| { | |
| "best_metric": 0.379541277885437, | |
| "best_model_checkpoint": "bert_tiny_lda_5_v1_book_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": 5.600667476654053, | |
| "learning_rate": 4.9e-05, | |
| "loss": 0.5111, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8081640124473732, | |
| "eval_loss": 0.4211691617965698, | |
| "eval_runtime": 1.683, | |
| "eval_samples_per_second": 3246.072, | |
| "eval_steps_per_second": 13.072, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 4.195663928985596, | |
| "learning_rate": 4.8e-05, | |
| "loss": 0.4041, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8251876258466044, | |
| "eval_loss": 0.38909679651260376, | |
| "eval_runtime": 1.7412, | |
| "eval_samples_per_second": 3137.417, | |
| "eval_steps_per_second": 12.635, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 5.590042591094971, | |
| "learning_rate": 4.7e-05, | |
| "loss": 0.3396, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.832875709317225, | |
| "eval_loss": 0.379541277885437, | |
| "eval_runtime": 1.7411, | |
| "eval_samples_per_second": 3137.606, | |
| "eval_steps_per_second": 12.635, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 5.420849800109863, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 0.2836, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8228079809628409, | |
| "eval_loss": 0.44419917464256287, | |
| "eval_runtime": 1.7526, | |
| "eval_samples_per_second": 3117.129, | |
| "eval_steps_per_second": 12.553, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 5.306129455566406, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.2279, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8154859967051071, | |
| "eval_loss": 0.48089849948883057, | |
| "eval_runtime": 1.7368, | |
| "eval_samples_per_second": 3145.495, | |
| "eval_steps_per_second": 12.667, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 5.556026935577393, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 0.1855, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8262859234852645, | |
| "eval_loss": 0.5276368856430054, | |
| "eval_runtime": 1.7009, | |
| "eval_samples_per_second": 3211.749, | |
| "eval_steps_per_second": 12.934, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 6.040602684020996, | |
| "learning_rate": 4.3e-05, | |
| "loss": 0.1468, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.823357129782171, | |
| "eval_loss": 0.5409088134765625, | |
| "eval_runtime": 1.7956, | |
| "eval_samples_per_second": 3042.377, | |
| "eval_steps_per_second": 12.252, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 6.397125244140625, | |
| "learning_rate": 4.2e-05, | |
| "loss": 0.1206, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8171334431630972, | |
| "eval_loss": 0.6155871748924255, | |
| "eval_runtime": 1.7094, | |
| "eval_samples_per_second": 3195.79, | |
| "eval_steps_per_second": 12.87, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "step": 3280, | |
| "total_flos": 2.197388993748173e+16, | |
| "train_loss": 0.27741835291792705, | |
| "train_runtime": 483.5483, | |
| "train_samples_per_second": 10830.666, | |
| "train_steps_per_second": 42.395 | |
| } | |
| ], | |
| "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": 2.197388993748173e+16, | |
| "train_batch_size": 256, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |