| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 30.0, | |
| "global_step": 10440, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.44, | |
| "learning_rate": 0.0003, | |
| "loss": NaN, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "eval_loss": 2.3965017795562744, | |
| "eval_runtime": 464.2782, | |
| "eval_samples_per_second": 9.996, | |
| "eval_wer": 0.9993478137582941, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "learning_rate": 0.00028490945674044263, | |
| "loss": NaN, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "eval_loss": 0.40979766845703125, | |
| "eval_runtime": 469.7691, | |
| "eval_samples_per_second": 9.879, | |
| "eval_wer": 0.4245448874269835, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.31, | |
| "learning_rate": 0.0002698189134808853, | |
| "loss": NaN, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.31, | |
| "eval_loss": 0.32825562357902527, | |
| "eval_runtime": 475.331, | |
| "eval_samples_per_second": 9.764, | |
| "eval_wer": 0.37052685305960414, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 5.75, | |
| "learning_rate": 0.00025472837022132795, | |
| "loss": NaN, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 5.75, | |
| "eval_loss": 0.30649319291114807, | |
| "eval_runtime": 474.8166, | |
| "eval_samples_per_second": 9.774, | |
| "eval_wer": 0.35002552033119716, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 7.18, | |
| "learning_rate": 0.0002396378269617706, | |
| "loss": NaN, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 7.18, | |
| "eval_loss": 0.3162720799446106, | |
| "eval_runtime": 475.8373, | |
| "eval_samples_per_second": 9.753, | |
| "eval_wer": 0.3463392502693813, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 8.62, | |
| "learning_rate": 0.00022454728370221327, | |
| "loss": NaN, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 8.62, | |
| "eval_loss": 0.30734783411026, | |
| "eval_runtime": 475.1293, | |
| "eval_samples_per_second": 9.768, | |
| "eval_wer": 0.3295808994498951, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 10.06, | |
| "learning_rate": 0.00020945674044265593, | |
| "loss": NaN, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 10.06, | |
| "eval_loss": 0.32298463582992554, | |
| "eval_runtime": 478.2831, | |
| "eval_samples_per_second": 9.703, | |
| "eval_wer": 0.32586627346452673, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 11.49, | |
| "learning_rate": 0.0001943661971830986, | |
| "loss": NaN, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 11.49, | |
| "eval_loss": 0.3256703317165375, | |
| "eval_runtime": 476.0024, | |
| "eval_samples_per_second": 9.75, | |
| "eval_wer": 0.3196279702829921, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 12.93, | |
| "learning_rate": 0.00017927565392354125, | |
| "loss": NaN, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 12.93, | |
| "eval_loss": 0.32864469289779663, | |
| "eval_runtime": 473.6083, | |
| "eval_samples_per_second": 9.799, | |
| "eval_wer": 0.32019508875404074, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 14.37, | |
| "learning_rate": 0.0001641851106639839, | |
| "loss": NaN, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 14.37, | |
| "eval_loss": 0.32181525230407715, | |
| "eval_runtime": 474.4211, | |
| "eval_samples_per_second": 9.782, | |
| "eval_wer": 0.31662224238643455, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 15.8, | |
| "learning_rate": 0.00014909456740442654, | |
| "loss": NaN, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 15.8, | |
| "eval_loss": 0.33631598949432373, | |
| "eval_runtime": 477.6284, | |
| "eval_samples_per_second": 9.717, | |
| "eval_wer": 0.3121136505415981, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 17.24, | |
| "learning_rate": 0.0001340040241448692, | |
| "loss": NaN, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 17.24, | |
| "eval_loss": 0.3525121510028839, | |
| "eval_runtime": 482.4872, | |
| "eval_samples_per_second": 9.619, | |
| "eval_wer": 0.30533658481256737, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 18.68, | |
| "learning_rate": 0.00011891348088531185, | |
| "loss": NaN, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 18.68, | |
| "eval_loss": 0.34829291701316833, | |
| "eval_runtime": 477.1506, | |
| "eval_samples_per_second": 9.726, | |
| "eval_wer": 0.3055634322009868, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 20.11, | |
| "learning_rate": 0.00010382293762575451, | |
| "loss": NaN, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 20.11, | |
| "eval_loss": 0.35680249333381653, | |
| "eval_runtime": 474.1059, | |
| "eval_samples_per_second": 9.789, | |
| "eval_wer": 0.30780355016162875, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 21.55, | |
| "learning_rate": 8.873239436619717e-05, | |
| "loss": NaN, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 21.55, | |
| "eval_loss": 0.36808809638023376, | |
| "eval_runtime": 477.9995, | |
| "eval_samples_per_second": 9.709, | |
| "eval_wer": 0.3064424658311121, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 22.99, | |
| "learning_rate": 7.364185110663983e-05, | |
| "loss": NaN, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 22.99, | |
| "eval_loss": 0.36075320839881897, | |
| "eval_runtime": 479.4594, | |
| "eval_samples_per_second": 9.68, | |
| "eval_wer": 0.29960868825497644, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 24.43, | |
| "learning_rate": 5.855130784708249e-05, | |
| "loss": NaN, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 24.43, | |
| "eval_loss": 0.36022624373435974, | |
| "eval_runtime": 482.5145, | |
| "eval_samples_per_second": 9.618, | |
| "eval_wer": 0.29745363806499175, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 25.86, | |
| "learning_rate": 4.346076458752515e-05, | |
| "loss": NaN, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 25.86, | |
| "eval_loss": 0.35646718740463257, | |
| "eval_runtime": 476.5808, | |
| "eval_samples_per_second": 9.738, | |
| "eval_wer": 0.2930301139908127, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 27.3, | |
| "learning_rate": 2.8370221327967802e-05, | |
| "loss": NaN, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 27.3, | |
| "eval_loss": 0.3554363548755646, | |
| "eval_runtime": 478.8286, | |
| "eval_samples_per_second": 9.692, | |
| "eval_wer": 0.29130040265411444, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 28.74, | |
| "learning_rate": 1.3279678068410461e-05, | |
| "loss": NaN, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 28.74, | |
| "eval_loss": 0.3562660217285156, | |
| "eval_runtime": 478.4327, | |
| "eval_samples_per_second": 9.7, | |
| "eval_wer": 0.29016616571201725, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "step": 10440, | |
| "total_flos": 4.532712593948488e+19, | |
| "train_runtime": 43862.7197, | |
| "train_samples_per_second": 0.238 | |
| } | |
| ], | |
| "max_steps": 10440, | |
| "num_train_epochs": 30, | |
| "total_flos": 4.532712593948488e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |