{ "best_metric": 8.569471791798646, "best_model_checkpoint": "./checkpoint-20000", "epoch": 9.08065, "global_step": 20000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0, "learning_rate": 4.4e-07, "loss": 1.1845, "step": 25 }, { "epoch": 0.0, "learning_rate": 9.400000000000001e-07, "loss": 1.048, "step": 50 }, { "epoch": 0.0, "learning_rate": 1.44e-06, "loss": 0.7852, "step": 75 }, { "epoch": 0.01, "learning_rate": 1.94e-06, "loss": 0.5364, "step": 100 }, { "epoch": 0.01, "learning_rate": 2.4400000000000004e-06, "loss": 0.4857, "step": 125 }, { "epoch": 0.01, "learning_rate": 2.9400000000000002e-06, "loss": 0.4276, "step": 150 }, { "epoch": 0.01, "learning_rate": 3.44e-06, "loss": 0.3452, "step": 175 }, { "epoch": 0.01, "learning_rate": 3.94e-06, "loss": 0.3893, "step": 200 }, { "epoch": 0.01, "learning_rate": 4.440000000000001e-06, "loss": 0.363, "step": 225 }, { "epoch": 0.01, "learning_rate": 4.94e-06, "loss": 0.3388, "step": 250 }, { "epoch": 0.01, "learning_rate": 5.4400000000000004e-06, "loss": 0.3354, "step": 275 }, { "epoch": 0.01, "learning_rate": 5.94e-06, "loss": 0.3401, "step": 300 }, { "epoch": 0.02, "learning_rate": 6.440000000000001e-06, "loss": 0.3235, "step": 325 }, { "epoch": 0.02, "learning_rate": 6.9400000000000005e-06, "loss": 0.3287, "step": 350 }, { "epoch": 0.02, "learning_rate": 7.440000000000001e-06, "loss": 0.3259, "step": 375 }, { "epoch": 0.02, "learning_rate": 7.94e-06, "loss": 0.3069, "step": 400 }, { "epoch": 0.02, "learning_rate": 8.44e-06, "loss": 0.3143, "step": 425 }, { "epoch": 0.02, "learning_rate": 8.94e-06, "loss": 0.2929, "step": 450 }, { "epoch": 0.02, "learning_rate": 9.440000000000001e-06, "loss": 0.3154, "step": 475 }, { "epoch": 0.03, "learning_rate": 9.940000000000001e-06, "loss": 0.2955, "step": 500 }, { "epoch": 0.03, "learning_rate": 9.98871794871795e-06, "loss": 0.3154, "step": 525 }, { "epoch": 0.03, "learning_rate": 9.975897435897436e-06, "loss": 0.3112, "step": 550 }, { "epoch": 0.03, "learning_rate": 9.963076923076925e-06, "loss": 0.2883, "step": 575 }, { "epoch": 0.03, "learning_rate": 9.950256410256412e-06, "loss": 0.2815, "step": 600 }, { "epoch": 0.03, "learning_rate": 9.937435897435898e-06, "loss": 0.2741, "step": 625 }, { "epoch": 0.03, "learning_rate": 9.924615384615385e-06, "loss": 0.2826, "step": 650 }, { "epoch": 0.03, "learning_rate": 9.911794871794874e-06, "loss": 0.2778, "step": 675 }, { "epoch": 0.04, "learning_rate": 9.89897435897436e-06, "loss": 0.2764, "step": 700 }, { "epoch": 0.04, "learning_rate": 9.886153846153846e-06, "loss": 0.2506, "step": 725 }, { "epoch": 0.04, "learning_rate": 9.873333333333334e-06, "loss": 0.2869, "step": 750 }, { "epoch": 0.04, "learning_rate": 9.860512820512821e-06, "loss": 0.2752, "step": 775 }, { "epoch": 0.04, "learning_rate": 9.847692307692308e-06, "loss": 0.2568, "step": 800 }, { "epoch": 0.04, "learning_rate": 9.834871794871795e-06, "loss": 0.2578, "step": 825 }, { "epoch": 0.04, "learning_rate": 9.822051282051283e-06, "loss": 0.2508, "step": 850 }, { "epoch": 0.04, "learning_rate": 9.80923076923077e-06, "loss": 0.2672, "step": 875 }, { "epoch": 0.04, "learning_rate": 9.796410256410257e-06, "loss": 0.241, "step": 900 }, { "epoch": 0.05, "learning_rate": 9.783589743589744e-06, "loss": 0.259, "step": 925 }, { "epoch": 0.05, "learning_rate": 9.770769230769232e-06, "loss": 0.2194, "step": 950 }, { "epoch": 0.05, "learning_rate": 9.757948717948719e-06, "loss": 0.2305, "step": 975 }, { "epoch": 0.05, "learning_rate": 9.745128205128206e-06, "loss": 0.2307, "step": 1000 }, { "epoch": 0.05, "learning_rate": 9.732307692307693e-06, "loss": 0.2145, "step": 1025 }, { "epoch": 0.05, "learning_rate": 9.71948717948718e-06, "loss": 0.1868, "step": 1050 }, { "epoch": 0.05, "learning_rate": 9.706666666666668e-06, "loss": 0.2109, "step": 1075 }, { "epoch": 0.06, "learning_rate": 9.693846153846155e-06, "loss": 0.2155, "step": 1100 }, { "epoch": 0.06, "learning_rate": 9.681025641025642e-06, "loss": 0.2174, "step": 1125 }, { "epoch": 0.06, "learning_rate": 9.668205128205129e-06, "loss": 0.2195, "step": 1150 }, { "epoch": 0.06, "learning_rate": 9.655384615384617e-06, "loss": 0.2136, "step": 1175 }, { "epoch": 0.06, "learning_rate": 9.642564102564104e-06, "loss": 0.2193, "step": 1200 }, { "epoch": 0.06, "learning_rate": 9.62974358974359e-06, "loss": 0.2027, "step": 1225 }, { "epoch": 0.06, "learning_rate": 9.616923076923077e-06, "loss": 0.2044, "step": 1250 }, { "epoch": 0.06, "learning_rate": 9.604102564102566e-06, "loss": 0.2021, "step": 1275 }, { "epoch": 0.07, "learning_rate": 9.591282051282053e-06, "loss": 0.2357, "step": 1300 }, { "epoch": 0.07, "learning_rate": 9.578461538461538e-06, "loss": 0.2255, "step": 1325 }, { "epoch": 0.07, "learning_rate": 9.565641025641026e-06, "loss": 0.2424, "step": 1350 }, { "epoch": 0.07, "learning_rate": 9.552820512820513e-06, "loss": 0.2354, "step": 1375 }, { "epoch": 0.07, "learning_rate": 9.54e-06, "loss": 0.2476, "step": 1400 }, { "epoch": 0.07, "learning_rate": 9.527179487179487e-06, "loss": 0.2391, "step": 1425 }, { "epoch": 0.07, "learning_rate": 9.514358974358975e-06, "loss": 0.2315, "step": 1450 }, { "epoch": 0.07, "learning_rate": 9.501538461538462e-06, "loss": 0.2347, "step": 1475 }, { "epoch": 0.07, "learning_rate": 9.488717948717949e-06, "loss": 0.2305, "step": 1500 }, { "epoch": 0.08, "learning_rate": 9.475897435897436e-06, "loss": 0.2338, "step": 1525 }, { "epoch": 0.08, "learning_rate": 9.463076923076924e-06, "loss": 0.2348, "step": 1550 }, { "epoch": 0.08, "learning_rate": 9.450256410256411e-06, "loss": 0.2087, "step": 1575 }, { "epoch": 0.08, "learning_rate": 9.437435897435898e-06, "loss": 0.2032, "step": 1600 }, { "epoch": 0.08, "learning_rate": 9.424615384615385e-06, "loss": 0.1865, "step": 1625 }, { "epoch": 0.08, "learning_rate": 9.411794871794872e-06, "loss": 0.1824, "step": 1650 }, { "epoch": 0.08, "learning_rate": 9.39897435897436e-06, "loss": 0.1776, "step": 1675 }, { "epoch": 0.09, "learning_rate": 9.386153846153847e-06, "loss": 0.1775, "step": 1700 }, { "epoch": 0.09, "learning_rate": 9.373333333333334e-06, "loss": 0.1828, "step": 1725 }, { "epoch": 0.09, "learning_rate": 9.36051282051282e-06, "loss": 0.1975, "step": 1750 }, { "epoch": 0.09, "learning_rate": 9.34769230769231e-06, "loss": 0.2122, "step": 1775 }, { "epoch": 0.09, "learning_rate": 9.334871794871796e-06, "loss": 0.2005, "step": 1800 }, { "epoch": 0.09, "learning_rate": 9.322051282051283e-06, "loss": 0.2373, "step": 1825 }, { "epoch": 0.09, "learning_rate": 9.30923076923077e-06, "loss": 0.2087, "step": 1850 }, { "epoch": 0.09, "learning_rate": 9.296410256410258e-06, "loss": 0.2028, "step": 1875 }, { "epoch": 0.1, "learning_rate": 9.283589743589745e-06, "loss": 0.1962, "step": 1900 }, { "epoch": 0.1, "learning_rate": 9.270769230769232e-06, "loss": 0.2066, "step": 1925 }, { "epoch": 0.1, "learning_rate": 9.257948717948719e-06, "loss": 0.2132, "step": 1950 }, { "epoch": 0.1, "learning_rate": 9.245128205128206e-06, "loss": 0.2177, "step": 1975 }, { "epoch": 0.1, "learning_rate": 9.232307692307692e-06, "loss": 0.2128, "step": 2000 }, { "epoch": 0.1, "eval_loss": 0.2644473910331726, "eval_runtime": 6687.6038, "eval_samples_per_second": 2.443, "eval_steps_per_second": 0.305, "eval_wer": 13.030277088148845, "step": 2000 }, { "epoch": 0.1, "learning_rate": 9.21948717948718e-06, "loss": 0.2023, "step": 2025 }, { "epoch": 1.0, "learning_rate": 9.206666666666668e-06, "loss": 0.2212, "step": 2050 }, { "epoch": 1.0, "learning_rate": 9.193846153846155e-06, "loss": 0.2242, "step": 2075 }, { "epoch": 1.0, "learning_rate": 9.181025641025641e-06, "loss": 0.2082, "step": 2100 }, { "epoch": 1.0, "learning_rate": 9.168205128205128e-06, "loss": 0.2491, "step": 2125 }, { "epoch": 1.01, "learning_rate": 9.155384615384617e-06, "loss": 0.2217, "step": 2150 }, { "epoch": 1.01, "learning_rate": 9.142564102564104e-06, "loss": 0.2181, "step": 2175 }, { "epoch": 1.01, "learning_rate": 9.12974358974359e-06, "loss": 0.2006, "step": 2200 }, { "epoch": 1.01, "learning_rate": 9.116923076923077e-06, "loss": 0.241, "step": 2225 }, { "epoch": 1.01, "learning_rate": 9.104102564102566e-06, "loss": 0.2108, "step": 2250 }, { "epoch": 1.01, "learning_rate": 9.091282051282053e-06, "loss": 0.1864, "step": 2275 }, { "epoch": 1.01, "learning_rate": 9.07846153846154e-06, "loss": 0.1984, "step": 2300 }, { "epoch": 1.01, "learning_rate": 9.065641025641026e-06, "loss": 0.1942, "step": 2325 }, { "epoch": 1.02, "learning_rate": 9.052820512820513e-06, "loss": 0.1931, "step": 2350 }, { "epoch": 1.02, "learning_rate": 9.040000000000002e-06, "loss": 0.1776, "step": 2375 }, { "epoch": 1.02, "learning_rate": 9.027179487179488e-06, "loss": 0.1912, "step": 2400 }, { "epoch": 1.02, "learning_rate": 9.014358974358975e-06, "loss": 0.1893, "step": 2425 }, { "epoch": 1.02, "learning_rate": 9.001538461538462e-06, "loss": 0.1657, "step": 2450 }, { "epoch": 1.02, "learning_rate": 8.98871794871795e-06, "loss": 0.1962, "step": 2475 }, { "epoch": 1.02, "learning_rate": 8.975897435897437e-06, "loss": 0.1816, "step": 2500 }, { "epoch": 1.02, "learning_rate": 8.963076923076924e-06, "loss": 0.1855, "step": 2525 }, { "epoch": 1.03, "learning_rate": 8.950256410256411e-06, "loss": 0.1754, "step": 2550 }, { "epoch": 1.03, "learning_rate": 8.937435897435898e-06, "loss": 0.1531, "step": 2575 }, { "epoch": 1.03, "learning_rate": 8.924615384615385e-06, "loss": 0.1735, "step": 2600 }, { "epoch": 1.03, "learning_rate": 8.911794871794871e-06, "loss": 0.1671, "step": 2625 }, { "epoch": 1.03, "learning_rate": 8.89897435897436e-06, "loss": 0.1787, "step": 2650 }, { "epoch": 1.03, "learning_rate": 8.886153846153847e-06, "loss": 0.1743, "step": 2675 }, { "epoch": 1.03, "learning_rate": 8.873333333333334e-06, "loss": 0.1828, "step": 2700 }, { "epoch": 1.03, "learning_rate": 8.86051282051282e-06, "loss": 0.1781, "step": 2725 }, { "epoch": 1.04, "learning_rate": 8.847692307692309e-06, "loss": 0.1594, "step": 2750 }, { "epoch": 1.04, "learning_rate": 8.834871794871796e-06, "loss": 0.1809, "step": 2775 }, { "epoch": 1.04, "learning_rate": 8.822051282051283e-06, "loss": 0.1761, "step": 2800 }, { "epoch": 1.04, "learning_rate": 8.80923076923077e-06, "loss": 0.188, "step": 2825 }, { "epoch": 1.04, "learning_rate": 8.796410256410258e-06, "loss": 0.1493, "step": 2850 }, { "epoch": 1.04, "learning_rate": 8.783589743589745e-06, "loss": 0.1794, "step": 2875 }, { "epoch": 1.04, "learning_rate": 8.770769230769232e-06, "loss": 0.1623, "step": 2900 }, { "epoch": 1.04, "learning_rate": 8.757948717948718e-06, "loss": 0.1759, "step": 2925 }, { "epoch": 1.05, "learning_rate": 8.745128205128205e-06, "loss": 0.1645, "step": 2950 }, { "epoch": 1.05, "learning_rate": 8.732307692307694e-06, "loss": 0.1725, "step": 2975 }, { "epoch": 1.05, "learning_rate": 8.71948717948718e-06, "loss": 0.1598, "step": 3000 }, { "epoch": 1.05, "learning_rate": 8.706666666666667e-06, "loss": 0.1579, "step": 3025 }, { "epoch": 1.05, "learning_rate": 8.693846153846154e-06, "loss": 0.1555, "step": 3050 }, { "epoch": 1.05, "learning_rate": 8.681025641025643e-06, "loss": 0.1481, "step": 3075 }, { "epoch": 1.05, "learning_rate": 8.66820512820513e-06, "loss": 0.1454, "step": 3100 }, { "epoch": 1.05, "learning_rate": 8.655384615384616e-06, "loss": 0.1647, "step": 3125 }, { "epoch": 1.06, "learning_rate": 8.642564102564103e-06, "loss": 0.1676, "step": 3150 }, { "epoch": 1.06, "learning_rate": 8.62974358974359e-06, "loss": 0.1679, "step": 3175 }, { "epoch": 1.06, "learning_rate": 8.616923076923077e-06, "loss": 0.1535, "step": 3200 }, { "epoch": 1.06, "learning_rate": 8.604102564102564e-06, "loss": 0.1428, "step": 3225 }, { "epoch": 1.06, "learning_rate": 8.591282051282052e-06, "loss": 0.1825, "step": 3250 }, { "epoch": 1.06, "learning_rate": 8.578461538461539e-06, "loss": 0.1708, "step": 3275 }, { "epoch": 1.06, "learning_rate": 8.565641025641026e-06, "loss": 0.1669, "step": 3300 }, { "epoch": 1.06, "learning_rate": 8.552820512820513e-06, "loss": 0.1803, "step": 3325 }, { "epoch": 1.07, "learning_rate": 8.540000000000001e-06, "loss": 0.1672, "step": 3350 }, { "epoch": 1.07, "learning_rate": 8.527179487179488e-06, "loss": 0.1638, "step": 3375 }, { "epoch": 1.07, "learning_rate": 8.514358974358975e-06, "loss": 0.1807, "step": 3400 }, { "epoch": 1.07, "learning_rate": 8.501538461538462e-06, "loss": 0.1651, "step": 3425 }, { "epoch": 1.07, "learning_rate": 8.48871794871795e-06, "loss": 0.1743, "step": 3450 }, { "epoch": 1.07, "learning_rate": 8.475897435897437e-06, "loss": 0.1704, "step": 3475 }, { "epoch": 1.07, "learning_rate": 8.463076923076924e-06, "loss": 0.1712, "step": 3500 }, { "epoch": 1.07, "learning_rate": 8.45025641025641e-06, "loss": 0.1604, "step": 3525 }, { "epoch": 1.08, "learning_rate": 8.437435897435898e-06, "loss": 0.1749, "step": 3550 }, { "epoch": 1.08, "learning_rate": 8.424615384615386e-06, "loss": 0.1577, "step": 3575 }, { "epoch": 1.08, "learning_rate": 8.411794871794873e-06, "loss": 0.1579, "step": 3600 }, { "epoch": 1.08, "learning_rate": 8.39897435897436e-06, "loss": 0.1467, "step": 3625 }, { "epoch": 1.08, "learning_rate": 8.386153846153847e-06, "loss": 0.1365, "step": 3650 }, { "epoch": 1.08, "learning_rate": 8.373333333333335e-06, "loss": 0.1776, "step": 3675 }, { "epoch": 1.08, "learning_rate": 8.360512820512822e-06, "loss": 0.1419, "step": 3700 }, { "epoch": 1.08, "learning_rate": 8.347692307692309e-06, "loss": 0.1903, "step": 3725 }, { "epoch": 1.09, "learning_rate": 8.334871794871796e-06, "loss": 0.1639, "step": 3750 }, { "epoch": 1.09, "learning_rate": 8.322051282051282e-06, "loss": 0.1695, "step": 3775 }, { "epoch": 1.09, "learning_rate": 8.30923076923077e-06, "loss": 0.1451, "step": 3800 }, { "epoch": 1.09, "learning_rate": 8.296410256410256e-06, "loss": 0.1569, "step": 3825 }, { "epoch": 1.09, "learning_rate": 8.283589743589745e-06, "loss": 0.1557, "step": 3850 }, { "epoch": 1.09, "learning_rate": 8.270769230769231e-06, "loss": 0.1677, "step": 3875 }, { "epoch": 1.09, "learning_rate": 8.257948717948718e-06, "loss": 0.1401, "step": 3900 }, { "epoch": 1.09, "learning_rate": 8.245128205128205e-06, "loss": 0.1566, "step": 3925 }, { "epoch": 1.1, "learning_rate": 8.232307692307694e-06, "loss": 0.1746, "step": 3950 }, { "epoch": 1.1, "learning_rate": 8.21948717948718e-06, "loss": 0.1521, "step": 3975 }, { "epoch": 1.1, "learning_rate": 8.206666666666667e-06, "loss": 0.1361, "step": 4000 }, { "epoch": 1.1, "eval_loss": 0.2300407737493515, "eval_runtime": 6609.9017, "eval_samples_per_second": 2.472, "eval_steps_per_second": 0.309, "eval_wer": 10.956824472790572, "step": 4000 }, { "epoch": 1.1, "learning_rate": 8.193846153846154e-06, "loss": 0.1406, "step": 4025 }, { "epoch": 1.1, "learning_rate": 8.181025641025642e-06, "loss": 0.1713, "step": 4050 }, { "epoch": 1.1, "learning_rate": 8.16820512820513e-06, "loss": 0.1638, "step": 4075 }, { "epoch": 2.0, "learning_rate": 8.155384615384616e-06, "loss": 0.1661, "step": 4100 }, { "epoch": 2.0, "learning_rate": 8.142564102564103e-06, "loss": 0.1623, "step": 4125 }, { "epoch": 2.0, "learning_rate": 8.12974358974359e-06, "loss": 0.159, "step": 4150 }, { "epoch": 2.0, "learning_rate": 8.116923076923078e-06, "loss": 0.1572, "step": 4175 }, { "epoch": 2.01, "learning_rate": 8.104102564102565e-06, "loss": 0.1526, "step": 4200 }, { "epoch": 2.01, "learning_rate": 8.091282051282052e-06, "loss": 0.1395, "step": 4225 }, { "epoch": 2.01, "learning_rate": 8.078461538461539e-06, "loss": 0.1363, "step": 4250 }, { "epoch": 2.01, "learning_rate": 8.065641025641027e-06, "loss": 0.1353, "step": 4275 }, { "epoch": 2.01, "learning_rate": 8.052820512820514e-06, "loss": 0.13, "step": 4300 }, { "epoch": 2.01, "learning_rate": 8.040000000000001e-06, "loss": 0.1246, "step": 4325 }, { "epoch": 2.01, "learning_rate": 8.027179487179488e-06, "loss": 0.1213, "step": 4350 }, { "epoch": 2.01, "learning_rate": 8.014358974358975e-06, "loss": 0.1263, "step": 4375 }, { "epoch": 2.02, "learning_rate": 8.001538461538461e-06, "loss": 0.1213, "step": 4400 }, { "epoch": 2.02, "learning_rate": 7.988717948717948e-06, "loss": 0.1103, "step": 4425 }, { "epoch": 2.02, "learning_rate": 7.975897435897437e-06, "loss": 0.1148, "step": 4450 }, { "epoch": 2.02, "learning_rate": 7.963076923076924e-06, "loss": 0.1015, "step": 4475 }, { "epoch": 2.02, "learning_rate": 7.95025641025641e-06, "loss": 0.1053, "step": 4500 }, { "epoch": 2.02, "learning_rate": 7.937435897435897e-06, "loss": 0.1087, "step": 4525 }, { "epoch": 2.02, "learning_rate": 7.924615384615386e-06, "loss": 0.1022, "step": 4550 }, { "epoch": 2.02, "learning_rate": 7.911794871794873e-06, "loss": 0.094, "step": 4575 }, { "epoch": 2.03, "learning_rate": 7.89897435897436e-06, "loss": 0.0887, "step": 4600 }, { "epoch": 2.03, "learning_rate": 7.886153846153846e-06, "loss": 0.099, "step": 4625 }, { "epoch": 2.03, "learning_rate": 7.873333333333335e-06, "loss": 0.0982, "step": 4650 }, { "epoch": 2.03, "learning_rate": 7.860512820512822e-06, "loss": 0.0981, "step": 4675 }, { "epoch": 2.03, "learning_rate": 7.847692307692308e-06, "loss": 0.0828, "step": 4700 }, { "epoch": 2.03, "learning_rate": 7.834871794871795e-06, "loss": 0.0834, "step": 4725 }, { "epoch": 2.03, "learning_rate": 7.822051282051282e-06, "loss": 0.0894, "step": 4750 }, { "epoch": 2.03, "learning_rate": 7.80923076923077e-06, "loss": 0.0792, "step": 4775 }, { "epoch": 2.04, "learning_rate": 7.796410256410257e-06, "loss": 0.0969, "step": 4800 }, { "epoch": 2.04, "learning_rate": 7.783589743589744e-06, "loss": 0.0841, "step": 4825 }, { "epoch": 2.04, "learning_rate": 7.770769230769231e-06, "loss": 0.0759, "step": 4850 }, { "epoch": 2.04, "learning_rate": 7.75794871794872e-06, "loss": 0.0836, "step": 4875 }, { "epoch": 2.04, "learning_rate": 7.745128205128206e-06, "loss": 0.0888, "step": 4900 }, { "epoch": 2.04, "learning_rate": 7.732307692307693e-06, "loss": 0.0859, "step": 4925 }, { "epoch": 2.04, "learning_rate": 7.71948717948718e-06, "loss": 0.08, "step": 4950 }, { "epoch": 2.04, "learning_rate": 7.706666666666669e-06, "loss": 0.0585, "step": 4975 }, { "epoch": 2.05, "learning_rate": 7.693846153846154e-06, "loss": 0.0627, "step": 5000 }, { "epoch": 2.05, "learning_rate": 7.68102564102564e-06, "loss": 0.0575, "step": 5025 }, { "epoch": 2.05, "learning_rate": 7.668205128205129e-06, "loss": 0.0511, "step": 5050 }, { "epoch": 2.05, "learning_rate": 7.655384615384616e-06, "loss": 0.0514, "step": 5075 }, { "epoch": 2.05, "learning_rate": 7.642564102564103e-06, "loss": 0.0476, "step": 5100 }, { "epoch": 2.05, "learning_rate": 7.62974358974359e-06, "loss": 0.0477, "step": 5125 }, { "epoch": 2.05, "learning_rate": 7.616923076923077e-06, "loss": 0.0548, "step": 5150 }, { "epoch": 2.05, "learning_rate": 7.604102564102565e-06, "loss": 0.0438, "step": 5175 }, { "epoch": 2.06, "learning_rate": 7.591282051282052e-06, "loss": 0.067, "step": 5200 }, { "epoch": 2.06, "learning_rate": 7.578461538461539e-06, "loss": 0.0502, "step": 5225 }, { "epoch": 2.06, "learning_rate": 7.565641025641026e-06, "loss": 0.051, "step": 5250 }, { "epoch": 2.06, "learning_rate": 7.552820512820514e-06, "loss": 0.0659, "step": 5275 }, { "epoch": 2.06, "learning_rate": 7.540000000000001e-06, "loss": 0.0516, "step": 5300 }, { "epoch": 2.06, "learning_rate": 7.5271794871794875e-06, "loss": 0.0613, "step": 5325 }, { "epoch": 2.06, "learning_rate": 7.514358974358975e-06, "loss": 0.0499, "step": 5350 }, { "epoch": 2.06, "learning_rate": 7.501538461538462e-06, "loss": 0.0647, "step": 5375 }, { "epoch": 2.07, "learning_rate": 7.48871794871795e-06, "loss": 0.0664, "step": 5400 }, { "epoch": 2.07, "learning_rate": 7.4758974358974365e-06, "loss": 0.0773, "step": 5425 }, { "epoch": 2.07, "learning_rate": 7.463076923076924e-06, "loss": 0.0835, "step": 5450 }, { "epoch": 2.07, "learning_rate": 7.450256410256411e-06, "loss": 0.0758, "step": 5475 }, { "epoch": 2.07, "learning_rate": 7.437435897435899e-06, "loss": 0.0738, "step": 5500 }, { "epoch": 2.07, "learning_rate": 7.4246153846153855e-06, "loss": 0.0725, "step": 5525 }, { "epoch": 2.07, "learning_rate": 7.411794871794873e-06, "loss": 0.0645, "step": 5550 }, { "epoch": 2.07, "learning_rate": 7.39897435897436e-06, "loss": 0.0754, "step": 5575 }, { "epoch": 2.08, "learning_rate": 7.386153846153846e-06, "loss": 0.0657, "step": 5600 }, { "epoch": 2.08, "learning_rate": 7.373333333333334e-06, "loss": 0.067, "step": 5625 }, { "epoch": 2.08, "learning_rate": 7.3605128205128204e-06, "loss": 0.0658, "step": 5650 }, { "epoch": 2.08, "learning_rate": 7.347692307692308e-06, "loss": 0.0521, "step": 5675 }, { "epoch": 2.08, "learning_rate": 7.334871794871795e-06, "loss": 0.046, "step": 5700 }, { "epoch": 2.08, "learning_rate": 7.322051282051283e-06, "loss": 0.0488, "step": 5725 }, { "epoch": 2.08, "learning_rate": 7.309230769230769e-06, "loss": 0.0529, "step": 5750 }, { "epoch": 2.08, "learning_rate": 7.296410256410257e-06, "loss": 0.0415, "step": 5775 }, { "epoch": 2.09, "learning_rate": 7.283589743589744e-06, "loss": 0.0328, "step": 5800 }, { "epoch": 2.09, "learning_rate": 7.270769230769232e-06, "loss": 0.0468, "step": 5825 }, { "epoch": 2.09, "learning_rate": 7.257948717948718e-06, "loss": 0.065, "step": 5850 }, { "epoch": 2.09, "learning_rate": 7.245128205128206e-06, "loss": 0.0658, "step": 5875 }, { "epoch": 2.09, "learning_rate": 7.232307692307693e-06, "loss": 0.0664, "step": 5900 }, { "epoch": 2.09, "learning_rate": 7.21948717948718e-06, "loss": 0.0613, "step": 5925 }, { "epoch": 2.09, "learning_rate": 7.206666666666667e-06, "loss": 0.0658, "step": 5950 }, { "epoch": 2.09, "learning_rate": 7.193846153846154e-06, "loss": 0.0569, "step": 5975 }, { "epoch": 2.1, "learning_rate": 7.181025641025642e-06, "loss": 0.0658, "step": 6000 }, { "epoch": 2.1, "eval_loss": 0.23763766884803772, "eval_runtime": 6733.6099, "eval_samples_per_second": 2.427, "eval_steps_per_second": 0.303, "eval_wer": 11.280997631481322, "step": 6000 }, { "epoch": 2.1, "learning_rate": 7.168205128205129e-06, "loss": 0.0664, "step": 6025 }, { "epoch": 2.1, "learning_rate": 7.155384615384616e-06, "loss": 0.0623, "step": 6050 }, { "epoch": 2.1, "learning_rate": 7.142564102564103e-06, "loss": 0.0654, "step": 6075 }, { "epoch": 2.1, "learning_rate": 7.129743589743591e-06, "loss": 0.0597, "step": 6100 }, { "epoch": 2.1, "learning_rate": 7.116923076923078e-06, "loss": 0.0655, "step": 6125 }, { "epoch": 3.0, "learning_rate": 7.104615384615385e-06, "loss": 0.1223, "step": 6150 }, { "epoch": 3.0, "learning_rate": 7.091794871794873e-06, "loss": 0.1452, "step": 6175 }, { "epoch": 3.0, "learning_rate": 7.07897435897436e-06, "loss": 0.1371, "step": 6200 }, { "epoch": 3.0, "learning_rate": 7.066153846153847e-06, "loss": 0.1189, "step": 6225 }, { "epoch": 3.01, "learning_rate": 7.053333333333334e-06, "loss": 0.1221, "step": 6250 }, { "epoch": 3.01, "learning_rate": 7.040512820512822e-06, "loss": 0.114, "step": 6275 }, { "epoch": 3.01, "learning_rate": 7.027692307692309e-06, "loss": 0.1174, "step": 6300 }, { "epoch": 3.01, "learning_rate": 7.014871794871796e-06, "loss": 0.1114, "step": 6325 }, { "epoch": 3.01, "learning_rate": 7.002051282051283e-06, "loss": 0.1253, "step": 6350 }, { "epoch": 3.01, "learning_rate": 6.989230769230769e-06, "loss": 0.1149, "step": 6375 }, { "epoch": 3.01, "learning_rate": 6.976410256410257e-06, "loss": 0.0962, "step": 6400 }, { "epoch": 3.01, "learning_rate": 6.963589743589744e-06, "loss": 0.1137, "step": 6425 }, { "epoch": 3.02, "learning_rate": 6.950769230769231e-06, "loss": 0.0988, "step": 6450 }, { "epoch": 3.02, "learning_rate": 6.937948717948718e-06, "loss": 0.1089, "step": 6475 }, { "epoch": 3.02, "learning_rate": 6.925128205128206e-06, "loss": 0.1228, "step": 6500 }, { "epoch": 3.02, "learning_rate": 6.912307692307693e-06, "loss": 0.119, "step": 6525 }, { "epoch": 3.02, "learning_rate": 6.899487179487179e-06, "loss": 0.1229, "step": 6550 }, { "epoch": 3.02, "learning_rate": 6.886666666666667e-06, "loss": 0.1138, "step": 6575 }, { "epoch": 3.02, "learning_rate": 6.873846153846154e-06, "loss": 0.1004, "step": 6600 }, { "epoch": 3.02, "learning_rate": 6.861025641025642e-06, "loss": 0.1238, "step": 6625 }, { "epoch": 3.03, "learning_rate": 6.848205128205128e-06, "loss": 0.1217, "step": 6650 }, { "epoch": 3.03, "learning_rate": 6.835384615384616e-06, "loss": 0.1008, "step": 6675 }, { "epoch": 3.03, "learning_rate": 6.822564102564103e-06, "loss": 0.1174, "step": 6700 }, { "epoch": 3.03, "learning_rate": 6.8097435897435906e-06, "loss": 0.1112, "step": 6725 }, { "epoch": 3.03, "learning_rate": 6.796923076923077e-06, "loss": 0.1061, "step": 6750 }, { "epoch": 3.03, "learning_rate": 6.784102564102565e-06, "loss": 0.1097, "step": 6775 }, { "epoch": 3.03, "learning_rate": 6.771282051282052e-06, "loss": 0.108, "step": 6800 }, { "epoch": 3.03, "learning_rate": 6.7584615384615396e-06, "loss": 0.0881, "step": 6825 }, { "epoch": 3.04, "learning_rate": 6.745641025641026e-06, "loss": 0.0979, "step": 6850 }, { "epoch": 3.04, "learning_rate": 6.732820512820514e-06, "loss": 0.1094, "step": 6875 }, { "epoch": 3.04, "learning_rate": 6.720000000000001e-06, "loss": 0.0872, "step": 6900 }, { "epoch": 3.04, "learning_rate": 6.7071794871794886e-06, "loss": 0.1266, "step": 6925 }, { "epoch": 3.04, "learning_rate": 6.694358974358975e-06, "loss": 0.1186, "step": 6950 }, { "epoch": 3.04, "learning_rate": 6.681538461538463e-06, "loss": 0.1044, "step": 6975 }, { "epoch": 3.04, "learning_rate": 6.668717948717949e-06, "loss": 0.1295, "step": 7000 }, { "epoch": 3.04, "learning_rate": 6.655897435897436e-06, "loss": 0.1242, "step": 7025 }, { "epoch": 3.05, "learning_rate": 6.6430769230769235e-06, "loss": 0.1105, "step": 7050 }, { "epoch": 3.05, "learning_rate": 6.63025641025641e-06, "loss": 0.1171, "step": 7075 }, { "epoch": 3.05, "learning_rate": 6.617435897435898e-06, "loss": 0.1247, "step": 7100 }, { "epoch": 3.05, "learning_rate": 6.604615384615385e-06, "loss": 0.1365, "step": 7125 }, { "epoch": 3.05, "learning_rate": 6.5917948717948725e-06, "loss": 0.1085, "step": 7150 }, { "epoch": 3.05, "learning_rate": 6.578974358974359e-06, "loss": 0.1359, "step": 7175 }, { "epoch": 3.05, "learning_rate": 6.566153846153846e-06, "loss": 0.1297, "step": 7200 }, { "epoch": 3.05, "learning_rate": 6.553333333333334e-06, "loss": 0.138, "step": 7225 }, { "epoch": 3.06, "learning_rate": 6.540512820512821e-06, "loss": 0.1113, "step": 7250 }, { "epoch": 3.06, "learning_rate": 6.527692307692308e-06, "loss": 0.1223, "step": 7275 }, { "epoch": 3.06, "learning_rate": 6.514871794871795e-06, "loss": 0.1113, "step": 7300 }, { "epoch": 3.06, "learning_rate": 6.502051282051283e-06, "loss": 0.1134, "step": 7325 }, { "epoch": 3.06, "learning_rate": 6.48923076923077e-06, "loss": 0.106, "step": 7350 }, { "epoch": 3.06, "learning_rate": 6.476410256410257e-06, "loss": 0.1245, "step": 7375 }, { "epoch": 3.06, "learning_rate": 6.463589743589744e-06, "loss": 0.1201, "step": 7400 }, { "epoch": 3.06, "learning_rate": 6.450769230769232e-06, "loss": 0.1048, "step": 7425 }, { "epoch": 3.07, "learning_rate": 6.437948717948719e-06, "loss": 0.1033, "step": 7450 }, { "epoch": 3.07, "learning_rate": 6.425128205128206e-06, "loss": 0.1045, "step": 7475 }, { "epoch": 3.07, "learning_rate": 6.412307692307693e-06, "loss": 0.1013, "step": 7500 }, { "epoch": 3.07, "learning_rate": 6.399487179487181e-06, "loss": 0.1129, "step": 7525 }, { "epoch": 3.07, "learning_rate": 6.386666666666668e-06, "loss": 0.1033, "step": 7550 }, { "epoch": 3.07, "learning_rate": 6.373846153846155e-06, "loss": 0.1045, "step": 7575 }, { "epoch": 3.07, "learning_rate": 6.361025641025641e-06, "loss": 0.0973, "step": 7600 }, { "epoch": 3.07, "learning_rate": 6.348205128205128e-06, "loss": 0.1397, "step": 7625 }, { "epoch": 3.08, "learning_rate": 6.335384615384616e-06, "loss": 0.1245, "step": 7650 }, { "epoch": 3.08, "learning_rate": 6.322564102564103e-06, "loss": 0.1176, "step": 7675 }, { "epoch": 3.08, "learning_rate": 6.30974358974359e-06, "loss": 0.1146, "step": 7700 }, { "epoch": 3.08, "learning_rate": 6.296923076923077e-06, "loss": 0.1309, "step": 7725 }, { "epoch": 3.08, "learning_rate": 6.284102564102565e-06, "loss": 0.1173, "step": 7750 }, { "epoch": 3.08, "learning_rate": 6.271282051282052e-06, "loss": 0.1022, "step": 7775 }, { "epoch": 3.08, "learning_rate": 6.258461538461538e-06, "loss": 0.1279, "step": 7800 }, { "epoch": 3.08, "learning_rate": 6.245641025641026e-06, "loss": 0.1176, "step": 7825 }, { "epoch": 3.09, "learning_rate": 6.232820512820513e-06, "loss": 0.1065, "step": 7850 }, { "epoch": 3.09, "learning_rate": 6.220000000000001e-06, "loss": 0.1144, "step": 7875 }, { "epoch": 3.09, "learning_rate": 6.207179487179487e-06, "loss": 0.1021, "step": 7900 }, { "epoch": 3.09, "learning_rate": 6.194358974358975e-06, "loss": 0.0878, "step": 7925 }, { "epoch": 3.09, "learning_rate": 6.181538461538462e-06, "loss": 0.1041, "step": 7950 }, { "epoch": 3.09, "learning_rate": 6.16871794871795e-06, "loss": 0.1046, "step": 7975 }, { "epoch": 3.09, "learning_rate": 6.155897435897436e-06, "loss": 0.102, "step": 8000 }, { "epoch": 3.09, "eval_loss": 0.21563656628131866, "eval_runtime": 6638.017, "eval_samples_per_second": 2.462, "eval_steps_per_second": 0.308, "eval_wer": 9.873013155266387, "step": 8000 }, { "epoch": 3.09, "learning_rate": 6.143076923076924e-06, "loss": 0.1017, "step": 8025 }, { "epoch": 3.1, "learning_rate": 6.130256410256411e-06, "loss": 0.1001, "step": 8050 }, { "epoch": 3.1, "learning_rate": 6.1174358974358986e-06, "loss": 0.1165, "step": 8075 }, { "epoch": 3.1, "learning_rate": 6.104615384615385e-06, "loss": 0.1091, "step": 8100 }, { "epoch": 3.1, "learning_rate": 6.091794871794873e-06, "loss": 0.1272, "step": 8125 }, { "epoch": 3.1, "learning_rate": 6.07897435897436e-06, "loss": 0.1186, "step": 8150 }, { "epoch": 4.0, "learning_rate": 6.066666666666667e-06, "loss": 0.1119, "step": 8175 }, { "epoch": 4.0, "learning_rate": 6.053846153846155e-06, "loss": 0.1089, "step": 8200 }, { "epoch": 4.0, "learning_rate": 6.041025641025642e-06, "loss": 0.1153, "step": 8225 }, { "epoch": 4.0, "learning_rate": 6.028205128205129e-06, "loss": 0.1282, "step": 8250 }, { "epoch": 4.01, "learning_rate": 6.015384615384616e-06, "loss": 0.0927, "step": 8275 }, { "epoch": 4.01, "learning_rate": 6.002564102564103e-06, "loss": 0.126, "step": 8300 }, { "epoch": 4.01, "learning_rate": 5.989743589743591e-06, "loss": 0.0954, "step": 8325 }, { "epoch": 4.01, "learning_rate": 5.976923076923078e-06, "loss": 0.0779, "step": 8350 }, { "epoch": 4.01, "learning_rate": 5.9641025641025644e-06, "loss": 0.1008, "step": 8375 }, { "epoch": 4.01, "learning_rate": 5.951282051282051e-06, "loss": 0.1075, "step": 8400 }, { "epoch": 4.01, "learning_rate": 5.938461538461538e-06, "loss": 0.1009, "step": 8425 }, { "epoch": 4.01, "learning_rate": 5.925641025641026e-06, "loss": 0.0909, "step": 8450 }, { "epoch": 4.02, "learning_rate": 5.912820512820513e-06, "loss": 0.0648, "step": 8475 }, { "epoch": 4.02, "learning_rate": 5.9e-06, "loss": 0.1017, "step": 8500 }, { "epoch": 4.02, "learning_rate": 5.887179487179487e-06, "loss": 0.0917, "step": 8525 }, { "epoch": 4.02, "learning_rate": 5.874358974358975e-06, "loss": 0.1215, "step": 8550 }, { "epoch": 4.02, "learning_rate": 5.861538461538462e-06, "loss": 0.1137, "step": 8575 }, { "epoch": 4.02, "learning_rate": 5.848717948717949e-06, "loss": 0.0983, "step": 8600 }, { "epoch": 4.02, "learning_rate": 5.835897435897436e-06, "loss": 0.0924, "step": 8625 }, { "epoch": 4.02, "learning_rate": 5.823076923076924e-06, "loss": 0.092, "step": 8650 }, { "epoch": 4.03, "learning_rate": 5.8102564102564106e-06, "loss": 0.1072, "step": 8675 }, { "epoch": 4.03, "learning_rate": 5.797435897435898e-06, "loss": 0.1197, "step": 8700 }, { "epoch": 4.03, "learning_rate": 5.784615384615385e-06, "loss": 0.0975, "step": 8725 }, { "epoch": 4.03, "learning_rate": 5.771794871794873e-06, "loss": 0.1013, "step": 8750 }, { "epoch": 4.03, "learning_rate": 5.7589743589743596e-06, "loss": 0.0867, "step": 8775 }, { "epoch": 4.03, "learning_rate": 5.746153846153847e-06, "loss": 0.1081, "step": 8800 }, { "epoch": 4.03, "learning_rate": 5.733333333333334e-06, "loss": 0.0966, "step": 8825 }, { "epoch": 4.03, "learning_rate": 5.720512820512821e-06, "loss": 0.1101, "step": 8850 }, { "epoch": 4.04, "learning_rate": 5.7076923076923086e-06, "loss": 0.1154, "step": 8875 }, { "epoch": 4.04, "learning_rate": 5.694871794871795e-06, "loss": 0.0817, "step": 8900 }, { "epoch": 4.04, "learning_rate": 5.682051282051283e-06, "loss": 0.0593, "step": 8925 }, { "epoch": 4.04, "learning_rate": 5.66923076923077e-06, "loss": 0.0578, "step": 8950 }, { "epoch": 4.04, "learning_rate": 5.6564102564102575e-06, "loss": 0.1105, "step": 8975 }, { "epoch": 4.04, "learning_rate": 5.6435897435897435e-06, "loss": 0.1078, "step": 9000 }, { "epoch": 4.04, "learning_rate": 5.63076923076923e-06, "loss": 0.1131, "step": 9025 }, { "epoch": 4.04, "learning_rate": 5.617948717948718e-06, "loss": 0.0791, "step": 9050 }, { "epoch": 4.05, "learning_rate": 5.605128205128205e-06, "loss": 0.0868, "step": 9075 }, { "epoch": 4.05, "learning_rate": 5.5923076923076925e-06, "loss": 0.0855, "step": 9100 }, { "epoch": 4.05, "learning_rate": 5.579487179487179e-06, "loss": 0.0802, "step": 9125 }, { "epoch": 4.05, "learning_rate": 5.566666666666667e-06, "loss": 0.0634, "step": 9150 }, { "epoch": 4.05, "learning_rate": 5.553846153846154e-06, "loss": 0.086, "step": 9175 }, { "epoch": 4.05, "learning_rate": 5.5410256410256415e-06, "loss": 0.0728, "step": 9200 }, { "epoch": 4.05, "learning_rate": 5.528205128205128e-06, "loss": 0.0829, "step": 9225 }, { "epoch": 4.05, "learning_rate": 5.515384615384616e-06, "loss": 0.1078, "step": 9250 }, { "epoch": 4.06, "learning_rate": 5.502564102564103e-06, "loss": 0.1151, "step": 9275 }, { "epoch": 4.06, "learning_rate": 5.4897435897435905e-06, "loss": 0.1386, "step": 9300 }, { "epoch": 4.06, "learning_rate": 5.476923076923077e-06, "loss": 0.0767, "step": 9325 }, { "epoch": 4.06, "learning_rate": 5.464102564102565e-06, "loss": 0.0681, "step": 9350 }, { "epoch": 4.06, "learning_rate": 5.451282051282052e-06, "loss": 0.0849, "step": 9375 }, { "epoch": 4.06, "learning_rate": 5.4384615384615395e-06, "loss": 0.0789, "step": 9400 }, { "epoch": 4.06, "learning_rate": 5.425641025641026e-06, "loss": 0.0867, "step": 9425 }, { "epoch": 4.06, "learning_rate": 5.412820512820514e-06, "loss": 0.0707, "step": 9450 }, { "epoch": 4.07, "learning_rate": 5.400000000000001e-06, "loss": 0.0718, "step": 9475 }, { "epoch": 4.07, "learning_rate": 5.387179487179488e-06, "loss": 0.0921, "step": 9500 }, { "epoch": 4.07, "learning_rate": 5.374358974358975e-06, "loss": 0.0984, "step": 9525 }, { "epoch": 4.07, "learning_rate": 5.361538461538462e-06, "loss": 0.1199, "step": 9550 }, { "epoch": 4.07, "learning_rate": 5.34871794871795e-06, "loss": 0.1196, "step": 9575 }, { "epoch": 4.07, "learning_rate": 5.335897435897436e-06, "loss": 0.1144, "step": 9600 }, { "epoch": 4.07, "learning_rate": 5.323076923076923e-06, "loss": 0.0895, "step": 9625 }, { "epoch": 4.07, "learning_rate": 5.31025641025641e-06, "loss": 0.1018, "step": 9650 }, { "epoch": 4.08, "learning_rate": 5.297435897435897e-06, "loss": 0.1285, "step": 9675 }, { "epoch": 4.08, "learning_rate": 5.284615384615385e-06, "loss": 0.0886, "step": 9700 }, { "epoch": 4.08, "learning_rate": 5.271794871794872e-06, "loss": 0.0819, "step": 9725 }, { "epoch": 4.08, "learning_rate": 5.258974358974359e-06, "loss": 0.1005, "step": 9750 }, { "epoch": 4.08, "learning_rate": 5.246153846153846e-06, "loss": 0.0895, "step": 9775 }, { "epoch": 4.08, "learning_rate": 5.233333333333334e-06, "loss": 0.0885, "step": 9800 }, { "epoch": 4.08, "learning_rate": 5.220512820512821e-06, "loss": 0.085, "step": 9825 }, { "epoch": 4.08, "learning_rate": 5.207692307692308e-06, "loss": 0.0726, "step": 9850 }, { "epoch": 4.09, "learning_rate": 5.194871794871795e-06, "loss": 0.0541, "step": 9875 }, { "epoch": 4.09, "learning_rate": 5.182051282051283e-06, "loss": 0.0707, "step": 9900 }, { "epoch": 4.09, "learning_rate": 5.16923076923077e-06, "loss": 0.0922, "step": 9925 }, { "epoch": 4.09, "learning_rate": 5.156410256410257e-06, "loss": 0.0859, "step": 9950 }, { "epoch": 4.09, "learning_rate": 5.143589743589744e-06, "loss": 0.0831, "step": 9975 }, { "epoch": 4.09, "learning_rate": 5.130769230769232e-06, "loss": 0.0706, "step": 10000 }, { "epoch": 4.09, "eval_loss": 0.212646484375, "eval_runtime": 6654.4763, "eval_samples_per_second": 2.455, "eval_steps_per_second": 0.307, "eval_wer": 9.617898011015038, "step": 10000 }, { "epoch": 4.09, "learning_rate": 5.1179487179487186e-06, "loss": 0.0725, "step": 10025 }, { "epoch": 4.09, "learning_rate": 5.105128205128206e-06, "loss": 0.0912, "step": 10050 }, { "epoch": 4.1, "learning_rate": 5.092307692307693e-06, "loss": 0.0808, "step": 10075 }, { "epoch": 4.1, "learning_rate": 5.07948717948718e-06, "loss": 0.0802, "step": 10100 }, { "epoch": 4.1, "learning_rate": 5.0666666666666676e-06, "loss": 0.0973, "step": 10125 }, { "epoch": 4.1, "learning_rate": 5.053846153846154e-06, "loss": 0.0966, "step": 10150 }, { "epoch": 4.1, "learning_rate": 5.041025641025642e-06, "loss": 0.0799, "step": 10175 }, { "epoch": 4.1, "learning_rate": 5.028205128205128e-06, "loss": 0.0858, "step": 10200 }, { "epoch": 5.0, "learning_rate": 5.015384615384616e-06, "loss": 0.0843, "step": 10225 }, { "epoch": 5.0, "learning_rate": 5.003076923076924e-06, "loss": 0.0578, "step": 10250 }, { "epoch": 5.0, "learning_rate": 4.990256410256411e-06, "loss": 0.0534, "step": 10275 }, { "epoch": 5.0, "learning_rate": 4.977435897435898e-06, "loss": 0.0478, "step": 10300 }, { "epoch": 5.01, "learning_rate": 4.964615384615385e-06, "loss": 0.0471, "step": 10325 }, { "epoch": 5.01, "learning_rate": 4.951794871794872e-06, "loss": 0.041, "step": 10350 }, { "epoch": 5.01, "learning_rate": 4.93897435897436e-06, "loss": 0.0465, "step": 10375 }, { "epoch": 5.01, "learning_rate": 4.926153846153847e-06, "loss": 0.0501, "step": 10400 }, { "epoch": 5.01, "learning_rate": 4.9133333333333334e-06, "loss": 0.0399, "step": 10425 }, { "epoch": 5.01, "learning_rate": 4.900512820512821e-06, "loss": 0.0416, "step": 10450 }, { "epoch": 5.01, "learning_rate": 4.887692307692308e-06, "loss": 0.0396, "step": 10475 }, { "epoch": 5.01, "learning_rate": 4.874871794871796e-06, "loss": 0.0393, "step": 10500 }, { "epoch": 5.02, "learning_rate": 4.8620512820512824e-06, "loss": 0.0416, "step": 10525 }, { "epoch": 5.02, "learning_rate": 4.849230769230769e-06, "loss": 0.0427, "step": 10550 }, { "epoch": 5.02, "learning_rate": 4.836410256410257e-06, "loss": 0.0398, "step": 10575 }, { "epoch": 5.02, "learning_rate": 4.823589743589744e-06, "loss": 0.0591, "step": 10600 }, { "epoch": 5.02, "learning_rate": 4.8107692307692314e-06, "loss": 0.0392, "step": 10625 }, { "epoch": 5.02, "learning_rate": 4.797948717948718e-06, "loss": 0.043, "step": 10650 }, { "epoch": 5.02, "learning_rate": 4.785128205128206e-06, "loss": 0.0479, "step": 10675 }, { "epoch": 5.02, "learning_rate": 4.772307692307693e-06, "loss": 0.043, "step": 10700 }, { "epoch": 5.03, "learning_rate": 4.7594871794871796e-06, "loss": 0.0446, "step": 10725 }, { "epoch": 5.03, "learning_rate": 4.746666666666667e-06, "loss": 0.0492, "step": 10750 }, { "epoch": 5.03, "learning_rate": 4.733846153846154e-06, "loss": 0.0389, "step": 10775 }, { "epoch": 5.03, "learning_rate": 4.721025641025642e-06, "loss": 0.0344, "step": 10800 }, { "epoch": 5.03, "learning_rate": 4.7082051282051286e-06, "loss": 0.0394, "step": 10825 }, { "epoch": 5.03, "learning_rate": 4.695384615384615e-06, "loss": 0.0449, "step": 10850 }, { "epoch": 5.03, "learning_rate": 4.682564102564103e-06, "loss": 0.0395, "step": 10875 }, { "epoch": 5.03, "learning_rate": 4.66974358974359e-06, "loss": 0.0348, "step": 10900 }, { "epoch": 5.04, "learning_rate": 4.6569230769230776e-06, "loss": 0.0327, "step": 10925 }, { "epoch": 5.04, "learning_rate": 4.644102564102564e-06, "loss": 0.043, "step": 10950 }, { "epoch": 5.04, "learning_rate": 4.631282051282052e-06, "loss": 0.0373, "step": 10975 }, { "epoch": 5.04, "learning_rate": 4.618461538461539e-06, "loss": 0.0464, "step": 11000 }, { "epoch": 5.04, "learning_rate": 4.605641025641026e-06, "loss": 0.0428, "step": 11025 }, { "epoch": 5.04, "learning_rate": 4.592820512820513e-06, "loss": 0.0395, "step": 11050 }, { "epoch": 5.04, "learning_rate": 4.58e-06, "loss": 0.0491, "step": 11075 }, { "epoch": 5.04, "learning_rate": 4.567179487179488e-06, "loss": 0.0543, "step": 11100 }, { "epoch": 5.05, "learning_rate": 4.554358974358975e-06, "loss": 0.0385, "step": 11125 }, { "epoch": 5.05, "learning_rate": 4.5415384615384615e-06, "loss": 0.0457, "step": 11150 }, { "epoch": 5.05, "learning_rate": 4.528717948717949e-06, "loss": 0.0589, "step": 11175 }, { "epoch": 5.05, "learning_rate": 4.515897435897436e-06, "loss": 0.0506, "step": 11200 }, { "epoch": 5.05, "learning_rate": 4.503076923076924e-06, "loss": 0.0353, "step": 11225 }, { "epoch": 5.05, "learning_rate": 4.4902564102564105e-06, "loss": 0.0464, "step": 11250 }, { "epoch": 5.05, "learning_rate": 4.477435897435898e-06, "loss": 0.0502, "step": 11275 }, { "epoch": 5.05, "learning_rate": 4.464615384615385e-06, "loss": 0.056, "step": 11300 }, { "epoch": 5.06, "learning_rate": 4.451794871794872e-06, "loss": 0.047, "step": 11325 }, { "epoch": 5.06, "learning_rate": 4.4389743589743595e-06, "loss": 0.0501, "step": 11350 }, { "epoch": 5.06, "learning_rate": 4.426153846153846e-06, "loss": 0.0432, "step": 11375 }, { "epoch": 5.06, "learning_rate": 4.413333333333334e-06, "loss": 0.0443, "step": 11400 }, { "epoch": 5.06, "learning_rate": 4.400512820512821e-06, "loss": 0.0399, "step": 11425 }, { "epoch": 5.06, "learning_rate": 4.387692307692308e-06, "loss": 0.0407, "step": 11450 }, { "epoch": 5.06, "learning_rate": 4.374871794871795e-06, "loss": 0.048, "step": 11475 }, { "epoch": 5.06, "learning_rate": 4.362051282051282e-06, "loss": 0.0479, "step": 11500 }, { "epoch": 5.07, "learning_rate": 4.34923076923077e-06, "loss": 0.0398, "step": 11525 }, { "epoch": 5.07, "learning_rate": 4.336410256410257e-06, "loss": 0.0322, "step": 11550 }, { "epoch": 5.07, "learning_rate": 4.323589743589744e-06, "loss": 0.0449, "step": 11575 }, { "epoch": 5.07, "learning_rate": 4.310769230769231e-06, "loss": 0.04, "step": 11600 }, { "epoch": 5.07, "learning_rate": 4.297948717948718e-06, "loss": 0.0385, "step": 11625 }, { "epoch": 5.07, "learning_rate": 4.285128205128206e-06, "loss": 0.0445, "step": 11650 }, { "epoch": 5.07, "learning_rate": 4.2723076923076925e-06, "loss": 0.0417, "step": 11675 }, { "epoch": 5.07, "learning_rate": 4.25948717948718e-06, "loss": 0.045, "step": 11700 }, { "epoch": 5.08, "learning_rate": 4.246666666666667e-06, "loss": 0.0476, "step": 11725 }, { "epoch": 5.08, "learning_rate": 4.233846153846154e-06, "loss": 0.0468, "step": 11750 }, { "epoch": 5.08, "learning_rate": 4.2210256410256414e-06, "loss": 0.0425, "step": 11775 }, { "epoch": 5.08, "learning_rate": 4.208205128205128e-06, "loss": 0.0451, "step": 11800 }, { "epoch": 5.08, "learning_rate": 4.195384615384616e-06, "loss": 0.0513, "step": 11825 }, { "epoch": 5.08, "learning_rate": 4.182564102564103e-06, "loss": 0.042, "step": 11850 }, { "epoch": 5.08, "learning_rate": 4.1697435897435904e-06, "loss": 0.038, "step": 11875 }, { "epoch": 5.08, "learning_rate": 4.156923076923077e-06, "loss": 0.0465, "step": 11900 }, { "epoch": 5.09, "learning_rate": 4.144102564102564e-06, "loss": 0.0486, "step": 11925 }, { "epoch": 5.09, "learning_rate": 4.131282051282052e-06, "loss": 0.0443, "step": 11950 }, { "epoch": 5.09, "learning_rate": 4.118461538461539e-06, "loss": 0.0402, "step": 11975 }, { "epoch": 5.09, "learning_rate": 4.105641025641026e-06, "loss": 0.0428, "step": 12000 }, { "epoch": 5.09, "eval_loss": 0.21781405806541443, "eval_runtime": 6662.3144, "eval_samples_per_second": 2.453, "eval_steps_per_second": 0.307, "eval_wer": 9.340524498473304, "step": 12000 }, { "epoch": 5.09, "learning_rate": 4.092820512820513e-06, "loss": 0.0358, "step": 12025 }, { "epoch": 5.09, "learning_rate": 4.08e-06, "loss": 0.0411, "step": 12050 }, { "epoch": 5.09, "learning_rate": 4.0671794871794876e-06, "loss": 0.0452, "step": 12075 }, { "epoch": 5.09, "learning_rate": 4.054358974358974e-06, "loss": 0.0439, "step": 12100 }, { "epoch": 5.1, "learning_rate": 4.041538461538462e-06, "loss": 0.0352, "step": 12125 }, { "epoch": 5.1, "learning_rate": 4.028717948717949e-06, "loss": 0.0429, "step": 12150 }, { "epoch": 5.1, "learning_rate": 4.0158974358974366e-06, "loss": 0.0421, "step": 12175 }, { "epoch": 5.1, "learning_rate": 4.003076923076923e-06, "loss": 0.0406, "step": 12200 }, { "epoch": 5.1, "learning_rate": 3.990256410256411e-06, "loss": 0.0493, "step": 12225 }, { "epoch": 5.1, "learning_rate": 3.977435897435898e-06, "loss": 0.0363, "step": 12250 }, { "epoch": 6.0, "learning_rate": 3.964615384615385e-06, "loss": 0.0482, "step": 12275 }, { "epoch": 6.0, "learning_rate": 3.951794871794872e-06, "loss": 0.0654, "step": 12300 }, { "epoch": 6.0, "learning_rate": 3.938974358974359e-06, "loss": 0.0544, "step": 12325 }, { "epoch": 6.0, "learning_rate": 3.926153846153846e-06, "loss": 0.046, "step": 12350 }, { "epoch": 6.01, "learning_rate": 3.913333333333334e-06, "loss": 0.0559, "step": 12375 }, { "epoch": 6.01, "learning_rate": 3.9005128205128205e-06, "loss": 0.0639, "step": 12400 }, { "epoch": 6.01, "learning_rate": 3.887692307692308e-06, "loss": 0.0614, "step": 12425 }, { "epoch": 6.01, "learning_rate": 3.874871794871795e-06, "loss": 0.0514, "step": 12450 }, { "epoch": 6.01, "learning_rate": 3.862051282051283e-06, "loss": 0.0534, "step": 12475 }, { "epoch": 6.01, "learning_rate": 3.8492307692307695e-06, "loss": 0.05, "step": 12500 }, { "epoch": 6.01, "learning_rate": 3.836410256410257e-06, "loss": 0.0528, "step": 12525 }, { "epoch": 6.01, "learning_rate": 3.823589743589744e-06, "loss": 0.0509, "step": 12550 }, { "epoch": 6.02, "learning_rate": 3.8107692307692313e-06, "loss": 0.0485, "step": 12575 }, { "epoch": 6.02, "learning_rate": 3.7979487179487185e-06, "loss": 0.0404, "step": 12600 }, { "epoch": 6.02, "learning_rate": 3.7851282051282058e-06, "loss": 0.0532, "step": 12625 }, { "epoch": 6.02, "learning_rate": 3.772307692307693e-06, "loss": 0.0535, "step": 12650 }, { "epoch": 6.02, "learning_rate": 3.7594871794871794e-06, "loss": 0.0458, "step": 12675 }, { "epoch": 6.02, "learning_rate": 3.7466666666666667e-06, "loss": 0.0449, "step": 12700 }, { "epoch": 6.02, "learning_rate": 3.733846153846154e-06, "loss": 0.0494, "step": 12725 }, { "epoch": 6.02, "learning_rate": 3.721025641025641e-06, "loss": 0.0518, "step": 12750 }, { "epoch": 6.03, "learning_rate": 3.7087179487179494e-06, "loss": 0.052, "step": 12775 }, { "epoch": 6.03, "learning_rate": 3.6958974358974362e-06, "loss": 0.0505, "step": 12800 }, { "epoch": 6.03, "learning_rate": 3.683076923076923e-06, "loss": 0.044, "step": 12825 }, { "epoch": 6.03, "learning_rate": 3.6702564102564103e-06, "loss": 0.0479, "step": 12850 }, { "epoch": 6.03, "learning_rate": 3.6574358974358976e-06, "loss": 0.0562, "step": 12875 }, { "epoch": 6.03, "learning_rate": 3.644615384615385e-06, "loss": 0.0498, "step": 12900 }, { "epoch": 6.03, "learning_rate": 3.631794871794872e-06, "loss": 0.0591, "step": 12925 }, { "epoch": 6.03, "learning_rate": 3.6189743589743593e-06, "loss": 0.0444, "step": 12950 }, { "epoch": 6.04, "learning_rate": 3.6061538461538466e-06, "loss": 0.0498, "step": 12975 }, { "epoch": 6.04, "learning_rate": 3.593333333333334e-06, "loss": 0.0538, "step": 13000 }, { "epoch": 6.04, "learning_rate": 3.580512820512821e-06, "loss": 0.049, "step": 13025 }, { "epoch": 6.04, "learning_rate": 3.5676923076923083e-06, "loss": 0.0404, "step": 13050 }, { "epoch": 6.04, "learning_rate": 3.5548717948717955e-06, "loss": 0.0493, "step": 13075 }, { "epoch": 6.04, "learning_rate": 3.5420512820512824e-06, "loss": 0.0591, "step": 13100 }, { "epoch": 6.04, "learning_rate": 3.5292307692307696e-06, "loss": 0.0589, "step": 13125 }, { "epoch": 6.04, "learning_rate": 3.5164102564102564e-06, "loss": 0.0385, "step": 13150 }, { "epoch": 6.05, "learning_rate": 3.5035897435897437e-06, "loss": 0.0441, "step": 13175 }, { "epoch": 6.05, "learning_rate": 3.490769230769231e-06, "loss": 0.044, "step": 13200 }, { "epoch": 6.05, "learning_rate": 3.477948717948718e-06, "loss": 0.044, "step": 13225 }, { "epoch": 6.05, "learning_rate": 3.4651282051282054e-06, "loss": 0.0392, "step": 13250 }, { "epoch": 6.05, "learning_rate": 3.4523076923076927e-06, "loss": 0.0421, "step": 13275 }, { "epoch": 6.05, "learning_rate": 3.43948717948718e-06, "loss": 0.0402, "step": 13300 }, { "epoch": 6.05, "learning_rate": 3.426666666666667e-06, "loss": 0.0498, "step": 13325 }, { "epoch": 6.05, "learning_rate": 3.4138461538461544e-06, "loss": 0.0439, "step": 13350 }, { "epoch": 6.06, "learning_rate": 3.4010256410256417e-06, "loss": 0.0532, "step": 13375 }, { "epoch": 6.06, "learning_rate": 3.3882051282051285e-06, "loss": 0.0386, "step": 13400 }, { "epoch": 6.06, "learning_rate": 3.3753846153846157e-06, "loss": 0.0431, "step": 13425 }, { "epoch": 6.06, "learning_rate": 3.3625641025641026e-06, "loss": 0.041, "step": 13450 }, { "epoch": 6.06, "learning_rate": 3.34974358974359e-06, "loss": 0.0471, "step": 13475 }, { "epoch": 6.06, "learning_rate": 3.336923076923077e-06, "loss": 0.0442, "step": 13500 }, { "epoch": 6.06, "learning_rate": 3.3241025641025643e-06, "loss": 0.051, "step": 13525 }, { "epoch": 6.06, "learning_rate": 3.3112820512820516e-06, "loss": 0.0501, "step": 13550 }, { "epoch": 6.07, "learning_rate": 3.298461538461539e-06, "loss": 0.0547, "step": 13575 }, { "epoch": 6.07, "learning_rate": 3.285641025641026e-06, "loss": 0.0568, "step": 13600 }, { "epoch": 6.07, "learning_rate": 3.2728205128205133e-06, "loss": 0.0537, "step": 13625 }, { "epoch": 6.07, "learning_rate": 3.2600000000000006e-06, "loss": 0.0532, "step": 13650 }, { "epoch": 6.07, "learning_rate": 3.247179487179488e-06, "loss": 0.0429, "step": 13675 }, { "epoch": 6.07, "learning_rate": 3.2343589743589746e-06, "loss": 0.0463, "step": 13700 }, { "epoch": 6.07, "learning_rate": 3.221538461538462e-06, "loss": 0.051, "step": 13725 }, { "epoch": 6.07, "learning_rate": 3.2087179487179487e-06, "loss": 0.0525, "step": 13750 }, { "epoch": 6.08, "learning_rate": 3.195897435897436e-06, "loss": 0.0522, "step": 13775 }, { "epoch": 6.08, "learning_rate": 3.183076923076923e-06, "loss": 0.0512, "step": 13800 }, { "epoch": 6.08, "learning_rate": 3.1702564102564104e-06, "loss": 0.0403, "step": 13825 }, { "epoch": 6.08, "learning_rate": 3.1574358974358977e-06, "loss": 0.0461, "step": 13850 }, { "epoch": 6.08, "learning_rate": 3.144615384615385e-06, "loss": 0.0398, "step": 13875 }, { "epoch": 6.08, "learning_rate": 3.131794871794872e-06, "loss": 0.0501, "step": 13900 }, { "epoch": 6.08, "learning_rate": 3.1189743589743594e-06, "loss": 0.0454, "step": 13925 }, { "epoch": 6.08, "learning_rate": 3.1061538461538467e-06, "loss": 0.0688, "step": 13950 }, { "epoch": 6.09, "learning_rate": 3.093333333333334e-06, "loss": 0.0597, "step": 13975 }, { "epoch": 6.09, "learning_rate": 3.080512820512821e-06, "loss": 0.0503, "step": 14000 }, { "epoch": 6.09, "eval_loss": 0.2109454870223999, "eval_runtime": 6639.7022, "eval_samples_per_second": 2.461, "eval_steps_per_second": 0.308, "eval_wer": 9.13563336472334, "step": 14000 }, { "epoch": 6.09, "learning_rate": 3.067692307692308e-06, "loss": 0.0394, "step": 14025 }, { "epoch": 6.09, "learning_rate": 3.054871794871795e-06, "loss": 0.0411, "step": 14050 }, { "epoch": 6.09, "learning_rate": 3.042051282051282e-06, "loss": 0.0518, "step": 14075 }, { "epoch": 6.09, "learning_rate": 3.0292307692307693e-06, "loss": 0.0459, "step": 14100 }, { "epoch": 6.09, "learning_rate": 3.0164102564102566e-06, "loss": 0.0491, "step": 14125 }, { "epoch": 6.09, "learning_rate": 3.003589743589744e-06, "loss": 0.0448, "step": 14150 }, { "epoch": 6.1, "learning_rate": 2.990769230769231e-06, "loss": 0.0507, "step": 14175 }, { "epoch": 6.1, "learning_rate": 2.9779487179487183e-06, "loss": 0.0461, "step": 14200 }, { "epoch": 6.1, "learning_rate": 2.9651282051282056e-06, "loss": 0.0436, "step": 14225 }, { "epoch": 6.1, "learning_rate": 2.952307692307693e-06, "loss": 0.0465, "step": 14250 }, { "epoch": 6.1, "learning_rate": 2.93948717948718e-06, "loss": 0.0427, "step": 14275 }, { "epoch": 6.1, "learning_rate": 2.9266666666666673e-06, "loss": 0.051, "step": 14300 }, { "epoch": 7.0, "learning_rate": 2.913846153846154e-06, "loss": 0.142, "step": 14325 }, { "epoch": 7.0, "learning_rate": 2.901025641025641e-06, "loss": 0.1518, "step": 14350 }, { "epoch": 7.0, "learning_rate": 2.888205128205128e-06, "loss": 0.1361, "step": 14375 }, { "epoch": 7.0, "learning_rate": 2.8753846153846154e-06, "loss": 0.135, "step": 14400 }, { "epoch": 7.01, "learning_rate": 2.8625641025641027e-06, "loss": 0.152, "step": 14425 }, { "epoch": 7.01, "learning_rate": 2.84974358974359e-06, "loss": 0.1646, "step": 14450 }, { "epoch": 7.01, "learning_rate": 2.836923076923077e-06, "loss": 0.1626, "step": 14475 }, { "epoch": 7.01, "learning_rate": 2.8241025641025644e-06, "loss": 0.1831, "step": 14500 }, { "epoch": 7.01, "learning_rate": 2.8112820512820517e-06, "loss": 0.1755, "step": 14525 }, { "epoch": 7.01, "learning_rate": 2.798461538461539e-06, "loss": 0.2315, "step": 14550 }, { "epoch": 7.01, "learning_rate": 2.785641025641026e-06, "loss": 0.1908, "step": 14575 }, { "epoch": 7.01, "learning_rate": 2.7728205128205134e-06, "loss": 0.145, "step": 14600 }, { "epoch": 7.02, "learning_rate": 2.7600000000000003e-06, "loss": 0.1282, "step": 14625 }, { "epoch": 7.02, "learning_rate": 2.7471794871794875e-06, "loss": 0.1179, "step": 14650 }, { "epoch": 7.02, "learning_rate": 2.7343589743589743e-06, "loss": 0.1075, "step": 14675 }, { "epoch": 7.02, "learning_rate": 2.7215384615384616e-06, "loss": 0.0984, "step": 14700 }, { "epoch": 7.02, "learning_rate": 2.708717948717949e-06, "loss": 0.0799, "step": 14725 }, { "epoch": 7.02, "learning_rate": 2.695897435897436e-06, "loss": 0.0683, "step": 14750 }, { "epoch": 7.02, "learning_rate": 2.6830769230769233e-06, "loss": 0.0858, "step": 14775 }, { "epoch": 7.02, "learning_rate": 2.6702564102564106e-06, "loss": 0.0797, "step": 14800 }, { "epoch": 7.03, "learning_rate": 2.657435897435898e-06, "loss": 0.0853, "step": 14825 }, { "epoch": 7.03, "learning_rate": 2.644615384615385e-06, "loss": 0.0662, "step": 14850 }, { "epoch": 7.03, "learning_rate": 2.6317948717948723e-06, "loss": 0.0706, "step": 14875 }, { "epoch": 7.03, "learning_rate": 2.6189743589743596e-06, "loss": 0.0691, "step": 14900 }, { "epoch": 7.03, "learning_rate": 2.6061538461538464e-06, "loss": 0.0637, "step": 14925 }, { "epoch": 7.03, "learning_rate": 2.5933333333333336e-06, "loss": 0.0705, "step": 14950 }, { "epoch": 7.03, "learning_rate": 2.5805128205128205e-06, "loss": 0.0596, "step": 14975 }, { "epoch": 7.03, "learning_rate": 2.5676923076923077e-06, "loss": 0.0728, "step": 15000 }, { "epoch": 7.04, "learning_rate": 2.554871794871795e-06, "loss": 0.0738, "step": 15025 }, { "epoch": 7.04, "learning_rate": 2.542051282051282e-06, "loss": 0.0694, "step": 15050 }, { "epoch": 7.04, "learning_rate": 2.5292307692307694e-06, "loss": 0.0671, "step": 15075 }, { "epoch": 7.04, "learning_rate": 2.5164102564102567e-06, "loss": 0.0722, "step": 15100 }, { "epoch": 7.04, "learning_rate": 2.503589743589744e-06, "loss": 0.0712, "step": 15125 }, { "epoch": 7.04, "learning_rate": 2.490769230769231e-06, "loss": 0.0869, "step": 15150 }, { "epoch": 7.04, "learning_rate": 2.4779487179487184e-06, "loss": 0.0673, "step": 15175 }, { "epoch": 7.04, "learning_rate": 2.4651282051282053e-06, "loss": 0.0779, "step": 15200 }, { "epoch": 7.05, "learning_rate": 2.4523076923076925e-06, "loss": 0.0661, "step": 15225 }, { "epoch": 7.05, "learning_rate": 2.4394871794871798e-06, "loss": 0.0725, "step": 15250 }, { "epoch": 7.05, "learning_rate": 2.426666666666667e-06, "loss": 0.1081, "step": 15275 }, { "epoch": 7.05, "learning_rate": 2.4138461538461543e-06, "loss": 0.1098, "step": 15300 }, { "epoch": 7.05, "learning_rate": 2.4010256410256415e-06, "loss": 0.0936, "step": 15325 }, { "epoch": 7.05, "learning_rate": 2.3882051282051283e-06, "loss": 0.1135, "step": 15350 }, { "epoch": 7.05, "learning_rate": 2.3753846153846156e-06, "loss": 0.0834, "step": 15375 }, { "epoch": 7.05, "learning_rate": 2.362564102564103e-06, "loss": 0.0623, "step": 15400 }, { "epoch": 7.06, "learning_rate": 2.34974358974359e-06, "loss": 0.0662, "step": 15425 }, { "epoch": 7.06, "learning_rate": 2.3374358974358975e-06, "loss": 0.0469, "step": 15450 }, { "epoch": 7.06, "learning_rate": 2.3246153846153847e-06, "loss": 0.0677, "step": 15475 }, { "epoch": 7.06, "learning_rate": 2.311794871794872e-06, "loss": 0.0712, "step": 15500 }, { "epoch": 7.06, "learning_rate": 2.2989743589743592e-06, "loss": 0.0799, "step": 15525 }, { "epoch": 7.06, "learning_rate": 2.2861538461538465e-06, "loss": 0.0841, "step": 15550 }, { "epoch": 7.06, "learning_rate": 2.2733333333333333e-06, "loss": 0.0785, "step": 15575 }, { "epoch": 7.06, "learning_rate": 2.2605128205128206e-06, "loss": 0.0658, "step": 15600 }, { "epoch": 7.07, "learning_rate": 2.247692307692308e-06, "loss": 0.0588, "step": 15625 }, { "epoch": 7.07, "learning_rate": 2.234871794871795e-06, "loss": 0.0562, "step": 15650 }, { "epoch": 7.07, "learning_rate": 2.2220512820512823e-06, "loss": 0.0512, "step": 15675 }, { "epoch": 7.07, "learning_rate": 2.2092307692307695e-06, "loss": 0.0599, "step": 15700 }, { "epoch": 7.07, "learning_rate": 2.1964102564102564e-06, "loss": 0.065, "step": 15725 }, { "epoch": 7.07, "learning_rate": 2.1835897435897436e-06, "loss": 0.051, "step": 15750 }, { "epoch": 7.07, "learning_rate": 2.170769230769231e-06, "loss": 0.0697, "step": 15775 }, { "epoch": 7.07, "learning_rate": 2.157948717948718e-06, "loss": 0.0543, "step": 15800 }, { "epoch": 7.08, "learning_rate": 2.1451282051282054e-06, "loss": 0.0547, "step": 15825 }, { "epoch": 7.08, "learning_rate": 2.1323076923076926e-06, "loss": 0.0661, "step": 15850 }, { "epoch": 7.08, "learning_rate": 2.1194871794871794e-06, "loss": 0.0724, "step": 15875 }, { "epoch": 7.08, "learning_rate": 2.1066666666666667e-06, "loss": 0.0718, "step": 15900 }, { "epoch": 7.08, "learning_rate": 2.093846153846154e-06, "loss": 0.0838, "step": 15925 }, { "epoch": 7.08, "learning_rate": 2.081025641025641e-06, "loss": 0.0672, "step": 15950 }, { "epoch": 7.08, "learning_rate": 2.0682051282051284e-06, "loss": 0.0766, "step": 15975 }, { "epoch": 7.08, "learning_rate": 2.0553846153846157e-06, "loss": 0.0778, "step": 16000 }, { "epoch": 7.08, "eval_loss": 0.2057988941669464, "eval_runtime": 6661.3277, "eval_samples_per_second": 2.453, "eval_steps_per_second": 0.307, "eval_wer": 9.200125560026255, "step": 16000 }, { "epoch": 7.09, "learning_rate": 2.0425641025641025e-06, "loss": 0.0652, "step": 16025 }, { "epoch": 7.09, "learning_rate": 2.0297435897435897e-06, "loss": 0.065, "step": 16050 }, { "epoch": 7.09, "learning_rate": 2.016923076923077e-06, "loss": 0.0793, "step": 16075 }, { "epoch": 7.09, "learning_rate": 2.0041025641025642e-06, "loss": 0.0676, "step": 16100 }, { "epoch": 7.09, "learning_rate": 1.9912820512820515e-06, "loss": 0.0646, "step": 16125 }, { "epoch": 7.09, "learning_rate": 1.9784615384615387e-06, "loss": 0.0694, "step": 16150 }, { "epoch": 7.09, "learning_rate": 1.9656410256410256e-06, "loss": 0.0744, "step": 16175 }, { "epoch": 7.09, "learning_rate": 1.952820512820513e-06, "loss": 0.0762, "step": 16200 }, { "epoch": 7.1, "learning_rate": 1.94e-06, "loss": 0.0628, "step": 16225 }, { "epoch": 7.1, "learning_rate": 1.9271794871794873e-06, "loss": 0.0637, "step": 16250 }, { "epoch": 7.1, "learning_rate": 1.9143589743589746e-06, "loss": 0.0784, "step": 16275 }, { "epoch": 7.1, "learning_rate": 1.9015384615384616e-06, "loss": 0.0708, "step": 16300 }, { "epoch": 7.1, "learning_rate": 1.8887179487179488e-06, "loss": 0.0644, "step": 16325 }, { "epoch": 8.0, "learning_rate": 1.875897435897436e-06, "loss": 0.0593, "step": 16350 }, { "epoch": 8.0, "learning_rate": 1.8630769230769233e-06, "loss": 0.0519, "step": 16375 }, { "epoch": 8.0, "learning_rate": 1.8502564102564106e-06, "loss": 0.0543, "step": 16400 }, { "epoch": 8.0, "learning_rate": 1.8374358974358974e-06, "loss": 0.0647, "step": 16425 }, { "epoch": 8.01, "learning_rate": 1.8246153846153847e-06, "loss": 0.0505, "step": 16450 }, { "epoch": 8.01, "learning_rate": 1.811794871794872e-06, "loss": 0.0614, "step": 16475 }, { "epoch": 8.01, "learning_rate": 1.7989743589743592e-06, "loss": 0.0439, "step": 16500 }, { "epoch": 8.01, "learning_rate": 1.7861538461538464e-06, "loss": 0.0504, "step": 16525 }, { "epoch": 8.01, "learning_rate": 1.7733333333333336e-06, "loss": 0.0499, "step": 16550 }, { "epoch": 8.01, "learning_rate": 1.7605128205128205e-06, "loss": 0.0355, "step": 16575 }, { "epoch": 8.01, "learning_rate": 1.7476923076923077e-06, "loss": 0.0351, "step": 16600 }, { "epoch": 8.01, "learning_rate": 1.734871794871795e-06, "loss": 0.0456, "step": 16625 }, { "epoch": 8.02, "learning_rate": 1.7220512820512822e-06, "loss": 0.041, "step": 16650 }, { "epoch": 8.02, "learning_rate": 1.7092307692307695e-06, "loss": 0.036, "step": 16675 }, { "epoch": 8.02, "learning_rate": 1.6964102564102567e-06, "loss": 0.0399, "step": 16700 }, { "epoch": 8.02, "learning_rate": 1.683589743589744e-06, "loss": 0.0347, "step": 16725 }, { "epoch": 8.02, "learning_rate": 1.6707692307692308e-06, "loss": 0.0397, "step": 16750 }, { "epoch": 8.02, "learning_rate": 1.657948717948718e-06, "loss": 0.031, "step": 16775 }, { "epoch": 8.02, "learning_rate": 1.6451282051282053e-06, "loss": 0.0408, "step": 16800 }, { "epoch": 8.02, "learning_rate": 1.6323076923076925e-06, "loss": 0.0317, "step": 16825 }, { "epoch": 8.03, "learning_rate": 1.6194871794871798e-06, "loss": 0.0344, "step": 16850 }, { "epoch": 8.03, "learning_rate": 1.606666666666667e-06, "loss": 0.0311, "step": 16875 }, { "epoch": 8.03, "learning_rate": 1.5938461538461538e-06, "loss": 0.0322, "step": 16900 }, { "epoch": 8.03, "learning_rate": 1.581025641025641e-06, "loss": 0.0275, "step": 16925 }, { "epoch": 8.03, "learning_rate": 1.5682051282051283e-06, "loss": 0.0215, "step": 16950 }, { "epoch": 8.03, "learning_rate": 1.5553846153846156e-06, "loss": 0.0325, "step": 16975 }, { "epoch": 8.03, "learning_rate": 1.5425641025641028e-06, "loss": 0.032, "step": 17000 }, { "epoch": 8.03, "learning_rate": 1.52974358974359e-06, "loss": 0.0307, "step": 17025 }, { "epoch": 8.04, "learning_rate": 1.516923076923077e-06, "loss": 0.0287, "step": 17050 }, { "epoch": 8.04, "learning_rate": 1.5041025641025642e-06, "loss": 0.03, "step": 17075 }, { "epoch": 8.04, "learning_rate": 1.4912820512820514e-06, "loss": 0.0241, "step": 17100 }, { "epoch": 8.04, "learning_rate": 1.4784615384615387e-06, "loss": 0.0246, "step": 17125 }, { "epoch": 8.04, "learning_rate": 1.465641025641026e-06, "loss": 0.0289, "step": 17150 }, { "epoch": 8.04, "learning_rate": 1.4528205128205132e-06, "loss": 0.0295, "step": 17175 }, { "epoch": 8.04, "learning_rate": 1.44e-06, "loss": 0.0267, "step": 17200 }, { "epoch": 8.04, "learning_rate": 1.4271794871794872e-06, "loss": 0.0174, "step": 17225 }, { "epoch": 8.05, "learning_rate": 1.4143589743589745e-06, "loss": 0.0199, "step": 17250 }, { "epoch": 8.05, "learning_rate": 1.4015384615384617e-06, "loss": 0.0094, "step": 17275 }, { "epoch": 8.05, "learning_rate": 1.388717948717949e-06, "loss": 0.0106, "step": 17300 }, { "epoch": 8.05, "learning_rate": 1.3758974358974362e-06, "loss": 0.0132, "step": 17325 }, { "epoch": 8.05, "learning_rate": 1.363076923076923e-06, "loss": 0.0122, "step": 17350 }, { "epoch": 8.05, "learning_rate": 1.3502564102564103e-06, "loss": 0.0085, "step": 17375 }, { "epoch": 8.05, "learning_rate": 1.3374358974358975e-06, "loss": 0.0093, "step": 17400 }, { "epoch": 8.05, "learning_rate": 1.3246153846153848e-06, "loss": 0.014, "step": 17425 }, { "epoch": 8.06, "learning_rate": 1.311794871794872e-06, "loss": 0.0102, "step": 17450 }, { "epoch": 8.06, "learning_rate": 1.2989743589743593e-06, "loss": 0.0142, "step": 17475 }, { "epoch": 8.06, "learning_rate": 1.286153846153846e-06, "loss": 0.0136, "step": 17500 }, { "epoch": 8.06, "learning_rate": 1.2733333333333334e-06, "loss": 0.0135, "step": 17525 }, { "epoch": 8.06, "learning_rate": 1.2605128205128206e-06, "loss": 0.015, "step": 17550 }, { "epoch": 8.06, "learning_rate": 1.2476923076923078e-06, "loss": 0.011, "step": 17575 }, { "epoch": 8.06, "learning_rate": 1.234871794871795e-06, "loss": 0.0108, "step": 17600 }, { "epoch": 8.06, "learning_rate": 1.2220512820512821e-06, "loss": 0.0119, "step": 17625 }, { "epoch": 8.07, "learning_rate": 1.2097435897435898e-06, "loss": 0.0186, "step": 17650 }, { "epoch": 8.07, "learning_rate": 1.196923076923077e-06, "loss": 0.0225, "step": 17675 }, { "epoch": 8.07, "learning_rate": 1.184102564102564e-06, "loss": 0.0251, "step": 17700 }, { "epoch": 8.07, "learning_rate": 1.1712820512820513e-06, "loss": 0.0264, "step": 17725 }, { "epoch": 8.07, "learning_rate": 1.1584615384615385e-06, "loss": 0.0265, "step": 17750 }, { "epoch": 8.07, "learning_rate": 1.1456410256410256e-06, "loss": 0.0254, "step": 17775 }, { "epoch": 8.07, "learning_rate": 1.1328205128205128e-06, "loss": 0.0202, "step": 17800 }, { "epoch": 8.07, "learning_rate": 1.12e-06, "loss": 0.0133, "step": 17825 }, { "epoch": 8.08, "learning_rate": 1.1071794871794873e-06, "loss": 0.0149, "step": 17850 }, { "epoch": 8.08, "learning_rate": 1.0943589743589744e-06, "loss": 0.016, "step": 17875 }, { "epoch": 8.08, "learning_rate": 1.0815384615384616e-06, "loss": 0.0198, "step": 17900 }, { "epoch": 8.08, "learning_rate": 1.0687179487179489e-06, "loss": 0.0119, "step": 17925 }, { "epoch": 8.08, "learning_rate": 1.0558974358974359e-06, "loss": 0.0092, "step": 17950 }, { "epoch": 8.08, "learning_rate": 1.0430769230769231e-06, "loss": 0.01, "step": 17975 }, { "epoch": 8.08, "learning_rate": 1.0302564102564104e-06, "loss": 0.0082, "step": 18000 }, { "epoch": 8.08, "eval_loss": 0.2172992080450058, "eval_runtime": 6735.6172, "eval_samples_per_second": 2.426, "eval_steps_per_second": 0.303, "eval_wer": 8.994092971492167, "step": 18000 }, { "epoch": 8.08, "learning_rate": 1.0174358974358974e-06, "loss": 0.0069, "step": 18025 }, { "epoch": 8.09, "learning_rate": 1.0046153846153847e-06, "loss": 0.0064, "step": 18050 }, { "epoch": 8.09, "learning_rate": 9.91794871794872e-07, "loss": 0.008, "step": 18075 }, { "epoch": 8.09, "learning_rate": 9.78974358974359e-07, "loss": 0.0123, "step": 18100 }, { "epoch": 8.09, "learning_rate": 9.661538461538462e-07, "loss": 0.0224, "step": 18125 }, { "epoch": 8.09, "learning_rate": 9.533333333333335e-07, "loss": 0.02, "step": 18150 }, { "epoch": 8.09, "learning_rate": 9.405128205128206e-07, "loss": 0.0193, "step": 18175 }, { "epoch": 8.09, "learning_rate": 9.276923076923077e-07, "loss": 0.0168, "step": 18200 }, { "epoch": 8.09, "learning_rate": 9.14871794871795e-07, "loss": 0.0147, "step": 18225 }, { "epoch": 8.1, "learning_rate": 9.020512820512821e-07, "loss": 0.0143, "step": 18250 }, { "epoch": 8.1, "learning_rate": 8.892307692307693e-07, "loss": 0.0186, "step": 18275 }, { "epoch": 8.1, "learning_rate": 8.764102564102565e-07, "loss": 0.019, "step": 18300 }, { "epoch": 8.1, "learning_rate": 8.635897435897437e-07, "loss": 0.0145, "step": 18325 }, { "epoch": 8.1, "learning_rate": 8.507692307692308e-07, "loss": 0.0182, "step": 18350 }, { "epoch": 8.1, "learning_rate": 8.37948717948718e-07, "loss": 0.0168, "step": 18375 }, { "epoch": 9.0, "learning_rate": 8.251282051282052e-07, "loss": 0.051, "step": 18400 }, { "epoch": 9.0, "learning_rate": 8.123076923076923e-07, "loss": 0.1094, "step": 18425 }, { "epoch": 9.0, "learning_rate": 7.994871794871796e-07, "loss": 0.0742, "step": 18450 }, { "epoch": 9.0, "learning_rate": 7.866666666666667e-07, "loss": 0.1018, "step": 18475 }, { "epoch": 9.01, "learning_rate": 7.738461538461539e-07, "loss": 0.1079, "step": 18500 }, { "epoch": 9.01, "learning_rate": 7.610256410256411e-07, "loss": 0.1147, "step": 18525 }, { "epoch": 9.01, "learning_rate": 7.482051282051283e-07, "loss": 0.0793, "step": 18550 }, { "epoch": 9.01, "learning_rate": 7.353846153846154e-07, "loss": 0.0913, "step": 18575 }, { "epoch": 9.01, "learning_rate": 7.225641025641026e-07, "loss": 0.0753, "step": 18600 }, { "epoch": 9.01, "learning_rate": 7.097435897435898e-07, "loss": 0.0841, "step": 18625 }, { "epoch": 9.01, "learning_rate": 6.969230769230769e-07, "loss": 0.096, "step": 18650 }, { "epoch": 9.01, "learning_rate": 6.841025641025642e-07, "loss": 0.0819, "step": 18675 }, { "epoch": 9.02, "learning_rate": 6.712820512820513e-07, "loss": 0.1015, "step": 18700 }, { "epoch": 9.02, "learning_rate": 6.584615384615385e-07, "loss": 0.0939, "step": 18725 }, { "epoch": 9.02, "learning_rate": 6.456410256410257e-07, "loss": 0.0884, "step": 18750 }, { "epoch": 9.02, "learning_rate": 6.328205128205129e-07, "loss": 0.0831, "step": 18775 }, { "epoch": 9.02, "learning_rate": 6.200000000000001e-07, "loss": 0.087, "step": 18800 }, { "epoch": 9.02, "learning_rate": 6.071794871794872e-07, "loss": 0.0908, "step": 18825 }, { "epoch": 9.02, "learning_rate": 5.943589743589744e-07, "loss": 0.0779, "step": 18850 }, { "epoch": 9.02, "learning_rate": 5.815384615384616e-07, "loss": 0.0898, "step": 18875 }, { "epoch": 9.03, "learning_rate": 5.687179487179488e-07, "loss": 0.0853, "step": 18900 }, { "epoch": 9.03, "learning_rate": 5.558974358974359e-07, "loss": 0.0921, "step": 18925 }, { "epoch": 9.03, "learning_rate": 5.430769230769232e-07, "loss": 0.0978, "step": 18950 }, { "epoch": 9.03, "learning_rate": 5.302564102564103e-07, "loss": 0.105, "step": 18975 }, { "epoch": 9.03, "learning_rate": 5.174358974358974e-07, "loss": 0.0825, "step": 19000 }, { "epoch": 9.03, "learning_rate": 5.046153846153847e-07, "loss": 0.0807, "step": 19025 }, { "epoch": 9.03, "learning_rate": 4.917948717948718e-07, "loss": 0.0925, "step": 19050 }, { "epoch": 9.03, "learning_rate": 4.78974358974359e-07, "loss": 0.0706, "step": 19075 }, { "epoch": 9.04, "learning_rate": 4.661538461538462e-07, "loss": 0.1002, "step": 19100 }, { "epoch": 9.04, "learning_rate": 4.533333333333334e-07, "loss": 0.1294, "step": 19125 }, { "epoch": 9.04, "learning_rate": 4.4051282051282056e-07, "loss": 0.0939, "step": 19150 }, { "epoch": 9.04, "learning_rate": 4.276923076923077e-07, "loss": 0.0974, "step": 19175 }, { "epoch": 9.04, "learning_rate": 4.1487179487179495e-07, "loss": 0.107, "step": 19200 }, { "epoch": 9.04, "learning_rate": 4.020512820512821e-07, "loss": 0.1042, "step": 19225 }, { "epoch": 9.04, "learning_rate": 3.8923076923076924e-07, "loss": 0.0934, "step": 19250 }, { "epoch": 9.04, "learning_rate": 3.764102564102565e-07, "loss": 0.0795, "step": 19275 }, { "epoch": 9.05, "learning_rate": 3.6358974358974363e-07, "loss": 0.0768, "step": 19300 }, { "epoch": 9.05, "learning_rate": 3.5076923076923077e-07, "loss": 0.0953, "step": 19325 }, { "epoch": 9.05, "learning_rate": 3.37948717948718e-07, "loss": 0.0998, "step": 19350 }, { "epoch": 9.05, "learning_rate": 3.2512820512820516e-07, "loss": 0.1177, "step": 19375 }, { "epoch": 9.05, "learning_rate": 3.1230769230769236e-07, "loss": 0.091, "step": 19400 }, { "epoch": 9.05, "learning_rate": 2.994871794871795e-07, "loss": 0.1083, "step": 19425 }, { "epoch": 9.05, "learning_rate": 2.866666666666667e-07, "loss": 0.0875, "step": 19450 }, { "epoch": 9.05, "learning_rate": 2.738461538461539e-07, "loss": 0.1051, "step": 19475 }, { "epoch": 9.06, "learning_rate": 2.6102564102564103e-07, "loss": 0.0907, "step": 19500 }, { "epoch": 9.06, "learning_rate": 2.482051282051282e-07, "loss": 0.0739, "step": 19525 }, { "epoch": 9.06, "learning_rate": 2.3538461538461542e-07, "loss": 0.1139, "step": 19550 }, { "epoch": 9.06, "learning_rate": 2.2256410256410256e-07, "loss": 0.0981, "step": 19575 }, { "epoch": 9.06, "learning_rate": 2.0974358974358976e-07, "loss": 0.1021, "step": 19600 }, { "epoch": 9.06, "learning_rate": 1.9692307692307695e-07, "loss": 0.1094, "step": 19625 }, { "epoch": 9.06, "learning_rate": 1.841025641025641e-07, "loss": 0.0846, "step": 19650 }, { "epoch": 9.06, "learning_rate": 1.712820512820513e-07, "loss": 0.0868, "step": 19675 }, { "epoch": 9.07, "learning_rate": 1.5846153846153848e-07, "loss": 0.0809, "step": 19700 }, { "epoch": 9.07, "learning_rate": 1.4564102564102565e-07, "loss": 0.0822, "step": 19725 }, { "epoch": 9.07, "learning_rate": 1.3282051282051282e-07, "loss": 0.1077, "step": 19750 }, { "epoch": 9.07, "learning_rate": 1.2000000000000002e-07, "loss": 0.0698, "step": 19775 }, { "epoch": 9.07, "learning_rate": 1.0717948717948718e-07, "loss": 0.0913, "step": 19800 }, { "epoch": 9.07, "learning_rate": 9.435897435897438e-08, "loss": 0.0927, "step": 19825 }, { "epoch": 9.07, "learning_rate": 8.153846153846155e-08, "loss": 0.0801, "step": 19850 }, { "epoch": 9.07, "learning_rate": 6.923076923076924e-08, "loss": 0.1086, "step": 19875 }, { "epoch": 9.08, "learning_rate": 5.641025641025642e-08, "loss": 0.104, "step": 19900 }, { "epoch": 9.08, "learning_rate": 4.358974358974359e-08, "loss": 0.0943, "step": 19925 }, { "epoch": 9.08, "learning_rate": 3.076923076923077e-08, "loss": 0.087, "step": 19950 }, { "epoch": 9.08, "learning_rate": 1.794871794871795e-08, "loss": 0.0851, "step": 19975 }, { "epoch": 9.08, "learning_rate": 5.128205128205129e-09, "loss": 0.0994, "step": 20000 }, { "epoch": 9.08, "eval_loss": 0.19801904261112213, "eval_runtime": 6631.1565, "eval_samples_per_second": 2.464, "eval_steps_per_second": 0.308, "eval_wer": 8.569471791798646, "step": 20000 }, { "epoch": 9.08, "step": 20000, "total_flos": 9.232654995357696e+19, "train_loss": 0.10471280402094126, "train_runtime": 140748.4205, "train_samples_per_second": 2.274, "train_steps_per_second": 0.142 } ], "max_steps": 20000, "num_train_epochs": 9223372036854775807, "total_flos": 9.232654995357696e+19, "trial_name": null, "trial_params": null }