| { | |
| "best_metric": 10.634930784232232, | |
| "best_model_checkpoint": "./whisper-small-mix-pt/checkpoint-4000", | |
| "epoch": 2.062, | |
| "eval_steps": 1000, | |
| "global_step": 5000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.005, | |
| "grad_norm": 8.937539100646973, | |
| "learning_rate": 4.6000000000000004e-07, | |
| "loss": 0.8175, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 4.877315044403076, | |
| "learning_rate": 9.600000000000001e-07, | |
| "loss": 0.7318, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.015, | |
| "grad_norm": 4.961461544036865, | |
| "learning_rate": 1.46e-06, | |
| "loss": 1.0495, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.423418045043945, | |
| "learning_rate": 1.9600000000000003e-06, | |
| "loss": 0.7303, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.025, | |
| "grad_norm": 3.701725482940674, | |
| "learning_rate": 2.46e-06, | |
| "loss": 0.5486, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 3.901665449142456, | |
| "learning_rate": 2.96e-06, | |
| "loss": 0.4327, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.035, | |
| "grad_norm": 2.7568914890289307, | |
| "learning_rate": 3.46e-06, | |
| "loss": 0.3445, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 3.0615532398223877, | |
| "learning_rate": 3.96e-06, | |
| "loss": 0.317, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.045, | |
| "grad_norm": 2.577845573425293, | |
| "learning_rate": 4.4600000000000005e-06, | |
| "loss": 0.2995, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 2.810206890106201, | |
| "learning_rate": 4.960000000000001e-06, | |
| "loss": 0.2701, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.055, | |
| "grad_norm": 2.9507219791412354, | |
| "learning_rate": 5.460000000000001e-06, | |
| "loss": 0.2275, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 2.8784899711608887, | |
| "learning_rate": 5.9600000000000005e-06, | |
| "loss": 0.4164, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.065, | |
| "grad_norm": 2.7221858501434326, | |
| "learning_rate": 6.460000000000001e-06, | |
| "loss": 0.2831, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 2.760021209716797, | |
| "learning_rate": 6.96e-06, | |
| "loss": 0.2202, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.075, | |
| "grad_norm": 2.8915979862213135, | |
| "learning_rate": 7.4600000000000006e-06, | |
| "loss": 0.2165, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 2.7558434009552, | |
| "learning_rate": 7.960000000000002e-06, | |
| "loss": 0.207, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.085, | |
| "grad_norm": 2.5287721157073975, | |
| "learning_rate": 8.46e-06, | |
| "loss": 0.2529, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 3.026515483856201, | |
| "learning_rate": 8.96e-06, | |
| "loss": 0.2134, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.095, | |
| "grad_norm": 2.3035366535186768, | |
| "learning_rate": 9.460000000000001e-06, | |
| "loss": 0.1968, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.298630952835083, | |
| "learning_rate": 9.960000000000001e-06, | |
| "loss": 0.1881, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.105, | |
| "grad_norm": 2.209261417388916, | |
| "learning_rate": 9.94888888888889e-06, | |
| "loss": 0.162, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 2.819951295852661, | |
| "learning_rate": 9.893333333333334e-06, | |
| "loss": 0.1646, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.115, | |
| "grad_norm": 2.437563896179199, | |
| "learning_rate": 9.837777777777778e-06, | |
| "loss": 0.1705, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 2.5952935218811035, | |
| "learning_rate": 9.782222222222222e-06, | |
| "loss": 0.1938, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.125, | |
| "grad_norm": 2.4692909717559814, | |
| "learning_rate": 9.726666666666668e-06, | |
| "loss": 0.1869, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 2.4430558681488037, | |
| "learning_rate": 9.671111111111112e-06, | |
| "loss": 0.1788, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.135, | |
| "grad_norm": 2.2733707427978516, | |
| "learning_rate": 9.615555555555558e-06, | |
| "loss": 0.1934, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 2.256990909576416, | |
| "learning_rate": 9.56e-06, | |
| "loss": 0.1622, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.145, | |
| "grad_norm": 2.216128349304199, | |
| "learning_rate": 9.504444444444446e-06, | |
| "loss": 0.157, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 2.0830190181732178, | |
| "learning_rate": 9.44888888888889e-06, | |
| "loss": 0.169, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.155, | |
| "grad_norm": 2.085129737854004, | |
| "learning_rate": 9.393333333333334e-06, | |
| "loss": 0.1548, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 2.3641719818115234, | |
| "learning_rate": 9.33777777777778e-06, | |
| "loss": 0.1722, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.165, | |
| "grad_norm": 3.0489625930786133, | |
| "learning_rate": 9.282222222222222e-06, | |
| "loss": 0.2306, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 2.5903120040893555, | |
| "learning_rate": 9.226666666666668e-06, | |
| "loss": 0.2255, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.175, | |
| "grad_norm": 2.470798969268799, | |
| "learning_rate": 9.171111111111112e-06, | |
| "loss": 0.1983, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.7610605955123901, | |
| "learning_rate": 9.115555555555556e-06, | |
| "loss": 0.1991, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.185, | |
| "grad_norm": 2.5426924228668213, | |
| "learning_rate": 9.060000000000001e-06, | |
| "loss": 0.1584, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 2.5950539112091064, | |
| "learning_rate": 9.004444444444445e-06, | |
| "loss": 0.1707, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.195, | |
| "grad_norm": 2.2430670261383057, | |
| "learning_rate": 8.94888888888889e-06, | |
| "loss": 0.1682, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 2.1364951133728027, | |
| "learning_rate": 8.893333333333333e-06, | |
| "loss": 0.1575, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "eval_loss": 0.21252164244651794, | |
| "eval_runtime": 385.7404, | |
| "eval_samples_per_second": 24.542, | |
| "eval_steps_per_second": 3.069, | |
| "eval_wer": 12.185491863550626, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.205, | |
| "grad_norm": 2.343320846557617, | |
| "learning_rate": 8.83777777777778e-06, | |
| "loss": 0.1679, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 2.078854560852051, | |
| "learning_rate": 8.782222222222223e-06, | |
| "loss": 0.1688, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.215, | |
| "grad_norm": 2.009870767593384, | |
| "learning_rate": 8.726666666666667e-06, | |
| "loss": 0.144, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 1.4615126848220825, | |
| "learning_rate": 8.671111111111113e-06, | |
| "loss": 0.127, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.225, | |
| "grad_norm": 2.175366163253784, | |
| "learning_rate": 8.615555555555555e-06, | |
| "loss": 0.1191, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 1.3462347984313965, | |
| "learning_rate": 8.560000000000001e-06, | |
| "loss": 0.1214, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.235, | |
| "grad_norm": 1.7791441679000854, | |
| "learning_rate": 8.504444444444445e-06, | |
| "loss": 0.1213, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 3.0184803009033203, | |
| "learning_rate": 8.448888888888889e-06, | |
| "loss": 0.3185, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.245, | |
| "grad_norm": 2.1637368202209473, | |
| "learning_rate": 8.393333333333335e-06, | |
| "loss": 0.2747, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 3.534830093383789, | |
| "learning_rate": 8.337777777777777e-06, | |
| "loss": 0.1535, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.255, | |
| "grad_norm": 3.3045480251312256, | |
| "learning_rate": 8.282222222222223e-06, | |
| "loss": 0.6384, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 2.566434383392334, | |
| "learning_rate": 8.226666666666667e-06, | |
| "loss": 0.4608, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.265, | |
| "grad_norm": 3.436089038848877, | |
| "learning_rate": 8.171111111111113e-06, | |
| "loss": 0.3286, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 2.8783295154571533, | |
