{ "best_metric": null, "best_model_checkpoint": null, "epoch": 40.0, "global_step": 483200, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.01, "learning_rate": 8.900000000000001e-07, "loss": 20.9197, "step": 100 }, { "epoch": 0.02, "learning_rate": 1.8900000000000001e-06, "loss": 5.3584, "step": 200 }, { "epoch": 0.02, "learning_rate": 2.89e-06, "loss": 3.9428, "step": 300 }, { "epoch": 0.03, "learning_rate": 3.89e-06, "loss": 3.4064, "step": 400 }, { "epoch": 0.04, "learning_rate": 4.890000000000001e-06, "loss": 3.3367, "step": 500 }, { "epoch": 0.04, "eval_loss": 3.416260242462158, "eval_runtime": 4566.4565, "eval_samples_per_second": 16.554, "eval_steps_per_second": 1.38, "eval_wer": 0.9938169325893345, "step": 500 }, { "epoch": 0.05, "learning_rate": 5.89e-06, "loss": 3.2118, "step": 600 }, { "epoch": 0.06, "learning_rate": 6.89e-06, "loss": 3.0688, "step": 700 }, { "epoch": 0.07, "learning_rate": 7.89e-06, "loss": 3.0779, "step": 800 }, { "epoch": 0.07, "learning_rate": 8.890000000000001e-06, "loss": 2.9562, "step": 900 }, { "epoch": 0.08, "learning_rate": 9.89e-06, "loss": 2.9738, "step": 1000 }, { "epoch": 0.08, "eval_loss": 2.9496805667877197, "eval_runtime": 4320.8906, "eval_samples_per_second": 17.495, "eval_steps_per_second": 1.458, "eval_wer": 0.9949830846414083, "step": 1000 }, { "epoch": 0.09, "learning_rate": 1.089e-05, "loss": 2.9354, "step": 1100 }, { "epoch": 0.1, "learning_rate": 1.1890000000000001e-05, "loss": 2.8981, "step": 1200 }, { "epoch": 0.11, "learning_rate": 1.289e-05, "loss": 2.8839, "step": 1300 }, { "epoch": 0.12, "learning_rate": 1.389e-05, "loss": 2.821, "step": 1400 }, { "epoch": 0.12, "learning_rate": 1.4890000000000001e-05, "loss": 2.8161, "step": 1500 }, { "epoch": 0.12, "eval_loss": 2.75079083442688, "eval_runtime": 4308.8364, "eval_samples_per_second": 17.544, "eval_steps_per_second": 1.462, "eval_wer": 1.0139076709112755, "step": 1500 }, { "epoch": 0.13, "learning_rate": 1.5890000000000002e-05, "loss": 2.8691, "step": 1600 }, { "epoch": 0.14, "learning_rate": 1.6890000000000003e-05, "loss": 2.8012, "step": 1700 }, { "epoch": 0.15, "learning_rate": 1.789e-05, "loss": 2.6498, "step": 1800 }, { "epoch": 0.16, "learning_rate": 1.889e-05, "loss": 2.4877, "step": 1900 }, { "epoch": 0.17, "learning_rate": 1.989e-05, "loss": 1.6716, "step": 2000 }, { "epoch": 0.17, "eval_loss": 1.2780412435531616, "eval_runtime": 4286.9494, "eval_samples_per_second": 17.634, "eval_steps_per_second": 1.47, "eval_wer": 0.951198536617636, "step": 2000 }, { "epoch": 0.17, "learning_rate": 1.9996300914380717e-05, "loss": 1.035, "step": 2100 }, { "epoch": 0.18, "learning_rate": 1.999214463840399e-05, "loss": 0.8341, "step": 2200 }, { "epoch": 0.19, "learning_rate": 1.9987988362427268e-05, "loss": 0.6403, "step": 2300 }, { "epoch": 0.2, "learning_rate": 1.998383208645054e-05, "loss": 0.5873, "step": 2400 }, { "epoch": 0.21, "learning_rate": 1.997967581047382e-05, "loss": 0.4444, "step": 2500 }, { "epoch": 0.21, "eval_loss": 0.2886000871658325, "eval_runtime": 4276.3997, "eval_samples_per_second": 17.677, "eval_steps_per_second": 1.473, "eval_wer": 0.34336562559134526, "step": 2500 }, { "epoch": 0.22, "learning_rate": 1.9975519534497093e-05, "loss": 0.5106, "step": 2600 }, { "epoch": 0.22, "learning_rate": 1.9971363258520366e-05, "loss": 0.483, "step": 2700 }, { "epoch": 0.23, "learning_rate": 1.9967206982543644e-05, "loss": 0.5262, "step": 2800 }, { "epoch": 0.24, "learning_rate": 1.9963050706566918e-05, "loss": 0.47, "step": 2900 }, { "epoch": 0.25, "learning_rate": 1.995889443059019e-05, "loss": 0.4584, "step": 3000 }, { "epoch": 0.25, "eval_loss": 0.2338569015264511, "eval_runtime": 4254.6189, "eval_samples_per_second": 17.768, "eval_steps_per_second": 1.481, "eval_wer": 0.25199345847160237, "step": 3000 }, { "epoch": 0.26, "learning_rate": 1.995473815461347e-05, "loss": 0.4232, "step": 3100 }, { "epoch": 0.26, "learning_rate": 1.9950581878636742e-05, "loss": 0.4209, "step": 3200 }, { "epoch": 0.27, "learning_rate": 1.9946425602660016e-05, "loss": 0.3426, "step": 3300 }, { "epoch": 0.28, "learning_rate": 1.9942269326683293e-05, "loss": 0.3656, "step": 3400 }, { "epoch": 0.29, "learning_rate": 1.993811305070657e-05, "loss": 0.3229, "step": 3500 }, { "epoch": 0.29, "eval_loss": 0.20768892765045166, "eval_runtime": 4265.7399, "eval_samples_per_second": 17.721, "eval_steps_per_second": 1.477, "eval_wer": 0.20467045435314715, "step": 3500 }, { "epoch": 0.3, "learning_rate": 1.9933956774729844e-05, "loss": 0.3463, "step": 3600 }, { "epoch": 0.31, "learning_rate": 1.9929800498753118e-05, "loss": 0.3308, "step": 3700 }, { "epoch": 0.31, "learning_rate": 1.9925644222776395e-05, "loss": 0.2989, "step": 3800 }, { "epoch": 0.32, "learning_rate": 1.992148794679967e-05, "loss": 0.2796, "step": 3900 }, { "epoch": 0.33, "learning_rate": 1.9917331670822943e-05, "loss": 0.3201, "step": 4000 }, { "epoch": 0.33, "eval_loss": 0.17527684569358826, "eval_runtime": 4256.9548, "eval_samples_per_second": 17.758, "eval_steps_per_second": 1.48, "eval_wer": 0.16469205431991643, "step": 4000 }, { "epoch": 0.34, "learning_rate": 1.991317539484622e-05, "loss": 0.3063, "step": 4100 }, { "epoch": 0.35, "learning_rate": 1.9909019118869494e-05, "loss": 0.2988, "step": 4200 }, { "epoch": 0.36, "learning_rate": 1.9904862842892768e-05, "loss": 0.2932, "step": 4300 }, { "epoch": 0.36, "learning_rate": 1.9900706566916045e-05, "loss": 0.2784, "step": 4400 }, { "epoch": 0.37, "learning_rate": 1.989655029093932e-05, "loss": 0.3029, "step": 4500 }, { "epoch": 0.37, "eval_loss": 0.20827600359916687, "eval_runtime": 4262.1717, "eval_samples_per_second": 17.736, "eval_steps_per_second": 1.478, "eval_wer": 0.19392431703951224, "step": 4500 }, { "epoch": 0.38, "learning_rate": 1.9892394014962596e-05, "loss": 0.2702, "step": 4600 }, { "epoch": 0.39, "learning_rate": 1.988823773898587e-05, "loss": 0.2938, "step": 4700 }, { "epoch": 0.4, "learning_rate": 1.9884081463009147e-05, "loss": 0.2672, "step": 4800 }, { "epoch": 0.41, "learning_rate": 1.987992518703242e-05, "loss": 0.2651, "step": 4900 }, { "epoch": 0.41, "learning_rate": 1.9875768911055695e-05, "loss": 0.2539, "step": 5000 }, { "epoch": 0.41, "eval_loss": 0.1839469075202942, "eval_runtime": 4257.558, "eval_samples_per_second": 17.755, "eval_steps_per_second": 1.48, "eval_wer": 0.13889363247133465, "step": 5000 }, { "epoch": 0.42, "learning_rate": 1.9871612635078972e-05, "loss": 0.2798, "step": 5100 }, { "epoch": 0.43, "learning_rate": 1.9867456359102246e-05, "loss": 0.2615, "step": 5200 }, { "epoch": 0.44, "learning_rate": 1.986330008312552e-05, "loss": 0.2498, "step": 5300 }, { "epoch": 0.45, "learning_rate": 1.9859143807148794e-05, "loss": 0.2435, "step": 5400 }, { "epoch": 0.46, "learning_rate": 1.985498753117207e-05, "loss": 0.2613, "step": 5500 }, { "epoch": 0.46, "eval_loss": 0.1560567021369934, "eval_runtime": 4268.4151, "eval_samples_per_second": 17.71, "eval_steps_per_second": 1.476, "eval_wer": 0.1327859495600776, "step": 5500 }, { "epoch": 0.46, "learning_rate": 1.9850831255195348e-05, "loss": 0.2448, "step": 5600 }, { "epoch": 0.47, "learning_rate": 1.9846674979218622e-05, "loss": 0.215, "step": 5700 }, { "epoch": 0.48, "learning_rate": 1.98425187032419e-05, "loss": 0.2359, "step": 5800 }, { "epoch": 0.49, "learning_rate": 1.9838362427265173e-05, "loss": 0.2321, "step": 5900 }, { "epoch": 0.5, "learning_rate": 1.9834206151288447e-05, "loss": 0.2288, "step": 6000 }, { "epoch": 0.5, "eval_loss": 0.15522293746471405, "eval_runtime": 4281.7155, "eval_samples_per_second": 17.655, "eval_steps_per_second": 1.471, "eval_wer": 0.14036132252104228, "step": 6000 }, { "epoch": 0.5, "learning_rate": 1.9830049875311724e-05, "loss": 0.2198, "step": 6100 }, { "epoch": 0.51, "learning_rate": 1.9825893599334998e-05, "loss": 0.2156, "step": 6200 }, { "epoch": 0.52, "learning_rate": 1.982173732335827e-05, "loss": 0.1949, "step": 6300 }, { "epoch": 0.53, "learning_rate": 1.981758104738155e-05, "loss": 0.2232, "step": 6400 }, { "epoch": 0.54, "learning_rate": 1.9813424771404823e-05, "loss": 0.2261, "step": 6500 }, { "epoch": 0.54, "eval_loss": 0.1471119225025177, "eval_runtime": 4282.3282, "eval_samples_per_second": 17.653, "eval_steps_per_second": 1.471, "eval_wer": 0.11719366585589869, "step": 6500 }, { "epoch": 0.55, "learning_rate": 1.9809268495428096e-05, "loss": 0.2148, "step": 6600 }, { "epoch": 0.55, "learning_rate": 1.9805112219451374e-05, "loss": 0.2234, "step": 6700 }, { "epoch": 0.56, "learning_rate": 1.9800955943474647e-05, "loss": 0.2332, "step": 6800 }, { "epoch": 0.57, "learning_rate": 1.9796799667497925e-05, "loss": 0.2034, "step": 6900 }, { "epoch": 0.58, "learning_rate": 1.97926433915212e-05, "loss": 0.1993, "step": 7000 }, { "epoch": 0.58, "eval_loss": 0.1662781983613968, "eval_runtime": 4272.2006, "eval_samples_per_second": 17.695, "eval_steps_per_second": 1.475, "eval_wer": 0.13439056247605774, "step": 7000 }, { "epoch": 0.59, "learning_rate": 1.9788487115544476e-05, "loss": 0.2135, "step": 7100 }, { "epoch": 0.6, "learning_rate": 1.978433083956775e-05, "loss": 0.2191, "step": 7200 }, { "epoch": 0.6, "learning_rate": 1.9780174563591023e-05, "loss": 0.2141, "step": 7300 }, { "epoch": 0.61, "learning_rate": 1.97760182876143e-05, "loss": 0.2127, "step": 7400 }, { "epoch": 0.62, "learning_rate": 1.9771862011637574e-05, "loss": 0.2082, "step": 7500 }, { "epoch": 0.62, "eval_loss": 0.14242374897003174, "eval_runtime": 4315.1545, "eval_samples_per_second": 17.518, "eval_steps_per_second": 1.46, "eval_wer": 0.10860906367836357, "step": 7500 }, { "epoch": 0.63, "learning_rate": 1.9767705735660848e-05, "loss": 0.2065, "step": 7600 }, { "epoch": 0.64, "learning_rate": 1.9763549459684125e-05, "loss": 0.2035, "step": 7700 }, { "epoch": 0.65, "learning_rate": 1.97593931837074e-05, "loss": 0.2058, "step": 7800 }, { "epoch": 0.65, "learning_rate": 1.9755236907730676e-05, "loss": 0.2109, "step": 7900 }, { "epoch": 0.66, "learning_rate": 1.975108063175395e-05, "loss": 0.1976, "step": 8000 }, { "epoch": 0.66, "eval_loss": 0.1394856870174408, "eval_runtime": 4316.0582, "eval_samples_per_second": 17.515, "eval_steps_per_second": 1.46, "eval_wer": 0.11024290731860413, "step": 8000 }, { "epoch": 0.67, "learning_rate": 1.9746924355777227e-05, "loss": 0.1988, "step": 8100 }, { "epoch": 0.68, "learning_rate": 1.97427680798005e-05, "loss": 0.2041, "step": 8200 }, { "epoch": 0.69, "learning_rate": 1.9738611803823775e-05, "loss": 0.1938, "step": 8300 }, { "epoch": 0.7, "learning_rate": 1.9734455527847052e-05, "loss": 0.2035, "step": 8400 }, { "epoch": 0.7, "learning_rate": 1.9730299251870326e-05, "loss": 0.1854, "step": 8500 }, { "epoch": 0.7, "eval_loss": 0.13305392861366272, "eval_runtime": 4290.88, "eval_samples_per_second": 17.618, "eval_steps_per_second": 1.468, "eval_wer": 0.11837366404051686, "step": 8500 }, { "epoch": 0.71, "learning_rate": 1.97261429758936e-05, "loss": 0.2031, "step": 8600 }, { "epoch": 0.72, "learning_rate": 1.9721986699916874e-05, "loss": 0.1914, "step": 8700 }, { "epoch": 0.73, "learning_rate": 1.971783042394015e-05, "loss": 0.1934, "step": 8800 }, { "epoch": 0.74, "learning_rate": 1.9713674147963425e-05, "loss": 0.1913, "step": 8900 }, { "epoch": 0.75, "learning_rate": 1.9709517871986702e-05, "loss": 0.1855, "step": 9000 }, { "epoch": 0.75, "eval_loss": 0.13057881593704224, "eval_runtime": 4328.6297, "eval_samples_per_second": 17.464, "eval_steps_per_second": 1.455, "eval_wer": 0.10493368471740813, "step": 9000 }, { "epoch": 0.75, "learning_rate": 1.970536159600998e-05, "loss": 0.1973, "step": 9100 }, { "epoch": 0.76, "learning_rate": 1.9701205320033253e-05, "loss": 0.1888, "step": 9200 }, { "epoch": 0.77, "learning_rate": 1.9697049044056527e-05, "loss": 0.1935, "step": 9300 }, { "epoch": 0.78, "learning_rate": 1.9692892768079804e-05, "loss": 0.185, "step": 9400 }, { "epoch": 0.79, "learning_rate": 1.9688736492103078e-05, "loss": 0.1898, "step": 9500 }, { "epoch": 0.79, "eval_loss": 0.1562846601009369, "eval_runtime": 4309.1194, "eval_samples_per_second": 17.543, "eval_steps_per_second": 1.462, "eval_wer": 0.10528291494936162, "step": 9500 }, { "epoch": 0.79, "learning_rate": 1.968458021612635e-05, "loss": 0.1864, "step": 9600 }, { "epoch": 0.8, "learning_rate": 1.9680423940149625e-05, "loss": 0.1856, "step": 9700 }, { "epoch": 0.81, "learning_rate": 1.9676267664172903e-05, "loss": 0.195, "step": 9800 }, { "epoch": 0.82, "learning_rate": 1.9672111388196177e-05, "loss": 0.1777, "step": 9900 }, { "epoch": 0.83, "learning_rate": 1.9667955112219454e-05, "loss": 0.1874, "step": 10000 }, { "epoch": 0.83, "eval_loss": 0.13609051704406738, "eval_runtime": 4306.6658, "eval_samples_per_second": 17.553, "eval_steps_per_second": 1.463, "eval_wer": 0.101149075155269, "step": 10000 }, { "epoch": 0.84, "learning_rate": 1.9663798836242728e-05, "loss": 0.1704, "step": 10100 }, { "epoch": 0.84, "learning_rate": 1.9659642560266005e-05, "loss": 0.1745, "step": 10200 }, { "epoch": 0.85, "learning_rate": 1.965548628428928e-05, "loss": 0.1787, "step": 10300 }, { "epoch": 0.86, "learning_rate": 1.9651330008312556e-05, "loss": 0.1765, "step": 10400 }, { "epoch": 0.87, "learning_rate": 1.964717373233583e-05, "loss": 0.1693, "step": 10500 }, { "epoch": 0.87, "eval_loss": 0.131689190864563, "eval_runtime": 4272.555, "eval_samples_per_second": 17.693, "eval_steps_per_second": 1.475, "eval_wer": 0.09750138845940237, "step": 10500 }, { "epoch": 0.88, "learning_rate": 1.9643017456359103e-05, "loss": 0.1794, "step": 10600 }, { "epoch": 0.89, "learning_rate": 1.9638861180382377e-05, "loss": 0.1787, "step": 10700 }, { "epoch": 0.89, "learning_rate": 1.9634704904405654e-05, "loss": 0.1674, "step": 10800 }, { "epoch": 0.9, "learning_rate": 1.9630548628428928e-05, "loss": 0.189, "step": 10900 }, { "epoch": 0.91, "learning_rate": 1.9626392352452202e-05, "loss": 0.1731, "step": 11000 }, { "epoch": 0.91, "eval_loss": 0.155191108584404, "eval_runtime": 4261.4303, "eval_samples_per_second": 17.739, "eval_steps_per_second": 1.478, "eval_wer": 0.11192444319316432, "step": 11000 }, { "epoch": 0.92, "learning_rate": 1.962223607647548e-05, "loss": 0.176, "step": 11100 }, { "epoch": 0.93, "learning_rate": 1.9618079800498757e-05, "loss": 0.1704, "step": 11200 }, { "epoch": 0.94, "learning_rate": 1.961392352452203e-05, "loss": 0.1704, "step": 11300 }, { "epoch": 0.94, "learning_rate": 1.9609767248545308e-05, "loss": 0.1753, "step": 11400 }, { "epoch": 0.95, "learning_rate": 1.960561097256858e-05, "loss": 0.1739, "step": 11500 }, { "epoch": 0.95, "eval_loss": 0.12559062242507935, "eval_runtime": 4256.5381, "eval_samples_per_second": 17.76, "eval_steps_per_second": 1.48, "eval_wer": 0.08628140572091428, "step": 11500 }, { "epoch": 0.96, "learning_rate": 1.9601454696591855e-05, "loss": 0.1689, "step": 11600 }, { "epoch": 0.97, "learning_rate": 1.959729842061513e-05, "loss": 0.1699, "step": 11700 }, { "epoch": 0.98, "learning_rate": 1.9593142144638406e-05, "loss": 0.1624, "step": 11800 }, { "epoch": 0.99, "learning_rate": 1.958898586866168e-05, "loss": 0.1653, "step": 11900 }, { "epoch": 0.99, "learning_rate": 1.9584829592684954e-05, "loss": 0.1621, "step": 12000 }, { "epoch": 0.99, "eval_loss": 0.13594742119312286, "eval_runtime": 4287.0797, "eval_samples_per_second": 17.633, "eval_steps_per_second": 1.47, "eval_wer": 0.08902755534222255, "step": 12000 }, { "epoch": 1.0, "learning_rate": 1.958067331670823e-05, "loss": 0.1652, "step": 12100 }, { "epoch": 1.01, "learning_rate": 1.9576517040731505e-05, "loss": 0.161, "step": 12200 }, { "epoch": 1.02, "learning_rate": 1.9572360764754782e-05, "loss": 0.1677, "step": 12300 }, { "epoch": 1.03, "learning_rate": 1.9568204488778056e-05, "loss": 0.1451, "step": 12400 }, { "epoch": 1.03, "learning_rate": 1.9564048212801333e-05, "loss": 0.1561, "step": 12500 }, { "epoch": 1.03, "eval_loss": 0.12517023086547852, "eval_runtime": 4279.7907, "eval_samples_per_second": 17.663, "eval_steps_per_second": 1.472, "eval_wer": 0.08327371804043378, "step": 12500 }, { "epoch": 1.04, "learning_rate": 1.9559891936824607e-05, "loss": 0.1567, "step": 12600 }, { "epoch": 1.05, "learning_rate": 1.955573566084788e-05, "loss": 0.1562, "step": 12700 }, { "epoch": 1.06, "learning_rate": 1.9551579384871158e-05, "loss": 0.1531, "step": 12800 }, { "epoch": 1.07, "learning_rate": 1.9547423108894432e-05, "loss": 0.136, "step": 12900 }, { "epoch": 1.08, "learning_rate": 1.9543266832917706e-05, "loss": 0.161, "step": 13000 }, { "epoch": 1.08, "eval_loss": 0.12426160275936127, "eval_runtime": 4270.6471, "eval_samples_per_second": 17.701, "eval_steps_per_second": 1.475, "eval_wer": 0.08019218431971643, "step": 13000 }, { "epoch": 1.08, "learning_rate": 1.9539110556940983e-05, "loss": 0.1518, "step": 13100 }, { "epoch": 1.09, "learning_rate": 1.9534954280964257e-05, "loss": 0.1529, "step": 13200 }, { "epoch": 1.1, "learning_rate": 1.9530798004987534e-05, "loss": 0.1455, "step": 13300 }, { "epoch": 1.11, "learning_rate": 1.9526641729010808e-05, "loss": 0.144, "step": 13400 }, { "epoch": 1.12, "learning_rate": 1.9522485453034085e-05, "loss": 0.1512, "step": 13500 }, { "epoch": 1.12, "eval_loss": 0.1287076622247696, "eval_runtime": 4315.5604, "eval_samples_per_second": 17.517, "eval_steps_per_second": 1.46, "eval_wer": 0.08290294938007788, "step": 13500 }, { "epoch": 1.13, "learning_rate": 1.951832917705736e-05, "loss": 0.1551, "step": 13600 }, { "epoch": 1.13, "learning_rate": 1.9514172901080633e-05, "loss": 0.1509, "step": 13700 }, { "epoch": 1.14, "learning_rate": 1.951001662510391e-05, "loss": 0.161, "step": 13800 }, { "epoch": 1.15, "learning_rate": 1.9505860349127184e-05, "loss": 0.1479, "step": 13900 }, { "epoch": 1.16, "learning_rate": 1.9501704073150457e-05, "loss": 0.145, "step": 14000 }, { "epoch": 1.16, "eval_loss": 0.12732306122779846, "eval_runtime": 4292.547, "eval_samples_per_second": 17.611, "eval_steps_per_second": 1.468, "eval_wer": 0.08094756777297266, "step": 14000 }, { "epoch": 1.17, "learning_rate": 1.9497547797173735e-05, "loss": 0.1588, "step": 14100 }, { "epoch": 1.18, "learning_rate": 1.949339152119701e-05, "loss": 0.159, "step": 14200 }, { "epoch": 1.18, "learning_rate": 1.9489235245220282e-05, "loss": 0.1658, "step": 14300 }, { "epoch": 1.19, "learning_rate": 1.948507896924356e-05, "loss": 0.1614, "step": 14400 }, { "epoch": 1.2, "learning_rate": 1.9480922693266833e-05, "loss": 0.1452, "step": 14500 }, { "epoch": 1.2, "eval_loss": 0.1276836395263672, "eval_runtime": 4288.5628, "eval_samples_per_second": 17.627, "eval_steps_per_second": 1.469, "eval_wer": 0.07965218515048439, "step": 14500 }, { "epoch": 1.21, "learning_rate": 1.947676641729011e-05, "loss": 0.1527, "step": 14600 }, { "epoch": 1.22, "learning_rate": 1.9472610141313384e-05, "loss": 0.1453, "step": 14700 }, { "epoch": 1.23, "learning_rate": 1.946845386533666e-05, "loss": 0.148, "step": 14800 }, { "epoch": 1.23, "learning_rate": 1.9464297589359935e-05, "loss": 0.1517, "step": 14900 }, { "epoch": 1.24, "learning_rate": 1.946014131338321e-05, "loss": 0.1506, "step": 15000 }, { "epoch": 1.24, "eval_loss": 0.13583926856517792, "eval_runtime": 4291.925, "eval_samples_per_second": 17.613, "eval_steps_per_second": 1.468, "eval_wer": 0.07947833926409344, "step": 15000 }, { "epoch": 1.25, "learning_rate": 1.9455985037406486e-05, "loss": 0.139, "step": 15100 }, { "epoch": 1.26, "learning_rate": 1.945182876142976e-05, "loss": 0.1292, "step": 15200 }, { "epoch": 1.27, "learning_rate": 1.9447672485453034e-05, "loss": 0.1536, "step": 15300 }, { "epoch": 1.27, "learning_rate": 1.944351620947631e-05, "loss": 0.1425, "step": 15400 }, { "epoch": 1.28, "learning_rate": 1.9439359933499585e-05, "loss": 0.1463, "step": 15500 }, { "epoch": 1.28, "eval_loss": 0.12759171426296234, "eval_runtime": 4295.3849, "eval_samples_per_second": 17.599, "eval_steps_per_second": 1.467, "eval_wer": 0.07744141932089335, "step": 15500 }, { "epoch": 1.29, "learning_rate": 1.9435203657522862e-05, "loss": 0.1526, "step": 15600 }, { "epoch": 1.3, "learning_rate": 1.9431047381546136e-05, "loss": 0.1444, "step": 15700 }, { "epoch": 1.31, "learning_rate": 1.9426891105569413e-05, "loss": 0.1455, "step": 15800 }, { "epoch": 1.32, "learning_rate": 1.9422734829592687e-05, "loss": 0.1376, "step": 15900 }, { "epoch": 1.32, "learning_rate": 1.941857855361596e-05, "loss": 0.1361, "step": 16000 }, { "epoch": 1.32, "eval_loss": 0.12064395844936371, "eval_runtime": 4298.884, "eval_samples_per_second": 17.585, "eval_steps_per_second": 1.465, "eval_wer": 0.07887064789131093, "step": 16000 }, { "epoch": 1.33, "learning_rate": 1.9414422277639238e-05, "loss": 0.1274, "step": 16100 }, { "epoch": 1.34, "learning_rate": 1.9410266001662512e-05, "loss": 0.1521, "step": 16200 }, { "epoch": 1.35, "learning_rate": 1.9406109725685786e-05, "loss": 0.1412, "step": 16300 }, { "epoch": 1.36, "learning_rate": 1.9401953449709063e-05, "loss": 0.1409, "step": 16400 }, { "epoch": 1.37, "learning_rate": 1.9397797173732337e-05, "loss": 0.1405, "step": 16500 }, { "epoch": 1.37, "eval_loss": 0.13032357394695282, "eval_runtime": 4300.715, "eval_samples_per_second": 17.577, "eval_steps_per_second": 1.465, "eval_wer": 0.07912141673628194, "step": 16500 }, { "epoch": 1.37, "learning_rate": 1.939364089775561e-05, "loss": 0.1439, "step": 16600 }, { "epoch": 1.38, "learning_rate": 1.9389484621778888e-05, "loss": 0.1427, "step": 16700 }, { "epoch": 1.39, "learning_rate": 1.9385328345802165e-05, "loss": 0.1261, "step": 16800 }, { "epoch": 1.4, "learning_rate": 1.938117206982544e-05, "loss": 0.1569, "step": 16900 }, { "epoch": 1.41, "learning_rate": 1.9377015793848713e-05, "loss": 0.1439, "step": 17000 }, { "epoch": 1.41, "eval_loss": 0.13192984461784363, "eval_runtime": 4306.5472, "eval_samples_per_second": 17.554, "eval_steps_per_second": 1.463, "eval_wer": 0.08043679932800103, "step": 17000 }, { "epoch": 1.42, "learning_rate": 1.937285951787199e-05, "loss": 0.1434, "step": 17100 }, { "epoch": 1.42, "learning_rate": 1.9368703241895264e-05, "loss": 0.1381, "step": 17200 }, { "epoch": 1.43, "learning_rate": 1.9364546965918538e-05, "loss": 0.1301, "step": 17300 }, { "epoch": 1.44, "learning_rate": 1.9360390689941815e-05, "loss": 0.1509, "step": 17400 }, { "epoch": 1.45, "learning_rate": 1.935623441396509e-05, "loss": 0.1429, "step": 17500 }, { "epoch": 1.45, "eval_loss": 0.11785099655389786, "eval_runtime": 4285.8122, "eval_samples_per_second": 17.638, "eval_steps_per_second": 1.47, "eval_wer": 0.07677218958124679, "step": 17500 }, { "epoch": 1.46, "learning_rate": 1.9352078137988362e-05, "loss": 0.1381, "step": 17600 }, { "epoch": 1.47, "learning_rate": 1.934792186201164e-05, "loss": 0.1389, "step": 17700 }, { "epoch": 1.47, "learning_rate": 1.9343765586034914e-05, "loss": 0.137, "step": 17800 }, { "epoch": 1.48, "learning_rate": 1.933960931005819e-05, "loss": 0.1338, "step": 17900 }, { "epoch": 1.49, "learning_rate": 1.9335453034081465e-05, "loss": 0.1299, "step": 18000 }, { "epoch": 1.49, "eval_loss": 0.13803158700466156, "eval_runtime": 4287.4199, "eval_samples_per_second": 17.632, "eval_steps_per_second": 1.469, "eval_wer": 0.08122910580137568, "step": 18000 }, { "epoch": 1.5, "learning_rate": 1.9331296758104742e-05, "loss": 0.1315, "step": 18100 }, { "epoch": 1.51, "learning_rate": 1.9327140482128016e-05, "loss": 0.1326, "step": 18200 }, { "epoch": 1.51, "learning_rate": 1.932298420615129e-05, "loss": 0.1522, "step": 18300 }, { "epoch": 1.52, "learning_rate": 1.9318827930174567e-05, "loss": 0.1457, "step": 18400 }, { "epoch": 1.53, "learning_rate": 1.931467165419784e-05, "loss": 0.1423, "step": 18500 }, { "epoch": 1.53, "eval_loss": 0.12282726168632507, "eval_runtime": 4298.6251, "eval_samples_per_second": 17.586, "eval_steps_per_second": 1.466, "eval_wer": 0.07682603565225284, "step": 18500 }, { "epoch": 1.54, "learning_rate": 1.9310515378221114e-05, "loss": 0.1349, "step": 18600 }, { "epoch": 1.55, "learning_rate": 1.9306359102244388e-05, "loss": 0.15, "step": 18700 }, { "epoch": 1.56, "learning_rate": 1.9302202826267665e-05, "loss": 0.1338, "step": 18800 }, { "epoch": 1.56, "learning_rate": 1.9298088113050707e-05, "loss": 0.1263, "step": 18900 }, { "epoch": 1.57, "learning_rate": 1.929393183707398e-05, "loss": 0.1432, "step": 19000 }, { "epoch": 1.57, "eval_loss": 0.12211327999830246, "eval_runtime": 4294.8551, "eval_samples_per_second": 17.601, "eval_steps_per_second": 1.467, "eval_wer": 0.07742141935166254, "step": 19000 }, { "epoch": 1.58, "learning_rate": 1.9289775561097258e-05, "loss": 0.1346, "step": 19100 }, { "epoch": 1.59, "learning_rate": 1.9285619285120535e-05, "loss": 0.1314, "step": 19200 }, { "epoch": 1.6, "learning_rate": 1.928146300914381e-05, "loss": 0.136, "step": 19300 }, { "epoch": 1.61, "learning_rate": 1.9277306733167086e-05, "loss": 0.1273, "step": 19400 }, { "epoch": 1.61, "learning_rate": 1.927315045719036e-05, "loss": 0.1261, "step": 19500 }, { "epoch": 1.61, "eval_loss": 0.13089631497859955, "eval_runtime": 4299.2737, "eval_samples_per_second": 17.583, "eval_steps_per_second": 1.465, "eval_wer": 0.07814141824397193, "step": 19500 }, { "epoch": 1.62, "learning_rate": 1.9268994181213634e-05, "loss": 0.1397, "step": 19600 }, { "epoch": 1.63, "learning_rate": 1.9264837905236908e-05, "loss": 0.138, "step": 19700 }, { "epoch": 1.64, "learning_rate": 1.9260681629260185e-05, "loss": 0.1275, "step": 19800 }, { "epoch": 1.65, "learning_rate": 1.925652535328346e-05, "loss": 0.1363, "step": 19900 }, { "epoch": 1.66, "learning_rate": 1.9252369077306733e-05, "loss": 0.1338, "step": 20000 }, { "epoch": 1.66, "eval_loss": 0.1331881582736969, "eval_runtime": 4303.9481, "eval_samples_per_second": 17.564, "eval_steps_per_second": 1.464, "eval_wer": 0.07612295981083106, "step": 20000 }, { "epoch": 1.66, "learning_rate": 1.924821280133001e-05, "loss": 0.1368, "step": 20100 }, { "epoch": 1.67, "learning_rate": 1.9244056525353284e-05, "loss": 0.1294, "step": 20200 }, { "epoch": 1.68, "learning_rate": 1.923990024937656e-05, "loss": 0.1304, "step": 20300 }, { "epoch": 1.69, "learning_rate": 1.9235743973399835e-05, "loss": 0.1336, "step": 20400 }, { "epoch": 1.7, "learning_rate": 1.9231587697423112e-05, "loss": 0.1345, "step": 20500 }, { "epoch": 1.7, "eval_loss": 0.11977848410606384, "eval_runtime": 4299.8051, "eval_samples_per_second": 17.581, "eval_steps_per_second": 1.465, "eval_wer": 0.0770152661303598, "step": 20500 }, { "epoch": 1.71, "learning_rate": 1.9227431421446386e-05, "loss": 0.1332, "step": 20600 }, { "epoch": 1.71, "learning_rate": 1.922327514546966e-05, "loss": 0.124, "step": 20700 }, { "epoch": 1.72, "learning_rate": 1.9219118869492937e-05, "loss": 0.135, "step": 20800 }, { "epoch": 1.73, "learning_rate": 1.921496259351621e-05, "loss": 0.1338, "step": 20900 }, { "epoch": 1.74, "learning_rate": 1.9210806317539484e-05, "loss": 0.1312, "step": 21000 }, { "epoch": 1.74, "eval_loss": 0.12255703657865524, "eval_runtime": 4290.5754, "eval_samples_per_second": 17.619, "eval_steps_per_second": 1.468, "eval_wer": 0.07461680828183341, "step": 21000 }, { "epoch": 1.75, "learning_rate": 1.920665004156276e-05, "loss": 0.1338, "step": 21100 }, { "epoch": 1.75, "learning_rate": 1.9202493765586036e-05, "loss": 0.1369, "step": 21200 }, { "epoch": 1.76, "learning_rate": 1.9198337489609313e-05, "loss": 0.1272, "step": 21300 }, { "epoch": 1.77, "learning_rate": 1.9194181213632587e-05, "loss": 0.1361, "step": 21400 }, { "epoch": 1.78, "learning_rate": 1.9190024937655864e-05, "loss": 0.1269, "step": 21500 }, { "epoch": 1.78, "eval_loss": 0.13120083510875702, "eval_runtime": 4289.6917, "eval_samples_per_second": 17.622, "eval_steps_per_second": 1.469, "eval_wer": 0.07586757558834525, "step": 21500 }, { "epoch": 1.79, "learning_rate": 1.9185868661679138e-05, "loss": 0.1275, "step": 21600 }, { "epoch": 1.8, "learning_rate": 1.918171238570241e-05, "loss": 0.1224, "step": 21700 }, { "epoch": 1.8, "learning_rate": 1.917755610972569e-05, "loss": 0.1349, "step": 21800 }, { "epoch": 1.81, "learning_rate": 1.9173399833748962e-05, "loss": 0.1278, "step": 21900 }, { "epoch": 1.82, "learning_rate": 1.9169243557772236e-05, "loss": 0.1322, "step": 22000 }, { "epoch": 1.82, "eval_loss": 0.12505650520324707, "eval_runtime": 4305.085, "eval_samples_per_second": 17.559, "eval_steps_per_second": 1.463, "eval_wer": 0.07506142298242618, "step": 22000 }, { "epoch": 1.83, "learning_rate": 1.9165087281795513e-05, "loss": 0.1403, "step": 22100 }, { "epoch": 1.84, "learning_rate": 1.9160931005818787e-05, "loss": 0.1244, "step": 22200 }, { "epoch": 1.85, "learning_rate": 1.915677472984206e-05, "loss": 0.1268, "step": 22300 }, { "epoch": 1.85, "learning_rate": 1.915261845386534e-05, "loss": 0.1318, "step": 22400 }, { "epoch": 1.86, "learning_rate": 1.9148462177888612e-05, "loss": 0.1329, "step": 22500 }, { "epoch": 1.86, "eval_loss": 0.11915752291679382, "eval_runtime": 4287.4958, "eval_samples_per_second": 17.632, "eval_steps_per_second": 1.469, "eval_wer": 0.07150758229602723, "step": 22500 }, { "epoch": 1.87, "learning_rate": 1.914430590191189e-05, "loss": 0.1365, "step": 22600 }, { "epoch": 1.88, "learning_rate": 1.9140149625935163e-05, "loss": 0.1339, "step": 22700 }, { "epoch": 1.89, "learning_rate": 1.913599334995844e-05, "loss": 0.1178, "step": 22800 }, { "epoch": 1.9, "learning_rate": 1.9131837073981714e-05, "loss": 0.1263, "step": 22900 }, { "epoch": 1.9, "learning_rate": 1.9127680798004988e-05, "loss": 0.1423, "step": 23000 }, { "epoch": 1.9, "eval_loss": 0.11137987673282623, "eval_runtime": 4310.8831, "eval_samples_per_second": 17.536, "eval_steps_per_second": 1.461, "eval_wer": 0.07183988947709312, "step": 23000 }, { "epoch": 1.91, "learning_rate": 1.9123524522028265e-05, "loss": 0.1275, "step": 23100 }, { "epoch": 1.92, "learning_rate": 1.911936824605154e-05, "loss": 0.126, "step": 23200 }, { "epoch": 1.93, "learning_rate": 1.9115211970074813e-05, "loss": 0.1217, "step": 23300 }, { "epoch": 1.94, "learning_rate": 1.9111097256857858e-05, "loss": 0.1262, "step": 23400 }, { "epoch": 1.95, "learning_rate": 1.9106940980881132e-05, "loss": 0.1223, "step": 23500 }, { "epoch": 1.95, "eval_loss": 0.12768128514289856, "eval_runtime": 4288.4349, "eval_samples_per_second": 17.628, "eval_steps_per_second": 1.469, "eval_wer": 0.07326142575165269, "step": 23500 }, { "epoch": 1.95, "learning_rate": 1.9102784704904406e-05, "loss": 0.1259, "step": 23600 }, { "epoch": 1.96, "learning_rate": 1.9098628428927683e-05, "loss": 0.1258, "step": 23700 }, { "epoch": 1.97, "learning_rate": 1.9094472152950957e-05, "loss": 0.1278, "step": 23800 }, { "epoch": 1.98, "learning_rate": 1.9090315876974234e-05, "loss": 0.1143, "step": 23900 }, { "epoch": 1.99, "learning_rate": 1.9086159600997508e-05, "loss": 0.1278, "step": 24000 }, { "epoch": 1.99, "eval_loss": 0.1287057250738144, "eval_runtime": 4304.0674, "eval_samples_per_second": 17.564, "eval_steps_per_second": 1.464, "eval_wer": 0.07507373065579899, "step": 24000 }, { "epoch": 2.0, "learning_rate": 1.9082003325020785e-05, "loss": 0.1344, "step": 24100 }, { "epoch": 2.0, "learning_rate": 1.907784704904406e-05, "loss": 0.1058, "step": 24200 }, { "epoch": 2.01, "learning_rate": 1.9073690773067333e-05, "loss": 0.1237, "step": 24300 }, { "epoch": 2.02, "learning_rate": 1.906953449709061e-05, "loss": 0.0997, "step": 24400 }, { "epoch": 2.03, "learning_rate": 1.9065378221113884e-05, "loss": 0.1128, "step": 24500 }, { "epoch": 2.03, "eval_loss": 0.1256413608789444, "eval_runtime": 4298.2074, "eval_samples_per_second": 17.588, "eval_steps_per_second": 1.466, "eval_wer": 0.07017527665342053, "step": 24500 }, { "epoch": 2.04, "learning_rate": 1.9061221945137158e-05, "loss": 0.1069, "step": 24600 }, { "epoch": 2.04, "learning_rate": 1.905706566916043e-05, "loss": 0.1266, "step": 24700 }, { "epoch": 2.05, "learning_rate": 1.905290939318371e-05, "loss": 0.1203, "step": 24800 }, { "epoch": 2.06, "learning_rate": 1.9048753117206982e-05, "loss": 0.1323, "step": 24900 }, { "epoch": 2.07, "learning_rate": 1.904459684123026e-05, "loss": 0.112, "step": 25000 }, { "epoch": 2.07, "eval_loss": 0.12580588459968567, "eval_runtime": 4282.766, "eval_samples_per_second": 17.651, "eval_steps_per_second": 1.471, "eval_wer": 0.07114758284987253, "step": 25000 }, { "epoch": 2.08, "learning_rate": 1.9040440565253537e-05, "loss": 0.1168, "step": 25100 }, { "epoch": 2.09, "learning_rate": 1.903628428927681e-05, "loss": 0.1125, "step": 25200 }, { "epoch": 2.09, "learning_rate": 1.9032128013300084e-05, "loss": 0.1355, "step": 25300 }, { "epoch": 2.1, "learning_rate": 1.902797173732336e-05, "loss": 0.1236, "step": 25400 }, { "epoch": 2.11, "learning_rate": 1.9023815461346635e-05, "loss": 0.115, "step": 25500 }, { "epoch": 2.11, "eval_loss": 0.12705326080322266, "eval_runtime": 4294.0288, "eval_samples_per_second": 17.605, "eval_steps_per_second": 1.467, "eval_wer": 0.07175988960016985, "step": 25500 }, { "epoch": 2.12, "learning_rate": 1.901965918536991e-05, "loss": 0.1137, "step": 25600 }, { "epoch": 2.13, "learning_rate": 1.9015502909393183e-05, "loss": 0.1106, "step": 25700 }, { "epoch": 2.14, "learning_rate": 1.901134663341646e-05, "loss": 0.1241, "step": 25800 }, { "epoch": 2.14, "learning_rate": 1.9007190357439734e-05, "loss": 0.1082, "step": 25900 }, { "epoch": 2.15, "learning_rate": 1.900303408146301e-05, "loss": 0.1131, "step": 26000 }, { "epoch": 2.15, "eval_loss": 0.12231649458408356, "eval_runtime": 4273.1327, "eval_samples_per_second": 17.691, "eval_steps_per_second": 1.474, "eval_wer": 0.07167988972324658, "step": 26000 }, { "epoch": 2.16, "learning_rate": 1.8998877805486285e-05, "loss": 0.1182, "step": 26100 }, { "epoch": 2.17, "learning_rate": 1.8994721529509562e-05, "loss": 0.1133, "step": 26200 }, { "epoch": 2.18, "learning_rate": 1.8990565253532836e-05, "loss": 0.1115, "step": 26300 }, { "epoch": 2.19, "learning_rate": 1.8986408977556113e-05, "loss": 0.1128, "step": 26400 }, { "epoch": 2.19, "learning_rate": 1.8982294264339155e-05, "loss": 0.1175, "step": 26500 }, { "epoch": 2.19, "eval_loss": 0.12289807945489883, "eval_runtime": 4287.184, "eval_samples_per_second": 17.633, "eval_steps_per_second": 1.469, "eval_wer": 0.0725460422368581, "step": 26500 }, { "epoch": 2.2, "learning_rate": 1.897813798836243e-05, "loss": 0.1174, "step": 26600 }, { "epoch": 2.21, "learning_rate": 1.8973981712385703e-05, "loss": 0.1084, "step": 26700 }, { "epoch": 2.22, "learning_rate": 1.896982543640898e-05, "loss": 0.1157, "step": 26800 }, { "epoch": 2.23, "learning_rate": 1.8965669160432254e-05, "loss": 0.13, "step": 26900 }, { "epoch": 2.24, "learning_rate": 1.8961512884455528e-05, "loss": 0.1156, "step": 27000 }, { "epoch": 2.24, "eval_loss": 0.12353357672691345, "eval_runtime": 4413.4944, "eval_samples_per_second": 17.128, "eval_steps_per_second": 1.427, "eval_wer": 0.07187988941555475, "step": 27000 }, { "epoch": 2.24, "learning_rate": 1.8957356608478805e-05, "loss": 0.1168, "step": 27100 }, { "epoch": 2.25, "learning_rate": 1.895320033250208e-05, "loss": 0.1143, "step": 27200 }, { "epoch": 2.26, "learning_rate": 1.8949044056525353e-05, "loss": 0.1121, "step": 27300 }, { "epoch": 2.27, "learning_rate": 1.894488778054863e-05, "loss": 0.1114, "step": 27400 }, { "epoch": 2.28, "learning_rate": 1.8940731504571907e-05, "loss": 0.104, "step": 27500 }, { "epoch": 2.28, "eval_loss": 0.1273384988307953, "eval_runtime": 4373.5476, "eval_samples_per_second": 17.285, "eval_steps_per_second": 1.44, "eval_wer": 0.06982143104395223, "step": 27500 }, { "epoch": 2.28, "learning_rate": 1.893657522859518e-05, "loss": 0.1136, "step": 27600 }, { "epoch": 2.29, "learning_rate": 1.8932418952618455e-05, "loss": 0.1172, "step": 27700 }, { "epoch": 2.3, "learning_rate": 1.8928262676641732e-05, "loss": 0.1181, "step": 27800 }, { "epoch": 2.31, "learning_rate": 1.8924106400665006e-05, "loss": 0.1178, "step": 27900 }, { "epoch": 2.32, "learning_rate": 1.891995012468828e-05, "loss": 0.1201, "step": 28000 }, { "epoch": 2.32, "eval_loss": 0.12364833056926727, "eval_runtime": 4323.8851, "eval_samples_per_second": 17.483, "eval_steps_per_second": 1.457, "eval_wer": 0.07097219850431, "step": 28000 }, { "epoch": 2.33, "learning_rate": 1.8915793848711557e-05, "loss": 0.1017, "step": 28100 }, { "epoch": 2.33, "learning_rate": 1.891163757273483e-05, "loss": 0.1161, "step": 28200 }, { "epoch": 2.34, "learning_rate": 1.8907481296758104e-05, "loss": 0.1079, "step": 28300 }, { "epoch": 2.35, "learning_rate": 1.890332502078138e-05, "loss": 0.1105, "step": 28400 }, { "epoch": 2.36, "learning_rate": 1.8899168744804655e-05, "loss": 0.1182, "step": 28500 }, { "epoch": 2.36, "eval_loss": 0.11907240003347397, "eval_runtime": 4340.453, "eval_samples_per_second": 17.416, "eval_steps_per_second": 1.451, "eval_wer": 0.0690568168356664, "step": 28500 }, { "epoch": 2.37, "learning_rate": 1.8895054031587697e-05, "loss": 0.107, "step": 28600 }, { "epoch": 2.38, "learning_rate": 1.8890897755610974e-05, "loss": 0.1176, "step": 28700 }, { "epoch": 2.38, "learning_rate": 1.8886741479634248e-05, "loss": 0.1091, "step": 28800 }, { "epoch": 2.39, "learning_rate": 1.8882585203657525e-05, "loss": 0.1036, "step": 28900 }, { "epoch": 2.4, "learning_rate": 1.88784289276808e-05, "loss": 0.123, "step": 29000 }, { "epoch": 2.4, "eval_loss": 0.12208979576826096, "eval_runtime": 4326.0887, "eval_samples_per_second": 17.474, "eval_steps_per_second": 1.456, "eval_wer": 0.07057527603803686, "step": 29000 }, { "epoch": 2.41, "learning_rate": 1.8874272651704077e-05, "loss": 0.1141, "step": 29100 }, { "epoch": 2.42, "learning_rate": 1.887011637572735e-05, "loss": 0.1123, "step": 29200 }, { "epoch": 2.43, "learning_rate": 1.8865960099750624e-05, "loss": 0.1159, "step": 29300 }, { "epoch": 2.43, "learning_rate": 1.88618038237739e-05, "loss": 0.1291, "step": 29400 }, { "epoch": 2.44, "learning_rate": 1.8857647547797175e-05, "loss": 0.1244, "step": 29500 }, { "epoch": 2.44, "eval_loss": 0.11539793759584427, "eval_runtime": 4339.188, "eval_samples_per_second": 17.421, "eval_steps_per_second": 1.452, "eval_wer": 0.07184142793626472, "step": 29500 }, { "epoch": 2.45, "learning_rate": 1.885349127182045e-05, "loss": 0.1135, "step": 29600 }, { "epoch": 2.46, "learning_rate": 1.8849334995843726e-05, "loss": 0.1041, "step": 29700 }, { "epoch": 2.47, "learning_rate": 1.8845178719867e-05, "loss": 0.1119, "step": 29800 }, { "epoch": 2.48, "learning_rate": 1.8841022443890277e-05, "loss": 0.1162, "step": 29900 }, { "epoch": 2.48, "learning_rate": 1.883686616791355e-05, "loss": 0.1143, "step": 30000 }, { "epoch": 2.48, "eval_loss": 0.10887276381254196, "eval_runtime": 4292.7588, "eval_samples_per_second": 17.61, "eval_steps_per_second": 1.468, "eval_wer": 0.0692752780380338, "step": 30000 }, { "epoch": 2.49, "learning_rate": 1.8832709891936828e-05, "loss": 0.1076, "step": 30100 }, { "epoch": 2.5, "learning_rate": 1.8828553615960102e-05, "loss": 0.1099, "step": 30200 }, { "epoch": 2.51, "learning_rate": 1.8824397339983376e-05, "loss": 0.1163, "step": 30300 }, { "epoch": 2.52, "learning_rate": 1.8820241064006653e-05, "loss": 0.1079, "step": 30400 }, { "epoch": 2.52, "learning_rate": 1.8816084788029927e-05, "loss": 0.1137, "step": 30500 }, { "epoch": 2.52, "eval_loss": 0.13602447509765625, "eval_runtime": 4325.8019, "eval_samples_per_second": 17.475, "eval_steps_per_second": 1.456, "eval_wer": 0.07623372887118635, "step": 30500 }, { "epoch": 2.53, "learning_rate": 1.88119285120532e-05, "loss": 0.1183, "step": 30600 }, { "epoch": 2.54, "learning_rate": 1.8807772236076478e-05, "loss": 0.115, "step": 30700 }, { "epoch": 2.55, "learning_rate": 1.8803615960099752e-05, "loss": 0.1023, "step": 30800 }, { "epoch": 2.56, "learning_rate": 1.8799459684123026e-05, "loss": 0.0994, "step": 30900 }, { "epoch": 2.57, "learning_rate": 1.8795303408146303e-05, "loss": 0.1056, "step": 31000 }, { "epoch": 2.57, "eval_loss": 0.1302049160003662, "eval_runtime": 4283.1765, "eval_samples_per_second": 17.649, "eval_steps_per_second": 1.471, "eval_wer": 0.0708675832806411, "step": 31000 }, { "epoch": 2.57, "learning_rate": 1.8791147132169577e-05, "loss": 0.1116, "step": 31100 }, { "epoch": 2.58, "learning_rate": 1.8786990856192854e-05, "loss": 0.1025, "step": 31200 }, { "epoch": 2.59, "learning_rate": 1.8782834580216128e-05, "loss": 0.107, "step": 31300 }, { "epoch": 2.6, "learning_rate": 1.8778678304239405e-05, "loss": 0.1061, "step": 31400 }, { "epoch": 2.61, "learning_rate": 1.877452202826268e-05, "loss": 0.1038, "step": 31500 }, { "epoch": 2.61, "eval_loss": 0.11699890345335007, "eval_runtime": 4303.5749, "eval_samples_per_second": 17.566, "eval_steps_per_second": 1.464, "eval_wer": 0.06898450925460115, "step": 31500 }, { "epoch": 2.62, "learning_rate": 1.8770365752285953e-05, "loss": 0.1079, "step": 31600 }, { "epoch": 2.62, "learning_rate": 1.876620947630923e-05, "loss": 0.097, "step": 31700 }, { "epoch": 2.63, "learning_rate": 1.8762053200332504e-05, "loss": 0.0961, "step": 31800 }, { "epoch": 2.64, "learning_rate": 1.8757896924355777e-05, "loss": 0.1075, "step": 31900 }, { "epoch": 2.65, "learning_rate": 1.8753740648379055e-05, "loss": 0.1095, "step": 32000 }, { "epoch": 2.65, "eval_loss": 0.11479315906763077, "eval_runtime": 4300.2811, "eval_samples_per_second": 17.579, "eval_steps_per_second": 1.465, "eval_wer": 0.06795374160962829, "step": 32000 }, { "epoch": 2.66, "learning_rate": 1.874958437240233e-05, "loss": 0.106, "step": 32100 }, { "epoch": 2.67, "learning_rate": 1.8745428096425606e-05, "loss": 0.101, "step": 32200 }, { "epoch": 2.67, "learning_rate": 1.874127182044888e-05, "loss": 0.1065, "step": 32300 }, { "epoch": 2.68, "learning_rate": 1.8737115544472157e-05, "loss": 0.1114, "step": 32400 }, { "epoch": 2.69, "learning_rate": 1.873295926849543e-05, "loss": 0.1109, "step": 32500 }, { "epoch": 2.69, "eval_loss": 0.12131709605455399, "eval_runtime": 4297.0699, "eval_samples_per_second": 17.592, "eval_steps_per_second": 1.466, "eval_wer": 0.07140912090904475, "step": 32500 }, { "epoch": 2.7, "learning_rate": 1.8728844555278472e-05, "loss": 0.0982, "step": 32600 }, { "epoch": 2.71, "learning_rate": 1.872468827930175e-05, "loss": 0.1039, "step": 32700 }, { "epoch": 2.72, "learning_rate": 1.8720532003325023e-05, "loss": 0.1045, "step": 32800 }, { "epoch": 2.72, "learning_rate": 1.8716375727348297e-05, "loss": 0.1134, "step": 32900 }, { "epoch": 2.73, "learning_rate": 1.871221945137157e-05, "loss": 0.1029, "step": 33000 }, { "epoch": 2.73, "eval_loss": 0.11449110507965088, "eval_runtime": 4330.1418, "eval_samples_per_second": 17.458, "eval_steps_per_second": 1.455, "eval_wer": 0.06798451079306032, "step": 33000 }, { "epoch": 2.74, "learning_rate": 1.8708063175394848e-05, "loss": 0.113, "step": 33100 }, { "epoch": 2.75, "learning_rate": 1.8703906899418122e-05, "loss": 0.1088, "step": 33200 }, { "epoch": 2.76, "learning_rate": 1.8699750623441396e-05, "loss": 0.1099, "step": 33300 }, { "epoch": 2.76, "learning_rate": 1.869563591022444e-05, "loss": 0.1057, "step": 33400 }, { "epoch": 2.77, "learning_rate": 1.8691479634247715e-05, "loss": 0.1089, "step": 33500 }, { "epoch": 2.77, "eval_loss": 0.127433642745018, "eval_runtime": 4310.5682, "eval_samples_per_second": 17.537, "eval_steps_per_second": 1.462, "eval_wer": 0.07011681520489968, "step": 33500 }, { "epoch": 2.78, "learning_rate": 1.868732335827099e-05, "loss": 0.1083, "step": 33600 }, { "epoch": 2.79, "learning_rate": 1.8683167082294266e-05, "loss": 0.1131, "step": 33700 }, { "epoch": 2.8, "learning_rate": 1.867901080631754e-05, "loss": 0.1023, "step": 33800 }, { "epoch": 2.81, "learning_rate": 1.8674854530340817e-05, "loss": 0.1039, "step": 33900 }, { "epoch": 2.81, "learning_rate": 1.867069825436409e-05, "loss": 0.1015, "step": 34000 }, { "epoch": 2.81, "eval_loss": 0.12738846242427826, "eval_runtime": 4313.9158, "eval_samples_per_second": 17.524, "eval_steps_per_second": 1.46, "eval_wer": 0.07029835338714864, "step": 34000 }, { "epoch": 2.82, "learning_rate": 1.8666541978387368e-05, "loss": 0.1134, "step": 34100 }, { "epoch": 2.83, "learning_rate": 1.8662385702410642e-05, "loss": 0.1143, "step": 34200 }, { "epoch": 2.84, "learning_rate": 1.8658229426433916e-05, "loss": 0.1008, "step": 34300 }, { "epoch": 2.85, "learning_rate": 1.8654073150457193e-05, "loss": 0.1101, "step": 34400 }, { "epoch": 2.86, "learning_rate": 1.8649916874480467e-05, "loss": 0.1043, "step": 34500 }, { "epoch": 2.86, "eval_loss": 0.12504072487354279, "eval_runtime": 4303.3701, "eval_samples_per_second": 17.566, "eval_steps_per_second": 1.464, "eval_wer": 0.07051373767117282, "step": 34500 }, { "epoch": 2.86, "learning_rate": 1.864576059850374e-05, "loss": 0.1124, "step": 34600 }, { "epoch": 2.87, "learning_rate": 1.8641604322527018e-05, "loss": 0.1114, "step": 34700 }, { "epoch": 2.88, "learning_rate": 1.863744804655029e-05, "loss": 0.1079, "step": 34800 }, { "epoch": 2.89, "learning_rate": 1.863329177057357e-05, "loss": 0.1036, "step": 34900 }, { "epoch": 2.9, "learning_rate": 1.8629135494596843e-05, "loss": 0.1307, "step": 35000 }, { "epoch": 2.9, "eval_loss": 0.12751950323581696, "eval_runtime": 4304.2505, "eval_samples_per_second": 17.563, "eval_steps_per_second": 1.464, "eval_wer": 0.0720721968120049, "step": 35000 }, { "epoch": 2.91, "learning_rate": 1.862497921862012e-05, "loss": 0.1072, "step": 35100 }, { "epoch": 2.91, "learning_rate": 1.8620822942643394e-05, "loss": 0.1084, "step": 35200 }, { "epoch": 2.92, "learning_rate": 1.8616666666666667e-05, "loss": 0.103, "step": 35300 }, { "epoch": 2.93, "learning_rate": 1.8612510390689945e-05, "loss": 0.108, "step": 35400 }, { "epoch": 2.94, "learning_rate": 1.860835411471322e-05, "loss": 0.1003, "step": 35500 }, { "epoch": 2.94, "eval_loss": 0.11630689352750778, "eval_runtime": 4349.6776, "eval_samples_per_second": 17.379, "eval_steps_per_second": 1.448, "eval_wer": 0.06557374527116112, "step": 35500 }, { "epoch": 2.95, "learning_rate": 1.8604197838736492e-05, "loss": 0.1073, "step": 35600 }, { "epoch": 2.96, "learning_rate": 1.860004156275977e-05, "loss": 0.1051, "step": 35700 }, { "epoch": 2.96, "learning_rate": 1.8595885286783043e-05, "loss": 0.1011, "step": 35800 }, { "epoch": 2.97, "learning_rate": 1.8591770573566085e-05, "loss": 0.1122, "step": 35900 }, { "epoch": 2.98, "learning_rate": 1.8587614297589362e-05, "loss": 0.0945, "step": 36000 }, { "epoch": 2.98, "eval_loss": 0.11436288803815842, "eval_runtime": 4296.7214, "eval_samples_per_second": 17.594, "eval_steps_per_second": 1.466, "eval_wer": 0.06728297341081013, "step": 36000 }, { "epoch": 2.99, "learning_rate": 1.8583458021612636e-05, "loss": 0.1152, "step": 36100 }, { "epoch": 3.0, "learning_rate": 1.857930174563591e-05, "loss": 0.1037, "step": 36200 }, { "epoch": 3.0, "learning_rate": 1.8575145469659187e-05, "loss": 0.0948, "step": 36300 }, { "epoch": 3.01, "learning_rate": 1.857098919368246e-05, "loss": 0.1216, "step": 36400 }, { "epoch": 3.02, "learning_rate": 1.8566832917705738e-05, "loss": 0.0886, "step": 36500 }, { "epoch": 3.02, "eval_loss": 0.11903050541877747, "eval_runtime": 4328.9528, "eval_samples_per_second": 17.463, "eval_steps_per_second": 1.455, "eval_wer": 0.06801527997649234, "step": 36500 }, { "epoch": 3.03, "learning_rate": 1.8562676641729012e-05, "loss": 0.0939, "step": 36600 }, { "epoch": 3.04, "learning_rate": 1.855852036575229e-05, "loss": 0.099, "step": 36700 }, { "epoch": 3.05, "learning_rate": 1.8554364089775563e-05, "loss": 0.0992, "step": 36800 }, { "epoch": 3.05, "learning_rate": 1.8550207813798837e-05, "loss": 0.0916, "step": 36900 }, { "epoch": 3.06, "learning_rate": 1.8546051537822114e-05, "loss": 0.0944, "step": 37000 }, { "epoch": 3.06, "eval_loss": 0.12649023532867432, "eval_runtime": 4296.7532, "eval_samples_per_second": 17.594, "eval_steps_per_second": 1.466, "eval_wer": 0.0667475896190929, "step": 37000 }, { "epoch": 3.07, "learning_rate": 1.8541895261845388e-05, "loss": 0.1023, "step": 37100 }, { "epoch": 3.08, "learning_rate": 1.8537738985868662e-05, "loss": 0.0989, "step": 37200 }, { "epoch": 3.09, "learning_rate": 1.853358270989194e-05, "loss": 0.1036, "step": 37300 }, { "epoch": 3.1, "learning_rate": 1.8529426433915213e-05, "loss": 0.091, "step": 37400 }, { "epoch": 3.1, "learning_rate": 1.852527015793849e-05, "loss": 0.105, "step": 37500 }, { "epoch": 3.1, "eval_loss": 0.12019772082567215, "eval_runtime": 4298.2163, "eval_samples_per_second": 17.588, "eval_steps_per_second": 1.466, "eval_wer": 0.0714691208167372, "step": 37500 }, { "epoch": 3.11, "learning_rate": 1.8521113881961764e-05, "loss": 0.1045, "step": 37600 }, { "epoch": 3.12, "learning_rate": 1.851695760598504e-05, "loss": 0.0926, "step": 37700 }, { "epoch": 3.13, "learning_rate": 1.8512801330008315e-05, "loss": 0.0982, "step": 37800 }, { "epoch": 3.14, "learning_rate": 1.850864505403159e-05, "loss": 0.0957, "step": 37900 }, { "epoch": 3.15, "learning_rate": 1.8504488778054866e-05, "loss": 0.0989, "step": 38000 }, { "epoch": 3.15, "eval_loss": 0.12159129977226257, "eval_runtime": 4315.9256, "eval_samples_per_second": 17.515, "eval_steps_per_second": 1.46, "eval_wer": 0.0682137412096289, "step": 38000 }, { "epoch": 3.15, "learning_rate": 1.850033250207814e-05, "loss": 0.0951, "step": 38100 }, { "epoch": 3.16, "learning_rate": 1.8496176226101414e-05, "loss": 0.0972, "step": 38200 }, { "epoch": 3.17, "learning_rate": 1.8492061512884455e-05, "loss": 0.0884, "step": 38300 }, { "epoch": 3.18, "learning_rate": 1.8487905236907733e-05, "loss": 0.0974, "step": 38400 }, { "epoch": 3.19, "learning_rate": 1.8483748960931006e-05, "loss": 0.1013, "step": 38500 }, { "epoch": 3.19, "eval_loss": 0.12531785666942596, "eval_runtime": 4319.5635, "eval_samples_per_second": 17.501, "eval_steps_per_second": 1.458, "eval_wer": 0.07043835317176435, "step": 38500 }, { "epoch": 3.2, "learning_rate": 1.847959268495428e-05, "loss": 0.0993, "step": 38600 }, { "epoch": 3.2, "learning_rate": 1.8475436408977557e-05, "loss": 0.0955, "step": 38700 }, { "epoch": 3.21, "learning_rate": 1.847128013300083e-05, "loss": 0.0951, "step": 38800 }, { "epoch": 3.22, "learning_rate": 1.846712385702411e-05, "loss": 0.0935, "step": 38900 }, { "epoch": 3.23, "learning_rate": 1.8462967581047386e-05, "loss": 0.1061, "step": 39000 }, { "epoch": 3.23, "eval_loss": 0.12737500667572021, "eval_runtime": 4334.8828, "eval_samples_per_second": 17.439, "eval_steps_per_second": 1.453, "eval_wer": 0.07019527662265135, "step": 39000 }, { "epoch": 3.24, "learning_rate": 1.845881130507066e-05, "loss": 0.0926, "step": 39100 }, { "epoch": 3.25, "learning_rate": 1.8454655029093933e-05, "loss": 0.0885, "step": 39200 }, { "epoch": 3.25, "learning_rate": 1.8450498753117207e-05, "loss": 0.0903, "step": 39300 }, { "epoch": 3.26, "learning_rate": 1.8446342477140484e-05, "loss": 0.0962, "step": 39400 }, { "epoch": 3.27, "learning_rate": 1.8442186201163758e-05, "loss": 0.1006, "step": 39500 }, { "epoch": 3.27, "eval_loss": 0.11653690785169601, "eval_runtime": 4290.7049, "eval_samples_per_second": 17.618, "eval_steps_per_second": 1.468, "eval_wer": 0.07048143002856919, "step": 39500 }, { "epoch": 3.28, "learning_rate": 1.8438029925187032e-05, "loss": 0.106, "step": 39600 }, { "epoch": 3.29, "learning_rate": 1.843387364921031e-05, "loss": 0.089, "step": 39700 }, { "epoch": 3.29, "learning_rate": 1.8429717373233583e-05, "loss": 0.1005, "step": 39800 }, { "epoch": 3.3, "learning_rate": 1.842556109725686e-05, "loss": 0.1062, "step": 39900 }, { "epoch": 3.31, "learning_rate": 1.8421404821280134e-05, "loss": 0.0976, "step": 40000 }, { "epoch": 3.31, "eval_loss": 0.12096834927797318, "eval_runtime": 4298.7824, "eval_samples_per_second": 17.585, "eval_steps_per_second": 1.466, "eval_wer": 0.06842451011613829, "step": 40000 }, { "epoch": 3.32, "learning_rate": 1.841724854530341e-05, "loss": 0.1069, "step": 40100 }, { "epoch": 3.33, "learning_rate": 1.8413092269326685e-05, "loss": 0.0937, "step": 40200 }, { "epoch": 3.34, "learning_rate": 1.840893599334996e-05, "loss": 0.0956, "step": 40300 }, { "epoch": 3.34, "learning_rate": 1.8404779717373236e-05, "loss": 0.0998, "step": 40400 }, { "epoch": 3.35, "learning_rate": 1.840062344139651e-05, "loss": 0.094, "step": 40500 }, { "epoch": 3.35, "eval_loss": 0.12432394176721573, "eval_runtime": 4306.3766, "eval_samples_per_second": 17.554, "eval_steps_per_second": 1.463, "eval_wer": 0.07025219961200059, "step": 40500 }, { "epoch": 3.36, "learning_rate": 1.8396467165419784e-05, "loss": 0.1036, "step": 40600 }, { "epoch": 3.37, "learning_rate": 1.839231088944306e-05, "loss": 0.0981, "step": 40700 }, { "epoch": 3.38, "learning_rate": 1.8388154613466335e-05, "loss": 0.1009, "step": 40800 }, { "epoch": 3.39, "learning_rate": 1.838399833748961e-05, "loss": 0.0919, "step": 40900 }, { "epoch": 3.39, "learning_rate": 1.8379842061512886e-05, "loss": 0.0918, "step": 41000 }, { "epoch": 3.39, "eval_loss": 0.12724146246910095, "eval_runtime": 4317.8074, "eval_samples_per_second": 17.508, "eval_steps_per_second": 1.459, "eval_wer": 0.06935066253744225, "step": 41000 }, { "epoch": 3.4, "learning_rate": 1.8375685785536163e-05, "loss": 0.0947, "step": 41100 }, { "epoch": 3.41, "learning_rate": 1.8371529509559437e-05, "loss": 0.0976, "step": 41200 }, { "epoch": 3.42, "learning_rate": 1.836741479634248e-05, "loss": 0.0989, "step": 41300 }, { "epoch": 3.43, "learning_rate": 1.8363258520365756e-05, "loss": 0.0984, "step": 41400 }, { "epoch": 3.44, "learning_rate": 1.835910224438903e-05, "loss": 0.0981, "step": 41500 }, { "epoch": 3.44, "eval_loss": 0.1341506540775299, "eval_runtime": 4294.8505, "eval_samples_per_second": 17.601, "eval_steps_per_second": 1.467, "eval_wer": 0.07259373447117774, "step": 41500 }, { "epoch": 3.44, "learning_rate": 1.8354945968412304e-05, "loss": 0.1001, "step": 41600 }, { "epoch": 3.45, "learning_rate": 1.835078969243558e-05, "loss": 0.0968, "step": 41700 }, { "epoch": 3.46, "learning_rate": 1.8346633416458855e-05, "loss": 0.0916, "step": 41800 }, { "epoch": 3.47, "learning_rate": 1.834247714048213e-05, "loss": 0.0925, "step": 41900 }, { "epoch": 3.48, "learning_rate": 1.8338320864505406e-05, "loss": 0.0919, "step": 42000 }, { "epoch": 3.48, "eval_loss": 0.12683247029781342, "eval_runtime": 4314.3412, "eval_samples_per_second": 17.522, "eval_steps_per_second": 1.46, "eval_wer": 0.07166296667235897, "step": 42000 }, { "epoch": 3.49, "learning_rate": 1.833416458852868e-05, "loss": 0.0962, "step": 42100 }, { "epoch": 3.49, "learning_rate": 1.8330008312551953e-05, "loss": 0.0959, "step": 42200 }, { "epoch": 3.5, "learning_rate": 1.832585203657523e-05, "loss": 0.1014, "step": 42300 }, { "epoch": 3.51, "learning_rate": 1.8321695760598504e-05, "loss": 0.1006, "step": 42400 }, { "epoch": 3.52, "learning_rate": 1.831753948462178e-05, "loss": 0.0987, "step": 42500 }, { "epoch": 3.52, "eval_loss": 0.1158851757645607, "eval_runtime": 4281.9663, "eval_samples_per_second": 17.654, "eval_steps_per_second": 1.471, "eval_wer": 0.0677152804380301, "step": 42500 }, { "epoch": 3.53, "learning_rate": 1.8313383208645055e-05, "loss": 0.0976, "step": 42600 }, { "epoch": 3.53, "learning_rate": 1.8309226932668333e-05, "loss": 0.0973, "step": 42700 }, { "epoch": 3.54, "learning_rate": 1.8305070656691606e-05, "loss": 0.0947, "step": 42800 }, { "epoch": 3.55, "learning_rate": 1.830091438071488e-05, "loss": 0.093, "step": 42900 }, { "epoch": 3.56, "learning_rate": 1.8296758104738157e-05, "loss": 0.095, "step": 43000 }, { "epoch": 3.56, "eval_loss": 0.11736118793487549, "eval_runtime": 4288.3242, "eval_samples_per_second": 17.628, "eval_steps_per_second": 1.469, "eval_wer": 0.06725374268654971, "step": 43000 }, { "epoch": 3.57, "learning_rate": 1.829260182876143e-05, "loss": 0.0971, "step": 43100 }, { "epoch": 3.58, "learning_rate": 1.8288445552784705e-05, "loss": 0.1007, "step": 43200 }, { "epoch": 3.58, "learning_rate": 1.828428927680798e-05, "loss": 0.0996, "step": 43300 }, { "epoch": 3.59, "learning_rate": 1.8280133000831256e-05, "loss": 0.1034, "step": 43400 }, { "epoch": 3.6, "learning_rate": 1.8275976724854533e-05, "loss": 0.0955, "step": 43500 }, { "epoch": 3.6, "eval_loss": 0.1226612851023674, "eval_runtime": 4298.1813, "eval_samples_per_second": 17.588, "eval_steps_per_second": 1.466, "eval_wer": 0.0667491280782645, "step": 43500 }, { "epoch": 3.61, "learning_rate": 1.8271820448877807e-05, "loss": 0.0864, "step": 43600 }, { "epoch": 3.62, "learning_rate": 1.8267664172901084e-05, "loss": 0.0983, "step": 43700 }, { "epoch": 3.63, "learning_rate": 1.8263507896924358e-05, "loss": 0.0867, "step": 43800 }, { "epoch": 3.63, "learning_rate": 1.8259351620947632e-05, "loss": 0.0909, "step": 43900 }, { "epoch": 3.64, "learning_rate": 1.825519534497091e-05, "loss": 0.0925, "step": 44000 }, { "epoch": 3.64, "eval_loss": 0.1232045590877533, "eval_runtime": 4273.2419, "eval_samples_per_second": 17.69, "eval_steps_per_second": 1.474, "eval_wer": 0.06785374176347421, "step": 44000 }, { "epoch": 3.65, "learning_rate": 1.8251039068994183e-05, "loss": 0.0884, "step": 44100 }, { "epoch": 3.66, "learning_rate": 1.8246882793017457e-05, "loss": 0.094, "step": 44200 }, { "epoch": 3.67, "learning_rate": 1.824272651704073e-05, "loss": 0.0872, "step": 44300 }, { "epoch": 3.68, "learning_rate": 1.8238570241064008e-05, "loss": 0.0935, "step": 44400 }, { "epoch": 3.68, "learning_rate": 1.823441396508728e-05, "loss": 0.1099, "step": 44500 }, { "epoch": 3.68, "eval_loss": 0.12568922340869904, "eval_runtime": 4283.0995, "eval_samples_per_second": 17.65, "eval_steps_per_second": 1.471, "eval_wer": 0.06664758977293882, "step": 44500 }, { "epoch": 3.69, "learning_rate": 1.823025768911056e-05, "loss": 0.0889, "step": 44600 }, { "epoch": 3.7, "learning_rate": 1.8226101413133833e-05, "loss": 0.0958, "step": 44700 }, { "epoch": 3.71, "learning_rate": 1.822194513715711e-05, "loss": 0.0911, "step": 44800 }, { "epoch": 3.72, "learning_rate": 1.8217788861180384e-05, "loss": 0.0894, "step": 44900 }, { "epoch": 3.73, "learning_rate": 1.821363258520366e-05, "loss": 0.0908, "step": 45000 }, { "epoch": 3.73, "eval_loss": 0.11977280676364899, "eval_runtime": 4286.0557, "eval_samples_per_second": 17.637, "eval_steps_per_second": 1.47, "eval_wer": 0.06597989849246386, "step": 45000 }, { "epoch": 3.73, "learning_rate": 1.8209517871986703e-05, "loss": 0.103, "step": 45100 }, { "epoch": 3.74, "learning_rate": 1.8205361596009977e-05, "loss": 0.0938, "step": 45200 }, { "epoch": 3.75, "learning_rate": 1.8201205320033254e-05, "loss": 0.0964, "step": 45300 }, { "epoch": 3.76, "learning_rate": 1.8197049044056528e-05, "loss": 0.0866, "step": 45400 }, { "epoch": 3.77, "learning_rate": 1.81928927680798e-05, "loss": 0.0923, "step": 45500 }, { "epoch": 3.77, "eval_loss": 0.13010576367378235, "eval_runtime": 4335.1699, "eval_samples_per_second": 17.438, "eval_steps_per_second": 1.453, "eval_wer": 0.06937681634335947, "step": 45500 }, { "epoch": 3.77, "learning_rate": 1.8188736492103075e-05, "loss": 0.0888, "step": 45600 }, { "epoch": 3.78, "learning_rate": 1.8184580216126352e-05, "loss": 0.0906, "step": 45700 }, { "epoch": 3.79, "learning_rate": 1.8180423940149626e-05, "loss": 0.0904, "step": 45800 }, { "epoch": 3.8, "learning_rate": 1.8176267664172904e-05, "loss": 0.0906, "step": 45900 }, { "epoch": 3.81, "learning_rate": 1.8172111388196177e-05, "loss": 0.0898, "step": 46000 }, { "epoch": 3.81, "eval_loss": 0.12169438600540161, "eval_runtime": 4295.4334, "eval_samples_per_second": 17.599, "eval_steps_per_second": 1.467, "eval_wer": 0.06838297171850505, "step": 46000 }, { "epoch": 3.82, "learning_rate": 1.8167955112219455e-05, "loss": 0.0826, "step": 46100 }, { "epoch": 3.82, "learning_rate": 1.816379883624273e-05, "loss": 0.091, "step": 46200 }, { "epoch": 3.83, "learning_rate": 1.8159642560266006e-05, "loss": 0.0901, "step": 46300 }, { "epoch": 3.84, "learning_rate": 1.815548628428928e-05, "loss": 0.0867, "step": 46400 }, { "epoch": 3.85, "learning_rate": 1.8151330008312553e-05, "loss": 0.095, "step": 46500 }, { "epoch": 3.85, "eval_loss": 0.12022671103477478, "eval_runtime": 4276.7534, "eval_samples_per_second": 17.676, "eval_steps_per_second": 1.473, "eval_wer": 0.06813374133270564, "step": 46500 }, { "epoch": 3.86, "learning_rate": 1.8147173732335827e-05, "loss": 0.0931, "step": 46600 }, { "epoch": 3.87, "learning_rate": 1.8143017456359104e-05, "loss": 0.0954, "step": 46700 }, { "epoch": 3.87, "learning_rate": 1.8138861180382378e-05, "loss": 0.0882, "step": 46800 }, { "epoch": 3.88, "learning_rate": 1.8134704904405652e-05, "loss": 0.099, "step": 46900 }, { "epoch": 3.89, "learning_rate": 1.813054862842893e-05, "loss": 0.0925, "step": 47000 }, { "epoch": 3.89, "eval_loss": 0.12550650537014008, "eval_runtime": 4294.1398, "eval_samples_per_second": 17.604, "eval_steps_per_second": 1.467, "eval_wer": 0.06745374237885787, "step": 47000 }, { "epoch": 3.9, "learning_rate": 1.8126392352452203e-05, "loss": 0.0928, "step": 47100 }, { "epoch": 3.91, "learning_rate": 1.812223607647548e-05, "loss": 0.0971, "step": 47200 }, { "epoch": 3.92, "learning_rate": 1.8118079800498757e-05, "loss": 0.0945, "step": 47300 }, { "epoch": 3.92, "learning_rate": 1.8113965087281796e-05, "loss": 0.0949, "step": 47400 }, { "epoch": 3.93, "learning_rate": 1.8109808811305073e-05, "loss": 0.0948, "step": 47500 }, { "epoch": 3.93, "eval_loss": 0.11757665127515793, "eval_runtime": 4277.9752, "eval_samples_per_second": 17.671, "eval_steps_per_second": 1.473, "eval_wer": 0.06651220536583789, "step": 47500 }, { "epoch": 3.94, "learning_rate": 1.8105652535328347e-05, "loss": 0.0907, "step": 47600 }, { "epoch": 3.95, "learning_rate": 1.8101496259351624e-05, "loss": 0.099, "step": 47700 }, { "epoch": 3.96, "learning_rate": 1.8097339983374898e-05, "loss": 0.0959, "step": 47800 }, { "epoch": 3.97, "learning_rate": 1.809318370739817e-05, "loss": 0.0965, "step": 47900 }, { "epoch": 3.97, "learning_rate": 1.808902743142145e-05, "loss": 0.0905, "step": 48000 }, { "epoch": 3.97, "eval_loss": 0.11655454337596893, "eval_runtime": 4280.8179, "eval_samples_per_second": 17.659, "eval_steps_per_second": 1.472, "eval_wer": 0.06458605448299311, "step": 48000 }, { "epoch": 3.98, "learning_rate": 1.8084871155444723e-05, "loss": 0.0903, "step": 48100 }, { "epoch": 3.99, "learning_rate": 1.8080714879467997e-05, "loss": 0.0861, "step": 48200 }, { "epoch": 4.0, "learning_rate": 1.8076558603491274e-05, "loss": 0.0919, "step": 48300 }, { "epoch": 4.01, "learning_rate": 1.8072402327514548e-05, "loss": 0.09, "step": 48400 }, { "epoch": 4.01, "learning_rate": 1.8068246051537825e-05, "loss": 0.0856, "step": 48500 }, { "epoch": 4.01, "eval_loss": 0.12168645858764648, "eval_runtime": 4363.3968, "eval_samples_per_second": 17.325, "eval_steps_per_second": 1.444, "eval_wer": 0.06436143944393932, "step": 48500 }, { "epoch": 4.02, "learning_rate": 1.80640897755611e-05, "loss": 0.0861, "step": 48600 }, { "epoch": 4.03, "learning_rate": 1.8059933499584376e-05, "loss": 0.086, "step": 48700 }, { "epoch": 4.04, "learning_rate": 1.805577722360765e-05, "loss": 0.0892, "step": 48800 }, { "epoch": 4.05, "learning_rate": 1.8051620947630923e-05, "loss": 0.0824, "step": 48900 }, { "epoch": 4.06, "learning_rate": 1.804750623441397e-05, "loss": 0.0864, "step": 49000 }, { "epoch": 4.06, "eval_loss": 0.11592572927474976, "eval_runtime": 4508.1981, "eval_samples_per_second": 16.768, "eval_steps_per_second": 1.397, "eval_wer": 0.0632522103812148, "step": 49000 }, { "epoch": 4.06, "learning_rate": 1.8043349958437242e-05, "loss": 0.0808, "step": 49100 }, { "epoch": 4.07, "learning_rate": 1.8039193682460516e-05, "loss": 0.0918, "step": 49200 }, { "epoch": 4.08, "learning_rate": 1.8035037406483794e-05, "loss": 0.0897, "step": 49300 }, { "epoch": 4.09, "learning_rate": 1.8030881130507067e-05, "loss": 0.0812, "step": 49400 }, { "epoch": 4.1, "learning_rate": 1.802672485453034e-05, "loss": 0.0891, "step": 49500 }, { "epoch": 4.1, "eval_loss": 0.12310642749071121, "eval_runtime": 4540.6147, "eval_samples_per_second": 16.649, "eval_steps_per_second": 1.387, "eval_wer": 0.06541066859897139, "step": 49500 }, { "epoch": 4.11, "learning_rate": 1.8022568578553615e-05, "loss": 0.0804, "step": 49600 }, { "epoch": 4.11, "learning_rate": 1.8018412302576892e-05, "loss": 0.0786, "step": 49700 }, { "epoch": 4.12, "learning_rate": 1.8014256026600166e-05, "loss": 0.0783, "step": 49800 }, { "epoch": 4.13, "learning_rate": 1.8010099750623443e-05, "loss": 0.083, "step": 49900 }, { "epoch": 4.14, "learning_rate": 1.800594347464672e-05, "loss": 0.1031, "step": 50000 }, { "epoch": 4.14, "eval_loss": 0.12009721249341965, "eval_runtime": 4556.6622, "eval_samples_per_second": 16.59, "eval_steps_per_second": 1.383, "eval_wer": 0.06431528566879127, "step": 50000 }, { "epoch": 4.15, "learning_rate": 1.8001787198669994e-05, "loss": 0.0868, "step": 50100 }, { "epoch": 4.16, "learning_rate": 1.7997630922693268e-05, "loss": 0.0903, "step": 50200 }, { "epoch": 4.16, "learning_rate": 1.7993474646716545e-05, "loss": 0.0856, "step": 50300 }, { "epoch": 4.17, "learning_rate": 1.798931837073982e-05, "loss": 0.0883, "step": 50400 }, { "epoch": 4.18, "learning_rate": 1.7985162094763093e-05, "loss": 0.0887, "step": 50500 }, { "epoch": 4.18, "eval_loss": 0.11914674937725067, "eval_runtime": 4505.2715, "eval_samples_per_second": 16.779, "eval_steps_per_second": 1.398, "eval_wer": 0.06479374647115928, "step": 50500 }, { "epoch": 4.19, "learning_rate": 1.7981005818786367e-05, "loss": 0.0927, "step": 50600 }, { "epoch": 4.2, "learning_rate": 1.7976849542809644e-05, "loss": 0.0899, "step": 50700 }, { "epoch": 4.21, "learning_rate": 1.7972693266832918e-05, "loss": 0.079, "step": 50800 }, { "epoch": 4.21, "learning_rate": 1.7968536990856195e-05, "loss": 0.0867, "step": 50900 }, { "epoch": 4.22, "learning_rate": 1.796438071487947e-05, "loss": 0.083, "step": 51000 }, { "epoch": 4.22, "eval_loss": 0.11644386500120163, "eval_runtime": 4464.1884, "eval_samples_per_second": 16.934, "eval_steps_per_second": 1.411, "eval_wer": 0.06439990092322935, "step": 51000 }, { "epoch": 4.23, "learning_rate": 1.7960224438902746e-05, "loss": 0.0831, "step": 51100 }, { "epoch": 4.24, "learning_rate": 1.795606816292602e-05, "loss": 0.0899, "step": 51200 }, { "epoch": 4.25, "learning_rate": 1.7951911886949297e-05, "loss": 0.0774, "step": 51300 }, { "epoch": 4.25, "learning_rate": 1.794775561097257e-05, "loss": 0.0883, "step": 51400 }, { "epoch": 4.26, "learning_rate": 1.7943599334995845e-05, "loss": 0.0904, "step": 51500 }, { "epoch": 4.26, "eval_loss": 0.11615677177906036, "eval_runtime": 4503.9486, "eval_samples_per_second": 16.784, "eval_steps_per_second": 1.399, "eval_wer": 0.06591220628891341, "step": 51500 }, { "epoch": 4.27, "learning_rate": 1.793944305901912e-05, "loss": 0.0832, "step": 51600 }, { "epoch": 4.28, "learning_rate": 1.7935286783042396e-05, "loss": 0.0804, "step": 51700 }, { "epoch": 4.29, "learning_rate": 1.793113050706567e-05, "loss": 0.0857, "step": 51800 }, { "epoch": 4.3, "learning_rate": 1.7926974231088943e-05, "loss": 0.0843, "step": 51900 }, { "epoch": 4.3, "learning_rate": 1.792281795511222e-05, "loss": 0.0819, "step": 52000 }, { "epoch": 4.3, "eval_loss": 0.11846227198839188, "eval_runtime": 4358.4478, "eval_samples_per_second": 17.344, "eval_steps_per_second": 1.445, "eval_wer": 0.06386913250902691, "step": 52000 }, { "epoch": 4.31, "learning_rate": 1.7918661679135498e-05, "loss": 0.0887, "step": 52100 }, { "epoch": 4.32, "learning_rate": 1.791450540315877e-05, "loss": 0.0814, "step": 52200 }, { "epoch": 4.33, "learning_rate": 1.791034912718205e-05, "loss": 0.0783, "step": 52300 }, { "epoch": 4.34, "learning_rate": 1.7906192851205323e-05, "loss": 0.0841, "step": 52400 }, { "epoch": 4.35, "learning_rate": 1.7902036575228596e-05, "loss": 0.0833, "step": 52500 }, { "epoch": 4.35, "eval_loss": 0.10962552577257156, "eval_runtime": 4310.6763, "eval_samples_per_second": 17.537, "eval_steps_per_second": 1.461, "eval_wer": 0.06353990224630424, "step": 52500 }, { "epoch": 4.35, "learning_rate": 1.789788029925187e-05, "loss": 0.0893, "step": 52600 }, { "epoch": 4.36, "learning_rate": 1.7893724023275148e-05, "loss": 0.084, "step": 52700 }, { "epoch": 4.37, "learning_rate": 1.788956774729842e-05, "loss": 0.0796, "step": 52800 }, { "epoch": 4.38, "learning_rate": 1.7885411471321695e-05, "loss": 0.0811, "step": 52900 }, { "epoch": 4.39, "learning_rate": 1.7881255195344972e-05, "loss": 0.0787, "step": 53000 }, { "epoch": 4.39, "eval_loss": 0.1263929158449173, "eval_runtime": 4488.6267, "eval_samples_per_second": 16.841, "eval_steps_per_second": 1.404, "eval_wer": 0.06621374428654725, "step": 53000 }, { "epoch": 4.4, "learning_rate": 1.7877098919368246e-05, "loss": 0.0802, "step": 53100 }, { "epoch": 4.4, "learning_rate": 1.7872942643391523e-05, "loss": 0.074, "step": 53200 }, { "epoch": 4.41, "learning_rate": 1.7868827930174565e-05, "loss": 0.0884, "step": 53300 }, { "epoch": 4.42, "learning_rate": 1.786467165419784e-05, "loss": 0.0858, "step": 53400 }, { "epoch": 4.43, "learning_rate": 1.7860515378221116e-05, "loss": 0.0845, "step": 53500 }, { "epoch": 4.43, "eval_loss": 0.12085627019405365, "eval_runtime": 4405.7188, "eval_samples_per_second": 17.158, "eval_steps_per_second": 1.43, "eval_wer": 0.0651875920190892, "step": 53500 }, { "epoch": 4.44, "learning_rate": 1.785635910224439e-05, "loss": 0.0866, "step": 53600 }, { "epoch": 4.45, "learning_rate": 1.7852202826267667e-05, "loss": 0.0955, "step": 53700 }, { "epoch": 4.45, "learning_rate": 1.784804655029094e-05, "loss": 0.0895, "step": 53800 }, { "epoch": 4.46, "learning_rate": 1.7843890274314215e-05, "loss": 0.0885, "step": 53900 }, { "epoch": 4.47, "learning_rate": 1.7839733998337492e-05, "loss": 0.0852, "step": 54000 }, { "epoch": 4.47, "eval_loss": 0.12128803133964539, "eval_runtime": 4470.4947, "eval_samples_per_second": 16.91, "eval_steps_per_second": 1.409, "eval_wer": 0.06452143919778584, "step": 54000 }, { "epoch": 4.48, "learning_rate": 1.7835577722360766e-05, "loss": 0.0908, "step": 54100 }, { "epoch": 4.49, "learning_rate": 1.783142144638404e-05, "loss": 0.0928, "step": 54200 }, { "epoch": 4.5, "learning_rate": 1.7827265170407317e-05, "loss": 0.0919, "step": 54300 }, { "epoch": 4.5, "learning_rate": 1.782310889443059e-05, "loss": 0.0847, "step": 54400 }, { "epoch": 4.51, "learning_rate": 1.7818952618453868e-05, "loss": 0.0772, "step": 54500 }, { "epoch": 4.51, "eval_loss": 0.12099339812994003, "eval_runtime": 4319.4127, "eval_samples_per_second": 17.501, "eval_steps_per_second": 1.459, "eval_wer": 0.06439067016819974, "step": 54500 }, { "epoch": 4.52, "learning_rate": 1.7814796342477142e-05, "loss": 0.0893, "step": 54600 }, { "epoch": 4.53, "learning_rate": 1.781064006650042e-05, "loss": 0.0893, "step": 54700 }, { "epoch": 4.54, "learning_rate": 1.7806483790523693e-05, "loss": 0.0943, "step": 54800 }, { "epoch": 4.54, "learning_rate": 1.7802327514546967e-05, "loss": 0.0883, "step": 54900 }, { "epoch": 4.55, "learning_rate": 1.7798171238570244e-05, "loss": 0.0817, "step": 55000 }, { "epoch": 4.55, "eval_loss": 0.1260470300912857, "eval_runtime": 4273.4988, "eval_samples_per_second": 17.689, "eval_steps_per_second": 1.474, "eval_wer": 0.06431528566879127, "step": 55000 }, { "epoch": 4.56, "learning_rate": 1.7794014962593518e-05, "loss": 0.0859, "step": 55100 }, { "epoch": 4.57, "learning_rate": 1.778985868661679e-05, "loss": 0.0799, "step": 55200 }, { "epoch": 4.58, "learning_rate": 1.778570241064007e-05, "loss": 0.094, "step": 55300 }, { "epoch": 4.59, "learning_rate": 1.7781546134663343e-05, "loss": 0.0789, "step": 55400 }, { "epoch": 4.59, "learning_rate": 1.7777389858686616e-05, "loss": 0.1003, "step": 55500 }, { "epoch": 4.59, "eval_loss": 0.12432339787483215, "eval_runtime": 4281.413, "eval_samples_per_second": 17.657, "eval_steps_per_second": 1.471, "eval_wer": 0.0664045132238258, "step": 55500 }, { "epoch": 4.6, "learning_rate": 1.7773233582709894e-05, "loss": 0.078, "step": 55600 }, { "epoch": 4.61, "learning_rate": 1.7769077306733167e-05, "loss": 0.0858, "step": 55700 }, { "epoch": 4.62, "learning_rate": 1.7764921030756445e-05, "loss": 0.0832, "step": 55800 }, { "epoch": 4.63, "learning_rate": 1.776076475477972e-05, "loss": 0.0877, "step": 55900 }, { "epoch": 4.64, "learning_rate": 1.775665004156276e-05, "loss": 0.089, "step": 56000 }, { "epoch": 4.64, "eval_loss": 0.115996815264225, "eval_runtime": 4268.4854, "eval_samples_per_second": 17.71, "eval_steps_per_second": 1.476, "eval_wer": 0.06369836354097917, "step": 56000 }, { "epoch": 4.64, "learning_rate": 1.7752493765586038e-05, "loss": 0.0806, "step": 56100 }, { "epoch": 4.65, "learning_rate": 1.774833748960931e-05, "loss": 0.0814, "step": 56200 }, { "epoch": 4.66, "learning_rate": 1.774418121363259e-05, "loss": 0.0811, "step": 56300 }, { "epoch": 4.67, "learning_rate": 1.7740024937655862e-05, "loss": 0.092, "step": 56400 }, { "epoch": 4.68, "learning_rate": 1.7735868661679136e-05, "loss": 0.0924, "step": 56500 }, { "epoch": 4.68, "eval_loss": 0.12010349333286285, "eval_runtime": 4395.0854, "eval_samples_per_second": 17.2, "eval_steps_per_second": 1.433, "eval_wer": 0.0661475905421684, "step": 56500 }, { "epoch": 4.69, "learning_rate": 1.7731712385702413e-05, "loss": 0.085, "step": 56600 }, { "epoch": 4.69, "learning_rate": 1.7727556109725687e-05, "loss": 0.0813, "step": 56700 }, { "epoch": 4.7, "learning_rate": 1.772339983374896e-05, "loss": 0.0803, "step": 56800 }, { "epoch": 4.71, "learning_rate": 1.7719243557772238e-05, "loss": 0.0822, "step": 56900 }, { "epoch": 4.72, "learning_rate": 1.7715087281795512e-05, "loss": 0.0782, "step": 57000 }, { "epoch": 4.72, "eval_loss": 0.13092364370822906, "eval_runtime": 4472.9611, "eval_samples_per_second": 16.9, "eval_steps_per_second": 1.408, "eval_wer": 0.06774297270311892, "step": 57000 }, { "epoch": 4.73, "learning_rate": 1.771093100581879e-05, "loss": 0.0875, "step": 57100 }, { "epoch": 4.74, "learning_rate": 1.7706774729842063e-05, "loss": 0.0823, "step": 57200 }, { "epoch": 4.74, "learning_rate": 1.770261845386534e-05, "loss": 0.0896, "step": 57300 }, { "epoch": 4.75, "learning_rate": 1.7698503740648382e-05, "loss": 0.0768, "step": 57400 }, { "epoch": 4.76, "learning_rate": 1.7694347464671656e-05, "loss": 0.0791, "step": 57500 }, { "epoch": 4.76, "eval_loss": 0.1266772747039795, "eval_runtime": 4493.7838, "eval_samples_per_second": 16.822, "eval_steps_per_second": 1.402, "eval_wer": 0.06614143670548199, "step": 57500 }, { "epoch": 4.77, "learning_rate": 1.7690191188694933e-05, "loss": 0.0838, "step": 57600 }, { "epoch": 4.78, "learning_rate": 1.7686034912718207e-05, "loss": 0.0815, "step": 57700 }, { "epoch": 4.78, "learning_rate": 1.768187863674148e-05, "loss": 0.0827, "step": 57800 }, { "epoch": 4.79, "learning_rate": 1.7677722360764755e-05, "loss": 0.079, "step": 57900 }, { "epoch": 4.8, "learning_rate": 1.7673566084788032e-05, "loss": 0.0873, "step": 58000 }, { "epoch": 4.8, "eval_loss": 0.12053040415048599, "eval_runtime": 4486.2793, "eval_samples_per_second": 16.85, "eval_steps_per_second": 1.404, "eval_wer": 0.06493374625577499, "step": 58000 }, { "epoch": 4.81, "learning_rate": 1.7669409808811306e-05, "loss": 0.0885, "step": 58100 }, { "epoch": 4.82, "learning_rate": 1.766525353283458e-05, "loss": 0.0932, "step": 58200 }, { "epoch": 4.83, "learning_rate": 1.7661097256857857e-05, "loss": 0.0785, "step": 58300 }, { "epoch": 4.83, "learning_rate": 1.765694098088113e-05, "loss": 0.075, "step": 58400 }, { "epoch": 4.84, "learning_rate": 1.7652784704904408e-05, "loss": 0.0808, "step": 58500 }, { "epoch": 4.84, "eval_loss": 0.12065942585468292, "eval_runtime": 4506.5513, "eval_samples_per_second": 16.774, "eval_steps_per_second": 1.398, "eval_wer": 0.0648306694912777, "step": 58500 }, { "epoch": 4.85, "learning_rate": 1.764862842892768e-05, "loss": 0.0844, "step": 58600 }, { "epoch": 4.86, "learning_rate": 1.764447215295096e-05, "loss": 0.0862, "step": 58700 }, { "epoch": 4.87, "learning_rate": 1.7640315876974233e-05, "loss": 0.0853, "step": 58800 }, { "epoch": 4.88, "learning_rate": 1.7636159600997506e-05, "loss": 0.0851, "step": 58900 }, { "epoch": 4.88, "learning_rate": 1.7632003325020784e-05, "loss": 0.0916, "step": 59000 }, { "epoch": 4.88, "eval_loss": 0.1208178922533989, "eval_runtime": 4527.3065, "eval_samples_per_second": 16.698, "eval_steps_per_second": 1.392, "eval_wer": 0.06226759651138998, "step": 59000 }, { "epoch": 4.89, "learning_rate": 1.7627847049044057e-05, "loss": 0.0848, "step": 59100 }, { "epoch": 4.9, "learning_rate": 1.762369077306733e-05, "loss": 0.0787, "step": 59200 }, { "epoch": 4.91, "learning_rate": 1.761953449709061e-05, "loss": 0.0788, "step": 59300 }, { "epoch": 4.92, "learning_rate": 1.761541978387365e-05, "loss": 0.0869, "step": 59400 }, { "epoch": 4.93, "learning_rate": 1.7611263507896924e-05, "loss": 0.0851, "step": 59500 }, { "epoch": 4.93, "eval_loss": 0.1335378885269165, "eval_runtime": 4481.0795, "eval_samples_per_second": 16.87, "eval_steps_per_second": 1.406, "eval_wer": 0.0649506693066626, "step": 59500 }, { "epoch": 4.93, "learning_rate": 1.76071072319202e-05, "loss": 0.0819, "step": 59600 }, { "epoch": 4.94, "learning_rate": 1.7602950955943475e-05, "loss": 0.0842, "step": 59700 }, { "epoch": 4.95, "learning_rate": 1.7598794679966752e-05, "loss": 0.0852, "step": 59800 }, { "epoch": 4.96, "learning_rate": 1.7594638403990026e-05, "loss": 0.0873, "step": 59900 }, { "epoch": 4.97, "learning_rate": 1.7590482128013303e-05, "loss": 0.0877, "step": 60000 }, { "epoch": 4.97, "eval_loss": 0.11959858983755112, "eval_runtime": 4487.6836, "eval_samples_per_second": 16.845, "eval_steps_per_second": 1.404, "eval_wer": 0.06187990480014646, "step": 60000 }, { "epoch": 4.98, "learning_rate": 1.7586325852036577e-05, "loss": 0.0845, "step": 60100 }, { "epoch": 4.98, "learning_rate": 1.758216957605985e-05, "loss": 0.085, "step": 60200 }, { "epoch": 4.99, "learning_rate": 1.7578013300083128e-05, "loss": 0.0816, "step": 60300 }, { "epoch": 5.0, "learning_rate": 1.7573857024106402e-05, "loss": 0.089, "step": 60400 }, { "epoch": 5.01, "learning_rate": 1.7569700748129676e-05, "loss": 0.0794, "step": 60500 }, { "epoch": 5.01, "eval_loss": 0.14357341825962067, "eval_runtime": 4418.5517, "eval_samples_per_second": 17.109, "eval_steps_per_second": 1.426, "eval_wer": 0.06889527862264827, "step": 60500 }, { "epoch": 5.02, "learning_rate": 1.7565544472152953e-05, "loss": 0.0846, "step": 60600 }, { "epoch": 5.02, "learning_rate": 1.7561388196176227e-05, "loss": 0.0777, "step": 60700 }, { "epoch": 5.03, "learning_rate": 1.75572319201995e-05, "loss": 0.078, "step": 60800 }, { "epoch": 5.04, "learning_rate": 1.7553075644222778e-05, "loss": 0.0823, "step": 60900 }, { "epoch": 5.05, "learning_rate": 1.7548919368246052e-05, "loss": 0.0819, "step": 61000 }, { "epoch": 5.05, "eval_loss": 0.13008062541484833, "eval_runtime": 4421.0264, "eval_samples_per_second": 17.099, "eval_steps_per_second": 1.425, "eval_wer": 0.06467836203328918, "step": 61000 }, { "epoch": 5.06, "learning_rate": 1.754476309226933e-05, "loss": 0.0749, "step": 61100 }, { "epoch": 5.07, "learning_rate": 1.7540606816292603e-05, "loss": 0.0867, "step": 61200 }, { "epoch": 5.07, "learning_rate": 1.753645054031588e-05, "loss": 0.07, "step": 61300 }, { "epoch": 5.08, "learning_rate": 1.7532294264339154e-05, "loss": 0.0744, "step": 61400 }, { "epoch": 5.09, "learning_rate": 1.7528137988362428e-05, "loss": 0.0767, "step": 61500 }, { "epoch": 5.09, "eval_loss": 0.13288891315460205, "eval_runtime": 4361.8552, "eval_samples_per_second": 17.331, "eval_steps_per_second": 1.444, "eval_wer": 0.06502451534689947, "step": 61500 }, { "epoch": 5.1, "learning_rate": 1.7523981712385705e-05, "loss": 0.0742, "step": 61600 }, { "epoch": 5.11, "learning_rate": 1.751982543640898e-05, "loss": 0.0763, "step": 61700 }, { "epoch": 5.12, "learning_rate": 1.7515669160432253e-05, "loss": 0.081, "step": 61800 }, { "epoch": 5.12, "learning_rate": 1.751151288445553e-05, "loss": 0.075, "step": 61900 }, { "epoch": 5.13, "learning_rate": 1.7507356608478804e-05, "loss": 0.0726, "step": 62000 }, { "epoch": 5.13, "eval_loss": 0.13207046687602997, "eval_runtime": 4292.8444, "eval_samples_per_second": 17.61, "eval_steps_per_second": 1.468, "eval_wer": 0.06535682252796535, "step": 62000 }, { "epoch": 5.14, "learning_rate": 1.750320033250208e-05, "loss": 0.0802, "step": 62100 }, { "epoch": 5.15, "learning_rate": 1.7499044056525355e-05, "loss": 0.0763, "step": 62200 }, { "epoch": 5.16, "learning_rate": 1.7494887780548632e-05, "loss": 0.0819, "step": 62300 }, { "epoch": 5.17, "learning_rate": 1.7490731504571906e-05, "loss": 0.0776, "step": 62400 }, { "epoch": 5.17, "learning_rate": 1.748657522859518e-05, "loss": 0.0767, "step": 62500 }, { "epoch": 5.17, "eval_loss": 0.12808410823345184, "eval_runtime": 4321.6201, "eval_samples_per_second": 17.492, "eval_steps_per_second": 1.458, "eval_wer": 0.06682297411850135, "step": 62500 }, { "epoch": 5.18, "learning_rate": 1.7482418952618457e-05, "loss": 0.0744, "step": 62600 }, { "epoch": 5.19, "learning_rate": 1.747826267664173e-05, "loss": 0.0831, "step": 62700 }, { "epoch": 5.2, "learning_rate": 1.7474106400665004e-05, "loss": 0.0722, "step": 62800 }, { "epoch": 5.21, "learning_rate": 1.7469950124688278e-05, "loss": 0.0772, "step": 62900 }, { "epoch": 5.22, "learning_rate": 1.7465793848711555e-05, "loss": 0.0749, "step": 63000 }, { "epoch": 5.22, "eval_loss": 0.1254168301820755, "eval_runtime": 4295.1312, "eval_samples_per_second": 17.6, "eval_steps_per_second": 1.467, "eval_wer": 0.06311836443328549, "step": 63000 }, { "epoch": 5.22, "learning_rate": 1.746163757273483e-05, "loss": 0.0734, "step": 63100 }, { "epoch": 5.23, "learning_rate": 1.7457481296758106e-05, "loss": 0.0743, "step": 63200 }, { "epoch": 5.24, "learning_rate": 1.7453325020781384e-05, "loss": 0.0769, "step": 63300 }, { "epoch": 5.25, "learning_rate": 1.7449168744804657e-05, "loss": 0.075, "step": 63400 }, { "epoch": 5.26, "learning_rate": 1.744501246882793e-05, "loss": 0.0782, "step": 63500 }, { "epoch": 5.26, "eval_loss": 0.11483746021986008, "eval_runtime": 4289.7785, "eval_samples_per_second": 17.622, "eval_steps_per_second": 1.469, "eval_wer": 0.06058298371848659, "step": 63500 }, { "epoch": 5.26, "learning_rate": 1.744085619285121e-05, "loss": 0.0841, "step": 63600 }, { "epoch": 5.27, "learning_rate": 1.7436699916874482e-05, "loss": 0.0772, "step": 63700 }, { "epoch": 5.28, "learning_rate": 1.7432543640897756e-05, "loss": 0.0844, "step": 63800 }, { "epoch": 5.29, "learning_rate": 1.742838736492103e-05, "loss": 0.0781, "step": 63900 }, { "epoch": 5.3, "learning_rate": 1.7424231088944307e-05, "loss": 0.0777, "step": 64000 }, { "epoch": 5.3, "eval_loss": 0.12916968762874603, "eval_runtime": 4318.31, "eval_samples_per_second": 17.506, "eval_steps_per_second": 1.459, "eval_wer": 0.06412297827234112, "step": 64000 }, { "epoch": 5.31, "learning_rate": 1.742007481296758e-05, "loss": 0.0788, "step": 64100 }, { "epoch": 5.31, "learning_rate": 1.7415918536990858e-05, "loss": 0.0752, "step": 64200 }, { "epoch": 5.32, "learning_rate": 1.7411762261014132e-05, "loss": 0.084, "step": 64300 }, { "epoch": 5.33, "learning_rate": 1.740760598503741e-05, "loss": 0.0782, "step": 64400 }, { "epoch": 5.34, "learning_rate": 1.740349127182045e-05, "loss": 0.0867, "step": 64500 }, { "epoch": 5.34, "eval_loss": 0.12179255485534668, "eval_runtime": 4304.3087, "eval_samples_per_second": 17.563, "eval_steps_per_second": 1.464, "eval_wer": 0.06442605472914657, "step": 64500 }, { "epoch": 5.35, "learning_rate": 1.7399334995843725e-05, "loss": 0.0797, "step": 64600 }, { "epoch": 5.36, "learning_rate": 1.7395178719867002e-05, "loss": 0.0844, "step": 64700 }, { "epoch": 5.36, "learning_rate": 1.7391022443890276e-05, "loss": 0.081, "step": 64800 }, { "epoch": 5.37, "learning_rate": 1.7386866167913553e-05, "loss": 0.0779, "step": 64900 }, { "epoch": 5.38, "learning_rate": 1.7382709891936827e-05, "loss": 0.0731, "step": 65000 }, { "epoch": 5.38, "eval_loss": 0.13472488522529602, "eval_runtime": 4292.7525, "eval_samples_per_second": 17.61, "eval_steps_per_second": 1.468, "eval_wer": 0.06532913026287653, "step": 65000 }, { "epoch": 5.39, "learning_rate": 1.73785536159601e-05, "loss": 0.0789, "step": 65100 }, { "epoch": 5.4, "learning_rate": 1.7374397339983375e-05, "loss": 0.0798, "step": 65200 }, { "epoch": 5.41, "learning_rate": 1.7370241064006652e-05, "loss": 0.0706, "step": 65300 }, { "epoch": 5.41, "learning_rate": 1.7366084788029926e-05, "loss": 0.0709, "step": 65400 }, { "epoch": 5.42, "learning_rate": 1.73619285120532e-05, "loss": 0.0791, "step": 65500 }, { "epoch": 5.42, "eval_loss": 0.12497570365667343, "eval_runtime": 4294.628, "eval_samples_per_second": 17.602, "eval_steps_per_second": 1.467, "eval_wer": 0.06109375216345821, "step": 65500 }, { "epoch": 5.43, "learning_rate": 1.7357772236076477e-05, "loss": 0.0756, "step": 65600 }, { "epoch": 5.44, "learning_rate": 1.7353615960099754e-05, "loss": 0.0719, "step": 65700 }, { "epoch": 5.45, "learning_rate": 1.7349459684123028e-05, "loss": 0.0787, "step": 65800 }, { "epoch": 5.46, "learning_rate": 1.7345303408146305e-05, "loss": 0.0749, "step": 65900 }, { "epoch": 5.46, "learning_rate": 1.734114713216958e-05, "loss": 0.0781, "step": 66000 }, { "epoch": 5.46, "eval_loss": 0.12787577509880066, "eval_runtime": 4290.5828, "eval_samples_per_second": 17.619, "eval_steps_per_second": 1.468, "eval_wer": 0.06468297741080398, "step": 66000 }, { "epoch": 5.47, "learning_rate": 1.7336990856192853e-05, "loss": 0.0753, "step": 66100 }, { "epoch": 5.48, "learning_rate": 1.7332834580216126e-05, "loss": 0.0829, "step": 66200 }, { "epoch": 5.49, "learning_rate": 1.7328678304239404e-05, "loss": 0.0808, "step": 66300 }, { "epoch": 5.5, "learning_rate": 1.7324522028262677e-05, "loss": 0.0734, "step": 66400 }, { "epoch": 5.5, "learning_rate": 1.732036575228595e-05, "loss": 0.0693, "step": 66500 }, { "epoch": 5.5, "eval_loss": 0.11359023302793503, "eval_runtime": 4300.2156, "eval_samples_per_second": 17.579, "eval_steps_per_second": 1.465, "eval_wer": 0.060736829635646715, "step": 66500 }, { "epoch": 5.51, "learning_rate": 1.731620947630923e-05, "loss": 0.079, "step": 66600 }, { "epoch": 5.52, "learning_rate": 1.731209476309227e-05, "loss": 0.0743, "step": 66700 }, { "epoch": 5.53, "learning_rate": 1.7307938487115544e-05, "loss": 0.074, "step": 66800 }, { "epoch": 5.54, "learning_rate": 1.730378221113882e-05, "loss": 0.0735, "step": 66900 }, { "epoch": 5.55, "learning_rate": 1.7299625935162095e-05, "loss": 0.0819, "step": 67000 }, { "epoch": 5.55, "eval_loss": 0.12357547134160995, "eval_runtime": 4294.4673, "eval_samples_per_second": 17.603, "eval_steps_per_second": 1.467, "eval_wer": 0.06294759546523775, "step": 67000 }, { "epoch": 5.55, "learning_rate": 1.7295469659185372e-05, "loss": 0.0762, "step": 67100 }, { "epoch": 5.56, "learning_rate": 1.7291313383208646e-05, "loss": 0.0737, "step": 67200 }, { "epoch": 5.57, "learning_rate": 1.7287157107231923e-05, "loss": 0.0745, "step": 67300 }, { "epoch": 5.58, "learning_rate": 1.7283000831255197e-05, "loss": 0.0804, "step": 67400 }, { "epoch": 5.59, "learning_rate": 1.727884455527847e-05, "loss": 0.0726, "step": 67500 }, { "epoch": 5.59, "eval_loss": 0.11987119913101196, "eval_runtime": 4281.6674, "eval_samples_per_second": 17.656, "eval_steps_per_second": 1.471, "eval_wer": 0.06187375096346005, "step": 67500 }, { "epoch": 5.6, "learning_rate": 1.7274688279301748e-05, "loss": 0.0706, "step": 67600 }, { "epoch": 5.6, "learning_rate": 1.7270532003325022e-05, "loss": 0.0817, "step": 67700 }, { "epoch": 5.61, "learning_rate": 1.7266375727348296e-05, "loss": 0.0752, "step": 67800 }, { "epoch": 5.62, "learning_rate": 1.7262219451371573e-05, "loss": 0.0747, "step": 67900 }, { "epoch": 5.63, "learning_rate": 1.7258063175394847e-05, "loss": 0.0792, "step": 68000 }, { "epoch": 5.63, "eval_loss": 0.12620863318443298, "eval_runtime": 4283.3813, "eval_samples_per_second": 17.648, "eval_steps_per_second": 1.471, "eval_wer": 0.06384144024393809, "step": 68000 }, { "epoch": 5.64, "learning_rate": 1.7253906899418124e-05, "loss": 0.0734, "step": 68100 }, { "epoch": 5.65, "learning_rate": 1.7249750623441398e-05, "loss": 0.0745, "step": 68200 }, { "epoch": 5.65, "learning_rate": 1.7245594347464675e-05, "loss": 0.0766, "step": 68300 }, { "epoch": 5.66, "learning_rate": 1.724143807148795e-05, "loss": 0.0769, "step": 68400 }, { "epoch": 5.67, "learning_rate": 1.7237281795511223e-05, "loss": 0.0728, "step": 68500 }, { "epoch": 5.67, "eval_loss": 0.12456958740949631, "eval_runtime": 4406.3603, "eval_samples_per_second": 17.156, "eval_steps_per_second": 1.43, "eval_wer": 0.063012210750445, "step": 68500 }, { "epoch": 5.68, "learning_rate": 1.72331255195345e-05, "loss": 0.0788, "step": 68600 }, { "epoch": 5.69, "learning_rate": 1.7228969243557774e-05, "loss": 0.0826, "step": 68700 }, { "epoch": 5.7, "learning_rate": 1.7224812967581048e-05, "loss": 0.0977, "step": 68800 }, { "epoch": 5.7, "learning_rate": 1.7220656691604325e-05, "loss": 0.0794, "step": 68900 }, { "epoch": 5.71, "learning_rate": 1.72165004156276e-05, "loss": 0.0785, "step": 69000 }, { "epoch": 5.71, "eval_loss": 0.12336178869009018, "eval_runtime": 4384.0636, "eval_samples_per_second": 17.243, "eval_steps_per_second": 1.437, "eval_wer": 0.06271221121198275, "step": 69000 }, { "epoch": 5.72, "learning_rate": 1.7212344139650872e-05, "loss": 0.0771, "step": 69100 }, { "epoch": 5.73, "learning_rate": 1.720818786367415e-05, "loss": 0.0755, "step": 69200 }, { "epoch": 5.74, "learning_rate": 1.7204031587697423e-05, "loss": 0.0746, "step": 69300 }, { "epoch": 5.75, "learning_rate": 1.71998753117207e-05, "loss": 0.0747, "step": 69400 }, { "epoch": 5.75, "learning_rate": 1.7195719035743975e-05, "loss": 0.0745, "step": 69500 }, { "epoch": 5.75, "eval_loss": 0.1184028834104538, "eval_runtime": 4387.3137, "eval_samples_per_second": 17.23, "eval_steps_per_second": 1.436, "eval_wer": 0.06272913426287037, "step": 69500 }, { "epoch": 5.76, "learning_rate": 1.7191604322527016e-05, "loss": 0.0711, "step": 69600 }, { "epoch": 5.77, "learning_rate": 1.7187448046550294e-05, "loss": 0.0763, "step": 69700 }, { "epoch": 5.78, "learning_rate": 1.7183291770573567e-05, "loss": 0.0831, "step": 69800 }, { "epoch": 5.79, "learning_rate": 1.7179135494596845e-05, "loss": 0.0815, "step": 69900 }, { "epoch": 5.79, "learning_rate": 1.717497921862012e-05, "loss": 0.0734, "step": 70000 }, { "epoch": 5.79, "eval_loss": 0.1255302131175995, "eval_runtime": 4383.9582, "eval_samples_per_second": 17.244, "eval_steps_per_second": 1.437, "eval_wer": 0.06284759561908367, "step": 70000 }, { "epoch": 5.8, "learning_rate": 1.7170822942643392e-05, "loss": 0.0776, "step": 70100 }, { "epoch": 5.81, "learning_rate": 1.7166666666666666e-05, "loss": 0.0754, "step": 70200 }, { "epoch": 5.82, "learning_rate": 1.7162510390689943e-05, "loss": 0.0746, "step": 70300 }, { "epoch": 5.83, "learning_rate": 1.7158354114713217e-05, "loss": 0.0769, "step": 70400 }, { "epoch": 5.84, "learning_rate": 1.7154197838736494e-05, "loss": 0.0743, "step": 70500 }, { "epoch": 5.84, "eval_loss": 0.1268174648284912, "eval_runtime": 4370.181, "eval_samples_per_second": 17.298, "eval_steps_per_second": 1.442, "eval_wer": 0.06152605919067817, "step": 70500 }, { "epoch": 5.84, "learning_rate": 1.7150041562759768e-05, "loss": 0.0741, "step": 70600 }, { "epoch": 5.85, "learning_rate": 1.7145885286783045e-05, "loss": 0.0764, "step": 70700 }, { "epoch": 5.86, "learning_rate": 1.714172901080632e-05, "loss": 0.07, "step": 70800 }, { "epoch": 5.87, "learning_rate": 1.7137572734829596e-05, "loss": 0.0741, "step": 70900 }, { "epoch": 5.88, "learning_rate": 1.713341645885287e-05, "loss": 0.0819, "step": 71000 }, { "epoch": 5.88, "eval_loss": 0.12211538106203079, "eval_runtime": 4366.281, "eval_samples_per_second": 17.313, "eval_steps_per_second": 1.443, "eval_wer": 0.06324144116701359, "step": 71000 }, { "epoch": 5.89, "learning_rate": 1.7129260182876144e-05, "loss": 0.0807, "step": 71100 }, { "epoch": 5.89, "learning_rate": 1.7125103906899418e-05, "loss": 0.0759, "step": 71200 }, { "epoch": 5.9, "learning_rate": 1.7120947630922695e-05, "loss": 0.0782, "step": 71300 }, { "epoch": 5.91, "learning_rate": 1.711679135494597e-05, "loss": 0.0745, "step": 71400 }, { "epoch": 5.92, "learning_rate": 1.7112635078969243e-05, "loss": 0.0745, "step": 71500 }, { "epoch": 5.92, "eval_loss": 0.12614519894123077, "eval_runtime": 4300.9661, "eval_samples_per_second": 17.576, "eval_steps_per_second": 1.465, "eval_wer": 0.06363682517411512, "step": 71500 }, { "epoch": 5.93, "learning_rate": 1.710847880299252e-05, "loss": 0.075, "step": 71600 }, { "epoch": 5.94, "learning_rate": 1.7104322527015794e-05, "loss": 0.0773, "step": 71700 }, { "epoch": 5.94, "learning_rate": 1.7100207813798836e-05, "loss": 0.0753, "step": 71800 }, { "epoch": 5.95, "learning_rate": 1.7096051537822113e-05, "loss": 0.075, "step": 71900 }, { "epoch": 5.96, "learning_rate": 1.7091895261845387e-05, "loss": 0.0684, "step": 72000 }, { "epoch": 5.96, "eval_loss": 0.12664571404457092, "eval_runtime": 4347.835, "eval_samples_per_second": 17.387, "eval_steps_per_second": 1.449, "eval_wer": 0.061713751209613525, "step": 72000 }, { "epoch": 5.97, "learning_rate": 1.7087738985868664e-05, "loss": 0.0732, "step": 72100 }, { "epoch": 5.98, "learning_rate": 1.7083582709891938e-05, "loss": 0.0787, "step": 72200 }, { "epoch": 5.99, "learning_rate": 1.7079426433915215e-05, "loss": 0.0734, "step": 72300 }, { "epoch": 5.99, "learning_rate": 1.707527015793849e-05, "loss": 0.0787, "step": 72400 }, { "epoch": 6.0, "learning_rate": 1.7071113881961762e-05, "loss": 0.0692, "step": 72500 }, { "epoch": 6.0, "eval_loss": 0.1245121881365776, "eval_runtime": 4296.7356, "eval_samples_per_second": 17.594, "eval_steps_per_second": 1.466, "eval_wer": 0.062041443013164596, "step": 72500 }, { "epoch": 6.01, "learning_rate": 1.706695760598504e-05, "loss": 0.0744, "step": 72600 }, { "epoch": 6.02, "learning_rate": 1.7062801330008314e-05, "loss": 0.0706, "step": 72700 }, { "epoch": 6.03, "learning_rate": 1.7058645054031587e-05, "loss": 0.0715, "step": 72800 }, { "epoch": 6.03, "learning_rate": 1.7054488778054865e-05, "loss": 0.0669, "step": 72900 }, { "epoch": 6.04, "learning_rate": 1.705033250207814e-05, "loss": 0.0649, "step": 73000 }, { "epoch": 6.04, "eval_loss": 0.12877026200294495, "eval_runtime": 4284.8763, "eval_samples_per_second": 17.642, "eval_steps_per_second": 1.47, "eval_wer": 0.06282605719068124, "step": 73000 }, { "epoch": 6.05, "learning_rate": 1.7046176226101416e-05, "loss": 0.0627, "step": 73100 }, { "epoch": 6.06, "learning_rate": 1.704201995012469e-05, "loss": 0.0799, "step": 73200 }, { "epoch": 6.07, "learning_rate": 1.7037863674147967e-05, "loss": 0.0711, "step": 73300 }, { "epoch": 6.08, "learning_rate": 1.703370739817124e-05, "loss": 0.0742, "step": 73400 }, { "epoch": 6.08, "learning_rate": 1.7029551122194514e-05, "loss": 0.0661, "step": 73500 }, { "epoch": 6.08, "eval_loss": 0.12214105576276779, "eval_runtime": 4275.1854, "eval_samples_per_second": 17.682, "eval_steps_per_second": 1.474, "eval_wer": 0.06225221191967397, "step": 73500 }, { "epoch": 6.09, "learning_rate": 1.702539484621779e-05, "loss": 0.0697, "step": 73600 }, { "epoch": 6.1, "learning_rate": 1.7021238570241065e-05, "loss": 0.0726, "step": 73700 }, { "epoch": 6.11, "learning_rate": 1.701708229426434e-05, "loss": 0.0702, "step": 73800 }, { "epoch": 6.12, "learning_rate": 1.7012926018287616e-05, "loss": 0.0653, "step": 73900 }, { "epoch": 6.13, "learning_rate": 1.700876974231089e-05, "loss": 0.0761, "step": 74000 }, { "epoch": 6.13, "eval_loss": 0.11997722834348679, "eval_runtime": 4282.3114, "eval_samples_per_second": 17.653, "eval_steps_per_second": 1.471, "eval_wer": 0.06376605574452962, "step": 74000 }, { "epoch": 6.13, "learning_rate": 1.7004613466334164e-05, "loss": 0.0705, "step": 74100 }, { "epoch": 6.14, "learning_rate": 1.700045719035744e-05, "loss": 0.0664, "step": 74200 }, { "epoch": 6.15, "learning_rate": 1.699630091438072e-05, "loss": 0.0734, "step": 74300 }, { "epoch": 6.16, "learning_rate": 1.6992144638403992e-05, "loss": 0.0646, "step": 74400 }, { "epoch": 6.17, "learning_rate": 1.6987988362427266e-05, "loss": 0.0629, "step": 74500 }, { "epoch": 6.17, "eval_loss": 0.11546548455953598, "eval_runtime": 4290.2654, "eval_samples_per_second": 17.62, "eval_steps_per_second": 1.468, "eval_wer": 0.060479906953989304, "step": 74500 }, { "epoch": 6.18, "learning_rate": 1.6983832086450543e-05, "loss": 0.0703, "step": 74600 }, { "epoch": 6.18, "learning_rate": 1.6979675810473817e-05, "loss": 0.0723, "step": 74700 }, { "epoch": 6.19, "learning_rate": 1.697551953449709e-05, "loss": 0.0698, "step": 74800 }, { "epoch": 6.2, "learning_rate": 1.6971363258520368e-05, "loss": 0.0753, "step": 74900 }, { "epoch": 6.21, "learning_rate": 1.6967206982543642e-05, "loss": 0.0718, "step": 75000 }, { "epoch": 6.21, "eval_loss": 0.11722259968519211, "eval_runtime": 4299.5817, "eval_samples_per_second": 17.582, "eval_steps_per_second": 1.465, "eval_wer": 0.06084913715517361, "step": 75000 }, { "epoch": 6.22, "learning_rate": 1.6963050706566916e-05, "loss": 0.0647, "step": 75100 }, { "epoch": 6.23, "learning_rate": 1.6958894430590193e-05, "loss": 0.0766, "step": 75200 }, { "epoch": 6.23, "learning_rate": 1.6954738154613467e-05, "loss": 0.0726, "step": 75300 }, { "epoch": 6.24, "learning_rate": 1.6950581878636744e-05, "loss": 0.0648, "step": 75400 }, { "epoch": 6.25, "learning_rate": 1.6946425602660018e-05, "loss": 0.071, "step": 75500 }, { "epoch": 6.25, "eval_loss": 0.13009411096572876, "eval_runtime": 4275.2697, "eval_samples_per_second": 17.682, "eval_steps_per_second": 1.474, "eval_wer": 0.06453836224867346, "step": 75500 }, { "epoch": 6.26, "learning_rate": 1.6942269326683295e-05, "loss": 0.0704, "step": 75600 }, { "epoch": 6.27, "learning_rate": 1.693811305070657e-05, "loss": 0.0749, "step": 75700 }, { "epoch": 6.27, "learning_rate": 1.6933956774729843e-05, "loss": 0.0738, "step": 75800 }, { "epoch": 6.28, "learning_rate": 1.692980049875312e-05, "loss": 0.0709, "step": 75900 }, { "epoch": 6.29, "learning_rate": 1.6925644222776394e-05, "loss": 0.0745, "step": 76000 }, { "epoch": 6.29, "eval_loss": 0.1395825296640396, "eval_runtime": 4306.4208, "eval_samples_per_second": 17.554, "eval_steps_per_second": 1.463, "eval_wer": 0.0661537443788548, "step": 76000 }, { "epoch": 6.3, "learning_rate": 1.6921487946799668e-05, "loss": 0.0712, "step": 76100 }, { "epoch": 6.31, "learning_rate": 1.691733167082294e-05, "loss": 0.0814, "step": 76200 }, { "epoch": 6.32, "learning_rate": 1.691317539484622e-05, "loss": 0.0725, "step": 76300 }, { "epoch": 6.32, "learning_rate": 1.6909019118869496e-05, "loss": 0.0767, "step": 76400 }, { "epoch": 6.33, "learning_rate": 1.690486284289277e-05, "loss": 0.0708, "step": 76500 }, { "epoch": 6.33, "eval_loss": 0.1320749670267105, "eval_runtime": 4280.6322, "eval_samples_per_second": 17.66, "eval_steps_per_second": 1.472, "eval_wer": 0.06319374893269394, "step": 76500 }, { "epoch": 6.34, "learning_rate": 1.6900706566916047e-05, "loss": 0.0701, "step": 76600 }, { "epoch": 6.35, "learning_rate": 1.689655029093932e-05, "loss": 0.0703, "step": 76700 }, { "epoch": 6.36, "learning_rate": 1.6892394014962594e-05, "loss": 0.0675, "step": 76800 }, { "epoch": 6.37, "learning_rate": 1.688823773898587e-05, "loss": 0.062, "step": 76900 }, { "epoch": 6.37, "learning_rate": 1.6884123025768913e-05, "loss": 0.0678, "step": 77000 }, { "epoch": 6.37, "eval_loss": 0.14557278156280518, "eval_runtime": 4284.2979, "eval_samples_per_second": 17.645, "eval_steps_per_second": 1.47, "eval_wer": 0.06619835969483123, "step": 77000 }, { "epoch": 6.38, "learning_rate": 1.6879966749792187e-05, "loss": 0.066, "step": 77100 }, { "epoch": 6.39, "learning_rate": 1.6875810473815464e-05, "loss": 0.0705, "step": 77200 }, { "epoch": 6.4, "learning_rate": 1.687165419783874e-05, "loss": 0.0687, "step": 77300 }, { "epoch": 6.41, "learning_rate": 1.6867497921862012e-05, "loss": 0.067, "step": 77400 }, { "epoch": 6.42, "learning_rate": 1.6863341645885286e-05, "loss": 0.0729, "step": 77500 }, { "epoch": 6.42, "eval_loss": 0.1397981494665146, "eval_runtime": 4293.6174, "eval_samples_per_second": 17.606, "eval_steps_per_second": 1.467, "eval_wer": 0.06638605171376659, "step": 77500 }, { "epoch": 6.42, "learning_rate": 1.6859185369908563e-05, "loss": 0.0895, "step": 77600 }, { "epoch": 6.43, "learning_rate": 1.6855029093931837e-05, "loss": 0.0752, "step": 77700 }, { "epoch": 6.44, "learning_rate": 1.6850872817955114e-05, "loss": 0.0725, "step": 77800 }, { "epoch": 6.45, "learning_rate": 1.6846716541978388e-05, "loss": 0.0734, "step": 77900 }, { "epoch": 6.46, "learning_rate": 1.6842560266001665e-05, "loss": 0.0672, "step": 78000 }, { "epoch": 6.46, "eval_loss": 0.1389104425907135, "eval_runtime": 4294.3084, "eval_samples_per_second": 17.604, "eval_steps_per_second": 1.467, "eval_wer": 0.0656691297398004, "step": 78000 }, { "epoch": 6.47, "learning_rate": 1.683840399002494e-05, "loss": 0.068, "step": 78100 }, { "epoch": 6.47, "learning_rate": 1.6834247714048216e-05, "loss": 0.0666, "step": 78200 }, { "epoch": 6.48, "learning_rate": 1.683009143807149e-05, "loss": 0.0712, "step": 78300 }, { "epoch": 6.49, "learning_rate": 1.6825935162094764e-05, "loss": 0.08, "step": 78400 }, { "epoch": 6.5, "learning_rate": 1.6821778886118038e-05, "loss": 0.0785, "step": 78500 }, { "epoch": 6.5, "eval_loss": 0.1259874850511551, "eval_runtime": 4334.9361, "eval_samples_per_second": 17.439, "eval_steps_per_second": 1.453, "eval_wer": 0.06345374853269456, "step": 78500 }, { "epoch": 6.51, "learning_rate": 1.6817622610141315e-05, "loss": 0.063, "step": 78600 }, { "epoch": 6.51, "learning_rate": 1.681346633416459e-05, "loss": 0.0668, "step": 78700 }, { "epoch": 6.52, "learning_rate": 1.6809310058187866e-05, "loss": 0.0655, "step": 78800 }, { "epoch": 6.53, "learning_rate": 1.680515378221114e-05, "loss": 0.065, "step": 78900 }, { "epoch": 6.54, "learning_rate": 1.6800997506234417e-05, "loss": 0.0677, "step": 79000 }, { "epoch": 6.54, "eval_loss": 0.13370338082313538, "eval_runtime": 4335.7123, "eval_samples_per_second": 17.435, "eval_steps_per_second": 1.453, "eval_wer": 0.061766058821447964, "step": 79000 }, { "epoch": 6.55, "learning_rate": 1.679684123025769e-05, "loss": 0.0709, "step": 79100 }, { "epoch": 6.56, "learning_rate": 1.6792684954280968e-05, "loss": 0.0676, "step": 79200 }, { "epoch": 6.56, "learning_rate": 1.6788528678304242e-05, "loss": 0.0663, "step": 79300 }, { "epoch": 6.57, "learning_rate": 1.6784372402327516e-05, "loss": 0.0701, "step": 79400 }, { "epoch": 6.58, "learning_rate": 1.678021612635079e-05, "loss": 0.0671, "step": 79500 }, { "epoch": 6.58, "eval_loss": 0.13135309517383575, "eval_runtime": 4345.9396, "eval_samples_per_second": 17.394, "eval_steps_per_second": 1.45, "eval_wer": 0.0638660555906837, "step": 79500 }, { "epoch": 6.59, "learning_rate": 1.6776059850374067e-05, "loss": 0.0746, "step": 79600 }, { "epoch": 6.6, "learning_rate": 1.677190357439734e-05, "loss": 0.0824, "step": 79700 }, { "epoch": 6.61, "learning_rate": 1.6767747298420614e-05, "loss": 0.0705, "step": 79800 }, { "epoch": 6.61, "learning_rate": 1.676359102244389e-05, "loss": 0.0775, "step": 79900 }, { "epoch": 6.62, "learning_rate": 1.6759434746467165e-05, "loss": 0.0725, "step": 80000 }, { "epoch": 6.62, "eval_loss": 0.1203729435801506, "eval_runtime": 4337.4553, "eval_samples_per_second": 17.428, "eval_steps_per_second": 1.452, "eval_wer": 0.06016298436463944, "step": 80000 }, { "epoch": 6.63, "learning_rate": 1.6755278470490443e-05, "loss": 0.0633, "step": 80100 }, { "epoch": 6.64, "learning_rate": 1.6751163757273484e-05, "loss": 0.071, "step": 80200 }, { "epoch": 6.65, "learning_rate": 1.6747007481296758e-05, "loss": 0.065, "step": 80300 }, { "epoch": 6.66, "learning_rate": 1.6742851205320035e-05, "loss": 0.0626, "step": 80400 }, { "epoch": 6.66, "learning_rate": 1.673869492934331e-05, "loss": 0.0686, "step": 80500 }, { "epoch": 6.66, "eval_loss": 0.127023845911026, "eval_runtime": 4327.3599, "eval_samples_per_second": 17.469, "eval_steps_per_second": 1.456, "eval_wer": 0.060287599557539144, "step": 80500 }, { "epoch": 6.67, "learning_rate": 1.6734538653366587e-05, "loss": 0.0741, "step": 80600 }, { "epoch": 6.68, "learning_rate": 1.673038237738986e-05, "loss": 0.0775, "step": 80700 }, { "epoch": 6.69, "learning_rate": 1.6726226101413134e-05, "loss": 0.0711, "step": 80800 }, { "epoch": 6.7, "learning_rate": 1.672206982543641e-05, "loss": 0.0773, "step": 80900 }, { "epoch": 6.71, "learning_rate": 1.6717913549459685e-05, "loss": 0.0672, "step": 81000 }, { "epoch": 6.71, "eval_loss": 0.12983781099319458, "eval_runtime": 4304.9899, "eval_samples_per_second": 17.56, "eval_steps_per_second": 1.463, "eval_wer": 0.06151682843564856, "step": 81000 }, { "epoch": 6.71, "learning_rate": 1.671375727348296e-05, "loss": 0.0661, "step": 81100 }, { "epoch": 6.72, "learning_rate": 1.6709600997506236e-05, "loss": 0.0674, "step": 81200 }, { "epoch": 6.73, "learning_rate": 1.670544472152951e-05, "loss": 0.0779, "step": 81300 }, { "epoch": 6.74, "learning_rate": 1.6701288445552787e-05, "loss": 0.069, "step": 81400 }, { "epoch": 6.75, "learning_rate": 1.669713216957606e-05, "loss": 0.0774, "step": 81500 }, { "epoch": 6.75, "eval_loss": 0.13896754384040833, "eval_runtime": 4304.8388, "eval_samples_per_second": 17.56, "eval_steps_per_second": 1.463, "eval_wer": 0.0649383616332898, "step": 81500 }, { "epoch": 6.75, "learning_rate": 1.6692975893599338e-05, "loss": 0.0701, "step": 81600 }, { "epoch": 6.76, "learning_rate": 1.6688819617622612e-05, "loss": 0.0713, "step": 81700 }, { "epoch": 6.77, "learning_rate": 1.6684663341645886e-05, "loss": 0.0666, "step": 81800 }, { "epoch": 6.78, "learning_rate": 1.6680507065669163e-05, "loss": 0.068, "step": 81900 }, { "epoch": 6.79, "learning_rate": 1.6676350789692437e-05, "loss": 0.0694, "step": 82000 }, { "epoch": 6.79, "eval_loss": 0.12394601851701736, "eval_runtime": 4285.5543, "eval_samples_per_second": 17.639, "eval_steps_per_second": 1.47, "eval_wer": 0.06193836624866731, "step": 82000 }, { "epoch": 6.8, "learning_rate": 1.667219451371571e-05, "loss": 0.0684, "step": 82100 }, { "epoch": 6.8, "learning_rate": 1.6668038237738988e-05, "loss": 0.0693, "step": 82200 }, { "epoch": 6.81, "learning_rate": 1.6663881961762262e-05, "loss": 0.0699, "step": 82300 }, { "epoch": 6.82, "learning_rate": 1.6659725685785536e-05, "loss": 0.0721, "step": 82400 }, { "epoch": 6.83, "learning_rate": 1.6655569409808813e-05, "loss": 0.0689, "step": 82500 }, { "epoch": 6.83, "eval_loss": 0.13068170845508575, "eval_runtime": 4277.9472, "eval_samples_per_second": 17.671, "eval_steps_per_second": 1.473, "eval_wer": 0.06275990344630239, "step": 82500 }, { "epoch": 6.84, "learning_rate": 1.6651413133832087e-05, "loss": 0.0744, "step": 82600 }, { "epoch": 6.85, "learning_rate": 1.6647256857855364e-05, "loss": 0.069, "step": 82700 }, { "epoch": 6.85, "learning_rate": 1.6643100581878638e-05, "loss": 0.0709, "step": 82800 }, { "epoch": 6.86, "learning_rate": 1.6638944305901915e-05, "loss": 0.0688, "step": 82900 }, { "epoch": 6.87, "learning_rate": 1.663478802992519e-05, "loss": 0.0693, "step": 83000 }, { "epoch": 6.87, "eval_loss": 0.12454091757535934, "eval_runtime": 4290.9182, "eval_samples_per_second": 17.617, "eval_steps_per_second": 1.468, "eval_wer": 0.06068144510546907, "step": 83000 }, { "epoch": 6.88, "learning_rate": 1.6630631753948463e-05, "loss": 0.0757, "step": 83100 }, { "epoch": 6.89, "learning_rate": 1.662647547797174e-05, "loss": 0.0715, "step": 83200 }, { "epoch": 6.9, "learning_rate": 1.6622319201995014e-05, "loss": 0.0724, "step": 83300 }, { "epoch": 6.9, "learning_rate": 1.6618162926018287e-05, "loss": 0.0665, "step": 83400 }, { "epoch": 6.91, "learning_rate": 1.6614006650041565e-05, "loss": 0.0669, "step": 83500 }, { "epoch": 6.91, "eval_loss": 0.12757913768291473, "eval_runtime": 4266.981, "eval_samples_per_second": 17.716, "eval_steps_per_second": 1.476, "eval_wer": 0.06224451962381596, "step": 83500 }, { "epoch": 6.92, "learning_rate": 1.6609891936824606e-05, "loss": 0.0667, "step": 83600 }, { "epoch": 6.93, "learning_rate": 1.660573566084788e-05, "loss": 0.0713, "step": 83700 }, { "epoch": 6.94, "learning_rate": 1.6601579384871157e-05, "loss": 0.0654, "step": 83800 }, { "epoch": 6.95, "learning_rate": 1.659742310889443e-05, "loss": 0.0687, "step": 83900 }, { "epoch": 6.95, "learning_rate": 1.659326683291771e-05, "loss": 0.0684, "step": 84000 }, { "epoch": 6.95, "eval_loss": 0.1295813024044037, "eval_runtime": 4265.7789, "eval_samples_per_second": 17.721, "eval_steps_per_second": 1.477, "eval_wer": 0.061209136601328304, "step": 84000 }, { "epoch": 6.96, "learning_rate": 1.6589110556940982e-05, "loss": 0.0715, "step": 84100 }, { "epoch": 6.97, "learning_rate": 1.658495428096426e-05, "loss": 0.0718, "step": 84200 }, { "epoch": 6.98, "learning_rate": 1.6580798004987533e-05, "loss": 0.0625, "step": 84300 }, { "epoch": 6.99, "learning_rate": 1.6576641729010807e-05, "loss": 0.0676, "step": 84400 }, { "epoch": 7.0, "learning_rate": 1.657248545303408e-05, "loss": 0.0656, "step": 84500 }, { "epoch": 7.0, "eval_loss": 0.12673963606357574, "eval_runtime": 4301.2218, "eval_samples_per_second": 17.575, "eval_steps_per_second": 1.465, "eval_wer": 0.06167375127115189, "step": 84500 }, { "epoch": 7.0, "learning_rate": 1.6568329177057358e-05, "loss": 0.0748, "step": 84600 }, { "epoch": 7.01, "learning_rate": 1.6564172901080632e-05, "loss": 0.0718, "step": 84700 }, { "epoch": 7.02, "learning_rate": 1.6560016625103906e-05, "loss": 0.0706, "step": 84800 }, { "epoch": 7.03, "learning_rate": 1.6555860349127183e-05, "loss": 0.0721, "step": 84900 }, { "epoch": 7.04, "learning_rate": 1.6551704073150457e-05, "loss": 0.064, "step": 85000 }, { "epoch": 7.04, "eval_loss": 0.12612000107765198, "eval_runtime": 4302.9753, "eval_samples_per_second": 17.568, "eval_steps_per_second": 1.464, "eval_wer": 0.06113221364274824, "step": 85000 }, { "epoch": 7.04, "learning_rate": 1.6547547797173734e-05, "loss": 0.0659, "step": 85100 }, { "epoch": 7.05, "learning_rate": 1.654339152119701e-05, "loss": 0.0663, "step": 85200 }, { "epoch": 7.06, "learning_rate": 1.6539235245220285e-05, "loss": 0.0636, "step": 85300 }, { "epoch": 7.07, "learning_rate": 1.653507896924356e-05, "loss": 0.0722, "step": 85400 }, { "epoch": 7.08, "learning_rate": 1.6530922693266833e-05, "loss": 0.0637, "step": 85500 }, { "epoch": 7.08, "eval_loss": 0.12599995732307434, "eval_runtime": 4285.9445, "eval_samples_per_second": 17.638, "eval_steps_per_second": 1.47, "eval_wer": 0.06062606057529142, "step": 85500 }, { "epoch": 7.09, "learning_rate": 1.652676641729011e-05, "loss": 0.0692, "step": 85600 }, { "epoch": 7.09, "learning_rate": 1.6522610141313384e-05, "loss": 0.0648, "step": 85700 }, { "epoch": 7.1, "learning_rate": 1.6518495428096426e-05, "loss": 0.062, "step": 85800 }, { "epoch": 7.11, "learning_rate": 1.6514339152119703e-05, "loss": 0.0701, "step": 85900 }, { "epoch": 7.12, "learning_rate": 1.6510182876142977e-05, "loss": 0.0607, "step": 86000 }, { "epoch": 7.12, "eval_loss": 0.12270276993513107, "eval_runtime": 4292.347, "eval_samples_per_second": 17.612, "eval_steps_per_second": 1.468, "eval_wer": 0.06008452294688777, "step": 86000 }, { "epoch": 7.13, "learning_rate": 1.650602660016625e-05, "loss": 0.064, "step": 86100 }, { "epoch": 7.14, "learning_rate": 1.6501870324189528e-05, "loss": 0.0692, "step": 86200 }, { "epoch": 7.14, "learning_rate": 1.64977140482128e-05, "loss": 0.0602, "step": 86300 }, { "epoch": 7.15, "learning_rate": 1.649355777223608e-05, "loss": 0.0622, "step": 86400 }, { "epoch": 7.16, "learning_rate": 1.6489401496259353e-05, "loss": 0.0621, "step": 86500 }, { "epoch": 7.16, "eval_loss": 0.13005945086479187, "eval_runtime": 4279.8179, "eval_samples_per_second": 17.663, "eval_steps_per_second": 1.472, "eval_wer": 0.061452213150441304, "step": 86500 }, { "epoch": 7.17, "learning_rate": 1.648524522028263e-05, "loss": 0.0735, "step": 86600 }, { "epoch": 7.18, "learning_rate": 1.6481088944305904e-05, "loss": 0.0598, "step": 86700 }, { "epoch": 7.19, "learning_rate": 1.6476932668329177e-05, "loss": 0.0647, "step": 86800 }, { "epoch": 7.19, "learning_rate": 1.6472776392352455e-05, "loss": 0.0681, "step": 86900 }, { "epoch": 7.2, "learning_rate": 1.646862011637573e-05, "loss": 0.0669, "step": 87000 }, { "epoch": 7.2, "eval_loss": 0.1312842071056366, "eval_runtime": 4296.4701, "eval_samples_per_second": 17.595, "eval_steps_per_second": 1.466, "eval_wer": 0.06153221302736457, "step": 87000 }, { "epoch": 7.21, "learning_rate": 1.6464463840399002e-05, "loss": 0.0796, "step": 87100 }, { "epoch": 7.22, "learning_rate": 1.646030756442228e-05, "loss": 0.0671, "step": 87200 }, { "epoch": 7.23, "learning_rate": 1.6456151288445553e-05, "loss": 0.0644, "step": 87300 }, { "epoch": 7.24, "learning_rate": 1.6451995012468827e-05, "loss": 0.0809, "step": 87400 }, { "epoch": 7.24, "learning_rate": 1.6447838736492104e-05, "loss": 0.0757, "step": 87500 }, { "epoch": 7.24, "eval_loss": 0.1288526952266693, "eval_runtime": 4307.6036, "eval_samples_per_second": 17.549, "eval_steps_per_second": 1.463, "eval_wer": 0.061786058790678784, "step": 87500 }, { "epoch": 7.25, "learning_rate": 1.644368246051538e-05, "loss": 0.0636, "step": 87600 }, { "epoch": 7.26, "learning_rate": 1.6439526184538655e-05, "loss": 0.0748, "step": 87700 }, { "epoch": 7.27, "learning_rate": 1.643536990856193e-05, "loss": 0.0643, "step": 87800 }, { "epoch": 7.28, "learning_rate": 1.6431213632585206e-05, "loss": 0.0627, "step": 87900 }, { "epoch": 7.28, "learning_rate": 1.642705735660848e-05, "loss": 0.0634, "step": 88000 }, { "epoch": 7.28, "eval_loss": 0.12834565341472626, "eval_runtime": 4284.736, "eval_samples_per_second": 17.643, "eval_steps_per_second": 1.47, "eval_wer": 0.062290673398964, "step": 88000 }, { "epoch": 7.29, "learning_rate": 1.6422901080631754e-05, "loss": 0.0671, "step": 88100 }, { "epoch": 7.3, "learning_rate": 1.64187863674148e-05, "loss": 0.0579, "step": 88200 }, { "epoch": 7.31, "learning_rate": 1.6414630091438073e-05, "loss": 0.0685, "step": 88300 }, { "epoch": 7.32, "learning_rate": 1.6410473815461347e-05, "loss": 0.0649, "step": 88400 }, { "epoch": 7.33, "learning_rate": 1.6406317539484624e-05, "loss": 0.0679, "step": 88500 }, { "epoch": 7.33, "eval_loss": 0.12121399492025375, "eval_runtime": 4268.9585, "eval_samples_per_second": 17.708, "eval_steps_per_second": 1.476, "eval_wer": 0.059496831543336084, "step": 88500 }, { "epoch": 7.33, "learning_rate": 1.6402161263507898e-05, "loss": 0.064, "step": 88600 }, { "epoch": 7.34, "learning_rate": 1.6398004987531172e-05, "loss": 0.0667, "step": 88700 }, { "epoch": 7.35, "learning_rate": 1.639384871155445e-05, "loss": 0.065, "step": 88800 }, { "epoch": 7.36, "learning_rate": 1.638973399833749e-05, "loss": 0.0683, "step": 88900 }, { "epoch": 7.37, "learning_rate": 1.6385577722360765e-05, "loss": 0.0666, "step": 89000 }, { "epoch": 7.37, "eval_loss": 0.12926992774009705, "eval_runtime": 4287.8557, "eval_samples_per_second": 17.63, "eval_steps_per_second": 1.469, "eval_wer": 0.06032913795517238, "step": 89000 }, { "epoch": 7.38, "learning_rate": 1.6381421446384042e-05, "loss": 0.0657, "step": 89100 }, { "epoch": 7.38, "learning_rate": 1.6377265170407316e-05, "loss": 0.0648, "step": 89200 }, { "epoch": 7.39, "learning_rate": 1.6373108894430593e-05, "loss": 0.0692, "step": 89300 }, { "epoch": 7.4, "learning_rate": 1.6368994181213635e-05, "loss": 0.0829, "step": 89400 }, { "epoch": 7.41, "learning_rate": 1.636483790523691e-05, "loss": 0.0653, "step": 89500 }, { "epoch": 7.41, "eval_loss": 0.13530230522155762, "eval_runtime": 4299.4196, "eval_samples_per_second": 17.583, "eval_steps_per_second": 1.465, "eval_wer": 0.06359682523565348, "step": 89500 }, { "epoch": 7.42, "learning_rate": 1.6360681629260186e-05, "loss": 0.0617, "step": 89600 }, { "epoch": 7.43, "learning_rate": 1.635652535328346e-05, "loss": 0.0624, "step": 89700 }, { "epoch": 7.43, "learning_rate": 1.6352369077306733e-05, "loss": 0.0591, "step": 89800 }, { "epoch": 7.44, "learning_rate": 1.634821280133001e-05, "loss": 0.065, "step": 89900 }, { "epoch": 7.45, "learning_rate": 1.6344056525353284e-05, "loss": 0.0663, "step": 90000 }, { "epoch": 7.45, "eval_loss": 0.13346536457538605, "eval_runtime": 4290.9202, "eval_samples_per_second": 17.617, "eval_steps_per_second": 1.468, "eval_wer": 0.06127529034570716, "step": 90000 }, { "epoch": 7.46, "learning_rate": 1.6339900249376558e-05, "loss": 0.0635, "step": 90100 }, { "epoch": 7.47, "learning_rate": 1.6335743973399835e-05, "loss": 0.0719, "step": 90200 }, { "epoch": 7.48, "learning_rate": 1.633158769742311e-05, "loss": 0.0702, "step": 90300 }, { "epoch": 7.48, "learning_rate": 1.6327431421446383e-05, "loss": 0.064, "step": 90400 }, { "epoch": 7.49, "learning_rate": 1.632327514546966e-05, "loss": 0.0772, "step": 90500 }, { "epoch": 7.49, "eval_loss": 0.13612253963947296, "eval_runtime": 4283.7161, "eval_samples_per_second": 17.647, "eval_steps_per_second": 1.471, "eval_wer": 0.06497836157175142, "step": 90500 }, { "epoch": 7.5, "learning_rate": 1.6319118869492937e-05, "loss": 0.0681, "step": 90600 }, { "epoch": 7.51, "learning_rate": 1.631496259351621e-05, "loss": 0.0635, "step": 90700 }, { "epoch": 7.52, "learning_rate": 1.6310806317539485e-05, "loss": 0.0639, "step": 90800 }, { "epoch": 7.52, "learning_rate": 1.6306650041562762e-05, "loss": 0.0667, "step": 90900 }, { "epoch": 7.53, "learning_rate": 1.6302493765586036e-05, "loss": 0.0595, "step": 91000 }, { "epoch": 7.53, "eval_loss": 0.13068097829818726, "eval_runtime": 4277.4765, "eval_samples_per_second": 17.673, "eval_steps_per_second": 1.473, "eval_wer": 0.06271990350784076, "step": 91000 }, { "epoch": 7.54, "learning_rate": 1.629833748960931e-05, "loss": 0.0768, "step": 91100 }, { "epoch": 7.55, "learning_rate": 1.6294181213632587e-05, "loss": 0.0744, "step": 91200 }, { "epoch": 7.56, "learning_rate": 1.629002493765586e-05, "loss": 0.0659, "step": 91300 }, { "epoch": 7.57, "learning_rate": 1.6285868661679135e-05, "loss": 0.0696, "step": 91400 }, { "epoch": 7.57, "learning_rate": 1.6281712385702412e-05, "loss": 0.0625, "step": 91500 }, { "epoch": 7.57, "eval_loss": 0.12392321974039078, "eval_runtime": 4330.5198, "eval_samples_per_second": 17.456, "eval_steps_per_second": 1.455, "eval_wer": 0.060344522546888386, "step": 91500 }, { "epoch": 7.58, "learning_rate": 1.6277556109725686e-05, "loss": 0.0642, "step": 91600 }, { "epoch": 7.59, "learning_rate": 1.6273399833748963e-05, "loss": 0.0644, "step": 91700 }, { "epoch": 7.6, "learning_rate": 1.6269243557772237e-05, "loss": 0.0701, "step": 91800 }, { "epoch": 7.61, "learning_rate": 1.6265087281795514e-05, "loss": 0.0686, "step": 91900 }, { "epoch": 7.62, "learning_rate": 1.6260931005818788e-05, "loss": 0.0654, "step": 92000 }, { "epoch": 7.62, "eval_loss": 0.12037347257137299, "eval_runtime": 4354.6079, "eval_samples_per_second": 17.36, "eval_steps_per_second": 1.447, "eval_wer": 0.06154759761908059, "step": 92000 }, { "epoch": 7.62, "learning_rate": 1.6256774729842062e-05, "loss": 0.0576, "step": 92100 }, { "epoch": 7.63, "learning_rate": 1.625261845386534e-05, "loss": 0.0667, "step": 92200 }, { "epoch": 7.64, "learning_rate": 1.6248462177888613e-05, "loss": 0.0704, "step": 92300 }, { "epoch": 7.65, "learning_rate": 1.6244305901911887e-05, "loss": 0.0654, "step": 92400 }, { "epoch": 7.66, "learning_rate": 1.6240149625935164e-05, "loss": 0.0659, "step": 92500 }, { "epoch": 7.66, "eval_loss": 0.13218365609645844, "eval_runtime": 4284.8283, "eval_samples_per_second": 17.642, "eval_steps_per_second": 1.47, "eval_wer": 0.06221528889955554, "step": 92500 }, { "epoch": 7.67, "learning_rate": 1.6235993349958438e-05, "loss": 0.0601, "step": 92600 }, { "epoch": 7.67, "learning_rate": 1.6231837073981715e-05, "loss": 0.0659, "step": 92700 }, { "epoch": 7.68, "learning_rate": 1.622768079800499e-05, "loss": 0.0648, "step": 92800 }, { "epoch": 7.69, "learning_rate": 1.6223524522028266e-05, "loss": 0.0637, "step": 92900 }, { "epoch": 7.7, "learning_rate": 1.621936824605154e-05, "loss": 0.0595, "step": 93000 }, { "epoch": 7.7, "eval_loss": 0.1267016977071762, "eval_runtime": 4276.9329, "eval_samples_per_second": 17.675, "eval_steps_per_second": 1.473, "eval_wer": 0.05946913927824726, "step": 93000 }, { "epoch": 7.71, "learning_rate": 1.6215211970074814e-05, "loss": 0.0695, "step": 93100 }, { "epoch": 7.72, "learning_rate": 1.621105569409809e-05, "loss": 0.0581, "step": 93200 }, { "epoch": 7.72, "learning_rate": 1.6206899418121365e-05, "loss": 0.0652, "step": 93300 }, { "epoch": 7.73, "learning_rate": 1.620274314214464e-05, "loss": 0.0695, "step": 93400 }, { "epoch": 7.74, "learning_rate": 1.6198586866167916e-05, "loss": 0.0683, "step": 93500 }, { "epoch": 7.74, "eval_loss": 0.12436062097549438, "eval_runtime": 4288.2163, "eval_samples_per_second": 17.629, "eval_steps_per_second": 1.469, "eval_wer": 0.060807598757540375, "step": 93500 }, { "epoch": 7.75, "learning_rate": 1.619443059019119e-05, "loss": 0.0593, "step": 93600 }, { "epoch": 7.76, "learning_rate": 1.6190274314214463e-05, "loss": 0.0701, "step": 93700 }, { "epoch": 7.76, "learning_rate": 1.618611803823774e-05, "loss": 0.0656, "step": 93800 }, { "epoch": 7.77, "learning_rate": 1.6182003325020782e-05, "loss": 0.061, "step": 93900 }, { "epoch": 7.78, "learning_rate": 1.6177847049044056e-05, "loss": 0.0703, "step": 94000 }, { "epoch": 7.78, "eval_loss": 0.11377289891242981, "eval_runtime": 4290.5334, "eval_samples_per_second": 17.619, "eval_steps_per_second": 1.468, "eval_wer": 0.05952452380842491, "step": 94000 }, { "epoch": 7.79, "learning_rate": 1.6173690773067333e-05, "loss": 0.0597, "step": 94100 }, { "epoch": 7.8, "learning_rate": 1.6169534497090607e-05, "loss": 0.0632, "step": 94200 }, { "epoch": 7.81, "learning_rate": 1.6165378221113884e-05, "loss": 0.0624, "step": 94300 }, { "epoch": 7.81, "learning_rate": 1.6161221945137158e-05, "loss": 0.0676, "step": 94400 }, { "epoch": 7.82, "learning_rate": 1.6157065669160435e-05, "loss": 0.0658, "step": 94500 }, { "epoch": 7.82, "eval_loss": 0.11505799740552902, "eval_runtime": 4281.6844, "eval_samples_per_second": 17.655, "eval_steps_per_second": 1.471, "eval_wer": 0.05796452620842122, "step": 94500 }, { "epoch": 7.83, "learning_rate": 1.615290939318371e-05, "loss": 0.0753, "step": 94600 }, { "epoch": 7.84, "learning_rate": 1.6148753117206983e-05, "loss": 0.0717, "step": 94700 }, { "epoch": 7.85, "learning_rate": 1.614459684123026e-05, "loss": 0.0648, "step": 94800 }, { "epoch": 7.86, "learning_rate": 1.6140440565253534e-05, "loss": 0.0652, "step": 94900 }, { "epoch": 7.86, "learning_rate": 1.6136284289276808e-05, "loss": 0.0614, "step": 95000 }, { "epoch": 7.86, "eval_loss": 0.1299211084842682, "eval_runtime": 4286.9743, "eval_samples_per_second": 17.634, "eval_steps_per_second": 1.47, "eval_wer": 0.061344521008429215, "step": 95000 }, { "epoch": 7.87, "learning_rate": 1.6132128013300085e-05, "loss": 0.0671, "step": 95100 }, { "epoch": 7.88, "learning_rate": 1.612797173732336e-05, "loss": 0.0631, "step": 95200 }, { "epoch": 7.89, "learning_rate": 1.6123815461346636e-05, "loss": 0.0669, "step": 95300 }, { "epoch": 7.9, "learning_rate": 1.611965918536991e-05, "loss": 0.0654, "step": 95400 }, { "epoch": 7.91, "learning_rate": 1.6115502909393187e-05, "loss": 0.069, "step": 95500 }, { "epoch": 7.91, "eval_loss": 0.1299111247062683, "eval_runtime": 4287.025, "eval_samples_per_second": 17.633, "eval_steps_per_second": 1.47, "eval_wer": 0.06205528914570901, "step": 95500 }, { "epoch": 7.91, "learning_rate": 1.611134663341646e-05, "loss": 0.0587, "step": 95600 }, { "epoch": 7.92, "learning_rate": 1.6107190357439735e-05, "loss": 0.0694, "step": 95700 }, { "epoch": 7.93, "learning_rate": 1.6103034081463012e-05, "loss": 0.0727, "step": 95800 }, { "epoch": 7.94, "learning_rate": 1.6098919368246054e-05, "loss": 0.0657, "step": 95900 }, { "epoch": 7.95, "learning_rate": 1.6094763092269328e-05, "loss": 0.0663, "step": 96000 }, { "epoch": 7.95, "eval_loss": 0.128158301115036, "eval_runtime": 4275.5886, "eval_samples_per_second": 17.681, "eval_steps_per_second": 1.473, "eval_wer": 0.05957990833860256, "step": 96000 }, { "epoch": 7.96, "learning_rate": 1.60906068162926e-05, "loss": 0.073, "step": 96100 }, { "epoch": 7.96, "learning_rate": 1.608645054031588e-05, "loss": 0.0637, "step": 96200 }, { "epoch": 7.97, "learning_rate": 1.6082294264339153e-05, "loss": 0.064, "step": 96300 }, { "epoch": 7.98, "learning_rate": 1.6078137988362426e-05, "loss": 0.068, "step": 96400 }, { "epoch": 7.99, "learning_rate": 1.6073981712385704e-05, "loss": 0.0665, "step": 96500 }, { "epoch": 7.99, "eval_loss": 0.12895143032073975, "eval_runtime": 4279.5787, "eval_samples_per_second": 17.664, "eval_steps_per_second": 1.472, "eval_wer": 0.059287601095998316, "step": 96500 }, { "epoch": 8.0, "learning_rate": 1.6069825436408977e-05, "loss": 0.0622, "step": 96600 }, { "epoch": 8.0, "learning_rate": 1.6065669160432255e-05, "loss": 0.0634, "step": 96700 }, { "epoch": 8.01, "learning_rate": 1.6061512884455532e-05, "loss": 0.0557, "step": 96800 }, { "epoch": 8.02, "learning_rate": 1.6057356608478806e-05, "loss": 0.0578, "step": 96900 }, { "epoch": 8.03, "learning_rate": 1.605320033250208e-05, "loss": 0.0631, "step": 97000 }, { "epoch": 8.03, "eval_loss": 0.13802798092365265, "eval_runtime": 4282.6877, "eval_samples_per_second": 17.651, "eval_steps_per_second": 1.471, "eval_wer": 0.06242144242855011, "step": 97000 }, { "epoch": 8.04, "learning_rate": 1.6049044056525353e-05, "loss": 0.0618, "step": 97100 }, { "epoch": 8.05, "learning_rate": 1.604488778054863e-05, "loss": 0.0572, "step": 97200 }, { "epoch": 8.05, "learning_rate": 1.6040731504571904e-05, "loss": 0.0654, "step": 97300 }, { "epoch": 8.06, "learning_rate": 1.6036575228595178e-05, "loss": 0.0578, "step": 97400 }, { "epoch": 8.07, "learning_rate": 1.6032418952618455e-05, "loss": 0.0617, "step": 97500 }, { "epoch": 8.07, "eval_loss": 0.14379365742206573, "eval_runtime": 4280.7629, "eval_samples_per_second": 17.659, "eval_steps_per_second": 1.472, "eval_wer": 0.06220144276701113, "step": 97500 }, { "epoch": 8.08, "learning_rate": 1.602826267664173e-05, "loss": 0.0641, "step": 97600 }, { "epoch": 8.09, "learning_rate": 1.6024106400665006e-05, "loss": 0.0593, "step": 97700 }, { "epoch": 8.1, "learning_rate": 1.601995012468828e-05, "loss": 0.0579, "step": 97800 }, { "epoch": 8.1, "learning_rate": 1.6015793848711557e-05, "loss": 0.0672, "step": 97900 }, { "epoch": 8.11, "learning_rate": 1.601163757273483e-05, "loss": 0.0673, "step": 98000 }, { "epoch": 8.11, "eval_loss": 0.13368521630764008, "eval_runtime": 4290.3526, "eval_samples_per_second": 17.62, "eval_steps_per_second": 1.468, "eval_wer": 0.0638537479173109, "step": 98000 }, { "epoch": 8.12, "learning_rate": 1.6007481296758105e-05, "loss": 0.0627, "step": 98100 }, { "epoch": 8.13, "learning_rate": 1.6003325020781382e-05, "loss": 0.0568, "step": 98200 }, { "epoch": 8.14, "learning_rate": 1.5999168744804656e-05, "loss": 0.0656, "step": 98300 }, { "epoch": 8.15, "learning_rate": 1.599501246882793e-05, "loss": 0.0547, "step": 98400 }, { "epoch": 8.15, "learning_rate": 1.5990856192851207e-05, "loss": 0.0643, "step": 98500 }, { "epoch": 8.15, "eval_loss": 0.13296905159950256, "eval_runtime": 4284.4779, "eval_samples_per_second": 17.644, "eval_steps_per_second": 1.47, "eval_wer": 0.061664520516122284, "step": 98500 }, { "epoch": 8.16, "learning_rate": 1.598669991687448e-05, "loss": 0.0593, "step": 98600 }, { "epoch": 8.17, "learning_rate": 1.5982543640897755e-05, "loss": 0.0613, "step": 98700 }, { "epoch": 8.18, "learning_rate": 1.5978387364921032e-05, "loss": 0.0623, "step": 98800 }, { "epoch": 8.19, "learning_rate": 1.597423108894431e-05, "loss": 0.059, "step": 98900 }, { "epoch": 8.2, "learning_rate": 1.5970074812967583e-05, "loss": 0.0637, "step": 99000 }, { "epoch": 8.2, "eval_loss": 0.13642705976963043, "eval_runtime": 4288.5819, "eval_samples_per_second": 17.627, "eval_steps_per_second": 1.469, "eval_wer": 0.06181836643328241, "step": 99000 }, { "epoch": 8.2, "learning_rate": 1.5965918536990857e-05, "loss": 0.0607, "step": 99100 }, { "epoch": 8.21, "learning_rate": 1.5961762261014134e-05, "loss": 0.0592, "step": 99200 }, { "epoch": 8.22, "learning_rate": 1.5957605985037408e-05, "loss": 0.0637, "step": 99300 }, { "epoch": 8.23, "learning_rate": 1.595344970906068e-05, "loss": 0.0671, "step": 99400 }, { "epoch": 8.24, "learning_rate": 1.594929343308396e-05, "loss": 0.0677, "step": 99500 }, { "epoch": 8.24, "eval_loss": 0.13001805543899536, "eval_runtime": 4314.4433, "eval_samples_per_second": 17.521, "eval_steps_per_second": 1.46, "eval_wer": 0.05905683222025812, "step": 99500 }, { "epoch": 8.25, "learning_rate": 1.5945137157107233e-05, "loss": 0.0565, "step": 99600 }, { "epoch": 8.25, "learning_rate": 1.5940980881130507e-05, "loss": 0.0589, "step": 99700 }, { "epoch": 8.26, "learning_rate": 1.5936824605153784e-05, "loss": 0.0625, "step": 99800 }, { "epoch": 8.27, "learning_rate": 1.5932709891936826e-05, "loss": 0.0594, "step": 99900 }, { "epoch": 8.28, "learning_rate": 1.59285536159601e-05, "loss": 0.0589, "step": 100000 }, { "epoch": 8.28, "eval_loss": 0.1326805055141449, "eval_runtime": 4280.4502, "eval_samples_per_second": 17.661, "eval_steps_per_second": 1.472, "eval_wer": 0.05978452340842553, "step": 100000 }, { "epoch": 8.29, "learning_rate": 1.5924397339983377e-05, "loss": 0.0623, "step": 100100 }, { "epoch": 8.29, "learning_rate": 1.592024106400665e-05, "loss": 0.0669, "step": 100200 }, { "epoch": 8.3, "learning_rate": 1.5916084788029928e-05, "loss": 0.0676, "step": 100300 }, { "epoch": 8.31, "learning_rate": 1.59119285120532e-05, "loss": 0.0558, "step": 100400 }, { "epoch": 8.32, "learning_rate": 1.590777223607648e-05, "loss": 0.0625, "step": 100500 }, { "epoch": 8.32, "eval_loss": 0.13208545744419098, "eval_runtime": 4274.2918, "eval_samples_per_second": 17.686, "eval_steps_per_second": 1.474, "eval_wer": 0.06069067586049868, "step": 100500 }, { "epoch": 8.33, "learning_rate": 1.5903615960099752e-05, "loss": 0.0627, "step": 100600 }, { "epoch": 8.34, "learning_rate": 1.5899459684123026e-05, "loss": 0.0537, "step": 100700 }, { "epoch": 8.34, "learning_rate": 1.5895303408146304e-05, "loss": 0.0784, "step": 100800 }, { "epoch": 8.35, "learning_rate": 1.5891147132169577e-05, "loss": 0.0579, "step": 100900 }, { "epoch": 8.36, "learning_rate": 1.588699085619285e-05, "loss": 0.0603, "step": 101000 }, { "epoch": 8.36, "eval_loss": 0.13600020110607147, "eval_runtime": 4287.5961, "eval_samples_per_second": 17.631, "eval_steps_per_second": 1.469, "eval_wer": 0.06327374880961721, "step": 101000 }, { "epoch": 8.37, "learning_rate": 1.5882834580216125e-05, "loss": 0.0713, "step": 101100 }, { "epoch": 8.38, "learning_rate": 1.5878678304239402e-05, "loss": 0.06, "step": 101200 }, { "epoch": 8.39, "learning_rate": 1.587452202826268e-05, "loss": 0.0589, "step": 101300 }, { "epoch": 8.39, "learning_rate": 1.5870365752285953e-05, "loss": 0.0598, "step": 101400 }, { "epoch": 8.4, "learning_rate": 1.586620947630923e-05, "loss": 0.0582, "step": 101500 }, { "epoch": 8.4, "eval_loss": 0.13653209805488586, "eval_runtime": 4304.7459, "eval_samples_per_second": 17.561, "eval_steps_per_second": 1.464, "eval_wer": 0.06213067364511747, "step": 101500 }, { "epoch": 8.41, "learning_rate": 1.5862053200332504e-05, "loss": 0.06, "step": 101600 }, { "epoch": 8.42, "learning_rate": 1.5857896924355778e-05, "loss": 0.0603, "step": 101700 }, { "epoch": 8.43, "learning_rate": 1.5853740648379055e-05, "loss": 0.0819, "step": 101800 }, { "epoch": 8.44, "learning_rate": 1.584958437240233e-05, "loss": 0.0543, "step": 101900 }, { "epoch": 8.44, "learning_rate": 1.5845428096425603e-05, "loss": 0.0748, "step": 102000 }, { "epoch": 8.44, "eval_loss": 0.1417304128408432, "eval_runtime": 4294.1991, "eval_samples_per_second": 17.604, "eval_steps_per_second": 1.467, "eval_wer": 0.06255374991730782, "step": 102000 }, { "epoch": 8.45, "learning_rate": 1.5841271820448877e-05, "loss": 0.0658, "step": 102100 }, { "epoch": 8.46, "learning_rate": 1.5837115544472154e-05, "loss": 0.0607, "step": 102200 }, { "epoch": 8.47, "learning_rate": 1.5832959268495428e-05, "loss": 0.0601, "step": 102300 }, { "epoch": 8.48, "learning_rate": 1.5828802992518705e-05, "loss": 0.0621, "step": 102400 }, { "epoch": 8.49, "learning_rate": 1.582464671654198e-05, "loss": 0.0608, "step": 102500 }, { "epoch": 8.49, "eval_loss": 0.12749695777893066, "eval_runtime": 4302.2866, "eval_samples_per_second": 17.571, "eval_steps_per_second": 1.464, "eval_wer": 0.059012216904281684, "step": 102500 }, { "epoch": 8.49, "learning_rate": 1.5820490440565256e-05, "loss": 0.0624, "step": 102600 }, { "epoch": 8.5, "learning_rate": 1.581633416458853e-05, "loss": 0.0578, "step": 102700 }, { "epoch": 8.51, "learning_rate": 1.5812177888611807e-05, "loss": 0.057, "step": 102800 }, { "epoch": 8.52, "learning_rate": 1.580802161263508e-05, "loss": 0.0593, "step": 102900 }, { "epoch": 8.53, "learning_rate": 1.5803865336658355e-05, "loss": 0.0581, "step": 103000 }, { "epoch": 8.53, "eval_loss": 0.13302412629127502, "eval_runtime": 4295.0185, "eval_samples_per_second": 17.601, "eval_steps_per_second": 1.467, "eval_wer": 0.06016144590546784, "step": 103000 }, { "epoch": 8.53, "learning_rate": 1.579970906068163e-05, "loss": 0.0707, "step": 103100 }, { "epoch": 8.54, "learning_rate": 1.5795552784704906e-05, "loss": 0.0591, "step": 103200 }, { "epoch": 8.55, "learning_rate": 1.579139650872818e-05, "loss": 0.0621, "step": 103300 }, { "epoch": 8.56, "learning_rate": 1.578728179551122e-05, "loss": 0.06, "step": 103400 }, { "epoch": 8.57, "learning_rate": 1.57831255195345e-05, "loss": 0.0589, "step": 103500 }, { "epoch": 8.57, "eval_loss": 0.1400037556886673, "eval_runtime": 4281.1553, "eval_samples_per_second": 17.658, "eval_steps_per_second": 1.472, "eval_wer": 0.06303682609719062, "step": 103500 }, { "epoch": 8.58, "learning_rate": 1.5778969243557772e-05, "loss": 0.0679, "step": 103600 }, { "epoch": 8.58, "learning_rate": 1.577481296758105e-05, "loss": 0.0621, "step": 103700 }, { "epoch": 8.59, "learning_rate": 1.5770656691604323e-05, "loss": 0.0576, "step": 103800 }, { "epoch": 8.6, "learning_rate": 1.57665004156276e-05, "loss": 0.064, "step": 103900 }, { "epoch": 8.61, "learning_rate": 1.5762344139650874e-05, "loss": 0.0642, "step": 104000 }, { "epoch": 8.61, "eval_loss": 0.1278255134820938, "eval_runtime": 4344.2675, "eval_samples_per_second": 17.401, "eval_steps_per_second": 1.45, "eval_wer": 0.06049067616819051, "step": 104000 }, { "epoch": 8.62, "learning_rate": 1.5758187863674152e-05, "loss": 0.0627, "step": 104100 }, { "epoch": 8.63, "learning_rate": 1.5754031587697426e-05, "loss": 0.0707, "step": 104200 }, { "epoch": 8.63, "learning_rate": 1.57498753117207e-05, "loss": 0.0552, "step": 104300 }, { "epoch": 8.64, "learning_rate": 1.5745719035743973e-05, "loss": 0.0593, "step": 104400 }, { "epoch": 8.65, "learning_rate": 1.574156275976725e-05, "loss": 0.0564, "step": 104500 }, { "epoch": 8.65, "eval_loss": 0.142498180270195, "eval_runtime": 4276.6305, "eval_samples_per_second": 17.676, "eval_steps_per_second": 1.473, "eval_wer": 0.06126759804984915, "step": 104500 }, { "epoch": 8.66, "learning_rate": 1.5737406483790524e-05, "loss": 0.0675, "step": 104600 }, { "epoch": 8.67, "learning_rate": 1.5733250207813798e-05, "loss": 0.0633, "step": 104700 }, { "epoch": 8.68, "learning_rate": 1.5729093931837075e-05, "loss": 0.0587, "step": 104800 }, { "epoch": 8.68, "learning_rate": 1.572493765586035e-05, "loss": 0.062, "step": 104900 }, { "epoch": 8.69, "learning_rate": 1.5720781379883626e-05, "loss": 0.0638, "step": 105000 }, { "epoch": 8.69, "eval_loss": 0.13117855787277222, "eval_runtime": 4285.1412, "eval_samples_per_second": 17.641, "eval_steps_per_second": 1.47, "eval_wer": 0.06027067650665153, "step": 105000 }, { "epoch": 8.7, "learning_rate": 1.57166251039069e-05, "loss": 0.072, "step": 105100 }, { "epoch": 8.71, "learning_rate": 1.5712468827930177e-05, "loss": 0.0581, "step": 105200 }, { "epoch": 8.72, "learning_rate": 1.570831255195345e-05, "loss": 0.0578, "step": 105300 }, { "epoch": 8.73, "learning_rate": 1.5704156275976725e-05, "loss": 0.0584, "step": 105400 }, { "epoch": 8.73, "learning_rate": 1.5700000000000002e-05, "loss": 0.0677, "step": 105500 }, { "epoch": 8.73, "eval_loss": 0.1252855658531189, "eval_runtime": 4279.8128, "eval_samples_per_second": 17.663, "eval_steps_per_second": 1.472, "eval_wer": 0.059218370433276254, "step": 105500 }, { "epoch": 8.74, "learning_rate": 1.5695843724023276e-05, "loss": 0.0602, "step": 105600 }, { "epoch": 8.75, "learning_rate": 1.569168744804655e-05, "loss": 0.0624, "step": 105700 }, { "epoch": 8.76, "learning_rate": 1.5687531172069827e-05, "loss": 0.063, "step": 105800 }, { "epoch": 8.77, "learning_rate": 1.56833748960931e-05, "loss": 0.0642, "step": 105900 }, { "epoch": 8.77, "learning_rate": 1.5679218620116378e-05, "loss": 0.0695, "step": 106000 }, { "epoch": 8.77, "eval_loss": 0.14521802961826324, "eval_runtime": 4306.5224, "eval_samples_per_second": 17.554, "eval_steps_per_second": 1.463, "eval_wer": 0.06364913284748792, "step": 106000 }, { "epoch": 8.78, "learning_rate": 1.5675062344139652e-05, "loss": 0.057, "step": 106100 }, { "epoch": 8.79, "learning_rate": 1.567090606816293e-05, "loss": 0.0694, "step": 106200 }, { "epoch": 8.8, "learning_rate": 1.5666749792186203e-05, "loss": 0.0536, "step": 106300 }, { "epoch": 8.81, "learning_rate": 1.5662593516209477e-05, "loss": 0.0613, "step": 106400 }, { "epoch": 8.82, "learning_rate": 1.5658437240232754e-05, "loss": 0.0581, "step": 106500 }, { "epoch": 8.82, "eval_loss": 0.13794204592704773, "eval_runtime": 4284.988, "eval_samples_per_second": 17.642, "eval_steps_per_second": 1.47, "eval_wer": 0.060715291207244296, "step": 106500 }, { "epoch": 8.82, "learning_rate": 1.5654280964256028e-05, "loss": 0.0641, "step": 106600 }, { "epoch": 8.83, "learning_rate": 1.56501246882793e-05, "loss": 0.0608, "step": 106700 }, { "epoch": 8.84, "learning_rate": 1.564596841230258e-05, "loss": 0.0623, "step": 106800 }, { "epoch": 8.85, "learning_rate": 1.5641812136325853e-05, "loss": 0.0665, "step": 106900 }, { "epoch": 8.86, "learning_rate": 1.5637655860349126e-05, "loss": 0.0593, "step": 107000 }, { "epoch": 8.86, "eval_loss": 0.12940241396427155, "eval_runtime": 4294.0176, "eval_samples_per_second": 17.605, "eval_steps_per_second": 1.467, "eval_wer": 0.058902986303097996, "step": 107000 }, { "epoch": 8.87, "learning_rate": 1.5633499584372404e-05, "loss": 0.0596, "step": 107100 }, { "epoch": 8.87, "learning_rate": 1.562934330839568e-05, "loss": 0.0619, "step": 107200 }, { "epoch": 8.88, "learning_rate": 1.5625187032418955e-05, "loss": 0.056, "step": 107300 }, { "epoch": 8.89, "learning_rate": 1.562103075644223e-05, "loss": 0.0572, "step": 107400 }, { "epoch": 8.9, "learning_rate": 1.5616874480465506e-05, "loss": 0.0597, "step": 107500 }, { "epoch": 8.9, "eval_loss": 0.12427941709756851, "eval_runtime": 4304.6905, "eval_samples_per_second": 17.561, "eval_steps_per_second": 1.464, "eval_wer": 0.059026063036826094, "step": 107500 }, { "epoch": 8.91, "learning_rate": 1.561271820448878e-05, "loss": 0.0583, "step": 107600 }, { "epoch": 8.92, "learning_rate": 1.560860349127182e-05, "loss": 0.0585, "step": 107700 }, { "epoch": 8.92, "learning_rate": 1.56044472152951e-05, "loss": 0.061, "step": 107800 }, { "epoch": 8.93, "learning_rate": 1.5600290939318372e-05, "loss": 0.0581, "step": 107900 }, { "epoch": 8.94, "learning_rate": 1.5596134663341646e-05, "loss": 0.0559, "step": 108000 }, { "epoch": 8.94, "eval_loss": 0.13426615297794342, "eval_runtime": 4320.9818, "eval_samples_per_second": 17.495, "eval_steps_per_second": 1.458, "eval_wer": 0.06017529203801225, "step": 108000 }, { "epoch": 8.95, "learning_rate": 1.5591978387364923e-05, "loss": 0.0662, "step": 108100 }, { "epoch": 8.96, "learning_rate": 1.5587822111388197e-05, "loss": 0.0578, "step": 108200 }, { "epoch": 8.97, "learning_rate": 1.558366583541147e-05, "loss": 0.0573, "step": 108300 }, { "epoch": 8.97, "learning_rate": 1.5579509559434748e-05, "loss": 0.0568, "step": 108400 }, { "epoch": 8.98, "learning_rate": 1.5575353283458022e-05, "loss": 0.0525, "step": 108500 }, { "epoch": 8.98, "eval_loss": 0.13602516055107117, "eval_runtime": 4288.3997, "eval_samples_per_second": 17.628, "eval_steps_per_second": 1.469, "eval_wer": 0.06058913755517299, "step": 108500 }, { "epoch": 8.99, "learning_rate": 1.55711970074813e-05, "loss": 0.0604, "step": 108600 }, { "epoch": 9.0, "learning_rate": 1.5567040731504573e-05, "loss": 0.0552, "step": 108700 }, { "epoch": 9.01, "learning_rate": 1.556288445552785e-05, "loss": 0.0571, "step": 108800 }, { "epoch": 9.01, "learning_rate": 1.5558728179551124e-05, "loss": 0.0638, "step": 108900 }, { "epoch": 9.02, "learning_rate": 1.5554571903574398e-05, "loss": 0.0558, "step": 109000 }, { "epoch": 9.02, "eval_loss": 0.13868845999240875, "eval_runtime": 4296.6045, "eval_samples_per_second": 17.594, "eval_steps_per_second": 1.466, "eval_wer": 0.05910144753623456, "step": 109000 }, { "epoch": 9.03, "learning_rate": 1.5550415627597675e-05, "loss": 0.0487, "step": 109100 }, { "epoch": 9.04, "learning_rate": 1.554625935162095e-05, "loss": 0.0547, "step": 109200 }, { "epoch": 9.05, "learning_rate": 1.5542103075644223e-05, "loss": 0.0606, "step": 109300 }, { "epoch": 9.06, "learning_rate": 1.5537946799667497e-05, "loss": 0.0535, "step": 109400 }, { "epoch": 9.06, "learning_rate": 1.5533790523690774e-05, "loss": 0.0491, "step": 109500 }, { "epoch": 9.06, "eval_loss": 0.14434558153152466, "eval_runtime": 4277.3038, "eval_samples_per_second": 17.674, "eval_steps_per_second": 1.473, "eval_wer": 0.05995836929481647, "step": 109500 }, { "epoch": 9.07, "learning_rate": 1.5529634247714048e-05, "loss": 0.0498, "step": 109600 }, { "epoch": 9.08, "learning_rate": 1.5525477971737325e-05, "loss": 0.0562, "step": 109700 }, { "epoch": 9.09, "learning_rate": 1.5521363258520367e-05, "loss": 0.0693, "step": 109800 }, { "epoch": 9.1, "learning_rate": 1.551720698254364e-05, "loss": 0.0566, "step": 109900 }, { "epoch": 9.11, "learning_rate": 1.5513050706566918e-05, "loss": 0.06, "step": 110000 }, { "epoch": 9.11, "eval_loss": 0.13651511073112488, "eval_runtime": 4286.4497, "eval_samples_per_second": 17.636, "eval_steps_per_second": 1.47, "eval_wer": 0.05873683271256506, "step": 110000 }, { "epoch": 9.11, "learning_rate": 1.5508894430590195e-05, "loss": 0.051, "step": 110100 }, { "epoch": 9.12, "learning_rate": 1.550473815461347e-05, "loss": 0.0534, "step": 110200 }, { "epoch": 9.13, "learning_rate": 1.5500581878636743e-05, "loss": 0.064, "step": 110300 }, { "epoch": 9.14, "learning_rate": 1.5496425602660016e-05, "loss": 0.0581, "step": 110400 }, { "epoch": 9.15, "learning_rate": 1.5492269326683294e-05, "loss": 0.0579, "step": 110500 }, { "epoch": 9.15, "eval_loss": 0.12653501331806183, "eval_runtime": 4277.0209, "eval_samples_per_second": 17.675, "eval_steps_per_second": 1.473, "eval_wer": 0.05863067902972457, "step": 110500 }, { "epoch": 9.16, "learning_rate": 1.5488113050706567e-05, "loss": 0.0533, "step": 110600 }, { "epoch": 9.16, "learning_rate": 1.548395677472984e-05, "loss": 0.0535, "step": 110700 }, { "epoch": 9.17, "learning_rate": 1.547980049875312e-05, "loss": 0.0541, "step": 110800 }, { "epoch": 9.18, "learning_rate": 1.5475644222776392e-05, "loss": 0.0549, "step": 110900 }, { "epoch": 9.19, "learning_rate": 1.547148794679967e-05, "loss": 0.0573, "step": 111000 }, { "epoch": 9.19, "eval_loss": 0.13596658408641815, "eval_runtime": 4298.7252, "eval_samples_per_second": 17.585, "eval_steps_per_second": 1.466, "eval_wer": 0.059352216381205566, "step": 111000 }, { "epoch": 9.2, "learning_rate": 1.5467331670822943e-05, "loss": 0.0627, "step": 111100 }, { "epoch": 9.21, "learning_rate": 1.546317539484622e-05, "loss": 0.0541, "step": 111200 }, { "epoch": 9.21, "learning_rate": 1.5459019118869494e-05, "loss": 0.0471, "step": 111300 }, { "epoch": 9.22, "learning_rate": 1.5454862842892768e-05, "loss": 0.0585, "step": 111400 }, { "epoch": 9.23, "learning_rate": 1.5450706566916045e-05, "loss": 0.0569, "step": 111500 }, { "epoch": 9.23, "eval_loss": 0.13165755569934845, "eval_runtime": 4276.0903, "eval_samples_per_second": 17.679, "eval_steps_per_second": 1.473, "eval_wer": 0.05985067715280438, "step": 111500 }, { "epoch": 9.24, "learning_rate": 1.544655029093932e-05, "loss": 0.059, "step": 111600 }, { "epoch": 9.25, "learning_rate": 1.5442394014962593e-05, "loss": 0.0575, "step": 111700 }, { "epoch": 9.25, "learning_rate": 1.543823773898587e-05, "loss": 0.0616, "step": 111800 }, { "epoch": 9.26, "learning_rate": 1.5434081463009144e-05, "loss": 0.0637, "step": 111900 }, { "epoch": 9.27, "learning_rate": 1.5429925187032418e-05, "loss": 0.0603, "step": 112000 }, { "epoch": 9.27, "eval_loss": 0.1299159675836563, "eval_runtime": 4284.8248, "eval_samples_per_second": 17.642, "eval_steps_per_second": 1.47, "eval_wer": 0.05984760023446118, "step": 112000 }, { "epoch": 9.28, "learning_rate": 1.5425768911055695e-05, "loss": 0.0528, "step": 112100 }, { "epoch": 9.29, "learning_rate": 1.5421612635078972e-05, "loss": 0.0526, "step": 112200 }, { "epoch": 9.3, "learning_rate": 1.5417456359102246e-05, "loss": 0.0583, "step": 112300 }, { "epoch": 9.3, "learning_rate": 1.541330008312552e-05, "loss": 0.0593, "step": 112400 }, { "epoch": 9.31, "learning_rate": 1.5409185369908565e-05, "loss": 0.065, "step": 112500 }, { "epoch": 9.31, "eval_loss": 0.12688618898391724, "eval_runtime": 4279.2584, "eval_samples_per_second": 17.665, "eval_steps_per_second": 1.472, "eval_wer": 0.05942913933978563, "step": 112500 }, { "epoch": 9.32, "learning_rate": 1.540502909393184e-05, "loss": 0.0637, "step": 112600 }, { "epoch": 9.33, "learning_rate": 1.5400872817955113e-05, "loss": 0.0543, "step": 112700 }, { "epoch": 9.34, "learning_rate": 1.539671654197839e-05, "loss": 0.0613, "step": 112800 }, { "epoch": 9.35, "learning_rate": 1.5392560266001664e-05, "loss": 0.0599, "step": 112900 }, { "epoch": 9.35, "learning_rate": 1.5388403990024938e-05, "loss": 0.0561, "step": 113000 }, { "epoch": 9.35, "eval_loss": 0.1300799548625946, "eval_runtime": 4279.4335, "eval_samples_per_second": 17.665, "eval_steps_per_second": 1.472, "eval_wer": 0.05858452525457653, "step": 113000 }, { "epoch": 9.36, "learning_rate": 1.5384247714048215e-05, "loss": 0.0563, "step": 113100 }, { "epoch": 9.37, "learning_rate": 1.538009143807149e-05, "loss": 0.059, "step": 113200 }, { "epoch": 9.38, "learning_rate": 1.5375935162094763e-05, "loss": 0.0594, "step": 113300 }, { "epoch": 9.39, "learning_rate": 1.537177888611804e-05, "loss": 0.0577, "step": 113400 }, { "epoch": 9.4, "learning_rate": 1.5367622610141314e-05, "loss": 0.0542, "step": 113500 }, { "epoch": 9.4, "eval_loss": 0.1333344727754593, "eval_runtime": 4283.9694, "eval_samples_per_second": 17.646, "eval_steps_per_second": 1.471, "eval_wer": 0.060002984610792905, "step": 113500 }, { "epoch": 9.4, "learning_rate": 1.536346633416459e-05, "loss": 0.0638, "step": 113600 }, { "epoch": 9.41, "learning_rate": 1.5359310058187865e-05, "loss": 0.0614, "step": 113700 }, { "epoch": 9.42, "learning_rate": 1.5355153782211142e-05, "loss": 0.0612, "step": 113800 }, { "epoch": 9.43, "learning_rate": 1.5350997506234416e-05, "loss": 0.0552, "step": 113900 }, { "epoch": 9.44, "learning_rate": 1.534684123025769e-05, "loss": 0.0622, "step": 114000 }, { "epoch": 9.44, "eval_loss": 0.12254803627729416, "eval_runtime": 4279.8284, "eval_samples_per_second": 17.663, "eval_steps_per_second": 1.472, "eval_wer": 0.05729221955043146, "step": 114000 }, { "epoch": 9.45, "learning_rate": 1.5342684954280967e-05, "loss": 0.0528, "step": 114100 }, { "epoch": 9.45, "learning_rate": 1.533852867830424e-05, "loss": 0.0608, "step": 114200 }, { "epoch": 9.46, "learning_rate": 1.5334372402327514e-05, "loss": 0.0624, "step": 114300 }, { "epoch": 9.47, "learning_rate": 1.533021612635079e-05, "loss": 0.0528, "step": 114400 }, { "epoch": 9.48, "learning_rate": 1.5326059850374065e-05, "loss": 0.0534, "step": 114500 }, { "epoch": 9.48, "eval_loss": 0.13141798973083496, "eval_runtime": 4286.299, "eval_samples_per_second": 17.636, "eval_steps_per_second": 1.47, "eval_wer": 0.05991375397884003, "step": 114500 }, { "epoch": 9.49, "learning_rate": 1.5321903574397343e-05, "loss": 0.0629, "step": 114600 }, { "epoch": 9.5, "learning_rate": 1.5317747298420616e-05, "loss": 0.0571, "step": 114700 }, { "epoch": 9.5, "learning_rate": 1.5313591022443894e-05, "loss": 0.0577, "step": 114800 }, { "epoch": 9.51, "learning_rate": 1.5309434746467167e-05, "loss": 0.0562, "step": 114900 }, { "epoch": 9.52, "learning_rate": 1.530527847049044e-05, "loss": 0.048, "step": 115000 }, { "epoch": 9.52, "eval_loss": 0.13801200687885284, "eval_runtime": 4307.0809, "eval_samples_per_second": 17.551, "eval_steps_per_second": 1.463, "eval_wer": 0.0589137555172992, "step": 115000 }, { "epoch": 9.53, "learning_rate": 1.530112219451372e-05, "loss": 0.0663, "step": 115100 }, { "epoch": 9.54, "learning_rate": 1.5296965918536992e-05, "loss": 0.0525, "step": 115200 }, { "epoch": 9.54, "learning_rate": 1.5292809642560266e-05, "loss": 0.0652, "step": 115300 }, { "epoch": 9.55, "learning_rate": 1.5288653366583543e-05, "loss": 0.0494, "step": 115400 }, { "epoch": 9.56, "learning_rate": 1.5284497090606817e-05, "loss": 0.0555, "step": 115500 }, { "epoch": 9.56, "eval_loss": 0.13024665415287018, "eval_runtime": 4332.9198, "eval_samples_per_second": 17.447, "eval_steps_per_second": 1.454, "eval_wer": 0.0592722165042823, "step": 115500 }, { "epoch": 9.57, "learning_rate": 1.528034081463009e-05, "loss": 0.0578, "step": 115600 }, { "epoch": 9.58, "learning_rate": 1.5276184538653368e-05, "loss": 0.0636, "step": 115700 }, { "epoch": 9.59, "learning_rate": 1.5272028262676642e-05, "loss": 0.0567, "step": 115800 }, { "epoch": 9.59, "learning_rate": 1.526787198669992e-05, "loss": 0.0556, "step": 115900 }, { "epoch": 9.6, "learning_rate": 1.5263715710723193e-05, "loss": 0.0534, "step": 116000 }, { "epoch": 9.6, "eval_loss": 0.12587133049964905, "eval_runtime": 4389.5673, "eval_samples_per_second": 17.222, "eval_steps_per_second": 1.435, "eval_wer": 0.05753529609954446, "step": 116000 }, { "epoch": 9.61, "learning_rate": 1.525955943474647e-05, "loss": 0.0617, "step": 116100 }, { "epoch": 9.62, "learning_rate": 1.5255403158769744e-05, "loss": 0.0547, "step": 116200 }, { "epoch": 9.63, "learning_rate": 1.5251288445552786e-05, "loss": 0.0605, "step": 116300 }, { "epoch": 9.64, "learning_rate": 1.5247132169576061e-05, "loss": 0.0562, "step": 116400 }, { "epoch": 9.64, "learning_rate": 1.5243017456359103e-05, "loss": 0.0559, "step": 116500 }, { "epoch": 9.64, "eval_loss": 0.1374831348657608, "eval_runtime": 4306.5678, "eval_samples_per_second": 17.553, "eval_steps_per_second": 1.463, "eval_wer": 0.058104525993036935, "step": 116500 }, { "epoch": 9.65, "learning_rate": 1.5238861180382379e-05, "loss": 0.0682, "step": 116600 }, { "epoch": 9.66, "learning_rate": 1.5234704904405653e-05, "loss": 0.0602, "step": 116700 }, { "epoch": 9.67, "learning_rate": 1.523054862842893e-05, "loss": 0.057, "step": 116800 }, { "epoch": 9.68, "learning_rate": 1.5226392352452204e-05, "loss": 0.0534, "step": 116900 }, { "epoch": 9.69, "learning_rate": 1.5222236076475479e-05, "loss": 0.0557, "step": 117000 }, { "epoch": 9.69, "eval_loss": 0.12481043487787247, "eval_runtime": 4355.4819, "eval_samples_per_second": 17.356, "eval_steps_per_second": 1.446, "eval_wer": 0.057984526177652033, "step": 117000 }, { "epoch": 9.69, "learning_rate": 1.5218079800498755e-05, "loss": 0.0575, "step": 117100 }, { "epoch": 9.7, "learning_rate": 1.521392352452203e-05, "loss": 0.0625, "step": 117200 }, { "epoch": 9.71, "learning_rate": 1.5209767248545304e-05, "loss": 0.0571, "step": 117300 }, { "epoch": 9.72, "learning_rate": 1.5205610972568581e-05, "loss": 0.0584, "step": 117400 }, { "epoch": 9.73, "learning_rate": 1.5201454696591855e-05, "loss": 0.0651, "step": 117500 }, { "epoch": 9.73, "eval_loss": 0.13871222734451294, "eval_runtime": 4296.8755, "eval_samples_per_second": 17.593, "eval_steps_per_second": 1.466, "eval_wer": 0.06031990720014277, "step": 117500 }, { "epoch": 9.74, "learning_rate": 1.519729842061513e-05, "loss": 0.0552, "step": 117600 }, { "epoch": 9.74, "learning_rate": 1.5193142144638404e-05, "loss": 0.0541, "step": 117700 }, { "epoch": 9.75, "learning_rate": 1.5188985868661682e-05, "loss": 0.0616, "step": 117800 }, { "epoch": 9.76, "learning_rate": 1.5184829592684955e-05, "loss": 0.0571, "step": 117900 }, { "epoch": 9.77, "learning_rate": 1.518067331670823e-05, "loss": 0.0582, "step": 118000 }, { "epoch": 9.77, "eval_loss": 0.12597554922103882, "eval_runtime": 4282.1712, "eval_samples_per_second": 17.653, "eval_steps_per_second": 1.471, "eval_wer": 0.057661449751615766, "step": 118000 }, { "epoch": 9.78, "learning_rate": 1.5176517040731506e-05, "loss": 0.0505, "step": 118100 }, { "epoch": 9.78, "learning_rate": 1.517236076475478e-05, "loss": 0.0605, "step": 118200 }, { "epoch": 9.79, "learning_rate": 1.5168204488778056e-05, "loss": 0.0543, "step": 118300 }, { "epoch": 9.8, "learning_rate": 1.5164048212801333e-05, "loss": 0.057, "step": 118400 }, { "epoch": 9.81, "learning_rate": 1.5159891936824607e-05, "loss": 0.0512, "step": 118500 }, { "epoch": 9.81, "eval_loss": 0.13426144421100616, "eval_runtime": 4298.6873, "eval_samples_per_second": 17.586, "eval_steps_per_second": 1.466, "eval_wer": 0.06004606146759774, "step": 118500 }, { "epoch": 9.82, "learning_rate": 1.515573566084788e-05, "loss": 0.0582, "step": 118600 }, { "epoch": 9.83, "learning_rate": 1.5151620947630922e-05, "loss": 0.0578, "step": 118700 }, { "epoch": 9.83, "learning_rate": 1.51474646716542e-05, "loss": 0.052, "step": 118800 }, { "epoch": 9.84, "learning_rate": 1.5143308395677473e-05, "loss": 0.0545, "step": 118900 }, { "epoch": 9.85, "learning_rate": 1.5139152119700749e-05, "loss": 0.061, "step": 119000 }, { "epoch": 9.85, "eval_loss": 0.13382236659526825, "eval_runtime": 4290.7168, "eval_samples_per_second": 17.618, "eval_steps_per_second": 1.468, "eval_wer": 0.0592645242084243, "step": 119000 }, { "epoch": 9.86, "learning_rate": 1.5134995843724024e-05, "loss": 0.0553, "step": 119100 }, { "epoch": 9.87, "learning_rate": 1.51308395677473e-05, "loss": 0.0604, "step": 119200 }, { "epoch": 9.88, "learning_rate": 1.5126683291770574e-05, "loss": 0.0555, "step": 119300 }, { "epoch": 9.88, "learning_rate": 1.5122527015793851e-05, "loss": 0.0548, "step": 119400 }, { "epoch": 9.89, "learning_rate": 1.5118370739817125e-05, "loss": 0.0628, "step": 119500 }, { "epoch": 9.89, "eval_loss": 0.13998854160308838, "eval_runtime": 4297.6447, "eval_samples_per_second": 17.59, "eval_steps_per_second": 1.466, "eval_wer": 0.059218370433276254, "step": 119500 }, { "epoch": 9.9, "learning_rate": 1.51142144638404e-05, "loss": 0.0554, "step": 119600 }, { "epoch": 9.91, "learning_rate": 1.5110058187863674e-05, "loss": 0.0627, "step": 119700 }, { "epoch": 9.92, "learning_rate": 1.5105901911886951e-05, "loss": 0.0533, "step": 119800 }, { "epoch": 9.93, "learning_rate": 1.5101745635910225e-05, "loss": 0.067, "step": 119900 }, { "epoch": 9.93, "learning_rate": 1.50975893599335e-05, "loss": 0.0605, "step": 120000 }, { "epoch": 9.93, "eval_loss": 0.14214302599430084, "eval_runtime": 4313.53, "eval_samples_per_second": 17.525, "eval_steps_per_second": 1.461, "eval_wer": 0.06019067662972826, "step": 120000 }, { "epoch": 9.94, "learning_rate": 1.5093433083956776e-05, "loss": 0.0586, "step": 120100 }, { "epoch": 9.95, "learning_rate": 1.5089276807980052e-05, "loss": 0.0628, "step": 120200 }, { "epoch": 9.96, "learning_rate": 1.5085120532003326e-05, "loss": 0.0572, "step": 120300 }, { "epoch": 9.97, "learning_rate": 1.5080964256026603e-05, "loss": 0.0534, "step": 120400 }, { "epoch": 9.98, "learning_rate": 1.5076807980049877e-05, "loss": 0.0541, "step": 120500 }, { "epoch": 9.98, "eval_loss": 0.12559331953525543, "eval_runtime": 4333.0941, "eval_samples_per_second": 17.446, "eval_steps_per_second": 1.454, "eval_wer": 0.057176835112561365, "step": 120500 }, { "epoch": 9.98, "learning_rate": 1.507265170407315e-05, "loss": 0.0642, "step": 120600 }, { "epoch": 9.99, "learning_rate": 1.5068495428096426e-05, "loss": 0.0591, "step": 120700 }, { "epoch": 10.0, "learning_rate": 1.5064339152119703e-05, "loss": 0.0591, "step": 120800 }, { "epoch": 10.01, "learning_rate": 1.5060182876142977e-05, "loss": 0.0509, "step": 120900 }, { "epoch": 10.02, "learning_rate": 1.505602660016625e-05, "loss": 0.0568, "step": 121000 }, { "epoch": 10.02, "eval_loss": 0.13634726405143738, "eval_runtime": 4315.633, "eval_samples_per_second": 17.517, "eval_steps_per_second": 1.46, "eval_wer": 0.05889529400723999, "step": 121000 }, { "epoch": 10.02, "learning_rate": 1.5051870324189528e-05, "loss": 0.0622, "step": 121100 }, { "epoch": 10.03, "learning_rate": 1.5047714048212802e-05, "loss": 0.0558, "step": 121200 }, { "epoch": 10.04, "learning_rate": 1.5043557772236077e-05, "loss": 0.0547, "step": 121300 }, { "epoch": 10.05, "learning_rate": 1.5039401496259355e-05, "loss": 0.0498, "step": 121400 }, { "epoch": 10.06, "learning_rate": 1.5035245220282628e-05, "loss": 0.0537, "step": 121500 }, { "epoch": 10.06, "eval_loss": 0.1358143836259842, "eval_runtime": 4319.5418, "eval_samples_per_second": 17.501, "eval_steps_per_second": 1.458, "eval_wer": 0.058659909753984996, "step": 121500 }, { "epoch": 10.07, "learning_rate": 1.5031088944305902e-05, "loss": 0.0498, "step": 121600 }, { "epoch": 10.07, "learning_rate": 1.5026932668329178e-05, "loss": 0.0548, "step": 121700 }, { "epoch": 10.08, "learning_rate": 1.5022776392352453e-05, "loss": 0.048, "step": 121800 }, { "epoch": 10.09, "learning_rate": 1.5018620116375729e-05, "loss": 0.0562, "step": 121900 }, { "epoch": 10.1, "learning_rate": 1.5014463840399003e-05, "loss": 0.0505, "step": 122000 }, { "epoch": 10.1, "eval_loss": 0.1300382912158966, "eval_runtime": 4295.9467, "eval_samples_per_second": 17.597, "eval_steps_per_second": 1.466, "eval_wer": 0.056484528485340794, "step": 122000 }, { "epoch": 10.11, "learning_rate": 1.5010349127182046e-05, "loss": 0.0538, "step": 122100 }, { "epoch": 10.12, "learning_rate": 1.5006192851205322e-05, "loss": 0.0557, "step": 122200 }, { "epoch": 10.12, "learning_rate": 1.5002036575228595e-05, "loss": 0.0558, "step": 122300 }, { "epoch": 10.13, "learning_rate": 1.4997880299251873e-05, "loss": 0.0483, "step": 122400 }, { "epoch": 10.14, "learning_rate": 