| { | |
| "best_metric": 0.6433334350585938, | |
| "best_model_checkpoint": "BERT-WMM/run-12/checkpoint-1066", | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 1599, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 10.519768714904785, | |
| "learning_rate": 2.0153542741677195e-05, | |
| "loss": 0.7357, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7380281690140845, | |
| "eval_loss": 0.6585038304328918, | |
| "eval_runtime": 2.092, | |
| "eval_samples_per_second": 1018.141, | |
| "eval_steps_per_second": 64.052, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 10.527291297912598, | |
| "learning_rate": 1.0984506007520144e-05, | |
| "loss": 0.4807, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7624413145539906, | |
| "eval_loss": 0.6433334350585938, | |
| "eval_runtime": 2.1058, | |
| "eval_samples_per_second": 1011.511, | |
| "eval_steps_per_second": 63.635, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "grad_norm": 16.033649444580078, | |
| "learning_rate": 1.8154692733630958e-06, | |
| "loss": 0.2957, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7713615023474178, | |
| "eval_loss": 0.7535221576690674, | |
| "eval_runtime": 2.0955, | |
| "eval_samples_per_second": 1016.443, | |
| "eval_steps_per_second": 63.945, | |
| "step": 1599 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 1599, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 507646505902536.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": { | |
| "learning_rate": 2.9322579475834245e-05 | |
| } | |
| } | |