| { | |
| "best_metric": 0.4154828190803528, | |
| "best_model_checkpoint": "bert_base_lda_100_v1_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": 3.7200541496276855, | |
| "learning_rate": 4.9e-05, | |
| "loss": 0.636, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7078528281164196, | |
| "eval_loss": 0.5723043084144592, | |
| "eval_runtime": 3.5417, | |
| "eval_samples_per_second": 1542.464, | |
| "eval_steps_per_second": 6.212, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 3.76737904548645, | |
| "learning_rate": 4.8e-05, | |
| "loss": 0.471, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.80907926047959, | |
| "eval_loss": 0.4177427887916565, | |
| "eval_runtime": 3.541, | |
| "eval_samples_per_second": 1542.774, | |
| "eval_steps_per_second": 6.213, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 4.409963607788086, | |
| "learning_rate": 4.7e-05, | |
| "loss": 0.3689, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8063335163829398, | |
| "eval_loss": 0.4154828190803528, | |
| "eval_runtime": 3.5825, | |
| "eval_samples_per_second": 1524.909, | |
| "eval_steps_per_second": 6.141, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 7.4005537033081055, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 0.2829, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7990115321252059, | |
| "eval_loss": 0.48499900102615356, | |
| "eval_runtime": 3.537, | |
| "eval_samples_per_second": 1544.511, | |
| "eval_steps_per_second": 6.22, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 3.301100254058838, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.2021, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.7889438037708218, | |
| "eval_loss": 0.5746128559112549, | |
| "eval_runtime": 3.5393, | |
| "eval_samples_per_second": 1543.523, | |
| "eval_steps_per_second": 6.216, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 3.967524290084839, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 0.1431, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.7920556470803588, | |
| "eval_loss": 0.6877047419548035, | |
| "eval_runtime": 3.5292, | |
| "eval_samples_per_second": 1547.962, | |
| "eval_steps_per_second": 6.234, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 6.012539386749268, | |
| "learning_rate": 4.3e-05, | |
| "loss": 0.1066, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.7841845140032949, | |
| "eval_loss": 0.7942188382148743, | |
| "eval_runtime": 3.5505, | |
| "eval_samples_per_second": 1538.67, | |
| "eval_steps_per_second": 6.196, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 9.717620849609375, | |
| "learning_rate": 4.2e-05, | |
| "loss": 0.0827, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.791689547867472, | |
| "eval_loss": 0.7339684963226318, | |
| "eval_runtime": 3.5897, | |
| "eval_samples_per_second": 1521.835, | |
| "eval_steps_per_second": 6.129, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "step": 3280, | |
| "total_flos": 1.1023616508628992e+17, | |
| "train_loss": 0.2866605537693675, | |
| "train_runtime": 1408.5203, | |
| "train_samples_per_second": 3718.193, | |
| "train_steps_per_second": 14.554 | |
| } | |
| ], | |
| "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": 1.1023616508628992e+17, | |
| "train_batch_size": 256, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |