{ "best_metric": 21.643241929604276, "best_model_checkpoint": "./checkpoint-2000", "epoch": 0.4, "global_step": 2000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.01, "learning_rate": 4.6000000000000004e-07, "loss": 1.4182, "step": 25 }, { "epoch": 0.01, "learning_rate": 9.400000000000001e-07, "loss": 1.292, "step": 50 }, { "epoch": 0.01, "learning_rate": 1.44e-06, "loss": 1.0018, "step": 75 }, { "epoch": 0.02, "learning_rate": 1.94e-06, "loss": 0.7765, "step": 100 }, { "epoch": 0.03, "learning_rate": 2.4400000000000004e-06, "loss": 0.7103, "step": 125 }, { "epoch": 0.03, "learning_rate": 2.9400000000000002e-06, "loss": 0.6597, "step": 150 }, { "epoch": 0.04, "learning_rate": 3.44e-06, "loss": 0.6657, "step": 175 }, { "epoch": 0.04, "learning_rate": 3.94e-06, "loss": 0.5853, "step": 200 }, { "epoch": 0.04, "learning_rate": 4.440000000000001e-06, "loss": 0.5273, "step": 225 }, { "epoch": 0.05, "learning_rate": 4.94e-06, "loss": 0.5979, "step": 250 }, { "epoch": 0.06, "learning_rate": 5.4400000000000004e-06, "loss": 0.5861, "step": 275 }, { "epoch": 0.06, "learning_rate": 5.94e-06, "loss": 0.5085, "step": 300 }, { "epoch": 0.07, "learning_rate": 6.440000000000001e-06, "loss": 0.4827, "step": 325 }, { "epoch": 0.07, "learning_rate": 6.9400000000000005e-06, "loss": 0.4909, "step": 350 }, { "epoch": 0.07, "learning_rate": 7.440000000000001e-06, "loss": 0.4651, "step": 375 }, { "epoch": 0.08, "learning_rate": 7.94e-06, "loss": 0.494, "step": 400 }, { "epoch": 0.09, "learning_rate": 8.44e-06, "loss": 0.4188, "step": 425 }, { "epoch": 0.09, "learning_rate": 8.94e-06, "loss": 0.3849, "step": 450 }, { "epoch": 0.1, "learning_rate": 9.440000000000001e-06, "loss": 0.4577, "step": 475 }, { "epoch": 0.1, "learning_rate": 9.940000000000001e-06, "loss": 0.4415, "step": 500 }, { "epoch": 0.1, "learning_rate": 9.951111111111111e-06, "loss": 0.4615, "step": 525 }, { "epoch": 0.11, "learning_rate": 9.895555555555557e-06, "loss": 0.4282, "step": 550 }, { "epoch": 0.12, "learning_rate": 9.842222222222223e-06, "loss": 0.4481, "step": 575 }, { "epoch": 0.12, "learning_rate": 9.786666666666667e-06, "loss": 0.4441, "step": 600 }, { "epoch": 0.12, "learning_rate": 9.731111111111113e-06, "loss": 0.4238, "step": 625 }, { "epoch": 0.13, "learning_rate": 9.675555555555555e-06, "loss": 0.4245, "step": 650 }, { "epoch": 0.14, "learning_rate": 9.620000000000001e-06, "loss": 0.4118, "step": 675 }, { "epoch": 0.14, "learning_rate": 9.564444444444445e-06, "loss": 0.4111, "step": 700 }, { "epoch": 0.14, "learning_rate": 9.508888888888889e-06, "loss": 0.3642, "step": 725 }, { "epoch": 0.15, "learning_rate": 9.453333333333335e-06, "loss": 0.401, "step": 750 }, { "epoch": 0.15, "learning_rate": 9.397777777777779e-06, "loss": 0.3855, "step": 775 }, { "epoch": 0.16, "learning_rate": 9.342222222222223e-06, "loss": 0.3668, "step": 800 }, { "epoch": 0.17, "learning_rate": 9.286666666666667e-06, "loss": 0.3794, "step": 825 }, { "epoch": 0.17, "learning_rate": 9.231111111111111e-06, "loss": 0.4296, "step": 850 }, { "epoch": 0.17, "learning_rate": 9.175555555555557e-06, "loss": 0.4003, "step": 875 }, { "epoch": 0.18, "learning_rate": 9.12e-06, "loss": 0.374, "step": 900 }, { "epoch": 0.18, "learning_rate": 9.064444444444447e-06, "loss": 0.4051, "step": 925 }, { "epoch": 0.19, "learning_rate": 9.008888888888889e-06, "loss": 0.3806, "step": 950 }, { "epoch": 0.2, "learning_rate": 8.953333333333335e-06, "loss": 0.4161, "step": 975 }, { "epoch": 0.2, "learning_rate": 8.897777777777779e-06, "loss": 0.4198, "step": 1000 }, { "epoch": 0.2, "eval_loss": 0.41016528010368347, "eval_runtime": 1814.0427, "eval_samples_per_second": 3.633, "eval_steps_per_second": 0.454, "eval_wer": 28.34865729677184, "step": 1000 }, { "epoch": 0.2, "learning_rate": 8.842222222222223e-06, "loss": 0.409, "step": 1025 }, { "epoch": 0.21, "learning_rate": 8.786666666666668e-06, "loss": 0.3674, "step": 1050 }, { "epoch": 0.21, "learning_rate": 8.73111111111111e-06, "loss": 0.3591, "step": 1075 }, { "epoch": 0.22, "learning_rate": 8.675555555555556e-06, "loss": 0.3892, "step": 1100 }, { "epoch": 0.23, "learning_rate": 8.62e-06, "loss": 0.3843, "step": 1125 }, { "epoch": 0.23, "learning_rate": 8.564444444444445e-06, "loss": 0.3605, "step": 1150 }, { "epoch": 0.23, "learning_rate": 8.50888888888889e-06, "loss": 0.326, "step": 1175 }, { "epoch": 0.24, "learning_rate": 8.453333333333334e-06, "loss": 0.3103, "step": 1200 }, { "epoch": 0.24, "learning_rate": 8.397777777777778e-06, "loss": 0.2766, "step": 1225 }, { "epoch": 0.25, "learning_rate": 8.342222222222222e-06, "loss": 0.3204, "step": 1250 }, { "epoch": 0.26, "learning_rate": 8.286666666666668e-06, "loss": 0.3426, "step": 1275 }, { "epoch": 0.26, "learning_rate": 8.231111111111112e-06, "loss": 0.3417, "step": 1300 }, { "epoch": 0.27, "learning_rate": 8.175555555555556e-06, "loss": 0.3179, "step": 1325 }, { "epoch": 0.27, "learning_rate": 8.120000000000002e-06, "loss": 0.2598, "step": 1350 }, { "epoch": 0.28, "learning_rate": 8.064444444444444e-06, "loss": 0.3453, "step": 1375 }, { "epoch": 0.28, "learning_rate": 8.00888888888889e-06, "loss": 0.2752, "step": 1400 }, { "epoch": 0.28, "learning_rate": 7.953333333333334e-06, "loss": 0.2927, "step": 1425 }, { "epoch": 0.29, "learning_rate": 7.897777777777778e-06, "loss": 0.3859, "step": 1450 }, { "epoch": 0.29, "learning_rate": 7.842222222222224e-06, "loss": 0.3137, "step": 1475 }, { "epoch": 0.3, "learning_rate": 7.786666666666666e-06, "loss": 0.2678, "step": 1500 }, { "epoch": 0.3, "learning_rate": 7.731111111111112e-06, "loss": 0.2803, "step": 1525 }, { "epoch": 0.31, "learning_rate": 7.675555555555556e-06, "loss": 0.2828, "step": 1550 }, { "epoch": 0.32, "learning_rate": 7.620000000000001e-06, "loss": 0.3655, "step": 1575 }, { "epoch": 0.32, "learning_rate": 7.564444444444446e-06, "loss": 0.3321, "step": 1600 }, { "epoch": 0.33, "learning_rate": 7.50888888888889e-06, "loss": 0.3649, "step": 1625 }, { "epoch": 0.33, "learning_rate": 7.453333333333334e-06, "loss": 0.3229, "step": 1650 }, { "epoch": 0.34, "learning_rate": 7.3977777777777786e-06, "loss": 0.3115, "step": 1675 }, { "epoch": 0.34, "learning_rate": 7.342222222222223e-06, "loss": 0.2925, "step": 1700 }, { "epoch": 0.34, "learning_rate": 7.2866666666666675e-06, "loss": 0.3014, "step": 1725 }, { "epoch": 0.35, "learning_rate": 7.231111111111112e-06, "loss": 0.3303, "step": 1750 }, { "epoch": 0.35, "learning_rate": 7.1755555555555556e-06, "loss": 0.3174, "step": 1775 }, { "epoch": 0.36, "learning_rate": 7.1200000000000004e-06, "loss": 0.3249, "step": 1800 }, { "epoch": 0.36, "learning_rate": 7.0644444444444445e-06, "loss": 0.2678, "step": 1825 }, { "epoch": 0.37, "learning_rate": 7.008888888888889e-06, "loss": 0.3088, "step": 1850 }, { "epoch": 0.38, "learning_rate": 6.953333333333334e-06, "loss": 0.2515, "step": 1875 }, { "epoch": 0.38, "learning_rate": 6.897777777777779e-06, "loss": 0.2838, "step": 1900 }, { "epoch": 0.39, "learning_rate": 6.842222222222222e-06, "loss": 0.2494, "step": 1925 }, { "epoch": 0.39, "learning_rate": 6.786666666666667e-06, "loss": 0.205, "step": 1950 }, { "epoch": 0.4, "learning_rate": 6.731111111111111e-06, "loss": 0.2439, "step": 1975 }, { "epoch": 0.4, "learning_rate": 6.675555555555556e-06, "loss": 0.2547, "step": 2000 }, { "epoch": 0.4, "eval_loss": 0.31417879462242126, "eval_runtime": 1807.9177, "eval_samples_per_second": 3.646, "eval_steps_per_second": 0.456, "eval_wer": 21.643241929604276, "step": 2000 } ], "max_steps": 5000, "num_train_epochs": 9223372036854775807, "total_flos": 8.16483926016e+18, "trial_name": null, "trial_params": null }