| { | |
| "best_metric": 0.8179571663920923, | |
| "best_model_checkpoint": "./bigbird-base-health-fact/checkpoint-2452", | |
| "epoch": 3.0, | |
| "global_step": 3678, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 2.6630434782608698e-06, | |
| "loss": 1.2792, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 5.380434782608695e-06, | |
| "loss": 1.0546, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 8.097826086956523e-06, | |
| "loss": 0.8857, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 9.909365558912388e-06, | |
| "loss": 0.7958, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 9.60725075528701e-06, | |
| "loss": 0.7246, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 9.305135951661632e-06, | |
| "loss": 0.7629, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 9.003021148036256e-06, | |
| "loss": 0.6498, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 8.700906344410877e-06, | |
| "loss": 0.6201, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 8.398791540785499e-06, | |
| "loss": 0.5809, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 8.099697885196374e-06, | |
| "loss": 0.647, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 7.797583081570997e-06, | |
| "loss": 0.5814, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 7.4954682779456205e-06, | |
| "loss": 0.5563, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_false_f1": 0.7926023778071335, | |
| "eval_loss": 0.5019509196281433, | |
| "eval_macro_f1": 0.6062122975261963, | |
| "eval_micro_f1": 0.7948929159802306, | |
| "eval_mixture_f1": 0.4591194968553459, | |
| "eval_runtime": 39.3928, | |
| "eval_samples_per_second": 30.818, | |
| "eval_steps_per_second": 0.965, | |
| "eval_true_f1": 0.8986175115207373, | |
| "eval_unproven_f1": 0.2745098039215686, | |
| "step": 1226 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "learning_rate": 7.193353474320243e-06, | |
| "loss": 0.5271, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "learning_rate": 6.891238670694864e-06, | |
| "loss": 0.4912, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 6.589123867069487e-06, | |
| "loss": 0.4844, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "learning_rate": 6.287009063444109e-06, | |
| "loss": 0.4816, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "learning_rate": 5.984894259818732e-06, | |
| "loss": 0.462, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "learning_rate": 5.682779456193354e-06, | |
| "loss": 0.4087, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "learning_rate": 5.380664652567976e-06, | |
| "loss": 0.5065, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "learning_rate": 5.078549848942599e-06, | |
| "loss": 0.4313, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "learning_rate": 4.776435045317221e-06, | |
| "loss": 0.5098, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "learning_rate": 4.4743202416918435e-06, | |
| "loss": 0.4699, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "learning_rate": 4.172205438066466e-06, | |
| "loss": 0.4408, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "learning_rate": 3.8700906344410875e-06, | |
| "loss": 0.5048, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_false_f1": 0.8201811125485123, | |
| "eval_loss": 0.4968700110912323, | |
| "eval_macro_f1": 0.684587518040316, | |
| "eval_micro_f1": 0.8179571663920923, | |
| "eval_mixture_f1": 0.43416370106761565, | |
| "eval_runtime": 39.4059, | |
| "eval_samples_per_second": 30.808, | |
| "eval_steps_per_second": 0.964, | |
| "eval_true_f1": 0.9125766871165646, | |
| "eval_unproven_f1": 0.5714285714285714, | |
| "step": 2452 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "learning_rate": 3.5679758308157103e-06, | |
| "loss": 0.378, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "learning_rate": 3.2658610271903322e-06, | |
| "loss": 0.3631, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "learning_rate": 2.963746223564955e-06, | |
| "loss": 0.3475, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "learning_rate": 2.661631419939577e-06, | |
| "loss": 0.3283, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "learning_rate": 2.3595166163142e-06, | |
| "loss": 0.317, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "learning_rate": 2.0604229607250755e-06, | |
| "loss": 0.3541, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "learning_rate": 1.758308157099698e-06, | |
| "loss": 0.3818, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "learning_rate": 1.4561933534743203e-06, | |
| "loss": 0.3467, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "learning_rate": 1.1540785498489427e-06, | |
| "loss": 0.3048, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "learning_rate": 8.51963746223565e-07, | |
| "loss": 0.3334, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "learning_rate": 5.498489425981874e-07, | |
| "loss": 0.3855, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "learning_rate": 2.477341389728097e-07, | |
| "loss": 0.3454, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_false_f1": 0.811443433029909, | |
| "eval_loss": 0.5863622426986694, | |
| "eval_macro_f1": 0.6874160790583576, | |
| "eval_micro_f1": 0.8130148270181219, | |
| "eval_mixture_f1": 0.4556962025316456, | |
| "eval_runtime": 39.6082, | |
| "eval_samples_per_second": 30.65, | |
| "eval_steps_per_second": 0.959, | |
| "eval_true_f1": 0.9153605015673981, | |
| "eval_unproven_f1": 0.5671641791044776, | |
| "step": 3678 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 3678, | |
| "total_flos": 2.106512041918464e+16, | |
| "train_loss": 0.5249485609128204, | |
| "train_runtime": 3291.1672, | |
| "train_samples_per_second": 8.937, | |
| "train_steps_per_second": 1.118 | |
| } | |
| ], | |
| "max_steps": 3678, | |
| "num_train_epochs": 3, | |
| "total_flos": 2.106512041918464e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |