| { | |
| "best_global_step": 7670, | |
| "best_metric": 0.6811214685440063, | |
| "best_model_checkpoint": "tinybert_base_train_book_ent_15p_s_init_mnli/checkpoint-7670", | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 15340, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.6886646747589111, | |
| "learning_rate": 4.90006518904824e-05, | |
| "loss": 0.8785, | |
| "step": 1534 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6660213958227204, | |
| "eval_loss": 0.7728874087333679, | |
| "eval_runtime": 5.2916, | |
| "eval_samples_per_second": 1854.821, | |
| "eval_steps_per_second": 7.37, | |
| "step": 1534 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.5701472759246826, | |
| "learning_rate": 4.80006518904824e-05, | |
| "loss": 0.7353, | |
| "step": 3068 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.691492613346918, | |
| "eval_loss": 0.7123497724533081, | |
| "eval_runtime": 5.2657, | |
| "eval_samples_per_second": 1863.958, | |
| "eval_steps_per_second": 7.406, | |
| "step": 3068 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.9975154399871826, | |
| "learning_rate": 4.70006518904824e-05, | |
| "loss": 0.6658, | |
| "step": 4602 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7072847682119205, | |
| "eval_loss": 0.6983441710472107, | |
| "eval_runtime": 5.2335, | |
| "eval_samples_per_second": 1875.435, | |
| "eval_steps_per_second": 7.452, | |
| "step": 4602 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 2.1961188316345215, | |
| "learning_rate": 4.60006518904824e-05, | |
| "loss": 0.6113, | |
| "step": 6136 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7168619460010188, | |
| "eval_loss": 0.700081467628479, | |
| "eval_runtime": 5.2018, | |
| "eval_samples_per_second": 1886.83, | |
| "eval_steps_per_second": 7.497, | |
| "step": 6136 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 2.4144446849823, | |
| "learning_rate": 4.50006518904824e-05, | |
| "loss": 0.5654, | |
| "step": 7670 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.7245033112582782, | |
| "eval_loss": 0.6811214685440063, | |
| "eval_runtime": 5.2441, | |
| "eval_samples_per_second": 1871.624, | |
| "eval_steps_per_second": 7.437, | |
| "step": 7670 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 2.4327688217163086, | |
| "learning_rate": 4.40006518904824e-05, | |
| "loss": 0.5207, | |
| "step": 9204 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.7257259296994396, | |
| "eval_loss": 0.7057485580444336, | |
| "eval_runtime": 5.236, | |
| "eval_samples_per_second": 1874.515, | |
| "eval_steps_per_second": 7.448, | |
| "step": 9204 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 2.349059581756592, | |
| "learning_rate": 4.30006518904824e-05, | |
| "loss": 0.4798, | |
| "step": 10738 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.7290881304126338, | |
| "eval_loss": 0.7188459038734436, | |
| "eval_runtime": 5.2738, | |
| "eval_samples_per_second": 1861.082, | |
| "eval_steps_per_second": 7.395, | |
| "step": 10738 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 3.1730854511260986, | |
| "learning_rate": 4.20006518904824e-05, | |
| "loss": 0.4403, | |
| "step": 12272 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.7230769230769231, | |
| "eval_loss": 0.7684288024902344, | |
| "eval_runtime": 5.2321, | |
| "eval_samples_per_second": 1875.918, | |
| "eval_steps_per_second": 7.454, | |
| "step": 12272 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 3.504354953765869, | |
| "learning_rate": 4.10006518904824e-05, | |
| "loss": 0.4036, | |
| "step": 13806 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.7164544065206316, | |
| "eval_loss": 0.8033895492553711, | |
| "eval_runtime": 5.2638, | |
| "eval_samples_per_second": 1864.638, | |
| "eval_steps_per_second": 7.409, | |
| "step": 13806 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 3.7538444995880127, | |
| "learning_rate": 4.00006518904824e-05, | |
| "loss": 0.3685, | |
| "step": 15340 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.7219561895058584, | |
| "eval_loss": 0.837572455406189, | |
| "eval_runtime": 5.2382, | |
| "eval_samples_per_second": 1873.732, | |
| "eval_steps_per_second": 7.445, | |
| "step": 15340 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 15340, | |
| "total_flos": 7.76592697038336e+16, | |
| "train_loss": 0.5669185474301255, | |
| "train_runtime": 2952.0721, | |
| "train_samples_per_second": 6651.294, | |
| "train_steps_per_second": 25.982 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 76700, | |
| "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": 7.76592697038336e+16, | |
| "train_batch_size": 256, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |