| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 7.861635220125786, | |
| "eval_steps": 500, | |
| "global_step": 2500, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6267741935483871, | |
| "eval_loss": 0.19981417059898376, | |
| "eval_runtime": 5.7251, | |
| "eval_samples_per_second": 541.471, | |
| "eval_steps_per_second": 11.353, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "learning_rate": 1.685534591194969e-05, | |
| "loss": 0.324, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8287096774193549, | |
| "eval_loss": 0.09645497798919678, | |
| "eval_runtime": 5.9908, | |
| "eval_samples_per_second": 517.464, | |
| "eval_steps_per_second": 10.85, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8841935483870967, | |
| "eval_loss": 0.061595458537340164, | |
| "eval_runtime": 5.711, | |
| "eval_samples_per_second": 542.813, | |
| "eval_steps_per_second": 11.382, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "learning_rate": 1.371069182389937e-05, | |
| "loss": 0.1107, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9035483870967742, | |
| "eval_loss": 0.046646881848573685, | |
| "eval_runtime": 5.7075, | |
| "eval_samples_per_second": 543.147, | |
| "eval_steps_per_second": 11.389, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 4.72, | |
| "learning_rate": 1.0566037735849058e-05, | |
| "loss": 0.0692, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9096774193548387, | |
| "eval_loss": 0.03824981674551964, | |
| "eval_runtime": 5.8862, | |
| "eval_samples_per_second": 526.657, | |
| "eval_steps_per_second": 11.043, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9193548387096774, | |
| "eval_loss": 0.033315952867269516, | |
| "eval_runtime": 5.7206, | |
| "eval_samples_per_second": 541.904, | |
| "eval_steps_per_second": 11.362, | |
| "step": 1908 | |
| }, | |
| { | |
| "epoch": 6.29, | |
| "learning_rate": 7.421383647798742e-06, | |
| "loss": 0.0539, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.9235483870967742, | |
| "eval_loss": 0.03077365830540657, | |
| "eval_runtime": 5.6787, | |
| "eval_samples_per_second": 545.903, | |
| "eval_steps_per_second": 11.446, | |
| "step": 2226 | |
| }, | |
| { | |
| "epoch": 7.86, | |
| "learning_rate": 4.276729559748428e-06, | |
| "loss": 0.0464, | |
| "step": 2500 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 3180, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 649131947564688.0, | |
| "train_batch_size": 48, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.924670627776918, | |
| "num_train_epochs": 10, | |
| "temperature": 10 | |
| } | |
| } | |