| "learning_rate": 8.115555555555557e-06, | |
| "loss": 0.3508, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.275, | |
| "grad_norm": 3.244107723236084, | |
| "learning_rate": 8.06e-06, | |
| "loss": 0.3174, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 2.4489357471466064, | |
| "learning_rate": 8.004444444444445e-06, | |
| "loss": 0.2287, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.285, | |
| "grad_norm": 2.270322561264038, | |
| "learning_rate": 7.948888888888889e-06, | |
| "loss": 0.1828, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 2.105621814727783, | |
| "learning_rate": 7.893333333333335e-06, | |
| "loss": 0.1336, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.295, | |
| "grad_norm": 1.8718715906143188, | |
| "learning_rate": 7.837777777777779e-06, | |
| "loss": 0.1358, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 1.595679759979248, | |
| "learning_rate": 7.782222222222223e-06, | |
| "loss": 0.1244, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.305, | |
| "grad_norm": 1.979271411895752, | |
| "learning_rate": 7.726666666666667e-06, | |
| "loss": 0.1222, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 2.024198532104492, | |
| "learning_rate": 7.67111111111111e-06, | |
| "loss": 0.1274, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.315, | |
| "grad_norm": 1.9602785110473633, | |
| "learning_rate": 7.6155555555555564e-06, | |
| "loss": 0.1243, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 2.508364677429199, | |
| "learning_rate": 7.5600000000000005e-06, | |
| "loss": 0.1896, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.325, | |
| "grad_norm": 2.0765221118927, | |
| "learning_rate": 7.504444444444445e-06, | |
| "loss": 0.1707, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 2.030231475830078, | |
| "learning_rate": 7.44888888888889e-06, | |
| "loss": 0.1341, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.335, | |
| "grad_norm": 2.9829025268554688, | |
| "learning_rate": 7.393333333333333e-06, | |
| "loss": 0.1305, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 2.1654741764068604, | |
| "learning_rate": 7.337777777777778e-06, | |
| "loss": 0.122, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.345, | |
| "grad_norm": 2.106933832168579, | |
| "learning_rate": 7.282222222222222e-06, | |
| "loss": 0.1427, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 3.380559206008911, | |
| "learning_rate": 7.226666666666667e-06, | |
| "loss": 0.3227, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.355, | |
| "grad_norm": 2.084108829498291, | |
| "learning_rate": 7.171111111111112e-06, | |
| "loss": 0.194, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 2.1483328342437744, | |
| "learning_rate": 7.115555555555557e-06, | |
| "loss": 0.1193, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.365, | |
| "grad_norm": 1.6093554496765137, | |
| "learning_rate": 7.06e-06, | |
| "loss": 0.0958, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 1.4673306941986084, | |
| "learning_rate": 7.004444444444445e-06, | |
| "loss": 0.0896, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.375, | |
| "grad_norm": 1.4364008903503418, | |
| "learning_rate": 6.948888888888889e-06, | |
| "loss": 0.0897, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 1.4021852016448975, | |
| "learning_rate": 6.893333333333334e-06, | |
| "loss": 0.0772, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.385, | |
| "grad_norm": 1.5660924911499023, | |
| "learning_rate": 6.837777777777779e-06, | |
| "loss": 0.0773, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 1.9443961381912231, | |
| "learning_rate": 6.782222222222222e-06, | |
| "loss": 0.1027, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.395, | |
| "grad_norm": 2.002857208251953, | |
| "learning_rate": 6.726666666666667e-06, | |
| "loss": 0.1945, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 2.1631336212158203, | |
| "learning_rate": 6.671111111111112e-06, | |
| "loss": 0.1986, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "eval_loss": 0.20620137453079224, | |
| "eval_runtime": 384.1444, | |
| "eval_samples_per_second": 24.644, | |
| "eval_steps_per_second": 3.082, | |
| "eval_wer": 11.570062981553118, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.405, | |
| "grad_norm": 1.9269295930862427, | |
| "learning_rate": 6.615555555555556e-06, | |
| "loss": 0.1477, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 1.8260035514831543, | |
| "learning_rate": 6.560000000000001e-06, | |
| "loss": 0.0935, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.415, | |
| "grad_norm": 1.3721944093704224, | |
| "learning_rate": 6.504444444444446e-06, | |
| "loss": 0.0801, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 1.8245339393615723, | |
| "learning_rate": 6.448888888888889e-06, | |
| "loss": 0.1047, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.425, | |
| "grad_norm": 1.5538787841796875, | |
| "learning_rate": 6.393333333333334e-06, | |
| "loss": 0.0879, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 1.519700527191162, | |
| "learning_rate": 6.3377777777777786e-06, | |
| "loss": 0.0813, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.435, | |
| "grad_norm": 3.4600226879119873, | |
| "learning_rate": 6.282222222222223e-06, | |
| "loss": 0.1835, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 3.3525190353393555, | |
| "learning_rate": 6.2266666666666675e-06, | |
| "loss": 0.6174, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.445, | |
| "grad_norm": 2.3610615730285645, | |
| "learning_rate": 6.171111111111112e-06, | |
| "loss": 0.6218, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 1.5530322790145874, | |
| "learning_rate": 6.1155555555555555e-06, | |
| "loss": 0.2095, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.455, | |
| "grad_norm": 1.4927952289581299, | |
| "learning_rate": 6.0600000000000004e-06, | |
| "loss": 0.0967, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 1.6570786237716675, | |
| "learning_rate": 6.004444444444445e-06, | |
| "loss": 0.0895, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.465, | |
| "grad_norm": 1.5751352310180664, | |
| "learning_rate": 5.948888888888889e-06, | |
| "loss": 0.0878, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 1.001, | |
| "grad_norm": 2.1557564735412598, | |
| "learning_rate": 5.893333333333334e-06, | |
| "loss": 0.095, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 1.006, | |
| "grad_norm": 2.489004611968994, | |
| "learning_rate": 5.837777777777777e-06, | |
| "loss": 0.1279, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 1.011, | |
| "grad_norm": 3.284184217453003, | |
| "learning_rate": 5.782222222222222e-06, | |
| "loss": 0.1914, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.016, | |
| "grad_norm": 3.20249080657959, | |
| "learning_rate": 5.726666666666667e-06, | |
| "loss": 0.526, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 1.021, | |
| "grad_norm": 2.9400064945220947, | |
| "learning_rate": 5.671111111111112e-06, | |
| "loss": 0.2781, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 1.026, | |
| "grad_norm": 1.9436590671539307, | |
| "learning_rate": 5.615555555555556e-06, | |
| "loss": 0.1756, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 1.031, | |
| "grad_norm": 2.086899757385254, | |
| "learning_rate": 5.560000000000001e-06, | |
| "loss": 0.1469, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.036, | |
| "grad_norm": 1.6339154243469238, | |
| "learning_rate": 5.504444444444444e-06, | |
| "loss": 0.1348, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 1.041, | |
| "grad_norm": 2.1945927143096924, | |
| "learning_rate": 5.448888888888889e-06, | |
| "loss": 0.112, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 1.046, | |
| "grad_norm": 1.639339804649353, | |
| "learning_rate": 5.393333333333334e-06, | |
| "loss": 0.1134, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 1.051, | |
| "grad_norm": 1.7829625606536865, | |
| "learning_rate": 5.337777777777779e-06, | |
| "loss": 0.098, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.056, | |
| "grad_norm": 1.5553560256958008, | |
| "learning_rate": 5.282222222222223e-06, | |
| "loss": 0.0706, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 1.061, | |
| "grad_norm": 2.2982029914855957, | |
| "learning_rate": 5.226666666666667e-06, | |
| "loss": 0.2427, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 1.066, | |
| "grad_norm": 1.794986367225647, | |
| "learning_rate": 5.171111111111111e-06, | |
| "loss": 0.116, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 1.071, | |
| "grad_norm": 1.2313064336776733, | |
| "learning_rate": 5.115555555555556e-06, | |
| "loss": 0.074, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 1.076, | |
| "grad_norm": 1.4959907531738281, | |
| "learning_rate": 5.060000000000001e-06, | |
| "loss": 0.0758, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 1.081, | |
| "grad_norm": 1.216587781906128, | |
| "learning_rate": 5.004444444444445e-06, | |
| "loss": 0.0787, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 1.086, | |
| "grad_norm": 1.8736680746078491, | |
| "learning_rate": 4.94888888888889e-06, | |
| "loss": 0.1223, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 1.091, | |
| "grad_norm": 1.2765681743621826, | |
| "learning_rate": 4.893333333333334e-06, | |
| "loss": 0.1025, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.096, | |
| "grad_norm": 1.614762544631958, | |
| "learning_rate": 4.837777777777778e-06, | |
| "loss": 0.089, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 1.101, | |
| "grad_norm": 1.6263494491577148, | |
| "learning_rate": 4.7822222222222226e-06, | |
| "loss": 0.1046, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 1.106, | |
| "grad_norm": 1.6685870885849, | |
| "learning_rate": 4.7266666666666674e-06, | |
| "loss": 0.0845, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 1.111, | |
| "grad_norm": 1.2902885675430298, | |
| "learning_rate": 4.6711111111111115e-06, | |
| "loss": 0.0806, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 1.116, | |
| "grad_norm": 1.8994227647781372, | |
| "learning_rate": 4.6155555555555555e-06, | |
| "loss": 0.0767, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 1.121, | |
| "grad_norm": 1.3723390102386475, | |
| "learning_rate": 4.56e-06, | |
| "loss": 0.0716, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 1.126, | |
| "grad_norm": 1.6283633708953857, | |
| "learning_rate": 4.504444444444444e-06, | |
| "loss": 0.0731, | |
| "step": 2975 | |
| }, | |
| { | |
| "epoch": 1.131, | |
| "grad_norm": 1.3962900638580322, | |
| "learning_rate": 4.448888888888889e-06, | |
| "loss": 0.0942, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.131, | |
| "eval_loss": 0.19787383079528809, | |
| "eval_runtime": 384.7545, | |
| "eval_samples_per_second": 24.605, | |
| "eval_steps_per_second": 3.077, | |
| "eval_wer": 11.015377729467055, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.1360000000000001, | |
| "grad_norm": 1.321205973625183, | |
| "learning_rate": 4.393333333333334e-06, | |
| "loss": 0.0731, | |
| "step": 3025 | |
| }, | |
| { | |
| "epoch": 1.141, | |
| "grad_norm": 1.2057939767837524, | |
| "learning_rate": 4.337777777777778e-06, | |
| "loss": 0.0646, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 1.146, | |
| "grad_norm": 1.0545603036880493, | |
| "learning_rate": 4.282222222222222e-06, | |
| "loss": 0.0628, | |
| "step": 3075 | |
| }, | |
| { | |
| "epoch": 1.151, | |
| "grad_norm": 1.363511562347412, | |
| "learning_rate": 4.226666666666667e-06, | |
| "loss": 0.0614, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 1.156, | |
| "grad_norm": 1.1930261850357056, | |
| "learning_rate": 4.171111111111111e-06, | |
| "loss": 0.0634, | |
| "step": 3125 | |
| }, | |
| { | |
| "epoch": 1.161, | |
| "grad_norm": 3.0981078147888184, | |
| "learning_rate": 4.115555555555556e-06, | |
| "loss": 0.2114, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 1.166, | |
| "grad_norm": 1.6411216259002686, | |
| "learning_rate": 4.060000000000001e-06, | |
| "loss": 0.1761, | |
| "step": 3175 | |
| }, | |
| { | |
| "epoch": 1.171, | |
| "grad_norm": 1.1803615093231201, | |
| "learning_rate": 4.004444444444445e-06, | |
| "loss": 0.0729, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 1.176, | |
| "grad_norm": 3.052395820617676, | |
| "learning_rate": 3.948888888888889e-06, | |
| "loss": 0.3801, | |
| "step": 3225 | |
| }, | |
| { | |
| "epoch": 1.181, | |
| "grad_norm": 2.7270925045013428, | |
| "learning_rate": 3.893333333333333e-06, | |
| "loss": 0.3596, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 1.186, | |
| "grad_norm": 1.7751537561416626, | |
| "learning_rate": 3.837777777777778e-06, | |
| "loss": 0.2001, | |
| "step": 3275 | |
| }, | |
| { | |
| "epoch": 1.191, | |
| "grad_norm": 1.7908082008361816, | |
| "learning_rate": 3.782222222222223e-06, | |
| "loss": 0.1933, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 1.196, | |
| "grad_norm": 1.8746870756149292, | |
| "learning_rate": 3.726666666666667e-06, | |
| "loss": 0.2034, | |
| "step": 3325 | |
| }, | |
| { | |
| "epoch": 1.201, | |
| "grad_norm": 1.8763881921768188, | |
| "learning_rate": 3.6711111111111113e-06, | |
| "loss": 0.1608, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 1.206, | |
| "grad_norm": 1.5297502279281616, | |
| "learning_rate": 3.615555555555556e-06, | |
| "loss": 0.1099, | |
| "step": 3375 | |
| }, | |
| { | |
| "epoch": 1.211, | |
| "grad_norm": 1.4193453788757324, | |
| "learning_rate": 3.5600000000000002e-06, | |
| "loss": 0.079, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 1.216, | |
| "grad_norm": 1.5194439888000488, | |
| "learning_rate": 3.5044444444444447e-06, | |
| "loss": 0.0652, | |
| "step": 3425 | |
| }, | |
| { | |
| "epoch": 1.221, | |
| "grad_norm": 1.3446123600006104, | |
| "learning_rate": 3.4488888888888896e-06, | |
| "loss": 0.0623, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 1.226, | |
| "grad_norm": 1.293591856956482, | |
| "learning_rate": 3.3933333333333336e-06, | |
| "loss": 0.063, | |
| "step": 3475 | |
| }, | |
| { | |
| "epoch": 1.231, | |
| "grad_norm": 1.3113939762115479, | |
| "learning_rate": 3.337777777777778e-06, | |
| "loss": 0.0691, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.236, | |
| "grad_norm": 1.8869807720184326, | |
| "learning_rate": 3.282222222222223e-06, | |
| "loss": 0.0697, | |
| "step": 3525 | |
| }, | |
| { | |
| "epoch": 1.241, | |
| "grad_norm": 1.311045527458191, | |
| "learning_rate": 3.226666666666667e-06, | |
| "loss": 0.1023, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 1.246, | |
| "grad_norm": 1.4343574047088623, | |
| "learning_rate": 3.1711111111111114e-06, | |
| "loss": 0.0833, | |
| "step": 3575 | |
| }, | |
| { | |
| "epoch": 1.251, | |
| "grad_norm": 1.6197088956832886, | |
| "learning_rate": 3.1155555555555555e-06, | |
| "loss": 0.0812, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 1.256, | |
| "grad_norm": 1.4370137453079224, | |
| "learning_rate": 3.0600000000000003e-06, | |
| "loss": 0.0666, | |
| "step": 3625 | |
| }, | |
| { | |
| "epoch": 1.2610000000000001, | |
| "grad_norm": 1.3375777006149292, | |
| "learning_rate": 3.004444444444445e-06, | |
| "loss": 0.058, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 1.266, | |
| "grad_norm": 1.6165860891342163, | |
| "learning_rate": 2.948888888888889e-06, | |
| "loss": 0.0736, | |
| "step": 3675 | |
| }, | |
| { | |
| "epoch": 1.271, | |
| "grad_norm": 2.2871458530426025, | |
| "learning_rate": 2.8933333333333337e-06, | |
| "loss": 0.0943, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 1.276, | |
| "grad_norm": 1.4543273448944092, | |
| "learning_rate": 2.837777777777778e-06, | |
| "loss": 0.0904, | |
| "step": 3725 | |
| }, | |
| { | |
| "epoch": 1.2810000000000001, | |
| "grad_norm": 1.64260995388031, | |
| "learning_rate": 2.7822222222222222e-06, | |
| "loss": 0.0736, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 1.286, | |
| "grad_norm": 1.1954628229141235, | |
| "learning_rate": 2.726666666666667e-06, | |
| "loss": 0.0634, | |
| "step": 3775 | |
| }, | |
| { | |
| "epoch": 1.291, | |
| "grad_norm": 1.245037317276001, | |
| "learning_rate": 2.6711111111111116e-06, | |
| "loss": 0.0759, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 1.296, | |
| "grad_norm": 1.1573712825775146, | |
| "learning_rate": 2.6155555555555556e-06, | |
| "loss": 0.0695, | |
| "step": 3825 | |
| }, | |
| { | |
| "epoch": 1.301, | |
| "grad_norm": 1.0521385669708252, | |
| "learning_rate": 2.56e-06, | |
| "loss": 0.058, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 1.306, | |
| "grad_norm": 1.6519237756729126, | |
| "learning_rate": 2.504444444444445e-06, | |
| "loss": 0.0706, | |
| "step": 3875 | |
| }, | |
| { | |
| "epoch": 1.311, | |
| "grad_norm": 1.3316805362701416, | |
| "learning_rate": 2.448888888888889e-06, | |
| "loss": 0.0681, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 1.316, | |
| "grad_norm": 1.8995634317398071, | |
| "learning_rate": 2.3933333333333334e-06, | |
| "loss": 0.1458, | |
| "step": 3925 | |
| }, | |
| { | |
| "epoch": 1.321, | |
| "grad_norm": 1.9890100955963135, | |
| "learning_rate": 2.337777777777778e-06, | |
| "loss": 0.1334, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 1.326, | |
| "grad_norm": 1.2592449188232422, | |
| "learning_rate": 2.2822222222222223e-06, | |
| "loss": 0.1247, | |
| "step": 3975 | |
| }, | |
| { | |
| "epoch": 1.331, | |
| "grad_norm": 1.3058019876480103, | |
| "learning_rate": 2.226666666666667e-06, | |
| "loss": 0.0577, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.331, | |
| "eval_loss": 0.2000385820865631, | |
| "eval_runtime": 386.3157, | |
| "eval_samples_per_second": 24.506, | |
| "eval_steps_per_second": 3.065, | |
| "eval_wer": 10.634930784232232, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.336, | |
| "grad_norm": 1.0941780805587769, | |
| "learning_rate": 2.1711111111111113e-06, | |
| "loss": 0.0498, | |
| "step": 4025 | |
| }, | |
| { | |
| "epoch": 1.341, | |
| "grad_norm": 1.407287836074829, | |
| "learning_rate": 2.1155555555555557e-06, | |
| "loss": 0.0479, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 1.346, | |
| "grad_norm": 0.8003798723220825, | |
| "learning_rate": 2.06e-06, | |
| "loss": 0.0518, | |
| "step": 4075 | |
| }, | |
| { | |
| "epoch": 1.351, | |
| "grad_norm": 1.1387590169906616, | |
| "learning_rate": 2.0044444444444446e-06, | |
| "loss": 0.0453, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 1.3559999999999999, | |
| "grad_norm": 2.531583786010742, | |
| "learning_rate": 1.948888888888889e-06, | |
| "loss": 0.1058, | |
| "step": 4125 | |
| }, | |
| { | |
| "epoch": 1.361, | |
| "grad_norm": 2.965607166290283, | |
| "learning_rate": 1.8933333333333333e-06, | |
| "loss": 0.4739, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 1.366, | |
| "grad_norm": 3.0888776779174805, | |
| "learning_rate": 1.837777777777778e-06, | |
| "loss": 0.5685, | |
| "step": 4175 | |
| }, | |
| { | |
| "epoch": 1.371, | |
| "grad_norm": 1.201741099357605, | |
| "learning_rate": 1.7822222222222225e-06, | |
| "loss": 0.203, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 1.376, | |
| "grad_norm": 1.1447727680206299, | |
| "learning_rate": 1.7266666666666667e-06, | |
| "loss": 0.0638, | |
| "step": 4225 | |
| }, | |
| { | |
| "epoch": 1.381, | |
| "grad_norm": 0.9738747477531433, | |
| "learning_rate": 1.6711111111111112e-06, | |
| "loss": 0.0554, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 1.3860000000000001, | |
| "grad_norm": 0.9229967594146729, | |
| "learning_rate": 1.6155555555555559e-06, | |
| "loss": 0.0513, | |
| "step": 4275 | |
| }, | |
| { | |
| "epoch": 1.391, | |
| "grad_norm": 1.1114579439163208, | |
| "learning_rate": 1.56e-06, | |
| "loss": 0.0515, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 1.396, | |
| "grad_norm": 2.101980447769165, | |
| "learning_rate": 1.5044444444444446e-06, | |
| "loss": 0.078, | |
| "step": 4325 | |
| }, | |
| { | |
| "epoch": 1.401, | |
| "grad_norm": 1.728270173072815, | |
| "learning_rate": 1.4488888888888892e-06, | |
| "loss": 0.1397, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 1.4060000000000001, | |
| "grad_norm": 1.6103638410568237, | |
| "learning_rate": 1.3933333333333335e-06, | |
| "loss": 0.0851, | |
| "step": 4375 | |
| }, | |
| { | |
| "epoch": 1.411, | |
| "grad_norm": 1.334058403968811, | |
| "learning_rate": 1.337777777777778e-06, | |
| "loss": 0.0715, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 1.416, | |
| "grad_norm": 1.138947606086731, | |
| "learning_rate": 1.2822222222222222e-06, | |
| "loss": 0.0652, | |
| "step": 4425 | |
| }, | |
| { | |
| "epoch": 1.421, | |
| "grad_norm": 1.4695849418640137, | |
| "learning_rate": 1.2266666666666666e-06, | |
| "loss": 0.0723, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 1.426, | |
| "grad_norm": 4.378330707550049, | |
| "learning_rate": 1.171111111111111e-06, | |
| "loss": 0.1358, | |
| "step": 4475 | |
| }, | |
| { | |
| "epoch": 1.431, | |
| "grad_norm": 1.7877916097640991, | |
| "learning_rate": 1.1155555555555558e-06, | |
| "loss": 0.2452, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.436, | |
| "grad_norm": 1.5879491567611694, | |
| "learning_rate": 1.06e-06, | |
| "loss": 0.1187, | |
| "step": 4525 | |
| }, | |
| { | |
| "epoch": 1.441, | |
| "grad_norm": 1.056880235671997, | |
| "learning_rate": 1.0044444444444445e-06, | |
| "loss": 0.0584, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 1.446, | |
| "grad_norm": 0.8786937594413757, | |
| "learning_rate": 9.488888888888889e-07, | |
| "loss": 0.0487, | |
| "step": 4575 | |
| }, | |
| { | |
| "epoch": 1.451, | |
| "grad_norm": 0.9723278880119324, | |
| "learning_rate": 8.933333333333334e-07, | |
| "loss": 0.0413, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 1.456, | |
| "grad_norm": 1.1325929164886475, | |
| "learning_rate": 8.37777777777778e-07, | |
| "loss": 0.0444, | |
| "step": 4625 | |
| }, | |
| { | |
| "epoch": 1.461, | |
| "grad_norm": 1.0554845333099365, | |
| "learning_rate": 7.822222222222223e-07, | |
| "loss": 0.0408, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 1.466, | |
| "grad_norm": 1.6001170873641968, | |
| "learning_rate": 7.266666666666668e-07, | |
| "loss": 0.0486, | |
| "step": 4675 | |
| }, | |
| { | |
| "epoch": 2.002, | |
| "grad_norm": 1.434520959854126, | |
| "learning_rate": 6.711111111111111e-07, | |
| "loss": 0.0605, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 2.007, | |
| "grad_norm": 2.3628861904144287, | |
| "learning_rate": 6.155555555555556e-07, | |
| "loss": 0.0848, | |
| "step": 4725 | |
| }, | |
| { | |
| "epoch": 2.012, | |
| "grad_norm": 1.8358659744262695, | |
| "learning_rate": 5.6e-07, | |
| "loss": 0.1237, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 2.017, | |
| "grad_norm": 1.7591222524642944, | |
| "learning_rate": 5.044444444444445e-07, | |
| "loss": 0.0846, | |
| "step": 4775 | |
| }, | |
| { | |
| "epoch": 2.022, | |
| "grad_norm": 1.3934890031814575, | |
| "learning_rate": 4.488888888888889e-07, | |
| "loss": 0.0855, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 2.027, | |
| "grad_norm": 1.0459132194519043, | |
| "learning_rate": 3.9333333333333336e-07, | |
| "loss": 0.0702, | |
| "step": 4825 | |
| }, | |
| { | |
| "epoch": 2.032, | |
| "grad_norm": 1.2503104209899902, | |
| "learning_rate": 3.3777777777777777e-07, | |
| "loss": 0.0637, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 2.037, | |
| "grad_norm": 1.1123186349868774, | |
| "learning_rate": 2.822222222222222e-07, | |
| "loss": 0.0555, | |
| "step": 4875 | |
| }, | |
| { | |
| "epoch": 2.042, | |
| "grad_norm": 1.0896210670471191, | |
| "learning_rate": 2.266666666666667e-07, | |
| "loss": 0.0538, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 2.047, | |
| "grad_norm": 1.138146162033081, | |
| "learning_rate": 1.7111111111111114e-07, | |
| "loss": 0.0486, | |
| "step": 4925 | |
| }, | |
| { | |
| "epoch": 2.052, | |
| "grad_norm": 1.1482701301574707, | |
| "learning_rate": 1.1555555555555556e-07, | |
| "loss": 0.0668, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 2.057, | |
| "grad_norm": 0.9316118955612183, | |
| "learning_rate": 6.000000000000001e-08, | |
| "loss": 0.0581, | |
| "step": 4975 | |
| }, | |
| { | |
| "epoch": 2.062, | |
| "grad_norm": 1.259885549545288, | |
| "learning_rate": 4.444444444444445e-09, | |
| "loss": 0.0516, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.062, | |
| "eval_loss": 0.20070870220661163, | |
| "eval_runtime": 383.0104, | |
| "eval_samples_per_second": 24.717, | |
| "eval_steps_per_second": 3.091, | |
| "eval_wer": 10.670098148917804, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.062, | |
| "step": 5000, | |
| "total_flos": 9.231442936676352e+19, | |
| "train_loss": 0.16698810555934906, | |
| "train_runtime": 16450.2071, | |
| "train_samples_per_second": 19.453, | |
| "train_steps_per_second": 0.304 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 5000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9223372036854775807, | |
| "save_steps": 1000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 9.231442936676352e+19, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |