, "loss": 0.1862, "step": 8864 }, { "epoch": 2.3589675359233633, "grad_norm": 0.2822009027004242, "learning_rate": 1.122360406427073e-07, "loss": 0.1969, "step": 8865 }, { "epoch": 2.3592336349121874, "grad_norm": 0.26912662386894226, "learning_rate": 1.1221910777279739e-07, "loss": 0.1961, "step": 8866 }, { "epoch": 2.359499733901011, "grad_norm": 0.2952110171318054, "learning_rate": 1.12202174547221e-07, "loss": 0.1943, "step": 8867 }, { "epoch": 2.3597658328898348, "grad_norm": 0.2674272954463959, "learning_rate": 1.1218524096647098e-07, "loss": 0.178, "step": 8868 }, { "epoch": 2.360031931878659, "grad_norm": 0.2859562337398529, "learning_rate": 1.1216830703104025e-07, "loss": 0.1832, "step": 8869 }, { "epoch": 2.3602980308674826, "grad_norm": 0.27523839473724365, "learning_rate": 1.1215137274142169e-07, "loss": 0.1903, "step": 8870 }, { "epoch": 2.3605641298563067, "grad_norm": 0.29723432660102844, "learning_rate": 1.1213443809810819e-07, "loss": 0.1805, "step": 8871 }, { "epoch": 2.3608302288451304, "grad_norm": 0.26378291845321655, "learning_rate": 1.1211750310159272e-07, "loss": 0.19, "step": 8872 }, { "epoch": 2.361096327833954, "grad_norm": 0.30959492921829224, "learning_rate": 1.1210056775236822e-07, "loss": 0.2124, "step": 8873 }, { "epoch": 2.361362426822778, "grad_norm": 0.2793654799461365, "learning_rate": 1.1208363205092759e-07, "loss": 0.1785, "step": 8874 }, { "epoch": 2.361628525811602, "grad_norm": 0.30234578251838684, "learning_rate": 1.120666959977638e-07, "loss": 0.1814, "step": 8875 }, { "epoch": 2.361894624800426, "grad_norm": 0.27337712049484253, "learning_rate": 1.1204975959336982e-07, "loss": 0.1966, "step": 8876 }, { "epoch": 2.3621607237892497, "grad_norm": 0.2801068425178528, "learning_rate": 1.1203282283823865e-07, "loss": 0.1923, "step": 8877 }, { "epoch": 2.3624268227780734, "grad_norm": 0.3029746115207672, "learning_rate": 1.1201588573286325e-07, "loss": 0.1964, "step": 8878 }, { "epoch": 2.3626929217668975, "grad_norm": 0.29962846636772156, "learning_rate": 1.1199894827773659e-07, "loss": 0.1872, "step": 8879 }, { "epoch": 2.362959020755721, "grad_norm": 0.3181588351726532, "learning_rate": 1.1198201047335172e-07, "loss": 0.2175, "step": 8880 }, { "epoch": 2.363225119744545, "grad_norm": 0.2847076654434204, "learning_rate": 1.1196507232020162e-07, "loss": 0.1903, "step": 8881 }, { "epoch": 2.363491218733369, "grad_norm": 0.27815309166908264, "learning_rate": 1.1194813381877937e-07, "loss": 0.1971, "step": 8882 }, { "epoch": 2.3637573177221927, "grad_norm": 0.2975272834300995, "learning_rate": 1.1193119496957793e-07, "loss": 0.1908, "step": 8883 }, { "epoch": 2.3640234167110163, "grad_norm": 0.3122478425502777, "learning_rate": 1.1191425577309038e-07, "loss": 0.1997, "step": 8884 }, { "epoch": 2.3642895156998405, "grad_norm": 0.30029842257499695, "learning_rate": 1.1189731622980977e-07, "loss": 0.2039, "step": 8885 }, { "epoch": 2.364555614688664, "grad_norm": 0.29055681824684143, "learning_rate": 1.118803763402292e-07, "loss": 0.1876, "step": 8886 }, { "epoch": 2.364821713677488, "grad_norm": 0.308069109916687, "learning_rate": 1.1186343610484172e-07, "loss": 0.1918, "step": 8887 }, { "epoch": 2.365087812666312, "grad_norm": 0.2740219533443451, "learning_rate": 1.1184649552414043e-07, "loss": 0.1915, "step": 8888 }, { "epoch": 2.3653539116551356, "grad_norm": 0.30026066303253174, "learning_rate": 1.118295545986184e-07, "loss": 0.2017, "step": 8889 }, { "epoch": 2.3656200106439593, "grad_norm": 0.31071361899375916, "learning_rate": 1.1181261332876874e-07, "loss": 0.2016, "step": 8890 }, { "epoch": 2.3658861096327835, "grad_norm": 0.2648687958717346, "learning_rate": 1.1179567171508462e-07, "loss": 0.1791, "step": 8891 }, { "epoch": 2.366152208621607, "grad_norm": 0.3074723184108734, "learning_rate": 1.1177872975805909e-07, "loss": 0.2055, "step": 8892 }, { "epoch": 2.3664183076104313, "grad_norm": 0.592007577419281, "learning_rate": 1.117617874581853e-07, "loss": 0.2011, "step": 8893 }, { "epoch": 2.366684406599255, "grad_norm": 0.2943512201309204, "learning_rate": 1.1174484481595644e-07, "loss": 0.1921, "step": 8894 }, { "epoch": 2.3669505055880786, "grad_norm": 0.3103181719779968, "learning_rate": 1.1172790183186562e-07, "loss": 0.2122, "step": 8895 }, { "epoch": 2.3672166045769027, "grad_norm": 0.292741596698761, "learning_rate": 1.1171095850640608e-07, "loss": 0.1936, "step": 8896 }, { "epoch": 2.3674827035657264, "grad_norm": 0.2888084948062897, "learning_rate": 1.1169401484007091e-07, "loss": 0.1981, "step": 8897 }, { "epoch": 2.36774880255455, "grad_norm": 0.27613410353660583, "learning_rate": 1.1167707083335334e-07, "loss": 0.1831, "step": 8898 }, { "epoch": 2.3680149015433742, "grad_norm": 0.27224302291870117, "learning_rate": 1.1166012648674655e-07, "loss": 0.1809, "step": 8899 }, { "epoch": 2.368281000532198, "grad_norm": 0.3113270699977875, "learning_rate": 1.1164318180074377e-07, "loss": 0.1942, "step": 8900 }, { "epoch": 2.368547099521022, "grad_norm": 0.28222066164016724, "learning_rate": 1.1162623677583818e-07, "loss": 0.1852, "step": 8901 }, { "epoch": 2.3688131985098457, "grad_norm": 0.2807347774505615, "learning_rate": 1.1160929141252302e-07, "loss": 0.1817, "step": 8902 }, { "epoch": 2.3690792974986694, "grad_norm": 0.29179447889328003, "learning_rate": 1.1159234571129154e-07, "loss": 0.1838, "step": 8903 }, { "epoch": 2.3693453964874935, "grad_norm": 0.2905271053314209, "learning_rate": 1.1157539967263696e-07, "loss": 0.1845, "step": 8904 }, { "epoch": 2.369611495476317, "grad_norm": 0.27604806423187256, "learning_rate": 1.115584532970526e-07, "loss": 0.1975, "step": 8905 }, { "epoch": 2.369877594465141, "grad_norm": 0.28413528203964233, "learning_rate": 1.1154150658503163e-07, "loss": 0.1979, "step": 8906 }, { "epoch": 2.370143693453965, "grad_norm": 0.32353007793426514, "learning_rate": 1.1152455953706737e-07, "loss": 0.2147, "step": 8907 }, { "epoch": 2.3704097924427887, "grad_norm": 0.27221161127090454, "learning_rate": 1.1150761215365314e-07, "loss": 0.1737, "step": 8908 }, { "epoch": 2.3706758914316124, "grad_norm": 0.30667009949684143, "learning_rate": 1.1149066443528217e-07, "loss": 0.2094, "step": 8909 }, { "epoch": 2.3709419904204365, "grad_norm": 0.29975438117980957, "learning_rate": 1.114737163824478e-07, "loss": 0.1936, "step": 8910 }, { "epoch": 2.37120808940926, "grad_norm": 0.3143932521343231, "learning_rate": 1.1145676799564335e-07, "loss": 0.198, "step": 8911 }, { "epoch": 2.371474188398084, "grad_norm": 0.29664120078086853, "learning_rate": 1.114398192753621e-07, "loss": 0.1775, "step": 8912 }, { "epoch": 2.371740287386908, "grad_norm": 0.2866249978542328, "learning_rate": 1.1142287022209745e-07, "loss": 0.1887, "step": 8913 }, { "epoch": 2.3720063863757317, "grad_norm": 0.27750810980796814, "learning_rate": 1.1140592083634268e-07, "loss": 0.1888, "step": 8914 }, { "epoch": 2.3722724853645554, "grad_norm": 0.2806650698184967, "learning_rate": 1.1138897111859118e-07, "loss": 0.1965, "step": 8915 }, { "epoch": 2.3725385843533795, "grad_norm": 0.29510584473609924, "learning_rate": 1.113720210693363e-07, "loss": 0.1895, "step": 8916 }, { "epoch": 2.372804683342203, "grad_norm": 0.2923104166984558, "learning_rate": 1.113550706890714e-07, "loss": 0.2111, "step": 8917 }, { "epoch": 2.3730707823310273, "grad_norm": 0.3212043344974518, "learning_rate": 1.113381199782899e-07, "loss": 0.2027, "step": 8918 }, { "epoch": 2.373336881319851, "grad_norm": 0.30434954166412354, "learning_rate": 1.1132116893748518e-07, "loss": 0.2083, "step": 8919 }, { "epoch": 2.3736029803086747, "grad_norm": 0.2864331007003784, "learning_rate": 1.113042175671506e-07, "loss": 0.1976, "step": 8920 }, { "epoch": 2.373869079297499, "grad_norm": 0.27502140402793884, "learning_rate": 1.112872658677796e-07, "loss": 0.1945, "step": 8921 }, { "epoch": 2.3741351782863225, "grad_norm": 0.29175329208374023, "learning_rate": 1.1127031383986558e-07, "loss": 0.1939, "step": 8922 }, { "epoch": 2.374401277275146, "grad_norm": 0.2899867594242096, "learning_rate": 1.1125336148390203e-07, "loss": 0.1872, "step": 8923 }, { "epoch": 2.3746673762639703, "grad_norm": 0.33928272128105164, "learning_rate": 1.1123640880038232e-07, "loss": 0.2164, "step": 8924 }, { "epoch": 2.374933475252794, "grad_norm": 0.2622031271457672, "learning_rate": 1.1121945578979992e-07, "loss": 0.1799, "step": 8925 }, { "epoch": 2.375199574241618, "grad_norm": 0.30329176783561707, "learning_rate": 1.112025024526483e-07, "loss": 0.1933, "step": 8926 }, { "epoch": 2.3754656732304418, "grad_norm": 0.324835866689682, "learning_rate": 1.1118554878942092e-07, "loss": 0.2073, "step": 8927 }, { "epoch": 2.3757317722192655, "grad_norm": 0.29138436913490295, "learning_rate": 1.1116859480061127e-07, "loss": 0.1897, "step": 8928 }, { "epoch": 2.3759978712080896, "grad_norm": 0.2804398834705353, "learning_rate": 1.1115164048671281e-07, "loss": 0.2005, "step": 8929 }, { "epoch": 2.3762639701969133, "grad_norm": 0.2859620749950409, "learning_rate": 1.1113468584821906e-07, "loss": 0.2014, "step": 8930 }, { "epoch": 2.376530069185737, "grad_norm": 0.3082054853439331, "learning_rate": 1.111177308856235e-07, "loss": 0.209, "step": 8931 }, { "epoch": 2.376796168174561, "grad_norm": 0.2997838854789734, "learning_rate": 1.111007755994197e-07, "loss": 0.1896, "step": 8932 }, { "epoch": 2.3770622671633848, "grad_norm": 0.3032371997833252, "learning_rate": 1.110838199901011e-07, "loss": 0.2105, "step": 8933 }, { "epoch": 2.3773283661522084, "grad_norm": 0.2832227349281311, "learning_rate": 1.110668640581613e-07, "loss": 0.1938, "step": 8934 }, { "epoch": 2.3775944651410326, "grad_norm": 0.2894972562789917, "learning_rate": 1.110499078040938e-07, "loss": 0.191, "step": 8935 }, { "epoch": 2.3778605641298562, "grad_norm": 0.303027480840683, "learning_rate": 1.110329512283922e-07, "loss": 0.197, "step": 8936 }, { "epoch": 2.37812666311868, "grad_norm": 0.2674036920070648, "learning_rate": 1.1101599433155003e-07, "loss": 0.1868, "step": 8937 }, { "epoch": 2.378392762107504, "grad_norm": 0.27799609303474426, "learning_rate": 1.1099903711406086e-07, "loss": 0.1792, "step": 8938 }, { "epoch": 2.3786588610963277, "grad_norm": 0.2801012694835663, "learning_rate": 1.109820795764183e-07, "loss": 0.1926, "step": 8939 }, { "epoch": 2.378924960085152, "grad_norm": 0.3086681067943573, "learning_rate": 1.109651217191159e-07, "loss": 0.1921, "step": 8940 }, { "epoch": 2.3791910590739755, "grad_norm": 0.30562201142311096, "learning_rate": 1.109481635426473e-07, "loss": 0.1877, "step": 8941 }, { "epoch": 2.3794571580627992, "grad_norm": 0.288135826587677, "learning_rate": 1.1093120504750606e-07, "loss": 0.1933, "step": 8942 }, { "epoch": 2.3797232570516234, "grad_norm": 0.31021827459335327, "learning_rate": 1.1091424623418582e-07, "loss": 0.1996, "step": 8943 }, { "epoch": 2.379989356040447, "grad_norm": 0.2606833577156067, "learning_rate": 1.108972871031802e-07, "loss": 0.1797, "step": 8944 }, { "epoch": 2.3802554550292707, "grad_norm": 0.3070383667945862, "learning_rate": 1.108803276549829e-07, "loss": 0.1888, "step": 8945 }, { "epoch": 2.380521554018095, "grad_norm": 0.28000566363334656, "learning_rate": 1.1086336789008747e-07, "loss": 0.1922, "step": 8946 }, { "epoch": 2.3807876530069185, "grad_norm": 0.29390063881874084, "learning_rate": 1.108464078089876e-07, "loss": 0.1942, "step": 8947 }, { "epoch": 2.3810537519957427, "grad_norm": 0.2855486571788788, "learning_rate": 1.1082944741217697e-07, "loss": 0.1921, "step": 8948 }, { "epoch": 2.3813198509845663, "grad_norm": 0.2805042564868927, "learning_rate": 1.1081248670014928e-07, "loss": 0.1989, "step": 8949 }, { "epoch": 2.38158594997339, "grad_norm": 0.28791356086730957, "learning_rate": 1.1079552567339816e-07, "loss": 0.1866, "step": 8950 }, { "epoch": 2.381852048962214, "grad_norm": 0.28343650698661804, "learning_rate": 1.107785643324173e-07, "loss": 0.1885, "step": 8951 }, { "epoch": 2.382118147951038, "grad_norm": 0.2905469536781311, "learning_rate": 1.1076160267770043e-07, "loss": 0.2039, "step": 8952 }, { "epoch": 2.3823842469398615, "grad_norm": 0.2835339307785034, "learning_rate": 1.1074464070974124e-07, "loss": 0.1917, "step": 8953 }, { "epoch": 2.3826503459286856, "grad_norm": 0.3089369535446167, "learning_rate": 1.1072767842903346e-07, "loss": 0.2041, "step": 8954 }, { "epoch": 2.3829164449175093, "grad_norm": 0.3317997455596924, "learning_rate": 1.1071071583607081e-07, "loss": 0.2305, "step": 8955 }, { "epoch": 2.383182543906333, "grad_norm": 0.30772173404693604, "learning_rate": 1.1069375293134704e-07, "loss": 0.1799, "step": 8956 }, { "epoch": 2.383448642895157, "grad_norm": 0.3095025420188904, "learning_rate": 1.1067678971535587e-07, "loss": 0.197, "step": 8957 }, { "epoch": 2.383714741883981, "grad_norm": 0.30206143856048584, "learning_rate": 1.1065982618859108e-07, "loss": 0.2032, "step": 8958 }, { "epoch": 2.3839808408728045, "grad_norm": 0.28977227210998535, "learning_rate": 1.1064286235154645e-07, "loss": 0.1798, "step": 8959 }, { "epoch": 2.3842469398616286, "grad_norm": 0.28886377811431885, "learning_rate": 1.1062589820471575e-07, "loss": 0.2042, "step": 8960 }, { "epoch": 2.3845130388504523, "grad_norm": 0.3194535970687866, "learning_rate": 1.106089337485927e-07, "loss": 0.2127, "step": 8961 }, { "epoch": 2.384779137839276, "grad_norm": 0.2926015257835388, "learning_rate": 1.1059196898367113e-07, "loss": 0.2084, "step": 8962 }, { "epoch": 2.3850452368281, "grad_norm": 0.27923691272735596, "learning_rate": 1.1057500391044487e-07, "loss": 0.2037, "step": 8963 }, { "epoch": 2.385311335816924, "grad_norm": 0.2872101366519928, "learning_rate": 1.1055803852940771e-07, "loss": 0.1925, "step": 8964 }, { "epoch": 2.385577434805748, "grad_norm": 0.30412837862968445, "learning_rate": 1.1054107284105343e-07, "loss": 0.1958, "step": 8965 }, { "epoch": 2.3858435337945716, "grad_norm": 0.31064724922180176, "learning_rate": 1.1052410684587589e-07, "loss": 0.2007, "step": 8966 }, { "epoch": 2.3861096327833953, "grad_norm": 0.29400524497032166, "learning_rate": 1.1050714054436894e-07, "loss": 0.2149, "step": 8967 }, { "epoch": 2.3863757317722194, "grad_norm": 0.28191936016082764, "learning_rate": 1.1049017393702642e-07, "loss": 0.1906, "step": 8968 }, { "epoch": 2.386641830761043, "grad_norm": 0.2891988754272461, "learning_rate": 1.1047320702434217e-07, "loss": 0.1826, "step": 8969 }, { "epoch": 2.3869079297498668, "grad_norm": 0.2698441743850708, "learning_rate": 1.1045623980681007e-07, "loss": 0.1857, "step": 8970 }, { "epoch": 2.387174028738691, "grad_norm": 0.2845393121242523, "learning_rate": 1.1043927228492393e-07, "loss": 0.1942, "step": 8971 }, { "epoch": 2.3874401277275146, "grad_norm": 0.29258424043655396, "learning_rate": 1.1042230445917775e-07, "loss": 0.192, "step": 8972 }, { "epoch": 2.3877062267163387, "grad_norm": 0.28202372789382935, "learning_rate": 1.104053363300653e-07, "loss": 0.1926, "step": 8973 }, { "epoch": 2.3879723257051624, "grad_norm": 0.309653103351593, "learning_rate": 1.1038836789808053e-07, "loss": 0.1973, "step": 8974 }, { "epoch": 2.388238424693986, "grad_norm": 0.29624396562576294, "learning_rate": 1.1037139916371734e-07, "loss": 0.2048, "step": 8975 }, { "epoch": 2.38850452368281, "grad_norm": 0.285939484834671, "learning_rate": 1.1035443012746964e-07, "loss": 0.1883, "step": 8976 }, { "epoch": 2.388770622671634, "grad_norm": 0.3109596073627472, "learning_rate": 1.1033746078983139e-07, "loss": 0.1908, "step": 8977 }, { "epoch": 2.3890367216604576, "grad_norm": 0.2983931303024292, "learning_rate": 1.1032049115129649e-07, "loss": 0.192, "step": 8978 }, { "epoch": 2.3893028206492817, "grad_norm": 0.28529974818229675, "learning_rate": 1.1030352121235889e-07, "loss": 0.1985, "step": 8979 }, { "epoch": 2.3895689196381054, "grad_norm": 0.28302955627441406, "learning_rate": 1.1028655097351256e-07, "loss": 0.2111, "step": 8980 }, { "epoch": 2.389835018626929, "grad_norm": 0.3202829360961914, "learning_rate": 1.1026958043525143e-07, "loss": 0.2126, "step": 8981 }, { "epoch": 2.390101117615753, "grad_norm": 0.28900521993637085, "learning_rate": 1.1025260959806946e-07, "loss": 0.2148, "step": 8982 }, { "epoch": 2.390367216604577, "grad_norm": 0.2721639573574066, "learning_rate": 1.1023563846246065e-07, "loss": 0.2003, "step": 8983 }, { "epoch": 2.3906333155934005, "grad_norm": 0.2913408875465393, "learning_rate": 1.1021866702891896e-07, "loss": 0.2169, "step": 8984 }, { "epoch": 2.3908994145822247, "grad_norm": 0.28957828879356384, "learning_rate": 1.1020169529793844e-07, "loss": 0.1884, "step": 8985 }, { "epoch": 2.3911655135710483, "grad_norm": 0.2738487720489502, "learning_rate": 1.1018472327001305e-07, "loss": 0.1904, "step": 8986 }, { "epoch": 2.3914316125598725, "grad_norm": 0.2771773338317871, "learning_rate": 1.101677509456368e-07, "loss": 0.2006, "step": 8987 }, { "epoch": 2.391697711548696, "grad_norm": 0.29612383246421814, "learning_rate": 1.1015077832530372e-07, "loss": 0.1879, "step": 8988 }, { "epoch": 2.39196381053752, "grad_norm": 0.32871198654174805, "learning_rate": 1.1013380540950785e-07, "loss": 0.1971, "step": 8989 }, { "epoch": 2.392229909526344, "grad_norm": 0.3149890601634979, "learning_rate": 1.1011683219874321e-07, "loss": 0.2195, "step": 8990 }, { "epoch": 2.3924960085151676, "grad_norm": 0.2859630882740021, "learning_rate": 1.1009985869350389e-07, "loss": 0.1885, "step": 8991 }, { "epoch": 2.3927621075039913, "grad_norm": 0.28773871064186096, "learning_rate": 1.1008288489428386e-07, "loss": 0.1935, "step": 8992 }, { "epoch": 2.3930282064928154, "grad_norm": 0.3055417835712433, "learning_rate": 1.1006591080157723e-07, "loss": 0.205, "step": 8993 }, { "epoch": 2.393294305481639, "grad_norm": 0.2801804542541504, "learning_rate": 1.1004893641587809e-07, "loss": 0.1882, "step": 8994 }, { "epoch": 2.3935604044704633, "grad_norm": 0.29821470379829407, "learning_rate": 1.100319617376805e-07, "loss": 0.1844, "step": 8995 }, { "epoch": 2.393826503459287, "grad_norm": 0.28384679555892944, "learning_rate": 1.1001498676747855e-07, "loss": 0.2105, "step": 8996 }, { "epoch": 2.3940926024481106, "grad_norm": 0.28753939270973206, "learning_rate": 1.0999801150576633e-07, "loss": 0.1971, "step": 8997 }, { "epoch": 2.3943587014369347, "grad_norm": 0.29156363010406494, "learning_rate": 1.0998103595303796e-07, "loss": 0.1841, "step": 8998 }, { "epoch": 2.3946248004257584, "grad_norm": 0.3065052926540375, "learning_rate": 1.0996406010978756e-07, "loss": 0.2155, "step": 8999 }, { "epoch": 2.394890899414582, "grad_norm": 0.27766430377960205, "learning_rate": 1.0994708397650922e-07, "loss": 0.179, "step": 9000 }, { "epoch": 2.3951569984034062, "grad_norm": 0.28003478050231934, "learning_rate": 1.0993010755369714e-07, "loss": 0.196, "step": 9001 }, { "epoch": 2.39542309739223, "grad_norm": 0.3105943500995636, "learning_rate": 1.0991313084184537e-07, "loss": 0.2025, "step": 9002 }, { "epoch": 2.3956891963810536, "grad_norm": 0.288010835647583, "learning_rate": 1.0989615384144811e-07, "loss": 0.1977, "step": 9003 }, { "epoch": 2.3959552953698777, "grad_norm": 0.30920881032943726, "learning_rate": 1.0987917655299952e-07, "loss": 0.2074, "step": 9004 }, { "epoch": 2.3962213943587014, "grad_norm": 0.3095327913761139, "learning_rate": 1.0986219897699374e-07, "loss": 0.2134, "step": 9005 }, { "epoch": 2.396487493347525, "grad_norm": 0.276978462934494, "learning_rate": 1.0984522111392494e-07, "loss": 0.1901, "step": 9006 }, { "epoch": 2.396753592336349, "grad_norm": 0.2880104184150696, "learning_rate": 1.0982824296428732e-07, "loss": 0.1859, "step": 9007 }, { "epoch": 2.397019691325173, "grad_norm": 0.28792551159858704, "learning_rate": 1.098112645285751e-07, "loss": 0.1843, "step": 9008 }, { "epoch": 2.3972857903139966, "grad_norm": 0.30170947313308716, "learning_rate": 1.097942858072824e-07, "loss": 0.2051, "step": 9009 }, { "epoch": 2.3975518893028207, "grad_norm": 0.2945517897605896, "learning_rate": 1.0977730680090351e-07, "loss": 0.1912, "step": 9010 }, { "epoch": 2.3978179882916444, "grad_norm": 0.30662766098976135, "learning_rate": 1.0976032750993258e-07, "loss": 0.2086, "step": 9011 }, { "epoch": 2.3980840872804685, "grad_norm": 0.28491222858428955, "learning_rate": 1.0974334793486386e-07, "loss": 0.1828, "step": 9012 }, { "epoch": 2.398350186269292, "grad_norm": 0.30029356479644775, "learning_rate": 1.097263680761916e-07, "loss": 0.2103, "step": 9013 }, { "epoch": 2.398616285258116, "grad_norm": 0.32570746541023254, "learning_rate": 1.0970938793441e-07, "loss": 0.2029, "step": 9014 }, { "epoch": 2.39888238424694, "grad_norm": 0.2887745797634125, "learning_rate": 1.0969240751001332e-07, "loss": 0.1996, "step": 9015 }, { "epoch": 2.3991484832357637, "grad_norm": 0.29650038480758667, "learning_rate": 1.0967542680349582e-07, "loss": 0.2064, "step": 9016 }, { "epoch": 2.3994145822245874, "grad_norm": 0.29699474573135376, "learning_rate": 1.0965844581535177e-07, "loss": 0.2055, "step": 9017 }, { "epoch": 2.3996806812134115, "grad_norm": 0.2980118989944458, "learning_rate": 1.0964146454607548e-07, "loss": 0.1969, "step": 9018 }, { "epoch": 2.399946780202235, "grad_norm": 0.3101623058319092, "learning_rate": 1.0962448299616116e-07, "loss": 0.2115, "step": 9019 }, { "epoch": 2.4002128791910593, "grad_norm": 0.2878580391407013, "learning_rate": 1.0960750116610316e-07, "loss": 0.1983, "step": 9020 }, { "epoch": 2.400478978179883, "grad_norm": 0.29676610231399536, "learning_rate": 1.0959051905639569e-07, "loss": 0.1964, "step": 9021 }, { "epoch": 2.4007450771687067, "grad_norm": 0.32578518986701965, "learning_rate": 1.0957353666753318e-07, "loss": 0.2086, "step": 9022 }, { "epoch": 2.401011176157531, "grad_norm": 0.30572617053985596, "learning_rate": 1.0955655400000982e-07, "loss": 0.1997, "step": 9023 }, { "epoch": 2.4012772751463545, "grad_norm": 0.28277164697647095, "learning_rate": 1.0953957105431999e-07, "loss": 0.2031, "step": 9024 }, { "epoch": 2.401543374135178, "grad_norm": 0.2942188084125519, "learning_rate": 1.0952258783095803e-07, "loss": 0.1896, "step": 9025 }, { "epoch": 2.4018094731240023, "grad_norm": 0.2801625728607178, "learning_rate": 1.0950560433041824e-07, "loss": 0.1888, "step": 9026 }, { "epoch": 2.402075572112826, "grad_norm": 0.2903992831707001, "learning_rate": 1.0948862055319503e-07, "loss": 0.1935, "step": 9027 }, { "epoch": 2.4023416711016496, "grad_norm": 0.3069938123226166, "learning_rate": 1.094716364997827e-07, "loss": 0.202, "step": 9028 }, { "epoch": 2.4026077700904738, "grad_norm": 0.29324889183044434, "learning_rate": 1.0945465217067562e-07, "loss": 0.1951, "step": 9029 }, { "epoch": 2.4028738690792975, "grad_norm": 0.305949866771698, "learning_rate": 1.0943766756636817e-07, "loss": 0.1958, "step": 9030 }, { "epoch": 2.403139968068121, "grad_norm": 0.32170751690864563, "learning_rate": 1.0942068268735471e-07, "loss": 0.2069, "step": 9031 }, { "epoch": 2.4034060670569453, "grad_norm": 0.2738732695579529, "learning_rate": 1.0940369753412966e-07, "loss": 0.182, "step": 9032 }, { "epoch": 2.403672166045769, "grad_norm": 0.316173255443573, "learning_rate": 1.0938671210718737e-07, "loss": 0.2029, "step": 9033 }, { "epoch": 2.4039382650345926, "grad_norm": 0.26776206493377686, "learning_rate": 1.0936972640702225e-07, "loss": 0.173, "step": 9034 }, { "epoch": 2.4042043640234168, "grad_norm": 0.31713995337486267, "learning_rate": 1.0935274043412875e-07, "loss": 0.2112, "step": 9035 }, { "epoch": 2.4044704630122404, "grad_norm": 0.29185575246810913, "learning_rate": 1.0933575418900126e-07, "loss": 0.2026, "step": 9036 }, { "epoch": 2.4047365620010646, "grad_norm": 0.29105445742607117, "learning_rate": 1.0931876767213417e-07, "loss": 0.2141, "step": 9037 }, { "epoch": 2.4050026609898882, "grad_norm": 0.28758516907691956, "learning_rate": 1.0930178088402198e-07, "loss": 0.1928, "step": 9038 }, { "epoch": 2.405268759978712, "grad_norm": 0.2843043804168701, "learning_rate": 1.0928479382515907e-07, "loss": 0.1957, "step": 9039 }, { "epoch": 2.405534858967536, "grad_norm": 0.289447546005249, "learning_rate": 1.0926780649603997e-07, "loss": 0.1863, "step": 9040 }, { "epoch": 2.4058009579563597, "grad_norm": 0.2830589711666107, "learning_rate": 1.0925081889715908e-07, "loss": 0.184, "step": 9041 }, { "epoch": 2.406067056945184, "grad_norm": 0.296242356300354, "learning_rate": 1.0923383102901085e-07, "loss": 0.2021, "step": 9042 }, { "epoch": 2.4063331559340075, "grad_norm": 0.28629788756370544, "learning_rate": 1.0921684289208977e-07, "loss": 0.2038, "step": 9043 }, { "epoch": 2.406599254922831, "grad_norm": 0.2527351975440979, "learning_rate": 1.0919985448689031e-07, "loss": 0.1675, "step": 9044 }, { "epoch": 2.4068653539116553, "grad_norm": 0.29364991188049316, "learning_rate": 1.0918286581390702e-07, "loss": 0.2055, "step": 9045 }, { "epoch": 2.407131452900479, "grad_norm": 0.27974894642829895, "learning_rate": 1.091658768736343e-07, "loss": 0.1919, "step": 9046 }, { "epoch": 2.4073975518893027, "grad_norm": 0.2838371992111206, "learning_rate": 1.0914888766656671e-07, "loss": 0.1985, "step": 9047 }, { "epoch": 2.407663650878127, "grad_norm": 0.28997695446014404, "learning_rate": 1.0913189819319877e-07, "loss": 0.1937, "step": 9048 }, { "epoch": 2.4079297498669505, "grad_norm": 0.27981674671173096, "learning_rate": 1.09114908454025e-07, "loss": 0.1901, "step": 9049 }, { "epoch": 2.408195848855774, "grad_norm": 0.3055601418018341, "learning_rate": 1.090979184495399e-07, "loss": 0.2095, "step": 9050 }, { "epoch": 2.4084619478445983, "grad_norm": 0.2996322214603424, "learning_rate": 1.0908092818023802e-07, "loss": 0.2058, "step": 9051 }, { "epoch": 2.408728046833422, "grad_norm": 0.3166055977344513, "learning_rate": 1.090639376466139e-07, "loss": 0.1865, "step": 9052 }, { "epoch": 2.4089941458222457, "grad_norm": 0.2882384657859802, "learning_rate": 1.0904694684916207e-07, "loss": 0.1973, "step": 9053 }, { "epoch": 2.40926024481107, "grad_norm": 0.2972894012928009, "learning_rate": 1.0902995578837712e-07, "loss": 0.2018, "step": 9054 }, { "epoch": 2.4095263437998935, "grad_norm": 0.27225261926651, "learning_rate": 1.0901296446475361e-07, "loss": 0.2016, "step": 9055 }, { "epoch": 2.409792442788717, "grad_norm": 0.3072970509529114, "learning_rate": 1.0899597287878608e-07, "loss": 0.1994, "step": 9056 }, { "epoch": 2.4100585417775413, "grad_norm": 0.29237887263298035, "learning_rate": 1.0897898103096916e-07, "loss": 0.1979, "step": 9057 }, { "epoch": 2.410324640766365, "grad_norm": 0.2769617736339569, "learning_rate": 1.0896198892179742e-07, "loss": 0.1959, "step": 9058 }, { "epoch": 2.410590739755189, "grad_norm": 0.2959739863872528, "learning_rate": 1.0894499655176543e-07, "loss": 0.1854, "step": 9059 }, { "epoch": 2.410856838744013, "grad_norm": 0.26450595259666443, "learning_rate": 1.0892800392136782e-07, "loss": 0.1871, "step": 9060 }, { "epoch": 2.4111229377328365, "grad_norm": 0.2855786681175232, "learning_rate": 1.0891101103109923e-07, "loss": 0.1879, "step": 9061 }, { "epoch": 2.4113890367216606, "grad_norm": 0.308162659406662, "learning_rate": 1.0889401788145423e-07, "loss": 0.2067, "step": 9062 }, { "epoch": 2.4116551357104843, "grad_norm": 0.29718881845474243, "learning_rate": 1.0887702447292746e-07, "loss": 0.1938, "step": 9063 }, { "epoch": 2.411921234699308, "grad_norm": 0.29436036944389343, "learning_rate": 1.0886003080601354e-07, "loss": 0.1877, "step": 9064 }, { "epoch": 2.412187333688132, "grad_norm": 0.2961132526397705, "learning_rate": 1.0884303688120713e-07, "loss": 0.2049, "step": 9065 }, { "epoch": 2.4124534326769558, "grad_norm": 0.3018084466457367, "learning_rate": 1.0882604269900288e-07, "loss": 0.1835, "step": 9066 }, { "epoch": 2.41271953166578, "grad_norm": 0.3072161376476288, "learning_rate": 1.0880904825989547e-07, "loss": 0.1899, "step": 9067 }, { "epoch": 2.4129856306546036, "grad_norm": 0.3081940710544586, "learning_rate": 1.087920535643795e-07, "loss": 0.1913, "step": 9068 }, { "epoch": 2.4132517296434273, "grad_norm": 0.2768070101737976, "learning_rate": 1.087750586129497e-07, "loss": 0.2004, "step": 9069 }, { "epoch": 2.4135178286322514, "grad_norm": 0.27357473969459534, "learning_rate": 1.0875806340610073e-07, "loss": 0.1881, "step": 9070 }, { "epoch": 2.413783927621075, "grad_norm": 0.28340381383895874, "learning_rate": 1.0874106794432727e-07, "loss": 0.1903, "step": 9071 }, { "epoch": 2.4140500266098988, "grad_norm": 0.2951710522174835, "learning_rate": 1.0872407222812404e-07, "loss": 0.1904, "step": 9072 }, { "epoch": 2.414316125598723, "grad_norm": 0.28470566868782043, "learning_rate": 1.087070762579857e-07, "loss": 0.2051, "step": 9073 }, { "epoch": 2.4145822245875466, "grad_norm": 0.2916002571582794, "learning_rate": 1.0869008003440696e-07, "loss": 0.1874, "step": 9074 }, { "epoch": 2.4148483235763702, "grad_norm": 0.29103952646255493, "learning_rate": 1.0867308355788256e-07, "loss": 0.1985, "step": 9075 }, { "epoch": 2.4151144225651944, "grad_norm": 0.312928169965744, "learning_rate": 1.0865608682890723e-07, "loss": 0.2046, "step": 9076 }, { "epoch": 2.415380521554018, "grad_norm": 0.30805328488349915, "learning_rate": 1.086390898479757e-07, "loss": 0.2162, "step": 9077 }, { "epoch": 2.4156466205428417, "grad_norm": 0.28005218505859375, "learning_rate": 1.0862209261558266e-07, "loss": 0.2043, "step": 9078 }, { "epoch": 2.415912719531666, "grad_norm": 0.30713316798210144, "learning_rate": 1.0860509513222289e-07, "loss": 0.1908, "step": 9079 }, { "epoch": 2.4161788185204895, "grad_norm": 0.28823840618133545, "learning_rate": 1.0858809739839116e-07, "loss": 0.1852, "step": 9080 }, { "epoch": 2.4164449175093132, "grad_norm": 0.3191516697406769, "learning_rate": 1.0857109941458225e-07, "loss": 0.1944, "step": 9081 }, { "epoch": 2.4167110164981374, "grad_norm": 0.2723444700241089, "learning_rate": 1.0855410118129087e-07, "loss": 0.1814, "step": 9082 }, { "epoch": 2.416977115486961, "grad_norm": 0.2766929566860199, "learning_rate": 1.085371026990118e-07, "loss": 0.1981, "step": 9083 }, { "epoch": 2.417243214475785, "grad_norm": 0.2876834571361542, "learning_rate": 1.0852010396823983e-07, "loss": 0.1969, "step": 9084 }, { "epoch": 2.417509313464609, "grad_norm": 0.2728942036628723, "learning_rate": 1.0850310498946978e-07, "loss": 0.175, "step": 9085 }, { "epoch": 2.4177754124534325, "grad_norm": 0.29420068860054016, "learning_rate": 1.0848610576319639e-07, "loss": 0.2013, "step": 9086 }, { "epoch": 2.4180415114422567, "grad_norm": 0.28709250688552856, "learning_rate": 1.084691062899145e-07, "loss": 0.1907, "step": 9087 }, { "epoch": 2.4183076104310803, "grad_norm": 0.29807981848716736, "learning_rate": 1.0845210657011892e-07, "loss": 0.1889, "step": 9088 }, { "epoch": 2.418573709419904, "grad_norm": 0.28642621636390686, "learning_rate": 1.0843510660430446e-07, "loss": 0.2075, "step": 9089 }, { "epoch": 2.418839808408728, "grad_norm": 0.43174752593040466, "learning_rate": 1.0841810639296598e-07, "loss": 0.2032, "step": 9090 }, { "epoch": 2.419105907397552, "grad_norm": 0.3072642982006073, "learning_rate": 1.0840110593659825e-07, "loss": 0.1924, "step": 9091 }, { "epoch": 2.419372006386376, "grad_norm": 0.3150838613510132, "learning_rate": 1.0838410523569617e-07, "loss": 0.212, "step": 9092 }, { "epoch": 2.4196381053751996, "grad_norm": 0.3236585259437561, "learning_rate": 1.0836710429075453e-07, "loss": 0.2124, "step": 9093 }, { "epoch": 2.4199042043640233, "grad_norm": 0.2810150980949402, "learning_rate": 1.0835010310226825e-07, "loss": 0.193, "step": 9094 }, { "epoch": 2.4201703033528474, "grad_norm": 0.2923889756202698, "learning_rate": 1.0833310167073208e-07, "loss": 0.2039, "step": 9095 }, { "epoch": 2.420436402341671, "grad_norm": 0.3140977621078491, "learning_rate": 1.0831609999664099e-07, "loss": 0.21, "step": 9096 }, { "epoch": 2.420702501330495, "grad_norm": 0.3042643368244171, "learning_rate": 1.0829909808048983e-07, "loss": 0.2093, "step": 9097 }, { "epoch": 2.420968600319319, "grad_norm": 0.2578776776790619, "learning_rate": 1.0828209592277345e-07, "loss": 0.1723, "step": 9098 }, { "epoch": 2.4212346993081426, "grad_norm": 0.2950414717197418, "learning_rate": 1.0826509352398679e-07, "loss": 0.1993, "step": 9099 }, { "epoch": 2.4215007982969663, "grad_norm": 0.2969196140766144, "learning_rate": 1.0824809088462469e-07, "loss": 0.2034, "step": 9100 }, { "epoch": 2.4217668972857904, "grad_norm": 0.30909326672554016, "learning_rate": 1.0823108800518211e-07, "loss": 0.1894, "step": 9101 }, { "epoch": 2.422032996274614, "grad_norm": 0.31931936740875244, "learning_rate": 1.0821408488615391e-07, "loss": 0.2058, "step": 9102 }, { "epoch": 2.422299095263438, "grad_norm": 0.2906549274921417, "learning_rate": 1.0819708152803506e-07, "loss": 0.1906, "step": 9103 }, { "epoch": 2.422565194252262, "grad_norm": 0.28502675890922546, "learning_rate": 1.081800779313204e-07, "loss": 0.2089, "step": 9104 }, { "epoch": 2.4228312932410856, "grad_norm": 0.31174901127815247, "learning_rate": 1.0816307409650493e-07, "loss": 0.2128, "step": 9105 }, { "epoch": 2.4230973922299097, "grad_norm": 0.30043482780456543, "learning_rate": 1.0814607002408358e-07, "loss": 0.1871, "step": 9106 }, { "epoch": 2.4233634912187334, "grad_norm": 0.27274197340011597, "learning_rate": 1.0812906571455128e-07, "loss": 0.1715, "step": 9107 }, { "epoch": 2.423629590207557, "grad_norm": 0.2947298288345337, "learning_rate": 1.0811206116840297e-07, "loss": 0.2127, "step": 9108 }, { "epoch": 2.423895689196381, "grad_norm": 0.28189903497695923, "learning_rate": 1.0809505638613363e-07, "loss": 0.1745, "step": 9109 }, { "epoch": 2.424161788185205, "grad_norm": 0.2764659523963928, "learning_rate": 1.080780513682382e-07, "loss": 0.1874, "step": 9110 }, { "epoch": 2.4244278871740286, "grad_norm": 0.25950607657432556, "learning_rate": 1.0806104611521171e-07, "loss": 0.1778, "step": 9111 }, { "epoch": 2.4246939861628527, "grad_norm": 0.2954733371734619, "learning_rate": 1.0804404062754907e-07, "loss": 0.1861, "step": 9112 }, { "epoch": 2.4249600851516764, "grad_norm": 0.31319597363471985, "learning_rate": 1.0802703490574532e-07, "loss": 0.1917, "step": 9113 }, { "epoch": 2.4252261841405005, "grad_norm": 0.30374449491500854, "learning_rate": 1.0801002895029539e-07, "loss": 0.21, "step": 9114 }, { "epoch": 2.425492283129324, "grad_norm": 0.30898475646972656, "learning_rate": 1.079930227616943e-07, "loss": 0.2011, "step": 9115 }, { "epoch": 2.425758382118148, "grad_norm": 0.3076159358024597, "learning_rate": 1.0797601634043712e-07, "loss": 0.1842, "step": 9116 }, { "epoch": 2.426024481106972, "grad_norm": 0.2906406819820404, "learning_rate": 1.0795900968701882e-07, "loss": 0.1812, "step": 9117 }, { "epoch": 2.4262905800957957, "grad_norm": 0.2867823839187622, "learning_rate": 1.0794200280193435e-07, "loss": 0.2039, "step": 9118 }, { "epoch": 2.4265566790846194, "grad_norm": 0.31527408957481384, "learning_rate": 1.0792499568567884e-07, "loss": 0.2001, "step": 9119 }, { "epoch": 2.4268227780734435, "grad_norm": 0.2851307988166809, "learning_rate": 1.0790798833874728e-07, "loss": 0.1877, "step": 9120 }, { "epoch": 2.427088877062267, "grad_norm": 0.3179270923137665, "learning_rate": 1.0789098076163472e-07, "loss": 0.2139, "step": 9121 }, { "epoch": 2.427354976051091, "grad_norm": 0.30767083168029785, "learning_rate": 1.0787397295483618e-07, "loss": 0.2092, "step": 9122 }, { "epoch": 2.427621075039915, "grad_norm": 0.31479862332344055, "learning_rate": 1.0785696491884677e-07, "loss": 0.2055, "step": 9123 }, { "epoch": 2.4278871740287387, "grad_norm": 0.2881034314632416, "learning_rate": 1.0783995665416146e-07, "loss": 0.1935, "step": 9124 }, { "epoch": 2.4281532730175623, "grad_norm": 0.34867748618125916, "learning_rate": 1.078229481612754e-07, "loss": 0.2039, "step": 9125 }, { "epoch": 2.4284193720063865, "grad_norm": 0.29629281163215637, "learning_rate": 1.0780593944068363e-07, "loss": 0.1978, "step": 9126 }, { "epoch": 2.42868547099521, "grad_norm": 0.3099342882633209, "learning_rate": 1.0778893049288123e-07, "loss": 0.1777, "step": 9127 }, { "epoch": 2.428951569984034, "grad_norm": 0.28068944811820984, "learning_rate": 1.0777192131836326e-07, "loss": 0.1942, "step": 9128 }, { "epoch": 2.429217668972858, "grad_norm": 0.2985071837902069, "learning_rate": 1.0775491191762484e-07, "loss": 0.2029, "step": 9129 }, { "epoch": 2.4294837679616816, "grad_norm": 0.2887870669364929, "learning_rate": 1.0773790229116113e-07, "loss": 0.1973, "step": 9130 }, { "epoch": 2.4297498669505058, "grad_norm": 0.3794042468070984, "learning_rate": 1.0772089243946712e-07, "loss": 0.2097, "step": 9131 }, { "epoch": 2.4300159659393294, "grad_norm": 0.336310476064682, "learning_rate": 1.07703882363038e-07, "loss": 0.1985, "step": 9132 }, { "epoch": 2.430282064928153, "grad_norm": 0.3085781931877136, "learning_rate": 1.0768687206236886e-07, "loss": 0.1944, "step": 9133 }, { "epoch": 2.4305481639169773, "grad_norm": 0.3094482719898224, "learning_rate": 1.0766986153795483e-07, "loss": 0.199, "step": 9134 }, { "epoch": 2.430814262905801, "grad_norm": 0.29133525490760803, "learning_rate": 1.0765285079029107e-07, "loss": 0.1937, "step": 9135 }, { "epoch": 2.4310803618946246, "grad_norm": 0.3016310930252075, "learning_rate": 1.0763583981987267e-07, "loss": 0.2035, "step": 9136 }, { "epoch": 2.4313464608834487, "grad_norm": 0.31074368953704834, "learning_rate": 1.076188286271948e-07, "loss": 0.1935, "step": 9137 }, { "epoch": 2.4316125598722724, "grad_norm": 0.2724955976009369, "learning_rate": 1.0760181721275258e-07, "loss": 0.1761, "step": 9138 }, { "epoch": 2.4318786588610966, "grad_norm": 0.2713221609592438, "learning_rate": 1.0758480557704124e-07, "loss": 0.1858, "step": 9139 }, { "epoch": 2.4321447578499202, "grad_norm": 0.2941557466983795, "learning_rate": 1.0756779372055591e-07, "loss": 0.2024, "step": 9140 }, { "epoch": 2.432410856838744, "grad_norm": 0.3205421566963196, "learning_rate": 1.0755078164379174e-07, "loss": 0.1934, "step": 9141 }, { "epoch": 2.432676955827568, "grad_norm": 0.34905320405960083, "learning_rate": 1.0753376934724395e-07, "loss": 0.1983, "step": 9142 }, { "epoch": 2.4329430548163917, "grad_norm": 0.29993611574172974, "learning_rate": 1.0751675683140764e-07, "loss": 0.19, "step": 9143 }, { "epoch": 2.4332091538052154, "grad_norm": 0.32719966769218445, "learning_rate": 1.0749974409677811e-07, "loss": 0.2211, "step": 9144 }, { "epoch": 2.4334752527940395, "grad_norm": 0.27968528866767883, "learning_rate": 1.0748273114385047e-07, "loss": 0.192, "step": 9145 }, { "epoch": 2.433741351782863, "grad_norm": 0.2795374393463135, "learning_rate": 1.0746571797311994e-07, "loss": 0.1931, "step": 9146 }, { "epoch": 2.434007450771687, "grad_norm": 0.287919282913208, "learning_rate": 1.0744870458508177e-07, "loss": 0.202, "step": 9147 }, { "epoch": 2.434273549760511, "grad_norm": 0.2897856533527374, "learning_rate": 1.0743169098023112e-07, "loss": 0.1817, "step": 9148 }, { "epoch": 2.4345396487493347, "grad_norm": 0.2638706862926483, "learning_rate": 1.0741467715906329e-07, "loss": 0.1775, "step": 9149 }, { "epoch": 2.4348057477381584, "grad_norm": 0.2999555468559265, "learning_rate": 1.0739766312207344e-07, "loss": 0.1896, "step": 9150 }, { "epoch": 2.4350718467269825, "grad_norm": 0.3213590979576111, "learning_rate": 1.0738064886975682e-07, "loss": 0.2058, "step": 9151 }, { "epoch": 2.435337945715806, "grad_norm": 0.3124833106994629, "learning_rate": 1.0736363440260867e-07, "loss": 0.2112, "step": 9152 }, { "epoch": 2.43560404470463, "grad_norm": 0.2859208583831787, "learning_rate": 1.0734661972112426e-07, "loss": 0.2007, "step": 9153 }, { "epoch": 2.435870143693454, "grad_norm": 0.2929692566394806, "learning_rate": 1.0732960482579883e-07, "loss": 0.1999, "step": 9154 }, { "epoch": 2.4361362426822777, "grad_norm": 0.28158751130104065, "learning_rate": 1.073125897171276e-07, "loss": 0.1988, "step": 9155 }, { "epoch": 2.436402341671102, "grad_norm": 0.3014938235282898, "learning_rate": 1.0729557439560588e-07, "loss": 0.1843, "step": 9156 }, { "epoch": 2.4366684406599255, "grad_norm": 0.31162622570991516, "learning_rate": 1.0727855886172894e-07, "loss": 0.1988, "step": 9157 }, { "epoch": 2.436934539648749, "grad_norm": 0.3053297698497772, "learning_rate": 1.0726154311599205e-07, "loss": 0.2051, "step": 9158 }, { "epoch": 2.4372006386375733, "grad_norm": 0.31345611810684204, "learning_rate": 1.072445271588905e-07, "loss": 0.1973, "step": 9159 }, { "epoch": 2.437466737626397, "grad_norm": 0.2883310616016388, "learning_rate": 1.0722751099091955e-07, "loss": 0.1955, "step": 9160 }, { "epoch": 2.437732836615221, "grad_norm": 0.276481568813324, "learning_rate": 1.0721049461257454e-07, "loss": 0.1855, "step": 9161 }, { "epoch": 2.437998935604045, "grad_norm": 0.28447532653808594, "learning_rate": 1.0719347802435077e-07, "loss": 0.1803, "step": 9162 }, { "epoch": 2.4382650345928685, "grad_norm": 0.283507376909256, "learning_rate": 1.0717646122674353e-07, "loss": 0.1924, "step": 9163 }, { "epoch": 2.4385311335816926, "grad_norm": 0.3318832218647003, "learning_rate": 1.0715944422024812e-07, "loss": 0.1871, "step": 9164 }, { "epoch": 2.4387972325705163, "grad_norm": 0.2599261999130249, "learning_rate": 1.0714242700535987e-07, "loss": 0.1814, "step": 9165 }, { "epoch": 2.43906333155934, "grad_norm": 0.27442049980163574, "learning_rate": 1.0712540958257412e-07, "loss": 0.1796, "step": 9166 }, { "epoch": 2.439329430548164, "grad_norm": 0.3314805328845978, "learning_rate": 1.0710839195238624e-07, "loss": 0.2003, "step": 9167 }, { "epoch": 2.4395955295369878, "grad_norm": 0.32206159830093384, "learning_rate": 1.0709137411529148e-07, "loss": 0.202, "step": 9168 }, { "epoch": 2.4398616285258115, "grad_norm": 0.27076300978660583, "learning_rate": 1.0707435607178523e-07, "loss": 0.1846, "step": 9169 }, { "epoch": 2.4401277275146356, "grad_norm": 0.32224026322364807, "learning_rate": 1.0705733782236283e-07, "loss": 0.2006, "step": 9170 }, { "epoch": 2.4403938265034593, "grad_norm": 0.27404576539993286, "learning_rate": 1.0704031936751971e-07, "loss": 0.177, "step": 9171 }, { "epoch": 2.440659925492283, "grad_norm": 0.29231661558151245, "learning_rate": 1.0702330070775115e-07, "loss": 0.1931, "step": 9172 }, { "epoch": 2.440926024481107, "grad_norm": 0.3053652346134186, "learning_rate": 1.0700628184355254e-07, "loss": 0.1976, "step": 9173 }, { "epoch": 2.4411921234699308, "grad_norm": 0.30134454369544983, "learning_rate": 1.0698926277541924e-07, "loss": 0.2026, "step": 9174 }, { "epoch": 2.4414582224587544, "grad_norm": 0.2942201495170593, "learning_rate": 1.0697224350384664e-07, "loss": 0.2024, "step": 9175 }, { "epoch": 2.4417243214475786, "grad_norm": 0.31046998500823975, "learning_rate": 1.0695522402933017e-07, "loss": 0.2105, "step": 9176 }, { "epoch": 2.4419904204364022, "grad_norm": 0.3029795289039612, "learning_rate": 1.0693820435236516e-07, "loss": 0.1964, "step": 9177 }, { "epoch": 2.4422565194252264, "grad_norm": 0.30111655592918396, "learning_rate": 1.0692118447344704e-07, "loss": 0.2143, "step": 9178 }, { "epoch": 2.44252261841405, "grad_norm": 0.34185075759887695, "learning_rate": 1.0690416439307121e-07, "loss": 0.2175, "step": 9179 }, { "epoch": 2.4427887174028737, "grad_norm": 0.2715390622615814, "learning_rate": 1.0688714411173312e-07, "loss": 0.1714, "step": 9180 }, { "epoch": 2.443054816391698, "grad_norm": 0.3013514578342438, "learning_rate": 1.0687012362992809e-07, "loss": 0.215, "step": 9181 }, { "epoch": 2.4433209153805215, "grad_norm": 0.2991926074028015, "learning_rate": 1.0685310294815162e-07, "loss": 0.1989, "step": 9182 }, { "epoch": 2.4435870143693452, "grad_norm": 0.3017837703227997, "learning_rate": 1.0683608206689914e-07, "loss": 0.2048, "step": 9183 }, { "epoch": 2.4438531133581693, "grad_norm": 0.291727215051651, "learning_rate": 1.0681906098666604e-07, "loss": 0.1921, "step": 9184 }, { "epoch": 2.444119212346993, "grad_norm": 0.26049765944480896, "learning_rate": 1.0680203970794778e-07, "loss": 0.1734, "step": 9185 }, { "epoch": 2.444385311335817, "grad_norm": 0.3047780990600586, "learning_rate": 1.0678501823123982e-07, "loss": 0.2057, "step": 9186 }, { "epoch": 2.444651410324641, "grad_norm": 0.2952757179737091, "learning_rate": 1.0676799655703756e-07, "loss": 0.2019, "step": 9187 }, { "epoch": 2.4449175093134645, "grad_norm": 0.29321032762527466, "learning_rate": 1.0675097468583651e-07, "loss": 0.2065, "step": 9188 }, { "epoch": 2.4451836083022886, "grad_norm": 0.2641034722328186, "learning_rate": 1.0673395261813216e-07, "loss": 0.1665, "step": 9189 }, { "epoch": 2.4454497072911123, "grad_norm": 0.2766808271408081, "learning_rate": 1.0671693035441989e-07, "loss": 0.1948, "step": 9190 }, { "epoch": 2.445715806279936, "grad_norm": 0.29183414578437805, "learning_rate": 1.066999078951952e-07, "loss": 0.1919, "step": 9191 }, { "epoch": 2.44598190526876, "grad_norm": 0.27194178104400635, "learning_rate": 1.0668288524095362e-07, "loss": 0.1733, "step": 9192 }, { "epoch": 2.446248004257584, "grad_norm": 0.286619633436203, "learning_rate": 1.066658623921906e-07, "loss": 0.1897, "step": 9193 }, { "epoch": 2.4465141032464075, "grad_norm": 0.31913596391677856, "learning_rate": 1.0664883934940165e-07, "loss": 0.2209, "step": 9194 }, { "epoch": 2.4467802022352316, "grad_norm": 0.33224865794181824, "learning_rate": 1.0663181611308223e-07, "loss": 0.196, "step": 9195 }, { "epoch": 2.4470463012240553, "grad_norm": 0.2820841372013092, "learning_rate": 1.0661479268372783e-07, "loss": 0.2049, "step": 9196 }, { "epoch": 2.447312400212879, "grad_norm": 0.27033960819244385, "learning_rate": 1.0659776906183402e-07, "loss": 0.188, "step": 9197 }, { "epoch": 2.447578499201703, "grad_norm": 0.30134785175323486, "learning_rate": 1.0658074524789631e-07, "loss": 0.1983, "step": 9198 }, { "epoch": 2.447844598190527, "grad_norm": 0.3023548424243927, "learning_rate": 1.0656372124241016e-07, "loss": 0.1895, "step": 9199 }, { "epoch": 2.4481106971793505, "grad_norm": 0.30274686217308044, "learning_rate": 1.0654669704587112e-07, "loss": 0.2229, "step": 9200 }, { "epoch": 2.4483767961681746, "grad_norm": 0.2579539716243744, "learning_rate": 1.0652967265877473e-07, "loss": 0.1705, "step": 9201 }, { "epoch": 2.4486428951569983, "grad_norm": 0.273698091506958, "learning_rate": 1.0651264808161655e-07, "loss": 0.1758, "step": 9202 }, { "epoch": 2.4489089941458224, "grad_norm": 0.29273495078086853, "learning_rate": 1.0649562331489208e-07, "loss": 0.2109, "step": 9203 }, { "epoch": 2.449175093134646, "grad_norm": 0.2957625985145569, "learning_rate": 1.064785983590969e-07, "loss": 0.2145, "step": 9204 }, { "epoch": 2.44944119212347, "grad_norm": 0.2808658182621002, "learning_rate": 1.0646157321472651e-07, "loss": 0.2007, "step": 9205 }, { "epoch": 2.449707291112294, "grad_norm": 0.2954835891723633, "learning_rate": 1.064445478822765e-07, "loss": 0.2147, "step": 9206 }, { "epoch": 2.4499733901011176, "grad_norm": 0.27962297201156616, "learning_rate": 1.0642752236224246e-07, "loss": 0.1877, "step": 9207 }, { "epoch": 2.4502394890899413, "grad_norm": 0.33042657375335693, "learning_rate": 1.0641049665511992e-07, "loss": 0.1729, "step": 9208 }, { "epoch": 2.4505055880787654, "grad_norm": 0.3099256157875061, "learning_rate": 1.0639347076140446e-07, "loss": 0.1922, "step": 9209 }, { "epoch": 2.450771687067589, "grad_norm": 0.2724290192127228, "learning_rate": 1.0637644468159166e-07, "loss": 0.1956, "step": 9210 }, { "epoch": 2.451037786056413, "grad_norm": 0.29302722215652466, "learning_rate": 1.0635941841617713e-07, "loss": 0.201, "step": 9211 }, { "epoch": 2.451303885045237, "grad_norm": 0.3057674169540405, "learning_rate": 1.0634239196565645e-07, "loss": 0.2127, "step": 9212 }, { "epoch": 2.4515699840340606, "grad_norm": 0.2921469509601593, "learning_rate": 1.0632536533052518e-07, "loss": 0.1994, "step": 9213 }, { "epoch": 2.4518360830228847, "grad_norm": 0.2865097224712372, "learning_rate": 1.0630833851127898e-07, "loss": 0.1827, "step": 9214 }, { "epoch": 2.4521021820117084, "grad_norm": 0.27722257375717163, "learning_rate": 1.0629131150841342e-07, "loss": 0.195, "step": 9215 }, { "epoch": 2.452368281000532, "grad_norm": 0.2964865565299988, "learning_rate": 1.0627428432242411e-07, "loss": 0.2046, "step": 9216 }, { "epoch": 2.452634379989356, "grad_norm": 0.2893384099006653, "learning_rate": 1.0625725695380668e-07, "loss": 0.1827, "step": 9217 }, { "epoch": 2.45290047897818, "grad_norm": 0.2724120020866394, "learning_rate": 1.0624022940305672e-07, "loss": 0.1863, "step": 9218 }, { "epoch": 2.4531665779670035, "grad_norm": 0.27879461646080017, "learning_rate": 1.0622320167066989e-07, "loss": 0.2057, "step": 9219 }, { "epoch": 2.4534326769558277, "grad_norm": 0.28890863060951233, "learning_rate": 1.0620617375714182e-07, "loss": 0.2084, "step": 9220 }, { "epoch": 2.4536987759446514, "grad_norm": 0.2939448356628418, "learning_rate": 1.061891456629682e-07, "loss": 0.2102, "step": 9221 }, { "epoch": 2.453964874933475, "grad_norm": 0.28911611437797546, "learning_rate": 1.0617211738864456e-07, "loss": 0.2008, "step": 9222 }, { "epoch": 2.454230973922299, "grad_norm": 0.2985973060131073, "learning_rate": 1.0615508893466661e-07, "loss": 0.214, "step": 9223 }, { "epoch": 2.454497072911123, "grad_norm": 0.3002115786075592, "learning_rate": 1.0613806030153003e-07, "loss": 0.1996, "step": 9224 }, { "epoch": 2.454763171899947, "grad_norm": 0.2995166480541229, "learning_rate": 1.0612103148973046e-07, "loss": 0.2085, "step": 9225 }, { "epoch": 2.4550292708887707, "grad_norm": 0.30213695764541626, "learning_rate": 1.0610400249976352e-07, "loss": 0.1913, "step": 9226 }, { "epoch": 2.4552953698775943, "grad_norm": 0.286940336227417, "learning_rate": 1.0608697333212492e-07, "loss": 0.1938, "step": 9227 }, { "epoch": 2.4555614688664185, "grad_norm": 0.29785314202308655, "learning_rate": 1.0606994398731035e-07, "loss": 0.2019, "step": 9228 }, { "epoch": 2.455827567855242, "grad_norm": 0.28963834047317505, "learning_rate": 1.0605291446581546e-07, "loss": 0.1998, "step": 9229 }, { "epoch": 2.456093666844066, "grad_norm": 0.29517102241516113, "learning_rate": 1.0603588476813596e-07, "loss": 0.2014, "step": 9230 }, { "epoch": 2.45635976583289, "grad_norm": 0.2811574637889862, "learning_rate": 1.0601885489476752e-07, "loss": 0.1988, "step": 9231 }, { "epoch": 2.4566258648217136, "grad_norm": 0.3171701729297638, "learning_rate": 1.0600182484620583e-07, "loss": 0.2121, "step": 9232 }, { "epoch": 2.4568919638105378, "grad_norm": 0.285934716463089, "learning_rate": 1.0598479462294662e-07, "loss": 0.1963, "step": 9233 }, { "epoch": 2.4571580627993614, "grad_norm": 0.3072398006916046, "learning_rate": 1.0596776422548556e-07, "loss": 0.1991, "step": 9234 }, { "epoch": 2.457424161788185, "grad_norm": 0.29835793375968933, "learning_rate": 1.0595073365431841e-07, "loss": 0.2114, "step": 9235 }, { "epoch": 2.4576902607770093, "grad_norm": 0.28162550926208496, "learning_rate": 1.0593370290994082e-07, "loss": 0.1999, "step": 9236 }, { "epoch": 2.457956359765833, "grad_norm": 0.28968554735183716, "learning_rate": 1.0591667199284855e-07, "loss": 0.1968, "step": 9237 }, { "epoch": 2.4582224587546566, "grad_norm": 0.27313700318336487, "learning_rate": 1.0589964090353732e-07, "loss": 0.1868, "step": 9238 }, { "epoch": 2.4584885577434807, "grad_norm": 0.2768264710903168, "learning_rate": 1.0588260964250289e-07, "loss": 0.1847, "step": 9239 }, { "epoch": 2.4587546567323044, "grad_norm": 0.25654202699661255, "learning_rate": 1.0586557821024093e-07, "loss": 0.1745, "step": 9240 }, { "epoch": 2.459020755721128, "grad_norm": 0.28591659665107727, "learning_rate": 1.0584854660724723e-07, "loss": 0.1982, "step": 9241 }, { "epoch": 2.4592868547099522, "grad_norm": 0.2925359606742859, "learning_rate": 1.0583151483401753e-07, "loss": 0.2018, "step": 9242 }, { "epoch": 2.459552953698776, "grad_norm": 0.31005823612213135, "learning_rate": 1.0581448289104757e-07, "loss": 0.1927, "step": 9243 }, { "epoch": 2.4598190526875996, "grad_norm": 0.31138211488723755, "learning_rate": 1.0579745077883313e-07, "loss": 0.1937, "step": 9244 }, { "epoch": 2.4600851516764237, "grad_norm": 0.32311251759529114, "learning_rate": 1.0578041849786994e-07, "loss": 0.2142, "step": 9245 }, { "epoch": 2.4603512506652474, "grad_norm": 0.2766056954860687, "learning_rate": 1.0576338604865376e-07, "loss": 0.1875, "step": 9246 }, { "epoch": 2.460617349654071, "grad_norm": 0.2934166193008423, "learning_rate": 1.0574635343168039e-07, "loss": 0.2004, "step": 9247 }, { "epoch": 2.460883448642895, "grad_norm": 0.278103232383728, "learning_rate": 1.057293206474456e-07, "loss": 0.1995, "step": 9248 }, { "epoch": 2.461149547631719, "grad_norm": 0.31363940238952637, "learning_rate": 1.0571228769644514e-07, "loss": 0.2168, "step": 9249 }, { "epoch": 2.461415646620543, "grad_norm": 0.29530152678489685, "learning_rate": 1.0569525457917483e-07, "loss": 0.1925, "step": 9250 }, { "epoch": 2.4616817456093667, "grad_norm": 0.3026600778102875, "learning_rate": 1.0567822129613042e-07, "loss": 0.1985, "step": 9251 }, { "epoch": 2.4619478445981904, "grad_norm": 0.3074595332145691, "learning_rate": 1.0566118784780777e-07, "loss": 0.1991, "step": 9252 }, { "epoch": 2.4622139435870145, "grad_norm": 0.2918050289154053, "learning_rate": 1.0564415423470263e-07, "loss": 0.1951, "step": 9253 }, { "epoch": 2.462480042575838, "grad_norm": 0.29835227131843567, "learning_rate": 1.0562712045731083e-07, "loss": 0.2265, "step": 9254 }, { "epoch": 2.462746141564662, "grad_norm": 0.3015149235725403, "learning_rate": 1.0561008651612813e-07, "loss": 0.2052, "step": 9255 }, { "epoch": 2.463012240553486, "grad_norm": 0.2567014694213867, "learning_rate": 1.0559305241165039e-07, "loss": 0.1748, "step": 9256 }, { "epoch": 2.4632783395423097, "grad_norm": 0.3142040967941284, "learning_rate": 1.0557601814437345e-07, "loss": 0.2044, "step": 9257 }, { "epoch": 2.463544438531134, "grad_norm": 0.3327663540840149, "learning_rate": 1.0555898371479306e-07, "loss": 0.1955, "step": 9258 }, { "epoch": 2.4638105375199575, "grad_norm": 0.286521315574646, "learning_rate": 1.0554194912340508e-07, "loss": 0.1873, "step": 9259 }, { "epoch": 2.464076636508781, "grad_norm": 0.31536322832107544, "learning_rate": 1.0552491437070536e-07, "loss": 0.1983, "step": 9260 }, { "epoch": 2.4643427354976053, "grad_norm": 0.2808834910392761, "learning_rate": 1.0550787945718976e-07, "loss": 0.1919, "step": 9261 }, { "epoch": 2.464608834486429, "grad_norm": 0.3559302091598511, "learning_rate": 1.0549084438335405e-07, "loss": 0.2034, "step": 9262 }, { "epoch": 2.4648749334752527, "grad_norm": 0.31518009305000305, "learning_rate": 1.0547380914969413e-07, "loss": 0.2121, "step": 9263 }, { "epoch": 2.465141032464077, "grad_norm": 0.2679135501384735, "learning_rate": 1.0545677375670586e-07, "loss": 0.1983, "step": 9264 }, { "epoch": 2.4654071314529005, "grad_norm": 0.2768208086490631, "learning_rate": 1.0543973820488505e-07, "loss": 0.1875, "step": 9265 }, { "epoch": 2.465673230441724, "grad_norm": 0.270641952753067, "learning_rate": 1.054227024947276e-07, "loss": 0.2049, "step": 9266 }, { "epoch": 2.4659393294305483, "grad_norm": 0.2893134653568268, "learning_rate": 1.0540566662672934e-07, "loss": 0.193, "step": 9267 }, { "epoch": 2.466205428419372, "grad_norm": 0.2658759355545044, "learning_rate": 1.0538863060138617e-07, "loss": 0.1665, "step": 9268 }, { "epoch": 2.4664715274081956, "grad_norm": 0.28615736961364746, "learning_rate": 1.0537159441919394e-07, "loss": 0.1827, "step": 9269 }, { "epoch": 2.4667376263970198, "grad_norm": 0.3005063533782959, "learning_rate": 1.0535455808064857e-07, "loss": 0.2047, "step": 9270 }, { "epoch": 2.4670037253858434, "grad_norm": 0.27768418192863464, "learning_rate": 1.0533752158624591e-07, "loss": 0.1846, "step": 9271 }, { "epoch": 2.467269824374667, "grad_norm": 0.2932039201259613, "learning_rate": 1.0532048493648185e-07, "loss": 0.1893, "step": 9272 }, { "epoch": 2.4675359233634913, "grad_norm": 0.3139360249042511, "learning_rate": 1.0530344813185228e-07, "loss": 0.2162, "step": 9273 }, { "epoch": 2.467802022352315, "grad_norm": 0.27420690655708313, "learning_rate": 1.0528641117285313e-07, "loss": 0.1853, "step": 9274 }, { "epoch": 2.468068121341139, "grad_norm": 0.3142856955528259, "learning_rate": 1.0526937405998029e-07, "loss": 0.2079, "step": 9275 }, { "epoch": 2.4683342203299627, "grad_norm": 0.2740411162376404, "learning_rate": 1.052523367937296e-07, "loss": 0.1775, "step": 9276 }, { "epoch": 2.4686003193187864, "grad_norm": 0.2902586758136749, "learning_rate": 1.0523529937459705e-07, "loss": 0.1815, "step": 9277 }, { "epoch": 2.4688664183076106, "grad_norm": 0.3061467707157135, "learning_rate": 1.0521826180307853e-07, "loss": 0.2015, "step": 9278 }, { "epoch": 2.4691325172964342, "grad_norm": 0.30227452516555786, "learning_rate": 1.0520122407966994e-07, "loss": 0.1994, "step": 9279 }, { "epoch": 2.4693986162852584, "grad_norm": 0.27045348286628723, "learning_rate": 1.0518418620486726e-07, "loss": 0.1863, "step": 9280 }, { "epoch": 2.469664715274082, "grad_norm": 0.2660072445869446, "learning_rate": 1.0516714817916634e-07, "loss": 0.1759, "step": 9281 }, { "epoch": 2.4699308142629057, "grad_norm": 0.29292839765548706, "learning_rate": 1.0515011000306318e-07, "loss": 0.1859, "step": 9282 }, { "epoch": 2.47019691325173, "grad_norm": 0.28930190205574036, "learning_rate": 1.0513307167705366e-07, "loss": 0.1975, "step": 9283 }, { "epoch": 2.4704630122405535, "grad_norm": 0.29473263025283813, "learning_rate": 1.0511603320163378e-07, "loss": 0.2007, "step": 9284 }, { "epoch": 2.470729111229377, "grad_norm": 0.2790910005569458, "learning_rate": 1.0509899457729947e-07, "loss": 0.2085, "step": 9285 }, { "epoch": 2.4709952102182013, "grad_norm": 0.32194429636001587, "learning_rate": 1.0508195580454664e-07, "loss": 0.2006, "step": 9286 }, { "epoch": 2.471261309207025, "grad_norm": 0.2981908917427063, "learning_rate": 1.0506491688387126e-07, "loss": 0.2031, "step": 9287 }, { "epoch": 2.4715274081958487, "grad_norm": 0.28398457169532776, "learning_rate": 1.0504787781576931e-07, "loss": 0.1854, "step": 9288 }, { "epoch": 2.471793507184673, "grad_norm": 0.27464956045150757, "learning_rate": 1.0503083860073679e-07, "loss": 0.1907, "step": 9289 }, { "epoch": 2.4720596061734965, "grad_norm": 0.2762002646923065, "learning_rate": 1.0501379923926956e-07, "loss": 0.1892, "step": 9290 }, { "epoch": 2.47232570516232, "grad_norm": 0.3020682632923126, "learning_rate": 1.0499675973186369e-07, "loss": 0.1998, "step": 9291 }, { "epoch": 2.4725918041511443, "grad_norm": 0.28278684616088867, "learning_rate": 1.0497972007901508e-07, "loss": 0.1843, "step": 9292 }, { "epoch": 2.472857903139968, "grad_norm": 0.32121458649635315, "learning_rate": 1.049626802812198e-07, "loss": 0.2251, "step": 9293 }, { "epoch": 2.4731240021287917, "grad_norm": 0.2953106164932251, "learning_rate": 1.0494564033897373e-07, "loss": 0.2055, "step": 9294 }, { "epoch": 2.473390101117616, "grad_norm": 0.2901684641838074, "learning_rate": 1.0492860025277295e-07, "loss": 0.2023, "step": 9295 }, { "epoch": 2.4736562001064395, "grad_norm": 0.35647380352020264, "learning_rate": 1.0491156002311338e-07, "loss": 0.1977, "step": 9296 }, { "epoch": 2.4739222990952636, "grad_norm": 0.29369303584098816, "learning_rate": 1.0489451965049108e-07, "loss": 0.1853, "step": 9297 }, { "epoch": 2.4741883980840873, "grad_norm": 0.2860049605369568, "learning_rate": 1.0487747913540201e-07, "loss": 0.1929, "step": 9298 }, { "epoch": 2.474454497072911, "grad_norm": 0.31950873136520386, "learning_rate": 1.0486043847834218e-07, "loss": 0.1977, "step": 9299 }, { "epoch": 2.474720596061735, "grad_norm": 0.3080255687236786, "learning_rate": 1.048433976798076e-07, "loss": 0.1888, "step": 9300 }, { "epoch": 2.474986695050559, "grad_norm": 0.29992273449897766, "learning_rate": 1.048263567402943e-07, "loss": 0.2056, "step": 9301 }, { "epoch": 2.4752527940393825, "grad_norm": 0.31109756231307983, "learning_rate": 1.0480931566029831e-07, "loss": 0.2139, "step": 9302 }, { "epoch": 2.4755188930282066, "grad_norm": 0.29892340302467346, "learning_rate": 1.047922744403156e-07, "loss": 0.1916, "step": 9303 }, { "epoch": 2.4757849920170303, "grad_norm": 0.3042890727519989, "learning_rate": 1.0477523308084222e-07, "loss": 0.2014, "step": 9304 }, { "epoch": 2.4760510910058544, "grad_norm": 0.3359890282154083, "learning_rate": 1.0475819158237424e-07, "loss": 0.213, "step": 9305 }, { "epoch": 2.476317189994678, "grad_norm": 0.29248398542404175, "learning_rate": 1.0474114994540761e-07, "loss": 0.1954, "step": 9306 }, { "epoch": 2.4765832889835018, "grad_norm": 0.2998009920120239, "learning_rate": 1.0472410817043846e-07, "loss": 0.1759, "step": 9307 }, { "epoch": 2.476849387972326, "grad_norm": 0.29290875792503357, "learning_rate": 1.0470706625796277e-07, "loss": 0.2177, "step": 9308 }, { "epoch": 2.4771154869611496, "grad_norm": 0.28244146704673767, "learning_rate": 1.046900242084766e-07, "loss": 0.1981, "step": 9309 }, { "epoch": 2.4773815859499733, "grad_norm": 0.2855718731880188, "learning_rate": 1.0467298202247598e-07, "loss": 0.2077, "step": 9310 }, { "epoch": 2.4776476849387974, "grad_norm": 0.3071414828300476, "learning_rate": 1.0465593970045703e-07, "loss": 0.2185, "step": 9311 }, { "epoch": 2.477913783927621, "grad_norm": 0.2825741469860077, "learning_rate": 1.0463889724291572e-07, "loss": 0.1902, "step": 9312 }, { "epoch": 2.4781798829164448, "grad_norm": 0.2955031991004944, "learning_rate": 1.0462185465034819e-07, "loss": 0.2022, "step": 9313 }, { "epoch": 2.478445981905269, "grad_norm": 0.2903554439544678, "learning_rate": 1.0460481192325043e-07, "loss": 0.1822, "step": 9314 }, { "epoch": 2.4787120808940926, "grad_norm": 0.26794368028640747, "learning_rate": 1.045877690621186e-07, "loss": 0.1876, "step": 9315 }, { "epoch": 2.4789781798829162, "grad_norm": 0.3001294434070587, "learning_rate": 1.0457072606744872e-07, "loss": 0.1953, "step": 9316 }, { "epoch": 2.4792442788717404, "grad_norm": 0.3147033452987671, "learning_rate": 1.0455368293973685e-07, "loss": 0.1982, "step": 9317 }, { "epoch": 2.479510377860564, "grad_norm": 0.2966737151145935, "learning_rate": 1.045366396794791e-07, "loss": 0.1945, "step": 9318 }, { "epoch": 2.4797764768493877, "grad_norm": 0.3337019681930542, "learning_rate": 1.0451959628717153e-07, "loss": 0.2323, "step": 9319 }, { "epoch": 2.480042575838212, "grad_norm": 0.3095190227031708, "learning_rate": 1.0450255276331029e-07, "loss": 0.2103, "step": 9320 }, { "epoch": 2.4803086748270355, "grad_norm": 0.30408239364624023, "learning_rate": 1.044855091083914e-07, "loss": 0.1981, "step": 9321 }, { "epoch": 2.4805747738158597, "grad_norm": 0.284359872341156, "learning_rate": 1.0446846532291098e-07, "loss": 0.1889, "step": 9322 }, { "epoch": 2.4808408728046834, "grad_norm": 0.29859453439712524, "learning_rate": 1.0445142140736514e-07, "loss": 0.1961, "step": 9323 }, { "epoch": 2.481106971793507, "grad_norm": 0.2963787615299225, "learning_rate": 1.0443437736225001e-07, "loss": 0.195, "step": 9324 }, { "epoch": 2.481373070782331, "grad_norm": 0.28733181953430176, "learning_rate": 1.0441733318806163e-07, "loss": 0.1822, "step": 9325 }, { "epoch": 2.481639169771155, "grad_norm": 0.29024019837379456, "learning_rate": 1.044002888852962e-07, "loss": 0.1968, "step": 9326 }, { "epoch": 2.4819052687599785, "grad_norm": 0.3010682165622711, "learning_rate": 1.0438324445444975e-07, "loss": 0.2143, "step": 9327 }, { "epoch": 2.4821713677488026, "grad_norm": 0.2711707353591919, "learning_rate": 1.0436619989601843e-07, "loss": 0.1858, "step": 9328 }, { "epoch": 2.4824374667376263, "grad_norm": 0.28720948100090027, "learning_rate": 1.0434915521049839e-07, "loss": 0.178, "step": 9329 }, { "epoch": 2.4827035657264505, "grad_norm": 0.28244146704673767, "learning_rate": 1.043321103983857e-07, "loss": 0.1873, "step": 9330 }, { "epoch": 2.482969664715274, "grad_norm": 0.32460492849349976, "learning_rate": 1.0431506546017654e-07, "loss": 0.2168, "step": 9331 }, { "epoch": 2.483235763704098, "grad_norm": 0.270982563495636, "learning_rate": 1.0429802039636703e-07, "loss": 0.1798, "step": 9332 }, { "epoch": 2.483501862692922, "grad_norm": 0.27185559272766113, "learning_rate": 1.0428097520745332e-07, "loss": 0.1735, "step": 9333 }, { "epoch": 2.4837679616817456, "grad_norm": 0.3079521059989929, "learning_rate": 1.0426392989393154e-07, "loss": 0.2063, "step": 9334 }, { "epoch": 2.4840340606705693, "grad_norm": 0.29230743646621704, "learning_rate": 1.0424688445629781e-07, "loss": 0.1908, "step": 9335 }, { "epoch": 2.4843001596593934, "grad_norm": 0.31001147627830505, "learning_rate": 1.042298388950483e-07, "loss": 0.1932, "step": 9336 }, { "epoch": 2.484566258648217, "grad_norm": 0.26964041590690613, "learning_rate": 1.0421279321067917e-07, "loss": 0.187, "step": 9337 }, { "epoch": 2.484832357637041, "grad_norm": 0.2868858873844147, "learning_rate": 1.041957474036866e-07, "loss": 0.1873, "step": 9338 }, { "epoch": 2.485098456625865, "grad_norm": 0.28977900743484497, "learning_rate": 1.0417870147456667e-07, "loss": 0.1826, "step": 9339 }, { "epoch": 2.4853645556146886, "grad_norm": 0.3098459541797638, "learning_rate": 1.041616554238156e-07, "loss": 0.2017, "step": 9340 }, { "epoch": 2.4856306546035123, "grad_norm": 0.2926102876663208, "learning_rate": 1.0414460925192956e-07, "loss": 0.186, "step": 9341 }, { "epoch": 2.4858967535923364, "grad_norm": 0.2749548852443695, "learning_rate": 1.0412756295940468e-07, "loss": 0.1823, "step": 9342 }, { "epoch": 2.48616285258116, "grad_norm": 0.29214581847190857, "learning_rate": 1.0411051654673719e-07, "loss": 0.2023, "step": 9343 }, { "epoch": 2.4864289515699842, "grad_norm": 0.299265593290329, "learning_rate": 1.0409347001442324e-07, "loss": 0.215, "step": 9344 }, { "epoch": 2.486695050558808, "grad_norm": 0.30288055539131165, "learning_rate": 1.0407642336295899e-07, "loss": 0.2031, "step": 9345 }, { "epoch": 2.4869611495476316, "grad_norm": 0.30767014622688293, "learning_rate": 1.0405937659284063e-07, "loss": 0.1801, "step": 9346 }, { "epoch": 2.4872272485364557, "grad_norm": 0.2944224178791046, "learning_rate": 1.0404232970456439e-07, "loss": 0.1822, "step": 9347 }, { "epoch": 2.4874933475252794, "grad_norm": 0.2828732430934906, "learning_rate": 1.040252826986264e-07, "loss": 0.1998, "step": 9348 }, { "epoch": 2.487759446514103, "grad_norm": 0.30153888463974, "learning_rate": 1.040082355755229e-07, "loss": 0.2115, "step": 9349 }, { "epoch": 2.488025545502927, "grad_norm": 0.2947537899017334, "learning_rate": 1.0399118833575004e-07, "loss": 0.2011, "step": 9350 }, { "epoch": 2.488291644491751, "grad_norm": 0.2901046574115753, "learning_rate": 1.0397414097980406e-07, "loss": 0.2096, "step": 9351 }, { "epoch": 2.488557743480575, "grad_norm": 0.27933627367019653, "learning_rate": 1.0395709350818118e-07, "loss": 0.1985, "step": 9352 }, { "epoch": 2.4888238424693987, "grad_norm": 0.3152606189250946, "learning_rate": 1.0394004592137755e-07, "loss": 0.1995, "step": 9353 }, { "epoch": 2.4890899414582224, "grad_norm": 0.27859440445899963, "learning_rate": 1.0392299821988944e-07, "loss": 0.1883, "step": 9354 }, { "epoch": 2.4893560404470465, "grad_norm": 0.28160637617111206, "learning_rate": 1.0390595040421304e-07, "loss": 0.2064, "step": 9355 }, { "epoch": 2.48962213943587, "grad_norm": 0.29968246817588806, "learning_rate": 1.0388890247484454e-07, "loss": 0.1994, "step": 9356 }, { "epoch": 2.489888238424694, "grad_norm": 0.29903092980384827, "learning_rate": 1.0387185443228021e-07, "loss": 0.2151, "step": 9357 }, { "epoch": 2.490154337413518, "grad_norm": 0.2868562340736389, "learning_rate": 1.0385480627701623e-07, "loss": 0.1961, "step": 9358 }, { "epoch": 2.4904204364023417, "grad_norm": 0.3207690119743347, "learning_rate": 1.0383775800954886e-07, "loss": 0.1878, "step": 9359 }, { "epoch": 2.4906865353911654, "grad_norm": 0.3014492094516754, "learning_rate": 1.0382070963037432e-07, "loss": 0.1966, "step": 9360 }, { "epoch": 2.4909526343799895, "grad_norm": 0.4008098542690277, "learning_rate": 1.0380366113998884e-07, "loss": 0.1915, "step": 9361 }, { "epoch": 2.491218733368813, "grad_norm": 0.2689894437789917, "learning_rate": 1.0378661253888867e-07, "loss": 0.1964, "step": 9362 }, { "epoch": 2.491484832357637, "grad_norm": 0.3086966872215271, "learning_rate": 1.0376956382757003e-07, "loss": 0.185, "step": 9363 }, { "epoch": 2.491750931346461, "grad_norm": 0.30031758546829224, "learning_rate": 1.0375251500652915e-07, "loss": 0.1828, "step": 9364 }, { "epoch": 2.4920170303352847, "grad_norm": 0.2875942885875702, "learning_rate": 1.0373546607626235e-07, "loss": 0.1927, "step": 9365 }, { "epoch": 2.4922831293241083, "grad_norm": 0.28443974256515503, "learning_rate": 1.037184170372658e-07, "loss": 0.1767, "step": 9366 }, { "epoch": 2.4925492283129325, "grad_norm": 0.30264395475387573, "learning_rate": 1.0370136789003581e-07, "loss": 0.1965, "step": 9367 }, { "epoch": 2.492815327301756, "grad_norm": 0.2961884140968323, "learning_rate": 1.036843186350686e-07, "loss": 0.1943, "step": 9368 }, { "epoch": 2.4930814262905803, "grad_norm": 0.2811256945133209, "learning_rate": 1.0366726927286042e-07, "loss": 0.2039, "step": 9369 }, { "epoch": 2.493347525279404, "grad_norm": 0.2802819311618805, "learning_rate": 1.0365021980390758e-07, "loss": 0.1932, "step": 9370 }, { "epoch": 2.4936136242682276, "grad_norm": 0.27154281735420227, "learning_rate": 1.036331702287063e-07, "loss": 0.173, "step": 9371 }, { "epoch": 2.4938797232570518, "grad_norm": 0.2938764989376068, "learning_rate": 1.0361612054775288e-07, "loss": 0.2026, "step": 9372 }, { "epoch": 2.4941458222458754, "grad_norm": 0.29870283603668213, "learning_rate": 1.0359907076154357e-07, "loss": 0.2067, "step": 9373 }, { "epoch": 2.494411921234699, "grad_norm": 0.3100726306438446, "learning_rate": 1.0358202087057468e-07, "loss": 0.1991, "step": 9374 }, { "epoch": 2.4946780202235233, "grad_norm": 0.2868695557117462, "learning_rate": 1.0356497087534244e-07, "loss": 0.1998, "step": 9375 }, { "epoch": 2.494944119212347, "grad_norm": 0.27917224168777466, "learning_rate": 1.0354792077634317e-07, "loss": 0.176, "step": 9376 }, { "epoch": 2.495210218201171, "grad_norm": 0.27611035108566284, "learning_rate": 1.0353087057407312e-07, "loss": 0.1921, "step": 9377 }, { "epoch": 2.4954763171899947, "grad_norm": 0.30757448077201843, "learning_rate": 1.035138202690286e-07, "loss": 0.1948, "step": 9378 }, { "epoch": 2.4957424161788184, "grad_norm": 0.2775898575782776, "learning_rate": 1.0349676986170593e-07, "loss": 0.1802, "step": 9379 }, { "epoch": 2.4960085151676425, "grad_norm": 0.3795239329338074, "learning_rate": 1.0347971935260131e-07, "loss": 0.2145, "step": 9380 }, { "epoch": 2.4962746141564662, "grad_norm": 0.3143330514431, "learning_rate": 1.0346266874221113e-07, "loss": 0.2229, "step": 9381 }, { "epoch": 2.49654071314529, "grad_norm": 0.2982024550437927, "learning_rate": 1.0344561803103162e-07, "loss": 0.1939, "step": 9382 }, { "epoch": 2.496806812134114, "grad_norm": 0.29581329226493835, "learning_rate": 1.0342856721955917e-07, "loss": 0.1861, "step": 9383 }, { "epoch": 2.4970729111229377, "grad_norm": 0.2873145639896393, "learning_rate": 1.0341151630829e-07, "loss": 0.196, "step": 9384 }, { "epoch": 2.4973390101117614, "grad_norm": 0.3043362498283386, "learning_rate": 1.0339446529772042e-07, "loss": 0.1818, "step": 9385 }, { "epoch": 2.4976051091005855, "grad_norm": 0.2592805027961731, "learning_rate": 1.0337741418834682e-07, "loss": 0.1732, "step": 9386 }, { "epoch": 2.497871208089409, "grad_norm": 0.32336774468421936, "learning_rate": 1.0336036298066544e-07, "loss": 0.2138, "step": 9387 }, { "epoch": 2.498137307078233, "grad_norm": 0.6936593055725098, "learning_rate": 1.0334331167517263e-07, "loss": 0.1835, "step": 9388 }, { "epoch": 2.498403406067057, "grad_norm": 0.2927296757698059, "learning_rate": 1.0332626027236469e-07, "loss": 0.1816, "step": 9389 }, { "epoch": 2.4986695050558807, "grad_norm": 0.2986987233161926, "learning_rate": 1.0330920877273794e-07, "loss": 0.19, "step": 9390 }, { "epoch": 2.4989356040447044, "grad_norm": 0.2904110550880432, "learning_rate": 1.0329215717678872e-07, "loss": 0.1904, "step": 9391 }, { "epoch": 2.4992017030335285, "grad_norm": 0.31432589888572693, "learning_rate": 1.0327510548501337e-07, "loss": 0.1914, "step": 9392 }, { "epoch": 2.499467802022352, "grad_norm": 0.2760288417339325, "learning_rate": 1.0325805369790817e-07, "loss": 0.202, "step": 9393 }, { "epoch": 2.4997339010111763, "grad_norm": 0.2869536578655243, "learning_rate": 1.032410018159695e-07, "loss": 0.1775, "step": 9394 }, { "epoch": 2.5, "grad_norm": 0.3134712874889374, "learning_rate": 1.0322394983969367e-07, "loss": 0.2181, "step": 9395 }, { "epoch": 2.5002660989888237, "grad_norm": 0.3007209599018097, "learning_rate": 1.0320689776957707e-07, "loss": 0.1872, "step": 9396 }, { "epoch": 2.500532197977648, "grad_norm": 0.37513166666030884, "learning_rate": 1.0318984560611599e-07, "loss": 0.2052, "step": 9397 }, { "epoch": 2.5007982969664715, "grad_norm": 0.2792762517929077, "learning_rate": 1.0317279334980678e-07, "loss": 0.194, "step": 9398 }, { "epoch": 2.5010643959552956, "grad_norm": 0.2706094980239868, "learning_rate": 1.0315574100114578e-07, "loss": 0.1912, "step": 9399 }, { "epoch": 2.5013304949441193, "grad_norm": 0.26659607887268066, "learning_rate": 1.0313868856062935e-07, "loss": 0.1817, "step": 9400 }, { "epoch": 2.501596593932943, "grad_norm": 0.27331721782684326, "learning_rate": 1.0312163602875386e-07, "loss": 0.1807, "step": 9401 }, { "epoch": 2.501862692921767, "grad_norm": 0.29068756103515625, "learning_rate": 1.0310458340601567e-07, "loss": 0.2031, "step": 9402 }, { "epoch": 2.502128791910591, "grad_norm": 0.2915784418582916, "learning_rate": 1.030875306929111e-07, "loss": 0.2102, "step": 9403 }, { "epoch": 2.5023948908994145, "grad_norm": 0.29679247736930847, "learning_rate": 1.0307047788993651e-07, "loss": 0.2088, "step": 9404 }, { "epoch": 2.5026609898882386, "grad_norm": 0.2998930811882019, "learning_rate": 1.030534249975883e-07, "loss": 0.2187, "step": 9405 }, { "epoch": 2.5029270888770623, "grad_norm": 0.2789885103702545, "learning_rate": 1.0303637201636282e-07, "loss": 0.1992, "step": 9406 }, { "epoch": 2.503193187865886, "grad_norm": 0.3259103298187256, "learning_rate": 1.0301931894675645e-07, "loss": 0.2293, "step": 9407 }, { "epoch": 2.50345928685471, "grad_norm": 0.28612568974494934, "learning_rate": 1.0300226578926549e-07, "loss": 0.1981, "step": 9408 }, { "epoch": 2.5037253858435338, "grad_norm": 0.26753103733062744, "learning_rate": 1.0298521254438641e-07, "loss": 0.1769, "step": 9409 }, { "epoch": 2.5039914848323575, "grad_norm": 0.30076292157173157, "learning_rate": 1.0296815921261552e-07, "loss": 0.1836, "step": 9410 }, { "epoch": 2.5042575838211816, "grad_norm": 0.30150434374809265, "learning_rate": 1.0295110579444925e-07, "loss": 0.1938, "step": 9411 }, { "epoch": 2.5045236828100053, "grad_norm": 0.3047023415565491, "learning_rate": 1.0293405229038394e-07, "loss": 0.2105, "step": 9412 }, { "epoch": 2.504789781798829, "grad_norm": 0.27251914143562317, "learning_rate": 1.0291699870091599e-07, "loss": 0.1948, "step": 9413 }, { "epoch": 2.505055880787653, "grad_norm": 0.28123190999031067, "learning_rate": 1.0289994502654175e-07, "loss": 0.1897, "step": 9414 }, { "epoch": 2.5053219797764767, "grad_norm": 0.2850694954395294, "learning_rate": 1.0288289126775769e-07, "loss": 0.1981, "step": 9415 }, { "epoch": 2.5055880787653004, "grad_norm": 0.2908112406730652, "learning_rate": 1.0286583742506012e-07, "loss": 0.1923, "step": 9416 }, { "epoch": 2.5058541777541246, "grad_norm": 0.3010086119174957, "learning_rate": 1.028487834989455e-07, "loss": 0.1995, "step": 9417 }, { "epoch": 2.5061202767429482, "grad_norm": 0.28674256801605225, "learning_rate": 1.0283172948991015e-07, "loss": 0.2004, "step": 9418 }, { "epoch": 2.5063863757317724, "grad_norm": 0.30207544565200806, "learning_rate": 1.028146753984505e-07, "loss": 0.204, "step": 9419 }, { "epoch": 2.506652474720596, "grad_norm": 0.2920713722705841, "learning_rate": 1.0279762122506299e-07, "loss": 0.1787, "step": 9420 }, { "epoch": 2.50691857370942, "grad_norm": 0.3003882169723511, "learning_rate": 1.0278056697024397e-07, "loss": 0.207, "step": 9421 }, { "epoch": 2.507184672698244, "grad_norm": 0.31050264835357666, "learning_rate": 1.0276351263448988e-07, "loss": 0.2254, "step": 9422 }, { "epoch": 2.5074507716870675, "grad_norm": 0.29959750175476074, "learning_rate": 1.0274645821829709e-07, "loss": 0.1978, "step": 9423 }, { "epoch": 2.5077168706758917, "grad_norm": 0.30373823642730713, "learning_rate": 1.0272940372216205e-07, "loss": 0.1884, "step": 9424 }, { "epoch": 2.5079829696647153, "grad_norm": 0.28101688623428345, "learning_rate": 1.0271234914658115e-07, "loss": 0.1805, "step": 9425 }, { "epoch": 2.508249068653539, "grad_norm": 0.2843727171421051, "learning_rate": 1.0269529449205081e-07, "loss": 0.1923, "step": 9426 }, { "epoch": 2.508515167642363, "grad_norm": 0.2804901897907257, "learning_rate": 1.0267823975906746e-07, "loss": 0.1919, "step": 9427 }, { "epoch": 2.508781266631187, "grad_norm": 0.28301259875297546, "learning_rate": 1.0266118494812748e-07, "loss": 0.1902, "step": 9428 }, { "epoch": 2.5090473656200105, "grad_norm": 0.2859644889831543, "learning_rate": 1.0264413005972733e-07, "loss": 0.1798, "step": 9429 }, { "epoch": 2.5093134646088346, "grad_norm": 0.32321304082870483, "learning_rate": 1.0262707509436342e-07, "loss": 0.1978, "step": 9430 }, { "epoch": 2.5095795635976583, "grad_norm": 0.29646795988082886, "learning_rate": 1.0261002005253216e-07, "loss": 0.1905, "step": 9431 }, { "epoch": 2.509845662586482, "grad_norm": 0.2801090478897095, "learning_rate": 1.0259296493473e-07, "loss": 0.1893, "step": 9432 }, { "epoch": 2.510111761575306, "grad_norm": 0.31618571281433105, "learning_rate": 1.0257590974145338e-07, "loss": 0.2178, "step": 9433 }, { "epoch": 2.51037786056413, "grad_norm": 0.2892795503139496, "learning_rate": 1.025588544731987e-07, "loss": 0.199, "step": 9434 }, { "epoch": 2.5106439595529535, "grad_norm": 0.2754785716533661, "learning_rate": 1.0254179913046239e-07, "loss": 0.1961, "step": 9435 }, { "epoch": 2.5109100585417776, "grad_norm": 0.28337109088897705, "learning_rate": 1.0252474371374094e-07, "loss": 0.1837, "step": 9436 }, { "epoch": 2.5111761575306013, "grad_norm": 0.28357642889022827, "learning_rate": 1.0250768822353075e-07, "loss": 0.2093, "step": 9437 }, { "epoch": 2.511442256519425, "grad_norm": 0.29627636075019836, "learning_rate": 1.0249063266032827e-07, "loss": 0.2004, "step": 9438 }, { "epoch": 2.511708355508249, "grad_norm": 0.3053172826766968, "learning_rate": 1.0247357702462993e-07, "loss": 0.2087, "step": 9439 }, { "epoch": 2.511974454497073, "grad_norm": 0.30054745078086853, "learning_rate": 1.0245652131693219e-07, "loss": 0.1981, "step": 9440 }, { "epoch": 2.5122405534858965, "grad_norm": 0.2877083420753479, "learning_rate": 1.0243946553773148e-07, "loss": 0.1898, "step": 9441 }, { "epoch": 2.5125066524747206, "grad_norm": 0.2880697250366211, "learning_rate": 1.0242240968752429e-07, "loss": 0.1659, "step": 9442 }, { "epoch": 2.5127727514635443, "grad_norm": 0.3131512999534607, "learning_rate": 1.0240535376680701e-07, "loss": 0.2155, "step": 9443 }, { "epoch": 2.5130388504523684, "grad_norm": 0.29843151569366455, "learning_rate": 1.0238829777607613e-07, "loss": 0.187, "step": 9444 }, { "epoch": 2.513304949441192, "grad_norm": 0.2887440323829651, "learning_rate": 1.0237124171582811e-07, "loss": 0.1851, "step": 9445 }, { "epoch": 2.513571048430016, "grad_norm": 0.28567105531692505, "learning_rate": 1.0235418558655944e-07, "loss": 0.1913, "step": 9446 }, { "epoch": 2.51383714741884, "grad_norm": 0.313656210899353, "learning_rate": 1.0233712938876649e-07, "loss": 0.2127, "step": 9447 }, { "epoch": 2.5141032464076636, "grad_norm": 0.30634719133377075, "learning_rate": 1.0232007312294582e-07, "loss": 0.1995, "step": 9448 }, { "epoch": 2.5143693453964877, "grad_norm": 0.3033616840839386, "learning_rate": 1.0230301678959378e-07, "loss": 0.1949, "step": 9449 }, { "epoch": 2.5146354443853114, "grad_norm": 0.34140270948410034, "learning_rate": 1.0228596038920693e-07, "loss": 0.2018, "step": 9450 }, { "epoch": 2.514901543374135, "grad_norm": 0.2569703459739685, "learning_rate": 1.0226890392228173e-07, "loss": 0.1858, "step": 9451 }, { "epoch": 2.515167642362959, "grad_norm": 0.26593726873397827, "learning_rate": 1.022518473893146e-07, "loss": 0.1735, "step": 9452 }, { "epoch": 2.515433741351783, "grad_norm": 0.2892343997955322, "learning_rate": 1.0223479079080206e-07, "loss": 0.2011, "step": 9453 }, { "epoch": 2.5156998403406066, "grad_norm": 0.28928256034851074, "learning_rate": 1.0221773412724055e-07, "loss": 0.1974, "step": 9454 }, { "epoch": 2.5159659393294307, "grad_norm": 0.2866228520870209, "learning_rate": 1.0220067739912654e-07, "loss": 0.1833, "step": 9455 }, { "epoch": 2.5162320383182544, "grad_norm": 0.30154430866241455, "learning_rate": 1.0218362060695658e-07, "loss": 0.196, "step": 9456 }, { "epoch": 2.516498137307078, "grad_norm": 0.2949485182762146, "learning_rate": 1.0216656375122706e-07, "loss": 0.1908, "step": 9457 }, { "epoch": 2.516764236295902, "grad_norm": 0.31693315505981445, "learning_rate": 1.0214950683243451e-07, "loss": 0.2168, "step": 9458 }, { "epoch": 2.517030335284726, "grad_norm": 0.29961511492729187, "learning_rate": 1.0213244985107536e-07, "loss": 0.2044, "step": 9459 }, { "epoch": 2.5172964342735495, "grad_norm": 0.27170708775520325, "learning_rate": 1.0211539280764616e-07, "loss": 0.1943, "step": 9460 }, { "epoch": 2.5175625332623737, "grad_norm": 0.2659403383731842, "learning_rate": 1.0209833570264337e-07, "loss": 0.1881, "step": 9461 }, { "epoch": 2.5178286322511974, "grad_norm": 0.27099794149398804, "learning_rate": 1.0208127853656345e-07, "loss": 0.1981, "step": 9462 }, { "epoch": 2.518094731240021, "grad_norm": 0.277765691280365, "learning_rate": 1.0206422130990293e-07, "loss": 0.1905, "step": 9463 }, { "epoch": 2.518360830228845, "grad_norm": 0.2869216203689575, "learning_rate": 1.0204716402315829e-07, "loss": 0.1911, "step": 9464 }, { "epoch": 2.518626929217669, "grad_norm": 0.28885549306869507, "learning_rate": 1.0203010667682604e-07, "loss": 0.1871, "step": 9465 }, { "epoch": 2.518893028206493, "grad_norm": 0.2907469570636749, "learning_rate": 1.0201304927140261e-07, "loss": 0.1972, "step": 9466 }, { "epoch": 2.5191591271953166, "grad_norm": 0.28563863039016724, "learning_rate": 1.0199599180738461e-07, "loss": 0.1807, "step": 9467 }, { "epoch": 2.5194252261841408, "grad_norm": 0.2939508557319641, "learning_rate": 1.019789342852684e-07, "loss": 0.2101, "step": 9468 }, { "epoch": 2.5196913251729645, "grad_norm": 0.28371110558509827, "learning_rate": 1.0196187670555061e-07, "loss": 0.2019, "step": 9469 }, { "epoch": 2.519957424161788, "grad_norm": 0.2714385688304901, "learning_rate": 1.0194481906872765e-07, "loss": 0.1792, "step": 9470 }, { "epoch": 2.5202235231506123, "grad_norm": 0.318323016166687, "learning_rate": 1.0192776137529606e-07, "loss": 0.2055, "step": 9471 }, { "epoch": 2.520489622139436, "grad_norm": 0.30952683091163635, "learning_rate": 1.0191070362575234e-07, "loss": 0.1945, "step": 9472 }, { "epoch": 2.5207557211282596, "grad_norm": 0.31440630555152893, "learning_rate": 1.01893645820593e-07, "loss": 0.2033, "step": 9473 }, { "epoch": 2.5210218201170838, "grad_norm": 0.2880775034427643, "learning_rate": 1.0187658796031455e-07, "loss": 0.1935, "step": 9474 }, { "epoch": 2.5212879191059074, "grad_norm": 0.2931425869464874, "learning_rate": 1.0185953004541348e-07, "loss": 0.192, "step": 9475 }, { "epoch": 2.521554018094731, "grad_norm": 0.2822575867176056, "learning_rate": 1.0184247207638634e-07, "loss": 0.1875, "step": 9476 }, { "epoch": 2.5218201170835552, "grad_norm": 0.313201904296875, "learning_rate": 1.0182541405372962e-07, "loss": 0.2123, "step": 9477 }, { "epoch": 2.522086216072379, "grad_norm": 0.27696382999420166, "learning_rate": 1.0180835597793984e-07, "loss": 0.171, "step": 9478 }, { "epoch": 2.5223523150612026, "grad_norm": 0.29593873023986816, "learning_rate": 1.0179129784951353e-07, "loss": 0.1957, "step": 9479 }, { "epoch": 2.5226184140500267, "grad_norm": 0.28632888197898865, "learning_rate": 1.0177423966894713e-07, "loss": 0.1838, "step": 9480 }, { "epoch": 2.5228845130388504, "grad_norm": 0.286095529794693, "learning_rate": 1.0175718143673727e-07, "loss": 0.1734, "step": 9481 }, { "epoch": 2.523150612027674, "grad_norm": 0.312656432390213, "learning_rate": 1.0174012315338039e-07, "loss": 0.2168, "step": 9482 }, { "epoch": 2.5234167110164982, "grad_norm": 0.3202497363090515, "learning_rate": 1.0172306481937307e-07, "loss": 0.1847, "step": 9483 }, { "epoch": 2.523682810005322, "grad_norm": 0.2707534730434418, "learning_rate": 1.0170600643521181e-07, "loss": 0.1834, "step": 9484 }, { "epoch": 2.5239489089941456, "grad_norm": 0.31568172574043274, "learning_rate": 1.0168894800139309e-07, "loss": 0.2093, "step": 9485 }, { "epoch": 2.5242150079829697, "grad_norm": 0.3039451241493225, "learning_rate": 1.0167188951841353e-07, "loss": 0.2008, "step": 9486 }, { "epoch": 2.5244811069717934, "grad_norm": 0.28294748067855835, "learning_rate": 1.0165483098676959e-07, "loss": 0.1839, "step": 9487 }, { "epoch": 2.524747205960617, "grad_norm": 0.27648240327835083, "learning_rate": 1.0163777240695781e-07, "loss": 0.2003, "step": 9488 }, { "epoch": 2.525013304949441, "grad_norm": 0.2659725844860077, "learning_rate": 1.0162071377947475e-07, "loss": 0.1811, "step": 9489 }, { "epoch": 2.525279403938265, "grad_norm": 0.2896175980567932, "learning_rate": 1.0160365510481688e-07, "loss": 0.182, "step": 9490 }, { "epoch": 2.525545502927089, "grad_norm": 0.27763786911964417, "learning_rate": 1.015865963834808e-07, "loss": 0.1828, "step": 9491 }, { "epoch": 2.5258116019159127, "grad_norm": 0.30146828293800354, "learning_rate": 1.0156953761596301e-07, "loss": 0.2044, "step": 9492 }, { "epoch": 2.526077700904737, "grad_norm": 0.2579682767391205, "learning_rate": 1.0155247880276005e-07, "loss": 0.1842, "step": 9493 }, { "epoch": 2.5263437998935605, "grad_norm": 0.3035089075565338, "learning_rate": 1.0153541994436847e-07, "loss": 0.1879, "step": 9494 }, { "epoch": 2.526609898882384, "grad_norm": 0.2890654504299164, "learning_rate": 1.0151836104128479e-07, "loss": 0.1953, "step": 9495 }, { "epoch": 2.5268759978712083, "grad_norm": 0.29952308535575867, "learning_rate": 1.0150130209400561e-07, "loss": 0.1928, "step": 9496 }, { "epoch": 2.527142096860032, "grad_norm": 0.286228209733963, "learning_rate": 1.0148424310302736e-07, "loss": 0.1944, "step": 9497 }, { "epoch": 2.5274081958488557, "grad_norm": 0.31014227867126465, "learning_rate": 1.0146718406884672e-07, "loss": 0.2042, "step": 9498 }, { "epoch": 2.52767429483768, "grad_norm": 0.3179403245449066, "learning_rate": 1.0145012499196011e-07, "loss": 0.203, "step": 9499 }, { "epoch": 2.5279403938265035, "grad_norm": 0.2820669114589691, "learning_rate": 1.0143306587286411e-07, "loss": 0.1828, "step": 9500 }, { "epoch": 2.528206492815327, "grad_norm": 0.2773382365703583, "learning_rate": 1.0141600671205533e-07, "loss": 0.1911, "step": 9501 }, { "epoch": 2.5284725918041513, "grad_norm": 0.31182190775871277, "learning_rate": 1.0139894751003024e-07, "loss": 0.1922, "step": 9502 }, { "epoch": 2.528738690792975, "grad_norm": 0.31462037563323975, "learning_rate": 1.0138188826728542e-07, "loss": 0.2134, "step": 9503 }, { "epoch": 2.5290047897817987, "grad_norm": 0.2835610806941986, "learning_rate": 1.0136482898431742e-07, "loss": 0.1945, "step": 9504 }, { "epoch": 2.529270888770623, "grad_norm": 0.30857113003730774, "learning_rate": 1.0134776966162282e-07, "loss": 0.2033, "step": 9505 }, { "epoch": 2.5295369877594465, "grad_norm": 0.29267534613609314, "learning_rate": 1.0133071029969813e-07, "loss": 0.1939, "step": 9506 }, { "epoch": 2.52980308674827, "grad_norm": 0.2987973690032959, "learning_rate": 1.0131365089903989e-07, "loss": 0.1816, "step": 9507 }, { "epoch": 2.5300691857370943, "grad_norm": 0.31265565752983093, "learning_rate": 1.0129659146014473e-07, "loss": 0.1894, "step": 9508 }, { "epoch": 2.530335284725918, "grad_norm": 0.2932252585887909, "learning_rate": 1.0127953198350914e-07, "loss": 0.1962, "step": 9509 }, { "epoch": 2.5306013837147416, "grad_norm": 0.2781011760234833, "learning_rate": 1.0126247246962972e-07, "loss": 0.2014, "step": 9510 }, { "epoch": 2.5308674827035658, "grad_norm": 0.2931744158267975, "learning_rate": 1.0124541291900294e-07, "loss": 0.2066, "step": 9511 }, { "epoch": 2.5311335816923894, "grad_norm": 0.3072233498096466, "learning_rate": 1.0122835333212546e-07, "loss": 0.1943, "step": 9512 }, { "epoch": 2.5313996806812136, "grad_norm": 0.30422481894493103, "learning_rate": 1.0121129370949381e-07, "loss": 0.1967, "step": 9513 }, { "epoch": 2.5316657796700373, "grad_norm": 0.32022449374198914, "learning_rate": 1.0119423405160457e-07, "loss": 0.2074, "step": 9514 }, { "epoch": 2.531931878658861, "grad_norm": 0.33717548847198486, "learning_rate": 1.0117717435895424e-07, "loss": 0.222, "step": 9515 }, { "epoch": 2.532197977647685, "grad_norm": 0.34961047768592834, "learning_rate": 1.0116011463203942e-07, "loss": 0.1947, "step": 9516 }, { "epoch": 2.5324640766365087, "grad_norm": 0.3629336655139923, "learning_rate": 1.0114305487135665e-07, "loss": 0.2052, "step": 9517 }, { "epoch": 2.532730175625333, "grad_norm": 0.30518120527267456, "learning_rate": 1.0112599507740258e-07, "loss": 0.2168, "step": 9518 }, { "epoch": 2.5329962746141566, "grad_norm": 0.28111353516578674, "learning_rate": 1.0110893525067372e-07, "loss": 0.1993, "step": 9519 }, { "epoch": 2.5332623736029802, "grad_norm": 0.2836512625217438, "learning_rate": 1.0109187539166659e-07, "loss": 0.1813, "step": 9520 }, { "epoch": 2.5335284725918044, "grad_norm": 0.2887375056743622, "learning_rate": 1.0107481550087782e-07, "loss": 0.1902, "step": 9521 }, { "epoch": 2.533794571580628, "grad_norm": 0.2899858355522156, "learning_rate": 1.0105775557880397e-07, "loss": 0.178, "step": 9522 }, { "epoch": 2.5340606705694517, "grad_norm": 0.31012779474258423, "learning_rate": 1.0104069562594158e-07, "loss": 0.2133, "step": 9523 }, { "epoch": 2.534326769558276, "grad_norm": 0.3097802996635437, "learning_rate": 1.0102363564278726e-07, "loss": 0.2211, "step": 9524 }, { "epoch": 2.5345928685470995, "grad_norm": 0.2979743480682373, "learning_rate": 1.0100657562983755e-07, "loss": 0.2064, "step": 9525 }, { "epoch": 2.534858967535923, "grad_norm": 0.29385340213775635, "learning_rate": 1.0098951558758904e-07, "loss": 0.1912, "step": 9526 }, { "epoch": 2.5351250665247473, "grad_norm": 0.30006715655326843, "learning_rate": 1.009724555165383e-07, "loss": 0.2123, "step": 9527 }, { "epoch": 2.535391165513571, "grad_norm": 0.27280673384666443, "learning_rate": 1.0095539541718192e-07, "loss": 0.1743, "step": 9528 }, { "epoch": 2.5356572645023947, "grad_norm": 0.2888484001159668, "learning_rate": 1.0093833529001647e-07, "loss": 0.1859, "step": 9529 }, { "epoch": 2.535923363491219, "grad_norm": 0.29527315497398376, "learning_rate": 1.0092127513553849e-07, "loss": 0.1865, "step": 9530 }, { "epoch": 2.5361894624800425, "grad_norm": 0.2827858328819275, "learning_rate": 1.0090421495424456e-07, "loss": 0.2025, "step": 9531 }, { "epoch": 2.536455561468866, "grad_norm": 0.28779157996177673, "learning_rate": 1.0088715474663132e-07, "loss": 0.1905, "step": 9532 }, { "epoch": 2.5367216604576903, "grad_norm": 0.2912345826625824, "learning_rate": 1.0087009451319532e-07, "loss": 0.1944, "step": 9533 }, { "epoch": 2.536987759446514, "grad_norm": 0.31403985619544983, "learning_rate": 1.008530342544331e-07, "loss": 0.2213, "step": 9534 }, { "epoch": 2.5372538584353377, "grad_norm": 0.2808779776096344, "learning_rate": 1.0083597397084128e-07, "loss": 0.1953, "step": 9535 }, { "epoch": 2.537519957424162, "grad_norm": 0.30451729893684387, "learning_rate": 1.008189136629164e-07, "loss": 0.2003, "step": 9536 }, { "epoch": 2.5377860564129855, "grad_norm": 0.3414231538772583, "learning_rate": 1.0080185333115512e-07, "loss": 0.2289, "step": 9537 }, { "epoch": 2.5380521554018096, "grad_norm": 0.31943801045417786, "learning_rate": 1.0078479297605393e-07, "loss": 0.2019, "step": 9538 }, { "epoch": 2.5383182543906333, "grad_norm": 0.3227391541004181, "learning_rate": 1.0076773259810949e-07, "loss": 0.1747, "step": 9539 }, { "epoch": 2.5385843533794574, "grad_norm": 0.32008031010627747, "learning_rate": 1.0075067219781833e-07, "loss": 0.1856, "step": 9540 }, { "epoch": 2.538850452368281, "grad_norm": 0.2679814100265503, "learning_rate": 1.0073361177567702e-07, "loss": 0.1826, "step": 9541 }, { "epoch": 2.539116551357105, "grad_norm": 0.3021682798862457, "learning_rate": 1.0071655133218222e-07, "loss": 0.1937, "step": 9542 }, { "epoch": 2.539382650345929, "grad_norm": 0.27975356578826904, "learning_rate": 1.0069949086783045e-07, "loss": 0.1944, "step": 9543 }, { "epoch": 2.5396487493347526, "grad_norm": 0.3105277717113495, "learning_rate": 1.0068243038311832e-07, "loss": 0.207, "step": 9544 }, { "epoch": 2.5399148483235763, "grad_norm": 0.3169531226158142, "learning_rate": 1.006653698785424e-07, "loss": 0.2013, "step": 9545 }, { "epoch": 2.5401809473124004, "grad_norm": 0.2736775577068329, "learning_rate": 1.0064830935459932e-07, "loss": 0.1936, "step": 9546 }, { "epoch": 2.540447046301224, "grad_norm": 0.29025155305862427, "learning_rate": 1.0063124881178561e-07, "loss": 0.2103, "step": 9547 }, { "epoch": 2.5407131452900478, "grad_norm": 0.307508260011673, "learning_rate": 1.0061418825059791e-07, "loss": 0.196, "step": 9548 }, { "epoch": 2.540979244278872, "grad_norm": 0.31263071298599243, "learning_rate": 1.0059712767153278e-07, "loss": 0.2113, "step": 9549 }, { "epoch": 2.5412453432676956, "grad_norm": 0.30936622619628906, "learning_rate": 1.005800670750868e-07, "loss": 0.2078, "step": 9550 }, { "epoch": 2.5415114422565193, "grad_norm": 0.2746991813182831, "learning_rate": 1.0056300646175659e-07, "loss": 0.1892, "step": 9551 }, { "epoch": 2.5417775412453434, "grad_norm": 0.3042471706867218, "learning_rate": 1.0054594583203872e-07, "loss": 0.2066, "step": 9552 }, { "epoch": 2.542043640234167, "grad_norm": 0.28037264943122864, "learning_rate": 1.0052888518642977e-07, "loss": 0.1877, "step": 9553 }, { "epoch": 2.5423097392229907, "grad_norm": 0.2761797308921814, "learning_rate": 1.0051182452542635e-07, "loss": 0.1922, "step": 9554 }, { "epoch": 2.542575838211815, "grad_norm": 0.37204232811927795, "learning_rate": 1.0049476384952507e-07, "loss": 0.2034, "step": 9555 }, { "epoch": 2.5428419372006386, "grad_norm": 0.37936490774154663, "learning_rate": 1.0047770315922246e-07, "loss": 0.2142, "step": 9556 }, { "epoch": 2.5431080361894622, "grad_norm": 0.2847452461719513, "learning_rate": 1.0046064245501518e-07, "loss": 0.1954, "step": 9557 }, { "epoch": 2.5433741351782864, "grad_norm": 0.26860612630844116, "learning_rate": 1.0044358173739978e-07, "loss": 0.2043, "step": 9558 }, { "epoch": 2.54364023416711, "grad_norm": 0.269957035779953, "learning_rate": 1.0042652100687288e-07, "loss": 0.1832, "step": 9559 }, { "epoch": 2.5439063331559337, "grad_norm": 0.264609158039093, "learning_rate": 1.0040946026393107e-07, "loss": 0.1788, "step": 9560 }, { "epoch": 2.544172432144758, "grad_norm": 0.33019670844078064, "learning_rate": 1.0039239950907093e-07, "loss": 0.2006, "step": 9561 }, { "epoch": 2.5444385311335815, "grad_norm": 0.2798272371292114, "learning_rate": 1.0037533874278903e-07, "loss": 0.1792, "step": 9562 }, { "epoch": 2.5447046301224057, "grad_norm": 0.3102656602859497, "learning_rate": 1.00358277965582e-07, "loss": 0.1934, "step": 9563 }, { "epoch": 2.5449707291112293, "grad_norm": 0.2784148156642914, "learning_rate": 1.0034121717794646e-07, "loss": 0.1913, "step": 9564 }, { "epoch": 2.5452368281000535, "grad_norm": 0.33264589309692383, "learning_rate": 1.0032415638037894e-07, "loss": 0.2159, "step": 9565 }, { "epoch": 2.545502927088877, "grad_norm": 0.2848680019378662, "learning_rate": 1.0030709557337608e-07, "loss": 0.1944, "step": 9566 }, { "epoch": 2.545769026077701, "grad_norm": 0.30196356773376465, "learning_rate": 1.0029003475743447e-07, "loss": 0.1869, "step": 9567 }, { "epoch": 2.546035125066525, "grad_norm": 0.2883496880531311, "learning_rate": 1.0027297393305071e-07, "loss": 0.1804, "step": 9568 }, { "epoch": 2.5463012240553486, "grad_norm": 0.31274861097335815, "learning_rate": 1.0025591310072139e-07, "loss": 0.2009, "step": 9569 }, { "epoch": 2.5465673230441723, "grad_norm": 0.292948842048645, "learning_rate": 1.002388522609431e-07, "loss": 0.194, "step": 9570 }, { "epoch": 2.5468334220329965, "grad_norm": 0.2952561378479004, "learning_rate": 1.0022179141421243e-07, "loss": 0.204, "step": 9571 }, { "epoch": 2.54709952102182, "grad_norm": 0.2733488976955414, "learning_rate": 1.0020473056102599e-07, "loss": 0.1887, "step": 9572 }, { "epoch": 2.547365620010644, "grad_norm": 0.2798174321651459, "learning_rate": 1.0018766970188039e-07, "loss": 0.1809, "step": 9573 }, { "epoch": 2.547631718999468, "grad_norm": 0.285706490278244, "learning_rate": 1.001706088372722e-07, "loss": 0.1974, "step": 9574 }, { "epoch": 2.5478978179882916, "grad_norm": 0.3030890226364136, "learning_rate": 1.0015354796769801e-07, "loss": 0.1979, "step": 9575 }, { "epoch": 2.5481639169771153, "grad_norm": 0.29094451665878296, "learning_rate": 1.0013648709365446e-07, "loss": 0.1903, "step": 9576 }, { "epoch": 2.5484300159659394, "grad_norm": 0.3098909556865692, "learning_rate": 1.0011942621563809e-07, "loss": 0.2125, "step": 9577 }, { "epoch": 2.548696114954763, "grad_norm": 0.2846946120262146, "learning_rate": 1.001023653341456e-07, "loss": 0.2048, "step": 9578 }, { "epoch": 2.548962213943587, "grad_norm": 0.2795884311199188, "learning_rate": 1.0008530444967346e-07, "loss": 0.199, "step": 9579 }, { "epoch": 2.549228312932411, "grad_norm": 0.2933695912361145, "learning_rate": 1.0006824356271837e-07, "loss": 0.1991, "step": 9580 }, { "epoch": 2.5494944119212346, "grad_norm": 0.31685084104537964, "learning_rate": 1.0005118267377688e-07, "loss": 0.2035, "step": 9581 }, { "epoch": 2.5497605109100583, "grad_norm": 0.29943519830703735, "learning_rate": 1.0003412178334561e-07, "loss": 0.201, "step": 9582 }, { "epoch": 2.5500266098988824, "grad_norm": 0.2950954735279083, "learning_rate": 1.000170608919211e-07, "loss": 0.2179, "step": 9583 }, { "epoch": 2.550292708887706, "grad_norm": 0.32464078068733215, "learning_rate": 1e-07, "loss": 0.234, "step": 9584 }, { "epoch": 2.55055880787653, "grad_norm": 0.28618428111076355, "learning_rate": 9.99829391080789e-08, "loss": 0.1898, "step": 9585 }, { "epoch": 2.550824906865354, "grad_norm": 0.2779059410095215, "learning_rate": 9.996587821665441e-08, "loss": 0.1812, "step": 9586 }, { "epoch": 2.551091005854178, "grad_norm": 0.2962338328361511, "learning_rate": 9.994881732622311e-08, "loss": 0.1936, "step": 9587 }, { "epoch": 2.5513571048430017, "grad_norm": 0.3080523908138275, "learning_rate": 9.993175643728162e-08, "loss": 0.1979, "step": 9588 }, { "epoch": 2.5516232038318254, "grad_norm": 0.27314725518226624, "learning_rate": 9.991469555032653e-08, "loss": 0.1876, "step": 9589 }, { "epoch": 2.5518893028206495, "grad_norm": 0.30793553590774536, "learning_rate": 9.989763466585441e-08, "loss": 0.2232, "step": 9590 }, { "epoch": 2.552155401809473, "grad_norm": 0.28656575083732605, "learning_rate": 9.98805737843619e-08, "loss": 0.2107, "step": 9591 }, { "epoch": 2.552421500798297, "grad_norm": 0.28299930691719055, "learning_rate": 9.986351290634556e-08, "loss": 0.1852, "step": 9592 }, { "epoch": 2.552687599787121, "grad_norm": 0.2801763713359833, "learning_rate": 9.9846452032302e-08, "loss": 0.2045, "step": 9593 }, { "epoch": 2.5529536987759447, "grad_norm": 0.28641489148139954, "learning_rate": 9.982939116272782e-08, "loss": 0.1844, "step": 9594 }, { "epoch": 2.5532197977647684, "grad_norm": 0.3002224564552307, "learning_rate": 9.981233029811961e-08, "loss": 0.2129, "step": 9595 }, { "epoch": 2.5534858967535925, "grad_norm": 0.3064732253551483, "learning_rate": 9.979526943897399e-08, "loss": 0.2113, "step": 9596 }, { "epoch": 2.553751995742416, "grad_norm": 0.30583178997039795, "learning_rate": 9.977820858578755e-08, "loss": 0.2239, "step": 9597 }, { "epoch": 2.55401809473124, "grad_norm": 0.30781012773513794, "learning_rate": 9.976114773905693e-08, "loss": 0.1969, "step": 9598 }, { "epoch": 2.554284193720064, "grad_norm": 0.2903083264827728, "learning_rate": 9.974408689927862e-08, "loss": 0.1821, "step": 9599 }, { "epoch": 2.5545502927088877, "grad_norm": 0.29900965094566345, "learning_rate": 9.972702606694928e-08, "loss": 0.1913, "step": 9600 }, { "epoch": 2.5548163916977114, "grad_norm": 0.2840671241283417, "learning_rate": 9.970996524256553e-08, "loss": 0.1867, "step": 9601 }, { "epoch": 2.5550824906865355, "grad_norm": 0.33872732520103455, "learning_rate": 9.969290442662392e-08, "loss": 0.2269, "step": 9602 }, { "epoch": 2.555348589675359, "grad_norm": 0.30907902121543884, "learning_rate": 9.967584361962105e-08, "loss": 0.2096, "step": 9603 }, { "epoch": 2.555614688664183, "grad_norm": 0.2965537905693054, "learning_rate": 9.965878282205354e-08, "loss": 0.1948, "step": 9604 }, { "epoch": 2.555880787653007, "grad_norm": 0.3246053457260132, "learning_rate": 9.964172203441798e-08, "loss": 0.2006, "step": 9605 }, { "epoch": 2.5561468866418307, "grad_norm": 0.30854812264442444, "learning_rate": 9.962466125721095e-08, "loss": 0.195, "step": 9606 }, { "epoch": 2.5564129856306543, "grad_norm": 0.301644891500473, "learning_rate": 9.96076004909291e-08, "loss": 0.1897, "step": 9607 }, { "epoch": 2.5566790846194785, "grad_norm": 0.3016291856765747, "learning_rate": 9.959053973606895e-08, "loss": 0.2105, "step": 9608 }, { "epoch": 2.556945183608302, "grad_norm": 0.2923157513141632, "learning_rate": 9.957347899312711e-08, "loss": 0.209, "step": 9609 }, { "epoch": 2.5572112825971263, "grad_norm": 0.29911401867866516, "learning_rate": 9.955641826260023e-08, "loss": 0.1783, "step": 9610 }, { "epoch": 2.55747738158595, "grad_norm": 0.2888638377189636, "learning_rate": 9.953935754498482e-08, "loss": 0.1866, "step": 9611 }, { "epoch": 2.557743480574774, "grad_norm": 0.320886492729187, "learning_rate": 9.952229684077753e-08, "loss": 0.185, "step": 9612 }, { "epoch": 2.5580095795635978, "grad_norm": 0.3035871088504791, "learning_rate": 9.950523615047494e-08, "loss": 0.1889, "step": 9613 }, { "epoch": 2.5582756785524214, "grad_norm": 0.29292261600494385, "learning_rate": 9.948817547457363e-08, "loss": 0.2156, "step": 9614 }, { "epoch": 2.5585417775412456, "grad_norm": 0.33630630373954773, "learning_rate": 9.947111481357022e-08, "loss": 0.1993, "step": 9615 }, { "epoch": 2.5588078765300692, "grad_norm": 0.3078562915325165, "learning_rate": 9.945405416796131e-08, "loss": 0.1973, "step": 9616 }, { "epoch": 2.559073975518893, "grad_norm": 0.2866417467594147, "learning_rate": 9.943699353824343e-08, "loss": 0.1827, "step": 9617 }, { "epoch": 2.559340074507717, "grad_norm": 0.281403124332428, "learning_rate": 9.941993292491321e-08, "loss": 0.1893, "step": 9618 }, { "epoch": 2.5596061734965407, "grad_norm": 0.3117430806159973, "learning_rate": 9.940287232846722e-08, "loss": 0.2131, "step": 9619 }, { "epoch": 2.5598722724853644, "grad_norm": 0.2757708728313446, "learning_rate": 9.938581174940211e-08, "loss": 0.1894, "step": 9620 }, { "epoch": 2.5601383714741885, "grad_norm": 0.28719839453697205, "learning_rate": 9.936875118821438e-08, "loss": 0.1868, "step": 9621 }, { "epoch": 2.5604044704630122, "grad_norm": 0.2795981764793396, "learning_rate": 9.935169064540069e-08, "loss": 0.1668, "step": 9622 }, { "epoch": 2.560670569451836, "grad_norm": 0.29977717995643616, "learning_rate": 9.933463012145758e-08, "loss": 0.2054, "step": 9623 }, { "epoch": 2.56093666844066, "grad_norm": 0.27536624670028687, "learning_rate": 9.931756961688167e-08, "loss": 0.1882, "step": 9624 }, { "epoch": 2.5612027674294837, "grad_norm": 0.2819059491157532, "learning_rate": 9.930050913216957e-08, "loss": 0.1961, "step": 9625 }, { "epoch": 2.5614688664183074, "grad_norm": 0.2878499925136566, "learning_rate": 9.928344866781781e-08, "loss": 0.1875, "step": 9626 }, { "epoch": 2.5617349654071315, "grad_norm": 0.3104204833507538, "learning_rate": 9.926638822432298e-08, "loss": 0.2072, "step": 9627 }, { "epoch": 2.562001064395955, "grad_norm": 0.307248055934906, "learning_rate": 9.924932780218169e-08, "loss": 0.2049, "step": 9628 }, { "epoch": 2.562267163384779, "grad_norm": 0.27166393399238586, "learning_rate": 9.923226740189052e-08, "loss": 0.1801, "step": 9629 }, { "epoch": 2.562533262373603, "grad_norm": 0.28341206908226013, "learning_rate": 9.921520702394607e-08, "loss": 0.1909, "step": 9630 }, { "epoch": 2.5627993613624267, "grad_norm": 0.2808336913585663, "learning_rate": 9.919814666884489e-08, "loss": 0.1958, "step": 9631 }, { "epoch": 2.563065460351251, "grad_norm": 0.3346012830734253, "learning_rate": 9.918108633708358e-08, "loss": 0.2053, "step": 9632 }, { "epoch": 2.5633315593400745, "grad_norm": 0.31265467405319214, "learning_rate": 9.916402602915871e-08, "loss": 0.217, "step": 9633 }, { "epoch": 2.563597658328898, "grad_norm": 0.2868967056274414, "learning_rate": 9.914696574556692e-08, "loss": 0.188, "step": 9634 }, { "epoch": 2.5638637573177223, "grad_norm": 0.289938360452652, "learning_rate": 9.912990548680471e-08, "loss": 0.2051, "step": 9635 }, { "epoch": 2.564129856306546, "grad_norm": 0.3121793270111084, "learning_rate": 9.911284525336868e-08, "loss": 0.1904, "step": 9636 }, { "epoch": 2.56439595529537, "grad_norm": 0.30123740434646606, "learning_rate": 9.909578504575543e-08, "loss": 0.2066, "step": 9637 }, { "epoch": 2.564662054284194, "grad_norm": 0.2938154637813568, "learning_rate": 9.907872486446151e-08, "loss": 0.1937, "step": 9638 }, { "epoch": 2.5649281532730175, "grad_norm": 0.30058297514915466, "learning_rate": 9.906166470998354e-08, "loss": 0.2082, "step": 9639 }, { "epoch": 2.5651942522618416, "grad_norm": 0.2999351918697357, "learning_rate": 9.904460458281808e-08, "loss": 0.2033, "step": 9640 }, { "epoch": 2.5654603512506653, "grad_norm": 0.29175397753715515, "learning_rate": 9.902754448346169e-08, "loss": 0.2039, "step": 9641 }, { "epoch": 2.565726450239489, "grad_norm": 0.2754492461681366, "learning_rate": 9.901048441241094e-08, "loss": 0.188, "step": 9642 }, { "epoch": 2.565992549228313, "grad_norm": 0.273639440536499, "learning_rate": 9.899342437016248e-08, "loss": 0.1848, "step": 9643 }, { "epoch": 2.566258648217137, "grad_norm": 0.2819407880306244, "learning_rate": 9.897636435721277e-08, "loss": 0.1849, "step": 9644 }, { "epoch": 2.5665247472059605, "grad_norm": 0.29118311405181885, "learning_rate": 9.895930437405844e-08, "loss": 0.1885, "step": 9645 }, { "epoch": 2.5667908461947846, "grad_norm": 0.2847785949707031, "learning_rate": 9.894224442119605e-08, "loss": 0.1901, "step": 9646 }, { "epoch": 2.5670569451836083, "grad_norm": 0.2792132496833801, "learning_rate": 9.892518449912218e-08, "loss": 0.1924, "step": 9647 }, { "epoch": 2.567323044172432, "grad_norm": 0.3070397675037384, "learning_rate": 9.890812460833341e-08, "loss": 0.1968, "step": 9648 }, { "epoch": 2.567589143161256, "grad_norm": 0.28208279609680176, "learning_rate": 9.88910647493263e-08, "loss": 0.2076, "step": 9649 }, { "epoch": 2.5678552421500798, "grad_norm": 0.2915283441543579, "learning_rate": 9.887400492259741e-08, "loss": 0.1915, "step": 9650 }, { "epoch": 2.5681213411389034, "grad_norm": 0.3031207025051117, "learning_rate": 9.885694512864331e-08, "loss": 0.1883, "step": 9651 }, { "epoch": 2.5683874401277276, "grad_norm": 0.31458863615989685, "learning_rate": 9.88398853679606e-08, "loss": 0.2092, "step": 9652 }, { "epoch": 2.5686535391165513, "grad_norm": 0.2948029339313507, "learning_rate": 9.88228256410458e-08, "loss": 0.1841, "step": 9653 }, { "epoch": 2.568919638105375, "grad_norm": 0.3206994831562042, "learning_rate": 9.880576594839547e-08, "loss": 0.1886, "step": 9654 }, { "epoch": 2.569185737094199, "grad_norm": 0.28807294368743896, "learning_rate": 9.878870629050619e-08, "loss": 0.1869, "step": 9655 }, { "epoch": 2.5694518360830227, "grad_norm": 0.3197080194950104, "learning_rate": 9.877164666787453e-08, "loss": 0.2017, "step": 9656 }, { "epoch": 2.569717935071847, "grad_norm": 0.2978452742099762, "learning_rate": 9.875458708099705e-08, "loss": 0.187, "step": 9657 }, { "epoch": 2.5699840340606706, "grad_norm": 0.2968303859233856, "learning_rate": 9.87375275303703e-08, "loss": 0.2066, "step": 9658 }, { "epoch": 2.5702501330494947, "grad_norm": 0.3233042061328888, "learning_rate": 9.872046801649086e-08, "loss": 0.1983, "step": 9659 }, { "epoch": 2.5705162320383184, "grad_norm": 0.3179689347743988, "learning_rate": 9.870340853985528e-08, "loss": 0.1937, "step": 9660 }, { "epoch": 2.570782331027142, "grad_norm": 0.32380905747413635, "learning_rate": 9.86863491009601e-08, "loss": 0.1979, "step": 9661 }, { "epoch": 2.571048430015966, "grad_norm": 0.2585102617740631, "learning_rate": 9.86692897003019e-08, "loss": 0.1676, "step": 9662 }, { "epoch": 2.57131452900479, "grad_norm": 0.28821712732315063, "learning_rate": 9.865223033837721e-08, "loss": 0.194, "step": 9663 }, { "epoch": 2.5715806279936135, "grad_norm": 0.3026398718357086, "learning_rate": 9.863517101568257e-08, "loss": 0.2049, "step": 9664 }, { "epoch": 2.5718467269824377, "grad_norm": 0.2957417368888855, "learning_rate": 9.861811173271458e-08, "loss": 0.1961, "step": 9665 }, { "epoch": 2.5721128259712613, "grad_norm": 0.29776889085769653, "learning_rate": 9.860105248996976e-08, "loss": 0.2181, "step": 9666 }, { "epoch": 2.572378924960085, "grad_norm": 0.2983642816543579, "learning_rate": 9.858399328794467e-08, "loss": 0.2103, "step": 9667 }, { "epoch": 2.572645023948909, "grad_norm": 0.29384273290634155, "learning_rate": 9.856693412713587e-08, "loss": 0.1965, "step": 9668 }, { "epoch": 2.572911122937733, "grad_norm": 0.2952907979488373, "learning_rate": 9.85498750080399e-08, "loss": 0.1834, "step": 9669 }, { "epoch": 2.5731772219265565, "grad_norm": 0.2886447608470917, "learning_rate": 9.853281593115333e-08, "loss": 0.1934, "step": 9670 }, { "epoch": 2.5734433209153806, "grad_norm": 0.3056894540786743, "learning_rate": 9.851575689697263e-08, "loss": 0.2103, "step": 9671 }, { "epoch": 2.5737094199042043, "grad_norm": 0.31173357367515564, "learning_rate": 9.849869790599442e-08, "loss": 0.2135, "step": 9672 }, { "epoch": 2.573975518893028, "grad_norm": 0.2904723584651947, "learning_rate": 9.848163895871521e-08, "loss": 0.2012, "step": 9673 }, { "epoch": 2.574241617881852, "grad_norm": 0.2792814373970032, "learning_rate": 9.846458005563154e-08, "loss": 0.1831, "step": 9674 }, { "epoch": 2.574507716870676, "grad_norm": 0.3011017143726349, "learning_rate": 9.844752119723995e-08, "loss": 0.2129, "step": 9675 }, { "epoch": 2.5747738158594995, "grad_norm": 0.30324211716651917, "learning_rate": 9.843046238403698e-08, "loss": 0.226, "step": 9676 }, { "epoch": 2.5750399148483236, "grad_norm": 0.3117813169956207, "learning_rate": 9.841340361651919e-08, "loss": 0.1768, "step": 9677 }, { "epoch": 2.5753060138371473, "grad_norm": 0.3157370686531067, "learning_rate": 9.83963448951831e-08, "loss": 0.2052, "step": 9678 }, { "epoch": 2.5755721128259714, "grad_norm": 0.2801477313041687, "learning_rate": 9.83792862205253e-08, "loss": 0.1843, "step": 9679 }, { "epoch": 2.575838211814795, "grad_norm": 0.28519704937934875, "learning_rate": 9.83622275930422e-08, "loss": 0.1808, "step": 9680 }, { "epoch": 2.576104310803619, "grad_norm": 0.2980009913444519, "learning_rate": 9.83451690132304e-08, "loss": 0.2021, "step": 9681 }, { "epoch": 2.576370409792443, "grad_norm": 0.30016136169433594, "learning_rate": 9.832811048158649e-08, "loss": 0.189, "step": 9682 }, { "epoch": 2.5766365087812666, "grad_norm": 0.32399794459342957, "learning_rate": 9.83110519986069e-08, "loss": 0.2033, "step": 9683 }, { "epoch": 2.5769026077700907, "grad_norm": 0.2960391044616699, "learning_rate": 9.82939935647882e-08, "loss": 0.1942, "step": 9684 }, { "epoch": 2.5771687067589144, "grad_norm": 0.2890849709510803, "learning_rate": 9.827693518062692e-08, "loss": 0.1923, "step": 9685 }, { "epoch": 2.577434805747738, "grad_norm": 0.2858369052410126, "learning_rate": 9.825987684661958e-08, "loss": 0.1993, "step": 9686 }, { "epoch": 2.577700904736562, "grad_norm": 0.2795419692993164, "learning_rate": 9.824281856326273e-08, "loss": 0.1816, "step": 9687 }, { "epoch": 2.577967003725386, "grad_norm": 0.2778724133968353, "learning_rate": 9.822576033105287e-08, "loss": 0.181, "step": 9688 }, { "epoch": 2.5782331027142096, "grad_norm": 0.3092010021209717, "learning_rate": 9.820870215048651e-08, "loss": 0.187, "step": 9689 }, { "epoch": 2.5784992017030337, "grad_norm": 0.2865777313709259, "learning_rate": 9.819164402206017e-08, "loss": 0.1985, "step": 9690 }, { "epoch": 2.5787653006918574, "grad_norm": 0.28376007080078125, "learning_rate": 9.817458594627037e-08, "loss": 0.185, "step": 9691 }, { "epoch": 2.579031399680681, "grad_norm": 0.33042359352111816, "learning_rate": 9.815752792361368e-08, "loss": 0.2088, "step": 9692 }, { "epoch": 2.579297498669505, "grad_norm": 0.2866283655166626, "learning_rate": 9.814046995458652e-08, "loss": 0.2067, "step": 9693 }, { "epoch": 2.579563597658329, "grad_norm": 0.27605563402175903, "learning_rate": 9.812341203968545e-08, "loss": 0.1809, "step": 9694 }, { "epoch": 2.5798296966471526, "grad_norm": 0.2853893041610718, "learning_rate": 9.8106354179407e-08, "loss": 0.2062, "step": 9695 }, { "epoch": 2.5800957956359767, "grad_norm": 0.2712154686450958, "learning_rate": 9.808929637424765e-08, "loss": 0.1712, "step": 9696 }, { "epoch": 2.5803618946248004, "grad_norm": 0.295810341835022, "learning_rate": 9.807223862470397e-08, "loss": 0.1916, "step": 9697 }, { "epoch": 2.580627993613624, "grad_norm": 0.30287081003189087, "learning_rate": 9.805518093127238e-08, "loss": 0.1971, "step": 9698 }, { "epoch": 2.580894092602448, "grad_norm": 0.28582999110221863, "learning_rate": 9.803812329444942e-08, "loss": 0.1884, "step": 9699 }, { "epoch": 2.581160191591272, "grad_norm": 0.28071218729019165, "learning_rate": 9.802106571473159e-08, "loss": 0.1859, "step": 9700 }, { "epoch": 2.5814262905800955, "grad_norm": 0.2882806658744812, "learning_rate": 9.800400819261541e-08, "loss": 0.2115, "step": 9701 }, { "epoch": 2.5816923895689197, "grad_norm": 0.3175196945667267, "learning_rate": 9.798695072859738e-08, "loss": 0.2051, "step": 9702 }, { "epoch": 2.5819584885577433, "grad_norm": 0.29229775071144104, "learning_rate": 9.796989332317397e-08, "loss": 0.1947, "step": 9703 }, { "epoch": 2.5822245875465675, "grad_norm": 0.2633800208568573, "learning_rate": 9.79528359768417e-08, "loss": 0.1738, "step": 9704 }, { "epoch": 2.582490686535391, "grad_norm": 0.2795445919036865, "learning_rate": 9.793577869009705e-08, "loss": 0.1879, "step": 9705 }, { "epoch": 2.5827567855242153, "grad_norm": 0.28161492943763733, "learning_rate": 9.791872146343657e-08, "loss": 0.2013, "step": 9706 }, { "epoch": 2.583022884513039, "grad_norm": 0.28678613901138306, "learning_rate": 9.790166429735666e-08, "loss": 0.1961, "step": 9707 }, { "epoch": 2.5832889835018626, "grad_norm": 0.29130253195762634, "learning_rate": 9.788460719235385e-08, "loss": 0.1737, "step": 9708 }, { "epoch": 2.5835550824906868, "grad_norm": 0.2688201665878296, "learning_rate": 9.786755014892464e-08, "loss": 0.1804, "step": 9709 }, { "epoch": 2.5838211814795105, "grad_norm": 0.29276761412620544, "learning_rate": 9.78504931675655e-08, "loss": 0.214, "step": 9710 }, { "epoch": 2.584087280468334, "grad_norm": 0.2810070514678955, "learning_rate": 9.783343624877296e-08, "loss": 0.1847, "step": 9711 }, { "epoch": 2.5843533794571583, "grad_norm": 0.33249008655548096, "learning_rate": 9.781637939304344e-08, "loss": 0.215, "step": 9712 }, { "epoch": 2.584619478445982, "grad_norm": 0.2943347990512848, "learning_rate": 9.779932260087344e-08, "loss": 0.1872, "step": 9713 }, { "epoch": 2.5848855774348056, "grad_norm": 0.3122358024120331, "learning_rate": 9.778226587275945e-08, "loss": 0.186, "step": 9714 }, { "epoch": 2.5851516764236298, "grad_norm": 0.2909439504146576, "learning_rate": 9.776520920919791e-08, "loss": 0.1948, "step": 9715 }, { "epoch": 2.5854177754124534, "grad_norm": 0.2757428288459778, "learning_rate": 9.77481526106854e-08, "loss": 0.1921, "step": 9716 }, { "epoch": 2.585683874401277, "grad_norm": 0.30277755856513977, "learning_rate": 9.773109607771828e-08, "loss": 0.1939, "step": 9717 }, { "epoch": 2.5859499733901012, "grad_norm": 0.2857336401939392, "learning_rate": 9.771403961079307e-08, "loss": 0.195, "step": 9718 }, { "epoch": 2.586216072378925, "grad_norm": 0.2800416052341461, "learning_rate": 9.769698321040621e-08, "loss": 0.208, "step": 9719 }, { "epoch": 2.5864821713677486, "grad_norm": 0.3102278709411621, "learning_rate": 9.76799268770542e-08, "loss": 0.2137, "step": 9720 }, { "epoch": 2.5867482703565727, "grad_norm": 0.29376623034477234, "learning_rate": 9.766287061123352e-08, "loss": 0.1928, "step": 9721 }, { "epoch": 2.5870143693453964, "grad_norm": 0.29465481638908386, "learning_rate": 9.764581441344058e-08, "loss": 0.1915, "step": 9722 }, { "epoch": 2.58728046833422, "grad_norm": 0.28693222999572754, "learning_rate": 9.762875828417187e-08, "loss": 0.1978, "step": 9723 }, { "epoch": 2.5875465673230442, "grad_norm": 0.2968639135360718, "learning_rate": 9.761170222392385e-08, "loss": 0.1963, "step": 9724 }, { "epoch": 2.587812666311868, "grad_norm": 0.3202323615550995, "learning_rate": 9.759464623319301e-08, "loss": 0.2032, "step": 9725 }, { "epoch": 2.5880787653006916, "grad_norm": 0.3329940736293793, "learning_rate": 9.757759031247575e-08, "loss": 0.1907, "step": 9726 }, { "epoch": 2.5883448642895157, "grad_norm": 0.2790798544883728, "learning_rate": 9.756053446226853e-08, "loss": 0.1838, "step": 9727 }, { "epoch": 2.5886109632783394, "grad_norm": 0.28474128246307373, "learning_rate": 9.754347868306783e-08, "loss": 0.2017, "step": 9728 }, { "epoch": 2.5888770622671635, "grad_norm": 0.30203118920326233, "learning_rate": 9.752642297537008e-08, "loss": 0.2121, "step": 9729 }, { "epoch": 2.589143161255987, "grad_norm": 0.27286630868911743, "learning_rate": 9.750936733967172e-08, "loss": 0.1854, "step": 9730 }, { "epoch": 2.5894092602448113, "grad_norm": 0.2793647348880768, "learning_rate": 9.749231177646925e-08, "loss": 0.1873, "step": 9731 }, { "epoch": 2.589675359233635, "grad_norm": 0.326133131980896, "learning_rate": 9.747525628625905e-08, "loss": 0.2004, "step": 9732 }, { "epoch": 2.5899414582224587, "grad_norm": 0.3018431067466736, "learning_rate": 9.745820086953759e-08, "loss": 0.204, "step": 9733 }, { "epoch": 2.590207557211283, "grad_norm": 0.31153249740600586, "learning_rate": 9.744114552680134e-08, "loss": 0.195, "step": 9734 }, { "epoch": 2.5904736562001065, "grad_norm": 0.2985233664512634, "learning_rate": 9.742409025854664e-08, "loss": 0.1871, "step": 9735 }, { "epoch": 2.59073975518893, "grad_norm": 0.29788094758987427, "learning_rate": 9.740703506527e-08, "loss": 0.1963, "step": 9736 }, { "epoch": 2.5910058541777543, "grad_norm": 0.27613648772239685, "learning_rate": 9.738997994746784e-08, "loss": 0.1843, "step": 9737 }, { "epoch": 2.591271953166578, "grad_norm": 0.26810020208358765, "learning_rate": 9.737292490563659e-08, "loss": 0.1806, "step": 9738 }, { "epoch": 2.5915380521554017, "grad_norm": 0.26596975326538086, "learning_rate": 9.735586994027266e-08, "loss": 0.1994, "step": 9739 }, { "epoch": 2.591804151144226, "grad_norm": 0.2722513675689697, "learning_rate": 9.733881505187251e-08, "loss": 0.169, "step": 9740 }, { "epoch": 2.5920702501330495, "grad_norm": 0.2884880006313324, "learning_rate": 9.732176024093255e-08, "loss": 0.1694, "step": 9741 }, { "epoch": 2.592336349121873, "grad_norm": 0.28923383355140686, "learning_rate": 9.730470550794917e-08, "loss": 0.1992, "step": 9742 }, { "epoch": 2.5926024481106973, "grad_norm": 0.2958422303199768, "learning_rate": 9.728765085341884e-08, "loss": 0.2004, "step": 9743 }, { "epoch": 2.592868547099521, "grad_norm": 0.26037460565567017, "learning_rate": 9.727059627783797e-08, "loss": 0.1722, "step": 9744 }, { "epoch": 2.5931346460883447, "grad_norm": 0.26531586050987244, "learning_rate": 9.72535417817029e-08, "loss": 0.1827, "step": 9745 }, { "epoch": 2.593400745077169, "grad_norm": 0.3036157488822937, "learning_rate": 9.723648736551014e-08, "loss": 0.2075, "step": 9746 }, { "epoch": 2.5936668440659925, "grad_norm": 0.2910501956939697, "learning_rate": 9.721943302975603e-08, "loss": 0.207, "step": 9747 }, { "epoch": 2.593932943054816, "grad_norm": 0.29981762170791626, "learning_rate": 9.7202378774937e-08, "loss": 0.1996, "step": 9748 }, { "epoch": 2.5941990420436403, "grad_norm": 0.2704954743385315, "learning_rate": 9.718532460154946e-08, "loss": 0.1931, "step": 9749 }, { "epoch": 2.594465141032464, "grad_norm": 0.2939280867576599, "learning_rate": 9.716827051008983e-08, "loss": 0.1924, "step": 9750 }, { "epoch": 2.594731240021288, "grad_norm": 0.27017635107040405, "learning_rate": 9.715121650105452e-08, "loss": 0.1787, "step": 9751 }, { "epoch": 2.5949973390101118, "grad_norm": 0.3278868794441223, "learning_rate": 9.713416257493987e-08, "loss": 0.21, "step": 9752 }, { "epoch": 2.5952634379989354, "grad_norm": 0.2893168032169342, "learning_rate": 9.711710873224233e-08, "loss": 0.1761, "step": 9753 }, { "epoch": 2.5955295369877596, "grad_norm": 0.2899131774902344, "learning_rate": 9.710005497345824e-08, "loss": 0.1874, "step": 9754 }, { "epoch": 2.5957956359765832, "grad_norm": 0.3006325960159302, "learning_rate": 9.708300129908403e-08, "loss": 0.1888, "step": 9755 }, { "epoch": 2.5960617349654074, "grad_norm": 0.2805866003036499, "learning_rate": 9.706594770961606e-08, "loss": 0.1966, "step": 9756 }, { "epoch": 2.596327833954231, "grad_norm": 0.3142143487930298, "learning_rate": 9.704889420555075e-08, "loss": 0.2013, "step": 9757 }, { "epoch": 2.5965939329430547, "grad_norm": 0.30812907218933105, "learning_rate": 9.703184078738446e-08, "loss": 0.2033, "step": 9758 }, { "epoch": 2.596860031931879, "grad_norm": 0.28511518239974976, "learning_rate": 9.701478745561356e-08, "loss": 0.1831, "step": 9759 }, { "epoch": 2.5971261309207025, "grad_norm": 0.2701876163482666, "learning_rate": 9.69977342107345e-08, "loss": 0.1709, "step": 9760 }, { "epoch": 2.5973922299095262, "grad_norm": 0.3172217905521393, "learning_rate": 9.698068105324359e-08, "loss": 0.2045, "step": 9761 }, { "epoch": 2.5976583288983504, "grad_norm": 0.27630239725112915, "learning_rate": 9.696362798363718e-08, "loss": 0.1894, "step": 9762 }, { "epoch": 2.597924427887174, "grad_norm": 0.327571839094162, "learning_rate": 9.694657500241171e-08, "loss": 0.2169, "step": 9763 }, { "epoch": 2.5981905268759977, "grad_norm": 0.2944542169570923, "learning_rate": 9.69295221100635e-08, "loss": 0.1985, "step": 9764 }, { "epoch": 2.598456625864822, "grad_norm": 0.2715780735015869, "learning_rate": 9.691246930708892e-08, "loss": 0.1741, "step": 9765 }, { "epoch": 2.5987227248536455, "grad_norm": 0.29791539907455444, "learning_rate": 9.689541659398434e-08, "loss": 0.2058, "step": 9766 }, { "epoch": 2.598988823842469, "grad_norm": 0.2964821755886078, "learning_rate": 9.687836397124613e-08, "loss": 0.2031, "step": 9767 }, { "epoch": 2.5992549228312933, "grad_norm": 0.2848888337612152, "learning_rate": 9.686131143937061e-08, "loss": 0.1834, "step": 9768 }, { "epoch": 2.599521021820117, "grad_norm": 0.2848014831542969, "learning_rate": 9.68442589988542e-08, "loss": 0.1791, "step": 9769 }, { "epoch": 2.5997871208089407, "grad_norm": 0.3160499334335327, "learning_rate": 9.682720665019325e-08, "loss": 0.1994, "step": 9770 }, { "epoch": 2.600053219797765, "grad_norm": 0.3041573762893677, "learning_rate": 9.681015439388403e-08, "loss": 0.2079, "step": 9771 }, { "epoch": 2.6003193187865885, "grad_norm": 0.29494917392730713, "learning_rate": 9.679310223042292e-08, "loss": 0.214, "step": 9772 }, { "epoch": 2.600585417775412, "grad_norm": 0.29662078619003296, "learning_rate": 9.677605016030632e-08, "loss": 0.2135, "step": 9773 }, { "epoch": 2.6008515167642363, "grad_norm": 0.27752313017845154, "learning_rate": 9.67589981840305e-08, "loss": 0.1892, "step": 9774 }, { "epoch": 2.60111761575306, "grad_norm": 0.3053909242153168, "learning_rate": 9.674194630209184e-08, "loss": 0.203, "step": 9775 }, { "epoch": 2.601383714741884, "grad_norm": 0.2877858579158783, "learning_rate": 9.672489451498665e-08, "loss": 0.197, "step": 9776 }, { "epoch": 2.601649813730708, "grad_norm": 0.32678818702697754, "learning_rate": 9.670784282321127e-08, "loss": 0.2179, "step": 9777 }, { "epoch": 2.601915912719532, "grad_norm": 0.2864231467247009, "learning_rate": 9.669079122726205e-08, "loss": 0.2072, "step": 9778 }, { "epoch": 2.6021820117083556, "grad_norm": 0.25913459062576294, "learning_rate": 9.667373972763534e-08, "loss": 0.1712, "step": 9779 }, { "epoch": 2.6024481106971793, "grad_norm": 0.32107600569725037, "learning_rate": 9.665668832482739e-08, "loss": 0.2018, "step": 9780 }, { "epoch": 2.6027142096860034, "grad_norm": 0.2937539517879486, "learning_rate": 9.663963701933456e-08, "loss": 0.1941, "step": 9781 }, { "epoch": 2.602980308674827, "grad_norm": 0.3276617228984833, "learning_rate": 9.662258581165318e-08, "loss": 0.2089, "step": 9782 }, { "epoch": 2.603246407663651, "grad_norm": 0.2838302254676819, "learning_rate": 9.660553470227957e-08, "loss": 0.1784, "step": 9783 }, { "epoch": 2.603512506652475, "grad_norm": 0.25705230236053467, "learning_rate": 9.658848369171002e-08, "loss": 0.1841, "step": 9784 }, { "epoch": 2.6037786056412986, "grad_norm": 0.2955755293369293, "learning_rate": 9.657143278044085e-08, "loss": 0.1921, "step": 9785 }, { "epoch": 2.6040447046301223, "grad_norm": 0.3040565550327301, "learning_rate": 9.655438196896835e-08, "loss": 0.1986, "step": 9786 }, { "epoch": 2.6043108036189464, "grad_norm": 0.30737370252609253, "learning_rate": 9.653733125778886e-08, "loss": 0.1883, "step": 9787 }, { "epoch": 2.60457690260777, "grad_norm": 0.32701149582862854, "learning_rate": 9.65202806473987e-08, "loss": 0.1941, "step": 9788 }, { "epoch": 2.6048430015965938, "grad_norm": 0.3086105287075043, "learning_rate": 9.650323013829412e-08, "loss": 0.1979, "step": 9789 }, { "epoch": 2.605109100585418, "grad_norm": 0.32331252098083496, "learning_rate": 9.64861797309714e-08, "loss": 0.2294, "step": 9790 }, { "epoch": 2.6053751995742416, "grad_norm": 0.30482226610183716, "learning_rate": 9.646912942592687e-08, "loss": 0.2009, "step": 9791 }, { "epoch": 2.6056412985630653, "grad_norm": 0.3187343180179596, "learning_rate": 9.645207922365683e-08, "loss": 0.2153, "step": 9792 }, { "epoch": 2.6059073975518894, "grad_norm": 0.28521403670310974, "learning_rate": 9.643502912465756e-08, "loss": 0.1885, "step": 9793 }, { "epoch": 2.606173496540713, "grad_norm": 0.2812175452709198, "learning_rate": 9.641797912942532e-08, "loss": 0.1945, "step": 9794 }, { "epoch": 2.6064395955295367, "grad_norm": 0.3033757209777832, "learning_rate": 9.640092923845641e-08, "loss": 0.2118, "step": 9795 }, { "epoch": 2.606705694518361, "grad_norm": 0.29285088181495667, "learning_rate": 9.63838794522471e-08, "loss": 0.1915, "step": 9796 }, { "epoch": 2.6069717935071846, "grad_norm": 0.28888413310050964, "learning_rate": 9.636682977129372e-08, "loss": 0.1898, "step": 9797 }, { "epoch": 2.6072378924960087, "grad_norm": 0.30065616965293884, "learning_rate": 9.634978019609244e-08, "loss": 0.1813, "step": 9798 }, { "epoch": 2.6075039914848324, "grad_norm": 0.290037602186203, "learning_rate": 9.633273072713958e-08, "loss": 0.2005, "step": 9799 }, { "epoch": 2.607770090473656, "grad_norm": 0.3033672571182251, "learning_rate": 9.631568136493141e-08, "loss": 0.2182, "step": 9800 }, { "epoch": 2.60803618946248, "grad_norm": 0.2858411967754364, "learning_rate": 9.629863210996418e-08, "loss": 0.2079, "step": 9801 }, { "epoch": 2.608302288451304, "grad_norm": 0.2984824776649475, "learning_rate": 9.62815829627342e-08, "loss": 0.1786, "step": 9802 }, { "epoch": 2.608568387440128, "grad_norm": 0.3771224617958069, "learning_rate": 9.626453392373764e-08, "loss": 0.1864, "step": 9803 }, { "epoch": 2.6088344864289517, "grad_norm": 0.2739388048648834, "learning_rate": 9.624748499347082e-08, "loss": 0.1775, "step": 9804 }, { "epoch": 2.6091005854177753, "grad_norm": 0.29884791374206543, "learning_rate": 9.623043617242996e-08, "loss": 0.1985, "step": 9805 }, { "epoch": 2.6093666844065995, "grad_norm": 0.29022249579429626, "learning_rate": 9.621338746111136e-08, "loss": 0.2013, "step": 9806 }, { "epoch": 2.609632783395423, "grad_norm": 0.2872214615345001, "learning_rate": 9.619633886001118e-08, "loss": 0.1953, "step": 9807 }, { "epoch": 2.609898882384247, "grad_norm": 0.29791679978370667, "learning_rate": 9.61792903696257e-08, "loss": 0.214, "step": 9808 }, { "epoch": 2.610164981373071, "grad_norm": 0.29955506324768066, "learning_rate": 9.616224199045115e-08, "loss": 0.1929, "step": 9809 }, { "epoch": 2.6104310803618946, "grad_norm": 0.3260681629180908, "learning_rate": 9.614519372298376e-08, "loss": 0.2182, "step": 9810 }, { "epoch": 2.6106971793507183, "grad_norm": 0.2757512927055359, "learning_rate": 9.612814556771979e-08, "loss": 0.1892, "step": 9811 }, { "epoch": 2.6109632783395424, "grad_norm": 0.29689398407936096, "learning_rate": 9.611109752515546e-08, "loss": 0.1863, "step": 9812 }, { "epoch": 2.611229377328366, "grad_norm": 0.3343140780925751, "learning_rate": 9.609404959578697e-08, "loss": 0.2092, "step": 9813 }, { "epoch": 2.61149547631719, "grad_norm": 0.3081148564815521, "learning_rate": 9.607700178011054e-08, "loss": 0.1889, "step": 9814 }, { "epoch": 2.611761575306014, "grad_norm": 0.27741968631744385, "learning_rate": 9.605995407862246e-08, "loss": 0.1913, "step": 9815 }, { "epoch": 2.6120276742948376, "grad_norm": 0.320487380027771, "learning_rate": 9.604290649181884e-08, "loss": 0.1959, "step": 9816 }, { "epoch": 2.6122937732836613, "grad_norm": 0.31056249141693115, "learning_rate": 9.602585902019595e-08, "loss": 0.2049, "step": 9817 }, { "epoch": 2.6125598722724854, "grad_norm": 0.30031490325927734, "learning_rate": 9.600881166424997e-08, "loss": 0.2024, "step": 9818 }, { "epoch": 2.612825971261309, "grad_norm": 0.32181257009506226, "learning_rate": 9.599176442447712e-08, "loss": 0.2047, "step": 9819 }, { "epoch": 2.613092070250133, "grad_norm": 0.3679294288158417, "learning_rate": 9.59747173013736e-08, "loss": 0.1891, "step": 9820 }, { "epoch": 2.613358169238957, "grad_norm": 0.30727341771125793, "learning_rate": 9.595767029543561e-08, "loss": 0.2105, "step": 9821 }, { "epoch": 2.6136242682277806, "grad_norm": 0.3323476016521454, "learning_rate": 9.594062340715936e-08, "loss": 0.1956, "step": 9822 }, { "epoch": 2.6138903672166047, "grad_norm": 0.3449176847934723, "learning_rate": 9.5923576637041e-08, "loss": 0.2244, "step": 9823 }, { "epoch": 2.6141564662054284, "grad_norm": 0.2921328842639923, "learning_rate": 9.590652998557678e-08, "loss": 0.2015, "step": 9824 }, { "epoch": 2.6144225651942525, "grad_norm": 0.31382501125335693, "learning_rate": 9.588948345326283e-08, "loss": 0.1875, "step": 9825 }, { "epoch": 2.614688664183076, "grad_norm": 0.2955707013607025, "learning_rate": 9.587243704059532e-08, "loss": 0.2097, "step": 9826 }, { "epoch": 2.6149547631719, "grad_norm": 0.31270983815193176, "learning_rate": 9.585539074807045e-08, "loss": 0.2131, "step": 9827 }, { "epoch": 2.615220862160724, "grad_norm": 0.29131120443344116, "learning_rate": 9.58383445761844e-08, "loss": 0.1974, "step": 9828 }, { "epoch": 2.6154869611495477, "grad_norm": 0.31903892755508423, "learning_rate": 9.582129852543332e-08, "loss": 0.1852, "step": 9829 }, { "epoch": 2.6157530601383714, "grad_norm": 0.31145042181015015, "learning_rate": 9.580425259631341e-08, "loss": 0.2041, "step": 9830 }, { "epoch": 2.6160191591271955, "grad_norm": 0.2939852774143219, "learning_rate": 9.57872067893208e-08, "loss": 0.1914, "step": 9831 }, { "epoch": 2.616285258116019, "grad_norm": 0.31937476992607117, "learning_rate": 9.577016110495169e-08, "loss": 0.1931, "step": 9832 }, { "epoch": 2.616551357104843, "grad_norm": 0.2521825134754181, "learning_rate": 9.57531155437022e-08, "loss": 0.1684, "step": 9833 }, { "epoch": 2.616817456093667, "grad_norm": 0.3101385533809662, "learning_rate": 9.573607010606848e-08, "loss": 0.2023, "step": 9834 }, { "epoch": 2.6170835550824907, "grad_norm": 0.2621942460536957, "learning_rate": 9.57190247925467e-08, "loss": 0.1714, "step": 9835 }, { "epoch": 2.6173496540713144, "grad_norm": 0.277555376291275, "learning_rate": 9.570197960363297e-08, "loss": 0.1796, "step": 9836 }, { "epoch": 2.6176157530601385, "grad_norm": 0.2753710150718689, "learning_rate": 9.568493453982345e-08, "loss": 0.1864, "step": 9837 }, { "epoch": 2.617881852048962, "grad_norm": 0.28599846363067627, "learning_rate": 9.56678896016143e-08, "loss": 0.195, "step": 9838 }, { "epoch": 2.618147951037786, "grad_norm": 0.28739333152770996, "learning_rate": 9.565084478950161e-08, "loss": 0.1797, "step": 9839 }, { "epoch": 2.61841405002661, "grad_norm": 0.28834080696105957, "learning_rate": 9.563380010398155e-08, "loss": 0.1999, "step": 9840 }, { "epoch": 2.6186801490154337, "grad_norm": 0.3192504942417145, "learning_rate": 9.561675554555024e-08, "loss": 0.2078, "step": 9841 }, { "epoch": 2.6189462480042573, "grad_norm": 0.32250142097473145, "learning_rate": 9.559971111470383e-08, "loss": 0.2088, "step": 9842 }, { "epoch": 2.6192123469930815, "grad_norm": 0.27718743681907654, "learning_rate": 9.558266681193836e-08, "loss": 0.1946, "step": 9843 }, { "epoch": 2.619478445981905, "grad_norm": 0.312534898519516, "learning_rate": 9.556562263775e-08, "loss": 0.2022, "step": 9844 }, { "epoch": 2.619744544970729, "grad_norm": 0.2823783755302429, "learning_rate": 9.554857859263486e-08, "loss": 0.1858, "step": 9845 }, { "epoch": 2.620010643959553, "grad_norm": 0.2642205059528351, "learning_rate": 9.553153467708902e-08, "loss": 0.1732, "step": 9846 }, { "epoch": 2.6202767429483766, "grad_norm": 0.28963977098464966, "learning_rate": 9.551449089160861e-08, "loss": 0.1933, "step": 9847 }, { "epoch": 2.6205428419372008, "grad_norm": 0.28195512294769287, "learning_rate": 9.54974472366897e-08, "loss": 0.1796, "step": 9848 }, { "epoch": 2.6208089409260245, "grad_norm": 0.28233101963996887, "learning_rate": 9.548040371282844e-08, "loss": 0.1883, "step": 9849 }, { "epoch": 2.6210750399148486, "grad_norm": 0.28333520889282227, "learning_rate": 9.546336032052088e-08, "loss": 0.1906, "step": 9850 }, { "epoch": 2.6213411389036723, "grad_norm": 0.288480281829834, "learning_rate": 9.544631706026317e-08, "loss": 0.2042, "step": 9851 }, { "epoch": 2.621607237892496, "grad_norm": 0.2948112189769745, "learning_rate": 9.54292739325513e-08, "loss": 0.1929, "step": 9852 }, { "epoch": 2.62187333688132, "grad_norm": 0.28577324748039246, "learning_rate": 9.54122309378814e-08, "loss": 0.1973, "step": 9853 }, { "epoch": 2.6221394358701438, "grad_norm": 0.29604777693748474, "learning_rate": 9.539518807674957e-08, "loss": 0.1954, "step": 9854 }, { "epoch": 2.6224055348589674, "grad_norm": 0.28745904564857483, "learning_rate": 9.537814534965183e-08, "loss": 0.1959, "step": 9855 }, { "epoch": 2.6226716338477916, "grad_norm": 0.28824156522750854, "learning_rate": 9.536110275708428e-08, "loss": 0.1929, "step": 9856 }, { "epoch": 2.6229377328366152, "grad_norm": 0.3020882308483124, "learning_rate": 9.534406029954299e-08, "loss": 0.2063, "step": 9857 }, { "epoch": 2.623203831825439, "grad_norm": 0.3164348900318146, "learning_rate": 9.532701797752401e-08, "loss": 0.2023, "step": 9858 }, { "epoch": 2.623469930814263, "grad_norm": 0.30601876974105835, "learning_rate": 9.53099757915234e-08, "loss": 0.1887, "step": 9859 }, { "epoch": 2.6237360298030867, "grad_norm": 0.3087588846683502, "learning_rate": 9.529293374203726e-08, "loss": 0.1994, "step": 9860 }, { "epoch": 2.6240021287919104, "grad_norm": 0.2646409273147583, "learning_rate": 9.527589182956156e-08, "loss": 0.1894, "step": 9861 }, { "epoch": 2.6242682277807345, "grad_norm": 0.3167863190174103, "learning_rate": 9.525885005459238e-08, "loss": 0.2259, "step": 9862 }, { "epoch": 2.6245343267695582, "grad_norm": 0.30532580614089966, "learning_rate": 9.524180841762577e-08, "loss": 0.2082, "step": 9863 }, { "epoch": 2.624800425758382, "grad_norm": 0.28326573967933655, "learning_rate": 9.522476691915777e-08, "loss": 0.184, "step": 9864 }, { "epoch": 2.625066524747206, "grad_norm": 0.2959205210208893, "learning_rate": 9.52077255596844e-08, "loss": 0.201, "step": 9865 }, { "epoch": 2.6253326237360297, "grad_norm": 0.2906176745891571, "learning_rate": 9.519068433970169e-08, "loss": 0.1958, "step": 9866 }, { "epoch": 2.6255987227248534, "grad_norm": 0.3021181523799896, "learning_rate": 9.517364325970567e-08, "loss": 0.2019, "step": 9867 }, { "epoch": 2.6258648217136775, "grad_norm": 0.27093252539634705, "learning_rate": 9.515660232019237e-08, "loss": 0.1738, "step": 9868 }, { "epoch": 2.626130920702501, "grad_norm": 0.29394853115081787, "learning_rate": 9.513956152165784e-08, "loss": 0.1945, "step": 9869 }, { "epoch": 2.6263970196913253, "grad_norm": 0.2957684397697449, "learning_rate": 9.5122520864598e-08, "loss": 0.2028, "step": 9870 }, { "epoch": 2.626663118680149, "grad_norm": 0.28395184874534607, "learning_rate": 9.510548034950893e-08, "loss": 0.192, "step": 9871 }, { "epoch": 2.6269292176689727, "grad_norm": 0.2610708475112915, "learning_rate": 9.508843997688662e-08, "loss": 0.1811, "step": 9872 }, { "epoch": 2.627195316657797, "grad_norm": 0.3006824553012848, "learning_rate": 9.507139974722705e-08, "loss": 0.1922, "step": 9873 }, { "epoch": 2.6274614156466205, "grad_norm": 0.2812713086605072, "learning_rate": 9.505435966102628e-08, "loss": 0.1819, "step": 9874 }, { "epoch": 2.6277275146354446, "grad_norm": 0.26224443316459656, "learning_rate": 9.503731971878022e-08, "loss": 0.1831, "step": 9875 }, { "epoch": 2.6279936136242683, "grad_norm": 0.2777775228023529, "learning_rate": 9.50202799209849e-08, "loss": 0.1875, "step": 9876 }, { "epoch": 2.628259712613092, "grad_norm": 0.29212486743927, "learning_rate": 9.50032402681363e-08, "loss": 0.2071, "step": 9877 }, { "epoch": 2.628525811601916, "grad_norm": 0.2965851128101349, "learning_rate": 9.498620076073047e-08, "loss": 0.194, "step": 9878 }, { "epoch": 2.62879191059074, "grad_norm": 0.2847842574119568, "learning_rate": 9.496916139926324e-08, "loss": 0.1966, "step": 9879 }, { "epoch": 2.6290580095795635, "grad_norm": 0.30290430784225464, "learning_rate": 9.495212218423068e-08, "loss": 0.2012, "step": 9880 }, { "epoch": 2.6293241085683876, "grad_norm": 0.289082795381546, "learning_rate": 9.493508311612874e-08, "loss": 0.2032, "step": 9881 }, { "epoch": 2.6295902075572113, "grad_norm": 0.3048672676086426, "learning_rate": 9.491804419545336e-08, "loss": 0.2138, "step": 9882 }, { "epoch": 2.629856306546035, "grad_norm": 0.2889997661113739, "learning_rate": 9.490100542270053e-08, "loss": 0.2037, "step": 9883 }, { "epoch": 2.630122405534859, "grad_norm": 0.2736967206001282, "learning_rate": 9.48839667983662e-08, "loss": 0.1917, "step": 9884 }, { "epoch": 2.630388504523683, "grad_norm": 0.29737091064453125, "learning_rate": 9.486692832294632e-08, "loss": 0.2084, "step": 9885 }, { "epoch": 2.6306546035125065, "grad_norm": 0.3297726809978485, "learning_rate": 9.484988999693681e-08, "loss": 0.1788, "step": 9886 }, { "epoch": 2.6309207025013306, "grad_norm": 0.28579387068748474, "learning_rate": 9.483285182083368e-08, "loss": 0.1995, "step": 9887 }, { "epoch": 2.6311868014901543, "grad_norm": 0.30761247873306274, "learning_rate": 9.481581379513277e-08, "loss": 0.1946, "step": 9888 }, { "epoch": 2.631452900478978, "grad_norm": 0.31770163774490356, "learning_rate": 9.479877592033006e-08, "loss": 0.1946, "step": 9889 }, { "epoch": 2.631718999467802, "grad_norm": 0.2958659529685974, "learning_rate": 9.478173819692147e-08, "loss": 0.1946, "step": 9890 }, { "epoch": 2.6319850984566258, "grad_norm": 0.2691918611526489, "learning_rate": 9.476470062540295e-08, "loss": 0.1958, "step": 9891 }, { "epoch": 2.6322511974454494, "grad_norm": 0.2638777196407318, "learning_rate": 9.47476632062704e-08, "loss": 0.1772, "step": 9892 }, { "epoch": 2.6325172964342736, "grad_norm": 0.27419912815093994, "learning_rate": 9.473062594001971e-08, "loss": 0.1942, "step": 9893 }, { "epoch": 2.6327833954230973, "grad_norm": 0.2849012315273285, "learning_rate": 9.471358882714686e-08, "loss": 0.1963, "step": 9894 }, { "epoch": 2.6330494944119214, "grad_norm": 0.2784726023674011, "learning_rate": 9.46965518681477e-08, "loss": 0.1839, "step": 9895 }, { "epoch": 2.633315593400745, "grad_norm": 0.2880219519138336, "learning_rate": 9.467951506351817e-08, "loss": 0.1737, "step": 9896 }, { "epoch": 2.633581692389569, "grad_norm": 0.2896842062473297, "learning_rate": 9.46624784137541e-08, "loss": 0.2008, "step": 9897 }, { "epoch": 2.633847791378393, "grad_norm": 0.2689674496650696, "learning_rate": 9.464544191935144e-08, "loss": 0.1824, "step": 9898 }, { "epoch": 2.6341138903672165, "grad_norm": 0.28224173188209534, "learning_rate": 9.462840558080605e-08, "loss": 0.1934, "step": 9899 }, { "epoch": 2.6343799893560407, "grad_norm": 0.29999664425849915, "learning_rate": 9.461136939861383e-08, "loss": 0.1923, "step": 9900 }, { "epoch": 2.6346460883448644, "grad_norm": 0.27272656559944153, "learning_rate": 9.459433337327065e-08, "loss": 0.1951, "step": 9901 }, { "epoch": 2.634912187333688, "grad_norm": 0.2973669469356537, "learning_rate": 9.45772975052724e-08, "loss": 0.1957, "step": 9902 }, { "epoch": 2.635178286322512, "grad_norm": 0.31009605526924133, "learning_rate": 9.456026179511496e-08, "loss": 0.1909, "step": 9903 }, { "epoch": 2.635444385311336, "grad_norm": 0.2951635420322418, "learning_rate": 9.454322624329413e-08, "loss": 0.2053, "step": 9904 }, { "epoch": 2.6357104843001595, "grad_norm": 0.30784475803375244, "learning_rate": 9.452619085030588e-08, "loss": 0.2027, "step": 9905 }, { "epoch": 2.6359765832889837, "grad_norm": 0.26457640528678894, "learning_rate": 9.450915561664598e-08, "loss": 0.1743, "step": 9906 }, { "epoch": 2.6362426822778073, "grad_norm": 0.32161158323287964, "learning_rate": 9.449212054281027e-08, "loss": 0.2057, "step": 9907 }, { "epoch": 2.636508781266631, "grad_norm": 0.28561872243881226, "learning_rate": 9.447508562929464e-08, "loss": 0.1983, "step": 9908 }, { "epoch": 2.636774880255455, "grad_norm": 0.28266164660453796, "learning_rate": 9.445805087659492e-08, "loss": 0.2009, "step": 9909 }, { "epoch": 2.637040979244279, "grad_norm": 0.27573129534721375, "learning_rate": 9.444101628520695e-08, "loss": 0.1907, "step": 9910 }, { "epoch": 2.6373070782331025, "grad_norm": 0.3330955505371094, "learning_rate": 9.442398185562657e-08, "loss": 0.2135, "step": 9911 }, { "epoch": 2.6375731772219266, "grad_norm": 0.29620179533958435, "learning_rate": 9.440694758834959e-08, "loss": 0.1989, "step": 9912 }, { "epoch": 2.6378392762107503, "grad_norm": 0.28195422887802124, "learning_rate": 9.438991348387186e-08, "loss": 0.1852, "step": 9913 }, { "epoch": 2.638105375199574, "grad_norm": 0.308366060256958, "learning_rate": 9.43728795426892e-08, "loss": 0.2136, "step": 9914 }, { "epoch": 2.638371474188398, "grad_norm": 0.3001715838909149, "learning_rate": 9.435584576529737e-08, "loss": 0.2044, "step": 9915 }, { "epoch": 2.638637573177222, "grad_norm": 0.29718706011772156, "learning_rate": 9.433881215219225e-08, "loss": 0.1993, "step": 9916 }, { "epoch": 2.638903672166046, "grad_norm": 0.32003670930862427, "learning_rate": 9.432177870386957e-08, "loss": 0.2053, "step": 9917 }, { "epoch": 2.6391697711548696, "grad_norm": 0.28847238421440125, "learning_rate": 9.430474542082518e-08, "loss": 0.1841, "step": 9918 }, { "epoch": 2.6394358701436933, "grad_norm": 0.2895463705062866, "learning_rate": 9.428771230355487e-08, "loss": 0.1968, "step": 9919 }, { "epoch": 2.6397019691325174, "grad_norm": 0.28271034359931946, "learning_rate": 9.427067935255441e-08, "loss": 0.1724, "step": 9920 }, { "epoch": 2.639968068121341, "grad_norm": 0.30430543422698975, "learning_rate": 9.42536465683196e-08, "loss": 0.2079, "step": 9921 }, { "epoch": 2.6402341671101652, "grad_norm": 0.2853877544403076, "learning_rate": 9.423661395134622e-08, "loss": 0.1878, "step": 9922 }, { "epoch": 2.640500266098989, "grad_norm": 0.2973951995372772, "learning_rate": 9.421958150213009e-08, "loss": 0.2148, "step": 9923 }, { "epoch": 2.6407663650878126, "grad_norm": 0.28036466240882874, "learning_rate": 9.420254922116689e-08, "loss": 0.1755, "step": 9924 }, { "epoch": 2.6410324640766367, "grad_norm": 0.2904036045074463, "learning_rate": 9.418551710895242e-08, "loss": 0.1849, "step": 9925 }, { "epoch": 2.6412985630654604, "grad_norm": 0.2946450114250183, "learning_rate": 9.416848516598248e-08, "loss": 0.1974, "step": 9926 }, { "epoch": 2.641564662054284, "grad_norm": 0.3157743811607361, "learning_rate": 9.415145339275276e-08, "loss": 0.2052, "step": 9927 }, { "epoch": 2.641830761043108, "grad_norm": 0.30967283248901367, "learning_rate": 9.413442178975906e-08, "loss": 0.1927, "step": 9928 }, { "epoch": 2.642096860031932, "grad_norm": 0.2924260199069977, "learning_rate": 9.411739035749712e-08, "loss": 0.1615, "step": 9929 }, { "epoch": 2.6423629590207556, "grad_norm": 0.2770316004753113, "learning_rate": 9.410035909646266e-08, "loss": 0.177, "step": 9930 }, { "epoch": 2.6426290580095797, "grad_norm": 0.28724536299705505, "learning_rate": 9.408332800715143e-08, "loss": 0.1946, "step": 9931 }, { "epoch": 2.6428951569984034, "grad_norm": 0.36555787920951843, "learning_rate": 9.40662970900592e-08, "loss": 0.2299, "step": 9932 }, { "epoch": 2.643161255987227, "grad_norm": 0.28488707542419434, "learning_rate": 9.404926634568162e-08, "loss": 0.1915, "step": 9933 }, { "epoch": 2.643427354976051, "grad_norm": 0.3301318883895874, "learning_rate": 9.403223577451443e-08, "loss": 0.2027, "step": 9934 }, { "epoch": 2.643693453964875, "grad_norm": 0.2875897288322449, "learning_rate": 9.401520537705337e-08, "loss": 0.2063, "step": 9935 }, { "epoch": 2.6439595529536986, "grad_norm": 0.3289218842983246, "learning_rate": 9.399817515379418e-08, "loss": 0.2051, "step": 9936 }, { "epoch": 2.6442256519425227, "grad_norm": 0.28112223744392395, "learning_rate": 9.398114510523248e-08, "loss": 0.1995, "step": 9937 }, { "epoch": 2.6444917509313464, "grad_norm": 0.2982581853866577, "learning_rate": 9.396411523186403e-08, "loss": 0.212, "step": 9938 }, { "epoch": 2.64475784992017, "grad_norm": 0.29779252409935, "learning_rate": 9.394708553418452e-08, "loss": 0.1965, "step": 9939 }, { "epoch": 2.645023948908994, "grad_norm": 0.29735907912254333, "learning_rate": 9.393005601268962e-08, "loss": 0.2093, "step": 9940 }, { "epoch": 2.645290047897818, "grad_norm": 0.3027251064777374, "learning_rate": 9.391302666787508e-08, "loss": 0.2002, "step": 9941 }, { "epoch": 2.645556146886642, "grad_norm": 0.3018028736114502, "learning_rate": 9.38959975002365e-08, "loss": 0.1704, "step": 9942 }, { "epoch": 2.6458222458754657, "grad_norm": 0.30260932445526123, "learning_rate": 9.387896851026957e-08, "loss": 0.1883, "step": 9943 }, { "epoch": 2.64608834486429, "grad_norm": 0.28472083806991577, "learning_rate": 9.386193969846997e-08, "loss": 0.1939, "step": 9944 }, { "epoch": 2.6463544438531135, "grad_norm": 0.2999827265739441, "learning_rate": 9.384491106533338e-08, "loss": 0.2054, "step": 9945 }, { "epoch": 2.646620542841937, "grad_norm": 0.2688619792461395, "learning_rate": 9.382788261135545e-08, "loss": 0.1861, "step": 9946 }, { "epoch": 2.6468866418307613, "grad_norm": 0.2733205556869507, "learning_rate": 9.381085433703182e-08, "loss": 0.197, "step": 9947 }, { "epoch": 2.647152740819585, "grad_norm": 0.29365313053131104, "learning_rate": 9.379382624285815e-08, "loss": 0.1837, "step": 9948 }, { "epoch": 2.6474188398084086, "grad_norm": 0.2870711088180542, "learning_rate": 9.377679832933009e-08, "loss": 0.2012, "step": 9949 }, { "epoch": 2.6476849387972328, "grad_norm": 0.27383047342300415, "learning_rate": 9.375977059694329e-08, "loss": 0.1906, "step": 9950 }, { "epoch": 2.6479510377860564, "grad_norm": 0.33478888869285583, "learning_rate": 9.374274304619335e-08, "loss": 0.2256, "step": 9951 }, { "epoch": 2.64821713677488, "grad_norm": 0.29917293787002563, "learning_rate": 9.37257156775759e-08, "loss": 0.2062, "step": 9952 }, { "epoch": 2.6484832357637043, "grad_norm": 0.31259870529174805, "learning_rate": 9.37086884915866e-08, "loss": 0.2095, "step": 9953 }, { "epoch": 2.648749334752528, "grad_norm": 0.3107747733592987, "learning_rate": 9.369166148872101e-08, "loss": 0.1866, "step": 9954 }, { "epoch": 2.6490154337413516, "grad_norm": 0.3054930865764618, "learning_rate": 9.367463466947482e-08, "loss": 0.2016, "step": 9955 }, { "epoch": 2.6492815327301757, "grad_norm": 0.2709258794784546, "learning_rate": 9.365760803434354e-08, "loss": 0.1859, "step": 9956 }, { "epoch": 2.6495476317189994, "grad_norm": 0.2801467180252075, "learning_rate": 9.364058158382286e-08, "loss": 0.2044, "step": 9957 }, { "epoch": 2.649813730707823, "grad_norm": 0.28945428133010864, "learning_rate": 9.36235553184083e-08, "loss": 0.1816, "step": 9958 }, { "epoch": 2.6500798296966472, "grad_norm": 0.3019697666168213, "learning_rate": 9.360652923859556e-08, "loss": 0.1902, "step": 9959 }, { "epoch": 2.650345928685471, "grad_norm": 0.32037484645843506, "learning_rate": 9.358950334488011e-08, "loss": 0.2275, "step": 9960 }, { "epoch": 2.6506120276742946, "grad_norm": 0.29168301820755005, "learning_rate": 9.357247763775756e-08, "loss": 0.1915, "step": 9961 }, { "epoch": 2.6508781266631187, "grad_norm": 0.3124184012413025, "learning_rate": 9.35554521177235e-08, "loss": 0.2105, "step": 9962 }, { "epoch": 2.6511442256519424, "grad_norm": 0.26574811339378357, "learning_rate": 9.35384267852735e-08, "loss": 0.1942, "step": 9963 }, { "epoch": 2.651410324640766, "grad_norm": 0.3133540153503418, "learning_rate": 9.35214016409031e-08, "loss": 0.194, "step": 9964 }, { "epoch": 2.65167642362959, "grad_norm": 0.311889111995697, "learning_rate": 9.350437668510794e-08, "loss": 0.1986, "step": 9965 }, { "epoch": 2.651942522618414, "grad_norm": 0.28874874114990234, "learning_rate": 9.348735191838345e-08, "loss": 0.1901, "step": 9966 }, { "epoch": 2.652208621607238, "grad_norm": 0.29967039823532104, "learning_rate": 9.347032734122524e-08, "loss": 0.1941, "step": 9967 }, { "epoch": 2.6524747205960617, "grad_norm": 0.2863479554653168, "learning_rate": 9.345330295412885e-08, "loss": 0.1794, "step": 9968 }, { "epoch": 2.652740819584886, "grad_norm": 0.3099868595600128, "learning_rate": 9.343627875758987e-08, "loss": 0.1952, "step": 9969 }, { "epoch": 2.6530069185737095, "grad_norm": 0.30344367027282715, "learning_rate": 9.341925475210372e-08, "loss": 0.1978, "step": 9970 }, { "epoch": 2.653273017562533, "grad_norm": 0.2869921922683716, "learning_rate": 9.340223093816598e-08, "loss": 0.1789, "step": 9971 }, { "epoch": 2.6535391165513573, "grad_norm": 0.2988170385360718, "learning_rate": 9.338520731627216e-08, "loss": 0.1823, "step": 9972 }, { "epoch": 2.653805215540181, "grad_norm": 0.30635857582092285, "learning_rate": 9.336818388691778e-08, "loss": 0.1943, "step": 9973 }, { "epoch": 2.6540713145290047, "grad_norm": 0.2839616537094116, "learning_rate": 9.335116065059836e-08, "loss": 0.1893, "step": 9974 }, { "epoch": 2.654337413517829, "grad_norm": 0.2733480632305145, "learning_rate": 9.333413760780939e-08, "loss": 0.1814, "step": 9975 }, { "epoch": 2.6546035125066525, "grad_norm": 0.3102288544178009, "learning_rate": 9.331711475904637e-08, "loss": 0.2023, "step": 9976 }, { "epoch": 2.654869611495476, "grad_norm": 0.2981685400009155, "learning_rate": 9.330009210480477e-08, "loss": 0.1986, "step": 9977 }, { "epoch": 2.6551357104843003, "grad_norm": 0.2900114059448242, "learning_rate": 9.328306964558015e-08, "loss": 0.1765, "step": 9978 }, { "epoch": 2.655401809473124, "grad_norm": 0.28738975524902344, "learning_rate": 9.326604738186787e-08, "loss": 0.2134, "step": 9979 }, { "epoch": 2.6556679084619477, "grad_norm": 0.29451796412467957, "learning_rate": 9.324902531416348e-08, "loss": 0.2065, "step": 9980 }, { "epoch": 2.655934007450772, "grad_norm": 0.28286972641944885, "learning_rate": 9.323200344296243e-08, "loss": 0.1917, "step": 9981 }, { "epoch": 2.6562001064395955, "grad_norm": 0.27418217062950134, "learning_rate": 9.32149817687602e-08, "loss": 0.1776, "step": 9982 }, { "epoch": 2.656466205428419, "grad_norm": 0.30949902534484863, "learning_rate": 9.319796029205221e-08, "loss": 0.1968, "step": 9983 }, { "epoch": 2.6567323044172433, "grad_norm": 0.2808607816696167, "learning_rate": 9.318093901333395e-08, "loss": 0.1942, "step": 9984 }, { "epoch": 2.656998403406067, "grad_norm": 0.30799755454063416, "learning_rate": 9.316391793310087e-08, "loss": 0.2001, "step": 9985 }, { "epoch": 2.6572645023948906, "grad_norm": 0.2686201333999634, "learning_rate": 9.314689705184836e-08, "loss": 0.1712, "step": 9986 }, { "epoch": 2.6575306013837148, "grad_norm": 0.2812924087047577, "learning_rate": 9.31298763700719e-08, "loss": 0.1983, "step": 9987 }, { "epoch": 2.6577967003725385, "grad_norm": 0.2488856166601181, "learning_rate": 9.311285588826692e-08, "loss": 0.1507, "step": 9988 }, { "epoch": 2.6580627993613626, "grad_norm": 0.2956167161464691, "learning_rate": 9.309583560692879e-08, "loss": 0.1992, "step": 9989 }, { "epoch": 2.6583288983501863, "grad_norm": 0.30900177359580994, "learning_rate": 9.307881552655296e-08, "loss": 0.2042, "step": 9990 }, { "epoch": 2.65859499733901, "grad_norm": 0.2931223511695862, "learning_rate": 9.306179564763483e-08, "loss": 0.2054, "step": 9991 }, { "epoch": 2.658861096327834, "grad_norm": 0.2673734724521637, "learning_rate": 9.304477597066983e-08, "loss": 0.171, "step": 9992 }, { "epoch": 2.6591271953166578, "grad_norm": 0.313991904258728, "learning_rate": 9.302775649615333e-08, "loss": 0.1958, "step": 9993 }, { "epoch": 2.659393294305482, "grad_norm": 0.2890240550041199, "learning_rate": 9.301073722458078e-08, "loss": 0.1977, "step": 9994 }, { "epoch": 2.6596593932943056, "grad_norm": 0.2756499648094177, "learning_rate": 9.299371815644748e-08, "loss": 0.1844, "step": 9995 }, { "epoch": 2.6599254922831292, "grad_norm": 0.3378642797470093, "learning_rate": 9.297669929224887e-08, "loss": 0.2229, "step": 9996 }, { "epoch": 2.6601915912719534, "grad_norm": 0.2973807156085968, "learning_rate": 9.295968063248032e-08, "loss": 0.2024, "step": 9997 }, { "epoch": 2.660457690260777, "grad_norm": 0.2985651195049286, "learning_rate": 9.294266217763716e-08, "loss": 0.202, "step": 9998 }, { "epoch": 2.6607237892496007, "grad_norm": 0.2969423830509186, "learning_rate": 9.292564392821478e-08, "loss": 0.1951, "step": 9999 }, { "epoch": 2.660989888238425, "grad_norm": 0.2936968505382538, "learning_rate": 9.290862588470854e-08, "loss": 0.1845, "step": 10000 }, { "epoch": 2.6612559872272485, "grad_norm": 0.29183685779571533, "learning_rate": 9.289160804761378e-08, "loss": 0.1795, "step": 10001 }, { "epoch": 2.6615220862160722, "grad_norm": 0.31102269887924194, "learning_rate": 9.287459041742585e-08, "loss": 0.2035, "step": 10002 }, { "epoch": 2.6617881852048964, "grad_norm": 0.30115199089050293, "learning_rate": 9.28575729946401e-08, "loss": 0.2072, "step": 10003 }, { "epoch": 2.66205428419372, "grad_norm": 0.3037875294685364, "learning_rate": 9.284055577975188e-08, "loss": 0.1958, "step": 10004 }, { "epoch": 2.6623203831825437, "grad_norm": 0.27721449732780457, "learning_rate": 9.28235387732565e-08, "loss": 0.1795, "step": 10005 }, { "epoch": 2.662586482171368, "grad_norm": 0.29364457726478577, "learning_rate": 9.280652197564924e-08, "loss": 0.2123, "step": 10006 }, { "epoch": 2.6628525811601915, "grad_norm": 0.27162012457847595, "learning_rate": 9.278950538742546e-08, "loss": 0.1753, "step": 10007 }, { "epoch": 2.663118680149015, "grad_norm": 0.3032427132129669, "learning_rate": 9.277248900908045e-08, "loss": 0.1983, "step": 10008 }, { "epoch": 2.6633847791378393, "grad_norm": 0.3033877909183502, "learning_rate": 9.27554728411095e-08, "loss": 0.1967, "step": 10009 }, { "epoch": 2.663650878126663, "grad_norm": 0.2670600414276123, "learning_rate": 9.273845688400794e-08, "loss": 0.1789, "step": 10010 }, { "epoch": 2.6639169771154867, "grad_norm": 0.29644298553466797, "learning_rate": 9.272144113827105e-08, "loss": 0.19, "step": 10011 }, { "epoch": 2.664183076104311, "grad_norm": 0.30068933963775635, "learning_rate": 9.27044256043941e-08, "loss": 0.1991, "step": 10012 }, { "epoch": 2.6644491750931345, "grad_norm": 0.30111175775527954, "learning_rate": 9.268741028287238e-08, "loss": 0.1979, "step": 10013 }, { "epoch": 2.6647152740819586, "grad_norm": 0.30399540066719055, "learning_rate": 9.26703951742012e-08, "loss": 0.2019, "step": 10014 }, { "epoch": 2.6649813730707823, "grad_norm": 0.3047797381877899, "learning_rate": 9.265338027887574e-08, "loss": 0.1974, "step": 10015 }, { "epoch": 2.6652474720596064, "grad_norm": 0.28744226694107056, "learning_rate": 9.263636559739132e-08, "loss": 0.1932, "step": 10016 }, { "epoch": 2.66551357104843, "grad_norm": 0.29343587160110474, "learning_rate": 9.26193511302432e-08, "loss": 0.2034, "step": 10017 }, { "epoch": 2.665779670037254, "grad_norm": 0.3230920135974884, "learning_rate": 9.260233687792657e-08, "loss": 0.2008, "step": 10018 }, { "epoch": 2.666045769026078, "grad_norm": 0.3068360686302185, "learning_rate": 9.258532284093671e-08, "loss": 0.2138, "step": 10019 }, { "epoch": 2.6663118680149016, "grad_norm": 0.28821542859077454, "learning_rate": 9.256830901976886e-08, "loss": 0.1937, "step": 10020 }, { "epoch": 2.6665779670037253, "grad_norm": 0.30668675899505615, "learning_rate": 9.255129541491822e-08, "loss": 0.2024, "step": 10021 }, { "epoch": 2.6668440659925494, "grad_norm": 0.29927176237106323, "learning_rate": 9.253428202688004e-08, "loss": 0.2022, "step": 10022 }, { "epoch": 2.667110164981373, "grad_norm": 0.2990668714046478, "learning_rate": 9.251726885614956e-08, "loss": 0.2107, "step": 10023 }, { "epoch": 2.667376263970197, "grad_norm": 0.2991805374622345, "learning_rate": 9.250025590322192e-08, "loss": 0.1862, "step": 10024 }, { "epoch": 2.667642362959021, "grad_norm": 0.28119605779647827, "learning_rate": 9.248324316859236e-08, "loss": 0.185, "step": 10025 }, { "epoch": 2.6679084619478446, "grad_norm": 0.2910069525241852, "learning_rate": 9.246623065275607e-08, "loss": 0.1878, "step": 10026 }, { "epoch": 2.6681745609366683, "grad_norm": 0.2861195206642151, "learning_rate": 9.244921835620826e-08, "loss": 0.1921, "step": 10027 }, { "epoch": 2.6684406599254924, "grad_norm": 0.33776217699050903, "learning_rate": 9.243220627944409e-08, "loss": 0.1976, "step": 10028 }, { "epoch": 2.668706758914316, "grad_norm": 0.29959848523139954, "learning_rate": 9.241519442295874e-08, "loss": 0.1993, "step": 10029 }, { "epoch": 2.6689728579031398, "grad_norm": 0.3136862516403198, "learning_rate": 9.239818278724738e-08, "loss": 0.2092, "step": 10030 }, { "epoch": 2.669238956891964, "grad_norm": 0.2731063663959503, "learning_rate": 9.238117137280519e-08, "loss": 0.1722, "step": 10031 }, { "epoch": 2.6695050558807876, "grad_norm": 0.30830439925193787, "learning_rate": 9.236416018012736e-08, "loss": 0.2006, "step": 10032 }, { "epoch": 2.6697711548696113, "grad_norm": 0.29865339398384094, "learning_rate": 9.234714920970896e-08, "loss": 0.1873, "step": 10033 }, { "epoch": 2.6700372538584354, "grad_norm": 0.3046357035636902, "learning_rate": 9.233013846204517e-08, "loss": 0.1989, "step": 10034 }, { "epoch": 2.670303352847259, "grad_norm": 0.2915939688682556, "learning_rate": 9.231312793763114e-08, "loss": 0.1696, "step": 10035 }, { "epoch": 2.670569451836083, "grad_norm": 0.27545884251594543, "learning_rate": 9.2296117636962e-08, "loss": 0.1777, "step": 10036 }, { "epoch": 2.670835550824907, "grad_norm": 0.2533074915409088, "learning_rate": 9.227910756053289e-08, "loss": 0.1602, "step": 10037 }, { "epoch": 2.6711016498137305, "grad_norm": 0.296502023935318, "learning_rate": 9.226209770883888e-08, "loss": 0.2098, "step": 10038 }, { "epoch": 2.6713677488025547, "grad_norm": 0.30265185236930847, "learning_rate": 9.224508808237512e-08, "loss": 0.1891, "step": 10039 }, { "epoch": 2.6716338477913784, "grad_norm": 0.2946624755859375, "learning_rate": 9.222807868163672e-08, "loss": 0.1957, "step": 10040 }, { "epoch": 2.6718999467802025, "grad_norm": 0.2748669981956482, "learning_rate": 9.221106950711881e-08, "loss": 0.1937, "step": 10041 }, { "epoch": 2.672166045769026, "grad_norm": 0.3024331033229828, "learning_rate": 9.21940605593164e-08, "loss": 0.1947, "step": 10042 }, { "epoch": 2.67243214475785, "grad_norm": 0.3138794004917145, "learning_rate": 9.21770518387246e-08, "loss": 0.1909, "step": 10043 }, { "epoch": 2.672698243746674, "grad_norm": 0.30611494183540344, "learning_rate": 9.216004334583853e-08, "loss": 0.219, "step": 10044 }, { "epoch": 2.6729643427354977, "grad_norm": 0.29197022318840027, "learning_rate": 9.214303508115325e-08, "loss": 0.1947, "step": 10045 }, { "epoch": 2.6732304417243213, "grad_norm": 0.27116474509239197, "learning_rate": 9.212602704516383e-08, "loss": 0.1918, "step": 10046 }, { "epoch": 2.6734965407131455, "grad_norm": 0.2881389260292053, "learning_rate": 9.210901923836529e-08, "loss": 0.1925, "step": 10047 }, { "epoch": 2.673762639701969, "grad_norm": 0.28575825691223145, "learning_rate": 9.209201166125271e-08, "loss": 0.2013, "step": 10048 }, { "epoch": 2.674028738690793, "grad_norm": 0.30997714400291443, "learning_rate": 9.207500431432114e-08, "loss": 0.1922, "step": 10049 }, { "epoch": 2.674294837679617, "grad_norm": 0.3126145005226135, "learning_rate": 9.205799719806566e-08, "loss": 0.2035, "step": 10050 }, { "epoch": 2.6745609366684406, "grad_norm": 0.29883885383605957, "learning_rate": 9.204099031298123e-08, "loss": 0.1992, "step": 10051 }, { "epoch": 2.6748270356572643, "grad_norm": 0.41498759388923645, "learning_rate": 9.20239836595629e-08, "loss": 0.2048, "step": 10052 }, { "epoch": 2.6750931346460884, "grad_norm": 0.28548291325569153, "learning_rate": 9.200697723830569e-08, "loss": 0.1945, "step": 10053 }, { "epoch": 2.675359233634912, "grad_norm": 0.25668102502822876, "learning_rate": 9.198997104970461e-08, "loss": 0.1676, "step": 10054 }, { "epoch": 2.675625332623736, "grad_norm": 0.30362382531166077, "learning_rate": 9.19729650942547e-08, "loss": 0.2038, "step": 10055 }, { "epoch": 2.67589143161256, "grad_norm": 0.2989717423915863, "learning_rate": 9.195595937245094e-08, "loss": 0.202, "step": 10056 }, { "epoch": 2.6761575306013836, "grad_norm": 0.2810022532939911, "learning_rate": 9.193895388478829e-08, "loss": 0.2045, "step": 10057 }, { "epoch": 2.6764236295902073, "grad_norm": 0.2957290709018707, "learning_rate": 9.192194863176176e-08, "loss": 0.2125, "step": 10058 }, { "epoch": 2.6766897285790314, "grad_norm": 0.28773486614227295, "learning_rate": 9.190494361386641e-08, "loss": 0.1864, "step": 10059 }, { "epoch": 2.676955827567855, "grad_norm": 0.32547527551651, "learning_rate": 9.188793883159705e-08, "loss": 0.1843, "step": 10060 }, { "epoch": 2.6772219265566792, "grad_norm": 0.33490079641342163, "learning_rate": 9.187093428544874e-08, "loss": 0.2126, "step": 10061 }, { "epoch": 2.677488025545503, "grad_norm": 0.26991307735443115, "learning_rate": 9.185392997591644e-08, "loss": 0.1697, "step": 10062 }, { "epoch": 2.677754124534327, "grad_norm": 0.28377532958984375, "learning_rate": 9.183692590349506e-08, "loss": 0.1908, "step": 10063 }, { "epoch": 2.6780202235231507, "grad_norm": 0.30166518688201904, "learning_rate": 9.18199220686796e-08, "loss": 0.2098, "step": 10064 }, { "epoch": 2.6782863225119744, "grad_norm": 0.3027515411376953, "learning_rate": 9.180291847196496e-08, "loss": 0.1954, "step": 10065 }, { "epoch": 2.6785524215007985, "grad_norm": 0.2798023819923401, "learning_rate": 9.178591511384608e-08, "loss": 0.1917, "step": 10066 }, { "epoch": 2.678818520489622, "grad_norm": 0.3164762258529663, "learning_rate": 9.176891199481788e-08, "loss": 0.205, "step": 10067 }, { "epoch": 2.679084619478446, "grad_norm": 0.2793607711791992, "learning_rate": 9.17519091153753e-08, "loss": 0.1885, "step": 10068 }, { "epoch": 2.67935071846727, "grad_norm": 0.2690056562423706, "learning_rate": 9.173490647601323e-08, "loss": 0.1967, "step": 10069 }, { "epoch": 2.6796168174560937, "grad_norm": 0.2707928419113159, "learning_rate": 9.171790407722656e-08, "loss": 0.1738, "step": 10070 }, { "epoch": 2.6798829164449174, "grad_norm": 0.263826847076416, "learning_rate": 9.170090191951019e-08, "loss": 0.1685, "step": 10071 }, { "epoch": 2.6801490154337415, "grad_norm": 0.29111939668655396, "learning_rate": 9.168390000335901e-08, "loss": 0.192, "step": 10072 }, { "epoch": 2.680415114422565, "grad_norm": 0.29431024193763733, "learning_rate": 9.166689832926792e-08, "loss": 0.209, "step": 10073 }, { "epoch": 2.680681213411389, "grad_norm": 0.27925682067871094, "learning_rate": 9.164989689773177e-08, "loss": 0.191, "step": 10074 }, { "epoch": 2.680947312400213, "grad_norm": 0.35181355476379395, "learning_rate": 9.163289570924545e-08, "loss": 0.2029, "step": 10075 }, { "epoch": 2.6812134113890367, "grad_norm": 0.29083994030952454, "learning_rate": 9.161589476430383e-08, "loss": 0.17, "step": 10076 }, { "epoch": 2.6814795103778604, "grad_norm": 0.2972864508628845, "learning_rate": 9.159889406340174e-08, "loss": 0.2052, "step": 10077 }, { "epoch": 2.6817456093666845, "grad_norm": 0.2851184606552124, "learning_rate": 9.158189360703401e-08, "loss": 0.1844, "step": 10078 }, { "epoch": 2.682011708355508, "grad_norm": 0.3204827606678009, "learning_rate": 9.156489339569553e-08, "loss": 0.1889, "step": 10079 }, { "epoch": 2.682277807344332, "grad_norm": 0.3200317621231079, "learning_rate": 9.154789342988107e-08, "loss": 0.2077, "step": 10080 }, { "epoch": 2.682543906333156, "grad_norm": 0.2945285737514496, "learning_rate": 9.15308937100855e-08, "loss": 0.1939, "step": 10081 }, { "epoch": 2.6828100053219797, "grad_norm": 0.2874632477760315, "learning_rate": 9.151389423680362e-08, "loss": 0.1858, "step": 10082 }, { "epoch": 2.6830761043108033, "grad_norm": 0.26742395758628845, "learning_rate": 9.149689501053023e-08, "loss": 0.185, "step": 10083 }, { "epoch": 2.6833422032996275, "grad_norm": 0.2985045611858368, "learning_rate": 9.147989603176016e-08, "loss": 0.1953, "step": 10084 }, { "epoch": 2.683608302288451, "grad_norm": 0.3197888731956482, "learning_rate": 9.14628973009882e-08, "loss": 0.2184, "step": 10085 }, { "epoch": 2.6838744012772753, "grad_norm": 0.3082524240016937, "learning_rate": 9.144589881870917e-08, "loss": 0.1975, "step": 10086 }, { "epoch": 2.684140500266099, "grad_norm": 0.3270261883735657, "learning_rate": 9.142890058541777e-08, "loss": 0.2116, "step": 10087 }, { "epoch": 2.684406599254923, "grad_norm": 0.27951592206954956, "learning_rate": 9.141190260160884e-08, "loss": 0.1879, "step": 10088 }, { "epoch": 2.6846726982437468, "grad_norm": 0.3175404965877533, "learning_rate": 9.13949048677771e-08, "loss": 0.2172, "step": 10089 }, { "epoch": 2.6849387972325705, "grad_norm": 0.2947024703025818, "learning_rate": 9.137790738441735e-08, "loss": 0.1821, "step": 10090 }, { "epoch": 2.6852048962213946, "grad_norm": 0.3347714841365814, "learning_rate": 9.136091015202432e-08, "loss": 0.1978, "step": 10091 }, { "epoch": 2.6854709952102183, "grad_norm": 0.3049811124801636, "learning_rate": 9.134391317109277e-08, "loss": 0.193, "step": 10092 }, { "epoch": 2.685737094199042, "grad_norm": 0.29317763447761536, "learning_rate": 9.132691644211742e-08, "loss": 0.1793, "step": 10093 }, { "epoch": 2.686003193187866, "grad_norm": 0.2955884635448456, "learning_rate": 9.130991996559303e-08, "loss": 0.2127, "step": 10094 }, { "epoch": 2.6862692921766897, "grad_norm": 0.29418185353279114, "learning_rate": 9.129292374201434e-08, "loss": 0.1925, "step": 10095 }, { "epoch": 2.6865353911655134, "grad_norm": 0.30996668338775635, "learning_rate": 9.127592777187598e-08, "loss": 0.1952, "step": 10096 }, { "epoch": 2.6868014901543376, "grad_norm": 0.27932363748550415, "learning_rate": 9.125893205567272e-08, "loss": 0.1694, "step": 10097 }, { "epoch": 2.6870675891431612, "grad_norm": 0.27427926659584045, "learning_rate": 9.124193659389928e-08, "loss": 0.182, "step": 10098 }, { "epoch": 2.687333688131985, "grad_norm": 0.27747392654418945, "learning_rate": 9.12249413870503e-08, "loss": 0.1786, "step": 10099 }, { "epoch": 2.687599787120809, "grad_norm": 0.2824425995349884, "learning_rate": 9.120794643562049e-08, "loss": 0.1993, "step": 10100 }, { "epoch": 2.6878658861096327, "grad_norm": 0.2984657287597656, "learning_rate": 9.119095174010453e-08, "loss": 0.1909, "step": 10101 }, { "epoch": 2.6881319850984564, "grad_norm": 0.2777099013328552, "learning_rate": 9.11739573009971e-08, "loss": 0.1856, "step": 10102 }, { "epoch": 2.6883980840872805, "grad_norm": 0.29414668679237366, "learning_rate": 9.115696311879284e-08, "loss": 0.1938, "step": 10103 }, { "epoch": 2.688664183076104, "grad_norm": 0.2878166437149048, "learning_rate": 9.113996919398648e-08, "loss": 0.1917, "step": 10104 }, { "epoch": 2.688930282064928, "grad_norm": 0.27235227823257446, "learning_rate": 9.112297552707257e-08, "loss": 0.1765, "step": 10105 }, { "epoch": 2.689196381053752, "grad_norm": 0.2866562604904175, "learning_rate": 9.110598211854579e-08, "loss": 0.1841, "step": 10106 }, { "epoch": 2.6894624800425757, "grad_norm": 0.2850109338760376, "learning_rate": 9.108898896890076e-08, "loss": 0.1774, "step": 10107 }, { "epoch": 2.6897285790314, "grad_norm": 0.2846860885620117, "learning_rate": 9.107199607863217e-08, "loss": 0.1852, "step": 10108 }, { "epoch": 2.6899946780202235, "grad_norm": 0.3090948164463043, "learning_rate": 9.105500344823456e-08, "loss": 0.2063, "step": 10109 }, { "epoch": 2.690260777009047, "grad_norm": 0.2906852066516876, "learning_rate": 9.103801107820258e-08, "loss": 0.1978, "step": 10110 }, { "epoch": 2.6905268759978713, "grad_norm": 0.2903880178928375, "learning_rate": 9.102101896903083e-08, "loss": 0.2183, "step": 10111 }, { "epoch": 2.690792974986695, "grad_norm": 0.3066238760948181, "learning_rate": 9.100402712121389e-08, "loss": 0.2, "step": 10112 }, { "epoch": 2.691059073975519, "grad_norm": 0.30340057611465454, "learning_rate": 9.098703553524642e-08, "loss": 0.1949, "step": 10113 }, { "epoch": 2.691325172964343, "grad_norm": 0.3055426776409149, "learning_rate": 9.09700442116229e-08, "loss": 0.2114, "step": 10114 }, { "epoch": 2.6915912719531665, "grad_norm": 0.2766842246055603, "learning_rate": 9.095305315083793e-08, "loss": 0.1882, "step": 10115 }, { "epoch": 2.6918573709419906, "grad_norm": 0.3432631194591522, "learning_rate": 9.093606235338612e-08, "loss": 0.2281, "step": 10116 }, { "epoch": 2.6921234699308143, "grad_norm": 0.3024308383464813, "learning_rate": 9.091907181976197e-08, "loss": 0.1884, "step": 10117 }, { "epoch": 2.692389568919638, "grad_norm": 0.2838062644004822, "learning_rate": 9.09020815504601e-08, "loss": 0.1841, "step": 10118 }, { "epoch": 2.692655667908462, "grad_norm": 0.3091967701911926, "learning_rate": 9.088509154597499e-08, "loss": 0.1927, "step": 10119 }, { "epoch": 2.692921766897286, "grad_norm": 0.28880175948143005, "learning_rate": 9.08681018068012e-08, "loss": 0.1877, "step": 10120 }, { "epoch": 2.6931878658861095, "grad_norm": 0.30146023631095886, "learning_rate": 9.085111233343326e-08, "loss": 0.199, "step": 10121 }, { "epoch": 2.6934539648749336, "grad_norm": 0.28660136461257935, "learning_rate": 9.083412312636571e-08, "loss": 0.211, "step": 10122 }, { "epoch": 2.6937200638637573, "grad_norm": 0.31434524059295654, "learning_rate": 9.081713418609301e-08, "loss": 0.1951, "step": 10123 }, { "epoch": 2.693986162852581, "grad_norm": 0.3087347149848938, "learning_rate": 9.080014551310969e-08, "loss": 0.202, "step": 10124 }, { "epoch": 2.694252261841405, "grad_norm": 0.30097150802612305, "learning_rate": 9.078315710791025e-08, "loss": 0.1821, "step": 10125 }, { "epoch": 2.6945183608302288, "grad_norm": 0.30194422602653503, "learning_rate": 9.076616897098916e-08, "loss": 0.1901, "step": 10126 }, { "epoch": 2.6947844598190525, "grad_norm": 0.3014296293258667, "learning_rate": 9.074918110284094e-08, "loss": 0.1988, "step": 10127 }, { "epoch": 2.6950505588078766, "grad_norm": 0.31884193420410156, "learning_rate": 9.073219350396003e-08, "loss": 0.1923, "step": 10128 }, { "epoch": 2.6953166577967003, "grad_norm": 0.28014901280403137, "learning_rate": 9.07152061748409e-08, "loss": 0.1939, "step": 10129 }, { "epoch": 2.695582756785524, "grad_norm": 0.2776657044887543, "learning_rate": 9.069821911597801e-08, "loss": 0.2013, "step": 10130 }, { "epoch": 2.695848855774348, "grad_norm": 0.3035949170589447, "learning_rate": 9.068123232786586e-08, "loss": 0.194, "step": 10131 }, { "epoch": 2.6961149547631718, "grad_norm": 0.3019736111164093, "learning_rate": 9.066424581099878e-08, "loss": 0.2124, "step": 10132 }, { "epoch": 2.696381053751996, "grad_norm": 0.2706938683986664, "learning_rate": 9.064725956587127e-08, "loss": 0.1622, "step": 10133 }, { "epoch": 2.6966471527408196, "grad_norm": 0.2984049320220947, "learning_rate": 9.063027359297775e-08, "loss": 0.1875, "step": 10134 }, { "epoch": 2.6969132517296437, "grad_norm": 0.30228161811828613, "learning_rate": 9.061328789281264e-08, "loss": 0.2032, "step": 10135 }, { "epoch": 2.6971793507184674, "grad_norm": 0.3024660348892212, "learning_rate": 9.059630246587034e-08, "loss": 0.203, "step": 10136 }, { "epoch": 2.697445449707291, "grad_norm": 0.3726619780063629, "learning_rate": 9.057931731264529e-08, "loss": 0.1922, "step": 10137 }, { "epoch": 2.697711548696115, "grad_norm": 0.3019651174545288, "learning_rate": 9.056233243363182e-08, "loss": 0.2012, "step": 10138 }, { "epoch": 2.697977647684939, "grad_norm": 0.28650030493736267, "learning_rate": 9.054534782932437e-08, "loss": 0.2032, "step": 10139 }, { "epoch": 2.6982437466737625, "grad_norm": 0.29200443625450134, "learning_rate": 9.052836350021733e-08, "loss": 0.1942, "step": 10140 }, { "epoch": 2.6985098456625867, "grad_norm": 0.28584250807762146, "learning_rate": 9.051137944680498e-08, "loss": 0.2121, "step": 10141 }, { "epoch": 2.6987759446514104, "grad_norm": 0.3202466666698456, "learning_rate": 9.049439566958175e-08, "loss": 0.2097, "step": 10142 }, { "epoch": 2.699042043640234, "grad_norm": 0.32020190358161926, "learning_rate": 9.047741216904198e-08, "loss": 0.204, "step": 10143 }, { "epoch": 2.699308142629058, "grad_norm": 0.32312285900115967, "learning_rate": 9.046042894568001e-08, "loss": 0.2221, "step": 10144 }, { "epoch": 2.699574241617882, "grad_norm": 0.29811742901802063, "learning_rate": 9.044344599999018e-08, "loss": 0.1961, "step": 10145 }, { "epoch": 2.6998403406067055, "grad_norm": 0.29029759764671326, "learning_rate": 9.042646333246684e-08, "loss": 0.1871, "step": 10146 }, { "epoch": 2.7001064395955297, "grad_norm": 0.3018447458744049, "learning_rate": 9.04094809436043e-08, "loss": 0.2066, "step": 10147 }, { "epoch": 2.7003725385843533, "grad_norm": 0.2746853232383728, "learning_rate": 9.039249883389686e-08, "loss": 0.2038, "step": 10148 }, { "epoch": 2.700638637573177, "grad_norm": 0.3248746991157532, "learning_rate": 9.037551700383884e-08, "loss": 0.2111, "step": 10149 }, { "epoch": 2.700904736562001, "grad_norm": 0.2975980341434479, "learning_rate": 9.035853545392455e-08, "loss": 0.1919, "step": 10150 }, { "epoch": 2.701170835550825, "grad_norm": 0.2735188603401184, "learning_rate": 9.034155418464822e-08, "loss": 0.175, "step": 10151 }, { "epoch": 2.7014369345396485, "grad_norm": 0.27865684032440186, "learning_rate": 9.032457319650418e-08, "loss": 0.1892, "step": 10152 }, { "epoch": 2.7017030335284726, "grad_norm": 0.30809715390205383, "learning_rate": 9.030759248998669e-08, "loss": 0.2041, "step": 10153 }, { "epoch": 2.7019691325172963, "grad_norm": 0.2923114001750946, "learning_rate": 9.029061206559002e-08, "loss": 0.194, "step": 10154 }, { "epoch": 2.7022352315061204, "grad_norm": 0.33494293689727783, "learning_rate": 9.027363192380841e-08, "loss": 0.1982, "step": 10155 }, { "epoch": 2.702501330494944, "grad_norm": 0.3369814157485962, "learning_rate": 9.025665206513613e-08, "loss": 0.2245, "step": 10156 }, { "epoch": 2.702767429483768, "grad_norm": 0.3138088583946228, "learning_rate": 9.023967249006744e-08, "loss": 0.1916, "step": 10157 }, { "epoch": 2.703033528472592, "grad_norm": 0.34888193011283875, "learning_rate": 9.022269319909652e-08, "loss": 0.2053, "step": 10158 }, { "epoch": 2.7032996274614156, "grad_norm": 0.29373809695243835, "learning_rate": 9.02057141927176e-08, "loss": 0.2075, "step": 10159 }, { "epoch": 2.7035657264502397, "grad_norm": 0.29124361276626587, "learning_rate": 9.018873547142494e-08, "loss": 0.2084, "step": 10160 }, { "epoch": 2.7038318254390634, "grad_norm": 0.28446656465530396, "learning_rate": 9.017175703571268e-08, "loss": 0.1989, "step": 10161 }, { "epoch": 2.704097924427887, "grad_norm": 0.3293708562850952, "learning_rate": 9.015477888607507e-08, "loss": 0.1827, "step": 10162 }, { "epoch": 2.7043640234167112, "grad_norm": 0.28419753909111023, "learning_rate": 9.013780102300628e-08, "loss": 0.1881, "step": 10163 }, { "epoch": 2.704630122405535, "grad_norm": 0.30014875531196594, "learning_rate": 9.012082344700049e-08, "loss": 0.2127, "step": 10164 }, { "epoch": 2.7048962213943586, "grad_norm": 0.28499162197113037, "learning_rate": 9.010384615855188e-08, "loss": 0.1825, "step": 10165 }, { "epoch": 2.7051623203831827, "grad_norm": 0.2857116758823395, "learning_rate": 9.008686915815461e-08, "loss": 0.177, "step": 10166 }, { "epoch": 2.7054284193720064, "grad_norm": 0.2957897186279297, "learning_rate": 9.006989244630288e-08, "loss": 0.182, "step": 10167 }, { "epoch": 2.70569451836083, "grad_norm": 0.27865028381347656, "learning_rate": 9.005291602349077e-08, "loss": 0.1842, "step": 10168 }, { "epoch": 2.705960617349654, "grad_norm": 0.28916674852371216, "learning_rate": 9.003593989021243e-08, "loss": 0.1923, "step": 10169 }, { "epoch": 2.706226716338478, "grad_norm": 0.28975871205329895, "learning_rate": 9.001896404696204e-08, "loss": 0.1805, "step": 10170 }, { "epoch": 2.7064928153273016, "grad_norm": 0.2976357936859131, "learning_rate": 9.000198849423368e-08, "loss": 0.1845, "step": 10171 }, { "epoch": 2.7067589143161257, "grad_norm": 0.3078511655330658, "learning_rate": 8.998501323252145e-08, "loss": 0.196, "step": 10172 }, { "epoch": 2.7070250133049494, "grad_norm": 0.2482679784297943, "learning_rate": 8.996803826231951e-08, "loss": 0.1657, "step": 10173 }, { "epoch": 2.707291112293773, "grad_norm": 0.2850145697593689, "learning_rate": 8.99510635841219e-08, "loss": 0.1865, "step": 10174 }, { "epoch": 2.707557211282597, "grad_norm": 0.2850516140460968, "learning_rate": 8.993408919842274e-08, "loss": 0.1955, "step": 10175 }, { "epoch": 2.707823310271421, "grad_norm": 0.3095419704914093, "learning_rate": 8.991711510571617e-08, "loss": 0.1953, "step": 10176 }, { "epoch": 2.7080894092602446, "grad_norm": 0.2898098826408386, "learning_rate": 8.990014130649615e-08, "loss": 0.1914, "step": 10177 }, { "epoch": 2.7083555082490687, "grad_norm": 0.25847846269607544, "learning_rate": 8.98831678012568e-08, "loss": 0.174, "step": 10178 }, { "epoch": 2.7086216072378924, "grad_norm": 0.3267362117767334, "learning_rate": 8.986619459049214e-08, "loss": 0.1814, "step": 10179 }, { "epoch": 2.7088877062267165, "grad_norm": 0.29090139269828796, "learning_rate": 8.984922167469628e-08, "loss": 0.2046, "step": 10180 }, { "epoch": 2.70915380521554, "grad_norm": 0.2989956736564636, "learning_rate": 8.98322490543632e-08, "loss": 0.1905, "step": 10181 }, { "epoch": 2.7094199042043643, "grad_norm": 0.30960679054260254, "learning_rate": 8.981527672998695e-08, "loss": 0.2086, "step": 10182 }, { "epoch": 2.709686003193188, "grad_norm": 0.2980270981788635, "learning_rate": 8.979830470206155e-08, "loss": 0.2106, "step": 10183 }, { "epoch": 2.7099521021820117, "grad_norm": 0.2860088050365448, "learning_rate": 8.978133297108102e-08, "loss": 0.1967, "step": 10184 }, { "epoch": 2.710218201170836, "grad_norm": 0.29628458619117737, "learning_rate": 8.976436153753939e-08, "loss": 0.1962, "step": 10185 }, { "epoch": 2.7104843001596595, "grad_norm": 0.2568384110927582, "learning_rate": 8.974739040193058e-08, "loss": 0.1659, "step": 10186 }, { "epoch": 2.710750399148483, "grad_norm": 0.28281691670417786, "learning_rate": 8.973041956474861e-08, "loss": 0.2034, "step": 10187 }, { "epoch": 2.7110164981373073, "grad_norm": 0.3079565167427063, "learning_rate": 8.971344902648745e-08, "loss": 0.1969, "step": 10188 }, { "epoch": 2.711282597126131, "grad_norm": 0.5156909227371216, "learning_rate": 8.969647878764111e-08, "loss": 0.2118, "step": 10189 }, { "epoch": 2.7115486961149546, "grad_norm": 0.2971632480621338, "learning_rate": 8.967950884870352e-08, "loss": 0.1941, "step": 10190 }, { "epoch": 2.7118147951037788, "grad_norm": 0.29460176825523376, "learning_rate": 8.96625392101686e-08, "loss": 0.1937, "step": 10191 }, { "epoch": 2.7120808940926024, "grad_norm": 0.2923174500465393, "learning_rate": 8.964556987253034e-08, "loss": 0.1932, "step": 10192 }, { "epoch": 2.712346993081426, "grad_norm": 0.2996905744075775, "learning_rate": 8.962860083628265e-08, "loss": 0.1921, "step": 10193 }, { "epoch": 2.7126130920702503, "grad_norm": 0.32723477482795715, "learning_rate": 8.961163210191951e-08, "loss": 0.1945, "step": 10194 }, { "epoch": 2.712879191059074, "grad_norm": 0.2670600712299347, "learning_rate": 8.959466366993473e-08, "loss": 0.1738, "step": 10195 }, { "epoch": 2.7131452900478976, "grad_norm": 0.28566524386405945, "learning_rate": 8.957769554082228e-08, "loss": 0.1938, "step": 10196 }, { "epoch": 2.7134113890367217, "grad_norm": 0.2798081636428833, "learning_rate": 8.956072771507606e-08, "loss": 0.1851, "step": 10197 }, { "epoch": 2.7136774880255454, "grad_norm": 0.3016435503959656, "learning_rate": 8.954376019318994e-08, "loss": 0.2111, "step": 10198 }, { "epoch": 2.713943587014369, "grad_norm": 0.28910452127456665, "learning_rate": 8.952679297565784e-08, "loss": 0.2102, "step": 10199 }, { "epoch": 2.7142096860031932, "grad_norm": 0.29427430033683777, "learning_rate": 8.950982606297358e-08, "loss": 0.1993, "step": 10200 }, { "epoch": 2.714475784992017, "grad_norm": 0.3086397647857666, "learning_rate": 8.949285945563104e-08, "loss": 0.2145, "step": 10201 }, { "epoch": 2.7147418839808406, "grad_norm": 0.2800162732601166, "learning_rate": 8.947589315412409e-08, "loss": 0.1874, "step": 10202 }, { "epoch": 2.7150079829696647, "grad_norm": 0.3297920823097229, "learning_rate": 8.94589271589466e-08, "loss": 0.2243, "step": 10203 }, { "epoch": 2.7152740819584884, "grad_norm": 0.27470073103904724, "learning_rate": 8.944196147059233e-08, "loss": 0.1975, "step": 10204 }, { "epoch": 2.7155401809473125, "grad_norm": 0.264668345451355, "learning_rate": 8.942499608955515e-08, "loss": 0.1852, "step": 10205 }, { "epoch": 2.715806279936136, "grad_norm": 0.2837870121002197, "learning_rate": 8.940803101632887e-08, "loss": 0.2047, "step": 10206 }, { "epoch": 2.7160723789249603, "grad_norm": 0.30277949571609497, "learning_rate": 8.93910662514073e-08, "loss": 0.195, "step": 10207 }, { "epoch": 2.716338477913784, "grad_norm": 0.32103443145751953, "learning_rate": 8.937410179528426e-08, "loss": 0.2148, "step": 10208 }, { "epoch": 2.7166045769026077, "grad_norm": 0.2840186357498169, "learning_rate": 8.935713764845354e-08, "loss": 0.1881, "step": 10209 }, { "epoch": 2.716870675891432, "grad_norm": 0.33972102403640747, "learning_rate": 8.93401738114089e-08, "loss": 0.2179, "step": 10210 }, { "epoch": 2.7171367748802555, "grad_norm": 0.28731569647789, "learning_rate": 8.932321028464412e-08, "loss": 0.1922, "step": 10211 }, { "epoch": 2.717402873869079, "grad_norm": 0.28710734844207764, "learning_rate": 8.9306247068653e-08, "loss": 0.2102, "step": 10212 }, { "epoch": 2.7176689728579033, "grad_norm": 0.3038859963417053, "learning_rate": 8.928928416392921e-08, "loss": 0.2081, "step": 10213 }, { "epoch": 2.717935071846727, "grad_norm": 0.31017693877220154, "learning_rate": 8.927232157096656e-08, "loss": 0.2085, "step": 10214 }, { "epoch": 2.7182011708355507, "grad_norm": 0.31343555450439453, "learning_rate": 8.925535929025878e-08, "loss": 0.213, "step": 10215 }, { "epoch": 2.718467269824375, "grad_norm": 0.29940518736839294, "learning_rate": 8.923839732229959e-08, "loss": 0.1978, "step": 10216 }, { "epoch": 2.7187333688131985, "grad_norm": 0.285751610994339, "learning_rate": 8.92214356675827e-08, "loss": 0.1945, "step": 10217 }, { "epoch": 2.718999467802022, "grad_norm": 0.28505071997642517, "learning_rate": 8.920447432660184e-08, "loss": 0.2006, "step": 10218 }, { "epoch": 2.7192655667908463, "grad_norm": 0.2819638252258301, "learning_rate": 8.918751329985072e-08, "loss": 0.1931, "step": 10219 }, { "epoch": 2.71953166577967, "grad_norm": 0.2862566411495209, "learning_rate": 8.9170552587823e-08, "loss": 0.2012, "step": 10220 }, { "epoch": 2.7197977647684937, "grad_norm": 0.28957027196884155, "learning_rate": 8.91535921910124e-08, "loss": 0.1917, "step": 10221 }, { "epoch": 2.720063863757318, "grad_norm": 0.3069075644016266, "learning_rate": 8.913663210991256e-08, "loss": 0.212, "step": 10222 }, { "epoch": 2.7203299627461415, "grad_norm": 0.2821287214756012, "learning_rate": 8.911967234501712e-08, "loss": 0.1956, "step": 10223 }, { "epoch": 2.720596061734965, "grad_norm": 0.300630122423172, "learning_rate": 8.910271289681979e-08, "loss": 0.1777, "step": 10224 }, { "epoch": 2.7208621607237893, "grad_norm": 0.2641216516494751, "learning_rate": 8.908575376581419e-08, "loss": 0.1787, "step": 10225 }, { "epoch": 2.721128259712613, "grad_norm": 0.2676613926887512, "learning_rate": 8.906879495249396e-08, "loss": 0.1753, "step": 10226 }, { "epoch": 2.721394358701437, "grad_norm": 0.27060467004776, "learning_rate": 8.90518364573527e-08, "loss": 0.1874, "step": 10227 }, { "epoch": 2.7216604576902608, "grad_norm": 0.289274126291275, "learning_rate": 8.903487828088409e-08, "loss": 0.1958, "step": 10228 }, { "epoch": 2.721926556679085, "grad_norm": 0.30116182565689087, "learning_rate": 8.901792042358169e-08, "loss": 0.2023, "step": 10229 }, { "epoch": 2.7221926556679086, "grad_norm": 0.2986699342727661, "learning_rate": 8.900096288593913e-08, "loss": 0.2082, "step": 10230 }, { "epoch": 2.7224587546567323, "grad_norm": 0.2649588882923126, "learning_rate": 8.898400566844996e-08, "loss": 0.2008, "step": 10231 }, { "epoch": 2.7227248536455564, "grad_norm": 0.29251518845558167, "learning_rate": 8.896704877160781e-08, "loss": 0.1765, "step": 10232 }, { "epoch": 2.72299095263438, "grad_norm": 0.2860904633998871, "learning_rate": 8.89500921959062e-08, "loss": 0.1943, "step": 10233 }, { "epoch": 2.7232570516232038, "grad_norm": 0.2982080280780792, "learning_rate": 8.89331359418387e-08, "loss": 0.1991, "step": 10234 }, { "epoch": 2.723523150612028, "grad_norm": 0.2859019339084625, "learning_rate": 8.89161800098989e-08, "loss": 0.1869, "step": 10235 }, { "epoch": 2.7237892496008516, "grad_norm": 0.29872044920921326, "learning_rate": 8.88992244005803e-08, "loss": 0.1922, "step": 10236 }, { "epoch": 2.7240553485896752, "grad_norm": 0.30212822556495667, "learning_rate": 8.888226911437647e-08, "loss": 0.2026, "step": 10237 }, { "epoch": 2.7243214475784994, "grad_norm": 0.3063866198062897, "learning_rate": 8.886531415178093e-08, "loss": 0.1923, "step": 10238 }, { "epoch": 2.724587546567323, "grad_norm": 0.2946607172489166, "learning_rate": 8.884835951328718e-08, "loss": 0.1897, "step": 10239 }, { "epoch": 2.7248536455561467, "grad_norm": 0.30352476239204407, "learning_rate": 8.883140519938873e-08, "loss": 0.2147, "step": 10240 }, { "epoch": 2.725119744544971, "grad_norm": 0.279876708984375, "learning_rate": 8.881445121057909e-08, "loss": 0.2064, "step": 10241 }, { "epoch": 2.7253858435337945, "grad_norm": 0.28396856784820557, "learning_rate": 8.879749754735171e-08, "loss": 0.1857, "step": 10242 }, { "epoch": 2.725651942522618, "grad_norm": 0.31126007437705994, "learning_rate": 8.878054421020009e-08, "loss": 0.1914, "step": 10243 }, { "epoch": 2.7259180415114423, "grad_norm": 0.37919899821281433, "learning_rate": 8.876359119961769e-08, "loss": 0.1839, "step": 10244 }, { "epoch": 2.726184140500266, "grad_norm": 0.29631277918815613, "learning_rate": 8.874663851609797e-08, "loss": 0.1951, "step": 10245 }, { "epoch": 2.7264502394890897, "grad_norm": 0.28685957193374634, "learning_rate": 8.872968616013438e-08, "loss": 0.1852, "step": 10246 }, { "epoch": 2.726716338477914, "grad_norm": 0.34556153416633606, "learning_rate": 8.871273413222038e-08, "loss": 0.2249, "step": 10247 }, { "epoch": 2.7269824374667375, "grad_norm": 0.2912866771221161, "learning_rate": 8.869578243284939e-08, "loss": 0.19, "step": 10248 }, { "epoch": 2.727248536455561, "grad_norm": 0.29188910126686096, "learning_rate": 8.867883106251484e-08, "loss": 0.18, "step": 10249 }, { "epoch": 2.7275146354443853, "grad_norm": 0.29217302799224854, "learning_rate": 8.86618800217101e-08, "loss": 0.1925, "step": 10250 }, { "epoch": 2.727780734433209, "grad_norm": 0.3092314600944519, "learning_rate": 8.86449293109286e-08, "loss": 0.1895, "step": 10251 }, { "epoch": 2.728046833422033, "grad_norm": 0.2893352806568146, "learning_rate": 8.86279789306637e-08, "loss": 0.1886, "step": 10252 }, { "epoch": 2.728312932410857, "grad_norm": 0.28762122988700867, "learning_rate": 8.861102888140882e-08, "loss": 0.1956, "step": 10253 }, { "epoch": 2.728579031399681, "grad_norm": 0.2952325642108917, "learning_rate": 8.859407916365731e-08, "loss": 0.206, "step": 10254 }, { "epoch": 2.7288451303885046, "grad_norm": 0.29086917638778687, "learning_rate": 8.857712977790254e-08, "loss": 0.2069, "step": 10255 }, { "epoch": 2.7291112293773283, "grad_norm": 0.26367971301078796, "learning_rate": 8.856018072463788e-08, "loss": 0.1724, "step": 10256 }, { "epoch": 2.7293773283661524, "grad_norm": 0.31176134943962097, "learning_rate": 8.854323200435664e-08, "loss": 0.2176, "step": 10257 }, { "epoch": 2.729643427354976, "grad_norm": 0.2834617495536804, "learning_rate": 8.852628361755222e-08, "loss": 0.194, "step": 10258 }, { "epoch": 2.7299095263438, "grad_norm": 0.29235878586769104, "learning_rate": 8.850933556471783e-08, "loss": 0.1878, "step": 10259 }, { "epoch": 2.730175625332624, "grad_norm": 0.3123387396335602, "learning_rate": 8.849238784634687e-08, "loss": 0.2096, "step": 10260 }, { "epoch": 2.7304417243214476, "grad_norm": 0.2694361209869385, "learning_rate": 8.847544046293262e-08, "loss": 0.1963, "step": 10261 }, { "epoch": 2.7307078233102713, "grad_norm": 0.2649155855178833, "learning_rate": 8.845849341496839e-08, "loss": 0.185, "step": 10262 }, { "epoch": 2.7309739222990954, "grad_norm": 0.2679440379142761, "learning_rate": 8.844154670294741e-08, "loss": 0.2031, "step": 10263 }, { "epoch": 2.731240021287919, "grad_norm": 0.29176145792007446, "learning_rate": 8.8424600327363e-08, "loss": 0.201, "step": 10264 }, { "epoch": 2.7315061202767428, "grad_norm": 0.258769154548645, "learning_rate": 8.840765428870845e-08, "loss": 0.1889, "step": 10265 }, { "epoch": 2.731772219265567, "grad_norm": 0.2832702398300171, "learning_rate": 8.839070858747696e-08, "loss": 0.1826, "step": 10266 }, { "epoch": 2.7320383182543906, "grad_norm": 0.3032168745994568, "learning_rate": 8.837376322416185e-08, "loss": 0.1987, "step": 10267 }, { "epoch": 2.7323044172432143, "grad_norm": 0.31064724922180176, "learning_rate": 8.835681819925626e-08, "loss": 0.2068, "step": 10268 }, { "epoch": 2.7325705162320384, "grad_norm": 0.2784923315048218, "learning_rate": 8.833987351325345e-08, "loss": 0.1843, "step": 10269 }, { "epoch": 2.732836615220862, "grad_norm": 0.32747307419776917, "learning_rate": 8.832292916664665e-08, "loss": 0.2083, "step": 10270 }, { "epoch": 2.7331027142096858, "grad_norm": 0.2750668227672577, "learning_rate": 8.83059851599291e-08, "loss": 0.1959, "step": 10271 }, { "epoch": 2.73336881319851, "grad_norm": 0.2968287169933319, "learning_rate": 8.828904149359393e-08, "loss": 0.1858, "step": 10272 }, { "epoch": 2.7336349121873336, "grad_norm": 0.29532691836357117, "learning_rate": 8.827209816813434e-08, "loss": 0.2032, "step": 10273 }, { "epoch": 2.7339010111761577, "grad_norm": 0.30659714341163635, "learning_rate": 8.825515518404354e-08, "loss": 0.2185, "step": 10274 }, { "epoch": 2.7341671101649814, "grad_norm": 0.3012174069881439, "learning_rate": 8.823821254181467e-08, "loss": 0.1858, "step": 10275 }, { "epoch": 2.734433209153805, "grad_norm": 0.3060581386089325, "learning_rate": 8.822127024194095e-08, "loss": 0.2164, "step": 10276 }, { "epoch": 2.734699308142629, "grad_norm": 0.28247159719467163, "learning_rate": 8.820432828491541e-08, "loss": 0.2072, "step": 10277 }, { "epoch": 2.734965407131453, "grad_norm": 0.2930125296115875, "learning_rate": 8.818738667123125e-08, "loss": 0.2153, "step": 10278 }, { "epoch": 2.735231506120277, "grad_norm": 0.27953046560287476, "learning_rate": 8.81704454013816e-08, "loss": 0.1906, "step": 10279 }, { "epoch": 2.7354976051091007, "grad_norm": 0.32572177052497864, "learning_rate": 8.815350447585959e-08, "loss": 0.2071, "step": 10280 }, { "epoch": 2.7357637040979244, "grad_norm": 0.2820779085159302, "learning_rate": 8.813656389515827e-08, "loss": 0.1947, "step": 10281 }, { "epoch": 2.7360298030867485, "grad_norm": 0.29333242774009705, "learning_rate": 8.811962365977078e-08, "loss": 0.2075, "step": 10282 }, { "epoch": 2.736295902075572, "grad_norm": 0.306974858045578, "learning_rate": 8.81026837701902e-08, "loss": 0.2015, "step": 10283 }, { "epoch": 2.736562001064396, "grad_norm": 0.3002369701862335, "learning_rate": 8.808574422690961e-08, "loss": 0.2013, "step": 10284 }, { "epoch": 2.73682810005322, "grad_norm": 0.2713250517845154, "learning_rate": 8.806880503042211e-08, "loss": 0.1938, "step": 10285 }, { "epoch": 2.7370941990420437, "grad_norm": 0.6906322240829468, "learning_rate": 8.805186618122066e-08, "loss": 0.2007, "step": 10286 }, { "epoch": 2.7373602980308673, "grad_norm": 0.3161742687225342, "learning_rate": 8.803492767979838e-08, "loss": 0.2108, "step": 10287 }, { "epoch": 2.7376263970196915, "grad_norm": 0.29928112030029297, "learning_rate": 8.801798952664828e-08, "loss": 0.2066, "step": 10288 }, { "epoch": 2.737892496008515, "grad_norm": 0.2722654342651367, "learning_rate": 8.80010517222634e-08, "loss": 0.186, "step": 10289 }, { "epoch": 2.738158594997339, "grad_norm": 0.28972330689430237, "learning_rate": 8.798411426713677e-08, "loss": 0.1839, "step": 10290 }, { "epoch": 2.738424693986163, "grad_norm": 0.3166373074054718, "learning_rate": 8.796717716176134e-08, "loss": 0.2211, "step": 10291 }, { "epoch": 2.7386907929749866, "grad_norm": 0.27995118498802185, "learning_rate": 8.795024040663014e-08, "loss": 0.1764, "step": 10292 }, { "epoch": 2.7389568919638103, "grad_norm": 0.28328320384025574, "learning_rate": 8.793330400223617e-08, "loss": 0.1887, "step": 10293 }, { "epoch": 2.7392229909526344, "grad_norm": 0.2966248095035553, "learning_rate": 8.791636794907244e-08, "loss": 0.1848, "step": 10294 }, { "epoch": 2.739489089941458, "grad_norm": 0.2813519537448883, "learning_rate": 8.78994322476318e-08, "loss": 0.1957, "step": 10295 }, { "epoch": 2.739755188930282, "grad_norm": 0.3309626281261444, "learning_rate": 8.788249689840727e-08, "loss": 0.2168, "step": 10296 }, { "epoch": 2.740021287919106, "grad_norm": 0.28506597876548767, "learning_rate": 8.786556190189181e-08, "loss": 0.1861, "step": 10297 }, { "epoch": 2.7402873869079296, "grad_norm": 0.3158491849899292, "learning_rate": 8.784862725857832e-08, "loss": 0.1874, "step": 10298 }, { "epoch": 2.7405534858967537, "grad_norm": 0.3216533064842224, "learning_rate": 8.783169296895976e-08, "loss": 0.2191, "step": 10299 }, { "epoch": 2.7408195848855774, "grad_norm": 0.287254273891449, "learning_rate": 8.781475903352901e-08, "loss": 0.1929, "step": 10300 }, { "epoch": 2.7410856838744015, "grad_norm": 0.28902024030685425, "learning_rate": 8.779782545277899e-08, "loss": 0.1978, "step": 10301 }, { "epoch": 2.7413517828632252, "grad_norm": 0.28916478157043457, "learning_rate": 8.778089222720259e-08, "loss": 0.1916, "step": 10302 }, { "epoch": 2.741617881852049, "grad_norm": 0.274957537651062, "learning_rate": 8.776395935729273e-08, "loss": 0.1741, "step": 10303 }, { "epoch": 2.741883980840873, "grad_norm": 0.2701354920864105, "learning_rate": 8.774702684354219e-08, "loss": 0.1649, "step": 10304 }, { "epoch": 2.7421500798296967, "grad_norm": 0.3012195825576782, "learning_rate": 8.77300946864439e-08, "loss": 0.2, "step": 10305 }, { "epoch": 2.7424161788185204, "grad_norm": 0.2937614619731903, "learning_rate": 8.771316288649067e-08, "loss": 0.1885, "step": 10306 }, { "epoch": 2.7426822778073445, "grad_norm": 0.27141156792640686, "learning_rate": 8.769623144417539e-08, "loss": 0.1961, "step": 10307 }, { "epoch": 2.742948376796168, "grad_norm": 0.29926908016204834, "learning_rate": 8.767930035999086e-08, "loss": 0.193, "step": 10308 }, { "epoch": 2.743214475784992, "grad_norm": 0.30240824818611145, "learning_rate": 8.76623696344299e-08, "loss": 0.2145, "step": 10309 }, { "epoch": 2.743480574773816, "grad_norm": 0.30408504605293274, "learning_rate": 8.764543926798535e-08, "loss": 0.1971, "step": 10310 }, { "epoch": 2.7437466737626397, "grad_norm": 0.2979758679866791, "learning_rate": 8.762850926114996e-08, "loss": 0.2055, "step": 10311 }, { "epoch": 2.7440127727514634, "grad_norm": 0.2884270250797272, "learning_rate": 8.761157961441658e-08, "loss": 0.2006, "step": 10312 }, { "epoch": 2.7442788717402875, "grad_norm": 0.2889450490474701, "learning_rate": 8.759465032827793e-08, "loss": 0.1836, "step": 10313 }, { "epoch": 2.744544970729111, "grad_norm": 0.2865791618824005, "learning_rate": 8.757772140322679e-08, "loss": 0.1841, "step": 10314 }, { "epoch": 2.744811069717935, "grad_norm": 0.32105639576911926, "learning_rate": 8.756079283975593e-08, "loss": 0.2064, "step": 10315 }, { "epoch": 2.745077168706759, "grad_norm": 0.3251301646232605, "learning_rate": 8.754386463835807e-08, "loss": 0.2017, "step": 10316 }, { "epoch": 2.7453432676955827, "grad_norm": 0.3232385218143463, "learning_rate": 8.752693679952599e-08, "loss": 0.2008, "step": 10317 }, { "epoch": 2.7456093666844064, "grad_norm": 0.297525018453598, "learning_rate": 8.751000932375239e-08, "loss": 0.2, "step": 10318 }, { "epoch": 2.7458754656732305, "grad_norm": 0.306011825799942, "learning_rate": 8.749308221152998e-08, "loss": 0.1821, "step": 10319 }, { "epoch": 2.746141564662054, "grad_norm": 0.28719040751457214, "learning_rate": 8.747615546335151e-08, "loss": 0.1974, "step": 10320 }, { "epoch": 2.746407663650878, "grad_norm": 0.2865429222583771, "learning_rate": 8.745922907970961e-08, "loss": 0.1847, "step": 10321 }, { "epoch": 2.746673762639702, "grad_norm": 0.28947752714157104, "learning_rate": 8.7442303061097e-08, "loss": 0.1946, "step": 10322 }, { "epoch": 2.7469398616285257, "grad_norm": 0.29588159918785095, "learning_rate": 8.742537740800636e-08, "loss": 0.1969, "step": 10323 }, { "epoch": 2.74720596061735, "grad_norm": 0.2579851448535919, "learning_rate": 8.740845212093028e-08, "loss": 0.1866, "step": 10324 }, { "epoch": 2.7474720596061735, "grad_norm": 0.28237345814704895, "learning_rate": 8.73915272003615e-08, "loss": 0.1819, "step": 10325 }, { "epoch": 2.7477381585949976, "grad_norm": 0.3040904998779297, "learning_rate": 8.737460264679261e-08, "loss": 0.2006, "step": 10326 }, { "epoch": 2.7480042575838213, "grad_norm": 0.2801330089569092, "learning_rate": 8.735767846071626e-08, "loss": 0.1828, "step": 10327 }, { "epoch": 2.748270356572645, "grad_norm": 0.2945619225502014, "learning_rate": 8.734075464262506e-08, "loss": 0.2002, "step": 10328 }, { "epoch": 2.748536455561469, "grad_norm": 0.2866697311401367, "learning_rate": 8.732383119301164e-08, "loss": 0.1893, "step": 10329 }, { "epoch": 2.7488025545502928, "grad_norm": 0.3054297864437103, "learning_rate": 8.73069081123686e-08, "loss": 0.202, "step": 10330 }, { "epoch": 2.7490686535391164, "grad_norm": 0.28041261434555054, "learning_rate": 8.728998540118845e-08, "loss": 0.2042, "step": 10331 }, { "epoch": 2.7493347525279406, "grad_norm": 0.2899165153503418, "learning_rate": 8.727306305996388e-08, "loss": 0.176, "step": 10332 }, { "epoch": 2.7496008515167643, "grad_norm": 0.2984828054904938, "learning_rate": 8.725614108918737e-08, "loss": 0.1994, "step": 10333 }, { "epoch": 2.749866950505588, "grad_norm": 0.28468945622444153, "learning_rate": 8.72392194893515e-08, "loss": 0.1795, "step": 10334 }, { "epoch": 2.750133049494412, "grad_norm": 0.283061146736145, "learning_rate": 8.722229826094881e-08, "loss": 0.1788, "step": 10335 }, { "epoch": 2.7503991484832357, "grad_norm": 0.2747710645198822, "learning_rate": 8.720537740447185e-08, "loss": 0.1885, "step": 10336 }, { "epoch": 2.7506652474720594, "grad_norm": 0.2963947355747223, "learning_rate": 8.718845692041313e-08, "loss": 0.2192, "step": 10337 }, { "epoch": 2.7509313464608836, "grad_norm": 0.30450278520584106, "learning_rate": 8.717153680926517e-08, "loss": 0.19, "step": 10338 }, { "epoch": 2.7511974454497072, "grad_norm": 0.29717743396759033, "learning_rate": 8.71546170715205e-08, "loss": 0.2056, "step": 10339 }, { "epoch": 2.751463544438531, "grad_norm": 0.26093533635139465, "learning_rate": 8.713769770767154e-08, "loss": 0.1719, "step": 10340 }, { "epoch": 2.751729643427355, "grad_norm": 0.2881002128124237, "learning_rate": 8.71207787182108e-08, "loss": 0.1949, "step": 10341 }, { "epoch": 2.7519957424161787, "grad_norm": 0.3062226176261902, "learning_rate": 8.710386010363077e-08, "loss": 0.1895, "step": 10342 }, { "epoch": 2.7522618414050024, "grad_norm": 0.2852414548397064, "learning_rate": 8.708694186442387e-08, "loss": 0.2105, "step": 10343 }, { "epoch": 2.7525279403938265, "grad_norm": 0.2924793064594269, "learning_rate": 8.707002400108255e-08, "loss": 0.2071, "step": 10344 }, { "epoch": 2.75279403938265, "grad_norm": 0.27949944138526917, "learning_rate": 8.705310651409927e-08, "loss": 0.1828, "step": 10345 }, { "epoch": 2.7530601383714743, "grad_norm": 0.2920910716056824, "learning_rate": 8.703618940396645e-08, "loss": 0.1811, "step": 10346 }, { "epoch": 2.753326237360298, "grad_norm": 0.279622882604599, "learning_rate": 8.701927267117648e-08, "loss": 0.1778, "step": 10347 }, { "epoch": 2.753592336349122, "grad_norm": 0.284160315990448, "learning_rate": 8.700235631622184e-08, "loss": 0.1966, "step": 10348 }, { "epoch": 2.753858435337946, "grad_norm": 0.37388190627098083, "learning_rate": 8.69854403395948e-08, "loss": 0.1985, "step": 10349 }, { "epoch": 2.7541245343267695, "grad_norm": 0.3090662658214569, "learning_rate": 8.69685247417878e-08, "loss": 0.2032, "step": 10350 }, { "epoch": 2.7543906333155936, "grad_norm": 0.28579315543174744, "learning_rate": 8.695160952329322e-08, "loss": 0.182, "step": 10351 }, { "epoch": 2.7546567323044173, "grad_norm": 0.2861779034137726, "learning_rate": 8.693469468460343e-08, "loss": 0.1842, "step": 10352 }, { "epoch": 2.754922831293241, "grad_norm": 0.2688957750797272, "learning_rate": 8.691778022621073e-08, "loss": 0.1725, "step": 10353 }, { "epoch": 2.755188930282065, "grad_norm": 0.2772398889064789, "learning_rate": 8.690086614860747e-08, "loss": 0.1821, "step": 10354 }, { "epoch": 2.755455029270889, "grad_norm": 0.29201966524124146, "learning_rate": 8.6883952452286e-08, "loss": 0.195, "step": 10355 }, { "epoch": 2.7557211282597125, "grad_norm": 0.3081376552581787, "learning_rate": 8.68670391377386e-08, "loss": 0.2104, "step": 10356 }, { "epoch": 2.7559872272485366, "grad_norm": 0.2659986615180969, "learning_rate": 8.685012620545764e-08, "loss": 0.1797, "step": 10357 }, { "epoch": 2.7562533262373603, "grad_norm": 0.3081170618534088, "learning_rate": 8.683321365593533e-08, "loss": 0.2305, "step": 10358 }, { "epoch": 2.756519425226184, "grad_norm": 0.28917378187179565, "learning_rate": 8.681630148966397e-08, "loss": 0.1915, "step": 10359 }, { "epoch": 2.756785524215008, "grad_norm": 0.354388564825058, "learning_rate": 8.679938970713585e-08, "loss": 0.1963, "step": 10360 }, { "epoch": 2.757051623203832, "grad_norm": 0.30508825182914734, "learning_rate": 8.678247830884322e-08, "loss": 0.1953, "step": 10361 }, { "epoch": 2.7573177221926555, "grad_norm": 0.31049177050590515, "learning_rate": 8.676556729527835e-08, "loss": 0.2082, "step": 10362 }, { "epoch": 2.7575838211814796, "grad_norm": 0.3806540369987488, "learning_rate": 8.674865666693343e-08, "loss": 0.199, "step": 10363 }, { "epoch": 2.7578499201703033, "grad_norm": 0.3029177188873291, "learning_rate": 8.673174642430071e-08, "loss": 0.2049, "step": 10364 }, { "epoch": 2.758116019159127, "grad_norm": 0.2732994556427002, "learning_rate": 8.67148365678724e-08, "loss": 0.1832, "step": 10365 }, { "epoch": 2.758382118147951, "grad_norm": 0.30758029222488403, "learning_rate": 8.669792709814074e-08, "loss": 0.1941, "step": 10366 }, { "epoch": 2.7586482171367748, "grad_norm": 0.3089453876018524, "learning_rate": 8.668101801559785e-08, "loss": 0.1858, "step": 10367 }, { "epoch": 2.7589143161255985, "grad_norm": 0.2957686185836792, "learning_rate": 8.666410932073593e-08, "loss": 0.1868, "step": 10368 }, { "epoch": 2.7591804151144226, "grad_norm": 0.2807006239891052, "learning_rate": 8.664720101404718e-08, "loss": 0.1894, "step": 10369 }, { "epoch": 2.7594465141032463, "grad_norm": 0.30016303062438965, "learning_rate": 8.663029309602374e-08, "loss": 0.198, "step": 10370 }, { "epoch": 2.7597126130920704, "grad_norm": 0.28926095366477966, "learning_rate": 8.661338556715774e-08, "loss": 0.1962, "step": 10371 }, { "epoch": 2.759978712080894, "grad_norm": 0.30937451124191284, "learning_rate": 8.659647842794134e-08, "loss": 0.2155, "step": 10372 }, { "epoch": 2.760244811069718, "grad_norm": 0.31505727767944336, "learning_rate": 8.657957167886664e-08, "loss": 0.2081, "step": 10373 }, { "epoch": 2.760510910058542, "grad_norm": 0.3081565499305725, "learning_rate": 8.656266532042578e-08, "loss": 0.2009, "step": 10374 }, { "epoch": 2.7607770090473656, "grad_norm": 0.29155632853507996, "learning_rate": 8.654575935311086e-08, "loss": 0.1876, "step": 10375 }, { "epoch": 2.7610431080361897, "grad_norm": 0.3247210383415222, "learning_rate": 8.652885377741392e-08, "loss": 0.2048, "step": 10376 }, { "epoch": 2.7613092070250134, "grad_norm": 0.28399354219436646, "learning_rate": 8.651194859382707e-08, "loss": 0.1955, "step": 10377 }, { "epoch": 2.761575306013837, "grad_norm": 0.2995959520339966, "learning_rate": 8.649504380284237e-08, "loss": 0.1887, "step": 10378 }, { "epoch": 2.761841405002661, "grad_norm": 0.2910744845867157, "learning_rate": 8.647813940495188e-08, "loss": 0.1968, "step": 10379 }, { "epoch": 2.762107503991485, "grad_norm": 0.2717624008655548, "learning_rate": 8.646123540064764e-08, "loss": 0.1878, "step": 10380 }, { "epoch": 2.7623736029803085, "grad_norm": 0.2851232588291168, "learning_rate": 8.644433179042172e-08, "loss": 0.1806, "step": 10381 }, { "epoch": 2.7626397019691327, "grad_norm": 0.300333172082901, "learning_rate": 8.642742857476608e-08, "loss": 0.1982, "step": 10382 }, { "epoch": 2.7629058009579563, "grad_norm": 0.2942471504211426, "learning_rate": 8.641052575417274e-08, "loss": 0.189, "step": 10383 }, { "epoch": 2.76317189994678, "grad_norm": 0.36712953448295593, "learning_rate": 8.639362332913377e-08, "loss": 0.1906, "step": 10384 }, { "epoch": 2.763437998935604, "grad_norm": 0.2798396646976471, "learning_rate": 8.637672130014104e-08, "loss": 0.1885, "step": 10385 }, { "epoch": 2.763704097924428, "grad_norm": 0.29142647981643677, "learning_rate": 8.635981966768659e-08, "loss": 0.2095, "step": 10386 }, { "epoch": 2.7639701969132515, "grad_norm": 0.2801045775413513, "learning_rate": 8.634291843226237e-08, "loss": 0.1797, "step": 10387 }, { "epoch": 2.7642362959020756, "grad_norm": 0.28928473591804504, "learning_rate": 8.632601759436033e-08, "loss": 0.2001, "step": 10388 }, { "epoch": 2.7645023948908993, "grad_norm": 0.29467839002609253, "learning_rate": 8.630911715447242e-08, "loss": 0.2027, "step": 10389 }, { "epoch": 2.764768493879723, "grad_norm": 0.28105202317237854, "learning_rate": 8.629221711309055e-08, "loss": 0.1978, "step": 10390 }, { "epoch": 2.765034592868547, "grad_norm": 0.2921455204486847, "learning_rate": 8.627531747070667e-08, "loss": 0.1776, "step": 10391 }, { "epoch": 2.765300691857371, "grad_norm": 0.2693135738372803, "learning_rate": 8.625841822781266e-08, "loss": 0.1693, "step": 10392 }, { "epoch": 2.765566790846195, "grad_norm": 0.27451154589653015, "learning_rate": 8.624151938490043e-08, "loss": 0.1766, "step": 10393 }, { "epoch": 2.7658328898350186, "grad_norm": 0.2948237657546997, "learning_rate": 8.622462094246184e-08, "loss": 0.1989, "step": 10394 }, { "epoch": 2.7660989888238423, "grad_norm": 0.2871150076389313, "learning_rate": 8.620772290098874e-08, "loss": 0.1903, "step": 10395 }, { "epoch": 2.7663650878126664, "grad_norm": 0.2878125309944153, "learning_rate": 8.619082526097303e-08, "loss": 0.2134, "step": 10396 }, { "epoch": 2.76663118680149, "grad_norm": 0.2766880691051483, "learning_rate": 8.617392802290653e-08, "loss": 0.1749, "step": 10397 }, { "epoch": 2.7668972857903142, "grad_norm": 0.2779392898082733, "learning_rate": 8.615703118728109e-08, "loss": 0.1876, "step": 10398 }, { "epoch": 2.767163384779138, "grad_norm": 0.29112479090690613, "learning_rate": 8.614013475458854e-08, "loss": 0.1822, "step": 10399 }, { "epoch": 2.7674294837679616, "grad_norm": 0.29855191707611084, "learning_rate": 8.612323872532066e-08, "loss": 0.2035, "step": 10400 }, { "epoch": 2.7676955827567857, "grad_norm": 0.26803526282310486, "learning_rate": 8.610634309996932e-08, "loss": 0.1817, "step": 10401 }, { "epoch": 2.7679616817456094, "grad_norm": 0.28624892234802246, "learning_rate": 8.608944787902623e-08, "loss": 0.1728, "step": 10402 }, { "epoch": 2.768227780734433, "grad_norm": 0.3083365857601166, "learning_rate": 8.607255306298317e-08, "loss": 0.1949, "step": 10403 }, { "epoch": 2.7684938797232572, "grad_norm": 0.2597641348838806, "learning_rate": 8.605565865233197e-08, "loss": 0.175, "step": 10404 }, { "epoch": 2.768759978712081, "grad_norm": 0.2812839448451996, "learning_rate": 8.603876464756432e-08, "loss": 0.1899, "step": 10405 }, { "epoch": 2.7690260777009046, "grad_norm": 0.2658478319644928, "learning_rate": 8.602187104917197e-08, "loss": 0.1754, "step": 10406 }, { "epoch": 2.7692921766897287, "grad_norm": 0.2939321994781494, "learning_rate": 8.600497785764667e-08, "loss": 0.201, "step": 10407 }, { "epoch": 2.7695582756785524, "grad_norm": 0.2767641842365265, "learning_rate": 8.598808507348011e-08, "loss": 0.1816, "step": 10408 }, { "epoch": 2.769824374667376, "grad_norm": 0.29196369647979736, "learning_rate": 8.597119269716402e-08, "loss": 0.1989, "step": 10409 }, { "epoch": 2.7700904736562, "grad_norm": 0.288913369178772, "learning_rate": 8.595430072919009e-08, "loss": 0.1932, "step": 10410 }, { "epoch": 2.770356572645024, "grad_norm": 0.2918581962585449, "learning_rate": 8.593740917005003e-08, "loss": 0.1997, "step": 10411 }, { "epoch": 2.7706226716338476, "grad_norm": 0.29259055852890015, "learning_rate": 8.592051802023545e-08, "loss": 0.1782, "step": 10412 }, { "epoch": 2.7708887706226717, "grad_norm": 0.2913621962070465, "learning_rate": 8.590362728023801e-08, "loss": 0.1935, "step": 10413 }, { "epoch": 2.7711548696114954, "grad_norm": 0.29762303829193115, "learning_rate": 8.588673695054941e-08, "loss": 0.2168, "step": 10414 }, { "epoch": 2.771420968600319, "grad_norm": 0.3415980935096741, "learning_rate": 8.586984703166125e-08, "loss": 0.1985, "step": 10415 }, { "epoch": 2.771687067589143, "grad_norm": 0.27345380187034607, "learning_rate": 8.585295752406513e-08, "loss": 0.1777, "step": 10416 }, { "epoch": 2.771953166577967, "grad_norm": 0.28732240200042725, "learning_rate": 8.583606842825268e-08, "loss": 0.194, "step": 10417 }, { "epoch": 2.772219265566791, "grad_norm": 0.2904008626937866, "learning_rate": 8.581917974471552e-08, "loss": 0.1865, "step": 10418 }, { "epoch": 2.7724853645556147, "grad_norm": 0.2966425120830536, "learning_rate": 8.580229147394522e-08, "loss": 0.1872, "step": 10419 }, { "epoch": 2.772751463544439, "grad_norm": 0.28320661187171936, "learning_rate": 8.578540361643338e-08, "loss": 0.1807, "step": 10420 }, { "epoch": 2.7730175625332625, "grad_norm": 0.2960817813873291, "learning_rate": 8.576851617267149e-08, "loss": 0.2018, "step": 10421 }, { "epoch": 2.773283661522086, "grad_norm": 0.26994484663009644, "learning_rate": 8.575162914315115e-08, "loss": 0.1771, "step": 10422 }, { "epoch": 2.7735497605109103, "grad_norm": 0.33311930298805237, "learning_rate": 8.573474252836389e-08, "loss": 0.2069, "step": 10423 }, { "epoch": 2.773815859499734, "grad_norm": 0.36389073729515076, "learning_rate": 8.571785632880127e-08, "loss": 0.2169, "step": 10424 }, { "epoch": 2.7740819584885577, "grad_norm": 0.2858095169067383, "learning_rate": 8.570097054495472e-08, "loss": 0.1709, "step": 10425 }, { "epoch": 2.774348057477382, "grad_norm": 0.3100167214870453, "learning_rate": 8.56840851773158e-08, "loss": 0.201, "step": 10426 }, { "epoch": 2.7746141564662055, "grad_norm": 0.2720201909542084, "learning_rate": 8.5667200226376e-08, "loss": 0.186, "step": 10427 }, { "epoch": 2.774880255455029, "grad_norm": 0.2796422839164734, "learning_rate": 8.565031569262678e-08, "loss": 0.1861, "step": 10428 }, { "epoch": 2.7751463544438533, "grad_norm": 0.30857422947883606, "learning_rate": 8.563343157655964e-08, "loss": 0.1942, "step": 10429 }, { "epoch": 2.775412453432677, "grad_norm": 0.3165830671787262, "learning_rate": 8.561654787866599e-08, "loss": 0.2245, "step": 10430 }, { "epoch": 2.7756785524215006, "grad_norm": 0.3094025254249573, "learning_rate": 8.559966459943726e-08, "loss": 0.2139, "step": 10431 }, { "epoch": 2.7759446514103248, "grad_norm": 0.3246416449546814, "learning_rate": 8.558278173936491e-08, "loss": 0.1933, "step": 10432 }, { "epoch": 2.7762107503991484, "grad_norm": 0.30562537908554077, "learning_rate": 8.556589929894038e-08, "loss": 0.1892, "step": 10433 }, { "epoch": 2.776476849387972, "grad_norm": 0.32253679633140564, "learning_rate": 8.554901727865502e-08, "loss": 0.201, "step": 10434 }, { "epoch": 2.7767429483767962, "grad_norm": 0.27668777108192444, "learning_rate": 8.553213567900023e-08, "loss": 0.1813, "step": 10435 }, { "epoch": 2.77700904736562, "grad_norm": 0.27104493975639343, "learning_rate": 8.551525450046743e-08, "loss": 0.1859, "step": 10436 }, { "epoch": 2.7772751463544436, "grad_norm": 0.2759884297847748, "learning_rate": 8.549837374354796e-08, "loss": 0.1868, "step": 10437 }, { "epoch": 2.7775412453432677, "grad_norm": 0.3313787877559662, "learning_rate": 8.54814934087332e-08, "loss": 0.2311, "step": 10438 }, { "epoch": 2.7778073443320914, "grad_norm": 0.2923206090927124, "learning_rate": 8.546461349651444e-08, "loss": 0.195, "step": 10439 }, { "epoch": 2.778073443320915, "grad_norm": 0.31058064103126526, "learning_rate": 8.544773400738303e-08, "loss": 0.1898, "step": 10440 }, { "epoch": 2.7783395423097392, "grad_norm": 0.3378142714500427, "learning_rate": 8.543085494183031e-08, "loss": 0.2055, "step": 10441 }, { "epoch": 2.778605641298563, "grad_norm": 0.2813400328159332, "learning_rate": 8.541397630034757e-08, "loss": 0.1836, "step": 10442 }, { "epoch": 2.778871740287387, "grad_norm": 0.29045993089675903, "learning_rate": 8.539709808342613e-08, "loss": 0.1807, "step": 10443 }, { "epoch": 2.7791378392762107, "grad_norm": 0.3118099868297577, "learning_rate": 8.538022029155723e-08, "loss": 0.1999, "step": 10444 }, { "epoch": 2.779403938265035, "grad_norm": 0.30540037155151367, "learning_rate": 8.536334292523215e-08, "loss": 0.2025, "step": 10445 }, { "epoch": 2.7796700372538585, "grad_norm": 0.27929598093032837, "learning_rate": 8.534646598494216e-08, "loss": 0.1795, "step": 10446 }, { "epoch": 2.779936136242682, "grad_norm": 0.30853137373924255, "learning_rate": 8.532958947117854e-08, "loss": 0.1905, "step": 10447 }, { "epoch": 2.7802022352315063, "grad_norm": 0.30660295486450195, "learning_rate": 8.531271338443244e-08, "loss": 0.1865, "step": 10448 }, { "epoch": 2.78046833422033, "grad_norm": 0.2872042953968048, "learning_rate": 8.529583772519511e-08, "loss": 0.1924, "step": 10449 }, { "epoch": 2.7807344332091537, "grad_norm": 0.34421488642692566, "learning_rate": 8.527896249395776e-08, "loss": 0.1884, "step": 10450 }, { "epoch": 2.781000532197978, "grad_norm": 0.30429932475090027, "learning_rate": 8.52620876912116e-08, "loss": 0.1999, "step": 10451 }, { "epoch": 2.7812666311868015, "grad_norm": 0.2970145046710968, "learning_rate": 8.524521331744779e-08, "loss": 0.2016, "step": 10452 }, { "epoch": 2.781532730175625, "grad_norm": 0.29658177495002747, "learning_rate": 8.522833937315755e-08, "loss": 0.1838, "step": 10453 }, { "epoch": 2.7817988291644493, "grad_norm": 0.30237406492233276, "learning_rate": 8.521146585883197e-08, "loss": 0.1888, "step": 10454 }, { "epoch": 2.782064928153273, "grad_norm": 0.2838260233402252, "learning_rate": 8.51945927749622e-08, "loss": 0.1891, "step": 10455 }, { "epoch": 2.7823310271420967, "grad_norm": 0.30452242493629456, "learning_rate": 8.517772012203945e-08, "loss": 0.2043, "step": 10456 }, { "epoch": 2.782597126130921, "grad_norm": 0.30314502120018005, "learning_rate": 8.516084790055474e-08, "loss": 0.1978, "step": 10457 }, { "epoch": 2.7828632251197445, "grad_norm": 0.2911030948162079, "learning_rate": 8.514397611099921e-08, "loss": 0.1874, "step": 10458 }, { "epoch": 2.783129324108568, "grad_norm": 0.33584773540496826, "learning_rate": 8.512710475386398e-08, "loss": 0.2096, "step": 10459 }, { "epoch": 2.7833954230973923, "grad_norm": 0.29037582874298096, "learning_rate": 8.51102338296401e-08, "loss": 0.1828, "step": 10460 }, { "epoch": 2.783661522086216, "grad_norm": 0.29344940185546875, "learning_rate": 8.509336333881866e-08, "loss": 0.1932, "step": 10461 }, { "epoch": 2.7839276210750397, "grad_norm": 0.2975143492221832, "learning_rate": 8.50764932818907e-08, "loss": 0.2124, "step": 10462 }, { "epoch": 2.784193720063864, "grad_norm": 0.3196769058704376, "learning_rate": 8.505962365934729e-08, "loss": 0.2024, "step": 10463 }, { "epoch": 2.7844598190526875, "grad_norm": 0.28905627131462097, "learning_rate": 8.504275447167944e-08, "loss": 0.1835, "step": 10464 }, { "epoch": 2.7847259180415116, "grad_norm": 0.28585782647132874, "learning_rate": 8.502588571937818e-08, "loss": 0.1908, "step": 10465 }, { "epoch": 2.7849920170303353, "grad_norm": 0.29508110880851746, "learning_rate": 8.50090174029345e-08, "loss": 0.2134, "step": 10466 }, { "epoch": 2.7852581160191594, "grad_norm": 0.3735520839691162, "learning_rate": 8.499214952283938e-08, "loss": 0.1793, "step": 10467 }, { "epoch": 2.785524215007983, "grad_norm": 0.28787365555763245, "learning_rate": 8.497528207958384e-08, "loss": 0.2108, "step": 10468 }, { "epoch": 2.7857903139968068, "grad_norm": 0.2743563652038574, "learning_rate": 8.49584150736588e-08, "loss": 0.1834, "step": 10469 }, { "epoch": 2.786056412985631, "grad_norm": 0.30151018500328064, "learning_rate": 8.494154850555526e-08, "loss": 0.1825, "step": 10470 }, { "epoch": 2.7863225119744546, "grad_norm": 0.3059569001197815, "learning_rate": 8.492468237576414e-08, "loss": 0.2065, "step": 10471 }, { "epoch": 2.7865886109632783, "grad_norm": 0.3080364763736725, "learning_rate": 8.49078166847764e-08, "loss": 0.2154, "step": 10472 }, { "epoch": 2.7868547099521024, "grad_norm": 0.30589455366134644, "learning_rate": 8.48909514330829e-08, "loss": 0.1897, "step": 10473 }, { "epoch": 2.787120808940926, "grad_norm": 0.3172919452190399, "learning_rate": 8.487408662117461e-08, "loss": 0.2044, "step": 10474 }, { "epoch": 2.7873869079297497, "grad_norm": 0.31553953886032104, "learning_rate": 8.485722224954235e-08, "loss": 0.1925, "step": 10475 }, { "epoch": 2.787653006918574, "grad_norm": 0.2818363606929779, "learning_rate": 8.484035831867705e-08, "loss": 0.1911, "step": 10476 }, { "epoch": 2.7879191059073976, "grad_norm": 0.29933151602745056, "learning_rate": 8.482349482906953e-08, "loss": 0.1812, "step": 10477 }, { "epoch": 2.7881852048962212, "grad_norm": 0.2665409743785858, "learning_rate": 8.480663178121068e-08, "loss": 0.1985, "step": 10478 }, { "epoch": 2.7884513038850454, "grad_norm": 0.2933849096298218, "learning_rate": 8.478976917559134e-08, "loss": 0.1725, "step": 10479 }, { "epoch": 2.788717402873869, "grad_norm": 0.30563005805015564, "learning_rate": 8.477290701270231e-08, "loss": 0.1901, "step": 10480 }, { "epoch": 2.7889835018626927, "grad_norm": 0.31839263439178467, "learning_rate": 8.475604529303442e-08, "loss": 0.1847, "step": 10481 }, { "epoch": 2.789249600851517, "grad_norm": 0.30711889266967773, "learning_rate": 8.473918401707851e-08, "loss": 0.2088, "step": 10482 }, { "epoch": 2.7895156998403405, "grad_norm": 0.2850828766822815, "learning_rate": 8.472232318532529e-08, "loss": 0.1782, "step": 10483 }, { "epoch": 2.789781798829164, "grad_norm": 0.28078794479370117, "learning_rate": 8.47054627982656e-08, "loss": 0.193, "step": 10484 }, { "epoch": 2.7900478978179883, "grad_norm": 0.28760626912117004, "learning_rate": 8.468860285639018e-08, "loss": 0.1915, "step": 10485 }, { "epoch": 2.790313996806812, "grad_norm": 0.2883279621601105, "learning_rate": 8.467174336018976e-08, "loss": 0.1934, "step": 10486 }, { "epoch": 2.7905800957956357, "grad_norm": 0.3148133158683777, "learning_rate": 8.465488431015508e-08, "loss": 0.2297, "step": 10487 }, { "epoch": 2.79084619478446, "grad_norm": 0.29984137415885925, "learning_rate": 8.463802570677688e-08, "loss": 0.2067, "step": 10488 }, { "epoch": 2.7911122937732835, "grad_norm": 0.27342700958251953, "learning_rate": 8.462116755054587e-08, "loss": 0.1806, "step": 10489 }, { "epoch": 2.7913783927621076, "grad_norm": 0.28995588421821594, "learning_rate": 8.460430984195273e-08, "loss": 0.1908, "step": 10490 }, { "epoch": 2.7916444917509313, "grad_norm": 0.3336830139160156, "learning_rate": 8.458745258148816e-08, "loss": 0.2025, "step": 10491 }, { "epoch": 2.7919105907397554, "grad_norm": 0.28300127387046814, "learning_rate": 8.457059576964285e-08, "loss": 0.1761, "step": 10492 }, { "epoch": 2.792176689728579, "grad_norm": 0.2924846410751343, "learning_rate": 8.455373940690744e-08, "loss": 0.1923, "step": 10493 }, { "epoch": 2.792442788717403, "grad_norm": 0.2922496199607849, "learning_rate": 8.453688349377253e-08, "loss": 0.1923, "step": 10494 }, { "epoch": 2.792708887706227, "grad_norm": 0.2995860278606415, "learning_rate": 8.452002803072883e-08, "loss": 0.2196, "step": 10495 }, { "epoch": 2.7929749866950506, "grad_norm": 0.29447415471076965, "learning_rate": 8.45031730182669e-08, "loss": 0.1852, "step": 10496 }, { "epoch": 2.7932410856838743, "grad_norm": 0.2994176149368286, "learning_rate": 8.448631845687736e-08, "loss": 0.1905, "step": 10497 }, { "epoch": 2.7935071846726984, "grad_norm": 0.2895939350128174, "learning_rate": 8.446946434705082e-08, "loss": 0.1816, "step": 10498 }, { "epoch": 2.793773283661522, "grad_norm": 0.31300944089889526, "learning_rate": 8.445261068927784e-08, "loss": 0.1861, "step": 10499 }, { "epoch": 2.794039382650346, "grad_norm": 0.2893325984477997, "learning_rate": 8.443575748404899e-08, "loss": 0.1919, "step": 10500 }, { "epoch": 2.79430548163917, "grad_norm": 0.27077528834342957, "learning_rate": 8.441890473185483e-08, "loss": 0.1957, "step": 10501 }, { "epoch": 2.7945715806279936, "grad_norm": 0.30324774980545044, "learning_rate": 8.440205243318594e-08, "loss": 0.1938, "step": 10502 }, { "epoch": 2.7948376796168173, "grad_norm": 0.3216325640678406, "learning_rate": 8.438520058853275e-08, "loss": 0.1981, "step": 10503 }, { "epoch": 2.7951037786056414, "grad_norm": 0.31974536180496216, "learning_rate": 8.436834919838585e-08, "loss": 0.2064, "step": 10504 }, { "epoch": 2.795369877594465, "grad_norm": 0.28026923537254333, "learning_rate": 8.435149826323573e-08, "loss": 0.1805, "step": 10505 }, { "epoch": 2.7956359765832888, "grad_norm": 0.2877550423145294, "learning_rate": 8.433464778357284e-08, "loss": 0.1885, "step": 10506 }, { "epoch": 2.795902075572113, "grad_norm": 0.28084540367126465, "learning_rate": 8.431779775988768e-08, "loss": 0.1878, "step": 10507 }, { "epoch": 2.7961681745609366, "grad_norm": 0.28806668519973755, "learning_rate": 8.430094819267071e-08, "loss": 0.1826, "step": 10508 }, { "epoch": 2.7964342735497603, "grad_norm": 0.37246957421302795, "learning_rate": 8.428409908241238e-08, "loss": 0.1912, "step": 10509 }, { "epoch": 2.7967003725385844, "grad_norm": 0.28403499722480774, "learning_rate": 8.426725042960311e-08, "loss": 0.1988, "step": 10510 }, { "epoch": 2.796966471527408, "grad_norm": 0.2924747169017792, "learning_rate": 8.425040223473339e-08, "loss": 0.183, "step": 10511 }, { "epoch": 2.797232570516232, "grad_norm": 0.3023376762866974, "learning_rate": 8.42335544982935e-08, "loss": 0.2229, "step": 10512 }, { "epoch": 2.797498669505056, "grad_norm": 0.27246227860450745, "learning_rate": 8.421670722077392e-08, "loss": 0.1793, "step": 10513 }, { "epoch": 2.7977647684938796, "grad_norm": 0.2983749806880951, "learning_rate": 8.4199860402665e-08, "loss": 0.1972, "step": 10514 }, { "epoch": 2.7980308674827037, "grad_norm": 0.28092846274375916, "learning_rate": 8.418301404445715e-08, "loss": 0.1931, "step": 10515 }, { "epoch": 2.7982969664715274, "grad_norm": 0.3257034718990326, "learning_rate": 8.416616814664067e-08, "loss": 0.2003, "step": 10516 }, { "epoch": 2.7985630654603515, "grad_norm": 0.2888447940349579, "learning_rate": 8.414932270970592e-08, "loss": 0.1792, "step": 10517 }, { "epoch": 2.798829164449175, "grad_norm": 0.2845219671726227, "learning_rate": 8.413247773414323e-08, "loss": 0.1802, "step": 10518 }, { "epoch": 2.799095263437999, "grad_norm": 0.3073056638240814, "learning_rate": 8.411563322044292e-08, "loss": 0.2019, "step": 10519 }, { "epoch": 2.799361362426823, "grad_norm": 0.3018055856227875, "learning_rate": 8.409878916909532e-08, "loss": 0.1987, "step": 10520 }, { "epoch": 2.7996274614156467, "grad_norm": 0.2931632399559021, "learning_rate": 8.408194558059063e-08, "loss": 0.1994, "step": 10521 }, { "epoch": 2.7998935604044703, "grad_norm": 0.3190029263496399, "learning_rate": 8.40651024554192e-08, "loss": 0.1726, "step": 10522 }, { "epoch": 2.8001596593932945, "grad_norm": 0.2915964424610138, "learning_rate": 8.404825979407123e-08, "loss": 0.1931, "step": 10523 }, { "epoch": 2.800425758382118, "grad_norm": 0.3013612926006317, "learning_rate": 8.403141759703703e-08, "loss": 0.2003, "step": 10524 }, { "epoch": 2.800691857370942, "grad_norm": 0.3388889729976654, "learning_rate": 8.401457586480679e-08, "loss": 0.2118, "step": 10525 }, { "epoch": 2.800957956359766, "grad_norm": 0.29631197452545166, "learning_rate": 8.399773459787074e-08, "loss": 0.1964, "step": 10526 }, { "epoch": 2.8012240553485896, "grad_norm": 0.31718432903289795, "learning_rate": 8.398089379671908e-08, "loss": 0.1837, "step": 10527 }, { "epoch": 2.8014901543374133, "grad_norm": 0.2983884811401367, "learning_rate": 8.3964053461842e-08, "loss": 0.1829, "step": 10528 }, { "epoch": 2.8017562533262375, "grad_norm": 0.3261459171772003, "learning_rate": 8.394721359372976e-08, "loss": 0.2183, "step": 10529 }, { "epoch": 2.802022352315061, "grad_norm": 0.24780787527561188, "learning_rate": 8.393037419287238e-08, "loss": 0.164, "step": 10530 }, { "epoch": 2.802288451303885, "grad_norm": 0.28829941153526306, "learning_rate": 8.391353525976011e-08, "loss": 0.1938, "step": 10531 }, { "epoch": 2.802554550292709, "grad_norm": 0.2714073061943054, "learning_rate": 8.389669679488303e-08, "loss": 0.1763, "step": 10532 }, { "epoch": 2.8028206492815326, "grad_norm": 0.28055593371391296, "learning_rate": 8.387985879873131e-08, "loss": 0.1823, "step": 10533 }, { "epoch": 2.8030867482703563, "grad_norm": 0.29878637194633484, "learning_rate": 8.386302127179507e-08, "loss": 0.1996, "step": 10534 }, { "epoch": 2.8033528472591804, "grad_norm": 0.2629626989364624, "learning_rate": 8.384618421456435e-08, "loss": 0.1841, "step": 10535 }, { "epoch": 2.803618946248004, "grad_norm": 0.2907545566558838, "learning_rate": 8.382934762752927e-08, "loss": 0.1774, "step": 10536 }, { "epoch": 2.8038850452368282, "grad_norm": 0.3367876708507538, "learning_rate": 8.381251151117988e-08, "loss": 0.2049, "step": 10537 }, { "epoch": 2.804151144225652, "grad_norm": 0.3231692910194397, "learning_rate": 8.37956758660063e-08, "loss": 0.1997, "step": 10538 }, { "epoch": 2.804417243214476, "grad_norm": 0.3372291922569275, "learning_rate": 8.37788406924985e-08, "loss": 0.2089, "step": 10539 }, { "epoch": 2.8046833422032997, "grad_norm": 0.28778180480003357, "learning_rate": 8.376200599114651e-08, "loss": 0.2059, "step": 10540 }, { "epoch": 2.8049494411921234, "grad_norm": 0.2955303490161896, "learning_rate": 8.374517176244037e-08, "loss": 0.1916, "step": 10541 }, { "epoch": 2.8052155401809475, "grad_norm": 0.2744690179824829, "learning_rate": 8.372833800687007e-08, "loss": 0.1734, "step": 10542 }, { "epoch": 2.8054816391697712, "grad_norm": 0.29530853033065796, "learning_rate": 8.371150472492561e-08, "loss": 0.1921, "step": 10543 }, { "epoch": 2.805747738158595, "grad_norm": 0.32203781604766846, "learning_rate": 8.369467191709697e-08, "loss": 0.2079, "step": 10544 }, { "epoch": 2.806013837147419, "grad_norm": 0.2993340790271759, "learning_rate": 8.367783958387407e-08, "loss": 0.1859, "step": 10545 }, { "epoch": 2.8062799361362427, "grad_norm": 0.30507516860961914, "learning_rate": 8.366100772574688e-08, "loss": 0.1951, "step": 10546 }, { "epoch": 2.8065460351250664, "grad_norm": 0.2894110381603241, "learning_rate": 8.364417634320537e-08, "loss": 0.1813, "step": 10547 }, { "epoch": 2.8068121341138905, "grad_norm": 0.3012150824069977, "learning_rate": 8.362734543673938e-08, "loss": 0.1865, "step": 10548 }, { "epoch": 2.807078233102714, "grad_norm": 0.2730230689048767, "learning_rate": 8.361051500683887e-08, "loss": 0.1721, "step": 10549 }, { "epoch": 2.807344332091538, "grad_norm": 0.27898097038269043, "learning_rate": 8.359368505399369e-08, "loss": 0.1909, "step": 10550 }, { "epoch": 2.807610431080362, "grad_norm": 0.30828219652175903, "learning_rate": 8.357685557869375e-08, "loss": 0.1929, "step": 10551 }, { "epoch": 2.8078765300691857, "grad_norm": 0.2957286834716797, "learning_rate": 8.35600265814289e-08, "loss": 0.2065, "step": 10552 }, { "epoch": 2.8081426290580094, "grad_norm": 0.27566951513290405, "learning_rate": 8.354319806268899e-08, "loss": 0.1809, "step": 10553 }, { "epoch": 2.8084087280468335, "grad_norm": 0.27973470091819763, "learning_rate": 8.352637002296388e-08, "loss": 0.1848, "step": 10554 }, { "epoch": 2.808674827035657, "grad_norm": 0.2940913140773773, "learning_rate": 8.350954246274333e-08, "loss": 0.176, "step": 10555 }, { "epoch": 2.808940926024481, "grad_norm": 0.28776097297668457, "learning_rate": 8.349271538251722e-08, "loss": 0.1876, "step": 10556 }, { "epoch": 2.809207025013305, "grad_norm": 0.299415647983551, "learning_rate": 8.34758887827753e-08, "loss": 0.2022, "step": 10557 }, { "epoch": 2.8094731240021287, "grad_norm": 0.2843327820301056, "learning_rate": 8.345906266400733e-08, "loss": 0.2005, "step": 10558 }, { "epoch": 2.809739222990953, "grad_norm": 0.3076317608356476, "learning_rate": 8.34422370267031e-08, "loss": 0.2063, "step": 10559 }, { "epoch": 2.8100053219797765, "grad_norm": 0.31493595242500305, "learning_rate": 8.342541187135235e-08, "loss": 0.1947, "step": 10560 }, { "epoch": 2.8102714209686, "grad_norm": 0.30815958976745605, "learning_rate": 8.340858719844484e-08, "loss": 0.2003, "step": 10561 }, { "epoch": 2.8105375199574243, "grad_norm": 0.27626287937164307, "learning_rate": 8.339176300847026e-08, "loss": 0.1759, "step": 10562 }, { "epoch": 2.810803618946248, "grad_norm": 0.274568647146225, "learning_rate": 8.337493930191835e-08, "loss": 0.1717, "step": 10563 }, { "epoch": 2.811069717935072, "grad_norm": 0.28794577717781067, "learning_rate": 8.335811607927881e-08, "loss": 0.1938, "step": 10564 }, { "epoch": 2.811335816923896, "grad_norm": 0.27080583572387695, "learning_rate": 8.33412933410413e-08, "loss": 0.1785, "step": 10565 }, { "epoch": 2.8116019159127195, "grad_norm": 0.30418238043785095, "learning_rate": 8.332447108769545e-08, "loss": 0.1891, "step": 10566 }, { "epoch": 2.8118680149015436, "grad_norm": 0.2976292073726654, "learning_rate": 8.330764931973099e-08, "loss": 0.1978, "step": 10567 }, { "epoch": 2.8121341138903673, "grad_norm": 0.30052968859672546, "learning_rate": 8.329082803763748e-08, "loss": 0.1912, "step": 10568 }, { "epoch": 2.812400212879191, "grad_norm": 0.3229994773864746, "learning_rate": 8.32740072419046e-08, "loss": 0.2048, "step": 10569 }, { "epoch": 2.812666311868015, "grad_norm": 0.2824496924877167, "learning_rate": 8.325718693302193e-08, "loss": 0.1884, "step": 10570 }, { "epoch": 2.8129324108568388, "grad_norm": 0.29438894987106323, "learning_rate": 8.324036711147907e-08, "loss": 0.1897, "step": 10571 }, { "epoch": 2.8131985098456624, "grad_norm": 0.3021443486213684, "learning_rate": 8.322354777776561e-08, "loss": 0.1965, "step": 10572 }, { "epoch": 2.8134646088344866, "grad_norm": 0.2892976999282837, "learning_rate": 8.320672893237115e-08, "loss": 0.1991, "step": 10573 }, { "epoch": 2.8137307078233103, "grad_norm": 0.29075804352760315, "learning_rate": 8.318991057578519e-08, "loss": 0.195, "step": 10574 }, { "epoch": 2.813996806812134, "grad_norm": 0.346346378326416, "learning_rate": 8.317309270849727e-08, "loss": 0.2238, "step": 10575 }, { "epoch": 2.814262905800958, "grad_norm": 0.2980723977088928, "learning_rate": 8.315627533099696e-08, "loss": 0.1816, "step": 10576 }, { "epoch": 2.8145290047897817, "grad_norm": 0.27334386110305786, "learning_rate": 8.313945844377371e-08, "loss": 0.182, "step": 10577 }, { "epoch": 2.8147951037786054, "grad_norm": 0.295747309923172, "learning_rate": 8.312264204731705e-08, "loss": 0.1862, "step": 10578 }, { "epoch": 2.8150612027674295, "grad_norm": 0.2753324508666992, "learning_rate": 8.310582614211647e-08, "loss": 0.1702, "step": 10579 }, { "epoch": 2.8153273017562532, "grad_norm": 0.3184739351272583, "learning_rate": 8.308901072866141e-08, "loss": 0.182, "step": 10580 }, { "epoch": 2.815593400745077, "grad_norm": 0.28981533646583557, "learning_rate": 8.307219580744134e-08, "loss": 0.1822, "step": 10581 }, { "epoch": 2.815859499733901, "grad_norm": 0.3057461678981781, "learning_rate": 8.30553813789457e-08, "loss": 0.2036, "step": 10582 }, { "epoch": 2.8161255987227247, "grad_norm": 0.3183242976665497, "learning_rate": 8.303856744366396e-08, "loss": 0.2006, "step": 10583 }, { "epoch": 2.816391697711549, "grad_norm": 0.3115222752094269, "learning_rate": 8.302175400208543e-08, "loss": 0.1902, "step": 10584 }, { "epoch": 2.8166577967003725, "grad_norm": 0.28603988885879517, "learning_rate": 8.300494105469957e-08, "loss": 0.1972, "step": 10585 }, { "epoch": 2.8169238956891967, "grad_norm": 0.28748273849487305, "learning_rate": 8.298812860199575e-08, "loss": 0.1795, "step": 10586 }, { "epoch": 2.8171899946780203, "grad_norm": 0.3002079129219055, "learning_rate": 8.297131664446333e-08, "loss": 0.2216, "step": 10587 }, { "epoch": 2.817456093666844, "grad_norm": 0.30550873279571533, "learning_rate": 8.295450518259166e-08, "loss": 0.2084, "step": 10588 }, { "epoch": 2.817722192655668, "grad_norm": 0.3067326843738556, "learning_rate": 8.29376942168701e-08, "loss": 0.192, "step": 10589 }, { "epoch": 2.817988291644492, "grad_norm": 0.27128544449806213, "learning_rate": 8.292088374778794e-08, "loss": 0.1782, "step": 10590 }, { "epoch": 2.8182543906333155, "grad_norm": 0.2997058629989624, "learning_rate": 8.290407377583452e-08, "loss": 0.2145, "step": 10591 }, { "epoch": 2.8185204896221396, "grad_norm": 0.297109991312027, "learning_rate": 8.288726430149917e-08, "loss": 0.1906, "step": 10592 }, { "epoch": 2.8187865886109633, "grad_norm": 0.30039381980895996, "learning_rate": 8.287045532527108e-08, "loss": 0.2142, "step": 10593 }, { "epoch": 2.819052687599787, "grad_norm": 0.2953723967075348, "learning_rate": 8.285364684763954e-08, "loss": 0.1913, "step": 10594 }, { "epoch": 2.819318786588611, "grad_norm": 0.3202798664569855, "learning_rate": 8.283683886909384e-08, "loss": 0.2017, "step": 10595 }, { "epoch": 2.819584885577435, "grad_norm": 0.3140588402748108, "learning_rate": 8.282003139012321e-08, "loss": 0.1997, "step": 10596 }, { "epoch": 2.8198509845662585, "grad_norm": 0.2839818596839905, "learning_rate": 8.280322441121684e-08, "loss": 0.198, "step": 10597 }, { "epoch": 2.8201170835550826, "grad_norm": 0.279539555311203, "learning_rate": 8.278641793286395e-08, "loss": 0.1937, "step": 10598 }, { "epoch": 2.8203831825439063, "grad_norm": 0.27024900913238525, "learning_rate": 8.276961195555375e-08, "loss": 0.1864, "step": 10599 }, { "epoch": 2.82064928153273, "grad_norm": 0.3122318685054779, "learning_rate": 8.275280647977541e-08, "loss": 0.1878, "step": 10600 }, { "epoch": 2.820915380521554, "grad_norm": 0.30983060598373413, "learning_rate": 8.273600150601812e-08, "loss": 0.2175, "step": 10601 }, { "epoch": 2.821181479510378, "grad_norm": 0.3032788932323456, "learning_rate": 8.271919703477095e-08, "loss": 0.2042, "step": 10602 }, { "epoch": 2.8214475784992015, "grad_norm": 0.28703054785728455, "learning_rate": 8.270239306652311e-08, "loss": 0.1746, "step": 10603 }, { "epoch": 2.8217136774880256, "grad_norm": 0.3140074610710144, "learning_rate": 8.268558960176369e-08, "loss": 0.2263, "step": 10604 }, { "epoch": 2.8219797764768493, "grad_norm": 0.2651137113571167, "learning_rate": 8.26687866409818e-08, "loss": 0.1938, "step": 10605 }, { "epoch": 2.822245875465673, "grad_norm": 0.2817825973033905, "learning_rate": 8.265198418466655e-08, "loss": 0.1795, "step": 10606 }, { "epoch": 2.822511974454497, "grad_norm": 0.3037800192832947, "learning_rate": 8.263518223330696e-08, "loss": 0.1885, "step": 10607 }, { "epoch": 2.8227780734433208, "grad_norm": 0.2960506081581116, "learning_rate": 8.261838078739215e-08, "loss": 0.1905, "step": 10608 }, { "epoch": 2.823044172432145, "grad_norm": 0.29057085514068604, "learning_rate": 8.260157984741113e-08, "loss": 0.2026, "step": 10609 }, { "epoch": 2.8233102714209686, "grad_norm": 0.31727927923202515, "learning_rate": 8.2584779413853e-08, "loss": 0.2166, "step": 10610 }, { "epoch": 2.8235763704097927, "grad_norm": 0.2880527079105377, "learning_rate": 8.256797948720667e-08, "loss": 0.1916, "step": 10611 }, { "epoch": 2.8238424693986164, "grad_norm": 0.314238965511322, "learning_rate": 8.25511800679612e-08, "loss": 0.2001, "step": 10612 }, { "epoch": 2.82410856838744, "grad_norm": 0.29843103885650635, "learning_rate": 8.253438115660557e-08, "loss": 0.2001, "step": 10613 }, { "epoch": 2.824374667376264, "grad_norm": 0.27986472845077515, "learning_rate": 8.251758275362876e-08, "loss": 0.1833, "step": 10614 }, { "epoch": 2.824640766365088, "grad_norm": 0.29694652557373047, "learning_rate": 8.250078485951975e-08, "loss": 0.1939, "step": 10615 }, { "epoch": 2.8249068653539116, "grad_norm": 0.2783285677433014, "learning_rate": 8.248398747476742e-08, "loss": 0.1927, "step": 10616 }, { "epoch": 2.8251729643427357, "grad_norm": 0.2773381471633911, "learning_rate": 8.246719059986075e-08, "loss": 0.184, "step": 10617 }, { "epoch": 2.8254390633315594, "grad_norm": 0.2861785590648651, "learning_rate": 8.245039423528863e-08, "loss": 0.189, "step": 10618 }, { "epoch": 2.825705162320383, "grad_norm": 0.31036442518234253, "learning_rate": 8.243359838154e-08, "loss": 0.2257, "step": 10619 }, { "epoch": 2.825971261309207, "grad_norm": 0.2724335491657257, "learning_rate": 8.241680303910369e-08, "loss": 0.1822, "step": 10620 }, { "epoch": 2.826237360298031, "grad_norm": 0.2831041216850281, "learning_rate": 8.240000820846857e-08, "loss": 0.1987, "step": 10621 }, { "epoch": 2.8265034592868545, "grad_norm": 0.306400865316391, "learning_rate": 8.238321389012352e-08, "loss": 0.2018, "step": 10622 }, { "epoch": 2.8267695582756787, "grad_norm": 0.3125542104244232, "learning_rate": 8.236642008455738e-08, "loss": 0.1934, "step": 10623 }, { "epoch": 2.8270356572645023, "grad_norm": 0.28816208243370056, "learning_rate": 8.234962679225895e-08, "loss": 0.1916, "step": 10624 }, { "epoch": 2.827301756253326, "grad_norm": 0.2815791666507721, "learning_rate": 8.233283401371708e-08, "loss": 0.1781, "step": 10625 }, { "epoch": 2.82756785524215, "grad_norm": 0.2914201319217682, "learning_rate": 8.231604174942052e-08, "loss": 0.1823, "step": 10626 }, { "epoch": 2.827833954230974, "grad_norm": 0.2708866596221924, "learning_rate": 8.229924999985809e-08, "loss": 0.1748, "step": 10627 }, { "epoch": 2.8281000532197975, "grad_norm": 0.289815217256546, "learning_rate": 8.228245876551856e-08, "loss": 0.191, "step": 10628 }, { "epoch": 2.8283661522086216, "grad_norm": 0.30338770151138306, "learning_rate": 8.22656680468906e-08, "loss": 0.1877, "step": 10629 }, { "epoch": 2.8286322511974453, "grad_norm": 0.27706000208854675, "learning_rate": 8.224887784446301e-08, "loss": 0.1786, "step": 10630 }, { "epoch": 2.8288983501862695, "grad_norm": 0.273551881313324, "learning_rate": 8.223208815872449e-08, "loss": 0.1934, "step": 10631 }, { "epoch": 2.829164449175093, "grad_norm": 0.2936823070049286, "learning_rate": 8.221529899016375e-08, "loss": 0.209, "step": 10632 }, { "epoch": 2.829430548163917, "grad_norm": 0.2700638771057129, "learning_rate": 8.219851033926948e-08, "loss": 0.203, "step": 10633 }, { "epoch": 2.829696647152741, "grad_norm": 0.29944753646850586, "learning_rate": 8.218172220653035e-08, "loss": 0.1927, "step": 10634 }, { "epoch": 2.8299627461415646, "grad_norm": 0.3076460063457489, "learning_rate": 8.216493459243505e-08, "loss": 0.1813, "step": 10635 }, { "epoch": 2.8302288451303887, "grad_norm": 0.3006119132041931, "learning_rate": 8.214814749747217e-08, "loss": 0.1884, "step": 10636 }, { "epoch": 2.8304949441192124, "grad_norm": 0.3071960210800171, "learning_rate": 8.213136092213038e-08, "loss": 0.1885, "step": 10637 }, { "epoch": 2.830761043108036, "grad_norm": 0.29242998361587524, "learning_rate": 8.211457486689827e-08, "loss": 0.194, "step": 10638 }, { "epoch": 2.8310271420968602, "grad_norm": 0.3194913864135742, "learning_rate": 8.209778933226445e-08, "loss": 0.2015, "step": 10639 }, { "epoch": 2.831293241085684, "grad_norm": 0.2987905740737915, "learning_rate": 8.208100431871749e-08, "loss": 0.2033, "step": 10640 }, { "epoch": 2.8315593400745076, "grad_norm": 0.297281950712204, "learning_rate": 8.206421982674597e-08, "loss": 0.1967, "step": 10641 }, { "epoch": 2.8318254390633317, "grad_norm": 0.289236456155777, "learning_rate": 8.204743585683842e-08, "loss": 0.1947, "step": 10642 }, { "epoch": 2.8320915380521554, "grad_norm": 0.332841694355011, "learning_rate": 8.203065240948342e-08, "loss": 0.2175, "step": 10643 }, { "epoch": 2.832357637040979, "grad_norm": 0.30967485904693604, "learning_rate": 8.201386948516946e-08, "loss": 0.1931, "step": 10644 }, { "epoch": 2.832623736029803, "grad_norm": 0.2964351773262024, "learning_rate": 8.19970870843851e-08, "loss": 0.1733, "step": 10645 }, { "epoch": 2.832889835018627, "grad_norm": 0.2962632179260254, "learning_rate": 8.198030520761878e-08, "loss": 0.1842, "step": 10646 }, { "epoch": 2.8331559340074506, "grad_norm": 0.2910691499710083, "learning_rate": 8.196352385535896e-08, "loss": 0.2021, "step": 10647 }, { "epoch": 2.8334220329962747, "grad_norm": 0.2922699451446533, "learning_rate": 8.194674302809416e-08, "loss": 0.2105, "step": 10648 }, { "epoch": 2.8336881319850984, "grad_norm": 0.2683624029159546, "learning_rate": 8.192996272631277e-08, "loss": 0.1759, "step": 10649 }, { "epoch": 2.833954230973922, "grad_norm": 0.2724379599094391, "learning_rate": 8.191318295050325e-08, "loss": 0.1882, "step": 10650 }, { "epoch": 2.834220329962746, "grad_norm": 0.3357524275779724, "learning_rate": 8.189640370115401e-08, "loss": 0.2121, "step": 10651 }, { "epoch": 2.83448642895157, "grad_norm": 0.312747061252594, "learning_rate": 8.187962497875344e-08, "loss": 0.2084, "step": 10652 }, { "epoch": 2.8347525279403936, "grad_norm": 0.3175438344478607, "learning_rate": 8.186284678378995e-08, "loss": 0.1961, "step": 10653 }, { "epoch": 2.8350186269292177, "grad_norm": 0.28457382321357727, "learning_rate": 8.184606911675189e-08, "loss": 0.1988, "step": 10654 }, { "epoch": 2.8352847259180414, "grad_norm": 0.2766687273979187, "learning_rate": 8.182929197812768e-08, "loss": 0.1807, "step": 10655 }, { "epoch": 2.8355508249068655, "grad_norm": 0.2894992232322693, "learning_rate": 8.181251536840554e-08, "loss": 0.1933, "step": 10656 }, { "epoch": 2.835816923895689, "grad_norm": 0.3043023943901062, "learning_rate": 8.179573928807387e-08, "loss": 0.1891, "step": 10657 }, { "epoch": 2.8360830228845133, "grad_norm": 0.30184206366539, "learning_rate": 8.177896373762098e-08, "loss": 0.1963, "step": 10658 }, { "epoch": 2.836349121873337, "grad_norm": 0.34041014313697815, "learning_rate": 8.176218871753512e-08, "loss": 0.2076, "step": 10659 }, { "epoch": 2.8366152208621607, "grad_norm": 0.27689701318740845, "learning_rate": 8.174541422830459e-08, "loss": 0.1846, "step": 10660 }, { "epoch": 2.836881319850985, "grad_norm": 0.3011351525783539, "learning_rate": 8.172864027041766e-08, "loss": 0.1924, "step": 10661 }, { "epoch": 2.8371474188398085, "grad_norm": 0.3097658157348633, "learning_rate": 8.171186684436257e-08, "loss": 0.2075, "step": 10662 }, { "epoch": 2.837413517828632, "grad_norm": 0.27428051829338074, "learning_rate": 8.169509395062755e-08, "loss": 0.2073, "step": 10663 }, { "epoch": 2.8376796168174563, "grad_norm": 0.2777220904827118, "learning_rate": 8.167832158970086e-08, "loss": 0.1841, "step": 10664 }, { "epoch": 2.83794571580628, "grad_norm": 0.85504549741745, "learning_rate": 8.166154976207062e-08, "loss": 0.1706, "step": 10665 }, { "epoch": 2.8382118147951036, "grad_norm": 0.28543785214424133, "learning_rate": 8.164477846822505e-08, "loss": 0.1981, "step": 10666 }, { "epoch": 2.8384779137839278, "grad_norm": 0.3004997968673706, "learning_rate": 8.162800770865232e-08, "loss": 0.1921, "step": 10667 }, { "epoch": 2.8387440127727515, "grad_norm": 0.3076854944229126, "learning_rate": 8.161123748384057e-08, "loss": 0.2033, "step": 10668 }, { "epoch": 2.839010111761575, "grad_norm": 0.30499744415283203, "learning_rate": 8.159446779427797e-08, "loss": 0.225, "step": 10669 }, { "epoch": 2.8392762107503993, "grad_norm": 0.29740989208221436, "learning_rate": 8.15776986404526e-08, "loss": 0.2017, "step": 10670 }, { "epoch": 2.839542309739223, "grad_norm": 0.32076048851013184, "learning_rate": 8.156093002285261e-08, "loss": 0.1933, "step": 10671 }, { "epoch": 2.8398084087280466, "grad_norm": 0.2958771586418152, "learning_rate": 8.154416194196608e-08, "loss": 0.1909, "step": 10672 }, { "epoch": 2.8400745077168708, "grad_norm": 0.31927624344825745, "learning_rate": 8.152739439828108e-08, "loss": 0.1818, "step": 10673 }, { "epoch": 2.8403406067056944, "grad_norm": 0.2711932063102722, "learning_rate": 8.151062739228565e-08, "loss": 0.1759, "step": 10674 }, { "epoch": 2.840606705694518, "grad_norm": 0.27148574590682983, "learning_rate": 8.149386092446784e-08, "loss": 0.1823, "step": 10675 }, { "epoch": 2.8408728046833422, "grad_norm": 0.3004918396472931, "learning_rate": 8.147709499531569e-08, "loss": 0.2206, "step": 10676 }, { "epoch": 2.841138903672166, "grad_norm": 0.3196414113044739, "learning_rate": 8.146032960531724e-08, "loss": 0.2024, "step": 10677 }, { "epoch": 2.84140500266099, "grad_norm": 0.3014391362667084, "learning_rate": 8.144356475496043e-08, "loss": 0.2031, "step": 10678 }, { "epoch": 2.8416711016498137, "grad_norm": 0.2704947888851166, "learning_rate": 8.142680044473325e-08, "loss": 0.1813, "step": 10679 }, { "epoch": 2.8419372006386374, "grad_norm": 0.2923547327518463, "learning_rate": 8.141003667512372e-08, "loss": 0.2004, "step": 10680 }, { "epoch": 2.8422032996274615, "grad_norm": 0.30759575963020325, "learning_rate": 8.139327344661972e-08, "loss": 0.2013, "step": 10681 }, { "epoch": 2.8424693986162852, "grad_norm": 0.2786657512187958, "learning_rate": 8.137651075970929e-08, "loss": 0.1797, "step": 10682 }, { "epoch": 2.8427354976051094, "grad_norm": 0.30180665850639343, "learning_rate": 8.135974861488022e-08, "loss": 0.1991, "step": 10683 }, { "epoch": 2.843001596593933, "grad_norm": 0.29852548241615295, "learning_rate": 8.134298701262046e-08, "loss": 0.1928, "step": 10684 }, { "epoch": 2.8432676955827567, "grad_norm": 0.29944950342178345, "learning_rate": 8.132622595341792e-08, "loss": 0.1972, "step": 10685 }, { "epoch": 2.843533794571581, "grad_norm": 0.3067605793476105, "learning_rate": 8.130946543776045e-08, "loss": 0.1987, "step": 10686 }, { "epoch": 2.8437998935604045, "grad_norm": 0.28341782093048096, "learning_rate": 8.129270546613593e-08, "loss": 0.1798, "step": 10687 }, { "epoch": 2.844065992549228, "grad_norm": 0.27177709341049194, "learning_rate": 8.127594603903216e-08, "loss": 0.1753, "step": 10688 }, { "epoch": 2.8443320915380523, "grad_norm": 0.28218722343444824, "learning_rate": 8.1259187156937e-08, "loss": 0.1845, "step": 10689 }, { "epoch": 2.844598190526876, "grad_norm": 0.3129928410053253, "learning_rate": 8.124242882033821e-08, "loss": 0.2029, "step": 10690 }, { "epoch": 2.8448642895156997, "grad_norm": 0.2910098731517792, "learning_rate": 8.122567102972368e-08, "loss": 0.1873, "step": 10691 }, { "epoch": 2.845130388504524, "grad_norm": 0.3094116449356079, "learning_rate": 8.120891378558108e-08, "loss": 0.2128, "step": 10692 }, { "epoch": 2.8453964874933475, "grad_norm": 0.2608225643634796, "learning_rate": 8.119215708839819e-08, "loss": 0.1804, "step": 10693 }, { "epoch": 2.845662586482171, "grad_norm": 0.2981460392475128, "learning_rate": 8.11754009386628e-08, "loss": 0.1927, "step": 10694 }, { "epoch": 2.8459286854709953, "grad_norm": 0.29553183913230896, "learning_rate": 8.115864533686258e-08, "loss": 0.1794, "step": 10695 }, { "epoch": 2.846194784459819, "grad_norm": 0.2940976619720459, "learning_rate": 8.114189028348529e-08, "loss": 0.1951, "step": 10696 }, { "epoch": 2.8464608834486427, "grad_norm": 0.3002425730228424, "learning_rate": 8.112513577901864e-08, "loss": 0.1915, "step": 10697 }, { "epoch": 2.846726982437467, "grad_norm": 0.2878725528717041, "learning_rate": 8.110838182395025e-08, "loss": 0.1796, "step": 10698 }, { "epoch": 2.8469930814262905, "grad_norm": 0.29818636178970337, "learning_rate": 8.10916284187678e-08, "loss": 0.1918, "step": 10699 }, { "epoch": 2.847259180415114, "grad_norm": 0.2944682836532593, "learning_rate": 8.107487556395901e-08, "loss": 0.1981, "step": 10700 }, { "epoch": 2.8475252794039383, "grad_norm": 0.27750641107559204, "learning_rate": 8.105812326001141e-08, "loss": 0.1795, "step": 10701 }, { "epoch": 2.847791378392762, "grad_norm": 0.2979058027267456, "learning_rate": 8.104137150741265e-08, "loss": 0.1945, "step": 10702 }, { "epoch": 2.848057477381586, "grad_norm": 0.2989223003387451, "learning_rate": 8.102462030665034e-08, "loss": 0.1947, "step": 10703 }, { "epoch": 2.84832357637041, "grad_norm": 0.29812687635421753, "learning_rate": 8.100786965821207e-08, "loss": 0.1915, "step": 10704 }, { "epoch": 2.848589675359234, "grad_norm": 0.29236769676208496, "learning_rate": 8.099111956258541e-08, "loss": 0.1773, "step": 10705 }, { "epoch": 2.8488557743480576, "grad_norm": 0.29633939266204834, "learning_rate": 8.097437002025789e-08, "loss": 0.2194, "step": 10706 }, { "epoch": 2.8491218733368813, "grad_norm": 0.3066290616989136, "learning_rate": 8.095762103171709e-08, "loss": 0.2014, "step": 10707 }, { "epoch": 2.8493879723257054, "grad_norm": 0.30064722895622253, "learning_rate": 8.094087259745048e-08, "loss": 0.1887, "step": 10708 }, { "epoch": 2.849654071314529, "grad_norm": 0.2890889644622803, "learning_rate": 8.092412471794558e-08, "loss": 0.1992, "step": 10709 }, { "epoch": 2.8499201703033528, "grad_norm": 0.3251712918281555, "learning_rate": 8.09073773936899e-08, "loss": 0.197, "step": 10710 }, { "epoch": 2.850186269292177, "grad_norm": 0.2867572605609894, "learning_rate": 8.089063062517087e-08, "loss": 0.1686, "step": 10711 }, { "epoch": 2.8504523682810006, "grad_norm": 0.3020190894603729, "learning_rate": 8.087388441287594e-08, "loss": 0.1902, "step": 10712 }, { "epoch": 2.8507184672698243, "grad_norm": 0.296502947807312, "learning_rate": 8.08571387572926e-08, "loss": 0.1999, "step": 10713 }, { "epoch": 2.8509845662586484, "grad_norm": 0.3030985891819, "learning_rate": 8.084039365890825e-08, "loss": 0.2095, "step": 10714 }, { "epoch": 2.851250665247472, "grad_norm": 0.2879403233528137, "learning_rate": 8.08236491182103e-08, "loss": 0.186, "step": 10715 }, { "epoch": 2.8515167642362957, "grad_norm": 0.2904673218727112, "learning_rate": 8.080690513568612e-08, "loss": 0.1791, "step": 10716 }, { "epoch": 2.85178286322512, "grad_norm": 0.32031309604644775, "learning_rate": 8.07901617118231e-08, "loss": 0.1957, "step": 10717 }, { "epoch": 2.8520489622139436, "grad_norm": 0.31487229466438293, "learning_rate": 8.077341884710861e-08, "loss": 0.2095, "step": 10718 }, { "epoch": 2.8523150612027672, "grad_norm": 0.28436070680618286, "learning_rate": 8.075667654202998e-08, "loss": 0.1775, "step": 10719 }, { "epoch": 2.8525811601915914, "grad_norm": 0.3007301688194275, "learning_rate": 8.07399347970745e-08, "loss": 0.201, "step": 10720 }, { "epoch": 2.852847259180415, "grad_norm": 0.30022159218788147, "learning_rate": 8.072319361272953e-08, "loss": 0.1883, "step": 10721 }, { "epoch": 2.8531133581692387, "grad_norm": 0.29582756757736206, "learning_rate": 8.070645298948232e-08, "loss": 0.2004, "step": 10722 }, { "epoch": 2.853379457158063, "grad_norm": 0.30560266971588135, "learning_rate": 8.068971292782017e-08, "loss": 0.2202, "step": 10723 }, { "epoch": 2.8536455561468865, "grad_norm": 0.29662179946899414, "learning_rate": 8.067297342823035e-08, "loss": 0.1965, "step": 10724 }, { "epoch": 2.85391165513571, "grad_norm": 0.2974649965763092, "learning_rate": 8.065623449120009e-08, "loss": 0.1931, "step": 10725 }, { "epoch": 2.8541777541245343, "grad_norm": 0.4139796197414398, "learning_rate": 8.063949611721663e-08, "loss": 0.1892, "step": 10726 }, { "epoch": 2.854443853113358, "grad_norm": 0.279682457447052, "learning_rate": 8.062275830676719e-08, "loss": 0.1845, "step": 10727 }, { "epoch": 2.854709952102182, "grad_norm": 0.3020443022251129, "learning_rate": 8.060602106033889e-08, "loss": 0.1895, "step": 10728 }, { "epoch": 2.854976051091006, "grad_norm": 0.33092695474624634, "learning_rate": 8.0589284378419e-08, "loss": 0.1976, "step": 10729 }, { "epoch": 2.85524215007983, "grad_norm": 0.3161627948284149, "learning_rate": 8.057254826149463e-08, "loss": 0.2002, "step": 10730 }, { "epoch": 2.8555082490686536, "grad_norm": 0.2958204448223114, "learning_rate": 8.055581271005291e-08, "loss": 0.1999, "step": 10731 }, { "epoch": 2.8557743480574773, "grad_norm": 0.33025288581848145, "learning_rate": 8.053907772458103e-08, "loss": 0.2233, "step": 10732 }, { "epoch": 2.8560404470463014, "grad_norm": 0.28145086765289307, "learning_rate": 8.052234330556605e-08, "loss": 0.1803, "step": 10733 }, { "epoch": 2.856306546035125, "grad_norm": 0.29037022590637207, "learning_rate": 8.050560945349508e-08, "loss": 0.1901, "step": 10734 }, { "epoch": 2.856572645023949, "grad_norm": 0.2854960858821869, "learning_rate": 8.04888761688552e-08, "loss": 0.1795, "step": 10735 }, { "epoch": 2.856838744012773, "grad_norm": 0.3135436773300171, "learning_rate": 8.047214345213352e-08, "loss": 0.1879, "step": 10736 }, { "epoch": 2.8571048430015966, "grad_norm": 0.27421504259109497, "learning_rate": 8.045541130381702e-08, "loss": 0.1779, "step": 10737 }, { "epoch": 2.8573709419904203, "grad_norm": 0.2902313470840454, "learning_rate": 8.043867972439272e-08, "loss": 0.1821, "step": 10738 }, { "epoch": 2.8576370409792444, "grad_norm": 0.2659573256969452, "learning_rate": 8.04219487143477e-08, "loss": 0.1793, "step": 10739 }, { "epoch": 2.857903139968068, "grad_norm": 0.31779441237449646, "learning_rate": 8.040521827416889e-08, "loss": 0.198, "step": 10740 }, { "epoch": 2.858169238956892, "grad_norm": 0.2852526009082794, "learning_rate": 8.03884884043433e-08, "loss": 0.1725, "step": 10741 }, { "epoch": 2.858435337945716, "grad_norm": 0.28257113695144653, "learning_rate": 8.03717591053579e-08, "loss": 0.1843, "step": 10742 }, { "epoch": 2.8587014369345396, "grad_norm": 0.29018688201904297, "learning_rate": 8.035503037769961e-08, "loss": 0.1858, "step": 10743 }, { "epoch": 2.8589675359233633, "grad_norm": 0.28919509053230286, "learning_rate": 8.03383022218554e-08, "loss": 0.2088, "step": 10744 }, { "epoch": 2.8592336349121874, "grad_norm": 0.30675119161605835, "learning_rate": 8.032157463831216e-08, "loss": 0.211, "step": 10745 }, { "epoch": 2.859499733901011, "grad_norm": 0.3413163125514984, "learning_rate": 8.03048476275568e-08, "loss": 0.1996, "step": 10746 }, { "epoch": 2.8597658328898348, "grad_norm": 0.2804313898086548, "learning_rate": 8.028812119007618e-08, "loss": 0.1825, "step": 10747 }, { "epoch": 2.860031931878659, "grad_norm": 0.2725500166416168, "learning_rate": 8.027139532635714e-08, "loss": 0.1777, "step": 10748 }, { "epoch": 2.8602980308674826, "grad_norm": 0.2827484607696533, "learning_rate": 8.025467003688661e-08, "loss": 0.1838, "step": 10749 }, { "epoch": 2.8605641298563067, "grad_norm": 0.2723684012889862, "learning_rate": 8.023794532215133e-08, "loss": 0.1814, "step": 10750 }, { "epoch": 2.8608302288451304, "grad_norm": 0.2999410033226013, "learning_rate": 8.022122118263815e-08, "loss": 0.2101, "step": 10751 }, { "epoch": 2.861096327833954, "grad_norm": 0.3159683048725128, "learning_rate": 8.020449761883386e-08, "loss": 0.2056, "step": 10752 }, { "epoch": 2.861362426822778, "grad_norm": 0.3030964732170105, "learning_rate": 8.018777463122526e-08, "loss": 0.2024, "step": 10753 }, { "epoch": 2.861628525811602, "grad_norm": 0.34276896715164185, "learning_rate": 8.01710522202991e-08, "loss": 0.208, "step": 10754 }, { "epoch": 2.861894624800426, "grad_norm": 0.3076513111591339, "learning_rate": 8.015433038654216e-08, "loss": 0.1917, "step": 10755 }, { "epoch": 2.8621607237892497, "grad_norm": 0.2825692892074585, "learning_rate": 8.013760913044109e-08, "loss": 0.1707, "step": 10756 }, { "epoch": 2.8624268227780734, "grad_norm": 0.3133767247200012, "learning_rate": 8.012088845248264e-08, "loss": 0.1754, "step": 10757 }, { "epoch": 2.8626929217668975, "grad_norm": 0.316623330116272, "learning_rate": 8.010416835315353e-08, "loss": 0.1928, "step": 10758 }, { "epoch": 2.862959020755721, "grad_norm": 0.30540770292282104, "learning_rate": 8.008744883294042e-08, "loss": 0.1964, "step": 10759 }, { "epoch": 2.863225119744545, "grad_norm": 0.29119792580604553, "learning_rate": 8.007072989232995e-08, "loss": 0.1758, "step": 10760 }, { "epoch": 2.863491218733369, "grad_norm": 0.2925816476345062, "learning_rate": 8.005401153180881e-08, "loss": 0.1889, "step": 10761 }, { "epoch": 2.8637573177221927, "grad_norm": 0.2943527102470398, "learning_rate": 8.003729375186359e-08, "loss": 0.1977, "step": 10762 }, { "epoch": 2.8640234167110163, "grad_norm": 0.29141512513160706, "learning_rate": 8.002057655298092e-08, "loss": 0.1796, "step": 10763 }, { "epoch": 2.8642895156998405, "grad_norm": 0.3049652576446533, "learning_rate": 8.000385993564743e-08, "loss": 0.2041, "step": 10764 }, { "epoch": 2.864555614688664, "grad_norm": 0.2891100347042084, "learning_rate": 7.998714390034962e-08, "loss": 0.1984, "step": 10765 }, { "epoch": 2.864821713677488, "grad_norm": 0.29714980721473694, "learning_rate": 7.997042844757407e-08, "loss": 0.1852, "step": 10766 }, { "epoch": 2.865087812666312, "grad_norm": 0.31174859404563904, "learning_rate": 7.995371357780737e-08, "loss": 0.1921, "step": 10767 }, { "epoch": 2.8653539116551356, "grad_norm": 0.30607300996780396, "learning_rate": 7.993699929153603e-08, "loss": 0.2059, "step": 10768 }, { "epoch": 2.8656200106439593, "grad_norm": 0.30601802468299866, "learning_rate": 7.992028558924652e-08, "loss": 0.2077, "step": 10769 }, { "epoch": 2.8658861096327835, "grad_norm": 0.2928089499473572, "learning_rate": 7.990357247142537e-08, "loss": 0.1919, "step": 10770 }, { "epoch": 2.866152208621607, "grad_norm": 0.2861524224281311, "learning_rate": 7.988685993855902e-08, "loss": 0.1809, "step": 10771 }, { "epoch": 2.866418307610431, "grad_norm": 0.31850531697273254, "learning_rate": 7.987014799113397e-08, "loss": 0.2021, "step": 10772 }, { "epoch": 2.866684406599255, "grad_norm": 0.29910239577293396, "learning_rate": 7.98534366296367e-08, "loss": 0.1956, "step": 10773 }, { "epoch": 2.8669505055880786, "grad_norm": 0.3211399018764496, "learning_rate": 7.983672585455352e-08, "loss": 0.2149, "step": 10774 }, { "epoch": 2.8672166045769027, "grad_norm": 0.2999607026576996, "learning_rate": 7.982001566637092e-08, "loss": 0.2041, "step": 10775 }, { "epoch": 2.8674827035657264, "grad_norm": 0.3077132999897003, "learning_rate": 7.980330606557524e-08, "loss": 0.1965, "step": 10776 }, { "epoch": 2.8677488025545506, "grad_norm": 0.2859199345111847, "learning_rate": 7.97865970526529e-08, "loss": 0.1872, "step": 10777 }, { "epoch": 2.8680149015433742, "grad_norm": 0.2796546220779419, "learning_rate": 7.976988862809024e-08, "loss": 0.1877, "step": 10778 }, { "epoch": 2.868281000532198, "grad_norm": 0.2810932397842407, "learning_rate": 7.975318079237358e-08, "loss": 0.1804, "step": 10779 }, { "epoch": 2.868547099521022, "grad_norm": 0.2919403612613678, "learning_rate": 7.973647354598926e-08, "loss": 0.2028, "step": 10780 }, { "epoch": 2.8688131985098457, "grad_norm": 0.31083908677101135, "learning_rate": 7.971976688942358e-08, "loss": 0.1837, "step": 10781 }, { "epoch": 2.8690792974986694, "grad_norm": 0.2869570553302765, "learning_rate": 7.970306082316287e-08, "loss": 0.1998, "step": 10782 }, { "epoch": 2.8693453964874935, "grad_norm": 0.29873570799827576, "learning_rate": 7.968635534769332e-08, "loss": 0.1997, "step": 10783 }, { "epoch": 2.869611495476317, "grad_norm": 0.2977922558784485, "learning_rate": 7.966965046350123e-08, "loss": 0.189, "step": 10784 }, { "epoch": 2.869877594465141, "grad_norm": 0.2722071707248688, "learning_rate": 7.96529461710728e-08, "loss": 0.1671, "step": 10785 }, { "epoch": 2.870143693453965, "grad_norm": 0.33762139081954956, "learning_rate": 7.96362424708943e-08, "loss": 0.2107, "step": 10786 }, { "epoch": 2.8704097924427887, "grad_norm": 0.28852030634880066, "learning_rate": 7.96195393634519e-08, "loss": 0.1873, "step": 10787 }, { "epoch": 2.8706758914316124, "grad_norm": 0.31075620651245117, "learning_rate": 7.960283684923182e-08, "loss": 0.1919, "step": 10788 }, { "epoch": 2.8709419904204365, "grad_norm": 0.2940230667591095, "learning_rate": 7.958613492872017e-08, "loss": 0.2, "step": 10789 }, { "epoch": 2.87120808940926, "grad_norm": 0.2781289219856262, "learning_rate": 7.956943360240313e-08, "loss": 0.1817, "step": 10790 }, { "epoch": 2.871474188398084, "grad_norm": 0.3116047978401184, "learning_rate": 7.955273287076688e-08, "loss": 0.1932, "step": 10791 }, { "epoch": 2.871740287386908, "grad_norm": 0.3067423105239868, "learning_rate": 7.953603273429743e-08, "loss": 0.2015, "step": 10792 }, { "epoch": 2.8720063863757317, "grad_norm": 0.2860473394393921, "learning_rate": 7.951933319348094e-08, "loss": 0.1883, "step": 10793 }, { "epoch": 2.8722724853645554, "grad_norm": 0.31181055307388306, "learning_rate": 7.950263424880348e-08, "loss": 0.2114, "step": 10794 }, { "epoch": 2.8725385843533795, "grad_norm": 0.3025546371936798, "learning_rate": 7.948593590075112e-08, "loss": 0.1971, "step": 10795 }, { "epoch": 2.872804683342203, "grad_norm": 0.2734859883785248, "learning_rate": 7.94692381498099e-08, "loss": 0.1845, "step": 10796 }, { "epoch": 2.8730707823310273, "grad_norm": 0.3035780191421509, "learning_rate": 7.945254099646585e-08, "loss": 0.218, "step": 10797 }, { "epoch": 2.873336881319851, "grad_norm": 0.27705296874046326, "learning_rate": 7.943584444120501e-08, "loss": 0.1918, "step": 10798 }, { "epoch": 2.8736029803086747, "grad_norm": 0.30668842792510986, "learning_rate": 7.941914848451331e-08, "loss": 0.2133, "step": 10799 }, { "epoch": 2.873869079297499, "grad_norm": 0.284564733505249, "learning_rate": 7.940245312687679e-08, "loss": 0.1823, "step": 10800 }, { "epoch": 2.8741351782863225, "grad_norm": 0.2674359381198883, "learning_rate": 7.938575836878138e-08, "loss": 0.1724, "step": 10801 }, { "epoch": 2.8744012772751466, "grad_norm": 0.2943330705165863, "learning_rate": 7.936906421071299e-08, "loss": 0.2067, "step": 10802 }, { "epoch": 2.8746673762639703, "grad_norm": 0.31097331643104553, "learning_rate": 7.935237065315757e-08, "loss": 0.1974, "step": 10803 }, { "epoch": 2.874933475252794, "grad_norm": 0.2967355251312256, "learning_rate": 7.933567769660103e-08, "loss": 0.1863, "step": 10804 }, { "epoch": 2.875199574241618, "grad_norm": 0.27347519993782043, "learning_rate": 7.931898534152928e-08, "loss": 0.1886, "step": 10805 }, { "epoch": 2.8754656732304418, "grad_norm": 0.28583210706710815, "learning_rate": 7.930229358842815e-08, "loss": 0.1876, "step": 10806 }, { "epoch": 2.8757317722192655, "grad_norm": 0.27383917570114136, "learning_rate": 7.928560243778352e-08, "loss": 0.1775, "step": 10807 }, { "epoch": 2.8759978712080896, "grad_norm": 0.2663789391517639, "learning_rate": 7.926891189008123e-08, "loss": 0.177, "step": 10808 }, { "epoch": 2.8762639701969133, "grad_norm": 0.2856944501399994, "learning_rate": 7.925222194580708e-08, "loss": 0.185, "step": 10809 }, { "epoch": 2.876530069185737, "grad_norm": 0.28127413988113403, "learning_rate": 7.92355326054469e-08, "loss": 0.1771, "step": 10810 }, { "epoch": 2.876796168174561, "grad_norm": 0.3067513108253479, "learning_rate": 7.921884386948642e-08, "loss": 0.1841, "step": 10811 }, { "epoch": 2.8770622671633848, "grad_norm": 0.2952110171318054, "learning_rate": 7.920215573841143e-08, "loss": 0.2003, "step": 10812 }, { "epoch": 2.8773283661522084, "grad_norm": 0.28973472118377686, "learning_rate": 7.91854682127077e-08, "loss": 0.2092, "step": 10813 }, { "epoch": 2.8775944651410326, "grad_norm": 0.28911787271499634, "learning_rate": 7.916878129286093e-08, "loss": 0.19, "step": 10814 }, { "epoch": 2.8778605641298562, "grad_norm": 0.2945402264595032, "learning_rate": 7.915209497935686e-08, "loss": 0.1929, "step": 10815 }, { "epoch": 2.87812666311868, "grad_norm": 0.31218215823173523, "learning_rate": 7.913540927268118e-08, "loss": 0.2006, "step": 10816 }, { "epoch": 2.878392762107504, "grad_norm": 0.2990468740463257, "learning_rate": 7.911872417331958e-08, "loss": 0.1973, "step": 10817 }, { "epoch": 2.8786588610963277, "grad_norm": 0.2919934391975403, "learning_rate": 7.910203968175768e-08, "loss": 0.183, "step": 10818 }, { "epoch": 2.8789249600851514, "grad_norm": 0.29182857275009155, "learning_rate": 7.908535579848115e-08, "loss": 0.1882, "step": 10819 }, { "epoch": 2.8791910590739755, "grad_norm": 0.29058918356895447, "learning_rate": 7.906867252397562e-08, "loss": 0.1911, "step": 10820 }, { "epoch": 2.8794571580627992, "grad_norm": 0.31137239933013916, "learning_rate": 7.905198985872666e-08, "loss": 0.2258, "step": 10821 }, { "epoch": 2.8797232570516234, "grad_norm": 0.30391907691955566, "learning_rate": 7.903530780321988e-08, "loss": 0.204, "step": 10822 }, { "epoch": 2.879989356040447, "grad_norm": 0.2965770363807678, "learning_rate": 7.901862635794087e-08, "loss": 0.1942, "step": 10823 }, { "epoch": 2.880255455029271, "grad_norm": 0.2951788902282715, "learning_rate": 7.900194552337515e-08, "loss": 0.1976, "step": 10824 }, { "epoch": 2.880521554018095, "grad_norm": 0.3150484263896942, "learning_rate": 7.89852653000083e-08, "loss": 0.2027, "step": 10825 }, { "epoch": 2.8807876530069185, "grad_norm": 0.29864266514778137, "learning_rate": 7.89685856883258e-08, "loss": 0.1819, "step": 10826 }, { "epoch": 2.8810537519957427, "grad_norm": 0.2979556918144226, "learning_rate": 7.89519066888132e-08, "loss": 0.2162, "step": 10827 }, { "epoch": 2.8813198509845663, "grad_norm": 0.30001935362815857, "learning_rate": 7.893522830195591e-08, "loss": 0.202, "step": 10828 }, { "epoch": 2.88158594997339, "grad_norm": 0.2678157091140747, "learning_rate": 7.891855052823942e-08, "loss": 0.1665, "step": 10829 }, { "epoch": 2.881852048962214, "grad_norm": 0.28672489523887634, "learning_rate": 7.890187336814922e-08, "loss": 0.1883, "step": 10830 }, { "epoch": 2.882118147951038, "grad_norm": 0.2834787368774414, "learning_rate": 7.888519682217067e-08, "loss": 0.1798, "step": 10831 }, { "epoch": 2.8823842469398615, "grad_norm": 0.2820465564727783, "learning_rate": 7.886852089078924e-08, "loss": 0.1815, "step": 10832 }, { "epoch": 2.8826503459286856, "grad_norm": 0.2707694470882416, "learning_rate": 7.885184557449028e-08, "loss": 0.177, "step": 10833 }, { "epoch": 2.8829164449175093, "grad_norm": 0.32680603861808777, "learning_rate": 7.883517087375919e-08, "loss": 0.2189, "step": 10834 }, { "epoch": 2.883182543906333, "grad_norm": 0.3810427784919739, "learning_rate": 7.881849678908132e-08, "loss": 0.2104, "step": 10835 }, { "epoch": 2.883448642895157, "grad_norm": 0.31004101037979126, "learning_rate": 7.880182332094207e-08, "loss": 0.199, "step": 10836 }, { "epoch": 2.883714741883981, "grad_norm": 0.3067008852958679, "learning_rate": 7.878515046982664e-08, "loss": 0.2014, "step": 10837 }, { "epoch": 2.8839808408728045, "grad_norm": 0.2938804030418396, "learning_rate": 7.87684782362204e-08, "loss": 0.2101, "step": 10838 }, { "epoch": 2.8842469398616286, "grad_norm": 0.29158130288124084, "learning_rate": 7.875180662060863e-08, "loss": 0.1904, "step": 10839 }, { "epoch": 2.8845130388504523, "grad_norm": 0.28449997305870056, "learning_rate": 7.873513562347664e-08, "loss": 0.2023, "step": 10840 }, { "epoch": 2.884779137839276, "grad_norm": 0.31623387336730957, "learning_rate": 7.871846524530958e-08, "loss": 0.2082, "step": 10841 }, { "epoch": 2.8850452368281, "grad_norm": 0.2876257300376892, "learning_rate": 7.870179548659276e-08, "loss": 0.1832, "step": 10842 }, { "epoch": 2.885311335816924, "grad_norm": 0.3006688058376312, "learning_rate": 7.86851263478114e-08, "loss": 0.1846, "step": 10843 }, { "epoch": 2.8855774348057475, "grad_norm": 0.31776824593544006, "learning_rate": 7.866845782945062e-08, "loss": 0.1829, "step": 10844 }, { "epoch": 2.8858435337945716, "grad_norm": 0.29019439220428467, "learning_rate": 7.865178993199572e-08, "loss": 0.1743, "step": 10845 }, { "epoch": 2.8861096327833953, "grad_norm": 0.3143932521343231, "learning_rate": 7.863512265593173e-08, "loss": 0.2011, "step": 10846 }, { "epoch": 2.8863757317722194, "grad_norm": 0.30543702840805054, "learning_rate": 7.861845600174385e-08, "loss": 0.1881, "step": 10847 }, { "epoch": 2.886641830761043, "grad_norm": 0.2875475585460663, "learning_rate": 7.86017899699172e-08, "loss": 0.1897, "step": 10848 }, { "epoch": 2.886907929749867, "grad_norm": 0.30945080518722534, "learning_rate": 7.858512456093689e-08, "loss": 0.206, "step": 10849 }, { "epoch": 2.887174028738691, "grad_norm": 0.2805710434913635, "learning_rate": 7.856845977528802e-08, "loss": 0.2036, "step": 10850 }, { "epoch": 2.8874401277275146, "grad_norm": 0.3033599257469177, "learning_rate": 7.85517956134556e-08, "loss": 0.194, "step": 10851 }, { "epoch": 2.8877062267163387, "grad_norm": 0.3060527741909027, "learning_rate": 7.853513207592474e-08, "loss": 0.1928, "step": 10852 }, { "epoch": 2.8879723257051624, "grad_norm": 0.2814183533191681, "learning_rate": 7.851846916318044e-08, "loss": 0.1979, "step": 10853 }, { "epoch": 2.888238424693986, "grad_norm": 0.2924398183822632, "learning_rate": 7.850180687570777e-08, "loss": 0.2093, "step": 10854 }, { "epoch": 2.88850452368281, "grad_norm": 0.28731560707092285, "learning_rate": 7.848514521399165e-08, "loss": 0.2052, "step": 10855 }, { "epoch": 2.888770622671634, "grad_norm": 0.3116503953933716, "learning_rate": 7.846848417851708e-08, "loss": 0.2087, "step": 10856 }, { "epoch": 2.8890367216604576, "grad_norm": 0.2808218002319336, "learning_rate": 7.845182376976904e-08, "loss": 0.1965, "step": 10857 }, { "epoch": 2.8893028206492817, "grad_norm": 0.2714845836162567, "learning_rate": 7.843516398823245e-08, "loss": 0.181, "step": 10858 }, { "epoch": 2.8895689196381054, "grad_norm": 0.27405691146850586, "learning_rate": 7.841850483439227e-08, "loss": 0.1634, "step": 10859 }, { "epoch": 2.889835018626929, "grad_norm": 0.279964804649353, "learning_rate": 7.840184630873337e-08, "loss": 0.197, "step": 10860 }, { "epoch": 2.890101117615753, "grad_norm": 0.2915344536304474, "learning_rate": 7.838518841174062e-08, "loss": 0.1851, "step": 10861 }, { "epoch": 2.890367216604577, "grad_norm": 0.3042716085910797, "learning_rate": 7.836853114389893e-08, "loss": 0.1953, "step": 10862 }, { "epoch": 2.8906333155934005, "grad_norm": 0.284498393535614, "learning_rate": 7.835187450569317e-08, "loss": 0.1983, "step": 10863 }, { "epoch": 2.8908994145822247, "grad_norm": 0.28415316343307495, "learning_rate": 7.833521849760809e-08, "loss": 0.1855, "step": 10864 }, { "epoch": 2.8911655135710483, "grad_norm": 0.2815403640270233, "learning_rate": 7.831856312012855e-08, "loss": 0.1908, "step": 10865 }, { "epoch": 2.891431612559872, "grad_norm": 0.31127631664276123, "learning_rate": 7.830190837373934e-08, "loss": 0.2128, "step": 10866 }, { "epoch": 2.891697711548696, "grad_norm": 0.2619934678077698, "learning_rate": 7.828525425892525e-08, "loss": 0.1718, "step": 10867 }, { "epoch": 2.89196381053752, "grad_norm": 0.29153427481651306, "learning_rate": 7.826860077617101e-08, "loss": 0.1815, "step": 10868 }, { "epoch": 2.892229909526344, "grad_norm": 0.2963833510875702, "learning_rate": 7.825194792596141e-08, "loss": 0.1853, "step": 10869 }, { "epoch": 2.8924960085151676, "grad_norm": 0.29697898030281067, "learning_rate": 7.823529570878111e-08, "loss": 0.1867, "step": 10870 }, { "epoch": 2.8927621075039913, "grad_norm": 0.32619795203208923, "learning_rate": 7.821864412511484e-08, "loss": 0.1796, "step": 10871 }, { "epoch": 2.8930282064928154, "grad_norm": 0.2925720810890198, "learning_rate": 7.820199317544734e-08, "loss": 0.1965, "step": 10872 }, { "epoch": 2.893294305481639, "grad_norm": 0.32064029574394226, "learning_rate": 7.818534286026314e-08, "loss": 0.2147, "step": 10873 }, { "epoch": 2.8935604044704633, "grad_norm": 0.27734294533729553, "learning_rate": 7.8168693180047e-08, "loss": 0.1927, "step": 10874 }, { "epoch": 2.893826503459287, "grad_norm": 0.3056836724281311, "learning_rate": 7.815204413528352e-08, "loss": 0.2148, "step": 10875 }, { "epoch": 2.8940926024481106, "grad_norm": 0.2696368098258972, "learning_rate": 7.813539572645731e-08, "loss": 0.1789, "step": 10876 }, { "epoch": 2.8943587014369347, "grad_norm": 0.31805387139320374, "learning_rate": 7.811874795405295e-08, "loss": 0.1883, "step": 10877 }, { "epoch": 2.8946248004257584, "grad_norm": 0.34931671619415283, "learning_rate": 7.810210081855501e-08, "loss": 0.1891, "step": 10878 }, { "epoch": 2.894890899414582, "grad_norm": 0.29576602578163147, "learning_rate": 7.80854543204481e-08, "loss": 0.189, "step": 10879 }, { "epoch": 2.8951569984034062, "grad_norm": 0.29213079810142517, "learning_rate": 7.806880846021668e-08, "loss": 0.1896, "step": 10880 }, { "epoch": 2.89542309739223, "grad_norm": 0.3040148913860321, "learning_rate": 7.805216323834533e-08, "loss": 0.1971, "step": 10881 }, { "epoch": 2.8956891963810536, "grad_norm": 0.3886711299419403, "learning_rate": 7.803551865531852e-08, "loss": 0.18, "step": 10882 }, { "epoch": 2.8959552953698777, "grad_norm": 0.2821389436721802, "learning_rate": 7.801887471162071e-08, "loss": 0.2016, "step": 10883 }, { "epoch": 2.8962213943587014, "grad_norm": 0.2922281324863434, "learning_rate": 7.800223140773639e-08, "loss": 0.1894, "step": 10884 }, { "epoch": 2.896487493347525, "grad_norm": 0.2892645299434662, "learning_rate": 7.798558874414998e-08, "loss": 0.1929, "step": 10885 }, { "epoch": 2.896753592336349, "grad_norm": 0.3110610246658325, "learning_rate": 7.796894672134593e-08, "loss": 0.2017, "step": 10886 }, { "epoch": 2.897019691325173, "grad_norm": 0.31164711713790894, "learning_rate": 7.795230533980864e-08, "loss": 0.2003, "step": 10887 }, { "epoch": 2.8972857903139966, "grad_norm": 0.3203131854534149, "learning_rate": 7.79356646000225e-08, "loss": 0.2036, "step": 10888 }, { "epoch": 2.8975518893028207, "grad_norm": 0.30928367376327515, "learning_rate": 7.791902450247189e-08, "loss": 0.197, "step": 10889 }, { "epoch": 2.8978179882916444, "grad_norm": 0.3019319772720337, "learning_rate": 7.790238504764116e-08, "loss": 0.2122, "step": 10890 }, { "epoch": 2.898084087280468, "grad_norm": 0.3007127642631531, "learning_rate": 7.788574623601457e-08, "loss": 0.2083, "step": 10891 }, { "epoch": 2.898350186269292, "grad_norm": 0.3208197057247162, "learning_rate": 7.786910806807653e-08, "loss": 0.206, "step": 10892 }, { "epoch": 2.898616285258116, "grad_norm": 0.3018113076686859, "learning_rate": 7.785247054431127e-08, "loss": 0.1946, "step": 10893 }, { "epoch": 2.89888238424694, "grad_norm": 0.3015112578868866, "learning_rate": 7.783583366520308e-08, "loss": 0.182, "step": 10894 }, { "epoch": 2.8991484832357637, "grad_norm": 0.29522451758384705, "learning_rate": 7.781919743123624e-08, "loss": 0.204, "step": 10895 }, { "epoch": 2.899414582224588, "grad_norm": 0.2888375222682953, "learning_rate": 7.780256184289495e-08, "loss": 0.1719, "step": 10896 }, { "epoch": 2.8996806812134115, "grad_norm": 0.27535149455070496, "learning_rate": 7.778592690066346e-08, "loss": 0.1875, "step": 10897 }, { "epoch": 2.899946780202235, "grad_norm": 0.32838913798332214, "learning_rate": 7.776929260502596e-08, "loss": 0.1889, "step": 10898 }, { "epoch": 2.9002128791910593, "grad_norm": 0.30014345049858093, "learning_rate": 7.775265895646668e-08, "loss": 0.2024, "step": 10899 }, { "epoch": 2.900478978179883, "grad_norm": 0.30020493268966675, "learning_rate": 7.773602595546968e-08, "loss": 0.1785, "step": 10900 }, { "epoch": 2.9007450771687067, "grad_norm": 0.29580530524253845, "learning_rate": 7.771939360251916e-08, "loss": 0.1925, "step": 10901 }, { "epoch": 2.901011176157531, "grad_norm": 0.26753154397010803, "learning_rate": 7.770276189809925e-08, "loss": 0.1713, "step": 10902 }, { "epoch": 2.9012772751463545, "grad_norm": 0.284746378660202, "learning_rate": 7.768613084269405e-08, "loss": 0.1903, "step": 10903 }, { "epoch": 2.901543374135178, "grad_norm": 0.2853130102157593, "learning_rate": 7.766950043678763e-08, "loss": 0.2041, "step": 10904 }, { "epoch": 2.9018094731240023, "grad_norm": 0.26912885904312134, "learning_rate": 7.765287068086408e-08, "loss": 0.1708, "step": 10905 }, { "epoch": 2.902075572112826, "grad_norm": 0.3019512891769409, "learning_rate": 7.763624157540742e-08, "loss": 0.2173, "step": 10906 }, { "epoch": 2.9023416711016496, "grad_norm": 0.29240959882736206, "learning_rate": 7.761961312090173e-08, "loss": 0.2067, "step": 10907 }, { "epoch": 2.9026077700904738, "grad_norm": 0.2787616550922394, "learning_rate": 7.760298531783102e-08, "loss": 0.1753, "step": 10908 }, { "epoch": 2.9028738690792975, "grad_norm": 0.310057669878006, "learning_rate": 7.75863581666792e-08, "loss": 0.1953, "step": 10909 }, { "epoch": 2.903139968068121, "grad_norm": 0.2875094413757324, "learning_rate": 7.756973166793032e-08, "loss": 0.1991, "step": 10910 }, { "epoch": 2.9034060670569453, "grad_norm": 0.29004615545272827, "learning_rate": 7.755310582206831e-08, "loss": 0.1861, "step": 10911 }, { "epoch": 2.903672166045769, "grad_norm": 0.2882637083530426, "learning_rate": 7.75364806295771e-08, "loss": 0.1832, "step": 10912 }, { "epoch": 2.9039382650345926, "grad_norm": 0.29533031582832336, "learning_rate": 7.751985609094061e-08, "loss": 0.176, "step": 10913 }, { "epoch": 2.9042043640234168, "grad_norm": 0.31050074100494385, "learning_rate": 7.750323220664274e-08, "loss": 0.1867, "step": 10914 }, { "epoch": 2.9044704630122404, "grad_norm": 0.27613505721092224, "learning_rate": 7.748660897716736e-08, "loss": 0.1809, "step": 10915 }, { "epoch": 2.9047365620010646, "grad_norm": 0.3000268042087555, "learning_rate": 7.746998640299835e-08, "loss": 0.2085, "step": 10916 }, { "epoch": 2.9050026609898882, "grad_norm": 0.29671311378479004, "learning_rate": 7.745336448461957e-08, "loss": 0.1872, "step": 10917 }, { "epoch": 2.905268759978712, "grad_norm": 0.30792686343193054, "learning_rate": 7.743674322251478e-08, "loss": 0.2044, "step": 10918 }, { "epoch": 2.905534858967536, "grad_norm": 0.2560020089149475, "learning_rate": 7.74201226171678e-08, "loss": 0.1709, "step": 10919 }, { "epoch": 2.9058009579563597, "grad_norm": 0.3123100996017456, "learning_rate": 7.740350266906242e-08, "loss": 0.1991, "step": 10920 }, { "epoch": 2.906067056945184, "grad_norm": 0.26802775263786316, "learning_rate": 7.738688337868243e-08, "loss": 0.1856, "step": 10921 }, { "epoch": 2.9063331559340075, "grad_norm": 0.28604596853256226, "learning_rate": 7.737026474651154e-08, "loss": 0.1816, "step": 10922 }, { "epoch": 2.906599254922831, "grad_norm": 0.3041936457157135, "learning_rate": 7.735364677303347e-08, "loss": 0.2038, "step": 10923 }, { "epoch": 2.9068653539116553, "grad_norm": 0.2994261085987091, "learning_rate": 7.733702945873195e-08, "loss": 0.1946, "step": 10924 }, { "epoch": 2.907131452900479, "grad_norm": 0.2847512364387512, "learning_rate": 7.732041280409065e-08, "loss": 0.2081, "step": 10925 }, { "epoch": 2.9073975518893027, "grad_norm": 0.3144665062427521, "learning_rate": 7.730379680959329e-08, "loss": 0.1994, "step": 10926 }, { "epoch": 2.907663650878127, "grad_norm": 0.2824505865573883, "learning_rate": 7.728718147572345e-08, "loss": 0.1848, "step": 10927 }, { "epoch": 2.9079297498669505, "grad_norm": 0.28350165486335754, "learning_rate": 7.727056680296477e-08, "loss": 0.2018, "step": 10928 }, { "epoch": 2.908195848855774, "grad_norm": 0.302267849445343, "learning_rate": 7.725395279180086e-08, "loss": 0.2107, "step": 10929 }, { "epoch": 2.9084619478445983, "grad_norm": 0.3050716519355774, "learning_rate": 7.723733944271534e-08, "loss": 0.206, "step": 10930 }, { "epoch": 2.908728046833422, "grad_norm": 0.28561753034591675, "learning_rate": 7.722072675619179e-08, "loss": 0.1853, "step": 10931 }, { "epoch": 2.9089941458222457, "grad_norm": 0.2845350503921509, "learning_rate": 7.720411473271371e-08, "loss": 0.1903, "step": 10932 }, { "epoch": 2.90926024481107, "grad_norm": 0.3055351972579956, "learning_rate": 7.718750337276467e-08, "loss": 0.2224, "step": 10933 }, { "epoch": 2.9095263437998935, "grad_norm": 0.2955787181854248, "learning_rate": 7.717089267682816e-08, "loss": 0.1889, "step": 10934 }, { "epoch": 2.909792442788717, "grad_norm": 0.2706684470176697, "learning_rate": 7.715428264538775e-08, "loss": 0.17, "step": 10935 }, { "epoch": 2.9100585417775413, "grad_norm": 0.2878163754940033, "learning_rate": 7.71376732789268e-08, "loss": 0.1811, "step": 10936 }, { "epoch": 2.910324640766365, "grad_norm": 0.27652114629745483, "learning_rate": 7.712106457792883e-08, "loss": 0.1791, "step": 10937 }, { "epoch": 2.9105907397551887, "grad_norm": 0.28568515181541443, "learning_rate": 7.710445654287725e-08, "loss": 0.1758, "step": 10938 }, { "epoch": 2.910856838744013, "grad_norm": 0.2843650281429291, "learning_rate": 7.70878491742555e-08, "loss": 0.1794, "step": 10939 }, { "epoch": 2.9111229377328365, "grad_norm": 0.2923468053340912, "learning_rate": 7.707124247254697e-08, "loss": 0.1949, "step": 10940 }, { "epoch": 2.9113890367216606, "grad_norm": 0.2922571301460266, "learning_rate": 7.705463643823506e-08, "loss": 0.2106, "step": 10941 }, { "epoch": 2.9116551357104843, "grad_norm": 0.30267322063446045, "learning_rate": 7.703803107180309e-08, "loss": 0.1981, "step": 10942 }, { "epoch": 2.9119212346993084, "grad_norm": 0.28829964995384216, "learning_rate": 7.702142637373441e-08, "loss": 0.1806, "step": 10943 }, { "epoch": 2.912187333688132, "grad_norm": 0.31026846170425415, "learning_rate": 7.700482234451238e-08, "loss": 0.2044, "step": 10944 }, { "epoch": 2.9124534326769558, "grad_norm": 0.3206580579280853, "learning_rate": 7.698821898462024e-08, "loss": 0.1828, "step": 10945 }, { "epoch": 2.91271953166578, "grad_norm": 0.2908343970775604, "learning_rate": 7.697161629454127e-08, "loss": 0.1903, "step": 10946 }, { "epoch": 2.9129856306546036, "grad_norm": 0.2602905035018921, "learning_rate": 7.695501427475877e-08, "loss": 0.168, "step": 10947 }, { "epoch": 2.9132517296434273, "grad_norm": 0.3061143755912781, "learning_rate": 7.693841292575597e-08, "loss": 0.2045, "step": 10948 }, { "epoch": 2.9135178286322514, "grad_norm": 0.3279709815979004, "learning_rate": 7.69218122480161e-08, "loss": 0.1917, "step": 10949 }, { "epoch": 2.913783927621075, "grad_norm": 0.2878625988960266, "learning_rate": 7.690521224202235e-08, "loss": 0.1768, "step": 10950 }, { "epoch": 2.9140500266098988, "grad_norm": 0.2831009030342102, "learning_rate": 7.688861290825792e-08, "loss": 0.1841, "step": 10951 }, { "epoch": 2.914316125598723, "grad_norm": 0.2903134524822235, "learning_rate": 7.687201424720596e-08, "loss": 0.1821, "step": 10952 }, { "epoch": 2.9145822245875466, "grad_norm": 0.2787277102470398, "learning_rate": 7.685541625934963e-08, "loss": 0.1915, "step": 10953 }, { "epoch": 2.9148483235763702, "grad_norm": 0.2865065932273865, "learning_rate": 7.6838818945172e-08, "loss": 0.1794, "step": 10954 }, { "epoch": 2.9151144225651944, "grad_norm": 0.30481064319610596, "learning_rate": 7.682222230515621e-08, "loss": 0.1898, "step": 10955 }, { "epoch": 2.915380521554018, "grad_norm": 0.3168529272079468, "learning_rate": 7.680562633978536e-08, "loss": 0.2005, "step": 10956 }, { "epoch": 2.9156466205428417, "grad_norm": 0.2899443805217743, "learning_rate": 7.678903104954251e-08, "loss": 0.1758, "step": 10957 }, { "epoch": 2.915912719531666, "grad_norm": 0.2759231626987457, "learning_rate": 7.677243643491068e-08, "loss": 0.1954, "step": 10958 }, { "epoch": 2.9161788185204895, "grad_norm": 0.29830703139305115, "learning_rate": 7.675584249637293e-08, "loss": 0.1893, "step": 10959 }, { "epoch": 2.9164449175093132, "grad_norm": 0.29595717787742615, "learning_rate": 7.673924923441227e-08, "loss": 0.1856, "step": 10960 }, { "epoch": 2.9167110164981374, "grad_norm": 0.30767419934272766, "learning_rate": 7.672265664951164e-08, "loss": 0.1986, "step": 10961 }, { "epoch": 2.916977115486961, "grad_norm": 0.2981856167316437, "learning_rate": 7.670606474215406e-08, "loss": 0.195, "step": 10962 }, { "epoch": 2.9172432144757847, "grad_norm": 0.3009563386440277, "learning_rate": 7.668947351282245e-08, "loss": 0.1931, "step": 10963 }, { "epoch": 2.917509313464609, "grad_norm": 0.28821927309036255, "learning_rate": 7.667288296199975e-08, "loss": 0.1922, "step": 10964 }, { "epoch": 2.9177754124534325, "grad_norm": 0.29286497831344604, "learning_rate": 7.665629309016883e-08, "loss": 0.1943, "step": 10965 }, { "epoch": 2.9180415114422567, "grad_norm": 0.2982037365436554, "learning_rate": 7.66397038978126e-08, "loss": 0.2034, "step": 10966 }, { "epoch": 2.9183076104310803, "grad_norm": 0.3147311806678772, "learning_rate": 7.662311538541394e-08, "loss": 0.1933, "step": 10967 }, { "epoch": 2.9185737094199045, "grad_norm": 0.2871447801589966, "learning_rate": 7.660652755345572e-08, "loss": 0.1722, "step": 10968 }, { "epoch": 2.918839808408728, "grad_norm": 0.27245616912841797, "learning_rate": 7.65899404024207e-08, "loss": 0.184, "step": 10969 }, { "epoch": 2.919105907397552, "grad_norm": 0.2978275716304779, "learning_rate": 7.657335393279178e-08, "loss": 0.1947, "step": 10970 }, { "epoch": 2.919372006386376, "grad_norm": 0.28785592317581177, "learning_rate": 7.65567681450517e-08, "loss": 0.1936, "step": 10971 }, { "epoch": 2.9196381053751996, "grad_norm": 0.29509568214416504, "learning_rate": 7.65401830396832e-08, "loss": 0.2076, "step": 10972 }, { "epoch": 2.9199042043640233, "grad_norm": 0.2922914922237396, "learning_rate": 7.652359861716908e-08, "loss": 0.2156, "step": 10973 }, { "epoch": 2.9201703033528474, "grad_norm": 0.29343181848526, "learning_rate": 7.650701487799204e-08, "loss": 0.1869, "step": 10974 }, { "epoch": 2.920436402341671, "grad_norm": 0.30960968136787415, "learning_rate": 7.649043182263479e-08, "loss": 0.1825, "step": 10975 }, { "epoch": 2.920702501330495, "grad_norm": 0.30073949694633484, "learning_rate": 7.647384945158002e-08, "loss": 0.2055, "step": 10976 }, { "epoch": 2.920968600319319, "grad_norm": 0.2788662612438202, "learning_rate": 7.645726776531042e-08, "loss": 0.192, "step": 10977 }, { "epoch": 2.9212346993081426, "grad_norm": 0.2969006896018982, "learning_rate": 7.644068676430861e-08, "loss": 0.1871, "step": 10978 }, { "epoch": 2.9215007982969663, "grad_norm": 0.27359291911125183, "learning_rate": 7.642410644905725e-08, "loss": 0.169, "step": 10979 }, { "epoch": 2.9217668972857904, "grad_norm": 0.32180944085121155, "learning_rate": 7.640752682003899e-08, "loss": 0.2121, "step": 10980 }, { "epoch": 2.922032996274614, "grad_norm": 0.3041050434112549, "learning_rate": 7.639094787773631e-08, "loss": 0.1959, "step": 10981 }, { "epoch": 2.922299095263438, "grad_norm": 0.2711631953716278, "learning_rate": 7.637436962263184e-08, "loss": 0.1849, "step": 10982 }, { "epoch": 2.922565194252262, "grad_norm": 0.2922423481941223, "learning_rate": 7.635779205520815e-08, "loss": 0.1845, "step": 10983 }, { "epoch": 2.9228312932410856, "grad_norm": 0.31275463104248047, "learning_rate": 7.634121517594772e-08, "loss": 0.1857, "step": 10984 }, { "epoch": 2.9230973922299093, "grad_norm": 0.32057827711105347, "learning_rate": 7.632463898533309e-08, "loss": 0.2134, "step": 10985 }, { "epoch": 2.9233634912187334, "grad_norm": 0.24135980010032654, "learning_rate": 7.630806348384675e-08, "loss": 0.1605, "step": 10986 }, { "epoch": 2.923629590207557, "grad_norm": 0.29566797614097595, "learning_rate": 7.629148867197116e-08, "loss": 0.2049, "step": 10987 }, { "epoch": 2.923895689196381, "grad_norm": 0.29551953077316284, "learning_rate": 7.627491455018877e-08, "loss": 0.209, "step": 10988 }, { "epoch": 2.924161788185205, "grad_norm": 0.2948092222213745, "learning_rate": 7.625834111898208e-08, "loss": 0.2142, "step": 10989 }, { "epoch": 2.9244278871740286, "grad_norm": 0.3166302740573883, "learning_rate": 7.624176837883338e-08, "loss": 0.2042, "step": 10990 }, { "epoch": 2.9246939861628527, "grad_norm": 0.2929497957229614, "learning_rate": 7.622519633022512e-08, "loss": 0.1995, "step": 10991 }, { "epoch": 2.9249600851516764, "grad_norm": 0.2750489115715027, "learning_rate": 7.620862497363965e-08, "loss": 0.1834, "step": 10992 }, { "epoch": 2.9252261841405005, "grad_norm": 0.332745760679245, "learning_rate": 7.619205430955937e-08, "loss": 0.2184, "step": 10993 }, { "epoch": 2.925492283129324, "grad_norm": 0.2969605326652527, "learning_rate": 7.617548433846654e-08, "loss": 0.197, "step": 10994 }, { "epoch": 2.925758382118148, "grad_norm": 0.34073197841644287, "learning_rate": 7.615891506084351e-08, "loss": 0.1784, "step": 10995 }, { "epoch": 2.926024481106972, "grad_norm": 0.284100741147995, "learning_rate": 7.614234647717256e-08, "loss": 0.1957, "step": 10996 }, { "epoch": 2.9262905800957957, "grad_norm": 0.37425315380096436, "learning_rate": 7.612577858793595e-08, "loss": 0.1719, "step": 10997 }, { "epoch": 2.9265566790846194, "grad_norm": 0.2945522367954254, "learning_rate": 7.610921139361594e-08, "loss": 0.215, "step": 10998 }, { "epoch": 2.9268227780734435, "grad_norm": 0.31383752822875977, "learning_rate": 7.60926448946948e-08, "loss": 0.2021, "step": 10999 }, { "epoch": 2.927088877062267, "grad_norm": 0.27603456377983093, "learning_rate": 7.607607909165465e-08, "loss": 0.1924, "step": 11000 }, { "epoch": 2.927354976051091, "grad_norm": 0.3070998191833496, "learning_rate": 7.605951398497772e-08, "loss": 0.1816, "step": 11001 }, { "epoch": 2.927621075039915, "grad_norm": 0.31703588366508484, "learning_rate": 7.60429495751462e-08, "loss": 0.2086, "step": 11002 }, { "epoch": 2.9278871740287387, "grad_norm": 0.30882105231285095, "learning_rate": 7.602638586264218e-08, "loss": 0.198, "step": 11003 }, { "epoch": 2.9281532730175623, "grad_norm": 0.2975292503833771, "learning_rate": 7.600982284794784e-08, "loss": 0.1791, "step": 11004 }, { "epoch": 2.9284193720063865, "grad_norm": 0.3104197084903717, "learning_rate": 7.599326053154524e-08, "loss": 0.1997, "step": 11005 }, { "epoch": 2.92868547099521, "grad_norm": 0.30741533637046814, "learning_rate": 7.597669891391652e-08, "loss": 0.2071, "step": 11006 }, { "epoch": 2.928951569984034, "grad_norm": 0.31381338834762573, "learning_rate": 7.59601379955437e-08, "loss": 0.197, "step": 11007 }, { "epoch": 2.929217668972858, "grad_norm": 0.2981858551502228, "learning_rate": 7.594357777690891e-08, "loss": 0.1982, "step": 11008 }, { "epoch": 2.9294837679616816, "grad_norm": 0.2913302183151245, "learning_rate": 7.592701825849405e-08, "loss": 0.183, "step": 11009 }, { "epoch": 2.9297498669505053, "grad_norm": 0.2977618873119354, "learning_rate": 7.591045944078118e-08, "loss": 0.2042, "step": 11010 }, { "epoch": 2.9300159659393294, "grad_norm": 0.2879282534122467, "learning_rate": 7.58939013242523e-08, "loss": 0.1885, "step": 11011 }, { "epoch": 2.930282064928153, "grad_norm": 0.3129747211933136, "learning_rate": 7.587734390938937e-08, "loss": 0.1907, "step": 11012 }, { "epoch": 2.9305481639169773, "grad_norm": 0.2635425627231598, "learning_rate": 7.586078719667432e-08, "loss": 0.1818, "step": 11013 }, { "epoch": 2.930814262905801, "grad_norm": 0.3114442527294159, "learning_rate": 7.584423118658905e-08, "loss": 0.2005, "step": 11014 }, { "epoch": 2.931080361894625, "grad_norm": 0.2682042121887207, "learning_rate": 7.582767587961552e-08, "loss": 0.1949, "step": 11015 }, { "epoch": 2.9313464608834487, "grad_norm": 0.3200417757034302, "learning_rate": 7.581112127623557e-08, "loss": 0.1913, "step": 11016 }, { "epoch": 2.9316125598722724, "grad_norm": 0.2986951768398285, "learning_rate": 7.57945673769311e-08, "loss": 0.2108, "step": 11017 }, { "epoch": 2.9318786588610966, "grad_norm": 0.30608126521110535, "learning_rate": 7.577801418218391e-08, "loss": 0.1936, "step": 11018 }, { "epoch": 2.9321447578499202, "grad_norm": 0.2978016138076782, "learning_rate": 7.576146169247583e-08, "loss": 0.1867, "step": 11019 }, { "epoch": 2.932410856838744, "grad_norm": 0.30680498480796814, "learning_rate": 7.574490990828866e-08, "loss": 0.195, "step": 11020 }, { "epoch": 2.932676955827568, "grad_norm": 0.32718291878700256, "learning_rate": 7.572835883010418e-08, "loss": 0.2158, "step": 11021 }, { "epoch": 2.9329430548163917, "grad_norm": 0.3071396052837372, "learning_rate": 7.571180845840417e-08, "loss": 0.2158, "step": 11022 }, { "epoch": 2.9332091538052154, "grad_norm": 0.3068293631076813, "learning_rate": 7.569525879367035e-08, "loss": 0.1912, "step": 11023 }, { "epoch": 2.9334752527940395, "grad_norm": 0.2886282801628113, "learning_rate": 7.567870983638442e-08, "loss": 0.1957, "step": 11024 }, { "epoch": 2.933741351782863, "grad_norm": 0.2879372537136078, "learning_rate": 7.566216158702812e-08, "loss": 0.2038, "step": 11025 }, { "epoch": 2.934007450771687, "grad_norm": 0.30169329047203064, "learning_rate": 7.564561404608312e-08, "loss": 0.1976, "step": 11026 }, { "epoch": 2.934273549760511, "grad_norm": 0.2998543381690979, "learning_rate": 7.562906721403104e-08, "loss": 0.1855, "step": 11027 }, { "epoch": 2.9345396487493347, "grad_norm": 0.2959960401058197, "learning_rate": 7.561252109135354e-08, "loss": 0.1838, "step": 11028 }, { "epoch": 2.9348057477381584, "grad_norm": 0.31615856289863586, "learning_rate": 7.559597567853219e-08, "loss": 0.1978, "step": 11029 }, { "epoch": 2.9350718467269825, "grad_norm": 0.2975878417491913, "learning_rate": 7.557943097604867e-08, "loss": 0.2049, "step": 11030 }, { "epoch": 2.935337945715806, "grad_norm": 0.27795398235321045, "learning_rate": 7.556288698438447e-08, "loss": 0.1762, "step": 11031 }, { "epoch": 2.93560404470463, "grad_norm": 0.3040209710597992, "learning_rate": 7.554634370402122e-08, "loss": 0.1911, "step": 11032 }, { "epoch": 2.935870143693454, "grad_norm": 0.2693542540073395, "learning_rate": 7.552980113544039e-08, "loss": 0.1866, "step": 11033 }, { "epoch": 2.9361362426822777, "grad_norm": 0.3212852478027344, "learning_rate": 7.55132592791235e-08, "loss": 0.2006, "step": 11034 }, { "epoch": 2.936402341671102, "grad_norm": 0.2720644176006317, "learning_rate": 7.549671813555211e-08, "loss": 0.186, "step": 11035 }, { "epoch": 2.9366684406599255, "grad_norm": 0.28149649500846863, "learning_rate": 7.54801777052076e-08, "loss": 0.1815, "step": 11036 }, { "epoch": 2.936934539648749, "grad_norm": 0.3100929260253906, "learning_rate": 7.546363798857142e-08, "loss": 0.1961, "step": 11037 }, { "epoch": 2.9372006386375733, "grad_norm": 0.2966582477092743, "learning_rate": 7.544709898612506e-08, "loss": 0.1907, "step": 11038 }, { "epoch": 2.937466737626397, "grad_norm": 0.28288179636001587, "learning_rate": 7.543056069834988e-08, "loss": 0.1949, "step": 11039 }, { "epoch": 2.937732836615221, "grad_norm": 0.3052208721637726, "learning_rate": 7.541402312572729e-08, "loss": 0.2076, "step": 11040 }, { "epoch": 2.937998935604045, "grad_norm": 0.32696452736854553, "learning_rate": 7.539748626873866e-08, "loss": 0.2061, "step": 11041 }, { "epoch": 2.9382650345928685, "grad_norm": 0.2928582429885864, "learning_rate": 7.538095012786532e-08, "loss": 0.1929, "step": 11042 }, { "epoch": 2.9385311335816926, "grad_norm": 0.30216720700263977, "learning_rate": 7.53644147035886e-08, "loss": 0.2025, "step": 11043 }, { "epoch": 2.9387972325705163, "grad_norm": 0.2996530532836914, "learning_rate": 7.534787999638982e-08, "loss": 0.2002, "step": 11044 }, { "epoch": 2.93906333155934, "grad_norm": 0.2740170657634735, "learning_rate": 7.533134600675024e-08, "loss": 0.1864, "step": 11045 }, { "epoch": 2.939329430548164, "grad_norm": 0.2943499982357025, "learning_rate": 7.53148127351511e-08, "loss": 0.1999, "step": 11046 }, { "epoch": 2.9395955295369878, "grad_norm": 0.290500670671463, "learning_rate": 7.529828018207368e-08, "loss": 0.1796, "step": 11047 }, { "epoch": 2.9398616285258115, "grad_norm": 0.27812814712524414, "learning_rate": 7.528174834799917e-08, "loss": 0.183, "step": 11048 }, { "epoch": 2.9401277275146356, "grad_norm": 0.27791911363601685, "learning_rate": 7.526521723340881e-08, "loss": 0.1861, "step": 11049 }, { "epoch": 2.9403938265034593, "grad_norm": 0.3135467767715454, "learning_rate": 7.524868683878373e-08, "loss": 0.2025, "step": 11050 }, { "epoch": 2.940659925492283, "grad_norm": 0.3284160792827606, "learning_rate": 7.523215716460514e-08, "loss": 0.2152, "step": 11051 }, { "epoch": 2.940926024481107, "grad_norm": 0.3354998230934143, "learning_rate": 7.521562821135413e-08, "loss": 0.2049, "step": 11052 }, { "epoch": 2.9411921234699308, "grad_norm": 0.28917983174324036, "learning_rate": 7.519909997951184e-08, "loss": 0.1847, "step": 11053 }, { "epoch": 2.9414582224587544, "grad_norm": 0.2906172275543213, "learning_rate": 7.518257246955937e-08, "loss": 0.1946, "step": 11054 }, { "epoch": 2.9417243214475786, "grad_norm": 0.29781678318977356, "learning_rate": 7.516604568197774e-08, "loss": 0.1971, "step": 11055 }, { "epoch": 2.9419904204364022, "grad_norm": 0.2797550857067108, "learning_rate": 7.514951961724805e-08, "loss": 0.1916, "step": 11056 }, { "epoch": 2.942256519425226, "grad_norm": 0.29964715242385864, "learning_rate": 7.513299427585131e-08, "loss": 0.1967, "step": 11057 }, { "epoch": 2.94252261841405, "grad_norm": 0.2700154483318329, "learning_rate": 7.511646965826856e-08, "loss": 0.1886, "step": 11058 }, { "epoch": 2.9427887174028737, "grad_norm": 0.29168304800987244, "learning_rate": 7.509994576498075e-08, "loss": 0.2052, "step": 11059 }, { "epoch": 2.943054816391698, "grad_norm": 0.31655585765838623, "learning_rate": 7.508342259646887e-08, "loss": 0.21, "step": 11060 }, { "epoch": 2.9433209153805215, "grad_norm": 0.311442494392395, "learning_rate": 7.506690015321389e-08, "loss": 0.2092, "step": 11061 }, { "epoch": 2.9435870143693457, "grad_norm": 0.33957427740097046, "learning_rate": 7.505037843569671e-08, "loss": 0.2044, "step": 11062 }, { "epoch": 2.9438531133581693, "grad_norm": 0.3004060685634613, "learning_rate": 7.50338574443982e-08, "loss": 0.1948, "step": 11063 }, { "epoch": 2.944119212346993, "grad_norm": 0.29076382517814636, "learning_rate": 7.50173371797993e-08, "loss": 0.1928, "step": 11064 }, { "epoch": 2.944385311335817, "grad_norm": 0.3128073513507843, "learning_rate": 7.500081764238084e-08, "loss": 0.2057, "step": 11065 }, { "epoch": 2.944651410324641, "grad_norm": 0.3104762136936188, "learning_rate": 7.498429883262365e-08, "loss": 0.2044, "step": 11066 }, { "epoch": 2.9449175093134645, "grad_norm": 0.3118605613708496, "learning_rate": 7.496778075100858e-08, "loss": 0.2189, "step": 11067 }, { "epoch": 2.9451836083022886, "grad_norm": 0.34436801075935364, "learning_rate": 7.495126339801641e-08, "loss": 0.1963, "step": 11068 }, { "epoch": 2.9454497072911123, "grad_norm": 0.2913665175437927, "learning_rate": 7.493474677412793e-08, "loss": 0.1858, "step": 11069 }, { "epoch": 2.945715806279936, "grad_norm": 0.3092983365058899, "learning_rate": 7.491823087982389e-08, "loss": 0.184, "step": 11070 }, { "epoch": 2.94598190526876, "grad_norm": 0.3043140470981598, "learning_rate": 7.490171571558507e-08, "loss": 0.207, "step": 11071 }, { "epoch": 2.946248004257584, "grad_norm": 0.29045018553733826, "learning_rate": 7.488520128189209e-08, "loss": 0.1969, "step": 11072 }, { "epoch": 2.9465141032464075, "grad_norm": 0.3231043517589569, "learning_rate": 7.486868757922569e-08, "loss": 0.1937, "step": 11073 }, { "epoch": 2.9467802022352316, "grad_norm": 0.27461573481559753, "learning_rate": 7.485217460806657e-08, "loss": 0.1821, "step": 11074 }, { "epoch": 2.9470463012240553, "grad_norm": 0.2956825792789459, "learning_rate": 7.483566236889533e-08, "loss": 0.2047, "step": 11075 }, { "epoch": 2.947312400212879, "grad_norm": 0.30239585041999817, "learning_rate": 7.481915086219262e-08, "loss": 0.1932, "step": 11076 }, { "epoch": 2.947578499201703, "grad_norm": 0.30055615305900574, "learning_rate": 7.480264008843903e-08, "loss": 0.2133, "step": 11077 }, { "epoch": 2.947844598190527, "grad_norm": 0.28424209356307983, "learning_rate": 7.478613004811518e-08, "loss": 0.1883, "step": 11078 }, { "epoch": 2.9481106971793505, "grad_norm": 0.2979631721973419, "learning_rate": 7.47696207417016e-08, "loss": 0.2021, "step": 11079 }, { "epoch": 2.9483767961681746, "grad_norm": 0.27166303992271423, "learning_rate": 7.475311216967893e-08, "loss": 0.1781, "step": 11080 }, { "epoch": 2.9486428951569983, "grad_norm": 0.2921275198459625, "learning_rate": 7.473660433252753e-08, "loss": 0.1842, "step": 11081 }, { "epoch": 2.948908994145822, "grad_norm": 0.30739739537239075, "learning_rate": 7.4720097230728e-08, "loss": 0.1916, "step": 11082 }, { "epoch": 2.949175093134646, "grad_norm": 0.2949550747871399, "learning_rate": 7.470359086476079e-08, "loss": 0.2043, "step": 11083 }, { "epoch": 2.94944119212347, "grad_norm": 0.3230539560317993, "learning_rate": 7.468708523510641e-08, "loss": 0.2083, "step": 11084 }, { "epoch": 2.949707291112294, "grad_norm": 0.2881339490413666, "learning_rate": 7.46705803422452e-08, "loss": 0.1882, "step": 11085 }, { "epoch": 2.9499733901011176, "grad_norm": 0.2816593050956726, "learning_rate": 7.465407618665767e-08, "loss": 0.1817, "step": 11086 }, { "epoch": 2.9502394890899417, "grad_norm": 0.2949644923210144, "learning_rate": 7.463757276882414e-08, "loss": 0.2021, "step": 11087 }, { "epoch": 2.9505055880787654, "grad_norm": 0.2863323390483856, "learning_rate": 7.462107008922503e-08, "loss": 0.1832, "step": 11088 }, { "epoch": 2.950771687067589, "grad_norm": 0.2836029827594757, "learning_rate": 7.46045681483407e-08, "loss": 0.1954, "step": 11089 }, { "epoch": 2.951037786056413, "grad_norm": 0.2757031321525574, "learning_rate": 7.458806694665143e-08, "loss": 0.1862, "step": 11090 }, { "epoch": 2.951303885045237, "grad_norm": 0.28676706552505493, "learning_rate": 7.457156648463754e-08, "loss": 0.181, "step": 11091 }, { "epoch": 2.9515699840340606, "grad_norm": 0.2874261438846588, "learning_rate": 7.455506676277934e-08, "loss": 0.1976, "step": 11092 }, { "epoch": 2.9518360830228847, "grad_norm": 0.27519115805625916, "learning_rate": 7.453856778155706e-08, "loss": 0.1953, "step": 11093 }, { "epoch": 2.9521021820117084, "grad_norm": 0.2926754355430603, "learning_rate": 7.4522069541451e-08, "loss": 0.1965, "step": 11094 }, { "epoch": 2.952368281000532, "grad_norm": 0.275546669960022, "learning_rate": 7.45055720429413e-08, "loss": 0.1826, "step": 11095 }, { "epoch": 2.952634379989356, "grad_norm": 0.3169638216495514, "learning_rate": 7.448907528650821e-08, "loss": 0.2168, "step": 11096 }, { "epoch": 2.95290047897818, "grad_norm": 0.3134723901748657, "learning_rate": 7.44725792726319e-08, "loss": 0.1984, "step": 11097 }, { "epoch": 2.9531665779670035, "grad_norm": 0.29785481095314026, "learning_rate": 7.445608400179259e-08, "loss": 0.1865, "step": 11098 }, { "epoch": 2.9534326769558277, "grad_norm": 0.310914546251297, "learning_rate": 7.443958947447028e-08, "loss": 0.185, "step": 11099 }, { "epoch": 2.9536987759446514, "grad_norm": 0.33208417892456055, "learning_rate": 7.442309569114517e-08, "loss": 0.2101, "step": 11100 }, { "epoch": 2.953964874933475, "grad_norm": 0.2872978150844574, "learning_rate": 7.440660265229732e-08, "loss": 0.1905, "step": 11101 }, { "epoch": 2.954230973922299, "grad_norm": 0.2855404019355774, "learning_rate": 7.439011035840683e-08, "loss": 0.1908, "step": 11102 }, { "epoch": 2.954497072911123, "grad_norm": 0.3128592073917389, "learning_rate": 7.437361880995374e-08, "loss": 0.1899, "step": 11103 }, { "epoch": 2.9547631718999465, "grad_norm": 0.3276713192462921, "learning_rate": 7.435712800741806e-08, "loss": 0.1933, "step": 11104 }, { "epoch": 2.9550292708887707, "grad_norm": 0.3083621561527252, "learning_rate": 7.434063795127979e-08, "loss": 0.2034, "step": 11105 }, { "epoch": 2.9552953698775943, "grad_norm": 0.3139045238494873, "learning_rate": 7.432414864201894e-08, "loss": 0.2132, "step": 11106 }, { "epoch": 2.9555614688664185, "grad_norm": 0.30552005767822266, "learning_rate": 7.430766008011549e-08, "loss": 0.1992, "step": 11107 }, { "epoch": 2.955827567855242, "grad_norm": 0.2859436571598053, "learning_rate": 7.42911722660493e-08, "loss": 0.2002, "step": 11108 }, { "epoch": 2.9560936668440663, "grad_norm": 0.3075792193412781, "learning_rate": 7.427468520030033e-08, "loss": 0.2064, "step": 11109 }, { "epoch": 2.95635976583289, "grad_norm": 0.29755377769470215, "learning_rate": 7.42581988833485e-08, "loss": 0.204, "step": 11110 }, { "epoch": 2.9566258648217136, "grad_norm": 0.3397372364997864, "learning_rate": 7.424171331567364e-08, "loss": 0.2028, "step": 11111 }, { "epoch": 2.9568919638105378, "grad_norm": 0.27815383672714233, "learning_rate": 7.422522849775563e-08, "loss": 0.1663, "step": 11112 }, { "epoch": 2.9571580627993614, "grad_norm": 0.31155577301979065, "learning_rate": 7.420874443007432e-08, "loss": 0.2082, "step": 11113 }, { "epoch": 2.957424161788185, "grad_norm": 0.28241023421287537, "learning_rate": 7.419226111310947e-08, "loss": 0.2002, "step": 11114 }, { "epoch": 2.9576902607770093, "grad_norm": 0.300374835729599, "learning_rate": 7.417577854734089e-08, "loss": 0.1924, "step": 11115 }, { "epoch": 2.957956359765833, "grad_norm": 0.2917901277542114, "learning_rate": 7.415929673324839e-08, "loss": 0.2025, "step": 11116 }, { "epoch": 2.9582224587546566, "grad_norm": 0.3281044065952301, "learning_rate": 7.414281567131162e-08, "loss": 0.192, "step": 11117 }, { "epoch": 2.9584885577434807, "grad_norm": 0.30879083275794983, "learning_rate": 7.412633536201035e-08, "loss": 0.2048, "step": 11118 }, { "epoch": 2.9587546567323044, "grad_norm": 0.2901454567909241, "learning_rate": 7.410985580582427e-08, "loss": 0.1731, "step": 11119 }, { "epoch": 2.959020755721128, "grad_norm": 0.3302680552005768, "learning_rate": 7.409337700323308e-08, "loss": 0.2117, "step": 11120 }, { "epoch": 2.9592868547099522, "grad_norm": 0.2952454388141632, "learning_rate": 7.40768989547164e-08, "loss": 0.1907, "step": 11121 }, { "epoch": 2.959552953698776, "grad_norm": 0.31544890999794006, "learning_rate": 7.40604216607539e-08, "loss": 0.2015, "step": 11122 }, { "epoch": 2.9598190526875996, "grad_norm": 0.30279701948165894, "learning_rate": 7.404394512182519e-08, "loss": 0.1985, "step": 11123 }, { "epoch": 2.9600851516764237, "grad_norm": 0.31507202982902527, "learning_rate": 7.402746933840983e-08, "loss": 0.2119, "step": 11124 }, { "epoch": 2.9603512506652474, "grad_norm": 0.31103846430778503, "learning_rate": 7.401099431098741e-08, "loss": 0.2224, "step": 11125 }, { "epoch": 2.960617349654071, "grad_norm": 0.2953406870365143, "learning_rate": 7.399452004003747e-08, "loss": 0.1784, "step": 11126 }, { "epoch": 2.960883448642895, "grad_norm": 0.2887861132621765, "learning_rate": 7.397804652603951e-08, "loss": 0.1874, "step": 11127 }, { "epoch": 2.961149547631719, "grad_norm": 0.2800036370754242, "learning_rate": 7.396157376947305e-08, "loss": 0.1676, "step": 11128 }, { "epoch": 2.9614156466205426, "grad_norm": 0.27167809009552, "learning_rate": 7.394510177081756e-08, "loss": 0.1785, "step": 11129 }, { "epoch": 2.9616817456093667, "grad_norm": 0.29688262939453125, "learning_rate": 7.392863053055251e-08, "loss": 0.1788, "step": 11130 }, { "epoch": 2.9619478445981904, "grad_norm": 0.2820795774459839, "learning_rate": 7.391216004915735e-08, "loss": 0.1939, "step": 11131 }, { "epoch": 2.9622139435870145, "grad_norm": 0.307461678981781, "learning_rate": 7.389569032711145e-08, "loss": 0.1993, "step": 11132 }, { "epoch": 2.962480042575838, "grad_norm": 0.2981698513031006, "learning_rate": 7.387922136489427e-08, "loss": 0.2101, "step": 11133 }, { "epoch": 2.9627461415646623, "grad_norm": 0.3191472291946411, "learning_rate": 7.386275316298511e-08, "loss": 0.2138, "step": 11134 }, { "epoch": 2.963012240553486, "grad_norm": 0.31682947278022766, "learning_rate": 7.384628572186333e-08, "loss": 0.2094, "step": 11135 }, { "epoch": 2.9632783395423097, "grad_norm": 0.30254602432250977, "learning_rate": 7.382981904200829e-08, "loss": 0.2049, "step": 11136 }, { "epoch": 2.963544438531134, "grad_norm": 0.3050227761268616, "learning_rate": 7.381335312389925e-08, "loss": 0.189, "step": 11137 }, { "epoch": 2.9638105375199575, "grad_norm": 0.2899463176727295, "learning_rate": 7.379688796801549e-08, "loss": 0.2006, "step": 11138 }, { "epoch": 2.964076636508781, "grad_norm": 0.2963721752166748, "learning_rate": 7.378042357483629e-08, "loss": 0.1909, "step": 11139 }, { "epoch": 2.9643427354976053, "grad_norm": 0.3070833683013916, "learning_rate": 7.376395994484091e-08, "loss": 0.2024, "step": 11140 }, { "epoch": 2.964608834486429, "grad_norm": 0.2820407450199127, "learning_rate": 7.374749707850849e-08, "loss": 0.1876, "step": 11141 }, { "epoch": 2.9648749334752527, "grad_norm": 0.30563122034072876, "learning_rate": 7.373103497631829e-08, "loss": 0.212, "step": 11142 }, { "epoch": 2.965141032464077, "grad_norm": 0.30180439352989197, "learning_rate": 7.371457363874951e-08, "loss": 0.1877, "step": 11143 }, { "epoch": 2.9654071314529005, "grad_norm": 0.30557921528816223, "learning_rate": 7.369811306628118e-08, "loss": 0.2049, "step": 11144 }, { "epoch": 2.965673230441724, "grad_norm": 0.26924481987953186, "learning_rate": 7.368165325939251e-08, "loss": 0.1806, "step": 11145 }, { "epoch": 2.9659393294305483, "grad_norm": 0.30214977264404297, "learning_rate": 7.366519421856257e-08, "loss": 0.1986, "step": 11146 }, { "epoch": 2.966205428419372, "grad_norm": 0.2922070622444153, "learning_rate": 7.364873594427043e-08, "loss": 0.1956, "step": 11147 }, { "epoch": 2.9664715274081956, "grad_norm": 0.2936665713787079, "learning_rate": 7.363227843699517e-08, "loss": 0.2056, "step": 11148 }, { "epoch": 2.9667376263970198, "grad_norm": 0.31704896688461304, "learning_rate": 7.361582169721584e-08, "loss": 0.2019, "step": 11149 }, { "epoch": 2.9670037253858434, "grad_norm": 0.3064086437225342, "learning_rate": 7.359936572541141e-08, "loss": 0.2026, "step": 11150 }, { "epoch": 2.967269824374667, "grad_norm": 0.27569642663002014, "learning_rate": 7.358291052206092e-08, "loss": 0.1791, "step": 11151 }, { "epoch": 2.9675359233634913, "grad_norm": 0.27835145592689514, "learning_rate": 7.356645608764333e-08, "loss": 0.1808, "step": 11152 }, { "epoch": 2.967802022352315, "grad_norm": 0.29086682200431824, "learning_rate": 7.355000242263755e-08, "loss": 0.1929, "step": 11153 }, { "epoch": 2.968068121341139, "grad_norm": 0.304595947265625, "learning_rate": 7.35335495275225e-08, "loss": 0.1989, "step": 11154 }, { "epoch": 2.9683342203299627, "grad_norm": 0.2727465033531189, "learning_rate": 7.351709740277714e-08, "loss": 0.185, "step": 11155 }, { "epoch": 2.9686003193187864, "grad_norm": 0.2821010947227478, "learning_rate": 7.350064604888029e-08, "loss": 0.1919, "step": 11156 }, { "epoch": 2.9688664183076106, "grad_norm": 0.28998711705207825, "learning_rate": 7.348419546631085e-08, "loss": 0.178, "step": 11157 }, { "epoch": 2.9691325172964342, "grad_norm": 0.2917974293231964, "learning_rate": 7.346774565554759e-08, "loss": 0.1827, "step": 11158 }, { "epoch": 2.9693986162852584, "grad_norm": 0.2725718319416046, "learning_rate": 7.34512966170694e-08, "loss": 0.1863, "step": 11159 }, { "epoch": 2.969664715274082, "grad_norm": 0.319010853767395, "learning_rate": 7.343484835135501e-08, "loss": 0.2082, "step": 11160 }, { "epoch": 2.9699308142629057, "grad_norm": 0.30565449595451355, "learning_rate": 7.341840085888324e-08, "loss": 0.2007, "step": 11161 }, { "epoch": 2.97019691325173, "grad_norm": 0.30013084411621094, "learning_rate": 7.340195414013278e-08, "loss": 0.1927, "step": 11162 }, { "epoch": 2.9704630122405535, "grad_norm": 0.3105136454105377, "learning_rate": 7.338550819558238e-08, "loss": 0.211, "step": 11163 }, { "epoch": 2.970729111229377, "grad_norm": 0.30615004897117615, "learning_rate": 7.336906302571072e-08, "loss": 0.2056, "step": 11164 }, { "epoch": 2.9709952102182013, "grad_norm": 0.3077731132507324, "learning_rate": 7.33526186309965e-08, "loss": 0.1941, "step": 11165 }, { "epoch": 2.971261309207025, "grad_norm": 0.2848653793334961, "learning_rate": 7.333617501191836e-08, "loss": 0.1862, "step": 11166 }, { "epoch": 2.9715274081958487, "grad_norm": 0.3091042935848236, "learning_rate": 7.331973216895493e-08, "loss": 0.206, "step": 11167 }, { "epoch": 2.971793507184673, "grad_norm": 0.26561403274536133, "learning_rate": 7.330329010258482e-08, "loss": 0.1773, "step": 11168 }, { "epoch": 2.9720596061734965, "grad_norm": 0.289388507604599, "learning_rate": 7.328684881328662e-08, "loss": 0.1819, "step": 11169 }, { "epoch": 2.97232570516232, "grad_norm": 0.28992390632629395, "learning_rate": 7.327040830153894e-08, "loss": 0.192, "step": 11170 }, { "epoch": 2.9725918041511443, "grad_norm": 0.2884923219680786, "learning_rate": 7.325396856782023e-08, "loss": 0.1899, "step": 11171 }, { "epoch": 2.972857903139968, "grad_norm": 0.3305363953113556, "learning_rate": 7.323752961260903e-08, "loss": 0.193, "step": 11172 }, { "epoch": 2.9731240021287917, "grad_norm": 0.31782934069633484, "learning_rate": 7.322109143638386e-08, "loss": 0.1944, "step": 11173 }, { "epoch": 2.973390101117616, "grad_norm": 0.2859014570713043, "learning_rate": 7.32046540396232e-08, "loss": 0.1883, "step": 11174 }, { "epoch": 2.9736562001064395, "grad_norm": 0.2964705526828766, "learning_rate": 7.31882174228055e-08, "loss": 0.2113, "step": 11175 }, { "epoch": 2.973922299095263, "grad_norm": 0.3393528461456299, "learning_rate": 7.317178158640916e-08, "loss": 0.1911, "step": 11176 }, { "epoch": 2.9741883980840873, "grad_norm": 0.2976015508174896, "learning_rate": 7.315534653091258e-08, "loss": 0.2043, "step": 11177 }, { "epoch": 2.974454497072911, "grad_norm": 0.28015047311782837, "learning_rate": 7.313891225679419e-08, "loss": 0.169, "step": 11178 }, { "epoch": 2.974720596061735, "grad_norm": 0.28682175278663635, "learning_rate": 7.312247876453235e-08, "loss": 0.1926, "step": 11179 }, { "epoch": 2.974986695050559, "grad_norm": 0.3009413480758667, "learning_rate": 7.31060460546053e-08, "loss": 0.188, "step": 11180 }, { "epoch": 2.975252794039383, "grad_norm": 0.30647537112236023, "learning_rate": 7.308961412749146e-08, "loss": 0.1988, "step": 11181 }, { "epoch": 2.9755188930282066, "grad_norm": 0.30040112137794495, "learning_rate": 7.307318298366906e-08, "loss": 0.205, "step": 11182 }, { "epoch": 2.9757849920170303, "grad_norm": 0.3159937262535095, "learning_rate": 7.305675262361639e-08, "loss": 0.216, "step": 11183 }, { "epoch": 2.9760510910058544, "grad_norm": 0.35128891468048096, "learning_rate": 7.304032304781169e-08, "loss": 0.2199, "step": 11184 }, { "epoch": 2.976317189994678, "grad_norm": 0.30831319093704224, "learning_rate": 7.30238942567332e-08, "loss": 0.2184, "step": 11185 }, { "epoch": 2.9765832889835018, "grad_norm": 0.28662481904029846, "learning_rate": 7.30074662508591e-08, "loss": 0.1885, "step": 11186 }, { "epoch": 2.976849387972326, "grad_norm": 0.3072698712348938, "learning_rate": 7.299103903066757e-08, "loss": 0.1911, "step": 11187 }, { "epoch": 2.9771154869611496, "grad_norm": 0.3003476858139038, "learning_rate": 7.297461259663682e-08, "loss": 0.1961, "step": 11188 }, { "epoch": 2.9773815859499733, "grad_norm": 0.2976401448249817, "learning_rate": 7.295818694924487e-08, "loss": 0.191, "step": 11189 }, { "epoch": 2.9776476849387974, "grad_norm": 0.3281787931919098, "learning_rate": 7.294176208896987e-08, "loss": 0.1891, "step": 11190 }, { "epoch": 2.977913783927621, "grad_norm": 0.29010099172592163, "learning_rate": 7.292533801628995e-08, "loss": 0.1932, "step": 11191 }, { "epoch": 2.9781798829164448, "grad_norm": 0.27383720874786377, "learning_rate": 7.290891473168313e-08, "loss": 0.1979, "step": 11192 }, { "epoch": 2.978445981905269, "grad_norm": 0.2980579137802124, "learning_rate": 7.289249223562744e-08, "loss": 0.1962, "step": 11193 }, { "epoch": 2.9787120808940926, "grad_norm": 0.2987568974494934, "learning_rate": 7.287607052860097e-08, "loss": 0.1922, "step": 11194 }, { "epoch": 2.9789781798829162, "grad_norm": 0.3087483048439026, "learning_rate": 7.285964961108163e-08, "loss": 0.1945, "step": 11195 }, { "epoch": 2.9792442788717404, "grad_norm": 0.28728538751602173, "learning_rate": 7.28432294835474e-08, "loss": 0.1942, "step": 11196 }, { "epoch": 2.979510377860564, "grad_norm": 0.29457539319992065, "learning_rate": 7.282681014647632e-08, "loss": 0.197, "step": 11197 }, { "epoch": 2.9797764768493877, "grad_norm": 0.2930183708667755, "learning_rate": 7.281039160034617e-08, "loss": 0.1815, "step": 11198 }, { "epoch": 2.980042575838212, "grad_norm": 0.30377858877182007, "learning_rate": 7.279397384563495e-08, "loss": 0.1989, "step": 11199 }, { "epoch": 2.9803086748270355, "grad_norm": 0.2863155007362366, "learning_rate": 7.27775568828205e-08, "loss": 0.1883, "step": 11200 }, { "epoch": 2.9805747738158592, "grad_norm": 0.30175715684890747, "learning_rate": 7.276114071238069e-08, "loss": 0.2131, "step": 11201 }, { "epoch": 2.9808408728046834, "grad_norm": 0.321059912443161, "learning_rate": 7.274472533479333e-08, "loss": 0.2051, "step": 11202 }, { "epoch": 2.981106971793507, "grad_norm": 0.33002781867980957, "learning_rate": 7.272831075053627e-08, "loss": 0.2055, "step": 11203 }, { "epoch": 2.981373070782331, "grad_norm": 0.29715004563331604, "learning_rate": 7.271189696008728e-08, "loss": 0.1784, "step": 11204 }, { "epoch": 2.981639169771155, "grad_norm": 0.2766312062740326, "learning_rate": 7.269548396392411e-08, "loss": 0.1792, "step": 11205 }, { "epoch": 2.981905268759979, "grad_norm": 0.5695600509643555, "learning_rate": 7.267907176252453e-08, "loss": 0.1823, "step": 11206 }, { "epoch": 2.9821713677488026, "grad_norm": 0.27604320645332336, "learning_rate": 7.266266035636621e-08, "loss": 0.1864, "step": 11207 }, { "epoch": 2.9824374667376263, "grad_norm": 0.29053258895874023, "learning_rate": 7.264624974592686e-08, "loss": 0.1806, "step": 11208 }, { "epoch": 2.9827035657264505, "grad_norm": 0.27814674377441406, "learning_rate": 7.262983993168417e-08, "loss": 0.1828, "step": 11209 }, { "epoch": 2.982969664715274, "grad_norm": 0.31088539958000183, "learning_rate": 7.261343091411576e-08, "loss": 0.2153, "step": 11210 }, { "epoch": 2.983235763704098, "grad_norm": 0.2882997989654541, "learning_rate": 7.259702269369926e-08, "loss": 0.2069, "step": 11211 }, { "epoch": 2.983501862692922, "grad_norm": 0.2866484224796295, "learning_rate": 7.258061527091228e-08, "loss": 0.2046, "step": 11212 }, { "epoch": 2.9837679616817456, "grad_norm": 0.2694719433784485, "learning_rate": 7.25642086462324e-08, "loss": 0.1832, "step": 11213 }, { "epoch": 2.9840340606705693, "grad_norm": 0.32200393080711365, "learning_rate": 7.254780282013721e-08, "loss": 0.1966, "step": 11214 }, { "epoch": 2.9843001596593934, "grad_norm": 0.29514211416244507, "learning_rate": 7.253139779310417e-08, "loss": 0.1928, "step": 11215 }, { "epoch": 2.984566258648217, "grad_norm": 0.30175861716270447, "learning_rate": 7.251499356561082e-08, "loss": 0.185, "step": 11216 }, { "epoch": 2.984832357637041, "grad_norm": 0.297961950302124, "learning_rate": 7.249859013813465e-08, "loss": 0.1972, "step": 11217 }, { "epoch": 2.985098456625865, "grad_norm": 0.2659684717655182, "learning_rate": 7.24821875111531e-08, "loss": 0.1871, "step": 11218 }, { "epoch": 2.9853645556146886, "grad_norm": 0.2880361080169678, "learning_rate": 7.246578568514363e-08, "loss": 0.1853, "step": 11219 }, { "epoch": 2.9856306546035123, "grad_norm": 0.31502994894981384, "learning_rate": 7.244938466058364e-08, "loss": 0.1976, "step": 11220 }, { "epoch": 2.9858967535923364, "grad_norm": 0.2771323621273041, "learning_rate": 7.243298443795052e-08, "loss": 0.1878, "step": 11221 }, { "epoch": 2.98616285258116, "grad_norm": 0.311691015958786, "learning_rate": 7.241658501772166e-08, "loss": 0.1896, "step": 11222 }, { "epoch": 2.986428951569984, "grad_norm": 0.3227006793022156, "learning_rate": 7.240018640037437e-08, "loss": 0.2051, "step": 11223 }, { "epoch": 2.986695050558808, "grad_norm": 0.2937394380569458, "learning_rate": 7.238378858638605e-08, "loss": 0.2016, "step": 11224 }, { "epoch": 2.9869611495476316, "grad_norm": 0.27238309383392334, "learning_rate": 7.23673915762339e-08, "loss": 0.1969, "step": 11225 }, { "epoch": 2.9872272485364557, "grad_norm": 0.30420705676078796, "learning_rate": 7.235099537039522e-08, "loss": 0.1837, "step": 11226 }, { "epoch": 2.9874933475252794, "grad_norm": 0.29168474674224854, "learning_rate": 7.23345999693473e-08, "loss": 0.1858, "step": 11227 }, { "epoch": 2.9877594465141035, "grad_norm": 0.2886228859424591, "learning_rate": 7.231820537356733e-08, "loss": 0.1726, "step": 11228 }, { "epoch": 2.988025545502927, "grad_norm": 0.9374451637268066, "learning_rate": 7.230181158353252e-08, "loss": 0.1975, "step": 11229 }, { "epoch": 2.988291644491751, "grad_norm": 0.2921004891395569, "learning_rate": 7.228541859972006e-08, "loss": 0.2006, "step": 11230 }, { "epoch": 2.988557743480575, "grad_norm": 0.3008808493614197, "learning_rate": 7.22690264226071e-08, "loss": 0.1972, "step": 11231 }, { "epoch": 2.9888238424693987, "grad_norm": 0.30036604404449463, "learning_rate": 7.225263505267079e-08, "loss": 0.196, "step": 11232 }, { "epoch": 2.9890899414582224, "grad_norm": 0.29754164814949036, "learning_rate": 7.223624449038825e-08, "loss": 0.2042, "step": 11233 }, { "epoch": 2.9893560404470465, "grad_norm": 0.2743931710720062, "learning_rate": 7.221985473623653e-08, "loss": 0.1777, "step": 11234 }, { "epoch": 2.98962213943587, "grad_norm": 0.29661303758621216, "learning_rate": 7.220346579069269e-08, "loss": 0.1925, "step": 11235 }, { "epoch": 2.989888238424694, "grad_norm": 0.2815399467945099, "learning_rate": 7.218707765423378e-08, "loss": 0.1858, "step": 11236 }, { "epoch": 2.990154337413518, "grad_norm": 0.2780326306819916, "learning_rate": 7.217069032733685e-08, "loss": 0.1897, "step": 11237 }, { "epoch": 2.9904204364023417, "grad_norm": 0.2995914816856384, "learning_rate": 7.215430381047887e-08, "loss": 0.1993, "step": 11238 }, { "epoch": 2.9906865353911654, "grad_norm": 0.28401923179626465, "learning_rate": 7.213791810413679e-08, "loss": 0.1941, "step": 11239 }, { "epoch": 2.9909526343799895, "grad_norm": 0.2815532088279724, "learning_rate": 7.212153320878755e-08, "loss": 0.1909, "step": 11240 }, { "epoch": 2.991218733368813, "grad_norm": 0.2932475209236145, "learning_rate": 7.210514912490812e-08, "loss": 0.1905, "step": 11241 }, { "epoch": 2.991484832357637, "grad_norm": 0.3161342442035675, "learning_rate": 7.20887658529754e-08, "loss": 0.1967, "step": 11242 }, { "epoch": 2.991750931346461, "grad_norm": 0.3280084729194641, "learning_rate": 7.207238339346619e-08, "loss": 0.2058, "step": 11243 }, { "epoch": 2.9920170303352847, "grad_norm": 0.2916021943092346, "learning_rate": 7.20560017468574e-08, "loss": 0.1786, "step": 11244 }, { "epoch": 2.9922831293241083, "grad_norm": 0.3200354278087616, "learning_rate": 7.203962091362583e-08, "loss": 0.203, "step": 11245 }, { "epoch": 2.9925492283129325, "grad_norm": 0.3076106011867523, "learning_rate": 7.202324089424833e-08, "loss": 0.2017, "step": 11246 }, { "epoch": 2.992815327301756, "grad_norm": 0.27876073122024536, "learning_rate": 7.200686168920161e-08, "loss": 0.1939, "step": 11247 }, { "epoch": 2.99308142629058, "grad_norm": 0.2618873417377472, "learning_rate": 7.199048329896247e-08, "loss": 0.1767, "step": 11248 }, { "epoch": 2.993347525279404, "grad_norm": 0.28433144092559814, "learning_rate": 7.197410572400764e-08, "loss": 0.1844, "step": 11249 }, { "epoch": 2.9936136242682276, "grad_norm": 0.2784794867038727, "learning_rate": 7.195772896481383e-08, "loss": 0.1847, "step": 11250 }, { "epoch": 2.9938797232570518, "grad_norm": 0.27027127146720886, "learning_rate": 7.19413530218577e-08, "loss": 0.191, "step": 11251 }, { "epoch": 2.9941458222458754, "grad_norm": 0.29507970809936523, "learning_rate": 7.192497789561598e-08, "loss": 0.1898, "step": 11252 }, { "epoch": 2.9944119212346996, "grad_norm": 0.32478436827659607, "learning_rate": 7.190860358656522e-08, "loss": 0.1988, "step": 11253 }, { "epoch": 2.9946780202235233, "grad_norm": 0.2800018787384033, "learning_rate": 7.189223009518207e-08, "loss": 0.1829, "step": 11254 }, { "epoch": 2.994944119212347, "grad_norm": 0.32307225465774536, "learning_rate": 7.187585742194311e-08, "loss": 0.2142, "step": 11255 }, { "epoch": 2.995210218201171, "grad_norm": 0.3214038014411926, "learning_rate": 7.185948556732495e-08, "loss": 0.1894, "step": 11256 }, { "epoch": 2.9954763171899947, "grad_norm": 0.29591816663742065, "learning_rate": 7.184311453180407e-08, "loss": 0.1978, "step": 11257 }, { "epoch": 2.9957424161788184, "grad_norm": 0.30646419525146484, "learning_rate": 7.182674431585702e-08, "loss": 0.1955, "step": 11258 }, { "epoch": 2.9960085151676425, "grad_norm": 0.2905493378639221, "learning_rate": 7.18103749199603e-08, "loss": 0.2059, "step": 11259 }, { "epoch": 2.9962746141564662, "grad_norm": 0.288863867521286, "learning_rate": 7.179400634459036e-08, "loss": 0.1724, "step": 11260 }, { "epoch": 2.99654071314529, "grad_norm": 0.3049732446670532, "learning_rate": 7.177763859022372e-08, "loss": 0.2036, "step": 11261 }, { "epoch": 2.996806812134114, "grad_norm": 0.3037424087524414, "learning_rate": 7.176127165733667e-08, "loss": 0.1931, "step": 11262 }, { "epoch": 2.9970729111229377, "grad_norm": 0.2895677983760834, "learning_rate": 7.17449055464057e-08, "loss": 0.19, "step": 11263 }, { "epoch": 2.9973390101117614, "grad_norm": 0.3263036012649536, "learning_rate": 7.172854025790718e-08, "loss": 0.2104, "step": 11264 }, { "epoch": 2.9976051091005855, "grad_norm": 0.32193565368652344, "learning_rate": 7.171217579231742e-08, "loss": 0.2073, "step": 11265 }, { "epoch": 2.997871208089409, "grad_norm": 0.30485039949417114, "learning_rate": 7.169581215011281e-08, "loss": 0.1925, "step": 11266 }, { "epoch": 2.998137307078233, "grad_norm": 0.31550318002700806, "learning_rate": 7.167944933176959e-08, "loss": 0.2067, "step": 11267 }, { "epoch": 2.998403406067057, "grad_norm": 0.2955566644668579, "learning_rate": 7.166308733776407e-08, "loss": 0.1983, "step": 11268 }, { "epoch": 2.9986695050558807, "grad_norm": 0.3150867223739624, "learning_rate": 7.16467261685725e-08, "loss": 0.197, "step": 11269 }, { "epoch": 2.9989356040447044, "grad_norm": 0.30212706327438354, "learning_rate": 7.163036582467116e-08, "loss": 0.1807, "step": 11270 }, { "epoch": 2.9992017030335285, "grad_norm": 0.2860099673271179, "learning_rate": 7.161400630653616e-08, "loss": 0.1847, "step": 11271 }, { "epoch": 2.999467802022352, "grad_norm": 0.27872639894485474, "learning_rate": 7.159764761464373e-08, "loss": 0.1923, "step": 11272 }, { "epoch": 2.9997339010111763, "grad_norm": 0.287439227104187, "learning_rate": 7.158128974947004e-08, "loss": 0.1934, "step": 11273 }, { "epoch": 3.0, "grad_norm": 0.2711164355278015, "learning_rate": 7.156493271149122e-08, "loss": 0.1825, "step": 11274 }, { "epoch": 3.0002660989888237, "grad_norm": 0.2802971601486206, "learning_rate": 7.154857650118337e-08, "loss": 0.1776, "step": 11275 }, { "epoch": 3.000532197977648, "grad_norm": 0.29268983006477356, "learning_rate": 7.153222111902262e-08, "loss": 0.1897, "step": 11276 }, { "epoch": 3.0007982969664715, "grad_norm": 0.2985761761665344, "learning_rate": 7.151586656548496e-08, "loss": 0.2027, "step": 11277 }, { "epoch": 3.001064395955295, "grad_norm": 0.2905530631542206, "learning_rate": 7.149951284104648e-08, "loss": 0.2039, "step": 11278 }, { "epoch": 3.0013304949441193, "grad_norm": 0.2744193971157074, "learning_rate": 7.14831599461832e-08, "loss": 0.175, "step": 11279 }, { "epoch": 3.001596593932943, "grad_norm": 0.29931363463401794, "learning_rate": 7.146680788137106e-08, "loss": 0.1935, "step": 11280 }, { "epoch": 3.001862692921767, "grad_norm": 0.2766936719417572, "learning_rate": 7.145045664708606e-08, "loss": 0.1865, "step": 11281 }, { "epoch": 3.002128791910591, "grad_norm": 0.29122477769851685, "learning_rate": 7.143410624380414e-08, "loss": 0.1916, "step": 11282 }, { "epoch": 3.0023948908994145, "grad_norm": 0.28185325860977173, "learning_rate": 7.141775667200121e-08, "loss": 0.1807, "step": 11283 }, { "epoch": 3.0026609898882386, "grad_norm": 0.2931601405143738, "learning_rate": 7.140140793215318e-08, "loss": 0.2047, "step": 11284 }, { "epoch": 3.0029270888770623, "grad_norm": 0.2983922064304352, "learning_rate": 7.13850600247359e-08, "loss": 0.199, "step": 11285 }, { "epoch": 3.003193187865886, "grad_norm": 0.3027064800262451, "learning_rate": 7.136871295022525e-08, "loss": 0.1935, "step": 11286 }, { "epoch": 3.00345928685471, "grad_norm": 0.3013591468334198, "learning_rate": 7.135236670909701e-08, "loss": 0.1708, "step": 11287 }, { "epoch": 3.0037253858435338, "grad_norm": 0.3268778622150421, "learning_rate": 7.133602130182703e-08, "loss": 0.2101, "step": 11288 }, { "epoch": 3.0039914848323575, "grad_norm": 0.2922589182853699, "learning_rate": 7.1319676728891e-08, "loss": 0.1956, "step": 11289 }, { "epoch": 3.0042575838211816, "grad_norm": 0.27986031770706177, "learning_rate": 7.130333299076471e-08, "loss": 0.1784, "step": 11290 }, { "epoch": 3.0045236828100053, "grad_norm": 0.26949137449264526, "learning_rate": 7.12869900879239e-08, "loss": 0.1702, "step": 11291 }, { "epoch": 3.0047897817988294, "grad_norm": 0.33144432306289673, "learning_rate": 7.127064802084423e-08, "loss": 0.1767, "step": 11292 }, { "epoch": 3.005055880787653, "grad_norm": 0.2711503803730011, "learning_rate": 7.125430679000142e-08, "loss": 0.1895, "step": 11293 }, { "epoch": 3.0053219797764767, "grad_norm": 0.27925875782966614, "learning_rate": 7.123796639587109e-08, "loss": 0.1761, "step": 11294 }, { "epoch": 3.005588078765301, "grad_norm": 0.2988312542438507, "learning_rate": 7.122162683892891e-08, "loss": 0.1861, "step": 11295 }, { "epoch": 3.0058541777541246, "grad_norm": 0.3012920618057251, "learning_rate": 7.120528811965044e-08, "loss": 0.2007, "step": 11296 }, { "epoch": 3.0061202767429482, "grad_norm": 0.29904651641845703, "learning_rate": 7.118895023851128e-08, "loss": 0.2012, "step": 11297 }, { "epoch": 3.0063863757317724, "grad_norm": 0.34422236680984497, "learning_rate": 7.117261319598697e-08, "loss": 0.2152, "step": 11298 }, { "epoch": 3.006652474720596, "grad_norm": 0.29538261890411377, "learning_rate": 7.115627699255302e-08, "loss": 0.1941, "step": 11299 }, { "epoch": 3.0069185737094197, "grad_norm": 0.2951267659664154, "learning_rate": 7.113994162868495e-08, "loss": 0.185, "step": 11300 }, { "epoch": 3.007184672698244, "grad_norm": 0.2975028455257416, "learning_rate": 7.112360710485826e-08, "loss": 0.1958, "step": 11301 }, { "epoch": 3.0074507716870675, "grad_norm": 0.265762060880661, "learning_rate": 7.110727342154837e-08, "loss": 0.1754, "step": 11302 }, { "epoch": 3.007716870675891, "grad_norm": 0.2702616751194, "learning_rate": 7.109094057923073e-08, "loss": 0.1815, "step": 11303 }, { "epoch": 3.0079829696647153, "grad_norm": 0.29403090476989746, "learning_rate": 7.107460857838076e-08, "loss": 0.1883, "step": 11304 }, { "epoch": 3.008249068653539, "grad_norm": 0.2705240845680237, "learning_rate": 7.105827741947385e-08, "loss": 0.183, "step": 11305 }, { "epoch": 3.008515167642363, "grad_norm": 0.30867117643356323, "learning_rate": 7.104194710298535e-08, "loss": 0.2054, "step": 11306 }, { "epoch": 3.008781266631187, "grad_norm": 0.28376975655555725, "learning_rate": 7.102561762939054e-08, "loss": 0.2105, "step": 11307 }, { "epoch": 3.0090473656200105, "grad_norm": 0.28578296303749084, "learning_rate": 7.100928899916479e-08, "loss": 0.1919, "step": 11308 }, { "epoch": 3.0093134646088346, "grad_norm": 0.2983150780200958, "learning_rate": 7.099296121278336e-08, "loss": 0.1877, "step": 11309 }, { "epoch": 3.0095795635976583, "grad_norm": 0.2851250171661377, "learning_rate": 7.09766342707215e-08, "loss": 0.1892, "step": 11310 }, { "epoch": 3.009845662586482, "grad_norm": 0.31195685267448425, "learning_rate": 7.096030817345445e-08, "loss": 0.2082, "step": 11311 }, { "epoch": 3.010111761575306, "grad_norm": 0.3233737051486969, "learning_rate": 7.094398292145745e-08, "loss": 0.2173, "step": 11312 }, { "epoch": 3.01037786056413, "grad_norm": 0.303960382938385, "learning_rate": 7.092765851520565e-08, "loss": 0.2006, "step": 11313 }, { "epoch": 3.0106439595529535, "grad_norm": 0.40842071175575256, "learning_rate": 7.091133495517423e-08, "loss": 0.1925, "step": 11314 }, { "epoch": 3.0109100585417776, "grad_norm": 0.2976071536540985, "learning_rate": 7.089501224183835e-08, "loss": 0.1981, "step": 11315 }, { "epoch": 3.0111761575306013, "grad_norm": 0.2784539759159088, "learning_rate": 7.087869037567307e-08, "loss": 0.183, "step": 11316 }, { "epoch": 3.0114422565194254, "grad_norm": 0.3004576861858368, "learning_rate": 7.08623693571535e-08, "loss": 0.2065, "step": 11317 }, { "epoch": 3.011708355508249, "grad_norm": 0.2535969913005829, "learning_rate": 7.084604918675471e-08, "loss": 0.1805, "step": 11318 }, { "epoch": 3.011974454497073, "grad_norm": 0.30555373430252075, "learning_rate": 7.082972986495172e-08, "loss": 0.1968, "step": 11319 }, { "epoch": 3.012240553485897, "grad_norm": 0.29854336380958557, "learning_rate": 7.081341139221954e-08, "loss": 0.198, "step": 11320 }, { "epoch": 3.0125066524747206, "grad_norm": 0.2779742479324341, "learning_rate": 7.07970937690332e-08, "loss": 0.1811, "step": 11321 }, { "epoch": 3.0127727514635443, "grad_norm": 0.28260597586631775, "learning_rate": 7.078077699586762e-08, "loss": 0.1846, "step": 11322 }, { "epoch": 3.0130388504523684, "grad_norm": 0.29025381803512573, "learning_rate": 7.076446107319774e-08, "loss": 0.183, "step": 11323 }, { "epoch": 3.013304949441192, "grad_norm": 0.29772117733955383, "learning_rate": 7.074814600149854e-08, "loss": 0.2003, "step": 11324 }, { "epoch": 3.0135710484300158, "grad_norm": 0.31187334656715393, "learning_rate": 7.073183178124483e-08, "loss": 0.2009, "step": 11325 }, { "epoch": 3.01383714741884, "grad_norm": 0.3056219816207886, "learning_rate": 7.07155184129115e-08, "loss": 0.2085, "step": 11326 }, { "epoch": 3.0141032464076636, "grad_norm": 0.2865046560764313, "learning_rate": 7.069920589697338e-08, "loss": 0.1849, "step": 11327 }, { "epoch": 3.0143693453964877, "grad_norm": 0.2947145104408264, "learning_rate": 7.068289423390533e-08, "loss": 0.1883, "step": 11328 }, { "epoch": 3.0146354443853114, "grad_norm": 0.3083798587322235, "learning_rate": 7.066658342418209e-08, "loss": 0.2032, "step": 11329 }, { "epoch": 3.014901543374135, "grad_norm": 0.288520485162735, "learning_rate": 7.065027346827842e-08, "loss": 0.1838, "step": 11330 }, { "epoch": 3.015167642362959, "grad_norm": 0.32287833094596863, "learning_rate": 7.063396436666911e-08, "loss": 0.2092, "step": 11331 }, { "epoch": 3.015433741351783, "grad_norm": 0.3147755265235901, "learning_rate": 7.061765611982882e-08, "loss": 0.1966, "step": 11332 }, { "epoch": 3.0156998403406066, "grad_norm": 0.3162626028060913, "learning_rate": 7.060134872823233e-08, "loss": 0.1938, "step": 11333 }, { "epoch": 3.0159659393294307, "grad_norm": 0.3088216781616211, "learning_rate": 7.058504219235422e-08, "loss": 0.2139, "step": 11334 }, { "epoch": 3.0162320383182544, "grad_norm": 0.2986215353012085, "learning_rate": 7.056873651266912e-08, "loss": 0.1954, "step": 11335 }, { "epoch": 3.016498137307078, "grad_norm": 0.31764596700668335, "learning_rate": 7.055243168965171e-08, "loss": 0.197, "step": 11336 }, { "epoch": 3.016764236295902, "grad_norm": 0.30826449394226074, "learning_rate": 7.053612772377658e-08, "loss": 0.2033, "step": 11337 }, { "epoch": 3.017030335284726, "grad_norm": 0.31376174092292786, "learning_rate": 7.051982461551822e-08, "loss": 0.2009, "step": 11338 }, { "epoch": 3.0172964342735495, "grad_norm": 0.28081122040748596, "learning_rate": 7.050352236535124e-08, "loss": 0.1839, "step": 11339 }, { "epoch": 3.0175625332623737, "grad_norm": 0.2694494128227234, "learning_rate": 7.048722097375013e-08, "loss": 0.1892, "step": 11340 }, { "epoch": 3.0178286322511974, "grad_norm": 0.3124610483646393, "learning_rate": 7.04709204411894e-08, "loss": 0.2017, "step": 11341 }, { "epoch": 3.0180947312400215, "grad_norm": 0.30144646763801575, "learning_rate": 7.045462076814353e-08, "loss": 0.1978, "step": 11342 }, { "epoch": 3.018360830228845, "grad_norm": 0.29781317710876465, "learning_rate": 7.04383219550869e-08, "loss": 0.188, "step": 11343 }, { "epoch": 3.018626929217669, "grad_norm": 0.335200697183609, "learning_rate": 7.042202400249397e-08, "loss": 0.1918, "step": 11344 }, { "epoch": 3.018893028206493, "grad_norm": 0.3275735676288605, "learning_rate": 7.040572691083912e-08, "loss": 0.2159, "step": 11345 }, { "epoch": 3.0191591271953166, "grad_norm": 0.2932007908821106, "learning_rate": 7.038943068059671e-08, "loss": 0.1837, "step": 11346 }, { "epoch": 3.0194252261841403, "grad_norm": 0.2802261710166931, "learning_rate": 7.037313531224111e-08, "loss": 0.175, "step": 11347 }, { "epoch": 3.0196913251729645, "grad_norm": 0.2999740540981293, "learning_rate": 7.03568408062466e-08, "loss": 0.1839, "step": 11348 }, { "epoch": 3.019957424161788, "grad_norm": 0.3045194149017334, "learning_rate": 7.034054716308749e-08, "loss": 0.1832, "step": 11349 }, { "epoch": 3.020223523150612, "grad_norm": 0.3202815651893616, "learning_rate": 7.032425438323804e-08, "loss": 0.2023, "step": 11350 }, { "epoch": 3.020489622139436, "grad_norm": 0.31298840045928955, "learning_rate": 7.030796246717254e-08, "loss": 0.2101, "step": 11351 }, { "epoch": 3.0207557211282596, "grad_norm": 0.29132238030433655, "learning_rate": 7.029167141536511e-08, "loss": 0.1841, "step": 11352 }, { "epoch": 3.0210218201170838, "grad_norm": 0.2779196500778198, "learning_rate": 7.027538122828998e-08, "loss": 0.1739, "step": 11353 }, { "epoch": 3.0212879191059074, "grad_norm": 0.29340505599975586, "learning_rate": 7.025909190642135e-08, "loss": 0.1921, "step": 11354 }, { "epoch": 3.021554018094731, "grad_norm": 0.2888182997703552, "learning_rate": 7.02428034502333e-08, "loss": 0.1995, "step": 11355 }, { "epoch": 3.0218201170835552, "grad_norm": 0.3584233224391937, "learning_rate": 7.022651586019998e-08, "loss": 0.2044, "step": 11356 }, { "epoch": 3.022086216072379, "grad_norm": 0.31318867206573486, "learning_rate": 7.021022913679553e-08, "loss": 0.2013, "step": 11357 }, { "epoch": 3.0223523150612026, "grad_norm": 0.2999723553657532, "learning_rate": 7.019394328049392e-08, "loss": 0.1849, "step": 11358 }, { "epoch": 3.0226184140500267, "grad_norm": 0.3098931312561035, "learning_rate": 7.017765829176923e-08, "loss": 0.1973, "step": 11359 }, { "epoch": 3.0228845130388504, "grad_norm": 0.2708367705345154, "learning_rate": 7.016137417109552e-08, "loss": 0.1718, "step": 11360 }, { "epoch": 3.023150612027674, "grad_norm": 0.280405193567276, "learning_rate": 7.014509091894669e-08, "loss": 0.1813, "step": 11361 }, { "epoch": 3.0234167110164982, "grad_norm": 0.29834482073783875, "learning_rate": 7.012880853579674e-08, "loss": 0.1753, "step": 11362 }, { "epoch": 3.023682810005322, "grad_norm": 0.30546334385871887, "learning_rate": 7.011252702211962e-08, "loss": 0.2019, "step": 11363 }, { "epoch": 3.023948908994146, "grad_norm": 0.3070395886898041, "learning_rate": 7.009624637838922e-08, "loss": 0.2063, "step": 11364 }, { "epoch": 3.0242150079829697, "grad_norm": 0.28705132007598877, "learning_rate": 7.007996660507947e-08, "loss": 0.1894, "step": 11365 }, { "epoch": 3.0244811069717934, "grad_norm": 0.313741534948349, "learning_rate": 7.00636877026642e-08, "loss": 0.2173, "step": 11366 }, { "epoch": 3.0247472059606175, "grad_norm": 0.2913731336593628, "learning_rate": 7.004740967161726e-08, "loss": 0.2175, "step": 11367 }, { "epoch": 3.025013304949441, "grad_norm": 0.28767409920692444, "learning_rate": 7.003113251241244e-08, "loss": 0.1772, "step": 11368 }, { "epoch": 3.025279403938265, "grad_norm": 0.30539214611053467, "learning_rate": 7.001485622552356e-08, "loss": 0.1784, "step": 11369 }, { "epoch": 3.025545502927089, "grad_norm": 0.268396258354187, "learning_rate": 6.999858081142436e-08, "loss": 0.178, "step": 11370 }, { "epoch": 3.0258116019159127, "grad_norm": 0.2921828329563141, "learning_rate": 6.998230627058855e-08, "loss": 0.1969, "step": 11371 }, { "epoch": 3.0260777009047364, "grad_norm": 0.2747023105621338, "learning_rate": 6.996603260348987e-08, "loss": 0.1817, "step": 11372 }, { "epoch": 3.0263437998935605, "grad_norm": 0.2849777936935425, "learning_rate": 6.994975981060198e-08, "loss": 0.1971, "step": 11373 }, { "epoch": 3.026609898882384, "grad_norm": 0.28812915086746216, "learning_rate": 6.993348789239858e-08, "loss": 0.1945, "step": 11374 }, { "epoch": 3.0268759978712083, "grad_norm": 0.287679523229599, "learning_rate": 6.991721684935328e-08, "loss": 0.1922, "step": 11375 }, { "epoch": 3.027142096860032, "grad_norm": 0.2974041700363159, "learning_rate": 6.990094668193968e-08, "loss": 0.1938, "step": 11376 }, { "epoch": 3.0274081958488557, "grad_norm": 0.3349026143550873, "learning_rate": 6.98846773906314e-08, "loss": 0.1993, "step": 11377 }, { "epoch": 3.02767429483768, "grad_norm": 0.2515183091163635, "learning_rate": 6.986840897590195e-08, "loss": 0.1742, "step": 11378 }, { "epoch": 3.0279403938265035, "grad_norm": 0.30018773674964905, "learning_rate": 6.985214143822486e-08, "loss": 0.1795, "step": 11379 }, { "epoch": 3.028206492815327, "grad_norm": 0.3043561577796936, "learning_rate": 6.983587477807369e-08, "loss": 0.191, "step": 11380 }, { "epoch": 3.0284725918041513, "grad_norm": 0.29445600509643555, "learning_rate": 6.981960899592185e-08, "loss": 0.185, "step": 11381 }, { "epoch": 3.028738690792975, "grad_norm": 0.2855032682418823, "learning_rate": 6.980334409224284e-08, "loss": 0.1846, "step": 11382 }, { "epoch": 3.0290047897817987, "grad_norm": 0.2781519591808319, "learning_rate": 6.978708006751009e-08, "loss": 0.1796, "step": 11383 }, { "epoch": 3.029270888770623, "grad_norm": 0.3218422532081604, "learning_rate": 6.977081692219697e-08, "loss": 0.1959, "step": 11384 }, { "epoch": 3.0295369877594465, "grad_norm": 0.2967997193336487, "learning_rate": 6.975455465677688e-08, "loss": 0.1897, "step": 11385 }, { "epoch": 3.02980308674827, "grad_norm": 0.277523934841156, "learning_rate": 6.973829327172321e-08, "loss": 0.1906, "step": 11386 }, { "epoch": 3.0300691857370943, "grad_norm": 0.31335756182670593, "learning_rate": 6.972203276750924e-08, "loss": 0.1855, "step": 11387 }, { "epoch": 3.030335284725918, "grad_norm": 0.27250149846076965, "learning_rate": 6.970577314460826e-08, "loss": 0.1723, "step": 11388 }, { "epoch": 3.030601383714742, "grad_norm": 0.3129645586013794, "learning_rate": 6.968951440349359e-08, "loss": 0.1954, "step": 11389 }, { "epoch": 3.0308674827035658, "grad_norm": 0.2822272479534149, "learning_rate": 6.967325654463842e-08, "loss": 0.1836, "step": 11390 }, { "epoch": 3.0311335816923894, "grad_norm": 0.2764626145362854, "learning_rate": 6.965699956851603e-08, "loss": 0.1704, "step": 11391 }, { "epoch": 3.0313996806812136, "grad_norm": 0.30746862292289734, "learning_rate": 6.964074347559958e-08, "loss": 0.199, "step": 11392 }, { "epoch": 3.0316657796700373, "grad_norm": 0.2792537808418274, "learning_rate": 6.962448826636227e-08, "loss": 0.1782, "step": 11393 }, { "epoch": 3.031931878658861, "grad_norm": 0.30792123079299927, "learning_rate": 6.960823394127724e-08, "loss": 0.1841, "step": 11394 }, { "epoch": 3.032197977647685, "grad_norm": 0.29978325963020325, "learning_rate": 6.95919805008176e-08, "loss": 0.1827, "step": 11395 }, { "epoch": 3.0324640766365087, "grad_norm": 0.275468111038208, "learning_rate": 6.957572794545651e-08, "loss": 0.1692, "step": 11396 }, { "epoch": 3.0327301756253324, "grad_norm": 0.3034757077693939, "learning_rate": 6.955947627566693e-08, "loss": 0.1909, "step": 11397 }, { "epoch": 3.0329962746141566, "grad_norm": 0.31233125925064087, "learning_rate": 6.954322549192195e-08, "loss": 0.2035, "step": 11398 }, { "epoch": 3.0332623736029802, "grad_norm": 0.7246724963188171, "learning_rate": 6.952697559469463e-08, "loss": 0.1659, "step": 11399 }, { "epoch": 3.0335284725918044, "grad_norm": 0.298354834318161, "learning_rate": 6.951072658445791e-08, "loss": 0.197, "step": 11400 }, { "epoch": 3.033794571580628, "grad_norm": 0.3028737008571625, "learning_rate": 6.949447846168477e-08, "loss": 0.1994, "step": 11401 }, { "epoch": 3.0340606705694517, "grad_norm": 0.2820373773574829, "learning_rate": 6.947823122684815e-08, "loss": 0.179, "step": 11402 }, { "epoch": 3.034326769558276, "grad_norm": 0.28645652532577515, "learning_rate": 6.946198488042097e-08, "loss": 0.1729, "step": 11403 }, { "epoch": 3.0345928685470995, "grad_norm": 0.32907721400260925, "learning_rate": 6.944573942287612e-08, "loss": 0.1979, "step": 11404 }, { "epoch": 3.034858967535923, "grad_norm": 0.2759127914905548, "learning_rate": 6.942949485468651e-08, "loss": 0.1812, "step": 11405 }, { "epoch": 3.0351250665247473, "grad_norm": 0.3090575635433197, "learning_rate": 6.941325117632489e-08, "loss": 0.2165, "step": 11406 }, { "epoch": 3.035391165513571, "grad_norm": 0.2931280732154846, "learning_rate": 6.93970083882641e-08, "loss": 0.1777, "step": 11407 }, { "epoch": 3.0356572645023947, "grad_norm": 0.2944234311580658, "learning_rate": 6.938076649097693e-08, "loss": 0.1948, "step": 11408 }, { "epoch": 3.035923363491219, "grad_norm": 0.3006397783756256, "learning_rate": 6.936452548493619e-08, "loss": 0.1909, "step": 11409 }, { "epoch": 3.0361894624800425, "grad_norm": 0.26786744594573975, "learning_rate": 6.934828537061452e-08, "loss": 0.1746, "step": 11410 }, { "epoch": 3.036455561468866, "grad_norm": 0.2877083122730255, "learning_rate": 6.93320461484847e-08, "loss": 0.2071, "step": 11411 }, { "epoch": 3.0367216604576903, "grad_norm": 0.2787586748600006, "learning_rate": 6.931580781901938e-08, "loss": 0.1845, "step": 11412 }, { "epoch": 3.036987759446514, "grad_norm": 0.29514870047569275, "learning_rate": 6.929957038269121e-08, "loss": 0.1894, "step": 11413 }, { "epoch": 3.037253858435338, "grad_norm": 0.3062792420387268, "learning_rate": 6.92833338399729e-08, "loss": 0.188, "step": 11414 }, { "epoch": 3.037519957424162, "grad_norm": 0.2934510409832001, "learning_rate": 6.926709819133694e-08, "loss": 0.213, "step": 11415 }, { "epoch": 3.0377860564129855, "grad_norm": 0.26780158281326294, "learning_rate": 6.925086343725595e-08, "loss": 0.1652, "step": 11416 }, { "epoch": 3.0380521554018096, "grad_norm": 0.30470848083496094, "learning_rate": 6.923462957820248e-08, "loss": 0.2042, "step": 11417 }, { "epoch": 3.0383182543906333, "grad_norm": 0.326185405254364, "learning_rate": 6.921839661464906e-08, "loss": 0.2159, "step": 11418 }, { "epoch": 3.038584353379457, "grad_norm": 0.310511976480484, "learning_rate": 6.920216454706822e-08, "loss": 0.1964, "step": 11419 }, { "epoch": 3.038850452368281, "grad_norm": 0.456766277551651, "learning_rate": 6.918593337593238e-08, "loss": 0.2047, "step": 11420 }, { "epoch": 3.039116551357105, "grad_norm": 0.31576719880104065, "learning_rate": 6.9169703101714e-08, "loss": 0.2054, "step": 11421 }, { "epoch": 3.0393826503459285, "grad_norm": 0.2897689938545227, "learning_rate": 6.915347372488553e-08, "loss": 0.1829, "step": 11422 }, { "epoch": 3.0396487493347526, "grad_norm": 0.3170423209667206, "learning_rate": 6.913724524591939e-08, "loss": 0.1969, "step": 11423 }, { "epoch": 3.0399148483235763, "grad_norm": 0.29528552293777466, "learning_rate": 6.912101766528785e-08, "loss": 0.1691, "step": 11424 }, { "epoch": 3.0401809473124004, "grad_norm": 0.34623652696609497, "learning_rate": 6.910479098346331e-08, "loss": 0.1891, "step": 11425 }, { "epoch": 3.040447046301224, "grad_norm": 0.3009033799171448, "learning_rate": 6.908856520091811e-08, "loss": 0.2247, "step": 11426 }, { "epoch": 3.0407131452900478, "grad_norm": 0.26990756392478943, "learning_rate": 6.90723403181245e-08, "loss": 0.1713, "step": 11427 }, { "epoch": 3.040979244278872, "grad_norm": 0.26904863119125366, "learning_rate": 6.905611633555475e-08, "loss": 0.1659, "step": 11428 }, { "epoch": 3.0412453432676956, "grad_norm": 0.2979658544063568, "learning_rate": 6.903989325368114e-08, "loss": 0.1849, "step": 11429 }, { "epoch": 3.0415114422565193, "grad_norm": 0.27537670731544495, "learning_rate": 6.902367107297583e-08, "loss": 0.1806, "step": 11430 }, { "epoch": 3.0417775412453434, "grad_norm": 0.31853657960891724, "learning_rate": 6.900744979391102e-08, "loss": 0.2019, "step": 11431 }, { "epoch": 3.042043640234167, "grad_norm": 0.3358848989009857, "learning_rate": 6.899122941695893e-08, "loss": 0.2118, "step": 11432 }, { "epoch": 3.0423097392229907, "grad_norm": 0.2993570864200592, "learning_rate": 6.897500994259158e-08, "loss": 0.2089, "step": 11433 }, { "epoch": 3.042575838211815, "grad_norm": 0.2806905210018158, "learning_rate": 6.895879137128115e-08, "loss": 0.196, "step": 11434 }, { "epoch": 3.0428419372006386, "grad_norm": 0.31748026609420776, "learning_rate": 6.894257370349969e-08, "loss": 0.2102, "step": 11435 }, { "epoch": 3.0431080361894627, "grad_norm": 0.31351086497306824, "learning_rate": 6.892635693971929e-08, "loss": 0.1969, "step": 11436 }, { "epoch": 3.0433741351782864, "grad_norm": 0.2838079035282135, "learning_rate": 6.891014108041194e-08, "loss": 0.1877, "step": 11437 }, { "epoch": 3.04364023416711, "grad_norm": 0.28534162044525146, "learning_rate": 6.889392612604969e-08, "loss": 0.1845, "step": 11438 }, { "epoch": 3.043906333155934, "grad_norm": 0.3073934316635132, "learning_rate": 6.887771207710446e-08, "loss": 0.1928, "step": 11439 }, { "epoch": 3.044172432144758, "grad_norm": 0.3118996322154999, "learning_rate": 6.886149893404822e-08, "loss": 0.2101, "step": 11440 }, { "epoch": 3.0444385311335815, "grad_norm": 0.2882118225097656, "learning_rate": 6.884528669735295e-08, "loss": 0.186, "step": 11441 }, { "epoch": 3.0447046301224057, "grad_norm": 0.28572195768356323, "learning_rate": 6.882907536749043e-08, "loss": 0.186, "step": 11442 }, { "epoch": 3.0449707291112293, "grad_norm": 0.29877087473869324, "learning_rate": 6.88128649449326e-08, "loss": 0.1899, "step": 11443 }, { "epoch": 3.045236828100053, "grad_norm": 0.3237127959728241, "learning_rate": 6.879665543015129e-08, "loss": 0.2005, "step": 11444 }, { "epoch": 3.045502927088877, "grad_norm": 0.38827887177467346, "learning_rate": 6.878044682361834e-08, "loss": 0.1979, "step": 11445 }, { "epoch": 3.045769026077701, "grad_norm": 0.3099222481250763, "learning_rate": 6.876423912580552e-08, "loss": 0.1784, "step": 11446 }, { "epoch": 3.046035125066525, "grad_norm": 0.30810829997062683, "learning_rate": 6.874803233718458e-08, "loss": 0.2112, "step": 11447 }, { "epoch": 3.0463012240553486, "grad_norm": 0.30710548162460327, "learning_rate": 6.873182645822729e-08, "loss": 0.1888, "step": 11448 }, { "epoch": 3.0465673230441723, "grad_norm": 0.2976829707622528, "learning_rate": 6.871562148940535e-08, "loss": 0.2099, "step": 11449 }, { "epoch": 3.0468334220329965, "grad_norm": 0.316023051738739, "learning_rate": 6.869941743119044e-08, "loss": 0.1779, "step": 11450 }, { "epoch": 3.04709952102182, "grad_norm": 0.2756997048854828, "learning_rate": 6.868321428405421e-08, "loss": 0.1871, "step": 11451 }, { "epoch": 3.047365620010644, "grad_norm": 0.27130717039108276, "learning_rate": 6.86670120484683e-08, "loss": 0.1797, "step": 11452 }, { "epoch": 3.047631718999468, "grad_norm": 0.27292537689208984, "learning_rate": 6.86508107249043e-08, "loss": 0.1837, "step": 11453 }, { "epoch": 3.0478978179882916, "grad_norm": 0.3019466996192932, "learning_rate": 6.863461031383381e-08, "loss": 0.1816, "step": 11454 }, { "epoch": 3.0481639169771153, "grad_norm": 0.3031793534755707, "learning_rate": 6.861841081572837e-08, "loss": 0.1883, "step": 11455 }, { "epoch": 3.0484300159659394, "grad_norm": 0.29830533266067505, "learning_rate": 6.860221223105952e-08, "loss": 0.1932, "step": 11456 }, { "epoch": 3.048696114954763, "grad_norm": 0.287760853767395, "learning_rate": 6.858601456029873e-08, "loss": 0.2044, "step": 11457 }, { "epoch": 3.048962213943587, "grad_norm": 0.2963608205318451, "learning_rate": 6.856981780391752e-08, "loss": 0.194, "step": 11458 }, { "epoch": 3.049228312932411, "grad_norm": 0.29341942071914673, "learning_rate": 6.855362196238731e-08, "loss": 0.193, "step": 11459 }, { "epoch": 3.0494944119212346, "grad_norm": 0.28427237272262573, "learning_rate": 6.853742703617949e-08, "loss": 0.1933, "step": 11460 }, { "epoch": 3.0497605109100587, "grad_norm": 0.31500479578971863, "learning_rate": 6.852123302576552e-08, "loss": 0.1881, "step": 11461 }, { "epoch": 3.0500266098988824, "grad_norm": 0.36226943135261536, "learning_rate": 6.850503993161666e-08, "loss": 0.2192, "step": 11462 }, { "epoch": 3.050292708887706, "grad_norm": 0.2792016267776489, "learning_rate": 6.848884775420434e-08, "loss": 0.174, "step": 11463 }, { "epoch": 3.05055880787653, "grad_norm": 0.3099125921726227, "learning_rate": 6.847265649399985e-08, "loss": 0.208, "step": 11464 }, { "epoch": 3.050824906865354, "grad_norm": 0.3460372984409332, "learning_rate": 6.845646615147445e-08, "loss": 0.2057, "step": 11465 }, { "epoch": 3.0510910058541776, "grad_norm": 0.3142945468425751, "learning_rate": 6.844027672709943e-08, "loss": 0.202, "step": 11466 }, { "epoch": 3.0513571048430017, "grad_norm": 0.3053029775619507, "learning_rate": 6.8424088221346e-08, "loss": 0.1994, "step": 11467 }, { "epoch": 3.0516232038318254, "grad_norm": 0.30114027857780457, "learning_rate": 6.840790063468542e-08, "loss": 0.2076, "step": 11468 }, { "epoch": 3.051889302820649, "grad_norm": 0.27352142333984375, "learning_rate": 6.839171396758878e-08, "loss": 0.185, "step": 11469 }, { "epoch": 3.052155401809473, "grad_norm": 0.29898229241371155, "learning_rate": 6.837552822052729e-08, "loss": 0.2018, "step": 11470 }, { "epoch": 3.052421500798297, "grad_norm": 0.28393658995628357, "learning_rate": 6.835934339397207e-08, "loss": 0.191, "step": 11471 }, { "epoch": 3.052687599787121, "grad_norm": 0.29259192943573, "learning_rate": 6.83431594883942e-08, "loss": 0.1901, "step": 11472 }, { "epoch": 3.0529536987759447, "grad_norm": 0.2983645498752594, "learning_rate": 6.832697650426475e-08, "loss": 0.1849, "step": 11473 }, { "epoch": 3.0532197977647684, "grad_norm": 0.30142942070961, "learning_rate": 6.831079444205478e-08, "loss": 0.2136, "step": 11474 }, { "epoch": 3.0534858967535925, "grad_norm": 0.30192694067955017, "learning_rate": 6.829461330223532e-08, "loss": 0.1975, "step": 11475 }, { "epoch": 3.053751995742416, "grad_norm": 0.3094893991947174, "learning_rate": 6.827843308527733e-08, "loss": 0.1937, "step": 11476 }, { "epoch": 3.05401809473124, "grad_norm": 0.29606226086616516, "learning_rate": 6.826225379165181e-08, "loss": 0.184, "step": 11477 }, { "epoch": 3.054284193720064, "grad_norm": 0.2728894352912903, "learning_rate": 6.824607542182966e-08, "loss": 0.1777, "step": 11478 }, { "epoch": 3.0545502927088877, "grad_norm": 0.2910973131656647, "learning_rate": 6.82298979762818e-08, "loss": 0.1821, "step": 11479 }, { "epoch": 3.0548163916977114, "grad_norm": 0.2835630476474762, "learning_rate": 6.821372145547914e-08, "loss": 0.1597, "step": 11480 }, { "epoch": 3.0550824906865355, "grad_norm": 0.2610424757003784, "learning_rate": 6.819754585989249e-08, "loss": 0.1924, "step": 11481 }, { "epoch": 3.055348589675359, "grad_norm": 0.2910638153553009, "learning_rate": 6.818137118999272e-08, "loss": 0.1845, "step": 11482 }, { "epoch": 3.0556146886641833, "grad_norm": 0.30506718158721924, "learning_rate": 6.81651974462506e-08, "loss": 0.2073, "step": 11483 }, { "epoch": 3.055880787653007, "grad_norm": 0.28977349400520325, "learning_rate": 6.814902462913694e-08, "loss": 0.2036, "step": 11484 }, { "epoch": 3.0561468866418307, "grad_norm": 0.2981196939945221, "learning_rate": 6.813285273912246e-08, "loss": 0.2046, "step": 11485 }, { "epoch": 3.0564129856306548, "grad_norm": 0.3000505268573761, "learning_rate": 6.811668177667795e-08, "loss": 0.1972, "step": 11486 }, { "epoch": 3.0566790846194785, "grad_norm": 0.29255250096321106, "learning_rate": 6.8100511742274e-08, "loss": 0.1887, "step": 11487 }, { "epoch": 3.056945183608302, "grad_norm": 0.298052579164505, "learning_rate": 6.808434263638132e-08, "loss": 0.1956, "step": 11488 }, { "epoch": 3.0572112825971263, "grad_norm": 0.3028098940849304, "learning_rate": 6.806817445947057e-08, "loss": 0.1807, "step": 11489 }, { "epoch": 3.05747738158595, "grad_norm": 0.3138989508152008, "learning_rate": 6.805200721201236e-08, "loss": 0.1954, "step": 11490 }, { "epoch": 3.0577434805747736, "grad_norm": 0.2694082260131836, "learning_rate": 6.803584089447726e-08, "loss": 0.1626, "step": 11491 }, { "epoch": 3.0580095795635978, "grad_norm": 0.29486942291259766, "learning_rate": 6.801967550733583e-08, "loss": 0.1815, "step": 11492 }, { "epoch": 3.0582756785524214, "grad_norm": 0.3127530515193939, "learning_rate": 6.800351105105862e-08, "loss": 0.1911, "step": 11493 }, { "epoch": 3.0585417775412456, "grad_norm": 0.3054407238960266, "learning_rate": 6.79873475261161e-08, "loss": 0.1971, "step": 11494 }, { "epoch": 3.0588078765300692, "grad_norm": 0.302768349647522, "learning_rate": 6.797118493297884e-08, "loss": 0.1894, "step": 11495 }, { "epoch": 3.059073975518893, "grad_norm": 0.3408339321613312, "learning_rate": 6.795502327211718e-08, "loss": 0.1913, "step": 11496 }, { "epoch": 3.059340074507717, "grad_norm": 0.2810031771659851, "learning_rate": 6.793886254400159e-08, "loss": 0.1759, "step": 11497 }, { "epoch": 3.0596061734965407, "grad_norm": 0.2946099042892456, "learning_rate": 6.792270274910246e-08, "loss": 0.1901, "step": 11498 }, { "epoch": 3.0598722724853644, "grad_norm": 0.3324200212955475, "learning_rate": 6.790654388789016e-08, "loss": 0.2077, "step": 11499 }, { "epoch": 3.0601383714741885, "grad_norm": 0.30792978405952454, "learning_rate": 6.789038596083505e-08, "loss": 0.2035, "step": 11500 }, { "epoch": 3.0604044704630122, "grad_norm": 0.31515201926231384, "learning_rate": 6.787422896840743e-08, "loss": 0.2022, "step": 11501 }, { "epoch": 3.060670569451836, "grad_norm": 0.2843989431858063, "learning_rate": 6.785807291107759e-08, "loss": 0.1851, "step": 11502 }, { "epoch": 3.06093666844066, "grad_norm": 0.2851492464542389, "learning_rate": 6.784191778931579e-08, "loss": 0.1826, "step": 11503 }, { "epoch": 3.0612027674294837, "grad_norm": 0.2699640393257141, "learning_rate": 6.782576360359225e-08, "loss": 0.175, "step": 11504 }, { "epoch": 3.0614688664183074, "grad_norm": 0.2764764726161957, "learning_rate": 6.780961035437725e-08, "loss": 0.173, "step": 11505 }, { "epoch": 3.0617349654071315, "grad_norm": 0.2950405478477478, "learning_rate": 6.779345804214087e-08, "loss": 0.1944, "step": 11506 }, { "epoch": 3.062001064395955, "grad_norm": 0.28038638830184937, "learning_rate": 6.77773066673533e-08, "loss": 0.1923, "step": 11507 }, { "epoch": 3.0622671633847793, "grad_norm": 0.2962670624256134, "learning_rate": 6.776115623048466e-08, "loss": 0.1944, "step": 11508 }, { "epoch": 3.062533262373603, "grad_norm": 0.28296947479248047, "learning_rate": 6.774500673200508e-08, "loss": 0.1731, "step": 11509 }, { "epoch": 3.0627993613624267, "grad_norm": 0.2877642512321472, "learning_rate": 6.772885817238458e-08, "loss": 0.1782, "step": 11510 }, { "epoch": 3.063065460351251, "grad_norm": 0.31705453991889954, "learning_rate": 6.771271055209324e-08, "loss": 0.1978, "step": 11511 }, { "epoch": 3.0633315593400745, "grad_norm": 0.2703797519207001, "learning_rate": 6.769656387160105e-08, "loss": 0.1925, "step": 11512 }, { "epoch": 3.063597658328898, "grad_norm": 0.2678728997707367, "learning_rate": 6.768041813137803e-08, "loss": 0.1729, "step": 11513 }, { "epoch": 3.0638637573177223, "grad_norm": 0.3167758285999298, "learning_rate": 6.766427333189415e-08, "loss": 0.1945, "step": 11514 }, { "epoch": 3.064129856306546, "grad_norm": 0.30353692173957825, "learning_rate": 6.764812947361927e-08, "loss": 0.1817, "step": 11515 }, { "epoch": 3.0643959552953697, "grad_norm": 0.2863709330558777, "learning_rate": 6.763198655702334e-08, "loss": 0.1884, "step": 11516 }, { "epoch": 3.064662054284194, "grad_norm": 0.29520392417907715, "learning_rate": 6.761584458257624e-08, "loss": 0.1863, "step": 11517 }, { "epoch": 3.0649281532730175, "grad_norm": 0.28053954243659973, "learning_rate": 6.759970355074783e-08, "loss": 0.1875, "step": 11518 }, { "epoch": 3.0651942522618416, "grad_norm": 0.307920902967453, "learning_rate": 6.758356346200791e-08, "loss": 0.1982, "step": 11519 }, { "epoch": 3.0654603512506653, "grad_norm": 0.283311128616333, "learning_rate": 6.756742431682631e-08, "loss": 0.1807, "step": 11520 }, { "epoch": 3.065726450239489, "grad_norm": 0.2701267898082733, "learning_rate": 6.755128611567276e-08, "loss": 0.1765, "step": 11521 }, { "epoch": 3.065992549228313, "grad_norm": 0.28468558192253113, "learning_rate": 6.753514885901703e-08, "loss": 0.1824, "step": 11522 }, { "epoch": 3.066258648217137, "grad_norm": 0.28449559211730957, "learning_rate": 6.751901254732886e-08, "loss": 0.1856, "step": 11523 }, { "epoch": 3.0665247472059605, "grad_norm": 0.28541386127471924, "learning_rate": 6.750287718107784e-08, "loss": 0.1978, "step": 11524 }, { "epoch": 3.0667908461947846, "grad_norm": 0.29754430055618286, "learning_rate": 6.748674276073369e-08, "loss": 0.2003, "step": 11525 }, { "epoch": 3.0670569451836083, "grad_norm": 0.2858324646949768, "learning_rate": 6.747060928676606e-08, "loss": 0.1903, "step": 11526 }, { "epoch": 3.067323044172432, "grad_norm": 0.29785940051078796, "learning_rate": 6.74544767596445e-08, "loss": 0.1815, "step": 11527 }, { "epoch": 3.067589143161256, "grad_norm": 0.288800984621048, "learning_rate": 6.743834517983865e-08, "loss": 0.2043, "step": 11528 }, { "epoch": 3.0678552421500798, "grad_norm": 0.2903235852718353, "learning_rate": 6.742221454781803e-08, "loss": 0.1867, "step": 11529 }, { "epoch": 3.0681213411389034, "grad_norm": 0.29719066619873047, "learning_rate": 6.740608486405214e-08, "loss": 0.1785, "step": 11530 }, { "epoch": 3.0683874401277276, "grad_norm": 0.296732634305954, "learning_rate": 6.738995612901051e-08, "loss": 0.1965, "step": 11531 }, { "epoch": 3.0686535391165513, "grad_norm": 0.27511870861053467, "learning_rate": 6.73738283431626e-08, "loss": 0.1855, "step": 11532 }, { "epoch": 3.0689196381053754, "grad_norm": 0.29012206196784973, "learning_rate": 6.73577015069778e-08, "loss": 0.1836, "step": 11533 }, { "epoch": 3.069185737094199, "grad_norm": 0.28639259934425354, "learning_rate": 6.734157562092555e-08, "loss": 0.1789, "step": 11534 }, { "epoch": 3.0694518360830227, "grad_norm": 0.2823973596096039, "learning_rate": 6.732545068547524e-08, "loss": 0.1949, "step": 11535 }, { "epoch": 3.069717935071847, "grad_norm": 0.3061840534210205, "learning_rate": 6.730932670109623e-08, "loss": 0.1817, "step": 11536 }, { "epoch": 3.0699840340606706, "grad_norm": 0.32360878586769104, "learning_rate": 6.729320366825783e-08, "loss": 0.2072, "step": 11537 }, { "epoch": 3.0702501330494942, "grad_norm": 0.2838860750198364, "learning_rate": 6.727708158742936e-08, "loss": 0.1762, "step": 11538 }, { "epoch": 3.0705162320383184, "grad_norm": 0.30475133657455444, "learning_rate": 6.726096045908009e-08, "loss": 0.2008, "step": 11539 }, { "epoch": 3.070782331027142, "grad_norm": 0.2921620011329651, "learning_rate": 6.724484028367923e-08, "loss": 0.2127, "step": 11540 }, { "epoch": 3.0710484300159657, "grad_norm": 0.2896413803100586, "learning_rate": 6.722872106169606e-08, "loss": 0.1873, "step": 11541 }, { "epoch": 3.07131452900479, "grad_norm": 0.30306917428970337, "learning_rate": 6.721260279359972e-08, "loss": 0.2016, "step": 11542 }, { "epoch": 3.0715806279936135, "grad_norm": 0.2879989743232727, "learning_rate": 6.719648547985936e-08, "loss": 0.1892, "step": 11543 }, { "epoch": 3.0718467269824377, "grad_norm": 0.2982316017150879, "learning_rate": 6.718036912094414e-08, "loss": 0.1879, "step": 11544 }, { "epoch": 3.0721128259712613, "grad_norm": 0.2890530824661255, "learning_rate": 6.716425371732315e-08, "loss": 0.1779, "step": 11545 }, { "epoch": 3.072378924960085, "grad_norm": 0.2919999361038208, "learning_rate": 6.714813926946548e-08, "loss": 0.1839, "step": 11546 }, { "epoch": 3.072645023948909, "grad_norm": 0.31943708658218384, "learning_rate": 6.713202577784017e-08, "loss": 0.1971, "step": 11547 }, { "epoch": 3.072911122937733, "grad_norm": 0.3068026900291443, "learning_rate": 6.711591324291626e-08, "loss": 0.1866, "step": 11548 }, { "epoch": 3.0731772219265565, "grad_norm": 0.2593258321285248, "learning_rate": 6.709980166516275e-08, "loss": 0.1663, "step": 11549 }, { "epoch": 3.0734433209153806, "grad_norm": 0.3056349456310272, "learning_rate": 6.708369104504858e-08, "loss": 0.1815, "step": 11550 }, { "epoch": 3.0737094199042043, "grad_norm": 0.3125321567058563, "learning_rate": 6.706758138304269e-08, "loss": 0.2034, "step": 11551 }, { "epoch": 3.073975518893028, "grad_norm": 0.3067212998867035, "learning_rate": 6.705147267961401e-08, "loss": 0.1981, "step": 11552 }, { "epoch": 3.074241617881852, "grad_norm": 0.2904260456562042, "learning_rate": 6.70353649352314e-08, "loss": 0.195, "step": 11553 }, { "epoch": 3.074507716870676, "grad_norm": 0.31347784399986267, "learning_rate": 6.701925815036372e-08, "loss": 0.1913, "step": 11554 }, { "epoch": 3.0747738158595, "grad_norm": 0.2999940514564514, "learning_rate": 6.70031523254798e-08, "loss": 0.1998, "step": 11555 }, { "epoch": 3.0750399148483236, "grad_norm": 0.3652932941913605, "learning_rate": 6.698704746104844e-08, "loss": 0.2079, "step": 11556 }, { "epoch": 3.0753060138371473, "grad_norm": 0.3116647005081177, "learning_rate": 6.697094355753842e-08, "loss": 0.2063, "step": 11557 }, { "epoch": 3.0755721128259714, "grad_norm": 0.3108910620212555, "learning_rate": 6.695484061541847e-08, "loss": 0.2047, "step": 11558 }, { "epoch": 3.075838211814795, "grad_norm": 0.2962067723274231, "learning_rate": 6.693873863515733e-08, "loss": 0.1897, "step": 11559 }, { "epoch": 3.076104310803619, "grad_norm": 0.2720482349395752, "learning_rate": 6.692263761722365e-08, "loss": 0.1785, "step": 11560 }, { "epoch": 3.076370409792443, "grad_norm": 0.29131922125816345, "learning_rate": 6.69065375620861e-08, "loss": 0.188, "step": 11561 }, { "epoch": 3.0766365087812666, "grad_norm": 0.3033633828163147, "learning_rate": 6.689043847021332e-08, "loss": 0.1984, "step": 11562 }, { "epoch": 3.0769026077700903, "grad_norm": 0.2955096662044525, "learning_rate": 6.687434034207391e-08, "loss": 0.1824, "step": 11563 }, { "epoch": 3.0771687067589144, "grad_norm": 0.298973023891449, "learning_rate": 6.685824317813643e-08, "loss": 0.2031, "step": 11564 }, { "epoch": 3.077434805747738, "grad_norm": 0.30768173933029175, "learning_rate": 6.684214697886943e-08, "loss": 0.2225, "step": 11565 }, { "epoch": 3.077700904736562, "grad_norm": 0.31777915358543396, "learning_rate": 6.682605174474146e-08, "loss": 0.2043, "step": 11566 }, { "epoch": 3.077967003725386, "grad_norm": 0.2865314781665802, "learning_rate": 6.680995747622097e-08, "loss": 0.1933, "step": 11567 }, { "epoch": 3.0782331027142096, "grad_norm": 0.3084949553012848, "learning_rate": 6.679386417377648e-08, "loss": 0.1918, "step": 11568 }, { "epoch": 3.0784992017030337, "grad_norm": 0.2885076403617859, "learning_rate": 6.677777183787636e-08, "loss": 0.183, "step": 11569 }, { "epoch": 3.0787653006918574, "grad_norm": 0.3052273094654083, "learning_rate": 6.676168046898903e-08, "loss": 0.1863, "step": 11570 }, { "epoch": 3.079031399680681, "grad_norm": 0.34541210532188416, "learning_rate": 6.67455900675829e-08, "loss": 0.1949, "step": 11571 }, { "epoch": 3.079297498669505, "grad_norm": 0.33250823616981506, "learning_rate": 6.67295006341263e-08, "loss": 0.2259, "step": 11572 }, { "epoch": 3.079563597658329, "grad_norm": 0.29752030968666077, "learning_rate": 6.671341216908752e-08, "loss": 0.2079, "step": 11573 }, { "epoch": 3.0798296966471526, "grad_norm": 0.29315802454948425, "learning_rate": 6.66973246729349e-08, "loss": 0.1849, "step": 11574 }, { "epoch": 3.0800957956359767, "grad_norm": 0.3049522042274475, "learning_rate": 6.668123814613669e-08, "loss": 0.2002, "step": 11575 }, { "epoch": 3.0803618946248004, "grad_norm": 0.28315821290016174, "learning_rate": 6.666515258916112e-08, "loss": 0.1889, "step": 11576 }, { "epoch": 3.080627993613624, "grad_norm": 0.29496264457702637, "learning_rate": 6.664906800247645e-08, "loss": 0.1832, "step": 11577 }, { "epoch": 3.080894092602448, "grad_norm": 0.31888118386268616, "learning_rate": 6.663298438655076e-08, "loss": 0.2031, "step": 11578 }, { "epoch": 3.081160191591272, "grad_norm": 0.3343154788017273, "learning_rate": 6.661690174185228e-08, "loss": 0.2199, "step": 11579 }, { "epoch": 3.081426290580096, "grad_norm": 0.2914772629737854, "learning_rate": 6.660082006884911e-08, "loss": 0.19, "step": 11580 }, { "epoch": 3.0816923895689197, "grad_norm": 0.28743690252304077, "learning_rate": 6.658473936800935e-08, "loss": 0.1965, "step": 11581 }, { "epoch": 3.0819584885577433, "grad_norm": 0.27369529008865356, "learning_rate": 6.656865963980106e-08, "loss": 0.1899, "step": 11582 }, { "epoch": 3.0822245875465675, "grad_norm": 0.2899593412876129, "learning_rate": 6.655258088469228e-08, "loss": 0.2004, "step": 11583 }, { "epoch": 3.082490686535391, "grad_norm": 0.2842787802219391, "learning_rate": 6.653650310315102e-08, "loss": 0.1788, "step": 11584 }, { "epoch": 3.082756785524215, "grad_norm": 0.29974815249443054, "learning_rate": 6.652042629564527e-08, "loss": 0.1934, "step": 11585 }, { "epoch": 3.083022884513039, "grad_norm": 0.30895429849624634, "learning_rate": 6.650435046264303e-08, "loss": 0.2069, "step": 11586 }, { "epoch": 3.0832889835018626, "grad_norm": 0.29551607370376587, "learning_rate": 6.648827560461212e-08, "loss": 0.1836, "step": 11587 }, { "epoch": 3.0835550824906863, "grad_norm": 0.26856452226638794, "learning_rate": 6.647220172202052e-08, "loss": 0.1881, "step": 11588 }, { "epoch": 3.0838211814795105, "grad_norm": 0.2929767966270447, "learning_rate": 6.645612881533606e-08, "loss": 0.1886, "step": 11589 }, { "epoch": 3.084087280468334, "grad_norm": 0.27910688519477844, "learning_rate": 6.644005688502658e-08, "loss": 0.1765, "step": 11590 }, { "epoch": 3.0843533794571583, "grad_norm": 0.30648085474967957, "learning_rate": 6.642398593155995e-08, "loss": 0.2138, "step": 11591 }, { "epoch": 3.084619478445982, "grad_norm": 0.2822621762752533, "learning_rate": 6.640791595540388e-08, "loss": 0.1863, "step": 11592 }, { "epoch": 3.0848855774348056, "grad_norm": 0.29523083567619324, "learning_rate": 6.639184695702616e-08, "loss": 0.184, "step": 11593 }, { "epoch": 3.0851516764236298, "grad_norm": 0.3090888559818268, "learning_rate": 6.63757789368945e-08, "loss": 0.2114, "step": 11594 }, { "epoch": 3.0854177754124534, "grad_norm": 0.30661576986312866, "learning_rate": 6.635971189547667e-08, "loss": 0.2086, "step": 11595 }, { "epoch": 3.085683874401277, "grad_norm": 0.275616854429245, "learning_rate": 6.634364583324023e-08, "loss": 0.1871, "step": 11596 }, { "epoch": 3.0859499733901012, "grad_norm": 0.2716244161128998, "learning_rate": 6.632758075065288e-08, "loss": 0.1918, "step": 11597 }, { "epoch": 3.086216072378925, "grad_norm": 0.28900521993637085, "learning_rate": 6.63115166481822e-08, "loss": 0.1911, "step": 11598 }, { "epoch": 3.0864821713677486, "grad_norm": 0.29802578687667847, "learning_rate": 6.629545352629582e-08, "loss": 0.1905, "step": 11599 }, { "epoch": 3.0867482703565727, "grad_norm": 0.30920663475990295, "learning_rate": 6.627939138546128e-08, "loss": 0.198, "step": 11600 }, { "epoch": 3.0870143693453964, "grad_norm": 0.3119189441204071, "learning_rate": 6.626333022614612e-08, "loss": 0.1988, "step": 11601 }, { "epoch": 3.0872804683342205, "grad_norm": 0.3151966631412506, "learning_rate": 6.624727004881779e-08, "loss": 0.2065, "step": 11602 }, { "epoch": 3.0875465673230442, "grad_norm": 0.3024337887763977, "learning_rate": 6.623121085394378e-08, "loss": 0.1932, "step": 11603 }, { "epoch": 3.087812666311868, "grad_norm": 0.3119645416736603, "learning_rate": 6.621515264199161e-08, "loss": 0.1966, "step": 11604 }, { "epoch": 3.088078765300692, "grad_norm": 0.31902647018432617, "learning_rate": 6.619909541342857e-08, "loss": 0.197, "step": 11605 }, { "epoch": 3.0883448642895157, "grad_norm": 0.313050776720047, "learning_rate": 6.61830391687221e-08, "loss": 0.1826, "step": 11606 }, { "epoch": 3.0886109632783394, "grad_norm": 0.32208070158958435, "learning_rate": 6.616698390833955e-08, "loss": 0.2129, "step": 11607 }, { "epoch": 3.0888770622671635, "grad_norm": 0.29585132002830505, "learning_rate": 6.615092963274826e-08, "loss": 0.1889, "step": 11608 }, { "epoch": 3.089143161255987, "grad_norm": 0.28717041015625, "learning_rate": 6.613487634241552e-08, "loss": 0.1984, "step": 11609 }, { "epoch": 3.089409260244811, "grad_norm": 0.3214147090911865, "learning_rate": 6.611882403780861e-08, "loss": 0.197, "step": 11610 }, { "epoch": 3.089675359233635, "grad_norm": 0.33162757754325867, "learning_rate": 6.610277271939477e-08, "loss": 0.2029, "step": 11611 }, { "epoch": 3.0899414582224587, "grad_norm": 0.3457496464252472, "learning_rate": 6.608672238764117e-08, "loss": 0.2111, "step": 11612 }, { "epoch": 3.090207557211283, "grad_norm": 0.2829594314098358, "learning_rate": 6.607067304301506e-08, "loss": 0.196, "step": 11613 }, { "epoch": 3.0904736562001065, "grad_norm": 0.28270429372787476, "learning_rate": 6.605462468598357e-08, "loss": 0.1847, "step": 11614 }, { "epoch": 3.09073975518893, "grad_norm": 0.31267666816711426, "learning_rate": 6.60385773170138e-08, "loss": 0.211, "step": 11615 }, { "epoch": 3.0910058541777543, "grad_norm": 0.30289241671562195, "learning_rate": 6.602253093657284e-08, "loss": 0.2111, "step": 11616 }, { "epoch": 3.091271953166578, "grad_norm": 0.3031289875507355, "learning_rate": 6.600648554512779e-08, "loss": 0.1994, "step": 11617 }, { "epoch": 3.0915380521554017, "grad_norm": 0.2971198558807373, "learning_rate": 6.599044114314568e-08, "loss": 0.1922, "step": 11618 }, { "epoch": 3.091804151144226, "grad_norm": 0.28122326731681824, "learning_rate": 6.597439773109353e-08, "loss": 0.1767, "step": 11619 }, { "epoch": 3.0920702501330495, "grad_norm": 0.33864617347717285, "learning_rate": 6.595835530943832e-08, "loss": 0.195, "step": 11620 }, { "epoch": 3.092336349121873, "grad_norm": 0.27873390913009644, "learning_rate": 6.594231387864702e-08, "loss": 0.1766, "step": 11621 }, { "epoch": 3.0926024481106973, "grad_norm": 0.2943953275680542, "learning_rate": 6.592627343918652e-08, "loss": 0.1982, "step": 11622 }, { "epoch": 3.092868547099521, "grad_norm": 0.27816441655158997, "learning_rate": 6.591023399152373e-08, "loss": 0.2002, "step": 11623 }, { "epoch": 3.0931346460883447, "grad_norm": 0.2843410074710846, "learning_rate": 6.589419553612551e-08, "loss": 0.1863, "step": 11624 }, { "epoch": 3.093400745077169, "grad_norm": 0.28466764092445374, "learning_rate": 6.587815807345869e-08, "loss": 0.1891, "step": 11625 }, { "epoch": 3.0936668440659925, "grad_norm": 0.268985390663147, "learning_rate": 6.586212160399009e-08, "loss": 0.1807, "step": 11626 }, { "epoch": 3.0939329430548166, "grad_norm": 0.3141425848007202, "learning_rate": 6.58460861281865e-08, "loss": 0.2221, "step": 11627 }, { "epoch": 3.0941990420436403, "grad_norm": 0.29741358757019043, "learning_rate": 6.583005164651467e-08, "loss": 0.1993, "step": 11628 }, { "epoch": 3.094465141032464, "grad_norm": 0.28981679677963257, "learning_rate": 6.58140181594413e-08, "loss": 0.1824, "step": 11629 }, { "epoch": 3.094731240021288, "grad_norm": 0.29393547773361206, "learning_rate": 6.579798566743313e-08, "loss": 0.2026, "step": 11630 }, { "epoch": 3.0949973390101118, "grad_norm": 0.27815935015678406, "learning_rate": 6.578195417095678e-08, "loss": 0.1841, "step": 11631 }, { "epoch": 3.0952634379989354, "grad_norm": 0.28528130054473877, "learning_rate": 6.576592367047889e-08, "loss": 0.1982, "step": 11632 }, { "epoch": 3.0955295369877596, "grad_norm": 0.3088056147098541, "learning_rate": 6.574989416646609e-08, "loss": 0.2023, "step": 11633 }, { "epoch": 3.0957956359765832, "grad_norm": 0.3192313015460968, "learning_rate": 6.573386565938492e-08, "loss": 0.2025, "step": 11634 }, { "epoch": 3.096061734965407, "grad_norm": 0.2767128050327301, "learning_rate": 6.571783814970196e-08, "loss": 0.1904, "step": 11635 }, { "epoch": 3.096327833954231, "grad_norm": 0.29536712169647217, "learning_rate": 6.57018116378837e-08, "loss": 0.1951, "step": 11636 }, { "epoch": 3.0965939329430547, "grad_norm": 0.30788108706474304, "learning_rate": 6.568578612439666e-08, "loss": 0.2049, "step": 11637 }, { "epoch": 3.096860031931879, "grad_norm": 0.2665921449661255, "learning_rate": 6.566976160970727e-08, "loss": 0.1726, "step": 11638 }, { "epoch": 3.0971261309207025, "grad_norm": 0.30738723278045654, "learning_rate": 6.565373809428201e-08, "loss": 0.19, "step": 11639 }, { "epoch": 3.0973922299095262, "grad_norm": 0.29606863856315613, "learning_rate": 6.563771557858729e-08, "loss": 0.1923, "step": 11640 }, { "epoch": 3.0976583288983504, "grad_norm": 0.2905563414096832, "learning_rate": 6.562169406308939e-08, "loss": 0.1907, "step": 11641 }, { "epoch": 3.097924427887174, "grad_norm": 0.3148794174194336, "learning_rate": 6.560567354825473e-08, "loss": 0.1777, "step": 11642 }, { "epoch": 3.0981905268759977, "grad_norm": 0.32402390241622925, "learning_rate": 6.558965403454964e-08, "loss": 0.1919, "step": 11643 }, { "epoch": 3.098456625864822, "grad_norm": 0.29485809803009033, "learning_rate": 6.557363552244033e-08, "loss": 0.1835, "step": 11644 }, { "epoch": 3.0987227248536455, "grad_norm": 0.3033834397792816, "learning_rate": 6.555761801239313e-08, "loss": 0.2019, "step": 11645 }, { "epoch": 3.098988823842469, "grad_norm": 0.29417914152145386, "learning_rate": 6.554160150487421e-08, "loss": 0.1811, "step": 11646 }, { "epoch": 3.0992549228312933, "grad_norm": 0.2859043478965759, "learning_rate": 6.552558600034982e-08, "loss": 0.1791, "step": 11647 }, { "epoch": 3.099521021820117, "grad_norm": 0.28325188159942627, "learning_rate": 6.55095714992861e-08, "loss": 0.184, "step": 11648 }, { "epoch": 3.0997871208089407, "grad_norm": 0.309068888425827, "learning_rate": 6.549355800214924e-08, "loss": 0.2017, "step": 11649 }, { "epoch": 3.100053219797765, "grad_norm": 0.28778979182243347, "learning_rate": 6.547754550940529e-08, "loss": 0.1736, "step": 11650 }, { "epoch": 3.1003193187865885, "grad_norm": 0.3112674653530121, "learning_rate": 6.546153402152033e-08, "loss": 0.2026, "step": 11651 }, { "epoch": 3.1005854177754126, "grad_norm": 0.3052460551261902, "learning_rate": 6.544552353896044e-08, "loss": 0.1984, "step": 11652 }, { "epoch": 3.1008515167642363, "grad_norm": 0.2800247073173523, "learning_rate": 6.542951406219168e-08, "loss": 0.1768, "step": 11653 }, { "epoch": 3.10111761575306, "grad_norm": 0.28776198625564575, "learning_rate": 6.541350559167996e-08, "loss": 0.1961, "step": 11654 }, { "epoch": 3.101383714741884, "grad_norm": 0.314396470785141, "learning_rate": 6.53974981278913e-08, "loss": 0.1849, "step": 11655 }, { "epoch": 3.101649813730708, "grad_norm": 0.29221636056900024, "learning_rate": 6.538149167129164e-08, "loss": 0.1867, "step": 11656 }, { "epoch": 3.1019159127195315, "grad_norm": 0.3125903904438019, "learning_rate": 6.536548622234684e-08, "loss": 0.2184, "step": 11657 }, { "epoch": 3.1021820117083556, "grad_norm": 0.30944889783859253, "learning_rate": 6.534948178152286e-08, "loss": 0.192, "step": 11658 }, { "epoch": 3.1024481106971793, "grad_norm": 0.3354684114456177, "learning_rate": 6.533347834928546e-08, "loss": 0.1851, "step": 11659 }, { "epoch": 3.1027142096860034, "grad_norm": 0.28921210765838623, "learning_rate": 6.531747592610051e-08, "loss": 0.1954, "step": 11660 }, { "epoch": 3.102980308674827, "grad_norm": 0.28804847598075867, "learning_rate": 6.530147451243377e-08, "loss": 0.2005, "step": 11661 }, { "epoch": 3.103246407663651, "grad_norm": 0.30220478773117065, "learning_rate": 6.5285474108751e-08, "loss": 0.1923, "step": 11662 }, { "epoch": 3.103512506652475, "grad_norm": 0.2883704900741577, "learning_rate": 6.526947471551798e-08, "loss": 0.1965, "step": 11663 }, { "epoch": 3.1037786056412986, "grad_norm": 0.26938194036483765, "learning_rate": 6.525347633320036e-08, "loss": 0.1821, "step": 11664 }, { "epoch": 3.1040447046301223, "grad_norm": 0.30771416425704956, "learning_rate": 6.52374789622638e-08, "loss": 0.2066, "step": 11665 }, { "epoch": 3.1043108036189464, "grad_norm": 0.3185529410839081, "learning_rate": 6.522148260317399e-08, "loss": 0.2128, "step": 11666 }, { "epoch": 3.10457690260777, "grad_norm": 0.2900647521018982, "learning_rate": 6.520548725639656e-08, "loss": 0.191, "step": 11667 }, { "epoch": 3.1048430015965938, "grad_norm": 0.2938987910747528, "learning_rate": 6.518949292239699e-08, "loss": 0.1912, "step": 11668 }, { "epoch": 3.105109100585418, "grad_norm": 0.28350144624710083, "learning_rate": 6.517349960164092e-08, "loss": 0.1752, "step": 11669 }, { "epoch": 3.1053751995742416, "grad_norm": 0.28462138772010803, "learning_rate": 6.515750729459384e-08, "loss": 0.1875, "step": 11670 }, { "epoch": 3.1056412985630653, "grad_norm": 0.29729995131492615, "learning_rate": 6.514151600172126e-08, "loss": 0.1813, "step": 11671 }, { "epoch": 3.1059073975518894, "grad_norm": 0.3095107078552246, "learning_rate": 6.512552572348865e-08, "loss": 0.1993, "step": 11672 }, { "epoch": 3.106173496540713, "grad_norm": 0.2874143123626709, "learning_rate": 6.510953646036143e-08, "loss": 0.2114, "step": 11673 }, { "epoch": 3.106439595529537, "grad_norm": 0.28405335545539856, "learning_rate": 6.5093548212805e-08, "loss": 0.1824, "step": 11674 }, { "epoch": 3.106705694518361, "grad_norm": 0.28603142499923706, "learning_rate": 6.507756098128475e-08, "loss": 0.1942, "step": 11675 }, { "epoch": 3.1069717935071846, "grad_norm": 0.2762472927570343, "learning_rate": 6.506157476626606e-08, "loss": 0.1821, "step": 11676 }, { "epoch": 3.1072378924960087, "grad_norm": 0.30630695819854736, "learning_rate": 6.504558956821419e-08, "loss": 0.1889, "step": 11677 }, { "epoch": 3.1075039914848324, "grad_norm": 0.30769580602645874, "learning_rate": 6.502960538759443e-08, "loss": 0.1988, "step": 11678 }, { "epoch": 3.107770090473656, "grad_norm": 0.3337559401988983, "learning_rate": 6.501362222487206e-08, "loss": 0.2012, "step": 11679 }, { "epoch": 3.10803618946248, "grad_norm": 0.2944246530532837, "learning_rate": 6.499764008051232e-08, "loss": 0.2041, "step": 11680 }, { "epoch": 3.108302288451304, "grad_norm": 0.31059449911117554, "learning_rate": 6.498165895498038e-08, "loss": 0.2, "step": 11681 }, { "epoch": 3.1085683874401275, "grad_norm": 0.30033212900161743, "learning_rate": 6.496567884874145e-08, "loss": 0.2031, "step": 11682 }, { "epoch": 3.1088344864289517, "grad_norm": 0.29973751306533813, "learning_rate": 6.494969976226062e-08, "loss": 0.1905, "step": 11683 }, { "epoch": 3.1091005854177753, "grad_norm": 0.2982700765132904, "learning_rate": 6.493372169600303e-08, "loss": 0.1918, "step": 11684 }, { "epoch": 3.1093666844065995, "grad_norm": 0.2804579436779022, "learning_rate": 6.491774465043378e-08, "loss": 0.1773, "step": 11685 }, { "epoch": 3.109632783395423, "grad_norm": 0.2879868149757385, "learning_rate": 6.490176862601787e-08, "loss": 0.1824, "step": 11686 }, { "epoch": 3.109898882384247, "grad_norm": 0.28695032000541687, "learning_rate": 6.488579362322033e-08, "loss": 0.1719, "step": 11687 }, { "epoch": 3.110164981373071, "grad_norm": 0.3072824478149414, "learning_rate": 6.486981964250618e-08, "loss": 0.2003, "step": 11688 }, { "epoch": 3.1104310803618946, "grad_norm": 0.29058244824409485, "learning_rate": 6.485384668434035e-08, "loss": 0.2028, "step": 11689 }, { "epoch": 3.1106971793507183, "grad_norm": 0.2994741201400757, "learning_rate": 6.483787474918778e-08, "loss": 0.1999, "step": 11690 }, { "epoch": 3.1109632783395424, "grad_norm": 0.2712177038192749, "learning_rate": 6.48219038375134e-08, "loss": 0.1814, "step": 11691 }, { "epoch": 3.111229377328366, "grad_norm": 0.3173598349094391, "learning_rate": 6.480593394978207e-08, "loss": 0.2138, "step": 11692 }, { "epoch": 3.11149547631719, "grad_norm": 0.2953314185142517, "learning_rate": 6.47899650864586e-08, "loss": 0.1973, "step": 11693 }, { "epoch": 3.111761575306014, "grad_norm": 0.28017187118530273, "learning_rate": 6.477399724800786e-08, "loss": 0.1832, "step": 11694 }, { "epoch": 3.1120276742948376, "grad_norm": 0.26629307866096497, "learning_rate": 6.475803043489457e-08, "loss": 0.1764, "step": 11695 }, { "epoch": 3.1122937732836613, "grad_norm": 0.27139905095100403, "learning_rate": 6.474206464758351e-08, "loss": 0.176, "step": 11696 }, { "epoch": 3.1125598722724854, "grad_norm": 0.3125045895576477, "learning_rate": 6.472609988653939e-08, "loss": 0.1972, "step": 11697 }, { "epoch": 3.112825971261309, "grad_norm": 0.29616066813468933, "learning_rate": 6.471013615222691e-08, "loss": 0.1868, "step": 11698 }, { "epoch": 3.1130920702501332, "grad_norm": 0.43921083211898804, "learning_rate": 6.469417344511076e-08, "loss": 0.1781, "step": 11699 }, { "epoch": 3.113358169238957, "grad_norm": 0.2697794735431671, "learning_rate": 6.467821176565553e-08, "loss": 0.1621, "step": 11700 }, { "epoch": 3.1136242682277806, "grad_norm": 0.311186283826828, "learning_rate": 6.466225111432585e-08, "loss": 0.2222, "step": 11701 }, { "epoch": 3.1138903672166047, "grad_norm": 0.2787412106990814, "learning_rate": 6.464629149158631e-08, "loss": 0.1896, "step": 11702 }, { "epoch": 3.1141564662054284, "grad_norm": 0.3336673676967621, "learning_rate": 6.463033289790144e-08, "loss": 0.1946, "step": 11703 }, { "epoch": 3.114422565194252, "grad_norm": 0.30454036593437195, "learning_rate": 6.461437533373569e-08, "loss": 0.1952, "step": 11704 }, { "epoch": 3.114688664183076, "grad_norm": 0.27446502447128296, "learning_rate": 6.459841879955364e-08, "loss": 0.1955, "step": 11705 }, { "epoch": 3.1149547631719, "grad_norm": 0.27410510182380676, "learning_rate": 6.458246329581968e-08, "loss": 0.1764, "step": 11706 }, { "epoch": 3.1152208621607236, "grad_norm": 0.47323334217071533, "learning_rate": 6.456650882299824e-08, "loss": 0.2281, "step": 11707 }, { "epoch": 3.1154869611495477, "grad_norm": 0.29444587230682373, "learning_rate": 6.455055538155374e-08, "loss": 0.1963, "step": 11708 }, { "epoch": 3.1157530601383714, "grad_norm": 0.32247090339660645, "learning_rate": 6.453460297195052e-08, "loss": 0.2171, "step": 11709 }, { "epoch": 3.1160191591271955, "grad_norm": 0.3010804355144501, "learning_rate": 6.451865159465293e-08, "loss": 0.1762, "step": 11710 }, { "epoch": 3.116285258116019, "grad_norm": 0.2904197871685028, "learning_rate": 6.450270125012526e-08, "loss": 0.2042, "step": 11711 }, { "epoch": 3.116551357104843, "grad_norm": 0.2705177664756775, "learning_rate": 6.448675193883183e-08, "loss": 0.1818, "step": 11712 }, { "epoch": 3.116817456093667, "grad_norm": 0.3121162950992584, "learning_rate": 6.447080366123681e-08, "loss": 0.1892, "step": 11713 }, { "epoch": 3.1170835550824907, "grad_norm": 0.29610300064086914, "learning_rate": 6.445485641780443e-08, "loss": 0.1899, "step": 11714 }, { "epoch": 3.1173496540713144, "grad_norm": 0.28711822628974915, "learning_rate": 6.443891020899892e-08, "loss": 0.1957, "step": 11715 }, { "epoch": 3.1176157530601385, "grad_norm": 0.29768994450569153, "learning_rate": 6.442296503528437e-08, "loss": 0.1973, "step": 11716 }, { "epoch": 3.117881852048962, "grad_norm": 0.2906254231929779, "learning_rate": 6.440702089712494e-08, "loss": 0.1893, "step": 11717 }, { "epoch": 3.118147951037786, "grad_norm": 0.2826876938343048, "learning_rate": 6.43910777949847e-08, "loss": 0.179, "step": 11718 }, { "epoch": 3.11841405002661, "grad_norm": 0.30023202300071716, "learning_rate": 6.437513572932776e-08, "loss": 0.2219, "step": 11719 }, { "epoch": 3.1186801490154337, "grad_norm": 0.2783626317977905, "learning_rate": 6.435919470061809e-08, "loss": 0.188, "step": 11720 }, { "epoch": 3.118946248004258, "grad_norm": 0.3048202395439148, "learning_rate": 6.434325470931977e-08, "loss": 0.2086, "step": 11721 }, { "epoch": 3.1192123469930815, "grad_norm": 0.32306501269340515, "learning_rate": 6.43273157558967e-08, "loss": 0.1941, "step": 11722 }, { "epoch": 3.119478445981905, "grad_norm": 0.2942084074020386, "learning_rate": 6.431137784081282e-08, "loss": 0.2078, "step": 11723 }, { "epoch": 3.1197445449707293, "grad_norm": 0.29320958256721497, "learning_rate": 6.42954409645321e-08, "loss": 0.1815, "step": 11724 }, { "epoch": 3.120010643959553, "grad_norm": 0.33570054173469543, "learning_rate": 6.427950512751836e-08, "loss": 0.1945, "step": 11725 }, { "epoch": 3.1202767429483766, "grad_norm": 0.2915850281715393, "learning_rate": 6.426357033023549e-08, "loss": 0.192, "step": 11726 }, { "epoch": 3.1205428419372008, "grad_norm": 0.28359317779541016, "learning_rate": 6.424763657314729e-08, "loss": 0.1925, "step": 11727 }, { "epoch": 3.1208089409260245, "grad_norm": 0.29370519518852234, "learning_rate": 6.423170385671756e-08, "loss": 0.186, "step": 11728 }, { "epoch": 3.121075039914848, "grad_norm": 0.3227192759513855, "learning_rate": 6.421577218141007e-08, "loss": 0.2162, "step": 11729 }, { "epoch": 3.1213411389036723, "grad_norm": 0.28988826274871826, "learning_rate": 6.419984154768856e-08, "loss": 0.1913, "step": 11730 }, { "epoch": 3.121607237892496, "grad_norm": 0.3051128387451172, "learning_rate": 6.418391195601669e-08, "loss": 0.1992, "step": 11731 }, { "epoch": 3.12187333688132, "grad_norm": 0.2903864085674286, "learning_rate": 6.416798340685813e-08, "loss": 0.1851, "step": 11732 }, { "epoch": 3.1221394358701438, "grad_norm": 0.28145405650138855, "learning_rate": 6.415205590067654e-08, "loss": 0.1894, "step": 11733 }, { "epoch": 3.1224055348589674, "grad_norm": 0.2995986342430115, "learning_rate": 6.413612943793555e-08, "loss": 0.2084, "step": 11734 }, { "epoch": 3.1226716338477916, "grad_norm": 0.2948741316795349, "learning_rate": 6.41202040190987e-08, "loss": 0.1789, "step": 11735 }, { "epoch": 3.1229377328366152, "grad_norm": 0.2944173216819763, "learning_rate": 6.410427964462952e-08, "loss": 0.187, "step": 11736 }, { "epoch": 3.123203831825439, "grad_norm": 0.28201812505722046, "learning_rate": 6.408835631499159e-08, "loss": 0.167, "step": 11737 }, { "epoch": 3.123469930814263, "grad_norm": 0.27093198895454407, "learning_rate": 6.407243403064834e-08, "loss": 0.1829, "step": 11738 }, { "epoch": 3.1237360298030867, "grad_norm": 0.29301655292510986, "learning_rate": 6.40565127920633e-08, "loss": 0.1706, "step": 11739 }, { "epoch": 3.1240021287919104, "grad_norm": 0.32088902592658997, "learning_rate": 6.404059259969982e-08, "loss": 0.1885, "step": 11740 }, { "epoch": 3.1242682277807345, "grad_norm": 0.3551754951477051, "learning_rate": 6.40246734540213e-08, "loss": 0.1853, "step": 11741 }, { "epoch": 3.1245343267695582, "grad_norm": 0.2958160936832428, "learning_rate": 6.400875535549114e-08, "loss": 0.1816, "step": 11742 }, { "epoch": 3.124800425758382, "grad_norm": 0.31061556935310364, "learning_rate": 6.399283830457264e-08, "loss": 0.2069, "step": 11743 }, { "epoch": 3.125066524747206, "grad_norm": 0.29052025079727173, "learning_rate": 6.397692230172918e-08, "loss": 0.1886, "step": 11744 }, { "epoch": 3.1253326237360297, "grad_norm": 0.2906312644481659, "learning_rate": 6.396100734742393e-08, "loss": 0.1787, "step": 11745 }, { "epoch": 3.125598722724854, "grad_norm": 0.295248806476593, "learning_rate": 6.39450934421202e-08, "loss": 0.1771, "step": 11746 }, { "epoch": 3.1258648217136775, "grad_norm": 0.3144305348396301, "learning_rate": 6.392918058628118e-08, "loss": 0.2, "step": 11747 }, { "epoch": 3.126130920702501, "grad_norm": 0.29985329508781433, "learning_rate": 6.39132687803701e-08, "loss": 0.2031, "step": 11748 }, { "epoch": 3.1263970196913253, "grad_norm": 0.29391682147979736, "learning_rate": 6.389735802485001e-08, "loss": 0.1923, "step": 11749 }, { "epoch": 3.126663118680149, "grad_norm": 0.3295348584651947, "learning_rate": 6.38814483201841e-08, "loss": 0.189, "step": 11750 }, { "epoch": 3.1269292176689727, "grad_norm": 0.3179134130477905, "learning_rate": 6.386553966683546e-08, "loss": 0.2115, "step": 11751 }, { "epoch": 3.127195316657797, "grad_norm": 0.28674688935279846, "learning_rate": 6.384963206526713e-08, "loss": 0.1779, "step": 11752 }, { "epoch": 3.1274614156466205, "grad_norm": 0.29033568501472473, "learning_rate": 6.383372551594212e-08, "loss": 0.1887, "step": 11753 }, { "epoch": 3.127727514635444, "grad_norm": 0.2833273112773895, "learning_rate": 6.381782001932351e-08, "loss": 0.1854, "step": 11754 }, { "epoch": 3.1279936136242683, "grad_norm": 0.26878485083580017, "learning_rate": 6.380191557587418e-08, "loss": 0.1815, "step": 11755 }, { "epoch": 3.128259712613092, "grad_norm": 0.3196609616279602, "learning_rate": 6.37860121860571e-08, "loss": 0.2133, "step": 11756 }, { "epoch": 3.128525811601916, "grad_norm": 0.27941644191741943, "learning_rate": 6.377010985033522e-08, "loss": 0.1898, "step": 11757 }, { "epoch": 3.12879191059074, "grad_norm": 0.2781100571155548, "learning_rate": 6.375420856917134e-08, "loss": 0.1563, "step": 11758 }, { "epoch": 3.1290580095795635, "grad_norm": 0.28643539547920227, "learning_rate": 6.373830834302832e-08, "loss": 0.1904, "step": 11759 }, { "epoch": 3.1293241085683876, "grad_norm": 0.28881651163101196, "learning_rate": 6.3722409172369e-08, "loss": 0.1989, "step": 11760 }, { "epoch": 3.1295902075572113, "grad_norm": 0.2792392671108246, "learning_rate": 6.370651105765616e-08, "loss": 0.1869, "step": 11761 }, { "epoch": 3.129856306546035, "grad_norm": 0.29551273584365845, "learning_rate": 6.369061399935254e-08, "loss": 0.1996, "step": 11762 }, { "epoch": 3.130122405534859, "grad_norm": 0.29656097292900085, "learning_rate": 6.367471799792088e-08, "loss": 0.1829, "step": 11763 }, { "epoch": 3.130388504523683, "grad_norm": 0.29785698652267456, "learning_rate": 6.365882305382386e-08, "loss": 0.1845, "step": 11764 }, { "epoch": 3.1306546035125065, "grad_norm": 0.3028927147388458, "learning_rate": 6.364292916752414e-08, "loss": 0.1963, "step": 11765 }, { "epoch": 3.1309207025013306, "grad_norm": 0.31337544322013855, "learning_rate": 6.362703633948435e-08, "loss": 0.2054, "step": 11766 }, { "epoch": 3.1311868014901543, "grad_norm": 0.29778939485549927, "learning_rate": 6.361114457016713e-08, "loss": 0.1984, "step": 11767 }, { "epoch": 3.131452900478978, "grad_norm": 0.29278481006622314, "learning_rate": 6.359525386003497e-08, "loss": 0.1817, "step": 11768 }, { "epoch": 3.131718999467802, "grad_norm": 0.2793750762939453, "learning_rate": 6.357936420955043e-08, "loss": 0.1719, "step": 11769 }, { "epoch": 3.1319850984566258, "grad_norm": 0.3064999580383301, "learning_rate": 6.356347561917606e-08, "loss": 0.1937, "step": 11770 }, { "epoch": 3.13225119744545, "grad_norm": 0.2985232174396515, "learning_rate": 6.354758808937429e-08, "loss": 0.2044, "step": 11771 }, { "epoch": 3.1325172964342736, "grad_norm": 0.3057096004486084, "learning_rate": 6.35317016206076e-08, "loss": 0.1964, "step": 11772 }, { "epoch": 3.1327833954230973, "grad_norm": 0.29148951172828674, "learning_rate": 6.35158162133384e-08, "loss": 0.1878, "step": 11773 }, { "epoch": 3.1330494944119214, "grad_norm": 0.28012529015541077, "learning_rate": 6.349993186802902e-08, "loss": 0.1589, "step": 11774 }, { "epoch": 3.133315593400745, "grad_norm": 0.2925456464290619, "learning_rate": 6.348404858514186e-08, "loss": 0.2032, "step": 11775 }, { "epoch": 3.1335816923895687, "grad_norm": 0.3202723562717438, "learning_rate": 6.346816636513928e-08, "loss": 0.1878, "step": 11776 }, { "epoch": 3.133847791378393, "grad_norm": 0.2811073362827301, "learning_rate": 6.34522852084835e-08, "loss": 0.189, "step": 11777 }, { "epoch": 3.1341138903672165, "grad_norm": 0.3013997972011566, "learning_rate": 6.343640511563677e-08, "loss": 0.1907, "step": 11778 }, { "epoch": 3.1343799893560407, "grad_norm": 0.3070630729198456, "learning_rate": 6.342052608706137e-08, "loss": 0.1911, "step": 11779 }, { "epoch": 3.1346460883448644, "grad_norm": 0.2812786102294922, "learning_rate": 6.340464812321947e-08, "loss": 0.1872, "step": 11780 }, { "epoch": 3.134912187333688, "grad_norm": 0.30138465762138367, "learning_rate": 6.338877122457324e-08, "loss": 0.1907, "step": 11781 }, { "epoch": 3.135178286322512, "grad_norm": 0.30955952405929565, "learning_rate": 6.337289539158484e-08, "loss": 0.199, "step": 11782 }, { "epoch": 3.135444385311336, "grad_norm": 0.28689903020858765, "learning_rate": 6.335702062471635e-08, "loss": 0.1914, "step": 11783 }, { "epoch": 3.1357104843001595, "grad_norm": 0.39157426357269287, "learning_rate": 6.334114692442984e-08, "loss": 0.1919, "step": 11784 }, { "epoch": 3.1359765832889837, "grad_norm": 0.3112666606903076, "learning_rate": 6.33252742911874e-08, "loss": 0.1937, "step": 11785 }, { "epoch": 3.1362426822778073, "grad_norm": 0.3159341812133789, "learning_rate": 6.330940272545095e-08, "loss": 0.202, "step": 11786 }, { "epoch": 3.136508781266631, "grad_norm": 0.2810916602611542, "learning_rate": 6.329353222768254e-08, "loss": 0.1946, "step": 11787 }, { "epoch": 3.136774880255455, "grad_norm": 0.30401095747947693, "learning_rate": 6.327766279834409e-08, "loss": 0.1882, "step": 11788 }, { "epoch": 3.137040979244279, "grad_norm": 0.3202120363712311, "learning_rate": 6.326179443789753e-08, "loss": 0.2066, "step": 11789 }, { "epoch": 3.1373070782331025, "grad_norm": 0.3060486912727356, "learning_rate": 6.324592714680473e-08, "loss": 0.177, "step": 11790 }, { "epoch": 3.1375731772219266, "grad_norm": 0.29600584506988525, "learning_rate": 6.323006092552757e-08, "loss": 0.1937, "step": 11791 }, { "epoch": 3.1378392762107503, "grad_norm": 0.3052923083305359, "learning_rate": 6.321419577452787e-08, "loss": 0.1907, "step": 11792 }, { "epoch": 3.1381053751995744, "grad_norm": 0.3045814633369446, "learning_rate": 6.319833169426742e-08, "loss": 0.1845, "step": 11793 }, { "epoch": 3.138371474188398, "grad_norm": 0.3028944730758667, "learning_rate": 6.3182468685208e-08, "loss": 0.1936, "step": 11794 }, { "epoch": 3.138637573177222, "grad_norm": 0.29800665378570557, "learning_rate": 6.31666067478113e-08, "loss": 0.1978, "step": 11795 }, { "epoch": 3.138903672166046, "grad_norm": 0.2586672008037567, "learning_rate": 6.315074588253904e-08, "loss": 0.1781, "step": 11796 }, { "epoch": 3.1391697711548696, "grad_norm": 0.299297034740448, "learning_rate": 6.313488608985288e-08, "loss": 0.1876, "step": 11797 }, { "epoch": 3.1394358701436933, "grad_norm": 0.28913629055023193, "learning_rate": 6.311902737021447e-08, "loss": 0.1834, "step": 11798 }, { "epoch": 3.1397019691325174, "grad_norm": 0.3185092806816101, "learning_rate": 6.31031697240854e-08, "loss": 0.2071, "step": 11799 }, { "epoch": 3.139968068121341, "grad_norm": 0.3029920756816864, "learning_rate": 6.308731315192728e-08, "loss": 0.2204, "step": 11800 }, { "epoch": 3.140234167110165, "grad_norm": 0.3007553815841675, "learning_rate": 6.307145765420161e-08, "loss": 0.1845, "step": 11801 }, { "epoch": 3.140500266098989, "grad_norm": 0.3043223023414612, "learning_rate": 6.305560323136993e-08, "loss": 0.2025, "step": 11802 }, { "epoch": 3.1407663650878126, "grad_norm": 0.2915854752063751, "learning_rate": 6.303974988389377e-08, "loss": 0.1961, "step": 11803 }, { "epoch": 3.1410324640766367, "grad_norm": 0.3018874526023865, "learning_rate": 6.302389761223448e-08, "loss": 0.2012, "step": 11804 }, { "epoch": 3.1412985630654604, "grad_norm": 0.28393974900245667, "learning_rate": 6.300804641685351e-08, "loss": 0.185, "step": 11805 }, { "epoch": 3.141564662054284, "grad_norm": 0.32814595103263855, "learning_rate": 6.299219629821229e-08, "loss": 0.2244, "step": 11806 }, { "epoch": 3.141830761043108, "grad_norm": 0.2852753698825836, "learning_rate": 6.297634725677212e-08, "loss": 0.1937, "step": 11807 }, { "epoch": 3.142096860031932, "grad_norm": 0.3019556403160095, "learning_rate": 6.296049929299435e-08, "loss": 0.1888, "step": 11808 }, { "epoch": 3.1423629590207556, "grad_norm": 0.29562291502952576, "learning_rate": 6.294465240734028e-08, "loss": 0.1989, "step": 11809 }, { "epoch": 3.1426290580095797, "grad_norm": 0.2997586727142334, "learning_rate": 6.292880660027116e-08, "loss": 0.1917, "step": 11810 }, { "epoch": 3.1428951569984034, "grad_norm": 0.3034747540950775, "learning_rate": 6.291296187224821e-08, "loss": 0.1754, "step": 11811 }, { "epoch": 3.143161255987227, "grad_norm": 0.3043699860572815, "learning_rate": 6.289711822373272e-08, "loss": 0.187, "step": 11812 }, { "epoch": 3.143427354976051, "grad_norm": 0.2667137384414673, "learning_rate": 6.288127565518571e-08, "loss": 0.1585, "step": 11813 }, { "epoch": 3.143693453964875, "grad_norm": 0.28599920868873596, "learning_rate": 6.28654341670684e-08, "loss": 0.1829, "step": 11814 }, { "epoch": 3.1439595529536986, "grad_norm": 0.3063119053840637, "learning_rate": 6.28495937598419e-08, "loss": 0.1943, "step": 11815 }, { "epoch": 3.1442256519425227, "grad_norm": 0.28865373134613037, "learning_rate": 6.283375443396726e-08, "loss": 0.1667, "step": 11816 }, { "epoch": 3.1444917509313464, "grad_norm": 0.30131980776786804, "learning_rate": 6.28179161899055e-08, "loss": 0.1898, "step": 11817 }, { "epoch": 3.1447578499201705, "grad_norm": 0.2921030521392822, "learning_rate": 6.28020790281177e-08, "loss": 0.2155, "step": 11818 }, { "epoch": 3.145023948908994, "grad_norm": 0.26944616436958313, "learning_rate": 6.278624294906476e-08, "loss": 0.1873, "step": 11819 }, { "epoch": 3.145290047897818, "grad_norm": 0.30459505319595337, "learning_rate": 6.277040795320767e-08, "loss": 0.1771, "step": 11820 }, { "epoch": 3.145556146886642, "grad_norm": 0.29230764508247375, "learning_rate": 6.275457404100738e-08, "loss": 0.2107, "step": 11821 }, { "epoch": 3.1458222458754657, "grad_norm": 0.29728713631629944, "learning_rate": 6.27387412129247e-08, "loss": 0.1958, "step": 11822 }, { "epoch": 3.1460883448642893, "grad_norm": 0.3017849028110504, "learning_rate": 6.272290946942051e-08, "loss": 0.1807, "step": 11823 }, { "epoch": 3.1463544438531135, "grad_norm": 0.30028045177459717, "learning_rate": 6.270707881095563e-08, "loss": 0.1808, "step": 11824 }, { "epoch": 3.146620542841937, "grad_norm": 0.2907055616378784, "learning_rate": 6.269124923799088e-08, "loss": 0.2025, "step": 11825 }, { "epoch": 3.1468866418307613, "grad_norm": 0.3046802878379822, "learning_rate": 6.267542075098698e-08, "loss": 0.2102, "step": 11826 }, { "epoch": 3.147152740819585, "grad_norm": 0.2948245704174042, "learning_rate": 6.265959335040467e-08, "loss": 0.2026, "step": 11827 }, { "epoch": 3.1474188398084086, "grad_norm": 0.30635207891464233, "learning_rate": 6.264376703670463e-08, "loss": 0.1899, "step": 11828 }, { "epoch": 3.1476849387972328, "grad_norm": 0.3181467056274414, "learning_rate": 6.262794181034753e-08, "loss": 0.2105, "step": 11829 }, { "epoch": 3.1479510377860564, "grad_norm": 0.2707045078277588, "learning_rate": 6.261211767179408e-08, "loss": 0.168, "step": 11830 }, { "epoch": 3.14821713677488, "grad_norm": 0.2955644428730011, "learning_rate": 6.259629462150476e-08, "loss": 0.1842, "step": 11831 }, { "epoch": 3.1484832357637043, "grad_norm": 0.2787947356700897, "learning_rate": 6.258047265994018e-08, "loss": 0.1703, "step": 11832 }, { "epoch": 3.148749334752528, "grad_norm": 0.30581316351890564, "learning_rate": 6.256465178756087e-08, "loss": 0.1947, "step": 11833 }, { "epoch": 3.1490154337413516, "grad_norm": 0.30336514115333557, "learning_rate": 6.254883200482737e-08, "loss": 0.2006, "step": 11834 }, { "epoch": 3.1492815327301757, "grad_norm": 0.27432531118392944, "learning_rate": 6.253301331220013e-08, "loss": 0.1936, "step": 11835 }, { "epoch": 3.1495476317189994, "grad_norm": 0.2902505397796631, "learning_rate": 6.251719571013958e-08, "loss": 0.1819, "step": 11836 }, { "epoch": 3.149813730707823, "grad_norm": 0.30165383219718933, "learning_rate": 6.250137919910613e-08, "loss": 0.1932, "step": 11837 }, { "epoch": 3.1500798296966472, "grad_norm": 0.40898919105529785, "learning_rate": 6.248556377956018e-08, "loss": 0.1762, "step": 11838 }, { "epoch": 3.150345928685471, "grad_norm": 0.2803047299385071, "learning_rate": 6.246974945196208e-08, "loss": 0.1843, "step": 11839 }, { "epoch": 3.1506120276742946, "grad_norm": 0.2865629196166992, "learning_rate": 6.245393621677211e-08, "loss": 0.1873, "step": 11840 }, { "epoch": 3.1508781266631187, "grad_norm": 0.3091674745082855, "learning_rate": 6.243812407445058e-08, "loss": 0.2152, "step": 11841 }, { "epoch": 3.1511442256519424, "grad_norm": 0.3014421761035919, "learning_rate": 6.24223130254577e-08, "loss": 0.1985, "step": 11842 }, { "epoch": 3.1514103246407665, "grad_norm": 0.2881520092487335, "learning_rate": 6.240650307025373e-08, "loss": 0.199, "step": 11843 }, { "epoch": 3.15167642362959, "grad_norm": 0.3075462281703949, "learning_rate": 6.239069420929886e-08, "loss": 0.1948, "step": 11844 }, { "epoch": 3.151942522618414, "grad_norm": 0.2797428369522095, "learning_rate": 6.237488644305322e-08, "loss": 0.1717, "step": 11845 }, { "epoch": 3.152208621607238, "grad_norm": 0.2910405993461609, "learning_rate": 6.235907977197694e-08, "loss": 0.1857, "step": 11846 }, { "epoch": 3.1524747205960617, "grad_norm": 0.3004874587059021, "learning_rate": 6.234327419653011e-08, "loss": 0.2, "step": 11847 }, { "epoch": 3.1527408195848854, "grad_norm": 0.2632330656051636, "learning_rate": 6.232746971717284e-08, "loss": 0.1691, "step": 11848 }, { "epoch": 3.1530069185737095, "grad_norm": 0.307327002286911, "learning_rate": 6.231166633436507e-08, "loss": 0.191, "step": 11849 }, { "epoch": 3.153273017562533, "grad_norm": 0.28662168979644775, "learning_rate": 6.229586404856683e-08, "loss": 0.184, "step": 11850 }, { "epoch": 3.1535391165513573, "grad_norm": 0.285532146692276, "learning_rate": 6.22800628602381e-08, "loss": 0.1851, "step": 11851 }, { "epoch": 3.153805215540181, "grad_norm": 0.286032497882843, "learning_rate": 6.226426276983879e-08, "loss": 0.1777, "step": 11852 }, { "epoch": 3.1540713145290047, "grad_norm": 0.7915868759155273, "learning_rate": 6.224846377782882e-08, "loss": 0.1838, "step": 11853 }, { "epoch": 3.154337413517829, "grad_norm": 0.314007431268692, "learning_rate": 6.223266588466805e-08, "loss": 0.2122, "step": 11854 }, { "epoch": 3.1546035125066525, "grad_norm": 0.2919567823410034, "learning_rate": 6.221686909081633e-08, "loss": 0.1884, "step": 11855 }, { "epoch": 3.154869611495476, "grad_norm": 0.30124643445014954, "learning_rate": 6.220107339673341e-08, "loss": 0.199, "step": 11856 }, { "epoch": 3.1551357104843003, "grad_norm": 0.33200880885124207, "learning_rate": 6.218527880287914e-08, "loss": 0.1898, "step": 11857 }, { "epoch": 3.155401809473124, "grad_norm": 0.3282276690006256, "learning_rate": 6.216948530971322e-08, "loss": 0.208, "step": 11858 }, { "epoch": 3.1556679084619477, "grad_norm": 0.2951206564903259, "learning_rate": 6.215369291769534e-08, "loss": 0.1822, "step": 11859 }, { "epoch": 3.155934007450772, "grad_norm": 0.26787951588630676, "learning_rate": 6.213790162728519e-08, "loss": 0.1754, "step": 11860 }, { "epoch": 3.1562001064395955, "grad_norm": 0.31006181240081787, "learning_rate": 6.21221114389424e-08, "loss": 0.2147, "step": 11861 }, { "epoch": 3.156466205428419, "grad_norm": 0.28002744913101196, "learning_rate": 6.210632235312661e-08, "loss": 0.1728, "step": 11862 }, { "epoch": 3.1567323044172433, "grad_norm": 0.30892786383628845, "learning_rate": 6.209053437029738e-08, "loss": 0.2036, "step": 11863 }, { "epoch": 3.156998403406067, "grad_norm": 0.28328797221183777, "learning_rate": 6.207474749091428e-08, "loss": 0.1875, "step": 11864 }, { "epoch": 3.157264502394891, "grad_norm": 0.37682098150253296, "learning_rate": 6.205896171543681e-08, "loss": 0.192, "step": 11865 }, { "epoch": 3.1575306013837148, "grad_norm": 0.2994341254234314, "learning_rate": 6.204317704432444e-08, "loss": 0.1862, "step": 11866 }, { "epoch": 3.1577967003725385, "grad_norm": 0.29280856251716614, "learning_rate": 6.202739347803666e-08, "loss": 0.1927, "step": 11867 }, { "epoch": 3.1580627993613626, "grad_norm": 0.2938173711299896, "learning_rate": 6.201161101703283e-08, "loss": 0.1956, "step": 11868 }, { "epoch": 3.1583288983501863, "grad_norm": 0.314243346452713, "learning_rate": 6.199582966177237e-08, "loss": 0.2017, "step": 11869 }, { "epoch": 3.15859499733901, "grad_norm": 0.2902550995349884, "learning_rate": 6.198004941271462e-08, "loss": 0.1757, "step": 11870 }, { "epoch": 3.158861096327834, "grad_norm": 0.2834068834781647, "learning_rate": 6.196427027031893e-08, "loss": 0.1922, "step": 11871 }, { "epoch": 3.1591271953166578, "grad_norm": 0.2989647090435028, "learning_rate": 6.194849223504457e-08, "loss": 0.1741, "step": 11872 }, { "epoch": 3.1593932943054814, "grad_norm": 0.2873101234436035, "learning_rate": 6.19327153073508e-08, "loss": 0.185, "step": 11873 }, { "epoch": 3.1596593932943056, "grad_norm": 0.32435786724090576, "learning_rate": 6.191693948769686e-08, "loss": 0.218, "step": 11874 }, { "epoch": 3.1599254922831292, "grad_norm": 0.28791379928588867, "learning_rate": 6.190116477654193e-08, "loss": 0.1751, "step": 11875 }, { "epoch": 3.1601915912719534, "grad_norm": 0.28984200954437256, "learning_rate": 6.188539117434515e-08, "loss": 0.1945, "step": 11876 }, { "epoch": 3.160457690260777, "grad_norm": 0.2979494631290436, "learning_rate": 6.186961868156571e-08, "loss": 0.2037, "step": 11877 }, { "epoch": 3.1607237892496007, "grad_norm": 0.2838287949562073, "learning_rate": 6.185384729866263e-08, "loss": 0.1752, "step": 11878 }, { "epoch": 3.160989888238425, "grad_norm": 0.3132784068584442, "learning_rate": 6.183807702609501e-08, "loss": 0.1795, "step": 11879 }, { "epoch": 3.1612559872272485, "grad_norm": 0.2820243835449219, "learning_rate": 6.182230786432188e-08, "loss": 0.1821, "step": 11880 }, { "epoch": 3.1615220862160722, "grad_norm": 0.27300402522087097, "learning_rate": 6.180653981380224e-08, "loss": 0.1781, "step": 11881 }, { "epoch": 3.1617881852048964, "grad_norm": 0.26328805088996887, "learning_rate": 6.179077287499506e-08, "loss": 0.1688, "step": 11882 }, { "epoch": 3.16205428419372, "grad_norm": 0.3104816973209381, "learning_rate": 6.177500704835928e-08, "loss": 0.2204, "step": 11883 }, { "epoch": 3.1623203831825437, "grad_norm": 0.298397421836853, "learning_rate": 6.175924233435381e-08, "loss": 0.1734, "step": 11884 }, { "epoch": 3.162586482171368, "grad_norm": 0.29672905802726746, "learning_rate": 6.174347873343748e-08, "loss": 0.2119, "step": 11885 }, { "epoch": 3.1628525811601915, "grad_norm": 0.2816697657108307, "learning_rate": 6.172771624606916e-08, "loss": 0.1837, "step": 11886 }, { "epoch": 3.163118680149015, "grad_norm": 0.29346561431884766, "learning_rate": 6.171195487270766e-08, "loss": 0.1811, "step": 11887 }, { "epoch": 3.1633847791378393, "grad_norm": 0.2736819088459015, "learning_rate": 6.169619461381173e-08, "loss": 0.1872, "step": 11888 }, { "epoch": 3.163650878126663, "grad_norm": 0.3002624809741974, "learning_rate": 6.16804354698401e-08, "loss": 0.1992, "step": 11889 }, { "epoch": 3.163916977115487, "grad_norm": 0.2846660315990448, "learning_rate": 6.166467744125151e-08, "loss": 0.1869, "step": 11890 }, { "epoch": 3.164183076104311, "grad_norm": 0.29328468441963196, "learning_rate": 6.164892052850462e-08, "loss": 0.1946, "step": 11891 }, { "epoch": 3.1644491750931345, "grad_norm": 0.27486109733581543, "learning_rate": 6.163316473205809e-08, "loss": 0.1824, "step": 11892 }, { "epoch": 3.1647152740819586, "grad_norm": 0.28875797986984253, "learning_rate": 6.161741005237053e-08, "loss": 0.1935, "step": 11893 }, { "epoch": 3.1649813730707823, "grad_norm": 0.31124553084373474, "learning_rate": 6.160165648990048e-08, "loss": 0.1832, "step": 11894 }, { "epoch": 3.165247472059606, "grad_norm": 0.2852957248687744, "learning_rate": 6.15859040451065e-08, "loss": 0.1914, "step": 11895 }, { "epoch": 3.16551357104843, "grad_norm": 0.3006126582622528, "learning_rate": 6.157015271844712e-08, "loss": 0.173, "step": 11896 }, { "epoch": 3.165779670037254, "grad_norm": 0.2914561927318573, "learning_rate": 6.155440251038083e-08, "loss": 0.2009, "step": 11897 }, { "epoch": 3.166045769026078, "grad_norm": 0.29299864172935486, "learning_rate": 6.153865342136603e-08, "loss": 0.1871, "step": 11898 }, { "epoch": 3.1663118680149016, "grad_norm": 0.32471221685409546, "learning_rate": 6.152290545186119e-08, "loss": 0.1829, "step": 11899 }, { "epoch": 3.1665779670037253, "grad_norm": 0.2996750771999359, "learning_rate": 6.150715860232465e-08, "loss": 0.1989, "step": 11900 }, { "epoch": 3.1668440659925494, "grad_norm": 0.30565696954727173, "learning_rate": 6.149141287321478e-08, "loss": 0.1875, "step": 11901 }, { "epoch": 3.167110164981373, "grad_norm": 0.2685380280017853, "learning_rate": 6.147566826498993e-08, "loss": 0.156, "step": 11902 }, { "epoch": 3.167376263970197, "grad_norm": 0.30956268310546875, "learning_rate": 6.145992477810833e-08, "loss": 0.1903, "step": 11903 }, { "epoch": 3.167642362959021, "grad_norm": 0.2987746298313141, "learning_rate": 6.144418241302822e-08, "loss": 0.2057, "step": 11904 }, { "epoch": 3.1679084619478446, "grad_norm": 0.32085010409355164, "learning_rate": 6.142844117020787e-08, "loss": 0.2083, "step": 11905 }, { "epoch": 3.1681745609366683, "grad_norm": 0.29966267943382263, "learning_rate": 6.141270105010545e-08, "loss": 0.2, "step": 11906 }, { "epoch": 3.1684406599254924, "grad_norm": 0.32665133476257324, "learning_rate": 6.139696205317912e-08, "loss": 0.201, "step": 11907 }, { "epoch": 3.168706758914316, "grad_norm": 0.2966950237751007, "learning_rate": 6.138122417988699e-08, "loss": 0.1867, "step": 11908 }, { "epoch": 3.1689728579031398, "grad_norm": 0.285781592130661, "learning_rate": 6.136548743068712e-08, "loss": 0.1933, "step": 11909 }, { "epoch": 3.169238956891964, "grad_norm": 0.2901575565338135, "learning_rate": 6.134975180603762e-08, "loss": 0.1889, "step": 11910 }, { "epoch": 3.1695050558807876, "grad_norm": 0.3144608438014984, "learning_rate": 6.133401730639653e-08, "loss": 0.1986, "step": 11911 }, { "epoch": 3.1697711548696117, "grad_norm": 0.2678353488445282, "learning_rate": 6.131828393222177e-08, "loss": 0.1667, "step": 11912 }, { "epoch": 3.1700372538584354, "grad_norm": 0.3052350878715515, "learning_rate": 6.130255168397131e-08, "loss": 0.2009, "step": 11913 }, { "epoch": 3.170303352847259, "grad_norm": 0.2962004244327545, "learning_rate": 6.12868205621031e-08, "loss": 0.1994, "step": 11914 }, { "epoch": 3.170569451836083, "grad_norm": 0.2972472310066223, "learning_rate": 6.127109056707503e-08, "loss": 0.2052, "step": 11915 }, { "epoch": 3.170835550824907, "grad_norm": 0.2990664839744568, "learning_rate": 6.125536169934497e-08, "loss": 0.1827, "step": 11916 }, { "epoch": 3.1711016498137305, "grad_norm": 0.3060995042324066, "learning_rate": 6.123963395937072e-08, "loss": 0.1901, "step": 11917 }, { "epoch": 3.1713677488025547, "grad_norm": 0.29385611414909363, "learning_rate": 6.122390734761007e-08, "loss": 0.197, "step": 11918 }, { "epoch": 3.1716338477913784, "grad_norm": 0.31888291239738464, "learning_rate": 6.120818186452081e-08, "loss": 0.186, "step": 11919 }, { "epoch": 3.171899946780202, "grad_norm": 0.3021220266819, "learning_rate": 6.119245751056069e-08, "loss": 0.1883, "step": 11920 }, { "epoch": 3.172166045769026, "grad_norm": 0.29984036087989807, "learning_rate": 6.117673428618732e-08, "loss": 0.1968, "step": 11921 }, { "epoch": 3.17243214475785, "grad_norm": 0.2968994677066803, "learning_rate": 6.116101219185843e-08, "loss": 0.1789, "step": 11922 }, { "epoch": 3.172698243746674, "grad_norm": 0.2878416180610657, "learning_rate": 6.114529122803161e-08, "loss": 0.1915, "step": 11923 }, { "epoch": 3.1729643427354977, "grad_norm": 0.35591790080070496, "learning_rate": 6.112957139516449e-08, "loss": 0.1934, "step": 11924 }, { "epoch": 3.1732304417243213, "grad_norm": 0.30383750796318054, "learning_rate": 6.111385269371462e-08, "loss": 0.1945, "step": 11925 }, { "epoch": 3.1734965407131455, "grad_norm": 0.3329671025276184, "learning_rate": 6.109813512413954e-08, "loss": 0.1789, "step": 11926 }, { "epoch": 3.173762639701969, "grad_norm": 0.30971938371658325, "learning_rate": 6.108241868689675e-08, "loss": 0.1991, "step": 11927 }, { "epoch": 3.174028738690793, "grad_norm": 0.2902931272983551, "learning_rate": 6.106670338244367e-08, "loss": 0.1904, "step": 11928 }, { "epoch": 3.174294837679617, "grad_norm": 0.3098791241645813, "learning_rate": 6.105098921123781e-08, "loss": 0.1949, "step": 11929 }, { "epoch": 3.1745609366684406, "grad_norm": 0.28706687688827515, "learning_rate": 6.103527617373649e-08, "loss": 0.2084, "step": 11930 }, { "epoch": 3.1748270356572643, "grad_norm": 0.2750474214553833, "learning_rate": 6.101956427039712e-08, "loss": 0.1746, "step": 11931 }, { "epoch": 3.1750931346460884, "grad_norm": 0.275570273399353, "learning_rate": 6.1003853501677e-08, "loss": 0.1737, "step": 11932 }, { "epoch": 3.175359233634912, "grad_norm": 0.3022004961967468, "learning_rate": 6.098814386803347e-08, "loss": 0.1879, "step": 11933 }, { "epoch": 3.175625332623736, "grad_norm": 0.29239508509635925, "learning_rate": 6.097243536992377e-08, "loss": 0.1932, "step": 11934 }, { "epoch": 3.17589143161256, "grad_norm": 0.27258315682411194, "learning_rate": 6.095672800780514e-08, "loss": 0.1774, "step": 11935 }, { "epoch": 3.1761575306013836, "grad_norm": 0.28591811656951904, "learning_rate": 6.094102178213482e-08, "loss": 0.2122, "step": 11936 }, { "epoch": 3.1764236295902077, "grad_norm": 0.3077285885810852, "learning_rate": 6.092531669336991e-08, "loss": 0.2013, "step": 11937 }, { "epoch": 3.1766897285790314, "grad_norm": 0.2929794490337372, "learning_rate": 6.090961274196759e-08, "loss": 0.1971, "step": 11938 }, { "epoch": 3.176955827567855, "grad_norm": 0.2997462749481201, "learning_rate": 6.089390992838495e-08, "loss": 0.1787, "step": 11939 }, { "epoch": 3.1772219265566792, "grad_norm": 0.3378627598285675, "learning_rate": 6.087820825307903e-08, "loss": 0.1979, "step": 11940 }, { "epoch": 3.177488025545503, "grad_norm": 0.2900421917438507, "learning_rate": 6.08625077165069e-08, "loss": 0.1831, "step": 11941 }, { "epoch": 3.1777541245343266, "grad_norm": 0.2812674939632416, "learning_rate": 6.084680831912555e-08, "loss": 0.1925, "step": 11942 }, { "epoch": 3.1780202235231507, "grad_norm": 0.28384774923324585, "learning_rate": 6.083111006139191e-08, "loss": 0.1812, "step": 11943 }, { "epoch": 3.1782863225119744, "grad_norm": 0.3339642882347107, "learning_rate": 6.0815412943763e-08, "loss": 0.2042, "step": 11944 }, { "epoch": 3.1785524215007985, "grad_norm": 0.3106270134449005, "learning_rate": 6.079971696669566e-08, "loss": 0.2027, "step": 11945 }, { "epoch": 3.178818520489622, "grad_norm": 0.32726842164993286, "learning_rate": 6.078402213064682e-08, "loss": 0.1963, "step": 11946 }, { "epoch": 3.179084619478446, "grad_norm": 0.3016764521598816, "learning_rate": 6.076832843607325e-08, "loss": 0.1985, "step": 11947 }, { "epoch": 3.17935071846727, "grad_norm": 0.2818711996078491, "learning_rate": 6.075263588343175e-08, "loss": 0.1808, "step": 11948 }, { "epoch": 3.1796168174560937, "grad_norm": 0.2980591058731079, "learning_rate": 6.073694447317916e-08, "loss": 0.1983, "step": 11949 }, { "epoch": 3.1798829164449174, "grad_norm": 0.3030446171760559, "learning_rate": 6.072125420577212e-08, "loss": 0.1919, "step": 11950 }, { "epoch": 3.1801490154337415, "grad_norm": 0.26106423139572144, "learning_rate": 6.07055650816674e-08, "loss": 0.1747, "step": 11951 }, { "epoch": 3.180415114422565, "grad_norm": 0.30499520897865295, "learning_rate": 6.068987710132166e-08, "loss": 0.1839, "step": 11952 }, { "epoch": 3.180681213411389, "grad_norm": 0.28741610050201416, "learning_rate": 6.067419026519152e-08, "loss": 0.168, "step": 11953 }, { "epoch": 3.180947312400213, "grad_norm": 0.47125744819641113, "learning_rate": 6.06585045737336e-08, "loss": 0.1904, "step": 11954 }, { "epoch": 3.1812134113890367, "grad_norm": 0.2998725175857544, "learning_rate": 6.064282002740445e-08, "loss": 0.1923, "step": 11955 }, { "epoch": 3.1814795103778604, "grad_norm": 0.29942432045936584, "learning_rate": 6.062713662666067e-08, "loss": 0.2016, "step": 11956 }, { "epoch": 3.1817456093666845, "grad_norm": 0.2775174379348755, "learning_rate": 6.061145437195867e-08, "loss": 0.1767, "step": 11957 }, { "epoch": 3.182011708355508, "grad_norm": 0.29673102498054504, "learning_rate": 6.059577326375497e-08, "loss": 0.1871, "step": 11958 }, { "epoch": 3.1822778073443323, "grad_norm": 0.31487661600112915, "learning_rate": 6.0580093302506e-08, "loss": 0.1821, "step": 11959 }, { "epoch": 3.182543906333156, "grad_norm": 0.2866194546222687, "learning_rate": 6.056441448866816e-08, "loss": 0.1839, "step": 11960 }, { "epoch": 3.1828100053219797, "grad_norm": 0.27188557386398315, "learning_rate": 6.054873682269781e-08, "loss": 0.1816, "step": 11961 }, { "epoch": 3.183076104310804, "grad_norm": 0.2892438769340515, "learning_rate": 6.05330603050513e-08, "loss": 0.1838, "step": 11962 }, { "epoch": 3.1833422032996275, "grad_norm": 0.275595098733902, "learning_rate": 6.051738493618493e-08, "loss": 0.188, "step": 11963 }, { "epoch": 3.183608302288451, "grad_norm": 0.3468238115310669, "learning_rate": 6.050171071655495e-08, "loss": 0.1743, "step": 11964 }, { "epoch": 3.1838744012772753, "grad_norm": 0.27922728657722473, "learning_rate": 6.048603764661768e-08, "loss": 0.1885, "step": 11965 }, { "epoch": 3.184140500266099, "grad_norm": 0.324724942445755, "learning_rate": 6.04703657268292e-08, "loss": 0.1893, "step": 11966 }, { "epoch": 3.1844065992549226, "grad_norm": 0.2951280474662781, "learning_rate": 6.045469495764575e-08, "loss": 0.199, "step": 11967 }, { "epoch": 3.1846726982437468, "grad_norm": 0.3017224967479706, "learning_rate": 6.043902533952347e-08, "loss": 0.1955, "step": 11968 }, { "epoch": 3.1849387972325705, "grad_norm": 0.28603628277778625, "learning_rate": 6.042335687291841e-08, "loss": 0.1713, "step": 11969 }, { "epoch": 3.1852048962213946, "grad_norm": 0.30795854330062866, "learning_rate": 6.040768955828669e-08, "loss": 0.1876, "step": 11970 }, { "epoch": 3.1854709952102183, "grad_norm": 0.28537890315055847, "learning_rate": 6.039202339608431e-08, "loss": 0.1906, "step": 11971 }, { "epoch": 3.185737094199042, "grad_norm": 0.28749629855155945, "learning_rate": 6.03763583867673e-08, "loss": 0.191, "step": 11972 }, { "epoch": 3.186003193187866, "grad_norm": 0.2940794825553894, "learning_rate": 6.036069453079162e-08, "loss": 0.2034, "step": 11973 }, { "epoch": 3.1862692921766897, "grad_norm": 0.26739615201950073, "learning_rate": 6.034503182861322e-08, "loss": 0.1868, "step": 11974 }, { "epoch": 3.1865353911655134, "grad_norm": 0.28333213925361633, "learning_rate": 6.032937028068797e-08, "loss": 0.2014, "step": 11975 }, { "epoch": 3.1868014901543376, "grad_norm": 0.2877621352672577, "learning_rate": 6.031370988747171e-08, "loss": 0.1798, "step": 11976 }, { "epoch": 3.1870675891431612, "grad_norm": 0.31608083844184875, "learning_rate": 6.029805064942033e-08, "loss": 0.2107, "step": 11977 }, { "epoch": 3.187333688131985, "grad_norm": 0.30937373638153076, "learning_rate": 6.028239256698964e-08, "loss": 0.2109, "step": 11978 }, { "epoch": 3.187599787120809, "grad_norm": 0.3066503405570984, "learning_rate": 6.026673564063534e-08, "loss": 0.1859, "step": 11979 }, { "epoch": 3.1878658861096327, "grad_norm": 0.2953994572162628, "learning_rate": 6.025107987081321e-08, "loss": 0.1761, "step": 11980 }, { "epoch": 3.1881319850984564, "grad_norm": 0.29202592372894287, "learning_rate": 6.023542525797894e-08, "loss": 0.1908, "step": 11981 }, { "epoch": 3.1883980840872805, "grad_norm": 0.2964071035385132, "learning_rate": 6.021977180258817e-08, "loss": 0.1823, "step": 11982 }, { "epoch": 3.188664183076104, "grad_norm": 0.3083723485469818, "learning_rate": 6.020411950509663e-08, "loss": 0.2042, "step": 11983 }, { "epoch": 3.1889302820649283, "grad_norm": 0.2885955572128296, "learning_rate": 6.018846836595978e-08, "loss": 0.1942, "step": 11984 }, { "epoch": 3.189196381053752, "grad_norm": 0.3195359408855438, "learning_rate": 6.017281838563324e-08, "loss": 0.2116, "step": 11985 }, { "epoch": 3.1894624800425757, "grad_norm": 0.2866694927215576, "learning_rate": 6.015716956457256e-08, "loss": 0.17, "step": 11986 }, { "epoch": 3.1897285790314, "grad_norm": 0.27705782651901245, "learning_rate": 6.014152190323323e-08, "loss": 0.188, "step": 11987 }, { "epoch": 3.1899946780202235, "grad_norm": 0.3081243932247162, "learning_rate": 6.01258754020707e-08, "loss": 0.1843, "step": 11988 }, { "epoch": 3.190260777009047, "grad_norm": 0.3056761920452118, "learning_rate": 6.011023006154041e-08, "loss": 0.177, "step": 11989 }, { "epoch": 3.1905268759978713, "grad_norm": 0.317143976688385, "learning_rate": 6.009458588209774e-08, "loss": 0.2099, "step": 11990 }, { "epoch": 3.190792974986695, "grad_norm": 0.282942533493042, "learning_rate": 6.007894286419806e-08, "loss": 0.2052, "step": 11991 }, { "epoch": 3.1910590739755187, "grad_norm": 0.28183984756469727, "learning_rate": 6.006330100829675e-08, "loss": 0.1817, "step": 11992 }, { "epoch": 3.191325172964343, "grad_norm": 0.296623557806015, "learning_rate": 6.0047660314849e-08, "loss": 0.1848, "step": 11993 }, { "epoch": 3.1915912719531665, "grad_norm": 0.2716952860355377, "learning_rate": 6.003202078431013e-08, "loss": 0.1693, "step": 11994 }, { "epoch": 3.1918573709419906, "grad_norm": 0.30164533853530884, "learning_rate": 6.001638241713535e-08, "loss": 0.2067, "step": 11995 }, { "epoch": 3.1921234699308143, "grad_norm": 0.29706040024757385, "learning_rate": 6.000074521377986e-08, "loss": 0.2053, "step": 11996 }, { "epoch": 3.192389568919638, "grad_norm": 0.2879098653793335, "learning_rate": 5.998510917469883e-08, "loss": 0.1966, "step": 11997 }, { "epoch": 3.192655667908462, "grad_norm": 0.2787838876247406, "learning_rate": 5.99694743003474e-08, "loss": 0.1896, "step": 11998 }, { "epoch": 3.192921766897286, "grad_norm": 0.3216349482536316, "learning_rate": 5.995384059118061e-08, "loss": 0.2124, "step": 11999 }, { "epoch": 3.1931878658861095, "grad_norm": 0.318841814994812, "learning_rate": 5.993820804765352e-08, "loss": 0.2092, "step": 12000 }, { "epoch": 3.1934539648749336, "grad_norm": 0.3054012358188629, "learning_rate": 5.992257667022125e-08, "loss": 0.1934, "step": 12001 }, { "epoch": 3.1937200638637573, "grad_norm": 0.2758338749408722, "learning_rate": 5.990694645933865e-08, "loss": 0.1908, "step": 12002 }, { "epoch": 3.193986162852581, "grad_norm": 0.29398834705352783, "learning_rate": 5.989131741546074e-08, "loss": 0.1882, "step": 12003 }, { "epoch": 3.194252261841405, "grad_norm": 0.29354438185691833, "learning_rate": 5.987568953904245e-08, "loss": 0.2023, "step": 12004 }, { "epoch": 3.1945183608302288, "grad_norm": 0.27943626046180725, "learning_rate": 5.986006283053864e-08, "loss": 0.1962, "step": 12005 }, { "epoch": 3.1947844598190525, "grad_norm": 0.3104410469532013, "learning_rate": 5.98444372904042e-08, "loss": 0.2174, "step": 12006 }, { "epoch": 3.1950505588078766, "grad_norm": 0.3046663999557495, "learning_rate": 5.982881291909392e-08, "loss": 0.2076, "step": 12007 }, { "epoch": 3.1953166577967003, "grad_norm": 0.2953067421913147, "learning_rate": 5.981318971706259e-08, "loss": 0.1927, "step": 12008 }, { "epoch": 3.1955827567855244, "grad_norm": 0.29151415824890137, "learning_rate": 5.979756768476497e-08, "loss": 0.2, "step": 12009 }, { "epoch": 3.195848855774348, "grad_norm": 0.3170233964920044, "learning_rate": 5.97819468226558e-08, "loss": 0.1988, "step": 12010 }, { "epoch": 3.1961149547631718, "grad_norm": 0.2760477662086487, "learning_rate": 5.97663271311897e-08, "loss": 0.1898, "step": 12011 }, { "epoch": 3.196381053751996, "grad_norm": 0.2798711955547333, "learning_rate": 5.975070861082135e-08, "loss": 0.1832, "step": 12012 }, { "epoch": 3.1966471527408196, "grad_norm": 0.3026156425476074, "learning_rate": 5.973509126200535e-08, "loss": 0.2108, "step": 12013 }, { "epoch": 3.1969132517296432, "grad_norm": 0.2909804880619049, "learning_rate": 5.97194750851963e-08, "loss": 0.2061, "step": 12014 }, { "epoch": 3.1971793507184674, "grad_norm": 0.2959434688091278, "learning_rate": 5.970386008084875e-08, "loss": 0.1935, "step": 12015 }, { "epoch": 3.197445449707291, "grad_norm": 0.2825208306312561, "learning_rate": 5.968824624941722e-08, "loss": 0.1905, "step": 12016 }, { "epoch": 3.197711548696115, "grad_norm": 0.3243113160133362, "learning_rate": 5.967263359135617e-08, "loss": 0.2048, "step": 12017 }, { "epoch": 3.197977647684939, "grad_norm": 0.3010031282901764, "learning_rate": 5.965702210712005e-08, "loss": 0.1979, "step": 12018 }, { "epoch": 3.1982437466737625, "grad_norm": 0.27379053831100464, "learning_rate": 5.964141179716324e-08, "loss": 0.1831, "step": 12019 }, { "epoch": 3.1985098456625867, "grad_norm": 0.28952422738075256, "learning_rate": 5.962580266194021e-08, "loss": 0.199, "step": 12020 }, { "epoch": 3.1987759446514104, "grad_norm": 0.31032681465148926, "learning_rate": 5.961019470190519e-08, "loss": 0.1915, "step": 12021 }, { "epoch": 3.199042043640234, "grad_norm": 0.31195756793022156, "learning_rate": 5.9594587917512503e-08, "loss": 0.2118, "step": 12022 }, { "epoch": 3.199308142629058, "grad_norm": 0.308694988489151, "learning_rate": 5.957898230921647e-08, "loss": 0.1803, "step": 12023 }, { "epoch": 3.199574241617882, "grad_norm": 0.283890962600708, "learning_rate": 5.956337787747131e-08, "loss": 0.1859, "step": 12024 }, { "epoch": 3.1998403406067055, "grad_norm": 0.27572259306907654, "learning_rate": 5.954777462273123e-08, "loss": 0.1804, "step": 12025 }, { "epoch": 3.2001064395955297, "grad_norm": 0.29351770877838135, "learning_rate": 5.953217254545039e-08, "loss": 0.1825, "step": 12026 }, { "epoch": 3.2003725385843533, "grad_norm": 0.30833253264427185, "learning_rate": 5.951657164608296e-08, "loss": 0.2016, "step": 12027 }, { "epoch": 3.200638637573177, "grad_norm": 0.31500035524368286, "learning_rate": 5.950097192508299e-08, "loss": 0.1932, "step": 12028 }, { "epoch": 3.200904736562001, "grad_norm": 0.3381083905696869, "learning_rate": 5.948537338290459e-08, "loss": 0.1722, "step": 12029 }, { "epoch": 3.201170835550825, "grad_norm": 0.30529502034187317, "learning_rate": 5.946977602000178e-08, "loss": 0.1844, "step": 12030 }, { "epoch": 3.201436934539649, "grad_norm": 0.30048155784606934, "learning_rate": 5.945417983682854e-08, "loss": 0.2152, "step": 12031 }, { "epoch": 3.2017030335284726, "grad_norm": 0.29382559657096863, "learning_rate": 5.9438584833838834e-08, "loss": 0.2026, "step": 12032 }, { "epoch": 3.2019691325172963, "grad_norm": 0.33758649230003357, "learning_rate": 5.942299101148662e-08, "loss": 0.2205, "step": 12033 }, { "epoch": 3.2022352315061204, "grad_norm": 0.2709081768989563, "learning_rate": 5.940739837022578e-08, "loss": 0.1774, "step": 12034 }, { "epoch": 3.202501330494944, "grad_norm": 0.2827433943748474, "learning_rate": 5.9391806910510176e-08, "loss": 0.1642, "step": 12035 }, { "epoch": 3.202767429483768, "grad_norm": 0.29823437333106995, "learning_rate": 5.937621663279363e-08, "loss": 0.2052, "step": 12036 }, { "epoch": 3.203033528472592, "grad_norm": 0.31732144951820374, "learning_rate": 5.936062753752996e-08, "loss": 0.1769, "step": 12037 }, { "epoch": 3.2032996274614156, "grad_norm": 0.2868728041648865, "learning_rate": 5.93450396251729e-08, "loss": 0.208, "step": 12038 }, { "epoch": 3.2035657264502393, "grad_norm": 0.29586681723594666, "learning_rate": 5.932945289617617e-08, "loss": 0.191, "step": 12039 }, { "epoch": 3.2038318254390634, "grad_norm": 0.3011143207550049, "learning_rate": 5.931386735099347e-08, "loss": 0.1813, "step": 12040 }, { "epoch": 3.204097924427887, "grad_norm": 0.27859416604042053, "learning_rate": 5.929828299007844e-08, "loss": 0.1951, "step": 12041 }, { "epoch": 3.2043640234167112, "grad_norm": 0.30765315890312195, "learning_rate": 5.9282699813884705e-08, "loss": 0.2079, "step": 12042 }, { "epoch": 3.204630122405535, "grad_norm": 0.3305794596672058, "learning_rate": 5.9267117822865865e-08, "loss": 0.1922, "step": 12043 }, { "epoch": 3.2048962213943586, "grad_norm": 0.27684471011161804, "learning_rate": 5.925153701747546e-08, "loss": 0.1844, "step": 12044 }, { "epoch": 3.2051623203831827, "grad_norm": 0.3117124140262604, "learning_rate": 5.9235957398167006e-08, "loss": 0.1806, "step": 12045 }, { "epoch": 3.2054284193720064, "grad_norm": 0.3079410493373871, "learning_rate": 5.9220378965394e-08, "loss": 0.1779, "step": 12046 }, { "epoch": 3.20569451836083, "grad_norm": 0.2968716323375702, "learning_rate": 5.920480171960991e-08, "loss": 0.1857, "step": 12047 }, { "epoch": 3.205960617349654, "grad_norm": 0.29786112904548645, "learning_rate": 5.918922566126807e-08, "loss": 0.1995, "step": 12048 }, { "epoch": 3.206226716338478, "grad_norm": 0.3425429165363312, "learning_rate": 5.917365079082193e-08, "loss": 0.218, "step": 12049 }, { "epoch": 3.2064928153273016, "grad_norm": 0.28374868631362915, "learning_rate": 5.915807710872481e-08, "loss": 0.1739, "step": 12050 }, { "epoch": 3.2067589143161257, "grad_norm": 0.28754204511642456, "learning_rate": 5.914250461543001e-08, "loss": 0.1897, "step": 12051 }, { "epoch": 3.2070250133049494, "grad_norm": 0.3204438388347626, "learning_rate": 5.912693331139082e-08, "loss": 0.2035, "step": 12052 }, { "epoch": 3.207291112293773, "grad_norm": 0.2985529601573944, "learning_rate": 5.9111363197060474e-08, "loss": 0.183, "step": 12053 }, { "epoch": 3.207557211282597, "grad_norm": 0.3576003909111023, "learning_rate": 5.9095794272892174e-08, "loss": 0.1941, "step": 12054 }, { "epoch": 3.207823310271421, "grad_norm": 0.2775062620639801, "learning_rate": 5.908022653933909e-08, "loss": 0.1824, "step": 12055 }, { "epoch": 3.208089409260245, "grad_norm": 0.3732728064060211, "learning_rate": 5.9064659996854415e-08, "loss": 0.1949, "step": 12056 }, { "epoch": 3.2083555082490687, "grad_norm": 0.2862898111343384, "learning_rate": 5.904909464589115e-08, "loss": 0.1973, "step": 12057 }, { "epoch": 3.2086216072378924, "grad_norm": 0.30267393589019775, "learning_rate": 5.903353048690242e-08, "loss": 0.1882, "step": 12058 }, { "epoch": 3.2088877062267165, "grad_norm": 0.3241662383079529, "learning_rate": 5.901796752034127e-08, "loss": 0.2178, "step": 12059 }, { "epoch": 3.20915380521554, "grad_norm": 0.27364107966423035, "learning_rate": 5.900240574666064e-08, "loss": 0.1797, "step": 12060 }, { "epoch": 3.209419904204364, "grad_norm": 0.29162007570266724, "learning_rate": 5.898684516631356e-08, "loss": 0.1835, "step": 12061 }, { "epoch": 3.209686003193188, "grad_norm": 0.3056678771972656, "learning_rate": 5.89712857797529e-08, "loss": 0.1964, "step": 12062 }, { "epoch": 3.2099521021820117, "grad_norm": 0.28521057963371277, "learning_rate": 5.8955727587431594e-08, "loss": 0.1856, "step": 12063 }, { "epoch": 3.210218201170836, "grad_norm": 0.3230350911617279, "learning_rate": 5.8940170589802486e-08, "loss": 0.2099, "step": 12064 }, { "epoch": 3.2104843001596595, "grad_norm": 0.31415441632270813, "learning_rate": 5.892461478731844e-08, "loss": 0.1875, "step": 12065 }, { "epoch": 3.210750399148483, "grad_norm": 0.2819766700267792, "learning_rate": 5.890906018043216e-08, "loss": 0.2056, "step": 12066 }, { "epoch": 3.2110164981373073, "grad_norm": 0.2956908941268921, "learning_rate": 5.889350676959646e-08, "loss": 0.182, "step": 12067 }, { "epoch": 3.211282597126131, "grad_norm": 0.2859857976436615, "learning_rate": 5.887795455526403e-08, "loss": 0.2083, "step": 12068 }, { "epoch": 3.2115486961149546, "grad_norm": 0.2889796197414398, "learning_rate": 5.886240353788761e-08, "loss": 0.1892, "step": 12069 }, { "epoch": 3.2118147951037788, "grad_norm": 0.3025193512439728, "learning_rate": 5.8846853717919775e-08, "loss": 0.1975, "step": 12070 }, { "epoch": 3.2120808940926024, "grad_norm": 0.2792225778102875, "learning_rate": 5.883130509581318e-08, "loss": 0.1856, "step": 12071 }, { "epoch": 3.212346993081426, "grad_norm": 0.32348641753196716, "learning_rate": 5.88157576720204e-08, "loss": 0.1884, "step": 12072 }, { "epoch": 3.2126130920702503, "grad_norm": 0.32559922337532043, "learning_rate": 5.8800211446993985e-08, "loss": 0.1958, "step": 12073 }, { "epoch": 3.212879191059074, "grad_norm": 0.2809082865715027, "learning_rate": 5.878466642118648e-08, "loss": 0.1882, "step": 12074 }, { "epoch": 3.2131452900478976, "grad_norm": 0.2877810001373291, "learning_rate": 5.876912259505027e-08, "loss": 0.1893, "step": 12075 }, { "epoch": 3.2134113890367217, "grad_norm": 0.31901612877845764, "learning_rate": 5.875357996903786e-08, "loss": 0.1983, "step": 12076 }, { "epoch": 3.2136774880255454, "grad_norm": 0.3124147057533264, "learning_rate": 5.873803854360164e-08, "loss": 0.2081, "step": 12077 }, { "epoch": 3.2139435870143696, "grad_norm": 0.2835468053817749, "learning_rate": 5.872249831919398e-08, "loss": 0.1846, "step": 12078 }, { "epoch": 3.2142096860031932, "grad_norm": 0.2665283977985382, "learning_rate": 5.870695929626723e-08, "loss": 0.1638, "step": 12079 }, { "epoch": 3.214475784992017, "grad_norm": 0.29358959197998047, "learning_rate": 5.869142147527366e-08, "loss": 0.1918, "step": 12080 }, { "epoch": 3.214741883980841, "grad_norm": 0.28982242941856384, "learning_rate": 5.867588485666557e-08, "loss": 0.1833, "step": 12081 }, { "epoch": 3.2150079829696647, "grad_norm": 0.3153739869594574, "learning_rate": 5.8660349440895154e-08, "loss": 0.2096, "step": 12082 }, { "epoch": 3.2152740819584884, "grad_norm": 0.29947909712791443, "learning_rate": 5.864481522841468e-08, "loss": 0.1997, "step": 12083 }, { "epoch": 3.2155401809473125, "grad_norm": 0.3108518123626709, "learning_rate": 5.862928221967623e-08, "loss": 0.1993, "step": 12084 }, { "epoch": 3.215806279936136, "grad_norm": 0.2767690122127533, "learning_rate": 5.8613750415131924e-08, "loss": 0.176, "step": 12085 }, { "epoch": 3.21607237892496, "grad_norm": 0.30119943618774414, "learning_rate": 5.859821981523391e-08, "loss": 0.1914, "step": 12086 }, { "epoch": 3.216338477913784, "grad_norm": 0.2904263436794281, "learning_rate": 5.858269042043421e-08, "loss": 0.1859, "step": 12087 }, { "epoch": 3.2166045769026077, "grad_norm": 0.2963078022003174, "learning_rate": 5.8567162231184854e-08, "loss": 0.1945, "step": 12088 }, { "epoch": 3.216870675891432, "grad_norm": 0.3820618987083435, "learning_rate": 5.8551635247937846e-08, "loss": 0.1976, "step": 12089 }, { "epoch": 3.2171367748802555, "grad_norm": 0.28234344720840454, "learning_rate": 5.853610947114509e-08, "loss": 0.2011, "step": 12090 }, { "epoch": 3.217402873869079, "grad_norm": 0.3180558979511261, "learning_rate": 5.852058490125854e-08, "loss": 0.2056, "step": 12091 }, { "epoch": 3.2176689728579033, "grad_norm": 0.32935628294944763, "learning_rate": 5.8505061538730097e-08, "loss": 0.2132, "step": 12092 }, { "epoch": 3.217935071846727, "grad_norm": 0.41666364669799805, "learning_rate": 5.848953938401153e-08, "loss": 0.1854, "step": 12093 }, { "epoch": 3.2182011708355507, "grad_norm": 0.29544156789779663, "learning_rate": 5.8474018437554705e-08, "loss": 0.2195, "step": 12094 }, { "epoch": 3.218467269824375, "grad_norm": 0.2889987826347351, "learning_rate": 5.845849869981137e-08, "loss": 0.1922, "step": 12095 }, { "epoch": 3.2187333688131985, "grad_norm": 0.26507505774497986, "learning_rate": 5.844298017123327e-08, "loss": 0.1884, "step": 12096 }, { "epoch": 3.218999467802022, "grad_norm": 0.3351002633571625, "learning_rate": 5.8427462852272123e-08, "loss": 0.205, "step": 12097 }, { "epoch": 3.2192655667908463, "grad_norm": 0.2893017828464508, "learning_rate": 5.841194674337959e-08, "loss": 0.1834, "step": 12098 }, { "epoch": 3.21953166577967, "grad_norm": 0.27172034978866577, "learning_rate": 5.839643184500733e-08, "loss": 0.1713, "step": 12099 }, { "epoch": 3.2197977647684937, "grad_norm": 0.3206422030925751, "learning_rate": 5.838091815760688e-08, "loss": 0.2071, "step": 12100 }, { "epoch": 3.220063863757318, "grad_norm": 0.2772023677825928, "learning_rate": 5.83654056816299e-08, "loss": 0.1753, "step": 12101 }, { "epoch": 3.2203299627461415, "grad_norm": 0.2935031056404114, "learning_rate": 5.83498944175278e-08, "loss": 0.1847, "step": 12102 }, { "epoch": 3.2205960617349656, "grad_norm": 0.29200613498687744, "learning_rate": 5.8334384365752143e-08, "loss": 0.1923, "step": 12103 }, { "epoch": 3.2208621607237893, "grad_norm": 0.3154045641422272, "learning_rate": 5.831887552675436e-08, "loss": 0.2092, "step": 12104 }, { "epoch": 3.221128259712613, "grad_norm": 0.29819485545158386, "learning_rate": 5.830336790098591e-08, "loss": 0.1979, "step": 12105 }, { "epoch": 3.221394358701437, "grad_norm": 0.2933197021484375, "learning_rate": 5.828786148889814e-08, "loss": 0.1926, "step": 12106 }, { "epoch": 3.2216604576902608, "grad_norm": 0.31316646933555603, "learning_rate": 5.827235629094241e-08, "loss": 0.1793, "step": 12107 }, { "epoch": 3.2219265566790845, "grad_norm": 0.29593566060066223, "learning_rate": 5.825685230757006e-08, "loss": 0.1748, "step": 12108 }, { "epoch": 3.2221926556679086, "grad_norm": 0.29201871156692505, "learning_rate": 5.8241349539232345e-08, "loss": 0.1876, "step": 12109 }, { "epoch": 3.2224587546567323, "grad_norm": 0.2651517689228058, "learning_rate": 5.822584798638056e-08, "loss": 0.1792, "step": 12110 }, { "epoch": 3.222724853645556, "grad_norm": 0.3145228624343872, "learning_rate": 5.821034764946584e-08, "loss": 0.2008, "step": 12111 }, { "epoch": 3.22299095263438, "grad_norm": 0.32533690333366394, "learning_rate": 5.819484852893939e-08, "loss": 0.1832, "step": 12112 }, { "epoch": 3.2232570516232038, "grad_norm": 0.2773013114929199, "learning_rate": 5.817935062525235e-08, "loss": 0.1903, "step": 12113 }, { "epoch": 3.223523150612028, "grad_norm": 0.29552069306373596, "learning_rate": 5.816385393885583e-08, "loss": 0.1741, "step": 12114 }, { "epoch": 3.2237892496008516, "grad_norm": 0.2858966290950775, "learning_rate": 5.814835847020094e-08, "loss": 0.1719, "step": 12115 }, { "epoch": 3.2240553485896752, "grad_norm": 0.2699735760688782, "learning_rate": 5.813286421973862e-08, "loss": 0.1775, "step": 12116 }, { "epoch": 3.2243214475784994, "grad_norm": 0.2895340323448181, "learning_rate": 5.811737118791993e-08, "loss": 0.1716, "step": 12117 }, { "epoch": 3.224587546567323, "grad_norm": 0.30062136054039, "learning_rate": 5.8101879375195796e-08, "loss": 0.1928, "step": 12118 }, { "epoch": 3.2248536455561467, "grad_norm": 0.29786041378974915, "learning_rate": 5.8086388782017225e-08, "loss": 0.2069, "step": 12119 }, { "epoch": 3.225119744544971, "grad_norm": 0.3212827444076538, "learning_rate": 5.8070899408835006e-08, "loss": 0.1843, "step": 12120 }, { "epoch": 3.2253858435337945, "grad_norm": 0.28562772274017334, "learning_rate": 5.805541125610003e-08, "loss": 0.1757, "step": 12121 }, { "epoch": 3.225651942522618, "grad_norm": 0.28197625279426575, "learning_rate": 5.803992432426312e-08, "loss": 0.2044, "step": 12122 }, { "epoch": 3.2259180415114423, "grad_norm": 0.2915897071361542, "learning_rate": 5.8024438613775087e-08, "loss": 0.1972, "step": 12123 }, { "epoch": 3.226184140500266, "grad_norm": 0.29118114709854126, "learning_rate": 5.800895412508664e-08, "loss": 0.2, "step": 12124 }, { "epoch": 3.2264502394890897, "grad_norm": 0.2969074845314026, "learning_rate": 5.799347085864851e-08, "loss": 0.2004, "step": 12125 }, { "epoch": 3.226716338477914, "grad_norm": 0.30826178193092346, "learning_rate": 5.7977988814911374e-08, "loss": 0.1944, "step": 12126 }, { "epoch": 3.2269824374667375, "grad_norm": 0.32264435291290283, "learning_rate": 5.796250799432587e-08, "loss": 0.1992, "step": 12127 }, { "epoch": 3.2272485364555616, "grad_norm": 0.31041815876960754, "learning_rate": 5.794702839734267e-08, "loss": 0.199, "step": 12128 }, { "epoch": 3.2275146354443853, "grad_norm": 0.287462443113327, "learning_rate": 5.7931550024412225e-08, "loss": 0.182, "step": 12129 }, { "epoch": 3.227780734433209, "grad_norm": 0.2916986644268036, "learning_rate": 5.791607287598513e-08, "loss": 0.174, "step": 12130 }, { "epoch": 3.228046833422033, "grad_norm": 0.3300602436065674, "learning_rate": 5.7900596952511896e-08, "loss": 0.2236, "step": 12131 }, { "epoch": 3.228312932410857, "grad_norm": 0.3272949755191803, "learning_rate": 5.7885122254442964e-08, "loss": 0.2097, "step": 12132 }, { "epoch": 3.2285790313996805, "grad_norm": 0.32734978199005127, "learning_rate": 5.7869648782228774e-08, "loss": 0.2051, "step": 12133 }, { "epoch": 3.2288451303885046, "grad_norm": 0.2738036811351776, "learning_rate": 5.785417653631977e-08, "loss": 0.1899, "step": 12134 }, { "epoch": 3.2291112293773283, "grad_norm": 0.27089402079582214, "learning_rate": 5.78387055171662e-08, "loss": 0.1801, "step": 12135 }, { "epoch": 3.2293773283661524, "grad_norm": 0.32134413719177246, "learning_rate": 5.7823235725218446e-08, "loss": 0.2105, "step": 12136 }, { "epoch": 3.229643427354976, "grad_norm": 0.3218243718147278, "learning_rate": 5.7807767160926836e-08, "loss": 0.2078, "step": 12137 }, { "epoch": 3.2299095263438, "grad_norm": 0.3076140582561493, "learning_rate": 5.779229982474152e-08, "loss": 0.1965, "step": 12138 }, { "epoch": 3.230175625332624, "grad_norm": 0.366604745388031, "learning_rate": 5.777683371711277e-08, "loss": 0.1725, "step": 12139 }, { "epoch": 3.2304417243214476, "grad_norm": 0.29447513818740845, "learning_rate": 5.776136883849077e-08, "loss": 0.1918, "step": 12140 }, { "epoch": 3.2307078233102713, "grad_norm": 0.278393417596817, "learning_rate": 5.774590518932564e-08, "loss": 0.184, "step": 12141 }, { "epoch": 3.2309739222990954, "grad_norm": 0.3217994272708893, "learning_rate": 5.77304427700675e-08, "loss": 0.1996, "step": 12142 }, { "epoch": 3.231240021287919, "grad_norm": 0.2886018753051758, "learning_rate": 5.771498158116642e-08, "loss": 0.1863, "step": 12143 }, { "epoch": 3.2315061202767428, "grad_norm": 0.2851216197013855, "learning_rate": 5.7699521623072437e-08, "loss": 0.198, "step": 12144 }, { "epoch": 3.231772219265567, "grad_norm": 0.26471537351608276, "learning_rate": 5.768406289623555e-08, "loss": 0.1768, "step": 12145 }, { "epoch": 3.2320383182543906, "grad_norm": 0.2913028597831726, "learning_rate": 5.766860540110575e-08, "loss": 0.2043, "step": 12146 }, { "epoch": 3.2323044172432143, "grad_norm": 0.2906161844730377, "learning_rate": 5.765314913813291e-08, "loss": 0.1729, "step": 12147 }, { "epoch": 3.2325705162320384, "grad_norm": 0.28620725870132446, "learning_rate": 5.7637694107766935e-08, "loss": 0.1959, "step": 12148 }, { "epoch": 3.232836615220862, "grad_norm": 0.2969423532485962, "learning_rate": 5.762224031045768e-08, "loss": 0.2059, "step": 12149 }, { "epoch": 3.233102714209686, "grad_norm": 0.26977312564849854, "learning_rate": 5.760678774665499e-08, "loss": 0.1913, "step": 12150 }, { "epoch": 3.23336881319851, "grad_norm": 0.3439186215400696, "learning_rate": 5.7591336416808645e-08, "loss": 0.2266, "step": 12151 }, { "epoch": 3.2336349121873336, "grad_norm": 0.3149082064628601, "learning_rate": 5.757588632136836e-08, "loss": 0.1819, "step": 12152 }, { "epoch": 3.2339010111761577, "grad_norm": 0.2629207372665405, "learning_rate": 5.7560437460783896e-08, "loss": 0.1705, "step": 12153 }, { "epoch": 3.2341671101649814, "grad_norm": 0.302812784910202, "learning_rate": 5.754498983550493e-08, "loss": 0.1805, "step": 12154 }, { "epoch": 3.234433209153805, "grad_norm": 0.29686087369918823, "learning_rate": 5.752954344598104e-08, "loss": 0.1706, "step": 12155 }, { "epoch": 3.234699308142629, "grad_norm": 0.39416414499282837, "learning_rate": 5.7514098292661874e-08, "loss": 0.1953, "step": 12156 }, { "epoch": 3.234965407131453, "grad_norm": 0.2582692801952362, "learning_rate": 5.7498654375997025e-08, "loss": 0.1641, "step": 12157 }, { "epoch": 3.2352315061202765, "grad_norm": 0.30138126015663147, "learning_rate": 5.748321169643595e-08, "loss": 0.2025, "step": 12158 }, { "epoch": 3.2354976051091007, "grad_norm": 0.3037846088409424, "learning_rate": 5.746777025442819e-08, "loss": 0.1996, "step": 12159 }, { "epoch": 3.2357637040979244, "grad_norm": 0.31538358330726624, "learning_rate": 5.745233005042321e-08, "loss": 0.1898, "step": 12160 }, { "epoch": 3.2360298030867485, "grad_norm": 0.2875726521015167, "learning_rate": 5.743689108487043e-08, "loss": 0.2058, "step": 12161 }, { "epoch": 3.236295902075572, "grad_norm": 0.3028354048728943, "learning_rate": 5.7421453358219244e-08, "loss": 0.199, "step": 12162 }, { "epoch": 3.236562001064396, "grad_norm": 0.2929554879665375, "learning_rate": 5.7406016870918975e-08, "loss": 0.2024, "step": 12163 }, { "epoch": 3.23682810005322, "grad_norm": 0.2819531559944153, "learning_rate": 5.739058162341901e-08, "loss": 0.1723, "step": 12164 }, { "epoch": 3.2370941990420437, "grad_norm": 0.2948697507381439, "learning_rate": 5.737514761616854e-08, "loss": 0.1967, "step": 12165 }, { "epoch": 3.2373602980308673, "grad_norm": 0.3010377883911133, "learning_rate": 5.735971484961686e-08, "loss": 0.1932, "step": 12166 }, { "epoch": 3.2376263970196915, "grad_norm": 0.29558444023132324, "learning_rate": 5.734428332421315e-08, "loss": 0.2063, "step": 12167 }, { "epoch": 3.237892496008515, "grad_norm": 0.3211166560649872, "learning_rate": 5.73288530404066e-08, "loss": 0.1974, "step": 12168 }, { "epoch": 3.238158594997339, "grad_norm": 0.34180304408073425, "learning_rate": 5.731342399864635e-08, "loss": 0.1935, "step": 12169 }, { "epoch": 3.238424693986163, "grad_norm": 0.27991002798080444, "learning_rate": 5.729799619938149e-08, "loss": 0.1814, "step": 12170 }, { "epoch": 3.2386907929749866, "grad_norm": 0.2977353036403656, "learning_rate": 5.728256964306108e-08, "loss": 0.1949, "step": 12171 }, { "epoch": 3.2389568919638103, "grad_norm": 0.27082473039627075, "learning_rate": 5.7267144330134154e-08, "loss": 0.1837, "step": 12172 }, { "epoch": 3.2392229909526344, "grad_norm": 0.31849026679992676, "learning_rate": 5.725172026104975e-08, "loss": 0.2038, "step": 12173 }, { "epoch": 3.239489089941458, "grad_norm": 0.3046323359012604, "learning_rate": 5.723629743625673e-08, "loss": 0.1978, "step": 12174 }, { "epoch": 3.2397551889302822, "grad_norm": 0.2882177233695984, "learning_rate": 5.722087585620407e-08, "loss": 0.2036, "step": 12175 }, { "epoch": 3.240021287919106, "grad_norm": 0.3548014461994171, "learning_rate": 5.7205455521340664e-08, "loss": 0.2031, "step": 12176 }, { "epoch": 3.2402873869079296, "grad_norm": 0.3076593577861786, "learning_rate": 5.7190036432115284e-08, "loss": 0.1816, "step": 12177 }, { "epoch": 3.2405534858967537, "grad_norm": 0.3052305281162262, "learning_rate": 5.717461858897681e-08, "loss": 0.1918, "step": 12178 }, { "epoch": 3.2408195848855774, "grad_norm": 0.2980218529701233, "learning_rate": 5.7159201992373984e-08, "loss": 0.2036, "step": 12179 }, { "epoch": 3.241085683874401, "grad_norm": 0.2934889793395996, "learning_rate": 5.714378664275554e-08, "loss": 0.1931, "step": 12180 }, { "epoch": 3.2413517828632252, "grad_norm": 0.28719544410705566, "learning_rate": 5.712837254057021e-08, "loss": 0.1911, "step": 12181 }, { "epoch": 3.241617881852049, "grad_norm": 0.2772689759731293, "learning_rate": 5.711295968626668e-08, "loss": 0.1869, "step": 12182 }, { "epoch": 3.241883980840873, "grad_norm": 0.29247039556503296, "learning_rate": 5.70975480802935e-08, "loss": 0.1924, "step": 12183 }, { "epoch": 3.2421500798296967, "grad_norm": 0.2997615933418274, "learning_rate": 5.708213772309929e-08, "loss": 0.2086, "step": 12184 }, { "epoch": 3.2424161788185204, "grad_norm": 0.33549365401268005, "learning_rate": 5.706672861513262e-08, "loss": 0.1781, "step": 12185 }, { "epoch": 3.2426822778073445, "grad_norm": 0.28547850251197815, "learning_rate": 5.7051320756842006e-08, "loss": 0.1892, "step": 12186 }, { "epoch": 3.242948376796168, "grad_norm": 0.3202246427536011, "learning_rate": 5.703591414867592e-08, "loss": 0.1991, "step": 12187 }, { "epoch": 3.243214475784992, "grad_norm": 0.3001640737056732, "learning_rate": 5.7020508791082835e-08, "loss": 0.1849, "step": 12188 }, { "epoch": 3.243480574773816, "grad_norm": 0.27929943799972534, "learning_rate": 5.700510468451113e-08, "loss": 0.1976, "step": 12189 }, { "epoch": 3.2437466737626397, "grad_norm": 0.3206113874912262, "learning_rate": 5.69897018294092e-08, "loss": 0.2026, "step": 12190 }, { "epoch": 3.2440127727514634, "grad_norm": 0.3031708598136902, "learning_rate": 5.6974300226225415e-08, "loss": 0.1941, "step": 12191 }, { "epoch": 3.2442788717402875, "grad_norm": 0.28692132234573364, "learning_rate": 5.695889987540799e-08, "loss": 0.1824, "step": 12192 }, { "epoch": 3.244544970729111, "grad_norm": 0.28840962052345276, "learning_rate": 5.694350077740524e-08, "loss": 0.1737, "step": 12193 }, { "epoch": 3.244811069717935, "grad_norm": 0.3140585422515869, "learning_rate": 5.6928102932665376e-08, "loss": 0.1852, "step": 12194 }, { "epoch": 3.245077168706759, "grad_norm": 0.314901739358902, "learning_rate": 5.6912706341636605e-08, "loss": 0.1906, "step": 12195 }, { "epoch": 3.2453432676955827, "grad_norm": 0.2928165793418884, "learning_rate": 5.6897311004767134e-08, "loss": 0.1825, "step": 12196 }, { "epoch": 3.245609366684407, "grad_norm": 0.2971515953540802, "learning_rate": 5.688191692250497e-08, "loss": 0.1907, "step": 12197 }, { "epoch": 3.2458754656732305, "grad_norm": 0.2914207875728607, "learning_rate": 5.686652409529825e-08, "loss": 0.1838, "step": 12198 }, { "epoch": 3.246141564662054, "grad_norm": 0.28667011857032776, "learning_rate": 5.685113252359501e-08, "loss": 0.1828, "step": 12199 }, { "epoch": 3.2464076636508783, "grad_norm": 0.30313846468925476, "learning_rate": 5.683574220784333e-08, "loss": 0.1955, "step": 12200 }, { "epoch": 3.246673762639702, "grad_norm": 0.30958694219589233, "learning_rate": 5.682035314849106e-08, "loss": 0.215, "step": 12201 }, { "epoch": 3.2469398616285257, "grad_norm": 0.2542703449726105, "learning_rate": 5.68049653459862e-08, "loss": 0.1805, "step": 12202 }, { "epoch": 3.24720596061735, "grad_norm": 0.32525554299354553, "learning_rate": 5.678957880077665e-08, "loss": 0.2053, "step": 12203 }, { "epoch": 3.2474720596061735, "grad_norm": 0.27802976965904236, "learning_rate": 5.6774193513310266e-08, "loss": 0.1726, "step": 12204 }, { "epoch": 3.247738158594997, "grad_norm": 0.3021235167980194, "learning_rate": 5.675880948403486e-08, "loss": 0.1928, "step": 12205 }, { "epoch": 3.2480042575838213, "grad_norm": 0.3065929412841797, "learning_rate": 5.6743426713398244e-08, "loss": 0.1883, "step": 12206 }, { "epoch": 3.248270356572645, "grad_norm": 0.3003710210323334, "learning_rate": 5.672804520184817e-08, "loss": 0.2035, "step": 12207 }, { "epoch": 3.248536455561469, "grad_norm": 0.29925858974456787, "learning_rate": 5.6712664949832345e-08, "loss": 0.1959, "step": 12208 }, { "epoch": 3.2488025545502928, "grad_norm": 0.3065144717693329, "learning_rate": 5.669728595779848e-08, "loss": 0.1911, "step": 12209 }, { "epoch": 3.2490686535391164, "grad_norm": 0.29285192489624023, "learning_rate": 5.668190822619414e-08, "loss": 0.1977, "step": 12210 }, { "epoch": 3.2493347525279406, "grad_norm": 0.27823394536972046, "learning_rate": 5.6666531755467e-08, "loss": 0.1931, "step": 12211 }, { "epoch": 3.2496008515167643, "grad_norm": 0.3121025860309601, "learning_rate": 5.6651156546064584e-08, "loss": 0.199, "step": 12212 }, { "epoch": 3.249866950505588, "grad_norm": 0.3577575087547302, "learning_rate": 5.663578259843447e-08, "loss": 0.2011, "step": 12213 }, { "epoch": 3.250133049494412, "grad_norm": 0.2954486906528473, "learning_rate": 5.66204099130241e-08, "loss": 0.1813, "step": 12214 }, { "epoch": 3.2503991484832357, "grad_norm": 0.29741132259368896, "learning_rate": 5.660503849028099e-08, "loss": 0.1876, "step": 12215 }, { "epoch": 3.2506652474720594, "grad_norm": 0.32325664162635803, "learning_rate": 5.6589668330652565e-08, "loss": 0.2184, "step": 12216 }, { "epoch": 3.2509313464608836, "grad_norm": 0.2850586175918579, "learning_rate": 5.6574299434586136e-08, "loss": 0.1985, "step": 12217 }, { "epoch": 3.2511974454497072, "grad_norm": 0.31237003207206726, "learning_rate": 5.6558931802529107e-08, "loss": 0.2117, "step": 12218 }, { "epoch": 3.251463544438531, "grad_norm": 0.2899893522262573, "learning_rate": 5.654356543492882e-08, "loss": 0.183, "step": 12219 }, { "epoch": 3.251729643427355, "grad_norm": 0.3049299716949463, "learning_rate": 5.6528200332232467e-08, "loss": 0.2049, "step": 12220 }, { "epoch": 3.2519957424161787, "grad_norm": 0.31139808893203735, "learning_rate": 5.6512836494887316e-08, "loss": 0.1782, "step": 12221 }, { "epoch": 3.252261841405003, "grad_norm": 0.30085673928260803, "learning_rate": 5.6497473923340585e-08, "loss": 0.1914, "step": 12222 }, { "epoch": 3.2525279403938265, "grad_norm": 0.29299941658973694, "learning_rate": 5.6482112618039455e-08, "loss": 0.1846, "step": 12223 }, { "epoch": 3.25279403938265, "grad_norm": 0.29921573400497437, "learning_rate": 5.646675257943101e-08, "loss": 0.1681, "step": 12224 }, { "epoch": 3.2530601383714743, "grad_norm": 0.37968143820762634, "learning_rate": 5.645139380796238e-08, "loss": 0.2067, "step": 12225 }, { "epoch": 3.253326237360298, "grad_norm": 0.3106781244277954, "learning_rate": 5.64360363040806e-08, "loss": 0.1981, "step": 12226 }, { "epoch": 3.2535923363491217, "grad_norm": 0.29731595516204834, "learning_rate": 5.642068006823273e-08, "loss": 0.2005, "step": 12227 }, { "epoch": 3.253858435337946, "grad_norm": 0.28155338764190674, "learning_rate": 5.6405325100865665e-08, "loss": 0.1742, "step": 12228 }, { "epoch": 3.2541245343267695, "grad_norm": 0.29734504222869873, "learning_rate": 5.6389971402426403e-08, "loss": 0.1813, "step": 12229 }, { "epoch": 3.2543906333155936, "grad_norm": 0.3385014832019806, "learning_rate": 5.637461897336184e-08, "loss": 0.2236, "step": 12230 }, { "epoch": 3.2546567323044173, "grad_norm": 0.3106648027896881, "learning_rate": 5.635926781411884e-08, "loss": 0.1786, "step": 12231 }, { "epoch": 3.254922831293241, "grad_norm": 0.2971597909927368, "learning_rate": 5.634391792514425e-08, "loss": 0.201, "step": 12232 }, { "epoch": 3.255188930282065, "grad_norm": 0.293300062417984, "learning_rate": 5.6328569306884854e-08, "loss": 0.1843, "step": 12233 }, { "epoch": 3.255455029270889, "grad_norm": 0.2762085497379303, "learning_rate": 5.63132219597874e-08, "loss": 0.1858, "step": 12234 }, { "epoch": 3.2557211282597125, "grad_norm": 0.3131121098995209, "learning_rate": 5.6297875884298687e-08, "loss": 0.1985, "step": 12235 }, { "epoch": 3.2559872272485366, "grad_norm": 0.28012576699256897, "learning_rate": 5.628253108086527e-08, "loss": 0.1921, "step": 12236 }, { "epoch": 3.2562533262373603, "grad_norm": 0.2881338894367218, "learning_rate": 5.626718754993389e-08, "loss": 0.1905, "step": 12237 }, { "epoch": 3.256519425226184, "grad_norm": 0.3392065167427063, "learning_rate": 5.625184529195115e-08, "loss": 0.1951, "step": 12238 }, { "epoch": 3.256785524215008, "grad_norm": 0.3136351406574249, "learning_rate": 5.623650430736358e-08, "loss": 0.2001, "step": 12239 }, { "epoch": 3.257051623203832, "grad_norm": 0.3246475160121918, "learning_rate": 5.622116459661772e-08, "loss": 0.2127, "step": 12240 }, { "epoch": 3.2573177221926555, "grad_norm": 0.2929929196834564, "learning_rate": 5.6205826160160085e-08, "loss": 0.1803, "step": 12241 }, { "epoch": 3.2575838211814796, "grad_norm": 0.34379950165748596, "learning_rate": 5.619048899843715e-08, "loss": 0.2026, "step": 12242 }, { "epoch": 3.2578499201703033, "grad_norm": 0.2942848205566406, "learning_rate": 5.617515311189533e-08, "loss": 0.1815, "step": 12243 }, { "epoch": 3.258116019159127, "grad_norm": 0.2994258999824524, "learning_rate": 5.6159818500981014e-08, "loss": 0.1994, "step": 12244 }, { "epoch": 3.258382118147951, "grad_norm": 0.2957448661327362, "learning_rate": 5.6144485166140585e-08, "loss": 0.2069, "step": 12245 }, { "epoch": 3.2586482171367748, "grad_norm": 0.295994371175766, "learning_rate": 5.612915310782029e-08, "loss": 0.2136, "step": 12246 }, { "epoch": 3.258914316125599, "grad_norm": 0.29043421149253845, "learning_rate": 5.611382232646644e-08, "loss": 0.1833, "step": 12247 }, { "epoch": 3.2591804151144226, "grad_norm": 0.2919135093688965, "learning_rate": 5.609849282252528e-08, "loss": 0.1904, "step": 12248 }, { "epoch": 3.2594465141032463, "grad_norm": 0.2909621596336365, "learning_rate": 5.608316459644298e-08, "loss": 0.1884, "step": 12249 }, { "epoch": 3.2597126130920704, "grad_norm": 0.2905309200286865, "learning_rate": 5.606783764866575e-08, "loss": 0.1889, "step": 12250 }, { "epoch": 3.259978712080894, "grad_norm": 0.2887742519378662, "learning_rate": 5.605251197963969e-08, "loss": 0.1772, "step": 12251 }, { "epoch": 3.2602448110697178, "grad_norm": 0.3204519748687744, "learning_rate": 5.6037187589810896e-08, "loss": 0.2112, "step": 12252 }, { "epoch": 3.260510910058542, "grad_norm": 0.3098318576812744, "learning_rate": 5.6021864479625425e-08, "loss": 0.1955, "step": 12253 }, { "epoch": 3.2607770090473656, "grad_norm": 0.2963419258594513, "learning_rate": 5.600654264952933e-08, "loss": 0.1937, "step": 12254 }, { "epoch": 3.2610431080361897, "grad_norm": 0.2801479995250702, "learning_rate": 5.5991222099968515e-08, "loss": 0.1872, "step": 12255 }, { "epoch": 3.2613092070250134, "grad_norm": 0.27733495831489563, "learning_rate": 5.5975902831388956e-08, "loss": 0.1759, "step": 12256 }, { "epoch": 3.261575306013837, "grad_norm": 0.27982452511787415, "learning_rate": 5.596058484423656e-08, "loss": 0.1833, "step": 12257 }, { "epoch": 3.261841405002661, "grad_norm": 0.30110350251197815, "learning_rate": 5.594526813895723e-08, "loss": 0.1762, "step": 12258 }, { "epoch": 3.262107503991485, "grad_norm": 0.28408750891685486, "learning_rate": 5.592995271599672e-08, "loss": 0.185, "step": 12259 }, { "epoch": 3.2623736029803085, "grad_norm": 0.29531827569007874, "learning_rate": 5.591463857580085e-08, "loss": 0.1773, "step": 12260 }, { "epoch": 3.2626397019691327, "grad_norm": 0.30670833587646484, "learning_rate": 5.589932571881539e-08, "loss": 0.1954, "step": 12261 }, { "epoch": 3.2629058009579563, "grad_norm": 0.28702765703201294, "learning_rate": 5.5884014145486045e-08, "loss": 0.1947, "step": 12262 }, { "epoch": 3.26317189994678, "grad_norm": 0.3016095459461212, "learning_rate": 5.5868703856258556e-08, "loss": 0.1941, "step": 12263 }, { "epoch": 3.263437998935604, "grad_norm": 0.29627373814582825, "learning_rate": 5.5853394851578474e-08, "loss": 0.202, "step": 12264 }, { "epoch": 3.263704097924428, "grad_norm": 0.32062840461730957, "learning_rate": 5.583808713189142e-08, "loss": 0.1976, "step": 12265 }, { "epoch": 3.2639701969132515, "grad_norm": 0.348104327917099, "learning_rate": 5.5822780697643016e-08, "loss": 0.1855, "step": 12266 }, { "epoch": 3.2642362959020756, "grad_norm": 0.2858991324901581, "learning_rate": 5.580747554927874e-08, "loss": 0.1843, "step": 12267 }, { "epoch": 3.2645023948908993, "grad_norm": 0.29684242606163025, "learning_rate": 5.5792171687244105e-08, "loss": 0.2016, "step": 12268 }, { "epoch": 3.264768493879723, "grad_norm": 0.31524452567100525, "learning_rate": 5.577686911198457e-08, "loss": 0.2053, "step": 12269 }, { "epoch": 3.265034592868547, "grad_norm": 0.29069972038269043, "learning_rate": 5.576156782394555e-08, "loss": 0.1783, "step": 12270 }, { "epoch": 3.265300691857371, "grad_norm": 0.2715303599834442, "learning_rate": 5.5746267823572435e-08, "loss": 0.1828, "step": 12271 }, { "epoch": 3.265566790846195, "grad_norm": 0.2823813259601593, "learning_rate": 5.5730969111310556e-08, "loss": 0.1899, "step": 12272 }, { "epoch": 3.2658328898350186, "grad_norm": 0.28758642077445984, "learning_rate": 5.571567168760526e-08, "loss": 0.1711, "step": 12273 }, { "epoch": 3.2660989888238423, "grad_norm": 0.31333664059638977, "learning_rate": 5.570037555290175e-08, "loss": 0.1937, "step": 12274 }, { "epoch": 3.2663650878126664, "grad_norm": 0.29700884222984314, "learning_rate": 5.568508070764526e-08, "loss": 0.1847, "step": 12275 }, { "epoch": 3.26663118680149, "grad_norm": 0.2970581352710724, "learning_rate": 5.5669787152281036e-08, "loss": 0.1815, "step": 12276 }, { "epoch": 3.2668972857903142, "grad_norm": 0.3031104505062103, "learning_rate": 5.565449488725422e-08, "loss": 0.1827, "step": 12277 }, { "epoch": 3.267163384779138, "grad_norm": 0.28851091861724854, "learning_rate": 5.563920391300989e-08, "loss": 0.1767, "step": 12278 }, { "epoch": 3.2674294837679616, "grad_norm": 0.2922408878803253, "learning_rate": 5.562391422999315e-08, "loss": 0.1786, "step": 12279 }, { "epoch": 3.2676955827567857, "grad_norm": 0.2993601858615875, "learning_rate": 5.560862583864905e-08, "loss": 0.1949, "step": 12280 }, { "epoch": 3.2679616817456094, "grad_norm": 0.33584606647491455, "learning_rate": 5.559333873942258e-08, "loss": 0.1906, "step": 12281 }, { "epoch": 3.268227780734433, "grad_norm": 0.27914491295814514, "learning_rate": 5.557805293275877e-08, "loss": 0.1897, "step": 12282 }, { "epoch": 3.2684938797232572, "grad_norm": 0.2918572425842285, "learning_rate": 5.556276841910246e-08, "loss": 0.1934, "step": 12283 }, { "epoch": 3.268759978712081, "grad_norm": 0.29705584049224854, "learning_rate": 5.554748519889857e-08, "loss": 0.1848, "step": 12284 }, { "epoch": 3.2690260777009046, "grad_norm": 0.2976590096950531, "learning_rate": 5.553220327259197e-08, "loss": 0.1952, "step": 12285 }, { "epoch": 3.2692921766897287, "grad_norm": 0.30522847175598145, "learning_rate": 5.5516922640627484e-08, "loss": 0.188, "step": 12286 }, { "epoch": 3.2695582756785524, "grad_norm": 0.305534303188324, "learning_rate": 5.550164330344986e-08, "loss": 0.1805, "step": 12287 }, { "epoch": 3.269824374667376, "grad_norm": 0.3038008213043213, "learning_rate": 5.548636526150388e-08, "loss": 0.2049, "step": 12288 }, { "epoch": 3.2700904736562, "grad_norm": 0.2835308015346527, "learning_rate": 5.547108851523421e-08, "loss": 0.1924, "step": 12289 }, { "epoch": 3.270356572645024, "grad_norm": 0.2887982130050659, "learning_rate": 5.5455813065085555e-08, "loss": 0.1912, "step": 12290 }, { "epoch": 3.2706226716338476, "grad_norm": 0.3087364137172699, "learning_rate": 5.5440538911502555e-08, "loss": 0.1848, "step": 12291 }, { "epoch": 3.2708887706226717, "grad_norm": 0.3079654574394226, "learning_rate": 5.542526605492972e-08, "loss": 0.1946, "step": 12292 }, { "epoch": 3.2711548696114954, "grad_norm": 0.3401772081851959, "learning_rate": 5.540999449581167e-08, "loss": 0.2152, "step": 12293 }, { "epoch": 3.2714209686003195, "grad_norm": 0.29638099670410156, "learning_rate": 5.5394724234592894e-08, "loss": 0.1865, "step": 12294 }, { "epoch": 3.271687067589143, "grad_norm": 0.29672881960868835, "learning_rate": 5.537945527171788e-08, "loss": 0.1882, "step": 12295 }, { "epoch": 3.271953166577967, "grad_norm": 0.2990221381187439, "learning_rate": 5.5364187607631064e-08, "loss": 0.1913, "step": 12296 }, { "epoch": 3.272219265566791, "grad_norm": 0.26671653985977173, "learning_rate": 5.53489212427769e-08, "loss": 0.1927, "step": 12297 }, { "epoch": 3.2724853645556147, "grad_norm": 0.2867092490196228, "learning_rate": 5.533365617759966e-08, "loss": 0.1745, "step": 12298 }, { "epoch": 3.2727514635444384, "grad_norm": 0.29959836602211, "learning_rate": 5.531839241254371e-08, "loss": 0.2017, "step": 12299 }, { "epoch": 3.2730175625332625, "grad_norm": 0.2846064269542694, "learning_rate": 5.5303129948053386e-08, "loss": 0.1751, "step": 12300 }, { "epoch": 3.273283661522086, "grad_norm": 0.2879510223865509, "learning_rate": 5.528786878457285e-08, "loss": 0.1932, "step": 12301 }, { "epoch": 3.2735497605109103, "grad_norm": 0.3338237404823303, "learning_rate": 5.5272608922546373e-08, "loss": 0.2173, "step": 12302 }, { "epoch": 3.273815859499734, "grad_norm": 0.30677157640457153, "learning_rate": 5.5257350362418097e-08, "loss": 0.1806, "step": 12303 }, { "epoch": 3.2740819584885577, "grad_norm": 0.2959423363208771, "learning_rate": 5.524209310463219e-08, "loss": 0.1896, "step": 12304 }, { "epoch": 3.274348057477382, "grad_norm": 0.30373042821884155, "learning_rate": 5.5226837149632744e-08, "loss": 0.1848, "step": 12305 }, { "epoch": 3.2746141564662055, "grad_norm": 0.2789378762245178, "learning_rate": 5.521158249786381e-08, "loss": 0.1878, "step": 12306 }, { "epoch": 3.274880255455029, "grad_norm": 0.3116642236709595, "learning_rate": 5.519632914976942e-08, "loss": 0.2102, "step": 12307 }, { "epoch": 3.2751463544438533, "grad_norm": 0.2983090281486511, "learning_rate": 5.518107710579356e-08, "loss": 0.1956, "step": 12308 }, { "epoch": 3.275412453432677, "grad_norm": 0.3062953054904938, "learning_rate": 5.516582636638021e-08, "loss": 0.1753, "step": 12309 }, { "epoch": 3.2756785524215006, "grad_norm": 0.2738018333911896, "learning_rate": 5.5150576931973205e-08, "loss": 0.1915, "step": 12310 }, { "epoch": 3.2759446514103248, "grad_norm": 0.3148665130138397, "learning_rate": 5.5135328803016446e-08, "loss": 0.2041, "step": 12311 }, { "epoch": 3.2762107503991484, "grad_norm": 0.3210901618003845, "learning_rate": 5.512008197995378e-08, "loss": 0.2192, "step": 12312 }, { "epoch": 3.276476849387972, "grad_norm": 0.29329079389572144, "learning_rate": 5.510483646322901e-08, "loss": 0.2221, "step": 12313 }, { "epoch": 3.2767429483767962, "grad_norm": 0.32175174355506897, "learning_rate": 5.508959225328587e-08, "loss": 0.1992, "step": 12314 }, { "epoch": 3.27700904736562, "grad_norm": 0.29919421672821045, "learning_rate": 5.507434935056809e-08, "loss": 0.1967, "step": 12315 }, { "epoch": 3.2772751463544436, "grad_norm": 0.328111469745636, "learning_rate": 5.505910775551935e-08, "loss": 0.1965, "step": 12316 }, { "epoch": 3.2775412453432677, "grad_norm": 0.3211434781551361, "learning_rate": 5.504386746858334e-08, "loss": 0.2001, "step": 12317 }, { "epoch": 3.2778073443320914, "grad_norm": 0.3167642056941986, "learning_rate": 5.5028628490203586e-08, "loss": 0.2008, "step": 12318 }, { "epoch": 3.2780734433209155, "grad_norm": 0.29531314969062805, "learning_rate": 5.5013390820823726e-08, "loss": 0.1909, "step": 12319 }, { "epoch": 3.2783395423097392, "grad_norm": 0.2706793248653412, "learning_rate": 5.49981544608872e-08, "loss": 0.1712, "step": 12320 }, { "epoch": 3.278605641298563, "grad_norm": 0.30389297008514404, "learning_rate": 5.498291941083757e-08, "loss": 0.1903, "step": 12321 }, { "epoch": 3.278871740287387, "grad_norm": 0.28419286012649536, "learning_rate": 5.496768567111827e-08, "loss": 0.1866, "step": 12322 }, { "epoch": 3.2791378392762107, "grad_norm": 0.314116895198822, "learning_rate": 5.4952453242172705e-08, "loss": 0.2055, "step": 12323 }, { "epoch": 3.2794039382650344, "grad_norm": 0.28328219056129456, "learning_rate": 5.493722212444427e-08, "loss": 0.1759, "step": 12324 }, { "epoch": 3.2796700372538585, "grad_norm": 0.30498045682907104, "learning_rate": 5.492199231837629e-08, "loss": 0.1861, "step": 12325 }, { "epoch": 3.279936136242682, "grad_norm": 0.2833634912967682, "learning_rate": 5.490676382441206e-08, "loss": 0.1756, "step": 12326 }, { "epoch": 3.2802022352315063, "grad_norm": 0.31119152903556824, "learning_rate": 5.489153664299491e-08, "loss": 0.1905, "step": 12327 }, { "epoch": 3.28046833422033, "grad_norm": 0.3053673505783081, "learning_rate": 5.4876310774567956e-08, "loss": 0.1931, "step": 12328 }, { "epoch": 3.2807344332091537, "grad_norm": 0.27546754479408264, "learning_rate": 5.486108621957444e-08, "loss": 0.1775, "step": 12329 }, { "epoch": 3.281000532197978, "grad_norm": 0.316272497177124, "learning_rate": 5.484586297845748e-08, "loss": 0.2007, "step": 12330 }, { "epoch": 3.2812666311868015, "grad_norm": 0.2936865985393524, "learning_rate": 5.483064105166022e-08, "loss": 0.1986, "step": 12331 }, { "epoch": 3.281532730175625, "grad_norm": 0.3197465240955353, "learning_rate": 5.481542043962572e-08, "loss": 0.2087, "step": 12332 }, { "epoch": 3.2817988291644493, "grad_norm": 0.30187398195266724, "learning_rate": 5.480020114279702e-08, "loss": 0.205, "step": 12333 }, { "epoch": 3.282064928153273, "grad_norm": 0.2866309583187103, "learning_rate": 5.478498316161709e-08, "loss": 0.1846, "step": 12334 }, { "epoch": 3.2823310271420967, "grad_norm": 0.2858516573905945, "learning_rate": 5.476976649652891e-08, "loss": 0.1788, "step": 12335 }, { "epoch": 3.282597126130921, "grad_norm": 0.28749287128448486, "learning_rate": 5.475455114797543e-08, "loss": 0.1829, "step": 12336 }, { "epoch": 3.2828632251197445, "grad_norm": 0.2780040204524994, "learning_rate": 5.473933711639944e-08, "loss": 0.1742, "step": 12337 }, { "epoch": 3.283129324108568, "grad_norm": 0.28062278032302856, "learning_rate": 5.472412440224383e-08, "loss": 0.1956, "step": 12338 }, { "epoch": 3.2833954230973923, "grad_norm": 0.28538787364959717, "learning_rate": 5.470891300595144e-08, "loss": 0.1904, "step": 12339 }, { "epoch": 3.283661522086216, "grad_norm": 0.2685031592845917, "learning_rate": 5.469370292796496e-08, "loss": 0.1951, "step": 12340 }, { "epoch": 3.28392762107504, "grad_norm": 0.28702443838119507, "learning_rate": 5.4678494168727144e-08, "loss": 0.1783, "step": 12341 }, { "epoch": 3.284193720063864, "grad_norm": 0.29200437664985657, "learning_rate": 5.466328672868068e-08, "loss": 0.1926, "step": 12342 }, { "epoch": 3.2844598190526875, "grad_norm": 0.3141414523124695, "learning_rate": 5.464808060826824e-08, "loss": 0.2008, "step": 12343 }, { "epoch": 3.2847259180415116, "grad_norm": 0.36420226097106934, "learning_rate": 5.463287580793241e-08, "loss": 0.1958, "step": 12344 }, { "epoch": 3.2849920170303353, "grad_norm": 0.2828785181045532, "learning_rate": 5.4617672328115816e-08, "loss": 0.2015, "step": 12345 }, { "epoch": 3.285258116019159, "grad_norm": 0.2881411910057068, "learning_rate": 5.460247016926089e-08, "loss": 0.1773, "step": 12346 }, { "epoch": 3.285524215007983, "grad_norm": 0.28486692905426025, "learning_rate": 5.4587269331810216e-08, "loss": 0.1823, "step": 12347 }, { "epoch": 3.2857903139968068, "grad_norm": 0.30195942521095276, "learning_rate": 5.457206981620619e-08, "loss": 0.2103, "step": 12348 }, { "epoch": 3.286056412985631, "grad_norm": 0.2916049659252167, "learning_rate": 5.4556871622891266e-08, "loss": 0.1953, "step": 12349 }, { "epoch": 3.2863225119744546, "grad_norm": 0.32045331597328186, "learning_rate": 5.454167475230782e-08, "loss": 0.1945, "step": 12350 }, { "epoch": 3.2865886109632783, "grad_norm": 0.2805752754211426, "learning_rate": 5.452647920489819e-08, "loss": 0.1726, "step": 12351 }, { "epoch": 3.2868547099521024, "grad_norm": 0.3071872293949127, "learning_rate": 5.4511284981104686e-08, "loss": 0.2011, "step": 12352 }, { "epoch": 3.287120808940926, "grad_norm": 0.30152812600135803, "learning_rate": 5.449609208136956e-08, "loss": 0.1941, "step": 12353 }, { "epoch": 3.2873869079297497, "grad_norm": 0.30398038029670715, "learning_rate": 5.4480900506135086e-08, "loss": 0.1992, "step": 12354 }, { "epoch": 3.287653006918574, "grad_norm": 0.28896695375442505, "learning_rate": 5.446571025584337e-08, "loss": 0.1814, "step": 12355 }, { "epoch": 3.2879191059073976, "grad_norm": 0.28394368290901184, "learning_rate": 5.445052133093659e-08, "loss": 0.1731, "step": 12356 }, { "epoch": 3.2881852048962212, "grad_norm": 0.3059513568878174, "learning_rate": 5.443533373185688e-08, "loss": 0.1836, "step": 12357 }, { "epoch": 3.2884513038850454, "grad_norm": 0.28050029277801514, "learning_rate": 5.442014745904629e-08, "loss": 0.1877, "step": 12358 }, { "epoch": 3.288717402873869, "grad_norm": 0.27358272671699524, "learning_rate": 5.4404962512946905e-08, "loss": 0.1713, "step": 12359 }, { "epoch": 3.2889835018626927, "grad_norm": 0.319775253534317, "learning_rate": 5.438977889400063e-08, "loss": 0.1969, "step": 12360 }, { "epoch": 3.289249600851517, "grad_norm": 0.3052007555961609, "learning_rate": 5.437459660264947e-08, "loss": 0.1877, "step": 12361 }, { "epoch": 3.2895156998403405, "grad_norm": 0.2964165508747101, "learning_rate": 5.435941563933533e-08, "loss": 0.189, "step": 12362 }, { "epoch": 3.289781798829164, "grad_norm": 0.3312038779258728, "learning_rate": 5.434423600450013e-08, "loss": 0.1934, "step": 12363 }, { "epoch": 3.2900478978179883, "grad_norm": 0.2813786268234253, "learning_rate": 5.4329057698585657e-08, "loss": 0.1753, "step": 12364 }, { "epoch": 3.290313996806812, "grad_norm": 0.27549058198928833, "learning_rate": 5.431388072203372e-08, "loss": 0.1742, "step": 12365 }, { "epoch": 3.290580095795636, "grad_norm": 0.32258936762809753, "learning_rate": 5.4298705075286086e-08, "loss": 0.2179, "step": 12366 }, { "epoch": 3.29084619478446, "grad_norm": 0.29754799604415894, "learning_rate": 5.42835307587845e-08, "loss": 0.2006, "step": 12367 }, { "epoch": 3.2911122937732835, "grad_norm": 0.3079933524131775, "learning_rate": 5.426835777297063e-08, "loss": 0.1899, "step": 12368 }, { "epoch": 3.2913783927621076, "grad_norm": 0.3135930895805359, "learning_rate": 5.425318611828611e-08, "loss": 0.1983, "step": 12369 }, { "epoch": 3.2916444917509313, "grad_norm": 0.3099971413612366, "learning_rate": 5.4238015795172575e-08, "loss": 0.19, "step": 12370 }, { "epoch": 3.291910590739755, "grad_norm": 0.2894536852836609, "learning_rate": 5.4222846804071574e-08, "loss": 0.2004, "step": 12371 }, { "epoch": 3.292176689728579, "grad_norm": 0.26557937264442444, "learning_rate": 5.4207679145424676e-08, "loss": 0.1764, "step": 12372 }, { "epoch": 3.292442788717403, "grad_norm": 0.3137916624546051, "learning_rate": 5.419251281967332e-08, "loss": 0.1784, "step": 12373 }, { "epoch": 3.292708887706227, "grad_norm": 0.3057956099510193, "learning_rate": 5.4177347827258954e-08, "loss": 0.1854, "step": 12374 }, { "epoch": 3.2929749866950506, "grad_norm": 0.30251628160476685, "learning_rate": 5.4162184168623036e-08, "loss": 0.1843, "step": 12375 }, { "epoch": 3.2932410856838743, "grad_norm": 0.30033209919929504, "learning_rate": 5.414702184420692e-08, "loss": 0.1957, "step": 12376 }, { "epoch": 3.2935071846726984, "grad_norm": 0.29373592138290405, "learning_rate": 5.413186085445193e-08, "loss": 0.1885, "step": 12377 }, { "epoch": 3.293773283661522, "grad_norm": 0.2865638732910156, "learning_rate": 5.411670119979941e-08, "loss": 0.1835, "step": 12378 }, { "epoch": 3.294039382650346, "grad_norm": 0.28799009323120117, "learning_rate": 5.410154288069055e-08, "loss": 0.1842, "step": 12379 }, { "epoch": 3.29430548163917, "grad_norm": 0.3175475299358368, "learning_rate": 5.408638589756661e-08, "loss": 0.2105, "step": 12380 }, { "epoch": 3.2945715806279936, "grad_norm": 0.3257367014884949, "learning_rate": 5.4071230250868794e-08, "loss": 0.1968, "step": 12381 }, { "epoch": 3.2948376796168173, "grad_norm": 0.2804051637649536, "learning_rate": 5.4056075941038184e-08, "loss": 0.1757, "step": 12382 }, { "epoch": 3.2951037786056414, "grad_norm": 0.2794119417667389, "learning_rate": 5.4040922968515904e-08, "loss": 0.1832, "step": 12383 }, { "epoch": 3.295369877594465, "grad_norm": 0.41382044553756714, "learning_rate": 5.402577133374302e-08, "loss": 0.2015, "step": 12384 }, { "epoch": 3.2956359765832888, "grad_norm": 0.33583346009254456, "learning_rate": 5.401062103716056e-08, "loss": 0.202, "step": 12385 }, { "epoch": 3.295902075572113, "grad_norm": 0.2981894612312317, "learning_rate": 5.3995472079209525e-08, "loss": 0.1774, "step": 12386 }, { "epoch": 3.2961681745609366, "grad_norm": 0.281504362821579, "learning_rate": 5.3980324460330837e-08, "loss": 0.1868, "step": 12387 }, { "epoch": 3.2964342735497603, "grad_norm": 0.285966157913208, "learning_rate": 5.3965178180965424e-08, "loss": 0.1851, "step": 12388 }, { "epoch": 3.2967003725385844, "grad_norm": 0.29934433102607727, "learning_rate": 5.395003324155414e-08, "loss": 0.1993, "step": 12389 }, { "epoch": 3.296966471527408, "grad_norm": 0.3277817666530609, "learning_rate": 5.393488964253786e-08, "loss": 0.2006, "step": 12390 }, { "epoch": 3.297232570516232, "grad_norm": 0.2630215883255005, "learning_rate": 5.39197473843573e-08, "loss": 0.1663, "step": 12391 }, { "epoch": 3.297498669505056, "grad_norm": 0.2964933216571808, "learning_rate": 5.390460646745325e-08, "loss": 0.189, "step": 12392 }, { "epoch": 3.2977647684938796, "grad_norm": 0.3162177801132202, "learning_rate": 5.388946689226641e-08, "loss": 0.1822, "step": 12393 }, { "epoch": 3.2980308674827037, "grad_norm": 0.30410489439964294, "learning_rate": 5.3874328659237466e-08, "loss": 0.2001, "step": 12394 }, { "epoch": 3.2982969664715274, "grad_norm": 0.31028303503990173, "learning_rate": 5.3859191768807056e-08, "loss": 0.1855, "step": 12395 }, { "epoch": 3.2985630654603515, "grad_norm": 0.3259815573692322, "learning_rate": 5.384405622141577e-08, "loss": 0.1959, "step": 12396 }, { "epoch": 3.298829164449175, "grad_norm": 0.3139954209327698, "learning_rate": 5.3828922017504154e-08, "loss": 0.2005, "step": 12397 }, { "epoch": 3.299095263437999, "grad_norm": 0.3067491352558136, "learning_rate": 5.381378915751279e-08, "loss": 0.1865, "step": 12398 }, { "epoch": 3.299361362426823, "grad_norm": 0.2794334292411804, "learning_rate": 5.379865764188205e-08, "loss": 0.1726, "step": 12399 }, { "epoch": 3.2996274614156467, "grad_norm": 0.30431994795799255, "learning_rate": 5.378352747105244e-08, "loss": 0.1965, "step": 12400 }, { "epoch": 3.2998935604044703, "grad_norm": 0.28950509428977966, "learning_rate": 5.3768398645464374e-08, "loss": 0.192, "step": 12401 }, { "epoch": 3.3001596593932945, "grad_norm": 0.3279157280921936, "learning_rate": 5.3753271165558144e-08, "loss": 0.1943, "step": 12402 }, { "epoch": 3.300425758382118, "grad_norm": 0.2865295708179474, "learning_rate": 5.3738145031774116e-08, "loss": 0.1962, "step": 12403 }, { "epoch": 3.300691857370942, "grad_norm": 0.31877920031547546, "learning_rate": 5.372302024455258e-08, "loss": 0.2115, "step": 12404 }, { "epoch": 3.300957956359766, "grad_norm": 0.28890854120254517, "learning_rate": 5.370789680433375e-08, "loss": 0.1755, "step": 12405 }, { "epoch": 3.3012240553485896, "grad_norm": 0.2825467586517334, "learning_rate": 5.369277471155785e-08, "loss": 0.1841, "step": 12406 }, { "epoch": 3.3014901543374133, "grad_norm": 0.28292718529701233, "learning_rate": 5.3677653966665046e-08, "loss": 0.1857, "step": 12407 }, { "epoch": 3.3017562533262375, "grad_norm": 0.3063387870788574, "learning_rate": 5.366253457009551e-08, "loss": 0.2022, "step": 12408 }, { "epoch": 3.302022352315061, "grad_norm": 0.2976977825164795, "learning_rate": 5.364741652228922e-08, "loss": 0.1862, "step": 12409 }, { "epoch": 3.302288451303885, "grad_norm": 0.29172268509864807, "learning_rate": 5.363229982368629e-08, "loss": 0.1888, "step": 12410 }, { "epoch": 3.302554550292709, "grad_norm": 0.30579182505607605, "learning_rate": 5.361718447472672e-08, "loss": 0.2203, "step": 12411 }, { "epoch": 3.3028206492815326, "grad_norm": 0.28570660948753357, "learning_rate": 5.360207047585048e-08, "loss": 0.1923, "step": 12412 }, { "epoch": 3.3030867482703568, "grad_norm": 0.29581356048583984, "learning_rate": 5.35869578274975e-08, "loss": 0.1778, "step": 12413 }, { "epoch": 3.3033528472591804, "grad_norm": 0.2931381165981293, "learning_rate": 5.3571846530107664e-08, "loss": 0.2013, "step": 12414 }, { "epoch": 3.303618946248004, "grad_norm": 0.3207260072231293, "learning_rate": 5.355673658412083e-08, "loss": 0.1948, "step": 12415 }, { "epoch": 3.3038850452368282, "grad_norm": 0.28267624974250793, "learning_rate": 5.354162798997679e-08, "loss": 0.1985, "step": 12416 }, { "epoch": 3.304151144225652, "grad_norm": 0.27480995655059814, "learning_rate": 5.352652074811539e-08, "loss": 0.1905, "step": 12417 }, { "epoch": 3.3044172432144756, "grad_norm": 0.2904636859893799, "learning_rate": 5.351141485897626e-08, "loss": 0.1914, "step": 12418 }, { "epoch": 3.3046833422032997, "grad_norm": 0.3020164668560028, "learning_rate": 5.3496310322999124e-08, "loss": 0.1851, "step": 12419 }, { "epoch": 3.3049494411921234, "grad_norm": 0.2937364876270294, "learning_rate": 5.34812071406237e-08, "loss": 0.1917, "step": 12420 }, { "epoch": 3.3052155401809475, "grad_norm": 0.30745285749435425, "learning_rate": 5.3466105312289514e-08, "loss": 0.2131, "step": 12421 }, { "epoch": 3.3054816391697712, "grad_norm": 0.2812592089176178, "learning_rate": 5.3451004838436166e-08, "loss": 0.1846, "step": 12422 }, { "epoch": 3.305747738158595, "grad_norm": 0.29294028878211975, "learning_rate": 5.343590571950322e-08, "loss": 0.1848, "step": 12423 }, { "epoch": 3.306013837147419, "grad_norm": 0.293780654668808, "learning_rate": 5.342080795593015e-08, "loss": 0.2002, "step": 12424 }, { "epoch": 3.3062799361362427, "grad_norm": 0.28731846809387207, "learning_rate": 5.3405711548156405e-08, "loss": 0.1849, "step": 12425 }, { "epoch": 3.3065460351250664, "grad_norm": 0.3988781273365021, "learning_rate": 5.339061649662148e-08, "loss": 0.2144, "step": 12426 }, { "epoch": 3.3068121341138905, "grad_norm": 0.31461670994758606, "learning_rate": 5.337552280176463e-08, "loss": 0.188, "step": 12427 }, { "epoch": 3.307078233102714, "grad_norm": 0.28166627883911133, "learning_rate": 5.336043046402525e-08, "loss": 0.1722, "step": 12428 }, { "epoch": 3.307344332091538, "grad_norm": 0.2822555899620056, "learning_rate": 5.3345339483842654e-08, "loss": 0.1724, "step": 12429 }, { "epoch": 3.307610431080362, "grad_norm": 0.28123924136161804, "learning_rate": 5.333024986165608e-08, "loss": 0.1667, "step": 12430 }, { "epoch": 3.3078765300691857, "grad_norm": 0.3071271777153015, "learning_rate": 5.331516159790477e-08, "loss": 0.2007, "step": 12431 }, { "epoch": 3.3081426290580094, "grad_norm": 0.3001479208469391, "learning_rate": 5.3300074693027876e-08, "loss": 0.2032, "step": 12432 }, { "epoch": 3.3084087280468335, "grad_norm": 0.3112996816635132, "learning_rate": 5.3284989147464555e-08, "loss": 0.1921, "step": 12433 }, { "epoch": 3.308674827035657, "grad_norm": 0.288335382938385, "learning_rate": 5.326990496165391e-08, "loss": 0.1894, "step": 12434 }, { "epoch": 3.308940926024481, "grad_norm": 0.32823890447616577, "learning_rate": 5.3254822136035026e-08, "loss": 0.2149, "step": 12435 }, { "epoch": 3.309207025013305, "grad_norm": 0.28713810443878174, "learning_rate": 5.323974067104686e-08, "loss": 0.1906, "step": 12436 }, { "epoch": 3.3094731240021287, "grad_norm": 0.31393715739250183, "learning_rate": 5.322466056712844e-08, "loss": 0.1925, "step": 12437 }, { "epoch": 3.309739222990953, "grad_norm": 0.31341198086738586, "learning_rate": 5.32095818247187e-08, "loss": 0.1935, "step": 12438 }, { "epoch": 3.3100053219797765, "grad_norm": 0.30732759833335876, "learning_rate": 5.319450444425654e-08, "loss": 0.1833, "step": 12439 }, { "epoch": 3.3102714209686, "grad_norm": 0.2922530472278595, "learning_rate": 5.317942842618086e-08, "loss": 0.2017, "step": 12440 }, { "epoch": 3.3105375199574243, "grad_norm": 0.30134549736976624, "learning_rate": 5.316435377093043e-08, "loss": 0.1932, "step": 12441 }, { "epoch": 3.310803618946248, "grad_norm": 0.3088366389274597, "learning_rate": 5.3149280478944025e-08, "loss": 0.187, "step": 12442 }, { "epoch": 3.3110697179350717, "grad_norm": 0.2711344361305237, "learning_rate": 5.3134208550660445e-08, "loss": 0.1984, "step": 12443 }, { "epoch": 3.311335816923896, "grad_norm": 0.3141315281391144, "learning_rate": 5.311913798651839e-08, "loss": 0.2026, "step": 12444 }, { "epoch": 3.3116019159127195, "grad_norm": 0.32671892642974854, "learning_rate": 5.3104068786956466e-08, "loss": 0.1926, "step": 12445 }, { "epoch": 3.3118680149015436, "grad_norm": 0.26140832901000977, "learning_rate": 5.308900095241334e-08, "loss": 0.1747, "step": 12446 }, { "epoch": 3.3121341138903673, "grad_norm": 0.2955365478992462, "learning_rate": 5.3073934483327596e-08, "loss": 0.1849, "step": 12447 }, { "epoch": 3.312400212879191, "grad_norm": 0.3302065134048462, "learning_rate": 5.305886938013777e-08, "loss": 0.2065, "step": 12448 }, { "epoch": 3.312666311868015, "grad_norm": 0.3190096318721771, "learning_rate": 5.304380564328238e-08, "loss": 0.1785, "step": 12449 }, { "epoch": 3.3129324108568388, "grad_norm": 0.3102613687515259, "learning_rate": 5.302874327319988e-08, "loss": 0.2016, "step": 12450 }, { "epoch": 3.3131985098456624, "grad_norm": 0.29140061140060425, "learning_rate": 5.301368227032872e-08, "loss": 0.2072, "step": 12451 }, { "epoch": 3.3134646088344866, "grad_norm": 0.28959956765174866, "learning_rate": 5.299862263510726e-08, "loss": 0.194, "step": 12452 }, { "epoch": 3.3137307078233103, "grad_norm": 0.30876970291137695, "learning_rate": 5.298356436797389e-08, "loss": 0.191, "step": 12453 }, { "epoch": 3.313996806812134, "grad_norm": 0.2987476885318756, "learning_rate": 5.296850746936685e-08, "loss": 0.1928, "step": 12454 }, { "epoch": 3.314262905800958, "grad_norm": 0.2889007329940796, "learning_rate": 5.295345193972445e-08, "loss": 0.1889, "step": 12455 }, { "epoch": 3.3145290047897817, "grad_norm": 0.31232786178588867, "learning_rate": 5.29383977794849e-08, "loss": 0.203, "step": 12456 }, { "epoch": 3.3147951037786054, "grad_norm": 0.2849438488483429, "learning_rate": 5.2923344989086395e-08, "loss": 0.1752, "step": 12457 }, { "epoch": 3.3150612027674295, "grad_norm": 0.29605939984321594, "learning_rate": 5.290829356896709e-08, "loss": 0.1911, "step": 12458 }, { "epoch": 3.3153273017562532, "grad_norm": 0.2940658926963806, "learning_rate": 5.289324351956508e-08, "loss": 0.1871, "step": 12459 }, { "epoch": 3.3155934007450774, "grad_norm": 0.27266034483909607, "learning_rate": 5.287819484131848e-08, "loss": 0.19, "step": 12460 }, { "epoch": 3.315859499733901, "grad_norm": 0.3066312074661255, "learning_rate": 5.2863147534665254e-08, "loss": 0.2033, "step": 12461 }, { "epoch": 3.3161255987227247, "grad_norm": 0.3305884003639221, "learning_rate": 5.284810160004343e-08, "loss": 0.2013, "step": 12462 }, { "epoch": 3.316391697711549, "grad_norm": 0.3009490966796875, "learning_rate": 5.2833057037890916e-08, "loss": 0.1877, "step": 12463 }, { "epoch": 3.3166577967003725, "grad_norm": 0.31188032031059265, "learning_rate": 5.2818013848645636e-08, "loss": 0.1926, "step": 12464 }, { "epoch": 3.316923895689196, "grad_norm": 0.2766810655593872, "learning_rate": 5.2802972032745464e-08, "loss": 0.1814, "step": 12465 }, { "epoch": 3.3171899946780203, "grad_norm": 0.3029083013534546, "learning_rate": 5.278793159062824e-08, "loss": 0.183, "step": 12466 }, { "epoch": 3.317456093666844, "grad_norm": 0.2956945598125458, "learning_rate": 5.2772892522731736e-08, "loss": 0.1826, "step": 12467 }, { "epoch": 3.317722192655668, "grad_norm": 0.31536877155303955, "learning_rate": 5.275785482949371e-08, "loss": 0.2005, "step": 12468 }, { "epoch": 3.317988291644492, "grad_norm": 0.30452781915664673, "learning_rate": 5.2742818511351864e-08, "loss": 0.1934, "step": 12469 }, { "epoch": 3.3182543906333155, "grad_norm": 0.2822684347629547, "learning_rate": 5.272778356874388e-08, "loss": 0.1633, "step": 12470 }, { "epoch": 3.3185204896221396, "grad_norm": 0.285033255815506, "learning_rate": 5.2712750002107395e-08, "loss": 0.1868, "step": 12471 }, { "epoch": 3.3187865886109633, "grad_norm": 0.2862895131111145, "learning_rate": 5.269771781187995e-08, "loss": 0.2042, "step": 12472 }, { "epoch": 3.319052687599787, "grad_norm": 0.30633553862571716, "learning_rate": 5.2682686998499116e-08, "loss": 0.188, "step": 12473 }, { "epoch": 3.319318786588611, "grad_norm": 0.2870502471923828, "learning_rate": 5.266765756240241e-08, "loss": 0.1855, "step": 12474 }, { "epoch": 3.319584885577435, "grad_norm": 0.2838262915611267, "learning_rate": 5.265262950402729e-08, "loss": 0.1741, "step": 12475 }, { "epoch": 3.3198509845662585, "grad_norm": 0.2791875898838043, "learning_rate": 5.263760282381119e-08, "loss": 0.1689, "step": 12476 }, { "epoch": 3.3201170835550826, "grad_norm": 0.6066334843635559, "learning_rate": 5.2622577522191506e-08, "loss": 0.2229, "step": 12477 }, { "epoch": 3.3203831825439063, "grad_norm": 0.3078697919845581, "learning_rate": 5.260755359960556e-08, "loss": 0.2032, "step": 12478 }, { "epoch": 3.32064928153273, "grad_norm": 0.30150869488716125, "learning_rate": 5.259253105649073e-08, "loss": 0.2013, "step": 12479 }, { "epoch": 3.320915380521554, "grad_norm": 0.2808416783809662, "learning_rate": 5.257750989328418e-08, "loss": 0.1799, "step": 12480 }, { "epoch": 3.321181479510378, "grad_norm": 0.3170238137245178, "learning_rate": 5.256249011042319e-08, "loss": 0.1816, "step": 12481 }, { "epoch": 3.3214475784992015, "grad_norm": 0.3013993203639984, "learning_rate": 5.254747170834497e-08, "loss": 0.1991, "step": 12482 }, { "epoch": 3.3217136774880256, "grad_norm": 0.2778134047985077, "learning_rate": 5.253245468748662e-08, "loss": 0.1861, "step": 12483 }, { "epoch": 3.3219797764768493, "grad_norm": 0.29987385869026184, "learning_rate": 5.251743904828525e-08, "loss": 0.1894, "step": 12484 }, { "epoch": 3.3222458754656734, "grad_norm": 0.2964349687099457, "learning_rate": 5.2502424791177946e-08, "loss": 0.2021, "step": 12485 }, { "epoch": 3.322511974454497, "grad_norm": 0.31072545051574707, "learning_rate": 5.248741191660173e-08, "loss": 0.1999, "step": 12486 }, { "epoch": 3.3227780734433208, "grad_norm": 0.2900964319705963, "learning_rate": 5.247240042499359e-08, "loss": 0.1925, "step": 12487 }, { "epoch": 3.323044172432145, "grad_norm": 0.2852060794830322, "learning_rate": 5.245739031679047e-08, "loss": 0.1939, "step": 12488 }, { "epoch": 3.3233102714209686, "grad_norm": 0.29620182514190674, "learning_rate": 5.244238159242932e-08, "loss": 0.1801, "step": 12489 }, { "epoch": 3.3235763704097923, "grad_norm": 0.3380484879016876, "learning_rate": 5.2427374252346914e-08, "loss": 0.2194, "step": 12490 }, { "epoch": 3.3238424693986164, "grad_norm": 0.3016730546951294, "learning_rate": 5.2412368296980125e-08, "loss": 0.1945, "step": 12491 }, { "epoch": 3.32410856838744, "grad_norm": 0.30471840500831604, "learning_rate": 5.239736372676573e-08, "loss": 0.1914, "step": 12492 }, { "epoch": 3.324374667376264, "grad_norm": 0.3165838122367859, "learning_rate": 5.238236054214047e-08, "loss": 0.2035, "step": 12493 }, { "epoch": 3.324640766365088, "grad_norm": 0.3142717182636261, "learning_rate": 5.236735874354108e-08, "loss": 0.2182, "step": 12494 }, { "epoch": 3.3249068653539116, "grad_norm": 0.3033415973186493, "learning_rate": 5.235235833140418e-08, "loss": 0.2276, "step": 12495 }, { "epoch": 3.3251729643427357, "grad_norm": 0.2742159068584442, "learning_rate": 5.233735930616642e-08, "loss": 0.1922, "step": 12496 }, { "epoch": 3.3254390633315594, "grad_norm": 0.2929607331752777, "learning_rate": 5.2322361668264383e-08, "loss": 0.1699, "step": 12497 }, { "epoch": 3.325705162320383, "grad_norm": 0.3093109130859375, "learning_rate": 5.230736541813463e-08, "loss": 0.2008, "step": 12498 }, { "epoch": 3.325971261309207, "grad_norm": 0.3582267165184021, "learning_rate": 5.22923705562136e-08, "loss": 0.203, "step": 12499 }, { "epoch": 3.326237360298031, "grad_norm": 0.30811044573783875, "learning_rate": 5.2277377082937803e-08, "loss": 0.1893, "step": 12500 }, { "epoch": 3.3265034592868545, "grad_norm": 0.28916433453559875, "learning_rate": 5.2262384998743634e-08, "loss": 0.1953, "step": 12501 }, { "epoch": 3.3267695582756787, "grad_norm": 0.31952419877052307, "learning_rate": 5.224739430406753e-08, "loss": 0.1848, "step": 12502 }, { "epoch": 3.3270356572645023, "grad_norm": 0.30768895149230957, "learning_rate": 5.223240499934576e-08, "loss": 0.191, "step": 12503 }, { "epoch": 3.327301756253326, "grad_norm": 0.29905781149864197, "learning_rate": 5.2217417085014646e-08, "loss": 0.2061, "step": 12504 }, { "epoch": 3.32756785524215, "grad_norm": 0.30780091881752014, "learning_rate": 5.220243056151046e-08, "loss": 0.1835, "step": 12505 }, { "epoch": 3.327833954230974, "grad_norm": 0.2778720259666443, "learning_rate": 5.2187445429269405e-08, "loss": 0.1836, "step": 12506 }, { "epoch": 3.3281000532197975, "grad_norm": 0.29658985137939453, "learning_rate": 5.217246168872772e-08, "loss": 0.1882, "step": 12507 }, { "epoch": 3.3283661522086216, "grad_norm": 0.31584322452545166, "learning_rate": 5.215747934032144e-08, "loss": 0.1882, "step": 12508 }, { "epoch": 3.3286322511974453, "grad_norm": 0.2942235767841339, "learning_rate": 5.214249838448672e-08, "loss": 0.2006, "step": 12509 }, { "epoch": 3.3288983501862695, "grad_norm": 0.29267194867134094, "learning_rate": 5.212751882165962e-08, "loss": 0.1816, "step": 12510 }, { "epoch": 3.329164449175093, "grad_norm": 0.3084852993488312, "learning_rate": 5.211254065227615e-08, "loss": 0.2208, "step": 12511 }, { "epoch": 3.329430548163917, "grad_norm": 0.27429866790771484, "learning_rate": 5.2097563876772265e-08, "loss": 0.1797, "step": 12512 }, { "epoch": 3.329696647152741, "grad_norm": 0.33131539821624756, "learning_rate": 5.2082588495583934e-08, "loss": 0.1864, "step": 12513 }, { "epoch": 3.3299627461415646, "grad_norm": 0.32429733872413635, "learning_rate": 5.206761450914703e-08, "loss": 0.2095, "step": 12514 }, { "epoch": 3.3302288451303887, "grad_norm": 0.2969873547554016, "learning_rate": 5.20526419178974e-08, "loss": 0.1928, "step": 12515 }, { "epoch": 3.3304949441192124, "grad_norm": 0.28115224838256836, "learning_rate": 5.203767072227092e-08, "loss": 0.1661, "step": 12516 }, { "epoch": 3.330761043108036, "grad_norm": 0.28385815024375916, "learning_rate": 5.202270092270329e-08, "loss": 0.1857, "step": 12517 }, { "epoch": 3.3310271420968602, "grad_norm": 0.2870449125766754, "learning_rate": 5.200773251963024e-08, "loss": 0.1808, "step": 12518 }, { "epoch": 3.331293241085684, "grad_norm": 0.3019932210445404, "learning_rate": 5.199276551348749e-08, "loss": 0.198, "step": 12519 }, { "epoch": 3.3315593400745076, "grad_norm": 0.32671016454696655, "learning_rate": 5.197779990471069e-08, "loss": 0.213, "step": 12520 }, { "epoch": 3.3318254390633317, "grad_norm": 0.27523940801620483, "learning_rate": 5.19628356937355e-08, "loss": 0.1761, "step": 12521 }, { "epoch": 3.3320915380521554, "grad_norm": 0.3143719732761383, "learning_rate": 5.1947872880997376e-08, "loss": 0.2198, "step": 12522 }, { "epoch": 3.332357637040979, "grad_norm": 0.29329177737236023, "learning_rate": 5.193291146693191e-08, "loss": 0.184, "step": 12523 }, { "epoch": 3.332623736029803, "grad_norm": 0.3036351799964905, "learning_rate": 5.1917951451974587e-08, "loss": 0.2004, "step": 12524 }, { "epoch": 3.332889835018627, "grad_norm": 0.2928776144981384, "learning_rate": 5.19029928365609e-08, "loss": 0.1869, "step": 12525 }, { "epoch": 3.3331559340074506, "grad_norm": 0.3995513916015625, "learning_rate": 5.188803562112616e-08, "loss": 0.194, "step": 12526 }, { "epoch": 3.3334220329962747, "grad_norm": 0.3126181364059448, "learning_rate": 5.187307980610578e-08, "loss": 0.2031, "step": 12527 }, { "epoch": 3.3336881319850984, "grad_norm": 0.29536861181259155, "learning_rate": 5.185812539193509e-08, "loss": 0.1938, "step": 12528 }, { "epoch": 3.333954230973922, "grad_norm": 0.30427154898643494, "learning_rate": 5.184317237904938e-08, "loss": 0.205, "step": 12529 }, { "epoch": 3.334220329962746, "grad_norm": 0.30053406953811646, "learning_rate": 5.182822076788387e-08, "loss": 0.1962, "step": 12530 }, { "epoch": 3.33448642895157, "grad_norm": 0.3262181878089905, "learning_rate": 5.181327055887379e-08, "loss": 0.2012, "step": 12531 }, { "epoch": 3.334752527940394, "grad_norm": 0.3086896240711212, "learning_rate": 5.1798321752454285e-08, "loss": 0.2065, "step": 12532 }, { "epoch": 3.3350186269292177, "grad_norm": 0.32780003547668457, "learning_rate": 5.178337434906047e-08, "loss": 0.203, "step": 12533 }, { "epoch": 3.3352847259180414, "grad_norm": 0.3141382336616516, "learning_rate": 5.176842834912745e-08, "loss": 0.1967, "step": 12534 }, { "epoch": 3.3355508249068655, "grad_norm": 0.32815641164779663, "learning_rate": 5.1753483753090275e-08, "loss": 0.19, "step": 12535 }, { "epoch": 3.335816923895689, "grad_norm": 0.3211548924446106, "learning_rate": 5.173854056138388e-08, "loss": 0.192, "step": 12536 }, { "epoch": 3.336083022884513, "grad_norm": 0.28197628259658813, "learning_rate": 5.1723598774443254e-08, "loss": 0.1886, "step": 12537 }, { "epoch": 3.336349121873337, "grad_norm": 0.2983662784099579, "learning_rate": 5.170865839270333e-08, "loss": 0.1858, "step": 12538 }, { "epoch": 3.3366152208621607, "grad_norm": 0.33161208033561707, "learning_rate": 5.1693719416598967e-08, "loss": 0.2091, "step": 12539 }, { "epoch": 3.336881319850985, "grad_norm": 0.2942845821380615, "learning_rate": 5.167878184656502e-08, "loss": 0.1876, "step": 12540 }, { "epoch": 3.3371474188398085, "grad_norm": 0.2960142493247986, "learning_rate": 5.166384568303629e-08, "loss": 0.1809, "step": 12541 }, { "epoch": 3.337413517828632, "grad_norm": 0.31158486008644104, "learning_rate": 5.1648910926447473e-08, "loss": 0.1953, "step": 12542 }, { "epoch": 3.3376796168174563, "grad_norm": 0.2848617136478424, "learning_rate": 5.163397757723332e-08, "loss": 0.1965, "step": 12543 }, { "epoch": 3.33794571580628, "grad_norm": 0.2936745584011078, "learning_rate": 5.161904563582854e-08, "loss": 0.2044, "step": 12544 }, { "epoch": 3.3382118147951036, "grad_norm": 0.3008826673030853, "learning_rate": 5.1604115102667676e-08, "loss": 0.2102, "step": 12545 }, { "epoch": 3.3384779137839278, "grad_norm": 0.2901475131511688, "learning_rate": 5.158918597818538e-08, "loss": 0.1774, "step": 12546 }, { "epoch": 3.3387440127727515, "grad_norm": 0.3041038513183594, "learning_rate": 5.1574258262816164e-08, "loss": 0.2062, "step": 12547 }, { "epoch": 3.339010111761575, "grad_norm": 0.29707151651382446, "learning_rate": 5.1559331956994554e-08, "loss": 0.1795, "step": 12548 }, { "epoch": 3.3392762107503993, "grad_norm": 0.3123239278793335, "learning_rate": 5.1544407061155013e-08, "loss": 0.1966, "step": 12549 }, { "epoch": 3.339542309739223, "grad_norm": 0.29150640964508057, "learning_rate": 5.152948357573199e-08, "loss": 0.1879, "step": 12550 }, { "epoch": 3.3398084087280466, "grad_norm": 0.320764422416687, "learning_rate": 5.1514561501159826e-08, "loss": 0.1925, "step": 12551 }, { "epoch": 3.3400745077168708, "grad_norm": 0.2766641676425934, "learning_rate": 5.1499640837872904e-08, "loss": 0.1789, "step": 12552 }, { "epoch": 3.3403406067056944, "grad_norm": 0.2807215452194214, "learning_rate": 5.148472158630553e-08, "loss": 0.1875, "step": 12553 }, { "epoch": 3.340606705694518, "grad_norm": 0.3009409010410309, "learning_rate": 5.146980374689192e-08, "loss": 0.1815, "step": 12554 }, { "epoch": 3.3408728046833422, "grad_norm": 0.2923336625099182, "learning_rate": 5.145488732006629e-08, "loss": 0.1659, "step": 12555 }, { "epoch": 3.341138903672166, "grad_norm": 0.3047599792480469, "learning_rate": 5.143997230626287e-08, "loss": 0.2001, "step": 12556 }, { "epoch": 3.34140500266099, "grad_norm": 0.291476845741272, "learning_rate": 5.142505870591576e-08, "loss": 0.2051, "step": 12557 }, { "epoch": 3.3416711016498137, "grad_norm": 0.307096928358078, "learning_rate": 5.141014651945905e-08, "loss": 0.1827, "step": 12558 }, { "epoch": 3.3419372006386374, "grad_norm": 0.32941895723342896, "learning_rate": 5.1395235747326836e-08, "loss": 0.1768, "step": 12559 }, { "epoch": 3.3422032996274615, "grad_norm": 0.29457247257232666, "learning_rate": 5.138032638995314e-08, "loss": 0.1915, "step": 12560 }, { "epoch": 3.3424693986162852, "grad_norm": 0.3023757338523865, "learning_rate": 5.136541844777187e-08, "loss": 0.1877, "step": 12561 }, { "epoch": 3.342735497605109, "grad_norm": 0.29949790239334106, "learning_rate": 5.135051192121699e-08, "loss": 0.2061, "step": 12562 }, { "epoch": 3.343001596593933, "grad_norm": 0.2854405641555786, "learning_rate": 5.133560681072242e-08, "loss": 0.1898, "step": 12563 }, { "epoch": 3.3432676955827567, "grad_norm": 0.27740660309791565, "learning_rate": 5.132070311672194e-08, "loss": 0.1897, "step": 12564 }, { "epoch": 3.343533794571581, "grad_norm": 0.29128462076187134, "learning_rate": 5.130580083964941e-08, "loss": 0.1928, "step": 12565 }, { "epoch": 3.3437998935604045, "grad_norm": 0.300605833530426, "learning_rate": 5.129089997993859e-08, "loss": 0.1951, "step": 12566 }, { "epoch": 3.344065992549228, "grad_norm": 0.2786582112312317, "learning_rate": 5.127600053802318e-08, "loss": 0.19, "step": 12567 }, { "epoch": 3.3443320915380523, "grad_norm": 0.2889583706855774, "learning_rate": 5.1261102514336886e-08, "loss": 0.1733, "step": 12568 }, { "epoch": 3.344598190526876, "grad_norm": 0.29017165303230286, "learning_rate": 5.1246205909313364e-08, "loss": 0.1928, "step": 12569 }, { "epoch": 3.3448642895156997, "grad_norm": 0.27611371874809265, "learning_rate": 5.123131072338619e-08, "loss": 0.1868, "step": 12570 }, { "epoch": 3.345130388504524, "grad_norm": 0.26480722427368164, "learning_rate": 5.1216416956988975e-08, "loss": 0.1776, "step": 12571 }, { "epoch": 3.3453964874933475, "grad_norm": 0.269778847694397, "learning_rate": 5.120152461055516e-08, "loss": 0.1737, "step": 12572 }, { "epoch": 3.345662586482171, "grad_norm": 0.28737154603004456, "learning_rate": 5.1186633684518254e-08, "loss": 0.1798, "step": 12573 }, { "epoch": 3.3459286854709953, "grad_norm": 0.30149927735328674, "learning_rate": 5.1171744179311694e-08, "loss": 0.1883, "step": 12574 }, { "epoch": 3.346194784459819, "grad_norm": 0.29070934653282166, "learning_rate": 5.115685609536888e-08, "loss": 0.1887, "step": 12575 }, { "epoch": 3.3464608834486427, "grad_norm": 0.304745614528656, "learning_rate": 5.1141969433123186e-08, "loss": 0.1935, "step": 12576 }, { "epoch": 3.346726982437467, "grad_norm": 0.3105025291442871, "learning_rate": 5.1127084193007884e-08, "loss": 0.1997, "step": 12577 }, { "epoch": 3.3469930814262905, "grad_norm": 0.3052283227443695, "learning_rate": 5.111220037545627e-08, "loss": 0.1931, "step": 12578 }, { "epoch": 3.3472591804151146, "grad_norm": 0.2944762408733368, "learning_rate": 5.1097317980901564e-08, "loss": 0.189, "step": 12579 }, { "epoch": 3.3475252794039383, "grad_norm": 0.2941136658191681, "learning_rate": 5.1082437009777e-08, "loss": 0.2093, "step": 12580 }, { "epoch": 3.347791378392762, "grad_norm": 0.2946356534957886, "learning_rate": 5.106755746251564e-08, "loss": 0.176, "step": 12581 }, { "epoch": 3.348057477381586, "grad_norm": 0.2842348515987396, "learning_rate": 5.105267933955064e-08, "loss": 0.1824, "step": 12582 }, { "epoch": 3.34832357637041, "grad_norm": 0.28284361958503723, "learning_rate": 5.1037802641315074e-08, "loss": 0.1715, "step": 12583 }, { "epoch": 3.3485896753592335, "grad_norm": 0.3129451870918274, "learning_rate": 5.102292736824192e-08, "loss": 0.1924, "step": 12584 }, { "epoch": 3.3488557743480576, "grad_norm": 0.3034847676753998, "learning_rate": 5.1008053520764184e-08, "loss": 0.2029, "step": 12585 }, { "epoch": 3.3491218733368813, "grad_norm": 0.28742438554763794, "learning_rate": 5.09931810993148e-08, "loss": 0.19, "step": 12586 }, { "epoch": 3.3493879723257054, "grad_norm": 0.2907610833644867, "learning_rate": 5.097831010432666e-08, "loss": 0.1886, "step": 12587 }, { "epoch": 3.349654071314529, "grad_norm": 0.29658666253089905, "learning_rate": 5.0963440536232636e-08, "loss": 0.1997, "step": 12588 }, { "epoch": 3.3499201703033528, "grad_norm": 0.29213565587997437, "learning_rate": 5.094857239546557e-08, "loss": 0.2051, "step": 12589 }, { "epoch": 3.350186269292177, "grad_norm": 0.2899998724460602, "learning_rate": 5.093370568245817e-08, "loss": 0.1814, "step": 12590 }, { "epoch": 3.3504523682810006, "grad_norm": 0.32239967584609985, "learning_rate": 5.0918840397643207e-08, "loss": 0.2072, "step": 12591 }, { "epoch": 3.3507184672698243, "grad_norm": 0.317466676235199, "learning_rate": 5.090397654145334e-08, "loss": 0.2076, "step": 12592 }, { "epoch": 3.3509845662586484, "grad_norm": 0.2874282896518707, "learning_rate": 5.0889114114321254e-08, "loss": 0.1773, "step": 12593 }, { "epoch": 3.351250665247472, "grad_norm": 0.27786752581596375, "learning_rate": 5.0874253116679544e-08, "loss": 0.179, "step": 12594 }, { "epoch": 3.3515167642362957, "grad_norm": 0.30659154057502747, "learning_rate": 5.085939354896076e-08, "loss": 0.1961, "step": 12595 }, { "epoch": 3.35178286322512, "grad_norm": 0.28140926361083984, "learning_rate": 5.0844535411597433e-08, "loss": 0.1764, "step": 12596 }, { "epoch": 3.3520489622139436, "grad_norm": 0.30168280005455017, "learning_rate": 5.082967870502206e-08, "loss": 0.1877, "step": 12597 }, { "epoch": 3.3523150612027672, "grad_norm": 0.28257817029953003, "learning_rate": 5.081482342966712e-08, "loss": 0.1682, "step": 12598 }, { "epoch": 3.3525811601915914, "grad_norm": 0.32449454069137573, "learning_rate": 5.079996958596491e-08, "loss": 0.2165, "step": 12599 }, { "epoch": 3.352847259180415, "grad_norm": 0.27705252170562744, "learning_rate": 5.078511717434784e-08, "loss": 0.1684, "step": 12600 }, { "epoch": 3.3531133581692387, "grad_norm": 0.2891294062137604, "learning_rate": 5.0770266195248224e-08, "loss": 0.1817, "step": 12601 }, { "epoch": 3.353379457158063, "grad_norm": 0.3449481725692749, "learning_rate": 5.075541664909833e-08, "loss": 0.1979, "step": 12602 }, { "epoch": 3.3536455561468865, "grad_norm": 0.2948434054851532, "learning_rate": 5.074056853633044e-08, "loss": 0.1852, "step": 12603 }, { "epoch": 3.3539116551357107, "grad_norm": 0.28611576557159424, "learning_rate": 5.0725721857376656e-08, "loss": 0.1872, "step": 12604 }, { "epoch": 3.3541777541245343, "grad_norm": 0.2986626625061035, "learning_rate": 5.071087661266917e-08, "loss": 0.202, "step": 12605 }, { "epoch": 3.354443853113358, "grad_norm": 0.27701622247695923, "learning_rate": 5.06960328026401e-08, "loss": 0.1797, "step": 12606 }, { "epoch": 3.354709952102182, "grad_norm": 0.28866246342658997, "learning_rate": 5.068119042772152e-08, "loss": 0.1888, "step": 12607 }, { "epoch": 3.354976051091006, "grad_norm": 0.2765878140926361, "learning_rate": 5.0666349488345405e-08, "loss": 0.1871, "step": 12608 }, { "epoch": 3.3552421500798295, "grad_norm": 0.28420066833496094, "learning_rate": 5.065150998494376e-08, "loss": 0.1699, "step": 12609 }, { "epoch": 3.3555082490686536, "grad_norm": 0.2882138788700104, "learning_rate": 5.0636671917948514e-08, "loss": 0.1985, "step": 12610 }, { "epoch": 3.3557743480574773, "grad_norm": 0.30397993326187134, "learning_rate": 5.0621835287791605e-08, "loss": 0.2137, "step": 12611 }, { "epoch": 3.3560404470463014, "grad_norm": 0.31156936287879944, "learning_rate": 5.060700009490484e-08, "loss": 0.1883, "step": 12612 }, { "epoch": 3.356306546035125, "grad_norm": 0.2863420248031616, "learning_rate": 5.0592166339720054e-08, "loss": 0.1807, "step": 12613 }, { "epoch": 3.356572645023949, "grad_norm": 0.2817707061767578, "learning_rate": 5.057733402266904e-08, "loss": 0.1959, "step": 12614 }, { "epoch": 3.356838744012773, "grad_norm": 0.3072241246700287, "learning_rate": 5.056250314418349e-08, "loss": 0.1859, "step": 12615 }, { "epoch": 3.3571048430015966, "grad_norm": 0.3307662904262543, "learning_rate": 5.054767370469515e-08, "loss": 0.2001, "step": 12616 }, { "epoch": 3.3573709419904203, "grad_norm": 0.5341781973838806, "learning_rate": 5.0532845704635585e-08, "loss": 0.2046, "step": 12617 }, { "epoch": 3.3576370409792444, "grad_norm": 0.3133532702922821, "learning_rate": 5.0518019144436454e-08, "loss": 0.1933, "step": 12618 }, { "epoch": 3.357903139968068, "grad_norm": 0.2943194508552551, "learning_rate": 5.0503194024529295e-08, "loss": 0.1928, "step": 12619 }, { "epoch": 3.358169238956892, "grad_norm": 0.26633429527282715, "learning_rate": 5.0488370345345656e-08, "loss": 0.1891, "step": 12620 }, { "epoch": 3.358435337945716, "grad_norm": 0.2768974006175995, "learning_rate": 5.047354810731699e-08, "loss": 0.1826, "step": 12621 }, { "epoch": 3.3587014369345396, "grad_norm": 0.2921583652496338, "learning_rate": 5.045872731087478e-08, "loss": 0.1977, "step": 12622 }, { "epoch": 3.3589675359233633, "grad_norm": 0.30514246225357056, "learning_rate": 5.0443907956450346e-08, "loss": 0.1911, "step": 12623 }, { "epoch": 3.3592336349121874, "grad_norm": 0.3037298619747162, "learning_rate": 5.0429090044475096e-08, "loss": 0.1976, "step": 12624 }, { "epoch": 3.359499733901011, "grad_norm": 0.32518401741981506, "learning_rate": 5.041427357538035e-08, "loss": 0.2111, "step": 12625 }, { "epoch": 3.3597658328898348, "grad_norm": 0.29672771692276, "learning_rate": 5.0399458549597315e-08, "loss": 0.1993, "step": 12626 }, { "epoch": 3.360031931878659, "grad_norm": 0.3005135655403137, "learning_rate": 5.0384644967557255e-08, "loss": 0.2113, "step": 12627 }, { "epoch": 3.3602980308674826, "grad_norm": 0.2955087125301361, "learning_rate": 5.0369832829691364e-08, "loss": 0.1988, "step": 12628 }, { "epoch": 3.3605641298563067, "grad_norm": 0.2976895570755005, "learning_rate": 5.0355022136430767e-08, "loss": 0.2051, "step": 12629 }, { "epoch": 3.3608302288451304, "grad_norm": 0.30098745226860046, "learning_rate": 5.034021288820658e-08, "loss": 0.1926, "step": 12630 }, { "epoch": 3.361096327833954, "grad_norm": 0.28729671239852905, "learning_rate": 5.0325405085449844e-08, "loss": 0.1894, "step": 12631 }, { "epoch": 3.361362426822778, "grad_norm": 0.33893319964408875, "learning_rate": 5.03105987285916e-08, "loss": 0.1839, "step": 12632 }, { "epoch": 3.361628525811602, "grad_norm": 0.29113298654556274, "learning_rate": 5.0295793818062805e-08, "loss": 0.1941, "step": 12633 }, { "epoch": 3.361894624800426, "grad_norm": 0.28016242384910583, "learning_rate": 5.028099035429443e-08, "loss": 0.1846, "step": 12634 }, { "epoch": 3.3621607237892497, "grad_norm": 0.2866830825805664, "learning_rate": 5.02661883377173e-08, "loss": 0.1994, "step": 12635 }, { "epoch": 3.3624268227780734, "grad_norm": 0.2964107394218445, "learning_rate": 5.02513877687623e-08, "loss": 0.1818, "step": 12636 }, { "epoch": 3.3626929217668975, "grad_norm": 0.30268800258636475, "learning_rate": 5.023658864786022e-08, "loss": 0.1944, "step": 12637 }, { "epoch": 3.362959020755721, "grad_norm": 0.31125134229660034, "learning_rate": 5.0221790975441836e-08, "loss": 0.1902, "step": 12638 }, { "epoch": 3.363225119744545, "grad_norm": 0.28927260637283325, "learning_rate": 5.020699475193787e-08, "loss": 0.1803, "step": 12639 }, { "epoch": 3.363491218733369, "grad_norm": 0.28057873249053955, "learning_rate": 5.019219997777899e-08, "loss": 0.2, "step": 12640 }, { "epoch": 3.3637573177221927, "grad_norm": 0.29020020365715027, "learning_rate": 5.017740665339587e-08, "loss": 0.1855, "step": 12641 }, { "epoch": 3.3640234167110163, "grad_norm": 0.2943904995918274, "learning_rate": 5.0162614779219106e-08, "loss": 0.2015, "step": 12642 }, { "epoch": 3.3642895156998405, "grad_norm": 0.31865164637565613, "learning_rate": 5.014782435567917e-08, "loss": 0.2023, "step": 12643 }, { "epoch": 3.364555614688664, "grad_norm": 0.32701531052589417, "learning_rate": 5.013303538320664e-08, "loss": 0.169, "step": 12644 }, { "epoch": 3.364821713677488, "grad_norm": 0.32028448581695557, "learning_rate": 5.011824786223202e-08, "loss": 0.2046, "step": 12645 }, { "epoch": 3.365087812666312, "grad_norm": 0.2834826707839966, "learning_rate": 5.0103461793185634e-08, "loss": 0.1824, "step": 12646 }, { "epoch": 3.3653539116551356, "grad_norm": 0.2969425320625305, "learning_rate": 5.008867717649793e-08, "loss": 0.1791, "step": 12647 }, { "epoch": 3.3656200106439593, "grad_norm": 0.3106900155544281, "learning_rate": 5.007389401259924e-08, "loss": 0.1972, "step": 12648 }, { "epoch": 3.3658861096327835, "grad_norm": 0.30727630853652954, "learning_rate": 5.005911230191986e-08, "loss": 0.1936, "step": 12649 }, { "epoch": 3.366152208621607, "grad_norm": 0.28268948197364807, "learning_rate": 5.004433204489005e-08, "loss": 0.1846, "step": 12650 }, { "epoch": 3.3664183076104313, "grad_norm": 0.2983185946941376, "learning_rate": 5.0029553241940024e-08, "loss": 0.1834, "step": 12651 }, { "epoch": 3.366684406599255, "grad_norm": 0.30247601866722107, "learning_rate": 5.0014775893499996e-08, "loss": 0.1984, "step": 12652 }, { "epoch": 3.3669505055880786, "grad_norm": 0.31425535678863525, "learning_rate": 5.000000000000002e-08, "loss": 0.1869, "step": 12653 }, { "epoch": 3.3672166045769027, "grad_norm": 0.2721482217311859, "learning_rate": 4.9985225561870216e-08, "loss": 0.1895, "step": 12654 }, { "epoch": 3.3674827035657264, "grad_norm": 0.29149171710014343, "learning_rate": 4.997045257954063e-08, "loss": 0.1796, "step": 12655 }, { "epoch": 3.36774880255455, "grad_norm": 0.2867054343223572, "learning_rate": 4.9955681053441276e-08, "loss": 0.1685, "step": 12656 }, { "epoch": 3.3680149015433742, "grad_norm": 0.3001534342765808, "learning_rate": 4.9940910984002114e-08, "loss": 0.1905, "step": 12657 }, { "epoch": 3.368281000532198, "grad_norm": 0.3011026382446289, "learning_rate": 4.992614237165305e-08, "loss": 0.1908, "step": 12658 }, { "epoch": 3.368547099521022, "grad_norm": 0.3020454943180084, "learning_rate": 4.991137521682397e-08, "loss": 0.1879, "step": 12659 }, { "epoch": 3.3688131985098457, "grad_norm": 0.2879796624183655, "learning_rate": 4.98966095199447e-08, "loss": 0.1872, "step": 12660 }, { "epoch": 3.3690792974986694, "grad_norm": 0.30175212025642395, "learning_rate": 4.988184528144507e-08, "loss": 0.1977, "step": 12661 }, { "epoch": 3.3693453964874935, "grad_norm": 0.2846330404281616, "learning_rate": 4.9867082501754755e-08, "loss": 0.2098, "step": 12662 }, { "epoch": 3.369611495476317, "grad_norm": 0.3144868314266205, "learning_rate": 4.98523211813035e-08, "loss": 0.2064, "step": 12663 }, { "epoch": 3.369877594465141, "grad_norm": 0.30143827199935913, "learning_rate": 4.9837561320521014e-08, "loss": 0.2035, "step": 12664 }, { "epoch": 3.370143693453965, "grad_norm": 0.2894722819328308, "learning_rate": 4.9822802919836824e-08, "loss": 0.1822, "step": 12665 }, { "epoch": 3.3704097924427887, "grad_norm": 0.30136042833328247, "learning_rate": 4.9808045979680566e-08, "loss": 0.2058, "step": 12666 }, { "epoch": 3.3706758914316124, "grad_norm": 0.3111685812473297, "learning_rate": 4.9793290500481766e-08, "loss": 0.2158, "step": 12667 }, { "epoch": 3.3709419904204365, "grad_norm": 0.28712382912635803, "learning_rate": 4.977853648266991e-08, "loss": 0.1801, "step": 12668 }, { "epoch": 3.37120808940926, "grad_norm": 0.29871058464050293, "learning_rate": 4.976378392667446e-08, "loss": 0.1932, "step": 12669 }, { "epoch": 3.371474188398084, "grad_norm": 0.27609166502952576, "learning_rate": 4.9749032832924874e-08, "loss": 0.191, "step": 12670 }, { "epoch": 3.371740287386908, "grad_norm": 0.2992389500141144, "learning_rate": 4.973428320185043e-08, "loss": 0.1803, "step": 12671 }, { "epoch": 3.3720063863757317, "grad_norm": 0.3005412817001343, "learning_rate": 4.971953503388047e-08, "loss": 0.1878, "step": 12672 }, { "epoch": 3.3722724853645554, "grad_norm": 0.334366112947464, "learning_rate": 4.9704788329444306e-08, "loss": 0.1906, "step": 12673 }, { "epoch": 3.3725385843533795, "grad_norm": 0.28355178236961365, "learning_rate": 4.969004308897116e-08, "loss": 0.1822, "step": 12674 }, { "epoch": 3.372804683342203, "grad_norm": 0.29816490411758423, "learning_rate": 4.967529931289023e-08, "loss": 0.2079, "step": 12675 }, { "epoch": 3.3730707823310273, "grad_norm": 0.2807747721672058, "learning_rate": 4.966055700163066e-08, "loss": 0.185, "step": 12676 }, { "epoch": 3.373336881319851, "grad_norm": 0.27627670764923096, "learning_rate": 4.964581615562159e-08, "loss": 0.1844, "step": 12677 }, { "epoch": 3.3736029803086747, "grad_norm": 0.27104437351226807, "learning_rate": 4.9631076775292056e-08, "loss": 0.1827, "step": 12678 }, { "epoch": 3.373869079297499, "grad_norm": 0.2798766791820526, "learning_rate": 4.9616338861071136e-08, "loss": 0.1828, "step": 12679 }, { "epoch": 3.3741351782863225, "grad_norm": 0.3197678327560425, "learning_rate": 4.960160241338774e-08, "loss": 0.1962, "step": 12680 }, { "epoch": 3.374401277275146, "grad_norm": 0.30328720808029175, "learning_rate": 4.958686743267083e-08, "loss": 0.1936, "step": 12681 }, { "epoch": 3.3746673762639703, "grad_norm": 0.2988664209842682, "learning_rate": 4.9572133919349326e-08, "loss": 0.2057, "step": 12682 }, { "epoch": 3.374933475252794, "grad_norm": 0.294968843460083, "learning_rate": 4.955740187385206e-08, "loss": 0.1896, "step": 12683 }, { "epoch": 3.375199574241618, "grad_norm": 0.29865145683288574, "learning_rate": 4.954267129660788e-08, "loss": 0.189, "step": 12684 }, { "epoch": 3.3754656732304418, "grad_norm": 0.28454723954200745, "learning_rate": 4.952794218804551e-08, "loss": 0.1907, "step": 12685 }, { "epoch": 3.3757317722192655, "grad_norm": 0.3158707916736603, "learning_rate": 4.951321454859368e-08, "loss": 0.2044, "step": 12686 }, { "epoch": 3.3759978712080896, "grad_norm": 0.3139215111732483, "learning_rate": 4.949848837868109e-08, "loss": 0.189, "step": 12687 }, { "epoch": 3.3762639701969133, "grad_norm": 0.2893131673336029, "learning_rate": 4.948376367873642e-08, "loss": 0.1713, "step": 12688 }, { "epoch": 3.376530069185737, "grad_norm": 0.31633487343788147, "learning_rate": 4.946904044918818e-08, "loss": 0.1928, "step": 12689 }, { "epoch": 3.376796168174561, "grad_norm": 0.2964870035648346, "learning_rate": 4.945431869046498e-08, "loss": 0.2147, "step": 12690 }, { "epoch": 3.3770622671633848, "grad_norm": 0.3131585121154785, "learning_rate": 4.943959840299532e-08, "loss": 0.2095, "step": 12691 }, { "epoch": 3.3773283661522084, "grad_norm": 0.294262170791626, "learning_rate": 4.942487958720767e-08, "loss": 0.2048, "step": 12692 }, { "epoch": 3.3775944651410326, "grad_norm": 0.2966375946998596, "learning_rate": 4.941016224353045e-08, "loss": 0.2074, "step": 12693 }, { "epoch": 3.3778605641298562, "grad_norm": 0.2875628173351288, "learning_rate": 4.9395446372392056e-08, "loss": 0.1765, "step": 12694 }, { "epoch": 3.37812666311868, "grad_norm": 0.3071123957633972, "learning_rate": 4.9380731974220835e-08, "loss": 0.1948, "step": 12695 }, { "epoch": 3.378392762107504, "grad_norm": 0.2851479649543762, "learning_rate": 4.936601904944506e-08, "loss": 0.2009, "step": 12696 }, { "epoch": 3.3786588610963277, "grad_norm": 0.29014045000076294, "learning_rate": 4.935130759849305e-08, "loss": 0.1983, "step": 12697 }, { "epoch": 3.378924960085152, "grad_norm": 0.2951735556125641, "learning_rate": 4.933659762179292e-08, "loss": 0.1686, "step": 12698 }, { "epoch": 3.3791910590739755, "grad_norm": 0.3185495436191559, "learning_rate": 4.93218891197729e-08, "loss": 0.2053, "step": 12699 }, { "epoch": 3.3794571580627992, "grad_norm": 0.3218490481376648, "learning_rate": 4.930718209286109e-08, "loss": 0.2115, "step": 12700 }, { "epoch": 3.3797232570516234, "grad_norm": 0.3276931941509247, "learning_rate": 4.929247654148559e-08, "loss": 0.1944, "step": 12701 }, { "epoch": 3.379989356040447, "grad_norm": 0.29852500557899475, "learning_rate": 4.927777246607444e-08, "loss": 0.1969, "step": 12702 }, { "epoch": 3.3802554550292707, "grad_norm": 0.2846297025680542, "learning_rate": 4.9263069867055676e-08, "loss": 0.1889, "step": 12703 }, { "epoch": 3.380521554018095, "grad_norm": 0.2970214784145355, "learning_rate": 4.924836874485717e-08, "loss": 0.1955, "step": 12704 }, { "epoch": 3.3807876530069185, "grad_norm": 0.29710716009140015, "learning_rate": 4.9233669099906874e-08, "loss": 0.1993, "step": 12705 }, { "epoch": 3.3810537519957427, "grad_norm": 0.3048202693462372, "learning_rate": 4.9218970932632697e-08, "loss": 0.2024, "step": 12706 }, { "epoch": 3.3813198509845663, "grad_norm": 0.30054160952568054, "learning_rate": 4.9204274243462385e-08, "loss": 0.1913, "step": 12707 }, { "epoch": 3.38158594997339, "grad_norm": 0.31752172112464905, "learning_rate": 4.918957903282376e-08, "loss": 0.2048, "step": 12708 }, { "epoch": 3.381852048962214, "grad_norm": 0.29783353209495544, "learning_rate": 4.9174885301144553e-08, "loss": 0.2042, "step": 12709 }, { "epoch": 3.382118147951038, "grad_norm": 0.30138707160949707, "learning_rate": 4.916019304885248e-08, "loss": 0.1928, "step": 12710 }, { "epoch": 3.3823842469398615, "grad_norm": 0.2996422350406647, "learning_rate": 4.914550227637517e-08, "loss": 0.2007, "step": 12711 }, { "epoch": 3.3826503459286856, "grad_norm": 0.3171217143535614, "learning_rate": 4.9130812984140256e-08, "loss": 0.1979, "step": 12712 }, { "epoch": 3.3829164449175093, "grad_norm": 0.2872687876224518, "learning_rate": 4.9116125172575305e-08, "loss": 0.1766, "step": 12713 }, { "epoch": 3.383182543906333, "grad_norm": 0.30633416771888733, "learning_rate": 4.9101438842107826e-08, "loss": 0.1861, "step": 12714 }, { "epoch": 3.383448642895157, "grad_norm": 0.30154964327812195, "learning_rate": 4.908675399316533e-08, "loss": 0.2071, "step": 12715 }, { "epoch": 3.383714741883981, "grad_norm": 0.2810789942741394, "learning_rate": 4.90720706261752e-08, "loss": 0.1793, "step": 12716 }, { "epoch": 3.3839808408728045, "grad_norm": 0.33471664786338806, "learning_rate": 4.905738874156486e-08, "loss": 0.1964, "step": 12717 }, { "epoch": 3.3842469398616286, "grad_norm": 0.31556034088134766, "learning_rate": 4.9042708339761675e-08, "loss": 0.1826, "step": 12718 }, { "epoch": 3.3845130388504523, "grad_norm": 0.2709023952484131, "learning_rate": 4.902802942119293e-08, "loss": 0.1805, "step": 12719 }, { "epoch": 3.384779137839276, "grad_norm": 0.2871451675891876, "learning_rate": 4.90133519862859e-08, "loss": 0.1742, "step": 12720 }, { "epoch": 3.3850452368281, "grad_norm": 0.2729576826095581, "learning_rate": 4.8998676035467814e-08, "loss": 0.1765, "step": 12721 }, { "epoch": 3.385311335816924, "grad_norm": 0.3000354468822479, "learning_rate": 4.898400156916585e-08, "loss": 0.1813, "step": 12722 }, { "epoch": 3.385577434805748, "grad_norm": 0.3110903799533844, "learning_rate": 4.8969328587807155e-08, "loss": 0.205, "step": 12723 }, { "epoch": 3.3858435337945716, "grad_norm": 0.2894182503223419, "learning_rate": 4.8954657091818786e-08, "loss": 0.185, "step": 12724 }, { "epoch": 3.3861096327833953, "grad_norm": 0.30576789379119873, "learning_rate": 4.89399870816278e-08, "loss": 0.195, "step": 12725 }, { "epoch": 3.3863757317722194, "grad_norm": 0.29708927869796753, "learning_rate": 4.892531855766125e-08, "loss": 0.1818, "step": 12726 }, { "epoch": 3.386641830761043, "grad_norm": 0.31987619400024414, "learning_rate": 4.891065152034605e-08, "loss": 0.201, "step": 12727 }, { "epoch": 3.3869079297498668, "grad_norm": 0.3042544424533844, "learning_rate": 4.889598597010911e-08, "loss": 0.1906, "step": 12728 }, { "epoch": 3.387174028738691, "grad_norm": 0.2764902412891388, "learning_rate": 4.888132190737733e-08, "loss": 0.1917, "step": 12729 }, { "epoch": 3.3874401277275146, "grad_norm": 0.26568159461021423, "learning_rate": 4.8866659332577545e-08, "loss": 0.1822, "step": 12730 }, { "epoch": 3.3877062267163387, "grad_norm": 0.31201425194740295, "learning_rate": 4.885199824613654e-08, "loss": 0.1931, "step": 12731 }, { "epoch": 3.3879723257051624, "grad_norm": 0.3151487112045288, "learning_rate": 4.883733864848105e-08, "loss": 0.1811, "step": 12732 }, { "epoch": 3.388238424693986, "grad_norm": 0.31218665838241577, "learning_rate": 4.8822680540037843e-08, "loss": 0.2081, "step": 12733 }, { "epoch": 3.38850452368281, "grad_norm": 0.30195584893226624, "learning_rate": 4.880802392123349e-08, "loss": 0.1889, "step": 12734 }, { "epoch": 3.388770622671634, "grad_norm": 0.30761298537254333, "learning_rate": 4.8793368792494624e-08, "loss": 0.1855, "step": 12735 }, { "epoch": 3.3890367216604576, "grad_norm": 0.3102484941482544, "learning_rate": 4.877871515424785e-08, "loss": 0.1908, "step": 12736 }, { "epoch": 3.3893028206492817, "grad_norm": 0.30522969365119934, "learning_rate": 4.876406300691967e-08, "loss": 0.1875, "step": 12737 }, { "epoch": 3.3895689196381054, "grad_norm": 0.2919481098651886, "learning_rate": 4.8749412350936604e-08, "loss": 0.1747, "step": 12738 }, { "epoch": 3.389835018626929, "grad_norm": 0.2929801046848297, "learning_rate": 4.873476318672506e-08, "loss": 0.1946, "step": 12739 }, { "epoch": 3.390101117615753, "grad_norm": 0.27897757291793823, "learning_rate": 4.8720115514711445e-08, "loss": 0.1751, "step": 12740 }, { "epoch": 3.390367216604577, "grad_norm": 0.2836616039276123, "learning_rate": 4.870546933532214e-08, "loss": 0.18, "step": 12741 }, { "epoch": 3.3906333155934005, "grad_norm": 0.33372825384140015, "learning_rate": 4.8690824648983455e-08, "loss": 0.199, "step": 12742 }, { "epoch": 3.3908994145822247, "grad_norm": 0.2961740493774414, "learning_rate": 4.867618145612161e-08, "loss": 0.1848, "step": 12743 }, { "epoch": 3.3911655135710483, "grad_norm": 0.3019292652606964, "learning_rate": 4.866153975716286e-08, "loss": 0.1991, "step": 12744 }, { "epoch": 3.3914316125598725, "grad_norm": 0.30461063981056213, "learning_rate": 4.864689955253339e-08, "loss": 0.2022, "step": 12745 }, { "epoch": 3.391697711548696, "grad_norm": 0.29890984296798706, "learning_rate": 4.863226084265939e-08, "loss": 0.1905, "step": 12746 }, { "epoch": 3.39196381053752, "grad_norm": 0.2985024154186249, "learning_rate": 4.8617623627966856e-08, "loss": 0.1943, "step": 12747 }, { "epoch": 3.392229909526344, "grad_norm": 0.29090818762779236, "learning_rate": 4.860298790888188e-08, "loss": 0.1915, "step": 12748 }, { "epoch": 3.3924960085151676, "grad_norm": 0.2995823621749878, "learning_rate": 4.858835368583046e-08, "loss": 0.2063, "step": 12749 }, { "epoch": 3.3927621075039913, "grad_norm": 0.3190643787384033, "learning_rate": 4.857372095923859e-08, "loss": 0.1909, "step": 12750 }, { "epoch": 3.3930282064928154, "grad_norm": 0.2848719656467438, "learning_rate": 4.855908972953222e-08, "loss": 0.1885, "step": 12751 }, { "epoch": 3.393294305481639, "grad_norm": 0.3338373303413391, "learning_rate": 4.854445999713714e-08, "loss": 0.1937, "step": 12752 }, { "epoch": 3.3935604044704633, "grad_norm": 0.30644646286964417, "learning_rate": 4.8529831762479234e-08, "loss": 0.2036, "step": 12753 }, { "epoch": 3.393826503459287, "grad_norm": 0.32979944348335266, "learning_rate": 4.8515205025984275e-08, "loss": 0.1855, "step": 12754 }, { "epoch": 3.3940926024481106, "grad_norm": 0.30413946509361267, "learning_rate": 4.8500579788078025e-08, "loss": 0.1916, "step": 12755 }, { "epoch": 3.3943587014369347, "grad_norm": 0.3030056357383728, "learning_rate": 4.848595604918617e-08, "loss": 0.1848, "step": 12756 }, { "epoch": 3.3946248004257584, "grad_norm": 0.2993912696838379, "learning_rate": 4.847133380973438e-08, "loss": 0.1931, "step": 12757 }, { "epoch": 3.394890899414582, "grad_norm": 0.28954729437828064, "learning_rate": 4.8456713070148284e-08, "loss": 0.1868, "step": 12758 }, { "epoch": 3.3951569984034062, "grad_norm": 0.28644993901252747, "learning_rate": 4.8442093830853425e-08, "loss": 0.185, "step": 12759 }, { "epoch": 3.39542309739223, "grad_norm": 0.320330947637558, "learning_rate": 4.8427476092275386e-08, "loss": 0.1922, "step": 12760 }, { "epoch": 3.3956891963810536, "grad_norm": 0.29904982447624207, "learning_rate": 4.841285985483958e-08, "loss": 0.2198, "step": 12761 }, { "epoch": 3.3959552953698777, "grad_norm": 0.3067435622215271, "learning_rate": 4.839824511897148e-08, "loss": 0.2066, "step": 12762 }, { "epoch": 3.3962213943587014, "grad_norm": 0.3094422221183777, "learning_rate": 4.8383631885096474e-08, "loss": 0.1928, "step": 12763 }, { "epoch": 3.396487493347525, "grad_norm": 0.2731805443763733, "learning_rate": 4.8369020153639926e-08, "loss": 0.1648, "step": 12764 }, { "epoch": 3.396753592336349, "grad_norm": 0.28578272461891174, "learning_rate": 4.8354409925027184e-08, "loss": 0.1868, "step": 12765 }, { "epoch": 3.397019691325173, "grad_norm": 0.2820766866207123, "learning_rate": 4.833980119968344e-08, "loss": 0.1718, "step": 12766 }, { "epoch": 3.3972857903139966, "grad_norm": 0.2864184081554413, "learning_rate": 4.832519397803394e-08, "loss": 0.1886, "step": 12767 }, { "epoch": 3.3975518893028207, "grad_norm": 0.30675867199897766, "learning_rate": 4.8310588260503874e-08, "loss": 0.2011, "step": 12768 }, { "epoch": 3.3978179882916444, "grad_norm": 0.2980903089046478, "learning_rate": 4.8295984047518415e-08, "loss": 0.1869, "step": 12769 }, { "epoch": 3.3980840872804685, "grad_norm": 0.39219874143600464, "learning_rate": 4.828138133950256e-08, "loss": 0.1752, "step": 12770 }, { "epoch": 3.398350186269292, "grad_norm": 0.29811620712280273, "learning_rate": 4.8266780136881425e-08, "loss": 0.1724, "step": 12771 }, { "epoch": 3.398616285258116, "grad_norm": 0.3672710955142975, "learning_rate": 4.825218044007998e-08, "loss": 0.1973, "step": 12772 }, { "epoch": 3.39888238424694, "grad_norm": 0.30680614709854126, "learning_rate": 4.823758224952321e-08, "loss": 0.1892, "step": 12773 }, { "epoch": 3.3991484832357637, "grad_norm": 0.3007359802722931, "learning_rate": 4.822298556563602e-08, "loss": 0.1998, "step": 12774 }, { "epoch": 3.3994145822245874, "grad_norm": 0.3066214621067047, "learning_rate": 4.820839038884329e-08, "loss": 0.1944, "step": 12775 }, { "epoch": 3.3996806812134115, "grad_norm": 0.328500360250473, "learning_rate": 4.819379671956984e-08, "loss": 0.2085, "step": 12776 }, { "epoch": 3.399946780202235, "grad_norm": 0.3017081022262573, "learning_rate": 4.8179204558240435e-08, "loss": 0.1894, "step": 12777 }, { "epoch": 3.4002128791910593, "grad_norm": 0.3068627715110779, "learning_rate": 4.816461390527988e-08, "loss": 0.1932, "step": 12778 }, { "epoch": 3.400478978179883, "grad_norm": 0.2887708842754364, "learning_rate": 4.8150024761112796e-08, "loss": 0.1852, "step": 12779 }, { "epoch": 3.4007450771687067, "grad_norm": 0.41582027077674866, "learning_rate": 4.813543712616386e-08, "loss": 0.2012, "step": 12780 }, { "epoch": 3.401011176157531, "grad_norm": 0.282180517911911, "learning_rate": 4.812085100085769e-08, "loss": 0.1791, "step": 12781 }, { "epoch": 3.4012772751463545, "grad_norm": 0.28526026010513306, "learning_rate": 4.810626638561883e-08, "loss": 0.1855, "step": 12782 }, { "epoch": 3.401543374135178, "grad_norm": 0.30629539489746094, "learning_rate": 4.809168328087182e-08, "loss": 0.224, "step": 12783 }, { "epoch": 3.4018094731240023, "grad_norm": 0.29905349016189575, "learning_rate": 4.8077101687041146e-08, "loss": 0.19, "step": 12784 }, { "epoch": 3.402075572112826, "grad_norm": 0.27972930669784546, "learning_rate": 4.8062521604551245e-08, "loss": 0.1853, "step": 12785 }, { "epoch": 3.4023416711016496, "grad_norm": 0.30209600925445557, "learning_rate": 4.804794303382645e-08, "loss": 0.1883, "step": 12786 }, { "epoch": 3.4026077700904738, "grad_norm": 0.30100443959236145, "learning_rate": 4.803336597529115e-08, "loss": 0.1922, "step": 12787 }, { "epoch": 3.4028738690792975, "grad_norm": 0.29967257380485535, "learning_rate": 4.801879042936967e-08, "loss": 0.1889, "step": 12788 }, { "epoch": 3.403139968068121, "grad_norm": 0.2694462239742279, "learning_rate": 4.80042163964862e-08, "loss": 0.176, "step": 12789 }, { "epoch": 3.4034060670569453, "grad_norm": 0.2690878212451935, "learning_rate": 4.798964387706497e-08, "loss": 0.1943, "step": 12790 }, { "epoch": 3.403672166045769, "grad_norm": 0.31331413984298706, "learning_rate": 4.7975072871530173e-08, "loss": 0.2044, "step": 12791 }, { "epoch": 3.4039382650345926, "grad_norm": 0.2656341791152954, "learning_rate": 4.796050338030593e-08, "loss": 0.171, "step": 12792 }, { "epoch": 3.4042043640234168, "grad_norm": 0.29407840967178345, "learning_rate": 4.794593540381631e-08, "loss": 0.1915, "step": 12793 }, { "epoch": 3.4044704630122404, "grad_norm": 0.29990723729133606, "learning_rate": 4.793136894248534e-08, "loss": 0.1986, "step": 12794 }, { "epoch": 3.4047365620010646, "grad_norm": 0.3011472821235657, "learning_rate": 4.7916803996737046e-08, "loss": 0.1859, "step": 12795 }, { "epoch": 3.4050026609898882, "grad_norm": 0.3379201591014862, "learning_rate": 4.7902240566995336e-08, "loss": 0.1985, "step": 12796 }, { "epoch": 3.405268759978712, "grad_norm": 0.3110799193382263, "learning_rate": 4.7887678653684186e-08, "loss": 0.186, "step": 12797 }, { "epoch": 3.405534858967536, "grad_norm": 0.3018118441104889, "learning_rate": 4.7873118257227355e-08, "loss": 0.1995, "step": 12798 }, { "epoch": 3.4058009579563597, "grad_norm": 0.2958218455314636, "learning_rate": 4.7858559378048705e-08, "loss": 0.1868, "step": 12799 }, { "epoch": 3.406067056945184, "grad_norm": 0.2993220090866089, "learning_rate": 4.7844002016572016e-08, "loss": 0.1997, "step": 12800 }, { "epoch": 3.4063331559340075, "grad_norm": 0.31846484541893005, "learning_rate": 4.7829446173221e-08, "loss": 0.2065, "step": 12801 }, { "epoch": 3.406599254922831, "grad_norm": 0.3135139048099518, "learning_rate": 4.7814891848419345e-08, "loss": 0.2017, "step": 12802 }, { "epoch": 3.4068653539116553, "grad_norm": 0.29916542768478394, "learning_rate": 4.78003390425907e-08, "loss": 0.2063, "step": 12803 }, { "epoch": 3.407131452900479, "grad_norm": 0.2994517982006073, "learning_rate": 4.778578775615868e-08, "loss": 0.1845, "step": 12804 }, { "epoch": 3.4073975518893027, "grad_norm": 0.277721107006073, "learning_rate": 4.777123798954677e-08, "loss": 0.1847, "step": 12805 }, { "epoch": 3.407663650878127, "grad_norm": 0.30760112404823303, "learning_rate": 4.7756689743178514e-08, "loss": 0.1981, "step": 12806 }, { "epoch": 3.4079297498669505, "grad_norm": 0.3285052478313446, "learning_rate": 4.77421430174774e-08, "loss": 0.2178, "step": 12807 }, { "epoch": 3.408195848855774, "grad_norm": 0.31228381395339966, "learning_rate": 4.7727597812866784e-08, "loss": 0.1852, "step": 12808 }, { "epoch": 3.4084619478445983, "grad_norm": 0.2971097230911255, "learning_rate": 4.771305412977007e-08, "loss": 0.1939, "step": 12809 }, { "epoch": 3.408728046833422, "grad_norm": 0.27016502618789673, "learning_rate": 4.76985119686106e-08, "loss": 0.1885, "step": 12810 }, { "epoch": 3.4089941458222457, "grad_norm": 0.2754317820072174, "learning_rate": 4.7683971329811633e-08, "loss": 0.1808, "step": 12811 }, { "epoch": 3.40926024481107, "grad_norm": 0.354219913482666, "learning_rate": 4.7669432213796437e-08, "loss": 0.2145, "step": 12812 }, { "epoch": 3.4095263437998935, "grad_norm": 0.27548667788505554, "learning_rate": 4.765489462098818e-08, "loss": 0.1686, "step": 12813 }, { "epoch": 3.409792442788717, "grad_norm": 0.28859421610832214, "learning_rate": 4.764035855181003e-08, "loss": 0.1857, "step": 12814 }, { "epoch": 3.4100585417775413, "grad_norm": 0.29215487837791443, "learning_rate": 4.762582400668513e-08, "loss": 0.1823, "step": 12815 }, { "epoch": 3.410324640766365, "grad_norm": 0.3030007481575012, "learning_rate": 4.7611290986036475e-08, "loss": 0.199, "step": 12816 }, { "epoch": 3.410590739755189, "grad_norm": 0.2621866762638092, "learning_rate": 4.759675949028712e-08, "loss": 0.1597, "step": 12817 }, { "epoch": 3.410856838744013, "grad_norm": 0.2879175543785095, "learning_rate": 4.758222951986002e-08, "loss": 0.173, "step": 12818 }, { "epoch": 3.4111229377328365, "grad_norm": 0.30288898944854736, "learning_rate": 4.756770107517811e-08, "loss": 0.2059, "step": 12819 }, { "epoch": 3.4113890367216606, "grad_norm": 0.3015115559101105, "learning_rate": 4.75531741566643e-08, "loss": 0.1748, "step": 12820 }, { "epoch": 3.4116551357104843, "grad_norm": 0.3027358055114746, "learning_rate": 4.753864876474141e-08, "loss": 0.1784, "step": 12821 }, { "epoch": 3.411921234699308, "grad_norm": 0.2905218303203583, "learning_rate": 4.752412489983223e-08, "loss": 0.1806, "step": 12822 }, { "epoch": 3.412187333688132, "grad_norm": 0.30171912908554077, "learning_rate": 4.750960256235953e-08, "loss": 0.1773, "step": 12823 }, { "epoch": 3.4124534326769558, "grad_norm": 0.28389590978622437, "learning_rate": 4.749508175274605e-08, "loss": 0.1867, "step": 12824 }, { "epoch": 3.41271953166578, "grad_norm": 0.3335951566696167, "learning_rate": 4.7480562471414366e-08, "loss": 0.2016, "step": 12825 }, { "epoch": 3.4129856306546036, "grad_norm": 0.2968043088912964, "learning_rate": 4.746604471878715e-08, "loss": 0.191, "step": 12826 }, { "epoch": 3.4132517296434273, "grad_norm": 0.29442286491394043, "learning_rate": 4.7451528495286993e-08, "loss": 0.1892, "step": 12827 }, { "epoch": 3.4135178286322514, "grad_norm": 0.27911409735679626, "learning_rate": 4.743701380133638e-08, "loss": 0.1825, "step": 12828 }, { "epoch": 3.413783927621075, "grad_norm": 0.2772868275642395, "learning_rate": 4.742250063735781e-08, "loss": 0.2047, "step": 12829 }, { "epoch": 3.4140500266098988, "grad_norm": 0.31782111525535583, "learning_rate": 4.740798900377373e-08, "loss": 0.1853, "step": 12830 }, { "epoch": 3.414316125598723, "grad_norm": 0.2854543924331665, "learning_rate": 4.7393478901006535e-08, "loss": 0.1948, "step": 12831 }, { "epoch": 3.4145822245875466, "grad_norm": 0.2854391634464264, "learning_rate": 4.7378970329478585e-08, "loss": 0.1718, "step": 12832 }, { "epoch": 3.4148483235763702, "grad_norm": 0.2700676918029785, "learning_rate": 4.736446328961221e-08, "loss": 0.1889, "step": 12833 }, { "epoch": 3.4151144225651944, "grad_norm": 0.33796852827072144, "learning_rate": 4.734995778182961e-08, "loss": 0.1893, "step": 12834 }, { "epoch": 3.415380521554018, "grad_norm": 0.3047862648963928, "learning_rate": 4.7335453806553026e-08, "loss": 0.1828, "step": 12835 }, { "epoch": 3.4156466205428417, "grad_norm": 0.2981768548488617, "learning_rate": 4.7320951364204644e-08, "loss": 0.1949, "step": 12836 }, { "epoch": 3.415912719531666, "grad_norm": 0.2861294150352478, "learning_rate": 4.7306450455206595e-08, "loss": 0.1752, "step": 12837 }, { "epoch": 3.4161788185204895, "grad_norm": 0.3207119107246399, "learning_rate": 4.729195107998094e-08, "loss": 0.1866, "step": 12838 }, { "epoch": 3.4164449175093132, "grad_norm": 0.2923467755317688, "learning_rate": 4.727745323894975e-08, "loss": 0.187, "step": 12839 }, { "epoch": 3.4167110164981374, "grad_norm": 0.2992873191833496, "learning_rate": 4.726295693253499e-08, "loss": 0.1947, "step": 12840 }, { "epoch": 3.416977115486961, "grad_norm": 0.287779301404953, "learning_rate": 4.724846216115863e-08, "loss": 0.1773, "step": 12841 }, { "epoch": 3.417243214475785, "grad_norm": 0.3324037194252014, "learning_rate": 4.7233968925242604e-08, "loss": 0.2081, "step": 12842 }, { "epoch": 3.417509313464609, "grad_norm": 0.3110204339027405, "learning_rate": 4.72194772252087e-08, "loss": 0.2137, "step": 12843 }, { "epoch": 3.4177754124534325, "grad_norm": 0.29383885860443115, "learning_rate": 4.720498706147879e-08, "loss": 0.1862, "step": 12844 }, { "epoch": 3.4180415114422567, "grad_norm": 0.28644418716430664, "learning_rate": 4.71904984344746e-08, "loss": 0.1844, "step": 12845 }, { "epoch": 3.4183076104310803, "grad_norm": 0.29556339979171753, "learning_rate": 4.717601134461793e-08, "loss": 0.1882, "step": 12846 }, { "epoch": 3.418573709419904, "grad_norm": 0.2865287959575653, "learning_rate": 4.716152579233037e-08, "loss": 0.1891, "step": 12847 }, { "epoch": 3.418839808408728, "grad_norm": 0.3041701018810272, "learning_rate": 4.7147041778033605e-08, "loss": 0.2019, "step": 12848 }, { "epoch": 3.419105907397552, "grad_norm": 0.29964855313301086, "learning_rate": 4.713255930214922e-08, "loss": 0.2019, "step": 12849 }, { "epoch": 3.419372006386376, "grad_norm": 0.287180632352829, "learning_rate": 4.7118078365098766e-08, "loss": 0.1939, "step": 12850 }, { "epoch": 3.4196381053751996, "grad_norm": 0.27533790469169617, "learning_rate": 4.7103598967303784e-08, "loss": 0.1801, "step": 12851 }, { "epoch": 3.4199042043640233, "grad_norm": 0.3090938925743103, "learning_rate": 4.708912110918566e-08, "loss": 0.1785, "step": 12852 }, { "epoch": 3.4201703033528474, "grad_norm": 0.29355862736701965, "learning_rate": 4.707464479116583e-08, "loss": 0.19, "step": 12853 }, { "epoch": 3.420436402341671, "grad_norm": 0.3104689121246338, "learning_rate": 4.7060170013665666e-08, "loss": 0.2154, "step": 12854 }, { "epoch": 3.420702501330495, "grad_norm": 0.3021823763847351, "learning_rate": 4.7045696777106505e-08, "loss": 0.1919, "step": 12855 }, { "epoch": 3.420968600319319, "grad_norm": 0.30811306834220886, "learning_rate": 4.703122508190961e-08, "loss": 0.2035, "step": 12856 }, { "epoch": 3.4212346993081426, "grad_norm": 0.3023792505264282, "learning_rate": 4.7016754928496226e-08, "loss": 0.2043, "step": 12857 }, { "epoch": 3.4215007982969663, "grad_norm": 0.3330523371696472, "learning_rate": 4.700228631728753e-08, "loss": 0.2223, "step": 12858 }, { "epoch": 3.4217668972857904, "grad_norm": 0.2988641858100891, "learning_rate": 4.6987819248704685e-08, "loss": 0.1813, "step": 12859 }, { "epoch": 3.422032996274614, "grad_norm": 0.2926095426082611, "learning_rate": 4.697335372316881e-08, "loss": 0.1791, "step": 12860 }, { "epoch": 3.422299095263438, "grad_norm": 0.30820658802986145, "learning_rate": 4.695888974110088e-08, "loss": 0.2007, "step": 12861 }, { "epoch": 3.422565194252262, "grad_norm": 0.275614470243454, "learning_rate": 4.694442730292195e-08, "loss": 0.1765, "step": 12862 }, { "epoch": 3.4228312932410856, "grad_norm": 0.28547921776771545, "learning_rate": 4.692996640905299e-08, "loss": 0.1995, "step": 12863 }, { "epoch": 3.4230973922299097, "grad_norm": 0.2906555235385895, "learning_rate": 4.6915507059914915e-08, "loss": 0.1899, "step": 12864 }, { "epoch": 3.4233634912187334, "grad_norm": 0.315923273563385, "learning_rate": 4.690104925592859e-08, "loss": 0.2039, "step": 12865 }, { "epoch": 3.423629590207557, "grad_norm": 0.27016618847846985, "learning_rate": 4.688659299751489e-08, "loss": 0.1737, "step": 12866 }, { "epoch": 3.423895689196381, "grad_norm": 0.28232863545417786, "learning_rate": 4.687213828509453e-08, "loss": 0.1866, "step": 12867 }, { "epoch": 3.424161788185205, "grad_norm": 0.2935064435005188, "learning_rate": 4.685768511908828e-08, "loss": 0.1895, "step": 12868 }, { "epoch": 3.4244278871740286, "grad_norm": 0.3034648597240448, "learning_rate": 4.6843233499916856e-08, "loss": 0.1928, "step": 12869 }, { "epoch": 3.4246939861628527, "grad_norm": 0.27671217918395996, "learning_rate": 4.682878342800086e-08, "loss": 0.1748, "step": 12870 }, { "epoch": 3.4249600851516764, "grad_norm": 0.2823885679244995, "learning_rate": 4.681433490376092e-08, "loss": 0.179, "step": 12871 }, { "epoch": 3.4252261841405005, "grad_norm": 0.2932959198951721, "learning_rate": 4.6799887927617587e-08, "loss": 0.1947, "step": 12872 }, { "epoch": 3.425492283129324, "grad_norm": 0.2750168442726135, "learning_rate": 4.678544249999139e-08, "loss": 0.178, "step": 12873 }, { "epoch": 3.425758382118148, "grad_norm": 0.29820510745048523, "learning_rate": 4.677099862130279e-08, "loss": 0.201, "step": 12874 }, { "epoch": 3.426024481106972, "grad_norm": 0.28178533911705017, "learning_rate": 4.6756556291972214e-08, "loss": 0.1857, "step": 12875 }, { "epoch": 3.4262905800957957, "grad_norm": 0.3026576340198517, "learning_rate": 4.674211551242004e-08, "loss": 0.1795, "step": 12876 }, { "epoch": 3.4265566790846194, "grad_norm": 0.3213546574115753, "learning_rate": 4.6727676283066595e-08, "loss": 0.2015, "step": 12877 }, { "epoch": 3.4268227780734435, "grad_norm": 0.32734236121177673, "learning_rate": 4.671323860433222e-08, "loss": 0.2135, "step": 12878 }, { "epoch": 3.427088877062267, "grad_norm": 0.30427807569503784, "learning_rate": 4.669880247663707e-08, "loss": 0.1876, "step": 12879 }, { "epoch": 3.427354976051091, "grad_norm": 0.2964710295200348, "learning_rate": 4.668436790040139e-08, "loss": 0.2046, "step": 12880 }, { "epoch": 3.427621075039915, "grad_norm": 0.3020710349082947, "learning_rate": 4.6669934876045314e-08, "loss": 0.1951, "step": 12881 }, { "epoch": 3.4278871740287387, "grad_norm": 0.30282673239707947, "learning_rate": 4.665550340398897e-08, "loss": 0.2059, "step": 12882 }, { "epoch": 3.4281532730175623, "grad_norm": 0.28937017917633057, "learning_rate": 4.6641073484652414e-08, "loss": 0.1794, "step": 12883 }, { "epoch": 3.4284193720063865, "grad_norm": 0.3015541434288025, "learning_rate": 4.662664511845568e-08, "loss": 0.1915, "step": 12884 }, { "epoch": 3.42868547099521, "grad_norm": 0.30499616265296936, "learning_rate": 4.6612218305818695e-08, "loss": 0.203, "step": 12885 }, { "epoch": 3.428951569984034, "grad_norm": 0.3006775677204132, "learning_rate": 4.659779304716147e-08, "loss": 0.197, "step": 12886 }, { "epoch": 3.429217668972858, "grad_norm": 0.32079005241394043, "learning_rate": 4.65833693429038e-08, "loss": 0.2011, "step": 12887 }, { "epoch": 3.4294837679616816, "grad_norm": 0.2657780945301056, "learning_rate": 4.656894719346553e-08, "loss": 0.1778, "step": 12888 }, { "epoch": 3.4297498669505058, "grad_norm": 0.31466421484947205, "learning_rate": 4.655452659926654e-08, "loss": 0.1992, "step": 12889 }, { "epoch": 3.4300159659393294, "grad_norm": 0.28360715508461, "learning_rate": 4.6540107560726463e-08, "loss": 0.1761, "step": 12890 }, { "epoch": 3.430282064928153, "grad_norm": 0.29282304644584656, "learning_rate": 4.6525690078265047e-08, "loss": 0.1961, "step": 12891 }, { "epoch": 3.4305481639169773, "grad_norm": 0.3005543351173401, "learning_rate": 4.651127415230194e-08, "loss": 0.1949, "step": 12892 }, { "epoch": 3.430814262905801, "grad_norm": 0.28868186473846436, "learning_rate": 4.649685978325677e-08, "loss": 0.1715, "step": 12893 }, { "epoch": 3.4310803618946246, "grad_norm": 0.28524184226989746, "learning_rate": 4.648244697154908e-08, "loss": 0.1853, "step": 12894 }, { "epoch": 3.4313464608834487, "grad_norm": 0.3137754797935486, "learning_rate": 4.646803571759842e-08, "loss": 0.1879, "step": 12895 }, { "epoch": 3.4316125598722724, "grad_norm": 1.16045081615448, "learning_rate": 4.6453626021824274e-08, "loss": 0.199, "step": 12896 }, { "epoch": 3.4318786588610966, "grad_norm": 0.31953731179237366, "learning_rate": 4.643921788464601e-08, "loss": 0.2031, "step": 12897 }, { "epoch": 3.4321447578499202, "grad_norm": 0.3098231852054596, "learning_rate": 4.642481130648303e-08, "loss": 0.2022, "step": 12898 }, { "epoch": 3.432410856838744, "grad_norm": 0.2931023836135864, "learning_rate": 4.64104062877547e-08, "loss": 0.1878, "step": 12899 }, { "epoch": 3.432676955827568, "grad_norm": 0.293687641620636, "learning_rate": 4.63960028288803e-08, "loss": 0.18, "step": 12900 }, { "epoch": 3.4329430548163917, "grad_norm": 0.3240000605583191, "learning_rate": 4.638160093027908e-08, "loss": 0.2248, "step": 12901 }, { "epoch": 3.4332091538052154, "grad_norm": 0.33451902866363525, "learning_rate": 4.6367200592370214e-08, "loss": 0.2012, "step": 12902 }, { "epoch": 3.4334752527940395, "grad_norm": 0.37272804975509644, "learning_rate": 4.6352801815572904e-08, "loss": 0.2011, "step": 12903 }, { "epoch": 3.433741351782863, "grad_norm": 0.2706755995750427, "learning_rate": 4.6338404600306237e-08, "loss": 0.1691, "step": 12904 }, { "epoch": 3.434007450771687, "grad_norm": 0.2938464283943176, "learning_rate": 4.6324008946989314e-08, "loss": 0.2047, "step": 12905 }, { "epoch": 3.434273549760511, "grad_norm": 0.28712427616119385, "learning_rate": 4.630961485604108e-08, "loss": 0.1889, "step": 12906 }, { "epoch": 3.4345396487493347, "grad_norm": 0.3150250017642975, "learning_rate": 4.629522232788057e-08, "loss": 0.2032, "step": 12907 }, { "epoch": 3.4348057477381584, "grad_norm": 0.32313835620880127, "learning_rate": 4.628083136292671e-08, "loss": 0.1973, "step": 12908 }, { "epoch": 3.4350718467269825, "grad_norm": 0.29401588439941406, "learning_rate": 4.6266441961598346e-08, "loss": 0.1966, "step": 12909 }, { "epoch": 3.435337945715806, "grad_norm": 0.29081183671951294, "learning_rate": 4.6252054124314334e-08, "loss": 0.2065, "step": 12910 }, { "epoch": 3.43560404470463, "grad_norm": 0.2943117320537567, "learning_rate": 4.6237667851493466e-08, "loss": 0.1882, "step": 12911 }, { "epoch": 3.435870143693454, "grad_norm": 0.3252730369567871, "learning_rate": 4.6223283143554506e-08, "loss": 0.228, "step": 12912 }, { "epoch": 3.4361362426822777, "grad_norm": 0.2938767671585083, "learning_rate": 4.620890000091614e-08, "loss": 0.1982, "step": 12913 }, { "epoch": 3.436402341671102, "grad_norm": 0.294938862323761, "learning_rate": 4.619451842399706e-08, "loss": 0.1986, "step": 12914 }, { "epoch": 3.4366684406599255, "grad_norm": 0.3093497157096863, "learning_rate": 4.618013841321581e-08, "loss": 0.1973, "step": 12915 }, { "epoch": 3.436934539648749, "grad_norm": 0.29812610149383545, "learning_rate": 4.616575996899099e-08, "loss": 0.1977, "step": 12916 }, { "epoch": 3.4372006386375733, "grad_norm": 0.2994501292705536, "learning_rate": 4.615138309174111e-08, "loss": 0.1917, "step": 12917 }, { "epoch": 3.437466737626397, "grad_norm": 0.28105148673057556, "learning_rate": 4.613700778188466e-08, "loss": 0.1835, "step": 12918 }, { "epoch": 3.437732836615221, "grad_norm": 0.3044438660144806, "learning_rate": 4.612263403984006e-08, "loss": 0.2036, "step": 12919 }, { "epoch": 3.437998935604045, "grad_norm": 0.289956271648407, "learning_rate": 4.610826186602569e-08, "loss": 0.1914, "step": 12920 }, { "epoch": 3.4382650345928685, "grad_norm": 0.28150761127471924, "learning_rate": 4.6093891260859894e-08, "loss": 0.1917, "step": 12921 }, { "epoch": 3.4385311335816926, "grad_norm": 0.31349778175354004, "learning_rate": 4.6079522224760946e-08, "loss": 0.2108, "step": 12922 }, { "epoch": 3.4387972325705163, "grad_norm": 0.308292418718338, "learning_rate": 4.606515475814715e-08, "loss": 0.191, "step": 12923 }, { "epoch": 3.43906333155934, "grad_norm": 0.3015095889568329, "learning_rate": 4.6050788861436615e-08, "loss": 0.19, "step": 12924 }, { "epoch": 3.439329430548164, "grad_norm": 0.3034827411174774, "learning_rate": 4.6036424535047544e-08, "loss": 0.2044, "step": 12925 }, { "epoch": 3.4395955295369878, "grad_norm": 0.2775036692619324, "learning_rate": 4.602206177939804e-08, "loss": 0.1792, "step": 12926 }, { "epoch": 3.4398616285258115, "grad_norm": 0.3036979138851166, "learning_rate": 4.600770059490616e-08, "loss": 0.2002, "step": 12927 }, { "epoch": 3.4401277275146356, "grad_norm": 0.3315128982067108, "learning_rate": 4.599334098198997e-08, "loss": 0.2151, "step": 12928 }, { "epoch": 3.4403938265034593, "grad_norm": 0.3163645565509796, "learning_rate": 4.5978982941067354e-08, "loss": 0.1927, "step": 12929 }, { "epoch": 3.440659925492283, "grad_norm": 0.29949021339416504, "learning_rate": 4.596462647255628e-08, "loss": 0.2044, "step": 12930 }, { "epoch": 3.440926024481107, "grad_norm": 0.30403974652290344, "learning_rate": 4.5950271576874636e-08, "loss": 0.1807, "step": 12931 }, { "epoch": 3.4411921234699308, "grad_norm": 0.2938075065612793, "learning_rate": 4.5935918254440276e-08, "loss": 0.1871, "step": 12932 }, { "epoch": 3.4414582224587544, "grad_norm": 0.28206562995910645, "learning_rate": 4.592156650567093e-08, "loss": 0.1789, "step": 12933 }, { "epoch": 3.4417243214475786, "grad_norm": 0.3033043444156647, "learning_rate": 4.5907216330984356e-08, "loss": 0.1801, "step": 12934 }, { "epoch": 3.4419904204364022, "grad_norm": 0.319553017616272, "learning_rate": 4.589286773079828e-08, "loss": 0.1993, "step": 12935 }, { "epoch": 3.4422565194252264, "grad_norm": 0.2817378044128418, "learning_rate": 4.587852070553032e-08, "loss": 0.1839, "step": 12936 }, { "epoch": 3.44252261841405, "grad_norm": 0.2775357961654663, "learning_rate": 4.5864175255598114e-08, "loss": 0.1682, "step": 12937 }, { "epoch": 3.4427887174028737, "grad_norm": 0.28880202770233154, "learning_rate": 4.5849831381419194e-08, "loss": 0.1845, "step": 12938 }, { "epoch": 3.443054816391698, "grad_norm": 0.298874169588089, "learning_rate": 4.583548908341108e-08, "loss": 0.178, "step": 12939 }, { "epoch": 3.4433209153805215, "grad_norm": 0.2737375795841217, "learning_rate": 4.582114836199126e-08, "loss": 0.19, "step": 12940 }, { "epoch": 3.4435870143693452, "grad_norm": 0.2839058041572571, "learning_rate": 4.580680921757716e-08, "loss": 0.1858, "step": 12941 }, { "epoch": 3.4438531133581693, "grad_norm": 0.29384320974349976, "learning_rate": 4.5792471650586094e-08, "loss": 0.1958, "step": 12942 }, { "epoch": 3.444119212346993, "grad_norm": 0.30861377716064453, "learning_rate": 4.577813566143544e-08, "loss": 0.1948, "step": 12943 }, { "epoch": 3.444385311335817, "grad_norm": 0.28922584652900696, "learning_rate": 4.5763801250542464e-08, "loss": 0.1894, "step": 12944 }, { "epoch": 3.444651410324641, "grad_norm": 0.29832515120506287, "learning_rate": 4.574946841832441e-08, "loss": 0.1963, "step": 12945 }, { "epoch": 3.4449175093134645, "grad_norm": 0.3002769947052002, "learning_rate": 4.573513716519847e-08, "loss": 0.1874, "step": 12946 }, { "epoch": 3.4451836083022886, "grad_norm": 0.3139705955982208, "learning_rate": 4.572080749158183e-08, "loss": 0.2014, "step": 12947 }, { "epoch": 3.4454497072911123, "grad_norm": 0.31430187821388245, "learning_rate": 4.5706479397891506e-08, "loss": 0.2261, "step": 12948 }, { "epoch": 3.445715806279936, "grad_norm": 0.30107003450393677, "learning_rate": 4.569215288454461e-08, "loss": 0.1921, "step": 12949 }, { "epoch": 3.44598190526876, "grad_norm": 0.3076641857624054, "learning_rate": 4.5677827951958156e-08, "loss": 0.1797, "step": 12950 }, { "epoch": 3.446248004257584, "grad_norm": 0.32076460123062134, "learning_rate": 4.566350460054905e-08, "loss": 0.2096, "step": 12951 }, { "epoch": 3.4465141032464075, "grad_norm": 0.33159446716308594, "learning_rate": 4.5649182830734246e-08, "loss": 0.209, "step": 12952 }, { "epoch": 3.4467802022352316, "grad_norm": 0.3331737220287323, "learning_rate": 4.563486264293062e-08, "loss": 0.214, "step": 12953 }, { "epoch": 3.4470463012240553, "grad_norm": 0.2957235276699066, "learning_rate": 4.562054403755496e-08, "loss": 0.1959, "step": 12954 }, { "epoch": 3.447312400212879, "grad_norm": 0.31826144456863403, "learning_rate": 4.560622701502409e-08, "loss": 0.1925, "step": 12955 }, { "epoch": 3.447578499201703, "grad_norm": 0.3193657696247101, "learning_rate": 4.55919115757547e-08, "loss": 0.1887, "step": 12956 }, { "epoch": 3.447844598190527, "grad_norm": 0.29027292132377625, "learning_rate": 4.5577597720163506e-08, "loss": 0.1821, "step": 12957 }, { "epoch": 3.4481106971793505, "grad_norm": 0.282061904668808, "learning_rate": 4.556328544866714e-08, "loss": 0.2038, "step": 12958 }, { "epoch": 3.4483767961681746, "grad_norm": 0.29890578985214233, "learning_rate": 4.5548974761682226e-08, "loss": 0.1848, "step": 12959 }, { "epoch": 3.4486428951569983, "grad_norm": 0.280588835477829, "learning_rate": 4.553466565962524e-08, "loss": 0.1742, "step": 12960 }, { "epoch": 3.4489089941458224, "grad_norm": 0.469747930765152, "learning_rate": 4.5520358142912704e-08, "loss": 0.1871, "step": 12961 }, { "epoch": 3.449175093134646, "grad_norm": 0.28882887959480286, "learning_rate": 4.550605221196111e-08, "loss": 0.1975, "step": 12962 }, { "epoch": 3.44944119212347, "grad_norm": 0.2978784143924713, "learning_rate": 4.5491747867186836e-08, "loss": 0.1892, "step": 12963 }, { "epoch": 3.449707291112294, "grad_norm": 0.30017879605293274, "learning_rate": 4.547744510900624e-08, "loss": 0.1912, "step": 12964 }, { "epoch": 3.4499733901011176, "grad_norm": 0.2899690866470337, "learning_rate": 4.546314393783567e-08, "loss": 0.1848, "step": 12965 }, { "epoch": 3.4502394890899413, "grad_norm": 0.3124080002307892, "learning_rate": 4.5448844354091366e-08, "loss": 0.2054, "step": 12966 }, { "epoch": 3.4505055880787654, "grad_norm": 0.31910911202430725, "learning_rate": 4.54345463581896e-08, "loss": 0.2041, "step": 12967 }, { "epoch": 3.450771687067589, "grad_norm": 0.29937973618507385, "learning_rate": 4.5420249950546464e-08, "loss": 0.1922, "step": 12968 }, { "epoch": 3.451037786056413, "grad_norm": 0.29326486587524414, "learning_rate": 4.5405955131578156e-08, "loss": 0.1745, "step": 12969 }, { "epoch": 3.451303885045237, "grad_norm": 0.30075880885124207, "learning_rate": 4.539166190170076e-08, "loss": 0.1978, "step": 12970 }, { "epoch": 3.4515699840340606, "grad_norm": 0.29079490900039673, "learning_rate": 4.537737026133027e-08, "loss": 0.1948, "step": 12971 }, { "epoch": 3.4518360830228847, "grad_norm": 0.297138512134552, "learning_rate": 4.536308021088271e-08, "loss": 0.1889, "step": 12972 }, { "epoch": 3.4521021820117084, "grad_norm": 0.29868200421333313, "learning_rate": 4.534879175077402e-08, "loss": 0.1727, "step": 12973 }, { "epoch": 3.452368281000532, "grad_norm": 0.29115092754364014, "learning_rate": 4.53345048814201e-08, "loss": 0.1902, "step": 12974 }, { "epoch": 3.452634379989356, "grad_norm": 0.2883329689502716, "learning_rate": 4.5320219603236806e-08, "loss": 0.1947, "step": 12975 }, { "epoch": 3.45290047897818, "grad_norm": 0.29694700241088867, "learning_rate": 4.5305935916639936e-08, "loss": 0.1853, "step": 12976 }, { "epoch": 3.4531665779670035, "grad_norm": 0.30747878551483154, "learning_rate": 4.529165382204531e-08, "loss": 0.1851, "step": 12977 }, { "epoch": 3.4534326769558277, "grad_norm": 0.2919085621833801, "learning_rate": 4.527737331986855e-08, "loss": 0.1861, "step": 12978 }, { "epoch": 3.4536987759446514, "grad_norm": 0.3045625686645508, "learning_rate": 4.526309441052536e-08, "loss": 0.2034, "step": 12979 }, { "epoch": 3.453964874933475, "grad_norm": 0.30373477935791016, "learning_rate": 4.524881709443138e-08, "loss": 0.1919, "step": 12980 }, { "epoch": 3.454230973922299, "grad_norm": 0.2830156683921814, "learning_rate": 4.5234541372002163e-08, "loss": 0.2025, "step": 12981 }, { "epoch": 3.454497072911123, "grad_norm": 0.29259559512138367, "learning_rate": 4.522026724365326e-08, "loss": 0.1835, "step": 12982 }, { "epoch": 3.454763171899947, "grad_norm": 0.30536532402038574, "learning_rate": 4.520599470980014e-08, "loss": 0.1888, "step": 12983 }, { "epoch": 3.4550292708887707, "grad_norm": 0.28821659088134766, "learning_rate": 4.519172377085825e-08, "loss": 0.1988, "step": 12984 }, { "epoch": 3.4552953698775943, "grad_norm": 0.30779266357421875, "learning_rate": 4.5177454427242956e-08, "loss": 0.1933, "step": 12985 }, { "epoch": 3.4555614688664185, "grad_norm": 0.3097136616706848, "learning_rate": 4.516318667936967e-08, "loss": 0.2126, "step": 12986 }, { "epoch": 3.455827567855242, "grad_norm": 0.28917914628982544, "learning_rate": 4.5148920527653587e-08, "loss": 0.1888, "step": 12987 }, { "epoch": 3.456093666844066, "grad_norm": 0.8671401143074036, "learning_rate": 4.513465597251002e-08, "loss": 0.2019, "step": 12988 }, { "epoch": 3.45635976583289, "grad_norm": 0.28241220116615295, "learning_rate": 4.51203930143542e-08, "loss": 0.1782, "step": 12989 }, { "epoch": 3.4566258648217136, "grad_norm": 0.3492102026939392, "learning_rate": 4.51061316536012e-08, "loss": 0.2029, "step": 12990 }, { "epoch": 3.4568919638105378, "grad_norm": 0.2830839157104492, "learning_rate": 4.5091871890666175e-08, "loss": 0.1858, "step": 12991 }, { "epoch": 3.4571580627993614, "grad_norm": 0.2947649359703064, "learning_rate": 4.5077613725964206e-08, "loss": 0.1827, "step": 12992 }, { "epoch": 3.457424161788185, "grad_norm": 0.3038511574268341, "learning_rate": 4.506335715991029e-08, "loss": 0.1841, "step": 12993 }, { "epoch": 3.4576902607770093, "grad_norm": 0.3168172538280487, "learning_rate": 4.50491021929194e-08, "loss": 0.1871, "step": 12994 }, { "epoch": 3.457956359765833, "grad_norm": 0.29267919063568115, "learning_rate": 4.50348488254065e-08, "loss": 0.1813, "step": 12995 }, { "epoch": 3.4582224587546566, "grad_norm": 0.2853470742702484, "learning_rate": 4.5020597057786405e-08, "loss": 0.1766, "step": 12996 }, { "epoch": 3.4584885577434807, "grad_norm": 0.3110971748828888, "learning_rate": 4.5006346890473976e-08, "loss": 0.2211, "step": 12997 }, { "epoch": 3.4587546567323044, "grad_norm": 0.33104759454727173, "learning_rate": 4.4992098323884e-08, "loss": 0.1919, "step": 12998 }, { "epoch": 3.459020755721128, "grad_norm": 0.2838345766067505, "learning_rate": 4.497785135843122e-08, "loss": 0.1781, "step": 12999 }, { "epoch": 3.4592868547099522, "grad_norm": 0.25084730982780457, "learning_rate": 4.4963605994530316e-08, "loss": 0.1594, "step": 13000 }, { "epoch": 3.459552953698776, "grad_norm": 0.2853078246116638, "learning_rate": 4.4949362232595946e-08, "loss": 0.1852, "step": 13001 }, { "epoch": 3.4598190526875996, "grad_norm": 0.2923905849456787, "learning_rate": 4.49351200730427e-08, "loss": 0.1831, "step": 13002 }, { "epoch": 3.4600851516764237, "grad_norm": 0.30935946106910706, "learning_rate": 4.492087951628514e-08, "loss": 0.18, "step": 13003 }, { "epoch": 3.4603512506652474, "grad_norm": 0.27850839495658875, "learning_rate": 4.49066405627378e-08, "loss": 0.194, "step": 13004 }, { "epoch": 3.460617349654071, "grad_norm": 0.2637026309967041, "learning_rate": 4.4892403212815065e-08, "loss": 0.1801, "step": 13005 }, { "epoch": 3.460883448642895, "grad_norm": 0.3162524104118347, "learning_rate": 4.487816746693138e-08, "loss": 0.1924, "step": 13006 }, { "epoch": 3.461149547631719, "grad_norm": 0.2857339382171631, "learning_rate": 4.4863933325501135e-08, "loss": 0.208, "step": 13007 }, { "epoch": 3.461415646620543, "grad_norm": 0.25830909609794617, "learning_rate": 4.4849700788938615e-08, "loss": 0.163, "step": 13008 }, { "epoch": 3.4616817456093667, "grad_norm": 0.3004642426967621, "learning_rate": 4.483546985765816e-08, "loss": 0.2233, "step": 13009 }, { "epoch": 3.4619478445981904, "grad_norm": 0.31335535645484924, "learning_rate": 4.48212405320739e-08, "loss": 0.2112, "step": 13010 }, { "epoch": 3.4622139435870145, "grad_norm": 0.3211962580680847, "learning_rate": 4.4807012812600055e-08, "loss": 0.2135, "step": 13011 }, { "epoch": 3.462480042575838, "grad_norm": 0.301874577999115, "learning_rate": 4.4792786699650764e-08, "loss": 0.1682, "step": 13012 }, { "epoch": 3.462746141564662, "grad_norm": 0.2726663053035736, "learning_rate": 4.477856219364015e-08, "loss": 0.1831, "step": 13013 }, { "epoch": 3.463012240553486, "grad_norm": 0.3026495575904846, "learning_rate": 4.476433929498216e-08, "loss": 0.1879, "step": 13014 }, { "epoch": 3.4632783395423097, "grad_norm": 0.26941293478012085, "learning_rate": 4.475011800409084e-08, "loss": 0.1856, "step": 13015 }, { "epoch": 3.463544438531134, "grad_norm": 0.3045472800731659, "learning_rate": 4.473589832138014e-08, "loss": 0.2039, "step": 13016 }, { "epoch": 3.4638105375199575, "grad_norm": 0.32093435525894165, "learning_rate": 4.472168024726394e-08, "loss": 0.1951, "step": 13017 }, { "epoch": 3.464076636508781, "grad_norm": 0.3461354374885559, "learning_rate": 4.4707463782156096e-08, "loss": 0.2041, "step": 13018 }, { "epoch": 3.4643427354976053, "grad_norm": 0.2976650595664978, "learning_rate": 4.469324892647042e-08, "loss": 0.1918, "step": 13019 }, { "epoch": 3.464608834486429, "grad_norm": 0.291533887386322, "learning_rate": 4.467903568062066e-08, "loss": 0.1925, "step": 13020 }, { "epoch": 3.4648749334752527, "grad_norm": 0.2866451144218445, "learning_rate": 4.466482404502054e-08, "loss": 0.2094, "step": 13021 }, { "epoch": 3.465141032464077, "grad_norm": 0.30110836029052734, "learning_rate": 4.4650614020083743e-08, "loss": 0.1873, "step": 13022 }, { "epoch": 3.4654071314529005, "grad_norm": 0.2801899015903473, "learning_rate": 4.463640560622384e-08, "loss": 0.1785, "step": 13023 }, { "epoch": 3.465673230441724, "grad_norm": 0.2995280921459198, "learning_rate": 4.462219880385441e-08, "loss": 0.1855, "step": 13024 }, { "epoch": 3.4659393294305483, "grad_norm": 0.3070882558822632, "learning_rate": 4.460799361338897e-08, "loss": 0.2044, "step": 13025 }, { "epoch": 3.466205428419372, "grad_norm": 0.2926807403564453, "learning_rate": 4.459379003524103e-08, "loss": 0.1798, "step": 13026 }, { "epoch": 3.4664715274081956, "grad_norm": 0.28904828429222107, "learning_rate": 4.457958806982399e-08, "loss": 0.1854, "step": 13027 }, { "epoch": 3.4667376263970198, "grad_norm": 0.2879863977432251, "learning_rate": 4.456538771755125e-08, "loss": 0.1764, "step": 13028 }, { "epoch": 3.4670037253858434, "grad_norm": 0.29139411449432373, "learning_rate": 4.4551188978836184e-08, "loss": 0.1927, "step": 13029 }, { "epoch": 3.467269824374667, "grad_norm": 0.3059544265270233, "learning_rate": 4.4536991854091985e-08, "loss": 0.2029, "step": 13030 }, { "epoch": 3.4675359233634913, "grad_norm": 0.3193403482437134, "learning_rate": 4.452279634373195e-08, "loss": 0.1942, "step": 13031 }, { "epoch": 3.467802022352315, "grad_norm": 0.2973202168941498, "learning_rate": 4.4508602448169305e-08, "loss": 0.1972, "step": 13032 }, { "epoch": 3.468068121341139, "grad_norm": 0.2968445122241974, "learning_rate": 4.449441016781713e-08, "loss": 0.1951, "step": 13033 }, { "epoch": 3.4683342203299627, "grad_norm": 0.29009777307510376, "learning_rate": 4.448021950308855e-08, "loss": 0.1826, "step": 13034 }, { "epoch": 3.4686003193187864, "grad_norm": 0.3206072151660919, "learning_rate": 4.446603045439662e-08, "loss": 0.2027, "step": 13035 }, { "epoch": 3.4688664183076106, "grad_norm": 0.3151213228702545, "learning_rate": 4.445184302215436e-08, "loss": 0.1771, "step": 13036 }, { "epoch": 3.4691325172964342, "grad_norm": 0.3265804350376129, "learning_rate": 4.4437657206774716e-08, "loss": 0.2151, "step": 13037 }, { "epoch": 3.4693986162852584, "grad_norm": 0.2964790165424347, "learning_rate": 4.442347300867061e-08, "loss": 0.2033, "step": 13038 }, { "epoch": 3.469664715274082, "grad_norm": 0.27309975028038025, "learning_rate": 4.4409290428254886e-08, "loss": 0.1903, "step": 13039 }, { "epoch": 3.4699308142629057, "grad_norm": 0.3005580008029938, "learning_rate": 4.4395109465940397e-08, "loss": 0.1961, "step": 13040 }, { "epoch": 3.47019691325173, "grad_norm": 0.29164955019950867, "learning_rate": 4.4380930122139924e-08, "loss": 0.1849, "step": 13041 }, { "epoch": 3.4704630122405535, "grad_norm": 0.26736247539520264, "learning_rate": 4.436675239726612e-08, "loss": 0.1767, "step": 13042 }, { "epoch": 3.470729111229377, "grad_norm": 0.2848007380962372, "learning_rate": 4.4352576291731704e-08, "loss": 0.194, "step": 13043 }, { "epoch": 3.4709952102182013, "grad_norm": 0.3269540071487427, "learning_rate": 4.433840180594931e-08, "loss": 0.2069, "step": 13044 }, { "epoch": 3.471261309207025, "grad_norm": 0.3470597565174103, "learning_rate": 4.432422894033152e-08, "loss": 0.2103, "step": 13045 }, { "epoch": 3.4715274081958487, "grad_norm": 0.2929776906967163, "learning_rate": 4.431005769529086e-08, "loss": 0.1704, "step": 13046 }, { "epoch": 3.471793507184673, "grad_norm": 0.28626880049705505, "learning_rate": 4.429588807123982e-08, "loss": 0.1927, "step": 13047 }, { "epoch": 3.4720596061734965, "grad_norm": 0.2960381805896759, "learning_rate": 4.428172006859088e-08, "loss": 0.1829, "step": 13048 }, { "epoch": 3.47232570516232, "grad_norm": 0.30319735407829285, "learning_rate": 4.4267553687756365e-08, "loss": 0.1758, "step": 13049 }, { "epoch": 3.4725918041511443, "grad_norm": 0.2966720759868622, "learning_rate": 4.425338892914865e-08, "loss": 0.1731, "step": 13050 }, { "epoch": 3.472857903139968, "grad_norm": 0.2944904863834381, "learning_rate": 4.423922579318008e-08, "loss": 0.1898, "step": 13051 }, { "epoch": 3.4731240021287917, "grad_norm": 0.28935325145721436, "learning_rate": 4.422506428026282e-08, "loss": 0.1969, "step": 13052 }, { "epoch": 3.473390101117616, "grad_norm": 0.3022843897342682, "learning_rate": 4.421090439080913e-08, "loss": 0.2152, "step": 13053 }, { "epoch": 3.4736562001064395, "grad_norm": 0.3034224808216095, "learning_rate": 4.419674612523117e-08, "loss": 0.1771, "step": 13054 }, { "epoch": 3.4739222990952636, "grad_norm": 0.29051104187965393, "learning_rate": 4.4182589483941015e-08, "loss": 0.1831, "step": 13055 }, { "epoch": 3.4741883980840873, "grad_norm": 0.2865974009037018, "learning_rate": 4.416843446735077e-08, "loss": 0.1768, "step": 13056 }, { "epoch": 3.474454497072911, "grad_norm": 0.2891337275505066, "learning_rate": 4.415428107587243e-08, "loss": 0.1727, "step": 13057 }, { "epoch": 3.474720596061735, "grad_norm": 0.2943684756755829, "learning_rate": 4.414012930991795e-08, "loss": 0.1906, "step": 13058 }, { "epoch": 3.474986695050559, "grad_norm": 0.27826476097106934, "learning_rate": 4.412597916989931e-08, "loss": 0.174, "step": 13059 }, { "epoch": 3.4752527940393825, "grad_norm": 0.30995798110961914, "learning_rate": 4.411183065622831e-08, "loss": 0.2105, "step": 13060 }, { "epoch": 3.4755188930282066, "grad_norm": 0.3067707121372223, "learning_rate": 4.4097683769316805e-08, "loss": 0.206, "step": 13061 }, { "epoch": 3.4757849920170303, "grad_norm": 0.2745855152606964, "learning_rate": 4.408353850957658e-08, "loss": 0.1729, "step": 13062 }, { "epoch": 3.4760510910058544, "grad_norm": 0.2970726788043976, "learning_rate": 4.406939487741936e-08, "loss": 0.193, "step": 13063 }, { "epoch": 3.476317189994678, "grad_norm": 0.30747491121292114, "learning_rate": 4.405525287325683e-08, "loss": 0.2014, "step": 13064 }, { "epoch": 3.4765832889835018, "grad_norm": 0.29439178109169006, "learning_rate": 4.404111249750063e-08, "loss": 0.1851, "step": 13065 }, { "epoch": 3.476849387972326, "grad_norm": 0.25178974866867065, "learning_rate": 4.4026973750562356e-08, "loss": 0.1752, "step": 13066 }, { "epoch": 3.4771154869611496, "grad_norm": 0.30815190076828003, "learning_rate": 4.4012836632853545e-08, "loss": 0.2045, "step": 13067 }, { "epoch": 3.4773815859499733, "grad_norm": 0.2999391555786133, "learning_rate": 4.399870114478572e-08, "loss": 0.2004, "step": 13068 }, { "epoch": 3.4776476849387974, "grad_norm": 0.3001325726509094, "learning_rate": 4.3984567286770254e-08, "loss": 0.1779, "step": 13069 }, { "epoch": 3.477913783927621, "grad_norm": 0.30317598581314087, "learning_rate": 4.3970435059218613e-08, "loss": 0.1854, "step": 13070 }, { "epoch": 3.4781798829164448, "grad_norm": 0.29344114661216736, "learning_rate": 4.395630446254215e-08, "loss": 0.2016, "step": 13071 }, { "epoch": 3.478445981905269, "grad_norm": 0.3006836473941803, "learning_rate": 4.3942175497152115e-08, "loss": 0.2069, "step": 13072 }, { "epoch": 3.4787120808940926, "grad_norm": 0.2763329744338989, "learning_rate": 4.39280481634598e-08, "loss": 0.1877, "step": 13073 }, { "epoch": 3.4789781798829162, "grad_norm": 0.2869108319282532, "learning_rate": 4.3913922461876406e-08, "loss": 0.196, "step": 13074 }, { "epoch": 3.4792442788717404, "grad_norm": 0.3034972548484802, "learning_rate": 4.3899798392813115e-08, "loss": 0.2023, "step": 13075 }, { "epoch": 3.479510377860564, "grad_norm": 0.29567751288414, "learning_rate": 4.3885675956681024e-08, "loss": 0.1923, "step": 13076 }, { "epoch": 3.4797764768493877, "grad_norm": 0.297864705324173, "learning_rate": 4.3871555153891247e-08, "loss": 0.1735, "step": 13077 }, { "epoch": 3.480042575838212, "grad_norm": 0.30890053510665894, "learning_rate": 4.385743598485472e-08, "loss": 0.1908, "step": 13078 }, { "epoch": 3.4803086748270355, "grad_norm": 0.2610379457473755, "learning_rate": 4.384331844998246e-08, "loss": 0.1687, "step": 13079 }, { "epoch": 3.4805747738158597, "grad_norm": 0.30303826928138733, "learning_rate": 4.382920254968539e-08, "loss": 0.1844, "step": 13080 }, { "epoch": 3.4808408728046834, "grad_norm": 0.2894327938556671, "learning_rate": 4.381508828437439e-08, "loss": 0.1924, "step": 13081 }, { "epoch": 3.481106971793507, "grad_norm": 0.2928641736507416, "learning_rate": 4.380097565446029e-08, "loss": 0.1847, "step": 13082 }, { "epoch": 3.481373070782331, "grad_norm": 0.3328501880168915, "learning_rate": 4.378686466035387e-08, "loss": 0.1958, "step": 13083 }, { "epoch": 3.481639169771155, "grad_norm": 0.3041445016860962, "learning_rate": 4.377275530246586e-08, "loss": 0.2051, "step": 13084 }, { "epoch": 3.4819052687599785, "grad_norm": 0.31825509667396545, "learning_rate": 4.3758647581206955e-08, "loss": 0.2043, "step": 13085 }, { "epoch": 3.4821713677488026, "grad_norm": 0.29068705439567566, "learning_rate": 4.374454149698782e-08, "loss": 0.1761, "step": 13086 }, { "epoch": 3.4824374667376263, "grad_norm": 0.29626432061195374, "learning_rate": 4.373043705021898e-08, "loss": 0.1869, "step": 13087 }, { "epoch": 3.4827035657264505, "grad_norm": 0.3385089635848999, "learning_rate": 4.3716334241311024e-08, "loss": 0.2055, "step": 13088 }, { "epoch": 3.482969664715274, "grad_norm": 0.34604108333587646, "learning_rate": 4.370223307067443e-08, "loss": 0.1747, "step": 13089 }, { "epoch": 3.483235763704098, "grad_norm": 0.3269534409046173, "learning_rate": 4.36881335387197e-08, "loss": 0.212, "step": 13090 }, { "epoch": 3.483501862692922, "grad_norm": 0.27216312289237976, "learning_rate": 4.367403564585714e-08, "loss": 0.1826, "step": 13091 }, { "epoch": 3.4837679616817456, "grad_norm": 0.2920125722885132, "learning_rate": 4.3659939392497167e-08, "loss": 0.1961, "step": 13092 }, { "epoch": 3.4840340606705693, "grad_norm": 0.2791695296764374, "learning_rate": 4.3645844779050065e-08, "loss": 0.1854, "step": 13093 }, { "epoch": 3.4843001596593934, "grad_norm": 0.3071378469467163, "learning_rate": 4.363175180592611e-08, "loss": 0.1983, "step": 13094 }, { "epoch": 3.484566258648217, "grad_norm": 0.27570950984954834, "learning_rate": 4.361766047353552e-08, "loss": 0.1743, "step": 13095 }, { "epoch": 3.484832357637041, "grad_norm": 0.31769487261772156, "learning_rate": 4.3603570782288415e-08, "loss": 0.2158, "step": 13096 }, { "epoch": 3.485098456625865, "grad_norm": 0.3324587345123291, "learning_rate": 4.358948273259493e-08, "loss": 0.2003, "step": 13097 }, { "epoch": 3.4853645556146886, "grad_norm": 0.2958568334579468, "learning_rate": 4.357539632486513e-08, "loss": 0.1991, "step": 13098 }, { "epoch": 3.4856306546035123, "grad_norm": 0.31430456042289734, "learning_rate": 4.356131155950905e-08, "loss": 0.214, "step": 13099 }, { "epoch": 3.4858967535923364, "grad_norm": 0.30379006266593933, "learning_rate": 4.354722843693663e-08, "loss": 0.1938, "step": 13100 }, { "epoch": 3.48616285258116, "grad_norm": 0.2914007306098938, "learning_rate": 4.353314695755783e-08, "loss": 0.1865, "step": 13101 }, { "epoch": 3.4864289515699842, "grad_norm": 0.31129953265190125, "learning_rate": 4.3519067121782505e-08, "loss": 0.1957, "step": 13102 }, { "epoch": 3.486695050558808, "grad_norm": 0.29824763536453247, "learning_rate": 4.350498893002048e-08, "loss": 0.181, "step": 13103 }, { "epoch": 3.4869611495476316, "grad_norm": 0.31596431136131287, "learning_rate": 4.349091238268158e-08, "loss": 0.1948, "step": 13104 }, { "epoch": 3.4872272485364557, "grad_norm": 0.3090001940727234, "learning_rate": 4.3476837480175444e-08, "loss": 0.1989, "step": 13105 }, { "epoch": 3.4874933475252794, "grad_norm": 0.30370989441871643, "learning_rate": 4.3462764222911815e-08, "loss": 0.1944, "step": 13106 }, { "epoch": 3.487759446514103, "grad_norm": 0.29422223567962646, "learning_rate": 4.344869261130033e-08, "loss": 0.1914, "step": 13107 }, { "epoch": 3.488025545502927, "grad_norm": 0.29010745882987976, "learning_rate": 4.343462264575057e-08, "loss": 0.1927, "step": 13108 }, { "epoch": 3.488291644491751, "grad_norm": 0.2969006299972534, "learning_rate": 4.342055432667208e-08, "loss": 0.1924, "step": 13109 }, { "epoch": 3.488557743480575, "grad_norm": 0.32204797863960266, "learning_rate": 4.3406487654474365e-08, "loss": 0.2084, "step": 13110 }, { "epoch": 3.4888238424693987, "grad_norm": 0.28119271993637085, "learning_rate": 4.339242262956682e-08, "loss": 0.1825, "step": 13111 }, { "epoch": 3.4890899414582224, "grad_norm": 0.3177731931209564, "learning_rate": 4.3378359252358874e-08, "loss": 0.2192, "step": 13112 }, { "epoch": 3.4893560404470465, "grad_norm": 0.3265717625617981, "learning_rate": 4.3364297523259906e-08, "loss": 0.2163, "step": 13113 }, { "epoch": 3.48962213943587, "grad_norm": 0.31586650013923645, "learning_rate": 4.335023744267916e-08, "loss": 0.1978, "step": 13114 }, { "epoch": 3.489888238424694, "grad_norm": 0.3009674549102783, "learning_rate": 4.3336179011025907e-08, "loss": 0.2076, "step": 13115 }, { "epoch": 3.490154337413518, "grad_norm": 0.29239511489868164, "learning_rate": 4.3322122228709346e-08, "loss": 0.1972, "step": 13116 }, { "epoch": 3.4904204364023417, "grad_norm": 0.29907485842704773, "learning_rate": 4.3308067096138654e-08, "loss": 0.2061, "step": 13117 }, { "epoch": 3.4906865353911654, "grad_norm": 0.2895635664463043, "learning_rate": 4.3294013613722934e-08, "loss": 0.1869, "step": 13118 }, { "epoch": 3.4909526343799895, "grad_norm": 0.30347928404808044, "learning_rate": 4.327996178187123e-08, "loss": 0.2071, "step": 13119 }, { "epoch": 3.491218733368813, "grad_norm": 0.29395630955696106, "learning_rate": 4.326591160099258e-08, "loss": 0.1803, "step": 13120 }, { "epoch": 3.491484832357637, "grad_norm": 0.2808205187320709, "learning_rate": 4.325186307149592e-08, "loss": 0.1966, "step": 13121 }, { "epoch": 3.491750931346461, "grad_norm": 0.2503916919231415, "learning_rate": 4.3237816193790233e-08, "loss": 0.1542, "step": 13122 }, { "epoch": 3.4920170303352847, "grad_norm": 0.28616970777511597, "learning_rate": 4.322377096828429e-08, "loss": 0.1949, "step": 13123 }, { "epoch": 3.4922831293241083, "grad_norm": 0.287710040807724, "learning_rate": 4.320972739538695e-08, "loss": 0.1769, "step": 13124 }, { "epoch": 3.4925492283129325, "grad_norm": 0.33625009655952454, "learning_rate": 4.3195685475507006e-08, "loss": 0.2, "step": 13125 }, { "epoch": 3.492815327301756, "grad_norm": 0.3080490529537201, "learning_rate": 4.318164520905315e-08, "loss": 0.1996, "step": 13126 }, { "epoch": 3.4930814262905803, "grad_norm": 0.31072375178337097, "learning_rate": 4.3167606596434094e-08, "loss": 0.1933, "step": 13127 }, { "epoch": 3.493347525279404, "grad_norm": 0.2914683520793915, "learning_rate": 4.315356963805843e-08, "loss": 0.1776, "step": 13128 }, { "epoch": 3.4936136242682276, "grad_norm": 0.3177751898765564, "learning_rate": 4.313953433433476e-08, "loss": 0.213, "step": 13129 }, { "epoch": 3.4938797232570518, "grad_norm": 0.28690654039382935, "learning_rate": 4.312550068567165e-08, "loss": 0.1751, "step": 13130 }, { "epoch": 3.4941458222458754, "grad_norm": 0.30310869216918945, "learning_rate": 4.311146869247749e-08, "loss": 0.1886, "step": 13131 }, { "epoch": 3.494411921234699, "grad_norm": 0.3076534867286682, "learning_rate": 4.3097438355160775e-08, "loss": 0.1933, "step": 13132 }, { "epoch": 3.4946780202235233, "grad_norm": 0.28338977694511414, "learning_rate": 4.308340967412992e-08, "loss": 0.1767, "step": 13133 }, { "epoch": 3.494944119212347, "grad_norm": 0.320621520280838, "learning_rate": 4.3069382649793195e-08, "loss": 0.2137, "step": 13134 }, { "epoch": 3.495210218201171, "grad_norm": 0.3033471703529358, "learning_rate": 4.3055357282558924e-08, "loss": 0.1823, "step": 13135 }, { "epoch": 3.4954763171899947, "grad_norm": 0.2979270815849304, "learning_rate": 4.304133357283534e-08, "loss": 0.2012, "step": 13136 }, { "epoch": 3.4957424161788184, "grad_norm": 0.3329060673713684, "learning_rate": 4.302731152103064e-08, "loss": 0.2001, "step": 13137 }, { "epoch": 3.4960085151676425, "grad_norm": 0.33174172043800354, "learning_rate": 4.3013291127552984e-08, "loss": 0.2203, "step": 13138 }, { "epoch": 3.4962746141564662, "grad_norm": 0.29234230518341064, "learning_rate": 4.299927239281045e-08, "loss": 0.2106, "step": 13139 }, { "epoch": 3.49654071314529, "grad_norm": 0.267817884683609, "learning_rate": 4.298525531721114e-08, "loss": 0.1805, "step": 13140 }, { "epoch": 3.496806812134114, "grad_norm": 0.30129143595695496, "learning_rate": 4.297123990116297e-08, "loss": 0.1846, "step": 13141 }, { "epoch": 3.4970729111229377, "grad_norm": 0.2760927081108093, "learning_rate": 4.295722614507393e-08, "loss": 0.182, "step": 13142 }, { "epoch": 3.4973390101117614, "grad_norm": 0.28168588876724243, "learning_rate": 4.2943214049351936e-08, "loss": 0.1908, "step": 13143 }, { "epoch": 3.4976051091005855, "grad_norm": 0.296619713306427, "learning_rate": 4.2929203614404826e-08, "loss": 0.1865, "step": 13144 }, { "epoch": 3.497871208089409, "grad_norm": 0.3064920902252197, "learning_rate": 4.291519484064042e-08, "loss": 0.1983, "step": 13145 }, { "epoch": 3.498137307078233, "grad_norm": 0.30035918951034546, "learning_rate": 4.290118772846647e-08, "loss": 0.192, "step": 13146 }, { "epoch": 3.498403406067057, "grad_norm": 0.2796039879322052, "learning_rate": 4.288718227829069e-08, "loss": 0.1845, "step": 13147 }, { "epoch": 3.4986695050558807, "grad_norm": 0.31185397505760193, "learning_rate": 4.2873178490520744e-08, "loss": 0.1995, "step": 13148 }, { "epoch": 3.4989356040447044, "grad_norm": 0.31083744764328003, "learning_rate": 4.285917636556429e-08, "loss": 0.2041, "step": 13149 }, { "epoch": 3.4992017030335285, "grad_norm": 0.33580896258354187, "learning_rate": 4.2845175903828786e-08, "loss": 0.2205, "step": 13150 }, { "epoch": 3.499467802022352, "grad_norm": 0.3372175097465515, "learning_rate": 4.2831177105721837e-08, "loss": 0.1996, "step": 13151 }, { "epoch": 3.4997339010111763, "grad_norm": 0.29102611541748047, "learning_rate": 4.281717997165091e-08, "loss": 0.1975, "step": 13152 }, { "epoch": 3.5, "grad_norm": 0.272541880607605, "learning_rate": 4.280318450202336e-08, "loss": 0.1847, "step": 13153 }, { "epoch": 3.5002660989888237, "grad_norm": 0.30946898460388184, "learning_rate": 4.27891906972466e-08, "loss": 0.1911, "step": 13154 }, { "epoch": 3.500532197977648, "grad_norm": 0.3038032054901123, "learning_rate": 4.277519855772795e-08, "loss": 0.1981, "step": 13155 }, { "epoch": 3.5007982969664715, "grad_norm": 0.34677430987358093, "learning_rate": 4.27612080838747e-08, "loss": 0.2097, "step": 13156 }, { "epoch": 3.5010643959552956, "grad_norm": 0.30024680495262146, "learning_rate": 4.274721927609406e-08, "loss": 0.1828, "step": 13157 }, { "epoch": 3.5013304949441193, "grad_norm": 0.3086603283882141, "learning_rate": 4.273323213479324e-08, "loss": 0.1927, "step": 13158 }, { "epoch": 3.501596593932943, "grad_norm": 0.31111428141593933, "learning_rate": 4.2719246660379304e-08, "loss": 0.1724, "step": 13159 }, { "epoch": 3.501862692921767, "grad_norm": 0.2785012423992157, "learning_rate": 4.270526285325938e-08, "loss": 0.1754, "step": 13160 }, { "epoch": 3.502128791910591, "grad_norm": 0.27920156717300415, "learning_rate": 4.269128071384048e-08, "loss": 0.184, "step": 13161 }, { "epoch": 3.5023948908994145, "grad_norm": 0.3068668246269226, "learning_rate": 4.26773002425296e-08, "loss": 0.183, "step": 13162 }, { "epoch": 3.5026609898882386, "grad_norm": 0.31341052055358887, "learning_rate": 4.2663321439733676e-08, "loss": 0.1846, "step": 13163 }, { "epoch": 3.5029270888770623, "grad_norm": 0.30758440494537354, "learning_rate": 4.264934430585959e-08, "loss": 0.1925, "step": 13164 }, { "epoch": 3.503193187865886, "grad_norm": 0.3096614480018616, "learning_rate": 4.263536884131419e-08, "loss": 0.1886, "step": 13165 }, { "epoch": 3.50345928685471, "grad_norm": 0.2890700399875641, "learning_rate": 4.262139504650425e-08, "loss": 0.1847, "step": 13166 }, { "epoch": 3.5037253858435338, "grad_norm": 0.2854529023170471, "learning_rate": 4.2607422921836546e-08, "loss": 0.2037, "step": 13167 }, { "epoch": 3.5039914848323575, "grad_norm": 0.2929900586605072, "learning_rate": 4.259345246771772e-08, "loss": 0.1888, "step": 13168 }, { "epoch": 3.5042575838211816, "grad_norm": 0.33013129234313965, "learning_rate": 4.257948368455443e-08, "loss": 0.2001, "step": 13169 }, { "epoch": 3.5045236828100053, "grad_norm": 0.2899761497974396, "learning_rate": 4.256551657275328e-08, "loss": 0.1851, "step": 13170 }, { "epoch": 3.504789781798829, "grad_norm": 0.29941925406455994, "learning_rate": 4.255155113272082e-08, "loss": 0.1822, "step": 13171 }, { "epoch": 3.505055880787653, "grad_norm": 0.27215078473091125, "learning_rate": 4.253758736486358e-08, "loss": 0.1666, "step": 13172 }, { "epoch": 3.5053219797764767, "grad_norm": 0.2769036293029785, "learning_rate": 4.252362526958793e-08, "loss": 0.1863, "step": 13173 }, { "epoch": 3.5055880787653004, "grad_norm": 0.3105587363243103, "learning_rate": 4.2509664847300306e-08, "loss": 0.2015, "step": 13174 }, { "epoch": 3.5058541777541246, "grad_norm": 0.2802845239639282, "learning_rate": 4.249570609840708e-08, "loss": 0.1869, "step": 13175 }, { "epoch": 3.5061202767429482, "grad_norm": 0.2974531352519989, "learning_rate": 4.2481749023314574e-08, "loss": 0.1863, "step": 13176 }, { "epoch": 3.5063863757317724, "grad_norm": 0.29572567343711853, "learning_rate": 4.246779362242897e-08, "loss": 0.1966, "step": 13177 }, { "epoch": 3.506652474720596, "grad_norm": 0.31395599246025085, "learning_rate": 4.245383989615652e-08, "loss": 0.1908, "step": 13178 }, { "epoch": 3.50691857370942, "grad_norm": 0.29185813665390015, "learning_rate": 4.243988784490336e-08, "loss": 0.1891, "step": 13179 }, { "epoch": 3.507184672698244, "grad_norm": 0.30540981888771057, "learning_rate": 4.2425937469075625e-08, "loss": 0.2098, "step": 13180 }, { "epoch": 3.5074507716870675, "grad_norm": 0.2915992736816406, "learning_rate": 4.241198876907935e-08, "loss": 0.1936, "step": 13181 }, { "epoch": 3.5077168706758917, "grad_norm": 0.28184714913368225, "learning_rate": 4.239804174532056e-08, "loss": 0.1796, "step": 13182 }, { "epoch": 3.5079829696647153, "grad_norm": 0.3176495432853699, "learning_rate": 4.238409639820522e-08, "loss": 0.1942, "step": 13183 }, { "epoch": 3.508249068653539, "grad_norm": 0.3551254868507385, "learning_rate": 4.237015272813923e-08, "loss": 0.1894, "step": 13184 }, { "epoch": 3.508515167642363, "grad_norm": 0.3073394298553467, "learning_rate": 4.2356210735528484e-08, "loss": 0.1856, "step": 13185 }, { "epoch": 3.508781266631187, "grad_norm": 0.36685243248939514, "learning_rate": 4.234227042077876e-08, "loss": 0.1925, "step": 13186 }, { "epoch": 3.5090473656200105, "grad_norm": 0.26832807064056396, "learning_rate": 4.232833178429582e-08, "loss": 0.1821, "step": 13187 }, { "epoch": 3.5093134646088346, "grad_norm": 0.31488335132598877, "learning_rate": 4.23143948264854e-08, "loss": 0.2003, "step": 13188 }, { "epoch": 3.5095795635976583, "grad_norm": 0.324947327375412, "learning_rate": 4.230045954775318e-08, "loss": 0.2015, "step": 13189 }, { "epoch": 3.509845662586482, "grad_norm": 0.29749616980552673, "learning_rate": 4.228652594850476e-08, "loss": 0.1829, "step": 13190 }, { "epoch": 3.510111761575306, "grad_norm": 0.2852365970611572, "learning_rate": 4.227259402914575e-08, "loss": 0.18, "step": 13191 }, { "epoch": 3.51037786056413, "grad_norm": 0.3225526511669159, "learning_rate": 4.2258663790081605e-08, "loss": 0.1915, "step": 13192 }, { "epoch": 3.5106439595529535, "grad_norm": 0.2925732135772705, "learning_rate": 4.224473523171783e-08, "loss": 0.1855, "step": 13193 }, { "epoch": 3.5109100585417776, "grad_norm": 0.29358750581741333, "learning_rate": 4.2230808354459903e-08, "loss": 0.1817, "step": 13194 }, { "epoch": 3.5111761575306013, "grad_norm": 0.28111666440963745, "learning_rate": 4.22168831587131e-08, "loss": 0.1914, "step": 13195 }, { "epoch": 3.511442256519425, "grad_norm": 0.29811087250709534, "learning_rate": 4.22029596448828e-08, "loss": 0.1881, "step": 13196 }, { "epoch": 3.511708355508249, "grad_norm": 0.28353527188301086, "learning_rate": 4.218903781337427e-08, "loss": 0.1912, "step": 13197 }, { "epoch": 3.511974454497073, "grad_norm": 0.3116469383239746, "learning_rate": 4.217511766459275e-08, "loss": 0.2019, "step": 13198 }, { "epoch": 3.5122405534858965, "grad_norm": 0.33289918303489685, "learning_rate": 4.216119919894342e-08, "loss": 0.1797, "step": 13199 }, { "epoch": 3.5125066524747206, "grad_norm": 0.320671021938324, "learning_rate": 4.2147282416831396e-08, "loss": 0.2091, "step": 13200 }, { "epoch": 3.5127727514635443, "grad_norm": 0.3088653087615967, "learning_rate": 4.2133367318661774e-08, "loss": 0.2003, "step": 13201 }, { "epoch": 3.5130388504523684, "grad_norm": 0.27512258291244507, "learning_rate": 4.211945390483956e-08, "loss": 0.1752, "step": 13202 }, { "epoch": 3.513304949441192, "grad_norm": 0.3310995101928711, "learning_rate": 4.210554217576981e-08, "loss": 0.2054, "step": 13203 }, { "epoch": 3.513571048430016, "grad_norm": 0.29426151514053345, "learning_rate": 4.209163213185738e-08, "loss": 0.1805, "step": 13204 }, { "epoch": 3.51383714741884, "grad_norm": 0.28193399310112, "learning_rate": 4.207772377350717e-08, "loss": 0.2027, "step": 13205 }, { "epoch": 3.5141032464076636, "grad_norm": 0.3065569996833801, "learning_rate": 4.2063817101124024e-08, "loss": 0.1974, "step": 13206 }, { "epoch": 3.5143693453964877, "grad_norm": 0.3100430369377136, "learning_rate": 4.204991211511274e-08, "loss": 0.1889, "step": 13207 }, { "epoch": 3.5146354443853114, "grad_norm": 0.3010364770889282, "learning_rate": 4.2036008815878046e-08, "loss": 0.1961, "step": 13208 }, { "epoch": 3.514901543374135, "grad_norm": 0.27824485301971436, "learning_rate": 4.202210720382463e-08, "loss": 0.1946, "step": 13209 }, { "epoch": 3.515167642362959, "grad_norm": 0.3276400864124298, "learning_rate": 4.2008207279357135e-08, "loss": 0.21, "step": 13210 }, { "epoch": 3.515433741351783, "grad_norm": 0.3131464123725891, "learning_rate": 4.199430904288019e-08, "loss": 0.198, "step": 13211 }, { "epoch": 3.5156998403406066, "grad_norm": 0.32737645506858826, "learning_rate": 4.198041249479827e-08, "loss": 0.1943, "step": 13212 }, { "epoch": 3.5159659393294307, "grad_norm": 0.27994200587272644, "learning_rate": 4.1966517635515886e-08, "loss": 0.2018, "step": 13213 }, { "epoch": 3.5162320383182544, "grad_norm": 0.31196358799934387, "learning_rate": 4.195262446543753e-08, "loss": 0.195, "step": 13214 }, { "epoch": 3.516498137307078, "grad_norm": 0.2666497528553009, "learning_rate": 4.1938732984967515e-08, "loss": 0.1837, "step": 13215 }, { "epoch": 3.516764236295902, "grad_norm": 0.29522228240966797, "learning_rate": 4.192484319451024e-08, "loss": 0.1906, "step": 13216 }, { "epoch": 3.517030335284726, "grad_norm": 0.30878961086273193, "learning_rate": 4.191095509446998e-08, "loss": 0.1901, "step": 13217 }, { "epoch": 3.5172964342735495, "grad_norm": 0.3043750822544098, "learning_rate": 4.1897068685250994e-08, "loss": 0.2002, "step": 13218 }, { "epoch": 3.5175625332623737, "grad_norm": 0.2739740312099457, "learning_rate": 4.1883183967257476e-08, "loss": 0.1738, "step": 13219 }, { "epoch": 3.5178286322511974, "grad_norm": 0.2729666531085968, "learning_rate": 4.1869300940893574e-08, "loss": 0.1667, "step": 13220 }, { "epoch": 3.518094731240021, "grad_norm": 0.5283333659172058, "learning_rate": 4.185541960656342e-08, "loss": 0.1841, "step": 13221 }, { "epoch": 3.518360830228845, "grad_norm": 0.32765164971351624, "learning_rate": 4.1841539964671e-08, "loss": 0.1945, "step": 13222 }, { "epoch": 3.518626929217669, "grad_norm": 0.28836289048194885, "learning_rate": 4.1827662015620334e-08, "loss": 0.1921, "step": 13223 }, { "epoch": 3.518893028206493, "grad_norm": 0.3275854289531708, "learning_rate": 4.18137857598154e-08, "loss": 0.2346, "step": 13224 }, { "epoch": 3.5191591271953166, "grad_norm": 0.3324703276157379, "learning_rate": 4.1799911197660073e-08, "loss": 0.1914, "step": 13225 }, { "epoch": 3.5194252261841408, "grad_norm": 0.28302785754203796, "learning_rate": 4.178603832955821e-08, "loss": 0.1883, "step": 13226 }, { "epoch": 3.5196913251729645, "grad_norm": 0.29777899384498596, "learning_rate": 4.177216715591363e-08, "loss": 0.1993, "step": 13227 }, { "epoch": 3.519957424161788, "grad_norm": 0.302831768989563, "learning_rate": 4.1758297677130084e-08, "loss": 0.203, "step": 13228 }, { "epoch": 3.5202235231506123, "grad_norm": 0.2819036841392517, "learning_rate": 4.174442989361126e-08, "loss": 0.1768, "step": 13229 }, { "epoch": 3.520489622139436, "grad_norm": 0.2859533727169037, "learning_rate": 4.173056380576085e-08, "loss": 0.1751, "step": 13230 }, { "epoch": 3.5207557211282596, "grad_norm": 0.3200201690196991, "learning_rate": 4.171669941398241e-08, "loss": 0.2304, "step": 13231 }, { "epoch": 3.5210218201170838, "grad_norm": 0.2840104401111603, "learning_rate": 4.170283671867951e-08, "loss": 0.1765, "step": 13232 }, { "epoch": 3.5212879191059074, "grad_norm": 0.30729731917381287, "learning_rate": 4.168897572025571e-08, "loss": 0.1808, "step": 13233 }, { "epoch": 3.521554018094731, "grad_norm": 0.3032781481742859, "learning_rate": 4.167511641911439e-08, "loss": 0.2048, "step": 13234 }, { "epoch": 3.5218201170835552, "grad_norm": 0.27776721119880676, "learning_rate": 4.166125881565899e-08, "loss": 0.1822, "step": 13235 }, { "epoch": 3.522086216072379, "grad_norm": 0.30236780643463135, "learning_rate": 4.164740291029286e-08, "loss": 0.1885, "step": 13236 }, { "epoch": 3.5223523150612026, "grad_norm": 0.3050132393836975, "learning_rate": 4.163354870341933e-08, "loss": 0.1973, "step": 13237 }, { "epoch": 3.5226184140500267, "grad_norm": 0.3334069848060608, "learning_rate": 4.161969619544164e-08, "loss": 0.2019, "step": 13238 }, { "epoch": 3.5228845130388504, "grad_norm": 0.2855297923088074, "learning_rate": 4.160584538676306e-08, "loss": 0.1835, "step": 13239 }, { "epoch": 3.523150612027674, "grad_norm": 0.3340846002101898, "learning_rate": 4.159199627778667e-08, "loss": 0.1859, "step": 13240 }, { "epoch": 3.5234167110164982, "grad_norm": 0.28483137488365173, "learning_rate": 4.15781488689156e-08, "loss": 0.1836, "step": 13241 }, { "epoch": 3.523682810005322, "grad_norm": 0.3012824058532715, "learning_rate": 4.1564303160552926e-08, "loss": 0.1985, "step": 13242 }, { "epoch": 3.5239489089941456, "grad_norm": 0.28441837430000305, "learning_rate": 4.155045915310167e-08, "loss": 0.188, "step": 13243 }, { "epoch": 3.5242150079829697, "grad_norm": 0.3078249394893646, "learning_rate": 4.153661684696478e-08, "loss": 0.1956, "step": 13244 }, { "epoch": 3.5244811069717934, "grad_norm": 0.29799506068229675, "learning_rate": 4.152277624254517e-08, "loss": 0.1743, "step": 13245 }, { "epoch": 3.524747205960617, "grad_norm": 0.3497644066810608, "learning_rate": 4.150893734024571e-08, "loss": 0.2182, "step": 13246 }, { "epoch": 3.525013304949441, "grad_norm": 0.2764475643634796, "learning_rate": 4.149510014046922e-08, "loss": 0.178, "step": 13247 }, { "epoch": 3.525279403938265, "grad_norm": 0.2833561897277832, "learning_rate": 4.1481264643618485e-08, "loss": 0.1896, "step": 13248 }, { "epoch": 3.525545502927089, "grad_norm": 0.3069465160369873, "learning_rate": 4.146743085009618e-08, "loss": 0.1996, "step": 13249 }, { "epoch": 3.5258116019159127, "grad_norm": 0.2714824378490448, "learning_rate": 4.1453598760304965e-08, "loss": 0.1933, "step": 13250 }, { "epoch": 3.526077700904737, "grad_norm": 0.3201304078102112, "learning_rate": 4.1439768374647475e-08, "loss": 0.2105, "step": 13251 }, { "epoch": 3.5263437998935605, "grad_norm": 0.31396377086639404, "learning_rate": 4.142593969352628e-08, "loss": 0.2217, "step": 13252 }, { "epoch": 3.526609898882384, "grad_norm": 0.29833748936653137, "learning_rate": 4.141211271734394e-08, "loss": 0.1771, "step": 13253 }, { "epoch": 3.5268759978712083, "grad_norm": 0.2828252613544464, "learning_rate": 4.1398287446502824e-08, "loss": 0.19, "step": 13254 }, { "epoch": 3.527142096860032, "grad_norm": 0.2921941578388214, "learning_rate": 4.138446388140542e-08, "loss": 0.1804, "step": 13255 }, { "epoch": 3.5274081958488557, "grad_norm": 0.30508285760879517, "learning_rate": 4.1370642022454075e-08, "loss": 0.1936, "step": 13256 }, { "epoch": 3.52767429483768, "grad_norm": 0.3075111210346222, "learning_rate": 4.135682187005114e-08, "loss": 0.1932, "step": 13257 }, { "epoch": 3.5279403938265035, "grad_norm": 0.2967357635498047, "learning_rate": 4.134300342459882e-08, "loss": 0.211, "step": 13258 }, { "epoch": 3.528206492815327, "grad_norm": 0.299209862947464, "learning_rate": 4.1329186686499364e-08, "loss": 0.2064, "step": 13259 }, { "epoch": 3.5284725918041513, "grad_norm": 0.3043988347053528, "learning_rate": 4.131537165615495e-08, "loss": 0.1766, "step": 13260 }, { "epoch": 3.528738690792975, "grad_norm": 0.3200879395008087, "learning_rate": 4.13015583339677e-08, "loss": 0.1986, "step": 13261 }, { "epoch": 3.5290047897817987, "grad_norm": 0.2931360900402069, "learning_rate": 4.128774672033968e-08, "loss": 0.1834, "step": 13262 }, { "epoch": 3.529270888770623, "grad_norm": 0.6699599623680115, "learning_rate": 4.1273936815672905e-08, "loss": 0.185, "step": 13263 }, { "epoch": 3.5295369877594465, "grad_norm": 0.29792559146881104, "learning_rate": 4.126012862036935e-08, "loss": 0.1901, "step": 13264 }, { "epoch": 3.52980308674827, "grad_norm": 0.3175484240055084, "learning_rate": 4.1246322134830924e-08, "loss": 0.2167, "step": 13265 }, { "epoch": 3.5300691857370943, "grad_norm": 0.31279507279396057, "learning_rate": 4.123251735945955e-08, "loss": 0.1829, "step": 13266 }, { "epoch": 3.530335284725918, "grad_norm": 0.3242698609828949, "learning_rate": 4.1218714294656976e-08, "loss": 0.1867, "step": 13267 }, { "epoch": 3.5306013837147416, "grad_norm": 0.27090945839881897, "learning_rate": 4.1204912940825e-08, "loss": 0.182, "step": 13268 }, { "epoch": 3.5308674827035658, "grad_norm": 0.3377320170402527, "learning_rate": 4.119111329836535e-08, "loss": 0.2099, "step": 13269 }, { "epoch": 3.5311335816923894, "grad_norm": 0.30733948945999146, "learning_rate": 4.11773153676797e-08, "loss": 0.2202, "step": 13270 }, { "epoch": 3.5313996806812136, "grad_norm": 0.2694568634033203, "learning_rate": 4.116351914916967e-08, "loss": 0.1634, "step": 13271 }, { "epoch": 3.5316657796700373, "grad_norm": 0.3586384057998657, "learning_rate": 4.114972464323682e-08, "loss": 0.204, "step": 13272 }, { "epoch": 3.531931878658861, "grad_norm": 0.28646737337112427, "learning_rate": 4.113593185028272e-08, "loss": 0.186, "step": 13273 }, { "epoch": 3.532197977647685, "grad_norm": 0.2884370982646942, "learning_rate": 4.112214077070877e-08, "loss": 0.1681, "step": 13274 }, { "epoch": 3.5324640766365087, "grad_norm": 0.30849263072013855, "learning_rate": 4.1108351404916433e-08, "loss": 0.1818, "step": 13275 }, { "epoch": 3.532730175625333, "grad_norm": 0.28146007657051086, "learning_rate": 4.109456375330711e-08, "loss": 0.1888, "step": 13276 }, { "epoch": 3.5329962746141566, "grad_norm": 0.29326754808425903, "learning_rate": 4.108077781628205e-08, "loss": 0.1941, "step": 13277 }, { "epoch": 3.5332623736029802, "grad_norm": 0.31395643949508667, "learning_rate": 4.106699359424256e-08, "loss": 0.2047, "step": 13278 }, { "epoch": 3.5335284725918044, "grad_norm": 0.2950383424758911, "learning_rate": 4.105321108758988e-08, "loss": 0.2045, "step": 13279 }, { "epoch": 3.533794571580628, "grad_norm": 0.30553537607192993, "learning_rate": 4.103943029672516e-08, "loss": 0.1835, "step": 13280 }, { "epoch": 3.5340606705694517, "grad_norm": 0.2779926657676697, "learning_rate": 4.102565122204955e-08, "loss": 0.1962, "step": 13281 }, { "epoch": 3.534326769558276, "grad_norm": 0.2827915549278259, "learning_rate": 4.1011873863964096e-08, "loss": 0.1715, "step": 13282 }, { "epoch": 3.5345928685470995, "grad_norm": 0.2844884991645813, "learning_rate": 4.099809822286984e-08, "loss": 0.1807, "step": 13283 }, { "epoch": 3.534858967535923, "grad_norm": 0.3014462888240814, "learning_rate": 4.0984324299167784e-08, "loss": 0.1959, "step": 13284 }, { "epoch": 3.5351250665247473, "grad_norm": 0.2906404137611389, "learning_rate": 4.097055209325877e-08, "loss": 0.1962, "step": 13285 }, { "epoch": 3.535391165513571, "grad_norm": 0.31680455803871155, "learning_rate": 4.095678160554372e-08, "loss": 0.1961, "step": 13286 }, { "epoch": 3.5356572645023947, "grad_norm": 0.2911186218261719, "learning_rate": 4.094301283642345e-08, "loss": 0.191, "step": 13287 }, { "epoch": 3.535923363491219, "grad_norm": 0.31924164295196533, "learning_rate": 4.092924578629875e-08, "loss": 0.1889, "step": 13288 }, { "epoch": 3.5361894624800425, "grad_norm": 0.2937607169151306, "learning_rate": 4.091548045557032e-08, "loss": 0.1857, "step": 13289 }, { "epoch": 3.536455561468866, "grad_norm": 0.4815433919429779, "learning_rate": 4.0901716844638835e-08, "loss": 0.2013, "step": 13290 }, { "epoch": 3.5367216604576903, "grad_norm": 0.3040670156478882, "learning_rate": 4.0887954953904934e-08, "loss": 0.2092, "step": 13291 }, { "epoch": 3.536987759446514, "grad_norm": 0.29256781935691833, "learning_rate": 4.087419478376922e-08, "loss": 0.1961, "step": 13292 }, { "epoch": 3.5372538584353377, "grad_norm": 0.28221067786216736, "learning_rate": 4.086043633463214e-08, "loss": 0.1816, "step": 13293 }, { "epoch": 3.537519957424162, "grad_norm": 0.28966501355171204, "learning_rate": 4.08466796068942e-08, "loss": 0.1982, "step": 13294 }, { "epoch": 3.5377860564129855, "grad_norm": 0.28253528475761414, "learning_rate": 4.083292460095586e-08, "loss": 0.1772, "step": 13295 }, { "epoch": 3.5380521554018096, "grad_norm": 0.30128711462020874, "learning_rate": 4.081917131721743e-08, "loss": 0.1868, "step": 13296 }, { "epoch": 3.5383182543906333, "grad_norm": 0.2939569056034088, "learning_rate": 4.080541975607926e-08, "loss": 0.1846, "step": 13297 }, { "epoch": 3.5385843533794574, "grad_norm": 0.2974978983402252, "learning_rate": 4.079166991794162e-08, "loss": 0.2084, "step": 13298 }, { "epoch": 3.538850452368281, "grad_norm": 0.2926815450191498, "learning_rate": 4.077792180320474e-08, "loss": 0.1894, "step": 13299 }, { "epoch": 3.539116551357105, "grad_norm": 0.3154156804084778, "learning_rate": 4.076417541226879e-08, "loss": 0.1839, "step": 13300 }, { "epoch": 3.539382650345929, "grad_norm": 0.3194619417190552, "learning_rate": 4.075043074553388e-08, "loss": 0.1897, "step": 13301 }, { "epoch": 3.5396487493347526, "grad_norm": 0.2959306240081787, "learning_rate": 4.07366878034001e-08, "loss": 0.1995, "step": 13302 }, { "epoch": 3.5399148483235763, "grad_norm": 0.31544309854507446, "learning_rate": 4.072294658626749e-08, "loss": 0.1787, "step": 13303 }, { "epoch": 3.5401809473124004, "grad_norm": 0.28459635376930237, "learning_rate": 4.070920709453597e-08, "loss": 0.1891, "step": 13304 }, { "epoch": 3.540447046301224, "grad_norm": 0.28889206051826477, "learning_rate": 4.069546932860548e-08, "loss": 0.1881, "step": 13305 }, { "epoch": 3.5407131452900478, "grad_norm": 0.2794574201107025, "learning_rate": 4.0681733288875886e-08, "loss": 0.1832, "step": 13306 }, { "epoch": 3.540979244278872, "grad_norm": 0.2908712327480316, "learning_rate": 4.066799897574703e-08, "loss": 0.1823, "step": 13307 }, { "epoch": 3.5412453432676956, "grad_norm": 0.2900185286998749, "learning_rate": 4.0654266389618665e-08, "loss": 0.1752, "step": 13308 }, { "epoch": 3.5415114422565193, "grad_norm": 0.28420209884643555, "learning_rate": 4.0640535530890506e-08, "loss": 0.1836, "step": 13309 }, { "epoch": 3.5417775412453434, "grad_norm": 0.2934598922729492, "learning_rate": 4.062680639996224e-08, "loss": 0.1985, "step": 13310 }, { "epoch": 3.542043640234167, "grad_norm": 0.28850576281547546, "learning_rate": 4.061307899723349e-08, "loss": 0.1698, "step": 13311 }, { "epoch": 3.5423097392229907, "grad_norm": 0.3093738853931427, "learning_rate": 4.0599353323103836e-08, "loss": 0.2004, "step": 13312 }, { "epoch": 3.542575838211815, "grad_norm": 0.2998066842556, "learning_rate": 4.0585629377972744e-08, "loss": 0.1974, "step": 13313 }, { "epoch": 3.5428419372006386, "grad_norm": 0.2882704734802246, "learning_rate": 4.0571907162239706e-08, "loss": 0.1846, "step": 13314 }, { "epoch": 3.5431080361894622, "grad_norm": 0.27818387746810913, "learning_rate": 4.0558186676304184e-08, "loss": 0.197, "step": 13315 }, { "epoch": 3.5433741351782864, "grad_norm": 0.28673356771469116, "learning_rate": 4.054446792056548e-08, "loss": 0.1893, "step": 13316 }, { "epoch": 3.54364023416711, "grad_norm": 0.30993136763572693, "learning_rate": 4.053075089542294e-08, "loss": 0.1926, "step": 13317 }, { "epoch": 3.5439063331559337, "grad_norm": 0.2958936393260956, "learning_rate": 4.0517035601275826e-08, "loss": 0.2071, "step": 13318 }, { "epoch": 3.544172432144758, "grad_norm": 0.29046398401260376, "learning_rate": 4.0503322038523355e-08, "loss": 0.1859, "step": 13319 }, { "epoch": 3.5444385311335815, "grad_norm": 0.32735058665275574, "learning_rate": 4.0489610207564704e-08, "loss": 0.1969, "step": 13320 }, { "epoch": 3.5447046301224057, "grad_norm": 0.3206937611103058, "learning_rate": 4.047590010879902e-08, "loss": 0.1968, "step": 13321 }, { "epoch": 3.5449707291112293, "grad_norm": 0.31321099400520325, "learning_rate": 4.04621917426253e-08, "loss": 0.2049, "step": 13322 }, { "epoch": 3.5452368281000535, "grad_norm": 0.2758359909057617, "learning_rate": 4.0448485109442584e-08, "loss": 0.1838, "step": 13323 }, { "epoch": 3.545502927088877, "grad_norm": 0.28333497047424316, "learning_rate": 4.0434780209649847e-08, "loss": 0.1918, "step": 13324 }, { "epoch": 3.545769026077701, "grad_norm": 0.2981727421283722, "learning_rate": 4.042107704364599e-08, "loss": 0.1921, "step": 13325 }, { "epoch": 3.546035125066525, "grad_norm": 0.29050394892692566, "learning_rate": 4.040737561182989e-08, "loss": 0.1837, "step": 13326 }, { "epoch": 3.5463012240553486, "grad_norm": 0.323896199464798, "learning_rate": 4.039367591460036e-08, "loss": 0.187, "step": 13327 }, { "epoch": 3.5465673230441723, "grad_norm": 0.31987884640693665, "learning_rate": 4.0379977952356145e-08, "loss": 0.2026, "step": 13328 }, { "epoch": 3.5468334220329965, "grad_norm": 0.32353273034095764, "learning_rate": 4.0366281725495974e-08, "loss": 0.2171, "step": 13329 }, { "epoch": 3.54709952102182, "grad_norm": 0.29635605216026306, "learning_rate": 4.035258723441855e-08, "loss": 0.1816, "step": 13330 }, { "epoch": 3.547365620010644, "grad_norm": 0.30316635966300964, "learning_rate": 4.03388944795224e-08, "loss": 0.181, "step": 13331 }, { "epoch": 3.547631718999468, "grad_norm": 0.3520119786262512, "learning_rate": 4.032520346120611e-08, "loss": 0.2179, "step": 13332 }, { "epoch": 3.5478978179882916, "grad_norm": 0.26017871499061584, "learning_rate": 4.031151417986822e-08, "loss": 0.1725, "step": 13333 }, { "epoch": 3.5481639169771153, "grad_norm": 0.3107374906539917, "learning_rate": 4.0297826635907205e-08, "loss": 0.2057, "step": 13334 }, { "epoch": 3.5484300159659394, "grad_norm": 0.28617268800735474, "learning_rate": 4.0284140829721404e-08, "loss": 0.195, "step": 13335 }, { "epoch": 3.548696114954763, "grad_norm": 0.28973180055618286, "learning_rate": 4.02704567617092e-08, "loss": 0.1974, "step": 13336 }, { "epoch": 3.548962213943587, "grad_norm": 0.27296414971351624, "learning_rate": 4.025677443226893e-08, "loss": 0.1763, "step": 13337 }, { "epoch": 3.549228312932411, "grad_norm": 0.32625532150268555, "learning_rate": 4.0243093841798825e-08, "loss": 0.2188, "step": 13338 }, { "epoch": 3.5494944119212346, "grad_norm": 0.26348215341567993, "learning_rate": 4.022941499069714e-08, "loss": 0.1684, "step": 13339 }, { "epoch": 3.5497605109100583, "grad_norm": 0.291016161441803, "learning_rate": 4.0215737879361965e-08, "loss": 0.1842, "step": 13340 }, { "epoch": 3.5500266098988824, "grad_norm": 0.317863404750824, "learning_rate": 4.0202062508191424e-08, "loss": 0.1988, "step": 13341 }, { "epoch": 3.550292708887706, "grad_norm": 0.2821309566497803, "learning_rate": 4.018838887758358e-08, "loss": 0.1709, "step": 13342 }, { "epoch": 3.55055880787653, "grad_norm": 0.27920618653297424, "learning_rate": 4.017471698793643e-08, "loss": 0.1704, "step": 13343 }, { "epoch": 3.550824906865354, "grad_norm": 0.27021360397338867, "learning_rate": 4.016104683964794e-08, "loss": 0.1759, "step": 13344 }, { "epoch": 3.551091005854178, "grad_norm": 0.2978534996509552, "learning_rate": 4.014737843311601e-08, "loss": 0.2032, "step": 13345 }, { "epoch": 3.5513571048430017, "grad_norm": 0.3097611367702484, "learning_rate": 4.013371176873849e-08, "loss": 0.1932, "step": 13346 }, { "epoch": 3.5516232038318254, "grad_norm": 0.31206580996513367, "learning_rate": 4.012004684691317e-08, "loss": 0.222, "step": 13347 }, { "epoch": 3.5518893028206495, "grad_norm": 0.27643102407455444, "learning_rate": 4.010638366803785e-08, "loss": 0.1883, "step": 13348 }, { "epoch": 3.552155401809473, "grad_norm": 0.3154110610485077, "learning_rate": 4.0092722232510145e-08, "loss": 0.2088, "step": 13349 }, { "epoch": 3.552421500798297, "grad_norm": 0.2944083511829376, "learning_rate": 4.007906254072776e-08, "loss": 0.1899, "step": 13350 }, { "epoch": 3.552687599787121, "grad_norm": 0.2870200574398041, "learning_rate": 4.006540459308827e-08, "loss": 0.184, "step": 13351 }, { "epoch": 3.5529536987759447, "grad_norm": 0.30055859684944153, "learning_rate": 4.005174838998925e-08, "loss": 0.1769, "step": 13352 }, { "epoch": 3.5532197977647684, "grad_norm": 0.3112301230430603, "learning_rate": 4.003809393182816e-08, "loss": 0.1949, "step": 13353 }, { "epoch": 3.5534858967535925, "grad_norm": 0.3380551040172577, "learning_rate": 4.002444121900251e-08, "loss": 0.2021, "step": 13354 }, { "epoch": 3.553751995742416, "grad_norm": 0.3206472098827362, "learning_rate": 4.001079025190962e-08, "loss": 0.2172, "step": 13355 }, { "epoch": 3.55401809473124, "grad_norm": 0.31299924850463867, "learning_rate": 3.999714103094687e-08, "loss": 0.1788, "step": 13356 }, { "epoch": 3.554284193720064, "grad_norm": 0.32069072127342224, "learning_rate": 3.9983493556511575e-08, "loss": 0.1808, "step": 13357 }, { "epoch": 3.5545502927088877, "grad_norm": 0.36444342136383057, "learning_rate": 3.9969847829000914e-08, "loss": 0.1966, "step": 13358 }, { "epoch": 3.5548163916977114, "grad_norm": 0.26732802391052246, "learning_rate": 3.995620384881212e-08, "loss": 0.1697, "step": 13359 }, { "epoch": 3.5550824906865355, "grad_norm": 0.29392653703689575, "learning_rate": 3.994256161634233e-08, "loss": 0.1924, "step": 13360 }, { "epoch": 3.555348589675359, "grad_norm": 0.27860990166664124, "learning_rate": 3.9928921131988635e-08, "loss": 0.1837, "step": 13361 }, { "epoch": 3.555614688664183, "grad_norm": 0.296502947807312, "learning_rate": 3.991528239614807e-08, "loss": 0.1757, "step": 13362 }, { "epoch": 3.555880787653007, "grad_norm": 0.2877727746963501, "learning_rate": 3.990164540921763e-08, "loss": 0.1892, "step": 13363 }, { "epoch": 3.5561468866418307, "grad_norm": 0.3073606491088867, "learning_rate": 3.988801017159425e-08, "loss": 0.2054, "step": 13364 }, { "epoch": 3.5564129856306543, "grad_norm": 0.2843794822692871, "learning_rate": 3.9874376683674806e-08, "loss": 0.1545, "step": 13365 }, { "epoch": 3.5566790846194785, "grad_norm": 0.30769404768943787, "learning_rate": 3.986074494585618e-08, "loss": 0.1845, "step": 13366 }, { "epoch": 3.556945183608302, "grad_norm": 0.2657890319824219, "learning_rate": 3.984711495853509e-08, "loss": 0.1685, "step": 13367 }, { "epoch": 3.5572112825971263, "grad_norm": 0.27731505036354065, "learning_rate": 3.983348672210829e-08, "loss": 0.1811, "step": 13368 }, { "epoch": 3.55747738158595, "grad_norm": 0.30086374282836914, "learning_rate": 3.981986023697248e-08, "loss": 0.1977, "step": 13369 }, { "epoch": 3.557743480574774, "grad_norm": 0.3140784502029419, "learning_rate": 3.980623550352428e-08, "loss": 0.1921, "step": 13370 }, { "epoch": 3.5580095795635978, "grad_norm": 0.29893437027931213, "learning_rate": 3.979261252216026e-08, "loss": 0.2094, "step": 13371 }, { "epoch": 3.5582756785524214, "grad_norm": 0.31550315022468567, "learning_rate": 3.9778991293276974e-08, "loss": 0.1932, "step": 13372 }, { "epoch": 3.5585417775412456, "grad_norm": 0.2786811590194702, "learning_rate": 3.976537181727092e-08, "loss": 0.1708, "step": 13373 }, { "epoch": 3.5588078765300692, "grad_norm": 0.28612270951271057, "learning_rate": 3.975175409453847e-08, "loss": 0.1842, "step": 13374 }, { "epoch": 3.559073975518893, "grad_norm": 0.2893364131450653, "learning_rate": 3.973813812547602e-08, "loss": 0.2043, "step": 13375 }, { "epoch": 3.559340074507717, "grad_norm": 0.2773779332637787, "learning_rate": 3.972452391047992e-08, "loss": 0.1859, "step": 13376 }, { "epoch": 3.5596061734965407, "grad_norm": 0.30729618668556213, "learning_rate": 3.9710911449946415e-08, "loss": 0.201, "step": 13377 }, { "epoch": 3.5598722724853644, "grad_norm": 0.2978876531124115, "learning_rate": 3.969730074427172e-08, "loss": 0.1804, "step": 13378 }, { "epoch": 3.5601383714741885, "grad_norm": 0.30620983242988586, "learning_rate": 3.9683691793852034e-08, "loss": 0.1876, "step": 13379 }, { "epoch": 3.5604044704630122, "grad_norm": 0.28629374504089355, "learning_rate": 3.967008459908348e-08, "loss": 0.1736, "step": 13380 }, { "epoch": 3.560670569451836, "grad_norm": 0.2900935411453247, "learning_rate": 3.96564791603621e-08, "loss": 0.1854, "step": 13381 }, { "epoch": 3.56093666844066, "grad_norm": 0.2733478546142578, "learning_rate": 3.9642875478083934e-08, "loss": 0.1681, "step": 13382 }, { "epoch": 3.5612027674294837, "grad_norm": 0.27783456444740295, "learning_rate": 3.962927355264496e-08, "loss": 0.1784, "step": 13383 }, { "epoch": 3.5614688664183074, "grad_norm": 0.2927946150302887, "learning_rate": 3.9615673384441104e-08, "loss": 0.192, "step": 13384 }, { "epoch": 3.5617349654071315, "grad_norm": 0.47753962874412537, "learning_rate": 3.960207497386818e-08, "loss": 0.1904, "step": 13385 }, { "epoch": 3.562001064395955, "grad_norm": 0.28594499826431274, "learning_rate": 3.9588478321322025e-08, "loss": 0.1931, "step": 13386 }, { "epoch": 3.562267163384779, "grad_norm": 0.2872535288333893, "learning_rate": 3.957488342719843e-08, "loss": 0.1984, "step": 13387 }, { "epoch": 3.562533262373603, "grad_norm": 0.33185845613479614, "learning_rate": 3.9561290291893065e-08, "loss": 0.199, "step": 13388 }, { "epoch": 3.5627993613624267, "grad_norm": 0.34021613001823425, "learning_rate": 3.954769891580162e-08, "loss": 0.1997, "step": 13389 }, { "epoch": 3.563065460351251, "grad_norm": 0.33338403701782227, "learning_rate": 3.95341092993197e-08, "loss": 0.1964, "step": 13390 }, { "epoch": 3.5633315593400745, "grad_norm": 0.2897398769855499, "learning_rate": 3.9520521442842844e-08, "loss": 0.1853, "step": 13391 }, { "epoch": 3.563597658328898, "grad_norm": 0.2927491366863251, "learning_rate": 3.950693534676659e-08, "loss": 0.1949, "step": 13392 }, { "epoch": 3.5638637573177223, "grad_norm": 0.30236488580703735, "learning_rate": 3.9493351011486406e-08, "loss": 0.2105, "step": 13393 }, { "epoch": 3.564129856306546, "grad_norm": 0.3093230128288269, "learning_rate": 3.947976843739764e-08, "loss": 0.2095, "step": 13394 }, { "epoch": 3.56439595529537, "grad_norm": 0.2990760803222656, "learning_rate": 3.9466187624895686e-08, "loss": 0.2012, "step": 13395 }, { "epoch": 3.564662054284194, "grad_norm": 0.3229445815086365, "learning_rate": 3.945260857437586e-08, "loss": 0.1931, "step": 13396 }, { "epoch": 3.5649281532730175, "grad_norm": 0.2692749500274658, "learning_rate": 3.943903128623335e-08, "loss": 0.1913, "step": 13397 }, { "epoch": 3.5651942522618416, "grad_norm": 0.28554269671440125, "learning_rate": 3.94254557608634e-08, "loss": 0.1901, "step": 13398 }, { "epoch": 3.5654603512506653, "grad_norm": 0.3069021999835968, "learning_rate": 3.9411881998661156e-08, "loss": 0.1999, "step": 13399 }, { "epoch": 3.565726450239489, "grad_norm": 0.29860541224479675, "learning_rate": 3.9398310000021705e-08, "loss": 0.1879, "step": 13400 }, { "epoch": 3.565992549228313, "grad_norm": 0.2853890061378479, "learning_rate": 3.9384739765340105e-08, "loss": 0.1979, "step": 13401 }, { "epoch": 3.566258648217137, "grad_norm": 0.3137316107749939, "learning_rate": 3.937117129501137e-08, "loss": 0.1917, "step": 13402 }, { "epoch": 3.5665247472059605, "grad_norm": 0.29586151242256165, "learning_rate": 3.9357604589430396e-08, "loss": 0.1817, "step": 13403 }, { "epoch": 3.5667908461947846, "grad_norm": 0.2968774437904358, "learning_rate": 3.934403964899208e-08, "loss": 0.1934, "step": 13404 }, { "epoch": 3.5670569451836083, "grad_norm": 0.29386359453201294, "learning_rate": 3.933047647409129e-08, "loss": 0.1925, "step": 13405 }, { "epoch": 3.567323044172432, "grad_norm": 0.28989067673683167, "learning_rate": 3.931691506512279e-08, "loss": 0.1832, "step": 13406 }, { "epoch": 3.567589143161256, "grad_norm": 0.31296664476394653, "learning_rate": 3.930335542248134e-08, "loss": 0.1995, "step": 13407 }, { "epoch": 3.5678552421500798, "grad_norm": 0.2812579870223999, "learning_rate": 3.928979754656162e-08, "loss": 0.1785, "step": 13408 }, { "epoch": 3.5681213411389034, "grad_norm": 0.32922831177711487, "learning_rate": 3.927624143775826e-08, "loss": 0.212, "step": 13409 }, { "epoch": 3.5683874401277276, "grad_norm": 0.28536054491996765, "learning_rate": 3.926268709646584e-08, "loss": 0.1866, "step": 13410 }, { "epoch": 3.5686535391165513, "grad_norm": 0.28100457787513733, "learning_rate": 3.924913452307893e-08, "loss": 0.1965, "step": 13411 }, { "epoch": 3.568919638105375, "grad_norm": 0.3166423439979553, "learning_rate": 3.9235583717991935e-08, "loss": 0.2036, "step": 13412 }, { "epoch": 3.569185737094199, "grad_norm": 0.30956679582595825, "learning_rate": 3.9222034681599336e-08, "loss": 0.2028, "step": 13413 }, { "epoch": 3.5694518360830227, "grad_norm": 0.2803342044353485, "learning_rate": 3.92084874142955e-08, "loss": 0.2016, "step": 13414 }, { "epoch": 3.569717935071847, "grad_norm": 0.30591312050819397, "learning_rate": 3.9194941916474745e-08, "loss": 0.2068, "step": 13415 }, { "epoch": 3.5699840340606706, "grad_norm": 0.2885376214981079, "learning_rate": 3.918139818853139e-08, "loss": 0.196, "step": 13416 }, { "epoch": 3.5702501330494947, "grad_norm": 0.2831115424633026, "learning_rate": 3.916785623085959e-08, "loss": 0.1683, "step": 13417 }, { "epoch": 3.5705162320383184, "grad_norm": 0.2784139811992645, "learning_rate": 3.915431604385354e-08, "loss": 0.18, "step": 13418 }, { "epoch": 3.570782331027142, "grad_norm": 0.2825392186641693, "learning_rate": 3.914077762790736e-08, "loss": 0.1876, "step": 13419 }, { "epoch": 3.571048430015966, "grad_norm": 0.29958394169807434, "learning_rate": 3.9127240983415165e-08, "loss": 0.1874, "step": 13420 }, { "epoch": 3.57131452900479, "grad_norm": 0.318238765001297, "learning_rate": 3.9113706110770893e-08, "loss": 0.1963, "step": 13421 }, { "epoch": 3.5715806279936135, "grad_norm": 0.3045644164085388, "learning_rate": 3.9100173010368555e-08, "loss": 0.1776, "step": 13422 }, { "epoch": 3.5718467269824377, "grad_norm": 0.2836754024028778, "learning_rate": 3.9086641682602054e-08, "loss": 0.1951, "step": 13423 }, { "epoch": 3.5721128259712613, "grad_norm": 0.3037772476673126, "learning_rate": 3.9073112127865246e-08, "loss": 0.1881, "step": 13424 }, { "epoch": 3.572378924960085, "grad_norm": 0.33155494928359985, "learning_rate": 3.9059584346551957e-08, "loss": 0.1898, "step": 13425 }, { "epoch": 3.572645023948909, "grad_norm": 0.29342272877693176, "learning_rate": 3.904605833905593e-08, "loss": 0.1824, "step": 13426 }, { "epoch": 3.572911122937733, "grad_norm": 0.31526055932044983, "learning_rate": 3.903253410577088e-08, "loss": 0.2069, "step": 13427 }, { "epoch": 3.5731772219265565, "grad_norm": 0.29429373145103455, "learning_rate": 3.901901164709046e-08, "loss": 0.1852, "step": 13428 }, { "epoch": 3.5734433209153806, "grad_norm": 0.2923470139503479, "learning_rate": 3.9005490963408314e-08, "loss": 0.1972, "step": 13429 }, { "epoch": 3.5737094199042043, "grad_norm": 0.28357994556427, "learning_rate": 3.899197205511793e-08, "loss": 0.1778, "step": 13430 }, { "epoch": 3.573975518893028, "grad_norm": 0.3033868968486786, "learning_rate": 3.8978454922612816e-08, "loss": 0.1943, "step": 13431 }, { "epoch": 3.574241617881852, "grad_norm": 0.29507705569267273, "learning_rate": 3.8964939566286447e-08, "loss": 0.1875, "step": 13432 }, { "epoch": 3.574507716870676, "grad_norm": 0.3361106812953949, "learning_rate": 3.8951425986532215e-08, "loss": 0.216, "step": 13433 }, { "epoch": 3.5747738158594995, "grad_norm": 0.300690233707428, "learning_rate": 3.8937914183743456e-08, "loss": 0.1846, "step": 13434 }, { "epoch": 3.5750399148483236, "grad_norm": 0.3071485161781311, "learning_rate": 3.89244041583135e-08, "loss": 0.2017, "step": 13435 }, { "epoch": 3.5753060138371473, "grad_norm": 0.29271018505096436, "learning_rate": 3.891089591063553e-08, "loss": 0.1909, "step": 13436 }, { "epoch": 3.5755721128259714, "grad_norm": 0.3233294188976288, "learning_rate": 3.8897389441102766e-08, "loss": 0.2206, "step": 13437 }, { "epoch": 3.575838211814795, "grad_norm": 0.29661139845848083, "learning_rate": 3.888388475010838e-08, "loss": 0.1706, "step": 13438 }, { "epoch": 3.576104310803619, "grad_norm": 0.32498565316200256, "learning_rate": 3.8870381838045374e-08, "loss": 0.1988, "step": 13439 }, { "epoch": 3.576370409792443, "grad_norm": 0.2884066104888916, "learning_rate": 3.8856880705306835e-08, "loss": 0.175, "step": 13440 }, { "epoch": 3.5766365087812666, "grad_norm": 0.3064008355140686, "learning_rate": 3.884338135228575e-08, "loss": 0.1856, "step": 13441 }, { "epoch": 3.5769026077700907, "grad_norm": 0.2942706346511841, "learning_rate": 3.882988377937505e-08, "loss": 0.1737, "step": 13442 }, { "epoch": 3.5771687067589144, "grad_norm": 0.3228016197681427, "learning_rate": 3.881638798696759e-08, "loss": 0.1891, "step": 13443 }, { "epoch": 3.577434805747738, "grad_norm": 0.31876319646835327, "learning_rate": 3.880289397545622e-08, "loss": 0.2061, "step": 13444 }, { "epoch": 3.577700904736562, "grad_norm": 0.3002914786338806, "learning_rate": 3.8789401745233705e-08, "loss": 0.1724, "step": 13445 }, { "epoch": 3.577967003725386, "grad_norm": 0.30432194471359253, "learning_rate": 3.877591129669279e-08, "loss": 0.1915, "step": 13446 }, { "epoch": 3.5782331027142096, "grad_norm": 0.30402520298957825, "learning_rate": 3.8762422630226145e-08, "loss": 0.2066, "step": 13447 }, { "epoch": 3.5784992017030337, "grad_norm": 0.2714035212993622, "learning_rate": 3.874893574622635e-08, "loss": 0.1845, "step": 13448 }, { "epoch": 3.5787653006918574, "grad_norm": 0.2879306375980377, "learning_rate": 3.8735450645086e-08, "loss": 0.1848, "step": 13449 }, { "epoch": 3.579031399680681, "grad_norm": 0.29589518904685974, "learning_rate": 3.8721967327197605e-08, "loss": 0.1931, "step": 13450 }, { "epoch": 3.579297498669505, "grad_norm": 0.28460127115249634, "learning_rate": 3.8708485792953636e-08, "loss": 0.1879, "step": 13451 }, { "epoch": 3.579563597658329, "grad_norm": 0.2856123745441437, "learning_rate": 3.86950060427465e-08, "loss": 0.1978, "step": 13452 }, { "epoch": 3.5798296966471526, "grad_norm": 0.31194007396698, "learning_rate": 3.868152807696857e-08, "loss": 0.1849, "step": 13453 }, { "epoch": 3.5800957956359767, "grad_norm": 0.3067214787006378, "learning_rate": 3.866805189601214e-08, "loss": 0.1994, "step": 13454 }, { "epoch": 3.5803618946248004, "grad_norm": 0.2846625745296478, "learning_rate": 3.8654577500269516e-08, "loss": 0.1764, "step": 13455 }, { "epoch": 3.580627993613624, "grad_norm": 0.2961916923522949, "learning_rate": 3.864110489013281e-08, "loss": 0.1835, "step": 13456 }, { "epoch": 3.580894092602448, "grad_norm": 0.31687843799591064, "learning_rate": 3.862763406599424e-08, "loss": 0.2079, "step": 13457 }, { "epoch": 3.581160191591272, "grad_norm": 0.3131938576698303, "learning_rate": 3.8614165028245916e-08, "loss": 0.1966, "step": 13458 }, { "epoch": 3.5814262905800955, "grad_norm": 0.2871093451976776, "learning_rate": 3.8600697777279826e-08, "loss": 0.1891, "step": 13459 }, { "epoch": 3.5816923895689197, "grad_norm": 0.28060758113861084, "learning_rate": 3.858723231348801e-08, "loss": 0.1762, "step": 13460 }, { "epoch": 3.5819584885577433, "grad_norm": 0.28141483664512634, "learning_rate": 3.8573768637262396e-08, "loss": 0.1721, "step": 13461 }, { "epoch": 3.5822245875465675, "grad_norm": 0.30587396025657654, "learning_rate": 3.8560306748994886e-08, "loss": 0.1882, "step": 13462 }, { "epoch": 3.582490686535391, "grad_norm": 0.28747597336769104, "learning_rate": 3.8546846649077316e-08, "loss": 0.1898, "step": 13463 }, { "epoch": 3.5827567855242153, "grad_norm": 0.33628731966018677, "learning_rate": 3.853338833790148e-08, "loss": 0.2092, "step": 13464 }, { "epoch": 3.583022884513039, "grad_norm": 0.2991437017917633, "learning_rate": 3.8519931815859155e-08, "loss": 0.2084, "step": 13465 }, { "epoch": 3.5832889835018626, "grad_norm": 0.28784120082855225, "learning_rate": 3.850647708334195e-08, "loss": 0.1779, "step": 13466 }, { "epoch": 3.5835550824906868, "grad_norm": 0.31560367345809937, "learning_rate": 3.8493024140741526e-08, "loss": 0.2117, "step": 13467 }, { "epoch": 3.5838211814795105, "grad_norm": 0.2806808352470398, "learning_rate": 3.847957298844946e-08, "loss": 0.1791, "step": 13468 }, { "epoch": 3.584087280468334, "grad_norm": 0.2976495921611786, "learning_rate": 3.8466123626857305e-08, "loss": 0.1862, "step": 13469 }, { "epoch": 3.5843533794571583, "grad_norm": 0.279233455657959, "learning_rate": 3.8452676056356506e-08, "loss": 0.1841, "step": 13470 }, { "epoch": 3.584619478445982, "grad_norm": 0.3030390441417694, "learning_rate": 3.843923027733851e-08, "loss": 0.1981, "step": 13471 }, { "epoch": 3.5848855774348056, "grad_norm": 0.29184091091156006, "learning_rate": 3.842578629019467e-08, "loss": 0.195, "step": 13472 }, { "epoch": 3.5851516764236298, "grad_norm": 0.3105842173099518, "learning_rate": 3.841234409531633e-08, "loss": 0.1864, "step": 13473 }, { "epoch": 3.5854177754124534, "grad_norm": 0.3028331398963928, "learning_rate": 3.8398903693094765e-08, "loss": 0.1906, "step": 13474 }, { "epoch": 3.585683874401277, "grad_norm": 0.29877975583076477, "learning_rate": 3.8385465083921135e-08, "loss": 0.198, "step": 13475 }, { "epoch": 3.5859499733901012, "grad_norm": 0.27568233013153076, "learning_rate": 3.837202826818664e-08, "loss": 0.1749, "step": 13476 }, { "epoch": 3.586216072378925, "grad_norm": 0.31177669763565063, "learning_rate": 3.8358593246282425e-08, "loss": 0.2096, "step": 13477 }, { "epoch": 3.5864821713677486, "grad_norm": 0.29267263412475586, "learning_rate": 3.834516001859946e-08, "loss": 0.193, "step": 13478 }, { "epoch": 3.5867482703565727, "grad_norm": 0.3087003827095032, "learning_rate": 3.833172858552881e-08, "loss": 0.1799, "step": 13479 }, { "epoch": 3.5870143693453964, "grad_norm": 0.28213363885879517, "learning_rate": 3.8318298947461424e-08, "loss": 0.1813, "step": 13480 }, { "epoch": 3.58728046833422, "grad_norm": 0.30821123719215393, "learning_rate": 3.8304871104788206e-08, "loss": 0.184, "step": 13481 }, { "epoch": 3.5875465673230442, "grad_norm": 0.29842984676361084, "learning_rate": 3.8291445057899984e-08, "loss": 0.1932, "step": 13482 }, { "epoch": 3.587812666311868, "grad_norm": 0.2801741361618042, "learning_rate": 3.827802080718762e-08, "loss": 0.1888, "step": 13483 }, { "epoch": 3.5880787653006916, "grad_norm": 0.2806750535964966, "learning_rate": 3.826459835304178e-08, "loss": 0.1649, "step": 13484 }, { "epoch": 3.5883448642895157, "grad_norm": 0.29895320534706116, "learning_rate": 3.825117769585317e-08, "loss": 0.1904, "step": 13485 }, { "epoch": 3.5886109632783394, "grad_norm": 0.3044433295726776, "learning_rate": 3.8237758836012457e-08, "loss": 0.2083, "step": 13486 }, { "epoch": 3.5888770622671635, "grad_norm": 0.2976921796798706, "learning_rate": 3.822434177391022e-08, "loss": 0.1948, "step": 13487 }, { "epoch": 3.589143161255987, "grad_norm": 0.29517948627471924, "learning_rate": 3.8210926509937e-08, "loss": 0.197, "step": 13488 }, { "epoch": 3.5894092602448113, "grad_norm": 0.2774837613105774, "learning_rate": 3.819751304448327e-08, "loss": 0.1829, "step": 13489 }, { "epoch": 3.589675359233635, "grad_norm": 0.3019517660140991, "learning_rate": 3.818410137793947e-08, "loss": 0.1934, "step": 13490 }, { "epoch": 3.5899414582224587, "grad_norm": 0.29655298590660095, "learning_rate": 3.817069151069597e-08, "loss": 0.1985, "step": 13491 }, { "epoch": 3.590207557211283, "grad_norm": 0.2882240116596222, "learning_rate": 3.815728344314314e-08, "loss": 0.1688, "step": 13492 }, { "epoch": 3.5904736562001065, "grad_norm": 0.2997923791408539, "learning_rate": 3.814387717567118e-08, "loss": 0.1993, "step": 13493 }, { "epoch": 3.59073975518893, "grad_norm": 0.31505435705184937, "learning_rate": 3.813047270867035e-08, "loss": 0.1998, "step": 13494 }, { "epoch": 3.5910058541777543, "grad_norm": 0.30144059658050537, "learning_rate": 3.811707004253082e-08, "loss": 0.1975, "step": 13495 }, { "epoch": 3.591271953166578, "grad_norm": 0.28729936480522156, "learning_rate": 3.810366917764269e-08, "loss": 0.1907, "step": 13496 }, { "epoch": 3.5915380521554017, "grad_norm": 0.3086014688014984, "learning_rate": 3.809027011439609e-08, "loss": 0.1945, "step": 13497 }, { "epoch": 3.591804151144226, "grad_norm": 0.3198757469654083, "learning_rate": 3.807687285318094e-08, "loss": 0.1984, "step": 13498 }, { "epoch": 3.5920702501330495, "grad_norm": 0.4513690173625946, "learning_rate": 3.8063477394387234e-08, "loss": 0.191, "step": 13499 }, { "epoch": 3.592336349121873, "grad_norm": 0.3023204803466797, "learning_rate": 3.805008373840489e-08, "loss": 0.1932, "step": 13500 }, { "epoch": 3.5926024481106973, "grad_norm": 0.3191502094268799, "learning_rate": 3.8036691885623784e-08, "loss": 0.2106, "step": 13501 }, { "epoch": 3.592868547099521, "grad_norm": 0.30616191029548645, "learning_rate": 3.802330183643366e-08, "loss": 0.1932, "step": 13502 }, { "epoch": 3.5931346460883447, "grad_norm": 0.3254445195198059, "learning_rate": 3.800991359122431e-08, "loss": 0.1969, "step": 13503 }, { "epoch": 3.593400745077169, "grad_norm": 0.2896772623062134, "learning_rate": 3.7996527150385403e-08, "loss": 0.1933, "step": 13504 }, { "epoch": 3.5936668440659925, "grad_norm": 0.30078577995300293, "learning_rate": 3.798314251430661e-08, "loss": 0.1909, "step": 13505 }, { "epoch": 3.593932943054816, "grad_norm": 0.27902016043663025, "learning_rate": 3.7969759683377514e-08, "loss": 0.1906, "step": 13506 }, { "epoch": 3.5941990420436403, "grad_norm": 0.30781289935112, "learning_rate": 3.795637865798764e-08, "loss": 0.1923, "step": 13507 }, { "epoch": 3.594465141032464, "grad_norm": 0.28906574845314026, "learning_rate": 3.79429994385265e-08, "loss": 0.2155, "step": 13508 }, { "epoch": 3.594731240021288, "grad_norm": 0.3090357482433319, "learning_rate": 3.792962202538352e-08, "loss": 0.1775, "step": 13509 }, { "epoch": 3.5949973390101118, "grad_norm": 0.27973946928977966, "learning_rate": 3.7916246418948106e-08, "loss": 0.1794, "step": 13510 }, { "epoch": 3.5952634379989354, "grad_norm": 0.31585004925727844, "learning_rate": 3.790287261960953e-08, "loss": 0.185, "step": 13511 }, { "epoch": 3.5955295369877596, "grad_norm": 0.26762890815734863, "learning_rate": 3.78895006277571e-08, "loss": 0.1797, "step": 13512 }, { "epoch": 3.5957956359765832, "grad_norm": 0.28798046708106995, "learning_rate": 3.787613044378003e-08, "loss": 0.1991, "step": 13513 }, { "epoch": 3.5960617349654074, "grad_norm": 0.29709500074386597, "learning_rate": 3.786276206806751e-08, "loss": 0.1893, "step": 13514 }, { "epoch": 3.596327833954231, "grad_norm": 0.3069719970226288, "learning_rate": 3.784939550100864e-08, "loss": 0.2017, "step": 13515 }, { "epoch": 3.5965939329430547, "grad_norm": 0.30023542046546936, "learning_rate": 3.78360307429925e-08, "loss": 0.2043, "step": 13516 }, { "epoch": 3.596860031931879, "grad_norm": 0.29272007942199707, "learning_rate": 3.782266779440814e-08, "loss": 0.1992, "step": 13517 }, { "epoch": 3.5971261309207025, "grad_norm": 0.27855730056762695, "learning_rate": 3.7809306655644426e-08, "loss": 0.1822, "step": 13518 }, { "epoch": 3.5973922299095262, "grad_norm": 0.29686397314071655, "learning_rate": 3.779594732709036e-08, "loss": 0.208, "step": 13519 }, { "epoch": 3.5976583288983504, "grad_norm": 0.3003292679786682, "learning_rate": 3.778258980913472e-08, "loss": 0.1901, "step": 13520 }, { "epoch": 3.597924427887174, "grad_norm": 0.2859437167644501, "learning_rate": 3.7769234102166355e-08, "loss": 0.1928, "step": 13521 }, { "epoch": 3.5981905268759977, "grad_norm": 0.2950004041194916, "learning_rate": 3.7755880206574e-08, "loss": 0.1866, "step": 13522 }, { "epoch": 3.598456625864822, "grad_norm": 0.27674204111099243, "learning_rate": 3.7742528122746354e-08, "loss": 0.1963, "step": 13523 }, { "epoch": 3.5987227248536455, "grad_norm": 0.30391353368759155, "learning_rate": 3.772917785107207e-08, "loss": 0.1581, "step": 13524 }, { "epoch": 3.598988823842469, "grad_norm": 0.2928353548049927, "learning_rate": 3.771582939193973e-08, "loss": 0.2108, "step": 13525 }, { "epoch": 3.5992549228312933, "grad_norm": 0.2975046932697296, "learning_rate": 3.7702482745737876e-08, "loss": 0.1891, "step": 13526 }, { "epoch": 3.599521021820117, "grad_norm": 0.25947487354278564, "learning_rate": 3.7689137912855e-08, "loss": 0.1774, "step": 13527 }, { "epoch": 3.5997871208089407, "grad_norm": 0.2860388457775116, "learning_rate": 3.767579489367956e-08, "loss": 0.1663, "step": 13528 }, { "epoch": 3.600053219797765, "grad_norm": 0.30377304553985596, "learning_rate": 3.7662453688599874e-08, "loss": 0.1971, "step": 13529 }, { "epoch": 3.6003193187865885, "grad_norm": 0.27415764331817627, "learning_rate": 3.76491142980043e-08, "loss": 0.1759, "step": 13530 }, { "epoch": 3.600585417775412, "grad_norm": 0.28338882327079773, "learning_rate": 3.763577672228112e-08, "loss": 0.184, "step": 13531 }, { "epoch": 3.6008515167642363, "grad_norm": 0.3101390600204468, "learning_rate": 3.7622440961818557e-08, "loss": 0.202, "step": 13532 }, { "epoch": 3.60111761575306, "grad_norm": 0.2909187376499176, "learning_rate": 3.760910701700477e-08, "loss": 0.1569, "step": 13533 }, { "epoch": 3.601383714741884, "grad_norm": 0.28101927042007446, "learning_rate": 3.759577488822788e-08, "loss": 0.1813, "step": 13534 }, { "epoch": 3.601649813730708, "grad_norm": 0.38756829500198364, "learning_rate": 3.758244457587596e-08, "loss": 0.2114, "step": 13535 }, { "epoch": 3.601915912719532, "grad_norm": 0.30227890610694885, "learning_rate": 3.756911608033704e-08, "loss": 0.1822, "step": 13536 }, { "epoch": 3.6021820117083556, "grad_norm": 0.31430432200431824, "learning_rate": 3.7555789401999016e-08, "loss": 0.1986, "step": 13537 }, { "epoch": 3.6024481106971793, "grad_norm": 0.28939929604530334, "learning_rate": 3.7542464541249827e-08, "loss": 0.1848, "step": 13538 }, { "epoch": 3.6027142096860034, "grad_norm": 0.31876036524772644, "learning_rate": 3.752914149847737e-08, "loss": 0.1804, "step": 13539 }, { "epoch": 3.602980308674827, "grad_norm": 0.3085983693599701, "learning_rate": 3.7515820274069354e-08, "loss": 0.2085, "step": 13540 }, { "epoch": 3.603246407663651, "grad_norm": 0.3019256889820099, "learning_rate": 3.7502500868413575e-08, "loss": 0.1907, "step": 13541 }, { "epoch": 3.603512506652475, "grad_norm": 0.30415627360343933, "learning_rate": 3.748918328189773e-08, "loss": 0.1845, "step": 13542 }, { "epoch": 3.6037786056412986, "grad_norm": 0.32124316692352295, "learning_rate": 3.747586751490944e-08, "loss": 0.2086, "step": 13543 }, { "epoch": 3.6040447046301223, "grad_norm": 0.31817618012428284, "learning_rate": 3.746255356783632e-08, "loss": 0.2061, "step": 13544 }, { "epoch": 3.6043108036189464, "grad_norm": 0.33409634232521057, "learning_rate": 3.744924144106588e-08, "loss": 0.1977, "step": 13545 }, { "epoch": 3.60457690260777, "grad_norm": 0.32814425230026245, "learning_rate": 3.7435931134985645e-08, "loss": 0.1956, "step": 13546 }, { "epoch": 3.6048430015965938, "grad_norm": 0.2865172326564789, "learning_rate": 3.742262264998297e-08, "loss": 0.1858, "step": 13547 }, { "epoch": 3.605109100585418, "grad_norm": 0.297323614358902, "learning_rate": 3.7409315986445265e-08, "loss": 0.1888, "step": 13548 }, { "epoch": 3.6053751995742416, "grad_norm": 0.274800181388855, "learning_rate": 3.7396011144759876e-08, "loss": 0.1667, "step": 13549 }, { "epoch": 3.6056412985630653, "grad_norm": 0.29203325510025024, "learning_rate": 3.7382708125314035e-08, "loss": 0.1997, "step": 13550 }, { "epoch": 3.6059073975518894, "grad_norm": 0.2915366590023041, "learning_rate": 3.7369406928494983e-08, "loss": 0.1833, "step": 13551 }, { "epoch": 3.606173496540713, "grad_norm": 0.3051019012928009, "learning_rate": 3.7356107554689873e-08, "loss": 0.1996, "step": 13552 }, { "epoch": 3.6064395955295367, "grad_norm": 0.29271313548088074, "learning_rate": 3.734281000428583e-08, "loss": 0.1891, "step": 13553 }, { "epoch": 3.606705694518361, "grad_norm": 0.31652647256851196, "learning_rate": 3.732951427766988e-08, "loss": 0.1967, "step": 13554 }, { "epoch": 3.6069717935071846, "grad_norm": 0.3015693128108978, "learning_rate": 3.731622037522907e-08, "loss": 0.2094, "step": 13555 }, { "epoch": 3.6072378924960087, "grad_norm": 0.31695789098739624, "learning_rate": 3.7302928297350344e-08, "loss": 0.2006, "step": 13556 }, { "epoch": 3.6075039914848324, "grad_norm": 0.2837160527706146, "learning_rate": 3.7289638044420554e-08, "loss": 0.1847, "step": 13557 }, { "epoch": 3.607770090473656, "grad_norm": 0.30310964584350586, "learning_rate": 3.727634961682659e-08, "loss": 0.164, "step": 13558 }, { "epoch": 3.60803618946248, "grad_norm": 0.3010418713092804, "learning_rate": 3.7263063014955245e-08, "loss": 0.1886, "step": 13559 }, { "epoch": 3.608302288451304, "grad_norm": 0.27994921803474426, "learning_rate": 3.7249778239193205e-08, "loss": 0.1853, "step": 13560 }, { "epoch": 3.608568387440128, "grad_norm": 0.3361634612083435, "learning_rate": 3.723649528992721e-08, "loss": 0.1992, "step": 13561 }, { "epoch": 3.6088344864289517, "grad_norm": 0.2888111174106598, "learning_rate": 3.722321416754386e-08, "loss": 0.2003, "step": 13562 }, { "epoch": 3.6091005854177753, "grad_norm": 0.3147554397583008, "learning_rate": 3.720993487242975e-08, "loss": 0.2171, "step": 13563 }, { "epoch": 3.6093666844065995, "grad_norm": 0.3162538707256317, "learning_rate": 3.71966574049714e-08, "loss": 0.1881, "step": 13564 }, { "epoch": 3.609632783395423, "grad_norm": 0.3082617223262787, "learning_rate": 3.718338176555532e-08, "loss": 0.1845, "step": 13565 }, { "epoch": 3.609898882384247, "grad_norm": 0.31284743547439575, "learning_rate": 3.717010795456786e-08, "loss": 0.1786, "step": 13566 }, { "epoch": 3.610164981373071, "grad_norm": 0.2988581359386444, "learning_rate": 3.715683597239543e-08, "loss": 0.175, "step": 13567 }, { "epoch": 3.6104310803618946, "grad_norm": 0.34918445348739624, "learning_rate": 3.7143565819424314e-08, "loss": 0.2264, "step": 13568 }, { "epoch": 3.6106971793507183, "grad_norm": 0.29344502091407776, "learning_rate": 3.7130297496040805e-08, "loss": 0.2049, "step": 13569 }, { "epoch": 3.6109632783395424, "grad_norm": 0.31282541155815125, "learning_rate": 3.7117031002631103e-08, "loss": 0.2053, "step": 13570 }, { "epoch": 3.611229377328366, "grad_norm": 0.3050135374069214, "learning_rate": 3.7103766339581356e-08, "loss": 0.215, "step": 13571 }, { "epoch": 3.61149547631719, "grad_norm": 0.3084397614002228, "learning_rate": 3.709050350727765e-08, "loss": 0.2044, "step": 13572 }, { "epoch": 3.611761575306014, "grad_norm": 0.3274678587913513, "learning_rate": 3.7077242506106057e-08, "loss": 0.1838, "step": 13573 }, { "epoch": 3.6120276742948376, "grad_norm": 0.3317786455154419, "learning_rate": 3.7063983336452576e-08, "loss": 0.1746, "step": 13574 }, { "epoch": 3.6122937732836613, "grad_norm": 0.2868100106716156, "learning_rate": 3.7050725998703104e-08, "loss": 0.1867, "step": 13575 }, { "epoch": 3.6125598722724854, "grad_norm": 0.3179353177547455, "learning_rate": 3.703747049324355e-08, "loss": 0.1962, "step": 13576 }, { "epoch": 3.612825971261309, "grad_norm": 0.3247286081314087, "learning_rate": 3.7024216820459754e-08, "loss": 0.1899, "step": 13577 }, { "epoch": 3.613092070250133, "grad_norm": 0.29733577370643616, "learning_rate": 3.7010964980737516e-08, "loss": 0.1773, "step": 13578 }, { "epoch": 3.613358169238957, "grad_norm": 0.3037869930267334, "learning_rate": 3.6997714974462516e-08, "loss": 0.1636, "step": 13579 }, { "epoch": 3.6136242682277806, "grad_norm": 0.2899090349674225, "learning_rate": 3.698446680202043e-08, "loss": 0.1895, "step": 13580 }, { "epoch": 3.6138903672166047, "grad_norm": 0.2947683334350586, "learning_rate": 3.697122046379691e-08, "loss": 0.1871, "step": 13581 }, { "epoch": 3.6141564662054284, "grad_norm": 0.3213302791118622, "learning_rate": 3.695797596017751e-08, "loss": 0.1956, "step": 13582 }, { "epoch": 3.6144225651942525, "grad_norm": 0.3298495411872864, "learning_rate": 3.694473329154778e-08, "loss": 0.1812, "step": 13583 }, { "epoch": 3.614688664183076, "grad_norm": 0.3024107813835144, "learning_rate": 3.69314924582931e-08, "loss": 0.1948, "step": 13584 }, { "epoch": 3.6149547631719, "grad_norm": 0.31441831588745117, "learning_rate": 3.691825346079892e-08, "loss": 0.1857, "step": 13585 }, { "epoch": 3.615220862160724, "grad_norm": 0.29679805040359497, "learning_rate": 3.6905016299450595e-08, "loss": 0.1878, "step": 13586 }, { "epoch": 3.6154869611495477, "grad_norm": 0.30424463748931885, "learning_rate": 3.6891780974633425e-08, "loss": 0.1904, "step": 13587 }, { "epoch": 3.6157530601383714, "grad_norm": 0.30651164054870605, "learning_rate": 3.687854748673265e-08, "loss": 0.2096, "step": 13588 }, { "epoch": 3.6160191591271955, "grad_norm": 0.31221166253089905, "learning_rate": 3.686531583613346e-08, "loss": 0.187, "step": 13589 }, { "epoch": 3.616285258116019, "grad_norm": 0.2973119020462036, "learning_rate": 3.6852086023221e-08, "loss": 0.1812, "step": 13590 }, { "epoch": 3.616551357104843, "grad_norm": 0.3048747181892395, "learning_rate": 3.683885804838036e-08, "loss": 0.1971, "step": 13591 }, { "epoch": 3.616817456093667, "grad_norm": 0.2928050458431244, "learning_rate": 3.682563191199659e-08, "loss": 0.1959, "step": 13592 }, { "epoch": 3.6170835550824907, "grad_norm": 0.29363641142845154, "learning_rate": 3.681240761445462e-08, "loss": 0.1917, "step": 13593 }, { "epoch": 3.6173496540713144, "grad_norm": 0.3011534810066223, "learning_rate": 3.679918515613939e-08, "loss": 0.2066, "step": 13594 }, { "epoch": 3.6176157530601385, "grad_norm": 0.2792910933494568, "learning_rate": 3.6785964537435785e-08, "loss": 0.1691, "step": 13595 }, { "epoch": 3.617881852048962, "grad_norm": 0.3108038008213043, "learning_rate": 3.677274575872862e-08, "loss": 0.1938, "step": 13596 }, { "epoch": 3.618147951037786, "grad_norm": 0.2981507480144501, "learning_rate": 3.675952882040265e-08, "loss": 0.202, "step": 13597 }, { "epoch": 3.61841405002661, "grad_norm": 0.28791823983192444, "learning_rate": 3.6746313722842645e-08, "loss": 0.1814, "step": 13598 }, { "epoch": 3.6186801490154337, "grad_norm": 0.3127327561378479, "learning_rate": 3.6733100466433166e-08, "loss": 0.1904, "step": 13599 }, { "epoch": 3.6189462480042573, "grad_norm": 0.31646281480789185, "learning_rate": 3.6719889051558864e-08, "loss": 0.1986, "step": 13600 }, { "epoch": 3.6192123469930815, "grad_norm": 0.2906528115272522, "learning_rate": 3.670667947860432e-08, "loss": 0.1992, "step": 13601 }, { "epoch": 3.619478445981905, "grad_norm": 0.2961946427822113, "learning_rate": 3.6693471747953966e-08, "loss": 0.19, "step": 13602 }, { "epoch": 3.619744544970729, "grad_norm": 0.2743295133113861, "learning_rate": 3.6680265859992263e-08, "loss": 0.173, "step": 13603 }, { "epoch": 3.620010643959553, "grad_norm": 0.29502537846565247, "learning_rate": 3.666706181510362e-08, "loss": 0.1927, "step": 13604 }, { "epoch": 3.6202767429483766, "grad_norm": 0.351716011762619, "learning_rate": 3.665385961367237e-08, "loss": 0.1952, "step": 13605 }, { "epoch": 3.6205428419372008, "grad_norm": 0.2869994044303894, "learning_rate": 3.664065925608279e-08, "loss": 0.1875, "step": 13606 }, { "epoch": 3.6208089409260245, "grad_norm": 0.3174950182437897, "learning_rate": 3.6627460742719095e-08, "loss": 0.195, "step": 13607 }, { "epoch": 3.6210750399148486, "grad_norm": 0.2994212210178375, "learning_rate": 3.661426407396548e-08, "loss": 0.1961, "step": 13608 }, { "epoch": 3.6213411389036723, "grad_norm": 0.2767588794231415, "learning_rate": 3.660106925020607e-08, "loss": 0.1759, "step": 13609 }, { "epoch": 3.621607237892496, "grad_norm": 0.3005558252334595, "learning_rate": 3.6587876271824946e-08, "loss": 0.185, "step": 13610 }, { "epoch": 3.62187333688132, "grad_norm": 0.644868016242981, "learning_rate": 3.6574685139206064e-08, "loss": 0.1936, "step": 13611 }, { "epoch": 3.6221394358701438, "grad_norm": 0.2946382164955139, "learning_rate": 3.65614958527334e-08, "loss": 0.188, "step": 13612 }, { "epoch": 3.6224055348589674, "grad_norm": 0.2960030138492584, "learning_rate": 3.6548308412790895e-08, "loss": 0.1809, "step": 13613 }, { "epoch": 3.6226716338477916, "grad_norm": 0.3138570785522461, "learning_rate": 3.653512281976238e-08, "loss": 0.1931, "step": 13614 }, { "epoch": 3.6229377328366152, "grad_norm": 0.30156078934669495, "learning_rate": 3.652193907403165e-08, "loss": 0.2023, "step": 13615 }, { "epoch": 3.623203831825439, "grad_norm": 0.30665379762649536, "learning_rate": 3.6508757175982455e-08, "loss": 0.1931, "step": 13616 }, { "epoch": 3.623469930814263, "grad_norm": 0.3221113383769989, "learning_rate": 3.649557712599851e-08, "loss": 0.211, "step": 13617 }, { "epoch": 3.6237360298030867, "grad_norm": 0.27966606616973877, "learning_rate": 3.6482398924463395e-08, "loss": 0.1794, "step": 13618 }, { "epoch": 3.6240021287919104, "grad_norm": 0.27695831656455994, "learning_rate": 3.646922257176074e-08, "loss": 0.1981, "step": 13619 }, { "epoch": 3.6242682277807345, "grad_norm": 0.2878272831439972, "learning_rate": 3.645604806827408e-08, "loss": 0.1744, "step": 13620 }, { "epoch": 3.6245343267695582, "grad_norm": 0.2947278320789337, "learning_rate": 3.644287541438684e-08, "loss": 0.1997, "step": 13621 }, { "epoch": 3.624800425758382, "grad_norm": 0.2935909330844879, "learning_rate": 3.642970461048248e-08, "loss": 0.1812, "step": 13622 }, { "epoch": 3.625066524747206, "grad_norm": 0.2857137620449066, "learning_rate": 3.6416535656944356e-08, "loss": 0.1918, "step": 13623 }, { "epoch": 3.6253326237360297, "grad_norm": 0.31415605545043945, "learning_rate": 3.640336855415579e-08, "loss": 0.188, "step": 13624 }, { "epoch": 3.6255987227248534, "grad_norm": 0.27658265829086304, "learning_rate": 3.6390203302500035e-08, "loss": 0.1742, "step": 13625 }, { "epoch": 3.6258648217136775, "grad_norm": 0.31239691376686096, "learning_rate": 3.63770399023603e-08, "loss": 0.1971, "step": 13626 }, { "epoch": 3.626130920702501, "grad_norm": 0.31162258982658386, "learning_rate": 3.636387835411973e-08, "loss": 0.2056, "step": 13627 }, { "epoch": 3.6263970196913253, "grad_norm": 0.297733873128891, "learning_rate": 3.635071865816147e-08, "loss": 0.1935, "step": 13628 }, { "epoch": 3.626663118680149, "grad_norm": 0.29761338233947754, "learning_rate": 3.6337560814868495e-08, "loss": 0.1786, "step": 13629 }, { "epoch": 3.6269292176689727, "grad_norm": 0.33333536982536316, "learning_rate": 3.632440482462382e-08, "loss": 0.2109, "step": 13630 }, { "epoch": 3.627195316657797, "grad_norm": 0.30714431405067444, "learning_rate": 3.63112506878104e-08, "loss": 0.1923, "step": 13631 }, { "epoch": 3.6274614156466205, "grad_norm": 0.29044878482818604, "learning_rate": 3.6298098404811104e-08, "loss": 0.1999, "step": 13632 }, { "epoch": 3.6277275146354446, "grad_norm": 0.2995046377182007, "learning_rate": 3.6284947976008765e-08, "loss": 0.1952, "step": 13633 }, { "epoch": 3.6279936136242683, "grad_norm": 0.3294979929924011, "learning_rate": 3.627179940178615e-08, "loss": 0.2068, "step": 13634 }, { "epoch": 3.628259712613092, "grad_norm": 0.2998814284801483, "learning_rate": 3.625865268252598e-08, "loss": 0.1782, "step": 13635 }, { "epoch": 3.628525811601916, "grad_norm": 0.2625125050544739, "learning_rate": 3.6245507818610946e-08, "loss": 0.1694, "step": 13636 }, { "epoch": 3.62879191059074, "grad_norm": 0.2981908321380615, "learning_rate": 3.6232364810423654e-08, "loss": 0.19, "step": 13637 }, { "epoch": 3.6290580095795635, "grad_norm": 0.27320533990859985, "learning_rate": 3.6219223658346634e-08, "loss": 0.1869, "step": 13638 }, { "epoch": 3.6293241085683876, "grad_norm": 0.32581421732902527, "learning_rate": 3.62060843627624e-08, "loss": 0.1982, "step": 13639 }, { "epoch": 3.6295902075572113, "grad_norm": 0.2971276044845581, "learning_rate": 3.619294692405345e-08, "loss": 0.2017, "step": 13640 }, { "epoch": 3.629856306546035, "grad_norm": 0.3097132742404938, "learning_rate": 3.617981134260211e-08, "loss": 0.1879, "step": 13641 }, { "epoch": 3.630122405534859, "grad_norm": 0.3613266050815582, "learning_rate": 3.616667761879074e-08, "loss": 0.2248, "step": 13642 }, { "epoch": 3.630388504523683, "grad_norm": 0.31593406200408936, "learning_rate": 3.615354575300166e-08, "loss": 0.2049, "step": 13643 }, { "epoch": 3.6306546035125065, "grad_norm": 0.2816426455974579, "learning_rate": 3.614041574561708e-08, "loss": 0.1904, "step": 13644 }, { "epoch": 3.6309207025013306, "grad_norm": 0.30228856205940247, "learning_rate": 3.612728759701919e-08, "loss": 0.1974, "step": 13645 }, { "epoch": 3.6311868014901543, "grad_norm": 0.3213527202606201, "learning_rate": 3.611416130759014e-08, "loss": 0.1942, "step": 13646 }, { "epoch": 3.631452900478978, "grad_norm": 0.31488752365112305, "learning_rate": 3.6101036877711956e-08, "loss": 0.1987, "step": 13647 }, { "epoch": 3.631718999467802, "grad_norm": 0.3260347545146942, "learning_rate": 3.6087914307766666e-08, "loss": 0.1994, "step": 13648 }, { "epoch": 3.6319850984566258, "grad_norm": 0.2949398159980774, "learning_rate": 3.607479359813625e-08, "loss": 0.1913, "step": 13649 }, { "epoch": 3.6322511974454494, "grad_norm": 0.38390469551086426, "learning_rate": 3.606167474920261e-08, "loss": 0.191, "step": 13650 }, { "epoch": 3.6325172964342736, "grad_norm": 0.2944716811180115, "learning_rate": 3.60485577613476e-08, "loss": 0.1814, "step": 13651 }, { "epoch": 3.6327833954230973, "grad_norm": 0.3025912344455719, "learning_rate": 3.603544263495303e-08, "loss": 0.2019, "step": 13652 }, { "epoch": 3.6330494944119214, "grad_norm": 0.30299079418182373, "learning_rate": 3.602232937040064e-08, "loss": 0.1837, "step": 13653 }, { "epoch": 3.633315593400745, "grad_norm": 0.31142422556877136, "learning_rate": 3.600921796807214e-08, "loss": 0.1974, "step": 13654 }, { "epoch": 3.633581692389569, "grad_norm": 0.29554855823516846, "learning_rate": 3.599610842834917e-08, "loss": 0.1871, "step": 13655 }, { "epoch": 3.633847791378393, "grad_norm": 0.29152631759643555, "learning_rate": 3.598300075161327e-08, "loss": 0.1902, "step": 13656 }, { "epoch": 3.6341138903672165, "grad_norm": 0.2925836741924286, "learning_rate": 3.5969894938246004e-08, "loss": 0.1949, "step": 13657 }, { "epoch": 3.6343799893560407, "grad_norm": 0.34890642762184143, "learning_rate": 3.595679098862884e-08, "loss": 0.2004, "step": 13658 }, { "epoch": 3.6346460883448644, "grad_norm": 0.3163938522338867, "learning_rate": 3.5943688903143196e-08, "loss": 0.2171, "step": 13659 }, { "epoch": 3.634912187333688, "grad_norm": 0.32501259446144104, "learning_rate": 3.593058868217049e-08, "loss": 0.2007, "step": 13660 }, { "epoch": 3.635178286322512, "grad_norm": 0.29832595586776733, "learning_rate": 3.591749032609197e-08, "loss": 0.203, "step": 13661 }, { "epoch": 3.635444385311336, "grad_norm": 0.3116350471973419, "learning_rate": 3.5904393835288906e-08, "loss": 0.1873, "step": 13662 }, { "epoch": 3.6357104843001595, "grad_norm": 0.3067229092121124, "learning_rate": 3.5891299210142514e-08, "loss": 0.2045, "step": 13663 }, { "epoch": 3.6359765832889837, "grad_norm": 0.2958590090274811, "learning_rate": 3.587820645103398e-08, "loss": 0.1947, "step": 13664 }, { "epoch": 3.6362426822778073, "grad_norm": 0.3007957637310028, "learning_rate": 3.586511555834434e-08, "loss": 0.1834, "step": 13665 }, { "epoch": 3.636508781266631, "grad_norm": 0.29869237542152405, "learning_rate": 3.585202653245465e-08, "loss": 0.2015, "step": 13666 }, { "epoch": 3.636774880255455, "grad_norm": 0.29248932003974915, "learning_rate": 3.5838939373745914e-08, "loss": 0.1845, "step": 13667 }, { "epoch": 3.637040979244279, "grad_norm": 0.3122803270816803, "learning_rate": 3.582585408259906e-08, "loss": 0.201, "step": 13668 }, { "epoch": 3.6373070782331025, "grad_norm": 0.3108362555503845, "learning_rate": 3.5812770659394964e-08, "loss": 0.2017, "step": 13669 }, { "epoch": 3.6375731772219266, "grad_norm": 0.29004189372062683, "learning_rate": 3.579968910451446e-08, "loss": 0.1859, "step": 13670 }, { "epoch": 3.6378392762107503, "grad_norm": 0.3169713020324707, "learning_rate": 3.578660941833831e-08, "loss": 0.1926, "step": 13671 }, { "epoch": 3.638105375199574, "grad_norm": 0.3090285062789917, "learning_rate": 3.577353160124723e-08, "loss": 0.1966, "step": 13672 }, { "epoch": 3.638371474188398, "grad_norm": 0.27887803316116333, "learning_rate": 3.5760455653621914e-08, "loss": 0.1913, "step": 13673 }, { "epoch": 3.638637573177222, "grad_norm": 0.2961256802082062, "learning_rate": 3.57473815758429e-08, "loss": 0.1935, "step": 13674 }, { "epoch": 3.638903672166046, "grad_norm": 0.290291428565979, "learning_rate": 3.5734309368290796e-08, "loss": 0.1849, "step": 13675 }, { "epoch": 3.6391697711548696, "grad_norm": 0.30654796957969666, "learning_rate": 3.572123903134606e-08, "loss": 0.1931, "step": 13676 }, { "epoch": 3.6394358701436933, "grad_norm": 0.30925652384757996, "learning_rate": 3.570817056538917e-08, "loss": 0.199, "step": 13677 }, { "epoch": 3.6397019691325174, "grad_norm": 0.3223821818828583, "learning_rate": 3.56951039708005e-08, "loss": 0.2008, "step": 13678 }, { "epoch": 3.639968068121341, "grad_norm": 0.28839004039764404, "learning_rate": 3.568203924796043e-08, "loss": 0.186, "step": 13679 }, { "epoch": 3.6402341671101652, "grad_norm": 0.3115370571613312, "learning_rate": 3.5668976397249154e-08, "loss": 0.2045, "step": 13680 }, { "epoch": 3.640500266098989, "grad_norm": 0.28323158621788025, "learning_rate": 3.5655915419046947e-08, "loss": 0.1706, "step": 13681 }, { "epoch": 3.6407663650878126, "grad_norm": 0.3160645365715027, "learning_rate": 3.5642856313734006e-08, "loss": 0.2198, "step": 13682 }, { "epoch": 3.6410324640766367, "grad_norm": 0.29657405614852905, "learning_rate": 3.562979908169037e-08, "loss": 0.1937, "step": 13683 }, { "epoch": 3.6412985630654604, "grad_norm": 0.30019840598106384, "learning_rate": 3.561674372329616e-08, "loss": 0.2005, "step": 13684 }, { "epoch": 3.641564662054284, "grad_norm": 0.2921920418739319, "learning_rate": 3.560369023893137e-08, "loss": 0.1873, "step": 13685 }, { "epoch": 3.641830761043108, "grad_norm": 0.2694137990474701, "learning_rate": 3.5590638628975956e-08, "loss": 0.1832, "step": 13686 }, { "epoch": 3.642096860031932, "grad_norm": 0.2866109311580658, "learning_rate": 3.557758889380981e-08, "loss": 0.1833, "step": 13687 }, { "epoch": 3.6423629590207556, "grad_norm": 0.2880009710788727, "learning_rate": 3.556454103381278e-08, "loss": 0.1977, "step": 13688 }, { "epoch": 3.6426290580095797, "grad_norm": 0.30098333954811096, "learning_rate": 3.555149504936466e-08, "loss": 0.1872, "step": 13689 }, { "epoch": 3.6428951569984034, "grad_norm": 0.2889758050441742, "learning_rate": 3.5538450940845176e-08, "loss": 0.1988, "step": 13690 }, { "epoch": 3.643161255987227, "grad_norm": 0.2811674475669861, "learning_rate": 3.5525408708634043e-08, "loss": 0.1774, "step": 13691 }, { "epoch": 3.643427354976051, "grad_norm": 0.2777928113937378, "learning_rate": 3.5512368353110836e-08, "loss": 0.1924, "step": 13692 }, { "epoch": 3.643693453964875, "grad_norm": 0.29232460260391235, "learning_rate": 3.549932987465512e-08, "loss": 0.1782, "step": 13693 }, { "epoch": 3.6439595529536986, "grad_norm": 0.28862935304641724, "learning_rate": 3.548629327364646e-08, "loss": 0.1913, "step": 13694 }, { "epoch": 3.6442256519425227, "grad_norm": 0.2829243838787079, "learning_rate": 3.547325855046429e-08, "loss": 0.1905, "step": 13695 }, { "epoch": 3.6444917509313464, "grad_norm": 0.2886488735675812, "learning_rate": 3.546022570548802e-08, "loss": 0.1953, "step": 13696 }, { "epoch": 3.64475784992017, "grad_norm": 0.2936088740825653, "learning_rate": 3.5447194739096997e-08, "loss": 0.1954, "step": 13697 }, { "epoch": 3.645023948908994, "grad_norm": 0.2927030622959137, "learning_rate": 3.5434165651670536e-08, "loss": 0.1927, "step": 13698 }, { "epoch": 3.645290047897818, "grad_norm": 0.2780410051345825, "learning_rate": 3.542113844358789e-08, "loss": 0.1834, "step": 13699 }, { "epoch": 3.645556146886642, "grad_norm": 0.27504462003707886, "learning_rate": 3.540811311522821e-08, "loss": 0.1812, "step": 13700 }, { "epoch": 3.6458222458754657, "grad_norm": 0.28823423385620117, "learning_rate": 3.5395089666970636e-08, "loss": 0.1654, "step": 13701 }, { "epoch": 3.64608834486429, "grad_norm": 0.29688793420791626, "learning_rate": 3.5382068099194295e-08, "loss": 0.1932, "step": 13702 }, { "epoch": 3.6463544438531135, "grad_norm": 0.3223836123943329, "learning_rate": 3.536904841227813e-08, "loss": 0.1901, "step": 13703 }, { "epoch": 3.646620542841937, "grad_norm": 0.28537502884864807, "learning_rate": 3.535603060660116e-08, "loss": 0.1802, "step": 13704 }, { "epoch": 3.6468866418307613, "grad_norm": 0.34337684512138367, "learning_rate": 3.534301468254229e-08, "loss": 0.1938, "step": 13705 }, { "epoch": 3.647152740819585, "grad_norm": 0.34209033846855164, "learning_rate": 3.5330000640480396e-08, "loss": 0.2216, "step": 13706 }, { "epoch": 3.6474188398084086, "grad_norm": 0.2876940667629242, "learning_rate": 3.531698848079425e-08, "loss": 0.1733, "step": 13707 }, { "epoch": 3.6476849387972328, "grad_norm": 0.2893463373184204, "learning_rate": 3.530397820386263e-08, "loss": 0.1923, "step": 13708 }, { "epoch": 3.6479510377860564, "grad_norm": 0.28389084339141846, "learning_rate": 3.5290969810064255e-08, "loss": 0.1871, "step": 13709 }, { "epoch": 3.64821713677488, "grad_norm": 0.29398107528686523, "learning_rate": 3.5277963299777703e-08, "loss": 0.1827, "step": 13710 }, { "epoch": 3.6484832357637043, "grad_norm": 0.286803275346756, "learning_rate": 3.526495867338158e-08, "loss": 0.1721, "step": 13711 }, { "epoch": 3.648749334752528, "grad_norm": 0.2776166796684265, "learning_rate": 3.525195593125444e-08, "loss": 0.1753, "step": 13712 }, { "epoch": 3.6490154337413516, "grad_norm": 0.2976242005825043, "learning_rate": 3.523895507377475e-08, "loss": 0.1953, "step": 13713 }, { "epoch": 3.6492815327301757, "grad_norm": 0.2948189675807953, "learning_rate": 3.5225956101320916e-08, "loss": 0.2068, "step": 13714 }, { "epoch": 3.6495476317189994, "grad_norm": 0.2740529477596283, "learning_rate": 3.5212959014271317e-08, "loss": 0.1856, "step": 13715 }, { "epoch": 3.649813730707823, "grad_norm": 0.28547403216362, "learning_rate": 3.519996381300426e-08, "loss": 0.1841, "step": 13716 }, { "epoch": 3.6500798296966472, "grad_norm": 0.29331809282302856, "learning_rate": 3.5186970497898016e-08, "loss": 0.1951, "step": 13717 }, { "epoch": 3.650345928685471, "grad_norm": 0.3196975588798523, "learning_rate": 3.51739790693308e-08, "loss": 0.2106, "step": 13718 }, { "epoch": 3.6506120276742946, "grad_norm": 0.2941480278968811, "learning_rate": 3.516098952768071e-08, "loss": 0.1998, "step": 13719 }, { "epoch": 3.6508781266631187, "grad_norm": 0.2960681915283203, "learning_rate": 3.514800187332585e-08, "loss": 0.1982, "step": 13720 }, { "epoch": 3.6511442256519424, "grad_norm": 0.2832346558570862, "learning_rate": 3.513501610664431e-08, "loss": 0.1862, "step": 13721 }, { "epoch": 3.651410324640766, "grad_norm": 0.2966080605983734, "learning_rate": 3.5122032228013996e-08, "loss": 0.1965, "step": 13722 }, { "epoch": 3.65167642362959, "grad_norm": 0.3294852077960968, "learning_rate": 3.510905023781287e-08, "loss": 0.2007, "step": 13723 }, { "epoch": 3.651942522618414, "grad_norm": 0.30232012271881104, "learning_rate": 3.50960701364188e-08, "loss": 0.1849, "step": 13724 }, { "epoch": 3.652208621607238, "grad_norm": 0.2738446891307831, "learning_rate": 3.5083091924209616e-08, "loss": 0.187, "step": 13725 }, { "epoch": 3.6524747205960617, "grad_norm": 0.2852874994277954, "learning_rate": 3.5070115601563066e-08, "loss": 0.1991, "step": 13726 }, { "epoch": 3.652740819584886, "grad_norm": 0.32047590613365173, "learning_rate": 3.5057141168856894e-08, "loss": 0.2073, "step": 13727 }, { "epoch": 3.6530069185737095, "grad_norm": 0.2918626368045807, "learning_rate": 3.5044168626468685e-08, "loss": 0.1955, "step": 13728 }, { "epoch": 3.653273017562533, "grad_norm": 0.3528236746788025, "learning_rate": 3.5031197974776075e-08, "loss": 0.1891, "step": 13729 }, { "epoch": 3.6535391165513573, "grad_norm": 0.2593407928943634, "learning_rate": 3.5018229214156604e-08, "loss": 0.1657, "step": 13730 }, { "epoch": 3.653805215540181, "grad_norm": 0.317417174577713, "learning_rate": 3.500526234498775e-08, "loss": 0.2036, "step": 13731 }, { "epoch": 3.6540713145290047, "grad_norm": 0.2979094386100769, "learning_rate": 3.4992297367646953e-08, "loss": 0.1744, "step": 13732 }, { "epoch": 3.654337413517829, "grad_norm": 0.293178915977478, "learning_rate": 3.49793342825116e-08, "loss": 0.196, "step": 13733 }, { "epoch": 3.6546035125066525, "grad_norm": 0.3026593327522278, "learning_rate": 3.496637308995899e-08, "loss": 0.1876, "step": 13734 }, { "epoch": 3.654869611495476, "grad_norm": 0.2947184443473816, "learning_rate": 3.49534137903664e-08, "loss": 0.2038, "step": 13735 }, { "epoch": 3.6551357104843003, "grad_norm": 0.2913855314254761, "learning_rate": 3.4940456384111085e-08, "loss": 0.1903, "step": 13736 }, { "epoch": 3.655401809473124, "grad_norm": 0.3046502470970154, "learning_rate": 3.4927500871570115e-08, "loss": 0.1944, "step": 13737 }, { "epoch": 3.6556679084619477, "grad_norm": 0.2691918611526489, "learning_rate": 3.491454725312065e-08, "loss": 0.1751, "step": 13738 }, { "epoch": 3.655934007450772, "grad_norm": 0.30984053015708923, "learning_rate": 3.490159552913972e-08, "loss": 0.1904, "step": 13739 }, { "epoch": 3.6562001064395955, "grad_norm": 0.3000321090221405, "learning_rate": 3.4888645700004305e-08, "loss": 0.1866, "step": 13740 }, { "epoch": 3.656466205428419, "grad_norm": 0.32914167642593384, "learning_rate": 3.487569776609139e-08, "loss": 0.1922, "step": 13741 }, { "epoch": 3.6567323044172433, "grad_norm": 0.2976192235946655, "learning_rate": 3.48627517277778e-08, "loss": 0.1822, "step": 13742 }, { "epoch": 3.656998403406067, "grad_norm": 0.300423264503479, "learning_rate": 3.484980758544036e-08, "loss": 0.1851, "step": 13743 }, { "epoch": 3.6572645023948906, "grad_norm": 0.3191455900669098, "learning_rate": 3.483686533945587e-08, "loss": 0.1978, "step": 13744 }, { "epoch": 3.6575306013837148, "grad_norm": 0.30914220213890076, "learning_rate": 3.482392499020107e-08, "loss": 0.2122, "step": 13745 }, { "epoch": 3.6577967003725385, "grad_norm": 0.30701011419296265, "learning_rate": 3.4810986538052555e-08, "loss": 0.2111, "step": 13746 }, { "epoch": 3.6580627993613626, "grad_norm": 0.28510499000549316, "learning_rate": 3.479804998338696e-08, "loss": 0.1847, "step": 13747 }, { "epoch": 3.6583288983501863, "grad_norm": 0.2955024838447571, "learning_rate": 3.478511532658083e-08, "loss": 0.1851, "step": 13748 }, { "epoch": 3.65859499733901, "grad_norm": 0.3129199147224426, "learning_rate": 3.4772182568010666e-08, "loss": 0.2005, "step": 13749 }, { "epoch": 3.658861096327834, "grad_norm": 0.2804308235645294, "learning_rate": 3.4759251708052904e-08, "loss": 0.1727, "step": 13750 }, { "epoch": 3.6591271953166578, "grad_norm": 0.2903203070163727, "learning_rate": 3.4746322747083935e-08, "loss": 0.1888, "step": 13751 }, { "epoch": 3.659393294305482, "grad_norm": 0.28582820296287537, "learning_rate": 3.4733395685480075e-08, "loss": 0.1745, "step": 13752 }, { "epoch": 3.6596593932943056, "grad_norm": 0.3109419047832489, "learning_rate": 3.4720470523617597e-08, "loss": 0.1928, "step": 13753 }, { "epoch": 3.6599254922831292, "grad_norm": 0.28272247314453125, "learning_rate": 3.470754726187277e-08, "loss": 0.1938, "step": 13754 }, { "epoch": 3.6601915912719534, "grad_norm": 0.31259310245513916, "learning_rate": 3.4694625900621675e-08, "loss": 0.1958, "step": 13755 }, { "epoch": 3.660457690260777, "grad_norm": 0.2942284941673279, "learning_rate": 3.4681706440240456e-08, "loss": 0.1703, "step": 13756 }, { "epoch": 3.6607237892496007, "grad_norm": 0.2876710593700409, "learning_rate": 3.466878888110517e-08, "loss": 0.1963, "step": 13757 }, { "epoch": 3.660989888238425, "grad_norm": 0.2774095833301544, "learning_rate": 3.46558732235918e-08, "loss": 0.1722, "step": 13758 }, { "epoch": 3.6612559872272485, "grad_norm": 0.2885681390762329, "learning_rate": 3.464295946807631e-08, "loss": 0.1995, "step": 13759 }, { "epoch": 3.6615220862160722, "grad_norm": 0.29820045828819275, "learning_rate": 3.463004761493459e-08, "loss": 0.1946, "step": 13760 }, { "epoch": 3.6617881852048964, "grad_norm": 0.32003262639045715, "learning_rate": 3.461713766454243e-08, "loss": 0.1855, "step": 13761 }, { "epoch": 3.66205428419372, "grad_norm": 0.2843213975429535, "learning_rate": 3.460422961727562e-08, "loss": 0.1824, "step": 13762 }, { "epoch": 3.6623203831825437, "grad_norm": 0.2903684079647064, "learning_rate": 3.459132347350993e-08, "loss": 0.191, "step": 13763 }, { "epoch": 3.662586482171368, "grad_norm": 0.28127601742744446, "learning_rate": 3.4578419233620936e-08, "loss": 0.1809, "step": 13764 }, { "epoch": 3.6628525811601915, "grad_norm": 0.26682135462760925, "learning_rate": 3.456551689798429e-08, "loss": 0.1711, "step": 13765 }, { "epoch": 3.663118680149015, "grad_norm": 0.2687990367412567, "learning_rate": 3.455261646697557e-08, "loss": 0.1747, "step": 13766 }, { "epoch": 3.6633847791378393, "grad_norm": 0.29073163866996765, "learning_rate": 3.4539717940970234e-08, "loss": 0.1922, "step": 13767 }, { "epoch": 3.663650878126663, "grad_norm": 0.28175443410873413, "learning_rate": 3.4526821320343746e-08, "loss": 0.1881, "step": 13768 }, { "epoch": 3.6639169771154867, "grad_norm": 0.3002089858055115, "learning_rate": 3.45139266054715e-08, "loss": 0.1954, "step": 13769 }, { "epoch": 3.664183076104311, "grad_norm": 0.6938426494598389, "learning_rate": 3.4501033796728805e-08, "loss": 0.1872, "step": 13770 }, { "epoch": 3.6644491750931345, "grad_norm": 0.29663190245628357, "learning_rate": 3.4488142894490956e-08, "loss": 0.1936, "step": 13771 }, { "epoch": 3.6647152740819586, "grad_norm": 0.2862900197505951, "learning_rate": 3.4475253899133197e-08, "loss": 0.192, "step": 13772 }, { "epoch": 3.6649813730707823, "grad_norm": 0.28269538283348083, "learning_rate": 3.446236681103064e-08, "loss": 0.1731, "step": 13773 }, { "epoch": 3.6652474720596064, "grad_norm": 0.34832534193992615, "learning_rate": 3.444948163055841e-08, "loss": 0.1925, "step": 13774 }, { "epoch": 3.66551357104843, "grad_norm": 0.3009616434574127, "learning_rate": 3.443659835809157e-08, "loss": 0.1908, "step": 13775 }, { "epoch": 3.665779670037254, "grad_norm": 0.3414000868797302, "learning_rate": 3.4423716994005115e-08, "loss": 0.1788, "step": 13776 }, { "epoch": 3.666045769026078, "grad_norm": 0.3098693788051605, "learning_rate": 3.4410837538674e-08, "loss": 0.2068, "step": 13777 }, { "epoch": 3.6663118680149016, "grad_norm": 0.3141530156135559, "learning_rate": 3.439795999247309e-08, "loss": 0.2049, "step": 13778 }, { "epoch": 3.6665779670037253, "grad_norm": 0.2874550223350525, "learning_rate": 3.438508435577725e-08, "loss": 0.1743, "step": 13779 }, { "epoch": 3.6668440659925494, "grad_norm": 0.32321202754974365, "learning_rate": 3.437221062896125e-08, "loss": 0.1962, "step": 13780 }, { "epoch": 3.667110164981373, "grad_norm": 0.31600049138069153, "learning_rate": 3.435933881239979e-08, "loss": 0.2104, "step": 13781 }, { "epoch": 3.667376263970197, "grad_norm": 0.3096664249897003, "learning_rate": 3.434646890646752e-08, "loss": 0.1844, "step": 13782 }, { "epoch": 3.667642362959021, "grad_norm": 0.3328140079975128, "learning_rate": 3.4333600911539105e-08, "loss": 0.2021, "step": 13783 }, { "epoch": 3.6679084619478446, "grad_norm": 0.29956454038619995, "learning_rate": 3.4320734827989054e-08, "loss": 0.1928, "step": 13784 }, { "epoch": 3.6681745609366683, "grad_norm": 0.298586905002594, "learning_rate": 3.430787065619186e-08, "loss": 0.1841, "step": 13785 }, { "epoch": 3.6684406599254924, "grad_norm": 0.3101212978363037, "learning_rate": 3.429500839652198e-08, "loss": 0.194, "step": 13786 }, { "epoch": 3.668706758914316, "grad_norm": 0.28542399406433105, "learning_rate": 3.428214804935382e-08, "loss": 0.1811, "step": 13787 }, { "epoch": 3.6689728579031398, "grad_norm": 0.3051784634590149, "learning_rate": 3.426928961506168e-08, "loss": 0.1943, "step": 13788 }, { "epoch": 3.669238956891964, "grad_norm": 0.31318917870521545, "learning_rate": 3.4256433094019865e-08, "loss": 0.2083, "step": 13789 }, { "epoch": 3.6695050558807876, "grad_norm": 0.30634307861328125, "learning_rate": 3.424357848660261e-08, "loss": 0.1827, "step": 13790 }, { "epoch": 3.6697711548696113, "grad_norm": 0.29137122631073, "learning_rate": 3.423072579318401e-08, "loss": 0.1814, "step": 13791 }, { "epoch": 3.6700372538584354, "grad_norm": 0.2802693843841553, "learning_rate": 3.421787501413822e-08, "loss": 0.1724, "step": 13792 }, { "epoch": 3.670303352847259, "grad_norm": 0.29486244916915894, "learning_rate": 3.420502614983929e-08, "loss": 0.1935, "step": 13793 }, { "epoch": 3.670569451836083, "grad_norm": 0.290243923664093, "learning_rate": 3.4192179200661205e-08, "loss": 0.2004, "step": 13794 }, { "epoch": 3.670835550824907, "grad_norm": 0.2970544099807739, "learning_rate": 3.4179334166977926e-08, "loss": 0.204, "step": 13795 }, { "epoch": 3.6711016498137305, "grad_norm": 0.29714077711105347, "learning_rate": 3.4166491049163325e-08, "loss": 0.1923, "step": 13796 }, { "epoch": 3.6713677488025547, "grad_norm": 0.3024020195007324, "learning_rate": 3.415364984759123e-08, "loss": 0.2156, "step": 13797 }, { "epoch": 3.6716338477913784, "grad_norm": 0.3414415717124939, "learning_rate": 3.414081056263542e-08, "loss": 0.199, "step": 13798 }, { "epoch": 3.6718999467802025, "grad_norm": 0.2932784855365753, "learning_rate": 3.4127973194669646e-08, "loss": 0.1839, "step": 13799 }, { "epoch": 3.672166045769026, "grad_norm": 0.28850576281547546, "learning_rate": 3.4115137744067515e-08, "loss": 0.2079, "step": 13800 }, { "epoch": 3.67243214475785, "grad_norm": 0.29129132628440857, "learning_rate": 3.410230421120264e-08, "loss": 0.1797, "step": 13801 }, { "epoch": 3.672698243746674, "grad_norm": 0.3347029983997345, "learning_rate": 3.40894725964486e-08, "loss": 0.2001, "step": 13802 }, { "epoch": 3.6729643427354977, "grad_norm": 0.29158303141593933, "learning_rate": 3.40766429001789e-08, "loss": 0.1761, "step": 13803 }, { "epoch": 3.6732304417243213, "grad_norm": 0.289782851934433, "learning_rate": 3.406381512276693e-08, "loss": 0.1906, "step": 13804 }, { "epoch": 3.6734965407131455, "grad_norm": 0.3087870180606842, "learning_rate": 3.405098926458609e-08, "loss": 0.1864, "step": 13805 }, { "epoch": 3.673762639701969, "grad_norm": 0.2939233183860779, "learning_rate": 3.403816532600974e-08, "loss": 0.1659, "step": 13806 }, { "epoch": 3.674028738690793, "grad_norm": 0.29765576124191284, "learning_rate": 3.4025343307411115e-08, "loss": 0.1911, "step": 13807 }, { "epoch": 3.674294837679617, "grad_norm": 0.2960168123245239, "learning_rate": 3.401252320916345e-08, "loss": 0.2001, "step": 13808 }, { "epoch": 3.6745609366684406, "grad_norm": 0.3144122362136841, "learning_rate": 3.3999705031639925e-08, "loss": 0.2074, "step": 13809 }, { "epoch": 3.6748270356572643, "grad_norm": 0.28814736008644104, "learning_rate": 3.398688877521358e-08, "loss": 0.1909, "step": 13810 }, { "epoch": 3.6750931346460884, "grad_norm": 0.3015974462032318, "learning_rate": 3.397407444025752e-08, "loss": 0.1865, "step": 13811 }, { "epoch": 3.675359233634912, "grad_norm": 0.28764259815216064, "learning_rate": 3.39612620271447e-08, "loss": 0.1749, "step": 13812 }, { "epoch": 3.675625332623736, "grad_norm": 0.299241840839386, "learning_rate": 3.394845153624809e-08, "loss": 0.191, "step": 13813 }, { "epoch": 3.67589143161256, "grad_norm": 0.30125969648361206, "learning_rate": 3.3935642967940546e-08, "loss": 0.2004, "step": 13814 }, { "epoch": 3.6761575306013836, "grad_norm": 0.3145284354686737, "learning_rate": 3.3922836322594906e-08, "loss": 0.1902, "step": 13815 }, { "epoch": 3.6764236295902073, "grad_norm": 0.29835692048072815, "learning_rate": 3.3910031600583924e-08, "loss": 0.1716, "step": 13816 }, { "epoch": 3.6766897285790314, "grad_norm": 0.2910650074481964, "learning_rate": 3.389722880228033e-08, "loss": 0.1743, "step": 13817 }, { "epoch": 3.676955827567855, "grad_norm": 0.3146950304508209, "learning_rate": 3.3884427928056805e-08, "loss": 0.1922, "step": 13818 }, { "epoch": 3.6772219265566792, "grad_norm": 0.3216951787471771, "learning_rate": 3.387162897828587e-08, "loss": 0.2107, "step": 13819 }, { "epoch": 3.677488025545503, "grad_norm": 0.2604914605617523, "learning_rate": 3.385883195334013e-08, "loss": 0.1693, "step": 13820 }, { "epoch": 3.677754124534327, "grad_norm": 0.308725506067276, "learning_rate": 3.3846036853592055e-08, "loss": 0.1952, "step": 13821 }, { "epoch": 3.6780202235231507, "grad_norm": 0.2785002291202545, "learning_rate": 3.3833243679414104e-08, "loss": 0.1848, "step": 13822 }, { "epoch": 3.6782863225119744, "grad_norm": 0.26357749104499817, "learning_rate": 3.38204524311786e-08, "loss": 0.1589, "step": 13823 }, { "epoch": 3.6785524215007985, "grad_norm": 0.3110492527484894, "learning_rate": 3.38076631092579e-08, "loss": 0.195, "step": 13824 }, { "epoch": 3.678818520489622, "grad_norm": 0.2910807132720947, "learning_rate": 3.379487571402425e-08, "loss": 0.1781, "step": 13825 }, { "epoch": 3.679084619478446, "grad_norm": 0.307800829410553, "learning_rate": 3.3782090245849883e-08, "loss": 0.194, "step": 13826 }, { "epoch": 3.67935071846727, "grad_norm": 0.31787005066871643, "learning_rate": 3.376930670510696e-08, "loss": 0.1966, "step": 13827 }, { "epoch": 3.6796168174560937, "grad_norm": 0.3014086186885834, "learning_rate": 3.375652509216751e-08, "loss": 0.1917, "step": 13828 }, { "epoch": 3.6798829164449174, "grad_norm": 0.2861451804637909, "learning_rate": 3.3743745407403626e-08, "loss": 0.1848, "step": 13829 }, { "epoch": 3.6801490154337415, "grad_norm": 0.3086618185043335, "learning_rate": 3.373096765118728e-08, "loss": 0.1992, "step": 13830 }, { "epoch": 3.680415114422565, "grad_norm": 0.2825751006603241, "learning_rate": 3.3718191823890395e-08, "loss": 0.1698, "step": 13831 }, { "epoch": 3.680681213411389, "grad_norm": 0.30939897894859314, "learning_rate": 3.3705417925884846e-08, "loss": 0.1966, "step": 13832 }, { "epoch": 3.680947312400213, "grad_norm": 0.31358417868614197, "learning_rate": 3.369264595754246e-08, "loss": 0.1956, "step": 13833 }, { "epoch": 3.6812134113890367, "grad_norm": 0.31380149722099304, "learning_rate": 3.367987591923498e-08, "loss": 0.1785, "step": 13834 }, { "epoch": 3.6814795103778604, "grad_norm": 0.2908332049846649, "learning_rate": 3.3667107811334106e-08, "loss": 0.1801, "step": 13835 }, { "epoch": 3.6817456093666845, "grad_norm": 0.2989042401313782, "learning_rate": 3.3654341634211536e-08, "loss": 0.1824, "step": 13836 }, { "epoch": 3.682011708355508, "grad_norm": 0.2940271496772766, "learning_rate": 3.3641577388238774e-08, "loss": 0.1855, "step": 13837 }, { "epoch": 3.682277807344332, "grad_norm": 0.2748432159423828, "learning_rate": 3.3628815073787396e-08, "loss": 0.1925, "step": 13838 }, { "epoch": 3.682543906333156, "grad_norm": 0.3109440505504608, "learning_rate": 3.361605469122889e-08, "loss": 0.2017, "step": 13839 }, { "epoch": 3.6828100053219797, "grad_norm": 0.29075321555137634, "learning_rate": 3.3603296240934665e-08, "loss": 0.1899, "step": 13840 }, { "epoch": 3.6830761043108033, "grad_norm": 0.3076053261756897, "learning_rate": 3.359053972327608e-08, "loss": 0.1858, "step": 13841 }, { "epoch": 3.6833422032996275, "grad_norm": 0.2912510633468628, "learning_rate": 3.357778513862449e-08, "loss": 0.1809, "step": 13842 }, { "epoch": 3.683608302288451, "grad_norm": 0.317072331905365, "learning_rate": 3.356503248735109e-08, "loss": 0.1895, "step": 13843 }, { "epoch": 3.6838744012772753, "grad_norm": 0.297168493270874, "learning_rate": 3.355228176982709e-08, "loss": 0.1783, "step": 13844 }, { "epoch": 3.684140500266099, "grad_norm": 0.30735406279563904, "learning_rate": 3.353953298642366e-08, "loss": 0.1939, "step": 13845 }, { "epoch": 3.684406599254923, "grad_norm": 0.28146034479141235, "learning_rate": 3.352678613751184e-08, "loss": 0.1883, "step": 13846 }, { "epoch": 3.6846726982437468, "grad_norm": 0.297282338142395, "learning_rate": 3.351404122346267e-08, "loss": 0.1882, "step": 13847 }, { "epoch": 3.6849387972325705, "grad_norm": 0.30670249462127686, "learning_rate": 3.350129824464715e-08, "loss": 0.1913, "step": 13848 }, { "epoch": 3.6852048962213946, "grad_norm": 0.31168869137763977, "learning_rate": 3.3488557201436165e-08, "loss": 0.2025, "step": 13849 }, { "epoch": 3.6854709952102183, "grad_norm": 0.30321967601776123, "learning_rate": 3.347581809420058e-08, "loss": 0.1937, "step": 13850 }, { "epoch": 3.685737094199042, "grad_norm": 0.2784610390663147, "learning_rate": 3.34630809233112e-08, "loss": 0.1795, "step": 13851 }, { "epoch": 3.686003193187866, "grad_norm": 0.26032090187072754, "learning_rate": 3.345034568913878e-08, "loss": 0.1722, "step": 13852 }, { "epoch": 3.6862692921766897, "grad_norm": 0.28983473777770996, "learning_rate": 3.3437612392054004e-08, "loss": 0.1924, "step": 13853 }, { "epoch": 3.6865353911655134, "grad_norm": 0.30727487802505493, "learning_rate": 3.342488103242753e-08, "loss": 0.1929, "step": 13854 }, { "epoch": 3.6868014901543376, "grad_norm": 0.27254971861839294, "learning_rate": 3.341215161062989e-08, "loss": 0.1668, "step": 13855 }, { "epoch": 3.6870675891431612, "grad_norm": 0.2871061861515045, "learning_rate": 3.3399424127031605e-08, "loss": 0.2008, "step": 13856 }, { "epoch": 3.687333688131985, "grad_norm": 0.31788119673728943, "learning_rate": 3.338669858200317e-08, "loss": 0.2192, "step": 13857 }, { "epoch": 3.687599787120809, "grad_norm": 0.3142581582069397, "learning_rate": 3.3373974975914976e-08, "loss": 0.1959, "step": 13858 }, { "epoch": 3.6878658861096327, "grad_norm": 0.31641897559165955, "learning_rate": 3.3361253309137374e-08, "loss": 0.2, "step": 13859 }, { "epoch": 3.6881319850984564, "grad_norm": 0.27926701307296753, "learning_rate": 3.334853358204066e-08, "loss": 0.1706, "step": 13860 }, { "epoch": 3.6883980840872805, "grad_norm": 0.30803871154785156, "learning_rate": 3.333581579499511e-08, "loss": 0.196, "step": 13861 }, { "epoch": 3.688664183076104, "grad_norm": 0.2938387095928192, "learning_rate": 3.3323099948370846e-08, "loss": 0.1904, "step": 13862 }, { "epoch": 3.688930282064928, "grad_norm": 0.28709641098976135, "learning_rate": 3.331038604253801e-08, "loss": 0.1852, "step": 13863 }, { "epoch": 3.689196381053752, "grad_norm": 0.30191320180892944, "learning_rate": 3.329767407786671e-08, "loss": 0.2017, "step": 13864 }, { "epoch": 3.6894624800425757, "grad_norm": 0.2955816984176636, "learning_rate": 3.3284964054726896e-08, "loss": 0.193, "step": 13865 }, { "epoch": 3.6897285790314, "grad_norm": 0.3028148114681244, "learning_rate": 3.327225597348856e-08, "loss": 0.2134, "step": 13866 }, { "epoch": 3.6899946780202235, "grad_norm": 0.33456534147262573, "learning_rate": 3.325954983452158e-08, "loss": 0.1901, "step": 13867 }, { "epoch": 3.690260777009047, "grad_norm": 0.3306833505630493, "learning_rate": 3.324684563819583e-08, "loss": 0.1997, "step": 13868 }, { "epoch": 3.6905268759978713, "grad_norm": 0.3002629280090332, "learning_rate": 3.323414338488107e-08, "loss": 0.1759, "step": 13869 }, { "epoch": 3.690792974986695, "grad_norm": 0.2903290390968323, "learning_rate": 3.322144307494705e-08, "loss": 0.1787, "step": 13870 }, { "epoch": 3.691059073975519, "grad_norm": 0.30443987250328064, "learning_rate": 3.3208744708763435e-08, "loss": 0.1932, "step": 13871 }, { "epoch": 3.691325172964343, "grad_norm": 0.29566267132759094, "learning_rate": 3.3196048286699864e-08, "loss": 0.1978, "step": 13872 }, { "epoch": 3.6915912719531665, "grad_norm": 0.3180205821990967, "learning_rate": 3.3183353809125844e-08, "loss": 0.1956, "step": 13873 }, { "epoch": 3.6918573709419906, "grad_norm": 0.30319130420684814, "learning_rate": 3.3170661276410906e-08, "loss": 0.1759, "step": 13874 }, { "epoch": 3.6921234699308143, "grad_norm": 0.2939799726009369, "learning_rate": 3.3157970688924495e-08, "loss": 0.1818, "step": 13875 }, { "epoch": 3.692389568919638, "grad_norm": 0.29299843311309814, "learning_rate": 3.3145282047036005e-08, "loss": 0.1806, "step": 13876 }, { "epoch": 3.692655667908462, "grad_norm": 0.3124901056289673, "learning_rate": 3.313259535111478e-08, "loss": 0.1871, "step": 13877 }, { "epoch": 3.692921766897286, "grad_norm": 0.2656475603580475, "learning_rate": 3.311991060153007e-08, "loss": 0.1833, "step": 13878 }, { "epoch": 3.6931878658861095, "grad_norm": 0.4523405432701111, "learning_rate": 3.3107227798651116e-08, "loss": 0.1957, "step": 13879 }, { "epoch": 3.6934539648749336, "grad_norm": 0.3087809681892395, "learning_rate": 3.309454694284707e-08, "loss": 0.1955, "step": 13880 }, { "epoch": 3.6937200638637573, "grad_norm": 0.2909778952598572, "learning_rate": 3.308186803448708e-08, "loss": 0.188, "step": 13881 }, { "epoch": 3.693986162852581, "grad_norm": 0.2775185704231262, "learning_rate": 3.306919107394013e-08, "loss": 0.1798, "step": 13882 }, { "epoch": 3.694252261841405, "grad_norm": 0.3074883818626404, "learning_rate": 3.305651606157523e-08, "loss": 0.1921, "step": 13883 }, { "epoch": 3.6945183608302288, "grad_norm": 0.2722242772579193, "learning_rate": 3.3043842997761385e-08, "loss": 0.1707, "step": 13884 }, { "epoch": 3.6947844598190525, "grad_norm": 0.30078956484794617, "learning_rate": 3.303117188286738e-08, "loss": 0.2001, "step": 13885 }, { "epoch": 3.6950505588078766, "grad_norm": 0.34027719497680664, "learning_rate": 3.301850271726207e-08, "loss": 0.2006, "step": 13886 }, { "epoch": 3.6953166577967003, "grad_norm": 0.41712823510169983, "learning_rate": 3.300583550131424e-08, "loss": 0.1941, "step": 13887 }, { "epoch": 3.695582756785524, "grad_norm": 0.2994147837162018, "learning_rate": 3.2993170235392585e-08, "loss": 0.1882, "step": 13888 }, { "epoch": 3.695848855774348, "grad_norm": 0.33283567428588867, "learning_rate": 3.298050691986577e-08, "loss": 0.2079, "step": 13889 }, { "epoch": 3.6961149547631718, "grad_norm": 0.33298784494400024, "learning_rate": 3.29678455551024e-08, "loss": 0.1858, "step": 13890 }, { "epoch": 3.696381053751996, "grad_norm": 0.31446364521980286, "learning_rate": 3.295518614147098e-08, "loss": 0.1763, "step": 13891 }, { "epoch": 3.6966471527408196, "grad_norm": 0.3013119101524353, "learning_rate": 3.294252867934001e-08, "loss": 0.1915, "step": 13892 }, { "epoch": 3.6969132517296437, "grad_norm": 0.3045412302017212, "learning_rate": 3.2929873169077914e-08, "loss": 0.2084, "step": 13893 }, { "epoch": 3.6971793507184674, "grad_norm": 0.31412556767463684, "learning_rate": 3.291721961105307e-08, "loss": 0.2133, "step": 13894 }, { "epoch": 3.697445449707291, "grad_norm": 0.28165775537490845, "learning_rate": 3.290456800563378e-08, "loss": 0.1975, "step": 13895 }, { "epoch": 3.697711548696115, "grad_norm": 0.2757234573364258, "learning_rate": 3.289191835318831e-08, "loss": 0.1941, "step": 13896 }, { "epoch": 3.697977647684939, "grad_norm": 0.28690168261528015, "learning_rate": 3.2879270654084843e-08, "loss": 0.1868, "step": 13897 }, { "epoch": 3.6982437466737625, "grad_norm": 0.31925222277641296, "learning_rate": 3.286662490869153e-08, "loss": 0.206, "step": 13898 }, { "epoch": 3.6985098456625867, "grad_norm": 0.31261345744132996, "learning_rate": 3.2853981117376494e-08, "loss": 0.2049, "step": 13899 }, { "epoch": 3.6987759446514104, "grad_norm": 0.3132200539112091, "learning_rate": 3.2841339280507694e-08, "loss": 0.2006, "step": 13900 }, { "epoch": 3.699042043640234, "grad_norm": 0.26870399713516235, "learning_rate": 3.282869939845313e-08, "loss": 0.1752, "step": 13901 }, { "epoch": 3.699308142629058, "grad_norm": 0.2835018038749695, "learning_rate": 3.2816061471580713e-08, "loss": 0.1782, "step": 13902 }, { "epoch": 3.699574241617882, "grad_norm": 0.30092570185661316, "learning_rate": 3.280342550025834e-08, "loss": 0.1979, "step": 13903 }, { "epoch": 3.6998403406067055, "grad_norm": 0.3272494971752167, "learning_rate": 3.2790791484853744e-08, "loss": 0.2096, "step": 13904 }, { "epoch": 3.7001064395955297, "grad_norm": 0.297542929649353, "learning_rate": 3.27781594257347e-08, "loss": 0.1833, "step": 13905 }, { "epoch": 3.7003725385843533, "grad_norm": 0.31194382905960083, "learning_rate": 3.27655293232689e-08, "loss": 0.2055, "step": 13906 }, { "epoch": 3.700638637573177, "grad_norm": 0.3014611303806305, "learning_rate": 3.2752901177823967e-08, "loss": 0.1749, "step": 13907 }, { "epoch": 3.700904736562001, "grad_norm": 0.3163463771343231, "learning_rate": 3.2740274989767504e-08, "loss": 0.2041, "step": 13908 }, { "epoch": 3.701170835550825, "grad_norm": 0.30004364252090454, "learning_rate": 3.2727650759466974e-08, "loss": 0.194, "step": 13909 }, { "epoch": 3.7014369345396485, "grad_norm": 0.3002164959907532, "learning_rate": 3.271502848728985e-08, "loss": 0.1891, "step": 13910 }, { "epoch": 3.7017030335284726, "grad_norm": 0.3119506239891052, "learning_rate": 3.2702408173603547e-08, "loss": 0.1982, "step": 13911 }, { "epoch": 3.7019691325172963, "grad_norm": 0.2876774072647095, "learning_rate": 3.2689789818775405e-08, "loss": 0.1862, "step": 13912 }, { "epoch": 3.7022352315061204, "grad_norm": 0.33038610219955444, "learning_rate": 3.267717342317271e-08, "loss": 0.1929, "step": 13913 }, { "epoch": 3.702501330494944, "grad_norm": 0.3079538345336914, "learning_rate": 3.2664558987162696e-08, "loss": 0.1906, "step": 13914 }, { "epoch": 3.702767429483768, "grad_norm": 0.2855537235736847, "learning_rate": 3.265194651111254e-08, "loss": 0.1975, "step": 13915 }, { "epoch": 3.703033528472592, "grad_norm": 0.30975326895713806, "learning_rate": 3.263933599538935e-08, "loss": 0.1934, "step": 13916 }, { "epoch": 3.7032996274614156, "grad_norm": 0.3056029975414276, "learning_rate": 3.262672744036023e-08, "loss": 0.2007, "step": 13917 }, { "epoch": 3.7035657264502397, "grad_norm": 0.30183738470077515, "learning_rate": 3.261412084639209e-08, "loss": 0.201, "step": 13918 }, { "epoch": 3.7038318254390634, "grad_norm": 0.31399938464164734, "learning_rate": 3.2601516213851944e-08, "loss": 0.2081, "step": 13919 }, { "epoch": 3.704097924427887, "grad_norm": 0.312946617603302, "learning_rate": 3.258891354310666e-08, "loss": 0.197, "step": 13920 }, { "epoch": 3.7043640234167112, "grad_norm": 0.30801424384117126, "learning_rate": 3.257631283452306e-08, "loss": 0.2074, "step": 13921 }, { "epoch": 3.704630122405535, "grad_norm": 0.2908603847026825, "learning_rate": 3.2563714088467934e-08, "loss": 0.2008, "step": 13922 }, { "epoch": 3.7048962213943586, "grad_norm": 0.30936378240585327, "learning_rate": 3.255111730530803e-08, "loss": 0.2033, "step": 13923 }, { "epoch": 3.7051623203831827, "grad_norm": 0.3069901466369629, "learning_rate": 3.2538522485409936e-08, "loss": 0.2141, "step": 13924 }, { "epoch": 3.7054284193720064, "grad_norm": 0.2904931902885437, "learning_rate": 3.252592962914029e-08, "loss": 0.1894, "step": 13925 }, { "epoch": 3.70569451836083, "grad_norm": 0.3070816099643707, "learning_rate": 3.251333873686568e-08, "loss": 0.2005, "step": 13926 }, { "epoch": 3.705960617349654, "grad_norm": 0.32094606757164, "learning_rate": 3.250074980895251e-08, "loss": 0.1943, "step": 13927 }, { "epoch": 3.706226716338478, "grad_norm": 0.2925177216529846, "learning_rate": 3.2488162845767266e-08, "loss": 0.1963, "step": 13928 }, { "epoch": 3.7064928153273016, "grad_norm": 0.2913637161254883, "learning_rate": 3.2475577847676295e-08, "loss": 0.1842, "step": 13929 }, { "epoch": 3.7067589143161257, "grad_norm": 0.27608436346054077, "learning_rate": 3.246299481504593e-08, "loss": 0.1822, "step": 13930 }, { "epoch": 3.7070250133049494, "grad_norm": 0.3211864233016968, "learning_rate": 3.2450413748242435e-08, "loss": 0.1975, "step": 13931 }, { "epoch": 3.707291112293773, "grad_norm": 0.28616341948509216, "learning_rate": 3.243783464763199e-08, "loss": 0.1755, "step": 13932 }, { "epoch": 3.707557211282597, "grad_norm": 0.31322696805000305, "learning_rate": 3.242525751358076e-08, "loss": 0.1972, "step": 13933 }, { "epoch": 3.707823310271421, "grad_norm": 0.30732670426368713, "learning_rate": 3.241268234645483e-08, "loss": 0.1955, "step": 13934 }, { "epoch": 3.7080894092602446, "grad_norm": 0.3171352446079254, "learning_rate": 3.240010914662027e-08, "loss": 0.2062, "step": 13935 }, { "epoch": 3.7083555082490687, "grad_norm": 0.28966575860977173, "learning_rate": 3.238753791444297e-08, "loss": 0.1801, "step": 13936 }, { "epoch": 3.7086216072378924, "grad_norm": 0.2914421558380127, "learning_rate": 3.237496865028889e-08, "loss": 0.1849, "step": 13937 }, { "epoch": 3.7088877062267165, "grad_norm": 0.31186866760253906, "learning_rate": 3.2362401354523894e-08, "loss": 0.1899, "step": 13938 }, { "epoch": 3.70915380521554, "grad_norm": 0.296801894903183, "learning_rate": 3.2349836027513763e-08, "loss": 0.1991, "step": 13939 }, { "epoch": 3.7094199042043643, "grad_norm": 0.3112637996673584, "learning_rate": 3.2337272669624247e-08, "loss": 0.1826, "step": 13940 }, { "epoch": 3.709686003193188, "grad_norm": 0.3022523522377014, "learning_rate": 3.232471128122105e-08, "loss": 0.1989, "step": 13941 }, { "epoch": 3.7099521021820117, "grad_norm": 0.3209155797958374, "learning_rate": 3.23121518626698e-08, "loss": 0.1903, "step": 13942 }, { "epoch": 3.710218201170836, "grad_norm": 0.2987728714942932, "learning_rate": 3.2299594414336073e-08, "loss": 0.1911, "step": 13943 }, { "epoch": 3.7104843001596595, "grad_norm": 0.292362779378891, "learning_rate": 3.2287038936585355e-08, "loss": 0.1963, "step": 13944 }, { "epoch": 3.710750399148483, "grad_norm": 0.30659520626068115, "learning_rate": 3.22744854297831e-08, "loss": 0.1986, "step": 13945 }, { "epoch": 3.7110164981373073, "grad_norm": 0.3011557161808014, "learning_rate": 3.226193389429478e-08, "loss": 0.1864, "step": 13946 }, { "epoch": 3.711282597126131, "grad_norm": 0.3858862817287445, "learning_rate": 3.224938433048563e-08, "loss": 0.1765, "step": 13947 }, { "epoch": 3.7115486961149546, "grad_norm": 0.30542391538619995, "learning_rate": 3.2236836738721005e-08, "loss": 0.18, "step": 13948 }, { "epoch": 3.7118147951037788, "grad_norm": 0.33882373571395874, "learning_rate": 3.222429111936611e-08, "loss": 0.2113, "step": 13949 }, { "epoch": 3.7120808940926024, "grad_norm": 0.2882823944091797, "learning_rate": 3.221174747278612e-08, "loss": 0.1785, "step": 13950 }, { "epoch": 3.712346993081426, "grad_norm": 0.3186173141002655, "learning_rate": 3.219920579934616e-08, "loss": 0.1874, "step": 13951 }, { "epoch": 3.7126130920702503, "grad_norm": 0.2995079755783081, "learning_rate": 3.2186666099411276e-08, "loss": 0.1838, "step": 13952 }, { "epoch": 3.712879191059074, "grad_norm": 0.3044196367263794, "learning_rate": 3.217412837334649e-08, "loss": 0.1962, "step": 13953 }, { "epoch": 3.7131452900478976, "grad_norm": 0.29463568329811096, "learning_rate": 3.2161592621516694e-08, "loss": 0.1977, "step": 13954 }, { "epoch": 3.7134113890367217, "grad_norm": 0.3466273844242096, "learning_rate": 3.214905884428679e-08, "loss": 0.1917, "step": 13955 }, { "epoch": 3.7136774880255454, "grad_norm": 0.31033557653427124, "learning_rate": 3.213652704202161e-08, "loss": 0.2006, "step": 13956 }, { "epoch": 3.713943587014369, "grad_norm": 0.3051753342151642, "learning_rate": 3.212399721508593e-08, "loss": 0.2019, "step": 13957 }, { "epoch": 3.7142096860031932, "grad_norm": 0.29744794964790344, "learning_rate": 3.211146936384445e-08, "loss": 0.1799, "step": 13958 }, { "epoch": 3.714475784992017, "grad_norm": 0.2849949300289154, "learning_rate": 3.2098943488661824e-08, "loss": 0.1784, "step": 13959 }, { "epoch": 3.7147418839808406, "grad_norm": 0.28739845752716064, "learning_rate": 3.208641958990265e-08, "loss": 0.1825, "step": 13960 }, { "epoch": 3.7150079829696647, "grad_norm": 0.31967392563819885, "learning_rate": 3.207389766793146e-08, "loss": 0.1902, "step": 13961 }, { "epoch": 3.7152740819584884, "grad_norm": 0.29754579067230225, "learning_rate": 3.2061377723112784e-08, "loss": 0.191, "step": 13962 }, { "epoch": 3.7155401809473125, "grad_norm": 0.31184306740760803, "learning_rate": 3.204885975581096e-08, "loss": 0.1947, "step": 13963 }, { "epoch": 3.715806279936136, "grad_norm": 0.289223313331604, "learning_rate": 3.2036343766390405e-08, "loss": 0.1942, "step": 13964 }, { "epoch": 3.7160723789249603, "grad_norm": 0.28650760650634766, "learning_rate": 3.2023829755215434e-08, "loss": 0.189, "step": 13965 }, { "epoch": 3.716338477913784, "grad_norm": 0.2726398706436157, "learning_rate": 3.201131772265027e-08, "loss": 0.1768, "step": 13966 }, { "epoch": 3.7166045769026077, "grad_norm": 0.28871026635169983, "learning_rate": 3.199880766905909e-08, "loss": 0.1812, "step": 13967 }, { "epoch": 3.716870675891432, "grad_norm": 0.3012828528881073, "learning_rate": 3.1986299594806076e-08, "loss": 0.1952, "step": 13968 }, { "epoch": 3.7171367748802555, "grad_norm": 0.3063753843307495, "learning_rate": 3.197379350025527e-08, "loss": 0.1935, "step": 13969 }, { "epoch": 3.717402873869079, "grad_norm": 0.31655827164649963, "learning_rate": 3.196128938577072e-08, "loss": 0.2041, "step": 13970 }, { "epoch": 3.7176689728579033, "grad_norm": 0.3095758557319641, "learning_rate": 3.19487872517164e-08, "loss": 0.1794, "step": 13971 }, { "epoch": 3.717935071846727, "grad_norm": 0.31161263585090637, "learning_rate": 3.193628709845616e-08, "loss": 0.1966, "step": 13972 }, { "epoch": 3.7182011708355507, "grad_norm": 0.2808794677257538, "learning_rate": 3.192378892635388e-08, "loss": 0.1653, "step": 13973 }, { "epoch": 3.718467269824375, "grad_norm": 0.32522496581077576, "learning_rate": 3.1911292735773344e-08, "loss": 0.2043, "step": 13974 }, { "epoch": 3.7187333688131985, "grad_norm": 0.30106431245803833, "learning_rate": 3.1898798527078284e-08, "loss": 0.2022, "step": 13975 }, { "epoch": 3.718999467802022, "grad_norm": 0.2661752998828888, "learning_rate": 3.1886306300632385e-08, "loss": 0.1679, "step": 13976 }, { "epoch": 3.7192655667908463, "grad_norm": 0.3080950081348419, "learning_rate": 3.187381605679924e-08, "loss": 0.1871, "step": 13977 }, { "epoch": 3.71953166577967, "grad_norm": 0.2958984971046448, "learning_rate": 3.1861327795942426e-08, "loss": 0.1907, "step": 13978 }, { "epoch": 3.7197977647684937, "grad_norm": 0.30492693185806274, "learning_rate": 3.184884151842544e-08, "loss": 0.1872, "step": 13979 }, { "epoch": 3.720063863757318, "grad_norm": 0.3336305022239685, "learning_rate": 3.183635722461176e-08, "loss": 0.2143, "step": 13980 }, { "epoch": 3.7203299627461415, "grad_norm": 0.29095521569252014, "learning_rate": 3.18238749148647e-08, "loss": 0.1762, "step": 13981 }, { "epoch": 3.720596061734965, "grad_norm": 0.3085625171661377, "learning_rate": 3.1811394589547636e-08, "loss": 0.184, "step": 13982 }, { "epoch": 3.7208621607237893, "grad_norm": 0.273908793926239, "learning_rate": 3.179891624902381e-08, "loss": 0.1715, "step": 13983 }, { "epoch": 3.721128259712613, "grad_norm": 0.3021586537361145, "learning_rate": 3.1786439893656455e-08, "loss": 0.1962, "step": 13984 }, { "epoch": 3.721394358701437, "grad_norm": 0.30241137742996216, "learning_rate": 3.177396552380876e-08, "loss": 0.2046, "step": 13985 }, { "epoch": 3.7216604576902608, "grad_norm": 0.29040518403053284, "learning_rate": 3.1761493139843734e-08, "loss": 0.1875, "step": 13986 }, { "epoch": 3.721926556679085, "grad_norm": 0.3274218738079071, "learning_rate": 3.174902274212448e-08, "loss": 0.1918, "step": 13987 }, { "epoch": 3.7221926556679086, "grad_norm": 0.2731555998325348, "learning_rate": 3.173655433101394e-08, "loss": 0.1782, "step": 13988 }, { "epoch": 3.7224587546567323, "grad_norm": 0.27299293875694275, "learning_rate": 3.172408790687511e-08, "loss": 0.1827, "step": 13989 }, { "epoch": 3.7227248536455564, "grad_norm": 0.31623509526252747, "learning_rate": 3.171162347007077e-08, "loss": 0.1925, "step": 13990 }, { "epoch": 3.72299095263438, "grad_norm": 0.333278089761734, "learning_rate": 3.169916102096376e-08, "loss": 0.184, "step": 13991 }, { "epoch": 3.7232570516232038, "grad_norm": 0.29889699816703796, "learning_rate": 3.1686700559916824e-08, "loss": 0.19, "step": 13992 }, { "epoch": 3.723523150612028, "grad_norm": 0.2906121611595154, "learning_rate": 3.167424208729267e-08, "loss": 0.1821, "step": 13993 }, { "epoch": 3.7237892496008516, "grad_norm": 0.33743107318878174, "learning_rate": 3.1661785603453916e-08, "loss": 0.2186, "step": 13994 }, { "epoch": 3.7240553485896752, "grad_norm": 0.2867896854877472, "learning_rate": 3.164933110876316e-08, "loss": 0.1854, "step": 13995 }, { "epoch": 3.7243214475784994, "grad_norm": 0.3030613660812378, "learning_rate": 3.163687860358288e-08, "loss": 0.1856, "step": 13996 }, { "epoch": 3.724587546567323, "grad_norm": 0.28317898511886597, "learning_rate": 3.162442808827558e-08, "loss": 0.1842, "step": 13997 }, { "epoch": 3.7248536455561467, "grad_norm": 0.30022314190864563, "learning_rate": 3.161197956320367e-08, "loss": 0.1967, "step": 13998 }, { "epoch": 3.725119744544971, "grad_norm": 0.2821776568889618, "learning_rate": 3.159953302872943e-08, "loss": 0.1833, "step": 13999 }, { "epoch": 3.7253858435337945, "grad_norm": 0.29810723662376404, "learning_rate": 3.1587088485215186e-08, "loss": 0.1999, "step": 14000 }, { "epoch": 3.725651942522618, "grad_norm": 0.2786598205566406, "learning_rate": 3.157464593302317e-08, "loss": 0.1837, "step": 14001 }, { "epoch": 3.7259180415114423, "grad_norm": 0.28703275322914124, "learning_rate": 3.156220537251554e-08, "loss": 0.1815, "step": 14002 }, { "epoch": 3.726184140500266, "grad_norm": 0.3092630207538605, "learning_rate": 3.154976680405441e-08, "loss": 0.1909, "step": 14003 }, { "epoch": 3.7264502394890897, "grad_norm": 0.2778889834880829, "learning_rate": 3.153733022800188e-08, "loss": 0.1761, "step": 14004 }, { "epoch": 3.726716338477914, "grad_norm": 0.29274117946624756, "learning_rate": 3.152489564471987e-08, "loss": 0.1902, "step": 14005 }, { "epoch": 3.7269824374667375, "grad_norm": 0.30145397782325745, "learning_rate": 3.151246305457036e-08, "loss": 0.182, "step": 14006 }, { "epoch": 3.727248536455561, "grad_norm": 0.30856359004974365, "learning_rate": 3.150003245791525e-08, "loss": 0.1911, "step": 14007 }, { "epoch": 3.7275146354443853, "grad_norm": 0.28822535276412964, "learning_rate": 3.148760385511631e-08, "loss": 0.1866, "step": 14008 }, { "epoch": 3.727780734433209, "grad_norm": 0.31802940368652344, "learning_rate": 3.147517724653532e-08, "loss": 0.1931, "step": 14009 }, { "epoch": 3.728046833422033, "grad_norm": 0.31116533279418945, "learning_rate": 3.1462752632534005e-08, "loss": 0.2173, "step": 14010 }, { "epoch": 3.728312932410857, "grad_norm": 0.28979435563087463, "learning_rate": 3.1450330013474004e-08, "loss": 0.1988, "step": 14011 }, { "epoch": 3.728579031399681, "grad_norm": 0.2855684459209442, "learning_rate": 3.143790938971691e-08, "loss": 0.1724, "step": 14012 }, { "epoch": 3.7288451303885046, "grad_norm": 0.3121791481971741, "learning_rate": 3.142549076162426e-08, "loss": 0.1782, "step": 14013 }, { "epoch": 3.7291112293773283, "grad_norm": 0.2912914752960205, "learning_rate": 3.1413074129557516e-08, "loss": 0.199, "step": 14014 }, { "epoch": 3.7293773283661524, "grad_norm": 0.3129883408546448, "learning_rate": 3.14006594938781e-08, "loss": 0.1575, "step": 14015 }, { "epoch": 3.729643427354976, "grad_norm": 0.2811844050884247, "learning_rate": 3.13882468549474e-08, "loss": 0.1945, "step": 14016 }, { "epoch": 3.7299095263438, "grad_norm": 0.28253987431526184, "learning_rate": 3.137583621312665e-08, "loss": 0.1907, "step": 14017 }, { "epoch": 3.730175625332624, "grad_norm": 0.32274729013442993, "learning_rate": 3.136342756877713e-08, "loss": 0.1906, "step": 14018 }, { "epoch": 3.7304417243214476, "grad_norm": 0.2889197766780853, "learning_rate": 3.135102092226003e-08, "loss": 0.1809, "step": 14019 }, { "epoch": 3.7307078233102713, "grad_norm": 0.30433011054992676, "learning_rate": 3.133861627393646e-08, "loss": 0.1839, "step": 14020 }, { "epoch": 3.7309739222990954, "grad_norm": 0.2828817367553711, "learning_rate": 3.1326213624167485e-08, "loss": 0.1876, "step": 14021 }, { "epoch": 3.731240021287919, "grad_norm": 0.28474321961402893, "learning_rate": 3.1313812973314136e-08, "loss": 0.171, "step": 14022 }, { "epoch": 3.7315061202767428, "grad_norm": 0.3045083284378052, "learning_rate": 3.130141432173734e-08, "loss": 0.1988, "step": 14023 }, { "epoch": 3.731772219265567, "grad_norm": 0.3727900981903076, "learning_rate": 3.128901766979803e-08, "loss": 0.1768, "step": 14024 }, { "epoch": 3.7320383182543906, "grad_norm": 0.3132655918598175, "learning_rate": 3.1276623017856975e-08, "loss": 0.1879, "step": 14025 }, { "epoch": 3.7323044172432143, "grad_norm": 0.28338348865509033, "learning_rate": 3.126423036627499e-08, "loss": 0.1851, "step": 14026 }, { "epoch": 3.7325705162320384, "grad_norm": 0.2969560921192169, "learning_rate": 3.125183971541282e-08, "loss": 0.1972, "step": 14027 }, { "epoch": 3.732836615220862, "grad_norm": 0.32632431387901306, "learning_rate": 3.1239451065631065e-08, "loss": 0.2232, "step": 14028 }, { "epoch": 3.7331027142096858, "grad_norm": 0.2953070402145386, "learning_rate": 3.122706441729035e-08, "loss": 0.1783, "step": 14029 }, { "epoch": 3.73336881319851, "grad_norm": 0.27537277340888977, "learning_rate": 3.121467977075123e-08, "loss": 0.1871, "step": 14030 }, { "epoch": 3.7336349121873336, "grad_norm": 0.2985110878944397, "learning_rate": 3.1202297126374175e-08, "loss": 0.183, "step": 14031 }, { "epoch": 3.7339010111761577, "grad_norm": 0.28726258873939514, "learning_rate": 3.118991648451962e-08, "loss": 0.1803, "step": 14032 }, { "epoch": 3.7341671101649814, "grad_norm": 0.2750134766101837, "learning_rate": 3.117753784554793e-08, "loss": 0.173, "step": 14033 }, { "epoch": 3.734433209153805, "grad_norm": 0.3281187415122986, "learning_rate": 3.1165161209819454e-08, "loss": 0.2051, "step": 14034 }, { "epoch": 3.734699308142629, "grad_norm": 0.2882572412490845, "learning_rate": 3.1152786577694376e-08, "loss": 0.18, "step": 14035 }, { "epoch": 3.734965407131453, "grad_norm": 0.28630873560905457, "learning_rate": 3.1140413949532916e-08, "loss": 0.1849, "step": 14036 }, { "epoch": 3.735231506120277, "grad_norm": 0.28396594524383545, "learning_rate": 3.112804332569522e-08, "loss": 0.1888, "step": 14037 }, { "epoch": 3.7354976051091007, "grad_norm": 0.2702459394931793, "learning_rate": 3.111567470654135e-08, "loss": 0.1671, "step": 14038 }, { "epoch": 3.7357637040979244, "grad_norm": 0.2845280170440674, "learning_rate": 3.1103308092431335e-08, "loss": 0.1987, "step": 14039 }, { "epoch": 3.7360298030867485, "grad_norm": 0.31163039803504944, "learning_rate": 3.109094348372514e-08, "loss": 0.1823, "step": 14040 }, { "epoch": 3.736295902075572, "grad_norm": 0.31088560819625854, "learning_rate": 3.107858088078266e-08, "loss": 0.1999, "step": 14041 }, { "epoch": 3.736562001064396, "grad_norm": 0.2979747951030731, "learning_rate": 3.106622028396373e-08, "loss": 0.1827, "step": 14042 }, { "epoch": 3.73682810005322, "grad_norm": 0.2885967195034027, "learning_rate": 3.105386169362818e-08, "loss": 0.1924, "step": 14043 }, { "epoch": 3.7370941990420437, "grad_norm": 0.308696448802948, "learning_rate": 3.104150511013567e-08, "loss": 0.187, "step": 14044 }, { "epoch": 3.7373602980308673, "grad_norm": 0.2870994210243225, "learning_rate": 3.1029150533845884e-08, "loss": 0.1926, "step": 14045 }, { "epoch": 3.7376263970196915, "grad_norm": 0.3051944077014923, "learning_rate": 3.101679796511849e-08, "loss": 0.197, "step": 14046 }, { "epoch": 3.737892496008515, "grad_norm": 0.29706698656082153, "learning_rate": 3.1004447404312964e-08, "loss": 0.1988, "step": 14047 }, { "epoch": 3.738158594997339, "grad_norm": 0.2963094413280487, "learning_rate": 3.0992098851788814e-08, "loss": 0.1916, "step": 14048 }, { "epoch": 3.738424693986163, "grad_norm": 0.2794313430786133, "learning_rate": 3.0979752307905506e-08, "loss": 0.1721, "step": 14049 }, { "epoch": 3.7386907929749866, "grad_norm": 0.3335626423358917, "learning_rate": 3.096740777302238e-08, "loss": 0.1824, "step": 14050 }, { "epoch": 3.7389568919638103, "grad_norm": 0.29462069272994995, "learning_rate": 3.095506524749879e-08, "loss": 0.1952, "step": 14051 }, { "epoch": 3.7392229909526344, "grad_norm": 0.3281588852405548, "learning_rate": 3.0942724731694005e-08, "loss": 0.1989, "step": 14052 }, { "epoch": 3.739489089941458, "grad_norm": 0.3135235607624054, "learning_rate": 3.0930386225967165e-08, "loss": 0.212, "step": 14053 }, { "epoch": 3.739755188930282, "grad_norm": 0.3229399025440216, "learning_rate": 3.091804973067744e-08, "loss": 0.1983, "step": 14054 }, { "epoch": 3.740021287919106, "grad_norm": 0.3204231858253479, "learning_rate": 3.0905715246183916e-08, "loss": 0.1926, "step": 14055 }, { "epoch": 3.7402873869079296, "grad_norm": 0.2829292118549347, "learning_rate": 3.089338277284562e-08, "loss": 0.1739, "step": 14056 }, { "epoch": 3.7405534858967537, "grad_norm": 0.3040929138660431, "learning_rate": 3.0881052311021525e-08, "loss": 0.1659, "step": 14057 }, { "epoch": 3.7408195848855774, "grad_norm": 0.2850804030895233, "learning_rate": 3.086872386107053e-08, "loss": 0.1897, "step": 14058 }, { "epoch": 3.7410856838744015, "grad_norm": 0.30138394236564636, "learning_rate": 3.085639742335149e-08, "loss": 0.1847, "step": 14059 }, { "epoch": 3.7413517828632252, "grad_norm": 0.3208393454551697, "learning_rate": 3.084407299822318e-08, "loss": 0.1976, "step": 14060 }, { "epoch": 3.741617881852049, "grad_norm": 0.3115290701389313, "learning_rate": 3.0831750586044356e-08, "loss": 0.1878, "step": 14061 }, { "epoch": 3.741883980840873, "grad_norm": 0.30969712138175964, "learning_rate": 3.08194301871737e-08, "loss": 0.1908, "step": 14062 }, { "epoch": 3.7421500798296967, "grad_norm": 0.30036020278930664, "learning_rate": 3.080711180196979e-08, "loss": 0.2133, "step": 14063 }, { "epoch": 3.7424161788185204, "grad_norm": 0.2757124602794647, "learning_rate": 3.079479543079118e-08, "loss": 0.1771, "step": 14064 }, { "epoch": 3.7426822778073445, "grad_norm": 0.3698836863040924, "learning_rate": 3.078248107399639e-08, "loss": 0.1926, "step": 14065 }, { "epoch": 3.742948376796168, "grad_norm": 0.33090755343437195, "learning_rate": 3.077016873194389e-08, "loss": 0.2151, "step": 14066 }, { "epoch": 3.743214475784992, "grad_norm": 0.2854974865913391, "learning_rate": 3.0757858404991986e-08, "loss": 0.1944, "step": 14067 }, { "epoch": 3.743480574773816, "grad_norm": 0.30340465903282166, "learning_rate": 3.074555009349905e-08, "loss": 0.1781, "step": 14068 }, { "epoch": 3.7437466737626397, "grad_norm": 0.29531624913215637, "learning_rate": 3.073324379782333e-08, "loss": 0.1995, "step": 14069 }, { "epoch": 3.7440127727514634, "grad_norm": 0.2847396731376648, "learning_rate": 3.072093951832303e-08, "loss": 0.1858, "step": 14070 }, { "epoch": 3.7442788717402875, "grad_norm": 0.3212289810180664, "learning_rate": 3.070863725535633e-08, "loss": 0.1974, "step": 14071 }, { "epoch": 3.744544970729111, "grad_norm": 0.28970158100128174, "learning_rate": 3.069633700928126e-08, "loss": 0.1872, "step": 14072 }, { "epoch": 3.744811069717935, "grad_norm": 0.36819592118263245, "learning_rate": 3.068403878045587e-08, "loss": 0.2052, "step": 14073 }, { "epoch": 3.745077168706759, "grad_norm": 0.30678442120552063, "learning_rate": 3.067174256923814e-08, "loss": 0.1864, "step": 14074 }, { "epoch": 3.7453432676955827, "grad_norm": 0.2917252779006958, "learning_rate": 3.065944837598596e-08, "loss": 0.1971, "step": 14075 }, { "epoch": 3.7456093666844064, "grad_norm": 0.3200603425502777, "learning_rate": 3.064715620105721e-08, "loss": 0.1963, "step": 14076 }, { "epoch": 3.7458754656732305, "grad_norm": 0.2909824550151825, "learning_rate": 3.063486604480966e-08, "loss": 0.1885, "step": 14077 }, { "epoch": 3.746141564662054, "grad_norm": 0.3118884861469269, "learning_rate": 3.062257790760107e-08, "loss": 0.1914, "step": 14078 }, { "epoch": 3.746407663650878, "grad_norm": 0.3081754148006439, "learning_rate": 3.061029178978909e-08, "loss": 0.2152, "step": 14079 }, { "epoch": 3.746673762639702, "grad_norm": 0.29026511311531067, "learning_rate": 3.0598007691731375e-08, "loss": 0.1986, "step": 14080 }, { "epoch": 3.7469398616285257, "grad_norm": 0.29383066296577454, "learning_rate": 3.058572561378543e-08, "loss": 0.1779, "step": 14081 }, { "epoch": 3.74720596061735, "grad_norm": 0.3005402684211731, "learning_rate": 3.0573445556308785e-08, "loss": 0.1934, "step": 14082 }, { "epoch": 3.7474720596061735, "grad_norm": 0.3067041039466858, "learning_rate": 3.056116751965887e-08, "loss": 0.1925, "step": 14083 }, { "epoch": 3.7477381585949976, "grad_norm": 0.3027140498161316, "learning_rate": 3.054889150419307e-08, "loss": 0.1882, "step": 14084 }, { "epoch": 3.7480042575838213, "grad_norm": 0.29757723212242126, "learning_rate": 3.053661751026871e-08, "loss": 0.1998, "step": 14085 }, { "epoch": 3.748270356572645, "grad_norm": 0.2772619128227234, "learning_rate": 3.052434553824309e-08, "loss": 0.1716, "step": 14086 }, { "epoch": 3.748536455561469, "grad_norm": 0.26297450065612793, "learning_rate": 3.051207558847335e-08, "loss": 0.1812, "step": 14087 }, { "epoch": 3.7488025545502928, "grad_norm": 0.3113076984882355, "learning_rate": 3.049980766131667e-08, "loss": 0.191, "step": 14088 }, { "epoch": 3.7490686535391164, "grad_norm": 0.3194781243801117, "learning_rate": 3.048754175713015e-08, "loss": 0.1904, "step": 14089 }, { "epoch": 3.7493347525279406, "grad_norm": 0.2805725932121277, "learning_rate": 3.0475277876270776e-08, "loss": 0.1921, "step": 14090 }, { "epoch": 3.7496008515167643, "grad_norm": 0.2849501371383667, "learning_rate": 3.0463016019095555e-08, "loss": 0.1784, "step": 14091 }, { "epoch": 3.749866950505588, "grad_norm": 0.29766666889190674, "learning_rate": 3.045075618596138e-08, "loss": 0.1899, "step": 14092 }, { "epoch": 3.750133049494412, "grad_norm": 0.32464516162872314, "learning_rate": 3.043849837722511e-08, "loss": 0.1827, "step": 14093 }, { "epoch": 3.7503991484832357, "grad_norm": 0.3248777687549591, "learning_rate": 3.0426242593243534e-08, "loss": 0.2116, "step": 14094 }, { "epoch": 3.7506652474720594, "grad_norm": 0.29851192235946655, "learning_rate": 3.041398883437339e-08, "loss": 0.1859, "step": 14095 }, { "epoch": 3.7509313464608836, "grad_norm": 0.29120394587516785, "learning_rate": 3.040173710097136e-08, "loss": 0.1945, "step": 14096 }, { "epoch": 3.7511974454497072, "grad_norm": 0.29550933837890625, "learning_rate": 3.0389487393394045e-08, "loss": 0.1963, "step": 14097 }, { "epoch": 3.751463544438531, "grad_norm": 0.3126433491706848, "learning_rate": 3.0377239711998036e-08, "loss": 0.2092, "step": 14098 }, { "epoch": 3.751729643427355, "grad_norm": 0.3064642548561096, "learning_rate": 3.036499405713978e-08, "loss": 0.2005, "step": 14099 }, { "epoch": 3.7519957424161787, "grad_norm": 0.3159639835357666, "learning_rate": 3.0352750429175746e-08, "loss": 0.1865, "step": 14100 }, { "epoch": 3.7522618414050024, "grad_norm": 0.29930979013442993, "learning_rate": 3.034050882846231e-08, "loss": 0.212, "step": 14101 }, { "epoch": 3.7525279403938265, "grad_norm": 0.2952471971511841, "learning_rate": 3.0328269255355785e-08, "loss": 0.1828, "step": 14102 }, { "epoch": 3.75279403938265, "grad_norm": 0.33203017711639404, "learning_rate": 3.0316031710212455e-08, "loss": 0.2277, "step": 14103 }, { "epoch": 3.7530601383714743, "grad_norm": 0.281292200088501, "learning_rate": 3.03037961933885e-08, "loss": 0.1801, "step": 14104 }, { "epoch": 3.753326237360298, "grad_norm": 0.2945399880409241, "learning_rate": 3.0291562705240105e-08, "loss": 0.1777, "step": 14105 }, { "epoch": 3.753592336349122, "grad_norm": 0.3031567931175232, "learning_rate": 3.027933124612328e-08, "loss": 0.1891, "step": 14106 }, { "epoch": 3.753858435337946, "grad_norm": 0.32479918003082275, "learning_rate": 3.0267101816394116e-08, "loss": 0.1983, "step": 14107 }, { "epoch": 3.7541245343267695, "grad_norm": 0.29915720224380493, "learning_rate": 3.0254874416408585e-08, "loss": 0.2032, "step": 14108 }, { "epoch": 3.7543906333155936, "grad_norm": 0.3021208941936493, "learning_rate": 3.024264904652253e-08, "loss": 0.188, "step": 14109 }, { "epoch": 3.7546567323044173, "grad_norm": 0.2908038794994354, "learning_rate": 3.023042570709184e-08, "loss": 0.1907, "step": 14110 }, { "epoch": 3.754922831293241, "grad_norm": 0.331864595413208, "learning_rate": 3.02182043984723e-08, "loss": 0.1948, "step": 14111 }, { "epoch": 3.755188930282065, "grad_norm": 0.2906266450881958, "learning_rate": 3.020598512101965e-08, "loss": 0.1798, "step": 14112 }, { "epoch": 3.755455029270889, "grad_norm": 0.30061855912208557, "learning_rate": 3.019376787508955e-08, "loss": 0.1955, "step": 14113 }, { "epoch": 3.7557211282597125, "grad_norm": 0.3162629306316376, "learning_rate": 3.018155266103762e-08, "loss": 0.1777, "step": 14114 }, { "epoch": 3.7559872272485366, "grad_norm": 0.29409846663475037, "learning_rate": 3.0169339479219414e-08, "loss": 0.208, "step": 14115 }, { "epoch": 3.7562533262373603, "grad_norm": 0.3084701597690582, "learning_rate": 3.015712832999044e-08, "loss": 0.1927, "step": 14116 }, { "epoch": 3.756519425226184, "grad_norm": 0.29408374428749084, "learning_rate": 3.01449192137061e-08, "loss": 0.2016, "step": 14117 }, { "epoch": 3.756785524215008, "grad_norm": 0.3100949823856354, "learning_rate": 3.013271213072177e-08, "loss": 0.2005, "step": 14118 }, { "epoch": 3.757051623203832, "grad_norm": 0.2584664225578308, "learning_rate": 3.01205070813928e-08, "loss": 0.1834, "step": 14119 }, { "epoch": 3.7573177221926555, "grad_norm": 0.3257899582386017, "learning_rate": 3.0108304066074405e-08, "loss": 0.2077, "step": 14120 }, { "epoch": 3.7575838211814796, "grad_norm": 0.29962751269340515, "learning_rate": 3.0096103085121825e-08, "loss": 0.1901, "step": 14121 }, { "epoch": 3.7578499201703033, "grad_norm": 0.32875221967697144, "learning_rate": 3.008390413889017e-08, "loss": 0.1953, "step": 14122 }, { "epoch": 3.758116019159127, "grad_norm": 0.2928643524646759, "learning_rate": 3.007170722773453e-08, "loss": 0.1843, "step": 14123 }, { "epoch": 3.758382118147951, "grad_norm": 0.2909708619117737, "learning_rate": 3.005951235200994e-08, "loss": 0.1874, "step": 14124 }, { "epoch": 3.7586482171367748, "grad_norm": 0.30616429448127747, "learning_rate": 3.004731951207137e-08, "loss": 0.1916, "step": 14125 }, { "epoch": 3.7589143161255985, "grad_norm": 0.27831152081489563, "learning_rate": 3.003512870827366e-08, "loss": 0.1929, "step": 14126 }, { "epoch": 3.7591804151144226, "grad_norm": 0.27515843510627747, "learning_rate": 3.00229399409717e-08, "loss": 0.1817, "step": 14127 }, { "epoch": 3.7594465141032463, "grad_norm": 0.307223379611969, "learning_rate": 3.001075321052029e-08, "loss": 0.1961, "step": 14128 }, { "epoch": 3.7597126130920704, "grad_norm": 0.29579958319664, "learning_rate": 2.9998568517274105e-08, "loss": 0.2043, "step": 14129 }, { "epoch": 3.759978712080894, "grad_norm": 0.30103299021720886, "learning_rate": 2.998638586158783e-08, "loss": 0.1817, "step": 14130 }, { "epoch": 3.760244811069718, "grad_norm": 0.2983150780200958, "learning_rate": 2.9974205243816075e-08, "loss": 0.2042, "step": 14131 }, { "epoch": 3.760510910058542, "grad_norm": 0.3111875355243683, "learning_rate": 2.996202666431338e-08, "loss": 0.1921, "step": 14132 }, { "epoch": 3.7607770090473656, "grad_norm": 0.2832881212234497, "learning_rate": 2.994985012343424e-08, "loss": 0.1905, "step": 14133 }, { "epoch": 3.7610431080361897, "grad_norm": 0.3439105451107025, "learning_rate": 2.993767562153311e-08, "loss": 0.194, "step": 14134 }, { "epoch": 3.7613092070250134, "grad_norm": 0.29096493124961853, "learning_rate": 2.992550315896429e-08, "loss": 0.188, "step": 14135 }, { "epoch": 3.761575306013837, "grad_norm": 0.2853882312774658, "learning_rate": 2.9913332736082144e-08, "loss": 0.1864, "step": 14136 }, { "epoch": 3.761841405002661, "grad_norm": 0.27142783999443054, "learning_rate": 2.990116435324089e-08, "loss": 0.1749, "step": 14137 }, { "epoch": 3.762107503991485, "grad_norm": 0.3108504116535187, "learning_rate": 2.988899801079474e-08, "loss": 0.1935, "step": 14138 }, { "epoch": 3.7623736029803085, "grad_norm": 0.3109520971775055, "learning_rate": 2.9876833709097814e-08, "loss": 0.1987, "step": 14139 }, { "epoch": 3.7626397019691327, "grad_norm": 0.29713550209999084, "learning_rate": 2.986467144850418e-08, "loss": 0.1909, "step": 14140 }, { "epoch": 3.7629058009579563, "grad_norm": 0.324602872133255, "learning_rate": 2.985251122936786e-08, "loss": 0.2103, "step": 14141 }, { "epoch": 3.76317189994678, "grad_norm": 0.286495566368103, "learning_rate": 2.98403530520428e-08, "loss": 0.1757, "step": 14142 }, { "epoch": 3.763437998935604, "grad_norm": 0.2937861680984497, "learning_rate": 2.982819691688293e-08, "loss": 0.1725, "step": 14143 }, { "epoch": 3.763704097924428, "grad_norm": 0.2827739715576172, "learning_rate": 2.981604282424202e-08, "loss": 0.1955, "step": 14144 }, { "epoch": 3.7639701969132515, "grad_norm": 0.3174063265323639, "learning_rate": 2.980389077447386e-08, "loss": 0.1959, "step": 14145 }, { "epoch": 3.7642362959020756, "grad_norm": 0.3102153539657593, "learning_rate": 2.9791740767932183e-08, "loss": 0.1861, "step": 14146 }, { "epoch": 3.7645023948908993, "grad_norm": 0.29116490483283997, "learning_rate": 2.9779592804970676e-08, "loss": 0.191, "step": 14147 }, { "epoch": 3.764768493879723, "grad_norm": 0.29094603657722473, "learning_rate": 2.976744688594286e-08, "loss": 0.1932, "step": 14148 }, { "epoch": 3.765034592868547, "grad_norm": 0.28680387139320374, "learning_rate": 2.9755303011202317e-08, "loss": 0.1922, "step": 14149 }, { "epoch": 3.765300691857371, "grad_norm": 0.2858532667160034, "learning_rate": 2.9743161181102505e-08, "loss": 0.1789, "step": 14150 }, { "epoch": 3.765566790846195, "grad_norm": 0.30277127027511597, "learning_rate": 2.9731021395996846e-08, "loss": 0.1787, "step": 14151 }, { "epoch": 3.7658328898350186, "grad_norm": 0.2845812737941742, "learning_rate": 2.9718883656238748e-08, "loss": 0.1842, "step": 14152 }, { "epoch": 3.7660989888238423, "grad_norm": 0.30914202332496643, "learning_rate": 2.9706747962181423e-08, "loss": 0.197, "step": 14153 }, { "epoch": 3.7663650878126664, "grad_norm": 0.3218402862548828, "learning_rate": 2.9694614314178157e-08, "loss": 0.2023, "step": 14154 }, { "epoch": 3.76663118680149, "grad_norm": 0.27209994196891785, "learning_rate": 2.9682482712582113e-08, "loss": 0.1831, "step": 14155 }, { "epoch": 3.7668972857903142, "grad_norm": 0.28778520226478577, "learning_rate": 2.967035315774642e-08, "loss": 0.1864, "step": 14156 }, { "epoch": 3.767163384779138, "grad_norm": 0.286777526140213, "learning_rate": 2.965822565002415e-08, "loss": 0.1786, "step": 14157 }, { "epoch": 3.7674294837679616, "grad_norm": 0.28304973244667053, "learning_rate": 2.9646100189768274e-08, "loss": 0.1787, "step": 14158 }, { "epoch": 3.7676955827567857, "grad_norm": 0.3077149987220764, "learning_rate": 2.963397677733176e-08, "loss": 0.1999, "step": 14159 }, { "epoch": 3.7679616817456094, "grad_norm": 0.29985082149505615, "learning_rate": 2.9621855413067477e-08, "loss": 0.1939, "step": 14160 }, { "epoch": 3.768227780734433, "grad_norm": 0.3211734890937805, "learning_rate": 2.9609736097328285e-08, "loss": 0.2069, "step": 14161 }, { "epoch": 3.7684938797232572, "grad_norm": 0.2649616301059723, "learning_rate": 2.9597618830466864e-08, "loss": 0.1792, "step": 14162 }, { "epoch": 3.768759978712081, "grad_norm": 0.28480812907218933, "learning_rate": 2.9585503612835972e-08, "loss": 0.1986, "step": 14163 }, { "epoch": 3.7690260777009046, "grad_norm": 0.31690070033073425, "learning_rate": 2.957339044478824e-08, "loss": 0.2012, "step": 14164 }, { "epoch": 3.7692921766897287, "grad_norm": 0.2900731861591339, "learning_rate": 2.9561279326676248e-08, "loss": 0.1858, "step": 14165 }, { "epoch": 3.7695582756785524, "grad_norm": 0.3123054802417755, "learning_rate": 2.9549170258852518e-08, "loss": 0.1941, "step": 14166 }, { "epoch": 3.769824374667376, "grad_norm": 0.321776807308197, "learning_rate": 2.9537063241669556e-08, "loss": 0.2095, "step": 14167 }, { "epoch": 3.7700904736562, "grad_norm": 0.30385822057724, "learning_rate": 2.9524958275479693e-08, "loss": 0.1892, "step": 14168 }, { "epoch": 3.770356572645024, "grad_norm": 0.3101469874382019, "learning_rate": 2.9512855360635313e-08, "loss": 0.1856, "step": 14169 }, { "epoch": 3.7706226716338476, "grad_norm": 0.2779906690120697, "learning_rate": 2.9500754497488723e-08, "loss": 0.1863, "step": 14170 }, { "epoch": 3.7708887706226717, "grad_norm": 0.30997416377067566, "learning_rate": 2.9488655686392083e-08, "loss": 0.1871, "step": 14171 }, { "epoch": 3.7711548696114954, "grad_norm": 0.29309016466140747, "learning_rate": 2.9476558927697603e-08, "loss": 0.1851, "step": 14172 }, { "epoch": 3.771420968600319, "grad_norm": 0.2926582992076874, "learning_rate": 2.946446422175738e-08, "loss": 0.1896, "step": 14173 }, { "epoch": 3.771687067589143, "grad_norm": 0.31067970395088196, "learning_rate": 2.945237156892345e-08, "loss": 0.1847, "step": 14174 }, { "epoch": 3.771953166577967, "grad_norm": 0.29342177510261536, "learning_rate": 2.9440280969547827e-08, "loss": 0.2064, "step": 14175 }, { "epoch": 3.772219265566791, "grad_norm": 0.3084103465080261, "learning_rate": 2.94281924239824e-08, "loss": 0.1803, "step": 14176 }, { "epoch": 3.7724853645556147, "grad_norm": 0.315878689289093, "learning_rate": 2.941610593257906e-08, "loss": 0.1985, "step": 14177 }, { "epoch": 3.772751463544439, "grad_norm": 0.30284035205841064, "learning_rate": 2.9404021495689613e-08, "loss": 0.1903, "step": 14178 }, { "epoch": 3.7730175625332625, "grad_norm": 0.29451537132263184, "learning_rate": 2.9391939113665833e-08, "loss": 0.1814, "step": 14179 }, { "epoch": 3.773283661522086, "grad_norm": 0.2864414155483246, "learning_rate": 2.9379858786859336e-08, "loss": 0.2067, "step": 14180 }, { "epoch": 3.7735497605109103, "grad_norm": 0.2970399260520935, "learning_rate": 2.936778051562179e-08, "loss": 0.1901, "step": 14181 }, { "epoch": 3.773815859499734, "grad_norm": 0.29909345507621765, "learning_rate": 2.935570430030475e-08, "loss": 0.2145, "step": 14182 }, { "epoch": 3.7740819584885577, "grad_norm": 0.2841445803642273, "learning_rate": 2.9343630141259733e-08, "loss": 0.1776, "step": 14183 }, { "epoch": 3.774348057477382, "grad_norm": 0.2926355004310608, "learning_rate": 2.933155803883819e-08, "loss": 0.2007, "step": 14184 }, { "epoch": 3.7746141564662055, "grad_norm": 0.3030164837837219, "learning_rate": 2.93194879933915e-08, "loss": 0.1917, "step": 14185 }, { "epoch": 3.774880255455029, "grad_norm": 0.30156615376472473, "learning_rate": 2.9307420005271e-08, "loss": 0.1841, "step": 14186 }, { "epoch": 3.7751463544438533, "grad_norm": 0.28330016136169434, "learning_rate": 2.929535407482798e-08, "loss": 0.1783, "step": 14187 }, { "epoch": 3.775412453432677, "grad_norm": 0.30250221490859985, "learning_rate": 2.9283290202413592e-08, "loss": 0.1866, "step": 14188 }, { "epoch": 3.7756785524215006, "grad_norm": 0.28186336159706116, "learning_rate": 2.9271228388379033e-08, "loss": 0.1854, "step": 14189 }, { "epoch": 3.7759446514103248, "grad_norm": 0.28026363253593445, "learning_rate": 2.9259168633075348e-08, "loss": 0.1905, "step": 14190 }, { "epoch": 3.7762107503991484, "grad_norm": 0.26326707005500793, "learning_rate": 2.924711093685359e-08, "loss": 0.1794, "step": 14191 }, { "epoch": 3.776476849387972, "grad_norm": 0.30655983090400696, "learning_rate": 2.9235055300064715e-08, "loss": 0.207, "step": 14192 }, { "epoch": 3.7767429483767962, "grad_norm": 0.3045268654823303, "learning_rate": 2.9223001723059647e-08, "loss": 0.1878, "step": 14193 }, { "epoch": 3.77700904736562, "grad_norm": 0.3039284348487854, "learning_rate": 2.9210950206189233e-08, "loss": 0.1884, "step": 14194 }, { "epoch": 3.7772751463544436, "grad_norm": 0.27418750524520874, "learning_rate": 2.9198900749804246e-08, "loss": 0.1636, "step": 14195 }, { "epoch": 3.7775412453432677, "grad_norm": 0.3266017436981201, "learning_rate": 2.9186853354255438e-08, "loss": 0.2099, "step": 14196 }, { "epoch": 3.7778073443320914, "grad_norm": 0.30133330821990967, "learning_rate": 2.9174808019893482e-08, "loss": 0.171, "step": 14197 }, { "epoch": 3.778073443320915, "grad_norm": 0.2923209071159363, "learning_rate": 2.9162764747068946e-08, "loss": 0.1937, "step": 14198 }, { "epoch": 3.7783395423097392, "grad_norm": 0.33505919575691223, "learning_rate": 2.9150723536132394e-08, "loss": 0.2113, "step": 14199 }, { "epoch": 3.778605641298563, "grad_norm": 0.3138546049594879, "learning_rate": 2.913868438743433e-08, "loss": 0.2077, "step": 14200 }, { "epoch": 3.778871740287387, "grad_norm": 0.311296671628952, "learning_rate": 2.9126647301325168e-08, "loss": 0.1878, "step": 14201 }, { "epoch": 3.7791378392762107, "grad_norm": 0.2931726276874542, "learning_rate": 2.9114612278155282e-08, "loss": 0.1864, "step": 14202 }, { "epoch": 3.779403938265035, "grad_norm": 0.28799617290496826, "learning_rate": 2.910257931827499e-08, "loss": 0.1831, "step": 14203 }, { "epoch": 3.7796700372538585, "grad_norm": 0.2829943001270294, "learning_rate": 2.9090548422034524e-08, "loss": 0.1803, "step": 14204 }, { "epoch": 3.779936136242682, "grad_norm": 0.3168276846408844, "learning_rate": 2.907851958978408e-08, "loss": 0.201, "step": 14205 }, { "epoch": 3.7802022352315063, "grad_norm": 0.3025496006011963, "learning_rate": 2.906649282187381e-08, "loss": 0.1994, "step": 14206 }, { "epoch": 3.78046833422033, "grad_norm": 0.30964407324790955, "learning_rate": 2.905446811865373e-08, "loss": 0.2082, "step": 14207 }, { "epoch": 3.7807344332091537, "grad_norm": 0.3064385950565338, "learning_rate": 2.9042445480473877e-08, "loss": 0.1921, "step": 14208 }, { "epoch": 3.781000532197978, "grad_norm": 0.28599387407302856, "learning_rate": 2.9030424907684224e-08, "loss": 0.1765, "step": 14209 }, { "epoch": 3.7812666311868015, "grad_norm": 0.3106589913368225, "learning_rate": 2.90184064006346e-08, "loss": 0.2174, "step": 14210 }, { "epoch": 3.781532730175625, "grad_norm": 0.29312556982040405, "learning_rate": 2.900638995967487e-08, "loss": 0.1993, "step": 14211 }, { "epoch": 3.7817988291644493, "grad_norm": 0.31339704990386963, "learning_rate": 2.8994375585154783e-08, "loss": 0.2079, "step": 14212 }, { "epoch": 3.782064928153273, "grad_norm": 0.28690388798713684, "learning_rate": 2.898236327742407e-08, "loss": 0.1831, "step": 14213 }, { "epoch": 3.7823310271420967, "grad_norm": 0.2946401834487915, "learning_rate": 2.897035303683236e-08, "loss": 0.189, "step": 14214 }, { "epoch": 3.782597126130921, "grad_norm": 0.29246786236763, "learning_rate": 2.895834486372929e-08, "loss": 0.2007, "step": 14215 }, { "epoch": 3.7828632251197445, "grad_norm": 0.3137560188770294, "learning_rate": 2.8946338758464296e-08, "loss": 0.1947, "step": 14216 }, { "epoch": 3.783129324108568, "grad_norm": 0.3134978115558624, "learning_rate": 2.89343347213869e-08, "loss": 0.205, "step": 14217 }, { "epoch": 3.7833954230973923, "grad_norm": 0.31173110008239746, "learning_rate": 2.8922332752846502e-08, "loss": 0.2, "step": 14218 }, { "epoch": 3.783661522086216, "grad_norm": 0.29386910796165466, "learning_rate": 2.8910332853192444e-08, "loss": 0.1855, "step": 14219 }, { "epoch": 3.7839276210750397, "grad_norm": 0.27430814504623413, "learning_rate": 2.8898335022774012e-08, "loss": 0.1729, "step": 14220 }, { "epoch": 3.784193720063864, "grad_norm": 0.31935185194015503, "learning_rate": 2.888633926194043e-08, "loss": 0.2099, "step": 14221 }, { "epoch": 3.7844598190526875, "grad_norm": 0.29374590516090393, "learning_rate": 2.887434557104087e-08, "loss": 0.1974, "step": 14222 }, { "epoch": 3.7847259180415116, "grad_norm": 0.30521464347839355, "learning_rate": 2.8862353950424446e-08, "loss": 0.1896, "step": 14223 }, { "epoch": 3.7849920170303353, "grad_norm": 0.3080472946166992, "learning_rate": 2.885036440044021e-08, "loss": 0.1994, "step": 14224 }, { "epoch": 3.7852581160191594, "grad_norm": 0.3088749945163727, "learning_rate": 2.8838376921437103e-08, "loss": 0.1839, "step": 14225 }, { "epoch": 3.785524215007983, "grad_norm": 0.30544552206993103, "learning_rate": 2.882639151376407e-08, "loss": 0.191, "step": 14226 }, { "epoch": 3.7857903139968068, "grad_norm": 0.2949149012565613, "learning_rate": 2.8814408177769978e-08, "loss": 0.1775, "step": 14227 }, { "epoch": 3.786056412985631, "grad_norm": 0.3012129068374634, "learning_rate": 2.8802426913803634e-08, "loss": 0.2044, "step": 14228 }, { "epoch": 3.7863225119744546, "grad_norm": 0.29688045382499695, "learning_rate": 2.8790447722213817e-08, "loss": 0.1923, "step": 14229 }, { "epoch": 3.7865886109632783, "grad_norm": 0.3055819571018219, "learning_rate": 2.8778470603349138e-08, "loss": 0.1918, "step": 14230 }, { "epoch": 3.7868547099521024, "grad_norm": 0.267720103263855, "learning_rate": 2.8766495557558256e-08, "loss": 0.1534, "step": 14231 }, { "epoch": 3.787120808940926, "grad_norm": 0.30546897649765015, "learning_rate": 2.8754522585189733e-08, "loss": 0.1954, "step": 14232 }, { "epoch": 3.7873869079297497, "grad_norm": 0.2972720265388489, "learning_rate": 2.87425516865921e-08, "loss": 0.1766, "step": 14233 }, { "epoch": 3.787653006918574, "grad_norm": 0.3231582045555115, "learning_rate": 2.873058286211374e-08, "loss": 0.1915, "step": 14234 }, { "epoch": 3.7879191059073976, "grad_norm": 0.2977791428565979, "learning_rate": 2.8718616112103054e-08, "loss": 0.1841, "step": 14235 }, { "epoch": 3.7881852048962212, "grad_norm": 0.29299038648605347, "learning_rate": 2.870665143690838e-08, "loss": 0.1751, "step": 14236 }, { "epoch": 3.7884513038850454, "grad_norm": 0.3162793517112732, "learning_rate": 2.8694688836877978e-08, "loss": 0.1751, "step": 14237 }, { "epoch": 3.788717402873869, "grad_norm": 0.2960895001888275, "learning_rate": 2.8682728312360038e-08, "loss": 0.1934, "step": 14238 }, { "epoch": 3.7889835018626927, "grad_norm": 0.33249223232269287, "learning_rate": 2.8670769863702693e-08, "loss": 0.1996, "step": 14239 }, { "epoch": 3.789249600851517, "grad_norm": 0.2918364107608795, "learning_rate": 2.865881349125404e-08, "loss": 0.1982, "step": 14240 }, { "epoch": 3.7895156998403405, "grad_norm": 0.3078337609767914, "learning_rate": 2.8646859195362094e-08, "loss": 0.1979, "step": 14241 }, { "epoch": 3.789781798829164, "grad_norm": 0.25461456179618835, "learning_rate": 2.863490697637484e-08, "loss": 0.1818, "step": 14242 }, { "epoch": 3.7900478978179883, "grad_norm": 0.2847828269004822, "learning_rate": 2.862295683464011e-08, "loss": 0.1825, "step": 14243 }, { "epoch": 3.790313996806812, "grad_norm": 0.2905771732330322, "learning_rate": 2.8611008770505783e-08, "loss": 0.1896, "step": 14244 }, { "epoch": 3.7905800957956357, "grad_norm": 0.3021577000617981, "learning_rate": 2.8599062784319627e-08, "loss": 0.1815, "step": 14245 }, { "epoch": 3.79084619478446, "grad_norm": 0.28217369318008423, "learning_rate": 2.8587118876429372e-08, "loss": 0.1774, "step": 14246 }, { "epoch": 3.7911122937732835, "grad_norm": 0.2615877389907837, "learning_rate": 2.8575177047182653e-08, "loss": 0.1614, "step": 14247 }, { "epoch": 3.7913783927621076, "grad_norm": 0.2883433401584625, "learning_rate": 2.8563237296927124e-08, "loss": 0.1976, "step": 14248 }, { "epoch": 3.7916444917509313, "grad_norm": 0.31322526931762695, "learning_rate": 2.855129962601023e-08, "loss": 0.1965, "step": 14249 }, { "epoch": 3.7919105907397554, "grad_norm": 0.2927665710449219, "learning_rate": 2.853936403477949e-08, "loss": 0.187, "step": 14250 }, { "epoch": 3.792176689728579, "grad_norm": 0.2863328158855438, "learning_rate": 2.852743052358235e-08, "loss": 0.1815, "step": 14251 }, { "epoch": 3.792442788717403, "grad_norm": 0.2988898456096649, "learning_rate": 2.8515499092766093e-08, "loss": 0.1894, "step": 14252 }, { "epoch": 3.792708887706227, "grad_norm": 0.31021177768707275, "learning_rate": 2.8503569742678058e-08, "loss": 0.1823, "step": 14253 }, { "epoch": 3.7929749866950506, "grad_norm": 0.29070422053337097, "learning_rate": 2.8491642473665466e-08, "loss": 0.1756, "step": 14254 }, { "epoch": 3.7932410856838743, "grad_norm": 0.28442931175231934, "learning_rate": 2.8479717286075498e-08, "loss": 0.1966, "step": 14255 }, { "epoch": 3.7935071846726984, "grad_norm": 0.2834007441997528, "learning_rate": 2.8467794180255256e-08, "loss": 0.1739, "step": 14256 }, { "epoch": 3.793773283661522, "grad_norm": 0.31258997321128845, "learning_rate": 2.8455873156551792e-08, "loss": 0.2034, "step": 14257 }, { "epoch": 3.794039382650346, "grad_norm": 0.2894824743270874, "learning_rate": 2.8443954215312104e-08, "loss": 0.1857, "step": 14258 }, { "epoch": 3.79430548163917, "grad_norm": 0.3067682683467865, "learning_rate": 2.843203735688311e-08, "loss": 0.1886, "step": 14259 }, { "epoch": 3.7945715806279936, "grad_norm": 0.2936301529407501, "learning_rate": 2.8420122581611717e-08, "loss": 0.1798, "step": 14260 }, { "epoch": 3.7948376796168173, "grad_norm": 0.28004518151283264, "learning_rate": 2.840820988984466e-08, "loss": 0.1787, "step": 14261 }, { "epoch": 3.7951037786056414, "grad_norm": 0.31050586700439453, "learning_rate": 2.8396299281928737e-08, "loss": 0.2073, "step": 14262 }, { "epoch": 3.795369877594465, "grad_norm": 0.3052668571472168, "learning_rate": 2.838439075821062e-08, "loss": 0.2048, "step": 14263 }, { "epoch": 3.7956359765832888, "grad_norm": 0.33365780115127563, "learning_rate": 2.8372484319036948e-08, "loss": 0.2065, "step": 14264 }, { "epoch": 3.795902075572113, "grad_norm": 0.35078293085098267, "learning_rate": 2.836057996475427e-08, "loss": 0.1967, "step": 14265 }, { "epoch": 3.7961681745609366, "grad_norm": 0.2957919239997864, "learning_rate": 2.83486776957091e-08, "loss": 0.1826, "step": 14266 }, { "epoch": 3.7964342735497603, "grad_norm": 0.3191911280155182, "learning_rate": 2.833677751224789e-08, "loss": 0.2018, "step": 14267 }, { "epoch": 3.7967003725385844, "grad_norm": 0.2939029335975647, "learning_rate": 2.832487941471704e-08, "loss": 0.1983, "step": 14268 }, { "epoch": 3.796966471527408, "grad_norm": 0.3121776580810547, "learning_rate": 2.8312983403462808e-08, "loss": 0.2014, "step": 14269 }, { "epoch": 3.797232570516232, "grad_norm": 0.27582278847694397, "learning_rate": 2.830108947883151e-08, "loss": 0.1679, "step": 14270 }, { "epoch": 3.797498669505056, "grad_norm": 0.31622573733329773, "learning_rate": 2.828919764116936e-08, "loss": 0.1989, "step": 14271 }, { "epoch": 3.7977647684938796, "grad_norm": 0.30041930079460144, "learning_rate": 2.8277307890822434e-08, "loss": 0.1966, "step": 14272 }, { "epoch": 3.7980308674827037, "grad_norm": 0.30254605412483215, "learning_rate": 2.826542022813685e-08, "loss": 0.1939, "step": 14273 }, { "epoch": 3.7982969664715274, "grad_norm": 0.3072763979434967, "learning_rate": 2.825353465345863e-08, "loss": 0.1948, "step": 14274 }, { "epoch": 3.7985630654603515, "grad_norm": 0.3104378283023834, "learning_rate": 2.824165116713373e-08, "loss": 0.2161, "step": 14275 }, { "epoch": 3.798829164449175, "grad_norm": 0.2941516041755676, "learning_rate": 2.822976976950805e-08, "loss": 0.1924, "step": 14276 }, { "epoch": 3.799095263437999, "grad_norm": 0.3127792775630951, "learning_rate": 2.8217890460927417e-08, "loss": 0.2016, "step": 14277 }, { "epoch": 3.799361362426823, "grad_norm": 0.29913607239723206, "learning_rate": 2.8206013241737637e-08, "loss": 0.2067, "step": 14278 }, { "epoch": 3.7996274614156467, "grad_norm": 0.3151604235172272, "learning_rate": 2.819413811228437e-08, "loss": 0.2141, "step": 14279 }, { "epoch": 3.7998935604044703, "grad_norm": 0.2893293499946594, "learning_rate": 2.8182265072913313e-08, "loss": 0.1962, "step": 14280 }, { "epoch": 3.8001596593932945, "grad_norm": 0.3156295120716095, "learning_rate": 2.817039412397004e-08, "loss": 0.2006, "step": 14281 }, { "epoch": 3.800425758382118, "grad_norm": 0.3132782280445099, "learning_rate": 2.8158525265800092e-08, "loss": 0.1862, "step": 14282 }, { "epoch": 3.800691857370942, "grad_norm": 0.30052080750465393, "learning_rate": 2.814665849874893e-08, "loss": 0.1866, "step": 14283 }, { "epoch": 3.800957956359766, "grad_norm": 0.2817952334880829, "learning_rate": 2.8134793823161984e-08, "loss": 0.1762, "step": 14284 }, { "epoch": 3.8012240553485896, "grad_norm": 0.3115536868572235, "learning_rate": 2.8122931239384595e-08, "loss": 0.1958, "step": 14285 }, { "epoch": 3.8014901543374133, "grad_norm": 0.2939499020576477, "learning_rate": 2.8111070747762044e-08, "loss": 0.1884, "step": 14286 }, { "epoch": 3.8017562533262375, "grad_norm": 0.289655864238739, "learning_rate": 2.8099212348639602e-08, "loss": 0.1733, "step": 14287 }, { "epoch": 3.802022352315061, "grad_norm": 0.2953912615776062, "learning_rate": 2.8087356042362364e-08, "loss": 0.1837, "step": 14288 }, { "epoch": 3.802288451303885, "grad_norm": 0.3055194616317749, "learning_rate": 2.8075501829275473e-08, "loss": 0.1881, "step": 14289 }, { "epoch": 3.802554550292709, "grad_norm": 0.32255810499191284, "learning_rate": 2.8063649709724003e-08, "loss": 0.2024, "step": 14290 }, { "epoch": 3.8028206492815326, "grad_norm": 0.3274955749511719, "learning_rate": 2.805179968405288e-08, "loss": 0.2073, "step": 14291 }, { "epoch": 3.8030867482703563, "grad_norm": 0.2807607650756836, "learning_rate": 2.803995175260706e-08, "loss": 0.1982, "step": 14292 }, { "epoch": 3.8033528472591804, "grad_norm": 0.32426077127456665, "learning_rate": 2.8028105915731394e-08, "loss": 0.2071, "step": 14293 }, { "epoch": 3.803618946248004, "grad_norm": 0.284660667181015, "learning_rate": 2.80162621737707e-08, "loss": 0.1652, "step": 14294 }, { "epoch": 3.8038850452368282, "grad_norm": 0.2970264256000519, "learning_rate": 2.8004420527069704e-08, "loss": 0.1895, "step": 14295 }, { "epoch": 3.804151144225652, "grad_norm": 0.31311753392219543, "learning_rate": 2.7992580975973136e-08, "loss": 0.1836, "step": 14296 }, { "epoch": 3.804417243214476, "grad_norm": 0.2941378951072693, "learning_rate": 2.7980743520825523e-08, "loss": 0.1852, "step": 14297 }, { "epoch": 3.8046833422032997, "grad_norm": 0.3159511983394623, "learning_rate": 2.7968908161971484e-08, "loss": 0.2166, "step": 14298 }, { "epoch": 3.8049494411921234, "grad_norm": 0.3016628623008728, "learning_rate": 2.7957074899755485e-08, "loss": 0.1928, "step": 14299 }, { "epoch": 3.8052155401809475, "grad_norm": 0.2760850489139557, "learning_rate": 2.7945243734521995e-08, "loss": 0.1898, "step": 14300 }, { "epoch": 3.8054816391697712, "grad_norm": 0.3098033666610718, "learning_rate": 2.7933414666615362e-08, "loss": 0.1926, "step": 14301 }, { "epoch": 3.805747738158595, "grad_norm": 0.2932116389274597, "learning_rate": 2.792158769637991e-08, "loss": 0.1808, "step": 14302 }, { "epoch": 3.806013837147419, "grad_norm": 0.299224317073822, "learning_rate": 2.7909762824159888e-08, "loss": 0.2038, "step": 14303 }, { "epoch": 3.8062799361362427, "grad_norm": 0.30435946583747864, "learning_rate": 2.7897940050299497e-08, "loss": 0.206, "step": 14304 }, { "epoch": 3.8065460351250664, "grad_norm": 0.2897014915943146, "learning_rate": 2.788611937514288e-08, "loss": 0.1795, "step": 14305 }, { "epoch": 3.8068121341138905, "grad_norm": 0.2990495562553406, "learning_rate": 2.7874300799034067e-08, "loss": 0.205, "step": 14306 }, { "epoch": 3.807078233102714, "grad_norm": 0.345167338848114, "learning_rate": 2.786248432231708e-08, "loss": 0.178, "step": 14307 }, { "epoch": 3.807344332091538, "grad_norm": 0.28618520498275757, "learning_rate": 2.7850669945335868e-08, "loss": 0.1797, "step": 14308 }, { "epoch": 3.807610431080362, "grad_norm": 0.26779741048812866, "learning_rate": 2.7838857668434323e-08, "loss": 0.1731, "step": 14309 }, { "epoch": 3.8078765300691857, "grad_norm": 0.26528576016426086, "learning_rate": 2.7827047491956302e-08, "loss": 0.1765, "step": 14310 }, { "epoch": 3.8081426290580094, "grad_norm": 0.2784794867038727, "learning_rate": 2.7815239416245508e-08, "loss": 0.1723, "step": 14311 }, { "epoch": 3.8084087280468335, "grad_norm": 0.3353399634361267, "learning_rate": 2.7803433441645652e-08, "loss": 0.1936, "step": 14312 }, { "epoch": 3.808674827035657, "grad_norm": 0.3214680254459381, "learning_rate": 2.7791629568500396e-08, "loss": 0.2028, "step": 14313 }, { "epoch": 3.808940926024481, "grad_norm": 0.27902674674987793, "learning_rate": 2.7779827797153353e-08, "loss": 0.1797, "step": 14314 }, { "epoch": 3.809207025013305, "grad_norm": 0.43413683772087097, "learning_rate": 2.7768028127947973e-08, "loss": 0.1907, "step": 14315 }, { "epoch": 3.8094731240021287, "grad_norm": 0.28458791971206665, "learning_rate": 2.7756230561227746e-08, "loss": 0.1855, "step": 14316 }, { "epoch": 3.809739222990953, "grad_norm": 0.2926250994205475, "learning_rate": 2.7744435097336073e-08, "loss": 0.1929, "step": 14317 }, { "epoch": 3.8100053219797765, "grad_norm": 0.2760438024997711, "learning_rate": 2.773264173661627e-08, "loss": 0.1756, "step": 14318 }, { "epoch": 3.8102714209686, "grad_norm": 0.2938324511051178, "learning_rate": 2.772085047941164e-08, "loss": 0.1895, "step": 14319 }, { "epoch": 3.8105375199574243, "grad_norm": 0.2950129210948944, "learning_rate": 2.7709061326065376e-08, "loss": 0.1958, "step": 14320 }, { "epoch": 3.810803618946248, "grad_norm": 0.29290294647216797, "learning_rate": 2.7697274276920635e-08, "loss": 0.1799, "step": 14321 }, { "epoch": 3.811069717935072, "grad_norm": 0.2985144257545471, "learning_rate": 2.76854893323205e-08, "loss": 0.1891, "step": 14322 }, { "epoch": 3.811335816923896, "grad_norm": 0.27732259035110474, "learning_rate": 2.7673706492608017e-08, "loss": 0.1772, "step": 14323 }, { "epoch": 3.8116019159127195, "grad_norm": 0.3052307963371277, "learning_rate": 2.766192575812618e-08, "loss": 0.2032, "step": 14324 }, { "epoch": 3.8118680149015436, "grad_norm": 0.2734672725200653, "learning_rate": 2.7650147129217814e-08, "loss": 0.1852, "step": 14325 }, { "epoch": 3.8121341138903673, "grad_norm": 0.3191491365432739, "learning_rate": 2.7638370606225826e-08, "loss": 0.1821, "step": 14326 }, { "epoch": 3.812400212879191, "grad_norm": 0.30018770694732666, "learning_rate": 2.762659618949298e-08, "loss": 0.1882, "step": 14327 }, { "epoch": 3.812666311868015, "grad_norm": 0.29108938574790955, "learning_rate": 2.7614823879362004e-08, "loss": 0.1961, "step": 14328 }, { "epoch": 3.8129324108568388, "grad_norm": 0.304357647895813, "learning_rate": 2.760305367617556e-08, "loss": 0.1837, "step": 14329 }, { "epoch": 3.8131985098456624, "grad_norm": 0.306894451379776, "learning_rate": 2.759128558027627e-08, "loss": 0.2098, "step": 14330 }, { "epoch": 3.8134646088344866, "grad_norm": 0.28747692704200745, "learning_rate": 2.7579519592006618e-08, "loss": 0.1728, "step": 14331 }, { "epoch": 3.8137307078233103, "grad_norm": 0.29949450492858887, "learning_rate": 2.7567755711709118e-08, "loss": 0.1993, "step": 14332 }, { "epoch": 3.813996806812134, "grad_norm": 0.34719887375831604, "learning_rate": 2.7555993939726218e-08, "loss": 0.1872, "step": 14333 }, { "epoch": 3.814262905800958, "grad_norm": 0.29710638523101807, "learning_rate": 2.754423427640019e-08, "loss": 0.1917, "step": 14334 }, { "epoch": 3.8145290047897817, "grad_norm": 0.3017393946647644, "learning_rate": 2.7532476722073384e-08, "loss": 0.1844, "step": 14335 }, { "epoch": 3.8147951037786054, "grad_norm": 0.2871372401714325, "learning_rate": 2.752072127708802e-08, "loss": 0.1932, "step": 14336 }, { "epoch": 3.8150612027674295, "grad_norm": 0.3308677673339844, "learning_rate": 2.7508967941786266e-08, "loss": 0.1847, "step": 14337 }, { "epoch": 3.8153273017562532, "grad_norm": 0.3091277480125427, "learning_rate": 2.749721671651023e-08, "loss": 0.1813, "step": 14338 }, { "epoch": 3.815593400745077, "grad_norm": 0.3061408996582031, "learning_rate": 2.7485467601601975e-08, "loss": 0.2095, "step": 14339 }, { "epoch": 3.815859499733901, "grad_norm": 0.35292860865592957, "learning_rate": 2.747372059740346e-08, "loss": 0.1955, "step": 14340 }, { "epoch": 3.8161255987227247, "grad_norm": 0.29936909675598145, "learning_rate": 2.7461975704256634e-08, "loss": 0.2001, "step": 14341 }, { "epoch": 3.816391697711549, "grad_norm": 0.2818065285682678, "learning_rate": 2.745023292250339e-08, "loss": 0.1877, "step": 14342 }, { "epoch": 3.8166577967003725, "grad_norm": 0.36351123452186584, "learning_rate": 2.7438492252485456e-08, "loss": 0.195, "step": 14343 }, { "epoch": 3.8169238956891967, "grad_norm": 0.29035961627960205, "learning_rate": 2.7426753694544602e-08, "loss": 0.187, "step": 14344 }, { "epoch": 3.8171899946780203, "grad_norm": 0.2697470486164093, "learning_rate": 2.741501724902252e-08, "loss": 0.1806, "step": 14345 }, { "epoch": 3.817456093666844, "grad_norm": 0.3265936076641083, "learning_rate": 2.740328291626083e-08, "loss": 0.2141, "step": 14346 }, { "epoch": 3.817722192655668, "grad_norm": 0.30336621403694153, "learning_rate": 2.739155069660106e-08, "loss": 0.1929, "step": 14347 }, { "epoch": 3.817988291644492, "grad_norm": 0.3021429181098938, "learning_rate": 2.737982059038474e-08, "loss": 0.1986, "step": 14348 }, { "epoch": 3.8182543906333155, "grad_norm": 0.30188262462615967, "learning_rate": 2.7368092597953307e-08, "loss": 0.2122, "step": 14349 }, { "epoch": 3.8185204896221396, "grad_norm": 0.2994624674320221, "learning_rate": 2.7356366719648084e-08, "loss": 0.1923, "step": 14350 }, { "epoch": 3.8187865886109633, "grad_norm": 0.32406726479530334, "learning_rate": 2.7344642955810403e-08, "loss": 0.2024, "step": 14351 }, { "epoch": 3.819052687599787, "grad_norm": 0.26788291335105896, "learning_rate": 2.7332921306781555e-08, "loss": 0.1661, "step": 14352 }, { "epoch": 3.819318786588611, "grad_norm": 0.2923773527145386, "learning_rate": 2.7321201772902657e-08, "loss": 0.1793, "step": 14353 }, { "epoch": 3.819584885577435, "grad_norm": 0.30030933022499084, "learning_rate": 2.7309484354514866e-08, "loss": 0.2031, "step": 14354 }, { "epoch": 3.8198509845662585, "grad_norm": 0.28764498233795166, "learning_rate": 2.729776905195924e-08, "loss": 0.1883, "step": 14355 }, { "epoch": 3.8201170835550826, "grad_norm": 0.29747965931892395, "learning_rate": 2.7286055865576796e-08, "loss": 0.1883, "step": 14356 }, { "epoch": 3.8203831825439063, "grad_norm": 0.2813895046710968, "learning_rate": 2.7274344795708472e-08, "loss": 0.1875, "step": 14357 }, { "epoch": 3.82064928153273, "grad_norm": 0.265621155500412, "learning_rate": 2.7262635842695126e-08, "loss": 0.1663, "step": 14358 }, { "epoch": 3.820915380521554, "grad_norm": 0.3101663887500763, "learning_rate": 2.7250929006877597e-08, "loss": 0.1974, "step": 14359 }, { "epoch": 3.821181479510378, "grad_norm": 0.3159961998462677, "learning_rate": 2.7239224288596664e-08, "loss": 0.2085, "step": 14360 }, { "epoch": 3.8214475784992015, "grad_norm": 0.3159301280975342, "learning_rate": 2.7227521688192968e-08, "loss": 0.2108, "step": 14361 }, { "epoch": 3.8217136774880256, "grad_norm": 0.2802208364009857, "learning_rate": 2.721582120600715e-08, "loss": 0.1701, "step": 14362 }, { "epoch": 3.8219797764768493, "grad_norm": 0.32550475001335144, "learning_rate": 2.7204122842379794e-08, "loss": 0.2055, "step": 14363 }, { "epoch": 3.822245875465673, "grad_norm": 0.2904084026813507, "learning_rate": 2.719242659765142e-08, "loss": 0.1782, "step": 14364 }, { "epoch": 3.822511974454497, "grad_norm": 0.3228214383125305, "learning_rate": 2.718073247216246e-08, "loss": 0.1919, "step": 14365 }, { "epoch": 3.8227780734433208, "grad_norm": 0.2792135775089264, "learning_rate": 2.7169040466253312e-08, "loss": 0.1883, "step": 14366 }, { "epoch": 3.823044172432145, "grad_norm": 0.2903168499469757, "learning_rate": 2.7157350580264283e-08, "loss": 0.1999, "step": 14367 }, { "epoch": 3.8233102714209686, "grad_norm": 0.30389901995658875, "learning_rate": 2.714566281453564e-08, "loss": 0.1998, "step": 14368 }, { "epoch": 3.8235763704097927, "grad_norm": 0.3009321689605713, "learning_rate": 2.7133977169407628e-08, "loss": 0.2, "step": 14369 }, { "epoch": 3.8238424693986164, "grad_norm": 0.3274156451225281, "learning_rate": 2.7122293645220317e-08, "loss": 0.1867, "step": 14370 }, { "epoch": 3.82410856838744, "grad_norm": 0.3085572421550751, "learning_rate": 2.711061224231381e-08, "loss": 0.1958, "step": 14371 }, { "epoch": 3.824374667376264, "grad_norm": 0.3122933804988861, "learning_rate": 2.7098932961028155e-08, "loss": 0.1933, "step": 14372 }, { "epoch": 3.824640766365088, "grad_norm": 0.3237762451171875, "learning_rate": 2.7087255801703246e-08, "loss": 0.1927, "step": 14373 }, { "epoch": 3.8249068653539116, "grad_norm": 0.31346529722213745, "learning_rate": 2.7075580764678997e-08, "loss": 0.2, "step": 14374 }, { "epoch": 3.8251729643427357, "grad_norm": 0.2757636606693268, "learning_rate": 2.7063907850295252e-08, "loss": 0.1951, "step": 14375 }, { "epoch": 3.8254390633315594, "grad_norm": 0.29159414768218994, "learning_rate": 2.705223705889178e-08, "loss": 0.193, "step": 14376 }, { "epoch": 3.825705162320383, "grad_norm": 0.2984609305858612, "learning_rate": 2.704056839080827e-08, "loss": 0.2045, "step": 14377 }, { "epoch": 3.825971261309207, "grad_norm": 0.2801584005355835, "learning_rate": 2.702890184638441e-08, "loss": 0.1665, "step": 14378 }, { "epoch": 3.826237360298031, "grad_norm": 0.3617696464061737, "learning_rate": 2.7017237425959716e-08, "loss": 0.2007, "step": 14379 }, { "epoch": 3.8265034592868545, "grad_norm": 0.31433790922164917, "learning_rate": 2.700557512987375e-08, "loss": 0.198, "step": 14380 }, { "epoch": 3.8267695582756787, "grad_norm": 0.3014620244503021, "learning_rate": 2.6993914958465957e-08, "loss": 0.1991, "step": 14381 }, { "epoch": 3.8270356572645023, "grad_norm": 0.2850915193557739, "learning_rate": 2.6982256912075752e-08, "loss": 0.175, "step": 14382 }, { "epoch": 3.827301756253326, "grad_norm": 0.3049326241016388, "learning_rate": 2.6970600991042447e-08, "loss": 0.2107, "step": 14383 }, { "epoch": 3.82756785524215, "grad_norm": 0.2843150198459625, "learning_rate": 2.6958947195705328e-08, "loss": 0.1929, "step": 14384 }, { "epoch": 3.827833954230974, "grad_norm": 0.3049503266811371, "learning_rate": 2.694729552640361e-08, "loss": 0.199, "step": 14385 }, { "epoch": 3.8281000532197975, "grad_norm": 0.3203991651535034, "learning_rate": 2.6935645983476452e-08, "loss": 0.2036, "step": 14386 }, { "epoch": 3.8283661522086216, "grad_norm": 0.34444743394851685, "learning_rate": 2.692399856726295e-08, "loss": 0.2019, "step": 14387 }, { "epoch": 3.8286322511974453, "grad_norm": 0.34413379430770874, "learning_rate": 2.6912353278102072e-08, "loss": 0.1758, "step": 14388 }, { "epoch": 3.8288983501862695, "grad_norm": 0.2739041745662689, "learning_rate": 2.6900710116332836e-08, "loss": 0.1714, "step": 14389 }, { "epoch": 3.829164449175093, "grad_norm": 0.30897751450538635, "learning_rate": 2.6889069082294114e-08, "loss": 0.193, "step": 14390 }, { "epoch": 3.829430548163917, "grad_norm": 0.28607410192489624, "learning_rate": 2.6877430176324802e-08, "loss": 0.1851, "step": 14391 }, { "epoch": 3.829696647152741, "grad_norm": 0.28121477365493774, "learning_rate": 2.686579339876359e-08, "loss": 0.1734, "step": 14392 }, { "epoch": 3.8299627461415646, "grad_norm": 0.34096500277519226, "learning_rate": 2.6854158749949264e-08, "loss": 0.1906, "step": 14393 }, { "epoch": 3.8302288451303887, "grad_norm": 0.30704206228256226, "learning_rate": 2.684252623022044e-08, "loss": 0.1968, "step": 14394 }, { "epoch": 3.8304949441192124, "grad_norm": 0.3474652171134949, "learning_rate": 2.6830895839915723e-08, "loss": 0.2209, "step": 14395 }, { "epoch": 3.830761043108036, "grad_norm": 0.2940840423107147, "learning_rate": 2.6819267579373682e-08, "loss": 0.1868, "step": 14396 }, { "epoch": 3.8310271420968602, "grad_norm": 0.2902325987815857, "learning_rate": 2.680764144893273e-08, "loss": 0.1875, "step": 14397 }, { "epoch": 3.831293241085684, "grad_norm": 0.3042011559009552, "learning_rate": 2.6796017448931273e-08, "loss": 0.1969, "step": 14398 }, { "epoch": 3.8315593400745076, "grad_norm": 0.28834426403045654, "learning_rate": 2.67843955797077e-08, "loss": 0.1921, "step": 14399 }, { "epoch": 3.8318254390633317, "grad_norm": 0.3160373568534851, "learning_rate": 2.677277584160025e-08, "loss": 0.1987, "step": 14400 }, { "epoch": 3.8320915380521554, "grad_norm": 0.2811100482940674, "learning_rate": 2.6761158234947168e-08, "loss": 0.1755, "step": 14401 }, { "epoch": 3.832357637040979, "grad_norm": 0.299010694026947, "learning_rate": 2.6749542760086608e-08, "loss": 0.1899, "step": 14402 }, { "epoch": 3.832623736029803, "grad_norm": 0.338848739862442, "learning_rate": 2.6737929417356663e-08, "loss": 0.2046, "step": 14403 }, { "epoch": 3.832889835018627, "grad_norm": 0.29746586084365845, "learning_rate": 2.6726318207095376e-08, "loss": 0.2042, "step": 14404 }, { "epoch": 3.8331559340074506, "grad_norm": 0.2829759120941162, "learning_rate": 2.6714709129640733e-08, "loss": 0.1921, "step": 14405 }, { "epoch": 3.8334220329962747, "grad_norm": 0.2819727063179016, "learning_rate": 2.6703102185330607e-08, "loss": 0.1743, "step": 14406 }, { "epoch": 3.8336881319850984, "grad_norm": 0.32049837708473206, "learning_rate": 2.6691497374502846e-08, "loss": 0.1866, "step": 14407 }, { "epoch": 3.833954230973922, "grad_norm": 0.30395564436912537, "learning_rate": 2.6679894697495263e-08, "loss": 0.1791, "step": 14408 }, { "epoch": 3.834220329962746, "grad_norm": 0.31937605142593384, "learning_rate": 2.6668294154645577e-08, "loss": 0.2071, "step": 14409 }, { "epoch": 3.83448642895157, "grad_norm": 0.3024088144302368, "learning_rate": 2.6656695746291435e-08, "loss": 0.1809, "step": 14410 }, { "epoch": 3.8347525279403936, "grad_norm": 0.29693880677223206, "learning_rate": 2.6645099472770483e-08, "loss": 0.1908, "step": 14411 }, { "epoch": 3.8350186269292177, "grad_norm": 0.3130301237106323, "learning_rate": 2.663350533442018e-08, "loss": 0.191, "step": 14412 }, { "epoch": 3.8352847259180414, "grad_norm": 0.2905879616737366, "learning_rate": 2.662191333157805e-08, "loss": 0.1803, "step": 14413 }, { "epoch": 3.8355508249068655, "grad_norm": 0.2903027832508087, "learning_rate": 2.661032346458153e-08, "loss": 0.1993, "step": 14414 }, { "epoch": 3.835816923895689, "grad_norm": 0.2796382009983063, "learning_rate": 2.6598735733767907e-08, "loss": 0.1849, "step": 14415 }, { "epoch": 3.8360830228845133, "grad_norm": 0.2932462990283966, "learning_rate": 2.6587150139474513e-08, "loss": 0.1921, "step": 14416 }, { "epoch": 3.836349121873337, "grad_norm": 0.24973538517951965, "learning_rate": 2.6575566682038552e-08, "loss": 0.1592, "step": 14417 }, { "epoch": 3.8366152208621607, "grad_norm": 0.296673446893692, "learning_rate": 2.6563985361797204e-08, "loss": 0.1784, "step": 14418 }, { "epoch": 3.836881319850985, "grad_norm": 0.27165889739990234, "learning_rate": 2.655240617908756e-08, "loss": 0.1748, "step": 14419 }, { "epoch": 3.8371474188398085, "grad_norm": 0.2856878936290741, "learning_rate": 2.654082913424668e-08, "loss": 0.1897, "step": 14420 }, { "epoch": 3.837413517828632, "grad_norm": 0.25865471363067627, "learning_rate": 2.652925422761152e-08, "loss": 0.1715, "step": 14421 }, { "epoch": 3.8376796168174563, "grad_norm": 0.3886295557022095, "learning_rate": 2.6517681459519004e-08, "loss": 0.2031, "step": 14422 }, { "epoch": 3.83794571580628, "grad_norm": 0.3063647449016571, "learning_rate": 2.6506110830306016e-08, "loss": 0.208, "step": 14423 }, { "epoch": 3.8382118147951036, "grad_norm": 0.33446812629699707, "learning_rate": 2.649454234030929e-08, "loss": 0.2059, "step": 14424 }, { "epoch": 3.8384779137839278, "grad_norm": 0.3320409655570984, "learning_rate": 2.6482975989865587e-08, "loss": 0.2198, "step": 14425 }, { "epoch": 3.8387440127727515, "grad_norm": 0.27599218487739563, "learning_rate": 2.647141177931156e-08, "loss": 0.1889, "step": 14426 }, { "epoch": 3.839010111761575, "grad_norm": 0.2865406274795532, "learning_rate": 2.6459849708983817e-08, "loss": 0.182, "step": 14427 }, { "epoch": 3.8392762107503993, "grad_norm": 0.30543166399002075, "learning_rate": 2.6448289779218913e-08, "loss": 0.1925, "step": 14428 }, { "epoch": 3.839542309739223, "grad_norm": 0.3712671101093292, "learning_rate": 2.643673199035331e-08, "loss": 0.1906, "step": 14429 }, { "epoch": 3.8398084087280466, "grad_norm": 0.27775007486343384, "learning_rate": 2.6425176342723477e-08, "loss": 0.1691, "step": 14430 }, { "epoch": 3.8400745077168708, "grad_norm": 0.30617937445640564, "learning_rate": 2.6413622836665684e-08, "loss": 0.1762, "step": 14431 }, { "epoch": 3.8403406067056944, "grad_norm": 0.2998538911342621, "learning_rate": 2.6402071472516264e-08, "loss": 0.1793, "step": 14432 }, { "epoch": 3.840606705694518, "grad_norm": 0.2843931317329407, "learning_rate": 2.639052225061149e-08, "loss": 0.1637, "step": 14433 }, { "epoch": 3.8408728046833422, "grad_norm": 0.30053937435150146, "learning_rate": 2.6378975171287442e-08, "loss": 0.1951, "step": 14434 }, { "epoch": 3.841138903672166, "grad_norm": 0.2837582230567932, "learning_rate": 2.6367430234880283e-08, "loss": 0.178, "step": 14435 }, { "epoch": 3.84140500266099, "grad_norm": 0.2772776782512665, "learning_rate": 2.6355887441726042e-08, "loss": 0.1685, "step": 14436 }, { "epoch": 3.8416711016498137, "grad_norm": 0.27937236428260803, "learning_rate": 2.634434679216071e-08, "loss": 0.1785, "step": 14437 }, { "epoch": 3.8419372006386374, "grad_norm": 0.31002405285835266, "learning_rate": 2.633280828652019e-08, "loss": 0.1911, "step": 14438 }, { "epoch": 3.8422032996274615, "grad_norm": 0.29631930589675903, "learning_rate": 2.6321271925140342e-08, "loss": 0.2041, "step": 14439 }, { "epoch": 3.8424693986162852, "grad_norm": 0.30124515295028687, "learning_rate": 2.6309737708356972e-08, "loss": 0.1786, "step": 14440 }, { "epoch": 3.8427354976051094, "grad_norm": 0.27848634123802185, "learning_rate": 2.6298205636505834e-08, "loss": 0.1895, "step": 14441 }, { "epoch": 3.843001596593933, "grad_norm": 0.31465309858322144, "learning_rate": 2.6286675709922524e-08, "loss": 0.2033, "step": 14442 }, { "epoch": 3.8432676955827567, "grad_norm": 0.3111461400985718, "learning_rate": 2.6275147928942697e-08, "loss": 0.1878, "step": 14443 }, { "epoch": 3.843533794571581, "grad_norm": 0.32842889428138733, "learning_rate": 2.626362229390189e-08, "loss": 0.2101, "step": 14444 }, { "epoch": 3.8437998935604045, "grad_norm": 0.3181668817996979, "learning_rate": 2.625209880513558e-08, "loss": 0.1914, "step": 14445 }, { "epoch": 3.844065992549228, "grad_norm": 0.30381280183792114, "learning_rate": 2.6240577462979187e-08, "loss": 0.197, "step": 14446 }, { "epoch": 3.8443320915380523, "grad_norm": 0.2854447662830353, "learning_rate": 2.6229058267768067e-08, "loss": 0.1769, "step": 14447 }, { "epoch": 3.844598190526876, "grad_norm": 0.29749950766563416, "learning_rate": 2.6217541219837524e-08, "loss": 0.179, "step": 14448 }, { "epoch": 3.8448642895156997, "grad_norm": 0.31189975142478943, "learning_rate": 2.620602631952278e-08, "loss": 0.1969, "step": 14449 }, { "epoch": 3.845130388504524, "grad_norm": 0.290598064661026, "learning_rate": 2.6194513567159038e-08, "loss": 0.1793, "step": 14450 }, { "epoch": 3.8453964874933475, "grad_norm": 0.319015234708786, "learning_rate": 2.6183002963081346e-08, "loss": 0.2048, "step": 14451 }, { "epoch": 3.845662586482171, "grad_norm": 0.33870309591293335, "learning_rate": 2.617149450762476e-08, "loss": 0.1973, "step": 14452 }, { "epoch": 3.8459286854709953, "grad_norm": 0.2714950442314148, "learning_rate": 2.6159988201124317e-08, "loss": 0.1819, "step": 14453 }, { "epoch": 3.846194784459819, "grad_norm": 0.30933499336242676, "learning_rate": 2.6148484043914852e-08, "loss": 0.1964, "step": 14454 }, { "epoch": 3.8464608834486427, "grad_norm": 0.29055503010749817, "learning_rate": 2.613698203633128e-08, "loss": 0.1826, "step": 14455 }, { "epoch": 3.846726982437467, "grad_norm": 0.3007749915122986, "learning_rate": 2.612548217870837e-08, "loss": 0.1814, "step": 14456 }, { "epoch": 3.8469930814262905, "grad_norm": 0.27317461371421814, "learning_rate": 2.6113984471380856e-08, "loss": 0.1751, "step": 14457 }, { "epoch": 3.847259180415114, "grad_norm": 0.2971697151660919, "learning_rate": 2.610248891468341e-08, "loss": 0.1724, "step": 14458 }, { "epoch": 3.8475252794039383, "grad_norm": 0.31652945280075073, "learning_rate": 2.6090995508950675e-08, "loss": 0.1992, "step": 14459 }, { "epoch": 3.847791378392762, "grad_norm": 0.285853773355484, "learning_rate": 2.6079504254517125e-08, "loss": 0.1879, "step": 14460 }, { "epoch": 3.848057477381586, "grad_norm": 0.30059316754341125, "learning_rate": 2.606801515171727e-08, "loss": 0.2076, "step": 14461 }, { "epoch": 3.84832357637041, "grad_norm": 0.3205145299434662, "learning_rate": 2.605652820088554e-08, "loss": 0.1902, "step": 14462 }, { "epoch": 3.848589675359234, "grad_norm": 0.31756100058555603, "learning_rate": 2.604504340235627e-08, "loss": 0.2072, "step": 14463 }, { "epoch": 3.8488557743480576, "grad_norm": 0.2802776098251343, "learning_rate": 2.603356075646378e-08, "loss": 0.1967, "step": 14464 }, { "epoch": 3.8491218733368813, "grad_norm": 0.3043570816516876, "learning_rate": 2.6022080263542266e-08, "loss": 0.1917, "step": 14465 }, { "epoch": 3.8493879723257054, "grad_norm": 0.3131057620048523, "learning_rate": 2.6010601923925922e-08, "loss": 0.1978, "step": 14466 }, { "epoch": 3.849654071314529, "grad_norm": 0.3250318467617035, "learning_rate": 2.5999125737948858e-08, "loss": 0.1984, "step": 14467 }, { "epoch": 3.8499201703033528, "grad_norm": 0.27887779474258423, "learning_rate": 2.5987651705945113e-08, "loss": 0.1847, "step": 14468 }, { "epoch": 3.850186269292177, "grad_norm": 0.2843048572540283, "learning_rate": 2.5976179828248633e-08, "loss": 0.1977, "step": 14469 }, { "epoch": 3.8504523682810006, "grad_norm": 0.2951229512691498, "learning_rate": 2.5964710105193354e-08, "loss": 0.195, "step": 14470 }, { "epoch": 3.8507184672698243, "grad_norm": 0.3035801947116852, "learning_rate": 2.595324253711314e-08, "loss": 0.1914, "step": 14471 }, { "epoch": 3.8509845662586484, "grad_norm": 0.2988489866256714, "learning_rate": 2.594177712434178e-08, "loss": 0.1928, "step": 14472 }, { "epoch": 3.851250665247472, "grad_norm": 0.31947827339172363, "learning_rate": 2.5930313867213016e-08, "loss": 0.1828, "step": 14473 }, { "epoch": 3.8515167642362957, "grad_norm": 0.31897929310798645, "learning_rate": 2.5918852766060463e-08, "loss": 0.1977, "step": 14474 }, { "epoch": 3.85178286322512, "grad_norm": 0.6119996309280396, "learning_rate": 2.590739382121777e-08, "loss": 0.1909, "step": 14475 }, { "epoch": 3.8520489622139436, "grad_norm": 0.2821000814437866, "learning_rate": 2.5895937033018455e-08, "loss": 0.172, "step": 14476 }, { "epoch": 3.8523150612027672, "grad_norm": 0.30248668789863586, "learning_rate": 2.5884482401796028e-08, "loss": 0.2003, "step": 14477 }, { "epoch": 3.8525811601915914, "grad_norm": 0.2789345383644104, "learning_rate": 2.587302992788386e-08, "loss": 0.1822, "step": 14478 }, { "epoch": 3.852847259180415, "grad_norm": 0.27956119179725647, "learning_rate": 2.5861579611615315e-08, "loss": 0.1798, "step": 14479 }, { "epoch": 3.8531133581692387, "grad_norm": 0.3026965856552124, "learning_rate": 2.5850131453323687e-08, "loss": 0.2084, "step": 14480 }, { "epoch": 3.853379457158063, "grad_norm": 0.274133563041687, "learning_rate": 2.5838685453342214e-08, "loss": 0.1744, "step": 14481 }, { "epoch": 3.8536455561468865, "grad_norm": 0.31483593583106995, "learning_rate": 2.5827241612004046e-08, "loss": 0.195, "step": 14482 }, { "epoch": 3.85391165513571, "grad_norm": 0.2661260962486267, "learning_rate": 2.581579992964228e-08, "loss": 0.1736, "step": 14483 }, { "epoch": 3.8541777541245343, "grad_norm": 0.3122768700122833, "learning_rate": 2.5804360406589966e-08, "loss": 0.1944, "step": 14484 }, { "epoch": 3.854443853113358, "grad_norm": 0.29061010479927063, "learning_rate": 2.5792923043180082e-08, "loss": 0.1925, "step": 14485 }, { "epoch": 3.854709952102182, "grad_norm": 0.3146178424358368, "learning_rate": 2.5781487839745553e-08, "loss": 0.1931, "step": 14486 }, { "epoch": 3.854976051091006, "grad_norm": 0.26889756321907043, "learning_rate": 2.5770054796619167e-08, "loss": 0.1784, "step": 14487 }, { "epoch": 3.85524215007983, "grad_norm": 0.31528565287590027, "learning_rate": 2.575862391413376e-08, "loss": 0.2084, "step": 14488 }, { "epoch": 3.8555082490686536, "grad_norm": 0.30327314138412476, "learning_rate": 2.574719519262205e-08, "loss": 0.1802, "step": 14489 }, { "epoch": 3.8557743480574773, "grad_norm": 0.30414214730262756, "learning_rate": 2.5735768632416687e-08, "loss": 0.1864, "step": 14490 }, { "epoch": 3.8560404470463014, "grad_norm": 0.2812018394470215, "learning_rate": 2.5724344233850272e-08, "loss": 0.1779, "step": 14491 }, { "epoch": 3.856306546035125, "grad_norm": 0.330976277589798, "learning_rate": 2.5712921997255377e-08, "loss": 0.1946, "step": 14492 }, { "epoch": 3.856572645023949, "grad_norm": 0.28965580463409424, "learning_rate": 2.5701501922964396e-08, "loss": 0.1903, "step": 14493 }, { "epoch": 3.856838744012773, "grad_norm": 0.2870928645133972, "learning_rate": 2.56900840113098e-08, "loss": 0.1849, "step": 14494 }, { "epoch": 3.8571048430015966, "grad_norm": 0.3133874237537384, "learning_rate": 2.5678668262623926e-08, "loss": 0.1994, "step": 14495 }, { "epoch": 3.8573709419904203, "grad_norm": 0.29403194785118103, "learning_rate": 2.5667254677239025e-08, "loss": 0.1916, "step": 14496 }, { "epoch": 3.8576370409792444, "grad_norm": 0.31955134868621826, "learning_rate": 2.565584325548733e-08, "loss": 0.2022, "step": 14497 }, { "epoch": 3.857903139968068, "grad_norm": 0.3047378659248352, "learning_rate": 2.564443399770101e-08, "loss": 0.1976, "step": 14498 }, { "epoch": 3.858169238956892, "grad_norm": 0.3281335234642029, "learning_rate": 2.563302690421215e-08, "loss": 0.2164, "step": 14499 }, { "epoch": 3.858435337945716, "grad_norm": 0.30793681740760803, "learning_rate": 2.5621621975352792e-08, "loss": 0.1948, "step": 14500 }, { "epoch": 3.8587014369345396, "grad_norm": 0.2933942377567291, "learning_rate": 2.56102192114549e-08, "loss": 0.1915, "step": 14501 }, { "epoch": 3.8589675359233633, "grad_norm": 0.3468838930130005, "learning_rate": 2.559881861285037e-08, "loss": 0.1965, "step": 14502 }, { "epoch": 3.8592336349121874, "grad_norm": 0.29600244760513306, "learning_rate": 2.5587420179871044e-08, "loss": 0.1964, "step": 14503 }, { "epoch": 3.859499733901011, "grad_norm": 0.3069950044155121, "learning_rate": 2.557602391284874e-08, "loss": 0.1912, "step": 14504 }, { "epoch": 3.8597658328898348, "grad_norm": 0.31170275807380676, "learning_rate": 2.5564629812115114e-08, "loss": 0.1956, "step": 14505 }, { "epoch": 3.860031931878659, "grad_norm": 0.3014833629131317, "learning_rate": 2.555323787800184e-08, "loss": 0.1971, "step": 14506 }, { "epoch": 3.8602980308674826, "grad_norm": 0.3122670352458954, "learning_rate": 2.5541848110840514e-08, "loss": 0.1875, "step": 14507 }, { "epoch": 3.8605641298563067, "grad_norm": 0.3035942316055298, "learning_rate": 2.5530460510962658e-08, "loss": 0.1815, "step": 14508 }, { "epoch": 3.8608302288451304, "grad_norm": 0.2907525599002838, "learning_rate": 2.551907507869975e-08, "loss": 0.1802, "step": 14509 }, { "epoch": 3.861096327833954, "grad_norm": 0.3709922134876251, "learning_rate": 2.550769181438317e-08, "loss": 0.2076, "step": 14510 }, { "epoch": 3.861362426822778, "grad_norm": 0.2842089831829071, "learning_rate": 2.549631071834426e-08, "loss": 0.1785, "step": 14511 }, { "epoch": 3.861628525811602, "grad_norm": 0.3074076473712921, "learning_rate": 2.5484931790914332e-08, "loss": 0.1922, "step": 14512 }, { "epoch": 3.861894624800426, "grad_norm": 0.3558569848537445, "learning_rate": 2.547355503242453e-08, "loss": 0.1941, "step": 14513 }, { "epoch": 3.8621607237892497, "grad_norm": 0.311478853225708, "learning_rate": 2.5462180443206037e-08, "loss": 0.2046, "step": 14514 }, { "epoch": 3.8624268227780734, "grad_norm": 0.28691112995147705, "learning_rate": 2.5450808023589976e-08, "loss": 0.1821, "step": 14515 }, { "epoch": 3.8626929217668975, "grad_norm": 0.3001019358634949, "learning_rate": 2.543943777390729e-08, "loss": 0.1934, "step": 14516 }, { "epoch": 3.862959020755721, "grad_norm": 0.2967360019683838, "learning_rate": 2.5428069694488974e-08, "loss": 0.1865, "step": 14517 }, { "epoch": 3.863225119744545, "grad_norm": 0.29479867219924927, "learning_rate": 2.5416703785665926e-08, "loss": 0.1956, "step": 14518 }, { "epoch": 3.863491218733369, "grad_norm": 0.27113738656044006, "learning_rate": 2.5405340047768976e-08, "loss": 0.1893, "step": 14519 }, { "epoch": 3.8637573177221927, "grad_norm": 0.3045903444290161, "learning_rate": 2.5393978481128897e-08, "loss": 0.1998, "step": 14520 }, { "epoch": 3.8640234167110163, "grad_norm": 0.2894510328769684, "learning_rate": 2.5382619086076385e-08, "loss": 0.1883, "step": 14521 }, { "epoch": 3.8642895156998405, "grad_norm": 0.320930153131485, "learning_rate": 2.537126186294213e-08, "loss": 0.2028, "step": 14522 }, { "epoch": 3.864555614688664, "grad_norm": 0.31561198830604553, "learning_rate": 2.535990681205663e-08, "loss": 0.2063, "step": 14523 }, { "epoch": 3.864821713677488, "grad_norm": 0.2890869379043579, "learning_rate": 2.534855393375045e-08, "loss": 0.1999, "step": 14524 }, { "epoch": 3.865087812666312, "grad_norm": 0.3245573937892914, "learning_rate": 2.533720322835403e-08, "loss": 0.1989, "step": 14525 }, { "epoch": 3.8653539116551356, "grad_norm": 0.2797115445137024, "learning_rate": 2.5325854696197768e-08, "loss": 0.1854, "step": 14526 }, { "epoch": 3.8656200106439593, "grad_norm": 0.27908074855804443, "learning_rate": 2.5314508337611984e-08, "loss": 0.1771, "step": 14527 }, { "epoch": 3.8658861096327835, "grad_norm": 0.3066927194595337, "learning_rate": 2.530316415292695e-08, "loss": 0.1927, "step": 14528 }, { "epoch": 3.866152208621607, "grad_norm": 0.2901553809642792, "learning_rate": 2.5291822142472853e-08, "loss": 0.1989, "step": 14529 }, { "epoch": 3.866418307610431, "grad_norm": 0.3028433620929718, "learning_rate": 2.528048230657983e-08, "loss": 0.18, "step": 14530 }, { "epoch": 3.866684406599255, "grad_norm": 0.2834811806678772, "learning_rate": 2.526914464557799e-08, "loss": 0.2065, "step": 14531 }, { "epoch": 3.8669505055880786, "grad_norm": 0.3547173738479614, "learning_rate": 2.5257809159797294e-08, "loss": 0.1791, "step": 14532 }, { "epoch": 3.8672166045769027, "grad_norm": 0.2960823178291321, "learning_rate": 2.5246475849567696e-08, "loss": 0.1764, "step": 14533 }, { "epoch": 3.8674827035657264, "grad_norm": 0.29587993025779724, "learning_rate": 2.5235144715219126e-08, "loss": 0.1944, "step": 14534 }, { "epoch": 3.8677488025545506, "grad_norm": 0.3139285743236542, "learning_rate": 2.5223815757081336e-08, "loss": 0.1966, "step": 14535 }, { "epoch": 3.8680149015433742, "grad_norm": 0.27349963784217834, "learning_rate": 2.5212488975484115e-08, "loss": 0.1762, "step": 14536 }, { "epoch": 3.868281000532198, "grad_norm": 0.2895015478134155, "learning_rate": 2.5201164370757155e-08, "loss": 0.1765, "step": 14537 }, { "epoch": 3.868547099521022, "grad_norm": 0.32672420144081116, "learning_rate": 2.5189841943230093e-08, "loss": 0.2053, "step": 14538 }, { "epoch": 3.8688131985098457, "grad_norm": 0.28490763902664185, "learning_rate": 2.517852169323249e-08, "loss": 0.1824, "step": 14539 }, { "epoch": 3.8690792974986694, "grad_norm": 0.28370776772499084, "learning_rate": 2.5167203621093868e-08, "loss": 0.1799, "step": 14540 }, { "epoch": 3.8693453964874935, "grad_norm": 0.3015349209308624, "learning_rate": 2.5155887727143633e-08, "loss": 0.1943, "step": 14541 }, { "epoch": 3.869611495476317, "grad_norm": 0.2866630256175995, "learning_rate": 2.514457401171116e-08, "loss": 0.1865, "step": 14542 }, { "epoch": 3.869877594465141, "grad_norm": 0.3000344932079315, "learning_rate": 2.5133262475125782e-08, "loss": 0.1864, "step": 14543 }, { "epoch": 3.870143693453965, "grad_norm": 0.30293723940849304, "learning_rate": 2.5121953117716744e-08, "loss": 0.1809, "step": 14544 }, { "epoch": 3.8704097924427887, "grad_norm": 0.28649047017097473, "learning_rate": 2.5110645939813235e-08, "loss": 0.1768, "step": 14545 }, { "epoch": 3.8706758914316124, "grad_norm": 0.281889945268631, "learning_rate": 2.5099340941744374e-08, "loss": 0.1778, "step": 14546 }, { "epoch": 3.8709419904204365, "grad_norm": 0.29307061433792114, "learning_rate": 2.5088038123839216e-08, "loss": 0.2025, "step": 14547 }, { "epoch": 3.87120808940926, "grad_norm": 0.30707404017448425, "learning_rate": 2.5076737486426768e-08, "loss": 0.2019, "step": 14548 }, { "epoch": 3.871474188398084, "grad_norm": 0.30135706067085266, "learning_rate": 2.506543902983598e-08, "loss": 0.1854, "step": 14549 }, { "epoch": 3.871740287386908, "grad_norm": 0.2758747339248657, "learning_rate": 2.505414275439567e-08, "loss": 0.1657, "step": 14550 }, { "epoch": 3.8720063863757317, "grad_norm": 0.3463132977485657, "learning_rate": 2.504284866043467e-08, "loss": 0.1751, "step": 14551 }, { "epoch": 3.8722724853645554, "grad_norm": 0.27171167731285095, "learning_rate": 2.5031556748281713e-08, "loss": 0.1697, "step": 14552 }, { "epoch": 3.8725385843533795, "grad_norm": 0.2842758297920227, "learning_rate": 2.5020267018265495e-08, "loss": 0.18, "step": 14553 }, { "epoch": 3.872804683342203, "grad_norm": 0.3032020628452301, "learning_rate": 2.500897947071464e-08, "loss": 0.2003, "step": 14554 }, { "epoch": 3.8730707823310273, "grad_norm": 0.2893627882003784, "learning_rate": 2.4997694105957666e-08, "loss": 0.1643, "step": 14555 }, { "epoch": 3.873336881319851, "grad_norm": 0.29035645723342896, "learning_rate": 2.4986410924323064e-08, "loss": 0.1819, "step": 14556 }, { "epoch": 3.8736029803086747, "grad_norm": 0.3081446886062622, "learning_rate": 2.4975129926139272e-08, "loss": 0.1784, "step": 14557 }, { "epoch": 3.873869079297499, "grad_norm": 0.2977047562599182, "learning_rate": 2.496385111173468e-08, "loss": 0.1997, "step": 14558 }, { "epoch": 3.8741351782863225, "grad_norm": 0.31181034445762634, "learning_rate": 2.4952574481437515e-08, "loss": 0.2067, "step": 14559 }, { "epoch": 3.8744012772751466, "grad_norm": 0.29790404438972473, "learning_rate": 2.4941300035576053e-08, "loss": 0.1908, "step": 14560 }, { "epoch": 3.8746673762639703, "grad_norm": 0.25679120421409607, "learning_rate": 2.493002777447846e-08, "loss": 0.174, "step": 14561 }, { "epoch": 3.874933475252794, "grad_norm": 0.3160361051559448, "learning_rate": 2.4918757698472846e-08, "loss": 0.2037, "step": 14562 }, { "epoch": 3.875199574241618, "grad_norm": 0.31099843978881836, "learning_rate": 2.490748980788724e-08, "loss": 0.1911, "step": 14563 }, { "epoch": 3.8754656732304418, "grad_norm": 0.3155432939529419, "learning_rate": 2.4896224103049633e-08, "loss": 0.1879, "step": 14564 }, { "epoch": 3.8757317722192655, "grad_norm": 0.2844476103782654, "learning_rate": 2.488496058428794e-08, "loss": 0.1908, "step": 14565 }, { "epoch": 3.8759978712080896, "grad_norm": 0.29402294754981995, "learning_rate": 2.4873699251930013e-08, "loss": 0.1926, "step": 14566 }, { "epoch": 3.8762639701969133, "grad_norm": 0.2920776605606079, "learning_rate": 2.4862440106303663e-08, "loss": 0.1976, "step": 14567 }, { "epoch": 3.876530069185737, "grad_norm": 0.2911014258861542, "learning_rate": 2.4851183147736553e-08, "loss": 0.1851, "step": 14568 }, { "epoch": 3.876796168174561, "grad_norm": 0.30032721161842346, "learning_rate": 2.4839928376556384e-08, "loss": 0.1827, "step": 14569 }, { "epoch": 3.8770622671633848, "grad_norm": 0.2838684618473053, "learning_rate": 2.482867579309075e-08, "loss": 0.1869, "step": 14570 }, { "epoch": 3.8773283661522084, "grad_norm": 0.3029843866825104, "learning_rate": 2.481742539766718e-08, "loss": 0.1947, "step": 14571 }, { "epoch": 3.8775944651410326, "grad_norm": 0.2860492467880249, "learning_rate": 2.4806177190613152e-08, "loss": 0.2027, "step": 14572 }, { "epoch": 3.8778605641298562, "grad_norm": 0.28189539909362793, "learning_rate": 2.4794931172256094e-08, "loss": 0.1819, "step": 14573 }, { "epoch": 3.87812666311868, "grad_norm": 0.2635709345340729, "learning_rate": 2.47836873429233e-08, "loss": 0.1782, "step": 14574 }, { "epoch": 3.878392762107504, "grad_norm": 0.30388495326042175, "learning_rate": 2.4772445702942056e-08, "loss": 0.1863, "step": 14575 }, { "epoch": 3.8786588610963277, "grad_norm": 0.3191526532173157, "learning_rate": 2.47612062526396e-08, "loss": 0.206, "step": 14576 }, { "epoch": 3.8789249600851514, "grad_norm": 0.31739866733551025, "learning_rate": 2.4749968992343107e-08, "loss": 0.1845, "step": 14577 }, { "epoch": 3.8791910590739755, "grad_norm": 0.3173653781414032, "learning_rate": 2.4738733922379597e-08, "loss": 0.2123, "step": 14578 }, { "epoch": 3.8794571580627992, "grad_norm": 0.2833060026168823, "learning_rate": 2.4727501043076126e-08, "loss": 0.1899, "step": 14579 }, { "epoch": 3.8797232570516234, "grad_norm": 0.2985784709453583, "learning_rate": 2.471627035475966e-08, "loss": 0.191, "step": 14580 }, { "epoch": 3.879989356040447, "grad_norm": 0.3214353919029236, "learning_rate": 2.470504185775709e-08, "loss": 0.2005, "step": 14581 }, { "epoch": 3.880255455029271, "grad_norm": 0.30012446641921997, "learning_rate": 2.469381555239526e-08, "loss": 0.1854, "step": 14582 }, { "epoch": 3.880521554018095, "grad_norm": 0.3050520122051239, "learning_rate": 2.468259143900091e-08, "loss": 0.181, "step": 14583 }, { "epoch": 3.8807876530069185, "grad_norm": 0.29981279373168945, "learning_rate": 2.4671369517900774e-08, "loss": 0.1975, "step": 14584 }, { "epoch": 3.8810537519957427, "grad_norm": 0.31197667121887207, "learning_rate": 2.466014978942149e-08, "loss": 0.2007, "step": 14585 }, { "epoch": 3.8813198509845663, "grad_norm": 0.2971017062664032, "learning_rate": 2.4648932253889642e-08, "loss": 0.183, "step": 14586 }, { "epoch": 3.88158594997339, "grad_norm": 0.3107544183731079, "learning_rate": 2.463771691163171e-08, "loss": 0.1894, "step": 14587 }, { "epoch": 3.881852048962214, "grad_norm": 0.2834124267101288, "learning_rate": 2.4626503762974162e-08, "loss": 0.1808, "step": 14588 }, { "epoch": 3.882118147951038, "grad_norm": 0.3121456801891327, "learning_rate": 2.461529280824338e-08, "loss": 0.2063, "step": 14589 }, { "epoch": 3.8823842469398615, "grad_norm": 0.26921194791793823, "learning_rate": 2.4604084047765693e-08, "loss": 0.1725, "step": 14590 }, { "epoch": 3.8826503459286856, "grad_norm": 0.34834015369415283, "learning_rate": 2.4592877481867357e-08, "loss": 0.1991, "step": 14591 }, { "epoch": 3.8829164449175093, "grad_norm": 0.28925713896751404, "learning_rate": 2.4581673110874558e-08, "loss": 0.1887, "step": 14592 }, { "epoch": 3.883182543906333, "grad_norm": 0.2761583924293518, "learning_rate": 2.4570470935113464e-08, "loss": 0.1808, "step": 14593 }, { "epoch": 3.883448642895157, "grad_norm": 0.29726722836494446, "learning_rate": 2.4559270954910083e-08, "loss": 0.1863, "step": 14594 }, { "epoch": 3.883714741883981, "grad_norm": 0.2834016978740692, "learning_rate": 2.4548073170590456e-08, "loss": 0.1935, "step": 14595 }, { "epoch": 3.8839808408728045, "grad_norm": 0.2951810956001282, "learning_rate": 2.453687758248052e-08, "loss": 0.1892, "step": 14596 }, { "epoch": 3.8842469398616286, "grad_norm": 0.28828340768814087, "learning_rate": 2.4525684190906128e-08, "loss": 0.1912, "step": 14597 }, { "epoch": 3.8845130388504523, "grad_norm": 0.31043052673339844, "learning_rate": 2.4514492996193092e-08, "loss": 0.2008, "step": 14598 }, { "epoch": 3.884779137839276, "grad_norm": 0.32431429624557495, "learning_rate": 2.450330399866717e-08, "loss": 0.2091, "step": 14599 }, { "epoch": 3.8850452368281, "grad_norm": 0.29629456996917725, "learning_rate": 2.4492117198654038e-08, "loss": 0.192, "step": 14600 }, { "epoch": 3.885311335816924, "grad_norm": 0.3011493682861328, "learning_rate": 2.4480932596479332e-08, "loss": 0.1938, "step": 14601 }, { "epoch": 3.8855774348057475, "grad_norm": 0.2981703281402588, "learning_rate": 2.4469750192468587e-08, "loss": 0.1899, "step": 14602 }, { "epoch": 3.8858435337945716, "grad_norm": 0.2943156361579895, "learning_rate": 2.4458569986947298e-08, "loss": 0.1992, "step": 14603 }, { "epoch": 3.8861096327833953, "grad_norm": 0.2886800765991211, "learning_rate": 2.4447391980240928e-08, "loss": 0.1794, "step": 14604 }, { "epoch": 3.8863757317722194, "grad_norm": 0.3042771518230438, "learning_rate": 2.4436216172674784e-08, "loss": 0.2077, "step": 14605 }, { "epoch": 3.886641830761043, "grad_norm": 0.30353933572769165, "learning_rate": 2.4425042564574183e-08, "loss": 0.1823, "step": 14606 }, { "epoch": 3.886907929749867, "grad_norm": 0.2977023124694824, "learning_rate": 2.4413871156264365e-08, "loss": 0.1893, "step": 14607 }, { "epoch": 3.887174028738691, "grad_norm": 0.3028176724910736, "learning_rate": 2.4402701948070492e-08, "loss": 0.2102, "step": 14608 }, { "epoch": 3.8874401277275146, "grad_norm": 0.34105950593948364, "learning_rate": 2.4391534940317693e-08, "loss": 0.2013, "step": 14609 }, { "epoch": 3.8877062267163387, "grad_norm": 0.32695019245147705, "learning_rate": 2.4380370133330986e-08, "loss": 0.1892, "step": 14610 }, { "epoch": 3.8879723257051624, "grad_norm": 0.2977851331233978, "learning_rate": 2.4369207527435364e-08, "loss": 0.1832, "step": 14611 }, { "epoch": 3.888238424693986, "grad_norm": 0.28821659088134766, "learning_rate": 2.4358047122955737e-08, "loss": 0.1723, "step": 14612 }, { "epoch": 3.88850452368281, "grad_norm": 0.3163782060146332, "learning_rate": 2.434688892021699e-08, "loss": 0.1943, "step": 14613 }, { "epoch": 3.888770622671634, "grad_norm": 0.2911611497402191, "learning_rate": 2.433573291954384e-08, "loss": 0.1959, "step": 14614 }, { "epoch": 3.8890367216604576, "grad_norm": 0.2917552590370178, "learning_rate": 2.4324579121261047e-08, "loss": 0.2022, "step": 14615 }, { "epoch": 3.8893028206492817, "grad_norm": 0.3175012171268463, "learning_rate": 2.4313427525693298e-08, "loss": 0.2053, "step": 14616 }, { "epoch": 3.8895689196381054, "grad_norm": 0.30843818187713623, "learning_rate": 2.430227813316512e-08, "loss": 0.2066, "step": 14617 }, { "epoch": 3.889835018626929, "grad_norm": 0.29378828406333923, "learning_rate": 2.429113094400108e-08, "loss": 0.1893, "step": 14618 }, { "epoch": 3.890101117615753, "grad_norm": 0.30470526218414307, "learning_rate": 2.427998595852564e-08, "loss": 0.2121, "step": 14619 }, { "epoch": 3.890367216604577, "grad_norm": 0.269731342792511, "learning_rate": 2.4268843177063215e-08, "loss": 0.1852, "step": 14620 }, { "epoch": 3.8906333155934005, "grad_norm": 0.2807960510253906, "learning_rate": 2.4257702599938123e-08, "loss": 0.1922, "step": 14621 }, { "epoch": 3.8908994145822247, "grad_norm": 0.28410401940345764, "learning_rate": 2.4246564227474674e-08, "loss": 0.1852, "step": 14622 }, { "epoch": 3.8911655135710483, "grad_norm": 0.29152223467826843, "learning_rate": 2.423542805999702e-08, "loss": 0.1899, "step": 14623 }, { "epoch": 3.891431612559872, "grad_norm": 0.2877659797668457, "learning_rate": 2.4224294097829334e-08, "loss": 0.1782, "step": 14624 }, { "epoch": 3.891697711548696, "grad_norm": 0.30219587683677673, "learning_rate": 2.4213162341295702e-08, "loss": 0.1961, "step": 14625 }, { "epoch": 3.89196381053752, "grad_norm": 0.3117356598377228, "learning_rate": 2.420203279072013e-08, "loss": 0.1869, "step": 14626 }, { "epoch": 3.892229909526344, "grad_norm": 0.3062775135040283, "learning_rate": 2.4190905446426578e-08, "loss": 0.1915, "step": 14627 }, { "epoch": 3.8924960085151676, "grad_norm": 0.2727998197078705, "learning_rate": 2.4179780308738928e-08, "loss": 0.1872, "step": 14628 }, { "epoch": 3.8927621075039913, "grad_norm": 0.28164035081863403, "learning_rate": 2.4168657377981017e-08, "loss": 0.1711, "step": 14629 }, { "epoch": 3.8930282064928154, "grad_norm": 0.29419922828674316, "learning_rate": 2.4157536654476585e-08, "loss": 0.1825, "step": 14630 }, { "epoch": 3.893294305481639, "grad_norm": 0.2914079427719116, "learning_rate": 2.4146418138549375e-08, "loss": 0.1899, "step": 14631 }, { "epoch": 3.8935604044704633, "grad_norm": 0.2936497628688812, "learning_rate": 2.4135301830522946e-08, "loss": 0.1835, "step": 14632 }, { "epoch": 3.893826503459287, "grad_norm": 0.29624226689338684, "learning_rate": 2.4124187730720912e-08, "loss": 0.193, "step": 14633 }, { "epoch": 3.8940926024481106, "grad_norm": 0.31317225098609924, "learning_rate": 2.4113075839466756e-08, "loss": 0.1979, "step": 14634 }, { "epoch": 3.8943587014369347, "grad_norm": 0.27298739552497864, "learning_rate": 2.4101966157083953e-08, "loss": 0.1784, "step": 14635 }, { "epoch": 3.8946248004257584, "grad_norm": 0.30366986989974976, "learning_rate": 2.409085868389581e-08, "loss": 0.1906, "step": 14636 }, { "epoch": 3.894890899414582, "grad_norm": 0.29991066455841064, "learning_rate": 2.407975342022569e-08, "loss": 0.1947, "step": 14637 }, { "epoch": 3.8951569984034062, "grad_norm": 0.30476686358451843, "learning_rate": 2.4068650366396813e-08, "loss": 0.178, "step": 14638 }, { "epoch": 3.89542309739223, "grad_norm": 0.2949827313423157, "learning_rate": 2.405754952273237e-08, "loss": 0.1878, "step": 14639 }, { "epoch": 3.8956891963810536, "grad_norm": 0.28713926672935486, "learning_rate": 2.4046450889555514e-08, "loss": 0.1638, "step": 14640 }, { "epoch": 3.8959552953698777, "grad_norm": 0.28856953978538513, "learning_rate": 2.4035354467189228e-08, "loss": 0.1907, "step": 14641 }, { "epoch": 3.8962213943587014, "grad_norm": 0.2951123118400574, "learning_rate": 2.402426025595653e-08, "loss": 0.1851, "step": 14642 }, { "epoch": 3.896487493347525, "grad_norm": 0.296022891998291, "learning_rate": 2.401316825618035e-08, "loss": 0.1761, "step": 14643 }, { "epoch": 3.896753592336349, "grad_norm": 0.326825350522995, "learning_rate": 2.400207846818354e-08, "loss": 0.1936, "step": 14644 }, { "epoch": 3.897019691325173, "grad_norm": 0.29254716634750366, "learning_rate": 2.3990990892288908e-08, "loss": 0.1796, "step": 14645 }, { "epoch": 3.8972857903139966, "grad_norm": 0.3170378506183624, "learning_rate": 2.3979905528819166e-08, "loss": 0.2006, "step": 14646 }, { "epoch": 3.8975518893028207, "grad_norm": 0.28277501463890076, "learning_rate": 2.3968822378097e-08, "loss": 0.187, "step": 14647 }, { "epoch": 3.8978179882916444, "grad_norm": 0.27887406945228577, "learning_rate": 2.3957741440445e-08, "loss": 0.1755, "step": 14648 }, { "epoch": 3.898084087280468, "grad_norm": 0.29673898220062256, "learning_rate": 2.3946662716185727e-08, "loss": 0.1912, "step": 14649 }, { "epoch": 3.898350186269292, "grad_norm": 0.3034519851207733, "learning_rate": 2.393558620564161e-08, "loss": 0.1954, "step": 14650 }, { "epoch": 3.898616285258116, "grad_norm": 0.3001004457473755, "learning_rate": 2.392451190913507e-08, "loss": 0.1752, "step": 14651 }, { "epoch": 3.89888238424694, "grad_norm": 0.29518958926200867, "learning_rate": 2.391343982698847e-08, "loss": 0.1842, "step": 14652 }, { "epoch": 3.8991484832357637, "grad_norm": 0.2955676019191742, "learning_rate": 2.390236995952407e-08, "loss": 0.1864, "step": 14653 }, { "epoch": 3.899414582224588, "grad_norm": 0.2932220995426178, "learning_rate": 2.389130230706411e-08, "loss": 0.1947, "step": 14654 }, { "epoch": 3.8996806812134115, "grad_norm": 0.3091318607330322, "learning_rate": 2.3880236869930738e-08, "loss": 0.2, "step": 14655 }, { "epoch": 3.899946780202235, "grad_norm": 0.2897256910800934, "learning_rate": 2.3869173648446006e-08, "loss": 0.1741, "step": 14656 }, { "epoch": 3.9002128791910593, "grad_norm": 0.351556658744812, "learning_rate": 2.385811264293196e-08, "loss": 0.1828, "step": 14657 }, { "epoch": 3.900478978179883, "grad_norm": 0.3267146646976471, "learning_rate": 2.3847053853710586e-08, "loss": 0.2104, "step": 14658 }, { "epoch": 3.9007450771687067, "grad_norm": 0.3131181001663208, "learning_rate": 2.3835997281103704e-08, "loss": 0.2049, "step": 14659 }, { "epoch": 3.901011176157531, "grad_norm": 0.31851527094841003, "learning_rate": 2.382494292543319e-08, "loss": 0.2037, "step": 14660 }, { "epoch": 3.9012772751463545, "grad_norm": 0.3052752614021301, "learning_rate": 2.3813890787020807e-08, "loss": 0.1913, "step": 14661 }, { "epoch": 3.901543374135178, "grad_norm": 0.2974010109901428, "learning_rate": 2.3802840866188247e-08, "loss": 0.1844, "step": 14662 }, { "epoch": 3.9018094731240023, "grad_norm": 0.29672595858573914, "learning_rate": 2.3791793163257136e-08, "loss": 0.1849, "step": 14663 }, { "epoch": 3.902075572112826, "grad_norm": 0.2768499255180359, "learning_rate": 2.378074767854905e-08, "loss": 0.1678, "step": 14664 }, { "epoch": 3.9023416711016496, "grad_norm": 0.280222624540329, "learning_rate": 2.3769704412385516e-08, "loss": 0.18, "step": 14665 }, { "epoch": 3.9026077700904738, "grad_norm": 0.27588918805122375, "learning_rate": 2.375866336508794e-08, "loss": 0.1834, "step": 14666 }, { "epoch": 3.9028738690792975, "grad_norm": 0.2883674204349518, "learning_rate": 2.374762453697775e-08, "loss": 0.183, "step": 14667 }, { "epoch": 3.903139968068121, "grad_norm": 0.290966659784317, "learning_rate": 2.3736587928376195e-08, "loss": 0.1821, "step": 14668 }, { "epoch": 3.9034060670569453, "grad_norm": 0.2932261526584625, "learning_rate": 2.372555353960455e-08, "loss": 0.185, "step": 14669 }, { "epoch": 3.903672166045769, "grad_norm": 0.30107584595680237, "learning_rate": 2.3714521370983997e-08, "loss": 0.2023, "step": 14670 }, { "epoch": 3.9039382650345926, "grad_norm": 0.2732861638069153, "learning_rate": 2.3703491422835653e-08, "loss": 0.1744, "step": 14671 }, { "epoch": 3.9042043640234168, "grad_norm": 0.2580694556236267, "learning_rate": 2.3692463695480568e-08, "loss": 0.1705, "step": 14672 }, { "epoch": 3.9044704630122404, "grad_norm": 0.29167139530181885, "learning_rate": 2.3681438189239733e-08, "loss": 0.1926, "step": 14673 }, { "epoch": 3.9047365620010646, "grad_norm": 0.2975851893424988, "learning_rate": 2.367041490443411e-08, "loss": 0.1899, "step": 14674 }, { "epoch": 3.9050026609898882, "grad_norm": 0.2984503507614136, "learning_rate": 2.3659393841384488e-08, "loss": 0.1897, "step": 14675 }, { "epoch": 3.905268759978712, "grad_norm": 0.28744813799858093, "learning_rate": 2.3648375000411693e-08, "loss": 0.1832, "step": 14676 }, { "epoch": 3.905534858967536, "grad_norm": 0.3138996958732605, "learning_rate": 2.363735838183649e-08, "loss": 0.2165, "step": 14677 }, { "epoch": 3.9058009579563597, "grad_norm": 0.28623825311660767, "learning_rate": 2.362634398597948e-08, "loss": 0.1879, "step": 14678 }, { "epoch": 3.906067056945184, "grad_norm": 0.2862323522567749, "learning_rate": 2.36153318131613e-08, "loss": 0.2018, "step": 14679 }, { "epoch": 3.9063331559340075, "grad_norm": 0.29026085138320923, "learning_rate": 2.360432186370247e-08, "loss": 0.1731, "step": 14680 }, { "epoch": 3.906599254922831, "grad_norm": 0.3011835217475891, "learning_rate": 2.359331413792347e-08, "loss": 0.1767, "step": 14681 }, { "epoch": 3.9068653539116553, "grad_norm": 0.27902084589004517, "learning_rate": 2.358230863614472e-08, "loss": 0.1822, "step": 14682 }, { "epoch": 3.907131452900479, "grad_norm": 0.25621309876441956, "learning_rate": 2.3571305358686545e-08, "loss": 0.1682, "step": 14683 }, { "epoch": 3.9073975518893027, "grad_norm": 0.28960371017456055, "learning_rate": 2.3560304305869217e-08, "loss": 0.1827, "step": 14684 }, { "epoch": 3.907663650878127, "grad_norm": 0.2989315390586853, "learning_rate": 2.3549305478012994e-08, "loss": 0.1741, "step": 14685 }, { "epoch": 3.9079297498669505, "grad_norm": 0.3019208014011383, "learning_rate": 2.353830887543795e-08, "loss": 0.1888, "step": 14686 }, { "epoch": 3.908195848855774, "grad_norm": 0.2750704884529114, "learning_rate": 2.3527314498464212e-08, "loss": 0.1829, "step": 14687 }, { "epoch": 3.9084619478445983, "grad_norm": 0.30780288577079773, "learning_rate": 2.3516322347411788e-08, "loss": 0.1866, "step": 14688 }, { "epoch": 3.908728046833422, "grad_norm": 0.29605820775032043, "learning_rate": 2.3505332422600633e-08, "loss": 0.193, "step": 14689 }, { "epoch": 3.9089941458222457, "grad_norm": 0.30805760622024536, "learning_rate": 2.3494344724350633e-08, "loss": 0.1906, "step": 14690 }, { "epoch": 3.90926024481107, "grad_norm": 0.3299652636051178, "learning_rate": 2.3483359252981618e-08, "loss": 0.1998, "step": 14691 }, { "epoch": 3.9095263437998935, "grad_norm": 0.3010775148868561, "learning_rate": 2.3472376008813332e-08, "loss": 0.2016, "step": 14692 }, { "epoch": 3.909792442788717, "grad_norm": 0.3133675754070282, "learning_rate": 2.3461394992165486e-08, "loss": 0.194, "step": 14693 }, { "epoch": 3.9100585417775413, "grad_norm": 0.3024483621120453, "learning_rate": 2.3450416203357726e-08, "loss": 0.2107, "step": 14694 }, { "epoch": 3.910324640766365, "grad_norm": 0.32749179005622864, "learning_rate": 2.343943964270957e-08, "loss": 0.1764, "step": 14695 }, { "epoch": 3.9105907397551887, "grad_norm": 0.3064405620098114, "learning_rate": 2.3428465310540525e-08, "loss": 0.2039, "step": 14696 }, { "epoch": 3.910856838744013, "grad_norm": 0.2954529821872711, "learning_rate": 2.341749320717008e-08, "loss": 0.1836, "step": 14697 }, { "epoch": 3.9111229377328365, "grad_norm": 0.3158511221408844, "learning_rate": 2.3406523332917528e-08, "loss": 0.1989, "step": 14698 }, { "epoch": 3.9113890367216606, "grad_norm": 0.3072682321071625, "learning_rate": 2.339555568810221e-08, "loss": 0.193, "step": 14699 }, { "epoch": 3.9116551357104843, "grad_norm": 0.29993489384651184, "learning_rate": 2.338459027304336e-08, "loss": 0.1912, "step": 14700 }, { "epoch": 3.9119212346993084, "grad_norm": 0.27926111221313477, "learning_rate": 2.3373627088060167e-08, "loss": 0.1873, "step": 14701 }, { "epoch": 3.912187333688132, "grad_norm": 0.3018615245819092, "learning_rate": 2.336266613347172e-08, "loss": 0.2031, "step": 14702 }, { "epoch": 3.9124534326769558, "grad_norm": 0.3302430510520935, "learning_rate": 2.335170740959711e-08, "loss": 0.1968, "step": 14703 }, { "epoch": 3.91271953166578, "grad_norm": 0.32320114970207214, "learning_rate": 2.334075091675526e-08, "loss": 0.2122, "step": 14704 }, { "epoch": 3.9129856306546036, "grad_norm": 0.30366408824920654, "learning_rate": 2.33297966552651e-08, "loss": 0.2083, "step": 14705 }, { "epoch": 3.9132517296434273, "grad_norm": 0.2784714102745056, "learning_rate": 2.3318844625445487e-08, "loss": 0.171, "step": 14706 }, { "epoch": 3.9135178286322514, "grad_norm": 0.29426509141921997, "learning_rate": 2.3307894827615205e-08, "loss": 0.1812, "step": 14707 }, { "epoch": 3.913783927621075, "grad_norm": 0.3059958815574646, "learning_rate": 2.329694726209298e-08, "loss": 0.1774, "step": 14708 }, { "epoch": 3.9140500266098988, "grad_norm": 0.29752200841903687, "learning_rate": 2.3286001929197464e-08, "loss": 0.1956, "step": 14709 }, { "epoch": 3.914316125598723, "grad_norm": 0.29609715938568115, "learning_rate": 2.327505882924724e-08, "loss": 0.1932, "step": 14710 }, { "epoch": 3.9145822245875466, "grad_norm": 0.27776533365249634, "learning_rate": 2.3264117962560836e-08, "loss": 0.1772, "step": 14711 }, { "epoch": 3.9148483235763702, "grad_norm": 0.2847668528556824, "learning_rate": 2.325317932945675e-08, "loss": 0.1696, "step": 14712 }, { "epoch": 3.9151144225651944, "grad_norm": 0.29491639137268066, "learning_rate": 2.3242242930253307e-08, "loss": 0.1976, "step": 14713 }, { "epoch": 3.915380521554018, "grad_norm": 0.2987893223762512, "learning_rate": 2.3231308765268887e-08, "loss": 0.1981, "step": 14714 }, { "epoch": 3.9156466205428417, "grad_norm": 0.3191193640232086, "learning_rate": 2.322037683482173e-08, "loss": 0.2066, "step": 14715 }, { "epoch": 3.915912719531666, "grad_norm": 0.3058733344078064, "learning_rate": 2.3209447139230078e-08, "loss": 0.2095, "step": 14716 }, { "epoch": 3.9161788185204895, "grad_norm": 0.28585562109947205, "learning_rate": 2.3198519678812012e-08, "loss": 0.1873, "step": 14717 }, { "epoch": 3.9164449175093132, "grad_norm": 0.2853280007839203, "learning_rate": 2.3187594453885618e-08, "loss": 0.1793, "step": 14718 }, { "epoch": 3.9167110164981374, "grad_norm": 0.28960710763931274, "learning_rate": 2.3176671464768914e-08, "loss": 0.1895, "step": 14719 }, { "epoch": 3.916977115486961, "grad_norm": 0.30210477113723755, "learning_rate": 2.316575071177983e-08, "loss": 0.1885, "step": 14720 }, { "epoch": 3.9172432144757847, "grad_norm": 0.29570192098617554, "learning_rate": 2.3154832195236274e-08, "loss": 0.1858, "step": 14721 }, { "epoch": 3.917509313464609, "grad_norm": 0.31312933564186096, "learning_rate": 2.3143915915455992e-08, "loss": 0.2012, "step": 14722 }, { "epoch": 3.9177754124534325, "grad_norm": 0.3148115873336792, "learning_rate": 2.3133001872756775e-08, "loss": 0.1811, "step": 14723 }, { "epoch": 3.9180415114422567, "grad_norm": 0.29014575481414795, "learning_rate": 2.312209006745628e-08, "loss": 0.1773, "step": 14724 }, { "epoch": 3.9183076104310803, "grad_norm": 0.2889425456523895, "learning_rate": 2.311118049987213e-08, "loss": 0.1929, "step": 14725 }, { "epoch": 3.9185737094199045, "grad_norm": 0.275583416223526, "learning_rate": 2.310027317032188e-08, "loss": 0.178, "step": 14726 }, { "epoch": 3.918839808408728, "grad_norm": 0.3055575489997864, "learning_rate": 2.3089368079123016e-08, "loss": 0.1943, "step": 14727 }, { "epoch": 3.919105907397552, "grad_norm": 0.2812022864818573, "learning_rate": 2.3078465226592946e-08, "loss": 0.1775, "step": 14728 }, { "epoch": 3.919372006386376, "grad_norm": 0.2986070215702057, "learning_rate": 2.3067564613049028e-08, "loss": 0.1848, "step": 14729 }, { "epoch": 3.9196381053751996, "grad_norm": 0.3392167389392853, "learning_rate": 2.3056666238808577e-08, "loss": 0.201, "step": 14730 }, { "epoch": 3.9199042043640233, "grad_norm": 0.323230117559433, "learning_rate": 2.3045770104188768e-08, "loss": 0.1908, "step": 14731 }, { "epoch": 3.9201703033528474, "grad_norm": 0.3065550625324249, "learning_rate": 2.303487620950677e-08, "loss": 0.2116, "step": 14732 }, { "epoch": 3.920436402341671, "grad_norm": 0.29861611127853394, "learning_rate": 2.3023984555079702e-08, "loss": 0.177, "step": 14733 }, { "epoch": 3.920702501330495, "grad_norm": 0.31751999258995056, "learning_rate": 2.301309514122457e-08, "loss": 0.1902, "step": 14734 }, { "epoch": 3.920968600319319, "grad_norm": 0.31570470333099365, "learning_rate": 2.3002207968258347e-08, "loss": 0.2027, "step": 14735 }, { "epoch": 3.9212346993081426, "grad_norm": 0.3135336935520172, "learning_rate": 2.2991323036497956e-08, "loss": 0.182, "step": 14736 }, { "epoch": 3.9215007982969663, "grad_norm": 0.2977560758590698, "learning_rate": 2.2980440346260166e-08, "loss": 0.1904, "step": 14737 }, { "epoch": 3.9217668972857904, "grad_norm": 0.31028828024864197, "learning_rate": 2.2969559897861778e-08, "loss": 0.178, "step": 14738 }, { "epoch": 3.922032996274614, "grad_norm": 0.31100085377693176, "learning_rate": 2.295868169161953e-08, "loss": 0.1979, "step": 14739 }, { "epoch": 3.922299095263438, "grad_norm": 0.2778160870075226, "learning_rate": 2.294780572784999e-08, "loss": 0.1796, "step": 14740 }, { "epoch": 3.922565194252262, "grad_norm": 0.3055420517921448, "learning_rate": 2.293693200686976e-08, "loss": 0.1948, "step": 14741 }, { "epoch": 3.9228312932410856, "grad_norm": 0.3121505677700043, "learning_rate": 2.292606052899535e-08, "loss": 0.1836, "step": 14742 }, { "epoch": 3.9230973922299093, "grad_norm": 0.2879633605480194, "learning_rate": 2.2915191294543202e-08, "loss": 0.169, "step": 14743 }, { "epoch": 3.9233634912187334, "grad_norm": 0.31161752343177795, "learning_rate": 2.290432430382968e-08, "loss": 0.1973, "step": 14744 }, { "epoch": 3.923629590207557, "grad_norm": 0.2937160134315491, "learning_rate": 2.2893459557171102e-08, "loss": 0.1862, "step": 14745 }, { "epoch": 3.923895689196381, "grad_norm": 0.2985934615135193, "learning_rate": 2.2882597054883703e-08, "loss": 0.1832, "step": 14746 }, { "epoch": 3.924161788185205, "grad_norm": 0.3321644365787506, "learning_rate": 2.287173679728368e-08, "loss": 0.2095, "step": 14747 }, { "epoch": 3.9244278871740286, "grad_norm": 0.3774718940258026, "learning_rate": 2.2860878784687166e-08, "loss": 0.1758, "step": 14748 }, { "epoch": 3.9246939861628527, "grad_norm": 0.29856765270233154, "learning_rate": 2.2850023017410147e-08, "loss": 0.1829, "step": 14749 }, { "epoch": 3.9249600851516764, "grad_norm": 0.3122999668121338, "learning_rate": 2.283916949576864e-08, "loss": 0.1892, "step": 14750 }, { "epoch": 3.9252261841405005, "grad_norm": 0.30384573340415955, "learning_rate": 2.282831822007857e-08, "loss": 0.2009, "step": 14751 }, { "epoch": 3.925492283129324, "grad_norm": 0.2872238755226135, "learning_rate": 2.281746919065578e-08, "loss": 0.1805, "step": 14752 }, { "epoch": 3.925758382118148, "grad_norm": 0.33381402492523193, "learning_rate": 2.2806622407816055e-08, "loss": 0.2154, "step": 14753 }, { "epoch": 3.926024481106972, "grad_norm": 0.29060155153274536, "learning_rate": 2.2795777871875123e-08, "loss": 0.1961, "step": 14754 }, { "epoch": 3.9262905800957957, "grad_norm": 0.29178935289382935, "learning_rate": 2.2784935583148645e-08, "loss": 0.1791, "step": 14755 }, { "epoch": 3.9265566790846194, "grad_norm": 0.2985253632068634, "learning_rate": 2.277409554195222e-08, "loss": 0.1932, "step": 14756 }, { "epoch": 3.9268227780734435, "grad_norm": 0.27982935309410095, "learning_rate": 2.276325774860134e-08, "loss": 0.1723, "step": 14757 }, { "epoch": 3.927088877062267, "grad_norm": 0.3171539306640625, "learning_rate": 2.275242220341148e-08, "loss": 0.2247, "step": 14758 }, { "epoch": 3.927354976051091, "grad_norm": 0.2826724052429199, "learning_rate": 2.2741588906698074e-08, "loss": 0.1931, "step": 14759 }, { "epoch": 3.927621075039915, "grad_norm": 0.31923016905784607, "learning_rate": 2.2730757858776383e-08, "loss": 0.1952, "step": 14760 }, { "epoch": 3.9278871740287387, "grad_norm": 0.2956431806087494, "learning_rate": 2.2719929059961695e-08, "loss": 0.192, "step": 14761 }, { "epoch": 3.9281532730175623, "grad_norm": 0.29637205600738525, "learning_rate": 2.2709102510569234e-08, "loss": 0.1973, "step": 14762 }, { "epoch": 3.9284193720063865, "grad_norm": 0.29453906416893005, "learning_rate": 2.26982782109141e-08, "loss": 0.2006, "step": 14763 }, { "epoch": 3.92868547099521, "grad_norm": 0.3033652901649475, "learning_rate": 2.268745616131138e-08, "loss": 0.2007, "step": 14764 }, { "epoch": 3.928951569984034, "grad_norm": 0.2585514485836029, "learning_rate": 2.2676636362076074e-08, "loss": 0.1821, "step": 14765 }, { "epoch": 3.929217668972858, "grad_norm": 0.3005974590778351, "learning_rate": 2.266581881352314e-08, "loss": 0.1853, "step": 14766 }, { "epoch": 3.9294837679616816, "grad_norm": 0.2905030846595764, "learning_rate": 2.2655003515967398e-08, "loss": 0.1969, "step": 14767 }, { "epoch": 3.9297498669505053, "grad_norm": 0.3246970474720001, "learning_rate": 2.2644190469723678e-08, "loss": 0.2074, "step": 14768 }, { "epoch": 3.9300159659393294, "grad_norm": 0.3023689389228821, "learning_rate": 2.2633379675106713e-08, "loss": 0.2076, "step": 14769 }, { "epoch": 3.930282064928153, "grad_norm": 0.28206154704093933, "learning_rate": 2.26225711324312e-08, "loss": 0.1817, "step": 14770 }, { "epoch": 3.9305481639169773, "grad_norm": 0.2665962278842926, "learning_rate": 2.2611764842011728e-08, "loss": 0.1853, "step": 14771 }, { "epoch": 3.930814262905801, "grad_norm": 0.3083944618701935, "learning_rate": 2.260096080416284e-08, "loss": 0.203, "step": 14772 }, { "epoch": 3.931080361894625, "grad_norm": 0.32900893688201904, "learning_rate": 2.2590159019199018e-08, "loss": 0.2153, "step": 14773 }, { "epoch": 3.9313464608834487, "grad_norm": 0.2953045070171356, "learning_rate": 2.257935948743468e-08, "loss": 0.2104, "step": 14774 }, { "epoch": 3.9316125598722724, "grad_norm": 0.28784769773483276, "learning_rate": 2.2568562209184194e-08, "loss": 0.1947, "step": 14775 }, { "epoch": 3.9318786588610966, "grad_norm": 0.29624179005622864, "learning_rate": 2.255776718476178e-08, "loss": 0.1951, "step": 14776 }, { "epoch": 3.9321447578499202, "grad_norm": 0.3023116886615753, "learning_rate": 2.254697441448169e-08, "loss": 0.1919, "step": 14777 }, { "epoch": 3.932410856838744, "grad_norm": 0.2878435254096985, "learning_rate": 2.25361838986581e-08, "loss": 0.1938, "step": 14778 }, { "epoch": 3.932676955827568, "grad_norm": 0.3125193119049072, "learning_rate": 2.252539563760504e-08, "loss": 0.2029, "step": 14779 }, { "epoch": 3.9329430548163917, "grad_norm": 0.2739975154399872, "learning_rate": 2.251460963163654e-08, "loss": 0.1902, "step": 14780 }, { "epoch": 3.9332091538052154, "grad_norm": 0.3185179829597473, "learning_rate": 2.2503825881066574e-08, "loss": 0.2093, "step": 14781 }, { "epoch": 3.9334752527940395, "grad_norm": 0.27942556142807007, "learning_rate": 2.2493044386209016e-08, "loss": 0.1813, "step": 14782 }, { "epoch": 3.933741351782863, "grad_norm": 0.3015945255756378, "learning_rate": 2.248226514737769e-08, "loss": 0.179, "step": 14783 }, { "epoch": 3.934007450771687, "grad_norm": 0.3182271122932434, "learning_rate": 2.2471488164886388e-08, "loss": 0.1917, "step": 14784 }, { "epoch": 3.934273549760511, "grad_norm": 0.27184367179870605, "learning_rate": 2.2460713439048728e-08, "loss": 0.1695, "step": 14785 }, { "epoch": 3.9345396487493347, "grad_norm": 0.28779810667037964, "learning_rate": 2.2449940970178382e-08, "loss": 0.1876, "step": 14786 }, { "epoch": 3.9348057477381584, "grad_norm": 0.3066202402114868, "learning_rate": 2.2439170758588888e-08, "loss": 0.2013, "step": 14787 }, { "epoch": 3.9350718467269825, "grad_norm": 0.30861496925354004, "learning_rate": 2.2428402804593748e-08, "loss": 0.1959, "step": 14788 }, { "epoch": 3.935337945715806, "grad_norm": 0.29449450969696045, "learning_rate": 2.2417637108506392e-08, "loss": 0.1961, "step": 14789 }, { "epoch": 3.93560404470463, "grad_norm": 0.2807082235813141, "learning_rate": 2.240687367064018e-08, "loss": 0.1829, "step": 14790 }, { "epoch": 3.935870143693454, "grad_norm": 0.3031291365623474, "learning_rate": 2.2396112491308405e-08, "loss": 0.2005, "step": 14791 }, { "epoch": 3.9361362426822777, "grad_norm": 0.30863046646118164, "learning_rate": 2.2385353570824306e-08, "loss": 0.1844, "step": 14792 }, { "epoch": 3.936402341671102, "grad_norm": 0.3008638620376587, "learning_rate": 2.2374596909501065e-08, "loss": 0.198, "step": 14793 }, { "epoch": 3.9366684406599255, "grad_norm": 0.31276652216911316, "learning_rate": 2.236384250765172e-08, "loss": 0.1959, "step": 14794 }, { "epoch": 3.936934539648749, "grad_norm": 0.3026772737503052, "learning_rate": 2.2353090365589344e-08, "loss": 0.1816, "step": 14795 }, { "epoch": 3.9372006386375733, "grad_norm": 0.2948720157146454, "learning_rate": 2.234234048362691e-08, "loss": 0.1865, "step": 14796 }, { "epoch": 3.937466737626397, "grad_norm": 0.31698474287986755, "learning_rate": 2.2331592862077298e-08, "loss": 0.2088, "step": 14797 }, { "epoch": 3.937732836615221, "grad_norm": 0.2906666398048401, "learning_rate": 2.232084750125338e-08, "loss": 0.1736, "step": 14798 }, { "epoch": 3.937998935604045, "grad_norm": 0.30383479595184326, "learning_rate": 2.2310104401467868e-08, "loss": 0.1973, "step": 14799 }, { "epoch": 3.9382650345928685, "grad_norm": 0.2983381748199463, "learning_rate": 2.22993635630335e-08, "loss": 0.1992, "step": 14800 }, { "epoch": 3.9385311335816926, "grad_norm": 0.2807394862174988, "learning_rate": 2.2288624986262917e-08, "loss": 0.1871, "step": 14801 }, { "epoch": 3.9387972325705163, "grad_norm": 0.283086895942688, "learning_rate": 2.2277888671468713e-08, "loss": 0.17, "step": 14802 }, { "epoch": 3.93906333155934, "grad_norm": 0.2948250472545624, "learning_rate": 2.2267154618963334e-08, "loss": 0.183, "step": 14803 }, { "epoch": 3.939329430548164, "grad_norm": 0.31266650557518005, "learning_rate": 2.2256422829059252e-08, "loss": 0.1987, "step": 14804 }, { "epoch": 3.9395955295369878, "grad_norm": 0.29915300011634827, "learning_rate": 2.2245693302068848e-08, "loss": 0.2019, "step": 14805 }, { "epoch": 3.9398616285258115, "grad_norm": 0.2940346598625183, "learning_rate": 2.2234966038304427e-08, "loss": 0.1808, "step": 14806 }, { "epoch": 3.9401277275146356, "grad_norm": 0.3038140535354614, "learning_rate": 2.222424103807822e-08, "loss": 0.1998, "step": 14807 }, { "epoch": 3.9403938265034593, "grad_norm": 0.3083913028240204, "learning_rate": 2.2213518301702417e-08, "loss": 0.1892, "step": 14808 }, { "epoch": 3.940659925492283, "grad_norm": 0.2799878716468811, "learning_rate": 2.220279782948913e-08, "loss": 0.1859, "step": 14809 }, { "epoch": 3.940926024481107, "grad_norm": 0.334277868270874, "learning_rate": 2.2192079621750393e-08, "loss": 0.2023, "step": 14810 }, { "epoch": 3.9411921234699308, "grad_norm": 0.2890993654727936, "learning_rate": 2.2181363678798226e-08, "loss": 0.1908, "step": 14811 }, { "epoch": 3.9414582224587544, "grad_norm": 0.28793928027153015, "learning_rate": 2.217065000094448e-08, "loss": 0.1855, "step": 14812 }, { "epoch": 3.9417243214475786, "grad_norm": 0.2712806761264801, "learning_rate": 2.2159938588501026e-08, "loss": 0.1904, "step": 14813 }, { "epoch": 3.9419904204364022, "grad_norm": 0.31120872497558594, "learning_rate": 2.2149229441779648e-08, "loss": 0.2052, "step": 14814 }, { "epoch": 3.942256519425226, "grad_norm": 0.287064790725708, "learning_rate": 2.2138522561092065e-08, "loss": 0.1744, "step": 14815 }, { "epoch": 3.94252261841405, "grad_norm": 0.3269481062889099, "learning_rate": 2.2127817946749927e-08, "loss": 0.1985, "step": 14816 }, { "epoch": 3.9427887174028737, "grad_norm": 0.31452715396881104, "learning_rate": 2.2117115599064852e-08, "loss": 0.2008, "step": 14817 }, { "epoch": 3.943054816391698, "grad_norm": 0.3015950918197632, "learning_rate": 2.2106415518348287e-08, "loss": 0.189, "step": 14818 }, { "epoch": 3.9433209153805215, "grad_norm": 0.308654248714447, "learning_rate": 2.2095717704911708e-08, "loss": 0.2033, "step": 14819 }, { "epoch": 3.9435870143693457, "grad_norm": 0.29697591066360474, "learning_rate": 2.2085022159066547e-08, "loss": 0.1902, "step": 14820 }, { "epoch": 3.9438531133581693, "grad_norm": 0.28719449043273926, "learning_rate": 2.207432888112406e-08, "loss": 0.1859, "step": 14821 }, { "epoch": 3.944119212346993, "grad_norm": 0.2782399356365204, "learning_rate": 2.2063637871395526e-08, "loss": 0.1763, "step": 14822 }, { "epoch": 3.944385311335817, "grad_norm": 0.3232516646385193, "learning_rate": 2.2052949130192133e-08, "loss": 0.2004, "step": 14823 }, { "epoch": 3.944651410324641, "grad_norm": 0.3099755346775055, "learning_rate": 2.2042262657825004e-08, "loss": 0.1841, "step": 14824 }, { "epoch": 3.9449175093134645, "grad_norm": 0.32218316197395325, "learning_rate": 2.2031578454605193e-08, "loss": 0.1995, "step": 14825 }, { "epoch": 3.9451836083022886, "grad_norm": 0.3349783420562744, "learning_rate": 2.2020896520843692e-08, "loss": 0.2156, "step": 14826 }, { "epoch": 3.9454497072911123, "grad_norm": 0.30727240443229675, "learning_rate": 2.2010216856851415e-08, "loss": 0.1963, "step": 14827 }, { "epoch": 3.945715806279936, "grad_norm": 0.3031664788722992, "learning_rate": 2.199953946293923e-08, "loss": 0.1731, "step": 14828 }, { "epoch": 3.94598190526876, "grad_norm": 0.3175657391548157, "learning_rate": 2.1988864339417933e-08, "loss": 0.2028, "step": 14829 }, { "epoch": 3.946248004257584, "grad_norm": 0.30555564165115356, "learning_rate": 2.1978191486598264e-08, "loss": 0.184, "step": 14830 }, { "epoch": 3.9465141032464075, "grad_norm": 0.30346763134002686, "learning_rate": 2.1967520904790825e-08, "loss": 0.1919, "step": 14831 }, { "epoch": 3.9467802022352316, "grad_norm": 0.2763991951942444, "learning_rate": 2.1956852594306253e-08, "loss": 0.187, "step": 14832 }, { "epoch": 3.9470463012240553, "grad_norm": 0.3146800100803375, "learning_rate": 2.194618655545506e-08, "loss": 0.2075, "step": 14833 }, { "epoch": 3.947312400212879, "grad_norm": 0.2945611774921417, "learning_rate": 2.193552278854771e-08, "loss": 0.1897, "step": 14834 }, { "epoch": 3.947578499201703, "grad_norm": 0.28725528717041016, "learning_rate": 2.1924861293894603e-08, "loss": 0.201, "step": 14835 }, { "epoch": 3.947844598190527, "grad_norm": 0.2939136028289795, "learning_rate": 2.1914202071806065e-08, "loss": 0.2006, "step": 14836 }, { "epoch": 3.9481106971793505, "grad_norm": 0.2872926890850067, "learning_rate": 2.190354512259238e-08, "loss": 0.1939, "step": 14837 }, { "epoch": 3.9483767961681746, "grad_norm": 0.3106127977371216, "learning_rate": 2.1892890446563695e-08, "loss": 0.1984, "step": 14838 }, { "epoch": 3.9486428951569983, "grad_norm": 0.313979834318161, "learning_rate": 2.1882238044030166e-08, "loss": 0.1929, "step": 14839 }, { "epoch": 3.948908994145822, "grad_norm": 0.2898484170436859, "learning_rate": 2.1871587915301893e-08, "loss": 0.1743, "step": 14840 }, { "epoch": 3.949175093134646, "grad_norm": 0.29106807708740234, "learning_rate": 2.1860940060688803e-08, "loss": 0.2007, "step": 14841 }, { "epoch": 3.94944119212347, "grad_norm": 0.28573077917099, "learning_rate": 2.185029448050085e-08, "loss": 0.1613, "step": 14842 }, { "epoch": 3.949707291112294, "grad_norm": 0.3079259693622589, "learning_rate": 2.1839651175047925e-08, "loss": 0.1965, "step": 14843 }, { "epoch": 3.9499733901011176, "grad_norm": 0.3238368332386017, "learning_rate": 2.1829010144639816e-08, "loss": 0.1981, "step": 14844 }, { "epoch": 3.9502394890899417, "grad_norm": 0.2965106666088104, "learning_rate": 2.1818371389586243e-08, "loss": 0.1996, "step": 14845 }, { "epoch": 3.9505055880787654, "grad_norm": 0.31944921612739563, "learning_rate": 2.1807734910196886e-08, "loss": 0.2112, "step": 14846 }, { "epoch": 3.950771687067589, "grad_norm": 0.29262152314186096, "learning_rate": 2.1797100706781346e-08, "loss": 0.1828, "step": 14847 }, { "epoch": 3.951037786056413, "grad_norm": 0.31182876229286194, "learning_rate": 2.178646877964917e-08, "loss": 0.1947, "step": 14848 }, { "epoch": 3.951303885045237, "grad_norm": 0.3086685240268707, "learning_rate": 2.1775839129109785e-08, "loss": 0.203, "step": 14849 }, { "epoch": 3.9515699840340606, "grad_norm": 0.2902146577835083, "learning_rate": 2.1765211755472624e-08, "loss": 0.1839, "step": 14850 }, { "epoch": 3.9518360830228847, "grad_norm": 0.27675989270210266, "learning_rate": 2.1754586659047003e-08, "loss": 0.1743, "step": 14851 }, { "epoch": 3.9521021820117084, "grad_norm": 0.29738250374794006, "learning_rate": 2.174396384014221e-08, "loss": 0.1828, "step": 14852 }, { "epoch": 3.952368281000532, "grad_norm": 0.2915045917034149, "learning_rate": 2.1733343299067442e-08, "loss": 0.1878, "step": 14853 }, { "epoch": 3.952634379989356, "grad_norm": 0.29369038343429565, "learning_rate": 2.172272503613183e-08, "loss": 0.1993, "step": 14854 }, { "epoch": 3.95290047897818, "grad_norm": 0.292983740568161, "learning_rate": 2.1712109051644434e-08, "loss": 0.2022, "step": 14855 }, { "epoch": 3.9531665779670035, "grad_norm": 0.29378947615623474, "learning_rate": 2.170149534591429e-08, "loss": 0.1871, "step": 14856 }, { "epoch": 3.9534326769558277, "grad_norm": 0.3019186556339264, "learning_rate": 2.1690883919250336e-08, "loss": 0.1837, "step": 14857 }, { "epoch": 3.9536987759446514, "grad_norm": 0.2809322774410248, "learning_rate": 2.1680274771961394e-08, "loss": 0.1703, "step": 14858 }, { "epoch": 3.953964874933475, "grad_norm": 0.29842761158943176, "learning_rate": 2.1669667904356327e-08, "loss": 0.1922, "step": 14859 }, { "epoch": 3.954230973922299, "grad_norm": 0.3592434227466583, "learning_rate": 2.165906331674382e-08, "loss": 0.1924, "step": 14860 }, { "epoch": 3.954497072911123, "grad_norm": 0.30814129114151, "learning_rate": 2.1648461009432573e-08, "loss": 0.2029, "step": 14861 }, { "epoch": 3.9547631718999465, "grad_norm": 0.32105058431625366, "learning_rate": 2.1637860982731182e-08, "loss": 0.2131, "step": 14862 }, { "epoch": 3.9550292708887707, "grad_norm": 0.2723998427391052, "learning_rate": 2.162726323694819e-08, "loss": 0.1914, "step": 14863 }, { "epoch": 3.9552953698775943, "grad_norm": 0.2969737648963928, "learning_rate": 2.161666777239207e-08, "loss": 0.1947, "step": 14864 }, { "epoch": 3.9555614688664185, "grad_norm": 0.30160558223724365, "learning_rate": 2.1606074589371248e-08, "loss": 0.1774, "step": 14865 }, { "epoch": 3.955827567855242, "grad_norm": 0.29838839173316956, "learning_rate": 2.159548368819406e-08, "loss": 0.186, "step": 14866 }, { "epoch": 3.9560936668440663, "grad_norm": 0.29601147770881653, "learning_rate": 2.158489506916874e-08, "loss": 0.1817, "step": 14867 }, { "epoch": 3.95635976583289, "grad_norm": 0.2932482659816742, "learning_rate": 2.157430873260353e-08, "loss": 0.1906, "step": 14868 }, { "epoch": 3.9566258648217136, "grad_norm": 0.307435542345047, "learning_rate": 2.156372467880655e-08, "loss": 0.1897, "step": 14869 }, { "epoch": 3.9568919638105378, "grad_norm": 0.2962658703327179, "learning_rate": 2.1553142908085898e-08, "loss": 0.1826, "step": 14870 }, { "epoch": 3.9571580627993614, "grad_norm": 0.3096923232078552, "learning_rate": 2.154256342074956e-08, "loss": 0.2199, "step": 14871 }, { "epoch": 3.957424161788185, "grad_norm": 0.2708991467952728, "learning_rate": 2.1531986217105502e-08, "loss": 0.189, "step": 14872 }, { "epoch": 3.9576902607770093, "grad_norm": 0.2837902903556824, "learning_rate": 2.152141129746158e-08, "loss": 0.1745, "step": 14873 }, { "epoch": 3.957956359765833, "grad_norm": 0.31054994463920593, "learning_rate": 2.1510838662125596e-08, "loss": 0.1864, "step": 14874 }, { "epoch": 3.9582224587546566, "grad_norm": 0.2959757149219513, "learning_rate": 2.1500268311405347e-08, "loss": 0.191, "step": 14875 }, { "epoch": 3.9584885577434807, "grad_norm": 0.2956463396549225, "learning_rate": 2.1489700245608432e-08, "loss": 0.1817, "step": 14876 }, { "epoch": 3.9587546567323044, "grad_norm": 0.30454152822494507, "learning_rate": 2.1479134465042492e-08, "loss": 0.1904, "step": 14877 }, { "epoch": 3.959020755721128, "grad_norm": 0.28253957629203796, "learning_rate": 2.1468570970015064e-08, "loss": 0.1728, "step": 14878 }, { "epoch": 3.9592868547099522, "grad_norm": 0.28500470519065857, "learning_rate": 2.1458009760833662e-08, "loss": 0.1913, "step": 14879 }, { "epoch": 3.959552953698776, "grad_norm": 0.2906327545642853, "learning_rate": 2.1447450837805636e-08, "loss": 0.1726, "step": 14880 }, { "epoch": 3.9598190526875996, "grad_norm": 0.3066767752170563, "learning_rate": 2.1436894201238364e-08, "loss": 0.2029, "step": 14881 }, { "epoch": 3.9600851516764237, "grad_norm": 0.31095317006111145, "learning_rate": 2.1426339851439112e-08, "loss": 0.195, "step": 14882 }, { "epoch": 3.9603512506652474, "grad_norm": 0.2946455478668213, "learning_rate": 2.141578778871509e-08, "loss": 0.1929, "step": 14883 }, { "epoch": 3.960617349654071, "grad_norm": 0.2954576015472412, "learning_rate": 2.140523801337346e-08, "loss": 0.1944, "step": 14884 }, { "epoch": 3.960883448642895, "grad_norm": 0.30496540665626526, "learning_rate": 2.139469052572127e-08, "loss": 0.1944, "step": 14885 }, { "epoch": 3.961149547631719, "grad_norm": 0.282336562871933, "learning_rate": 2.138414532606554e-08, "loss": 0.1655, "step": 14886 }, { "epoch": 3.9614156466205426, "grad_norm": 0.2742499113082886, "learning_rate": 2.137360241471321e-08, "loss": 0.1799, "step": 14887 }, { "epoch": 3.9616817456093667, "grad_norm": 0.3095497786998749, "learning_rate": 2.136306179197117e-08, "loss": 0.1809, "step": 14888 }, { "epoch": 3.9619478445981904, "grad_norm": 0.29259851574897766, "learning_rate": 2.135252345814621e-08, "loss": 0.1836, "step": 14889 }, { "epoch": 3.9622139435870145, "grad_norm": 0.28284457325935364, "learning_rate": 2.1341987413545105e-08, "loss": 0.1739, "step": 14890 }, { "epoch": 3.962480042575838, "grad_norm": 0.28210723400115967, "learning_rate": 2.1331453658474496e-08, "loss": 0.1844, "step": 14891 }, { "epoch": 3.9627461415646623, "grad_norm": 0.2880859076976776, "learning_rate": 2.132092219324102e-08, "loss": 0.1988, "step": 14892 }, { "epoch": 3.963012240553486, "grad_norm": 0.2971457839012146, "learning_rate": 2.1310393018151228e-08, "loss": 0.2029, "step": 14893 }, { "epoch": 3.9632783395423097, "grad_norm": 0.30179211497306824, "learning_rate": 2.1299866133511556e-08, "loss": 0.1948, "step": 14894 }, { "epoch": 3.963544438531134, "grad_norm": 0.2978350818157196, "learning_rate": 2.1289341539628447e-08, "loss": 0.1783, "step": 14895 }, { "epoch": 3.9638105375199575, "grad_norm": 0.2900627851486206, "learning_rate": 2.127881923680823e-08, "loss": 0.1852, "step": 14896 }, { "epoch": 3.964076636508781, "grad_norm": 0.2947036325931549, "learning_rate": 2.1268299225357177e-08, "loss": 0.1803, "step": 14897 }, { "epoch": 3.9643427354976053, "grad_norm": 0.3077940046787262, "learning_rate": 2.125778150558153e-08, "loss": 0.189, "step": 14898 }, { "epoch": 3.964608834486429, "grad_norm": 0.31427833437919617, "learning_rate": 2.1247266077787417e-08, "loss": 0.2044, "step": 14899 }, { "epoch": 3.9648749334752527, "grad_norm": 0.3048447370529175, "learning_rate": 2.12367529422809e-08, "loss": 0.1924, "step": 14900 }, { "epoch": 3.965141032464077, "grad_norm": 0.33553794026374817, "learning_rate": 2.1226242099367985e-08, "loss": 0.2047, "step": 14901 }, { "epoch": 3.9654071314529005, "grad_norm": 0.28232651948928833, "learning_rate": 2.1215733549354663e-08, "loss": 0.1707, "step": 14902 }, { "epoch": 3.965673230441724, "grad_norm": 0.2856953740119934, "learning_rate": 2.1205227292546747e-08, "loss": 0.1886, "step": 14903 }, { "epoch": 3.9659393294305483, "grad_norm": 0.3013323247432709, "learning_rate": 2.1194723329250074e-08, "loss": 0.203, "step": 14904 }, { "epoch": 3.966205428419372, "grad_norm": 0.2853599488735199, "learning_rate": 2.1184221659770384e-08, "loss": 0.1882, "step": 14905 }, { "epoch": 3.9664715274081956, "grad_norm": 0.2893584370613098, "learning_rate": 2.1173722284413377e-08, "loss": 0.181, "step": 14906 }, { "epoch": 3.9667376263970198, "grad_norm": 0.3267238140106201, "learning_rate": 2.1163225203484625e-08, "loss": 0.21, "step": 14907 }, { "epoch": 3.9670037253858434, "grad_norm": 0.3027190864086151, "learning_rate": 2.11527304172897e-08, "loss": 0.1957, "step": 14908 }, { "epoch": 3.967269824374667, "grad_norm": 0.31064340472221375, "learning_rate": 2.1142237926134066e-08, "loss": 0.1935, "step": 14909 }, { "epoch": 3.9675359233634913, "grad_norm": 0.3064180314540863, "learning_rate": 2.1131747730323125e-08, "loss": 0.179, "step": 14910 }, { "epoch": 3.967802022352315, "grad_norm": 0.2870473563671112, "learning_rate": 2.1121259830162262e-08, "loss": 0.1748, "step": 14911 }, { "epoch": 3.968068121341139, "grad_norm": 0.3301200270652771, "learning_rate": 2.1110774225956696e-08, "loss": 0.2282, "step": 14912 }, { "epoch": 3.9683342203299627, "grad_norm": 0.3090327978134155, "learning_rate": 2.1100290918011655e-08, "loss": 0.1997, "step": 14913 }, { "epoch": 3.9686003193187864, "grad_norm": 0.30415078997612, "learning_rate": 2.1089809906632284e-08, "loss": 0.1807, "step": 14914 }, { "epoch": 3.9688664183076106, "grad_norm": 0.2875346541404724, "learning_rate": 2.1079331192123652e-08, "loss": 0.1802, "step": 14915 }, { "epoch": 3.9691325172964342, "grad_norm": 0.3214878737926483, "learning_rate": 2.106885477479078e-08, "loss": 0.2173, "step": 14916 }, { "epoch": 3.9693986162852584, "grad_norm": 0.2708263397216797, "learning_rate": 2.10583806549386e-08, "loss": 0.168, "step": 14917 }, { "epoch": 3.969664715274082, "grad_norm": 0.2978089153766632, "learning_rate": 2.104790883287202e-08, "loss": 0.2071, "step": 14918 }, { "epoch": 3.9699308142629057, "grad_norm": 0.30369725823402405, "learning_rate": 2.103743930889579e-08, "loss": 0.1807, "step": 14919 }, { "epoch": 3.97019691325173, "grad_norm": 0.2895180583000183, "learning_rate": 2.1026972083314675e-08, "loss": 0.2007, "step": 14920 }, { "epoch": 3.9704630122405535, "grad_norm": 0.3070825934410095, "learning_rate": 2.1016507156433384e-08, "loss": 0.201, "step": 14921 }, { "epoch": 3.970729111229377, "grad_norm": 0.30438464879989624, "learning_rate": 2.1006044528556454e-08, "loss": 0.1878, "step": 14922 }, { "epoch": 3.9709952102182013, "grad_norm": 0.3030012845993042, "learning_rate": 2.099558419998847e-08, "loss": 0.1659, "step": 14923 }, { "epoch": 3.971261309207025, "grad_norm": 0.2730826437473297, "learning_rate": 2.0985126171033895e-08, "loss": 0.1693, "step": 14924 }, { "epoch": 3.9715274081958487, "grad_norm": 0.30868273973464966, "learning_rate": 2.0974670441997133e-08, "loss": 0.1864, "step": 14925 }, { "epoch": 3.971793507184673, "grad_norm": 0.294010192155838, "learning_rate": 2.096421701318254e-08, "loss": 0.1945, "step": 14926 }, { "epoch": 3.9720596061734965, "grad_norm": 0.3128400444984436, "learning_rate": 2.095376588489436e-08, "loss": 0.1784, "step": 14927 }, { "epoch": 3.97232570516232, "grad_norm": 0.2905682921409607, "learning_rate": 2.0943317057436817e-08, "loss": 0.2026, "step": 14928 }, { "epoch": 3.9725918041511443, "grad_norm": 0.3068739175796509, "learning_rate": 2.0932870531114077e-08, "loss": 0.2086, "step": 14929 }, { "epoch": 3.972857903139968, "grad_norm": 0.27567338943481445, "learning_rate": 2.0922426306230158e-08, "loss": 0.183, "step": 14930 }, { "epoch": 3.9731240021287917, "grad_norm": 0.30362260341644287, "learning_rate": 2.0911984383089077e-08, "loss": 0.1869, "step": 14931 }, { "epoch": 3.973390101117616, "grad_norm": 0.3327155113220215, "learning_rate": 2.090154476199478e-08, "loss": 0.1932, "step": 14932 }, { "epoch": 3.9736562001064395, "grad_norm": 0.32339879870414734, "learning_rate": 2.0891107443251143e-08, "loss": 0.2143, "step": 14933 }, { "epoch": 3.973922299095263, "grad_norm": 0.287534236907959, "learning_rate": 2.088067242716196e-08, "loss": 0.1958, "step": 14934 }, { "epoch": 3.9741883980840873, "grad_norm": 0.30287089943885803, "learning_rate": 2.0870239714030978e-08, "loss": 0.2057, "step": 14935 }, { "epoch": 3.974454497072911, "grad_norm": 0.3146795630455017, "learning_rate": 2.0859809304161857e-08, "loss": 0.2014, "step": 14936 }, { "epoch": 3.974720596061735, "grad_norm": 0.28312742710113525, "learning_rate": 2.0849381197858197e-08, "loss": 0.1781, "step": 14937 }, { "epoch": 3.974986695050559, "grad_norm": 0.29034197330474854, "learning_rate": 2.0838955395423575e-08, "loss": 0.1814, "step": 14938 }, { "epoch": 3.975252794039383, "grad_norm": 0.29550960659980774, "learning_rate": 2.0828531897161383e-08, "loss": 0.1867, "step": 14939 }, { "epoch": 3.9755188930282066, "grad_norm": 0.29364922642707825, "learning_rate": 2.0818110703375068e-08, "loss": 0.1879, "step": 14940 }, { "epoch": 3.9757849920170303, "grad_norm": 0.2900139093399048, "learning_rate": 2.080769181436798e-08, "loss": 0.1943, "step": 14941 }, { "epoch": 3.9760510910058544, "grad_norm": 0.29829058051109314, "learning_rate": 2.0797275230443333e-08, "loss": 0.1872, "step": 14942 }, { "epoch": 3.976317189994678, "grad_norm": 0.30480581521987915, "learning_rate": 2.0786860951904362e-08, "loss": 0.199, "step": 14943 }, { "epoch": 3.9765832889835018, "grad_norm": 0.3105356693267822, "learning_rate": 2.0776448979054196e-08, "loss": 0.2038, "step": 14944 }, { "epoch": 3.976849387972326, "grad_norm": 0.30000513792037964, "learning_rate": 2.0766039312195894e-08, "loss": 0.183, "step": 14945 }, { "epoch": 3.9771154869611496, "grad_norm": 0.26569995284080505, "learning_rate": 2.0755631951632456e-08, "loss": 0.1686, "step": 14946 }, { "epoch": 3.9773815859499733, "grad_norm": 0.30085381865501404, "learning_rate": 2.0745226897666857e-08, "loss": 0.1858, "step": 14947 }, { "epoch": 3.9776476849387974, "grad_norm": 0.34696340560913086, "learning_rate": 2.0734824150601883e-08, "loss": 0.202, "step": 14948 }, { "epoch": 3.977913783927621, "grad_norm": 0.28678250312805176, "learning_rate": 2.0724423710740356e-08, "loss": 0.1857, "step": 14949 }, { "epoch": 3.9781798829164448, "grad_norm": 0.28772854804992676, "learning_rate": 2.071402557838503e-08, "loss": 0.1811, "step": 14950 }, { "epoch": 3.978445981905269, "grad_norm": 0.30966952443122864, "learning_rate": 2.070362975383856e-08, "loss": 0.197, "step": 14951 }, { "epoch": 3.9787120808940926, "grad_norm": 0.30130526423454285, "learning_rate": 2.0693236237403534e-08, "loss": 0.1878, "step": 14952 }, { "epoch": 3.9789781798829162, "grad_norm": 0.2770409882068634, "learning_rate": 2.068284502938248e-08, "loss": 0.1676, "step": 14953 }, { "epoch": 3.9792442788717404, "grad_norm": 0.29621535539627075, "learning_rate": 2.0672456130077865e-08, "loss": 0.1971, "step": 14954 }, { "epoch": 3.979510377860564, "grad_norm": 0.32010379433631897, "learning_rate": 2.066206953979207e-08, "loss": 0.1892, "step": 14955 }, { "epoch": 3.9797764768493877, "grad_norm": 0.28132277727127075, "learning_rate": 2.0651685258827468e-08, "loss": 0.1841, "step": 14956 }, { "epoch": 3.980042575838212, "grad_norm": 0.29890987277030945, "learning_rate": 2.0641303287486255e-08, "loss": 0.1973, "step": 14957 }, { "epoch": 3.9803086748270355, "grad_norm": 0.31118330359458923, "learning_rate": 2.0630923626070652e-08, "loss": 0.1909, "step": 14958 }, { "epoch": 3.9805747738158592, "grad_norm": 0.29427292943000793, "learning_rate": 2.062054627488278e-08, "loss": 0.1892, "step": 14959 }, { "epoch": 3.9808408728046834, "grad_norm": 0.31099167466163635, "learning_rate": 2.0610171234224726e-08, "loss": 0.1874, "step": 14960 }, { "epoch": 3.981106971793507, "grad_norm": 0.2900144159793854, "learning_rate": 2.0599798504398423e-08, "loss": 0.189, "step": 14961 }, { "epoch": 3.981373070782331, "grad_norm": 0.2801012396812439, "learning_rate": 2.058942808570584e-08, "loss": 0.1714, "step": 14962 }, { "epoch": 3.981639169771155, "grad_norm": 0.33901742100715637, "learning_rate": 2.05790599784488e-08, "loss": 0.2187, "step": 14963 }, { "epoch": 3.981905268759979, "grad_norm": 0.2690727114677429, "learning_rate": 2.0568694182929115e-08, "loss": 0.176, "step": 14964 }, { "epoch": 3.9821713677488026, "grad_norm": 0.33794882893562317, "learning_rate": 2.0558330699448534e-08, "loss": 0.1935, "step": 14965 }, { "epoch": 3.9824374667376263, "grad_norm": 0.2862667739391327, "learning_rate": 2.054796952830865e-08, "loss": 0.1839, "step": 14966 }, { "epoch": 3.9827035657264505, "grad_norm": 0.3100830018520355, "learning_rate": 2.0537610669811066e-08, "loss": 0.1785, "step": 14967 }, { "epoch": 3.982969664715274, "grad_norm": 0.329610675573349, "learning_rate": 2.0527254124257332e-08, "loss": 0.2035, "step": 14968 }, { "epoch": 3.983235763704098, "grad_norm": 0.28893864154815674, "learning_rate": 2.051689989194887e-08, "loss": 0.194, "step": 14969 }, { "epoch": 3.983501862692922, "grad_norm": 0.27445173263549805, "learning_rate": 2.050654797318707e-08, "loss": 0.179, "step": 14970 }, { "epoch": 3.9837679616817456, "grad_norm": 0.31794798374176025, "learning_rate": 2.0496198368273267e-08, "loss": 0.1954, "step": 14971 }, { "epoch": 3.9840340606705693, "grad_norm": 0.295952707529068, "learning_rate": 2.0485851077508697e-08, "loss": 0.1984, "step": 14972 }, { "epoch": 3.9843001596593934, "grad_norm": 0.28208377957344055, "learning_rate": 2.0475506101194552e-08, "loss": 0.1789, "step": 14973 }, { "epoch": 3.984566258648217, "grad_norm": 0.32342156767845154, "learning_rate": 2.046516343963196e-08, "loss": 0.1876, "step": 14974 }, { "epoch": 3.984832357637041, "grad_norm": 0.2958686053752899, "learning_rate": 2.045482309312192e-08, "loss": 0.1857, "step": 14975 }, { "epoch": 3.985098456625865, "grad_norm": 0.3228478729724884, "learning_rate": 2.0444485061965456e-08, "loss": 0.2068, "step": 14976 }, { "epoch": 3.9853645556146886, "grad_norm": 0.32759279012680054, "learning_rate": 2.043414934646346e-08, "loss": 0.1943, "step": 14977 }, { "epoch": 3.9856306546035123, "grad_norm": 0.3022361397743225, "learning_rate": 2.042381594691678e-08, "loss": 0.188, "step": 14978 }, { "epoch": 3.9858967535923364, "grad_norm": 0.29206520318984985, "learning_rate": 2.0413484863626208e-08, "loss": 0.1922, "step": 14979 }, { "epoch": 3.98616285258116, "grad_norm": 0.29586362838745117, "learning_rate": 2.040315609689247e-08, "loss": 0.1846, "step": 14980 }, { "epoch": 3.986428951569984, "grad_norm": 0.28074386715888977, "learning_rate": 2.0392829647016163e-08, "loss": 0.1806, "step": 14981 }, { "epoch": 3.986695050558808, "grad_norm": 0.30780351161956787, "learning_rate": 2.0382505514297887e-08, "loss": 0.2113, "step": 14982 }, { "epoch": 3.9869611495476316, "grad_norm": 0.29943856596946716, "learning_rate": 2.037218369903818e-08, "loss": 0.2028, "step": 14983 }, { "epoch": 3.9872272485364557, "grad_norm": 0.3332586884498596, "learning_rate": 2.0361864201537428e-08, "loss": 0.2042, "step": 14984 }, { "epoch": 3.9874933475252794, "grad_norm": 0.2938138246536255, "learning_rate": 2.0351547022096016e-08, "loss": 0.1808, "step": 14985 }, { "epoch": 3.9877594465141035, "grad_norm": 0.2867070436477661, "learning_rate": 2.0341232161014277e-08, "loss": 0.1848, "step": 14986 }, { "epoch": 3.988025545502927, "grad_norm": 0.3067185580730438, "learning_rate": 2.0330919618592435e-08, "loss": 0.1948, "step": 14987 }, { "epoch": 3.988291644491751, "grad_norm": 0.31879428029060364, "learning_rate": 2.032060939513067e-08, "loss": 0.1907, "step": 14988 }, { "epoch": 3.988557743480575, "grad_norm": 0.29478955268859863, "learning_rate": 2.031030149092907e-08, "loss": 0.1879, "step": 14989 }, { "epoch": 3.9888238424693987, "grad_norm": 0.30274733901023865, "learning_rate": 2.029999590628769e-08, "loss": 0.1933, "step": 14990 }, { "epoch": 3.9890899414582224, "grad_norm": 0.3073115646839142, "learning_rate": 2.0289692641506493e-08, "loss": 0.1825, "step": 14991 }, { "epoch": 3.9893560404470465, "grad_norm": 0.2729243040084839, "learning_rate": 2.0279391696885394e-08, "loss": 0.1706, "step": 14992 }, { "epoch": 3.98962213943587, "grad_norm": 0.301504522562027, "learning_rate": 2.02690930727242e-08, "loss": 0.2013, "step": 14993 }, { "epoch": 3.989888238424694, "grad_norm": 0.28668370842933655, "learning_rate": 2.025879676932267e-08, "loss": 0.1834, "step": 14994 }, { "epoch": 3.990154337413518, "grad_norm": 0.31356507539749146, "learning_rate": 2.0248502786980526e-08, "loss": 0.1946, "step": 14995 }, { "epoch": 3.9904204364023417, "grad_norm": 0.28243911266326904, "learning_rate": 2.0238211125997395e-08, "loss": 0.1974, "step": 14996 }, { "epoch": 3.9906865353911654, "grad_norm": 0.26441049575805664, "learning_rate": 2.022792178667283e-08, "loss": 0.1534, "step": 14997 }, { "epoch": 3.9909526343799895, "grad_norm": 0.30152440071105957, "learning_rate": 2.0217634769306347e-08, "loss": 0.1963, "step": 14998 }, { "epoch": 3.991218733368813, "grad_norm": 0.2989377975463867, "learning_rate": 2.020735007419736e-08, "loss": 0.2019, "step": 14999 }, { "epoch": 3.991484832357637, "grad_norm": 0.29250702261924744, "learning_rate": 2.0197067701645255e-08, "loss": 0.1779, "step": 15000 }, { "epoch": 3.991750931346461, "grad_norm": 0.29313114285469055, "learning_rate": 2.0186787651949278e-08, "loss": 0.1786, "step": 15001 }, { "epoch": 3.9920170303352847, "grad_norm": 0.3209103047847748, "learning_rate": 2.017650992540868e-08, "loss": 0.205, "step": 15002 }, { "epoch": 3.9922831293241083, "grad_norm": 0.3047346770763397, "learning_rate": 2.0166234522322644e-08, "loss": 0.1819, "step": 15003 }, { "epoch": 3.9925492283129325, "grad_norm": 0.3239666223526001, "learning_rate": 2.015596144299021e-08, "loss": 0.1884, "step": 15004 }, { "epoch": 3.992815327301756, "grad_norm": 0.3165817856788635, "learning_rate": 2.0145690687710425e-08, "loss": 0.206, "step": 15005 }, { "epoch": 3.99308142629058, "grad_norm": 0.3000262975692749, "learning_rate": 2.0135422256782253e-08, "loss": 0.2102, "step": 15006 }, { "epoch": 3.993347525279404, "grad_norm": 0.29881614446640015, "learning_rate": 2.0125156150504563e-08, "loss": 0.2079, "step": 15007 }, { "epoch": 3.9936136242682276, "grad_norm": 0.31610307097435, "learning_rate": 2.0114892369176184e-08, "loss": 0.207, "step": 15008 }, { "epoch": 3.9938797232570518, "grad_norm": 0.30668875575065613, "learning_rate": 2.0104630913095867e-08, "loss": 0.1958, "step": 15009 }, { "epoch": 3.9941458222458754, "grad_norm": 0.2882729470729828, "learning_rate": 2.0094371782562325e-08, "loss": 0.208, "step": 15010 }, { "epoch": 3.9944119212346996, "grad_norm": 0.2946227490901947, "learning_rate": 2.0084114977874133e-08, "loss": 0.1907, "step": 15011 }, { "epoch": 3.9946780202235233, "grad_norm": 0.3371238708496094, "learning_rate": 2.0073860499329843e-08, "loss": 0.2147, "step": 15012 }, { "epoch": 3.994944119212347, "grad_norm": 0.27367040514945984, "learning_rate": 2.0063608347227946e-08, "loss": 0.1801, "step": 15013 }, { "epoch": 3.995210218201171, "grad_norm": 0.31872761249542236, "learning_rate": 2.005335852186687e-08, "loss": 0.2091, "step": 15014 }, { "epoch": 3.9954763171899947, "grad_norm": 0.32730498909950256, "learning_rate": 2.0043111023544934e-08, "loss": 0.2016, "step": 15015 }, { "epoch": 3.9957424161788184, "grad_norm": 0.2902812957763672, "learning_rate": 2.003286585256043e-08, "loss": 0.1868, "step": 15016 }, { "epoch": 3.9960085151676425, "grad_norm": 0.28621402382850647, "learning_rate": 2.002262300921158e-08, "loss": 0.1733, "step": 15017 }, { "epoch": 3.9962746141564662, "grad_norm": 0.3025347888469696, "learning_rate": 2.0012382493796508e-08, "loss": 0.1876, "step": 15018 }, { "epoch": 3.99654071314529, "grad_norm": 0.2850509285926819, "learning_rate": 2.0002144306613323e-08, "loss": 0.1902, "step": 15019 }, { "epoch": 3.996806812134114, "grad_norm": 0.2909678518772125, "learning_rate": 1.9991908447959983e-08, "loss": 0.189, "step": 15020 }, { "epoch": 3.9970729111229377, "grad_norm": 0.3595559000968933, "learning_rate": 1.998167491813444e-08, "loss": 0.2204, "step": 15021 }, { "epoch": 3.9973390101117614, "grad_norm": 0.28762948513031006, "learning_rate": 1.997144371743461e-08, "loss": 0.1809, "step": 15022 }, { "epoch": 3.9976051091005855, "grad_norm": 0.27656835317611694, "learning_rate": 1.9961214846158235e-08, "loss": 0.1818, "step": 15023 }, { "epoch": 3.997871208089409, "grad_norm": 0.33480945229530334, "learning_rate": 1.9950988304603067e-08, "loss": 0.2059, "step": 15024 }, { "epoch": 3.998137307078233, "grad_norm": 0.36534208059310913, "learning_rate": 1.9940764093066796e-08, "loss": 0.2033, "step": 15025 }, { "epoch": 3.998403406067057, "grad_norm": 0.29788461327552795, "learning_rate": 1.9930542211847e-08, "loss": 0.2, "step": 15026 }, { "epoch": 3.9986695050558807, "grad_norm": 0.2910445034503937, "learning_rate": 1.992032266124123e-08, "loss": 0.182, "step": 15027 }, { "epoch": 3.9989356040447044, "grad_norm": 0.28127381205558777, "learning_rate": 1.9910105441546964e-08, "loss": 0.1891, "step": 15028 }, { "epoch": 3.9992017030335285, "grad_norm": 0.33353105187416077, "learning_rate": 1.989989055306156e-08, "loss": 0.2022, "step": 15029 }, { "epoch": 3.999467802022352, "grad_norm": 0.29958489537239075, "learning_rate": 1.9889677996082355e-08, "loss": 0.1957, "step": 15030 }, { "epoch": 3.9997339010111763, "grad_norm": 0.3147068917751312, "learning_rate": 1.987946777090662e-08, "loss": 0.2079, "step": 15031 }, { "epoch": 4.0, "grad_norm": 0.2879391312599182, "learning_rate": 1.9869259877831546e-08, "loss": 0.1765, "step": 15032 }, { "epoch": 4.000266098988824, "grad_norm": 0.2817676067352295, "learning_rate": 1.9859054317154265e-08, "loss": 0.2051, "step": 15033 }, { "epoch": 4.000532197977647, "grad_norm": 0.3153369724750519, "learning_rate": 1.9848851089171824e-08, "loss": 0.202, "step": 15034 }, { "epoch": 4.0007982969664715, "grad_norm": 0.3221234679222107, "learning_rate": 1.9838650194181215e-08, "loss": 0.2012, "step": 15035 }, { "epoch": 4.001064395955296, "grad_norm": 0.2968790531158447, "learning_rate": 1.982845163247937e-08, "loss": 0.1728, "step": 15036 }, { "epoch": 4.001330494944119, "grad_norm": 0.31144601106643677, "learning_rate": 1.981825540436316e-08, "loss": 0.2051, "step": 15037 }, { "epoch": 4.001596593932943, "grad_norm": 0.29883521795272827, "learning_rate": 1.9808061510129315e-08, "loss": 0.1739, "step": 15038 }, { "epoch": 4.001862692921767, "grad_norm": 0.3138100802898407, "learning_rate": 1.9797869950074586e-08, "loss": 0.1975, "step": 15039 }, { "epoch": 4.00212879191059, "grad_norm": 0.32317450642585754, "learning_rate": 1.9787680724495615e-08, "loss": 0.1909, "step": 15040 }, { "epoch": 4.0023948908994145, "grad_norm": 0.2855346202850342, "learning_rate": 1.9777493833688996e-08, "loss": 0.1734, "step": 15041 }, { "epoch": 4.002660989888239, "grad_norm": 0.3084987998008728, "learning_rate": 1.976730927795126e-08, "loss": 0.1949, "step": 15042 }, { "epoch": 4.002927088877062, "grad_norm": 0.3030807375907898, "learning_rate": 1.9757127057578805e-08, "loss": 0.1922, "step": 15043 }, { "epoch": 4.003193187865886, "grad_norm": 0.3023544251918793, "learning_rate": 1.9746947172868024e-08, "loss": 0.1905, "step": 15044 }, { "epoch": 4.00345928685471, "grad_norm": 0.3115883767604828, "learning_rate": 1.9736769624115246e-08, "loss": 0.1918, "step": 15045 }, { "epoch": 4.003725385843534, "grad_norm": 0.29026859998703003, "learning_rate": 1.9726594411616728e-08, "loss": 0.1769, "step": 15046 }, { "epoch": 4.0039914848323575, "grad_norm": 0.3356688320636749, "learning_rate": 1.971642153566858e-08, "loss": 0.1953, "step": 15047 }, { "epoch": 4.004257583821182, "grad_norm": 0.2919880747795105, "learning_rate": 1.9706250996566954e-08, "loss": 0.1774, "step": 15048 }, { "epoch": 4.004523682810006, "grad_norm": 0.28338587284088135, "learning_rate": 1.969608279460788e-08, "loss": 0.1726, "step": 15049 }, { "epoch": 4.004789781798829, "grad_norm": 0.2962515950202942, "learning_rate": 1.968591693008733e-08, "loss": 0.1941, "step": 15050 }, { "epoch": 4.005055880787653, "grad_norm": 0.29977768659591675, "learning_rate": 1.967575340330121e-08, "loss": 0.1895, "step": 15051 }, { "epoch": 4.005321979776477, "grad_norm": 0.3121107816696167, "learning_rate": 1.9665592214545347e-08, "loss": 0.2066, "step": 15052 }, { "epoch": 4.0055880787653, "grad_norm": 0.30895453691482544, "learning_rate": 1.9655433364115504e-08, "loss": 0.1827, "step": 15053 }, { "epoch": 4.005854177754125, "grad_norm": 0.2735651433467865, "learning_rate": 1.9645276852307392e-08, "loss": 0.1882, "step": 15054 }, { "epoch": 4.006120276742949, "grad_norm": 0.3206486999988556, "learning_rate": 1.963512267941665e-08, "loss": 0.2264, "step": 15055 }, { "epoch": 4.006386375731772, "grad_norm": 0.2806089222431183, "learning_rate": 1.96249708457388e-08, "loss": 0.1963, "step": 15056 }, { "epoch": 4.006652474720596, "grad_norm": 0.6061239838600159, "learning_rate": 1.9614821351569343e-08, "loss": 0.2149, "step": 15057 }, { "epoch": 4.00691857370942, "grad_norm": 0.29280510544776917, "learning_rate": 1.960467419720373e-08, "loss": 0.1864, "step": 15058 }, { "epoch": 4.007184672698243, "grad_norm": 0.321481317281723, "learning_rate": 1.959452938293731e-08, "loss": 0.2011, "step": 15059 }, { "epoch": 4.0074507716870675, "grad_norm": 0.2936200201511383, "learning_rate": 1.9584386909065365e-08, "loss": 0.1895, "step": 15060 }, { "epoch": 4.007716870675892, "grad_norm": 0.27025166153907776, "learning_rate": 1.9574246775883152e-08, "loss": 0.1581, "step": 15061 }, { "epoch": 4.007982969664715, "grad_norm": 0.29776617884635925, "learning_rate": 1.9564108983685756e-08, "loss": 0.2, "step": 15062 }, { "epoch": 4.008249068653539, "grad_norm": 0.2942911684513092, "learning_rate": 1.9553973532768297e-08, "loss": 0.1834, "step": 15063 }, { "epoch": 4.008515167642363, "grad_norm": 0.2908310294151306, "learning_rate": 1.954384042342583e-08, "loss": 0.1656, "step": 15064 }, { "epoch": 4.008781266631186, "grad_norm": 0.2896577715873718, "learning_rate": 1.9533709655953234e-08, "loss": 0.1804, "step": 15065 }, { "epoch": 4.0090473656200105, "grad_norm": 0.2846387028694153, "learning_rate": 1.9523581230645413e-08, "loss": 0.1976, "step": 15066 }, { "epoch": 4.009313464608835, "grad_norm": 0.27374640107154846, "learning_rate": 1.9513455147797198e-08, "loss": 0.1814, "step": 15067 }, { "epoch": 4.009579563597659, "grad_norm": 0.3080528974533081, "learning_rate": 1.9503331407703317e-08, "loss": 0.1863, "step": 15068 }, { "epoch": 4.009845662586482, "grad_norm": 0.33058348298072815, "learning_rate": 1.9493210010658456e-08, "loss": 0.1989, "step": 15069 }, { "epoch": 4.010111761575306, "grad_norm": 0.2971341013908386, "learning_rate": 1.9483090956957216e-08, "loss": 0.1972, "step": 15070 }, { "epoch": 4.01037786056413, "grad_norm": 0.2986631691455841, "learning_rate": 1.9472974246894136e-08, "loss": 0.1973, "step": 15071 }, { "epoch": 4.0106439595529535, "grad_norm": 0.2970145046710968, "learning_rate": 1.9462859880763693e-08, "loss": 0.1778, "step": 15072 }, { "epoch": 4.010910058541778, "grad_norm": 0.3002634048461914, "learning_rate": 1.9452747858860307e-08, "loss": 0.2045, "step": 15073 }, { "epoch": 4.011176157530602, "grad_norm": 0.28516849875450134, "learning_rate": 1.9442638181478276e-08, "loss": 0.1669, "step": 15074 }, { "epoch": 4.011442256519425, "grad_norm": 0.2915102541446686, "learning_rate": 1.9432530848911877e-08, "loss": 0.1855, "step": 15075 }, { "epoch": 4.011708355508249, "grad_norm": 0.297514945268631, "learning_rate": 1.9422425861455304e-08, "loss": 0.1723, "step": 15076 }, { "epoch": 4.011974454497073, "grad_norm": 0.27041539549827576, "learning_rate": 1.9412323219402705e-08, "loss": 0.1878, "step": 15077 }, { "epoch": 4.0122405534858965, "grad_norm": 0.2993571162223816, "learning_rate": 1.940222292304814e-08, "loss": 0.2005, "step": 15078 }, { "epoch": 4.012506652474721, "grad_norm": 0.32027849555015564, "learning_rate": 1.9392124972685586e-08, "loss": 0.1672, "step": 15079 }, { "epoch": 4.012772751463545, "grad_norm": 0.29650211334228516, "learning_rate": 1.9382029368608976e-08, "loss": 0.1951, "step": 15080 }, { "epoch": 4.013038850452368, "grad_norm": 0.29699215292930603, "learning_rate": 1.937193611111221e-08, "loss": 0.1756, "step": 15081 }, { "epoch": 4.013304949441192, "grad_norm": 0.3107917904853821, "learning_rate": 1.936184520048899e-08, "loss": 0.1898, "step": 15082 }, { "epoch": 4.013571048430016, "grad_norm": 0.2780263423919678, "learning_rate": 1.935175663703309e-08, "loss": 0.1625, "step": 15083 }, { "epoch": 4.0138371474188395, "grad_norm": 0.2918245494365692, "learning_rate": 1.934167042103818e-08, "loss": 0.1818, "step": 15084 }, { "epoch": 4.014103246407664, "grad_norm": 0.3009835481643677, "learning_rate": 1.9331586552797797e-08, "loss": 0.1829, "step": 15085 }, { "epoch": 4.014369345396488, "grad_norm": 0.31370946764945984, "learning_rate": 1.9321505032605468e-08, "loss": 0.1756, "step": 15086 }, { "epoch": 4.014635444385311, "grad_norm": 0.3020119369029999, "learning_rate": 1.931142586075465e-08, "loss": 0.1884, "step": 15087 }, { "epoch": 4.014901543374135, "grad_norm": 0.28212299942970276, "learning_rate": 1.9301349037538717e-08, "loss": 0.1844, "step": 15088 }, { "epoch": 4.015167642362959, "grad_norm": 0.294092059135437, "learning_rate": 1.929127456325098e-08, "loss": 0.1796, "step": 15089 }, { "epoch": 4.015433741351782, "grad_norm": 0.28105756640434265, "learning_rate": 1.9281202438184686e-08, "loss": 0.1883, "step": 15090 }, { "epoch": 4.015699840340607, "grad_norm": 0.30449867248535156, "learning_rate": 1.9271132662633004e-08, "loss": 0.2018, "step": 15091 }, { "epoch": 4.015965939329431, "grad_norm": 0.29339542984962463, "learning_rate": 1.9261065236889063e-08, "loss": 0.1933, "step": 15092 }, { "epoch": 4.016232038318255, "grad_norm": 0.33596116304397583, "learning_rate": 1.9251000161245855e-08, "loss": 0.2106, "step": 15093 }, { "epoch": 4.016498137307078, "grad_norm": 0.32182854413986206, "learning_rate": 1.9240937435996362e-08, "loss": 0.1989, "step": 15094 }, { "epoch": 4.016764236295902, "grad_norm": 0.3149864971637726, "learning_rate": 1.9230877061433502e-08, "loss": 0.201, "step": 15095 }, { "epoch": 4.017030335284726, "grad_norm": 0.3252885937690735, "learning_rate": 1.922081903785009e-08, "loss": 0.2127, "step": 15096 }, { "epoch": 4.0172964342735495, "grad_norm": 0.27147504687309265, "learning_rate": 1.9210763365538894e-08, "loss": 0.179, "step": 15097 }, { "epoch": 4.017562533262374, "grad_norm": 0.2815079092979431, "learning_rate": 1.9200710044792612e-08, "loss": 0.1935, "step": 15098 }, { "epoch": 4.017828632251198, "grad_norm": 0.27968624234199524, "learning_rate": 1.919065907590387e-08, "loss": 0.1838, "step": 15099 }, { "epoch": 4.018094731240021, "grad_norm": 0.30568554997444153, "learning_rate": 1.9180610459165248e-08, "loss": 0.2005, "step": 15100 }, { "epoch": 4.018360830228845, "grad_norm": 0.31619498133659363, "learning_rate": 1.9170564194869177e-08, "loss": 0.1915, "step": 15101 }, { "epoch": 4.018626929217669, "grad_norm": 0.3053704798221588, "learning_rate": 1.9160520283308114e-08, "loss": 0.1891, "step": 15102 }, { "epoch": 4.0188930282064925, "grad_norm": 0.2910172641277313, "learning_rate": 1.9150478724774445e-08, "loss": 0.1783, "step": 15103 }, { "epoch": 4.019159127195317, "grad_norm": 0.2879372537136078, "learning_rate": 1.914043951956037e-08, "loss": 0.1865, "step": 15104 }, { "epoch": 4.019425226184141, "grad_norm": 0.3209002614021301, "learning_rate": 1.9130402667958167e-08, "loss": 0.1827, "step": 15105 }, { "epoch": 4.019691325172964, "grad_norm": 0.26699545979499817, "learning_rate": 1.912036817025996e-08, "loss": 0.1759, "step": 15106 }, { "epoch": 4.019957424161788, "grad_norm": 0.31174659729003906, "learning_rate": 1.9110336026757835e-08, "loss": 0.1829, "step": 15107 }, { "epoch": 4.020223523150612, "grad_norm": 0.2999320924282074, "learning_rate": 1.9100306237743802e-08, "loss": 0.1897, "step": 15108 }, { "epoch": 4.0204896221394355, "grad_norm": 0.3882080912590027, "learning_rate": 1.9090278803509797e-08, "loss": 0.202, "step": 15109 }, { "epoch": 4.02075572112826, "grad_norm": 0.25826749205589294, "learning_rate": 1.9080253724347726e-08, "loss": 0.1697, "step": 15110 }, { "epoch": 4.021021820117084, "grad_norm": 0.2795042097568512, "learning_rate": 1.907023100054933e-08, "loss": 0.1815, "step": 15111 }, { "epoch": 4.021287919105907, "grad_norm": 0.303286075592041, "learning_rate": 1.9060210632406383e-08, "loss": 0.2026, "step": 15112 }, { "epoch": 4.021554018094731, "grad_norm": 0.2882431149482727, "learning_rate": 1.9050192620210547e-08, "loss": 0.1845, "step": 15113 }, { "epoch": 4.021820117083555, "grad_norm": 0.3758770823478699, "learning_rate": 1.9040176964253417e-08, "loss": 0.1779, "step": 15114 }, { "epoch": 4.0220862160723785, "grad_norm": 0.31062084436416626, "learning_rate": 1.9030163664826537e-08, "loss": 0.1745, "step": 15115 }, { "epoch": 4.022352315061203, "grad_norm": 0.3765328526496887, "learning_rate": 1.902015272222134e-08, "loss": 0.1954, "step": 15116 }, { "epoch": 4.022618414050027, "grad_norm": 0.3115042746067047, "learning_rate": 1.9010144136729245e-08, "loss": 0.1864, "step": 15117 }, { "epoch": 4.022884513038851, "grad_norm": 0.3276275396347046, "learning_rate": 1.9000137908641566e-08, "loss": 0.2098, "step": 15118 }, { "epoch": 4.023150612027674, "grad_norm": 3.0325772762298584, "learning_rate": 1.8990134038249582e-08, "loss": 0.1833, "step": 15119 }, { "epoch": 4.023416711016498, "grad_norm": 0.3210701048374176, "learning_rate": 1.8980132525844427e-08, "loss": 0.1876, "step": 15120 }, { "epoch": 4.023682810005322, "grad_norm": 0.29858019948005676, "learning_rate": 1.8970133371717246e-08, "loss": 0.1948, "step": 15121 }, { "epoch": 4.023948908994146, "grad_norm": 0.2902229428291321, "learning_rate": 1.8960136576159103e-08, "loss": 0.1937, "step": 15122 }, { "epoch": 4.02421500798297, "grad_norm": 0.34674185514450073, "learning_rate": 1.8950142139460977e-08, "loss": 0.1777, "step": 15123 }, { "epoch": 4.024481106971794, "grad_norm": 0.34348636865615845, "learning_rate": 1.8940150061913752e-08, "loss": 0.2032, "step": 15124 }, { "epoch": 4.024747205960617, "grad_norm": 0.3260689973831177, "learning_rate": 1.8930160343808287e-08, "loss": 0.1816, "step": 15125 }, { "epoch": 4.025013304949441, "grad_norm": 0.31263282895088196, "learning_rate": 1.8920172985435367e-08, "loss": 0.2016, "step": 15126 }, { "epoch": 4.025279403938265, "grad_norm": 0.32028889656066895, "learning_rate": 1.8910187987085678e-08, "loss": 0.2051, "step": 15127 }, { "epoch": 4.025545502927089, "grad_norm": 0.3073887526988983, "learning_rate": 1.8900205349049903e-08, "loss": 0.1905, "step": 15128 }, { "epoch": 4.025811601915913, "grad_norm": 0.2916789948940277, "learning_rate": 1.889022507161855e-08, "loss": 0.1842, "step": 15129 }, { "epoch": 4.026077700904737, "grad_norm": 0.2867863178253174, "learning_rate": 1.888024715508214e-08, "loss": 0.1825, "step": 15130 }, { "epoch": 4.02634379989356, "grad_norm": 0.2761531174182892, "learning_rate": 1.8870271599731113e-08, "loss": 0.1793, "step": 15131 }, { "epoch": 4.026609898882384, "grad_norm": 0.30952170491218567, "learning_rate": 1.8860298405855834e-08, "loss": 0.188, "step": 15132 }, { "epoch": 4.026875997871208, "grad_norm": 0.2680613696575165, "learning_rate": 1.8850327573746583e-08, "loss": 0.1701, "step": 15133 }, { "epoch": 4.0271420968600316, "grad_norm": 0.29604408144950867, "learning_rate": 1.8840359103693592e-08, "loss": 0.177, "step": 15134 }, { "epoch": 4.027408195848856, "grad_norm": 0.2752539813518524, "learning_rate": 1.8830392995987033e-08, "loss": 0.1755, "step": 15135 }, { "epoch": 4.02767429483768, "grad_norm": 0.2994175851345062, "learning_rate": 1.882042925091697e-08, "loss": 0.1902, "step": 15136 }, { "epoch": 4.027940393826503, "grad_norm": 0.2903673052787781, "learning_rate": 1.881046786877345e-08, "loss": 0.2061, "step": 15137 }, { "epoch": 4.028206492815327, "grad_norm": 0.3105187714099884, "learning_rate": 1.8800508849846386e-08, "loss": 0.1959, "step": 15138 }, { "epoch": 4.028472591804151, "grad_norm": 0.2924853563308716, "learning_rate": 1.8790552194425678e-08, "loss": 0.1877, "step": 15139 }, { "epoch": 4.028738690792975, "grad_norm": 0.26864099502563477, "learning_rate": 1.8780597902801132e-08, "loss": 0.1719, "step": 15140 }, { "epoch": 4.029004789781799, "grad_norm": 0.3212112784385681, "learning_rate": 1.87706459752625e-08, "loss": 0.1949, "step": 15141 }, { "epoch": 4.029270888770623, "grad_norm": 0.2837810516357422, "learning_rate": 1.8760696412099453e-08, "loss": 0.176, "step": 15142 }, { "epoch": 4.029536987759447, "grad_norm": 0.3355604112148285, "learning_rate": 1.8750749213601625e-08, "loss": 0.2283, "step": 15143 }, { "epoch": 4.02980308674827, "grad_norm": 0.3403922915458679, "learning_rate": 1.8740804380058495e-08, "loss": 0.2068, "step": 15144 }, { "epoch": 4.030069185737094, "grad_norm": 0.2811851501464844, "learning_rate": 1.8730861911759565e-08, "loss": 0.1767, "step": 15145 }, { "epoch": 4.030335284725918, "grad_norm": 0.2727055549621582, "learning_rate": 1.872092180899426e-08, "loss": 0.1792, "step": 15146 }, { "epoch": 4.030601383714742, "grad_norm": 0.3217625319957733, "learning_rate": 1.871098407205186e-08, "loss": 0.1912, "step": 15147 }, { "epoch": 4.030867482703566, "grad_norm": 0.2942333221435547, "learning_rate": 1.8701048701221644e-08, "loss": 0.1818, "step": 15148 }, { "epoch": 4.03113358169239, "grad_norm": 0.30458754301071167, "learning_rate": 1.8691115696792803e-08, "loss": 0.1935, "step": 15149 }, { "epoch": 4.031399680681213, "grad_norm": 0.3155501186847687, "learning_rate": 1.8681185059054472e-08, "loss": 0.192, "step": 15150 }, { "epoch": 4.031665779670037, "grad_norm": 0.30668139457702637, "learning_rate": 1.867125678829571e-08, "loss": 0.1862, "step": 15151 }, { "epoch": 4.031931878658861, "grad_norm": 0.2853943407535553, "learning_rate": 1.8661330884805482e-08, "loss": 0.1784, "step": 15152 }, { "epoch": 4.032197977647685, "grad_norm": 0.31356191635131836, "learning_rate": 1.8651407348872728e-08, "loss": 0.1816, "step": 15153 }, { "epoch": 4.032464076636509, "grad_norm": 0.3152729868888855, "learning_rate": 1.8641486180786283e-08, "loss": 0.1967, "step": 15154 }, { "epoch": 4.032730175625333, "grad_norm": 0.31526654958724976, "learning_rate": 1.8631567380834955e-08, "loss": 0.2022, "step": 15155 }, { "epoch": 4.032996274614156, "grad_norm": 0.31690484285354614, "learning_rate": 1.862165094930741e-08, "loss": 0.2201, "step": 15156 }, { "epoch": 4.03326237360298, "grad_norm": 0.2980385720729828, "learning_rate": 1.86117368864923e-08, "loss": 0.1905, "step": 15157 }, { "epoch": 4.033528472591804, "grad_norm": 0.2871812880039215, "learning_rate": 1.860182519267821e-08, "loss": 0.172, "step": 15158 }, { "epoch": 4.033794571580628, "grad_norm": 0.3255405128002167, "learning_rate": 1.859191586815364e-08, "loss": 0.2013, "step": 15159 }, { "epoch": 4.034060670569452, "grad_norm": 0.29987338185310364, "learning_rate": 1.8582008913207025e-08, "loss": 0.1951, "step": 15160 }, { "epoch": 4.034326769558276, "grad_norm": 0.32545775175094604, "learning_rate": 1.8572104328126736e-08, "loss": 0.1949, "step": 15161 }, { "epoch": 4.034592868547099, "grad_norm": 0.34274721145629883, "learning_rate": 1.8562202113201086e-08, "loss": 0.1966, "step": 15162 }, { "epoch": 4.034858967535923, "grad_norm": 0.29299020767211914, "learning_rate": 1.8552302268718258e-08, "loss": 0.1897, "step": 15163 }, { "epoch": 4.035125066524747, "grad_norm": 0.2879461944103241, "learning_rate": 1.8542404794966426e-08, "loss": 0.1734, "step": 15164 }, { "epoch": 4.0353911655135715, "grad_norm": 0.3397575318813324, "learning_rate": 1.8532509692233722e-08, "loss": 0.2181, "step": 15165 }, { "epoch": 4.035657264502395, "grad_norm": 0.30727335810661316, "learning_rate": 1.8522616960808113e-08, "loss": 0.1963, "step": 15166 }, { "epoch": 4.035923363491219, "grad_norm": 0.30252805352211, "learning_rate": 1.8512726600977556e-08, "loss": 0.1859, "step": 15167 }, { "epoch": 4.036189462480043, "grad_norm": 0.28363344073295593, "learning_rate": 1.850283861302996e-08, "loss": 0.1906, "step": 15168 }, { "epoch": 4.036455561468866, "grad_norm": 0.2788446247577667, "learning_rate": 1.849295299725312e-08, "loss": 0.177, "step": 15169 }, { "epoch": 4.03672166045769, "grad_norm": 0.3321244418621063, "learning_rate": 1.8483069753934788e-08, "loss": 0.1966, "step": 15170 }, { "epoch": 4.0369877594465144, "grad_norm": 0.3101147413253784, "learning_rate": 1.8473188883362633e-08, "loss": 0.2093, "step": 15171 }, { "epoch": 4.037253858435338, "grad_norm": 0.2915489077568054, "learning_rate": 1.8463310385824273e-08, "loss": 0.2079, "step": 15172 }, { "epoch": 4.037519957424162, "grad_norm": 0.3071044981479645, "learning_rate": 1.845343426160727e-08, "loss": 0.1899, "step": 15173 }, { "epoch": 4.037786056412986, "grad_norm": 0.29382532835006714, "learning_rate": 1.8443560510999034e-08, "loss": 0.2009, "step": 15174 }, { "epoch": 4.038052155401809, "grad_norm": 0.317773699760437, "learning_rate": 1.843368913428699e-08, "loss": 0.1916, "step": 15175 }, { "epoch": 4.038318254390633, "grad_norm": 0.3186979591846466, "learning_rate": 1.8423820131758472e-08, "loss": 0.2051, "step": 15176 }, { "epoch": 4.038584353379457, "grad_norm": 0.2747167646884918, "learning_rate": 1.8413953503700742e-08, "loss": 0.1674, "step": 15177 }, { "epoch": 4.038850452368281, "grad_norm": 0.28892654180526733, "learning_rate": 1.8404089250400978e-08, "loss": 0.1996, "step": 15178 }, { "epoch": 4.039116551357105, "grad_norm": 0.31067296862602234, "learning_rate": 1.839422737214632e-08, "loss": 0.2007, "step": 15179 }, { "epoch": 4.039382650345929, "grad_norm": 0.31120872497558594, "learning_rate": 1.8384367869223826e-08, "loss": 0.2033, "step": 15180 }, { "epoch": 4.039648749334752, "grad_norm": 0.31540438532829285, "learning_rate": 1.8374510741920457e-08, "loss": 0.196, "step": 15181 }, { "epoch": 4.039914848323576, "grad_norm": 0.31253018975257874, "learning_rate": 1.8364655990523183e-08, "loss": 0.196, "step": 15182 }, { "epoch": 4.0401809473124, "grad_norm": 0.30678945779800415, "learning_rate": 1.835480361531877e-08, "loss": 0.1985, "step": 15183 }, { "epoch": 4.040447046301224, "grad_norm": 0.28547146916389465, "learning_rate": 1.834495361659404e-08, "loss": 0.1697, "step": 15184 }, { "epoch": 4.040713145290048, "grad_norm": 0.30239519476890564, "learning_rate": 1.8335105994635713e-08, "loss": 0.1697, "step": 15185 }, { "epoch": 4.040979244278872, "grad_norm": 0.2964491546154022, "learning_rate": 1.8325260749730385e-08, "loss": 0.1848, "step": 15186 }, { "epoch": 4.041245343267696, "grad_norm": 0.2968927025794983, "learning_rate": 1.8315417882164664e-08, "loss": 0.18, "step": 15187 }, { "epoch": 4.041511442256519, "grad_norm": 0.3188728094100952, "learning_rate": 1.8305577392225025e-08, "loss": 0.2, "step": 15188 }, { "epoch": 4.041777541245343, "grad_norm": 0.29680466651916504, "learning_rate": 1.829573928019792e-08, "loss": 0.1986, "step": 15189 }, { "epoch": 4.0420436402341675, "grad_norm": 0.2721911072731018, "learning_rate": 1.8285903546369687e-08, "loss": 0.1872, "step": 15190 }, { "epoch": 4.042309739222991, "grad_norm": 0.305408775806427, "learning_rate": 1.8276070191026672e-08, "loss": 0.1944, "step": 15191 }, { "epoch": 4.042575838211815, "grad_norm": 0.29566484689712524, "learning_rate": 1.8266239214455027e-08, "loss": 0.1869, "step": 15192 }, { "epoch": 4.042841937200639, "grad_norm": 0.3051067590713501, "learning_rate": 1.8256410616940944e-08, "loss": 0.1874, "step": 15193 }, { "epoch": 4.043108036189462, "grad_norm": 0.3038381338119507, "learning_rate": 1.8246584398770493e-08, "loss": 0.1939, "step": 15194 }, { "epoch": 4.043374135178286, "grad_norm": 0.3067580759525299, "learning_rate": 1.8236760560229715e-08, "loss": 0.1874, "step": 15195 }, { "epoch": 4.0436402341671105, "grad_norm": 0.3030587136745453, "learning_rate": 1.8226939101604522e-08, "loss": 0.1981, "step": 15196 }, { "epoch": 4.043906333155934, "grad_norm": 0.3247995376586914, "learning_rate": 1.8217120023180822e-08, "loss": 0.1998, "step": 15197 }, { "epoch": 4.044172432144758, "grad_norm": 0.3000451624393463, "learning_rate": 1.8207303325244415e-08, "loss": 0.186, "step": 15198 }, { "epoch": 4.044438531133582, "grad_norm": 0.2982993423938751, "learning_rate": 1.8197489008081023e-08, "loss": 0.1952, "step": 15199 }, { "epoch": 4.044704630122405, "grad_norm": 0.2921949625015259, "learning_rate": 1.818767707197636e-08, "loss": 0.1754, "step": 15200 }, { "epoch": 4.044970729111229, "grad_norm": 0.27431434392929077, "learning_rate": 1.8177867517215973e-08, "loss": 0.1769, "step": 15201 }, { "epoch": 4.0452368281000535, "grad_norm": 0.2875366806983948, "learning_rate": 1.8168060344085413e-08, "loss": 0.1748, "step": 15202 }, { "epoch": 4.045502927088877, "grad_norm": 0.33590564131736755, "learning_rate": 1.815825555287014e-08, "loss": 0.1977, "step": 15203 }, { "epoch": 4.045769026077701, "grad_norm": 0.2813149392604828, "learning_rate": 1.814845314385557e-08, "loss": 0.1699, "step": 15204 }, { "epoch": 4.046035125066525, "grad_norm": 0.28449469804763794, "learning_rate": 1.8138653117326974e-08, "loss": 0.1886, "step": 15205 }, { "epoch": 4.046301224055348, "grad_norm": 0.32194972038269043, "learning_rate": 1.812885547356964e-08, "loss": 0.2196, "step": 15206 }, { "epoch": 4.046567323044172, "grad_norm": 0.3043003976345062, "learning_rate": 1.8119060212868754e-08, "loss": 0.1943, "step": 15207 }, { "epoch": 4.0468334220329965, "grad_norm": 0.3263835608959198, "learning_rate": 1.8109267335509416e-08, "loss": 0.1943, "step": 15208 }, { "epoch": 4.04709952102182, "grad_norm": 0.28608229756355286, "learning_rate": 1.8099476841776696e-08, "loss": 0.171, "step": 15209 }, { "epoch": 4.047365620010644, "grad_norm": 0.3703271746635437, "learning_rate": 1.8089688731955542e-08, "loss": 0.1977, "step": 15210 }, { "epoch": 4.047631718999468, "grad_norm": 0.3153410255908966, "learning_rate": 1.8079903006330866e-08, "loss": 0.1916, "step": 15211 }, { "epoch": 4.047897817988292, "grad_norm": 0.29815733432769775, "learning_rate": 1.807011966518751e-08, "loss": 0.1906, "step": 15212 }, { "epoch": 4.048163916977115, "grad_norm": 0.3145900070667267, "learning_rate": 1.8060338708810242e-08, "loss": 0.1918, "step": 15213 }, { "epoch": 4.048430015965939, "grad_norm": 0.326630175113678, "learning_rate": 1.805056013748375e-08, "loss": 0.1933, "step": 15214 }, { "epoch": 4.048696114954764, "grad_norm": 0.3023689389228821, "learning_rate": 1.804078395149269e-08, "loss": 0.1824, "step": 15215 }, { "epoch": 4.048962213943587, "grad_norm": 0.29251447319984436, "learning_rate": 1.8031010151121594e-08, "loss": 0.1867, "step": 15216 }, { "epoch": 4.049228312932411, "grad_norm": 0.29189401865005493, "learning_rate": 1.8021238736654965e-08, "loss": 0.1732, "step": 15217 }, { "epoch": 4.049494411921235, "grad_norm": 0.2867453396320343, "learning_rate": 1.8011469708377246e-08, "loss": 0.1765, "step": 15218 }, { "epoch": 4.049760510910058, "grad_norm": 0.29427221417427063, "learning_rate": 1.800170306657274e-08, "loss": 0.1876, "step": 15219 }, { "epoch": 4.050026609898882, "grad_norm": 0.3196590542793274, "learning_rate": 1.7991938811525743e-08, "loss": 0.1976, "step": 15220 }, { "epoch": 4.0502927088877065, "grad_norm": 0.31055665016174316, "learning_rate": 1.7982176943520477e-08, "loss": 0.2037, "step": 15221 }, { "epoch": 4.05055880787653, "grad_norm": 0.26872536540031433, "learning_rate": 1.7972417462841084e-08, "loss": 0.1676, "step": 15222 }, { "epoch": 4.050824906865354, "grad_norm": 0.30564776062965393, "learning_rate": 1.7962660369771642e-08, "loss": 0.1746, "step": 15223 }, { "epoch": 4.051091005854178, "grad_norm": 0.3068183362483978, "learning_rate": 1.7952905664596174e-08, "loss": 0.1863, "step": 15224 }, { "epoch": 4.051357104843001, "grad_norm": 0.2813256084918976, "learning_rate": 1.794315334759856e-08, "loss": 0.1793, "step": 15225 }, { "epoch": 4.051623203831825, "grad_norm": 0.30489426851272583, "learning_rate": 1.7933403419062687e-08, "loss": 0.1929, "step": 15226 }, { "epoch": 4.0518893028206495, "grad_norm": 0.3004753291606903, "learning_rate": 1.7923655879272392e-08, "loss": 0.206, "step": 15227 }, { "epoch": 4.052155401809473, "grad_norm": 0.29520195722579956, "learning_rate": 1.7913910728511338e-08, "loss": 0.1933, "step": 15228 }, { "epoch": 4.052421500798297, "grad_norm": 0.295751690864563, "learning_rate": 1.790416796706321e-08, "loss": 0.1925, "step": 15229 }, { "epoch": 4.052687599787121, "grad_norm": 0.32368239760398865, "learning_rate": 1.7894427595211592e-08, "loss": 0.1969, "step": 15230 }, { "epoch": 4.052953698775944, "grad_norm": 0.2944517731666565, "learning_rate": 1.788468961323999e-08, "loss": 0.1937, "step": 15231 }, { "epoch": 4.053219797764768, "grad_norm": 0.2890969216823578, "learning_rate": 1.7874954021431877e-08, "loss": 0.1688, "step": 15232 }, { "epoch": 4.0534858967535925, "grad_norm": 0.29277193546295166, "learning_rate": 1.7865220820070614e-08, "loss": 0.1777, "step": 15233 }, { "epoch": 4.053751995742417, "grad_norm": 0.2905530333518982, "learning_rate": 1.785549000943951e-08, "loss": 0.1827, "step": 15234 }, { "epoch": 4.05401809473124, "grad_norm": 0.29731684923171997, "learning_rate": 1.7845761589821807e-08, "loss": 0.194, "step": 15235 }, { "epoch": 4.054284193720064, "grad_norm": 0.3296726942062378, "learning_rate": 1.7836035561500695e-08, "loss": 0.2028, "step": 15236 }, { "epoch": 4.054550292708888, "grad_norm": 0.34926387667655945, "learning_rate": 1.782631192475923e-08, "loss": 0.1803, "step": 15237 }, { "epoch": 4.054816391697711, "grad_norm": 0.2832760214805603, "learning_rate": 1.781659067988046e-08, "loss": 0.1804, "step": 15238 }, { "epoch": 4.0550824906865355, "grad_norm": 0.2625047266483307, "learning_rate": 1.780687182714735e-08, "loss": 0.1577, "step": 15239 }, { "epoch": 4.05534858967536, "grad_norm": 0.2986902594566345, "learning_rate": 1.7797155366842786e-08, "loss": 0.1777, "step": 15240 }, { "epoch": 4.055614688664183, "grad_norm": 0.2893392741680145, "learning_rate": 1.7787441299249594e-08, "loss": 0.1972, "step": 15241 }, { "epoch": 4.055880787653007, "grad_norm": 0.27074548602104187, "learning_rate": 1.7777729624650527e-08, "loss": 0.1758, "step": 15242 }, { "epoch": 4.056146886641831, "grad_norm": 0.3365980386734009, "learning_rate": 1.7768020343328282e-08, "loss": 0.1938, "step": 15243 }, { "epoch": 4.056412985630654, "grad_norm": 0.36590802669525146, "learning_rate": 1.775831345556543e-08, "loss": 0.1781, "step": 15244 }, { "epoch": 4.0566790846194785, "grad_norm": 0.30443257093429565, "learning_rate": 1.7748608961644542e-08, "loss": 0.2052, "step": 15245 }, { "epoch": 4.056945183608303, "grad_norm": 0.3180275857448578, "learning_rate": 1.773890686184809e-08, "loss": 0.178, "step": 15246 }, { "epoch": 4.057211282597126, "grad_norm": 0.2798972725868225, "learning_rate": 1.772920715645846e-08, "loss": 0.1748, "step": 15247 }, { "epoch": 4.05747738158595, "grad_norm": 0.26858800649642944, "learning_rate": 1.7719509845757985e-08, "loss": 0.1825, "step": 15248 }, { "epoch": 4.057743480574774, "grad_norm": 0.314413845539093, "learning_rate": 1.7709814930028956e-08, "loss": 0.2002, "step": 15249 }, { "epoch": 4.058009579563597, "grad_norm": 0.28465768694877625, "learning_rate": 1.770012240955353e-08, "loss": 0.1947, "step": 15250 }, { "epoch": 4.058275678552421, "grad_norm": 0.345344603061676, "learning_rate": 1.7690432284613867e-08, "loss": 0.1946, "step": 15251 }, { "epoch": 4.058541777541246, "grad_norm": 0.3083972930908203, "learning_rate": 1.7680744555492e-08, "loss": 0.1897, "step": 15252 }, { "epoch": 4.058807876530069, "grad_norm": 0.3067673146724701, "learning_rate": 1.767105922246992e-08, "loss": 0.2085, "step": 15253 }, { "epoch": 4.059073975518893, "grad_norm": 0.30104860663414, "learning_rate": 1.7661376285829567e-08, "loss": 0.193, "step": 15254 }, { "epoch": 4.059340074507717, "grad_norm": 0.3369310796260834, "learning_rate": 1.7651695745852724e-08, "loss": 0.2052, "step": 15255 }, { "epoch": 4.05960617349654, "grad_norm": 0.3081555962562561, "learning_rate": 1.7642017602821214e-08, "loss": 0.1969, "step": 15256 }, { "epoch": 4.059872272485364, "grad_norm": 0.30703839659690857, "learning_rate": 1.763234185701673e-08, "loss": 0.1896, "step": 15257 }, { "epoch": 4.0601383714741885, "grad_norm": 0.2994717061519623, "learning_rate": 1.7622668508720906e-08, "loss": 0.1998, "step": 15258 }, { "epoch": 4.060404470463013, "grad_norm": 0.28287121653556824, "learning_rate": 1.7612997558215315e-08, "loss": 0.1743, "step": 15259 }, { "epoch": 4.060670569451836, "grad_norm": 0.27832451462745667, "learning_rate": 1.760332900578144e-08, "loss": 0.1877, "step": 15260 }, { "epoch": 4.06093666844066, "grad_norm": 0.29359668493270874, "learning_rate": 1.759366285170073e-08, "loss": 0.1873, "step": 15261 }, { "epoch": 4.061202767429484, "grad_norm": 0.3099531829357147, "learning_rate": 1.758399909625452e-08, "loss": 0.1872, "step": 15262 }, { "epoch": 4.061468866418307, "grad_norm": 0.2915686070919037, "learning_rate": 1.757433773972413e-08, "loss": 0.1795, "step": 15263 }, { "epoch": 4.0617349654071315, "grad_norm": 0.2803097367286682, "learning_rate": 1.756467878239074e-08, "loss": 0.1717, "step": 15264 }, { "epoch": 4.062001064395956, "grad_norm": 0.32062944769859314, "learning_rate": 1.7555022224535498e-08, "loss": 0.2013, "step": 15265 }, { "epoch": 4.062267163384779, "grad_norm": 0.2776228189468384, "learning_rate": 1.7545368066439515e-08, "loss": 0.1836, "step": 15266 }, { "epoch": 4.062533262373603, "grad_norm": 0.3153837025165558, "learning_rate": 1.7535716308383763e-08, "loss": 0.1861, "step": 15267 }, { "epoch": 4.062799361362427, "grad_norm": 0.28712502121925354, "learning_rate": 1.7526066950649176e-08, "loss": 0.1721, "step": 15268 }, { "epoch": 4.06306546035125, "grad_norm": 0.34527069330215454, "learning_rate": 1.7516419993516652e-08, "loss": 0.2072, "step": 15269 }, { "epoch": 4.0633315593400745, "grad_norm": 0.30686986446380615, "learning_rate": 1.7506775437266975e-08, "loss": 0.1922, "step": 15270 }, { "epoch": 4.063597658328899, "grad_norm": 0.3126676678657532, "learning_rate": 1.7497133282180864e-08, "loss": 0.1964, "step": 15271 }, { "epoch": 4.063863757317722, "grad_norm": 0.29571402072906494, "learning_rate": 1.748749352853902e-08, "loss": 0.1869, "step": 15272 }, { "epoch": 4.064129856306546, "grad_norm": 0.3137513995170593, "learning_rate": 1.7477856176621973e-08, "loss": 0.2152, "step": 15273 }, { "epoch": 4.06439595529537, "grad_norm": 0.28237950801849365, "learning_rate": 1.7468221226710266e-08, "loss": 0.1652, "step": 15274 }, { "epoch": 4.064662054284193, "grad_norm": 0.2890740633010864, "learning_rate": 1.7458588679084342e-08, "loss": 0.1829, "step": 15275 }, { "epoch": 4.0649281532730175, "grad_norm": 0.2732013165950775, "learning_rate": 1.7448958534024593e-08, "loss": 0.1835, "step": 15276 }, { "epoch": 4.065194252261842, "grad_norm": 0.3204996883869171, "learning_rate": 1.7439330791811313e-08, "loss": 0.1707, "step": 15277 }, { "epoch": 4.065460351250665, "grad_norm": 0.3159067928791046, "learning_rate": 1.7429705452724754e-08, "loss": 0.218, "step": 15278 }, { "epoch": 4.065726450239489, "grad_norm": 0.2892514765262604, "learning_rate": 1.7420082517045064e-08, "loss": 0.188, "step": 15279 }, { "epoch": 4.065992549228313, "grad_norm": 0.2777458131313324, "learning_rate": 1.741046198505236e-08, "loss": 0.1927, "step": 15280 }, { "epoch": 4.066258648217136, "grad_norm": 0.28492218255996704, "learning_rate": 1.7400843857026703e-08, "loss": 0.1732, "step": 15281 }, { "epoch": 4.0665247472059605, "grad_norm": 0.3073310852050781, "learning_rate": 1.739122813324798e-08, "loss": 0.1993, "step": 15282 }, { "epoch": 4.066790846194785, "grad_norm": 0.299784392118454, "learning_rate": 1.7381614813996114e-08, "loss": 0.1982, "step": 15283 }, { "epoch": 4.067056945183609, "grad_norm": 0.2951756417751312, "learning_rate": 1.7372003899550925e-08, "loss": 0.1909, "step": 15284 }, { "epoch": 4.067323044172432, "grad_norm": 0.2854057252407074, "learning_rate": 1.7362395390192163e-08, "loss": 0.1881, "step": 15285 }, { "epoch": 4.067589143161256, "grad_norm": 0.2848786413669586, "learning_rate": 1.735278928619953e-08, "loss": 0.184, "step": 15286 }, { "epoch": 4.06785524215008, "grad_norm": 0.3055401146411896, "learning_rate": 1.7343185587852583e-08, "loss": 0.1913, "step": 15287 }, { "epoch": 4.0681213411389034, "grad_norm": 0.3222517967224121, "learning_rate": 1.7333584295430892e-08, "loss": 0.1976, "step": 15288 }, { "epoch": 4.068387440127728, "grad_norm": 0.26692500710487366, "learning_rate": 1.7323985409213914e-08, "loss": 0.1778, "step": 15289 }, { "epoch": 4.068653539116552, "grad_norm": 0.29531970620155334, "learning_rate": 1.7314388929481083e-08, "loss": 0.1951, "step": 15290 }, { "epoch": 4.068919638105375, "grad_norm": 0.30617234110832214, "learning_rate": 1.7304794856511672e-08, "loss": 0.1874, "step": 15291 }, { "epoch": 4.069185737094199, "grad_norm": 0.30004340410232544, "learning_rate": 1.7295203190584973e-08, "loss": 0.2028, "step": 15292 }, { "epoch": 4.069451836083023, "grad_norm": 0.30369114875793457, "learning_rate": 1.728561393198016e-08, "loss": 0.1953, "step": 15293 }, { "epoch": 4.069717935071846, "grad_norm": 0.29006320238113403, "learning_rate": 1.7276027080976364e-08, "loss": 0.1832, "step": 15294 }, { "epoch": 4.0699840340606706, "grad_norm": 0.38189759850502014, "learning_rate": 1.7266442637852628e-08, "loss": 0.2115, "step": 15295 }, { "epoch": 4.070250133049495, "grad_norm": 0.31075620651245117, "learning_rate": 1.7256860602887935e-08, "loss": 0.1864, "step": 15296 }, { "epoch": 4.070516232038318, "grad_norm": 0.29199957847595215, "learning_rate": 1.724728097636119e-08, "loss": 0.1842, "step": 15297 }, { "epoch": 4.070782331027142, "grad_norm": 0.3018523156642914, "learning_rate": 1.7237703758551224e-08, "loss": 0.1935, "step": 15298 }, { "epoch": 4.071048430015966, "grad_norm": 0.33097904920578003, "learning_rate": 1.7228128949736842e-08, "loss": 0.1803, "step": 15299 }, { "epoch": 4.071314529004789, "grad_norm": 0.28751394152641296, "learning_rate": 1.7218556550196684e-08, "loss": 0.1849, "step": 15300 }, { "epoch": 4.0715806279936135, "grad_norm": 0.31693628430366516, "learning_rate": 1.7208986560209404e-08, "loss": 0.1963, "step": 15301 }, { "epoch": 4.071846726982438, "grad_norm": 0.3676755428314209, "learning_rate": 1.719941898005356e-08, "loss": 0.21, "step": 15302 }, { "epoch": 4.072112825971261, "grad_norm": 0.32023149728775024, "learning_rate": 1.7189853810007638e-08, "loss": 0.2063, "step": 15303 }, { "epoch": 4.072378924960085, "grad_norm": 0.31329426169395447, "learning_rate": 1.718029105035006e-08, "loss": 0.1949, "step": 15304 }, { "epoch": 4.072645023948909, "grad_norm": 0.300643652677536, "learning_rate": 1.7170730701359204e-08, "loss": 0.2009, "step": 15305 }, { "epoch": 4.072911122937732, "grad_norm": 0.2950117290019989, "learning_rate": 1.716117276331327e-08, "loss": 0.2057, "step": 15306 }, { "epoch": 4.0731772219265565, "grad_norm": 0.2650645077228546, "learning_rate": 1.7151617236490524e-08, "loss": 0.1774, "step": 15307 }, { "epoch": 4.073443320915381, "grad_norm": 0.3155284523963928, "learning_rate": 1.7142064121169108e-08, "loss": 0.1996, "step": 15308 }, { "epoch": 4.073709419904205, "grad_norm": 0.30053067207336426, "learning_rate": 1.7132513417627036e-08, "loss": 0.1913, "step": 15309 }, { "epoch": 4.073975518893028, "grad_norm": 0.3017136752605438, "learning_rate": 1.7122965126142343e-08, "loss": 0.1857, "step": 15310 }, { "epoch": 4.074241617881852, "grad_norm": 0.3041384518146515, "learning_rate": 1.7113419246992945e-08, "loss": 0.1892, "step": 15311 }, { "epoch": 4.074507716870676, "grad_norm": 0.28098952770233154, "learning_rate": 1.710387578045669e-08, "loss": 0.1669, "step": 15312 }, { "epoch": 4.0747738158594995, "grad_norm": 0.28892767429351807, "learning_rate": 1.7094334726811375e-08, "loss": 0.1806, "step": 15313 }, { "epoch": 4.075039914848324, "grad_norm": 0.32506322860717773, "learning_rate": 1.7084796086334728e-08, "loss": 0.1954, "step": 15314 }, { "epoch": 4.075306013837148, "grad_norm": 0.42472726106643677, "learning_rate": 1.707525985930437e-08, "loss": 0.1908, "step": 15315 }, { "epoch": 4.075572112825971, "grad_norm": 0.31065571308135986, "learning_rate": 1.7065726045997887e-08, "loss": 0.1889, "step": 15316 }, { "epoch": 4.075838211814795, "grad_norm": 0.2949526607990265, "learning_rate": 1.7056194646692813e-08, "loss": 0.1962, "step": 15317 }, { "epoch": 4.076104310803619, "grad_norm": 0.2972749173641205, "learning_rate": 1.7046665661666516e-08, "loss": 0.1958, "step": 15318 }, { "epoch": 4.0763704097924425, "grad_norm": 0.29111090302467346, "learning_rate": 1.7037139091196395e-08, "loss": 0.1859, "step": 15319 }, { "epoch": 4.076636508781267, "grad_norm": 0.2861417531967163, "learning_rate": 1.7027614935559753e-08, "loss": 0.1912, "step": 15320 }, { "epoch": 4.076902607770091, "grad_norm": 0.3055020570755005, "learning_rate": 1.70180931950338e-08, "loss": 0.2013, "step": 15321 }, { "epoch": 4.077168706758914, "grad_norm": 0.2679222524166107, "learning_rate": 1.7008573869895693e-08, "loss": 0.1697, "step": 15322 }, { "epoch": 4.077434805747738, "grad_norm": 0.2807113230228424, "learning_rate": 1.699905696042252e-08, "loss": 0.1703, "step": 15323 }, { "epoch": 4.077700904736562, "grad_norm": 0.3024751842021942, "learning_rate": 1.6989542466891282e-08, "loss": 0.1814, "step": 15324 }, { "epoch": 4.0779670037253855, "grad_norm": 0.2817886769771576, "learning_rate": 1.6980030389578957e-08, "loss": 0.1841, "step": 15325 }, { "epoch": 4.07823310271421, "grad_norm": 0.32659944891929626, "learning_rate": 1.697052072876237e-08, "loss": 0.2062, "step": 15326 }, { "epoch": 4.078499201703034, "grad_norm": 0.34372520446777344, "learning_rate": 1.6961013484718333e-08, "loss": 0.1862, "step": 15327 }, { "epoch": 4.078765300691857, "grad_norm": 0.28533631563186646, "learning_rate": 1.6951508657723612e-08, "loss": 0.1975, "step": 15328 }, { "epoch": 4.079031399680681, "grad_norm": 0.31851738691329956, "learning_rate": 1.694200624805482e-08, "loss": 0.1885, "step": 15329 }, { "epoch": 4.079297498669505, "grad_norm": 0.27904263138771057, "learning_rate": 1.6932506255988555e-08, "loss": 0.1728, "step": 15330 }, { "epoch": 4.079563597658329, "grad_norm": 0.3434476852416992, "learning_rate": 1.6923008681801365e-08, "loss": 0.178, "step": 15331 }, { "epoch": 4.079829696647153, "grad_norm": 0.27314653992652893, "learning_rate": 1.6913513525769674e-08, "loss": 0.1734, "step": 15332 }, { "epoch": 4.080095795635977, "grad_norm": 0.29878491163253784, "learning_rate": 1.690402078816988e-08, "loss": 0.2022, "step": 15333 }, { "epoch": 4.080361894624801, "grad_norm": 0.2973249554634094, "learning_rate": 1.6894530469278278e-08, "loss": 0.1943, "step": 15334 }, { "epoch": 4.080627993613624, "grad_norm": 0.2824706733226776, "learning_rate": 1.6885042569371145e-08, "loss": 0.1854, "step": 15335 }, { "epoch": 4.080894092602448, "grad_norm": 0.27916741371154785, "learning_rate": 1.687555708872459e-08, "loss": 0.1678, "step": 15336 }, { "epoch": 4.081160191591272, "grad_norm": 0.29746586084365845, "learning_rate": 1.686607402761473e-08, "loss": 0.1966, "step": 15337 }, { "epoch": 4.0814262905800955, "grad_norm": 0.32795578241348267, "learning_rate": 1.6856593386317608e-08, "loss": 0.1999, "step": 15338 }, { "epoch": 4.08169238956892, "grad_norm": 0.33560264110565186, "learning_rate": 1.684711516510917e-08, "loss": 0.1905, "step": 15339 }, { "epoch": 4.081958488557744, "grad_norm": 0.29886946082115173, "learning_rate": 1.683763936426531e-08, "loss": 0.1869, "step": 15340 }, { "epoch": 4.082224587546567, "grad_norm": 0.3121001422405243, "learning_rate": 1.6828165984061825e-08, "loss": 0.205, "step": 15341 }, { "epoch": 4.082490686535391, "grad_norm": 0.30078086256980896, "learning_rate": 1.6818695024774488e-08, "loss": 0.2, "step": 15342 }, { "epoch": 4.082756785524215, "grad_norm": 0.2935734689235687, "learning_rate": 1.6809226486678964e-08, "loss": 0.1755, "step": 15343 }, { "epoch": 4.0830228845130385, "grad_norm": 0.27878832817077637, "learning_rate": 1.6799760370050875e-08, "loss": 0.1774, "step": 15344 }, { "epoch": 4.083288983501863, "grad_norm": 0.2914532423019409, "learning_rate": 1.6790296675165706e-08, "loss": 0.178, "step": 15345 }, { "epoch": 4.083555082490687, "grad_norm": 0.27750465273857117, "learning_rate": 1.6780835402298944e-08, "loss": 0.1649, "step": 15346 }, { "epoch": 4.08382118147951, "grad_norm": 0.282744824886322, "learning_rate": 1.6771376551726023e-08, "loss": 0.1826, "step": 15347 }, { "epoch": 4.084087280468334, "grad_norm": 0.3091132640838623, "learning_rate": 1.6761920123722194e-08, "loss": 0.1805, "step": 15348 }, { "epoch": 4.084353379457158, "grad_norm": 0.2868088185787201, "learning_rate": 1.6752466118562758e-08, "loss": 0.1772, "step": 15349 }, { "epoch": 4.0846194784459815, "grad_norm": 0.47101545333862305, "learning_rate": 1.674301453652287e-08, "loss": 0.1822, "step": 15350 }, { "epoch": 4.084885577434806, "grad_norm": 0.29432299733161926, "learning_rate": 1.6733565377877667e-08, "loss": 0.1761, "step": 15351 }, { "epoch": 4.08515167642363, "grad_norm": 0.32416486740112305, "learning_rate": 1.6724118642902162e-08, "loss": 0.2009, "step": 15352 }, { "epoch": 4.085417775412454, "grad_norm": 0.3207657039165497, "learning_rate": 1.6714674331871347e-08, "loss": 0.193, "step": 15353 }, { "epoch": 4.085683874401277, "grad_norm": 0.30223602056503296, "learning_rate": 1.6705232445060145e-08, "loss": 0.19, "step": 15354 }, { "epoch": 4.085949973390101, "grad_norm": 0.32026898860931396, "learning_rate": 1.6695792982743318e-08, "loss": 0.1999, "step": 15355 }, { "epoch": 4.086216072378925, "grad_norm": 0.3084488809108734, "learning_rate": 1.6686355945195674e-08, "loss": 0.1925, "step": 15356 }, { "epoch": 4.086482171367749, "grad_norm": 0.3425135612487793, "learning_rate": 1.6676921332691873e-08, "loss": 0.2096, "step": 15357 }, { "epoch": 4.086748270356573, "grad_norm": 0.3042745888233185, "learning_rate": 1.6667489145506552e-08, "loss": 0.2024, "step": 15358 }, { "epoch": 4.087014369345397, "grad_norm": 0.30046576261520386, "learning_rate": 1.6658059383914247e-08, "loss": 0.2016, "step": 15359 }, { "epoch": 4.08728046833422, "grad_norm": 0.31360766291618347, "learning_rate": 1.664863204818945e-08, "loss": 0.1989, "step": 15360 }, { "epoch": 4.087546567323044, "grad_norm": 0.3025093674659729, "learning_rate": 1.6639207138606537e-08, "loss": 0.1945, "step": 15361 }, { "epoch": 4.087812666311868, "grad_norm": 0.30095937848091125, "learning_rate": 1.6629784655439873e-08, "loss": 0.1863, "step": 15362 }, { "epoch": 4.088078765300692, "grad_norm": 0.3099764287471771, "learning_rate": 1.662036459896373e-08, "loss": 0.1862, "step": 15363 }, { "epoch": 4.088344864289516, "grad_norm": 0.29848310351371765, "learning_rate": 1.6610946969452255e-08, "loss": 0.1921, "step": 15364 }, { "epoch": 4.08861096327834, "grad_norm": 0.3186548352241516, "learning_rate": 1.66015317671796e-08, "loss": 0.2092, "step": 15365 }, { "epoch": 4.088877062267163, "grad_norm": 0.2884225845336914, "learning_rate": 1.6592118992419802e-08, "loss": 0.1944, "step": 15366 }, { "epoch": 4.089143161255987, "grad_norm": 0.28699713945388794, "learning_rate": 1.6582708645446885e-08, "loss": 0.1724, "step": 15367 }, { "epoch": 4.089409260244811, "grad_norm": 0.31919029355049133, "learning_rate": 1.657330072653471e-08, "loss": 0.2018, "step": 15368 }, { "epoch": 4.089675359233635, "grad_norm": 0.2915687561035156, "learning_rate": 1.6563895235957126e-08, "loss": 0.1836, "step": 15369 }, { "epoch": 4.089941458222459, "grad_norm": 0.28667721152305603, "learning_rate": 1.655449217398792e-08, "loss": 0.1774, "step": 15370 }, { "epoch": 4.090207557211283, "grad_norm": 0.2931099236011505, "learning_rate": 1.654509154090078e-08, "loss": 0.1927, "step": 15371 }, { "epoch": 4.090473656200106, "grad_norm": 0.29491621255874634, "learning_rate": 1.6535693336969358e-08, "loss": 0.206, "step": 15372 }, { "epoch": 4.09073975518893, "grad_norm": 0.3045044243335724, "learning_rate": 1.6526297562467173e-08, "loss": 0.2062, "step": 15373 }, { "epoch": 4.091005854177754, "grad_norm": 0.2976183593273163, "learning_rate": 1.6516904217667725e-08, "loss": 0.1823, "step": 15374 }, { "epoch": 4.0912719531665775, "grad_norm": 0.30483102798461914, "learning_rate": 1.6507513302844433e-08, "loss": 0.1798, "step": 15375 }, { "epoch": 4.091538052155402, "grad_norm": 0.3038092851638794, "learning_rate": 1.6498124818270643e-08, "loss": 0.2114, "step": 15376 }, { "epoch": 4.091804151144226, "grad_norm": 0.3097550570964813, "learning_rate": 1.6488738764219623e-08, "loss": 0.1866, "step": 15377 }, { "epoch": 4.09207025013305, "grad_norm": 0.2984451651573181, "learning_rate": 1.6479355140964592e-08, "loss": 0.1682, "step": 15378 }, { "epoch": 4.092336349121873, "grad_norm": 0.3015495538711548, "learning_rate": 1.646997394877867e-08, "loss": 0.1754, "step": 15379 }, { "epoch": 4.092602448110697, "grad_norm": 0.3708617389202118, "learning_rate": 1.6460595187934923e-08, "loss": 0.1969, "step": 15380 }, { "epoch": 4.092868547099521, "grad_norm": 0.2886691689491272, "learning_rate": 1.645121885870637e-08, "loss": 0.1817, "step": 15381 }, { "epoch": 4.093134646088345, "grad_norm": 0.28881874680519104, "learning_rate": 1.644184496136587e-08, "loss": 0.1699, "step": 15382 }, { "epoch": 4.093400745077169, "grad_norm": 0.2905147671699524, "learning_rate": 1.6432473496186315e-08, "loss": 0.1859, "step": 15383 }, { "epoch": 4.093666844065993, "grad_norm": 0.2813310921192169, "learning_rate": 1.6423104463440473e-08, "loss": 0.174, "step": 15384 }, { "epoch": 4.093932943054816, "grad_norm": 0.34608086943626404, "learning_rate": 1.6413737863401054e-08, "loss": 0.214, "step": 15385 }, { "epoch": 4.09419904204364, "grad_norm": 0.3168424069881439, "learning_rate": 1.64043736963407e-08, "loss": 0.198, "step": 15386 }, { "epoch": 4.094465141032464, "grad_norm": 0.2881380319595337, "learning_rate": 1.639501196253199e-08, "loss": 0.1723, "step": 15387 }, { "epoch": 4.094731240021288, "grad_norm": 0.3260376453399658, "learning_rate": 1.6385652662247383e-08, "loss": 0.1972, "step": 15388 }, { "epoch": 4.094997339010112, "grad_norm": 0.28944286704063416, "learning_rate": 1.6376295795759332e-08, "loss": 0.2031, "step": 15389 }, { "epoch": 4.095263437998936, "grad_norm": 0.28999796509742737, "learning_rate": 1.6366941363340204e-08, "loss": 0.1728, "step": 15390 }, { "epoch": 4.095529536987759, "grad_norm": 0.27825459837913513, "learning_rate": 1.6357589365262237e-08, "loss": 0.1933, "step": 15391 }, { "epoch": 4.095795635976583, "grad_norm": 0.2858765125274658, "learning_rate": 1.6348239801797658e-08, "loss": 0.1862, "step": 15392 }, { "epoch": 4.096061734965407, "grad_norm": 0.2986477315425873, "learning_rate": 1.633889267321863e-08, "loss": 0.1975, "step": 15393 }, { "epoch": 4.096327833954231, "grad_norm": 0.321044921875, "learning_rate": 1.6329547979797208e-08, "loss": 0.1898, "step": 15394 }, { "epoch": 4.096593932943055, "grad_norm": 0.2936387062072754, "learning_rate": 1.6320205721805402e-08, "loss": 0.1846, "step": 15395 }, { "epoch": 4.096860031931879, "grad_norm": 0.31803691387176514, "learning_rate": 1.6310865899515135e-08, "loss": 0.2052, "step": 15396 }, { "epoch": 4.097126130920702, "grad_norm": 0.28247761726379395, "learning_rate": 1.630152851319826e-08, "loss": 0.1853, "step": 15397 }, { "epoch": 4.097392229909526, "grad_norm": 0.30344632267951965, "learning_rate": 1.6292193563126565e-08, "loss": 0.1965, "step": 15398 }, { "epoch": 4.09765832889835, "grad_norm": 0.3073122203350067, "learning_rate": 1.6282861049571805e-08, "loss": 0.1887, "step": 15399 }, { "epoch": 4.097924427887174, "grad_norm": 0.3131873309612274, "learning_rate": 1.627353097280556e-08, "loss": 0.2038, "step": 15400 }, { "epoch": 4.098190526875998, "grad_norm": 0.29412856698036194, "learning_rate": 1.6264203333099437e-08, "loss": 0.1911, "step": 15401 }, { "epoch": 4.098456625864822, "grad_norm": 0.30144643783569336, "learning_rate": 1.6254878130724937e-08, "loss": 0.2003, "step": 15402 }, { "epoch": 4.098722724853646, "grad_norm": 0.2767614424228668, "learning_rate": 1.624555536595349e-08, "loss": 0.173, "step": 15403 }, { "epoch": 4.098988823842469, "grad_norm": 0.27897515892982483, "learning_rate": 1.6236235039056457e-08, "loss": 0.1834, "step": 15404 }, { "epoch": 4.099254922831293, "grad_norm": 0.3170761168003082, "learning_rate": 1.6226917150305143e-08, "loss": 0.1927, "step": 15405 }, { "epoch": 4.0995210218201175, "grad_norm": 0.2954314649105072, "learning_rate": 1.6217601699970775e-08, "loss": 0.1953, "step": 15406 }, { "epoch": 4.099787120808941, "grad_norm": 0.2983807921409607, "learning_rate": 1.6208288688324458e-08, "loss": 0.1885, "step": 15407 }, { "epoch": 4.100053219797765, "grad_norm": 0.3005152642726898, "learning_rate": 1.6198978115637287e-08, "loss": 0.1885, "step": 15408 }, { "epoch": 4.100319318786589, "grad_norm": 0.2923549711704254, "learning_rate": 1.6189669982180308e-08, "loss": 0.1702, "step": 15409 }, { "epoch": 4.100585417775412, "grad_norm": 0.34130173921585083, "learning_rate": 1.61803642882244e-08, "loss": 0.1957, "step": 15410 }, { "epoch": 4.100851516764236, "grad_norm": 0.30889248847961426, "learning_rate": 1.6171061034040444e-08, "loss": 0.1873, "step": 15411 }, { "epoch": 4.10111761575306, "grad_norm": 0.2790437936782837, "learning_rate": 1.6161760219899258e-08, "loss": 0.1716, "step": 15412 }, { "epoch": 4.101383714741884, "grad_norm": 0.3060436248779297, "learning_rate": 1.615246184607153e-08, "loss": 0.1869, "step": 15413 }, { "epoch": 4.101649813730708, "grad_norm": 0.29730847477912903, "learning_rate": 1.614316591282794e-08, "loss": 0.1965, "step": 15414 }, { "epoch": 4.101915912719532, "grad_norm": 0.3114114999771118, "learning_rate": 1.6133872420439054e-08, "loss": 0.1972, "step": 15415 }, { "epoch": 4.102182011708355, "grad_norm": 0.3015221953392029, "learning_rate": 1.6124581369175395e-08, "loss": 0.2002, "step": 15416 }, { "epoch": 4.102448110697179, "grad_norm": 0.2740911543369293, "learning_rate": 1.611529275930741e-08, "loss": 0.1778, "step": 15417 }, { "epoch": 4.102714209686003, "grad_norm": 0.2993439733982086, "learning_rate": 1.6106006591105427e-08, "loss": 0.1881, "step": 15418 }, { "epoch": 4.102980308674827, "grad_norm": 0.41645100712776184, "learning_rate": 1.6096722864839753e-08, "loss": 0.1843, "step": 15419 }, { "epoch": 4.103246407663651, "grad_norm": 0.29909083247184753, "learning_rate": 1.6087441580780636e-08, "loss": 0.2011, "step": 15420 }, { "epoch": 4.103512506652475, "grad_norm": 0.31329217553138733, "learning_rate": 1.6078162739198207e-08, "loss": 0.1968, "step": 15421 }, { "epoch": 4.103778605641298, "grad_norm": 0.2892545461654663, "learning_rate": 1.6068886340362564e-08, "loss": 0.1913, "step": 15422 }, { "epoch": 4.104044704630122, "grad_norm": 0.2824421525001526, "learning_rate": 1.605961238454372e-08, "loss": 0.1864, "step": 15423 }, { "epoch": 4.104310803618946, "grad_norm": 0.30826741456985474, "learning_rate": 1.6050340872011602e-08, "loss": 0.1943, "step": 15424 }, { "epoch": 4.10457690260777, "grad_norm": 0.2922345995903015, "learning_rate": 1.6041071803036098e-08, "loss": 0.1841, "step": 15425 }, { "epoch": 4.104843001596594, "grad_norm": 0.29083052277565, "learning_rate": 1.6031805177887015e-08, "loss": 0.1814, "step": 15426 }, { "epoch": 4.105109100585418, "grad_norm": 0.3166593015193939, "learning_rate": 1.6022540996834045e-08, "loss": 0.1951, "step": 15427 }, { "epoch": 4.105375199574242, "grad_norm": 0.3068024218082428, "learning_rate": 1.601327926014685e-08, "loss": 0.1697, "step": 15428 }, { "epoch": 4.105641298563065, "grad_norm": 0.30170661211013794, "learning_rate": 1.600401996809505e-08, "loss": 0.2169, "step": 15429 }, { "epoch": 4.105907397551889, "grad_norm": 0.3144501745700836, "learning_rate": 1.5994763120948118e-08, "loss": 0.198, "step": 15430 }, { "epoch": 4.1061734965407135, "grad_norm": 0.2944015860557556, "learning_rate": 1.5985508718975516e-08, "loss": 0.1899, "step": 15431 }, { "epoch": 4.106439595529537, "grad_norm": 0.3051125705242157, "learning_rate": 1.5976256762446605e-08, "loss": 0.1808, "step": 15432 }, { "epoch": 4.106705694518361, "grad_norm": 0.27089595794677734, "learning_rate": 1.5967007251630704e-08, "loss": 0.164, "step": 15433 }, { "epoch": 4.106971793507185, "grad_norm": 0.2792819142341614, "learning_rate": 1.5957760186797032e-08, "loss": 0.1694, "step": 15434 }, { "epoch": 4.107237892496008, "grad_norm": 0.3330535292625427, "learning_rate": 1.5948515568214758e-08, "loss": 0.2206, "step": 15435 }, { "epoch": 4.107503991484832, "grad_norm": 0.301670640707016, "learning_rate": 1.593927339615295e-08, "loss": 0.1851, "step": 15436 }, { "epoch": 4.1077700904736565, "grad_norm": 0.3016396462917328, "learning_rate": 1.593003367088063e-08, "loss": 0.212, "step": 15437 }, { "epoch": 4.10803618946248, "grad_norm": 0.30856868624687195, "learning_rate": 1.5920796392666735e-08, "loss": 0.1954, "step": 15438 }, { "epoch": 4.108302288451304, "grad_norm": 0.34246334433555603, "learning_rate": 1.5911561561780153e-08, "loss": 0.2133, "step": 15439 }, { "epoch": 4.108568387440128, "grad_norm": 0.2837793529033661, "learning_rate": 1.5902329178489682e-08, "loss": 0.1751, "step": 15440 }, { "epoch": 4.108834486428951, "grad_norm": 0.3231039345264435, "learning_rate": 1.5893099243064046e-08, "loss": 0.2047, "step": 15441 }, { "epoch": 4.109100585417775, "grad_norm": 0.32983946800231934, "learning_rate": 1.5883871755771917e-08, "loss": 0.2047, "step": 15442 }, { "epoch": 4.1093666844065995, "grad_norm": 0.3108147978782654, "learning_rate": 1.5874646716881866e-08, "loss": 0.1993, "step": 15443 }, { "epoch": 4.109632783395423, "grad_norm": 0.29532232880592346, "learning_rate": 1.586542412666245e-08, "loss": 0.1933, "step": 15444 }, { "epoch": 4.109898882384247, "grad_norm": 0.2764620780944824, "learning_rate": 1.5856203985382066e-08, "loss": 0.1812, "step": 15445 }, { "epoch": 4.110164981373071, "grad_norm": 0.3051339387893677, "learning_rate": 1.5846986293309095e-08, "loss": 0.2017, "step": 15446 }, { "epoch": 4.110431080361894, "grad_norm": 0.29497143626213074, "learning_rate": 1.5837771050711868e-08, "loss": 0.1845, "step": 15447 }, { "epoch": 4.110697179350718, "grad_norm": 0.28486162424087524, "learning_rate": 1.5828558257858604e-08, "loss": 0.1819, "step": 15448 }, { "epoch": 4.1109632783395424, "grad_norm": 0.28886398673057556, "learning_rate": 1.5819347915017455e-08, "loss": 0.1829, "step": 15449 }, { "epoch": 4.111229377328367, "grad_norm": 0.2931828796863556, "learning_rate": 1.581014002245651e-08, "loss": 0.1994, "step": 15450 }, { "epoch": 4.11149547631719, "grad_norm": 0.2963891327381134, "learning_rate": 1.5800934580443782e-08, "loss": 0.2053, "step": 15451 }, { "epoch": 4.111761575306014, "grad_norm": 0.3648051917552948, "learning_rate": 1.5791731589247238e-08, "loss": 0.179, "step": 15452 }, { "epoch": 4.112027674294838, "grad_norm": 0.32377415895462036, "learning_rate": 1.5782531049134762e-08, "loss": 0.1884, "step": 15453 }, { "epoch": 4.112293773283661, "grad_norm": 0.27972611784935, "learning_rate": 1.577333296037412e-08, "loss": 0.182, "step": 15454 }, { "epoch": 4.112559872272485, "grad_norm": 0.3106258809566498, "learning_rate": 1.5764137323233052e-08, "loss": 0.1907, "step": 15455 }, { "epoch": 4.1128259712613096, "grad_norm": 0.28703540563583374, "learning_rate": 1.575494413797924e-08, "loss": 0.1803, "step": 15456 }, { "epoch": 4.113092070250133, "grad_norm": 0.3127412796020508, "learning_rate": 1.5745753404880268e-08, "loss": 0.1991, "step": 15457 }, { "epoch": 4.113358169238957, "grad_norm": 0.3069339692592621, "learning_rate": 1.5736565124203638e-08, "loss": 0.2027, "step": 15458 }, { "epoch": 4.113624268227781, "grad_norm": 0.3244415819644928, "learning_rate": 1.5727379296216813e-08, "loss": 0.2021, "step": 15459 }, { "epoch": 4.113890367216604, "grad_norm": 0.29736757278442383, "learning_rate": 1.5718195921187173e-08, "loss": 0.1907, "step": 15460 }, { "epoch": 4.114156466205428, "grad_norm": 0.3051803708076477, "learning_rate": 1.570901499938201e-08, "loss": 0.1932, "step": 15461 }, { "epoch": 4.1144225651942525, "grad_norm": 0.3453417122364044, "learning_rate": 1.569983653106859e-08, "loss": 0.2009, "step": 15462 }, { "epoch": 4.114688664183076, "grad_norm": 0.327569842338562, "learning_rate": 1.569066051651401e-08, "loss": 0.1955, "step": 15463 }, { "epoch": 4.1149547631719, "grad_norm": 0.2998402416706085, "learning_rate": 1.56814869559854e-08, "loss": 0.1909, "step": 15464 }, { "epoch": 4.115220862160724, "grad_norm": 0.2835899591445923, "learning_rate": 1.5672315849749773e-08, "loss": 0.169, "step": 15465 }, { "epoch": 4.115486961149547, "grad_norm": 0.29179054498672485, "learning_rate": 1.566314719807408e-08, "loss": 0.1948, "step": 15466 }, { "epoch": 4.115753060138371, "grad_norm": 0.320586621761322, "learning_rate": 1.5653981001225193e-08, "loss": 0.1979, "step": 15467 }, { "epoch": 4.1160191591271955, "grad_norm": 0.32384008169174194, "learning_rate": 1.5644817259469945e-08, "loss": 0.2029, "step": 15468 }, { "epoch": 4.116285258116019, "grad_norm": 0.2992018759250641, "learning_rate": 1.563565597307501e-08, "loss": 0.1973, "step": 15469 }, { "epoch": 4.116551357104843, "grad_norm": 0.31278711557388306, "learning_rate": 1.5626497142307084e-08, "loss": 0.1814, "step": 15470 }, { "epoch": 4.116817456093667, "grad_norm": 0.2719525992870331, "learning_rate": 1.5617340767432774e-08, "loss": 0.178, "step": 15471 }, { "epoch": 4.117083555082491, "grad_norm": 0.285255491733551, "learning_rate": 1.5608186848718552e-08, "loss": 0.1946, "step": 15472 }, { "epoch": 4.117349654071314, "grad_norm": 0.28153204917907715, "learning_rate": 1.559903538643089e-08, "loss": 0.1769, "step": 15473 }, { "epoch": 4.1176157530601385, "grad_norm": 0.31292420625686646, "learning_rate": 1.558988638083616e-08, "loss": 0.1828, "step": 15474 }, { "epoch": 4.117881852048963, "grad_norm": 0.2935466170310974, "learning_rate": 1.5580739832200674e-08, "loss": 0.1989, "step": 15475 }, { "epoch": 4.118147951037786, "grad_norm": 0.2914458215236664, "learning_rate": 1.5571595740790656e-08, "loss": 0.1803, "step": 15476 }, { "epoch": 4.11841405002661, "grad_norm": 0.322081059217453, "learning_rate": 1.556245410687228e-08, "loss": 0.1998, "step": 15477 }, { "epoch": 4.118680149015434, "grad_norm": 0.3062897324562073, "learning_rate": 1.5553314930711615e-08, "loss": 0.191, "step": 15478 }, { "epoch": 4.118946248004257, "grad_norm": 0.3167594075202942, "learning_rate": 1.5544178212574688e-08, "loss": 0.1968, "step": 15479 }, { "epoch": 4.1192123469930815, "grad_norm": 0.29137566685676575, "learning_rate": 1.5535043952727466e-08, "loss": 0.177, "step": 15480 }, { "epoch": 4.119478445981906, "grad_norm": 0.2861465811729431, "learning_rate": 1.5525912151435794e-08, "loss": 0.1763, "step": 15481 }, { "epoch": 4.119744544970729, "grad_norm": 0.32280415296554565, "learning_rate": 1.551678280896548e-08, "loss": 0.2026, "step": 15482 }, { "epoch": 4.120010643959553, "grad_norm": 0.3022271990776062, "learning_rate": 1.550765592558225e-08, "loss": 0.1909, "step": 15483 }, { "epoch": 4.120276742948377, "grad_norm": 0.32300156354904175, "learning_rate": 1.549853150155178e-08, "loss": 0.206, "step": 15484 }, { "epoch": 4.1205428419372, "grad_norm": 0.3117654025554657, "learning_rate": 1.5489409537139652e-08, "loss": 0.2071, "step": 15485 }, { "epoch": 4.1208089409260245, "grad_norm": 0.29703977704048157, "learning_rate": 1.5480290032611387e-08, "loss": 0.1876, "step": 15486 }, { "epoch": 4.121075039914849, "grad_norm": 0.32057324051856995, "learning_rate": 1.5471172988232438e-08, "loss": 0.2047, "step": 15487 }, { "epoch": 4.121341138903672, "grad_norm": 0.31506046652793884, "learning_rate": 1.546205840426815e-08, "loss": 0.2146, "step": 15488 }, { "epoch": 4.121607237892496, "grad_norm": 0.3041343092918396, "learning_rate": 1.5452946280983846e-08, "loss": 0.1877, "step": 15489 }, { "epoch": 4.12187333688132, "grad_norm": 0.3061790466308594, "learning_rate": 1.5443836618644767e-08, "loss": 0.1885, "step": 15490 }, { "epoch": 4.122139435870143, "grad_norm": 0.30207088589668274, "learning_rate": 1.5434729417516044e-08, "loss": 0.1958, "step": 15491 }, { "epoch": 4.122405534858967, "grad_norm": 0.27981850504875183, "learning_rate": 1.542562467786277e-08, "loss": 0.1762, "step": 15492 }, { "epoch": 4.122671633847792, "grad_norm": 0.2870655357837677, "learning_rate": 1.541652239994996e-08, "loss": 0.1871, "step": 15493 }, { "epoch": 4.122937732836615, "grad_norm": 0.28311267495155334, "learning_rate": 1.5407422584042574e-08, "loss": 0.1813, "step": 15494 }, { "epoch": 4.123203831825439, "grad_norm": 0.28291329741477966, "learning_rate": 1.5398325230405463e-08, "loss": 0.1829, "step": 15495 }, { "epoch": 4.123469930814263, "grad_norm": 0.3044435977935791, "learning_rate": 1.538923033930345e-08, "loss": 0.1978, "step": 15496 }, { "epoch": 4.123736029803087, "grad_norm": 0.3058732748031616, "learning_rate": 1.5380137911001246e-08, "loss": 0.1983, "step": 15497 }, { "epoch": 4.12400212879191, "grad_norm": 0.289408802986145, "learning_rate": 1.537104794576354e-08, "loss": 0.1862, "step": 15498 }, { "epoch": 4.1242682277807345, "grad_norm": 0.2843458950519562, "learning_rate": 1.536196044385486e-08, "loss": 0.1706, "step": 15499 }, { "epoch": 4.124534326769559, "grad_norm": 0.2962064743041992, "learning_rate": 1.5352875405539766e-08, "loss": 0.1853, "step": 15500 }, { "epoch": 4.124800425758382, "grad_norm": 0.34524932503700256, "learning_rate": 1.534379283108267e-08, "loss": 0.1902, "step": 15501 }, { "epoch": 4.125066524747206, "grad_norm": 0.32632026076316833, "learning_rate": 1.5334712720747967e-08, "loss": 0.1904, "step": 15502 }, { "epoch": 4.12533262373603, "grad_norm": 0.29594799876213074, "learning_rate": 1.5325635074799935e-08, "loss": 0.1747, "step": 15503 }, { "epoch": 4.125598722724853, "grad_norm": 0.3021031618118286, "learning_rate": 1.531655989350282e-08, "loss": 0.2013, "step": 15504 }, { "epoch": 4.1258648217136775, "grad_norm": 0.3087970018386841, "learning_rate": 1.530748717712077e-08, "loss": 0.1794, "step": 15505 }, { "epoch": 4.126130920702502, "grad_norm": 0.310367614030838, "learning_rate": 1.5298416925917866e-08, "loss": 0.2, "step": 15506 }, { "epoch": 4.126397019691325, "grad_norm": 0.30802106857299805, "learning_rate": 1.5289349140158136e-08, "loss": 0.1932, "step": 15507 }, { "epoch": 4.126663118680149, "grad_norm": 0.2851751148700714, "learning_rate": 1.5280283820105488e-08, "loss": 0.1833, "step": 15508 }, { "epoch": 4.126929217668973, "grad_norm": 0.29062309861183167, "learning_rate": 1.52712209660238e-08, "loss": 0.1826, "step": 15509 }, { "epoch": 4.127195316657796, "grad_norm": 0.3272567689418793, "learning_rate": 1.5262160578176897e-08, "loss": 0.1963, "step": 15510 }, { "epoch": 4.1274614156466205, "grad_norm": 0.29281094670295715, "learning_rate": 1.5253102656828463e-08, "loss": 0.1782, "step": 15511 }, { "epoch": 4.127727514635445, "grad_norm": 0.3175693154335022, "learning_rate": 1.524404720224216e-08, "loss": 0.1878, "step": 15512 }, { "epoch": 4.127993613624268, "grad_norm": 0.2925410270690918, "learning_rate": 1.5234994214681574e-08, "loss": 0.1819, "step": 15513 }, { "epoch": 4.128259712613092, "grad_norm": 0.2756064236164093, "learning_rate": 1.5225943694410225e-08, "loss": 0.1814, "step": 15514 }, { "epoch": 4.128525811601916, "grad_norm": 0.30819568037986755, "learning_rate": 1.521689564169154e-08, "loss": 0.1918, "step": 15515 }, { "epoch": 4.128791910590739, "grad_norm": 0.302456796169281, "learning_rate": 1.5207850056788905e-08, "loss": 0.1921, "step": 15516 }, { "epoch": 4.1290580095795635, "grad_norm": 0.2937290072441101, "learning_rate": 1.5198806939965568e-08, "loss": 0.1777, "step": 15517 }, { "epoch": 4.129324108568388, "grad_norm": 0.3138205409049988, "learning_rate": 1.5189766291484783e-08, "loss": 0.1974, "step": 15518 }, { "epoch": 4.129590207557211, "grad_norm": 0.30355262756347656, "learning_rate": 1.5180728111609698e-08, "loss": 0.1993, "step": 15519 }, { "epoch": 4.129856306546035, "grad_norm": 0.3040081262588501, "learning_rate": 1.5171692400603385e-08, "loss": 0.1898, "step": 15520 }, { "epoch": 4.130122405534859, "grad_norm": 0.29579249024391174, "learning_rate": 1.516265915872884e-08, "loss": 0.2018, "step": 15521 }, { "epoch": 4.130388504523683, "grad_norm": 0.3148496747016907, "learning_rate": 1.5153628386249018e-08, "loss": 0.1936, "step": 15522 }, { "epoch": 4.1306546035125065, "grad_norm": 0.3054586350917816, "learning_rate": 1.5144600083426772e-08, "loss": 0.1938, "step": 15523 }, { "epoch": 4.130920702501331, "grad_norm": 0.37595802545547485, "learning_rate": 1.5135574250524897e-08, "loss": 0.202, "step": 15524 }, { "epoch": 4.131186801490155, "grad_norm": 0.29665929079055786, "learning_rate": 1.512655088780612e-08, "loss": 0.1785, "step": 15525 }, { "epoch": 4.131452900478978, "grad_norm": 0.322192907333374, "learning_rate": 1.5117529995533062e-08, "loss": 0.2161, "step": 15526 }, { "epoch": 4.131718999467802, "grad_norm": 0.29921478033065796, "learning_rate": 1.5108511573968308e-08, "loss": 0.1856, "step": 15527 }, { "epoch": 4.131985098456626, "grad_norm": 0.29664167761802673, "learning_rate": 1.5099495623374358e-08, "loss": 0.1925, "step": 15528 }, { "epoch": 4.132251197445449, "grad_norm": 0.3121269643306732, "learning_rate": 1.509048214401365e-08, "loss": 0.1792, "step": 15529 }, { "epoch": 4.132517296434274, "grad_norm": 0.2889983057975769, "learning_rate": 1.508147113614857e-08, "loss": 0.1818, "step": 15530 }, { "epoch": 4.132783395423098, "grad_norm": 0.29159361124038696, "learning_rate": 1.5072462600041345e-08, "loss": 0.1938, "step": 15531 }, { "epoch": 4.133049494411921, "grad_norm": 0.2974458634853363, "learning_rate": 1.5063456535954235e-08, "loss": 0.2064, "step": 15532 }, { "epoch": 4.133315593400745, "grad_norm": 0.28895103931427, "learning_rate": 1.5054452944149353e-08, "loss": 0.1776, "step": 15533 }, { "epoch": 4.133581692389569, "grad_norm": 0.289870023727417, "learning_rate": 1.504545182488882e-08, "loss": 0.1901, "step": 15534 }, { "epoch": 4.133847791378392, "grad_norm": 0.3070356249809265, "learning_rate": 1.5036453178434573e-08, "loss": 0.1917, "step": 15535 }, { "epoch": 4.1341138903672165, "grad_norm": 0.29711177945137024, "learning_rate": 1.502745700504857e-08, "loss": 0.1927, "step": 15536 }, { "epoch": 4.134379989356041, "grad_norm": 0.3122800290584564, "learning_rate": 1.501846330499268e-08, "loss": 0.1943, "step": 15537 }, { "epoch": 4.134646088344864, "grad_norm": 0.3012572228908539, "learning_rate": 1.500947207852865e-08, "loss": 0.1825, "step": 15538 }, { "epoch": 4.134912187333688, "grad_norm": 0.2662403881549835, "learning_rate": 1.5000483325918224e-08, "loss": 0.1809, "step": 15539 }, { "epoch": 4.135178286322512, "grad_norm": 0.2848401665687561, "learning_rate": 1.4991497047423017e-08, "loss": 0.1884, "step": 15540 }, { "epoch": 4.135444385311335, "grad_norm": 0.271447092294693, "learning_rate": 1.498251324330462e-08, "loss": 0.1774, "step": 15541 }, { "epoch": 4.1357104843001595, "grad_norm": 0.29773983359336853, "learning_rate": 1.497353191382452e-08, "loss": 0.1908, "step": 15542 }, { "epoch": 4.135976583288984, "grad_norm": 0.28152188658714294, "learning_rate": 1.4964553059244156e-08, "loss": 0.185, "step": 15543 }, { "epoch": 4.136242682277807, "grad_norm": 0.3027522563934326, "learning_rate": 1.4955576679824834e-08, "loss": 0.1984, "step": 15544 }, { "epoch": 4.136508781266631, "grad_norm": 0.29239803552627563, "learning_rate": 1.4946602775827854e-08, "loss": 0.1904, "step": 15545 }, { "epoch": 4.136774880255455, "grad_norm": 0.2873234152793884, "learning_rate": 1.4937631347514433e-08, "loss": 0.1726, "step": 15546 }, { "epoch": 4.137040979244279, "grad_norm": 0.3088943064212799, "learning_rate": 1.492866239514571e-08, "loss": 0.1971, "step": 15547 }, { "epoch": 4.1373070782331025, "grad_norm": 0.3274195194244385, "learning_rate": 1.4919695918982734e-08, "loss": 0.2044, "step": 15548 }, { "epoch": 4.137573177221927, "grad_norm": 0.32574620842933655, "learning_rate": 1.4910731919286513e-08, "loss": 0.2067, "step": 15549 }, { "epoch": 4.137839276210751, "grad_norm": 0.3320377469062805, "learning_rate": 1.4901770396317947e-08, "loss": 0.2253, "step": 15550 }, { "epoch": 4.138105375199574, "grad_norm": 0.2894071638584137, "learning_rate": 1.4892811350337875e-08, "loss": 0.1849, "step": 15551 }, { "epoch": 4.138371474188398, "grad_norm": 0.28374773263931274, "learning_rate": 1.4883854781607119e-08, "loss": 0.1827, "step": 15552 }, { "epoch": 4.138637573177222, "grad_norm": 0.33454808592796326, "learning_rate": 1.4874900690386327e-08, "loss": 0.2065, "step": 15553 }, { "epoch": 4.1389036721660455, "grad_norm": 0.3126623034477234, "learning_rate": 1.4865949076936146e-08, "loss": 0.1882, "step": 15554 }, { "epoch": 4.13916977115487, "grad_norm": 0.27355697751045227, "learning_rate": 1.4856999941517135e-08, "loss": 0.1695, "step": 15555 }, { "epoch": 4.139435870143694, "grad_norm": 0.3066370189189911, "learning_rate": 1.4848053284389784e-08, "loss": 0.1839, "step": 15556 }, { "epoch": 4.139701969132517, "grad_norm": 0.3646029829978943, "learning_rate": 1.4839109105814519e-08, "loss": 0.1877, "step": 15557 }, { "epoch": 4.139968068121341, "grad_norm": 0.3079218566417694, "learning_rate": 1.4830167406051663e-08, "loss": 0.2031, "step": 15558 }, { "epoch": 4.140234167110165, "grad_norm": 0.28625962138175964, "learning_rate": 1.482122818536149e-08, "loss": 0.1922, "step": 15559 }, { "epoch": 4.1405002660989885, "grad_norm": 0.2916910648345947, "learning_rate": 1.48122914440042e-08, "loss": 0.188, "step": 15560 }, { "epoch": 4.140766365087813, "grad_norm": 0.3201514482498169, "learning_rate": 1.480335718223994e-08, "loss": 0.1973, "step": 15561 }, { "epoch": 4.141032464076637, "grad_norm": 0.29305270314216614, "learning_rate": 1.4794425400328725e-08, "loss": 0.2002, "step": 15562 }, { "epoch": 4.14129856306546, "grad_norm": 0.33238938450813293, "learning_rate": 1.4785496098530547e-08, "loss": 0.1843, "step": 15563 }, { "epoch": 4.141564662054284, "grad_norm": 0.3067389726638794, "learning_rate": 1.4776569277105322e-08, "loss": 0.1778, "step": 15564 }, { "epoch": 4.141830761043108, "grad_norm": 0.29947009682655334, "learning_rate": 1.4767644936312896e-08, "loss": 0.1856, "step": 15565 }, { "epoch": 4.1420968600319314, "grad_norm": 0.27224498987197876, "learning_rate": 1.4758723076413004e-08, "loss": 0.177, "step": 15566 }, { "epoch": 4.142362959020756, "grad_norm": 0.2990736663341522, "learning_rate": 1.4749803697665364e-08, "loss": 0.2032, "step": 15567 }, { "epoch": 4.14262905800958, "grad_norm": 0.2991224229335785, "learning_rate": 1.4740886800329588e-08, "loss": 0.1908, "step": 15568 }, { "epoch": 4.142895156998404, "grad_norm": 0.29259100556373596, "learning_rate": 1.4731972384665248e-08, "loss": 0.1907, "step": 15569 }, { "epoch": 4.143161255987227, "grad_norm": 0.29428401589393616, "learning_rate": 1.4723060450931778e-08, "loss": 0.1848, "step": 15570 }, { "epoch": 4.143427354976051, "grad_norm": 0.2992461025714874, "learning_rate": 1.4714150999388596e-08, "loss": 0.1797, "step": 15571 }, { "epoch": 4.143693453964875, "grad_norm": 0.2952878475189209, "learning_rate": 1.470524403029505e-08, "loss": 0.1863, "step": 15572 }, { "epoch": 4.1439595529536986, "grad_norm": 0.28517016768455505, "learning_rate": 1.4696339543910374e-08, "loss": 0.1867, "step": 15573 }, { "epoch": 4.144225651942523, "grad_norm": 0.2887687385082245, "learning_rate": 1.4687437540493763e-08, "loss": 0.1809, "step": 15574 }, { "epoch": 4.144491750931347, "grad_norm": 0.27210766077041626, "learning_rate": 1.4678538020304321e-08, "loss": 0.1582, "step": 15575 }, { "epoch": 4.14475784992017, "grad_norm": 0.27690359950065613, "learning_rate": 1.4669640983601118e-08, "loss": 0.1738, "step": 15576 }, { "epoch": 4.145023948908994, "grad_norm": 0.2887670695781708, "learning_rate": 1.4660746430643101e-08, "loss": 0.1797, "step": 15577 }, { "epoch": 4.145290047897818, "grad_norm": 0.31142619252204895, "learning_rate": 1.4651854361689176e-08, "loss": 0.2134, "step": 15578 }, { "epoch": 4.1455561468866415, "grad_norm": 0.3444034457206726, "learning_rate": 1.464296477699819e-08, "loss": 0.1939, "step": 15579 }, { "epoch": 4.145822245875466, "grad_norm": 0.3141412138938904, "learning_rate": 1.4634077676828837e-08, "loss": 0.1961, "step": 15580 }, { "epoch": 4.14608834486429, "grad_norm": 0.30561086535453796, "learning_rate": 1.4625193061439845e-08, "loss": 0.1746, "step": 15581 }, { "epoch": 4.146354443853113, "grad_norm": 0.27624455094337463, "learning_rate": 1.4616310931089803e-08, "loss": 0.1755, "step": 15582 }, { "epoch": 4.146620542841937, "grad_norm": 0.31278282403945923, "learning_rate": 1.4607431286037264e-08, "loss": 0.2021, "step": 15583 }, { "epoch": 4.146886641830761, "grad_norm": 0.2805924415588379, "learning_rate": 1.4598554126540675e-08, "loss": 0.1716, "step": 15584 }, { "epoch": 4.1471527408195845, "grad_norm": 0.29548683762550354, "learning_rate": 1.4589679452858428e-08, "loss": 0.1889, "step": 15585 }, { "epoch": 4.147418839808409, "grad_norm": 0.2908594012260437, "learning_rate": 1.4580807265248852e-08, "loss": 0.1838, "step": 15586 }, { "epoch": 4.147684938797233, "grad_norm": 0.28353989124298096, "learning_rate": 1.4571937563970183e-08, "loss": 0.1803, "step": 15587 }, { "epoch": 4.147951037786056, "grad_norm": 0.31564003229141235, "learning_rate": 1.4563070349280625e-08, "loss": 0.2037, "step": 15588 }, { "epoch": 4.14821713677488, "grad_norm": 0.3041022717952728, "learning_rate": 1.4554205621438232e-08, "loss": 0.1945, "step": 15589 }, { "epoch": 4.148483235763704, "grad_norm": 0.3055446743965149, "learning_rate": 1.4545343380701058e-08, "loss": 0.1965, "step": 15590 }, { "epoch": 4.148749334752528, "grad_norm": 0.30170756578445435, "learning_rate": 1.4536483627327078e-08, "loss": 0.1934, "step": 15591 }, { "epoch": 4.149015433741352, "grad_norm": 0.3602495491504669, "learning_rate": 1.452762636157413e-08, "loss": 0.2022, "step": 15592 }, { "epoch": 4.149281532730176, "grad_norm": 0.2844228446483612, "learning_rate": 1.4518771583700063e-08, "loss": 0.1834, "step": 15593 }, { "epoch": 4.149547631719, "grad_norm": 0.30863943696022034, "learning_rate": 1.4509919293962592e-08, "loss": 0.1896, "step": 15594 }, { "epoch": 4.149813730707823, "grad_norm": 0.32135677337646484, "learning_rate": 1.45010694926194e-08, "loss": 0.21, "step": 15595 }, { "epoch": 4.150079829696647, "grad_norm": 0.31043973565101624, "learning_rate": 1.4492222179928082e-08, "loss": 0.195, "step": 15596 }, { "epoch": 4.150345928685471, "grad_norm": 0.2958332896232605, "learning_rate": 1.4483377356146165e-08, "loss": 0.1752, "step": 15597 }, { "epoch": 4.150612027674295, "grad_norm": 0.2775978744029999, "learning_rate": 1.4474535021531097e-08, "loss": 0.1654, "step": 15598 }, { "epoch": 4.150878126663119, "grad_norm": 0.2870965600013733, "learning_rate": 1.4465695176340243e-08, "loss": 0.1851, "step": 15599 }, { "epoch": 4.151144225651943, "grad_norm": 0.3373699486255646, "learning_rate": 1.4456857820830904e-08, "loss": 0.188, "step": 15600 }, { "epoch": 4.151410324640766, "grad_norm": 0.29432934522628784, "learning_rate": 1.444802295526033e-08, "loss": 0.1921, "step": 15601 }, { "epoch": 4.15167642362959, "grad_norm": 0.2688073515892029, "learning_rate": 1.4439190579885675e-08, "loss": 0.1582, "step": 15602 }, { "epoch": 4.151942522618414, "grad_norm": 0.27987757325172424, "learning_rate": 1.4430360694964017e-08, "loss": 0.1998, "step": 15603 }, { "epoch": 4.152208621607238, "grad_norm": 0.2829645574092865, "learning_rate": 1.4421533300752387e-08, "loss": 0.1799, "step": 15604 }, { "epoch": 4.152474720596062, "grad_norm": 0.2903292179107666, "learning_rate": 1.4412708397507722e-08, "loss": 0.1907, "step": 15605 }, { "epoch": 4.152740819584886, "grad_norm": 0.2989315390586853, "learning_rate": 1.4403885985486886e-08, "loss": 0.1898, "step": 15606 }, { "epoch": 4.153006918573709, "grad_norm": 0.3009895086288452, "learning_rate": 1.4395066064946703e-08, "loss": 0.1889, "step": 15607 }, { "epoch": 4.153273017562533, "grad_norm": 0.421624094247818, "learning_rate": 1.438624863614386e-08, "loss": 0.1823, "step": 15608 }, { "epoch": 4.153539116551357, "grad_norm": 0.3237406313419342, "learning_rate": 1.4377433699335018e-08, "loss": 0.2058, "step": 15609 }, { "epoch": 4.153805215540181, "grad_norm": 0.29629236459732056, "learning_rate": 1.436862125477677e-08, "loss": 0.1721, "step": 15610 }, { "epoch": 4.154071314529005, "grad_norm": 0.3194151520729065, "learning_rate": 1.4359811302725645e-08, "loss": 0.1964, "step": 15611 }, { "epoch": 4.154337413517829, "grad_norm": 0.29110991954803467, "learning_rate": 1.4351003843438014e-08, "loss": 0.1981, "step": 15612 }, { "epoch": 4.154603512506652, "grad_norm": 0.30219918489456177, "learning_rate": 1.4342198877170286e-08, "loss": 0.188, "step": 15613 }, { "epoch": 4.154869611495476, "grad_norm": 0.33454278111457825, "learning_rate": 1.4333396404178732e-08, "loss": 0.1973, "step": 15614 }, { "epoch": 4.1551357104843, "grad_norm": 0.28837066888809204, "learning_rate": 1.4324596424719581e-08, "loss": 0.2046, "step": 15615 }, { "epoch": 4.155401809473124, "grad_norm": 0.31831690669059753, "learning_rate": 1.4315798939048996e-08, "loss": 0.2064, "step": 15616 }, { "epoch": 4.155667908461948, "grad_norm": 0.29303017258644104, "learning_rate": 1.4307003947423014e-08, "loss": 0.1874, "step": 15617 }, { "epoch": 4.155934007450772, "grad_norm": 0.3053959310054779, "learning_rate": 1.4298211450097631e-08, "loss": 0.1756, "step": 15618 }, { "epoch": 4.156200106439596, "grad_norm": 0.29356783628463745, "learning_rate": 1.4289421447328808e-08, "loss": 0.1957, "step": 15619 }, { "epoch": 4.156466205428419, "grad_norm": 0.31279483437538147, "learning_rate": 1.4280633939372366e-08, "loss": 0.2, "step": 15620 }, { "epoch": 4.156732304417243, "grad_norm": 0.31855306029319763, "learning_rate": 1.427184892648411e-08, "loss": 0.1926, "step": 15621 }, { "epoch": 4.156998403406067, "grad_norm": 0.28912845253944397, "learning_rate": 1.4263066408919745e-08, "loss": 0.1777, "step": 15622 }, { "epoch": 4.157264502394891, "grad_norm": 0.31086987257003784, "learning_rate": 1.4254286386934889e-08, "loss": 0.1762, "step": 15623 }, { "epoch": 4.157530601383715, "grad_norm": 0.31329894065856934, "learning_rate": 1.4245508860785127e-08, "loss": 0.1911, "step": 15624 }, { "epoch": 4.157796700372539, "grad_norm": 0.2728824317455292, "learning_rate": 1.4236733830725966e-08, "loss": 0.1605, "step": 15625 }, { "epoch": 4.158062799361362, "grad_norm": 0.29327282309532166, "learning_rate": 1.422796129701278e-08, "loss": 0.1893, "step": 15626 }, { "epoch": 4.158328898350186, "grad_norm": 0.2899979054927826, "learning_rate": 1.4219191259900942e-08, "loss": 0.1978, "step": 15627 }, { "epoch": 4.15859499733901, "grad_norm": 0.3297022581100464, "learning_rate": 1.4210423719645704e-08, "loss": 0.2012, "step": 15628 }, { "epoch": 4.158861096327834, "grad_norm": 0.30766311287879944, "learning_rate": 1.4201658676502293e-08, "loss": 0.1978, "step": 15629 }, { "epoch": 4.159127195316658, "grad_norm": 0.29742661118507385, "learning_rate": 1.419289613072584e-08, "loss": 0.1869, "step": 15630 }, { "epoch": 4.159393294305482, "grad_norm": 0.3173011839389801, "learning_rate": 1.4184136082571363e-08, "loss": 0.2067, "step": 15631 }, { "epoch": 4.159659393294305, "grad_norm": 0.34957069158554077, "learning_rate": 1.4175378532293868e-08, "loss": 0.2029, "step": 15632 }, { "epoch": 4.159925492283129, "grad_norm": 0.31655237078666687, "learning_rate": 1.4166623480148265e-08, "loss": 0.1977, "step": 15633 }, { "epoch": 4.160191591271953, "grad_norm": 0.3079972267150879, "learning_rate": 1.4157870926389414e-08, "loss": 0.189, "step": 15634 }, { "epoch": 4.160457690260777, "grad_norm": 0.26846563816070557, "learning_rate": 1.4149120871272024e-08, "loss": 0.171, "step": 15635 }, { "epoch": 4.160723789249601, "grad_norm": 0.309559166431427, "learning_rate": 1.4140373315050814e-08, "loss": 0.2094, "step": 15636 }, { "epoch": 4.160989888238425, "grad_norm": 0.29670006036758423, "learning_rate": 1.413162825798041e-08, "loss": 0.1806, "step": 15637 }, { "epoch": 4.161255987227248, "grad_norm": 0.30639490485191345, "learning_rate": 1.4122885700315346e-08, "loss": 0.1914, "step": 15638 }, { "epoch": 4.161522086216072, "grad_norm": 0.31063178181648254, "learning_rate": 1.4114145642310105e-08, "loss": 0.1925, "step": 15639 }, { "epoch": 4.161788185204896, "grad_norm": 0.28755736351013184, "learning_rate": 1.4105408084219084e-08, "loss": 0.177, "step": 15640 }, { "epoch": 4.1620542841937205, "grad_norm": 0.2888134717941284, "learning_rate": 1.4096673026296613e-08, "loss": 0.1804, "step": 15641 }, { "epoch": 4.162320383182544, "grad_norm": 0.31808793544769287, "learning_rate": 1.4087940468796943e-08, "loss": 0.2115, "step": 15642 }, { "epoch": 4.162586482171368, "grad_norm": 0.3008369505405426, "learning_rate": 1.4079210411974286e-08, "loss": 0.1854, "step": 15643 }, { "epoch": 4.162852581160192, "grad_norm": 0.3124009072780609, "learning_rate": 1.4070482856082688e-08, "loss": 0.1877, "step": 15644 }, { "epoch": 4.163118680149015, "grad_norm": 0.27889055013656616, "learning_rate": 1.4061757801376228e-08, "loss": 0.1569, "step": 15645 }, { "epoch": 4.163384779137839, "grad_norm": 0.31994134187698364, "learning_rate": 1.4053035248108857e-08, "loss": 0.1883, "step": 15646 }, { "epoch": 4.1636508781266635, "grad_norm": 0.31077897548675537, "learning_rate": 1.4044315196534485e-08, "loss": 0.1947, "step": 15647 }, { "epoch": 4.163916977115487, "grad_norm": 0.3247716724872589, "learning_rate": 1.4035597646906904e-08, "loss": 0.1901, "step": 15648 }, { "epoch": 4.164183076104311, "grad_norm": 0.3118721842765808, "learning_rate": 1.4026882599479872e-08, "loss": 0.1955, "step": 15649 }, { "epoch": 4.164449175093135, "grad_norm": 0.30173274874687195, "learning_rate": 1.4018170054507083e-08, "loss": 0.1886, "step": 15650 }, { "epoch": 4.164715274081958, "grad_norm": 0.2562245726585388, "learning_rate": 1.4009460012242092e-08, "loss": 0.175, "step": 15651 }, { "epoch": 4.164981373070782, "grad_norm": 0.3213803172111511, "learning_rate": 1.4000752472938448e-08, "loss": 0.2007, "step": 15652 }, { "epoch": 4.165247472059606, "grad_norm": 0.29595354199409485, "learning_rate": 1.399204743684963e-08, "loss": 0.1899, "step": 15653 }, { "epoch": 4.16551357104843, "grad_norm": 0.3184867203235626, "learning_rate": 1.3983344904228967e-08, "loss": 0.21, "step": 15654 }, { "epoch": 4.165779670037254, "grad_norm": 0.31077325344085693, "learning_rate": 1.3974644875329789e-08, "loss": 0.1912, "step": 15655 }, { "epoch": 4.166045769026078, "grad_norm": 0.30357468128204346, "learning_rate": 1.396594735040535e-08, "loss": 0.1735, "step": 15656 }, { "epoch": 4.166311868014901, "grad_norm": 0.28217849135398865, "learning_rate": 1.395725232970879e-08, "loss": 0.1656, "step": 15657 }, { "epoch": 4.166577967003725, "grad_norm": 0.30041518807411194, "learning_rate": 1.3948559813493198e-08, "loss": 0.193, "step": 15658 }, { "epoch": 4.166844065992549, "grad_norm": 0.30205339193344116, "learning_rate": 1.3939869802011617e-08, "loss": 0.1719, "step": 15659 }, { "epoch": 4.167110164981373, "grad_norm": 0.26217907667160034, "learning_rate": 1.3931182295516963e-08, "loss": 0.1689, "step": 15660 }, { "epoch": 4.167376263970197, "grad_norm": 0.3584619462490082, "learning_rate": 1.3922497294262148e-08, "loss": 0.209, "step": 15661 }, { "epoch": 4.167642362959021, "grad_norm": 0.3001333773136139, "learning_rate": 1.3913814798499902e-08, "loss": 0.1726, "step": 15662 }, { "epoch": 4.167908461947844, "grad_norm": 0.3288152515888214, "learning_rate": 1.3905134808483e-08, "loss": 0.198, "step": 15663 }, { "epoch": 4.168174560936668, "grad_norm": 0.3137367069721222, "learning_rate": 1.3896457324464073e-08, "loss": 0.1701, "step": 15664 }, { "epoch": 4.168440659925492, "grad_norm": 0.30325764417648315, "learning_rate": 1.3887782346695709e-08, "loss": 0.1743, "step": 15665 }, { "epoch": 4.1687067589143165, "grad_norm": 0.2671993672847748, "learning_rate": 1.3879109875430416e-08, "loss": 0.1676, "step": 15666 }, { "epoch": 4.16897285790314, "grad_norm": 0.3015911281108856, "learning_rate": 1.3870439910920618e-08, "loss": 0.1892, "step": 15667 }, { "epoch": 4.169238956891964, "grad_norm": 0.27704736590385437, "learning_rate": 1.3861772453418686e-08, "loss": 0.1846, "step": 15668 }, { "epoch": 4.169505055880788, "grad_norm": 0.30520859360694885, "learning_rate": 1.3853107503176909e-08, "loss": 0.1969, "step": 15669 }, { "epoch": 4.169771154869611, "grad_norm": 0.2908307909965515, "learning_rate": 1.3844445060447506e-08, "loss": 0.1948, "step": 15670 }, { "epoch": 4.170037253858435, "grad_norm": 0.27985286712646484, "learning_rate": 1.383578512548259e-08, "loss": 0.1674, "step": 15671 }, { "epoch": 4.1703033528472595, "grad_norm": 0.2975655496120453, "learning_rate": 1.3827127698534236e-08, "loss": 0.1842, "step": 15672 }, { "epoch": 4.170569451836083, "grad_norm": 0.38542360067367554, "learning_rate": 1.3818472779854485e-08, "loss": 0.1988, "step": 15673 }, { "epoch": 4.170835550824907, "grad_norm": 0.29786449670791626, "learning_rate": 1.3809820369695191e-08, "loss": 0.1869, "step": 15674 }, { "epoch": 4.171101649813731, "grad_norm": 0.33675843477249146, "learning_rate": 1.3801170468308243e-08, "loss": 0.2077, "step": 15675 }, { "epoch": 4.171367748802554, "grad_norm": 0.3230394423007965, "learning_rate": 1.3792523075945406e-08, "loss": 0.1932, "step": 15676 }, { "epoch": 4.171633847791378, "grad_norm": 0.2911384701728821, "learning_rate": 1.3783878192858378e-08, "loss": 0.1736, "step": 15677 }, { "epoch": 4.1718999467802025, "grad_norm": 0.3423920273780823, "learning_rate": 1.3775235819298814e-08, "loss": 0.2037, "step": 15678 }, { "epoch": 4.172166045769026, "grad_norm": 0.33414533734321594, "learning_rate": 1.3766595955518267e-08, "loss": 0.2106, "step": 15679 }, { "epoch": 4.17243214475785, "grad_norm": 0.31696611642837524, "learning_rate": 1.3757958601768194e-08, "loss": 0.1857, "step": 15680 }, { "epoch": 4.172698243746674, "grad_norm": 0.303351491689682, "learning_rate": 1.3749323758300014e-08, "loss": 0.1916, "step": 15681 }, { "epoch": 4.172964342735497, "grad_norm": 0.29906198382377625, "learning_rate": 1.374069142536507e-08, "loss": 0.1803, "step": 15682 }, { "epoch": 4.173230441724321, "grad_norm": 0.30143535137176514, "learning_rate": 1.3732061603214629e-08, "loss": 0.1898, "step": 15683 }, { "epoch": 4.1734965407131455, "grad_norm": 0.29056447744369507, "learning_rate": 1.3723434292099888e-08, "loss": 0.1935, "step": 15684 }, { "epoch": 4.17376263970197, "grad_norm": 0.2985278367996216, "learning_rate": 1.3714809492271962e-08, "loss": 0.1844, "step": 15685 }, { "epoch": 4.174028738690793, "grad_norm": 0.29415392875671387, "learning_rate": 1.370618720398189e-08, "loss": 0.1857, "step": 15686 }, { "epoch": 4.174294837679617, "grad_norm": 0.31849074363708496, "learning_rate": 1.369756742748065e-08, "loss": 0.1983, "step": 15687 }, { "epoch": 4.174560936668441, "grad_norm": 0.33844509720802307, "learning_rate": 1.3688950163019164e-08, "loss": 0.2038, "step": 15688 }, { "epoch": 4.174827035657264, "grad_norm": 0.3240344822406769, "learning_rate": 1.3680335410848209e-08, "loss": 0.2008, "step": 15689 }, { "epoch": 4.1750931346460884, "grad_norm": 0.32340559363365173, "learning_rate": 1.3671723171218563e-08, "loss": 0.1923, "step": 15690 }, { "epoch": 4.175359233634913, "grad_norm": 0.3052440285682678, "learning_rate": 1.3663113444380903e-08, "loss": 0.1801, "step": 15691 }, { "epoch": 4.175625332623736, "grad_norm": 0.282501220703125, "learning_rate": 1.3654506230585872e-08, "loss": 0.1673, "step": 15692 }, { "epoch": 4.17589143161256, "grad_norm": 0.31665486097335815, "learning_rate": 1.3645901530083937e-08, "loss": 0.1993, "step": 15693 }, { "epoch": 4.176157530601384, "grad_norm": 0.322297602891922, "learning_rate": 1.3637299343125597e-08, "loss": 0.2102, "step": 15694 }, { "epoch": 4.176423629590207, "grad_norm": 0.27766090631484985, "learning_rate": 1.3628699669961241e-08, "loss": 0.1704, "step": 15695 }, { "epoch": 4.176689728579031, "grad_norm": 0.340549498796463, "learning_rate": 1.362010251084117e-08, "loss": 0.2101, "step": 15696 }, { "epoch": 4.1769558275678556, "grad_norm": 0.30673640966415405, "learning_rate": 1.3611507866015649e-08, "loss": 0.1888, "step": 15697 }, { "epoch": 4.177221926556679, "grad_norm": 0.2960909605026245, "learning_rate": 1.360291573573481e-08, "loss": 0.1873, "step": 15698 }, { "epoch": 4.177488025545503, "grad_norm": 0.5204905867576599, "learning_rate": 1.3594326120248778e-08, "loss": 0.1918, "step": 15699 }, { "epoch": 4.177754124534327, "grad_norm": 0.26985952258110046, "learning_rate": 1.3585739019807552e-08, "loss": 0.1823, "step": 15700 }, { "epoch": 4.17802022352315, "grad_norm": 0.2971092760562897, "learning_rate": 1.3577154434661087e-08, "loss": 0.2101, "step": 15701 }, { "epoch": 4.178286322511974, "grad_norm": 0.29907986521720886, "learning_rate": 1.3568572365059272e-08, "loss": 0.1989, "step": 15702 }, { "epoch": 4.1785524215007985, "grad_norm": 0.28270187973976135, "learning_rate": 1.3559992811251886e-08, "loss": 0.1823, "step": 15703 }, { "epoch": 4.178818520489622, "grad_norm": 0.29667332768440247, "learning_rate": 1.3551415773488683e-08, "loss": 0.1831, "step": 15704 }, { "epoch": 4.179084619478446, "grad_norm": 0.2708129286766052, "learning_rate": 1.35428412520193e-08, "loss": 0.1822, "step": 15705 }, { "epoch": 4.17935071846727, "grad_norm": 0.3091406226158142, "learning_rate": 1.3534269247093343e-08, "loss": 0.1992, "step": 15706 }, { "epoch": 4.179616817456093, "grad_norm": 0.2831333875656128, "learning_rate": 1.3525699758960284e-08, "loss": 0.1768, "step": 15707 }, { "epoch": 4.179882916444917, "grad_norm": 0.2987636923789978, "learning_rate": 1.3517132787869579e-08, "loss": 0.1887, "step": 15708 }, { "epoch": 4.1801490154337415, "grad_norm": 0.2973686158657074, "learning_rate": 1.350856833407058e-08, "loss": 0.1937, "step": 15709 }, { "epoch": 4.180415114422566, "grad_norm": 0.3225371837615967, "learning_rate": 1.3500006397812591e-08, "loss": 0.2023, "step": 15710 }, { "epoch": 4.180681213411389, "grad_norm": 0.3260001242160797, "learning_rate": 1.3491446979344822e-08, "loss": 0.2034, "step": 15711 }, { "epoch": 4.180947312400213, "grad_norm": 0.28932520747184753, "learning_rate": 1.348289007891643e-08, "loss": 0.2021, "step": 15712 }, { "epoch": 4.181213411389037, "grad_norm": 0.30907705426216125, "learning_rate": 1.347433569677645e-08, "loss": 0.2031, "step": 15713 }, { "epoch": 4.18147951037786, "grad_norm": 0.2940256893634796, "learning_rate": 1.3465783833173893e-08, "loss": 0.1862, "step": 15714 }, { "epoch": 4.1817456093666845, "grad_norm": 0.2639364004135132, "learning_rate": 1.3457234488357705e-08, "loss": 0.1808, "step": 15715 }, { "epoch": 4.182011708355509, "grad_norm": 0.28597548604011536, "learning_rate": 1.3448687662576708e-08, "loss": 0.1899, "step": 15716 }, { "epoch": 4.182277807344332, "grad_norm": 0.309940367937088, "learning_rate": 1.344014335607967e-08, "loss": 0.2163, "step": 15717 }, { "epoch": 4.182543906333156, "grad_norm": 0.3133711814880371, "learning_rate": 1.3431601569115314e-08, "loss": 0.2032, "step": 15718 }, { "epoch": 4.18281000532198, "grad_norm": 0.28798985481262207, "learning_rate": 1.3423062301932265e-08, "loss": 0.1866, "step": 15719 }, { "epoch": 4.183076104310803, "grad_norm": 0.30000340938568115, "learning_rate": 1.3414525554779066e-08, "loss": 0.1718, "step": 15720 }, { "epoch": 4.1833422032996275, "grad_norm": 0.2967289090156555, "learning_rate": 1.340599132790422e-08, "loss": 0.1846, "step": 15721 }, { "epoch": 4.183608302288452, "grad_norm": 0.3213893473148346, "learning_rate": 1.3397459621556128e-08, "loss": 0.1863, "step": 15722 }, { "epoch": 4.183874401277275, "grad_norm": 0.3251339793205261, "learning_rate": 1.3388930435983125e-08, "loss": 0.194, "step": 15723 }, { "epoch": 4.184140500266099, "grad_norm": 0.30249977111816406, "learning_rate": 1.3380403771433502e-08, "loss": 0.1875, "step": 15724 }, { "epoch": 4.184406599254923, "grad_norm": 0.2960187792778015, "learning_rate": 1.3371879628155392e-08, "loss": 0.1802, "step": 15725 }, { "epoch": 4.184672698243746, "grad_norm": 0.2891601324081421, "learning_rate": 1.336335800639693e-08, "loss": 0.1806, "step": 15726 }, { "epoch": 4.1849387972325705, "grad_norm": 0.283200204372406, "learning_rate": 1.3354838906406174e-08, "loss": 0.1785, "step": 15727 }, { "epoch": 4.185204896221395, "grad_norm": 0.2750890254974365, "learning_rate": 1.3346322328431092e-08, "loss": 0.1869, "step": 15728 }, { "epoch": 4.185470995210218, "grad_norm": 0.29145434498786926, "learning_rate": 1.3337808272719564e-08, "loss": 0.1836, "step": 15729 }, { "epoch": 4.185737094199042, "grad_norm": 0.3077506124973297, "learning_rate": 1.3329296739519424e-08, "loss": 0.185, "step": 15730 }, { "epoch": 4.186003193187866, "grad_norm": 0.3029008209705353, "learning_rate": 1.3320787729078442e-08, "loss": 0.2053, "step": 15731 }, { "epoch": 4.186269292176689, "grad_norm": 0.2825261056423187, "learning_rate": 1.3312281241644242e-08, "loss": 0.1758, "step": 15732 }, { "epoch": 4.186535391165513, "grad_norm": 0.2919633388519287, "learning_rate": 1.3303777277464457e-08, "loss": 0.1799, "step": 15733 }, { "epoch": 4.186801490154338, "grad_norm": 0.29528605937957764, "learning_rate": 1.3295275836786635e-08, "loss": 0.1795, "step": 15734 }, { "epoch": 4.187067589143162, "grad_norm": 0.2952996790409088, "learning_rate": 1.3286776919858178e-08, "loss": 0.1725, "step": 15735 }, { "epoch": 4.187333688131985, "grad_norm": 0.271380215883255, "learning_rate": 1.3278280526926499e-08, "loss": 0.1769, "step": 15736 }, { "epoch": 4.187599787120809, "grad_norm": 0.3249484598636627, "learning_rate": 1.3269786658238902e-08, "loss": 0.2047, "step": 15737 }, { "epoch": 4.187865886109633, "grad_norm": 0.28622934222221375, "learning_rate": 1.326129531404262e-08, "loss": 0.1846, "step": 15738 }, { "epoch": 4.188131985098456, "grad_norm": 0.2957800328731537, "learning_rate": 1.3252806494584823e-08, "loss": 0.1872, "step": 15739 }, { "epoch": 4.1883980840872805, "grad_norm": 0.31785354018211365, "learning_rate": 1.3244320200112591e-08, "loss": 0.2077, "step": 15740 }, { "epoch": 4.188664183076105, "grad_norm": 0.31083810329437256, "learning_rate": 1.3235836430872937e-08, "loss": 0.1939, "step": 15741 }, { "epoch": 4.188930282064928, "grad_norm": 0.31243109703063965, "learning_rate": 1.3227355187112831e-08, "loss": 0.1949, "step": 15742 }, { "epoch": 4.189196381053752, "grad_norm": 0.3013336956501007, "learning_rate": 1.3218876469079088e-08, "loss": 0.1883, "step": 15743 }, { "epoch": 4.189462480042576, "grad_norm": 0.29854363203048706, "learning_rate": 1.3210400277018518e-08, "loss": 0.1943, "step": 15744 }, { "epoch": 4.189728579031399, "grad_norm": 0.3541388213634491, "learning_rate": 1.320192661117786e-08, "loss": 0.2096, "step": 15745 }, { "epoch": 4.1899946780202235, "grad_norm": 0.29892218112945557, "learning_rate": 1.3193455471803738e-08, "loss": 0.1883, "step": 15746 }, { "epoch": 4.190260777009048, "grad_norm": 0.2891243100166321, "learning_rate": 1.3184986859142743e-08, "loss": 0.191, "step": 15747 }, { "epoch": 4.190526875997871, "grad_norm": 0.2896229922771454, "learning_rate": 1.3176520773441369e-08, "loss": 0.1762, "step": 15748 }, { "epoch": 4.190792974986695, "grad_norm": 0.32141250371932983, "learning_rate": 1.3168057214946038e-08, "loss": 0.2048, "step": 15749 }, { "epoch": 4.191059073975519, "grad_norm": 0.2923101484775543, "learning_rate": 1.315959618390311e-08, "loss": 0.2027, "step": 15750 }, { "epoch": 4.191325172964342, "grad_norm": 0.29837241768836975, "learning_rate": 1.3151137680558877e-08, "loss": 0.1851, "step": 15751 }, { "epoch": 4.1915912719531665, "grad_norm": 0.2967860698699951, "learning_rate": 1.3142681705159498e-08, "loss": 0.1847, "step": 15752 }, { "epoch": 4.191857370941991, "grad_norm": 0.3218958377838135, "learning_rate": 1.313422825795114e-08, "loss": 0.2305, "step": 15753 }, { "epoch": 4.192123469930814, "grad_norm": 0.29339879751205444, "learning_rate": 1.3125777339179866e-08, "loss": 0.1899, "step": 15754 }, { "epoch": 4.192389568919638, "grad_norm": 0.306209921836853, "learning_rate": 1.3117328949091632e-08, "loss": 0.1852, "step": 15755 }, { "epoch": 4.192655667908462, "grad_norm": 0.3300478756427765, "learning_rate": 1.3108883087932366e-08, "loss": 0.2142, "step": 15756 }, { "epoch": 4.192921766897285, "grad_norm": 0.3091743290424347, "learning_rate": 1.310043975594789e-08, "loss": 0.1863, "step": 15757 }, { "epoch": 4.1931878658861095, "grad_norm": 0.3076476752758026, "learning_rate": 1.3091998953383987e-08, "loss": 0.1943, "step": 15758 }, { "epoch": 4.193453964874934, "grad_norm": 0.28691789507865906, "learning_rate": 1.308356068048635e-08, "loss": 0.1784, "step": 15759 }, { "epoch": 4.193720063863758, "grad_norm": 0.30473440885543823, "learning_rate": 1.3075124937500592e-08, "loss": 0.2093, "step": 15760 }, { "epoch": 4.193986162852581, "grad_norm": 0.2974090874195099, "learning_rate": 1.3066691724672241e-08, "loss": 0.1896, "step": 15761 }, { "epoch": 4.194252261841405, "grad_norm": 0.2838122248649597, "learning_rate": 1.3058261042246765e-08, "loss": 0.1997, "step": 15762 }, { "epoch": 4.194518360830229, "grad_norm": 0.32227835059165955, "learning_rate": 1.304983289046957e-08, "loss": 0.1859, "step": 15763 }, { "epoch": 4.1947844598190525, "grad_norm": 0.2892279326915741, "learning_rate": 1.3041407269585968e-08, "loss": 0.189, "step": 15764 }, { "epoch": 4.195050558807877, "grad_norm": 0.31873345375061035, "learning_rate": 1.3032984179841223e-08, "loss": 0.1963, "step": 15765 }, { "epoch": 4.195316657796701, "grad_norm": 0.281190425157547, "learning_rate": 1.30245636214805e-08, "loss": 0.1718, "step": 15766 }, { "epoch": 4.195582756785524, "grad_norm": 0.3032282590866089, "learning_rate": 1.3016145594748906e-08, "loss": 0.1888, "step": 15767 }, { "epoch": 4.195848855774348, "grad_norm": 0.30949270725250244, "learning_rate": 1.3007730099891457e-08, "loss": 0.1764, "step": 15768 }, { "epoch": 4.196114954763172, "grad_norm": 0.29747453331947327, "learning_rate": 1.2999317137153131e-08, "loss": 0.182, "step": 15769 }, { "epoch": 4.196381053751995, "grad_norm": 0.30766957998275757, "learning_rate": 1.299090670677877e-08, "loss": 0.2031, "step": 15770 }, { "epoch": 4.19664715274082, "grad_norm": 0.3035983443260193, "learning_rate": 1.2982498809013198e-08, "loss": 0.1943, "step": 15771 }, { "epoch": 4.196913251729644, "grad_norm": 0.31873762607574463, "learning_rate": 1.2974093444101142e-08, "loss": 0.1824, "step": 15772 }, { "epoch": 4.197179350718467, "grad_norm": 0.3571311831474304, "learning_rate": 1.2965690612287294e-08, "loss": 0.1953, "step": 15773 }, { "epoch": 4.197445449707291, "grad_norm": 0.2989235520362854, "learning_rate": 1.2957290313816182e-08, "loss": 0.1885, "step": 15774 }, { "epoch": 4.197711548696115, "grad_norm": 0.30247658491134644, "learning_rate": 1.2948892548932343e-08, "loss": 0.1979, "step": 15775 }, { "epoch": 4.197977647684938, "grad_norm": 0.30162954330444336, "learning_rate": 1.2940497317880229e-08, "loss": 0.2041, "step": 15776 }, { "epoch": 4.1982437466737625, "grad_norm": 0.3291032016277313, "learning_rate": 1.2932104620904182e-08, "loss": 0.1835, "step": 15777 }, { "epoch": 4.198509845662587, "grad_norm": 0.3211611211299896, "learning_rate": 1.2923714458248514e-08, "loss": 0.2012, "step": 15778 }, { "epoch": 4.19877594465141, "grad_norm": 0.29851725697517395, "learning_rate": 1.2915326830157414e-08, "loss": 0.1978, "step": 15779 }, { "epoch": 4.199042043640234, "grad_norm": 0.27860623598098755, "learning_rate": 1.2906941736875044e-08, "loss": 0.1752, "step": 15780 }, { "epoch": 4.199308142629058, "grad_norm": 0.3004891872406006, "learning_rate": 1.2898559178645451e-08, "loss": 0.1974, "step": 15781 }, { "epoch": 4.199574241617881, "grad_norm": 0.29553207755088806, "learning_rate": 1.2890179155712655e-08, "loss": 0.1806, "step": 15782 }, { "epoch": 4.1998403406067055, "grad_norm": 0.30012640357017517, "learning_rate": 1.2881801668320558e-08, "loss": 0.1722, "step": 15783 }, { "epoch": 4.20010643959553, "grad_norm": 0.30522167682647705, "learning_rate": 1.287342671671301e-08, "loss": 0.2037, "step": 15784 }, { "epoch": 4.200372538584354, "grad_norm": 0.30928924679756165, "learning_rate": 1.2865054301133804e-08, "loss": 0.1845, "step": 15785 }, { "epoch": 4.200638637573177, "grad_norm": 0.31068259477615356, "learning_rate": 1.2856684421826614e-08, "loss": 0.1996, "step": 15786 }, { "epoch": 4.200904736562001, "grad_norm": 0.27563104033470154, "learning_rate": 1.2848317079035098e-08, "loss": 0.1809, "step": 15787 }, { "epoch": 4.201170835550825, "grad_norm": 0.2869395315647125, "learning_rate": 1.2839952273002751e-08, "loss": 0.1914, "step": 15788 }, { "epoch": 4.2014369345396485, "grad_norm": 0.2928507626056671, "learning_rate": 1.2831590003973103e-08, "loss": 0.1731, "step": 15789 }, { "epoch": 4.201703033528473, "grad_norm": 0.316851407289505, "learning_rate": 1.2823230272189533e-08, "loss": 0.2111, "step": 15790 }, { "epoch": 4.201969132517297, "grad_norm": 0.29786574840545654, "learning_rate": 1.2814873077895372e-08, "loss": 0.1931, "step": 15791 }, { "epoch": 4.20223523150612, "grad_norm": 0.26819270849227905, "learning_rate": 1.280651842133389e-08, "loss": 0.179, "step": 15792 }, { "epoch": 4.202501330494944, "grad_norm": 0.2831394374370575, "learning_rate": 1.2798166302748281e-08, "loss": 0.1814, "step": 15793 }, { "epoch": 4.202767429483768, "grad_norm": 0.3120565712451935, "learning_rate": 1.2789816722381608e-08, "loss": 0.1875, "step": 15794 }, { "epoch": 4.2030335284725915, "grad_norm": 0.27552101016044617, "learning_rate": 1.278146968047693e-08, "loss": 0.1933, "step": 15795 }, { "epoch": 4.203299627461416, "grad_norm": 0.3064480423927307, "learning_rate": 1.2773125177277244e-08, "loss": 0.2092, "step": 15796 }, { "epoch": 4.20356572645024, "grad_norm": 0.3257083296775818, "learning_rate": 1.2764783213025366e-08, "loss": 0.2047, "step": 15797 }, { "epoch": 4.203831825439063, "grad_norm": 0.3191569745540619, "learning_rate": 1.2756443787964144e-08, "loss": 0.1984, "step": 15798 }, { "epoch": 4.204097924427887, "grad_norm": 0.29513782262802124, "learning_rate": 1.274810690233633e-08, "loss": 0.1737, "step": 15799 }, { "epoch": 4.204364023416711, "grad_norm": 0.3356194794178009, "learning_rate": 1.2739772556384565e-08, "loss": 0.208, "step": 15800 }, { "epoch": 4.2046301224055345, "grad_norm": 0.28369128704071045, "learning_rate": 1.2731440750351463e-08, "loss": 0.1815, "step": 15801 }, { "epoch": 4.204896221394359, "grad_norm": 0.30620014667510986, "learning_rate": 1.272311148447952e-08, "loss": 0.186, "step": 15802 }, { "epoch": 4.205162320383183, "grad_norm": 0.3081642985343933, "learning_rate": 1.2714784759011198e-08, "loss": 0.1942, "step": 15803 }, { "epoch": 4.205428419372007, "grad_norm": 0.2781318426132202, "learning_rate": 1.2706460574188849e-08, "loss": 0.1772, "step": 15804 }, { "epoch": 4.20569451836083, "grad_norm": 0.28865987062454224, "learning_rate": 1.269813893025481e-08, "loss": 0.2121, "step": 15805 }, { "epoch": 4.205960617349654, "grad_norm": 0.3299844264984131, "learning_rate": 1.2689819827451231e-08, "loss": 0.2089, "step": 15806 }, { "epoch": 4.206226716338478, "grad_norm": 0.2951864004135132, "learning_rate": 1.268150326602031e-08, "loss": 0.1847, "step": 15807 }, { "epoch": 4.206492815327302, "grad_norm": 0.3084244728088379, "learning_rate": 1.2673189246204108e-08, "loss": 0.2173, "step": 15808 }, { "epoch": 4.206758914316126, "grad_norm": 0.2975757122039795, "learning_rate": 1.2664877768244631e-08, "loss": 0.1979, "step": 15809 }, { "epoch": 4.20702501330495, "grad_norm": 0.3022695481777191, "learning_rate": 1.2656568832383785e-08, "loss": 0.1987, "step": 15810 }, { "epoch": 4.207291112293773, "grad_norm": 0.2897850275039673, "learning_rate": 1.2648262438863444e-08, "loss": 0.1947, "step": 15811 }, { "epoch": 4.207557211282597, "grad_norm": 0.28156012296676636, "learning_rate": 1.2639958587925381e-08, "loss": 0.1709, "step": 15812 }, { "epoch": 4.207823310271421, "grad_norm": 0.28369787335395813, "learning_rate": 1.2631657279811314e-08, "loss": 0.1788, "step": 15813 }, { "epoch": 4.2080894092602446, "grad_norm": 0.30777549743652344, "learning_rate": 1.2623358514762828e-08, "loss": 0.1814, "step": 15814 }, { "epoch": 4.208355508249069, "grad_norm": 0.28784647583961487, "learning_rate": 1.2615062293021505e-08, "loss": 0.1914, "step": 15815 }, { "epoch": 4.208621607237893, "grad_norm": 0.2978781461715698, "learning_rate": 1.2606768614828844e-08, "loss": 0.1916, "step": 15816 }, { "epoch": 4.208887706226716, "grad_norm": 0.2991897165775299, "learning_rate": 1.2598477480426218e-08, "loss": 0.1876, "step": 15817 }, { "epoch": 4.20915380521554, "grad_norm": 0.30286720395088196, "learning_rate": 1.2590188890054976e-08, "loss": 0.1865, "step": 15818 }, { "epoch": 4.209419904204364, "grad_norm": 0.3239957094192505, "learning_rate": 1.2581902843956383e-08, "loss": 0.1958, "step": 15819 }, { "epoch": 4.2096860031931875, "grad_norm": 0.3150308132171631, "learning_rate": 1.257361934237161e-08, "loss": 0.1921, "step": 15820 }, { "epoch": 4.209952102182012, "grad_norm": 0.33015331625938416, "learning_rate": 1.256533838554179e-08, "loss": 0.1902, "step": 15821 }, { "epoch": 4.210218201170836, "grad_norm": 0.2918618321418762, "learning_rate": 1.2557059973707939e-08, "loss": 0.1859, "step": 15822 }, { "epoch": 4.210484300159659, "grad_norm": 0.300931841135025, "learning_rate": 1.2548784107111054e-08, "loss": 0.1883, "step": 15823 }, { "epoch": 4.210750399148483, "grad_norm": 0.30614638328552246, "learning_rate": 1.2540510785991975e-08, "loss": 0.1835, "step": 15824 }, { "epoch": 4.211016498137307, "grad_norm": 0.2956545650959015, "learning_rate": 1.2532240010591545e-08, "loss": 0.1943, "step": 15825 }, { "epoch": 4.2112825971261305, "grad_norm": 0.30081233382225037, "learning_rate": 1.2523971781150489e-08, "loss": 0.1961, "step": 15826 }, { "epoch": 4.211548696114955, "grad_norm": 0.30088239908218384, "learning_rate": 1.2515706097909495e-08, "loss": 0.1875, "step": 15827 }, { "epoch": 4.211814795103779, "grad_norm": 0.28375858068466187, "learning_rate": 1.2507442961109149e-08, "loss": 0.182, "step": 15828 }, { "epoch": 4.212080894092603, "grad_norm": 0.27592360973358154, "learning_rate": 1.2499182370989957e-08, "loss": 0.1712, "step": 15829 }, { "epoch": 4.212346993081426, "grad_norm": 0.316301554441452, "learning_rate": 1.2490924327792373e-08, "loss": 0.2113, "step": 15830 }, { "epoch": 4.21261309207025, "grad_norm": 0.2827289402484894, "learning_rate": 1.2482668831756771e-08, "loss": 0.1887, "step": 15831 }, { "epoch": 4.212879191059074, "grad_norm": 0.2830166816711426, "learning_rate": 1.2474415883123467e-08, "loss": 0.1817, "step": 15832 }, { "epoch": 4.213145290047898, "grad_norm": 0.30546677112579346, "learning_rate": 1.246616548213263e-08, "loss": 0.2043, "step": 15833 }, { "epoch": 4.213411389036722, "grad_norm": 0.2937859892845154, "learning_rate": 1.245791762902444e-08, "loss": 0.1847, "step": 15834 }, { "epoch": 4.213677488025546, "grad_norm": 0.31696033477783203, "learning_rate": 1.2449672324038996e-08, "loss": 0.2027, "step": 15835 }, { "epoch": 4.213943587014369, "grad_norm": 0.2836441695690155, "learning_rate": 1.2441429567416239e-08, "loss": 0.1805, "step": 15836 }, { "epoch": 4.214209686003193, "grad_norm": 0.32849252223968506, "learning_rate": 1.2433189359396123e-08, "loss": 0.21, "step": 15837 }, { "epoch": 4.214475784992017, "grad_norm": 0.27205052971839905, "learning_rate": 1.24249517002185e-08, "loss": 0.1618, "step": 15838 }, { "epoch": 4.214741883980841, "grad_norm": 0.29525113105773926, "learning_rate": 1.2416716590123144e-08, "loss": 0.1873, "step": 15839 }, { "epoch": 4.215007982969665, "grad_norm": 0.3064463138580322, "learning_rate": 1.2408484029349764e-08, "loss": 0.1947, "step": 15840 }, { "epoch": 4.215274081958489, "grad_norm": 0.3267022371292114, "learning_rate": 1.2400254018138013e-08, "loss": 0.2247, "step": 15841 }, { "epoch": 4.215540180947312, "grad_norm": 0.3100523352622986, "learning_rate": 1.2392026556727386e-08, "loss": 0.1974, "step": 15842 }, { "epoch": 4.215806279936136, "grad_norm": 0.3020722270011902, "learning_rate": 1.2383801645357395e-08, "loss": 0.1988, "step": 15843 }, { "epoch": 4.21607237892496, "grad_norm": 0.3173386752605438, "learning_rate": 1.2375579284267456e-08, "loss": 0.1901, "step": 15844 }, { "epoch": 4.216338477913784, "grad_norm": 0.3796420991420746, "learning_rate": 1.236735947369688e-08, "loss": 0.2091, "step": 15845 }, { "epoch": 4.216604576902608, "grad_norm": 0.3246004581451416, "learning_rate": 1.2359142213884932e-08, "loss": 0.1829, "step": 15846 }, { "epoch": 4.216870675891432, "grad_norm": 0.30805182456970215, "learning_rate": 1.2350927505070797e-08, "loss": 0.2025, "step": 15847 }, { "epoch": 4.217136774880255, "grad_norm": 0.3038491904735565, "learning_rate": 1.2342715347493593e-08, "loss": 0.199, "step": 15848 }, { "epoch": 4.217402873869079, "grad_norm": 0.2974189519882202, "learning_rate": 1.2334505741392332e-08, "loss": 0.1862, "step": 15849 }, { "epoch": 4.217668972857903, "grad_norm": 0.2995971143245697, "learning_rate": 1.2326298687005998e-08, "loss": 0.1835, "step": 15850 }, { "epoch": 4.217935071846727, "grad_norm": 0.31104743480682373, "learning_rate": 1.231809418457349e-08, "loss": 0.1897, "step": 15851 }, { "epoch": 4.218201170835551, "grad_norm": 0.28502780199050903, "learning_rate": 1.2309892234333563e-08, "loss": 0.1605, "step": 15852 }, { "epoch": 4.218467269824375, "grad_norm": 0.28760436177253723, "learning_rate": 1.2301692836525002e-08, "loss": 0.1864, "step": 15853 }, { "epoch": 4.218733368813199, "grad_norm": 0.276520699262619, "learning_rate": 1.2293495991386448e-08, "loss": 0.1725, "step": 15854 }, { "epoch": 4.218999467802022, "grad_norm": 0.28142932057380676, "learning_rate": 1.2285301699156515e-08, "loss": 0.1725, "step": 15855 }, { "epoch": 4.219265566790846, "grad_norm": 0.3205374777317047, "learning_rate": 1.2277109960073684e-08, "loss": 0.1922, "step": 15856 }, { "epoch": 4.21953166577967, "grad_norm": 0.2731729745864868, "learning_rate": 1.2268920774376413e-08, "loss": 0.1839, "step": 15857 }, { "epoch": 4.219797764768494, "grad_norm": 0.2630573809146881, "learning_rate": 1.2260734142303075e-08, "loss": 0.1655, "step": 15858 }, { "epoch": 4.220063863757318, "grad_norm": 0.2929573059082031, "learning_rate": 1.2252550064091939e-08, "loss": 0.1846, "step": 15859 }, { "epoch": 4.220329962746142, "grad_norm": 0.2910377085208893, "learning_rate": 1.2244368539981264e-08, "loss": 0.1755, "step": 15860 }, { "epoch": 4.220596061734965, "grad_norm": 0.3010345995426178, "learning_rate": 1.2236189570209144e-08, "loss": 0.2041, "step": 15861 }, { "epoch": 4.220862160723789, "grad_norm": 0.31165197491645813, "learning_rate": 1.2228013155013673e-08, "loss": 0.1932, "step": 15862 }, { "epoch": 4.221128259712613, "grad_norm": 0.30263614654541016, "learning_rate": 1.2219839294632838e-08, "loss": 0.1968, "step": 15863 }, { "epoch": 4.221394358701437, "grad_norm": 0.32027047872543335, "learning_rate": 1.2211667989304563e-08, "loss": 0.1965, "step": 15864 }, { "epoch": 4.221660457690261, "grad_norm": 0.30377379059791565, "learning_rate": 1.2203499239266691e-08, "loss": 0.1899, "step": 15865 }, { "epoch": 4.221926556679085, "grad_norm": 0.31438305974006653, "learning_rate": 1.2195333044756994e-08, "loss": 0.2207, "step": 15866 }, { "epoch": 4.222192655667908, "grad_norm": 0.29252126812934875, "learning_rate": 1.2187169406013164e-08, "loss": 0.1879, "step": 15867 }, { "epoch": 4.222458754656732, "grad_norm": 0.28372806310653687, "learning_rate": 1.2179008323272843e-08, "loss": 0.1857, "step": 15868 }, { "epoch": 4.222724853645556, "grad_norm": 0.32703444361686707, "learning_rate": 1.2170849796773574e-08, "loss": 0.2084, "step": 15869 }, { "epoch": 4.22299095263438, "grad_norm": 0.2969551980495453, "learning_rate": 1.2162693826752812e-08, "loss": 0.1794, "step": 15870 }, { "epoch": 4.223257051623204, "grad_norm": 0.4275880753993988, "learning_rate": 1.2154540413447956e-08, "loss": 0.188, "step": 15871 }, { "epoch": 4.223523150612028, "grad_norm": 0.3112579882144928, "learning_rate": 1.2146389557096337e-08, "loss": 0.19, "step": 15872 }, { "epoch": 4.223789249600851, "grad_norm": 0.33070436120033264, "learning_rate": 1.2138241257935222e-08, "loss": 0.2187, "step": 15873 }, { "epoch": 4.224055348589675, "grad_norm": 0.2974364161491394, "learning_rate": 1.2130095516201789e-08, "loss": 0.1977, "step": 15874 }, { "epoch": 4.224321447578499, "grad_norm": 0.2904689311981201, "learning_rate": 1.2121952332133089e-08, "loss": 0.1841, "step": 15875 }, { "epoch": 4.224587546567323, "grad_norm": 0.2889224588871002, "learning_rate": 1.2113811705966204e-08, "loss": 0.1724, "step": 15876 }, { "epoch": 4.224853645556147, "grad_norm": 0.2739708721637726, "learning_rate": 1.2105673637938053e-08, "loss": 0.1663, "step": 15877 }, { "epoch": 4.225119744544971, "grad_norm": 0.3107408285140991, "learning_rate": 1.2097538128285555e-08, "loss": 0.2107, "step": 15878 }, { "epoch": 4.225385843533795, "grad_norm": 0.2907096743583679, "learning_rate": 1.2089405177245459e-08, "loss": 0.1832, "step": 15879 }, { "epoch": 4.225651942522618, "grad_norm": 0.2980996072292328, "learning_rate": 1.2081274785054518e-08, "loss": 0.1847, "step": 15880 }, { "epoch": 4.225918041511442, "grad_norm": 0.3202560245990753, "learning_rate": 1.2073146951949398e-08, "loss": 0.2089, "step": 15881 }, { "epoch": 4.2261841405002665, "grad_norm": 0.29132235050201416, "learning_rate": 1.2065021678166665e-08, "loss": 0.1828, "step": 15882 }, { "epoch": 4.22645023948909, "grad_norm": 0.2812774181365967, "learning_rate": 1.2056898963942829e-08, "loss": 0.1845, "step": 15883 }, { "epoch": 4.226716338477914, "grad_norm": 0.2804855704307556, "learning_rate": 1.2048778809514326e-08, "loss": 0.1676, "step": 15884 }, { "epoch": 4.226982437466738, "grad_norm": 0.3193986117839813, "learning_rate": 1.204066121511751e-08, "loss": 0.1775, "step": 15885 }, { "epoch": 4.227248536455561, "grad_norm": 0.32279732823371887, "learning_rate": 1.2032546180988657e-08, "loss": 0.2095, "step": 15886 }, { "epoch": 4.227514635444385, "grad_norm": 0.26975226402282715, "learning_rate": 1.2024433707364002e-08, "loss": 0.1617, "step": 15887 }, { "epoch": 4.2277807344332095, "grad_norm": 0.3329581022262573, "learning_rate": 1.2016323794479644e-08, "loss": 0.1862, "step": 15888 }, { "epoch": 4.228046833422033, "grad_norm": 0.2834254503250122, "learning_rate": 1.200821644257164e-08, "loss": 0.1871, "step": 15889 }, { "epoch": 4.228312932410857, "grad_norm": 0.3171207904815674, "learning_rate": 1.2000111651875988e-08, "loss": 0.2141, "step": 15890 }, { "epoch": 4.228579031399681, "grad_norm": 0.2893223464488983, "learning_rate": 1.1992009422628601e-08, "loss": 0.1917, "step": 15891 }, { "epoch": 4.228845130388504, "grad_norm": 0.2982235848903656, "learning_rate": 1.1983909755065313e-08, "loss": 0.2066, "step": 15892 }, { "epoch": 4.229111229377328, "grad_norm": 0.291236013174057, "learning_rate": 1.197581264942188e-08, "loss": 0.1832, "step": 15893 }, { "epoch": 4.229377328366152, "grad_norm": 0.2793480455875397, "learning_rate": 1.1967718105934e-08, "loss": 0.1831, "step": 15894 }, { "epoch": 4.229643427354976, "grad_norm": 0.28427135944366455, "learning_rate": 1.1959626124837252e-08, "loss": 0.1693, "step": 15895 }, { "epoch": 4.2299095263438, "grad_norm": 0.3406815826892853, "learning_rate": 1.1951536706367204e-08, "loss": 0.1965, "step": 15896 }, { "epoch": 4.230175625332624, "grad_norm": 0.2896782457828522, "learning_rate": 1.1943449850759324e-08, "loss": 0.192, "step": 15897 }, { "epoch": 4.230441724321447, "grad_norm": 0.30241715908050537, "learning_rate": 1.1935365558248955e-08, "loss": 0.1841, "step": 15898 }, { "epoch": 4.230707823310271, "grad_norm": 0.2979249358177185, "learning_rate": 1.1927283829071444e-08, "loss": 0.1965, "step": 15899 }, { "epoch": 4.230973922299095, "grad_norm": 0.28887951374053955, "learning_rate": 1.1919204663462013e-08, "loss": 0.1879, "step": 15900 }, { "epoch": 4.231240021287919, "grad_norm": 0.30860990285873413, "learning_rate": 1.191112806165584e-08, "loss": 0.1957, "step": 15901 }, { "epoch": 4.231506120276743, "grad_norm": 0.29683175683021545, "learning_rate": 1.1903054023888015e-08, "loss": 0.1782, "step": 15902 }, { "epoch": 4.231772219265567, "grad_norm": 0.2937030792236328, "learning_rate": 1.189498255039354e-08, "loss": 0.1797, "step": 15903 }, { "epoch": 4.232038318254391, "grad_norm": 0.2855883240699768, "learning_rate": 1.188691364140736e-08, "loss": 0.1925, "step": 15904 }, { "epoch": 4.232304417243214, "grad_norm": 0.30942127108573914, "learning_rate": 1.1878847297164363e-08, "loss": 0.2093, "step": 15905 }, { "epoch": 4.232570516232038, "grad_norm": 0.2860352396965027, "learning_rate": 1.1870783517899296e-08, "loss": 0.1824, "step": 15906 }, { "epoch": 4.2328366152208625, "grad_norm": 0.31103819608688354, "learning_rate": 1.1862722303846895e-08, "loss": 0.1964, "step": 15907 }, { "epoch": 4.233102714209686, "grad_norm": 0.32271480560302734, "learning_rate": 1.1854663655241803e-08, "loss": 0.2093, "step": 15908 }, { "epoch": 4.23336881319851, "grad_norm": 0.2876739203929901, "learning_rate": 1.1846607572318579e-08, "loss": 0.1925, "step": 15909 }, { "epoch": 4.233634912187334, "grad_norm": 0.3065698742866516, "learning_rate": 1.1838554055311722e-08, "loss": 0.1988, "step": 15910 }, { "epoch": 4.233901011176157, "grad_norm": 0.3007362484931946, "learning_rate": 1.1830503104455659e-08, "loss": 0.1855, "step": 15911 }, { "epoch": 4.234167110164981, "grad_norm": 0.29726317524909973, "learning_rate": 1.182245471998471e-08, "loss": 0.193, "step": 15912 }, { "epoch": 4.2344332091538055, "grad_norm": 0.3046054244041443, "learning_rate": 1.1814408902133155e-08, "loss": 0.1847, "step": 15913 }, { "epoch": 4.234699308142629, "grad_norm": 0.28472959995269775, "learning_rate": 1.1806365651135208e-08, "loss": 0.1859, "step": 15914 }, { "epoch": 4.234965407131453, "grad_norm": 0.3095085024833679, "learning_rate": 1.1798324967224937e-08, "loss": 0.192, "step": 15915 }, { "epoch": 4.235231506120277, "grad_norm": 0.296328604221344, "learning_rate": 1.1790286850636444e-08, "loss": 0.1892, "step": 15916 }, { "epoch": 4.2354976051091, "grad_norm": 0.3032933473587036, "learning_rate": 1.178225130160364e-08, "loss": 0.1776, "step": 15917 }, { "epoch": 4.235763704097924, "grad_norm": 0.2757885754108429, "learning_rate": 1.1774218320360452e-08, "loss": 0.1804, "step": 15918 }, { "epoch": 4.2360298030867485, "grad_norm": 0.29588782787323, "learning_rate": 1.176618790714069e-08, "loss": 0.1868, "step": 15919 }, { "epoch": 4.236295902075572, "grad_norm": 0.2743972837924957, "learning_rate": 1.1758160062178091e-08, "loss": 0.1702, "step": 15920 }, { "epoch": 4.236562001064396, "grad_norm": 0.2901364862918854, "learning_rate": 1.1750134785706345e-08, "loss": 0.1883, "step": 15921 }, { "epoch": 4.23682810005322, "grad_norm": 0.28897470235824585, "learning_rate": 1.1742112077959033e-08, "loss": 0.1793, "step": 15922 }, { "epoch": 4.237094199042044, "grad_norm": 0.31733739376068115, "learning_rate": 1.17340919391697e-08, "loss": 0.1818, "step": 15923 }, { "epoch": 4.237360298030867, "grad_norm": 0.309022456407547, "learning_rate": 1.1726074369571758e-08, "loss": 0.2, "step": 15924 }, { "epoch": 4.2376263970196915, "grad_norm": 0.2771216034889221, "learning_rate": 1.171805936939858e-08, "loss": 0.1796, "step": 15925 }, { "epoch": 4.237892496008516, "grad_norm": 0.31736230850219727, "learning_rate": 1.1710046938883478e-08, "loss": 0.1848, "step": 15926 }, { "epoch": 4.238158594997339, "grad_norm": 0.2892654538154602, "learning_rate": 1.1702037078259663e-08, "loss": 0.181, "step": 15927 }, { "epoch": 4.238424693986163, "grad_norm": 0.28629788756370544, "learning_rate": 1.1694029787760285e-08, "loss": 0.1704, "step": 15928 }, { "epoch": 4.238690792974987, "grad_norm": 0.3061569631099701, "learning_rate": 1.1686025067618421e-08, "loss": 0.1904, "step": 15929 }, { "epoch": 4.23895689196381, "grad_norm": 0.31181567907333374, "learning_rate": 1.1678022918067053e-08, "loss": 0.1986, "step": 15930 }, { "epoch": 4.239222990952634, "grad_norm": 0.3083740770816803, "learning_rate": 1.1670023339339119e-08, "loss": 0.2022, "step": 15931 }, { "epoch": 4.239489089941459, "grad_norm": 0.31158947944641113, "learning_rate": 1.1662026331667485e-08, "loss": 0.196, "step": 15932 }, { "epoch": 4.239755188930282, "grad_norm": 0.2861368656158447, "learning_rate": 1.1654031895284865e-08, "loss": 0.187, "step": 15933 }, { "epoch": 4.240021287919106, "grad_norm": 0.3114456832408905, "learning_rate": 1.1646040030423997e-08, "loss": 0.1988, "step": 15934 }, { "epoch": 4.24028738690793, "grad_norm": 0.2769545614719391, "learning_rate": 1.1638050737317495e-08, "loss": 0.1782, "step": 15935 }, { "epoch": 4.240553485896753, "grad_norm": 0.3067481219768524, "learning_rate": 1.1630064016197916e-08, "loss": 0.198, "step": 15936 }, { "epoch": 4.240819584885577, "grad_norm": 0.29867079854011536, "learning_rate": 1.1622079867297718e-08, "loss": 0.1892, "step": 15937 }, { "epoch": 4.2410856838744015, "grad_norm": 0.32500705122947693, "learning_rate": 1.1614098290849295e-08, "loss": 0.1989, "step": 15938 }, { "epoch": 4.241351782863225, "grad_norm": 0.2773030400276184, "learning_rate": 1.160611928708498e-08, "loss": 0.1851, "step": 15939 }, { "epoch": 4.241617881852049, "grad_norm": 0.27914562821388245, "learning_rate": 1.1598142856237014e-08, "loss": 0.1655, "step": 15940 }, { "epoch": 4.241883980840873, "grad_norm": 0.2873470187187195, "learning_rate": 1.1590168998537608e-08, "loss": 0.1921, "step": 15941 }, { "epoch": 4.242150079829696, "grad_norm": 0.2937721610069275, "learning_rate": 1.15821977142188e-08, "loss": 0.1889, "step": 15942 }, { "epoch": 4.24241617881852, "grad_norm": 0.32439717650413513, "learning_rate": 1.1574229003512647e-08, "loss": 0.2079, "step": 15943 }, { "epoch": 4.2426822778073445, "grad_norm": 0.28980502486228943, "learning_rate": 1.15662628666511e-08, "loss": 0.1784, "step": 15944 }, { "epoch": 4.242948376796168, "grad_norm": 0.29237422347068787, "learning_rate": 1.1558299303866015e-08, "loss": 0.1943, "step": 15945 }, { "epoch": 4.243214475784992, "grad_norm": 0.31407544016838074, "learning_rate": 1.1550338315389197e-08, "loss": 0.1898, "step": 15946 }, { "epoch": 4.243480574773816, "grad_norm": 0.27837473154067993, "learning_rate": 1.1542379901452382e-08, "loss": 0.1901, "step": 15947 }, { "epoch": 4.24374667376264, "grad_norm": 0.3145534098148346, "learning_rate": 1.1534424062287207e-08, "loss": 0.209, "step": 15948 }, { "epoch": 4.244012772751463, "grad_norm": 0.28381985425949097, "learning_rate": 1.1526470798125243e-08, "loss": 0.1694, "step": 15949 }, { "epoch": 4.2442788717402875, "grad_norm": 0.299656480550766, "learning_rate": 1.1518520109198026e-08, "loss": 0.1782, "step": 15950 }, { "epoch": 4.244544970729112, "grad_norm": 0.26706036925315857, "learning_rate": 1.1510571995736928e-08, "loss": 0.1769, "step": 15951 }, { "epoch": 4.244811069717935, "grad_norm": 0.31530171632766724, "learning_rate": 1.1502626457973308e-08, "loss": 0.2256, "step": 15952 }, { "epoch": 4.245077168706759, "grad_norm": 0.2960880398750305, "learning_rate": 1.1494683496138458e-08, "loss": 0.1794, "step": 15953 }, { "epoch": 4.245343267695583, "grad_norm": 0.31715336441993713, "learning_rate": 1.148674311046357e-08, "loss": 0.1907, "step": 15954 }, { "epoch": 4.245609366684406, "grad_norm": 0.2874879240989685, "learning_rate": 1.1478805301179762e-08, "loss": 0.1964, "step": 15955 }, { "epoch": 4.2458754656732305, "grad_norm": 0.29651081562042236, "learning_rate": 1.1470870068518112e-08, "loss": 0.1886, "step": 15956 }, { "epoch": 4.246141564662055, "grad_norm": 0.31353759765625, "learning_rate": 1.1462937412709562e-08, "loss": 0.1891, "step": 15957 }, { "epoch": 4.246407663650878, "grad_norm": 0.2960517406463623, "learning_rate": 1.1455007333985001e-08, "loss": 0.184, "step": 15958 }, { "epoch": 4.246673762639702, "grad_norm": 0.2637540102005005, "learning_rate": 1.1447079832575301e-08, "loss": 0.1744, "step": 15959 }, { "epoch": 4.246939861628526, "grad_norm": 0.30493661761283875, "learning_rate": 1.1439154908711169e-08, "loss": 0.1862, "step": 15960 }, { "epoch": 4.247205960617349, "grad_norm": 0.2895970046520233, "learning_rate": 1.1431232562623283e-08, "loss": 0.1848, "step": 15961 }, { "epoch": 4.2474720596061735, "grad_norm": 0.3200891613960266, "learning_rate": 1.142331279454225e-08, "loss": 0.1807, "step": 15962 }, { "epoch": 4.247738158594998, "grad_norm": 0.28449586033821106, "learning_rate": 1.1415395604698596e-08, "loss": 0.1906, "step": 15963 }, { "epoch": 4.248004257583821, "grad_norm": 0.26913878321647644, "learning_rate": 1.140748099332276e-08, "loss": 0.165, "step": 15964 }, { "epoch": 4.248270356572645, "grad_norm": 0.277697890996933, "learning_rate": 1.1399568960645135e-08, "loss": 0.1803, "step": 15965 }, { "epoch": 4.248536455561469, "grad_norm": 0.3046564757823944, "learning_rate": 1.1391659506896012e-08, "loss": 0.2055, "step": 15966 }, { "epoch": 4.248802554550292, "grad_norm": 0.30507516860961914, "learning_rate": 1.1383752632305598e-08, "loss": 0.1969, "step": 15967 }, { "epoch": 4.2490686535391164, "grad_norm": 0.3155054748058319, "learning_rate": 1.1375848337104088e-08, "loss": 0.1994, "step": 15968 }, { "epoch": 4.249334752527941, "grad_norm": 0.3133808374404907, "learning_rate": 1.1367946621521497e-08, "loss": 0.1745, "step": 15969 }, { "epoch": 4.249600851516764, "grad_norm": 0.33393824100494385, "learning_rate": 1.136004748578785e-08, "loss": 0.2029, "step": 15970 }, { "epoch": 4.249866950505588, "grad_norm": 0.27662193775177, "learning_rate": 1.1352150930133065e-08, "loss": 0.1537, "step": 15971 }, { "epoch": 4.250133049494412, "grad_norm": 0.2842680811882019, "learning_rate": 1.1344256954787002e-08, "loss": 0.1738, "step": 15972 }, { "epoch": 4.250399148483236, "grad_norm": 0.3017492890357971, "learning_rate": 1.1336365559979421e-08, "loss": 0.1877, "step": 15973 }, { "epoch": 4.250665247472059, "grad_norm": 0.28780922293663025, "learning_rate": 1.132847674594003e-08, "loss": 0.2007, "step": 15974 }, { "epoch": 4.2509313464608836, "grad_norm": 0.2918495237827301, "learning_rate": 1.1320590512898465e-08, "loss": 0.1807, "step": 15975 }, { "epoch": 4.251197445449708, "grad_norm": 0.28331443667411804, "learning_rate": 1.1312706861084232e-08, "loss": 0.1806, "step": 15976 }, { "epoch": 4.251463544438531, "grad_norm": 0.290290504693985, "learning_rate": 1.1304825790726836e-08, "loss": 0.1852, "step": 15977 }, { "epoch": 4.251729643427355, "grad_norm": 0.29445329308509827, "learning_rate": 1.1296947302055681e-08, "loss": 0.1789, "step": 15978 }, { "epoch": 4.251995742416179, "grad_norm": 0.2977920472621918, "learning_rate": 1.1289071395300053e-08, "loss": 0.1854, "step": 15979 }, { "epoch": 4.252261841405002, "grad_norm": 0.2659049928188324, "learning_rate": 1.1281198070689234e-08, "loss": 0.1791, "step": 15980 }, { "epoch": 4.2525279403938265, "grad_norm": 0.29597142338752747, "learning_rate": 1.1273327328452376e-08, "loss": 0.2065, "step": 15981 }, { "epoch": 4.252794039382651, "grad_norm": 0.28860801458358765, "learning_rate": 1.1265459168818581e-08, "loss": 0.1966, "step": 15982 }, { "epoch": 4.253060138371474, "grad_norm": 0.3289570212364197, "learning_rate": 1.1257593592016867e-08, "loss": 0.196, "step": 15983 }, { "epoch": 4.253326237360298, "grad_norm": 0.29878100752830505, "learning_rate": 1.1249730598276197e-08, "loss": 0.216, "step": 15984 }, { "epoch": 4.253592336349122, "grad_norm": 0.2653738856315613, "learning_rate": 1.1241870187825419e-08, "loss": 0.1801, "step": 15985 }, { "epoch": 4.253858435337945, "grad_norm": 0.2924254238605499, "learning_rate": 1.1234012360893374e-08, "loss": 0.1791, "step": 15986 }, { "epoch": 4.2541245343267695, "grad_norm": 0.31196001172065735, "learning_rate": 1.1226157117708724e-08, "loss": 0.1743, "step": 15987 }, { "epoch": 4.254390633315594, "grad_norm": 0.3085087537765503, "learning_rate": 1.121830445850015e-08, "loss": 0.1988, "step": 15988 }, { "epoch": 4.254656732304417, "grad_norm": 0.29194122552871704, "learning_rate": 1.1210454383496203e-08, "loss": 0.186, "step": 15989 }, { "epoch": 4.254922831293241, "grad_norm": 0.31356048583984375, "learning_rate": 1.1202606892925392e-08, "loss": 0.2045, "step": 15990 }, { "epoch": 4.255188930282065, "grad_norm": 0.29879188537597656, "learning_rate": 1.1194761987016132e-08, "loss": 0.1998, "step": 15991 }, { "epoch": 4.255455029270888, "grad_norm": 0.30804911255836487, "learning_rate": 1.1186919665996774e-08, "loss": 0.2054, "step": 15992 }, { "epoch": 4.2557211282597125, "grad_norm": 0.2908015847206116, "learning_rate": 1.1179079930095591e-08, "loss": 0.1735, "step": 15993 }, { "epoch": 4.255987227248537, "grad_norm": 0.31000807881355286, "learning_rate": 1.1171242779540757e-08, "loss": 0.1812, "step": 15994 }, { "epoch": 4.25625332623736, "grad_norm": 0.3135233223438263, "learning_rate": 1.1163408214560432e-08, "loss": 0.2041, "step": 15995 }, { "epoch": 4.256519425226184, "grad_norm": 0.2603212893009186, "learning_rate": 1.1155576235382614e-08, "loss": 0.1755, "step": 15996 }, { "epoch": 4.256785524215008, "grad_norm": 0.28389546275138855, "learning_rate": 1.1147746842235283e-08, "loss": 0.1892, "step": 15997 }, { "epoch": 4.257051623203832, "grad_norm": 0.2753848731517792, "learning_rate": 1.1139920035346362e-08, "loss": 0.1676, "step": 15998 }, { "epoch": 4.2573177221926555, "grad_norm": 0.29410243034362793, "learning_rate": 1.113209581494362e-08, "loss": 0.1815, "step": 15999 }, { "epoch": 4.25758382118148, "grad_norm": 0.29664140939712524, "learning_rate": 1.1124274181254834e-08, "loss": 0.1829, "step": 16000 }, { "epoch": 4.257849920170304, "grad_norm": 0.30851417779922485, "learning_rate": 1.1116455134507664e-08, "loss": 0.2154, "step": 16001 }, { "epoch": 4.258116019159127, "grad_norm": 0.3145752549171448, "learning_rate": 1.1108638674929694e-08, "loss": 0.1941, "step": 16002 }, { "epoch": 4.258382118147951, "grad_norm": 0.30846109986305237, "learning_rate": 1.1100824802748443e-08, "loss": 0.1975, "step": 16003 }, { "epoch": 4.258648217136775, "grad_norm": 0.29060444235801697, "learning_rate": 1.1093013518191397e-08, "loss": 0.1692, "step": 16004 }, { "epoch": 4.2589143161255985, "grad_norm": 0.2963078022003174, "learning_rate": 1.108520482148585e-08, "loss": 0.1752, "step": 16005 }, { "epoch": 4.259180415114423, "grad_norm": 0.31263771653175354, "learning_rate": 1.107739871285912e-08, "loss": 0.1763, "step": 16006 }, { "epoch": 4.259446514103247, "grad_norm": 0.29094427824020386, "learning_rate": 1.106959519253844e-08, "loss": 0.1812, "step": 16007 }, { "epoch": 4.25971261309207, "grad_norm": 0.30169329047203064, "learning_rate": 1.1061794260750934e-08, "loss": 0.207, "step": 16008 }, { "epoch": 4.259978712080894, "grad_norm": 0.2879827916622162, "learning_rate": 1.105399591772367e-08, "loss": 0.1931, "step": 16009 }, { "epoch": 4.260244811069718, "grad_norm": 0.3073805868625641, "learning_rate": 1.1046200163683639e-08, "loss": 0.169, "step": 16010 }, { "epoch": 4.260510910058541, "grad_norm": 0.3080954849720001, "learning_rate": 1.1038406998857753e-08, "loss": 0.1922, "step": 16011 }, { "epoch": 4.260777009047366, "grad_norm": 0.3020250201225281, "learning_rate": 1.1030616423472861e-08, "loss": 0.1981, "step": 16012 }, { "epoch": 4.26104310803619, "grad_norm": 0.316672682762146, "learning_rate": 1.1022828437755726e-08, "loss": 0.1786, "step": 16013 }, { "epoch": 4.261309207025013, "grad_norm": 0.30422794818878174, "learning_rate": 1.1015043041933025e-08, "loss": 0.186, "step": 16014 }, { "epoch": 4.261575306013837, "grad_norm": 0.2918505370616913, "learning_rate": 1.1007260236231364e-08, "loss": 0.1907, "step": 16015 }, { "epoch": 4.261841405002661, "grad_norm": 0.30631157755851746, "learning_rate": 1.0999480020877283e-08, "loss": 0.194, "step": 16016 }, { "epoch": 4.262107503991485, "grad_norm": 0.3186366856098175, "learning_rate": 1.099170239609728e-08, "loss": 0.198, "step": 16017 }, { "epoch": 4.2623736029803085, "grad_norm": 0.3055863082408905, "learning_rate": 1.0983927362117695e-08, "loss": 0.1888, "step": 16018 }, { "epoch": 4.262639701969133, "grad_norm": 0.31369549036026, "learning_rate": 1.0976154919164848e-08, "loss": 0.2077, "step": 16019 }, { "epoch": 4.262905800957956, "grad_norm": 0.2895728647708893, "learning_rate": 1.0968385067464991e-08, "loss": 0.1966, "step": 16020 }, { "epoch": 4.26317189994678, "grad_norm": 0.2616560757160187, "learning_rate": 1.0960617807244276e-08, "loss": 0.172, "step": 16021 }, { "epoch": 4.263437998935604, "grad_norm": 0.32149964570999146, "learning_rate": 1.0952853138728801e-08, "loss": 0.2124, "step": 16022 }, { "epoch": 4.263704097924428, "grad_norm": 0.31325647234916687, "learning_rate": 1.0945091062144551e-08, "loss": 0.1977, "step": 16023 }, { "epoch": 4.2639701969132515, "grad_norm": 0.28704914450645447, "learning_rate": 1.0937331577717469e-08, "loss": 0.1958, "step": 16024 }, { "epoch": 4.264236295902076, "grad_norm": 0.2997637689113617, "learning_rate": 1.0929574685673403e-08, "loss": 0.1755, "step": 16025 }, { "epoch": 4.2645023948909, "grad_norm": 0.27997541427612305, "learning_rate": 1.0921820386238167e-08, "loss": 0.1879, "step": 16026 }, { "epoch": 4.264768493879723, "grad_norm": 0.31854185461997986, "learning_rate": 1.0914068679637445e-08, "loss": 0.1944, "step": 16027 }, { "epoch": 4.265034592868547, "grad_norm": 0.29372337460517883, "learning_rate": 1.0906319566096867e-08, "loss": 0.1924, "step": 16028 }, { "epoch": 4.265300691857371, "grad_norm": 0.3090228736400604, "learning_rate": 1.0898573045842007e-08, "loss": 0.1854, "step": 16029 }, { "epoch": 4.2655667908461945, "grad_norm": 0.3068271279335022, "learning_rate": 1.0890829119098333e-08, "loss": 0.1818, "step": 16030 }, { "epoch": 4.265832889835019, "grad_norm": 0.2684980630874634, "learning_rate": 1.0883087786091271e-08, "loss": 0.1754, "step": 16031 }, { "epoch": 4.266098988823843, "grad_norm": 0.294203519821167, "learning_rate": 1.0875349047046113e-08, "loss": 0.1862, "step": 16032 }, { "epoch": 4.266365087812666, "grad_norm": 0.3501581847667694, "learning_rate": 1.0867612902188138e-08, "loss": 0.2041, "step": 16033 }, { "epoch": 4.26663118680149, "grad_norm": 0.28569814562797546, "learning_rate": 1.0859879351742528e-08, "loss": 0.1776, "step": 16034 }, { "epoch": 4.266897285790314, "grad_norm": 0.29004937410354614, "learning_rate": 1.0852148395934369e-08, "loss": 0.1872, "step": 16035 }, { "epoch": 4.2671633847791375, "grad_norm": 0.3014187216758728, "learning_rate": 1.08444200349887e-08, "loss": 0.2057, "step": 16036 }, { "epoch": 4.267429483767962, "grad_norm": 0.3001822233200073, "learning_rate": 1.0836694269130498e-08, "loss": 0.1881, "step": 16037 }, { "epoch": 4.267695582756786, "grad_norm": 0.29380497336387634, "learning_rate": 1.0828971098584594e-08, "loss": 0.1659, "step": 16038 }, { "epoch": 4.267961681745609, "grad_norm": 0.3284164369106293, "learning_rate": 1.082125052357582e-08, "loss": 0.2057, "step": 16039 }, { "epoch": 4.268227780734433, "grad_norm": 0.30059072375297546, "learning_rate": 1.0813532544328897e-08, "loss": 0.1965, "step": 16040 }, { "epoch": 4.268493879723257, "grad_norm": 0.2992050051689148, "learning_rate": 1.0805817161068464e-08, "loss": 0.1995, "step": 16041 }, { "epoch": 4.268759978712081, "grad_norm": 0.28495681285858154, "learning_rate": 1.0798104374019101e-08, "loss": 0.176, "step": 16042 }, { "epoch": 4.269026077700905, "grad_norm": 0.3155985176563263, "learning_rate": 1.0790394183405305e-08, "loss": 0.2116, "step": 16043 }, { "epoch": 4.269292176689729, "grad_norm": 0.30929768085479736, "learning_rate": 1.0782686589451506e-08, "loss": 0.1993, "step": 16044 }, { "epoch": 4.269558275678552, "grad_norm": 0.30066266655921936, "learning_rate": 1.0774981592382048e-08, "loss": 0.1797, "step": 16045 }, { "epoch": 4.269824374667376, "grad_norm": 0.33746233582496643, "learning_rate": 1.0767279192421208e-08, "loss": 0.217, "step": 16046 }, { "epoch": 4.2700904736562, "grad_norm": 0.3021667003631592, "learning_rate": 1.0759579389793183e-08, "loss": 0.1852, "step": 16047 }, { "epoch": 4.270356572645024, "grad_norm": 0.31846925616264343, "learning_rate": 1.0751882184722083e-08, "loss": 0.1914, "step": 16048 }, { "epoch": 4.270622671633848, "grad_norm": 0.2959347367286682, "learning_rate": 1.0744187577431984e-08, "loss": 0.2007, "step": 16049 }, { "epoch": 4.270888770622672, "grad_norm": 0.286724716424942, "learning_rate": 1.073649556814682e-08, "loss": 0.1779, "step": 16050 }, { "epoch": 4.271154869611496, "grad_norm": 0.2934783697128296, "learning_rate": 1.0728806157090498e-08, "loss": 0.1933, "step": 16051 }, { "epoch": 4.271420968600319, "grad_norm": 0.30994153022766113, "learning_rate": 1.072111934448684e-08, "loss": 0.1873, "step": 16052 }, { "epoch": 4.271687067589143, "grad_norm": 0.28002995252609253, "learning_rate": 1.0713435130559578e-08, "loss": 0.1792, "step": 16053 }, { "epoch": 4.271953166577967, "grad_norm": 0.30179861187934875, "learning_rate": 1.0705753515532401e-08, "loss": 0.1733, "step": 16054 }, { "epoch": 4.2722192655667905, "grad_norm": 0.3371522128582001, "learning_rate": 1.0698074499628884e-08, "loss": 0.1812, "step": 16055 }, { "epoch": 4.272485364555615, "grad_norm": 0.30850255489349365, "learning_rate": 1.0690398083072549e-08, "loss": 0.191, "step": 16056 }, { "epoch": 4.272751463544439, "grad_norm": 0.3017881214618683, "learning_rate": 1.068272426608685e-08, "loss": 0.177, "step": 16057 }, { "epoch": 4.273017562533262, "grad_norm": 0.30416756868362427, "learning_rate": 1.0675053048895122e-08, "loss": 0.1842, "step": 16058 }, { "epoch": 4.273283661522086, "grad_norm": 0.31137755513191223, "learning_rate": 1.0667384431720695e-08, "loss": 0.1998, "step": 16059 }, { "epoch": 4.27354976051091, "grad_norm": 0.286668986082077, "learning_rate": 1.0659718414786734e-08, "loss": 0.1787, "step": 16060 }, { "epoch": 4.2738158594997335, "grad_norm": 0.2803723216056824, "learning_rate": 1.0652054998316396e-08, "loss": 0.1924, "step": 16061 }, { "epoch": 4.274081958488558, "grad_norm": 0.3506108522415161, "learning_rate": 1.0644394182532757e-08, "loss": 0.2175, "step": 16062 }, { "epoch": 4.274348057477382, "grad_norm": 0.30174797773361206, "learning_rate": 1.0636735967658783e-08, "loss": 0.1954, "step": 16063 }, { "epoch": 4.274614156466205, "grad_norm": 0.2923990786075592, "learning_rate": 1.0629080353917397e-08, "loss": 0.1833, "step": 16064 }, { "epoch": 4.274880255455029, "grad_norm": 0.2946988344192505, "learning_rate": 1.0621427341531442e-08, "loss": 0.1935, "step": 16065 }, { "epoch": 4.275146354443853, "grad_norm": 0.28835341334342957, "learning_rate": 1.0613776930723662e-08, "loss": 0.1722, "step": 16066 }, { "epoch": 4.275412453432677, "grad_norm": 0.3275534212589264, "learning_rate": 1.0606129121716767e-08, "loss": 0.206, "step": 16067 }, { "epoch": 4.275678552421501, "grad_norm": 0.30107271671295166, "learning_rate": 1.0598483914733315e-08, "loss": 0.1984, "step": 16068 }, { "epoch": 4.275944651410325, "grad_norm": 0.302753210067749, "learning_rate": 1.059084130999588e-08, "loss": 0.1865, "step": 16069 }, { "epoch": 4.276210750399149, "grad_norm": 0.31696826219558716, "learning_rate": 1.0583201307726897e-08, "loss": 0.1817, "step": 16070 }, { "epoch": 4.276476849387972, "grad_norm": 0.2954995632171631, "learning_rate": 1.0575563908148766e-08, "loss": 0.1882, "step": 16071 }, { "epoch": 4.276742948376796, "grad_norm": 0.3314721882343292, "learning_rate": 1.0567929111483776e-08, "loss": 0.2035, "step": 16072 }, { "epoch": 4.27700904736562, "grad_norm": 0.31012052297592163, "learning_rate": 1.0560296917954159e-08, "loss": 0.2066, "step": 16073 }, { "epoch": 4.277275146354444, "grad_norm": 0.29804855585098267, "learning_rate": 1.0552667327782072e-08, "loss": 0.2087, "step": 16074 }, { "epoch": 4.277541245343268, "grad_norm": 0.30138441920280457, "learning_rate": 1.0545040341189593e-08, "loss": 0.1929, "step": 16075 }, { "epoch": 4.277807344332092, "grad_norm": 0.3153237998485565, "learning_rate": 1.0537415958398743e-08, "loss": 0.1906, "step": 16076 }, { "epoch": 4.278073443320915, "grad_norm": 0.2954239547252655, "learning_rate": 1.0529794179631413e-08, "loss": 0.177, "step": 16077 }, { "epoch": 4.278339542309739, "grad_norm": 0.30209535360336304, "learning_rate": 1.0522175005109458e-08, "loss": 0.1871, "step": 16078 }, { "epoch": 4.278605641298563, "grad_norm": 0.29089221358299255, "learning_rate": 1.0514558435054677e-08, "loss": 0.1968, "step": 16079 }, { "epoch": 4.278871740287387, "grad_norm": 0.336489737033844, "learning_rate": 1.050694446968875e-08, "loss": 0.2153, "step": 16080 }, { "epoch": 4.279137839276211, "grad_norm": 0.3119986057281494, "learning_rate": 1.0499333109233288e-08, "loss": 0.2122, "step": 16081 }, { "epoch": 4.279403938265035, "grad_norm": 0.27564242482185364, "learning_rate": 1.0491724353909859e-08, "loss": 0.1832, "step": 16082 }, { "epoch": 4.279670037253858, "grad_norm": 0.31060028076171875, "learning_rate": 1.0484118203939928e-08, "loss": 0.2064, "step": 16083 }, { "epoch": 4.279936136242682, "grad_norm": 0.3237895965576172, "learning_rate": 1.0476514659544888e-08, "loss": 0.1831, "step": 16084 }, { "epoch": 4.280202235231506, "grad_norm": 0.2902282774448395, "learning_rate": 1.0468913720946082e-08, "loss": 0.1828, "step": 16085 }, { "epoch": 4.28046833422033, "grad_norm": 0.3055419921875, "learning_rate": 1.0461315388364722e-08, "loss": 0.1789, "step": 16086 }, { "epoch": 4.280734433209154, "grad_norm": 0.30130475759506226, "learning_rate": 1.0453719662021976e-08, "loss": 0.1884, "step": 16087 }, { "epoch": 4.281000532197978, "grad_norm": 0.30246496200561523, "learning_rate": 1.0446126542138945e-08, "loss": 0.1987, "step": 16088 }, { "epoch": 4.281266631186801, "grad_norm": 0.2705972492694855, "learning_rate": 1.0438536028936652e-08, "loss": 0.1832, "step": 16089 }, { "epoch": 4.281532730175625, "grad_norm": 0.3141982853412628, "learning_rate": 1.043094812263603e-08, "loss": 0.1856, "step": 16090 }, { "epoch": 4.281798829164449, "grad_norm": 0.293525755405426, "learning_rate": 1.0423362823457937e-08, "loss": 0.1806, "step": 16091 }, { "epoch": 4.282064928153273, "grad_norm": 0.32046887278556824, "learning_rate": 1.0415780131623176e-08, "loss": 0.2264, "step": 16092 }, { "epoch": 4.282331027142097, "grad_norm": 0.31760045886039734, "learning_rate": 1.0408200047352456e-08, "loss": 0.184, "step": 16093 }, { "epoch": 4.282597126130921, "grad_norm": 0.28548663854599, "learning_rate": 1.0400622570866425e-08, "loss": 0.1813, "step": 16094 }, { "epoch": 4.282863225119745, "grad_norm": 0.2985931932926178, "learning_rate": 1.0393047702385605e-08, "loss": 0.1949, "step": 16095 }, { "epoch": 4.283129324108568, "grad_norm": 0.3251204192638397, "learning_rate": 1.0385475442130508e-08, "loss": 0.207, "step": 16096 }, { "epoch": 4.283395423097392, "grad_norm": 0.2706426680088043, "learning_rate": 1.037790579032154e-08, "loss": 0.1637, "step": 16097 }, { "epoch": 4.283661522086216, "grad_norm": 0.2807133197784424, "learning_rate": 1.037033874717903e-08, "loss": 0.1871, "step": 16098 }, { "epoch": 4.28392762107504, "grad_norm": 0.2947465479373932, "learning_rate": 1.0362774312923262e-08, "loss": 0.195, "step": 16099 }, { "epoch": 4.284193720063864, "grad_norm": 0.2966321110725403, "learning_rate": 1.0355212487774367e-08, "loss": 0.1955, "step": 16100 }, { "epoch": 4.284459819052688, "grad_norm": 0.3065868318080902, "learning_rate": 1.0347653271952484e-08, "loss": 0.1849, "step": 16101 }, { "epoch": 4.284725918041511, "grad_norm": 0.296851247549057, "learning_rate": 1.0340096665677622e-08, "loss": 0.1888, "step": 16102 }, { "epoch": 4.284992017030335, "grad_norm": 0.2987528443336487, "learning_rate": 1.0332542669169775e-08, "loss": 0.1749, "step": 16103 }, { "epoch": 4.285258116019159, "grad_norm": 0.30717042088508606, "learning_rate": 1.0324991282648777e-08, "loss": 0.1952, "step": 16104 }, { "epoch": 4.285524215007983, "grad_norm": 0.2880634367465973, "learning_rate": 1.031744250633444e-08, "loss": 0.1921, "step": 16105 }, { "epoch": 4.285790313996807, "grad_norm": 0.3020859658718109, "learning_rate": 1.030989634044649e-08, "loss": 0.1907, "step": 16106 }, { "epoch": 4.286056412985631, "grad_norm": 0.27855560183525085, "learning_rate": 1.0302352785204583e-08, "loss": 0.1847, "step": 16107 }, { "epoch": 4.286322511974454, "grad_norm": 0.29089441895484924, "learning_rate": 1.0294811840828277e-08, "loss": 0.194, "step": 16108 }, { "epoch": 4.286588610963278, "grad_norm": 0.2702421545982361, "learning_rate": 1.0287273507537097e-08, "loss": 0.1703, "step": 16109 }, { "epoch": 4.286854709952102, "grad_norm": 0.30178508162498474, "learning_rate": 1.0279737785550436e-08, "loss": 0.2066, "step": 16110 }, { "epoch": 4.287120808940926, "grad_norm": 0.29280319809913635, "learning_rate": 1.027220467508766e-08, "loss": 0.1929, "step": 16111 }, { "epoch": 4.28738690792975, "grad_norm": 0.2884262204170227, "learning_rate": 1.0264674176368027e-08, "loss": 0.1843, "step": 16112 }, { "epoch": 4.287653006918574, "grad_norm": 0.3021067976951599, "learning_rate": 1.0257146289610751e-08, "loss": 0.1943, "step": 16113 }, { "epoch": 4.287919105907397, "grad_norm": 0.2897292673587799, "learning_rate": 1.0249621015034927e-08, "loss": 0.1958, "step": 16114 }, { "epoch": 4.288185204896221, "grad_norm": 0.2887483835220337, "learning_rate": 1.0242098352859586e-08, "loss": 0.1877, "step": 16115 }, { "epoch": 4.288451303885045, "grad_norm": 0.2879716455936432, "learning_rate": 1.023457830330371e-08, "loss": 0.1892, "step": 16116 }, { "epoch": 4.2887174028738695, "grad_norm": 0.3185684382915497, "learning_rate": 1.0227060866586179e-08, "loss": 0.198, "step": 16117 }, { "epoch": 4.288983501862693, "grad_norm": 0.2932780086994171, "learning_rate": 1.0219546042925842e-08, "loss": 0.1781, "step": 16118 }, { "epoch": 4.289249600851517, "grad_norm": 0.3106536567211151, "learning_rate": 1.021203383254139e-08, "loss": 0.1938, "step": 16119 }, { "epoch": 4.289515699840341, "grad_norm": 0.2972419559955597, "learning_rate": 1.0204524235651491e-08, "loss": 0.1877, "step": 16120 }, { "epoch": 4.289781798829164, "grad_norm": 0.2712860703468323, "learning_rate": 1.0197017252474749e-08, "loss": 0.1759, "step": 16121 }, { "epoch": 4.290047897817988, "grad_norm": 0.27952179312705994, "learning_rate": 1.0189512883229678e-08, "loss": 0.1755, "step": 16122 }, { "epoch": 4.2903139968068125, "grad_norm": 0.3454585671424866, "learning_rate": 1.0182011128134682e-08, "loss": 0.201, "step": 16123 }, { "epoch": 4.290580095795636, "grad_norm": 0.3038947880268097, "learning_rate": 1.0174511987408129e-08, "loss": 0.1932, "step": 16124 }, { "epoch": 4.29084619478446, "grad_norm": 0.3739728033542633, "learning_rate": 1.0167015461268303e-08, "loss": 0.1757, "step": 16125 }, { "epoch": 4.291112293773284, "grad_norm": 0.2850547730922699, "learning_rate": 1.0159521549933404e-08, "loss": 0.1942, "step": 16126 }, { "epoch": 4.291378392762107, "grad_norm": 0.2963232398033142, "learning_rate": 1.0152030253621568e-08, "loss": 0.2076, "step": 16127 }, { "epoch": 4.291644491750931, "grad_norm": 0.27730002999305725, "learning_rate": 1.0144541572550847e-08, "loss": 0.1894, "step": 16128 }, { "epoch": 4.2919105907397554, "grad_norm": 0.3196701109409332, "learning_rate": 1.0137055506939208e-08, "loss": 0.2071, "step": 16129 }, { "epoch": 4.292176689728579, "grad_norm": 0.3241759240627289, "learning_rate": 1.0129572057004554e-08, "loss": 0.1901, "step": 16130 }, { "epoch": 4.292442788717403, "grad_norm": 0.29224804043769836, "learning_rate": 1.0122091222964735e-08, "loss": 0.1878, "step": 16131 }, { "epoch": 4.292708887706227, "grad_norm": 0.3273252844810486, "learning_rate": 1.0114613005037453e-08, "loss": 0.2124, "step": 16132 }, { "epoch": 4.29297498669505, "grad_norm": 0.2952656149864197, "learning_rate": 1.0107137403440403e-08, "loss": 0.1929, "step": 16133 }, { "epoch": 4.293241085683874, "grad_norm": 0.2942025661468506, "learning_rate": 1.0099664418391185e-08, "loss": 0.1875, "step": 16134 }, { "epoch": 4.293507184672698, "grad_norm": 0.26398739218711853, "learning_rate": 1.0092194050107305e-08, "loss": 0.1615, "step": 16135 }, { "epoch": 4.2937732836615226, "grad_norm": 0.29766613245010376, "learning_rate": 1.0084726298806212e-08, "loss": 0.1748, "step": 16136 }, { "epoch": 4.294039382650346, "grad_norm": 0.31533893942832947, "learning_rate": 1.0077261164705276e-08, "loss": 0.179, "step": 16137 }, { "epoch": 4.29430548163917, "grad_norm": 0.30185914039611816, "learning_rate": 1.0069798648021799e-08, "loss": 0.1843, "step": 16138 }, { "epoch": 4.294571580627993, "grad_norm": 0.2997833788394928, "learning_rate": 1.0062338748972977e-08, "loss": 0.2127, "step": 16139 }, { "epoch": 4.294837679616817, "grad_norm": 0.307654470205307, "learning_rate": 1.0054881467775943e-08, "loss": 0.1956, "step": 16140 }, { "epoch": 4.295103778605641, "grad_norm": 0.3115728795528412, "learning_rate": 1.0047426804647785e-08, "loss": 0.1985, "step": 16141 }, { "epoch": 4.2953698775944655, "grad_norm": 0.2925126850605011, "learning_rate": 1.0039974759805459e-08, "loss": 0.2125, "step": 16142 }, { "epoch": 4.295635976583289, "grad_norm": 0.294901043176651, "learning_rate": 1.0032525333465892e-08, "loss": 0.177, "step": 16143 }, { "epoch": 4.295902075572113, "grad_norm": 0.29676541686058044, "learning_rate": 1.0025078525845909e-08, "loss": 0.188, "step": 16144 }, { "epoch": 4.296168174560937, "grad_norm": 0.298700749874115, "learning_rate": 1.0017634337162274e-08, "loss": 0.204, "step": 16145 }, { "epoch": 4.29643427354976, "grad_norm": 0.30840009450912476, "learning_rate": 1.0010192767631665e-08, "loss": 0.1952, "step": 16146 }, { "epoch": 4.296700372538584, "grad_norm": 0.31403008103370667, "learning_rate": 1.0002753817470688e-08, "loss": 0.2036, "step": 16147 }, { "epoch": 4.2969664715274085, "grad_norm": 0.27879980206489563, "learning_rate": 9.995317486895882e-09, "loss": 0.1683, "step": 16148 }, { "epoch": 4.297232570516232, "grad_norm": 0.29707643389701843, "learning_rate": 9.987883776123696e-09, "loss": 0.1952, "step": 16149 }, { "epoch": 4.297498669505056, "grad_norm": 0.2874751389026642, "learning_rate": 9.980452685370488e-09, "loss": 0.1778, "step": 16150 }, { "epoch": 4.29776476849388, "grad_norm": 0.2808016836643219, "learning_rate": 9.973024214852566e-09, "loss": 0.1839, "step": 16151 }, { "epoch": 4.298030867482703, "grad_norm": 0.304453045129776, "learning_rate": 9.965598364786154e-09, "loss": 0.1741, "step": 16152 }, { "epoch": 4.298296966471527, "grad_norm": 0.2762756049633026, "learning_rate": 9.958175135387404e-09, "loss": 0.1751, "step": 16153 }, { "epoch": 4.2985630654603515, "grad_norm": 0.2652944326400757, "learning_rate": 9.950754526872385e-09, "loss": 0.1729, "step": 16154 }, { "epoch": 4.298829164449175, "grad_norm": 0.3027874231338501, "learning_rate": 9.943336539457092e-09, "loss": 0.1945, "step": 16155 }, { "epoch": 4.299095263437999, "grad_norm": 0.30944719910621643, "learning_rate": 9.935921173357442e-09, "loss": 0.1759, "step": 16156 }, { "epoch": 4.299361362426823, "grad_norm": 0.31802991032600403, "learning_rate": 9.928508428789295e-09, "loss": 0.2028, "step": 16157 }, { "epoch": 4.299627461415646, "grad_norm": 0.32236313819885254, "learning_rate": 9.921098305968378e-09, "loss": 0.2082, "step": 16158 }, { "epoch": 4.29989356040447, "grad_norm": 0.2972993850708008, "learning_rate": 9.913690805110409e-09, "loss": 0.172, "step": 16159 }, { "epoch": 4.3001596593932945, "grad_norm": 0.3177741765975952, "learning_rate": 9.906285926431001e-09, "loss": 0.2194, "step": 16160 }, { "epoch": 4.300425758382119, "grad_norm": 0.29042303562164307, "learning_rate": 9.898883670145674e-09, "loss": 0.1871, "step": 16161 }, { "epoch": 4.300691857370942, "grad_norm": 0.3302144408226013, "learning_rate": 9.8914840364699e-09, "loss": 0.1909, "step": 16162 }, { "epoch": 4.300957956359766, "grad_norm": 0.29169392585754395, "learning_rate": 9.88408702561906e-09, "loss": 0.1833, "step": 16163 }, { "epoch": 4.301224055348589, "grad_norm": 0.32871684432029724, "learning_rate": 9.87669263780846e-09, "loss": 0.2077, "step": 16164 }, { "epoch": 4.301490154337413, "grad_norm": 0.2773464024066925, "learning_rate": 9.869300873253327e-09, "loss": 0.1742, "step": 16165 }, { "epoch": 4.3017562533262375, "grad_norm": 0.29405516386032104, "learning_rate": 9.861911732168837e-09, "loss": 0.1955, "step": 16166 }, { "epoch": 4.302022352315062, "grad_norm": 0.302795946598053, "learning_rate": 9.854525214770071e-09, "loss": 0.1995, "step": 16167 }, { "epoch": 4.302288451303885, "grad_norm": 0.28762131929397583, "learning_rate": 9.847141321271991e-09, "loss": 0.1958, "step": 16168 }, { "epoch": 4.302554550292709, "grad_norm": 0.30418360233306885, "learning_rate": 9.839760051889556e-09, "loss": 0.2106, "step": 16169 }, { "epoch": 4.302820649281533, "grad_norm": 0.3116316795349121, "learning_rate": 9.832381406837598e-09, "loss": 0.1942, "step": 16170 }, { "epoch": 4.303086748270356, "grad_norm": 0.28781870007514954, "learning_rate": 9.825005386330909e-09, "loss": 0.1659, "step": 16171 }, { "epoch": 4.30335284725918, "grad_norm": 0.29659390449523926, "learning_rate": 9.817631990584163e-09, "loss": 0.1957, "step": 16172 }, { "epoch": 4.303618946248005, "grad_norm": 0.30546456575393677, "learning_rate": 9.81026121981201e-09, "loss": 0.1984, "step": 16173 }, { "epoch": 4.303885045236828, "grad_norm": 0.27672696113586426, "learning_rate": 9.802893074228969e-09, "loss": 0.1724, "step": 16174 }, { "epoch": 4.304151144225652, "grad_norm": 0.3179207146167755, "learning_rate": 9.79552755404951e-09, "loss": 0.1908, "step": 16175 }, { "epoch": 4.304417243214476, "grad_norm": 0.29016613960266113, "learning_rate": 9.788164659488063e-09, "loss": 0.1963, "step": 16176 }, { "epoch": 4.304683342203299, "grad_norm": 0.31884944438934326, "learning_rate": 9.78080439075888e-09, "loss": 0.1867, "step": 16177 }, { "epoch": 4.304949441192123, "grad_norm": 0.31982406973838806, "learning_rate": 9.773446748076231e-09, "loss": 0.2006, "step": 16178 }, { "epoch": 4.3052155401809475, "grad_norm": 0.30843740701675415, "learning_rate": 9.76609173165428e-09, "loss": 0.1886, "step": 16179 }, { "epoch": 4.305481639169771, "grad_norm": 0.3142317831516266, "learning_rate": 9.758739341707123e-09, "loss": 0.1848, "step": 16180 }, { "epoch": 4.305747738158595, "grad_norm": 0.29100555181503296, "learning_rate": 9.751389578448743e-09, "loss": 0.1871, "step": 16181 }, { "epoch": 4.306013837147419, "grad_norm": 0.31485944986343384, "learning_rate": 9.74404244209308e-09, "loss": 0.1928, "step": 16182 }, { "epoch": 4.306279936136242, "grad_norm": 0.28875255584716797, "learning_rate": 9.736697932853999e-09, "loss": 0.1771, "step": 16183 }, { "epoch": 4.306546035125066, "grad_norm": 0.28588297963142395, "learning_rate": 9.72935605094527e-09, "loss": 0.1778, "step": 16184 }, { "epoch": 4.3068121341138905, "grad_norm": 0.27106717228889465, "learning_rate": 9.722016796580612e-09, "loss": 0.1813, "step": 16185 }, { "epoch": 4.307078233102715, "grad_norm": 0.2862781584262848, "learning_rate": 9.714680169973632e-09, "loss": 0.1753, "step": 16186 }, { "epoch": 4.307344332091538, "grad_norm": 0.2884294092655182, "learning_rate": 9.707346171337894e-09, "loss": 0.1769, "step": 16187 }, { "epoch": 4.307610431080362, "grad_norm": 0.3089805543422699, "learning_rate": 9.700014800886857e-09, "loss": 0.182, "step": 16188 }, { "epoch": 4.307876530069186, "grad_norm": 0.28252434730529785, "learning_rate": 9.692686058833933e-09, "loss": 0.1807, "step": 16189 }, { "epoch": 4.308142629058009, "grad_norm": 0.3466586172580719, "learning_rate": 9.685359945392424e-09, "loss": 0.2112, "step": 16190 }, { "epoch": 4.3084087280468335, "grad_norm": 0.3000757694244385, "learning_rate": 9.678036460775596e-09, "loss": 0.1942, "step": 16191 }, { "epoch": 4.308674827035658, "grad_norm": 0.27627116441726685, "learning_rate": 9.670715605196611e-09, "loss": 0.1879, "step": 16192 }, { "epoch": 4.308940926024481, "grad_norm": 0.3136330544948578, "learning_rate": 9.663397378868554e-09, "loss": 0.1823, "step": 16193 }, { "epoch": 4.309207025013305, "grad_norm": 0.2708417773246765, "learning_rate": 9.656081782004466e-09, "loss": 0.174, "step": 16194 }, { "epoch": 4.309473124002129, "grad_norm": 0.31370264291763306, "learning_rate": 9.648768814817243e-09, "loss": 0.2026, "step": 16195 }, { "epoch": 4.309739222990952, "grad_norm": 0.29398682713508606, "learning_rate": 9.64145847751976e-09, "loss": 0.1896, "step": 16196 }, { "epoch": 4.3100053219797765, "grad_norm": 0.2938845157623291, "learning_rate": 9.634150770324812e-09, "loss": 0.1788, "step": 16197 }, { "epoch": 4.310271420968601, "grad_norm": 0.2799726724624634, "learning_rate": 9.626845693445096e-09, "loss": 0.1848, "step": 16198 }, { "epoch": 4.310537519957424, "grad_norm": 0.27762725949287415, "learning_rate": 9.619543247093253e-09, "loss": 0.1837, "step": 16199 }, { "epoch": 4.310803618946248, "grad_norm": 0.2928820252418518, "learning_rate": 9.612243431481859e-09, "loss": 0.1804, "step": 16200 }, { "epoch": 4.311069717935072, "grad_norm": 0.2728865146636963, "learning_rate": 9.604946246823343e-09, "loss": 0.1757, "step": 16201 }, { "epoch": 4.311335816923895, "grad_norm": 0.2997020483016968, "learning_rate": 9.597651693330144e-09, "loss": 0.1928, "step": 16202 }, { "epoch": 4.3116019159127195, "grad_norm": 0.30674028396606445, "learning_rate": 9.590359771214585e-09, "loss": 0.1998, "step": 16203 }, { "epoch": 4.311868014901544, "grad_norm": 0.29672279953956604, "learning_rate": 9.583070480688893e-09, "loss": 0.1845, "step": 16204 }, { "epoch": 4.312134113890367, "grad_norm": 0.3047477900981903, "learning_rate": 9.575783821965255e-09, "loss": 0.1901, "step": 16205 }, { "epoch": 4.312400212879191, "grad_norm": 0.3301297426223755, "learning_rate": 9.568499795255757e-09, "loss": 0.2, "step": 16206 }, { "epoch": 4.312666311868015, "grad_norm": 0.3075311481952667, "learning_rate": 9.561218400772442e-09, "loss": 0.2047, "step": 16207 }, { "epoch": 4.312932410856838, "grad_norm": 0.30321547389030457, "learning_rate": 9.553939638727226e-09, "loss": 0.1841, "step": 16208 }, { "epoch": 4.313198509845662, "grad_norm": 0.29972589015960693, "learning_rate": 9.546663509331987e-09, "loss": 0.1853, "step": 16209 }, { "epoch": 4.313464608834487, "grad_norm": 0.29168471693992615, "learning_rate": 9.539390012798521e-09, "loss": 0.1993, "step": 16210 }, { "epoch": 4.313730707823311, "grad_norm": 0.3124208450317383, "learning_rate": 9.532119149338525e-09, "loss": 0.1908, "step": 16211 }, { "epoch": 4.313996806812134, "grad_norm": 0.29690447449684143, "learning_rate": 9.524850919163664e-09, "loss": 0.2005, "step": 16212 }, { "epoch": 4.314262905800958, "grad_norm": 0.3012607991695404, "learning_rate": 9.517585322485455e-09, "loss": 0.1932, "step": 16213 }, { "epoch": 4.314529004789782, "grad_norm": 0.28423449397087097, "learning_rate": 9.51032235951541e-09, "loss": 0.1826, "step": 16214 }, { "epoch": 4.314795103778605, "grad_norm": 0.30906519293785095, "learning_rate": 9.503062030464925e-09, "loss": 0.2142, "step": 16215 }, { "epoch": 4.3150612027674295, "grad_norm": 0.32094141840934753, "learning_rate": 9.495804335545332e-09, "loss": 0.2109, "step": 16216 }, { "epoch": 4.315327301756254, "grad_norm": 0.33011555671691895, "learning_rate": 9.488549274967872e-09, "loss": 0.1783, "step": 16217 }, { "epoch": 4.315593400745077, "grad_norm": 0.3305089771747589, "learning_rate": 9.481296848943743e-09, "loss": 0.1977, "step": 16218 }, { "epoch": 4.315859499733901, "grad_norm": 0.2970682978630066, "learning_rate": 9.474047057684053e-09, "loss": 0.1796, "step": 16219 }, { "epoch": 4.316125598722725, "grad_norm": 0.27837809920310974, "learning_rate": 9.46679990139978e-09, "loss": 0.1698, "step": 16220 }, { "epoch": 4.316391697711548, "grad_norm": 0.45083701610565186, "learning_rate": 9.459555380301899e-09, "loss": 0.1808, "step": 16221 }, { "epoch": 4.3166577967003725, "grad_norm": 0.29116231203079224, "learning_rate": 9.452313494601293e-09, "loss": 0.2038, "step": 16222 }, { "epoch": 4.316923895689197, "grad_norm": 0.31697046756744385, "learning_rate": 9.445074244508711e-09, "loss": 0.1886, "step": 16223 }, { "epoch": 4.31718999467802, "grad_norm": 0.3545614182949066, "learning_rate": 9.437837630234902e-09, "loss": 0.1989, "step": 16224 }, { "epoch": 4.317456093666844, "grad_norm": 0.34541213512420654, "learning_rate": 9.430603651990498e-09, "loss": 0.1925, "step": 16225 }, { "epoch": 4.317722192655668, "grad_norm": 0.2957431375980377, "learning_rate": 9.423372309986056e-09, "loss": 0.1826, "step": 16226 }, { "epoch": 4.317988291644491, "grad_norm": 0.31752872467041016, "learning_rate": 9.41614360443207e-09, "loss": 0.1882, "step": 16227 }, { "epoch": 4.3182543906333155, "grad_norm": 0.2912338674068451, "learning_rate": 9.40891753553894e-09, "loss": 0.1794, "step": 16228 }, { "epoch": 4.31852048962214, "grad_norm": 0.2949257493019104, "learning_rate": 9.401694103516999e-09, "loss": 0.179, "step": 16229 }, { "epoch": 4.318786588610963, "grad_norm": 0.26390311121940613, "learning_rate": 9.394473308576534e-09, "loss": 0.1709, "step": 16230 }, { "epoch": 4.319052687599787, "grad_norm": 0.3076511323451996, "learning_rate": 9.387255150927676e-09, "loss": 0.2018, "step": 16231 }, { "epoch": 4.319318786588611, "grad_norm": 0.31508275866508484, "learning_rate": 9.380039630780545e-09, "loss": 0.1792, "step": 16232 }, { "epoch": 4.319584885577434, "grad_norm": 0.3076978027820587, "learning_rate": 9.372826748345175e-09, "loss": 0.2014, "step": 16233 }, { "epoch": 4.3198509845662585, "grad_norm": 0.35930225253105164, "learning_rate": 9.365616503831497e-09, "loss": 0.2072, "step": 16234 }, { "epoch": 4.320117083555083, "grad_norm": 0.29336798191070557, "learning_rate": 9.358408897449398e-09, "loss": 0.1853, "step": 16235 }, { "epoch": 4.320383182543907, "grad_norm": 0.31171396374702454, "learning_rate": 9.351203929408669e-09, "loss": 0.1866, "step": 16236 }, { "epoch": 4.32064928153273, "grad_norm": 0.28009727597236633, "learning_rate": 9.344001599919027e-09, "loss": 0.1833, "step": 16237 }, { "epoch": 4.320915380521554, "grad_norm": 0.2921917736530304, "learning_rate": 9.336801909190107e-09, "loss": 0.1933, "step": 16238 }, { "epoch": 4.321181479510378, "grad_norm": 0.328044056892395, "learning_rate": 9.329604857431505e-09, "loss": 0.2058, "step": 16239 }, { "epoch": 4.3214475784992015, "grad_norm": 0.2846648097038269, "learning_rate": 9.322410444852657e-09, "loss": 0.1838, "step": 16240 }, { "epoch": 4.321713677488026, "grad_norm": 0.28932809829711914, "learning_rate": 9.315218671663005e-09, "loss": 0.1812, "step": 16241 }, { "epoch": 4.32197977647685, "grad_norm": 0.2641311585903168, "learning_rate": 9.308029538071882e-09, "loss": 0.1579, "step": 16242 }, { "epoch": 4.322245875465673, "grad_norm": 0.3072119355201721, "learning_rate": 9.300843044288532e-09, "loss": 0.1831, "step": 16243 }, { "epoch": 4.322511974454497, "grad_norm": 0.3021095097064972, "learning_rate": 9.29365919052214e-09, "loss": 0.1965, "step": 16244 }, { "epoch": 4.322778073443321, "grad_norm": 0.30103442072868347, "learning_rate": 9.286477976981822e-09, "loss": 0.2003, "step": 16245 }, { "epoch": 4.3230441724321444, "grad_norm": 0.3117576539516449, "learning_rate": 9.279299403876583e-09, "loss": 0.1853, "step": 16246 }, { "epoch": 4.323310271420969, "grad_norm": 0.29085367918014526, "learning_rate": 9.272123471415383e-09, "loss": 0.1937, "step": 16247 }, { "epoch": 4.323576370409793, "grad_norm": 0.28446561098098755, "learning_rate": 9.26495017980713e-09, "loss": 0.1654, "step": 16248 }, { "epoch": 4.323842469398616, "grad_norm": 0.29121658205986023, "learning_rate": 9.257779529260556e-09, "loss": 0.1842, "step": 16249 }, { "epoch": 4.32410856838744, "grad_norm": 0.2754536271095276, "learning_rate": 9.250611519984409e-09, "loss": 0.1825, "step": 16250 }, { "epoch": 4.324374667376264, "grad_norm": 0.291769415140152, "learning_rate": 9.243446152187328e-09, "loss": 0.1979, "step": 16251 }, { "epoch": 4.324640766365087, "grad_norm": 0.29453936219215393, "learning_rate": 9.236283426077895e-09, "loss": 0.19, "step": 16252 }, { "epoch": 4.3249068653539116, "grad_norm": 0.3077327311038971, "learning_rate": 9.229123341864576e-09, "loss": 0.2058, "step": 16253 }, { "epoch": 4.325172964342736, "grad_norm": 0.3172057867050171, "learning_rate": 9.221965899755802e-09, "loss": 0.2146, "step": 16254 }, { "epoch": 4.32543906333156, "grad_norm": 0.35009655356407166, "learning_rate": 9.214811099959896e-09, "loss": 0.1782, "step": 16255 }, { "epoch": 4.325705162320383, "grad_norm": 0.3006059229373932, "learning_rate": 9.207658942685103e-09, "loss": 0.1915, "step": 16256 }, { "epoch": 4.325971261309207, "grad_norm": 0.2875864803791046, "learning_rate": 9.200509428139657e-09, "loss": 0.1937, "step": 16257 }, { "epoch": 4.32623736029803, "grad_norm": 0.31167104840278625, "learning_rate": 9.193362556531592e-09, "loss": 0.1967, "step": 16258 }, { "epoch": 4.3265034592868545, "grad_norm": 0.2941826581954956, "learning_rate": 9.186218328068973e-09, "loss": 0.2015, "step": 16259 }, { "epoch": 4.326769558275679, "grad_norm": 0.2968471944332123, "learning_rate": 9.179076742959735e-09, "loss": 0.1975, "step": 16260 }, { "epoch": 4.327035657264503, "grad_norm": 0.29602447152137756, "learning_rate": 9.171937801411778e-09, "loss": 0.1932, "step": 16261 }, { "epoch": 4.327301756253326, "grad_norm": 0.3049710690975189, "learning_rate": 9.164801503632869e-09, "loss": 0.1875, "step": 16262 }, { "epoch": 4.32756785524215, "grad_norm": 0.29969021677970886, "learning_rate": 9.15766784983073e-09, "loss": 0.1903, "step": 16263 }, { "epoch": 4.327833954230974, "grad_norm": 0.2926918864250183, "learning_rate": 9.150536840213008e-09, "loss": 0.1837, "step": 16264 }, { "epoch": 4.3281000532197975, "grad_norm": 0.3005487620830536, "learning_rate": 9.14340847498728e-09, "loss": 0.1914, "step": 16265 }, { "epoch": 4.328366152208622, "grad_norm": 0.319676011800766, "learning_rate": 9.136282754361035e-09, "loss": 0.1934, "step": 16266 }, { "epoch": 4.328632251197446, "grad_norm": 0.30186766386032104, "learning_rate": 9.129159678541654e-09, "loss": 0.1746, "step": 16267 }, { "epoch": 4.328898350186269, "grad_norm": 0.2987520694732666, "learning_rate": 9.12203924773649e-09, "loss": 0.1839, "step": 16268 }, { "epoch": 4.329164449175093, "grad_norm": 0.2878348231315613, "learning_rate": 9.114921462152813e-09, "loss": 0.1839, "step": 16269 }, { "epoch": 4.329430548163917, "grad_norm": 0.2999449670314789, "learning_rate": 9.107806321997779e-09, "loss": 0.2052, "step": 16270 }, { "epoch": 4.3296966471527405, "grad_norm": 0.2868591248989105, "learning_rate": 9.100693827478511e-09, "loss": 0.1799, "step": 16271 }, { "epoch": 4.329962746141565, "grad_norm": 0.2979740500450134, "learning_rate": 9.093583978802021e-09, "loss": 0.186, "step": 16272 }, { "epoch": 4.330228845130389, "grad_norm": 0.32040175795555115, "learning_rate": 9.086476776175279e-09, "loss": 0.2046, "step": 16273 }, { "epoch": 4.330494944119212, "grad_norm": 0.3004835844039917, "learning_rate": 9.079372219805125e-09, "loss": 0.1924, "step": 16274 }, { "epoch": 4.330761043108036, "grad_norm": 0.3106862008571625, "learning_rate": 9.07227030989841e-09, "loss": 0.1942, "step": 16275 }, { "epoch": 4.33102714209686, "grad_norm": 0.30510878562927246, "learning_rate": 9.06517104666178e-09, "loss": 0.1996, "step": 16276 }, { "epoch": 4.3312932410856835, "grad_norm": 0.3013097643852234, "learning_rate": 9.05807443030192e-09, "loss": 0.1909, "step": 16277 }, { "epoch": 4.331559340074508, "grad_norm": 0.29489168524742126, "learning_rate": 9.050980461025382e-09, "loss": 0.1793, "step": 16278 }, { "epoch": 4.331825439063332, "grad_norm": 0.3011491000652313, "learning_rate": 9.043889139038662e-09, "loss": 0.1849, "step": 16279 }, { "epoch": 4.332091538052156, "grad_norm": 0.31249895691871643, "learning_rate": 9.036800464548155e-09, "loss": 0.198, "step": 16280 }, { "epoch": 4.332357637040979, "grad_norm": 0.31745681166648865, "learning_rate": 9.029714437760228e-09, "loss": 0.2084, "step": 16281 }, { "epoch": 4.332623736029803, "grad_norm": 0.30162712931632996, "learning_rate": 9.022631058881092e-09, "loss": 0.189, "step": 16282 }, { "epoch": 4.332889835018627, "grad_norm": 0.32317066192626953, "learning_rate": 9.01555032811694e-09, "loss": 0.1757, "step": 16283 }, { "epoch": 4.333155934007451, "grad_norm": 0.32135409116744995, "learning_rate": 9.008472245673904e-09, "loss": 0.2147, "step": 16284 }, { "epoch": 4.333422032996275, "grad_norm": 0.307566374540329, "learning_rate": 9.001396811757967e-09, "loss": 0.1895, "step": 16285 }, { "epoch": 4.333688131985099, "grad_norm": 0.28216224908828735, "learning_rate": 8.994324026575085e-09, "loss": 0.1834, "step": 16286 }, { "epoch": 4.333954230973922, "grad_norm": 0.2840820550918579, "learning_rate": 8.987253890331137e-09, "loss": 0.1802, "step": 16287 }, { "epoch": 4.334220329962746, "grad_norm": 0.33259668946266174, "learning_rate": 8.980186403231927e-09, "loss": 0.2042, "step": 16288 }, { "epoch": 4.33448642895157, "grad_norm": 0.2899925410747528, "learning_rate": 8.973121565483155e-09, "loss": 0.1999, "step": 16289 }, { "epoch": 4.334752527940394, "grad_norm": 0.31559038162231445, "learning_rate": 8.966059377290457e-09, "loss": 0.2005, "step": 16290 }, { "epoch": 4.335018626929218, "grad_norm": 0.3206759989261627, "learning_rate": 8.958999838859416e-09, "loss": 0.1748, "step": 16291 }, { "epoch": 4.335284725918042, "grad_norm": 0.31680506467819214, "learning_rate": 8.951942950395497e-09, "loss": 0.1846, "step": 16292 }, { "epoch": 4.335550824906865, "grad_norm": 0.27084630727767944, "learning_rate": 8.944888712104137e-09, "loss": 0.1733, "step": 16293 }, { "epoch": 4.335816923895689, "grad_norm": 0.2936408519744873, "learning_rate": 8.937837124190617e-09, "loss": 0.1838, "step": 16294 }, { "epoch": 4.336083022884513, "grad_norm": 0.286358118057251, "learning_rate": 8.930788186860227e-09, "loss": 0.2105, "step": 16295 }, { "epoch": 4.3363491218733365, "grad_norm": 0.3106076419353485, "learning_rate": 8.923741900318127e-09, "loss": 0.1939, "step": 16296 }, { "epoch": 4.336615220862161, "grad_norm": 0.305978000164032, "learning_rate": 8.91669826476943e-09, "loss": 0.1863, "step": 16297 }, { "epoch": 4.336881319850985, "grad_norm": 0.32627007365226746, "learning_rate": 8.909657280419148e-09, "loss": 0.1978, "step": 16298 }, { "epoch": 4.337147418839808, "grad_norm": 0.29875022172927856, "learning_rate": 8.90261894747223e-09, "loss": 0.183, "step": 16299 }, { "epoch": 4.337413517828632, "grad_norm": 0.2852087914943695, "learning_rate": 8.895583266133555e-09, "loss": 0.1878, "step": 16300 }, { "epoch": 4.337679616817456, "grad_norm": 0.2666836678981781, "learning_rate": 8.888550236607895e-09, "loss": 0.1693, "step": 16301 }, { "epoch": 4.3379457158062795, "grad_norm": 0.3063143789768219, "learning_rate": 8.881519859099963e-09, "loss": 0.194, "step": 16302 }, { "epoch": 4.338211814795104, "grad_norm": 0.3203032314777374, "learning_rate": 8.874492133814416e-09, "loss": 0.19, "step": 16303 }, { "epoch": 4.338477913783928, "grad_norm": 0.2855737805366516, "learning_rate": 8.86746706095578e-09, "loss": 0.1782, "step": 16304 }, { "epoch": 4.338744012772752, "grad_norm": 0.3244806230068207, "learning_rate": 8.86044464072856e-09, "loss": 0.1875, "step": 16305 }, { "epoch": 4.339010111761575, "grad_norm": 0.33305788040161133, "learning_rate": 8.853424873337145e-09, "loss": 0.2119, "step": 16306 }, { "epoch": 4.339276210750399, "grad_norm": 0.2917881906032562, "learning_rate": 8.846407758985885e-09, "loss": 0.1889, "step": 16307 }, { "epoch": 4.339542309739223, "grad_norm": 0.29746896028518677, "learning_rate": 8.839393297879005e-09, "loss": 0.197, "step": 16308 }, { "epoch": 4.339808408728047, "grad_norm": 0.30826106667518616, "learning_rate": 8.832381490220697e-09, "loss": 0.1887, "step": 16309 }, { "epoch": 4.340074507716871, "grad_norm": 0.28477415442466736, "learning_rate": 8.825372336215053e-09, "loss": 0.1813, "step": 16310 }, { "epoch": 4.340340606705695, "grad_norm": 0.29678937792778015, "learning_rate": 8.8183658360661e-09, "loss": 0.1932, "step": 16311 }, { "epoch": 4.340606705694518, "grad_norm": 0.3026164174079895, "learning_rate": 8.81136198997775e-09, "loss": 0.1882, "step": 16312 }, { "epoch": 4.340872804683342, "grad_norm": 0.3003731966018677, "learning_rate": 8.804360798153887e-09, "loss": 0.2125, "step": 16313 }, { "epoch": 4.341138903672166, "grad_norm": 0.3039722144603729, "learning_rate": 8.797362260798292e-09, "loss": 0.2016, "step": 16314 }, { "epoch": 4.34140500266099, "grad_norm": 0.3037361800670624, "learning_rate": 8.790366378114666e-09, "loss": 0.1802, "step": 16315 }, { "epoch": 4.341671101649814, "grad_norm": 0.297610878944397, "learning_rate": 8.783373150306661e-09, "loss": 0.1845, "step": 16316 }, { "epoch": 4.341937200638638, "grad_norm": 0.3102847635746002, "learning_rate": 8.776382577577823e-09, "loss": 0.1752, "step": 16317 }, { "epoch": 4.342203299627461, "grad_norm": 0.37669408321380615, "learning_rate": 8.769394660131624e-09, "loss": 0.174, "step": 16318 }, { "epoch": 4.342469398616285, "grad_norm": 0.3038609027862549, "learning_rate": 8.762409398171478e-09, "loss": 0.1827, "step": 16319 }, { "epoch": 4.342735497605109, "grad_norm": 0.3283623456954956, "learning_rate": 8.75542679190071e-09, "loss": 0.2166, "step": 16320 }, { "epoch": 4.343001596593933, "grad_norm": 0.3190031051635742, "learning_rate": 8.748446841522528e-09, "loss": 0.1875, "step": 16321 }, { "epoch": 4.343267695582757, "grad_norm": 0.3112242817878723, "learning_rate": 8.741469547240132e-09, "loss": 0.2133, "step": 16322 }, { "epoch": 4.343533794571581, "grad_norm": 0.2903011739253998, "learning_rate": 8.734494909256629e-09, "loss": 0.1862, "step": 16323 }, { "epoch": 4.343799893560404, "grad_norm": 0.31468263268470764, "learning_rate": 8.727522927774988e-09, "loss": 0.2113, "step": 16324 }, { "epoch": 4.344065992549228, "grad_norm": 0.32050448656082153, "learning_rate": 8.720553602998171e-09, "loss": 0.1894, "step": 16325 }, { "epoch": 4.344332091538052, "grad_norm": 0.34893810749053955, "learning_rate": 8.713586935129036e-09, "loss": 0.208, "step": 16326 }, { "epoch": 4.344598190526876, "grad_norm": 0.2890302240848541, "learning_rate": 8.706622924370366e-09, "loss": 0.1791, "step": 16327 }, { "epoch": 4.3448642895157, "grad_norm": 0.3221621811389923, "learning_rate": 8.699661570924854e-09, "loss": 0.1808, "step": 16328 }, { "epoch": 4.345130388504524, "grad_norm": 0.2963820993900299, "learning_rate": 8.69270287499515e-09, "loss": 0.2008, "step": 16329 }, { "epoch": 4.345396487493348, "grad_norm": 0.321520060300827, "learning_rate": 8.68574683678378e-09, "loss": 0.2035, "step": 16330 }, { "epoch": 4.345662586482171, "grad_norm": 0.30167722702026367, "learning_rate": 8.678793456493216e-09, "loss": 0.1889, "step": 16331 }, { "epoch": 4.345928685470995, "grad_norm": 0.2918670177459717, "learning_rate": 8.671842734325862e-09, "loss": 0.1966, "step": 16332 }, { "epoch": 4.346194784459819, "grad_norm": 0.2932356894016266, "learning_rate": 8.664894670484035e-09, "loss": 0.19, "step": 16333 }, { "epoch": 4.346460883448643, "grad_norm": 0.2899181842803955, "learning_rate": 8.657949265169983e-09, "loss": 0.2008, "step": 16334 }, { "epoch": 4.346726982437467, "grad_norm": 0.2807013690471649, "learning_rate": 8.651006518585858e-09, "loss": 0.1755, "step": 16335 }, { "epoch": 4.346993081426291, "grad_norm": 0.3004627823829651, "learning_rate": 8.644066430933738e-09, "loss": 0.1763, "step": 16336 }, { "epoch": 4.347259180415114, "grad_norm": 0.2990252375602722, "learning_rate": 8.637129002415655e-09, "loss": 0.1891, "step": 16337 }, { "epoch": 4.347525279403938, "grad_norm": 0.31733280420303345, "learning_rate": 8.630194233233534e-09, "loss": 0.209, "step": 16338 }, { "epoch": 4.347791378392762, "grad_norm": 0.27153071761131287, "learning_rate": 8.623262123589204e-09, "loss": 0.1653, "step": 16339 }, { "epoch": 4.348057477381586, "grad_norm": 0.2890167236328125, "learning_rate": 8.616332673684456e-09, "loss": 0.2, "step": 16340 }, { "epoch": 4.34832357637041, "grad_norm": 0.2925736606121063, "learning_rate": 8.609405883720988e-09, "loss": 0.175, "step": 16341 }, { "epoch": 4.348589675359234, "grad_norm": 0.2846527099609375, "learning_rate": 8.602481753900425e-09, "loss": 0.1836, "step": 16342 }, { "epoch": 4.348855774348057, "grad_norm": 0.30256766080856323, "learning_rate": 8.59556028442432e-09, "loss": 0.2073, "step": 16343 }, { "epoch": 4.349121873336881, "grad_norm": 0.27593541145324707, "learning_rate": 8.588641475494108e-09, "loss": 0.1822, "step": 16344 }, { "epoch": 4.349387972325705, "grad_norm": 0.3025299608707428, "learning_rate": 8.581725327311195e-09, "loss": 0.1894, "step": 16345 }, { "epoch": 4.349654071314529, "grad_norm": 0.3384433388710022, "learning_rate": 8.574811840076901e-09, "loss": 0.2249, "step": 16346 }, { "epoch": 4.349920170303353, "grad_norm": 0.4059697389602661, "learning_rate": 8.567901013992462e-09, "loss": 0.1875, "step": 16347 }, { "epoch": 4.350186269292177, "grad_norm": 0.3110412657260895, "learning_rate": 8.560992849259008e-09, "loss": 0.1894, "step": 16348 }, { "epoch": 4.350452368281, "grad_norm": 0.3009033799171448, "learning_rate": 8.554087346077633e-09, "loss": 0.1909, "step": 16349 }, { "epoch": 4.350718467269824, "grad_norm": 0.2995396554470062, "learning_rate": 8.54718450464933e-09, "loss": 0.1982, "step": 16350 }, { "epoch": 4.350984566258648, "grad_norm": 0.3211521804332733, "learning_rate": 8.54028432517504e-09, "loss": 0.1815, "step": 16351 }, { "epoch": 4.351250665247472, "grad_norm": 0.29802682995796204, "learning_rate": 8.533386807855602e-09, "loss": 0.1806, "step": 16352 }, { "epoch": 4.351516764236296, "grad_norm": 0.3175867795944214, "learning_rate": 8.526491952891779e-09, "loss": 0.1924, "step": 16353 }, { "epoch": 4.35178286322512, "grad_norm": 0.30757176876068115, "learning_rate": 8.519599760484265e-09, "loss": 0.213, "step": 16354 }, { "epoch": 4.352048962213944, "grad_norm": 0.2910744845867157, "learning_rate": 8.512710230833687e-09, "loss": 0.1728, "step": 16355 }, { "epoch": 4.352315061202767, "grad_norm": 0.32400989532470703, "learning_rate": 8.505823364140574e-09, "loss": 0.1957, "step": 16356 }, { "epoch": 4.352581160191591, "grad_norm": 0.2921878695487976, "learning_rate": 8.498939160605367e-09, "loss": 0.1841, "step": 16357 }, { "epoch": 4.3528472591804155, "grad_norm": 0.2941671311855316, "learning_rate": 8.49205762042846e-09, "loss": 0.1924, "step": 16358 }, { "epoch": 4.353113358169239, "grad_norm": 0.30344733595848083, "learning_rate": 8.485178743810162e-09, "loss": 0.1819, "step": 16359 }, { "epoch": 4.353379457158063, "grad_norm": 0.3019130825996399, "learning_rate": 8.478302530950687e-09, "loss": 0.1769, "step": 16360 }, { "epoch": 4.353645556146887, "grad_norm": 0.2851985991001129, "learning_rate": 8.4714289820502e-09, "loss": 0.2048, "step": 16361 }, { "epoch": 4.35391165513571, "grad_norm": 0.2926129996776581, "learning_rate": 8.464558097308783e-09, "loss": 0.1824, "step": 16362 }, { "epoch": 4.354177754124534, "grad_norm": 0.30285272002220154, "learning_rate": 8.457689876926388e-09, "loss": 0.2058, "step": 16363 }, { "epoch": 4.3544438531133585, "grad_norm": 0.3304292857646942, "learning_rate": 8.450824321102957e-09, "loss": 0.2297, "step": 16364 }, { "epoch": 4.354709952102182, "grad_norm": 0.31267523765563965, "learning_rate": 8.443961430038315e-09, "loss": 0.1948, "step": 16365 }, { "epoch": 4.354976051091006, "grad_norm": 0.2920478284358978, "learning_rate": 8.437101203932262e-09, "loss": 0.1964, "step": 16366 }, { "epoch": 4.35524215007983, "grad_norm": 0.2971596419811249, "learning_rate": 8.430243642984426e-09, "loss": 0.1962, "step": 16367 }, { "epoch": 4.355508249068653, "grad_norm": 0.29438117146492004, "learning_rate": 8.423388747394445e-09, "loss": 0.202, "step": 16368 }, { "epoch": 4.355774348057477, "grad_norm": 0.268381267786026, "learning_rate": 8.416536517361839e-09, "loss": 0.1611, "step": 16369 }, { "epoch": 4.3560404470463014, "grad_norm": 0.35308000445365906, "learning_rate": 8.40968695308606e-09, "loss": 0.2143, "step": 16370 }, { "epoch": 4.356306546035125, "grad_norm": 0.3086749017238617, "learning_rate": 8.40284005476648e-09, "loss": 0.1926, "step": 16371 }, { "epoch": 4.356572645023949, "grad_norm": 0.26662299036979675, "learning_rate": 8.395995822602397e-09, "loss": 0.1657, "step": 16372 }, { "epoch": 4.356838744012773, "grad_norm": 0.2883439362049103, "learning_rate": 8.38915425679304e-09, "loss": 0.198, "step": 16373 }, { "epoch": 4.357104843001597, "grad_norm": 0.317060649394989, "learning_rate": 8.382315357537528e-09, "loss": 0.1756, "step": 16374 }, { "epoch": 4.35737094199042, "grad_norm": 0.26489585638046265, "learning_rate": 8.375479125034945e-09, "loss": 0.1812, "step": 16375 }, { "epoch": 4.357637040979244, "grad_norm": 0.326824426651001, "learning_rate": 8.368645559484255e-09, "loss": 0.194, "step": 16376 }, { "epoch": 4.357903139968068, "grad_norm": 0.2931162416934967, "learning_rate": 8.361814661084366e-09, "loss": 0.2023, "step": 16377 }, { "epoch": 4.358169238956892, "grad_norm": 0.2803453803062439, "learning_rate": 8.354986430034127e-09, "loss": 0.179, "step": 16378 }, { "epoch": 4.358435337945716, "grad_norm": 0.31094396114349365, "learning_rate": 8.348160866532272e-09, "loss": 0.2167, "step": 16379 }, { "epoch": 4.35870143693454, "grad_norm": 0.2733193635940552, "learning_rate": 8.341337970777484e-09, "loss": 0.1714, "step": 16380 }, { "epoch": 4.358967535923363, "grad_norm": 0.29405099153518677, "learning_rate": 8.334517742968362e-09, "loss": 0.1937, "step": 16381 }, { "epoch": 4.359233634912187, "grad_norm": 0.2946379482746124, "learning_rate": 8.327700183303433e-09, "loss": 0.191, "step": 16382 }, { "epoch": 4.3594997339010115, "grad_norm": 0.2884482443332672, "learning_rate": 8.320885291981116e-09, "loss": 0.1914, "step": 16383 }, { "epoch": 4.359765832889835, "grad_norm": 0.3001362979412079, "learning_rate": 8.314073069199779e-09, "loss": 0.1855, "step": 16384 }, { "epoch": 4.360031931878659, "grad_norm": 0.30541184544563293, "learning_rate": 8.307263515157736e-09, "loss": 0.1992, "step": 16385 }, { "epoch": 4.360298030867483, "grad_norm": 0.29311826825141907, "learning_rate": 8.300456630053154e-09, "loss": 0.1754, "step": 16386 }, { "epoch": 4.360564129856306, "grad_norm": 0.27835091948509216, "learning_rate": 8.293652414084184e-09, "loss": 0.1772, "step": 16387 }, { "epoch": 4.36083022884513, "grad_norm": 0.28536534309387207, "learning_rate": 8.28685086744888e-09, "loss": 0.1829, "step": 16388 }, { "epoch": 4.3610963278339545, "grad_norm": 0.2901393175125122, "learning_rate": 8.280051990345216e-09, "loss": 0.1807, "step": 16389 }, { "epoch": 4.361362426822778, "grad_norm": 0.30575352907180786, "learning_rate": 8.2732557829711e-09, "loss": 0.1829, "step": 16390 }, { "epoch": 4.361628525811602, "grad_norm": 0.2811282277107239, "learning_rate": 8.266462245524331e-09, "loss": 0.1878, "step": 16391 }, { "epoch": 4.361894624800426, "grad_norm": 0.30390188097953796, "learning_rate": 8.25967137820266e-09, "loss": 0.1945, "step": 16392 }, { "epoch": 4.362160723789249, "grad_norm": 0.3060331642627716, "learning_rate": 8.252883181203785e-09, "loss": 0.21, "step": 16393 }, { "epoch": 4.362426822778073, "grad_norm": 0.30319538712501526, "learning_rate": 8.246097654725227e-09, "loss": 0.2245, "step": 16394 }, { "epoch": 4.3626929217668975, "grad_norm": 0.27634966373443604, "learning_rate": 8.239314798964537e-09, "loss": 0.1686, "step": 16395 }, { "epoch": 4.362959020755721, "grad_norm": 0.29572156071662903, "learning_rate": 8.232534614119146e-09, "loss": 0.1774, "step": 16396 }, { "epoch": 4.363225119744545, "grad_norm": 0.2950867712497711, "learning_rate": 8.225757100386388e-09, "loss": 0.2003, "step": 16397 }, { "epoch": 4.363491218733369, "grad_norm": 0.3044168949127197, "learning_rate": 8.218982257963558e-09, "loss": 0.2042, "step": 16398 }, { "epoch": 4.363757317722193, "grad_norm": 0.31537172198295593, "learning_rate": 8.212210087047856e-09, "loss": 0.2022, "step": 16399 }, { "epoch": 4.364023416711016, "grad_norm": 0.2887437343597412, "learning_rate": 8.205440587836376e-09, "loss": 0.1785, "step": 16400 }, { "epoch": 4.3642895156998405, "grad_norm": 0.29899880290031433, "learning_rate": 8.19867376052621e-09, "loss": 0.1817, "step": 16401 }, { "epoch": 4.364555614688665, "grad_norm": 0.30027395486831665, "learning_rate": 8.191909605314262e-09, "loss": 0.1891, "step": 16402 }, { "epoch": 4.364821713677488, "grad_norm": 0.2981601059436798, "learning_rate": 8.185148122397456e-09, "loss": 0.1887, "step": 16403 }, { "epoch": 4.365087812666312, "grad_norm": 0.3087146282196045, "learning_rate": 8.178389311972611e-09, "loss": 0.2107, "step": 16404 }, { "epoch": 4.365353911655136, "grad_norm": 0.2873038947582245, "learning_rate": 8.171633174236425e-09, "loss": 0.1803, "step": 16405 }, { "epoch": 4.365620010643959, "grad_norm": 0.30739620327949524, "learning_rate": 8.164879709385564e-09, "loss": 0.1833, "step": 16406 }, { "epoch": 4.3658861096327835, "grad_norm": 0.32608944177627563, "learning_rate": 8.158128917616602e-09, "loss": 0.1909, "step": 16407 }, { "epoch": 4.366152208621608, "grad_norm": 0.34772413969039917, "learning_rate": 8.151380799126051e-09, "loss": 0.1894, "step": 16408 }, { "epoch": 4.366418307610431, "grad_norm": 0.3021889925003052, "learning_rate": 8.144635354110318e-09, "loss": 0.1862, "step": 16409 }, { "epoch": 4.366684406599255, "grad_norm": 0.2851155996322632, "learning_rate": 8.137892582765749e-09, "loss": 0.1777, "step": 16410 }, { "epoch": 4.366950505588079, "grad_norm": 0.3094400465488434, "learning_rate": 8.131152485288628e-09, "loss": 0.1753, "step": 16411 }, { "epoch": 4.367216604576902, "grad_norm": 0.302873432636261, "learning_rate": 8.124415061875112e-09, "loss": 0.1996, "step": 16412 }, { "epoch": 4.367482703565726, "grad_norm": 0.29017549753189087, "learning_rate": 8.11768031272131e-09, "loss": 0.1933, "step": 16413 }, { "epoch": 4.367748802554551, "grad_norm": 0.2971510887145996, "learning_rate": 8.110948238023275e-09, "loss": 0.1938, "step": 16414 }, { "epoch": 4.368014901543374, "grad_norm": 0.2746948301792145, "learning_rate": 8.10421883797694e-09, "loss": 0.1806, "step": 16415 }, { "epoch": 4.368281000532198, "grad_norm": 0.29567021131515503, "learning_rate": 8.097492112778193e-09, "loss": 0.1763, "step": 16416 }, { "epoch": 4.368547099521022, "grad_norm": 0.31796595454216003, "learning_rate": 8.090768062622822e-09, "loss": 0.2008, "step": 16417 }, { "epoch": 4.368813198509845, "grad_norm": 0.3022434115409851, "learning_rate": 8.084046687706559e-09, "loss": 0.1784, "step": 16418 }, { "epoch": 4.369079297498669, "grad_norm": 0.3138270676136017, "learning_rate": 8.077327988225046e-09, "loss": 0.2042, "step": 16419 }, { "epoch": 4.3693453964874935, "grad_norm": 0.30937445163726807, "learning_rate": 8.070611964373852e-09, "loss": 0.197, "step": 16420 }, { "epoch": 4.369611495476317, "grad_norm": 0.3211160898208618, "learning_rate": 8.06389861634843e-09, "loss": 0.206, "step": 16421 }, { "epoch": 4.369877594465141, "grad_norm": 0.3060791492462158, "learning_rate": 8.057187944344212e-09, "loss": 0.1832, "step": 16422 }, { "epoch": 4.370143693453965, "grad_norm": 0.3411019742488861, "learning_rate": 8.050479948556533e-09, "loss": 0.2012, "step": 16423 }, { "epoch": 4.370409792442789, "grad_norm": 0.27612197399139404, "learning_rate": 8.043774629180644e-09, "loss": 0.1817, "step": 16424 }, { "epoch": 4.370675891431612, "grad_norm": 0.3236585259437561, "learning_rate": 8.037071986411703e-09, "loss": 0.2112, "step": 16425 }, { "epoch": 4.3709419904204365, "grad_norm": 0.3090505301952362, "learning_rate": 8.030372020444809e-09, "loss": 0.1682, "step": 16426 }, { "epoch": 4.371208089409261, "grad_norm": 0.3355942666530609, "learning_rate": 8.023674731475005e-09, "loss": 0.1886, "step": 16427 }, { "epoch": 4.371474188398084, "grad_norm": 0.28990718722343445, "learning_rate": 8.016980119697204e-09, "loss": 0.191, "step": 16428 }, { "epoch": 4.371740287386908, "grad_norm": 0.30464285612106323, "learning_rate": 8.01028818530629e-09, "loss": 0.1699, "step": 16429 }, { "epoch": 4.372006386375732, "grad_norm": 0.30094224214553833, "learning_rate": 8.003598928497035e-09, "loss": 0.2031, "step": 16430 }, { "epoch": 4.372272485364555, "grad_norm": 0.298542320728302, "learning_rate": 7.996912349464147e-09, "loss": 0.1982, "step": 16431 }, { "epoch": 4.3725385843533795, "grad_norm": 0.300658643245697, "learning_rate": 7.990228448402248e-09, "loss": 0.1876, "step": 16432 }, { "epoch": 4.372804683342204, "grad_norm": 0.28331613540649414, "learning_rate": 7.983547225505904e-09, "loss": 0.1802, "step": 16433 }, { "epoch": 4.373070782331027, "grad_norm": 0.3035441040992737, "learning_rate": 7.976868680969584e-09, "loss": 0.1983, "step": 16434 }, { "epoch": 4.373336881319851, "grad_norm": 0.28227779269218445, "learning_rate": 7.970192814987675e-09, "loss": 0.1845, "step": 16435 }, { "epoch": 4.373602980308675, "grad_norm": 0.31621676683425903, "learning_rate": 7.963519627754512e-09, "loss": 0.1859, "step": 16436 }, { "epoch": 4.373869079297498, "grad_norm": 0.30938392877578735, "learning_rate": 7.956849119464315e-09, "loss": 0.2044, "step": 16437 }, { "epoch": 4.3741351782863225, "grad_norm": 0.29950857162475586, "learning_rate": 7.950181290311265e-09, "loss": 0.1812, "step": 16438 }, { "epoch": 4.374401277275147, "grad_norm": 0.31783872842788696, "learning_rate": 7.943516140489415e-09, "loss": 0.2027, "step": 16439 }, { "epoch": 4.37466737626397, "grad_norm": 0.29155057668685913, "learning_rate": 7.9368536701928e-09, "loss": 0.1709, "step": 16440 }, { "epoch": 4.374933475252794, "grad_norm": 0.521140992641449, "learning_rate": 7.930193879615321e-09, "loss": 0.1857, "step": 16441 }, { "epoch": 4.375199574241618, "grad_norm": 0.28299733996391296, "learning_rate": 7.923536768950856e-09, "loss": 0.175, "step": 16442 }, { "epoch": 4.375465673230441, "grad_norm": 0.31385454535484314, "learning_rate": 7.916882338393161e-09, "loss": 0.2111, "step": 16443 }, { "epoch": 4.3757317722192655, "grad_norm": 0.2891943156719208, "learning_rate": 7.910230588135924e-09, "loss": 0.1814, "step": 16444 }, { "epoch": 4.37599787120809, "grad_norm": 0.33209994435310364, "learning_rate": 7.90358151837277e-09, "loss": 0.2202, "step": 16445 }, { "epoch": 4.376263970196913, "grad_norm": 0.3061325252056122, "learning_rate": 7.896935129297222e-09, "loss": 0.2062, "step": 16446 }, { "epoch": 4.376530069185737, "grad_norm": 0.27852776646614075, "learning_rate": 7.890291421102769e-09, "loss": 0.1802, "step": 16447 }, { "epoch": 4.376796168174561, "grad_norm": 0.31535467505455017, "learning_rate": 7.883650393982755e-09, "loss": 0.2155, "step": 16448 }, { "epoch": 4.377062267163385, "grad_norm": 0.29940396547317505, "learning_rate": 7.877012048130505e-09, "loss": 0.1848, "step": 16449 }, { "epoch": 4.377328366152208, "grad_norm": 0.29794156551361084, "learning_rate": 7.870376383739242e-09, "loss": 0.1965, "step": 16450 }, { "epoch": 4.377594465141033, "grad_norm": 0.3007708191871643, "learning_rate": 7.863743401002099e-09, "loss": 0.197, "step": 16451 }, { "epoch": 4.377860564129857, "grad_norm": 0.28286412358283997, "learning_rate": 7.857113100112156e-09, "loss": 0.1825, "step": 16452 }, { "epoch": 4.37812666311868, "grad_norm": 0.31733620166778564, "learning_rate": 7.850485481262414e-09, "loss": 0.1936, "step": 16453 }, { "epoch": 4.378392762107504, "grad_norm": 0.27547746896743774, "learning_rate": 7.843860544645776e-09, "loss": 0.1882, "step": 16454 }, { "epoch": 4.378658861096328, "grad_norm": 0.29598432779312134, "learning_rate": 7.837238290455073e-09, "loss": 0.1763, "step": 16455 }, { "epoch": 4.378924960085151, "grad_norm": 0.3124958872795105, "learning_rate": 7.830618718883076e-09, "loss": 0.1927, "step": 16456 }, { "epoch": 4.3791910590739755, "grad_norm": 0.3120875656604767, "learning_rate": 7.82400183012244e-09, "loss": 0.1983, "step": 16457 }, { "epoch": 4.3794571580628, "grad_norm": 0.3041799068450928, "learning_rate": 7.817387624365779e-09, "loss": 0.2054, "step": 16458 }, { "epoch": 4.379723257051623, "grad_norm": 0.3030855059623718, "learning_rate": 7.810776101805616e-09, "loss": 0.1873, "step": 16459 }, { "epoch": 4.379989356040447, "grad_norm": 0.3254372477531433, "learning_rate": 7.804167262634388e-09, "loss": 0.2097, "step": 16460 }, { "epoch": 4.380255455029271, "grad_norm": 0.29010340571403503, "learning_rate": 7.79756110704447e-09, "loss": 0.1934, "step": 16461 }, { "epoch": 4.380521554018094, "grad_norm": 0.29049810767173767, "learning_rate": 7.790957635228156e-09, "loss": 0.1813, "step": 16462 }, { "epoch": 4.3807876530069185, "grad_norm": 0.32415273785591125, "learning_rate": 7.784356847377648e-09, "loss": 0.2024, "step": 16463 }, { "epoch": 4.381053751995743, "grad_norm": 0.28550252318382263, "learning_rate": 7.77775874368507e-09, "loss": 0.1721, "step": 16464 }, { "epoch": 4.381319850984566, "grad_norm": 0.3297935724258423, "learning_rate": 7.771163324342478e-09, "loss": 0.1972, "step": 16465 }, { "epoch": 4.38158594997339, "grad_norm": 0.27494531869888306, "learning_rate": 7.764570589541875e-09, "loss": 0.1815, "step": 16466 }, { "epoch": 4.381852048962214, "grad_norm": 0.2947549819946289, "learning_rate": 7.757980539475118e-09, "loss": 0.1816, "step": 16467 }, { "epoch": 4.382118147951037, "grad_norm": 0.29266357421875, "learning_rate": 7.751393174334042e-09, "loss": 0.1933, "step": 16468 }, { "epoch": 4.3823842469398615, "grad_norm": 0.28244155645370483, "learning_rate": 7.744808494310385e-09, "loss": 0.1719, "step": 16469 }, { "epoch": 4.382650345928686, "grad_norm": 0.2920219600200653, "learning_rate": 7.738226499595812e-09, "loss": 0.2048, "step": 16470 }, { "epoch": 4.382916444917509, "grad_norm": 0.3049633502960205, "learning_rate": 7.731647190381919e-09, "loss": 0.2054, "step": 16471 }, { "epoch": 4.383182543906333, "grad_norm": 0.29962360858917236, "learning_rate": 7.725070566860204e-09, "loss": 0.1868, "step": 16472 }, { "epoch": 4.383448642895157, "grad_norm": 0.49196338653564453, "learning_rate": 7.718496629222082e-09, "loss": 0.1854, "step": 16473 }, { "epoch": 4.383714741883981, "grad_norm": 0.30210620164871216, "learning_rate": 7.711925377658946e-09, "loss": 0.1874, "step": 16474 }, { "epoch": 4.3839808408728045, "grad_norm": 0.3027087450027466, "learning_rate": 7.705356812362019e-09, "loss": 0.2036, "step": 16475 }, { "epoch": 4.384246939861629, "grad_norm": 0.2951517403125763, "learning_rate": 7.698790933522515e-09, "loss": 0.2034, "step": 16476 }, { "epoch": 4.384513038850453, "grad_norm": 0.3054961562156677, "learning_rate": 7.692227741331547e-09, "loss": 0.1877, "step": 16477 }, { "epoch": 4.384779137839276, "grad_norm": 0.29511758685112, "learning_rate": 7.685667235980153e-09, "loss": 0.195, "step": 16478 }, { "epoch": 4.3850452368281, "grad_norm": 0.30778512358665466, "learning_rate": 7.679109417659302e-09, "loss": 0.2019, "step": 16479 }, { "epoch": 4.385311335816924, "grad_norm": 0.28654205799102783, "learning_rate": 7.672554286559862e-09, "loss": 0.1829, "step": 16480 }, { "epoch": 4.3855774348057475, "grad_norm": 0.288974404335022, "learning_rate": 7.666001842872638e-09, "loss": 0.1864, "step": 16481 }, { "epoch": 4.385843533794572, "grad_norm": 0.2819720208644867, "learning_rate": 7.659452086788365e-09, "loss": 0.1833, "step": 16482 }, { "epoch": 4.386109632783396, "grad_norm": 0.29957422614097595, "learning_rate": 7.652905018497691e-09, "loss": 0.1955, "step": 16483 }, { "epoch": 4.386375731772219, "grad_norm": 0.3442877531051636, "learning_rate": 7.646360638191162e-09, "loss": 0.2131, "step": 16484 }, { "epoch": 4.386641830761043, "grad_norm": 0.27463340759277344, "learning_rate": 7.639818946059284e-09, "loss": 0.1577, "step": 16485 }, { "epoch": 4.386907929749867, "grad_norm": 0.29967349767684937, "learning_rate": 7.63327994229248e-09, "loss": 0.1914, "step": 16486 }, { "epoch": 4.3871740287386904, "grad_norm": 0.2809702157974243, "learning_rate": 7.626743627081057e-09, "loss": 0.1939, "step": 16487 }, { "epoch": 4.387440127727515, "grad_norm": 0.3753601312637329, "learning_rate": 7.620210000615267e-09, "loss": 0.1899, "step": 16488 }, { "epoch": 4.387706226716339, "grad_norm": 0.2748008668422699, "learning_rate": 7.61367906308531e-09, "loss": 0.1881, "step": 16489 }, { "epoch": 4.387972325705162, "grad_norm": 0.26478326320648193, "learning_rate": 7.607150814681285e-09, "loss": 0.1818, "step": 16490 }, { "epoch": 4.388238424693986, "grad_norm": 0.2790495455265045, "learning_rate": 7.600625255593195e-09, "loss": 0.174, "step": 16491 }, { "epoch": 4.38850452368281, "grad_norm": 0.36095014214515686, "learning_rate": 7.594102386011014e-09, "loss": 0.1919, "step": 16492 }, { "epoch": 4.388770622671634, "grad_norm": 0.32228219509124756, "learning_rate": 7.587582206124565e-09, "loss": 0.196, "step": 16493 }, { "epoch": 4.3890367216604576, "grad_norm": 0.3299301564693451, "learning_rate": 7.581064716123654e-09, "loss": 0.2019, "step": 16494 }, { "epoch": 4.389302820649282, "grad_norm": 0.29505351185798645, "learning_rate": 7.57454991619797e-09, "loss": 0.1897, "step": 16495 }, { "epoch": 4.389568919638105, "grad_norm": 0.32887136936187744, "learning_rate": 7.568037806537176e-09, "loss": 0.1969, "step": 16496 }, { "epoch": 4.389835018626929, "grad_norm": 0.3123182952404022, "learning_rate": 7.561528387330796e-09, "loss": 0.2002, "step": 16497 }, { "epoch": 4.390101117615753, "grad_norm": 0.3052566647529602, "learning_rate": 7.555021658768313e-09, "loss": 0.1897, "step": 16498 }, { "epoch": 4.390367216604577, "grad_norm": 0.2927473187446594, "learning_rate": 7.54851762103912e-09, "loss": 0.1919, "step": 16499 }, { "epoch": 4.3906333155934005, "grad_norm": 0.28051692247390747, "learning_rate": 7.54201627433253e-09, "loss": 0.1772, "step": 16500 }, { "epoch": 4.390899414582225, "grad_norm": 0.3141613304615021, "learning_rate": 7.535517618837795e-09, "loss": 0.2054, "step": 16501 }, { "epoch": 4.391165513571049, "grad_norm": 0.3085402250289917, "learning_rate": 7.529021654744038e-09, "loss": 0.2145, "step": 16502 }, { "epoch": 4.391431612559872, "grad_norm": 0.2970435619354248, "learning_rate": 7.522528382240378e-09, "loss": 0.1894, "step": 16503 }, { "epoch": 4.391697711548696, "grad_norm": 0.3180775046348572, "learning_rate": 7.516037801515795e-09, "loss": 0.2014, "step": 16504 }, { "epoch": 4.39196381053752, "grad_norm": 0.30649223923683167, "learning_rate": 7.509549912759227e-09, "loss": 0.1991, "step": 16505 }, { "epoch": 4.3922299095263435, "grad_norm": 0.28274065256118774, "learning_rate": 7.503064716159501e-09, "loss": 0.1841, "step": 16506 }, { "epoch": 4.392496008515168, "grad_norm": 0.2778826653957367, "learning_rate": 7.4965822119054e-09, "loss": 0.1949, "step": 16507 }, { "epoch": 4.392762107503992, "grad_norm": 0.27871695160865784, "learning_rate": 7.490102400185605e-09, "loss": 0.183, "step": 16508 }, { "epoch": 4.393028206492815, "grad_norm": 0.27922314405441284, "learning_rate": 7.483625281188732e-09, "loss": 0.1894, "step": 16509 }, { "epoch": 4.393294305481639, "grad_norm": 0.28605136275291443, "learning_rate": 7.477150855103331e-09, "loss": 0.178, "step": 16510 }, { "epoch": 4.393560404470463, "grad_norm": 0.30589133501052856, "learning_rate": 7.470679122117807e-09, "loss": 0.1872, "step": 16511 }, { "epoch": 4.3938265034592865, "grad_norm": 0.2836717367172241, "learning_rate": 7.464210082420575e-09, "loss": 0.1822, "step": 16512 }, { "epoch": 4.394092602448111, "grad_norm": 0.3075672388076782, "learning_rate": 7.457743736199906e-09, "loss": 0.1787, "step": 16513 }, { "epoch": 4.394358701436935, "grad_norm": 0.2934107184410095, "learning_rate": 7.451280083644052e-09, "loss": 0.197, "step": 16514 }, { "epoch": 4.394624800425758, "grad_norm": 0.2752162218093872, "learning_rate": 7.444819124941126e-09, "loss": 0.181, "step": 16515 }, { "epoch": 4.394890899414582, "grad_norm": 0.28128841519355774, "learning_rate": 7.438360860279192e-09, "loss": 0.1786, "step": 16516 }, { "epoch": 4.395156998403406, "grad_norm": 0.29025599360466003, "learning_rate": 7.431905289846241e-09, "loss": 0.1901, "step": 16517 }, { "epoch": 4.39542309739223, "grad_norm": 0.31975895166397095, "learning_rate": 7.42545241383018e-09, "loss": 0.2019, "step": 16518 }, { "epoch": 4.395689196381054, "grad_norm": 0.31081530451774597, "learning_rate": 7.4190022324188584e-09, "loss": 0.1995, "step": 16519 }, { "epoch": 4.395955295369878, "grad_norm": 0.5026330947875977, "learning_rate": 7.41255474579997e-09, "loss": 0.2007, "step": 16520 }, { "epoch": 4.396221394358702, "grad_norm": 0.29483839869499207, "learning_rate": 7.40610995416121e-09, "loss": 0.1906, "step": 16521 }, { "epoch": 4.396487493347525, "grad_norm": 0.3060276210308075, "learning_rate": 7.3996678576901725e-09, "loss": 0.1995, "step": 16522 }, { "epoch": 4.396753592336349, "grad_norm": 0.2954787611961365, "learning_rate": 7.393228456574374e-09, "loss": 0.19, "step": 16523 }, { "epoch": 4.397019691325173, "grad_norm": 0.3064907491207123, "learning_rate": 7.386791751001243e-09, "loss": 0.1851, "step": 16524 }, { "epoch": 4.397285790313997, "grad_norm": 0.28171321749687195, "learning_rate": 7.3803577411581494e-09, "loss": 0.1767, "step": 16525 }, { "epoch": 4.397551889302821, "grad_norm": 0.26949894428253174, "learning_rate": 7.373926427232347e-09, "loss": 0.1727, "step": 16526 }, { "epoch": 4.397817988291645, "grad_norm": 0.29636237025260925, "learning_rate": 7.367497809411039e-09, "loss": 0.1969, "step": 16527 }, { "epoch": 4.398084087280468, "grad_norm": 0.29348674416542053, "learning_rate": 7.361071887881376e-09, "loss": 0.1914, "step": 16528 }, { "epoch": 4.398350186269292, "grad_norm": 0.3037845194339752, "learning_rate": 7.354648662830354e-09, "loss": 0.207, "step": 16529 }, { "epoch": 4.398616285258116, "grad_norm": 0.277438759803772, "learning_rate": 7.348228134444956e-09, "loss": 0.2006, "step": 16530 }, { "epoch": 4.39888238424694, "grad_norm": 0.3074255585670471, "learning_rate": 7.341810302912066e-09, "loss": 0.1778, "step": 16531 }, { "epoch": 4.399148483235764, "grad_norm": 0.30387189984321594, "learning_rate": 7.335395168418501e-09, "loss": 0.1754, "step": 16532 }, { "epoch": 4.399414582224588, "grad_norm": 0.28429529070854187, "learning_rate": 7.328982731150979e-09, "loss": 0.1872, "step": 16533 }, { "epoch": 4.399680681213411, "grad_norm": 0.3107511103153229, "learning_rate": 7.32257299129615e-09, "loss": 0.2083, "step": 16534 }, { "epoch": 4.399946780202235, "grad_norm": 0.29617151618003845, "learning_rate": 7.316165949040598e-09, "loss": 0.1853, "step": 16535 }, { "epoch": 4.400212879191059, "grad_norm": 0.3102778494358063, "learning_rate": 7.309761604570786e-09, "loss": 0.2004, "step": 16536 }, { "epoch": 4.4004789781798825, "grad_norm": 0.33926838636398315, "learning_rate": 7.303359958073174e-09, "loss": 0.1983, "step": 16537 }, { "epoch": 4.400745077168707, "grad_norm": 0.3026040494441986, "learning_rate": 7.296961009734049e-09, "loss": 0.1829, "step": 16538 }, { "epoch": 4.401011176157531, "grad_norm": 0.2963828146457672, "learning_rate": 7.290564759739693e-09, "loss": 0.1887, "step": 16539 }, { "epoch": 4.401277275146354, "grad_norm": 0.2962549030780792, "learning_rate": 7.2841712082762796e-09, "loss": 0.1769, "step": 16540 }, { "epoch": 4.401543374135178, "grad_norm": 0.28675559163093567, "learning_rate": 7.2777803555298945e-09, "loss": 0.1766, "step": 16541 }, { "epoch": 4.401809473124002, "grad_norm": 0.28849029541015625, "learning_rate": 7.271392201686588e-09, "loss": 0.1899, "step": 16542 }, { "epoch": 4.402075572112826, "grad_norm": 0.3026171922683716, "learning_rate": 7.265006746932279e-09, "loss": 0.2044, "step": 16543 }, { "epoch": 4.40234167110165, "grad_norm": 0.3045876622200012, "learning_rate": 7.258623991452861e-09, "loss": 0.2194, "step": 16544 }, { "epoch": 4.402607770090474, "grad_norm": 0.31227728724479675, "learning_rate": 7.252243935434077e-09, "loss": 0.2236, "step": 16545 }, { "epoch": 4.402873869079298, "grad_norm": 0.3086562156677246, "learning_rate": 7.2458665790616546e-09, "loss": 0.1902, "step": 16546 }, { "epoch": 4.403139968068121, "grad_norm": 0.30141180753707886, "learning_rate": 7.239491922521246e-09, "loss": 0.2026, "step": 16547 }, { "epoch": 4.403406067056945, "grad_norm": 0.32465964555740356, "learning_rate": 7.233119965998347e-09, "loss": 0.2048, "step": 16548 }, { "epoch": 4.403672166045769, "grad_norm": 0.2839810848236084, "learning_rate": 7.2267507096784755e-09, "loss": 0.1737, "step": 16549 }, { "epoch": 4.403938265034593, "grad_norm": 0.32215556502342224, "learning_rate": 7.2203841537469945e-09, "loss": 0.1853, "step": 16550 }, { "epoch": 4.404204364023417, "grad_norm": 0.2821880578994751, "learning_rate": 7.214020298389234e-09, "loss": 0.1803, "step": 16551 }, { "epoch": 4.404470463012241, "grad_norm": 0.2957623600959778, "learning_rate": 7.207659143790434e-09, "loss": 0.2002, "step": 16552 }, { "epoch": 4.404736562001064, "grad_norm": 0.3073584735393524, "learning_rate": 7.2013006901357365e-09, "loss": 0.2026, "step": 16553 }, { "epoch": 4.405002660989888, "grad_norm": 0.278702974319458, "learning_rate": 7.194944937610226e-09, "loss": 0.1734, "step": 16554 }, { "epoch": 4.405268759978712, "grad_norm": 0.31618547439575195, "learning_rate": 7.18859188639892e-09, "loss": 0.1949, "step": 16555 }, { "epoch": 4.405534858967536, "grad_norm": 0.3173487186431885, "learning_rate": 7.182241536686706e-09, "loss": 0.1877, "step": 16556 }, { "epoch": 4.40580095795636, "grad_norm": 0.31965363025665283, "learning_rate": 7.1758938886584465e-09, "loss": 0.174, "step": 16557 }, { "epoch": 4.406067056945184, "grad_norm": 0.3038910925388336, "learning_rate": 7.169548942498893e-09, "loss": 0.202, "step": 16558 }, { "epoch": 4.406333155934007, "grad_norm": 0.310176819562912, "learning_rate": 7.163206698392743e-09, "loss": 0.1974, "step": 16559 }, { "epoch": 4.406599254922831, "grad_norm": 0.30463576316833496, "learning_rate": 7.156867156524593e-09, "loss": 0.2019, "step": 16560 }, { "epoch": 4.406865353911655, "grad_norm": 0.27989432215690613, "learning_rate": 7.150530317078973e-09, "loss": 0.186, "step": 16561 }, { "epoch": 4.407131452900479, "grad_norm": 0.297069787979126, "learning_rate": 7.144196180240347e-09, "loss": 0.1945, "step": 16562 }, { "epoch": 4.407397551889303, "grad_norm": 0.3020130395889282, "learning_rate": 7.137864746193057e-09, "loss": 0.182, "step": 16563 }, { "epoch": 4.407663650878127, "grad_norm": 0.3053734004497528, "learning_rate": 7.131536015121442e-09, "loss": 0.1824, "step": 16564 }, { "epoch": 4.40792974986695, "grad_norm": 0.31153205037117004, "learning_rate": 7.125209987209657e-09, "loss": 0.1824, "step": 16565 }, { "epoch": 4.408195848855774, "grad_norm": 0.2919328510761261, "learning_rate": 7.118886662641865e-09, "loss": 0.1829, "step": 16566 }, { "epoch": 4.408461947844598, "grad_norm": 0.2904180884361267, "learning_rate": 7.11256604160213e-09, "loss": 0.1933, "step": 16567 }, { "epoch": 4.4087280468334225, "grad_norm": 0.2788953483104706, "learning_rate": 7.106248124274406e-09, "loss": 0.1849, "step": 16568 }, { "epoch": 4.408994145822246, "grad_norm": 0.2827506959438324, "learning_rate": 7.09993291084261e-09, "loss": 0.194, "step": 16569 }, { "epoch": 4.40926024481107, "grad_norm": 0.32376018166542053, "learning_rate": 7.093620401490541e-09, "loss": 0.1926, "step": 16570 }, { "epoch": 4.409526343799894, "grad_norm": 0.3165643811225891, "learning_rate": 7.087310596401963e-09, "loss": 0.1896, "step": 16571 }, { "epoch": 4.409792442788717, "grad_norm": 0.27533936500549316, "learning_rate": 7.081003495760518e-09, "loss": 0.1695, "step": 16572 }, { "epoch": 4.410058541777541, "grad_norm": 0.30197998881340027, "learning_rate": 7.0746990997498255e-09, "loss": 0.2007, "step": 16573 }, { "epoch": 4.410324640766365, "grad_norm": 0.2875036597251892, "learning_rate": 7.06839740855334e-09, "loss": 0.1836, "step": 16574 }, { "epoch": 4.410590739755189, "grad_norm": 0.2802007794380188, "learning_rate": 7.0620984223545235e-09, "loss": 0.1737, "step": 16575 }, { "epoch": 4.410856838744013, "grad_norm": 0.315940260887146, "learning_rate": 7.055802141336698e-09, "loss": 0.1789, "step": 16576 }, { "epoch": 4.411122937732837, "grad_norm": 0.33208680152893066, "learning_rate": 7.049508565683149e-09, "loss": 0.2086, "step": 16577 }, { "epoch": 4.41138903672166, "grad_norm": 0.2871570289134979, "learning_rate": 7.043217695577064e-09, "loss": 0.1891, "step": 16578 }, { "epoch": 4.411655135710484, "grad_norm": 0.29315185546875, "learning_rate": 7.036929531201552e-09, "loss": 0.1925, "step": 16579 }, { "epoch": 4.411921234699308, "grad_norm": 0.3020434081554413, "learning_rate": 7.030644072739644e-09, "loss": 0.2055, "step": 16580 }, { "epoch": 4.412187333688132, "grad_norm": 0.2934534549713135, "learning_rate": 7.024361320374295e-09, "loss": 0.1733, "step": 16581 }, { "epoch": 4.412453432676956, "grad_norm": 0.2832041084766388, "learning_rate": 7.018081274288401e-09, "loss": 0.1749, "step": 16582 }, { "epoch": 4.41271953166578, "grad_norm": 0.25801199674606323, "learning_rate": 7.011803934664717e-09, "loss": 0.1733, "step": 16583 }, { "epoch": 4.412985630654603, "grad_norm": 0.29381147027015686, "learning_rate": 7.005529301685975e-09, "loss": 0.1842, "step": 16584 }, { "epoch": 4.413251729643427, "grad_norm": 0.3357222080230713, "learning_rate": 6.999257375534817e-09, "loss": 0.205, "step": 16585 }, { "epoch": 4.413517828632251, "grad_norm": 0.332941472530365, "learning_rate": 6.99298815639382e-09, "loss": 0.2203, "step": 16586 }, { "epoch": 4.413783927621075, "grad_norm": 0.28888726234436035, "learning_rate": 6.986721644445437e-09, "loss": 0.1848, "step": 16587 }, { "epoch": 4.414050026609899, "grad_norm": 0.2999139428138733, "learning_rate": 6.9804578398720674e-09, "loss": 0.1897, "step": 16588 }, { "epoch": 4.414316125598723, "grad_norm": 0.3120892643928528, "learning_rate": 6.9741967428560645e-09, "loss": 0.1781, "step": 16589 }, { "epoch": 4.414582224587546, "grad_norm": 0.29688844084739685, "learning_rate": 6.967938353579639e-09, "loss": 0.1908, "step": 16590 }, { "epoch": 4.41484832357637, "grad_norm": 0.2889192998409271, "learning_rate": 6.961682672225e-09, "loss": 0.1838, "step": 16591 }, { "epoch": 4.415114422565194, "grad_norm": 0.3091980218887329, "learning_rate": 6.95542969897418e-09, "loss": 0.2027, "step": 16592 }, { "epoch": 4.4153805215540185, "grad_norm": 0.2979685962200165, "learning_rate": 6.949179434009222e-09, "loss": 0.1867, "step": 16593 }, { "epoch": 4.415646620542842, "grad_norm": 0.2935374677181244, "learning_rate": 6.942931877512048e-09, "loss": 0.1896, "step": 16594 }, { "epoch": 4.415912719531666, "grad_norm": 0.30270621180534363, "learning_rate": 6.936687029664501e-09, "loss": 0.2037, "step": 16595 }, { "epoch": 4.41617881852049, "grad_norm": 0.2722228765487671, "learning_rate": 6.9304448906483685e-09, "loss": 0.17, "step": 16596 }, { "epoch": 4.416444917509313, "grad_norm": 0.28984013199806213, "learning_rate": 6.924205460645316e-09, "loss": 0.1796, "step": 16597 }, { "epoch": 4.416711016498137, "grad_norm": 0.2987452447414398, "learning_rate": 6.917968739836988e-09, "loss": 0.1998, "step": 16598 }, { "epoch": 4.4169771154869615, "grad_norm": 0.2922700047492981, "learning_rate": 6.911734728404894e-09, "loss": 0.1962, "step": 16599 }, { "epoch": 4.417243214475785, "grad_norm": 0.2842026948928833, "learning_rate": 6.905503426530523e-09, "loss": 0.1851, "step": 16600 }, { "epoch": 4.417509313464609, "grad_norm": 0.31535452604293823, "learning_rate": 6.899274834395219e-09, "loss": 0.1927, "step": 16601 }, { "epoch": 4.417775412453433, "grad_norm": 0.2923545241355896, "learning_rate": 6.89304895218028e-09, "loss": 0.1976, "step": 16602 }, { "epoch": 4.418041511442256, "grad_norm": 0.2801973521709442, "learning_rate": 6.8868257800669405e-09, "loss": 0.1766, "step": 16603 }, { "epoch": 4.41830761043108, "grad_norm": 0.2923032343387604, "learning_rate": 6.8806053182363435e-09, "loss": 0.1889, "step": 16604 }, { "epoch": 4.4185737094199045, "grad_norm": 0.2886609733104706, "learning_rate": 6.8743875668695325e-09, "loss": 0.1898, "step": 16605 }, { "epoch": 4.418839808408728, "grad_norm": 0.2934746742248535, "learning_rate": 6.868172526147531e-09, "loss": 0.186, "step": 16606 }, { "epoch": 4.419105907397552, "grad_norm": 0.29127010703086853, "learning_rate": 6.861960196251182e-09, "loss": 0.1872, "step": 16607 }, { "epoch": 4.419372006386376, "grad_norm": 0.3075162470340729, "learning_rate": 6.855750577361352e-09, "loss": 0.2064, "step": 16608 }, { "epoch": 4.419638105375199, "grad_norm": 0.2848946154117584, "learning_rate": 6.849543669658786e-09, "loss": 0.1819, "step": 16609 }, { "epoch": 4.419904204364023, "grad_norm": 0.30261075496673584, "learning_rate": 6.843339473324128e-09, "loss": 0.1996, "step": 16610 }, { "epoch": 4.420170303352847, "grad_norm": 0.3101196885108948, "learning_rate": 6.837137988537978e-09, "loss": 0.1845, "step": 16611 }, { "epoch": 4.420436402341672, "grad_norm": 0.32139596343040466, "learning_rate": 6.830939215480846e-09, "loss": 0.1947, "step": 16612 }, { "epoch": 4.420702501330495, "grad_norm": 0.2807718813419342, "learning_rate": 6.824743154333157e-09, "loss": 0.1857, "step": 16613 }, { "epoch": 4.420968600319319, "grad_norm": 0.3078245222568512, "learning_rate": 6.818549805275276e-09, "loss": 0.1915, "step": 16614 }, { "epoch": 4.421234699308142, "grad_norm": 0.29524752497673035, "learning_rate": 6.812359168487469e-09, "loss": 0.1885, "step": 16615 }, { "epoch": 4.421500798296966, "grad_norm": 0.3028770089149475, "learning_rate": 6.806171244149916e-09, "loss": 0.2033, "step": 16616 }, { "epoch": 4.42176689728579, "grad_norm": 0.3532603979110718, "learning_rate": 6.799986032442751e-09, "loss": 0.193, "step": 16617 }, { "epoch": 4.4220329962746145, "grad_norm": 0.3066873550415039, "learning_rate": 6.793803533545994e-09, "loss": 0.2002, "step": 16618 }, { "epoch": 4.422299095263438, "grad_norm": 0.3039869964122772, "learning_rate": 6.787623747639637e-09, "loss": 0.1832, "step": 16619 }, { "epoch": 4.422565194252262, "grad_norm": 0.2841264307498932, "learning_rate": 6.781446674903501e-09, "loss": 0.1824, "step": 16620 }, { "epoch": 4.422831293241086, "grad_norm": 0.3023211658000946, "learning_rate": 6.775272315517422e-09, "loss": 0.1857, "step": 16621 }, { "epoch": 4.423097392229909, "grad_norm": 0.297610878944397, "learning_rate": 6.7691006696611095e-09, "loss": 0.1856, "step": 16622 }, { "epoch": 4.423363491218733, "grad_norm": 0.3006495237350464, "learning_rate": 6.762931737514199e-09, "loss": 0.1944, "step": 16623 }, { "epoch": 4.4236295902075575, "grad_norm": 0.30107131600379944, "learning_rate": 6.756765519256258e-09, "loss": 0.1983, "step": 16624 }, { "epoch": 4.423895689196381, "grad_norm": 0.3045599162578583, "learning_rate": 6.750602015066776e-09, "loss": 0.1986, "step": 16625 }, { "epoch": 4.424161788185205, "grad_norm": 0.2955782115459442, "learning_rate": 6.7444412251251635e-09, "loss": 0.1873, "step": 16626 }, { "epoch": 4.424427887174029, "grad_norm": 0.2896685302257538, "learning_rate": 6.738283149610713e-09, "loss": 0.1965, "step": 16627 }, { "epoch": 4.424693986162852, "grad_norm": 0.29362472891807556, "learning_rate": 6.7321277887026796e-09, "loss": 0.1849, "step": 16628 }, { "epoch": 4.424960085151676, "grad_norm": 0.263202041387558, "learning_rate": 6.725975142580265e-09, "loss": 0.1723, "step": 16629 }, { "epoch": 4.4252261841405005, "grad_norm": 0.38319534063339233, "learning_rate": 6.719825211422514e-09, "loss": 0.1999, "step": 16630 }, { "epoch": 4.425492283129324, "grad_norm": 0.324553906917572, "learning_rate": 6.713677995408451e-09, "loss": 0.1866, "step": 16631 }, { "epoch": 4.425758382118148, "grad_norm": 0.29944419860839844, "learning_rate": 6.707533494717011e-09, "loss": 0.1694, "step": 16632 }, { "epoch": 4.426024481106972, "grad_norm": 0.29620328545570374, "learning_rate": 6.701391709527027e-09, "loss": 0.1946, "step": 16633 }, { "epoch": 4.426290580095795, "grad_norm": 0.3428947627544403, "learning_rate": 6.695252640017291e-09, "loss": 0.1987, "step": 16634 }, { "epoch": 4.426556679084619, "grad_norm": 0.2792493999004364, "learning_rate": 6.689116286366492e-09, "loss": 0.1985, "step": 16635 }, { "epoch": 4.4268227780734435, "grad_norm": 0.29810386896133423, "learning_rate": 6.6829826487532325e-09, "loss": 0.1908, "step": 16636 }, { "epoch": 4.427088877062268, "grad_norm": 0.2976628839969635, "learning_rate": 6.67685172735607e-09, "loss": 0.1994, "step": 16637 }, { "epoch": 4.427354976051091, "grad_norm": 0.2929835617542267, "learning_rate": 6.670723522353427e-09, "loss": 0.1933, "step": 16638 }, { "epoch": 4.427621075039915, "grad_norm": 0.29871872067451477, "learning_rate": 6.6645980339236965e-09, "loss": 0.1824, "step": 16639 }, { "epoch": 4.427887174028739, "grad_norm": 0.30400848388671875, "learning_rate": 6.658475262245167e-09, "loss": 0.1902, "step": 16640 }, { "epoch": 4.428153273017562, "grad_norm": 0.2795804738998413, "learning_rate": 6.652355207496074e-09, "loss": 0.1847, "step": 16641 }, { "epoch": 4.4284193720063865, "grad_norm": 0.31521502137184143, "learning_rate": 6.6462378698545426e-09, "loss": 0.1871, "step": 16642 }, { "epoch": 4.428685470995211, "grad_norm": 0.28692418336868286, "learning_rate": 6.64012324949863e-09, "loss": 0.1793, "step": 16643 }, { "epoch": 4.428951569984034, "grad_norm": 0.2889063060283661, "learning_rate": 6.634011346606327e-09, "loss": 0.18, "step": 16644 }, { "epoch": 4.429217668972858, "grad_norm": 0.3113390803337097, "learning_rate": 6.627902161355547e-09, "loss": 0.2023, "step": 16645 }, { "epoch": 4.429483767961682, "grad_norm": 0.32393455505371094, "learning_rate": 6.621795693924081e-09, "loss": 0.2112, "step": 16646 }, { "epoch": 4.429749866950505, "grad_norm": 0.30629870295524597, "learning_rate": 6.615691944489687e-09, "loss": 0.2016, "step": 16647 }, { "epoch": 4.4300159659393294, "grad_norm": 0.32419633865356445, "learning_rate": 6.609590913230056e-09, "loss": 0.2074, "step": 16648 }, { "epoch": 4.430282064928154, "grad_norm": 0.28357455134391785, "learning_rate": 6.603492600322724e-09, "loss": 0.1962, "step": 16649 }, { "epoch": 4.430548163916977, "grad_norm": 0.29188457131385803, "learning_rate": 6.597397005945227e-09, "loss": 0.1823, "step": 16650 }, { "epoch": 4.430814262905801, "grad_norm": 0.3063153326511383, "learning_rate": 6.591304130274977e-09, "loss": 0.2133, "step": 16651 }, { "epoch": 4.431080361894625, "grad_norm": 0.3042115867137909, "learning_rate": 6.585213973489334e-09, "loss": 0.1935, "step": 16652 }, { "epoch": 4.431346460883448, "grad_norm": 0.34728506207466125, "learning_rate": 6.5791265357655554e-09, "loss": 0.2046, "step": 16653 }, { "epoch": 4.431612559872272, "grad_norm": 0.29711464047431946, "learning_rate": 6.573041817280856e-09, "loss": 0.1964, "step": 16654 }, { "epoch": 4.4318786588610966, "grad_norm": 0.2968237102031708, "learning_rate": 6.566959818212326e-09, "loss": 0.191, "step": 16655 }, { "epoch": 4.43214475784992, "grad_norm": 0.3133593499660492, "learning_rate": 6.560880538737001e-09, "loss": 0.2114, "step": 16656 }, { "epoch": 4.432410856838744, "grad_norm": 0.30855172872543335, "learning_rate": 6.5548039790318195e-09, "loss": 0.1946, "step": 16657 }, { "epoch": 4.432676955827568, "grad_norm": 0.2971022427082062, "learning_rate": 6.548730139273662e-09, "loss": 0.1743, "step": 16658 }, { "epoch": 4.432943054816391, "grad_norm": 0.3056495785713196, "learning_rate": 6.5426590196393406e-09, "loss": 0.1841, "step": 16659 }, { "epoch": 4.433209153805215, "grad_norm": 0.3090517222881317, "learning_rate": 6.536590620305549e-09, "loss": 0.2091, "step": 16660 }, { "epoch": 4.4334752527940395, "grad_norm": 0.2889688014984131, "learning_rate": 6.530524941448923e-09, "loss": 0.1815, "step": 16661 }, { "epoch": 4.433741351782864, "grad_norm": 0.29067227244377136, "learning_rate": 6.524461983246033e-09, "loss": 0.1899, "step": 16662 }, { "epoch": 4.434007450771687, "grad_norm": 0.49194520711898804, "learning_rate": 6.518401745873347e-09, "loss": 0.1998, "step": 16663 }, { "epoch": 4.434273549760511, "grad_norm": 0.2645859122276306, "learning_rate": 6.5123442295072716e-09, "loss": 0.1803, "step": 16664 }, { "epoch": 4.434539648749335, "grad_norm": 0.31603312492370605, "learning_rate": 6.506289434324108e-09, "loss": 0.1982, "step": 16665 }, { "epoch": 4.434805747738158, "grad_norm": 0.2736424207687378, "learning_rate": 6.500237360500105e-09, "loss": 0.1798, "step": 16666 }, { "epoch": 4.4350718467269825, "grad_norm": 0.3096180558204651, "learning_rate": 6.494188008211421e-09, "loss": 0.2013, "step": 16667 }, { "epoch": 4.435337945715807, "grad_norm": 0.29402992129325867, "learning_rate": 6.48814137763416e-09, "loss": 0.175, "step": 16668 }, { "epoch": 4.43560404470463, "grad_norm": 0.3196949064731598, "learning_rate": 6.482097468944281e-09, "loss": 0.2048, "step": 16669 }, { "epoch": 4.435870143693454, "grad_norm": 0.2824499011039734, "learning_rate": 6.476056282317732e-09, "loss": 0.1757, "step": 16670 }, { "epoch": 4.436136242682278, "grad_norm": 0.2755313217639923, "learning_rate": 6.470017817930351e-09, "loss": 0.186, "step": 16671 }, { "epoch": 4.436402341671101, "grad_norm": 0.28389638662338257, "learning_rate": 6.463982075957896e-09, "loss": 0.1644, "step": 16672 }, { "epoch": 4.4366684406599255, "grad_norm": 0.3400058150291443, "learning_rate": 6.457949056576084e-09, "loss": 0.2059, "step": 16673 }, { "epoch": 4.43693453964875, "grad_norm": 0.3113766610622406, "learning_rate": 6.451918759960473e-09, "loss": 0.1989, "step": 16674 }, { "epoch": 4.437200638637573, "grad_norm": 0.2488037347793579, "learning_rate": 6.445891186286623e-09, "loss": 0.1719, "step": 16675 }, { "epoch": 4.437466737626397, "grad_norm": 0.3018220067024231, "learning_rate": 6.4398663357299595e-09, "loss": 0.1824, "step": 16676 }, { "epoch": 4.437732836615221, "grad_norm": 0.29701632261276245, "learning_rate": 6.433844208465866e-09, "loss": 0.1835, "step": 16677 }, { "epoch": 4.437998935604044, "grad_norm": 0.3048081398010254, "learning_rate": 6.4278248046696125e-09, "loss": 0.1805, "step": 16678 }, { "epoch": 4.4382650345928685, "grad_norm": 0.28247007727622986, "learning_rate": 6.421808124516437e-09, "loss": 0.1822, "step": 16679 }, { "epoch": 4.438531133581693, "grad_norm": 0.29212015867233276, "learning_rate": 6.415794168181443e-09, "loss": 0.1941, "step": 16680 }, { "epoch": 4.438797232570516, "grad_norm": 0.3100220561027527, "learning_rate": 6.409782935839703e-09, "loss": 0.2001, "step": 16681 }, { "epoch": 4.43906333155934, "grad_norm": 0.29152607917785645, "learning_rate": 6.403774427666186e-09, "loss": 0.1764, "step": 16682 }, { "epoch": 4.439329430548164, "grad_norm": 0.31073206663131714, "learning_rate": 6.3977686438357544e-09, "loss": 0.1978, "step": 16683 }, { "epoch": 4.439595529536987, "grad_norm": 0.31192100048065186, "learning_rate": 6.3917655845232456e-09, "loss": 0.1949, "step": 16684 }, { "epoch": 4.4398616285258115, "grad_norm": 0.30823227763175964, "learning_rate": 6.3857652499033966e-09, "loss": 0.1926, "step": 16685 }, { "epoch": 4.440127727514636, "grad_norm": 0.31126436591148376, "learning_rate": 6.379767640150846e-09, "loss": 0.1892, "step": 16686 }, { "epoch": 4.44039382650346, "grad_norm": 0.31447193026542664, "learning_rate": 6.373772755440199e-09, "loss": 0.1884, "step": 16687 }, { "epoch": 4.440659925492283, "grad_norm": 0.2890680432319641, "learning_rate": 6.367780595945904e-09, "loss": 0.1758, "step": 16688 }, { "epoch": 4.440926024481107, "grad_norm": 0.2940415143966675, "learning_rate": 6.361791161842411e-09, "loss": 0.1781, "step": 16689 }, { "epoch": 4.441192123469931, "grad_norm": 0.312480628490448, "learning_rate": 6.355804453304048e-09, "loss": 0.2025, "step": 16690 }, { "epoch": 4.441458222458754, "grad_norm": 0.2857486307621002, "learning_rate": 6.349820470505085e-09, "loss": 0.1888, "step": 16691 }, { "epoch": 4.441724321447579, "grad_norm": 0.3091951012611389, "learning_rate": 6.343839213619673e-09, "loss": 0.2008, "step": 16692 }, { "epoch": 4.441990420436403, "grad_norm": 0.3259669244289398, "learning_rate": 6.3378606828219275e-09, "loss": 0.1968, "step": 16693 }, { "epoch": 4.442256519425226, "grad_norm": 0.2936009168624878, "learning_rate": 6.3318848782858536e-09, "loss": 0.1838, "step": 16694 }, { "epoch": 4.44252261841405, "grad_norm": 0.30684101581573486, "learning_rate": 6.325911800185424e-09, "loss": 0.1846, "step": 16695 }, { "epoch": 4.442788717402874, "grad_norm": 0.3172135651111603, "learning_rate": 6.319941448694466e-09, "loss": 0.2087, "step": 16696 }, { "epoch": 4.443054816391697, "grad_norm": 0.2954677939414978, "learning_rate": 6.313973823986773e-09, "loss": 0.1912, "step": 16697 }, { "epoch": 4.4433209153805215, "grad_norm": 0.34774941205978394, "learning_rate": 6.3080089262360635e-09, "loss": 0.2158, "step": 16698 }, { "epoch": 4.443587014369346, "grad_norm": 0.29661285877227783, "learning_rate": 6.302046755615931e-09, "loss": 0.186, "step": 16699 }, { "epoch": 4.443853113358169, "grad_norm": 0.2953079044818878, "learning_rate": 6.296087312299958e-09, "loss": 0.1937, "step": 16700 }, { "epoch": 4.444119212346993, "grad_norm": 0.283582866191864, "learning_rate": 6.290130596461563e-09, "loss": 0.1921, "step": 16701 }, { "epoch": 4.444385311335817, "grad_norm": 0.31091058254241943, "learning_rate": 6.28417660827415e-09, "loss": 0.1916, "step": 16702 }, { "epoch": 4.44465141032464, "grad_norm": 0.3030044138431549, "learning_rate": 6.2782253479110256e-09, "loss": 0.201, "step": 16703 }, { "epoch": 4.4449175093134645, "grad_norm": 0.2939567565917969, "learning_rate": 6.272276815545419e-09, "loss": 0.1821, "step": 16704 }, { "epoch": 4.445183608302289, "grad_norm": 0.2791639268398285, "learning_rate": 6.2663310113504674e-09, "loss": 0.1947, "step": 16705 }, { "epoch": 4.445449707291112, "grad_norm": 0.27969327569007874, "learning_rate": 6.260387935499245e-09, "loss": 0.1742, "step": 16706 }, { "epoch": 4.445715806279936, "grad_norm": 0.28564491868019104, "learning_rate": 6.254447588164757e-09, "loss": 0.1841, "step": 16707 }, { "epoch": 4.44598190526876, "grad_norm": 0.310406893491745, "learning_rate": 6.248509969519877e-09, "loss": 0.1899, "step": 16708 }, { "epoch": 4.446248004257583, "grad_norm": 0.3332180380821228, "learning_rate": 6.242575079737444e-09, "loss": 0.1983, "step": 16709 }, { "epoch": 4.4465141032464075, "grad_norm": 0.29858168959617615, "learning_rate": 6.23664291899022e-09, "loss": 0.1898, "step": 16710 }, { "epoch": 4.446780202235232, "grad_norm": 0.31675025820732117, "learning_rate": 6.230713487450867e-09, "loss": 0.1861, "step": 16711 }, { "epoch": 4.447046301224056, "grad_norm": 0.31882140040397644, "learning_rate": 6.224786785291969e-09, "loss": 0.198, "step": 16712 }, { "epoch": 4.447312400212879, "grad_norm": 0.28882619738578796, "learning_rate": 6.218862812686032e-09, "loss": 0.2022, "step": 16713 }, { "epoch": 4.447578499201703, "grad_norm": 0.30770933628082275, "learning_rate": 6.2129415698055075e-09, "loss": 0.1951, "step": 16714 }, { "epoch": 4.447844598190527, "grad_norm": 0.3181227743625641, "learning_rate": 6.207023056822736e-09, "loss": 0.2115, "step": 16715 }, { "epoch": 4.4481106971793505, "grad_norm": 0.2670800983905792, "learning_rate": 6.20110727391e-09, "loss": 0.1615, "step": 16716 }, { "epoch": 4.448376796168175, "grad_norm": 0.2838623523712158, "learning_rate": 6.1951942212394745e-09, "loss": 0.191, "step": 16717 }, { "epoch": 4.448642895156999, "grad_norm": 0.3138120472431183, "learning_rate": 6.189283898983311e-09, "loss": 0.1905, "step": 16718 }, { "epoch": 4.448908994145822, "grad_norm": 0.2778444290161133, "learning_rate": 6.1833763073134924e-09, "loss": 0.1681, "step": 16719 }, { "epoch": 4.449175093134646, "grad_norm": 0.3109523057937622, "learning_rate": 6.177471446401994e-09, "loss": 0.1944, "step": 16720 }, { "epoch": 4.44944119212347, "grad_norm": 0.3064553439617157, "learning_rate": 6.1715693164207e-09, "loss": 0.1867, "step": 16721 }, { "epoch": 4.4497072911122935, "grad_norm": 0.2835722267627716, "learning_rate": 6.165669917541394e-09, "loss": 0.1951, "step": 16722 }, { "epoch": 4.449973390101118, "grad_norm": 0.2677622437477112, "learning_rate": 6.1597732499358066e-09, "loss": 0.1684, "step": 16723 }, { "epoch": 4.450239489089942, "grad_norm": 0.31559646129608154, "learning_rate": 6.153879313775556e-09, "loss": 0.1933, "step": 16724 }, { "epoch": 4.450505588078765, "grad_norm": 0.32632434368133545, "learning_rate": 6.147988109232216e-09, "loss": 0.213, "step": 16725 }, { "epoch": 4.450771687067589, "grad_norm": 0.2930199205875397, "learning_rate": 6.1420996364772605e-09, "loss": 0.1923, "step": 16726 }, { "epoch": 4.451037786056413, "grad_norm": 0.2874249219894409, "learning_rate": 6.136213895682096e-09, "loss": 0.1943, "step": 16727 }, { "epoch": 4.451303885045236, "grad_norm": 0.31101012229919434, "learning_rate": 6.13033088701802e-09, "loss": 0.1752, "step": 16728 }, { "epoch": 4.451569984034061, "grad_norm": 0.29334691166877747, "learning_rate": 6.124450610656273e-09, "loss": 0.1731, "step": 16729 }, { "epoch": 4.451836083022885, "grad_norm": 0.28933873772621155, "learning_rate": 6.11857306676804e-09, "loss": 0.1926, "step": 16730 }, { "epoch": 4.452102182011709, "grad_norm": 0.2888033390045166, "learning_rate": 6.112698255524373e-09, "loss": 0.184, "step": 16731 }, { "epoch": 4.452368281000532, "grad_norm": 0.329263299703598, "learning_rate": 6.10682617709628e-09, "loss": 0.2049, "step": 16732 }, { "epoch": 4.452634379989356, "grad_norm": 0.28963613510131836, "learning_rate": 6.100956831654691e-09, "loss": 0.1849, "step": 16733 }, { "epoch": 4.452900478978179, "grad_norm": 0.28455400466918945, "learning_rate": 6.0950902193704465e-09, "loss": 0.174, "step": 16734 }, { "epoch": 4.4531665779670035, "grad_norm": 0.28370654582977295, "learning_rate": 6.0892263404142995e-09, "loss": 0.1959, "step": 16735 }, { "epoch": 4.453432676955828, "grad_norm": 0.3066006302833557, "learning_rate": 6.0833651949569465e-09, "loss": 0.1846, "step": 16736 }, { "epoch": 4.453698775944652, "grad_norm": 0.3057693839073181, "learning_rate": 6.077506783168973e-09, "loss": 0.1988, "step": 16737 }, { "epoch": 4.453964874933475, "grad_norm": 0.29057052731513977, "learning_rate": 6.071651105220899e-09, "loss": 0.1762, "step": 16738 }, { "epoch": 4.454230973922299, "grad_norm": 0.29550275206565857, "learning_rate": 6.065798161283187e-09, "loss": 0.1774, "step": 16739 }, { "epoch": 4.454497072911123, "grad_norm": 0.27147001028060913, "learning_rate": 6.05994795152619e-09, "loss": 0.1799, "step": 16740 }, { "epoch": 4.4547631718999465, "grad_norm": 0.2907756567001343, "learning_rate": 6.054100476120194e-09, "loss": 0.1846, "step": 16741 }, { "epoch": 4.455029270888771, "grad_norm": 0.31705135107040405, "learning_rate": 6.048255735235408e-09, "loss": 0.1914, "step": 16742 }, { "epoch": 4.455295369877595, "grad_norm": 0.3113205134868622, "learning_rate": 6.042413729041962e-09, "loss": 0.207, "step": 16743 }, { "epoch": 4.455561468866418, "grad_norm": 0.29892536997795105, "learning_rate": 6.036574457709886e-09, "loss": 0.195, "step": 16744 }, { "epoch": 4.455827567855242, "grad_norm": 0.2903590798377991, "learning_rate": 6.030737921409168e-09, "loss": 0.1782, "step": 16745 }, { "epoch": 4.456093666844066, "grad_norm": 0.3064362406730652, "learning_rate": 6.024904120309671e-09, "loss": 0.1724, "step": 16746 }, { "epoch": 4.4563597658328895, "grad_norm": 0.30426225066185, "learning_rate": 6.019073054581203e-09, "loss": 0.1832, "step": 16747 }, { "epoch": 4.456625864821714, "grad_norm": 0.28309810161590576, "learning_rate": 6.013244724393496e-09, "loss": 0.1834, "step": 16748 }, { "epoch": 4.456891963810538, "grad_norm": 0.31065496802330017, "learning_rate": 6.007419129916224e-09, "loss": 0.1908, "step": 16749 }, { "epoch": 4.457158062799361, "grad_norm": 0.3093542456626892, "learning_rate": 6.00159627131892e-09, "loss": 0.1925, "step": 16750 }, { "epoch": 4.457424161788185, "grad_norm": 0.2784349322319031, "learning_rate": 5.995776148771081e-09, "loss": 0.1837, "step": 16751 }, { "epoch": 4.457690260777009, "grad_norm": 0.2819385230541229, "learning_rate": 5.989958762442115e-09, "loss": 0.1714, "step": 16752 }, { "epoch": 4.4579563597658325, "grad_norm": 0.2920096516609192, "learning_rate": 5.984144112501355e-09, "loss": 0.1926, "step": 16753 }, { "epoch": 4.458222458754657, "grad_norm": 0.30142927169799805, "learning_rate": 5.978332199118075e-09, "loss": 0.1991, "step": 16754 }, { "epoch": 4.458488557743481, "grad_norm": 0.2917122542858124, "learning_rate": 5.972523022461396e-09, "loss": 0.2037, "step": 16755 }, { "epoch": 4.458754656732305, "grad_norm": 0.29200050234794617, "learning_rate": 5.966716582700426e-09, "loss": 0.1904, "step": 16756 }, { "epoch": 4.459020755721128, "grad_norm": 0.3087920844554901, "learning_rate": 5.960912880004198e-09, "loss": 0.1868, "step": 16757 }, { "epoch": 4.459286854709952, "grad_norm": 0.6350594162940979, "learning_rate": 5.9551119145416105e-09, "loss": 0.185, "step": 16758 }, { "epoch": 4.459552953698776, "grad_norm": 0.29628506302833557, "learning_rate": 5.9493136864815386e-09, "loss": 0.1935, "step": 16759 }, { "epoch": 4.4598190526876, "grad_norm": 0.29346171021461487, "learning_rate": 5.943518195992747e-09, "loss": 0.1828, "step": 16760 }, { "epoch": 4.460085151676424, "grad_norm": 0.31999847292900085, "learning_rate": 5.937725443243913e-09, "loss": 0.1913, "step": 16761 }, { "epoch": 4.460351250665248, "grad_norm": 0.3186878263950348, "learning_rate": 5.931935428403678e-09, "loss": 0.1882, "step": 16762 }, { "epoch": 4.460617349654071, "grad_norm": 0.31027135252952576, "learning_rate": 5.926148151640564e-09, "loss": 0.2061, "step": 16763 }, { "epoch": 4.460883448642895, "grad_norm": 0.2774486541748047, "learning_rate": 5.920363613123003e-09, "loss": 0.195, "step": 16764 }, { "epoch": 4.461149547631719, "grad_norm": 0.28738752007484436, "learning_rate": 5.914581813019371e-09, "loss": 0.1842, "step": 16765 }, { "epoch": 4.461415646620543, "grad_norm": 0.31985723972320557, "learning_rate": 5.908802751497988e-09, "loss": 0.1925, "step": 16766 }, { "epoch": 4.461681745609367, "grad_norm": 0.29859641194343567, "learning_rate": 5.903026428727042e-09, "loss": 0.1872, "step": 16767 }, { "epoch": 4.461947844598191, "grad_norm": 0.3178517520427704, "learning_rate": 5.897252844874678e-09, "loss": 0.1927, "step": 16768 }, { "epoch": 4.462213943587014, "grad_norm": 0.2966476380825043, "learning_rate": 5.891482000108971e-09, "loss": 0.1884, "step": 16769 }, { "epoch": 4.462480042575838, "grad_norm": 0.3965606987476349, "learning_rate": 5.885713894597855e-09, "loss": 0.1826, "step": 16770 }, { "epoch": 4.462746141564662, "grad_norm": 0.27863481640815735, "learning_rate": 5.879948528509238e-09, "loss": 0.1703, "step": 16771 }, { "epoch": 4.4630122405534856, "grad_norm": 0.2753538191318512, "learning_rate": 5.874185902010953e-09, "loss": 0.1843, "step": 16772 }, { "epoch": 4.46327833954231, "grad_norm": 0.2845134139060974, "learning_rate": 5.868426015270711e-09, "loss": 0.1713, "step": 16773 }, { "epoch": 4.463544438531134, "grad_norm": 0.3027113974094391, "learning_rate": 5.862668868456167e-09, "loss": 0.2003, "step": 16774 }, { "epoch": 4.463810537519957, "grad_norm": 0.3109835088253021, "learning_rate": 5.8569144617349185e-09, "loss": 0.2017, "step": 16775 }, { "epoch": 4.464076636508781, "grad_norm": 0.30218419432640076, "learning_rate": 5.851162795274445e-09, "loss": 0.1911, "step": 16776 }, { "epoch": 4.464342735497605, "grad_norm": 0.2969152629375458, "learning_rate": 5.8454138692421665e-09, "loss": 0.1867, "step": 16777 }, { "epoch": 4.4646088344864285, "grad_norm": 0.3156706988811493, "learning_rate": 5.839667683805427e-09, "loss": 0.2116, "step": 16778 }, { "epoch": 4.464874933475253, "grad_norm": 0.2983585596084595, "learning_rate": 5.8339242391314714e-09, "loss": 0.196, "step": 16779 }, { "epoch": 4.465141032464077, "grad_norm": 0.30674225091934204, "learning_rate": 5.828183535387487e-09, "loss": 0.2076, "step": 16780 }, { "epoch": 4.465407131452901, "grad_norm": 0.30028030276298523, "learning_rate": 5.822445572740575e-09, "loss": 0.199, "step": 16781 }, { "epoch": 4.465673230441724, "grad_norm": 0.3088424503803253, "learning_rate": 5.816710351357723e-09, "loss": 0.1971, "step": 16782 }, { "epoch": 4.465939329430548, "grad_norm": 0.3125987946987152, "learning_rate": 5.810977871405898e-09, "loss": 0.1808, "step": 16783 }, { "epoch": 4.466205428419372, "grad_norm": 0.2969750165939331, "learning_rate": 5.805248133051943e-09, "loss": 0.187, "step": 16784 }, { "epoch": 4.466471527408196, "grad_norm": 0.33363744616508484, "learning_rate": 5.799521136462637e-09, "loss": 0.1684, "step": 16785 }, { "epoch": 4.46673762639702, "grad_norm": 0.28734949231147766, "learning_rate": 5.793796881804691e-09, "loss": 0.1762, "step": 16786 }, { "epoch": 4.467003725385844, "grad_norm": 0.31673964858055115, "learning_rate": 5.788075369244705e-09, "loss": 0.193, "step": 16787 }, { "epoch": 4.467269824374667, "grad_norm": 0.3005916178226471, "learning_rate": 5.782356598949245e-09, "loss": 0.1863, "step": 16788 }, { "epoch": 4.467535923363491, "grad_norm": 0.27698057889938354, "learning_rate": 5.776640571084734e-09, "loss": 0.1813, "step": 16789 }, { "epoch": 4.467802022352315, "grad_norm": 0.29594358801841736, "learning_rate": 5.770927285817573e-09, "loss": 0.1994, "step": 16790 }, { "epoch": 4.468068121341139, "grad_norm": 0.2968551814556122, "learning_rate": 5.765216743314061e-09, "loss": 0.1949, "step": 16791 }, { "epoch": 4.468334220329963, "grad_norm": 0.3119654953479767, "learning_rate": 5.759508943740399e-09, "loss": 0.2176, "step": 16792 }, { "epoch": 4.468600319318787, "grad_norm": 0.32146963477134705, "learning_rate": 5.753803887262742e-09, "loss": 0.1904, "step": 16793 }, { "epoch": 4.46886641830761, "grad_norm": 0.2988821566104889, "learning_rate": 5.7481015740471484e-09, "loss": 0.1921, "step": 16794 }, { "epoch": 4.469132517296434, "grad_norm": 0.3324722945690155, "learning_rate": 5.742402004259594e-09, "loss": 0.2, "step": 16795 }, { "epoch": 4.469398616285258, "grad_norm": 0.2967604100704193, "learning_rate": 5.736705178065982e-09, "loss": 0.1861, "step": 16796 }, { "epoch": 4.469664715274082, "grad_norm": 0.3484119474887848, "learning_rate": 5.731011095632121e-09, "loss": 0.1804, "step": 16797 }, { "epoch": 4.469930814262906, "grad_norm": 0.30097389221191406, "learning_rate": 5.72531975712377e-09, "loss": 0.1865, "step": 16798 }, { "epoch": 4.47019691325173, "grad_norm": 0.28900790214538574, "learning_rate": 5.719631162706584e-09, "loss": 0.2006, "step": 16799 }, { "epoch": 4.470463012240553, "grad_norm": 0.31066152453422546, "learning_rate": 5.713945312546131e-09, "loss": 0.2107, "step": 16800 }, { "epoch": 4.470729111229377, "grad_norm": 0.31483545899391174, "learning_rate": 5.708262206807912e-09, "loss": 0.1952, "step": 16801 }, { "epoch": 4.470995210218201, "grad_norm": 0.2803029417991638, "learning_rate": 5.702581845657362e-09, "loss": 0.1733, "step": 16802 }, { "epoch": 4.471261309207025, "grad_norm": 0.2963142693042755, "learning_rate": 5.696904229259813e-09, "loss": 0.1872, "step": 16803 }, { "epoch": 4.471527408195849, "grad_norm": 0.2818208336830139, "learning_rate": 5.691229357780525e-09, "loss": 0.1914, "step": 16804 }, { "epoch": 4.471793507184673, "grad_norm": 0.2896735966205597, "learning_rate": 5.685557231384674e-09, "loss": 0.1777, "step": 16805 }, { "epoch": 4.472059606173497, "grad_norm": 0.30962997674942017, "learning_rate": 5.6798878502373744e-09, "loss": 0.1943, "step": 16806 }, { "epoch": 4.47232570516232, "grad_norm": 0.3144949674606323, "learning_rate": 5.674221214503639e-09, "loss": 0.1892, "step": 16807 }, { "epoch": 4.472591804151144, "grad_norm": 0.3261994421482086, "learning_rate": 5.668557324348422e-09, "loss": 0.1762, "step": 16808 }, { "epoch": 4.4728579031399684, "grad_norm": 0.31596213579177856, "learning_rate": 5.6628961799365605e-09, "loss": 0.2017, "step": 16809 }, { "epoch": 4.473124002128792, "grad_norm": 0.2839461863040924, "learning_rate": 5.6572377814328446e-09, "loss": 0.1805, "step": 16810 }, { "epoch": 4.473390101117616, "grad_norm": 0.3180898427963257, "learning_rate": 5.651582129001986e-09, "loss": 0.2014, "step": 16811 }, { "epoch": 4.47365620010644, "grad_norm": 0.2932920753955841, "learning_rate": 5.645929222808598e-09, "loss": 0.1899, "step": 16812 }, { "epoch": 4.473922299095263, "grad_norm": 0.2653487026691437, "learning_rate": 5.640279063017217e-09, "loss": 0.1581, "step": 16813 }, { "epoch": 4.474188398084087, "grad_norm": 0.3126658499240875, "learning_rate": 5.634631649792299e-09, "loss": 0.1988, "step": 16814 }, { "epoch": 4.474454497072911, "grad_norm": 0.28678590059280396, "learning_rate": 5.628986983298245e-09, "loss": 0.1865, "step": 16815 }, { "epoch": 4.474720596061735, "grad_norm": 0.29562148451805115, "learning_rate": 5.6233450636993584e-09, "loss": 0.1986, "step": 16816 }, { "epoch": 4.474986695050559, "grad_norm": 0.32948872447013855, "learning_rate": 5.617705891159852e-09, "loss": 0.2064, "step": 16817 }, { "epoch": 4.475252794039383, "grad_norm": 0.2884546220302582, "learning_rate": 5.6120694658438496e-09, "loss": 0.1952, "step": 16818 }, { "epoch": 4.475518893028206, "grad_norm": 0.2971055805683136, "learning_rate": 5.606435787915442e-09, "loss": 0.1895, "step": 16819 }, { "epoch": 4.47578499201703, "grad_norm": 0.2912813425064087, "learning_rate": 5.600804857538588e-09, "loss": 0.184, "step": 16820 }, { "epoch": 4.476051091005854, "grad_norm": 0.26760828495025635, "learning_rate": 5.5951766748772e-09, "loss": 0.1747, "step": 16821 }, { "epoch": 4.476317189994678, "grad_norm": 0.2999931871891022, "learning_rate": 5.589551240095103e-09, "loss": 0.1899, "step": 16822 }, { "epoch": 4.476583288983502, "grad_norm": 0.28988030552864075, "learning_rate": 5.583928553356032e-09, "loss": 0.1887, "step": 16823 }, { "epoch": 4.476849387972326, "grad_norm": 0.30407169461250305, "learning_rate": 5.578308614823657e-09, "loss": 0.1928, "step": 16824 }, { "epoch": 4.477115486961149, "grad_norm": 0.28425461053848267, "learning_rate": 5.5726914246615576e-09, "loss": 0.1789, "step": 16825 }, { "epoch": 4.477381585949973, "grad_norm": 0.3045605719089508, "learning_rate": 5.567076983033236e-09, "loss": 0.1914, "step": 16826 }, { "epoch": 4.477647684938797, "grad_norm": 0.2834033668041229, "learning_rate": 5.561465290102096e-09, "loss": 0.1824, "step": 16827 }, { "epoch": 4.477913783927621, "grad_norm": 0.2766401767730713, "learning_rate": 5.555856346031507e-09, "loss": 0.1669, "step": 16828 }, { "epoch": 4.478179882916445, "grad_norm": 0.2941203713417053, "learning_rate": 5.550250150984714e-09, "loss": 0.1805, "step": 16829 }, { "epoch": 4.478445981905269, "grad_norm": 0.2915722727775574, "learning_rate": 5.544646705124911e-09, "loss": 0.1768, "step": 16830 }, { "epoch": 4.478712080894093, "grad_norm": 0.3768651485443115, "learning_rate": 5.539046008615178e-09, "loss": 0.1932, "step": 16831 }, { "epoch": 4.478978179882916, "grad_norm": 0.3009687662124634, "learning_rate": 5.533448061618551e-09, "loss": 0.1902, "step": 16832 }, { "epoch": 4.47924427887174, "grad_norm": 0.2853451669216156, "learning_rate": 5.5278528642979775e-09, "loss": 0.1841, "step": 16833 }, { "epoch": 4.4795103778605645, "grad_norm": 0.3253103196620941, "learning_rate": 5.522260416816305e-09, "loss": 0.1992, "step": 16834 }, { "epoch": 4.479776476849388, "grad_norm": 0.28601282835006714, "learning_rate": 5.516670719336336e-09, "loss": 0.1929, "step": 16835 }, { "epoch": 4.480042575838212, "grad_norm": 0.28246116638183594, "learning_rate": 5.511083772020752e-09, "loss": 0.1827, "step": 16836 }, { "epoch": 4.480308674827036, "grad_norm": 0.28803732991218567, "learning_rate": 5.50549957503218e-09, "loss": 0.1915, "step": 16837 }, { "epoch": 4.480574773815859, "grad_norm": 0.3356206715106964, "learning_rate": 5.499918128533154e-09, "loss": 0.1915, "step": 16838 }, { "epoch": 4.480840872804683, "grad_norm": 0.304649293422699, "learning_rate": 5.494339432686157e-09, "loss": 0.184, "step": 16839 }, { "epoch": 4.4811069717935075, "grad_norm": 0.27382203936576843, "learning_rate": 5.488763487653547e-09, "loss": 0.18, "step": 16840 }, { "epoch": 4.481373070782331, "grad_norm": 0.3005729913711548, "learning_rate": 5.483190293597639e-09, "loss": 0.193, "step": 16841 }, { "epoch": 4.481639169771155, "grad_norm": 0.3327980041503906, "learning_rate": 5.4776198506806595e-09, "loss": 0.2067, "step": 16842 }, { "epoch": 4.481905268759979, "grad_norm": 0.310577392578125, "learning_rate": 5.4720521590647326e-09, "loss": 0.194, "step": 16843 }, { "epoch": 4.482171367748802, "grad_norm": 0.3136884272098541, "learning_rate": 5.466487218911942e-09, "loss": 0.2042, "step": 16844 }, { "epoch": 4.482437466737626, "grad_norm": 0.33163371682167053, "learning_rate": 5.460925030384245e-09, "loss": 0.2036, "step": 16845 }, { "epoch": 4.4827035657264505, "grad_norm": 0.2857239544391632, "learning_rate": 5.455365593643546e-09, "loss": 0.1821, "step": 16846 }, { "epoch": 4.482969664715274, "grad_norm": 0.2972033619880676, "learning_rate": 5.449808908851672e-09, "loss": 0.19, "step": 16847 }, { "epoch": 4.483235763704098, "grad_norm": 0.3013891875743866, "learning_rate": 5.444254976170371e-09, "loss": 0.2007, "step": 16848 }, { "epoch": 4.483501862692922, "grad_norm": 0.28966429829597473, "learning_rate": 5.438703795761291e-09, "loss": 0.1809, "step": 16849 }, { "epoch": 4.483767961681746, "grad_norm": 0.3192131519317627, "learning_rate": 5.4331553677860355e-09, "loss": 0.2046, "step": 16850 }, { "epoch": 4.484034060670569, "grad_norm": 0.31594452261924744, "learning_rate": 5.4276096924060654e-09, "loss": 0.1777, "step": 16851 }, { "epoch": 4.484300159659393, "grad_norm": 0.28415852785110474, "learning_rate": 5.4220667697828295e-09, "loss": 0.1686, "step": 16852 }, { "epoch": 4.484566258648217, "grad_norm": 0.3211568295955658, "learning_rate": 5.416526600077676e-09, "loss": 0.2211, "step": 16853 }, { "epoch": 4.484832357637041, "grad_norm": 0.2988649308681488, "learning_rate": 5.41098918345183e-09, "loss": 0.1871, "step": 16854 }, { "epoch": 4.485098456625865, "grad_norm": 0.30854788422584534, "learning_rate": 5.405454520066488e-09, "loss": 0.1942, "step": 16855 }, { "epoch": 4.485364555614689, "grad_norm": 0.2910269498825073, "learning_rate": 5.399922610082763e-09, "loss": 0.1814, "step": 16856 }, { "epoch": 4.485630654603512, "grad_norm": 0.2852509617805481, "learning_rate": 5.394393453661661e-09, "loss": 0.1794, "step": 16857 }, { "epoch": 4.485896753592336, "grad_norm": 0.28042545914649963, "learning_rate": 5.388867050964119e-09, "loss": 0.1815, "step": 16858 }, { "epoch": 4.4861628525811605, "grad_norm": 0.33473432064056396, "learning_rate": 5.383343402151008e-09, "loss": 0.2128, "step": 16859 }, { "epoch": 4.486428951569984, "grad_norm": 0.26001977920532227, "learning_rate": 5.3778225073831006e-09, "loss": 0.1739, "step": 16860 }, { "epoch": 4.486695050558808, "grad_norm": 0.3186209201812744, "learning_rate": 5.37230436682109e-09, "loss": 0.1898, "step": 16861 }, { "epoch": 4.486961149547632, "grad_norm": 0.3020398020744324, "learning_rate": 5.366788980625614e-09, "loss": 0.1928, "step": 16862 }, { "epoch": 4.487227248536455, "grad_norm": 0.2968953251838684, "learning_rate": 5.361276348957189e-09, "loss": 0.1825, "step": 16863 }, { "epoch": 4.487493347525279, "grad_norm": 0.3250948190689087, "learning_rate": 5.3557664719762865e-09, "loss": 0.2118, "step": 16864 }, { "epoch": 4.4877594465141035, "grad_norm": 0.3060382306575775, "learning_rate": 5.350259349843278e-09, "loss": 0.1974, "step": 16865 }, { "epoch": 4.488025545502927, "grad_norm": 0.30149850249290466, "learning_rate": 5.344754982718458e-09, "loss": 0.1814, "step": 16866 }, { "epoch": 4.488291644491751, "grad_norm": 0.3190793991088867, "learning_rate": 5.3392533707620534e-09, "loss": 0.1958, "step": 16867 }, { "epoch": 4.488557743480575, "grad_norm": 0.2867649495601654, "learning_rate": 5.333754514134203e-09, "loss": 0.1757, "step": 16868 }, { "epoch": 4.488823842469398, "grad_norm": 0.2851067781448364, "learning_rate": 5.328258412994957e-09, "loss": 0.195, "step": 16869 }, { "epoch": 4.489089941458222, "grad_norm": 0.2987173795700073, "learning_rate": 5.322765067504309e-09, "loss": 0.1755, "step": 16870 }, { "epoch": 4.4893560404470465, "grad_norm": 0.2790147364139557, "learning_rate": 5.317274477822131e-09, "loss": 0.1849, "step": 16871 }, { "epoch": 4.48962213943587, "grad_norm": 0.2756970226764679, "learning_rate": 5.311786644108252e-09, "loss": 0.1826, "step": 16872 }, { "epoch": 4.489888238424694, "grad_norm": 0.30494365096092224, "learning_rate": 5.3063015665224305e-09, "loss": 0.1938, "step": 16873 }, { "epoch": 4.490154337413518, "grad_norm": 0.31371498107910156, "learning_rate": 5.300819245224275e-09, "loss": 0.2099, "step": 16874 }, { "epoch": 4.490420436402342, "grad_norm": 0.29050886631011963, "learning_rate": 5.29533968037339e-09, "loss": 0.1774, "step": 16875 }, { "epoch": 4.490686535391165, "grad_norm": 0.28572800755500793, "learning_rate": 5.289862872129269e-09, "loss": 0.1757, "step": 16876 }, { "epoch": 4.4909526343799895, "grad_norm": 0.294056236743927, "learning_rate": 5.284388820651331e-09, "loss": 0.2019, "step": 16877 }, { "epoch": 4.491218733368814, "grad_norm": 0.3007306754589081, "learning_rate": 5.278917526098903e-09, "loss": 0.1844, "step": 16878 }, { "epoch": 4.491484832357637, "grad_norm": 0.29151129722595215, "learning_rate": 5.273448988631246e-09, "loss": 0.1839, "step": 16879 }, { "epoch": 4.491750931346461, "grad_norm": 0.30549514293670654, "learning_rate": 5.267983208407534e-09, "loss": 0.1803, "step": 16880 }, { "epoch": 4.492017030335285, "grad_norm": 0.3510120213031769, "learning_rate": 5.2625201855868714e-09, "loss": 0.2175, "step": 16881 }, { "epoch": 4.492283129324108, "grad_norm": 0.31007885932922363, "learning_rate": 5.257059920328255e-09, "loss": 0.1847, "step": 16882 }, { "epoch": 4.4925492283129325, "grad_norm": 0.29627206921577454, "learning_rate": 5.251602412790623e-09, "loss": 0.2023, "step": 16883 }, { "epoch": 4.492815327301757, "grad_norm": 0.30168241262435913, "learning_rate": 5.246147663132838e-09, "loss": 0.189, "step": 16884 }, { "epoch": 4.49308142629058, "grad_norm": 0.296568363904953, "learning_rate": 5.240695671513662e-09, "loss": 0.2021, "step": 16885 }, { "epoch": 4.493347525279404, "grad_norm": 0.30652520060539246, "learning_rate": 5.2352464380918006e-09, "loss": 0.173, "step": 16886 }, { "epoch": 4.493613624268228, "grad_norm": 0.304329514503479, "learning_rate": 5.229799963025861e-09, "loss": 0.1997, "step": 16887 }, { "epoch": 4.493879723257051, "grad_norm": 0.2833746373653412, "learning_rate": 5.2243562464743705e-09, "loss": 0.1749, "step": 16888 }, { "epoch": 4.494145822245875, "grad_norm": 0.2903938293457031, "learning_rate": 5.2189152885958046e-09, "loss": 0.2072, "step": 16889 }, { "epoch": 4.4944119212347, "grad_norm": 0.2944124639034271, "learning_rate": 5.213477089548502e-09, "loss": 0.1895, "step": 16890 }, { "epoch": 4.494678020223523, "grad_norm": 0.3577301800251007, "learning_rate": 5.20804164949078e-09, "loss": 0.2207, "step": 16891 }, { "epoch": 4.494944119212347, "grad_norm": 0.27749231457710266, "learning_rate": 5.202608968580846e-09, "loss": 0.168, "step": 16892 }, { "epoch": 4.495210218201171, "grad_norm": 0.27364876866340637, "learning_rate": 5.197179046976819e-09, "loss": 0.172, "step": 16893 }, { "epoch": 4.495476317189994, "grad_norm": 0.30884262919425964, "learning_rate": 5.1917518848367485e-09, "loss": 0.1938, "step": 16894 }, { "epoch": 4.495742416178818, "grad_norm": 0.2933993339538574, "learning_rate": 5.1863274823186085e-09, "loss": 0.193, "step": 16895 }, { "epoch": 4.4960085151676425, "grad_norm": 0.2795248329639435, "learning_rate": 5.180905839580308e-09, "loss": 0.174, "step": 16896 }, { "epoch": 4.496274614156466, "grad_norm": 0.291518896818161, "learning_rate": 5.17548695677964e-09, "loss": 0.1915, "step": 16897 }, { "epoch": 4.49654071314529, "grad_norm": 0.2934696674346924, "learning_rate": 5.170070834074336e-09, "loss": 0.1917, "step": 16898 }, { "epoch": 4.496806812134114, "grad_norm": 0.3031530976295471, "learning_rate": 5.164657471622058e-09, "loss": 0.1967, "step": 16899 }, { "epoch": 4.497072911122938, "grad_norm": 0.3003334403038025, "learning_rate": 5.159246869580347e-09, "loss": 0.1937, "step": 16900 }, { "epoch": 4.497339010111761, "grad_norm": 0.2916385531425476, "learning_rate": 5.15383902810671e-09, "loss": 0.2013, "step": 16901 }, { "epoch": 4.4976051091005855, "grad_norm": 0.2999199628829956, "learning_rate": 5.1484339473585544e-09, "loss": 0.186, "step": 16902 }, { "epoch": 4.49787120808941, "grad_norm": 0.3031773865222931, "learning_rate": 5.143031627493211e-09, "loss": 0.1912, "step": 16903 }, { "epoch": 4.498137307078233, "grad_norm": 0.30567654967308044, "learning_rate": 5.137632068667918e-09, "loss": 0.2033, "step": 16904 }, { "epoch": 4.498403406067057, "grad_norm": 0.2928844690322876, "learning_rate": 5.132235271039853e-09, "loss": 0.1681, "step": 16905 }, { "epoch": 4.498669505055881, "grad_norm": 0.2901036739349365, "learning_rate": 5.126841234766099e-09, "loss": 0.1749, "step": 16906 }, { "epoch": 4.498935604044704, "grad_norm": 0.29809921979904175, "learning_rate": 5.121449960003655e-09, "loss": 0.1968, "step": 16907 }, { "epoch": 4.4992017030335285, "grad_norm": 0.30750513076782227, "learning_rate": 5.1160614469094695e-09, "loss": 0.1902, "step": 16908 }, { "epoch": 4.499467802022353, "grad_norm": 0.28505444526672363, "learning_rate": 5.110675695640354e-09, "loss": 0.1669, "step": 16909 }, { "epoch": 4.499733901011176, "grad_norm": 0.28345584869384766, "learning_rate": 5.105292706353093e-09, "loss": 0.167, "step": 16910 }, { "epoch": 4.5, "grad_norm": 0.3067342936992645, "learning_rate": 5.099912479204371e-09, "loss": 0.2039, "step": 16911 }, { "epoch": 4.500266098988824, "grad_norm": 0.3381754755973816, "learning_rate": 5.0945350143508095e-09, "loss": 0.2032, "step": 16912 }, { "epoch": 4.500532197977647, "grad_norm": 0.27265292406082153, "learning_rate": 5.089160311948893e-09, "loss": 0.1717, "step": 16913 }, { "epoch": 4.5007982969664715, "grad_norm": 0.2796947658061981, "learning_rate": 5.0837883721550845e-09, "loss": 0.1724, "step": 16914 }, { "epoch": 4.501064395955296, "grad_norm": 0.257771372795105, "learning_rate": 5.07841919512576e-09, "loss": 0.1685, "step": 16915 }, { "epoch": 4.501330494944119, "grad_norm": 0.29961007833480835, "learning_rate": 5.073052781017184e-09, "loss": 0.1975, "step": 16916 }, { "epoch": 4.501596593932943, "grad_norm": 0.3240768313407898, "learning_rate": 5.067689129985575e-09, "loss": 0.2017, "step": 16917 }, { "epoch": 4.501862692921767, "grad_norm": 0.27847105264663696, "learning_rate": 5.062328242187041e-09, "loss": 0.1762, "step": 16918 }, { "epoch": 4.502128791910591, "grad_norm": 0.30029064416885376, "learning_rate": 5.056970117777626e-09, "loss": 0.1836, "step": 16919 }, { "epoch": 4.5023948908994145, "grad_norm": 0.29625219106674194, "learning_rate": 5.051614756913292e-09, "loss": 0.178, "step": 16920 }, { "epoch": 4.502660989888239, "grad_norm": 0.3105551600456238, "learning_rate": 5.046262159749926e-09, "loss": 0.1901, "step": 16921 }, { "epoch": 4.502927088877062, "grad_norm": 0.30155494809150696, "learning_rate": 5.040912326443325e-09, "loss": 0.1888, "step": 16922 }, { "epoch": 4.503193187865886, "grad_norm": 0.30189386010169983, "learning_rate": 5.035565257149199e-09, "loss": 0.1819, "step": 16923 }, { "epoch": 4.50345928685471, "grad_norm": 0.2937409579753876, "learning_rate": 5.0302209520232006e-09, "loss": 0.1717, "step": 16924 }, { "epoch": 4.503725385843534, "grad_norm": 0.27940088510513306, "learning_rate": 5.024879411220884e-09, "loss": 0.1744, "step": 16925 }, { "epoch": 4.5039914848323575, "grad_norm": 0.28500914573669434, "learning_rate": 5.019540634897735e-09, "loss": 0.1873, "step": 16926 }, { "epoch": 4.504257583821182, "grad_norm": 0.29984238743782043, "learning_rate": 5.01420462320914e-09, "loss": 0.191, "step": 16927 }, { "epoch": 4.504523682810006, "grad_norm": 0.30552729964256287, "learning_rate": 5.008871376310408e-09, "loss": 0.1833, "step": 16928 }, { "epoch": 4.504789781798829, "grad_norm": 0.34846237301826477, "learning_rate": 5.003540894356795e-09, "loss": 0.2195, "step": 16929 }, { "epoch": 4.505055880787653, "grad_norm": 0.28944388031959534, "learning_rate": 4.998213177503452e-09, "loss": 0.1854, "step": 16930 }, { "epoch": 4.505321979776477, "grad_norm": 0.29989251494407654, "learning_rate": 4.9928882259054674e-09, "loss": 0.1905, "step": 16931 }, { "epoch": 4.5055880787653, "grad_norm": 0.2934589385986328, "learning_rate": 4.987566039717816e-09, "loss": 0.191, "step": 16932 }, { "epoch": 4.505854177754125, "grad_norm": 0.30720359086990356, "learning_rate": 4.9822466190954094e-09, "loss": 0.194, "step": 16933 }, { "epoch": 4.506120276742949, "grad_norm": 0.30755215883255005, "learning_rate": 4.9769299641931e-09, "loss": 0.2044, "step": 16934 }, { "epoch": 4.506386375731772, "grad_norm": 0.2923334538936615, "learning_rate": 4.971616075165641e-09, "loss": 0.1961, "step": 16935 }, { "epoch": 4.506652474720596, "grad_norm": 0.2950831353664398, "learning_rate": 4.966304952167688e-09, "loss": 0.1814, "step": 16936 }, { "epoch": 4.50691857370942, "grad_norm": 0.3188458979129791, "learning_rate": 4.960996595353839e-09, "loss": 0.2138, "step": 16937 }, { "epoch": 4.507184672698243, "grad_norm": 0.2998858094215393, "learning_rate": 4.955691004878626e-09, "loss": 0.2099, "step": 16938 }, { "epoch": 4.5074507716870675, "grad_norm": 0.28272196650505066, "learning_rate": 4.950388180896459e-09, "loss": 0.179, "step": 16939 }, { "epoch": 4.507716870675892, "grad_norm": 0.3325027823448181, "learning_rate": 4.945088123561703e-09, "loss": 0.2103, "step": 16940 }, { "epoch": 4.507982969664715, "grad_norm": 0.29942113161087036, "learning_rate": 4.939790833028623e-09, "loss": 0.1963, "step": 16941 }, { "epoch": 4.508249068653539, "grad_norm": 0.3104401230812073, "learning_rate": 4.934496309451408e-09, "loss": 0.1892, "step": 16942 }, { "epoch": 4.508515167642363, "grad_norm": 0.30887746810913086, "learning_rate": 4.929204552984167e-09, "loss": 0.1894, "step": 16943 }, { "epoch": 4.508781266631187, "grad_norm": 0.32871291041374207, "learning_rate": 4.923915563780956e-09, "loss": 0.1905, "step": 16944 }, { "epoch": 4.5090473656200105, "grad_norm": 0.2896025478839874, "learning_rate": 4.918629341995684e-09, "loss": 0.1855, "step": 16945 }, { "epoch": 4.509313464608835, "grad_norm": 0.2778574824333191, "learning_rate": 4.913345887782227e-09, "loss": 0.1803, "step": 16946 }, { "epoch": 4.509579563597658, "grad_norm": 0.31138014793395996, "learning_rate": 4.908065201294387e-09, "loss": 0.1799, "step": 16947 }, { "epoch": 4.509845662586482, "grad_norm": 0.29930582642555237, "learning_rate": 4.902787282685872e-09, "loss": 0.1889, "step": 16948 }, { "epoch": 4.510111761575306, "grad_norm": 0.28947609663009644, "learning_rate": 4.897512132110293e-09, "loss": 0.1729, "step": 16949 }, { "epoch": 4.51037786056413, "grad_norm": 0.32130974531173706, "learning_rate": 4.892239749721205e-09, "loss": 0.1962, "step": 16950 }, { "epoch": 4.5106439595529535, "grad_norm": 0.3169848620891571, "learning_rate": 4.886970135672086e-09, "loss": 0.1865, "step": 16951 }, { "epoch": 4.510910058541778, "grad_norm": 0.27939411997795105, "learning_rate": 4.881703290116301e-09, "loss": 0.1721, "step": 16952 }, { "epoch": 4.511176157530602, "grad_norm": 0.3200021982192993, "learning_rate": 4.876439213207162e-09, "loss": 0.1784, "step": 16953 }, { "epoch": 4.511442256519425, "grad_norm": 0.28555941581726074, "learning_rate": 4.871177905097901e-09, "loss": 0.1904, "step": 16954 }, { "epoch": 4.511708355508249, "grad_norm": 0.29088449478149414, "learning_rate": 4.8659193659416285e-09, "loss": 0.1852, "step": 16955 }, { "epoch": 4.511974454497073, "grad_norm": 0.2890036702156067, "learning_rate": 4.860663595891445e-09, "loss": 0.1861, "step": 16956 }, { "epoch": 4.5122405534858965, "grad_norm": 0.29118260741233826, "learning_rate": 4.855410595100307e-09, "loss": 0.1809, "step": 16957 }, { "epoch": 4.512506652474721, "grad_norm": 0.2976479232311249, "learning_rate": 4.850160363721134e-09, "loss": 0.1869, "step": 16958 }, { "epoch": 4.512772751463545, "grad_norm": 0.288968026638031, "learning_rate": 4.844912901906739e-09, "loss": 0.192, "step": 16959 }, { "epoch": 4.513038850452368, "grad_norm": 0.2970655560493469, "learning_rate": 4.839668209809855e-09, "loss": 0.181, "step": 16960 }, { "epoch": 4.513304949441192, "grad_norm": 0.3006112575531006, "learning_rate": 4.834426287583149e-09, "loss": 0.1908, "step": 16961 }, { "epoch": 4.513571048430016, "grad_norm": 0.2741609513759613, "learning_rate": 4.82918713537922e-09, "loss": 0.1701, "step": 16962 }, { "epoch": 4.5138371474188395, "grad_norm": 0.29525110125541687, "learning_rate": 4.8239507533505256e-09, "loss": 0.1764, "step": 16963 }, { "epoch": 4.514103246407664, "grad_norm": 0.2934997081756592, "learning_rate": 4.8187171416494976e-09, "loss": 0.1975, "step": 16964 }, { "epoch": 4.514369345396488, "grad_norm": 0.3100847601890564, "learning_rate": 4.813486300428482e-09, "loss": 0.2161, "step": 16965 }, { "epoch": 4.514635444385311, "grad_norm": 0.305244505405426, "learning_rate": 4.8082582298397345e-09, "loss": 0.1914, "step": 16966 }, { "epoch": 4.514901543374135, "grad_norm": 0.2970462143421173, "learning_rate": 4.8030329300354225e-09, "loss": 0.1911, "step": 16967 }, { "epoch": 4.515167642362959, "grad_norm": 0.3010159134864807, "learning_rate": 4.797810401167646e-09, "loss": 0.1793, "step": 16968 }, { "epoch": 4.515433741351783, "grad_norm": 0.2634839117527008, "learning_rate": 4.792590643388428e-09, "loss": 0.1754, "step": 16969 }, { "epoch": 4.515699840340607, "grad_norm": 0.28860434889793396, "learning_rate": 4.787373656849692e-09, "loss": 0.1831, "step": 16970 }, { "epoch": 4.515965939329431, "grad_norm": 0.3019898235797882, "learning_rate": 4.782159441703293e-09, "loss": 0.1809, "step": 16971 }, { "epoch": 4.516232038318254, "grad_norm": 0.31326204538345337, "learning_rate": 4.776947998100989e-09, "loss": 0.2091, "step": 16972 }, { "epoch": 4.516498137307078, "grad_norm": 0.31225818395614624, "learning_rate": 4.7717393261945014e-09, "loss": 0.198, "step": 16973 }, { "epoch": 4.516764236295902, "grad_norm": 0.2959151566028595, "learning_rate": 4.766533426135411e-09, "loss": 0.1917, "step": 16974 }, { "epoch": 4.517030335284726, "grad_norm": 0.27900058031082153, "learning_rate": 4.761330298075261e-09, "loss": 0.1892, "step": 16975 }, { "epoch": 4.5172964342735495, "grad_norm": 0.2831149995326996, "learning_rate": 4.75612994216551e-09, "loss": 0.1655, "step": 16976 }, { "epoch": 4.517562533262374, "grad_norm": 0.2841302752494812, "learning_rate": 4.750932358557513e-09, "loss": 0.1918, "step": 16977 }, { "epoch": 4.517828632251198, "grad_norm": 0.2946605682373047, "learning_rate": 4.745737547402562e-09, "loss": 0.1949, "step": 16978 }, { "epoch": 4.518094731240021, "grad_norm": 0.2867688834667206, "learning_rate": 4.740545508851867e-09, "loss": 0.1788, "step": 16979 }, { "epoch": 4.518360830228845, "grad_norm": 0.2982594668865204, "learning_rate": 4.735356243056565e-09, "loss": 0.188, "step": 16980 }, { "epoch": 4.518626929217669, "grad_norm": 0.30499404668807983, "learning_rate": 4.730169750167678e-09, "loss": 0.1874, "step": 16981 }, { "epoch": 4.5188930282064925, "grad_norm": 0.31537511944770813, "learning_rate": 4.7249860303361754e-09, "loss": 0.1849, "step": 16982 }, { "epoch": 4.519159127195317, "grad_norm": 0.31095173954963684, "learning_rate": 4.719805083712958e-09, "loss": 0.1803, "step": 16983 }, { "epoch": 4.519425226184141, "grad_norm": 0.3059348464012146, "learning_rate": 4.714626910448827e-09, "loss": 0.1771, "step": 16984 }, { "epoch": 4.519691325172964, "grad_norm": 0.3021383285522461, "learning_rate": 4.709451510694495e-09, "loss": 0.1884, "step": 16985 }, { "epoch": 4.519957424161788, "grad_norm": 0.3128170073032379, "learning_rate": 4.70427888460061e-09, "loss": 0.2114, "step": 16986 }, { "epoch": 4.520223523150612, "grad_norm": 0.29356661438941956, "learning_rate": 4.6991090323177385e-09, "loss": 0.1986, "step": 16987 }, { "epoch": 4.5204896221394355, "grad_norm": 0.2997266352176666, "learning_rate": 4.69394195399635e-09, "loss": 0.1903, "step": 16988 }, { "epoch": 4.52075572112826, "grad_norm": 0.31192201375961304, "learning_rate": 4.688777649786868e-09, "loss": 0.2008, "step": 16989 }, { "epoch": 4.521021820117084, "grad_norm": 0.28848206996917725, "learning_rate": 4.683616119839584e-09, "loss": 0.1788, "step": 16990 }, { "epoch": 4.521287919105907, "grad_norm": 0.31196480989456177, "learning_rate": 4.6784573643047445e-09, "loss": 0.201, "step": 16991 }, { "epoch": 4.521554018094731, "grad_norm": 0.3071448504924774, "learning_rate": 4.6733013833325175e-09, "loss": 0.1988, "step": 16992 }, { "epoch": 4.521820117083555, "grad_norm": 0.3004017174243927, "learning_rate": 4.668148177072984e-09, "loss": 0.2031, "step": 16993 }, { "epoch": 4.522086216072379, "grad_norm": 0.27400708198547363, "learning_rate": 4.662997745676111e-09, "loss": 0.1764, "step": 16994 }, { "epoch": 4.522352315061203, "grad_norm": 0.29425913095474243, "learning_rate": 4.657850089291848e-09, "loss": 0.2103, "step": 16995 }, { "epoch": 4.522618414050027, "grad_norm": 0.3012206554412842, "learning_rate": 4.652705208070007e-09, "loss": 0.1833, "step": 16996 }, { "epoch": 4.52288451303885, "grad_norm": 0.3042266070842743, "learning_rate": 4.647563102160345e-09, "loss": 0.1991, "step": 16997 }, { "epoch": 4.523150612027674, "grad_norm": 0.31033581495285034, "learning_rate": 4.642423771712556e-09, "loss": 0.1854, "step": 16998 }, { "epoch": 4.523416711016498, "grad_norm": 0.2746228277683258, "learning_rate": 4.637287216876218e-09, "loss": 0.189, "step": 16999 }, { "epoch": 4.523682810005322, "grad_norm": 0.2954103648662567, "learning_rate": 4.632153437800823e-09, "loss": 0.1716, "step": 17000 }, { "epoch": 4.523948908994146, "grad_norm": 0.30197790265083313, "learning_rate": 4.62702243463583e-09, "loss": 0.192, "step": 17001 }, { "epoch": 4.52421500798297, "grad_norm": 0.28954803943634033, "learning_rate": 4.6218942075305854e-09, "loss": 0.1893, "step": 17002 }, { "epoch": 4.524481106971794, "grad_norm": 0.2928517162799835, "learning_rate": 4.616768756634348e-09, "loss": 0.1893, "step": 17003 }, { "epoch": 4.524747205960617, "grad_norm": 0.2738316059112549, "learning_rate": 4.61164608209631e-09, "loss": 0.1728, "step": 17004 }, { "epoch": 4.525013304949441, "grad_norm": 0.28850480914115906, "learning_rate": 4.606526184065585e-09, "loss": 0.1875, "step": 17005 }, { "epoch": 4.525279403938265, "grad_norm": 0.31120869517326355, "learning_rate": 4.601409062691197e-09, "loss": 0.2072, "step": 17006 }, { "epoch": 4.525545502927089, "grad_norm": 0.33342140913009644, "learning_rate": 4.596294718122107e-09, "loss": 0.1697, "step": 17007 }, { "epoch": 4.525811601915913, "grad_norm": 0.3052528202533722, "learning_rate": 4.5911831505071494e-09, "loss": 0.1924, "step": 17008 }, { "epoch": 4.526077700904737, "grad_norm": 0.2912253439426422, "learning_rate": 4.586074359995118e-09, "loss": 0.1958, "step": 17009 }, { "epoch": 4.52634379989356, "grad_norm": 0.30847030878067017, "learning_rate": 4.580968346734726e-09, "loss": 0.1922, "step": 17010 }, { "epoch": 4.526609898882384, "grad_norm": 0.3362357020378113, "learning_rate": 4.5758651108745886e-09, "loss": 0.1962, "step": 17011 }, { "epoch": 4.526875997871208, "grad_norm": 0.2907564043998718, "learning_rate": 4.570764652563253e-09, "loss": 0.1701, "step": 17012 }, { "epoch": 4.5271420968600316, "grad_norm": 0.29036983847618103, "learning_rate": 4.5656669719492e-09, "loss": 0.19, "step": 17013 }, { "epoch": 4.527408195848856, "grad_norm": 0.322237104177475, "learning_rate": 4.560572069180768e-09, "loss": 0.2032, "step": 17014 }, { "epoch": 4.52767429483768, "grad_norm": 0.2608666718006134, "learning_rate": 4.555479944406282e-09, "loss": 0.1626, "step": 17015 }, { "epoch": 4.527940393826503, "grad_norm": 0.3090400695800781, "learning_rate": 4.5503905977739675e-09, "loss": 0.2116, "step": 17016 }, { "epoch": 4.528206492815327, "grad_norm": 0.37163665890693665, "learning_rate": 4.545304029431929e-09, "loss": 0.191, "step": 17017 }, { "epoch": 4.528472591804151, "grad_norm": 0.3065171539783478, "learning_rate": 4.540220239528247e-09, "loss": 0.188, "step": 17018 }, { "epoch": 4.528738690792975, "grad_norm": 0.2906561493873596, "learning_rate": 4.535139228210905e-09, "loss": 0.2039, "step": 17019 }, { "epoch": 4.529004789781799, "grad_norm": 0.2994215488433838, "learning_rate": 4.5300609956277825e-09, "loss": 0.1761, "step": 17020 }, { "epoch": 4.529270888770623, "grad_norm": 0.28518545627593994, "learning_rate": 4.524985541926696e-09, "loss": 0.1792, "step": 17021 }, { "epoch": 4.529536987759446, "grad_norm": 0.29981693625450134, "learning_rate": 4.519912867255382e-09, "loss": 0.1934, "step": 17022 }, { "epoch": 4.52980308674827, "grad_norm": 0.3005761504173279, "learning_rate": 4.514842971761501e-09, "loss": 0.2062, "step": 17023 }, { "epoch": 4.530069185737094, "grad_norm": 0.31212908029556274, "learning_rate": 4.509775855592613e-09, "loss": 0.2006, "step": 17024 }, { "epoch": 4.530335284725918, "grad_norm": 0.29033035039901733, "learning_rate": 4.50471151889622e-09, "loss": 0.1791, "step": 17025 }, { "epoch": 4.530601383714742, "grad_norm": 0.2785864472389221, "learning_rate": 4.499649961819707e-09, "loss": 0.1798, "step": 17026 }, { "epoch": 4.530867482703566, "grad_norm": 0.32215794920921326, "learning_rate": 4.494591184510432e-09, "loss": 0.1927, "step": 17027 }, { "epoch": 4.53113358169239, "grad_norm": 0.3021482527256012, "learning_rate": 4.489535187115623e-09, "loss": 0.2072, "step": 17028 }, { "epoch": 4.531399680681213, "grad_norm": 0.3341998755931854, "learning_rate": 4.48448196978245e-09, "loss": 0.1965, "step": 17029 }, { "epoch": 4.531665779670037, "grad_norm": 0.29128730297088623, "learning_rate": 4.479431532658018e-09, "loss": 0.1804, "step": 17030 }, { "epoch": 4.531931878658861, "grad_norm": 0.28922024369239807, "learning_rate": 4.474383875889298e-09, "loss": 0.1911, "step": 17031 }, { "epoch": 4.532197977647685, "grad_norm": 0.29691243171691895, "learning_rate": 4.469338999623262e-09, "loss": 0.1747, "step": 17032 }, { "epoch": 4.532464076636509, "grad_norm": 0.2823645770549774, "learning_rate": 4.464296904006704e-09, "loss": 0.1913, "step": 17033 }, { "epoch": 4.532730175625333, "grad_norm": 0.29937201738357544, "learning_rate": 4.459257589186416e-09, "loss": 0.1843, "step": 17034 }, { "epoch": 4.532996274614156, "grad_norm": 0.30376750230789185, "learning_rate": 4.454221055309071e-09, "loss": 0.1942, "step": 17035 }, { "epoch": 4.53326237360298, "grad_norm": 0.28878217935562134, "learning_rate": 4.449187302521262e-09, "loss": 0.2034, "step": 17036 }, { "epoch": 4.533528472591804, "grad_norm": 0.30170488357543945, "learning_rate": 4.444156330969517e-09, "loss": 0.1808, "step": 17037 }, { "epoch": 4.5337945715806285, "grad_norm": 0.31149396300315857, "learning_rate": 4.439128140800275e-09, "loss": 0.1967, "step": 17038 }, { "epoch": 4.534060670569452, "grad_norm": 0.28699323534965515, "learning_rate": 4.434102732159884e-09, "loss": 0.1667, "step": 17039 }, { "epoch": 4.534326769558276, "grad_norm": 0.2861810326576233, "learning_rate": 4.429080105194638e-09, "loss": 0.1884, "step": 17040 }, { "epoch": 4.534592868547099, "grad_norm": 0.31527945399284363, "learning_rate": 4.424060260050711e-09, "loss": 0.1999, "step": 17041 }, { "epoch": 4.534858967535923, "grad_norm": 0.33015766739845276, "learning_rate": 4.4190431968742394e-09, "loss": 0.2191, "step": 17042 }, { "epoch": 4.535125066524747, "grad_norm": 0.31335893273353577, "learning_rate": 4.414028915811252e-09, "loss": 0.1997, "step": 17043 }, { "epoch": 4.5353911655135715, "grad_norm": 0.3239162266254425, "learning_rate": 4.409017417007688e-09, "loss": 0.2102, "step": 17044 }, { "epoch": 4.535657264502395, "grad_norm": 0.3142719566822052, "learning_rate": 4.40400870060943e-09, "loss": 0.1885, "step": 17045 }, { "epoch": 4.535923363491219, "grad_norm": 0.2934645414352417, "learning_rate": 4.399002766762261e-09, "loss": 0.1855, "step": 17046 }, { "epoch": 4.536189462480043, "grad_norm": 0.28079450130462646, "learning_rate": 4.3939996156118986e-09, "loss": 0.1964, "step": 17047 }, { "epoch": 4.536455561468866, "grad_norm": 0.30193015933036804, "learning_rate": 4.3889992473039704e-09, "loss": 0.1976, "step": 17048 }, { "epoch": 4.53672166045769, "grad_norm": 0.3868529200553894, "learning_rate": 4.384001661984016e-09, "loss": 0.2147, "step": 17049 }, { "epoch": 4.5369877594465144, "grad_norm": 0.3037090003490448, "learning_rate": 4.379006859797518e-09, "loss": 0.19, "step": 17050 }, { "epoch": 4.537253858435338, "grad_norm": 0.2971878945827484, "learning_rate": 4.37401484088985e-09, "loss": 0.209, "step": 17051 }, { "epoch": 4.537519957424162, "grad_norm": 0.2986566424369812, "learning_rate": 4.369025605406329e-09, "loss": 0.1853, "step": 17052 }, { "epoch": 4.537786056412986, "grad_norm": 0.29016000032424927, "learning_rate": 4.364039153492161e-09, "loss": 0.1778, "step": 17053 }, { "epoch": 4.538052155401809, "grad_norm": 0.29268261790275574, "learning_rate": 4.359055485292495e-09, "loss": 0.1768, "step": 17054 }, { "epoch": 4.538318254390633, "grad_norm": 0.32173776626586914, "learning_rate": 4.354074600952407e-09, "loss": 0.2054, "step": 17055 }, { "epoch": 4.538584353379457, "grad_norm": 0.33251550793647766, "learning_rate": 4.349096500616844e-09, "loss": 0.1812, "step": 17056 }, { "epoch": 4.538850452368281, "grad_norm": 0.26864054799079895, "learning_rate": 4.344121184430738e-09, "loss": 0.1889, "step": 17057 }, { "epoch": 4.539116551357105, "grad_norm": 0.2830483317375183, "learning_rate": 4.339148652538893e-09, "loss": 0.1572, "step": 17058 }, { "epoch": 4.539382650345929, "grad_norm": 0.31417185068130493, "learning_rate": 4.33417890508605e-09, "loss": 0.1889, "step": 17059 }, { "epoch": 4.539648749334752, "grad_norm": 0.30978795886039734, "learning_rate": 4.329211942216859e-09, "loss": 0.1926, "step": 17060 }, { "epoch": 4.539914848323576, "grad_norm": 0.28085950016975403, "learning_rate": 4.324247764075928e-09, "loss": 0.1671, "step": 17061 }, { "epoch": 4.5401809473124, "grad_norm": 0.300293505191803, "learning_rate": 4.319286370807707e-09, "loss": 0.2024, "step": 17062 }, { "epoch": 4.5404470463012245, "grad_norm": 0.29869771003723145, "learning_rate": 4.314327762556624e-09, "loss": 0.1849, "step": 17063 }, { "epoch": 4.540713145290048, "grad_norm": 0.2893633246421814, "learning_rate": 4.309371939467011e-09, "loss": 0.1823, "step": 17064 }, { "epoch": 4.540979244278872, "grad_norm": 0.3154151737689972, "learning_rate": 4.304418901683127e-09, "loss": 0.1999, "step": 17065 }, { "epoch": 4.541245343267695, "grad_norm": 0.2869836688041687, "learning_rate": 4.299468649349136e-09, "loss": 0.1715, "step": 17066 }, { "epoch": 4.541511442256519, "grad_norm": 0.28500065207481384, "learning_rate": 4.294521182609134e-09, "loss": 0.1891, "step": 17067 }, { "epoch": 4.541777541245343, "grad_norm": 0.30536773800849915, "learning_rate": 4.289576501607117e-09, "loss": 0.1907, "step": 17068 }, { "epoch": 4.5420436402341675, "grad_norm": 0.3298015892505646, "learning_rate": 4.284634606487025e-09, "loss": 0.1989, "step": 17069 }, { "epoch": 4.542309739222991, "grad_norm": 0.2775557339191437, "learning_rate": 4.27969549739271e-09, "loss": 0.1813, "step": 17070 }, { "epoch": 4.542575838211815, "grad_norm": 0.3206798732280731, "learning_rate": 4.274759174467901e-09, "loss": 0.1684, "step": 17071 }, { "epoch": 4.542841937200639, "grad_norm": 0.3138729929924011, "learning_rate": 4.269825637856317e-09, "loss": 0.1948, "step": 17072 }, { "epoch": 4.543108036189462, "grad_norm": 0.31642401218414307, "learning_rate": 4.264894887701542e-09, "loss": 0.1819, "step": 17073 }, { "epoch": 4.543374135178286, "grad_norm": 0.2968199849128723, "learning_rate": 4.2599669241471184e-09, "loss": 0.1907, "step": 17074 }, { "epoch": 4.5436402341671105, "grad_norm": 0.28842028975486755, "learning_rate": 4.255041747336452e-09, "loss": 0.1796, "step": 17075 }, { "epoch": 4.543906333155934, "grad_norm": 0.3006766438484192, "learning_rate": 4.250119357412918e-09, "loss": 0.1768, "step": 17076 }, { "epoch": 4.544172432144758, "grad_norm": 0.3216701149940491, "learning_rate": 4.2451997545198126e-09, "loss": 0.2006, "step": 17077 }, { "epoch": 4.544438531133582, "grad_norm": 0.31148916482925415, "learning_rate": 4.240282938800311e-09, "loss": 0.2028, "step": 17078 }, { "epoch": 4.544704630122405, "grad_norm": 0.3332855999469757, "learning_rate": 4.235368910397541e-09, "loss": 0.1979, "step": 17079 }, { "epoch": 4.544970729111229, "grad_norm": 0.2822750508785248, "learning_rate": 4.230457669454524e-09, "loss": 0.1923, "step": 17080 }, { "epoch": 4.5452368281000535, "grad_norm": 0.2748866379261017, "learning_rate": 4.2255492161142215e-09, "loss": 0.1778, "step": 17081 }, { "epoch": 4.545502927088877, "grad_norm": 0.26579925417900085, "learning_rate": 4.220643550519509e-09, "loss": 0.1831, "step": 17082 }, { "epoch": 4.545769026077701, "grad_norm": 0.2833840250968933, "learning_rate": 4.215740672813162e-09, "loss": 0.1811, "step": 17083 }, { "epoch": 4.546035125066525, "grad_norm": 0.2927905023097992, "learning_rate": 4.21084058313792e-09, "loss": 0.201, "step": 17084 }, { "epoch": 4.546301224055348, "grad_norm": 0.3107866048812866, "learning_rate": 4.205943281636382e-09, "loss": 0.1907, "step": 17085 }, { "epoch": 4.546567323044172, "grad_norm": 0.2957768440246582, "learning_rate": 4.2010487684511095e-09, "loss": 0.186, "step": 17086 }, { "epoch": 4.5468334220329965, "grad_norm": 0.37806251645088196, "learning_rate": 4.196157043724568e-09, "loss": 0.1884, "step": 17087 }, { "epoch": 4.547099521021821, "grad_norm": 0.31029102206230164, "learning_rate": 4.191268107599166e-09, "loss": 0.1891, "step": 17088 }, { "epoch": 4.547365620010644, "grad_norm": 0.3168851137161255, "learning_rate": 4.186381960217156e-09, "loss": 0.2104, "step": 17089 }, { "epoch": 4.547631718999468, "grad_norm": 0.270095556974411, "learning_rate": 4.181498601720801e-09, "loss": 0.1801, "step": 17090 }, { "epoch": 4.547897817988291, "grad_norm": 0.31092679500579834, "learning_rate": 4.176618032252222e-09, "loss": 0.1887, "step": 17091 }, { "epoch": 4.548163916977115, "grad_norm": 0.27732983231544495, "learning_rate": 4.171740251953504e-09, "loss": 0.1792, "step": 17092 }, { "epoch": 4.548430015965939, "grad_norm": 0.2991446852684021, "learning_rate": 4.166865260966601e-09, "loss": 0.1874, "step": 17093 }, { "epoch": 4.548696114954764, "grad_norm": 0.2695271372795105, "learning_rate": 4.161993059433433e-09, "loss": 0.1861, "step": 17094 }, { "epoch": 4.548962213943587, "grad_norm": 0.2983093559741974, "learning_rate": 4.157123647495797e-09, "loss": 0.1783, "step": 17095 }, { "epoch": 4.549228312932411, "grad_norm": 0.31619518995285034, "learning_rate": 4.152257025295436e-09, "loss": 0.1817, "step": 17096 }, { "epoch": 4.549494411921235, "grad_norm": 0.2875846028327942, "learning_rate": 4.147393192974025e-09, "loss": 0.1899, "step": 17097 }, { "epoch": 4.549760510910058, "grad_norm": 0.2921355664730072, "learning_rate": 4.142532150673095e-09, "loss": 0.1804, "step": 17098 }, { "epoch": 4.550026609898882, "grad_norm": 0.29556170105934143, "learning_rate": 4.137673898534177e-09, "loss": 0.1878, "step": 17099 }, { "epoch": 4.5502927088877065, "grad_norm": 0.3129917085170746, "learning_rate": 4.132818436698659e-09, "loss": 0.1919, "step": 17100 }, { "epoch": 4.55055880787653, "grad_norm": 0.3068302273750305, "learning_rate": 4.1279657653078835e-09, "loss": 0.1877, "step": 17101 }, { "epoch": 4.550824906865354, "grad_norm": 0.2947317063808441, "learning_rate": 4.123115884503092e-09, "loss": 0.2078, "step": 17102 }, { "epoch": 4.551091005854178, "grad_norm": 0.297559916973114, "learning_rate": 4.118268794425462e-09, "loss": 0.1975, "step": 17103 }, { "epoch": 4.551357104843001, "grad_norm": 0.33791905641555786, "learning_rate": 4.113424495216067e-09, "loss": 0.207, "step": 17104 }, { "epoch": 4.551623203831825, "grad_norm": 0.2821420729160309, "learning_rate": 4.108582987015918e-09, "loss": 0.1882, "step": 17105 }, { "epoch": 4.5518893028206495, "grad_norm": 0.29189664125442505, "learning_rate": 4.103744269965947e-09, "loss": 0.1956, "step": 17106 }, { "epoch": 4.552155401809473, "grad_norm": 0.29062125086784363, "learning_rate": 4.098908344206986e-09, "loss": 0.183, "step": 17107 }, { "epoch": 4.552421500798297, "grad_norm": 0.31526899337768555, "learning_rate": 4.094075209879788e-09, "loss": 0.1847, "step": 17108 }, { "epoch": 4.552687599787121, "grad_norm": 0.28604331612586975, "learning_rate": 4.089244867125052e-09, "loss": 0.1846, "step": 17109 }, { "epoch": 4.552953698775944, "grad_norm": 0.27925345301628113, "learning_rate": 4.084417316083366e-09, "loss": 0.1946, "step": 17110 }, { "epoch": 4.553219797764768, "grad_norm": 0.3362392783164978, "learning_rate": 4.07959255689525e-09, "loss": 0.1788, "step": 17111 }, { "epoch": 4.5534858967535925, "grad_norm": 0.3039853870868683, "learning_rate": 4.074770589701138e-09, "loss": 0.2001, "step": 17112 }, { "epoch": 4.553751995742417, "grad_norm": 0.2946634590625763, "learning_rate": 4.069951414641404e-09, "loss": 0.1925, "step": 17113 }, { "epoch": 4.55401809473124, "grad_norm": 0.2680324912071228, "learning_rate": 4.065135031856282e-09, "loss": 0.1762, "step": 17114 }, { "epoch": 4.554284193720064, "grad_norm": 0.31133440136909485, "learning_rate": 4.0603214414859924e-09, "loss": 0.1731, "step": 17115 }, { "epoch": 4.554550292708887, "grad_norm": 0.2922325134277344, "learning_rate": 4.055510643670657e-09, "loss": 0.1904, "step": 17116 }, { "epoch": 4.554816391697711, "grad_norm": 0.2993568181991577, "learning_rate": 4.050702638550274e-09, "loss": 0.186, "step": 17117 }, { "epoch": 4.5550824906865355, "grad_norm": 0.33369383215904236, "learning_rate": 4.045897426264799e-09, "loss": 0.1881, "step": 17118 }, { "epoch": 4.55534858967536, "grad_norm": 0.31193119287490845, "learning_rate": 4.041095006954121e-09, "loss": 0.1966, "step": 17119 }, { "epoch": 4.555614688664183, "grad_norm": 0.3100355863571167, "learning_rate": 4.036295380758003e-09, "loss": 0.2058, "step": 17120 }, { "epoch": 4.555880787653007, "grad_norm": 0.3053269386291504, "learning_rate": 4.03149854781617e-09, "loss": 0.1952, "step": 17121 }, { "epoch": 4.556146886641831, "grad_norm": 0.29034504294395447, "learning_rate": 4.026704508268219e-09, "loss": 0.1971, "step": 17122 }, { "epoch": 4.556412985630654, "grad_norm": 0.28529879450798035, "learning_rate": 4.021913262253718e-09, "loss": 0.1904, "step": 17123 }, { "epoch": 4.5566790846194785, "grad_norm": 0.33611589670181274, "learning_rate": 4.0171248099121315e-09, "loss": 0.2073, "step": 17124 }, { "epoch": 4.556945183608303, "grad_norm": 0.2977314591407776, "learning_rate": 4.012339151382804e-09, "loss": 0.175, "step": 17125 }, { "epoch": 4.557211282597126, "grad_norm": 0.2872021496295929, "learning_rate": 4.007556286805069e-09, "loss": 0.1706, "step": 17126 }, { "epoch": 4.55747738158595, "grad_norm": 0.3247890770435333, "learning_rate": 4.0027762163181155e-09, "loss": 0.1989, "step": 17127 }, { "epoch": 4.557743480574774, "grad_norm": 0.3010537922382355, "learning_rate": 3.997998940061098e-09, "loss": 0.1961, "step": 17128 }, { "epoch": 4.558009579563597, "grad_norm": 0.34178248047828674, "learning_rate": 3.993224458173062e-09, "loss": 0.2043, "step": 17129 }, { "epoch": 4.558275678552421, "grad_norm": 0.3602190911769867, "learning_rate": 3.9884527707929824e-09, "loss": 0.2292, "step": 17130 }, { "epoch": 4.558541777541246, "grad_norm": 0.32258933782577515, "learning_rate": 3.983683878059752e-09, "loss": 0.1871, "step": 17131 }, { "epoch": 4.558807876530069, "grad_norm": 0.3219045400619507, "learning_rate": 3.978917780112179e-09, "loss": 0.1883, "step": 17132 }, { "epoch": 4.559073975518893, "grad_norm": 0.2712460458278656, "learning_rate": 3.974154477088998e-09, "loss": 0.1589, "step": 17133 }, { "epoch": 4.559340074507717, "grad_norm": 0.26719897985458374, "learning_rate": 3.969393969128853e-09, "loss": 0.1749, "step": 17134 }, { "epoch": 4.55960617349654, "grad_norm": 0.3393428921699524, "learning_rate": 3.964636256370302e-09, "loss": 0.2221, "step": 17135 }, { "epoch": 4.559872272485364, "grad_norm": 0.36460110545158386, "learning_rate": 3.959881338951854e-09, "loss": 0.2146, "step": 17136 }, { "epoch": 4.5601383714741885, "grad_norm": 0.28081661462783813, "learning_rate": 3.955129217011876e-09, "loss": 0.1753, "step": 17137 }, { "epoch": 4.560404470463013, "grad_norm": 0.29200640320777893, "learning_rate": 3.950379890688704e-09, "loss": 0.1751, "step": 17138 }, { "epoch": 4.560670569451836, "grad_norm": 0.30219054222106934, "learning_rate": 3.945633360120593e-09, "loss": 0.1939, "step": 17139 }, { "epoch": 4.56093666844066, "grad_norm": 0.3034994602203369, "learning_rate": 3.940889625445687e-09, "loss": 0.2064, "step": 17140 }, { "epoch": 4.561202767429483, "grad_norm": 0.3208039104938507, "learning_rate": 3.936148686802077e-09, "loss": 0.2076, "step": 17141 }, { "epoch": 4.561468866418307, "grad_norm": 0.3172178268432617, "learning_rate": 3.93141054432774e-09, "loss": 0.2027, "step": 17142 }, { "epoch": 4.5617349654071315, "grad_norm": 0.32450059056282043, "learning_rate": 3.9266751981606225e-09, "loss": 0.2239, "step": 17143 }, { "epoch": 4.562001064395956, "grad_norm": 0.30724453926086426, "learning_rate": 3.921942648438526e-09, "loss": 0.1788, "step": 17144 }, { "epoch": 4.562267163384779, "grad_norm": 0.3139008581638336, "learning_rate": 3.917212895299204e-09, "loss": 0.2078, "step": 17145 }, { "epoch": 4.562533262373603, "grad_norm": 0.2905728220939636, "learning_rate": 3.91248593888035e-09, "loss": 0.1889, "step": 17146 }, { "epoch": 4.562799361362427, "grad_norm": 0.3093746304512024, "learning_rate": 3.90776177931954e-09, "loss": 0.2139, "step": 17147 }, { "epoch": 4.56306546035125, "grad_norm": 0.2970680296421051, "learning_rate": 3.903040416754278e-09, "loss": 0.1883, "step": 17148 }, { "epoch": 4.5633315593400745, "grad_norm": 0.3208879828453064, "learning_rate": 3.898321851322006e-09, "loss": 0.1961, "step": 17149 }, { "epoch": 4.563597658328899, "grad_norm": 0.2856004238128662, "learning_rate": 3.8936060831600505e-09, "loss": 0.1865, "step": 17150 }, { "epoch": 4.563863757317722, "grad_norm": 0.29712313413619995, "learning_rate": 3.888893112405689e-09, "loss": 0.1908, "step": 17151 }, { "epoch": 4.564129856306546, "grad_norm": 0.27062472701072693, "learning_rate": 3.884182939196112e-09, "loss": 0.1677, "step": 17152 }, { "epoch": 4.56439595529537, "grad_norm": 0.2841787338256836, "learning_rate": 3.8794755636683884e-09, "loss": 0.1834, "step": 17153 }, { "epoch": 4.564662054284193, "grad_norm": 0.3124311864376068, "learning_rate": 3.8747709859595634e-09, "loss": 0.1883, "step": 17154 }, { "epoch": 4.5649281532730175, "grad_norm": 0.309818834066391, "learning_rate": 3.870069206206561e-09, "loss": 0.2027, "step": 17155 }, { "epoch": 4.565194252261842, "grad_norm": 0.32839345932006836, "learning_rate": 3.86537022454626e-09, "loss": 0.206, "step": 17156 }, { "epoch": 4.565460351250666, "grad_norm": 0.30162742733955383, "learning_rate": 3.860674041115408e-09, "loss": 0.1893, "step": 17157 }, { "epoch": 4.565726450239489, "grad_norm": 0.28439727425575256, "learning_rate": 3.8559806560507165e-09, "loss": 0.1793, "step": 17158 }, { "epoch": 4.565992549228313, "grad_norm": 0.2960314452648163, "learning_rate": 3.851290069488789e-09, "loss": 0.1915, "step": 17159 }, { "epoch": 4.566258648217136, "grad_norm": 0.30676135420799255, "learning_rate": 3.8466022815661695e-09, "loss": 0.1976, "step": 17160 }, { "epoch": 4.5665247472059605, "grad_norm": 0.3180488646030426, "learning_rate": 3.841917292419294e-09, "loss": 0.1982, "step": 17161 }, { "epoch": 4.566790846194785, "grad_norm": 0.28752753138542175, "learning_rate": 3.837235102184533e-09, "loss": 0.1903, "step": 17162 }, { "epoch": 4.567056945183609, "grad_norm": 0.3214662969112396, "learning_rate": 3.832555710998165e-09, "loss": 0.1909, "step": 17163 }, { "epoch": 4.567323044172432, "grad_norm": 0.31361788511276245, "learning_rate": 3.827879118996414e-09, "loss": 0.1983, "step": 17164 }, { "epoch": 4.567589143161256, "grad_norm": 0.28524550795555115, "learning_rate": 3.823205326315393e-09, "loss": 0.1798, "step": 17165 }, { "epoch": 4.56785524215008, "grad_norm": 0.28683844208717346, "learning_rate": 3.818534333091139e-09, "loss": 0.1779, "step": 17166 }, { "epoch": 4.5681213411389034, "grad_norm": 0.31471309065818787, "learning_rate": 3.81386613945962e-09, "loss": 0.2071, "step": 17167 }, { "epoch": 4.568387440127728, "grad_norm": 0.2846739888191223, "learning_rate": 3.8092007455567066e-09, "loss": 0.182, "step": 17168 }, { "epoch": 4.568653539116552, "grad_norm": 0.2855880856513977, "learning_rate": 3.804538151518211e-09, "loss": 0.1855, "step": 17169 }, { "epoch": 4.568919638105375, "grad_norm": 0.31751519441604614, "learning_rate": 3.799878357479846e-09, "loss": 0.187, "step": 17170 }, { "epoch": 4.569185737094199, "grad_norm": 0.32307544350624084, "learning_rate": 3.795221363577239e-09, "loss": 0.1972, "step": 17171 }, { "epoch": 4.569451836083023, "grad_norm": 0.2942754328250885, "learning_rate": 3.790567169945935e-09, "loss": 0.1879, "step": 17172 }, { "epoch": 4.569717935071846, "grad_norm": 1.6704151630401611, "learning_rate": 3.785915776721416e-09, "loss": 0.1893, "step": 17173 }, { "epoch": 4.5699840340606706, "grad_norm": 0.3038491904735565, "learning_rate": 3.781267184039083e-09, "loss": 0.2054, "step": 17174 }, { "epoch": 4.570250133049495, "grad_norm": 0.27177056670188904, "learning_rate": 3.776621392034229e-09, "loss": 0.1803, "step": 17175 }, { "epoch": 4.570516232038318, "grad_norm": 0.3253997266292572, "learning_rate": 3.771978400842079e-09, "loss": 0.1906, "step": 17176 }, { "epoch": 4.570782331027142, "grad_norm": 0.28750094771385193, "learning_rate": 3.767338210597793e-09, "loss": 0.1946, "step": 17177 }, { "epoch": 4.571048430015966, "grad_norm": 0.287388414144516, "learning_rate": 3.7627008214364154e-09, "loss": 0.1862, "step": 17178 }, { "epoch": 4.571314529004789, "grad_norm": 0.32208582758903503, "learning_rate": 3.758066233492952e-09, "loss": 0.1974, "step": 17179 }, { "epoch": 4.5715806279936135, "grad_norm": 0.29381248354911804, "learning_rate": 3.7534344469022815e-09, "loss": 0.1884, "step": 17180 }, { "epoch": 4.571846726982438, "grad_norm": 0.3019408881664276, "learning_rate": 3.748805461799231e-09, "loss": 0.1937, "step": 17181 }, { "epoch": 4.572112825971262, "grad_norm": 0.30944862961769104, "learning_rate": 3.7441792783185485e-09, "loss": 0.2015, "step": 17182 }, { "epoch": 4.572378924960085, "grad_norm": 0.2755585312843323, "learning_rate": 3.73955589659487e-09, "loss": 0.1838, "step": 17183 }, { "epoch": 4.572645023948909, "grad_norm": 0.3412737548351288, "learning_rate": 3.734935316762789e-09, "loss": 0.1935, "step": 17184 }, { "epoch": 4.572911122937732, "grad_norm": 0.2914769649505615, "learning_rate": 3.7303175389567844e-09, "loss": 0.1911, "step": 17185 }, { "epoch": 4.5731772219265565, "grad_norm": 0.3299455940723419, "learning_rate": 3.725702563311284e-09, "loss": 0.2136, "step": 17186 }, { "epoch": 4.573443320915381, "grad_norm": 0.32444244623184204, "learning_rate": 3.7210903899606017e-09, "loss": 0.2037, "step": 17187 }, { "epoch": 4.573709419904205, "grad_norm": 0.3076153099536896, "learning_rate": 3.7164810190389972e-09, "loss": 0.1791, "step": 17188 }, { "epoch": 4.573975518893028, "grad_norm": 0.30935409665107727, "learning_rate": 3.71187445068063e-09, "loss": 0.1953, "step": 17189 }, { "epoch": 4.574241617881852, "grad_norm": 0.3034173548221588, "learning_rate": 3.7072706850195924e-09, "loss": 0.1927, "step": 17190 }, { "epoch": 4.574507716870676, "grad_norm": 0.29759469628334045, "learning_rate": 3.702669722189877e-09, "loss": 0.1826, "step": 17191 }, { "epoch": 4.5747738158594995, "grad_norm": 0.30017897486686707, "learning_rate": 3.6980715623254112e-09, "loss": 0.1877, "step": 17192 }, { "epoch": 4.575039914848324, "grad_norm": 0.2869594693183899, "learning_rate": 3.693476205560042e-09, "loss": 0.1834, "step": 17193 }, { "epoch": 4.575306013837148, "grad_norm": 0.29647931456565857, "learning_rate": 3.688883652027519e-09, "loss": 0.1839, "step": 17194 }, { "epoch": 4.575572112825971, "grad_norm": 0.2938395142555237, "learning_rate": 3.6842939018615347e-09, "loss": 0.174, "step": 17195 }, { "epoch": 4.575838211814795, "grad_norm": 0.3050118088722229, "learning_rate": 3.6797069551956606e-09, "loss": 0.1869, "step": 17196 }, { "epoch": 4.576104310803619, "grad_norm": 0.32049891352653503, "learning_rate": 3.6751228121634226e-09, "loss": 0.2136, "step": 17197 }, { "epoch": 4.5763704097924425, "grad_norm": 0.29484835267066956, "learning_rate": 3.67054147289827e-09, "loss": 0.1854, "step": 17198 }, { "epoch": 4.576636508781267, "grad_norm": 0.3071902096271515, "learning_rate": 3.665962937533529e-09, "loss": 0.1882, "step": 17199 }, { "epoch": 4.576902607770091, "grad_norm": 0.2892853915691376, "learning_rate": 3.661387206202482e-09, "loss": 0.1953, "step": 17200 }, { "epoch": 4.577168706758914, "grad_norm": 0.2879592776298523, "learning_rate": 3.6568142790383003e-09, "loss": 0.2019, "step": 17201 }, { "epoch": 4.577434805747738, "grad_norm": 0.30730581283569336, "learning_rate": 3.652244156174111e-09, "loss": 0.1796, "step": 17202 }, { "epoch": 4.577700904736562, "grad_norm": 0.3014034032821655, "learning_rate": 3.64767683774293e-09, "loss": 0.1924, "step": 17203 }, { "epoch": 4.5779670037253855, "grad_norm": 0.2874353229999542, "learning_rate": 3.643112323877695e-09, "loss": 0.178, "step": 17204 }, { "epoch": 4.57823310271421, "grad_norm": 0.26830530166625977, "learning_rate": 3.6385506147112775e-09, "loss": 0.1625, "step": 17205 }, { "epoch": 4.578499201703034, "grad_norm": 0.2893565595149994, "learning_rate": 3.6339917103764604e-09, "loss": 0.1805, "step": 17206 }, { "epoch": 4.578765300691858, "grad_norm": 0.30649638175964355, "learning_rate": 3.6294356110059156e-09, "loss": 0.1995, "step": 17207 }, { "epoch": 4.579031399680681, "grad_norm": 0.2951407730579376, "learning_rate": 3.6248823167322805e-09, "loss": 0.1782, "step": 17208 }, { "epoch": 4.579297498669505, "grad_norm": 0.28678736090660095, "learning_rate": 3.6203318276880944e-09, "loss": 0.1773, "step": 17209 }, { "epoch": 4.579563597658328, "grad_norm": 0.2887110114097595, "learning_rate": 3.6157841440057958e-09, "loss": 0.1784, "step": 17210 }, { "epoch": 4.579829696647153, "grad_norm": 0.3147416412830353, "learning_rate": 3.611239265817756e-09, "loss": 0.1925, "step": 17211 }, { "epoch": 4.580095795635977, "grad_norm": 0.30130136013031006, "learning_rate": 3.6066971932562805e-09, "loss": 0.172, "step": 17212 }, { "epoch": 4.580361894624801, "grad_norm": 0.2953716516494751, "learning_rate": 3.602157926453553e-09, "loss": 0.1796, "step": 17213 }, { "epoch": 4.580627993613624, "grad_norm": 0.289598286151886, "learning_rate": 3.5976214655417334e-09, "loss": 0.1834, "step": 17214 }, { "epoch": 4.580894092602448, "grad_norm": 0.31236326694488525, "learning_rate": 3.5930878106528395e-09, "loss": 0.1862, "step": 17215 }, { "epoch": 4.581160191591272, "grad_norm": 0.3351087272167206, "learning_rate": 3.5885569619188315e-09, "loss": 0.1971, "step": 17216 }, { "epoch": 4.5814262905800955, "grad_norm": 0.29243093729019165, "learning_rate": 3.5840289194716266e-09, "loss": 0.1745, "step": 17217 }, { "epoch": 4.58169238956892, "grad_norm": 0.30213141441345215, "learning_rate": 3.579503683442975e-09, "loss": 0.1876, "step": 17218 }, { "epoch": 4.581958488557744, "grad_norm": 0.2884557843208313, "learning_rate": 3.574981253964615e-09, "loss": 0.2045, "step": 17219 }, { "epoch": 4.582224587546567, "grad_norm": 0.28508031368255615, "learning_rate": 3.570461631168198e-09, "loss": 0.1968, "step": 17220 }, { "epoch": 4.582490686535391, "grad_norm": 0.30640798807144165, "learning_rate": 3.5659448151852622e-09, "loss": 0.1985, "step": 17221 }, { "epoch": 4.582756785524215, "grad_norm": 0.31372374296188354, "learning_rate": 3.561430806147281e-09, "loss": 0.1749, "step": 17222 }, { "epoch": 4.5830228845130385, "grad_norm": 0.2872796356678009, "learning_rate": 3.5569196041856485e-09, "loss": 0.1851, "step": 17223 }, { "epoch": 4.583288983501863, "grad_norm": 0.2883966565132141, "learning_rate": 3.5524112094316826e-09, "loss": 0.1815, "step": 17224 }, { "epoch": 4.583555082490687, "grad_norm": 0.3194229304790497, "learning_rate": 3.5479056220166005e-09, "loss": 0.1771, "step": 17225 }, { "epoch": 4.58382118147951, "grad_norm": 0.29768380522727966, "learning_rate": 3.5434028420715413e-09, "loss": 0.1719, "step": 17226 }, { "epoch": 4.584087280468334, "grad_norm": 0.30266597867012024, "learning_rate": 3.538902869727578e-09, "loss": 0.1929, "step": 17227 }, { "epoch": 4.584353379457158, "grad_norm": 0.3174716532230377, "learning_rate": 3.5344057051157062e-09, "loss": 0.1944, "step": 17228 }, { "epoch": 4.5846194784459815, "grad_norm": 0.2875393331050873, "learning_rate": 3.5299113483667988e-09, "loss": 0.1821, "step": 17229 }, { "epoch": 4.584885577434806, "grad_norm": 0.28571557998657227, "learning_rate": 3.5254197996116953e-09, "loss": 0.1767, "step": 17230 }, { "epoch": 4.58515167642363, "grad_norm": 0.3228575885295868, "learning_rate": 3.520931058981136e-09, "loss": 0.1899, "step": 17231 }, { "epoch": 4.585417775412454, "grad_norm": 0.2935628294944763, "learning_rate": 3.5164451266057605e-09, "loss": 0.1855, "step": 17232 }, { "epoch": 4.585683874401277, "grad_norm": 0.32444295287132263, "learning_rate": 3.5119620026161644e-09, "loss": 0.1957, "step": 17233 }, { "epoch": 4.585949973390101, "grad_norm": 0.3203735947608948, "learning_rate": 3.5074816871428105e-09, "loss": 0.1931, "step": 17234 }, { "epoch": 4.5862160723789245, "grad_norm": 0.2922413647174835, "learning_rate": 3.503004180316127e-09, "loss": 0.2021, "step": 17235 }, { "epoch": 4.586482171367749, "grad_norm": 0.34277328848838806, "learning_rate": 3.4985294822664324e-09, "loss": 0.1926, "step": 17236 }, { "epoch": 4.586748270356573, "grad_norm": 0.31031519174575806, "learning_rate": 3.4940575931240002e-09, "loss": 0.2056, "step": 17237 }, { "epoch": 4.587014369345397, "grad_norm": 0.2880367338657379, "learning_rate": 3.4895885130189705e-09, "loss": 0.1818, "step": 17238 }, { "epoch": 4.58728046833422, "grad_norm": 0.3181264400482178, "learning_rate": 3.4851222420814173e-09, "loss": 0.1806, "step": 17239 }, { "epoch": 4.587546567323044, "grad_norm": 0.29756295680999756, "learning_rate": 3.48065878044137e-09, "loss": 0.2096, "step": 17240 }, { "epoch": 4.587812666311868, "grad_norm": 0.311400443315506, "learning_rate": 3.4761981282287356e-09, "loss": 0.1926, "step": 17241 }, { "epoch": 4.588078765300692, "grad_norm": 0.28587964177131653, "learning_rate": 3.4717402855733546e-09, "loss": 0.1981, "step": 17242 }, { "epoch": 4.588344864289516, "grad_norm": 0.30630362033843994, "learning_rate": 3.4672852526049788e-09, "loss": 0.199, "step": 17243 }, { "epoch": 4.58861096327834, "grad_norm": 0.27760231494903564, "learning_rate": 3.4628330294532826e-09, "loss": 0.1789, "step": 17244 }, { "epoch": 4.588877062267163, "grad_norm": 0.2879256010055542, "learning_rate": 3.4583836162478506e-09, "loss": 0.18, "step": 17245 }, { "epoch": 4.589143161255987, "grad_norm": 0.2986876666545868, "learning_rate": 3.453937013118213e-09, "loss": 0.1978, "step": 17246 }, { "epoch": 4.589409260244811, "grad_norm": 0.3121926784515381, "learning_rate": 3.4494932201937887e-09, "loss": 0.1754, "step": 17247 }, { "epoch": 4.589675359233635, "grad_norm": 0.30383989214897156, "learning_rate": 3.4450522376039293e-09, "loss": 0.2066, "step": 17248 }, { "epoch": 4.589941458222459, "grad_norm": 0.320679247379303, "learning_rate": 3.440614065477887e-09, "loss": 0.1896, "step": 17249 }, { "epoch": 4.590207557211283, "grad_norm": 0.2910504937171936, "learning_rate": 3.4361787039448587e-09, "loss": 0.1887, "step": 17250 }, { "epoch": 4.590473656200106, "grad_norm": 0.30303657054901123, "learning_rate": 3.431746153133963e-09, "loss": 0.1755, "step": 17251 }, { "epoch": 4.59073975518893, "grad_norm": 0.3274526298046112, "learning_rate": 3.427316413174175e-09, "loss": 0.1861, "step": 17252 }, { "epoch": 4.591005854177754, "grad_norm": 0.30453839898109436, "learning_rate": 3.422889484194469e-09, "loss": 0.1873, "step": 17253 }, { "epoch": 4.5912719531665775, "grad_norm": 0.2877132296562195, "learning_rate": 3.4184653663236863e-09, "loss": 0.1851, "step": 17254 }, { "epoch": 4.591538052155402, "grad_norm": 0.29108625650405884, "learning_rate": 3.4140440596906016e-09, "loss": 0.1905, "step": 17255 }, { "epoch": 4.591804151144226, "grad_norm": 0.2958393692970276, "learning_rate": 3.4096255644239125e-09, "loss": 0.1857, "step": 17256 }, { "epoch": 4.59207025013305, "grad_norm": 0.2967596650123596, "learning_rate": 3.405209880652249e-09, "loss": 0.1928, "step": 17257 }, { "epoch": 4.592336349121873, "grad_norm": 0.3179309070110321, "learning_rate": 3.400797008504097e-09, "loss": 0.188, "step": 17258 }, { "epoch": 4.592602448110697, "grad_norm": 0.2924671471118927, "learning_rate": 3.3963869481079322e-09, "loss": 0.1884, "step": 17259 }, { "epoch": 4.5928685470995205, "grad_norm": 0.2999130189418793, "learning_rate": 3.391979699592129e-09, "loss": 0.2005, "step": 17260 }, { "epoch": 4.593134646088345, "grad_norm": 0.2852519154548645, "learning_rate": 3.3875752630849407e-09, "loss": 0.1681, "step": 17261 }, { "epoch": 4.593400745077169, "grad_norm": 0.3381808400154114, "learning_rate": 3.383173638714587e-09, "loss": 0.2146, "step": 17262 }, { "epoch": 4.593666844065993, "grad_norm": 0.2958924174308777, "learning_rate": 3.3787748266091874e-09, "loss": 0.1883, "step": 17263 }, { "epoch": 4.593932943054816, "grad_norm": 0.3049115240573883, "learning_rate": 3.3743788268967732e-09, "loss": 0.182, "step": 17264 }, { "epoch": 4.59419904204364, "grad_norm": 0.311662882566452, "learning_rate": 3.3699856397053085e-09, "loss": 0.1934, "step": 17265 }, { "epoch": 4.594465141032464, "grad_norm": 0.31119173765182495, "learning_rate": 3.3655952651626683e-09, "loss": 0.2071, "step": 17266 }, { "epoch": 4.594731240021288, "grad_norm": 0.2940139174461365, "learning_rate": 3.361207703396629e-09, "loss": 0.1863, "step": 17267 }, { "epoch": 4.594997339010112, "grad_norm": 0.29134756326675415, "learning_rate": 3.3568229545349214e-09, "loss": 0.1783, "step": 17268 }, { "epoch": 4.595263437998936, "grad_norm": 0.2916785478591919, "learning_rate": 3.3524410187051766e-09, "loss": 0.1717, "step": 17269 }, { "epoch": 4.595529536987759, "grad_norm": 0.2945447862148285, "learning_rate": 3.348061896034926e-09, "loss": 0.1862, "step": 17270 }, { "epoch": 4.595795635976583, "grad_norm": 0.2631802558898926, "learning_rate": 3.3436855866516346e-09, "loss": 0.1757, "step": 17271 }, { "epoch": 4.596061734965407, "grad_norm": 0.28583505749702454, "learning_rate": 3.339312090682689e-09, "loss": 0.1827, "step": 17272 }, { "epoch": 4.596327833954231, "grad_norm": 0.2729727029800415, "learning_rate": 3.334941408255387e-09, "loss": 0.1738, "step": 17273 }, { "epoch": 4.596593932943055, "grad_norm": 0.2749752998352051, "learning_rate": 3.33057353949695e-09, "loss": 0.1742, "step": 17274 }, { "epoch": 4.596860031931879, "grad_norm": 0.30880919098854065, "learning_rate": 3.326208484534532e-09, "loss": 0.2055, "step": 17275 }, { "epoch": 4.597126130920703, "grad_norm": 0.317277193069458, "learning_rate": 3.321846243495174e-09, "loss": 0.1942, "step": 17276 }, { "epoch": 4.597392229909526, "grad_norm": 0.28189700841903687, "learning_rate": 3.3174868165058433e-09, "loss": 0.1696, "step": 17277 }, { "epoch": 4.59765832889835, "grad_norm": 0.30869612097740173, "learning_rate": 3.3131302036934373e-09, "loss": 0.1801, "step": 17278 }, { "epoch": 4.597924427887174, "grad_norm": 0.3310002386569977, "learning_rate": 3.3087764051847767e-09, "loss": 0.1955, "step": 17279 }, { "epoch": 4.598190526875998, "grad_norm": 0.3246479332447052, "learning_rate": 3.3044254211065714e-09, "loss": 0.2225, "step": 17280 }, { "epoch": 4.598456625864822, "grad_norm": 0.28848010301589966, "learning_rate": 3.3000772515854645e-09, "loss": 0.1881, "step": 17281 }, { "epoch": 4.598722724853646, "grad_norm": 0.3094601631164551, "learning_rate": 3.295731896748033e-09, "loss": 0.1811, "step": 17282 }, { "epoch": 4.598988823842469, "grad_norm": 0.2873231768608093, "learning_rate": 3.291389356720764e-09, "loss": 0.184, "step": 17283 }, { "epoch": 4.599254922831293, "grad_norm": 0.28207632899284363, "learning_rate": 3.2870496316300454e-09, "loss": 0.1741, "step": 17284 }, { "epoch": 4.5995210218201175, "grad_norm": 0.2931702733039856, "learning_rate": 3.2827127216021986e-09, "loss": 0.1867, "step": 17285 }, { "epoch": 4.599787120808941, "grad_norm": 0.34052377939224243, "learning_rate": 3.278378626763456e-09, "loss": 0.2148, "step": 17286 }, { "epoch": 4.600053219797765, "grad_norm": 0.3028867840766907, "learning_rate": 3.2740473472399943e-09, "loss": 0.197, "step": 17287 }, { "epoch": 4.600319318786589, "grad_norm": 0.3239157795906067, "learning_rate": 3.2697188831578572e-09, "loss": 0.1939, "step": 17288 }, { "epoch": 4.600585417775412, "grad_norm": 0.30221179127693176, "learning_rate": 3.2653932346430323e-09, "loss": 0.2063, "step": 17289 }, { "epoch": 4.600851516764236, "grad_norm": 0.3160322904586792, "learning_rate": 3.261070401821453e-09, "loss": 0.2045, "step": 17290 }, { "epoch": 4.60111761575306, "grad_norm": 0.3231625556945801, "learning_rate": 3.2567503848189404e-09, "loss": 0.1841, "step": 17291 }, { "epoch": 4.601383714741884, "grad_norm": 0.29387712478637695, "learning_rate": 3.2524331837612273e-09, "loss": 0.1823, "step": 17292 }, { "epoch": 4.601649813730708, "grad_norm": 0.2902500331401825, "learning_rate": 3.248118798773969e-09, "loss": 0.1946, "step": 17293 }, { "epoch": 4.601915912719532, "grad_norm": 0.3043900728225708, "learning_rate": 3.2438072299827757e-09, "loss": 0.2029, "step": 17294 }, { "epoch": 4.602182011708355, "grad_norm": 0.290779173374176, "learning_rate": 3.239498477513125e-09, "loss": 0.1778, "step": 17295 }, { "epoch": 4.602448110697179, "grad_norm": 0.30692705512046814, "learning_rate": 3.2351925414904393e-09, "loss": 0.2063, "step": 17296 }, { "epoch": 4.602714209686003, "grad_norm": 0.3277023434638977, "learning_rate": 3.230889422040051e-09, "loss": 0.2092, "step": 17297 }, { "epoch": 4.602980308674827, "grad_norm": 0.2971958816051483, "learning_rate": 3.226589119287215e-09, "loss": 0.1915, "step": 17298 }, { "epoch": 4.603246407663651, "grad_norm": 0.316224068403244, "learning_rate": 3.2222916333570994e-09, "loss": 0.1853, "step": 17299 }, { "epoch": 4.603512506652475, "grad_norm": 0.30917811393737793, "learning_rate": 3.2179969643747915e-09, "loss": 0.188, "step": 17300 }, { "epoch": 4.603778605641299, "grad_norm": 0.277628093957901, "learning_rate": 3.2137051124652927e-09, "loss": 0.166, "step": 17301 }, { "epoch": 4.604044704630122, "grad_norm": 0.2976731061935425, "learning_rate": 3.209416077753546e-09, "loss": 0.1732, "step": 17302 }, { "epoch": 4.604310803618946, "grad_norm": 0.31361907720565796, "learning_rate": 3.205129860364375e-09, "loss": 0.2033, "step": 17303 }, { "epoch": 4.60457690260777, "grad_norm": 0.31397634744644165, "learning_rate": 3.200846460422546e-09, "loss": 0.1968, "step": 17304 }, { "epoch": 4.604843001596594, "grad_norm": 0.3060181736946106, "learning_rate": 3.1965658780527595e-09, "loss": 0.1964, "step": 17305 }, { "epoch": 4.605109100585418, "grad_norm": 0.31154632568359375, "learning_rate": 3.192288113379582e-09, "loss": 0.194, "step": 17306 }, { "epoch": 4.605375199574242, "grad_norm": 0.296684592962265, "learning_rate": 3.1880131665275255e-09, "loss": 0.1898, "step": 17307 }, { "epoch": 4.605641298563065, "grad_norm": 0.2988927960395813, "learning_rate": 3.183741037621046e-09, "loss": 0.18, "step": 17308 }, { "epoch": 4.605907397551889, "grad_norm": 0.2851238250732422, "learning_rate": 3.1794717267844884e-09, "loss": 0.1805, "step": 17309 }, { "epoch": 4.6061734965407135, "grad_norm": 0.3056563436985016, "learning_rate": 3.175205234142109e-09, "loss": 0.1953, "step": 17310 }, { "epoch": 4.606439595529537, "grad_norm": 0.302670955657959, "learning_rate": 3.1709415598181077e-09, "loss": 0.1775, "step": 17311 }, { "epoch": 4.606705694518361, "grad_norm": 0.3366966247558594, "learning_rate": 3.166680703936575e-09, "loss": 0.2001, "step": 17312 }, { "epoch": 4.606971793507185, "grad_norm": 0.2728709876537323, "learning_rate": 3.1624226666215446e-09, "loss": 0.1635, "step": 17313 }, { "epoch": 4.607237892496008, "grad_norm": 0.30597618222236633, "learning_rate": 3.158167447996973e-09, "loss": 0.203, "step": 17314 }, { "epoch": 4.607503991484832, "grad_norm": 0.2995619475841522, "learning_rate": 3.153915048186684e-09, "loss": 0.1776, "step": 17315 }, { "epoch": 4.6077700904736565, "grad_norm": 0.31628477573394775, "learning_rate": 3.149665467314466e-09, "loss": 0.2093, "step": 17316 }, { "epoch": 4.60803618946248, "grad_norm": 0.31206589937210083, "learning_rate": 3.1454187055040214e-09, "loss": 0.2052, "step": 17317 }, { "epoch": 4.608302288451304, "grad_norm": 0.285136878490448, "learning_rate": 3.141174762878962e-09, "loss": 0.1792, "step": 17318 }, { "epoch": 4.608568387440128, "grad_norm": 0.275995671749115, "learning_rate": 3.136933639562811e-09, "loss": 0.1685, "step": 17319 }, { "epoch": 4.608834486428951, "grad_norm": 0.2844887971878052, "learning_rate": 3.132695335679014e-09, "loss": 0.1842, "step": 17320 }, { "epoch": 4.609100585417775, "grad_norm": 0.32362979650497437, "learning_rate": 3.1284598513509507e-09, "loss": 0.198, "step": 17321 }, { "epoch": 4.6093666844065995, "grad_norm": 0.3013255000114441, "learning_rate": 3.124227186701889e-09, "loss": 0.1947, "step": 17322 }, { "epoch": 4.609632783395423, "grad_norm": 0.29220619797706604, "learning_rate": 3.119997341855052e-09, "loss": 0.192, "step": 17323 }, { "epoch": 4.609898882384247, "grad_norm": 0.2793041467666626, "learning_rate": 3.1157703169335303e-09, "loss": 0.164, "step": 17324 }, { "epoch": 4.610164981373071, "grad_norm": 0.29348334670066833, "learning_rate": 3.1115461120603815e-09, "loss": 0.2024, "step": 17325 }, { "epoch": 4.610431080361895, "grad_norm": 0.29183968901634216, "learning_rate": 3.1073247273585622e-09, "loss": 0.1787, "step": 17326 }, { "epoch": 4.610697179350718, "grad_norm": 0.3046170771121979, "learning_rate": 3.1031061629509303e-09, "loss": 0.1885, "step": 17327 }, { "epoch": 4.6109632783395424, "grad_norm": 0.2739183306694031, "learning_rate": 3.098890418960298e-09, "loss": 0.1626, "step": 17328 }, { "epoch": 4.611229377328366, "grad_norm": 0.2930525243282318, "learning_rate": 3.0946774955093567e-09, "loss": 0.2043, "step": 17329 }, { "epoch": 4.61149547631719, "grad_norm": 0.27512380480766296, "learning_rate": 3.0904673927207415e-09, "loss": 0.1853, "step": 17330 }, { "epoch": 4.611761575306014, "grad_norm": 0.29650822281837463, "learning_rate": 3.0862601107169984e-09, "loss": 0.1923, "step": 17331 }, { "epoch": 4.612027674294838, "grad_norm": 0.2906056344509125, "learning_rate": 3.0820556496206073e-09, "loss": 0.1936, "step": 17332 }, { "epoch": 4.612293773283661, "grad_norm": 0.2745620608329773, "learning_rate": 3.077854009553915e-09, "loss": 0.1692, "step": 17333 }, { "epoch": 4.612559872272485, "grad_norm": 0.32120922207832336, "learning_rate": 3.073655190639235e-09, "loss": 0.199, "step": 17334 }, { "epoch": 4.6128259712613096, "grad_norm": 0.2997715473175049, "learning_rate": 3.06945919299878e-09, "loss": 0.1917, "step": 17335 }, { "epoch": 4.613092070250133, "grad_norm": 0.2852616310119629, "learning_rate": 3.0652660167546972e-09, "loss": 0.1897, "step": 17336 }, { "epoch": 4.613358169238957, "grad_norm": 0.35392794013023376, "learning_rate": 3.0610756620290336e-09, "loss": 0.1993, "step": 17337 }, { "epoch": 4.613624268227781, "grad_norm": 0.26889875531196594, "learning_rate": 3.056888128943769e-09, "loss": 0.1723, "step": 17338 }, { "epoch": 4.613890367216604, "grad_norm": 0.310209184885025, "learning_rate": 3.052703417620772e-09, "loss": 0.1862, "step": 17339 }, { "epoch": 4.614156466205428, "grad_norm": 0.3011919856071472, "learning_rate": 3.0485215281818576e-09, "loss": 0.1816, "step": 17340 }, { "epoch": 4.6144225651942525, "grad_norm": 0.2835847735404968, "learning_rate": 3.04434246074875e-09, "loss": 0.1841, "step": 17341 }, { "epoch": 4.614688664183076, "grad_norm": 0.32287707924842834, "learning_rate": 3.040166215443085e-09, "loss": 0.1919, "step": 17342 }, { "epoch": 4.6149547631719, "grad_norm": 0.2985996901988983, "learning_rate": 3.0359927923864324e-09, "loss": 0.185, "step": 17343 }, { "epoch": 4.615220862160724, "grad_norm": 0.29250895977020264, "learning_rate": 3.031822191700262e-09, "loss": 0.1766, "step": 17344 }, { "epoch": 4.615486961149547, "grad_norm": 0.32532286643981934, "learning_rate": 3.0276544135059756e-09, "loss": 0.2154, "step": 17345 }, { "epoch": 4.615753060138371, "grad_norm": 0.3112773001194, "learning_rate": 3.023489457924877e-09, "loss": 0.2007, "step": 17346 }, { "epoch": 4.6160191591271955, "grad_norm": 0.3072455823421478, "learning_rate": 3.0193273250782026e-09, "loss": 0.2057, "step": 17347 }, { "epoch": 4.616285258116019, "grad_norm": 0.3088287115097046, "learning_rate": 3.0151680150871108e-09, "loss": 0.1889, "step": 17348 }, { "epoch": 4.616551357104843, "grad_norm": 0.2821464240550995, "learning_rate": 3.011011528072649e-09, "loss": 0.1697, "step": 17349 }, { "epoch": 4.616817456093667, "grad_norm": 0.30713537335395813, "learning_rate": 3.0068578641558316e-09, "loss": 0.1768, "step": 17350 }, { "epoch": 4.617083555082491, "grad_norm": 0.27381131052970886, "learning_rate": 3.002707023457529e-09, "loss": 0.1796, "step": 17351 }, { "epoch": 4.617349654071314, "grad_norm": 0.3005678653717041, "learning_rate": 2.9985590060985666e-09, "loss": 0.1996, "step": 17352 }, { "epoch": 4.6176157530601385, "grad_norm": 0.33622559905052185, "learning_rate": 2.9944138121997032e-09, "loss": 0.1987, "step": 17353 }, { "epoch": 4.617881852048962, "grad_norm": 0.297324538230896, "learning_rate": 2.9902714418815646e-09, "loss": 0.1913, "step": 17354 }, { "epoch": 4.618147951037786, "grad_norm": 0.29206740856170654, "learning_rate": 2.9861318952647542e-09, "loss": 0.1897, "step": 17355 }, { "epoch": 4.61841405002661, "grad_norm": 0.2965684235095978, "learning_rate": 2.981995172469742e-09, "loss": 0.1807, "step": 17356 }, { "epoch": 4.618680149015434, "grad_norm": 0.29398220777511597, "learning_rate": 2.9778612736169663e-09, "loss": 0.1829, "step": 17357 }, { "epoch": 4.618946248004257, "grad_norm": 0.28786277770996094, "learning_rate": 2.9737301988267182e-09, "loss": 0.1704, "step": 17358 }, { "epoch": 4.6192123469930815, "grad_norm": 0.3066064715385437, "learning_rate": 2.969601948219258e-09, "loss": 0.1912, "step": 17359 }, { "epoch": 4.619478445981906, "grad_norm": 0.33934587240219116, "learning_rate": 2.965476521914756e-09, "loss": 0.2018, "step": 17360 }, { "epoch": 4.619744544970729, "grad_norm": 0.3220568895339966, "learning_rate": 2.9613539200332717e-09, "loss": 0.2075, "step": 17361 }, { "epoch": 4.620010643959553, "grad_norm": 0.3111827075481415, "learning_rate": 2.95723414269482e-09, "loss": 0.1743, "step": 17362 }, { "epoch": 4.620276742948377, "grad_norm": 0.31484195590019226, "learning_rate": 2.9531171900193162e-09, "loss": 0.1922, "step": 17363 }, { "epoch": 4.6205428419372, "grad_norm": 0.27018019556999207, "learning_rate": 2.949003062126587e-09, "loss": 0.1672, "step": 17364 }, { "epoch": 4.6208089409260245, "grad_norm": 0.31550922989845276, "learning_rate": 2.944891759136392e-09, "loss": 0.1936, "step": 17365 }, { "epoch": 4.621075039914849, "grad_norm": 0.30888882279396057, "learning_rate": 2.940783281168391e-09, "loss": 0.2006, "step": 17366 }, { "epoch": 4.621341138903672, "grad_norm": 0.28784918785095215, "learning_rate": 2.936677628342188e-09, "loss": 0.1965, "step": 17367 }, { "epoch": 4.621607237892496, "grad_norm": 0.2969963550567627, "learning_rate": 2.9325748007772766e-09, "loss": 0.1906, "step": 17368 }, { "epoch": 4.62187333688132, "grad_norm": 0.28416821360588074, "learning_rate": 2.9284747985930723e-09, "loss": 0.1884, "step": 17369 }, { "epoch": 4.622139435870143, "grad_norm": 0.27783167362213135, "learning_rate": 2.9243776219089243e-09, "loss": 0.187, "step": 17370 }, { "epoch": 4.622405534858967, "grad_norm": 0.28238481283187866, "learning_rate": 2.920283270844093e-09, "loss": 0.1892, "step": 17371 }, { "epoch": 4.622671633847792, "grad_norm": 0.3123825788497925, "learning_rate": 2.9161917455177487e-09, "loss": 0.1894, "step": 17372 }, { "epoch": 4.622937732836615, "grad_norm": 0.3439878523349762, "learning_rate": 2.912103046048986e-09, "loss": 0.2216, "step": 17373 }, { "epoch": 4.623203831825439, "grad_norm": 0.2999209463596344, "learning_rate": 2.908017172556809e-09, "loss": 0.1845, "step": 17374 }, { "epoch": 4.623469930814263, "grad_norm": 0.3023751974105835, "learning_rate": 2.903934125160168e-09, "loss": 0.1832, "step": 17375 }, { "epoch": 4.623736029803087, "grad_norm": 0.26877328753471375, "learning_rate": 2.899853903977889e-09, "loss": 0.1709, "step": 17376 }, { "epoch": 4.62400212879191, "grad_norm": 0.310979962348938, "learning_rate": 2.8957765091287555e-09, "loss": 0.19, "step": 17377 }, { "epoch": 4.6242682277807345, "grad_norm": 0.3102288544178009, "learning_rate": 2.8917019407314392e-09, "loss": 0.2014, "step": 17378 }, { "epoch": 4.624534326769558, "grad_norm": 0.27815428376197815, "learning_rate": 2.8876301989045336e-09, "loss": 0.1716, "step": 17379 }, { "epoch": 4.624800425758382, "grad_norm": 0.3222835659980774, "learning_rate": 2.8835612837665667e-09, "loss": 0.1947, "step": 17380 }, { "epoch": 4.625066524747206, "grad_norm": 0.28549477458000183, "learning_rate": 2.879495195435966e-09, "loss": 0.1761, "step": 17381 }, { "epoch": 4.62533262373603, "grad_norm": 0.28079235553741455, "learning_rate": 2.875431934031092e-09, "loss": 0.1804, "step": 17382 }, { "epoch": 4.625598722724853, "grad_norm": 0.29112347960472107, "learning_rate": 2.8713714996702054e-09, "loss": 0.1721, "step": 17383 }, { "epoch": 4.6258648217136775, "grad_norm": 0.29709985852241516, "learning_rate": 2.8673138924715125e-09, "loss": 0.1878, "step": 17384 }, { "epoch": 4.626130920702502, "grad_norm": 0.3055919110774994, "learning_rate": 2.863259112553107e-09, "loss": 0.1888, "step": 17385 }, { "epoch": 4.626397019691325, "grad_norm": 0.2863892912864685, "learning_rate": 2.859207160033006e-09, "loss": 0.19, "step": 17386 }, { "epoch": 4.626663118680149, "grad_norm": 0.3337599039077759, "learning_rate": 2.8551580350291816e-09, "loss": 0.2163, "step": 17387 }, { "epoch": 4.626929217668973, "grad_norm": 0.28988227248191833, "learning_rate": 2.851111737659462e-09, "loss": 0.1804, "step": 17388 }, { "epoch": 4.627195316657796, "grad_norm": 0.30682772397994995, "learning_rate": 2.84706826804163e-09, "loss": 0.1977, "step": 17389 }, { "epoch": 4.6274614156466205, "grad_norm": 0.2906365394592285, "learning_rate": 2.8430276262933928e-09, "loss": 0.1831, "step": 17390 }, { "epoch": 4.627727514635445, "grad_norm": 0.28945669531822205, "learning_rate": 2.838989812532344e-09, "loss": 0.1895, "step": 17391 }, { "epoch": 4.627993613624268, "grad_norm": 0.28389543294906616, "learning_rate": 2.8349548268760348e-09, "loss": 0.1934, "step": 17392 }, { "epoch": 4.628259712613092, "grad_norm": 0.3015410304069519, "learning_rate": 2.8309226694419042e-09, "loss": 0.1789, "step": 17393 }, { "epoch": 4.628525811601916, "grad_norm": 0.30129727721214294, "learning_rate": 2.826893340347314e-09, "loss": 0.1904, "step": 17394 }, { "epoch": 4.62879191059074, "grad_norm": 0.30262744426727295, "learning_rate": 2.822866839709559e-09, "loss": 0.1785, "step": 17395 }, { "epoch": 4.6290580095795635, "grad_norm": 0.31497058272361755, "learning_rate": 2.8188431676458345e-09, "loss": 0.1863, "step": 17396 }, { "epoch": 4.629324108568388, "grad_norm": 0.3445032238960266, "learning_rate": 2.8148223242732473e-09, "loss": 0.1803, "step": 17397 }, { "epoch": 4.629590207557211, "grad_norm": 0.3046934902667999, "learning_rate": 2.8108043097088475e-09, "loss": 0.1877, "step": 17398 }, { "epoch": 4.629856306546035, "grad_norm": 0.290572851896286, "learning_rate": 2.806789124069586e-09, "loss": 0.1932, "step": 17399 }, { "epoch": 4.630122405534859, "grad_norm": 0.3490256071090698, "learning_rate": 2.8027767674723366e-09, "loss": 0.2044, "step": 17400 }, { "epoch": 4.630388504523683, "grad_norm": 0.298782080411911, "learning_rate": 2.7987672400338833e-09, "loss": 0.1717, "step": 17401 }, { "epoch": 4.6306546035125065, "grad_norm": 0.31424224376678467, "learning_rate": 2.794760541870933e-09, "loss": 0.184, "step": 17402 }, { "epoch": 4.630920702501331, "grad_norm": 0.3206009864807129, "learning_rate": 2.790756673100103e-09, "loss": 0.2194, "step": 17403 }, { "epoch": 4.631186801490155, "grad_norm": 0.30195608735084534, "learning_rate": 2.786755633837956e-09, "loss": 0.2164, "step": 17404 }, { "epoch": 4.631452900478978, "grad_norm": 0.3008001744747162, "learning_rate": 2.782757424200943e-09, "loss": 0.164, "step": 17405 }, { "epoch": 4.631718999467802, "grad_norm": 0.32674646377563477, "learning_rate": 2.7787620443054383e-09, "loss": 0.2039, "step": 17406 }, { "epoch": 4.631985098456626, "grad_norm": 0.27266091108322144, "learning_rate": 2.774769494267737e-09, "loss": 0.1937, "step": 17407 }, { "epoch": 4.632251197445449, "grad_norm": 0.30880051851272583, "learning_rate": 2.770779774204046e-09, "loss": 0.1992, "step": 17408 }, { "epoch": 4.632517296434274, "grad_norm": 0.33743828535079956, "learning_rate": 2.7667928842305066e-09, "loss": 0.1994, "step": 17409 }, { "epoch": 4.632783395423098, "grad_norm": 0.2877972722053528, "learning_rate": 2.7628088244631585e-09, "loss": 0.2005, "step": 17410 }, { "epoch": 4.633049494411921, "grad_norm": 0.294467031955719, "learning_rate": 2.758827595017976e-09, "loss": 0.1769, "step": 17411 }, { "epoch": 4.633315593400745, "grad_norm": 0.3232618570327759, "learning_rate": 2.7548491960108446e-09, "loss": 0.2087, "step": 17412 }, { "epoch": 4.633581692389569, "grad_norm": 0.29897838830947876, "learning_rate": 2.7508736275575485e-09, "loss": 0.1841, "step": 17413 }, { "epoch": 4.633847791378392, "grad_norm": 0.35533496737480164, "learning_rate": 2.7469008897738287e-09, "loss": 0.2, "step": 17414 }, { "epoch": 4.6341138903672165, "grad_norm": 0.3160697817802429, "learning_rate": 2.7429309827752935e-09, "loss": 0.2078, "step": 17415 }, { "epoch": 4.634379989356041, "grad_norm": 0.3119843304157257, "learning_rate": 2.7389639066775274e-09, "loss": 0.1914, "step": 17416 }, { "epoch": 4.634646088344864, "grad_norm": 0.28080999851226807, "learning_rate": 2.7349996615959713e-09, "loss": 0.1737, "step": 17417 }, { "epoch": 4.634912187333688, "grad_norm": 0.2924683392047882, "learning_rate": 2.7310382476460338e-09, "loss": 0.1864, "step": 17418 }, { "epoch": 4.635178286322512, "grad_norm": 0.29108428955078125, "learning_rate": 2.7270796649430326e-09, "loss": 0.2045, "step": 17419 }, { "epoch": 4.635444385311336, "grad_norm": 0.2932884991168976, "learning_rate": 2.7231239136021543e-09, "loss": 0.2116, "step": 17420 }, { "epoch": 4.6357104843001595, "grad_norm": 0.2935510575771332, "learning_rate": 2.7191709937385734e-09, "loss": 0.1981, "step": 17421 }, { "epoch": 4.635976583288984, "grad_norm": 0.31452423334121704, "learning_rate": 2.7152209054673304e-09, "loss": 0.1759, "step": 17422 }, { "epoch": 4.636242682277807, "grad_norm": 0.32215645909309387, "learning_rate": 2.7112736489034226e-09, "loss": 0.1957, "step": 17423 }, { "epoch": 4.636508781266631, "grad_norm": 0.28263115882873535, "learning_rate": 2.707329224161725e-09, "loss": 0.1692, "step": 17424 }, { "epoch": 4.636774880255455, "grad_norm": 0.29741546511650085, "learning_rate": 2.7033876313570457e-09, "loss": 0.2127, "step": 17425 }, { "epoch": 4.637040979244279, "grad_norm": 0.29860275983810425, "learning_rate": 2.6994488706041263e-09, "loss": 0.1956, "step": 17426 }, { "epoch": 4.6373070782331025, "grad_norm": 0.269059419631958, "learning_rate": 2.6955129420176194e-09, "loss": 0.1816, "step": 17427 }, { "epoch": 4.637573177221927, "grad_norm": 0.3095425069332123, "learning_rate": 2.6915798457120776e-09, "loss": 0.184, "step": 17428 }, { "epoch": 4.637839276210751, "grad_norm": 0.320204496383667, "learning_rate": 2.6876495818019874e-09, "loss": 0.2012, "step": 17429 }, { "epoch": 4.638105375199574, "grad_norm": 0.2987173795700073, "learning_rate": 2.683722150401746e-09, "loss": 0.1937, "step": 17430 }, { "epoch": 4.638371474188398, "grad_norm": 0.3186711370944977, "learning_rate": 2.6797975516256844e-09, "loss": 0.1913, "step": 17431 }, { "epoch": 4.638637573177222, "grad_norm": 0.3082047402858734, "learning_rate": 2.675875785588033e-09, "loss": 0.1954, "step": 17432 }, { "epoch": 4.6389036721660455, "grad_norm": 0.2878193259239197, "learning_rate": 2.6719568524029235e-09, "loss": 0.1879, "step": 17433 }, { "epoch": 4.63916977115487, "grad_norm": 0.2901051342487335, "learning_rate": 2.6680407521844418e-09, "loss": 0.1843, "step": 17434 }, { "epoch": 4.639435870143694, "grad_norm": 0.2771926522254944, "learning_rate": 2.6641274850465747e-09, "loss": 0.1975, "step": 17435 }, { "epoch": 4.639701969132517, "grad_norm": 0.30208244919776917, "learning_rate": 2.6602170511032195e-09, "loss": 0.1828, "step": 17436 }, { "epoch": 4.639968068121341, "grad_norm": 0.2978552281856537, "learning_rate": 2.6563094504682084e-09, "loss": 0.1949, "step": 17437 }, { "epoch": 4.640234167110165, "grad_norm": 0.3031984865665436, "learning_rate": 2.6524046832552827e-09, "loss": 0.187, "step": 17438 }, { "epoch": 4.6405002660989885, "grad_norm": 0.3044798970222473, "learning_rate": 2.6485027495781074e-09, "loss": 0.2136, "step": 17439 }, { "epoch": 4.640766365087813, "grad_norm": 0.306977242231369, "learning_rate": 2.6446036495502367e-09, "loss": 0.1842, "step": 17440 }, { "epoch": 4.641032464076637, "grad_norm": 0.3040834069252014, "learning_rate": 2.6407073832851676e-09, "loss": 0.213, "step": 17441 }, { "epoch": 4.64129856306546, "grad_norm": 0.2833418548107147, "learning_rate": 2.6368139508963326e-09, "loss": 0.1627, "step": 17442 }, { "epoch": 4.641564662054284, "grad_norm": 0.304221510887146, "learning_rate": 2.6329233524970294e-09, "loss": 0.1901, "step": 17443 }, { "epoch": 4.641830761043108, "grad_norm": 0.2960973381996155, "learning_rate": 2.6290355882005123e-09, "loss": 0.19, "step": 17444 }, { "epoch": 4.642096860031932, "grad_norm": 0.3150951564311981, "learning_rate": 2.625150658119957e-09, "loss": 0.1845, "step": 17445 }, { "epoch": 4.642362959020756, "grad_norm": 0.357639878988266, "learning_rate": 2.6212685623684284e-09, "loss": 0.2068, "step": 17446 }, { "epoch": 4.64262905800958, "grad_norm": 0.3107392191886902, "learning_rate": 2.6173893010589256e-09, "loss": 0.2072, "step": 17447 }, { "epoch": 4.642895156998403, "grad_norm": 0.2968963086605072, "learning_rate": 2.61351287430438e-09, "loss": 0.1994, "step": 17448 }, { "epoch": 4.643161255987227, "grad_norm": 0.3049814701080322, "learning_rate": 2.6096392822176015e-09, "loss": 0.1955, "step": 17449 }, { "epoch": 4.643427354976051, "grad_norm": 0.3313276469707489, "learning_rate": 2.605768524911378e-09, "loss": 0.1963, "step": 17450 }, { "epoch": 4.643693453964875, "grad_norm": 0.2910810112953186, "learning_rate": 2.6019006024983304e-09, "loss": 0.1855, "step": 17451 }, { "epoch": 4.6439595529536986, "grad_norm": 0.3017297387123108, "learning_rate": 2.5980355150910683e-09, "loss": 0.1888, "step": 17452 }, { "epoch": 4.644225651942523, "grad_norm": 0.3051857054233551, "learning_rate": 2.5941732628020797e-09, "loss": 0.1881, "step": 17453 }, { "epoch": 4.644491750931347, "grad_norm": 0.3092077970504761, "learning_rate": 2.590313845743808e-09, "loss": 0.1938, "step": 17454 }, { "epoch": 4.64475784992017, "grad_norm": 0.3049316704273224, "learning_rate": 2.5864572640285743e-09, "loss": 0.1881, "step": 17455 }, { "epoch": 4.645023948908994, "grad_norm": 0.27344223856925964, "learning_rate": 2.5826035177686333e-09, "loss": 0.1808, "step": 17456 }, { "epoch": 4.645290047897818, "grad_norm": 0.2956206202507019, "learning_rate": 2.5787526070761732e-09, "loss": 0.1682, "step": 17457 }, { "epoch": 4.6455561468866415, "grad_norm": 0.2656215727329254, "learning_rate": 2.574904532063282e-09, "loss": 0.1594, "step": 17458 }, { "epoch": 4.645822245875466, "grad_norm": 0.29738670587539673, "learning_rate": 2.5710592928419374e-09, "loss": 0.1961, "step": 17459 }, { "epoch": 4.64608834486429, "grad_norm": 0.31926509737968445, "learning_rate": 2.567216889524093e-09, "loss": 0.2047, "step": 17460 }, { "epoch": 4.646354443853113, "grad_norm": 0.29716914892196655, "learning_rate": 2.5633773222215937e-09, "loss": 0.1925, "step": 17461 }, { "epoch": 4.646620542841937, "grad_norm": 0.2819048762321472, "learning_rate": 2.559540591046172e-09, "loss": 0.1891, "step": 17462 }, { "epoch": 4.646886641830761, "grad_norm": 0.29682084918022156, "learning_rate": 2.5557066961095275e-09, "loss": 0.1902, "step": 17463 }, { "epoch": 4.6471527408195845, "grad_norm": 0.29192692041397095, "learning_rate": 2.5518756375232485e-09, "loss": 0.1842, "step": 17464 }, { "epoch": 4.647418839808409, "grad_norm": 0.30425912141799927, "learning_rate": 2.548047415398846e-09, "loss": 0.1849, "step": 17465 }, { "epoch": 4.647684938797233, "grad_norm": 0.28140461444854736, "learning_rate": 2.5442220298477536e-09, "loss": 0.1847, "step": 17466 }, { "epoch": 4.647951037786056, "grad_norm": 0.2913988530635834, "learning_rate": 2.5403994809813143e-09, "loss": 0.184, "step": 17467 }, { "epoch": 4.64821713677488, "grad_norm": 0.30263715982437134, "learning_rate": 2.536579768910818e-09, "loss": 0.1866, "step": 17468 }, { "epoch": 4.648483235763704, "grad_norm": 0.31382206082344055, "learning_rate": 2.5327628937473975e-09, "loss": 0.1999, "step": 17469 }, { "epoch": 4.648749334752528, "grad_norm": 0.31560418009757996, "learning_rate": 2.5289488556021867e-09, "loss": 0.2019, "step": 17470 }, { "epoch": 4.649015433741352, "grad_norm": 0.3126167058944702, "learning_rate": 2.5251376545861845e-09, "loss": 0.1855, "step": 17471 }, { "epoch": 4.649281532730176, "grad_norm": 0.32279691100120544, "learning_rate": 2.521329290810348e-09, "loss": 0.2026, "step": 17472 }, { "epoch": 4.649547631718999, "grad_norm": 0.30380213260650635, "learning_rate": 2.5175237643854986e-09, "loss": 0.1919, "step": 17473 }, { "epoch": 4.649813730707823, "grad_norm": 0.3142165541648865, "learning_rate": 2.513721075422437e-09, "loss": 0.1985, "step": 17474 }, { "epoch": 4.650079829696647, "grad_norm": 0.29341232776641846, "learning_rate": 2.509921224031819e-09, "loss": 0.1746, "step": 17475 }, { "epoch": 4.650345928685471, "grad_norm": 0.32996270060539246, "learning_rate": 2.5061242103242788e-09, "loss": 0.2006, "step": 17476 }, { "epoch": 4.650612027674295, "grad_norm": 0.2858651578426361, "learning_rate": 2.502330034410327e-09, "loss": 0.1778, "step": 17477 }, { "epoch": 4.650878126663119, "grad_norm": 0.3970326781272888, "learning_rate": 2.4985386964003874e-09, "loss": 0.1848, "step": 17478 }, { "epoch": 4.651144225651943, "grad_norm": 0.3087795674800873, "learning_rate": 2.494750196404827e-09, "loss": 0.2095, "step": 17479 }, { "epoch": 4.651410324640766, "grad_norm": 0.2960794270038605, "learning_rate": 2.4909645345339126e-09, "loss": 0.1791, "step": 17480 }, { "epoch": 4.65167642362959, "grad_norm": 0.3119101822376251, "learning_rate": 2.4871817108978566e-09, "loss": 0.1928, "step": 17481 }, { "epoch": 4.651942522618414, "grad_norm": 0.2878955602645874, "learning_rate": 2.4834017256067375e-09, "loss": 0.184, "step": 17482 }, { "epoch": 4.652208621607238, "grad_norm": 0.31074586510658264, "learning_rate": 2.4796245787706006e-09, "loss": 0.2097, "step": 17483 }, { "epoch": 4.652474720596062, "grad_norm": 0.3136346936225891, "learning_rate": 2.47585027049938e-09, "loss": 0.2064, "step": 17484 }, { "epoch": 4.652740819584886, "grad_norm": 0.2846944034099579, "learning_rate": 2.4720788009029438e-09, "loss": 0.1743, "step": 17485 }, { "epoch": 4.653006918573709, "grad_norm": 0.3185928761959076, "learning_rate": 2.4683101700910704e-09, "loss": 0.2134, "step": 17486 }, { "epoch": 4.653273017562533, "grad_norm": 0.32236114144325256, "learning_rate": 2.4645443781734387e-09, "loss": 0.2168, "step": 17487 }, { "epoch": 4.653539116551357, "grad_norm": 0.297017365694046, "learning_rate": 2.4607814252596726e-09, "loss": 0.1875, "step": 17488 }, { "epoch": 4.653805215540181, "grad_norm": 0.30844640731811523, "learning_rate": 2.457021311459295e-09, "loss": 0.1974, "step": 17489 }, { "epoch": 4.654071314529005, "grad_norm": 0.29922544956207275, "learning_rate": 2.453264036881775e-09, "loss": 0.1877, "step": 17490 }, { "epoch": 4.654337413517829, "grad_norm": 0.2915477752685547, "learning_rate": 2.4495096016364458e-09, "loss": 0.1878, "step": 17491 }, { "epoch": 4.654603512506652, "grad_norm": 0.29965656995773315, "learning_rate": 2.44575800583261e-09, "loss": 0.1644, "step": 17492 }, { "epoch": 4.654869611495476, "grad_norm": 0.294796347618103, "learning_rate": 2.4420092495794685e-09, "loss": 0.19, "step": 17493 }, { "epoch": 4.6551357104843, "grad_norm": 0.277290016412735, "learning_rate": 2.4382633329861236e-09, "loss": 0.1816, "step": 17494 }, { "epoch": 4.655401809473124, "grad_norm": 0.2936781048774719, "learning_rate": 2.4345202561616316e-09, "loss": 0.1828, "step": 17495 }, { "epoch": 4.655667908461948, "grad_norm": 0.28857919573783875, "learning_rate": 2.4307800192149066e-09, "loss": 0.1803, "step": 17496 }, { "epoch": 4.655934007450772, "grad_norm": 0.3142538368701935, "learning_rate": 2.427042622254849e-09, "loss": 0.2058, "step": 17497 }, { "epoch": 4.656200106439595, "grad_norm": 0.3372769355773926, "learning_rate": 2.4233080653902394e-09, "loss": 0.1873, "step": 17498 }, { "epoch": 4.656466205428419, "grad_norm": 0.3301999270915985, "learning_rate": 2.4195763487297683e-09, "loss": 0.2218, "step": 17499 }, { "epoch": 4.656732304417243, "grad_norm": 0.2932753562927246, "learning_rate": 2.4158474723820822e-09, "loss": 0.1789, "step": 17500 }, { "epoch": 4.656998403406067, "grad_norm": 0.3203001320362091, "learning_rate": 2.412121436455683e-09, "loss": 0.189, "step": 17501 }, { "epoch": 4.657264502394891, "grad_norm": 0.3019447922706604, "learning_rate": 2.408398241059051e-09, "loss": 0.2046, "step": 17502 }, { "epoch": 4.657530601383715, "grad_norm": 0.29149675369262695, "learning_rate": 2.4046778863005434e-09, "loss": 0.2049, "step": 17503 }, { "epoch": 4.657796700372539, "grad_norm": 0.2947021722793579, "learning_rate": 2.4009603722884743e-09, "loss": 0.1952, "step": 17504 }, { "epoch": 4.658062799361362, "grad_norm": 0.29537394642829895, "learning_rate": 2.397245699131034e-09, "loss": 0.1961, "step": 17505 }, { "epoch": 4.658328898350186, "grad_norm": 0.31857749819755554, "learning_rate": 2.393533866936337e-09, "loss": 0.2019, "step": 17506 }, { "epoch": 4.65859499733901, "grad_norm": 0.3013744056224823, "learning_rate": 2.3898248758124405e-09, "loss": 0.1925, "step": 17507 }, { "epoch": 4.658861096327834, "grad_norm": 0.3039514124393463, "learning_rate": 2.3861187258672923e-09, "loss": 0.1959, "step": 17508 }, { "epoch": 4.659127195316658, "grad_norm": 0.3035339117050171, "learning_rate": 2.382415417208783e-09, "loss": 0.1912, "step": 17509 }, { "epoch": 4.659393294305482, "grad_norm": 0.2887188792228699, "learning_rate": 2.3787149499447045e-09, "loss": 0.186, "step": 17510 }, { "epoch": 4.659659393294305, "grad_norm": 0.3199455142021179, "learning_rate": 2.37501732418276e-09, "loss": 0.2013, "step": 17511 }, { "epoch": 4.659925492283129, "grad_norm": 0.2956911623477936, "learning_rate": 2.3713225400305736e-09, "loss": 0.1862, "step": 17512 }, { "epoch": 4.660191591271953, "grad_norm": 0.2883439362049103, "learning_rate": 2.3676305975957154e-09, "loss": 0.1938, "step": 17513 }, { "epoch": 4.6604576902607775, "grad_norm": 0.3052901029586792, "learning_rate": 2.363941496985611e-09, "loss": 0.1806, "step": 17514 }, { "epoch": 4.660723789249601, "grad_norm": 0.32156914472579956, "learning_rate": 2.3602552383076734e-09, "loss": 0.189, "step": 17515 }, { "epoch": 4.660989888238425, "grad_norm": 0.29557937383651733, "learning_rate": 2.3565718216691732e-09, "loss": 0.1782, "step": 17516 }, { "epoch": 4.661255987227248, "grad_norm": 0.32140618562698364, "learning_rate": 2.3528912471773464e-09, "loss": 0.1765, "step": 17517 }, { "epoch": 4.661522086216072, "grad_norm": 0.3109695315361023, "learning_rate": 2.349213514939319e-09, "loss": 0.2021, "step": 17518 }, { "epoch": 4.661788185204896, "grad_norm": 0.30148881673812866, "learning_rate": 2.345538625062138e-09, "loss": 0.1912, "step": 17519 }, { "epoch": 4.6620542841937205, "grad_norm": 0.2766697406768799, "learning_rate": 2.3418665776527736e-09, "loss": 0.1894, "step": 17520 }, { "epoch": 4.662320383182544, "grad_norm": 0.2953968048095703, "learning_rate": 2.338197372818107e-09, "loss": 0.1745, "step": 17521 }, { "epoch": 4.662586482171368, "grad_norm": 0.31620854139328003, "learning_rate": 2.3345310106649307e-09, "loss": 0.2024, "step": 17522 }, { "epoch": 4.662852581160192, "grad_norm": 0.29480651021003723, "learning_rate": 2.3308674912999814e-09, "loss": 0.1883, "step": 17523 }, { "epoch": 4.663118680149015, "grad_norm": 0.30370327830314636, "learning_rate": 2.327206814829874e-09, "loss": 0.1816, "step": 17524 }, { "epoch": 4.663384779137839, "grad_norm": 0.2935096323490143, "learning_rate": 2.3235489813611674e-09, "loss": 0.1898, "step": 17525 }, { "epoch": 4.6636508781266635, "grad_norm": 0.2971576154232025, "learning_rate": 2.3198939910003436e-09, "loss": 0.1744, "step": 17526 }, { "epoch": 4.663916977115487, "grad_norm": 0.2702561318874359, "learning_rate": 2.3162418438537834e-09, "loss": 0.1788, "step": 17527 }, { "epoch": 4.664183076104311, "grad_norm": 0.3354839086532593, "learning_rate": 2.31259254002778e-09, "loss": 0.2181, "step": 17528 }, { "epoch": 4.664449175093135, "grad_norm": 0.31797298789024353, "learning_rate": 2.308946079628571e-09, "loss": 0.2124, "step": 17529 }, { "epoch": 4.664715274081958, "grad_norm": 0.32943207025527954, "learning_rate": 2.305302462762293e-09, "loss": 0.2151, "step": 17530 }, { "epoch": 4.664981373070782, "grad_norm": 0.2896844744682312, "learning_rate": 2.301661689534995e-09, "loss": 0.1896, "step": 17531 }, { "epoch": 4.665247472059606, "grad_norm": 0.3130292594432831, "learning_rate": 2.298023760052659e-09, "loss": 0.2024, "step": 17532 }, { "epoch": 4.66551357104843, "grad_norm": 0.3234444856643677, "learning_rate": 2.294388674421166e-09, "loss": 0.2091, "step": 17533 }, { "epoch": 4.665779670037254, "grad_norm": 0.2984817326068878, "learning_rate": 2.290756432746321e-09, "loss": 0.1858, "step": 17534 }, { "epoch": 4.666045769026078, "grad_norm": 0.33048415184020996, "learning_rate": 2.287127035133862e-09, "loss": 0.2096, "step": 17535 }, { "epoch": 4.666311868014901, "grad_norm": 0.29427310824394226, "learning_rate": 2.2835004816894377e-09, "loss": 0.188, "step": 17536 }, { "epoch": 4.666577967003725, "grad_norm": 0.29778483510017395, "learning_rate": 2.279876772518585e-09, "loss": 0.1841, "step": 17537 }, { "epoch": 4.666844065992549, "grad_norm": 0.319185346364975, "learning_rate": 2.2762559077267873e-09, "loss": 0.2001, "step": 17538 }, { "epoch": 4.6671101649813735, "grad_norm": 0.30735987424850464, "learning_rate": 2.2726378874194484e-09, "loss": 0.1914, "step": 17539 }, { "epoch": 4.667376263970197, "grad_norm": 0.3148311376571655, "learning_rate": 2.2690227117018844e-09, "loss": 0.2091, "step": 17540 }, { "epoch": 4.667642362959021, "grad_norm": 0.2979126572608948, "learning_rate": 2.2654103806793e-09, "loss": 0.1657, "step": 17541 }, { "epoch": 4.667908461947844, "grad_norm": 0.27790015935897827, "learning_rate": 2.2618008944568556e-09, "loss": 0.1863, "step": 17542 }, { "epoch": 4.668174560936668, "grad_norm": 0.3095160722732544, "learning_rate": 2.2581942531396225e-09, "loss": 0.1957, "step": 17543 }, { "epoch": 4.668440659925492, "grad_norm": 0.3042064309120178, "learning_rate": 2.2545904568325614e-09, "loss": 0.1957, "step": 17544 }, { "epoch": 4.6687067589143165, "grad_norm": 0.3176018297672272, "learning_rate": 2.250989505640588e-09, "loss": 0.1914, "step": 17545 }, { "epoch": 4.66897285790314, "grad_norm": 0.31190019845962524, "learning_rate": 2.2473913996684967e-09, "loss": 0.1941, "step": 17546 }, { "epoch": 4.669238956891964, "grad_norm": 0.2889198362827301, "learning_rate": 2.2437961390210258e-09, "loss": 0.189, "step": 17547 }, { "epoch": 4.669505055880788, "grad_norm": 0.3152230679988861, "learning_rate": 2.240203723802836e-09, "loss": 0.1959, "step": 17548 }, { "epoch": 4.669771154869611, "grad_norm": 0.278279185295105, "learning_rate": 2.236614154118488e-09, "loss": 0.1721, "step": 17549 }, { "epoch": 4.670037253858435, "grad_norm": 0.30218198895454407, "learning_rate": 2.2330274300724648e-09, "loss": 0.1754, "step": 17550 }, { "epoch": 4.6703033528472595, "grad_norm": 0.2951742708683014, "learning_rate": 2.22944355176915e-09, "loss": 0.1993, "step": 17551 }, { "epoch": 4.670569451836083, "grad_norm": 0.2935323119163513, "learning_rate": 2.2258625193128822e-09, "loss": 0.2052, "step": 17552 }, { "epoch": 4.670835550824907, "grad_norm": 0.31204503774642944, "learning_rate": 2.222284332807889e-09, "loss": 0.1967, "step": 17553 }, { "epoch": 4.671101649813731, "grad_norm": 0.2866077721118927, "learning_rate": 2.2187089923583204e-09, "loss": 0.1889, "step": 17554 }, { "epoch": 4.671367748802554, "grad_norm": 0.2814111113548279, "learning_rate": 2.2151364980682375e-09, "loss": 0.1733, "step": 17555 }, { "epoch": 4.671633847791378, "grad_norm": 0.33488935232162476, "learning_rate": 2.2115668500416463e-09, "loss": 0.216, "step": 17556 }, { "epoch": 4.6718999467802025, "grad_norm": 0.3197843134403229, "learning_rate": 2.20800004838243e-09, "loss": 0.1942, "step": 17557 }, { "epoch": 4.672166045769026, "grad_norm": 0.30554333329200745, "learning_rate": 2.2044360931944396e-09, "loss": 0.1878, "step": 17558 }, { "epoch": 4.67243214475785, "grad_norm": 0.2816126048564911, "learning_rate": 2.200874984581369e-09, "loss": 0.1902, "step": 17559 }, { "epoch": 4.672698243746674, "grad_norm": 0.2984377145767212, "learning_rate": 2.1973167226469026e-09, "loss": 0.196, "step": 17560 }, { "epoch": 4.672964342735497, "grad_norm": 0.3117702305316925, "learning_rate": 2.1937613074946014e-09, "loss": 0.2046, "step": 17561 }, { "epoch": 4.673230441724321, "grad_norm": 0.3020521104335785, "learning_rate": 2.1902087392279723e-09, "loss": 0.1948, "step": 17562 }, { "epoch": 4.6734965407131455, "grad_norm": 0.34214791655540466, "learning_rate": 2.1866590179503876e-09, "loss": 0.2127, "step": 17563 }, { "epoch": 4.67376263970197, "grad_norm": 0.2880845069885254, "learning_rate": 2.183112143765198e-09, "loss": 0.1865, "step": 17564 }, { "epoch": 4.674028738690793, "grad_norm": 0.28490349650382996, "learning_rate": 2.1795681167756207e-09, "loss": 0.1921, "step": 17565 }, { "epoch": 4.674294837679617, "grad_norm": 0.2933596074581146, "learning_rate": 2.1760269370848406e-09, "loss": 0.1684, "step": 17566 }, { "epoch": 4.67456093666844, "grad_norm": 0.306832492351532, "learning_rate": 2.17248860479593e-09, "loss": 0.1901, "step": 17567 }, { "epoch": 4.674827035657264, "grad_norm": 0.32459965348243713, "learning_rate": 2.1689531200118517e-09, "loss": 0.1923, "step": 17568 }, { "epoch": 4.6750931346460884, "grad_norm": 0.29355746507644653, "learning_rate": 2.1654204828355336e-09, "loss": 0.1823, "step": 17569 }, { "epoch": 4.675359233634913, "grad_norm": 0.32093313336372375, "learning_rate": 2.1618906933698054e-09, "loss": 0.2245, "step": 17570 }, { "epoch": 4.675625332623736, "grad_norm": 0.2945428788661957, "learning_rate": 2.1583637517173956e-09, "loss": 0.1817, "step": 17571 }, { "epoch": 4.67589143161256, "grad_norm": 0.28811439871788025, "learning_rate": 2.154839657980989e-09, "loss": 0.1848, "step": 17572 }, { "epoch": 4.676157530601384, "grad_norm": 0.29524460434913635, "learning_rate": 2.1513184122631365e-09, "loss": 0.1791, "step": 17573 }, { "epoch": 4.676423629590207, "grad_norm": 0.3056647777557373, "learning_rate": 2.1478000146663457e-09, "loss": 0.1934, "step": 17574 }, { "epoch": 4.676689728579031, "grad_norm": 0.3025456666946411, "learning_rate": 2.1442844652930224e-09, "loss": 0.186, "step": 17575 }, { "epoch": 4.6769558275678556, "grad_norm": 0.2946697473526001, "learning_rate": 2.140771764245508e-09, "loss": 0.176, "step": 17576 }, { "epoch": 4.677221926556679, "grad_norm": 0.2912387251853943, "learning_rate": 2.1372619116260317e-09, "loss": 0.1987, "step": 17577 }, { "epoch": 4.677488025545503, "grad_norm": 0.29972484707832336, "learning_rate": 2.1337549075367667e-09, "loss": 0.1973, "step": 17578 }, { "epoch": 4.677754124534327, "grad_norm": 0.2852610647678375, "learning_rate": 2.1302507520797872e-09, "loss": 0.1776, "step": 17579 }, { "epoch": 4.67802022352315, "grad_norm": 0.2981356680393219, "learning_rate": 2.12674944535709e-09, "loss": 0.1846, "step": 17580 }, { "epoch": 4.678286322511974, "grad_norm": 0.2922133505344391, "learning_rate": 2.1232509874705927e-09, "loss": 0.187, "step": 17581 }, { "epoch": 4.6785524215007985, "grad_norm": 0.2977994382381439, "learning_rate": 2.1197553785221366e-09, "loss": 0.226, "step": 17582 }, { "epoch": 4.678818520489622, "grad_norm": 0.31003808975219727, "learning_rate": 2.11626261861344e-09, "loss": 0.1812, "step": 17583 }, { "epoch": 4.679084619478446, "grad_norm": 0.2706122398376465, "learning_rate": 2.1127727078461886e-09, "loss": 0.1814, "step": 17584 }, { "epoch": 4.67935071846727, "grad_norm": 0.32166528701782227, "learning_rate": 2.1092856463219788e-09, "loss": 0.1848, "step": 17585 }, { "epoch": 4.679616817456093, "grad_norm": 0.2783065140247345, "learning_rate": 2.1058014341422735e-09, "loss": 0.1802, "step": 17586 }, { "epoch": 4.679882916444917, "grad_norm": 0.32858577370643616, "learning_rate": 2.1023200714085144e-09, "loss": 0.18, "step": 17587 }, { "epoch": 4.6801490154337415, "grad_norm": 0.3188079595565796, "learning_rate": 2.0988415582220308e-09, "loss": 0.1997, "step": 17588 }, { "epoch": 4.680415114422566, "grad_norm": 0.28698235750198364, "learning_rate": 2.0953658946840757e-09, "loss": 0.1969, "step": 17589 }, { "epoch": 4.680681213411389, "grad_norm": 0.31148606538772583, "learning_rate": 2.0918930808958014e-09, "loss": 0.193, "step": 17590 }, { "epoch": 4.680947312400213, "grad_norm": 0.321458637714386, "learning_rate": 2.0884231169583154e-09, "loss": 0.2206, "step": 17591 }, { "epoch": 4.681213411389036, "grad_norm": 0.2791142761707306, "learning_rate": 2.0849560029726044e-09, "loss": 0.1754, "step": 17592 }, { "epoch": 4.68147951037786, "grad_norm": 0.28047874569892883, "learning_rate": 2.081491739039587e-09, "loss": 0.1825, "step": 17593 }, { "epoch": 4.6817456093666845, "grad_norm": 0.3357508182525635, "learning_rate": 2.078030325260105e-09, "loss": 0.1877, "step": 17594 }, { "epoch": 4.682011708355509, "grad_norm": 0.3096841871738434, "learning_rate": 2.074571761734911e-09, "loss": 0.186, "step": 17595 }, { "epoch": 4.682277807344332, "grad_norm": 0.30090266466140747, "learning_rate": 2.0711160485646693e-09, "loss": 0.1903, "step": 17596 }, { "epoch": 4.682543906333156, "grad_norm": 0.30494821071624756, "learning_rate": 2.067663185849966e-09, "loss": 0.1802, "step": 17597 }, { "epoch": 4.68281000532198, "grad_norm": 0.30746564269065857, "learning_rate": 2.0642131736913093e-09, "loss": 0.2042, "step": 17598 }, { "epoch": 4.683076104310803, "grad_norm": 0.2986954152584076, "learning_rate": 2.060766012189119e-09, "loss": 0.1885, "step": 17599 }, { "epoch": 4.6833422032996275, "grad_norm": 0.31677836179733276, "learning_rate": 2.0573217014437373e-09, "loss": 0.2147, "step": 17600 }, { "epoch": 4.683608302288452, "grad_norm": 0.3128487765789032, "learning_rate": 2.053880241555417e-09, "loss": 0.2164, "step": 17601 }, { "epoch": 4.683874401277275, "grad_norm": 0.3012669086456299, "learning_rate": 2.0504416326243225e-09, "loss": 0.1776, "step": 17602 }, { "epoch": 4.684140500266099, "grad_norm": 0.2975651025772095, "learning_rate": 2.0470058747505516e-09, "loss": 0.1777, "step": 17603 }, { "epoch": 4.684406599254923, "grad_norm": 0.2852623462677002, "learning_rate": 2.043572968034113e-09, "loss": 0.1946, "step": 17604 }, { "epoch": 4.684672698243746, "grad_norm": 0.30709248781204224, "learning_rate": 2.040142912574916e-09, "loss": 0.1907, "step": 17605 }, { "epoch": 4.6849387972325705, "grad_norm": 0.34427690505981445, "learning_rate": 2.0367157084728138e-09, "loss": 0.1959, "step": 17606 }, { "epoch": 4.685204896221395, "grad_norm": 0.3192392587661743, "learning_rate": 2.0332913558275488e-09, "loss": 0.1918, "step": 17607 }, { "epoch": 4.685470995210218, "grad_norm": 0.3082454800605774, "learning_rate": 2.0298698547388082e-09, "loss": 0.192, "step": 17608 }, { "epoch": 4.685737094199042, "grad_norm": 0.29753684997558594, "learning_rate": 2.02645120530619e-09, "loss": 0.1779, "step": 17609 }, { "epoch": 4.686003193187866, "grad_norm": 0.34442564845085144, "learning_rate": 2.023035407629181e-09, "loss": 0.2068, "step": 17610 }, { "epoch": 4.686269292176689, "grad_norm": 0.30712538957595825, "learning_rate": 2.0196224618072133e-09, "loss": 0.1859, "step": 17611 }, { "epoch": 4.686535391165513, "grad_norm": 0.32180294394493103, "learning_rate": 2.0162123679396515e-09, "loss": 0.202, "step": 17612 }, { "epoch": 4.686801490154338, "grad_norm": 0.29145383834838867, "learning_rate": 2.012805126125716e-09, "loss": 0.196, "step": 17613 }, { "epoch": 4.687067589143162, "grad_norm": 0.2972506582736969, "learning_rate": 2.0094007364646058e-09, "loss": 0.2035, "step": 17614 }, { "epoch": 4.687333688131985, "grad_norm": 0.2996663451194763, "learning_rate": 2.005999199055408e-09, "loss": 0.1952, "step": 17615 }, { "epoch": 4.687599787120809, "grad_norm": 0.28844210505485535, "learning_rate": 2.0026005139971436e-09, "loss": 0.1854, "step": 17616 }, { "epoch": 4.687865886109632, "grad_norm": 0.30222028493881226, "learning_rate": 1.999204681388722e-09, "loss": 0.1925, "step": 17617 }, { "epoch": 4.688131985098456, "grad_norm": 0.3028039038181305, "learning_rate": 1.9958117013289977e-09, "loss": 0.1964, "step": 17618 }, { "epoch": 4.6883980840872805, "grad_norm": 0.32086634635925293, "learning_rate": 1.992421573916725e-09, "loss": 0.1704, "step": 17619 }, { "epoch": 4.688664183076105, "grad_norm": 0.3274880349636078, "learning_rate": 1.9890342992505916e-09, "loss": 0.2065, "step": 17620 }, { "epoch": 4.688930282064928, "grad_norm": 0.34689030051231384, "learning_rate": 1.9856498774291962e-09, "loss": 0.1908, "step": 17621 }, { "epoch": 4.689196381053752, "grad_norm": 0.3157990872859955, "learning_rate": 1.982268308551038e-09, "loss": 0.209, "step": 17622 }, { "epoch": 4.689462480042576, "grad_norm": 0.29004693031311035, "learning_rate": 1.9788895927145388e-09, "loss": 0.1763, "step": 17623 }, { "epoch": 4.689728579031399, "grad_norm": 0.28366202116012573, "learning_rate": 1.9755137300180634e-09, "loss": 0.1807, "step": 17624 }, { "epoch": 4.6899946780202235, "grad_norm": 0.2949695587158203, "learning_rate": 1.972140720559856e-09, "loss": 0.1871, "step": 17625 }, { "epoch": 4.690260777009048, "grad_norm": 0.2955667972564697, "learning_rate": 1.968770564438116e-09, "loss": 0.201, "step": 17626 }, { "epoch": 4.690526875997871, "grad_norm": 0.2836347222328186, "learning_rate": 1.9654032617509196e-09, "loss": 0.1998, "step": 17627 }, { "epoch": 4.690792974986695, "grad_norm": 0.28462928533554077, "learning_rate": 1.9620388125962894e-09, "loss": 0.1799, "step": 17628 }, { "epoch": 4.691059073975519, "grad_norm": 0.3647019565105438, "learning_rate": 1.9586772170721687e-09, "loss": 0.2011, "step": 17629 }, { "epoch": 4.691325172964342, "grad_norm": 0.3461199104785919, "learning_rate": 1.9553184752763906e-09, "loss": 0.2033, "step": 17630 }, { "epoch": 4.6915912719531665, "grad_norm": 0.33180513978004456, "learning_rate": 1.9519625873067104e-09, "loss": 0.1976, "step": 17631 }, { "epoch": 4.691857370941991, "grad_norm": 0.3067106604576111, "learning_rate": 1.9486095532608273e-09, "loss": 0.1957, "step": 17632 }, { "epoch": 4.692123469930815, "grad_norm": 0.29776453971862793, "learning_rate": 1.94525937323633e-09, "loss": 0.2005, "step": 17633 }, { "epoch": 4.692389568919638, "grad_norm": 0.29467615485191345, "learning_rate": 1.9419120473307293e-09, "loss": 0.1862, "step": 17634 }, { "epoch": 4.692655667908462, "grad_norm": 0.31796231865882874, "learning_rate": 1.9385675756414698e-09, "loss": 0.2068, "step": 17635 }, { "epoch": 4.692921766897285, "grad_norm": 0.30571961402893066, "learning_rate": 1.935225958265896e-09, "loss": 0.2104, "step": 17636 }, { "epoch": 4.6931878658861095, "grad_norm": 0.29894113540649414, "learning_rate": 1.931887195301263e-09, "loss": 0.1984, "step": 17637 }, { "epoch": 4.693453964874934, "grad_norm": 0.2762083113193512, "learning_rate": 1.9285512868447705e-09, "loss": 0.1774, "step": 17638 }, { "epoch": 4.693720063863758, "grad_norm": 0.2759381830692291, "learning_rate": 1.9252182329935198e-09, "loss": 0.1954, "step": 17639 }, { "epoch": 4.693986162852581, "grad_norm": 0.30733153223991394, "learning_rate": 1.921888033844499e-09, "loss": 0.1906, "step": 17640 }, { "epoch": 4.694252261841405, "grad_norm": 0.2944008708000183, "learning_rate": 1.918560689494675e-09, "loss": 0.1864, "step": 17641 }, { "epoch": 4.694518360830229, "grad_norm": 0.28197044134140015, "learning_rate": 1.9152362000408707e-09, "loss": 0.1837, "step": 17642 }, { "epoch": 4.6947844598190525, "grad_norm": 0.29442790150642395, "learning_rate": 1.911914565579886e-09, "loss": 0.1849, "step": 17643 }, { "epoch": 4.695050558807877, "grad_norm": 0.2824145555496216, "learning_rate": 1.908595786208367e-09, "loss": 0.1729, "step": 17644 }, { "epoch": 4.695316657796701, "grad_norm": 0.3002772033214569, "learning_rate": 1.9052798620229348e-09, "loss": 0.1852, "step": 17645 }, { "epoch": 4.695582756785524, "grad_norm": 0.2887158989906311, "learning_rate": 1.9019667931201023e-09, "loss": 0.1876, "step": 17646 }, { "epoch": 4.695848855774348, "grad_norm": 0.27402108907699585, "learning_rate": 1.8986565795963137e-09, "loss": 0.1671, "step": 17647 }, { "epoch": 4.696114954763172, "grad_norm": 0.29408594965934753, "learning_rate": 1.895349221547915e-09, "loss": 0.1974, "step": 17648 }, { "epoch": 4.696381053751995, "grad_norm": 0.3075665533542633, "learning_rate": 1.892044719071184e-09, "loss": 0.2123, "step": 17649 }, { "epoch": 4.69664715274082, "grad_norm": 0.2796425521373749, "learning_rate": 1.888743072262278e-09, "loss": 0.1867, "step": 17650 }, { "epoch": 4.696913251729644, "grad_norm": 0.29311394691467285, "learning_rate": 1.88544428121733e-09, "loss": 0.194, "step": 17651 }, { "epoch": 4.697179350718467, "grad_norm": 0.30267196893692017, "learning_rate": 1.882148346032342e-09, "loss": 0.2014, "step": 17652 }, { "epoch": 4.697445449707291, "grad_norm": 0.307852566242218, "learning_rate": 1.878855266803259e-09, "loss": 0.2027, "step": 17653 }, { "epoch": 4.697711548696115, "grad_norm": 0.3196294605731964, "learning_rate": 1.875565043625915e-09, "loss": 0.191, "step": 17654 }, { "epoch": 4.697977647684938, "grad_norm": 0.2863142192363739, "learning_rate": 1.872277676596112e-09, "loss": 0.1996, "step": 17655 }, { "epoch": 4.6982437466737625, "grad_norm": 0.3109986484050751, "learning_rate": 1.8689931658095184e-09, "loss": 0.1842, "step": 17656 }, { "epoch": 4.698509845662587, "grad_norm": 0.30080774426460266, "learning_rate": 1.8657115113617337e-09, "loss": 0.1969, "step": 17657 }, { "epoch": 4.698775944651411, "grad_norm": 0.2948683500289917, "learning_rate": 1.8624327133482832e-09, "loss": 0.1916, "step": 17658 }, { "epoch": 4.699042043640234, "grad_norm": 0.30444642901420593, "learning_rate": 1.8591567718646118e-09, "loss": 0.1863, "step": 17659 }, { "epoch": 4.699308142629058, "grad_norm": 0.3048110008239746, "learning_rate": 1.8558836870060656e-09, "loss": 0.1827, "step": 17660 }, { "epoch": 4.699574241617881, "grad_norm": 0.3032423257827759, "learning_rate": 1.8526134588679022e-09, "loss": 0.1942, "step": 17661 }, { "epoch": 4.6998403406067055, "grad_norm": 0.3072943091392517, "learning_rate": 1.8493460875453337e-09, "loss": 0.2077, "step": 17662 }, { "epoch": 4.70010643959553, "grad_norm": 0.2874540686607361, "learning_rate": 1.8460815731334733e-09, "loss": 0.195, "step": 17663 }, { "epoch": 4.700372538584354, "grad_norm": 0.3237508535385132, "learning_rate": 1.8428199157273006e-09, "loss": 0.2114, "step": 17664 }, { "epoch": 4.700638637573177, "grad_norm": 0.29988375306129456, "learning_rate": 1.8395611154217838e-09, "loss": 0.1769, "step": 17665 }, { "epoch": 4.700904736562001, "grad_norm": 0.29322853684425354, "learning_rate": 1.8363051723117695e-09, "loss": 0.1805, "step": 17666 }, { "epoch": 4.701170835550825, "grad_norm": 0.2939044237136841, "learning_rate": 1.8330520864920484e-09, "loss": 0.1714, "step": 17667 }, { "epoch": 4.7014369345396485, "grad_norm": 0.27868664264678955, "learning_rate": 1.8298018580572783e-09, "loss": 0.1906, "step": 17668 }, { "epoch": 4.701703033528473, "grad_norm": 0.33027052879333496, "learning_rate": 1.8265544871020722e-09, "loss": 0.1939, "step": 17669 }, { "epoch": 4.701969132517297, "grad_norm": 0.3128136694431305, "learning_rate": 1.8233099737209655e-09, "loss": 0.19, "step": 17670 }, { "epoch": 4.70223523150612, "grad_norm": 0.3090677261352539, "learning_rate": 1.8200683180084054e-09, "loss": 0.178, "step": 17671 }, { "epoch": 4.702501330494944, "grad_norm": 0.31411680579185486, "learning_rate": 1.8168295200587157e-09, "loss": 0.1846, "step": 17672 }, { "epoch": 4.702767429483768, "grad_norm": 0.31455036997795105, "learning_rate": 1.8135935799661995e-09, "loss": 0.1984, "step": 17673 }, { "epoch": 4.7030335284725915, "grad_norm": 0.2936711609363556, "learning_rate": 1.8103604978250364e-09, "loss": 0.1702, "step": 17674 }, { "epoch": 4.703299627461416, "grad_norm": 0.30257731676101685, "learning_rate": 1.8071302737293292e-09, "loss": 0.183, "step": 17675 }, { "epoch": 4.70356572645024, "grad_norm": 0.34178659319877625, "learning_rate": 1.8039029077731139e-09, "loss": 0.2036, "step": 17676 }, { "epoch": 4.703831825439063, "grad_norm": 0.320377379655838, "learning_rate": 1.8006784000503039e-09, "loss": 0.1906, "step": 17677 }, { "epoch": 4.704097924427887, "grad_norm": 0.306199848651886, "learning_rate": 1.7974567506547911e-09, "loss": 0.1936, "step": 17678 }, { "epoch": 4.704364023416711, "grad_norm": 0.3025422692298889, "learning_rate": 1.7942379596803226e-09, "loss": 0.1852, "step": 17679 }, { "epoch": 4.7046301224055345, "grad_norm": 0.29657235741615295, "learning_rate": 1.79102202722059e-09, "loss": 0.1923, "step": 17680 }, { "epoch": 4.704896221394359, "grad_norm": 0.2943035066127777, "learning_rate": 1.7878089533692186e-09, "loss": 0.1842, "step": 17681 }, { "epoch": 4.705162320383183, "grad_norm": 0.37296801805496216, "learning_rate": 1.7845987382197224e-09, "loss": 0.1866, "step": 17682 }, { "epoch": 4.705428419372007, "grad_norm": 0.3083569407463074, "learning_rate": 1.7813913818655602e-09, "loss": 0.1957, "step": 17683 }, { "epoch": 4.70569451836083, "grad_norm": 0.31807032227516174, "learning_rate": 1.7781868844000458e-09, "loss": 0.1895, "step": 17684 }, { "epoch": 4.705960617349654, "grad_norm": 0.28825682401657104, "learning_rate": 1.7749852459164939e-09, "loss": 0.1834, "step": 17685 }, { "epoch": 4.7062267163384774, "grad_norm": 0.3478679656982422, "learning_rate": 1.771786466508085e-09, "loss": 0.1981, "step": 17686 }, { "epoch": 4.706492815327302, "grad_norm": 0.2998620867729187, "learning_rate": 1.768590546267923e-09, "loss": 0.1994, "step": 17687 }, { "epoch": 4.706758914316126, "grad_norm": 0.29364532232284546, "learning_rate": 1.765397485289033e-09, "loss": 0.1825, "step": 17688 }, { "epoch": 4.70702501330495, "grad_norm": 0.29075998067855835, "learning_rate": 1.7622072836643631e-09, "loss": 0.182, "step": 17689 }, { "epoch": 4.707291112293773, "grad_norm": 0.290530264377594, "learning_rate": 1.7590199414867613e-09, "loss": 0.187, "step": 17690 }, { "epoch": 4.707557211282597, "grad_norm": 0.30007117986679077, "learning_rate": 1.7558354588490088e-09, "loss": 0.1817, "step": 17691 }, { "epoch": 4.707823310271421, "grad_norm": 0.2898051142692566, "learning_rate": 1.752653835843798e-09, "loss": 0.1873, "step": 17692 }, { "epoch": 4.7080894092602446, "grad_norm": 0.294839084148407, "learning_rate": 1.749475072563733e-09, "loss": 0.1708, "step": 17693 }, { "epoch": 4.708355508249069, "grad_norm": 0.2822641134262085, "learning_rate": 1.7462991691013507e-09, "loss": 0.1684, "step": 17694 }, { "epoch": 4.708621607237893, "grad_norm": 0.2800692319869995, "learning_rate": 1.743126125549088e-09, "loss": 0.193, "step": 17695 }, { "epoch": 4.708887706226716, "grad_norm": 0.30355018377304077, "learning_rate": 1.7399559419992936e-09, "loss": 0.2034, "step": 17696 }, { "epoch": 4.70915380521554, "grad_norm": 0.34946396946907043, "learning_rate": 1.7367886185442604e-09, "loss": 0.2066, "step": 17697 }, { "epoch": 4.709419904204364, "grad_norm": 0.2787940204143524, "learning_rate": 1.7336241552761698e-09, "loss": 0.1625, "step": 17698 }, { "epoch": 4.7096860031931875, "grad_norm": 0.29080522060394287, "learning_rate": 1.7304625522871375e-09, "loss": 0.1781, "step": 17699 }, { "epoch": 4.709952102182012, "grad_norm": 0.28315261006355286, "learning_rate": 1.7273038096691783e-09, "loss": 0.1844, "step": 17700 }, { "epoch": 4.710218201170836, "grad_norm": 0.3021635413169861, "learning_rate": 1.7241479275142412e-09, "loss": 0.1966, "step": 17701 }, { "epoch": 4.710484300159659, "grad_norm": 0.2966468930244446, "learning_rate": 1.7209949059142082e-09, "loss": 0.1911, "step": 17702 }, { "epoch": 4.710750399148483, "grad_norm": 0.32740887999534607, "learning_rate": 1.717844744960817e-09, "loss": 0.1903, "step": 17703 }, { "epoch": 4.711016498137307, "grad_norm": 0.30000266432762146, "learning_rate": 1.7146974447457718e-09, "loss": 0.184, "step": 17704 }, { "epoch": 4.7112825971261305, "grad_norm": 0.2724337875843048, "learning_rate": 1.7115530053607107e-09, "loss": 0.1597, "step": 17705 }, { "epoch": 4.711548696114955, "grad_norm": 0.3023325800895691, "learning_rate": 1.7084114268971273e-09, "loss": 0.1808, "step": 17706 }, { "epoch": 4.711814795103779, "grad_norm": 0.2962903380393982, "learning_rate": 1.7052727094464703e-09, "loss": 0.1695, "step": 17707 }, { "epoch": 4.712080894092603, "grad_norm": 0.2879208028316498, "learning_rate": 1.7021368531001113e-09, "loss": 0.1689, "step": 17708 }, { "epoch": 4.712346993081426, "grad_norm": 0.3019930422306061, "learning_rate": 1.6990038579493105e-09, "loss": 0.188, "step": 17709 }, { "epoch": 4.71261309207025, "grad_norm": 0.33511000871658325, "learning_rate": 1.6958737240852838e-09, "loss": 0.2031, "step": 17710 }, { "epoch": 4.7128791910590735, "grad_norm": 0.31438982486724854, "learning_rate": 1.6927464515991141e-09, "loss": 0.1937, "step": 17711 }, { "epoch": 4.713145290047898, "grad_norm": 0.299487829208374, "learning_rate": 1.6896220405818618e-09, "loss": 0.1783, "step": 17712 }, { "epoch": 4.713411389036722, "grad_norm": 0.2781984508037567, "learning_rate": 1.6865004911244429e-09, "loss": 0.1877, "step": 17713 }, { "epoch": 4.713677488025546, "grad_norm": 0.32705825567245483, "learning_rate": 1.6833818033177294e-09, "loss": 0.1897, "step": 17714 }, { "epoch": 4.713943587014369, "grad_norm": 0.2907677888870239, "learning_rate": 1.6802659772524819e-09, "loss": 0.1882, "step": 17715 }, { "epoch": 4.714209686003193, "grad_norm": 0.2962554693222046, "learning_rate": 1.6771530130194168e-09, "loss": 0.1984, "step": 17716 }, { "epoch": 4.714475784992017, "grad_norm": 0.2940623164176941, "learning_rate": 1.6740429107091391e-09, "loss": 0.1728, "step": 17717 }, { "epoch": 4.714741883980841, "grad_norm": 0.3170594274997711, "learning_rate": 1.6709356704121659e-09, "loss": 0.1876, "step": 17718 }, { "epoch": 4.715007982969665, "grad_norm": 0.29208239912986755, "learning_rate": 1.6678312922189463e-09, "loss": 0.1903, "step": 17719 }, { "epoch": 4.715274081958489, "grad_norm": 0.29675787687301636, "learning_rate": 1.664729776219842e-09, "loss": 0.1811, "step": 17720 }, { "epoch": 4.715540180947312, "grad_norm": 0.28469130396842957, "learning_rate": 1.6616311225051361e-09, "loss": 0.2078, "step": 17721 }, { "epoch": 4.715806279936136, "grad_norm": 0.2764645516872406, "learning_rate": 1.6585353311650008e-09, "loss": 0.1573, "step": 17722 }, { "epoch": 4.71607237892496, "grad_norm": 0.31679314374923706, "learning_rate": 1.6554424022895753e-09, "loss": 0.1919, "step": 17723 }, { "epoch": 4.716338477913784, "grad_norm": 0.29070034623146057, "learning_rate": 1.6523523359688652e-09, "loss": 0.1804, "step": 17724 }, { "epoch": 4.716604576902608, "grad_norm": 0.28010937571525574, "learning_rate": 1.6492651322928208e-09, "loss": 0.191, "step": 17725 }, { "epoch": 4.716870675891432, "grad_norm": 0.30055463314056396, "learning_rate": 1.6461807913513038e-09, "loss": 0.2031, "step": 17726 }, { "epoch": 4.717136774880255, "grad_norm": 0.29443225264549255, "learning_rate": 1.6430993132340976e-09, "loss": 0.1803, "step": 17727 }, { "epoch": 4.717402873869079, "grad_norm": 0.3048163950443268, "learning_rate": 1.6400206980308752e-09, "loss": 0.1891, "step": 17728 }, { "epoch": 4.717668972857903, "grad_norm": 0.3021255135536194, "learning_rate": 1.6369449458312757e-09, "loss": 0.1811, "step": 17729 }, { "epoch": 4.717935071846727, "grad_norm": 0.32140904664993286, "learning_rate": 1.6338720567248054e-09, "loss": 0.184, "step": 17730 }, { "epoch": 4.718201170835551, "grad_norm": 0.28134885430336, "learning_rate": 1.630802030800915e-09, "loss": 0.1703, "step": 17731 }, { "epoch": 4.718467269824375, "grad_norm": 0.3201770782470703, "learning_rate": 1.6277348681489666e-09, "loss": 0.1937, "step": 17732 }, { "epoch": 4.718733368813199, "grad_norm": 0.3045731484889984, "learning_rate": 1.624670568858222e-09, "loss": 0.1795, "step": 17733 }, { "epoch": 4.718999467802022, "grad_norm": 0.29027119278907776, "learning_rate": 1.6216091330178982e-09, "loss": 0.1939, "step": 17734 }, { "epoch": 4.719265566790846, "grad_norm": 0.2827020287513733, "learning_rate": 1.6185505607171025e-09, "loss": 0.1843, "step": 17735 }, { "epoch": 4.7195316657796695, "grad_norm": 0.3177674412727356, "learning_rate": 1.6154948520448408e-09, "loss": 0.193, "step": 17736 }, { "epoch": 4.719797764768494, "grad_norm": 0.3256179988384247, "learning_rate": 1.6124420070900757e-09, "loss": 0.206, "step": 17737 }, { "epoch": 4.720063863757318, "grad_norm": 0.29003530740737915, "learning_rate": 1.6093920259416694e-09, "loss": 0.1862, "step": 17738 }, { "epoch": 4.720329962746142, "grad_norm": 0.30504506826400757, "learning_rate": 1.6063449086883951e-09, "loss": 0.1918, "step": 17739 }, { "epoch": 4.720596061734965, "grad_norm": 0.32325682044029236, "learning_rate": 1.6033006554189376e-09, "loss": 0.213, "step": 17740 }, { "epoch": 4.720862160723789, "grad_norm": 0.31965798139572144, "learning_rate": 1.600259266221915e-09, "loss": 0.1972, "step": 17741 }, { "epoch": 4.721128259712613, "grad_norm": 0.2926989793777466, "learning_rate": 1.597220741185845e-09, "loss": 0.1833, "step": 17742 }, { "epoch": 4.721394358701437, "grad_norm": 0.27979129552841187, "learning_rate": 1.5941850803991908e-09, "loss": 0.1795, "step": 17743 }, { "epoch": 4.721660457690261, "grad_norm": 0.3179989457130432, "learning_rate": 1.5911522839503032e-09, "loss": 0.2079, "step": 17744 }, { "epoch": 4.721926556679085, "grad_norm": 0.3178718686103821, "learning_rate": 1.5881223519274456e-09, "loss": 0.2248, "step": 17745 }, { "epoch": 4.722192655667908, "grad_norm": 0.28695982694625854, "learning_rate": 1.5850952844188248e-09, "loss": 0.184, "step": 17746 }, { "epoch": 4.722458754656732, "grad_norm": 0.31588688492774963, "learning_rate": 1.5820710815125483e-09, "loss": 0.1852, "step": 17747 }, { "epoch": 4.722724853645556, "grad_norm": 0.3192441165447235, "learning_rate": 1.5790497432966565e-09, "loss": 0.188, "step": 17748 }, { "epoch": 4.72299095263438, "grad_norm": 0.2849709689617157, "learning_rate": 1.5760312698590572e-09, "loss": 0.176, "step": 17749 }, { "epoch": 4.723257051623204, "grad_norm": 0.2953890562057495, "learning_rate": 1.5730156612876465e-09, "loss": 0.1848, "step": 17750 }, { "epoch": 4.723523150612028, "grad_norm": 0.31634071469306946, "learning_rate": 1.5700029176701769e-09, "loss": 0.2121, "step": 17751 }, { "epoch": 4.723789249600852, "grad_norm": 0.31053856015205383, "learning_rate": 1.5669930390943554e-09, "loss": 0.2071, "step": 17752 }, { "epoch": 4.724055348589675, "grad_norm": 0.32083967328071594, "learning_rate": 1.5639860256477898e-09, "loss": 0.1767, "step": 17753 }, { "epoch": 4.724321447578499, "grad_norm": 0.29518043994903564, "learning_rate": 1.5609818774180105e-09, "loss": 0.1792, "step": 17754 }, { "epoch": 4.724587546567323, "grad_norm": 0.3107297122478485, "learning_rate": 1.5579805944924472e-09, "loss": 0.2019, "step": 17755 }, { "epoch": 4.724853645556147, "grad_norm": 0.3291359543800354, "learning_rate": 1.5549821769584636e-09, "loss": 0.2011, "step": 17756 }, { "epoch": 4.725119744544971, "grad_norm": 0.3633023202419281, "learning_rate": 1.5519866249033453e-09, "loss": 0.1791, "step": 17757 }, { "epoch": 4.725385843533795, "grad_norm": 0.27494943141937256, "learning_rate": 1.548993938414278e-09, "loss": 0.1705, "step": 17758 }, { "epoch": 4.725651942522618, "grad_norm": 0.3124629557132721, "learning_rate": 1.54600411757837e-09, "loss": 0.2046, "step": 17759 }, { "epoch": 4.725918041511442, "grad_norm": 0.33332768082618713, "learning_rate": 1.5430171624826516e-09, "loss": 0.2011, "step": 17760 }, { "epoch": 4.7261841405002665, "grad_norm": 0.29869747161865234, "learning_rate": 1.5400330732140532e-09, "loss": 0.2003, "step": 17761 }, { "epoch": 4.72645023948909, "grad_norm": 0.3025286793708801, "learning_rate": 1.5370518498594498e-09, "loss": 0.1993, "step": 17762 }, { "epoch": 4.726716338477914, "grad_norm": 0.27817481756210327, "learning_rate": 1.5340734925056164e-09, "loss": 0.1801, "step": 17763 }, { "epoch": 4.726982437466738, "grad_norm": 0.3036274313926697, "learning_rate": 1.5310980012392394e-09, "loss": 0.18, "step": 17764 }, { "epoch": 4.727248536455561, "grad_norm": 0.2947532832622528, "learning_rate": 1.5281253761469159e-09, "loss": 0.1966, "step": 17765 }, { "epoch": 4.727514635444385, "grad_norm": 0.2966509163379669, "learning_rate": 1.5251556173151881e-09, "loss": 0.1834, "step": 17766 }, { "epoch": 4.7277807344332095, "grad_norm": 0.28427866101264954, "learning_rate": 1.5221887248305087e-09, "loss": 0.1821, "step": 17767 }, { "epoch": 4.728046833422033, "grad_norm": 0.30415472388267517, "learning_rate": 1.519224698779198e-09, "loss": 0.1984, "step": 17768 }, { "epoch": 4.728312932410857, "grad_norm": 0.3209559917449951, "learning_rate": 1.5162635392475642e-09, "loss": 0.1776, "step": 17769 }, { "epoch": 4.728579031399681, "grad_norm": 0.29458633065223694, "learning_rate": 1.5133052463217832e-09, "loss": 0.1951, "step": 17770 }, { "epoch": 4.728845130388504, "grad_norm": 0.318891316652298, "learning_rate": 1.5103498200879749e-09, "loss": 0.1772, "step": 17771 }, { "epoch": 4.729111229377328, "grad_norm": 0.32952824234962463, "learning_rate": 1.5073972606321484e-09, "loss": 0.1951, "step": 17772 }, { "epoch": 4.729377328366152, "grad_norm": 0.27458783984184265, "learning_rate": 1.5044475680402568e-09, "loss": 0.1668, "step": 17773 }, { "epoch": 4.729643427354976, "grad_norm": 0.30381321907043457, "learning_rate": 1.501500742398154e-09, "loss": 0.1907, "step": 17774 }, { "epoch": 4.7299095263438, "grad_norm": 0.3000441789627075, "learning_rate": 1.498556783791627e-09, "loss": 0.1936, "step": 17775 }, { "epoch": 4.730175625332624, "grad_norm": 0.30558323860168457, "learning_rate": 1.4956156923063513e-09, "loss": 0.2025, "step": 17776 }, { "epoch": 4.730441724321448, "grad_norm": 0.316203773021698, "learning_rate": 1.4926774680279364e-09, "loss": 0.2033, "step": 17777 }, { "epoch": 4.730707823310271, "grad_norm": 0.3030838966369629, "learning_rate": 1.4897421110419028e-09, "loss": 0.1895, "step": 17778 }, { "epoch": 4.730973922299095, "grad_norm": 0.2731136977672577, "learning_rate": 1.4868096214337045e-09, "loss": 0.1714, "step": 17779 }, { "epoch": 4.731240021287919, "grad_norm": 0.28174450993537903, "learning_rate": 1.483879999288684e-09, "loss": 0.1869, "step": 17780 }, { "epoch": 4.731506120276743, "grad_norm": 0.3080122172832489, "learning_rate": 1.4809532446921292e-09, "loss": 0.1875, "step": 17781 }, { "epoch": 4.731772219265567, "grad_norm": 0.30009397864341736, "learning_rate": 1.4780293577292159e-09, "loss": 0.1875, "step": 17782 }, { "epoch": 4.732038318254391, "grad_norm": 0.27896982431411743, "learning_rate": 1.475108338485065e-09, "loss": 0.1791, "step": 17783 }, { "epoch": 4.732304417243214, "grad_norm": 0.29422304034233093, "learning_rate": 1.4721901870446974e-09, "loss": 0.1872, "step": 17784 }, { "epoch": 4.732570516232038, "grad_norm": 0.286335825920105, "learning_rate": 1.4692749034930451e-09, "loss": 0.1729, "step": 17785 }, { "epoch": 4.7328366152208625, "grad_norm": 0.3021477162837982, "learning_rate": 1.4663624879149738e-09, "loss": 0.1924, "step": 17786 }, { "epoch": 4.733102714209686, "grad_norm": 0.3326890170574188, "learning_rate": 1.4634529403952379e-09, "loss": 0.1735, "step": 17787 }, { "epoch": 4.73336881319851, "grad_norm": 0.3060012757778168, "learning_rate": 1.460546261018536e-09, "loss": 0.1801, "step": 17788 }, { "epoch": 4.733634912187334, "grad_norm": 0.29930374026298523, "learning_rate": 1.4576424498694894e-09, "loss": 0.1731, "step": 17789 }, { "epoch": 4.733901011176157, "grad_norm": 0.3227235972881317, "learning_rate": 1.4547415070325975e-09, "loss": 0.2006, "step": 17790 }, { "epoch": 4.734167110164981, "grad_norm": 0.2856941521167755, "learning_rate": 1.4518434325923257e-09, "loss": 0.1731, "step": 17791 }, { "epoch": 4.7344332091538055, "grad_norm": 0.3016877770423889, "learning_rate": 1.4489482266329955e-09, "loss": 0.2069, "step": 17792 }, { "epoch": 4.734699308142629, "grad_norm": 0.2930026650428772, "learning_rate": 1.4460558892389175e-09, "loss": 0.1798, "step": 17793 }, { "epoch": 4.734965407131453, "grad_norm": 0.32044142484664917, "learning_rate": 1.4431664204942463e-09, "loss": 0.203, "step": 17794 }, { "epoch": 4.735231506120277, "grad_norm": 0.312536358833313, "learning_rate": 1.4402798204831034e-09, "loss": 0.2072, "step": 17795 }, { "epoch": 4.7354976051091, "grad_norm": 0.3501836359500885, "learning_rate": 1.4373960892894998e-09, "loss": 0.2094, "step": 17796 }, { "epoch": 4.735763704097924, "grad_norm": 0.3159533441066742, "learning_rate": 1.4345152269973904e-09, "loss": 0.1882, "step": 17797 }, { "epoch": 4.7360298030867485, "grad_norm": 0.2964015007019043, "learning_rate": 1.431637233690619e-09, "loss": 0.1726, "step": 17798 }, { "epoch": 4.736295902075572, "grad_norm": 0.2950148582458496, "learning_rate": 1.4287621094529522e-09, "loss": 0.1872, "step": 17799 }, { "epoch": 4.736562001064396, "grad_norm": 0.294464647769928, "learning_rate": 1.4258898543680896e-09, "loss": 0.1861, "step": 17800 }, { "epoch": 4.73682810005322, "grad_norm": 0.3267112374305725, "learning_rate": 1.42302046851962e-09, "loss": 0.2107, "step": 17801 }, { "epoch": 4.737094199042044, "grad_norm": 0.3099227249622345, "learning_rate": 1.4201539519910878e-09, "loss": 0.1965, "step": 17802 }, { "epoch": 4.737360298030867, "grad_norm": 0.2866852581501007, "learning_rate": 1.4172903048658925e-09, "loss": 0.1772, "step": 17803 }, { "epoch": 4.7376263970196915, "grad_norm": 0.29573631286621094, "learning_rate": 1.4144295272274232e-09, "loss": 0.2056, "step": 17804 }, { "epoch": 4.737892496008515, "grad_norm": 0.32210665941238403, "learning_rate": 1.4115716191589355e-09, "loss": 0.205, "step": 17805 }, { "epoch": 4.738158594997339, "grad_norm": 0.26711153984069824, "learning_rate": 1.4087165807436186e-09, "loss": 0.1869, "step": 17806 }, { "epoch": 4.738424693986163, "grad_norm": 0.2825743854045868, "learning_rate": 1.4058644120645613e-09, "loss": 0.2052, "step": 17807 }, { "epoch": 4.738690792974987, "grad_norm": 0.3148232400417328, "learning_rate": 1.4030151132047975e-09, "loss": 0.2002, "step": 17808 }, { "epoch": 4.73895689196381, "grad_norm": 0.40282636880874634, "learning_rate": 1.4001686842472604e-09, "loss": 0.1809, "step": 17809 }, { "epoch": 4.739222990952634, "grad_norm": 0.33827728033065796, "learning_rate": 1.3973251252747953e-09, "loss": 0.1905, "step": 17810 }, { "epoch": 4.739489089941459, "grad_norm": 0.2814609706401825, "learning_rate": 1.3944844363701913e-09, "loss": 0.1807, "step": 17811 }, { "epoch": 4.739755188930282, "grad_norm": 0.29885944724082947, "learning_rate": 1.3916466176161046e-09, "loss": 0.1766, "step": 17812 }, { "epoch": 4.740021287919106, "grad_norm": 0.3620537221431732, "learning_rate": 1.388811669095158e-09, "loss": 0.2077, "step": 17813 }, { "epoch": 4.74028738690793, "grad_norm": 0.2627928853034973, "learning_rate": 1.385979590889863e-09, "loss": 0.1689, "step": 17814 }, { "epoch": 4.740553485896753, "grad_norm": 0.3404746949672699, "learning_rate": 1.383150383082643e-09, "loss": 0.1875, "step": 17815 }, { "epoch": 4.740819584885577, "grad_norm": 0.30776914954185486, "learning_rate": 1.3803240457558762e-09, "loss": 0.1746, "step": 17816 }, { "epoch": 4.7410856838744015, "grad_norm": 0.2961495518684387, "learning_rate": 1.377500578991797e-09, "loss": 0.1952, "step": 17817 }, { "epoch": 4.741351782863225, "grad_norm": 0.2890540659427643, "learning_rate": 1.3746799828726063e-09, "loss": 0.1851, "step": 17818 }, { "epoch": 4.741617881852049, "grad_norm": 0.31706932187080383, "learning_rate": 1.3718622574804161e-09, "loss": 0.2233, "step": 17819 }, { "epoch": 4.741883980840873, "grad_norm": 0.3209097683429718, "learning_rate": 1.3690474028972277e-09, "loss": 0.2209, "step": 17820 }, { "epoch": 4.742150079829696, "grad_norm": 0.2881035804748535, "learning_rate": 1.3662354192049641e-09, "loss": 0.1921, "step": 17821 }, { "epoch": 4.74241617881852, "grad_norm": 0.29850906133651733, "learning_rate": 1.3634263064855044e-09, "loss": 0.1997, "step": 17822 }, { "epoch": 4.7426822778073445, "grad_norm": 0.30636081099510193, "learning_rate": 1.3606200648205835e-09, "loss": 0.1809, "step": 17823 }, { "epoch": 4.742948376796168, "grad_norm": 0.2765592336654663, "learning_rate": 1.3578166942919022e-09, "loss": 0.1808, "step": 17824 }, { "epoch": 4.743214475784992, "grad_norm": 0.32514768838882446, "learning_rate": 1.3550161949810512e-09, "loss": 0.1883, "step": 17825 }, { "epoch": 4.743480574773816, "grad_norm": 0.2946750819683075, "learning_rate": 1.3522185669695653e-09, "loss": 0.1754, "step": 17826 }, { "epoch": 4.74374667376264, "grad_norm": 0.32610023021698, "learning_rate": 1.3494238103388455e-09, "loss": 0.2175, "step": 17827 }, { "epoch": 4.744012772751463, "grad_norm": 0.2997816503047943, "learning_rate": 1.3466319251702608e-09, "loss": 0.2001, "step": 17828 }, { "epoch": 4.7442788717402875, "grad_norm": 0.2964790463447571, "learning_rate": 1.3438429115450678e-09, "loss": 0.1787, "step": 17829 }, { "epoch": 4.744544970729111, "grad_norm": 0.3000880181789398, "learning_rate": 1.3410567695444574e-09, "loss": 0.1934, "step": 17830 }, { "epoch": 4.744811069717935, "grad_norm": 0.28884926438331604, "learning_rate": 1.3382734992495092e-09, "loss": 0.1692, "step": 17831 }, { "epoch": 4.745077168706759, "grad_norm": 0.31865963339805603, "learning_rate": 1.3354931007412474e-09, "loss": 0.2131, "step": 17832 }, { "epoch": 4.745343267695583, "grad_norm": 0.30985358357429504, "learning_rate": 1.3327155741005957e-09, "loss": 0.1886, "step": 17833 }, { "epoch": 4.745609366684406, "grad_norm": 0.320894718170166, "learning_rate": 1.329940919408412e-09, "loss": 0.1845, "step": 17834 }, { "epoch": 4.7458754656732305, "grad_norm": 0.2995074391365051, "learning_rate": 1.327169136745465e-09, "loss": 0.1858, "step": 17835 }, { "epoch": 4.746141564662055, "grad_norm": 0.3000018298625946, "learning_rate": 1.3244002261924125e-09, "loss": 0.1628, "step": 17836 }, { "epoch": 4.746407663650878, "grad_norm": 0.28877437114715576, "learning_rate": 1.3216341878298565e-09, "loss": 0.181, "step": 17837 }, { "epoch": 4.746673762639702, "grad_norm": 0.2910558581352234, "learning_rate": 1.3188710217383215e-09, "loss": 0.1872, "step": 17838 }, { "epoch": 4.746939861628526, "grad_norm": 0.31718748807907104, "learning_rate": 1.3161107279982319e-09, "loss": 0.2098, "step": 17839 }, { "epoch": 4.747205960617349, "grad_norm": 0.28458407521247864, "learning_rate": 1.3133533066899237e-09, "loss": 0.1907, "step": 17840 }, { "epoch": 4.7474720596061735, "grad_norm": 0.2840389013290405, "learning_rate": 1.310598757893655e-09, "loss": 0.1754, "step": 17841 }, { "epoch": 4.747738158594998, "grad_norm": 0.2980172634124756, "learning_rate": 1.307847081689617e-09, "loss": 0.1872, "step": 17842 }, { "epoch": 4.748004257583821, "grad_norm": 0.28740033507347107, "learning_rate": 1.3050982781579012e-09, "loss": 0.173, "step": 17843 }, { "epoch": 4.748270356572645, "grad_norm": 0.33380192518234253, "learning_rate": 1.3023523473785214e-09, "loss": 0.2106, "step": 17844 }, { "epoch": 4.748536455561469, "grad_norm": 0.2880871593952179, "learning_rate": 1.2996092894313915e-09, "loss": 0.1982, "step": 17845 }, { "epoch": 4.748802554550292, "grad_norm": 0.3219437599182129, "learning_rate": 1.2968691043963698e-09, "loss": 0.1925, "step": 17846 }, { "epoch": 4.7490686535391164, "grad_norm": 0.3234109580516815, "learning_rate": 1.2941317923532036e-09, "loss": 0.1937, "step": 17847 }, { "epoch": 4.749334752527941, "grad_norm": 0.27540841698646545, "learning_rate": 1.2913973533815736e-09, "loss": 0.1785, "step": 17848 }, { "epoch": 4.749600851516764, "grad_norm": 0.31107768416404724, "learning_rate": 1.2886657875610717e-09, "loss": 0.1982, "step": 17849 }, { "epoch": 4.749866950505588, "grad_norm": 0.3034299910068512, "learning_rate": 1.2859370949712122e-09, "loss": 0.1996, "step": 17850 }, { "epoch": 4.750133049494412, "grad_norm": 0.2878860533237457, "learning_rate": 1.2832112756914092e-09, "loss": 0.1801, "step": 17851 }, { "epoch": 4.750399148483236, "grad_norm": 0.2965973913669586, "learning_rate": 1.2804883298010105e-09, "loss": 0.1945, "step": 17852 }, { "epoch": 4.750665247472059, "grad_norm": 0.31540447473526, "learning_rate": 1.2777682573792748e-09, "loss": 0.2211, "step": 17853 }, { "epoch": 4.7509313464608836, "grad_norm": 0.3126133680343628, "learning_rate": 1.275051058505372e-09, "loss": 0.2024, "step": 17854 }, { "epoch": 4.751197445449708, "grad_norm": 0.2852059006690979, "learning_rate": 1.2723367332583946e-09, "loss": 0.1895, "step": 17855 }, { "epoch": 4.751463544438531, "grad_norm": 0.28226375579833984, "learning_rate": 1.2696252817173681e-09, "loss": 0.1741, "step": 17856 }, { "epoch": 4.751729643427355, "grad_norm": 0.3046046793460846, "learning_rate": 1.266916703961185e-09, "loss": 0.1873, "step": 17857 }, { "epoch": 4.751995742416179, "grad_norm": 0.28207990527153015, "learning_rate": 1.2642110000686935e-09, "loss": 0.1859, "step": 17858 }, { "epoch": 4.752261841405002, "grad_norm": 0.30671125650405884, "learning_rate": 1.2615081701186636e-09, "loss": 0.1848, "step": 17859 }, { "epoch": 4.7525279403938265, "grad_norm": 0.3068973422050476, "learning_rate": 1.258808214189755e-09, "loss": 0.1795, "step": 17860 }, { "epoch": 4.752794039382651, "grad_norm": 0.2908942699432373, "learning_rate": 1.256111132360571e-09, "loss": 0.1982, "step": 17861 }, { "epoch": 4.753060138371474, "grad_norm": 0.283154159784317, "learning_rate": 1.253416924709594e-09, "loss": 0.1949, "step": 17862 }, { "epoch": 4.753326237360298, "grad_norm": 0.3015051782131195, "learning_rate": 1.2507255913152603e-09, "loss": 0.1909, "step": 17863 }, { "epoch": 4.753592336349122, "grad_norm": 0.3011425733566284, "learning_rate": 1.248037132255908e-09, "loss": 0.1982, "step": 17864 }, { "epoch": 4.753858435337945, "grad_norm": 0.32049086689949036, "learning_rate": 1.2453515476097852e-09, "loss": 0.196, "step": 17865 }, { "epoch": 4.7541245343267695, "grad_norm": 0.30267444252967834, "learning_rate": 1.242668837455063e-09, "loss": 0.1993, "step": 17866 }, { "epoch": 4.754390633315594, "grad_norm": 0.30913445353507996, "learning_rate": 1.2399890018698344e-09, "loss": 0.2212, "step": 17867 }, { "epoch": 4.754656732304417, "grad_norm": 0.30944955348968506, "learning_rate": 1.2373120409321036e-09, "loss": 0.2107, "step": 17868 }, { "epoch": 4.754922831293241, "grad_norm": 0.32764381170272827, "learning_rate": 1.2346379547197749e-09, "loss": 0.189, "step": 17869 }, { "epoch": 4.755188930282065, "grad_norm": 0.2994249165058136, "learning_rate": 1.2319667433106973e-09, "loss": 0.1682, "step": 17870 }, { "epoch": 4.755455029270889, "grad_norm": 0.29264310002326965, "learning_rate": 1.2292984067826196e-09, "loss": 0.1823, "step": 17871 }, { "epoch": 4.7557211282597125, "grad_norm": 0.3023701608181, "learning_rate": 1.2266329452132018e-09, "loss": 0.193, "step": 17872 }, { "epoch": 4.755987227248537, "grad_norm": 0.2973717153072357, "learning_rate": 1.2239703586800377e-09, "loss": 0.2001, "step": 17873 }, { "epoch": 4.75625332623736, "grad_norm": 0.30002257227897644, "learning_rate": 1.2213106472606427e-09, "loss": 0.199, "step": 17874 }, { "epoch": 4.756519425226184, "grad_norm": 0.31727343797683716, "learning_rate": 1.2186538110324107e-09, "loss": 0.2004, "step": 17875 }, { "epoch": 4.756785524215008, "grad_norm": 0.32206282019615173, "learning_rate": 1.2159998500726797e-09, "loss": 0.1948, "step": 17876 }, { "epoch": 4.757051623203832, "grad_norm": 0.29692694544792175, "learning_rate": 1.2133487644586993e-09, "loss": 0.2058, "step": 17877 }, { "epoch": 4.7573177221926555, "grad_norm": 0.32781216502189636, "learning_rate": 1.2107005542676408e-09, "loss": 0.1987, "step": 17878 }, { "epoch": 4.75758382118148, "grad_norm": 0.30255383253097534, "learning_rate": 1.208055219576587e-09, "loss": 0.1833, "step": 17879 }, { "epoch": 4.757849920170304, "grad_norm": 0.32323455810546875, "learning_rate": 1.2054127604625319e-09, "loss": 0.189, "step": 17880 }, { "epoch": 4.758116019159127, "grad_norm": 0.30691346526145935, "learning_rate": 1.202773177002403e-09, "loss": 0.183, "step": 17881 }, { "epoch": 4.758382118147951, "grad_norm": 0.30004602670669556, "learning_rate": 1.2001364692730165e-09, "loss": 0.2008, "step": 17882 }, { "epoch": 4.758648217136775, "grad_norm": 0.3225141167640686, "learning_rate": 1.1975026373511332e-09, "loss": 0.2002, "step": 17883 }, { "epoch": 4.7589143161255985, "grad_norm": 0.31441396474838257, "learning_rate": 1.194871681313403e-09, "loss": 0.1944, "step": 17884 }, { "epoch": 4.759180415114423, "grad_norm": 0.3315030038356781, "learning_rate": 1.192243601236409e-09, "loss": 0.2092, "step": 17885 }, { "epoch": 4.759446514103247, "grad_norm": 0.3108427822589874, "learning_rate": 1.1896183971966568e-09, "loss": 0.202, "step": 17886 }, { "epoch": 4.75971261309207, "grad_norm": 0.2964847981929779, "learning_rate": 1.1869960692705628e-09, "loss": 0.177, "step": 17887 }, { "epoch": 4.759978712080894, "grad_norm": 0.2735958993434906, "learning_rate": 1.1843766175344439e-09, "loss": 0.1803, "step": 17888 }, { "epoch": 4.760244811069718, "grad_norm": 0.3004409670829773, "learning_rate": 1.181760042064539e-09, "loss": 0.1922, "step": 17889 }, { "epoch": 4.760510910058541, "grad_norm": 0.3519755005836487, "learning_rate": 1.1791463429370318e-09, "loss": 0.2214, "step": 17890 }, { "epoch": 4.760777009047366, "grad_norm": 0.2845865488052368, "learning_rate": 1.1765355202279836e-09, "loss": 0.1907, "step": 17891 }, { "epoch": 4.76104310803619, "grad_norm": 0.2922903001308441, "learning_rate": 1.1739275740134003e-09, "loss": 0.1915, "step": 17892 }, { "epoch": 4.761309207025013, "grad_norm": 0.27103888988494873, "learning_rate": 1.1713225043691765e-09, "loss": 0.1737, "step": 17893 }, { "epoch": 4.761575306013837, "grad_norm": 0.29270032048225403, "learning_rate": 1.168720311371152e-09, "loss": 0.2042, "step": 17894 }, { "epoch": 4.761841405002661, "grad_norm": 0.3207316994667053, "learning_rate": 1.1661209950950657e-09, "loss": 0.2063, "step": 17895 }, { "epoch": 4.762107503991485, "grad_norm": 0.3831127882003784, "learning_rate": 1.1635245556165685e-09, "loss": 0.1999, "step": 17896 }, { "epoch": 4.7623736029803085, "grad_norm": 0.34926295280456543, "learning_rate": 1.1609309930112554e-09, "loss": 0.213, "step": 17897 }, { "epoch": 4.762639701969133, "grad_norm": 0.2850615680217743, "learning_rate": 1.1583403073546105e-09, "loss": 0.1814, "step": 17898 }, { "epoch": 4.762905800957956, "grad_norm": 0.2858264744281769, "learning_rate": 1.1557524987220401e-09, "loss": 0.1717, "step": 17899 }, { "epoch": 4.76317189994678, "grad_norm": 0.32701751589775085, "learning_rate": 1.1531675671888618e-09, "loss": 0.1934, "step": 17900 }, { "epoch": 4.763437998935604, "grad_norm": 0.30520713329315186, "learning_rate": 1.1505855128303154e-09, "loss": 0.1796, "step": 17901 }, { "epoch": 4.763704097924428, "grad_norm": 0.29107630252838135, "learning_rate": 1.1480063357215852e-09, "loss": 0.1801, "step": 17902 }, { "epoch": 4.7639701969132515, "grad_norm": 0.28168293833732605, "learning_rate": 1.145430035937711e-09, "loss": 0.1762, "step": 17903 }, { "epoch": 4.764236295902076, "grad_norm": 0.2850582003593445, "learning_rate": 1.1428566135537e-09, "loss": 0.1811, "step": 17904 }, { "epoch": 4.7645023948909, "grad_norm": 0.31121423840522766, "learning_rate": 1.140286068644447e-09, "loss": 0.1951, "step": 17905 }, { "epoch": 4.764768493879723, "grad_norm": 0.32750874757766724, "learning_rate": 1.1377184012847819e-09, "loss": 0.1822, "step": 17906 }, { "epoch": 4.765034592868547, "grad_norm": 0.29122236371040344, "learning_rate": 1.1351536115494442e-09, "loss": 0.2133, "step": 17907 }, { "epoch": 4.765300691857371, "grad_norm": 0.29833897948265076, "learning_rate": 1.1325916995130858e-09, "loss": 0.1816, "step": 17908 }, { "epoch": 4.7655667908461945, "grad_norm": 0.28095296025276184, "learning_rate": 1.130032665250269e-09, "loss": 0.1862, "step": 17909 }, { "epoch": 4.765832889835019, "grad_norm": 0.706034243106842, "learning_rate": 1.1274765088354898e-09, "loss": 0.1875, "step": 17910 }, { "epoch": 4.766098988823843, "grad_norm": 0.3224725127220154, "learning_rate": 1.1249232303431666e-09, "loss": 0.1914, "step": 17911 }, { "epoch": 4.766365087812666, "grad_norm": 0.29971614480018616, "learning_rate": 1.1223728298475842e-09, "loss": 0.1998, "step": 17912 }, { "epoch": 4.76663118680149, "grad_norm": 0.303641676902771, "learning_rate": 1.1198253074229945e-09, "loss": 0.1907, "step": 17913 }, { "epoch": 4.766897285790314, "grad_norm": 0.28813686966896057, "learning_rate": 1.1172806631435493e-09, "loss": 0.1844, "step": 17914 }, { "epoch": 4.7671633847791375, "grad_norm": 0.31200119853019714, "learning_rate": 1.1147388970833228e-09, "loss": 0.1855, "step": 17915 }, { "epoch": 4.767429483767962, "grad_norm": 0.28517821431159973, "learning_rate": 1.1122000093162998e-09, "loss": 0.1874, "step": 17916 }, { "epoch": 4.767695582756786, "grad_norm": 0.29331421852111816, "learning_rate": 1.109663999916366e-09, "loss": 0.1885, "step": 17917 }, { "epoch": 4.767961681745609, "grad_norm": 0.2813337445259094, "learning_rate": 1.107130868957351e-09, "loss": 0.184, "step": 17918 }, { "epoch": 4.768227780734433, "grad_norm": 0.3053528070449829, "learning_rate": 1.104600616512974e-09, "loss": 0.1899, "step": 17919 }, { "epoch": 4.768493879723257, "grad_norm": 0.29569587111473083, "learning_rate": 1.102073242656909e-09, "loss": 0.208, "step": 17920 }, { "epoch": 4.768759978712081, "grad_norm": 0.28831103444099426, "learning_rate": 1.0995487474626975e-09, "loss": 0.1852, "step": 17921 }, { "epoch": 4.769026077700905, "grad_norm": 0.2789076566696167, "learning_rate": 1.097027131003836e-09, "loss": 0.182, "step": 17922 }, { "epoch": 4.769292176689729, "grad_norm": 0.30118557810783386, "learning_rate": 1.0945083933537102e-09, "loss": 0.2119, "step": 17923 }, { "epoch": 4.769558275678552, "grad_norm": 0.3080034554004669, "learning_rate": 1.0919925345856395e-09, "loss": 0.2101, "step": 17924 }, { "epoch": 4.769824374667376, "grad_norm": 0.2987576425075531, "learning_rate": 1.0894795547728542e-09, "loss": 0.1841, "step": 17925 }, { "epoch": 4.7700904736562, "grad_norm": 0.29283803701400757, "learning_rate": 1.0869694539884954e-09, "loss": 0.1925, "step": 17926 }, { "epoch": 4.770356572645024, "grad_norm": 0.2975948452949524, "learning_rate": 1.0844622323056386e-09, "loss": 0.1924, "step": 17927 }, { "epoch": 4.770622671633848, "grad_norm": 0.30702850222587585, "learning_rate": 1.0819578897972582e-09, "loss": 0.1879, "step": 17928 }, { "epoch": 4.770888770622672, "grad_norm": 0.2893216907978058, "learning_rate": 1.0794564265362405e-09, "loss": 0.1752, "step": 17929 }, { "epoch": 4.771154869611496, "grad_norm": 0.35051429271698, "learning_rate": 1.0769578425954052e-09, "loss": 0.1943, "step": 17930 }, { "epoch": 4.771420968600319, "grad_norm": 0.30034753680229187, "learning_rate": 1.0744621380474716e-09, "loss": 0.1969, "step": 17931 }, { "epoch": 4.771687067589143, "grad_norm": 0.31760042905807495, "learning_rate": 1.071969312965082e-09, "loss": 0.2003, "step": 17932 }, { "epoch": 4.771953166577967, "grad_norm": 0.2996245324611664, "learning_rate": 1.0694793674208114e-09, "loss": 0.1791, "step": 17933 }, { "epoch": 4.7722192655667905, "grad_norm": 0.32102084159851074, "learning_rate": 1.066992301487113e-09, "loss": 0.1984, "step": 17934 }, { "epoch": 4.772485364555615, "grad_norm": 0.3112804889678955, "learning_rate": 1.0645081152364066e-09, "loss": 0.1995, "step": 17935 }, { "epoch": 4.772751463544439, "grad_norm": 0.29158496856689453, "learning_rate": 1.062026808740979e-09, "loss": 0.1786, "step": 17936 }, { "epoch": 4.773017562533262, "grad_norm": 0.2907736897468567, "learning_rate": 1.0595483820730612e-09, "loss": 0.1822, "step": 17937 }, { "epoch": 4.773283661522086, "grad_norm": 0.33304867148399353, "learning_rate": 1.0570728353047952e-09, "loss": 0.1931, "step": 17938 }, { "epoch": 4.77354976051091, "grad_norm": 0.2984831929206848, "learning_rate": 1.0546001685082352e-09, "loss": 0.1966, "step": 17939 }, { "epoch": 4.7738158594997335, "grad_norm": 0.3065325617790222, "learning_rate": 1.0521303817553563e-09, "loss": 0.207, "step": 17940 }, { "epoch": 4.774081958488558, "grad_norm": 0.2882901132106781, "learning_rate": 1.0496634751180454e-09, "loss": 0.1862, "step": 17941 }, { "epoch": 4.774348057477382, "grad_norm": 0.2957538068294525, "learning_rate": 1.0471994486681014e-09, "loss": 0.1818, "step": 17942 }, { "epoch": 4.774614156466205, "grad_norm": 0.32386717200279236, "learning_rate": 1.044738302477266e-09, "loss": 0.1816, "step": 17943 }, { "epoch": 4.774880255455029, "grad_norm": 0.28918495774269104, "learning_rate": 1.0422800366171492e-09, "loss": 0.1587, "step": 17944 }, { "epoch": 4.775146354443853, "grad_norm": 0.3072220981121063, "learning_rate": 1.0398246511593268e-09, "loss": 0.1947, "step": 17945 }, { "epoch": 4.775412453432677, "grad_norm": 0.2996099591255188, "learning_rate": 1.0373721461752748e-09, "loss": 0.1881, "step": 17946 }, { "epoch": 4.775678552421501, "grad_norm": 0.3056359887123108, "learning_rate": 1.0349225217363477e-09, "loss": 0.1969, "step": 17947 }, { "epoch": 4.775944651410325, "grad_norm": 0.2985369563102722, "learning_rate": 1.0324757779138771e-09, "loss": 0.1736, "step": 17948 }, { "epoch": 4.776210750399148, "grad_norm": 0.3268560767173767, "learning_rate": 1.0300319147790726e-09, "loss": 0.1939, "step": 17949 }, { "epoch": 4.776476849387972, "grad_norm": 0.32928043603897095, "learning_rate": 1.0275909324030552e-09, "loss": 0.1969, "step": 17950 }, { "epoch": 4.776742948376796, "grad_norm": 0.2912999987602234, "learning_rate": 1.0251528308568902e-09, "loss": 0.1804, "step": 17951 }, { "epoch": 4.77700904736562, "grad_norm": 0.2966216504573822, "learning_rate": 1.022717610211532e-09, "loss": 0.1972, "step": 17952 }, { "epoch": 4.777275146354444, "grad_norm": 0.3270748555660248, "learning_rate": 1.0202852705378796e-09, "loss": 0.1943, "step": 17953 }, { "epoch": 4.777541245343268, "grad_norm": 0.3066524565219879, "learning_rate": 1.0178558119067315e-09, "loss": 0.1897, "step": 17954 }, { "epoch": 4.777807344332092, "grad_norm": 0.3168278634548187, "learning_rate": 1.0154292343887872e-09, "loss": 0.191, "step": 17955 }, { "epoch": 4.778073443320915, "grad_norm": 0.27696511149406433, "learning_rate": 1.0130055380546898e-09, "loss": 0.1596, "step": 17956 }, { "epoch": 4.778339542309739, "grad_norm": 0.2870175838470459, "learning_rate": 1.0105847229749832e-09, "loss": 0.1866, "step": 17957 }, { "epoch": 4.778605641298563, "grad_norm": 0.30861175060272217, "learning_rate": 1.0081667892201329e-09, "loss": 0.2119, "step": 17958 }, { "epoch": 4.778871740287387, "grad_norm": 0.3584563434123993, "learning_rate": 1.005751736860516e-09, "loss": 0.2142, "step": 17959 }, { "epoch": 4.779137839276211, "grad_norm": 0.3019482493400574, "learning_rate": 1.0033395659664323e-09, "loss": 0.203, "step": 17960 }, { "epoch": 4.779403938265035, "grad_norm": 0.29036131501197815, "learning_rate": 1.000930276608092e-09, "loss": 0.1723, "step": 17961 }, { "epoch": 4.779670037253858, "grad_norm": 0.2874140739440918, "learning_rate": 9.985238688556164e-10, "loss": 0.1754, "step": 17962 }, { "epoch": 4.779936136242682, "grad_norm": 0.2708258628845215, "learning_rate": 9.96120342779061e-10, "loss": 0.1715, "step": 17963 }, { "epoch": 4.780202235231506, "grad_norm": 0.2725723385810852, "learning_rate": 9.937196984483809e-10, "loss": 0.1789, "step": 17964 }, { "epoch": 4.7804683342203305, "grad_norm": 0.32278895378112793, "learning_rate": 9.913219359334645e-10, "loss": 0.201, "step": 17965 }, { "epoch": 4.780734433209154, "grad_norm": 0.30019116401672363, "learning_rate": 9.889270553040784e-10, "loss": 0.2093, "step": 17966 }, { "epoch": 4.781000532197978, "grad_norm": 0.29796740412712097, "learning_rate": 9.865350566299557e-10, "loss": 0.19, "step": 17967 }, { "epoch": 4.781266631186801, "grad_norm": 0.33931005001068115, "learning_rate": 9.841459399807073e-10, "loss": 0.186, "step": 17968 }, { "epoch": 4.781532730175625, "grad_norm": 0.3183441162109375, "learning_rate": 9.817597054258774e-10, "loss": 0.1974, "step": 17969 }, { "epoch": 4.781798829164449, "grad_norm": 0.28571969270706177, "learning_rate": 9.79376353034922e-10, "loss": 0.1809, "step": 17970 }, { "epoch": 4.782064928153273, "grad_norm": 0.3151979446411133, "learning_rate": 9.769958828772184e-10, "loss": 0.1923, "step": 17971 }, { "epoch": 4.782331027142097, "grad_norm": 0.29062291979789734, "learning_rate": 9.74618295022056e-10, "loss": 0.1818, "step": 17972 }, { "epoch": 4.782597126130921, "grad_norm": 0.28563621640205383, "learning_rate": 9.722435895386349e-10, "loss": 0.1775, "step": 17973 }, { "epoch": 4.782863225119745, "grad_norm": 0.2981509864330292, "learning_rate": 9.698717664960887e-10, "loss": 0.1984, "step": 17974 }, { "epoch": 4.783129324108568, "grad_norm": 0.29336145520210266, "learning_rate": 9.675028259634399e-10, "loss": 0.1987, "step": 17975 }, { "epoch": 4.783395423097392, "grad_norm": 0.30470141768455505, "learning_rate": 9.651367680096445e-10, "loss": 0.1795, "step": 17976 }, { "epoch": 4.783661522086216, "grad_norm": 0.28773707151412964, "learning_rate": 9.627735927035807e-10, "loss": 0.1744, "step": 17977 }, { "epoch": 4.78392762107504, "grad_norm": 0.2963818311691284, "learning_rate": 9.604133001140269e-10, "loss": 0.2097, "step": 17978 }, { "epoch": 4.784193720063864, "grad_norm": 0.7053566575050354, "learning_rate": 9.580558903096948e-10, "loss": 0.189, "step": 17979 }, { "epoch": 4.784459819052688, "grad_norm": 0.2721591293811798, "learning_rate": 9.557013633591848e-10, "loss": 0.1708, "step": 17980 }, { "epoch": 4.784725918041511, "grad_norm": 0.30670860409736633, "learning_rate": 9.533497193310535e-10, "loss": 0.1851, "step": 17981 }, { "epoch": 4.784992017030335, "grad_norm": 0.2928827702999115, "learning_rate": 9.510009582937351e-10, "loss": 0.1836, "step": 17982 }, { "epoch": 4.785258116019159, "grad_norm": 0.3020060956478119, "learning_rate": 9.486550803155968e-10, "loss": 0.1868, "step": 17983 }, { "epoch": 4.785524215007983, "grad_norm": 0.3291422128677368, "learning_rate": 9.463120854649287e-10, "loss": 0.2073, "step": 17984 }, { "epoch": 4.785790313996807, "grad_norm": 0.2988809645175934, "learning_rate": 9.439719738099317e-10, "loss": 0.1828, "step": 17985 }, { "epoch": 4.786056412985631, "grad_norm": 0.3223429322242737, "learning_rate": 9.416347454187068e-10, "loss": 0.2028, "step": 17986 }, { "epoch": 4.786322511974454, "grad_norm": 0.28447237610816956, "learning_rate": 9.393004003592886e-10, "loss": 0.1866, "step": 17987 }, { "epoch": 4.786588610963278, "grad_norm": 0.2993759214878082, "learning_rate": 9.36968938699645e-10, "loss": 0.1988, "step": 17988 }, { "epoch": 4.786854709952102, "grad_norm": 0.2777254283428192, "learning_rate": 9.346403605075991e-10, "loss": 0.1726, "step": 17989 }, { "epoch": 4.7871208089409265, "grad_norm": 0.2853977680206299, "learning_rate": 9.32314665850964e-10, "loss": 0.1756, "step": 17990 }, { "epoch": 4.78738690792975, "grad_norm": 0.2994762659072876, "learning_rate": 9.299918547974184e-10, "loss": 0.2029, "step": 17991 }, { "epoch": 4.787653006918574, "grad_norm": 0.30014553666114807, "learning_rate": 9.276719274145861e-10, "loss": 0.1956, "step": 17992 }, { "epoch": 4.787919105907397, "grad_norm": 0.3036828339099884, "learning_rate": 9.253548837699687e-10, "loss": 0.1927, "step": 17993 }, { "epoch": 4.788185204896221, "grad_norm": 0.2937643527984619, "learning_rate": 9.230407239310345e-10, "loss": 0.183, "step": 17994 }, { "epoch": 4.788451303885045, "grad_norm": 0.2904314398765564, "learning_rate": 9.207294479651295e-10, "loss": 0.1853, "step": 17995 }, { "epoch": 4.7887174028738695, "grad_norm": 0.3337799906730652, "learning_rate": 9.184210559395445e-10, "loss": 0.2104, "step": 17996 }, { "epoch": 4.788983501862693, "grad_norm": 0.29233017563819885, "learning_rate": 9.161155479214477e-10, "loss": 0.1962, "step": 17997 }, { "epoch": 4.789249600851517, "grad_norm": 0.29770228266716003, "learning_rate": 9.138129239779635e-10, "loss": 0.1929, "step": 17998 }, { "epoch": 4.789515699840341, "grad_norm": 0.30851835012435913, "learning_rate": 9.115131841761159e-10, "loss": 0.1931, "step": 17999 }, { "epoch": 4.789781798829164, "grad_norm": 0.3204856812953949, "learning_rate": 9.092163285828292e-10, "loss": 0.1798, "step": 18000 }, { "epoch": 4.790047897817988, "grad_norm": 0.3040373623371124, "learning_rate": 9.069223572649831e-10, "loss": 0.1729, "step": 18001 }, { "epoch": 4.7903139968068125, "grad_norm": 0.32673338055610657, "learning_rate": 9.046312702893244e-10, "loss": 0.2063, "step": 18002 }, { "epoch": 4.790580095795636, "grad_norm": 0.3254874050617218, "learning_rate": 9.023430677225441e-10, "loss": 0.2132, "step": 18003 }, { "epoch": 4.79084619478446, "grad_norm": 0.28350725769996643, "learning_rate": 9.000577496312556e-10, "loss": 0.1891, "step": 18004 }, { "epoch": 4.791112293773284, "grad_norm": 0.2862967848777771, "learning_rate": 8.977753160819835e-10, "loss": 0.1599, "step": 18005 }, { "epoch": 4.791378392762107, "grad_norm": 0.31121349334716797, "learning_rate": 8.954957671411523e-10, "loss": 0.189, "step": 18006 }, { "epoch": 4.791644491750931, "grad_norm": 0.2952388525009155, "learning_rate": 8.932191028751091e-10, "loss": 0.1888, "step": 18007 }, { "epoch": 4.7919105907397554, "grad_norm": 0.2681407332420349, "learning_rate": 8.909453233501451e-10, "loss": 0.1659, "step": 18008 }, { "epoch": 4.792176689728579, "grad_norm": 0.30398356914520264, "learning_rate": 8.886744286324189e-10, "loss": 0.1941, "step": 18009 }, { "epoch": 4.792442788717403, "grad_norm": 0.2818802297115326, "learning_rate": 8.864064187880438e-10, "loss": 0.1941, "step": 18010 }, { "epoch": 4.792708887706227, "grad_norm": 0.30530646443367004, "learning_rate": 8.841412938830339e-10, "loss": 0.2065, "step": 18011 }, { "epoch": 4.79297498669505, "grad_norm": 0.2898639142513275, "learning_rate": 8.818790539833143e-10, "loss": 0.1924, "step": 18012 }, { "epoch": 4.793241085683874, "grad_norm": 0.3276275396347046, "learning_rate": 8.796196991547433e-10, "loss": 0.1984, "step": 18013 }, { "epoch": 4.793507184672698, "grad_norm": 0.30833062529563904, "learning_rate": 8.773632294630684e-10, "loss": 0.184, "step": 18014 }, { "epoch": 4.7937732836615226, "grad_norm": 0.32611986994743347, "learning_rate": 8.751096449739926e-10, "loss": 0.1757, "step": 18015 }, { "epoch": 4.794039382650346, "grad_norm": 0.3116942048072815, "learning_rate": 8.728589457530855e-10, "loss": 0.2032, "step": 18016 }, { "epoch": 4.79430548163917, "grad_norm": 0.2982458472251892, "learning_rate": 8.706111318658837e-10, "loss": 0.1782, "step": 18017 }, { "epoch": 4.794571580627993, "grad_norm": 0.3057154417037964, "learning_rate": 8.683662033778016e-10, "loss": 0.1855, "step": 18018 }, { "epoch": 4.794837679616817, "grad_norm": 0.30597805976867676, "learning_rate": 8.66124160354198e-10, "loss": 0.1874, "step": 18019 }, { "epoch": 4.795103778605641, "grad_norm": 0.2800348699092865, "learning_rate": 8.638850028603096e-10, "loss": 0.1802, "step": 18020 }, { "epoch": 4.7953698775944655, "grad_norm": 0.28636154532432556, "learning_rate": 8.616487309613285e-10, "loss": 0.1835, "step": 18021 }, { "epoch": 4.795635976583289, "grad_norm": 0.2914332151412964, "learning_rate": 8.594153447223362e-10, "loss": 0.1841, "step": 18022 }, { "epoch": 4.795902075572113, "grad_norm": 0.3096632957458496, "learning_rate": 8.571848442083585e-10, "loss": 0.1893, "step": 18023 }, { "epoch": 4.796168174560937, "grad_norm": 0.280335932970047, "learning_rate": 8.549572294842988e-10, "loss": 0.176, "step": 18024 }, { "epoch": 4.79643427354976, "grad_norm": 0.30760547518730164, "learning_rate": 8.527325006150055e-10, "loss": 0.2036, "step": 18025 }, { "epoch": 4.796700372538584, "grad_norm": 0.3151033818721771, "learning_rate": 8.505106576652377e-10, "loss": 0.2017, "step": 18026 }, { "epoch": 4.7969664715274085, "grad_norm": 0.30252423882484436, "learning_rate": 8.48291700699677e-10, "loss": 0.1742, "step": 18027 }, { "epoch": 4.797232570516232, "grad_norm": 0.3180505931377411, "learning_rate": 8.460756297828942e-10, "loss": 0.2028, "step": 18028 }, { "epoch": 4.797498669505056, "grad_norm": 0.3230452835559845, "learning_rate": 8.438624449793929e-10, "loss": 0.1903, "step": 18029 }, { "epoch": 4.79776476849388, "grad_norm": 0.2952783703804016, "learning_rate": 8.416521463536108e-10, "loss": 0.1899, "step": 18030 }, { "epoch": 4.798030867482703, "grad_norm": 0.26924145221710205, "learning_rate": 8.394447339698629e-10, "loss": 0.1749, "step": 18031 }, { "epoch": 4.798296966471527, "grad_norm": 0.29673877358436584, "learning_rate": 8.37240207892409e-10, "loss": 0.184, "step": 18032 }, { "epoch": 4.7985630654603515, "grad_norm": 0.31664901971817017, "learning_rate": 8.3503856818542e-10, "loss": 0.1953, "step": 18033 }, { "epoch": 4.798829164449175, "grad_norm": 0.2527719736099243, "learning_rate": 8.328398149129778e-10, "loss": 0.1528, "step": 18034 }, { "epoch": 4.799095263437999, "grad_norm": 0.3352119028568268, "learning_rate": 8.30643948139087e-10, "loss": 0.1948, "step": 18035 }, { "epoch": 4.799361362426823, "grad_norm": 0.28787922859191895, "learning_rate": 8.28450967927663e-10, "loss": 0.1944, "step": 18036 }, { "epoch": 4.799627461415646, "grad_norm": 0.30858296155929565, "learning_rate": 8.262608743425437e-10, "loss": 0.2018, "step": 18037 }, { "epoch": 4.79989356040447, "grad_norm": 0.29988014698028564, "learning_rate": 8.240736674474558e-10, "loss": 0.2091, "step": 18038 }, { "epoch": 4.8001596593932945, "grad_norm": 0.3182743191719055, "learning_rate": 8.218893473060706e-10, "loss": 0.1982, "step": 18039 }, { "epoch": 4.800425758382119, "grad_norm": 0.3257894217967987, "learning_rate": 8.197079139819818e-10, "loss": 0.1915, "step": 18040 }, { "epoch": 4.800691857370942, "grad_norm": 0.2993874251842499, "learning_rate": 8.175293675386718e-10, "loss": 0.1964, "step": 18041 }, { "epoch": 4.800957956359766, "grad_norm": 0.29614120721817017, "learning_rate": 8.153537080395679e-10, "loss": 0.1841, "step": 18042 }, { "epoch": 4.801224055348589, "grad_norm": 0.31925061345100403, "learning_rate": 8.131809355479746e-10, "loss": 0.1692, "step": 18043 }, { "epoch": 4.801490154337413, "grad_norm": 0.3305957317352295, "learning_rate": 8.110110501271528e-10, "loss": 0.2202, "step": 18044 }, { "epoch": 4.8017562533262375, "grad_norm": 0.27546730637550354, "learning_rate": 8.088440518402628e-10, "loss": 0.1642, "step": 18045 }, { "epoch": 4.802022352315062, "grad_norm": 0.30374547839164734, "learning_rate": 8.066799407503655e-10, "loss": 0.1844, "step": 18046 }, { "epoch": 4.802288451303885, "grad_norm": 0.2961849570274353, "learning_rate": 8.045187169204659e-10, "loss": 0.1843, "step": 18047 }, { "epoch": 4.802554550292709, "grad_norm": 0.34592700004577637, "learning_rate": 8.023603804134693e-10, "loss": 0.1959, "step": 18048 }, { "epoch": 4.802820649281533, "grad_norm": 0.30013924837112427, "learning_rate": 8.002049312922033e-10, "loss": 0.1814, "step": 18049 }, { "epoch": 4.803086748270356, "grad_norm": 0.3122280240058899, "learning_rate": 7.980523696193952e-10, "loss": 0.2136, "step": 18050 }, { "epoch": 4.80335284725918, "grad_norm": 0.30388835072517395, "learning_rate": 7.959026954577175e-10, "loss": 0.1855, "step": 18051 }, { "epoch": 4.803618946248005, "grad_norm": 0.3026224672794342, "learning_rate": 7.937559088697199e-10, "loss": 0.2053, "step": 18052 }, { "epoch": 4.803885045236828, "grad_norm": 0.25449374318122864, "learning_rate": 7.916120099178969e-10, "loss": 0.1573, "step": 18053 }, { "epoch": 4.804151144225652, "grad_norm": 0.3053365647792816, "learning_rate": 7.894709986646652e-10, "loss": 0.1957, "step": 18054 }, { "epoch": 4.804417243214476, "grad_norm": 0.288463294506073, "learning_rate": 7.873328751723418e-10, "loss": 0.186, "step": 18055 }, { "epoch": 4.804683342203299, "grad_norm": 0.2794193625450134, "learning_rate": 7.851976395031434e-10, "loss": 0.1761, "step": 18056 }, { "epoch": 4.804949441192123, "grad_norm": 0.2975601255893707, "learning_rate": 7.830652917192427e-10, "loss": 0.2053, "step": 18057 }, { "epoch": 4.8052155401809475, "grad_norm": 0.2935267388820648, "learning_rate": 7.809358318826898e-10, "loss": 0.1754, "step": 18058 }, { "epoch": 4.805481639169771, "grad_norm": 0.3272092044353485, "learning_rate": 7.788092600554797e-10, "loss": 0.1973, "step": 18059 }, { "epoch": 4.805747738158595, "grad_norm": 0.2878219783306122, "learning_rate": 7.766855762994962e-10, "loss": 0.1804, "step": 18060 }, { "epoch": 4.806013837147419, "grad_norm": 0.2808850407600403, "learning_rate": 7.745647806765787e-10, "loss": 0.1742, "step": 18061 }, { "epoch": 4.806279936136242, "grad_norm": 0.27295735478401184, "learning_rate": 7.724468732484335e-10, "loss": 0.1817, "step": 18062 }, { "epoch": 4.806546035125066, "grad_norm": 0.2822936475276947, "learning_rate": 7.703318540767223e-10, "loss": 0.1829, "step": 18063 }, { "epoch": 4.8068121341138905, "grad_norm": 0.727479875087738, "learning_rate": 7.68219723223007e-10, "loss": 0.2005, "step": 18064 }, { "epoch": 4.807078233102715, "grad_norm": 0.3049604892730713, "learning_rate": 7.661104807487606e-10, "loss": 0.188, "step": 18065 }, { "epoch": 4.807344332091538, "grad_norm": 0.2944961190223694, "learning_rate": 7.640041267153785e-10, "loss": 0.1836, "step": 18066 }, { "epoch": 4.807610431080362, "grad_norm": 0.2833147346973419, "learning_rate": 7.619006611841672e-10, "loss": 0.1803, "step": 18067 }, { "epoch": 4.807876530069185, "grad_norm": 0.2884778380393982, "learning_rate": 7.598000842163666e-10, "loss": 0.1997, "step": 18068 }, { "epoch": 4.808142629058009, "grad_norm": 0.3401561975479126, "learning_rate": 7.577023958731055e-10, "loss": 0.1968, "step": 18069 }, { "epoch": 4.8084087280468335, "grad_norm": 0.30572110414505005, "learning_rate": 7.556075962154573e-10, "loss": 0.1852, "step": 18070 }, { "epoch": 4.808674827035658, "grad_norm": 0.28396686911582947, "learning_rate": 7.535156853043845e-10, "loss": 0.1917, "step": 18071 }, { "epoch": 4.808940926024481, "grad_norm": 0.3193773925304413, "learning_rate": 7.514266632007715e-10, "loss": 0.1748, "step": 18072 }, { "epoch": 4.809207025013305, "grad_norm": 0.2714765667915344, "learning_rate": 7.493405299654365e-10, "loss": 0.177, "step": 18073 }, { "epoch": 4.809473124002129, "grad_norm": 0.31175392866134644, "learning_rate": 7.472572856590975e-10, "loss": 0.1921, "step": 18074 }, { "epoch": 4.809739222990952, "grad_norm": 0.29476356506347656, "learning_rate": 7.45176930342395e-10, "loss": 0.1974, "step": 18075 }, { "epoch": 4.8100053219797765, "grad_norm": 0.2941574454307556, "learning_rate": 7.430994640758692e-10, "loss": 0.1902, "step": 18076 }, { "epoch": 4.810271420968601, "grad_norm": 0.2948009967803955, "learning_rate": 7.410248869200054e-10, "loss": 0.1799, "step": 18077 }, { "epoch": 4.810537519957424, "grad_norm": 0.31325721740722656, "learning_rate": 7.389531989351771e-10, "loss": 0.218, "step": 18078 }, { "epoch": 4.810803618946248, "grad_norm": 0.2922903597354889, "learning_rate": 7.368844001817031e-10, "loss": 0.1769, "step": 18079 }, { "epoch": 4.811069717935072, "grad_norm": 0.31630924344062805, "learning_rate": 7.348184907197907e-10, "loss": 0.2021, "step": 18080 }, { "epoch": 4.811335816923895, "grad_norm": 0.3045790493488312, "learning_rate": 7.327554706095584e-10, "loss": 0.1879, "step": 18081 }, { "epoch": 4.8116019159127195, "grad_norm": 0.29181545972824097, "learning_rate": 7.306953399110915e-10, "loss": 0.1894, "step": 18082 }, { "epoch": 4.811868014901544, "grad_norm": 0.2783309519290924, "learning_rate": 7.286380986843199e-10, "loss": 0.171, "step": 18083 }, { "epoch": 4.812134113890368, "grad_norm": 0.28600025177001953, "learning_rate": 7.265837469891289e-10, "loss": 0.183, "step": 18084 }, { "epoch": 4.812400212879191, "grad_norm": 0.28094691038131714, "learning_rate": 7.245322848853375e-10, "loss": 0.1805, "step": 18085 }, { "epoch": 4.812666311868015, "grad_norm": 0.32423967123031616, "learning_rate": 7.224837124326422e-10, "loss": 0.2071, "step": 18086 }, { "epoch": 4.812932410856838, "grad_norm": 0.37156495451927185, "learning_rate": 7.204380296906732e-10, "loss": 0.2075, "step": 18087 }, { "epoch": 4.813198509845662, "grad_norm": 0.29890531301498413, "learning_rate": 7.183952367189716e-10, "loss": 0.1812, "step": 18088 }, { "epoch": 4.813464608834487, "grad_norm": 0.2958775758743286, "learning_rate": 7.163553335770123e-10, "loss": 0.1827, "step": 18089 }, { "epoch": 4.813730707823311, "grad_norm": 0.3018261790275574, "learning_rate": 7.143183203241588e-10, "loss": 0.178, "step": 18090 }, { "epoch": 4.813996806812134, "grad_norm": 0.30630478262901306, "learning_rate": 7.12284197019708e-10, "loss": 0.2008, "step": 18091 }, { "epoch": 4.814262905800958, "grad_norm": 0.3027082085609436, "learning_rate": 7.102529637228683e-10, "loss": 0.2029, "step": 18092 }, { "epoch": 4.814529004789782, "grad_norm": 0.32637736201286316, "learning_rate": 7.082246204927589e-10, "loss": 0.2173, "step": 18093 }, { "epoch": 4.814795103778605, "grad_norm": 0.29960644245147705, "learning_rate": 7.061991673884104e-10, "loss": 0.1932, "step": 18094 }, { "epoch": 4.8150612027674295, "grad_norm": 0.2989920973777771, "learning_rate": 7.041766044688091e-10, "loss": 0.194, "step": 18095 }, { "epoch": 4.815327301756254, "grad_norm": 0.2821574807167053, "learning_rate": 7.021569317927967e-10, "loss": 0.1844, "step": 18096 }, { "epoch": 4.815593400745077, "grad_norm": 0.28999531269073486, "learning_rate": 7.001401494191705e-10, "loss": 0.1964, "step": 18097 }, { "epoch": 4.815859499733901, "grad_norm": 0.30217650532722473, "learning_rate": 6.981262574066393e-10, "loss": 0.1784, "step": 18098 }, { "epoch": 4.816125598722725, "grad_norm": 0.3049311935901642, "learning_rate": 6.961152558138117e-10, "loss": 0.203, "step": 18099 }, { "epoch": 4.816391697711548, "grad_norm": 0.2933395802974701, "learning_rate": 6.941071446992408e-10, "loss": 0.1799, "step": 18100 }, { "epoch": 4.8166577967003725, "grad_norm": 0.3043488562107086, "learning_rate": 6.921019241213577e-10, "loss": 0.1843, "step": 18101 }, { "epoch": 4.816923895689197, "grad_norm": 0.2946685552597046, "learning_rate": 6.900995941385379e-10, "loss": 0.1964, "step": 18102 }, { "epoch": 4.81718999467802, "grad_norm": 0.31079933047294617, "learning_rate": 6.881001548090681e-10, "loss": 0.1891, "step": 18103 }, { "epoch": 4.817456093666844, "grad_norm": 0.2898842394351959, "learning_rate": 6.861036061911352e-10, "loss": 0.1806, "step": 18104 }, { "epoch": 4.817722192655668, "grad_norm": 0.2899443805217743, "learning_rate": 6.841099483428592e-10, "loss": 0.1788, "step": 18105 }, { "epoch": 4.817988291644491, "grad_norm": 0.3007807433605194, "learning_rate": 6.821191813222826e-10, "loss": 0.1903, "step": 18106 }, { "epoch": 4.8182543906333155, "grad_norm": 0.3085962235927582, "learning_rate": 6.801313051873258e-10, "loss": 0.1716, "step": 18107 }, { "epoch": 4.81852048962214, "grad_norm": 0.3045879006385803, "learning_rate": 6.781463199958759e-10, "loss": 0.1786, "step": 18108 }, { "epoch": 4.818786588610964, "grad_norm": 0.3100505769252777, "learning_rate": 6.761642258056977e-10, "loss": 0.1968, "step": 18109 }, { "epoch": 4.819052687599787, "grad_norm": 0.29491183161735535, "learning_rate": 6.741850226744894e-10, "loss": 0.1873, "step": 18110 }, { "epoch": 4.819318786588611, "grad_norm": 0.304276704788208, "learning_rate": 6.722087106598606e-10, "loss": 0.1894, "step": 18111 }, { "epoch": 4.819584885577434, "grad_norm": 0.30115997791290283, "learning_rate": 6.702352898193319e-10, "loss": 0.1971, "step": 18112 }, { "epoch": 4.8198509845662585, "grad_norm": 0.3104836046695709, "learning_rate": 6.682647602103353e-10, "loss": 0.1909, "step": 18113 }, { "epoch": 4.820117083555083, "grad_norm": 0.30548006296157837, "learning_rate": 6.66297121890258e-10, "loss": 0.176, "step": 18114 }, { "epoch": 4.820383182543907, "grad_norm": 0.30896130204200745, "learning_rate": 6.643323749163432e-10, "loss": 0.1917, "step": 18115 }, { "epoch": 4.82064928153273, "grad_norm": 0.30814415216445923, "learning_rate": 6.623705193457896e-10, "loss": 0.1893, "step": 18116 }, { "epoch": 4.820915380521554, "grad_norm": 0.3160328269004822, "learning_rate": 6.604115552356959e-10, "loss": 0.1899, "step": 18117 }, { "epoch": 4.821181479510378, "grad_norm": 0.32517102360725403, "learning_rate": 6.584554826431055e-10, "loss": 0.1852, "step": 18118 }, { "epoch": 4.8214475784992015, "grad_norm": 0.29033514857292175, "learning_rate": 6.565023016249282e-10, "loss": 0.1812, "step": 18119 }, { "epoch": 4.821713677488026, "grad_norm": 0.2952576279640198, "learning_rate": 6.545520122380299e-10, "loss": 0.1959, "step": 18120 }, { "epoch": 4.82197977647685, "grad_norm": 0.29571154713630676, "learning_rate": 6.526046145391761e-10, "loss": 0.1954, "step": 18121 }, { "epoch": 4.822245875465673, "grad_norm": 0.30315935611724854, "learning_rate": 6.506601085850438e-10, "loss": 0.1695, "step": 18122 }, { "epoch": 4.822511974454497, "grad_norm": 0.30697572231292725, "learning_rate": 6.487184944322321e-10, "loss": 0.1959, "step": 18123 }, { "epoch": 4.822778073443321, "grad_norm": 0.31460264325141907, "learning_rate": 6.467797721372736e-10, "loss": 0.1771, "step": 18124 }, { "epoch": 4.8230441724321444, "grad_norm": 0.3278025984764099, "learning_rate": 6.448439417565787e-10, "loss": 0.2157, "step": 18125 }, { "epoch": 4.823310271420969, "grad_norm": 0.30104348063468933, "learning_rate": 6.429110033465135e-10, "loss": 0.187, "step": 18126 }, { "epoch": 4.823576370409793, "grad_norm": 0.2957039773464203, "learning_rate": 6.409809569633328e-10, "loss": 0.1972, "step": 18127 }, { "epoch": 4.823842469398616, "grad_norm": 0.33959394693374634, "learning_rate": 6.390538026632031e-10, "loss": 0.1993, "step": 18128 }, { "epoch": 4.82410856838744, "grad_norm": 0.2973523736000061, "learning_rate": 6.371295405022348e-10, "loss": 0.18, "step": 18129 }, { "epoch": 4.824374667376264, "grad_norm": 0.31171146035194397, "learning_rate": 6.352081705364276e-10, "loss": 0.1922, "step": 18130 }, { "epoch": 4.824640766365087, "grad_norm": 0.30443742871284485, "learning_rate": 6.332896928217257e-10, "loss": 0.1894, "step": 18131 }, { "epoch": 4.8249068653539116, "grad_norm": 0.2868706285953522, "learning_rate": 6.3137410741394e-10, "loss": 0.179, "step": 18132 }, { "epoch": 4.825172964342736, "grad_norm": 0.32803210616111755, "learning_rate": 6.294614143688592e-10, "loss": 0.2131, "step": 18133 }, { "epoch": 4.82543906333156, "grad_norm": 0.28523263335227966, "learning_rate": 6.275516137421388e-10, "loss": 0.1581, "step": 18134 }, { "epoch": 4.825705162320383, "grad_norm": 0.3123294413089752, "learning_rate": 6.256447055893676e-10, "loss": 0.1715, "step": 18135 }, { "epoch": 4.825971261309207, "grad_norm": 0.3158649802207947, "learning_rate": 6.237406899660679e-10, "loss": 0.212, "step": 18136 }, { "epoch": 4.82623736029803, "grad_norm": 0.3128940463066101, "learning_rate": 6.218395669276399e-10, "loss": 0.1758, "step": 18137 }, { "epoch": 4.8265034592868545, "grad_norm": 0.2904883623123169, "learning_rate": 6.199413365294282e-10, "loss": 0.1828, "step": 18138 }, { "epoch": 4.826769558275679, "grad_norm": 0.27067241072654724, "learning_rate": 6.180459988266884e-10, "loss": 0.1639, "step": 18139 }, { "epoch": 4.827035657264503, "grad_norm": 0.29956087470054626, "learning_rate": 6.161535538745877e-10, "loss": 0.1887, "step": 18140 }, { "epoch": 4.827301756253326, "grad_norm": 0.31668269634246826, "learning_rate": 6.142640017282041e-10, "loss": 0.1774, "step": 18141 }, { "epoch": 4.82756785524215, "grad_norm": 0.31312668323516846, "learning_rate": 6.123773424425493e-10, "loss": 0.2057, "step": 18142 }, { "epoch": 4.827833954230974, "grad_norm": 0.2921992540359497, "learning_rate": 6.104935760725349e-10, "loss": 0.1787, "step": 18143 }, { "epoch": 4.8281000532197975, "grad_norm": 0.31109145283699036, "learning_rate": 6.086127026729837e-10, "loss": 0.1853, "step": 18144 }, { "epoch": 4.828366152208622, "grad_norm": 0.2962035834789276, "learning_rate": 6.067347222986629e-10, "loss": 0.1925, "step": 18145 }, { "epoch": 4.828632251197446, "grad_norm": 0.30438223481178284, "learning_rate": 6.048596350042068e-10, "loss": 0.1978, "step": 18146 }, { "epoch": 4.828898350186269, "grad_norm": 0.30209678411483765, "learning_rate": 6.02987440844216e-10, "loss": 0.2078, "step": 18147 }, { "epoch": 4.829164449175093, "grad_norm": 0.3031829595565796, "learning_rate": 6.011181398731802e-10, "loss": 0.1821, "step": 18148 }, { "epoch": 4.829430548163917, "grad_norm": 0.3275560736656189, "learning_rate": 5.992517321455004e-10, "loss": 0.1818, "step": 18149 }, { "epoch": 4.8296966471527405, "grad_norm": 0.30937761068344116, "learning_rate": 5.97388217715522e-10, "loss": 0.1951, "step": 18150 }, { "epoch": 4.829962746141565, "grad_norm": 0.2866133153438568, "learning_rate": 5.955275966374796e-10, "loss": 0.1847, "step": 18151 }, { "epoch": 4.830228845130389, "grad_norm": 0.29086682200431824, "learning_rate": 5.936698689655184e-10, "loss": 0.1855, "step": 18152 }, { "epoch": 4.830494944119212, "grad_norm": 0.3020339906215668, "learning_rate": 5.918150347537177e-10, "loss": 0.1923, "step": 18153 }, { "epoch": 4.830761043108036, "grad_norm": 0.31942036747932434, "learning_rate": 5.899630940560785e-10, "loss": 0.2131, "step": 18154 }, { "epoch": 4.83102714209686, "grad_norm": 0.30991989374160767, "learning_rate": 5.881140469265023e-10, "loss": 0.1944, "step": 18155 }, { "epoch": 4.8312932410856835, "grad_norm": 0.29134300351142883, "learning_rate": 5.862678934188015e-10, "loss": 0.181, "step": 18156 }, { "epoch": 4.831559340074508, "grad_norm": 0.3338521420955658, "learning_rate": 5.84424633586711e-10, "loss": 0.2176, "step": 18157 }, { "epoch": 4.831825439063332, "grad_norm": 0.3037935793399811, "learning_rate": 5.825842674838877e-10, "loss": 0.1879, "step": 18158 }, { "epoch": 4.832091538052156, "grad_norm": 0.2753458619117737, "learning_rate": 5.807467951639111e-10, "loss": 0.1647, "step": 18159 }, { "epoch": 4.832357637040979, "grad_norm": 0.30200693011283875, "learning_rate": 5.789122166802496e-10, "loss": 0.1932, "step": 18160 }, { "epoch": 4.832623736029803, "grad_norm": 0.3339569568634033, "learning_rate": 5.770805320863159e-10, "loss": 0.2033, "step": 18161 }, { "epoch": 4.8328898350186265, "grad_norm": 0.3325735032558441, "learning_rate": 5.752517414354118e-10, "loss": 0.1866, "step": 18162 }, { "epoch": 4.833155934007451, "grad_norm": 0.29704102873802185, "learning_rate": 5.734258447807727e-10, "loss": 0.1905, "step": 18163 }, { "epoch": 4.833422032996275, "grad_norm": 0.905833899974823, "learning_rate": 5.71602842175567e-10, "loss": 0.1759, "step": 18164 }, { "epoch": 4.833688131985099, "grad_norm": 0.315051794052124, "learning_rate": 5.69782733672819e-10, "loss": 0.2007, "step": 18165 }, { "epoch": 4.833954230973922, "grad_norm": 0.29078713059425354, "learning_rate": 5.679655193255417e-10, "loss": 0.1804, "step": 18166 }, { "epoch": 4.834220329962746, "grad_norm": 0.3274259567260742, "learning_rate": 5.661511991866042e-10, "loss": 0.1925, "step": 18167 }, { "epoch": 4.83448642895157, "grad_norm": 0.30300092697143555, "learning_rate": 5.643397733088307e-10, "loss": 0.1876, "step": 18168 }, { "epoch": 4.834752527940394, "grad_norm": 0.30191436409950256, "learning_rate": 5.625312417449457e-10, "loss": 0.1856, "step": 18169 }, { "epoch": 4.835018626929218, "grad_norm": 0.3040039539337158, "learning_rate": 5.607256045475961e-10, "loss": 0.1797, "step": 18170 }, { "epoch": 4.835284725918042, "grad_norm": 0.25967463850975037, "learning_rate": 5.589228617693287e-10, "loss": 0.1672, "step": 18171 }, { "epoch": 4.835550824906865, "grad_norm": 0.297478049993515, "learning_rate": 5.571230134626236e-10, "loss": 0.1773, "step": 18172 }, { "epoch": 4.835816923895689, "grad_norm": 0.36152613162994385, "learning_rate": 5.553260596798615e-10, "loss": 0.1959, "step": 18173 }, { "epoch": 4.836083022884513, "grad_norm": 0.3133690655231476, "learning_rate": 5.535320004733557e-10, "loss": 0.1857, "step": 18174 }, { "epoch": 4.8363491218733365, "grad_norm": 0.29171833395957947, "learning_rate": 5.517408358953201e-10, "loss": 0.1877, "step": 18175 }, { "epoch": 4.836615220862161, "grad_norm": 0.3126582205295563, "learning_rate": 5.499525659978909e-10, "loss": 0.2069, "step": 18176 }, { "epoch": 4.836881319850985, "grad_norm": 0.33182060718536377, "learning_rate": 5.481671908331265e-10, "loss": 0.1999, "step": 18177 }, { "epoch": 4.837147418839808, "grad_norm": 0.2955508232116699, "learning_rate": 5.463847104529962e-10, "loss": 0.1939, "step": 18178 }, { "epoch": 4.837413517828632, "grad_norm": 0.3320784568786621, "learning_rate": 5.446051249093697e-10, "loss": 0.1954, "step": 18179 }, { "epoch": 4.837679616817456, "grad_norm": 0.29241323471069336, "learning_rate": 5.428284342540501e-10, "loss": 0.1975, "step": 18180 }, { "epoch": 4.8379457158062795, "grad_norm": 0.29923221468925476, "learning_rate": 5.410546385387738e-10, "loss": 0.1815, "step": 18181 }, { "epoch": 4.838211814795104, "grad_norm": 0.3282654285430908, "learning_rate": 5.392837378151438e-10, "loss": 0.1931, "step": 18182 }, { "epoch": 4.838477913783928, "grad_norm": 0.29692092537879944, "learning_rate": 5.37515732134719e-10, "loss": 0.1875, "step": 18183 }, { "epoch": 4.838744012772752, "grad_norm": 0.3021465241909027, "learning_rate": 5.357506215489583e-10, "loss": 0.202, "step": 18184 }, { "epoch": 4.839010111761575, "grad_norm": 0.2781126797199249, "learning_rate": 5.339884061092425e-10, "loss": 0.1784, "step": 18185 }, { "epoch": 4.839276210750399, "grad_norm": 0.2798377573490143, "learning_rate": 5.322290858668643e-10, "loss": 0.1788, "step": 18186 }, { "epoch": 4.8395423097392225, "grad_norm": 0.315326452255249, "learning_rate": 5.30472660873027e-10, "loss": 0.2032, "step": 18187 }, { "epoch": 4.839808408728047, "grad_norm": 0.28472790122032166, "learning_rate": 5.287191311788675e-10, "loss": 0.1889, "step": 18188 }, { "epoch": 4.840074507716871, "grad_norm": 0.2925773859024048, "learning_rate": 5.269684968354115e-10, "loss": 0.1851, "step": 18189 }, { "epoch": 4.840340606705695, "grad_norm": 0.29247429966926575, "learning_rate": 5.252207578936296e-10, "loss": 0.181, "step": 18190 }, { "epoch": 4.840606705694518, "grad_norm": 0.2878042459487915, "learning_rate": 5.23475914404381e-10, "loss": 0.1801, "step": 18191 }, { "epoch": 4.840872804683342, "grad_norm": 0.31351348757743835, "learning_rate": 5.217339664184695e-10, "loss": 0.1949, "step": 18192 }, { "epoch": 4.841138903672166, "grad_norm": 0.30746573209762573, "learning_rate": 5.199949139865878e-10, "loss": 0.1991, "step": 18193 }, { "epoch": 4.84140500266099, "grad_norm": 0.2994884252548218, "learning_rate": 5.182587571593511e-10, "loss": 0.1898, "step": 18194 }, { "epoch": 4.841671101649814, "grad_norm": 0.30853787064552307, "learning_rate": 5.165254959873078e-10, "loss": 0.2101, "step": 18195 }, { "epoch": 4.841937200638638, "grad_norm": 0.2893742322921753, "learning_rate": 5.147951305208953e-10, "loss": 0.1747, "step": 18196 }, { "epoch": 4.842203299627461, "grad_norm": 0.28994596004486084, "learning_rate": 5.130676608104845e-10, "loss": 0.1842, "step": 18197 }, { "epoch": 4.842469398616285, "grad_norm": 0.29157665371894836, "learning_rate": 5.113430869063573e-10, "loss": 0.1976, "step": 18198 }, { "epoch": 4.842735497605109, "grad_norm": 0.2972981929779053, "learning_rate": 5.09621408858718e-10, "loss": 0.1835, "step": 18199 }, { "epoch": 4.843001596593933, "grad_norm": 0.3017518222332001, "learning_rate": 5.079026267176823e-10, "loss": 0.1956, "step": 18200 }, { "epoch": 4.843267695582757, "grad_norm": 0.3337205648422241, "learning_rate": 5.061867405332654e-10, "loss": 0.2038, "step": 18201 }, { "epoch": 4.843533794571581, "grad_norm": 0.3278258740901947, "learning_rate": 5.044737503554164e-10, "loss": 0.2123, "step": 18202 }, { "epoch": 4.843799893560405, "grad_norm": 0.31522423028945923, "learning_rate": 5.027636562340065e-10, "loss": 0.1976, "step": 18203 }, { "epoch": 4.844065992549228, "grad_norm": 0.30970731377601624, "learning_rate": 5.01056458218796e-10, "loss": 0.184, "step": 18204 }, { "epoch": 4.844332091538052, "grad_norm": 0.29887405037879944, "learning_rate": 4.993521563594894e-10, "loss": 0.1778, "step": 18205 }, { "epoch": 4.844598190526876, "grad_norm": 0.32037603855133057, "learning_rate": 4.976507507056915e-10, "loss": 0.2056, "step": 18206 }, { "epoch": 4.8448642895157, "grad_norm": 0.30384328961372375, "learning_rate": 4.959522413069184e-10, "loss": 0.1888, "step": 18207 }, { "epoch": 4.845130388504524, "grad_norm": 0.3047689199447632, "learning_rate": 4.942566282126192e-10, "loss": 0.204, "step": 18208 }, { "epoch": 4.845396487493348, "grad_norm": 0.31110289692878723, "learning_rate": 4.925639114721547e-10, "loss": 0.1903, "step": 18209 }, { "epoch": 4.845662586482171, "grad_norm": 0.3125694692134857, "learning_rate": 4.908740911347741e-10, "loss": 0.1921, "step": 18210 }, { "epoch": 4.845928685470995, "grad_norm": 0.3072899580001831, "learning_rate": 4.891871672496828e-10, "loss": 0.1945, "step": 18211 }, { "epoch": 4.846194784459819, "grad_norm": 0.31465813517570496, "learning_rate": 4.875031398659746e-10, "loss": 0.2063, "step": 18212 }, { "epoch": 4.846460883448643, "grad_norm": 0.27212056517601013, "learning_rate": 4.85822009032677e-10, "loss": 0.1781, "step": 18213 }, { "epoch": 4.846726982437467, "grad_norm": 0.28705689311027527, "learning_rate": 4.841437747986954e-10, "loss": 0.1856, "step": 18214 }, { "epoch": 4.846993081426291, "grad_norm": 0.260343462228775, "learning_rate": 4.824684372129128e-10, "loss": 0.1674, "step": 18215 }, { "epoch": 4.847259180415114, "grad_norm": 0.2867841422557831, "learning_rate": 4.807959963240793e-10, "loss": 0.1886, "step": 18216 }, { "epoch": 4.847525279403938, "grad_norm": 0.3144204020500183, "learning_rate": 4.791264521808669e-10, "loss": 0.167, "step": 18217 }, { "epoch": 4.847791378392762, "grad_norm": 0.3042307496070862, "learning_rate": 4.774598048318923e-10, "loss": 0.1962, "step": 18218 }, { "epoch": 4.848057477381586, "grad_norm": 0.3083774447441101, "learning_rate": 4.75796054325639e-10, "loss": 0.1904, "step": 18219 }, { "epoch": 4.84832357637041, "grad_norm": 0.3104056119918823, "learning_rate": 4.74135200710557e-10, "loss": 0.1983, "step": 18220 }, { "epoch": 4.848589675359234, "grad_norm": 0.3351457417011261, "learning_rate": 4.724772440349856e-10, "loss": 0.2121, "step": 18221 }, { "epoch": 4.848855774348057, "grad_norm": 0.3003915846347809, "learning_rate": 4.70822184347186e-10, "loss": 0.1974, "step": 18222 }, { "epoch": 4.849121873336881, "grad_norm": 0.29303622245788574, "learning_rate": 4.691700216953198e-10, "loss": 0.1911, "step": 18223 }, { "epoch": 4.849387972325705, "grad_norm": 0.27615538239479065, "learning_rate": 4.675207561274819e-10, "loss": 0.1787, "step": 18224 }, { "epoch": 4.849654071314529, "grad_norm": 0.32034802436828613, "learning_rate": 4.658743876916893e-10, "loss": 0.2112, "step": 18225 }, { "epoch": 4.849920170303353, "grad_norm": 0.29183557629585266, "learning_rate": 4.6423091643584823e-10, "loss": 0.1745, "step": 18226 }, { "epoch": 4.850186269292177, "grad_norm": 0.27558332681655884, "learning_rate": 4.6259034240780925e-10, "loss": 0.1902, "step": 18227 }, { "epoch": 4.850452368281001, "grad_norm": 0.28949037194252014, "learning_rate": 4.609526656553231e-10, "loss": 0.1828, "step": 18228 }, { "epoch": 4.850718467269824, "grad_norm": 0.30476686358451843, "learning_rate": 4.593178862260405e-10, "loss": 0.2035, "step": 18229 }, { "epoch": 4.850984566258648, "grad_norm": 0.2820374667644501, "learning_rate": 4.5768600416756786e-10, "loss": 0.1868, "step": 18230 }, { "epoch": 4.851250665247472, "grad_norm": 0.30668675899505615, "learning_rate": 4.560570195274005e-10, "loss": 0.1936, "step": 18231 }, { "epoch": 4.851516764236296, "grad_norm": 0.3042137026786804, "learning_rate": 4.5443093235294494e-10, "loss": 0.1747, "step": 18232 }, { "epoch": 4.85178286322512, "grad_norm": 0.33440494537353516, "learning_rate": 4.528077426915411e-10, "loss": 0.205, "step": 18233 }, { "epoch": 4.852048962213944, "grad_norm": 0.30501797795295715, "learning_rate": 4.5118745059042894e-10, "loss": 0.1898, "step": 18234 }, { "epoch": 4.852315061202767, "grad_norm": 0.34111249446868896, "learning_rate": 4.495700560967708e-10, "loss": 0.1943, "step": 18235 }, { "epoch": 4.852581160191591, "grad_norm": 0.2940026521682739, "learning_rate": 4.4795555925766226e-10, "loss": 0.1819, "step": 18236 }, { "epoch": 4.8528472591804155, "grad_norm": 0.2998597025871277, "learning_rate": 4.4634396012006583e-10, "loss": 0.1811, "step": 18237 }, { "epoch": 4.853113358169239, "grad_norm": 0.2961556017398834, "learning_rate": 4.4473525873092165e-10, "loss": 0.1721, "step": 18238 }, { "epoch": 4.853379457158063, "grad_norm": 0.32997721433639526, "learning_rate": 4.431294551370368e-10, "loss": 0.1992, "step": 18239 }, { "epoch": 4.853645556146887, "grad_norm": 0.28806668519973755, "learning_rate": 4.4152654938515165e-10, "loss": 0.1794, "step": 18240 }, { "epoch": 4.85391165513571, "grad_norm": 0.2937168776988983, "learning_rate": 4.399265415219289e-10, "loss": 0.1944, "step": 18241 }, { "epoch": 4.854177754124534, "grad_norm": 0.3084243834018707, "learning_rate": 4.383294315939423e-10, "loss": 0.1863, "step": 18242 }, { "epoch": 4.8544438531133585, "grad_norm": 0.3162509799003601, "learning_rate": 4.36735219647677e-10, "loss": 0.2183, "step": 18243 }, { "epoch": 4.854709952102182, "grad_norm": 0.31431126594543457, "learning_rate": 4.351439057295292e-10, "loss": 0.2036, "step": 18244 }, { "epoch": 4.854976051091006, "grad_norm": 0.3026432394981384, "learning_rate": 4.335554898858396e-10, "loss": 0.1842, "step": 18245 }, { "epoch": 4.85524215007983, "grad_norm": 0.32084473967552185, "learning_rate": 4.3196997216281563e-10, "loss": 0.1828, "step": 18246 }, { "epoch": 4.855508249068653, "grad_norm": 0.3076423704624176, "learning_rate": 4.303873526066204e-10, "loss": 0.1902, "step": 18247 }, { "epoch": 4.855774348057477, "grad_norm": 0.2863953411579132, "learning_rate": 4.288076312633282e-10, "loss": 0.1876, "step": 18248 }, { "epoch": 4.8560404470463014, "grad_norm": 0.29597020149230957, "learning_rate": 4.2723080817890224e-10, "loss": 0.1884, "step": 18249 }, { "epoch": 4.856306546035125, "grad_norm": 0.2986955940723419, "learning_rate": 4.256568833992502e-10, "loss": 0.172, "step": 18250 }, { "epoch": 4.856572645023949, "grad_norm": 0.3125002384185791, "learning_rate": 4.2408585697019106e-10, "loss": 0.1867, "step": 18251 }, { "epoch": 4.856838744012773, "grad_norm": 0.3060840368270874, "learning_rate": 4.225177289374438e-10, "loss": 0.1953, "step": 18252 }, { "epoch": 4.857104843001597, "grad_norm": 0.30110520124435425, "learning_rate": 4.209524993466496e-10, "loss": 0.193, "step": 18253 }, { "epoch": 4.85737094199042, "grad_norm": 0.32955965399742126, "learning_rate": 4.1939016824338314e-10, "loss": 0.2221, "step": 18254 }, { "epoch": 4.857637040979244, "grad_norm": 0.2978127896785736, "learning_rate": 4.178307356731081e-10, "loss": 0.1794, "step": 18255 }, { "epoch": 4.857903139968068, "grad_norm": 0.33173757791519165, "learning_rate": 4.162742016812104e-10, "loss": 0.2217, "step": 18256 }, { "epoch": 4.858169238956892, "grad_norm": 0.3103583753108978, "learning_rate": 4.1472056631300935e-10, "loss": 0.2011, "step": 18257 }, { "epoch": 4.858435337945716, "grad_norm": 0.28016403317451477, "learning_rate": 4.131698296137243e-10, "loss": 0.1754, "step": 18258 }, { "epoch": 4.85870143693454, "grad_norm": 0.3159780204296112, "learning_rate": 4.1162199162848577e-10, "loss": 0.2016, "step": 18259 }, { "epoch": 4.858967535923363, "grad_norm": 0.27045896649360657, "learning_rate": 4.100770524023578e-10, "loss": 0.1707, "step": 18260 }, { "epoch": 4.859233634912187, "grad_norm": 0.28691965341567993, "learning_rate": 4.0853501198029326e-10, "loss": 0.1793, "step": 18261 }, { "epoch": 4.8594997339010115, "grad_norm": 0.29969772696495056, "learning_rate": 4.0699587040720075e-10, "loss": 0.1932, "step": 18262 }, { "epoch": 4.859765832889835, "grad_norm": 0.34985029697418213, "learning_rate": 4.0545962772786655e-10, "loss": 0.2011, "step": 18263 }, { "epoch": 4.860031931878659, "grad_norm": 0.27440062165260315, "learning_rate": 4.039262839869995e-10, "loss": 0.1775, "step": 18264 }, { "epoch": 4.860298030867483, "grad_norm": 0.28339430689811707, "learning_rate": 4.023958392292415e-10, "loss": 0.1745, "step": 18265 }, { "epoch": 4.860564129856306, "grad_norm": 0.3080093264579773, "learning_rate": 4.0086829349914585e-10, "loss": 0.175, "step": 18266 }, { "epoch": 4.86083022884513, "grad_norm": 0.4030227065086365, "learning_rate": 3.9934364684115483e-10, "loss": 0.1806, "step": 18267 }, { "epoch": 4.8610963278339545, "grad_norm": 0.2838766276836395, "learning_rate": 3.9782189929967737e-10, "loss": 0.191, "step": 18268 }, { "epoch": 4.861362426822778, "grad_norm": 0.30372750759124756, "learning_rate": 3.9630305091897797e-10, "loss": 0.1886, "step": 18269 }, { "epoch": 4.861628525811602, "grad_norm": 0.30808430910110474, "learning_rate": 3.947871017432769e-10, "loss": 0.2084, "step": 18270 }, { "epoch": 4.861894624800426, "grad_norm": 0.2875027358531952, "learning_rate": 3.932740518167055e-10, "loss": 0.194, "step": 18271 }, { "epoch": 4.862160723789249, "grad_norm": 0.2794892191886902, "learning_rate": 3.9176390118330627e-10, "loss": 0.1828, "step": 18272 }, { "epoch": 4.862426822778073, "grad_norm": 0.31131961941719055, "learning_rate": 3.902566498870219e-10, "loss": 0.2082, "step": 18273 }, { "epoch": 4.8626929217668975, "grad_norm": 0.2702406644821167, "learning_rate": 3.8875229797172836e-10, "loss": 0.1716, "step": 18274 }, { "epoch": 4.862959020755721, "grad_norm": 0.30390438437461853, "learning_rate": 3.87250845481224e-10, "loss": 0.1815, "step": 18275 }, { "epoch": 4.863225119744545, "grad_norm": 0.30593323707580566, "learning_rate": 3.8575229245920714e-10, "loss": 0.192, "step": 18276 }, { "epoch": 4.863491218733369, "grad_norm": 0.29130905866622925, "learning_rate": 3.8425663894928737e-10, "loss": 0.1791, "step": 18277 }, { "epoch": 4.863757317722193, "grad_norm": 0.2817822992801666, "learning_rate": 3.8276388499500767e-10, "loss": 0.1726, "step": 18278 }, { "epoch": 4.864023416711016, "grad_norm": 0.28292369842529297, "learning_rate": 3.812740306398221e-10, "loss": 0.1926, "step": 18279 }, { "epoch": 4.8642895156998405, "grad_norm": 0.27973565459251404, "learning_rate": 3.79787075927096e-10, "loss": 0.1849, "step": 18280 }, { "epoch": 4.864555614688664, "grad_norm": 0.29728803038597107, "learning_rate": 3.7830302090010593e-10, "loss": 0.2006, "step": 18281 }, { "epoch": 4.864821713677488, "grad_norm": 0.32192665338516235, "learning_rate": 3.768218656020505e-10, "loss": 0.1882, "step": 18282 }, { "epoch": 4.865087812666312, "grad_norm": 0.29476436972618103, "learning_rate": 3.753436100760288e-10, "loss": 0.1916, "step": 18283 }, { "epoch": 4.865353911655136, "grad_norm": 0.31794798374176025, "learning_rate": 3.7386825436509504e-10, "loss": 0.1977, "step": 18284 }, { "epoch": 4.865620010643959, "grad_norm": 0.29793354868888855, "learning_rate": 3.7239579851217063e-10, "loss": 0.2159, "step": 18285 }, { "epoch": 4.8658861096327835, "grad_norm": 0.3072263300418854, "learning_rate": 3.709262425601323e-10, "loss": 0.2008, "step": 18286 }, { "epoch": 4.866152208621608, "grad_norm": 0.30919188261032104, "learning_rate": 3.694595865517458e-10, "loss": 0.1884, "step": 18287 }, { "epoch": 4.866418307610431, "grad_norm": 0.27617961168289185, "learning_rate": 3.679958305296993e-10, "loss": 0.1898, "step": 18288 }, { "epoch": 4.866684406599255, "grad_norm": 0.2882675230503082, "learning_rate": 3.665349745366031e-10, "loss": 0.1849, "step": 18289 }, { "epoch": 4.866950505588079, "grad_norm": 0.30491262674331665, "learning_rate": 3.6507701861497877e-10, "loss": 0.2006, "step": 18290 }, { "epoch": 4.867216604576902, "grad_norm": 0.30008599162101746, "learning_rate": 3.6362196280725896e-10, "loss": 0.1927, "step": 18291 }, { "epoch": 4.867482703565726, "grad_norm": 0.30373725295066833, "learning_rate": 3.621698071557988e-10, "loss": 0.2136, "step": 18292 }, { "epoch": 4.867748802554551, "grad_norm": 0.33004558086395264, "learning_rate": 3.607205517028755e-10, "loss": 0.1874, "step": 18293 }, { "epoch": 4.868014901543374, "grad_norm": 0.28574925661087036, "learning_rate": 3.592741964906665e-10, "loss": 0.1771, "step": 18294 }, { "epoch": 4.868281000532198, "grad_norm": 0.2927454113960266, "learning_rate": 3.5783074156127133e-10, "loss": 0.1834, "step": 18295 }, { "epoch": 4.868547099521022, "grad_norm": 0.3058396279811859, "learning_rate": 3.5639018695670095e-10, "loss": 0.2022, "step": 18296 }, { "epoch": 4.868813198509845, "grad_norm": 0.30067870020866394, "learning_rate": 3.549525327188885e-10, "loss": 0.181, "step": 18297 }, { "epoch": 4.869079297498669, "grad_norm": 0.3377009630203247, "learning_rate": 3.5351777888968924e-10, "loss": 0.2143, "step": 18298 }, { "epoch": 4.8693453964874935, "grad_norm": 0.2870309352874756, "learning_rate": 3.520859255108588e-10, "loss": 0.1928, "step": 18299 }, { "epoch": 4.869611495476317, "grad_norm": 0.3018728494644165, "learning_rate": 3.506569726240638e-10, "loss": 0.1829, "step": 18300 }, { "epoch": 4.869877594465141, "grad_norm": 0.30270060896873474, "learning_rate": 3.492309202709154e-10, "loss": 0.1883, "step": 18301 }, { "epoch": 4.870143693453965, "grad_norm": 0.3104242980480194, "learning_rate": 3.478077684929137e-10, "loss": 0.21, "step": 18302 }, { "epoch": 4.870409792442789, "grad_norm": 0.29846304655075073, "learning_rate": 3.463875173314812e-10, "loss": 0.1757, "step": 18303 }, { "epoch": 4.870675891431612, "grad_norm": 0.3072642683982849, "learning_rate": 3.4497016682795145e-10, "loss": 0.1903, "step": 18304 }, { "epoch": 4.8709419904204365, "grad_norm": 0.2912832200527191, "learning_rate": 3.4355571702360254e-10, "loss": 0.1961, "step": 18305 }, { "epoch": 4.87120808940926, "grad_norm": 0.27291780710220337, "learning_rate": 3.4214416795957935e-10, "loss": 0.1734, "step": 18306 }, { "epoch": 4.871474188398084, "grad_norm": 0.33406445384025574, "learning_rate": 3.407355196769934e-10, "loss": 0.203, "step": 18307 }, { "epoch": 4.871740287386908, "grad_norm": 0.32007259130477905, "learning_rate": 3.3932977221683423e-10, "loss": 0.2044, "step": 18308 }, { "epoch": 4.872006386375732, "grad_norm": 0.2710858881473541, "learning_rate": 3.3792692562000236e-10, "loss": 0.1681, "step": 18309 }, { "epoch": 4.872272485364555, "grad_norm": 0.2629937529563904, "learning_rate": 3.3652697992736513e-10, "loss": 0.1614, "step": 18310 }, { "epoch": 4.8725385843533795, "grad_norm": 0.282345175743103, "learning_rate": 3.351299351796455e-10, "loss": 0.1899, "step": 18311 }, { "epoch": 4.872804683342204, "grad_norm": 0.2755945026874542, "learning_rate": 3.3373579141751097e-10, "loss": 0.179, "step": 18312 }, { "epoch": 4.873070782331027, "grad_norm": 0.3026883006095886, "learning_rate": 3.3234454868156237e-10, "loss": 0.1752, "step": 18313 }, { "epoch": 4.873336881319851, "grad_norm": 0.3305718004703522, "learning_rate": 3.3095620701226735e-10, "loss": 0.1921, "step": 18314 }, { "epoch": 4.873602980308675, "grad_norm": 0.3344559371471405, "learning_rate": 3.295707664500491e-10, "loss": 0.202, "step": 18315 }, { "epoch": 4.873869079297498, "grad_norm": 0.28596097230911255, "learning_rate": 3.2818822703524205e-10, "loss": 0.1834, "step": 18316 }, { "epoch": 4.8741351782863225, "grad_norm": 0.30026161670684814, "learning_rate": 3.268085888080696e-10, "loss": 0.1812, "step": 18317 }, { "epoch": 4.874401277275147, "grad_norm": 0.31702059507369995, "learning_rate": 3.254318518087107e-10, "loss": 0.2041, "step": 18318 }, { "epoch": 4.87466737626397, "grad_norm": 0.306405246257782, "learning_rate": 3.240580160772222e-10, "loss": 0.1861, "step": 18319 }, { "epoch": 4.874933475252794, "grad_norm": 0.29182863235473633, "learning_rate": 3.2268708165359424e-10, "loss": 0.1766, "step": 18320 }, { "epoch": 4.875199574241618, "grad_norm": 0.2693941593170166, "learning_rate": 3.213190485777506e-10, "loss": 0.1635, "step": 18321 }, { "epoch": 4.875465673230442, "grad_norm": 0.2855377495288849, "learning_rate": 3.199539168894816e-10, "loss": 0.1894, "step": 18322 }, { "epoch": 4.8757317722192655, "grad_norm": 0.30623361468315125, "learning_rate": 3.185916866285443e-10, "loss": 0.2106, "step": 18323 }, { "epoch": 4.87599787120809, "grad_norm": 0.3009551763534546, "learning_rate": 3.172323578345848e-10, "loss": 0.1939, "step": 18324 }, { "epoch": 4.876263970196913, "grad_norm": 0.28218650817871094, "learning_rate": 3.1587593054716034e-10, "loss": 0.19, "step": 18325 }, { "epoch": 4.876530069185737, "grad_norm": 0.31983017921447754, "learning_rate": 3.1452240480577264e-10, "loss": 0.1921, "step": 18326 }, { "epoch": 4.876796168174561, "grad_norm": 0.283497154712677, "learning_rate": 3.131717806497902e-10, "loss": 0.1755, "step": 18327 }, { "epoch": 4.877062267163385, "grad_norm": 0.28508052229881287, "learning_rate": 3.118240581185594e-10, "loss": 0.1887, "step": 18328 }, { "epoch": 4.877328366152208, "grad_norm": 0.2916746437549591, "learning_rate": 3.104792372512821e-10, "loss": 0.1843, "step": 18329 }, { "epoch": 4.877594465141033, "grad_norm": 0.31323686242103577, "learning_rate": 3.0913731808711597e-10, "loss": 0.1971, "step": 18330 }, { "epoch": 4.877860564129857, "grad_norm": 0.27395007014274597, "learning_rate": 3.077983006651186e-10, "loss": 0.1691, "step": 18331 }, { "epoch": 4.87812666311868, "grad_norm": 0.30148956179618835, "learning_rate": 3.064621850242588e-10, "loss": 0.2062, "step": 18332 }, { "epoch": 4.878392762107504, "grad_norm": 0.31779909133911133, "learning_rate": 3.0512897120343883e-10, "loss": 0.195, "step": 18333 }, { "epoch": 4.878658861096328, "grad_norm": 0.3030569851398468, "learning_rate": 3.0379865924146096e-10, "loss": 0.2005, "step": 18334 }, { "epoch": 4.878924960085151, "grad_norm": 0.28502950072288513, "learning_rate": 3.024712491770387e-10, "loss": 0.185, "step": 18335 }, { "epoch": 4.8791910590739755, "grad_norm": 0.30192750692367554, "learning_rate": 3.0114674104882996e-10, "loss": 0.1945, "step": 18336 }, { "epoch": 4.8794571580628, "grad_norm": 0.28279969096183777, "learning_rate": 2.9982513489535956e-10, "loss": 0.1847, "step": 18337 }, { "epoch": 4.879723257051623, "grad_norm": 0.29796069860458374, "learning_rate": 2.985064307551188e-10, "loss": 0.2008, "step": 18338 }, { "epoch": 4.879989356040447, "grad_norm": 0.29114073514938354, "learning_rate": 2.9719062866647716e-10, "loss": 0.1908, "step": 18339 }, { "epoch": 4.880255455029271, "grad_norm": 0.29610195755958557, "learning_rate": 2.958777286677483e-10, "loss": 0.1878, "step": 18340 }, { "epoch": 4.880521554018094, "grad_norm": 0.29460781812667847, "learning_rate": 2.9456773079713505e-10, "loss": 0.2013, "step": 18341 }, { "epoch": 4.8807876530069185, "grad_norm": 0.29612118005752563, "learning_rate": 2.9326063509276243e-10, "loss": 0.1771, "step": 18342 }, { "epoch": 4.881053751995743, "grad_norm": 0.2924855351448059, "learning_rate": 2.9195644159270003e-10, "loss": 0.1909, "step": 18343 }, { "epoch": 4.881319850984566, "grad_norm": 0.301611065864563, "learning_rate": 2.906551503348953e-10, "loss": 0.1851, "step": 18344 }, { "epoch": 4.88158594997339, "grad_norm": 0.4591231048107147, "learning_rate": 2.8935676135722897e-10, "loss": 0.202, "step": 18345 }, { "epoch": 4.881852048962214, "grad_norm": 0.3286616802215576, "learning_rate": 2.88061274697482e-10, "loss": 0.1916, "step": 18346 }, { "epoch": 4.882118147951038, "grad_norm": 0.29383862018585205, "learning_rate": 2.867686903933797e-10, "loss": 0.177, "step": 18347 }, { "epoch": 4.8823842469398615, "grad_norm": 0.2976371645927429, "learning_rate": 2.8547900848253646e-10, "loss": 0.1984, "step": 18348 }, { "epoch": 4.882650345928686, "grad_norm": 0.2728227376937866, "learning_rate": 2.84192229002489e-10, "loss": 0.1779, "step": 18349 }, { "epoch": 4.882916444917509, "grad_norm": 0.3180117905139923, "learning_rate": 2.8290835199069607e-10, "loss": 0.192, "step": 18350 }, { "epoch": 4.883182543906333, "grad_norm": 0.3387514650821686, "learning_rate": 2.8162737748452794e-10, "loss": 0.1982, "step": 18351 }, { "epoch": 4.883448642895157, "grad_norm": 0.2819625735282898, "learning_rate": 2.803493055212769e-10, "loss": 0.1888, "step": 18352 }, { "epoch": 4.883714741883981, "grad_norm": 0.30586230754852295, "learning_rate": 2.7907413613813545e-10, "loss": 0.1759, "step": 18353 }, { "epoch": 4.8839808408728045, "grad_norm": 0.29758885502815247, "learning_rate": 2.778018693722295e-10, "loss": 0.1839, "step": 18354 }, { "epoch": 4.884246939861629, "grad_norm": 0.2934686839580536, "learning_rate": 2.765325052605738e-10, "loss": 0.1851, "step": 18355 }, { "epoch": 4.884513038850453, "grad_norm": 0.29942357540130615, "learning_rate": 2.7526604384013886e-10, "loss": 0.1688, "step": 18356 }, { "epoch": 4.884779137839276, "grad_norm": 0.34079480171203613, "learning_rate": 2.740024851477618e-10, "loss": 0.2094, "step": 18357 }, { "epoch": 4.8850452368281, "grad_norm": 0.27855831384658813, "learning_rate": 2.727418292202466e-10, "loss": 0.1776, "step": 18358 }, { "epoch": 4.885311335816924, "grad_norm": 0.3091683089733124, "learning_rate": 2.7148407609427494e-10, "loss": 0.1977, "step": 18359 }, { "epoch": 4.8855774348057475, "grad_norm": 0.3045039176940918, "learning_rate": 2.7022922580646203e-10, "loss": 0.1948, "step": 18360 }, { "epoch": 4.885843533794572, "grad_norm": 0.28997325897216797, "learning_rate": 2.6897727839332307e-10, "loss": 0.1966, "step": 18361 }, { "epoch": 4.886109632783396, "grad_norm": 0.2923903465270996, "learning_rate": 2.6772823389131784e-10, "loss": 0.1901, "step": 18362 }, { "epoch": 4.886375731772219, "grad_norm": 0.2928865849971771, "learning_rate": 2.664820923367839e-10, "loss": 0.1854, "step": 18363 }, { "epoch": 4.886641830761043, "grad_norm": 0.2915275990962982, "learning_rate": 2.652388537660033e-10, "loss": 0.1781, "step": 18364 }, { "epoch": 4.886907929749867, "grad_norm": 0.2981119453907013, "learning_rate": 2.6399851821515826e-10, "loss": 0.2065, "step": 18365 }, { "epoch": 4.8871740287386904, "grad_norm": 0.3425242602825165, "learning_rate": 2.627610857203533e-10, "loss": 0.1908, "step": 18366 }, { "epoch": 4.887440127727515, "grad_norm": 0.31371375918388367, "learning_rate": 2.61526556317615e-10, "loss": 0.208, "step": 18367 }, { "epoch": 4.887706226716339, "grad_norm": 0.3148134648799896, "learning_rate": 2.602949300428592e-10, "loss": 0.1895, "step": 18368 }, { "epoch": 4.887972325705162, "grad_norm": 0.3249076306819916, "learning_rate": 2.590662069319571e-10, "loss": 0.1941, "step": 18369 }, { "epoch": 4.888238424693986, "grad_norm": 0.2999076843261719, "learning_rate": 2.57840387020658e-10, "loss": 0.1999, "step": 18370 }, { "epoch": 4.88850452368281, "grad_norm": 0.28667643666267395, "learning_rate": 2.5661747034465554e-10, "loss": 0.19, "step": 18371 }, { "epoch": 4.888770622671634, "grad_norm": 0.2909872531890869, "learning_rate": 2.553974569395323e-10, "loss": 0.1661, "step": 18372 }, { "epoch": 4.8890367216604576, "grad_norm": 0.299763947725296, "learning_rate": 2.5418034684079324e-10, "loss": 0.1831, "step": 18373 }, { "epoch": 4.889302820649282, "grad_norm": 0.3039919435977936, "learning_rate": 2.529661400838878e-10, "loss": 0.1944, "step": 18374 }, { "epoch": 4.889568919638105, "grad_norm": 0.2794129252433777, "learning_rate": 2.517548367041544e-10, "loss": 0.1866, "step": 18375 }, { "epoch": 4.889835018626929, "grad_norm": 0.2728897035121918, "learning_rate": 2.505464367368315e-10, "loss": 0.1792, "step": 18376 }, { "epoch": 4.890101117615753, "grad_norm": 0.3107883334159851, "learning_rate": 2.493409402171132e-10, "loss": 0.1915, "step": 18377 }, { "epoch": 4.890367216604577, "grad_norm": 0.3141060769557953, "learning_rate": 2.481383471800713e-10, "loss": 0.1982, "step": 18378 }, { "epoch": 4.8906333155934005, "grad_norm": 0.27069729566574097, "learning_rate": 2.4693865766072243e-10, "loss": 0.1803, "step": 18379 }, { "epoch": 4.890899414582225, "grad_norm": 0.2896899878978729, "learning_rate": 2.4574187169399406e-10, "loss": 0.1959, "step": 18380 }, { "epoch": 4.891165513571049, "grad_norm": 0.29005298018455505, "learning_rate": 2.4454798931469176e-10, "loss": 0.1882, "step": 18381 }, { "epoch": 4.891431612559872, "grad_norm": 0.31955546140670776, "learning_rate": 2.433570105575988e-10, "loss": 0.1884, "step": 18382 }, { "epoch": 4.891697711548696, "grad_norm": 0.3154554069042206, "learning_rate": 2.421689354573653e-10, "loss": 0.2078, "step": 18383 }, { "epoch": 4.89196381053752, "grad_norm": 0.31551632285118103, "learning_rate": 2.409837640485746e-10, "loss": 0.1962, "step": 18384 }, { "epoch": 4.8922299095263435, "grad_norm": 0.2782782018184662, "learning_rate": 2.3980149636572134e-10, "loss": 0.1893, "step": 18385 }, { "epoch": 4.892496008515168, "grad_norm": 0.29902517795562744, "learning_rate": 2.386221324432225e-10, "loss": 0.1905, "step": 18386 }, { "epoch": 4.892762107503992, "grad_norm": 0.2936541438102722, "learning_rate": 2.374456723154061e-10, "loss": 0.2018, "step": 18387 }, { "epoch": 4.893028206492815, "grad_norm": 0.27842867374420166, "learning_rate": 2.362721160165115e-10, "loss": 0.1738, "step": 18388 }, { "epoch": 4.893294305481639, "grad_norm": 0.30731984972953796, "learning_rate": 2.3510146358071136e-10, "loss": 0.1891, "step": 18389 }, { "epoch": 4.893560404470463, "grad_norm": 0.3288733959197998, "learning_rate": 2.3393371504206725e-10, "loss": 0.2291, "step": 18390 }, { "epoch": 4.8938265034592865, "grad_norm": 0.31356266140937805, "learning_rate": 2.3276887043457427e-10, "loss": 0.194, "step": 18391 }, { "epoch": 4.894092602448111, "grad_norm": 0.28531721234321594, "learning_rate": 2.3160692979212748e-10, "loss": 0.1673, "step": 18392 }, { "epoch": 4.894358701436935, "grad_norm": 0.2964000403881073, "learning_rate": 2.304478931485665e-10, "loss": 0.1613, "step": 18393 }, { "epoch": 4.894624800425758, "grad_norm": 0.27924075722694397, "learning_rate": 2.292917605376088e-10, "loss": 0.1856, "step": 18394 }, { "epoch": 4.894890899414582, "grad_norm": 0.3366377353668213, "learning_rate": 2.2813853199292743e-10, "loss": 0.1895, "step": 18395 }, { "epoch": 4.895156998403406, "grad_norm": 0.30516645312309265, "learning_rate": 2.2698820754807334e-10, "loss": 0.2062, "step": 18396 }, { "epoch": 4.89542309739223, "grad_norm": 0.32408833503723145, "learning_rate": 2.258407872365309e-10, "loss": 0.1817, "step": 18397 }, { "epoch": 4.895689196381054, "grad_norm": 0.30184206366539, "learning_rate": 2.2469627109170663e-10, "loss": 0.1805, "step": 18398 }, { "epoch": 4.895955295369878, "grad_norm": 0.2880866229534149, "learning_rate": 2.235546591469184e-10, "loss": 0.1733, "step": 18399 }, { "epoch": 4.896221394358701, "grad_norm": 0.2779730558395386, "learning_rate": 2.2241595143537296e-10, "loss": 0.1626, "step": 18400 }, { "epoch": 4.896487493347525, "grad_norm": 0.30361807346343994, "learning_rate": 2.2128014799023264e-10, "loss": 0.2067, "step": 18401 }, { "epoch": 4.896753592336349, "grad_norm": 0.2860822379589081, "learning_rate": 2.2014724884455993e-10, "loss": 0.1837, "step": 18402 }, { "epoch": 4.897019691325173, "grad_norm": 0.3057304918766022, "learning_rate": 2.1901725403131733e-10, "loss": 0.2084, "step": 18403 }, { "epoch": 4.897285790313997, "grad_norm": 0.29055243730545044, "learning_rate": 2.1789016358340072e-10, "loss": 0.1893, "step": 18404 }, { "epoch": 4.897551889302821, "grad_norm": 0.29345259070396423, "learning_rate": 2.1676597753362834e-10, "loss": 0.1923, "step": 18405 }, { "epoch": 4.897817988291645, "grad_norm": 0.27397048473358154, "learning_rate": 2.1564469591470735e-10, "loss": 0.1804, "step": 18406 }, { "epoch": 4.898084087280468, "grad_norm": 0.27694839239120483, "learning_rate": 2.145263187592894e-10, "loss": 0.187, "step": 18407 }, { "epoch": 4.898350186269292, "grad_norm": 0.2930940091609955, "learning_rate": 2.1341084609991512e-10, "loss": 0.1816, "step": 18408 }, { "epoch": 4.898616285258116, "grad_norm": 0.3077024221420288, "learning_rate": 2.1229827796905852e-10, "loss": 0.1991, "step": 18409 }, { "epoch": 4.89888238424694, "grad_norm": 0.29363659024238586, "learning_rate": 2.1118861439910486e-10, "loss": 0.1897, "step": 18410 }, { "epoch": 4.899148483235764, "grad_norm": 0.26168808341026306, "learning_rate": 2.1008185542235046e-10, "loss": 0.1729, "step": 18411 }, { "epoch": 4.899414582224588, "grad_norm": 0.41078799962997437, "learning_rate": 2.0897800107101404e-10, "loss": 0.2096, "step": 18412 }, { "epoch": 4.899680681213411, "grad_norm": 0.30742886662483215, "learning_rate": 2.0787705137721434e-10, "loss": 0.1849, "step": 18413 }, { "epoch": 4.899946780202235, "grad_norm": 0.2908216714859009, "learning_rate": 2.067790063730146e-10, "loss": 0.1913, "step": 18414 }, { "epoch": 4.900212879191059, "grad_norm": 0.29317763447761536, "learning_rate": 2.0568386609036703e-10, "loss": 0.172, "step": 18415 }, { "epoch": 4.9004789781798825, "grad_norm": 0.305318683385849, "learning_rate": 2.0459163056114616e-10, "loss": 0.1987, "step": 18416 }, { "epoch": 4.900745077168707, "grad_norm": 0.2878947854042053, "learning_rate": 2.0350229981714873e-10, "loss": 0.1925, "step": 18417 }, { "epoch": 4.901011176157531, "grad_norm": 0.30962806940078735, "learning_rate": 2.0241587389008273e-10, "loss": 0.2046, "step": 18418 }, { "epoch": 4.901277275146354, "grad_norm": 0.30945199728012085, "learning_rate": 2.0133235281156735e-10, "loss": 0.2013, "step": 18419 }, { "epoch": 4.901543374135178, "grad_norm": 0.30217245221138, "learning_rate": 2.0025173661313287e-10, "loss": 0.1958, "step": 18420 }, { "epoch": 4.901809473124002, "grad_norm": 0.3177815079689026, "learning_rate": 1.9917402532625416e-10, "loss": 0.1939, "step": 18421 }, { "epoch": 4.902075572112826, "grad_norm": 0.3016432523727417, "learning_rate": 1.9809921898228387e-10, "loss": 0.1933, "step": 18422 }, { "epoch": 4.90234167110165, "grad_norm": 0.3304864466190338, "learning_rate": 1.9702731761250813e-10, "loss": 0.2042, "step": 18423 }, { "epoch": 4.902607770090474, "grad_norm": 0.2952777147293091, "learning_rate": 1.9595832124814637e-10, "loss": 0.1805, "step": 18424 }, { "epoch": 4.902873869079297, "grad_norm": 0.2841630280017853, "learning_rate": 1.9489222992028486e-10, "loss": 0.1878, "step": 18425 }, { "epoch": 4.903139968068121, "grad_norm": 0.29325997829437256, "learning_rate": 1.9382904365997653e-10, "loss": 0.1933, "step": 18426 }, { "epoch": 4.903406067056945, "grad_norm": 0.3063047230243683, "learning_rate": 1.9276876249815222e-10, "loss": 0.1934, "step": 18427 }, { "epoch": 4.903672166045769, "grad_norm": 0.3186180889606476, "learning_rate": 1.9171138646569828e-10, "loss": 0.1959, "step": 18428 }, { "epoch": 4.903938265034593, "grad_norm": 0.28000184893608093, "learning_rate": 1.9065691559335684e-10, "loss": 0.1756, "step": 18429 }, { "epoch": 4.904204364023417, "grad_norm": 0.30174344778060913, "learning_rate": 1.8960534991185884e-10, "loss": 0.1878, "step": 18430 }, { "epoch": 4.904470463012241, "grad_norm": 0.2999070882797241, "learning_rate": 1.8855668945177983e-10, "loss": 0.1789, "step": 18431 }, { "epoch": 4.904736562001064, "grad_norm": 0.29529130458831787, "learning_rate": 1.8751093424366204e-10, "loss": 0.1816, "step": 18432 }, { "epoch": 4.905002660989888, "grad_norm": 0.2870076894760132, "learning_rate": 1.864680843179367e-10, "loss": 0.1784, "step": 18433 }, { "epoch": 4.905268759978712, "grad_norm": 0.29535576701164246, "learning_rate": 1.854281397049684e-10, "loss": 0.1863, "step": 18434 }, { "epoch": 4.905534858967536, "grad_norm": 0.3098396956920624, "learning_rate": 1.8439110043502182e-10, "loss": 0.1996, "step": 18435 }, { "epoch": 4.90580095795636, "grad_norm": 0.3172910213470459, "learning_rate": 1.8335696653828393e-10, "loss": 0.2009, "step": 18436 }, { "epoch": 4.906067056945184, "grad_norm": 0.2864147126674652, "learning_rate": 1.8232573804485285e-10, "loss": 0.1894, "step": 18437 }, { "epoch": 4.906333155934007, "grad_norm": 0.30300426483154297, "learning_rate": 1.8129741498473795e-10, "loss": 0.1965, "step": 18438 }, { "epoch": 4.906599254922831, "grad_norm": 0.2970436215400696, "learning_rate": 1.802719973878819e-10, "loss": 0.1981, "step": 18439 }, { "epoch": 4.906865353911655, "grad_norm": 0.3098726272583008, "learning_rate": 1.7924948528412753e-10, "loss": 0.196, "step": 18440 }, { "epoch": 4.9071314529004795, "grad_norm": 0.2918415367603302, "learning_rate": 1.782298787032399e-10, "loss": 0.191, "step": 18441 }, { "epoch": 4.907397551889303, "grad_norm": 0.2838175892829895, "learning_rate": 1.7721317767489529e-10, "loss": 0.1973, "step": 18442 }, { "epoch": 4.907663650878127, "grad_norm": 0.2965428829193115, "learning_rate": 1.7619938222869224e-10, "loss": 0.1918, "step": 18443 }, { "epoch": 4.90792974986695, "grad_norm": 0.2853854298591614, "learning_rate": 1.7518849239412936e-10, "loss": 0.2001, "step": 18444 }, { "epoch": 4.908195848855774, "grad_norm": 0.3030586540699005, "learning_rate": 1.7418050820063868e-10, "loss": 0.1999, "step": 18445 }, { "epoch": 4.908461947844598, "grad_norm": 0.29188022017478943, "learning_rate": 1.7317542967756337e-10, "loss": 0.1648, "step": 18446 }, { "epoch": 4.9087280468334225, "grad_norm": 0.3050466775894165, "learning_rate": 1.7217325685414673e-10, "loss": 0.2062, "step": 18447 }, { "epoch": 4.908994145822246, "grad_norm": 0.28953883051872253, "learning_rate": 1.7117398975956543e-10, "loss": 0.178, "step": 18448 }, { "epoch": 4.90926024481107, "grad_norm": 0.299944132566452, "learning_rate": 1.7017762842290728e-10, "loss": 0.199, "step": 18449 }, { "epoch": 4.909526343799894, "grad_norm": 0.29586148262023926, "learning_rate": 1.6918417287318244e-10, "loss": 0.1759, "step": 18450 }, { "epoch": 4.909792442788717, "grad_norm": 0.3032758831977844, "learning_rate": 1.6819362313928997e-10, "loss": 0.1952, "step": 18451 }, { "epoch": 4.910058541777541, "grad_norm": 0.32104283571243286, "learning_rate": 1.672059792500624e-10, "loss": 0.179, "step": 18452 }, { "epoch": 4.910324640766365, "grad_norm": 0.31085795164108276, "learning_rate": 1.662212412342767e-10, "loss": 0.1952, "step": 18453 }, { "epoch": 4.910590739755189, "grad_norm": 0.3007752299308777, "learning_rate": 1.6523940912055444e-10, "loss": 0.1927, "step": 18454 }, { "epoch": 4.910856838744013, "grad_norm": 0.29007163643836975, "learning_rate": 1.6426048293750605e-10, "loss": 0.1788, "step": 18455 }, { "epoch": 4.911122937732837, "grad_norm": 0.2980814576148987, "learning_rate": 1.6328446271360875e-10, "loss": 0.1917, "step": 18456 }, { "epoch": 4.91138903672166, "grad_norm": 0.28053799271583557, "learning_rate": 1.6231134847728423e-10, "loss": 0.1719, "step": 18457 }, { "epoch": 4.911655135710484, "grad_norm": 0.29136428236961365, "learning_rate": 1.613411402568432e-10, "loss": 0.189, "step": 18458 }, { "epoch": 4.911921234699308, "grad_norm": 0.29290950298309326, "learning_rate": 1.6037383808054083e-10, "loss": 0.2053, "step": 18459 }, { "epoch": 4.912187333688132, "grad_norm": 0.31133124232292175, "learning_rate": 1.5940944197651019e-10, "loss": 0.1983, "step": 18460 }, { "epoch": 4.912453432676956, "grad_norm": 0.3186502754688263, "learning_rate": 1.58447951972851e-10, "loss": 0.2098, "step": 18461 }, { "epoch": 4.91271953166578, "grad_norm": 0.3413541913032532, "learning_rate": 1.5748936809752978e-10, "loss": 0.1971, "step": 18462 }, { "epoch": 4.912985630654603, "grad_norm": 0.30598020553588867, "learning_rate": 1.5653369037845754e-10, "loss": 0.1884, "step": 18463 }, { "epoch": 4.913251729643427, "grad_norm": 0.31565791368484497, "learning_rate": 1.5558091884344538e-10, "loss": 0.1861, "step": 18464 }, { "epoch": 4.913517828632251, "grad_norm": 0.27259254455566406, "learning_rate": 1.5463105352023774e-10, "loss": 0.1564, "step": 18465 }, { "epoch": 4.9137839276210755, "grad_norm": 0.31286609172821045, "learning_rate": 1.5368409443646814e-10, "loss": 0.1999, "step": 18466 }, { "epoch": 4.914050026609899, "grad_norm": 0.3124084174633026, "learning_rate": 1.5274004161970332e-10, "loss": 0.202, "step": 18467 }, { "epoch": 4.914316125598723, "grad_norm": 0.2903044819831848, "learning_rate": 1.517988950974325e-10, "loss": 0.1912, "step": 18468 }, { "epoch": 4.914582224587546, "grad_norm": 0.26452359557151794, "learning_rate": 1.508606548970337e-10, "loss": 0.1843, "step": 18469 }, { "epoch": 4.91484832357637, "grad_norm": 0.3007436990737915, "learning_rate": 1.4992532104582955e-10, "loss": 0.2001, "step": 18470 }, { "epoch": 4.915114422565194, "grad_norm": 0.2910817861557007, "learning_rate": 1.4899289357104272e-10, "loss": 0.185, "step": 18471 }, { "epoch": 4.9153805215540185, "grad_norm": 0.2760818600654602, "learning_rate": 1.4806337249980704e-10, "loss": 0.1823, "step": 18472 }, { "epoch": 4.915646620542842, "grad_norm": 0.28262078762054443, "learning_rate": 1.4713675785918977e-10, "loss": 0.1786, "step": 18473 }, { "epoch": 4.915912719531666, "grad_norm": 0.28444167971611023, "learning_rate": 1.4621304967615822e-10, "loss": 0.1826, "step": 18474 }, { "epoch": 4.91617881852049, "grad_norm": 0.3213438093662262, "learning_rate": 1.4529224797759088e-10, "loss": 0.2079, "step": 18475 }, { "epoch": 4.916444917509313, "grad_norm": 0.30850011110305786, "learning_rate": 1.4437435279029965e-10, "loss": 0.1951, "step": 18476 }, { "epoch": 4.916711016498137, "grad_norm": 0.2784685492515564, "learning_rate": 1.4345936414099646e-10, "loss": 0.1798, "step": 18477 }, { "epoch": 4.9169771154869615, "grad_norm": 0.30688101053237915, "learning_rate": 1.425472820563156e-10, "loss": 0.1867, "step": 18478 }, { "epoch": 4.917243214475785, "grad_norm": 0.302877277135849, "learning_rate": 1.4163810656280252e-10, "loss": 0.1837, "step": 18479 }, { "epoch": 4.917509313464609, "grad_norm": 0.3040710985660553, "learning_rate": 1.4073183768693597e-10, "loss": 0.1892, "step": 18480 }, { "epoch": 4.917775412453433, "grad_norm": 0.2751975357532501, "learning_rate": 1.398284754550727e-10, "loss": 0.1693, "step": 18481 }, { "epoch": 4.918041511442256, "grad_norm": 0.32650625705718994, "learning_rate": 1.3892801989352498e-10, "loss": 0.1826, "step": 18482 }, { "epoch": 4.91830761043108, "grad_norm": 0.3202764093875885, "learning_rate": 1.3803047102849408e-10, "loss": 0.1857, "step": 18483 }, { "epoch": 4.9185737094199045, "grad_norm": 0.2917061150074005, "learning_rate": 1.3713582888610353e-10, "loss": 0.1919, "step": 18484 }, { "epoch": 4.918839808408728, "grad_norm": 0.31070148944854736, "learning_rate": 1.3624409349239918e-10, "loss": 0.202, "step": 18485 }, { "epoch": 4.919105907397552, "grad_norm": 0.31637313961982727, "learning_rate": 1.3535526487333803e-10, "loss": 0.1949, "step": 18486 }, { "epoch": 4.919372006386376, "grad_norm": 0.3008381724357605, "learning_rate": 1.344693430547883e-10, "loss": 0.1879, "step": 18487 }, { "epoch": 4.919638105375199, "grad_norm": 0.30968570709228516, "learning_rate": 1.3358632806254044e-10, "loss": 0.1846, "step": 18488 }, { "epoch": 4.919904204364023, "grad_norm": 0.3088178336620331, "learning_rate": 1.3270621992228503e-10, "loss": 0.178, "step": 18489 }, { "epoch": 4.920170303352847, "grad_norm": 0.29426077008247375, "learning_rate": 1.3182901865965711e-10, "loss": 0.2037, "step": 18490 }, { "epoch": 4.920436402341672, "grad_norm": 0.3002553880214691, "learning_rate": 1.3095472430018073e-10, "loss": 0.1936, "step": 18491 }, { "epoch": 4.920702501330495, "grad_norm": 0.2848266661167145, "learning_rate": 1.3008333686930217e-10, "loss": 0.1857, "step": 18492 }, { "epoch": 4.920968600319319, "grad_norm": 0.30008426308631897, "learning_rate": 1.2921485639239005e-10, "loss": 0.1975, "step": 18493 }, { "epoch": 4.921234699308142, "grad_norm": 0.2884708344936371, "learning_rate": 1.2834928289472414e-10, "loss": 0.1839, "step": 18494 }, { "epoch": 4.921500798296966, "grad_norm": 0.3016155958175659, "learning_rate": 1.274866164014843e-10, "loss": 0.1923, "step": 18495 }, { "epoch": 4.92176689728579, "grad_norm": 0.3064568340778351, "learning_rate": 1.2662685693780594e-10, "loss": 0.2099, "step": 18496 }, { "epoch": 4.9220329962746145, "grad_norm": 0.3121053874492645, "learning_rate": 1.2577000452869135e-10, "loss": 0.1858, "step": 18497 }, { "epoch": 4.922299095263438, "grad_norm": 0.28787413239479065, "learning_rate": 1.2491605919909831e-10, "loss": 0.1778, "step": 18498 }, { "epoch": 4.922565194252262, "grad_norm": 0.287299782037735, "learning_rate": 1.240650209738736e-10, "loss": 0.1771, "step": 18499 }, { "epoch": 4.922831293241086, "grad_norm": 0.3266220986843109, "learning_rate": 1.2321688987779744e-10, "loss": 0.205, "step": 18500 }, { "epoch": 4.923097392229909, "grad_norm": 0.3084475100040436, "learning_rate": 1.2237166593553893e-10, "loss": 0.1941, "step": 18501 }, { "epoch": 4.923363491218733, "grad_norm": 0.30537694692611694, "learning_rate": 1.2152934917172285e-10, "loss": 0.1818, "step": 18502 }, { "epoch": 4.9236295902075575, "grad_norm": 0.29805988073349, "learning_rate": 1.2068993961084072e-10, "loss": 0.1982, "step": 18503 }, { "epoch": 4.923895689196381, "grad_norm": 0.2989754378795624, "learning_rate": 1.1985343727735076e-10, "loss": 0.1744, "step": 18504 }, { "epoch": 4.924161788185205, "grad_norm": 0.3045249283313751, "learning_rate": 1.1901984219558902e-10, "loss": 0.1796, "step": 18505 }, { "epoch": 4.924427887174029, "grad_norm": 0.2791639268398285, "learning_rate": 1.1818915438982502e-10, "loss": 0.1897, "step": 18506 }, { "epoch": 4.924693986162852, "grad_norm": 0.2843552827835083, "learning_rate": 1.173613738842283e-10, "loss": 0.1792, "step": 18507 }, { "epoch": 4.924960085151676, "grad_norm": 0.28697219491004944, "learning_rate": 1.1653650070290178e-10, "loss": 0.1727, "step": 18508 }, { "epoch": 4.9252261841405005, "grad_norm": 0.31288108229637146, "learning_rate": 1.157145348698596e-10, "loss": 0.2084, "step": 18509 }, { "epoch": 4.925492283129324, "grad_norm": 0.3068991005420685, "learning_rate": 1.1489547640900488e-10, "loss": 0.1814, "step": 18510 }, { "epoch": 4.925758382118148, "grad_norm": 0.2979295551776886, "learning_rate": 1.1407932534420739e-10, "loss": 0.1736, "step": 18511 }, { "epoch": 4.926024481106972, "grad_norm": 0.30303171277046204, "learning_rate": 1.132660816992037e-10, "loss": 0.1987, "step": 18512 }, { "epoch": 4.926290580095795, "grad_norm": 0.32095956802368164, "learning_rate": 1.124557454976749e-10, "loss": 0.1808, "step": 18513 }, { "epoch": 4.926556679084619, "grad_norm": 0.3030645251274109, "learning_rate": 1.1164831676319097e-10, "loss": 0.2068, "step": 18514 }, { "epoch": 4.9268227780734435, "grad_norm": 0.30267754197120667, "learning_rate": 1.1084379551928869e-10, "loss": 0.2068, "step": 18515 }, { "epoch": 4.927088877062268, "grad_norm": 0.29511070251464844, "learning_rate": 1.1004218178934931e-10, "loss": 0.195, "step": 18516 }, { "epoch": 4.927354976051091, "grad_norm": 0.31029605865478516, "learning_rate": 1.0924347559672087e-10, "loss": 0.2023, "step": 18517 }, { "epoch": 4.927621075039915, "grad_norm": 0.29313281178474426, "learning_rate": 1.084476769646514e-10, "loss": 0.1992, "step": 18518 }, { "epoch": 4.927887174028738, "grad_norm": 0.2826337516307831, "learning_rate": 1.0765478591631128e-10, "loss": 0.1884, "step": 18519 }, { "epoch": 4.928153273017562, "grad_norm": 0.30609312653541565, "learning_rate": 1.0686480247477092e-10, "loss": 0.1973, "step": 18520 }, { "epoch": 4.9284193720063865, "grad_norm": 0.2969426214694977, "learning_rate": 1.0607772666302306e-10, "loss": 0.1805, "step": 18521 }, { "epoch": 4.928685470995211, "grad_norm": 0.3053213059902191, "learning_rate": 1.0529355850397159e-10, "loss": 0.1952, "step": 18522 }, { "epoch": 4.928951569984034, "grad_norm": 0.29173943400382996, "learning_rate": 1.045122980204649e-10, "loss": 0.1892, "step": 18523 }, { "epoch": 4.929217668972858, "grad_norm": 0.33531454205513, "learning_rate": 1.0373394523521817e-10, "loss": 0.1722, "step": 18524 }, { "epoch": 4.929483767961682, "grad_norm": 0.31951242685317993, "learning_rate": 1.0295850017091323e-10, "loss": 0.2039, "step": 18525 }, { "epoch": 4.929749866950505, "grad_norm": 0.37189045548439026, "learning_rate": 1.0218596285009873e-10, "loss": 0.1891, "step": 18526 }, { "epoch": 4.9300159659393294, "grad_norm": 0.30492520332336426, "learning_rate": 1.0141633329525667e-10, "loss": 0.1943, "step": 18527 }, { "epoch": 4.930282064928154, "grad_norm": 0.3653099238872528, "learning_rate": 1.0064961152881358e-10, "loss": 0.1853, "step": 18528 }, { "epoch": 4.930548163916977, "grad_norm": 0.30294641852378845, "learning_rate": 9.98857975730738e-11, "loss": 0.1917, "step": 18529 }, { "epoch": 4.930814262905801, "grad_norm": 0.270704060792923, "learning_rate": 9.912489145026403e-11, "loss": 0.1704, "step": 18530 }, { "epoch": 4.931080361894625, "grad_norm": 0.32555341720581055, "learning_rate": 9.836689318254431e-11, "loss": 0.1932, "step": 18531 }, { "epoch": 4.931346460883448, "grad_norm": 0.33392655849456787, "learning_rate": 9.761180279197479e-11, "loss": 0.2133, "step": 18532 }, { "epoch": 4.931612559872272, "grad_norm": 0.3026379942893982, "learning_rate": 9.685962030052674e-11, "loss": 0.2009, "step": 18533 }, { "epoch": 4.9318786588610966, "grad_norm": 0.27075210213661194, "learning_rate": 9.61103457301049e-11, "loss": 0.1704, "step": 18534 }, { "epoch": 4.93214475784992, "grad_norm": 0.3002595603466034, "learning_rate": 9.536397910250294e-11, "loss": 0.1925, "step": 18535 }, { "epoch": 4.932410856838744, "grad_norm": 0.29769954085350037, "learning_rate": 9.462052043947011e-11, "loss": 0.1923, "step": 18536 }, { "epoch": 4.932676955827568, "grad_norm": 0.2881838083267212, "learning_rate": 9.387996976262247e-11, "loss": 0.1839, "step": 18537 }, { "epoch": 4.932943054816391, "grad_norm": 0.2984301745891571, "learning_rate": 9.314232709353164e-11, "loss": 0.2003, "step": 18538 }, { "epoch": 4.933209153805215, "grad_norm": 0.2980403006076813, "learning_rate": 9.240759245364715e-11, "loss": 0.2039, "step": 18539 }, { "epoch": 4.9334752527940395, "grad_norm": 0.30141526460647583, "learning_rate": 9.167576586438519e-11, "loss": 0.205, "step": 18540 }, { "epoch": 4.933741351782864, "grad_norm": 0.28958019614219666, "learning_rate": 9.094684734702873e-11, "loss": 0.2012, "step": 18541 }, { "epoch": 4.934007450771687, "grad_norm": 0.29781532287597656, "learning_rate": 9.022083692279414e-11, "loss": 0.1802, "step": 18542 }, { "epoch": 4.934273549760511, "grad_norm": 0.3025088310241699, "learning_rate": 8.949773461282006e-11, "loss": 0.1916, "step": 18543 }, { "epoch": 4.934539648749334, "grad_norm": 0.3099420964717865, "learning_rate": 8.877754043815633e-11, "loss": 0.1946, "step": 18544 }, { "epoch": 4.934805747738158, "grad_norm": 0.3043050467967987, "learning_rate": 8.806025441975285e-11, "loss": 0.1845, "step": 18545 }, { "epoch": 4.9350718467269825, "grad_norm": 0.2860504984855652, "learning_rate": 8.734587657850401e-11, "loss": 0.1725, "step": 18546 }, { "epoch": 4.935337945715807, "grad_norm": 0.28987058997154236, "learning_rate": 8.66344069351932e-11, "loss": 0.1863, "step": 18547 }, { "epoch": 4.93560404470463, "grad_norm": 0.28680935502052307, "learning_rate": 8.592584551053716e-11, "loss": 0.17, "step": 18548 }, { "epoch": 4.935870143693454, "grad_norm": 0.2768939733505249, "learning_rate": 8.522019232515277e-11, "loss": 0.1861, "step": 18549 }, { "epoch": 4.936136242682278, "grad_norm": 0.30849653482437134, "learning_rate": 8.451744739957911e-11, "loss": 0.2137, "step": 18550 }, { "epoch": 4.936402341671101, "grad_norm": 0.2876201868057251, "learning_rate": 8.381761075427762e-11, "loss": 0.1618, "step": 18551 }, { "epoch": 4.9366684406599255, "grad_norm": 0.29006749391555786, "learning_rate": 8.312068240960978e-11, "loss": 0.1856, "step": 18552 }, { "epoch": 4.93693453964875, "grad_norm": 0.3455030024051666, "learning_rate": 8.242666238587048e-11, "loss": 0.2035, "step": 18553 }, { "epoch": 4.937200638637573, "grad_norm": 0.3017522692680359, "learning_rate": 8.173555070326576e-11, "loss": 0.1883, "step": 18554 }, { "epoch": 4.937466737626397, "grad_norm": 0.29162198305130005, "learning_rate": 8.104734738189067e-11, "loss": 0.1833, "step": 18555 }, { "epoch": 4.937732836615221, "grad_norm": 0.312807559967041, "learning_rate": 8.036205244179584e-11, "loss": 0.1994, "step": 18556 }, { "epoch": 4.937998935604044, "grad_norm": 0.318136990070343, "learning_rate": 7.967966590293196e-11, "loss": 0.2, "step": 18557 }, { "epoch": 4.9382650345928685, "grad_norm": 0.3144402801990509, "learning_rate": 7.900018778513873e-11, "loss": 0.1778, "step": 18558 }, { "epoch": 4.938531133581693, "grad_norm": 0.3090205192565918, "learning_rate": 7.832361810821142e-11, "loss": 0.2032, "step": 18559 }, { "epoch": 4.938797232570517, "grad_norm": 0.27467525005340576, "learning_rate": 7.764995689184539e-11, "loss": 0.1874, "step": 18560 }, { "epoch": 4.93906333155934, "grad_norm": 0.3345978558063507, "learning_rate": 7.697920415564718e-11, "loss": 0.1968, "step": 18561 }, { "epoch": 4.939329430548164, "grad_norm": 0.2798305153846741, "learning_rate": 7.631135991912341e-11, "loss": 0.1666, "step": 18562 }, { "epoch": 4.939595529536987, "grad_norm": 0.3074668049812317, "learning_rate": 7.564642420173628e-11, "loss": 0.1971, "step": 18563 }, { "epoch": 4.9398616285258115, "grad_norm": 0.3053053021430969, "learning_rate": 7.498439702282589e-11, "loss": 0.1827, "step": 18564 }, { "epoch": 4.940127727514636, "grad_norm": 0.2951902747154236, "learning_rate": 7.432527840166569e-11, "loss": 0.1739, "step": 18565 }, { "epoch": 4.94039382650346, "grad_norm": 0.28618180751800537, "learning_rate": 7.366906835745145e-11, "loss": 0.1852, "step": 18566 }, { "epoch": 4.940659925492283, "grad_norm": 0.2883571684360504, "learning_rate": 7.301576690925682e-11, "loss": 0.1766, "step": 18567 }, { "epoch": 4.940926024481107, "grad_norm": 0.3333991467952728, "learning_rate": 7.23653740761332e-11, "loss": 0.1979, "step": 18568 }, { "epoch": 4.941192123469931, "grad_norm": 0.31195804476737976, "learning_rate": 7.171788987698768e-11, "loss": 0.1853, "step": 18569 }, { "epoch": 4.941458222458754, "grad_norm": 0.30976006388664246, "learning_rate": 7.107331433067188e-11, "loss": 0.1861, "step": 18570 }, { "epoch": 4.941724321447579, "grad_norm": 0.28904393315315247, "learning_rate": 7.043164745594854e-11, "loss": 0.1783, "step": 18571 }, { "epoch": 4.941990420436403, "grad_norm": 0.3067225217819214, "learning_rate": 6.979288927150273e-11, "loss": 0.1835, "step": 18572 }, { "epoch": 4.942256519425226, "grad_norm": 0.30634793639183044, "learning_rate": 6.915703979590848e-11, "loss": 0.2042, "step": 18573 }, { "epoch": 4.94252261841405, "grad_norm": 0.28919517993927, "learning_rate": 6.852409904769541e-11, "loss": 0.1827, "step": 18574 }, { "epoch": 4.942788717402874, "grad_norm": 0.2944565415382385, "learning_rate": 6.789406704527101e-11, "loss": 0.1863, "step": 18575 }, { "epoch": 4.943054816391697, "grad_norm": 0.2829148471355438, "learning_rate": 6.726694380697617e-11, "loss": 0.1902, "step": 18576 }, { "epoch": 4.9433209153805215, "grad_norm": 0.27773457765579224, "learning_rate": 6.664272935107407e-11, "loss": 0.1757, "step": 18577 }, { "epoch": 4.943587014369346, "grad_norm": 0.2876487672328949, "learning_rate": 6.602142369572793e-11, "loss": 0.1791, "step": 18578 }, { "epoch": 4.943853113358169, "grad_norm": 0.2958492338657379, "learning_rate": 6.540302685901222e-11, "loss": 0.1982, "step": 18579 }, { "epoch": 4.944119212346993, "grad_norm": 0.31144487857818604, "learning_rate": 6.478753885894584e-11, "loss": 0.1814, "step": 18580 }, { "epoch": 4.944385311335817, "grad_norm": 0.33422422409057617, "learning_rate": 6.417495971343667e-11, "loss": 0.1969, "step": 18581 }, { "epoch": 4.94465141032464, "grad_norm": 0.2779046893119812, "learning_rate": 6.356528944030381e-11, "loss": 0.1641, "step": 18582 }, { "epoch": 4.9449175093134645, "grad_norm": 0.3080863356590271, "learning_rate": 6.295852805731083e-11, "loss": 0.1941, "step": 18583 }, { "epoch": 4.945183608302289, "grad_norm": 0.2825024127960205, "learning_rate": 6.235467558211028e-11, "loss": 0.1847, "step": 18584 }, { "epoch": 4.945449707291113, "grad_norm": 0.3065972328186035, "learning_rate": 6.175373203227696e-11, "loss": 0.2036, "step": 18585 }, { "epoch": 4.945715806279936, "grad_norm": 0.2902739644050598, "learning_rate": 6.115569742530802e-11, "loss": 0.1952, "step": 18586 }, { "epoch": 4.94598190526876, "grad_norm": 0.2814626097679138, "learning_rate": 6.056057177861173e-11, "loss": 0.1885, "step": 18587 }, { "epoch": 4.946248004257583, "grad_norm": 0.3156077563762665, "learning_rate": 5.99683551095076e-11, "loss": 0.1997, "step": 18588 }, { "epoch": 4.9465141032464075, "grad_norm": 0.29302477836608887, "learning_rate": 5.937904743522625e-11, "loss": 0.1731, "step": 18589 }, { "epoch": 4.946780202235232, "grad_norm": 0.3297177255153656, "learning_rate": 5.879264877293177e-11, "loss": 0.21, "step": 18590 }, { "epoch": 4.947046301224056, "grad_norm": 0.2826448976993561, "learning_rate": 5.820915913967716e-11, "loss": 0.182, "step": 18591 }, { "epoch": 4.947312400212879, "grad_norm": 0.29484817385673523, "learning_rate": 5.7628578552471054e-11, "loss": 0.2073, "step": 18592 }, { "epoch": 4.947578499201703, "grad_norm": 0.31044164299964905, "learning_rate": 5.7050907028199924e-11, "loss": 0.1818, "step": 18593 }, { "epoch": 4.947844598190527, "grad_norm": 0.29884257912635803, "learning_rate": 5.647614458367256e-11, "loss": 0.2053, "step": 18594 }, { "epoch": 4.9481106971793505, "grad_norm": 0.32388728857040405, "learning_rate": 5.590429123562002e-11, "loss": 0.1798, "step": 18595 }, { "epoch": 4.948376796168175, "grad_norm": 0.2916512191295624, "learning_rate": 5.533534700069564e-11, "loss": 0.1987, "step": 18596 }, { "epoch": 4.948642895156999, "grad_norm": 0.31213754415512085, "learning_rate": 5.476931189545286e-11, "loss": 0.1999, "step": 18597 }, { "epoch": 4.948908994145822, "grad_norm": 0.33047202229499817, "learning_rate": 5.420618593637849e-11, "loss": 0.2027, "step": 18598 }, { "epoch": 4.949175093134646, "grad_norm": 0.2912018299102783, "learning_rate": 5.3645969139848314e-11, "loss": 0.1826, "step": 18599 }, { "epoch": 4.94944119212347, "grad_norm": 0.3919611871242523, "learning_rate": 5.3088661522182606e-11, "loss": 0.1791, "step": 18600 }, { "epoch": 4.9497072911122935, "grad_norm": 0.2956198453903198, "learning_rate": 5.2534263099590637e-11, "loss": 0.2039, "step": 18601 }, { "epoch": 4.949973390101118, "grad_norm": 0.2941761016845703, "learning_rate": 5.1982773888215035e-11, "loss": 0.1742, "step": 18602 }, { "epoch": 4.950239489089942, "grad_norm": 0.3141532242298126, "learning_rate": 5.143419390410964e-11, "loss": 0.1822, "step": 18603 }, { "epoch": 4.950505588078765, "grad_norm": 0.3207624554634094, "learning_rate": 5.088852316323944e-11, "loss": 0.2062, "step": 18604 }, { "epoch": 4.950771687067589, "grad_norm": 0.32369038462638855, "learning_rate": 5.0345761681491736e-11, "loss": 0.2153, "step": 18605 }, { "epoch": 4.951037786056413, "grad_norm": 0.3149227201938629, "learning_rate": 4.980590947465391e-11, "loss": 0.1984, "step": 18606 }, { "epoch": 4.951303885045236, "grad_norm": 0.2803778648376465, "learning_rate": 4.92689665584578e-11, "loss": 0.1918, "step": 18607 }, { "epoch": 4.951569984034061, "grad_norm": 0.27320000529289246, "learning_rate": 4.8734932948524266e-11, "loss": 0.168, "step": 18608 }, { "epoch": 4.951836083022885, "grad_norm": 0.32121819257736206, "learning_rate": 4.8203808660385316e-11, "loss": 0.2206, "step": 18609 }, { "epoch": 4.952102182011709, "grad_norm": 0.29617226123809814, "learning_rate": 4.767559370951746e-11, "loss": 0.1834, "step": 18610 }, { "epoch": 4.952368281000532, "grad_norm": 0.30976182222366333, "learning_rate": 4.715028811128618e-11, "loss": 0.2025, "step": 18611 }, { "epoch": 4.952634379989356, "grad_norm": 0.325028657913208, "learning_rate": 4.662789188099037e-11, "loss": 0.2023, "step": 18612 }, { "epoch": 4.952900478978179, "grad_norm": 0.31700700521469116, "learning_rate": 4.610840503382896e-11, "loss": 0.2104, "step": 18613 }, { "epoch": 4.9531665779670035, "grad_norm": 0.30111294984817505, "learning_rate": 4.5591827584923195e-11, "loss": 0.1945, "step": 18614 }, { "epoch": 4.953432676955828, "grad_norm": 0.281722754240036, "learning_rate": 4.5078159549316594e-11, "loss": 0.1667, "step": 18615 }, { "epoch": 4.953698775944652, "grad_norm": 0.3443453907966614, "learning_rate": 4.4567400941941666e-11, "loss": 0.1979, "step": 18616 }, { "epoch": 4.953964874933475, "grad_norm": 0.2845161259174347, "learning_rate": 4.40595517776865e-11, "loss": 0.218, "step": 18617 }, { "epoch": 4.954230973922299, "grad_norm": 0.346772164106369, "learning_rate": 4.3554612071328153e-11, "loss": 0.2003, "step": 18618 }, { "epoch": 4.954497072911123, "grad_norm": 0.31022822856903076, "learning_rate": 4.305258183755489e-11, "loss": 0.1832, "step": 18619 }, { "epoch": 4.9547631718999465, "grad_norm": 0.30651232600212097, "learning_rate": 4.255346109097724e-11, "loss": 0.2007, "step": 18620 }, { "epoch": 4.955029270888771, "grad_norm": 0.3054785430431366, "learning_rate": 4.205724984615022e-11, "loss": 0.2132, "step": 18621 }, { "epoch": 4.955295369877595, "grad_norm": 0.2891823649406433, "learning_rate": 4.1563948117484534e-11, "loss": 0.1861, "step": 18622 }, { "epoch": 4.955561468866418, "grad_norm": 0.3330530822277069, "learning_rate": 4.107355591935757e-11, "loss": 0.2218, "step": 18623 }, { "epoch": 4.955827567855242, "grad_norm": 0.30392855405807495, "learning_rate": 4.058607326603569e-11, "loss": 0.2066, "step": 18624 }, { "epoch": 4.956093666844066, "grad_norm": 0.29829347133636475, "learning_rate": 4.010150017171865e-11, "loss": 0.1849, "step": 18625 }, { "epoch": 4.9563597658328895, "grad_norm": 0.28520458936691284, "learning_rate": 3.9619836650495173e-11, "loss": 0.1827, "step": 18626 }, { "epoch": 4.956625864821714, "grad_norm": 0.28633204102516174, "learning_rate": 3.914108271639849e-11, "loss": 0.1598, "step": 18627 }, { "epoch": 4.956891963810538, "grad_norm": 0.296580970287323, "learning_rate": 3.8665238383350785e-11, "loss": 0.1875, "step": 18628 }, { "epoch": 4.957158062799361, "grad_norm": 0.2625484764575958, "learning_rate": 3.819230366521875e-11, "loss": 0.1782, "step": 18629 }, { "epoch": 4.957424161788185, "grad_norm": 0.30158066749572754, "learning_rate": 3.772227857575805e-11, "loss": 0.1945, "step": 18630 }, { "epoch": 4.957690260777009, "grad_norm": 0.299958735704422, "learning_rate": 3.725516312865773e-11, "loss": 0.1801, "step": 18631 }, { "epoch": 4.9579563597658325, "grad_norm": 0.2954237163066864, "learning_rate": 3.6790957337506925e-11, "loss": 0.1815, "step": 18632 }, { "epoch": 4.958222458754657, "grad_norm": 0.26809602975845337, "learning_rate": 3.632966121581704e-11, "loss": 0.1729, "step": 18633 }, { "epoch": 4.958488557743481, "grad_norm": 0.29899877309799194, "learning_rate": 3.587127477702179e-11, "loss": 0.205, "step": 18634 }, { "epoch": 4.958754656732305, "grad_norm": 0.30895280838012695, "learning_rate": 3.541579803445494e-11, "loss": 0.1792, "step": 18635 }, { "epoch": 4.959020755721128, "grad_norm": 0.2849825322628021, "learning_rate": 3.496323100138365e-11, "loss": 0.1821, "step": 18636 }, { "epoch": 4.959286854709952, "grad_norm": 0.3001309335231781, "learning_rate": 3.451357369096408e-11, "loss": 0.1726, "step": 18637 }, { "epoch": 4.9595529536987755, "grad_norm": 0.30483749508857727, "learning_rate": 3.4066826116307955e-11, "loss": 0.1846, "step": 18638 }, { "epoch": 4.9598190526876, "grad_norm": 0.300247460603714, "learning_rate": 3.3622988290404885e-11, "loss": 0.1928, "step": 18639 }, { "epoch": 4.960085151676424, "grad_norm": 0.32987338304519653, "learning_rate": 3.3182060226177865e-11, "loss": 0.1974, "step": 18640 }, { "epoch": 4.960351250665248, "grad_norm": 0.27581530809402466, "learning_rate": 3.274404193646107e-11, "loss": 0.1896, "step": 18641 }, { "epoch": 4.960617349654071, "grad_norm": 0.3184918165206909, "learning_rate": 3.230893343399988e-11, "loss": 0.1963, "step": 18642 }, { "epoch": 4.960883448642895, "grad_norm": 0.3003775477409363, "learning_rate": 3.187673473146191e-11, "loss": 0.1715, "step": 18643 }, { "epoch": 4.961149547631719, "grad_norm": 0.319837749004364, "learning_rate": 3.1447445841426004e-11, "loss": 0.1885, "step": 18644 }, { "epoch": 4.961415646620543, "grad_norm": 0.27717167139053345, "learning_rate": 3.102106677639327e-11, "loss": 0.1761, "step": 18645 }, { "epoch": 4.961681745609367, "grad_norm": 0.2973215878009796, "learning_rate": 3.05975975487649e-11, "loss": 0.2017, "step": 18646 }, { "epoch": 4.961947844598191, "grad_norm": 0.3047614097595215, "learning_rate": 3.017703817087547e-11, "loss": 0.2001, "step": 18647 }, { "epoch": 4.962213943587014, "grad_norm": 0.29341650009155273, "learning_rate": 2.9759388654959636e-11, "loss": 0.1767, "step": 18648 }, { "epoch": 4.962480042575838, "grad_norm": 0.31183555722236633, "learning_rate": 2.9344649013185454e-11, "loss": 0.1847, "step": 18649 }, { "epoch": 4.962746141564662, "grad_norm": 0.28164857625961304, "learning_rate": 2.893281925760993e-11, "loss": 0.1735, "step": 18650 }, { "epoch": 4.9630122405534856, "grad_norm": 0.31185653805732727, "learning_rate": 2.8523899400234584e-11, "loss": 0.2099, "step": 18651 }, { "epoch": 4.96327833954231, "grad_norm": 0.29738593101501465, "learning_rate": 2.8117889452949904e-11, "loss": 0.1927, "step": 18652 }, { "epoch": 4.963544438531134, "grad_norm": 0.3050462603569031, "learning_rate": 2.7714789427590867e-11, "loss": 0.2068, "step": 18653 }, { "epoch": 4.963810537519957, "grad_norm": 0.28524965047836304, "learning_rate": 2.7314599335870325e-11, "loss": 0.1778, "step": 18654 }, { "epoch": 4.964076636508781, "grad_norm": 0.2970822751522064, "learning_rate": 2.6917319189434518e-11, "loss": 0.1987, "step": 18655 }, { "epoch": 4.964342735497605, "grad_norm": 0.31201010942459106, "learning_rate": 2.6522948999874174e-11, "loss": 0.2061, "step": 18656 }, { "epoch": 4.9646088344864285, "grad_norm": 0.3242849111557007, "learning_rate": 2.6131488778646792e-11, "loss": 0.2021, "step": 18657 }, { "epoch": 4.964874933475253, "grad_norm": 0.2821778655052185, "learning_rate": 2.5742938537154368e-11, "loss": 0.1763, "step": 18658 }, { "epoch": 4.965141032464077, "grad_norm": 0.28852880001068115, "learning_rate": 2.535729828669897e-11, "loss": 0.1854, "step": 18659 }, { "epoch": 4.965407131452901, "grad_norm": 0.31494030356407166, "learning_rate": 2.4974568038516052e-11, "loss": 0.2113, "step": 18660 }, { "epoch": 4.965673230441724, "grad_norm": 0.2975262403488159, "learning_rate": 2.4594747803741157e-11, "loss": 0.1922, "step": 18661 }, { "epoch": 4.965939329430548, "grad_norm": 0.2963707447052002, "learning_rate": 2.4217837593420997e-11, "loss": 0.1949, "step": 18662 }, { "epoch": 4.9662054284193715, "grad_norm": 0.3126547038555145, "learning_rate": 2.3843837418546786e-11, "loss": 0.2006, "step": 18663 }, { "epoch": 4.966471527408196, "grad_norm": 0.294515460729599, "learning_rate": 2.3472747289987606e-11, "loss": 0.188, "step": 18664 }, { "epoch": 4.96673762639702, "grad_norm": 0.2813899517059326, "learning_rate": 2.3104567218545922e-11, "loss": 0.1694, "step": 18665 }, { "epoch": 4.967003725385844, "grad_norm": 0.2932601571083069, "learning_rate": 2.2739297214946496e-11, "loss": 0.1803, "step": 18666 }, { "epoch": 4.967269824374667, "grad_norm": 0.30292752385139465, "learning_rate": 2.2376937289814157e-11, "loss": 0.1889, "step": 18667 }, { "epoch": 4.967535923363491, "grad_norm": 0.27662280201911926, "learning_rate": 2.201748745370713e-11, "loss": 0.1793, "step": 18668 }, { "epoch": 4.967802022352315, "grad_norm": 0.3280746638774872, "learning_rate": 2.166094771708371e-11, "loss": 0.2113, "step": 18669 }, { "epoch": 4.968068121341139, "grad_norm": 0.28814733028411865, "learning_rate": 2.1307318090313387e-11, "loss": 0.1734, "step": 18670 }, { "epoch": 4.968334220329963, "grad_norm": 0.32004618644714355, "learning_rate": 2.0956598583687923e-11, "loss": 0.198, "step": 18671 }, { "epoch": 4.968600319318787, "grad_norm": 0.29448217153549194, "learning_rate": 2.0608789207432476e-11, "loss": 0.1998, "step": 18672 }, { "epoch": 4.96886641830761, "grad_norm": 0.31289947032928467, "learning_rate": 2.0263889971661173e-11, "loss": 0.2032, "step": 18673 }, { "epoch": 4.969132517296434, "grad_norm": 0.3138636648654938, "learning_rate": 1.9921900886410437e-11, "loss": 0.1958, "step": 18674 }, { "epoch": 4.969398616285258, "grad_norm": 0.29666846990585327, "learning_rate": 1.9582821961638962e-11, "loss": 0.1981, "step": 18675 }, { "epoch": 4.969664715274082, "grad_norm": 0.3000621497631073, "learning_rate": 1.9246653207205533e-11, "loss": 0.1901, "step": 18676 }, { "epoch": 4.969930814262906, "grad_norm": 0.2811009883880615, "learning_rate": 1.8913394632913414e-11, "loss": 0.1733, "step": 18677 }, { "epoch": 4.97019691325173, "grad_norm": 0.3017438054084778, "learning_rate": 1.858304624845486e-11, "loss": 0.1873, "step": 18678 }, { "epoch": 4.970463012240554, "grad_norm": 0.28191661834716797, "learning_rate": 1.8255608063444396e-11, "loss": 0.1841, "step": 18679 }, { "epoch": 4.970729111229377, "grad_norm": 0.28951552510261536, "learning_rate": 1.7931080087407735e-11, "loss": 0.1906, "step": 18680 }, { "epoch": 4.970995210218201, "grad_norm": 0.3079663813114166, "learning_rate": 1.760946232980398e-11, "loss": 0.1757, "step": 18681 }, { "epoch": 4.971261309207025, "grad_norm": 0.29356831312179565, "learning_rate": 1.729075479998121e-11, "loss": 0.2009, "step": 18682 }, { "epoch": 4.971527408195849, "grad_norm": 0.40017953515052795, "learning_rate": 1.697495750723199e-11, "loss": 0.2055, "step": 18683 }, { "epoch": 4.971793507184673, "grad_norm": 0.2935781478881836, "learning_rate": 1.6662070460726764e-11, "loss": 0.1885, "step": 18684 }, { "epoch": 4.972059606173497, "grad_norm": 0.2951190769672394, "learning_rate": 1.6352093669591562e-11, "loss": 0.1879, "step": 18685 }, { "epoch": 4.97232570516232, "grad_norm": 0.3065086603164673, "learning_rate": 1.6045027142841394e-11, "loss": 0.2107, "step": 18686 }, { "epoch": 4.972591804151144, "grad_norm": 0.29791247844696045, "learning_rate": 1.574087088941356e-11, "loss": 0.1752, "step": 18687 }, { "epoch": 4.9728579031399684, "grad_norm": 0.28271573781967163, "learning_rate": 1.5439624918156534e-11, "loss": 0.1724, "step": 18688 }, { "epoch": 4.973124002128792, "grad_norm": 0.3128052353858948, "learning_rate": 1.5141289237852185e-11, "loss": 0.199, "step": 18689 }, { "epoch": 4.973390101117616, "grad_norm": 0.5731390118598938, "learning_rate": 1.4845863857171347e-11, "loss": 0.1819, "step": 18690 }, { "epoch": 4.97365620010644, "grad_norm": 0.3076389729976654, "learning_rate": 1.455334878472936e-11, "loss": 0.1835, "step": 18691 }, { "epoch": 4.973922299095263, "grad_norm": 0.31365546584129333, "learning_rate": 1.426374402901942e-11, "loss": 0.1948, "step": 18692 }, { "epoch": 4.974188398084087, "grad_norm": 0.2995395064353943, "learning_rate": 1.397704959847923e-11, "loss": 0.1844, "step": 18693 }, { "epoch": 4.974454497072911, "grad_norm": 0.307594358921051, "learning_rate": 1.3693265501468766e-11, "loss": 0.1871, "step": 18694 }, { "epoch": 4.974720596061735, "grad_norm": 0.3159145712852478, "learning_rate": 1.3412391746225882e-11, "loss": 0.1887, "step": 18695 }, { "epoch": 4.974986695050559, "grad_norm": 0.32009604573249817, "learning_rate": 1.3134428340932924e-11, "loss": 0.1974, "step": 18696 }, { "epoch": 4.975252794039383, "grad_norm": 0.3031020164489746, "learning_rate": 1.285937529369452e-11, "loss": 0.1908, "step": 18697 }, { "epoch": 4.975518893028206, "grad_norm": 0.3231106400489807, "learning_rate": 1.2587232612493171e-11, "loss": 0.2098, "step": 18698 }, { "epoch": 4.97578499201703, "grad_norm": 0.3168588876724243, "learning_rate": 1.2318000305266974e-11, "loss": 0.2051, "step": 18699 }, { "epoch": 4.976051091005854, "grad_norm": 0.28669583797454834, "learning_rate": 1.2051678379854103e-11, "loss": 0.1902, "step": 18700 }, { "epoch": 4.976317189994678, "grad_norm": 0.3230595886707306, "learning_rate": 1.1788266844003913e-11, "loss": 0.2085, "step": 18701 }, { "epoch": 4.976583288983502, "grad_norm": 0.3180759847164154, "learning_rate": 1.1527765705365844e-11, "loss": 0.2124, "step": 18702 }, { "epoch": 4.976849387972326, "grad_norm": 0.2987428605556488, "learning_rate": 1.1270174971544922e-11, "loss": 0.1882, "step": 18703 }, { "epoch": 4.97711548696115, "grad_norm": 0.3019896149635315, "learning_rate": 1.101549465003515e-11, "loss": 0.1989, "step": 18704 }, { "epoch": 4.977381585949973, "grad_norm": 0.31066492199897766, "learning_rate": 1.0763724748241721e-11, "loss": 0.189, "step": 18705 }, { "epoch": 4.977647684938797, "grad_norm": 0.29999569058418274, "learning_rate": 1.0514865273492102e-11, "loss": 0.1885, "step": 18706 }, { "epoch": 4.977913783927621, "grad_norm": 0.3121890723705292, "learning_rate": 1.0268916233047153e-11, "loss": 0.2032, "step": 18707 }, { "epoch": 4.978179882916445, "grad_norm": 0.29379478096961975, "learning_rate": 1.002587763404561e-11, "loss": 0.2005, "step": 18708 }, { "epoch": 4.978445981905269, "grad_norm": 0.2896963655948639, "learning_rate": 9.785749483559591e-12, "loss": 0.1814, "step": 18709 }, { "epoch": 4.978712080894093, "grad_norm": 0.2796841859817505, "learning_rate": 9.548531788605708e-12, "loss": 0.1758, "step": 18710 }, { "epoch": 4.978978179882916, "grad_norm": 0.3060343265533447, "learning_rate": 9.314224556056239e-12, "loss": 0.1911, "step": 18711 }, { "epoch": 4.97924427887174, "grad_norm": 0.3147569000720978, "learning_rate": 9.08282779276126e-12, "loss": 0.1837, "step": 18712 }, { "epoch": 4.9795103778605645, "grad_norm": 0.3058799207210541, "learning_rate": 8.85434150542652e-12, "loss": 0.2211, "step": 18713 }, { "epoch": 4.979776476849388, "grad_norm": 0.3104289472103119, "learning_rate": 8.62876570072446e-12, "loss": 0.1718, "step": 18714 }, { "epoch": 4.980042575838212, "grad_norm": 0.2978777289390564, "learning_rate": 8.406100385216497e-12, "loss": 0.1941, "step": 18715 }, { "epoch": 4.980308674827036, "grad_norm": 0.3011627495288849, "learning_rate": 8.186345565386332e-12, "loss": 0.2102, "step": 18716 }, { "epoch": 4.980574773815859, "grad_norm": 0.2897833585739136, "learning_rate": 7.96950124761775e-12, "loss": 0.1713, "step": 18717 }, { "epoch": 4.980840872804683, "grad_norm": 0.2872862219810486, "learning_rate": 7.755567438227917e-12, "loss": 0.1828, "step": 18718 }, { "epoch": 4.9811069717935075, "grad_norm": 0.31726956367492676, "learning_rate": 7.544544143445187e-12, "loss": 0.2019, "step": 18719 }, { "epoch": 4.981373070782331, "grad_norm": 0.31278836727142334, "learning_rate": 7.336431369420193e-12, "loss": 0.1977, "step": 18720 }, { "epoch": 4.981639169771155, "grad_norm": 0.3092527687549591, "learning_rate": 7.131229122203652e-12, "loss": 0.1857, "step": 18721 }, { "epoch": 4.981905268759979, "grad_norm": 0.29237669706344604, "learning_rate": 6.928937407768565e-12, "loss": 0.1951, "step": 18722 }, { "epoch": 4.982171367748802, "grad_norm": 0.2715529203414917, "learning_rate": 6.729556231999112e-12, "loss": 0.1752, "step": 18723 }, { "epoch": 4.982437466737626, "grad_norm": 0.305517315864563, "learning_rate": 6.53308560070176e-12, "loss": 0.2166, "step": 18724 }, { "epoch": 4.9827035657264505, "grad_norm": 0.30832672119140625, "learning_rate": 6.339525519594157e-12, "loss": 0.1883, "step": 18725 }, { "epoch": 4.982969664715274, "grad_norm": 0.2955923080444336, "learning_rate": 6.1488759943162385e-12, "loss": 0.1835, "step": 18726 }, { "epoch": 4.983235763704098, "grad_norm": 0.3243430256843567, "learning_rate": 5.961137030408014e-12, "loss": 0.182, "step": 18727 }, { "epoch": 4.983501862692922, "grad_norm": 0.3098168969154358, "learning_rate": 5.7763086333428856e-12, "loss": 0.2029, "step": 18728 }, { "epoch": 4.983767961681746, "grad_norm": 0.2979908287525177, "learning_rate": 5.5943908084943315e-12, "loss": 0.1974, "step": 18729 }, { "epoch": 4.984034060670569, "grad_norm": 0.2883261740207672, "learning_rate": 5.415383561169218e-12, "loss": 0.1976, "step": 18730 }, { "epoch": 4.984300159659393, "grad_norm": 0.2931622564792633, "learning_rate": 5.239286896563388e-12, "loss": 0.1886, "step": 18731 }, { "epoch": 4.984566258648217, "grad_norm": 0.2864530682563782, "learning_rate": 5.066100819806074e-12, "loss": 0.1762, "step": 18732 }, { "epoch": 4.984832357637041, "grad_norm": 0.293373703956604, "learning_rate": 4.895825335937687e-12, "loss": 0.1751, "step": 18733 }, { "epoch": 4.985098456625865, "grad_norm": 0.2872990369796753, "learning_rate": 4.7284604499209236e-12, "loss": 0.1936, "step": 18734 }, { "epoch": 4.985364555614689, "grad_norm": 0.3283519148826599, "learning_rate": 4.564006166618562e-12, "loss": 0.2166, "step": 18735 }, { "epoch": 4.985630654603512, "grad_norm": 0.4379405975341797, "learning_rate": 4.402462490826764e-12, "loss": 0.2155, "step": 18736 }, { "epoch": 4.985896753592336, "grad_norm": 0.2882833480834961, "learning_rate": 4.243829427252877e-12, "loss": 0.1855, "step": 18737 }, { "epoch": 4.9861628525811605, "grad_norm": 0.3024487793445587, "learning_rate": 4.088106980493222e-12, "loss": 0.1998, "step": 18738 }, { "epoch": 4.986428951569984, "grad_norm": 0.3053094744682312, "learning_rate": 3.935295155088614e-12, "loss": 0.1784, "step": 18739 }, { "epoch": 4.986695050558808, "grad_norm": 0.2855670154094696, "learning_rate": 3.785393955502147e-12, "loss": 0.1862, "step": 18740 }, { "epoch": 4.986961149547632, "grad_norm": 0.28487950563430786, "learning_rate": 3.6384033860747955e-12, "loss": 0.181, "step": 18741 }, { "epoch": 4.987227248536455, "grad_norm": 0.3236205577850342, "learning_rate": 3.4943234511031204e-12, "loss": 0.2094, "step": 18742 }, { "epoch": 4.987493347525279, "grad_norm": 0.29796838760375977, "learning_rate": 3.3531541547726638e-12, "loss": 0.2003, "step": 18743 }, { "epoch": 4.9877594465141035, "grad_norm": 0.2920244634151459, "learning_rate": 3.2148955011912505e-12, "loss": 0.1683, "step": 18744 }, { "epoch": 4.988025545502927, "grad_norm": 0.2934262156486511, "learning_rate": 3.0795474943778875e-12, "loss": 0.1795, "step": 18745 }, { "epoch": 4.988291644491751, "grad_norm": 0.29583418369293213, "learning_rate": 2.9471101382960717e-12, "loss": 0.1852, "step": 18746 }, { "epoch": 4.988557743480575, "grad_norm": 0.28092288970947266, "learning_rate": 2.81758343676497e-12, "loss": 0.1766, "step": 18747 }, { "epoch": 4.988823842469398, "grad_norm": 0.29224181175231934, "learning_rate": 2.6909673935926468e-12, "loss": 0.1934, "step": 18748 }, { "epoch": 4.989089941458222, "grad_norm": 0.32259616255760193, "learning_rate": 2.5672620124317367e-12, "loss": 0.1848, "step": 18749 }, { "epoch": 4.9893560404470465, "grad_norm": 0.29985690116882324, "learning_rate": 2.4464672969015666e-12, "loss": 0.183, "step": 18750 }, { "epoch": 4.98962213943587, "grad_norm": 0.28816741704940796, "learning_rate": 2.3285832505104407e-12, "loss": 0.1751, "step": 18751 }, { "epoch": 4.989888238424694, "grad_norm": 0.3259725570678711, "learning_rate": 2.213609876700051e-12, "loss": 0.1689, "step": 18752 }, { "epoch": 4.990154337413518, "grad_norm": 0.30655407905578613, "learning_rate": 2.1015471788121685e-12, "loss": 0.1779, "step": 18753 }, { "epoch": 4.990420436402342, "grad_norm": 0.2945082187652588, "learning_rate": 1.9923951600997467e-12, "loss": 0.2063, "step": 18754 }, { "epoch": 4.990686535391165, "grad_norm": 0.30058541893959045, "learning_rate": 1.8861538237491258e-12, "loss": 0.193, "step": 18755 }, { "epoch": 4.9909526343799895, "grad_norm": 0.2776656448841095, "learning_rate": 1.7828231728467258e-12, "loss": 0.1692, "step": 18756 }, { "epoch": 4.991218733368813, "grad_norm": 0.2868916392326355, "learning_rate": 1.682403210401251e-12, "loss": 0.1924, "step": 18757 }, { "epoch": 4.991484832357637, "grad_norm": 0.31572458148002625, "learning_rate": 1.58489393934369e-12, "loss": 0.2123, "step": 18758 }, { "epoch": 4.991750931346461, "grad_norm": 0.269774466753006, "learning_rate": 1.4902953625051117e-12, "loss": 0.1663, "step": 18759 }, { "epoch": 4.992017030335285, "grad_norm": 0.30006980895996094, "learning_rate": 1.3986074826388694e-12, "loss": 0.1763, "step": 18760 }, { "epoch": 4.992283129324108, "grad_norm": 0.2971031367778778, "learning_rate": 1.3098303024206003e-12, "loss": 0.1723, "step": 18761 }, { "epoch": 4.9925492283129325, "grad_norm": 0.4298919141292572, "learning_rate": 1.223963824426022e-12, "loss": 0.2026, "step": 18762 }, { "epoch": 4.992815327301757, "grad_norm": 0.3181355893611908, "learning_rate": 1.1410080511531361e-12, "loss": 0.1876, "step": 18763 }, { "epoch": 4.99308142629058, "grad_norm": 0.2876644730567932, "learning_rate": 1.060962985022229e-12, "loss": 0.1873, "step": 18764 }, { "epoch": 4.993347525279404, "grad_norm": 0.3168115019798279, "learning_rate": 9.83828628364769e-13, "loss": 0.2259, "step": 18765 }, { "epoch": 4.993613624268228, "grad_norm": 0.30202198028564453, "learning_rate": 9.096049834234065e-13, "loss": 0.1958, "step": 18766 }, { "epoch": 4.993879723257051, "grad_norm": 0.29616186022758484, "learning_rate": 8.382920523630765e-13, "loss": 0.1974, "step": 18767 }, { "epoch": 4.994145822245875, "grad_norm": 0.299511194229126, "learning_rate": 7.698898372487938e-13, "loss": 0.1849, "step": 18768 }, { "epoch": 4.9944119212347, "grad_norm": 0.2901143729686737, "learning_rate": 7.043983400789599e-13, "loss": 0.1902, "step": 18769 }, { "epoch": 4.994678020223523, "grad_norm": 0.3051586151123047, "learning_rate": 6.418175627520561e-13, "loss": 0.1897, "step": 18770 }, { "epoch": 4.994944119212347, "grad_norm": 0.31149736046791077, "learning_rate": 5.821475071110527e-13, "loss": 0.193, "step": 18771 }, { "epoch": 4.995210218201171, "grad_norm": 0.3284060060977936, "learning_rate": 5.25388174865693e-13, "loss": 0.1967, "step": 18772 }, { "epoch": 4.995476317189994, "grad_norm": 0.29555052518844604, "learning_rate": 4.715395676813116e-13, "loss": 0.1753, "step": 18773 }, { "epoch": 4.995742416178818, "grad_norm": 0.3292394280433655, "learning_rate": 4.2060168713442536e-13, "loss": 0.1965, "step": 18774 }, { "epoch": 4.9960085151676425, "grad_norm": 0.3069424629211426, "learning_rate": 3.7257453467942623e-13, "loss": 0.1771, "step": 18775 }, { "epoch": 4.996274614156466, "grad_norm": 0.2867438793182373, "learning_rate": 3.2745811174850204e-13, "loss": 0.1735, "step": 18776 }, { "epoch": 4.99654071314529, "grad_norm": 0.3576580286026001, "learning_rate": 2.852524196295114e-13, "loss": 0.234, "step": 18777 }, { "epoch": 4.996806812134114, "grad_norm": 0.2737654149532318, "learning_rate": 2.459574595659042e-13, "loss": 0.1955, "step": 18778 }, { "epoch": 4.997072911122938, "grad_norm": 0.2906860113143921, "learning_rate": 2.0957323269010785e-13, "loss": 0.1743, "step": 18779 }, { "epoch": 4.997339010111761, "grad_norm": 0.32635819911956787, "learning_rate": 1.760997400790387e-13, "loss": 0.2054, "step": 18780 }, { "epoch": 4.9976051091005855, "grad_norm": 0.29705750942230225, "learning_rate": 1.4553698267638636e-13, "loss": 0.1979, "step": 18781 }, { "epoch": 4.997871208089409, "grad_norm": 0.2904598116874695, "learning_rate": 1.178849614036359e-13, "loss": 0.2013, "step": 18782 }, { "epoch": 4.998137307078233, "grad_norm": 0.27896788716316223, "learning_rate": 9.314367704904569e-14, "loss": 0.2014, "step": 18783 }, { "epoch": 4.998403406067057, "grad_norm": 0.2829393446445465, "learning_rate": 7.131313033426068e-14, "loss": 0.176, "step": 18784 }, { "epoch": 4.998669505055881, "grad_norm": 0.30721598863601685, "learning_rate": 5.239332190321022e-14, "loss": 0.194, "step": 18785 }, { "epoch": 4.998935604044704, "grad_norm": 0.28402408957481384, "learning_rate": 3.6384252288801375e-14, "loss": 0.1862, "step": 18786 }, { "epoch": 4.9992017030335285, "grad_norm": 0.304360568523407, "learning_rate": 2.3285921968430045e-14, "loss": 0.2138, "step": 18787 }, { "epoch": 4.999467802022353, "grad_norm": 0.2927015423774719, "learning_rate": 1.3098331330674283e-14, "loss": 0.2114, "step": 18788 }, { "epoch": 4.999733901011176, "grad_norm": 0.2910849452018738, "learning_rate": 5.82148066419208e-15, "loss": 0.198, "step": 18789 }, { "epoch": 5.0, "grad_norm": 0.3695437014102936, "learning_rate": 1.4553701799258078e-15, "loss": 0.1815, "step": 18790 }, { "epoch": 5.0, "step": 18790, "total_flos": 3.766642029157679e+20, "train_loss": 0.07627738880755865, "train_runtime": 166255.2229, "train_samples_per_second": 7.233, "train_steps_per_second": 0.113 } ], "logging_steps": 1.0, "max_steps": 18790, "num_input_tokens_seen": 0, "num_train_epochs": 5, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 3.766642029157679e+20, "train_batch_size": 2, "trial_name": null, "trial_params": null }