|
{ |
|
"best_metric": 12.361457422226865, |
|
"best_model_checkpoint": "./checkpoint-1000", |
|
"epoch": 13.012, |
|
"global_step": 3000, |
|
"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": 2.2525, |
|
"step": 25 |
|
}, |
|
{ |
|
"epoch": 0.02, |
|
"learning_rate": 9.400000000000001e-07, |
|
"loss": 1.9074, |
|
"step": 50 |
|
}, |
|
{ |
|
"epoch": 0.03, |
|
"learning_rate": 1.44e-06, |
|
"loss": 1.4407, |
|
"step": 75 |
|
}, |
|
{ |
|
"epoch": 0.03, |
|
"learning_rate": 1.94e-06, |
|
"loss": 0.9036, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 0.04, |
|
"learning_rate": 2.4400000000000004e-06, |
|
"loss": 1.0942, |
|
"step": 125 |
|
}, |
|
{ |
|
"epoch": 0.05, |
|
"learning_rate": 2.9400000000000002e-06, |
|
"loss": 1.1163, |
|
"step": 150 |
|
}, |
|
{ |
|
"epoch": 0.06, |
|
"learning_rate": 3.44e-06, |
|
"loss": 0.7416, |
|
"step": 175 |
|
}, |
|
{ |
|
"epoch": 0.07, |
|
"learning_rate": 3.94e-06, |
|
"loss": 0.7039, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 0.07, |
|
"learning_rate": 4.440000000000001e-06, |
|
"loss": 0.6301, |
|
"step": 225 |
|
}, |
|
{ |
|
"epoch": 1.01, |
|
"learning_rate": 4.94e-06, |
|
"loss": 0.7837, |
|
"step": 250 |
|
}, |
|
{ |
|
"epoch": 1.02, |
|
"learning_rate": 5.4400000000000004e-06, |
|
"loss": 0.5005, |
|
"step": 275 |
|
}, |
|
{ |
|
"epoch": 1.02, |
|
"learning_rate": 5.94e-06, |
|
"loss": 0.3661, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 1.03, |
|
"learning_rate": 6.440000000000001e-06, |
|
"loss": 0.3475, |
|
"step": 325 |
|
}, |
|
{ |
|
"epoch": 1.04, |
|
"learning_rate": 6.9400000000000005e-06, |
|
"loss": 0.2912, |
|
"step": 350 |
|
}, |
|
{ |
|
"epoch": 1.05, |
|
"learning_rate": 7.440000000000001e-06, |
|
"loss": 0.4989, |
|
"step": 375 |
|
}, |
|
{ |
|
"epoch": 1.06, |
|
"learning_rate": 7.94e-06, |
|
"loss": 0.2618, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 1.07, |
|
"learning_rate": 8.44e-06, |
|
"loss": 0.2507, |
|
"step": 425 |
|
}, |
|
{ |
|
"epoch": 1.07, |
|
"learning_rate": 8.94e-06, |
|
"loss": 0.3651, |
|
"step": 450 |
|
}, |
|
{ |
|
"epoch": 2.01, |
|
"learning_rate": 9.440000000000001e-06, |
|
"loss": 0.2203, |
|
"step": 475 |
|
}, |
|
{ |
|
"epoch": 2.01, |
|
"learning_rate": 9.940000000000001e-06, |
|
"loss": 0.1886, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 2.01, |
|
"eval_loss": 0.36212778091430664, |
|
"eval_runtime": 442.4584, |
|
"eval_samples_per_second": 5.784, |
|
"eval_steps_per_second": 0.181, |
|
"eval_wer": 11.997232412109502, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 2.02, |
|
"learning_rate": 9.912000000000001e-06, |
|
"loss": 0.1614, |
|
"step": 525 |
|
}, |
|
{ |
|
"epoch": 2.03, |
|
"learning_rate": 9.812e-06, |
|
"loss": 0.1866, |
|
"step": 550 |
|
}, |
|
{ |
|
"epoch": 2.04, |
|
"learning_rate": 9.712e-06, |
|
"loss": 0.2434, |
|
"step": 575 |
|
}, |
|
{ |
|
"epoch": 2.05, |
|
"learning_rate": 9.612000000000002e-06, |
|
"loss": 0.3595, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 2.06, |
|
"learning_rate": 9.512000000000001e-06, |
|
"loss": 0.3434, |
|
"step": 625 |
|
}, |
|
{ |
|
"epoch": 2.06, |
|
"learning_rate": 9.412e-06, |
|
"loss": 0.3429, |
|
"step": 650 |
|
}, |
|
{ |
|
"epoch": 2.07, |
|
"learning_rate": 9.312000000000002e-06, |
|
"loss": 0.2233, |
|
"step": 675 |
|
}, |
|
{ |
|
"epoch": 3.01, |
|
"learning_rate": 9.212000000000001e-06, |
|
"loss": 0.1943, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 3.01, |
|
"learning_rate": 9.112e-06, |
|
"loss": 0.255, |
|
"step": 725 |
|
}, |
|
{ |
|
"epoch": 3.02, |
|
"learning_rate": 9.012000000000001e-06, |
|
"loss": 0.1173, |
|
"step": 750 |
|
}, |
|
{ |
|
"epoch": 3.03, |
|
"learning_rate": 8.912e-06, |
|
"loss": 0.133, |
|
"step": 775 |
|
}, |
|
{ |
|
"epoch": 3.04, |
|
"learning_rate": 8.812000000000001e-06, |
|
"loss": 0.0991, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 3.05, |
|
"learning_rate": 8.712e-06, |
|
"loss": 0.1382, |
|
"step": 825 |
|
}, |
|
{ |
|
"epoch": 3.06, |
|
"learning_rate": 8.612e-06, |
|
"loss": 0.1949, |
|
"step": 850 |
|
}, |
|
{ |
|
"epoch": 3.06, |
|
"learning_rate": 8.512e-06, |
|
"loss": 0.272, |
|
"step": 875 |
|
}, |
|
{ |
|
"epoch": 3.07, |
|
"learning_rate": 8.412e-06, |
|
"loss": 0.1639, |
|
"step": 900 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"learning_rate": 8.312000000000001e-06, |
|
"loss": 0.1378, |
|
"step": 925 |
|
}, |
|
{ |
|
"epoch": 4.01, |
|
"learning_rate": 8.212e-06, |
|
"loss": 0.0636, |
|
"step": 950 |
|
}, |
|
{ |
|
"epoch": 4.02, |
|
"learning_rate": 8.112000000000001e-06, |
|
"loss": 0.099, |
|
"step": 975 |
|
}, |
|
{ |
|
"epoch": 4.03, |
|
"learning_rate": 8.012e-06, |
|
"loss": 0.187, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 4.03, |
|
"eval_loss": 0.3786308169364929, |
|
"eval_runtime": 447.1416, |
|
"eval_samples_per_second": 5.723, |
|
"eval_steps_per_second": 0.179, |
|
"eval_wer": 12.361457422226865, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 4.04, |
|
"learning_rate": 7.912000000000001e-06, |
|
"loss": 0.2418, |
|
"step": 1025 |
|
}, |
|
{ |
|
"epoch": 4.05, |
|
"learning_rate": 7.812e-06, |
|
"loss": 0.1562, |
|
"step": 1050 |
|
}, |
|
{ |
|
"epoch": 4.05, |
|
"learning_rate": 7.712e-06, |
|
"loss": 0.0962, |
|
"step": 1075 |
|
}, |
|
{ |
|
"epoch": 4.06, |
|
"learning_rate": 7.612e-06, |
|
"loss": 0.0856, |
|
"step": 1100 |
|
}, |
|
{ |
|
"epoch": 4.07, |
|
"learning_rate": 7.512e-06, |
|
"loss": 0.1057, |
|
"step": 1125 |
|
}, |
|
{ |
|
"epoch": 5.0, |
|
"learning_rate": 7.412e-06, |
|
"loss": 0.0839, |
|
"step": 1150 |
|
}, |
|
{ |
|
"epoch": 5.01, |
|
"learning_rate": 7.3120000000000005e-06, |
|
"loss": 0.0567, |
|
"step": 1175 |
|
}, |
|
{ |
|
"epoch": 5.02, |
|
"learning_rate": 7.212e-06, |
|
"loss": 0.1323, |
|
"step": 1200 |
|
}, |
|
{ |
|
"epoch": 5.03, |
|
"learning_rate": 7.1120000000000015e-06, |
|
"loss": 0.0973, |
|
"step": 1225 |
|
}, |
|
{ |
|
"epoch": 5.04, |
|
"learning_rate": 7.012000000000001e-06, |
|
"loss": 0.1396, |
|
"step": 1250 |
|
}, |
|
{ |
|
"epoch": 5.04, |
|
"learning_rate": 6.912000000000001e-06, |
|
"loss": 0.1016, |
|
"step": 1275 |
|
}, |
|
{ |
|
"epoch": 5.05, |
|
"learning_rate": 6.812000000000001e-06, |
|
"loss": 0.0617, |
|
"step": 1300 |
|
}, |
|
{ |
|
"epoch": 5.06, |
|
"learning_rate": 6.712000000000001e-06, |
|
"loss": 0.0431, |
|
"step": 1325 |
|
}, |
|
{ |
|
"epoch": 5.07, |
|
"learning_rate": 6.612e-06, |
|
"loss": 0.0489, |
|
"step": 1350 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"learning_rate": 6.5120000000000005e-06, |
|
"loss": 0.085, |
|
"step": 1375 |
|
}, |
|
{ |
|
"epoch": 6.01, |
|
"learning_rate": 6.412000000000001e-06, |
|
"loss": 0.0573, |
|
"step": 1400 |
|
}, |
|
{ |
|
"epoch": 6.02, |
|
"learning_rate": 6.312000000000001e-06, |
|
"loss": 0.0385, |
|
"step": 1425 |
|
}, |
|
{ |
|
"epoch": 6.03, |
|
"learning_rate": 6.212e-06, |
|
"loss": 0.0348, |
|
"step": 1450 |
|
}, |
|
{ |
|
"epoch": 6.04, |
|
"learning_rate": 6.112e-06, |
|
"loss": 0.0282, |
|
"step": 1475 |
|
}, |
|
{ |
|
"epoch": 6.04, |
|
"learning_rate": 6.012e-06, |
|
"loss": 0.0285, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 6.04, |
|
"eval_loss": 0.4436163902282715, |
|
"eval_runtime": 560.6152, |
|
"eval_samples_per_second": 4.565, |
|
"eval_steps_per_second": 0.143, |
|
"eval_wer": 12.595135833735851, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 6.05, |
|
"learning_rate": 5.912e-06, |
|
"loss": 0.1004, |
|
"step": 1525 |
|
}, |
|
{ |
|
"epoch": 6.06, |
|
"learning_rate": 5.812000000000001e-06, |
|
"loss": 0.0921, |
|
"step": 1550 |
|
}, |
|
{ |
|
"epoch": 6.07, |
|
"learning_rate": 5.7120000000000005e-06, |
|
"loss": 0.0294, |
|
"step": 1575 |
|
}, |
|
{ |
|
"epoch": 7.0, |
|
"learning_rate": 5.612000000000001e-06, |
|
"loss": 0.0263, |
|
"step": 1600 |
|
}, |
|
{ |
|
"epoch": 7.01, |
|
"learning_rate": 5.512000000000001e-06, |
|
"loss": 0.017, |
|
"step": 1625 |
|
}, |
|
{ |
|
"epoch": 7.02, |
|
"learning_rate": 5.412000000000001e-06, |
|
"loss": 0.0637, |
|
"step": 1650 |
|
}, |
|
{ |
|
"epoch": 7.03, |
|
"learning_rate": 5.312e-06, |
|
"loss": 0.0308, |
|
"step": 1675 |
|
}, |
|
{ |
|
"epoch": 7.03, |
|
"learning_rate": 5.212e-06, |
|
"loss": 0.0465, |
|
"step": 1700 |
|
}, |
|
{ |
|
"epoch": 7.04, |
|
"learning_rate": 5.112e-06, |
|
"loss": 0.0616, |
|
"step": 1725 |
|
}, |
|
{ |
|
"epoch": 7.05, |
|
"learning_rate": 5.0120000000000005e-06, |
|
"loss": 0.0251, |
|
"step": 1750 |
|
}, |
|
{ |
|
"epoch": 7.06, |
|
"learning_rate": 4.9120000000000006e-06, |
|
"loss": 0.016, |
|
"step": 1775 |
|
}, |
|
{ |
|
"epoch": 7.07, |
|
"learning_rate": 4.812000000000001e-06, |
|
"loss": 0.0183, |
|
"step": 1800 |
|
}, |
|
{ |
|
"epoch": 8.0, |
|
"learning_rate": 4.712000000000001e-06, |
|
"loss": 0.0182, |
|
"step": 1825 |
|
}, |
|
{ |
|
"epoch": 8.01, |
|
"learning_rate": 4.612e-06, |
|
"loss": 0.0219, |
|
"step": 1850 |
|
}, |
|
{ |
|
"epoch": 8.02, |
|
"learning_rate": 4.512e-06, |
|
"loss": 0.0153, |
|
"step": 1875 |
|
}, |
|
{ |
|
"epoch": 8.03, |
|
"learning_rate": 4.412e-06, |
|
"loss": 0.0126, |
|
"step": 1900 |
|
}, |
|
{ |
|
"epoch": 8.03, |
|
"learning_rate": 4.312e-06, |
|
"loss": 0.0228, |
|
"step": 1925 |
|
}, |
|
{ |
|
"epoch": 8.04, |
|
"learning_rate": 4.2120000000000005e-06, |
|
"loss": 0.043, |
|
"step": 1950 |
|
}, |
|
{ |
|
"epoch": 8.05, |
|
"learning_rate": 4.112000000000001e-06, |
|
"loss": 0.0153, |
|
"step": 1975 |
|
}, |
|
{ |
|
"epoch": 8.06, |
|
"learning_rate": 4.012000000000001e-06, |
|
"loss": 0.0319, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 8.06, |
|
"eval_loss": 0.5057264566421509, |
|
"eval_runtime": 546.7668, |
|
"eval_samples_per_second": 4.68, |
|
"eval_steps_per_second": 0.146, |
|
"eval_wer": 12.826203313272671, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 8.07, |
|
"learning_rate": 3.912e-06, |
|
"loss": 0.0151, |
|
"step": 2025 |
|
}, |
|
{ |
|
"epoch": 8.08, |
|
"learning_rate": 3.812e-06, |
|
"loss": 0.0105, |
|
"step": 2050 |
|
}, |
|
{ |
|
"epoch": 9.01, |
|
"learning_rate": 3.712e-06, |
|
"loss": 0.0069, |
|
"step": 2075 |
|
}, |
|
{ |
|
"epoch": 9.02, |
|
"learning_rate": 3.6120000000000003e-06, |
|
"loss": 0.0068, |
|
"step": 2100 |
|
}, |
|
{ |
|
"epoch": 9.02, |
|
"learning_rate": 3.5120000000000004e-06, |
|
"loss": 0.0124, |
|
"step": 2125 |
|
}, |
|
{ |
|
"epoch": 9.03, |
|
"learning_rate": 3.412e-06, |
|
"loss": 0.0343, |
|
"step": 2150 |
|
}, |
|
{ |
|
"epoch": 9.04, |
|
"learning_rate": 3.3120000000000002e-06, |
|
"loss": 0.0123, |
|
"step": 2175 |
|
}, |
|
{ |
|
"epoch": 9.05, |
|
"learning_rate": 3.212e-06, |
|
"loss": 0.0219, |
|
"step": 2200 |
|
}, |
|
{ |
|
"epoch": 9.06, |
|
"learning_rate": 3.112e-06, |
|
"loss": 0.0065, |
|
"step": 2225 |
|
}, |
|
{ |
|
"epoch": 9.07, |
|
"learning_rate": 3.0120000000000006e-06, |
|
"loss": 0.0052, |
|
"step": 2250 |
|
}, |
|
{ |
|
"epoch": 9.07, |
|
"learning_rate": 2.9120000000000002e-06, |
|
"loss": 0.0079, |
|
"step": 2275 |
|
}, |
|
{ |
|
"epoch": 10.01, |
|
"learning_rate": 2.8120000000000004e-06, |
|
"loss": 0.0046, |
|
"step": 2300 |
|
}, |
|
{ |
|
"epoch": 10.02, |
|
"learning_rate": 2.712e-06, |
|
"loss": 0.0112, |
|
"step": 2325 |
|
}, |
|
{ |
|
"epoch": 10.02, |
|
"learning_rate": 2.612e-06, |
|
"loss": 0.0065, |
|
"step": 2350 |
|
}, |
|
{ |
|
"epoch": 10.03, |
|
"learning_rate": 2.512e-06, |
|
"loss": 0.0182, |
|
"step": 2375 |
|
}, |
|
{ |
|
"epoch": 10.04, |
|
"learning_rate": 2.4120000000000004e-06, |
|
"loss": 0.0093, |
|
"step": 2400 |
|
}, |
|
{ |
|
"epoch": 10.05, |
|
"learning_rate": 2.312e-06, |
|
"loss": 0.0107, |
|
"step": 2425 |
|
}, |
|
{ |
|
"epoch": 10.06, |
|
"learning_rate": 2.212e-06, |
|
"loss": 0.0063, |
|
"step": 2450 |
|
}, |
|
{ |
|
"epoch": 10.06, |
|
"learning_rate": 2.1120000000000003e-06, |
|
"loss": 0.0048, |
|
"step": 2475 |
|
}, |
|
{ |
|
"epoch": 10.07, |
|
"learning_rate": 2.012e-06, |
|
"loss": 0.0044, |
|
"step": 2500 |
|
}, |
|
{ |
|
"epoch": 10.07, |
|
"eval_loss": 0.5620025396347046, |
|
"eval_runtime": 564.8934, |
|
"eval_samples_per_second": 4.53, |
|
"eval_steps_per_second": 0.142, |
|
"eval_wer": 13.123849558099765, |
|
"step": 2500 |
|
}, |
|
{ |
|
"epoch": 11.01, |
|
"learning_rate": 1.912e-06, |
|
"loss": 0.0104, |
|
"step": 2525 |
|
}, |
|
{ |
|
"epoch": 11.01, |
|
"learning_rate": 1.8120000000000002e-06, |
|
"loss": 0.0044, |
|
"step": 2550 |
|
}, |
|
{ |
|
"epoch": 11.02, |
|
"learning_rate": 1.712e-06, |
|
"loss": 0.0036, |
|
"step": 2575 |
|
}, |
|
{ |
|
"epoch": 11.03, |
|
"learning_rate": 1.6120000000000002e-06, |
|
"loss": 0.0067, |
|
"step": 2600 |
|
}, |
|
{ |
|
"epoch": 11.04, |
|
"learning_rate": 1.512e-06, |
|
"loss": 0.0056, |
|
"step": 2625 |
|
}, |
|
{ |
|
"epoch": 11.05, |
|
"learning_rate": 1.412e-06, |
|
"loss": 0.0033, |
|
"step": 2650 |
|
}, |
|
{ |
|
"epoch": 11.06, |
|
"learning_rate": 1.3120000000000003e-06, |
|
"loss": 0.0038, |
|
"step": 2675 |
|
}, |
|
{ |
|
"epoch": 11.06, |
|
"learning_rate": 1.2120000000000002e-06, |
|
"loss": 0.0037, |
|
"step": 2700 |
|
}, |
|
{ |
|
"epoch": 11.07, |
|
"learning_rate": 1.1120000000000001e-06, |
|
"loss": 0.0154, |
|
"step": 2725 |
|
}, |
|
{ |
|
"epoch": 12.0, |
|
"learning_rate": 1.012e-06, |
|
"loss": 0.0056, |
|
"step": 2750 |
|
}, |
|
{ |
|
"epoch": 12.01, |
|
"learning_rate": 9.120000000000001e-07, |
|
"loss": 0.0078, |
|
"step": 2775 |
|
}, |
|
{ |
|
"epoch": 12.02, |
|
"learning_rate": 8.12e-07, |
|
"loss": 0.0085, |
|
"step": 2800 |
|
}, |
|
{ |
|
"epoch": 12.03, |
|
"learning_rate": 7.12e-07, |
|
"loss": 0.0049, |
|
"step": 2825 |
|
}, |
|
{ |
|
"epoch": 12.04, |
|
"learning_rate": 6.12e-07, |
|
"loss": 0.0031, |
|
"step": 2850 |
|
}, |
|
{ |
|
"epoch": 12.05, |
|
"learning_rate": 5.12e-07, |
|
"loss": 0.003, |
|
"step": 2875 |
|
}, |
|
{ |
|
"epoch": 12.05, |
|
"learning_rate": 4.1200000000000004e-07, |
|
"loss": 0.0034, |
|
"step": 2900 |
|
}, |
|
{ |
|
"epoch": 12.06, |
|
"learning_rate": 3.12e-07, |
|
"loss": 0.0099, |
|
"step": 2925 |
|
}, |
|
{ |
|
"epoch": 12.07, |
|
"learning_rate": 2.1200000000000002e-07, |
|
"loss": 0.0045, |
|
"step": 2950 |
|
}, |
|
{ |
|
"epoch": 13.0, |
|
"learning_rate": 1.1200000000000001e-07, |
|
"loss": 0.0033, |
|
"step": 2975 |
|
}, |
|
{ |
|
"epoch": 13.01, |
|
"learning_rate": 1.2e-08, |
|
"loss": 0.0054, |
|
"step": 3000 |
|
}, |
|
{ |
|
"epoch": 13.01, |
|
"eval_loss": 0.5805584192276001, |
|
"eval_runtime": 543.2649, |
|
"eval_samples_per_second": 4.71, |
|
"eval_steps_per_second": 0.147, |
|
"eval_wer": 12.980248299630553, |
|
"step": 3000 |
|
}, |
|
{ |
|
"epoch": 13.01, |
|
"step": 3000, |
|
"total_flos": 5.528084211449856e+19, |
|
"train_loss": 0.17952798489481211, |
|
"train_runtime": 38488.0043, |
|
"train_samples_per_second": 4.989, |
|
"train_steps_per_second": 0.078 |
|
} |
|
], |
|
"max_steps": 3000, |
|
"num_train_epochs": 9223372036854775807, |
|
"total_flos": 5.528084211449856e+19, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|