|
{ |
|
"best_metric": 0.7635990420800547, |
|
"best_model_checkpoint": "./whisper-small-dialect_levantine/checkpoint-4000", |
|
"epoch": 3.5087719298245617, |
|
"eval_steps": 1000, |
|
"global_step": 5000, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.02, |
|
"grad_norm": 105.01622009277344, |
|
"learning_rate": 5.000000000000001e-07, |
|
"loss": 4.4756, |
|
"step": 25 |
|
}, |
|
{ |
|
"epoch": 0.04, |
|
"grad_norm": 38.415096282958984, |
|
"learning_rate": 1.0000000000000002e-06, |
|
"loss": 3.8473, |
|
"step": 50 |
|
}, |
|
{ |
|
"epoch": 0.05, |
|
"grad_norm": 32.70545959472656, |
|
"learning_rate": 1.5e-06, |
|
"loss": 3.1106, |
|
"step": 75 |
|
}, |
|
{ |
|
"epoch": 0.07, |
|
"grad_norm": 24.912517547607422, |
|
"learning_rate": 2.0000000000000003e-06, |
|
"loss": 2.3441, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 0.09, |
|
"grad_norm": 21.164583206176758, |
|
"learning_rate": 2.5e-06, |
|
"loss": 2.0976, |
|
"step": 125 |
|
}, |
|
{ |
|
"epoch": 0.11, |
|
"grad_norm": 28.40788459777832, |
|
"learning_rate": 3e-06, |
|
"loss": 1.9012, |
|
"step": 150 |
|
}, |
|
{ |
|
"epoch": 0.12, |
|
"grad_norm": 27.274938583374023, |
|
"learning_rate": 3.5e-06, |
|
"loss": 1.8956, |
|
"step": 175 |
|
}, |
|
{ |
|
"epoch": 0.14, |
|
"grad_norm": 20.819931030273438, |
|
"learning_rate": 4.000000000000001e-06, |
|
"loss": 1.6958, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 0.16, |
|
"grad_norm": 27.277860641479492, |
|
"learning_rate": 4.5e-06, |
|
"loss": 1.7819, |
|
"step": 225 |
|
}, |
|
{ |
|
"epoch": 0.18, |
|
"grad_norm": 22.838720321655273, |
|
"learning_rate": 5e-06, |
|
"loss": 1.6489, |
|
"step": 250 |
|
}, |
|
{ |
|
"epoch": 0.19, |
|
"grad_norm": 23.70868492126465, |
|
"learning_rate": 5.500000000000001e-06, |
|
"loss": 1.5876, |
|
"step": 275 |
|
}, |
|
{ |
|
"epoch": 0.21, |
|
"grad_norm": 25.0347900390625, |
|
"learning_rate": 6e-06, |
|
"loss": 1.6304, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 0.23, |
|
"grad_norm": 23.519012451171875, |
|
"learning_rate": 6.5000000000000004e-06, |
|
"loss": 1.324, |
|
"step": 325 |
|
}, |
|
{ |
|
"epoch": 0.25, |
|
"grad_norm": 21.76858901977539, |
|
"learning_rate": 7e-06, |
|
"loss": 1.3019, |
|
"step": 350 |
|
}, |
|
{ |
|
"epoch": 0.26, |
|
"grad_norm": 27.329273223876953, |
|
"learning_rate": 7.500000000000001e-06, |
|
"loss": 1.1865, |
|
"step": 375 |
|
}, |
|
{ |
|
"epoch": 0.28, |
|
"grad_norm": 23.374053955078125, |
|
"learning_rate": 8.000000000000001e-06, |
|
"loss": 1.2699, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 0.3, |
|
"grad_norm": 23.69354820251465, |
|
"learning_rate": 8.5e-06, |
|
"loss": 1.2327, |
|
"step": 425 |
|
}, |
|
{ |
|
"epoch": 0.32, |
|
"grad_norm": 22.70684814453125, |
|
"learning_rate": 9e-06, |
|
"loss": 1.2815, |
|
"step": 450 |
|
}, |
|
{ |
|
"epoch": 0.33, |
|
"grad_norm": 20.093311309814453, |
|
"learning_rate": 9.5e-06, |
|
"loss": 1.2482, |
|
"step": 475 |
|
}, |
|
{ |
|
"epoch": 0.35, |
|
"grad_norm": 21.54157066345215, |
|
"learning_rate": 1e-05, |
|
"loss": 1.2527, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 0.37, |
|
"grad_norm": 20.645353317260742, |
|
"learning_rate": 9.944444444444445e-06, |
|
"loss": 1.2954, |
|
"step": 525 |
|
}, |
|
{ |
|
"epoch": 0.39, |
|
"grad_norm": 17.942337036132812, |
|
"learning_rate": 9.88888888888889e-06, |
|
"loss": 1.2958, |
|
"step": 550 |
|
}, |
|
{ |
|
"epoch": 0.4, |
|
"grad_norm": 30.65532112121582, |
|
"learning_rate": 9.833333333333333e-06, |
|
"loss": 1.2974, |
|
"step": 575 |
|
}, |
|
{ |
|
"epoch": 0.42, |
|
"grad_norm": 25.07595443725586, |
|
"learning_rate": 9.777777777777779e-06, |
|
"loss": 1.289, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 0.44, |
|
"grad_norm": 20.494924545288086, |
|
"learning_rate": 9.722222222222223e-06, |
|
"loss": 1.1602, |
|
"step": 625 |
|
}, |
|
{ |
|
"epoch": 0.46, |
|
"grad_norm": 15.94078540802002, |
|
"learning_rate": 9.666666666666667e-06, |
|
"loss": 1.2113, |
|
"step": 650 |
|
}, |
|
{ |
|
"epoch": 0.47, |
|
"grad_norm": 21.14436149597168, |
|
"learning_rate": 9.611111111111112e-06, |
|
"loss": 1.2642, |
|
"step": 675 |
|
}, |
|
{ |
|
"epoch": 0.49, |
|
"grad_norm": 23.884052276611328, |
|
"learning_rate": 9.555555555555556e-06, |
|
"loss": 1.1667, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 0.51, |
|
"grad_norm": 23.286977767944336, |
|
"learning_rate": 9.5e-06, |
|
"loss": 1.1886, |
|
"step": 725 |
|
}, |
|
{ |
|
"epoch": 0.53, |
|
"grad_norm": 25.439273834228516, |
|
"learning_rate": 9.444444444444445e-06, |
|
"loss": 1.1659, |
|
"step": 750 |
|
}, |
|
{ |
|
"epoch": 0.54, |
|
"grad_norm": 21.63466453552246, |
|
"learning_rate": 9.38888888888889e-06, |
|
"loss": 1.3864, |
|
"step": 775 |
|
}, |
|
{ |
|
"epoch": 0.56, |
|
"grad_norm": 19.906169891357422, |
|
"learning_rate": 9.333333333333334e-06, |
|
"loss": 1.1403, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 0.58, |
|
"grad_norm": 24.829113006591797, |
|
"learning_rate": 9.277777777777778e-06, |
|
"loss": 1.1671, |
|
"step": 825 |
|
}, |
|
{ |
|
"epoch": 0.6, |
|
"grad_norm": 17.034282684326172, |
|
"learning_rate": 9.222222222222224e-06, |
|
"loss": 1.1379, |
|
"step": 850 |
|
}, |
|
{ |
|
"epoch": 0.61, |
|
"grad_norm": 18.528812408447266, |
|
"learning_rate": 9.166666666666666e-06, |
|
"loss": 1.1063, |
|
"step": 875 |
|
}, |
|
{ |
|
"epoch": 0.63, |
|
"grad_norm": 23.627670288085938, |
|
"learning_rate": 9.111111111111112e-06, |
|
"loss": 1.2375, |
|
"step": 900 |
|
}, |
|
{ |
|
"epoch": 0.65, |
|
"grad_norm": 17.76607894897461, |
|
"learning_rate": 9.055555555555556e-06, |
|
"loss": 1.1779, |
|
"step": 925 |
|
}, |
|
{ |
|
"epoch": 0.67, |
|
"grad_norm": 21.990543365478516, |
|
"learning_rate": 9e-06, |
|
"loss": 1.1007, |
|
"step": 950 |
|
}, |
|
{ |
|
"epoch": 0.68, |
|
"grad_norm": 21.411767959594727, |
|
"learning_rate": 8.944444444444446e-06, |
|
"loss": 1.1753, |
|
"step": 975 |
|
}, |
|
{ |
|
"epoch": 0.7, |
|
"grad_norm": 22.59610939025879, |
|
"learning_rate": 8.888888888888888e-06, |
|
"loss": 1.1945, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 0.7, |
|
"eval_loss": 1.1679444313049316, |
|
"eval_runtime": 866.814, |
|
"eval_samples_per_second": 3.288, |
|
"eval_steps_per_second": 0.412, |
|
"eval_wer": 0.8150872391378721, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 0.72, |
|
"grad_norm": 17.922718048095703, |
|
"learning_rate": 8.833333333333334e-06, |
|
"loss": 1.1264, |
|
"step": 1025 |
|
}, |
|
{ |
|
"epoch": 0.74, |
|
"grad_norm": 22.84149742126465, |
|
"learning_rate": 8.777777777777778e-06, |
|
"loss": 1.1382, |
|
"step": 1050 |
|
}, |
|
{ |
|
"epoch": 0.75, |
|
"grad_norm": 21.807554244995117, |
|
"learning_rate": 8.722222222222224e-06, |
|
"loss": 1.1199, |
|
"step": 1075 |
|
}, |
|
{ |
|
"epoch": 0.77, |
|
"grad_norm": 15.542016983032227, |
|
"learning_rate": 8.666666666666668e-06, |
|
"loss": 1.1283, |
|
"step": 1100 |
|
}, |
|
{ |
|
"epoch": 0.79, |
|
"grad_norm": 19.53961181640625, |
|
"learning_rate": 8.611111111111112e-06, |
|
"loss": 1.1479, |
|
"step": 1125 |
|
}, |
|
{ |
|
"epoch": 0.81, |
|
"grad_norm": 14.435591697692871, |
|
"learning_rate": 8.555555555555556e-06, |
|
"loss": 1.2046, |
|
"step": 1150 |
|
}, |
|
{ |
|
"epoch": 0.82, |
|
"grad_norm": 15.91525650024414, |
|
"learning_rate": 8.5e-06, |
|
"loss": 1.2012, |
|
"step": 1175 |
|
}, |
|
{ |
|
"epoch": 0.84, |
|
"grad_norm": 25.986412048339844, |
|
"learning_rate": 8.444444444444446e-06, |
|
"loss": 1.1793, |
|
"step": 1200 |
|
}, |
|
{ |
|
"epoch": 0.86, |
|
"grad_norm": 19.79583740234375, |
|
"learning_rate": 8.38888888888889e-06, |
|
"loss": 1.1427, |
|
"step": 1225 |
|
}, |
|
{ |
|
"epoch": 0.88, |
|
"grad_norm": 23.57172966003418, |
|
"learning_rate": 8.333333333333334e-06, |
|
"loss": 1.1072, |
|
"step": 1250 |
|
}, |
|
{ |
|
"epoch": 0.89, |
|
"grad_norm": 22.90821647644043, |
|
"learning_rate": 8.277777777777778e-06, |
|
"loss": 1.1761, |
|
"step": 1275 |
|
}, |
|
{ |
|
"epoch": 0.91, |
|
"grad_norm": 27.41729736328125, |
|
"learning_rate": 8.222222222222222e-06, |
|
"loss": 1.1739, |
|
"step": 1300 |
|
}, |
|
{ |
|
"epoch": 0.93, |
|
"grad_norm": 30.00156593322754, |
|
"learning_rate": 8.166666666666668e-06, |
|
"loss": 1.2369, |
|
"step": 1325 |
|
}, |
|
{ |
|
"epoch": 0.95, |
|
"grad_norm": 22.23008155822754, |
|
"learning_rate": 8.111111111111112e-06, |
|
"loss": 1.1685, |
|
"step": 1350 |
|
}, |
|
{ |
|
"epoch": 0.96, |
|
"grad_norm": 18.34453582763672, |
|
"learning_rate": 8.055555555555557e-06, |
|
"loss": 1.1862, |
|
"step": 1375 |
|
}, |
|
{ |
|
"epoch": 0.98, |
|
"grad_norm": 21.947595596313477, |
|
"learning_rate": 8.000000000000001e-06, |
|
"loss": 1.238, |
|
"step": 1400 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"grad_norm": 28.159698486328125, |
|
"learning_rate": 7.944444444444445e-06, |
|
"loss": 1.0259, |
|
"step": 1425 |
|
}, |
|
{ |
|
"epoch": 1.02, |
|
"grad_norm": 13.005045890808105, |
|
"learning_rate": 7.88888888888889e-06, |
|
"loss": 0.843, |
|
"step": 1450 |
|
}, |
|
{ |
|
"epoch": 1.04, |
|
"grad_norm": 14.627166748046875, |
|
"learning_rate": 7.833333333333333e-06, |
|
"loss": 0.8975, |
|
"step": 1475 |
|
}, |
|
{ |
|
"epoch": 1.05, |
|
"grad_norm": 12.167830467224121, |
|
"learning_rate": 7.77777777777778e-06, |
|
"loss": 0.857, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 1.07, |
|
"grad_norm": 20.749853134155273, |
|
"learning_rate": 7.722222222222223e-06, |
|
"loss": 0.8968, |
|
"step": 1525 |
|
}, |
|
{ |
|
"epoch": 1.09, |
|
"grad_norm": 16.837162017822266, |
|
"learning_rate": 7.666666666666667e-06, |
|
"loss": 0.8388, |
|
"step": 1550 |
|
}, |
|
{ |
|
"epoch": 1.11, |
|
"grad_norm": 17.441505432128906, |
|
"learning_rate": 7.611111111111111e-06, |
|
"loss": 0.8734, |
|
"step": 1575 |
|
}, |
|
{ |
|
"epoch": 1.12, |
|
"grad_norm": 18.170475006103516, |
|
"learning_rate": 7.555555555555556e-06, |
|
"loss": 0.8283, |
|
"step": 1600 |
|
}, |
|
{ |
|
"epoch": 1.14, |
|
"grad_norm": 19.032150268554688, |
|
"learning_rate": 7.500000000000001e-06, |
|
"loss": 0.7661, |
|
"step": 1625 |
|
}, |
|
{ |
|
"epoch": 1.16, |
|
"grad_norm": 15.91515827178955, |
|
"learning_rate": 7.444444444444445e-06, |
|
"loss": 0.8611, |
|
"step": 1650 |
|
}, |
|
{ |
|
"epoch": 1.18, |
|
"grad_norm": 18.5009765625, |
|
"learning_rate": 7.38888888888889e-06, |
|
"loss": 0.8536, |
|
"step": 1675 |
|
}, |
|
{ |
|
"epoch": 1.19, |
|
"grad_norm": 22.158689498901367, |
|
"learning_rate": 7.333333333333333e-06, |
|
"loss": 0.9115, |
|
"step": 1700 |
|
}, |
|
{ |
|
"epoch": 1.21, |
|
"grad_norm": 13.969917297363281, |
|
"learning_rate": 7.277777777777778e-06, |
|
"loss": 0.9389, |
|
"step": 1725 |
|
}, |
|
{ |
|
"epoch": 1.23, |
|
"grad_norm": 14.708273887634277, |
|
"learning_rate": 7.222222222222223e-06, |
|
"loss": 0.8358, |
|
"step": 1750 |
|
}, |
|
{ |
|
"epoch": 1.25, |
|
"grad_norm": 18.655982971191406, |
|
"learning_rate": 7.166666666666667e-06, |
|
"loss": 0.9696, |
|
"step": 1775 |
|
}, |
|
{ |
|
"epoch": 1.26, |
|
"grad_norm": 24.355396270751953, |
|
"learning_rate": 7.111111111111112e-06, |
|
"loss": 0.8242, |
|
"step": 1800 |
|
}, |
|
{ |
|
"epoch": 1.28, |
|
"grad_norm": 15.554370880126953, |
|
"learning_rate": 7.055555555555557e-06, |
|
"loss": 0.9143, |
|
"step": 1825 |
|
}, |
|
{ |
|
"epoch": 1.3, |
|
"grad_norm": 20.58561897277832, |
|
"learning_rate": 7e-06, |
|
"loss": 0.8439, |
|
"step": 1850 |
|
}, |
|
{ |
|
"epoch": 1.32, |
|
"grad_norm": 20.269287109375, |
|
"learning_rate": 6.944444444444445e-06, |
|
"loss": 0.8884, |
|
"step": 1875 |
|
}, |
|
{ |
|
"epoch": 1.33, |
|
"grad_norm": 11.557083129882812, |
|
"learning_rate": 6.88888888888889e-06, |
|
"loss": 0.9044, |
|
"step": 1900 |
|
}, |
|
{ |
|
"epoch": 1.35, |
|
"grad_norm": 23.037826538085938, |
|
"learning_rate": 6.833333333333334e-06, |
|
"loss": 0.8734, |
|
"step": 1925 |
|
}, |
|
{ |
|
"epoch": 1.37, |
|
"grad_norm": 20.72858428955078, |
|
"learning_rate": 6.777777777777779e-06, |
|
"loss": 0.8953, |
|
"step": 1950 |
|
}, |
|
{ |
|
"epoch": 1.39, |
|
"grad_norm": 18.743228912353516, |
|
"learning_rate": 6.7222222222222235e-06, |
|
"loss": 0.8319, |
|
"step": 1975 |
|
}, |
|
{ |
|
"epoch": 1.4, |
|
"grad_norm": 17.591102600097656, |
|
"learning_rate": 6.666666666666667e-06, |
|
"loss": 0.8778, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 1.4, |
|
"eval_loss": 1.1207834482192993, |
|
"eval_runtime": 750.3921, |
|
"eval_samples_per_second": 3.798, |
|
"eval_steps_per_second": 0.476, |
|
"eval_wer": 0.7828144600296499, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 1.42, |
|
"grad_norm": 18.505678176879883, |
|
"learning_rate": 6.6111111111111115e-06, |
|
"loss": 0.8456, |
|
"step": 2025 |
|
}, |
|
{ |
|
"epoch": 1.44, |
|
"grad_norm": 19.876806259155273, |
|
"learning_rate": 6.555555555555556e-06, |
|
"loss": 0.9283, |
|
"step": 2050 |
|
}, |
|
{ |
|
"epoch": 1.46, |
|
"grad_norm": 23.498214721679688, |
|
"learning_rate": 6.5000000000000004e-06, |
|
"loss": 0.9199, |
|
"step": 2075 |
|
}, |
|
{ |
|
"epoch": 1.47, |
|
"grad_norm": 19.818599700927734, |
|
"learning_rate": 6.444444444444445e-06, |
|
"loss": 0.8638, |
|
"step": 2100 |
|
}, |
|
{ |
|
"epoch": 1.49, |
|
"grad_norm": 17.756738662719727, |
|
"learning_rate": 6.3888888888888885e-06, |
|
"loss": 0.9432, |
|
"step": 2125 |
|
}, |
|
{ |
|
"epoch": 1.51, |
|
"grad_norm": 13.811593055725098, |
|
"learning_rate": 6.333333333333333e-06, |
|
"loss": 0.7516, |
|
"step": 2150 |
|
}, |
|
{ |
|
"epoch": 1.53, |
|
"grad_norm": 17.024320602416992, |
|
"learning_rate": 6.277777777777778e-06, |
|
"loss": 0.9079, |
|
"step": 2175 |
|
}, |
|
{ |
|
"epoch": 1.54, |
|
"grad_norm": 22.73233413696289, |
|
"learning_rate": 6.222222222222223e-06, |
|
"loss": 0.8704, |
|
"step": 2200 |
|
}, |
|
{ |
|
"epoch": 1.56, |
|
"grad_norm": 18.211219787597656, |
|
"learning_rate": 6.166666666666667e-06, |
|
"loss": 0.8635, |
|
"step": 2225 |
|
}, |
|
{ |
|
"epoch": 1.58, |
|
"grad_norm": 14.778497695922852, |
|
"learning_rate": 6.111111111111112e-06, |
|
"loss": 0.8439, |
|
"step": 2250 |
|
}, |
|
{ |
|
"epoch": 1.6, |
|
"grad_norm": 22.68735694885254, |
|
"learning_rate": 6.055555555555555e-06, |
|
"loss": 0.8572, |
|
"step": 2275 |
|
}, |
|
{ |
|
"epoch": 1.61, |
|
"grad_norm": 19.818214416503906, |
|
"learning_rate": 6e-06, |
|
"loss": 0.8196, |
|
"step": 2300 |
|
}, |
|
{ |
|
"epoch": 1.63, |
|
"grad_norm": 22.226512908935547, |
|
"learning_rate": 5.944444444444445e-06, |
|
"loss": 0.8498, |
|
"step": 2325 |
|
}, |
|
{ |
|
"epoch": 1.65, |
|
"grad_norm": 15.627732276916504, |
|
"learning_rate": 5.88888888888889e-06, |
|
"loss": 0.905, |
|
"step": 2350 |
|
}, |
|
{ |
|
"epoch": 1.67, |
|
"grad_norm": 12.331079483032227, |
|
"learning_rate": 5.833333333333334e-06, |
|
"loss": 0.8842, |
|
"step": 2375 |
|
}, |
|
{ |
|
"epoch": 1.68, |
|
"grad_norm": 17.217872619628906, |
|
"learning_rate": 5.777777777777778e-06, |
|
"loss": 0.7816, |
|
"step": 2400 |
|
}, |
|
{ |
|
"epoch": 1.7, |
|
"grad_norm": 19.87596893310547, |
|
"learning_rate": 5.722222222222222e-06, |
|
"loss": 0.884, |
|
"step": 2425 |
|
}, |
|
{ |
|
"epoch": 1.72, |
|
"grad_norm": 15.646110534667969, |
|
"learning_rate": 5.666666666666667e-06, |
|
"loss": 0.8369, |
|
"step": 2450 |
|
}, |
|
{ |
|
"epoch": 1.74, |
|
"grad_norm": 15.3438138961792, |
|
"learning_rate": 5.611111111111112e-06, |
|
"loss": 0.8858, |
|
"step": 2475 |
|
}, |
|
{ |
|
"epoch": 1.75, |
|
"grad_norm": 26.44184112548828, |
|
"learning_rate": 5.555555555555557e-06, |
|
"loss": 0.9208, |
|
"step": 2500 |
|
}, |
|
{ |
|
"epoch": 1.77, |
|
"grad_norm": 16.2983455657959, |
|
"learning_rate": 5.500000000000001e-06, |
|
"loss": 0.8959, |
|
"step": 2525 |
|
}, |
|
{ |
|
"epoch": 1.79, |
|
"grad_norm": 17.188587188720703, |
|
"learning_rate": 5.444444444444445e-06, |
|
"loss": 0.8785, |
|
"step": 2550 |
|
}, |
|
{ |
|
"epoch": 1.81, |
|
"grad_norm": 18.34478759765625, |
|
"learning_rate": 5.388888888888889e-06, |
|
"loss": 0.8116, |
|
"step": 2575 |
|
}, |
|
{ |
|
"epoch": 1.82, |
|
"grad_norm": 14.330605506896973, |
|
"learning_rate": 5.333333333333334e-06, |
|
"loss": 0.8592, |
|
"step": 2600 |
|
}, |
|
{ |
|
"epoch": 1.84, |
|
"grad_norm": 17.461301803588867, |
|
"learning_rate": 5.2777777777777785e-06, |
|
"loss": 0.8729, |
|
"step": 2625 |
|
}, |
|
{ |
|
"epoch": 1.86, |
|
"grad_norm": 17.23073387145996, |
|
"learning_rate": 5.2222222222222226e-06, |
|
"loss": 0.8403, |
|
"step": 2650 |
|
}, |
|
{ |
|
"epoch": 1.88, |
|
"grad_norm": 16.816059112548828, |
|
"learning_rate": 5.1666666666666675e-06, |
|
"loss": 0.8266, |
|
"step": 2675 |
|
}, |
|
{ |
|
"epoch": 1.89, |
|
"grad_norm": 25.448545455932617, |
|
"learning_rate": 5.1111111111111115e-06, |
|
"loss": 0.8047, |
|
"step": 2700 |
|
}, |
|
{ |
|
"epoch": 1.91, |
|
"grad_norm": 16.244035720825195, |
|
"learning_rate": 5.0555555555555555e-06, |
|
"loss": 0.8034, |
|
"step": 2725 |
|
}, |
|
{ |
|
"epoch": 1.93, |
|
"grad_norm": 19.600101470947266, |
|
"learning_rate": 5e-06, |
|
"loss": 0.8126, |
|
"step": 2750 |
|
}, |
|
{ |
|
"epoch": 1.95, |
|
"grad_norm": 19.725372314453125, |
|
"learning_rate": 4.944444444444445e-06, |
|
"loss": 0.7724, |
|
"step": 2775 |
|
}, |
|
{ |
|
"epoch": 1.96, |
|
"grad_norm": 13.881664276123047, |
|
"learning_rate": 4.888888888888889e-06, |
|
"loss": 0.8932, |
|
"step": 2800 |
|
}, |
|
{ |
|
"epoch": 1.98, |
|
"grad_norm": 14.711101531982422, |
|
"learning_rate": 4.833333333333333e-06, |
|
"loss": 0.8111, |
|
"step": 2825 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"grad_norm": 22.943696975708008, |
|
"learning_rate": 4.777777777777778e-06, |
|
"loss": 0.7908, |
|
"step": 2850 |
|
}, |
|
{ |
|
"epoch": 2.02, |
|
"grad_norm": 17.254667282104492, |
|
"learning_rate": 4.722222222222222e-06, |
|
"loss": 0.7, |
|
"step": 2875 |
|
}, |
|
{ |
|
"epoch": 2.04, |
|
"grad_norm": 14.545769691467285, |
|
"learning_rate": 4.666666666666667e-06, |
|
"loss": 0.645, |
|
"step": 2900 |
|
}, |
|
{ |
|
"epoch": 2.05, |
|
"grad_norm": 11.072650909423828, |
|
"learning_rate": 4.611111111111112e-06, |
|
"loss": 0.5782, |
|
"step": 2925 |
|
}, |
|
{ |
|
"epoch": 2.07, |
|
"grad_norm": 14.558616638183594, |
|
"learning_rate": 4.555555555555556e-06, |
|
"loss": 0.5763, |
|
"step": 2950 |
|
}, |
|
{ |
|
"epoch": 2.09, |
|
"grad_norm": 16.14755630493164, |
|
"learning_rate": 4.5e-06, |
|
"loss": 0.5983, |
|
"step": 2975 |
|
}, |
|
{ |
|
"epoch": 2.11, |
|
"grad_norm": 18.195594787597656, |
|
"learning_rate": 4.444444444444444e-06, |
|
"loss": 0.6046, |
|
"step": 3000 |
|
}, |
|
{ |
|
"epoch": 2.11, |
|
"eval_loss": 1.1167398691177368, |
|
"eval_runtime": 778.5794, |
|
"eval_samples_per_second": 3.661, |
|
"eval_steps_per_second": 0.459, |
|
"eval_wer": 0.9134450906602806, |
|
"step": 3000 |
|
}, |
|
{ |
|
"epoch": 2.12, |
|
"grad_norm": 18.549034118652344, |
|
"learning_rate": 4.388888888888889e-06, |
|
"loss": 0.5426, |
|
"step": 3025 |
|
}, |
|
{ |
|
"epoch": 2.14, |
|
"grad_norm": 11.774942398071289, |
|
"learning_rate": 4.333333333333334e-06, |
|
"loss": 0.6074, |
|
"step": 3050 |
|
}, |
|
{ |
|
"epoch": 2.16, |
|
"grad_norm": 21.260597229003906, |
|
"learning_rate": 4.277777777777778e-06, |
|
"loss": 0.5798, |
|
"step": 3075 |
|
}, |
|
{ |
|
"epoch": 2.18, |
|
"grad_norm": 18.69565773010254, |
|
"learning_rate": 4.222222222222223e-06, |
|
"loss": 0.5168, |
|
"step": 3100 |
|
}, |
|
{ |
|
"epoch": 2.19, |
|
"grad_norm": 19.02937889099121, |
|
"learning_rate": 4.166666666666667e-06, |
|
"loss": 0.6275, |
|
"step": 3125 |
|
}, |
|
{ |
|
"epoch": 2.21, |
|
"grad_norm": 12.80723762512207, |
|
"learning_rate": 4.111111111111111e-06, |
|
"loss": 0.6773, |
|
"step": 3150 |
|
}, |
|
{ |
|
"epoch": 2.23, |
|
"grad_norm": 14.226038932800293, |
|
"learning_rate": 4.055555555555556e-06, |
|
"loss": 0.6184, |
|
"step": 3175 |
|
}, |
|
{ |
|
"epoch": 2.25, |
|
"grad_norm": 15.16125774383545, |
|
"learning_rate": 4.000000000000001e-06, |
|
"loss": 0.5668, |
|
"step": 3200 |
|
}, |
|
{ |
|
"epoch": 2.26, |
|
"grad_norm": 9.370119094848633, |
|
"learning_rate": 3.944444444444445e-06, |
|
"loss": 0.5718, |
|
"step": 3225 |
|
}, |
|
{ |
|
"epoch": 2.28, |
|
"grad_norm": 15.369794845581055, |
|
"learning_rate": 3.88888888888889e-06, |
|
"loss": 0.6725, |
|
"step": 3250 |
|
}, |
|
{ |
|
"epoch": 2.3, |
|
"grad_norm": 13.635235786437988, |
|
"learning_rate": 3.833333333333334e-06, |
|
"loss": 0.5705, |
|
"step": 3275 |
|
}, |
|
{ |
|
"epoch": 2.32, |
|
"grad_norm": 12.524065971374512, |
|
"learning_rate": 3.777777777777778e-06, |
|
"loss": 0.6573, |
|
"step": 3300 |
|
}, |
|
{ |
|
"epoch": 2.33, |
|
"grad_norm": 14.678699493408203, |
|
"learning_rate": 3.7222222222222225e-06, |
|
"loss": 0.6684, |
|
"step": 3325 |
|
}, |
|
{ |
|
"epoch": 2.35, |
|
"grad_norm": 18.842226028442383, |
|
"learning_rate": 3.6666666666666666e-06, |
|
"loss": 0.5352, |
|
"step": 3350 |
|
}, |
|
{ |
|
"epoch": 2.37, |
|
"grad_norm": 16.38367462158203, |
|
"learning_rate": 3.6111111111111115e-06, |
|
"loss": 0.6098, |
|
"step": 3375 |
|
}, |
|
{ |
|
"epoch": 2.39, |
|
"grad_norm": 15.753807067871094, |
|
"learning_rate": 3.555555555555556e-06, |
|
"loss": 0.6081, |
|
"step": 3400 |
|
}, |
|
{ |
|
"epoch": 2.4, |
|
"grad_norm": 12.830479621887207, |
|
"learning_rate": 3.5e-06, |
|
"loss": 0.5804, |
|
"step": 3425 |
|
}, |
|
{ |
|
"epoch": 2.42, |
|
"grad_norm": 13.615287780761719, |
|
"learning_rate": 3.444444444444445e-06, |
|
"loss": 0.6099, |
|
"step": 3450 |
|
}, |
|
{ |
|
"epoch": 2.44, |
|
"grad_norm": 13.487171173095703, |
|
"learning_rate": 3.3888888888888893e-06, |
|
"loss": 0.6227, |
|
"step": 3475 |
|
}, |
|
{ |
|
"epoch": 2.46, |
|
"grad_norm": 14.19842529296875, |
|
"learning_rate": 3.3333333333333333e-06, |
|
"loss": 0.6085, |
|
"step": 3500 |
|
}, |
|
{ |
|
"epoch": 2.47, |
|
"grad_norm": 19.63158416748047, |
|
"learning_rate": 3.277777777777778e-06, |
|
"loss": 0.6429, |
|
"step": 3525 |
|
}, |
|
{ |
|
"epoch": 2.49, |
|
"grad_norm": 15.446227073669434, |
|
"learning_rate": 3.2222222222222227e-06, |
|
"loss": 0.5227, |
|
"step": 3550 |
|
}, |
|
{ |
|
"epoch": 2.51, |
|
"grad_norm": 14.606199264526367, |
|
"learning_rate": 3.1666666666666667e-06, |
|
"loss": 0.5511, |
|
"step": 3575 |
|
}, |
|
{ |
|
"epoch": 2.53, |
|
"grad_norm": 14.071135520935059, |
|
"learning_rate": 3.1111111111111116e-06, |
|
"loss": 0.5825, |
|
"step": 3600 |
|
}, |
|
{ |
|
"epoch": 2.54, |
|
"grad_norm": 10.74336051940918, |
|
"learning_rate": 3.055555555555556e-06, |
|
"loss": 0.6515, |
|
"step": 3625 |
|
}, |
|
{ |
|
"epoch": 2.56, |
|
"grad_norm": 13.432340621948242, |
|
"learning_rate": 3e-06, |
|
"loss": 0.5975, |
|
"step": 3650 |
|
}, |
|
{ |
|
"epoch": 2.58, |
|
"grad_norm": 15.705540657043457, |
|
"learning_rate": 2.944444444444445e-06, |
|
"loss": 0.5845, |
|
"step": 3675 |
|
}, |
|
{ |
|
"epoch": 2.6, |
|
"grad_norm": 15.272418022155762, |
|
"learning_rate": 2.888888888888889e-06, |
|
"loss": 0.5825, |
|
"step": 3700 |
|
}, |
|
{ |
|
"epoch": 2.61, |
|
"grad_norm": 23.63802719116211, |
|
"learning_rate": 2.8333333333333335e-06, |
|
"loss": 0.5883, |
|
"step": 3725 |
|
}, |
|
{ |
|
"epoch": 2.63, |
|
"grad_norm": 15.042671203613281, |
|
"learning_rate": 2.7777777777777783e-06, |
|
"loss": 0.5646, |
|
"step": 3750 |
|
}, |
|
{ |
|
"epoch": 2.65, |
|
"grad_norm": 15.876602172851562, |
|
"learning_rate": 2.7222222222222224e-06, |
|
"loss": 0.5859, |
|
"step": 3775 |
|
}, |
|
{ |
|
"epoch": 2.67, |
|
"grad_norm": 15.712118148803711, |
|
"learning_rate": 2.666666666666667e-06, |
|
"loss": 0.6506, |
|
"step": 3800 |
|
}, |
|
{ |
|
"epoch": 2.68, |
|
"grad_norm": 12.527118682861328, |
|
"learning_rate": 2.6111111111111113e-06, |
|
"loss": 0.6143, |
|
"step": 3825 |
|
}, |
|
{ |
|
"epoch": 2.7, |
|
"grad_norm": 15.08745002746582, |
|
"learning_rate": 2.5555555555555557e-06, |
|
"loss": 0.6117, |
|
"step": 3850 |
|
}, |
|
{ |
|
"epoch": 2.72, |
|
"grad_norm": 10.62087345123291, |
|
"learning_rate": 2.5e-06, |
|
"loss": 0.573, |
|
"step": 3875 |
|
}, |
|
{ |
|
"epoch": 2.74, |
|
"grad_norm": 15.184548377990723, |
|
"learning_rate": 2.4444444444444447e-06, |
|
"loss": 0.6388, |
|
"step": 3900 |
|
}, |
|
{ |
|
"epoch": 2.75, |
|
"grad_norm": 11.05632209777832, |
|
"learning_rate": 2.388888888888889e-06, |
|
"loss": 0.6476, |
|
"step": 3925 |
|
}, |
|
{ |
|
"epoch": 2.77, |
|
"grad_norm": 21.41426658630371, |
|
"learning_rate": 2.3333333333333336e-06, |
|
"loss": 0.601, |
|
"step": 3950 |
|
}, |
|
{ |
|
"epoch": 2.79, |
|
"grad_norm": 14.866922378540039, |
|
"learning_rate": 2.277777777777778e-06, |
|
"loss": 0.5879, |
|
"step": 3975 |
|
}, |
|
{ |
|
"epoch": 2.81, |
|
"grad_norm": 13.731112480163574, |
|
"learning_rate": 2.222222222222222e-06, |
|
"loss": 0.5933, |
|
"step": 4000 |
|
}, |
|
{ |
|
"epoch": 2.81, |
|
"eval_loss": 1.1129027605056763, |
|
"eval_runtime": 830.7696, |
|
"eval_samples_per_second": 3.431, |
|
"eval_steps_per_second": 0.43, |
|
"eval_wer": 0.7635990420800547, |
|
"step": 4000 |
|
}, |
|
{ |
|
"epoch": 2.82, |
|
"grad_norm": 16.265804290771484, |
|
"learning_rate": 2.166666666666667e-06, |
|
"loss": 0.6687, |
|
"step": 4025 |
|
}, |
|
{ |
|
"epoch": 2.84, |
|
"grad_norm": 16.1932315826416, |
|
"learning_rate": 2.1111111111111114e-06, |
|
"loss": 0.5827, |
|
"step": 4050 |
|
}, |
|
{ |
|
"epoch": 2.86, |
|
"grad_norm": 14.049725532531738, |
|
"learning_rate": 2.0555555555555555e-06, |
|
"loss": 0.5921, |
|
"step": 4075 |
|
}, |
|
{ |
|
"epoch": 2.88, |
|
"grad_norm": 16.718233108520508, |
|
"learning_rate": 2.0000000000000003e-06, |
|
"loss": 0.5516, |
|
"step": 4100 |
|
}, |
|
{ |
|
"epoch": 2.89, |
|
"grad_norm": 16.140684127807617, |
|
"learning_rate": 1.944444444444445e-06, |
|
"loss": 0.5442, |
|
"step": 4125 |
|
}, |
|
{ |
|
"epoch": 2.91, |
|
"grad_norm": 15.428065299987793, |
|
"learning_rate": 1.888888888888889e-06, |
|
"loss": 0.6088, |
|
"step": 4150 |
|
}, |
|
{ |
|
"epoch": 2.93, |
|
"grad_norm": 16.217172622680664, |
|
"learning_rate": 1.8333333333333333e-06, |
|
"loss": 0.6119, |
|
"step": 4175 |
|
}, |
|
{ |
|
"epoch": 2.95, |
|
"grad_norm": 23.09720230102539, |
|
"learning_rate": 1.777777777777778e-06, |
|
"loss": 0.5694, |
|
"step": 4200 |
|
}, |
|
{ |
|
"epoch": 2.96, |
|
"grad_norm": 18.135848999023438, |
|
"learning_rate": 1.7222222222222224e-06, |
|
"loss": 0.6148, |
|
"step": 4225 |
|
}, |
|
{ |
|
"epoch": 2.98, |
|
"grad_norm": 19.2158260345459, |
|
"learning_rate": 1.6666666666666667e-06, |
|
"loss": 0.6248, |
|
"step": 4250 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"grad_norm": 25.484533309936523, |
|
"learning_rate": 1.6111111111111113e-06, |
|
"loss": 0.5624, |
|
"step": 4275 |
|
}, |
|
{ |
|
"epoch": 3.02, |
|
"grad_norm": 11.158307075500488, |
|
"learning_rate": 1.5555555555555558e-06, |
|
"loss": 0.475, |
|
"step": 4300 |
|
}, |
|
{ |
|
"epoch": 3.04, |
|
"grad_norm": 14.582618713378906, |
|
"learning_rate": 1.5e-06, |
|
"loss": 0.4581, |
|
"step": 4325 |
|
}, |
|
{ |
|
"epoch": 3.05, |
|
"grad_norm": 11.259454727172852, |
|
"learning_rate": 1.4444444444444445e-06, |
|
"loss": 0.4561, |
|
"step": 4350 |
|
}, |
|
{ |
|
"epoch": 3.07, |
|
"grad_norm": 12.487770080566406, |
|
"learning_rate": 1.3888888888888892e-06, |
|
"loss": 0.4852, |
|
"step": 4375 |
|
}, |
|
{ |
|
"epoch": 3.09, |
|
"grad_norm": 10.27232837677002, |
|
"learning_rate": 1.3333333333333334e-06, |
|
"loss": 0.4099, |
|
"step": 4400 |
|
}, |
|
{ |
|
"epoch": 3.11, |
|
"grad_norm": 10.88982105255127, |
|
"learning_rate": 1.2777777777777779e-06, |
|
"loss": 0.4292, |
|
"step": 4425 |
|
}, |
|
{ |
|
"epoch": 3.12, |
|
"grad_norm": 12.418704986572266, |
|
"learning_rate": 1.2222222222222223e-06, |
|
"loss": 0.4237, |
|
"step": 4450 |
|
}, |
|
{ |
|
"epoch": 3.14, |
|
"grad_norm": 9.705329895019531, |
|
"learning_rate": 1.1666666666666668e-06, |
|
"loss": 0.4299, |
|
"step": 4475 |
|
}, |
|
{ |
|
"epoch": 3.16, |
|
"grad_norm": 11.134821891784668, |
|
"learning_rate": 1.111111111111111e-06, |
|
"loss": 0.4061, |
|
"step": 4500 |
|
}, |
|
{ |
|
"epoch": 3.18, |
|
"grad_norm": 13.798453330993652, |
|
"learning_rate": 1.0555555555555557e-06, |
|
"loss": 0.4039, |
|
"step": 4525 |
|
}, |
|
{ |
|
"epoch": 3.19, |
|
"grad_norm": 13.317481994628906, |
|
"learning_rate": 1.0000000000000002e-06, |
|
"loss": 0.4498, |
|
"step": 4550 |
|
}, |
|
{ |
|
"epoch": 3.21, |
|
"grad_norm": 13.943197250366211, |
|
"learning_rate": 9.444444444444445e-07, |
|
"loss": 0.4514, |
|
"step": 4575 |
|
}, |
|
{ |
|
"epoch": 3.23, |
|
"grad_norm": 9.061113357543945, |
|
"learning_rate": 8.88888888888889e-07, |
|
"loss": 0.436, |
|
"step": 4600 |
|
}, |
|
{ |
|
"epoch": 3.25, |
|
"grad_norm": 13.598713874816895, |
|
"learning_rate": 8.333333333333333e-07, |
|
"loss": 0.4427, |
|
"step": 4625 |
|
}, |
|
{ |
|
"epoch": 3.26, |
|
"grad_norm": 18.128005981445312, |
|
"learning_rate": 7.777777777777779e-07, |
|
"loss": 0.4219, |
|
"step": 4650 |
|
}, |
|
{ |
|
"epoch": 3.28, |
|
"grad_norm": 10.521560668945312, |
|
"learning_rate": 7.222222222222222e-07, |
|
"loss": 0.4455, |
|
"step": 4675 |
|
}, |
|
{ |
|
"epoch": 3.3, |
|
"grad_norm": 13.369185447692871, |
|
"learning_rate": 6.666666666666667e-07, |
|
"loss": 0.4168, |
|
"step": 4700 |
|
}, |
|
{ |
|
"epoch": 3.32, |
|
"grad_norm": 8.334280014038086, |
|
"learning_rate": 6.111111111111112e-07, |
|
"loss": 0.4713, |
|
"step": 4725 |
|
}, |
|
{ |
|
"epoch": 3.33, |
|
"grad_norm": 9.726198196411133, |
|
"learning_rate": 5.555555555555555e-07, |
|
"loss": 0.4642, |
|
"step": 4750 |
|
}, |
|
{ |
|
"epoch": 3.35, |
|
"grad_norm": 16.844810485839844, |
|
"learning_rate": 5.000000000000001e-07, |
|
"loss": 0.5192, |
|
"step": 4775 |
|
}, |
|
{ |
|
"epoch": 3.37, |
|
"grad_norm": 14.9822998046875, |
|
"learning_rate": 4.444444444444445e-07, |
|
"loss": 0.4789, |
|
"step": 4800 |
|
}, |
|
{ |
|
"epoch": 3.39, |
|
"grad_norm": 10.031481742858887, |
|
"learning_rate": 3.8888888888888895e-07, |
|
"loss": 0.4984, |
|
"step": 4825 |
|
}, |
|
{ |
|
"epoch": 3.4, |
|
"grad_norm": 9.072463035583496, |
|
"learning_rate": 3.3333333333333335e-07, |
|
"loss": 0.3774, |
|
"step": 4850 |
|
}, |
|
{ |
|
"epoch": 3.42, |
|
"grad_norm": 13.830499649047852, |
|
"learning_rate": 2.7777777777777776e-07, |
|
"loss": 0.456, |
|
"step": 4875 |
|
}, |
|
{ |
|
"epoch": 3.44, |
|
"grad_norm": 12.41981315612793, |
|
"learning_rate": 2.2222222222222224e-07, |
|
"loss": 0.5071, |
|
"step": 4900 |
|
}, |
|
{ |
|
"epoch": 3.46, |
|
"grad_norm": 14.112458229064941, |
|
"learning_rate": 1.6666666666666668e-07, |
|
"loss": 0.504, |
|
"step": 4925 |
|
}, |
|
{ |
|
"epoch": 3.47, |
|
"grad_norm": 12.638689994812012, |
|
"learning_rate": 1.1111111111111112e-07, |
|
"loss": 0.4269, |
|
"step": 4950 |
|
}, |
|
{ |
|
"epoch": 3.49, |
|
"grad_norm": 11.225275039672852, |
|
"learning_rate": 5.555555555555556e-08, |
|
"loss": 0.4078, |
|
"step": 4975 |
|
}, |
|
{ |
|
"epoch": 3.51, |
|
"grad_norm": 13.034343719482422, |
|
"learning_rate": 0.0, |
|
"loss": 0.3848, |
|
"step": 5000 |
|
}, |
|
{ |
|
"epoch": 3.51, |
|
"eval_loss": 1.1417452096939087, |
|
"eval_runtime": 847.7864, |
|
"eval_samples_per_second": 3.362, |
|
"eval_steps_per_second": 0.421, |
|
"eval_wer": 0.8866461398106967, |
|
"step": 5000 |
|
} |
|
], |
|
"logging_steps": 25, |
|
"max_steps": 5000, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 4, |
|
"save_steps": 1000, |
|
"total_flos": 1.154081874419712e+19, |
|
"train_batch_size": 8, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|