| { | |
| "best_metric": 73.63219561895059, | |
| "best_model_checkpoint": "outputs/bert-small/weak_tiny_poe/mnli_21/checkpoint-36816", | |
| "epoch": 3.0, | |
| "global_step": 36816, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 1.9728378965667103e-05, | |
| "loss": 1.0604, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 1.9456757931334205e-05, | |
| "loss": 1.0403, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 1.9185136897001307e-05, | |
| "loss": 1.0406, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 1.8913515862668405e-05, | |
| "loss": 1.0413, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 1.8641894828335507e-05, | |
| "loss": 1.033, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 1.837027379400261e-05, | |
| "loss": 1.0259, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 1.809865275966971e-05, | |
| "loss": 1.0157, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 1.7827031725336812e-05, | |
| "loss": 1.0122, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 1.7555410691003914e-05, | |
| "loss": 0.9959, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 1.7283789656671015e-05, | |
| "loss": 0.993, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 1.7012168622338114e-05, | |
| "loss": 0.9812, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 1.6740547588005215e-05, | |
| "loss": 0.9753, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 1.6468926553672317e-05, | |
| "loss": 0.9835, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 1.619730551933942e-05, | |
| "loss": 0.9689, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 1.592568448500652e-05, | |
| "loss": 0.9522, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 1.5654063450673622e-05, | |
| "loss": 0.9631, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 1.5382442416340724e-05, | |
| "loss": 0.9604, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 1.5110821382007822e-05, | |
| "loss": 0.9531, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 1.4839200347674924e-05, | |
| "loss": 0.9446, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 1.4567579313342026e-05, | |
| "loss": 0.937, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 1.4295958279009128e-05, | |
| "loss": 0.9305, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.402433724467623e-05, | |
| "loss": 0.9372, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 1.3752716210343331e-05, | |
| "loss": 0.9268, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 1.3481095176010431e-05, | |
| "loss": 0.9348, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 70.20886398369842, | |
| "eval_combined_score": 70.20886398369842, | |
| "eval_loss": 0.7318788170814514, | |
| "eval_runtime": 23.7297, | |
| "eval_samples_per_second": 413.617, | |
| "eval_steps_per_second": 1.644, | |
| "step": 12272 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "learning_rate": 1.3209474141677531e-05, | |
| "loss": 0.9, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "learning_rate": 1.2937853107344633e-05, | |
| "loss": 0.8926, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "learning_rate": 1.2666232073011735e-05, | |
| "loss": 0.8989, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "learning_rate": 1.2394611038678836e-05, | |
| "loss": 0.8987, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "learning_rate": 1.2122990004345938e-05, | |
| "loss": 0.8936, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 1.185136897001304e-05, | |
| "loss": 0.8854, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "learning_rate": 1.157974793568014e-05, | |
| "loss": 0.8695, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "learning_rate": 1.130812690134724e-05, | |
| "loss": 0.8697, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "learning_rate": 1.1036505867014341e-05, | |
| "loss": 0.8856, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "learning_rate": 1.0764884832681443e-05, | |
| "loss": 0.8774, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "learning_rate": 1.0493263798348545e-05, | |
| "loss": 0.8622, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "learning_rate": 1.0221642764015647e-05, | |
| "loss": 0.8697, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "learning_rate": 9.950021729682747e-06, | |
| "loss": 0.8712, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "learning_rate": 9.678400695349848e-06, | |
| "loss": 0.8648, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "learning_rate": 9.40677966101695e-06, | |
| "loss": 0.8784, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "learning_rate": 9.13515862668405e-06, | |
| "loss": 0.8646, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "learning_rate": 8.863537592351152e-06, | |
| "loss": 0.869, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "learning_rate": 8.591916558018254e-06, | |
| "loss": 0.857, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "learning_rate": 8.320295523685355e-06, | |
| "loss": 0.8614, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "learning_rate": 8.048674489352455e-06, | |
| "loss": 0.8684, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "learning_rate": 7.777053455019557e-06, | |
| "loss": 0.8772, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "learning_rate": 7.505432420686659e-06, | |
| "loss": 0.8688, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "learning_rate": 7.23381138635376e-06, | |
| "loss": 0.8593, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "learning_rate": 6.962190352020861e-06, | |
| "loss": 0.8482, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 6.690569317687963e-06, | |
| "loss": 0.8727, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 72.57259296994395, | |
| "eval_combined_score": 72.57259296994395, | |
| "eval_loss": 0.6770668029785156, | |
| "eval_runtime": 21.9469, | |
| "eval_samples_per_second": 447.216, | |
| "eval_steps_per_second": 1.777, | |
| "step": 24544 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "learning_rate": 6.418948283355063e-06, | |
| "loss": 0.8268, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "learning_rate": 6.147327249022165e-06, | |
| "loss": 0.8269, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "learning_rate": 5.8757062146892665e-06, | |
| "loss": 0.8091, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "learning_rate": 5.6040851803563665e-06, | |
| "loss": 0.8131, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "learning_rate": 5.332464146023468e-06, | |
| "loss": 0.8151, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "learning_rate": 5.06084311169057e-06, | |
| "loss": 0.8163, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "learning_rate": 4.789222077357671e-06, | |
| "loss": 0.8122, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "learning_rate": 4.5176010430247726e-06, | |
| "loss": 0.808, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "learning_rate": 4.2459800086918734e-06, | |
| "loss": 0.8017, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "learning_rate": 3.974358974358974e-06, | |
| "loss": 0.795, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "learning_rate": 3.702737940026076e-06, | |
| "loss": 0.8065, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "learning_rate": 3.431116905693177e-06, | |
| "loss": 0.8099, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "learning_rate": 3.1594958713602786e-06, | |
| "loss": 0.8143, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "learning_rate": 2.8878748370273795e-06, | |
| "loss": 0.8089, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "learning_rate": 2.6162538026944812e-06, | |
| "loss": 0.8189, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "learning_rate": 2.344632768361582e-06, | |
| "loss": 0.8137, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "learning_rate": 2.0730117340286834e-06, | |
| "loss": 0.804, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "learning_rate": 1.8013906996957845e-06, | |
| "loss": 0.8167, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "learning_rate": 1.5297696653628858e-06, | |
| "loss": 0.8123, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "learning_rate": 1.258148631029987e-06, | |
| "loss": 0.8131, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "learning_rate": 9.865275966970884e-07, | |
| "loss": 0.8024, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "learning_rate": 7.149065623641896e-07, | |
| "loss": 0.8063, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "learning_rate": 4.432855280312908e-07, | |
| "loss": 0.8108, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "learning_rate": 1.7166449369839202e-07, | |
| "loss": 0.8284, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 73.63219561895059, | |
| "eval_combined_score": 73.63219561895059, | |
| "eval_loss": 0.6550651788711548, | |
| "eval_runtime": 23.928, | |
| "eval_samples_per_second": 410.19, | |
| "eval_steps_per_second": 1.63, | |
| "step": 36816 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 36816, | |
| "total_flos": 0.0, | |
| "train_loss": 0.8892150490970935, | |
| "train_runtime": 6789.2004, | |
| "train_samples_per_second": 173.526, | |
| "train_steps_per_second": 5.423 | |
| } | |
| ], | |
| "max_steps": 36816, | |
| "num_train_epochs": 3, | |
| "total_flos": 0.0, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |