| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 38.46153846153846, | |
| "global_step": 5500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 3.5, | |
| "learning_rate": 4.9800000000000004e-05, | |
| "loss": 4.6717, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "eval_cer": 0.9966954216782684, | |
| "eval_loss": 3.3382508754730225, | |
| "eval_runtime": 157.3429, | |
| "eval_samples_per_second": 7.258, | |
| "eval_steps_per_second": 0.909, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 6.99, | |
| "learning_rate": 9.98e-05, | |
| "loss": 3.1712, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 6.99, | |
| "eval_cer": 0.9966954216782684, | |
| "eval_loss": 3.281019449234009, | |
| "eval_runtime": 163.1426, | |
| "eval_samples_per_second": 7.0, | |
| "eval_steps_per_second": 0.877, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 10.49, | |
| "learning_rate": 9.190243902439025e-05, | |
| "loss": 3.1588, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 10.49, | |
| "eval_cer": 0.9966954216782684, | |
| "eval_loss": 3.2297120094299316, | |
| "eval_runtime": 161.5142, | |
| "eval_samples_per_second": 7.071, | |
| "eval_steps_per_second": 0.885, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 13.99, | |
| "learning_rate": 8.377235772357724e-05, | |
| "loss": 2.9826, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 13.99, | |
| "eval_cer": 0.2508371402935709, | |
| "eval_loss": 1.0106703042984009, | |
| "eval_runtime": 160.9136, | |
| "eval_samples_per_second": 7.097, | |
| "eval_steps_per_second": 0.889, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 17.48, | |
| "learning_rate": 7.564227642276424e-05, | |
| "loss": 0.576, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 17.48, | |
| "eval_cer": 0.048671424633916056, | |
| "eval_loss": 0.23353584110736847, | |
| "eval_runtime": 161.7107, | |
| "eval_samples_per_second": 7.062, | |
| "eval_steps_per_second": 0.884, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 20.98, | |
| "learning_rate": 6.751219512195123e-05, | |
| "loss": 0.251, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 20.98, | |
| "eval_cer": 0.031453603954938104, | |
| "eval_loss": 0.15631325542926788, | |
| "eval_runtime": 162.3596, | |
| "eval_samples_per_second": 7.034, | |
| "eval_steps_per_second": 0.881, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 24.48, | |
| "learning_rate": 5.938211382113822e-05, | |
| "loss": 0.1958, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 24.48, | |
| "eval_cer": 0.026911641381413432, | |
| "eval_loss": 0.14399459958076477, | |
| "eval_runtime": 161.0224, | |
| "eval_samples_per_second": 7.092, | |
| "eval_steps_per_second": 0.888, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 27.97, | |
| "learning_rate": 5.125203252032521e-05, | |
| "loss": 0.1513, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 27.97, | |
| "eval_cer": 0.024079145677072034, | |
| "eval_loss": 0.13698184490203857, | |
| "eval_runtime": 157.5139, | |
| "eval_samples_per_second": 7.25, | |
| "eval_steps_per_second": 0.908, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 31.47, | |
| "learning_rate": 4.31219512195122e-05, | |
| "loss": 0.1178, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 31.47, | |
| "eval_cer": 0.022228933679722732, | |
| "eval_loss": 0.13266856968402863, | |
| "eval_runtime": 158.1185, | |
| "eval_samples_per_second": 7.222, | |
| "eval_steps_per_second": 0.904, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 34.97, | |
| "learning_rate": 3.4991869918699187e-05, | |
| "loss": 0.1008, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 34.97, | |
| "eval_cer": 0.02185068114777652, | |
| "eval_loss": 0.134142205119133, | |
| "eval_runtime": 161.0792, | |
| "eval_samples_per_second": 7.09, | |
| "eval_steps_per_second": 0.888, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 38.46, | |
| "learning_rate": 2.686178861788618e-05, | |
| "loss": 0.0933, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 38.46, | |
| "eval_cer": 0.020815618017722158, | |
| "eval_loss": 0.13403166830539703, | |
| "eval_runtime": 159.8014, | |
| "eval_samples_per_second": 7.146, | |
| "eval_steps_per_second": 0.895, | |
| "step": 5500 | |
| } | |
| ], | |
| "max_steps": 7150, | |
| "num_train_epochs": 50, | |
| "total_flos": 2.045078058974218e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |