| { | |
| "best_metric": 51.78571428571429, | |
| "best_model_checkpoint": "./whisper-small-bn/checkpoint-4000", | |
| "epoch": 3.3333333333333335, | |
| "eval_steps": 1000, | |
| "global_step": 4000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 27.802248001098633, | |
| "learning_rate": 3.2e-07, | |
| "loss": 2.2811, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 18.253421783447266, | |
| "learning_rate": 7.2e-07, | |
| "loss": 2.1964, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 13.585515975952148, | |
| "learning_rate": 1.12e-06, | |
| "loss": 1.8635, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 22.400362014770508, | |
| "learning_rate": 1.52e-06, | |
| "loss": 1.703, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 13.77535343170166, | |
| "learning_rate": 1.9200000000000003e-06, | |
| "loss": 1.5336, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 14.989215850830078, | |
| "learning_rate": 2.3200000000000002e-06, | |
| "loss": 1.4044, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 22.01858139038086, | |
| "learning_rate": 2.7200000000000002e-06, | |
| "loss": 1.2402, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 16.482311248779297, | |
| "learning_rate": 3.12e-06, | |
| "loss": 1.0922, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 17.529874801635742, | |
| "learning_rate": 3.52e-06, | |
| "loss": 0.9583, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 18.252761840820312, | |
| "learning_rate": 3.920000000000001e-06, | |
| "loss": 0.8783, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 16.48067283630371, | |
| "learning_rate": 4.32e-06, | |
| "loss": 0.7643, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 15.062904357910156, | |
| "learning_rate": 4.7200000000000005e-06, | |
| "loss": 0.6991, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 14.231403350830078, | |
| "learning_rate": 5.12e-06, | |
| "loss": 0.6594, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 18.425230026245117, | |
| "learning_rate": 5.5200000000000005e-06, | |
| "loss": 0.5724, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 11.786844253540039, | |
| "learning_rate": 5.92e-06, | |
| "loss": 0.5687, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 15.346015930175781, | |
| "learning_rate": 6.3200000000000005e-06, | |
| "loss": 0.4696, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 7.936849594116211, | |
| "learning_rate": 6.720000000000001e-06, | |
| "loss": 0.4569, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 13.894779205322266, | |
| "learning_rate": 7.1200000000000004e-06, | |
| "loss": 0.4089, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 11.665081977844238, | |
| "learning_rate": 7.520000000000001e-06, | |
| "loss": 0.4168, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 16.7696475982666, | |
| "learning_rate": 7.92e-06, | |
| "loss": 0.4433, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 15.60254192352295, | |
| "learning_rate": 8.32e-06, | |
| "loss": 0.4013, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 13.72472858428955, | |
| "learning_rate": 8.720000000000001e-06, | |
| "loss": 0.3903, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 10.506444931030273, | |
| "learning_rate": 9.12e-06, | |
| "loss": 0.3581, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 13.643900871276855, | |
| "learning_rate": 9.52e-06, | |
| "loss": 0.3605, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 11.669689178466797, | |
| "learning_rate": 9.920000000000002e-06, | |
| "loss": 0.3548, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 9.907474517822266, | |
| "learning_rate": 9.936000000000001e-06, | |
| "loss": 0.3535, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 11.366904258728027, | |
| "learning_rate": 9.856000000000002e-06, | |
| "loss": 0.2943, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 7.5114240646362305, | |
| "learning_rate": 9.776000000000001e-06, | |
| "loss": 0.3057, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "grad_norm": 11.717443466186523, | |
| "learning_rate": 9.696000000000002e-06, | |
| "loss": 0.2919, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 9.275129318237305, | |
| "learning_rate": 9.616e-06, | |
| "loss": 0.3115, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "grad_norm": 9.388382911682129, | |
| "learning_rate": 9.536000000000002e-06, | |
| "loss": 0.1926, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "grad_norm": 8.574433326721191, | |
| "learning_rate": 9.456e-06, | |
| "loss": 0.2193, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 5.512338638305664, | |
| "learning_rate": 9.376000000000001e-06, | |
| "loss": 0.18, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "grad_norm": 9.644317626953125, | |
| "learning_rate": 9.296e-06, | |
| "loss": 0.1916, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "grad_norm": 9.321474075317383, | |
| "learning_rate": 9.216000000000001e-06, | |
| "loss": 0.2055, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 4.868476390838623, | |
| "learning_rate": 9.136e-06, | |
| "loss": 0.1843, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.47, | |
| "grad_norm": 6.592445373535156, | |
| "learning_rate": 9.056000000000001e-06, | |
| "loss": 0.1713, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "grad_norm": 8.138116836547852, | |
| "learning_rate": 8.976e-06, | |
| "loss": 0.1792, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 9.782510757446289, | |
| "learning_rate": 8.896000000000001e-06, | |
| "loss": 0.1796, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "grad_norm": 5.867671012878418, | |
| "learning_rate": 8.816000000000002e-06, | |
| "loss": 0.1652, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "grad_norm": 6.637588977813721, | |
| "learning_rate": 8.736e-06, | |
| "loss": 0.1735, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 11.652266502380371, | |
| "learning_rate": 8.656000000000001e-06, | |
| "loss": 0.1734, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "grad_norm": 6.051203727722168, | |
| "learning_rate": 8.576e-06, | |
| "loss": 0.1557, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "grad_norm": 11.018319129943848, | |
| "learning_rate": 8.496000000000001e-06, | |
| "loss": 0.1668, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 7.878730297088623, | |
| "learning_rate": 8.416e-06, | |
| "loss": 0.1706, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 3.07, | |
| "grad_norm": 6.223515510559082, | |
| "learning_rate": 8.336000000000001e-06, | |
| "loss": 0.0917, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 3.13, | |
| "grad_norm": 7.449841022491455, | |
| "learning_rate": 8.256e-06, | |
| "loss": 0.0853, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 3.824934244155884, | |
| "learning_rate": 8.176000000000001e-06, | |
| "loss": 0.1033, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 3.27, | |
| "grad_norm": 6.369926929473877, | |
| "learning_rate": 8.096e-06, | |
| "loss": 0.0997, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "grad_norm": 6.938141822814941, | |
| "learning_rate": 8.016e-06, | |
| "loss": 0.095, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "eval_loss": 0.27550819516181946, | |
| "eval_runtime": 345.1723, | |
| "eval_samples_per_second": 0.29, | |
| "eval_steps_per_second": 0.29, | |
| "eval_wer": 60.18099547511312, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "grad_norm": 7.869150638580322, | |
| "learning_rate": 7.936e-06, | |
| "loss": 0.1443, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "grad_norm": 3.6447396278381348, | |
| "learning_rate": 7.860000000000001e-06, | |
| "loss": 0.1974, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "grad_norm": 9.588157653808594, | |
| "learning_rate": 7.78e-06, | |
| "loss": 0.1427, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "grad_norm": 10.796669006347656, | |
| "learning_rate": 7.7e-06, | |
| "loss": 0.187, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "grad_norm": 4.941339015960693, | |
| "learning_rate": 7.620000000000001e-06, | |
| "loss": 0.1568, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 11.114888191223145, | |
| "learning_rate": 7.540000000000001e-06, | |
| "loss": 0.1697, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "grad_norm": 10.200749397277832, | |
| "learning_rate": 7.4600000000000006e-06, | |
| "loss": 0.1675, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 5.67927360534668, | |
| "learning_rate": 7.3800000000000005e-06, | |
| "loss": 0.1557, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "grad_norm": 7.413832664489746, | |
| "learning_rate": 7.304000000000001e-06, | |
| "loss": 0.1629, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "grad_norm": 6.933262348175049, | |
| "learning_rate": 7.224000000000001e-06, | |
| "loss": 0.17, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "grad_norm": 4.847285747528076, | |
| "learning_rate": 7.1440000000000005e-06, | |
| "loss": 0.1389, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "grad_norm": 19.55495834350586, | |
| "learning_rate": 7.0640000000000005e-06, | |
| "loss": 0.1338, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 7.148222923278809, | |
| "learning_rate": 6.984e-06, | |
| "loss": 0.1296, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "grad_norm": 6.323164939880371, | |
| "learning_rate": 6.904e-06, | |
| "loss": 0.163, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "grad_norm": 9.096046447753906, | |
| "learning_rate": 6.824e-06, | |
| "loss": 0.1942, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "grad_norm": 8.462915420532227, | |
| "learning_rate": 6.744e-06, | |
| "loss": 0.1576, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "grad_norm": 7.848359107971191, | |
| "learning_rate": 6.664e-06, | |
| "loss": 0.1219, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.91, | |
| "grad_norm": 6.30383825302124, | |
| "learning_rate": 6.584e-06, | |
| "loss": 0.1404, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 5.657530784606934, | |
| "learning_rate": 6.504e-06, | |
| "loss": 0.1449, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 7.4172234535217285, | |
| "learning_rate": 6.424e-06, | |
| "loss": 0.1234, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "grad_norm": 6.550841331481934, | |
| "learning_rate": 6.344e-06, | |
| "loss": 0.0885, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 3.08, | |
| "grad_norm": 6.212351322174072, | |
| "learning_rate": 6.264e-06, | |
| "loss": 0.0725, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 3.13, | |
| "grad_norm": 5.2562575340271, | |
| "learning_rate": 6.184e-06, | |
| "loss": 0.0886, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "grad_norm": 7.921740531921387, | |
| "learning_rate": 6.104000000000001e-06, | |
| "loss": 0.0805, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 3.21, | |
| "grad_norm": 3.5878069400787354, | |
| "learning_rate": 6.024000000000001e-06, | |
| "loss": 0.0628, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "grad_norm": 8.585208892822266, | |
| "learning_rate": 5.944000000000001e-06, | |
| "loss": 0.0812, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "grad_norm": 4.111368656158447, | |
| "learning_rate": 5.868e-06, | |
| "loss": 0.0694, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 3.34, | |
| "grad_norm": 5.304944038391113, | |
| "learning_rate": 5.788e-06, | |
| "loss": 0.0739, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 3.38, | |
| "grad_norm": 6.168578147888184, | |
| "learning_rate": 5.708e-06, | |
| "loss": 0.0821, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 3.43, | |
| "grad_norm": 12.2472505569458, | |
| "learning_rate": 5.628e-06, | |
| "loss": 0.0885, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 3.47, | |
| "grad_norm": 9.349952697753906, | |
| "learning_rate": 5.548e-06, | |
| "loss": 0.0607, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 3.51, | |
| "grad_norm": 5.98253870010376, | |
| "learning_rate": 5.468e-06, | |
| "loss": 0.0619, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 3.55, | |
| "grad_norm": 7.106723785400391, | |
| "learning_rate": 5.388e-06, | |
| "loss": 0.0992, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 3.693020820617676, | |
| "learning_rate": 5.308000000000001e-06, | |
| "loss": 0.0855, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 3.64, | |
| "grad_norm": 7.3649678230285645, | |
| "learning_rate": 5.228000000000001e-06, | |
| "loss": 0.081, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 3.68, | |
| "grad_norm": 5.4454240798950195, | |
| "learning_rate": 5.1480000000000005e-06, | |
| "loss": 0.0872, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 3.73, | |
| "grad_norm": 8.854460716247559, | |
| "learning_rate": 5.0680000000000004e-06, | |
| "loss": 0.0869, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 3.77, | |
| "grad_norm": 3.8615269660949707, | |
| "learning_rate": 4.988e-06, | |
| "loss": 0.0821, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 3.81, | |
| "grad_norm": 3.7190287113189697, | |
| "learning_rate": 4.908e-06, | |
| "loss": 0.0636, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 3.85, | |
| "grad_norm": 7.683879375457764, | |
| "learning_rate": 4.828e-06, | |
| "loss": 0.0892, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "grad_norm": 3.645141839981079, | |
| "learning_rate": 4.748e-06, | |
| "loss": 0.0803, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 3.94, | |
| "grad_norm": 3.8167316913604736, | |
| "learning_rate": 4.668e-06, | |
| "loss": 0.1012, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 3.98, | |
| "grad_norm": 7.247605323791504, | |
| "learning_rate": 4.588e-06, | |
| "loss": 0.0728, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 4.03, | |
| "grad_norm": 3.5495615005493164, | |
| "learning_rate": 4.508e-06, | |
| "loss": 0.0439, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 4.07, | |
| "grad_norm": 4.105050086975098, | |
| "learning_rate": 4.428000000000001e-06, | |
| "loss": 0.0357, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 4.11, | |
| "grad_norm": 9.399913787841797, | |
| "learning_rate": 4.3480000000000006e-06, | |
| "loss": 0.0363, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 4.15, | |
| "grad_norm": 3.0376510620117188, | |
| "learning_rate": 4.2680000000000005e-06, | |
| "loss": 0.0221, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "grad_norm": 0.6908012628555298, | |
| "learning_rate": 4.188e-06, | |
| "loss": 0.0323, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 4.24, | |
| "grad_norm": 6.255786895751953, | |
| "learning_rate": 4.108e-06, | |
| "loss": 0.0327, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 4.28, | |
| "grad_norm": 7.1336669921875, | |
| "learning_rate": 4.028e-06, | |
| "loss": 0.0416, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 4.28, | |
| "eval_loss": 0.27948442101478577, | |
| "eval_runtime": 349.9899, | |
| "eval_samples_per_second": 0.286, | |
| "eval_steps_per_second": 0.286, | |
| "eval_wer": 59.0146750524109, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 13.214738845825195, | |
| "learning_rate": 7.250909090909092e-06, | |
| "loss": 0.2839, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 8.846616744995117, | |
| "learning_rate": 7.214545454545455e-06, | |
| "loss": 0.2221, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 11.778766632080078, | |
| "learning_rate": 7.178181818181818e-06, | |
| "loss": 0.2394, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 4.484768867492676, | |
| "learning_rate": 7.141818181818182e-06, | |
| "loss": 0.2299, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 4.385435104370117, | |
| "learning_rate": 7.105454545454546e-06, | |
| "loss": 0.2221, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 8.751326560974121, | |
| "learning_rate": 7.06909090909091e-06, | |
| "loss": 0.1966, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 8.837333679199219, | |
| "learning_rate": 7.032727272727273e-06, | |
| "loss": 0.1893, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 9.331002235412598, | |
| "learning_rate": 6.998181818181818e-06, | |
| "loss": 0.2365, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 6.599325180053711, | |
| "learning_rate": 6.961818181818183e-06, | |
| "loss": 0.2152, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 11.716184616088867, | |
| "learning_rate": 6.9254545454545464e-06, | |
| "loss": 0.203, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 6.506683349609375, | |
| "learning_rate": 6.88909090909091e-06, | |
| "loss": 0.2162, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 4.64941930770874, | |
| "learning_rate": 6.852727272727273e-06, | |
| "loss": 0.214, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 8.344452857971191, | |
| "learning_rate": 6.816363636363637e-06, | |
| "loss": 0.2106, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 5.504006385803223, | |
| "learning_rate": 6.780000000000001e-06, | |
| "loss": 0.196, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 9.84432601928711, | |
| "learning_rate": 6.743636363636365e-06, | |
| "loss": 0.2207, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "grad_norm": 7.112894535064697, | |
| "learning_rate": 6.707272727272728e-06, | |
| "loss": 0.2072, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 7.10872745513916, | |
| "learning_rate": 6.670909090909091e-06, | |
| "loss": 0.1925, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 8.612652778625488, | |
| "learning_rate": 6.634545454545455e-06, | |
| "loss": 0.2023, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 6.083488941192627, | |
| "learning_rate": 6.5981818181818195e-06, | |
| "loss": 0.2111, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 7.59145975112915, | |
| "learning_rate": 6.561818181818182e-06, | |
| "loss": 0.1934, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 6.529219627380371, | |
| "learning_rate": 6.525454545454546e-06, | |
| "loss": 0.1756, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 13.685791015625, | |
| "learning_rate": 6.48909090909091e-06, | |
| "loss": 0.1872, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "grad_norm": 6.508156776428223, | |
| "learning_rate": 6.4527272727272725e-06, | |
| "loss": 0.1424, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "grad_norm": 4.847611427307129, | |
| "learning_rate": 6.416363636363637e-06, | |
| "loss": 0.1588, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 5.715017795562744, | |
| "learning_rate": 6.380000000000001e-06, | |
| "loss": 0.1896, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 10.562045097351074, | |
| "learning_rate": 6.3436363636363644e-06, | |
| "loss": 0.2007, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "grad_norm": 3.2870736122131348, | |
| "learning_rate": 6.307272727272727e-06, | |
| "loss": 0.179, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "grad_norm": 11.021241188049316, | |
| "learning_rate": 6.270909090909092e-06, | |
| "loss": 0.1573, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "grad_norm": 7.566544055938721, | |
| "learning_rate": 6.2345454545454555e-06, | |
| "loss": 0.1832, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "grad_norm": 5.980641841888428, | |
| "learning_rate": 6.198181818181819e-06, | |
| "loss": 0.1498, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "grad_norm": 5.697175025939941, | |
| "learning_rate": 6.161818181818182e-06, | |
| "loss": 0.1378, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 4.451406002044678, | |
| "learning_rate": 6.125454545454546e-06, | |
| "loss": 0.152, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "grad_norm": 5.854424476623535, | |
| "learning_rate": 6.08909090909091e-06, | |
| "loss": 0.1492, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "grad_norm": 6.480546951293945, | |
| "learning_rate": 6.052727272727274e-06, | |
| "loss": 0.1501, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "grad_norm": 3.3347442150115967, | |
| "learning_rate": 6.016363636363637e-06, | |
| "loss": 0.132, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "grad_norm": 5.9716081619262695, | |
| "learning_rate": 5.98e-06, | |
| "loss": 0.1483, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "grad_norm": 5.842201232910156, | |
| "learning_rate": 5.943636363636364e-06, | |
| "loss": 0.1688, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 5.079952716827393, | |
| "learning_rate": 5.9072727272727285e-06, | |
| "loss": 0.167, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 9.07481575012207, | |
| "learning_rate": 5.870909090909091e-06, | |
| "loss": 0.1581, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "grad_norm": 9.30324935913086, | |
| "learning_rate": 5.834545454545455e-06, | |
| "loss": 0.1704, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "grad_norm": 4.848841667175293, | |
| "learning_rate": 5.798181818181819e-06, | |
| "loss": 0.1894, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "grad_norm": 7.52855110168457, | |
| "learning_rate": 5.7618181818181816e-06, | |
| "loss": 0.1702, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "grad_norm": 6.624313831329346, | |
| "learning_rate": 5.725454545454546e-06, | |
| "loss": 0.1628, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 6.889793872833252, | |
| "learning_rate": 5.68909090909091e-06, | |
| "loss": 0.1705, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "grad_norm": 5.635292053222656, | |
| "learning_rate": 5.6527272727272734e-06, | |
| "loss": 0.1531, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "grad_norm": 4.1545729637146, | |
| "learning_rate": 5.616363636363636e-06, | |
| "loss": 0.1371, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "grad_norm": 7.1985697746276855, | |
| "learning_rate": 5.580000000000001e-06, | |
| "loss": 0.1571, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 2.47, | |
| "grad_norm": 7.112518310546875, | |
| "learning_rate": 5.5436363636363645e-06, | |
| "loss": 0.1525, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 9.503728866577148, | |
| "learning_rate": 5.507272727272728e-06, | |
| "loss": 0.1426, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 5.497040271759033, | |
| "learning_rate": 5.470909090909091e-06, | |
| "loss": 0.1501, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "eval_loss": 0.2098405510187149, | |
| "eval_runtime": 325.0771, | |
| "eval_samples_per_second": 0.308, | |
| "eval_steps_per_second": 0.308, | |
| "eval_wer": 55.5672268907563, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "grad_norm": 6.7488603591918945, | |
| "learning_rate": 5.434545454545455e-06, | |
| "loss": 0.1647, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "grad_norm": 5.16122579574585, | |
| "learning_rate": 5.398181818181819e-06, | |
| "loss": 0.1617, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "grad_norm": 4.19830322265625, | |
| "learning_rate": 5.361818181818183e-06, | |
| "loss": 0.1484, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "grad_norm": 9.339705467224121, | |
| "learning_rate": 5.325454545454546e-06, | |
| "loss": 0.153, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "grad_norm": 5.106137752532959, | |
| "learning_rate": 5.289090909090909e-06, | |
| "loss": 0.141, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 8.092510223388672, | |
| "learning_rate": 5.252727272727273e-06, | |
| "loss": 0.1443, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "grad_norm": 5.475297451019287, | |
| "learning_rate": 5.2163636363636376e-06, | |
| "loss": 0.1638, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 2.63, | |
| "grad_norm": 8.545796394348145, | |
| "learning_rate": 5.18e-06, | |
| "loss": 0.1382, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "grad_norm": 5.189711093902588, | |
| "learning_rate": 5.143636363636364e-06, | |
| "loss": 0.1492, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "grad_norm": 7.558894157409668, | |
| "learning_rate": 5.107272727272728e-06, | |
| "loss": 0.1559, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 2.68, | |
| "grad_norm": 5.985530853271484, | |
| "learning_rate": 5.0709090909090906e-06, | |
| "loss": 0.1605, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 6.780598163604736, | |
| "learning_rate": 5.034545454545455e-06, | |
| "loss": 0.148, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "grad_norm": 6.225440979003906, | |
| "learning_rate": 4.998181818181819e-06, | |
| "loss": 0.1451, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "grad_norm": 5.706241607666016, | |
| "learning_rate": 4.9618181818181824e-06, | |
| "loss": 0.1814, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 7.860342502593994, | |
| "learning_rate": 4.925454545454546e-06, | |
| "loss": 0.1407, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "grad_norm": 4.965442180633545, | |
| "learning_rate": 4.88909090909091e-06, | |
| "loss": 0.1494, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "grad_norm": 9.562829971313477, | |
| "learning_rate": 4.8527272727272735e-06, | |
| "loss": 0.1577, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 6.261291027069092, | |
| "learning_rate": 4.816363636363637e-06, | |
| "loss": 0.1561, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "grad_norm": 5.747939109802246, | |
| "learning_rate": 4.78e-06, | |
| "loss": 0.1344, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "grad_norm": 8.154770851135254, | |
| "learning_rate": 4.7436363636363645e-06, | |
| "loss": 0.1419, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "grad_norm": 5.976426601409912, | |
| "learning_rate": 4.707272727272727e-06, | |
| "loss": 0.1314, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "grad_norm": 4.834322452545166, | |
| "learning_rate": 4.670909090909092e-06, | |
| "loss": 0.1128, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 8.85062026977539, | |
| "learning_rate": 4.634545454545455e-06, | |
| "loss": 0.1718, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 3.992954969406128, | |
| "learning_rate": 4.598181818181818e-06, | |
| "loss": 0.1341, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 6.488000392913818, | |
| "learning_rate": 4.561818181818182e-06, | |
| "loss": 0.1401, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "grad_norm": 4.876968860626221, | |
| "learning_rate": 4.525454545454546e-06, | |
| "loss": 0.1319, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 2.95, | |
| "grad_norm": 7.407408237457275, | |
| "learning_rate": 4.489090909090909e-06, | |
| "loss": 0.1377, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "grad_norm": 5.017177104949951, | |
| "learning_rate": 4.452727272727273e-06, | |
| "loss": 0.1732, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 2.98, | |
| "grad_norm": 8.190740585327148, | |
| "learning_rate": 4.416363636363637e-06, | |
| "loss": 0.1225, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 7.947383403778076, | |
| "learning_rate": 4.38e-06, | |
| "loss": 0.1454, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "grad_norm": 5.404466152191162, | |
| "learning_rate": 4.343636363636364e-06, | |
| "loss": 0.0746, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "grad_norm": 5.470146179199219, | |
| "learning_rate": 4.307272727272728e-06, | |
| "loss": 0.0725, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 3.05, | |
| "grad_norm": 4.404048919677734, | |
| "learning_rate": 4.2709090909090914e-06, | |
| "loss": 0.088, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 3.07, | |
| "grad_norm": 3.169548273086548, | |
| "learning_rate": 4.234545454545455e-06, | |
| "loss": 0.0811, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 3.08, | |
| "grad_norm": 2.710052967071533, | |
| "learning_rate": 4.198181818181819e-06, | |
| "loss": 0.0646, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 5.970617294311523, | |
| "learning_rate": 4.1618181818181825e-06, | |
| "loss": 0.0789, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "grad_norm": 4.657149791717529, | |
| "learning_rate": 4.125454545454546e-06, | |
| "loss": 0.0814, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 3.13, | |
| "grad_norm": 4.4107666015625, | |
| "learning_rate": 4.089090909090909e-06, | |
| "loss": 0.0831, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 3.15, | |
| "grad_norm": 6.185523509979248, | |
| "learning_rate": 4.0527272727272735e-06, | |
| "loss": 0.0672, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "grad_norm": 3.5439445972442627, | |
| "learning_rate": 4.016363636363636e-06, | |
| "loss": 0.0839, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 3.18, | |
| "grad_norm": 5.79164981842041, | |
| "learning_rate": 3.980000000000001e-06, | |
| "loss": 0.0692, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 6.883912086486816, | |
| "learning_rate": 3.943636363636364e-06, | |
| "loss": 0.0798, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "grad_norm": 4.028432846069336, | |
| "learning_rate": 3.907272727272727e-06, | |
| "loss": 0.0808, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "grad_norm": 4.477746486663818, | |
| "learning_rate": 3.870909090909091e-06, | |
| "loss": 0.0781, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "grad_norm": 2.0900027751922607, | |
| "learning_rate": 3.834545454545455e-06, | |
| "loss": 0.073, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 3.27, | |
| "grad_norm": 2.609961748123169, | |
| "learning_rate": 3.798181818181819e-06, | |
| "loss": 0.0961, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 3.28, | |
| "grad_norm": 6.023408889770508, | |
| "learning_rate": 3.761818181818182e-06, | |
| "loss": 0.0806, | |
| "step": 3940 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "grad_norm": 5.091029167175293, | |
| "learning_rate": 3.725454545454546e-06, | |
| "loss": 0.0681, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 3.32, | |
| "grad_norm": 5.474444389343262, | |
| "learning_rate": 3.6890909090909094e-06, | |
| "loss": 0.0817, | |
| "step": 3980 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "grad_norm": 5.174444198608398, | |
| "learning_rate": 3.6527272727272727e-06, | |
| "loss": 0.0827, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "eval_loss": 0.20179511606693268, | |
| "eval_runtime": 319.617, | |
| "eval_samples_per_second": 0.313, | |
| "eval_steps_per_second": 0.313, | |
| "eval_wer": 51.78571428571429, | |
| "step": 4000 | |
| } | |
| ], | |
| "logging_steps": 20, | |
| "max_steps": 6000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 1000, | |
| "total_flos": 3.46244763303936e+18, | |
| "train_batch_size": 3, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |