| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.8334419817470664, | |
| "global_step": 22500, | |
| "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": 0.755, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 1.9456757931334205e-05, | |
| "loss": 0.5771, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 1.9185136897001307e-05, | |
| "loss": 0.5303, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 1.8913515862668405e-05, | |
| "loss": 0.5153, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 1.8641894828335507e-05, | |
| "loss": 0.5023, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "eval_accuracy": 0.8247580234335201, | |
| "eval_loss": 0.4500664174556732, | |
| "eval_runtime": 38.363, | |
| "eval_samples_per_second": 255.846, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 1.837027379400261e-05, | |
| "loss": 0.4797, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 1.809865275966971e-05, | |
| "loss": 0.4613, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 1.7827031725336812e-05, | |
| "loss": 0.4665, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 1.7555410691003914e-05, | |
| "loss": 0.4516, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 1.7283789656671015e-05, | |
| "loss": 0.4485, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "eval_accuracy": 0.8420784513499745, | |
| "eval_loss": 0.40373361110687256, | |
| "eval_runtime": 38.331, | |
| "eval_samples_per_second": 256.059, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 1.7012168622338114e-05, | |
| "loss": 0.4394, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 1.6740547588005215e-05, | |
| "loss": 0.4262, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 1.6468926553672317e-05, | |
| "loss": 0.4374, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 1.619730551933942e-05, | |
| "loss": 0.4239, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 1.592568448500652e-05, | |
| "loss": 0.4311, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "eval_accuracy": 0.845746306673459, | |
| "eval_loss": 0.43105846643447876, | |
| "eval_runtime": 38.3496, | |
| "eval_samples_per_second": 255.935, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 1.5654063450673622e-05, | |
| "loss": 0.41, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 1.5382442416340724e-05, | |
| "loss": 0.4248, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 1.5110821382007822e-05, | |
| "loss": 0.4041, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 1.4839200347674924e-05, | |
| "loss": 0.4047, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 1.4567579313342026e-05, | |
| "loss": 0.4062, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "eval_accuracy": 0.8529801324503311, | |
| "eval_loss": 0.38434121012687683, | |
| "eval_runtime": 38.3942, | |
| "eval_samples_per_second": 255.638, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 1.4295958279009128e-05, | |
| "loss": 0.4095, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.402433724467623e-05, | |
| "loss": 0.3963, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 1.3752716210343331e-05, | |
| "loss": 0.3997, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 1.3481095176010431e-05, | |
| "loss": 0.4054, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "learning_rate": 1.3209474141677531e-05, | |
| "loss": 0.3341, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "eval_accuracy": 0.8573611818644932, | |
| "eval_loss": 0.4230915606021881, | |
| "eval_runtime": 38.3424, | |
| "eval_samples_per_second": 255.983, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "learning_rate": 1.2937853107344633e-05, | |
| "loss": 0.2561, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "learning_rate": 1.2666232073011735e-05, | |
| "loss": 0.2567, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "learning_rate": 1.2394611038678836e-05, | |
| "loss": 0.2583, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "learning_rate": 1.2122990004345938e-05, | |
| "loss": 0.2511, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 1.185136897001304e-05, | |
| "loss": 0.2584, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "eval_accuracy": 0.8566479877738156, | |
| "eval_loss": 0.4263673424720764, | |
| "eval_runtime": 38.3835, | |
| "eval_samples_per_second": 255.709, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "learning_rate": 1.157974793568014e-05, | |
| "loss": 0.2588, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "learning_rate": 1.130812690134724e-05, | |
| "loss": 0.2712, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "learning_rate": 1.1036505867014341e-05, | |
| "loss": 0.2543, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "learning_rate": 1.0764884832681443e-05, | |
| "loss": 0.2602, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "learning_rate": 1.0493263798348545e-05, | |
| "loss": 0.2615, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "eval_accuracy": 0.85206316861946, | |
| "eval_loss": 0.43541985750198364, | |
| "eval_runtime": 38.3513, | |
| "eval_samples_per_second": 255.924, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "learning_rate": 1.0221642764015647e-05, | |
| "loss": 0.2601, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "learning_rate": 9.950021729682747e-06, | |
| "loss": 0.2624, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "learning_rate": 9.678400695349848e-06, | |
| "loss": 0.2611, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "learning_rate": 9.40677966101695e-06, | |
| "loss": 0.264, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "learning_rate": 9.13515862668405e-06, | |
| "loss": 0.2589, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "eval_accuracy": 0.8614365766683647, | |
| "eval_loss": 0.40027427673339844, | |
| "eval_runtime": 38.3743, | |
| "eval_samples_per_second": 255.77, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "learning_rate": 8.863537592351152e-06, | |
| "loss": 0.2596, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "learning_rate": 8.591916558018254e-06, | |
| "loss": 0.2558, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "learning_rate": 8.320295523685355e-06, | |
| "loss": 0.2599, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "learning_rate": 8.048674489352455e-06, | |
| "loss": 0.2562, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "learning_rate": 7.777053455019557e-06, | |
| "loss": 0.2641, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "eval_accuracy": 0.8667345899133979, | |
| "eval_loss": 0.3984006643295288, | |
| "eval_runtime": 38.3342, | |
| "eval_samples_per_second": 256.038, | |
| "step": 22500 | |
| } | |
| ], | |
| "max_steps": 36816, | |
| "num_train_epochs": 3, | |
| "total_flos": 1.8532124395781683e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |