{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 17.793594306049823, "eval_steps": 500, "global_step": 10000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.017793594306049824, "grad_norm": 8.256741523742676, "learning_rate": 1.8e-06, "loss": 0.7098, "step": 10 }, { "epoch": 0.03558718861209965, "grad_norm": 5.343027114868164, "learning_rate": 3.8e-06, "loss": 0.6435, "step": 20 }, { "epoch": 0.05338078291814947, "grad_norm": 3.1888511180877686, "learning_rate": 5.8e-06, "loss": 0.3916, "step": 30 }, { "epoch": 0.0711743772241993, "grad_norm": 2.084213972091675, "learning_rate": 7.8e-06, "loss": 0.2048, "step": 40 }, { "epoch": 0.08896797153024912, "grad_norm": 0.9424139261245728, "learning_rate": 9.800000000000001e-06, "loss": 0.1579, "step": 50 }, { "epoch": 0.10676156583629894, "grad_norm": 1.0634982585906982, "learning_rate": 1.18e-05, "loss": 0.1331, "step": 60 }, { "epoch": 0.12455516014234876, "grad_norm": 0.9179893136024475, "learning_rate": 1.3800000000000002e-05, "loss": 0.091, "step": 70 }, { "epoch": 0.1423487544483986, "grad_norm": 0.7661588191986084, "learning_rate": 1.58e-05, "loss": 0.099, "step": 80 }, { "epoch": 0.1601423487544484, "grad_norm": 1.236345887184143, "learning_rate": 1.78e-05, "loss": 0.0858, "step": 90 }, { "epoch": 0.17793594306049823, "grad_norm": 0.5287535786628723, "learning_rate": 1.9800000000000004e-05, "loss": 0.0801, "step": 100 }, { "epoch": 0.19572953736654805, "grad_norm": 0.6885743141174316, "learning_rate": 2.18e-05, "loss": 0.0749, "step": 110 }, { "epoch": 0.21352313167259787, "grad_norm": 0.9239180684089661, "learning_rate": 2.38e-05, "loss": 0.0617, "step": 120 }, { "epoch": 0.2313167259786477, "grad_norm": 0.502554714679718, "learning_rate": 2.58e-05, "loss": 0.0683, "step": 130 }, { "epoch": 0.2491103202846975, "grad_norm": 0.6825931072235107, "learning_rate": 2.7800000000000005e-05, "loss": 0.0668, "step": 140 }, { "epoch": 0.2669039145907473, "grad_norm": 1.147830843925476, "learning_rate": 2.98e-05, "loss": 0.0629, "step": 150 }, { "epoch": 0.2846975088967972, "grad_norm": 0.8498034477233887, "learning_rate": 3.18e-05, "loss": 0.0575, "step": 160 }, { "epoch": 0.302491103202847, "grad_norm": 0.9111298322677612, "learning_rate": 3.38e-05, "loss": 0.0551, "step": 170 }, { "epoch": 0.3202846975088968, "grad_norm": 0.4847106635570526, "learning_rate": 3.58e-05, "loss": 0.053, "step": 180 }, { "epoch": 0.33807829181494664, "grad_norm": 0.5012326836585999, "learning_rate": 3.7800000000000004e-05, "loss": 0.0418, "step": 190 }, { "epoch": 0.35587188612099646, "grad_norm": 0.43026965856552124, "learning_rate": 3.9800000000000005e-05, "loss": 0.0444, "step": 200 }, { "epoch": 0.3736654804270463, "grad_norm": 0.5495039820671082, "learning_rate": 4.18e-05, "loss": 0.0472, "step": 210 }, { "epoch": 0.3914590747330961, "grad_norm": 0.6074822545051575, "learning_rate": 4.38e-05, "loss": 0.0409, "step": 220 }, { "epoch": 0.4092526690391459, "grad_norm": 0.5337981581687927, "learning_rate": 4.58e-05, "loss": 0.039, "step": 230 }, { "epoch": 0.42704626334519574, "grad_norm": 0.6981046795845032, "learning_rate": 4.78e-05, "loss": 0.0436, "step": 240 }, { "epoch": 0.44483985765124556, "grad_norm": 0.6591747999191284, "learning_rate": 4.9800000000000004e-05, "loss": 0.0347, "step": 250 }, { "epoch": 0.4626334519572954, "grad_norm": 0.9619963765144348, "learning_rate": 5.1800000000000005e-05, "loss": 0.0377, "step": 260 }, { "epoch": 0.4804270462633452, "grad_norm": 0.6122570633888245, "learning_rate": 5.380000000000001e-05, "loss": 0.0467, "step": 270 }, { "epoch": 0.498220640569395, "grad_norm": 0.6114277243614197, "learning_rate": 5.580000000000001e-05, "loss": 0.0344, "step": 280 }, { "epoch": 0.5160142348754448, "grad_norm": 0.48597225546836853, "learning_rate": 5.7799999999999995e-05, "loss": 0.0416, "step": 290 }, { "epoch": 0.5338078291814946, "grad_norm": 0.5121509432792664, "learning_rate": 5.9800000000000003e-05, "loss": 0.0339, "step": 300 }, { "epoch": 0.5516014234875445, "grad_norm": 0.7633475065231323, "learning_rate": 6.18e-05, "loss": 0.0535, "step": 310 }, { "epoch": 0.5693950177935944, "grad_norm": 0.43353763222694397, "learning_rate": 6.38e-05, "loss": 0.0346, "step": 320 }, { "epoch": 0.5871886120996441, "grad_norm": 0.4473856985569, "learning_rate": 6.58e-05, "loss": 0.0358, "step": 330 }, { "epoch": 0.604982206405694, "grad_norm": 0.8218453526496887, "learning_rate": 6.780000000000001e-05, "loss": 0.0351, "step": 340 }, { "epoch": 0.6227758007117438, "grad_norm": 0.6579710841178894, "learning_rate": 6.98e-05, "loss": 0.0317, "step": 350 }, { "epoch": 0.6405693950177936, "grad_norm": 0.7301086187362671, "learning_rate": 7.18e-05, "loss": 0.0326, "step": 360 }, { "epoch": 0.6583629893238434, "grad_norm": 0.6139028668403625, "learning_rate": 7.38e-05, "loss": 0.0422, "step": 370 }, { "epoch": 0.6761565836298933, "grad_norm": 0.9374700784683228, "learning_rate": 7.58e-05, "loss": 0.0358, "step": 380 }, { "epoch": 0.693950177935943, "grad_norm": 0.4101995825767517, "learning_rate": 7.780000000000001e-05, "loss": 0.0349, "step": 390 }, { "epoch": 0.7117437722419929, "grad_norm": 0.8889083862304688, "learning_rate": 7.98e-05, "loss": 0.028, "step": 400 }, { "epoch": 0.7295373665480427, "grad_norm": 0.6061159372329712, "learning_rate": 8.18e-05, "loss": 0.0339, "step": 410 }, { "epoch": 0.7473309608540926, "grad_norm": 0.49921879172325134, "learning_rate": 8.38e-05, "loss": 0.0288, "step": 420 }, { "epoch": 0.7651245551601423, "grad_norm": 0.6122949719429016, "learning_rate": 8.58e-05, "loss": 0.033, "step": 430 }, { "epoch": 0.7829181494661922, "grad_norm": 0.513400137424469, "learning_rate": 8.78e-05, "loss": 0.0283, "step": 440 }, { "epoch": 0.800711743772242, "grad_norm": 0.5969704985618591, "learning_rate": 8.98e-05, "loss": 0.0278, "step": 450 }, { "epoch": 0.8185053380782918, "grad_norm": 0.7151464819908142, "learning_rate": 9.180000000000001e-05, "loss": 0.0345, "step": 460 }, { "epoch": 0.8362989323843416, "grad_norm": 0.8436950445175171, "learning_rate": 9.38e-05, "loss": 0.0342, "step": 470 }, { "epoch": 0.8540925266903915, "grad_norm": 0.32498398423194885, "learning_rate": 9.58e-05, "loss": 0.0311, "step": 480 }, { "epoch": 0.8718861209964412, "grad_norm": 0.5976306200027466, "learning_rate": 9.78e-05, "loss": 0.028, "step": 490 }, { "epoch": 0.8896797153024911, "grad_norm": 0.7325876951217651, "learning_rate": 9.98e-05, "loss": 0.0273, "step": 500 }, { "epoch": 0.9074733096085409, "grad_norm": 0.5371050834655762, "learning_rate": 9.9999778549206e-05, "loss": 0.0317, "step": 510 }, { "epoch": 0.9252669039145908, "grad_norm": 0.4035126864910126, "learning_rate": 9.999901304280685e-05, "loss": 0.0276, "step": 520 }, { "epoch": 0.9430604982206405, "grad_norm": 0.6664149165153503, "learning_rate": 9.999770075521164e-05, "loss": 0.034, "step": 530 }, { "epoch": 0.9608540925266904, "grad_norm": 0.7558611631393433, "learning_rate": 9.99958417007713e-05, "loss": 0.0279, "step": 540 }, { "epoch": 0.9786476868327402, "grad_norm": 0.7639133334159851, "learning_rate": 9.999343589981615e-05, "loss": 0.0299, "step": 550 }, { "epoch": 0.99644128113879, "grad_norm": 0.4322656989097595, "learning_rate": 9.999048337865568e-05, "loss": 0.0263, "step": 560 }, { "epoch": 1.0142348754448398, "grad_norm": 0.556773841381073, "learning_rate": 9.998698416957815e-05, "loss": 0.0236, "step": 570 }, { "epoch": 1.0320284697508897, "grad_norm": 0.4819738268852234, "learning_rate": 9.998293831085037e-05, "loss": 0.0266, "step": 580 }, { "epoch": 1.0498220640569396, "grad_norm": 0.5747538805007935, "learning_rate": 9.997834584671719e-05, "loss": 0.0274, "step": 590 }, { "epoch": 1.0676156583629894, "grad_norm": 0.6461822986602783, "learning_rate": 9.997320682740107e-05, "loss": 0.032, "step": 600 }, { "epoch": 1.085409252669039, "grad_norm": 0.5052154064178467, "learning_rate": 9.996752130910149e-05, "loss": 0.0354, "step": 610 }, { "epoch": 1.103202846975089, "grad_norm": 0.4525640308856964, "learning_rate": 9.99612893539944e-05, "loss": 0.0271, "step": 620 }, { "epoch": 1.1209964412811388, "grad_norm": 0.37099045515060425, "learning_rate": 9.995451103023144e-05, "loss": 0.0213, "step": 630 }, { "epoch": 1.1387900355871885, "grad_norm": 0.4526245594024658, "learning_rate": 9.994718641193928e-05, "loss": 0.0216, "step": 640 }, { "epoch": 1.1565836298932384, "grad_norm": 0.40107694268226624, "learning_rate": 9.993931557921874e-05, "loss": 0.0238, "step": 650 }, { "epoch": 1.1743772241992882, "grad_norm": 0.4636555016040802, "learning_rate": 9.993089861814402e-05, "loss": 0.0235, "step": 660 }, { "epoch": 1.1921708185053381, "grad_norm": 0.638054370880127, "learning_rate": 9.992193562076166e-05, "loss": 0.0364, "step": 670 }, { "epoch": 1.209964412811388, "grad_norm": 0.3661559522151947, "learning_rate": 9.991242668508954e-05, "loss": 0.022, "step": 680 }, { "epoch": 1.2277580071174377, "grad_norm": 0.39624184370040894, "learning_rate": 9.990237191511587e-05, "loss": 0.0274, "step": 690 }, { "epoch": 1.2455516014234875, "grad_norm": 0.6092879772186279, "learning_rate": 9.989177142079802e-05, "loss": 0.0251, "step": 700 }, { "epoch": 1.2633451957295374, "grad_norm": 0.7996970415115356, "learning_rate": 9.988062531806126e-05, "loss": 0.0313, "step": 710 }, { "epoch": 1.281138790035587, "grad_norm": 0.7109901905059814, "learning_rate": 9.986893372879762e-05, "loss": 0.0251, "step": 720 }, { "epoch": 1.298932384341637, "grad_norm": 0.6405479311943054, "learning_rate": 9.985669678086443e-05, "loss": 0.0303, "step": 730 }, { "epoch": 1.3167259786476868, "grad_norm": 0.38322243094444275, "learning_rate": 9.984391460808298e-05, "loss": 0.0214, "step": 740 }, { "epoch": 1.3345195729537367, "grad_norm": 0.4705495238304138, "learning_rate": 9.983058735023709e-05, "loss": 0.0236, "step": 750 }, { "epoch": 1.3523131672597866, "grad_norm": 0.5615078210830688, "learning_rate": 9.98167151530715e-05, "loss": 0.0207, "step": 760 }, { "epoch": 1.3701067615658362, "grad_norm": 0.41926711797714233, "learning_rate": 9.980229816829034e-05, "loss": 0.0179, "step": 770 }, { "epoch": 1.387900355871886, "grad_norm": 0.26638129353523254, "learning_rate": 9.978733655355544e-05, "loss": 0.0222, "step": 780 }, { "epoch": 1.405693950177936, "grad_norm": 0.32848984003067017, "learning_rate": 9.977183047248464e-05, "loss": 0.0233, "step": 790 }, { "epoch": 1.4234875444839858, "grad_norm": 0.41427695751190186, "learning_rate": 9.975578009464992e-05, "loss": 0.023, "step": 800 }, { "epoch": 1.4412811387900355, "grad_norm": 0.6090890765190125, "learning_rate": 9.97391855955757e-05, "loss": 0.0283, "step": 810 }, { "epoch": 1.4590747330960854, "grad_norm": 0.40939247608184814, "learning_rate": 9.972204715673669e-05, "loss": 0.0228, "step": 820 }, { "epoch": 1.4768683274021353, "grad_norm": 0.44156694412231445, "learning_rate": 9.970436496555617e-05, "loss": 0.0241, "step": 830 }, { "epoch": 1.4946619217081851, "grad_norm": 0.5506104826927185, "learning_rate": 9.968613921540373e-05, "loss": 0.018, "step": 840 }, { "epoch": 1.512455516014235, "grad_norm": 0.6918988823890686, "learning_rate": 9.966737010559326e-05, "loss": 0.0332, "step": 850 }, { "epoch": 1.5302491103202847, "grad_norm": 0.500394344329834, "learning_rate": 9.964805784138072e-05, "loss": 0.0249, "step": 860 }, { "epoch": 1.5480427046263345, "grad_norm": 0.7790364623069763, "learning_rate": 9.962820263396195e-05, "loss": 0.0237, "step": 870 }, { "epoch": 1.5658362989323842, "grad_norm": 0.4147443473339081, "learning_rate": 9.960780470047033e-05, "loss": 0.0218, "step": 880 }, { "epoch": 1.583629893238434, "grad_norm": 0.5203830599784851, "learning_rate": 9.958686426397437e-05, "loss": 0.0195, "step": 890 }, { "epoch": 1.601423487544484, "grad_norm": 0.3408946096897125, "learning_rate": 9.956538155347534e-05, "loss": 0.0255, "step": 900 }, { "epoch": 1.6192170818505338, "grad_norm": 0.3399748206138611, "learning_rate": 9.95433568039047e-05, "loss": 0.0199, "step": 910 }, { "epoch": 1.6370106761565837, "grad_norm": 0.5042240023612976, "learning_rate": 9.952079025612162e-05, "loss": 0.0198, "step": 920 }, { "epoch": 1.6548042704626336, "grad_norm": 0.32233956456184387, "learning_rate": 9.949768215691022e-05, "loss": 0.0184, "step": 930 }, { "epoch": 1.6725978647686834, "grad_norm": 0.2752324342727661, "learning_rate": 9.9474032758977e-05, "loss": 0.0183, "step": 940 }, { "epoch": 1.690391459074733, "grad_norm": 0.33111658692359924, "learning_rate": 9.944984232094794e-05, "loss": 0.0276, "step": 950 }, { "epoch": 1.708185053380783, "grad_norm": 0.5611407160758972, "learning_rate": 9.942511110736584e-05, "loss": 0.0172, "step": 960 }, { "epoch": 1.7259786476868326, "grad_norm": 0.5938228964805603, "learning_rate": 9.939983938868726e-05, "loss": 0.0194, "step": 970 }, { "epoch": 1.7437722419928825, "grad_norm": 0.34952178597450256, "learning_rate": 9.93740274412797e-05, "loss": 0.0204, "step": 980 }, { "epoch": 1.7615658362989324, "grad_norm": 0.3609432280063629, "learning_rate": 9.934767554741846e-05, "loss": 0.0247, "step": 990 }, { "epoch": 1.7793594306049823, "grad_norm": 0.44361451268196106, "learning_rate": 9.932078399528361e-05, "loss": 0.0218, "step": 1000 }, { "epoch": 1.7971530249110321, "grad_norm": 0.305938720703125, "learning_rate": 9.929335307895689e-05, "loss": 0.0212, "step": 1010 }, { "epoch": 1.814946619217082, "grad_norm": 0.28131330013275146, "learning_rate": 9.926538309841839e-05, "loss": 0.0175, "step": 1020 }, { "epoch": 1.8327402135231317, "grad_norm": 0.5200617909431458, "learning_rate": 9.923687435954334e-05, "loss": 0.0187, "step": 1030 }, { "epoch": 1.8505338078291815, "grad_norm": 0.31232354044914246, "learning_rate": 9.920782717409873e-05, "loss": 0.0197, "step": 1040 }, { "epoch": 1.8683274021352312, "grad_norm": 0.43739986419677734, "learning_rate": 9.917824185973994e-05, "loss": 0.0164, "step": 1050 }, { "epoch": 1.886120996441281, "grad_norm": 0.474654883146286, "learning_rate": 9.914811874000723e-05, "loss": 0.0248, "step": 1060 }, { "epoch": 1.903914590747331, "grad_norm": 0.32275819778442383, "learning_rate": 9.911745814432218e-05, "loss": 0.0212, "step": 1070 }, { "epoch": 1.9217081850533808, "grad_norm": 0.4485851228237152, "learning_rate": 9.90862604079842e-05, "loss": 0.0207, "step": 1080 }, { "epoch": 1.9395017793594307, "grad_norm": 0.4017932116985321, "learning_rate": 9.90545258721667e-05, "loss": 0.0129, "step": 1090 }, { "epoch": 1.9572953736654806, "grad_norm": 0.43503502011299133, "learning_rate": 9.90222548839135e-05, "loss": 0.0183, "step": 1100 }, { "epoch": 1.9750889679715302, "grad_norm": 0.39594727754592896, "learning_rate": 9.898944779613495e-05, "loss": 0.0191, "step": 1110 }, { "epoch": 1.99288256227758, "grad_norm": 0.37174543738365173, "learning_rate": 9.89561049676041e-05, "loss": 0.0201, "step": 1120 }, { "epoch": 2.0106761565836297, "grad_norm": 0.24639452993869781, "learning_rate": 9.89222267629528e-05, "loss": 0.0213, "step": 1130 }, { "epoch": 2.0284697508896796, "grad_norm": 0.4444299638271332, "learning_rate": 9.888781355266763e-05, "loss": 0.0203, "step": 1140 }, { "epoch": 2.0462633451957295, "grad_norm": 0.3867044448852539, "learning_rate": 9.885286571308598e-05, "loss": 0.0222, "step": 1150 }, { "epoch": 2.0640569395017794, "grad_norm": 0.9543397426605225, "learning_rate": 9.881738362639182e-05, "loss": 0.0315, "step": 1160 }, { "epoch": 2.0818505338078293, "grad_norm": 0.4074014723300934, "learning_rate": 9.878136768061154e-05, "loss": 0.022, "step": 1170 }, { "epoch": 2.099644128113879, "grad_norm": 0.320652574300766, "learning_rate": 9.874481826960979e-05, "loss": 0.0207, "step": 1180 }, { "epoch": 2.117437722419929, "grad_norm": 0.3749418258666992, "learning_rate": 9.870773579308503e-05, "loss": 0.0206, "step": 1190 }, { "epoch": 2.135231316725979, "grad_norm": 0.3456991910934448, "learning_rate": 9.867012065656533e-05, "loss": 0.0199, "step": 1200 }, { "epoch": 2.1530249110320283, "grad_norm": 0.3032795190811157, "learning_rate": 9.863197327140376e-05, "loss": 0.0199, "step": 1210 }, { "epoch": 2.170818505338078, "grad_norm": 0.6455495953559875, "learning_rate": 9.859329405477403e-05, "loss": 0.016, "step": 1220 }, { "epoch": 2.188612099644128, "grad_norm": 0.4881193935871124, "learning_rate": 9.855408342966585e-05, "loss": 0.0207, "step": 1230 }, { "epoch": 2.206405693950178, "grad_norm": 0.6028072834014893, "learning_rate": 9.851434182488033e-05, "loss": 0.0215, "step": 1240 }, { "epoch": 2.224199288256228, "grad_norm": 0.4521181583404541, "learning_rate": 9.84740696750253e-05, "loss": 0.025, "step": 1250 }, { "epoch": 2.2419928825622777, "grad_norm": 0.5232805013656616, "learning_rate": 9.843326742051055e-05, "loss": 0.0205, "step": 1260 }, { "epoch": 2.2597864768683276, "grad_norm": 0.42357897758483887, "learning_rate": 9.839193550754297e-05, "loss": 0.0237, "step": 1270 }, { "epoch": 2.277580071174377, "grad_norm": 0.576448380947113, "learning_rate": 9.835007438812177e-05, "loss": 0.02, "step": 1280 }, { "epoch": 2.295373665480427, "grad_norm": 0.6486237049102783, "learning_rate": 9.830768452003341e-05, "loss": 0.0276, "step": 1290 }, { "epoch": 2.3131672597864767, "grad_norm": 0.498912513256073, "learning_rate": 9.826476636684671e-05, "loss": 0.0218, "step": 1300 }, { "epoch": 2.3309608540925266, "grad_norm": 0.8010467290878296, "learning_rate": 9.822132039790773e-05, "loss": 0.02, "step": 1310 }, { "epoch": 2.3487544483985765, "grad_norm": 0.4010263681411743, "learning_rate": 9.817734708833461e-05, "loss": 0.022, "step": 1320 }, { "epoch": 2.3665480427046264, "grad_norm": 0.49060800671577454, "learning_rate": 9.813284691901243e-05, "loss": 0.0224, "step": 1330 }, { "epoch": 2.3843416370106763, "grad_norm": 0.6511074900627136, "learning_rate": 9.808782037658792e-05, "loss": 0.0201, "step": 1340 }, { "epoch": 2.402135231316726, "grad_norm": 0.545953631401062, "learning_rate": 9.804226795346411e-05, "loss": 0.0189, "step": 1350 }, { "epoch": 2.419928825622776, "grad_norm": 0.36032336950302124, "learning_rate": 9.799619014779503e-05, "loss": 0.0173, "step": 1360 }, { "epoch": 2.4377224199288254, "grad_norm": 0.4656877815723419, "learning_rate": 9.794958746348013e-05, "loss": 0.0132, "step": 1370 }, { "epoch": 2.4555160142348753, "grad_norm": 0.4230228364467621, "learning_rate": 9.790246041015896e-05, "loss": 0.0196, "step": 1380 }, { "epoch": 2.473309608540925, "grad_norm": 0.40642184019088745, "learning_rate": 9.785480950320538e-05, "loss": 0.0167, "step": 1390 }, { "epoch": 2.491103202846975, "grad_norm": 0.3487657904624939, "learning_rate": 9.78066352637221e-05, "loss": 0.0196, "step": 1400 }, { "epoch": 2.508896797153025, "grad_norm": 0.43371060490608215, "learning_rate": 9.775793821853488e-05, "loss": 0.0189, "step": 1410 }, { "epoch": 2.526690391459075, "grad_norm": 0.3415305018424988, "learning_rate": 9.77087189001868e-05, "loss": 0.0205, "step": 1420 }, { "epoch": 2.5444839857651247, "grad_norm": 0.38975775241851807, "learning_rate": 9.765897784693243e-05, "loss": 0.0177, "step": 1430 }, { "epoch": 2.562277580071174, "grad_norm": 0.4732246398925781, "learning_rate": 9.760871560273197e-05, "loss": 0.0225, "step": 1440 }, { "epoch": 2.580071174377224, "grad_norm": 0.4898048937320709, "learning_rate": 9.755793271724526e-05, "loss": 0.0213, "step": 1450 }, { "epoch": 2.597864768683274, "grad_norm": 0.34239596128463745, "learning_rate": 9.750662974582584e-05, "loss": 0.0193, "step": 1460 }, { "epoch": 2.6156583629893237, "grad_norm": 0.5860516428947449, "learning_rate": 9.745480724951473e-05, "loss": 0.0194, "step": 1470 }, { "epoch": 2.6334519572953736, "grad_norm": 0.30503103137016296, "learning_rate": 9.740246579503447e-05, "loss": 0.0244, "step": 1480 }, { "epoch": 2.6512455516014235, "grad_norm": 0.3923785984516144, "learning_rate": 9.734960595478284e-05, "loss": 0.0159, "step": 1490 }, { "epoch": 2.6690391459074734, "grad_norm": 0.37029045820236206, "learning_rate": 9.729622830682657e-05, "loss": 0.0209, "step": 1500 }, { "epoch": 2.6868327402135233, "grad_norm": 0.2239556461572647, "learning_rate": 9.724233343489504e-05, "loss": 0.0231, "step": 1510 }, { "epoch": 2.704626334519573, "grad_norm": 0.4776449203491211, "learning_rate": 9.718792192837396e-05, "loss": 0.0175, "step": 1520 }, { "epoch": 2.722419928825623, "grad_norm": 0.3160976469516754, "learning_rate": 9.713299438229886e-05, "loss": 0.0196, "step": 1530 }, { "epoch": 2.7402135231316724, "grad_norm": 0.3945485055446625, "learning_rate": 9.707755139734855e-05, "loss": 0.0189, "step": 1540 }, { "epoch": 2.7580071174377223, "grad_norm": 0.6034446358680725, "learning_rate": 9.702159357983866e-05, "loss": 0.026, "step": 1550 }, { "epoch": 2.775800711743772, "grad_norm": 0.4630682170391083, "learning_rate": 9.696512154171492e-05, "loss": 0.0171, "step": 1560 }, { "epoch": 2.793594306049822, "grad_norm": 0.43598613142967224, "learning_rate": 9.690813590054645e-05, "loss": 0.0171, "step": 1570 }, { "epoch": 2.811387900355872, "grad_norm": 0.3855835497379303, "learning_rate": 9.685063727951914e-05, "loss": 0.017, "step": 1580 }, { "epoch": 2.829181494661922, "grad_norm": 0.37535423040390015, "learning_rate": 9.679262630742865e-05, "loss": 0.0209, "step": 1590 }, { "epoch": 2.8469750889679717, "grad_norm": 0.35578790307044983, "learning_rate": 9.673410361867373e-05, "loss": 0.0162, "step": 1600 }, { "epoch": 2.864768683274021, "grad_norm": 0.4471454620361328, "learning_rate": 9.667506985324909e-05, "loss": 0.0158, "step": 1610 }, { "epoch": 2.882562277580071, "grad_norm": 0.5354124903678894, "learning_rate": 9.661552565673855e-05, "loss": 0.0152, "step": 1620 }, { "epoch": 2.900355871886121, "grad_norm": 0.3263882100582123, "learning_rate": 9.655547168030789e-05, "loss": 0.0144, "step": 1630 }, { "epoch": 2.9181494661921707, "grad_norm": 0.2756742537021637, "learning_rate": 9.649490858069777e-05, "loss": 0.015, "step": 1640 }, { "epoch": 2.9359430604982206, "grad_norm": 0.3291882276535034, "learning_rate": 9.643383702021658e-05, "loss": 0.0219, "step": 1650 }, { "epoch": 2.9537366548042705, "grad_norm": 0.19895005226135254, "learning_rate": 9.637225766673307e-05, "loss": 0.0124, "step": 1660 }, { "epoch": 2.9715302491103204, "grad_norm": 0.3430637717247009, "learning_rate": 9.631017119366922e-05, "loss": 0.0166, "step": 1670 }, { "epoch": 2.9893238434163703, "grad_norm": 0.3490820825099945, "learning_rate": 9.624757827999273e-05, "loss": 0.0232, "step": 1680 }, { "epoch": 3.00711743772242, "grad_norm": 0.5379742383956909, "learning_rate": 9.618447961020971e-05, "loss": 0.0234, "step": 1690 }, { "epoch": 3.0249110320284696, "grad_norm": 0.26694798469543457, "learning_rate": 9.612087587435707e-05, "loss": 0.0147, "step": 1700 }, { "epoch": 3.0427046263345194, "grad_norm": 0.39126265048980713, "learning_rate": 9.605676776799508e-05, "loss": 0.0213, "step": 1710 }, { "epoch": 3.0604982206405693, "grad_norm": 0.3498062491416931, "learning_rate": 9.599215599219973e-05, "loss": 0.0207, "step": 1720 }, { "epoch": 3.078291814946619, "grad_norm": 0.46342992782592773, "learning_rate": 9.592704125355505e-05, "loss": 0.0193, "step": 1730 }, { "epoch": 3.096085409252669, "grad_norm": 0.36067822575569153, "learning_rate": 9.586142426414538e-05, "loss": 0.0177, "step": 1740 }, { "epoch": 3.113879003558719, "grad_norm": 0.49401432275772095, "learning_rate": 9.57953057415476e-05, "loss": 0.0196, "step": 1750 }, { "epoch": 3.131672597864769, "grad_norm": 0.3961247205734253, "learning_rate": 9.572868640882328e-05, "loss": 0.02, "step": 1760 }, { "epoch": 3.1494661921708187, "grad_norm": 0.40128588676452637, "learning_rate": 9.56615669945108e-05, "loss": 0.0167, "step": 1770 }, { "epoch": 3.167259786476868, "grad_norm": 0.28565290570259094, "learning_rate": 9.55939482326173e-05, "loss": 0.0196, "step": 1780 }, { "epoch": 3.185053380782918, "grad_norm": 0.2395537942647934, "learning_rate": 9.552583086261069e-05, "loss": 0.0206, "step": 1790 }, { "epoch": 3.202846975088968, "grad_norm": 0.28574904799461365, "learning_rate": 9.545721562941168e-05, "loss": 0.0198, "step": 1800 }, { "epoch": 3.2206405693950177, "grad_norm": 0.35048720240592957, "learning_rate": 9.538810328338543e-05, "loss": 0.0142, "step": 1810 }, { "epoch": 3.2384341637010676, "grad_norm": 0.5313221216201782, "learning_rate": 9.531849458033349e-05, "loss": 0.0132, "step": 1820 }, { "epoch": 3.2562277580071175, "grad_norm": 0.3386707305908203, "learning_rate": 9.524839028148547e-05, "loss": 0.0161, "step": 1830 }, { "epoch": 3.2740213523131674, "grad_norm": 0.39070066809654236, "learning_rate": 9.517779115349077e-05, "loss": 0.0158, "step": 1840 }, { "epoch": 3.2918149466192173, "grad_norm": 0.4264601767063141, "learning_rate": 9.510669796841014e-05, "loss": 0.0147, "step": 1850 }, { "epoch": 3.309608540925267, "grad_norm": 0.45584696531295776, "learning_rate": 9.503511150370727e-05, "loss": 0.0183, "step": 1860 }, { "epoch": 3.3274021352313166, "grad_norm": 0.47540706396102905, "learning_rate": 9.496303254224024e-05, "loss": 0.0194, "step": 1870 }, { "epoch": 3.3451957295373664, "grad_norm": 0.4083940386772156, "learning_rate": 9.489046187225306e-05, "loss": 0.0161, "step": 1880 }, { "epoch": 3.3629893238434163, "grad_norm": 0.27993685007095337, "learning_rate": 9.481740028736692e-05, "loss": 0.0159, "step": 1890 }, { "epoch": 3.380782918149466, "grad_norm": 0.3538760244846344, "learning_rate": 9.474384858657164e-05, "loss": 0.0237, "step": 1900 }, { "epoch": 3.398576512455516, "grad_norm": 0.23415428400039673, "learning_rate": 9.466980757421679e-05, "loss": 0.0201, "step": 1910 }, { "epoch": 3.416370106761566, "grad_norm": 0.26716089248657227, "learning_rate": 9.459527806000305e-05, "loss": 0.0163, "step": 1920 }, { "epoch": 3.434163701067616, "grad_norm": 0.31686195731163025, "learning_rate": 9.452026085897325e-05, "loss": 0.0183, "step": 1930 }, { "epoch": 3.4519572953736652, "grad_norm": 0.4365226626396179, "learning_rate": 9.444475679150348e-05, "loss": 0.017, "step": 1940 }, { "epoch": 3.469750889679715, "grad_norm": 0.2929478585720062, "learning_rate": 9.436876668329411e-05, "loss": 0.0174, "step": 1950 }, { "epoch": 3.487544483985765, "grad_norm": 0.21563659608364105, "learning_rate": 9.429229136536079e-05, "loss": 0.0127, "step": 1960 }, { "epoch": 3.505338078291815, "grad_norm": 0.37227290868759155, "learning_rate": 9.421533167402534e-05, "loss": 0.0195, "step": 1970 }, { "epoch": 3.5231316725978647, "grad_norm": 0.3725384771823883, "learning_rate": 9.413788845090666e-05, "loss": 0.0213, "step": 1980 }, { "epoch": 3.5409252669039146, "grad_norm": 0.28063350915908813, "learning_rate": 9.405996254291136e-05, "loss": 0.017, "step": 1990 }, { "epoch": 3.5587188612099645, "grad_norm": 0.4616221785545349, "learning_rate": 9.398155480222474e-05, "loss": 0.0156, "step": 2000 }, { "epoch": 3.5765124555160144, "grad_norm": 0.24089166522026062, "learning_rate": 9.390266608630128e-05, "loss": 0.0126, "step": 2010 }, { "epoch": 3.5943060498220643, "grad_norm": 0.32936349511146545, "learning_rate": 9.38232972578553e-05, "loss": 0.0147, "step": 2020 }, { "epoch": 3.612099644128114, "grad_norm": 0.3568757176399231, "learning_rate": 9.374344918485164e-05, "loss": 0.0228, "step": 2030 }, { "epoch": 3.6298932384341636, "grad_norm": 0.3555702269077301, "learning_rate": 9.366312274049602e-05, "loss": 0.0201, "step": 2040 }, { "epoch": 3.6476868327402134, "grad_norm": 0.3663824200630188, "learning_rate": 9.358231880322554e-05, "loss": 0.0166, "step": 2050 }, { "epoch": 3.6654804270462633, "grad_norm": 0.27771082520484924, "learning_rate": 9.350103825669916e-05, "loss": 0.0133, "step": 2060 }, { "epoch": 3.683274021352313, "grad_norm": 0.3489212095737457, "learning_rate": 9.341928198978787e-05, "loss": 0.0121, "step": 2070 }, { "epoch": 3.701067615658363, "grad_norm": 0.2707537114620209, "learning_rate": 9.333705089656512e-05, "loss": 0.0148, "step": 2080 }, { "epoch": 3.718861209964413, "grad_norm": 0.3518145680427551, "learning_rate": 9.325434587629698e-05, "loss": 0.0157, "step": 2090 }, { "epoch": 3.7366548042704624, "grad_norm": 0.4352695345878601, "learning_rate": 9.31711678334323e-05, "loss": 0.0141, "step": 2100 }, { "epoch": 3.7544483985765122, "grad_norm": 0.37550631165504456, "learning_rate": 9.308751767759282e-05, "loss": 0.0128, "step": 2110 }, { "epoch": 3.772241992882562, "grad_norm": 0.3806985020637512, "learning_rate": 9.300339632356325e-05, "loss": 0.0128, "step": 2120 }, { "epoch": 3.790035587188612, "grad_norm": 0.3301794230937958, "learning_rate": 9.291880469128124e-05, "loss": 0.0171, "step": 2130 }, { "epoch": 3.807829181494662, "grad_norm": 0.33358946442604065, "learning_rate": 9.283374370582732e-05, "loss": 0.0171, "step": 2140 }, { "epoch": 3.8256227758007118, "grad_norm": 0.22197243571281433, "learning_rate": 9.274821429741482e-05, "loss": 0.0235, "step": 2150 }, { "epoch": 3.8434163701067616, "grad_norm": 0.38814395666122437, "learning_rate": 9.266221740137961e-05, "loss": 0.0211, "step": 2160 }, { "epoch": 3.8612099644128115, "grad_norm": 0.29497620463371277, "learning_rate": 9.257575395817001e-05, "loss": 0.016, "step": 2170 }, { "epoch": 3.8790035587188614, "grad_norm": 0.24132850766181946, "learning_rate": 9.248882491333637e-05, "loss": 0.0146, "step": 2180 }, { "epoch": 3.8967971530249113, "grad_norm": 0.21246039867401123, "learning_rate": 9.240143121752076e-05, "loss": 0.0168, "step": 2190 }, { "epoch": 3.914590747330961, "grad_norm": 0.2822973430156708, "learning_rate": 9.23135738264467e-05, "loss": 0.0143, "step": 2200 }, { "epoch": 3.9323843416370106, "grad_norm": 0.23143430054187775, "learning_rate": 9.222525370090849e-05, "loss": 0.0123, "step": 2210 }, { "epoch": 3.9501779359430604, "grad_norm": 0.41190528869628906, "learning_rate": 9.213647180676088e-05, "loss": 0.013, "step": 2220 }, { "epoch": 3.9679715302491103, "grad_norm": 0.3731071352958679, "learning_rate": 9.204722911490846e-05, "loss": 0.0139, "step": 2230 }, { "epoch": 3.98576512455516, "grad_norm": 0.36592987179756165, "learning_rate": 9.1957526601295e-05, "loss": 0.015, "step": 2240 }, { "epoch": 4.00355871886121, "grad_norm": 0.24950368702411652, "learning_rate": 9.186736524689281e-05, "loss": 0.0173, "step": 2250 }, { "epoch": 4.0213523131672595, "grad_norm": 0.4144071042537689, "learning_rate": 9.177674603769204e-05, "loss": 0.0153, "step": 2260 }, { "epoch": 4.039145907473309, "grad_norm": 0.30726197361946106, "learning_rate": 9.168566996468983e-05, "loss": 0.0156, "step": 2270 }, { "epoch": 4.056939501779359, "grad_norm": 0.9211875200271606, "learning_rate": 9.159413802387951e-05, "loss": 0.0197, "step": 2280 }, { "epoch": 4.074733096085409, "grad_norm": 0.25200194120407104, "learning_rate": 9.150215121623974e-05, "loss": 0.0196, "step": 2290 }, { "epoch": 4.092526690391459, "grad_norm": 0.4888349175453186, "learning_rate": 9.140971054772349e-05, "loss": 0.0114, "step": 2300 }, { "epoch": 4.110320284697509, "grad_norm": 0.36884981393814087, "learning_rate": 9.131681702924713e-05, "loss": 0.0193, "step": 2310 }, { "epoch": 4.128113879003559, "grad_norm": 0.3222210705280304, "learning_rate": 9.122347167667926e-05, "loss": 0.0141, "step": 2320 }, { "epoch": 4.145907473309609, "grad_norm": 0.2928904592990875, "learning_rate": 9.112967551082973e-05, "loss": 0.0209, "step": 2330 }, { "epoch": 4.1637010676156585, "grad_norm": 0.20041736960411072, "learning_rate": 9.103542955743835e-05, "loss": 0.0175, "step": 2340 }, { "epoch": 4.181494661921708, "grad_norm": 0.5178707242012024, "learning_rate": 9.094073484716381e-05, "loss": 0.0181, "step": 2350 }, { "epoch": 4.199288256227758, "grad_norm": 0.44691911339759827, "learning_rate": 9.084559241557226e-05, "loss": 0.0175, "step": 2360 }, { "epoch": 4.217081850533808, "grad_norm": 0.3252614140510559, "learning_rate": 9.075000330312608e-05, "loss": 0.0154, "step": 2370 }, { "epoch": 4.234875444839858, "grad_norm": 0.2034541368484497, "learning_rate": 9.065396855517253e-05, "loss": 0.0128, "step": 2380 }, { "epoch": 4.252669039145908, "grad_norm": 0.6165895462036133, "learning_rate": 9.055748922193219e-05, "loss": 0.0213, "step": 2390 }, { "epoch": 4.270462633451958, "grad_norm": 0.37853437662124634, "learning_rate": 9.046056635848761e-05, "loss": 0.0177, "step": 2400 }, { "epoch": 4.288256227758007, "grad_norm": 0.2273794710636139, "learning_rate": 9.036320102477169e-05, "loss": 0.0133, "step": 2410 }, { "epoch": 4.306049822064057, "grad_norm": 0.4300830364227295, "learning_rate": 9.02653942855561e-05, "loss": 0.015, "step": 2420 }, { "epoch": 4.3238434163701065, "grad_norm": 0.5935473442077637, "learning_rate": 9.016714721043971e-05, "loss": 0.0131, "step": 2430 }, { "epoch": 4.341637010676156, "grad_norm": 0.41647306084632874, "learning_rate": 9.006846087383675e-05, "loss": 0.0192, "step": 2440 }, { "epoch": 4.359430604982206, "grad_norm": 0.3729289174079895, "learning_rate": 8.996933635496523e-05, "loss": 0.0137, "step": 2450 }, { "epoch": 4.377224199288256, "grad_norm": 0.312363862991333, "learning_rate": 8.986977473783498e-05, "loss": 0.014, "step": 2460 }, { "epoch": 4.395017793594306, "grad_norm": 0.49942725896835327, "learning_rate": 8.97697771112359e-05, "loss": 0.0209, "step": 2470 }, { "epoch": 4.412811387900356, "grad_norm": 0.4198131263256073, "learning_rate": 8.966934456872602e-05, "loss": 0.0136, "step": 2480 }, { "epoch": 4.430604982206406, "grad_norm": 0.44993939995765686, "learning_rate": 8.95684782086195e-05, "loss": 0.0149, "step": 2490 }, { "epoch": 4.448398576512456, "grad_norm": 0.43005722761154175, "learning_rate": 8.946717913397476e-05, "loss": 0.0179, "step": 2500 }, { "epoch": 4.4661921708185055, "grad_norm": 0.3152318596839905, "learning_rate": 8.93654484525822e-05, "loss": 0.0149, "step": 2510 }, { "epoch": 4.483985765124555, "grad_norm": 0.2560085654258728, "learning_rate": 8.926328727695226e-05, "loss": 0.0145, "step": 2520 }, { "epoch": 4.501779359430605, "grad_norm": 0.3447585999965668, "learning_rate": 8.916069672430319e-05, "loss": 0.0132, "step": 2530 }, { "epoch": 4.519572953736655, "grad_norm": 0.24789801239967346, "learning_rate": 8.905767791654884e-05, "loss": 0.0166, "step": 2540 }, { "epoch": 4.537366548042705, "grad_norm": 0.6095348596572876, "learning_rate": 8.895423198028638e-05, "loss": 0.0217, "step": 2550 }, { "epoch": 4.555160142348754, "grad_norm": 0.4355863630771637, "learning_rate": 8.885036004678402e-05, "loss": 0.0184, "step": 2560 }, { "epoch": 4.572953736654805, "grad_norm": 0.3435424566268921, "learning_rate": 8.874606325196857e-05, "loss": 0.015, "step": 2570 }, { "epoch": 4.590747330960854, "grad_norm": 0.4828678071498871, "learning_rate": 8.864134273641304e-05, "loss": 0.0232, "step": 2580 }, { "epoch": 4.608540925266904, "grad_norm": 0.5187134146690369, "learning_rate": 8.853619964532427e-05, "loss": 0.0161, "step": 2590 }, { "epoch": 4.6263345195729535, "grad_norm": 0.5699332356452942, "learning_rate": 8.843063512853019e-05, "loss": 0.0148, "step": 2600 }, { "epoch": 4.644128113879003, "grad_norm": 0.33006542921066284, "learning_rate": 8.832465034046749e-05, "loss": 0.0154, "step": 2610 }, { "epoch": 4.661921708185053, "grad_norm": 0.42953774333000183, "learning_rate": 8.821824644016882e-05, "loss": 0.0148, "step": 2620 }, { "epoch": 4.679715302491103, "grad_norm": 0.23779122531414032, "learning_rate": 8.811142459125019e-05, "loss": 0.0119, "step": 2630 }, { "epoch": 4.697508896797153, "grad_norm": 0.1330065131187439, "learning_rate": 8.800418596189822e-05, "loss": 0.0108, "step": 2640 }, { "epoch": 4.715302491103203, "grad_norm": 0.26739972829818726, "learning_rate": 8.789653172485737e-05, "loss": 0.0115, "step": 2650 }, { "epoch": 4.733096085409253, "grad_norm": 0.2631899118423462, "learning_rate": 8.778846305741715e-05, "loss": 0.0217, "step": 2660 }, { "epoch": 4.750889679715303, "grad_norm": 0.4090069532394409, "learning_rate": 8.767998114139918e-05, "loss": 0.0175, "step": 2670 }, { "epoch": 4.7686832740213525, "grad_norm": 0.24645604193210602, "learning_rate": 8.757108716314429e-05, "loss": 0.0171, "step": 2680 }, { "epoch": 4.786476868327402, "grad_norm": 0.2608324885368347, "learning_rate": 8.746178231349962e-05, "loss": 0.0136, "step": 2690 }, { "epoch": 4.804270462633452, "grad_norm": 0.49711763858795166, "learning_rate": 8.735206778780549e-05, "loss": 0.0157, "step": 2700 }, { "epoch": 4.822064056939502, "grad_norm": 0.21173803508281708, "learning_rate": 8.724194478588234e-05, "loss": 0.0168, "step": 2710 }, { "epoch": 4.839857651245552, "grad_norm": 0.3102663457393646, "learning_rate": 8.713141451201772e-05, "loss": 0.0135, "step": 2720 }, { "epoch": 4.857651245551601, "grad_norm": 0.4581625461578369, "learning_rate": 8.702047817495295e-05, "loss": 0.0132, "step": 2730 }, { "epoch": 4.875444839857651, "grad_norm": 0.23893003165721893, "learning_rate": 8.69091369878701e-05, "loss": 0.0137, "step": 2740 }, { "epoch": 4.893238434163701, "grad_norm": 0.36692094802856445, "learning_rate": 8.679739216837849e-05, "loss": 0.0111, "step": 2750 }, { "epoch": 4.911032028469751, "grad_norm": 0.20873861014842987, "learning_rate": 8.66852449385016e-05, "loss": 0.0119, "step": 2760 }, { "epoch": 4.9288256227758005, "grad_norm": 0.4027778208255768, "learning_rate": 8.657269652466356e-05, "loss": 0.012, "step": 2770 }, { "epoch": 4.94661921708185, "grad_norm": 0.24668477475643158, "learning_rate": 8.645974815767577e-05, "loss": 0.0124, "step": 2780 }, { "epoch": 4.9644128113879, "grad_norm": 0.2619366943836212, "learning_rate": 8.634640107272351e-05, "loss": 0.0111, "step": 2790 }, { "epoch": 4.98220640569395, "grad_norm": 0.23273354768753052, "learning_rate": 8.623265650935234e-05, "loss": 0.0133, "step": 2800 }, { "epoch": 5.0, "grad_norm": 0.5402395129203796, "learning_rate": 8.611851571145456e-05, "loss": 0.015, "step": 2810 }, { "epoch": 5.01779359430605, "grad_norm": 0.3934841752052307, "learning_rate": 8.600397992725566e-05, "loss": 0.0105, "step": 2820 }, { "epoch": 5.0355871886121, "grad_norm": 0.31666693091392517, "learning_rate": 8.588905040930061e-05, "loss": 0.0135, "step": 2830 }, { "epoch": 5.05338078291815, "grad_norm": 0.4594133794307709, "learning_rate": 8.577372841444022e-05, "loss": 0.0163, "step": 2840 }, { "epoch": 5.0711743772241995, "grad_norm": 0.23171374201774597, "learning_rate": 8.565801520381736e-05, "loss": 0.0135, "step": 2850 }, { "epoch": 5.088967971530249, "grad_norm": 0.3793932795524597, "learning_rate": 8.554191204285313e-05, "loss": 0.0149, "step": 2860 }, { "epoch": 5.106761565836299, "grad_norm": 0.3394503593444824, "learning_rate": 8.542542020123315e-05, "loss": 0.0119, "step": 2870 }, { "epoch": 5.124555160142349, "grad_norm": 0.2711711823940277, "learning_rate": 8.530854095289347e-05, "loss": 0.0107, "step": 2880 }, { "epoch": 5.142348754448399, "grad_norm": 0.2837259769439697, "learning_rate": 8.519127557600688e-05, "loss": 0.0111, "step": 2890 }, { "epoch": 5.160142348754448, "grad_norm": 0.3499346673488617, "learning_rate": 8.507362535296871e-05, "loss": 0.0126, "step": 2900 }, { "epoch": 5.177935943060498, "grad_norm": 0.4011555016040802, "learning_rate": 8.495559157038299e-05, "loss": 0.0234, "step": 2910 }, { "epoch": 5.195729537366548, "grad_norm": 0.3094343841075897, "learning_rate": 8.483717551904823e-05, "loss": 0.0135, "step": 2920 }, { "epoch": 5.213523131672598, "grad_norm": 0.23437032103538513, "learning_rate": 8.47183784939434e-05, "loss": 0.0121, "step": 2930 }, { "epoch": 5.2313167259786475, "grad_norm": 0.35597503185272217, "learning_rate": 8.459920179421374e-05, "loss": 0.0145, "step": 2940 }, { "epoch": 5.249110320284697, "grad_norm": 0.2975935637950897, "learning_rate": 8.447964672315656e-05, "loss": 0.0121, "step": 2950 }, { "epoch": 5.266903914590747, "grad_norm": 0.3539317846298218, "learning_rate": 8.435971458820692e-05, "loss": 0.0173, "step": 2960 }, { "epoch": 5.284697508896797, "grad_norm": 0.2830093502998352, "learning_rate": 8.423940670092345e-05, "loss": 0.0151, "step": 2970 }, { "epoch": 5.302491103202847, "grad_norm": 0.3167116641998291, "learning_rate": 8.411872437697394e-05, "loss": 0.0126, "step": 2980 }, { "epoch": 5.320284697508897, "grad_norm": 0.3638388514518738, "learning_rate": 8.399766893612096e-05, "loss": 0.0095, "step": 2990 }, { "epoch": 5.338078291814947, "grad_norm": 0.3457958698272705, "learning_rate": 8.38762417022074e-05, "loss": 0.013, "step": 3000 }, { "epoch": 5.355871886120997, "grad_norm": 0.48569247126579285, "learning_rate": 8.375444400314204e-05, "loss": 0.0131, "step": 3010 }, { "epoch": 5.3736654804270465, "grad_norm": 0.426494836807251, "learning_rate": 8.3632277170885e-05, "loss": 0.0206, "step": 3020 }, { "epoch": 5.391459074733096, "grad_norm": 0.2801465690135956, "learning_rate": 8.350974254143318e-05, "loss": 0.0122, "step": 3030 }, { "epoch": 5.409252669039146, "grad_norm": 0.4153662919998169, "learning_rate": 8.338684145480566e-05, "loss": 0.0121, "step": 3040 }, { "epoch": 5.427046263345196, "grad_norm": 0.350985050201416, "learning_rate": 8.326357525502904e-05, "loss": 0.0147, "step": 3050 }, { "epoch": 5.444839857651246, "grad_norm": 0.36608701944351196, "learning_rate": 8.313994529012273e-05, "loss": 0.0175, "step": 3060 }, { "epoch": 5.462633451957295, "grad_norm": 0.44421863555908203, "learning_rate": 8.301595291208422e-05, "loss": 0.0167, "step": 3070 }, { "epoch": 5.480427046263345, "grad_norm": 0.18505853414535522, "learning_rate": 8.289159947687427e-05, "loss": 0.0138, "step": 3080 }, { "epoch": 5.498220640569395, "grad_norm": 0.2705553472042084, "learning_rate": 8.276688634440216e-05, "loss": 0.013, "step": 3090 }, { "epoch": 5.516014234875445, "grad_norm": 0.3491264879703522, "learning_rate": 8.26418148785107e-05, "loss": 0.0139, "step": 3100 }, { "epoch": 5.5338078291814945, "grad_norm": 0.3588181734085083, "learning_rate": 8.251638644696141e-05, "loss": 0.016, "step": 3110 }, { "epoch": 5.551601423487544, "grad_norm": 0.2247844785451889, "learning_rate": 8.23906024214195e-05, "loss": 0.015, "step": 3120 }, { "epoch": 5.569395017793594, "grad_norm": 0.3206322491168976, "learning_rate": 8.226446417743897e-05, "loss": 0.0137, "step": 3130 }, { "epoch": 5.587188612099644, "grad_norm": 0.3421669602394104, "learning_rate": 8.213797309444742e-05, "loss": 0.0151, "step": 3140 }, { "epoch": 5.604982206405694, "grad_norm": 0.21748651564121246, "learning_rate": 8.201113055573105e-05, "loss": 0.0177, "step": 3150 }, { "epoch": 5.622775800711744, "grad_norm": 0.3844865560531616, "learning_rate": 8.188393794841958e-05, "loss": 0.0108, "step": 3160 }, { "epoch": 5.640569395017794, "grad_norm": 0.23741115629673004, "learning_rate": 8.175639666347094e-05, "loss": 0.0192, "step": 3170 }, { "epoch": 5.658362989323844, "grad_norm": 0.41678863763809204, "learning_rate": 8.162850809565623e-05, "loss": 0.0148, "step": 3180 }, { "epoch": 5.6761565836298935, "grad_norm": 0.36877238750457764, "learning_rate": 8.150027364354431e-05, "loss": 0.0133, "step": 3190 }, { "epoch": 5.693950177935943, "grad_norm": 0.19399970769882202, "learning_rate": 8.137169470948662e-05, "loss": 0.0185, "step": 3200 }, { "epoch": 5.711743772241993, "grad_norm": 0.34747499227523804, "learning_rate": 8.124277269960179e-05, "loss": 0.0164, "step": 3210 }, { "epoch": 5.729537366548042, "grad_norm": 0.21102990210056305, "learning_rate": 8.111350902376023e-05, "loss": 0.0139, "step": 3220 }, { "epoch": 5.747330960854093, "grad_norm": 0.5020536184310913, "learning_rate": 8.098390509556883e-05, "loss": 0.0142, "step": 3230 }, { "epoch": 5.765124555160142, "grad_norm": 0.25244399905204773, "learning_rate": 8.085396233235536e-05, "loss": 0.0108, "step": 3240 }, { "epoch": 5.782918149466192, "grad_norm": 0.28897625207901, "learning_rate": 8.072368215515306e-05, "loss": 0.009, "step": 3250 }, { "epoch": 5.800711743772242, "grad_norm": 0.287878155708313, "learning_rate": 8.059306598868506e-05, "loss": 0.0216, "step": 3260 }, { "epoch": 5.818505338078292, "grad_norm": 0.249873548746109, "learning_rate": 8.046211526134888e-05, "loss": 0.0108, "step": 3270 }, { "epoch": 5.8362989323843415, "grad_norm": 0.297770619392395, "learning_rate": 8.033083140520065e-05, "loss": 0.0133, "step": 3280 }, { "epoch": 5.854092526690391, "grad_norm": 0.4646328389644623, "learning_rate": 8.019921585593962e-05, "loss": 0.0139, "step": 3290 }, { "epoch": 5.871886120996441, "grad_norm": 0.3265191614627838, "learning_rate": 8.006727005289232e-05, "loss": 0.0093, "step": 3300 }, { "epoch": 5.889679715302491, "grad_norm": 0.4838326871395111, "learning_rate": 7.993499543899692e-05, "loss": 0.013, "step": 3310 }, { "epoch": 5.907473309608541, "grad_norm": 0.34666940569877625, "learning_rate": 7.980239346078742e-05, "loss": 0.0121, "step": 3320 }, { "epoch": 5.925266903914591, "grad_norm": 0.2872714698314667, "learning_rate": 7.966946556837778e-05, "loss": 0.0115, "step": 3330 }, { "epoch": 5.943060498220641, "grad_norm": 0.3573853373527527, "learning_rate": 7.953621321544616e-05, "loss": 0.0125, "step": 3340 }, { "epoch": 5.960854092526691, "grad_norm": 0.1685630977153778, "learning_rate": 7.940263785921896e-05, "loss": 0.0161, "step": 3350 }, { "epoch": 5.9786476868327405, "grad_norm": 0.26834341883659363, "learning_rate": 7.926874096045482e-05, "loss": 0.0135, "step": 3360 }, { "epoch": 5.99644128113879, "grad_norm": 0.17882370948791504, "learning_rate": 7.913452398342881e-05, "loss": 0.0186, "step": 3370 }, { "epoch": 6.01423487544484, "grad_norm": 0.35554710030555725, "learning_rate": 7.89999883959163e-05, "loss": 0.0156, "step": 3380 }, { "epoch": 6.032028469750889, "grad_norm": 0.30835315585136414, "learning_rate": 7.886513566917687e-05, "loss": 0.0179, "step": 3390 }, { "epoch": 6.049822064056939, "grad_norm": 0.3074568510055542, "learning_rate": 7.872996727793838e-05, "loss": 0.0162, "step": 3400 }, { "epoch": 6.067615658362989, "grad_norm": 0.3852033019065857, "learning_rate": 7.859448470038069e-05, "loss": 0.0168, "step": 3410 }, { "epoch": 6.085409252669039, "grad_norm": 0.424195259809494, "learning_rate": 7.845868941811956e-05, "loss": 0.0135, "step": 3420 }, { "epoch": 6.103202846975089, "grad_norm": 0.28172874450683594, "learning_rate": 7.832258291619043e-05, "loss": 0.0119, "step": 3430 }, { "epoch": 6.120996441281139, "grad_norm": 0.26583024859428406, "learning_rate": 7.81861666830322e-05, "loss": 0.0115, "step": 3440 }, { "epoch": 6.1387900355871885, "grad_norm": 0.16795125603675842, "learning_rate": 7.804944221047097e-05, "loss": 0.0156, "step": 3450 }, { "epoch": 6.156583629893238, "grad_norm": 0.21690520644187927, "learning_rate": 7.791241099370364e-05, "loss": 0.0148, "step": 3460 }, { "epoch": 6.174377224199288, "grad_norm": 0.4513839781284332, "learning_rate": 7.777507453128163e-05, "loss": 0.0132, "step": 3470 }, { "epoch": 6.192170818505338, "grad_norm": 0.28137388825416565, "learning_rate": 7.763743432509451e-05, "loss": 0.0117, "step": 3480 }, { "epoch": 6.209964412811388, "grad_norm": 0.3489830195903778, "learning_rate": 7.749949188035353e-05, "loss": 0.0121, "step": 3490 }, { "epoch": 6.227758007117438, "grad_norm": 0.3131566047668457, "learning_rate": 7.736124870557516e-05, "loss": 0.0159, "step": 3500 }, { "epoch": 6.245551601423488, "grad_norm": 0.2563199996948242, "learning_rate": 7.722270631256459e-05, "loss": 0.0138, "step": 3510 }, { "epoch": 6.263345195729538, "grad_norm": 0.33804214000701904, "learning_rate": 7.708386621639925e-05, "loss": 0.0109, "step": 3520 }, { "epoch": 6.2811387900355875, "grad_norm": 0.3322550356388092, "learning_rate": 7.694472993541219e-05, "loss": 0.0151, "step": 3530 }, { "epoch": 6.298932384341637, "grad_norm": 0.4328867197036743, "learning_rate": 7.680529899117547e-05, "loss": 0.0141, "step": 3540 }, { "epoch": 6.316725978647687, "grad_norm": 0.35369932651519775, "learning_rate": 7.666557490848358e-05, "loss": 0.016, "step": 3550 }, { "epoch": 6.334519572953736, "grad_norm": 0.17900612950325012, "learning_rate": 7.65255592153367e-05, "loss": 0.0133, "step": 3560 }, { "epoch": 6.352313167259786, "grad_norm": 0.326280802488327, "learning_rate": 7.638525344292402e-05, "loss": 0.0145, "step": 3570 }, { "epoch": 6.370106761565836, "grad_norm": 0.3691059947013855, "learning_rate": 7.624465912560697e-05, "loss": 0.0115, "step": 3580 }, { "epoch": 6.387900355871886, "grad_norm": 0.31320035457611084, "learning_rate": 7.610377780090249e-05, "loss": 0.0133, "step": 3590 }, { "epoch": 6.405693950177936, "grad_norm": 0.3357943892478943, "learning_rate": 7.596261100946618e-05, "loss": 0.0133, "step": 3600 }, { "epoch": 6.423487544483986, "grad_norm": 0.31567659974098206, "learning_rate": 7.582116029507542e-05, "loss": 0.0125, "step": 3610 }, { "epoch": 6.4412811387900355, "grad_norm": 0.25522446632385254, "learning_rate": 7.56794272046126e-05, "loss": 0.0126, "step": 3620 }, { "epoch": 6.459074733096085, "grad_norm": 0.21432431042194366, "learning_rate": 7.55374132880481e-05, "loss": 0.0158, "step": 3630 }, { "epoch": 6.476868327402135, "grad_norm": 0.40561428666114807, "learning_rate": 7.539512009842333e-05, "loss": 0.0151, "step": 3640 }, { "epoch": 6.494661921708185, "grad_norm": 0.4114283621311188, "learning_rate": 7.525254919183382e-05, "loss": 0.0145, "step": 3650 }, { "epoch": 6.512455516014235, "grad_norm": 0.45906561613082886, "learning_rate": 7.510970212741215e-05, "loss": 0.0109, "step": 3660 }, { "epoch": 6.530249110320285, "grad_norm": 0.1999790519475937, "learning_rate": 7.496658046731096e-05, "loss": 0.0128, "step": 3670 }, { "epoch": 6.548042704626335, "grad_norm": 0.291906476020813, "learning_rate": 7.482318577668578e-05, "loss": 0.0117, "step": 3680 }, { "epoch": 6.565836298932385, "grad_norm": 0.23814351856708527, "learning_rate": 7.467951962367796e-05, "loss": 0.0171, "step": 3690 }, { "epoch": 6.5836298932384345, "grad_norm": 0.20695213973522186, "learning_rate": 7.453558357939755e-05, "loss": 0.0115, "step": 3700 }, { "epoch": 6.6014234875444835, "grad_norm": 0.2081184834241867, "learning_rate": 7.439137921790606e-05, "loss": 0.0132, "step": 3710 }, { "epoch": 6.619217081850534, "grad_norm": 0.19175657629966736, "learning_rate": 7.42469081161993e-05, "loss": 0.0159, "step": 3720 }, { "epoch": 6.637010676156583, "grad_norm": 0.2653181850910187, "learning_rate": 7.410217185419006e-05, "loss": 0.0099, "step": 3730 }, { "epoch": 6.654804270462633, "grad_norm": 0.23116546869277954, "learning_rate": 7.395717201469095e-05, "loss": 0.0126, "step": 3740 }, { "epoch": 6.672597864768683, "grad_norm": 0.3256244659423828, "learning_rate": 7.381191018339696e-05, "loss": 0.0154, "step": 3750 }, { "epoch": 6.690391459074733, "grad_norm": 0.12761980295181274, "learning_rate": 7.36663879488682e-05, "loss": 0.0107, "step": 3760 }, { "epoch": 6.708185053380783, "grad_norm": 0.2808552086353302, "learning_rate": 7.352060690251254e-05, "loss": 0.0177, "step": 3770 }, { "epoch": 6.725978647686833, "grad_norm": 0.25418514013290405, "learning_rate": 7.337456863856811e-05, "loss": 0.0136, "step": 3780 }, { "epoch": 6.7437722419928825, "grad_norm": 0.3004860579967499, "learning_rate": 7.3228274754086e-05, "loss": 0.0114, "step": 3790 }, { "epoch": 6.761565836298932, "grad_norm": 0.251720130443573, "learning_rate": 7.308172684891267e-05, "loss": 0.0134, "step": 3800 }, { "epoch": 6.779359430604982, "grad_norm": 0.19469593465328217, "learning_rate": 7.293492652567255e-05, "loss": 0.0136, "step": 3810 }, { "epoch": 6.797153024911032, "grad_norm": 0.19448034465312958, "learning_rate": 7.278787538975043e-05, "loss": 0.0147, "step": 3820 }, { "epoch": 6.814946619217082, "grad_norm": 0.2681104838848114, "learning_rate": 7.2640575049274e-05, "loss": 0.016, "step": 3830 }, { "epoch": 6.832740213523132, "grad_norm": 0.2622634768486023, "learning_rate": 7.249302711509616e-05, "loss": 0.0103, "step": 3840 }, { "epoch": 6.850533807829182, "grad_norm": 0.21388547122478485, "learning_rate": 7.23452332007775e-05, "loss": 0.0101, "step": 3850 }, { "epoch": 6.868327402135232, "grad_norm": 0.23964089155197144, "learning_rate": 7.219719492256858e-05, "loss": 0.0156, "step": 3860 }, { "epoch": 6.8861209964412815, "grad_norm": 0.24294117093086243, "learning_rate": 7.20489138993923e-05, "loss": 0.012, "step": 3870 }, { "epoch": 6.9039145907473305, "grad_norm": 0.23408496379852295, "learning_rate": 7.190039175282614e-05, "loss": 0.0148, "step": 3880 }, { "epoch": 6.921708185053381, "grad_norm": 0.2788211405277252, "learning_rate": 7.175163010708455e-05, "loss": 0.0096, "step": 3890 }, { "epoch": 6.93950177935943, "grad_norm": 0.22868235409259796, "learning_rate": 7.1602630589001e-05, "loss": 0.0166, "step": 3900 }, { "epoch": 6.95729537366548, "grad_norm": 0.3270765542984009, "learning_rate": 7.14533948280104e-05, "loss": 0.0116, "step": 3910 }, { "epoch": 6.97508896797153, "grad_norm": 0.1552087366580963, "learning_rate": 7.130392445613109e-05, "loss": 0.0121, "step": 3920 }, { "epoch": 6.99288256227758, "grad_norm": 0.3435005247592926, "learning_rate": 7.115422110794711e-05, "loss": 0.0111, "step": 3930 }, { "epoch": 7.01067615658363, "grad_norm": 0.30654627084732056, "learning_rate": 7.100428642059033e-05, "loss": 0.0146, "step": 3940 }, { "epoch": 7.02846975088968, "grad_norm": 0.18930654227733612, "learning_rate": 7.08541220337224e-05, "loss": 0.0135, "step": 3950 }, { "epoch": 7.0462633451957295, "grad_norm": 0.3636874258518219, "learning_rate": 7.070372958951706e-05, "loss": 0.0131, "step": 3960 }, { "epoch": 7.064056939501779, "grad_norm": 0.3094032108783722, "learning_rate": 7.055311073264194e-05, "loss": 0.0126, "step": 3970 }, { "epoch": 7.081850533807829, "grad_norm": 0.24257999658584595, "learning_rate": 7.040226711024077e-05, "loss": 0.0125, "step": 3980 }, { "epoch": 7.099644128113879, "grad_norm": 0.18561193346977234, "learning_rate": 7.02512003719152e-05, "loss": 0.0136, "step": 3990 }, { "epoch": 7.117437722419929, "grad_norm": 0.3578704297542572, "learning_rate": 7.00999121697069e-05, "loss": 0.0116, "step": 4000 }, { "epoch": 7.135231316725979, "grad_norm": 0.20149308443069458, "learning_rate": 6.99484041580794e-05, "loss": 0.012, "step": 4010 }, { "epoch": 7.153024911032029, "grad_norm": 0.28345271944999695, "learning_rate": 6.979667799390004e-05, "loss": 0.011, "step": 4020 }, { "epoch": 7.170818505338079, "grad_norm": 0.22687336802482605, "learning_rate": 6.964473533642185e-05, "loss": 0.0134, "step": 4030 }, { "epoch": 7.1886120996441285, "grad_norm": 0.3412753939628601, "learning_rate": 6.949257784726539e-05, "loss": 0.0173, "step": 4040 }, { "epoch": 7.2064056939501775, "grad_norm": 0.26786401867866516, "learning_rate": 6.934020719040056e-05, "loss": 0.0088, "step": 4050 }, { "epoch": 7.224199288256227, "grad_norm": 0.36585181951522827, "learning_rate": 6.918762503212848e-05, "loss": 0.0168, "step": 4060 }, { "epoch": 7.241992882562277, "grad_norm": 0.31331494450569153, "learning_rate": 6.903483304106319e-05, "loss": 0.014, "step": 4070 }, { "epoch": 7.259786476868327, "grad_norm": 0.21219559013843536, "learning_rate": 6.888183288811341e-05, "loss": 0.0116, "step": 4080 }, { "epoch": 7.277580071174377, "grad_norm": 0.27953457832336426, "learning_rate": 6.87286262464643e-05, "loss": 0.0122, "step": 4090 }, { "epoch": 7.295373665480427, "grad_norm": 0.24714434146881104, "learning_rate": 6.857521479155915e-05, "loss": 0.0127, "step": 4100 }, { "epoch": 7.313167259786477, "grad_norm": 0.2595514953136444, "learning_rate": 6.842160020108104e-05, "loss": 0.0139, "step": 4110 }, { "epoch": 7.330960854092527, "grad_norm": 0.3924713432788849, "learning_rate": 6.826778415493455e-05, "loss": 0.0114, "step": 4120 }, { "epoch": 7.3487544483985765, "grad_norm": 0.20177221298217773, "learning_rate": 6.811376833522729e-05, "loss": 0.0122, "step": 4130 }, { "epoch": 7.366548042704626, "grad_norm": 0.1646234393119812, "learning_rate": 6.795955442625159e-05, "loss": 0.0127, "step": 4140 }, { "epoch": 7.384341637010676, "grad_norm": 0.2915242612361908, "learning_rate": 6.780514411446608e-05, "loss": 0.0123, "step": 4150 }, { "epoch": 7.402135231316726, "grad_norm": 0.15717151761054993, "learning_rate": 6.765053908847716e-05, "loss": 0.0071, "step": 4160 }, { "epoch": 7.419928825622776, "grad_norm": 0.2122805118560791, "learning_rate": 6.749574103902064e-05, "loss": 0.0163, "step": 4170 }, { "epoch": 7.437722419928826, "grad_norm": 0.310825377702713, "learning_rate": 6.734075165894317e-05, "loss": 0.0146, "step": 4180 }, { "epoch": 7.455516014234876, "grad_norm": 0.20916928350925446, "learning_rate": 6.71855726431838e-05, "loss": 0.0177, "step": 4190 }, { "epoch": 7.473309608540926, "grad_norm": 0.23749059438705444, "learning_rate": 6.703020568875538e-05, "loss": 0.01, "step": 4200 }, { "epoch": 7.4911032028469755, "grad_norm": 0.3732492923736572, "learning_rate": 6.687465249472603e-05, "loss": 0.0115, "step": 4210 }, { "epoch": 7.5088967971530245, "grad_norm": 0.17961043119430542, "learning_rate": 6.671891476220055e-05, "loss": 0.0105, "step": 4220 }, { "epoch": 7.526690391459074, "grad_norm": 0.2919195294380188, "learning_rate": 6.656299419430183e-05, "loss": 0.0091, "step": 4230 }, { "epoch": 7.544483985765124, "grad_norm": 0.2580617666244507, "learning_rate": 6.640689249615223e-05, "loss": 0.0128, "step": 4240 }, { "epoch": 7.562277580071174, "grad_norm": 0.4530206322669983, "learning_rate": 6.625061137485491e-05, "loss": 0.015, "step": 4250 }, { "epoch": 7.580071174377224, "grad_norm": 0.31996697187423706, "learning_rate": 6.609415253947517e-05, "loss": 0.0174, "step": 4260 }, { "epoch": 7.597864768683274, "grad_norm": 0.22932684421539307, "learning_rate": 6.593751770102178e-05, "loss": 0.0136, "step": 4270 }, { "epoch": 7.615658362989324, "grad_norm": 0.2675830125808716, "learning_rate": 6.578070857242823e-05, "loss": 0.0094, "step": 4280 }, { "epoch": 7.633451957295374, "grad_norm": 0.3280281126499176, "learning_rate": 6.562372686853402e-05, "loss": 0.0109, "step": 4290 }, { "epoch": 7.6512455516014235, "grad_norm": 0.4405723512172699, "learning_rate": 6.546657430606593e-05, "loss": 0.0111, "step": 4300 }, { "epoch": 7.669039145907473, "grad_norm": 0.304009348154068, "learning_rate": 6.530925260361918e-05, "loss": 0.0111, "step": 4310 }, { "epoch": 7.686832740213523, "grad_norm": 0.39385437965393066, "learning_rate": 6.515176348163871e-05, "loss": 0.0142, "step": 4320 }, { "epoch": 7.704626334519573, "grad_norm": 0.39983391761779785, "learning_rate": 6.499410866240032e-05, "loss": 0.0113, "step": 4330 }, { "epoch": 7.722419928825623, "grad_norm": 0.30925992131233215, "learning_rate": 6.48362898699919e-05, "loss": 0.0099, "step": 4340 }, { "epoch": 7.740213523131673, "grad_norm": 0.2269662767648697, "learning_rate": 6.467830883029443e-05, "loss": 0.0102, "step": 4350 }, { "epoch": 7.758007117437723, "grad_norm": 0.22585412859916687, "learning_rate": 6.452016727096326e-05, "loss": 0.0104, "step": 4360 }, { "epoch": 7.775800711743772, "grad_norm": 0.22858357429504395, "learning_rate": 6.436186692140916e-05, "loss": 0.0105, "step": 4370 }, { "epoch": 7.7935943060498225, "grad_norm": 0.26793235540390015, "learning_rate": 6.420340951277938e-05, "loss": 0.0097, "step": 4380 }, { "epoch": 7.8113879003558715, "grad_norm": 0.22688648104667664, "learning_rate": 6.404479677793874e-05, "loss": 0.0121, "step": 4390 }, { "epoch": 7.829181494661921, "grad_norm": 0.3997623920440674, "learning_rate": 6.388603045145075e-05, "loss": 0.0161, "step": 4400 }, { "epoch": 7.846975088967971, "grad_norm": 0.2152387946844101, "learning_rate": 6.372711226955843e-05, "loss": 0.0135, "step": 4410 }, { "epoch": 7.864768683274021, "grad_norm": 0.20011481642723083, "learning_rate": 6.356804397016564e-05, "loss": 0.0131, "step": 4420 }, { "epoch": 7.882562277580071, "grad_norm": 0.2608214318752289, "learning_rate": 6.340882729281779e-05, "loss": 0.0165, "step": 4430 }, { "epoch": 7.900355871886121, "grad_norm": 0.34395548701286316, "learning_rate": 6.324946397868294e-05, "loss": 0.0151, "step": 4440 }, { "epoch": 7.918149466192171, "grad_norm": 0.20370933413505554, "learning_rate": 6.308995577053276e-05, "loss": 0.0109, "step": 4450 }, { "epoch": 7.935943060498221, "grad_norm": 0.2904776930809021, "learning_rate": 6.293030441272347e-05, "loss": 0.0118, "step": 4460 }, { "epoch": 7.9537366548042705, "grad_norm": 0.3407762944698334, "learning_rate": 6.277051165117677e-05, "loss": 0.0095, "step": 4470 }, { "epoch": 7.97153024911032, "grad_norm": 0.2798351049423218, "learning_rate": 6.261057923336064e-05, "loss": 0.0087, "step": 4480 }, { "epoch": 7.98932384341637, "grad_norm": 0.2502301335334778, "learning_rate": 6.245050890827042e-05, "loss": 0.0122, "step": 4490 }, { "epoch": 8.00711743772242, "grad_norm": 0.23939184844493866, "learning_rate": 6.229030242640952e-05, "loss": 0.0109, "step": 4500 }, { "epoch": 8.02491103202847, "grad_norm": 0.29999208450317383, "learning_rate": 6.212996153977037e-05, "loss": 0.0162, "step": 4510 }, { "epoch": 8.042704626334519, "grad_norm": 0.24488234519958496, "learning_rate": 6.196948800181523e-05, "loss": 0.0121, "step": 4520 }, { "epoch": 8.06049822064057, "grad_norm": 0.36686190962791443, "learning_rate": 6.180888356745695e-05, "loss": 0.0125, "step": 4530 }, { "epoch": 8.078291814946619, "grad_norm": 0.3249785602092743, "learning_rate": 6.164814999303995e-05, "loss": 0.0094, "step": 4540 }, { "epoch": 8.09608540925267, "grad_norm": 0.3318764567375183, "learning_rate": 6.148728903632081e-05, "loss": 0.0109, "step": 4550 }, { "epoch": 8.113879003558718, "grad_norm": 0.2541491687297821, "learning_rate": 6.132630245644921e-05, "loss": 0.0107, "step": 4560 }, { "epoch": 8.13167259786477, "grad_norm": 0.296268105506897, "learning_rate": 6.116519201394857e-05, "loss": 0.0114, "step": 4570 }, { "epoch": 8.149466192170818, "grad_norm": 0.19967666268348694, "learning_rate": 6.10039594706969e-05, "loss": 0.0147, "step": 4580 }, { "epoch": 8.167259786476869, "grad_norm": 0.34265848994255066, "learning_rate": 6.084260658990744e-05, "loss": 0.011, "step": 4590 }, { "epoch": 8.185053380782918, "grad_norm": 0.31476011872291565, "learning_rate": 6.068113513610943e-05, "loss": 0.0103, "step": 4600 }, { "epoch": 8.202846975088969, "grad_norm": 0.21745525300502777, "learning_rate": 6.0519546875128876e-05, "loss": 0.0135, "step": 4610 }, { "epoch": 8.220640569395018, "grad_norm": 0.3754664659500122, "learning_rate": 6.035784357406906e-05, "loss": 0.0123, "step": 4620 }, { "epoch": 8.238434163701069, "grad_norm": 0.2650909125804901, "learning_rate": 6.01960270012914e-05, "loss": 0.0166, "step": 4630 }, { "epoch": 8.256227758007118, "grad_norm": 0.23287297785282135, "learning_rate": 6.003409892639599e-05, "loss": 0.0095, "step": 4640 }, { "epoch": 8.274021352313166, "grad_norm": 0.1946777105331421, "learning_rate": 5.9872061120202336e-05, "loss": 0.0131, "step": 4650 }, { "epoch": 8.291814946619217, "grad_norm": 0.3016112148761749, "learning_rate": 5.9709915354729914e-05, "loss": 0.0183, "step": 4660 }, { "epoch": 8.309608540925266, "grad_norm": 0.19713705778121948, "learning_rate": 5.9547663403178824e-05, "loss": 0.0151, "step": 4670 }, { "epoch": 8.327402135231317, "grad_norm": 0.32332122325897217, "learning_rate": 5.9385307039910445e-05, "loss": 0.0103, "step": 4680 }, { "epoch": 8.345195729537366, "grad_norm": 0.19220270216464996, "learning_rate": 5.922284804042792e-05, "loss": 0.013, "step": 4690 }, { "epoch": 8.362989323843417, "grad_norm": 0.36306634545326233, "learning_rate": 5.906028818135687e-05, "loss": 0.0177, "step": 4700 }, { "epoch": 8.380782918149466, "grad_norm": 0.296560674905777, "learning_rate": 5.889762924042585e-05, "loss": 0.0123, "step": 4710 }, { "epoch": 8.398576512455517, "grad_norm": 0.15354299545288086, "learning_rate": 5.873487299644699e-05, "loss": 0.0094, "step": 4720 }, { "epoch": 8.416370106761565, "grad_norm": 0.20477841794490814, "learning_rate": 5.857202122929649e-05, "loss": 0.0136, "step": 4730 }, { "epoch": 8.434163701067616, "grad_norm": 0.24429598450660706, "learning_rate": 5.840907571989518e-05, "loss": 0.0095, "step": 4740 }, { "epoch": 8.451957295373665, "grad_norm": 0.14784900844097137, "learning_rate": 5.824603825018904e-05, "loss": 0.0096, "step": 4750 }, { "epoch": 8.469750889679716, "grad_norm": 0.21253588795661926, "learning_rate": 5.808291060312975e-05, "loss": 0.0106, "step": 4760 }, { "epoch": 8.487544483985765, "grad_norm": 0.29630398750305176, "learning_rate": 5.7919694562655083e-05, "loss": 0.0145, "step": 4770 }, { "epoch": 8.505338078291816, "grad_norm": 0.3592516779899597, "learning_rate": 5.775639191366954e-05, "loss": 0.0127, "step": 4780 }, { "epoch": 8.523131672597865, "grad_norm": 0.14543798565864563, "learning_rate": 5.75930044420247e-05, "loss": 0.0105, "step": 4790 }, { "epoch": 8.540925266903916, "grad_norm": 0.28967422246932983, "learning_rate": 5.74295339344998e-05, "loss": 0.0128, "step": 4800 }, { "epoch": 8.558718861209965, "grad_norm": 0.21376687288284302, "learning_rate": 5.726598217878211e-05, "loss": 0.0087, "step": 4810 }, { "epoch": 8.576512455516013, "grad_norm": 0.11675862967967987, "learning_rate": 5.71023509634474e-05, "loss": 0.0093, "step": 4820 }, { "epoch": 8.594306049822064, "grad_norm": 0.21051128208637238, "learning_rate": 5.693864207794049e-05, "loss": 0.0116, "step": 4830 }, { "epoch": 8.612099644128113, "grad_norm": 0.4002830386161804, "learning_rate": 5.677485731255545e-05, "loss": 0.0208, "step": 4840 }, { "epoch": 8.629893238434164, "grad_norm": 0.3562188744544983, "learning_rate": 5.6610998458416296e-05, "loss": 0.0125, "step": 4850 }, { "epoch": 8.647686832740213, "grad_norm": 0.37460166215896606, "learning_rate": 5.644706730745716e-05, "loss": 0.0112, "step": 4860 }, { "epoch": 8.665480427046264, "grad_norm": 0.2684260606765747, "learning_rate": 5.628306565240287e-05, "loss": 0.0132, "step": 4870 }, { "epoch": 8.683274021352313, "grad_norm": 0.3357107639312744, "learning_rate": 5.611899528674923e-05, "loss": 0.0137, "step": 4880 }, { "epoch": 8.701067615658364, "grad_norm": 0.38067126274108887, "learning_rate": 5.595485800474349e-05, "loss": 0.0103, "step": 4890 }, { "epoch": 8.718861209964412, "grad_norm": 0.2616477310657501, "learning_rate": 5.579065560136467e-05, "loss": 0.0129, "step": 4900 }, { "epoch": 8.736654804270463, "grad_norm": 0.20944181084632874, "learning_rate": 5.562638987230392e-05, "loss": 0.012, "step": 4910 }, { "epoch": 8.754448398576512, "grad_norm": 0.24873778223991394, "learning_rate": 5.546206261394498e-05, "loss": 0.0077, "step": 4920 }, { "epoch": 8.772241992882563, "grad_norm": 0.34095433354377747, "learning_rate": 5.529767562334437e-05, "loss": 0.0109, "step": 4930 }, { "epoch": 8.790035587188612, "grad_norm": 0.3209082782268524, "learning_rate": 5.5133230698211926e-05, "loss": 0.0114, "step": 4940 }, { "epoch": 8.807829181494661, "grad_norm": 0.2815232574939728, "learning_rate": 5.496872963689096e-05, "loss": 0.0149, "step": 4950 }, { "epoch": 8.825622775800712, "grad_norm": 0.1784064769744873, "learning_rate": 5.4804174238338756e-05, "loss": 0.0094, "step": 4960 }, { "epoch": 8.84341637010676, "grad_norm": 0.19197413325309753, "learning_rate": 5.463956630210678e-05, "loss": 0.0128, "step": 4970 }, { "epoch": 8.861209964412812, "grad_norm": 0.21905569732189178, "learning_rate": 5.4474907628321046e-05, "loss": 0.0182, "step": 4980 }, { "epoch": 8.87900355871886, "grad_norm": 0.16814273595809937, "learning_rate": 5.431020001766244e-05, "loss": 0.0101, "step": 4990 }, { "epoch": 8.896797153024911, "grad_norm": 0.37039893865585327, "learning_rate": 5.4145445271346986e-05, "loss": 0.0117, "step": 5000 }, { "epoch": 8.91459074733096, "grad_norm": 0.32573556900024414, "learning_rate": 5.398064519110622e-05, "loss": 0.0116, "step": 5010 }, { "epoch": 8.932384341637011, "grad_norm": 0.22297902405261993, "learning_rate": 5.3815801579167394e-05, "loss": 0.0143, "step": 5020 }, { "epoch": 8.95017793594306, "grad_norm": 0.44576072692871094, "learning_rate": 5.365091623823382e-05, "loss": 0.0144, "step": 5030 }, { "epoch": 8.96797153024911, "grad_norm": 0.26499807834625244, "learning_rate": 5.348599097146521e-05, "loss": 0.0111, "step": 5040 }, { "epoch": 8.98576512455516, "grad_norm": 0.21782799065113068, "learning_rate": 5.3321027582457836e-05, "loss": 0.0127, "step": 5050 }, { "epoch": 9.00355871886121, "grad_norm": 0.22396863996982574, "learning_rate": 5.315602787522491e-05, "loss": 0.0084, "step": 5060 }, { "epoch": 9.02135231316726, "grad_norm": 0.14507652819156647, "learning_rate": 5.299099365417678e-05, "loss": 0.0072, "step": 5070 }, { "epoch": 9.03914590747331, "grad_norm": 0.19003397226333618, "learning_rate": 5.2825926724101236e-05, "loss": 0.0119, "step": 5080 }, { "epoch": 9.05693950177936, "grad_norm": 0.17321667075157166, "learning_rate": 5.26608288901438e-05, "loss": 0.0095, "step": 5090 }, { "epoch": 9.07473309608541, "grad_norm": 0.28541091084480286, "learning_rate": 5.24957019577879e-05, "loss": 0.014, "step": 5100 }, { "epoch": 9.092526690391459, "grad_norm": 0.1327798217535019, "learning_rate": 5.2330547732835266e-05, "loss": 0.0087, "step": 5110 }, { "epoch": 9.11032028469751, "grad_norm": 0.2860845625400543, "learning_rate": 5.2165368021385996e-05, "loss": 0.0185, "step": 5120 }, { "epoch": 9.128113879003559, "grad_norm": 0.17205515503883362, "learning_rate": 5.200016462981897e-05, "loss": 0.0113, "step": 5130 }, { "epoch": 9.145907473309608, "grad_norm": 0.2530641257762909, "learning_rate": 5.1834939364772015e-05, "loss": 0.0104, "step": 5140 }, { "epoch": 9.163701067615659, "grad_norm": 0.2918217182159424, "learning_rate": 5.166969403312214e-05, "loss": 0.0088, "step": 5150 }, { "epoch": 9.181494661921707, "grad_norm": 0.22969265282154083, "learning_rate": 5.1504430441965844e-05, "loss": 0.0133, "step": 5160 }, { "epoch": 9.199288256227758, "grad_norm": 0.2440153956413269, "learning_rate": 5.133915039859923e-05, "loss": 0.0086, "step": 5170 }, { "epoch": 9.217081850533807, "grad_norm": 0.1535383015871048, "learning_rate": 5.1173855710498444e-05, "loss": 0.0092, "step": 5180 }, { "epoch": 9.234875444839858, "grad_norm": 0.23259951174259186, "learning_rate": 5.100854818529967e-05, "loss": 0.0122, "step": 5190 }, { "epoch": 9.252669039145907, "grad_norm": 0.29800716042518616, "learning_rate": 5.084322963077951e-05, "loss": 0.0145, "step": 5200 }, { "epoch": 9.270462633451958, "grad_norm": 0.2709343433380127, "learning_rate": 5.067790185483522e-05, "loss": 0.0109, "step": 5210 }, { "epoch": 9.288256227758007, "grad_norm": 0.3267824649810791, "learning_rate": 5.0512566665464844e-05, "loss": 0.0124, "step": 5220 }, { "epoch": 9.306049822064058, "grad_norm": 0.17874538898468018, "learning_rate": 5.034722587074755e-05, "loss": 0.0116, "step": 5230 }, { "epoch": 9.323843416370106, "grad_norm": 0.3466548025608063, "learning_rate": 5.018188127882375e-05, "loss": 0.0102, "step": 5240 }, { "epoch": 9.341637010676157, "grad_norm": 0.34574735164642334, "learning_rate": 5.0016534697875417e-05, "loss": 0.0112, "step": 5250 }, { "epoch": 9.359430604982206, "grad_norm": 0.1662270426750183, "learning_rate": 4.9851187936106294e-05, "loss": 0.0102, "step": 5260 }, { "epoch": 9.377224199288257, "grad_norm": 0.22003521025180817, "learning_rate": 4.968584280172206e-05, "loss": 0.015, "step": 5270 }, { "epoch": 9.395017793594306, "grad_norm": 0.2559777498245239, "learning_rate": 4.95205011029106e-05, "loss": 0.0097, "step": 5280 }, { "epoch": 9.412811387900355, "grad_norm": 0.39615070819854736, "learning_rate": 4.935516464782227e-05, "loss": 0.0116, "step": 5290 }, { "epoch": 9.430604982206406, "grad_norm": 0.24687030911445618, "learning_rate": 4.918983524455003e-05, "loss": 0.0102, "step": 5300 }, { "epoch": 9.448398576512455, "grad_norm": 0.29982423782348633, "learning_rate": 4.9024514701109766e-05, "loss": 0.0116, "step": 5310 }, { "epoch": 9.466192170818506, "grad_norm": 0.24275128543376923, "learning_rate": 4.885920482542043e-05, "loss": 0.011, "step": 5320 }, { "epoch": 9.483985765124554, "grad_norm": 0.23122717440128326, "learning_rate": 4.869390742528438e-05, "loss": 0.0093, "step": 5330 }, { "epoch": 9.501779359430605, "grad_norm": 0.3967891335487366, "learning_rate": 4.852862430836744e-05, "loss": 0.0135, "step": 5340 }, { "epoch": 9.519572953736654, "grad_norm": 0.3189123570919037, "learning_rate": 4.836335728217933e-05, "loss": 0.0113, "step": 5350 }, { "epoch": 9.537366548042705, "grad_norm": 0.2779250741004944, "learning_rate": 4.819810815405379e-05, "loss": 0.0146, "step": 5360 }, { "epoch": 9.555160142348754, "grad_norm": 0.3087226450443268, "learning_rate": 4.803287873112877e-05, "loss": 0.0109, "step": 5370 }, { "epoch": 9.572953736654805, "grad_norm": 0.6251241564750671, "learning_rate": 4.786767082032681e-05, "loss": 0.0143, "step": 5380 }, { "epoch": 9.590747330960854, "grad_norm": 0.34025272727012634, "learning_rate": 4.77024862283351e-05, "loss": 0.0082, "step": 5390 }, { "epoch": 9.608540925266905, "grad_norm": 0.47951599955558777, "learning_rate": 4.753732676158593e-05, "loss": 0.0129, "step": 5400 }, { "epoch": 9.626334519572953, "grad_norm": 0.4520055651664734, "learning_rate": 4.737219422623672e-05, "loss": 0.0158, "step": 5410 }, { "epoch": 9.644128113879004, "grad_norm": 0.20497407019138336, "learning_rate": 4.720709042815044e-05, "loss": 0.0095, "step": 5420 }, { "epoch": 9.661921708185053, "grad_norm": 0.26534706354141235, "learning_rate": 4.704201717287578e-05, "loss": 0.0094, "step": 5430 }, { "epoch": 9.679715302491104, "grad_norm": 0.21294786036014557, "learning_rate": 4.6876976265627404e-05, "loss": 0.0106, "step": 5440 }, { "epoch": 9.697508896797153, "grad_norm": 0.2418232560157776, "learning_rate": 4.671196951126626e-05, "loss": 0.0161, "step": 5450 }, { "epoch": 9.715302491103202, "grad_norm": 0.26775938272476196, "learning_rate": 4.654699871427971e-05, "loss": 0.0089, "step": 5460 }, { "epoch": 9.733096085409253, "grad_norm": 0.26764431595802307, "learning_rate": 4.6382065678762034e-05, "loss": 0.0141, "step": 5470 }, { "epoch": 9.750889679715302, "grad_norm": 0.30992841720581055, "learning_rate": 4.6217172208394424e-05, "loss": 0.008, "step": 5480 }, { "epoch": 9.768683274021353, "grad_norm": 0.18100039660930634, "learning_rate": 4.605232010642549e-05, "loss": 0.0086, "step": 5490 }, { "epoch": 9.786476868327401, "grad_norm": 0.12815871834754944, "learning_rate": 4.588751117565142e-05, "loss": 0.0067, "step": 5500 }, { "epoch": 9.804270462633452, "grad_norm": 0.19295573234558105, "learning_rate": 4.5722747218396214e-05, "loss": 0.0122, "step": 5510 }, { "epoch": 9.822064056939501, "grad_norm": 0.19486500322818756, "learning_rate": 4.5558030036492194e-05, "loss": 0.0084, "step": 5520 }, { "epoch": 9.839857651245552, "grad_norm": 0.2748246192932129, "learning_rate": 4.539336143125999e-05, "loss": 0.0086, "step": 5530 }, { "epoch": 9.857651245551601, "grad_norm": 0.22238148748874664, "learning_rate": 4.522874320348916e-05, "loss": 0.0146, "step": 5540 }, { "epoch": 9.875444839857652, "grad_norm": 0.32224124670028687, "learning_rate": 4.506417715341821e-05, "loss": 0.0095, "step": 5550 }, { "epoch": 9.8932384341637, "grad_norm": 0.10665499418973923, "learning_rate": 4.489966508071511e-05, "loss": 0.0077, "step": 5560 }, { "epoch": 9.911032028469752, "grad_norm": 0.16376017034053802, "learning_rate": 4.4735208784457575e-05, "loss": 0.0082, "step": 5570 }, { "epoch": 9.9288256227758, "grad_norm": 0.282703161239624, "learning_rate": 4.457081006311325e-05, "loss": 0.0093, "step": 5580 }, { "epoch": 9.946619217081851, "grad_norm": 0.12352361530065536, "learning_rate": 4.440647071452027e-05, "loss": 0.0117, "step": 5590 }, { "epoch": 9.9644128113879, "grad_norm": 0.25517991185188293, "learning_rate": 4.424219253586737e-05, "loss": 0.0105, "step": 5600 }, { "epoch": 9.982206405693951, "grad_norm": 0.3717750012874603, "learning_rate": 4.407797732367443e-05, "loss": 0.015, "step": 5610 }, { "epoch": 10.0, "grad_norm": 0.2974468469619751, "learning_rate": 4.391382687377268e-05, "loss": 0.0105, "step": 5620 }, { "epoch": 10.017793594306049, "grad_norm": 0.24511711299419403, "learning_rate": 4.374974298128512e-05, "loss": 0.0109, "step": 5630 }, { "epoch": 10.0355871886121, "grad_norm": 0.2662994861602783, "learning_rate": 4.358572744060699e-05, "loss": 0.0099, "step": 5640 }, { "epoch": 10.053380782918149, "grad_norm": 0.1978130042552948, "learning_rate": 4.342178204538588e-05, "loss": 0.0114, "step": 5650 }, { "epoch": 10.0711743772242, "grad_norm": 0.1516188681125641, "learning_rate": 4.325790858850241e-05, "loss": 0.0062, "step": 5660 }, { "epoch": 10.088967971530248, "grad_norm": 0.2778376042842865, "learning_rate": 4.309410886205043e-05, "loss": 0.0105, "step": 5670 }, { "epoch": 10.1067615658363, "grad_norm": 0.16069914400577545, "learning_rate": 4.293038465731752e-05, "loss": 0.0082, "step": 5680 }, { "epoch": 10.124555160142348, "grad_norm": 0.1466551274061203, "learning_rate": 4.276673776476533e-05, "loss": 0.0099, "step": 5690 }, { "epoch": 10.142348754448399, "grad_norm": 0.2549195885658264, "learning_rate": 4.260316997401007e-05, "loss": 0.0106, "step": 5700 }, { "epoch": 10.160142348754448, "grad_norm": 0.3441568613052368, "learning_rate": 4.243968307380293e-05, "loss": 0.011, "step": 5710 }, { "epoch": 10.177935943060499, "grad_norm": 0.21854907274246216, "learning_rate": 4.22762788520104e-05, "loss": 0.0144, "step": 5720 }, { "epoch": 10.195729537366548, "grad_norm": 0.3372473418712616, "learning_rate": 4.211295909559491e-05, "loss": 0.0142, "step": 5730 }, { "epoch": 10.213523131672599, "grad_norm": 0.26563113927841187, "learning_rate": 4.194972559059511e-05, "loss": 0.008, "step": 5740 }, { "epoch": 10.231316725978647, "grad_norm": 0.2662186920642853, "learning_rate": 4.178658012210651e-05, "loss": 0.0097, "step": 5750 }, { "epoch": 10.249110320284698, "grad_norm": 0.2725781798362732, "learning_rate": 4.162352447426177e-05, "loss": 0.0093, "step": 5760 }, { "epoch": 10.266903914590747, "grad_norm": 0.1814977526664734, "learning_rate": 4.146056043021135e-05, "loss": 0.0137, "step": 5770 }, { "epoch": 10.284697508896798, "grad_norm": 0.3420964777469635, "learning_rate": 4.1297689772103944e-05, "loss": 0.0121, "step": 5780 }, { "epoch": 10.302491103202847, "grad_norm": 0.14642766118049622, "learning_rate": 4.113491428106694e-05, "loss": 0.0113, "step": 5790 }, { "epoch": 10.320284697508896, "grad_norm": 0.26963359117507935, "learning_rate": 4.0972235737187055e-05, "loss": 0.0106, "step": 5800 }, { "epoch": 10.338078291814947, "grad_norm": 0.26372790336608887, "learning_rate": 4.080965591949076e-05, "loss": 0.008, "step": 5810 }, { "epoch": 10.355871886120996, "grad_norm": 0.17928752303123474, "learning_rate": 4.0647176605924924e-05, "loss": 0.0086, "step": 5820 }, { "epoch": 10.373665480427047, "grad_norm": 0.3506091237068176, "learning_rate": 4.0484799573337255e-05, "loss": 0.0113, "step": 5830 }, { "epoch": 10.391459074733095, "grad_norm": 0.18861903250217438, "learning_rate": 4.032252659745699e-05, "loss": 0.0107, "step": 5840 }, { "epoch": 10.409252669039146, "grad_norm": 0.26331573724746704, "learning_rate": 4.016035945287539e-05, "loss": 0.0124, "step": 5850 }, { "epoch": 10.427046263345195, "grad_norm": 0.11781974136829376, "learning_rate": 3.999829991302635e-05, "loss": 0.0133, "step": 5860 }, { "epoch": 10.444839857651246, "grad_norm": 0.24894948303699493, "learning_rate": 3.983634975016707e-05, "loss": 0.0086, "step": 5870 }, { "epoch": 10.462633451957295, "grad_norm": 0.2436850666999817, "learning_rate": 3.967451073535854e-05, "loss": 0.0123, "step": 5880 }, { "epoch": 10.480427046263346, "grad_norm": 0.1779191941022873, "learning_rate": 3.951278463844633e-05, "loss": 0.0112, "step": 5890 }, { "epoch": 10.498220640569395, "grad_norm": 0.3594033718109131, "learning_rate": 3.935117322804111e-05, "loss": 0.0089, "step": 5900 }, { "epoch": 10.516014234875446, "grad_norm": 0.18209725618362427, "learning_rate": 3.918967827149938e-05, "loss": 0.0106, "step": 5910 }, { "epoch": 10.533807829181494, "grad_norm": 0.17099666595458984, "learning_rate": 3.9028301534904094e-05, "loss": 0.0108, "step": 5920 }, { "epoch": 10.551601423487545, "grad_norm": 0.35403206944465637, "learning_rate": 3.88670447830454e-05, "loss": 0.0101, "step": 5930 }, { "epoch": 10.569395017793594, "grad_norm": 0.259818434715271, "learning_rate": 3.870590977940132e-05, "loss": 0.008, "step": 5940 }, { "epoch": 10.587188612099645, "grad_norm": 0.2787943482398987, "learning_rate": 3.8544898286118404e-05, "loss": 0.0075, "step": 5950 }, { "epoch": 10.604982206405694, "grad_norm": 0.2341775745153427, "learning_rate": 3.838401206399257e-05, "loss": 0.0082, "step": 5960 }, { "epoch": 10.622775800711743, "grad_norm": 0.3597370982170105, "learning_rate": 3.822325287244975e-05, "loss": 0.0082, "step": 5970 }, { "epoch": 10.640569395017794, "grad_norm": 0.3382866680622101, "learning_rate": 3.8062622469526725e-05, "loss": 0.009, "step": 5980 }, { "epoch": 10.658362989323843, "grad_norm": 0.33738914132118225, "learning_rate": 3.790212261185183e-05, "loss": 0.0077, "step": 5990 }, { "epoch": 10.676156583629894, "grad_norm": 0.43008390069007874, "learning_rate": 3.7741755054625794e-05, "loss": 0.011, "step": 6000 }, { "epoch": 10.693950177935942, "grad_norm": 0.1997474730014801, "learning_rate": 3.758152155160255e-05, "loss": 0.0089, "step": 6010 }, { "epoch": 10.711743772241993, "grad_norm": 0.1688026487827301, "learning_rate": 3.742142385506999e-05, "loss": 0.0097, "step": 6020 }, { "epoch": 10.729537366548042, "grad_norm": 0.1905738264322281, "learning_rate": 3.72614637158309e-05, "loss": 0.0088, "step": 6030 }, { "epoch": 10.747330960854093, "grad_norm": 0.1853916496038437, "learning_rate": 3.710164288318371e-05, "loss": 0.0121, "step": 6040 }, { "epoch": 10.765124555160142, "grad_norm": 0.23017893731594086, "learning_rate": 3.694196310490345e-05, "loss": 0.0062, "step": 6050 }, { "epoch": 10.782918149466193, "grad_norm": 0.21578657627105713, "learning_rate": 3.678242612722259e-05, "loss": 0.008, "step": 6060 }, { "epoch": 10.800711743772242, "grad_norm": 0.19698543846607208, "learning_rate": 3.6623033694811953e-05, "loss": 0.0071, "step": 6070 }, { "epoch": 10.818505338078293, "grad_norm": 0.1377231329679489, "learning_rate": 3.6463787550761665e-05, "loss": 0.009, "step": 6080 }, { "epoch": 10.836298932384341, "grad_norm": 0.17074494063854218, "learning_rate": 3.630468943656202e-05, "loss": 0.0086, "step": 6090 }, { "epoch": 10.854092526690392, "grad_norm": 0.16820582747459412, "learning_rate": 3.6145741092084523e-05, "loss": 0.0071, "step": 6100 }, { "epoch": 10.871886120996441, "grad_norm": 0.27742069959640503, "learning_rate": 3.598694425556278e-05, "loss": 0.007, "step": 6110 }, { "epoch": 10.889679715302492, "grad_norm": 0.18640372157096863, "learning_rate": 3.58283006635736e-05, "loss": 0.0088, "step": 6120 }, { "epoch": 10.907473309608541, "grad_norm": 0.19590039551258087, "learning_rate": 3.566981205101781e-05, "loss": 0.0072, "step": 6130 }, { "epoch": 10.92526690391459, "grad_norm": 0.24782133102416992, "learning_rate": 3.5511480151101556e-05, "loss": 0.0086, "step": 6140 }, { "epoch": 10.94306049822064, "grad_norm": 0.1208200454711914, "learning_rate": 3.5353306695317104e-05, "loss": 0.0071, "step": 6150 }, { "epoch": 10.96085409252669, "grad_norm": 0.19201305508613586, "learning_rate": 3.519529341342402e-05, "loss": 0.0091, "step": 6160 }, { "epoch": 10.97864768683274, "grad_norm": 0.3206014037132263, "learning_rate": 3.503744203343026e-05, "loss": 0.0096, "step": 6170 }, { "epoch": 10.99644128113879, "grad_norm": 0.25889068841934204, "learning_rate": 3.487975428157318e-05, "loss": 0.0114, "step": 6180 }, { "epoch": 11.01423487544484, "grad_norm": 0.2371494472026825, "learning_rate": 3.472223188230083e-05, "loss": 0.0094, "step": 6190 }, { "epoch": 11.03202846975089, "grad_norm": 0.26126569509506226, "learning_rate": 3.4564876558252866e-05, "loss": 0.0124, "step": 6200 }, { "epoch": 11.04982206405694, "grad_norm": 0.2658976912498474, "learning_rate": 3.440769003024195e-05, "loss": 0.0077, "step": 6210 }, { "epoch": 11.067615658362989, "grad_norm": 0.21916937828063965, "learning_rate": 3.425067401723477e-05, "loss": 0.0089, "step": 6220 }, { "epoch": 11.08540925266904, "grad_norm": 0.2990090847015381, "learning_rate": 3.409383023633325e-05, "loss": 0.0127, "step": 6230 }, { "epoch": 11.103202846975089, "grad_norm": 0.2553524374961853, "learning_rate": 3.3937160402755894e-05, "loss": 0.0065, "step": 6240 }, { "epoch": 11.12099644128114, "grad_norm": 0.3186771869659424, "learning_rate": 3.378066622981885e-05, "loss": 0.009, "step": 6250 }, { "epoch": 11.138790035587188, "grad_norm": 0.19516108930110931, "learning_rate": 3.362434942891738e-05, "loss": 0.0116, "step": 6260 }, { "epoch": 11.15658362989324, "grad_norm": 0.18056991696357727, "learning_rate": 3.346821170950693e-05, "loss": 0.0113, "step": 6270 }, { "epoch": 11.174377224199288, "grad_norm": 0.16609880328178406, "learning_rate": 3.3312254779084585e-05, "loss": 0.0105, "step": 6280 }, { "epoch": 11.192170818505337, "grad_norm": 0.2682517468929291, "learning_rate": 3.315648034317039e-05, "loss": 0.0121, "step": 6290 }, { "epoch": 11.209964412811388, "grad_norm": 0.2861747145652771, "learning_rate": 3.3000890105288564e-05, "loss": 0.0085, "step": 6300 }, { "epoch": 11.227758007117437, "grad_norm": 0.13003288209438324, "learning_rate": 3.284548576694908e-05, "loss": 0.0082, "step": 6310 }, { "epoch": 11.245551601423488, "grad_norm": 0.18294209241867065, "learning_rate": 3.2690269027628815e-05, "loss": 0.007, "step": 6320 }, { "epoch": 11.263345195729537, "grad_norm": 0.15584713220596313, "learning_rate": 3.253524158475324e-05, "loss": 0.0085, "step": 6330 }, { "epoch": 11.281138790035588, "grad_norm": 0.2869352400302887, "learning_rate": 3.238040513367757e-05, "loss": 0.0118, "step": 6340 }, { "epoch": 11.298932384341636, "grad_norm": 0.3261018395423889, "learning_rate": 3.222576136766843e-05, "loss": 0.0076, "step": 6350 }, { "epoch": 11.316725978647687, "grad_norm": 0.27041152119636536, "learning_rate": 3.2071311977885324e-05, "loss": 0.0068, "step": 6360 }, { "epoch": 11.334519572953736, "grad_norm": 0.1218111589550972, "learning_rate": 3.191705865336197e-05, "loss": 0.0099, "step": 6370 }, { "epoch": 11.352313167259787, "grad_norm": 0.2986414134502411, "learning_rate": 3.1763003080988075e-05, "loss": 0.0075, "step": 6380 }, { "epoch": 11.370106761565836, "grad_norm": 0.1739029437303543, "learning_rate": 3.160914694549063e-05, "loss": 0.0093, "step": 6390 }, { "epoch": 11.387900355871887, "grad_norm": 0.25803279876708984, "learning_rate": 3.145549192941573e-05, "loss": 0.0083, "step": 6400 }, { "epoch": 11.405693950177936, "grad_norm": 0.37231940031051636, "learning_rate": 3.130203971310999e-05, "loss": 0.0092, "step": 6410 }, { "epoch": 11.423487544483987, "grad_norm": 0.13861826062202454, "learning_rate": 3.114879197470225e-05, "loss": 0.0074, "step": 6420 }, { "epoch": 11.441281138790035, "grad_norm": 0.23372088372707367, "learning_rate": 3.0995750390085285e-05, "loss": 0.0094, "step": 6430 }, { "epoch": 11.459074733096086, "grad_norm": 0.16732391715049744, "learning_rate": 3.084291663289728e-05, "loss": 0.0081, "step": 6440 }, { "epoch": 11.476868327402135, "grad_norm": 0.23261557519435883, "learning_rate": 3.069029237450375e-05, "loss": 0.0061, "step": 6450 }, { "epoch": 11.494661921708184, "grad_norm": 0.16935811936855316, "learning_rate": 3.053787928397911e-05, "loss": 0.007, "step": 6460 }, { "epoch": 11.512455516014235, "grad_norm": 0.2779266834259033, "learning_rate": 3.0385679028088526e-05, "loss": 0.0076, "step": 6470 }, { "epoch": 11.530249110320284, "grad_norm": 0.2547450065612793, "learning_rate": 3.023369327126959e-05, "loss": 0.0109, "step": 6480 }, { "epoch": 11.548042704626335, "grad_norm": 0.08240263909101486, "learning_rate": 3.0081923675614198e-05, "loss": 0.006, "step": 6490 }, { "epoch": 11.565836298932384, "grad_norm": 0.18201076984405518, "learning_rate": 2.993037190085034e-05, "loss": 0.0083, "step": 6500 }, { "epoch": 11.583629893238435, "grad_norm": 0.1898127645254135, "learning_rate": 2.977903960432392e-05, "loss": 0.0092, "step": 6510 }, { "epoch": 11.601423487544483, "grad_norm": 0.3015061318874359, "learning_rate": 2.9627928440980722e-05, "loss": 0.0097, "step": 6520 }, { "epoch": 11.619217081850534, "grad_norm": 0.2096952348947525, "learning_rate": 2.9477040063348183e-05, "loss": 0.0123, "step": 6530 }, { "epoch": 11.637010676156583, "grad_norm": 0.15504148602485657, "learning_rate": 2.9326376121517456e-05, "loss": 0.0098, "step": 6540 }, { "epoch": 11.654804270462634, "grad_norm": 0.1393892914056778, "learning_rate": 2.9175938263125236e-05, "loss": 0.0056, "step": 6550 }, { "epoch": 11.672597864768683, "grad_norm": 0.2806447446346283, "learning_rate": 2.9025728133335873e-05, "loss": 0.0093, "step": 6560 }, { "epoch": 11.690391459074734, "grad_norm": 0.28365856409072876, "learning_rate": 2.8875747374823288e-05, "loss": 0.0092, "step": 6570 }, { "epoch": 11.708185053380783, "grad_norm": 0.17643526196479797, "learning_rate": 2.872599762775298e-05, "loss": 0.0145, "step": 6580 }, { "epoch": 11.725978647686834, "grad_norm": 0.17914247512817383, "learning_rate": 2.857648052976425e-05, "loss": 0.0119, "step": 6590 }, { "epoch": 11.743772241992882, "grad_norm": 0.3211177587509155, "learning_rate": 2.8427197715952047e-05, "loss": 0.0081, "step": 6600 }, { "epoch": 11.761565836298931, "grad_norm": 0.16622333228588104, "learning_rate": 2.8278150818849393e-05, "loss": 0.0076, "step": 6610 }, { "epoch": 11.779359430604982, "grad_norm": 0.13399113714694977, "learning_rate": 2.812934146840922e-05, "loss": 0.0108, "step": 6620 }, { "epoch": 11.797153024911031, "grad_norm": 0.23551350831985474, "learning_rate": 2.7980771291986764e-05, "loss": 0.0096, "step": 6630 }, { "epoch": 11.814946619217082, "grad_norm": 0.1992315798997879, "learning_rate": 2.783244191432167e-05, "loss": 0.0086, "step": 6640 }, { "epoch": 11.832740213523131, "grad_norm": 0.1778637170791626, "learning_rate": 2.768435495752022e-05, "loss": 0.0061, "step": 6650 }, { "epoch": 11.850533807829182, "grad_norm": 0.32973238825798035, "learning_rate": 2.753651204103771e-05, "loss": 0.0088, "step": 6660 }, { "epoch": 11.86832740213523, "grad_norm": 0.23835435509681702, "learning_rate": 2.7388914781660523e-05, "loss": 0.0071, "step": 6670 }, { "epoch": 11.886120996441282, "grad_norm": 0.34990188479423523, "learning_rate": 2.7241564793488693e-05, "loss": 0.009, "step": 6680 }, { "epoch": 11.90391459074733, "grad_norm": 0.15917867422103882, "learning_rate": 2.7094463687918037e-05, "loss": 0.0063, "step": 6690 }, { "epoch": 11.921708185053381, "grad_norm": 0.3595600128173828, "learning_rate": 2.694761307362268e-05, "loss": 0.0101, "step": 6700 }, { "epoch": 11.93950177935943, "grad_norm": 0.15483443439006805, "learning_rate": 2.6801014556537467e-05, "loss": 0.006, "step": 6710 }, { "epoch": 11.957295373665481, "grad_norm": 0.10688208788633347, "learning_rate": 2.6654669739840243e-05, "loss": 0.007, "step": 6720 }, { "epoch": 11.97508896797153, "grad_norm": 0.24212613701820374, "learning_rate": 2.650858022393451e-05, "loss": 0.0074, "step": 6730 }, { "epoch": 11.99288256227758, "grad_norm": 0.26816338300704956, "learning_rate": 2.6362747606431747e-05, "loss": 0.0141, "step": 6740 }, { "epoch": 12.01067615658363, "grad_norm": 0.1278560906648636, "learning_rate": 2.6217173482134172e-05, "loss": 0.0083, "step": 6750 }, { "epoch": 12.02846975088968, "grad_norm": 0.19196155667304993, "learning_rate": 2.6071859443017044e-05, "loss": 0.0067, "step": 6760 }, { "epoch": 12.04626334519573, "grad_norm": 0.15388615429401398, "learning_rate": 2.5926807078211414e-05, "loss": 0.0076, "step": 6770 }, { "epoch": 12.064056939501778, "grad_norm": 0.1559678018093109, "learning_rate": 2.5782017973986728e-05, "loss": 0.006, "step": 6780 }, { "epoch": 12.08185053380783, "grad_norm": 0.18299314379692078, "learning_rate": 2.5637493713733374e-05, "loss": 0.0085, "step": 6790 }, { "epoch": 12.099644128113878, "grad_norm": 0.26192229986190796, "learning_rate": 2.549323587794559e-05, "loss": 0.0081, "step": 6800 }, { "epoch": 12.117437722419929, "grad_norm": 0.1605634093284607, "learning_rate": 2.5349246044203895e-05, "loss": 0.01, "step": 6810 }, { "epoch": 12.135231316725978, "grad_norm": 0.13856644928455353, "learning_rate": 2.520552578715808e-05, "loss": 0.0075, "step": 6820 }, { "epoch": 12.153024911032029, "grad_norm": 0.16835758090019226, "learning_rate": 2.506207667850981e-05, "loss": 0.004, "step": 6830 }, { "epoch": 12.170818505338078, "grad_norm": 0.1715366244316101, "learning_rate": 2.4918900286995555e-05, "loss": 0.0065, "step": 6840 }, { "epoch": 12.188612099644129, "grad_norm": 0.21099534630775452, "learning_rate": 2.4775998178369458e-05, "loss": 0.0087, "step": 6850 }, { "epoch": 12.206405693950177, "grad_norm": 0.19617988169193268, "learning_rate": 2.4633371915386017e-05, "loss": 0.0056, "step": 6860 }, { "epoch": 12.224199288256228, "grad_norm": 0.16397397220134735, "learning_rate": 2.4491023057783235e-05, "loss": 0.0111, "step": 6870 }, { "epoch": 12.241992882562277, "grad_norm": 0.20086123049259186, "learning_rate": 2.4348953162265375e-05, "loss": 0.0122, "step": 6880 }, { "epoch": 12.259786476868328, "grad_norm": 0.14804822206497192, "learning_rate": 2.420716378248607e-05, "loss": 0.0109, "step": 6890 }, { "epoch": 12.277580071174377, "grad_norm": 0.20101533830165863, "learning_rate": 2.4065656469031266e-05, "loss": 0.0084, "step": 6900 }, { "epoch": 12.295373665480428, "grad_norm": 0.25633594393730164, "learning_rate": 2.3924432769402268e-05, "loss": 0.0093, "step": 6910 }, { "epoch": 12.313167259786477, "grad_norm": 0.21354562044143677, "learning_rate": 2.3783494227998844e-05, "loss": 0.0108, "step": 6920 }, { "epoch": 12.330960854092528, "grad_norm": 0.18450871109962463, "learning_rate": 2.3642842386102264e-05, "loss": 0.008, "step": 6930 }, { "epoch": 12.348754448398576, "grad_norm": 0.1975587159395218, "learning_rate": 2.3502478781858567e-05, "loss": 0.0063, "step": 6940 }, { "epoch": 12.366548042704625, "grad_norm": 0.13471993803977966, "learning_rate": 2.3362404950261628e-05, "loss": 0.0073, "step": 6950 }, { "epoch": 12.384341637010676, "grad_norm": 0.1467139720916748, "learning_rate": 2.3222622423136458e-05, "loss": 0.0085, "step": 6960 }, { "epoch": 12.402135231316725, "grad_norm": 0.16831523180007935, "learning_rate": 2.3083132729122332e-05, "loss": 0.0062, "step": 6970 }, { "epoch": 12.419928825622776, "grad_norm": 0.19095368683338165, "learning_rate": 2.294393739365621e-05, "loss": 0.0113, "step": 6980 }, { "epoch": 12.437722419928825, "grad_norm": 0.2602579891681671, "learning_rate": 2.2805037938956e-05, "loss": 0.0097, "step": 6990 }, { "epoch": 12.455516014234876, "grad_norm": 0.10121659934520721, "learning_rate": 2.266643588400386e-05, "loss": 0.0085, "step": 7000 }, { "epoch": 12.473309608540925, "grad_norm": 0.14880746603012085, "learning_rate": 2.252813274452969e-05, "loss": 0.008, "step": 7010 }, { "epoch": 12.491103202846976, "grad_norm": 0.1044023260474205, "learning_rate": 2.2390130032994427e-05, "loss": 0.0082, "step": 7020 }, { "epoch": 12.508896797153024, "grad_norm": 0.12160605192184448, "learning_rate": 2.2252429258573633e-05, "loss": 0.0071, "step": 7030 }, { "epoch": 12.526690391459075, "grad_norm": 0.15826143324375153, "learning_rate": 2.2115031927140904e-05, "loss": 0.0062, "step": 7040 }, { "epoch": 12.544483985765124, "grad_norm": 0.15788325667381287, "learning_rate": 2.1977939541251463e-05, "loss": 0.013, "step": 7050 }, { "epoch": 12.562277580071175, "grad_norm": 0.1755935400724411, "learning_rate": 2.1841153600125684e-05, "loss": 0.0109, "step": 7060 }, { "epoch": 12.580071174377224, "grad_norm": 0.25362086296081543, "learning_rate": 2.170467559963267e-05, "loss": 0.0087, "step": 7070 }, { "epoch": 12.597864768683275, "grad_norm": 0.2530089318752289, "learning_rate": 2.1568507032273982e-05, "loss": 0.0062, "step": 7080 }, { "epoch": 12.615658362989324, "grad_norm": 0.1937112808227539, "learning_rate": 2.1432649387167264e-05, "loss": 0.0061, "step": 7090 }, { "epoch": 12.633451957295375, "grad_norm": 0.2581865191459656, "learning_rate": 2.1297104150029973e-05, "loss": 0.0062, "step": 7100 }, { "epoch": 12.651245551601424, "grad_norm": 0.1914801299571991, "learning_rate": 2.116187280316307e-05, "loss": 0.0082, "step": 7110 }, { "epoch": 12.669039145907472, "grad_norm": 0.13688324391841888, "learning_rate": 2.1026956825434908e-05, "loss": 0.005, "step": 7120 }, { "epoch": 12.686832740213523, "grad_norm": 0.12988261878490448, "learning_rate": 2.0892357692265017e-05, "loss": 0.0071, "step": 7130 }, { "epoch": 12.704626334519572, "grad_norm": 0.13537898659706116, "learning_rate": 2.0758076875607947e-05, "loss": 0.008, "step": 7140 }, { "epoch": 12.722419928825623, "grad_norm": 0.16035237908363342, "learning_rate": 2.0624115843937207e-05, "loss": 0.0129, "step": 7150 }, { "epoch": 12.740213523131672, "grad_norm": 0.12505704164505005, "learning_rate": 2.0490476062229157e-05, "loss": 0.0086, "step": 7160 }, { "epoch": 12.758007117437723, "grad_norm": 0.18027375638484955, "learning_rate": 2.035715899194704e-05, "loss": 0.0062, "step": 7170 }, { "epoch": 12.775800711743772, "grad_norm": 0.11907955259084702, "learning_rate": 2.022416609102499e-05, "loss": 0.0079, "step": 7180 }, { "epoch": 12.793594306049823, "grad_norm": 0.140675887465477, "learning_rate": 2.009149881385205e-05, "loss": 0.0105, "step": 7190 }, { "epoch": 12.811387900355871, "grad_norm": 0.21867775917053223, "learning_rate": 1.995915861125634e-05, "loss": 0.0091, "step": 7200 }, { "epoch": 12.829181494661922, "grad_norm": 0.1701361984014511, "learning_rate": 1.9827146930489065e-05, "loss": 0.0072, "step": 7210 }, { "epoch": 12.846975088967971, "grad_norm": 0.17031005024909973, "learning_rate": 1.9695465215208848e-05, "loss": 0.0073, "step": 7220 }, { "epoch": 12.864768683274022, "grad_norm": 0.1395309567451477, "learning_rate": 1.9564114905465813e-05, "loss": 0.009, "step": 7230 }, { "epoch": 12.882562277580071, "grad_norm": 0.15697839856147766, "learning_rate": 1.9433097437685936e-05, "loss": 0.0056, "step": 7240 }, { "epoch": 12.900355871886122, "grad_norm": 0.16651837527751923, "learning_rate": 1.930241424465521e-05, "loss": 0.0091, "step": 7250 }, { "epoch": 12.91814946619217, "grad_norm": 0.21154910326004028, "learning_rate": 1.9172066755504115e-05, "loss": 0.0061, "step": 7260 }, { "epoch": 12.935943060498222, "grad_norm": 0.33214208483695984, "learning_rate": 1.9042056395691914e-05, "loss": 0.0078, "step": 7270 }, { "epoch": 12.95373665480427, "grad_norm": 0.26672059297561646, "learning_rate": 1.8912384586991066e-05, "loss": 0.0073, "step": 7280 }, { "epoch": 12.97153024911032, "grad_norm": 0.1947115957736969, "learning_rate": 1.8783052747471717e-05, "loss": 0.0071, "step": 7290 }, { "epoch": 12.98932384341637, "grad_norm": 0.18347632884979248, "learning_rate": 1.865406229148611e-05, "loss": 0.0101, "step": 7300 }, { "epoch": 13.00711743772242, "grad_norm": 0.1259283721446991, "learning_rate": 1.8525414629653233e-05, "loss": 0.0071, "step": 7310 }, { "epoch": 13.02491103202847, "grad_norm": 0.10032038390636444, "learning_rate": 1.8397111168843255e-05, "loss": 0.0096, "step": 7320 }, { "epoch": 13.042704626334519, "grad_norm": 0.17091962695121765, "learning_rate": 1.8269153312162323e-05, "loss": 0.0064, "step": 7330 }, { "epoch": 13.06049822064057, "grad_norm": 0.11359452456235886, "learning_rate": 1.8141542458937054e-05, "loss": 0.0093, "step": 7340 }, { "epoch": 13.078291814946619, "grad_norm": 0.17523515224456787, "learning_rate": 1.8014280004699268e-05, "loss": 0.0076, "step": 7350 }, { "epoch": 13.09608540925267, "grad_norm": 0.10500986129045486, "learning_rate": 1.788736734117078e-05, "loss": 0.0063, "step": 7360 }, { "epoch": 13.113879003558718, "grad_norm": 0.13552160561084747, "learning_rate": 1.7760805856248152e-05, "loss": 0.0061, "step": 7370 }, { "epoch": 13.13167259786477, "grad_norm": 0.1696113795042038, "learning_rate": 1.7634596933987518e-05, "loss": 0.0096, "step": 7380 }, { "epoch": 13.149466192170818, "grad_norm": 0.2435390204191208, "learning_rate": 1.7508741954589404e-05, "loss": 0.007, "step": 7390 }, { "epoch": 13.167259786476869, "grad_norm": 0.25069230794906616, "learning_rate": 1.7383242294383717e-05, "loss": 0.01, "step": 7400 }, { "epoch": 13.185053380782918, "grad_norm": 0.14684581756591797, "learning_rate": 1.7258099325814632e-05, "loss": 0.0035, "step": 7410 }, { "epoch": 13.202846975088969, "grad_norm": 0.11808019131422043, "learning_rate": 1.7133314417425594e-05, "loss": 0.0084, "step": 7420 }, { "epoch": 13.220640569395018, "grad_norm": 0.2643025517463684, "learning_rate": 1.7008888933844408e-05, "loss": 0.0069, "step": 7430 }, { "epoch": 13.238434163701069, "grad_norm": 0.16348512470722198, "learning_rate": 1.6884824235768172e-05, "loss": 0.0104, "step": 7440 }, { "epoch": 13.256227758007118, "grad_norm": 0.14586317539215088, "learning_rate": 1.6761121679948592e-05, "loss": 0.0065, "step": 7450 }, { "epoch": 13.274021352313166, "grad_norm": 0.18598070740699768, "learning_rate": 1.663778261917695e-05, "loss": 0.0048, "step": 7460 }, { "epoch": 13.291814946619217, "grad_norm": 0.07809760421514511, "learning_rate": 1.651480840226952e-05, "loss": 0.0104, "step": 7470 }, { "epoch": 13.309608540925266, "grad_norm": 0.12334582209587097, "learning_rate": 1.639220037405258e-05, "loss": 0.0094, "step": 7480 }, { "epoch": 13.327402135231317, "grad_norm": 0.26063403487205505, "learning_rate": 1.6269959875347906e-05, "loss": 0.0088, "step": 7490 }, { "epoch": 13.345195729537366, "grad_norm": 0.20475943386554718, "learning_rate": 1.614808824295802e-05, "loss": 0.0066, "step": 7500 }, { "epoch": 13.362989323843417, "grad_norm": 0.13111452758312225, "learning_rate": 1.602658680965152e-05, "loss": 0.0058, "step": 7510 }, { "epoch": 13.380782918149466, "grad_norm": 0.1845695823431015, "learning_rate": 1.5905456904148686e-05, "loss": 0.011, "step": 7520 }, { "epoch": 13.398576512455517, "grad_norm": 0.17362673580646515, "learning_rate": 1.57846998511067e-05, "loss": 0.0069, "step": 7530 }, { "epoch": 13.416370106761565, "grad_norm": 0.15618179738521576, "learning_rate": 1.566431697110538e-05, "loss": 0.0097, "step": 7540 }, { "epoch": 13.434163701067616, "grad_norm": 0.13760599493980408, "learning_rate": 1.554430958063259e-05, "loss": 0.0061, "step": 7550 }, { "epoch": 13.451957295373665, "grad_norm": 0.1948143094778061, "learning_rate": 1.5424678992069912e-05, "loss": 0.0063, "step": 7560 }, { "epoch": 13.469750889679716, "grad_norm": 0.0962040051817894, "learning_rate": 1.5305426513678362e-05, "loss": 0.006, "step": 7570 }, { "epoch": 13.487544483985765, "grad_norm": 0.06810682266950607, "learning_rate": 1.518655344958388e-05, "loss": 0.0099, "step": 7580 }, { "epoch": 13.505338078291816, "grad_norm": 0.22294467687606812, "learning_rate": 1.5068061099763275e-05, "loss": 0.0102, "step": 7590 }, { "epoch": 13.523131672597865, "grad_norm": 0.20120318233966827, "learning_rate": 1.494995076002988e-05, "loss": 0.0076, "step": 7600 }, { "epoch": 13.540925266903916, "grad_norm": 0.23899489641189575, "learning_rate": 1.4832223722019456e-05, "loss": 0.0075, "step": 7610 }, { "epoch": 13.558718861209965, "grad_norm": 0.1266239881515503, "learning_rate": 1.4714881273176035e-05, "loss": 0.0069, "step": 7620 }, { "epoch": 13.576512455516013, "grad_norm": 0.10396558046340942, "learning_rate": 1.4597924696737835e-05, "loss": 0.0054, "step": 7630 }, { "epoch": 13.594306049822064, "grad_norm": 0.17402786016464233, "learning_rate": 1.4481355271723252e-05, "loss": 0.0082, "step": 7640 }, { "epoch": 13.612099644128113, "grad_norm": 0.2635336220264435, "learning_rate": 1.4365174272916809e-05, "loss": 0.0091, "step": 7650 }, { "epoch": 13.629893238434164, "grad_norm": 0.16081690788269043, "learning_rate": 1.4249382970855319e-05, "loss": 0.0068, "step": 7660 }, { "epoch": 13.647686832740213, "grad_norm": 0.1956338882446289, "learning_rate": 1.4133982631813903e-05, "loss": 0.009, "step": 7670 }, { "epoch": 13.665480427046264, "grad_norm": 0.26059967279434204, "learning_rate": 1.4018974517792194e-05, "loss": 0.0057, "step": 7680 }, { "epoch": 13.683274021352313, "grad_norm": 0.11822003126144409, "learning_rate": 1.390435988650048e-05, "loss": 0.0064, "step": 7690 }, { "epoch": 13.701067615658364, "grad_norm": 0.1905195415019989, "learning_rate": 1.3790139991346006e-05, "loss": 0.0046, "step": 7700 }, { "epoch": 13.718861209964412, "grad_norm": 0.15209944546222687, "learning_rate": 1.367631608141926e-05, "loss": 0.0066, "step": 7710 }, { "epoch": 13.736654804270463, "grad_norm": 0.1900908648967743, "learning_rate": 1.3562889401480278e-05, "loss": 0.0079, "step": 7720 }, { "epoch": 13.754448398576512, "grad_norm": 0.0607878603041172, "learning_rate": 1.3449861191945074e-05, "loss": 0.008, "step": 7730 }, { "epoch": 13.772241992882563, "grad_norm": 0.14666680991649628, "learning_rate": 1.3337232688872009e-05, "loss": 0.0072, "step": 7740 }, { "epoch": 13.790035587188612, "grad_norm": 0.08438649773597717, "learning_rate": 1.3225005123948364e-05, "loss": 0.0061, "step": 7750 }, { "epoch": 13.807829181494661, "grad_norm": 0.10742039978504181, "learning_rate": 1.311317972447681e-05, "loss": 0.0066, "step": 7760 }, { "epoch": 13.825622775800712, "grad_norm": 0.06236627325415611, "learning_rate": 1.3001757713361996e-05, "loss": 0.0052, "step": 7770 }, { "epoch": 13.84341637010676, "grad_norm": 0.15915493667125702, "learning_rate": 1.2890740309097204e-05, "loss": 0.0064, "step": 7780 }, { "epoch": 13.861209964412812, "grad_norm": 0.12230442464351654, "learning_rate": 1.2780128725750944e-05, "loss": 0.006, "step": 7790 }, { "epoch": 13.87900355871886, "grad_norm": 0.17348918318748474, "learning_rate": 1.266992417295379e-05, "loss": 0.0074, "step": 7800 }, { "epoch": 13.896797153024911, "grad_norm": 0.1712888926267624, "learning_rate": 1.2560127855885073e-05, "loss": 0.0055, "step": 7810 }, { "epoch": 13.91459074733096, "grad_norm": 0.1703440248966217, "learning_rate": 1.2450740975259745e-05, "loss": 0.0106, "step": 7820 }, { "epoch": 13.932384341637011, "grad_norm": 0.20247185230255127, "learning_rate": 1.234176472731517e-05, "loss": 0.0109, "step": 7830 }, { "epoch": 13.95017793594306, "grad_norm": 0.19216491281986237, "learning_rate": 1.2233200303798158e-05, "loss": 0.0102, "step": 7840 }, { "epoch": 13.96797153024911, "grad_norm": 0.11944342404603958, "learning_rate": 1.2125048891951846e-05, "loss": 0.0081, "step": 7850 }, { "epoch": 13.98576512455516, "grad_norm": 0.06938405334949493, "learning_rate": 1.2017311674502745e-05, "loss": 0.0038, "step": 7860 }, { "epoch": 14.00355871886121, "grad_norm": 0.1549089550971985, "learning_rate": 1.1909989829647822e-05, "loss": 0.0056, "step": 7870 }, { "epoch": 14.02135231316726, "grad_norm": 0.24149680137634277, "learning_rate": 1.1803084531041553e-05, "loss": 0.0078, "step": 7880 }, { "epoch": 14.03914590747331, "grad_norm": 0.08600204437971115, "learning_rate": 1.1696596947783162e-05, "loss": 0.0068, "step": 7890 }, { "epoch": 14.05693950177936, "grad_norm": 0.12699651718139648, "learning_rate": 1.1590528244403803e-05, "loss": 0.0072, "step": 7900 }, { "epoch": 14.07473309608541, "grad_norm": 0.16090741753578186, "learning_rate": 1.148487958085382e-05, "loss": 0.0068, "step": 7910 }, { "epoch": 14.092526690391459, "grad_norm": 0.20414701104164124, "learning_rate": 1.1379652112490086e-05, "loss": 0.0072, "step": 7920 }, { "epoch": 14.11032028469751, "grad_norm": 0.1307162493467331, "learning_rate": 1.1274846990063315e-05, "loss": 0.0056, "step": 7930 }, { "epoch": 14.128113879003559, "grad_norm": 0.23150752484798431, "learning_rate": 1.117046535970554e-05, "loss": 0.0084, "step": 7940 }, { "epoch": 14.145907473309608, "grad_norm": 0.22152088582515717, "learning_rate": 1.106650836291755e-05, "loss": 0.0078, "step": 7950 }, { "epoch": 14.163701067615659, "grad_norm": 0.12757408618927002, "learning_rate": 1.0962977136556418e-05, "loss": 0.0124, "step": 7960 }, { "epoch": 14.181494661921707, "grad_norm": 0.08678414672613144, "learning_rate": 1.0859872812823024e-05, "loss": 0.0097, "step": 7970 }, { "epoch": 14.199288256227758, "grad_norm": 0.16517916321754456, "learning_rate": 1.0757196519249747e-05, "loss": 0.0094, "step": 7980 }, { "epoch": 14.217081850533807, "grad_norm": 0.2503165006637573, "learning_rate": 1.0654949378688077e-05, "loss": 0.0102, "step": 7990 }, { "epoch": 14.234875444839858, "grad_norm": 0.19046436250209808, "learning_rate": 1.0553132509296376e-05, "loss": 0.0086, "step": 8000 }, { "epoch": 14.252669039145907, "grad_norm": 0.16274245083332062, "learning_rate": 1.0451747024527613e-05, "loss": 0.0077, "step": 8010 }, { "epoch": 14.270462633451958, "grad_norm": 0.1391967236995697, "learning_rate": 1.0350794033117189e-05, "loss": 0.0077, "step": 8020 }, { "epoch": 14.288256227758007, "grad_norm": 0.1625727415084839, "learning_rate": 1.0250274639070856e-05, "loss": 0.008, "step": 8030 }, { "epoch": 14.306049822064058, "grad_norm": 0.11398770660161972, "learning_rate": 1.0150189941652599e-05, "loss": 0.0061, "step": 8040 }, { "epoch": 14.323843416370106, "grad_norm": 0.08820738643407822, "learning_rate": 1.0050541035372635e-05, "loss": 0.0057, "step": 8050 }, { "epoch": 14.341637010676157, "grad_norm": 0.09258157014846802, "learning_rate": 9.951329009975458e-06, "loss": 0.0054, "step": 8060 }, { "epoch": 14.359430604982206, "grad_norm": 0.09987417608499527, "learning_rate": 9.852554950427845e-06, "loss": 0.0086, "step": 8070 }, { "epoch": 14.377224199288257, "grad_norm": 0.2878149747848511, "learning_rate": 9.754219936907105e-06, "loss": 0.0059, "step": 8080 }, { "epoch": 14.395017793594306, "grad_norm": 0.2295035421848297, "learning_rate": 9.656325044789194e-06, "loss": 0.0054, "step": 8090 }, { "epoch": 14.412811387900355, "grad_norm": 0.0969391018152237, "learning_rate": 9.55887134463697e-06, "loss": 0.0045, "step": 8100 }, { "epoch": 14.430604982206406, "grad_norm": 0.14281509816646576, "learning_rate": 9.461859902188475e-06, "loss": 0.0067, "step": 8110 }, { "epoch": 14.448398576512455, "grad_norm": 0.15115158259868622, "learning_rate": 9.365291778345303e-06, "loss": 0.0063, "step": 8120 }, { "epoch": 14.466192170818506, "grad_norm": 0.07826753705739975, "learning_rate": 9.269168029160991e-06, "loss": 0.0062, "step": 8130 }, { "epoch": 14.483985765124554, "grad_norm": 0.05976415425539017, "learning_rate": 9.173489705829447e-06, "loss": 0.0039, "step": 8140 }, { "epoch": 14.501779359430605, "grad_norm": 0.11851614713668823, "learning_rate": 9.078257854673516e-06, "loss": 0.0084, "step": 8150 }, { "epoch": 14.519572953736654, "grad_norm": 0.13051682710647583, "learning_rate": 8.983473517133429e-06, "loss": 0.0062, "step": 8160 }, { "epoch": 14.537366548042705, "grad_norm": 0.10752177983522415, "learning_rate": 8.889137729755537e-06, "loss": 0.006, "step": 8170 }, { "epoch": 14.555160142348754, "grad_norm": 0.17306986451148987, "learning_rate": 8.79525152418087e-06, "loss": 0.0079, "step": 8180 }, { "epoch": 14.572953736654805, "grad_norm": 0.18458189070224762, "learning_rate": 8.701815927133961e-06, "loss": 0.0083, "step": 8190 }, { "epoch": 14.590747330960854, "grad_norm": 0.14504797756671906, "learning_rate": 8.608831960411534e-06, "loss": 0.0075, "step": 8200 }, { "epoch": 14.608540925266905, "grad_norm": 0.2952544093132019, "learning_rate": 8.516300640871321e-06, "loss": 0.0055, "step": 8210 }, { "epoch": 14.626334519572953, "grad_norm": 0.15258938074111938, "learning_rate": 8.424222980421038e-06, "loss": 0.0078, "step": 8220 }, { "epoch": 14.644128113879004, "grad_norm": 0.25653886795043945, "learning_rate": 8.332599986007184e-06, "loss": 0.0063, "step": 8230 }, { "epoch": 14.661921708185053, "grad_norm": 0.12304195761680603, "learning_rate": 8.241432659604203e-06, "loss": 0.0069, "step": 8240 }, { "epoch": 14.679715302491104, "grad_norm": 0.0941120833158493, "learning_rate": 8.150721998203331e-06, "loss": 0.0047, "step": 8250 }, { "epoch": 14.697508896797153, "grad_norm": 0.0756886824965477, "learning_rate": 8.06046899380184e-06, "loss": 0.0056, "step": 8260 }, { "epoch": 14.715302491103202, "grad_norm": 0.10911428183317184, "learning_rate": 7.970674633392133e-06, "loss": 0.0086, "step": 8270 }, { "epoch": 14.733096085409253, "grad_norm": 0.18531003594398499, "learning_rate": 7.881339898950924e-06, "loss": 0.0122, "step": 8280 }, { "epoch": 14.750889679715302, "grad_norm": 0.25072163343429565, "learning_rate": 7.792465767428597e-06, "loss": 0.0076, "step": 8290 }, { "epoch": 14.768683274021353, "grad_norm": 0.11867135763168335, "learning_rate": 7.704053210738376e-06, "loss": 0.0047, "step": 8300 }, { "epoch": 14.786476868327401, "grad_norm": 0.11052953451871872, "learning_rate": 7.6161031957458494e-06, "loss": 0.0068, "step": 8310 }, { "epoch": 14.804270462633452, "grad_norm": 0.1394992172718048, "learning_rate": 7.5286166842582605e-06, "loss": 0.0094, "step": 8320 }, { "epoch": 14.822064056939501, "grad_norm": 0.14540903270244598, "learning_rate": 7.4415946330140814e-06, "loss": 0.0054, "step": 8330 }, { "epoch": 14.839857651245552, "grad_norm": 0.14087249338626862, "learning_rate": 7.3550379936725644e-06, "loss": 0.0066, "step": 8340 }, { "epoch": 14.857651245551601, "grad_norm": 0.11690076440572739, "learning_rate": 7.2689477128032035e-06, "loss": 0.0073, "step": 8350 }, { "epoch": 14.875444839857652, "grad_norm": 0.2452922910451889, "learning_rate": 7.183324731875551e-06, "loss": 0.0057, "step": 8360 }, { "epoch": 14.8932384341637, "grad_norm": 0.14778119325637817, "learning_rate": 7.098169987248782e-06, "loss": 0.0057, "step": 8370 }, { "epoch": 14.911032028469752, "grad_norm": 0.23892836272716522, "learning_rate": 7.013484410161553e-06, "loss": 0.0058, "step": 8380 }, { "epoch": 14.9288256227758, "grad_norm": 0.21200525760650635, "learning_rate": 6.92926892672176e-06, "loss": 0.0079, "step": 8390 }, { "epoch": 14.946619217081851, "grad_norm": 0.0676952600479126, "learning_rate": 6.845524457896446e-06, "loss": 0.0067, "step": 8400 }, { "epoch": 14.9644128113879, "grad_norm": 0.2176414430141449, "learning_rate": 6.7622519195017165e-06, "loss": 0.0047, "step": 8410 }, { "epoch": 14.982206405693951, "grad_norm": 0.14390335977077484, "learning_rate": 6.679452222192684e-06, "loss": 0.0066, "step": 8420 }, { "epoch": 15.0, "grad_norm": 0.1431676149368286, "learning_rate": 6.597126271453579e-06, "loss": 0.0052, "step": 8430 }, { "epoch": 15.017793594306049, "grad_norm": 0.09316446632146835, "learning_rate": 6.51527496758782e-06, "loss": 0.006, "step": 8440 }, { "epoch": 15.0355871886121, "grad_norm": 0.06405527144670486, "learning_rate": 6.433899205708155e-06, "loss": 0.007, "step": 8450 }, { "epoch": 15.053380782918149, "grad_norm": 0.1082666888833046, "learning_rate": 6.352999875726856e-06, "loss": 0.0047, "step": 8460 }, { "epoch": 15.0711743772242, "grad_norm": 0.11947515606880188, "learning_rate": 6.272577862346052e-06, "loss": 0.0049, "step": 8470 }, { "epoch": 15.088967971530248, "grad_norm": 0.10116983205080032, "learning_rate": 6.192634045047996e-06, "loss": 0.0051, "step": 8480 }, { "epoch": 15.1067615658363, "grad_norm": 0.07082372903823853, "learning_rate": 6.113169298085458e-06, "loss": 0.0098, "step": 8490 }, { "epoch": 15.124555160142348, "grad_norm": 0.0615873746573925, "learning_rate": 6.034184490472195e-06, "loss": 0.0073, "step": 8500 }, { "epoch": 15.142348754448399, "grad_norm": 0.07893116027116776, "learning_rate": 5.955680485973386e-06, "loss": 0.0061, "step": 8510 }, { "epoch": 15.160142348754448, "grad_norm": 0.13956958055496216, "learning_rate": 5.877658143096265e-06, "loss": 0.0063, "step": 8520 }, { "epoch": 15.177935943060499, "grad_norm": 0.13331711292266846, "learning_rate": 5.800118315080661e-06, "loss": 0.0045, "step": 8530 }, { "epoch": 15.195729537366548, "grad_norm": 0.06119225174188614, "learning_rate": 5.723061849889716e-06, "loss": 0.0045, "step": 8540 }, { "epoch": 15.213523131672599, "grad_norm": 0.1262703239917755, "learning_rate": 5.646489590200604e-06, "loss": 0.0058, "step": 8550 }, { "epoch": 15.231316725978647, "grad_norm": 0.11567643284797668, "learning_rate": 5.570402373395256e-06, "loss": 0.0052, "step": 8560 }, { "epoch": 15.249110320284698, "grad_norm": 0.2528938055038452, "learning_rate": 5.494801031551305e-06, "loss": 0.0058, "step": 8570 }, { "epoch": 15.266903914590747, "grad_norm": 0.048760950565338135, "learning_rate": 5.41968639143291e-06, "loss": 0.0063, "step": 8580 }, { "epoch": 15.284697508896798, "grad_norm": 0.07633157819509506, "learning_rate": 5.345059274481751e-06, "loss": 0.0051, "step": 8590 }, { "epoch": 15.302491103202847, "grad_norm": 0.16065268218517303, "learning_rate": 5.270920496808002e-06, "loss": 0.0056, "step": 8600 }, { "epoch": 15.320284697508896, "grad_norm": 0.08078490197658539, "learning_rate": 5.1972708691814695e-06, "loss": 0.0078, "step": 8610 }, { "epoch": 15.338078291814947, "grad_norm": 0.17307817935943604, "learning_rate": 5.124111197022674e-06, "loss": 0.0084, "step": 8620 }, { "epoch": 15.355871886120996, "grad_norm": 0.14141802489757538, "learning_rate": 5.051442280394081e-06, "loss": 0.008, "step": 8630 }, { "epoch": 15.373665480427047, "grad_norm": 0.08857766538858414, "learning_rate": 4.979264913991322e-06, "loss": 0.0053, "step": 8640 }, { "epoch": 15.391459074733095, "grad_norm": 0.13246099650859833, "learning_rate": 4.907579887134489e-06, "loss": 0.0044, "step": 8650 }, { "epoch": 15.409252669039146, "grad_norm": 0.11752492934465408, "learning_rate": 4.836387983759572e-06, "loss": 0.0084, "step": 8660 }, { "epoch": 15.427046263345195, "grad_norm": 0.07969677448272705, "learning_rate": 4.765689982409816e-06, "loss": 0.0054, "step": 8670 }, { "epoch": 15.444839857651246, "grad_norm": 0.1713133305311203, "learning_rate": 4.695486656227233e-06, "loss": 0.0062, "step": 8680 }, { "epoch": 15.462633451957295, "grad_norm": 0.0910022109746933, "learning_rate": 4.625778772944156e-06, "loss": 0.009, "step": 8690 }, { "epoch": 15.480427046263346, "grad_norm": 0.18115879595279694, "learning_rate": 4.556567094874825e-06, "loss": 0.0068, "step": 8700 }, { "epoch": 15.498220640569395, "grad_norm": 0.20949910581111908, "learning_rate": 4.487852378907059e-06, "loss": 0.0051, "step": 8710 }, { "epoch": 15.516014234875446, "grad_norm": 0.21307167410850525, "learning_rate": 4.419635376493986e-06, "loss": 0.0041, "step": 8720 }, { "epoch": 15.533807829181494, "grad_norm": 0.16194497048854828, "learning_rate": 4.351916833645825e-06, "loss": 0.0066, "step": 8730 }, { "epoch": 15.551601423487545, "grad_norm": 0.11840556561946869, "learning_rate": 4.284697490921691e-06, "loss": 0.0052, "step": 8740 }, { "epoch": 15.569395017793594, "grad_norm": 0.09030645340681076, "learning_rate": 4.2179780834215585e-06, "loss": 0.0074, "step": 8750 }, { "epoch": 15.587188612099645, "grad_norm": 0.1824294924736023, "learning_rate": 4.151759340778178e-06, "loss": 0.0054, "step": 8760 }, { "epoch": 15.604982206405694, "grad_norm": 0.16402997076511383, "learning_rate": 4.086041987149109e-06, "loss": 0.0061, "step": 8770 }, { "epoch": 15.622775800711743, "grad_norm": 0.13904333114624023, "learning_rate": 4.020826741208811e-06, "loss": 0.0071, "step": 8780 }, { "epoch": 15.640569395017794, "grad_norm": 0.1926247626543045, "learning_rate": 3.956114316140746e-06, "loss": 0.0073, "step": 8790 }, { "epoch": 15.658362989323843, "grad_norm": 0.07429147511720657, "learning_rate": 3.891905419629643e-06, "loss": 0.0083, "step": 8800 }, { "epoch": 15.676156583629894, "grad_norm": 0.0966663584113121, "learning_rate": 3.8282007538536946e-06, "loss": 0.0119, "step": 8810 }, { "epoch": 15.693950177935942, "grad_norm": 0.11520890146493912, "learning_rate": 3.7650010154769265e-06, "loss": 0.0054, "step": 8820 }, { "epoch": 15.711743772241993, "grad_norm": 0.21386560797691345, "learning_rate": 3.7023068956415608e-06, "loss": 0.0084, "step": 8830 }, { "epoch": 15.729537366548042, "grad_norm": 0.08490584790706635, "learning_rate": 3.6401190799604303e-06, "loss": 0.0043, "step": 8840 }, { "epoch": 15.747330960854093, "grad_norm": 0.11275563389062881, "learning_rate": 3.578438248509536e-06, "loss": 0.0043, "step": 8850 }, { "epoch": 15.765124555160142, "grad_norm": 0.1284712255001068, "learning_rate": 3.5172650758205583e-06, "loss": 0.0087, "step": 8860 }, { "epoch": 15.782918149466193, "grad_norm": 0.23638872802257538, "learning_rate": 3.45660023087353e-06, "loss": 0.0064, "step": 8870 }, { "epoch": 15.800711743772242, "grad_norm": 0.12647084891796112, "learning_rate": 3.3964443770894528e-06, "loss": 0.0038, "step": 8880 }, { "epoch": 15.818505338078293, "grad_norm": 0.13691189885139465, "learning_rate": 3.3367981723231245e-06, "loss": 0.0069, "step": 8890 }, { "epoch": 15.836298932384341, "grad_norm": 0.10518136620521545, "learning_rate": 3.2776622688558746e-06, "loss": 0.0042, "step": 8900 }, { "epoch": 15.854092526690392, "grad_norm": 0.11015290021896362, "learning_rate": 3.2190373133884677e-06, "loss": 0.0088, "step": 8910 }, { "epoch": 15.871886120996441, "grad_norm": 0.1718825399875641, "learning_rate": 3.1609239470340446e-06, "loss": 0.005, "step": 8920 }, { "epoch": 15.889679715302492, "grad_norm": 0.15669891238212585, "learning_rate": 3.1033228053110373e-06, "loss": 0.0057, "step": 8930 }, { "epoch": 15.907473309608541, "grad_norm": 0.08638143539428711, "learning_rate": 3.0462345181363314e-06, "loss": 0.004, "step": 8940 }, { "epoch": 15.92526690391459, "grad_norm": 0.21767888963222504, "learning_rate": 2.9896597098182654e-06, "loss": 0.0078, "step": 8950 }, { "epoch": 15.94306049822064, "grad_norm": 0.10943535715341568, "learning_rate": 2.933598999049891e-06, "loss": 0.0075, "step": 8960 }, { "epoch": 15.96085409252669, "grad_norm": 0.07237322628498077, "learning_rate": 2.8780529989021697e-06, "loss": 0.0061, "step": 8970 }, { "epoch": 15.97864768683274, "grad_norm": 0.2045833170413971, "learning_rate": 2.823022316817242e-06, "loss": 0.0047, "step": 8980 }, { "epoch": 15.99644128113879, "grad_norm": 0.13137757778167725, "learning_rate": 2.7685075546018456e-06, "loss": 0.0092, "step": 8990 }, { "epoch": 16.01423487544484, "grad_norm": 0.05556589737534523, "learning_rate": 2.7145093084206598e-06, "loss": 0.0082, "step": 9000 }, { "epoch": 16.03202846975089, "grad_norm": 0.1200408861041069, "learning_rate": 2.661028168789892e-06, "loss": 0.0075, "step": 9010 }, { "epoch": 16.04982206405694, "grad_norm": 0.17064574360847473, "learning_rate": 2.6080647205706855e-06, "loss": 0.0058, "step": 9020 }, { "epoch": 16.06761565836299, "grad_norm": 0.15872745215892792, "learning_rate": 2.555619542962834e-06, "loss": 0.0053, "step": 9030 }, { "epoch": 16.085409252669038, "grad_norm": 0.16841872036457062, "learning_rate": 2.503693209498409e-06, "loss": 0.0061, "step": 9040 }, { "epoch": 16.10320284697509, "grad_norm": 0.07748531550168991, "learning_rate": 2.452286288035449e-06, "loss": 0.0047, "step": 9050 }, { "epoch": 16.12099644128114, "grad_norm": 0.15680713951587677, "learning_rate": 2.4013993407518363e-06, "loss": 0.0087, "step": 9060 }, { "epoch": 16.13879003558719, "grad_norm": 0.12604057788848877, "learning_rate": 2.351032924139063e-06, "loss": 0.0037, "step": 9070 }, { "epoch": 16.156583629893237, "grad_norm": 0.09285897016525269, "learning_rate": 2.30118758899619e-06, "loss": 0.0081, "step": 9080 }, { "epoch": 16.17437722419929, "grad_norm": 0.07407069951295853, "learning_rate": 2.2518638804238157e-06, "loss": 0.0038, "step": 9090 }, { "epoch": 16.19217081850534, "grad_norm": 0.10429558157920837, "learning_rate": 2.203062337818118e-06, "loss": 0.005, "step": 9100 }, { "epoch": 16.209964412811388, "grad_norm": 0.09953689575195312, "learning_rate": 2.1547834948649483e-06, "loss": 0.0081, "step": 9110 }, { "epoch": 16.227758007117437, "grad_norm": 0.1587396264076233, "learning_rate": 2.1070278795340017e-06, "loss": 0.0074, "step": 9120 }, { "epoch": 16.245551601423486, "grad_norm": 0.08077317476272583, "learning_rate": 2.059796014073029e-06, "loss": 0.005, "step": 9130 }, { "epoch": 16.26334519572954, "grad_norm": 0.1258343905210495, "learning_rate": 2.01308841500214e-06, "loss": 0.0032, "step": 9140 }, { "epoch": 16.281138790035588, "grad_norm": 0.11051934212446213, "learning_rate": 1.9669055931081704e-06, "loss": 0.005, "step": 9150 }, { "epoch": 16.298932384341636, "grad_norm": 0.05631111562252045, "learning_rate": 1.9212480534390507e-06, "loss": 0.0102, "step": 9160 }, { "epoch": 16.316725978647685, "grad_norm": 0.06051097810268402, "learning_rate": 1.8761162952983246e-06, "loss": 0.0057, "step": 9170 }, { "epoch": 16.334519572953738, "grad_norm": 0.19867028295993805, "learning_rate": 1.8315108122396618e-06, "loss": 0.009, "step": 9180 }, { "epoch": 16.352313167259787, "grad_norm": 0.19927223026752472, "learning_rate": 1.787432092061475e-06, "loss": 0.0052, "step": 9190 }, { "epoch": 16.370106761565836, "grad_norm": 0.21529878675937653, "learning_rate": 1.743880616801602e-06, "loss": 0.0044, "step": 9200 }, { "epoch": 16.387900355871885, "grad_norm": 0.12427757680416107, "learning_rate": 1.7008568627319865e-06, "loss": 0.0078, "step": 9210 }, { "epoch": 16.405693950177938, "grad_norm": 0.18077917397022247, "learning_rate": 1.6583613003535226e-06, "loss": 0.0077, "step": 9220 }, { "epoch": 16.423487544483987, "grad_norm": 0.17289398610591888, "learning_rate": 1.6163943943908522e-06, "loss": 0.0069, "step": 9230 }, { "epoch": 16.441281138790035, "grad_norm": 0.11186295747756958, "learning_rate": 1.5749566037873476e-06, "loss": 0.0068, "step": 9240 }, { "epoch": 16.459074733096084, "grad_norm": 0.2473910003900528, "learning_rate": 1.5340483817000428e-06, "loss": 0.0099, "step": 9250 }, { "epoch": 16.476868327402137, "grad_norm": 0.13086803257465363, "learning_rate": 1.4936701754947101e-06, "loss": 0.0059, "step": 9260 }, { "epoch": 16.494661921708186, "grad_norm": 0.17718632519245148, "learning_rate": 1.4538224267409361e-06, "loss": 0.0066, "step": 9270 }, { "epoch": 16.512455516014235, "grad_norm": 0.05542243272066116, "learning_rate": 1.414505571207314e-06, "loss": 0.0047, "step": 9280 }, { "epoch": 16.530249110320284, "grad_norm": 0.20820865035057068, "learning_rate": 1.3757200388566816e-06, "loss": 0.0057, "step": 9290 }, { "epoch": 16.548042704626333, "grad_norm": 0.07538910955190659, "learning_rate": 1.3374662538414074e-06, "loss": 0.0042, "step": 9300 }, { "epoch": 16.565836298932386, "grad_norm": 0.21573886275291443, "learning_rate": 1.2997446344987617e-06, "loss": 0.0084, "step": 9310 }, { "epoch": 16.583629893238435, "grad_norm": 0.20143724977970123, "learning_rate": 1.262555593346315e-06, "loss": 0.0074, "step": 9320 }, { "epoch": 16.601423487544483, "grad_norm": 0.08944947272539139, "learning_rate": 1.2258995370774685e-06, "loss": 0.0054, "step": 9330 }, { "epoch": 16.619217081850532, "grad_norm": 0.14856110513210297, "learning_rate": 1.1897768665569798e-06, "loss": 0.0075, "step": 9340 }, { "epoch": 16.637010676156585, "grad_norm": 0.12041742354631424, "learning_rate": 1.1541879768165954e-06, "loss": 0.0073, "step": 9350 }, { "epoch": 16.654804270462634, "grad_norm": 0.10676651448011398, "learning_rate": 1.1191332570507085e-06, "loss": 0.0046, "step": 9360 }, { "epoch": 16.672597864768683, "grad_norm": 0.19305583834648132, "learning_rate": 1.0846130906121132e-06, "loss": 0.0095, "step": 9370 }, { "epoch": 16.690391459074732, "grad_norm": 0.26436492800712585, "learning_rate": 1.0506278550078131e-06, "loss": 0.0102, "step": 9380 }, { "epoch": 16.708185053380785, "grad_norm": 0.23873686790466309, "learning_rate": 1.0171779218949185e-06, "loss": 0.0073, "step": 9390 }, { "epoch": 16.725978647686834, "grad_norm": 0.1502653956413269, "learning_rate": 9.842636570765174e-07, "loss": 0.0066, "step": 9400 }, { "epoch": 16.743772241992882, "grad_norm": 0.14922446012496948, "learning_rate": 9.518854204977612e-07, "loss": 0.0043, "step": 9410 }, { "epoch": 16.76156583629893, "grad_norm": 0.17765751481056213, "learning_rate": 9.200435662418349e-07, "loss": 0.0048, "step": 9420 }, { "epoch": 16.779359430604984, "grad_norm": 0.04855192452669144, "learning_rate": 8.887384425261658e-07, "loss": 0.0062, "step": 9430 }, { "epoch": 16.797153024911033, "grad_norm": 0.10483010113239288, "learning_rate": 8.579703916985648e-07, "loss": 0.005, "step": 9440 }, { "epoch": 16.814946619217082, "grad_norm": 0.12176404893398285, "learning_rate": 8.277397502335194e-07, "loss": 0.0111, "step": 9450 }, { "epoch": 16.83274021352313, "grad_norm": 0.18805243074893951, "learning_rate": 7.980468487284675e-07, "loss": 0.0051, "step": 9460 }, { "epoch": 16.85053380782918, "grad_norm": 0.07495567202568054, "learning_rate": 7.688920119002297e-07, "loss": 0.0054, "step": 9470 }, { "epoch": 16.868327402135233, "grad_norm": 0.13131652772426605, "learning_rate": 7.402755585814269e-07, "loss": 0.0064, "step": 9480 }, { "epoch": 16.88612099644128, "grad_norm": 0.16626645624637604, "learning_rate": 7.121978017170073e-07, "loss": 0.0055, "step": 9490 }, { "epoch": 16.90391459074733, "grad_norm": 0.130281463265419, "learning_rate": 6.846590483608306e-07, "loss": 0.007, "step": 9500 }, { "epoch": 16.92170818505338, "grad_norm": 0.12625665962696075, "learning_rate": 6.576595996722834e-07, "loss": 0.0087, "step": 9510 }, { "epoch": 16.939501779359432, "grad_norm": 0.1332959085702896, "learning_rate": 6.311997509130141e-07, "loss": 0.0061, "step": 9520 }, { "epoch": 16.95729537366548, "grad_norm": 0.1283436268568039, "learning_rate": 6.052797914436803e-07, "loss": 0.0076, "step": 9530 }, { "epoch": 16.97508896797153, "grad_norm": 0.12841516733169556, "learning_rate": 5.799000047208181e-07, "loss": 0.0083, "step": 9540 }, { "epoch": 16.99288256227758, "grad_norm": 0.15479148924350739, "learning_rate": 5.550606682937054e-07, "loss": 0.0085, "step": 9550 }, { "epoch": 17.01067615658363, "grad_norm": 0.15748296678066254, "learning_rate": 5.307620538013481e-07, "loss": 0.0075, "step": 9560 }, { "epoch": 17.02846975088968, "grad_norm": 0.056218549609184265, "learning_rate": 5.070044269694874e-07, "loss": 0.0069, "step": 9570 }, { "epoch": 17.04626334519573, "grad_norm": 0.26711803674697876, "learning_rate": 4.837880476077417e-07, "loss": 0.0069, "step": 9580 }, { "epoch": 17.06405693950178, "grad_norm": 0.24658702313899994, "learning_rate": 4.6111316960670835e-07, "loss": 0.0091, "step": 9590 }, { "epoch": 17.081850533807827, "grad_norm": 0.05740724876523018, "learning_rate": 4.389800409352218e-07, "loss": 0.003, "step": 9600 }, { "epoch": 17.09964412811388, "grad_norm": 0.17005211114883423, "learning_rate": 4.173889036376277e-07, "loss": 0.0065, "step": 9610 }, { "epoch": 17.11743772241993, "grad_norm": 0.1728227436542511, "learning_rate": 3.963399938311463e-07, "loss": 0.0093, "step": 9620 }, { "epoch": 17.135231316725978, "grad_norm": 0.09242857992649078, "learning_rate": 3.7583354170328545e-07, "loss": 0.0082, "step": 9630 }, { "epoch": 17.153024911032027, "grad_norm": 0.2696179747581482, "learning_rate": 3.558697715093207e-07, "loss": 0.0069, "step": 9640 }, { "epoch": 17.17081850533808, "grad_norm": 0.2052176296710968, "learning_rate": 3.3644890156983576e-07, "loss": 0.0048, "step": 9650 }, { "epoch": 17.18861209964413, "grad_norm": 0.17072510719299316, "learning_rate": 3.175711442683638e-07, "loss": 0.0067, "step": 9660 }, { "epoch": 17.206405693950177, "grad_norm": 0.14615271985530853, "learning_rate": 2.9923670604902197e-07, "loss": 0.0067, "step": 9670 }, { "epoch": 17.224199288256226, "grad_norm": 0.043361760675907135, "learning_rate": 2.814457874143028e-07, "loss": 0.0054, "step": 9680 }, { "epoch": 17.24199288256228, "grad_norm": 0.15768305957317352, "learning_rate": 2.641985829228366e-07, "loss": 0.0076, "step": 9690 }, { "epoch": 17.259786476868328, "grad_norm": 0.08170788735151291, "learning_rate": 2.474952811872877e-07, "loss": 0.0048, "step": 9700 }, { "epoch": 17.277580071174377, "grad_norm": 0.10405512154102325, "learning_rate": 2.3133606487228397e-07, "loss": 0.0056, "step": 9710 }, { "epoch": 17.295373665480426, "grad_norm": 0.07582027465105057, "learning_rate": 2.157211106924295e-07, "loss": 0.0096, "step": 9720 }, { "epoch": 17.31316725978648, "grad_norm": 0.08911514282226562, "learning_rate": 2.006505894103672e-07, "loss": 0.0053, "step": 9730 }, { "epoch": 17.330960854092528, "grad_norm": 0.0798618420958519, "learning_rate": 1.8612466583489696e-07, "loss": 0.0064, "step": 9740 }, { "epoch": 17.348754448398576, "grad_norm": 0.12333443015813828, "learning_rate": 1.7214349881918834e-07, "loss": 0.0046, "step": 9750 }, { "epoch": 17.366548042704625, "grad_norm": 0.0852411612868309, "learning_rate": 1.5870724125904845e-07, "loss": 0.0062, "step": 9760 }, { "epoch": 17.384341637010674, "grad_norm": 0.23348741233348846, "learning_rate": 1.4581604009124006e-07, "loss": 0.006, "step": 9770 }, { "epoch": 17.402135231316727, "grad_norm": 0.23921631276607513, "learning_rate": 1.334700362918717e-07, "loss": 0.0054, "step": 9780 }, { "epoch": 17.419928825622776, "grad_norm": 0.11547928303480148, "learning_rate": 1.2166936487486015e-07, "loss": 0.0053, "step": 9790 }, { "epoch": 17.437722419928825, "grad_norm": 0.21013370156288147, "learning_rate": 1.1041415489045914e-07, "loss": 0.0077, "step": 9800 }, { "epoch": 17.455516014234874, "grad_norm": 0.11394015699625015, "learning_rate": 9.970452942384412e-08, "loss": 0.005, "step": 9810 }, { "epoch": 17.473309608540927, "grad_norm": 0.046961430460214615, "learning_rate": 8.954060559375754e-08, "loss": 0.0051, "step": 9820 }, { "epoch": 17.491103202846976, "grad_norm": 0.16323858499526978, "learning_rate": 7.99224945512489e-08, "loss": 0.0081, "step": 9830 }, { "epoch": 17.508896797153024, "grad_norm": 0.1260339915752411, "learning_rate": 7.085030147843675e-08, "loss": 0.0065, "step": 9840 }, { "epoch": 17.526690391459073, "grad_norm": 0.08852656185626984, "learning_rate": 6.232412558736523e-08, "loss": 0.0101, "step": 9850 }, { "epoch": 17.544483985765126, "grad_norm": 0.0741061270236969, "learning_rate": 5.434406011893822e-08, "loss": 0.0126, "step": 9860 }, { "epoch": 17.562277580071175, "grad_norm": 0.12436532974243164, "learning_rate": 4.6910192341864664e-08, "loss": 0.0043, "step": 9870 }, { "epoch": 17.580071174377224, "grad_norm": 0.11481741815805435, "learning_rate": 4.0022603551737035e-08, "loss": 0.0084, "step": 9880 }, { "epoch": 17.597864768683273, "grad_norm": 0.12135223299264908, "learning_rate": 3.3681369070120985e-08, "loss": 0.0045, "step": 9890 }, { "epoch": 17.615658362989326, "grad_norm": 0.042236149311065674, "learning_rate": 2.7886558243744866e-08, "loss": 0.0083, "step": 9900 }, { "epoch": 17.633451957295375, "grad_norm": 0.09669643640518188, "learning_rate": 2.2638234443722596e-08, "loss": 0.0072, "step": 9910 }, { "epoch": 17.651245551601424, "grad_norm": 0.13173197209835052, "learning_rate": 1.7936455064887504e-08, "loss": 0.0127, "step": 9920 }, { "epoch": 17.669039145907472, "grad_norm": 0.09635408222675323, "learning_rate": 1.378127152514841e-08, "loss": 0.0042, "step": 9930 }, { "epoch": 17.68683274021352, "grad_norm": 0.1304931640625, "learning_rate": 1.0172729264917857e-08, "loss": 0.0067, "step": 9940 }, { "epoch": 17.704626334519574, "grad_norm": 0.15431125462055206, "learning_rate": 7.1108677466458215e-09, "loss": 0.0064, "step": 9950 }, { "epoch": 17.722419928825623, "grad_norm": 0.1137930229306221, "learning_rate": 4.595720454353414e-09, "loss": 0.0077, "step": 9960 }, { "epoch": 17.740213523131672, "grad_norm": 0.1015796884894371, "learning_rate": 2.627314893294264e-09, "loss": 0.005, "step": 9970 }, { "epoch": 17.75800711743772, "grad_norm": 0.0713748186826706, "learning_rate": 1.2056725896270048e-09, "loss": 0.0059, "step": 9980 }, { "epoch": 17.775800711743774, "grad_norm": 0.11912448704242706, "learning_rate": 3.308090902098826e-10, "loss": 0.0092, "step": 9990 }, { "epoch": 17.793594306049823, "grad_norm": 0.09733171761035919, "learning_rate": 2.7339624120159555e-12, "loss": 0.007, "step": 10000 }, { "epoch": 17.793594306049823, "step": 10000, "total_flos": 0.0, "train_loss": 0.0020938181336969137, "train_runtime": 3857.1416, "train_samples_per_second": 82.963, "train_steps_per_second": 2.593 } ], "logging_steps": 10, "max_steps": 10000, "num_input_tokens_seen": 0, "num_train_epochs": 18, "save_steps": 1000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 32, "trial_name": null, "trial_params": null }