| { | |
| "best_metric": 0.87736, | |
| "best_model_checkpoint": "outputs/checkpoint-702", | |
| "epoch": 3.0, | |
| "eval_steps": 78, | |
| "global_step": 1173, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 1.6949152542372883e-07, | |
| "loss": 0.6918, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 1.3220338983050848e-05, | |
| "loss": 0.6487, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "eval_accuracy": 0.81168, | |
| "eval_loss": 0.4401787221431732, | |
| "eval_runtime": 30.2854, | |
| "eval_samples_per_second": 825.48, | |
| "eval_steps_per_second": 12.911, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 1.9279620853080568e-05, | |
| "loss": 0.402, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "eval_accuracy": 0.8416, | |
| "eval_loss": 0.356289267539978, | |
| "eval_runtime": 30.4052, | |
| "eval_samples_per_second": 822.228, | |
| "eval_steps_per_second": 12.86, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 1.780094786729858e-05, | |
| "loss": 0.3528, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "eval_accuracy": 0.8434, | |
| "eval_loss": 0.3521649241447449, | |
| "eval_runtime": 30.6429, | |
| "eval_samples_per_second": 815.849, | |
| "eval_steps_per_second": 12.76, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 1.632227488151659e-05, | |
| "loss": 0.3362, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "eval_accuracy": 0.86516, | |
| "eval_loss": 0.30993545055389404, | |
| "eval_runtime": 30.6544, | |
| "eval_samples_per_second": 815.543, | |
| "eval_steps_per_second": 12.755, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 1.4843601895734598e-05, | |
| "loss": 0.3184, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.86996, | |
| "eval_loss": 0.30280688405036926, | |
| "eval_runtime": 30.6512, | |
| "eval_samples_per_second": 815.629, | |
| "eval_steps_per_second": 12.756, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 1.3364928909952607e-05, | |
| "loss": 0.265, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "eval_accuracy": 0.8738, | |
| "eval_loss": 0.305215448141098, | |
| "eval_runtime": 30.5331, | |
| "eval_samples_per_second": 818.783, | |
| "eval_steps_per_second": 12.806, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "learning_rate": 1.1886255924170618e-05, | |
| "loss": 0.2593, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "eval_accuracy": 0.87348, | |
| "eval_loss": 0.29833072423934937, | |
| "eval_runtime": 30.6351, | |
| "eval_samples_per_second": 816.057, | |
| "eval_steps_per_second": 12.763, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "learning_rate": 1.0407582938388628e-05, | |
| "loss": 0.2537, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "eval_accuracy": 0.87536, | |
| "eval_loss": 0.2977478802204132, | |
| "eval_runtime": 30.5918, | |
| "eval_samples_per_second": 817.213, | |
| "eval_steps_per_second": 12.781, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "learning_rate": 8.928909952606636e-06, | |
| "loss": 0.2558, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "eval_accuracy": 0.87736, | |
| "eval_loss": 0.29114434123039246, | |
| "eval_runtime": 30.9298, | |
| "eval_samples_per_second": 808.283, | |
| "eval_steps_per_second": 12.642, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "learning_rate": 7.450236966824646e-06, | |
| "loss": 0.2476, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "eval_accuracy": 0.87508, | |
| "eval_loss": 0.2907171845436096, | |
| "eval_runtime": 30.4929, | |
| "eval_samples_per_second": 819.862, | |
| "eval_steps_per_second": 12.823, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "learning_rate": 5.971563981042654e-06, | |
| "loss": 0.1941, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "eval_accuracy": 0.87736, | |
| "eval_loss": 0.3151108920574188, | |
| "eval_runtime": 30.5003, | |
| "eval_samples_per_second": 819.664, | |
| "eval_steps_per_second": 12.82, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "learning_rate": 4.492890995260664e-06, | |
| "loss": 0.1873, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "eval_accuracy": 0.87644, | |
| "eval_loss": 0.31038883328437805, | |
| "eval_runtime": 30.5889, | |
| "eval_samples_per_second": 817.29, | |
| "eval_steps_per_second": 12.782, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "learning_rate": 3.0142180094786734e-06, | |
| "loss": 0.1869, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "eval_accuracy": 0.87708, | |
| "eval_loss": 0.3180868625640869, | |
| "eval_runtime": 30.5304, | |
| "eval_samples_per_second": 818.855, | |
| "eval_steps_per_second": 12.807, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "learning_rate": 1.5355450236966826e-06, | |
| "loss": 0.1807, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "eval_accuracy": 0.87636, | |
| "eval_loss": 0.3148181140422821, | |
| "eval_runtime": 30.541, | |
| "eval_samples_per_second": 818.572, | |
| "eval_steps_per_second": 12.802, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "learning_rate": 5.6872037914691944e-08, | |
| "loss": 0.1967, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "eval_accuracy": 0.8766, | |
| "eval_loss": 0.3140537142753601, | |
| "eval_runtime": 30.4059, | |
| "eval_samples_per_second": 822.21, | |
| "eval_steps_per_second": 12.859, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 1173, | |
| "total_flos": 2483763724800000.0, | |
| "train_loss": 0.2856195556334929, | |
| "train_runtime": 750.4873, | |
| "train_samples_per_second": 99.935, | |
| "train_steps_per_second": 1.563 | |
| } | |
| ], | |
| "logging_steps": 78, | |
| "max_steps": 1173, | |
| "num_train_epochs": 3, | |
| "save_steps": 78, | |
| "total_flos": 2483763724800000.0, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |