{ "best_metric": 60.07326007326007, "best_model_checkpoint": "./checkpoint-70", "epoch": 1.0, "global_step": 100, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.1, "learning_rate": 7e-05, "loss": 2.5622, "step": 10 }, { "epoch": 0.1, "eval_loss": 1.5401501655578613, "eval_runtime": 62.3188, "eval_samples_per_second": 1.027, "eval_steps_per_second": 0.032, "eval_wer": 94.5054945054945, "step": 10 }, { "epoch": 0.2, "learning_rate": 9.222222222222223e-05, "loss": 1.3719, "step": 20 }, { "epoch": 0.2, "eval_loss": 1.001212239265442, "eval_runtime": 18.9902, "eval_samples_per_second": 3.37, "eval_steps_per_second": 0.105, "eval_wer": 75.27472527472527, "step": 20 }, { "epoch": 0.3, "learning_rate": 8.111111111111112e-05, "loss": 0.9898, "step": 30 }, { "epoch": 0.3, "eval_loss": 0.8217034339904785, "eval_runtime": 17.7847, "eval_samples_per_second": 3.599, "eval_steps_per_second": 0.112, "eval_wer": 72.7106227106227, "step": 30 }, { "epoch": 0.4, "learning_rate": 7e-05, "loss": 0.9742, "step": 40 }, { "epoch": 0.4, "eval_loss": 0.7924289107322693, "eval_runtime": 19.063, "eval_samples_per_second": 3.357, "eval_steps_per_second": 0.105, "eval_wer": 72.52747252747253, "step": 40 }, { "epoch": 0.5, "learning_rate": 5.8888888888888896e-05, "loss": 0.6951, "step": 50 }, { "epoch": 0.5, "eval_loss": 0.7628086805343628, "eval_runtime": 17.8327, "eval_samples_per_second": 3.589, "eval_steps_per_second": 0.112, "eval_wer": 76.19047619047619, "step": 50 }, { "epoch": 0.6, "learning_rate": 4.7777777777777784e-05, "loss": 0.7824, "step": 60 }, { "epoch": 0.6, "eval_loss": 0.6737741231918335, "eval_runtime": 18.8876, "eval_samples_per_second": 3.388, "eval_steps_per_second": 0.106, "eval_wer": 65.38461538461539, "step": 60 }, { "epoch": 0.7, "learning_rate": 3.6666666666666666e-05, "loss": 0.6818, "step": 70 }, { "epoch": 0.7, "eval_loss": 0.6388838291168213, "eval_runtime": 18.9502, "eval_samples_per_second": 3.377, "eval_steps_per_second": 0.106, "eval_wer": 60.07326007326007, "step": 70 }, { "epoch": 0.8, "learning_rate": 2.5555555555555554e-05, "loss": 0.7823, "step": 80 }, { "epoch": 0.8, "eval_loss": 0.6208388805389404, "eval_runtime": 18.1881, "eval_samples_per_second": 3.519, "eval_steps_per_second": 0.11, "eval_wer": 65.75091575091575, "step": 80 }, { "epoch": 0.9, "learning_rate": 1.4444444444444444e-05, "loss": 0.5994, "step": 90 }, { "epoch": 0.9, "eval_loss": 0.5900620818138123, "eval_runtime": 17.489, "eval_samples_per_second": 3.659, "eval_steps_per_second": 0.114, "eval_wer": 61.904761904761905, "step": 90 }, { "epoch": 1.0, "learning_rate": 3.3333333333333333e-06, "loss": 0.6647, "step": 100 }, { "epoch": 1.0, "eval_loss": 0.5789934992790222, "eval_runtime": 18.4962, "eval_samples_per_second": 3.46, "eval_steps_per_second": 0.108, "eval_wer": 61.72161172161172, "step": 100 }, { "epoch": 1.0, "step": 100, "total_flos": 7.8780432384e+16, "train_loss": 1.0103698587417602, "train_runtime": 873.4716, "train_samples_per_second": 3.664, "train_steps_per_second": 0.114 } ], "max_steps": 100, "num_train_epochs": 9223372036854775807, "total_flos": 7.8780432384e+16, "trial_name": null, "trial_params": null }