| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 39.01170351105332, | |
| "eval_steps": 1500, | |
| "global_step": 22500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 2.6, | |
| "learning_rate": 0.0002932622504537205, | |
| "loss": 6.742, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "eval_cer": 0.2330206972312148, | |
| "eval_loss": 0.9037942886352539, | |
| "eval_runtime": 202.8675, | |
| "eval_samples_per_second": 22.704, | |
| "eval_steps_per_second": 2.839, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "learning_rate": 0.00027284482758620687, | |
| "loss": 0.9228, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "eval_cer": 0.1655612053827781, | |
| "eval_loss": 0.6193090677261353, | |
| "eval_runtime": 202.6276, | |
| "eval_samples_per_second": 22.731, | |
| "eval_steps_per_second": 2.843, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 7.8, | |
| "learning_rate": 0.00025242740471869324, | |
| "loss": 0.6805, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 7.8, | |
| "eval_cer": 0.14813458609698654, | |
| "eval_loss": 0.5522382259368896, | |
| "eval_runtime": 201.9891, | |
| "eval_samples_per_second": 22.803, | |
| "eval_steps_per_second": 2.852, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 10.4, | |
| "learning_rate": 0.00023200998185117966, | |
| "loss": 0.5577, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 10.4, | |
| "eval_cer": 0.1348940171935986, | |
| "eval_loss": 0.5135979652404785, | |
| "eval_runtime": 208.2736, | |
| "eval_samples_per_second": 22.115, | |
| "eval_steps_per_second": 2.766, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "learning_rate": 0.00021160617059891106, | |
| "loss": 0.4797, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_cer": 0.1312322633799652, | |
| "eval_loss": 0.5074160099029541, | |
| "eval_runtime": 207.3615, | |
| "eval_samples_per_second": 22.212, | |
| "eval_steps_per_second": 2.778, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 15.6, | |
| "learning_rate": 0.00019120235934664247, | |
| "loss": 0.4161, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 15.6, | |
| "eval_cer": 0.12432486414061135, | |
| "eval_loss": 0.4958959221839905, | |
| "eval_runtime": 198.4858, | |
| "eval_samples_per_second": 23.206, | |
| "eval_steps_per_second": 2.902, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 18.21, | |
| "learning_rate": 0.00017078493647912883, | |
| "loss": 0.3701, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 18.21, | |
| "eval_cer": 0.12240244338845382, | |
| "eval_loss": 0.4947836697101593, | |
| "eval_runtime": 198.4077, | |
| "eval_samples_per_second": 23.215, | |
| "eval_steps_per_second": 2.903, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 20.81, | |
| "learning_rate": 0.00015038112522686024, | |
| "loss": 0.3307, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 20.81, | |
| "eval_cer": 0.11993908173200955, | |
| "eval_loss": 0.48810431361198425, | |
| "eval_runtime": 197.2477, | |
| "eval_samples_per_second": 23.351, | |
| "eval_steps_per_second": 2.92, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 23.41, | |
| "learning_rate": 0.00012996370235934663, | |
| "loss": 0.2946, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 23.41, | |
| "eval_cer": 0.1179251171345112, | |
| "eval_loss": 0.4970262050628662, | |
| "eval_runtime": 201.6098, | |
| "eval_samples_per_second": 22.846, | |
| "eval_steps_per_second": 2.857, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 26.01, | |
| "learning_rate": 0.00010954627949183302, | |
| "loss": 0.2636, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 26.01, | |
| "eval_cer": 0.11448806185035078, | |
| "eval_loss": 0.4949529767036438, | |
| "eval_runtime": 204.2568, | |
| "eval_samples_per_second": 22.55, | |
| "eval_steps_per_second": 2.82, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 28.61, | |
| "learning_rate": 8.915607985480944e-05, | |
| "loss": 0.2367, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 28.61, | |
| "eval_cer": 0.11192483418080741, | |
| "eval_loss": 0.49051955342292786, | |
| "eval_runtime": 200.659, | |
| "eval_samples_per_second": 22.954, | |
| "eval_steps_per_second": 2.871, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 31.21, | |
| "learning_rate": 6.875226860254083e-05, | |
| "loss": 0.2157, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 31.21, | |
| "eval_cer": 0.11102604006291558, | |
| "eval_loss": 0.5065664649009705, | |
| "eval_runtime": 198.3314, | |
| "eval_samples_per_second": 23.224, | |
| "eval_steps_per_second": 2.904, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 33.81, | |
| "learning_rate": 4.833484573502722e-05, | |
| "loss": 0.1979, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 33.81, | |
| "eval_cer": 0.11034362230673846, | |
| "eval_loss": 0.5133121609687805, | |
| "eval_runtime": 198.0306, | |
| "eval_samples_per_second": 23.259, | |
| "eval_steps_per_second": 2.909, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 36.41, | |
| "learning_rate": 2.791742286751361e-05, | |
| "loss": 0.1808, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 36.41, | |
| "eval_cer": 0.10908697497524156, | |
| "eval_loss": 0.516022801399231, | |
| "eval_runtime": 198.8874, | |
| "eval_samples_per_second": 23.159, | |
| "eval_steps_per_second": 2.896, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 39.01, | |
| "learning_rate": 7.527223230490017e-06, | |
| "loss": 0.17, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 39.01, | |
| "eval_cer": 0.10736428624928221, | |
| "eval_loss": 0.5128792524337769, | |
| "eval_runtime": 200.4673, | |
| "eval_samples_per_second": 22.976, | |
| "eval_steps_per_second": 2.873, | |
| "step": 22500 | |
| } | |
| ], | |
| "logging_steps": 1500, | |
| "max_steps": 23040, | |
| "num_train_epochs": 40, | |
| "save_steps": 1500, | |
| "total_flos": 1.6536314253048755e+20, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |