| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 13.0, | |
| "global_step": 13273, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 1.9347045380346067e-05, | |
| "loss": 0.1777, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 1.8694090760692135e-05, | |
| "loss": 0.1437, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.9603556990623474, | |
| "eval_loss": 0.13507625460624695, | |
| "eval_runtime": 3.6085, | |
| "eval_samples_per_second": 2586.408, | |
| "eval_steps_per_second": 80.921, | |
| "step": 1021 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "learning_rate": 1.80411361410382e-05, | |
| "loss": 0.1001, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "learning_rate": 1.7388181521384265e-05, | |
| "loss": 0.11, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9631415605545044, | |
| "eval_loss": 0.12850655615329742, | |
| "eval_runtime": 3.5831, | |
| "eval_samples_per_second": 2604.746, | |
| "eval_steps_per_second": 81.494, | |
| "step": 2042 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "learning_rate": 1.673522690173033e-05, | |
| "loss": 0.0762, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "learning_rate": 1.6082272282076395e-05, | |
| "loss": 0.0722, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9621772170066833, | |
| "eval_loss": 0.1381947100162506, | |
| "eval_runtime": 3.5883, | |
| "eval_samples_per_second": 2600.987, | |
| "eval_steps_per_second": 81.377, | |
| "step": 3063 | |
| }, | |
| { | |
| "epoch": 3.43, | |
| "learning_rate": 1.5429317662422463e-05, | |
| "loss": 0.0569, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 3.92, | |
| "learning_rate": 1.4776363042768528e-05, | |
| "loss": 0.0442, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.966034471988678, | |
| "eval_loss": 0.1748303323984146, | |
| "eval_runtime": 3.5824, | |
| "eval_samples_per_second": 2605.239, | |
| "eval_steps_per_second": 81.51, | |
| "step": 4084 | |
| }, | |
| { | |
| "epoch": 4.41, | |
| "learning_rate": 1.4123408423114595e-05, | |
| "loss": 0.0301, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 4.9, | |
| "learning_rate": 1.3470453803460662e-05, | |
| "loss": 0.031, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9617486596107483, | |
| "eval_loss": 0.21482647955417633, | |
| "eval_runtime": 3.6096, | |
| "eval_samples_per_second": 2585.638, | |
| "eval_steps_per_second": 80.896, | |
| "step": 5105 | |
| }, | |
| { | |
| "epoch": 5.39, | |
| "learning_rate": 1.2817499183806728e-05, | |
| "loss": 0.0222, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 5.88, | |
| "learning_rate": 1.2164544564152792e-05, | |
| "loss": 0.0236, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9650701880455017, | |
| "eval_loss": 0.20721212029457092, | |
| "eval_runtime": 3.59, | |
| "eval_samples_per_second": 2599.744, | |
| "eval_steps_per_second": 81.338, | |
| "step": 6126 | |
| }, | |
| { | |
| "epoch": 6.37, | |
| "learning_rate": 1.1511589944498858e-05, | |
| "loss": 0.0135, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 6.86, | |
| "learning_rate": 1.0858635324844923e-05, | |
| "loss": 0.0123, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.9661416411399841, | |
| "eval_loss": 0.23626652359962463, | |
| "eval_runtime": 3.6087, | |
| "eval_samples_per_second": 2586.26, | |
| "eval_steps_per_second": 80.916, | |
| "step": 7147 | |
| }, | |
| { | |
| "epoch": 7.35, | |
| "learning_rate": 1.020568070519099e-05, | |
| "loss": 0.0085, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 7.84, | |
| "learning_rate": 9.552726085537057e-06, | |
| "loss": 0.0069, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.966034471988678, | |
| "eval_loss": 0.28718122839927673, | |
| "eval_runtime": 3.6111, | |
| "eval_samples_per_second": 2584.509, | |
| "eval_steps_per_second": 80.861, | |
| "step": 8168 | |
| }, | |
| { | |
| "epoch": 8.33, | |
| "learning_rate": 8.899771465883122e-06, | |
| "loss": 0.0053, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 8.81, | |
| "learning_rate": 8.246816846229188e-06, | |
| "loss": 0.0051, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.9598199725151062, | |
| "eval_loss": 0.30353665351867676, | |
| "eval_runtime": 3.5877, | |
| "eval_samples_per_second": 2601.423, | |
| "eval_steps_per_second": 81.39, | |
| "step": 9189 | |
| }, | |
| { | |
| "epoch": 9.3, | |
| "learning_rate": 7.593862226575253e-06, | |
| "loss": 0.0052, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 9.79, | |
| "learning_rate": 6.94090760692132e-06, | |
| "loss": 0.0036, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.9662488102912903, | |
| "eval_loss": 0.32062527537345886, | |
| "eval_runtime": 3.5765, | |
| "eval_samples_per_second": 2609.519, | |
| "eval_steps_per_second": 81.644, | |
| "step": 10210 | |
| }, | |
| { | |
| "epoch": 10.28, | |
| "learning_rate": 6.287952987267385e-06, | |
| "loss": 0.0042, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 10.77, | |
| "learning_rate": 5.634998367613451e-06, | |
| "loss": 0.0052, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.9659273624420166, | |
| "eval_loss": 0.3190840780735016, | |
| "eval_runtime": 3.6013, | |
| "eval_samples_per_second": 2591.577, | |
| "eval_steps_per_second": 81.082, | |
| "step": 11231 | |
| }, | |
| { | |
| "epoch": 11.26, | |
| "learning_rate": 4.9820437479595175e-06, | |
| "loss": 0.0029, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 11.75, | |
| "learning_rate": 4.329089128305583e-06, | |
| "loss": 0.002, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.966570258140564, | |
| "eval_loss": 0.3279273509979248, | |
| "eval_runtime": 3.5932, | |
| "eval_samples_per_second": 2597.412, | |
| "eval_steps_per_second": 81.265, | |
| "step": 12252 | |
| }, | |
| { | |
| "epoch": 12.24, | |
| "learning_rate": 3.676134508651649e-06, | |
| "loss": 0.0023, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 12.73, | |
| "learning_rate": 3.023179888997715e-06, | |
| "loss": 0.0011, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.9678559899330139, | |
| "eval_loss": 0.331949383020401, | |
| "eval_runtime": 3.605, | |
| "eval_samples_per_second": 2588.877, | |
| "eval_steps_per_second": 80.998, | |
| "step": 13273 | |
| } | |
| ], | |
| "max_steps": 15315, | |
| "num_train_epochs": 15, | |
| "total_flos": 2.7931343434285056e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |