| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.992, | |
| "eval_steps": 500, | |
| "global_step": 390, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0128, | |
| "grad_norm": 6.179561739341508, | |
| "learning_rate": 1.0256410256410257e-06, | |
| "loss": 1.0206, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 6.520240490637618, | |
| "learning_rate": 2.0512820512820513e-06, | |
| "loss": 1.0474, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 6.068063758320805, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 0.983, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 5.420704263826881, | |
| "learning_rate": 4.102564102564103e-06, | |
| "loss": 0.9415, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 4.293969172521014, | |
| "learning_rate": 5.128205128205128e-06, | |
| "loss": 0.904, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 2.1974274882655007, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.9037, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 2.0228833163950886, | |
| "learning_rate": 7.17948717948718e-06, | |
| "loss": 0.8888, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 4.538063805395591, | |
| "learning_rate": 8.205128205128205e-06, | |
| "loss": 0.8801, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 4.252389202011524, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.836, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 4.436110605549327, | |
| "learning_rate": 1.0256410256410256e-05, | |
| "loss": 0.8296, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 3.8823041402567484, | |
| "learning_rate": 1.1282051282051283e-05, | |
| "loss": 0.8095, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 2.9717526689555442, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.8073, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 2.354865609515145, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.8039, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 2.552282377468162, | |
| "learning_rate": 1.435897435897436e-05, | |
| "loss": 0.7822, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 2.265689603309999, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.7659, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 1.6984108092753578, | |
| "learning_rate": 1.641025641025641e-05, | |
| "loss": 0.7272, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 1.5449269365274938, | |
| "learning_rate": 1.7435897435897438e-05, | |
| "loss": 0.6942, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 1.7079136610440926, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.7241, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 1.3819915185227318, | |
| "learning_rate": 1.9487179487179488e-05, | |
| "loss": 0.7641, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 0.9489351834076887, | |
| "learning_rate": 2.0512820512820512e-05, | |
| "loss": 0.6843, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 1.3593283222883354, | |
| "learning_rate": 2.153846153846154e-05, | |
| "loss": 0.6939, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 1.2477608711140469, | |
| "learning_rate": 2.2564102564102566e-05, | |
| "loss": 0.7347, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 0.8548745456957831, | |
| "learning_rate": 2.3589743589743593e-05, | |
| "loss": 0.6858, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 0.9174288380688765, | |
| "learning_rate": 2.461538461538462e-05, | |
| "loss": 0.7, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.8954594842318306, | |
| "learning_rate": 2.5641025641025646e-05, | |
| "loss": 0.7038, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 0.9787546312270949, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 0.6978, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 0.7943164122223401, | |
| "learning_rate": 2.7692307692307694e-05, | |
| "loss": 0.6901, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 0.8374325358413578, | |
| "learning_rate": 2.871794871794872e-05, | |
| "loss": 0.6963, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 0.7571602636130476, | |
| "learning_rate": 2.9743589743589747e-05, | |
| "loss": 0.6699, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.6510997903360709, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.6468, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 0.7266803459006894, | |
| "learning_rate": 3.1794871794871795e-05, | |
| "loss": 0.6653, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 0.7932643426137853, | |
| "learning_rate": 3.282051282051282e-05, | |
| "loss": 0.6599, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 0.64415005790164, | |
| "learning_rate": 3.384615384615385e-05, | |
| "loss": 0.6647, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 0.8406246458031713, | |
| "learning_rate": 3.4871794871794875e-05, | |
| "loss": 0.6577, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 0.64205205498832, | |
| "learning_rate": 3.58974358974359e-05, | |
| "loss": 0.6263, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 0.6792379946600755, | |
| "learning_rate": 3.692307692307693e-05, | |
| "loss": 0.634, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 0.8629078617807411, | |
| "learning_rate": 3.794871794871795e-05, | |
| "loss": 0.6673, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 0.9726240368310539, | |
| "learning_rate": 3.8974358974358976e-05, | |
| "loss": 0.6695, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 0.9421365503793089, | |
| "learning_rate": 4e-05, | |
| "loss": 0.6477, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 0.9965349379230883, | |
| "learning_rate": 3.9999198907597046e-05, | |
| "loss": 0.6677, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 0.9856308684105368, | |
| "learning_rate": 3.9996795694563096e-05, | |
| "loss": 0.667, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 1.2056604991358801, | |
| "learning_rate": 3.999279055341771e-05, | |
| "loss": 0.6459, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 0.7541204045235476, | |
| "learning_rate": 3.998718380500971e-05, | |
| "loss": 0.6583, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 0.739711876587551, | |
| "learning_rate": 3.997997589849145e-05, | |
| "loss": 0.6331, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.9466329102332102, | |
| "learning_rate": 3.9971167411282835e-05, | |
| "loss": 0.6406, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 0.7604311408723609, | |
| "learning_rate": 3.99607590490251e-05, | |
| "loss": 0.6123, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 0.7909343643072346, | |
| "learning_rate": 3.9948751645524235e-05, | |
| "loss": 0.6565, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 0.9449377386303025, | |
| "learning_rate": 3.9935146162684206e-05, | |
| "loss": 0.6583, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 1.1050016277798747, | |
| "learning_rate": 3.9919943690429906e-05, | |
| "loss": 0.6316, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.6785469671661221, | |
| "learning_rate": 3.9903145446619837e-05, | |
| "loss": 0.6636, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 1.372081476401045, | |
| "learning_rate": 3.9884752776948564e-05, | |
| "loss": 0.631, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 0.8389788605624093, | |
| "learning_rate": 3.9864767154838864e-05, | |
| "loss": 0.6556, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 1.2190488114682667, | |
| "learning_rate": 3.9843190181323744e-05, | |
| "loss": 0.6199, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 0.9851675522850577, | |
| "learning_rate": 3.982002358491817e-05, | |
| "loss": 0.6163, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 1.076411750809413, | |
| "learning_rate": 3.979526922148058e-05, | |
| "loss": 0.6944, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 0.9840283450499598, | |
| "learning_rate": 3.9768929074064206e-05, | |
| "loss": 0.6081, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 0.9717644407164979, | |
| "learning_rate": 3.9741005252758255e-05, | |
| "loss": 0.6324, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 0.8085077940297515, | |
| "learning_rate": 3.971149999451886e-05, | |
| "loss": 0.5804, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 1.0320968667424966, | |
| "learning_rate": 3.9680415662989806e-05, | |
| "loss": 0.6227, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.7684424543370763, | |
| "learning_rate": 3.9647754748313294e-05, | |
| "loss": 0.6618, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 0.9061342327336713, | |
| "learning_rate": 3.96135198669304e-05, | |
| "loss": 0.6427, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 0.7043290606779539, | |
| "learning_rate": 3.957771376137144e-05, | |
| "loss": 0.6218, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 0.6871318467690204, | |
| "learning_rate": 3.954033930003634e-05, | |
| "loss": 0.6368, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 0.6153165855360422, | |
| "learning_rate": 3.9501399476964806e-05, | |
| "loss": 0.6368, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 0.7055520957293149, | |
| "learning_rate": 3.946089741159648e-05, | |
| "loss": 0.6358, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 0.792766727754026, | |
| "learning_rate": 3.9418836348521045e-05, | |
| "loss": 0.6311, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 0.6537700030481322, | |
| "learning_rate": 3.937521965721831e-05, | |
| "loss": 0.6181, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 0.84433593784499, | |
| "learning_rate": 3.933005083178828e-05, | |
| "loss": 0.6214, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 0.566702911682452, | |
| "learning_rate": 3.928333349067125e-05, | |
| "loss": 0.6213, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 0.6267860552106422, | |
| "learning_rate": 3.923507137635792e-05, | |
| "loss": 0.6138, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 0.5804101939145874, | |
| "learning_rate": 3.9185268355089606e-05, | |
| "loss": 0.6112, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 0.5980589876184129, | |
| "learning_rate": 3.913392841654851e-05, | |
| "loss": 0.6363, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 0.6069227182508037, | |
| "learning_rate": 3.9081055673538093e-05, | |
| "loss": 0.6332, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 0.7130878004642133, | |
| "learning_rate": 3.902665436165364e-05, | |
| "loss": 0.6747, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.6531508371926328, | |
| "learning_rate": 3.897072883894291e-05, | |
| "loss": 0.6192, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 0.554386372121544, | |
| "learning_rate": 3.8913283585557054e-05, | |
| "loss": 0.6315, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 0.7125395385908339, | |
| "learning_rate": 3.885432320339167e-05, | |
| "loss": 0.6014, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 0.5012119695263991, | |
| "learning_rate": 3.879385241571817e-05, | |
| "loss": 0.5926, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.0112, | |
| "grad_norm": 1.4447729618541627, | |
| "learning_rate": 3.873187606680543e-05, | |
| "loss": 0.9997, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.024, | |
| "grad_norm": 0.9903543602977524, | |
| "learning_rate": 3.866839912153168e-05, | |
| "loss": 0.5298, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 0.8034247049191908, | |
| "learning_rate": 3.860342666498677e-05, | |
| "loss": 0.4968, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.0496, | |
| "grad_norm": 0.7815324725895815, | |
| "learning_rate": 3.853696390206484e-05, | |
| "loss": 0.506, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.0624, | |
| "grad_norm": 0.9527322152613892, | |
| "learning_rate": 3.846901615704734e-05, | |
| "loss": 0.4821, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 0.6405612921103615, | |
| "learning_rate": 3.839958887317649e-05, | |
| "loss": 0.5149, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.088, | |
| "grad_norm": 1.000605901037876, | |
| "learning_rate": 3.832868761221926e-05, | |
| "loss": 0.502, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.1008, | |
| "grad_norm": 0.7298928491488215, | |
| "learning_rate": 3.825631805402182e-05, | |
| "loss": 0.5091, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 0.7434601142178421, | |
| "learning_rate": 3.818248599605448e-05, | |
| "loss": 0.4983, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.1264, | |
| "grad_norm": 0.8520291545307982, | |
| "learning_rate": 3.810719735294731e-05, | |
| "loss": 0.5416, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.1392, | |
| "grad_norm": 0.7583935832652922, | |
| "learning_rate": 3.8030458156016326e-05, | |
| "loss": 0.5169, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.6174380158073125, | |
| "learning_rate": 3.795227455278029e-05, | |
| "loss": 0.5042, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1648, | |
| "grad_norm": 0.6893018374805242, | |
| "learning_rate": 3.787265280646825e-05, | |
| "loss": 0.4989, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.1776, | |
| "grad_norm": 0.5949027926652132, | |
| "learning_rate": 3.7791599295517825e-05, | |
| "loss": 0.5028, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 0.7040774307991322, | |
| "learning_rate": 3.7709120513064196e-05, | |
| "loss": 0.5051, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.2032, | |
| "grad_norm": 0.7971263532509502, | |
| "learning_rate": 3.762522306641998e-05, | |
| "loss": 0.4837, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.216, | |
| "grad_norm": 0.7999309515165074, | |
| "learning_rate": 3.7539913676545874e-05, | |
| "loss": 0.5355, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 0.6050496036379237, | |
| "learning_rate": 3.745319917751229e-05, | |
| "loss": 0.487, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.2416, | |
| "grad_norm": 0.5940425839558177, | |
| "learning_rate": 3.736508651595188e-05, | |
| "loss": 0.5234, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.2544, | |
| "grad_norm": 0.5385807798263836, | |
| "learning_rate": 3.727558275050301e-05, | |
| "loss": 0.479, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 0.6562464674620533, | |
| "learning_rate": 3.718469505124434e-05, | |
| "loss": 0.5157, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.49329024684263056, | |
| "learning_rate": 3.709243069912041e-05, | |
| "loss": 0.4721, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.2928, | |
| "grad_norm": 0.722778254832613, | |
| "learning_rate": 3.699879708535838e-05, | |
| "loss": 0.5309, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 0.6005460274063595, | |
| "learning_rate": 3.69038017108759e-05, | |
| "loss": 0.4795, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.3184, | |
| "grad_norm": 0.6600589258358752, | |
| "learning_rate": 3.680745218568026e-05, | |
| "loss": 0.5234, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.3312, | |
| "grad_norm": 0.6375742104370463, | |
| "learning_rate": 3.6709756228258735e-05, | |
| "loss": 0.5016, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.6205515891818901, | |
| "learning_rate": 3.6610721664960236e-05, | |
| "loss": 0.4771, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.3568, | |
| "grad_norm": 0.5312197395315701, | |
| "learning_rate": 3.65103564293684e-05, | |
| "loss": 0.5092, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.3696, | |
| "grad_norm": 0.7696367874703324, | |
