{ "best_metric": null, "best_model_checkpoint": null, "epoch": 12.064, "global_step": 1000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.03, "learning_rate": 6.611096473669596e-06, "loss": 1.6963, "step": 25 }, { "epoch": 0.05, "learning_rate": 8.31378915840787e-06, "loss": 0.7824, "step": 50 }, { "epoch": 0.07, "learning_rate": 9.256291743595376e-06, "loss": 0.576, "step": 75 }, { "epoch": 1.02, "learning_rate": 9.888618026444238e-06, "loss": 0.4233, "step": 100 }, { "epoch": 1.05, "learning_rate": 9.800000000000001e-06, "loss": 0.3407, "step": 125 }, { "epoch": 1.07, "learning_rate": 9.522222222222223e-06, "loss": 0.286, "step": 150 }, { "epoch": 2.02, "learning_rate": 9.244444444444445e-06, "loss": 0.2156, "step": 175 }, { "epoch": 2.04, "learning_rate": 8.966666666666667e-06, "loss": 0.193, "step": 200 }, { "epoch": 2.07, "learning_rate": 8.68888888888889e-06, "loss": 0.1665, "step": 225 }, { "epoch": 3.02, "learning_rate": 8.411111111111112e-06, "loss": 0.124, "step": 250 }, { "epoch": 3.04, "learning_rate": 8.133333333333334e-06, "loss": 0.1063, "step": 275 }, { "epoch": 3.07, "learning_rate": 7.855555555555556e-06, "loss": 0.0879, "step": 300 }, { "epoch": 4.01, "learning_rate": 7.5777777777777785e-06, "loss": 0.0733, "step": 325 }, { "epoch": 4.04, "learning_rate": 7.3e-06, "loss": 0.0535, "step": 350 }, { "epoch": 4.06, "learning_rate": 7.022222222222222e-06, "loss": 0.0525, "step": 375 }, { "epoch": 5.01, "learning_rate": 6.744444444444444e-06, "loss": 0.0462, "step": 400 }, { "epoch": 5.04, "learning_rate": 6.466666666666667e-06, "loss": 0.0295, "step": 425 }, { "epoch": 5.06, "learning_rate": 6.18888888888889e-06, "loss": 0.0308, "step": 450 }, { "epoch": 6.01, "learning_rate": 5.911111111111112e-06, "loss": 0.0267, "step": 475 }, { "epoch": 6.03, "learning_rate": 5.633333333333334e-06, "loss": 0.0189, "step": 500 }, { "epoch": 6.06, "learning_rate": 5.355555555555556e-06, "loss": 0.0175, "step": 525 }, { "epoch": 7.0, "learning_rate": 5.077777777777778e-06, "loss": 0.0165, "step": 550 }, { "epoch": 7.03, "learning_rate": 4.800000000000001e-06, "loss": 0.0121, "step": 575 }, { "epoch": 7.05, "learning_rate": 4.5222222222222225e-06, "loss": 0.01, "step": 600 }, { "epoch": 8.0, "learning_rate": 4.244444444444445e-06, "loss": 0.0086, "step": 625 }, { "epoch": 8.03, "learning_rate": 3.966666666666667e-06, "loss": 0.0076, "step": 650 }, { "epoch": 8.05, "learning_rate": 3.688888888888889e-06, "loss": 0.0055, "step": 675 }, { "epoch": 8.08, "learning_rate": 3.4111111111111113e-06, "loss": 0.0053, "step": 700 }, { "epoch": 9.02, "learning_rate": 3.133333333333334e-06, "loss": 0.0035, "step": 725 }, { "epoch": 9.05, "learning_rate": 2.855555555555556e-06, "loss": 0.0026, "step": 750 }, { "epoch": 9.07, "learning_rate": 2.577777777777778e-06, "loss": 0.0024, "step": 775 }, { "epoch": 10.02, "learning_rate": 2.3000000000000004e-06, "loss": 0.0021, "step": 800 }, { "epoch": 10.04, "learning_rate": 2.0222222222222223e-06, "loss": 0.0018, "step": 825 }, { "epoch": 10.07, "learning_rate": 1.7444444444444448e-06, "loss": 0.0018, "step": 850 }, { "epoch": 11.02, "learning_rate": 1.4666666666666669e-06, "loss": 0.0016, "step": 875 }, { "epoch": 11.04, "learning_rate": 1.188888888888889e-06, "loss": 0.0016, "step": 900 }, { "epoch": 11.07, "learning_rate": 9.111111111111113e-07, "loss": 0.0016, "step": 925 }, { "epoch": 12.01, "learning_rate": 6.333333333333334e-07, "loss": 0.0015, "step": 950 }, { "epoch": 12.04, "learning_rate": 3.555555555555556e-07, "loss": 0.0014, "step": 975 }, { "epoch": 12.06, "learning_rate": 7.777777777777778e-08, "loss": 0.0014, "step": 1000 }, { "epoch": 12.06, "eval_loss": 0.82470703125, "eval_runtime": 163.5422, "eval_samples_per_second": 3.314, "eval_steps_per_second": 0.208, "eval_wer": 37.406890130353815, "step": 1000 }, { "epoch": 12.06, "step": 1000, "total_flos": 6.7788604353401586e+19, "train_loss": 0.13589781951904298, "train_runtime": 7855.0034, "train_samples_per_second": 4.074, "train_steps_per_second": 0.127 } ], "max_steps": 1000, "num_train_epochs": 9223372036854775807, "total_flos": 6.7788604353401586e+19, "trial_name": null, "trial_params": null }