1.4993724023275146e-05, "loss": 0.0545, "step": 122500 }, { "epoch": 10.14, "eval_loss": 0.13645365834236145, "eval_runtime": 4297.572, "eval_samples_per_second": 17.59, "eval_steps_per_second": 1.466, "eval_wer": 0.06152759764984977, "step": 122500 }, { "epoch": 10.15, "learning_rate": 1.4989567747298422e-05, "loss": 0.0513, "step": 122600 }, { "epoch": 10.16, "learning_rate": 1.4985411471321698e-05, "loss": 0.046, "step": 122700 }, { "epoch": 10.17, "learning_rate": 1.4981255195344973e-05, "loss": 0.052, "step": 122800 }, { "epoch": 10.17, "learning_rate": 1.4977098919368247e-05, "loss": 0.0513, "step": 122900 }, { "epoch": 10.18, "learning_rate": 1.497294264339152e-05, "loss": 0.0541, "step": 123000 }, { "epoch": 10.18, "eval_loss": 0.13865052163600922, "eval_runtime": 4297.6826, "eval_samples_per_second": 17.59, "eval_steps_per_second": 1.466, "eval_wer": 0.05972913887824788, "step": 123000 }, { "epoch": 10.19, "learning_rate": 1.4968786367414798e-05, "loss": 0.0514, "step": 123100 }, { "epoch": 10.2, "learning_rate": 1.4964630091438073e-05, "loss": 0.0516, "step": 123200 }, { "epoch": 10.21, "learning_rate": 1.4960473815461347e-05, "loss": 0.0566, "step": 123300 }, { "epoch": 10.22, "learning_rate": 1.4956317539484624e-05, "loss": 0.0488, "step": 123400 }, { "epoch": 10.22, "learning_rate": 1.4952161263507898e-05, "loss": 0.0472, "step": 123500 }, { "epoch": 10.22, "eval_loss": 0.1292572021484375, "eval_runtime": 4300.0088, "eval_samples_per_second": 17.58, "eval_steps_per_second": 1.465, "eval_wer": 0.05779221878120187, "step": 123500 }, { "epoch": 10.23, "learning_rate": 1.4948004987531172e-05, "loss": 0.0566, "step": 123600 }, { "epoch": 10.24, "learning_rate": 1.494384871155445e-05, "loss": 0.0561, "step": 123700 }, { "epoch": 10.25, "learning_rate": 1.4939692435577725e-05, "loss": 0.0555, "step": 123800 }, { "epoch": 10.26, "learning_rate": 1.4935536159600999e-05, "loss": 0.0494, "step": 123900 }, { "epoch": 10.26, "learning_rate": 1.4931379883624272e-05, "loss": 0.0494, "step": 124000 }, { "epoch": 10.26, "eval_loss": 0.12953031063079834, "eval_runtime": 4322.7914, "eval_samples_per_second": 17.488, "eval_steps_per_second": 1.457, "eval_wer": 0.05909990907706296, "step": 124000 }, { "epoch": 10.27, "learning_rate": 1.492722360764755e-05, "loss": 0.0569, "step": 124100 }, { "epoch": 10.28, "learning_rate": 1.4923067331670823e-05, "loss": 0.0598, "step": 124200 }, { "epoch": 10.29, "learning_rate": 1.4918911055694099e-05, "loss": 0.0559, "step": 124300 }, { "epoch": 10.3, "learning_rate": 1.4914754779717375e-05, "loss": 0.0509, "step": 124400 }, { "epoch": 10.31, "learning_rate": 1.491059850374065e-05, "loss": 0.0566, "step": 124500 }, { "epoch": 10.31, "eval_loss": 0.14169026911258698, "eval_runtime": 4306.1968, "eval_samples_per_second": 17.555, "eval_steps_per_second": 1.463, "eval_wer": 0.06043529163801287, "step": 124500 }, { "epoch": 10.31, "learning_rate": 1.4906442227763924e-05, "loss": 0.049, "step": 124600 }, { "epoch": 10.32, "learning_rate": 1.4902285951787201e-05, "loss": 0.0664, "step": 124700 }, { "epoch": 10.33, "learning_rate": 1.4898129675810475e-05, "loss": 0.0535, "step": 124800 }, { "epoch": 10.34, "learning_rate": 1.489397339983375e-05, "loss": 0.0566, "step": 124900 }, { "epoch": 10.35, "learning_rate": 1.4889817123857024e-05, "loss": 0.0497, "step": 125000 }, { "epoch": 10.35, "eval_loss": 0.14685194194316864, "eval_runtime": 4290.4515, "eval_samples_per_second": 17.619, "eval_steps_per_second": 1.468, "eval_wer": 0.058067602972918506, "step": 125000 }, { "epoch": 10.36, "learning_rate": 1.4885660847880301e-05, "loss": 0.0632, "step": 125100 }, { "epoch": 10.36, "learning_rate": 1.4881504571903575e-05, "loss": 0.0542, "step": 125200 }, { "epoch": 10.37, "learning_rate": 1.487734829592685e-05, "loss": 0.0584, "step": 125300 }, { "epoch": 10.38, "learning_rate": 1.4873192019950126e-05, "loss": 0.0588, "step": 125400 }, { "epoch": 10.39, "learning_rate": 1.4869035743973402e-05, "loss": 0.0519, "step": 125500 }, { "epoch": 10.39, "eval_loss": 0.13356956839561462, "eval_runtime": 4297.5221, "eval_samples_per_second": 17.59, "eval_steps_per_second": 1.466, "eval_wer": 0.05773683425102423, "step": 125500 }, { "epoch": 10.4, "learning_rate": 1.4864879467996676e-05, "loss": 0.0537, "step": 125600 }, { "epoch": 10.41, "learning_rate": 1.4860723192019953e-05, "loss": 0.0572, "step": 125700 }, { "epoch": 10.41, "learning_rate": 1.4856566916043227e-05, "loss": 0.0487, "step": 125800 }, { "epoch": 10.42, "learning_rate": 1.4852410640066502e-05, "loss": 0.0564, "step": 125900 }, { "epoch": 10.43, "learning_rate": 1.4848254364089776e-05, "loss": 0.0467, "step": 126000 }, { "epoch": 10.43, "eval_loss": 0.1457975059747696, "eval_runtime": 4304.9238, "eval_samples_per_second": 17.56, "eval_steps_per_second": 1.463, "eval_wer": 0.06117836741789628, "step": 126000 }, { "epoch": 10.44, "learning_rate": 1.4844098088113053e-05, "loss": 0.0523, "step": 126100 }, { "epoch": 10.45, "learning_rate": 1.4839941812136327e-05, "loss": 0.0525, "step": 126200 }, { "epoch": 10.46, "learning_rate": 1.4835785536159601e-05, "loss": 0.0509, "step": 126300 }, { "epoch": 10.46, "learning_rate": 1.4831629260182878e-05, "loss": 0.0507, "step": 126400 }, { "epoch": 10.47, "learning_rate": 1.4827472984206152e-05, "loss": 0.0547, "step": 126500 }, { "epoch": 10.47, "eval_loss": 0.14238734543323517, "eval_runtime": 4304.8213, "eval_samples_per_second": 17.561, "eval_steps_per_second": 1.463, "eval_wer": 0.06094144470546969, "step": 126500 }, { "epoch": 10.48, "learning_rate": 1.4823316708229427e-05, "loss": 0.0602, "step": 126600 }, { "epoch": 10.49, "learning_rate": 1.4819160432252705e-05, "loss": 0.0572, "step": 126700 }, { "epoch": 10.5, "learning_rate": 1.4815004156275978e-05, "loss": 0.0514, "step": 126800 }, { "epoch": 10.5, "learning_rate": 1.4810847880299252e-05, "loss": 0.051, "step": 126900 }, { "epoch": 10.51, "learning_rate": 1.4806691604322528e-05, "loss": 0.0484, "step": 127000 }, { "epoch": 10.51, "eval_loss": 0.12180278450250626, "eval_runtime": 4308.1548, "eval_samples_per_second": 17.547, "eval_steps_per_second": 1.462, "eval_wer": 0.056656835912560134, "step": 127000 }, { "epoch": 10.52, "learning_rate": 1.4802535328345803e-05, "loss": 0.0522, "step": 127100 }, { "epoch": 10.53, "learning_rate": 1.4798379052369079e-05, "loss": 0.0573, "step": 127200 }, { "epoch": 10.54, "learning_rate": 1.4794222776392353e-05, "loss": 0.0558, "step": 127300 }, { "epoch": 10.55, "learning_rate": 1.479006650041563e-05, "loss": 0.048, "step": 127400 }, { "epoch": 10.55, "learning_rate": 1.4785910224438904e-05, "loss": 0.0541, "step": 127500 }, { "epoch": 10.55, "eval_loss": 0.12812846899032593, "eval_runtime": 4283.1468, "eval_samples_per_second": 17.649, "eval_steps_per_second": 1.471, "eval_wer": 0.0580645260545753, "step": 127500 }, { "epoch": 10.56, "learning_rate": 1.478175394846218e-05, "loss": 0.0496, "step": 127600 }, { "epoch": 10.57, "learning_rate": 1.4777597672485455e-05, "loss": 0.0536, "step": 127700 }, { "epoch": 10.58, "learning_rate": 1.477344139650873e-05, "loss": 0.055, "step": 127800 }, { "epoch": 10.59, "learning_rate": 1.4769285120532004e-05, "loss": 0.0541, "step": 127900 }, { "epoch": 10.6, "learning_rate": 1.476512884455528e-05, "loss": 0.0518, "step": 128000 }, { "epoch": 10.6, "eval_loss": 0.12462115287780762, "eval_runtime": 4277.626, "eval_samples_per_second": 17.672, "eval_steps_per_second": 1.473, "eval_wer": 0.056898374002501535, "step": 128000 }, { "epoch": 10.6, "learning_rate": 1.4761014131338323e-05, "loss": 0.054, "step": 128100 }, { "epoch": 10.61, "learning_rate": 1.4756857855361597e-05, "loss": 0.0545, "step": 128200 }, { "epoch": 10.62, "learning_rate": 1.4752701579384872e-05, "loss": 0.0579, "step": 128300 }, { "epoch": 10.63, "learning_rate": 1.4748545303408148e-05, "loss": 0.0509, "step": 128400 }, { "epoch": 10.64, "learning_rate": 1.4744389027431423e-05, "loss": 0.0542, "step": 128500 }, { "epoch": 10.64, "eval_loss": 0.13268114626407623, "eval_runtime": 4289.8982, "eval_samples_per_second": 17.622, "eval_steps_per_second": 1.469, "eval_wer": 0.057790680322030274, "step": 128500 }, { "epoch": 10.65, "learning_rate": 1.4740232751454697e-05, "loss": 0.0556, "step": 128600 }, { "epoch": 10.65, "learning_rate": 1.4736076475477974e-05, "loss": 0.0543, "step": 128700 }, { "epoch": 10.66, "learning_rate": 1.4731920199501248e-05, "loss": 0.0561, "step": 128800 }, { "epoch": 10.67, "learning_rate": 1.4727763923524522e-05, "loss": 0.0535, "step": 128900 }, { "epoch": 10.68, "learning_rate": 1.4723607647547798e-05, "loss": 0.0553, "step": 129000 }, { "epoch": 10.68, "eval_loss": 0.13587293028831482, "eval_runtime": 4292.1367, "eval_samples_per_second": 17.612, "eval_steps_per_second": 1.468, "eval_wer": 0.05804914146285929, "step": 129000 }, { "epoch": 10.69, "learning_rate": 1.4719492934330841e-05, "loss": 0.0624, "step": 129100 }, { "epoch": 10.7, "learning_rate": 1.4715336658354115e-05, "loss": 0.0526, "step": 129200 }, { "epoch": 10.7, "learning_rate": 1.471118038237739e-05, "loss": 0.0622, "step": 129300 }, { "epoch": 10.71, "learning_rate": 1.4707024106400668e-05, "loss": 0.0561, "step": 129400 }, { "epoch": 10.72, "learning_rate": 1.4702867830423942e-05, "loss": 0.0567, "step": 129500 }, { "epoch": 10.72, "eval_loss": 0.12789109349250793, "eval_runtime": 4269.6673, "eval_samples_per_second": 17.705, "eval_steps_per_second": 1.476, "eval_wer": 0.05837221788889556, "step": 129500 }, { "epoch": 10.73, "learning_rate": 1.4698711554447215e-05, "loss": 0.0521, "step": 129600 }, { "epoch": 10.74, "learning_rate": 1.4694555278470493e-05, "loss": 0.0562, "step": 129700 }, { "epoch": 10.75, "learning_rate": 1.4690399002493766e-05, "loss": 0.0527, "step": 129800 }, { "epoch": 10.75, "learning_rate": 1.4686242726517042e-05, "loss": 0.0486, "step": 129900 }, { "epoch": 10.76, "learning_rate": 1.4682086450540316e-05, "loss": 0.047, "step": 130000 }, { "epoch": 10.76, "eval_loss": 0.13897189497947693, "eval_runtime": 4288.6094, "eval_samples_per_second": 17.627, "eval_steps_per_second": 1.469, "eval_wer": 0.05899837077173727, "step": 130000 }, { "epoch": 10.77, "learning_rate": 1.4677930174563593e-05, "loss": 0.0538, "step": 130100 }, { "epoch": 10.78, "learning_rate": 1.4673773898586867e-05, "loss": 0.0601, "step": 130200 }, { "epoch": 10.79, "learning_rate": 1.4669617622610142e-05, "loss": 0.0537, "step": 130300 }, { "epoch": 10.79, "learning_rate": 1.4665461346633418e-05, "loss": 0.0552, "step": 130400 }, { "epoch": 10.8, "learning_rate": 1.4661305070656693e-05, "loss": 0.0494, "step": 130500 }, { "epoch": 10.8, "eval_loss": 0.1310400515794754, "eval_runtime": 4281.1005, "eval_samples_per_second": 17.658, "eval_steps_per_second": 1.472, "eval_wer": 0.05865837129481339, "step": 130500 }, { "epoch": 10.81, "learning_rate": 1.4657148794679967e-05, "loss": 0.0552, "step": 130600 }, { "epoch": 10.82, "learning_rate": 1.4652992518703244e-05, "loss": 0.0661, "step": 130700 }, { "epoch": 10.83, "learning_rate": 1.4648836242726518e-05, "loss": 0.0573, "step": 130800 }, { "epoch": 10.84, "learning_rate": 1.464472152950956e-05, "loss": 0.0544, "step": 130900 }, { "epoch": 10.84, "learning_rate": 1.4640565253532837e-05, "loss": 0.0548, "step": 131000 }, { "epoch": 10.84, "eval_loss": 0.13375192880630493, "eval_runtime": 4270.3662, "eval_samples_per_second": 17.702, "eval_steps_per_second": 1.475, "eval_wer": 0.056492220781198796, "step": 131000 }, { "epoch": 10.85, "learning_rate": 1.4636408977556111e-05, "loss": 0.0575, "step": 131100 }, { "epoch": 10.86, "learning_rate": 1.4632252701579387e-05, "loss": 0.0497, "step": 131200 }, { "epoch": 10.87, "learning_rate": 1.462809642560266e-05, "loss": 0.064, "step": 131300 }, { "epoch": 10.88, "learning_rate": 1.4623940149625938e-05, "loss": 0.0541, "step": 131400 }, { "epoch": 10.89, "learning_rate": 1.4619783873649211e-05, "loss": 0.0596, "step": 131500 }, { "epoch": 10.89, "eval_loss": 0.12919747829437256, "eval_runtime": 4477.9839, "eval_samples_per_second": 16.881, "eval_steps_per_second": 1.407, "eval_wer": 0.05645375930190877, "step": 131500 }, { "epoch": 10.89, "learning_rate": 1.4615627597672485e-05, "loss": 0.0531, "step": 131600 }, { "epoch": 10.9, "learning_rate": 1.4611471321695762e-05, "loss": 0.0463, "step": 131700 }, { "epoch": 10.91, "learning_rate": 1.4607315045719038e-05, "loss": 0.0546, "step": 131800 }, { "epoch": 10.92, "learning_rate": 1.4603200332502078e-05, "loss": 0.0544, "step": 131900 }, { "epoch": 10.93, "learning_rate": 1.4599044056525355e-05, "loss": 0.0576, "step": 132000 }, { "epoch": 10.93, "eval_loss": 0.1348622739315033, "eval_runtime": 4262.7912, "eval_samples_per_second": 17.734, "eval_steps_per_second": 1.478, "eval_wer": 0.058453756224990425, "step": 132000 }, { "epoch": 10.94, "learning_rate": 1.459488778054863e-05, "loss": 0.0506, "step": 132100 }, { "epoch": 10.94, "learning_rate": 1.4590731504571905e-05, "loss": 0.0563, "step": 132200 }, { "epoch": 10.95, "learning_rate": 1.4586575228595178e-05, "loss": 0.0533, "step": 132300 }, { "epoch": 10.96, "learning_rate": 1.4582418952618456e-05, "loss": 0.0574, "step": 132400 }, { "epoch": 10.97, "learning_rate": 1.457826267664173e-05, "loss": 0.055, "step": 132500 }, { "epoch": 10.97, "eval_loss": 0.13642220199108124, "eval_runtime": 4286.6049, "eval_samples_per_second": 17.635, "eval_steps_per_second": 1.47, "eval_wer": 0.05751529613031364, "step": 132500 }, { "epoch": 10.98, "learning_rate": 1.4574106400665005e-05, "loss": 0.0527, "step": 132600 }, { "epoch": 10.99, "learning_rate": 1.456995012468828e-05, "loss": 0.0583, "step": 132700 }, { "epoch": 10.99, "learning_rate": 1.4565793848711556e-05, "loss": 0.0541, "step": 132800 }, { "epoch": 11.0, "learning_rate": 1.456163757273483e-05, "loss": 0.0457, "step": 132900 }, { "epoch": 11.01, "learning_rate": 1.4557481296758107e-05, "loss": 0.0496, "step": 133000 }, { "epoch": 11.01, "eval_loss": 0.13280907273292542, "eval_runtime": 4441.6729, "eval_samples_per_second": 17.019, "eval_steps_per_second": 1.418, "eval_wer": 0.0570999121539813, "step": 133000 }, { "epoch": 11.02, "learning_rate": 1.4553325020781381e-05, "loss": 0.0447, "step": 133100 }, { "epoch": 11.03, "learning_rate": 1.4549168744804656e-05, "loss": 0.0503, "step": 133200 }, { "epoch": 11.03, "learning_rate": 1.454501246882793e-05, "loss": 0.0481, "step": 133300 }, { "epoch": 11.04, "learning_rate": 1.4540856192851207e-05, "loss": 0.0523, "step": 133400 }, { "epoch": 11.05, "learning_rate": 1.4536699916874481e-05, "loss": 0.0537, "step": 133500 }, { "epoch": 11.05, "eval_loss": 0.13865238428115845, "eval_runtime": 4356.0636, "eval_samples_per_second": 17.354, "eval_steps_per_second": 1.446, "eval_wer": 0.0569768354202532, "step": 133500 }, { "epoch": 11.06, "learning_rate": 1.4532543640897757e-05, "loss": 0.0464, "step": 133600 }, { "epoch": 11.07, "learning_rate": 1.4528387364921032e-05, "loss": 0.0526, "step": 133700 }, { "epoch": 11.08, "learning_rate": 1.4524231088944308e-05, "loss": 0.0453, "step": 133800 }, { "epoch": 11.08, "learning_rate": 1.4520074812967582e-05, "loss": 0.0542, "step": 133900 }, { "epoch": 11.09, "learning_rate": 1.4515918536990859e-05, "loss": 0.0526, "step": 134000 }, { "epoch": 11.09, "eval_loss": 0.1281552016735077, "eval_runtime": 4559.7914, "eval_samples_per_second": 16.579, "eval_steps_per_second": 1.382, "eval_wer": 0.05633068256818066, "step": 134000 }, { "epoch": 11.1, "learning_rate": 1.4511762261014133e-05, "loss": 0.0569, "step": 134100 }, { "epoch": 11.11, "learning_rate": 1.4507605985037408e-05, "loss": 0.0495, "step": 134200 }, { "epoch": 11.12, "learning_rate": 1.4503449709060682e-05, "loss": 0.0454, "step": 134300 }, { "epoch": 11.13, "learning_rate": 1.449929343308396e-05, "loss": 0.0459, "step": 134400 }, { "epoch": 11.13, "learning_rate": 1.4495137157107233e-05, "loss": 0.0481, "step": 134500 }, { "epoch": 11.13, "eval_loss": 0.13253799080848694, "eval_runtime": 4400.0945, "eval_samples_per_second": 17.18, "eval_steps_per_second": 1.432, "eval_wer": 0.05700145076699882, "step": 134500 }, { "epoch": 11.14, "learning_rate": 1.4490980881130507e-05, "loss": 0.052, "step": 134600 }, { "epoch": 11.15, "learning_rate": 1.4486824605153784e-05, "loss": 0.0478, "step": 134700 }, { "epoch": 11.16, "learning_rate": 1.4482668329177058e-05, "loss": 0.0484, "step": 134800 }, { "epoch": 11.17, "learning_rate": 1.4478512053200333e-05, "loss": 0.0477, "step": 134900 }, { "epoch": 11.18, "learning_rate": 1.447435577722361e-05, "loss": 0.0553, "step": 135000 }, { "epoch": 11.18, "eval_loss": 0.13499818742275238, "eval_runtime": 4322.909, "eval_samples_per_second": 17.487, "eval_steps_per_second": 1.457, "eval_wer": 0.05720760429599339, "step": 135000 }, { "epoch": 11.18, "learning_rate": 1.4470199501246884e-05, "loss": 0.0566, "step": 135100 }, { "epoch": 11.19, "learning_rate": 1.4466043225270158e-05, "loss": 0.05, "step": 135200 }, { "epoch": 11.2, "learning_rate": 1.4461886949293434e-05, "loss": 0.0545, "step": 135300 }, { "epoch": 11.21, "learning_rate": 1.445773067331671e-05, "loss": 0.0495, "step": 135400 }, { "epoch": 11.22, "learning_rate": 1.4453574397339985e-05, "loss": 0.0496, "step": 135500 }, { "epoch": 11.22, "eval_loss": 0.13316142559051514, "eval_runtime": 4332.813, "eval_samples_per_second": 17.447, "eval_steps_per_second": 1.454, "eval_wer": 0.05672606657528219, "step": 135500 }, { "epoch": 11.23, "learning_rate": 1.4449418121363259e-05, "loss": 0.0525, "step": 135600 }, { "epoch": 11.23, "learning_rate": 1.4445261845386536e-05, "loss": 0.0449, "step": 135700 }, { "epoch": 11.24, "learning_rate": 1.444110556940981e-05, "loss": 0.054, "step": 135800 }, { "epoch": 11.25, "learning_rate": 1.4436949293433085e-05, "loss": 0.0504, "step": 135900 }, { "epoch": 11.26, "learning_rate": 1.443279301745636e-05, "loss": 0.0471, "step": 136000 }, { "epoch": 11.26, "eval_loss": 0.1493394523859024, "eval_runtime": 5481.991, "eval_samples_per_second": 13.79, "eval_steps_per_second": 1.149, "eval_wer": 0.059167601280613415, "step": 136000 }, { "epoch": 11.27, "learning_rate": 1.4428636741479636e-05, "loss": 0.0503, "step": 136100 }, { "epoch": 11.27, "learning_rate": 1.442448046550291e-05, "loss": 0.0462, "step": 136200 }, { "epoch": 11.28, "learning_rate": 1.4420324189526186e-05, "loss": 0.0515, "step": 136300 }, { "epoch": 11.29, "learning_rate": 1.4416167913549461e-05, "loss": 0.0521, "step": 136400 }, { "epoch": 11.3, "learning_rate": 1.4412011637572737e-05, "loss": 0.0518, "step": 136500 }, { "epoch": 11.3, "eval_loss": 0.1276169717311859, "eval_runtime": 6657.8221, "eval_samples_per_second": 11.354, "eval_steps_per_second": 0.946, "eval_wer": 0.05539529939184709, "step": 136500 }, { "epoch": 11.31, "learning_rate": 1.440785536159601e-05, "loss": 0.048, "step": 136600 }, { "epoch": 11.32, "learning_rate": 1.4403699085619288e-05, "loss": 0.0458, "step": 136700 }, { "epoch": 11.32, "learning_rate": 1.4399542809642561e-05, "loss": 0.0489, "step": 136800 }, { "epoch": 11.33, "learning_rate": 1.4395386533665835e-05, "loss": 0.0458, "step": 136900 }, { "epoch": 11.34, "learning_rate": 1.4391230257689112e-05, "loss": 0.0513, "step": 137000 }, { "epoch": 11.34, "eval_loss": 0.14221999049186707, "eval_runtime": 4318.354, "eval_samples_per_second": 17.506, "eval_steps_per_second": 1.459, "eval_wer": 0.05703683532794565, "step": 137000 }, { "epoch": 11.35, "learning_rate": 1.4387073981712388e-05, "loss": 0.0433, "step": 137100 }, { "epoch": 11.36, "learning_rate": 1.4382917705735662e-05, "loss": 0.0529, "step": 137200 }, { "epoch": 11.37, "learning_rate": 1.4378761429758936e-05, "loss": 0.0537, "step": 137300 }, { "epoch": 11.37, "learning_rate": 1.4374605153782213e-05, "loss": 0.0442, "step": 137400 }, { "epoch": 11.38, "learning_rate": 1.4370448877805487e-05, "loss": 0.0468, "step": 137500 }, { "epoch": 11.38, "eval_loss": 0.13952383399009705, "eval_runtime": 4807.5747, "eval_samples_per_second": 15.724, "eval_steps_per_second": 1.31, "eval_wer": 0.056776835727945035, "step": 137500 }, { "epoch": 11.39, "learning_rate": 1.4366292601828762e-05, "loss": 0.0504, "step": 137600 }, { "epoch": 11.4, "learning_rate": 1.436213632585204e-05, "loss": 0.0515, "step": 137700 }, { "epoch": 11.41, "learning_rate": 1.4357980049875313e-05, "loss": 0.0485, "step": 137800 }, { "epoch": 11.42, "learning_rate": 1.4353823773898587e-05, "loss": 0.053, "step": 137900 }, { "epoch": 11.42, "learning_rate": 1.434970906068163e-05, "loss": 0.0538, "step": 138000 }, { "epoch": 11.42, "eval_loss": 0.13265180587768555, "eval_runtime": 6428.6127, "eval_samples_per_second": 11.759, "eval_steps_per_second": 0.98, "eval_wer": 0.05725375807114143, "step": 138000 }, { "epoch": 11.43, "learning_rate": 1.4345552784704906e-05, "loss": 0.0519, "step": 138100 }, { "epoch": 11.44, "learning_rate": 1.434139650872818e-05, "loss": 0.0521, "step": 138200 }, { "epoch": 11.45, "learning_rate": 1.4337240232751455e-05, "loss": 0.0487, "step": 138300 }, { "epoch": 11.46, "learning_rate": 1.4333083956774731e-05, "loss": 0.0431, "step": 138400 }, { "epoch": 11.47, "learning_rate": 1.4328927680798006e-05, "loss": 0.0445, "step": 138500 }, { "epoch": 11.47, "eval_loss": 0.1409018188714981, "eval_runtime": 6321.6147, "eval_samples_per_second": 11.958, "eval_steps_per_second": 0.997, "eval_wer": 0.05544606854450993, "step": 138500 }, { "epoch": 11.47, "learning_rate": 1.432477140482128e-05, "loss": 0.0521, "step": 138600 }, { "epoch": 11.48, "learning_rate": 1.4320615128844557e-05, "loss": 0.0455, "step": 138700 }, { "epoch": 11.49, "learning_rate": 1.4316458852867831e-05, "loss": 0.0481, "step": 138800 }, { "epoch": 11.5, "learning_rate": 1.4312302576891107e-05, "loss": 0.0473, "step": 138900 }, { "epoch": 11.51, "learning_rate": 1.4308146300914382e-05, "loss": 0.0473, "step": 139000 }, { "epoch": 11.51, "eval_loss": 0.14671126008033752, "eval_runtime": 5636.0424, "eval_samples_per_second": 13.413, "eval_steps_per_second": 1.118, "eval_wer": 0.05847683311256444, "step": 139000 }, { "epoch": 11.51, "learning_rate": 1.4303990024937658e-05, "loss": 0.0505, "step": 139100 }, { "epoch": 11.52, "learning_rate": 1.4299833748960932e-05, "loss": 0.0523, "step": 139200 }, { "epoch": 11.53, "learning_rate": 1.4295677472984205e-05, "loss": 0.0514, "step": 139300 }, { "epoch": 11.54, "learning_rate": 1.4291521197007483e-05, "loss": 0.0486, "step": 139400 }, { "epoch": 11.55, "learning_rate": 1.4287364921030758e-05, "loss": 0.0556, "step": 139500 }, { "epoch": 11.55, "eval_loss": 0.15508776903152466, "eval_runtime": 6363.3174, "eval_samples_per_second": 11.88, "eval_steps_per_second": 0.99, "eval_wer": 0.059544523777655727, "step": 139500 }, { "epoch": 11.56, "learning_rate": 1.4283208645054032e-05, "loss": 0.0507, "step": 139600 }, { "epoch": 11.56, "learning_rate": 1.427905236907731e-05, "loss": 0.06, "step": 139700 }, { "epoch": 11.57, "learning_rate": 1.4274896093100583e-05, "loss": 0.0476, "step": 139800 }, { "epoch": 11.58, "learning_rate": 1.4270739817123857e-05, "loss": 0.051, "step": 139900 }, { "epoch": 11.59, "learning_rate": 1.4266583541147134e-05, "loss": 0.0468, "step": 140000 }, { "epoch": 11.59, "eval_loss": 0.13970208168029785, "eval_runtime": 6076.0887, "eval_samples_per_second": 12.441, "eval_steps_per_second": 1.037, "eval_wer": 0.05646145159776677, "step": 140000 }, { "epoch": 11.6, "learning_rate": 1.4262468827930176e-05, "loss": 0.0554, "step": 140100 }, { "epoch": 11.61, "learning_rate": 1.425831255195345e-05, "loss": 0.06, "step": 140200 }, { "epoch": 11.61, "learning_rate": 1.4254156275976725e-05, "loss": 0.0626, "step": 140300 }, { "epoch": 11.62, "learning_rate": 1.425e-05, "loss": 0.052, "step": 140400 }, { "epoch": 11.63, "learning_rate": 1.4245843724023276e-05, "loss": 0.0509, "step": 140500 }, { "epoch": 11.63, "eval_loss": 0.1370314508676529, "eval_runtime": 6045.4823, "eval_samples_per_second": 12.504, "eval_steps_per_second": 1.042, "eval_wer": 0.058469140816706434, "step": 140500 }, { "epoch": 11.64, "learning_rate": 1.424168744804655e-05, "loss": 0.0571, "step": 140600 }, { "epoch": 11.65, "learning_rate": 1.4237531172069827e-05, "loss": 0.0468, "step": 140700 }, { "epoch": 11.66, "learning_rate": 1.4233374896093101e-05, "loss": 0.0456, "step": 140800 }, { "epoch": 11.66, "learning_rate": 1.4229218620116377e-05, "loss": 0.0479, "step": 140900 }, { "epoch": 11.67, "learning_rate": 1.4225062344139652e-05, "loss": 0.0481, "step": 141000 }, { "epoch": 11.67, "eval_loss": 0.13343821465969086, "eval_runtime": 6303.8955, "eval_samples_per_second": 11.992, "eval_steps_per_second": 0.999, "eval_wer": 0.05789221862735596, "step": 141000 }, { "epoch": 11.68, "learning_rate": 1.4220906068162928e-05, "loss": 0.0501, "step": 141100 }, { "epoch": 11.69, "learning_rate": 1.4216749792186202e-05, "loss": 0.051, "step": 141200 }, { "epoch": 11.7, "learning_rate": 1.4212593516209479e-05, "loss": 0.0485, "step": 141300 }, { "epoch": 11.71, "learning_rate": 1.4208437240232753e-05, "loss": 0.049, "step": 141400 }, { "epoch": 11.71, "learning_rate": 1.4204322527015794e-05, "loss": 0.0499, "step": 141500 }, { "epoch": 11.71, "eval_loss": 0.12790416181087494, "eval_runtime": 6503.8782, "eval_samples_per_second": 11.623, "eval_steps_per_second": 0.969, "eval_wer": 0.056635297484157715, "step": 141500 }, { "epoch": 11.72, "learning_rate": 1.420016625103907e-05, "loss": 0.0496, "step": 141600 }, { "epoch": 11.73, "learning_rate": 1.4196009975062345e-05, "loss": 0.0483, "step": 141700 }, { "epoch": 11.74, "learning_rate": 1.4191853699085621e-05, "loss": 0.0501, "step": 141800 }, { "epoch": 11.75, "learning_rate": 1.4187697423108895e-05, "loss": 0.0458, "step": 141900 }, { "epoch": 11.75, "learning_rate": 1.4183541147132172e-05, "loss": 0.0562, "step": 142000 }, { "epoch": 11.75, "eval_loss": 0.14323097467422485, "eval_runtime": 6494.3171, "eval_samples_per_second": 11.64, "eval_steps_per_second": 0.97, "eval_wer": 0.05825837191019706, "step": 142000 }, { "epoch": 11.76, "learning_rate": 1.4179384871155446e-05, "loss": 0.0501, "step": 142100 }, { "epoch": 11.77, "learning_rate": 1.4175228595178721e-05, "loss": 0.0465, "step": 142200 }, { "epoch": 11.78, "learning_rate": 1.4171072319201997e-05, "loss": 0.0484, "step": 142300 }, { "epoch": 11.79, "learning_rate": 1.4166916043225272e-05, "loss": 0.052, "step": 142400 }, { "epoch": 11.8, "learning_rate": 1.4162759767248546e-05, "loss": 0.0488, "step": 142500 }, { "epoch": 11.8, "eval_loss": 0.1447732299566269, "eval_runtime": 6383.6673, "eval_samples_per_second": 11.842, "eval_steps_per_second": 0.987, "eval_wer": 0.05821375659422062, "step": 142500 }, { "epoch": 11.8, "learning_rate": 1.415860349127182e-05, "loss": 0.0497, "step": 142600 }, { "epoch": 11.81, "learning_rate": 1.4154447215295097e-05, "loss": 0.0504, "step": 142700 }, { "epoch": 11.82, "learning_rate": 1.4150290939318371e-05, "loss": 0.0532, "step": 142800 }, { "epoch": 11.83, "learning_rate": 1.4146134663341647e-05, "loss": 0.0512, "step": 142900 }, { "epoch": 11.84, "learning_rate": 1.4141978387364924e-05, "loss": 0.0547, "step": 143000 }, { "epoch": 11.84, "eval_loss": 0.13579151034355164, "eval_runtime": 6383.1814, "eval_samples_per_second": 11.843, "eval_steps_per_second": 0.987, "eval_wer": 0.057789141862858676, "step": 143000 }, { "epoch": 11.85, "learning_rate": 1.4137822111388198e-05, "loss": 0.047, "step": 143100 }, { "epoch": 11.85, "learning_rate": 1.4133665835411471e-05, "loss": 0.0441, "step": 143200 }, { "epoch": 11.86, "learning_rate": 1.4129509559434749e-05, "loss": 0.054, "step": 143300 }, { "epoch": 11.87, "learning_rate": 1.4125353283458022e-05, "loss": 0.0567, "step": 143400 }, { "epoch": 11.88, "learning_rate": 1.4121197007481298e-05, "loss": 0.0464, "step": 143500 }, { "epoch": 11.88, "eval_loss": 0.13994088768959045, "eval_runtime": 6332.8048, "eval_samples_per_second": 11.937, "eval_steps_per_second": 0.995, "eval_wer": 0.05798914155516684, "step": 143500 }, { "epoch": 11.89, "learning_rate": 1.4117040731504572e-05, "loss": 0.0484, "step": 143600 }, { "epoch": 11.9, "learning_rate": 1.4112884455527849e-05, "loss": 0.0556, "step": 143700 }, { "epoch": 11.9, "learning_rate": 1.4108728179551123e-05, "loss": 0.0524, "step": 143800 }, { "epoch": 11.91, "learning_rate": 1.4104571903574398e-05, "loss": 0.0456, "step": 143900 }, { "epoch": 11.92, "learning_rate": 1.4100415627597674e-05, "loss": 0.0507, "step": 144000 }, { "epoch": 11.92, "eval_loss": 0.1418793797492981, "eval_runtime": 6284.6136, "eval_samples_per_second": 12.029, "eval_steps_per_second": 1.002, "eval_wer": 0.059326062575288344, "step": 144000 }, { "epoch": 11.93, "learning_rate": 1.409625935162095e-05, "loss": 0.0537, "step": 144100 }, { "epoch": 11.94, "learning_rate": 1.4092103075644223e-05, "loss": 0.0555, "step": 144200 }, { "epoch": 11.95, "learning_rate": 1.40879467996675e-05, "loss": 0.0511, "step": 144300 }, { "epoch": 11.95, "learning_rate": 1.4083790523690774e-05, "loss": 0.0506, "step": 144400 }, { "epoch": 11.96, "learning_rate": 1.407963424771405e-05, "loss": 0.0509, "step": 144500 }, { "epoch": 11.96, "eval_loss": 0.13393011689186096, "eval_runtime": 6414.118, "eval_samples_per_second": 11.786, "eval_steps_per_second": 0.982, "eval_wer": 0.05620452891610936, "step": 144500 }, { "epoch": 11.97, "learning_rate": 1.4075477971737324e-05, "loss": 0.0488, "step": 144600 }, { "epoch": 11.98, "learning_rate": 1.40713216957606e-05, "loss": 0.0511, "step": 144700 }, { "epoch": 11.99, "learning_rate": 1.4067165419783875e-05, "loss": 0.0426, "step": 144800 }, { "epoch": 12.0, "learning_rate": 1.4063009143807148e-05, "loss": 0.0472, "step": 144900 }, { "epoch": 12.0, "learning_rate": 1.4058894430590194e-05, "loss": 0.0447, "step": 145000 }, { "epoch": 12.0, "eval_loss": 0.13015855848789215, "eval_runtime": 6420.9995, "eval_samples_per_second": 11.773, "eval_steps_per_second": 0.981, "eval_wer": 0.05536145329007186, "step": 145000 }, { "epoch": 12.01, "learning_rate": 1.4054738154613467e-05, "loss": 0.0496, "step": 145100 }, { "epoch": 12.02, "learning_rate": 1.4050581878636741e-05, "loss": 0.0418, "step": 145200 }, { "epoch": 12.03, "learning_rate": 1.4046425602660018e-05, "loss": 0.0502, "step": 145300 }, { "epoch": 12.04, "learning_rate": 1.4042269326683294e-05, "loss": 0.0509, "step": 145400 }, { "epoch": 12.04, "learning_rate": 1.4038113050706568e-05, "loss": 0.044, "step": 145500 }, { "epoch": 12.04, "eval_loss": 0.13767455518245697, "eval_runtime": 4467.4878, "eval_samples_per_second": 16.921, "eval_steps_per_second": 1.41, "eval_wer": 0.056044529162262824, "step": 145500 }, { "epoch": 12.05, "learning_rate": 1.4033956774729842e-05, "loss": 0.041, "step": 145600 }, { "epoch": 12.06, "learning_rate": 1.4029800498753119e-05, "loss": 0.0597, "step": 145700 }, { "epoch": 12.07, "learning_rate": 1.4025644222776393e-05, "loss": 0.0441, "step": 145800 }, { "epoch": 12.08, "learning_rate": 1.4021487946799668e-05, "loss": 0.0469, "step": 145900 }, { "epoch": 12.09, "learning_rate": 1.4017331670822945e-05, "loss": 0.0435, "step": 146000 }, { "epoch": 12.09, "eval_loss": 0.13885104656219482, "eval_runtime": 4265.0544, "eval_samples_per_second": 17.724, "eval_steps_per_second": 1.477, "eval_wer": 0.05841529474570039, "step": 146000 }, { "epoch": 12.09, "learning_rate": 1.401317539484622e-05, "loss": 0.0498, "step": 146100 }, { "epoch": 12.1, "learning_rate": 1.4009019118869493e-05, "loss": 0.0442, "step": 146200 }, { "epoch": 12.11, "learning_rate": 1.400486284289277e-05, "loss": 0.0491, "step": 146300 }, { "epoch": 12.12, "learning_rate": 1.4000706566916044e-05, "loss": 0.0469, "step": 146400 }, { "epoch": 12.13, "learning_rate": 1.399655029093932e-05, "loss": 0.0451, "step": 146500 }, { "epoch": 12.13, "eval_loss": 0.14748530089855194, "eval_runtime": 4289.5577, "eval_samples_per_second": 17.623, "eval_steps_per_second": 1.469, "eval_wer": 0.059201447382388646, "step": 146500 }, { "epoch": 12.14, "learning_rate": 1.3992394014962593e-05, "loss": 0.049, "step": 146600 }, { "epoch": 12.14, "learning_rate": 1.398823773898587e-05, "loss": 0.0478, "step": 146700 }, { "epoch": 12.15, "learning_rate": 1.3984081463009144e-05, "loss": 0.0454, "step": 146800 }, { "epoch": 12.16, "learning_rate": 1.397992518703242e-05, "loss": 0.0493, "step": 146900 }, { "epoch": 12.17, "learning_rate": 1.3975768911055695e-05, "loss": 0.0494, "step": 147000 }, { "epoch": 12.17, "eval_loss": 0.14632566273212433, "eval_runtime": 4805.1893, "eval_samples_per_second": 15.732, "eval_steps_per_second": 1.311, "eval_wer": 0.057126065959898525, "step": 147000 }, { "epoch": 12.18, "learning_rate": 1.3971612635078971e-05, "loss": 0.0532, "step": 147100 }, { "epoch": 12.19, "learning_rate": 1.3967456359102245e-05, "loss": 0.0482, "step": 147200 }, { "epoch": 12.19, "learning_rate": 1.3963300083125522e-05, "loss": 0.0548, "step": 147300 }, { "epoch": 12.2, "learning_rate": 1.3959143807148796e-05, "loss": 0.0494, "step": 147400 }, { "epoch": 12.21, "learning_rate": 1.3954987531172071e-05, "loss": 0.0723, "step": 147500 }, { "epoch": 12.21, "eval_loss": 0.13052673637866974, "eval_runtime": 6437.6289, "eval_samples_per_second": 11.743, "eval_steps_per_second": 0.979, "eval_wer": 0.05506299221078122, "step": 147500 }, { "epoch": 12.22, "learning_rate": 1.3950831255195345e-05, "loss": 0.0503, "step": 147600 }, { "epoch": 12.23, "learning_rate": 1.3946674979218622e-05, "loss": 0.0482, "step": 147700 }, { "epoch": 12.24, "learning_rate": 1.3942518703241896e-05, "loss": 0.0457, "step": 147800 }, { "epoch": 12.24, "learning_rate": 1.393836242726517e-05, "loss": 0.0519, "step": 147900 }, { "epoch": 12.25, "learning_rate": 1.3934206151288447e-05, "loss": 0.0414, "step": 148000 }, { "epoch": 12.25, "eval_loss": 0.13862648606300354, "eval_runtime": 6337.7187, "eval_samples_per_second": 11.928, "eval_steps_per_second": 0.994, "eval_wer": 0.05499683846640236, "step": 148000 }, { "epoch": 12.26, "learning_rate": 1.3930049875311723e-05, "loss": 0.0466, "step": 148100 }, { "epoch": 12.27, "learning_rate": 1.3925893599334997e-05, "loss": 0.049, "step": 148200 }, { "epoch": 12.28, "learning_rate": 1.3921737323358274e-05, "loss": 0.0463, "step": 148300 }, { "epoch": 12.28, "learning_rate": 1.3917581047381548e-05, "loss": 0.0499, "step": 148400 }, { "epoch": 12.29, "learning_rate": 1.3913424771404821e-05, "loss": 0.0479, "step": 148500 }, { "epoch": 12.29, "eval_loss": 0.15571652352809906, "eval_runtime": 6243.0304, "eval_samples_per_second": 12.109, "eval_steps_per_second": 1.009, "eval_wer": 0.05654145147469004, "step": 148500 }, { "epoch": 12.3, "learning_rate": 1.3909268495428097e-05, "loss": 0.0444, "step": 148600 }, { "epoch": 12.31, "learning_rate": 1.3905112219451372e-05, "loss": 0.0538, "step": 148700 }, { "epoch": 12.32, "learning_rate": 1.3900955943474648e-05, "loss": 0.05, "step": 148800 }, { "epoch": 12.33, "learning_rate": 1.3896799667497922e-05, "loss": 0.052, "step": 148900 }, { "epoch": 12.33, "learning_rate": 1.3892643391521199e-05, "loss": 0.0489, "step": 149000 }, { "epoch": 12.33, "eval_loss": 0.12930600345134735, "eval_runtime": 5959.0002, "eval_samples_per_second": 12.686, "eval_steps_per_second": 1.057, "eval_wer": 0.05471222351965612, "step": 149000 }, { "epoch": 12.34, "learning_rate": 1.3888487115544473e-05, "loss": 0.0519, "step": 149100 }, { "epoch": 12.35, "learning_rate": 1.3884330839567748e-05, "loss": 0.0416, "step": 149200 }, { "epoch": 12.36, "learning_rate": 1.3880174563591024e-05, "loss": 0.0455, "step": 149300 }, { "epoch": 12.37, "learning_rate": 1.38760182876143e-05, "loss": 0.0481, "step": 149400 }, { "epoch": 12.38, "learning_rate": 1.3871862011637573e-05, "loss": 0.0461, "step": 149500 }, { "epoch": 12.38, "eval_loss": 0.14204691350460052, "eval_runtime": 5492.1814, "eval_samples_per_second": 13.764, "eval_steps_per_second": 1.147, "eval_wer": 0.05698760463445441, "step": 149500 }, { "epoch": 12.38, "learning_rate": 1.3867705735660849e-05, "loss": 0.0547, "step": 149600 }, { "epoch": 12.39, "learning_rate": 1.3863549459684124e-05, "loss": 0.0473, "step": 149700 }, { "epoch": 12.4, "learning_rate": 1.38593931837074e-05, "loss": 0.0497, "step": 149800 }, { "epoch": 12.41, "learning_rate": 1.3855236907730674e-05, "loss": 0.0448, "step": 149900 }, { "epoch": 12.42, "learning_rate": 1.385108063175395e-05, "loss": 0.0462, "step": 150000 }, { "epoch": 12.42, "eval_loss": 0.13579699397087097, "eval_runtime": 5156.7726, "eval_samples_per_second": 14.659, "eval_steps_per_second": 1.222, "eval_wer": 0.05657529757646527, "step": 150000 }, { "epoch": 12.43, "learning_rate": 1.3846924355777225e-05, "loss": 0.0542, "step": 150100 }, { "epoch": 12.43, "learning_rate": 1.38427680798005e-05, "loss": 0.0491, "step": 150200 }, { "epoch": 12.44, "learning_rate": 1.3838611803823776e-05, "loss": 0.0494, "step": 150300 }, { "epoch": 12.45, "learning_rate": 1.3834455527847051e-05, "loss": 0.051, "step": 150400 }, { "epoch": 12.46, "learning_rate": 1.3830299251870325e-05, "loss": 0.0431, "step": 150500 }, { "epoch": 12.46, "eval_loss": 0.15288777649402618, "eval_runtime": 5203.0504, "eval_samples_per_second": 14.529, "eval_steps_per_second": 1.211, "eval_wer": 0.05870144815161823, "step": 150500 }, { "epoch": 12.47, "learning_rate": 1.3826142975893599e-05, "loss": 0.0425, "step": 150600 }, { "epoch": 12.48, "learning_rate": 1.3821986699916876e-05, "loss": 0.0497, "step": 150700 }, { "epoch": 12.48, "learning_rate": 1.381783042394015e-05, "loss": 0.0429, "step": 150800 }, { "epoch": 12.49, "learning_rate": 1.3813674147963425e-05, "loss": 0.042, "step": 150900 }, { "epoch": 12.5, "learning_rate": 1.3809517871986703e-05, "loss": 0.0439, "step": 151000 }, { "epoch": 12.5, "eval_loss": 0.1448381543159485, "eval_runtime": 5129.9952, "eval_samples_per_second": 14.736, "eval_steps_per_second": 1.228, "eval_wer": 0.05713529671492813, "step": 151000 }, { "epoch": 12.51, "learning_rate": 1.3805403158769743e-05, "loss": 0.0493, "step": 151100 }, { "epoch": 12.52, "learning_rate": 1.3801246882793018e-05, "loss": 0.0449, "step": 151200 }, { "epoch": 12.52, "learning_rate": 1.3797090606816295e-05, "loss": 0.05, "step": 151300 }, { "epoch": 12.53, "learning_rate": 1.379293433083957e-05, "loss": 0.0463, "step": 151400 }, { "epoch": 12.54, "learning_rate": 1.3788778054862843e-05, "loss": 0.0384, "step": 151500 }, { "epoch": 12.54, "eval_loss": 0.1331833451986313, "eval_runtime": 5433.2254, "eval_samples_per_second": 13.913, "eval_steps_per_second": 1.16, "eval_wer": 0.05534760715752745, "step": 151500 }, { "epoch": 12.55, "learning_rate": 1.3784621778886119e-05, "loss": 0.0469, "step": 151600 }, { "epoch": 12.56, "learning_rate": 1.3780465502909394e-05, "loss": 0.047, "step": 151700 }, { "epoch": 12.57, "learning_rate": 1.377630922693267e-05, "loss": 0.0458, "step": 151800 }, { "epoch": 12.57, "learning_rate": 1.3772152950955943e-05, "loss": 0.0492, "step": 151900 }, { "epoch": 12.58, "learning_rate": 1.376799667497922e-05, "loss": 0.0498, "step": 152000 }, { "epoch": 12.58, "eval_loss": 0.1323612481355667, "eval_runtime": 5102.5693, "eval_samples_per_second": 14.815, "eval_steps_per_second": 1.235, "eval_wer": 0.05547529926877036, "step": 152000 }, { "epoch": 12.59, "learning_rate": 1.3763840399002494e-05, "loss": 0.0574, "step": 152100 }, { "epoch": 12.6, "learning_rate": 1.375968412302577e-05, "loss": 0.0457, "step": 152200 }, { "epoch": 12.61, "learning_rate": 1.3755527847049045e-05, "loss": 0.0444, "step": 152300 }, { "epoch": 12.62, "learning_rate": 1.3751371571072321e-05, "loss": 0.0443, "step": 152400 }, { "epoch": 12.62, "learning_rate": 1.3747215295095595e-05, "loss": 0.0458, "step": 152500 }, { "epoch": 12.62, "eval_loss": 0.13542887568473816, "eval_runtime": 4924.5532, "eval_samples_per_second": 15.351, "eval_steps_per_second": 1.279, "eval_wer": 0.054872223273502656, "step": 152500 }, { "epoch": 12.63, "learning_rate": 1.374305901911887e-05, "loss": 0.0505, "step": 152600 }, { "epoch": 12.64, "learning_rate": 1.3738902743142146e-05, "loss": 0.0542, "step": 152700 }, { "epoch": 12.65, "learning_rate": 1.3734746467165421e-05, "loss": 0.0486, "step": 152800 }, { "epoch": 12.66, "learning_rate": 1.3730590191188695e-05, "loss": 0.0459, "step": 152900 }, { "epoch": 12.67, "learning_rate": 1.3726433915211972e-05, "loss": 0.0475, "step": 153000 }, { "epoch": 12.67, "eval_loss": 0.13288748264312744, "eval_runtime": 4512.9096, "eval_samples_per_second": 16.751, "eval_steps_per_second": 1.396, "eval_wer": 0.055515299207231986, "step": 153000 }, { "epoch": 12.67, "learning_rate": 1.3722277639235246e-05, "loss": 0.0514, "step": 153100 }, { "epoch": 12.68, "learning_rate": 1.3718162926018288e-05, "loss": 0.0467, "step": 153200 }, { "epoch": 12.69, "learning_rate": 1.3714006650041565e-05, "loss": 0.0493, "step": 153300 }, { "epoch": 12.7, "learning_rate": 1.3709850374064839e-05, "loss": 0.045, "step": 153400 }, { "epoch": 12.71, "learning_rate": 1.3705694098088113e-05, "loss": 0.0487, "step": 153500 }, { "epoch": 12.71, "eval_loss": 0.13235388696193695, "eval_runtime": 4508.4334, "eval_samples_per_second": 16.767, "eval_steps_per_second": 1.397, "eval_wer": 0.05652145150545922, "step": 153500 }, { "epoch": 12.72, "learning_rate": 1.370153782211139e-05, "loss": 0.0487, "step": 153600 }, { "epoch": 12.72, "learning_rate": 1.3697381546134666e-05, "loss": 0.0445, "step": 153700 }, { "epoch": 12.73, "learning_rate": 1.369322527015794e-05, "loss": 0.044, "step": 153800 }, { "epoch": 12.74, "learning_rate": 1.3689068994181213e-05, "loss": 0.046, "step": 153900 }, { "epoch": 12.75, "learning_rate": 1.368491271820449e-05, "loss": 0.0425, "step": 154000 }, { "epoch": 12.75, "eval_loss": 0.13745063543319702, "eval_runtime": 4551.4839, "eval_samples_per_second": 16.609, "eval_steps_per_second": 1.384, "eval_wer": 0.05529991492320781, "step": 154000 }, { "epoch": 12.76, "learning_rate": 1.3680756442227764e-05, "loss": 0.0442, "step": 154100 }, { "epoch": 12.76, "learning_rate": 1.3676641729010806e-05, "loss": 0.053, "step": 154200 }, { "epoch": 12.77, "learning_rate": 1.3672485453034083e-05, "loss": 0.0488, "step": 154300 }, { "epoch": 12.78, "learning_rate": 1.3668329177057357e-05, "loss": 0.0479, "step": 154400 }, { "epoch": 12.79, "learning_rate": 1.3664172901080633e-05, "loss": 0.043, "step": 154500 }, { "epoch": 12.79, "eval_loss": 0.13538089394569397, "eval_runtime": 4531.1582, "eval_samples_per_second": 16.683, "eval_steps_per_second": 1.39, "eval_wer": 0.056021452274688806, "step": 154500 }, { "epoch": 12.8, "learning_rate": 1.3660016625103908e-05, "loss": 0.0485, "step": 154600 }, { "epoch": 12.81, "learning_rate": 1.3655860349127184e-05, "loss": 0.0477, "step": 154700 }, { "epoch": 12.81, "learning_rate": 1.3651704073150458e-05, "loss": 0.0515, "step": 154800 }, { "epoch": 12.82, "learning_rate": 1.3647547797173733e-05, "loss": 0.0512, "step": 154900 }, { "epoch": 12.83, "learning_rate": 1.3643391521197009e-05, "loss": 0.0515, "step": 155000 }, { "epoch": 12.83, "eval_loss": 0.1378927379846573, "eval_runtime": 4518.9554, "eval_samples_per_second": 16.728, "eval_steps_per_second": 1.394, "eval_wer": 0.05600145230545799, "step": 155000 }, { "epoch": 12.84, "learning_rate": 1.3639235245220284e-05, "loss": 0.0627, "step": 155100 }, { "epoch": 12.85, "learning_rate": 1.3635078969243558e-05, "loss": 0.0524, "step": 155200 }, { "epoch": 12.86, "learning_rate": 1.3630922693266835e-05, "loss": 0.0458, "step": 155300 }, { "epoch": 12.86, "learning_rate": 1.3626766417290109e-05, "loss": 0.0467, "step": 155400 }, { "epoch": 12.87, "learning_rate": 1.3622610141313384e-05, "loss": 0.0494, "step": 155500 }, { "epoch": 12.87, "eval_loss": 0.14550597965717316, "eval_runtime": 4350.1461, "eval_samples_per_second": 17.378, "eval_steps_per_second": 1.448, "eval_wer": 0.057112219827354115, "step": 155500 }, { "epoch": 12.88, "learning_rate": 1.361845386533666e-05, "loss": 0.0467, "step": 155600 }, { "epoch": 12.89, "learning_rate": 1.3614297589359936e-05, "loss": 0.051, "step": 155700 }, { "epoch": 12.9, "learning_rate": 1.361014131338321e-05, "loss": 0.0464, "step": 155800 }, { "epoch": 12.91, "learning_rate": 1.3605985037406483e-05, "loss": 0.0409, "step": 155900 }, { "epoch": 12.91, "learning_rate": 1.360182876142976e-05, "loss": 0.0525, "step": 156000 }, { "epoch": 12.91, "eval_loss": 0.13452914357185364, "eval_runtime": 4376.7319, "eval_samples_per_second": 17.272, "eval_steps_per_second": 1.439, "eval_wer": 0.056206067375280964, "step": 156000 }, { "epoch": 12.92, "learning_rate": 1.3597672485453036e-05, "loss": 0.0563, "step": 156100 }, { "epoch": 12.93, "learning_rate": 1.359351620947631e-05, "loss": 0.0547, "step": 156200 }, { "epoch": 12.94, "learning_rate": 1.3589359933499587e-05, "loss": 0.0509, "step": 156300 }, { "epoch": 12.95, "learning_rate": 1.358520365752286e-05, "loss": 0.0504, "step": 156400 }, { "epoch": 12.96, "learning_rate": 1.3581047381546135e-05, "loss": 0.048, "step": 156500 }, { "epoch": 12.96, "eval_loss": 0.13944144546985626, "eval_runtime": 4358.368, "eval_samples_per_second": 17.345, "eval_steps_per_second": 1.445, "eval_wer": 0.054993761548059156, "step": 156500 }, { "epoch": 12.96, "learning_rate": 1.3576891105569412e-05, "loss": 0.0517, "step": 156600 }, { "epoch": 12.97, "learning_rate": 1.3572734829592686e-05, "loss": 0.0459, "step": 156700 }, { "epoch": 12.98, "learning_rate": 1.3568578553615961e-05, "loss": 0.0452, "step": 156800 }, { "epoch": 12.99, "learning_rate": 1.3564422277639235e-05, "loss": 0.0481, "step": 156900 }, { "epoch": 13.0, "learning_rate": 1.3560266001662512e-05, "loss": 0.0462, "step": 157000 }, { "epoch": 13.0, "eval_loss": 0.13642911612987518, "eval_runtime": 4388.4463, "eval_samples_per_second": 17.226, "eval_steps_per_second": 1.436, "eval_wer": 0.05641837474096194, "step": 157000 }, { "epoch": 13.0, "learning_rate": 1.3556151288445554e-05, "loss": 0.0421, "step": 157100 }, { "epoch": 13.01, "learning_rate": 1.3551995012468828e-05, "loss": 0.044, "step": 157200 }, { "epoch": 13.02, "learning_rate": 1.3547838736492105e-05, "loss": 0.0438, "step": 157300 }, { "epoch": 13.03, "learning_rate": 1.3543682460515379e-05, "loss": 0.0483, "step": 157400 }, { "epoch": 13.04, "learning_rate": 1.3539526184538654e-05, "loss": 0.0495, "step": 157500 }, { "epoch": 13.04, "eval_loss": 0.15104050934314728, "eval_runtime": 4405.3132, "eval_samples_per_second": 17.16, "eval_steps_per_second": 1.43, "eval_wer": 0.057198373540963784, "step": 157500 }, { "epoch": 13.05, "learning_rate": 1.353536990856193e-05, "loss": 0.0413, "step": 157600 }, { "epoch": 13.05, "learning_rate": 1.3531213632585205e-05, "loss": 0.0447, "step": 157700 }, { "epoch": 13.06, "learning_rate": 1.352705735660848e-05, "loss": 0.0455, "step": 157800 }, { "epoch": 13.07, "learning_rate": 1.3522901080631755e-05, "loss": 0.0456, "step": 157900 }, { "epoch": 13.08, "learning_rate": 1.351874480465503e-05, "loss": 0.0433, "step": 158000 }, { "epoch": 13.08, "eval_loss": 0.13567404448986053, "eval_runtime": 4409.8991, "eval_samples_per_second": 17.142, "eval_steps_per_second": 1.429, "eval_wer": 0.05470760814214132, "step": 158000 }, { "epoch": 13.09, "learning_rate": 1.3514588528678306e-05, "loss": 0.0454, "step": 158100 }, { "epoch": 13.1, "learning_rate": 1.351043225270158e-05, "loss": 0.0462, "step": 158200 }, { "epoch": 13.1, "learning_rate": 1.3506275976724857e-05, "loss": 0.0461, "step": 158300 }, { "epoch": 13.11, "learning_rate": 1.350211970074813e-05, "loss": 0.0456, "step": 158400 }, { "epoch": 13.12, "learning_rate": 1.3497963424771406e-05, "loss": 0.0419, "step": 158500 }, { "epoch": 13.12, "eval_loss": 0.14733904600143433, "eval_runtime": 4378.7602, "eval_samples_per_second": 17.264, "eval_steps_per_second": 1.439, "eval_wer": 0.05543376087113712, "step": 158500 }, { "epoch": 13.13, "learning_rate": 1.3493807148794682e-05, "loss": 0.0504, "step": 158600 }, { "epoch": 13.14, "learning_rate": 1.3489650872817957e-05, "loss": 0.0498, "step": 158700 }, { "epoch": 13.15, "learning_rate": 1.3485494596841231e-05, "loss": 0.0538, "step": 158800 }, { "epoch": 13.15, "learning_rate": 1.3481338320864505e-05, "loss": 0.043, "step": 158900 }, { "epoch": 13.16, "learning_rate": 1.3477182044887782e-05, "loss": 0.0453, "step": 159000 }, { "epoch": 13.16, "eval_loss": 0.1443016678094864, "eval_runtime": 4376.0209, "eval_samples_per_second": 17.275, "eval_steps_per_second": 1.44, "eval_wer": 0.05646914389362478, "step": 159000 }, { "epoch": 13.17, "learning_rate": 1.3473025768911056e-05, "loss": 0.0446, "step": 159100 }, { "epoch": 13.18, "learning_rate": 1.3468869492934331e-05, "loss": 0.051, "step": 159200 }, { "epoch": 13.19, "learning_rate": 1.3464713216957609e-05, "loss": 0.05, "step": 159300 }, { "epoch": 13.2, "learning_rate": 1.3460556940980882e-05, "loss": 0.0397, "step": 159400 }, { "epoch": 13.2, "learning_rate": 1.3456400665004156e-05, "loss": 0.043, "step": 159500 }, { "epoch": 13.2, "eval_loss": 0.16219967603683472, "eval_runtime": 4404.3353, "eval_samples_per_second": 17.164, "eval_steps_per_second": 1.43, "eval_wer": 0.058241448859309444, "step": 159500 }, { "epoch": 13.21, "learning_rate": 1.3452327514546968e-05, "loss": 0.0485, "step": 159600 }, { "epoch": 13.22, "learning_rate": 1.3448171238570242e-05, "loss": 0.0509, "step": 159700 }, { "epoch": 13.23, "learning_rate": 1.3444014962593517e-05, "loss": 0.0491, "step": 159800 }, { "epoch": 13.24, "learning_rate": 1.3439858686616794e-05, "loss": 0.0494, "step": 159900 }, { "epoch": 13.25, "learning_rate": 1.3435702410640068e-05, "loss": 0.0404, "step": 160000 }, { "epoch": 13.25, "eval_loss": 0.1548277884721756, "eval_runtime": 4400.9331, "eval_samples_per_second": 17.177, "eval_steps_per_second": 1.432, "eval_wer": 0.05657222065812206, "step": 160000 }, { "epoch": 13.25, "learning_rate": 1.3431546134663342e-05, "loss": 0.0391, "step": 160100 }, { "epoch": 13.26, "learning_rate": 1.3427389858686617e-05, "loss": 0.0402, "step": 160200 }, { "epoch": 13.27, "learning_rate": 1.3423233582709893e-05, "loss": 0.0405, "step": 160300 }, { "epoch": 13.28, "learning_rate": 1.3419077306733168e-05, "loss": 0.0474, "step": 160400 }, { "epoch": 13.29, "learning_rate": 1.3414921030756442e-05, "loss": 0.0396, "step": 160500 }, { "epoch": 13.29, "eval_loss": 0.14695149660110474, "eval_runtime": 4467.2107, "eval_samples_per_second": 16.922, "eval_steps_per_second": 1.41, "eval_wer": 0.0563737594249855, "step": 160500 }, { "epoch": 13.29, "learning_rate": 1.341076475477972e-05, "loss": 0.0475, "step": 160600 }, { "epoch": 13.3, "learning_rate": 1.3406650041562761e-05, "loss": 0.0502, "step": 160700 }, { "epoch": 13.31, "learning_rate": 1.3402493765586035e-05, "loss": 0.0408, "step": 160800 }, { "epoch": 13.32, "learning_rate": 1.3398337489609312e-05, "loss": 0.0488, "step": 160900 }, { "epoch": 13.33, "learning_rate": 1.3394181213632586e-05, "loss": 0.041, "step": 161000 }, { "epoch": 13.33, "eval_loss": 0.14018584787845612, "eval_runtime": 4397.4367, "eval_samples_per_second": 17.191, "eval_steps_per_second": 1.433, "eval_wer": 0.055658375910190905, "step": 161000 }, { "epoch": 13.34, "learning_rate": 1.3390024937655862e-05, "loss": 0.0432, "step": 161100 }, { "epoch": 13.34, "learning_rate": 1.3385868661679136e-05, "loss": 0.0462, "step": 161200 }, { "epoch": 13.35, "learning_rate": 1.3381712385702413e-05, "loss": 0.0479, "step": 161300 }, { "epoch": 13.36, "learning_rate": 1.3377556109725687e-05, "loss": 0.0434, "step": 161400 }, { "epoch": 13.37, "learning_rate": 1.3373399833748962e-05, "loss": 0.0468, "step": 161500 }, { "epoch": 13.37, "eval_loss": 0.14445674419403076, "eval_runtime": 4465.5833, "eval_samples_per_second": 16.928, "eval_steps_per_second": 1.411, "eval_wer": 0.056849143309010294, "step": 161500 }, { "epoch": 13.38, "learning_rate": 1.3369243557772238e-05, "loss": 0.0459, "step": 161600 }, { "epoch": 13.39, "learning_rate": 1.3365087281795513e-05, "loss": 0.0451, "step": 161700 }, { "epoch": 13.39, "learning_rate": 1.3360931005818787e-05, "loss": 0.0474, "step": 161800 }, { "epoch": 13.4, "learning_rate": 1.3356774729842064e-05, "loss": 0.0434, "step": 161900 }, { "epoch": 13.41, "learning_rate": 1.3352618453865338e-05, "loss": 0.0481, "step": 162000 }, { "epoch": 13.41, "eval_loss": 0.1446288824081421, "eval_runtime": 4308.2738, "eval_samples_per_second": 17.546, "eval_steps_per_second": 1.462, "eval_wer": 0.05779683415871668, "step": 162000 }, { "epoch": 13.42, "learning_rate": 1.3348462177888612e-05, "loss": 0.0455, "step": 162100 }, { "epoch": 13.43, "learning_rate": 1.3344305901911887e-05, "loss": 0.0415, "step": 162200 }, { "epoch": 13.44, "learning_rate": 1.3340149625935164e-05, "loss": 0.0441, "step": 162300 }, { "epoch": 13.44, "learning_rate": 1.3335993349958438e-05, "loss": 0.0433, "step": 162400 }, { "epoch": 13.45, "learning_rate": 1.3331837073981712e-05, "loss": 0.0472, "step": 162500 }, { "epoch": 13.45, "eval_loss": 0.14025430381298065, "eval_runtime": 4285.0898, "eval_samples_per_second": 17.641, "eval_steps_per_second": 1.47, "eval_wer": 0.055327607188296636, "step": 162500 }, { "epoch": 13.46, "learning_rate": 1.332768079800499e-05, "loss": 0.0488, "step": 162600 }, { "epoch": 13.47, "learning_rate": 1.3323524522028263e-05, "loss": 0.0489, "step": 162700 }, { "epoch": 13.48, "learning_rate": 1.3319368246051539e-05, "loss": 0.0443, "step": 162800 }, { "epoch": 13.49, "learning_rate": 1.3315211970074814e-05, "loss": 0.0471, "step": 162900 }, { "epoch": 13.49, "learning_rate": 1.331105569409809e-05, "loss": 0.0437, "step": 163000 }, { "epoch": 13.49, "eval_loss": 0.14942488074302673, "eval_runtime": 4495.5325, "eval_samples_per_second": 16.816, "eval_steps_per_second": 1.401, "eval_wer": 0.05656914373977886, "step": 163000 }, { "epoch": 13.5, "learning_rate": 1.3306899418121364e-05, "loss": 0.0446, "step": 163100 }, { "epoch": 13.51, "learning_rate": 1.3302743142144639e-05, "loss": 0.0438, "step": 163200 }, { "epoch": 13.52, "learning_rate": 1.3298586866167915e-05, "loss": 0.0438, "step": 163300 }, { "epoch": 13.53, "learning_rate": 1.329443059019119e-05, "loss": 0.0433, "step": 163400 }, { "epoch": 13.53, "learning_rate": 1.3290274314214464e-05, "loss": 0.0379, "step": 163500 }, { "epoch": 13.53, "eval_loss": 0.15515358746051788, "eval_runtime": 4754.6221, "eval_samples_per_second": 15.899, "eval_steps_per_second": 1.325, "eval_wer": 0.05634606715989668, "step": 163500 }, { "epoch": 13.54, "learning_rate": 1.3286118038237741e-05, "loss": 0.0471, "step": 163600 }, { "epoch": 13.55, "learning_rate": 1.3281961762261015e-05, "loss": 0.0426, "step": 163700 }, { "epoch": 13.56, "learning_rate": 1.327780548628429e-05, "loss": 0.0432, "step": 163800 }, { "epoch": 13.57, "learning_rate": 1.3273649210307566e-05, "loss": 0.0495, "step": 163900 }, { "epoch": 13.58, "learning_rate": 1.3269492934330841e-05, "loss": 0.0401, "step": 164000 }, { "epoch": 13.58, "eval_loss": 0.16145916283130646, "eval_runtime": 4757.8925, "eval_samples_per_second": 15.888, "eval_steps_per_second": 1.324, "eval_wer": 0.060993752317304126, "step": 164000 }, { "epoch": 13.58, "learning_rate": 1.3265336658354115e-05, "loss": 0.0427, "step": 164100 }, { "epoch": 13.59, "learning_rate": 1.3261180382377389e-05, "loss": 0.0392, "step": 164200 }, { "epoch": 13.6, "learning_rate": 1.3257024106400666e-05, "loss": 0.0452, "step": 164300 }, { "epoch": 13.61, "learning_rate": 1.3252867830423942e-05, "loss": 0.0453, "step": 164400 }, { "epoch": 13.62, "learning_rate": 1.3248711554447216e-05, "loss": 0.0504, "step": 164500 }, { "epoch": 13.62, "eval_loss": 0.15362077951431274, "eval_runtime": 4976.7181, "eval_samples_per_second": 15.19, "eval_steps_per_second": 1.266, "eval_wer": 0.05772760349599462, "step": 164500 }, { "epoch": 13.63, "learning_rate": 1.3244555278470493e-05, "loss": 0.0426, "step": 164600 }, { "epoch": 13.63, "learning_rate": 1.3240399002493767e-05, "loss": 0.0428, "step": 164700 }, { "epoch": 13.64, "learning_rate": 1.323624272651704e-05, "loss": 0.0424, "step": 164800 }, { "epoch": 13.65, "learning_rate": 1.3232086450540318e-05, "loss": 0.0465, "step": 164900 }, { "epoch": 13.66, "learning_rate": 1.3227930174563592e-05, "loss": 0.0425, "step": 165000 }, { "epoch": 13.66, "eval_loss": 0.1512785255908966, "eval_runtime": 4532.89, "eval_samples_per_second": 16.677, "eval_steps_per_second": 1.39, "eval_wer": 0.05828760263445749, "step": 165000 }, { "epoch": 13.67, "learning_rate": 1.3223773898586867e-05, "loss": 0.0496, "step": 165100 }, { "epoch": 13.68, "learning_rate": 1.3219617622610141e-05, "loss": 0.0539, "step": 165200 }, { "epoch": 13.68, "learning_rate": 1.3215461346633418e-05, "loss": 0.043, "step": 165300 }, { "epoch": 13.69, "learning_rate": 1.3211305070656692e-05, "loss": 0.0489, "step": 165400 }, { "epoch": 13.7, "learning_rate": 1.3207148794679967e-05, "loss": 0.0467, "step": 165500 }, { "epoch": 13.7, "eval_loss": 0.14248988032341003, "eval_runtime": 4270.3581, "eval_samples_per_second": 17.702, "eval_steps_per_second": 1.475, "eval_wer": 0.05749991153859763, "step": 165500 }, { "epoch": 13.71, "learning_rate": 1.3202992518703243e-05, "loss": 0.0503, "step": 165600 }, { "epoch": 13.72, "learning_rate": 1.3198836242726518e-05, "loss": 0.0453, "step": 165700 }, { "epoch": 13.73, "learning_rate": 1.3194679966749792e-05, "loss": 0.0482, "step": 165800 }, { "epoch": 13.73, "learning_rate": 1.319052369077307e-05, "loss": 0.0457, "step": 165900 }, { "epoch": 13.74, "learning_rate": 1.3186367414796343e-05, "loss": 0.0459, "step": 166000 }, { "epoch": 13.74, "eval_loss": 0.13585519790649414, "eval_runtime": 4275.5938, "eval_samples_per_second": 17.681, "eval_steps_per_second": 1.473, "eval_wer": 0.05511068444510085, "step": 166000 }, { "epoch": 13.75, "learning_rate": 1.3182211138819619e-05, "loss": 0.0465, "step": 166100 }, { "epoch": 13.76, "learning_rate": 1.3178054862842893e-05, "loss": 0.0432, "step": 166200 }, { "epoch": 13.77, "learning_rate": 1.317389858686617e-05, "loss": 0.0539, "step": 166300 }, { "epoch": 13.77, "learning_rate": 1.3169742310889444e-05, "loss": 0.0426, "step": 166400 }, { "epoch": 13.78, "learning_rate": 1.316558603491272e-05, "loss": 0.0416, "step": 166500 }, { "epoch": 13.78, "eval_loss": 0.1490047723054886, "eval_runtime": 4272.3321, "eval_samples_per_second": 17.694, "eval_steps_per_second": 1.475, "eval_wer": 0.056586066790666474, "step": 166500 }, { "epoch": 13.79, "learning_rate": 1.3161429758935995e-05, "loss": 0.0425, "step": 166600 }, { "epoch": 13.8, "learning_rate": 1.315727348295927e-05, "loss": 0.0395, "step": 166700 }, { "epoch": 13.81, "learning_rate": 1.3153117206982544e-05, "loss": 0.0414, "step": 166800 }, { "epoch": 13.82, "learning_rate": 1.3148960931005821e-05, "loss": 0.0484, "step": 166900 }, { "epoch": 13.82, "learning_rate": 1.3144846217788863e-05, "loss": 0.0457, "step": 167000 }, { "epoch": 13.82, "eval_loss": 0.14718973636627197, "eval_runtime": 4725.0948, "eval_samples_per_second": 15.999, "eval_steps_per_second": 1.333, "eval_wer": 0.05596760620368276, "step": 167000 }, { "epoch": 13.83, "learning_rate": 1.3140689941812137e-05, "loss": 0.0531, "step": 167100 }, { "epoch": 13.84, "learning_rate": 1.313653366583541e-05, "loss": 0.0473, "step": 167200 }, { "epoch": 13.85, "learning_rate": 1.3132377389858688e-05, "loss": 0.0409, "step": 167300 }, { "epoch": 13.86, "learning_rate": 1.3128221113881962e-05, "loss": 0.044, "step": 167400 }, { "epoch": 13.87, "learning_rate": 1.3124064837905237e-05, "loss": 0.0484, "step": 167500 }, { "epoch": 13.87, "eval_loss": 0.1357632577419281, "eval_runtime": 4554.3686, "eval_samples_per_second": 16.598, "eval_steps_per_second": 1.383, "eval_wer": 0.0554106839835631, "step": 167500 }, { "epoch": 13.87, "learning_rate": 1.3119908561928515e-05, "loss": 0.0456, "step": 167600 }, { "epoch": 13.88, "learning_rate": 1.3115752285951788e-05, "loss": 0.0471, "step": 167700 }, { "epoch": 13.89, "learning_rate": 1.3111596009975062e-05, "loss": 0.0471, "step": 167800 }, { "epoch": 13.9, "learning_rate": 1.3107481296758107e-05, "loss": 0.0443, "step": 167900 }, { "epoch": 13.91, "learning_rate": 1.3103325020781381e-05, "loss": 0.0574, "step": 168000 }, { "epoch": 13.91, "eval_loss": 0.13568097352981567, "eval_runtime": 4571.8814, "eval_samples_per_second": 16.535, "eval_steps_per_second": 1.378, "eval_wer": 0.056427605495991545, "step": 168000 }, { "epoch": 13.92, "learning_rate": 1.3099168744804655e-05, "loss": 0.0399, "step": 168100 }, { "epoch": 13.92, "learning_rate": 1.3095012468827932e-05, "loss": 0.0451, "step": 168200 }, { "epoch": 13.93, "learning_rate": 1.3090856192851206e-05, "loss": 0.0465, "step": 168300 }, { "epoch": 13.94, "learning_rate": 1.3086699916874482e-05, "loss": 0.0424, "step": 168400 }, { "epoch": 13.95, "learning_rate": 1.3082543640897755e-05, "loss": 0.0468, "step": 168500 }, { "epoch": 13.95, "eval_loss": 0.13916262984275818, "eval_runtime": 4502.5222, "eval_samples_per_second": 16.789, "eval_steps_per_second": 1.399, "eval_wer": 0.05687529711492752, "step": 168500 }, { "epoch": 13.96, "learning_rate": 1.3078387364921033e-05, "loss": 0.0435, "step": 168600 }, { "epoch": 13.97, "learning_rate": 1.3074231088944306e-05, "loss": 0.042, "step": 168700 }, { "epoch": 13.97, "learning_rate": 1.3070074812967582e-05, "loss": 0.0436, "step": 168800 }, { "epoch": 13.98, "learning_rate": 1.3065918536990857e-05, "loss": 0.0414, "step": 168900 }, { "epoch": 13.99, "learning_rate": 1.3061762261014133e-05, "loss": 0.0462, "step": 169000 }, { "epoch": 13.99, "eval_loss": 0.12307950854301453, "eval_runtime": 4550.7016, "eval_samples_per_second": 16.612, "eval_steps_per_second": 1.384, "eval_wer": 0.05413530133030565, "step": 169000 }, { "epoch": 14.0, "learning_rate": 1.3057605985037407e-05, "loss": 0.0414, "step": 169100 }, { "epoch": 14.01, "learning_rate": 1.3053449709060684e-05, "loss": 0.0432, "step": 169200 }, { "epoch": 14.01, "learning_rate": 1.3049293433083958e-05, "loss": 0.0467, "step": 169300 }, { "epoch": 14.02, "learning_rate": 1.3045137157107233e-05, "loss": 0.0483, "step": 169400 }, { "epoch": 14.03, "learning_rate": 1.3040980881130507e-05, "loss": 0.0395, "step": 169500 }, { "epoch": 14.03, "eval_loss": 0.14031049609184265, "eval_runtime": 4518.8267, "eval_samples_per_second": 16.729, "eval_steps_per_second": 1.394, "eval_wer": 0.05584452946995466, "step": 169500 }, { "epoch": 14.04, "learning_rate": 1.3036824605153784e-05, "loss": 0.0396, "step": 169600 }, { "epoch": 14.05, "learning_rate": 1.3032668329177058e-05, "loss": 0.0397, "step": 169700 }, { "epoch": 14.06, "learning_rate": 1.3028512053200332e-05, "loss": 0.0463, "step": 169800 }, { "epoch": 14.06, "learning_rate": 1.302435577722361e-05, "loss": 0.0379, "step": 169900 }, { "epoch": 14.07, "learning_rate": 1.3020199501246885e-05, "loss": 0.0351, "step": 170000 }, { "epoch": 14.07, "eval_loss": 0.1400744915008545, "eval_runtime": 4525.4812, "eval_samples_per_second": 16.704, "eval_steps_per_second": 1.392, "eval_wer": 0.05364914823207964, "step": 170000 }, { "epoch": 14.08, "learning_rate": 1.3016043225270159e-05, "loss": 0.0478, "step": 170100 }, { "epoch": 14.09, "learning_rate": 1.3011886949293436e-05, "loss": 0.0431, "step": 170200 }, { "epoch": 14.1, "learning_rate": 1.300773067331671e-05, "loss": 0.0417, "step": 170300 }, { "epoch": 14.11, "learning_rate": 1.3003574397339983e-05, "loss": 0.0467, "step": 170400 }, { "epoch": 14.11, "learning_rate": 1.2999418121363259e-05, "loss": 0.0439, "step": 170500 }, { "epoch": 14.11, "eval_loss": 0.13535217940807343, "eval_runtime": 4515.3915, "eval_samples_per_second": 16.742, "eval_steps_per_second": 1.395, "eval_wer": 0.05464606977527727, "step": 170500 }, { "epoch": 14.12, "learning_rate": 1.2995261845386536e-05, "loss": 0.0487, "step": 170600 }, { "epoch": 14.13, "learning_rate": 1.299110556940981e-05, "loss": 0.0445, "step": 170700 }, { "epoch": 14.14, "learning_rate": 1.2986949293433084e-05, "loss": 0.0448, "step": 170800 }, { "epoch": 14.15, "learning_rate": 1.2982793017456361e-05, "loss": 0.0453, "step": 170900 }, { "epoch": 14.16, "learning_rate": 1.2978636741479635e-05, "loss": 0.0369, "step": 171000 }, { "epoch": 14.16, "eval_loss": 0.14505808055400848, "eval_runtime": 4523.3986, "eval_samples_per_second": 16.712, "eval_steps_per_second": 1.393, "eval_wer": 0.055692222011966136, "step": 171000 }, { "epoch": 14.16, "learning_rate": 1.297448046550291e-05, "loss": 0.0444, "step": 171100 }, { "epoch": 14.17, "learning_rate": 1.2970324189526186e-05, "loss": 0.0426, "step": 171200 }, { "epoch": 14.18, "learning_rate": 1.2966167913549461e-05, "loss": 0.0513, "step": 171300 }, { "epoch": 14.19, "learning_rate": 1.2962011637572735e-05, "loss": 0.042, "step": 171400 }, { "epoch": 14.2, "learning_rate": 1.295785536159601e-05, "loss": 0.0367, "step": 171500 }, { "epoch": 14.2, "eval_loss": 0.13588671386241913, "eval_runtime": 4534.9748, "eval_samples_per_second": 16.669, "eval_steps_per_second": 1.389, "eval_wer": 0.05554760684983562, "step": 171500 }, { "epoch": 14.21, "learning_rate": 1.2953699085619286e-05, "loss": 0.0462, "step": 171600 }, { "epoch": 14.21, "learning_rate": 1.2949542809642562e-05, "loss": 0.046, "step": 171700 }, { "epoch": 14.22, "learning_rate": 1.2945386533665836e-05, "loss": 0.0401, "step": 171800 }, { "epoch": 14.23, "learning_rate": 1.2941230257689113e-05, "loss": 0.0444, "step": 171900 }, { "epoch": 14.24, "learning_rate": 1.2937073981712387e-05, "loss": 0.041, "step": 172000 }, { "epoch": 14.24, "eval_loss": 0.14004744589328766, "eval_runtime": 4548.1259, "eval_samples_per_second": 16.621, "eval_steps_per_second": 1.385, "eval_wer": 0.055886067867587895, "step": 172000 }, { "epoch": 14.25, "learning_rate": 1.2932917705735662e-05, "loss": 0.0418, "step": 172100 }, { "epoch": 14.25, "learning_rate": 1.2928761429758938e-05, "loss": 0.0485, "step": 172200 }, { "epoch": 14.26, "learning_rate": 1.2924605153782213e-05, "loss": 0.0544, "step": 172300 }, { "epoch": 14.27, "learning_rate": 1.2920448877805487e-05, "loss": 0.0417, "step": 172400 }, { "epoch": 14.28, "learning_rate": 1.291629260182876e-05, "loss": 0.0414, "step": 172500 }, { "epoch": 14.28, "eval_loss": 0.14940544962882996, "eval_runtime": 4518.2465, "eval_samples_per_second": 16.731, "eval_steps_per_second": 1.394, "eval_wer": 0.0595122161350521, "step": 172500 }, { "epoch": 14.29, "learning_rate": 1.2912136325852038e-05, "loss": 0.0418, "step": 172600 }, { "epoch": 14.3, "learning_rate": 1.2907980049875314e-05, "loss": 0.0422, "step": 172700 }, { "epoch": 14.3, "learning_rate": 1.2903823773898587e-05, "loss": 0.044, "step": 172800 }, { "epoch": 14.31, "learning_rate": 1.2899667497921865e-05, "loss": 0.0458, "step": 172900 }, { "epoch": 14.32, "learning_rate": 1.2895511221945138e-05, "loss": 0.0443, "step": 173000 }, { "epoch": 14.32, "eval_loss": 0.1440606713294983, "eval_runtime": 4699.3809, "eval_samples_per_second": 16.086, "eval_steps_per_second": 1.341, "eval_wer": 0.055621452890072476, "step": 173000 }, { "epoch": 14.33, "learning_rate": 1.2891354945968412e-05, "loss": 0.0409, "step": 173100 }, { "epoch": 14.34, "learning_rate": 1.288719866999169e-05, "loss": 0.0386, "step": 173200 }, { "epoch": 14.35, "learning_rate": 1.2883083956774731e-05, "loss": 0.0429, "step": 173300 }, { "epoch": 14.35, "learning_rate": 1.2878927680798005e-05, "loss": 0.0401, "step": 173400 }, { "epoch": 14.36, "learning_rate": 1.287477140482128e-05, "loss": 0.0456, "step": 173500 }, { "epoch": 14.36, "eval_loss": 0.14035511016845703, "eval_runtime": 5057.2152, "eval_samples_per_second": 14.948, "eval_steps_per_second": 1.246, "eval_wer": 0.056599912923210885, "step": 173500 }, { "epoch": 14.37, "learning_rate": 1.2870615128844556e-05, "loss": 0.0397, "step": 173600 }, { "epoch": 14.38, "learning_rate": 1.2866458852867832e-05, "loss": 0.0495, "step": 173700 }, { "epoch": 14.39, "learning_rate": 1.2862302576891105e-05, "loss": 0.0528, "step": 173800 }, { "epoch": 14.4, "learning_rate": 1.2858146300914383e-05, "loss": 0.0409, "step": 173900 }, { "epoch": 14.4, "learning_rate": 1.2853990024937656e-05, "loss": 0.0441, "step": 174000 }, { "epoch": 14.4, "eval_loss": 0.13616320490837097, "eval_runtime": 5002.2066, "eval_samples_per_second": 15.112, "eval_steps_per_second": 1.259, "eval_wer": 0.05525376114805977, "step": 174000 }, { "epoch": 14.41, "learning_rate": 1.2849833748960932e-05, "loss": 0.0433, "step": 174100 }, { "epoch": 14.42, "learning_rate": 1.2845677472984208e-05, "loss": 0.0435, "step": 174200 }, { "epoch": 14.43, "learning_rate": 1.2841521197007483e-05, "loss": 0.0438, "step": 174300 }, { "epoch": 14.44, "learning_rate": 1.2837364921030757e-05, "loss": 0.0453, "step": 174400 }, { "epoch": 14.45, "learning_rate": 1.2833208645054032e-05, "loss": 0.0536, "step": 174500 }, { "epoch": 14.45, "eval_loss": 0.13783754408359528, "eval_runtime": 5010.5284, "eval_samples_per_second": 15.087, "eval_steps_per_second": 1.257, "eval_wer": 0.05723221964273901, "step": 174500 }, { "epoch": 14.45, "learning_rate": 1.2829052369077308e-05, "loss": 0.0422, "step": 174600 }, { "epoch": 14.46, "learning_rate": 1.2824896093100583e-05, "loss": 0.0411, "step": 174700 }, { "epoch": 14.47, "learning_rate": 1.2820739817123857e-05, "loss": 0.0369, "step": 174800 }, { "epoch": 14.48, "learning_rate": 1.2816583541147134e-05, "loss": 0.045, "step": 174900 }, { "epoch": 14.49, "learning_rate": 1.2812427265170408e-05, "loss": 0.0394, "step": 175000 }, { "epoch": 14.49, "eval_loss": 0.1492939591407776, "eval_runtime": 4793.0764, "eval_samples_per_second": 15.772, "eval_steps_per_second": 1.314, "eval_wer": 0.05800760306522605, "step": 175000 }, { "epoch": 14.5, "learning_rate": 1.2808270989193684e-05, "loss": 0.0448, "step": 175100 }, { "epoch": 14.5, "learning_rate": 1.280411471321696e-05, "loss": 0.0432, "step": 175200 }, { "epoch": 14.51, "learning_rate": 1.2800000000000001e-05, "loss": 0.0427, "step": 175300 }, { "epoch": 14.52, "learning_rate": 1.2795843724023277e-05, "loss": 0.0406, "step": 175400 }, { "epoch": 14.53, "learning_rate": 1.279168744804655e-05, "loss": 0.0401, "step": 175500 }, { "epoch": 14.53, "eval_loss": 0.14772075414657593, "eval_runtime": 4681.1791, "eval_samples_per_second": 16.149, "eval_steps_per_second": 1.346, "eval_wer": 0.05725837344865623, "step": 175500 }, { "epoch": 14.54, "learning_rate": 1.2787531172069828e-05, "loss": 0.0398, "step": 175600 }, { "epoch": 14.54, "learning_rate": 1.2783374896093101e-05, "loss": 0.0437, "step": 175700 }, { "epoch": 14.55, "learning_rate": 1.2779218620116375e-05, "loss": 0.0395, "step": 175800 }, { "epoch": 14.56, "learning_rate": 1.2775062344139653e-05, "loss": 0.0394, "step": 175900 }, { "epoch": 14.57, "learning_rate": 1.2770906068162926e-05, "loss": 0.0408, "step": 176000 }, { "epoch": 14.57, "eval_loss": 0.14988575875759125, "eval_runtime": 5018.6468, "eval_samples_per_second": 15.063, "eval_steps_per_second": 1.255, "eval_wer": 0.05723991193859702, "step": 176000 }, { "epoch": 14.58, "learning_rate": 1.2766749792186202e-05, "loss": 0.0376, "step": 176100 }, { "epoch": 14.59, "learning_rate": 1.2762593516209479e-05, "loss": 0.0453, "step": 176200 }, { "epoch": 14.59, "learning_rate": 1.2758437240232753e-05, "loss": 0.0445, "step": 176300 }, { "epoch": 14.6, "learning_rate": 1.2754280964256027e-05, "loss": 0.0402, "step": 176400 }, { "epoch": 14.61, "learning_rate": 1.2750124688279302e-05, "loss": 0.0405, "step": 176500 }, { "epoch": 14.61, "eval_loss": 0.14350935816764832, "eval_runtime": 4683.3367, "eval_samples_per_second": 16.141, "eval_steps_per_second": 1.345, "eval_wer": 0.05476299267231897, "step": 176500 }, { "epoch": 14.62, "learning_rate": 1.2745968412302578e-05, "loss": 0.0478, "step": 176600 }, { "epoch": 14.63, "learning_rate": 1.2741812136325853e-05, "loss": 0.0394, "step": 176700 }, { "epoch": 14.64, "learning_rate": 1.2737655860349127e-05, "loss": 0.0383, "step": 176800 }, { "epoch": 14.64, "learning_rate": 1.2733499584372404e-05, "loss": 0.0484, "step": 176900 }, { "epoch": 14.65, "learning_rate": 1.2729343308395678e-05, "loss": 0.0476, "step": 177000 }, { "epoch": 14.65, "eval_loss": 0.1424773782491684, "eval_runtime": 4710.1392, "eval_samples_per_second": 16.049, "eval_steps_per_second": 1.338, "eval_wer": 0.05570914506285375, "step": 177000 }, { "epoch": 14.66, "learning_rate": 1.2725187032418954e-05, "loss": 0.0411, "step": 177100 }, { "epoch": 14.67, "learning_rate": 1.2721030756442229e-05, "loss": 0.0375, "step": 177200 }, { "epoch": 14.68, "learning_rate": 1.2716874480465505e-05, "loss": 0.0426, "step": 177300 }, { "epoch": 14.69, "learning_rate": 1.2712718204488778e-05, "loss": 0.0457, "step": 177400 }, { "epoch": 14.69, "learning_rate": 1.2708561928512054e-05, "loss": 0.0439, "step": 177500 }, { "epoch": 14.69, "eval_loss": 0.1372547596693039, "eval_runtime": 4868.2023, "eval_samples_per_second": 15.528, "eval_steps_per_second": 1.294, "eval_wer": 0.055013761517289976, "step": 177500 }, { "epoch": 14.7, "learning_rate": 1.270440565253533e-05, "loss": 0.0473, "step": 177600 }, { "epoch": 14.71, "learning_rate": 1.2700249376558605e-05, "loss": 0.039, "step": 177700 }, { "epoch": 14.72, "learning_rate": 1.2696093100581879e-05, "loss": 0.0458, "step": 177800 }, { "epoch": 14.73, "learning_rate": 1.2691936824605156e-05, "loss": 0.043, "step": 177900 }, { "epoch": 14.74, "learning_rate": 1.2687822111388198e-05, "loss": 0.0397, "step": 178000 }, { "epoch": 14.74, "eval_loss": 0.1475825309753418, "eval_runtime": 4708.5663, "eval_samples_per_second": 16.055, "eval_steps_per_second": 1.338, "eval_wer": 0.058610679060493755, "step": 178000 }, { "epoch": 14.74, "learning_rate": 1.2683665835411472e-05, "loss": 0.0438, "step": 178100 }, { "epoch": 14.75, "learning_rate": 1.2679509559434749e-05, "loss": 0.0461, "step": 178200 }, { "epoch": 14.76, "learning_rate": 1.2675353283458023e-05, "loss": 0.045, "step": 178300 }, { "epoch": 14.77, "learning_rate": 1.2671197007481297e-05, "loss": 0.045, "step": 178400 }, { "epoch": 14.78, "learning_rate": 1.2667040731504574e-05, "loss": 0.041, "step": 178500 }, { "epoch": 14.78, "eval_loss": 0.15099816024303436, "eval_runtime": 4461.7179, "eval_samples_per_second": 16.943, "eval_steps_per_second": 1.412, "eval_wer": 0.05799068001433844, "step": 178500 }, { "epoch": 14.78, "learning_rate": 1.266288445552785e-05, "loss": 0.041, "step": 178600 }, { "epoch": 14.79, "learning_rate": 1.2658728179551123e-05, "loss": 0.043, "step": 178700 }, { "epoch": 14.8, "learning_rate": 1.2654571903574397e-05, "loss": 0.0402, "step": 178800 }, { "epoch": 14.81, "learning_rate": 1.2650415627597674e-05, "loss": 0.0448, "step": 178900 }, { "epoch": 14.82, "learning_rate": 1.2646259351620948e-05, "loss": 0.0399, "step": 179000 }, { "epoch": 14.82, "eval_loss": 0.1468614786863327, "eval_runtime": 4514.8927, "eval_samples_per_second": 16.743, "eval_steps_per_second": 1.395, "eval_wer": 0.055835298714925055, "step": 179000 }, { "epoch": 14.83, "learning_rate": 1.2642103075644223e-05, "loss": 0.0381, "step": 179100 }, { "epoch": 14.83, "learning_rate": 1.2637946799667499e-05, "loss": 0.0433, "step": 179200 }, { "epoch": 14.84, "learning_rate": 1.2633790523690775e-05, "loss": 0.0433, "step": 179300 }, { "epoch": 14.85, "learning_rate": 1.2629634247714048e-05, "loss": 0.0378, "step": 179400 }, { "epoch": 14.86, "learning_rate": 1.2625477971737326e-05, "loss": 0.0424, "step": 179500 }, { "epoch": 14.86, "eval_loss": 0.13825999200344086, "eval_runtime": 4504.5295, "eval_samples_per_second": 16.782, "eval_steps_per_second": 1.399, "eval_wer": 0.05606914450900845, "step": 179500 }, { "epoch": 14.87, "learning_rate": 1.26213216957606e-05, "loss": 0.0429, "step": 179600 }, { "epoch": 14.88, "learning_rate": 1.2617165419783875e-05, "loss": 0.0406, "step": 179700 }, { "epoch": 14.88, "learning_rate": 1.2613009143807149e-05, "loss": 0.0404, "step": 179800 }, { "epoch": 14.89, "learning_rate": 1.2608852867830426e-05, "loss": 0.0433, "step": 179900 }, { "epoch": 14.9, "learning_rate": 1.26046965918537e-05, "loss": 0.0383, "step": 180000 }, { "epoch": 14.9, "eval_loss": 0.1468030959367752, "eval_runtime": 4400.1728, "eval_samples_per_second": 17.18, "eval_steps_per_second": 1.432, "eval_wer": 0.05610606752912688, "step": 180000 }, { "epoch": 14.91, "learning_rate": 1.2600540315876975e-05, "loss": 0.041, "step": 180100 }, { "epoch": 14.92, "learning_rate": 1.259638403990025e-05, "loss": 0.0452, "step": 180200 }, { "epoch": 14.93, "learning_rate": 1.2592227763923526e-05, "loss": 0.0469, "step": 180300 }, { "epoch": 14.93, "learning_rate": 1.25880714879468e-05, "loss": 0.0408, "step": 180400 }, { "epoch": 14.94, "learning_rate": 1.2583915211970077e-05, "loss": 0.0494, "step": 180500 }, { "epoch": 14.94, "eval_loss": 0.13648249208927155, "eval_runtime": 4474.5888, "eval_samples_per_second": 16.894, "eval_steps_per_second": 1.408, "eval_wer": 0.05615529822261812, "step": 180500 }, { "epoch": 14.95, "learning_rate": 1.2579758935993351e-05, "loss": 0.0492, "step": 180600 }, { "epoch": 14.96, "learning_rate": 1.2575602660016627e-05, "loss": 0.0462, "step": 180700 }, { "epoch": 14.97, "learning_rate": 1.25714463840399e-05, "loss": 0.0454, "step": 180800 }, { "epoch": 14.98, "learning_rate": 1.2567290108063178e-05, "loss": 0.0424, "step": 180900 }, { "epoch": 14.98, "learning_rate": 1.2563133832086452e-05, "loss": 0.0379, "step": 181000 }, { "epoch": 14.98, "eval_loss": 0.13876716792583466, "eval_runtime": 4423.5892, "eval_samples_per_second": 17.089, "eval_steps_per_second": 1.424, "eval_wer": 0.05439376247113466, "step": 181000 }, { "epoch": 14.99, "learning_rate": 1.2558977556109725e-05, "loss": 0.0398, "step": 181100 }, { "epoch": 15.0, "learning_rate": 1.2554821280133003e-05, "loss": 0.0444, "step": 181200 }, { "epoch": 15.01, "learning_rate": 1.2550665004156276e-05, "loss": 0.0436, "step": 181300 }, { "epoch": 15.02, "learning_rate": 1.2546508728179552e-05, "loss": 0.0437, "step": 181400 }, { "epoch": 15.02, "learning_rate": 1.2542352452202829e-05, "loss": 0.0416, "step": 181500 }, { "epoch": 15.02, "eval_loss": 0.14462417364120483, "eval_runtime": 4459.5144, "eval_samples_per_second": 16.951, "eval_steps_per_second": 1.413, "eval_wer": 0.05537068404510147, "step": 181500 }, { "epoch": 15.03, "learning_rate": 1.2538196176226103e-05, "loss": 0.0416, "step": 181600 }, { "epoch": 15.04, "learning_rate": 1.2534039900249377e-05, "loss": 0.042, "step": 181700 }, { "epoch": 15.05, "learning_rate": 1.2529883624272652e-05, "loss": 0.0396, "step": 181800 }, { "epoch": 15.06, "learning_rate": 1.2525727348295928e-05, "loss": 0.0396, "step": 181900 }, { "epoch": 15.07, "learning_rate": 1.2521571072319203e-05, "loss": 0.0391, "step": 182000 }, { "epoch": 15.07, "eval_loss": 0.13681894540786743, "eval_runtime": 4427.2139, "eval_samples_per_second": 17.075, "eval_steps_per_second": 1.423, "eval_wer": 0.054699915846283316, "step": 182000 }, { "epoch": 15.07, "learning_rate": 1.2517414796342477e-05, "loss": 0.0433, "step": 182100 }, { "epoch": 15.08, "learning_rate": 1.2513258520365754e-05, "loss": 0.0445, "step": 182200 }, { "epoch": 15.09, "learning_rate": 1.2509102244389028e-05, "loss": 0.0425, "step": 182300 }, { "epoch": 15.1, "learning_rate": 1.2504945968412304e-05, "loss": 0.04, "step": 182400 }, { "epoch": 15.11, "learning_rate": 1.2500831255195347e-05, "loss": 0.0381, "step": 182500 }, { "epoch": 15.11, "eval_loss": 0.1466192752122879, "eval_runtime": 4423.3341, "eval_samples_per_second": 17.09, "eval_steps_per_second": 1.424, "eval_wer": 0.055892221704274304, "step": 182500 }, { "epoch": 15.12, "learning_rate": 1.2496674979218621e-05, "loss": 0.0492, "step": 182600 }, { "epoch": 15.12, "learning_rate": 1.2492518703241897e-05, "loss": 0.0428, "step": 182700 }, { "epoch": 15.13, "learning_rate": 1.248836242726517e-05, "loss": 0.0443, "step": 182800 }, { "epoch": 15.14, "learning_rate": 1.2484206151288448e-05, "loss": 0.0457, "step": 182900 }, { "epoch": 15.15, "learning_rate": 1.2480049875311721e-05, "loss": 0.0398, "step": 183000 }, { "epoch": 15.15, "eval_loss": 0.14347638189792633, "eval_runtime": 4421.7332, "eval_samples_per_second": 17.096, "eval_steps_per_second": 1.425, "eval_wer": 0.05763068056818374, "step": 183000 }, { "epoch": 15.16, "learning_rate": 1.2475893599334997e-05, "loss": 0.0385, "step": 183100 }, { "epoch": 15.17, "learning_rate": 1.2471737323358272e-05, "loss": 0.0441, "step": 183200 }, { "epoch": 15.17, "learning_rate": 1.2467581047381548e-05, "loss": 0.0428, "step": 183300 }, { "epoch": 15.18, "learning_rate": 1.2463424771404822e-05, "loss": 0.0422, "step": 183400 }, { "epoch": 15.19, "learning_rate": 1.2459268495428099e-05, "loss": 0.0359, "step": 183500 }, { "epoch": 15.19, "eval_loss": 0.15226058661937714, "eval_runtime": 4450.7908, "eval_samples_per_second": 16.985, "eval_steps_per_second": 1.415, "eval_wer": 0.05690760475753114, "step": 183500 }, { "epoch": 15.2, "learning_rate": 1.2455112219451373e-05, "loss": 0.0441, "step": 183600 }, { "epoch": 15.21, "learning_rate": 1.2450955943474647e-05, "loss": 0.0368, "step": 183700 }, { "epoch": 15.22, "learning_rate": 1.2446799667497922e-05, "loss": 0.0456, "step": 183800 }, { "epoch": 15.22, "learning_rate": 1.24426433915212e-05, "loss": 0.0534, "step": 183900 }, { "epoch": 15.23, "learning_rate": 1.2438487115544473e-05, "loss": 0.0368, "step": 184000 }, { "epoch": 15.23, "eval_loss": 0.14243130385875702, "eval_runtime": 4304.6104, "eval_samples_per_second": 17.561, "eval_steps_per_second": 1.464, "eval_wer": 0.05531683797409542, "step": 184000 }, { "epoch": 15.24, "learning_rate": 1.2434330839567747e-05, "loss": 0.0391, "step": 184100 }, { "epoch": 15.25, "learning_rate": 1.2430174563591024e-05, "loss": 0.0387, "step": 184200 }, { "epoch": 15.26, "learning_rate": 1.2426018287614298e-05, "loss": 0.0389, "step": 184300 }, { "epoch": 15.26, "learning_rate": 1.2421862011637574e-05, "loss": 0.0377, "step": 184400 }, { "epoch": 15.27, "learning_rate": 1.241770573566085e-05, "loss": 0.0395, "step": 184500 }, { "epoch": 15.27, "eval_loss": 0.14944978058338165, "eval_runtime": 4339.957, "eval_samples_per_second": 17.418, "eval_steps_per_second": 1.452, "eval_wer": 0.05531222259658062, "step": 184500 }, { "epoch": 15.28, "learning_rate": 1.2413591022443891e-05, "loss": 0.0398, "step": 184600 }, { "epoch": 15.29, "learning_rate": 1.2409434746467166e-05, "loss": 0.0418, "step": 184700 }, { "epoch": 15.3, "learning_rate": 1.240527847049044e-05, "loss": 0.0412, "step": 184800 }, { "epoch": 15.31, "learning_rate": 1.2401122194513717e-05, "loss": 0.0379, "step": 184900 }, { "epoch": 15.31, "learning_rate": 1.2396965918536991e-05, "loss": 0.0474, "step": 185000 }, { "epoch": 15.31, "eval_loss": 0.14591313898563385, "eval_runtime": 4406.8776, "eval_samples_per_second": 17.154, "eval_steps_per_second": 1.43, "eval_wer": 0.05662914364747131, "step": 185000 }, { "epoch": 15.32, "learning_rate": 1.2392809642560267e-05, "loss": 0.0419, "step": 185100 }, { "epoch": 15.33, "learning_rate": 1.2388653366583542e-05, "loss": 0.0364, "step": 185200 }, { "epoch": 15.34, "learning_rate": 1.2384497090606818e-05, "loss": 0.0429, "step": 185300 }, { "epoch": 15.35, "learning_rate": 1.2380340814630092e-05, "loss": 0.0462, "step": 185400 }, { "epoch": 15.36, "learning_rate": 1.2376184538653369e-05, "loss": 0.045, "step": 185500 }, { "epoch": 15.36, "eval_loss": 0.14320936799049377, "eval_runtime": 4309.2129, "eval_samples_per_second": 17.543, "eval_steps_per_second": 1.462, "eval_wer": 0.056219913507825375, "step": 185500 }, { "epoch": 15.36, "learning_rate": 1.2372028262676643e-05, "loss": 0.0386, "step": 185600 }, { "epoch": 15.37, "learning_rate": 1.2367913549459684e-05, "loss": 0.0424, "step": 185700 }, { "epoch": 15.38, "learning_rate": 1.236375727348296e-05, "loss": 0.036, "step": 185800 }, { "epoch": 15.39, "learning_rate": 1.2359642560266004e-05, "loss": 0.0417, "step": 185900 }, { "epoch": 15.4, "learning_rate": 1.2355486284289277e-05, "loss": 0.0358, "step": 186000 }, { "epoch": 15.4, "eval_loss": 0.1542247086763382, "eval_runtime": 4326.2543, "eval_samples_per_second": 17.474, "eval_steps_per_second": 1.456, "eval_wer": 0.05686606635989791, "step": 186000 }, { "epoch": 15.41, "learning_rate": 1.2351330008312553e-05, "loss": 0.0472, "step": 186100 }, { "epoch": 15.41, "learning_rate": 1.2347173732335828e-05, "loss": 0.0484, "step": 186200 }, { "epoch": 15.42, "learning_rate": 1.2343017456359104e-05, "loss": 0.0542, "step": 186300 }, { "epoch": 15.43, "learning_rate": 1.2338861180382378e-05, "loss": 0.0383, "step": 186400 }, { "epoch": 15.44, "learning_rate": 1.2334704904405655e-05, "loss": 0.0391, "step": 186500 }, { "epoch": 15.44, "eval_loss": 0.14198127388954163, "eval_runtime": 4325.297, "eval_samples_per_second": 17.477, "eval_steps_per_second": 1.457, "eval_wer": 0.054416839358708675, "step": 186500 }, { "epoch": 15.45, "learning_rate": 1.2330548628428929e-05, "loss": 0.041, "step": 186600 }, { "epoch": 15.46, "learning_rate": 1.2326392352452203e-05, "loss": 0.0425, "step": 186700 }, { "epoch": 15.46, "learning_rate": 1.232223607647548e-05, "loss": 0.0448, "step": 186800 }, { "epoch": 15.47, "learning_rate": 1.2318079800498755e-05, "loss": 0.0432, "step": 186900 }, { "epoch": 15.48, "learning_rate": 1.2313923524522029e-05, "loss": 0.0401, "step": 187000 }, { "epoch": 15.48, "eval_loss": 0.13979791104793549, "eval_runtime": 4410.1921, "eval_samples_per_second": 17.141, "eval_steps_per_second": 1.429, "eval_wer": 0.056099913692440474, "step": 187000 }, { "epoch": 15.49, "learning_rate": 1.2309767248545303e-05, "loss": 0.0399, "step": 187100 }, { "epoch": 15.5, "learning_rate": 1.230561097256858e-05, "loss": 0.0418, "step": 187200 }, { "epoch": 15.5, "learning_rate": 1.2301454696591854e-05, "loss": 0.0447, "step": 187300 }, { "epoch": 15.51, "learning_rate": 1.229729842061513e-05, "loss": 0.0428, "step": 187400 }, { "epoch": 15.52, "learning_rate": 1.2293142144638405e-05, "loss": 0.0356, "step": 187500 }, { "epoch": 15.52, "eval_loss": 0.14980462193489075, "eval_runtime": 4333.3482, "eval_samples_per_second": 17.445, "eval_steps_per_second": 1.454, "eval_wer": 0.05524606885220176, "step": 187500 }, { "epoch": 15.53, "learning_rate": 1.228898586866168e-05, "loss": 0.0463, "step": 187600 }, { "epoch": 15.54, "learning_rate": 1.2284829592684954e-05, "loss": 0.0403, "step": 187700 }, { "epoch": 15.55, "learning_rate": 1.2280673316708232e-05, "loss": 0.0367, "step": 187800 }, { "epoch": 15.55, "learning_rate": 1.2276517040731505e-05, "loss": 0.0393, "step": 187900 }, { "epoch": 15.56, "learning_rate": 1.2272360764754781e-05, "loss": 0.0385, "step": 188000 }, { "epoch": 15.56, "eval_loss": 0.15155218541622162, "eval_runtime": 4322.2554, "eval_samples_per_second": 17.49, "eval_steps_per_second": 1.458, "eval_wer": 0.05777375727114266, "step": 188000 }, { "epoch": 15.57, "learning_rate": 1.2268204488778055e-05, "loss": 0.0407, "step": 188100 }, { "epoch": 15.58, "learning_rate": 1.2264048212801332e-05, "loss": 0.0396, "step": 188200 }, { "epoch": 15.59, "learning_rate": 1.2259891936824606e-05, "loss": 0.037, "step": 188300 }, { "epoch": 15.6, "learning_rate": 1.2255735660847881e-05, "loss": 0.0381, "step": 188400 }, { "epoch": 15.6, "learning_rate": 1.2251579384871157e-05, "loss": 0.0423, "step": 188500 }, { "epoch": 15.6, "eval_loss": 0.14062604308128357, "eval_runtime": 4323.2118, "eval_samples_per_second": 17.486, "eval_steps_per_second": 1.457, "eval_wer": 0.05568145279776493, "step": 188500 }, { "epoch": 15.61, "learning_rate": 1.2247423108894432e-05, "loss": 0.0422, "step": 188600 }, { "epoch": 15.62, "learning_rate": 1.2243266832917706e-05, "loss": 0.0391, "step": 188700 }, { "epoch": 15.63, "learning_rate": 1.2239110556940983e-05, "loss": 0.0447, "step": 188800 }, { "epoch": 15.64, "learning_rate": 1.2234954280964257e-05, "loss": 0.0471, "step": 188900 }, { "epoch": 15.65, "learning_rate": 1.2230798004987533e-05, "loss": 0.0416, "step": 189000 }, { "epoch": 15.65, "eval_loss": 0.1473342925310135, "eval_runtime": 4308.4558, "eval_samples_per_second": 17.546, "eval_steps_per_second": 1.462, "eval_wer": 0.0563691440474707, "step": 189000 }, { "epoch": 15.65, "learning_rate": 1.2226641729010806e-05, "loss": 0.0385, "step": 189100 }, { "epoch": 15.66, "learning_rate": 1.2222485453034084e-05, "loss": 0.0353, "step": 189200 }, { "epoch": 15.67, "learning_rate": 1.2218329177057358e-05, "loss": 0.0424, "step": 189300 }, { "epoch": 15.68, "learning_rate": 1.2214172901080631e-05, "loss": 0.0455, "step": 189400 }, { "epoch": 15.69, "learning_rate": 1.2210016625103909e-05, "loss": 0.0407, "step": 189500 }, { "epoch": 15.69, "eval_loss": 0.13586293160915375, "eval_runtime": 4297.9086, "eval_samples_per_second": 17.589, "eval_steps_per_second": 1.466, "eval_wer": 0.056955296991850785, "step": 189500 }, { "epoch": 15.7, "learning_rate": 1.2205860349127182e-05, "loss": 0.0411, "step": 189600 }, { "epoch": 15.7, "learning_rate": 1.2201704073150458e-05, "loss": 0.0427, "step": 189700 }, { "epoch": 15.71, "learning_rate": 1.2197547797173735e-05, "loss": 0.0396, "step": 189800 }, { "epoch": 15.72, "learning_rate": 1.2193391521197009e-05, "loss": 0.0446, "step": 189900 }, { "epoch": 15.73, "learning_rate": 1.2189235245220283e-05, "loss": 0.0459, "step": 190000 }, { "epoch": 15.73, "eval_loss": 0.1482781618833542, "eval_runtime": 4379.6251, "eval_samples_per_second": 17.261, "eval_steps_per_second": 1.438, "eval_wer": 0.05750144999776923, "step": 190000 }, { "epoch": 15.74, "learning_rate": 1.2185078969243558e-05, "loss": 0.0446, "step": 190100 }, { "epoch": 15.75, "learning_rate": 1.2180964256026602e-05, "loss": 0.0381, "step": 190200 }, { "epoch": 15.75, "learning_rate": 1.2176807980049876e-05, "loss": 0.0454, "step": 190300 }, { "epoch": 15.76, "learning_rate": 1.2172651704073151e-05, "loss": 0.0456, "step": 190400 }, { "epoch": 15.77, "learning_rate": 1.2168495428096427e-05, "loss": 0.0363, "step": 190500 }, { "epoch": 15.77, "eval_loss": 0.14273369312286377, "eval_runtime": 4328.1672, "eval_samples_per_second": 17.466, "eval_steps_per_second": 1.456, "eval_wer": 0.05556606835989483, "step": 190500 }, { "epoch": 15.78, "learning_rate": 1.2164339152119702e-05, "loss": 0.0397, "step": 190600 }, { "epoch": 15.79, "learning_rate": 1.2160182876142976e-05, "loss": 0.043, "step": 190700 }, { "epoch": 15.79, "learning_rate": 1.2156026600166253e-05, "loss": 0.0452, "step": 190800 }, { "epoch": 15.8, "learning_rate": 1.2151870324189527e-05, "loss": 0.0374, "step": 190900 }, { "epoch": 15.81, "learning_rate": 1.2147714048212803e-05, "loss": 0.0379, "step": 191000 }, { "epoch": 15.81, "eval_loss": 0.14714951813220978, "eval_runtime": 4379.3528, "eval_samples_per_second": 17.262, "eval_steps_per_second": 1.439, "eval_wer": 0.05569683738948094, "step": 191000 }, { "epoch": 15.82, "learning_rate": 1.2143557772236076e-05, "loss": 0.0446, "step": 191100 }, { "epoch": 15.83, "learning_rate": 1.2139401496259354e-05, "loss": 0.0417, "step": 191200 }, { "epoch": 15.84, "learning_rate": 1.2135286783042395e-05, "loss": 0.0458, "step": 191300 }, { "epoch": 15.84, "learning_rate": 1.213113050706567e-05, "loss": 0.0426, "step": 191400 }, { "epoch": 15.85, "learning_rate": 1.2126974231088946e-05, "loss": 0.0438, "step": 191500 }, { "epoch": 15.85, "eval_loss": 0.14119766652584076, "eval_runtime": 4404.6145, "eval_samples_per_second": 17.163, "eval_steps_per_second": 1.43, "eval_wer": 0.05451530074569116, "step": 191500 }, { "epoch": 15.86, "learning_rate": 1.212281795511222e-05, "loss": 0.04, "step": 191600 }, { "epoch": 15.87, "learning_rate": 1.2118661679135496e-05, "loss": 0.0408, "step": 191700 }, { "epoch": 15.88, "learning_rate": 1.2114505403158771e-05, "loss": 0.0367, "step": 191800 }, { "epoch": 15.89, "learning_rate": 1.2110349127182047e-05, "loss": 0.0439, "step": 191900 }, { "epoch": 15.89, "learning_rate": 1.210619285120532e-05, "loss": 0.0408, "step": 192000 }, { "epoch": 15.89, "eval_loss": 0.14833366870880127, "eval_runtime": 4366.2057, "eval_samples_per_second": 17.314, "eval_steps_per_second": 1.443, "eval_wer": 0.057106065990667705, "step": 192000 }, { "epoch": 15.9, "learning_rate": 1.2102036575228594e-05, "loss": 0.0437, "step": 192100 }, { "epoch": 15.91, "learning_rate": 1.2097880299251872e-05, "loss": 0.0449, "step": 192200 }, { "epoch": 15.92, "learning_rate": 1.2093724023275145e-05, "loss": 0.0406, "step": 192300 }, { "epoch": 15.93, "learning_rate": 1.2089567747298421e-05, "loss": 0.0455, "step": 192400 }, { "epoch": 15.94, "learning_rate": 1.2085411471321698e-05, "loss": 0.0458, "step": 192500 }, { "epoch": 15.94, "eval_loss": 0.146177276968956, "eval_runtime": 4363.913, "eval_samples_per_second": 17.323, "eval_steps_per_second": 1.444, "eval_wer": 0.06036298405694761, "step": 192500 }, { "epoch": 15.94, "learning_rate": 1.2081255195344972e-05, "loss": 0.0365, "step": 192600 }, { "epoch": 15.95, "learning_rate": 1.2077098919368246e-05, "loss": 0.0417, "step": 192700 }, { "epoch": 15.96, "learning_rate": 1.2072942643391523e-05, "loss": 0.0371, "step": 192800 }, { "epoch": 15.97, "learning_rate": 1.2068786367414797e-05, "loss": 0.0375, "step": 192900 }, { "epoch": 15.98, "learning_rate": 1.2064630091438072e-05, "loss": 0.0398, "step": 193000 }, { "epoch": 15.98, "eval_loss": 0.15597130358219147, "eval_runtime": 4313.853, "eval_samples_per_second": 17.524, "eval_steps_per_second": 1.46, "eval_wer": 0.05779837261788828, "step": 193000 }, { "epoch": 15.99, "learning_rate": 1.2060473815461346e-05, "loss": 0.0391, "step": 193100 }, { "epoch": 15.99, "learning_rate": 1.2056317539484623e-05, "loss": 0.0352, "step": 193200 }, { "epoch": 16.0, "learning_rate": 1.2052161263507897e-05, "loss": 0.0378, "step": 193300 }, { "epoch": 16.01, "learning_rate": 1.2048004987531173e-05, "loss": 0.0372, "step": 193400 }, { "epoch": 16.02, "learning_rate": 1.2043848711554448e-05, "loss": 0.0391, "step": 193500 }, { "epoch": 16.02, "eval_loss": 0.14235103130340576, "eval_runtime": 4330.2222, "eval_samples_per_second": 17.458, "eval_steps_per_second": 1.455, "eval_wer": 0.05660298984155409, "step": 193500 }, { "epoch": 16.03, "learning_rate": 1.2039692435577724e-05, "loss": 0.0316, "step": 193600 }, { "epoch": 16.03, "learning_rate": 1.2035536159600998e-05, "loss": 0.0348, "step": 193700 }, { "epoch": 16.04, "learning_rate": 1.2031379883624275e-05, "loss": 0.0327, "step": 193800 }, { "epoch": 16.05, "learning_rate": 1.2027223607647549e-05, "loss": 0.0423, "step": 193900 }, { "epoch": 16.06, "learning_rate": 1.2023067331670824e-05, "loss": 0.0371, "step": 194000 }, { "epoch": 16.06, "eval_loss": 0.1485888510942459, "eval_runtime": 4340.0807, "eval_samples_per_second": 17.418, "eval_steps_per_second": 1.452, "eval_wer": 0.05643222087350635, "step": 194000 }, { "epoch": 16.07, "learning_rate": 1.2018911055694098e-05, "loss": 0.0387, "step": 194100 }, { "epoch": 16.08, "learning_rate": 1.2014754779717375e-05, "loss": 0.0361, "step": 194200 }, { "epoch": 16.08, "learning_rate": 1.2010598503740649e-05, "loss": 0.0356, "step": 194300 }, { "epoch": 16.09, "learning_rate": 1.2006442227763923e-05, "loss": 0.0404, "step": 194400 }, { "epoch": 16.1, "learning_rate": 1.20022859517872e-05, "loss": 0.0359, "step": 194500 }, { "epoch": 16.1, "eval_loss": 0.1516532599925995, "eval_runtime": 4308.7002, "eval_samples_per_second": 17.545, "eval_steps_per_second": 1.462, "eval_wer": 0.05633991332321027, "step": 194500 }, { "epoch": 16.11, "learning_rate": 1.1998129675810476e-05, "loss": 0.0362, "step": 194600 }, { "epoch": 16.12, "learning_rate": 1.199397339983375e-05, "loss": 0.0403, "step": 194700 }, { "epoch": 16.13, "learning_rate": 1.1989817123857027e-05, "loss": 0.0352, "step": 194800 }, { "epoch": 16.13, "learning_rate": 1.19856608478803e-05, "loss": 0.0424, "step": 194900 }, { "epoch": 16.14, "learning_rate": 1.1981504571903574e-05, "loss": 0.0363, "step": 195000 }, { "epoch": 16.14, "eval_loss": 0.137625053524971, "eval_runtime": 4306.0027, "eval_samples_per_second": 17.556, "eval_steps_per_second": 1.463, "eval_wer": 0.05427376265574976, "step": 195000 }, { "epoch": 16.15, "learning_rate": 1.197734829592685e-05, "loss": 0.0419, "step": 195100 }, { "epoch": 16.16, "learning_rate": 1.1973192019950127e-05, "loss": 0.0337, "step": 195200 }, { "epoch": 16.17, "learning_rate": 1.19690357439734e-05, "loss": 0.0352, "step": 195300 }, { "epoch": 16.18, "learning_rate": 1.1964879467996675e-05, "loss": 0.0355, "step": 195400 }, { "epoch": 16.18, "learning_rate": 1.1960723192019952e-05, "loss": 0.0401, "step": 195500 }, { "epoch": 16.18, "eval_loss": 0.14431744813919067, "eval_runtime": 4337.0911, "eval_samples_per_second": 17.43, "eval_steps_per_second": 1.453, "eval_wer": 0.05654145147469004, "step": 195500 }, { "epoch": 16.19, "learning_rate": 1.1956566916043226e-05, "loss": 0.0411, "step": 195600 }, { "epoch": 16.2, "learning_rate": 1.1952410640066501e-05, "loss": 0.0369, "step": 195700 }, { "epoch": 16.21, "learning_rate": 1.1948254364089777e-05, "loss": 0.045, "step": 195800 }, { "epoch": 16.22, "learning_rate": 1.1944098088113052e-05, "loss": 0.0376, "step": 195900 }, { "epoch": 16.23, "learning_rate": 1.1939941812136326e-05, "loss": 0.0423, "step": 196000 }, { "epoch": 16.23, "eval_loss": 0.14515773952007294, "eval_runtime": 4280.5155, "eval_samples_per_second": 17.66, "eval_steps_per_second": 1.472, "eval_wer": 0.05532453026995343, "step": 196000 }, { "epoch": 16.23, "learning_rate": 1.1935785536159602e-05, "loss": 0.0362, "step": 196100 }, { "epoch": 16.24, "learning_rate": 1.1931629260182877e-05, "loss": 0.042, "step": 196200 }, { "epoch": 16.25, "learning_rate": 1.1927472984206153e-05, "loss": 0.0395, "step": 196300 }, { "epoch": 16.26, "learning_rate": 1.1923316708229426e-05, "loss": 0.0412, "step": 196400 }, { "epoch": 16.27, "learning_rate": 1.1919160432252704e-05, "loss": 0.041, "step": 196500 }, { "epoch": 16.27, "eval_loss": 0.1451353132724762, "eval_runtime": 4277.9203, "eval_samples_per_second": 17.671, "eval_steps_per_second": 1.473, "eval_wer": 0.05523837655634376, "step": 196500 }, { "epoch": 16.27, "learning_rate": 1.1915004156275977e-05, "loss": 0.0434, "step": 196600 }, { "epoch": 16.28, "learning_rate": 1.1910847880299253e-05, "loss": 0.0342, "step": 196700 }, { "epoch": 16.29, "learning_rate": 1.1906691604322528e-05, "loss": 0.0405, "step": 196800 }, { "epoch": 16.3, "learning_rate": 1.1902535328345804e-05, "loss": 0.0364, "step": 196900 }, { "epoch": 16.31, "learning_rate": 1.1898420615128846e-05, "loss": 0.0409, "step": 197000 }, { "epoch": 16.31, "eval_loss": 0.15655146539211273, "eval_runtime": 4249.4062, "eval_samples_per_second": 17.79, "eval_steps_per_second": 1.483, "eval_wer": 0.05763529594569854, "step": 197000 }, { "epoch": 16.32, "learning_rate": 1.1894264339152121e-05, "loss": 0.0346, "step": 197100 }, { "epoch": 16.32, "learning_rate": 1.1890108063175397e-05, "loss": 0.0374, "step": 197200 }, { "epoch": 16.33, "learning_rate": 1.188595178719867e-05, "loss": 0.0394, "step": 197300 }, { "epoch": 16.34, "learning_rate": 1.1881795511221944e-05, "loss": 0.0405, "step": 197400 }, { "epoch": 16.35, "learning_rate": 1.1877639235245222e-05, "loss": 0.0401, "step": 197500 }, { "epoch": 16.35, "eval_loss": 0.15672236680984497, "eval_runtime": 4274.9371, "eval_samples_per_second": 17.683, "eval_steps_per_second": 1.474, "eval_wer": 0.057418373202502766, "step": 197500 }, { "epoch": 16.36, "learning_rate": 1.1873482959268497e-05, "loss": 0.038, "step": 197600 }, { "epoch": 16.37, "learning_rate": 1.1869326683291771e-05, "loss": 0.0397, "step": 197700 }, { "epoch": 16.37, "learning_rate": 1.1865170407315048e-05, "loss": 0.037, "step": 197800 }, { "epoch": 16.38, "learning_rate": 1.1861014131338322e-05, "loss": 0.0389, "step": 197900 }, { "epoch": 16.39, "learning_rate": 1.1856857855361596e-05, "loss": 0.047, "step": 198000 }, { "epoch": 16.39, "eval_loss": 0.13389697670936584, "eval_runtime": 4243.1095, "eval_samples_per_second": 17.816, "eval_steps_per_second": 1.485, "eval_wer": 0.054807607988295405, "step": 198000 }, { "epoch": 16.4, "learning_rate": 1.1852701579384873e-05, "loss": 0.0473, "step": 198100 }, { "epoch": 16.41, "learning_rate": 1.1848545303408147e-05, "loss": 0.0344, "step": 198200 }, { "epoch": 16.42, "learning_rate": 1.1844389027431422e-05, "loss": 0.0369, "step": 198300 }, { "epoch": 16.42, "learning_rate": 1.1840232751454696e-05, "loss": 0.0385, "step": 198400 }, { "epoch": 16.43, "learning_rate": 1.1836076475477973e-05, "loss": 0.0399, "step": 198500 }, { "epoch": 16.43, "eval_loss": 0.14400769770145416, "eval_runtime": 4242.7886, "eval_samples_per_second": 17.817, "eval_steps_per_second": 1.485, "eval_wer": 0.05427837803326457, "step": 198500 }, { "epoch": 16.44, "learning_rate": 1.1831920199501247e-05, "loss": 0.0365, "step": 198600 }, { "epoch": 16.45, "learning_rate": 1.1827763923524523e-05, "loss": 0.0352, "step": 198700 }, { "epoch": 16.46, "learning_rate": 1.1823607647547798e-05, "loss": 0.0424, "step": 198800 }, { "epoch": 16.47, "learning_rate": 1.1819451371571074e-05, "loss": 0.0416, "step": 198900 }, { "epoch": 16.47, "learning_rate": 1.1815295095594348e-05, "loss": 0.041, "step": 199000 }, { "epoch": 16.47, "eval_loss": 0.15036717057228088, "eval_runtime": 4262.9289, "eval_samples_per_second": 17.733, "eval_steps_per_second": 1.478, "eval_wer": 0.0575737575788345, "step": 199000 }, { "epoch": 16.48, "learning_rate": 1.1811180382377391e-05, "loss": 0.0421, "step": 199100 }, { "epoch": 16.49, "learning_rate": 1.1807024106400667e-05, "loss": 0.0368, "step": 199200 }, { "epoch": 16.5, "learning_rate": 1.180286783042394e-05, "loss": 0.0381, "step": 199300 }, { "epoch": 16.51, "learning_rate": 1.1798711554447216e-05, "loss": 0.0358, "step": 199400 }, { "epoch": 16.51, "learning_rate": 1.1794555278470492e-05, "loss": 0.0394, "step": 199500 }, { "epoch": 16.51, "eval_loss": 0.15342549979686737, "eval_runtime": 4264.9374, "eval_samples_per_second": 17.725, "eval_steps_per_second": 1.477, "eval_wer": 0.05676606651374382, "step": 199500 }, { "epoch": 16.52, "learning_rate": 1.1790399002493767e-05, "loss": 0.0328, "step": 199600 }, { "epoch": 16.53, "learning_rate": 1.1786242726517041e-05, "loss": 0.0401, "step": 199700 }, { "epoch": 16.54, "learning_rate": 1.1782086450540318e-05, "loss": 0.0407, "step": 199800 }, { "epoch": 16.55, "learning_rate": 1.1777930174563592e-05, "loss": 0.0466, "step": 199900 }, { "epoch": 16.56, "learning_rate": 1.1773773898586867e-05, "loss": 0.0386, "step": 200000 }, { "epoch": 16.56, "eval_loss": 0.14965596795082092, "eval_runtime": 4267.3685, "eval_samples_per_second": 17.715, "eval_steps_per_second": 1.476, "eval_wer": 0.05829375647114389, "step": 200000 }, { "epoch": 16.56, "learning_rate": 1.1769617622610143e-05, "loss": 0.0426, "step": 200100 }, { "epoch": 16.57, "learning_rate": 1.1765461346633418e-05, "loss": 0.0373, "step": 200200 }, { "epoch": 16.58, "learning_rate": 1.1761305070656692e-05, "loss": 0.0365, "step": 200300 }, { "epoch": 16.59, "learning_rate": 1.1757148794679966e-05, "loss": 0.0361, "step": 200400 }, { "epoch": 16.6, "learning_rate": 1.1752992518703243e-05, "loss": 0.0313, "step": 200500 }, { "epoch": 16.6, "eval_loss": 0.16236110031604767, "eval_runtime": 4260.8179, "eval_samples_per_second": 17.742, "eval_steps_per_second": 1.479, "eval_wer": 0.058055295299545694, "step": 200500 }, { "epoch": 16.61, "learning_rate": 1.1748836242726517e-05, "loss": 0.0388, "step": 200600 }, { "epoch": 16.61, "learning_rate": 1.1744679966749793e-05, "loss": 0.0385, "step": 200700 }, { "epoch": 16.62, "learning_rate": 1.174052369077307e-05, "loss": 0.0381, "step": 200800 }, { "epoch": 16.63, "learning_rate": 1.1736367414796344e-05, "loss": 0.0363, "step": 200900 }, { "epoch": 16.64, "learning_rate": 1.1732211138819618e-05, "loss": 0.0418, "step": 201000 }, { "epoch": 16.64, "eval_loss": 0.14727430045604706, "eval_runtime": 4245.4404, "eval_samples_per_second": 17.806, "eval_steps_per_second": 1.484, "eval_wer": 0.055272222658118986, "step": 201000 }, { "epoch": 16.65, "learning_rate": 1.1728054862842895e-05, "loss": 0.0401, "step": 201100 }, { "epoch": 16.66, "learning_rate": 1.1723898586866169e-05, "loss": 0.0343, "step": 201200 }, { "epoch": 16.66, "learning_rate": 1.1719742310889444e-05, "loss": 0.0394, "step": 201300 }, { "epoch": 16.67, "learning_rate": 1.1715627597672486e-05, "loss": 0.0355, "step": 201400 }, { "epoch": 16.68, "learning_rate": 1.1711471321695761e-05, "loss": 0.0401, "step": 201500 }, { "epoch": 16.68, "eval_loss": 0.1419604867696762, "eval_runtime": 4283.1988, "eval_samples_per_second": 17.649, "eval_steps_per_second": 1.471, "eval_wer": 0.05541683782024951, "step": 201500 }, { "epoch": 16.69, "learning_rate": 1.1707315045719037e-05, "loss": 0.0553, "step": 201600 }, { "epoch": 16.7, "learning_rate": 1.170315876974231e-05, "loss": 0.0423, "step": 201700 }, { "epoch": 16.71, "learning_rate": 1.1699002493765588e-05, "loss": 0.037, "step": 201800 }, { "epoch": 16.71, "learning_rate": 1.1694846217788862e-05, "loss": 0.0411, "step": 201900 }, { "epoch": 16.72, "learning_rate": 1.1690689941812137e-05, "loss": 0.0429, "step": 202000 }, { "epoch": 16.72, "eval_loss": 0.1533859819173813, "eval_runtime": 4255.7651, "eval_samples_per_second": 17.763, "eval_steps_per_second": 1.48, "eval_wer": 0.055922990887706324, "step": 202000 }, { "epoch": 16.73, "learning_rate": 1.1686533665835413e-05, "loss": 0.0376, "step": 202100 }, { "epoch": 16.74, "learning_rate": 1.1682377389858688e-05, "loss": 0.036, "step": 202200 }, { "epoch": 16.75, "learning_rate": 1.1678221113881962e-05, "loss": 0.0351, "step": 202300 }, { "epoch": 16.75, "learning_rate": 1.1674064837905238e-05, "loss": 0.0394, "step": 202400 }, { "epoch": 16.76, "learning_rate": 1.1669908561928513e-05, "loss": 0.0424, "step": 202500 }, { "epoch": 16.76, "eval_loss": 0.14160937070846558, "eval_runtime": 4245.9554, "eval_samples_per_second": 17.804, "eval_steps_per_second": 1.484, "eval_wer": 0.05458299294924162, "step": 202500 }, { "epoch": 16.77, "learning_rate": 1.1665752285951789e-05, "loss": 0.0309, "step": 202600 }, { "epoch": 16.78, "learning_rate": 1.1661596009975063e-05, "loss": 0.0451, "step": 202700 }, { "epoch": 16.79, "learning_rate": 1.165743973399834e-05, "loss": 0.0379, "step": 202800 }, { "epoch": 16.8, "learning_rate": 1.1653283458021614e-05, "loss": 0.0413, "step": 202900 }, { "epoch": 16.8, "learning_rate": 1.1649127182044887e-05, "loss": 0.0487, "step": 203000 }, { "epoch": 16.8, "eval_loss": 0.1487305462360382, "eval_runtime": 4254.5989, "eval_samples_per_second": 17.768, "eval_steps_per_second": 1.481, "eval_wer": 0.05680452799303386, "step": 203000 }, { "epoch": 16.81, "learning_rate": 1.1644970906068165e-05, "loss": 0.0354, "step": 203100 }, { "epoch": 16.82, "learning_rate": 1.164081463009144e-05, "loss": 0.0393, "step": 203200 }, { "epoch": 16.83, "learning_rate": 1.1636658354114714e-05, "loss": 0.0359, "step": 203300 }, { "epoch": 16.84, "learning_rate": 1.1632502078137988e-05, "loss": 0.04, "step": 203400 }, { "epoch": 16.85, "learning_rate": 1.1628345802161265e-05, "loss": 0.0434, "step": 203500 }, { "epoch": 16.85, "eval_loss": 0.15244609117507935, "eval_runtime": 4268.4963, "eval_samples_per_second": 17.71, "eval_steps_per_second": 1.476, "eval_wer": 0.05635222099658308, "step": 203500 }, { "epoch": 16.85, "learning_rate": 1.1624231088944307e-05, "loss": 0.0363, "step": 203600 }, { "epoch": 16.86, "learning_rate": 1.162007481296758e-05, "loss": 0.0396, "step": 203700 }, { "epoch": 16.87, "learning_rate": 1.1615918536990858e-05, "loss": 0.0373, "step": 203800 }, { "epoch": 16.88, "learning_rate": 1.1611762261014132e-05, "loss": 0.038, "step": 203900 }, { "epoch": 16.89, "learning_rate": 1.1607605985037407e-05, "loss": 0.0388, "step": 204000 }, { "epoch": 16.89, "eval_loss": 0.1624448299407959, "eval_runtime": 4274.6622, "eval_samples_per_second": 17.684, "eval_steps_per_second": 1.474, "eval_wer": 0.05901529382262489, "step": 204000 }, { "epoch": 16.9, "learning_rate": 1.1603449709060683e-05, "loss": 0.0363, "step": 204100 }, { "epoch": 16.9, "learning_rate": 1.1599293433083958e-05, "loss": 0.0381, "step": 204200 }, { "epoch": 16.91, "learning_rate": 1.1595137157107232e-05, "loss": 0.0354, "step": 204300 }, { "epoch": 16.92, "learning_rate": 1.159098088113051e-05, "loss": 0.0448, "step": 204400 }, { "epoch": 16.93, "learning_rate": 1.1586824605153783e-05, "loss": 0.0393, "step": 204500 }, { "epoch": 16.93, "eval_loss": 0.15931500494480133, "eval_runtime": 4258.7268, "eval_samples_per_second": 17.751, "eval_steps_per_second": 1.479, "eval_wer": 0.055779914184747405, "step": 204500 }, { "epoch": 16.94, "learning_rate": 1.1582668329177059e-05, "loss": 0.0373, "step": 204600 }, { "epoch": 16.95, "learning_rate": 1.1578512053200332e-05, "loss": 0.0383, "step": 204700 }, { "epoch": 16.95, "learning_rate": 1.157435577722361e-05, "loss": 0.0381, "step": 204800 }, { "epoch": 16.96, "learning_rate": 1.1570199501246883e-05, "loss": 0.0393, "step": 204900 }, { "epoch": 16.97, "learning_rate": 1.1566043225270159e-05, "loss": 0.0457, "step": 205000 }, { "epoch": 16.97, "eval_loss": 0.15160848200321198, "eval_runtime": 4303.2587, "eval_samples_per_second": 17.567, "eval_steps_per_second": 1.464, "eval_wer": 0.057441450090076784, "step": 205000 }, { "epoch": 16.98, "learning_rate": 1.1561886949293434e-05, "loss": 0.0323, "step": 205100 }, { "epoch": 16.99, "learning_rate": 1.155773067331671e-05, "loss": 0.0389, "step": 205200 }, { "epoch": 17.0, "learning_rate": 1.1553574397339984e-05, "loss": 0.0361, "step": 205300 }, { "epoch": 17.0, "learning_rate": 1.1549418121363261e-05, "loss": 0.0343, "step": 205400 }, { "epoch": 17.01, "learning_rate": 1.1545261845386535e-05, "loss": 0.0413, "step": 205500 }, { "epoch": 17.01, "eval_loss": 0.14971312880516052, "eval_runtime": 4476.0625, "eval_samples_per_second": 16.889, "eval_steps_per_second": 1.407, "eval_wer": 0.055813760286522636, "step": 205500 }, { "epoch": 17.02, "learning_rate": 1.154110556940981e-05, "loss": 0.0355, "step": 205600 }, { "epoch": 17.03, "learning_rate": 1.1536949293433084e-05, "loss": 0.0357, "step": 205700 }, { "epoch": 17.04, "learning_rate": 1.1532793017456361e-05, "loss": 0.045, "step": 205800 }, { "epoch": 17.04, "learning_rate": 1.1528636741479635e-05, "loss": 0.0383, "step": 205900 }, { "epoch": 17.05, "learning_rate": 1.1524480465502909e-05, "loss": 0.0367, "step": 206000 }, { "epoch": 17.05, "eval_loss": 0.15126442909240723, "eval_runtime": 4389.5757, "eval_samples_per_second": 17.221, "eval_steps_per_second": 1.435, "eval_wer": 0.05487683865101746, "step": 206000 }, { "epoch": 17.06, "learning_rate": 1.1520324189526186e-05, "loss": 0.0397, "step": 206100 }, { "epoch": 17.07, "learning_rate": 1.151616791354946e-05, "loss": 0.0382, "step": 206200 }, { "epoch": 17.08, "learning_rate": 1.1512011637572736e-05, "loss": 0.0385, "step": 206300 }, { "epoch": 17.09, "learning_rate": 1.1507855361596013e-05, "loss": 0.0406, "step": 206400 }, { "epoch": 17.09, "learning_rate": 1.1503699085619287e-05, "loss": 0.0402, "step": 206500 }, { "epoch": 17.09, "eval_loss": 0.1535288691520691, "eval_runtime": 4391.8366, "eval_samples_per_second": 17.213, "eval_steps_per_second": 1.434, "eval_wer": 0.056427605495991545, "step": 206500 }, { "epoch": 17.1, "learning_rate": 1.149954280964256e-05, "loss": 0.0459, "step": 206600 }, { "epoch": 17.11, "learning_rate": 1.1495386533665836e-05, "loss": 0.0337, "step": 206700 }, { "epoch": 17.12, "learning_rate": 1.1491230257689111e-05, "loss": 0.0383, "step": 206800 }, { "epoch": 17.13, "learning_rate": 1.1487073981712387e-05, "loss": 0.0392, "step": 206900 }, { "epoch": 17.14, "learning_rate": 1.148291770573566e-05, "loss": 0.0349, "step": 207000 }, { "epoch": 17.14, "eval_loss": 0.1539006382226944, "eval_runtime": 4291.2918, "eval_samples_per_second": 17.616, "eval_steps_per_second": 1.468, "eval_wer": 0.0540814552592996, "step": 207000 }, { "epoch": 17.14, "learning_rate": 1.1478761429758938e-05, "loss": 0.0381, "step": 207100 }, { "epoch": 17.15, "learning_rate": 1.1474605153782212e-05, "loss": 0.038, "step": 207200 }, { "epoch": 17.16, "learning_rate": 1.1470448877805487e-05, "loss": 0.0325, "step": 207300 }, { "epoch": 17.17, "learning_rate": 1.1466292601828763e-05, "loss": 0.0385, "step": 207400 }, { "epoch": 17.18, "learning_rate": 1.1462136325852038e-05, "loss": 0.0384, "step": 207500 }, { "epoch": 17.18, "eval_loss": 0.14295735955238342, "eval_runtime": 4266.6281, "eval_samples_per_second": 17.718, "eval_steps_per_second": 1.477, "eval_wer": 0.053361456366990204, "step": 207500 }, { "epoch": 17.19, "learning_rate": 1.1457980049875312e-05, "loss": 0.0374, "step": 207600 }, { "epoch": 17.19, "learning_rate": 1.1453823773898588e-05, "loss": 0.0355, "step": 207700 }, { "epoch": 17.2, "learning_rate": 1.1449667497921863e-05, "loss": 0.0395, "step": 207800 }, { "epoch": 17.21, "learning_rate": 1.1445511221945139e-05, "loss": 0.0377, "step": 207900 }, { "epoch": 17.22, "learning_rate": 1.1441354945968413e-05, "loss": 0.0399, "step": 208000 }, { "epoch": 17.22, "eval_loss": 0.15152186155319214, "eval_runtime": 4274.0119, "eval_samples_per_second": 17.687, "eval_steps_per_second": 1.474, "eval_wer": 0.05326299498000772, "step": 208000 }, { "epoch": 17.23, "learning_rate": 1.143719866999169e-05, "loss": 0.0383, "step": 208100 }, { "epoch": 17.24, "learning_rate": 1.1433042394014964e-05, "loss": 0.0351, "step": 208200 }, { "epoch": 17.24, "learning_rate": 1.1428886118038237e-05, "loss": 0.0329, "step": 208300 }, { "epoch": 17.25, "learning_rate": 1.1424729842061515e-05, "loss": 0.0357, "step": 208400 }, { "epoch": 17.26, "learning_rate": 1.142057356608479e-05, "loss": 0.0393, "step": 208500 }, { "epoch": 17.26, "eval_loss": 0.15290088951587677, "eval_runtime": 4278.3285, "eval_samples_per_second": 17.669, "eval_steps_per_second": 1.473, "eval_wer": 0.05381222490426938, "step": 208500 }, { "epoch": 17.27, "learning_rate": 1.1416417290108064e-05, "loss": 0.0328, "step": 208600 }, { "epoch": 17.28, "learning_rate": 1.1412261014131338e-05, "loss": 0.0325, "step": 208700 }, { "epoch": 17.28, "learning_rate": 1.1408104738154615e-05, "loss": 0.0376, "step": 208800 }, { "epoch": 17.29, "learning_rate": 1.1403948462177889e-05, "loss": 0.0302, "step": 208900 }, { "epoch": 17.3, "learning_rate": 1.1399792186201164e-05, "loss": 0.0344, "step": 209000 }, { "epoch": 17.3, "eval_loss": 0.14451298117637634, "eval_runtime": 4263.4528, "eval_samples_per_second": 17.731, "eval_steps_per_second": 1.478, "eval_wer": 0.05345222545811468, "step": 209000 }, { "epoch": 17.31, "learning_rate": 1.1395635910224442e-05, "loss": 0.0407, "step": 209100 }, { "epoch": 17.32, "learning_rate": 1.1391479634247715e-05, "loss": 0.0345, "step": 209200 }, { "epoch": 17.33, "learning_rate": 1.138732335827099e-05, "loss": 0.0358, "step": 209300 }, { "epoch": 17.33, "learning_rate": 1.1383167082294266e-05, "loss": 0.0418, "step": 209400 }, { "epoch": 17.34, "learning_rate": 1.1379052369077308e-05, "loss": 0.0394, "step": 209500 }, { "epoch": 17.34, "eval_loss": 0.147212952375412, "eval_runtime": 4263.9703, "eval_samples_per_second": 17.729, "eval_steps_per_second": 1.477, "eval_wer": 0.05417530126876728, "step": 209500 }, { "epoch": 17.35, "learning_rate": 1.1374896093100582e-05, "loss": 0.0353, "step": 209600 }, { "epoch": 17.36, "learning_rate": 1.1370739817123858e-05, "loss": 0.0337, "step": 209700 }, { "epoch": 17.37, "learning_rate": 1.1366583541147133e-05, "loss": 0.0377, "step": 209800 }, { "epoch": 17.38, "learning_rate": 1.1362427265170409e-05, "loss": 0.0345, "step": 209900 }, { "epoch": 17.38, "learning_rate": 1.1358270989193682e-05, "loss": 0.0496, "step": 210000 }, { "epoch": 17.38, "eval_loss": 0.16750968992710114, "eval_runtime": 4244.1762, "eval_samples_per_second": 17.811, "eval_steps_per_second": 1.484, "eval_wer": 0.058012218442740855, "step": 210000 }, { "epoch": 17.39, "learning_rate": 1.135411471321696e-05, "loss": 0.036, "step": 210100 }, { "epoch": 17.4, "learning_rate": 1.1349958437240233e-05, "loss": 0.0376, "step": 210200 }, { "epoch": 17.41, "learning_rate": 1.1345802161263509e-05, "loss": 0.0338, "step": 210300 }, { "epoch": 17.42, "learning_rate": 1.1341687448046552e-05, "loss": 0.0424, "step": 210400 }, { "epoch": 17.43, "learning_rate": 1.1337531172069826e-05, "loss": 0.0355, "step": 210500 }, { "epoch": 17.43, "eval_loss": 0.16486169397830963, "eval_runtime": 4260.9717, "eval_samples_per_second": 17.741, "eval_steps_per_second": 1.479, "eval_wer": 0.055086069098355235, "step": 210500 }, { "epoch": 17.43, "learning_rate": 1.1333374896093102e-05, "loss": 0.0386, "step": 210600 }, { "epoch": 17.44, "learning_rate": 1.1329218620116376e-05, "loss": 0.0313, "step": 210700 }, { "epoch": 17.45, "learning_rate": 1.1325062344139653e-05, "loss": 0.0344, "step": 210800 }, { "epoch": 17.46, "learning_rate": 1.1320906068162927e-05, "loss": 0.0379, "step": 210900 }, { "epoch": 17.47, "learning_rate": 1.13167497921862e-05, "loss": 0.0322, "step": 211000 }, { "epoch": 17.47, "eval_loss": 0.165803000330925, "eval_runtime": 4246.2025, "eval_samples_per_second": 17.803, "eval_steps_per_second": 1.484, "eval_wer": 0.0579414493208472, "step": 211000 }, { "epoch": 17.48, "learning_rate": 1.1312593516209478e-05, "loss": 0.0392, "step": 211100 }, { "epoch": 17.48, "learning_rate": 1.1308437240232753e-05, "loss": 0.0369, "step": 211200 }, { "epoch": 17.49, "learning_rate": 1.1304280964256027e-05, "loss": 0.037, "step": 211300 }, { "epoch": 17.5, "learning_rate": 1.1300124688279304e-05, "loss": 0.0397, "step": 211400 }, { "epoch": 17.51, "learning_rate": 1.1295968412302578e-05, "loss": 0.0358, "step": 211500 }, { "epoch": 17.51, "eval_loss": 0.15967398881912231, "eval_runtime": 4253.6894, "eval_samples_per_second": 17.772, "eval_steps_per_second": 1.481, "eval_wer": 0.05579376031729182, "step": 211500 }, { "epoch": 17.52, "learning_rate": 1.1291812136325852e-05, "loss": 0.0402, "step": 211600 }, { "epoch": 17.52, "learning_rate": 1.1287655860349127e-05, "loss": 0.039, "step": 211700 }, { "epoch": 17.53, "learning_rate": 1.1283499584372405e-05, "loss": 0.0339, "step": 211800 }, { "epoch": 17.54, "learning_rate": 1.1279343308395678e-05, "loss": 0.039, "step": 211900 }, { "epoch": 17.55, "learning_rate": 1.1275187032418952e-05, "loss": 0.0345, "step": 212000 }, { "epoch": 17.55, "eval_loss": 0.15865014493465424, "eval_runtime": 4260.1226, "eval_samples_per_second": 17.745, "eval_steps_per_second": 1.479, "eval_wer": 0.05472453119302893, "step": 212000 }, { "epoch": 17.56, "learning_rate": 1.127103075644223e-05, "loss": 0.038, "step": 212100 }, { "epoch": 17.57, "learning_rate": 1.1266874480465503e-05, "loss": 0.035, "step": 212200 }, { "epoch": 17.57, "learning_rate": 1.1262718204488779e-05, "loss": 0.0317, "step": 212300 }, { "epoch": 17.58, "learning_rate": 1.1258561928512054e-05, "loss": 0.0343, "step": 212400 }, { "epoch": 17.59, "learning_rate": 1.125440565253533e-05, "loss": 0.0387, "step": 212500 }, { "epoch": 17.59, "eval_loss": 0.1570260226726532, "eval_runtime": 4253.3728, "eval_samples_per_second": 17.773, "eval_steps_per_second": 1.481, "eval_wer": 0.054566069898354004, "step": 212500 }, { "epoch": 17.6, "learning_rate": 1.1250290939318372e-05, "loss": 0.035, "step": 212600 }, { "epoch": 17.61, "learning_rate": 1.1246134663341645e-05, "loss": 0.0398, "step": 212700 }, { "epoch": 17.62, "learning_rate": 1.1241978387364923e-05, "loss": 0.0367, "step": 212800 }, { "epoch": 17.62, "learning_rate": 1.1237822111388197e-05, "loss": 0.0384, "step": 212900 }, { "epoch": 17.63, "learning_rate": 1.1233665835411472e-05, "loss": 0.0369, "step": 213000 }, { "epoch": 17.63, "eval_loss": 0.15911221504211426, "eval_runtime": 4243.044, "eval_samples_per_second": 17.816, "eval_steps_per_second": 1.485, "eval_wer": 0.05460606983681564, "step": 213000 }, { "epoch": 17.64, "learning_rate": 1.1229509559434748e-05, "loss": 0.0488, "step": 213100 }, { "epoch": 17.65, "learning_rate": 1.1225353283458023e-05, "loss": 0.0388, "step": 213200 }, { "epoch": 17.66, "learning_rate": 1.1221197007481297e-05, "loss": 0.0329, "step": 213300 }, { "epoch": 17.67, "learning_rate": 1.1217040731504574e-05, "loss": 0.0396, "step": 213400 }, { "epoch": 17.67, "learning_rate": 1.1212884455527848e-05, "loss": 0.0397, "step": 213500 }, { "epoch": 17.67, "eval_loss": 0.1563698649406433, "eval_runtime": 4260.9852, "eval_samples_per_second": 17.741, "eval_steps_per_second": 1.479, "eval_wer": 0.05477068496817697, "step": 213500 }, { "epoch": 17.68, "learning_rate": 1.1208728179551123e-05, "loss": 0.038, "step": 213600 }, { "epoch": 17.69, "learning_rate": 1.1204571903574397e-05, "loss": 0.0312, "step": 213700 }, { "epoch": 17.7, "learning_rate": 1.1200415627597674e-05, "loss": 0.0375, "step": 213800 }, { "epoch": 17.71, "learning_rate": 1.1196259351620948e-05, "loss": 0.0306, "step": 213900 }, { "epoch": 17.72, "learning_rate": 1.1192103075644222e-05, "loss": 0.0369, "step": 214000 }, { "epoch": 17.72, "eval_loss": 0.15147362649440765, "eval_runtime": 4250.9969, "eval_samples_per_second": 17.783, "eval_steps_per_second": 1.482, "eval_wer": 0.05409991676935882, "step": 214000 }, { "epoch": 17.72, "learning_rate": 1.11879467996675e-05, "loss": 0.0359, "step": 214100 }, { "epoch": 17.73, "learning_rate": 1.1183790523690773e-05, "loss": 0.0362, "step": 214200 }, { "epoch": 17.74, "learning_rate": 1.1179634247714049e-05, "loss": 0.036, "step": 214300 }, { "epoch": 17.75, "learning_rate": 1.1175477971737326e-05, "loss": 0.0424, "step": 214400 }, { "epoch": 17.76, "learning_rate": 1.11713216957606e-05, "loss": 0.0392, "step": 214500 }, { "epoch": 17.76, "eval_loss": 0.15435676276683807, "eval_runtime": 4376.2349, "eval_samples_per_second": 17.274, "eval_steps_per_second": 1.44, "eval_wer": 0.053881455566991435, "step": 214500 }, { "epoch": 17.76, "learning_rate": 1.1167165419783874e-05, "loss": 0.0326, "step": 214600 }, { "epoch": 17.77, "learning_rate": 1.1163009143807149e-05, "loss": 0.0372, "step": 214700 }, { "epoch": 17.78, "learning_rate": 1.1158852867830425e-05, "loss": 0.0367, "step": 214800 }, { "epoch": 17.79, "learning_rate": 1.11546965918537e-05, "loss": 0.0369, "step": 214900 }, { "epoch": 17.8, "learning_rate": 1.1150540315876974e-05, "loss": 0.0345, "step": 215000 }, { "epoch": 17.8, "eval_loss": 0.15085247159004211, "eval_runtime": 4404.7512, "eval_samples_per_second": 17.162, "eval_steps_per_second": 1.43, "eval_wer": 0.05423991655397453, "step": 215000 }, { "epoch": 17.81, "learning_rate": 1.1146384039900251e-05, "loss": 0.0369, "step": 215100 }, { "epoch": 17.81, "learning_rate": 1.1142227763923525e-05, "loss": 0.0449, "step": 215200 }, { "epoch": 17.82, "learning_rate": 1.11380714879468e-05, "loss": 0.0348, "step": 215300 }, { "epoch": 17.83, "learning_rate": 1.1133915211970076e-05, "loss": 0.0328, "step": 215400 }, { "epoch": 17.84, "learning_rate": 1.1129758935993351e-05, "loss": 0.0397, "step": 215500 }, { "epoch": 17.84, "eval_loss": 0.13771255314350128, "eval_runtime": 4437.8405, "eval_samples_per_second": 17.034, "eval_steps_per_second": 1.42, "eval_wer": 0.05389837861787905, "step": 215500 }, { "epoch": 17.85, "learning_rate": 1.1125602660016625e-05, "loss": 0.0353, "step": 215600 }, { "epoch": 17.86, "learning_rate": 1.11214463840399e-05, "loss": 0.0367, "step": 215700 }, { "epoch": 17.86, "learning_rate": 1.1117331670822944e-05, "loss": 0.0377, "step": 215800 }, { "epoch": 17.87, "learning_rate": 1.1113175394846218e-05, "loss": 0.0395, "step": 215900 }, { "epoch": 17.88, "learning_rate": 1.1109019118869494e-05, "loss": 0.0385, "step": 216000 }, { "epoch": 17.88, "eval_loss": 0.15229590237140656, "eval_runtime": 4381.5572, "eval_samples_per_second": 17.253, "eval_steps_per_second": 1.438, "eval_wer": 0.05390760937290866, "step": 216000 }, { "epoch": 17.89, "learning_rate": 1.110486284289277e-05, "loss": 0.0341, "step": 216100 }, { "epoch": 17.9, "learning_rate": 1.1100706566916045e-05, "loss": 0.0402, "step": 216200 }, { "epoch": 17.91, "learning_rate": 1.1096550290939319e-05, "loss": 0.042, "step": 216300 }, { "epoch": 17.91, "learning_rate": 1.1092394014962596e-05, "loss": 0.0354, "step": 216400 }, { "epoch": 17.92, "learning_rate": 1.108823773898587e-05, "loss": 0.0374, "step": 216500 }, { "epoch": 17.92, "eval_loss": 0.15816286206245422, "eval_runtime": 4460.5829, "eval_samples_per_second": 16.947, "eval_steps_per_second": 1.412, "eval_wer": 0.054801454151608996, "step": 216500 }, { "epoch": 17.93, "learning_rate": 1.1084081463009143e-05, "loss": 0.0313, "step": 216600 }, { "epoch": 17.94, "learning_rate": 1.107992518703242e-05, "loss": 0.0347, "step": 216700 }, { "epoch": 17.95, "learning_rate": 1.1075768911055696e-05, "loss": 0.0333, "step": 216800 }, { "epoch": 17.96, "learning_rate": 1.107161263507897e-05, "loss": 0.0371, "step": 216900 }, { "epoch": 17.96, "learning_rate": 1.1067456359102244e-05, "loss": 0.0415, "step": 217000 }, { "epoch": 17.96, "eval_loss": 0.15913133323192596, "eval_runtime": 4518.8945, "eval_samples_per_second": 16.729, "eval_steps_per_second": 1.394, "eval_wer": 0.05471222351965612, "step": 217000 }, { "epoch": 17.97, "learning_rate": 1.1063300083125521e-05, "loss": 0.0381, "step": 217100 }, { "epoch": 17.98, "learning_rate": 1.1059143807148795e-05, "loss": 0.0397, "step": 217200 }, { "epoch": 17.99, "learning_rate": 1.105498753117207e-05, "loss": 0.0359, "step": 217300 }, { "epoch": 18.0, "learning_rate": 1.1050831255195348e-05, "loss": 0.0442, "step": 217400 }, { "epoch": 18.0, "learning_rate": 1.1046674979218621e-05, "loss": 0.0282, "step": 217500 }, { "epoch": 18.0, "eval_loss": 0.15394429862499237, "eval_runtime": 4529.1056, "eval_samples_per_second": 16.691, "eval_steps_per_second": 1.391, "eval_wer": 0.05349837923326272, "step": 217500 }, { "epoch": 18.01, "learning_rate": 1.1042518703241895e-05, "loss": 0.0348, "step": 217600 }, { "epoch": 18.02, "learning_rate": 1.1038362427265172e-05, "loss": 0.0351, "step": 217700 }, { "epoch": 18.03, "learning_rate": 1.1034206151288446e-05, "loss": 0.0335, "step": 217800 }, { "epoch": 18.04, "learning_rate": 1.1030049875311722e-05, "loss": 0.0346, "step": 217900 }, { "epoch": 18.05, "learning_rate": 1.1025893599334996e-05, "loss": 0.0461, "step": 218000 }, { "epoch": 18.05, "eval_loss": 0.1549045890569687, "eval_runtime": 4441.4918, "eval_samples_per_second": 17.02, "eval_steps_per_second": 1.418, "eval_wer": 0.0551660689752785, "step": 218000 }, { "epoch": 18.05, "learning_rate": 1.1021737323358273e-05, "loss": 0.0315, "step": 218100 }, { "epoch": 18.06, "learning_rate": 1.1017581047381547e-05, "loss": 0.0386, "step": 218200 }, { "epoch": 18.07, "learning_rate": 1.1013424771404822e-05, "loss": 0.036, "step": 218300 }, { "epoch": 18.08, "learning_rate": 1.1009268495428098e-05, "loss": 0.0311, "step": 218400 }, { "epoch": 18.09, "learning_rate": 1.1005112219451373e-05, "loss": 0.0312, "step": 218500 }, { "epoch": 18.09, "eval_loss": 0.1631767600774765, "eval_runtime": 4483.5874, "eval_samples_per_second": 16.86, "eval_steps_per_second": 1.405, "eval_wer": 0.0554060686060483, "step": 218500 }, { "epoch": 18.1, "learning_rate": 1.1000955943474647e-05, "loss": 0.0405, "step": 218600 }, { "epoch": 18.1, "learning_rate": 1.0996799667497924e-05, "loss": 0.0377, "step": 218700 }, { "epoch": 18.11, "learning_rate": 1.0992643391521198e-05, "loss": 0.0379, "step": 218800 }, { "epoch": 18.12, "learning_rate": 1.0988487115544473e-05, "loss": 0.0307, "step": 218900 }, { "epoch": 18.13, "learning_rate": 1.0984330839567747e-05, "loss": 0.0317, "step": 219000 }, { "epoch": 18.13, "eval_loss": 0.15618892014026642, "eval_runtime": 4413.6853, "eval_samples_per_second": 17.127, "eval_steps_per_second": 1.427, "eval_wer": 0.054815300284153406, "step": 219000 }, { "epoch": 18.14, "learning_rate": 1.0980174563591025e-05, "loss": 0.0418, "step": 219100 }, { "epoch": 18.15, "learning_rate": 1.0976018287614298e-05, "loss": 0.0394, "step": 219200 }, { "epoch": 18.15, "learning_rate": 1.0971862011637572e-05, "loss": 0.0335, "step": 219300 }, { "epoch": 18.16, "learning_rate": 1.096770573566085e-05, "loss": 0.046, "step": 219400 }, { "epoch": 18.17, "learning_rate": 1.0963549459684125e-05, "loss": 0.0354, "step": 219500 }, { "epoch": 18.17, "eval_loss": 0.1510472595691681, "eval_runtime": 4424.4239, "eval_samples_per_second": 17.086, "eval_steps_per_second": 1.424, "eval_wer": 0.05384299408770141, "step": 219500 }, { "epoch": 18.18, "learning_rate": 1.0959393183707399e-05, "loss": 0.042, "step": 219600 }, { "epoch": 18.19, "learning_rate": 1.0955236907730676e-05, "loss": 0.0365, "step": 219700 }, { "epoch": 18.2, "learning_rate": 1.095108063175395e-05, "loss": 0.0364, "step": 219800 }, { "epoch": 18.2, "learning_rate": 1.0946924355777224e-05, "loss": 0.0374, "step": 219900 }, { "epoch": 18.21, "learning_rate": 1.0942768079800499e-05, "loss": 0.0396, "step": 220000 }, { "epoch": 18.21, "eval_loss": 0.16656936705112457, "eval_runtime": 4504.532, "eval_samples_per_second": 16.782, "eval_steps_per_second": 1.399, "eval_wer": 0.056027606111375215, "step": 220000 }, { "epoch": 18.22, "learning_rate": 1.0938611803823775e-05, "loss": 0.0342, "step": 220100 }, { "epoch": 18.23, "learning_rate": 1.093445552784705e-05, "loss": 0.0389, "step": 220200 }, { "epoch": 18.24, "learning_rate": 1.0930299251870324e-05, "loss": 0.0373, "step": 220300 }, { "epoch": 18.25, "learning_rate": 1.0926142975893601e-05, "loss": 0.0294, "step": 220400 }, { "epoch": 18.25, "learning_rate": 1.0921986699916875e-05, "loss": 0.0451, "step": 220500 }, { "epoch": 18.25, "eval_loss": 0.1614394634962082, "eval_runtime": 4498.975, "eval_samples_per_second": 16.803, "eval_steps_per_second": 1.4, "eval_wer": 0.05714452746995774, "step": 220500 }, { "epoch": 18.26, "learning_rate": 1.091783042394015e-05, "loss": 0.0348, "step": 220600 }, { "epoch": 18.27, "learning_rate": 1.0913674147963426e-05, "loss": 0.0362, "step": 220700 }, { "epoch": 18.28, "learning_rate": 1.0909517871986702e-05, "loss": 0.0369, "step": 220800 }, { "epoch": 18.29, "learning_rate": 1.0905361596009975e-05, "loss": 0.0321, "step": 220900 }, { "epoch": 18.29, "learning_rate": 1.0901205320033251e-05, "loss": 0.0302, "step": 221000 }, { "epoch": 18.29, "eval_loss": 0.15397481620311737, "eval_runtime": 4457.2496, "eval_samples_per_second": 16.96, "eval_steps_per_second": 1.413, "eval_wer": 0.05402760918829356, "step": 221000 }, { "epoch": 18.3, "learning_rate": 1.0897049044056526e-05, "loss": 0.0337, "step": 221100 }, { "epoch": 18.31, "learning_rate": 1.0892892768079802e-05, "loss": 0.0333, "step": 221200 }, { "epoch": 18.32, "learning_rate": 1.0888778054862844e-05, "loss": 0.0381, "step": 221300 }, { "epoch": 18.33, "learning_rate": 1.088462177888612e-05, "loss": 0.0349, "step": 221400 }, { "epoch": 18.34, "learning_rate": 1.0880465502909395e-05, "loss": 0.0372, "step": 221500 }, { "epoch": 18.34, "eval_loss": 0.15029260516166687, "eval_runtime": 4513.3121, "eval_samples_per_second": 16.749, "eval_steps_per_second": 1.396, "eval_wer": 0.05402145535160715, "step": 221500 }, { "epoch": 18.34, "learning_rate": 1.0876309226932669e-05, "loss": 0.0339, "step": 221600 }, { "epoch": 18.35, "learning_rate": 1.0872152950955946e-05, "loss": 0.0351, "step": 221700 }, { "epoch": 18.36, "learning_rate": 1.086799667497922e-05, "loss": 0.0379, "step": 221800 }, { "epoch": 18.37, "learning_rate": 1.0863840399002495e-05, "loss": 0.0367, "step": 221900 }, { "epoch": 18.38, "learning_rate": 1.0859684123025769e-05, "loss": 0.0314, "step": 222000 }, { "epoch": 18.38, "eval_loss": 0.15877769887447357, "eval_runtime": 4450.723, "eval_samples_per_second": 16.985, "eval_steps_per_second": 1.416, "eval_wer": 0.056435297791849554, "step": 222000 }, { "epoch": 18.39, "learning_rate": 1.0855527847049046e-05, "loss": 0.0396, "step": 222100 }, { "epoch": 18.39, "learning_rate": 1.085137157107232e-05, "loss": 0.0375, "step": 222200 }, { "epoch": 18.4, "learning_rate": 1.0847215295095594e-05, "loss": 0.0304, "step": 222300 }, { "epoch": 18.41, "learning_rate": 1.0843059019118871e-05, "loss": 0.0349, "step": 222400 }, { "epoch": 18.42, "learning_rate": 1.0838902743142145e-05, "loss": 0.0318, "step": 222500 }, { "epoch": 18.42, "eval_loss": 0.16047616302967072, "eval_runtime": 4444.0043, "eval_samples_per_second": 17.011, "eval_steps_per_second": 1.418, "eval_wer": 0.05562452980841568, "step": 222500 }, { "epoch": 18.43, "learning_rate": 1.083474646716542e-05, "loss": 0.0316, "step": 222600 }, { "epoch": 18.44, "learning_rate": 1.0830590191188698e-05, "loss": 0.0337, "step": 222700 }, { "epoch": 18.44, "learning_rate": 1.0826433915211971e-05, "loss": 0.0405, "step": 222800 }, { "epoch": 18.45, "learning_rate": 1.0822277639235245e-05, "loss": 0.0389, "step": 222900 }, { "epoch": 18.46, "learning_rate": 1.081812136325852e-05, "loss": 0.0421, "step": 223000 }, { "epoch": 18.46, "eval_loss": 0.16630135476589203, "eval_runtime": 4475.5739, "eval_samples_per_second": 16.891, "eval_steps_per_second": 1.408, "eval_wer": 0.05677375880960183, "step": 223000 }, { "epoch": 18.47, "learning_rate": 1.0813965087281796e-05, "loss": 0.0316, "step": 223100 }, { "epoch": 18.48, "learning_rate": 1.0809808811305072e-05, "loss": 0.0369, "step": 223200 }, { "epoch": 18.49, "learning_rate": 1.0805652535328346e-05, "loss": 0.0396, "step": 223300 }, { "epoch": 18.49, "learning_rate": 1.0801496259351623e-05, "loss": 0.0351, "step": 223400 }, { "epoch": 18.5, "learning_rate": 1.0797339983374897e-05, "loss": 0.0391, "step": 223500 }, { "epoch": 18.5, "eval_loss": 0.16064849495887756, "eval_runtime": 4529.6353, "eval_samples_per_second": 16.689, "eval_steps_per_second": 1.391, "eval_wer": 0.05699683538948402, "step": 223500 }, { "epoch": 18.51, "learning_rate": 1.0793183707398172e-05, "loss": 0.0399, "step": 223600 }, { "epoch": 18.52, "learning_rate": 1.0789068994181216e-05, "loss": 0.0302, "step": 223700 }, { "epoch": 18.53, "learning_rate": 1.078491271820449e-05, "loss": 0.0351, "step": 223800 }, { "epoch": 18.53, "learning_rate": 1.0780756442227765e-05, "loss": 0.0352, "step": 223900 }, { "epoch": 18.54, "learning_rate": 1.0776600166251039e-05, "loss": 0.0343, "step": 224000 }, { "epoch": 18.54, "eval_loss": 0.16169136762619019, "eval_runtime": 4511.7855, "eval_samples_per_second": 16.755, "eval_steps_per_second": 1.396, "eval_wer": 0.0560152984380024, "step": 224000 }, { "epoch": 18.55, "learning_rate": 1.0772443890274316e-05, "loss": 0.0364, "step": 224100 }, { "epoch": 18.56, "learning_rate": 1.076828761429759e-05, "loss": 0.0344, "step": 224200 }, { "epoch": 18.57, "learning_rate": 1.0764131338320865e-05, "loss": 0.0331, "step": 224300 }, { "epoch": 18.58, "learning_rate": 1.0759975062344141e-05, "loss": 0.0371, "step": 224400 }, { "epoch": 18.58, "learning_rate": 1.0755818786367416e-05, "loss": 0.0402, "step": 224500 }, { "epoch": 18.58, "eval_loss": 0.14505957067012787, "eval_runtime": 4410.8628, "eval_samples_per_second": 17.138, "eval_steps_per_second": 1.428, "eval_wer": 0.05455222376580959, "step": 224500 }, { "epoch": 18.59, "learning_rate": 1.075166251039069e-05, "loss": 0.0303, "step": 224600 }, { "epoch": 18.6, "learning_rate": 1.0747506234413967e-05, "loss": 0.0325, "step": 224700 }, { "epoch": 18.61, "learning_rate": 1.0743349958437241e-05, "loss": 0.0339, "step": 224800 }, { "epoch": 18.62, "learning_rate": 1.0739193682460515e-05, "loss": 0.0343, "step": 224900 }, { "epoch": 18.63, "learning_rate": 1.073503740648379e-05, "loss": 0.035, "step": 225000 }, { "epoch": 18.63, "eval_loss": 0.1485747992992401, "eval_runtime": 4360.2812, "eval_samples_per_second": 17.337, "eval_steps_per_second": 1.445, "eval_wer": 0.052296842620242125, "step": 225000 }, { "epoch": 18.63, "learning_rate": 1.0730881130507068e-05, "loss": 0.0336, "step": 225100 }, { "epoch": 18.64, "learning_rate": 1.0726724854530342e-05, "loss": 0.0343, "step": 225200 }, { "epoch": 18.65, "learning_rate": 1.0722568578553615e-05, "loss": 0.0378, "step": 225300 }, { "epoch": 18.66, "learning_rate": 1.0718412302576893e-05, "loss": 0.0361, "step": 225400 }, { "epoch": 18.67, "learning_rate": 1.0714256026600166e-05, "loss": 0.0289, "step": 225500 }, { "epoch": 18.67, "eval_loss": 0.15769605338573456, "eval_runtime": 4384.1853, "eval_samples_per_second": 17.243, "eval_steps_per_second": 1.437, "eval_wer": 0.05374914807823373, "step": 225500 }, { "epoch": 18.68, "learning_rate": 1.0710099750623442e-05, "loss": 0.0298, "step": 225600 }, { "epoch": 18.68, "learning_rate": 1.070594347464672e-05, "loss": 0.0326, "step": 225700 }, { "epoch": 18.69, "learning_rate": 1.070182876142976e-05, "loss": 0.0382, "step": 225800 }, { "epoch": 18.7, "learning_rate": 1.0697672485453035e-05, "loss": 0.0339, "step": 225900 }, { "epoch": 18.71, "learning_rate": 1.069351620947631e-05, "loss": 0.0307, "step": 226000 }, { "epoch": 18.71, "eval_loss": 0.16679580509662628, "eval_runtime": 4410.4096, "eval_samples_per_second": 17.14, "eval_steps_per_second": 1.428, "eval_wer": 0.055821452582380644, "step": 226000 }, { "epoch": 18.72, "learning_rate": 1.0689359933499586e-05, "loss": 0.0358, "step": 226100 }, { "epoch": 18.73, "learning_rate": 1.068520365752286e-05, "loss": 0.0398, "step": 226200 }, { "epoch": 18.73, "learning_rate": 1.0681047381546135e-05, "loss": 0.04, "step": 226300 }, { "epoch": 18.74, "learning_rate": 1.067689110556941e-05, "loss": 0.0378, "step": 226400 }, { "epoch": 18.75, "learning_rate": 1.0672734829592686e-05, "loss": 0.0332, "step": 226500 }, { "epoch": 18.75, "eval_loss": 0.16404318809509277, "eval_runtime": 4408.1367, "eval_samples_per_second": 17.149, "eval_steps_per_second": 1.429, "eval_wer": 0.054786069559892986, "step": 226500 }, { "epoch": 18.76, "learning_rate": 1.066857855361596e-05, "loss": 0.0402, "step": 226600 }, { "epoch": 18.77, "learning_rate": 1.0664422277639237e-05, "loss": 0.0409, "step": 226700 }, { "epoch": 18.77, "learning_rate": 1.0660307564422279e-05, "loss": 0.0383, "step": 226800 }, { "epoch": 18.78, "learning_rate": 1.0656151288445553e-05, "loss": 0.0318, "step": 226900 }, { "epoch": 18.79, "learning_rate": 1.065199501246883e-05, "loss": 0.034, "step": 227000 }, { "epoch": 18.79, "eval_loss": 0.1552259922027588, "eval_runtime": 4343.4133, "eval_samples_per_second": 17.405, "eval_steps_per_second": 1.45, "eval_wer": 0.05425376268651894, "step": 227000 }, { "epoch": 18.8, "learning_rate": 1.0647838736492104e-05, "loss": 0.034, "step": 227100 }, { "epoch": 18.81, "learning_rate": 1.064368246051538e-05, "loss": 0.0311, "step": 227200 }, { "epoch": 18.82, "learning_rate": 1.0639526184538653e-05, "loss": 0.0393, "step": 227300 }, { "epoch": 18.82, "learning_rate": 1.063536990856193e-05, "loss": 0.0404, "step": 227400 }, { "epoch": 18.83, "learning_rate": 1.0631213632585204e-05, "loss": 0.0346, "step": 227500 }, { "epoch": 18.83, "eval_loss": 0.1626134067773819, "eval_runtime": 4391.7917, "eval_samples_per_second": 17.213, "eval_steps_per_second": 1.434, "eval_wer": 0.05549837615634438, "step": 227500 }, { "epoch": 18.84, "learning_rate": 1.0627057356608478e-05, "loss": 0.0338, "step": 227600 }, { "epoch": 18.85, "learning_rate": 1.0622901080631755e-05, "loss": 0.0355, "step": 227700 }, { "epoch": 18.86, "learning_rate": 1.0618744804655031e-05, "loss": 0.04, "step": 227800 }, { "epoch": 18.87, "learning_rate": 1.0614588528678305e-05, "loss": 0.0383, "step": 227900 }, { "epoch": 18.87, "learning_rate": 1.0610432252701582e-05, "loss": 0.0361, "step": 228000 }, { "epoch": 18.87, "eval_loss": 0.1563984453678131, "eval_runtime": 4366.2216, "eval_samples_per_second": 17.314, "eval_steps_per_second": 1.443, "eval_wer": 0.05614145209007371, "step": 228000 }, { "epoch": 18.88, "learning_rate": 1.0606275976724856e-05, "loss": 0.0362, "step": 228100 }, { "epoch": 18.89, "learning_rate": 1.060211970074813e-05, "loss": 0.0351, "step": 228200 }, { "epoch": 18.9, "learning_rate": 1.0597963424771405e-05, "loss": 0.0303, "step": 228300 }, { "epoch": 18.91, "learning_rate": 1.059380714879468e-05, "loss": 0.0377, "step": 228400 }, { "epoch": 18.92, "learning_rate": 1.0589650872817956e-05, "loss": 0.0343, "step": 228500 }, { "epoch": 18.92, "eval_loss": 0.14722448587417603, "eval_runtime": 4349.8809, "eval_samples_per_second": 17.379, "eval_steps_per_second": 1.448, "eval_wer": 0.05400760921906274, "step": 228500 }, { "epoch": 18.92, "learning_rate": 1.058549459684123e-05, "loss": 0.0338, "step": 228600 }, { "epoch": 18.93, "learning_rate": 1.0581338320864507e-05, "loss": 0.0365, "step": 228700 }, { "epoch": 18.94, "learning_rate": 1.0577182044887781e-05, "loss": 0.0368, "step": 228800 }, { "epoch": 18.95, "learning_rate": 1.0573025768911056e-05, "loss": 0.0382, "step": 228900 }, { "epoch": 18.96, "learning_rate": 1.0568869492934332e-05, "loss": 0.0324, "step": 229000 }, { "epoch": 18.96, "eval_loss": 0.15999796986579895, "eval_runtime": 4360.3438, "eval_samples_per_second": 17.337, "eval_steps_per_second": 1.445, "eval_wer": 0.055207607372911735, "step": 229000 }, { "epoch": 18.97, "learning_rate": 1.0564713216957608e-05, "loss": 0.0296, "step": 229100 }, { "epoch": 18.97, "learning_rate": 1.0560556940980881e-05, "loss": 0.0329, "step": 229200 }, { "epoch": 18.98, "learning_rate": 1.0556400665004157e-05, "loss": 0.0314, "step": 229300 }, { "epoch": 18.99, "learning_rate": 1.05522859517872e-05, "loss": 0.032, "step": 229400 }, { "epoch": 19.0, "learning_rate": 1.0548129675810474e-05, "loss": 0.0304, "step": 229500 }, { "epoch": 19.0, "eval_loss": 0.15255214273929596, "eval_runtime": 4328.7856, "eval_samples_per_second": 17.463, "eval_steps_per_second": 1.455, "eval_wer": 0.05331222567349896, "step": 229500 }, { "epoch": 19.01, "learning_rate": 1.054397339983375e-05, "loss": 0.037, "step": 229600 }, { "epoch": 19.01, "learning_rate": 1.0539817123857025e-05, "loss": 0.0387, "step": 229700 }, { "epoch": 19.02, "learning_rate": 1.05356608478803e-05, "loss": 0.0345, "step": 229800 }, { "epoch": 19.03, "learning_rate": 1.0531504571903575e-05, "loss": 0.0318, "step": 229900 }, { "epoch": 19.04, "learning_rate": 1.0527348295926852e-05, "loss": 0.036, "step": 230000 }, { "epoch": 19.04, "eval_loss": 0.15271221101284027, "eval_runtime": 4349.7612, "eval_samples_per_second": 17.379, "eval_steps_per_second": 1.448, "eval_wer": 0.05401068613740594, "step": 230000 }, { "epoch": 19.05, "learning_rate": 1.0523192019950126e-05, "loss": 0.0312, "step": 230100 }, { "epoch": 19.06, "learning_rate": 1.0519035743973401e-05, "loss": 0.0326, "step": 230200 }, { "epoch": 19.06, "learning_rate": 1.0514879467996675e-05, "loss": 0.0339, "step": 230300 }, { "epoch": 19.07, "learning_rate": 1.0510723192019952e-05, "loss": 0.0332, "step": 230400 }, { "epoch": 19.08, "learning_rate": 1.0506566916043226e-05, "loss": 0.0303, "step": 230500 }, { "epoch": 19.08, "eval_loss": 0.17356747388839722, "eval_runtime": 4240.9467, "eval_samples_per_second": 17.825, "eval_steps_per_second": 1.486, "eval_wer": 0.05586606789835708, "step": 230500 }, { "epoch": 19.09, "learning_rate": 1.05024106400665e-05, "loss": 0.0382, "step": 230600 }, { "epoch": 19.1, "learning_rate": 1.0498254364089777e-05, "loss": 0.0319, "step": 230700 }, { "epoch": 19.11, "learning_rate": 1.049409808811305e-05, "loss": 0.0372, "step": 230800 }, { "epoch": 19.11, "learning_rate": 1.0489941812136326e-05, "loss": 0.0333, "step": 230900 }, { "epoch": 19.12, "learning_rate": 1.0485785536159604e-05, "loss": 0.0389, "step": 231000 }, { "epoch": 19.12, "eval_loss": 0.16221372783184052, "eval_runtime": 4235.1153, "eval_samples_per_second": 17.85, "eval_steps_per_second": 1.488, "eval_wer": 0.055378376340959476, "step": 231000 }, { "epoch": 19.13, "learning_rate": 1.0481629260182877e-05, "loss": 0.0315, "step": 231100 }, { "epoch": 19.14, "learning_rate": 1.0477472984206151e-05, "loss": 0.0297, "step": 231200 }, { "epoch": 19.15, "learning_rate": 1.0473316708229427e-05, "loss": 0.0306, "step": 231300 }, { "epoch": 19.16, "learning_rate": 1.0469160432252702e-05, "loss": 0.0315, "step": 231400 }, { "epoch": 19.16, "learning_rate": 1.0465087281795512e-05, "loss": 0.0346, "step": 231500 }, { "epoch": 19.16, "eval_loss": 0.16887778043746948, "eval_runtime": 4246.6584, "eval_samples_per_second": 17.801, "eval_steps_per_second": 1.484, "eval_wer": 0.054302993380010184, "step": 231500 }, { "epoch": 19.17, "learning_rate": 1.0460931005818786e-05, "loss": 0.0331, "step": 231600 }, { "epoch": 19.18, "learning_rate": 1.0456774729842063e-05, "loss": 0.032, "step": 231700 }, { "epoch": 19.19, "learning_rate": 1.0452618453865337e-05, "loss": 0.0381, "step": 231800 }, { "epoch": 19.2, "learning_rate": 1.0448462177888612e-05, "loss": 0.0306, "step": 231900 }, { "epoch": 19.21, "learning_rate": 1.0444305901911888e-05, "loss": 0.0328, "step": 232000 }, { "epoch": 19.21, "eval_loss": 0.16647648811340332, "eval_runtime": 4229.7655, "eval_samples_per_second": 17.872, "eval_steps_per_second": 1.489, "eval_wer": 0.056042990703091225, "step": 232000 }, { "epoch": 19.21, "learning_rate": 1.0440149625935163e-05, "loss": 0.037, "step": 232100 }, { "epoch": 19.22, "learning_rate": 1.0435993349958437e-05, "loss": 0.0325, "step": 232200 }, { "epoch": 19.23, "learning_rate": 1.0431837073981714e-05, "loss": 0.0337, "step": 232300 }, { "epoch": 19.24, "learning_rate": 1.0427680798004988e-05, "loss": 0.0372, "step": 232400 }, { "epoch": 19.25, "learning_rate": 1.0423524522028264e-05, "loss": 0.0321, "step": 232500 }, { "epoch": 19.25, "eval_loss": 0.1617848426103592, "eval_runtime": 4240.4198, "eval_samples_per_second": 17.827, "eval_steps_per_second": 1.486, "eval_wer": 0.05617222127350573, "step": 232500 }, { "epoch": 19.25, "learning_rate": 1.0419368246051538e-05, "loss": 0.0292, "step": 232600 }, { "epoch": 19.26, "learning_rate": 1.0415211970074815e-05, "loss": 0.0327, "step": 232700 }, { "epoch": 19.27, "learning_rate": 1.0411055694098089e-05, "loss": 0.0339, "step": 232800 }, { "epoch": 19.28, "learning_rate": 1.0406899418121364e-05, "loss": 0.0331, "step": 232900 }, { "epoch": 19.29, "learning_rate": 1.040274314214464e-05, "loss": 0.0354, "step": 233000 }, { "epoch": 19.29, "eval_loss": 0.15175674855709076, "eval_runtime": 4249.1456, "eval_samples_per_second": 17.791, "eval_steps_per_second": 1.483, "eval_wer": 0.05365068669125125, "step": 233000 }, { "epoch": 19.3, "learning_rate": 1.0398586866167915e-05, "loss": 0.036, "step": 233100 }, { "epoch": 19.3, "learning_rate": 1.0394430590191189e-05, "loss": 0.0357, "step": 233200 }, { "epoch": 19.31, "learning_rate": 1.0390274314214466e-05, "loss": 0.0339, "step": 233300 }, { "epoch": 19.32, "learning_rate": 1.038611803823774e-05, "loss": 0.0442, "step": 233400 }, { "epoch": 19.33, "learning_rate": 1.0381961762261014e-05, "loss": 0.0352, "step": 233500 }, { "epoch": 19.33, "eval_loss": 0.14967215061187744, "eval_runtime": 4251.1927, "eval_samples_per_second": 17.782, "eval_steps_per_second": 1.482, "eval_wer": 0.052976841574089883, "step": 233500 }, { "epoch": 19.34, "learning_rate": 1.037780548628429e-05, "loss": 0.0365, "step": 233600 }, { "epoch": 19.35, "learning_rate": 1.0373649210307567e-05, "loss": 0.0403, "step": 233700 }, { "epoch": 19.35, "learning_rate": 1.036949293433084e-05, "loss": 0.0365, "step": 233800 }, { "epoch": 19.36, "learning_rate": 1.0365336658354114e-05, "loss": 0.0383, "step": 233900 }, { "epoch": 19.37, "learning_rate": 1.0361180382377391e-05, "loss": 0.0378, "step": 234000 }, { "epoch": 19.37, "eval_loss": 0.1584049016237259, "eval_runtime": 4254.6421, "eval_samples_per_second": 17.768, "eval_steps_per_second": 1.481, "eval_wer": 0.05367068666048206, "step": 234000 }, { "epoch": 19.38, "learning_rate": 1.0357024106400665e-05, "loss": 0.0324, "step": 234100 }, { "epoch": 19.39, "learning_rate": 1.035286783042394e-05, "loss": 0.0328, "step": 234200 }, { "epoch": 19.4, "learning_rate": 1.0348711554447218e-05, "loss": 0.0367, "step": 234300 }, { "epoch": 19.4, "learning_rate": 1.0344555278470492e-05, "loss": 0.0352, "step": 234400 }, { "epoch": 19.41, "learning_rate": 1.0340399002493766e-05, "loss": 0.0367, "step": 234500 }, { "epoch": 19.41, "eval_loss": 0.14730526506900787, "eval_runtime": 4239.633, "eval_samples_per_second": 17.831, "eval_steps_per_second": 1.486, "eval_wer": 0.053202995072315275, "step": 234500 }, { "epoch": 19.42, "learning_rate": 1.0336242726517041e-05, "loss": 0.0357, "step": 234600 }, { "epoch": 19.43, "learning_rate": 1.0332086450540317e-05, "loss": 0.0319, "step": 234700 }, { "epoch": 19.44, "learning_rate": 1.0327930174563592e-05, "loss": 0.0357, "step": 234800 }, { "epoch": 19.45, "learning_rate": 1.0323773898586866e-05, "loss": 0.029, "step": 234900 }, { "epoch": 19.45, "learning_rate": 1.0319617622610143e-05, "loss": 0.0351, "step": 235000 }, { "epoch": 19.45, "eval_loss": 0.1574089080095291, "eval_runtime": 4258.209, "eval_samples_per_second": 17.753, "eval_steps_per_second": 1.479, "eval_wer": 0.05415068592202166, "step": 235000 }, { "epoch": 19.46, "learning_rate": 1.0315461346633417e-05, "loss": 0.031, "step": 235100 }, { "epoch": 19.47, "learning_rate": 1.0311305070656693e-05, "loss": 0.0316, "step": 235200 }, { "epoch": 19.48, "learning_rate": 1.0307148794679968e-05, "loss": 0.0285, "step": 235300 }, { "epoch": 19.49, "learning_rate": 1.0302992518703244e-05, "loss": 0.0354, "step": 235400 }, { "epoch": 19.5, "learning_rate": 1.0298836242726517e-05, "loss": 0.0281, "step": 235500 }, { "epoch": 19.5, "eval_loss": 0.16067633032798767, "eval_runtime": 4304.0016, "eval_samples_per_second": 17.564, "eval_steps_per_second": 1.464, "eval_wer": 0.05339837938710863, "step": 235500 }, { "epoch": 19.5, "learning_rate": 1.0294679966749791e-05, "loss": 0.0294, "step": 235600 }, { "epoch": 19.51, "learning_rate": 1.0290523690773068e-05, "loss": 0.0352, "step": 235700 }, { "epoch": 19.52, "learning_rate": 1.0286367414796344e-05, "loss": 0.0322, "step": 235800 }, { "epoch": 19.53, "learning_rate": 1.0282211138819618e-05, "loss": 0.0328, "step": 235900 }, { "epoch": 19.54, "learning_rate": 1.0278054862842895e-05, "loss": 0.029, "step": 236000 }, { "epoch": 19.54, "eval_loss": 0.15970279276371002, "eval_runtime": 4338.3071, "eval_samples_per_second": 17.425, "eval_steps_per_second": 1.452, "eval_wer": 0.053395302468765435, "step": 236000 }, { "epoch": 19.54, "learning_rate": 1.0273898586866169e-05, "loss": 0.0351, "step": 236100 }, { "epoch": 19.55, "learning_rate": 1.0269742310889443e-05, "loss": 0.0305, "step": 236200 }, { "epoch": 19.56, "learning_rate": 1.026558603491272e-05, "loss": 0.0389, "step": 236300 }, { "epoch": 19.57, "learning_rate": 1.0261429758935995e-05, "loss": 0.0277, "step": 236400 }, { "epoch": 19.58, "learning_rate": 1.025727348295927e-05, "loss": 0.0327, "step": 236500 }, { "epoch": 19.58, "eval_loss": 0.1514918953180313, "eval_runtime": 4295.7895, "eval_samples_per_second": 17.597, "eval_steps_per_second": 1.467, "eval_wer": 0.05342607165219746, "step": 236500 }, { "epoch": 19.59, "learning_rate": 1.0253117206982543e-05, "loss": 0.0413, "step": 236600 }, { "epoch": 19.59, "learning_rate": 1.024896093100582e-05, "loss": 0.0314, "step": 236700 }, { "epoch": 19.6, "learning_rate": 1.0244804655029094e-05, "loss": 0.0354, "step": 236800 }, { "epoch": 19.61, "learning_rate": 1.024064837905237e-05, "loss": 0.0325, "step": 236900 }, { "epoch": 19.62, "learning_rate": 1.0236492103075645e-05, "loss": 0.0355, "step": 237000 }, { "epoch": 19.62, "eval_loss": 0.1611497849225998, "eval_runtime": 4250.5968, "eval_samples_per_second": 17.785, "eval_steps_per_second": 1.482, "eval_wer": 0.05681529720723506, "step": 237000 }, { "epoch": 19.63, "learning_rate": 1.023233582709892e-05, "loss": 0.03, "step": 237100 }, { "epoch": 19.64, "learning_rate": 1.0228179551122194e-05, "loss": 0.029, "step": 237200 }, { "epoch": 19.64, "learning_rate": 1.0224023275145472e-05, "loss": 0.0338, "step": 237300 }, { "epoch": 19.65, "learning_rate": 1.0219866999168745e-05, "loss": 0.0427, "step": 237400 }, { "epoch": 19.66, "learning_rate": 1.0215752285951787e-05, "loss": 0.0308, "step": 237500 }, { "epoch": 19.66, "eval_loss": 0.1428430825471878, "eval_runtime": 4250.3608, "eval_samples_per_second": 17.786, "eval_steps_per_second": 1.482, "eval_wer": 0.05273991886166329, "step": 237500 }, { "epoch": 19.67, "learning_rate": 1.0211596009975063e-05, "loss": 0.0354, "step": 237600 }, { "epoch": 19.68, "learning_rate": 1.0207439733998338e-05, "loss": 0.0357, "step": 237700 }, { "epoch": 19.69, "learning_rate": 1.0203283458021614e-05, "loss": 0.0333, "step": 237800 }, { "epoch": 19.69, "learning_rate": 1.0199127182044888e-05, "loss": 0.0354, "step": 237900 }, { "epoch": 19.7, "learning_rate": 1.0194970906068165e-05, "loss": 0.0308, "step": 238000 }, { "epoch": 19.7, "eval_loss": 0.15038935840129852, "eval_runtime": 4265.6391, "eval_samples_per_second": 17.722, "eval_steps_per_second": 1.477, "eval_wer": 0.05253376533266872, "step": 238000 }, { "epoch": 19.71, "learning_rate": 1.0190814630091439e-05, "loss": 0.031, "step": 238100 }, { "epoch": 19.72, "learning_rate": 1.0186658354114714e-05, "loss": 0.0274, "step": 238200 }, { "epoch": 19.73, "learning_rate": 1.018250207813799e-05, "loss": 0.0335, "step": 238300 }, { "epoch": 19.74, "learning_rate": 1.0178345802161265e-05, "loss": 0.0332, "step": 238400 }, { "epoch": 19.74, "learning_rate": 1.0174189526184539e-05, "loss": 0.0338, "step": 238500 }, { "epoch": 19.74, "eval_loss": 0.16093917191028595, "eval_runtime": 4253.5071, "eval_samples_per_second": 17.772, "eval_steps_per_second": 1.481, "eval_wer": 0.054187608942140086, "step": 238500 }, { "epoch": 19.75, "learning_rate": 1.0170033250207813e-05, "loss": 0.0354, "step": 238600 }, { "epoch": 19.76, "learning_rate": 1.016587697423109e-05, "loss": 0.0336, "step": 238700 }, { "epoch": 19.77, "learning_rate": 1.0161720698254366e-05, "loss": 0.0304, "step": 238800 }, { "epoch": 19.78, "learning_rate": 1.015756442227764e-05, "loss": 0.0326, "step": 238900 }, { "epoch": 19.78, "learning_rate": 1.0153408146300917e-05, "loss": 0.0341, "step": 239000 }, { "epoch": 19.78, "eval_loss": 0.1463533192873001, "eval_runtime": 4252.3292, "eval_samples_per_second": 17.777, "eval_steps_per_second": 1.482, "eval_wer": 0.05409222447350081, "step": 239000 }, { "epoch": 19.79, "learning_rate": 1.014925187032419e-05, "loss": 0.032, "step": 239100 }, { "epoch": 19.8, "learning_rate": 1.0145095594347464e-05, "loss": 0.0329, "step": 239200 }, { "epoch": 19.81, "learning_rate": 1.014098088113051e-05, "loss": 0.0328, "step": 239300 }, { "epoch": 19.82, "learning_rate": 1.0136824605153783e-05, "loss": 0.0333, "step": 239400 }, { "epoch": 19.83, "learning_rate": 1.0132668329177057e-05, "loss": 0.0349, "step": 239500 }, { "epoch": 19.83, "eval_loss": 0.15494167804718018, "eval_runtime": 4243.2655, "eval_samples_per_second": 17.815, "eval_steps_per_second": 1.485, "eval_wer": 0.05317376434805485, "step": 239500 }, { "epoch": 19.83, "learning_rate": 1.0128512053200333e-05, "loss": 0.0298, "step": 239600 }, { "epoch": 19.84, "learning_rate": 1.0124355777223608e-05, "loss": 0.0337, "step": 239700 }, { "epoch": 19.85, "learning_rate": 1.0120199501246884e-05, "loss": 0.0343, "step": 239800 }, { "epoch": 19.86, "learning_rate": 1.0116043225270158e-05, "loss": 0.0311, "step": 239900 }, { "epoch": 19.87, "learning_rate": 1.0111886949293435e-05, "loss": 0.0352, "step": 240000 }, { "epoch": 19.87, "eval_loss": 0.15909014642238617, "eval_runtime": 4278.6728, "eval_samples_per_second": 17.668, "eval_steps_per_second": 1.472, "eval_wer": 0.054958376987112326, "step": 240000 }, { "epoch": 19.88, "learning_rate": 1.0107730673316709e-05, "loss": 0.0367, "step": 240100 }, { "epoch": 19.88, "learning_rate": 1.0103574397339984e-05, "loss": 0.0326, "step": 240200 }, { "epoch": 19.89, "learning_rate": 1.009941812136326e-05, "loss": 0.0342, "step": 240300 }, { "epoch": 19.9, "learning_rate": 1.0095261845386535e-05, "loss": 0.0402, "step": 240400 }, { "epoch": 19.91, "learning_rate": 1.0091105569409809e-05, "loss": 0.0377, "step": 240500 }, { "epoch": 19.91, "eval_loss": 0.15816356241703033, "eval_runtime": 4263.6199, "eval_samples_per_second": 17.73, "eval_steps_per_second": 1.478, "eval_wer": 0.05565991436936251, "step": 240500 }, { "epoch": 19.92, "learning_rate": 1.0086949293433084e-05, "loss": 0.0325, "step": 240600 }, { "epoch": 19.93, "learning_rate": 1.008279301745636e-05, "loss": 0.0339, "step": 240700 }, { "epoch": 19.93, "learning_rate": 1.0078636741479636e-05, "loss": 0.0513, "step": 240800 }, { "epoch": 19.94, "learning_rate": 1.007448046550291e-05, "loss": 0.0339, "step": 240900 }, { "epoch": 19.95, "learning_rate": 1.0070324189526187e-05, "loss": 0.0283, "step": 241000 }, { "epoch": 19.95, "eval_loss": 0.14816011488437653, "eval_runtime": 4277.8751, "eval_samples_per_second": 17.671, "eval_steps_per_second": 1.473, "eval_wer": 0.052555303761071134, "step": 241000 }, { "epoch": 19.96, "learning_rate": 1.006616791354946e-05, "loss": 0.0339, "step": 241100 }, { "epoch": 19.97, "learning_rate": 1.0062011637572734e-05, "loss": 0.0309, "step": 241200 }, { "epoch": 19.98, "learning_rate": 1.0057855361596011e-05, "loss": 0.0346, "step": 241300 }, { "epoch": 19.98, "learning_rate": 1.0053699085619287e-05, "loss": 0.0314, "step": 241400 }, { "epoch": 19.99, "learning_rate": 1.004954280964256e-05, "loss": 0.0344, "step": 241500 }, { "epoch": 19.99, "eval_loss": 0.14264987409114838, "eval_runtime": 4245.9253, "eval_samples_per_second": 17.804, "eval_steps_per_second": 1.484, "eval_wer": 0.0528783801871074, "step": 241500 }, { "epoch": 20.0, "learning_rate": 1.0045386533665835e-05, "loss": 0.0343, "step": 241600 }, { "epoch": 20.01, "learning_rate": 1.0041230257689112e-05, "loss": 0.0347, "step": 241700 }, { "epoch": 20.02, "learning_rate": 1.0037073981712386e-05, "loss": 0.0323, "step": 241800 }, { "epoch": 20.02, "learning_rate": 1.0032917705735661e-05, "loss": 0.0331, "step": 241900 }, { "epoch": 20.03, "learning_rate": 1.0028761429758938e-05, "loss": 0.0329, "step": 242000 }, { "epoch": 20.03, "eval_loss": 0.15965133905410767, "eval_runtime": 4271.0084, "eval_samples_per_second": 17.7, "eval_steps_per_second": 1.475, "eval_wer": 0.05615991360013292, "step": 242000 }, { "epoch": 20.04, "learning_rate": 1.0024605153782212e-05, "loss": 0.029, "step": 242100 }, { "epoch": 20.05, "learning_rate": 1.0020490440565254e-05, "loss": 0.0276, "step": 242200 }, { "epoch": 20.06, "learning_rate": 1.0016334164588531e-05, "loss": 0.0301, "step": 242300 }, { "epoch": 20.07, "learning_rate": 1.0012177888611805e-05, "loss": 0.0434, "step": 242400 }, { "epoch": 20.07, "learning_rate": 1.0008021612635079e-05, "loss": 0.0293, "step": 242500 }, { "epoch": 20.07, "eval_loss": 0.16469527781009674, "eval_runtime": 4256.3213, "eval_samples_per_second": 17.761, "eval_steps_per_second": 1.48, "eval_wer": 0.054573762194212005, "step": 242500 }, { "epoch": 20.08, "learning_rate": 1.0003865336658356e-05, "loss": 0.0326, "step": 242600 }, { "epoch": 20.09, "learning_rate": 9.99970906068163e-06, "loss": 0.0352, "step": 242700 }, { "epoch": 20.1, "learning_rate": 9.995552784704905e-06, "loss": 0.0316, "step": 242800 }, { "epoch": 20.11, "learning_rate": 9.991396508728181e-06, "loss": 0.0321, "step": 242900 }, { "epoch": 20.12, "learning_rate": 9.987240232751455e-06, "loss": 0.029, "step": 243000 }, { "epoch": 20.12, "eval_loss": 0.15357014536857605, "eval_runtime": 4254.7754, "eval_samples_per_second": 17.767, "eval_steps_per_second": 1.481, "eval_wer": 0.05153068995278469, "step": 243000 }, { "epoch": 20.12, "learning_rate": 9.98308395677473e-06, "loss": 0.0327, "step": 243100 }, { "epoch": 20.13, "learning_rate": 9.978927680798006e-06, "loss": 0.0302, "step": 243200 }, { "epoch": 20.14, "learning_rate": 9.974771404821281e-06, "loss": 0.0335, "step": 243300 }, { "epoch": 20.15, "learning_rate": 9.970615128844557e-06, "loss": 0.0343, "step": 243400 }, { "epoch": 20.16, "learning_rate": 9.96645885286783e-06, "loss": 0.0336, "step": 243500 }, { "epoch": 20.16, "eval_loss": 0.14965561032295227, "eval_runtime": 4242.4975, "eval_samples_per_second": 17.819, "eval_steps_per_second": 1.485, "eval_wer": 0.05402145535160715, "step": 243500 }, { "epoch": 20.17, "learning_rate": 9.962302576891106e-06, "loss": 0.0314, "step": 243600 }, { "epoch": 20.17, "learning_rate": 9.958146300914382e-06, "loss": 0.0283, "step": 243700 }, { "epoch": 20.18, "learning_rate": 9.953990024937657e-06, "loss": 0.0296, "step": 243800 }, { "epoch": 20.19, "learning_rate": 9.949833748960933e-06, "loss": 0.0329, "step": 243900 }, { "epoch": 20.2, "learning_rate": 9.945677472984206e-06, "loss": 0.0325, "step": 244000 }, { "epoch": 20.2, "eval_loss": 0.15507319569587708, "eval_runtime": 4286.2749, "eval_samples_per_second": 17.637, "eval_steps_per_second": 1.47, "eval_wer": 0.054699915846283316, "step": 244000 }, { "epoch": 20.21, "learning_rate": 9.941521197007482e-06, "loss": 0.0309, "step": 244100 }, { "epoch": 20.22, "learning_rate": 9.937364921030758e-06, "loss": 0.0313, "step": 244200 }, { "epoch": 20.22, "learning_rate": 9.933208645054031e-06, "loss": 0.0327, "step": 244300 }, { "epoch": 20.23, "learning_rate": 9.929052369077309e-06, "loss": 0.0298, "step": 244400 }, { "epoch": 20.24, "learning_rate": 9.924896093100582e-06, "loss": 0.0268, "step": 244500 }, { "epoch": 20.24, "eval_loss": 0.15281596779823303, "eval_runtime": 4263.4282, "eval_samples_per_second": 17.731, "eval_steps_per_second": 1.478, "eval_wer": 0.05511376136344406, "step": 244500 }, { "epoch": 20.25, "learning_rate": 9.920739817123858e-06, "loss": 0.0351, "step": 244600 }, { "epoch": 20.26, "learning_rate": 9.916583541147133e-06, "loss": 0.0333, "step": 244700 }, { "epoch": 20.26, "learning_rate": 9.912427265170407e-06, "loss": 0.034, "step": 244800 }, { "epoch": 20.27, "learning_rate": 9.908270989193683e-06, "loss": 0.0316, "step": 244900 }, { "epoch": 20.28, "learning_rate": 9.904114713216958e-06, "loss": 0.0298, "step": 245000 }, { "epoch": 20.28, "eval_loss": 0.15852180123329163, "eval_runtime": 4254.2244, "eval_samples_per_second": 17.769, "eval_steps_per_second": 1.481, "eval_wer": 0.05426145498237695, "step": 245000 }, { "epoch": 20.29, "learning_rate": 9.899958437240234e-06, "loss": 0.029, "step": 245100 }, { "epoch": 20.3, "learning_rate": 9.89580216126351e-06, "loss": 0.0317, "step": 245200 }, { "epoch": 20.31, "learning_rate": 9.891645885286783e-06, "loss": 0.0344, "step": 245300 }, { "epoch": 20.31, "learning_rate": 9.887489609310059e-06, "loss": 0.0315, "step": 245400 }, { "epoch": 20.32, "learning_rate": 9.883333333333334e-06, "loss": 0.0323, "step": 245500 }, { "epoch": 20.32, "eval_loss": 0.16052758693695068, "eval_runtime": 4265.3914, "eval_samples_per_second": 17.723, "eval_steps_per_second": 1.477, "eval_wer": 0.05476453113149057, "step": 245500 }, { "epoch": 20.33, "learning_rate": 9.87917705735661e-06, "loss": 0.0373, "step": 245600 }, { "epoch": 20.34, "learning_rate": 9.875020781379885e-06, "loss": 0.0347, "step": 245700 }, { "epoch": 20.35, "learning_rate": 9.870864505403159e-06, "loss": 0.0298, "step": 245800 }, { "epoch": 20.36, "learning_rate": 9.866708229426435e-06, "loss": 0.0308, "step": 245900 }, { "epoch": 20.36, "learning_rate": 9.86255195344971e-06, "loss": 0.0298, "step": 246000 }, { "epoch": 20.36, "eval_loss": 0.16174006462097168, "eval_runtime": 4255.2596, "eval_samples_per_second": 17.765, "eval_steps_per_second": 1.481, "eval_wer": 0.05519683815871052, "step": 246000 }, { "epoch": 20.37, "learning_rate": 9.858395677472986e-06, "loss": 0.0373, "step": 246100 }, { "epoch": 20.38, "learning_rate": 9.854239401496261e-06, "loss": 0.0291, "step": 246200 }, { "epoch": 20.39, "learning_rate": 9.850083125519535e-06, "loss": 0.0349, "step": 246300 }, { "epoch": 20.4, "learning_rate": 9.84592684954281e-06, "loss": 0.0307, "step": 246400 }, { "epoch": 20.41, "learning_rate": 9.841770573566086e-06, "loss": 0.027, "step": 246500 }, { "epoch": 20.41, "eval_loss": 0.17689533531665802, "eval_runtime": 4258.8351, "eval_samples_per_second": 17.75, "eval_steps_per_second": 1.479, "eval_wer": 0.05707837372557888, "step": 246500 }, { "epoch": 20.41, "learning_rate": 9.837614297589361e-06, "loss": 0.0293, "step": 246600 }, { "epoch": 20.42, "learning_rate": 9.833458021612637e-06, "loss": 0.0294, "step": 246700 }, { "epoch": 20.43, "learning_rate": 9.82930174563591e-06, "loss": 0.0351, "step": 246800 }, { "epoch": 20.44, "learning_rate": 9.825145469659186e-06, "loss": 0.0314, "step": 246900 }, { "epoch": 20.45, "learning_rate": 9.82098919368246e-06, "loss": 0.0286, "step": 247000 }, { "epoch": 20.45, "eval_loss": 0.16199079155921936, "eval_runtime": 4264.7046, "eval_samples_per_second": 17.726, "eval_steps_per_second": 1.477, "eval_wer": 0.05441222398119387, "step": 247000 }, { "epoch": 20.46, "learning_rate": 9.816832917705736e-06, "loss": 0.0324, "step": 247100 }, { "epoch": 20.46, "learning_rate": 9.812676641729013e-06, "loss": 0.0335, "step": 247200 }, { "epoch": 20.47, "learning_rate": 9.808520365752287e-06, "loss": 0.0367, "step": 247300 }, { "epoch": 20.48, "learning_rate": 9.804364089775562e-06, "loss": 0.0334, "step": 247400 }, { "epoch": 20.49, "learning_rate": 9.800207813798836e-06, "loss": 0.0344, "step": 247500 }, { "epoch": 20.49, "eval_loss": 0.15828540921211243, "eval_runtime": 4246.3663, "eval_samples_per_second": 17.802, "eval_steps_per_second": 1.484, "eval_wer": 0.05470914660131292, "step": 247500 }, { "epoch": 20.5, "learning_rate": 9.796051537822112e-06, "loss": 0.0351, "step": 247600 }, { "epoch": 20.5, "learning_rate": 9.791895261845387e-06, "loss": 0.031, "step": 247700 }, { "epoch": 20.51, "learning_rate": 9.787738985868663e-06, "loss": 0.0306, "step": 247800 }, { "epoch": 20.52, "learning_rate": 9.783582709891938e-06, "loss": 0.0428, "step": 247900 }, { "epoch": 20.53, "learning_rate": 9.779426433915212e-06, "loss": 0.0369, "step": 248000 }, { "epoch": 20.53, "eval_loss": 0.15957312285900116, "eval_runtime": 4251.2772, "eval_samples_per_second": 17.782, "eval_steps_per_second": 1.482, "eval_wer": 0.053726071190659704, "step": 248000 }, { "epoch": 20.54, "learning_rate": 9.775270157938487e-06, "loss": 0.0305, "step": 248100 }, { "epoch": 20.55, "learning_rate": 9.771155444721531e-06, "loss": 0.0358, "step": 248200 }, { "epoch": 20.55, "learning_rate": 9.766999168744805e-06, "loss": 0.0309, "step": 248300 }, { "epoch": 20.56, "learning_rate": 9.76284289276808e-06, "loss": 0.0323, "step": 248400 }, { "epoch": 20.57, "learning_rate": 9.758686616791356e-06, "loss": 0.0357, "step": 248500 }, { "epoch": 20.57, "eval_loss": 0.16622671484947205, "eval_runtime": 4263.501, "eval_samples_per_second": 17.731, "eval_steps_per_second": 1.478, "eval_wer": 0.055278376494805395, "step": 248500 }, { "epoch": 20.58, "learning_rate": 9.754530340814631e-06, "loss": 0.0338, "step": 248600 }, { "epoch": 20.59, "learning_rate": 9.750374064837907e-06, "loss": 0.0342, "step": 248700 }, { "epoch": 20.6, "learning_rate": 9.74621778886118e-06, "loss": 0.0336, "step": 248800 }, { "epoch": 20.6, "learning_rate": 9.742061512884456e-06, "loss": 0.0278, "step": 248900 }, { "epoch": 20.61, "learning_rate": 9.7379467996675e-06, "loss": 0.031, "step": 249000 }, { "epoch": 20.61, "eval_loss": 0.16190360486507416, "eval_runtime": 4267.368, "eval_samples_per_second": 17.715, "eval_steps_per_second": 1.476, "eval_wer": 0.05397530157645911, "step": 249000 }, { "epoch": 20.62, "learning_rate": 9.733790523690773e-06, "loss": 0.0293, "step": 249100 }, { "epoch": 20.63, "learning_rate": 9.729634247714049e-06, "loss": 0.0393, "step": 249200 }, { "epoch": 20.64, "learning_rate": 9.725477971737325e-06, "loss": 0.0393, "step": 249300 }, { "epoch": 20.65, "learning_rate": 9.7213216957606e-06, "loss": 0.0359, "step": 249400 }, { "epoch": 20.65, "learning_rate": 9.717165419783876e-06, "loss": 0.042, "step": 249500 }, { "epoch": 20.65, "eval_loss": 0.14939779043197632, "eval_runtime": 4263.492, "eval_samples_per_second": 17.731, "eval_steps_per_second": 1.478, "eval_wer": 0.05309837984864638, "step": 249500 }, { "epoch": 20.66, "learning_rate": 9.71300914380715e-06, "loss": 0.0324, "step": 249600 }, { "epoch": 20.67, "learning_rate": 9.708852867830425e-06, "loss": 0.0332, "step": 249700 }, { "epoch": 20.68, "learning_rate": 9.704696591853699e-06, "loss": 0.0299, "step": 249800 }, { "epoch": 20.69, "learning_rate": 9.700540315876976e-06, "loss": 0.0382, "step": 249900 }, { "epoch": 20.7, "learning_rate": 9.696384039900251e-06, "loss": 0.0342, "step": 250000 }, { "epoch": 20.7, "eval_loss": 0.15561692416667938, "eval_runtime": 4255.903, "eval_samples_per_second": 17.762, "eval_steps_per_second": 1.48, "eval_wer": 0.053473763886517096, "step": 250000 }, { "epoch": 20.7, "learning_rate": 9.692227763923525e-06, "loss": 0.0296, "step": 250100 }, { "epoch": 20.71, "learning_rate": 9.6880714879468e-06, "loss": 0.0362, "step": 250200 }, { "epoch": 20.72, "learning_rate": 9.683915211970075e-06, "loss": 0.0309, "step": 250300 }, { "epoch": 20.73, "learning_rate": 9.67975893599335e-06, "loss": 0.0325, "step": 250400 }, { "epoch": 20.74, "learning_rate": 9.675602660016626e-06, "loss": 0.0304, "step": 250500 }, { "epoch": 20.74, "eval_loss": 0.15056154131889343, "eval_runtime": 4269.4573, "eval_samples_per_second": 17.706, "eval_steps_per_second": 1.476, "eval_wer": 0.0531137644403624, "step": 250500 }, { "epoch": 20.75, "learning_rate": 9.671446384039901e-06, "loss": 0.0326, "step": 250600 }, { "epoch": 20.75, "learning_rate": 9.667290108063177e-06, "loss": 0.0392, "step": 250700 }, { "epoch": 20.76, "learning_rate": 9.66313383208645e-06, "loss": 0.0294, "step": 250800 }, { "epoch": 20.77, "learning_rate": 9.658977556109726e-06, "loss": 0.0333, "step": 250900 }, { "epoch": 20.78, "learning_rate": 9.654821280133002e-06, "loss": 0.0339, "step": 251000 }, { "epoch": 20.78, "eval_loss": 0.15236234664916992, "eval_runtime": 4262.5725, "eval_samples_per_second": 17.735, "eval_steps_per_second": 1.478, "eval_wer": 0.052990687706634294, "step": 251000 }, { "epoch": 20.79, "learning_rate": 9.650665004156277e-06, "loss": 0.0314, "step": 251100 }, { "epoch": 20.79, "learning_rate": 9.646508728179553e-06, "loss": 0.0305, "step": 251200 }, { "epoch": 20.8, "learning_rate": 9.642352452202826e-06, "loss": 0.0342, "step": 251300 }, { "epoch": 20.81, "learning_rate": 9.63823773898587e-06, "loss": 0.0313, "step": 251400 }, { "epoch": 20.82, "learning_rate": 9.634081463009145e-06, "loss": 0.0305, "step": 251500 }, { "epoch": 20.82, "eval_loss": 0.16681845486164093, "eval_runtime": 4260.4055, "eval_samples_per_second": 17.744, "eval_steps_per_second": 1.479, "eval_wer": 0.05632452873149426, "step": 251500 }, { "epoch": 20.83, "learning_rate": 9.62992518703242e-06, "loss": 0.0291, "step": 251600 }, { "epoch": 20.84, "learning_rate": 9.625768911055695e-06, "loss": 0.0266, "step": 251700 }, { "epoch": 20.84, "learning_rate": 9.62161263507897e-06, "loss": 0.0327, "step": 251800 }, { "epoch": 20.85, "learning_rate": 9.617456359102246e-06, "loss": 0.0321, "step": 251900 }, { "epoch": 20.86, "learning_rate": 9.613341645885288e-06, "loss": 0.0308, "step": 252000 }, { "epoch": 20.86, "eval_loss": 0.16326159238815308, "eval_runtime": 4265.1464, "eval_samples_per_second": 17.724, "eval_steps_per_second": 1.477, "eval_wer": 0.05487837711018906, "step": 252000 }, { "epoch": 20.87, "learning_rate": 9.609185369908563e-06, "loss": 0.0336, "step": 252100 }, { "epoch": 20.88, "learning_rate": 9.605029093931839e-06, "loss": 0.0327, "step": 252200 }, { "epoch": 20.89, "learning_rate": 9.600872817955112e-06, "loss": 0.0329, "step": 252300 }, { "epoch": 20.89, "learning_rate": 9.596716541978388e-06, "loss": 0.0342, "step": 252400 }, { "epoch": 20.9, "learning_rate": 9.592560266001664e-06, "loss": 0.0322, "step": 252500 }, { "epoch": 20.9, "eval_loss": 0.1632661074399948, "eval_runtime": 4264.8616, "eval_samples_per_second": 17.725, "eval_steps_per_second": 1.477, "eval_wer": 0.05404453223918117, "step": 252500 }, { "epoch": 20.91, "learning_rate": 9.588403990024937e-06, "loss": 0.0318, "step": 252600 }, { "epoch": 20.92, "learning_rate": 9.584247714048215e-06, "loss": 0.0275, "step": 252700 }, { "epoch": 20.93, "learning_rate": 9.580091438071488e-06, "loss": 0.0307, "step": 252800 }, { "epoch": 20.94, "learning_rate": 9.575935162094764e-06, "loss": 0.0299, "step": 252900 }, { "epoch": 20.94, "learning_rate": 9.57177888611804e-06, "loss": 0.0268, "step": 253000 }, { "epoch": 20.94, "eval_loss": 0.1592932790517807, "eval_runtime": 4259.2009, "eval_samples_per_second": 17.749, "eval_steps_per_second": 1.479, "eval_wer": 0.05209530446876236, "step": 253000 }, { "epoch": 20.95, "learning_rate": 9.567622610141313e-06, "loss": 0.0347, "step": 253100 }, { "epoch": 20.96, "learning_rate": 9.563466334164589e-06, "loss": 0.0341, "step": 253200 }, { "epoch": 20.97, "learning_rate": 9.559310058187864e-06, "loss": 0.032, "step": 253300 }, { "epoch": 20.98, "learning_rate": 9.55515378221114e-06, "loss": 0.0361, "step": 253400 }, { "epoch": 20.99, "learning_rate": 9.550997506234415e-06, "loss": 0.0352, "step": 253500 }, { "epoch": 20.99, "eval_loss": 0.1567884236574173, "eval_runtime": 4247.361, "eval_samples_per_second": 17.798, "eval_steps_per_second": 1.483, "eval_wer": 0.05333376410190138, "step": 253500 }, { "epoch": 20.99, "learning_rate": 9.546841230257689e-06, "loss": 0.0315, "step": 253600 }, { "epoch": 21.0, "learning_rate": 9.542684954280965e-06, "loss": 0.0275, "step": 253700 }, { "epoch": 21.01, "learning_rate": 9.53852867830424e-06, "loss": 0.0277, "step": 253800 }, { "epoch": 21.02, "learning_rate": 9.534372402327516e-06, "loss": 0.0302, "step": 253900 }, { "epoch": 21.03, "learning_rate": 9.530216126350791e-06, "loss": 0.0247, "step": 254000 }, { "epoch": 21.03, "eval_loss": 0.17209503054618835, "eval_runtime": 4269.5179, "eval_samples_per_second": 17.706, "eval_steps_per_second": 1.476, "eval_wer": 0.053039918400125535, "step": 254000 }, { "epoch": 21.03, "learning_rate": 9.526059850374065e-06, "loss": 0.0307, "step": 254100 }, { "epoch": 21.04, "learning_rate": 9.52190357439734e-06, "loss": 0.0269, "step": 254200 }, { "epoch": 21.05, "learning_rate": 9.517747298420616e-06, "loss": 0.0309, "step": 254300 }, { "epoch": 21.06, "learning_rate": 9.513591022443892e-06, "loss": 0.0315, "step": 254400 }, { "epoch": 21.07, "learning_rate": 9.509434746467167e-06, "loss": 0.0342, "step": 254500 }, { "epoch": 21.07, "eval_loss": 0.17056623101234436, "eval_runtime": 4276.8124, "eval_samples_per_second": 17.676, "eval_steps_per_second": 1.473, "eval_wer": 0.055101453690071245, "step": 254500 }, { "epoch": 21.08, "learning_rate": 9.505278470490441e-06, "loss": 0.029, "step": 254600 }, { "epoch": 21.08, "learning_rate": 9.501122194513716e-06, "loss": 0.0272, "step": 254700 }, { "epoch": 21.09, "learning_rate": 9.496965918536992e-06, "loss": 0.0273, "step": 254800 }, { "epoch": 21.1, "learning_rate": 9.492809642560267e-06, "loss": 0.0334, "step": 254900 }, { "epoch": 21.11, "learning_rate": 9.488653366583543e-06, "loss": 0.0296, "step": 255000 }, { "epoch": 21.11, "eval_loss": 0.16262391209602356, "eval_runtime": 4273.7247, "eval_samples_per_second": 17.688, "eval_steps_per_second": 1.474, "eval_wer": 0.052698380464030053, "step": 255000 }, { "epoch": 21.12, "learning_rate": 9.484497090606817e-06, "loss": 0.0287, "step": 255100 }, { "epoch": 21.13, "learning_rate": 9.480340814630092e-06, "loss": 0.0334, "step": 255200 }, { "epoch": 21.13, "learning_rate": 9.476184538653366e-06, "loss": 0.0265, "step": 255300 }, { "epoch": 21.14, "learning_rate": 9.472028262676642e-06, "loss": 0.0302, "step": 255400 }, { "epoch": 21.15, "learning_rate": 9.467871986699919e-06, "loss": 0.032, "step": 255500 }, { "epoch": 21.15, "eval_loss": 0.1463223248720169, "eval_runtime": 4266.4689, "eval_samples_per_second": 17.718, "eval_steps_per_second": 1.477, "eval_wer": 0.051782997256927296, "step": 255500 }, { "epoch": 21.16, "learning_rate": 9.463715710723193e-06, "loss": 0.0294, "step": 255600 }, { "epoch": 21.17, "learning_rate": 9.459559434746468e-06, "loss": 0.0281, "step": 255700 }, { "epoch": 21.18, "learning_rate": 9.455403158769742e-06, "loss": 0.0246, "step": 255800 }, { "epoch": 21.18, "learning_rate": 9.451246882793017e-06, "loss": 0.0291, "step": 255900 }, { "epoch": 21.19, "learning_rate": 9.447090606816293e-06, "loss": 0.0349, "step": 256000 }, { "epoch": 21.19, "eval_loss": 0.14798389375209808, "eval_runtime": 4258.2723, "eval_samples_per_second": 17.753, "eval_steps_per_second": 1.479, "eval_wer": 0.05267991895397084, "step": 256000 }, { "epoch": 21.2, "learning_rate": 9.442934330839569e-06, "loss": 0.0316, "step": 256100 }, { "epoch": 21.21, "learning_rate": 9.438778054862844e-06, "loss": 0.0294, "step": 256200 }, { "epoch": 21.22, "learning_rate": 9.434621778886118e-06, "loss": 0.0334, "step": 256300 }, { "epoch": 21.23, "learning_rate": 9.430465502909393e-06, "loss": 0.0292, "step": 256400 }, { "epoch": 21.23, "learning_rate": 9.426309226932669e-06, "loss": 0.034, "step": 256500 }, { "epoch": 21.23, "eval_loss": 0.14688608050346375, "eval_runtime": 4249.3983, "eval_samples_per_second": 17.79, "eval_steps_per_second": 1.483, "eval_wer": 0.05175838191018168, "step": 256500 }, { "epoch": 21.24, "learning_rate": 9.422152950955944e-06, "loss": 0.0314, "step": 256600 }, { "epoch": 21.25, "learning_rate": 9.41799667497922e-06, "loss": 0.0284, "step": 256700 }, { "epoch": 21.26, "learning_rate": 9.413840399002494e-06, "loss": 0.0278, "step": 256800 }, { "epoch": 21.27, "learning_rate": 9.40968412302577e-06, "loss": 0.0321, "step": 256900 }, { "epoch": 21.27, "learning_rate": 9.405527847049045e-06, "loss": 0.0338, "step": 257000 }, { "epoch": 21.27, "eval_loss": 0.14212098717689514, "eval_runtime": 4275.6388, "eval_samples_per_second": 17.68, "eval_steps_per_second": 1.473, "eval_wer": 0.05195222776580344, "step": 257000 }, { "epoch": 21.28, "learning_rate": 9.40137157107232e-06, "loss": 0.0292, "step": 257100 }, { "epoch": 21.29, "learning_rate": 9.397215295095596e-06, "loss": 0.032, "step": 257200 }, { "epoch": 21.3, "learning_rate": 9.39305901911887e-06, "loss": 0.0334, "step": 257300 }, { "epoch": 21.31, "learning_rate": 9.388902743142145e-06, "loss": 0.03, "step": 257400 }, { "epoch": 21.32, "learning_rate": 9.38474646716542e-06, "loss": 0.0289, "step": 257500 }, { "epoch": 21.32, "eval_loss": 0.15305761992931366, "eval_runtime": 4261.1618, "eval_samples_per_second": 17.74, "eval_steps_per_second": 1.478, "eval_wer": 0.053567609895984775, "step": 257500 }, { "epoch": 21.32, "learning_rate": 9.380590191188696e-06, "loss": 0.0337, "step": 257600 }, { "epoch": 21.33, "learning_rate": 9.376433915211972e-06, "loss": 0.0336, "step": 257700 }, { "epoch": 21.34, "learning_rate": 9.372277639235246e-06, "loss": 0.026, "step": 257800 }, { "epoch": 21.35, "learning_rate": 9.368121363258521e-06, "loss": 0.0254, "step": 257900 }, { "epoch": 21.36, "learning_rate": 9.363965087281797e-06, "loss": 0.0253, "step": 258000 }, { "epoch": 21.36, "eval_loss": 0.15871678292751312, "eval_runtime": 4259.7506, "eval_samples_per_second": 17.746, "eval_steps_per_second": 1.479, "eval_wer": 0.05341376397882465, "step": 258000 }, { "epoch": 21.37, "learning_rate": 9.35980881130507e-06, "loss": 0.0289, "step": 258100 }, { "epoch": 21.37, "learning_rate": 9.355652535328346e-06, "loss": 0.0316, "step": 258200 }, { "epoch": 21.38, "learning_rate": 9.351496259351621e-06, "loss": 0.0277, "step": 258300 }, { "epoch": 21.39, "learning_rate": 9.347339983374897e-06, "loss": 0.029, "step": 258400 }, { "epoch": 21.4, "learning_rate": 9.343225270157939e-06, "loss": 0.0287, "step": 258500 }, { "epoch": 21.4, "eval_loss": 0.15664973855018616, "eval_runtime": 4276.7902, "eval_samples_per_second": 17.676, "eval_steps_per_second": 1.473, "eval_wer": 0.05321068736817328, "step": 258500 }, { "epoch": 21.41, "learning_rate": 9.339068994181214e-06, "loss": 0.0285, "step": 258600 }, { "epoch": 21.42, "learning_rate": 9.33491271820449e-06, "loss": 0.0296, "step": 258700 }, { "epoch": 21.42, "learning_rate": 9.330798004987532e-06, "loss": 0.026, "step": 258800 }, { "epoch": 21.43, "learning_rate": 9.326641729010807e-06, "loss": 0.0322, "step": 258900 }, { "epoch": 21.44, "learning_rate": 9.322485453034083e-06, "loss": 0.0279, "step": 259000 }, { "epoch": 21.44, "eval_loss": 0.1633937507867813, "eval_runtime": 4256.698, "eval_samples_per_second": 17.759, "eval_steps_per_second": 1.48, "eval_wer": 0.0535999175385884, "step": 259000 }, { "epoch": 21.45, "learning_rate": 9.318329177057356e-06, "loss": 0.0326, "step": 259100 }, { "epoch": 21.46, "learning_rate": 9.314172901080632e-06, "loss": 0.028, "step": 259200 }, { "epoch": 21.47, "learning_rate": 9.310016625103908e-06, "loss": 0.0315, "step": 259300 }, { "epoch": 21.47, "learning_rate": 9.305860349127183e-06, "loss": 0.0336, "step": 259400 }, { "epoch": 21.48, "learning_rate": 9.301704073150459e-06, "loss": 0.0318, "step": 259500 }, { "epoch": 21.48, "eval_loss": 0.1576114445924759, "eval_runtime": 4271.1643, "eval_samples_per_second": 17.699, "eval_steps_per_second": 1.475, "eval_wer": 0.0534722254273455, "step": 259500 }, { "epoch": 21.49, "learning_rate": 9.297547797173732e-06, "loss": 0.0376, "step": 259600 }, { "epoch": 21.5, "learning_rate": 9.293391521197008e-06, "loss": 0.0354, "step": 259700 }, { "epoch": 21.51, "learning_rate": 9.289235245220283e-06, "loss": 0.0306, "step": 259800 }, { "epoch": 21.51, "learning_rate": 9.285078969243559e-06, "loss": 0.0294, "step": 259900 }, { "epoch": 21.52, "learning_rate": 9.280922693266834e-06, "loss": 0.028, "step": 260000 }, { "epoch": 21.52, "eval_loss": 0.1622873842716217, "eval_runtime": 4284.388, "eval_samples_per_second": 17.644, "eval_steps_per_second": 1.47, "eval_wer": 0.05457530065338361, "step": 260000 }, { "epoch": 21.53, "learning_rate": 9.276766417290108e-06, "loss": 0.0284, "step": 260100 }, { "epoch": 21.54, "learning_rate": 9.272610141313384e-06, "loss": 0.0278, "step": 260200 }, { "epoch": 21.55, "learning_rate": 9.26845386533666e-06, "loss": 0.0301, "step": 260300 }, { "epoch": 21.56, "learning_rate": 9.264297589359935e-06, "loss": 0.033, "step": 260400 }, { "epoch": 21.56, "learning_rate": 9.26014131338321e-06, "loss": 0.0303, "step": 260500 }, { "epoch": 21.56, "eval_loss": 0.15286412835121155, "eval_runtime": 4256.1178, "eval_samples_per_second": 17.761, "eval_steps_per_second": 1.48, "eval_wer": 0.05230761183444333, "step": 260500 }, { "epoch": 21.57, "learning_rate": 9.255985037406484e-06, "loss": 0.028, "step": 260600 }, { "epoch": 21.58, "learning_rate": 9.25182876142976e-06, "loss": 0.0268, "step": 260700 }, { "epoch": 21.59, "learning_rate": 9.247672485453035e-06, "loss": 0.0304, "step": 260800 }, { "epoch": 21.6, "learning_rate": 9.243516209476309e-06, "loss": 0.0302, "step": 260900 }, { "epoch": 21.61, "learning_rate": 9.239359933499586e-06, "loss": 0.0304, "step": 261000 }, { "epoch": 21.61, "eval_loss": 0.16833233833312988, "eval_runtime": 4252.3263, "eval_samples_per_second": 17.777, "eval_steps_per_second": 1.482, "eval_wer": 0.055313761055752225, "step": 261000 }, { "epoch": 21.61, "learning_rate": 9.23520365752286e-06, "loss": 0.0398, "step": 261100 }, { "epoch": 21.62, "learning_rate": 9.231047381546136e-06, "loss": 0.0294, "step": 261200 }, { "epoch": 21.63, "learning_rate": 9.226891105569411e-06, "loss": 0.0298, "step": 261300 }, { "epoch": 21.64, "learning_rate": 9.222734829592685e-06, "loss": 0.0322, "step": 261400 }, { "epoch": 21.65, "learning_rate": 9.21857855361596e-06, "loss": 0.034, "step": 261500 }, { "epoch": 21.65, "eval_loss": 0.17345653474330902, "eval_runtime": 4261.3939, "eval_samples_per_second": 17.74, "eval_steps_per_second": 1.478, "eval_wer": 0.055016838435633174, "step": 261500 }, { "epoch": 21.66, "learning_rate": 9.214422277639236e-06, "loss": 0.0363, "step": 261600 }, { "epoch": 21.66, "learning_rate": 9.210266001662511e-06, "loss": 0.0306, "step": 261700 }, { "epoch": 21.67, "learning_rate": 9.206109725685787e-06, "loss": 0.0291, "step": 261800 }, { "epoch": 21.68, "learning_rate": 9.20195344970906e-06, "loss": 0.034, "step": 261900 }, { "epoch": 21.69, "learning_rate": 9.197797173732336e-06, "loss": 0.03, "step": 262000 }, { "epoch": 21.69, "eval_loss": 0.17540572583675385, "eval_runtime": 4249.579, "eval_samples_per_second": 17.789, "eval_steps_per_second": 1.482, "eval_wer": 0.057261450366999436, "step": 262000 }, { "epoch": 21.7, "learning_rate": 9.193640897755612e-06, "loss": 0.0355, "step": 262100 }, { "epoch": 21.71, "learning_rate": 9.189484621778887e-06, "loss": 0.0287, "step": 262200 }, { "epoch": 21.71, "learning_rate": 9.185328345802163e-06, "loss": 0.0294, "step": 262300 }, { "epoch": 21.72, "learning_rate": 9.181172069825437e-06, "loss": 0.0306, "step": 262400 }, { "epoch": 21.73, "learning_rate": 9.177015793848712e-06, "loss": 0.0308, "step": 262500 }, { "epoch": 21.73, "eval_loss": 0.16141650080680847, "eval_runtime": 4247.498, "eval_samples_per_second": 17.798, "eval_steps_per_second": 1.483, "eval_wer": 0.053310687214327364, "step": 262500 }, { "epoch": 21.74, "learning_rate": 9.172859517871988e-06, "loss": 0.031, "step": 262600 }, { "epoch": 21.75, "learning_rate": 9.168703241895263e-06, "loss": 0.0311, "step": 262700 }, { "epoch": 21.75, "learning_rate": 9.164546965918539e-06, "loss": 0.0323, "step": 262800 }, { "epoch": 21.76, "learning_rate": 9.160390689941813e-06, "loss": 0.0261, "step": 262900 }, { "epoch": 21.77, "learning_rate": 9.156234413965088e-06, "loss": 0.0292, "step": 263000 }, { "epoch": 21.77, "eval_loss": 0.15404871106147766, "eval_runtime": 4268.8031, "eval_samples_per_second": 17.709, "eval_steps_per_second": 1.476, "eval_wer": 0.0520291507243835, "step": 263000 }, { "epoch": 21.78, "learning_rate": 9.152078137988364e-06, "loss": 0.0341, "step": 263100 }, { "epoch": 21.79, "learning_rate": 9.147921862011639e-06, "loss": 0.04, "step": 263200 }, { "epoch": 21.8, "learning_rate": 9.143765586034915e-06, "loss": 0.0267, "step": 263300 }, { "epoch": 21.8, "learning_rate": 9.139650872817956e-06, "loss": 0.0282, "step": 263400 }, { "epoch": 21.81, "learning_rate": 9.135494596841232e-06, "loss": 0.0274, "step": 263500 }, { "epoch": 21.81, "eval_loss": 0.16027560830116272, "eval_runtime": 4267.5106, "eval_samples_per_second": 17.714, "eval_steps_per_second": 1.476, "eval_wer": 0.05231530413030134, "step": 263500 }, { "epoch": 21.82, "learning_rate": 9.131338320864506e-06, "loss": 0.0324, "step": 263600 }, { "epoch": 21.83, "learning_rate": 9.127182044887781e-06, "loss": 0.0291, "step": 263700 }, { "epoch": 21.84, "learning_rate": 9.123025768911057e-06, "loss": 0.0267, "step": 263800 }, { "epoch": 21.85, "learning_rate": 9.118911055694099e-06, "loss": 0.0289, "step": 263900 }, { "epoch": 21.85, "learning_rate": 9.114754779717374e-06, "loss": 0.0318, "step": 264000 }, { "epoch": 21.85, "eval_loss": 0.1559901088476181, "eval_runtime": 4255.6508, "eval_samples_per_second": 17.763, "eval_steps_per_second": 1.48, "eval_wer": 0.05220761198828925, "step": 264000 }, { "epoch": 21.86, "learning_rate": 9.110598503740648e-06, "loss": 0.0359, "step": 264100 }, { "epoch": 21.87, "learning_rate": 9.106442227763923e-06, "loss": 0.0294, "step": 264200 }, { "epoch": 21.88, "learning_rate": 9.102285951787199e-06, "loss": 0.0254, "step": 264300 }, { "epoch": 21.89, "learning_rate": 9.098129675810475e-06, "loss": 0.035, "step": 264400 }, { "epoch": 21.9, "learning_rate": 9.09397339983375e-06, "loss": 0.0302, "step": 264500 }, { "epoch": 21.9, "eval_loss": 0.15430544316768646, "eval_runtime": 4251.9046, "eval_samples_per_second": 17.779, "eval_steps_per_second": 1.482, "eval_wer": 0.05306914912438596, "step": 264500 }, { "epoch": 21.9, "learning_rate": 9.089817123857024e-06, "loss": 0.0278, "step": 264600 }, { "epoch": 21.91, "learning_rate": 9.0856608478803e-06, "loss": 0.0335, "step": 264700 }, { "epoch": 21.92, "learning_rate": 9.081504571903575e-06, "loss": 0.029, "step": 264800 }, { "epoch": 21.93, "learning_rate": 9.07734829592685e-06, "loss": 0.033, "step": 264900 }, { "epoch": 21.94, "learning_rate": 9.073192019950126e-06, "loss": 0.0263, "step": 265000 }, { "epoch": 21.94, "eval_loss": 0.16334763169288635, "eval_runtime": 4267.5312, "eval_samples_per_second": 17.714, "eval_steps_per_second": 1.476, "eval_wer": 0.05297068773740348, "step": 265000 }, { "epoch": 21.95, "learning_rate": 9.0690357439734e-06, "loss": 0.0335, "step": 265100 }, { "epoch": 21.95, "learning_rate": 9.064879467996675e-06, "loss": 0.0297, "step": 265200 }, { "epoch": 21.96, "learning_rate": 9.06072319201995e-06, "loss": 0.0302, "step": 265300 }, { "epoch": 21.97, "learning_rate": 9.056566916043226e-06, "loss": 0.0322, "step": 265400 }, { "epoch": 21.98, "learning_rate": 9.052410640066502e-06, "loss": 0.0292, "step": 265500 }, { "epoch": 21.98, "eval_loss": 0.15081118047237396, "eval_runtime": 4264.5063, "eval_samples_per_second": 17.727, "eval_steps_per_second": 1.477, "eval_wer": 0.0516660743598856, "step": 265500 }, { "epoch": 21.99, "learning_rate": 9.048254364089776e-06, "loss": 0.0313, "step": 265600 }, { "epoch": 22.0, "learning_rate": 9.044098088113051e-06, "loss": 0.0333, "step": 265700 }, { "epoch": 22.0, "learning_rate": 9.039941812136327e-06, "loss": 0.0335, "step": 265800 }, { "epoch": 22.01, "learning_rate": 9.035785536159602e-06, "loss": 0.0266, "step": 265900 }, { "epoch": 22.02, "learning_rate": 9.031629260182878e-06, "loss": 0.0255, "step": 266000 }, { "epoch": 22.02, "eval_loss": 0.1707092672586441, "eval_runtime": 4259.3792, "eval_samples_per_second": 17.748, "eval_steps_per_second": 1.479, "eval_wer": 0.052682995872314044, "step": 266000 }, { "epoch": 22.03, "learning_rate": 9.027472984206152e-06, "loss": 0.0291, "step": 266100 }, { "epoch": 22.04, "learning_rate": 9.023316708229427e-06, "loss": 0.0232, "step": 266200 }, { "epoch": 22.04, "learning_rate": 9.019160432252703e-06, "loss": 0.0332, "step": 266300 }, { "epoch": 22.05, "learning_rate": 9.015004156275976e-06, "loss": 0.0297, "step": 266400 }, { "epoch": 22.06, "learning_rate": 9.010847880299252e-06, "loss": 0.0279, "step": 266500 }, { "epoch": 22.06, "eval_loss": 0.16501082479953766, "eval_runtime": 4266.6477, "eval_samples_per_second": 17.718, "eval_steps_per_second": 1.477, "eval_wer": 0.05280145722852734, "step": 266500 }, { "epoch": 22.07, "learning_rate": 9.006691604322527e-06, "loss": 0.0275, "step": 266600 }, { "epoch": 22.08, "learning_rate": 9.002535328345803e-06, "loss": 0.0294, "step": 266700 }, { "epoch": 22.09, "learning_rate": 8.998379052369078e-06, "loss": 0.0309, "step": 266800 }, { "epoch": 22.09, "learning_rate": 8.994222776392352e-06, "loss": 0.0278, "step": 266900 }, { "epoch": 22.1, "learning_rate": 8.990066500415628e-06, "loss": 0.0307, "step": 267000 }, { "epoch": 22.1, "eval_loss": 0.15756945312023163, "eval_runtime": 4248.1733, "eval_samples_per_second": 17.795, "eval_steps_per_second": 1.483, "eval_wer": 0.05104299839538708, "step": 267000 }, { "epoch": 22.11, "learning_rate": 8.985910224438903e-06, "loss": 0.0311, "step": 267100 }, { "epoch": 22.12, "learning_rate": 8.981753948462179e-06, "loss": 0.0283, "step": 267200 }, { "epoch": 22.13, "learning_rate": 8.977597672485454e-06, "loss": 0.0295, "step": 267300 }, { "epoch": 22.14, "learning_rate": 8.973441396508728e-06, "loss": 0.0247, "step": 267400 }, { "epoch": 22.14, "learning_rate": 8.969285120532004e-06, "loss": 0.0303, "step": 267500 }, { "epoch": 22.14, "eval_loss": 0.15765732526779175, "eval_runtime": 4241.4146, "eval_samples_per_second": 17.823, "eval_steps_per_second": 1.485, "eval_wer": 0.05195068930663184, "step": 267500 }, { "epoch": 22.15, "learning_rate": 8.96512884455528e-06, "loss": 0.0257, "step": 267600 }, { "epoch": 22.16, "learning_rate": 8.960972568578555e-06, "loss": 0.0262, "step": 267700 }, { "epoch": 22.17, "learning_rate": 8.95681629260183e-06, "loss": 0.027, "step": 267800 }, { "epoch": 22.18, "learning_rate": 8.952660016625104e-06, "loss": 0.0252, "step": 267900 }, { "epoch": 22.19, "learning_rate": 8.94850374064838e-06, "loss": 0.0283, "step": 268000 }, { "epoch": 22.19, "eval_loss": 0.16180028021335602, "eval_runtime": 4283.5513, "eval_samples_per_second": 17.648, "eval_steps_per_second": 1.471, "eval_wer": 0.05235376560959137, "step": 268000 }, { "epoch": 22.19, "learning_rate": 8.944347464671655e-06, "loss": 0.0314, "step": 268100 }, { "epoch": 22.2, "learning_rate": 8.94019118869493e-06, "loss": 0.0289, "step": 268200 }, { "epoch": 22.21, "learning_rate": 8.936034912718206e-06, "loss": 0.0325, "step": 268300 }, { "epoch": 22.22, "learning_rate": 8.931920199501248e-06, "loss": 0.0326, "step": 268400 }, { "epoch": 22.23, "learning_rate": 8.927763923524523e-06, "loss": 0.026, "step": 268500 }, { "epoch": 22.23, "eval_loss": 0.15640631318092346, "eval_runtime": 4294.1938, "eval_samples_per_second": 17.604, "eval_steps_per_second": 1.467, "eval_wer": 0.05216607359065602, "step": 268500 }, { "epoch": 22.24, "learning_rate": 8.923607647547799e-06, "loss": 0.0316, "step": 268600 }, { "epoch": 22.24, "learning_rate": 8.919451371571073e-06, "loss": 0.0314, "step": 268700 }, { "epoch": 22.25, "learning_rate": 8.915295095594348e-06, "loss": 0.0246, "step": 268800 }, { "epoch": 22.26, "learning_rate": 8.911138819617622e-06, "loss": 0.0292, "step": 268900 }, { "epoch": 22.27, "learning_rate": 8.9069825436409e-06, "loss": 0.0284, "step": 269000 }, { "epoch": 22.27, "eval_loss": 0.15949949622154236, "eval_runtime": 4259.9764, "eval_samples_per_second": 17.745, "eval_steps_per_second": 1.479, "eval_wer": 0.05390453245456545, "step": 269000 }, { "epoch": 22.28, "learning_rate": 8.902826267664175e-06, "loss": 0.0312, "step": 269100 }, { "epoch": 22.28, "learning_rate": 8.898669991687449e-06, "loss": 0.0338, "step": 269200 }, { "epoch": 22.29, "learning_rate": 8.894513715710724e-06, "loss": 0.0284, "step": 269300 }, { "epoch": 22.3, "learning_rate": 8.890357439733998e-06, "loss": 0.0317, "step": 269400 }, { "epoch": 22.31, "learning_rate": 8.886201163757274e-06, "loss": 0.0275, "step": 269500 }, { "epoch": 22.31, "eval_loss": 0.16498588025569916, "eval_runtime": 4239.7794, "eval_samples_per_second": 17.83, "eval_steps_per_second": 1.486, "eval_wer": 0.05313684132793642, "step": 269500 }, { "epoch": 22.32, "learning_rate": 8.882044887780549e-06, "loss": 0.0281, "step": 269600 }, { "epoch": 22.33, "learning_rate": 8.877888611803825e-06, "loss": 0.0279, "step": 269700 }, { "epoch": 22.33, "learning_rate": 8.8737323358271e-06, "loss": 0.0279, "step": 269800 }, { "epoch": 22.34, "learning_rate": 8.869576059850374e-06, "loss": 0.027, "step": 269900 }, { "epoch": 22.35, "learning_rate": 8.865461346633417e-06, "loss": 0.0356, "step": 270000 }, { "epoch": 22.35, "eval_loss": 0.16060078144073486, "eval_runtime": 4251.9086, "eval_samples_per_second": 17.779, "eval_steps_per_second": 1.482, "eval_wer": 0.054389147093619854, "step": 270000 }, { "epoch": 22.36, "learning_rate": 8.861305070656693e-06, "loss": 0.0278, "step": 270100 }, { "epoch": 22.37, "learning_rate": 8.857148794679967e-06, "loss": 0.032, "step": 270200 }, { "epoch": 22.38, "learning_rate": 8.852992518703242e-06, "loss": 0.0305, "step": 270300 }, { "epoch": 22.38, "learning_rate": 8.848836242726518e-06, "loss": 0.0298, "step": 270400 }, { "epoch": 22.39, "learning_rate": 8.844679966749793e-06, "loss": 0.0309, "step": 270500 }, { "epoch": 22.39, "eval_loss": 0.16174378991127014, "eval_runtime": 4253.7883, "eval_samples_per_second": 17.771, "eval_steps_per_second": 1.481, "eval_wer": 0.05469837738711171, "step": 270500 }, { "epoch": 22.4, "learning_rate": 8.840523690773069e-06, "loss": 0.0438, "step": 270600 }, { "epoch": 22.41, "learning_rate": 8.836367414796343e-06, "loss": 0.0266, "step": 270700 }, { "epoch": 22.42, "learning_rate": 8.832211138819618e-06, "loss": 0.0252, "step": 270800 }, { "epoch": 22.43, "learning_rate": 8.828054862842894e-06, "loss": 0.0254, "step": 270900 }, { "epoch": 22.43, "learning_rate": 8.82389858686617e-06, "loss": 0.0294, "step": 271000 }, { "epoch": 22.43, "eval_loss": 0.1526549756526947, "eval_runtime": 4247.2699, "eval_samples_per_second": 17.798, "eval_steps_per_second": 1.483, "eval_wer": 0.05270914967823127, "step": 271000 }, { "epoch": 22.44, "learning_rate": 8.819742310889445e-06, "loss": 0.0331, "step": 271100 }, { "epoch": 22.45, "learning_rate": 8.815586034912719e-06, "loss": 0.0369, "step": 271200 }, { "epoch": 22.46, "learning_rate": 8.811429758935994e-06, "loss": 0.0289, "step": 271300 }, { "epoch": 22.47, "learning_rate": 8.80727348295927e-06, "loss": 0.0306, "step": 271400 }, { "epoch": 22.48, "learning_rate": 8.803117206982545e-06, "loss": 0.0273, "step": 271500 }, { "epoch": 22.48, "eval_loss": 0.15402893722057343, "eval_runtime": 4238.9287, "eval_samples_per_second": 17.834, "eval_steps_per_second": 1.486, "eval_wer": 0.05223068887586327, "step": 271500 }, { "epoch": 22.48, "learning_rate": 8.79896093100582e-06, "loss": 0.0319, "step": 271600 }, { "epoch": 22.49, "learning_rate": 8.794804655029094e-06, "loss": 0.0264, "step": 271700 }, { "epoch": 22.5, "learning_rate": 8.79064837905237e-06, "loss": 0.0322, "step": 271800 }, { "epoch": 22.51, "learning_rate": 8.786492103075644e-06, "loss": 0.0346, "step": 271900 }, { "epoch": 22.52, "learning_rate": 8.78233582709892e-06, "loss": 0.0225, "step": 272000 }, { "epoch": 22.52, "eval_loss": 0.15175247192382812, "eval_runtime": 4254.1604, "eval_samples_per_second": 17.77, "eval_steps_per_second": 1.481, "eval_wer": 0.05137992095396776, "step": 272000 }, { "epoch": 22.52, "learning_rate": 8.778179551122196e-06, "loss": 0.0261, "step": 272100 }, { "epoch": 22.53, "learning_rate": 8.77402327514547e-06, "loss": 0.0259, "step": 272200 }, { "epoch": 22.54, "learning_rate": 8.769866999168746e-06, "loss": 0.0283, "step": 272300 }, { "epoch": 22.55, "learning_rate": 8.76571072319202e-06, "loss": 0.0287, "step": 272400 }, { "epoch": 22.56, "learning_rate": 8.761554447215295e-06, "loss": 0.0273, "step": 272500 }, { "epoch": 22.56, "eval_loss": 0.15179097652435303, "eval_runtime": 4266.3924, "eval_samples_per_second": 17.719, "eval_steps_per_second": 1.477, "eval_wer": 0.05211376597882157, "step": 272500 }, { "epoch": 22.57, "learning_rate": 8.75739817123857e-06, "loss": 0.0285, "step": 272600 }, { "epoch": 22.57, "learning_rate": 8.753241895261846e-06, "loss": 0.0341, "step": 272700 }, { "epoch": 22.58, "learning_rate": 8.749085619285122e-06, "loss": 0.0347, "step": 272800 }, { "epoch": 22.59, "learning_rate": 8.744929343308396e-06, "loss": 0.0273, "step": 272900 }, { "epoch": 22.6, "learning_rate": 8.740773067331671e-06, "loss": 0.0269, "step": 273000 }, { "epoch": 22.6, "eval_loss": 0.15478336811065674, "eval_runtime": 4253.9942, "eval_samples_per_second": 17.77, "eval_steps_per_second": 1.481, "eval_wer": 0.051647612849826385, "step": 273000 }, { "epoch": 22.61, "learning_rate": 8.736616791354947e-06, "loss": 0.027, "step": 273100 }, { "epoch": 22.62, "learning_rate": 8.732460515378222e-06, "loss": 0.0299, "step": 273200 }, { "epoch": 22.62, "learning_rate": 8.728304239401498e-06, "loss": 0.0298, "step": 273300 }, { "epoch": 22.63, "learning_rate": 8.724147963424771e-06, "loss": 0.0286, "step": 273400 }, { "epoch": 22.64, "learning_rate": 8.719991687448047e-06, "loss": 0.0228, "step": 273500 }, { "epoch": 22.64, "eval_loss": 0.15459494292736053, "eval_runtime": 4243.9387, "eval_samples_per_second": 17.812, "eval_steps_per_second": 1.484, "eval_wer": 0.05193684317408742, "step": 273500 }, { "epoch": 22.65, "learning_rate": 8.715835411471322e-06, "loss": 0.0324, "step": 273600 }, { "epoch": 22.66, "learning_rate": 8.711679135494598e-06, "loss": 0.0299, "step": 273700 }, { "epoch": 22.67, "learning_rate": 8.707522859517873e-06, "loss": 0.0267, "step": 273800 }, { "epoch": 22.67, "learning_rate": 8.703366583541147e-06, "loss": 0.0325, "step": 273900 }, { "epoch": 22.68, "learning_rate": 8.69925187032419e-06, "loss": 0.0265, "step": 274000 }, { "epoch": 22.68, "eval_loss": 0.15476132929325104, "eval_runtime": 4277.8522, "eval_samples_per_second": 17.671, "eval_steps_per_second": 1.473, "eval_wer": 0.052310688752786536, "step": 274000 }, { "epoch": 22.69, "learning_rate": 8.695095594347466e-06, "loss": 0.0288, "step": 274100 }, { "epoch": 22.7, "learning_rate": 8.69093931837074e-06, "loss": 0.0269, "step": 274200 }, { "epoch": 22.71, "learning_rate": 8.686783042394016e-06, "loss": 0.0298, "step": 274300 }, { "epoch": 22.72, "learning_rate": 8.68262676641729e-06, "loss": 0.0264, "step": 274400 }, { "epoch": 22.72, "learning_rate": 8.678470490440567e-06, "loss": 0.0287, "step": 274500 }, { "epoch": 22.72, "eval_loss": 0.1556084305047989, "eval_runtime": 4265.4048, "eval_samples_per_second": 17.723, "eval_steps_per_second": 1.477, "eval_wer": 0.05138761324982577, "step": 274500 }, { "epoch": 22.73, "learning_rate": 8.674314214463842e-06, "loss": 0.0267, "step": 274600 }, { "epoch": 22.74, "learning_rate": 8.670157938487116e-06, "loss": 0.0345, "step": 274700 }, { "epoch": 22.75, "learning_rate": 8.666001662510392e-06, "loss": 0.0315, "step": 274800 }, { "epoch": 22.76, "learning_rate": 8.661845386533665e-06, "loss": 0.0284, "step": 274900 }, { "epoch": 22.76, "learning_rate": 8.657689110556941e-06, "loss": 0.029, "step": 275000 }, { "epoch": 22.76, "eval_loss": 0.16708585619926453, "eval_runtime": 4293.3383, "eval_samples_per_second": 17.608, "eval_steps_per_second": 1.467, "eval_wer": 0.05252145765929591, "step": 275000 }, { "epoch": 22.77, "learning_rate": 8.653532834580216e-06, "loss": 0.0321, "step": 275100 }, { "epoch": 22.78, "learning_rate": 8.649376558603492e-06, "loss": 0.0276, "step": 275200 }, { "epoch": 22.79, "learning_rate": 8.645220282626767e-06, "loss": 0.0248, "step": 275300 }, { "epoch": 22.8, "learning_rate": 8.641064006650041e-06, "loss": 0.0264, "step": 275400 }, { "epoch": 22.81, "learning_rate": 8.636907730673317e-06, "loss": 0.0301, "step": 275500 }, { "epoch": 22.81, "eval_loss": 0.1547713726758957, "eval_runtime": 4266.6297, "eval_samples_per_second": 17.718, "eval_steps_per_second": 1.477, "eval_wer": 0.051869150970536966, "step": 275500 }, { "epoch": 22.81, "learning_rate": 8.632751454696592e-06, "loss": 0.0256, "step": 275600 }, { "epoch": 22.82, "learning_rate": 8.628595178719868e-06, "loss": 0.0263, "step": 275700 }, { "epoch": 22.83, "learning_rate": 8.624438902743143e-06, "loss": 0.026, "step": 275800 }, { "epoch": 22.84, "learning_rate": 8.620282626766417e-06, "loss": 0.0286, "step": 275900 }, { "epoch": 22.85, "learning_rate": 8.616126350789693e-06, "loss": 0.0274, "step": 276000 }, { "epoch": 22.85, "eval_loss": 0.15673740208148956, "eval_runtime": 4258.8534, "eval_samples_per_second": 17.75, "eval_steps_per_second": 1.479, "eval_wer": 0.05222761195752006, "step": 276000 }, { "epoch": 22.86, "learning_rate": 8.611970074812968e-06, "loss": 0.0287, "step": 276100 }, { "epoch": 22.86, "learning_rate": 8.607813798836244e-06, "loss": 0.0317, "step": 276200 }, { "epoch": 22.87, "learning_rate": 8.60365752285952e-06, "loss": 0.0315, "step": 276300 }, { "epoch": 22.88, "learning_rate": 8.599501246882793e-06, "loss": 0.0277, "step": 276400 }, { "epoch": 22.89, "learning_rate": 8.595344970906069e-06, "loss": 0.027, "step": 276500 }, { "epoch": 22.89, "eval_loss": 0.16561152040958405, "eval_runtime": 4253.7373, "eval_samples_per_second": 17.771, "eval_steps_per_second": 1.481, "eval_wer": 0.05103992147704388, "step": 276500 }, { "epoch": 22.9, "learning_rate": 8.591188694929344e-06, "loss": 0.0232, "step": 276600 }, { "epoch": 22.91, "learning_rate": 8.58703241895262e-06, "loss": 0.0333, "step": 276700 }, { "epoch": 22.91, "learning_rate": 8.582876142975895e-06, "loss": 0.0301, "step": 276800 }, { "epoch": 22.92, "learning_rate": 8.578719866999169e-06, "loss": 0.0314, "step": 276900 }, { "epoch": 22.93, "learning_rate": 8.574563591022444e-06, "loss": 0.0317, "step": 277000 }, { "epoch": 22.93, "eval_loss": 0.15551182627677917, "eval_runtime": 4263.1272, "eval_samples_per_second": 17.732, "eval_steps_per_second": 1.478, "eval_wer": 0.0519014586131406, "step": 277000 }, { "epoch": 22.94, "learning_rate": 8.57040731504572e-06, "loss": 0.025, "step": 277100 }, { "epoch": 22.95, "learning_rate": 8.566251039068994e-06, "loss": 0.0277, "step": 277200 }, { "epoch": 22.96, "learning_rate": 8.562094763092271e-06, "loss": 0.03, "step": 277300 }, { "epoch": 22.96, "learning_rate": 8.557980049875313e-06, "loss": 0.0295, "step": 277400 }, { "epoch": 22.97, "learning_rate": 8.553823773898587e-06, "loss": 0.0314, "step": 277500 }, { "epoch": 22.97, "eval_loss": 0.1548997014760971, "eval_runtime": 4255.5503, "eval_samples_per_second": 17.764, "eval_steps_per_second": 1.48, "eval_wer": 0.05182145873621733, "step": 277500 }, { "epoch": 22.98, "learning_rate": 8.549667497921862e-06, "loss": 0.0297, "step": 277600 }, { "epoch": 22.99, "learning_rate": 8.545511221945138e-06, "loss": 0.0284, "step": 277700 }, { "epoch": 23.0, "learning_rate": 8.541354945968413e-06, "loss": 0.0258, "step": 277800 }, { "epoch": 23.0, "learning_rate": 8.537198669991689e-06, "loss": 0.0282, "step": 277900 }, { "epoch": 23.01, "learning_rate": 8.533042394014963e-06, "loss": 0.0262, "step": 278000 }, { "epoch": 23.01, "eval_loss": 0.15159553289413452, "eval_runtime": 4254.8884, "eval_samples_per_second": 17.767, "eval_steps_per_second": 1.481, "eval_wer": 0.051426074729115805, "step": 278000 }, { "epoch": 23.02, "learning_rate": 8.528927680798006e-06, "loss": 0.0258, "step": 278100 }, { "epoch": 23.03, "learning_rate": 8.52477140482128e-06, "loss": 0.024, "step": 278200 }, { "epoch": 23.04, "learning_rate": 8.520615128844555e-06, "loss": 0.0287, "step": 278300 }, { "epoch": 23.05, "learning_rate": 8.516458852867831e-06, "loss": 0.0257, "step": 278400 }, { "epoch": 23.05, "learning_rate": 8.512302576891106e-06, "loss": 0.0258, "step": 278500 }, { "epoch": 23.05, "eval_loss": 0.1660708636045456, "eval_runtime": 4248.2444, "eval_samples_per_second": 17.794, "eval_steps_per_second": 1.483, "eval_wer": 0.053253764224978115, "step": 278500 }, { "epoch": 23.06, "learning_rate": 8.508146300914382e-06, "loss": 0.0292, "step": 278600 }, { "epoch": 23.07, "learning_rate": 8.503990024937656e-06, "loss": 0.0236, "step": 278700 }, { "epoch": 23.08, "learning_rate": 8.499833748960931e-06, "loss": 0.0296, "step": 278800 }, { "epoch": 23.09, "learning_rate": 8.495677472984207e-06, "loss": 0.0262, "step": 278900 }, { "epoch": 23.1, "learning_rate": 8.491521197007482e-06, "loss": 0.0252, "step": 279000 }, { "epoch": 23.1, "eval_loss": 0.16304850578308105, "eval_runtime": 4271.4877, "eval_samples_per_second": 17.698, "eval_steps_per_second": 1.475, "eval_wer": 0.05222915041669167, "step": 279000 }, { "epoch": 23.1, "learning_rate": 8.487364921030758e-06, "loss": 0.0307, "step": 279100 }, { "epoch": 23.11, "learning_rate": 8.483208645054032e-06, "loss": 0.0393, "step": 279200 }, { "epoch": 23.12, "learning_rate": 8.479052369077307e-06, "loss": 0.0262, "step": 279300 }, { "epoch": 23.13, "learning_rate": 8.474896093100583e-06, "loss": 0.031, "step": 279400 }, { "epoch": 23.14, "learning_rate": 8.470739817123858e-06, "loss": 0.0295, "step": 279500 }, { "epoch": 23.14, "eval_loss": 0.16331173479557037, "eval_runtime": 4268.1345, "eval_samples_per_second": 17.711, "eval_steps_per_second": 1.476, "eval_wer": 0.05402607072912195, "step": 279500 }, { "epoch": 23.15, "learning_rate": 8.466583541147134e-06, "loss": 0.026, "step": 279600 }, { "epoch": 23.15, "learning_rate": 8.462427265170408e-06, "loss": 0.0261, "step": 279700 }, { "epoch": 23.16, "learning_rate": 8.458270989193683e-06, "loss": 0.0286, "step": 279800 }, { "epoch": 23.17, "learning_rate": 8.454114713216959e-06, "loss": 0.0256, "step": 279900 }, { "epoch": 23.18, "learning_rate": 8.449958437240232e-06, "loss": 0.0261, "step": 280000 }, { "epoch": 23.18, "eval_loss": 0.16794565320014954, "eval_runtime": 4280.9153, "eval_samples_per_second": 17.659, "eval_steps_per_second": 1.472, "eval_wer": 0.05381222490426938, "step": 280000 }, { "epoch": 23.19, "learning_rate": 8.44580216126351e-06, "loss": 0.0298, "step": 280100 }, { "epoch": 23.2, "learning_rate": 8.441645885286783e-06, "loss": 0.0264, "step": 280200 }, { "epoch": 23.2, "learning_rate": 8.437489609310059e-06, "loss": 0.0282, "step": 280300 }, { "epoch": 23.21, "learning_rate": 8.433333333333334e-06, "loss": 0.0381, "step": 280400 }, { "epoch": 23.22, "learning_rate": 8.429177057356608e-06, "loss": 0.0254, "step": 280500 }, { "epoch": 23.22, "eval_loss": 0.16154265403747559, "eval_runtime": 4253.2888, "eval_samples_per_second": 17.773, "eval_steps_per_second": 1.481, "eval_wer": 0.05284299562616058, "step": 280500 }, { "epoch": 23.23, "learning_rate": 8.425020781379884e-06, "loss": 0.0236, "step": 280600 }, { "epoch": 23.24, "learning_rate": 8.42086450540316e-06, "loss": 0.024, "step": 280700 }, { "epoch": 23.25, "learning_rate": 8.416708229426435e-06, "loss": 0.029, "step": 280800 }, { "epoch": 23.25, "learning_rate": 8.41255195344971e-06, "loss": 0.0277, "step": 280900 }, { "epoch": 23.26, "learning_rate": 8.408395677472984e-06, "loss": 0.024, "step": 281000 }, { "epoch": 23.26, "eval_loss": 0.15463851392269135, "eval_runtime": 4257.7604, "eval_samples_per_second": 17.755, "eval_steps_per_second": 1.48, "eval_wer": 0.052212227365804054, "step": 281000 }, { "epoch": 23.27, "learning_rate": 8.40423940149626e-06, "loss": 0.0302, "step": 281100 }, { "epoch": 23.28, "learning_rate": 8.400083125519535e-06, "loss": 0.0278, "step": 281200 }, { "epoch": 23.29, "learning_rate": 8.39592684954281e-06, "loss": 0.0247, "step": 281300 }, { "epoch": 23.29, "learning_rate": 8.391770573566086e-06, "loss": 0.0226, "step": 281400 }, { "epoch": 23.3, "learning_rate": 8.38761429758936e-06, "loss": 0.0269, "step": 281500 }, { "epoch": 23.3, "eval_loss": 0.15263642370700836, "eval_runtime": 4308.1607, "eval_samples_per_second": 17.547, "eval_steps_per_second": 1.462, "eval_wer": 0.05172915118592125, "step": 281500 }, { "epoch": 23.31, "learning_rate": 8.383458021612636e-06, "loss": 0.0252, "step": 281600 }, { "epoch": 23.32, "learning_rate": 8.379301745635911e-06, "loss": 0.0287, "step": 281700 }, { "epoch": 23.33, "learning_rate": 8.375145469659187e-06, "loss": 0.0258, "step": 281800 }, { "epoch": 23.34, "learning_rate": 8.370989193682462e-06, "loss": 0.0247, "step": 281900 }, { "epoch": 23.34, "learning_rate": 8.366874480465504e-06, "loss": 0.0273, "step": 282000 }, { "epoch": 23.34, "eval_loss": 0.1539817750453949, "eval_runtime": 4246.5129, "eval_samples_per_second": 17.802, "eval_steps_per_second": 1.484, "eval_wer": 0.05088607555988375, "step": 282000 }, { "epoch": 23.35, "learning_rate": 8.36271820448878e-06, "loss": 0.0241, "step": 282100 }, { "epoch": 23.36, "learning_rate": 8.358561928512053e-06, "loss": 0.0307, "step": 282200 }, { "epoch": 23.37, "learning_rate": 8.354405652535329e-06, "loss": 0.0303, "step": 282300 }, { "epoch": 23.38, "learning_rate": 8.350249376558604e-06, "loss": 0.0207, "step": 282400 }, { "epoch": 23.39, "learning_rate": 8.34609310058188e-06, "loss": 0.0246, "step": 282500 }, { "epoch": 23.39, "eval_loss": 0.16459651291370392, "eval_runtime": 4259.0674, "eval_samples_per_second": 17.749, "eval_steps_per_second": 1.479, "eval_wer": 0.052761457290065705, "step": 282500 }, { "epoch": 23.39, "learning_rate": 8.341936824605155e-06, "loss": 0.0313, "step": 282600 }, { "epoch": 23.4, "learning_rate": 8.33778054862843e-06, "loss": 0.0285, "step": 282700 }, { "epoch": 23.41, "learning_rate": 8.333624272651705e-06, "loss": 0.0289, "step": 282800 }, { "epoch": 23.42, "learning_rate": 8.32946799667498e-06, "loss": 0.0289, "step": 282900 }, { "epoch": 23.43, "learning_rate": 8.325311720698254e-06, "loss": 0.0246, "step": 283000 }, { "epoch": 23.43, "eval_loss": 0.15874142944812775, "eval_runtime": 4251.5944, "eval_samples_per_second": 17.78, "eval_steps_per_second": 1.482, "eval_wer": 0.051470690045092235, "step": 283000 }, { "epoch": 23.44, "learning_rate": 8.32115544472153e-06, "loss": 0.0279, "step": 283100 }, { "epoch": 23.44, "learning_rate": 8.316999168744805e-06, "loss": 0.0253, "step": 283200 }, { "epoch": 23.45, "learning_rate": 8.31284289276808e-06, "loss": 0.032, "step": 283300 }, { "epoch": 23.46, "learning_rate": 8.308686616791356e-06, "loss": 0.0237, "step": 283400 }, { "epoch": 23.47, "learning_rate": 8.30453034081463e-06, "loss": 0.027, "step": 283500 }, { "epoch": 23.47, "eval_loss": 0.1602151244878769, "eval_runtime": 4238.9968, "eval_samples_per_second": 17.833, "eval_steps_per_second": 1.486, "eval_wer": 0.05205068915278592, "step": 283500 }, { "epoch": 23.48, "learning_rate": 8.300415627597673e-06, "loss": 0.0281, "step": 283600 }, { "epoch": 23.49, "learning_rate": 8.296259351620949e-06, "loss": 0.0325, "step": 283700 }, { "epoch": 23.49, "learning_rate": 8.292103075644223e-06, "loss": 0.0246, "step": 283800 }, { "epoch": 23.5, "learning_rate": 8.287946799667498e-06, "loss": 0.0278, "step": 283900 }, { "epoch": 23.51, "learning_rate": 8.283790523690774e-06, "loss": 0.0259, "step": 284000 }, { "epoch": 23.51, "eval_loss": 0.1659906506538391, "eval_runtime": 4264.991, "eval_samples_per_second": 17.725, "eval_steps_per_second": 1.477, "eval_wer": 0.0532414565516053, "step": 284000 }, { "epoch": 23.52, "learning_rate": 8.27963424771405e-06, "loss": 0.0267, "step": 284100 }, { "epoch": 23.53, "learning_rate": 8.275477971737325e-06, "loss": 0.0256, "step": 284200 }, { "epoch": 23.53, "learning_rate": 8.271321695760599e-06, "loss": 0.0288, "step": 284300 }, { "epoch": 23.54, "learning_rate": 8.267165419783874e-06, "loss": 0.0361, "step": 284400 }, { "epoch": 23.55, "learning_rate": 8.26300914380715e-06, "loss": 0.0223, "step": 284500 }, { "epoch": 23.55, "eval_loss": 0.16784659028053284, "eval_runtime": 4253.999, "eval_samples_per_second": 17.77, "eval_steps_per_second": 1.481, "eval_wer": 0.05385376330190261, "step": 284500 }, { "epoch": 23.56, "learning_rate": 8.258852867830425e-06, "loss": 0.026, "step": 284600 }, { "epoch": 23.57, "learning_rate": 8.2546965918537e-06, "loss": 0.0275, "step": 284700 }, { "epoch": 23.58, "learning_rate": 8.250540315876975e-06, "loss": 0.027, "step": 284800 }, { "epoch": 23.58, "learning_rate": 8.24638403990025e-06, "loss": 0.0272, "step": 284900 }, { "epoch": 23.59, "learning_rate": 8.242227763923526e-06, "loss": 0.0299, "step": 285000 }, { "epoch": 23.59, "eval_loss": 0.14981313049793243, "eval_runtime": 4274.6177, "eval_samples_per_second": 17.685, "eval_steps_per_second": 1.474, "eval_wer": 0.051526074575269885, "step": 285000 }, { "epoch": 23.6, "learning_rate": 8.238071487946801e-06, "loss": 0.0275, "step": 285100 }, { "epoch": 23.61, "learning_rate": 8.233915211970077e-06, "loss": 0.0276, "step": 285200 }, { "epoch": 23.62, "learning_rate": 8.22975893599335e-06, "loss": 0.026, "step": 285300 }, { "epoch": 23.63, "learning_rate": 8.225602660016626e-06, "loss": 0.0239, "step": 285400 }, { "epoch": 23.63, "learning_rate": 8.2214463840399e-06, "loss": 0.0271, "step": 285500 }, { "epoch": 23.63, "eval_loss": 0.1506432145833969, "eval_runtime": 4285.5135, "eval_samples_per_second": 17.64, "eval_steps_per_second": 1.47, "eval_wer": 0.05062607595988314, "step": 285500 }, { "epoch": 23.64, "learning_rate": 8.217290108063177e-06, "loss": 0.0294, "step": 285600 }, { "epoch": 23.65, "learning_rate": 8.213133832086453e-06, "loss": 0.0256, "step": 285700 }, { "epoch": 23.66, "learning_rate": 8.208977556109726e-06, "loss": 0.0284, "step": 285800 }, { "epoch": 23.67, "learning_rate": 8.204821280133002e-06, "loss": 0.0235, "step": 285900 }, { "epoch": 23.68, "learning_rate": 8.200665004156276e-06, "loss": 0.0295, "step": 286000 }, { "epoch": 23.68, "eval_loss": 0.1595706194639206, "eval_runtime": 4271.4241, "eval_samples_per_second": 17.698, "eval_steps_per_second": 1.475, "eval_wer": 0.0531199182770488, "step": 286000 }, { "epoch": 23.68, "learning_rate": 8.196508728179551e-06, "loss": 0.0241, "step": 286100 }, { "epoch": 23.69, "learning_rate": 8.192352452202827e-06, "loss": 0.0291, "step": 286200 }, { "epoch": 23.7, "learning_rate": 8.188196176226102e-06, "loss": 0.0247, "step": 286300 }, { "epoch": 23.71, "learning_rate": 8.184039900249378e-06, "loss": 0.0264, "step": 286400 }, { "epoch": 23.72, "learning_rate": 8.179883624272652e-06, "loss": 0.024, "step": 286500 }, { "epoch": 23.72, "eval_loss": 0.15698347985744476, "eval_runtime": 4273.1238, "eval_samples_per_second": 17.691, "eval_steps_per_second": 1.474, "eval_wer": 0.05228607340604091, "step": 286500 }, { "epoch": 23.73, "learning_rate": 8.175727348295927e-06, "loss": 0.0272, "step": 286600 }, { "epoch": 23.73, "learning_rate": 8.171571072319203e-06, "loss": 0.0331, "step": 286700 }, { "epoch": 23.74, "learning_rate": 8.167414796342478e-06, "loss": 0.0235, "step": 286800 }, { "epoch": 23.75, "learning_rate": 8.163258520365754e-06, "loss": 0.0286, "step": 286900 }, { "epoch": 23.76, "learning_rate": 8.159102244389027e-06, "loss": 0.025, "step": 287000 }, { "epoch": 23.76, "eval_loss": 0.15459580719470978, "eval_runtime": 4242.4992, "eval_samples_per_second": 17.819, "eval_steps_per_second": 1.485, "eval_wer": 0.05205991990781553, "step": 287000 }, { "epoch": 23.77, "learning_rate": 8.154945968412303e-06, "loss": 0.0257, "step": 287100 }, { "epoch": 23.77, "learning_rate": 8.150789692435578e-06, "loss": 0.0335, "step": 287200 }, { "epoch": 23.78, "learning_rate": 8.146633416458854e-06, "loss": 0.0261, "step": 287300 }, { "epoch": 23.79, "learning_rate": 8.14247714048213e-06, "loss": 0.0264, "step": 287400 }, { "epoch": 23.8, "learning_rate": 8.138320864505403e-06, "loss": 0.0254, "step": 287500 }, { "epoch": 23.8, "eval_loss": 0.16357311606407166, "eval_runtime": 4263.9505, "eval_samples_per_second": 17.729, "eval_steps_per_second": 1.478, "eval_wer": 0.05288453402379381, "step": 287500 }, { "epoch": 23.81, "learning_rate": 8.134164588528679e-06, "loss": 0.0271, "step": 287600 }, { "epoch": 23.82, "learning_rate": 8.130008312551954e-06, "loss": 0.031, "step": 287700 }, { "epoch": 23.82, "learning_rate": 8.12585203657523e-06, "loss": 0.0278, "step": 287800 }, { "epoch": 23.83, "learning_rate": 8.121695760598505e-06, "loss": 0.0243, "step": 287900 }, { "epoch": 23.84, "learning_rate": 8.11753948462178e-06, "loss": 0.0293, "step": 288000 }, { "epoch": 23.84, "eval_loss": 0.16621161997318268, "eval_runtime": 4252.7601, "eval_samples_per_second": 17.776, "eval_steps_per_second": 1.481, "eval_wer": 0.0527506880758645, "step": 288000 }, { "epoch": 23.85, "learning_rate": 8.113383208645055e-06, "loss": 0.0267, "step": 288100 }, { "epoch": 23.86, "learning_rate": 8.10922693266833e-06, "loss": 0.0306, "step": 288200 }, { "epoch": 23.87, "learning_rate": 8.105112219451372e-06, "loss": 0.0258, "step": 288300 }, { "epoch": 23.87, "learning_rate": 8.100955943474648e-06, "loss": 0.0249, "step": 288400 }, { "epoch": 23.88, "learning_rate": 8.096799667497921e-06, "loss": 0.0243, "step": 288500 }, { "epoch": 23.88, "eval_loss": 0.16768045723438263, "eval_runtime": 4250.2003, "eval_samples_per_second": 17.786, "eval_steps_per_second": 1.482, "eval_wer": 0.054159916677051265, "step": 288500 }, { "epoch": 23.89, "learning_rate": 8.092684954280965e-06, "loss": 0.0306, "step": 288600 }, { "epoch": 23.9, "learning_rate": 8.08852867830424e-06, "loss": 0.0252, "step": 288700 }, { "epoch": 23.91, "learning_rate": 8.084372402327514e-06, "loss": 0.0239, "step": 288800 }, { "epoch": 23.92, "learning_rate": 8.08021612635079e-06, "loss": 0.0336, "step": 288900 }, { "epoch": 23.92, "learning_rate": 8.076059850374065e-06, "loss": 0.0258, "step": 289000 }, { "epoch": 23.92, "eval_loss": 0.16301842033863068, "eval_runtime": 4255.2228, "eval_samples_per_second": 17.765, "eval_steps_per_second": 1.481, "eval_wer": 0.05230761183444333, "step": 289000 }, { "epoch": 23.93, "learning_rate": 8.07190357439734e-06, "loss": 0.0317, "step": 289100 }, { "epoch": 23.94, "learning_rate": 8.067747298420616e-06, "loss": 0.0275, "step": 289200 }, { "epoch": 23.95, "learning_rate": 8.06359102244389e-06, "loss": 0.0283, "step": 289300 }, { "epoch": 23.96, "learning_rate": 8.059434746467166e-06, "loss": 0.0272, "step": 289400 }, { "epoch": 23.97, "learning_rate": 8.055278470490441e-06, "loss": 0.0308, "step": 289500 }, { "epoch": 23.97, "eval_loss": 0.16474178433418274, "eval_runtime": 4252.6864, "eval_samples_per_second": 17.776, "eval_steps_per_second": 1.481, "eval_wer": 0.05408607063681441, "step": 289500 }, { "epoch": 23.97, "learning_rate": 8.051122194513717e-06, "loss": 0.0313, "step": 289600 }, { "epoch": 23.98, "learning_rate": 8.046965918536992e-06, "loss": 0.0252, "step": 289700 }, { "epoch": 23.99, "learning_rate": 8.042809642560266e-06, "loss": 0.0277, "step": 289800 }, { "epoch": 24.0, "learning_rate": 8.038653366583542e-06, "loss": 0.0226, "step": 289900 }, { "epoch": 24.01, "learning_rate": 8.034497090606817e-06, "loss": 0.0258, "step": 290000 }, { "epoch": 24.01, "eval_loss": 0.17375166714191437, "eval_runtime": 4247.4153, "eval_samples_per_second": 17.798, "eval_steps_per_second": 1.483, "eval_wer": 0.0532368411740905, "step": 290000 }, { "epoch": 24.01, "learning_rate": 8.030340814630093e-06, "loss": 0.0277, "step": 290100 }, { "epoch": 24.02, "learning_rate": 8.026184538653368e-06, "loss": 0.0265, "step": 290200 }, { "epoch": 24.03, "learning_rate": 8.022028262676642e-06, "loss": 0.0272, "step": 290300 }, { "epoch": 24.04, "learning_rate": 8.017871986699917e-06, "loss": 0.0278, "step": 290400 }, { "epoch": 24.05, "learning_rate": 8.013715710723193e-06, "loss": 0.0209, "step": 290500 }, { "epoch": 24.05, "eval_loss": 0.17175936698913574, "eval_runtime": 4261.5883, "eval_samples_per_second": 17.739, "eval_steps_per_second": 1.478, "eval_wer": 0.054036839943323166, "step": 290500 }, { "epoch": 24.06, "learning_rate": 8.009559434746468e-06, "loss": 0.027, "step": 290600 }, { "epoch": 24.06, "learning_rate": 8.005403158769744e-06, "loss": 0.0233, "step": 290700 }, { "epoch": 24.07, "learning_rate": 8.001246882793018e-06, "loss": 0.0278, "step": 290800 }, { "epoch": 24.08, "learning_rate": 7.997090606816293e-06, "loss": 0.0304, "step": 290900 }, { "epoch": 24.09, "learning_rate": 7.992934330839567e-06, "loss": 0.0253, "step": 291000 }, { "epoch": 24.09, "eval_loss": 0.17227056622505188, "eval_runtime": 4253.1668, "eval_samples_per_second": 17.774, "eval_steps_per_second": 1.481, "eval_wer": 0.05425376268651894, "step": 291000 }, { "epoch": 24.1, "learning_rate": 7.988778054862843e-06, "loss": 0.0268, "step": 291100 }, { "epoch": 24.11, "learning_rate": 7.98462177888612e-06, "loss": 0.0233, "step": 291200 }, { "epoch": 24.11, "learning_rate": 7.980465502909394e-06, "loss": 0.0246, "step": 291300 }, { "epoch": 24.12, "learning_rate": 7.97630922693267e-06, "loss": 0.0277, "step": 291400 }, { "epoch": 24.13, "learning_rate": 7.972152950955943e-06, "loss": 0.0275, "step": 291500 }, { "epoch": 24.13, "eval_loss": 0.16872729361057281, "eval_runtime": 4284.5435, "eval_samples_per_second": 17.644, "eval_steps_per_second": 1.47, "eval_wer": 0.053446071621428275, "step": 291500 }, { "epoch": 24.14, "learning_rate": 7.967996674979219e-06, "loss": 0.0233, "step": 291600 }, { "epoch": 24.15, "learning_rate": 7.963840399002494e-06, "loss": 0.0278, "step": 291700 }, { "epoch": 24.16, "learning_rate": 7.95968412302577e-06, "loss": 0.0264, "step": 291800 }, { "epoch": 24.16, "learning_rate": 7.955527847049045e-06, "loss": 0.0217, "step": 291900 }, { "epoch": 24.17, "learning_rate": 7.951371571072319e-06, "loss": 0.0297, "step": 292000 }, { "epoch": 24.17, "eval_loss": 0.16061381995677948, "eval_runtime": 4249.2488, "eval_samples_per_second": 17.79, "eval_steps_per_second": 1.483, "eval_wer": 0.05290607245219623, "step": 292000 }, { "epoch": 24.18, "learning_rate": 7.947215295095594e-06, "loss": 0.0261, "step": 292100 }, { "epoch": 24.19, "learning_rate": 7.94305901911887e-06, "loss": 0.0264, "step": 292200 }, { "epoch": 24.2, "learning_rate": 7.938902743142145e-06, "loss": 0.023, "step": 292300 }, { "epoch": 24.21, "learning_rate": 7.934746467165421e-06, "loss": 0.028, "step": 292400 }, { "epoch": 24.21, "learning_rate": 7.930590191188695e-06, "loss": 0.0321, "step": 292500 }, { "epoch": 24.21, "eval_loss": 0.15714974701404572, "eval_runtime": 4250.869, "eval_samples_per_second": 17.783, "eval_steps_per_second": 1.482, "eval_wer": 0.05388453248533464, "step": 292500 }, { "epoch": 24.22, "learning_rate": 7.92643391521197e-06, "loss": 0.0282, "step": 292600 }, { "epoch": 24.23, "learning_rate": 7.922277639235246e-06, "loss": 0.0246, "step": 292700 }, { "epoch": 24.24, "learning_rate": 7.918121363258521e-06, "loss": 0.029, "step": 292800 }, { "epoch": 24.25, "learning_rate": 7.913965087281797e-06, "loss": 0.0255, "step": 292900 }, { "epoch": 24.25, "learning_rate": 7.90980881130507e-06, "loss": 0.0258, "step": 293000 }, { "epoch": 24.25, "eval_loss": 0.1592789739370346, "eval_runtime": 4267.1968, "eval_samples_per_second": 17.715, "eval_steps_per_second": 1.476, "eval_wer": 0.053538379171724354, "step": 293000 }, { "epoch": 24.26, "learning_rate": 7.905652535328346e-06, "loss": 0.0292, "step": 293100 }, { "epoch": 24.27, "learning_rate": 7.901496259351622e-06, "loss": 0.0246, "step": 293200 }, { "epoch": 24.28, "learning_rate": 7.897339983374897e-06, "loss": 0.0244, "step": 293300 }, { "epoch": 24.29, "learning_rate": 7.893183707398173e-06, "loss": 0.0244, "step": 293400 }, { "epoch": 24.3, "learning_rate": 7.889027431421447e-06, "loss": 0.0327, "step": 293500 }, { "epoch": 24.3, "eval_loss": 0.16882435977458954, "eval_runtime": 4247.0109, "eval_samples_per_second": 17.8, "eval_steps_per_second": 1.483, "eval_wer": 0.053758378833263336, "step": 293500 }, { "epoch": 24.3, "learning_rate": 7.884871155444722e-06, "loss": 0.0256, "step": 293600 }, { "epoch": 24.31, "learning_rate": 7.880714879467998e-06, "loss": 0.0315, "step": 293700 }, { "epoch": 24.32, "learning_rate": 7.876558603491271e-06, "loss": 0.0292, "step": 293800 }, { "epoch": 24.33, "learning_rate": 7.872402327514547e-06, "loss": 0.0276, "step": 293900 }, { "epoch": 24.34, "learning_rate": 7.868246051537822e-06, "loss": 0.0285, "step": 294000 }, { "epoch": 24.34, "eval_loss": 0.17667077481746674, "eval_runtime": 4245.4587, "eval_samples_per_second": 17.806, "eval_steps_per_second": 1.484, "eval_wer": 0.053601455997760006, "step": 294000 }, { "epoch": 24.35, "learning_rate": 7.864089775561098e-06, "loss": 0.0237, "step": 294100 }, { "epoch": 24.35, "learning_rate": 7.859933499584374e-06, "loss": 0.0235, "step": 294200 }, { "epoch": 24.36, "learning_rate": 7.855777223607647e-06, "loss": 0.0235, "step": 294300 }, { "epoch": 24.37, "learning_rate": 7.851620947630923e-06, "loss": 0.0343, "step": 294400 }, { "epoch": 24.38, "learning_rate": 7.847464671654198e-06, "loss": 0.0221, "step": 294500 }, { "epoch": 24.38, "eval_loss": 0.17418146133422852, "eval_runtime": 4265.8012, "eval_samples_per_second": 17.721, "eval_steps_per_second": 1.477, "eval_wer": 0.05325837960249292, "step": 294500 }, { "epoch": 24.39, "learning_rate": 7.843308395677474e-06, "loss": 0.0292, "step": 294600 }, { "epoch": 24.4, "learning_rate": 7.83915211970075e-06, "loss": 0.026, "step": 294700 }, { "epoch": 24.4, "learning_rate": 7.834995843724023e-06, "loss": 0.0247, "step": 294800 }, { "epoch": 24.41, "learning_rate": 7.830839567747299e-06, "loss": 0.025, "step": 294900 }, { "epoch": 24.42, "learning_rate": 7.82672485453034e-06, "loss": 0.0302, "step": 295000 }, { "epoch": 24.42, "eval_loss": 0.16957047581672668, "eval_runtime": 4243.394, "eval_samples_per_second": 17.815, "eval_steps_per_second": 1.485, "eval_wer": 0.05382760949598539, "step": 295000 }, { "epoch": 24.43, "learning_rate": 7.822568578553616e-06, "loss": 0.0207, "step": 295100 }, { "epoch": 24.44, "learning_rate": 7.818412302576892e-06, "loss": 0.0295, "step": 295200 }, { "epoch": 24.45, "learning_rate": 7.814256026600167e-06, "loss": 0.0333, "step": 295300 }, { "epoch": 24.45, "learning_rate": 7.810099750623443e-06, "loss": 0.0204, "step": 295400 }, { "epoch": 24.46, "learning_rate": 7.805943474646716e-06, "loss": 0.0298, "step": 295500 }, { "epoch": 24.46, "eval_loss": 0.16215559840202332, "eval_runtime": 4251.4266, "eval_samples_per_second": 17.781, "eval_steps_per_second": 1.482, "eval_wer": 0.051770689583554484, "step": 295500 }, { "epoch": 24.47, "learning_rate": 7.80182876142976e-06, "loss": 0.0275, "step": 295600 }, { "epoch": 24.48, "learning_rate": 7.797672485453036e-06, "loss": 0.0241, "step": 295700 }, { "epoch": 24.49, "learning_rate": 7.79351620947631e-06, "loss": 0.023, "step": 295800 }, { "epoch": 24.5, "learning_rate": 7.789359933499585e-06, "loss": 0.0234, "step": 295900 }, { "epoch": 24.5, "learning_rate": 7.78520365752286e-06, "loss": 0.0233, "step": 296000 }, { "epoch": 24.5, "eval_loss": 0.16756777465343475, "eval_runtime": 4250.5762, "eval_samples_per_second": 17.785, "eval_steps_per_second": 1.482, "eval_wer": 0.05380299414923977, "step": 296000 }, { "epoch": 24.51, "learning_rate": 7.781047381546136e-06, "loss": 0.0253, "step": 296100 }, { "epoch": 24.52, "learning_rate": 7.776891105569411e-06, "loss": 0.0279, "step": 296200 }, { "epoch": 24.53, "learning_rate": 7.772734829592685e-06, "loss": 0.0226, "step": 296300 }, { "epoch": 24.54, "learning_rate": 7.76857855361596e-06, "loss": 0.0341, "step": 296400 }, { "epoch": 24.54, "learning_rate": 7.764422277639236e-06, "loss": 0.0272, "step": 296500 }, { "epoch": 24.54, "eval_loss": 0.1701800376176834, "eval_runtime": 4251.9693, "eval_samples_per_second": 17.779, "eval_steps_per_second": 1.482, "eval_wer": 0.05344914853977148, "step": 296500 }, { "epoch": 24.55, "learning_rate": 7.76026600166251e-06, "loss": 0.0291, "step": 296600 }, { "epoch": 24.56, "learning_rate": 7.756109725685787e-06, "loss": 0.0264, "step": 296700 }, { "epoch": 24.57, "learning_rate": 7.751953449709061e-06, "loss": 0.0261, "step": 296800 }, { "epoch": 24.58, "learning_rate": 7.747797173732337e-06, "loss": 0.0275, "step": 296900 }, { "epoch": 24.59, "learning_rate": 7.743640897755612e-06, "loss": 0.0277, "step": 297000 }, { "epoch": 24.59, "eval_loss": 0.16309253871440887, "eval_runtime": 4268.6173, "eval_samples_per_second": 17.709, "eval_steps_per_second": 1.476, "eval_wer": 0.05325530268414971, "step": 297000 }, { "epoch": 24.59, "learning_rate": 7.739484621778886e-06, "loss": 0.0284, "step": 297100 }, { "epoch": 24.6, "learning_rate": 7.735328345802161e-06, "loss": 0.0263, "step": 297200 }, { "epoch": 24.61, "learning_rate": 7.731172069825437e-06, "loss": 0.0267, "step": 297300 }, { "epoch": 24.62, "learning_rate": 7.727015793848713e-06, "loss": 0.0222, "step": 297400 }, { "epoch": 24.63, "learning_rate": 7.722859517871988e-06, "loss": 0.0225, "step": 297500 }, { "epoch": 24.63, "eval_loss": 0.16012592613697052, "eval_runtime": 4244.9495, "eval_samples_per_second": 17.808, "eval_steps_per_second": 1.484, "eval_wer": 0.05089530631491336, "step": 297500 }, { "epoch": 24.64, "learning_rate": 7.718703241895262e-06, "loss": 0.0236, "step": 297600 }, { "epoch": 24.64, "learning_rate": 7.714546965918537e-06, "loss": 0.0272, "step": 297700 }, { "epoch": 24.65, "learning_rate": 7.710390689941813e-06, "loss": 0.0287, "step": 297800 }, { "epoch": 24.66, "learning_rate": 7.706275976724855e-06, "loss": 0.0268, "step": 297900 }, { "epoch": 24.67, "learning_rate": 7.70211970074813e-06, "loss": 0.025, "step": 298000 }, { "epoch": 24.67, "eval_loss": 0.15973058342933655, "eval_runtime": 4248.8222, "eval_samples_per_second": 17.792, "eval_steps_per_second": 1.483, "eval_wer": 0.05186607405219377, "step": 298000 }, { "epoch": 24.68, "learning_rate": 7.697963424771406e-06, "loss": 0.0225, "step": 298100 }, { "epoch": 24.69, "learning_rate": 7.693807148794681e-06, "loss": 0.0265, "step": 298200 }, { "epoch": 24.69, "learning_rate": 7.689650872817955e-06, "loss": 0.0288, "step": 298300 }, { "epoch": 24.7, "learning_rate": 7.68549459684123e-06, "loss": 0.0275, "step": 298400 }, { "epoch": 24.71, "learning_rate": 7.681338320864506e-06, "loss": 0.0275, "step": 298500 }, { "epoch": 24.71, "eval_loss": 0.151408851146698, "eval_runtime": 4279.3065, "eval_samples_per_second": 17.665, "eval_steps_per_second": 1.472, "eval_wer": 0.05171376659420524, "step": 298500 }, { "epoch": 24.72, "learning_rate": 7.677182044887782e-06, "loss": 0.0239, "step": 298600 }, { "epoch": 24.73, "learning_rate": 7.673025768911057e-06, "loss": 0.026, "step": 298700 }, { "epoch": 24.74, "learning_rate": 7.668869492934331e-06, "loss": 0.0266, "step": 298800 }, { "epoch": 24.74, "learning_rate": 7.664713216957606e-06, "loss": 0.0284, "step": 298900 }, { "epoch": 24.75, "learning_rate": 7.660556940980882e-06, "loss": 0.029, "step": 299000 }, { "epoch": 24.75, "eval_loss": 0.15697653591632843, "eval_runtime": 4268.837, "eval_samples_per_second": 17.709, "eval_steps_per_second": 1.476, "eval_wer": 0.05150761306521067, "step": 299000 }, { "epoch": 24.76, "learning_rate": 7.656400665004158e-06, "loss": 0.0287, "step": 299100 }, { "epoch": 24.77, "learning_rate": 7.652244389027433e-06, "loss": 0.0284, "step": 299200 }, { "epoch": 24.78, "learning_rate": 7.648088113050707e-06, "loss": 0.0308, "step": 299300 }, { "epoch": 24.78, "learning_rate": 7.643931837073982e-06, "loss": 0.0232, "step": 299400 }, { "epoch": 24.79, "learning_rate": 7.639775561097258e-06, "loss": 0.0271, "step": 299500 }, { "epoch": 24.79, "eval_loss": 0.15025997161865234, "eval_runtime": 4261.9859, "eval_samples_per_second": 17.737, "eval_steps_per_second": 1.478, "eval_wer": 0.05092146012083058, "step": 299500 }, { "epoch": 24.8, "learning_rate": 7.635619285120532e-06, "loss": 0.0241, "step": 299600 }, { "epoch": 24.81, "learning_rate": 7.631463009143807e-06, "loss": 0.0281, "step": 299700 }, { "epoch": 24.82, "learning_rate": 7.627306733167083e-06, "loss": 0.0248, "step": 299800 }, { "epoch": 24.83, "learning_rate": 7.623150457190358e-06, "loss": 0.026, "step": 299900 }, { "epoch": 24.83, "learning_rate": 7.618994181213634e-06, "loss": 0.0218, "step": 300000 }, { "epoch": 24.83, "eval_loss": 0.16333267092704773, "eval_runtime": 4240.0181, "eval_samples_per_second": 17.829, "eval_steps_per_second": 1.486, "eval_wer": 0.05224299654923608, "step": 300000 }, { "epoch": 24.84, "learning_rate": 7.6148379052369084e-06, "loss": 0.0238, "step": 300100 }, { "epoch": 24.85, "learning_rate": 7.610681629260184e-06, "loss": 0.0244, "step": 300200 }, { "epoch": 24.86, "learning_rate": 7.606525353283459e-06, "loss": 0.0225, "step": 300300 }, { "epoch": 24.87, "learning_rate": 7.602369077306734e-06, "loss": 0.0243, "step": 300400 }, { "epoch": 24.88, "learning_rate": 7.59821280133001e-06, "loss": 0.027, "step": 300500 }, { "epoch": 24.88, "eval_loss": 0.1585514098405838, "eval_runtime": 4243.0002, "eval_samples_per_second": 17.816, "eval_steps_per_second": 1.485, "eval_wer": 0.05171530505337684, "step": 300500 }, { "epoch": 24.88, "learning_rate": 7.5940565253532835e-06, "loss": 0.0251, "step": 300600 }, { "epoch": 24.89, "learning_rate": 7.58990024937656e-06, "loss": 0.0248, "step": 300700 }, { "epoch": 24.9, "learning_rate": 7.585743973399834e-06, "loss": 0.0245, "step": 300800 }, { "epoch": 24.91, "learning_rate": 7.581587697423109e-06, "loss": 0.0255, "step": 300900 }, { "epoch": 24.92, "learning_rate": 7.577431421446385e-06, "loss": 0.0223, "step": 301000 }, { "epoch": 24.92, "eval_loss": 0.1582973748445511, "eval_runtime": 4241.3168, "eval_samples_per_second": 17.823, "eval_steps_per_second": 1.485, "eval_wer": 0.05132915180130492, "step": 301000 }, { "epoch": 24.93, "learning_rate": 7.573275145469659e-06, "loss": 0.022, "step": 301100 }, { "epoch": 24.93, "learning_rate": 7.569118869492935e-06, "loss": 0.0246, "step": 301200 }, { "epoch": 24.94, "learning_rate": 7.5649625935162095e-06, "loss": 0.0286, "step": 301300 }, { "epoch": 24.95, "learning_rate": 7.560806317539485e-06, "loss": 0.0261, "step": 301400 }, { "epoch": 24.96, "learning_rate": 7.556650041562761e-06, "loss": 0.028, "step": 301500 }, { "epoch": 24.96, "eval_loss": 0.15908832848072052, "eval_runtime": 4251.3328, "eval_samples_per_second": 17.781, "eval_steps_per_second": 1.482, "eval_wer": 0.05174299731846566, "step": 301500 }, { "epoch": 24.97, "learning_rate": 7.552493765586035e-06, "loss": 0.0295, "step": 301600 }, { "epoch": 24.98, "learning_rate": 7.548337489609311e-06, "loss": 0.0226, "step": 301700 }, { "epoch": 24.98, "learning_rate": 7.5441812136325854e-06, "loss": 0.0235, "step": 301800 }, { "epoch": 24.99, "learning_rate": 7.540024937655861e-06, "loss": 0.0322, "step": 301900 }, { "epoch": 25.0, "learning_rate": 7.5358686616791365e-06, "loss": 0.0273, "step": 302000 }, { "epoch": 25.0, "eval_loss": 0.1565355509519577, "eval_runtime": 4308.997, "eval_samples_per_second": 17.544, "eval_steps_per_second": 1.462, "eval_wer": 0.050129153647455925, "step": 302000 }, { "epoch": 25.01, "learning_rate": 7.531712385702411e-06, "loss": 0.0265, "step": 302100 }, { "epoch": 25.02, "learning_rate": 7.527556109725687e-06, "loss": 0.0276, "step": 302200 }, { "epoch": 25.02, "learning_rate": 7.523399833748961e-06, "loss": 0.0268, "step": 302300 }, { "epoch": 25.03, "learning_rate": 7.519285120532004e-06, "loss": 0.0265, "step": 302400 }, { "epoch": 25.04, "learning_rate": 7.5151288445552795e-06, "loss": 0.0324, "step": 302500 }, { "epoch": 25.04, "eval_loss": 0.15976493060588837, "eval_runtime": 4258.6202, "eval_samples_per_second": 17.751, "eval_steps_per_second": 1.479, "eval_wer": 0.052504534608408295, "step": 302500 }, { "epoch": 25.05, "learning_rate": 7.510972568578554e-06, "loss": 0.0229, "step": 302600 }, { "epoch": 25.06, "learning_rate": 7.50681629260183e-06, "loss": 0.0297, "step": 302700 }, { "epoch": 25.07, "learning_rate": 7.502660016625104e-06, "loss": 0.0245, "step": 302800 }, { "epoch": 25.07, "learning_rate": 7.49850374064838e-06, "loss": 0.0241, "step": 302900 }, { "epoch": 25.08, "learning_rate": 7.494347464671655e-06, "loss": 0.0204, "step": 303000 }, { "epoch": 25.08, "eval_loss": 0.17353515326976776, "eval_runtime": 4277.9677, "eval_samples_per_second": 17.671, "eval_steps_per_second": 1.473, "eval_wer": 0.05205838144864393, "step": 303000 }, { "epoch": 25.09, "learning_rate": 7.49019118869493e-06, "loss": 0.0242, "step": 303100 }, { "epoch": 25.1, "learning_rate": 7.486034912718206e-06, "loss": 0.0223, "step": 303200 }, { "epoch": 25.11, "learning_rate": 7.481878636741479e-06, "loss": 0.0248, "step": 303300 }, { "epoch": 25.12, "learning_rate": 7.477722360764755e-06, "loss": 0.0229, "step": 303400 }, { "epoch": 25.12, "learning_rate": 7.473566084788031e-06, "loss": 0.0254, "step": 303500 }, { "epoch": 25.12, "eval_loss": 0.16294944286346436, "eval_runtime": 4250.6252, "eval_samples_per_second": 17.784, "eval_steps_per_second": 1.482, "eval_wer": 0.051961458520833045, "step": 303500 }, { "epoch": 25.13, "learning_rate": 7.469409808811305e-06, "loss": 0.0225, "step": 303600 }, { "epoch": 25.14, "learning_rate": 7.465253532834581e-06, "loss": 0.0272, "step": 303700 }, { "epoch": 25.15, "learning_rate": 7.461097256857855e-06, "loss": 0.0263, "step": 303800 }, { "epoch": 25.16, "learning_rate": 7.456940980881131e-06, "loss": 0.0324, "step": 303900 }, { "epoch": 25.17, "learning_rate": 7.452784704904406e-06, "loss": 0.0284, "step": 304000 }, { "epoch": 25.17, "eval_loss": 0.16522598266601562, "eval_runtime": 4245.8797, "eval_samples_per_second": 17.804, "eval_steps_per_second": 1.484, "eval_wer": 0.05139222862734057, "step": 304000 }, { "epoch": 25.17, "learning_rate": 7.448628428927681e-06, "loss": 0.0282, "step": 304100 }, { "epoch": 25.18, "learning_rate": 7.4444721529509565e-06, "loss": 0.0258, "step": 304200 }, { "epoch": 25.19, "learning_rate": 7.440315876974231e-06, "loss": 0.0246, "step": 304300 }, { "epoch": 25.2, "learning_rate": 7.436159600997507e-06, "loss": 0.024, "step": 304400 }, { "epoch": 25.21, "learning_rate": 7.432003325020782e-06, "loss": 0.0206, "step": 304500 }, { "epoch": 25.21, "eval_loss": 0.17054599523544312, "eval_runtime": 4255.6689, "eval_samples_per_second": 17.763, "eval_steps_per_second": 1.48, "eval_wer": 0.05174915115515207, "step": 304500 }, { "epoch": 25.22, "learning_rate": 7.427847049044057e-06, "loss": 0.0299, "step": 304600 }, { "epoch": 25.22, "learning_rate": 7.423690773067332e-06, "loss": 0.0274, "step": 304700 }, { "epoch": 25.23, "learning_rate": 7.419534497090607e-06, "loss": 0.0238, "step": 304800 }, { "epoch": 25.24, "learning_rate": 7.415461346633417e-06, "loss": 0.0278, "step": 304900 }, { "epoch": 25.25, "learning_rate": 7.411305070656692e-06, "loss": 0.0259, "step": 305000 }, { "epoch": 25.25, "eval_loss": 0.16055528819561005, "eval_runtime": 4243.8025, "eval_samples_per_second": 17.813, "eval_steps_per_second": 1.485, "eval_wer": 0.05099069078355264, "step": 305000 }, { "epoch": 25.26, "learning_rate": 7.407148794679967e-06, "loss": 0.0266, "step": 305100 }, { "epoch": 25.26, "learning_rate": 7.402992518703243e-06, "loss": 0.0251, "step": 305200 }, { "epoch": 25.27, "learning_rate": 7.398836242726518e-06, "loss": 0.0284, "step": 305300 }, { "epoch": 25.28, "learning_rate": 7.394679966749793e-06, "loss": 0.0278, "step": 305400 }, { "epoch": 25.29, "learning_rate": 7.390523690773068e-06, "loss": 0.0229, "step": 305500 }, { "epoch": 25.29, "eval_loss": 0.1652679443359375, "eval_runtime": 4249.0631, "eval_samples_per_second": 17.791, "eval_steps_per_second": 1.483, "eval_wer": 0.05037223019656893, "step": 305500 }, { "epoch": 25.3, "learning_rate": 7.386367414796343e-06, "loss": 0.0231, "step": 305600 }, { "epoch": 25.31, "learning_rate": 7.3822111388196185e-06, "loss": 0.024, "step": 305700 }, { "epoch": 25.31, "learning_rate": 7.378054862842894e-06, "loss": 0.0257, "step": 305800 }, { "epoch": 25.32, "learning_rate": 7.373898586866169e-06, "loss": 0.0247, "step": 305900 }, { "epoch": 25.33, "learning_rate": 7.369742310889444e-06, "loss": 0.0204, "step": 306000 }, { "epoch": 25.33, "eval_loss": 0.16478034853935242, "eval_runtime": 4247.3753, "eval_samples_per_second": 17.798, "eval_steps_per_second": 1.483, "eval_wer": 0.049259924215501204, "step": 306000 }, { "epoch": 25.34, "learning_rate": 7.365586034912718e-06, "loss": 0.0253, "step": 306100 }, { "epoch": 25.35, "learning_rate": 7.361429758935994e-06, "loss": 0.0262, "step": 306200 }, { "epoch": 25.36, "learning_rate": 7.35727348295927e-06, "loss": 0.0258, "step": 306300 }, { "epoch": 25.36, "learning_rate": 7.353117206982544e-06, "loss": 0.0289, "step": 306400 }, { "epoch": 25.37, "learning_rate": 7.348960931005819e-06, "loss": 0.0304, "step": 306500 }, { "epoch": 25.37, "eval_loss": 0.1615319550037384, "eval_runtime": 4260.8478, "eval_samples_per_second": 17.742, "eval_steps_per_second": 1.479, "eval_wer": 0.04914607823680271, "step": 306500 }, { "epoch": 25.38, "learning_rate": 7.344804655029094e-06, "loss": 0.0226, "step": 306600 }, { "epoch": 25.39, "learning_rate": 7.340648379052369e-06, "loss": 0.0283, "step": 306700 }, { "epoch": 25.4, "learning_rate": 7.336492103075645e-06, "loss": 0.0298, "step": 306800 }, { "epoch": 25.41, "learning_rate": 7.33233582709892e-06, "loss": 0.0211, "step": 306900 }, { "epoch": 25.41, "learning_rate": 7.328179551122195e-06, "loss": 0.0267, "step": 307000 }, { "epoch": 25.41, "eval_loss": 0.15540775656700134, "eval_runtime": 4247.4516, "eval_samples_per_second": 17.798, "eval_steps_per_second": 1.483, "eval_wer": 0.05029069186047406, "step": 307000 }, { "epoch": 25.42, "learning_rate": 7.32402327514547e-06, "loss": 0.0258, "step": 307100 }, { "epoch": 25.43, "learning_rate": 7.319866999168745e-06, "loss": 0.0251, "step": 307200 }, { "epoch": 25.44, "learning_rate": 7.315710723192021e-06, "loss": 0.0287, "step": 307300 }, { "epoch": 25.45, "learning_rate": 7.3115544472152955e-06, "loss": 0.027, "step": 307400 }, { "epoch": 25.46, "learning_rate": 7.307398171238571e-06, "loss": 0.0247, "step": 307500 }, { "epoch": 25.46, "eval_loss": 0.15595640242099762, "eval_runtime": 4259.137, "eval_samples_per_second": 17.749, "eval_steps_per_second": 1.479, "eval_wer": 0.049013770748045006, "step": 307500 }, { "epoch": 25.46, "learning_rate": 7.303241895261846e-06, "loss": 0.0243, "step": 307600 }, { "epoch": 25.47, "learning_rate": 7.299085619285121e-06, "loss": 0.0233, "step": 307700 }, { "epoch": 25.48, "learning_rate": 7.294929343308397e-06, "loss": 0.0218, "step": 307800 }, { "epoch": 25.49, "learning_rate": 7.290773067331671e-06, "loss": 0.0229, "step": 307900 }, { "epoch": 25.5, "learning_rate": 7.286616791354947e-06, "loss": 0.0288, "step": 308000 }, { "epoch": 25.5, "eval_loss": 0.17048393189907074, "eval_runtime": 4240.0376, "eval_samples_per_second": 17.829, "eval_steps_per_second": 1.486, "eval_wer": 0.051776843420240894, "step": 308000 }, { "epoch": 25.5, "learning_rate": 7.2824605153782216e-06, "loss": 0.02, "step": 308100 }, { "epoch": 25.51, "learning_rate": 7.278304239401497e-06, "loss": 0.0232, "step": 308200 }, { "epoch": 25.52, "learning_rate": 7.274147963424773e-06, "loss": 0.0268, "step": 308300 }, { "epoch": 25.53, "learning_rate": 7.269991687448047e-06, "loss": 0.0284, "step": 308400 }, { "epoch": 25.54, "learning_rate": 7.265835411471323e-06, "loss": 0.0289, "step": 308500 }, { "epoch": 25.54, "eval_loss": 0.15603603422641754, "eval_runtime": 4263.0198, "eval_samples_per_second": 17.733, "eval_steps_per_second": 1.478, "eval_wer": 0.05104761377290189, "step": 308500 }, { "epoch": 25.55, "learning_rate": 7.261679135494597e-06, "loss": 0.0222, "step": 308600 }, { "epoch": 25.55, "learning_rate": 7.257522859517873e-06, "loss": 0.0258, "step": 308700 }, { "epoch": 25.56, "learning_rate": 7.2533665835411485e-06, "loss": 0.0223, "step": 308800 }, { "epoch": 25.57, "learning_rate": 7.249210307564422e-06, "loss": 0.0217, "step": 308900 }, { "epoch": 25.58, "learning_rate": 7.245054031587699e-06, "loss": 0.0213, "step": 309000 }, { "epoch": 25.58, "eval_loss": 0.16434399783611298, "eval_runtime": 4263.8686, "eval_samples_per_second": 17.729, "eval_steps_per_second": 1.478, "eval_wer": 0.0520229968876971, "step": 309000 }, { "epoch": 25.59, "learning_rate": 7.2408977556109725e-06, "loss": 0.0244, "step": 309100 }, { "epoch": 25.6, "learning_rate": 7.236741479634248e-06, "loss": 0.0238, "step": 309200 }, { "epoch": 25.6, "learning_rate": 7.2325852036575235e-06, "loss": 0.0268, "step": 309300 }, { "epoch": 25.61, "learning_rate": 7.228428927680798e-06, "loss": 0.0307, "step": 309400 }, { "epoch": 25.62, "learning_rate": 7.224272651704074e-06, "loss": 0.03, "step": 309500 }, { "epoch": 25.62, "eval_loss": 0.15077020227909088, "eval_runtime": 4256.2452, "eval_samples_per_second": 17.761, "eval_steps_per_second": 1.48, "eval_wer": 0.04996453851609459, "step": 309500 }, { "epoch": 25.63, "learning_rate": 7.220116375727348e-06, "loss": 0.0308, "step": 309600 }, { "epoch": 25.64, "learning_rate": 7.215960099750624e-06, "loss": 0.0254, "step": 309700 }, { "epoch": 25.65, "learning_rate": 7.211803823773899e-06, "loss": 0.0237, "step": 309800 }, { "epoch": 25.65, "learning_rate": 7.207647547797174e-06, "loss": 0.0298, "step": 309900 }, { "epoch": 25.66, "learning_rate": 7.20349127182045e-06, "loss": 0.0255, "step": 310000 }, { "epoch": 25.66, "eval_loss": 0.15968406200408936, "eval_runtime": 4247.5154, "eval_samples_per_second": 17.797, "eval_steps_per_second": 1.483, "eval_wer": 0.05038761478828494, "step": 310000 }, { "epoch": 25.67, "learning_rate": 7.199334995843724e-06, "loss": 0.0253, "step": 310100 }, { "epoch": 25.68, "learning_rate": 7.195178719867e-06, "loss": 0.0353, "step": 310200 }, { "epoch": 25.69, "learning_rate": 7.1910640066500425e-06, "loss": 0.0251, "step": 310300 }, { "epoch": 25.7, "learning_rate": 7.186907730673317e-06, "loss": 0.0261, "step": 310400 }, { "epoch": 25.7, "learning_rate": 7.182751454696593e-06, "loss": 0.0302, "step": 310500 }, { "epoch": 25.7, "eval_loss": 0.15765096247196198, "eval_runtime": 4258.929, "eval_samples_per_second": 17.75, "eval_steps_per_second": 1.479, "eval_wer": 0.05195530468414664, "step": 310500 }, { "epoch": 25.71, "learning_rate": 7.1786367414796345e-06, "loss": 0.0217, "step": 310600 }, { "epoch": 25.72, "learning_rate": 7.17448046550291e-06, "loss": 0.0242, "step": 310700 }, { "epoch": 25.73, "learning_rate": 7.1703241895261855e-06, "loss": 0.0263, "step": 310800 }, { "epoch": 25.74, "learning_rate": 7.16616791354946e-06, "loss": 0.0263, "step": 310900 }, { "epoch": 25.75, "learning_rate": 7.162011637572736e-06, "loss": 0.0252, "step": 311000 }, { "epoch": 25.75, "eval_loss": 0.15734127163887024, "eval_runtime": 4254.4766, "eval_samples_per_second": 17.768, "eval_steps_per_second": 1.481, "eval_wer": 0.05053992224627347, "step": 311000 }, { "epoch": 25.75, "learning_rate": 7.15785536159601e-06, "loss": 0.026, "step": 311100 }, { "epoch": 25.76, "learning_rate": 7.153699085619286e-06, "loss": 0.0233, "step": 311200 }, { "epoch": 25.77, "learning_rate": 7.149542809642561e-06, "loss": 0.0239, "step": 311300 }, { "epoch": 25.78, "learning_rate": 7.145386533665836e-06, "loss": 0.0247, "step": 311400 }, { "epoch": 25.79, "learning_rate": 7.141230257689112e-06, "loss": 0.0275, "step": 311500 }, { "epoch": 25.79, "eval_loss": 0.1494714617729187, "eval_runtime": 4259.8875, "eval_samples_per_second": 17.746, "eval_steps_per_second": 1.479, "eval_wer": 0.04979530800721845, "step": 311500 }, { "epoch": 25.79, "learning_rate": 7.137073981712385e-06, "loss": 0.0242, "step": 311600 }, { "epoch": 25.8, "learning_rate": 7.132917705735661e-06, "loss": 0.0255, "step": 311700 }, { "epoch": 25.81, "learning_rate": 7.128761429758937e-06, "loss": 0.0229, "step": 311800 }, { "epoch": 25.82, "learning_rate": 7.124605153782211e-06, "loss": 0.0228, "step": 311900 }, { "epoch": 25.83, "learning_rate": 7.120448877805487e-06, "loss": 0.0297, "step": 312000 }, { "epoch": 25.83, "eval_loss": 0.1524471938610077, "eval_runtime": 4258.0783, "eval_samples_per_second": 17.753, "eval_steps_per_second": 1.48, "eval_wer": 0.0499691538936094, "step": 312000 }, { "epoch": 25.84, "learning_rate": 7.116292601828761e-06, "loss": 0.0259, "step": 312100 }, { "epoch": 25.84, "learning_rate": 7.112136325852037e-06, "loss": 0.0279, "step": 312200 }, { "epoch": 25.85, "learning_rate": 7.107980049875312e-06, "loss": 0.0235, "step": 312300 }, { "epoch": 25.86, "learning_rate": 7.103823773898587e-06, "loss": 0.0301, "step": 312400 }, { "epoch": 25.87, "learning_rate": 7.0996674979218625e-06, "loss": 0.0251, "step": 312500 }, { "epoch": 25.87, "eval_loss": 0.1585235744714737, "eval_runtime": 4263.5764, "eval_samples_per_second": 17.73, "eval_steps_per_second": 1.478, "eval_wer": 0.04946915466283898, "step": 312500 }, { "epoch": 25.88, "learning_rate": 7.095511221945137e-06, "loss": 0.0229, "step": 312600 }, { "epoch": 25.89, "learning_rate": 7.091354945968413e-06, "loss": 0.0211, "step": 312700 }, { "epoch": 25.89, "learning_rate": 7.087198669991688e-06, "loss": 0.0291, "step": 312800 }, { "epoch": 25.9, "learning_rate": 7.08308395677473e-06, "loss": 0.0249, "step": 312900 }, { "epoch": 25.91, "learning_rate": 7.0789276807980056e-06, "loss": 0.023, "step": 313000 }, { "epoch": 25.91, "eval_loss": 0.15815819799900055, "eval_runtime": 4275.061, "eval_samples_per_second": 17.683, "eval_steps_per_second": 1.474, "eval_wer": 0.05005376914804746, "step": 313000 }, { "epoch": 25.92, "learning_rate": 7.074771404821281e-06, "loss": 0.0266, "step": 313100 }, { "epoch": 25.93, "learning_rate": 7.070615128844556e-06, "loss": 0.0267, "step": 313200 }, { "epoch": 25.94, "learning_rate": 7.066458852867831e-06, "loss": 0.0236, "step": 313300 }, { "epoch": 25.94, "learning_rate": 7.062302576891106e-06, "loss": 0.0264, "step": 313400 }, { "epoch": 25.95, "learning_rate": 7.0581463009143814e-06, "loss": 0.0236, "step": 313500 }, { "epoch": 25.95, "eval_loss": 0.15578600764274597, "eval_runtime": 4255.7413, "eval_samples_per_second": 17.763, "eval_steps_per_second": 1.48, "eval_wer": 0.04866146359774831, "step": 313500 }, { "epoch": 25.96, "learning_rate": 7.053990024937657e-06, "loss": 0.0201, "step": 313600 }, { "epoch": 25.97, "learning_rate": 7.049833748960932e-06, "loss": 0.0226, "step": 313700 }, { "epoch": 25.98, "learning_rate": 7.045677472984207e-06, "loss": 0.0252, "step": 313800 }, { "epoch": 25.99, "learning_rate": 7.041521197007482e-06, "loss": 0.0215, "step": 313900 }, { "epoch": 25.99, "learning_rate": 7.037364921030757e-06, "loss": 0.0287, "step": 314000 }, { "epoch": 25.99, "eval_loss": 0.16192090511322021, "eval_runtime": 4301.8527, "eval_samples_per_second": 17.573, "eval_steps_per_second": 1.464, "eval_wer": 0.04974761577289881, "step": 314000 }, { "epoch": 26.0, "learning_rate": 7.033208645054033e-06, "loss": 0.0229, "step": 314100 }, { "epoch": 26.01, "learning_rate": 7.0290523690773075e-06, "loss": 0.0267, "step": 314200 }, { "epoch": 26.02, "learning_rate": 7.024896093100583e-06, "loss": 0.0215, "step": 314300 }, { "epoch": 26.03, "learning_rate": 7.020739817123857e-06, "loss": 0.027, "step": 314400 }, { "epoch": 26.03, "learning_rate": 7.016583541147133e-06, "loss": 0.0215, "step": 314500 }, { "epoch": 26.03, "eval_loss": 0.1698404997587204, "eval_runtime": 4262.2463, "eval_samples_per_second": 17.736, "eval_steps_per_second": 1.478, "eval_wer": 0.050669152816687976, "step": 314500 }, { "epoch": 26.04, "learning_rate": 7.012427265170409e-06, "loss": 0.0193, "step": 314600 }, { "epoch": 26.05, "learning_rate": 7.0082709891936825e-06, "loss": 0.0221, "step": 314700 }, { "epoch": 26.06, "learning_rate": 7.004114713216958e-06, "loss": 0.0211, "step": 314800 }, { "epoch": 26.07, "learning_rate": 6.999958437240233e-06, "loss": 0.0223, "step": 314900 }, { "epoch": 26.08, "learning_rate": 6.995802161263508e-06, "loss": 0.0244, "step": 315000 }, { "epoch": 26.08, "eval_loss": 0.16744618117809296, "eval_runtime": 4274.2266, "eval_samples_per_second": 17.686, "eval_steps_per_second": 1.474, "eval_wer": 0.048981463105441374, "step": 315000 }, { "epoch": 26.08, "learning_rate": 6.991645885286784e-06, "loss": 0.0203, "step": 315100 }, { "epoch": 26.09, "learning_rate": 6.9874896093100584e-06, "loss": 0.0192, "step": 315200 }, { "epoch": 26.1, "learning_rate": 6.983333333333334e-06, "loss": 0.0231, "step": 315300 }, { "epoch": 26.11, "learning_rate": 6.979177057356609e-06, "loss": 0.0257, "step": 315400 }, { "epoch": 26.12, "learning_rate": 6.975020781379884e-06, "loss": 0.0243, "step": 315500 }, { "epoch": 26.12, "eval_loss": 0.15585178136825562, "eval_runtime": 4268.5829, "eval_samples_per_second": 17.71, "eval_steps_per_second": 1.476, "eval_wer": 0.049364539439170095, "step": 315500 }, { "epoch": 26.13, "learning_rate": 6.97086450540316e-06, "loss": 0.0278, "step": 315600 }, { "epoch": 26.13, "learning_rate": 6.966708229426434e-06, "loss": 0.0214, "step": 315700 }, { "epoch": 26.14, "learning_rate": 6.96255195344971e-06, "loss": 0.0227, "step": 315800 }, { "epoch": 26.15, "learning_rate": 6.9583956774729845e-06, "loss": 0.0248, "step": 315900 }, { "epoch": 26.16, "learning_rate": 6.954280964256027e-06, "loss": 0.0236, "step": 316000 }, { "epoch": 26.16, "eval_loss": 0.17228305339813232, "eval_runtime": 4248.1838, "eval_samples_per_second": 17.795, "eval_steps_per_second": 1.483, "eval_wer": 0.05028146110544445, "step": 316000 }, { "epoch": 26.17, "learning_rate": 6.950124688279303e-06, "loss": 0.0246, "step": 316100 }, { "epoch": 26.18, "learning_rate": 6.945968412302577e-06, "loss": 0.0241, "step": 316200 }, { "epoch": 26.18, "learning_rate": 6.941812136325853e-06, "loss": 0.023, "step": 316300 }, { "epoch": 26.19, "learning_rate": 6.9376558603491276e-06, "loss": 0.0238, "step": 316400 }, { "epoch": 26.2, "learning_rate": 6.933499584372403e-06, "loss": 0.021, "step": 316500 }, { "epoch": 26.2, "eval_loss": 0.16230708360671997, "eval_runtime": 4250.1322, "eval_samples_per_second": 17.787, "eval_steps_per_second": 1.482, "eval_wer": 0.050076846035621486, "step": 316500 }, { "epoch": 26.21, "learning_rate": 6.929343308395679e-06, "loss": 0.0221, "step": 316600 }, { "epoch": 26.22, "learning_rate": 6.925187032418953e-06, "loss": 0.0212, "step": 316700 }, { "epoch": 26.23, "learning_rate": 6.921030756442229e-06, "loss": 0.0197, "step": 316800 }, { "epoch": 26.23, "learning_rate": 6.9168744804655034e-06, "loss": 0.0213, "step": 316900 }, { "epoch": 26.24, "learning_rate": 6.912718204488779e-06, "loss": 0.0242, "step": 317000 }, { "epoch": 26.24, "eval_loss": 0.1656346172094345, "eval_runtime": 4254.6965, "eval_samples_per_second": 17.767, "eval_steps_per_second": 1.481, "eval_wer": 0.050335307176450496, "step": 317000 }, { "epoch": 26.25, "learning_rate": 6.9085619285120545e-06, "loss": 0.0226, "step": 317100 }, { "epoch": 26.26, "learning_rate": 6.904405652535328e-06, "loss": 0.0231, "step": 317200 }, { "epoch": 26.27, "learning_rate": 6.900249376558605e-06, "loss": 0.0222, "step": 317300 }, { "epoch": 26.27, "learning_rate": 6.8960931005818785e-06, "loss": 0.0206, "step": 317400 }, { "epoch": 26.28, "learning_rate": 6.891936824605154e-06, "loss": 0.0243, "step": 317500 }, { "epoch": 26.28, "eval_loss": 0.15832772850990295, "eval_runtime": 4267.413, "eval_samples_per_second": 17.714, "eval_steps_per_second": 1.476, "eval_wer": 0.0494968469279278, "step": 317500 }, { "epoch": 26.29, "learning_rate": 6.8877805486284295e-06, "loss": 0.0262, "step": 317600 }, { "epoch": 26.3, "learning_rate": 6.883624272651704e-06, "loss": 0.022, "step": 317700 }, { "epoch": 26.31, "learning_rate": 6.87946799667498e-06, "loss": 0.0236, "step": 317800 }, { "epoch": 26.32, "learning_rate": 6.875311720698254e-06, "loss": 0.0229, "step": 317900 }, { "epoch": 26.32, "learning_rate": 6.87115544472153e-06, "loss": 0.0196, "step": 318000 }, { "epoch": 26.32, "eval_loss": 0.1700233817100525, "eval_runtime": 4262.9785, "eval_samples_per_second": 17.733, "eval_steps_per_second": 1.478, "eval_wer": 0.0490122322888734, "step": 318000 }, { "epoch": 26.33, "learning_rate": 6.8670407315045726e-06, "loss": 0.0235, "step": 318100 }, { "epoch": 26.34, "learning_rate": 6.862884455527847e-06, "loss": 0.0236, "step": 318200 }, { "epoch": 26.35, "learning_rate": 6.858728179551123e-06, "loss": 0.0209, "step": 318300 }, { "epoch": 26.36, "learning_rate": 6.854571903574397e-06, "loss": 0.0235, "step": 318400 }, { "epoch": 26.37, "learning_rate": 6.850415627597673e-06, "loss": 0.0191, "step": 318500 }, { "epoch": 26.37, "eval_loss": 0.16589322686195374, "eval_runtime": 4258.2415, "eval_samples_per_second": 17.753, "eval_steps_per_second": 1.479, "eval_wer": 0.04958915447822388, "step": 318500 }, { "epoch": 26.37, "learning_rate": 6.8462593516209484e-06, "loss": 0.0264, "step": 318600 }, { "epoch": 26.38, "learning_rate": 6.842103075644223e-06, "loss": 0.029, "step": 318700 }, { "epoch": 26.39, "learning_rate": 6.837946799667499e-06, "loss": 0.0257, "step": 318800 }, { "epoch": 26.4, "learning_rate": 6.833790523690773e-06, "loss": 0.0245, "step": 318900 }, { "epoch": 26.41, "learning_rate": 6.829634247714049e-06, "loss": 0.0242, "step": 319000 }, { "epoch": 26.41, "eval_loss": 0.16057461500167847, "eval_runtime": 4330.7788, "eval_samples_per_second": 17.455, "eval_steps_per_second": 1.455, "eval_wer": 0.04920300122615196, "step": 319000 }, { "epoch": 26.42, "learning_rate": 6.825477971737324e-06, "loss": 0.0238, "step": 319100 }, { "epoch": 26.42, "learning_rate": 6.821321695760599e-06, "loss": 0.0236, "step": 319200 }, { "epoch": 26.43, "learning_rate": 6.8171654197838745e-06, "loss": 0.0263, "step": 319300 }, { "epoch": 26.44, "learning_rate": 6.813009143807149e-06, "loss": 0.0211, "step": 319400 }, { "epoch": 26.45, "learning_rate": 6.808852867830425e-06, "loss": 0.0227, "step": 319500 }, { "epoch": 26.45, "eval_loss": 0.15528564155101776, "eval_runtime": 4251.473, "eval_samples_per_second": 17.781, "eval_steps_per_second": 1.482, "eval_wer": 0.04915684745100392, "step": 319500 }, { "epoch": 26.46, "learning_rate": 6.8046965918537e-06, "loss": 0.025, "step": 319600 }, { "epoch": 26.47, "learning_rate": 6.800540315876975e-06, "loss": 0.0268, "step": 319700 }, { "epoch": 26.47, "learning_rate": 6.79638403990025e-06, "loss": 0.0266, "step": 319800 }, { "epoch": 26.48, "learning_rate": 6.792227763923524e-06, "loss": 0.0243, "step": 319900 }, { "epoch": 26.49, "learning_rate": 6.7880714879468e-06, "loss": 0.0211, "step": 320000 }, { "epoch": 26.49, "eval_loss": 0.15347783267498016, "eval_runtime": 4276.4934, "eval_samples_per_second": 17.677, "eval_steps_per_second": 1.473, "eval_wer": 0.049170693583548336, "step": 320000 }, { "epoch": 26.5, "learning_rate": 6.783956774729843e-06, "loss": 0.0228, "step": 320100 }, { "epoch": 26.51, "learning_rate": 6.779800498753117e-06, "loss": 0.0202, "step": 320200 }, { "epoch": 26.51, "learning_rate": 6.775644222776393e-06, "loss": 0.0188, "step": 320300 }, { "epoch": 26.52, "learning_rate": 6.771487946799669e-06, "loss": 0.0233, "step": 320400 }, { "epoch": 26.53, "learning_rate": 6.767331670822943e-06, "loss": 0.0206, "step": 320500 }, { "epoch": 26.53, "eval_loss": 0.16099123656749725, "eval_runtime": 4248.6935, "eval_samples_per_second": 17.793, "eval_steps_per_second": 1.483, "eval_wer": 0.04923223195041238, "step": 320500 }, { "epoch": 26.54, "learning_rate": 6.763175394846218e-06, "loss": 0.0215, "step": 320600 }, { "epoch": 26.55, "learning_rate": 6.759019118869493e-06, "loss": 0.023, "step": 320700 }, { "epoch": 26.56, "learning_rate": 6.7548628428927685e-06, "loss": 0.0213, "step": 320800 }, { "epoch": 26.56, "learning_rate": 6.750706566916044e-06, "loss": 0.0224, "step": 320900 }, { "epoch": 26.57, "learning_rate": 6.746550290939319e-06, "loss": 0.0211, "step": 321000 }, { "epoch": 26.57, "eval_loss": 0.1596706658601761, "eval_runtime": 4279.5558, "eval_samples_per_second": 17.664, "eval_steps_per_second": 1.472, "eval_wer": 0.04857069450662384, "step": 321000 }, { "epoch": 26.58, "learning_rate": 6.742394014962594e-06, "loss": 0.0246, "step": 321100 }, { "epoch": 26.59, "learning_rate": 6.738237738985869e-06, "loss": 0.0221, "step": 321200 }, { "epoch": 26.6, "learning_rate": 6.734081463009144e-06, "loss": 0.0239, "step": 321300 }, { "epoch": 26.61, "learning_rate": 6.72992518703242e-06, "loss": 0.0254, "step": 321400 }, { "epoch": 26.61, "learning_rate": 6.7257689110556946e-06, "loss": 0.023, "step": 321500 }, { "epoch": 26.61, "eval_loss": 0.1543499082326889, "eval_runtime": 4247.2687, "eval_samples_per_second": 17.798, "eval_steps_per_second": 1.483, "eval_wer": 0.048036849174078196, "step": 321500 }, { "epoch": 26.62, "learning_rate": 6.72161263507897e-06, "loss": 0.0255, "step": 321600 }, { "epoch": 26.63, "learning_rate": 6.717456359102245e-06, "loss": 0.0276, "step": 321700 }, { "epoch": 26.64, "learning_rate": 6.71330008312552e-06, "loss": 0.0224, "step": 321800 }, { "epoch": 26.65, "learning_rate": 6.709143807148796e-06, "loss": 0.0248, "step": 321900 }, { "epoch": 26.66, "learning_rate": 6.7049875311720705e-06, "loss": 0.0224, "step": 322000 }, { "epoch": 26.66, "eval_loss": 0.16775822639465332, "eval_runtime": 4259.236, "eval_samples_per_second": 17.748, "eval_steps_per_second": 1.479, "eval_wer": 0.049464539285324176, "step": 322000 }, { "epoch": 26.66, "learning_rate": 6.700831255195346e-06, "loss": 0.0273, "step": 322100 }, { "epoch": 26.67, "learning_rate": 6.696674979218621e-06, "loss": 0.0207, "step": 322200 }, { "epoch": 26.68, "learning_rate": 6.692518703241896e-06, "loss": 0.0214, "step": 322300 }, { "epoch": 26.69, "learning_rate": 6.688362427265172e-06, "loss": 0.0258, "step": 322400 }, { "epoch": 26.7, "learning_rate": 6.684206151288446e-06, "loss": 0.025, "step": 322500 }, { "epoch": 26.7, "eval_loss": 0.16594603657722473, "eval_runtime": 4257.2492, "eval_samples_per_second": 17.757, "eval_steps_per_second": 1.48, "eval_wer": 0.04994915392437858, "step": 322500 }, { "epoch": 26.71, "learning_rate": 6.680049875311722e-06, "loss": 0.0308, "step": 322600 }, { "epoch": 26.71, "learning_rate": 6.675893599334996e-06, "loss": 0.0271, "step": 322700 }, { "epoch": 26.72, "learning_rate": 6.671737323358272e-06, "loss": 0.0216, "step": 322800 }, { "epoch": 26.73, "learning_rate": 6.6675810473815476e-06, "loss": 0.0217, "step": 322900 }, { "epoch": 26.74, "learning_rate": 6.663424771404821e-06, "loss": 0.0235, "step": 323000 }, { "epoch": 26.74, "eval_loss": 0.16272631287574768, "eval_runtime": 4274.5414, "eval_samples_per_second": 17.685, "eval_steps_per_second": 1.474, "eval_wer": 0.049013770748045006, "step": 323000 }, { "epoch": 26.75, "learning_rate": 6.659268495428097e-06, "loss": 0.0242, "step": 323100 }, { "epoch": 26.75, "learning_rate": 6.6551122194513716e-06, "loss": 0.0275, "step": 323200 }, { "epoch": 26.76, "learning_rate": 6.650955943474647e-06, "loss": 0.0215, "step": 323300 }, { "epoch": 26.77, "learning_rate": 6.646799667497923e-06, "loss": 0.0243, "step": 323400 }, { "epoch": 26.78, "learning_rate": 6.642643391521197e-06, "loss": 0.0253, "step": 323500 }, { "epoch": 26.78, "eval_loss": 0.17328812181949615, "eval_runtime": 4258.0943, "eval_samples_per_second": 17.753, "eval_steps_per_second": 1.48, "eval_wer": 0.05128145956698528, "step": 323500 }, { "epoch": 26.79, "learning_rate": 6.638487115544473e-06, "loss": 0.0199, "step": 323600 }, { "epoch": 26.8, "learning_rate": 6.6343308395677475e-06, "loss": 0.0248, "step": 323700 }, { "epoch": 26.8, "learning_rate": 6.63021612635079e-06, "loss": 0.0209, "step": 323800 }, { "epoch": 26.81, "learning_rate": 6.626059850374066e-06, "loss": 0.0234, "step": 323900 }, { "epoch": 26.82, "learning_rate": 6.62190357439734e-06, "loss": 0.0217, "step": 324000 }, { "epoch": 26.82, "eval_loss": 0.16965433955192566, "eval_runtime": 4258.1937, "eval_samples_per_second": 17.753, "eval_steps_per_second": 1.48, "eval_wer": 0.05061530674568193, "step": 324000 }, { "epoch": 26.83, "learning_rate": 6.617747298420616e-06, "loss": 0.0209, "step": 324100 }, { "epoch": 26.84, "learning_rate": 6.6135910224438905e-06, "loss": 0.0213, "step": 324200 }, { "epoch": 26.85, "learning_rate": 6.609434746467166e-06, "loss": 0.0221, "step": 324300 }, { "epoch": 26.85, "learning_rate": 6.6052784704904415e-06, "loss": 0.022, "step": 324400 }, { "epoch": 26.86, "learning_rate": 6.601122194513716e-06, "loss": 0.0209, "step": 324500 }, { "epoch": 26.86, "eval_loss": 0.16843178868293762, "eval_runtime": 4251.6049, "eval_samples_per_second": 17.78, "eval_steps_per_second": 1.482, "eval_wer": 0.05068761432674719, "step": 324500 }, { "epoch": 26.87, "learning_rate": 6.596965918536992e-06, "loss": 0.0261, "step": 324600 }, { "epoch": 26.88, "learning_rate": 6.592809642560266e-06, "loss": 0.0218, "step": 324700 }, { "epoch": 26.89, "learning_rate": 6.588653366583542e-06, "loss": 0.0221, "step": 324800 }, { "epoch": 26.9, "learning_rate": 6.584497090606817e-06, "loss": 0.0279, "step": 324900 }, { "epoch": 26.9, "learning_rate": 6.580340814630092e-06, "loss": 0.0243, "step": 325000 }, { "epoch": 26.9, "eval_loss": 0.16332215070724487, "eval_runtime": 4263.2635, "eval_samples_per_second": 17.732, "eval_steps_per_second": 1.478, "eval_wer": 0.0501029998415387, "step": 325000 }, { "epoch": 26.91, "learning_rate": 6.576184538653368e-06, "loss": 0.0206, "step": 325100 }, { "epoch": 26.92, "learning_rate": 6.5720282626766414e-06, "loss": 0.0251, "step": 325200 }, { "epoch": 26.93, "learning_rate": 6.567871986699918e-06, "loss": 0.0205, "step": 325300 }, { "epoch": 26.94, "learning_rate": 6.563715710723193e-06, "loss": 0.0242, "step": 325400 }, { "epoch": 26.95, "learning_rate": 6.559559434746467e-06, "loss": 0.026, "step": 325500 }, { "epoch": 26.95, "eval_loss": 0.16982166469097137, "eval_runtime": 4257.8652, "eval_samples_per_second": 17.754, "eval_steps_per_second": 1.48, "eval_wer": 0.04991838474094655, "step": 325500 }, { "epoch": 26.95, "learning_rate": 6.5554031587697435e-06, "loss": 0.0211, "step": 325600 }, { "epoch": 26.96, "learning_rate": 6.551246882793017e-06, "loss": 0.0231, "step": 325700 }, { "epoch": 26.97, "learning_rate": 6.547090606816293e-06, "loss": 0.0238, "step": 325800 }, { "epoch": 26.98, "learning_rate": 6.542934330839568e-06, "loss": 0.0215, "step": 325900 }, { "epoch": 26.99, "learning_rate": 6.538778054862843e-06, "loss": 0.024, "step": 326000 }, { "epoch": 26.99, "eval_loss": 0.16048327088356018, "eval_runtime": 4268.2195, "eval_samples_per_second": 17.711, "eval_steps_per_second": 1.476, "eval_wer": 0.04981223105810607, "step": 326000 }, { "epoch": 27.0, "learning_rate": 6.5346217788861185e-06, "loss": 0.0219, "step": 326100 }, { "epoch": 27.0, "learning_rate": 6.530465502909393e-06, "loss": 0.0229, "step": 326200 }, { "epoch": 27.01, "learning_rate": 6.526309226932669e-06, "loss": 0.0216, "step": 326300 }, { "epoch": 27.02, "learning_rate": 6.522152950955944e-06, "loss": 0.02, "step": 326400 }, { "epoch": 27.03, "learning_rate": 6.517996674979219e-06, "loss": 0.0199, "step": 326500 }, { "epoch": 27.03, "eval_loss": 0.166230246424675, "eval_runtime": 4273.1153, "eval_samples_per_second": 17.691, "eval_steps_per_second": 1.474, "eval_wer": 0.04982300027230727, "step": 326500 }, { "epoch": 27.04, "learning_rate": 6.513840399002494e-06, "loss": 0.022, "step": 326600 }, { "epoch": 27.04, "learning_rate": 6.509684123025769e-06, "loss": 0.0245, "step": 326700 }, { "epoch": 27.05, "learning_rate": 6.505527847049045e-06, "loss": 0.0205, "step": 326800 }, { "epoch": 27.06, "learning_rate": 6.50137157107232e-06, "loss": 0.0218, "step": 326900 }, { "epoch": 27.07, "learning_rate": 6.497215295095595e-06, "loss": 0.0231, "step": 327000 }, { "epoch": 27.07, "eval_loss": 0.16446614265441895, "eval_runtime": 4267.7619, "eval_samples_per_second": 17.713, "eval_steps_per_second": 1.476, "eval_wer": 0.04995223084272178, "step": 327000 }, { "epoch": 27.08, "learning_rate": 6.49305901911887e-06, "loss": 0.0198, "step": 327100 }, { "epoch": 27.09, "learning_rate": 6.488902743142145e-06, "loss": 0.024, "step": 327200 }, { "epoch": 27.09, "learning_rate": 6.4847464671654205e-06, "loss": 0.0214, "step": 327300 }, { "epoch": 27.1, "learning_rate": 6.480590191188696e-06, "loss": 0.0225, "step": 327400 }, { "epoch": 27.11, "learning_rate": 6.476433915211971e-06, "loss": 0.0268, "step": 327500 }, { "epoch": 27.11, "eval_loss": 0.164895698428154, "eval_runtime": 4260.7279, "eval_samples_per_second": 17.742, "eval_steps_per_second": 1.479, "eval_wer": 0.05015684591254475, "step": 327500 }, { "epoch": 27.12, "learning_rate": 6.472277639235246e-06, "loss": 0.0205, "step": 327600 }, { "epoch": 27.13, "learning_rate": 6.468121363258521e-06, "loss": 0.0273, "step": 327700 }, { "epoch": 27.14, "learning_rate": 6.463965087281796e-06, "loss": 0.0241, "step": 327800 }, { "epoch": 27.14, "learning_rate": 6.459808811305072e-06, "loss": 0.0216, "step": 327900 }, { "epoch": 27.15, "learning_rate": 6.455652535328346e-06, "loss": 0.026, "step": 328000 }, { "epoch": 27.15, "eval_loss": 0.16703662276268005, "eval_runtime": 4242.1585, "eval_samples_per_second": 17.82, "eval_steps_per_second": 1.485, "eval_wer": 0.04926300113384441, "step": 328000 }, { "epoch": 27.16, "learning_rate": 6.451496259351622e-06, "loss": 0.0237, "step": 328100 }, { "epoch": 27.17, "learning_rate": 6.447381546134663e-06, "loss": 0.0248, "step": 328200 }, { "epoch": 27.18, "learning_rate": 6.443225270157939e-06, "loss": 0.0191, "step": 328300 }, { "epoch": 27.19, "learning_rate": 6.439068994181215e-06, "loss": 0.022, "step": 328400 }, { "epoch": 27.19, "learning_rate": 6.434912718204489e-06, "loss": 0.0231, "step": 328500 }, { "epoch": 27.19, "eval_loss": 0.16620615124702454, "eval_runtime": 4275.6313, "eval_samples_per_second": 17.68, "eval_steps_per_second": 1.473, "eval_wer": 0.05023838424863962, "step": 328500 }, { "epoch": 27.2, "learning_rate": 6.430756442227764e-06, "loss": 0.0243, "step": 328600 }, { "epoch": 27.21, "learning_rate": 6.426600166251039e-06, "loss": 0.0221, "step": 328700 }, { "epoch": 27.22, "learning_rate": 6.4224438902743145e-06, "loss": 0.0232, "step": 328800 }, { "epoch": 27.23, "learning_rate": 6.41828761429759e-06, "loss": 0.024, "step": 328900 }, { "epoch": 27.24, "learning_rate": 6.414131338320865e-06, "loss": 0.0227, "step": 329000 }, { "epoch": 27.24, "eval_loss": 0.1704210340976715, "eval_runtime": 4292.7891, "eval_samples_per_second": 17.61, "eval_steps_per_second": 1.468, "eval_wer": 0.050998383079410646, "step": 329000 }, { "epoch": 27.24, "learning_rate": 6.40997506234414e-06, "loss": 0.0233, "step": 329100 }, { "epoch": 27.25, "learning_rate": 6.405818786367415e-06, "loss": 0.0243, "step": 329200 }, { "epoch": 27.26, "learning_rate": 6.40166251039069e-06, "loss": 0.0176, "step": 329300 }, { "epoch": 27.27, "learning_rate": 6.397506234413966e-06, "loss": 0.0211, "step": 329400 }, { "epoch": 27.28, "learning_rate": 6.3933499584372405e-06, "loss": 0.0215, "step": 329500 }, { "epoch": 27.28, "eval_loss": 0.1616627424955368, "eval_runtime": 4276.5221, "eval_samples_per_second": 17.677, "eval_steps_per_second": 1.473, "eval_wer": 0.04783531102259843, "step": 329500 }, { "epoch": 27.28, "learning_rate": 6.389193682460516e-06, "loss": 0.0245, "step": 329600 }, { "epoch": 27.29, "learning_rate": 6.385037406483791e-06, "loss": 0.0217, "step": 329700 }, { "epoch": 27.3, "learning_rate": 6.380881130507066e-06, "loss": 0.0215, "step": 329800 }, { "epoch": 27.31, "learning_rate": 6.376724854530342e-06, "loss": 0.0217, "step": 329900 }, { "epoch": 27.32, "learning_rate": 6.372568578553616e-06, "loss": 0.025, "step": 330000 }, { "epoch": 27.32, "eval_loss": 0.17023000121116638, "eval_runtime": 4252.0142, "eval_samples_per_second": 17.779, "eval_steps_per_second": 1.482, "eval_wer": 0.04990453860840214, "step": 330000 }, { "epoch": 27.33, "learning_rate": 6.368412302576892e-06, "loss": 0.0173, "step": 330100 }, { "epoch": 27.33, "learning_rate": 6.364256026600167e-06, "loss": 0.0252, "step": 330200 }, { "epoch": 27.34, "learning_rate": 6.360099750623442e-06, "loss": 0.0199, "step": 330300 }, { "epoch": 27.35, "learning_rate": 6.355943474646718e-06, "loss": 0.0198, "step": 330400 }, { "epoch": 27.36, "learning_rate": 6.351787198669992e-06, "loss": 0.0249, "step": 330500 }, { "epoch": 27.36, "eval_loss": 0.16947472095489502, "eval_runtime": 4267.443, "eval_samples_per_second": 17.714, "eval_steps_per_second": 1.476, "eval_wer": 0.05030915337053327, "step": 330500 }, { "epoch": 27.37, "learning_rate": 6.347630922693268e-06, "loss": 0.0227, "step": 330600 }, { "epoch": 27.38, "learning_rate": 6.343516209476309e-06, "loss": 0.0299, "step": 330700 }, { "epoch": 27.38, "learning_rate": 6.339359933499585e-06, "loss": 0.021, "step": 330800 }, { "epoch": 27.39, "learning_rate": 6.335203657522861e-06, "loss": 0.0226, "step": 330900 }, { "epoch": 27.4, "learning_rate": 6.3310473815461345e-06, "loss": 0.0235, "step": 331000 }, { "epoch": 27.4, "eval_loss": 0.16203533113002777, "eval_runtime": 4265.8289, "eval_samples_per_second": 17.721, "eval_steps_per_second": 1.477, "eval_wer": 0.0492537703788148, "step": 331000 }, { "epoch": 27.41, "learning_rate": 6.32689110556941e-06, "loss": 0.0263, "step": 331100 }, { "epoch": 27.42, "learning_rate": 6.3227763923524535e-06, "loss": 0.0229, "step": 331200 }, { "epoch": 27.43, "learning_rate": 6.318620116375727e-06, "loss": 0.0238, "step": 331300 }, { "epoch": 27.43, "learning_rate": 6.314463840399003e-06, "loss": 0.0212, "step": 331400 }, { "epoch": 27.44, "learning_rate": 6.3103075644222776e-06, "loss": 0.03, "step": 331500 }, { "epoch": 27.44, "eval_loss": 0.17722351849079132, "eval_runtime": 4251.6307, "eval_samples_per_second": 17.78, "eval_steps_per_second": 1.482, "eval_wer": 0.050112230596568316, "step": 331500 }, { "epoch": 27.45, "learning_rate": 6.306151288445553e-06, "loss": 0.0203, "step": 331600 }, { "epoch": 27.46, "learning_rate": 6.301995012468829e-06, "loss": 0.0231, "step": 331700 }, { "epoch": 27.47, "learning_rate": 6.297838736492103e-06, "loss": 0.0273, "step": 331800 }, { "epoch": 27.48, "learning_rate": 6.293682460515379e-06, "loss": 0.0202, "step": 331900 }, { "epoch": 27.48, "learning_rate": 6.2895261845386534e-06, "loss": 0.027, "step": 332000 }, { "epoch": 27.48, "eval_loss": 0.17277346551418304, "eval_runtime": 4273.801, "eval_samples_per_second": 17.688, "eval_steps_per_second": 1.474, "eval_wer": 0.050862998672309735, "step": 332000 }, { "epoch": 27.49, "learning_rate": 6.285411471321696e-06, "loss": 0.0199, "step": 332100 }, { "epoch": 27.5, "learning_rate": 6.281255195344972e-06, "loss": 0.0259, "step": 332200 }, { "epoch": 27.51, "learning_rate": 6.277098919368246e-06, "loss": 0.0224, "step": 332300 }, { "epoch": 27.52, "learning_rate": 6.272942643391522e-06, "loss": 0.0236, "step": 332400 }, { "epoch": 27.52, "learning_rate": 6.2687863674147965e-06, "loss": 0.0301, "step": 332500 }, { "epoch": 27.52, "eval_loss": 0.16610358655452728, "eval_runtime": 4270.2521, "eval_samples_per_second": 17.703, "eval_steps_per_second": 1.475, "eval_wer": 0.05024299962615442, "step": 332500 }, { "epoch": 27.53, "learning_rate": 6.264630091438072e-06, "loss": 0.0339, "step": 332600 }, { "epoch": 27.54, "learning_rate": 6.2604738154613475e-06, "loss": 0.022, "step": 332700 }, { "epoch": 27.55, "learning_rate": 6.256317539484622e-06, "loss": 0.0209, "step": 332800 }, { "epoch": 27.56, "learning_rate": 6.252161263507898e-06, "loss": 0.0244, "step": 332900 }, { "epoch": 27.57, "learning_rate": 6.248004987531172e-06, "loss": 0.0208, "step": 333000 }, { "epoch": 27.57, "eval_loss": 0.171222984790802, "eval_runtime": 4253.7579, "eval_samples_per_second": 17.771, "eval_steps_per_second": 1.481, "eval_wer": 0.050927613957516986, "step": 333000 }, { "epoch": 27.57, "learning_rate": 6.243848711554448e-06, "loss": 0.0224, "step": 333100 }, { "epoch": 27.58, "learning_rate": 6.239692435577723e-06, "loss": 0.0245, "step": 333200 }, { "epoch": 27.59, "learning_rate": 6.235536159600998e-06, "loss": 0.0224, "step": 333300 }, { "epoch": 27.6, "learning_rate": 6.231379883624274e-06, "loss": 0.0234, "step": 333400 }, { "epoch": 27.61, "learning_rate": 6.227223607647548e-06, "loss": 0.0226, "step": 333500 }, { "epoch": 27.61, "eval_loss": 0.17040465772151947, "eval_runtime": 4258.8698, "eval_samples_per_second": 17.75, "eval_steps_per_second": 1.479, "eval_wer": 0.050306076452190075, "step": 333500 }, { "epoch": 27.62, "learning_rate": 6.223067331670824e-06, "loss": 0.0243, "step": 333600 }, { "epoch": 27.62, "learning_rate": 6.218911055694099e-06, "loss": 0.022, "step": 333700 }, { "epoch": 27.63, "learning_rate": 6.214754779717373e-06, "loss": 0.0239, "step": 333800 }, { "epoch": 27.64, "learning_rate": 6.2105985037406495e-06, "loss": 0.0182, "step": 333900 }, { "epoch": 27.65, "learning_rate": 6.206442227763923e-06, "loss": 0.0241, "step": 334000 }, { "epoch": 27.65, "eval_loss": 0.1608024686574936, "eval_runtime": 4263.1051, "eval_samples_per_second": 17.732, "eval_steps_per_second": 1.478, "eval_wer": 0.04909838600248307, "step": 334000 }, { "epoch": 27.66, "learning_rate": 6.202285951787199e-06, "loss": 0.0268, "step": 334100 }, { "epoch": 27.67, "learning_rate": 6.198129675810474e-06, "loss": 0.02, "step": 334200 }, { "epoch": 27.67, "learning_rate": 6.193973399833749e-06, "loss": 0.0235, "step": 334300 }, { "epoch": 27.68, "learning_rate": 6.1898171238570245e-06, "loss": 0.021, "step": 334400 }, { "epoch": 27.69, "learning_rate": 6.185660847880299e-06, "loss": 0.0217, "step": 334500 }, { "epoch": 27.69, "eval_loss": 0.17031249403953552, "eval_runtime": 4258.415, "eval_samples_per_second": 17.752, "eval_steps_per_second": 1.479, "eval_wer": 0.049330693337394864, "step": 334500 }, { "epoch": 27.7, "learning_rate": 6.181504571903575e-06, "loss": 0.023, "step": 334600 }, { "epoch": 27.71, "learning_rate": 6.17734829592685e-06, "loss": 0.022, "step": 334700 }, { "epoch": 27.72, "learning_rate": 6.173192019950125e-06, "loss": 0.0216, "step": 334800 }, { "epoch": 27.72, "learning_rate": 6.1690357439734e-06, "loss": 0.0197, "step": 334900 }, { "epoch": 27.73, "learning_rate": 6.164879467996675e-06, "loss": 0.0254, "step": 335000 }, { "epoch": 27.73, "eval_loss": 0.17127148807048798, "eval_runtime": 4286.0818, "eval_samples_per_second": 17.637, "eval_steps_per_second": 1.47, "eval_wer": 0.050390691706628146, "step": 335000 }, { "epoch": 27.74, "learning_rate": 6.160723192019951e-06, "loss": 0.0198, "step": 335100 }, { "epoch": 27.75, "learning_rate": 6.156566916043226e-06, "loss": 0.0249, "step": 335200 }, { "epoch": 27.76, "learning_rate": 6.152410640066501e-06, "loss": 0.0204, "step": 335300 }, { "epoch": 27.76, "learning_rate": 6.148254364089776e-06, "loss": 0.0228, "step": 335400 }, { "epoch": 27.77, "learning_rate": 6.144098088113051e-06, "loss": 0.0196, "step": 335500 }, { "epoch": 27.77, "eval_loss": 0.17418259382247925, "eval_runtime": 4284.2534, "eval_samples_per_second": 17.645, "eval_steps_per_second": 1.471, "eval_wer": 0.050770691122013656, "step": 335500 }, { "epoch": 27.78, "learning_rate": 6.1399418121363265e-06, "loss": 0.0241, "step": 335600 }, { "epoch": 27.79, "learning_rate": 6.135785536159602e-06, "loss": 0.025, "step": 335700 }, { "epoch": 27.8, "learning_rate": 6.131629260182877e-06, "loss": 0.0246, "step": 335800 }, { "epoch": 27.81, "learning_rate": 6.127472984206152e-06, "loss": 0.0191, "step": 335900 }, { "epoch": 27.81, "learning_rate": 6.123316708229427e-06, "loss": 0.0253, "step": 336000 }, { "epoch": 27.81, "eval_loss": 0.167551189661026, "eval_runtime": 4252.6618, "eval_samples_per_second": 17.776, "eval_steps_per_second": 1.481, "eval_wer": 0.05031838412556289, "step": 336000 }, { "epoch": 27.82, "learning_rate": 6.119160432252702e-06, "loss": 0.0211, "step": 336100 }, { "epoch": 27.83, "learning_rate": 6.115004156275978e-06, "loss": 0.0263, "step": 336200 }, { "epoch": 27.84, "learning_rate": 6.1108478802992526e-06, "loss": 0.0197, "step": 336300 }, { "epoch": 27.85, "learning_rate": 6.106691604322528e-06, "loss": 0.0247, "step": 336400 }, { "epoch": 27.86, "learning_rate": 6.102535328345802e-06, "loss": 0.0279, "step": 336500 }, { "epoch": 27.86, "eval_loss": 0.17374154925346375, "eval_runtime": 4248.6356, "eval_samples_per_second": 17.793, "eval_steps_per_second": 1.483, "eval_wer": 0.05107992141550551, "step": 336500 }, { "epoch": 27.86, "learning_rate": 6.098379052369077e-06, "loss": 0.0216, "step": 336600 }, { "epoch": 27.87, "learning_rate": 6.094222776392354e-06, "loss": 0.0245, "step": 336700 }, { "epoch": 27.88, "learning_rate": 6.090066500415628e-06, "loss": 0.0219, "step": 336800 }, { "epoch": 27.89, "learning_rate": 6.085910224438903e-06, "loss": 0.0226, "step": 336900 }, { "epoch": 27.9, "learning_rate": 6.081753948462178e-06, "loss": 0.0212, "step": 337000 }, { "epoch": 27.9, "eval_loss": 0.16492845118045807, "eval_runtime": 4238.2956, "eval_samples_per_second": 17.836, "eval_steps_per_second": 1.486, "eval_wer": 0.05030299953384687, "step": 337000 }, { "epoch": 27.91, "learning_rate": 6.077597672485453e-06, "loss": 0.0215, "step": 337100 }, { "epoch": 27.91, "learning_rate": 6.073441396508729e-06, "loss": 0.0214, "step": 337200 }, { "epoch": 27.92, "learning_rate": 6.0692851205320035e-06, "loss": 0.0234, "step": 337300 }, { "epoch": 27.93, "learning_rate": 6.065128844555279e-06, "loss": 0.0218, "step": 337400 }, { "epoch": 27.94, "learning_rate": 6.060972568578554e-06, "loss": 0.0262, "step": 337500 }, { "epoch": 27.94, "eval_loss": 0.1631237417459488, "eval_runtime": 4257.42, "eval_samples_per_second": 17.756, "eval_steps_per_second": 1.48, "eval_wer": 0.04909223216579667, "step": 337500 }, { "epoch": 27.95, "learning_rate": 6.056816292601829e-06, "loss": 0.0211, "step": 337600 }, { "epoch": 27.96, "learning_rate": 6.052660016625105e-06, "loss": 0.0217, "step": 337700 }, { "epoch": 27.96, "learning_rate": 6.048503740648379e-06, "loss": 0.0194, "step": 337800 }, { "epoch": 27.97, "learning_rate": 6.044347464671655e-06, "loss": 0.0212, "step": 337900 }, { "epoch": 27.98, "learning_rate": 6.0401911886949296e-06, "loss": 0.0257, "step": 338000 }, { "epoch": 27.98, "eval_loss": 0.16542497277259827, "eval_runtime": 4265.9103, "eval_samples_per_second": 17.721, "eval_steps_per_second": 1.477, "eval_wer": 0.04970607737526558, "step": 338000 }, { "epoch": 27.99, "learning_rate": 6.036034912718205e-06, "loss": 0.021, "step": 338100 }, { "epoch": 28.0, "learning_rate": 6.031878636741481e-06, "loss": 0.0212, "step": 338200 }, { "epoch": 28.0, "learning_rate": 6.027722360764755e-06, "loss": 0.0202, "step": 338300 }, { "epoch": 28.01, "learning_rate": 6.023566084788031e-06, "loss": 0.0213, "step": 338400 }, { "epoch": 28.02, "learning_rate": 6.0194098088113054e-06, "loss": 0.0201, "step": 338500 }, { "epoch": 28.02, "eval_loss": 0.17354296147823334, "eval_runtime": 4279.7559, "eval_samples_per_second": 17.663, "eval_steps_per_second": 1.472, "eval_wer": 0.04956453913147826, "step": 338500 }, { "epoch": 28.03, "learning_rate": 6.015253532834581e-06, "loss": 0.0206, "step": 338600 }, { "epoch": 28.04, "learning_rate": 6.0110972568578565e-06, "loss": 0.0198, "step": 338700 }, { "epoch": 28.05, "learning_rate": 6.006940980881131e-06, "loss": 0.0197, "step": 338800 }, { "epoch": 28.05, "learning_rate": 6.002784704904407e-06, "loss": 0.0194, "step": 338900 }, { "epoch": 28.06, "learning_rate": 5.9986284289276805e-06, "loss": 0.0265, "step": 339000 }, { "epoch": 28.06, "eval_loss": 0.16222169995307922, "eval_runtime": 4279.0396, "eval_samples_per_second": 17.666, "eval_steps_per_second": 1.472, "eval_wer": 0.04845377160958214, "step": 339000 }, { "epoch": 28.07, "learning_rate": 5.994472152950956e-06, "loss": 0.0195, "step": 339100 }, { "epoch": 28.08, "learning_rate": 5.9903574397339995e-06, "loss": 0.0172, "step": 339200 }, { "epoch": 28.09, "learning_rate": 5.986201163757273e-06, "loss": 0.0222, "step": 339300 }, { "epoch": 28.1, "learning_rate": 5.982086450540317e-06, "loss": 0.0324, "step": 339400 }, { "epoch": 28.1, "learning_rate": 5.977930174563592e-06, "loss": 0.0244, "step": 339500 }, { "epoch": 28.1, "eval_loss": 0.17535507678985596, "eval_runtime": 4272.0926, "eval_samples_per_second": 17.695, "eval_steps_per_second": 1.475, "eval_wer": 0.05008146141313629, "step": 339500 }, { "epoch": 28.11, "learning_rate": 5.973773898586866e-06, "loss": 0.021, "step": 339600 }, { "epoch": 28.12, "learning_rate": 5.969617622610142e-06, "loss": 0.0201, "step": 339700 }, { "epoch": 28.13, "learning_rate": 5.965461346633416e-06, "loss": 0.0228, "step": 339800 }, { "epoch": 28.14, "learning_rate": 5.961305070656692e-06, "loss": 0.0215, "step": 339900 }, { "epoch": 28.15, "learning_rate": 5.957148794679967e-06, "loss": 0.0241, "step": 340000 }, { "epoch": 28.15, "eval_loss": 0.1689147651195526, "eval_runtime": 4243.5895, "eval_samples_per_second": 17.814, "eval_steps_per_second": 1.485, "eval_wer": 0.05017992280011877, "step": 340000 }, { "epoch": 28.15, "learning_rate": 5.952992518703242e-06, "loss": 0.0241, "step": 340100 }, { "epoch": 28.16, "learning_rate": 5.948836242726518e-06, "loss": 0.0235, "step": 340200 }, { "epoch": 28.17, "learning_rate": 5.944679966749792e-06, "loss": 0.0213, "step": 340300 }, { "epoch": 28.18, "learning_rate": 5.940523690773068e-06, "loss": 0.0253, "step": 340400 }, { "epoch": 28.19, "learning_rate": 5.936367414796343e-06, "loss": 0.0227, "step": 340500 }, { "epoch": 28.19, "eval_loss": 0.17368170619010925, "eval_runtime": 4248.3542, "eval_samples_per_second": 17.794, "eval_steps_per_second": 1.483, "eval_wer": 0.04887069404508609, "step": 340500 }, { "epoch": 28.2, "learning_rate": 5.932211138819618e-06, "loss": 0.0209, "step": 340600 }, { "epoch": 28.2, "learning_rate": 5.9280548628428935e-06, "loss": 0.0223, "step": 340700 }, { "epoch": 28.21, "learning_rate": 5.923898586866168e-06, "loss": 0.021, "step": 340800 }, { "epoch": 28.22, "learning_rate": 5.919742310889444e-06, "loss": 0.0396, "step": 340900 }, { "epoch": 28.23, "learning_rate": 5.915586034912719e-06, "loss": 0.0218, "step": 341000 }, { "epoch": 28.23, "eval_loss": 0.16852885484695435, "eval_runtime": 4265.4638, "eval_samples_per_second": 17.723, "eval_steps_per_second": 1.477, "eval_wer": 0.04956453913147826, "step": 341000 }, { "epoch": 28.24, "learning_rate": 5.911429758935994e-06, "loss": 0.0233, "step": 341100 }, { "epoch": 28.25, "learning_rate": 5.907273482959269e-06, "loss": 0.0223, "step": 341200 }, { "epoch": 28.25, "learning_rate": 5.903117206982544e-06, "loss": 0.0195, "step": 341300 }, { "epoch": 28.26, "learning_rate": 5.8989609310058196e-06, "loss": 0.0214, "step": 341400 }, { "epoch": 28.27, "learning_rate": 5.894804655029095e-06, "loss": 0.0261, "step": 341500 }, { "epoch": 28.27, "eval_loss": 0.16817937791347504, "eval_runtime": 4245.9551, "eval_samples_per_second": 17.804, "eval_steps_per_second": 1.484, "eval_wer": 0.04935684714331209, "step": 341500 }, { "epoch": 28.28, "learning_rate": 5.89064837905237e-06, "loss": 0.0285, "step": 341600 }, { "epoch": 28.29, "learning_rate": 5.886492103075645e-06, "loss": 0.0242, "step": 341700 }, { "epoch": 28.29, "learning_rate": 5.882335827098919e-06, "loss": 0.0221, "step": 341800 }, { "epoch": 28.3, "learning_rate": 5.8781795511221955e-06, "loss": 0.0212, "step": 341900 }, { "epoch": 28.31, "learning_rate": 5.874023275145471e-06, "loss": 0.0207, "step": 342000 }, { "epoch": 28.31, "eval_loss": 0.16235408186912537, "eval_runtime": 4259.6162, "eval_samples_per_second": 17.747, "eval_steps_per_second": 1.479, "eval_wer": 0.04850300230307338, "step": 342000 }, { "epoch": 28.32, "learning_rate": 5.869866999168745e-06, "loss": 0.0212, "step": 342100 }, { "epoch": 28.33, "learning_rate": 5.86571072319202e-06, "loss": 0.0244, "step": 342200 }, { "epoch": 28.34, "learning_rate": 5.861554447215295e-06, "loss": 0.0212, "step": 342300 }, { "epoch": 28.34, "learning_rate": 5.8573981712385705e-06, "loss": 0.0193, "step": 342400 }, { "epoch": 28.35, "learning_rate": 5.853241895261846e-06, "loss": 0.0273, "step": 342500 }, { "epoch": 28.35, "eval_loss": 0.17012634873390198, "eval_runtime": 4259.4029, "eval_samples_per_second": 17.748, "eval_steps_per_second": 1.479, "eval_wer": 0.049743000395384006, "step": 342500 }, { "epoch": 28.36, "learning_rate": 5.849085619285121e-06, "loss": 0.0207, "step": 342600 }, { "epoch": 28.37, "learning_rate": 5.844929343308396e-06, "loss": 0.0215, "step": 342700 }, { "epoch": 28.38, "learning_rate": 5.840773067331671e-06, "loss": 0.0251, "step": 342800 }, { "epoch": 28.39, "learning_rate": 5.836616791354946e-06, "loss": 0.0191, "step": 342900 }, { "epoch": 28.39, "learning_rate": 5.832460515378222e-06, "loss": 0.0203, "step": 343000 }, { "epoch": 28.39, "eval_loss": 0.17441004514694214, "eval_runtime": 4251.6286, "eval_samples_per_second": 17.78, "eval_steps_per_second": 1.482, "eval_wer": 0.051076844497162315, "step": 343000 }, { "epoch": 28.4, "learning_rate": 5.8283042394014966e-06, "loss": 0.0203, "step": 343100 }, { "epoch": 28.41, "learning_rate": 5.824147963424772e-06, "loss": 0.0215, "step": 343200 }, { "epoch": 28.42, "learning_rate": 5.819991687448047e-06, "loss": 0.0204, "step": 343300 }, { "epoch": 28.43, "learning_rate": 5.815835411471322e-06, "loss": 0.021, "step": 343400 }, { "epoch": 28.44, "learning_rate": 5.811679135494598e-06, "loss": 0.0181, "step": 343500 }, { "epoch": 28.44, "eval_loss": 0.18557117879390717, "eval_runtime": 4241.8502, "eval_samples_per_second": 17.821, "eval_steps_per_second": 1.485, "eval_wer": 0.05054607608295987, "step": 343500 }, { "epoch": 28.44, "learning_rate": 5.80756442227764e-06, "loss": 0.021, "step": 343600 }, { "epoch": 28.45, "learning_rate": 5.803408146300915e-06, "loss": 0.0196, "step": 343700 }, { "epoch": 28.46, "learning_rate": 5.79925187032419e-06, "loss": 0.0224, "step": 343800 }, { "epoch": 28.47, "learning_rate": 5.795095594347465e-06, "loss": 0.0193, "step": 343900 }, { "epoch": 28.48, "learning_rate": 5.790939318370741e-06, "loss": 0.023, "step": 344000 }, { "epoch": 28.48, "eval_loss": 0.16727718710899353, "eval_runtime": 4276.5549, "eval_samples_per_second": 17.677, "eval_steps_per_second": 1.473, "eval_wer": 0.0491306936450867, "step": 344000 }, { "epoch": 28.49, "learning_rate": 5.7867830423940155e-06, "loss": 0.0215, "step": 344100 }, { "epoch": 28.49, "learning_rate": 5.782626766417291e-06, "loss": 0.0171, "step": 344200 }, { "epoch": 28.5, "learning_rate": 5.778470490440566e-06, "loss": 0.0239, "step": 344300 }, { "epoch": 28.51, "learning_rate": 5.774314214463841e-06, "loss": 0.0216, "step": 344400 }, { "epoch": 28.52, "learning_rate": 5.770157938487117e-06, "loss": 0.022, "step": 344500 }, { "epoch": 28.52, "eval_loss": 0.1729736179113388, "eval_runtime": 4276.2349, "eval_samples_per_second": 17.678, "eval_steps_per_second": 1.473, "eval_wer": 0.05051684535869945, "step": 344500 }, { "epoch": 28.53, "learning_rate": 5.7660016625103905e-06, "loss": 0.0187, "step": 344600 }, { "epoch": 28.53, "learning_rate": 5.761845386533667e-06, "loss": 0.0205, "step": 344700 }, { "epoch": 28.54, "learning_rate": 5.757730673316708e-06, "loss": 0.0235, "step": 344800 }, { "epoch": 28.55, "learning_rate": 5.753574397339983e-06, "loss": 0.0197, "step": 344900 }, { "epoch": 28.56, "learning_rate": 5.74941812136326e-06, "loss": 0.0209, "step": 345000 }, { "epoch": 28.56, "eval_loss": 0.17523859441280365, "eval_runtime": 4274.2934, "eval_samples_per_second": 17.686, "eval_steps_per_second": 1.474, "eval_wer": 0.04974761577289881, "step": 345000 }, { "epoch": 28.57, "learning_rate": 5.745261845386534e-06, "loss": 0.0185, "step": 345100 }, { "epoch": 28.58, "learning_rate": 5.741105569409809e-06, "loss": 0.0237, "step": 345200 }, { "epoch": 28.58, "learning_rate": 5.736949293433084e-06, "loss": 0.0195, "step": 345300 }, { "epoch": 28.59, "learning_rate": 5.732793017456359e-06, "loss": 0.0219, "step": 345400 }, { "epoch": 28.6, "learning_rate": 5.728678304239402e-06, "loss": 0.0191, "step": 345500 }, { "epoch": 28.6, "eval_loss": 0.17928774654865265, "eval_runtime": 4282.6545, "eval_samples_per_second": 17.651, "eval_steps_per_second": 1.471, "eval_wer": 0.05020453814686439, "step": 345500 }, { "epoch": 28.61, "learning_rate": 5.724522028262677e-06, "loss": 0.0175, "step": 345600 }, { "epoch": 28.62, "learning_rate": 5.720365752285952e-06, "loss": 0.0225, "step": 345700 }, { "epoch": 28.63, "learning_rate": 5.716209476309227e-06, "loss": 0.0163, "step": 345800 }, { "epoch": 28.63, "learning_rate": 5.712053200332502e-06, "loss": 0.0178, "step": 345900 }, { "epoch": 28.64, "learning_rate": 5.707896924355778e-06, "loss": 0.0186, "step": 346000 }, { "epoch": 28.64, "eval_loss": 0.17991885542869568, "eval_runtime": 4272.8505, "eval_samples_per_second": 17.692, "eval_steps_per_second": 1.474, "eval_wer": 0.04990915398591694, "step": 346000 }, { "epoch": 28.65, "learning_rate": 5.7037406483790525e-06, "loss": 0.0192, "step": 346100 }, { "epoch": 28.66, "learning_rate": 5.699584372402328e-06, "loss": 0.0206, "step": 346200 }, { "epoch": 28.67, "learning_rate": 5.6954280964256036e-06, "loss": 0.0208, "step": 346300 }, { "epoch": 28.68, "learning_rate": 5.691271820448878e-06, "loss": 0.0215, "step": 346400 }, { "epoch": 28.68, "learning_rate": 5.687115544472154e-06, "loss": 0.026, "step": 346500 }, { "epoch": 28.68, "eval_loss": 0.1725488156080246, "eval_runtime": 4267.3632, "eval_samples_per_second": 17.715, "eval_steps_per_second": 1.476, "eval_wer": 0.0496106929066263, "step": 346500 }, { "epoch": 28.69, "learning_rate": 5.682959268495428e-06, "loss": 0.0192, "step": 346600 }, { "epoch": 28.7, "learning_rate": 5.678802992518704e-06, "loss": 0.0235, "step": 346700 }, { "epoch": 28.71, "learning_rate": 5.6746467165419794e-06, "loss": 0.0292, "step": 346800 }, { "epoch": 28.72, "learning_rate": 5.670490440565254e-06, "loss": 0.022, "step": 346900 }, { "epoch": 28.73, "learning_rate": 5.66633416458853e-06, "loss": 0.0233, "step": 347000 }, { "epoch": 28.73, "eval_loss": 0.1716708540916443, "eval_runtime": 4255.5017, "eval_samples_per_second": 17.764, "eval_steps_per_second": 1.48, "eval_wer": 0.0497368465586976, "step": 347000 }, { "epoch": 28.73, "learning_rate": 5.662177888611804e-06, "loss": 0.0215, "step": 347100 }, { "epoch": 28.74, "learning_rate": 5.65802161263508e-06, "loss": 0.0251, "step": 347200 }, { "epoch": 28.75, "learning_rate": 5.653865336658355e-06, "loss": 0.0186, "step": 347300 }, { "epoch": 28.76, "learning_rate": 5.64970906068163e-06, "loss": 0.0227, "step": 347400 }, { "epoch": 28.77, "learning_rate": 5.6455527847049055e-06, "loss": 0.024, "step": 347500 }, { "epoch": 28.77, "eval_loss": 0.17180359363555908, "eval_runtime": 4266.2193, "eval_samples_per_second": 17.719, "eval_steps_per_second": 1.477, "eval_wer": 0.049692231242721166, "step": 347500 }, { "epoch": 28.77, "learning_rate": 5.641396508728179e-06, "loss": 0.0222, "step": 347600 }, { "epoch": 28.78, "learning_rate": 5.637240232751455e-06, "loss": 0.0206, "step": 347700 }, { "epoch": 28.79, "learning_rate": 5.633083956774731e-06, "loss": 0.0172, "step": 347800 }, { "epoch": 28.8, "learning_rate": 5.628927680798005e-06, "loss": 0.0218, "step": 347900 }, { "epoch": 28.81, "learning_rate": 5.6247714048212806e-06, "loss": 0.0175, "step": 348000 }, { "epoch": 28.81, "eval_loss": 0.1797322779893875, "eval_runtime": 4259.3197, "eval_samples_per_second": 17.748, "eval_steps_per_second": 1.479, "eval_wer": 0.05071684505100761, "step": 348000 }, { "epoch": 28.82, "learning_rate": 5.620615128844555e-06, "loss": 0.0268, "step": 348100 }, { "epoch": 28.82, "learning_rate": 5.616458852867831e-06, "loss": 0.0187, "step": 348200 }, { "epoch": 28.83, "learning_rate": 5.612302576891106e-06, "loss": 0.0189, "step": 348300 }, { "epoch": 28.84, "learning_rate": 5.608146300914381e-06, "loss": 0.0207, "step": 348400 }, { "epoch": 28.85, "learning_rate": 5.6039900249376564e-06, "loss": 0.0228, "step": 348500 }, { "epoch": 28.85, "eval_loss": 0.1775551438331604, "eval_runtime": 4250.0844, "eval_samples_per_second": 17.787, "eval_steps_per_second": 1.482, "eval_wer": 0.050421460890060166, "step": 348500 }, { "epoch": 28.86, "learning_rate": 5.599875311720698e-06, "loss": 0.0203, "step": 348600 }, { "epoch": 28.87, "learning_rate": 5.595719035743974e-06, "loss": 0.0215, "step": 348700 }, { "epoch": 28.87, "learning_rate": 5.591562759767249e-06, "loss": 0.0216, "step": 348800 }, { "epoch": 28.88, "learning_rate": 5.587406483790524e-06, "loss": 0.026, "step": 348900 }, { "epoch": 28.89, "learning_rate": 5.5832502078137995e-06, "loss": 0.0197, "step": 349000 }, { "epoch": 28.89, "eval_loss": 0.17536646127700806, "eval_runtime": 4290.7781, "eval_samples_per_second": 17.618, "eval_steps_per_second": 1.468, "eval_wer": 0.05100915229361186, "step": 349000 }, { "epoch": 28.9, "learning_rate": 5.579093931837074e-06, "loss": 0.0192, "step": 349100 }, { "epoch": 28.91, "learning_rate": 5.57493765586035e-06, "loss": 0.0277, "step": 349200 }, { "epoch": 28.92, "learning_rate": 5.570781379883625e-06, "loss": 0.019, "step": 349300 }, { "epoch": 28.92, "learning_rate": 5.5666251039069e-06, "loss": 0.0225, "step": 349400 }, { "epoch": 28.93, "learning_rate": 5.562468827930175e-06, "loss": 0.0221, "step": 349500 }, { "epoch": 28.93, "eval_loss": 0.17973721027374268, "eval_runtime": 4276.6315, "eval_samples_per_second": 17.676, "eval_steps_per_second": 1.473, "eval_wer": 0.051006075375268654, "step": 349500 }, { "epoch": 28.94, "learning_rate": 5.55831255195345e-06, "loss": 0.023, "step": 349600 }, { "epoch": 28.95, "learning_rate": 5.5541562759767256e-06, "loss": 0.024, "step": 349700 }, { "epoch": 28.96, "learning_rate": 5.550000000000001e-06, "loss": 0.0185, "step": 349800 }, { "epoch": 28.97, "learning_rate": 5.545843724023276e-06, "loss": 0.0217, "step": 349900 }, { "epoch": 28.97, "learning_rate": 5.541687448046551e-06, "loss": 0.0206, "step": 350000 }, { "epoch": 28.97, "eval_loss": 0.16925954818725586, "eval_runtime": 4313.24, "eval_samples_per_second": 17.526, "eval_steps_per_second": 1.461, "eval_wer": 0.049193770471122354, "step": 350000 }, { "epoch": 28.98, "learning_rate": 5.537531172069825e-06, "loss": 0.0194, "step": 350100 }, { "epoch": 28.99, "learning_rate": 5.5333748960931014e-06, "loss": 0.0225, "step": 350200 }, { "epoch": 29.0, "learning_rate": 5.529218620116377e-06, "loss": 0.019, "step": 350300 }, { "epoch": 29.01, "learning_rate": 5.525062344139651e-06, "loss": 0.0208, "step": 350400 }, { "epoch": 29.01, "learning_rate": 5.520906068162927e-06, "loss": 0.0205, "step": 350500 }, { "epoch": 29.01, "eval_loss": 0.17347750067710876, "eval_runtime": 4259.7726, "eval_samples_per_second": 17.746, "eval_steps_per_second": 1.479, "eval_wer": 0.04986300021076891, "step": 350500 }, { "epoch": 29.02, "learning_rate": 5.516749792186201e-06, "loss": 0.0208, "step": 350600 }, { "epoch": 29.03, "learning_rate": 5.5125935162094765e-06, "loss": 0.02, "step": 350700 }, { "epoch": 29.04, "learning_rate": 5.508437240232752e-06, "loss": 0.0196, "step": 350800 }, { "epoch": 29.05, "learning_rate": 5.504280964256027e-06, "loss": 0.0178, "step": 350900 }, { "epoch": 29.06, "learning_rate": 5.500124688279302e-06, "loss": 0.0214, "step": 351000 }, { "epoch": 29.06, "eval_loss": 0.17631380259990692, "eval_runtime": 4259.4281, "eval_samples_per_second": 17.748, "eval_steps_per_second": 1.479, "eval_wer": 0.04969530816106437, "step": 351000 }, { "epoch": 29.06, "learning_rate": 5.495968412302577e-06, "loss": 0.0231, "step": 351100 }, { "epoch": 29.07, "learning_rate": 5.491812136325852e-06, "loss": 0.0202, "step": 351200 }, { "epoch": 29.08, "learning_rate": 5.487655860349128e-06, "loss": 0.0219, "step": 351300 }, { "epoch": 29.09, "learning_rate": 5.4834995843724026e-06, "loss": 0.0174, "step": 351400 }, { "epoch": 29.1, "learning_rate": 5.479343308395678e-06, "loss": 0.0219, "step": 351500 }, { "epoch": 29.1, "eval_loss": 0.1813071370124817, "eval_runtime": 4251.3836, "eval_samples_per_second": 17.781, "eval_steps_per_second": 1.482, "eval_wer": 0.05017376896343236, "step": 351500 }, { "epoch": 29.11, "learning_rate": 5.475187032418953e-06, "loss": 0.0194, "step": 351600 }, { "epoch": 29.11, "learning_rate": 5.471030756442228e-06, "loss": 0.0157, "step": 351700 }, { "epoch": 29.12, "learning_rate": 5.466874480465504e-06, "loss": 0.0181, "step": 351800 }, { "epoch": 29.13, "learning_rate": 5.4627182044887784e-06, "loss": 0.023, "step": 351900 }, { "epoch": 29.14, "learning_rate": 5.458561928512054e-06, "loss": 0.023, "step": 352000 }, { "epoch": 29.14, "eval_loss": 0.1716579794883728, "eval_runtime": 4277.7165, "eval_samples_per_second": 17.672, "eval_steps_per_second": 1.473, "eval_wer": 0.05003223071964505, "step": 352000 }, { "epoch": 29.15, "learning_rate": 5.454405652535329e-06, "loss": 0.0213, "step": 352100 }, { "epoch": 29.16, "learning_rate": 5.450249376558604e-06, "loss": 0.0258, "step": 352200 }, { "epoch": 29.16, "learning_rate": 5.44609310058188e-06, "loss": 0.0232, "step": 352300 }, { "epoch": 29.17, "learning_rate": 5.441936824605154e-06, "loss": 0.0241, "step": 352400 }, { "epoch": 29.18, "learning_rate": 5.43778054862843e-06, "loss": 0.0233, "step": 352500 }, { "epoch": 29.18, "eval_loss": 0.16904759407043457, "eval_runtime": 4298.5236, "eval_samples_per_second": 17.586, "eval_steps_per_second": 1.466, "eval_wer": 0.04957992372319427, "step": 352500 }, { "epoch": 29.19, "learning_rate": 5.4336242726517045e-06, "loss": 0.0237, "step": 352600 }, { "epoch": 29.2, "learning_rate": 5.429509559434747e-06, "loss": 0.0215, "step": 352700 }, { "epoch": 29.21, "learning_rate": 5.425353283458023e-06, "loss": 0.0168, "step": 352800 }, { "epoch": 29.21, "learning_rate": 5.421197007481297e-06, "loss": 0.0208, "step": 352900 }, { "epoch": 29.22, "learning_rate": 5.417040731504573e-06, "loss": 0.0226, "step": 353000 }, { "epoch": 29.22, "eval_loss": 0.18332967162132263, "eval_runtime": 4259.1553, "eval_samples_per_second": 17.749, "eval_steps_per_second": 1.479, "eval_wer": 0.051833766409590136, "step": 353000 }, { "epoch": 29.23, "learning_rate": 5.412884455527847e-06, "loss": 0.0193, "step": 353100 }, { "epoch": 29.24, "learning_rate": 5.408728179551122e-06, "loss": 0.019, "step": 353200 }, { "epoch": 29.25, "learning_rate": 5.404571903574399e-06, "loss": 0.0198, "step": 353300 }, { "epoch": 29.25, "learning_rate": 5.400415627597672e-06, "loss": 0.0179, "step": 353400 }, { "epoch": 29.26, "learning_rate": 5.396259351620948e-06, "loss": 0.0177, "step": 353500 }, { "epoch": 29.26, "eval_loss": 0.18135038018226624, "eval_runtime": 4247.9784, "eval_samples_per_second": 17.796, "eval_steps_per_second": 1.483, "eval_wer": 0.05156453605455991, "step": 353500 }, { "epoch": 29.27, "learning_rate": 5.392103075644223e-06, "loss": 0.0183, "step": 353600 }, { "epoch": 29.28, "learning_rate": 5.387946799667498e-06, "loss": 0.0198, "step": 353700 }, { "epoch": 29.29, "learning_rate": 5.383790523690774e-06, "loss": 0.0232, "step": 353800 }, { "epoch": 29.3, "learning_rate": 5.379634247714048e-06, "loss": 0.0202, "step": 353900 }, { "epoch": 29.3, "learning_rate": 5.375477971737324e-06, "loss": 0.0218, "step": 354000 }, { "epoch": 29.3, "eval_loss": 0.17454756796360016, "eval_runtime": 4326.9918, "eval_samples_per_second": 17.471, "eval_steps_per_second": 1.456, "eval_wer": 0.05055838375633268, "step": 354000 }, { "epoch": 29.31, "learning_rate": 5.3713216957605985e-06, "loss": 0.0209, "step": 354100 }, { "epoch": 29.32, "learning_rate": 5.367165419783874e-06, "loss": 0.0215, "step": 354200 }, { "epoch": 29.33, "learning_rate": 5.3630091438071495e-06, "loss": 0.024, "step": 354300 }, { "epoch": 29.34, "learning_rate": 5.358852867830424e-06, "loss": 0.0193, "step": 354400 }, { "epoch": 29.35, "learning_rate": 5.3546965918537e-06, "loss": 0.0222, "step": 354500 }, { "epoch": 29.35, "eval_loss": 0.17360937595367432, "eval_runtime": 4258.1221, "eval_samples_per_second": 17.753, "eval_steps_per_second": 1.48, "eval_wer": 0.05054299916461667, "step": 354500 }, { "epoch": 29.35, "learning_rate": 5.350540315876974e-06, "loss": 0.022, "step": 354600 }, { "epoch": 29.36, "learning_rate": 5.34638403990025e-06, "loss": 0.0241, "step": 354700 }, { "epoch": 29.37, "learning_rate": 5.342227763923525e-06, "loss": 0.0252, "step": 354800 }, { "epoch": 29.38, "learning_rate": 5.3380714879468e-06, "loss": 0.0202, "step": 354900 }, { "epoch": 29.39, "learning_rate": 5.333915211970076e-06, "loss": 0.0209, "step": 355000 }, { "epoch": 29.39, "eval_loss": 0.16637356579303741, "eval_runtime": 4247.5355, "eval_samples_per_second": 17.797, "eval_steps_per_second": 1.483, "eval_wer": 0.049573769886507864, "step": 355000 }, { "epoch": 29.4, "learning_rate": 5.32975893599335e-06, "loss": 0.0189, "step": 355100 }, { "epoch": 29.4, "learning_rate": 5.325602660016626e-06, "loss": 0.0198, "step": 355200 }, { "epoch": 29.41, "learning_rate": 5.321446384039901e-06, "loss": 0.0204, "step": 355300 }, { "epoch": 29.42, "learning_rate": 5.317290108063176e-06, "loss": 0.0224, "step": 355400 }, { "epoch": 29.43, "learning_rate": 5.3131338320864515e-06, "loss": 0.0165, "step": 355500 }, { "epoch": 29.43, "eval_loss": 0.16730983555316925, "eval_runtime": 4265.3804, "eval_samples_per_second": 17.723, "eval_steps_per_second": 1.477, "eval_wer": 0.04953223148887463, "step": 355500 }, { "epoch": 29.44, "learning_rate": 5.308977556109725e-06, "loss": 0.0247, "step": 355600 }, { "epoch": 29.45, "learning_rate": 5.304821280133001e-06, "loss": 0.0207, "step": 355700 }, { "epoch": 29.45, "learning_rate": 5.300665004156277e-06, "loss": 0.0247, "step": 355800 }, { "epoch": 29.46, "learning_rate": 5.296508728179551e-06, "loss": 0.0193, "step": 355900 }, { "epoch": 29.47, "learning_rate": 5.2923524522028265e-06, "loss": 0.0221, "step": 356000 }, { "epoch": 29.47, "eval_loss": 0.17361171543598175, "eval_runtime": 4249.0151, "eval_samples_per_second": 17.791, "eval_steps_per_second": 1.483, "eval_wer": 0.04980300030307646, "step": 356000 }, { "epoch": 29.48, "learning_rate": 5.288196176226101e-06, "loss": 0.0246, "step": 356100 }, { "epoch": 29.49, "learning_rate": 5.284039900249377e-06, "loss": 0.0192, "step": 356200 }, { "epoch": 29.5, "learning_rate": 5.279883624272652e-06, "loss": 0.018, "step": 356300 }, { "epoch": 29.5, "learning_rate": 5.275727348295927e-06, "loss": 0.022, "step": 356400 }, { "epoch": 29.51, "learning_rate": 5.271571072319202e-06, "loss": 0.018, "step": 356500 }, { "epoch": 29.51, "eval_loss": 0.17254520952701569, "eval_runtime": 4323.354, "eval_samples_per_second": 17.485, "eval_steps_per_second": 1.457, "eval_wer": 0.049472231581182184, "step": 356500 }, { "epoch": 29.52, "learning_rate": 5.267414796342477e-06, "loss": 0.019, "step": 356600 }, { "epoch": 29.53, "learning_rate": 5.263258520365753e-06, "loss": 0.0199, "step": 356700 }, { "epoch": 29.54, "learning_rate": 5.259102244389028e-06, "loss": 0.0208, "step": 356800 }, { "epoch": 29.54, "learning_rate": 5.254945968412303e-06, "loss": 0.0204, "step": 356900 }, { "epoch": 29.55, "learning_rate": 5.2508312551953455e-06, "loss": 0.0233, "step": 357000 }, { "epoch": 29.55, "eval_loss": 0.17152801156044006, "eval_runtime": 4301.2427, "eval_samples_per_second": 17.575, "eval_steps_per_second": 1.465, "eval_wer": 0.05053992224627347, "step": 357000 }, { "epoch": 29.56, "learning_rate": 5.24667497921862e-06, "loss": 0.0222, "step": 357100 }, { "epoch": 29.57, "learning_rate": 5.242518703241896e-06, "loss": 0.025, "step": 357200 }, { "epoch": 29.58, "learning_rate": 5.238362427265171e-06, "loss": 0.0187, "step": 357300 }, { "epoch": 29.59, "learning_rate": 5.234206151288446e-06, "loss": 0.0191, "step": 357400 }, { "epoch": 29.59, "learning_rate": 5.230049875311721e-06, "loss": 0.0201, "step": 357500 }, { "epoch": 29.59, "eval_loss": 0.17231164872646332, "eval_runtime": 4262.071, "eval_samples_per_second": 17.737, "eval_steps_per_second": 1.478, "eval_wer": 0.0507014604592916, "step": 357500 }, { "epoch": 29.6, "learning_rate": 5.225893599334996e-06, "loss": 0.0181, "step": 357600 }, { "epoch": 29.61, "learning_rate": 5.2217373233582715e-06, "loss": 0.0157, "step": 357700 }, { "epoch": 29.62, "learning_rate": 5.217581047381547e-06, "loss": 0.0199, "step": 357800 }, { "epoch": 29.63, "learning_rate": 5.213424771404822e-06, "loss": 0.0242, "step": 357900 }, { "epoch": 29.64, "learning_rate": 5.209268495428097e-06, "loss": 0.0234, "step": 358000 }, { "epoch": 29.64, "eval_loss": 0.16698582470417023, "eval_runtime": 4252.1518, "eval_samples_per_second": 17.778, "eval_steps_per_second": 1.482, "eval_wer": 0.05036915327822573, "step": 358000 }, { "epoch": 29.64, "learning_rate": 5.205112219451371e-06, "loss": 0.0207, "step": 358100 }, { "epoch": 29.65, "learning_rate": 5.200955943474647e-06, "loss": 0.0237, "step": 358200 }, { "epoch": 29.66, "learning_rate": 5.196799667497923e-06, "loss": 0.0221, "step": 358300 }, { "epoch": 29.67, "learning_rate": 5.192643391521197e-06, "loss": 0.0226, "step": 358400 }, { "epoch": 29.68, "learning_rate": 5.188487115544473e-06, "loss": 0.0232, "step": 358500 }, { "epoch": 29.68, "eval_loss": 0.17163337767124176, "eval_runtime": 4275.6924, "eval_samples_per_second": 17.68, "eval_steps_per_second": 1.473, "eval_wer": 0.05018761509597677, "step": 358500 }, { "epoch": 29.69, "learning_rate": 5.184330839567747e-06, "loss": 0.0233, "step": 358600 }, { "epoch": 29.69, "learning_rate": 5.1801745635910225e-06, "loss": 0.0193, "step": 358700 }, { "epoch": 29.7, "learning_rate": 5.176018287614298e-06, "loss": 0.0204, "step": 358800 }, { "epoch": 29.71, "learning_rate": 5.171862011637573e-06, "loss": 0.0219, "step": 358900 }, { "epoch": 29.72, "learning_rate": 5.167705735660848e-06, "loss": 0.0193, "step": 359000 }, { "epoch": 29.72, "eval_loss": 0.17436064779758453, "eval_runtime": 4244.7219, "eval_samples_per_second": 17.809, "eval_steps_per_second": 1.484, "eval_wer": 0.05063992209242755, "step": 359000 }, { "epoch": 29.73, "learning_rate": 5.163549459684123e-06, "loss": 0.0173, "step": 359100 }, { "epoch": 29.74, "learning_rate": 5.159393183707398e-06, "loss": 0.0219, "step": 359200 }, { "epoch": 29.74, "learning_rate": 5.155236907730674e-06, "loss": 0.022, "step": 359300 }, { "epoch": 29.75, "learning_rate": 5.1510806317539485e-06, "loss": 0.0203, "step": 359400 }, { "epoch": 29.76, "learning_rate": 5.146965918536991e-06, "loss": 0.0193, "step": 359500 }, { "epoch": 29.76, "eval_loss": 0.17065052688121796, "eval_runtime": 4264.4361, "eval_samples_per_second": 17.727, "eval_steps_per_second": 1.477, "eval_wer": 0.04915992436934712, "step": 359500 }, { "epoch": 29.77, "learning_rate": 5.142809642560267e-06, "loss": 0.0172, "step": 359600 }, { "epoch": 29.78, "learning_rate": 5.138653366583541e-06, "loss": 0.0191, "step": 359700 }, { "epoch": 29.78, "learning_rate": 5.134497090606817e-06, "loss": 0.0166, "step": 359800 }, { "epoch": 29.79, "learning_rate": 5.1303408146300916e-06, "loss": 0.0196, "step": 359900 }, { "epoch": 29.8, "learning_rate": 5.126184538653367e-06, "loss": 0.0192, "step": 360000 }, { "epoch": 29.8, "eval_loss": 0.17323929071426392, "eval_runtime": 4260.5363, "eval_samples_per_second": 17.743, "eval_steps_per_second": 1.479, "eval_wer": 0.049079924492423856, "step": 360000 }, { "epoch": 29.81, "learning_rate": 5.122028262676643e-06, "loss": 0.0196, "step": 360100 }, { "epoch": 29.82, "learning_rate": 5.117871986699917e-06, "loss": 0.0198, "step": 360200 }, { "epoch": 29.83, "learning_rate": 5.113715710723193e-06, "loss": 0.0248, "step": 360300 }, { "epoch": 29.83, "learning_rate": 5.1095594347464675e-06, "loss": 0.022, "step": 360400 }, { "epoch": 29.84, "learning_rate": 5.105403158769743e-06, "loss": 0.0197, "step": 360500 }, { "epoch": 29.84, "eval_loss": 0.17387284338474274, "eval_runtime": 4261.1541, "eval_samples_per_second": 17.74, "eval_steps_per_second": 1.478, "eval_wer": 0.050162999749231156, "step": 360500 }, { "epoch": 29.85, "learning_rate": 5.1012468827930185e-06, "loss": 0.0168, "step": 360600 }, { "epoch": 29.86, "learning_rate": 5.097090606816293e-06, "loss": 0.0206, "step": 360700 }, { "epoch": 29.87, "learning_rate": 5.092934330839569e-06, "loss": 0.0194, "step": 360800 }, { "epoch": 29.88, "learning_rate": 5.088778054862843e-06, "loss": 0.0198, "step": 360900 }, { "epoch": 29.88, "learning_rate": 5.084621778886119e-06, "loss": 0.0196, "step": 361000 }, { "epoch": 29.88, "eval_loss": 0.1785241812467575, "eval_runtime": 4327.8083, "eval_samples_per_second": 17.467, "eval_steps_per_second": 1.456, "eval_wer": 0.04989376939420093, "step": 361000 }, { "epoch": 29.89, "learning_rate": 5.080465502909394e-06, "loss": 0.0185, "step": 361100 }, { "epoch": 29.9, "learning_rate": 5.076309226932668e-06, "loss": 0.0216, "step": 361200 }, { "epoch": 29.91, "learning_rate": 5.0721529509559446e-06, "loss": 0.0218, "step": 361300 }, { "epoch": 29.92, "learning_rate": 5.067996674979218e-06, "loss": 0.0186, "step": 361400 }, { "epoch": 29.93, "learning_rate": 5.063840399002494e-06, "loss": 0.0159, "step": 361500 }, { "epoch": 29.93, "eval_loss": 0.17497943341732025, "eval_runtime": 4259.3358, "eval_samples_per_second": 17.748, "eval_steps_per_second": 1.479, "eval_wer": 0.048601463690055864, "step": 361500 }, { "epoch": 29.93, "learning_rate": 5.059684123025769e-06, "loss": 0.0233, "step": 361600 }, { "epoch": 29.94, "learning_rate": 5.055527847049044e-06, "loss": 0.0198, "step": 361700 }, { "epoch": 29.95, "learning_rate": 5.05137157107232e-06, "loss": 0.0208, "step": 361800 }, { "epoch": 29.96, "learning_rate": 5.0472568578553614e-06, "loss": 0.0202, "step": 361900 }, { "epoch": 29.97, "learning_rate": 5.043100581878637e-06, "loss": 0.0201, "step": 362000 }, { "epoch": 29.97, "eval_loss": 0.1696506291627884, "eval_runtime": 4278.4179, "eval_samples_per_second": 17.669, "eval_steps_per_second": 1.473, "eval_wer": 0.04942761626520575, "step": 362000 }, { "epoch": 29.98, "learning_rate": 5.0389443059019125e-06, "loss": 0.0215, "step": 362100 }, { "epoch": 29.98, "learning_rate": 5.034788029925187e-06, "loss": 0.0189, "step": 362200 }, { "epoch": 29.99, "learning_rate": 5.030631753948463e-06, "loss": 0.0201, "step": 362300 }, { "epoch": 30.0, "learning_rate": 5.026475477971737e-06, "loss": 0.023, "step": 362400 }, { "epoch": 30.01, "learning_rate": 5.022319201995013e-06, "loss": 0.024, "step": 362500 }, { "epoch": 30.01, "eval_loss": 0.17689953744411469, "eval_runtime": 4277.2785, "eval_samples_per_second": 17.674, "eval_steps_per_second": 1.473, "eval_wer": 0.049278385725560425, "step": 362500 }, { "epoch": 30.02, "learning_rate": 5.018162926018288e-06, "loss": 0.019, "step": 362600 }, { "epoch": 30.02, "learning_rate": 5.014006650041563e-06, "loss": 0.0159, "step": 362700 }, { "epoch": 30.03, "learning_rate": 5.0098503740648385e-06, "loss": 0.022, "step": 362800 }, { "epoch": 30.04, "learning_rate": 5.005694098088113e-06, "loss": 0.0186, "step": 362900 }, { "epoch": 30.05, "learning_rate": 5.001537822111389e-06, "loss": 0.0196, "step": 363000 }, { "epoch": 30.05, "eval_loss": 0.17565162479877472, "eval_runtime": 4273.05, "eval_samples_per_second": 17.691, "eval_steps_per_second": 1.474, "eval_wer": 0.04914915515514592, "step": 363000 }, { "epoch": 30.06, "learning_rate": 4.997381546134663e-06, "loss": 0.0172, "step": 363100 }, { "epoch": 30.07, "learning_rate": 4.993225270157939e-06, "loss": 0.018, "step": 363200 }, { "epoch": 30.07, "learning_rate": 4.9890689941812144e-06, "loss": 0.0183, "step": 363300 }, { "epoch": 30.08, "learning_rate": 4.984912718204489e-06, "loss": 0.0173, "step": 363400 }, { "epoch": 30.09, "learning_rate": 4.980756442227765e-06, "loss": 0.0185, "step": 363500 }, { "epoch": 30.09, "eval_loss": 0.18026015162467957, "eval_runtime": 4268.4549, "eval_samples_per_second": 17.71, "eval_steps_per_second": 1.476, "eval_wer": 0.04938300094922931, "step": 363500 }, { "epoch": 30.1, "learning_rate": 4.976600166251039e-06, "loss": 0.0216, "step": 363600 }, { "epoch": 30.11, "learning_rate": 4.972443890274315e-06, "loss": 0.0215, "step": 363700 }, { "epoch": 30.12, "learning_rate": 4.96828761429759e-06, "loss": 0.0205, "step": 363800 }, { "epoch": 30.12, "learning_rate": 4.964131338320865e-06, "loss": 0.0216, "step": 363900 }, { "epoch": 30.13, "learning_rate": 4.95997506234414e-06, "loss": 0.0194, "step": 364000 }, { "epoch": 30.13, "eval_loss": 0.17700572311878204, "eval_runtime": 4247.9801, "eval_samples_per_second": 17.796, "eval_steps_per_second": 1.483, "eval_wer": 0.04963223133502871, "step": 364000 }, { "epoch": 30.14, "learning_rate": 4.955860349127182e-06, "loss": 0.0179, "step": 364100 }, { "epoch": 30.15, "learning_rate": 4.951704073150458e-06, "loss": 0.0174, "step": 364200 }, { "epoch": 30.16, "learning_rate": 4.9475893599335e-06, "loss": 0.0311, "step": 364300 }, { "epoch": 30.17, "learning_rate": 4.943433083956775e-06, "loss": 0.024, "step": 364400 }, { "epoch": 30.17, "learning_rate": 4.93927680798005e-06, "loss": 0.0191, "step": 364500 }, { "epoch": 30.17, "eval_loss": 0.1768103837966919, "eval_runtime": 4241.3088, "eval_samples_per_second": 17.824, "eval_steps_per_second": 1.485, "eval_wer": 0.04841531013029211, "step": 364500 }, { "epoch": 30.18, "learning_rate": 4.935120532003325e-06, "loss": 0.0187, "step": 364600 }, { "epoch": 30.19, "learning_rate": 4.930964256026601e-06, "loss": 0.0198, "step": 364700 }, { "epoch": 30.2, "learning_rate": 4.9268079800498756e-06, "loss": 0.0194, "step": 364800 }, { "epoch": 30.21, "learning_rate": 4.922651704073151e-06, "loss": 0.0187, "step": 364900 }, { "epoch": 30.22, "learning_rate": 4.918495428096426e-06, "loss": 0.025, "step": 365000 }, { "epoch": 30.22, "eval_loss": 0.1808198094367981, "eval_runtime": 4265.0427, "eval_samples_per_second": 17.724, "eval_steps_per_second": 1.477, "eval_wer": 0.05052607611372906, "step": 365000 }, { "epoch": 30.22, "learning_rate": 4.914339152119701e-06, "loss": 0.0202, "step": 365100 }, { "epoch": 30.23, "learning_rate": 4.910182876142977e-06, "loss": 0.0216, "step": 365200 }, { "epoch": 30.24, "learning_rate": 4.9060266001662514e-06, "loss": 0.0166, "step": 365300 }, { "epoch": 30.25, "learning_rate": 4.901870324189527e-06, "loss": 0.0182, "step": 365400 }, { "epoch": 30.26, "learning_rate": 4.897714048212802e-06, "loss": 0.0274, "step": 365500 }, { "epoch": 30.26, "eval_loss": 0.17277929186820984, "eval_runtime": 4260.0289, "eval_samples_per_second": 17.745, "eval_steps_per_second": 1.479, "eval_wer": 0.04850454076224498, "step": 365500 }, { "epoch": 30.26, "learning_rate": 4.893557772236076e-06, "loss": 0.0202, "step": 365600 }, { "epoch": 30.27, "learning_rate": 4.889401496259352e-06, "loss": 0.018, "step": 365700 }, { "epoch": 30.28, "learning_rate": 4.885245220282627e-06, "loss": 0.0186, "step": 365800 }, { "epoch": 30.29, "learning_rate": 4.881088944305902e-06, "loss": 0.0229, "step": 365900 }, { "epoch": 30.3, "learning_rate": 4.8769326683291775e-06, "loss": 0.017, "step": 366000 }, { "epoch": 30.3, "eval_loss": 0.1774836629629135, "eval_runtime": 4288.4443, "eval_samples_per_second": 17.628, "eval_steps_per_second": 1.469, "eval_wer": 0.04970607737526558, "step": 366000 }, { "epoch": 30.31, "learning_rate": 4.872776392352452e-06, "loss": 0.0192, "step": 366100 }, { "epoch": 30.31, "learning_rate": 4.868620116375728e-06, "loss": 0.0232, "step": 366200 }, { "epoch": 30.32, "learning_rate": 4.864463840399003e-06, "loss": 0.0189, "step": 366300 }, { "epoch": 30.33, "learning_rate": 4.860307564422278e-06, "loss": 0.0148, "step": 366400 }, { "epoch": 30.34, "learning_rate": 4.856151288445553e-06, "loss": 0.0255, "step": 366500 }, { "epoch": 30.34, "eval_loss": 0.17642787098884583, "eval_runtime": 4264.8867, "eval_samples_per_second": 17.725, "eval_steps_per_second": 1.477, "eval_wer": 0.05017684588177557, "step": 366500 }, { "epoch": 30.35, "learning_rate": 4.851995012468828e-06, "loss": 0.0204, "step": 366600 }, { "epoch": 30.36, "learning_rate": 4.847838736492104e-06, "loss": 0.0232, "step": 366700 }, { "epoch": 30.36, "learning_rate": 4.843682460515379e-06, "loss": 0.0204, "step": 366800 }, { "epoch": 30.37, "learning_rate": 4.839526184538654e-06, "loss": 0.0201, "step": 366900 }, { "epoch": 30.38, "learning_rate": 4.8353699085619284e-06, "loss": 0.0226, "step": 367000 }, { "epoch": 30.38, "eval_loss": 0.17326509952545166, "eval_runtime": 4335.0068, "eval_samples_per_second": 17.438, "eval_steps_per_second": 1.453, "eval_wer": 0.04921223198118157, "step": 367000 }, { "epoch": 30.39, "learning_rate": 4.831213632585204e-06, "loss": 0.0192, "step": 367100 }, { "epoch": 30.4, "learning_rate": 4.8270573566084795e-06, "loss": 0.0178, "step": 367200 }, { "epoch": 30.41, "learning_rate": 4.822901080631754e-06, "loss": 0.0207, "step": 367300 }, { "epoch": 30.41, "learning_rate": 4.81874480465503e-06, "loss": 0.0197, "step": 367400 }, { "epoch": 30.42, "learning_rate": 4.814588528678304e-06, "loss": 0.0194, "step": 367500 }, { "epoch": 30.42, "eval_loss": 0.18370088934898376, "eval_runtime": 4270.6421, "eval_samples_per_second": 17.701, "eval_steps_per_second": 1.475, "eval_wer": 0.05062607595988314, "step": 367500 }, { "epoch": 30.43, "learning_rate": 4.81043225270158e-06, "loss": 0.0213, "step": 367600 }, { "epoch": 30.44, "learning_rate": 4.806275976724855e-06, "loss": 0.0189, "step": 367700 }, { "epoch": 30.45, "learning_rate": 4.80211970074813e-06, "loss": 0.0188, "step": 367800 }, { "epoch": 30.46, "learning_rate": 4.7979634247714055e-06, "loss": 0.0203, "step": 367900 }, { "epoch": 30.46, "learning_rate": 4.79380714879468e-06, "loss": 0.0198, "step": 368000 }, { "epoch": 30.46, "eval_loss": 0.1802985519170761, "eval_runtime": 4268.9654, "eval_samples_per_second": 17.708, "eval_steps_per_second": 1.476, "eval_wer": 0.04929992415396284, "step": 368000 }, { "epoch": 30.47, "learning_rate": 4.789650872817955e-06, "loss": 0.0168, "step": 368100 }, { "epoch": 30.48, "learning_rate": 4.785494596841231e-06, "loss": 0.0234, "step": 368200 }, { "epoch": 30.49, "learning_rate": 4.781338320864506e-06, "loss": 0.0216, "step": 368300 }, { "epoch": 30.5, "learning_rate": 4.777223607647548e-06, "loss": 0.0199, "step": 368400 }, { "epoch": 30.5, "learning_rate": 4.773067331670823e-06, "loss": 0.0173, "step": 368500 }, { "epoch": 30.5, "eval_loss": 0.18490323424339294, "eval_runtime": 4263.9151, "eval_samples_per_second": 17.729, "eval_steps_per_second": 1.478, "eval_wer": 0.04954915453976225, "step": 368500 }, { "epoch": 30.51, "learning_rate": 4.768911055694099e-06, "loss": 0.0174, "step": 368600 }, { "epoch": 30.52, "learning_rate": 4.7647547797173734e-06, "loss": 0.0197, "step": 368700 }, { "epoch": 30.53, "learning_rate": 4.760598503740649e-06, "loss": 0.0195, "step": 368800 }, { "epoch": 30.54, "learning_rate": 4.756442227763924e-06, "loss": 0.0223, "step": 368900 }, { "epoch": 30.55, "learning_rate": 4.752285951787199e-06, "loss": 0.0203, "step": 369000 }, { "epoch": 30.55, "eval_loss": 0.18108709156513214, "eval_runtime": 4267.4925, "eval_samples_per_second": 17.714, "eval_steps_per_second": 1.476, "eval_wer": 0.04891684782023412, "step": 369000 }, { "epoch": 30.55, "learning_rate": 4.748129675810475e-06, "loss": 0.0185, "step": 369100 }, { "epoch": 30.56, "learning_rate": 4.743973399833749e-06, "loss": 0.0195, "step": 369200 }, { "epoch": 30.57, "learning_rate": 4.739817123857025e-06, "loss": 0.0216, "step": 369300 }, { "epoch": 30.58, "learning_rate": 4.7356608478802995e-06, "loss": 0.0217, "step": 369400 }, { "epoch": 30.59, "learning_rate": 4.731504571903574e-06, "loss": 0.0205, "step": 369500 }, { "epoch": 30.59, "eval_loss": 0.17219138145446777, "eval_runtime": 4279.1966, "eval_samples_per_second": 17.666, "eval_steps_per_second": 1.472, "eval_wer": 0.04909684754331147, "step": 369500 }, { "epoch": 30.6, "learning_rate": 4.7273482959268506e-06, "loss": 0.0179, "step": 369600 }, { "epoch": 30.6, "learning_rate": 4.723192019950125e-06, "loss": 0.0205, "step": 369700 }, { "epoch": 30.61, "learning_rate": 4.7190357439734e-06, "loss": 0.0169, "step": 369800 }, { "epoch": 30.62, "learning_rate": 4.714879467996675e-06, "loss": 0.0174, "step": 369900 }, { "epoch": 30.63, "learning_rate": 4.71072319201995e-06, "loss": 0.0191, "step": 370000 }, { "epoch": 30.63, "eval_loss": 0.1743849515914917, "eval_runtime": 4288.076, "eval_samples_per_second": 17.629, "eval_steps_per_second": 1.469, "eval_wer": 0.04883069410662445, "step": 370000 }, { "epoch": 30.64, "learning_rate": 4.706566916043226e-06, "loss": 0.0201, "step": 370100 }, { "epoch": 30.65, "learning_rate": 4.702410640066501e-06, "loss": 0.0173, "step": 370200 }, { "epoch": 30.65, "learning_rate": 4.698254364089776e-06, "loss": 0.0204, "step": 370300 }, { "epoch": 30.66, "learning_rate": 4.694098088113051e-06, "loss": 0.0218, "step": 370400 }, { "epoch": 30.67, "learning_rate": 4.689941812136326e-06, "loss": 0.0149, "step": 370500 }, { "epoch": 30.67, "eval_loss": 0.17745113372802734, "eval_runtime": 4262.3129, "eval_samples_per_second": 17.736, "eval_steps_per_second": 1.478, "eval_wer": 0.048279925723191196, "step": 370500 }, { "epoch": 30.68, "learning_rate": 4.6857855361596015e-06, "loss": 0.0191, "step": 370600 }, { "epoch": 30.69, "learning_rate": 4.681629260182877e-06, "loss": 0.0224, "step": 370700 }, { "epoch": 30.7, "learning_rate": 4.677472984206152e-06, "loss": 0.0169, "step": 370800 }, { "epoch": 30.7, "learning_rate": 4.673316708229426e-06, "loss": 0.0167, "step": 370900 }, { "epoch": 30.71, "learning_rate": 4.669160432252702e-06, "loss": 0.0216, "step": 371000 }, { "epoch": 30.71, "eval_loss": 0.17567051947116852, "eval_runtime": 4250.6346, "eval_samples_per_second": 17.784, "eval_steps_per_second": 1.482, "eval_wer": 0.04838146402851688, "step": 371000 }, { "epoch": 30.72, "learning_rate": 4.665004156275977e-06, "loss": 0.0238, "step": 371100 }, { "epoch": 30.73, "learning_rate": 4.660847880299252e-06, "loss": 0.0188, "step": 371200 }, { "epoch": 30.74, "learning_rate": 4.6566916043225276e-06, "loss": 0.0156, "step": 371300 }, { "epoch": 30.75, "learning_rate": 4.652576891105569e-06, "loss": 0.0181, "step": 371400 }, { "epoch": 30.75, "learning_rate": 4.648420615128845e-06, "loss": 0.0206, "step": 371500 }, { "epoch": 30.75, "eval_loss": 0.17855584621429443, "eval_runtime": 4255.5376, "eval_samples_per_second": 17.764, "eval_steps_per_second": 1.48, "eval_wer": 0.04801223382733257, "step": 371500 }, { "epoch": 30.76, "learning_rate": 4.64426433915212e-06, "loss": 0.0166, "step": 371600 }, { "epoch": 30.77, "learning_rate": 4.640108063175395e-06, "loss": 0.02, "step": 371700 }, { "epoch": 30.78, "learning_rate": 4.635951787198671e-06, "loss": 0.0206, "step": 371800 }, { "epoch": 30.79, "learning_rate": 4.631795511221945e-06, "loss": 0.0194, "step": 371900 }, { "epoch": 30.79, "learning_rate": 4.627639235245221e-06, "loss": 0.0169, "step": 372000 }, { "epoch": 30.79, "eval_loss": 0.1798625886440277, "eval_runtime": 4278.4342, "eval_samples_per_second": 17.669, "eval_steps_per_second": 1.473, "eval_wer": 0.04892761703443533, "step": 372000 }, { "epoch": 30.8, "learning_rate": 4.623482959268496e-06, "loss": 0.02, "step": 372100 }, { "epoch": 30.81, "learning_rate": 4.619326683291771e-06, "loss": 0.0204, "step": 372200 }, { "epoch": 30.82, "learning_rate": 4.6151704073150465e-06, "loss": 0.0267, "step": 372300 }, { "epoch": 30.83, "learning_rate": 4.611014131338321e-06, "loss": 0.0267, "step": 372400 }, { "epoch": 30.84, "learning_rate": 4.606857855361596e-06, "loss": 0.0237, "step": 372500 }, { "epoch": 30.84, "eval_loss": 0.17742688953876495, "eval_runtime": 4251.674, "eval_samples_per_second": 17.78, "eval_steps_per_second": 1.482, "eval_wer": 0.04905838606402144, "step": 372500 }, { "epoch": 30.84, "learning_rate": 4.602701579384871e-06, "loss": 0.0163, "step": 372600 }, { "epoch": 30.85, "learning_rate": 4.598545303408147e-06, "loss": 0.0188, "step": 372700 }, { "epoch": 30.86, "learning_rate": 4.5943890274314215e-06, "loss": 0.0207, "step": 372800 }, { "epoch": 30.87, "learning_rate": 4.590232751454697e-06, "loss": 0.0176, "step": 372900 }, { "epoch": 30.88, "learning_rate": 4.586076475477972e-06, "loss": 0.0187, "step": 373000 }, { "epoch": 30.88, "eval_loss": 0.1776452362537384, "eval_runtime": 4261.3982, "eval_samples_per_second": 17.739, "eval_steps_per_second": 1.478, "eval_wer": 0.04793684932792411, "step": 373000 }, { "epoch": 30.89, "learning_rate": 4.581920199501247e-06, "loss": 0.019, "step": 373100 }, { "epoch": 30.89, "learning_rate": 4.577763923524523e-06, "loss": 0.0246, "step": 373200 }, { "epoch": 30.9, "learning_rate": 4.573607647547797e-06, "loss": 0.0166, "step": 373300 }, { "epoch": 30.91, "learning_rate": 4.56949293433084e-06, "loss": 0.0165, "step": 373400 }, { "epoch": 30.92, "learning_rate": 4.565336658354116e-06, "loss": 0.0201, "step": 373500 }, { "epoch": 30.92, "eval_loss": 0.18362769484519958, "eval_runtime": 4268.3729, "eval_samples_per_second": 17.71, "eval_steps_per_second": 1.476, "eval_wer": 0.050492230011953826, "step": 373500 }, { "epoch": 30.93, "learning_rate": 4.56118038237739e-06, "loss": 0.0253, "step": 373600 }, { "epoch": 30.94, "learning_rate": 4.557024106400666e-06, "loss": 0.0171, "step": 373700 }, { "epoch": 30.94, "learning_rate": 4.5528678304239405e-06, "loss": 0.0239, "step": 373800 }, { "epoch": 30.95, "learning_rate": 4.548711554447215e-06, "loss": 0.0203, "step": 373900 }, { "epoch": 30.96, "learning_rate": 4.544555278470491e-06, "loss": 0.0181, "step": 374000 }, { "epoch": 30.96, "eval_loss": 0.17726416885852814, "eval_runtime": 4263.4222, "eval_samples_per_second": 17.731, "eval_steps_per_second": 1.478, "eval_wer": 0.0485260791906474, "step": 374000 }, { "epoch": 30.97, "learning_rate": 4.540399002493766e-06, "loss": 0.0208, "step": 374100 }, { "epoch": 30.98, "learning_rate": 4.536242726517041e-06, "loss": 0.0195, "step": 374200 }, { "epoch": 30.99, "learning_rate": 4.532086450540316e-06, "loss": 0.0178, "step": 374300 }, { "epoch": 30.99, "learning_rate": 4.527930174563591e-06, "loss": 0.0185, "step": 374400 }, { "epoch": 31.0, "learning_rate": 4.5237738985868665e-06, "loss": 0.0157, "step": 374500 }, { "epoch": 31.0, "eval_loss": 0.17786812782287598, "eval_runtime": 4253.8076, "eval_samples_per_second": 17.771, "eval_steps_per_second": 1.481, "eval_wer": 0.04812146442851627, "step": 374500 }, { "epoch": 31.01, "learning_rate": 4.519617622610142e-06, "loss": 0.0218, "step": 374600 }, { "epoch": 31.02, "learning_rate": 4.515461346633417e-06, "loss": 0.0152, "step": 374700 }, { "epoch": 31.03, "learning_rate": 4.511305070656692e-06, "loss": 0.023, "step": 374800 }, { "epoch": 31.03, "learning_rate": 4.507148794679967e-06, "loss": 0.0183, "step": 374900 }, { "epoch": 31.04, "learning_rate": 4.502992518703242e-06, "loss": 0.022, "step": 375000 }, { "epoch": 31.04, "eval_loss": 0.17090477049350739, "eval_runtime": 4258.9829, "eval_samples_per_second": 17.75, "eval_steps_per_second": 1.479, "eval_wer": 0.04740146553620687, "step": 375000 }, { "epoch": 31.05, "learning_rate": 4.498836242726518e-06, "loss": 0.019, "step": 375100 }, { "epoch": 31.06, "learning_rate": 4.494679966749793e-06, "loss": 0.0219, "step": 375200 }, { "epoch": 31.07, "learning_rate": 4.490523690773067e-06, "loss": 0.0198, "step": 375300 }, { "epoch": 31.08, "learning_rate": 4.48640897755611e-06, "loss": 0.0171, "step": 375400 }, { "epoch": 31.08, "learning_rate": 4.4822527015793855e-06, "loss": 0.0196, "step": 375500 }, { "epoch": 31.08, "eval_loss": 0.17020238935947418, "eval_runtime": 4265.6976, "eval_samples_per_second": 17.722, "eval_steps_per_second": 1.477, "eval_wer": 0.04810454137762865, "step": 375500 }, { "epoch": 31.09, "learning_rate": 4.47809642560266e-06, "loss": 0.0249, "step": 375600 }, { "epoch": 31.1, "learning_rate": 4.473940149625936e-06, "loss": 0.02, "step": 375700 }, { "epoch": 31.11, "learning_rate": 4.4698254364089775e-06, "loss": 0.0209, "step": 375800 }, { "epoch": 31.12, "learning_rate": 4.465669160432253e-06, "loss": 0.0179, "step": 375900 }, { "epoch": 31.13, "learning_rate": 4.4615128844555285e-06, "loss": 0.0167, "step": 376000 }, { "epoch": 31.13, "eval_loss": 0.1841665804386139, "eval_runtime": 4263.2176, "eval_samples_per_second": 17.732, "eval_steps_per_second": 1.478, "eval_wer": 0.048944540085322945, "step": 376000 }, { "epoch": 31.13, "learning_rate": 4.457356608478803e-06, "loss": 0.0189, "step": 376100 }, { "epoch": 31.14, "learning_rate": 4.453200332502079e-06, "loss": 0.0189, "step": 376200 }, { "epoch": 31.15, "learning_rate": 4.449044056525353e-06, "loss": 0.0185, "step": 376300 }, { "epoch": 31.16, "learning_rate": 4.444887780548629e-06, "loss": 0.0188, "step": 376400 }, { "epoch": 31.17, "learning_rate": 4.440731504571904e-06, "loss": 0.018, "step": 376500 }, { "epoch": 31.17, "eval_loss": 0.18488512933254242, "eval_runtime": 4264.2663, "eval_samples_per_second": 17.728, "eval_steps_per_second": 1.477, "eval_wer": 0.04867223281194952, "step": 376500 }, { "epoch": 31.18, "learning_rate": 4.436575228595179e-06, "loss": 0.0188, "step": 376600 }, { "epoch": 31.18, "learning_rate": 4.432418952618454e-06, "loss": 0.0171, "step": 376700 }, { "epoch": 31.19, "learning_rate": 4.428262676641729e-06, "loss": 0.0187, "step": 376800 }, { "epoch": 31.2, "learning_rate": 4.424106400665005e-06, "loss": 0.0215, "step": 376900 }, { "epoch": 31.21, "learning_rate": 4.4199501246882794e-06, "loss": 0.0168, "step": 377000 }, { "epoch": 31.21, "eval_loss": 0.18045353889465332, "eval_runtime": 4253.1017, "eval_samples_per_second": 17.774, "eval_steps_per_second": 1.481, "eval_wer": 0.04928453956224683, "step": 377000 }, { "epoch": 31.22, "learning_rate": 4.415793848711555e-06, "loss": 0.0223, "step": 377100 }, { "epoch": 31.23, "learning_rate": 4.41163757273483e-06, "loss": 0.0168, "step": 377200 }, { "epoch": 31.23, "learning_rate": 4.407481296758105e-06, "loss": 0.0163, "step": 377300 }, { "epoch": 31.24, "learning_rate": 4.403325020781381e-06, "loss": 0.0161, "step": 377400 }, { "epoch": 31.25, "learning_rate": 4.399168744804655e-06, "loss": 0.0175, "step": 377500 }, { "epoch": 31.25, "eval_loss": 0.18924129009246826, "eval_runtime": 4258.7776, "eval_samples_per_second": 17.75, "eval_steps_per_second": 1.479, "eval_wer": 0.04984300024153809, "step": 377500 }, { "epoch": 31.26, "learning_rate": 4.395012468827931e-06, "loss": 0.0174, "step": 377600 }, { "epoch": 31.27, "learning_rate": 4.3908561928512055e-06, "loss": 0.0198, "step": 377700 }, { "epoch": 31.27, "learning_rate": 4.38669991687448e-06, "loss": 0.0187, "step": 377800 }, { "epoch": 31.28, "learning_rate": 4.3825436408977565e-06, "loss": 0.0211, "step": 377900 }, { "epoch": 31.29, "learning_rate": 4.378387364921031e-06, "loss": 0.0188, "step": 378000 }, { "epoch": 31.29, "eval_loss": 0.18072986602783203, "eval_runtime": 4269.7007, "eval_samples_per_second": 17.705, "eval_steps_per_second": 1.476, "eval_wer": 0.04839223324271809, "step": 378000 }, { "epoch": 31.3, "learning_rate": 4.374231088944306e-06, "loss": 0.0209, "step": 378100 }, { "epoch": 31.31, "learning_rate": 4.370074812967581e-06, "loss": 0.018, "step": 378200 }, { "epoch": 31.32, "learning_rate": 4.365918536990856e-06, "loss": 0.0181, "step": 378300 }, { "epoch": 31.32, "learning_rate": 4.361762261014132e-06, "loss": 0.018, "step": 378400 }, { "epoch": 31.33, "learning_rate": 4.357605985037407e-06, "loss": 0.0179, "step": 378500 }, { "epoch": 31.33, "eval_loss": 0.1798219531774521, "eval_runtime": 4288.731, "eval_samples_per_second": 17.626, "eval_steps_per_second": 1.469, "eval_wer": 0.04919223201195075, "step": 378500 }, { "epoch": 31.34, "learning_rate": 4.353449709060682e-06, "loss": 0.0199, "step": 378600 }, { "epoch": 31.35, "learning_rate": 4.349293433083957e-06, "loss": 0.02, "step": 378700 }, { "epoch": 31.36, "learning_rate": 4.345137157107232e-06, "loss": 0.0178, "step": 378800 }, { "epoch": 31.37, "learning_rate": 4.3409808811305075e-06, "loss": 0.0173, "step": 378900 }, { "epoch": 31.37, "learning_rate": 4.336824605153783e-06, "loss": 0.0159, "step": 379000 }, { "epoch": 31.37, "eval_loss": 0.18698078393936157, "eval_runtime": 4264.1081, "eval_samples_per_second": 17.728, "eval_steps_per_second": 1.477, "eval_wer": 0.0491260782675719, "step": 379000 }, { "epoch": 31.38, "learning_rate": 4.332668329177058e-06, "loss": 0.016, "step": 379100 }, { "epoch": 31.39, "learning_rate": 4.328512053200332e-06, "loss": 0.0163, "step": 379200 }, { "epoch": 31.4, "learning_rate": 4.324355777223608e-06, "loss": 0.0158, "step": 379300 }, { "epoch": 31.41, "learning_rate": 4.3202410640066505e-06, "loss": 0.0209, "step": 379400 }, { "epoch": 31.42, "learning_rate": 4.316084788029925e-06, "loss": 0.0205, "step": 379500 }, { "epoch": 31.42, "eval_loss": 0.1824035495519638, "eval_runtime": 4292.7259, "eval_samples_per_second": 17.61, "eval_steps_per_second": 1.468, "eval_wer": 0.04886300174922808, "step": 379500 }, { "epoch": 31.42, "learning_rate": 4.311928512053201e-06, "loss": 0.0204, "step": 379600 }, { "epoch": 31.43, "learning_rate": 4.307772236076475e-06, "loss": 0.0203, "step": 379700 }, { "epoch": 31.44, "learning_rate": 4.303615960099751e-06, "loss": 0.0151, "step": 379800 }, { "epoch": 31.45, "learning_rate": 4.299459684123026e-06, "loss": 0.0187, "step": 379900 }, { "epoch": 31.46, "learning_rate": 4.295303408146301e-06, "loss": 0.019, "step": 380000 }, { "epoch": 31.46, "eval_loss": 0.18226408958435059, "eval_runtime": 4575.2019, "eval_samples_per_second": 16.523, "eval_steps_per_second": 1.377, "eval_wer": 0.04934607792911088, "step": 380000 }, { "epoch": 31.47, "learning_rate": 4.291147132169577e-06, "loss": 0.0206, "step": 380100 }, { "epoch": 31.47, "learning_rate": 4.286990856192851e-06, "loss": 0.0202, "step": 380200 }, { "epoch": 31.48, "learning_rate": 4.282834580216127e-06, "loss": 0.0192, "step": 380300 }, { "epoch": 31.49, "learning_rate": 4.278678304239402e-06, "loss": 0.0188, "step": 380400 }, { "epoch": 31.5, "learning_rate": 4.274522028262677e-06, "loss": 0.0234, "step": 380500 }, { "epoch": 31.5, "eval_loss": 0.17939870059490204, "eval_runtime": 4359.7119, "eval_samples_per_second": 17.339, "eval_steps_per_second": 1.445, "eval_wer": 0.04820761814212594, "step": 380500 }, { "epoch": 31.51, "learning_rate": 4.2703657522859525e-06, "loss": 0.0177, "step": 380600 }, { "epoch": 31.51, "learning_rate": 4.266209476309227e-06, "loss": 0.0155, "step": 380700 }, { "epoch": 31.52, "learning_rate": 4.262053200332503e-06, "loss": 0.0175, "step": 380800 }, { "epoch": 31.53, "learning_rate": 4.257896924355777e-06, "loss": 0.0172, "step": 380900 }, { "epoch": 31.54, "learning_rate": 4.253740648379053e-06, "loss": 0.0209, "step": 381000 }, { "epoch": 31.54, "eval_loss": 0.18396584689617157, "eval_runtime": 4337.3149, "eval_samples_per_second": 17.429, "eval_steps_per_second": 1.453, "eval_wer": 0.04906915527822265, "step": 381000 }, { "epoch": 31.55, "learning_rate": 4.2495843724023275e-06, "loss": 0.0202, "step": 381100 }, { "epoch": 31.56, "learning_rate": 4.245428096425603e-06, "loss": 0.0193, "step": 381200 }, { "epoch": 31.56, "learning_rate": 4.2412718204488786e-06, "loss": 0.0208, "step": 381300 }, { "epoch": 31.57, "learning_rate": 4.237115544472153e-06, "loss": 0.0151, "step": 381400 }, { "epoch": 31.58, "learning_rate": 4.232959268495429e-06, "loss": 0.0179, "step": 381500 }, { "epoch": 31.58, "eval_loss": 0.1791081726551056, "eval_runtime": 4339.1452, "eval_samples_per_second": 17.422, "eval_steps_per_second": 1.452, "eval_wer": 0.04830761798828002, "step": 381500 }, { "epoch": 31.59, "learning_rate": 4.228802992518703e-06, "loss": 0.0219, "step": 381600 }, { "epoch": 31.6, "learning_rate": 4.224646716541979e-06, "loss": 0.0166, "step": 381700 }, { "epoch": 31.61, "learning_rate": 4.2204904405652544e-06, "loss": 0.0169, "step": 381800 }, { "epoch": 31.61, "learning_rate": 4.216334164588529e-06, "loss": 0.019, "step": 381900 }, { "epoch": 31.62, "learning_rate": 4.212177888611805e-06, "loss": 0.017, "step": 382000 }, { "epoch": 31.62, "eval_loss": 0.1858493983745575, "eval_runtime": 4334.6385, "eval_samples_per_second": 17.44, "eval_steps_per_second": 1.453, "eval_wer": 0.04897377080958337, "step": 382000 }, { "epoch": 31.63, "learning_rate": 4.208021612635079e-06, "loss": 0.0244, "step": 382100 }, { "epoch": 31.64, "learning_rate": 4.203865336658354e-06, "loss": 0.0213, "step": 382200 }, { "epoch": 31.65, "learning_rate": 4.1997090606816295e-06, "loss": 0.0176, "step": 382300 }, { "epoch": 31.66, "learning_rate": 4.195552784704905e-06, "loss": 0.0221, "step": 382400 }, { "epoch": 31.66, "learning_rate": 4.19139650872818e-06, "loss": 0.0194, "step": 382500 }, { "epoch": 31.66, "eval_loss": 0.1883159577846527, "eval_runtime": 4338.2867, "eval_samples_per_second": 17.425, "eval_steps_per_second": 1.452, "eval_wer": 0.049641462090058326, "step": 382500 }, { "epoch": 31.67, "learning_rate": 4.187240232751455e-06, "loss": 0.0198, "step": 382600 }, { "epoch": 31.68, "learning_rate": 4.18308395677473e-06, "loss": 0.0234, "step": 382700 }, { "epoch": 31.69, "learning_rate": 4.1789692435577725e-06, "loss": 0.0207, "step": 382800 }, { "epoch": 31.7, "learning_rate": 4.174812967581048e-06, "loss": 0.018, "step": 382900 }, { "epoch": 31.71, "learning_rate": 4.170656691604323e-06, "loss": 0.0194, "step": 383000 }, { "epoch": 31.71, "eval_loss": 0.18673014640808105, "eval_runtime": 4346.7245, "eval_samples_per_second": 17.391, "eval_steps_per_second": 1.449, "eval_wer": 0.049087616788281864, "step": 383000 }, { "epoch": 31.71, "learning_rate": 4.166500415627598e-06, "loss": 0.0156, "step": 383100 }, { "epoch": 31.72, "learning_rate": 4.162344139650873e-06, "loss": 0.0165, "step": 383200 }, { "epoch": 31.73, "learning_rate": 4.158187863674148e-06, "loss": 0.0266, "step": 383300 }, { "epoch": 31.74, "learning_rate": 4.154031587697424e-06, "loss": 0.0184, "step": 383400 }, { "epoch": 31.75, "learning_rate": 4.149875311720699e-06, "loss": 0.0202, "step": 383500 }, { "epoch": 31.75, "eval_loss": 0.18335460126399994, "eval_runtime": 4355.0378, "eval_samples_per_second": 17.358, "eval_steps_per_second": 1.447, "eval_wer": 0.04839223324271809, "step": 383500 }, { "epoch": 31.75, "learning_rate": 4.145719035743973e-06, "loss": 0.0202, "step": 383600 }, { "epoch": 31.76, "learning_rate": 4.141562759767249e-06, "loss": 0.0231, "step": 383700 }, { "epoch": 31.77, "learning_rate": 4.137406483790524e-06, "loss": 0.0221, "step": 383800 }, { "epoch": 31.78, "learning_rate": 4.133250207813799e-06, "loss": 0.0183, "step": 383900 }, { "epoch": 31.79, "learning_rate": 4.1290939318370745e-06, "loss": 0.0162, "step": 384000 }, { "epoch": 31.79, "eval_loss": 0.1811029314994812, "eval_runtime": 4453.0436, "eval_samples_per_second": 16.976, "eval_steps_per_second": 1.415, "eval_wer": 0.04879684800484922, "step": 384000 }, { "epoch": 31.8, "learning_rate": 4.124937655860349e-06, "loss": 0.0166, "step": 384100 }, { "epoch": 31.8, "learning_rate": 4.120781379883625e-06, "loss": 0.02, "step": 384200 }, { "epoch": 31.81, "learning_rate": 4.1166251039069e-06, "loss": 0.0169, "step": 384300 }, { "epoch": 31.82, "learning_rate": 4.112468827930175e-06, "loss": 0.0152, "step": 384400 }, { "epoch": 31.83, "learning_rate": 4.10831255195345e-06, "loss": 0.0183, "step": 384500 }, { "epoch": 31.83, "eval_loss": 0.17847386002540588, "eval_runtime": 4458.3908, "eval_samples_per_second": 16.956, "eval_steps_per_second": 1.413, "eval_wer": 0.04783377256342682, "step": 384500 }, { "epoch": 31.84, "learning_rate": 4.104156275976725e-06, "loss": 0.0189, "step": 384600 }, { "epoch": 31.85, "learning_rate": 4.1e-06, "loss": 0.0217, "step": 384700 }, { "epoch": 31.85, "learning_rate": 4.095843724023276e-06, "loss": 0.0198, "step": 384800 }, { "epoch": 31.86, "learning_rate": 4.091687448046551e-06, "loss": 0.0197, "step": 384900 }, { "epoch": 31.87, "learning_rate": 4.087531172069825e-06, "loss": 0.0172, "step": 385000 }, { "epoch": 31.87, "eval_loss": 0.17977607250213623, "eval_runtime": 4458.2623, "eval_samples_per_second": 16.956, "eval_steps_per_second": 1.413, "eval_wer": 0.0479276185728945, "step": 385000 }, { "epoch": 31.88, "learning_rate": 4.083374896093101e-06, "loss": 0.0156, "step": 385100 }, { "epoch": 31.89, "learning_rate": 4.079260182876144e-06, "loss": 0.0158, "step": 385200 }, { "epoch": 31.9, "learning_rate": 4.075103906899418e-06, "loss": 0.0169, "step": 385300 }, { "epoch": 31.9, "learning_rate": 4.070947630922694e-06, "loss": 0.0181, "step": 385400 }, { "epoch": 31.91, "learning_rate": 4.0667913549459685e-06, "loss": 0.0184, "step": 385500 }, { "epoch": 31.91, "eval_loss": 0.17723698914051056, "eval_runtime": 4450.8959, "eval_samples_per_second": 16.984, "eval_steps_per_second": 1.415, "eval_wer": 0.04792915703206611, "step": 385500 }, { "epoch": 31.92, "learning_rate": 4.062635078969244e-06, "loss": 0.017, "step": 385600 }, { "epoch": 31.93, "learning_rate": 4.0584788029925195e-06, "loss": 0.0192, "step": 385700 }, { "epoch": 31.94, "learning_rate": 4.054322527015794e-06, "loss": 0.0214, "step": 385800 }, { "epoch": 31.95, "learning_rate": 4.050207813798837e-06, "loss": 0.0216, "step": 385900 }, { "epoch": 31.95, "learning_rate": 4.0460515378221115e-06, "loss": 0.0178, "step": 386000 }, { "epoch": 31.95, "eval_loss": 0.17839497327804565, "eval_runtime": 4494.0105, "eval_samples_per_second": 16.821, "eval_steps_per_second": 1.402, "eval_wer": 0.04844300239538093, "step": 386000 }, { "epoch": 31.96, "learning_rate": 4.041895261845387e-06, "loss": 0.0188, "step": 386100 }, { "epoch": 31.97, "learning_rate": 4.0377389858686625e-06, "loss": 0.0193, "step": 386200 }, { "epoch": 31.98, "learning_rate": 4.033582709891937e-06, "loss": 0.0179, "step": 386300 }, { "epoch": 31.99, "learning_rate": 4.029426433915212e-06, "loss": 0.0203, "step": 386400 }, { "epoch": 32.0, "learning_rate": 4.025270157938487e-06, "loss": 0.0145, "step": 386500 }, { "epoch": 32.0, "eval_loss": 0.19245170056819916, "eval_runtime": 4440.3173, "eval_samples_per_second": 17.025, "eval_steps_per_second": 1.419, "eval_wer": 0.04926146267467281, "step": 386500 }, { "epoch": 32.0, "learning_rate": 4.021113881961762e-06, "loss": 0.0231, "step": 386600 }, { "epoch": 32.01, "learning_rate": 4.0169576059850376e-06, "loss": 0.0179, "step": 386700 }, { "epoch": 32.02, "learning_rate": 4.012801330008313e-06, "loss": 0.0227, "step": 386800 }, { "epoch": 32.03, "learning_rate": 4.008645054031588e-06, "loss": 0.0198, "step": 386900 }, { "epoch": 32.04, "learning_rate": 4.004488778054863e-06, "loss": 0.0168, "step": 387000 }, { "epoch": 32.04, "eval_loss": 0.19617211818695068, "eval_runtime": 4441.1058, "eval_samples_per_second": 17.022, "eval_steps_per_second": 1.419, "eval_wer": 0.04945684698946617, "step": 387000 }, { "epoch": 32.04, "learning_rate": 4.000332502078138e-06, "loss": 0.0171, "step": 387100 }, { "epoch": 32.05, "learning_rate": 3.9961762261014135e-06, "loss": 0.0229, "step": 387200 }, { "epoch": 32.06, "learning_rate": 3.992019950124689e-06, "loss": 0.0156, "step": 387300 }, { "epoch": 32.07, "learning_rate": 3.987863674147964e-06, "loss": 0.0219, "step": 387400 }, { "epoch": 32.08, "learning_rate": 3.983707398171239e-06, "loss": 0.021, "step": 387500 }, { "epoch": 32.08, "eval_loss": 0.18802021443843842, "eval_runtime": 4456.3183, "eval_samples_per_second": 16.964, "eval_steps_per_second": 1.414, "eval_wer": 0.050064538362248674, "step": 387500 }, { "epoch": 32.09, "learning_rate": 3.979551122194514e-06, "loss": 0.0157, "step": 387600 }, { "epoch": 32.09, "learning_rate": 3.975394846217789e-06, "loss": 0.0186, "step": 387700 }, { "epoch": 32.1, "learning_rate": 3.971238570241064e-06, "loss": 0.0189, "step": 387800 }, { "epoch": 32.11, "learning_rate": 3.9670822942643395e-06, "loss": 0.0191, "step": 387900 }, { "epoch": 32.12, "learning_rate": 3.962926018287614e-06, "loss": 0.0194, "step": 388000 }, { "epoch": 32.12, "eval_loss": 0.18474696576595306, "eval_runtime": 4418.49, "eval_samples_per_second": 17.109, "eval_steps_per_second": 1.426, "eval_wer": 0.04902300150307461, "step": 388000 }, { "epoch": 32.13, "learning_rate": 3.95876974231089e-06, "loss": 0.016, "step": 388100 }, { "epoch": 32.14, "learning_rate": 3.954613466334165e-06, "loss": 0.0199, "step": 388200 }, { "epoch": 32.14, "learning_rate": 3.95045719035744e-06, "loss": 0.0167, "step": 388300 }, { "epoch": 32.15, "learning_rate": 3.946300914380715e-06, "loss": 0.0198, "step": 388400 }, { "epoch": 32.16, "learning_rate": 3.94214463840399e-06, "loss": 0.0184, "step": 388500 }, { "epoch": 32.16, "eval_loss": 0.18392613530158997, "eval_runtime": 4436.8769, "eval_samples_per_second": 17.038, "eval_steps_per_second": 1.42, "eval_wer": 0.04890607860603292, "step": 388500 }, { "epoch": 32.17, "learning_rate": 3.937988362427266e-06, "loss": 0.0191, "step": 388600 }, { "epoch": 32.18, "learning_rate": 3.933832086450541e-06, "loss": 0.0164, "step": 388700 }, { "epoch": 32.19, "learning_rate": 3.929675810473816e-06, "loss": 0.021, "step": 388800 }, { "epoch": 32.19, "learning_rate": 3.925519534497091e-06, "loss": 0.0179, "step": 388900 }, { "epoch": 32.2, "learning_rate": 3.921363258520366e-06, "loss": 0.0185, "step": 389000 }, { "epoch": 32.2, "eval_loss": 0.18545901775360107, "eval_runtime": 4424.2899, "eval_samples_per_second": 17.086, "eval_steps_per_second": 1.424, "eval_wer": 0.04959992369242509, "step": 389000 }, { "epoch": 32.21, "learning_rate": 3.917206982543641e-06, "loss": 0.0198, "step": 389100 }, { "epoch": 32.22, "learning_rate": 3.913050706566916e-06, "loss": 0.0167, "step": 389200 }, { "epoch": 32.23, "learning_rate": 3.908894430590192e-06, "loss": 0.0191, "step": 389300 }, { "epoch": 32.24, "learning_rate": 3.904738154613466e-06, "loss": 0.0168, "step": 389400 }, { "epoch": 32.24, "learning_rate": 3.900581878636742e-06, "loss": 0.0239, "step": 389500 }, { "epoch": 32.24, "eval_loss": 0.1817357838153839, "eval_runtime": 4454.9463, "eval_samples_per_second": 16.969, "eval_steps_per_second": 1.414, "eval_wer": 0.049393770163430516, "step": 389500 }, { "epoch": 32.25, "learning_rate": 3.8964256026600165e-06, "loss": 0.0189, "step": 389600 }, { "epoch": 32.26, "learning_rate": 3.892269326683292e-06, "loss": 0.0195, "step": 389700 }, { "epoch": 32.27, "learning_rate": 3.8881130507065676e-06, "loss": 0.0142, "step": 389800 }, { "epoch": 32.28, "learning_rate": 3.883956774729842e-06, "loss": 0.0161, "step": 389900 }, { "epoch": 32.28, "learning_rate": 3.879842061512885e-06, "loss": 0.0196, "step": 390000 }, { "epoch": 32.28, "eval_loss": 0.185058131814003, "eval_runtime": 4483.3014, "eval_samples_per_second": 16.861, "eval_steps_per_second": 1.405, "eval_wer": 0.049343001010767676, "step": 390000 }, { "epoch": 32.29, "learning_rate": 3.8756857855361604e-06, "loss": 0.0151, "step": 390100 }, { "epoch": 32.3, "learning_rate": 3.871529509559435e-06, "loss": 0.0165, "step": 390200 }, { "epoch": 32.31, "learning_rate": 3.867373233582711e-06, "loss": 0.0159, "step": 390300 }, { "epoch": 32.32, "learning_rate": 3.863216957605985e-06, "loss": 0.0148, "step": 390400 }, { "epoch": 32.33, "learning_rate": 3.85906068162926e-06, "loss": 0.0193, "step": 390500 }, { "epoch": 32.33, "eval_loss": 0.1858201026916504, "eval_runtime": 4650.2158, "eval_samples_per_second": 16.256, "eval_steps_per_second": 1.355, "eval_wer": 0.04967530819183355, "step": 390500 }, { "epoch": 32.33, "learning_rate": 3.8549044056525355e-06, "loss": 0.0216, "step": 390600 }, { "epoch": 32.34, "learning_rate": 3.850748129675811e-06, "loss": 0.0178, "step": 390700 }, { "epoch": 32.35, "learning_rate": 3.846591853699086e-06, "loss": 0.0201, "step": 390800 }, { "epoch": 32.36, "learning_rate": 3.842435577722361e-06, "loss": 0.0167, "step": 390900 }, { "epoch": 32.37, "learning_rate": 3.838279301745636e-06, "loss": 0.0218, "step": 391000 }, { "epoch": 32.37, "eval_loss": 0.1770581752061844, "eval_runtime": 4855.3537, "eval_samples_per_second": 15.569, "eval_steps_per_second": 1.298, "eval_wer": 0.04871069429123955, "step": 391000 }, { "epoch": 32.38, "learning_rate": 3.834123025768911e-06, "loss": 0.0149, "step": 391100 }, { "epoch": 32.38, "learning_rate": 3.829966749792187e-06, "loss": 0.0163, "step": 391200 }, { "epoch": 32.39, "learning_rate": 3.8258104738154615e-06, "loss": 0.0192, "step": 391300 }, { "epoch": 32.4, "learning_rate": 3.821654197838737e-06, "loss": 0.016, "step": 391400 }, { "epoch": 32.41, "learning_rate": 3.817497921862012e-06, "loss": 0.017, "step": 391500 }, { "epoch": 32.41, "eval_loss": 0.18436600267887115, "eval_runtime": 4894.5463, "eval_samples_per_second": 15.445, "eval_steps_per_second": 1.287, "eval_wer": 0.04871992504626916, "step": 391500 }, { "epoch": 32.42, "learning_rate": 3.8133416458852872e-06, "loss": 0.0164, "step": 391600 }, { "epoch": 32.43, "learning_rate": 3.8091853699085623e-06, "loss": 0.0156, "step": 391700 }, { "epoch": 32.43, "learning_rate": 3.8050290939318374e-06, "loss": 0.0169, "step": 391800 }, { "epoch": 32.44, "learning_rate": 3.8008728179551125e-06, "loss": 0.0216, "step": 391900 }, { "epoch": 32.45, "learning_rate": 3.7967165419783876e-06, "loss": 0.0195, "step": 392000 }, { "epoch": 32.45, "eval_loss": 0.1789231151342392, "eval_runtime": 4784.7076, "eval_samples_per_second": 15.799, "eval_steps_per_second": 1.317, "eval_wer": 0.04798300310307215, "step": 392000 }, { "epoch": 32.46, "learning_rate": 3.792560266001663e-06, "loss": 0.019, "step": 392100 }, { "epoch": 32.47, "learning_rate": 3.7884039900249382e-06, "loss": 0.0159, "step": 392200 }, { "epoch": 32.48, "learning_rate": 3.7842477140482133e-06, "loss": 0.0175, "step": 392300 }, { "epoch": 32.48, "learning_rate": 3.780091438071488e-06, "loss": 0.0195, "step": 392400 }, { "epoch": 32.49, "learning_rate": 3.775935162094763e-06, "loss": 0.0194, "step": 392500 }, { "epoch": 32.49, "eval_loss": 0.17805065214633942, "eval_runtime": 4694.2088, "eval_samples_per_second": 16.104, "eval_steps_per_second": 1.342, "eval_wer": 0.0482860795598776, "step": 392500 }, { "epoch": 32.5, "learning_rate": 3.7718204488778058e-06, "loss": 0.0177, "step": 392600 }, { "epoch": 32.51, "learning_rate": 3.767664172901081e-06, "loss": 0.016, "step": 392700 }, { "epoch": 32.52, "learning_rate": 3.763507896924356e-06, "loss": 0.0186, "step": 392800 }, { "epoch": 32.52, "learning_rate": 3.759351620947631e-06, "loss": 0.017, "step": 392900 }, { "epoch": 32.53, "learning_rate": 3.755195344970906e-06, "loss": 0.0136, "step": 393000 }, { "epoch": 32.53, "eval_loss": 0.1806943565607071, "eval_runtime": 4674.5871, "eval_samples_per_second": 16.171, "eval_steps_per_second": 1.348, "eval_wer": 0.04881069413739363, "step": 393000 }, { "epoch": 32.54, "learning_rate": 3.7510390689941816e-06, "loss": 0.016, "step": 393100 }, { "epoch": 32.55, "learning_rate": 3.7468827930174567e-06, "loss": 0.0206, "step": 393200 }, { "epoch": 32.56, "learning_rate": 3.742726517040732e-06, "loss": 0.0178, "step": 393300 }, { "epoch": 32.57, "learning_rate": 3.738570241064007e-06, "loss": 0.0165, "step": 393400 }, { "epoch": 32.57, "learning_rate": 3.7344139650872824e-06, "loss": 0.0191, "step": 393500 }, { "epoch": 32.57, "eval_loss": 0.18050839006900787, "eval_runtime": 4939.7503, "eval_samples_per_second": 15.303, "eval_steps_per_second": 1.275, "eval_wer": 0.04927684726638882, "step": 393500 }, { "epoch": 32.58, "learning_rate": 3.7302576891105575e-06, "loss": 0.0155, "step": 393600 }, { "epoch": 32.59, "learning_rate": 3.7261014131338326e-06, "loss": 0.0162, "step": 393700 }, { "epoch": 32.6, "learning_rate": 3.7219451371571073e-06, "loss": 0.0185, "step": 393800 }, { "epoch": 32.61, "learning_rate": 3.7177888611803824e-06, "loss": 0.0184, "step": 393900 }, { "epoch": 32.62, "learning_rate": 3.713632585203658e-06, "loss": 0.0156, "step": 394000 }, { "epoch": 32.62, "eval_loss": 0.1852325201034546, "eval_runtime": 4749.4735, "eval_samples_per_second": 15.917, "eval_steps_per_second": 1.326, "eval_wer": 0.049066078359879445, "step": 394000 }, { "epoch": 32.62, "learning_rate": 3.709476309226933e-06, "loss": 0.0189, "step": 394100 }, { "epoch": 32.63, "learning_rate": 3.705320033250208e-06, "loss": 0.0166, "step": 394200 }, { "epoch": 32.64, "learning_rate": 3.701163757273483e-06, "loss": 0.0181, "step": 394300 }, { "epoch": 32.65, "learning_rate": 3.6970074812967583e-06, "loss": 0.017, "step": 394400 }, { "epoch": 32.66, "learning_rate": 3.6928512053200338e-06, "loss": 0.0156, "step": 394500 }, { "epoch": 32.66, "eval_loss": 0.18618862330913544, "eval_runtime": 4653.2137, "eval_samples_per_second": 16.246, "eval_steps_per_second": 1.354, "eval_wer": 0.04922453965455438, "step": 394500 }, { "epoch": 32.67, "learning_rate": 3.688694929343309e-06, "loss": 0.0183, "step": 394600 }, { "epoch": 32.67, "learning_rate": 3.684538653366584e-06, "loss": 0.0209, "step": 394700 }, { "epoch": 32.68, "learning_rate": 3.680382377389859e-06, "loss": 0.0151, "step": 394800 }, { "epoch": 32.69, "learning_rate": 3.676226101413134e-06, "loss": 0.0156, "step": 394900 }, { "epoch": 32.7, "learning_rate": 3.6720698254364097e-06, "loss": 0.0182, "step": 395000 }, { "epoch": 32.7, "eval_loss": 0.18999694287776947, "eval_runtime": 4687.3447, "eval_samples_per_second": 16.127, "eval_steps_per_second": 1.344, "eval_wer": 0.04994453854686377, "step": 395000 }, { "epoch": 32.71, "learning_rate": 3.6679135494596848e-06, "loss": 0.0187, "step": 395100 }, { "epoch": 32.72, "learning_rate": 3.6637572734829594e-06, "loss": 0.0189, "step": 395200 }, { "epoch": 32.72, "learning_rate": 3.6596009975062345e-06, "loss": 0.0183, "step": 395300 }, { "epoch": 32.73, "learning_rate": 3.6554447215295096e-06, "loss": 0.0174, "step": 395400 }, { "epoch": 32.74, "learning_rate": 3.651288445552785e-06, "loss": 0.0158, "step": 395500 }, { "epoch": 32.74, "eval_loss": 0.192576065659523, "eval_runtime": 4673.505, "eval_samples_per_second": 16.175, "eval_steps_per_second": 1.348, "eval_wer": 0.05012607672911273, "step": 395500 }, { "epoch": 32.75, "learning_rate": 3.6471321695760602e-06, "loss": 0.0197, "step": 395600 }, { "epoch": 32.76, "learning_rate": 3.6429758935993353e-06, "loss": 0.0174, "step": 395700 }, { "epoch": 32.76, "learning_rate": 3.6388196176226104e-06, "loss": 0.0169, "step": 395800 }, { "epoch": 32.77, "learning_rate": 3.6346633416458855e-06, "loss": 0.0191, "step": 395900 }, { "epoch": 32.78, "learning_rate": 3.630507065669161e-06, "loss": 0.0195, "step": 396000 }, { "epoch": 32.78, "eval_loss": 0.19045275449752808, "eval_runtime": 4743.5937, "eval_samples_per_second": 15.936, "eval_steps_per_second": 1.328, "eval_wer": 0.04950607768295741, "step": 396000 }, { "epoch": 32.79, "learning_rate": 3.626350789692436e-06, "loss": 0.019, "step": 396100 }, { "epoch": 32.8, "learning_rate": 3.622194513715711e-06, "loss": 0.0169, "step": 396200 }, { "epoch": 32.81, "learning_rate": 3.6180382377389863e-06, "loss": 0.0176, "step": 396300 }, { "epoch": 32.81, "learning_rate": 3.613881961762261e-06, "loss": 0.0169, "step": 396400 }, { "epoch": 32.82, "learning_rate": 3.609725685785537e-06, "loss": 0.0196, "step": 396500 }, { "epoch": 32.82, "eval_loss": 0.18402999639511108, "eval_runtime": 4693.9881, "eval_samples_per_second": 16.105, "eval_steps_per_second": 1.342, "eval_wer": 0.048993770778814186, "step": 396500 }, { "epoch": 32.83, "learning_rate": 3.6055694098088116e-06, "loss": 0.0191, "step": 396600 }, { "epoch": 32.84, "learning_rate": 3.6014131338320867e-06, "loss": 0.0163, "step": 396700 }, { "epoch": 32.85, "learning_rate": 3.5972568578553618e-06, "loss": 0.019, "step": 396800 }, { "epoch": 32.86, "learning_rate": 3.593100581878637e-06, "loss": 0.0201, "step": 396900 }, { "epoch": 32.86, "learning_rate": 3.5889443059019124e-06, "loss": 0.0169, "step": 397000 }, { "epoch": 32.86, "eval_loss": 0.18464869260787964, "eval_runtime": 4786.0502, "eval_samples_per_second": 15.795, "eval_steps_per_second": 1.316, "eval_wer": 0.0488860786368021, "step": 397000 }, { "epoch": 32.87, "learning_rate": 3.5848295926849546e-06, "loss": 0.0234, "step": 397100 }, { "epoch": 32.88, "learning_rate": 3.5806733167082297e-06, "loss": 0.0219, "step": 397200 }, { "epoch": 32.89, "learning_rate": 3.576517040731505e-06, "loss": 0.0188, "step": 397300 }, { "epoch": 32.9, "learning_rate": 3.57236076475478e-06, "loss": 0.0152, "step": 397400 }, { "epoch": 32.91, "learning_rate": 3.5682044887780554e-06, "loss": 0.0187, "step": 397500 }, { "epoch": 32.91, "eval_loss": 0.18593889474868774, "eval_runtime": 4761.0358, "eval_samples_per_second": 15.878, "eval_steps_per_second": 1.323, "eval_wer": 0.050464537746865004, "step": 397500 }, { "epoch": 32.91, "learning_rate": 3.5640482128013305e-06, "loss": 0.021, "step": 397600 }, { "epoch": 32.92, "learning_rate": 3.5598919368246056e-06, "loss": 0.0162, "step": 397700 }, { "epoch": 32.93, "learning_rate": 3.5557356608478803e-06, "loss": 0.0141, "step": 397800 }, { "epoch": 32.94, "learning_rate": 3.5515793848711554e-06, "loss": 0.0183, "step": 397900 }, { "epoch": 32.95, "learning_rate": 3.547423108894431e-06, "loss": 0.0204, "step": 398000 }, { "epoch": 32.95, "eval_loss": 0.18958240747451782, "eval_runtime": 4691.4108, "eval_samples_per_second": 16.113, "eval_steps_per_second": 1.343, "eval_wer": 0.050762998826155654, "step": 398000 }, { "epoch": 32.96, "learning_rate": 3.543266832917706e-06, "loss": 0.018, "step": 398100 }, { "epoch": 32.96, "learning_rate": 3.539110556940981e-06, "loss": 0.0156, "step": 398200 }, { "epoch": 32.97, "learning_rate": 3.534954280964256e-06, "loss": 0.016, "step": 398300 }, { "epoch": 32.98, "learning_rate": 3.5307980049875313e-06, "loss": 0.0214, "step": 398400 }, { "epoch": 32.99, "learning_rate": 3.526683291770574e-06, "loss": 0.0189, "step": 398500 }, { "epoch": 32.99, "eval_loss": 0.1872534453868866, "eval_runtime": 4675.8167, "eval_samples_per_second": 16.167, "eval_steps_per_second": 1.347, "eval_wer": 0.05050915306284144, "step": 398500 }, { "epoch": 33.0, "learning_rate": 3.522527015793849e-06, "loss": 0.0145, "step": 398600 }, { "epoch": 33.0, "learning_rate": 3.518370739817124e-06, "loss": 0.0173, "step": 398700 }, { "epoch": 33.01, "learning_rate": 3.514214463840399e-06, "loss": 0.0161, "step": 398800 }, { "epoch": 33.02, "learning_rate": 3.5100581878636747e-06, "loss": 0.0167, "step": 398900 }, { "epoch": 33.03, "learning_rate": 3.50590191188695e-06, "loss": 0.0191, "step": 399000 }, { "epoch": 33.03, "eval_loss": 0.19028130173683167, "eval_runtime": 4700.6452, "eval_samples_per_second": 16.082, "eval_steps_per_second": 1.34, "eval_wer": 0.05023069195278161, "step": 399000 }, { "epoch": 33.04, "learning_rate": 3.501745635910225e-06, "loss": 0.0173, "step": 399100 }, { "epoch": 33.05, "learning_rate": 3.4975893599335e-06, "loss": 0.0173, "step": 399200 }, { "epoch": 33.05, "learning_rate": 3.4934330839567747e-06, "loss": 0.0151, "step": 399300 }, { "epoch": 33.06, "learning_rate": 3.4892768079800506e-06, "loss": 0.0176, "step": 399400 }, { "epoch": 33.07, "learning_rate": 3.4851205320033253e-06, "loss": 0.017, "step": 399500 }, { "epoch": 33.07, "eval_loss": 0.1891067773103714, "eval_runtime": 4749.424, "eval_samples_per_second": 15.917, "eval_steps_per_second": 1.326, "eval_wer": 0.049730692722011194, "step": 399500 }, { "epoch": 33.08, "learning_rate": 3.4809642560266004e-06, "loss": 0.0181, "step": 399600 }, { "epoch": 33.09, "learning_rate": 3.4768079800498755e-06, "loss": 0.0147, "step": 399700 }, { "epoch": 33.1, "learning_rate": 3.4726517040731506e-06, "loss": 0.0171, "step": 399800 }, { "epoch": 33.1, "learning_rate": 3.468495428096426e-06, "loss": 0.0163, "step": 399900 }, { "epoch": 33.11, "learning_rate": 3.464339152119701e-06, "loss": 0.0171, "step": 400000 }, { "epoch": 33.11, "eval_loss": 0.1897532343864441, "eval_runtime": 4648.6227, "eval_samples_per_second": 16.262, "eval_steps_per_second": 1.355, "eval_wer": 0.049523000733845024, "step": 400000 }, { "epoch": 33.12, "learning_rate": 3.4601828761429763e-06, "loss": 0.0166, "step": 400100 }, { "epoch": 33.13, "learning_rate": 3.4560266001662514e-06, "loss": 0.0174, "step": 400200 }, { "epoch": 33.14, "learning_rate": 3.4518703241895264e-06, "loss": 0.0191, "step": 400300 }, { "epoch": 33.15, "learning_rate": 3.447714048212802e-06, "loss": 0.0184, "step": 400400 }, { "epoch": 33.15, "learning_rate": 3.443557772236077e-06, "loss": 0.0146, "step": 400500 }, { "epoch": 33.15, "eval_loss": 0.18746218085289001, "eval_runtime": 4703.2049, "eval_samples_per_second": 16.073, "eval_steps_per_second": 1.34, "eval_wer": 0.05069376816343359, "step": 400500 }, { "epoch": 33.16, "learning_rate": 3.439401496259352e-06, "loss": 0.0152, "step": 400600 }, { "epoch": 33.17, "learning_rate": 3.435245220282627e-06, "loss": 0.018, "step": 400700 }, { "epoch": 33.18, "learning_rate": 3.431088944305902e-06, "loss": 0.0181, "step": 400800 }, { "epoch": 33.19, "learning_rate": 3.4269326683291774e-06, "loss": 0.0193, "step": 400900 }, { "epoch": 33.2, "learning_rate": 3.4227763923524525e-06, "loss": 0.014, "step": 401000 }, { "epoch": 33.2, "eval_loss": 0.18578752875328064, "eval_runtime": 4867.6907, "eval_samples_per_second": 15.53, "eval_steps_per_second": 1.294, "eval_wer": 0.04965530822260274, "step": 401000 }, { "epoch": 33.2, "learning_rate": 3.4186201163757276e-06, "loss": 0.0179, "step": 401100 }, { "epoch": 33.21, "learning_rate": 3.4144638403990027e-06, "loss": 0.0175, "step": 401200 }, { "epoch": 33.22, "learning_rate": 3.410349127182045e-06, "loss": 0.0159, "step": 401300 }, { "epoch": 33.23, "learning_rate": 3.4061928512053205e-06, "loss": 0.0196, "step": 401400 }, { "epoch": 33.24, "learning_rate": 3.4020365752285956e-06, "loss": 0.0176, "step": 401500 }, { "epoch": 33.24, "eval_loss": 0.1860053688287735, "eval_runtime": 4756.4375, "eval_samples_per_second": 15.893, "eval_steps_per_second": 1.325, "eval_wer": 0.04986607712911211, "step": 401500 }, { "epoch": 33.25, "learning_rate": 3.3978802992518707e-06, "loss": 0.0193, "step": 401600 }, { "epoch": 33.25, "learning_rate": 3.3937240232751457e-06, "loss": 0.0152, "step": 401700 }, { "epoch": 33.26, "learning_rate": 3.389567747298421e-06, "loss": 0.0196, "step": 401800 }, { "epoch": 33.27, "learning_rate": 3.3854114713216964e-06, "loss": 0.0144, "step": 401900 }, { "epoch": 33.28, "learning_rate": 3.3812551953449715e-06, "loss": 0.0212, "step": 402000 }, { "epoch": 33.28, "eval_loss": 0.18674153089523315, "eval_runtime": 4858.7366, "eval_samples_per_second": 15.559, "eval_steps_per_second": 1.297, "eval_wer": 0.04893223241195013, "step": 402000 }, { "epoch": 33.29, "learning_rate": 3.377098919368246e-06, "loss": 0.0187, "step": 402100 }, { "epoch": 33.29, "learning_rate": 3.3729426433915212e-06, "loss": 0.015, "step": 402200 }, { "epoch": 33.3, "learning_rate": 3.3687863674147963e-06, "loss": 0.0152, "step": 402300 }, { "epoch": 33.31, "learning_rate": 3.364630091438072e-06, "loss": 0.0181, "step": 402400 }, { "epoch": 33.32, "learning_rate": 3.360473815461347e-06, "loss": 0.0149, "step": 402500 }, { "epoch": 33.32, "eval_loss": 0.18385359644889832, "eval_runtime": 4637.1504, "eval_samples_per_second": 16.302, "eval_steps_per_second": 1.359, "eval_wer": 0.04873377117881357, "step": 402500 }, { "epoch": 33.33, "learning_rate": 3.356317539484622e-06, "loss": 0.015, "step": 402600 }, { "epoch": 33.34, "learning_rate": 3.352161263507897e-06, "loss": 0.0138, "step": 402700 }, { "epoch": 33.34, "learning_rate": 3.348004987531172e-06, "loss": 0.0146, "step": 402800 }, { "epoch": 33.35, "learning_rate": 3.3438487115544477e-06, "loss": 0.0213, "step": 402900 }, { "epoch": 33.36, "learning_rate": 3.339692435577723e-06, "loss": 0.0169, "step": 403000 }, { "epoch": 33.36, "eval_loss": 0.18301208317279816, "eval_runtime": 4831.3452, "eval_samples_per_second": 15.647, "eval_steps_per_second": 1.304, "eval_wer": 0.04872915580129877, "step": 403000 }, { "epoch": 33.37, "learning_rate": 3.335536159600998e-06, "loss": 0.015, "step": 403100 }, { "epoch": 33.38, "learning_rate": 3.331379883624273e-06, "loss": 0.0152, "step": 403200 }, { "epoch": 33.39, "learning_rate": 3.3272236076475477e-06, "loss": 0.0156, "step": 403300 }, { "epoch": 33.39, "learning_rate": 3.3230673316708236e-06, "loss": 0.0152, "step": 403400 }, { "epoch": 33.4, "learning_rate": 3.3189110556940983e-06, "loss": 0.0189, "step": 403500 }, { "epoch": 33.4, "eval_loss": 0.18443149328231812, "eval_runtime": 4964.5004, "eval_samples_per_second": 15.227, "eval_steps_per_second": 1.269, "eval_wer": 0.04851530997644619, "step": 403500 }, { "epoch": 33.41, "learning_rate": 3.3147547797173734e-06, "loss": 0.0175, "step": 403600 }, { "epoch": 33.42, "learning_rate": 3.3105985037406484e-06, "loss": 0.0187, "step": 403700 }, { "epoch": 33.43, "learning_rate": 3.3064422277639235e-06, "loss": 0.0134, "step": 403800 }, { "epoch": 33.44, "learning_rate": 3.302285951787199e-06, "loss": 0.0238, "step": 403900 }, { "epoch": 33.44, "learning_rate": 3.298129675810474e-06, "loss": 0.0194, "step": 404000 }, { "epoch": 33.44, "eval_loss": 0.1864852011203766, "eval_runtime": 4804.1081, "eval_samples_per_second": 15.735, "eval_steps_per_second": 1.311, "eval_wer": 0.04902453996224621, "step": 404000 }, { "epoch": 33.45, "learning_rate": 3.2940149625935164e-06, "loss": 0.0171, "step": 404100 }, { "epoch": 33.46, "learning_rate": 3.2898586866167915e-06, "loss": 0.0162, "step": 404200 }, { "epoch": 33.47, "learning_rate": 3.285702410640067e-06, "loss": 0.0167, "step": 404300 }, { "epoch": 33.48, "learning_rate": 3.281546134663342e-06, "loss": 0.0169, "step": 404400 }, { "epoch": 33.49, "learning_rate": 3.277389858686617e-06, "loss": 0.0184, "step": 404500 }, { "epoch": 33.49, "eval_loss": 0.18476155400276184, "eval_runtime": 4709.0298, "eval_samples_per_second": 16.053, "eval_steps_per_second": 1.338, "eval_wer": 0.04950300076461421, "step": 404500 }, { "epoch": 33.49, "learning_rate": 3.2732335827098923e-06, "loss": 0.0144, "step": 404600 }, { "epoch": 33.5, "learning_rate": 3.269077306733167e-06, "loss": 0.0191, "step": 404700 }, { "epoch": 33.51, "learning_rate": 3.264921030756443e-06, "loss": 0.0166, "step": 404800 }, { "epoch": 33.52, "learning_rate": 3.2607647547797176e-06, "loss": 0.0163, "step": 404900 }, { "epoch": 33.53, "learning_rate": 3.2566084788029927e-06, "loss": 0.0185, "step": 405000 }, { "epoch": 33.53, "eval_loss": 0.18381834030151367, "eval_runtime": 4780.9275, "eval_samples_per_second": 15.812, "eval_steps_per_second": 1.318, "eval_wer": 0.04944300085692176, "step": 405000 }, { "epoch": 33.53, "learning_rate": 3.2524522028262678e-06, "loss": 0.0155, "step": 405100 }, { "epoch": 33.54, "learning_rate": 3.248295926849543e-06, "loss": 0.0205, "step": 405200 }, { "epoch": 33.55, "learning_rate": 3.2441396508728184e-06, "loss": 0.0158, "step": 405300 }, { "epoch": 33.56, "learning_rate": 3.2399833748960935e-06, "loss": 0.0178, "step": 405400 }, { "epoch": 33.57, "learning_rate": 3.2358270989193685e-06, "loss": 0.0184, "step": 405500 }, { "epoch": 33.57, "eval_loss": 0.1834350973367691, "eval_runtime": 4868.1041, "eval_samples_per_second": 15.529, "eval_steps_per_second": 1.294, "eval_wer": 0.0488891555551453, "step": 405500 }, { "epoch": 33.58, "learning_rate": 3.2316708229426436e-06, "loss": 0.0186, "step": 405600 }, { "epoch": 33.58, "learning_rate": 3.2275145469659187e-06, "loss": 0.0167, "step": 405700 }, { "epoch": 33.59, "learning_rate": 3.2233582709891942e-06, "loss": 0.0175, "step": 405800 }, { "epoch": 33.6, "learning_rate": 3.2192019950124693e-06, "loss": 0.0167, "step": 405900 }, { "epoch": 33.61, "learning_rate": 3.2150457190357444e-06, "loss": 0.019, "step": 406000 }, { "epoch": 33.61, "eval_loss": 0.17691895365715027, "eval_runtime": 4873.8188, "eval_samples_per_second": 15.51, "eval_steps_per_second": 1.293, "eval_wer": 0.04817838741786551, "step": 406000 }, { "epoch": 33.62, "learning_rate": 3.210889443059019e-06, "loss": 0.0172, "step": 406100 }, { "epoch": 33.63, "learning_rate": 3.2067747298420614e-06, "loss": 0.0184, "step": 406200 }, { "epoch": 33.63, "learning_rate": 3.2026184538653373e-06, "loss": 0.017, "step": 406300 }, { "epoch": 33.64, "learning_rate": 3.198462177888612e-06, "loss": 0.0193, "step": 406400 }, { "epoch": 33.65, "learning_rate": 3.194305901911887e-06, "loss": 0.0174, "step": 406500 }, { "epoch": 33.65, "eval_loss": 0.1825142204761505, "eval_runtime": 4793.4079, "eval_samples_per_second": 15.771, "eval_steps_per_second": 1.314, "eval_wer": 0.048181464336208714, "step": 406500 }, { "epoch": 33.66, "learning_rate": 3.190149625935162e-06, "loss": 0.0165, "step": 406600 }, { "epoch": 33.67, "learning_rate": 3.1859933499584372e-06, "loss": 0.0171, "step": 406700 }, { "epoch": 33.68, "learning_rate": 3.1818370739817128e-06, "loss": 0.0162, "step": 406800 }, { "epoch": 33.68, "learning_rate": 3.177680798004988e-06, "loss": 0.0176, "step": 406900 }, { "epoch": 33.69, "learning_rate": 3.173524522028263e-06, "loss": 0.0215, "step": 407000 }, { "epoch": 33.69, "eval_loss": 0.1819218546152115, "eval_runtime": 4975.1453, "eval_samples_per_second": 15.195, "eval_steps_per_second": 1.266, "eval_wer": 0.048452233150410536, "step": 407000 }, { "epoch": 33.7, "learning_rate": 3.169368246051538e-06, "loss": 0.0179, "step": 407100 }, { "epoch": 33.71, "learning_rate": 3.165211970074813e-06, "loss": 0.021, "step": 407200 }, { "epoch": 33.72, "learning_rate": 3.1610556940980886e-06, "loss": 0.018, "step": 407300 }, { "epoch": 33.73, "learning_rate": 3.1568994181213637e-06, "loss": 0.0189, "step": 407400 }, { "epoch": 33.73, "learning_rate": 3.1527431421446384e-06, "loss": 0.0166, "step": 407500 }, { "epoch": 33.73, "eval_loss": 0.1855279952287674, "eval_runtime": 4885.325, "eval_samples_per_second": 15.474, "eval_steps_per_second": 1.29, "eval_wer": 0.049050693768163435, "step": 407500 }, { "epoch": 33.74, "learning_rate": 3.1485868661679135e-06, "loss": 0.0145, "step": 407600 }, { "epoch": 33.75, "learning_rate": 3.1444305901911886e-06, "loss": 0.0179, "step": 407700 }, { "epoch": 33.76, "learning_rate": 3.140274314214464e-06, "loss": 0.0211, "step": 407800 }, { "epoch": 33.77, "learning_rate": 3.136118038237739e-06, "loss": 0.0151, "step": 407900 }, { "epoch": 33.77, "learning_rate": 3.1319617622610143e-06, "loss": 0.0134, "step": 408000 }, { "epoch": 33.77, "eval_loss": 0.18771491944789886, "eval_runtime": 4810.4282, "eval_samples_per_second": 15.715, "eval_steps_per_second": 1.31, "eval_wer": 0.0481630028261495, "step": 408000 }, { "epoch": 33.78, "learning_rate": 3.1278054862842894e-06, "loss": 0.0166, "step": 408100 }, { "epoch": 33.79, "learning_rate": 3.123690773067332e-06, "loss": 0.0189, "step": 408200 }, { "epoch": 33.8, "learning_rate": 3.119534497090607e-06, "loss": 0.0141, "step": 408300 }, { "epoch": 33.81, "learning_rate": 3.1153782211138822e-06, "loss": 0.0167, "step": 408400 }, { "epoch": 33.82, "learning_rate": 3.1112219451371573e-06, "loss": 0.0212, "step": 408500 }, { "epoch": 33.82, "eval_loss": 0.1877739280462265, "eval_runtime": 5024.1464, "eval_samples_per_second": 15.046, "eval_steps_per_second": 1.254, "eval_wer": 0.04952607765218823, "step": 408500 }, { "epoch": 33.82, "learning_rate": 3.1070656691604324e-06, "loss": 0.017, "step": 408600 }, { "epoch": 33.83, "learning_rate": 3.102909393183708e-06, "loss": 0.0162, "step": 408700 }, { "epoch": 33.84, "learning_rate": 3.098753117206983e-06, "loss": 0.0187, "step": 408800 }, { "epoch": 33.85, "learning_rate": 3.094596841230258e-06, "loss": 0.0151, "step": 408900 }, { "epoch": 33.86, "learning_rate": 3.090440565253533e-06, "loss": 0.0176, "step": 409000 }, { "epoch": 33.86, "eval_loss": 0.18734630942344666, "eval_runtime": 4693.8074, "eval_samples_per_second": 16.105, "eval_steps_per_second": 1.342, "eval_wer": 0.04906761681905105, "step": 409000 }, { "epoch": 33.87, "learning_rate": 3.086284289276808e-06, "loss": 0.0145, "step": 409100 }, { "epoch": 33.87, "learning_rate": 3.0821280133000834e-06, "loss": 0.0162, "step": 409200 }, { "epoch": 33.88, "learning_rate": 3.0779717373233585e-06, "loss": 0.0145, "step": 409300 }, { "epoch": 33.89, "learning_rate": 3.0738154613466336e-06, "loss": 0.0195, "step": 409400 }, { "epoch": 33.9, "learning_rate": 3.0696591853699087e-06, "loss": 0.0156, "step": 409500 }, { "epoch": 33.9, "eval_loss": 0.18688324093818665, "eval_runtime": 4608.5602, "eval_samples_per_second": 16.403, "eval_steps_per_second": 1.367, "eval_wer": 0.048267618049818384, "step": 409500 }, { "epoch": 33.91, "learning_rate": 3.0655029093931838e-06, "loss": 0.0157, "step": 409600 }, { "epoch": 33.92, "learning_rate": 3.0613466334164593e-06, "loss": 0.0199, "step": 409700 }, { "epoch": 33.92, "learning_rate": 3.0571903574397344e-06, "loss": 0.0152, "step": 409800 }, { "epoch": 33.93, "learning_rate": 3.0530340814630095e-06, "loss": 0.0167, "step": 409900 }, { "epoch": 33.94, "learning_rate": 3.0488778054862846e-06, "loss": 0.013, "step": 410000 }, { "epoch": 33.94, "eval_loss": 0.18626342713832855, "eval_runtime": 4644.0816, "eval_samples_per_second": 16.278, "eval_steps_per_second": 1.357, "eval_wer": 0.0490076169113586, "step": 410000 }, { "epoch": 33.95, "learning_rate": 3.0447215295095592e-06, "loss": 0.0163, "step": 410100 }, { "epoch": 33.96, "learning_rate": 3.040565253532835e-06, "loss": 0.0199, "step": 410200 }, { "epoch": 33.97, "learning_rate": 3.0364089775561103e-06, "loss": 0.0171, "step": 410300 }, { "epoch": 33.97, "learning_rate": 3.032252701579385e-06, "loss": 0.0154, "step": 410400 }, { "epoch": 33.98, "learning_rate": 3.02809642560266e-06, "loss": 0.0182, "step": 410500 }, { "epoch": 33.98, "eval_loss": 0.18997299671173096, "eval_runtime": 4779.1577, "eval_samples_per_second": 15.818, "eval_steps_per_second": 1.318, "eval_wer": 0.049850692537396095, "step": 410500 }, { "epoch": 33.99, "learning_rate": 3.023940149625935e-06, "loss": 0.0164, "step": 410600 }, { "epoch": 34.0, "learning_rate": 3.0197838736492107e-06, "loss": 0.0159, "step": 410700 }, { "epoch": 34.01, "learning_rate": 3.0156275976724857e-06, "loss": 0.0168, "step": 410800 }, { "epoch": 34.01, "learning_rate": 3.011471321695761e-06, "loss": 0.0167, "step": 410900 }, { "epoch": 34.02, "learning_rate": 3.007315045719036e-06, "loss": 0.0173, "step": 411000 }, { "epoch": 34.02, "eval_loss": 0.18745605647563934, "eval_runtime": 4747.1359, "eval_samples_per_second": 15.924, "eval_steps_per_second": 1.327, "eval_wer": 0.04883992486165406, "step": 411000 }, { "epoch": 34.03, "learning_rate": 3.003158769742311e-06, "loss": 0.0161, "step": 411100 }, { "epoch": 34.04, "learning_rate": 2.9990024937655865e-06, "loss": 0.0202, "step": 411200 }, { "epoch": 34.05, "learning_rate": 2.9948462177888616e-06, "loss": 0.0216, "step": 411300 }, { "epoch": 34.06, "learning_rate": 2.9906899418121367e-06, "loss": 0.0161, "step": 411400 }, { "epoch": 34.06, "learning_rate": 2.9865336658354114e-06, "loss": 0.0152, "step": 411500 }, { "epoch": 34.06, "eval_loss": 0.1894349902868271, "eval_runtime": 4866.7682, "eval_samples_per_second": 15.533, "eval_steps_per_second": 1.294, "eval_wer": 0.04872454042378396, "step": 411500 }, { "epoch": 34.07, "learning_rate": 2.9823773898586865e-06, "loss": 0.0179, "step": 411600 }, { "epoch": 34.08, "learning_rate": 2.9782211138819624e-06, "loss": 0.0156, "step": 411700 }, { "epoch": 34.09, "learning_rate": 2.974064837905237e-06, "loss": 0.0166, "step": 411800 }, { "epoch": 34.1, "learning_rate": 2.969908561928512e-06, "loss": 0.0182, "step": 411900 }, { "epoch": 34.11, "learning_rate": 2.9657522859517873e-06, "loss": 0.0158, "step": 412000 }, { "epoch": 34.11, "eval_loss": 0.18681329488754272, "eval_runtime": 4713.8686, "eval_samples_per_second": 16.037, "eval_steps_per_second": 1.336, "eval_wer": 0.04857684834331024, "step": 412000 }, { "epoch": 34.11, "learning_rate": 2.9615960099750624e-06, "loss": 0.0149, "step": 412100 }, { "epoch": 34.12, "learning_rate": 2.957439733998338e-06, "loss": 0.0185, "step": 412200 }, { "epoch": 34.13, "learning_rate": 2.953283458021613e-06, "loss": 0.018, "step": 412300 }, { "epoch": 34.14, "learning_rate": 2.949127182044888e-06, "loss": 0.0173, "step": 412400 }, { "epoch": 34.15, "learning_rate": 2.944970906068163e-06, "loss": 0.0144, "step": 412500 }, { "epoch": 34.15, "eval_loss": 0.19083596765995026, "eval_runtime": 4733.1488, "eval_samples_per_second": 15.971, "eval_steps_per_second": 1.331, "eval_wer": 0.04819838738709633, "step": 412500 }, { "epoch": 34.16, "learning_rate": 2.9408146300914383e-06, "loss": 0.019, "step": 412600 }, { "epoch": 34.16, "learning_rate": 2.9366583541147138e-06, "loss": 0.0158, "step": 412700 }, { "epoch": 34.17, "learning_rate": 2.932502078137989e-06, "loss": 0.0183, "step": 412800 }, { "epoch": 34.18, "learning_rate": 2.9283458021612635e-06, "loss": 0.0184, "step": 412900 }, { "epoch": 34.19, "learning_rate": 2.9241895261845386e-06, "loss": 0.0198, "step": 413000 }, { "epoch": 34.19, "eval_loss": 0.18740113079547882, "eval_runtime": 4793.0142, "eval_samples_per_second": 15.772, "eval_steps_per_second": 1.314, "eval_wer": 0.048758386525559194, "step": 413000 }, { "epoch": 34.2, "learning_rate": 2.9200332502078137e-06, "loss": 0.0182, "step": 413100 }, { "epoch": 34.21, "learning_rate": 2.9158769742310892e-06, "loss": 0.0168, "step": 413200 }, { "epoch": 34.21, "learning_rate": 2.9117206982543643e-06, "loss": 0.0188, "step": 413300 }, { "epoch": 34.22, "learning_rate": 2.9075644222776394e-06, "loss": 0.0178, "step": 413400 }, { "epoch": 34.23, "learning_rate": 2.9034081463009145e-06, "loss": 0.0146, "step": 413500 }, { "epoch": 34.23, "eval_loss": 0.19411829113960266, "eval_runtime": 4757.2901, "eval_samples_per_second": 15.89, "eval_steps_per_second": 1.324, "eval_wer": 0.04886607866757128, "step": 413500 }, { "epoch": 34.24, "learning_rate": 2.8992518703241896e-06, "loss": 0.0193, "step": 413600 }, { "epoch": 34.25, "learning_rate": 2.8951371571072323e-06, "loss": 0.0167, "step": 413700 }, { "epoch": 34.25, "learning_rate": 2.8909808811305074e-06, "loss": 0.023, "step": 413800 }, { "epoch": 34.26, "learning_rate": 2.8868246051537825e-06, "loss": 0.0239, "step": 413900 }, { "epoch": 34.27, "learning_rate": 2.8826683291770576e-06, "loss": 0.0186, "step": 414000 }, { "epoch": 34.27, "eval_loss": 0.18190579116344452, "eval_runtime": 4761.0269, "eval_samples_per_second": 15.878, "eval_steps_per_second": 1.323, "eval_wer": 0.04910607829834108, "step": 414000 }, { "epoch": 34.28, "learning_rate": 2.878512053200333e-06, "loss": 0.0195, "step": 414100 }, { "epoch": 34.29, "learning_rate": 2.874355777223608e-06, "loss": 0.0174, "step": 414200 }, { "epoch": 34.3, "learning_rate": 2.8701995012468833e-06, "loss": 0.0164, "step": 414300 }, { "epoch": 34.3, "learning_rate": 2.866043225270158e-06, "loss": 0.0183, "step": 414400 }, { "epoch": 34.31, "learning_rate": 2.861886949293433e-06, "loss": 0.0168, "step": 414500 }, { "epoch": 34.31, "eval_loss": 0.1872846931219101, "eval_runtime": 4673.0737, "eval_samples_per_second": 16.177, "eval_steps_per_second": 1.348, "eval_wer": 0.049487616172898194, "step": 414500 }, { "epoch": 34.32, "learning_rate": 2.8577306733167085e-06, "loss": 0.0195, "step": 414600 }, { "epoch": 34.33, "learning_rate": 2.8535743973399836e-06, "loss": 0.0161, "step": 414700 }, { "epoch": 34.34, "learning_rate": 2.8494181213632587e-06, "loss": 0.0164, "step": 414800 }, { "epoch": 34.35, "learning_rate": 2.845261845386534e-06, "loss": 0.0153, "step": 414900 }, { "epoch": 34.35, "learning_rate": 2.841105569409809e-06, "loss": 0.0152, "step": 415000 }, { "epoch": 34.35, "eval_loss": 0.19325360655784607, "eval_runtime": 4587.675, "eval_samples_per_second": 16.478, "eval_steps_per_second": 1.373, "eval_wer": 0.04956607759064986, "step": 415000 }, { "epoch": 34.36, "learning_rate": 2.8369492934330844e-06, "loss": 0.0164, "step": 415100 }, { "epoch": 34.37, "learning_rate": 2.8327930174563595e-06, "loss": 0.0164, "step": 415200 }, { "epoch": 34.38, "learning_rate": 2.8286367414796346e-06, "loss": 0.022, "step": 415300 }, { "epoch": 34.39, "learning_rate": 2.8244804655029097e-06, "loss": 0.0167, "step": 415400 }, { "epoch": 34.4, "learning_rate": 2.8203241895261844e-06, "loss": 0.016, "step": 415500 }, { "epoch": 34.4, "eval_loss": 0.18899288773536682, "eval_runtime": 4630.4591, "eval_samples_per_second": 16.326, "eval_steps_per_second": 1.361, "eval_wer": 0.04873377117881357, "step": 415500 }, { "epoch": 34.4, "learning_rate": 2.8161679135494603e-06, "loss": 0.0189, "step": 415600 }, { "epoch": 34.41, "learning_rate": 2.8120116375727354e-06, "loss": 0.0149, "step": 415700 }, { "epoch": 34.42, "learning_rate": 2.80785536159601e-06, "loss": 0.0177, "step": 415800 }, { "epoch": 34.43, "learning_rate": 2.803699085619285e-06, "loss": 0.0148, "step": 415900 }, { "epoch": 34.44, "learning_rate": 2.7995428096425603e-06, "loss": 0.0185, "step": 416000 }, { "epoch": 34.44, "eval_loss": 0.18481364846229553, "eval_runtime": 4828.8287, "eval_samples_per_second": 15.655, "eval_steps_per_second": 1.305, "eval_wer": 0.048501463843901776, "step": 416000 }, { "epoch": 34.45, "learning_rate": 2.7953865336658358e-06, "loss": 0.0159, "step": 416100 }, { "epoch": 34.45, "learning_rate": 2.791230257689111e-06, "loss": 0.0236, "step": 416200 }, { "epoch": 34.46, "learning_rate": 2.787073981712386e-06, "loss": 0.0192, "step": 416300 }, { "epoch": 34.47, "learning_rate": 2.782917705735661e-06, "loss": 0.0155, "step": 416400 }, { "epoch": 34.48, "learning_rate": 2.778761429758936e-06, "loss": 0.0159, "step": 416500 }, { "epoch": 34.48, "eval_loss": 0.18218894302845, "eval_runtime": 4692.0835, "eval_samples_per_second": 16.111, "eval_steps_per_second": 1.343, "eval_wer": 0.047846080236799635, "step": 416500 }, { "epoch": 34.49, "learning_rate": 2.7746051537822117e-06, "loss": 0.0166, "step": 416600 }, { "epoch": 34.5, "learning_rate": 2.7704488778054868e-06, "loss": 0.0153, "step": 416700 }, { "epoch": 34.5, "learning_rate": 2.766292601828762e-06, "loss": 0.012, "step": 416800 }, { "epoch": 34.51, "learning_rate": 2.7621363258520365e-06, "loss": 0.0196, "step": 416900 }, { "epoch": 34.52, "learning_rate": 2.7579800498753116e-06, "loss": 0.0166, "step": 417000 }, { "epoch": 34.52, "eval_loss": 0.1857985258102417, "eval_runtime": 4694.0756, "eval_samples_per_second": 16.104, "eval_steps_per_second": 1.342, "eval_wer": 0.04842915626283652, "step": 417000 }, { "epoch": 34.53, "learning_rate": 2.7538237738985876e-06, "loss": 0.0243, "step": 417100 }, { "epoch": 34.54, "learning_rate": 2.7496674979218622e-06, "loss": 0.019, "step": 417200 }, { "epoch": 34.54, "learning_rate": 2.7455527847049045e-06, "loss": 0.017, "step": 417300 }, { "epoch": 34.55, "learning_rate": 2.7413965087281796e-06, "loss": 0.0138, "step": 417400 }, { "epoch": 34.56, "learning_rate": 2.7372402327514547e-06, "loss": 0.0173, "step": 417500 }, { "epoch": 34.56, "eval_loss": 0.1884331852197647, "eval_runtime": 4711.4269, "eval_samples_per_second": 16.045, "eval_steps_per_second": 1.337, "eval_wer": 0.048786078790648016, "step": 417500 }, { "epoch": 34.57, "learning_rate": 2.73308395677473e-06, "loss": 0.0157, "step": 417600 }, { "epoch": 34.58, "learning_rate": 2.7289276807980053e-06, "loss": 0.0162, "step": 417700 }, { "epoch": 34.59, "learning_rate": 2.7247714048212804e-06, "loss": 0.015, "step": 417800 }, { "epoch": 34.59, "learning_rate": 2.7206151288445555e-06, "loss": 0.0167, "step": 417900 }, { "epoch": 34.6, "learning_rate": 2.7164588528678305e-06, "loss": 0.0178, "step": 418000 }, { "epoch": 34.6, "eval_loss": 0.18625645339488983, "eval_runtime": 4604.4832, "eval_samples_per_second": 16.418, "eval_steps_per_second": 1.368, "eval_wer": 0.04777838803324918, "step": 418000 }, { "epoch": 34.61, "learning_rate": 2.712302576891106e-06, "loss": 0.0163, "step": 418100 }, { "epoch": 34.62, "learning_rate": 2.708146300914381e-06, "loss": 0.0184, "step": 418200 }, { "epoch": 34.63, "learning_rate": 2.7039900249376563e-06, "loss": 0.0181, "step": 418300 }, { "epoch": 34.64, "learning_rate": 2.699833748960931e-06, "loss": 0.0161, "step": 418400 }, { "epoch": 34.64, "learning_rate": 2.695677472984206e-06, "loss": 0.0156, "step": 418500 }, { "epoch": 34.64, "eval_loss": 0.19056209921836853, "eval_runtime": 4716.9129, "eval_samples_per_second": 16.026, "eval_steps_per_second": 1.336, "eval_wer": 0.048161464366977894, "step": 418500 }, { "epoch": 34.65, "learning_rate": 2.6915211970074815e-06, "loss": 0.0158, "step": 418600 }, { "epoch": 34.66, "learning_rate": 2.6873649210307566e-06, "loss": 0.0137, "step": 418700 }, { "epoch": 34.67, "learning_rate": 2.6832086450540317e-06, "loss": 0.0172, "step": 418800 }, { "epoch": 34.68, "learning_rate": 2.679052369077307e-06, "loss": 0.0185, "step": 418900 }, { "epoch": 34.69, "learning_rate": 2.674896093100582e-06, "loss": 0.0184, "step": 419000 }, { "epoch": 34.69, "eval_loss": 0.18717414140701294, "eval_runtime": 4628.8344, "eval_samples_per_second": 16.331, "eval_steps_per_second": 1.361, "eval_wer": 0.048543002241535016, "step": 419000 }, { "epoch": 34.69, "learning_rate": 2.6707398171238574e-06, "loss": 0.0188, "step": 419100 }, { "epoch": 34.7, "learning_rate": 2.6665835411471325e-06, "loss": 0.0161, "step": 419200 }, { "epoch": 34.71, "learning_rate": 2.6624272651704076e-06, "loss": 0.017, "step": 419300 }, { "epoch": 34.72, "learning_rate": 2.6582709891936827e-06, "loss": 0.0173, "step": 419400 }, { "epoch": 34.73, "learning_rate": 2.6541147132169574e-06, "loss": 0.015, "step": 419500 }, { "epoch": 34.73, "eval_loss": 0.18294526636600494, "eval_runtime": 4637.2364, "eval_samples_per_second": 16.302, "eval_steps_per_second": 1.359, "eval_wer": 0.04801377228650418, "step": 419500 }, { "epoch": 34.74, "learning_rate": 2.6499584372402333e-06, "loss": 0.0144, "step": 419600 }, { "epoch": 34.74, "learning_rate": 2.6458021612635084e-06, "loss": 0.0205, "step": 419700 }, { "epoch": 34.75, "learning_rate": 2.6416874480465502e-06, "loss": 0.0258, "step": 419800 }, { "epoch": 34.76, "learning_rate": 2.6375311720698253e-06, "loss": 0.0144, "step": 419900 }, { "epoch": 34.77, "learning_rate": 2.633374896093101e-06, "loss": 0.018, "step": 420000 }, { "epoch": 34.77, "eval_loss": 0.18076634407043457, "eval_runtime": 4656.5948, "eval_samples_per_second": 16.234, "eval_steps_per_second": 1.353, "eval_wer": 0.04788608017526127, "step": 420000 }, { "epoch": 34.78, "learning_rate": 2.629218620116376e-06, "loss": 0.0173, "step": 420100 }, { "epoch": 34.78, "learning_rate": 2.625062344139651e-06, "loss": 0.0157, "step": 420200 }, { "epoch": 34.79, "learning_rate": 2.620906068162926e-06, "loss": 0.0182, "step": 420300 }, { "epoch": 34.8, "learning_rate": 2.616749792186201e-06, "loss": 0.0161, "step": 420400 }, { "epoch": 34.81, "learning_rate": 2.6125935162094767e-06, "loss": 0.0177, "step": 420500 }, { "epoch": 34.81, "eval_loss": 0.17866896092891693, "eval_runtime": 4680.7992, "eval_samples_per_second": 16.15, "eval_steps_per_second": 1.346, "eval_wer": 0.04807684911253983, "step": 420500 }, { "epoch": 34.82, "learning_rate": 2.608437240232752e-06, "loss": 0.0157, "step": 420600 }, { "epoch": 34.83, "learning_rate": 2.604280964256027e-06, "loss": 0.0124, "step": 420700 }, { "epoch": 34.83, "learning_rate": 2.600124688279302e-06, "loss": 0.0174, "step": 420800 }, { "epoch": 34.84, "learning_rate": 2.595968412302577e-06, "loss": 0.0168, "step": 420900 }, { "epoch": 34.85, "learning_rate": 2.5918121363258526e-06, "loss": 0.0163, "step": 421000 }, { "epoch": 34.85, "eval_loss": 0.184172585606575, "eval_runtime": 4673.3913, "eval_samples_per_second": 16.176, "eval_steps_per_second": 1.348, "eval_wer": 0.04901838612555981, "step": 421000 }, { "epoch": 34.86, "learning_rate": 2.5876558603491277e-06, "loss": 0.0172, "step": 421100 }, { "epoch": 34.87, "learning_rate": 2.5834995843724024e-06, "loss": 0.0161, "step": 421200 }, { "epoch": 34.88, "learning_rate": 2.5793433083956775e-06, "loss": 0.0152, "step": 421300 }, { "epoch": 34.88, "learning_rate": 2.5751870324189526e-06, "loss": 0.0155, "step": 421400 }, { "epoch": 34.89, "learning_rate": 2.571030756442228e-06, "loss": 0.0143, "step": 421500 }, { "epoch": 34.89, "eval_loss": 0.18483568727970123, "eval_runtime": 4735.7666, "eval_samples_per_second": 15.963, "eval_steps_per_second": 1.33, "eval_wer": 0.048773771117275204, "step": 421500 }, { "epoch": 34.9, "learning_rate": 2.566874480465503e-06, "loss": 0.0139, "step": 421600 }, { "epoch": 34.91, "learning_rate": 2.5627182044887783e-06, "loss": 0.0183, "step": 421700 }, { "epoch": 34.92, "learning_rate": 2.5585619285120533e-06, "loss": 0.0161, "step": 421800 }, { "epoch": 34.93, "learning_rate": 2.5544056525353284e-06, "loss": 0.0159, "step": 421900 }, { "epoch": 34.93, "learning_rate": 2.550249376558604e-06, "loss": 0.0136, "step": 422000 }, { "epoch": 34.93, "eval_loss": 0.18828418850898743, "eval_runtime": 4938.1352, "eval_samples_per_second": 15.308, "eval_steps_per_second": 1.276, "eval_wer": 0.048944540085322945, "step": 422000 }, { "epoch": 34.94, "learning_rate": 2.546093100581879e-06, "loss": 0.0159, "step": 422100 }, { "epoch": 34.95, "learning_rate": 2.541936824605154e-06, "loss": 0.013, "step": 422200 }, { "epoch": 34.96, "learning_rate": 2.5377805486284292e-06, "loss": 0.0153, "step": 422300 }, { "epoch": 34.97, "learning_rate": 2.533624272651704e-06, "loss": 0.0175, "step": 422400 }, { "epoch": 34.98, "learning_rate": 2.52946799667498e-06, "loss": 0.0183, "step": 422500 }, { "epoch": 34.98, "eval_loss": 0.18762393295764923, "eval_runtime": 5002.4829, "eval_samples_per_second": 15.111, "eval_steps_per_second": 1.259, "eval_wer": 0.04856761758828063, "step": 422500 }, { "epoch": 34.98, "learning_rate": 2.5253117206982545e-06, "loss": 0.0171, "step": 422600 }, { "epoch": 34.99, "learning_rate": 2.5211554447215296e-06, "loss": 0.0172, "step": 422700 }, { "epoch": 35.0, "learning_rate": 2.5169991687448047e-06, "loss": 0.014, "step": 422800 }, { "epoch": 35.01, "learning_rate": 2.51284289276808e-06, "loss": 0.0129, "step": 422900 }, { "epoch": 35.02, "learning_rate": 2.5086866167913553e-06, "loss": 0.017, "step": 423000 }, { "epoch": 35.02, "eval_loss": 0.1899639517068863, "eval_runtime": 4772.8029, "eval_samples_per_second": 15.839, "eval_steps_per_second": 1.32, "eval_wer": 0.04849684846638697, "step": 423000 }, { "epoch": 35.02, "learning_rate": 2.5045303408146304e-06, "loss": 0.0155, "step": 423100 }, { "epoch": 35.03, "learning_rate": 2.5003740648379055e-06, "loss": 0.0144, "step": 423200 }, { "epoch": 35.04, "learning_rate": 2.4962177888611806e-06, "loss": 0.0174, "step": 423300 }, { "epoch": 35.05, "learning_rate": 2.4920615128844557e-06, "loss": 0.014, "step": 423400 }, { "epoch": 35.06, "learning_rate": 2.4879052369077308e-06, "loss": 0.016, "step": 423500 }, { "epoch": 35.06, "eval_loss": 0.18816962838172913, "eval_runtime": 4651.6293, "eval_samples_per_second": 16.251, "eval_steps_per_second": 1.354, "eval_wer": 0.04902607842141781, "step": 423500 }, { "epoch": 35.07, "learning_rate": 2.4837489609310063e-06, "loss": 0.0141, "step": 423600 }, { "epoch": 35.07, "learning_rate": 2.4795926849542814e-06, "loss": 0.0172, "step": 423700 }, { "epoch": 35.08, "learning_rate": 2.475436408977556e-06, "loss": 0.0243, "step": 423800 }, { "epoch": 35.09, "learning_rate": 2.4712801330008316e-06, "loss": 0.0164, "step": 423900 }, { "epoch": 35.1, "learning_rate": 2.4671238570241067e-06, "loss": 0.0155, "step": 424000 }, { "epoch": 35.1, "eval_loss": 0.186177596449852, "eval_runtime": 4781.7558, "eval_samples_per_second": 15.809, "eval_steps_per_second": 1.318, "eval_wer": 0.04848146387467096, "step": 424000 }, { "epoch": 35.11, "learning_rate": 2.4629675810473818e-06, "loss": 0.0166, "step": 424100 }, { "epoch": 35.12, "learning_rate": 2.458811305070657e-06, "loss": 0.0156, "step": 424200 }, { "epoch": 35.12, "learning_rate": 2.454655029093932e-06, "loss": 0.0179, "step": 424300 }, { "epoch": 35.13, "learning_rate": 2.4504987531172075e-06, "loss": 0.0169, "step": 424400 }, { "epoch": 35.14, "learning_rate": 2.446342477140482e-06, "loss": 0.0154, "step": 424500 }, { "epoch": 35.14, "eval_loss": 0.18243065476417542, "eval_runtime": 4694.6118, "eval_samples_per_second": 16.103, "eval_steps_per_second": 1.342, "eval_wer": 0.04831069490662322, "step": 424500 }, { "epoch": 35.15, "learning_rate": 2.4421862011637576e-06, "loss": 0.0152, "step": 424600 }, { "epoch": 35.16, "learning_rate": 2.4380299251870327e-06, "loss": 0.016, "step": 424700 }, { "epoch": 35.17, "learning_rate": 2.433873649210308e-06, "loss": 0.0173, "step": 424800 }, { "epoch": 35.17, "learning_rate": 2.429717373233583e-06, "loss": 0.0162, "step": 424900 }, { "epoch": 35.18, "learning_rate": 2.425561097256858e-06, "loss": 0.0223, "step": 425000 }, { "epoch": 35.18, "eval_loss": 0.18454158306121826, "eval_runtime": 4770.9234, "eval_samples_per_second": 15.845, "eval_steps_per_second": 1.32, "eval_wer": 0.048730694260470366, "step": 425000 }, { "epoch": 35.19, "learning_rate": 2.4214048212801335e-06, "loss": 0.0178, "step": 425100 }, { "epoch": 35.2, "learning_rate": 2.417248545303408e-06, "loss": 0.0155, "step": 425200 }, { "epoch": 35.21, "learning_rate": 2.4130922693266833e-06, "loss": 0.0124, "step": 425300 }, { "epoch": 35.22, "learning_rate": 2.408935993349959e-06, "loss": 0.0136, "step": 425400 }, { "epoch": 35.22, "learning_rate": 2.404779717373234e-06, "loss": 0.016, "step": 425500 }, { "epoch": 35.22, "eval_loss": 0.18696685135364532, "eval_runtime": 4717.4581, "eval_samples_per_second": 16.025, "eval_steps_per_second": 1.335, "eval_wer": 0.04918915509360755, "step": 425500 }, { "epoch": 35.23, "learning_rate": 2.400623441396509e-06, "loss": 0.0186, "step": 425600 }, { "epoch": 35.24, "learning_rate": 2.396467165419784e-06, "loss": 0.016, "step": 425700 }, { "epoch": 35.25, "learning_rate": 2.3923524522028268e-06, "loss": 0.0143, "step": 425800 }, { "epoch": 35.26, "learning_rate": 2.3881961762261014e-06, "loss": 0.0139, "step": 425900 }, { "epoch": 35.26, "learning_rate": 2.3840399002493765e-06, "loss": 0.0126, "step": 426000 }, { "epoch": 35.26, "eval_loss": 0.1873481124639511, "eval_runtime": 4757.2176, "eval_samples_per_second": 15.891, "eval_steps_per_second": 1.324, "eval_wer": 0.04874454039301478, "step": 426000 }, { "epoch": 35.27, "learning_rate": 2.379883624272652e-06, "loss": 0.0147, "step": 426100 }, { "epoch": 35.28, "learning_rate": 2.375727348295927e-06, "loss": 0.0131, "step": 426200 }, { "epoch": 35.29, "learning_rate": 2.3715710723192022e-06, "loss": 0.0151, "step": 426300 }, { "epoch": 35.3, "learning_rate": 2.3674563591022445e-06, "loss": 0.0236, "step": 426400 }, { "epoch": 35.31, "learning_rate": 2.36330008312552e-06, "loss": 0.0143, "step": 426500 }, { "epoch": 35.31, "eval_loss": 0.1857605129480362, "eval_runtime": 5102.2717, "eval_samples_per_second": 14.816, "eval_steps_per_second": 1.235, "eval_wer": 0.04813992593857548, "step": 426500 }, { "epoch": 35.31, "learning_rate": 2.359143807148795e-06, "loss": 0.0134, "step": 426600 }, { "epoch": 35.32, "learning_rate": 2.3549875311720698e-06, "loss": 0.0156, "step": 426700 }, { "epoch": 35.33, "learning_rate": 2.3508312551953453e-06, "loss": 0.0181, "step": 426800 }, { "epoch": 35.34, "learning_rate": 2.3466749792186204e-06, "loss": 0.0173, "step": 426900 }, { "epoch": 35.35, "learning_rate": 2.3425187032418955e-06, "loss": 0.0147, "step": 427000 }, { "epoch": 35.35, "eval_loss": 0.18606716394424438, "eval_runtime": 5015.0447, "eval_samples_per_second": 15.074, "eval_steps_per_second": 1.256, "eval_wer": 0.04837531019183047, "step": 427000 }, { "epoch": 35.36, "learning_rate": 2.3383624272651705e-06, "loss": 0.0153, "step": 427100 }, { "epoch": 35.36, "learning_rate": 2.3342061512884456e-06, "loss": 0.0129, "step": 427200 }, { "epoch": 35.37, "learning_rate": 2.330049875311721e-06, "loss": 0.0171, "step": 427300 }, { "epoch": 35.38, "learning_rate": 2.325893599334996e-06, "loss": 0.0152, "step": 427400 }, { "epoch": 35.39, "learning_rate": 2.3217373233582713e-06, "loss": 0.015, "step": 427500 }, { "epoch": 35.39, "eval_loss": 0.18783314526081085, "eval_runtime": 4851.4675, "eval_samples_per_second": 15.582, "eval_steps_per_second": 1.299, "eval_wer": 0.04860454060839906, "step": 427500 }, { "epoch": 35.4, "learning_rate": 2.3175810473815464e-06, "loss": 0.0175, "step": 427600 }, { "epoch": 35.41, "learning_rate": 2.3134247714048215e-06, "loss": 0.0156, "step": 427700 }, { "epoch": 35.41, "learning_rate": 2.3092684954280966e-06, "loss": 0.0136, "step": 427800 }, { "epoch": 35.42, "learning_rate": 2.3051122194513717e-06, "loss": 0.023, "step": 427900 }, { "epoch": 35.43, "learning_rate": 2.300955943474647e-06, "loss": 0.0206, "step": 428000 }, { "epoch": 35.43, "eval_loss": 0.18829987943172455, "eval_runtime": 4798.1705, "eval_samples_per_second": 15.755, "eval_steps_per_second": 1.313, "eval_wer": 0.04953069302970303, "step": 428000 }, { "epoch": 35.44, "learning_rate": 2.296799667497922e-06, "loss": 0.0222, "step": 428100 }, { "epoch": 35.45, "learning_rate": 2.292643391521197e-06, "loss": 0.013, "step": 428200 }, { "epoch": 35.46, "learning_rate": 2.2884871155444725e-06, "loss": 0.0305, "step": 428300 }, { "epoch": 35.46, "learning_rate": 2.2843308395677476e-06, "loss": 0.0186, "step": 428400 }, { "epoch": 35.47, "learning_rate": 2.2801745635910227e-06, "loss": 0.0216, "step": 428500 }, { "epoch": 35.47, "eval_loss": 0.18424326181411743, "eval_runtime": 5037.7667, "eval_samples_per_second": 15.006, "eval_steps_per_second": 1.251, "eval_wer": 0.04787223404271686, "step": 428500 }, { "epoch": 35.48, "learning_rate": 2.2760182876142978e-06, "loss": 0.0169, "step": 428600 }, { "epoch": 35.49, "learning_rate": 2.2719035743973405e-06, "loss": 0.0144, "step": 428700 }, { "epoch": 35.5, "learning_rate": 2.267747298420615e-06, "loss": 0.0176, "step": 428800 }, { "epoch": 35.5, "learning_rate": 2.2635910224438902e-06, "loss": 0.0126, "step": 428900 }, { "epoch": 35.51, "learning_rate": 2.2594347464671657e-06, "loss": 0.0146, "step": 429000 }, { "epoch": 35.51, "eval_loss": 0.19004803895950317, "eval_runtime": 4753.8045, "eval_samples_per_second": 15.902, "eval_steps_per_second": 1.325, "eval_wer": 0.048872232504257686, "step": 429000 }, { "epoch": 35.52, "learning_rate": 2.255278470490441e-06, "loss": 0.0141, "step": 429100 }, { "epoch": 35.53, "learning_rate": 2.251122194513716e-06, "loss": 0.0136, "step": 429200 }, { "epoch": 35.54, "learning_rate": 2.246965918536991e-06, "loss": 0.0139, "step": 429300 }, { "epoch": 35.55, "learning_rate": 2.242809642560266e-06, "loss": 0.0159, "step": 429400 }, { "epoch": 35.55, "learning_rate": 2.238653366583541e-06, "loss": 0.0191, "step": 429500 }, { "epoch": 35.55, "eval_loss": 0.18872034549713135, "eval_runtime": 4774.1994, "eval_samples_per_second": 15.834, "eval_steps_per_second": 1.32, "eval_wer": 0.04817223358117911, "step": 429500 }, { "epoch": 35.56, "learning_rate": 2.2344970906068163e-06, "loss": 0.0205, "step": 429600 }, { "epoch": 35.57, "learning_rate": 2.230340814630092e-06, "loss": 0.017, "step": 429700 }, { "epoch": 35.58, "learning_rate": 2.226184538653367e-06, "loss": 0.0153, "step": 429800 }, { "epoch": 35.59, "learning_rate": 2.222028262676642e-06, "loss": 0.0134, "step": 429900 }, { "epoch": 35.6, "learning_rate": 2.217871986699917e-06, "loss": 0.0166, "step": 430000 }, { "epoch": 35.6, "eval_loss": 0.18626925349235535, "eval_runtime": 4841.9518, "eval_samples_per_second": 15.613, "eval_steps_per_second": 1.301, "eval_wer": 0.047966080052184536, "step": 430000 }, { "epoch": 35.6, "learning_rate": 2.213715710723192e-06, "loss": 0.0131, "step": 430100 }, { "epoch": 35.61, "learning_rate": 2.2095594347464673e-06, "loss": 0.0181, "step": 430200 }, { "epoch": 35.62, "learning_rate": 2.2054031587697424e-06, "loss": 0.019, "step": 430300 }, { "epoch": 35.63, "learning_rate": 2.2012468827930175e-06, "loss": 0.014, "step": 430400 }, { "epoch": 35.64, "learning_rate": 2.197090606816293e-06, "loss": 0.0145, "step": 430500 }, { "epoch": 35.64, "eval_loss": 0.18772979080677032, "eval_runtime": 5352.0092, "eval_samples_per_second": 14.125, "eval_steps_per_second": 1.177, "eval_wer": 0.047846080236799635, "step": 430500 }, { "epoch": 35.65, "learning_rate": 2.192934330839568e-06, "loss": 0.0181, "step": 430600 }, { "epoch": 35.65, "learning_rate": 2.188778054862843e-06, "loss": 0.017, "step": 430700 }, { "epoch": 35.66, "learning_rate": 2.1846217788861183e-06, "loss": 0.016, "step": 430800 }, { "epoch": 35.67, "learning_rate": 2.1804655029093933e-06, "loss": 0.0153, "step": 430900 }, { "epoch": 35.68, "learning_rate": 2.1763092269326684e-06, "loss": 0.0136, "step": 431000 }, { "epoch": 35.68, "eval_loss": 0.18888543546199799, "eval_runtime": 4619.468, "eval_samples_per_second": 16.364, "eval_steps_per_second": 1.364, "eval_wer": 0.047752234227331956, "step": 431000 }, { "epoch": 35.69, "learning_rate": 2.1721529509559435e-06, "loss": 0.0148, "step": 431100 }, { "epoch": 35.7, "learning_rate": 2.167996674979219e-06, "loss": 0.0172, "step": 431200 }, { "epoch": 35.7, "learning_rate": 2.163840399002494e-06, "loss": 0.0155, "step": 431300 }, { "epoch": 35.71, "learning_rate": 2.159684123025769e-06, "loss": 0.0168, "step": 431400 }, { "epoch": 35.72, "learning_rate": 2.1555278470490443e-06, "loss": 0.0134, "step": 431500 }, { "epoch": 35.72, "eval_loss": 0.18361884355545044, "eval_runtime": 6103.4018, "eval_samples_per_second": 12.386, "eval_steps_per_second": 1.032, "eval_wer": 0.04774761884981715, "step": 431500 }, { "epoch": 35.73, "learning_rate": 2.1513715710723194e-06, "loss": 0.0146, "step": 431600 }, { "epoch": 35.74, "learning_rate": 2.1472152950955945e-06, "loss": 0.0159, "step": 431700 }, { "epoch": 35.75, "learning_rate": 2.1431005818786368e-06, "loss": 0.0135, "step": 431800 }, { "epoch": 35.75, "learning_rate": 2.1389443059019123e-06, "loss": 0.0178, "step": 431900 }, { "epoch": 35.76, "learning_rate": 2.1347880299251874e-06, "loss": 0.0125, "step": 432000 }, { "epoch": 35.76, "eval_loss": 0.1898612380027771, "eval_runtime": 4988.451, "eval_samples_per_second": 15.154, "eval_steps_per_second": 1.263, "eval_wer": 0.048036849174078196, "step": 432000 }, { "epoch": 35.77, "learning_rate": 2.130631753948462e-06, "loss": 0.0147, "step": 432100 }, { "epoch": 35.78, "learning_rate": 2.1264754779717376e-06, "loss": 0.015, "step": 432200 }, { "epoch": 35.79, "learning_rate": 2.1223192019950126e-06, "loss": 0.0134, "step": 432300 }, { "epoch": 35.79, "learning_rate": 2.1181629260182877e-06, "loss": 0.0145, "step": 432400 }, { "epoch": 35.8, "learning_rate": 2.114006650041563e-06, "loss": 0.0156, "step": 432500 }, { "epoch": 35.8, "eval_loss": 0.18623985350131989, "eval_runtime": 4673.1113, "eval_samples_per_second": 16.177, "eval_steps_per_second": 1.348, "eval_wer": 0.0480430030107646, "step": 432500 }, { "epoch": 35.81, "learning_rate": 2.109850374064838e-06, "loss": 0.0178, "step": 432600 }, { "epoch": 35.82, "learning_rate": 2.1056940980881134e-06, "loss": 0.0194, "step": 432700 }, { "epoch": 35.83, "learning_rate": 2.101537822111388e-06, "loss": 0.0166, "step": 432800 }, { "epoch": 35.84, "learning_rate": 2.0973815461346636e-06, "loss": 0.0161, "step": 432900 }, { "epoch": 35.84, "learning_rate": 2.0932252701579387e-06, "loss": 0.0214, "step": 433000 }, { "epoch": 35.84, "eval_loss": 0.18444742262363434, "eval_runtime": 4987.2961, "eval_samples_per_second": 15.158, "eval_steps_per_second": 1.263, "eval_wer": 0.04809992600011385, "step": 433000 }, { "epoch": 35.85, "learning_rate": 2.089068994181214e-06, "loss": 0.0175, "step": 433100 }, { "epoch": 35.86, "learning_rate": 2.084912718204489e-06, "loss": 0.0143, "step": 433200 }, { "epoch": 35.87, "learning_rate": 2.080756442227764e-06, "loss": 0.0158, "step": 433300 }, { "epoch": 35.88, "learning_rate": 2.0766001662510395e-06, "loss": 0.0166, "step": 433400 }, { "epoch": 35.89, "learning_rate": 2.072443890274314e-06, "loss": 0.0142, "step": 433500 }, { "epoch": 35.89, "eval_loss": 0.1824430674314499, "eval_runtime": 4940.676, "eval_samples_per_second": 15.301, "eval_steps_per_second": 1.275, "eval_wer": 0.04713685055869145, "step": 433500 }, { "epoch": 35.89, "learning_rate": 2.0682876142975893e-06, "loss": 0.0209, "step": 433600 }, { "epoch": 35.9, "learning_rate": 2.064131338320865e-06, "loss": 0.015, "step": 433700 }, { "epoch": 35.91, "learning_rate": 2.05997506234414e-06, "loss": 0.0172, "step": 433800 }, { "epoch": 35.92, "learning_rate": 2.055818786367415e-06, "loss": 0.0145, "step": 433900 }, { "epoch": 35.93, "learning_rate": 2.05166251039069e-06, "loss": 0.0168, "step": 434000 }, { "epoch": 35.93, "eval_loss": 0.18661919236183167, "eval_runtime": 4727.1706, "eval_samples_per_second": 15.992, "eval_steps_per_second": 1.333, "eval_wer": 0.047624542116089054, "step": 434000 }, { "epoch": 35.94, "learning_rate": 2.047506234413965e-06, "loss": 0.0175, "step": 434100 }, { "epoch": 35.94, "learning_rate": 2.0433499584372403e-06, "loss": 0.0167, "step": 434200 }, { "epoch": 35.95, "learning_rate": 2.0391936824605153e-06, "loss": 0.0142, "step": 434300 }, { "epoch": 35.96, "learning_rate": 2.035037406483791e-06, "loss": 0.017, "step": 434400 }, { "epoch": 35.97, "learning_rate": 2.030881130507066e-06, "loss": 0.0144, "step": 434500 }, { "epoch": 35.97, "eval_loss": 0.18267148733139038, "eval_runtime": 4749.4792, "eval_samples_per_second": 15.916, "eval_steps_per_second": 1.326, "eval_wer": 0.04750608075987575, "step": 434500 }, { "epoch": 35.98, "learning_rate": 2.026724854530341e-06, "loss": 0.0139, "step": 434600 }, { "epoch": 35.99, "learning_rate": 2.022568578553616e-06, "loss": 0.0163, "step": 434700 }, { "epoch": 35.99, "learning_rate": 2.0184123025768912e-06, "loss": 0.0159, "step": 434800 }, { "epoch": 36.0, "learning_rate": 2.0142560266001663e-06, "loss": 0.0191, "step": 434900 }, { "epoch": 36.01, "learning_rate": 2.0100997506234414e-06, "loss": 0.0128, "step": 435000 }, { "epoch": 36.01, "eval_loss": 0.18690939247608185, "eval_runtime": 4691.3081, "eval_samples_per_second": 16.114, "eval_steps_per_second": 1.343, "eval_wer": 0.04817069512200751, "step": 435000 }, { "epoch": 36.02, "learning_rate": 2.005985037406484e-06, "loss": 0.013, "step": 435100 }, { "epoch": 36.03, "learning_rate": 2.001828761429759e-06, "loss": 0.0143, "step": 435200 }, { "epoch": 36.03, "learning_rate": 1.9976724854530343e-06, "loss": 0.0137, "step": 435300 }, { "epoch": 36.04, "learning_rate": 1.9935162094763094e-06, "loss": 0.0161, "step": 435400 }, { "epoch": 36.05, "learning_rate": 1.9893599334995845e-06, "loss": 0.0135, "step": 435500 }, { "epoch": 36.05, "eval_loss": 0.18987324833869934, "eval_runtime": 4746.3476, "eval_samples_per_second": 15.927, "eval_steps_per_second": 1.327, "eval_wer": 0.048629155955144686, "step": 435500 }, { "epoch": 36.06, "learning_rate": 1.9852036575228596e-06, "loss": 0.0149, "step": 435600 }, { "epoch": 36.07, "learning_rate": 1.9810473815461347e-06, "loss": 0.0189, "step": 435700 }, { "epoch": 36.08, "learning_rate": 1.97689110556941e-06, "loss": 0.0155, "step": 435800 }, { "epoch": 36.08, "learning_rate": 1.9727348295926853e-06, "loss": 0.0153, "step": 435900 }, { "epoch": 36.09, "learning_rate": 1.9685785536159604e-06, "loss": 0.0139, "step": 436000 }, { "epoch": 36.09, "eval_loss": 0.19107340276241302, "eval_runtime": 4868.7275, "eval_samples_per_second": 15.527, "eval_steps_per_second": 1.294, "eval_wer": 0.04841992550780691, "step": 436000 }, { "epoch": 36.1, "learning_rate": 1.9644222776392354e-06, "loss": 0.0143, "step": 436100 }, { "epoch": 36.11, "learning_rate": 1.9602660016625105e-06, "loss": 0.0153, "step": 436200 }, { "epoch": 36.12, "learning_rate": 1.9561097256857856e-06, "loss": 0.0138, "step": 436300 }, { "epoch": 36.13, "learning_rate": 1.9519534497090607e-06, "loss": 0.0181, "step": 436400 }, { "epoch": 36.13, "learning_rate": 1.947797173732336e-06, "loss": 0.0128, "step": 436500 }, { "epoch": 36.13, "eval_loss": 0.18756169080734253, "eval_runtime": 4816.6563, "eval_samples_per_second": 15.694, "eval_steps_per_second": 1.308, "eval_wer": 0.04821223351964074, "step": 436500 }, { "epoch": 36.14, "learning_rate": 1.9436408977556113e-06, "loss": 0.0128, "step": 436600 }, { "epoch": 36.15, "learning_rate": 1.9394846217788864e-06, "loss": 0.0146, "step": 436700 }, { "epoch": 36.16, "learning_rate": 1.9353283458021615e-06, "loss": 0.0177, "step": 436800 }, { "epoch": 36.17, "learning_rate": 1.9311720698254366e-06, "loss": 0.0157, "step": 436900 }, { "epoch": 36.18, "learning_rate": 1.9270157938487117e-06, "loss": 0.0114, "step": 437000 }, { "epoch": 36.18, "eval_loss": 0.18915089964866638, "eval_runtime": 4818.0324, "eval_samples_per_second": 15.69, "eval_steps_per_second": 1.308, "eval_wer": 0.048738386556328374, "step": 437000 }, { "epoch": 36.18, "learning_rate": 1.922859517871987e-06, "loss": 0.0144, "step": 437100 }, { "epoch": 36.19, "learning_rate": 1.918703241895262e-06, "loss": 0.0168, "step": 437200 }, { "epoch": 36.2, "learning_rate": 1.9145469659185374e-06, "loss": 0.018, "step": 437300 }, { "epoch": 36.21, "learning_rate": 1.9103906899418125e-06, "loss": 0.0153, "step": 437400 }, { "epoch": 36.22, "learning_rate": 1.9062344139650874e-06, "loss": 0.0137, "step": 437500 }, { "epoch": 36.22, "eval_loss": 0.19094204902648926, "eval_runtime": 5927.001, "eval_samples_per_second": 12.754, "eval_steps_per_second": 1.063, "eval_wer": 0.04826300267230358, "step": 437500 }, { "epoch": 36.23, "learning_rate": 1.9020781379883627e-06, "loss": 0.0144, "step": 437600 }, { "epoch": 36.23, "learning_rate": 1.8979218620116378e-06, "loss": 0.0157, "step": 437700 }, { "epoch": 36.24, "learning_rate": 1.893765586034913e-06, "loss": 0.0166, "step": 437800 }, { "epoch": 36.25, "learning_rate": 1.889609310058188e-06, "loss": 0.0158, "step": 437900 }, { "epoch": 36.26, "learning_rate": 1.885453034081463e-06, "loss": 0.0161, "step": 438000 }, { "epoch": 36.26, "eval_loss": 0.19109764695167542, "eval_runtime": 4693.2096, "eval_samples_per_second": 16.107, "eval_steps_per_second": 1.342, "eval_wer": 0.048267618049818384, "step": 438000 }, { "epoch": 36.27, "learning_rate": 1.8813383208645055e-06, "loss": 0.0146, "step": 438100 }, { "epoch": 36.27, "learning_rate": 1.8771820448877806e-06, "loss": 0.0181, "step": 438200 }, { "epoch": 36.28, "learning_rate": 1.873025768911056e-06, "loss": 0.0133, "step": 438300 }, { "epoch": 36.29, "learning_rate": 1.868869492934331e-06, "loss": 0.0184, "step": 438400 }, { "epoch": 36.3, "learning_rate": 1.8647132169576063e-06, "loss": 0.0128, "step": 438500 }, { "epoch": 36.3, "eval_loss": 0.1889868676662445, "eval_runtime": 4978.7234, "eval_samples_per_second": 15.184, "eval_steps_per_second": 1.265, "eval_wer": 0.04798454156224375, "step": 438500 }, { "epoch": 36.31, "learning_rate": 1.8605569409808812e-06, "loss": 0.0151, "step": 438600 }, { "epoch": 36.32, "learning_rate": 1.8564006650041563e-06, "loss": 0.0141, "step": 438700 }, { "epoch": 36.32, "learning_rate": 1.8522859517871988e-06, "loss": 0.0176, "step": 438800 }, { "epoch": 36.33, "learning_rate": 1.8481296758104738e-06, "loss": 0.0129, "step": 438900 }, { "epoch": 36.34, "learning_rate": 1.8439733998337492e-06, "loss": 0.0128, "step": 439000 }, { "epoch": 36.34, "eval_loss": 0.19092309474945068, "eval_runtime": 4752.9164, "eval_samples_per_second": 15.905, "eval_steps_per_second": 1.326, "eval_wer": 0.04789069555277607, "step": 439000 }, { "epoch": 36.35, "learning_rate": 1.8398171238570242e-06, "loss": 0.0126, "step": 439100 }, { "epoch": 36.36, "learning_rate": 1.8356608478802995e-06, "loss": 0.0181, "step": 439200 }, { "epoch": 36.37, "learning_rate": 1.8315045719035746e-06, "loss": 0.0138, "step": 439300 }, { "epoch": 36.37, "learning_rate": 1.8273482959268495e-06, "loss": 0.0136, "step": 439400 }, { "epoch": 36.38, "learning_rate": 1.8231920199501248e-06, "loss": 0.0157, "step": 439500 }, { "epoch": 36.38, "eval_loss": 0.18842723965644836, "eval_runtime": 5025.0902, "eval_samples_per_second": 15.044, "eval_steps_per_second": 1.254, "eval_wer": 0.04808454140839783, "step": 439500 }, { "epoch": 36.39, "learning_rate": 1.8190357439734e-06, "loss": 0.0128, "step": 439600 }, { "epoch": 36.4, "learning_rate": 1.8148794679966752e-06, "loss": 0.0176, "step": 439700 }, { "epoch": 36.41, "learning_rate": 1.8107231920199503e-06, "loss": 0.0158, "step": 439800 }, { "epoch": 36.42, "learning_rate": 1.8065669160432254e-06, "loss": 0.0149, "step": 439900 }, { "epoch": 36.42, "learning_rate": 1.8024106400665007e-06, "loss": 0.0116, "step": 440000 }, { "epoch": 36.42, "eval_loss": 0.186118945479393, "eval_runtime": 4709.8186, "eval_samples_per_second": 16.051, "eval_steps_per_second": 1.338, "eval_wer": 0.04794915700129692, "step": 440000 }, { "epoch": 36.43, "learning_rate": 1.7982543640897756e-06, "loss": 0.016, "step": 440100 }, { "epoch": 36.44, "learning_rate": 1.794098088113051e-06, "loss": 0.0136, "step": 440200 }, { "epoch": 36.45, "learning_rate": 1.789941812136326e-06, "loss": 0.0133, "step": 440300 }, { "epoch": 36.46, "learning_rate": 1.7858270989193685e-06, "loss": 0.013, "step": 440400 }, { "epoch": 36.47, "learning_rate": 1.7817123857024107e-06, "loss": 0.0166, "step": 440500 }, { "epoch": 36.47, "eval_loss": 0.1861451268196106, "eval_runtime": 4953.5962, "eval_samples_per_second": 15.261, "eval_steps_per_second": 1.272, "eval_wer": 0.047953772378811724, "step": 440500 }, { "epoch": 36.47, "learning_rate": 1.777556109725686e-06, "loss": 0.0174, "step": 440600 }, { "epoch": 36.48, "learning_rate": 1.773399833748961e-06, "loss": 0.0119, "step": 440700 }, { "epoch": 36.49, "learning_rate": 1.7692435577722362e-06, "loss": 0.0144, "step": 440800 }, { "epoch": 36.5, "learning_rate": 1.7650872817955115e-06, "loss": 0.0216, "step": 440900 }, { "epoch": 36.51, "learning_rate": 1.7609310058187864e-06, "loss": 0.013, "step": 441000 }, { "epoch": 36.51, "eval_loss": 0.19139191508293152, "eval_runtime": 5840.876, "eval_samples_per_second": 12.942, "eval_steps_per_second": 1.079, "eval_wer": 0.04836915635514407, "step": 441000 }, { "epoch": 36.51, "learning_rate": 1.7567747298420617e-06, "loss": 0.0162, "step": 441100 }, { "epoch": 36.52, "learning_rate": 1.7526184538653368e-06, "loss": 0.0137, "step": 441200 }, { "epoch": 36.53, "learning_rate": 1.7484621778886119e-06, "loss": 0.0126, "step": 441300 }, { "epoch": 36.54, "learning_rate": 1.7443059019118872e-06, "loss": 0.014, "step": 441400 }, { "epoch": 36.55, "learning_rate": 1.7401496259351623e-06, "loss": 0.0154, "step": 441500 }, { "epoch": 36.55, "eval_loss": 0.1931745857000351, "eval_runtime": 5892.1899, "eval_samples_per_second": 12.83, "eval_steps_per_second": 1.069, "eval_wer": 0.04834607946757005, "step": 441500 }, { "epoch": 36.56, "learning_rate": 1.7359933499584376e-06, "loss": 0.0144, "step": 441600 }, { "epoch": 36.56, "learning_rate": 1.7318370739817125e-06, "loss": 0.014, "step": 441700 }, { "epoch": 36.57, "learning_rate": 1.7276807980049875e-06, "loss": 0.0162, "step": 441800 }, { "epoch": 36.58, "learning_rate": 1.7235245220282629e-06, "loss": 0.0145, "step": 441900 }, { "epoch": 36.59, "learning_rate": 1.719368246051538e-06, "loss": 0.0156, "step": 442000 }, { "epoch": 36.59, "eval_loss": 0.19164663553237915, "eval_runtime": 5904.9887, "eval_samples_per_second": 12.802, "eval_steps_per_second": 1.067, "eval_wer": 0.04848300233384256, "step": 442000 }, { "epoch": 36.6, "learning_rate": 1.7152119700748132e-06, "loss": 0.0161, "step": 442100 }, { "epoch": 36.61, "learning_rate": 1.7110556940980883e-06, "loss": 0.0165, "step": 442200 }, { "epoch": 36.61, "learning_rate": 1.7068994181213632e-06, "loss": 0.0189, "step": 442300 }, { "epoch": 36.62, "learning_rate": 1.7027431421446385e-06, "loss": 0.0154, "step": 442400 }, { "epoch": 36.63, "learning_rate": 1.6985868661679136e-06, "loss": 0.0162, "step": 442500 }, { "epoch": 36.63, "eval_loss": 0.19317083060741425, "eval_runtime": 5957.0804, "eval_samples_per_second": 12.69, "eval_steps_per_second": 1.058, "eval_wer": 0.04845531006875374, "step": 442500 }, { "epoch": 36.64, "learning_rate": 1.694430590191189e-06, "loss": 0.0162, "step": 442600 }, { "epoch": 36.65, "learning_rate": 1.690274314214464e-06, "loss": 0.0149, "step": 442700 }, { "epoch": 36.66, "learning_rate": 1.6861180382377391e-06, "loss": 0.0125, "step": 442800 }, { "epoch": 36.66, "learning_rate": 1.6819617622610144e-06, "loss": 0.0135, "step": 442900 }, { "epoch": 36.67, "learning_rate": 1.6778054862842893e-06, "loss": 0.0137, "step": 443000 }, { "epoch": 36.67, "eval_loss": 0.19152706861495972, "eval_runtime": 5965.6405, "eval_samples_per_second": 12.672, "eval_steps_per_second": 1.056, "eval_wer": 0.04790454168532048, "step": 443000 }, { "epoch": 36.68, "learning_rate": 1.6736492103075646e-06, "loss": 0.0153, "step": 443100 }, { "epoch": 36.69, "learning_rate": 1.6694929343308397e-06, "loss": 0.016, "step": 443200 }, { "epoch": 36.7, "learning_rate": 1.6653366583541148e-06, "loss": 0.014, "step": 443300 }, { "epoch": 36.71, "learning_rate": 1.66118038237739e-06, "loss": 0.0132, "step": 443400 }, { "epoch": 36.71, "learning_rate": 1.6570241064006652e-06, "loss": 0.0177, "step": 443500 }, { "epoch": 36.71, "eval_loss": 0.19010721147060394, "eval_runtime": 6073.7641, "eval_samples_per_second": 12.446, "eval_steps_per_second": 1.037, "eval_wer": 0.047687618942124706, "step": 443500 }, { "epoch": 36.72, "learning_rate": 1.6528678304239405e-06, "loss": 0.015, "step": 443600 }, { "epoch": 36.73, "learning_rate": 1.6487115544472154e-06, "loss": 0.0128, "step": 443700 }, { "epoch": 36.74, "learning_rate": 1.6445552784704905e-06, "loss": 0.0141, "step": 443800 }, { "epoch": 36.75, "learning_rate": 1.6403990024937658e-06, "loss": 0.0137, "step": 443900 }, { "epoch": 36.75, "learning_rate": 1.6362427265170409e-06, "loss": 0.0161, "step": 444000 }, { "epoch": 36.75, "eval_loss": 0.18940725922584534, "eval_runtime": 6309.1223, "eval_samples_per_second": 11.982, "eval_steps_per_second": 0.999, "eval_wer": 0.04788761863443287, "step": 444000 }, { "epoch": 36.76, "learning_rate": 1.6320864505403162e-06, "loss": 0.0147, "step": 444100 }, { "epoch": 36.77, "learning_rate": 1.627930174563591e-06, "loss": 0.0156, "step": 444200 }, { "epoch": 36.78, "learning_rate": 1.6237738985868661e-06, "loss": 0.0167, "step": 444300 }, { "epoch": 36.79, "learning_rate": 1.6196176226101414e-06, "loss": 0.0131, "step": 444400 }, { "epoch": 36.8, "learning_rate": 1.6154613466334165e-06, "loss": 0.0151, "step": 444500 }, { "epoch": 36.8, "eval_loss": 0.1907188594341278, "eval_runtime": 4684.2097, "eval_samples_per_second": 16.138, "eval_steps_per_second": 1.345, "eval_wer": 0.04784146485928483, "step": 444500 }, { "epoch": 36.8, "learning_rate": 1.6113050706566918e-06, "loss": 0.0162, "step": 444600 }, { "epoch": 36.81, "learning_rate": 1.607148794679967e-06, "loss": 0.0156, "step": 444700 }, { "epoch": 36.82, "learning_rate": 1.6029925187032418e-06, "loss": 0.0135, "step": 444800 }, { "epoch": 36.83, "learning_rate": 1.5988362427265171e-06, "loss": 0.0155, "step": 444900 }, { "epoch": 36.84, "learning_rate": 1.5946799667497922e-06, "loss": 0.0135, "step": 445000 }, { "epoch": 36.84, "eval_loss": 0.19115544855594635, "eval_runtime": 5583.8375, "eval_samples_per_second": 13.538, "eval_steps_per_second": 1.128, "eval_wer": 0.04779377262496519, "step": 445000 }, { "epoch": 36.85, "learning_rate": 1.5905236907730675e-06, "loss": 0.0165, "step": 445100 }, { "epoch": 36.85, "learning_rate": 1.5863674147963426e-06, "loss": 0.0157, "step": 445200 }, { "epoch": 36.86, "learning_rate": 1.582252701579385e-06, "loss": 0.0137, "step": 445300 }, { "epoch": 36.87, "learning_rate": 1.5780964256026602e-06, "loss": 0.0161, "step": 445400 }, { "epoch": 36.88, "learning_rate": 1.5739401496259353e-06, "loss": 0.013, "step": 445500 }, { "epoch": 36.88, "eval_loss": 0.1881672888994217, "eval_runtime": 4948.1716, "eval_samples_per_second": 15.277, "eval_steps_per_second": 1.273, "eval_wer": 0.047816849512539214, "step": 445500 }, { "epoch": 36.89, "learning_rate": 1.5697838736492106e-06, "loss": 0.0122, "step": 445600 }, { "epoch": 36.9, "learning_rate": 1.5656275976724854e-06, "loss": 0.0128, "step": 445700 }, { "epoch": 36.9, "learning_rate": 1.5614713216957607e-06, "loss": 0.0164, "step": 445800 }, { "epoch": 36.91, "learning_rate": 1.5573150457190358e-06, "loss": 0.0173, "step": 445900 }, { "epoch": 36.92, "learning_rate": 1.553158769742311e-06, "loss": 0.0151, "step": 446000 }, { "epoch": 36.92, "eval_loss": 0.19035682082176208, "eval_runtime": 4693.6422, "eval_samples_per_second": 16.106, "eval_steps_per_second": 1.342, "eval_wer": 0.047595311391828626, "step": 446000 }, { "epoch": 36.93, "learning_rate": 1.5490024937655862e-06, "loss": 0.0121, "step": 446100 }, { "epoch": 36.94, "learning_rate": 1.5448462177888613e-06, "loss": 0.0176, "step": 446200 }, { "epoch": 36.95, "learning_rate": 1.5406899418121366e-06, "loss": 0.0165, "step": 446300 }, { "epoch": 36.95, "learning_rate": 1.5365336658354115e-06, "loss": 0.0161, "step": 446400 }, { "epoch": 36.96, "learning_rate": 1.5323773898586866e-06, "loss": 0.0143, "step": 446500 }, { "epoch": 36.96, "eval_loss": 0.19050593674182892, "eval_runtime": 4872.0391, "eval_samples_per_second": 15.516, "eval_steps_per_second": 1.293, "eval_wer": 0.04749377308650295, "step": 446500 }, { "epoch": 36.97, "learning_rate": 1.528221113881962e-06, "loss": 0.0113, "step": 446600 }, { "epoch": 36.98, "learning_rate": 1.524064837905237e-06, "loss": 0.0129, "step": 446700 }, { "epoch": 36.99, "learning_rate": 1.5199085619285123e-06, "loss": 0.0124, "step": 446800 }, { "epoch": 37.0, "learning_rate": 1.5157522859517874e-06, "loss": 0.0145, "step": 446900 }, { "epoch": 37.0, "learning_rate": 1.5115960099750623e-06, "loss": 0.0125, "step": 447000 }, { "epoch": 37.0, "eval_loss": 0.19229859113693237, "eval_runtime": 5981.5202, "eval_samples_per_second": 12.638, "eval_steps_per_second": 1.053, "eval_wer": 0.04808607986756944, "step": 447000 }, { "epoch": 37.01, "learning_rate": 1.5074397339983376e-06, "loss": 0.0134, "step": 447100 }, { "epoch": 37.02, "learning_rate": 1.5032834580216127e-06, "loss": 0.019, "step": 447200 }, { "epoch": 37.03, "learning_rate": 1.499127182044888e-06, "loss": 0.0135, "step": 447300 }, { "epoch": 37.04, "learning_rate": 1.494970906068163e-06, "loss": 0.0144, "step": 447400 }, { "epoch": 37.04, "learning_rate": 1.4908146300914382e-06, "loss": 0.0137, "step": 447500 }, { "epoch": 37.04, "eval_loss": 0.1908280998468399, "eval_runtime": 4706.0067, "eval_samples_per_second": 16.064, "eval_steps_per_second": 1.339, "eval_wer": 0.047695311237982714, "step": 447500 }, { "epoch": 37.05, "learning_rate": 1.4866583541147135e-06, "loss": 0.0141, "step": 447600 }, { "epoch": 37.06, "learning_rate": 1.4825020781379884e-06, "loss": 0.0133, "step": 447700 }, { "epoch": 37.07, "learning_rate": 1.4783458021612637e-06, "loss": 0.0159, "step": 447800 }, { "epoch": 37.08, "learning_rate": 1.4741895261845387e-06, "loss": 0.0153, "step": 447900 }, { "epoch": 37.09, "learning_rate": 1.4700332502078138e-06, "loss": 0.0144, "step": 448000 }, { "epoch": 37.09, "eval_loss": 0.18677328526973724, "eval_runtime": 4638.5777, "eval_samples_per_second": 16.297, "eval_steps_per_second": 1.358, "eval_wer": 0.04780761875750961, "step": 448000 }, { "epoch": 37.09, "learning_rate": 1.4658769742310891e-06, "loss": 0.014, "step": 448100 }, { "epoch": 37.1, "learning_rate": 1.4617206982543642e-06, "loss": 0.0185, "step": 448200 }, { "epoch": 37.11, "learning_rate": 1.4575644222776395e-06, "loss": 0.014, "step": 448300 }, { "epoch": 37.12, "learning_rate": 1.4534081463009144e-06, "loss": 0.0153, "step": 448400 }, { "epoch": 37.13, "learning_rate": 1.4492518703241895e-06, "loss": 0.0167, "step": 448500 }, { "epoch": 37.13, "eval_loss": 0.18684706091880798, "eval_runtime": 4689.0813, "eval_samples_per_second": 16.121, "eval_steps_per_second": 1.344, "eval_wer": 0.0477999264616516, "step": 448500 }, { "epoch": 37.14, "learning_rate": 1.4450955943474648e-06, "loss": 0.0145, "step": 448600 }, { "epoch": 37.14, "learning_rate": 1.44093931837074e-06, "loss": 0.0123, "step": 448700 }, { "epoch": 37.15, "learning_rate": 1.4367830423940152e-06, "loss": 0.0169, "step": 448800 }, { "epoch": 37.16, "learning_rate": 1.43262676641729e-06, "loss": 0.0173, "step": 448900 }, { "epoch": 37.17, "learning_rate": 1.4284704904405652e-06, "loss": 0.0158, "step": 449000 }, { "epoch": 37.17, "eval_loss": 0.18813903629779816, "eval_runtime": 4882.8978, "eval_samples_per_second": 15.482, "eval_steps_per_second": 1.29, "eval_wer": 0.047695311237982714, "step": 449000 }, { "epoch": 37.18, "learning_rate": 1.4243142144638405e-06, "loss": 0.0161, "step": 449100 }, { "epoch": 37.19, "learning_rate": 1.4201579384871156e-06, "loss": 0.0163, "step": 449200 }, { "epoch": 37.19, "learning_rate": 1.4160016625103909e-06, "loss": 0.0109, "step": 449300 }, { "epoch": 37.2, "learning_rate": 1.411845386533666e-06, "loss": 0.015, "step": 449400 }, { "epoch": 37.21, "learning_rate": 1.4076891105569409e-06, "loss": 0.0168, "step": 449500 }, { "epoch": 37.21, "eval_loss": 0.18819968402385712, "eval_runtime": 4784.1367, "eval_samples_per_second": 15.801, "eval_steps_per_second": 1.317, "eval_wer": 0.04786146482851565, "step": 449500 }, { "epoch": 37.22, "learning_rate": 1.4035328345802162e-06, "loss": 0.0106, "step": 449600 }, { "epoch": 37.23, "learning_rate": 1.3993765586034913e-06, "loss": 0.0142, "step": 449700 }, { "epoch": 37.24, "learning_rate": 1.3952202826267666e-06, "loss": 0.0131, "step": 449800 }, { "epoch": 37.24, "learning_rate": 1.3910640066500417e-06, "loss": 0.0159, "step": 449900 }, { "epoch": 37.25, "learning_rate": 1.3869077306733168e-06, "loss": 0.0128, "step": 450000 }, { "epoch": 37.25, "eval_loss": 0.18858300149440765, "eval_runtime": 5779.5025, "eval_samples_per_second": 13.08, "eval_steps_per_second": 1.09, "eval_wer": 0.04776454190070477, "step": 450000 }, { "epoch": 37.26, "learning_rate": 1.382751454696592e-06, "loss": 0.0133, "step": 450100 }, { "epoch": 37.27, "learning_rate": 1.378595178719867e-06, "loss": 0.0149, "step": 450200 }, { "epoch": 37.28, "learning_rate": 1.3744389027431422e-06, "loss": 0.0173, "step": 450300 }, { "epoch": 37.28, "learning_rate": 1.3702826267664173e-06, "loss": 0.0177, "step": 450400 }, { "epoch": 37.29, "learning_rate": 1.3661263507896924e-06, "loss": 0.0145, "step": 450500 }, { "epoch": 37.29, "eval_loss": 0.1862291693687439, "eval_runtime": 5445.287, "eval_samples_per_second": 13.883, "eval_steps_per_second": 1.157, "eval_wer": 0.047695311237982714, "step": 450500 }, { "epoch": 37.3, "learning_rate": 1.3619700748129677e-06, "loss": 0.0119, "step": 450600 }, { "epoch": 37.31, "learning_rate": 1.3578137988362428e-06, "loss": 0.0173, "step": 450700 }, { "epoch": 37.32, "learning_rate": 1.3536575228595181e-06, "loss": 0.0151, "step": 450800 }, { "epoch": 37.33, "learning_rate": 1.349501246882793e-06, "loss": 0.0147, "step": 450900 }, { "epoch": 37.33, "learning_rate": 1.345344970906068e-06, "loss": 0.016, "step": 451000 }, { "epoch": 37.33, "eval_loss": 0.18831786513328552, "eval_runtime": 4890.0166, "eval_samples_per_second": 15.459, "eval_steps_per_second": 1.288, "eval_wer": 0.04759069601431382, "step": 451000 }, { "epoch": 37.34, "learning_rate": 1.3411886949293434e-06, "loss": 0.0136, "step": 451100 }, { "epoch": 37.35, "learning_rate": 1.3370324189526185e-06, "loss": 0.018, "step": 451200 }, { "epoch": 37.36, "learning_rate": 1.3328761429758938e-06, "loss": 0.0144, "step": 451300 }, { "epoch": 37.37, "learning_rate": 1.328719866999169e-06, "loss": 0.0171, "step": 451400 }, { "epoch": 37.38, "learning_rate": 1.3245635910224438e-06, "loss": 0.0132, "step": 451500 }, { "epoch": 37.38, "eval_loss": 0.1871754229068756, "eval_runtime": 5631.7877, "eval_samples_per_second": 13.423, "eval_steps_per_second": 1.119, "eval_wer": 0.047809157216681206, "step": 451500 }, { "epoch": 37.38, "learning_rate": 1.320407315045719e-06, "loss": 0.0149, "step": 451600 }, { "epoch": 37.39, "learning_rate": 1.3162510390689942e-06, "loss": 0.0142, "step": 451700 }, { "epoch": 37.4, "learning_rate": 1.3120947630922695e-06, "loss": 0.0176, "step": 451800 }, { "epoch": 37.41, "learning_rate": 1.3079384871155446e-06, "loss": 0.0154, "step": 451900 }, { "epoch": 37.42, "learning_rate": 1.3037822111388197e-06, "loss": 0.0165, "step": 452000 }, { "epoch": 37.42, "eval_loss": 0.18738719820976257, "eval_runtime": 4680.9114, "eval_samples_per_second": 16.15, "eval_steps_per_second": 1.346, "eval_wer": 0.04775531114567516, "step": 452000 }, { "epoch": 37.43, "learning_rate": 1.299625935162095e-06, "loss": 0.0176, "step": 452100 }, { "epoch": 37.43, "learning_rate": 1.2954696591853699e-06, "loss": 0.0152, "step": 452200 }, { "epoch": 37.44, "learning_rate": 1.2913133832086452e-06, "loss": 0.0168, "step": 452300 }, { "epoch": 37.45, "learning_rate": 1.2871571072319202e-06, "loss": 0.0145, "step": 452400 }, { "epoch": 37.46, "learning_rate": 1.2830008312551953e-06, "loss": 0.014, "step": 452500 }, { "epoch": 37.46, "eval_loss": 0.18841210007667542, "eval_runtime": 4789.6319, "eval_samples_per_second": 15.783, "eval_steps_per_second": 1.315, "eval_wer": 0.048030695337391786, "step": 452500 }, { "epoch": 37.47, "learning_rate": 1.2788445552784706e-06, "loss": 0.0145, "step": 452600 }, { "epoch": 37.48, "learning_rate": 1.2746882793017457e-06, "loss": 0.0153, "step": 452700 }, { "epoch": 37.48, "learning_rate": 1.270532003325021e-06, "loss": 0.0161, "step": 452800 }, { "epoch": 37.49, "learning_rate": 1.266375727348296e-06, "loss": 0.0144, "step": 452900 }, { "epoch": 37.5, "learning_rate": 1.2622194513715712e-06, "loss": 0.0146, "step": 453000 }, { "epoch": 37.5, "eval_loss": 0.1888255923986435, "eval_runtime": 5195.98, "eval_samples_per_second": 14.549, "eval_steps_per_second": 1.212, "eval_wer": 0.047796849543308394, "step": 453000 }, { "epoch": 37.51, "learning_rate": 1.2580631753948463e-06, "loss": 0.0136, "step": 453100 }, { "epoch": 37.52, "learning_rate": 1.2539068994181214e-06, "loss": 0.0174, "step": 453200 }, { "epoch": 37.52, "learning_rate": 1.2497506234413965e-06, "loss": 0.0147, "step": 453300 }, { "epoch": 37.53, "learning_rate": 1.2455943474646718e-06, "loss": 0.0206, "step": 453400 }, { "epoch": 37.54, "learning_rate": 1.241438071487947e-06, "loss": 0.0142, "step": 453500 }, { "epoch": 37.54, "eval_loss": 0.18833860754966736, "eval_runtime": 4712.3964, "eval_samples_per_second": 16.042, "eval_steps_per_second": 1.337, "eval_wer": 0.04796915697052774, "step": 453500 }, { "epoch": 37.55, "learning_rate": 1.237281795511222e-06, "loss": 0.0166, "step": 453600 }, { "epoch": 37.56, "learning_rate": 1.2331255195344973e-06, "loss": 0.0317, "step": 453700 }, { "epoch": 37.57, "learning_rate": 1.2289692435577722e-06, "loss": 0.013, "step": 453800 }, { "epoch": 37.57, "learning_rate": 1.2248129675810475e-06, "loss": 0.0183, "step": 453900 }, { "epoch": 37.58, "learning_rate": 1.2206566916043226e-06, "loss": 0.0153, "step": 454000 }, { "epoch": 37.58, "eval_loss": 0.18771396577358246, "eval_runtime": 4706.8792, "eval_samples_per_second": 16.061, "eval_steps_per_second": 1.338, "eval_wer": 0.04763684978946186, "step": 454000 }, { "epoch": 37.59, "learning_rate": 1.2165004156275979e-06, "loss": 0.0152, "step": 454100 }, { "epoch": 37.6, "learning_rate": 1.2123857024106401e-06, "loss": 0.0151, "step": 454200 }, { "epoch": 37.61, "learning_rate": 1.2082294264339152e-06, "loss": 0.0154, "step": 454300 }, { "epoch": 37.62, "learning_rate": 1.2040731504571905e-06, "loss": 0.0117, "step": 454400 }, { "epoch": 37.62, "learning_rate": 1.1999584372402328e-06, "loss": 0.0171, "step": 454500 }, { "epoch": 37.62, "eval_loss": 0.19031791388988495, "eval_runtime": 4655.0252, "eval_samples_per_second": 16.239, "eval_steps_per_second": 1.353, "eval_wer": 0.0480476183882794, "step": 454500 }, { "epoch": 37.63, "learning_rate": 1.195802161263508e-06, "loss": 0.0163, "step": 454600 }, { "epoch": 37.64, "learning_rate": 1.1916874480465503e-06, "loss": 0.0138, "step": 454700 }, { "epoch": 37.65, "learning_rate": 1.1875311720698256e-06, "loss": 0.0139, "step": 454800 }, { "epoch": 37.66, "learning_rate": 1.1833748960931007e-06, "loss": 0.0145, "step": 454900 }, { "epoch": 37.67, "learning_rate": 1.1792186201163758e-06, "loss": 0.013, "step": 455000 }, { "epoch": 37.67, "eval_loss": 0.19344937801361084, "eval_runtime": 4711.2716, "eval_samples_per_second": 16.046, "eval_steps_per_second": 1.337, "eval_wer": 0.04784300331845643, "step": 455000 }, { "epoch": 37.67, "learning_rate": 1.175062344139651e-06, "loss": 0.0172, "step": 455100 }, { "epoch": 37.68, "learning_rate": 1.1709060681629262e-06, "loss": 0.0147, "step": 455200 }, { "epoch": 37.69, "learning_rate": 1.1667497921862013e-06, "loss": 0.0171, "step": 455300 }, { "epoch": 37.7, "learning_rate": 1.1625935162094764e-06, "loss": 0.0162, "step": 455400 }, { "epoch": 37.71, "learning_rate": 1.1584372402327515e-06, "loss": 0.0135, "step": 455500 }, { "epoch": 37.71, "eval_loss": 0.18963535130023956, "eval_runtime": 4713.2672, "eval_samples_per_second": 16.039, "eval_steps_per_second": 1.337, "eval_wer": 0.047703003533840715, "step": 455500 }, { "epoch": 37.72, "learning_rate": 1.1542809642560266e-06, "loss": 0.013, "step": 455600 }, { "epoch": 37.72, "learning_rate": 1.150124688279302e-06, "loss": 0.0132, "step": 455700 }, { "epoch": 37.73, "learning_rate": 1.145968412302577e-06, "loss": 0.0147, "step": 455800 }, { "epoch": 37.74, "learning_rate": 1.141812136325852e-06, "loss": 0.0185, "step": 455900 }, { "epoch": 37.75, "learning_rate": 1.1376558603491272e-06, "loss": 0.0151, "step": 456000 }, { "epoch": 37.75, "eval_loss": 0.19113990664482117, "eval_runtime": 4773.6386, "eval_samples_per_second": 15.836, "eval_steps_per_second": 1.32, "eval_wer": 0.04771531120721353, "step": 456000 }, { "epoch": 37.76, "learning_rate": 1.1334995843724025e-06, "loss": 0.0242, "step": 456100 }, { "epoch": 37.76, "learning_rate": 1.1293433083956776e-06, "loss": 0.0135, "step": 456200 }, { "epoch": 37.77, "learning_rate": 1.1251870324189527e-06, "loss": 0.0143, "step": 456300 }, { "epoch": 37.78, "learning_rate": 1.121030756442228e-06, "loss": 0.0145, "step": 456400 }, { "epoch": 37.79, "learning_rate": 1.1168744804655029e-06, "loss": 0.0159, "step": 456500 }, { "epoch": 37.79, "eval_loss": 0.19025032222270966, "eval_runtime": 5611.9939, "eval_samples_per_second": 13.47, "eval_steps_per_second": 1.123, "eval_wer": 0.04742761934212409, "step": 456500 }, { "epoch": 37.8, "learning_rate": 1.1127182044887782e-06, "loss": 0.0116, "step": 456600 }, { "epoch": 37.81, "learning_rate": 1.1085619285120533e-06, "loss": 0.0153, "step": 456700 }, { "epoch": 37.81, "learning_rate": 1.1044056525353286e-06, "loss": 0.0116, "step": 456800 }, { "epoch": 37.82, "learning_rate": 1.1002493765586037e-06, "loss": 0.0121, "step": 456900 }, { "epoch": 37.83, "learning_rate": 1.0960931005818787e-06, "loss": 0.0151, "step": 457000 }, { "epoch": 37.83, "eval_loss": 0.19265969097614288, "eval_runtime": 4785.6444, "eval_samples_per_second": 15.796, "eval_steps_per_second": 1.316, "eval_wer": 0.04770761891135552, "step": 457000 }, { "epoch": 37.84, "learning_rate": 1.0919368246051538e-06, "loss": 0.0147, "step": 457100 }, { "epoch": 37.85, "learning_rate": 1.087780548628429e-06, "loss": 0.0179, "step": 457200 }, { "epoch": 37.86, "learning_rate": 1.0836242726517042e-06, "loss": 0.0111, "step": 457300 }, { "epoch": 37.86, "learning_rate": 1.0794679966749793e-06, "loss": 0.0122, "step": 457400 }, { "epoch": 37.87, "learning_rate": 1.0753117206982544e-06, "loss": 0.0128, "step": 457500 }, { "epoch": 37.87, "eval_loss": 0.19400928914546967, "eval_runtime": 4710.8023, "eval_samples_per_second": 16.047, "eval_steps_per_second": 1.337, "eval_wer": 0.047473773117272126, "step": 457500 }, { "epoch": 37.88, "learning_rate": 1.0711970074812969e-06, "loss": 0.0171, "step": 457600 }, { "epoch": 37.89, "learning_rate": 1.067040731504572e-06, "loss": 0.0125, "step": 457700 }, { "epoch": 37.9, "learning_rate": 1.062884455527847e-06, "loss": 0.0146, "step": 457800 }, { "epoch": 37.91, "learning_rate": 1.0587281795511224e-06, "loss": 0.0155, "step": 457900 }, { "epoch": 37.91, "learning_rate": 1.0545719035743975e-06, "loss": 0.0154, "step": 458000 }, { "epoch": 37.91, "eval_loss": 0.1928563416004181, "eval_runtime": 4708.9258, "eval_samples_per_second": 16.054, "eval_steps_per_second": 1.338, "eval_wer": 0.04786915712437365, "step": 458000 }, { "epoch": 37.92, "learning_rate": 1.0504156275976726e-06, "loss": 0.0127, "step": 458100 }, { "epoch": 37.93, "learning_rate": 1.0462593516209477e-06, "loss": 0.0136, "step": 458200 }, { "epoch": 37.94, "learning_rate": 1.0421030756442227e-06, "loss": 0.0163, "step": 458300 }, { "epoch": 37.95, "learning_rate": 1.037946799667498e-06, "loss": 0.0177, "step": 458400 }, { "epoch": 37.96, "learning_rate": 1.0337905236907731e-06, "loss": 0.0119, "step": 458500 }, { "epoch": 37.96, "eval_loss": 0.191275492310524, "eval_runtime": 4815.3796, "eval_samples_per_second": 15.699, "eval_steps_per_second": 1.308, "eval_wer": 0.047438388556325296, "step": 458500 }, { "epoch": 37.96, "learning_rate": 1.0296342477140484e-06, "loss": 0.0167, "step": 458600 }, { "epoch": 37.97, "learning_rate": 1.0254779717373233e-06, "loss": 0.0156, "step": 458700 }, { "epoch": 37.98, "learning_rate": 1.0213216957605986e-06, "loss": 0.0163, "step": 458800 }, { "epoch": 37.99, "learning_rate": 1.0171654197838737e-06, "loss": 0.0138, "step": 458900 }, { "epoch": 38.0, "learning_rate": 1.0130091438071488e-06, "loss": 0.0141, "step": 459000 }, { "epoch": 38.0, "eval_loss": 0.18814916908740997, "eval_runtime": 4681.0272, "eval_samples_per_second": 16.149, "eval_steps_per_second": 1.346, "eval_wer": 0.04732761949597001, "step": 459000 }, { "epoch": 38.0, "learning_rate": 1.0088528678304241e-06, "loss": 0.0169, "step": 459100 }, { "epoch": 38.01, "learning_rate": 1.0046965918536992e-06, "loss": 0.0121, "step": 459200 }, { "epoch": 38.02, "learning_rate": 1.0005403158769743e-06, "loss": 0.0141, "step": 459300 }, { "epoch": 38.03, "learning_rate": 9.963840399002494e-07, "loss": 0.0152, "step": 459400 }, { "epoch": 38.04, "learning_rate": 9.922277639235247e-07, "loss": 0.0135, "step": 459500 }, { "epoch": 38.04, "eval_loss": 0.1907120943069458, "eval_runtime": 4770.4215, "eval_samples_per_second": 15.847, "eval_steps_per_second": 1.321, "eval_wer": 0.04719223508886909, "step": 459500 }, { "epoch": 38.05, "learning_rate": 9.880714879467998e-07, "loss": 0.0136, "step": 459600 }, { "epoch": 38.05, "learning_rate": 9.839152119700749e-07, "loss": 0.0121, "step": 459700 }, { "epoch": 38.06, "learning_rate": 9.7975893599335e-07, "loss": 0.0152, "step": 459800 }, { "epoch": 38.07, "learning_rate": 9.75602660016625e-07, "loss": 0.0127, "step": 459900 }, { "epoch": 38.08, "learning_rate": 9.714463840399004e-07, "loss": 0.014, "step": 460000 }, { "epoch": 38.08, "eval_loss": 0.19126340746879578, "eval_runtime": 4820.1135, "eval_samples_per_second": 15.683, "eval_steps_per_second": 1.307, "eval_wer": 0.04759684985100023, "step": 460000 }, { "epoch": 38.09, "learning_rate": 9.672901080631755e-07, "loss": 0.0106, "step": 460100 }, { "epoch": 38.1, "learning_rate": 9.631338320864506e-07, "loss": 0.0144, "step": 460200 }, { "epoch": 38.1, "learning_rate": 9.589775561097257e-07, "loss": 0.0161, "step": 460300 }, { "epoch": 38.11, "learning_rate": 9.54821280133001e-07, "loss": 0.0137, "step": 460400 }, { "epoch": 38.12, "learning_rate": 9.506650041562761e-07, "loss": 0.0146, "step": 460500 }, { "epoch": 38.12, "eval_loss": 0.19125477969646454, "eval_runtime": 4754.7807, "eval_samples_per_second": 15.899, "eval_steps_per_second": 1.325, "eval_wer": 0.047638388248633465, "step": 460500 }, { "epoch": 38.13, "learning_rate": 9.465087281795513e-07, "loss": 0.0131, "step": 460600 }, { "epoch": 38.14, "learning_rate": 9.423524522028262e-07, "loss": 0.0149, "step": 460700 }, { "epoch": 38.15, "learning_rate": 9.381961762261014e-07, "loss": 0.0149, "step": 460800 }, { "epoch": 38.15, "learning_rate": 9.340399002493766e-07, "loss": 0.0145, "step": 460900 }, { "epoch": 38.16, "learning_rate": 9.298836242726518e-07, "loss": 0.0187, "step": 461000 }, { "epoch": 38.16, "eval_loss": 0.19156520068645477, "eval_runtime": 4731.3205, "eval_samples_per_second": 15.978, "eval_steps_per_second": 1.332, "eval_wer": 0.0474430039338401, "step": 461000 }, { "epoch": 38.17, "learning_rate": 9.257273482959269e-07, "loss": 0.0136, "step": 461100 }, { "epoch": 38.18, "learning_rate": 9.21571072319202e-07, "loss": 0.0173, "step": 461200 }, { "epoch": 38.19, "learning_rate": 9.174147963424772e-07, "loss": 0.0133, "step": 461300 }, { "epoch": 38.2, "learning_rate": 9.132585203657523e-07, "loss": 0.0102, "step": 461400 }, { "epoch": 38.2, "learning_rate": 9.091022443890275e-07, "loss": 0.0142, "step": 461500 }, { "epoch": 38.2, "eval_loss": 0.1934869885444641, "eval_runtime": 4751.4709, "eval_samples_per_second": 15.91, "eval_steps_per_second": 1.326, "eval_wer": 0.047509157678218956, "step": 461500 }, { "epoch": 38.21, "learning_rate": 9.049459684123027e-07, "loss": 0.0155, "step": 461600 }, { "epoch": 38.22, "learning_rate": 9.007896924355777e-07, "loss": 0.0164, "step": 461700 }, { "epoch": 38.23, "learning_rate": 8.966334164588529e-07, "loss": 0.0124, "step": 461800 }, { "epoch": 38.24, "learning_rate": 8.925187032418953e-07, "loss": 0.0146, "step": 461900 }, { "epoch": 38.25, "learning_rate": 8.884039900249377e-07, "loss": 0.0144, "step": 462000 }, { "epoch": 38.25, "eval_loss": 0.19136488437652588, "eval_runtime": 4732.7081, "eval_samples_per_second": 15.973, "eval_steps_per_second": 1.331, "eval_wer": 0.047304542608395984, "step": 462000 }, { "epoch": 38.25, "learning_rate": 8.842477140482129e-07, "loss": 0.0163, "step": 462100 }, { "epoch": 38.26, "learning_rate": 8.80091438071488e-07, "loss": 0.015, "step": 462200 }, { "epoch": 38.27, "learning_rate": 8.759351620947631e-07, "loss": 0.0177, "step": 462300 }, { "epoch": 38.28, "learning_rate": 8.717788861180383e-07, "loss": 0.0132, "step": 462400 }, { "epoch": 38.29, "learning_rate": 8.676226101413135e-07, "loss": 0.0138, "step": 462500 }, { "epoch": 38.29, "eval_loss": 0.19278642535209656, "eval_runtime": 4836.9634, "eval_samples_per_second": 15.629, "eval_steps_per_second": 1.302, "eval_wer": 0.04750454230070415, "step": 462500 }, { "epoch": 38.29, "learning_rate": 8.634663341645887e-07, "loss": 0.0177, "step": 462600 }, { "epoch": 38.3, "learning_rate": 8.593100581878637e-07, "loss": 0.0164, "step": 462700 }, { "epoch": 38.31, "learning_rate": 8.551537822111389e-07, "loss": 0.0133, "step": 462800 }, { "epoch": 38.32, "learning_rate": 8.509975062344141e-07, "loss": 0.0152, "step": 462900 }, { "epoch": 38.33, "learning_rate": 8.468412302576892e-07, "loss": 0.0131, "step": 463000 }, { "epoch": 38.33, "eval_loss": 0.19196464121341705, "eval_runtime": 4750.0313, "eval_samples_per_second": 15.915, "eval_steps_per_second": 1.326, "eval_wer": 0.047273773424963965, "step": 463000 }, { "epoch": 38.34, "learning_rate": 8.426849542809644e-07, "loss": 0.0136, "step": 463100 }, { "epoch": 38.34, "learning_rate": 8.385286783042395e-07, "loss": 0.014, "step": 463200 }, { "epoch": 38.35, "learning_rate": 8.343724023275146e-07, "loss": 0.0152, "step": 463300 }, { "epoch": 38.36, "learning_rate": 8.302161263507898e-07, "loss": 0.015, "step": 463400 }, { "epoch": 38.37, "learning_rate": 8.26059850374065e-07, "loss": 0.013, "step": 463500 }, { "epoch": 38.37, "eval_loss": 0.1904260665178299, "eval_runtime": 4750.869, "eval_samples_per_second": 15.912, "eval_steps_per_second": 1.326, "eval_wer": 0.04694146624389808, "step": 463500 }, { "epoch": 38.38, "learning_rate": 8.219035743973402e-07, "loss": 0.0124, "step": 463600 }, { "epoch": 38.39, "learning_rate": 8.177472984206151e-07, "loss": 0.0125, "step": 463700 }, { "epoch": 38.39, "learning_rate": 8.135910224438903e-07, "loss": 0.0188, "step": 463800 }, { "epoch": 38.4, "learning_rate": 8.094347464671655e-07, "loss": 0.0128, "step": 463900 }, { "epoch": 38.41, "learning_rate": 8.052784704904406e-07, "loss": 0.0139, "step": 464000 }, { "epoch": 38.41, "eval_loss": 0.19182735681533813, "eval_runtime": 4960.4811, "eval_samples_per_second": 15.239, "eval_steps_per_second": 1.27, "eval_wer": 0.04736300405691683, "step": 464000 }, { "epoch": 38.42, "learning_rate": 8.011221945137158e-07, "loss": 0.0156, "step": 464100 }, { "epoch": 38.43, "learning_rate": 7.969659185369909e-07, "loss": 0.0111, "step": 464200 }, { "epoch": 38.44, "learning_rate": 7.92809642560266e-07, "loss": 0.0158, "step": 464300 }, { "epoch": 38.44, "learning_rate": 7.886533665835412e-07, "loss": 0.0166, "step": 464400 }, { "epoch": 38.45, "learning_rate": 7.844970906068164e-07, "loss": 0.0135, "step": 464500 }, { "epoch": 38.45, "eval_loss": 0.19227033853530884, "eval_runtime": 4774.3861, "eval_samples_per_second": 15.833, "eval_steps_per_second": 1.32, "eval_wer": 0.04716454282378027, "step": 464500 }, { "epoch": 38.46, "learning_rate": 7.803408146300916e-07, "loss": 0.0176, "step": 464600 }, { "epoch": 38.47, "learning_rate": 7.761845386533666e-07, "loss": 0.0162, "step": 464700 }, { "epoch": 38.48, "learning_rate": 7.720282626766418e-07, "loss": 0.012, "step": 464800 }, { "epoch": 38.49, "learning_rate": 7.67871986699917e-07, "loss": 0.0128, "step": 464900 }, { "epoch": 38.49, "learning_rate": 7.637157107231921e-07, "loss": 0.0149, "step": 465000 }, { "epoch": 38.49, "eval_loss": 0.1919548064470291, "eval_runtime": 4726.898, "eval_samples_per_second": 15.993, "eval_steps_per_second": 1.333, "eval_wer": 0.04709377370188661, "step": 465000 }, { "epoch": 38.5, "learning_rate": 7.595594347464673e-07, "loss": 0.0139, "step": 465100 }, { "epoch": 38.51, "learning_rate": 7.554031587697423e-07, "loss": 0.0144, "step": 465200 }, { "epoch": 38.52, "learning_rate": 7.512468827930175e-07, "loss": 0.0142, "step": 465300 }, { "epoch": 38.53, "learning_rate": 7.470906068162927e-07, "loss": 0.0151, "step": 465400 }, { "epoch": 38.53, "learning_rate": 7.429343308395679e-07, "loss": 0.0133, "step": 465500 }, { "epoch": 38.53, "eval_loss": 0.19052021205425262, "eval_runtime": 5468.9967, "eval_samples_per_second": 13.822, "eval_steps_per_second": 1.152, "eval_wer": 0.04709223524271501, "step": 465500 }, { "epoch": 38.54, "learning_rate": 7.387780548628431e-07, "loss": 0.0201, "step": 465600 }, { "epoch": 38.55, "learning_rate": 7.34621778886118e-07, "loss": 0.0135, "step": 465700 }, { "epoch": 38.56, "learning_rate": 7.304655029093933e-07, "loss": 0.0137, "step": 465800 }, { "epoch": 38.57, "learning_rate": 7.263092269326683e-07, "loss": 0.0115, "step": 465900 }, { "epoch": 38.58, "learning_rate": 7.221529509559435e-07, "loss": 0.0147, "step": 466000 }, { "epoch": 38.58, "eval_loss": 0.1911328136920929, "eval_runtime": 4932.9282, "eval_samples_per_second": 15.325, "eval_steps_per_second": 1.277, "eval_wer": 0.04723531194567393, "step": 466000 }, { "epoch": 38.58, "learning_rate": 7.179966749792187e-07, "loss": 0.0127, "step": 466100 }, { "epoch": 38.59, "learning_rate": 7.138403990024937e-07, "loss": 0.0122, "step": 466200 }, { "epoch": 38.6, "learning_rate": 7.096841230257689e-07, "loss": 0.0196, "step": 466300 }, { "epoch": 38.61, "learning_rate": 7.055278470490441e-07, "loss": 0.0148, "step": 466400 }, { "epoch": 38.62, "learning_rate": 7.013715710723193e-07, "loss": 0.0161, "step": 466500 }, { "epoch": 38.62, "eval_loss": 0.19189859926700592, "eval_runtime": 4709.7544, "eval_samples_per_second": 16.051, "eval_steps_per_second": 1.338, "eval_wer": 0.047353773301887225, "step": 466500 }, { "epoch": 38.63, "learning_rate": 6.972152950955944e-07, "loss": 0.0177, "step": 466600 }, { "epoch": 38.63, "learning_rate": 6.931005818786368e-07, "loss": 0.0149, "step": 466700 }, { "epoch": 38.64, "learning_rate": 6.88944305901912e-07, "loss": 0.0118, "step": 466800 }, { "epoch": 38.65, "learning_rate": 6.847880299251871e-07, "loss": 0.0158, "step": 466900 }, { "epoch": 38.66, "learning_rate": 6.806317539484622e-07, "loss": 0.0174, "step": 467000 }, { "epoch": 38.66, "eval_loss": 0.19175074994564056, "eval_runtime": 5377.2761, "eval_samples_per_second": 14.058, "eval_steps_per_second": 1.172, "eval_wer": 0.0472060812214135, "step": 467000 }, { "epoch": 38.67, "learning_rate": 6.764754779717374e-07, "loss": 0.0178, "step": 467100 }, { "epoch": 38.68, "learning_rate": 6.723192019950126e-07, "loss": 0.0122, "step": 467200 }, { "epoch": 38.68, "learning_rate": 6.682044887780549e-07, "loss": 0.0131, "step": 467300 }, { "epoch": 38.69, "learning_rate": 6.640482128013301e-07, "loss": 0.0118, "step": 467400 }, { "epoch": 38.7, "learning_rate": 6.598919368246052e-07, "loss": 0.0123, "step": 467500 }, { "epoch": 38.7, "eval_loss": 0.19163398444652557, "eval_runtime": 4655.9798, "eval_samples_per_second": 16.236, "eval_steps_per_second": 1.353, "eval_wer": 0.047316850281768796, "step": 467500 }, { "epoch": 38.71, "learning_rate": 6.557356608478803e-07, "loss": 0.0158, "step": 467600 }, { "epoch": 38.72, "learning_rate": 6.515793848711555e-07, "loss": 0.0128, "step": 467700 }, { "epoch": 38.73, "learning_rate": 6.474231088944306e-07, "loss": 0.0156, "step": 467800 }, { "epoch": 38.73, "learning_rate": 6.432668329177058e-07, "loss": 0.0154, "step": 467900 }, { "epoch": 38.74, "learning_rate": 6.39110556940981e-07, "loss": 0.0143, "step": 468000 }, { "epoch": 38.74, "eval_loss": 0.19149768352508545, "eval_runtime": 4642.0004, "eval_samples_per_second": 16.285, "eval_steps_per_second": 1.357, "eval_wer": 0.04699992769241893, "step": 468000 }, { "epoch": 38.75, "learning_rate": 6.34954280964256e-07, "loss": 0.0124, "step": 468100 }, { "epoch": 38.76, "learning_rate": 6.307980049875312e-07, "loss": 0.022, "step": 468200 }, { "epoch": 38.77, "learning_rate": 6.266417290108064e-07, "loss": 0.0122, "step": 468300 }, { "epoch": 38.77, "learning_rate": 6.224854530340816e-07, "loss": 0.0132, "step": 468400 }, { "epoch": 38.78, "learning_rate": 6.183291770573567e-07, "loss": 0.0112, "step": 468500 }, { "epoch": 38.78, "eval_loss": 0.19026269018650055, "eval_runtime": 4729.4876, "eval_samples_per_second": 15.984, "eval_steps_per_second": 1.332, "eval_wer": 0.046904543223779654, "step": 468500 }, { "epoch": 38.79, "learning_rate": 6.141729010806319e-07, "loss": 0.0128, "step": 468600 }, { "epoch": 38.8, "learning_rate": 6.10016625103907e-07, "loss": 0.0139, "step": 468700 }, { "epoch": 38.81, "learning_rate": 6.05860349127182e-07, "loss": 0.013, "step": 468800 }, { "epoch": 38.82, "learning_rate": 6.017040731504572e-07, "loss": 0.0145, "step": 468900 }, { "epoch": 38.82, "learning_rate": 5.975477971737323e-07, "loss": 0.0126, "step": 469000 }, { "epoch": 38.82, "eval_loss": 0.19231946766376495, "eval_runtime": 4783.4587, "eval_samples_per_second": 15.803, "eval_steps_per_second": 1.317, "eval_wer": 0.0469630046723005, "step": 469000 }, { "epoch": 38.83, "learning_rate": 5.933915211970075e-07, "loss": 0.013, "step": 469100 }, { "epoch": 38.84, "learning_rate": 5.892352452202827e-07, "loss": 0.0162, "step": 469200 }, { "epoch": 38.85, "learning_rate": 5.850789692435578e-07, "loss": 0.0165, "step": 469300 }, { "epoch": 38.86, "learning_rate": 5.80922693266833e-07, "loss": 0.0145, "step": 469400 }, { "epoch": 38.87, "learning_rate": 5.767664172901081e-07, "loss": 0.0138, "step": 469500 }, { "epoch": 38.87, "eval_loss": 0.19286265969276428, "eval_runtime": 4827.9942, "eval_samples_per_second": 15.658, "eval_steps_per_second": 1.305, "eval_wer": 0.04713992747703465, "step": 469500 }, { "epoch": 38.87, "learning_rate": 5.726101413133832e-07, "loss": 0.0143, "step": 469600 }, { "epoch": 38.88, "learning_rate": 5.684538653366584e-07, "loss": 0.0119, "step": 469700 }, { "epoch": 38.89, "learning_rate": 5.642975893599335e-07, "loss": 0.0132, "step": 469800 }, { "epoch": 38.9, "learning_rate": 5.601413133832087e-07, "loss": 0.016, "step": 469900 }, { "epoch": 38.91, "learning_rate": 5.559850374064838e-07, "loss": 0.014, "step": 470000 }, { "epoch": 38.91, "eval_loss": 0.19285814464092255, "eval_runtime": 5656.6425, "eval_samples_per_second": 13.364, "eval_steps_per_second": 1.114, "eval_wer": 0.04719223508886909, "step": 470000 }, { "epoch": 38.92, "learning_rate": 5.51828761429759e-07, "loss": 0.0181, "step": 470100 }, { "epoch": 38.92, "learning_rate": 5.476724854530342e-07, "loss": 0.0145, "step": 470200 }, { "epoch": 38.93, "learning_rate": 5.435577722360765e-07, "loss": 0.012, "step": 470300 }, { "epoch": 38.94, "learning_rate": 5.394014962593516e-07, "loss": 0.012, "step": 470400 }, { "epoch": 38.95, "learning_rate": 5.352452202826268e-07, "loss": 0.0152, "step": 470500 }, { "epoch": 38.95, "eval_loss": 0.19387485086917877, "eval_runtime": 5527.2311, "eval_samples_per_second": 13.677, "eval_steps_per_second": 1.14, "eval_wer": 0.04711992750780383, "step": 470500 }, { "epoch": 38.96, "learning_rate": 5.310889443059019e-07, "loss": 0.0134, "step": 470600 }, { "epoch": 38.97, "learning_rate": 5.269742310889443e-07, "loss": 0.0125, "step": 470700 }, { "epoch": 38.97, "learning_rate": 5.228179551122195e-07, "loss": 0.0127, "step": 470800 }, { "epoch": 38.98, "learning_rate": 5.186616791354947e-07, "loss": 0.0141, "step": 470900 }, { "epoch": 38.99, "learning_rate": 5.145054031587698e-07, "loss": 0.0124, "step": 471000 }, { "epoch": 38.99, "eval_loss": 0.19433453679084778, "eval_runtime": 6020.9311, "eval_samples_per_second": 12.555, "eval_steps_per_second": 1.046, "eval_wer": 0.047096850620229815, "step": 471000 }, { "epoch": 39.0, "learning_rate": 5.10349127182045e-07, "loss": 0.0144, "step": 471100 }, { "epoch": 39.01, "learning_rate": 5.061928512053201e-07, "loss": 0.0143, "step": 471200 }, { "epoch": 39.01, "learning_rate": 5.020365752285953e-07, "loss": 0.0124, "step": 471300 }, { "epoch": 39.02, "learning_rate": 4.978802992518704e-07, "loss": 0.0146, "step": 471400 }, { "epoch": 39.03, "learning_rate": 4.937240232751455e-07, "loss": 0.0103, "step": 471500 }, { "epoch": 39.03, "eval_loss": 0.19349665939807892, "eval_runtime": 5420.7564, "eval_samples_per_second": 13.945, "eval_steps_per_second": 1.162, "eval_wer": 0.047032235335022564, "step": 471500 }, { "epoch": 39.04, "learning_rate": 4.895677472984207e-07, "loss": 0.0097, "step": 471600 }, { "epoch": 39.05, "learning_rate": 4.854114713216957e-07, "loss": 0.0136, "step": 471700 }, { "epoch": 39.06, "learning_rate": 4.812551953449709e-07, "loss": 0.0149, "step": 471800 }, { "epoch": 39.06, "learning_rate": 4.770989193682461e-07, "loss": 0.0157, "step": 471900 }, { "epoch": 39.07, "learning_rate": 4.7294264339152124e-07, "loss": 0.0143, "step": 472000 }, { "epoch": 39.07, "eval_loss": 0.19397491216659546, "eval_runtime": 5221.0408, "eval_samples_per_second": 14.479, "eval_steps_per_second": 1.207, "eval_wer": 0.046996850774075734, "step": 472000 }, { "epoch": 39.08, "learning_rate": 4.687863674147964e-07, "loss": 0.0127, "step": 472100 }, { "epoch": 39.09, "learning_rate": 4.646300914380715e-07, "loss": 0.0182, "step": 472200 }, { "epoch": 39.1, "learning_rate": 4.604738154613467e-07, "loss": 0.0165, "step": 472300 }, { "epoch": 39.11, "learning_rate": 4.5631753948462187e-07, "loss": 0.0167, "step": 472400 }, { "epoch": 39.11, "learning_rate": 4.5216126350789697e-07, "loss": 0.0174, "step": 472500 }, { "epoch": 39.11, "eval_loss": 0.19229955971240997, "eval_runtime": 5950.6691, "eval_samples_per_second": 12.704, "eval_steps_per_second": 1.059, "eval_wer": 0.047083004487685404, "step": 472500 }, { "epoch": 39.12, "learning_rate": 4.480049875311721e-07, "loss": 0.0153, "step": 472600 }, { "epoch": 39.13, "learning_rate": 4.438487115544472e-07, "loss": 0.0161, "step": 472700 }, { "epoch": 39.14, "learning_rate": 4.396924355777224e-07, "loss": 0.0136, "step": 472800 }, { "epoch": 39.15, "learning_rate": 4.3553615960099755e-07, "loss": 0.0159, "step": 472900 }, { "epoch": 39.16, "learning_rate": 4.313798836242727e-07, "loss": 0.0152, "step": 473000 }, { "epoch": 39.16, "eval_loss": 0.19175788760185242, "eval_runtime": 5973.5802, "eval_samples_per_second": 12.655, "eval_steps_per_second": 1.055, "eval_wer": 0.047189158170525894, "step": 473000 }, { "epoch": 39.16, "learning_rate": 4.2722360764754784e-07, "loss": 0.0145, "step": 473100 }, { "epoch": 39.17, "learning_rate": 4.2306733167082293e-07, "loss": 0.0124, "step": 473200 }, { "epoch": 39.18, "learning_rate": 4.1891105569409813e-07, "loss": 0.0149, "step": 473300 }, { "epoch": 39.19, "learning_rate": 4.147547797173733e-07, "loss": 0.0153, "step": 473400 }, { "epoch": 39.2, "learning_rate": 4.105985037406484e-07, "loss": 0.0153, "step": 473500 }, { "epoch": 39.2, "eval_loss": 0.19086137413978577, "eval_runtime": 6024.8325, "eval_samples_per_second": 12.547, "eval_steps_per_second": 1.046, "eval_wer": 0.046996850774075734, "step": 473500 }, { "epoch": 39.21, "learning_rate": 4.0644222776392357e-07, "loss": 0.0164, "step": 473600 }, { "epoch": 39.21, "learning_rate": 4.0228595178719866e-07, "loss": 0.013, "step": 473700 }, { "epoch": 39.22, "learning_rate": 3.9812967581047386e-07, "loss": 0.0132, "step": 473800 }, { "epoch": 39.23, "learning_rate": 3.93973399833749e-07, "loss": 0.0141, "step": 473900 }, { "epoch": 39.24, "learning_rate": 3.8981712385702415e-07, "loss": 0.0161, "step": 474000 }, { "epoch": 39.24, "eval_loss": 0.19134704768657684, "eval_runtime": 5971.9817, "eval_samples_per_second": 12.658, "eval_steps_per_second": 1.055, "eval_wer": 0.046976850804844913, "step": 474000 }, { "epoch": 39.25, "learning_rate": 3.856608478802993e-07, "loss": 0.0146, "step": 474100 }, { "epoch": 39.25, "learning_rate": 3.815045719035744e-07, "loss": 0.0181, "step": 474200 }, { "epoch": 39.26, "learning_rate": 3.773482959268496e-07, "loss": 0.0136, "step": 474300 }, { "epoch": 39.27, "learning_rate": 3.7319201995012473e-07, "loss": 0.0132, "step": 474400 }, { "epoch": 39.28, "learning_rate": 3.690357439733998e-07, "loss": 0.0133, "step": 474500 }, { "epoch": 39.28, "eval_loss": 0.19125811755657196, "eval_runtime": 4536.4067, "eval_samples_per_second": 16.664, "eval_steps_per_second": 1.389, "eval_wer": 0.046959927753957305, "step": 474500 }, { "epoch": 39.29, "learning_rate": 3.64879467996675e-07, "loss": 0.018, "step": 474600 }, { "epoch": 39.3, "learning_rate": 3.607231920199501e-07, "loss": 0.0164, "step": 474700 }, { "epoch": 39.3, "learning_rate": 3.565669160432253e-07, "loss": 0.0129, "step": 474800 }, { "epoch": 39.31, "learning_rate": 3.5241064006650046e-07, "loss": 0.014, "step": 474900 }, { "epoch": 39.32, "learning_rate": 3.4825436408977555e-07, "loss": 0.0126, "step": 475000 }, { "epoch": 39.32, "eval_loss": 0.19118857383728027, "eval_runtime": 4763.258, "eval_samples_per_second": 15.87, "eval_steps_per_second": 1.323, "eval_wer": 0.04682300488768479, "step": 475000 }, { "epoch": 39.33, "learning_rate": 3.4409808811305075e-07, "loss": 0.0143, "step": 475100 }, { "epoch": 39.34, "learning_rate": 3.3994181213632584e-07, "loss": 0.0136, "step": 475200 }, { "epoch": 39.35, "learning_rate": 3.3578553615960104e-07, "loss": 0.0135, "step": 475300 }, { "epoch": 39.35, "learning_rate": 3.316292601828762e-07, "loss": 0.0154, "step": 475400 }, { "epoch": 39.36, "learning_rate": 3.274729842061513e-07, "loss": 0.0162, "step": 475500 }, { "epoch": 39.36, "eval_loss": 0.19141505658626556, "eval_runtime": 4695.1421, "eval_samples_per_second": 16.101, "eval_steps_per_second": 1.342, "eval_wer": 0.0467091589089863, "step": 475500 }, { "epoch": 39.37, "learning_rate": 3.233167082294265e-07, "loss": 0.0153, "step": 475600 }, { "epoch": 39.38, "learning_rate": 3.1916043225270157e-07, "loss": 0.0167, "step": 475700 }, { "epoch": 39.39, "learning_rate": 3.1500415627597677e-07, "loss": 0.0125, "step": 475800 }, { "epoch": 39.4, "learning_rate": 3.1084788029925186e-07, "loss": 0.0139, "step": 475900 }, { "epoch": 39.4, "learning_rate": 3.0669160432252706e-07, "loss": 0.0134, "step": 476000 }, { "epoch": 39.4, "eval_loss": 0.19058671593666077, "eval_runtime": 4828.3655, "eval_samples_per_second": 15.656, "eval_steps_per_second": 1.305, "eval_wer": 0.046690697398927075, "step": 476000 }, { "epoch": 39.41, "learning_rate": 3.025353283458022e-07, "loss": 0.0171, "step": 476100 }, { "epoch": 39.42, "learning_rate": 2.9837905236907735e-07, "loss": 0.0138, "step": 476200 }, { "epoch": 39.43, "learning_rate": 2.9422277639235245e-07, "loss": 0.0135, "step": 476300 }, { "epoch": 39.44, "learning_rate": 2.900665004156276e-07, "loss": 0.0178, "step": 476400 }, { "epoch": 39.45, "learning_rate": 2.859102244389028e-07, "loss": 0.013, "step": 476500 }, { "epoch": 39.45, "eval_loss": 0.19052618741989136, "eval_runtime": 5318.3793, "eval_samples_per_second": 14.214, "eval_steps_per_second": 1.185, "eval_wer": 0.04675069730661953, "step": 476500 }, { "epoch": 39.45, "learning_rate": 2.8175394846217794e-07, "loss": 0.0132, "step": 476600 }, { "epoch": 39.46, "learning_rate": 2.775976724854531e-07, "loss": 0.0174, "step": 476700 }, { "epoch": 39.47, "learning_rate": 2.734413965087282e-07, "loss": 0.0147, "step": 476800 }, { "epoch": 39.48, "learning_rate": 2.692851205320033e-07, "loss": 0.0137, "step": 476900 }, { "epoch": 39.49, "learning_rate": 2.651288445552785e-07, "loss": 0.016, "step": 477000 }, { "epoch": 39.49, "eval_loss": 0.191084623336792, "eval_runtime": 5102.8067, "eval_samples_per_second": 14.814, "eval_steps_per_second": 1.235, "eval_wer": 0.046795312622595966, "step": 477000 }, { "epoch": 39.5, "learning_rate": 2.6097256857855366e-07, "loss": 0.014, "step": 477100 }, { "epoch": 39.5, "learning_rate": 2.568162926018288e-07, "loss": 0.0116, "step": 477200 }, { "epoch": 39.51, "learning_rate": 2.526600166251039e-07, "loss": 0.0125, "step": 477300 }, { "epoch": 39.52, "learning_rate": 2.4850374064837905e-07, "loss": 0.0135, "step": 477400 }, { "epoch": 39.53, "learning_rate": 2.4438902743142146e-07, "loss": 0.0149, "step": 477500 }, { "epoch": 39.53, "eval_loss": 0.19123640656471252, "eval_runtime": 4908.9194, "eval_samples_per_second": 15.4, "eval_steps_per_second": 1.283, "eval_wer": 0.046690697398927075, "step": 477500 }, { "epoch": 39.54, "learning_rate": 2.402327514546966e-07, "loss": 0.0166, "step": 477600 }, { "epoch": 39.54, "learning_rate": 2.3607647547797178e-07, "loss": 0.0151, "step": 477700 }, { "epoch": 39.55, "learning_rate": 2.319201995012469e-07, "loss": 0.0125, "step": 477800 }, { "epoch": 39.56, "learning_rate": 2.2776392352452204e-07, "loss": 0.0113, "step": 477900 }, { "epoch": 39.57, "learning_rate": 2.236076475477972e-07, "loss": 0.0132, "step": 478000 }, { "epoch": 39.57, "eval_loss": 0.1917223334312439, "eval_runtime": 4939.0518, "eval_samples_per_second": 15.306, "eval_steps_per_second": 1.276, "eval_wer": 0.04672915887821711, "step": 478000 }, { "epoch": 39.58, "learning_rate": 2.1945137157107236e-07, "loss": 0.0146, "step": 478100 }, { "epoch": 39.59, "learning_rate": 2.1529509559434748e-07, "loss": 0.0155, "step": 478200 }, { "epoch": 39.59, "learning_rate": 2.1113881961762262e-07, "loss": 0.0137, "step": 478300 }, { "epoch": 39.6, "learning_rate": 2.0698254364089777e-07, "loss": 0.0137, "step": 478400 }, { "epoch": 39.61, "learning_rate": 2.0282626766417292e-07, "loss": 0.0289, "step": 478500 }, { "epoch": 39.61, "eval_loss": 0.19159676134586334, "eval_runtime": 5736.0382, "eval_samples_per_second": 13.179, "eval_steps_per_second": 1.098, "eval_wer": 0.04665531283798025, "step": 478500 }, { "epoch": 39.62, "learning_rate": 1.986699916874481e-07, "loss": 0.0139, "step": 478600 }, { "epoch": 39.63, "learning_rate": 1.945137157107232e-07, "loss": 0.0107, "step": 478700 }, { "epoch": 39.64, "learning_rate": 1.9035743973399835e-07, "loss": 0.0143, "step": 478800 }, { "epoch": 39.64, "learning_rate": 1.862011637572735e-07, "loss": 0.0153, "step": 478900 }, { "epoch": 39.65, "learning_rate": 1.8204488778054862e-07, "loss": 0.0129, "step": 479000 }, { "epoch": 39.65, "eval_loss": 0.19163928925991058, "eval_runtime": 5947.7386, "eval_samples_per_second": 12.71, "eval_steps_per_second": 1.059, "eval_wer": 0.04675838960247754, "step": 479000 }, { "epoch": 39.66, "learning_rate": 1.778886118038238e-07, "loss": 0.0169, "step": 479100 }, { "epoch": 39.67, "learning_rate": 1.7373233582709893e-07, "loss": 0.0184, "step": 479200 }, { "epoch": 39.68, "learning_rate": 1.6957605985037408e-07, "loss": 0.0128, "step": 479300 }, { "epoch": 39.69, "learning_rate": 1.6541978387364923e-07, "loss": 0.0143, "step": 479400 }, { "epoch": 39.69, "learning_rate": 1.6126350789692434e-07, "loss": 0.0122, "step": 479500 }, { "epoch": 39.69, "eval_loss": 0.1916171908378601, "eval_runtime": 5936.5531, "eval_samples_per_second": 12.734, "eval_steps_per_second": 1.061, "eval_wer": 0.046712235827329494, "step": 479500 }, { "epoch": 39.7, "learning_rate": 1.5710723192019952e-07, "loss": 0.015, "step": 479600 }, { "epoch": 39.71, "learning_rate": 1.5295095594347466e-07, "loss": 0.0131, "step": 479700 }, { "epoch": 39.72, "learning_rate": 1.4883624272651705e-07, "loss": 0.0128, "step": 479800 }, { "epoch": 39.73, "learning_rate": 1.446799667497922e-07, "loss": 0.013, "step": 479900 }, { "epoch": 39.74, "learning_rate": 1.4052369077306734e-07, "loss": 0.0159, "step": 480000 }, { "epoch": 39.74, "eval_loss": 0.19114597141742706, "eval_runtime": 5443.284, "eval_samples_per_second": 13.888, "eval_steps_per_second": 1.157, "eval_wer": 0.046659928215495056, "step": 480000 }, { "epoch": 39.74, "learning_rate": 1.3636741479634248e-07, "loss": 0.0119, "step": 480100 }, { "epoch": 39.75, "learning_rate": 1.3221113881961763e-07, "loss": 0.0156, "step": 480200 }, { "epoch": 39.76, "learning_rate": 1.2805486284289278e-07, "loss": 0.0147, "step": 480300 }, { "epoch": 39.77, "learning_rate": 1.2389858686616792e-07, "loss": 0.0126, "step": 480400 }, { "epoch": 39.78, "learning_rate": 1.1974231088944307e-07, "loss": 0.0126, "step": 480500 }, { "epoch": 39.78, "eval_loss": 0.19150681793689728, "eval_runtime": 4665.1573, "eval_samples_per_second": 16.204, "eval_steps_per_second": 1.35, "eval_wer": 0.04679223570425276, "step": 480500 }, { "epoch": 39.78, "learning_rate": 1.1558603491271821e-07, "loss": 0.0161, "step": 480600 }, { "epoch": 39.79, "learning_rate": 1.1147132169576061e-07, "loss": 0.0127, "step": 480700 }, { "epoch": 39.8, "learning_rate": 1.0731504571903576e-07, "loss": 0.0125, "step": 480800 }, { "epoch": 39.81, "learning_rate": 1.031587697423109e-07, "loss": 0.0145, "step": 480900 }, { "epoch": 39.82, "learning_rate": 9.900249376558605e-08, "loss": 0.0164, "step": 481000 }, { "epoch": 39.82, "eval_loss": 0.19145222008228302, "eval_runtime": 4631.5526, "eval_samples_per_second": 16.322, "eval_steps_per_second": 1.36, "eval_wer": 0.0468430048569156, "step": 481000 }, { "epoch": 39.83, "learning_rate": 9.484621778886118e-08, "loss": 0.015, "step": 481100 }, { "epoch": 39.83, "learning_rate": 9.068994181213634e-08, "loss": 0.0121, "step": 481200 }, { "epoch": 39.84, "learning_rate": 8.653366583541147e-08, "loss": 0.0136, "step": 481300 }, { "epoch": 39.85, "learning_rate": 8.237738985868663e-08, "loss": 0.015, "step": 481400 }, { "epoch": 39.86, "learning_rate": 7.822111388196176e-08, "loss": 0.0139, "step": 481500 }, { "epoch": 39.86, "eval_loss": 0.19163356721401215, "eval_runtime": 4692.0213, "eval_samples_per_second": 16.111, "eval_steps_per_second": 1.343, "eval_wer": 0.04686454328531802, "step": 481500 }, { "epoch": 39.87, "learning_rate": 7.406483790523692e-08, "loss": 0.0112, "step": 481600 }, { "epoch": 39.88, "learning_rate": 6.990856192851207e-08, "loss": 0.014, "step": 481700 }, { "epoch": 39.88, "learning_rate": 6.57522859517872e-08, "loss": 0.0198, "step": 481800 }, { "epoch": 39.89, "learning_rate": 6.159600997506235e-08, "loss": 0.0124, "step": 481900 }, { "epoch": 39.9, "learning_rate": 5.743973399833749e-08, "loss": 0.0122, "step": 482000 }, { "epoch": 39.9, "eval_loss": 0.19193434715270996, "eval_runtime": 4634.6234, "eval_samples_per_second": 16.311, "eval_steps_per_second": 1.359, "eval_wer": 0.046875312499519234, "step": 482000 }, { "epoch": 39.91, "learning_rate": 5.328345802161264e-08, "loss": 0.0137, "step": 482100 }, { "epoch": 39.92, "learning_rate": 4.912718204488779e-08, "loss": 0.0136, "step": 482200 }, { "epoch": 39.93, "learning_rate": 4.4970906068162934e-08, "loss": 0.0128, "step": 482300 }, { "epoch": 39.93, "learning_rate": 4.081463009143807e-08, "loss": 0.0139, "step": 482400 }, { "epoch": 39.94, "learning_rate": 3.665835411471322e-08, "loss": 0.0154, "step": 482500 }, { "epoch": 39.94, "eval_loss": 0.1920659840106964, "eval_runtime": 4575.8559, "eval_samples_per_second": 16.52, "eval_steps_per_second": 1.377, "eval_wer": 0.04685838944863162, "step": 482500 }, { "epoch": 39.95, "learning_rate": 3.2502078137988364e-08, "loss": 0.0141, "step": 482600 }, { "epoch": 39.96, "learning_rate": 2.834580216126351e-08, "loss": 0.0158, "step": 482700 }, { "epoch": 39.97, "learning_rate": 2.4189526184538652e-08, "loss": 0.0136, "step": 482800 }, { "epoch": 39.98, "learning_rate": 2.00332502078138e-08, "loss": 0.0167, "step": 482900 }, { "epoch": 39.98, "learning_rate": 1.5876974231088944e-08, "loss": 0.0124, "step": 483000 }, { "epoch": 39.98, "eval_loss": 0.19205854833126068, "eval_runtime": 4790.8246, "eval_samples_per_second": 15.779, "eval_steps_per_second": 1.315, "eval_wer": 0.04687685095869083, "step": 483000 }, { "epoch": 39.99, "learning_rate": 1.172069825436409e-08, "loss": 0.0141, "step": 483100 }, { "epoch": 40.0, "learning_rate": 7.564422277639237e-09, "loss": 0.0134, "step": 483200 }, { "epoch": 40.0, "step": 483200, "total_flos": 5.645472564958761e+21, "train_loss": 0.0034612330223205466, "train_runtime": 1304716.1332, "train_samples_per_second": 8.888, "train_steps_per_second": 0.37 } ], "max_steps": 483200, "num_train_epochs": 40, "total_flos": 5.645472564958761e+21, "trial_name": null, "trial_params": null }