| { | |
| "best_metric": 0.5934821963310242, | |
| "best_model_checkpoint": "bert_tiny_lda_50_v1_book_cola/checkpoint-102", | |
| "epoch": 8.0, | |
| "eval_steps": 500, | |
| "global_step": 272, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.7861921191215515, | |
| "learning_rate": 4.9e-05, | |
| "loss": 0.6133, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6912751793861389, | |
| "eval_loss": 0.6133261919021606, | |
| "eval_matthews_correlation": 0.0, | |
| "eval_runtime": 0.3311, | |
| "eval_samples_per_second": 3150.274, | |
| "eval_steps_per_second": 15.102, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.130760908126831, | |
| "learning_rate": 4.8e-05, | |
| "loss": 0.5995, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.6893576383590698, | |
| "eval_loss": 0.6048817038536072, | |
| "eval_matthews_correlation": -0.02929206145132745, | |
| "eval_runtime": 0.3328, | |
| "eval_samples_per_second": 3134.435, | |
| "eval_steps_per_second": 15.026, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.8091745376586914, | |
| "learning_rate": 4.7e-05, | |
| "loss": 0.57, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.6989453434944153, | |
| "eval_loss": 0.5934821963310242, | |
| "eval_matthews_correlation": 0.1366972617887581, | |
| "eval_runtime": 0.332, | |
| "eval_samples_per_second": 3141.232, | |
| "eval_steps_per_second": 15.059, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 2.9426748752593994, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 0.5209, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.6845637559890747, | |
| "eval_loss": 0.6036127805709839, | |
| "eval_matthews_correlation": 0.1838806266083814, | |
| "eval_runtime": 0.3338, | |
| "eval_samples_per_second": 3124.801, | |
| "eval_steps_per_second": 14.98, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 3.014277935028076, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.466, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.6826462149620056, | |
| "eval_loss": 0.6264468431472778, | |
| "eval_matthews_correlation": 0.20738864502723908, | |
| "eval_runtime": 0.3336, | |
| "eval_samples_per_second": 3126.624, | |
| "eval_steps_per_second": 14.989, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 3.3172547817230225, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 0.3922, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.6979866027832031, | |
| "eval_loss": 0.712081789970398, | |
| "eval_matthews_correlation": 0.1825111909284102, | |
| "eval_runtime": 0.355, | |
| "eval_samples_per_second": 2938.433, | |
| "eval_steps_per_second": 14.086, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 5.064127445220947, | |
| "learning_rate": 4.3e-05, | |
| "loss": 0.3492, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.6989453434944153, | |
| "eval_loss": 0.7019482851028442, | |
| "eval_matthews_correlation": 0.22042544777190046, | |
| "eval_runtime": 0.3311, | |
| "eval_samples_per_second": 3149.922, | |
| "eval_steps_per_second": 15.1, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 7.477691650390625, | |
| "learning_rate": 4.2e-05, | |
| "loss": 0.307, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.6999041438102722, | |
| "eval_loss": 0.7502853274345398, | |
| "eval_matthews_correlation": 0.2271320033748531, | |
| "eval_runtime": 0.3593, | |
| "eval_samples_per_second": 2902.519, | |
| "eval_steps_per_second": 13.914, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "step": 272, | |
| "total_flos": 1793902531485696.0, | |
| "train_loss": 0.4772484197336085, | |
| "train_runtime": 50.017, | |
| "train_samples_per_second": 8548.097, | |
| "train_steps_per_second": 33.988 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1700, | |
| "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": 1793902531485696.0, | |
| "train_batch_size": 256, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |