{ "best_metric": 16.9811320754717, "best_model_checkpoint": "./checkpoint-500", "epoch": 10.06, "global_step": 500, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.05, "learning_rate": 1.25e-06, "loss": 2.0148, "step": 25 }, { "epoch": 1.01, "learning_rate": 2.5e-06, "loss": 1.0925, "step": 50 }, { "epoch": 1.06, "learning_rate": 3.7500000000000005e-06, "loss": 0.4214, "step": 75 }, { "epoch": 2.01, "learning_rate": 5e-06, "loss": 0.1978, "step": 100 }, { "epoch": 2.01, "eval_loss": 0.5473833680152893, "eval_runtime": 81.8567, "eval_samples_per_second": 0.684, "eval_steps_per_second": 0.342, "eval_wer": 21.069182389937108, "step": 100 }, { "epoch": 2.06, "learning_rate": 6.25e-06, "loss": 0.0536, "step": 125 }, { "epoch": 3.02, "learning_rate": 7.500000000000001e-06, "loss": 0.0218, "step": 150 }, { "epoch": 3.07, "learning_rate": 8.750000000000001e-06, "loss": 0.0099, "step": 175 }, { "epoch": 4.02, "learning_rate": 1e-05, "loss": 0.0087, "step": 200 }, { "epoch": 4.02, "eval_loss": 0.6202297210693359, "eval_runtime": 80.6319, "eval_samples_per_second": 0.695, "eval_steps_per_second": 0.347, "eval_wer": 19.49685534591195, "step": 200 }, { "epoch": 4.07, "learning_rate": 9.166666666666666e-06, "loss": 0.0134, "step": 225 }, { "epoch": 5.03, "learning_rate": 8.333333333333334e-06, "loss": 0.0069, "step": 250 }, { "epoch": 5.08, "learning_rate": 7.500000000000001e-06, "loss": 0.003, "step": 275 }, { "epoch": 6.04, "learning_rate": 6.666666666666667e-06, "loss": 0.0029, "step": 300 }, { "epoch": 6.04, "eval_loss": 0.6264036893844604, "eval_runtime": 79.2868, "eval_samples_per_second": 0.706, "eval_steps_per_second": 0.353, "eval_wer": 18.238993710691823, "step": 300 }, { "epoch": 6.09, "learning_rate": 5.833333333333334e-06, "loss": 0.0012, "step": 325 }, { "epoch": 7.04, "learning_rate": 5e-06, "loss": 0.001, "step": 350 }, { "epoch": 7.09, "learning_rate": 4.166666666666667e-06, "loss": 0.0005, "step": 375 }, { "epoch": 8.05, "learning_rate": 3.3333333333333333e-06, "loss": 0.0003, "step": 400 }, { "epoch": 8.05, "eval_loss": 0.6659052968025208, "eval_runtime": 82.053, "eval_samples_per_second": 0.682, "eval_steps_per_second": 0.341, "eval_wer": 17.19077568134172, "step": 400 }, { "epoch": 9.0, "learning_rate": 2.5e-06, "loss": 0.0003, "step": 425 }, { "epoch": 9.05, "learning_rate": 1.6666666666666667e-06, "loss": 0.0003, "step": 450 }, { "epoch": 10.01, "learning_rate": 8.333333333333333e-07, "loss": 0.0003, "step": 475 }, { "epoch": 10.06, "learning_rate": 0.0, "loss": 0.0002, "step": 500 }, { "epoch": 10.06, "eval_loss": 0.6749836206436157, "eval_runtime": 80.483, "eval_samples_per_second": 0.696, "eval_steps_per_second": 0.348, "eval_wer": 16.9811320754717, "step": 500 }, { "epoch": 10.06, "step": 500, "total_flos": 3.4098126594048e+19, "train_loss": 0.19253770965430886, "train_runtime": 5880.7795, "train_samples_per_second": 2.721, "train_steps_per_second": 0.085 } ], "max_steps": 500, "num_train_epochs": 9223372036854775807, "total_flos": 3.4098126594048e+19, "trial_name": null, "trial_params": null }