| { | |
| "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": 6.043263430499764, | |
| "learning_rate": 1.0256410256410257e-06, | |
| "loss": 0.7836, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 6.310560948487243, | |
| "learning_rate": 2.0512820512820513e-06, | |
| "loss": 0.8034, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 5.8939073795241885, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 0.7847, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 5.553026986804903, | |
| "learning_rate": 4.102564102564103e-06, | |
| "loss": 0.7928, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 4.010298225268131, | |
| "learning_rate": 5.128205128205128e-06, | |
| "loss": 0.7599, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 2.1935384971557226, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.7161, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 1.9440013995862226, | |
| "learning_rate": 7.17948717948718e-06, | |
| "loss": 0.7034, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 3.985864671611728, | |
| "learning_rate": 8.205128205128205e-06, | |
| "loss": 0.721, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 4.069900917632356, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.7091, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 4.126279356367066, | |
| "learning_rate": 1.0256410256410256e-05, | |
| "loss": 0.672, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 3.644371774862054, | |
| "learning_rate": 1.1282051282051283e-05, | |
| "loss": 0.671, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 1.9902380119897582, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.6379, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 1.6301743800145354, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.6346, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 2.2053366104355407, | |
| "learning_rate": 1.435897435897436e-05, | |
| "loss": 0.6231, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 1.6854029572552016, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.6133, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 1.1011671884029717, | |
| "learning_rate": 1.641025641025641e-05, | |
| "loss": 0.6029, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 0.9640548560162108, | |
| "learning_rate": 1.7435897435897438e-05, | |
| "loss": 0.5699, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 0.9146734654838942, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.5778, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 0.8553855988115932, | |
| "learning_rate": 1.9487179487179488e-05, | |
| "loss": 0.5547, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 0.8454309875322201, | |
| "learning_rate": 2.0512820512820512e-05, | |
| "loss": 0.5492, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 0.7568213118638362, | |
| "learning_rate": 2.153846153846154e-05, | |
| "loss": 0.5533, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 1.0266647925634058, | |
| "learning_rate": 2.2564102564102566e-05, | |
| "loss": 0.5435, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 0.6665073043898262, | |
| "learning_rate": 2.3589743589743593e-05, | |
| "loss": 0.5424, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 0.8665766094094925, | |
| "learning_rate": 2.461538461538462e-05, | |
| "loss": 0.5344, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.6922518001452801, | |
| "learning_rate": 2.5641025641025646e-05, | |
| "loss": 0.543, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 0.5712640638465311, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 0.5298, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 0.7837047704560033, | |
| "learning_rate": 2.7692307692307694e-05, | |
| "loss": 0.5139, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 0.7189297685531614, | |
| "learning_rate": 2.871794871794872e-05, | |
| "loss": 0.5094, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 0.7336799483880951, | |
| "learning_rate": 2.9743589743589747e-05, | |
| "loss": 0.5148, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.6705489035504345, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.5088, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 0.8470116729183148, | |
| "learning_rate": 3.1794871794871795e-05, | |
| "loss": 0.5247, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 0.7125507443505569, | |
| "learning_rate": 3.282051282051282e-05, | |
| "loss": 0.496, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 0.7089066688749583, | |
| "learning_rate": 3.384615384615385e-05, | |
| "loss": 0.5109, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 0.7018992525781913, | |
| "learning_rate": 3.4871794871794875e-05, | |
| "loss": 0.5185, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 0.7341991583825473, | |
| "learning_rate": 3.58974358974359e-05, | |
| "loss": 0.4933, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 0.7310749857020874, | |
| "learning_rate": 3.692307692307693e-05, | |
| "loss": 0.4994, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 0.667438034843974, | |
| "learning_rate": 3.794871794871795e-05, | |
| "loss": 0.5051, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 0.5322067520611612, | |
| "learning_rate": 3.8974358974358976e-05, | |
| "loss": 0.5131, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 0.7335357575732063, | |
| "learning_rate": 4e-05, | |
| "loss": 0.5017, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 0.9672918971294094, | |
| "learning_rate": 3.9999198907597046e-05, | |
| "loss": 0.49, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 1.0463017393667762, | |
| "learning_rate": 3.9996795694563096e-05, | |
| "loss": 0.4992, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 0.6077670911152491, | |
| "learning_rate": 3.999279055341771e-05, | |
| "loss": 0.4907, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 0.7347274055455547, | |
| "learning_rate": 3.998718380500971e-05, | |
| "loss": 0.503, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 0.8765937307198725, | |
| "learning_rate": 3.997997589849145e-05, | |
| "loss": 0.4854, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.7294640060630456, | |
| "learning_rate": 3.9971167411282835e-05, | |
| "loss": 0.4932, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 0.6233284774906399, | |
| "learning_rate": 3.99607590490251e-05, | |
| "loss": 0.4795, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 0.44824617229904506, | |
| "learning_rate": 3.9948751645524235e-05, | |
| "loss": 0.4873, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 0.5165123281093215, | |
| "learning_rate": 3.9935146162684206e-05, | |
| "loss": 0.4832, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 0.508901612448862, | |
| "learning_rate": 3.9919943690429906e-05, | |
| "loss": 0.4692, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.4744837852539331, | |
| "learning_rate": 3.9903145446619837e-05, | |
| "loss": 0.497, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 0.5367594129509015, | |
| "learning_rate": 3.9884752776948564e-05, | |
| "loss": 0.4783, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 0.6188596617991571, | |
| "learning_rate": 3.9864767154838864e-05, | |
| "loss": 0.4717, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 0.45499137251329597, | |
| "learning_rate": 3.9843190181323744e-05, | |
| "loss": 0.4824, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 0.6857021856152895, | |
| "learning_rate": 3.982002358491817e-05, | |
| "loss": 0.483, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 0.8143450724362253, | |
| "learning_rate": 3.979526922148058e-05, | |
| "loss": 0.4654, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 0.47538243349853354, | |
| "learning_rate": 3.9768929074064206e-05, | |
| "loss": 0.4834, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 0.5778480967529612, | |
| "learning_rate": 3.9741005252758255e-05, | |
| "loss": 0.4802, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 0.6585225185189254, | |
| "learning_rate": 3.971149999451886e-05, | |
| "loss": 0.4915, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 0.8345475848946361, | |
| "learning_rate": 3.9680415662989806e-05, | |
| "loss": 0.4741, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.5473551682440083, | |
| "learning_rate": 3.9647754748313294e-05, | |
| "loss": 0.47, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 0.5671642894915455, | |
| "learning_rate": 3.96135198669304e-05, | |
| "loss": 0.4593, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 0.836342823987611, | |
| "learning_rate": 3.957771376137144e-05, | |
| "loss": 0.4676, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 0.7423876977001905, | |
| "learning_rate": 3.954033930003634e-05, | |
| "loss": 0.4589, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 0.5475000201138382, | |
| "learning_rate": 3.9501399476964806e-05, | |
| "loss": 0.4583, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 0.6691648857012322, | |
| "learning_rate": 3.946089741159648e-05, | |
| "loss": 0.4804, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 0.4822853096365147, | |
| "learning_rate": 3.9418836348521045e-05, | |
| "loss": 0.4644, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 0.6188943131593398, | |
| "learning_rate": 3.937521965721831e-05, | |
| "loss": 0.4765, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 0.7414546337363769, | |
| "learning_rate": 3.933005083178828e-05, | |
| "loss": 0.4724, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 0.7988134061999206, | |
| "learning_rate": 3.928333349067125e-05, | |
| "loss": 0.4674, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 0.5893484755837565, | |
| "learning_rate": 3.923507137635792e-05, | |
| "loss": 0.4716, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 0.5371932066523464, | |
| "learning_rate": 3.9185268355089606e-05, | |
| "loss": 0.4677, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 0.7251817376254791, | |
| "learning_rate": 3.913392841654851e-05, | |
| "loss": 0.4673, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 0.5536359711767566, | |
| "learning_rate": 3.9081055673538093e-05, | |
| "loss": 0.4669, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 0.5527745257450913, | |
| "learning_rate": 3.902665436165364e-05, | |
| "loss": 0.4627, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.46041638151149444, | |
| "learning_rate": 3.897072883894291e-05, | |
| "loss": 0.4543, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 0.516634006665692, | |
| "learning_rate": 3.8913283585557054e-05, | |
| "loss": 0.4628, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 0.5051437079463834, | |
| "learning_rate": 3.885432320339167e-05, | |
| "loss": 0.46, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 0.5412224987857337, | |
| "learning_rate": 3.879385241571817e-05, | |
| "loss": 0.4634, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.0112, | |
| "grad_norm": 1.1617596045136462, | |
| "learning_rate": 3.873187606680543e-05, | |
| "loss": 0.8202, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.024, | |
| "grad_norm": 1.0856645039820778, | |
| "learning_rate": 3.866839912153168e-05, | |
| "loss": 0.4194, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 0.8959919804726858, | |
| "learning_rate": 3.860342666498677e-05, | |
| "loss": 0.4463, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.0496, | |
| "grad_norm": 1.0674266209473993, | |
| "learning_rate": 3.853696390206484e-05, | |
| "loss": 0.4267, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.0624, | |
| "grad_norm": 0.6965125717910822, | |
| "learning_rate": 3.846901615704734e-05, | |
| "loss": 0.4317, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 1.0733609299373061, | |
| "learning_rate": 3.839958887317649e-05, | |
| "loss": 0.4298, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.088, | |
| "grad_norm": 0.6906292461101706, | |
| "learning_rate": 3.832868761221926e-05, | |
| "loss": 0.4273, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.1008, | |
| "grad_norm": 0.8510701025692062, | |
| "learning_rate": 3.825631805402182e-05, | |
| "loss": 0.4139, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 0.6428450767785712, | |
| "learning_rate": 3.818248599605448e-05, | |
| "loss": 0.4325, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.1264, | |
| "grad_norm": 0.7220771480940529, | |
| "learning_rate": 3.810719735294731e-05, | |
| "loss": 0.4266, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.1392, | |
| "grad_norm": 0.6926687729509802, | |
| "learning_rate": 3.8030458156016326e-05, | |
| "loss": 0.4261, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.6644935156163283, | |
| "learning_rate": 3.795227455278029e-05, | |
| "loss": 0.4169, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1648, | |
| "grad_norm": 0.5160114938799306, | |
| "learning_rate": 3.787265280646825e-05, | |
| "loss": 0.408, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.1776, | |
| "grad_norm": 0.5772479443597772, | |
| "learning_rate": 3.7791599295517825e-05, | |
| "loss": 0.4042, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 0.4990992384712611, | |
| "learning_rate": 3.7709120513064196e-05, | |
| "loss": 0.432, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.2032, | |
| "grad_norm": 0.6008408343032209, | |
| "learning_rate": 3.762522306641998e-05, | |
| "loss": 0.4217, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.216, | |
| "grad_norm": 0.5030345534824172, | |
| "learning_rate": 3.7539913676545874e-05, | |
| "loss": 0.4237, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 0.5789101137412095, | |
| "learning_rate": 3.745319917751229e-05, | |
| "loss": 0.4326, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.2416, | |
| "grad_norm": 0.5536413045796079, | |
| "learning_rate": 3.736508651595188e-05, | |
| "loss": 0.4179, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.2544, | |
| "grad_norm": 0.5359427728023507, | |
| "learning_rate": 3.727558275050301e-05, | |
| "loss": 0.4181, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 0.5663150162094858, | |
| "learning_rate": 3.718469505124434e-05, | |
| "loss": 0.4184, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.5072843070203581, | |
| "learning_rate": 3.709243069912041e-05, | |
| "loss": 0.4281, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.2928, | |
| "grad_norm": 0.4503117371146354, | |
| "learning_rate": 3.699879708535838e-05, | |
| "loss": 0.4143, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 0.55160278777352, | |
| "learning_rate": 3.69038017108759e-05, | |
| "loss": 0.429, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.3184, | |
| "grad_norm": 0.45406641940754455, | |
| "learning_rate": 3.680745218568026e-05, | |
| "loss": 0.4071, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.3312, | |
| "grad_norm": 0.40694896153269045, | |
| "learning_rate": 3.6709756228258735e-05, | |
| "loss": 0.4203, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.44986497750864923, | |
| "learning_rate": 3.6610721664960236e-05, | |
| "loss": 0.4045, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.3568, | |
| "grad_norm": 0.37721415537529546, | |
| "learning_rate": 3.65103564293684e-05, | |
| "loss": 0.4211, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.3696, | |
| "grad_norm": 0.368659245329271, | |
| "learning_rate": 3.640866856166601e-05, | |
| "loss": 0.4276, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 0.3338767825735265, | |
| "learning_rate": 3.6305666207990886e-05, | |
| "loss": 0.411, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.3952, | |
| "grad_norm": 0.3633865115842757, | |
| "learning_rate": 3.6201357619783336e-05, | |
| "loss": 0.4089, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.408, | |
| "grad_norm": 0.4374442499655877, | |
| "learning_rate": 3.609575115312511e-05, | |
| "loss": 0.4226, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 0.4004164644698159, | |
| "learning_rate": 3.598885526807003e-05, | |
| "loss": 0.4121, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.4336, | |
| "grad_norm": 0.375110117398818, | |
| "learning_rate": 3.5880678527966224e-05, | |
| "loss": 0.4157, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.4464000000000001, | |
| "grad_norm": 0.3731005692384888, | |
| "learning_rate": 3.577122959877017e-05, | |
| "loss": 0.3933, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 0.3277842841378337, | |
| "learning_rate": 3.566051724835245e-05, | |
| "loss": 0.4202, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.472, | |
| "grad_norm": 0.33499183413983424, | |
| "learning_rate": 3.554855034579532e-05, | |
| "loss": 0.4172, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.4848, | |
| "grad_norm": 0.3748179696878851, | |
| "learning_rate": 3.5435337860682304e-05, | |
| "loss": 0.4029, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 0.4254097187751942, | |
| "learning_rate": 3.532088886237956e-05, | |
| "loss": 0.422, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.5104, | |
| "grad_norm": 0.5179617507141738, | |
| "learning_rate": 3.520521251930941e-05, | |
| "loss": 0.4103, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.5232, | |
| "grad_norm": 0.3875635520166566, | |
| "learning_rate": 3.5088318098215805e-05, | |
| "loss": 0.4155, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.47883477189205015, | |
| "learning_rate": 3.497021496342203e-05, | |
| "loss": 0.4114, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.5488, | |
| "grad_norm": 0.34661418607863886, | |
| "learning_rate": 3.485091257608047e-05, | |
| "loss": 0.4069, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.5615999999999999, | |
| "grad_norm": 0.47775160975563835, | |
| "learning_rate": 3.473042049341474e-05, | |
| "loss": 0.4032, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.41755319036242544, | |
| "learning_rate": 3.4608748367954064e-05, | |
| "loss": 0.4222, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.5872000000000002, | |
| "grad_norm": 0.36696142618261907, | |
| "learning_rate": 3.4485905946759965e-05, | |
| "loss": 0.4108, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.37958909151546943, | |
| "learning_rate": 3.4361903070645484e-05, | |
| "loss": 0.4085, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.3812085467183527, | |
| "learning_rate": 3.423674967338681e-05, | |
| "loss": 0.4141, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.6256, | |
| "grad_norm": 0.4593367556431078, | |
| "learning_rate": 3.411045578092754e-05, | |
| "loss": 0.4298, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.6383999999999999, | |
| "grad_norm": 0.4331143102681411, | |
| "learning_rate": 3.398303151057543e-05, | |
| "loss": 0.406, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.39194276694765806, | |
| "learning_rate": 3.385448707019199e-05, | |
| "loss": 0.4222, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 0.42420930748931346, | |
| "learning_rate": 3.372483275737468e-05, | |
| "loss": 0.4174, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.6768, | |
| "grad_norm": 0.47235884487449303, | |
| "learning_rate": 3.359407895863199e-05, | |
| "loss": 0.4034, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 0.39115166726965117, | |
| "learning_rate": 3.34622361485514e-05, | |
| "loss": 0.4154, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.7024, | |
| "grad_norm": 0.3451636569499237, | |
| "learning_rate": 3.332931488896029e-05, | |
| "loss": 0.4105, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.7151999999999998, | |
| "grad_norm": 0.3161116844063196, | |
| "learning_rate": 3.319532582807977e-05, | |
| "loss": 0.4183, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.3679004495473946, | |
| "learning_rate": 3.30602796996717e-05, | |
| "loss": 0.4191, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.7408000000000001, | |
| "grad_norm": 0.34726956073288306, | |
| "learning_rate": 3.2924187322178865e-05, | |
| "loss": 0.4257, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.7536, | |
| "grad_norm": 0.31060248978831084, | |
| "learning_rate": 3.278705959785821e-05, | |
| "loss": 0.4282, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 0.3146125528995013, | |
| "learning_rate": 3.2648907511907544e-05, | |
| "loss": 0.4032, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.7792, | |
| "grad_norm": 0.28346266626815353, | |
| "learning_rate": 3.250974213158555e-05, | |
| "loss": 0.4211, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.792, | |
| "grad_norm": 0.3373383894036537, | |
| "learning_rate": 3.23695746053251e-05, | |
| "loss": 0.4011, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 0.34395210564673195, | |
| "learning_rate": 3.222841616184025e-05, | |
| "loss": 0.4205, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.8176, | |
| "grad_norm": 0.3304361665571028, | |
| "learning_rate": 3.208627810922665e-05, | |
| "loss": 0.4151, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.8304, | |
| "grad_norm": 0.35253423378913656, | |
| "learning_rate": 3.194317183405573e-05, | |
| "loss": 0.403, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 0.32968583518217465, | |
| "learning_rate": 3.1799108800462466e-05, | |
| "loss": 0.4018, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.8559999999999999, | |
| "grad_norm": 0.3185495540407445, | |
| "learning_rate": 3.1654100549227024e-05, | |
| "loss": 0.4059, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.8688, | |
| "grad_norm": 0.3333565733934174, | |
| "learning_rate": 3.1508158696850275e-05, | |
| "loss": 0.4063, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 0.36834449770897526, | |
| "learning_rate": 3.136129493462312e-05, | |
| "loss": 0.4094, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.8944, | |
| "grad_norm": 0.33600546421388605, | |
| "learning_rate": 3.121352102768998e-05, | |
| "loss": 0.3885, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.9072, | |
| "grad_norm": 0.39221742726595227, | |
| "learning_rate": 3.106484881410628e-05, | |
| "loss": 0.4265, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.3420689584780464, | |
| "learning_rate": 3.091529020389009e-05, | |
| "loss": 0.3906, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9327999999999999, | |
| "grad_norm": 0.30516892060327233, | |
| "learning_rate": 3.076485717806808e-05, | |
| "loss": 0.4124, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.9456, | |
| "grad_norm": 0.2897850815596238, | |
| "learning_rate": 3.061356178771564e-05, | |
| "loss": 0.4082, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.3491682379379715, | |
| "learning_rate": 3.0461416152991555e-05, | |
| "loss": 0.4131, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.9712, | |
| "grad_norm": 0.3321799614030019, | |
| "learning_rate": 3.0308432462167045e-05, | |
| "loss": 0.414, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.984, | |
| "grad_norm": 0.34275541557645073, | |
| "learning_rate": 3.015462297064936e-05, | |
| "loss": 0.4246, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 0.3196110492871807, | |
| "learning_rate": 3.0000000000000004e-05, | |
| "loss": 0.4096, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.0096, | |
| "grad_norm": 0.934928679444364, | |
| "learning_rate": 2.98445759369477e-05, | |
| "loss": 0.7759, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.0224, | |
| "grad_norm": 1.0236079452171798, | |
| "learning_rate": 2.9688363232396056e-05, | |
| "loss": 0.3658, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 0.6907905080882383, | |
| "learning_rate": 2.9531374400426158e-05, | |
| "loss": 0.361, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 0.614016794612752, | |
| "learning_rate": 2.9373622017294075e-05, | |
| "loss": 0.3473, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.0608, | |
| "grad_norm": 0.7357077840047502, | |
| "learning_rate": 2.9215118720423375e-05, | |
| "loss": 0.3597, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 0.5871826378074856, | |
| "learning_rate": 2.9055877207392752e-05, | |
| "loss": 0.3611, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.0864, | |
| "grad_norm": 0.5973189355148211, | |
| "learning_rate": 2.8895910234918828e-05, | |
| "loss": 0.3577, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.0992, | |
| "grad_norm": 0.5305834501976533, | |
| "learning_rate": 2.873523061783426e-05, | |
| "loss": 0.3667, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.6200972108019946, | |
| "learning_rate": 2.8573851228061084e-05, | |
| "loss": 0.3584, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.1248, | |
| "grad_norm": 0.4456354026006885, | |
| "learning_rate": 2.8411784993579633e-05, | |
| "loss": 0.359, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.1376, | |
| "grad_norm": 0.606392060552971, | |
| "learning_rate": 2.8249044897392814e-05, | |
| "loss": 0.3636, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.4543237788961414, | |
| "learning_rate": 2.80856439764861e-05, | |
| "loss": 0.3651, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.1632, | |
| "grad_norm": 0.4203540347085503, | |
| "learning_rate": 2.792159532078314e-05, | |
| "loss": 0.3523, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 0.4879252666919352, | |
| "learning_rate": 2.77569120720971e-05, | |
| "loss": 0.3728, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 0.39553858494174166, | |
| "learning_rate": 2.7591607423077932e-05, | |
| "loss": 0.3473, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.2016, | |
| "grad_norm": 0.4708728706692069, | |
| "learning_rate": 2.7425694616155474e-05, | |
| "loss": 0.3651, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.2144, | |
| "grad_norm": 0.38129855381591565, | |
| "learning_rate": 2.7259186942478656e-05, | |
| "loss": 0.3642, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 0.4045113626468126, | |
| "learning_rate": 2.7092097740850712e-05, | |
| "loss": 0.3444, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.3725252563936457, | |
| "learning_rate": 2.692444039666066e-05, | |
| "loss": 0.349, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.2528, | |
| "grad_norm": 0.43541464400993846, | |
| "learning_rate": 2.6756228340810946e-05, | |
| "loss": 0.367, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 0.36746613679395396, | |
| "learning_rate": 2.6587475048641596e-05, | |
| "loss": 0.3524, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.2784, | |
| "grad_norm": 0.41600433055006897, | |
| "learning_rate": 2.6418194038850634e-05, | |
| "loss": 0.3637, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.2912, | |
| "grad_norm": 0.42485141621329564, | |
| "learning_rate": 2.624839887241115e-05, | |
| "loss": 0.33, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.3355524372237687, | |
| "learning_rate": 2.607810315148494e-05, | |
| "loss": 0.3509, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.3168, | |
| "grad_norm": 0.39312247650879356, | |
| "learning_rate": 2.5907320518332827e-05, | |
| "loss": 0.3585, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.3296, | |
| "grad_norm": 0.34147429361884374, | |
| "learning_rate": 2.5736064654221808e-05, | |
| "loss": 0.3476, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.3913611625996246, | |
| "learning_rate": 2.5564349278329056e-05, | |
| "loss": 0.3623, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.3552, | |
| "grad_norm": 0.35016818850603587, | |
| "learning_rate": 2.539218814664288e-05, | |
| "loss": 0.367, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 0.3390125142485565, | |
| "learning_rate": 2.521959505086075e-05, | |
| "loss": 0.3482, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 0.3989973226987953, | |
| "learning_rate": 2.5046583817284437e-05, | |
| "loss": 0.3475, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.3936, | |
| "grad_norm": 0.27839268304068504, | |
| "learning_rate": 2.487316830571244e-05, | |
| "loss": 0.3487, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.4064, | |
| "grad_norm": 0.3741956707296818, | |
| "learning_rate": 2.4699362408329646e-05, | |
| "loss": 0.3523, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 0.24703942380951768, | |
| "learning_rate": 2.4525180048594452e-05, | |
| "loss": 0.3474, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.432, | |
| "grad_norm": 0.36272455587292735, | |
| "learning_rate": 2.435063518012335e-05, | |
| "loss": 0.3696, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.4448, | |
| "grad_norm": 0.257371299726762, | |
| "learning_rate": 2.4175741785573177e-05, | |
| "loss": 0.3501, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 0.3730160171190589, | |
| "learning_rate": 2.4000513875520892e-05, | |
| "loss": 0.3604, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.4704, | |
| "grad_norm": 0.23591573396310475, | |
| "learning_rate": 2.3824965487341247e-05, | |
| "loss": 0.3417, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.4832, | |
| "grad_norm": 0.33186202870800685, | |
| "learning_rate": 2.3649110684082258e-05, | |
| "loss": 0.3565, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.2952048814996425, | |
| "learning_rate": 2.3472963553338614e-05, | |
| "loss": 0.3577, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.5088, | |
| "grad_norm": 0.30462445713853137, | |
| "learning_rate": 2.3296538206123134e-05, | |
| "loss": 0.3535, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.5216, | |
| "grad_norm": 0.31780543779325826, | |
| "learning_rate": 2.311984877573636e-05, | |
| "loss": 0.3742, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.26698422088860363, | |
| "learning_rate": 2.2942909416634326e-05, | |
| "loss": 0.3446, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.5472, | |
| "grad_norm": 0.2738562371187304, | |
| "learning_rate": 2.2765734303294666e-05, | |
| "loss": 0.3551, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.27866823737832547, | |
| "learning_rate": 2.2588337629081107e-05, | |
| "loss": 0.3544, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.26045914326744124, | |
| "learning_rate": 2.2410733605106462e-05, | |
| "loss": 0.3477, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.5856, | |
| "grad_norm": 0.31014120584401034, | |
| "learning_rate": 2.2232936459094158e-05, | |
| "loss": 0.3697, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.5984, | |
| "grad_norm": 0.27841335766852626, | |
| "learning_rate": 2.205496043423849e-05, | |
| "loss": 0.3456, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.2862907595537163, | |
| "learning_rate": 2.1876819788063586e-05, | |
| "loss": 0.3476, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.624, | |
| "grad_norm": 0.29080076931320287, | |
| "learning_rate": 2.16985287912813e-05, | |
| "loss": 0.3574, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.6368, | |
| "grad_norm": 0.2594251736323389, | |
| "learning_rate": 2.1520101726647922e-05, | |
| "loss": 0.3539, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.30418822140778956, | |
| "learning_rate": 2.1341552887820048e-05, | |
| "loss": 0.3455, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.6624, | |
| "grad_norm": 0.2847830777699319, | |
| "learning_rate": 2.1162896578209517e-05, | |
| "loss": 0.364, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.6752000000000002, | |
| "grad_norm": 0.2985995943000321, | |
| "learning_rate": 2.0984147109837564e-05, | |
| "loss": 0.3475, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.26987472155424863, | |
| "learning_rate": 2.0805318802188307e-05, | |
| "loss": 0.3579, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.7008, | |
| "grad_norm": 0.3035529447970014, | |
| "learning_rate": 2.0626425981061608e-05, | |
| "loss": 0.3498, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.7136, | |
| "grad_norm": 0.2868524536433486, | |
| "learning_rate": 2.0447482977425465e-05, | |
| "loss": 0.3541, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.2938698686450696, | |
| "learning_rate": 2.0268504126267952e-05, | |
| "loss": 0.3501, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.7392, | |
| "grad_norm": 0.31108071495771955, | |
| "learning_rate": 2.008950376544887e-05, | |
| "loss": 0.3443, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 0.29191778627949005, | |
| "learning_rate": 1.9910496234551132e-05, | |
| "loss": 0.3554, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.3130472348817152, | |
| "learning_rate": 1.9731495873732055e-05, | |
| "loss": 0.3511, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.7776, | |
| "grad_norm": 0.23938283248245062, | |
| "learning_rate": 1.9552517022574542e-05, | |
| "loss": 0.3457, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.7904, | |
| "grad_norm": 0.33638139528089983, | |
| "learning_rate": 1.93735740189384e-05, | |
| "loss": 0.3574, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.2725263439035572, | |
| "learning_rate": 1.9194681197811703e-05, | |
| "loss": 0.3509, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 0.34846190643169006, | |
| "learning_rate": 1.901585289016244e-05, | |
| "loss": 0.3602, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.8288, | |
| "grad_norm": 0.282445618365205, | |
| "learning_rate": 1.8837103421790486e-05, | |
| "loss": 0.3641, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.298546235479727, | |
| "learning_rate": 1.8658447112179952e-05, | |
| "loss": 0.3477, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.8544, | |
| "grad_norm": 0.2830491571348469, | |
| "learning_rate": 1.8479898273352084e-05, | |
| "loss": 0.3528, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.8672, | |
| "grad_norm": 0.30120732024157043, | |
| "learning_rate": 1.83014712087187e-05, | |
| "loss": 0.3663, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.278625164747294, | |
| "learning_rate": 1.8123180211936417e-05, | |
| "loss": 0.34, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.8928000000000003, | |
| "grad_norm": 0.31432746327290567, | |
| "learning_rate": 1.794503956576152e-05, | |
| "loss": 0.3649, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.9055999999999997, | |
| "grad_norm": 0.27925289524129054, | |
| "learning_rate": 1.776706354090585e-05, | |
| "loss": 0.3456, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.2930736777187307, | |
| "learning_rate": 1.758926639489354e-05, | |
| "loss": 0.3374, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.9312, | |
| "grad_norm": 0.28631397214328114, | |
| "learning_rate": 1.7411662370918893e-05, | |
| "loss": 0.3508, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 0.28855685187775515, | |
| "learning_rate": 1.7234265696705344e-05, | |
| "loss": 0.3641, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 0.2574983753185641, | |
| "learning_rate": 1.7057090583365678e-05, | |
| "loss": 0.3508, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.9696, | |
| "grad_norm": 0.3135454054501316, | |
| "learning_rate": 1.6880151224263646e-05, | |
| "loss": 0.354, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.9824, | |
| "grad_norm": 0.23669837052280993, | |
| "learning_rate": 1.6703461793876876e-05, | |
| "loss": 0.3619, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.2654003694215428, | |
| "learning_rate": 1.6527036446661396e-05, | |
| "loss": 0.3463, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 3.008, | |
| "grad_norm": 0.8110914524977084, | |
| "learning_rate": 1.635088931591775e-05, | |
| "loss": 0.6571, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 3.0208, | |
| "grad_norm": 0.4927579107099172, | |
| "learning_rate": 1.6175034512658753e-05, | |
| "loss": 0.3053, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 3.0336, | |
| "grad_norm": 0.5343601607955114, | |
| "learning_rate": 1.5999486124479115e-05, | |
| "loss": 0.3081, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 3.0464, | |
| "grad_norm": 0.47616019797663545, | |
| "learning_rate": 1.5824258214426833e-05, | |
| "loss": 0.3074, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 3.0592, | |
| "grad_norm": 0.5092719956633589, | |
| "learning_rate": 1.5649364819876655e-05, | |
| "loss": 0.3087, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 3.072, | |
| "grad_norm": 0.3976744545046491, | |
| "learning_rate": 1.547481995140556e-05, | |
| "loss": 0.2961, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.0848, | |
| "grad_norm": 0.43135072736718333, | |
| "learning_rate": 1.5300637591670357e-05, | |
| "loss": 0.3018, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 3.0976, | |
| "grad_norm": 0.41934440770187326, | |
| "learning_rate": 1.5126831694287564e-05, | |
| "loss": 0.3156, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 3.1104, | |
| "grad_norm": 0.3432402459454663, | |
| "learning_rate": 1.4953416182715566e-05, | |
| "loss": 0.2941, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 3.1232, | |
| "grad_norm": 0.42487103004264876, | |
| "learning_rate": 1.478040494913926e-05, | |
| "loss": 0.3091, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.136, | |
| "grad_norm": 0.28096000394784443, | |
| "learning_rate": 1.460781185335713e-05, | |
| "loss": 0.3008, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 3.1488, | |
| "grad_norm": 0.35282592668823415, | |
| "learning_rate": 1.443565072167095e-05, | |
| "loss": 0.2965, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 3.1616, | |
| "grad_norm": 0.3099781879452714, | |
| "learning_rate": 1.4263935345778202e-05, | |
| "loss": 0.3041, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 3.1744, | |
| "grad_norm": 0.2879517791583705, | |
| "learning_rate": 1.409267948166718e-05, | |
| "loss": 0.2932, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 3.1872, | |
| "grad_norm": 0.29599606790636535, | |
| "learning_rate": 1.3921896848515064e-05, | |
| "loss": 0.3103, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.32115873426543345, | |
| "learning_rate": 1.3751601127588849e-05, | |
| "loss": 0.3099, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.2128, | |
| "grad_norm": 0.3240905938409589, | |
| "learning_rate": 1.3581805961149371e-05, | |
| "loss": 0.3018, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 3.2256, | |
| "grad_norm": 0.27810782641762516, | |
| "learning_rate": 1.341252495135841e-05, | |
| "loss": 0.291, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 3.2384, | |
| "grad_norm": 0.27103514655866207, | |
| "learning_rate": 1.324377165918906e-05, | |
| "loss": 0.3068, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 3.2512, | |
| "grad_norm": 0.31232808385242766, | |
| "learning_rate": 1.3075559603339354e-05, | |
| "loss": 0.2985, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 3.2640000000000002, | |
| "grad_norm": 0.27191837257705587, | |
| "learning_rate": 1.2907902259149287e-05, | |
| "loss": 0.3134, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 3.2768, | |
| "grad_norm": 0.30363186300170436, | |
| "learning_rate": 1.274081305752135e-05, | |
| "loss": 0.2996, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 3.2896, | |
| "grad_norm": 0.2963966610539313, | |
| "learning_rate": 1.2574305383844528e-05, | |
| "loss": 0.2794, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 3.3024, | |
| "grad_norm": 0.24622465411557967, | |
| "learning_rate": 1.2408392576922075e-05, | |
| "loss": 0.2927, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 3.3152, | |
| "grad_norm": 0.28706174627781483, | |
| "learning_rate": 1.2243087927902905e-05, | |
| "loss": 0.3127, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 3.328, | |
| "grad_norm": 0.2688560097423624, | |
| "learning_rate": 1.2078404679216864e-05, | |
| "loss": 0.2949, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 3.3407999999999998, | |
| "grad_norm": 0.24420629943447866, | |
| "learning_rate": 1.1914356023513904e-05, | |
| "loss": 0.2912, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 3.3536, | |
| "grad_norm": 0.25231007929180993, | |
| "learning_rate": 1.1750955102607193e-05, | |
| "loss": 0.3051, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 3.3664, | |
| "grad_norm": 0.28772972204632646, | |
| "learning_rate": 1.1588215006420374e-05, | |
| "loss": 0.3059, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 3.3792, | |
| "grad_norm": 0.2430502606869639, | |
| "learning_rate": 1.1426148771938915e-05, | |
| "loss": 0.2894, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 3.392, | |
| "grad_norm": 0.25181830565633234, | |
| "learning_rate": 1.1264769382165748e-05, | |
| "loss": 0.3018, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 3.4048, | |
| "grad_norm": 0.26536806247510114, | |
| "learning_rate": 1.110408976508118e-05, | |
| "loss": 0.2992, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 3.4176, | |
| "grad_norm": 0.22464906686193997, | |
| "learning_rate": 1.094412279260726e-05, | |
| "loss": 0.2897, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 3.4304, | |
| "grad_norm": 0.240850912306661, | |
| "learning_rate": 1.0784881279576635e-05, | |
| "loss": 0.298, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 3.4432, | |
| "grad_norm": 0.23488709472818292, | |
| "learning_rate": 1.0626377982705929e-05, | |
| "loss": 0.2986, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 3.456, | |
| "grad_norm": 0.22104731994056562, | |
| "learning_rate": 1.0468625599573842e-05, | |
| "loss": 0.2944, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.4688, | |
| "grad_norm": 0.22765225207348563, | |
| "learning_rate": 1.0311636767603952e-05, | |
| "loss": 0.2951, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 3.4816, | |
| "grad_norm": 0.24080246044291082, | |
| "learning_rate": 1.0155424063052306e-05, | |
| "loss": 0.3088, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 3.4944, | |
| "grad_norm": 0.21816199855649074, | |
| "learning_rate": 1.0000000000000006e-05, | |
| "loss": 0.303, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 3.5072, | |
| "grad_norm": 0.22701897913136335, | |
| "learning_rate": 9.84537702935065e-06, | |
| "loss": 0.2961, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 0.22669490142947354, | |
| "learning_rate": 9.691567537832964e-06, | |
| "loss": 0.2983, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.5328, | |
| "grad_norm": 0.22238583592444078, | |
| "learning_rate": 9.538583847008452e-06, | |
| "loss": 0.3136, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 3.5456, | |
| "grad_norm": 0.20680399388103746, | |
| "learning_rate": 9.386438212284372e-06, | |
| "loss": 0.2874, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 3.5584, | |
| "grad_norm": 0.21841901308514358, | |
| "learning_rate": 9.235142821931928e-06, | |
| "loss": 0.2995, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 3.5712, | |
| "grad_norm": 0.20992286180370623, | |
| "learning_rate": 9.084709796109907e-06, | |
| "loss": 0.2929, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 3.584, | |
| "grad_norm": 0.20535086517609702, | |
| "learning_rate": 8.93515118589373e-06, | |
| "loss": 0.2882, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.5968, | |
| "grad_norm": 0.2040545303229864, | |
| "learning_rate": 8.786478972310023e-06, | |
| "loss": 0.2999, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 3.6096, | |
| "grad_norm": 0.1972696173771072, | |
| "learning_rate": 8.638705065376887e-06, | |
| "loss": 0.2792, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 3.6224, | |
| "grad_norm": 0.20595086570259513, | |
| "learning_rate": 8.491841303149728e-06, | |
| "loss": 0.3, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 3.6352, | |
| "grad_norm": 0.21559665046648244, | |
| "learning_rate": 8.345899450772975e-06, | |
| "loss": 0.3054, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 3.648, | |
| "grad_norm": 0.22914067240229374, | |
| "learning_rate": 8.200891199537549e-06, | |
| "loss": 0.2977, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 3.6608, | |
| "grad_norm": 0.19325134978427533, | |
| "learning_rate": 8.056828165944282e-06, | |
| "loss": 0.2859, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 3.6736, | |
| "grad_norm": 0.2235828066384403, | |
| "learning_rate": 7.913721890773354e-06, | |
| "loss": 0.2943, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 3.6864, | |
| "grad_norm": 0.21218831537621788, | |
| "learning_rate": 7.771583838159756e-06, | |
| "loss": 0.3061, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.6992000000000003, | |
| "grad_norm": 0.2019458235697608, | |
| "learning_rate": 7.630425394674903e-06, | |
| "loss": 0.2887, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 3.7119999999999997, | |
| "grad_norm": 0.21772427651342774, | |
| "learning_rate": 7.49025786841445e-06, | |
| "loss": 0.3052, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.7248, | |
| "grad_norm": 0.21168547061512974, | |
| "learning_rate": 7.3510924880924575e-06, | |
| "loss": 0.2992, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 3.7376, | |
| "grad_norm": 0.20419575992313294, | |
| "learning_rate": 7.212940402141808e-06, | |
| "loss": 0.2981, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 3.7504, | |
| "grad_norm": 0.2003967083437701, | |
| "learning_rate": 7.075812677821145e-06, | |
| "loss": 0.2954, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 3.7632, | |
| "grad_norm": 0.18999075019544526, | |
| "learning_rate": 6.939720300328303e-06, | |
| "loss": 0.2833, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 3.776, | |
| "grad_norm": 0.19849448218291585, | |
| "learning_rate": 6.8046741719202385e-06, | |
| "loss": 0.3033, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 3.7888, | |
| "grad_norm": 0.20865535018018086, | |
| "learning_rate": 6.67068511103971e-06, | |
| "loss": 0.3067, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 3.8016, | |
| "grad_norm": 0.187437835814437, | |
| "learning_rate": 6.537763851448593e-06, | |
| "loss": 0.2866, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 3.8144, | |
| "grad_norm": 0.20175582821334903, | |
| "learning_rate": 6.4059210413680175e-06, | |
| "loss": 0.3021, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 3.8272, | |
| "grad_norm": 0.19881896333131485, | |
| "learning_rate": 6.275167242625331e-06, | |
| "loss": 0.2911, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.19788736357342163, | |
| "learning_rate": 6.145512929808013e-06, | |
| "loss": 0.3074, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.8528000000000002, | |
| "grad_norm": 0.18594814195497833, | |
| "learning_rate": 6.016968489424572e-06, | |
| "loss": 0.2889, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 3.8656, | |
| "grad_norm": 0.19626956102544968, | |
| "learning_rate": 5.889544219072465e-06, | |
| "loss": 0.286, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 3.8784, | |
| "grad_norm": 0.19335467289370822, | |
| "learning_rate": 5.7632503266131925e-06, | |
| "loss": 0.301, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 3.8912, | |
| "grad_norm": 0.20329232849955123, | |
| "learning_rate": 5.638096929354522e-06, | |
| "loss": 0.3018, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 3.904, | |
| "grad_norm": 0.19623380102316826, | |
| "learning_rate": 5.514094053240035e-06, | |
| "loss": 0.2932, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 3.9168, | |
| "grad_norm": 0.20040771058487258, | |
| "learning_rate": 5.39125163204594e-06, | |
| "loss": 0.2982, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 3.9295999999999998, | |
| "grad_norm": 0.19742272558147897, | |
| "learning_rate": 5.269579506585259e-06, | |
| "loss": 0.3095, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 3.9424, | |
| "grad_norm": 0.19803710395959664, | |
| "learning_rate": 5.149087423919541e-06, | |
| "loss": 0.3022, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 3.9552, | |
| "grad_norm": 0.19236143447031787, | |
| "learning_rate": 5.029785036577976e-06, | |
| "loss": 0.3059, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 3.968, | |
| "grad_norm": 0.19117604630998866, | |
| "learning_rate": 4.911681901784198e-06, | |
| "loss": 0.3072, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.9808, | |
| "grad_norm": 0.18988071468477521, | |
| "learning_rate": 4.794787480690597e-06, | |
| "loss": 0.2913, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 3.9936, | |
| "grad_norm": 0.2009831779426468, | |
| "learning_rate": 4.679111137620442e-06, | |
| "loss": 0.2993, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 4.0064, | |
| "grad_norm": 0.5632280925124512, | |
| "learning_rate": 4.5646621393177e-06, | |
| "loss": 0.5714, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 4.0192, | |
| "grad_norm": 0.30295580961224955, | |
| "learning_rate": 4.451449654204685e-06, | |
| "loss": 0.2803, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 4.032, | |
| "grad_norm": 0.23959599119308816, | |
| "learning_rate": 4.339482751647557e-06, | |
| "loss": 0.2555, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 4.0448, | |
| "grad_norm": 0.2204796751672579, | |
| "learning_rate": 4.228770401229824e-06, | |
| "loss": 0.2697, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 4.0576, | |
| "grad_norm": 0.26164117620868194, | |
| "learning_rate": 4.119321472033779e-06, | |
| "loss": 0.2689, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 4.0704, | |
| "grad_norm": 0.31254641059599036, | |
| "learning_rate": 4.011144731929981e-06, | |
| "loss": 0.2631, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 4.0832, | |
| "grad_norm": 0.2959531711212858, | |
| "learning_rate": 3.904248846874894e-06, | |
| "loss": 0.2723, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 4.096, | |
| "grad_norm": 0.23920286570294827, | |
| "learning_rate": 3.7986423802166705e-06, | |
| "loss": 0.2584, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 4.1088, | |
| "grad_norm": 0.22015501609003424, | |
| "learning_rate": 3.694333792009115e-06, | |
| "loss": 0.2692, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 4.1216, | |
| "grad_norm": 0.2387964460667242, | |
| "learning_rate": 3.5913314383339937e-06, | |
| "loss": 0.2599, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 4.1344, | |
| "grad_norm": 0.2511860059043027, | |
| "learning_rate": 3.4896435706316e-06, | |
| "loss": 0.2831, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 4.1472, | |
| "grad_norm": 0.23659326838197345, | |
| "learning_rate": 3.3892783350397675e-06, | |
| "loss": 0.263, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 0.2209162945694483, | |
| "learning_rate": 3.290243771741275e-06, | |
| "loss": 0.2712, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 4.1728, | |
| "grad_norm": 0.20199866391033341, | |
| "learning_rate": 3.1925478143197418e-06, | |
| "loss": 0.2598, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 4.1856, | |
| "grad_norm": 0.2014882396839415, | |
| "learning_rate": 3.0961982891241083e-06, | |
| "loss": 0.2585, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 4.1984, | |
| "grad_norm": 0.23344394144325656, | |
| "learning_rate": 3.001202914641628e-06, | |
| "loss": 0.2748, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 4.2112, | |
| "grad_norm": 0.23612419288080505, | |
| "learning_rate": 2.907569300879596e-06, | |
| "loss": 0.2499, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 4.224, | |
| "grad_norm": 0.20885975695664077, | |
| "learning_rate": 2.815304948755664e-06, | |
| "loss": 0.2646, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 4.2368, | |
| "grad_norm": 0.21529597575216605, | |
| "learning_rate": 2.7244172494969978e-06, | |
| "loss": 0.2673, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 4.2496, | |
| "grad_norm": 0.21164720639106868, | |
| "learning_rate": 2.6349134840481294e-06, | |
| "loss": 0.268, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 4.2624, | |
| "grad_norm": 0.19501771560133913, | |
| "learning_rate": 2.546800822487714e-06, | |
| "loss": 0.2647, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 4.2752, | |
| "grad_norm": 0.19182972592946143, | |
| "learning_rate": 2.4600863234541338e-06, | |
| "loss": 0.259, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 4.288, | |
| "grad_norm": 0.20555554000403917, | |
| "learning_rate": 2.374776933580025e-06, | |
| "loss": 0.2518, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 4.3008, | |
| "grad_norm": 0.20584681289289913, | |
| "learning_rate": 2.2908794869358044e-06, | |
| "loss": 0.273, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 4.3136, | |
| "grad_norm": 0.18865108411490195, | |
| "learning_rate": 2.2084007044821764e-06, | |
| "loss": 0.2553, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 4.3264, | |
| "grad_norm": 0.1868823005242205, | |
| "learning_rate": 2.127347193531757e-06, | |
| "loss": 0.2631, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 4.3392, | |
| "grad_norm": 0.18248303291457005, | |
| "learning_rate": 2.0477254472197237e-06, | |
| "loss": 0.2637, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 4.352, | |
| "grad_norm": 0.1881614520651855, | |
| "learning_rate": 1.96954184398368e-06, | |
| "loss": 0.2555, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 4.3648, | |
| "grad_norm": 0.1908904332943584, | |
| "learning_rate": 1.8928026470526917e-06, | |
| "loss": 0.2626, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 4.3776, | |
| "grad_norm": 0.20460087483221504, | |
| "learning_rate": 1.817514003945524e-06, | |
| "loss": 0.263, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 4.3904, | |
| "grad_norm": 0.1978994172877242, | |
| "learning_rate": 1.743681945978184e-06, | |
| "loss": 0.2592, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 4.4032, | |
| "grad_norm": 0.18553288366651352, | |
| "learning_rate": 1.6713123877807413e-06, | |
| "loss": 0.2599, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 4.416, | |
| "grad_norm": 0.18501368725594813, | |
| "learning_rate": 1.6004111268235156e-06, | |
| "loss": 0.267, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 4.4288, | |
| "grad_norm": 0.1802623353473316, | |
| "learning_rate": 1.5309838429526714e-06, | |
| "loss": 0.2489, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 4.4416, | |
| "grad_norm": 0.1801867415427244, | |
| "learning_rate": 1.4630360979351644e-06, | |
| "loss": 0.249, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 4.4544, | |
| "grad_norm": 0.18621397159511183, | |
| "learning_rate": 1.396573335013236e-06, | |
| "loss": 0.2578, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 4.4672, | |
| "grad_norm": 0.1853442065705862, | |
| "learning_rate": 1.3316008784683265e-06, | |
| "loss": 0.2715, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 0.17739351043616997, | |
| "learning_rate": 1.2681239331945695e-06, | |
| "loss": 0.258, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.4928, | |
| "grad_norm": 0.18711453533839706, | |
| "learning_rate": 1.2061475842818337e-06, | |
| "loss": 0.2662, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 4.5056, | |
| "grad_norm": 0.1752807343835462, | |
| "learning_rate": 1.1456767966083393e-06, | |
| "loss": 0.2631, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 4.5184, | |
| "grad_norm": 0.16875907478026356, | |
| "learning_rate": 1.086716414442952e-06, | |
| "loss": 0.2676, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 4.5312, | |
| "grad_norm": 0.16860487528620643, | |
| "learning_rate": 1.0292711610570904e-06, | |
| "loss": 0.2581, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 4.5440000000000005, | |
| "grad_norm": 0.17559979853919078, | |
| "learning_rate": 9.733456383463658e-07, | |
| "loss": 0.2583, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 4.5568, | |
| "grad_norm": 0.17827232369186016, | |
| "learning_rate": 9.189443264619102e-07, | |
| "loss": 0.2579, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 4.5696, | |
| "grad_norm": 0.17242076586765648, | |
| "learning_rate": 8.660715834514977e-07, | |
| "loss": 0.2513, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 4.5824, | |
| "grad_norm": 0.17426052915343207, | |
| "learning_rate": 8.147316449103959e-07, | |
| "loss": 0.2636, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 4.5952, | |
| "grad_norm": 0.17983994762413885, | |
| "learning_rate": 7.649286236420806e-07, | |
| "loss": 0.2748, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 4.608, | |
| "grad_norm": 0.17392285097462207, | |
| "learning_rate": 7.166665093287539e-07, | |
| "loss": 0.2567, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.6208, | |
| "grad_norm": 0.1779507092373814, | |
| "learning_rate": 6.69949168211721e-07, | |
| "loss": 0.2703, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 4.6336, | |
| "grad_norm": 0.1709659693009413, | |
| "learning_rate": 6.247803427816945e-07, | |
| "loss": 0.2712, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 4.6464, | |
| "grad_norm": 0.16795556154011698, | |
| "learning_rate": 5.811636514789598e-07, | |
| "loss": 0.2566, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 4.6592, | |
| "grad_norm": 0.16765383887805013, | |
| "learning_rate": 5.391025884035239e-07, | |
| "loss": 0.2698, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 4.672, | |
| "grad_norm": 0.1692473094904187, | |
| "learning_rate": 4.986005230351954e-07, | |
| "loss": 0.2563, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 4.6848, | |
| "grad_norm": 0.1625051854077313, | |
| "learning_rate": 4.5966069996365993e-07, | |
| "loss": 0.2503, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.6975999999999996, | |
| "grad_norm": 0.16992361814444382, | |
| "learning_rate": 4.22286238628562e-07, | |
| "loss": 0.2723, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 4.7104, | |
| "grad_norm": 0.1663062650696023, | |
| "learning_rate": 3.8648013306960664e-07, | |
| "loss": 0.2604, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 4.7232, | |
| "grad_norm": 0.17303492851675578, | |
| "learning_rate": 3.522452516867048e-07, | |
| "loss": 0.2619, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 4.736, | |
| "grad_norm": 0.16718963063684367, | |
| "learning_rate": 3.1958433701019697e-07, | |
| "loss": 0.2611, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.7488, | |
| "grad_norm": 0.17371519183251058, | |
| "learning_rate": 2.8850000548115155e-07, | |
| "loss": 0.267, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 4.7616, | |
| "grad_norm": 0.1712546787334409, | |
| "learning_rate": 2.5899474724174313e-07, | |
| "loss": 0.2694, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 4.7744, | |
| "grad_norm": 0.16756141777351802, | |
| "learning_rate": 2.3107092593579905e-07, | |
| "loss": 0.2563, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 4.7872, | |
| "grad_norm": 0.17046484427263547, | |
| "learning_rate": 2.0473077851942858e-07, | |
| "loss": 0.2738, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.16790481452672923, | |
| "learning_rate": 1.799764150818306e-07, | |
| "loss": 0.2606, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 4.8128, | |
| "grad_norm": 0.17116193260418378, | |
| "learning_rate": 1.5680981867625566e-07, | |
| "loss": 0.2723, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 4.8256, | |
| "grad_norm": 0.16411944775543716, | |
| "learning_rate": 1.3523284516113955e-07, | |
| "loss": 0.2526, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 4.8384, | |
| "grad_norm": 0.1673768080163684, | |
| "learning_rate": 1.1524722305144231e-07, | |
| "loss": 0.256, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 4.8512, | |
| "grad_norm": 0.17018197801975496, | |
| "learning_rate": 9.685455338016347e-08, | |
| "loss": 0.271, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 4.864, | |
| "grad_norm": 0.1698134662220362, | |
| "learning_rate": 8.005630957010014e-08, | |
| "loss": 0.2588, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.8768, | |
| "grad_norm": 0.17333190629165726, | |
| "learning_rate": 6.485383731580142e-08, | |
| "loss": 0.2664, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 4.8896, | |
| "grad_norm": 0.16940841305166987, | |
| "learning_rate": 5.1248354475768034e-08, | |
| "loss": 0.2607, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 4.9024, | |
| "grad_norm": 0.16665475939777138, | |
| "learning_rate": 3.924095097489922e-08, | |
| "loss": 0.2481, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 4.9152000000000005, | |
| "grad_norm": 0.17746993670022432, | |
| "learning_rate": 2.8832588717164766e-08, | |
| "loss": 0.2733, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.928, | |
| "grad_norm": 0.1702060704982818, | |
| "learning_rate": 2.0024101508555604e-08, | |
| "loss": 0.2615, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 4.9408, | |
| "grad_norm": 0.16469226784772867, | |
| "learning_rate": 1.281619499029274e-08, | |
| "loss": 0.2514, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 4.9536, | |
| "grad_norm": 0.1666195244598944, | |
| "learning_rate": 7.209446582292501e-09, | |
| "loss": 0.2591, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 4.9664, | |
| "grad_norm": 0.16614035020896728, | |
| "learning_rate": 3.2043054369057523e-09, | |
| "loss": 0.2585, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 4.9792, | |
| "grad_norm": 0.16541385437942008, | |
| "learning_rate": 8.010924029533406e-10, | |
| "loss": 0.2586, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "grad_norm": 0.17045021794260856, | |
| "learning_rate": 0.0, | |
| "loss": 0.2621, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "step": 390, | |
| "total_flos": 2.0752858298565263e+18, | |
| "train_loss": 0.3777036047516725, | |
| "train_runtime": 60864.0199, | |
| "train_samples_per_second": 0.822, | |
| "train_steps_per_second": 0.006 | |
| } | |
| ], | |
| "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": 2.0752858298565263e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |