{ "best_metric": null, "best_model_checkpoint": null, "epoch": 2.4096385542168672, "global_step": 3000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0, "learning_rate": 0.0, "loss": 36.5036, "step": 1 }, { "epoch": 0.4, "learning_rate": 4.96e-05, "loss": 8.5484, "step": 500 }, { "epoch": 0.4, "eval_cer": 1.0, "eval_clean_cer": 1.0, "eval_loss": 3.6234467029571533, "eval_runtime": 384.6662, "eval_samples_per_second": 19.9, "eval_steps_per_second": 1.245, "eval_wer": 1.0, "learning_rate": 4.96e-05, "step": 500 }, { "epoch": 0.8, "learning_rate": 4.792887029288703e-05, "loss": 3.2275, "step": 1000 }, { "epoch": 0.8, "eval_cer": 0.7080718890168877, "eval_clean_cer": 0.653960461430081, "eval_loss": 2.2960469722747803, "eval_runtime": 383.0083, "eval_samples_per_second": 19.987, "eval_steps_per_second": 1.251, "eval_wer": 0.9997886281969985, "learning_rate": 4.792887029288703e-05, "step": 1000 }, { "epoch": 1.2, "learning_rate": 4.5841004184100424e-05, "loss": 0.9955, "step": 1500 }, { "epoch": 1.2, "eval_cer": 0.43273425819260275, "eval_clean_cer": 0.37557605377480335, "eval_loss": 1.2224403619766235, "eval_runtime": 382.1035, "eval_samples_per_second": 20.034, "eval_steps_per_second": 1.254, "eval_wer": 0.9549127684828382, "learning_rate": 4.5841004184100424e-05, "step": 1500 }, { "epoch": 1.61, "learning_rate": 4.37489539748954e-05, "loss": 0.66, "step": 2000 }, { "epoch": 1.61, "eval_cer": 0.36509166891345374, "eval_clean_cer": 0.304027973464835, "eval_loss": 0.9558963775634766, "eval_runtime": 380.573, "eval_samples_per_second": 20.114, "eval_steps_per_second": 1.259, "eval_wer": 0.9231907386631546, "learning_rate": 4.37489539748954e-05, "step": 2000 }, { "epoch": 2.01, "learning_rate": 4.1665271966527195e-05, "loss": 0.546, "step": 2500 }, { "epoch": 2.01, "eval_cer": 0.3585320718645065, "eval_clean_cer": 0.2825745567688153, "eval_loss": 0.920676589012146, "eval_runtime": 379.317, "eval_samples_per_second": 20.181, "eval_steps_per_second": 1.263, "eval_wer": 0.948051314570021, "learning_rate": 4.1665271966527195e-05, "step": 2500 }, { "epoch": 2.41, "learning_rate": 3.957740585774059e-05, "loss": 0.4459, "step": 3000 }, { "epoch": 2.41, "eval_cer": 0.29398698497512193, "eval_clean_cer": 0.23832115181401903, "eval_loss": 0.7700594067573547, "eval_runtime": 382.0036, "eval_samples_per_second": 20.039, "eval_steps_per_second": 1.254, "eval_wer": 0.869307188267239, "learning_rate": 3.957740585774059e-05, "step": 3000 } ], "max_steps": 12450, "num_train_epochs": 10, "total_flos": 3.2256636479360102e+19, "trial_name": null, "trial_params": null }