| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.669540229885057, | |
| "global_step": 3250, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 4.94e-05, | |
| "loss": 7.315, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "eval_cer": 1.0, | |
| "eval_loss": 3.364396810531616, | |
| "eval_runtime": 411.3291, | |
| "eval_samples_per_second": 18.839, | |
| "eval_steps_per_second": 2.356, | |
| "eval_wer": 1.0, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 9.94e-05, | |
| "loss": 3.2126, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "eval_cer": 1.0, | |
| "eval_loss": 3.178248405456543, | |
| "eval_runtime": 409.0228, | |
| "eval_samples_per_second": 18.945, | |
| "eval_steps_per_second": 2.369, | |
| "eval_wer": 1.0, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "learning_rate": 9.992892086330936e-05, | |
| "loss": 3.1414, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "eval_cer": 1.0, | |
| "eval_loss": 3.1066977977752686, | |
| "eval_runtime": 414.126, | |
| "eval_samples_per_second": 18.712, | |
| "eval_steps_per_second": 2.34, | |
| "eval_wer": 1.0, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "learning_rate": 9.985697841726619e-05, | |
| "loss": 1.9013, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "eval_cer": 0.12787848954973915, | |
| "eval_loss": 0.5506780743598938, | |
| "eval_runtime": 412.7461, | |
| "eval_samples_per_second": 18.774, | |
| "eval_steps_per_second": 2.348, | |
| "eval_wer": 0.5514430132072395, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "learning_rate": 9.978503597122302e-05, | |
| "loss": 0.5632, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "eval_cer": 0.06747688721413134, | |
| "eval_loss": 0.28180211782455444, | |
| "eval_runtime": 412.6671, | |
| "eval_samples_per_second": 18.778, | |
| "eval_steps_per_second": 2.348, | |
| "eval_wer": 0.30760729749805243, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "learning_rate": 9.971309352517985e-05, | |
| "loss": 0.388, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "eval_cer": 0.053711059057835714, | |
| "eval_loss": 0.2308780699968338, | |
| "eval_runtime": 524.2327, | |
| "eval_samples_per_second": 14.782, | |
| "eval_steps_per_second": 1.848, | |
| "eval_wer": 0.24573799300686633, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "learning_rate": 9.96411510791367e-05, | |
| "loss": 0.3175, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "eval_cer": 0.0491677133901393, | |
| "eval_loss": 0.2099951058626175, | |
| "eval_runtime": 529.8067, | |
| "eval_samples_per_second": 14.626, | |
| "eval_steps_per_second": 1.829, | |
| "eval_wer": 0.22468612424588294, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "learning_rate": 9.956920863309353e-05, | |
| "loss": 0.294, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "eval_cer": 0.046197281913749985, | |
| "eval_loss": 0.19443732500076294, | |
| "eval_runtime": 533.3896, | |
| "eval_samples_per_second": 14.528, | |
| "eval_steps_per_second": 1.817, | |
| "eval_wer": 0.21001141366378606, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "learning_rate": 9.949726618705036e-05, | |
| "loss": 0.2543, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "eval_cer": 0.04413495377442826, | |
| "eval_loss": 0.19428253173828125, | |
| "eval_runtime": 534.1775, | |
| "eval_samples_per_second": 14.506, | |
| "eval_steps_per_second": 1.814, | |
| "eval_wer": 0.19475696142906318, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 3.59, | |
| "learning_rate": 9.94253237410072e-05, | |
| "loss": 0.2493, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.59, | |
| "eval_cer": 0.04232723404736848, | |
| "eval_loss": 0.1892683357000351, | |
| "eval_runtime": 528.5117, | |
| "eval_samples_per_second": 14.662, | |
| "eval_steps_per_second": 1.833, | |
| "eval_wer": 0.18792688008406255, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.95, | |
| "learning_rate": 9.935338129496404e-05, | |
| "loss": 0.2333, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 3.95, | |
| "eval_cer": 0.04126919464530219, | |
| "eval_loss": 0.1779049038887024, | |
| "eval_runtime": 417.6945, | |
| "eval_samples_per_second": 18.552, | |
| "eval_steps_per_second": 2.32, | |
| "eval_wer": 0.18323459608312045, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 4.31, | |
| "learning_rate": 9.928143884892087e-05, | |
| "loss": 0.2131, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.31, | |
| "eval_cer": 0.040952348621153996, | |
| "eval_loss": 0.18109484016895294, | |
| "eval_runtime": 409.5311, | |
| "eval_samples_per_second": 18.922, | |
| "eval_steps_per_second": 2.366, | |
| "eval_wer": 0.17970179538743047, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.67, | |
| "learning_rate": 9.92094964028777e-05, | |
| "loss": 0.1992, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 4.67, | |
| "eval_cer": 0.03845435719862851, | |
| "eval_loss": 0.17126505076885223, | |
| "eval_runtime": 422.692, | |
| "eval_samples_per_second": 18.332, | |
| "eval_steps_per_second": 2.292, | |
| "eval_wer": 0.169665018026342, | |
| "step": 3250 | |
| } | |
| ], | |
| "max_steps": 348000, | |
| "num_train_epochs": 500, | |
| "total_flos": 1.643409928882224e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |