| { | |
| "best_metric": 0.16926079988479614, | |
| "best_model_checkpoint": "./checkpoint-2000", | |
| "epoch": 28.97666068222621, | |
| "global_step": 2000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.45, | |
| "learning_rate": 1.0202e-05, | |
| "loss": 6.1347, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "learning_rate": 2.0102e-05, | |
| "loss": 3.041, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 4.34, | |
| "learning_rate": 3.0002000000000002e-05, | |
| "loss": 1.8943, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 5.79, | |
| "learning_rate": 3.9902e-05, | |
| "loss": 1.473, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 7.24, | |
| "learning_rate": 4.9802e-05, | |
| "loss": 1.255, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 7.24, | |
| "eval_cer": 0.07577137076378351, | |
| "eval_loss": 0.29775065183639526, | |
| "eval_runtime": 14.4182, | |
| "eval_samples_per_second": 23.234, | |
| "eval_steps_per_second": 0.416, | |
| "eval_wer": 0.42935206869633097, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 8.69, | |
| "learning_rate": 5e-05, | |
| "loss": 1.1289, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 10.14, | |
| "learning_rate": 5e-05, | |
| "loss": 1.0964, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 11.59, | |
| "learning_rate": 5e-05, | |
| "loss": 1.0491, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 13.04, | |
| "learning_rate": 5e-05, | |
| "loss": 1.0342, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 14.49, | |
| "learning_rate": 5e-05, | |
| "loss": 1.0058, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 14.49, | |
| "eval_cer": 0.04825493171471927, | |
| "eval_loss": 0.18830506503582, | |
| "eval_runtime": 13.5348, | |
| "eval_samples_per_second": 24.751, | |
| "eval_steps_per_second": 0.443, | |
| "eval_wer": 0.28376268540202965, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 15.93, | |
| "learning_rate": 5e-05, | |
| "loss": 0.9735, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 17.39, | |
| "learning_rate": 5e-05, | |
| "loss": 0.9772, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 18.83, | |
| "learning_rate": 5e-05, | |
| "loss": 0.9625, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 20.29, | |
| "learning_rate": 5e-05, | |
| "loss": 0.9492, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 21.73, | |
| "learning_rate": 5e-05, | |
| "loss": 0.9371, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 21.73, | |
| "eval_cer": 0.045725847243297923, | |
| "eval_loss": 0.1812673956155777, | |
| "eval_runtime": 13.3792, | |
| "eval_samples_per_second": 25.039, | |
| "eval_steps_per_second": 0.448, | |
| "eval_wer": 0.26268540202966434, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 23.19, | |
| "learning_rate": 5e-05, | |
| "loss": 0.9325, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 24.63, | |
| "learning_rate": 5e-05, | |
| "loss": 0.9169, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 26.09, | |
| "learning_rate": 5e-05, | |
| "loss": 0.9168, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 27.53, | |
| "learning_rate": 5e-05, | |
| "loss": 0.9111, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 28.98, | |
| "learning_rate": 5e-05, | |
| "loss": 0.8999, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 28.98, | |
| "eval_cer": 0.042943854324734444, | |
| "eval_loss": 0.16926079988479614, | |
| "eval_runtime": 13.6397, | |
| "eval_samples_per_second": 24.561, | |
| "eval_steps_per_second": 0.44, | |
| "eval_wer": 0.2373145979703357, | |
| "step": 2000 | |
| } | |
| ], | |
| "max_steps": 5000, | |
| "num_train_epochs": 73, | |
| "total_flos": 1.8112953812816295e+20, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |