| { | |
| "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": 6.4772331570390875, | |
| "learning_rate": 1.0256410256410257e-06, | |
| "loss": 1.0416, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 6.4609236851006, | |
| "learning_rate": 2.0512820512820513e-06, | |
| "loss": 1.0573, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 6.372060996532902, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 1.0529, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 5.701931751869838, | |
| "learning_rate": 4.102564102564103e-06, | |
| "loss": 1.0094, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 4.1986234763258246, | |
| "learning_rate": 5.128205128205128e-06, | |
| "loss": 0.9876, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 2.7072608868905674, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.958, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 2.51886534968088, | |
| "learning_rate": 7.17948717948718e-06, | |
| "loss": 0.9648, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 4.742818068062914, | |
| "learning_rate": 8.205128205128205e-06, | |
| "loss": 1.0247, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 4.444536744411092, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.9769, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 4.275427982968736, | |
| "learning_rate": 1.0256410256410256e-05, | |
| "loss": 0.9297, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 3.9760160145038355, | |
| "learning_rate": 1.1282051282051283e-05, | |
| "loss": 0.9136, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 2.8019221819330355, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.89, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 1.753942180394664, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.8703, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 1.6142618309782335, | |
| "learning_rate": 1.435897435897436e-05, | |
| "loss": 0.8429, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 1.3980658367087893, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.8324, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 1.1235618442879347, | |
| "learning_rate": 1.641025641025641e-05, | |
| "loss": 0.8216, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 1.1082025953461083, | |
| "learning_rate": 1.7435897435897438e-05, | |
| "loss": 0.7861, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 1.09823619020249, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.7886, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 1.1592942939257511, | |
| "learning_rate": 1.9487179487179488e-05, | |
| "loss": 0.7775, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 1.1537321253823973, | |
| "learning_rate": 2.0512820512820512e-05, | |
| "loss": 0.7678, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 1.025780911787061, | |
| "learning_rate": 2.153846153846154e-05, | |
| "loss": 0.8155, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 0.8844299276395537, | |
| "learning_rate": 2.2564102564102566e-05, | |
| "loss": 0.7646, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 1.0917684001067873, | |
| "learning_rate": 2.3589743589743593e-05, | |
| "loss": 0.7478, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 0.8784543182070157, | |
| "learning_rate": 2.461538461538462e-05, | |
| "loss": 0.7521, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.8339568504968727, | |
| "learning_rate": 2.5641025641025646e-05, | |
| "loss": 0.7351, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 0.7629063462860947, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 0.7358, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 1.724297577703023, | |
| "learning_rate": 2.7692307692307694e-05, | |
| "loss": 0.7795, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 0.8347055237941182, | |
| "learning_rate": 2.871794871794872e-05, | |
| "loss": 0.7439, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 0.9911885442550583, | |
| "learning_rate": 2.9743589743589747e-05, | |
| "loss": 0.7185, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 1.10395635295512, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.7387, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 1.0269546333971975, | |
| "learning_rate": 3.1794871794871795e-05, | |
| "loss": 0.7326, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 0.993562417635811, | |
| "learning_rate": 3.282051282051282e-05, | |
| "loss": 0.7421, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 1.639712778960047, | |
| "learning_rate": 3.384615384615385e-05, | |
| "loss": 0.7129, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 0.8674659776449699, | |
| "learning_rate": 3.4871794871794875e-05, | |
| "loss": 0.7309, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 1.5845124468239566, | |
| "learning_rate": 3.58974358974359e-05, | |
| "loss": 0.7093, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 1.1027837163333147, | |
| "learning_rate": 3.692307692307693e-05, | |
| "loss": 0.7215, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 1.2908004992313573, | |
| "learning_rate": 3.794871794871795e-05, | |
| "loss": 0.713, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 0.8727435676535944, | |
| "learning_rate": 3.8974358974358976e-05, | |
| "loss": 0.7094, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 0.9908035623322232, | |
| "learning_rate": 4e-05, | |
| "loss": 0.7114, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 1.4556065089073809, | |
| "learning_rate": 3.9999198907597046e-05, | |
| "loss": 0.7129, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 1.172279035409775, | |
| "learning_rate": 3.9996795694563096e-05, | |
| "loss": 0.7151, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 1.1267488528065377, | |
| "learning_rate": 3.999279055341771e-05, | |
| "loss": 0.697, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 0.8508223922961116, | |
| "learning_rate": 3.998718380500971e-05, | |
| "loss": 0.6984, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 0.7832878953186442, | |
| "learning_rate": 3.997997589849145e-05, | |
| "loss": 0.6938, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.8903650351894811, | |
| "learning_rate": 3.9971167411282835e-05, | |
| "loss": 0.7026, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 0.8088996234219692, | |
| "learning_rate": 3.99607590490251e-05, | |
| "loss": 0.7019, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 0.9141557760083519, | |
| "learning_rate": 3.9948751645524235e-05, | |
| "loss": 0.7074, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 1.110663777049766, | |
| "learning_rate": 3.9935146162684206e-05, | |
| "loss": 0.6862, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 1.1923838691808681, | |
| "learning_rate": 3.9919943690429906e-05, | |
| "loss": 0.6936, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 1.0597972011189978, | |
| "learning_rate": 3.9903145446619837e-05, | |
| "loss": 0.6885, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 1.0827234123938607, | |
| "learning_rate": 3.9884752776948564e-05, | |
| "loss": 0.6867, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 1.0103677441115202, | |
| "learning_rate": 3.9864767154838864e-05, | |
| "loss": 0.6813, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 1.438741141096182, | |
| "learning_rate": 3.9843190181323744e-05, | |
| "loss": 0.6986, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 1.0023451887913393, | |
| "learning_rate": 3.982002358491817e-05, | |
| "loss": 0.6845, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 1.7956087884014165, | |
| "learning_rate": 3.979526922148058e-05, | |
| "loss": 0.6902, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 1.3274409636573652, | |
| "learning_rate": 3.9768929074064206e-05, | |
| "loss": 0.6842, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 1.7394872777116857, | |
| "learning_rate": 3.9741005252758255e-05, | |
| "loss": 0.6915, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 1.378284034093402, | |
| "learning_rate": 3.971149999451886e-05, | |
| "loss": 0.6866, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 1.534673427940463, | |
| "learning_rate": 3.9680415662989806e-05, | |
| "loss": 0.6926, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 1.1153884104735996, | |
| "learning_rate": 3.9647754748313294e-05, | |
| "loss": 0.6754, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 1.36585767908662, | |
| "learning_rate": 3.96135198669304e-05, | |
| "loss": 0.6838, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 1.0483392103832203, | |
| "learning_rate": 3.957771376137144e-05, | |
| "loss": 0.6951, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 1.2080550973097068, | |
| "learning_rate": 3.954033930003634e-05, | |
| "loss": 0.6726, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 0.9385909186084465, | |
| "learning_rate": 3.9501399476964806e-05, | |
| "loss": 0.6751, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 1.219354144790043, | |
| "learning_rate": 3.946089741159648e-05, | |
| "loss": 0.6666, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 0.8868700154132488, | |
| "learning_rate": 3.9418836348521045e-05, | |
| "loss": 0.6823, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 1.1162551726609178, | |
| "learning_rate": 3.937521965721831e-05, | |
| "loss": 0.6503, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 0.8860936844471101, | |
| "learning_rate": 3.933005083178828e-05, | |
| "loss": 0.6637, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 1.012932179918778, | |
| "learning_rate": 3.928333349067125e-05, | |
| "loss": 0.6853, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 0.9055796845994227, | |
| "learning_rate": 3.923507137635792e-05, | |
| "loss": 0.6724, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 0.8635814480648324, | |
| "learning_rate": 3.9185268355089606e-05, | |
| "loss": 0.6845, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 0.8167363090854673, | |
| "learning_rate": 3.913392841654851e-05, | |
| "loss": 0.6658, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 0.7948905380773769, | |
| "learning_rate": 3.9081055673538093e-05, | |
| "loss": 0.6689, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 0.7140080639952283, | |
| "learning_rate": 3.902665436165364e-05, | |
| "loss": 0.668, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.6331506069616595, | |
| "learning_rate": 3.897072883894291e-05, | |
| "loss": 0.664, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 0.7962346309402896, | |
| "learning_rate": 3.8913283585557054e-05, | |
| "loss": 0.6641, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 0.7026370754498099, | |
| "learning_rate": 3.885432320339167e-05, | |
| "loss": 0.6538, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 0.7631291550658298, | |
| "learning_rate": 3.879385241571817e-05, | |
| "loss": 0.6864, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.0112, | |
| "grad_norm": 0.851784124709734, | |
| "learning_rate": 3.873187606680543e-05, | |
| "loss": 0.6196, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.024, | |
| "grad_norm": 0.6307389925589272, | |
| "learning_rate": 3.866839912153168e-05, | |
| "loss": 0.6021, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 0.7259669079516865, | |
| "learning_rate": 3.860342666498677e-05, | |
| "loss": 0.6101, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.0496, | |
| "grad_norm": 1.267499389834184, | |
| "learning_rate": 3.853696390206484e-05, | |
| "loss": 0.6131, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.0624, | |
| "grad_norm": 1.0494169261183413, | |
| "learning_rate": 3.846901615704734e-05, | |
| "loss": 0.6133, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 0.6050361569284982, | |
| "learning_rate": 3.839958887317649e-05, | |
| "loss": 0.6134, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.088, | |
| "grad_norm": 0.7628296952015292, | |
| "learning_rate": 3.832868761221926e-05, | |
| "loss": 0.6011, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.1008, | |
| "grad_norm": 1.0944287759562783, | |
| "learning_rate": 3.825631805402182e-05, | |
| "loss": 0.6088, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 0.7617167303570096, | |
| "learning_rate": 3.818248599605448e-05, | |
| "loss": 0.6034, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.1264, | |
| "grad_norm": 0.7078174928967245, | |
| "learning_rate": 3.810719735294731e-05, | |
| "loss": 0.6116, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.1392, | |
| "grad_norm": 0.8593342874747895, | |
| "learning_rate": 3.8030458156016326e-05, | |
| "loss": 0.6075, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.9136480226583094, | |
| "learning_rate": 3.795227455278029e-05, | |
| "loss": 0.6272, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1648, | |
| "grad_norm": 0.6448822801640876, | |
| "learning_rate": 3.787265280646825e-05, | |
| "loss": 0.5964, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.1776, | |
| "grad_norm": 0.6130018934012975, | |
| "learning_rate": 3.7791599295517825e-05, | |
| "loss": 0.61, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 0.7940913926445108, | |
| "learning_rate": 3.7709120513064196e-05, | |
| "loss": 0.6066, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.2032, | |
| "grad_norm": 0.7900375287324176, | |
| "learning_rate": 3.762522306641998e-05, | |
| "loss": 0.6248, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.216, | |
| "grad_norm": 1.3846465374358778, | |
| "learning_rate": 3.7539913676545874e-05, | |
| "loss": 0.6419, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 0.7604171538360377, | |
| "learning_rate": 3.745319917751229e-05, | |
| "loss": 0.6071, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.2416, | |
| "grad_norm": 1.1907289468782911, | |
| "learning_rate": 3.736508651595188e-05, | |
| "loss": 0.6131, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.2544, | |
| "grad_norm": 0.4755617311625865, | |
| "learning_rate": 3.727558275050301e-05, | |
| "loss": 0.6221, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 0.7750767344395085, | |
| "learning_rate": 3.718469505124434e-05, | |
| "loss": 0.5973, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.7985285588114858, | |
| "learning_rate": 3.709243069912041e-05, | |
| "loss": 0.597, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.2928, | |
| "grad_norm": 0.5174241535960574, | |
| "learning_rate": 3.699879708535838e-05, | |
| "loss": 0.6183, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 0.5606339078615882, | |
| "learning_rate": 3.69038017108759e-05, | |
| "loss": 0.6003, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.3184, | |
| "grad_norm": 0.6062393541277423, | |
| "learning_rate": 3.680745218568026e-05, | |
| "loss": 0.587, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.3312, | |
| "grad_norm": 0.6391795997589044, | |
| "learning_rate": 3.6709756228258735e-05, | |
| "loss": 0.6024, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.6817091963921497, | |
| "learning_rate": 3.6610721664960236e-05, | |
| "loss": 0.6119, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.3568, | |
| "grad_norm": 0.6202669908353199, | |
| "learning_rate": 3.65103564293684e-05, | |
| "loss": 0.5991, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.3696, | |
| "grad_norm": 0.7683520783743589, | |
| "learning_rate": 3.640866856166601e-05, | |
| "loss": 0.6079, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 0.6415282934164027, | |
| "learning_rate": 3.6305666207990886e-05, | |
| "loss": 0.5985, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.3952, | |
| "grad_norm": 0.6265527820310741, | |
| "learning_rate": 3.6201357619783336e-05, | |
| "loss": 0.6, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.408, | |
| "grad_norm": 0.7002458988543002, | |
| "learning_rate": 3.609575115312511e-05, | |
| "loss": 0.5953, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 0.7070127497108548, | |
| "learning_rate": 3.598885526807003e-05, | |
| "loss": 0.618, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.4336, | |
| "grad_norm": 0.7796274005308657, | |
| "learning_rate": 3.5880678527966224e-05, | |
| "loss": 0.6133, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.4464000000000001, | |
| "grad_norm": 0.6250805424349141, | |
| "learning_rate": 3.577122959877017e-05, | |
| "loss": 0.5831, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 0.5780614589643612, | |
| "learning_rate": 3.566051724835245e-05, | |
| "loss": 0.5992, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.472, | |
| "grad_norm": 0.4390012120658671, | |
| "learning_rate": 3.554855034579532e-05, | |
| "loss": 0.5939, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.4848, | |
| "grad_norm": 0.5307581134262008, | |
| "learning_rate": 3.5435337860682304e-05, | |
| "loss": 0.5971, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 0.49327020749453704, | |
| "learning_rate": 3.532088886237956e-05, | |
| "loss": 0.6088, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.5104, | |
| "grad_norm": 0.49054325181953373, | |
| "learning_rate": 3.520521251930941e-05, | |
| "loss": 0.5962, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.5232, | |
| "grad_norm": 0.523142272392975, | |
| "learning_rate": 3.5088318098215805e-05, | |
| "loss": 0.6103, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.39279963818343633, | |
| "learning_rate": 3.497021496342203e-05, | |
| "loss": 0.5967, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.5488, | |
| "grad_norm": 1.2743347614995277, | |
| "learning_rate": 3.485091257608047e-05, | |
| "loss": 0.6203, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.5615999999999999, | |
| "grad_norm": 0.4309032896655364, | |
| "learning_rate": 3.473042049341474e-05, | |
| "loss": 0.6116, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.5644891227864532, | |
| "learning_rate": 3.4608748367954064e-05, | |
| "loss": 0.5942, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.5872000000000002, | |
| "grad_norm": 0.5027406013499266, | |
| "learning_rate": 3.4485905946759965e-05, | |
| "loss": 0.6123, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.41067238794464184, | |
| "learning_rate": 3.4361903070645484e-05, | |
| "loss": 0.5865, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.601531071854875, | |
| "learning_rate": 3.423674967338681e-05, | |
| "loss": 0.6066, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.6256, | |
| "grad_norm": 0.5870807871707389, | |
| "learning_rate": 3.411045578092754e-05, | |
| "loss": 0.6024, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.6383999999999999, | |
| "grad_norm": 0.6083613677782111, | |
| "learning_rate": 3.398303151057543e-05, | |
| "loss": 0.608, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.5725717450042651, | |
| "learning_rate": 3.385448707019199e-05, | |
| "loss": 0.6003, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 0.5466944334380455, | |
| "learning_rate": 3.372483275737468e-05, | |
| "loss": 0.6009, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.6768, | |
| "grad_norm": 0.6581581051771994, | |
| "learning_rate": 3.359407895863199e-05, | |
| "loss": 0.5946, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 0.5170494043869698, | |
| "learning_rate": 3.34622361485514e-05, | |
| "loss": 0.5979, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.7024, | |
| "grad_norm": 0.4272177060967506, | |
| "learning_rate": 3.332931488896029e-05, | |
| "loss": 0.6126, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.7151999999999998, | |
| "grad_norm": 0.5854858600800749, | |
| "learning_rate": 3.319532582807977e-05, | |
| "loss": 0.6298, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.4603126803626732, | |
| "learning_rate": 3.30602796996717e-05, | |
| "loss": 0.6068, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.7408000000000001, | |
| "grad_norm": 0.4030643609116338, | |
| "learning_rate": 3.2924187322178865e-05, | |
| "loss": 0.5922, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.7536, | |
| "grad_norm": 0.521872401717636, | |
| "learning_rate": 3.278705959785821e-05, | |
| "loss": 0.6036, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 0.43323160460128696, | |
| "learning_rate": 3.2648907511907544e-05, | |
| "loss": 0.5933, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.7792, | |
| "grad_norm": 0.42147414713760073, | |
| "learning_rate": 3.250974213158555e-05, | |
| "loss": 0.603, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.792, | |
| "grad_norm": 0.44305552805403464, | |
| "learning_rate": 3.23695746053251e-05, | |
| "loss": 0.6024, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 0.4313818392118414, | |
| "learning_rate": 3.222841616184025e-05, | |
| "loss": 0.6029, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.8176, | |
| "grad_norm": 0.45442842158234487, | |
| "learning_rate": 3.208627810922665e-05, | |
| "loss": 0.5893, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.8304, | |
| "grad_norm": 0.43166573142906883, | |
| "learning_rate": 3.194317183405573e-05, | |
| "loss": 0.5966, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 0.4233581157421932, | |
| "learning_rate": 3.1799108800462466e-05, | |
| "loss": 0.597, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.8559999999999999, | |
| "grad_norm": 0.41983825635373967, | |
| "learning_rate": 3.1654100549227024e-05, | |
| "loss": 0.6016, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.8688, | |
| "grad_norm": 0.4775077223960118, | |
| "learning_rate": 3.1508158696850275e-05, | |
| "loss": 0.5935, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 0.49137137164056993, | |
| "learning_rate": 3.136129493462312e-05, | |
| "loss": 0.6167, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.8944, | |
| "grad_norm": 0.4327335425374451, | |
| "learning_rate": 3.121352102768998e-05, | |
| "loss": 0.593, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.9072, | |
| "grad_norm": 0.5490054155478232, | |
| "learning_rate": 3.106484881410628e-05, | |
| "loss": 0.6117, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.3910639442615754, | |
| "learning_rate": 3.091529020389009e-05, | |
| "loss": 0.598, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9327999999999999, | |
| "grad_norm": 0.44967337833903454, | |
| "learning_rate": 3.076485717806808e-05, | |
| "loss": 0.5908, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.9456, | |
| "grad_norm": 0.3412735134639727, | |
| "learning_rate": 3.061356178771564e-05, | |
| "loss": 0.6083, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.5062101356795918, | |
| "learning_rate": 3.0461416152991555e-05, | |
| "loss": 0.5845, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.9712, | |
| "grad_norm": 0.44404668482752346, | |
| "learning_rate": 3.0308432462167045e-05, | |
| "loss": 0.5864, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.984, | |
| "grad_norm": 0.4316637116575441, | |
| "learning_rate": 3.015462297064936e-05, | |
| "loss": 0.6027, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 0.3916492660806835, | |
| "learning_rate": 3.0000000000000004e-05, | |
| "loss": 0.5818, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.0096, | |
| "grad_norm": 0.4888543656887078, | |
| "learning_rate": 2.98445759369477e-05, | |
| "loss": 0.5543, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.0224, | |
| "grad_norm": 0.4815500557431066, | |
| "learning_rate": 2.9688363232396056e-05, | |
| "loss": 0.5258, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 0.8571848258445681, | |
| "learning_rate": 2.9531374400426158e-05, | |
| "loss": 0.5121, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 0.6441379504775101, | |
| "learning_rate": 2.9373622017294075e-05, | |
| "loss": 0.5064, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.0608, | |
| "grad_norm": 0.6704481350354612, | |
| "learning_rate": 2.9215118720423375e-05, | |
| "loss": 0.5216, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 1.1087985376379281, | |
| "learning_rate": 2.9055877207392752e-05, | |
| "loss": 0.5238, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.0864, | |
| "grad_norm": 0.5947531487765395, | |
| "learning_rate": 2.8895910234918828e-05, | |
| "loss": 0.5643, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.0992, | |
| "grad_norm": 0.5380072667042054, | |
| "learning_rate": 2.873523061783426e-05, | |
| "loss": 0.5395, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.6285029176514783, | |
| "learning_rate": 2.8573851228061084e-05, | |
| "loss": 0.5292, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.1248, | |
| "grad_norm": 0.5080988917676031, | |
| "learning_rate": 2.8411784993579633e-05, | |
| "loss": 0.5323, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.1376, | |
| "grad_norm": 0.6925953902952061, | |
| "learning_rate": 2.8249044897392814e-05, | |
| "loss": 0.5387, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.6144791118974732, | |
| "learning_rate": 2.80856439764861e-05, | |
| "loss": 0.527, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.1632, | |
| "grad_norm": 0.5641525509874317, | |
| "learning_rate": 2.792159532078314e-05, | |
| "loss": 0.5346, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 0.5822330796778116, | |
| "learning_rate": 2.77569120720971e-05, | |
| "loss": 0.5412, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 0.5139066091526103, | |
| "learning_rate": 2.7591607423077932e-05, | |
| "loss": 0.5195, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.2016, | |
| "grad_norm": 0.42590295273400064, | |
| "learning_rate": 2.7425694616155474e-05, | |
| "loss": 0.5198, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.2144, | |
| "grad_norm": 0.5042344292315121, | |
| "learning_rate": 2.7259186942478656e-05, | |
| "loss": 0.5078, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 0.49776813233689815, | |
| "learning_rate": 2.7092097740850712e-05, | |
| "loss": 0.5279, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.6478103171305779, | |
| "learning_rate": 2.692444039666066e-05, | |
| "loss": 0.5281, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.2528, | |
| "grad_norm": 0.4808064612691258, | |
| "learning_rate": 2.6756228340810946e-05, | |
| "loss": 0.5114, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 0.5599610114705597, | |
| "learning_rate": 2.6587475048641596e-05, | |
| "loss": 0.5248, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.2784, | |
| "grad_norm": 0.5115581463086503, | |
| "learning_rate": 2.6418194038850634e-05, | |
| "loss": 0.5205, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.2912, | |
| "grad_norm": 1.1296918134293887, | |
| "learning_rate": 2.624839887241115e-05, | |
| "loss": 0.5119, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 2.981819583523617, | |
| "learning_rate": 2.607810315148494e-05, | |
| "loss": 0.5256, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.3168, | |
| "grad_norm": 0.9514646089409182, | |
| "learning_rate": 2.5907320518332827e-05, | |
| "loss": 0.5256, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.3296, | |
| "grad_norm": 0.7760907258850646, | |
| "learning_rate": 2.5736064654221808e-05, | |
| "loss": 0.53, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.6226637778314663, | |
| "learning_rate": 2.5564349278329056e-05, | |
| "loss": 0.5216, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.3552, | |
| "grad_norm": 0.6771255180025393, | |
| "learning_rate": 2.539218814664288e-05, | |
| "loss": 0.5102, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 11.05011318668379, | |
| "learning_rate": 2.521959505086075e-05, | |
| "loss": 0.5422, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 1.0244472713681818, | |
| "learning_rate": 2.5046583817284437e-05, | |
| "loss": 0.5194, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.3936, | |
| "grad_norm": 1.2557015062288144, | |
| "learning_rate": 2.487316830571244e-05, | |
| "loss": 0.5246, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.4064, | |
| "grad_norm": 0.6907933031791822, | |
| "learning_rate": 2.4699362408329646e-05, | |
| "loss": 0.5253, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 1.5167808710335784, | |
| "learning_rate": 2.4525180048594452e-05, | |
| "loss": 0.5312, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.432, | |
| "grad_norm": 0.8383632337756304, | |
| "learning_rate": 2.435063518012335e-05, | |
| "loss": 0.5227, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.4448, | |
| "grad_norm": 1.5873251457007391, | |
| "learning_rate": 2.4175741785573177e-05, | |
| "loss": 0.5389, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 1.2763322421660372, | |
| "learning_rate": 2.4000513875520892e-05, | |
| "loss": 0.5291, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.4704, | |
| "grad_norm": 1.151744306753643, | |
| "learning_rate": 2.3824965487341247e-05, | |
| "loss": 0.5183, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.4832, | |
| "grad_norm": 1.1106685552386601, | |
| "learning_rate": 2.3649110684082258e-05, | |
| "loss": 0.5148, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.9018675862280014, | |
| "learning_rate": 2.3472963553338614e-05, | |
| "loss": 0.521, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.5088, | |
| "grad_norm": 0.9207401583573775, | |
| "learning_rate": 2.3296538206123134e-05, | |
| "loss": 0.521, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.5216, | |
| "grad_norm": 0.780896793219167, | |
| "learning_rate": 2.311984877573636e-05, | |
| "loss": 0.5205, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.81296689027176, | |
| "learning_rate": 2.2942909416634326e-05, | |
| "loss": 0.517, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.5472, | |
| "grad_norm": 0.7652123891830968, | |
| "learning_rate": 2.2765734303294666e-05, | |
| "loss": 0.5181, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.7472526183951649, | |
| "learning_rate": 2.2588337629081107e-05, | |
| "loss": 0.5154, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.6900588717281774, | |
| "learning_rate": 2.2410733605106462e-05, | |
| "loss": 0.5313, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.5856, | |
| "grad_norm": 0.7014835732432961, | |
| "learning_rate": 2.2232936459094158e-05, | |
| "loss": 0.5264, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.5984, | |
| "grad_norm": 0.6334650137837722, | |
| "learning_rate": 2.205496043423849e-05, | |
| "loss": 0.5133, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.6068938258440488, | |
| "learning_rate": 2.1876819788063586e-05, | |
| "loss": 0.5251, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.624, | |
| "grad_norm": 0.6420854105273656, | |
| "learning_rate": 2.16985287912813e-05, | |
| "loss": 0.5142, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.6368, | |
| "grad_norm": 0.47359144811086606, | |
| "learning_rate": 2.1520101726647922e-05, | |
| "loss": 0.5145, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.7107111946039573, | |
| "learning_rate": 2.1341552887820048e-05, | |
| "loss": 0.5384, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.6624, | |
| "grad_norm": 0.36445714525911443, | |
| "learning_rate": 2.1162896578209517e-05, | |
| "loss": 0.5345, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.6752000000000002, | |
| "grad_norm": 0.6123793686108365, | |
| "learning_rate": 2.0984147109837564e-05, | |
| "loss": 0.5161, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.4157114262816254, | |
| "learning_rate": 2.0805318802188307e-05, | |
| "loss": 0.522, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.7008, | |
| "grad_norm": 0.48438044456699125, | |
| "learning_rate": 2.0626425981061608e-05, | |
| "loss": 0.5242, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.7136, | |
| "grad_norm": 0.5191429309709934, | |
| "learning_rate": 2.0447482977425465e-05, | |
| "loss": 0.5208, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.38699267500264917, | |
| "learning_rate": 2.0268504126267952e-05, | |
| "loss": 0.5226, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.7392, | |
| "grad_norm": 0.46715551010372003, | |
| "learning_rate": 2.008950376544887e-05, | |
| "loss": 0.511, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 0.41115294696217414, | |
| "learning_rate": 1.9910496234551132e-05, | |
| "loss": 0.5185, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.3842758319982132, | |
| "learning_rate": 1.9731495873732055e-05, | |
| "loss": 0.5237, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.7776, | |
| "grad_norm": 0.41930663467606166, | |
| "learning_rate": 1.9552517022574542e-05, | |
| "loss": 0.5228, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.7904, | |
| "grad_norm": 0.31054850171535525, | |
| "learning_rate": 1.93735740189384e-05, | |
| "loss": 0.5077, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.3940170724243544, | |
| "learning_rate": 1.9194681197811703e-05, | |
| "loss": 0.5161, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 0.3713210620901307, | |
| "learning_rate": 1.901585289016244e-05, | |
| "loss": 0.5171, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.8288, | |
| "grad_norm": 0.3571748502702753, | |
| "learning_rate": 1.8837103421790486e-05, | |
| "loss": 0.5298, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.37813522602033334, | |
| "learning_rate": 1.8658447112179952e-05, | |
| "loss": 0.5258, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.8544, | |
| "grad_norm": 0.32175707737347425, | |
| "learning_rate": 1.8479898273352084e-05, | |
| "loss": 0.5228, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.8672, | |
| "grad_norm": 0.3995308189650411, | |
| "learning_rate": 1.83014712087187e-05, | |
| "loss": 0.5194, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.3136395219390525, | |
| "learning_rate": 1.8123180211936417e-05, | |
| "loss": 0.5122, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.8928000000000003, | |
| "grad_norm": 0.4112786760550883, | |
| "learning_rate": 1.794503956576152e-05, | |
| "loss": 0.5194, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.9055999999999997, | |
| "grad_norm": 0.3719752981770789, | |
| "learning_rate": 1.776706354090585e-05, | |
| "loss": 0.5168, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.33626436393752196, | |
| "learning_rate": 1.758926639489354e-05, | |
| "loss": 0.532, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.9312, | |
| "grad_norm": 0.3733036341893251, | |
| "learning_rate": 1.7411662370918893e-05, | |
| "loss": 0.5329, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 0.32622867849007553, | |
| "learning_rate": 1.7234265696705344e-05, | |
| "loss": 0.5099, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 0.31745223013022084, | |
| "learning_rate": 1.7057090583365678e-05, | |
| "loss": 0.527, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.9696, | |
| "grad_norm": 0.28328283021260636, | |
| "learning_rate": 1.6880151224263646e-05, | |
| "loss": 0.5046, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.9824, | |
| "grad_norm": 0.2913715628130577, | |
| "learning_rate": 1.6703461793876876e-05, | |
| "loss": 0.5114, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.29466681227825775, | |
| "learning_rate": 1.6527036446661396e-05, | |
| "loss": 0.5212, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 3.008, | |
| "grad_norm": 0.3690756941642924, | |
| "learning_rate": 1.635088931591775e-05, | |
| "loss": 0.4906, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 3.0208, | |
| "grad_norm": 0.3781793273396284, | |
| "learning_rate": 1.6175034512658753e-05, | |
| "loss": 0.4553, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 3.0336, | |
| "grad_norm": 0.3959199001237425, | |
| "learning_rate": 1.5999486124479115e-05, | |
| "loss": 0.454, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 3.0464, | |
| "grad_norm": 0.5327881021520445, | |
| "learning_rate": 1.5824258214426833e-05, | |
| "loss": 0.4601, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 3.0592, | |
| "grad_norm": 0.44602841924437264, | |
| "learning_rate": 1.5649364819876655e-05, | |
| "loss": 0.471, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 3.072, | |
| "grad_norm": 0.4244406743987455, | |
| "learning_rate": 1.547481995140556e-05, | |
| "loss": 0.4619, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.0848, | |
| "grad_norm": 0.49615891983164706, | |
| "learning_rate": 1.5300637591670357e-05, | |
| "loss": 0.4714, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 3.0976, | |
| "grad_norm": 0.37527641868046685, | |
| "learning_rate": 1.5126831694287564e-05, | |
| "loss": 0.4487, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 3.1104, | |
| "grad_norm": 0.5002935031588264, | |
| "learning_rate": 1.4953416182715566e-05, | |
| "loss": 0.4881, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 3.1232, | |
| "grad_norm": 0.4182656757519464, | |
| "learning_rate": 1.478040494913926e-05, | |
| "loss": 0.465, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.136, | |
| "grad_norm": 0.36889800297143954, | |
| "learning_rate": 1.460781185335713e-05, | |
| "loss": 0.4477, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 3.1488, | |
| "grad_norm": 0.4671266896308205, | |
| "learning_rate": 1.443565072167095e-05, | |
| "loss": 0.4608, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 3.1616, | |
| "grad_norm": 0.33080249079185275, | |
| "learning_rate": 1.4263935345778202e-05, | |
| "loss": 0.4537, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 3.1744, | |
| "grad_norm": 0.35142616139830696, | |
| "learning_rate": 1.409267948166718e-05, | |
| "loss": 0.4671, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 3.1872, | |
| "grad_norm": 0.3869870153646901, | |
| "learning_rate": 1.3921896848515064e-05, | |
| "loss": 0.4523, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.3101503604126732, | |
| "learning_rate": 1.3751601127588849e-05, | |
| "loss": 0.4448, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.2128, | |
| "grad_norm": 0.3431117745891633, | |
| "learning_rate": 1.3581805961149371e-05, | |
| "loss": 0.4551, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 3.2256, | |
| "grad_norm": 0.3009185171263903, | |
| "learning_rate": 1.341252495135841e-05, | |
| "loss": 0.4471, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 3.2384, | |
| "grad_norm": 0.35198592063045986, | |
| "learning_rate": 1.324377165918906e-05, | |
| "loss": 0.4467, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 3.2512, | |
| "grad_norm": 0.305334219012278, | |
| "learning_rate": 1.3075559603339354e-05, | |
| "loss": 0.4422, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 3.2640000000000002, | |
| "grad_norm": 0.33463011724595837, | |
| "learning_rate": 1.2907902259149287e-05, | |
| "loss": 0.4526, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 3.2768, | |
| "grad_norm": 0.3307553833226135, | |
| "learning_rate": 1.274081305752135e-05, | |
| "loss": 0.4606, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 3.2896, | |
| "grad_norm": 0.29944988009729223, | |
| "learning_rate": 1.2574305383844528e-05, | |
| "loss": 0.4624, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 3.3024, | |
| "grad_norm": 0.33125913079360175, | |
| "learning_rate": 1.2408392576922075e-05, | |
| "loss": 0.4488, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 3.3152, | |
| "grad_norm": 0.3157204486791319, | |
| "learning_rate": 1.2243087927902905e-05, | |
| "loss": 0.4703, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 3.328, | |
| "grad_norm": 0.49629783732261507, | |
| "learning_rate": 1.2078404679216864e-05, | |
| "loss": 0.4415, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 3.3407999999999998, | |
| "grad_norm": 0.33875848926814134, | |
| "learning_rate": 1.1914356023513904e-05, | |
| "loss": 0.4519, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 3.3536, | |
| "grad_norm": 0.30958638670838906, | |
| "learning_rate": 1.1750955102607193e-05, | |
| "loss": 0.4592, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 3.3664, | |
| "grad_norm": 0.3163752620814582, | |
| "learning_rate": 1.1588215006420374e-05, | |
| "loss": 0.457, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 3.3792, | |
| "grad_norm": 0.29159197402925424, | |
| "learning_rate": 1.1426148771938915e-05, | |
| "loss": 0.4661, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 3.392, | |
| "grad_norm": 0.3282379766739495, | |
| "learning_rate": 1.1264769382165748e-05, | |
| "loss": 0.4341, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 3.4048, | |
| "grad_norm": 0.30438472052025084, | |
| "learning_rate": 1.110408976508118e-05, | |
| "loss": 0.4523, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 3.4176, | |
| "grad_norm": 0.2771592723904903, | |
| "learning_rate": 1.094412279260726e-05, | |
| "loss": 0.4591, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 3.4304, | |
| "grad_norm": 0.2793279801320436, | |
| "learning_rate": 1.0784881279576635e-05, | |
| "loss": 0.4519, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 3.4432, | |
| "grad_norm": 0.2885870595457907, | |
| "learning_rate": 1.0626377982705929e-05, | |
| "loss": 0.4525, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 3.456, | |
| "grad_norm": 0.2769749165466486, | |
| "learning_rate": 1.0468625599573842e-05, | |
| "loss": 0.4643, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.4688, | |
| "grad_norm": 0.26576176960473713, | |
| "learning_rate": 1.0311636767603952e-05, | |
| "loss": 0.4458, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 3.4816, | |
| "grad_norm": 0.2927848520728921, | |
| "learning_rate": 1.0155424063052306e-05, | |
| "loss": 0.4512, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 3.4944, | |
| "grad_norm": 0.2904017409587212, | |
| "learning_rate": 1.0000000000000006e-05, | |
| "loss": 0.4686, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 3.5072, | |
| "grad_norm": 0.30287993097797533, | |
| "learning_rate": 9.84537702935065e-06, | |
| "loss": 0.4536, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 0.27295541325769884, | |
| "learning_rate": 9.691567537832964e-06, | |
| "loss": 0.462, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.5328, | |
| "grad_norm": 0.2683263073893058, | |
| "learning_rate": 9.538583847008452e-06, | |
| "loss": 0.4429, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 3.5456, | |
| "grad_norm": 0.2569896613414521, | |
| "learning_rate": 9.386438212284372e-06, | |
| "loss": 0.4694, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 3.5584, | |
| "grad_norm": 0.2706926067299209, | |
| "learning_rate": 9.235142821931928e-06, | |
| "loss": 0.4628, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 3.5712, | |
| "grad_norm": 0.276525439907565, | |
| "learning_rate": 9.084709796109907e-06, | |
| "loss": 0.4647, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 3.584, | |
| "grad_norm": 0.8636642822521984, | |
| "learning_rate": 8.93515118589373e-06, | |
| "loss": 0.472, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.5968, | |
| "grad_norm": 0.25402516720905227, | |
| "learning_rate": 8.786478972310023e-06, | |
| "loss": 0.4602, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 3.6096, | |
| "grad_norm": 0.2813512087588087, | |
| "learning_rate": 8.638705065376887e-06, | |
| "loss": 0.4512, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 3.6224, | |
| "grad_norm": 0.2655867756240358, | |
| "learning_rate": 8.491841303149728e-06, | |
| "loss": 0.4484, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 3.6352, | |
| "grad_norm": 0.2454900100726722, | |
| "learning_rate": 8.345899450772975e-06, | |
| "loss": 0.4444, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 3.648, | |
| "grad_norm": 0.2540614684282997, | |
| "learning_rate": 8.200891199537549e-06, | |
| "loss": 0.4513, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 3.6608, | |
| "grad_norm": 0.2503768007575843, | |
| "learning_rate": 8.056828165944282e-06, | |
| "loss": 0.4466, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 3.6736, | |
| "grad_norm": 0.24413788351950494, | |
| "learning_rate": 7.913721890773354e-06, | |
| "loss": 0.4663, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 3.6864, | |
| "grad_norm": 0.24737629699824315, | |
| "learning_rate": 7.771583838159756e-06, | |
| "loss": 0.446, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.6992000000000003, | |
| "grad_norm": 0.247884049405193, | |
| "learning_rate": 7.630425394674903e-06, | |
| "loss": 0.4723, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 3.7119999999999997, | |
| "grad_norm": 0.23896001680765466, | |
| "learning_rate": 7.49025786841445e-06, | |
| "loss": 0.468, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.7248, | |
| "grad_norm": 0.24349299333550553, | |
| "learning_rate": 7.3510924880924575e-06, | |
| "loss": 0.4662, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 3.7376, | |
| "grad_norm": 0.25759618368339365, | |
| "learning_rate": 7.212940402141808e-06, | |
| "loss": 0.456, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 3.7504, | |
| "grad_norm": 0.2783925560786573, | |
| "learning_rate": 7.075812677821145e-06, | |
| "loss": 0.4386, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 3.7632, | |
| "grad_norm": 0.24606421907763323, | |
| "learning_rate": 6.939720300328303e-06, | |
| "loss": 0.4441, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 3.776, | |
| "grad_norm": 0.2655587047503392, | |
| "learning_rate": 6.8046741719202385e-06, | |
| "loss": 0.4503, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 3.7888, | |
| "grad_norm": 0.25711481251601814, | |
| "learning_rate": 6.67068511103971e-06, | |
| "loss": 0.4538, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 3.8016, | |
| "grad_norm": 0.26078745541115567, | |
| "learning_rate": 6.537763851448593e-06, | |
| "loss": 0.4527, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 3.8144, | |
| "grad_norm": 0.24844160542227053, | |
| "learning_rate": 6.4059210413680175e-06, | |
| "loss": 0.4583, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 3.8272, | |
| "grad_norm": 0.24768515864599866, | |
| "learning_rate": 6.275167242625331e-06, | |
| "loss": 0.4373, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.239785192282427, | |
| "learning_rate": 6.145512929808013e-06, | |
| "loss": 0.4692, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.8528000000000002, | |
| "grad_norm": 0.23890682030137445, | |
| "learning_rate": 6.016968489424572e-06, | |
| "loss": 0.4558, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 3.8656, | |
| "grad_norm": 0.2371949740196673, | |
| "learning_rate": 5.889544219072465e-06, | |
| "loss": 0.4711, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 3.8784, | |
| "grad_norm": 0.23607828766214778, | |
| "learning_rate": 5.7632503266131925e-06, | |
| "loss": 0.4354, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 3.8912, | |
| "grad_norm": 0.22506466995573687, | |
| "learning_rate": 5.638096929354522e-06, | |
| "loss": 0.469, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 3.904, | |
| "grad_norm": 0.24540851264258934, | |
| "learning_rate": 5.514094053240035e-06, | |
| "loss": 0.4586, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 3.9168, | |
| "grad_norm": 0.2188922440355333, | |
| "learning_rate": 5.39125163204594e-06, | |
| "loss": 0.443, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 3.9295999999999998, | |
| "grad_norm": 0.2393711334973867, | |
| "learning_rate": 5.269579506585259e-06, | |
| "loss": 0.453, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 3.9424, | |
| "grad_norm": 0.22674654339309047, | |
| "learning_rate": 5.149087423919541e-06, | |
| "loss": 0.4607, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 3.9552, | |
| "grad_norm": 0.21287599788811704, | |
| "learning_rate": 5.029785036577976e-06, | |
| "loss": 0.4514, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 3.968, | |
| "grad_norm": 0.22761687227523666, | |
| "learning_rate": 4.911681901784198e-06, | |
| "loss": 0.4487, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.9808, | |
| "grad_norm": 0.2402884859858742, | |
| "learning_rate": 4.794787480690597e-06, | |
| "loss": 0.442, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 3.9936, | |
| "grad_norm": 0.2313528713603373, | |
| "learning_rate": 4.679111137620442e-06, | |
| "loss": 0.448, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 4.0064, | |
| "grad_norm": 0.25923647921403925, | |
| "learning_rate": 4.5646621393177e-06, | |
| "loss": 0.4229, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 4.0192, | |
| "grad_norm": 0.33285765144302826, | |
| "learning_rate": 4.451449654204685e-06, | |
| "loss": 0.4092, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 4.032, | |
| "grad_norm": 0.2599331525350334, | |
| "learning_rate": 4.339482751647557e-06, | |
| "loss": 0.4061, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 4.0448, | |
| "grad_norm": 0.2432945188441946, | |
| "learning_rate": 4.228770401229824e-06, | |
| "loss": 0.4082, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 4.0576, | |
| "grad_norm": 0.32677652351404196, | |
| "learning_rate": 4.119321472033779e-06, | |
| "loss": 0.4184, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 4.0704, | |
| "grad_norm": 0.3407478280987734, | |
| "learning_rate": 4.011144731929981e-06, | |
| "loss": 0.4122, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 4.0832, | |
| "grad_norm": 0.2778844197348087, | |
| "learning_rate": 3.904248846874894e-06, | |
| "loss": 0.4147, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 4.096, | |
| "grad_norm": 0.26746213254568313, | |
| "learning_rate": 3.7986423802166705e-06, | |
| "loss": 0.4223, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 4.1088, | |
| "grad_norm": 0.29699977757607166, | |
| "learning_rate": 3.694333792009115e-06, | |
| "loss": 0.4017, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 4.1216, | |
| "grad_norm": 0.2879614691371465, | |
| "learning_rate": 3.5913314383339937e-06, | |
| "loss": 0.4135, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 4.1344, | |
| "grad_norm": 0.260955019719919, | |
| "learning_rate": 3.4896435706316e-06, | |
| "loss": 0.4203, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 4.1472, | |
| "grad_norm": 0.6200209971332792, | |
| "learning_rate": 3.3892783350397675e-06, | |
| "loss": 0.438, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 0.26191920958488224, | |
| "learning_rate": 3.290243771741275e-06, | |
| "loss": 0.4173, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 4.1728, | |
| "grad_norm": 0.2428343438447167, | |
| "learning_rate": 3.1925478143197418e-06, | |
| "loss": 0.4085, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 4.1856, | |
| "grad_norm": 0.5249320022495597, | |
| "learning_rate": 3.0961982891241083e-06, | |
| "loss": 0.4216, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 4.1984, | |
| "grad_norm": 0.26986557816676215, | |
| "learning_rate": 3.001202914641628e-06, | |
| "loss": 0.4106, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 4.2112, | |
| "grad_norm": 0.2559336288037187, | |
| "learning_rate": 2.907569300879596e-06, | |
| "loss": 0.437, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 4.224, | |
| "grad_norm": 0.23889646716135812, | |
| "learning_rate": 2.815304948755664e-06, | |
| "loss": 0.4056, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 4.2368, | |
| "grad_norm": 0.23493043476873637, | |
| "learning_rate": 2.7244172494969978e-06, | |
| "loss": 0.4022, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 4.2496, | |
| "grad_norm": 0.23133955451257487, | |
| "learning_rate": 2.6349134840481294e-06, | |
| "loss": 0.416, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 4.2624, | |
| "grad_norm": 0.22891191276988695, | |
| "learning_rate": 2.546800822487714e-06, | |
| "loss": 0.4171, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 4.2752, | |
| "grad_norm": 0.24809037402458942, | |
| "learning_rate": 2.4600863234541338e-06, | |
| "loss": 0.4228, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 4.288, | |
| "grad_norm": 0.24100785127402222, | |
| "learning_rate": 2.374776933580025e-06, | |
| "loss": 0.4026, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 4.3008, | |
| "grad_norm": 0.2247433535482438, | |
| "learning_rate": 2.2908794869358044e-06, | |
| "loss": 0.4114, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 4.3136, | |
| "grad_norm": 0.21331618403246796, | |
| "learning_rate": 2.2084007044821764e-06, | |
| "loss": 0.4204, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 4.3264, | |
| "grad_norm": 0.2283106560914472, | |
| "learning_rate": 2.127347193531757e-06, | |
| "loss": 0.4249, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 4.3392, | |
| "grad_norm": 0.230287391007796, | |
| "learning_rate": 2.0477254472197237e-06, | |
| "loss": 0.3998, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 4.352, | |
| "grad_norm": 0.23909156095067574, | |
| "learning_rate": 1.96954184398368e-06, | |
| "loss": 0.4043, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 4.3648, | |
| "grad_norm": 0.2203142411781777, | |
| "learning_rate": 1.8928026470526917e-06, | |
| "loss": 0.4255, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 4.3776, | |
| "grad_norm": 0.22933571285080048, | |
| "learning_rate": 1.817514003945524e-06, | |
| "loss": 0.4102, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 4.3904, | |
| "grad_norm": 0.21544013688903116, | |
| "learning_rate": 1.743681945978184e-06, | |
| "loss": 0.4184, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 4.4032, | |
| "grad_norm": 0.2039050932176269, | |
| "learning_rate": 1.6713123877807413e-06, | |
| "loss": 0.4288, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 4.416, | |
| "grad_norm": 0.20580730917625772, | |
| "learning_rate": 1.6004111268235156e-06, | |
| "loss": 0.4364, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 4.4288, | |
| "grad_norm": 0.21995247755306507, | |
| "learning_rate": 1.5309838429526714e-06, | |
| "loss": 0.3962, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 4.4416, | |
| "grad_norm": 0.21924100411014943, | |
| "learning_rate": 1.4630360979351644e-06, | |
| "loss": 0.4039, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 4.4544, | |
| "grad_norm": 0.21523101722080548, | |
| "learning_rate": 1.396573335013236e-06, | |
| "loss": 0.4121, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 4.4672, | |
| "grad_norm": 0.20920136764612174, | |
| "learning_rate": 1.3316008784683265e-06, | |
| "loss": 0.4292, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 0.21066440981135176, | |
| "learning_rate": 1.2681239331945695e-06, | |
| "loss": 0.421, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.4928, | |
| "grad_norm": 0.20371784575370688, | |
| "learning_rate": 1.2061475842818337e-06, | |
| "loss": 0.4134, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 4.5056, | |
| "grad_norm": 0.2057701483966716, | |
| "learning_rate": 1.1456767966083393e-06, | |
| "loss": 0.4113, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 4.5184, | |
| "grad_norm": 0.20842750880006203, | |
| "learning_rate": 1.086716414442952e-06, | |
| "loss": 0.4098, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 4.5312, | |
| "grad_norm": 0.2075288087992605, | |
| "learning_rate": 1.0292711610570904e-06, | |
| "loss": 0.4252, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 4.5440000000000005, | |
| "grad_norm": 0.2086258319606851, | |
| "learning_rate": 9.733456383463658e-07, | |
| "loss": 0.4083, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 4.5568, | |
| "grad_norm": 0.20645550024356008, | |
| "learning_rate": 9.189443264619102e-07, | |
| "loss": 0.3998, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 4.5696, | |
| "grad_norm": 0.20767967607750082, | |
| "learning_rate": 8.660715834514977e-07, | |
| "loss": 0.4105, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 4.5824, | |
| "grad_norm": 0.19974979009292662, | |
| "learning_rate": 8.147316449103959e-07, | |
| "loss": 0.4172, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 4.5952, | |
| "grad_norm": 0.20069080791679, | |
| "learning_rate": 7.649286236420806e-07, | |
| "loss": 0.4091, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 4.608, | |
| "grad_norm": 0.2030678221429786, | |
| "learning_rate": 7.166665093287539e-07, | |
| "loss": 0.4077, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.6208, | |
| "grad_norm": 0.2003773469477473, | |
| "learning_rate": 6.69949168211721e-07, | |
| "loss": 0.4317, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 4.6336, | |
| "grad_norm": 0.2074219498499131, | |
| "learning_rate": 6.247803427816945e-07, | |
| "loss": 0.4306, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 4.6464, | |
| "grad_norm": 0.2025643419254546, | |
| "learning_rate": 5.811636514789598e-07, | |
| "loss": 0.4123, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 4.6592, | |
| "grad_norm": 0.21023127440408973, | |
| "learning_rate": 5.391025884035239e-07, | |
| "loss": 0.4072, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 4.672, | |
| "grad_norm": 0.2038288102542892, | |
| "learning_rate": 4.986005230351954e-07, | |
| "loss": 0.4173, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 4.6848, | |
| "grad_norm": 0.20144566262199845, | |
| "learning_rate": 4.5966069996365993e-07, | |
| "loss": 0.4, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.6975999999999996, | |
| "grad_norm": 0.2041251627848762, | |
| "learning_rate": 4.22286238628562e-07, | |
| "loss": 0.4123, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 4.7104, | |
| "grad_norm": 0.20040962022672074, | |
| "learning_rate": 3.8648013306960664e-07, | |
| "loss": 0.4203, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 4.7232, | |
| "grad_norm": 0.2068938635693079, | |
| "learning_rate": 3.522452516867048e-07, | |
| "loss": 0.4118, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 4.736, | |
| "grad_norm": 0.20837284537704034, | |
| "learning_rate": 3.1958433701019697e-07, | |
| "loss": 0.4079, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.7488, | |
| "grad_norm": 0.20865068614826446, | |
| "learning_rate": 2.8850000548115155e-07, | |
| "loss": 0.3994, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 4.7616, | |
| "grad_norm": 0.19773627334509386, | |
| "learning_rate": 2.5899474724174313e-07, | |
| "loss": 0.4306, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 4.7744, | |
| "grad_norm": 0.19653283513789593, | |
| "learning_rate": 2.3107092593579905e-07, | |
| "loss": 0.4251, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 4.7872, | |
| "grad_norm": 0.20228291782179902, | |
| "learning_rate": 2.0473077851942858e-07, | |
| "loss": 0.4051, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.19840013104168644, | |
| "learning_rate": 1.799764150818306e-07, | |
| "loss": 0.4216, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 4.8128, | |
| "grad_norm": 0.2145103959909319, | |
| "learning_rate": 1.5680981867625566e-07, | |
| "loss": 0.4017, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 4.8256, | |
| "grad_norm": 0.20746316609970863, | |
| "learning_rate": 1.3523284516113955e-07, | |
| "loss": 0.4075, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 4.8384, | |
| "grad_norm": 0.20627956280546833, | |
| "learning_rate": 1.1524722305144231e-07, | |
| "loss": 0.4153, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 4.8512, | |
| "grad_norm": 0.2033869537362194, | |
| "learning_rate": 9.685455338016347e-08, | |
| "loss": 0.4284, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 4.864, | |
| "grad_norm": 0.20191829942597284, | |
| "learning_rate": 8.005630957010014e-08, | |
| "loss": 0.4379, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.8768, | |
| "grad_norm": 0.2015559976085769, | |
| "learning_rate": 6.485383731580142e-08, | |
| "loss": 0.4223, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 4.8896, | |
| "grad_norm": 0.19932967545655483, | |
| "learning_rate": 5.1248354475768034e-08, | |
| "loss": 0.4091, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 4.9024, | |
| "grad_norm": 0.20043951940947408, | |
| "learning_rate": 3.924095097489922e-08, | |
| "loss": 0.4092, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 4.9152000000000005, | |
| "grad_norm": 0.20344949870903242, | |
| "learning_rate": 2.8832588717164766e-08, | |
| "loss": 0.4231, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.928, | |
| "grad_norm": 0.2102344094357479, | |
| "learning_rate": 2.0024101508555604e-08, | |
| "loss": 0.3954, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 4.9408, | |
| "grad_norm": 0.19895757088144383, | |
| "learning_rate": 1.281619499029274e-08, | |
| "loss": 0.4179, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 4.9536, | |
| "grad_norm": 0.20118845143462785, | |
| "learning_rate": 7.209446582292501e-09, | |
| "loss": 0.4213, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 4.9664, | |
| "grad_norm": 0.19855226255255506, | |
| "learning_rate": 3.2043054369057523e-09, | |
| "loss": 0.416, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 4.9792, | |
| "grad_norm": 0.2022857920371411, | |
| "learning_rate": 8.010924029533406e-10, | |
| "loss": 0.4095, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "grad_norm": 0.20162895663254343, | |
| "learning_rate": 0.0, | |
| "loss": 0.4022, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "step": 390, | |
| "total_flos": 2.1313081500059566e+18, | |
| "train_loss": 0.5509019647653286, | |
| "train_runtime": 63715.1552, | |
| "train_samples_per_second": 0.785, | |
| "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": 2.1313081500059566e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |