| { |
| "best_global_step": 1230, |
| "best_metric": 0.6439675688743591, |
| "best_model_checkpoint": "distilbert_km_50_v2_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": 1.5073614120483398, |
| "learning_rate": 4.9e-05, |
| "loss": 0.6688, |
| "step": 410 |
| }, |
| { |
| "epoch": 1.0, |
| "eval_accuracy": 0.6080907926047959, |
| "eval_loss": 0.6536266207695007, |
| "eval_runtime": 2.301, |
| "eval_samples_per_second": 2374.143, |
| "eval_steps_per_second": 9.561, |
| "step": 410 |
| }, |
| { |
| "epoch": 2.0, |
| "grad_norm": 1.545426607131958, |
| "learning_rate": 4.8e-05, |
| "loss": 0.6354, |
| "step": 820 |
| }, |
| { |
| "epoch": 2.0, |
| "eval_accuracy": 0.615046677649643, |
| "eval_loss": 0.6506986618041992, |
| "eval_runtime": 2.267, |
| "eval_samples_per_second": 2409.836, |
| "eval_steps_per_second": 9.705, |
| "step": 820 |
| }, |
| { |
| "epoch": 3.0, |
| "grad_norm": 2.5010275840759277, |
| "learning_rate": 4.7e-05, |
| "loss": 0.5863, |
| "step": 1230 |
| }, |
| { |
| "epoch": 3.0, |
| "eval_accuracy": 0.6326194398682042, |
| "eval_loss": 0.6439675688743591, |
| "eval_runtime": 2.3046, |
| "eval_samples_per_second": 2370.441, |
| "eval_steps_per_second": 9.546, |
| "step": 1230 |
| }, |
| { |
| "epoch": 4.0, |
| "grad_norm": 3.375784158706665, |
| "learning_rate": 4.600000000000001e-05, |
| "loss": 0.5027, |
| "step": 1640 |
| }, |
| { |
| "epoch": 4.0, |
| "eval_accuracy": 0.6231008603331503, |
| "eval_loss": 0.7137660384178162, |
| "eval_runtime": 2.2641, |
| "eval_samples_per_second": 2412.889, |
| "eval_steps_per_second": 9.717, |
| "step": 1640 |
| }, |
| { |
| "epoch": 5.0, |
| "grad_norm": 7.077524662017822, |
| "learning_rate": 4.5e-05, |
| "loss": 0.4002, |
| "step": 2050 |
| }, |
| { |
| "epoch": 5.0, |
| "eval_accuracy": 0.6209042650558302, |
| "eval_loss": 0.8596444725990295, |
| "eval_runtime": 2.2596, |
| "eval_samples_per_second": 2417.631, |
| "eval_steps_per_second": 9.736, |
| "step": 2050 |
| }, |
| { |
| "epoch": 6.0, |
| "grad_norm": 5.8376946449279785, |
| "learning_rate": 4.4000000000000006e-05, |
| "loss": 0.3023, |
| "step": 2460 |
| }, |
| { |
| "epoch": 6.0, |
| "eval_accuracy": 0.6161449752883031, |
| "eval_loss": 1.0441951751708984, |
| "eval_runtime": 2.2665, |
| "eval_samples_per_second": 2410.341, |
| "eval_steps_per_second": 9.707, |
| "step": 2460 |
| }, |
| { |
| "epoch": 7.0, |
| "grad_norm": 7.014330863952637, |
| "learning_rate": 4.3e-05, |
| "loss": 0.2266, |
| "step": 2870 |
| }, |
| { |
| "epoch": 7.0, |
| "eval_accuracy": 0.6110195863078894, |
| "eval_loss": 1.2264615297317505, |
| "eval_runtime": 2.2732, |
| "eval_samples_per_second": 2403.218, |
| "eval_steps_per_second": 9.678, |
| "step": 2870 |
| }, |
| { |
| "epoch": 8.0, |
| "grad_norm": 4.5718770027160645, |
| "learning_rate": 4.2e-05, |
| "loss": 0.1741, |
| "step": 3280 |
| }, |
| { |
| "epoch": 8.0, |
| "eval_accuracy": 0.6121178839465495, |
| "eval_loss": 1.2523001432418823, |
| "eval_runtime": 2.2906, |
| "eval_samples_per_second": 2384.982, |
| "eval_steps_per_second": 9.605, |
| "step": 3280 |
| }, |
| { |
| "epoch": 8.0, |
| "step": 3280, |
| "total_flos": 5.550013094970163e+16, |
| "train_loss": 0.4370324088305962, |
| "train_runtime": 801.8009, |
| "train_samples_per_second": 6531.734, |
| "train_steps_per_second": 25.567 |
| } |
| ], |
| "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": 5.550013094970163e+16, |
| "train_batch_size": 256, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
