| { | |
| "best_metric": 16.632698616044568, | |
| "best_model_checkpoint": "./checkpoint-4000", | |
| "epoch": 7.1138, | |
| "global_step": 5000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 9.200000000000001e-07, | |
| "loss": 1.1412, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 1.9200000000000003e-06, | |
| "loss": 0.5256, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 2.92e-06, | |
| "loss": 0.1965, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 3.920000000000001e-06, | |
| "loss": 0.1692, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 4.92e-06, | |
| "loss": 0.1573, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 5.92e-06, | |
| "loss": 0.1267, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 6.92e-06, | |
| "loss": 0.1248, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 7.92e-06, | |
| "loss": 0.1412, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "eval_cer": 5.2393487430695105, | |
| "eval_loss": 0.2656303644180298, | |
| "eval_runtime": 3483.4243, | |
| "eval_samples_per_second": 2.912, | |
| "eval_steps_per_second": 0.182, | |
| "eval_wer": 19.8334595003785, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 8.920000000000001e-06, | |
| "loss": 0.1807, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 9.920000000000002e-06, | |
| "loss": 0.1606, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 9.89777777777778e-06, | |
| "loss": 0.1767, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 9.786666666666667e-06, | |
| "loss": 0.1367, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 9.675555555555555e-06, | |
| "loss": 0.1498, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 9.564444444444445e-06, | |
| "loss": 0.1635, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "learning_rate": 9.453333333333335e-06, | |
| "loss": 0.1518, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "learning_rate": 9.342222222222223e-06, | |
| "loss": 0.0851, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "eval_cer": 4.891634975507227, | |
| "eval_loss": 0.23817911744117737, | |
| "eval_runtime": 3529.8171, | |
| "eval_samples_per_second": 2.874, | |
| "eval_steps_per_second": 0.18, | |
| "eval_wer": 18.63001999262408, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "learning_rate": 9.231111111111111e-06, | |
| "loss": 0.0817, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "learning_rate": 9.12e-06, | |
| "loss": 0.0717, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "learning_rate": 9.008888888888889e-06, | |
| "loss": 0.0514, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 8.897777777777779e-06, | |
| "loss": 0.0455, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "learning_rate": 8.786666666666668e-06, | |
| "loss": 0.0604, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "learning_rate": 8.675555555555556e-06, | |
| "loss": 0.0633, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "learning_rate": 8.564444444444445e-06, | |
| "loss": 0.0533, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "learning_rate": 8.453333333333334e-06, | |
| "loss": 0.0525, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "eval_cer": 5.223779469895081, | |
| "eval_loss": 0.253223180770874, | |
| "eval_runtime": 3540.1304, | |
| "eval_samples_per_second": 2.865, | |
| "eval_steps_per_second": 0.179, | |
| "eval_wer": 19.169626739649452, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "learning_rate": 8.342222222222222e-06, | |
| "loss": 0.0427, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "learning_rate": 8.231111111111112e-06, | |
| "loss": 0.0513, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "learning_rate": 8.120000000000002e-06, | |
| "loss": 0.058, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "learning_rate": 8.00888888888889e-06, | |
| "loss": 0.0368, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "learning_rate": 7.897777777777778e-06, | |
| "loss": 0.0268, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "learning_rate": 7.786666666666666e-06, | |
| "loss": 0.0265, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "learning_rate": 7.675555555555556e-06, | |
| "loss": 0.0202, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "learning_rate": 7.564444444444446e-06, | |
| "loss": 0.0163, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "eval_cer": 4.583997670375421, | |
| "eval_loss": 0.24465897679328918, | |
| "eval_runtime": 3488.6047, | |
| "eval_samples_per_second": 2.907, | |
| "eval_steps_per_second": 0.182, | |
| "eval_wer": 17.201420835031737, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "learning_rate": 7.453333333333334e-06, | |
| "loss": 0.0171, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "learning_rate": 7.342222222222223e-06, | |
| "loss": 0.0244, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "learning_rate": 7.231111111111112e-06, | |
| "loss": 0.0202, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "learning_rate": 7.1200000000000004e-06, | |
| "loss": 0.0171, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "learning_rate": 7.008888888888889e-06, | |
| "loss": 0.0175, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 6.897777777777779e-06, | |
| "loss": 0.0135, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "learning_rate": 6.786666666666667e-06, | |
| "loss": 0.0219, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "learning_rate": 6.675555555555556e-06, | |
| "loss": 0.0202, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "eval_cer": 4.493465230064843, | |
| "eval_loss": 0.2471683919429779, | |
| "eval_runtime": 3536.1026, | |
| "eval_samples_per_second": 2.868, | |
| "eval_steps_per_second": 0.179, | |
| "eval_wer": 17.10631029328985, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "learning_rate": 6.564444444444446e-06, | |
| "loss": 0.0114, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "learning_rate": 6.453333333333334e-06, | |
| "loss": 0.01, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.05, | |
| "learning_rate": 6.342222222222223e-06, | |
| "loss": 0.0101, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 3.06, | |
| "learning_rate": 6.231111111111111e-06, | |
| "loss": 0.008, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.07, | |
| "learning_rate": 6.120000000000001e-06, | |
| "loss": 0.005, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 3.08, | |
| "learning_rate": 6.00888888888889e-06, | |
| "loss": 0.0083, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.09, | |
| "learning_rate": 5.897777777777778e-06, | |
| "loss": 0.0096, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "learning_rate": 5.7866666666666674e-06, | |
| "loss": 0.0075, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "eval_cer": 4.4317647771143225, | |
| "eval_loss": 0.25027787685394287, | |
| "eval_runtime": 3586.7167, | |
| "eval_samples_per_second": 2.828, | |
| "eval_steps_per_second": 0.177, | |
| "eval_wer": 17.01508181447621, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.11, | |
| "learning_rate": 5.675555555555556e-06, | |
| "loss": 0.0068, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "learning_rate": 5.5644444444444444e-06, | |
| "loss": 0.0051, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 5.453333333333334e-06, | |
| "loss": 0.0072, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 4.01, | |
| "learning_rate": 5.342222222222223e-06, | |
| "loss": 0.0077, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "learning_rate": 5.231111111111111e-06, | |
| "loss": 0.0062, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 4.03, | |
| "learning_rate": 5.12e-06, | |
| "loss": 0.0053, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 4.04, | |
| "learning_rate": 5.00888888888889e-06, | |
| "loss": 0.0041, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 4.05, | |
| "learning_rate": 4.897777777777778e-06, | |
| "loss": 0.0039, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 4.05, | |
| "eval_cer": 4.365451206186191, | |
| "eval_loss": 0.2513594627380371, | |
| "eval_runtime": 3531.9344, | |
| "eval_samples_per_second": 2.872, | |
| "eval_steps_per_second": 0.18, | |
| "eval_wer": 16.74333740949941, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 4.06, | |
| "learning_rate": 4.786666666666667e-06, | |
| "loss": 0.0033, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 4.07, | |
| "learning_rate": 4.675555555555556e-06, | |
| "loss": 0.0026, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 4.08, | |
| "learning_rate": 4.564444444444445e-06, | |
| "loss": 0.0035, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 4.09, | |
| "learning_rate": 4.453333333333334e-06, | |
| "loss": 0.0033, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.1, | |
| "learning_rate": 4.3422222222222225e-06, | |
| "loss": 0.0028, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 4.11, | |
| "learning_rate": 4.2311111111111114e-06, | |
| "loss": 0.0032, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 4.12, | |
| "learning_rate": 4.12e-06, | |
| "loss": 0.0023, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 5.01, | |
| "learning_rate": 4.008888888888889e-06, | |
| "loss": 0.0038, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 5.01, | |
| "eval_cer": 4.358243209346177, | |
| "eval_loss": 0.2565072774887085, | |
| "eval_runtime": 3521.7324, | |
| "eval_samples_per_second": 2.88, | |
| "eval_steps_per_second": 0.18, | |
| "eval_wer": 16.88697373784429, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 5.02, | |
| "learning_rate": 3.897777777777778e-06, | |
| "loss": 0.0037, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 5.03, | |
| "learning_rate": 3.7866666666666667e-06, | |
| "loss": 0.0025, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 5.04, | |
| "learning_rate": 3.675555555555556e-06, | |
| "loss": 0.002, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 5.05, | |
| "learning_rate": 3.564444444444445e-06, | |
| "loss": 0.0023, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 5.06, | |
| "learning_rate": 3.4533333333333334e-06, | |
| "loss": 0.0029, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 5.07, | |
| "learning_rate": 3.3422222222222224e-06, | |
| "loss": 0.0016, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 5.08, | |
| "learning_rate": 3.2311111111111117e-06, | |
| "loss": 0.0017, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 5.09, | |
| "learning_rate": 3.12e-06, | |
| "loss": 0.0023, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 5.09, | |
| "eval_cer": 4.333736020090129, | |
| "eval_loss": 0.2590169310569763, | |
| "eval_runtime": 3522.239, | |
| "eval_samples_per_second": 2.88, | |
| "eval_steps_per_second": 0.18, | |
| "eval_wer": 16.698693685824647, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "learning_rate": 3.008888888888889e-06, | |
| "loss": 0.0018, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 5.11, | |
| "learning_rate": 2.8977777777777785e-06, | |
| "loss": 0.0017, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 5.12, | |
| "learning_rate": 2.786666666666667e-06, | |
| "loss": 0.0016, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 2.675555555555556e-06, | |
| "loss": 0.0014, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 6.01, | |
| "learning_rate": 2.5644444444444444e-06, | |
| "loss": 0.0024, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 6.02, | |
| "learning_rate": 2.4533333333333333e-06, | |
| "loss": 0.002, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 6.03, | |
| "learning_rate": 2.342222222222222e-06, | |
| "loss": 0.0014, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 6.04, | |
| "learning_rate": 2.2311111111111115e-06, | |
| "loss": 0.0013, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 6.04, | |
| "eval_cer": 4.285298281325233, | |
| "eval_loss": 0.2575719356536865, | |
| "eval_runtime": 3539.1562, | |
| "eval_samples_per_second": 2.866, | |
| "eval_steps_per_second": 0.179, | |
| "eval_wer": 16.632698616044568, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 6.05, | |
| "learning_rate": 2.12e-06, | |
| "loss": 0.0013, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 6.06, | |
| "learning_rate": 2.008888888888889e-06, | |
| "loss": 0.0011, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 6.07, | |
| "learning_rate": 1.8977777777777779e-06, | |
| "loss": 0.0011, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 6.08, | |
| "learning_rate": 1.7866666666666668e-06, | |
| "loss": 0.0013, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 6.09, | |
| "learning_rate": 1.675555555555556e-06, | |
| "loss": 0.0016, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 6.1, | |
| "learning_rate": 1.5644444444444446e-06, | |
| "loss": 0.0013, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 6.11, | |
| "learning_rate": 1.4533333333333335e-06, | |
| "loss": 0.0014, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 6.12, | |
| "learning_rate": 1.3422222222222222e-06, | |
| "loss": 0.0011, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 6.12, | |
| "eval_cer": 4.355648330483771, | |
| "eval_loss": 0.26468709111213684, | |
| "eval_runtime": 3546.9734, | |
| "eval_samples_per_second": 2.86, | |
| "eval_steps_per_second": 0.179, | |
| "eval_wer": 16.91220714687785, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "learning_rate": 1.2311111111111112e-06, | |
| "loss": 0.0013, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 7.01, | |
| "learning_rate": 1.12e-06, | |
| "loss": 0.0016, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 7.02, | |
| "learning_rate": 1.008888888888889e-06, | |
| "loss": 0.0013, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 7.03, | |
| "learning_rate": 8.977777777777778e-07, | |
| "loss": 0.0011, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 7.04, | |
| "learning_rate": 7.866666666666667e-07, | |
| "loss": 0.0011, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 7.05, | |
| "learning_rate": 6.755555555555555e-07, | |
| "loss": 0.001, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 7.06, | |
| "learning_rate": 5.644444444444445e-07, | |
| "loss": 0.001, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 7.07, | |
| "learning_rate": 4.533333333333334e-07, | |
| "loss": 0.001, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 7.07, | |
| "eval_cer": 4.28385668195723, | |
| "eval_loss": 0.26154619455337524, | |
| "eval_runtime": 3542.9557, | |
| "eval_samples_per_second": 2.863, | |
| "eval_steps_per_second": 0.179, | |
| "eval_wer": 16.634639647508685, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 7.08, | |
| "learning_rate": 3.422222222222223e-07, | |
| "loss": 0.0012, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 7.09, | |
| "learning_rate": 2.3111111111111112e-07, | |
| "loss": 0.0011, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 7.1, | |
| "learning_rate": 1.2000000000000002e-07, | |
| "loss": 0.0012, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 7.11, | |
| "learning_rate": 8.88888888888889e-09, | |
| "loss": 0.0012, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 7.11, | |
| "step": 5000, | |
| "total_flos": 4.616962385559552e+19, | |
| "train_loss": 0.04895092924162745, | |
| "train_runtime": 68546.6851, | |
| "train_samples_per_second": 2.334, | |
| "train_steps_per_second": 0.073 | |
| } | |
| ], | |
| "max_steps": 5000, | |
| "num_train_epochs": 9223372036854775807, | |
| "total_flos": 4.616962385559552e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |