| { | |
| "best_metric": 0.812191103789127, | |
| "best_model_checkpoint": "./longformer-base-health-fact2/checkpoint-1839", | |
| "epoch": 3.0, | |
| "global_step": 1839, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 1.597826086956522e-05, | |
| "loss": 1.1814, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 2.9764350453172204e-05, | |
| "loss": 0.7803, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 2.795166163141994e-05, | |
| "loss": 0.695, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 2.61570996978852e-05, | |
| "loss": 0.6093, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 2.434441087613293e-05, | |
| "loss": 0.5988, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 2.2531722054380667e-05, | |
| "loss": 0.555, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_false_f1": 0.7698209718670076, | |
| "eval_loss": 0.5242577791213989, | |
| "eval_macro_f1": 0.5534734720356734, | |
| "eval_micro_f1": 0.7841845140032949, | |
| "eval_mixture_f1": 0.4169611307420495, | |
| "eval_runtime": 35.7477, | |
| "eval_samples_per_second": 33.96, | |
| "eval_steps_per_second": 1.063, | |
| "eval_true_f1": 0.8937784522003035, | |
| "eval_unproven_f1": 0.13333333333333333, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "learning_rate": 2.07190332326284e-05, | |
| "loss": 0.5169, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "learning_rate": 1.8906344410876132e-05, | |
| "loss": 0.4659, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "learning_rate": 1.709365558912387e-05, | |
| "loss": 0.4781, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "learning_rate": 1.52809667673716e-05, | |
| "loss": 0.4591, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "learning_rate": 1.3468277945619335e-05, | |
| "loss": 0.4591, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "learning_rate": 1.165558912386707e-05, | |
| "loss": 0.4282, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_false_f1": 0.7829360100376411, | |
| "eval_loss": 0.5008112788200378, | |
| "eval_macro_f1": 0.6393266392992799, | |
| "eval_micro_f1": 0.8031301482701811, | |
| "eval_mixture_f1": 0.46052631578947373, | |
| "eval_runtime": 35.7682, | |
| "eval_samples_per_second": 33.941, | |
| "eval_steps_per_second": 1.062, | |
| "eval_true_f1": 0.9199048374306107, | |
| "eval_unproven_f1": 0.3939393939393939, | |
| "step": 1226 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "learning_rate": 9.842900302114804e-06, | |
| "loss": 0.3421, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "learning_rate": 8.030211480362539e-06, | |
| "loss": 0.3261, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "learning_rate": 6.217522658610272e-06, | |
| "loss": 0.3269, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "learning_rate": 4.404833836858006e-06, | |
| "loss": 0.3102, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "learning_rate": 2.5921450151057403e-06, | |
| "loss": 0.2872, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "learning_rate": 7.794561933534744e-07, | |
| "loss": 0.2897, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_false_f1": 0.7941176470588236, | |
| "eval_loss": 0.5857986211776733, | |
| "eval_macro_f1": 0.6829679742192657, | |
| "eval_micro_f1": 0.812191103789127, | |
| "eval_mixture_f1": 0.5014925373134329, | |
| "eval_runtime": 35.8572, | |
| "eval_samples_per_second": 33.857, | |
| "eval_steps_per_second": 1.06, | |
| "eval_true_f1": 0.9234411996842935, | |
| "eval_unproven_f1": 0.5128205128205128, | |
| "step": 1839 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 1839, | |
| "total_flos": 2.18669047246848e+16, | |
| "train_loss": 0.49966207906692944, | |
| "train_runtime": 2323.0594, | |
| "train_samples_per_second": 12.661, | |
| "train_steps_per_second": 0.792 | |
| } | |
| ], | |
| "max_steps": 1839, | |
| "num_train_epochs": 3, | |
| "total_flos": 2.18669047246848e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |