| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.2597631645250693, | |
| "eval_steps": 500, | |
| "global_step": 10000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 9.5e-05, | |
| "loss": 57.2732, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "eval_loss": 51.029930114746094, | |
| "eval_runtime": 3.1759, | |
| "eval_samples_per_second": 31.487, | |
| "eval_steps_per_second": 4.093, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 9e-05, | |
| "loss": 56.0705, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "eval_loss": 49.898155212402344, | |
| "eval_runtime": 4.4269, | |
| "eval_samples_per_second": 22.589, | |
| "eval_steps_per_second": 2.937, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 8.5e-05, | |
| "loss": 55.3144, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "eval_loss": 49.3874397277832, | |
| "eval_runtime": 4.228, | |
| "eval_samples_per_second": 23.652, | |
| "eval_steps_per_second": 3.075, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 8e-05, | |
| "loss": 54.301, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "eval_loss": 48.694358825683594, | |
| "eval_runtime": 3.1949, | |
| "eval_samples_per_second": 31.299, | |
| "eval_steps_per_second": 4.069, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 54.8258, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "eval_loss": 49.120418548583984, | |
| "eval_runtime": 4.0702, | |
| "eval_samples_per_second": 24.569, | |
| "eval_steps_per_second": 3.194, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 7e-05, | |
| "loss": 55.0078, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "eval_loss": 48.479034423828125, | |
| "eval_runtime": 3.9143, | |
| "eval_samples_per_second": 25.547, | |
| "eval_steps_per_second": 3.321, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 55.0228, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "eval_loss": 48.023162841796875, | |
| "eval_runtime": 3.9112, | |
| "eval_samples_per_second": 25.568, | |
| "eval_steps_per_second": 3.324, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 6e-05, | |
| "loss": 53.2773, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "eval_loss": 47.419307708740234, | |
| "eval_runtime": 4.7859, | |
| "eval_samples_per_second": 20.895, | |
| "eval_steps_per_second": 2.716, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 54.9082, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "eval_loss": 46.841548919677734, | |
| "eval_runtime": 4.8111, | |
| "eval_samples_per_second": 20.785, | |
| "eval_steps_per_second": 2.702, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 5e-05, | |
| "loss": 53.3684, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "eval_loss": 46.429256439208984, | |
| "eval_runtime": 4.525, | |
| "eval_samples_per_second": 22.099, | |
| "eval_steps_per_second": 2.873, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 4.5e-05, | |
| "loss": 54.4457, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "eval_loss": 46.04310607910156, | |
| "eval_runtime": 4.6243, | |
| "eval_samples_per_second": 21.625, | |
| "eval_steps_per_second": 2.811, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 4e-05, | |
| "loss": 53.9792, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "eval_loss": 45.82972717285156, | |
| "eval_runtime": 3.6612, | |
| "eval_samples_per_second": 27.314, | |
| "eval_steps_per_second": 3.551, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 3.5e-05, | |
| "loss": 53.924, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "eval_loss": 45.8700065612793, | |
| "eval_runtime": 4.1117, | |
| "eval_samples_per_second": 24.321, | |
| "eval_steps_per_second": 3.162, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 3e-05, | |
| "loss": 54.2944, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "eval_loss": 45.5869140625, | |
| "eval_runtime": 4.5188, | |
| "eval_samples_per_second": 22.13, | |
| "eval_steps_per_second": 2.877, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 2.5e-05, | |
| "loss": 53.9664, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "eval_loss": 45.52823257446289, | |
| "eval_runtime": 4.5014, | |
| "eval_samples_per_second": 22.215, | |
| "eval_steps_per_second": 2.888, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 2e-05, | |
| "loss": 52.0975, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "eval_loss": 45.37757873535156, | |
| "eval_runtime": 4.4906, | |
| "eval_samples_per_second": 22.269, | |
| "eval_steps_per_second": 2.895, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 1.5e-05, | |
| "loss": 50.0576, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "eval_loss": 45.175968170166016, | |
| "eval_runtime": 4.5314, | |
| "eval_samples_per_second": 22.068, | |
| "eval_steps_per_second": 2.869, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "learning_rate": 1e-05, | |
| "loss": 50.6407, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "eval_loss": 45.02992630004883, | |
| "eval_runtime": 4.196, | |
| "eval_samples_per_second": 23.832, | |
| "eval_steps_per_second": 3.098, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 5e-06, | |
| "loss": 50.3118, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "eval_loss": 44.964359283447266, | |
| "eval_runtime": 4.7953, | |
| "eval_samples_per_second": 20.854, | |
| "eval_steps_per_second": 2.711, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "learning_rate": 0.0, | |
| "loss": 48.9218, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "eval_loss": 44.92063903808594, | |
| "eval_runtime": 4.609, | |
| "eval_samples_per_second": 21.697, | |
| "eval_steps_per_second": 2.821, | |
| "step": 10000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 10000, | |
| "num_train_epochs": 2, | |
| "save_steps": 5000, | |
| "total_flos": 1.3694633480945664e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |