| { | |
| "best_metric": 0.31482580304145813, | |
| "best_model_checkpoint": "bert_uncased_L-4_H-512_A-8_qnli/checkpoint-1230", | |
| "epoch": 8.0, | |
| "eval_steps": 500, | |
| "global_step": 3280, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 5.487641334533691, | |
| "learning_rate": 4.9e-05, | |
| "loss": 0.437, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8520959179937763, | |
| "eval_loss": 0.34476375579833984, | |
| "eval_runtime": 2.1739, | |
| "eval_samples_per_second": 2513.023, | |
| "eval_steps_per_second": 10.12, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 5.546387195587158, | |
| "learning_rate": 4.8e-05, | |
| "loss": 0.3392, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8606992494966136, | |
| "eval_loss": 0.3215360939502716, | |
| "eval_runtime": 2.1782, | |
| "eval_samples_per_second": 2508.088, | |
| "eval_steps_per_second": 10.1, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 4.7529730796813965, | |
| "learning_rate": 4.7e-05, | |
| "loss": 0.2746, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8735127219476478, | |
| "eval_loss": 0.31482580304145813, | |
| "eval_runtime": 2.2051, | |
| "eval_samples_per_second": 2477.472, | |
| "eval_steps_per_second": 9.977, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 4.871633529663086, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 0.2175, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8702178290316676, | |
| "eval_loss": 0.3549071252346039, | |
| "eval_runtime": 2.1724, | |
| "eval_samples_per_second": 2514.741, | |
| "eval_steps_per_second": 10.127, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 7.13149881362915, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.1712, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8579535053999634, | |
| "eval_loss": 0.39998722076416016, | |
| "eval_runtime": 2.1979, | |
| "eval_samples_per_second": 2485.602, | |
| "eval_steps_per_second": 10.01, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 9.117788314819336, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 0.1311, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8649093904448105, | |
| "eval_loss": 0.43354788422584534, | |
| "eval_runtime": 2.1658, | |
| "eval_samples_per_second": 2522.423, | |
| "eval_steps_per_second": 10.158, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 4.943286418914795, | |
| "learning_rate": 4.3e-05, | |
| "loss": 0.1065, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.8641771920190372, | |
| "eval_loss": 0.4818989932537079, | |
| "eval_runtime": 2.1604, | |
| "eval_samples_per_second": 2528.736, | |
| "eval_steps_per_second": 10.183, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 4.7664289474487305, | |
| "learning_rate": 4.2e-05, | |
| "loss": 0.0849, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.866739886509244, | |
| "eval_loss": 0.5126516222953796, | |
| "eval_runtime": 2.2037, | |
| "eval_samples_per_second": 2478.978, | |
| "eval_steps_per_second": 9.983, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "step": 3280, | |
| "total_flos": 1.6570204935856128e+16, | |
| "train_loss": 0.22026773196894947, | |
| "train_runtime": 535.1431, | |
| "train_samples_per_second": 9786.448, | |
| "train_steps_per_second": 38.308 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 20500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 5 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.6570204935856128e+16, | |
| "train_batch_size": 256, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |