| { | |
| "best_metric": 0.695, | |
| "best_model_checkpoint": "chord-final-model/checkpoint-40", | |
| "epoch": 4.912280701754386, | |
| "eval_steps": 500, | |
| "global_step": 40, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.98, | |
| "eval_accuracy": 0.4, | |
| "eval_loss": 1.5147000551223755, | |
| "eval_runtime": 2.3038, | |
| "eval_samples_per_second": 86.815, | |
| "eval_steps_per_second": 3.039, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 2.2252283096313477, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 1.6354, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "eval_accuracy": 0.585, | |
| "eval_loss": 1.3898262977600098, | |
| "eval_runtime": 2.2865, | |
| "eval_samples_per_second": 87.469, | |
| "eval_steps_per_second": 3.061, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "grad_norm": 2.2747433185577393, | |
| "learning_rate": 5.555555555555557e-06, | |
| "loss": 1.4272, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.95, | |
| "eval_accuracy": 0.67, | |
| "eval_loss": 1.296023964881897, | |
| "eval_runtime": 2.2818, | |
| "eval_samples_per_second": 87.65, | |
| "eval_steps_per_second": 3.068, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 3.68, | |
| "grad_norm": 2.199195623397827, | |
| "learning_rate": 2.7777777777777783e-06, | |
| "loss": 1.2884, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.93, | |
| "eval_accuracy": 0.67, | |
| "eval_loss": 1.2369245290756226, | |
| "eval_runtime": 2.2809, | |
| "eval_samples_per_second": 87.684, | |
| "eval_steps_per_second": 3.069, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 4.91, | |
| "grad_norm": 2.335061550140381, | |
| "learning_rate": 0.0, | |
| "loss": 1.2176, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.91, | |
| "eval_accuracy": 0.695, | |
| "eval_loss": 1.213279128074646, | |
| "eval_runtime": 2.345, | |
| "eval_samples_per_second": 85.287, | |
| "eval_steps_per_second": 2.985, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.91, | |
| "step": 40, | |
| "total_flos": 6.869074650501612e+17, | |
| "train_loss": 1.3921312093734741, | |
| "train_runtime": 216.3192, | |
| "train_samples_per_second": 41.605, | |
| "train_steps_per_second": 0.185 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 40, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 6.869074650501612e+17, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |