| { | |
| "best_metric": 0.41116195917129517, | |
| "best_model_checkpoint": "bert_tiny_lda_20_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": 4.762260437011719, | |
| "learning_rate": 4.9e-05, | |
| "loss": 0.5552, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7704557935200439, | |
| "eval_loss": 0.5005637407302856, | |
| "eval_runtime": 1.7019, | |
| "eval_samples_per_second": 3209.871, | |
| "eval_steps_per_second": 12.926, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 3.692357063293457, | |
| "learning_rate": 4.8e-05, | |
| "loss": 0.4442, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8154859967051071, | |
| "eval_loss": 0.41485995054244995, | |
| "eval_runtime": 1.7455, | |
| "eval_samples_per_second": 3129.809, | |
| "eval_steps_per_second": 12.604, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 4.518798351287842, | |
| "learning_rate": 4.7e-05, | |
| "loss": 0.378, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8169503935566539, | |
| "eval_loss": 0.41116195917129517, | |
| "eval_runtime": 1.6679, | |
| "eval_samples_per_second": 3275.375, | |
| "eval_steps_per_second": 13.19, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 5.528214931488037, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 0.3178, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8021233754347428, | |
| "eval_loss": 0.4807521402835846, | |
| "eval_runtime": 1.7665, | |
| "eval_samples_per_second": 3092.621, | |
| "eval_steps_per_second": 12.454, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 4.699852466583252, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.2591, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.7991945817316493, | |
| "eval_loss": 0.4954220652580261, | |
| "eval_runtime": 1.8189, | |
| "eval_samples_per_second": 3003.458, | |
| "eval_steps_per_second": 12.095, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 7.655298233032227, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 0.2115, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8072487644151565, | |
| "eval_loss": 0.5165395140647888, | |
| "eval_runtime": 1.7281, | |
| "eval_samples_per_second": 3161.244, | |
| "eval_steps_per_second": 12.731, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 7.057610511779785, | |
| "learning_rate": 4.3e-05, | |
| "loss": 0.1671, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.7982793336994325, | |
| "eval_loss": 0.6288537383079529, | |
| "eval_runtime": 1.7554, | |
| "eval_samples_per_second": 3112.04, | |
| "eval_steps_per_second": 12.532, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 11.527206420898438, | |
| "learning_rate": 4.2e-05, | |
| "loss": 0.1377, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.7957166392092258, | |
| "eval_loss": 0.6667070388793945, | |
| "eval_runtime": 1.9325, | |
| "eval_samples_per_second": 2826.959, | |
| "eval_steps_per_second": 11.384, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "step": 3280, | |
| "total_flos": 2.197388993748173e+16, | |
| "train_loss": 0.3088277851662985, | |
| "train_runtime": 480.2738, | |
| "train_samples_per_second": 10904.51, | |
| "train_steps_per_second": 42.684 | |
| } | |
| ], | |
| "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 | |
| } | |