| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.961038961038961, | |
| "eval_steps": 500, | |
| "global_step": 57, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.4391919003064584, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 2.7465, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 1.01878785429204, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 2.3521, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.7432627531351906, | |
| "learning_rate": 1.9697969360350098e-05, | |
| "loss": 2.3746, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.6306502028658637, | |
| "learning_rate": 1.8502171357296144e-05, | |
| "loss": 2.2849, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "eval_loss": 1.2240214347839355, | |
| "eval_runtime": 38.055, | |
| "eval_samples_per_second": 9.618, | |
| "eval_steps_per_second": 0.604, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 0.8323697335437488, | |
| "learning_rate": 1.650618300204242e-05, | |
| "loss": 2.41, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 0.8356330876376933, | |
| "learning_rate": 1.3897858732926794e-05, | |
| "loss": 2.078, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 0.8229304096804662, | |
| "learning_rate": 1.092268359463302e-05, | |
| "loss": 2.0295, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 0.6758717282694323, | |
| "learning_rate": 7.860669167935028e-06, | |
| "loss": 1.9826, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "eval_loss": 1.252795934677124, | |
| "eval_runtime": 38.2606, | |
| "eval_samples_per_second": 9.566, | |
| "eval_steps_per_second": 0.601, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 1.0361066126350802, | |
| "learning_rate": 5.000000000000003e-06, | |
| "loss": 1.9803, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 2.34, | |
| "grad_norm": 0.9815477883010376, | |
| "learning_rate": 2.6099108277934105e-06, | |
| "loss": 1.8108, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 0.894384535047576, | |
| "learning_rate": 9.153472818047627e-07, | |
| "loss": 1.8236, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "grad_norm": 0.8145920730047358, | |
| "learning_rate": 7.579490328064265e-08, | |
| "loss": 1.7301, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "eval_loss": 1.304147720336914, | |
| "eval_runtime": 38.5373, | |
| "eval_samples_per_second": 9.497, | |
| "eval_steps_per_second": 0.597, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "step": 57, | |
| "total_flos": 11882295459840.0, | |
| "train_loss": 2.079949947825649, | |
| "train_runtime": 1266.1143, | |
| "train_samples_per_second": 2.912, | |
| "train_steps_per_second": 0.045 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 57, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 11882295459840.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |