| { | |
| "best_global_step": 1230, | |
| "best_metric": 0.44436830282211304, | |
| "best_model_checkpoint": "tinybert_base_train_book_ent_15p_s_init_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": 4.298269271850586, | |
| "learning_rate": 4.900243902439024e-05, | |
| "loss": 0.6384, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6985172981878089, | |
| "eval_loss": 0.5731965899467468, | |
| "eval_runtime": 2.9466, | |
| "eval_samples_per_second": 1853.975, | |
| "eval_steps_per_second": 7.466, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 4.312274932861328, | |
| "learning_rate": 4.8002439024390246e-05, | |
| "loss": 0.5415, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7642321069009702, | |
| "eval_loss": 0.4735166132450104, | |
| "eval_runtime": 2.8912, | |
| "eval_samples_per_second": 1889.532, | |
| "eval_steps_per_second": 7.609, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 6.964209079742432, | |
| "learning_rate": 4.700243902439024e-05, | |
| "loss": 0.4624, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7882116053450485, | |
| "eval_loss": 0.44436830282211304, | |
| "eval_runtime": 2.9004, | |
| "eval_samples_per_second": 1883.526, | |
| "eval_steps_per_second": 7.585, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 7.770699977874756, | |
| "learning_rate": 4.600243902439025e-05, | |
| "loss": 0.4039, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7964488376349991, | |
| "eval_loss": 0.4634305238723755, | |
| "eval_runtime": 3.0779, | |
| "eval_samples_per_second": 1774.904, | |
| "eval_steps_per_second": 7.148, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 5.855748653411865, | |
| "learning_rate": 4.500243902439025e-05, | |
| "loss": 0.3496, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.7951674903898956, | |
| "eval_loss": 0.49477097392082214, | |
| "eval_runtime": 2.9325, | |
| "eval_samples_per_second": 1862.892, | |
| "eval_steps_per_second": 7.502, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 8.18688678741455, | |
| "learning_rate": 4.400243902439024e-05, | |
| "loss": 0.2985, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.7946183415705657, | |
| "eval_loss": 0.5067967176437378, | |
| "eval_runtime": 2.8942, | |
| "eval_samples_per_second": 1887.586, | |
| "eval_steps_per_second": 7.601, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 7.592705249786377, | |
| "learning_rate": 4.3002439024390246e-05, | |
| "loss": 0.2571, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.7838184147904081, | |
| "eval_loss": 0.5432331562042236, | |
| "eval_runtime": 2.9329, | |
| "eval_samples_per_second": 1862.65, | |
| "eval_steps_per_second": 7.501, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 4.670092582702637, | |
| "learning_rate": 4.200243902439024e-05, | |
| "loss": 0.2182, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.7872963573128318, | |
| "eval_loss": 0.6078780889511108, | |
| "eval_runtime": 2.9158, | |
| "eval_samples_per_second": 1873.589, | |
| "eval_steps_per_second": 7.545, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "step": 3280, | |
| "total_flos": 1.6570204935856128e+16, | |
| "train_loss": 0.39620263867261934, | |
| "train_runtime": 645.0572, | |
| "train_samples_per_second": 8118.893, | |
| "train_steps_per_second": 31.78 | |
| } | |
| ], | |
| "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 | |
| } | |