| { | |
| "best_global_step": 6136, | |
| "best_metric": 0.7984881401062012, | |
| "best_model_checkpoint": "bert_base_km_5_v1_mnli/checkpoint-6136", | |
| "epoch": 9.0, | |
| "eval_steps": 500, | |
| "global_step": 13806, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.2298744916915894, | |
| "learning_rate": 4.9e-05, | |
| "loss": 0.9844, | |
| "step": 1534 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.5690269994905757, | |
| "eval_loss": 0.9082422256469727, | |
| "eval_runtime": 6.2931, | |
| "eval_samples_per_second": 1559.643, | |
| "eval_steps_per_second": 6.197, | |
| "step": 1534 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.6216703653335571, | |
| "learning_rate": 4.8e-05, | |
| "loss": 0.8695, | |
| "step": 3068 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.6127356087620989, | |
| "eval_loss": 0.8509761691093445, | |
| "eval_runtime": 6.2158, | |
| "eval_samples_per_second": 1579.046, | |
| "eval_steps_per_second": 6.274, | |
| "step": 3068 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.8060803413391113, | |
| "learning_rate": 4.7e-05, | |
| "loss": 0.7774, | |
| "step": 4602 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.6435048395313296, | |
| "eval_loss": 0.8041264414787292, | |
| "eval_runtime": 6.2225, | |
| "eval_samples_per_second": 1577.334, | |
| "eval_steps_per_second": 6.268, | |
| "step": 4602 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 2.345538854598999, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 0.6904, | |
| "step": 6136 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.6587875700458482, | |
| "eval_loss": 0.7984881401062012, | |
| "eval_runtime": 6.2467, | |
| "eval_samples_per_second": 1571.228, | |
| "eval_steps_per_second": 6.243, | |
| "step": 6136 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 2.3710386753082275, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.6068, | |
| "step": 7670 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.6659195109526236, | |
| "eval_loss": 0.8158076405525208, | |
| "eval_runtime": 6.2317, | |
| "eval_samples_per_second": 1575.016, | |
| "eval_steps_per_second": 6.258, | |
| "step": 7670 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 3.218816041946411, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 0.521, | |
| "step": 9204 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.6599083036169129, | |
| "eval_loss": 0.9025458097457886, | |
| "eval_runtime": 6.2247, | |
| "eval_samples_per_second": 1576.787, | |
| "eval_steps_per_second": 6.265, | |
| "step": 9204 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 2.895859956741333, | |
| "learning_rate": 4.3e-05, | |
| "loss": 0.4345, | |
| "step": 10738 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.6590932246561385, | |
| "eval_loss": 0.947009801864624, | |
| "eval_runtime": 6.2527, | |
| "eval_samples_per_second": 1569.728, | |
| "eval_steps_per_second": 6.237, | |
| "step": 10738 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 4.124659061431885, | |
| "learning_rate": 4.2e-05, | |
| "loss": 0.3593, | |
| "step": 12272 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.6571574121242996, | |
| "eval_loss": 1.0819920301437378, | |
| "eval_runtime": 6.2552, | |
| "eval_samples_per_second": 1569.088, | |
| "eval_steps_per_second": 6.235, | |
| "step": 12272 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 4.0520501136779785, | |
| "learning_rate": 4.1e-05, | |
| "loss": 0.2939, | |
| "step": 13806 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.6549159449821701, | |
| "eval_loss": 1.2040681838989258, | |
| "eval_runtime": 6.2329, | |
| "eval_samples_per_second": 1574.711, | |
| "eval_steps_per_second": 6.257, | |
| "step": 13806 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "step": 13806, | |
| "total_flos": 4.6496324415879475e+17, | |
| "train_loss": 0.6152484290896508, | |
| "train_runtime": 5805.2431, | |
| "train_samples_per_second": 3382.305, | |
| "train_steps_per_second": 13.212 | |
| } | |
| ], | |
| "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": 4.6496324415879475e+17, | |
| "train_batch_size": 256, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |