| { | |
| "best_metric": 0.75532066822052, | |
| "best_model_checkpoint": "BERT-WMM/run-0/checkpoint-1599", | |
| "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": 10.492278099060059, | |
| "learning_rate": 7.698774119953696e-07, | |
| "loss": 0.9956, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6281690140845071, | |
| "eval_loss": 0.8787437081336975, | |
| "eval_runtime": 2.0952, | |
| "eval_samples_per_second": 1016.614, | |
| "eval_steps_per_second": 63.956, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 16.712677001953125, | |
| "learning_rate": 4.1961471318036977e-07, | |
| "loss": 0.858, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.6732394366197183, | |
| "eval_loss": 0.791917085647583, | |
| "eval_runtime": 2.1109, | |
| "eval_samples_per_second": 1009.044, | |
| "eval_steps_per_second": 63.48, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "grad_norm": 29.26333999633789, | |
| "learning_rate": 6.935201436536996e-08, | |
| "loss": 0.7846, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.6892018779342723, | |
| "eval_loss": 0.75532066822052, | |
| "eval_runtime": 2.0969, | |
| "eval_samples_per_second": 1015.8, | |
| "eval_steps_per_second": 63.905, | |
| "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": 1.1201401108103694e-06 | |
| } | |
| } | |