| "learning_rate": 3.640866856166601e-05, | |
| "loss": 0.4899, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 0.5424057352550463, | |
| "learning_rate": 3.6305666207990886e-05, | |
| "loss": 0.4703, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.3952, | |
| "grad_norm": 0.6081113880779749, | |
| "learning_rate": 3.6201357619783336e-05, | |
| "loss": 0.5012, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.408, | |
| "grad_norm": 0.7418918752168742, | |
| "learning_rate": 3.609575115312511e-05, | |
| "loss": 0.5053, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 0.5684643365962214, | |
| "learning_rate": 3.598885526807003e-05, | |
| "loss": 0.4743, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.4336, | |
| "grad_norm": 0.6369847185988619, | |
| "learning_rate": 3.5880678527966224e-05, | |
| "loss": 0.5135, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.4464000000000001, | |
| "grad_norm": 0.5103147374874601, | |
| "learning_rate": 3.577122959877017e-05, | |
| "loss": 0.491, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 0.5602518499821942, | |
| "learning_rate": 3.566051724835245e-05, | |
| "loss": 0.4814, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.472, | |
| "grad_norm": 0.4876125649946477, | |
| "learning_rate": 3.554855034579532e-05, | |
| "loss": 0.4959, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.4848, | |
| "grad_norm": 0.528135239505157, | |
| "learning_rate": 3.5435337860682304e-05, | |
| "loss": 0.4864, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 0.4754221222261812, | |
| "learning_rate": 3.532088886237956e-05, | |
| "loss": 0.5042, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.5104, | |
| "grad_norm": 0.5181285089123091, | |
| "learning_rate": 3.520521251930941e-05, | |
| "loss": 0.4605, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.5232, | |
| "grad_norm": 0.5032145271296318, | |
| "learning_rate": 3.5088318098215805e-05, | |
| "loss": 0.5083, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.47093215359716967, | |
| "learning_rate": 3.497021496342203e-05, | |
| "loss": 0.513, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.5488, | |
| "grad_norm": 0.44732411449460957, | |
| "learning_rate": 3.485091257608047e-05, | |
| "loss": 0.4776, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.5615999999999999, | |
| "grad_norm": 0.5383669109851341, | |
| "learning_rate": 3.473042049341474e-05, | |
| "loss": 0.4808, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.5487331298122498, | |
| "learning_rate": 3.4608748367954064e-05, | |
| "loss": 0.4703, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.5872000000000002, | |
| "grad_norm": 0.5842231137225898, | |
| "learning_rate": 3.4485905946759965e-05, | |
| "loss": 0.5264, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.5321475952949951, | |
| "learning_rate": 3.4361903070645484e-05, | |
| "loss": 0.4737, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.6298659006916985, | |
| "learning_rate": 3.423674967338681e-05, | |
| "loss": 0.4976, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.6256, | |
| "grad_norm": 0.5640964104142328, | |
| "learning_rate": 3.411045578092754e-05, | |
| "loss": 0.4977, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.6383999999999999, | |
| "grad_norm": 0.5379975293339242, | |
| "learning_rate": 3.398303151057543e-05, | |
| "loss": 0.5002, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.6083548007744284, | |
| "learning_rate": 3.385448707019199e-05, | |
| "loss": 0.4834, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 0.6013368220314496, | |
| "learning_rate": 3.372483275737468e-05, | |
| "loss": 0.5038, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.6768, | |
| "grad_norm": 0.5363991902303341, | |
| "learning_rate": 3.359407895863199e-05, | |
| "loss": 0.4893, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 0.6261620728966693, | |
| "learning_rate": 3.34622361485514e-05, | |
| "loss": 0.5012, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.7024, | |
| "grad_norm": 0.482280647172146, | |
| "learning_rate": 3.332931488896029e-05, | |
| "loss": 0.5094, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.7151999999999998, | |
| "grad_norm": 0.5300823472336182, | |
| "learning_rate": 3.319532582807977e-05, | |
| "loss": 0.4967, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.5444378904723253, | |
| "learning_rate": 3.30602796996717e-05, | |
| "loss": 0.5016, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.7408000000000001, | |
| "grad_norm": 0.6551851476731805, | |
| "learning_rate": 3.2924187322178865e-05, | |
| "loss": 0.5036, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.7536, | |
| "grad_norm": 0.519435106622402, | |
| "learning_rate": 3.278705959785821e-05, | |
| "loss": 0.5061, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 0.5785324392841291, | |
| "learning_rate": 3.2648907511907544e-05, | |
| "loss": 0.4532, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.7792, | |
| "grad_norm": 0.49668827319033826, | |
| "learning_rate": 3.250974213158555e-05, | |
| "loss": 0.5211, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.792, | |
| "grad_norm": 0.593251076907279, | |
| "learning_rate": 3.23695746053251e-05, | |
| "loss": 0.4884, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 0.46902446635324346, | |
| "learning_rate": 3.222841616184025e-05, | |
| "loss": 0.5061, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.8176, | |
| "grad_norm": 0.6207683659059231, | |
| "learning_rate": 3.208627810922665e-05, | |
| "loss": 0.4993, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.8304, | |
| "grad_norm": 0.5195253279958499, | |
| "learning_rate": 3.194317183405573e-05, | |
| "loss": 0.4595, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 0.6975819449169954, | |
| "learning_rate": 3.1799108800462466e-05, | |
| "loss": 0.5538, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.8559999999999999, | |
| "grad_norm": 0.5648243358548889, | |
| "learning_rate": 3.1654100549227024e-05, | |
| "loss": 0.5118, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.8688, | |
| "grad_norm": 0.5615221961071623, | |
| "learning_rate": 3.1508158696850275e-05, | |
| "loss": 0.471, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 0.566095744943867, | |
| "learning_rate": 3.136129493462312e-05, | |
| "loss": 0.5015, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.8944, | |
| "grad_norm": 0.5577281686421962, | |
| "learning_rate": 3.121352102768998e-05, | |
| "loss": 0.5121, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.9072, | |
| "grad_norm": 0.5372555017484676, | |
| "learning_rate": 3.106484881410628e-05, | |
| "loss": 0.47, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.5209354962432491, | |
| "learning_rate": 3.091529020389009e-05, | |
| "loss": 0.4997, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9327999999999999, | |
| "grad_norm": 0.5074275592528913, | |
| "learning_rate": 3.076485717806808e-05, | |
| "loss": 0.5021, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.9456, | |
| "grad_norm": 0.5412134644804545, | |
| "learning_rate": 3.061356178771564e-05, | |
| "loss": 0.5033, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.49553921510731663, | |
| "learning_rate": 3.0461416152991555e-05, | |
| "loss": 0.5171, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.9712, | |
| "grad_norm": 0.522448464010999, | |
| "learning_rate": 3.0308432462167045e-05, | |
| "loss": 0.4864, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.984, | |
| "grad_norm": 0.4339330690006661, | |
| "learning_rate": 3.015462297064936e-05, | |
| "loss": 0.5059, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 0.5223922068945513, | |
| "learning_rate": 3.0000000000000004e-05, | |
| "loss": 0.4814, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.0096, | |
| "grad_norm": 1.3882629873785546, | |
| "learning_rate": 2.98445759369477e-05, | |
| "loss": 0.8654, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.0224, | |
| "grad_norm": 0.6901059071615732, | |
| "learning_rate": 2.9688363232396056e-05, | |
| "loss": 0.3626, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 1.1912948395438292, | |
| "learning_rate": 2.9531374400426158e-05, | |
| "loss": 0.3672, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 0.7804515731155103, | |
| "learning_rate": 2.9373622017294075e-05, | |
| "loss": 0.3536, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.0608, | |
| "grad_norm": 0.6521838733062567, | |
| "learning_rate": 2.9215118720423375e-05, | |
| "loss": 0.3692, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 0.6868619158588961, | |
| "learning_rate": 2.9055877207392752e-05, | |
| "loss": 0.3614, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.0864, | |
| "grad_norm": 0.6645524923761431, | |
| "learning_rate": 2.8895910234918828e-05, | |
| "loss": 0.3434, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.0992, | |
| "grad_norm": 0.577155980754643, | |
| "learning_rate": 2.873523061783426e-05, | |
| "loss": 0.3531, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.7249683831059074, | |
| "learning_rate": 2.8573851228061084e-05, | |
| "loss": 0.3517, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.1248, | |
| "grad_norm": 0.4943363838527499, | |
| "learning_rate": 2.8411784993579633e-05, | |
| "loss": 0.3329, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.1376, | |
| "grad_norm": 0.6108485564325224, | |
| "learning_rate": 2.8249044897392814e-05, | |
| "loss": 0.3626, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.6314863366824466, | |
| "learning_rate": 2.80856439764861e-05, | |
| "loss": 0.355, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.1632, | |
| "grad_norm": 0.49228112393356865, | |
| "learning_rate": 2.792159532078314e-05, | |
| "loss": 0.3397, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 0.6192502285309187, | |
| "learning_rate": 2.77569120720971e-05, | |
| "loss": 0.3838, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 0.5216023225417783, | |
| "learning_rate": 2.7591607423077932e-05, | |
| "loss": 0.3499, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.2016, | |
| "grad_norm": 0.5266014185867368, | |
| "learning_rate": 2.7425694616155474e-05, | |
| "loss": 0.3613, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.2144, | |
| "grad_norm": 0.4535790218801368, | |
| "learning_rate": 2.7259186942478656e-05, | |
| "loss": 0.3168, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 0.5827840204126353, | |
| "learning_rate": 2.7092097740850712e-05, | |
| "loss": 0.3655, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.479082257940368, | |
| "learning_rate": 2.692444039666066e-05, | |
| "loss": 0.3602, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.2528, | |
| "grad_norm": 0.4732493002443038, | |
| "learning_rate": 2.6756228340810946e-05, | |
| "loss": 0.3331, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 0.47513056534076864, | |
| "learning_rate": 2.6587475048641596e-05, | |
| "loss": 0.362, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.2784, | |
| "grad_norm": 0.516837377176613, | |
| "learning_rate": 2.6418194038850634e-05, | |
| "loss": 0.3822, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.2912, | |
| "grad_norm": 0.4097408252564964, | |
| "learning_rate": 2.624839887241115e-05, | |
| "loss": 0.3088, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.48086569877322843, | |
| "learning_rate": 2.607810315148494e-05, | |
| "loss": 0.3717, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.3168, | |
| "grad_norm": 0.40289696710989986, | |
| "learning_rate": 2.5907320518332827e-05, | |
| "loss": 0.3468, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.3296, | |
| "grad_norm": 0.5154848826104516, | |
| "learning_rate": 2.5736064654221808e-05, | |
| "loss": 0.3669, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.4513912189875944, | |
| "learning_rate": 2.5564349278329056e-05, | |
| "loss": 0.3518, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.3552, | |
| "grad_norm": 0.4471685202605346, | |
| "learning_rate": 2.539218814664288e-05, | |
| "loss": 0.3557, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 0.447659485804986, | |
| "learning_rate": 2.521959505086075e-05, | |
| "loss": 0.349, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 0.38932951142327926, | |
| "learning_rate": 2.5046583817284437e-05, | |
| "loss": 0.3442, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.3936, | |
| "grad_norm": 0.5367485624531029, | |
| "learning_rate": 2.487316830571244e-05, | |
| "loss": 0.3335, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.4064, | |
| "grad_norm": 0.46973513616141344, | |
| "learning_rate": 2.4699362408329646e-05, | |
| "loss": 0.381, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 0.5280136305258875, | |
| "learning_rate": 2.4525180048594452e-05, | |
| "loss": 0.35, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.432, | |
| "grad_norm": 0.4051059425176286, | |
| "learning_rate": 2.435063518012335e-05, | |
| "loss": 0.3173, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.4448, | |
| "grad_norm": 0.46858016440367734, | |
| "learning_rate": 2.4175741785573177e-05, | |
| "loss": 0.3326, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 0.43682960790770536, | |
| "learning_rate": 2.4000513875520892e-05, | |
| "loss": 0.3583, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.4704, | |
| "grad_norm": 0.41443963171610004, | |
| "learning_rate": 2.3824965487341247e-05, | |
| "loss": 0.3237, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.4832, | |
| "grad_norm": 0.5103228438139114, | |
| "learning_rate": 2.3649110684082258e-05, | |
| "loss": 0.3652, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.47234890491598913, | |
| "learning_rate": 2.3472963553338614e-05, | |
| "loss": 0.3579, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.5088, | |
| "grad_norm": 0.516116984183649, | |
| "learning_rate": 2.3296538206123134e-05, | |
| "loss": 0.3459, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.5216, | |
| "grad_norm": 0.4229111827853508, | |
| "learning_rate": 2.311984877573636e-05, | |
| "loss": 0.3422, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.6082786723459409, | |
| "learning_rate": 2.2942909416634326e-05, | |
| "loss": 0.3711, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.5472, | |
| "grad_norm": 0.3800199518916105, | |
| "learning_rate": 2.2765734303294666e-05, | |
| "loss": 0.3187, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.5260626023878103, | |
| "learning_rate": 2.2588337629081107e-05, | |
| "loss": 0.3442, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.44418070354528305, | |
| "learning_rate": 2.2410733605106462e-05, | |
| "loss": 0.3479, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.5856, | |
| "grad_norm": 0.49045850716372746, | |
| "learning_rate": 2.2232936459094158e-05, | |
| "loss": 0.3457, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.5984, | |
| "grad_norm": 0.4382230367074546, | |
| "learning_rate": 2.205496043423849e-05, | |
| "loss": 0.3559, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.45380064851396545, | |
| "learning_rate": 2.1876819788063586e-05, | |
| "loss": 0.373, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.624, | |
| "grad_norm": 0.391472045997739, | |
| "learning_rate": 2.16985287912813e-05, | |
| "loss": 0.3353, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.6368, | |
| "grad_norm": 0.3998972133145308, | |
| "learning_rate": 2.1520101726647922e-05, | |
| "loss": 0.3429, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.43828417009133586, | |
| "learning_rate": 2.1341552887820048e-05, | |
| "loss": 0.3697, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.6624, | |
| "grad_norm": 0.38412354348721145, | |
| "learning_rate": 2.1162896578209517e-05, | |
| "loss": 0.3438, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.6752000000000002, | |
| "grad_norm": 0.4723043681439649, | |
| "learning_rate": 2.0984147109837564e-05, | |
| "loss": 0.3343, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.4054934390511446, | |
| "learning_rate": 2.0805318802188307e-05, | |
| "loss": 0.3687, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.7008, | |
| "grad_norm": 0.460243031951981, | |
| "learning_rate": 2.0626425981061608e-05, | |
| "loss": 0.3307, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.7136, | |
| "grad_norm": 0.419602761666458, | |
| "learning_rate": 2.0447482977425465e-05, | |
| "loss": 0.3636, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.42714466560266806, | |
| "learning_rate": 2.0268504126267952e-05, | |
| "loss": 0.3237, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.7392, | |
| "grad_norm": 0.4492085063244264, | |
| "learning_rate": 2.008950376544887e-05, | |
| "loss": 0.3591, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 0.4653464265780463, | |
| "learning_rate": 1.9910496234551132e-05, | |
| "loss": 0.3481, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.4047349537583972, | |
| "learning_rate": 1.9731495873732055e-05, | |
| "loss": 0.3439, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.7776, | |
| "grad_norm": 0.48336600266602925, | |
| "learning_rate": 1.9552517022574542e-05, | |
| "loss": 0.3561, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.7904, | |
| "grad_norm": 0.35256313810730716, | |
| "learning_rate": 1.93735740189384e-05, | |
| "loss": 0.3142, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.4417416551756251, | |
| "learning_rate": 1.9194681197811703e-05, | |
| "loss": 0.3609, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 0.3886379515067039, | |
| "learning_rate": 1.901585289016244e-05, | |
| "loss": 0.329, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.8288, | |
| "grad_norm": 0.4446771840373405, | |
| "learning_rate": 1.8837103421790486e-05, | |
| "loss": 0.3608, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.39819863600110067, | |
| "learning_rate": 1.8658447112179952e-05, | |
| "loss": 0.3522, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.8544, | |
| "grad_norm": 0.46193440190039115, | |
| "learning_rate": 1.8479898273352084e-05, | |
| "loss": 0.3637, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.8672, | |
| "grad_norm": 0.39295346395938, | |
| "learning_rate": 1.83014712087187e-05, | |
| "loss": 0.349, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.36387726269917464, | |
| "learning_rate": 1.8123180211936417e-05, | |
| "loss": 0.3286, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.8928000000000003, | |
| "grad_norm": 0.40463142835548216, | |
| "learning_rate": 1.794503956576152e-05, | |
| "loss": 0.381, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.9055999999999997, | |
| "grad_norm": 0.4234441068480991, | |
| "learning_rate": 1.776706354090585e-05, | |
| "loss": 0.3583, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.4192684392573933, | |
| "learning_rate": 1.758926639489354e-05, | |
| "loss": 0.3431, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.9312, | |
| "grad_norm": 0.46444829833284157, | |
| "learning_rate": 1.7411662370918893e-05, | |
| "loss": 0.3282, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 0.43797887325694834, | |
| "learning_rate": 1.7234265696705344e-05, | |
| "loss": 0.3448, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 0.5058462923269963, | |
| "learning_rate": 1.7057090583365678e-05, | |
| "loss": 0.3866, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.9696, | |
| "grad_norm": 0.446807531387623, | |
| "learning_rate": 1.6880151224263646e-05, | |
| "loss": 0.3628, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.9824, | |
| "grad_norm": 0.4009829004062254, | |
| "learning_rate": 1.6703461793876876e-05, | |
| "loss": 0.3247, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.45180930517606976, | |
| "learning_rate": 1.6527036446661396e-05, | |
| "loss": 0.3628, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 3.008, | |
| "grad_norm": 1.2983208112035076, | |
| "learning_rate": 1.635088931591775e-05, | |
| "loss": 0.5371, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 3.0208, | |
| "grad_norm": 0.62457365577063, | |
| "learning_rate": 1.6175034512658753e-05, | |
| "loss": 0.248, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 3.0336, | |
| "grad_norm": 0.5049511917356196, | |
| "learning_rate": 1.5999486124479115e-05, | |
| "loss": 0.2372, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 3.0464, | |
| "grad_norm": 1.0037467976985412, | |
| "learning_rate": 1.5824258214426833e-05, | |
| "loss": 0.2524, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 3.0592, | |
| "grad_norm": 0.610024081562144, | |
| "learning_rate": 1.5649364819876655e-05, | |
| "loss": 0.2348, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 3.072, | |
| "grad_norm": 0.5191183934301941, | |
| "learning_rate": 1.547481995140556e-05, | |
| "loss": 0.2282, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.0848, | |
| "grad_norm": 0.5661449137991017, | |
| "learning_rate": 1.5300637591670357e-05, | |
| "loss": 0.2386, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 3.0976, | |
| "grad_norm": 0.5242450041931568, | |
| "learning_rate": 1.5126831694287564e-05, | |
| "loss": 0.2305, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 3.1104, | |
| "grad_norm": 0.4124677940285404, | |
| "learning_rate": 1.4953416182715566e-05, | |
| "loss": 0.2249, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 3.1232, | |
| "grad_norm": 0.4555633745952224, | |
| "learning_rate": 1.478040494913926e-05, | |
| "loss": 0.2231, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.136, | |
| "grad_norm": 0.46533807836935737, | |
| "learning_rate": 1.460781185335713e-05, | |
| "loss": 0.2274, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 3.1488, | |
| "grad_norm": 0.4916655288312457, | |
| "learning_rate": 1.443565072167095e-05, | |
| "loss": 0.224, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 3.1616, | |
| "grad_norm": 0.4246464250321595, | |
| "learning_rate": 1.4263935345778202e-05, | |
| "loss": 0.2148, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 3.1744, | |
| "grad_norm": 0.4040331561623456, | |
| "learning_rate": 1.409267948166718e-05, | |
| "loss": 0.2246, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 3.1872, | |
| "grad_norm": 0.43503689101205745, | |
| "learning_rate": 1.3921896848515064e-05, | |
| "loss": 0.233, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.440595981783615, | |
| "learning_rate": 1.3751601127588849e-05, | |
| "loss": 0.2421, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.2128, | |
| "grad_norm": 0.4664622899483918, | |
| "learning_rate": 1.3581805961149371e-05, | |
| "loss": 0.2377, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 3.2256, | |
| "grad_norm": 0.35873260242235505, | |
| "learning_rate": 1.341252495135841e-05, | |
| "loss": 0.2034, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 3.2384, | |
| "grad_norm": 0.35624864290116753, | |
| "learning_rate": 1.324377165918906e-05, | |
| "loss": 0.2178, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 3.2512, | |
| "grad_norm": 0.41061918901293076, | |
| "learning_rate": 1.3075559603339354e-05, | |
| "loss": 0.2328, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 3.2640000000000002, | |
| "grad_norm": 0.4116441819384426, | |
| "learning_rate": 1.2907902259149287e-05, | |
| "loss": 0.243, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 3.2768, | |
| "grad_norm": 0.3659258653061729, | |
| "learning_rate": 1.274081305752135e-05, | |
| "loss": 0.1931, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 3.2896, | |
| "grad_norm": 0.41164535784552553, | |
| "learning_rate": 1.2574305383844528e-05, | |
| "loss": 0.2395, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 3.3024, | |
| "grad_norm": 0.3714357168831428, | |
| "learning_rate": 1.2408392576922075e-05, | |
| "loss": 0.2095, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 3.3152, | |
| "grad_norm": 0.36157473075455976, | |
| "learning_rate": 1.2243087927902905e-05, | |
| "loss": 0.217, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 3.328, | |
| "grad_norm": 0.3946714303991262, | |
| "learning_rate": 1.2078404679216864e-05, | |
| "loss": 0.2278, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 3.3407999999999998, | |
| "grad_norm": 0.3655965980173958, | |
| "learning_rate": 1.1914356023513904e-05, | |
| "loss": 0.2188, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 3.3536, | |
| "grad_norm": 0.3579101411723521, | |
| "learning_rate": 1.1750955102607193e-05, | |
| "loss": 0.2292, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 3.3664, | |
| "grad_norm": 0.34660229710321105, | |
| "learning_rate": 1.1588215006420374e-05, | |
| "loss": 0.222, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 3.3792, | |
| "grad_norm": 0.35643645617481806, | |
| "learning_rate": 1.1426148771938915e-05, | |
| "loss": 0.2208, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 3.392, | |
| "grad_norm": 0.37206681339121644, | |
| "learning_rate": 1.1264769382165748e-05, | |
| "loss": 0.2322, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 3.4048, | |
| "grad_norm": 0.3472660222316565, | |
| "learning_rate": 1.110408976508118e-05, | |
| "loss": 0.2149, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 3.4176, | |
| "grad_norm": 0.3529717167350539, | |
| "learning_rate": 1.094412279260726e-05, | |
| "loss": 0.2244, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 3.4304, | |
| "grad_norm": 0.3330845729201185, | |
| "learning_rate": 1.0784881279576635e-05, | |
| "loss": 0.1943, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 3.4432, | |
| "grad_norm": 0.3461833589727701, | |
| "learning_rate": 1.0626377982705929e-05, | |
| "loss": 0.2298, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 3.456, | |
| "grad_norm": 0.36693572508198286, | |
| "learning_rate": 1.0468625599573842e-05, | |
| "loss": 0.2378, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.4688, | |
| "grad_norm": 0.3297809019911421, | |
| "learning_rate": 1.0311636767603952e-05, | |
| "loss": 0.2365, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 3.4816, | |
| "grad_norm": 0.3371706825027726, | |
| "learning_rate": 1.0155424063052306e-05, | |
| "loss": 0.2169, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 3.4944, | |
| "grad_norm": 0.3548778903019076, | |
| "learning_rate": 1.0000000000000006e-05, | |
| "loss": 0.2145, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 3.5072, | |
| "grad_norm": 0.35605619738770694, | |
| "learning_rate": 9.84537702935065e-06, | |
| "loss": 0.2221, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 0.33800439377127767, | |
| "learning_rate": 9.691567537832964e-06, | |
| "loss": 0.2299, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.5328, | |
| "grad_norm": 0.3480460703456158, | |
| "learning_rate": 9.538583847008452e-06, | |
| "loss": 0.2392, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 3.5456, | |
| "grad_norm": 0.3281483980080044, | |
| "learning_rate": 9.386438212284372e-06, | |
| "loss": 0.2, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 3.5584, | |
| "grad_norm": 0.3270952161858396, | |
| "learning_rate": 9.235142821931928e-06, | |
| "loss": 0.1984, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 3.5712, | |
| "grad_norm": 0.3519365249097174, | |
| "learning_rate": 9.084709796109907e-06, | |
| "loss": 0.2284, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 3.584, | |
| "grad_norm": 0.33509494620403457, | |
| "learning_rate": 8.93515118589373e-06, | |
| "loss": 0.2324, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.5968, | |
| "grad_norm": 0.35667650227094605, | |
| "learning_rate": 8.786478972310023e-06, | |
| "loss": 0.2427, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 3.6096, | |
| "grad_norm": 0.3402455499252984, | |
| "learning_rate": 8.638705065376887e-06, | |
| "loss": 0.2194, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 3.6224, | |
| "grad_norm": 0.3529815166501557, | |
| "learning_rate": 8.491841303149728e-06, | |
| "loss": 0.2301, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 3.6352, | |
| "grad_norm": 0.34858234800103355, | |
| "learning_rate": 8.345899450772975e-06, | |
| "loss": 0.2286, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 3.648, | |
| "grad_norm": 0.3116517591197441, | |
| "learning_rate": 8.200891199537549e-06, | |
| "loss": 0.218, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 3.6608, | |
| "grad_norm": 0.362614220554881, | |
| "learning_rate": 8.056828165944282e-06, | |
| "loss": 0.2489, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 3.6736, | |
| "grad_norm": 0.3193597242694849, | |
| "learning_rate": 7.913721890773354e-06, | |
| "loss": 0.2087, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 3.6864, | |
| "grad_norm": 0.3541019741295966, | |
| "learning_rate": 7.771583838159756e-06, | |
| "loss": 0.233, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.6992000000000003, | |
| "grad_norm": 0.3383994716720103, | |
| "learning_rate": 7.630425394674903e-06, | |
| "loss": 0.1964, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 3.7119999999999997, | |
| "grad_norm": 0.36274143319698593, | |
| "learning_rate": 7.49025786841445e-06, | |
| "loss": 0.2288, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.7248, | |
| "grad_norm": 0.3348702462797562, | |
| "learning_rate": 7.3510924880924575e-06, | |
| "loss": 0.2146, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 3.7376, | |
| "grad_norm": 0.3695748421027966, | |
| "learning_rate": 7.212940402141808e-06, | |
| "loss": 0.228, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 3.7504, | |
| "grad_norm": 0.33021262094626064, | |
| "learning_rate": 7.075812677821145e-06, | |
| "loss": 0.2009, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 3.7632, | |
| "grad_norm": 0.36604120586687305, | |
| "learning_rate": 6.939720300328303e-06, | |
| "loss": 0.2591, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 3.776, | |
| "grad_norm": 0.31871783826834016, | |
| "learning_rate": 6.8046741719202385e-06, | |
| "loss": 0.2118, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 3.7888, | |
| "grad_norm": 0.32460570514714143, | |
| "learning_rate": 6.67068511103971e-06, | |
| "loss": 0.2105, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 3.8016, | |
| "grad_norm": 0.338337791171013, | |
| "learning_rate": 6.537763851448593e-06, | |
| "loss": 0.2168, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 3.8144, | |
| "grad_norm": 0.3234769353488224, | |
| "learning_rate": 6.4059210413680175e-06, | |
| "loss": 0.2047, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 3.8272, | |
| "grad_norm": 0.3094397848308515, | |
| "learning_rate": 6.275167242625331e-06, | |
| "loss": 0.2186, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.33047327473001153, | |
| "learning_rate": 6.145512929808013e-06, | |
| "loss": 0.2244, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.8528000000000002, | |
| "grad_norm": 0.3285191192492445, | |
| "learning_rate": 6.016968489424572e-06, | |
| "loss": 0.218, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 3.8656, | |
| "grad_norm": 0.31471450241774845, | |
| "learning_rate": 5.889544219072465e-06, | |
| "loss": 0.2093, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 3.8784, | |
| "grad_norm": 0.331897029066441, | |
| "learning_rate": 5.7632503266131925e-06, | |
| "loss": 0.2326, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 3.8912, | |
| "grad_norm": 0.3047595000934823, | |
| "learning_rate": 5.638096929354522e-06, | |
| "loss": 0.2102, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 3.904, | |
| "grad_norm": 0.3433370172144795, | |
| "learning_rate": 5.514094053240035e-06, | |
| "loss": 0.2268, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 3.9168, | |
| "grad_norm": 0.3109256937258501, | |
| "learning_rate": 5.39125163204594e-06, | |
| "loss": 0.2138, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 3.9295999999999998, | |
| "grad_norm": 0.3141390213523075, | |
| "learning_rate": 5.269579506585259e-06, | |
| "loss": 0.2198, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 3.9424, | |
| "grad_norm": 0.35766115277650484, | |
| "learning_rate": 5.149087423919541e-06, | |
| "loss": 0.2478, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 3.9552, | |
| "grad_norm": 0.3300632217120457, | |
| "learning_rate": 5.029785036577976e-06, | |
| "loss": 0.2244, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 3.968, | |
| "grad_norm": 0.3288021090948234, | |
| "learning_rate": 4.911681901784198e-06, | |
| "loss": 0.2298, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.9808, | |
| "grad_norm": 0.330907353122105, | |
| "learning_rate": 4.794787480690597e-06, | |
| "loss": 0.2108, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 3.9936, | |
| "grad_norm": 0.31884871326446496, | |
| "learning_rate": 4.679111137620442e-06, | |
| "loss": 0.2036, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 4.0064, | |
| "grad_norm": 1.2719412720472383, | |
| "learning_rate": 4.5646621393177e-06, | |
| "loss": 0.3723, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 4.0192, | |
| "grad_norm": 0.5636831657755029, | |
| "learning_rate": 4.451449654204685e-06, | |
| "loss": 0.1749, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 4.032, | |
| "grad_norm": 0.4396489339437205, | |
| "learning_rate": 4.339482751647557e-06, | |
| "loss": 0.1563, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 4.0448, | |
| "grad_norm": 0.32697340504946937, | |
| "learning_rate": 4.228770401229824e-06, | |
| "loss": 0.1616, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 4.0576, | |
| "grad_norm": 0.31967035026341384, | |
| "learning_rate": 4.119321472033779e-06, | |
| "loss": 0.1572, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 4.0704, | |
| "grad_norm": 0.3873647336085365, | |
| "learning_rate": 4.011144731929981e-06, | |
| "loss": 0.162, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 4.0832, | |
| "grad_norm": 0.44005462772164966, | |
| "learning_rate": 3.904248846874894e-06, | |
| "loss": 0.1548, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 4.096, | |
| "grad_norm": 0.4787246213275118, | |
| "learning_rate": 3.7986423802166705e-06, | |
| "loss": 0.1439, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 4.1088, | |
| "grad_norm": 0.4701137147753205, | |
| "learning_rate": 3.694333792009115e-06, | |
| "loss": 0.1515, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 4.1216, | |
| "grad_norm": 0.363840918865789, | |
| "learning_rate": 3.5913314383339937e-06, | |
| "loss": 0.1431, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 4.1344, | |
| "grad_norm": 0.32330692016940504, | |
| "learning_rate": 3.4896435706316e-06, | |
| "loss": 0.1564, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 4.1472, | |
| "grad_norm": 0.31871053706494956, | |
| "learning_rate": 3.3892783350397675e-06, | |
| "loss": 0.164, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 0.3532258948335278, | |
| "learning_rate": 3.290243771741275e-06, | |
| "loss": 0.1583, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 4.1728, | |
| "grad_norm": 0.34451700230557614, | |
| "learning_rate": 3.1925478143197418e-06, | |
| "loss": 0.1304, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 4.1856, | |
| "grad_norm": 0.35470123133395637, | |
| "learning_rate": 3.0961982891241083e-06, | |
| "loss": 0.1456, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 4.1984, | |
| "grad_norm": 0.34182176798380015, | |
| "learning_rate": 3.001202914641628e-06, | |
| "loss": 0.1541, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 4.2112, | |
| "grad_norm": 0.3082607040683442, | |
| "learning_rate": 2.907569300879596e-06, | |
| "loss": 0.1427, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 4.224, | |
| "grad_norm": 0.31172848120144736, | |
| "learning_rate": 2.815304948755664e-06, | |
| "loss": 0.1573, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 4.2368, | |
| "grad_norm": 0.31793858235851286, | |
| "learning_rate": 2.7244172494969978e-06, | |
| "loss": 0.1624, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 4.2496, | |
| "grad_norm": 0.2978587996579456, | |
| "learning_rate": 2.6349134840481294e-06, | |
| "loss": 0.1501, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 4.2624, | |
| "grad_norm": 0.32035478026467956, | |
| "learning_rate": 2.546800822487714e-06, | |
| "loss": 0.1566, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 4.2752, | |
| "grad_norm": 0.344577009739516, | |
| "learning_rate": 2.4600863234541338e-06, | |
| "loss": 0.1539, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 4.288, | |
| "grad_norm": 0.3214238404135311, | |
| "learning_rate": 2.374776933580025e-06, | |
| "loss": 0.1489, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 4.3008, | |
| "grad_norm": 0.31140702615395655, | |
| "learning_rate": 2.2908794869358044e-06, | |
| "loss": 0.1423, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 4.3136, | |
| "grad_norm": 0.31395589480464403, | |
| "learning_rate": 2.2084007044821764e-06, | |
| "loss": 0.1531, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 4.3264, | |
| "grad_norm": 0.30646569711003113, | |
| "learning_rate": 2.127347193531757e-06, | |
| "loss": 0.1342, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 4.3392, | |
| "grad_norm": 0.30423009716580257, | |
| "learning_rate": 2.0477254472197237e-06, | |
| "loss": 0.1535, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 4.352, | |
| "grad_norm": 0.2931307624762192, | |
| "learning_rate": 1.96954184398368e-06, | |
| "loss": 0.1508, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 4.3648, | |
| "grad_norm": 0.2960591674519737, | |
| "learning_rate": 1.8928026470526917e-06, | |
| "loss": 0.1353, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 4.3776, | |
| "grad_norm": 0.2958903566927966, | |
| "learning_rate": 1.817514003945524e-06, | |
| "loss": 0.1471, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 4.3904, | |
| "grad_norm": 0.31694782910594493, | |
| "learning_rate": 1.743681945978184e-06, | |
| "loss": 0.17, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 4.4032, | |
| "grad_norm": 0.2742196092341873, | |
| "learning_rate": 1.6713123877807413e-06, | |
| "loss": 0.1239, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 4.416, | |
| "grad_norm": 0.30223757951718977, | |
| "learning_rate": 1.6004111268235156e-06, | |
| "loss": 0.144, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 4.4288, | |
| "grad_norm": 0.29879390315765, | |
| "learning_rate": 1.5309838429526714e-06, | |
| "loss": 0.1486, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 4.4416, | |
| "grad_norm": 0.28594016356565966, | |
| "learning_rate": 1.4630360979351644e-06, | |
| "loss": 0.1392, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 4.4544, | |
| "grad_norm": 0.29888745723626703, | |
| "learning_rate": 1.396573335013236e-06, | |
| "loss": 0.1596, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 4.4672, | |
| "grad_norm": 0.28634236366209714, | |
| "learning_rate": 1.3316008784683265e-06, | |
| "loss": 0.1449, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 0.29616342467675494, | |
| "learning_rate": 1.2681239331945695e-06, | |
| "loss": 0.1454, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.4928, | |
| "grad_norm": 0.2942141713639158, | |
| "learning_rate": 1.2061475842818337e-06, | |
| "loss": 0.1637, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 4.5056, | |
| "grad_norm": 0.2949946537785864, | |
| "learning_rate": 1.1456767966083393e-06, | |
| "loss": 0.1502, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 4.5184, | |
| "grad_norm": 0.3101641767916716, | |
| "learning_rate": 1.086716414442952e-06, | |
| "loss": 0.1675, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 4.5312, | |
| "grad_norm": 0.26935606117473687, | |
| "learning_rate": 1.0292711610570904e-06, | |
| "loss": 0.1335, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 4.5440000000000005, | |
| "grad_norm": 0.3052308309740731, | |
| "learning_rate": 9.733456383463658e-07, | |
| "loss": 0.172, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 4.5568, | |
| "grad_norm": 0.28686005733447584, | |
| "learning_rate": 9.189443264619102e-07, | |
| "loss": 0.1482, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 4.5696, | |
| "grad_norm": 0.2848297639328367, | |
| "learning_rate": 8.660715834514977e-07, | |
| "loss": 0.1363, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 4.5824, | |
| "grad_norm": 0.30644214390080593, | |
| "learning_rate": 8.147316449103959e-07, | |
| "loss": 0.1565, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 4.5952, | |
| "grad_norm": 0.2990242198882543, | |
| "learning_rate": 7.649286236420806e-07, | |
| "loss": 0.1621, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 4.608, | |
| "grad_norm": 0.28804995237437764, | |
| "learning_rate": 7.166665093287539e-07, | |
| "loss": 0.145, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.6208, | |
| "grad_norm": 0.2920307039959378, | |
| "learning_rate": 6.69949168211721e-07, | |
| "loss": 0.1476, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 4.6336, | |
| "grad_norm": 0.27507951821247006, | |
| "learning_rate": 6.247803427816945e-07, | |
| "loss": 0.1459, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 4.6464, | |
| "grad_norm": 0.2895934268479852, | |
| "learning_rate": 5.811636514789598e-07, | |
| "loss": 0.1552, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 4.6592, | |
| "grad_norm": 0.29210409891497313, | |
| "learning_rate": 5.391025884035239e-07, | |
| "loss": 0.1529, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 4.672, | |
| "grad_norm": 0.28599592201300683, | |
| "learning_rate": 4.986005230351954e-07, | |
| "loss": 0.1328, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 4.6848, | |
| "grad_norm": 0.3001037075276377, | |
| "learning_rate": 4.5966069996365993e-07, | |
| "loss": 0.1425, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.6975999999999996, | |
| "grad_norm": 0.28100520397245593, | |
| "learning_rate": 4.22286238628562e-07, | |
| "loss": 0.1464, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 4.7104, | |
| "grad_norm": 0.2787932239006293, | |
| "learning_rate": 3.8648013306960664e-07, | |
| "loss": 0.1468, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 4.7232, | |
| "grad_norm": 0.30019840991084656, | |
| "learning_rate": 3.522452516867048e-07, | |
| "loss": 0.1632, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 4.736, | |
| "grad_norm": 0.29460001059032875, | |
| "learning_rate": 3.1958433701019697e-07, | |
| "loss": 0.1595, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.7488, | |
| "grad_norm": 0.28180559660815996, | |
| "learning_rate": 2.8850000548115155e-07, | |
| "loss": 0.136, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 4.7616, | |
| "grad_norm": 0.2828671042606847, | |
| "learning_rate": 2.5899474724174313e-07, | |
| "loss": 0.1461, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 4.7744, | |
| "grad_norm": 0.29019191527470195, | |
| "learning_rate": 2.3107092593579905e-07, | |
| "loss": 0.1504, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 4.7872, | |
| "grad_norm": 0.2903778270872329, | |
| "learning_rate": 2.0473077851942858e-07, | |
| "loss": 0.1465, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.2860541971867706, | |
| "learning_rate": 1.799764150818306e-07, | |
| "loss": 0.1333, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 4.8128, | |
| "grad_norm": 0.2827493957180667, | |
| "learning_rate": 1.5680981867625566e-07, | |
| "loss": 0.1427, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 4.8256, | |
| "grad_norm": 0.28683067998312556, | |
| "learning_rate": 1.3523284516113955e-07, | |
| "loss": 0.1534, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 4.8384, | |
| "grad_norm": 0.273758499132957, | |
| "learning_rate": 1.1524722305144231e-07, | |
| "loss": 0.1403, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 4.8512, | |
| "grad_norm": 0.29326741405244794, | |
| "learning_rate": 9.685455338016347e-08, | |
| "loss": 0.1604, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 4.864, | |
| "grad_norm": 0.2799867747524855, | |
| "learning_rate": 8.005630957010014e-08, | |
| "loss": 0.1527, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.8768, | |
| "grad_norm": 0.31253700269776347, | |
| "learning_rate": 6.485383731580142e-08, | |
| "loss": 0.1905, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 4.8896, | |
| "grad_norm": 0.26936018760609604, | |
| "learning_rate": 5.1248354475768034e-08, | |
| "loss": 0.1339, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 4.9024, | |
| "grad_norm": 0.2771980240463275, | |
| "learning_rate": 3.924095097489922e-08, | |
| "loss": 0.1319, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 4.9152000000000005, | |
| "grad_norm": 0.30101024285223715, | |
| "learning_rate": 2.8832588717164766e-08, | |
| "loss": 0.1623, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.928, | |
| "grad_norm": 0.28925832042872623, | |
| "learning_rate": 2.0024101508555604e-08, | |
| "loss": 0.1515, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 4.9408, | |
| "grad_norm": 0.2708006774131746, | |
| "learning_rate": 1.281619499029274e-08, | |
| "loss": 0.1404, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 4.9536, | |
| "grad_norm": 0.29273982914754165, | |
| "learning_rate": 7.209446582292501e-09, | |
| "loss": 0.1565, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 4.9664, | |
| "grad_norm": 0.2915386140542476, | |
| "learning_rate": 3.2043054369057523e-09, | |
| "loss": 0.1517, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 4.9792, | |
| "grad_norm": 0.29280128130146194, | |
| "learning_rate": 8.010924029533406e-10, | |
| "loss": 0.1374, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "grad_norm": 0.29339802546054505, | |
| "learning_rate": 0.0, | |
| "loss": 0.144, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "step": 390, | |
| "total_flos": 7.185110241892106e+17, | |
| "train_loss": 0.3879696773030819, | |
| "train_runtime": 13223.07, | |
| "train_samples_per_second": 3.781, | |
| "train_steps_per_second": 0.029 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 390, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 7.185110241892106e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |