| { | |
| "best_metric": 0.848714292049408, | |
| "best_model_checkpoint": "distilbert_sa_GLUE_Experiment_mnli_384/checkpoint-7670", | |
| "epoch": 10.0, | |
| "global_step": 15340, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 4.9e-05, | |
| "loss": 1.0075, | |
| "step": 1534 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.5303107488537953, | |
| "eval_loss": 0.9587203860282898, | |
| "eval_runtime": 5.6935, | |
| "eval_samples_per_second": 1723.896, | |
| "eval_steps_per_second": 6.85, | |
| "step": 1534 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 4.8e-05, | |
| "loss": 0.9233, | |
| "step": 3068 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.5728986245542537, | |
| "eval_loss": 0.9005416631698608, | |
| "eval_runtime": 5.5438, | |
| "eval_samples_per_second": 1770.436, | |
| "eval_steps_per_second": 7.035, | |
| "step": 3068 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 4.7e-05, | |
| "loss": 0.8749, | |
| "step": 4602 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.5887926642893531, | |
| "eval_loss": 0.8833683133125305, | |
| "eval_runtime": 5.9778, | |
| "eval_samples_per_second": 1641.912, | |
| "eval_steps_per_second": 6.524, | |
| "step": 4602 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 0.8389, | |
| "step": 6136 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.610697911360163, | |
| "eval_loss": 0.8563822507858276, | |
| "eval_runtime": 5.6621, | |
| "eval_samples_per_second": 1733.461, | |
| "eval_steps_per_second": 6.888, | |
| "step": 6136 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 4.50006518904824e-05, | |
| "loss": 0.8058, | |
| "step": 7670 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.6141619969434539, | |
| "eval_loss": 0.848714292049408, | |
| "eval_runtime": 5.565, | |
| "eval_samples_per_second": 1763.688, | |
| "eval_steps_per_second": 7.008, | |
| "step": 7670 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 4.40013037809648e-05, | |
| "loss": 0.776, | |
| "step": 9204 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.6220071319409067, | |
| "eval_loss": 0.8578256368637085, | |
| "eval_runtime": 5.8496, | |
| "eval_samples_per_second": 1677.901, | |
| "eval_steps_per_second": 6.667, | |
| "step": 9204 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "learning_rate": 4.3001955671447194e-05, | |
| "loss": 0.7467, | |
| "step": 10738 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.6187468160978095, | |
| "eval_loss": 0.8617885708808899, | |
| "eval_runtime": 5.8167, | |
| "eval_samples_per_second": 1687.379, | |
| "eval_steps_per_second": 6.705, | |
| "step": 10738 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "learning_rate": 4.20019556714472e-05, | |
| "loss": 0.7171, | |
| "step": 12272 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.6206826286296485, | |
| "eval_loss": 0.8828079104423523, | |
| "eval_runtime": 5.7638, | |
| "eval_samples_per_second": 1702.865, | |
| "eval_steps_per_second": 6.766, | |
| "step": 12272 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "learning_rate": 4.10026075619296e-05, | |
| "loss": 0.6876, | |
| "step": 13806 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.6292409577177789, | |
| "eval_loss": 0.890081524848938, | |
| "eval_runtime": 5.7709, | |
| "eval_samples_per_second": 1700.785, | |
| "eval_steps_per_second": 6.758, | |
| "step": 13806 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 4.00026075619296e-05, | |
| "loss": 0.6589, | |
| "step": 15340 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.62190524707081, | |
| "eval_loss": 0.8952736854553223, | |
| "eval_runtime": 5.7281, | |
| "eval_samples_per_second": 1713.472, | |
| "eval_steps_per_second": 6.808, | |
| "step": 15340 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 15340, | |
| "total_flos": 1.0787244155600896e+17, | |
| "train_loss": 0.8036706678248757, | |
| "train_runtime": 5092.3208, | |
| "train_samples_per_second": 3855.825, | |
| "train_steps_per_second": 15.062 | |
| } | |
| ], | |
| "max_steps": 76700, | |
| "num_train_epochs": 50, | |
| "total_flos": 1.0787244155600896e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |