| { | |
| "best_metric": 0.6178650259971619, | |
| "best_model_checkpoint": "BERT-WMM/run-11/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": 10.442337036132812, | |
| "learning_rate": 1.5945617993693323e-05, | |
| "loss": 0.7201, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7488262910798122, | |
| "eval_loss": 0.6336334943771362, | |
| "eval_runtime": 2.089, | |
| "eval_samples_per_second": 1019.614, | |
| "eval_steps_per_second": 64.145, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 9.159146308898926, | |
| "learning_rate": 8.691014720857415e-06, | |
| "loss": 0.4449, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7647887323943662, | |
| "eval_loss": 0.6178650259971619, | |
| "eval_runtime": 2.1564, | |
| "eval_samples_per_second": 987.735, | |
| "eval_steps_per_second": 62.139, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "grad_norm": 26.11417007446289, | |
| "learning_rate": 1.4364114480215095e-06, | |
| "loss": 0.2581, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7652582159624414, | |
| "eval_loss": 0.8162142634391785, | |
| "eval_runtime": 2.1257, | |
| "eval_samples_per_second": 1002.034, | |
| "eval_steps_per_second": 63.039, | |
| "step": 1599 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 1599, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 518314932761976.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": { | |
| "learning_rate": 2.320022126652923e-05 | |
| } | |
| } | |