{ "best_metric": 17.449157344638696, "best_model_checkpoint": "./openai/whisper-small-gom-LDC-v1.0/checkpoint-11000", "epoch": 2.846790890269151, "global_step": 11000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.01, "learning_rate": 3.36e-07, "loss": 2.8574, "step": 25 }, { "epoch": 0.01, "learning_rate": 7.359999999999999e-07, "loss": 2.3034, "step": 50 }, { "epoch": 0.02, "learning_rate": 1.1359999999999998e-06, "loss": 1.5867, "step": 75 }, { "epoch": 0.03, "learning_rate": 1.536e-06, "loss": 1.0482, "step": 100 }, { "epoch": 0.03, "learning_rate": 1.9359999999999998e-06, "loss": 0.8649, "step": 125 }, { "epoch": 0.04, "learning_rate": 2.3359999999999997e-06, "loss": 0.7442, "step": 150 }, { "epoch": 0.05, "learning_rate": 2.736e-06, "loss": 0.7494, "step": 175 }, { "epoch": 0.05, "learning_rate": 3.136e-06, "loss": 0.6801, "step": 200 }, { "epoch": 0.06, "learning_rate": 3.5359999999999997e-06, "loss": 0.6158, "step": 225 }, { "epoch": 0.06, "learning_rate": 3.936e-06, "loss": 0.6206, "step": 250 }, { "epoch": 0.07, "learning_rate": 4.3360000000000005e-06, "loss": 0.5527, "step": 275 }, { "epoch": 0.08, "learning_rate": 4.735999999999999e-06, "loss": 0.5373, "step": 300 }, { "epoch": 0.08, "learning_rate": 5.136e-06, "loss": 0.4616, "step": 325 }, { "epoch": 0.09, "learning_rate": 5.535999999999999e-06, "loss": 0.4252, "step": 350 }, { "epoch": 0.1, "learning_rate": 5.936e-06, "loss": 0.3137, "step": 375 }, { "epoch": 0.1, "learning_rate": 6.336e-06, "loss": 0.2815, "step": 400 }, { "epoch": 0.11, "learning_rate": 6.7359999999999995e-06, "loss": 0.2738, "step": 425 }, { "epoch": 0.12, "learning_rate": 7.136e-06, "loss": 0.3058, "step": 450 }, { "epoch": 0.12, "learning_rate": 7.5359999999999995e-06, "loss": 0.2604, "step": 475 }, { "epoch": 0.13, "learning_rate": 7.936e-06, "loss": 0.2547, "step": 500 }, { "epoch": 0.13, "eval_loss": 0.2495306134223938, "eval_runtime": 4903.6311, "eval_samples_per_second": 2.224, "eval_steps_per_second": 0.278, "eval_wer": 61.019581505656184, "step": 500 }, { "epoch": 0.14, "learning_rate": 7.984853948791921e-06, "loss": 0.2135, "step": 525 }, { "epoch": 0.14, "learning_rate": 7.966822935448972e-06, "loss": 0.2514, "step": 550 }, { "epoch": 0.15, "learning_rate": 7.948791922106022e-06, "loss": 0.238, "step": 575 }, { "epoch": 0.16, "learning_rate": 7.930760908763071e-06, "loss": 0.197, "step": 600 }, { "epoch": 0.16, "learning_rate": 7.912729895420122e-06, "loss": 0.2452, "step": 625 }, { "epoch": 0.17, "learning_rate": 7.894698882077172e-06, "loss": 0.2489, "step": 650 }, { "epoch": 0.17, "learning_rate": 7.876667868734223e-06, "loss": 0.2276, "step": 675 }, { "epoch": 0.18, "learning_rate": 7.858636855391273e-06, "loss": 0.2024, "step": 700 }, { "epoch": 0.19, "learning_rate": 7.840605842048324e-06, "loss": 0.2235, "step": 725 }, { "epoch": 0.19, "learning_rate": 7.822574828705373e-06, "loss": 0.1964, "step": 750 }, { "epoch": 0.2, "learning_rate": 7.804543815362423e-06, "loss": 0.205, "step": 775 }, { "epoch": 0.21, "learning_rate": 7.786512802019472e-06, "loss": 0.2044, "step": 800 }, { "epoch": 0.21, "learning_rate": 7.768481788676523e-06, "loss": 0.1805, "step": 825 }, { "epoch": 0.22, "learning_rate": 7.750450775333573e-06, "loss": 0.1725, "step": 850 }, { "epoch": 0.23, "learning_rate": 7.732419761990624e-06, "loss": 0.1888, "step": 875 }, { "epoch": 0.23, "learning_rate": 7.714388748647674e-06, "loss": 0.1867, "step": 900 }, { "epoch": 0.24, "learning_rate": 7.696357735304725e-06, "loss": 0.2047, "step": 925 }, { "epoch": 0.25, "learning_rate": 7.678326721961773e-06, "loss": 0.1637, "step": 950 }, { "epoch": 0.25, "learning_rate": 7.660295708618824e-06, "loss": 0.1941, "step": 975 }, { "epoch": 0.26, "learning_rate": 7.642264695275874e-06, "loss": 0.1726, "step": 1000 }, { "epoch": 0.26, "eval_loss": 0.18191906809806824, "eval_runtime": 5156.4322, "eval_samples_per_second": 2.115, "eval_steps_per_second": 0.265, "eval_wer": 47.49931790039247, "step": 1000 }, { "epoch": 0.27, "learning_rate": 7.624233681932924e-06, "loss": 0.1846, "step": 1025 }, { "epoch": 0.27, "learning_rate": 7.606202668589975e-06, "loss": 0.1819, "step": 1050 }, { "epoch": 0.28, "learning_rate": 7.588171655247024e-06, "loss": 0.1714, "step": 1075 }, { "epoch": 0.28, "learning_rate": 7.570140641904075e-06, "loss": 0.1706, "step": 1100 }, { "epoch": 0.29, "learning_rate": 7.552109628561125e-06, "loss": 0.1503, "step": 1125 }, { "epoch": 0.3, "learning_rate": 7.534078615218175e-06, "loss": 0.1544, "step": 1150 }, { "epoch": 0.3, "learning_rate": 7.5160476018752255e-06, "loss": 0.1811, "step": 1175 }, { "epoch": 0.31, "learning_rate": 7.498016588532275e-06, "loss": 0.1715, "step": 1200 }, { "epoch": 0.32, "learning_rate": 7.479985575189326e-06, "loss": 0.1732, "step": 1225 }, { "epoch": 0.32, "learning_rate": 7.461954561846375e-06, "loss": 0.1624, "step": 1250 }, { "epoch": 0.33, "learning_rate": 7.443923548503425e-06, "loss": 0.1469, "step": 1275 }, { "epoch": 0.34, "learning_rate": 7.4258925351604755e-06, "loss": 0.1764, "step": 1300 }, { "epoch": 0.34, "learning_rate": 7.407861521817526e-06, "loss": 0.1604, "step": 1325 }, { "epoch": 0.35, "learning_rate": 7.389830508474576e-06, "loss": 0.1561, "step": 1350 }, { "epoch": 0.36, "learning_rate": 7.371799495131626e-06, "loss": 0.1661, "step": 1375 }, { "epoch": 0.36, "learning_rate": 7.353768481788676e-06, "loss": 0.1534, "step": 1400 }, { "epoch": 0.37, "learning_rate": 7.335737468445726e-06, "loss": 0.1423, "step": 1425 }, { "epoch": 0.38, "learning_rate": 7.317706455102777e-06, "loss": 0.1806, "step": 1450 }, { "epoch": 0.38, "learning_rate": 7.299675441759827e-06, "loss": 0.1488, "step": 1475 }, { "epoch": 0.39, "learning_rate": 7.281644428416877e-06, "loss": 0.1877, "step": 1500 }, { "epoch": 0.39, "eval_loss": 0.1481647789478302, "eval_runtime": 4672.6288, "eval_samples_per_second": 2.334, "eval_steps_per_second": 0.292, "eval_wer": 40.243876844292394, "step": 1500 }, { "epoch": 0.39, "learning_rate": 7.263613415073928e-06, "loss": 0.1544, "step": 1525 }, { "epoch": 0.4, "learning_rate": 7.2455824017309764e-06, "loss": 0.1771, "step": 1550 }, { "epoch": 0.41, "learning_rate": 7.227551388388027e-06, "loss": 0.141, "step": 1575 }, { "epoch": 0.41, "learning_rate": 7.209520375045077e-06, "loss": 0.1359, "step": 1600 }, { "epoch": 0.42, "learning_rate": 7.191489361702127e-06, "loss": 0.1242, "step": 1625 }, { "epoch": 0.43, "learning_rate": 7.173458348359178e-06, "loss": 0.1281, "step": 1650 }, { "epoch": 0.43, "learning_rate": 7.155427335016227e-06, "loss": 0.1404, "step": 1675 }, { "epoch": 0.44, "learning_rate": 7.137396321673278e-06, "loss": 0.1502, "step": 1700 }, { "epoch": 0.45, "learning_rate": 7.119365308330328e-06, "loss": 0.1348, "step": 1725 }, { "epoch": 0.45, "learning_rate": 7.101334294987378e-06, "loss": 0.1473, "step": 1750 }, { "epoch": 0.46, "learning_rate": 7.0833032816444285e-06, "loss": 0.1401, "step": 1775 }, { "epoch": 0.47, "learning_rate": 7.065272268301478e-06, "loss": 0.1089, "step": 1800 }, { "epoch": 0.47, "learning_rate": 7.047241254958528e-06, "loss": 0.1416, "step": 1825 }, { "epoch": 0.48, "learning_rate": 7.029210241615578e-06, "loss": 0.1239, "step": 1850 }, { "epoch": 0.49, "learning_rate": 7.011179228272628e-06, "loss": 0.133, "step": 1875 }, { "epoch": 0.49, "learning_rate": 6.993148214929679e-06, "loss": 0.1305, "step": 1900 }, { "epoch": 0.5, "learning_rate": 6.975117201586729e-06, "loss": 0.1352, "step": 1925 }, { "epoch": 0.5, "learning_rate": 6.957086188243779e-06, "loss": 0.1567, "step": 1950 }, { "epoch": 0.51, "learning_rate": 6.939055174900829e-06, "loss": 0.1324, "step": 1975 }, { "epoch": 0.52, "learning_rate": 6.921024161557879e-06, "loss": 0.1354, "step": 2000 }, { "epoch": 0.52, "eval_loss": 0.12738725543022156, "eval_runtime": 4615.525, "eval_samples_per_second": 2.363, "eval_steps_per_second": 0.296, "eval_wer": 34.36732637941528, "step": 2000 }, { "epoch": 0.52, "learning_rate": 6.9029931482149295e-06, "loss": 0.14, "step": 2025 }, { "epoch": 0.53, "learning_rate": 6.88496213487198e-06, "loss": 0.1316, "step": 2050 }, { "epoch": 0.54, "learning_rate": 6.86693112152903e-06, "loss": 0.1072, "step": 2075 }, { "epoch": 0.54, "learning_rate": 6.84890010818608e-06, "loss": 0.108, "step": 2100 }, { "epoch": 0.55, "learning_rate": 6.83086909484313e-06, "loss": 0.0992, "step": 2125 }, { "epoch": 0.56, "learning_rate": 6.8128380815001795e-06, "loss": 0.1377, "step": 2150 }, { "epoch": 0.56, "learning_rate": 6.79480706815723e-06, "loss": 0.1154, "step": 2175 }, { "epoch": 0.57, "learning_rate": 6.77677605481428e-06, "loss": 0.1272, "step": 2200 }, { "epoch": 0.58, "learning_rate": 6.75874504147133e-06, "loss": 0.1183, "step": 2225 }, { "epoch": 0.58, "learning_rate": 6.740714028128381e-06, "loss": 0.1131, "step": 2250 }, { "epoch": 0.59, "learning_rate": 6.72268301478543e-06, "loss": 0.1208, "step": 2275 }, { "epoch": 0.6, "learning_rate": 6.704652001442481e-06, "loss": 0.1224, "step": 2300 }, { "epoch": 0.6, "learning_rate": 6.686620988099531e-06, "loss": 0.1313, "step": 2325 }, { "epoch": 0.61, "learning_rate": 6.668589974756581e-06, "loss": 0.1284, "step": 2350 }, { "epoch": 0.61, "learning_rate": 6.650558961413632e-06, "loss": 0.1191, "step": 2375 }, { "epoch": 0.62, "learning_rate": 6.6325279480706804e-06, "loss": 0.1103, "step": 2400 }, { "epoch": 0.63, "learning_rate": 6.614496934727731e-06, "loss": 0.1218, "step": 2425 }, { "epoch": 0.63, "learning_rate": 6.5964659213847815e-06, "loss": 0.1555, "step": 2450 }, { "epoch": 0.64, "learning_rate": 6.578434908041831e-06, "loss": 0.1221, "step": 2475 }, { "epoch": 0.65, "learning_rate": 6.560403894698882e-06, "loss": 0.1186, "step": 2500 }, { "epoch": 0.65, "eval_loss": 0.11205455660820007, "eval_runtime": 4616.7031, "eval_samples_per_second": 2.363, "eval_steps_per_second": 0.295, "eval_wer": 31.03028522257435, "step": 2500 }, { "epoch": 0.65, "learning_rate": 6.542372881355932e-06, "loss": 0.1156, "step": 2525 }, { "epoch": 0.66, "learning_rate": 6.524341868012982e-06, "loss": 0.1085, "step": 2550 }, { "epoch": 0.67, "learning_rate": 6.506310854670032e-06, "loss": 0.1327, "step": 2575 }, { "epoch": 0.67, "learning_rate": 6.488279841327083e-06, "loss": 0.1383, "step": 2600 }, { "epoch": 0.68, "learning_rate": 6.4702488279841325e-06, "loss": 0.1128, "step": 2625 }, { "epoch": 0.69, "learning_rate": 6.452217814641183e-06, "loss": 0.1116, "step": 2650 }, { "epoch": 0.69, "learning_rate": 6.434186801298233e-06, "loss": 0.1082, "step": 2675 }, { "epoch": 0.7, "learning_rate": 6.416155787955282e-06, "loss": 0.1235, "step": 2700 }, { "epoch": 0.71, "learning_rate": 6.398124774612333e-06, "loss": 0.1142, "step": 2725 }, { "epoch": 0.71, "learning_rate": 6.3800937612693826e-06, "loss": 0.1218, "step": 2750 }, { "epoch": 0.72, "learning_rate": 6.362062747926433e-06, "loss": 0.1019, "step": 2775 }, { "epoch": 0.72, "learning_rate": 6.344031734583484e-06, "loss": 0.1024, "step": 2800 }, { "epoch": 0.73, "learning_rate": 6.326000721240533e-06, "loss": 0.1283, "step": 2825 }, { "epoch": 0.74, "learning_rate": 6.307969707897584e-06, "loss": 0.0894, "step": 2850 }, { "epoch": 0.74, "learning_rate": 6.2899386945546334e-06, "loss": 0.1031, "step": 2875 }, { "epoch": 0.75, "learning_rate": 6.271907681211684e-06, "loss": 0.0935, "step": 2900 }, { "epoch": 0.76, "learning_rate": 6.2538766678687345e-06, "loss": 0.1256, "step": 2925 }, { "epoch": 0.76, "learning_rate": 6.235845654525784e-06, "loss": 0.0871, "step": 2950 }, { "epoch": 0.77, "learning_rate": 6.217814641182835e-06, "loss": 0.099, "step": 2975 }, { "epoch": 0.78, "learning_rate": 6.199783627839884e-06, "loss": 0.1367, "step": 3000 }, { "epoch": 0.78, "eval_loss": 0.10175555944442749, "eval_runtime": 4617.2532, "eval_samples_per_second": 2.362, "eval_steps_per_second": 0.295, "eval_wer": 28.322874472684532, "step": 3000 }, { "epoch": 0.78, "learning_rate": 6.181752614496934e-06, "loss": 0.1023, "step": 3025 }, { "epoch": 0.79, "learning_rate": 6.1637216011539845e-06, "loss": 0.0861, "step": 3050 }, { "epoch": 0.8, "learning_rate": 6.145690587811034e-06, "loss": 0.1037, "step": 3075 }, { "epoch": 0.8, "learning_rate": 6.127659574468085e-06, "loss": 0.098, "step": 3100 }, { "epoch": 0.81, "learning_rate": 6.109628561125135e-06, "loss": 0.1019, "step": 3125 }, { "epoch": 0.82, "learning_rate": 6.091597547782185e-06, "loss": 0.0791, "step": 3150 }, { "epoch": 0.82, "learning_rate": 6.073566534439235e-06, "loss": 0.1013, "step": 3175 }, { "epoch": 0.83, "learning_rate": 6.055535521096286e-06, "loss": 0.0994, "step": 3200 }, { "epoch": 0.83, "learning_rate": 6.037504507753336e-06, "loss": 0.0995, "step": 3225 }, { "epoch": 0.84, "learning_rate": 6.019473494410386e-06, "loss": 0.1041, "step": 3250 }, { "epoch": 0.85, "learning_rate": 6.001442481067435e-06, "loss": 0.1038, "step": 3275 }, { "epoch": 0.85, "learning_rate": 5.9834114677244854e-06, "loss": 0.0966, "step": 3300 }, { "epoch": 0.86, "learning_rate": 5.965380454381536e-06, "loss": 0.0625, "step": 3325 }, { "epoch": 0.87, "learning_rate": 5.947349441038586e-06, "loss": 0.0919, "step": 3350 }, { "epoch": 0.87, "learning_rate": 5.929318427695636e-06, "loss": 0.0843, "step": 3375 }, { "epoch": 0.88, "learning_rate": 5.911287414352687e-06, "loss": 0.0989, "step": 3400 }, { "epoch": 0.89, "learning_rate": 5.893256401009736e-06, "loss": 0.0968, "step": 3425 }, { "epoch": 0.89, "learning_rate": 5.875225387666787e-06, "loss": 0.0817, "step": 3450 }, { "epoch": 0.9, "learning_rate": 5.8571943743238365e-06, "loss": 0.0778, "step": 3475 }, { "epoch": 0.91, "learning_rate": 5.839163360980887e-06, "loss": 0.1058, "step": 3500 }, { "epoch": 0.91, "eval_loss": 0.09466591477394104, "eval_runtime": 4606.7225, "eval_samples_per_second": 2.368, "eval_steps_per_second": 0.296, "eval_wer": 27.83805905933217, "step": 3500 }, { "epoch": 0.91, "learning_rate": 5.8211323476379375e-06, "loss": 0.1071, "step": 3525 }, { "epoch": 0.92, "learning_rate": 5.803101334294987e-06, "loss": 0.088, "step": 3550 }, { "epoch": 0.93, "learning_rate": 5.785070320952037e-06, "loss": 0.0747, "step": 3575 }, { "epoch": 0.93, "learning_rate": 5.767039307609087e-06, "loss": 0.0716, "step": 3600 }, { "epoch": 0.94, "learning_rate": 5.749008294266137e-06, "loss": 0.0809, "step": 3625 }, { "epoch": 0.94, "learning_rate": 5.730977280923188e-06, "loss": 0.0824, "step": 3650 }, { "epoch": 0.95, "learning_rate": 5.712946267580237e-06, "loss": 0.0794, "step": 3675 }, { "epoch": 0.96, "learning_rate": 5.694915254237288e-06, "loss": 0.0896, "step": 3700 }, { "epoch": 0.96, "learning_rate": 5.676884240894338e-06, "loss": 0.0888, "step": 3725 }, { "epoch": 0.97, "learning_rate": 5.658853227551388e-06, "loss": 0.0932, "step": 3750 }, { "epoch": 0.98, "learning_rate": 5.6408222142084385e-06, "loss": 0.1099, "step": 3775 }, { "epoch": 0.98, "learning_rate": 5.622791200865489e-06, "loss": 0.0872, "step": 3800 }, { "epoch": 0.99, "learning_rate": 5.604760187522539e-06, "loss": 0.0971, "step": 3825 }, { "epoch": 1.0, "learning_rate": 5.586729174179589e-06, "loss": 0.0888, "step": 3850 }, { "epoch": 1.0, "learning_rate": 5.568698160836638e-06, "loss": 0.074, "step": 3875 }, { "epoch": 1.01, "learning_rate": 5.5506671474936885e-06, "loss": 0.0521, "step": 3900 }, { "epoch": 1.02, "learning_rate": 5.532636134150739e-06, "loss": 0.0707, "step": 3925 }, { "epoch": 1.02, "learning_rate": 5.514605120807789e-06, "loss": 0.0581, "step": 3950 }, { "epoch": 1.03, "learning_rate": 5.496574107464839e-06, "loss": 0.0725, "step": 3975 }, { "epoch": 1.04, "learning_rate": 5.47854309412189e-06, "loss": 0.0807, "step": 4000 }, { "epoch": 1.04, "eval_loss": 0.08684225380420685, "eval_runtime": 4613.5658, "eval_samples_per_second": 2.364, "eval_steps_per_second": 0.296, "eval_wer": 24.383486893193695, "step": 4000 }, { "epoch": 1.04, "learning_rate": 5.460512080778939e-06, "loss": 0.0662, "step": 4025 }, { "epoch": 1.05, "learning_rate": 5.44248106743599e-06, "loss": 0.0821, "step": 4050 }, { "epoch": 1.05, "learning_rate": 5.4244500540930396e-06, "loss": 0.0688, "step": 4075 }, { "epoch": 1.06, "learning_rate": 5.40641904075009e-06, "loss": 0.0563, "step": 4100 }, { "epoch": 1.07, "learning_rate": 5.388388027407141e-06, "loss": 0.0651, "step": 4125 }, { "epoch": 1.07, "learning_rate": 5.370357014064189e-06, "loss": 0.0619, "step": 4150 }, { "epoch": 1.08, "learning_rate": 5.35232600072124e-06, "loss": 0.0745, "step": 4175 }, { "epoch": 1.09, "learning_rate": 5.3342949873782905e-06, "loss": 0.0664, "step": 4200 }, { "epoch": 1.09, "learning_rate": 5.31626397403534e-06, "loss": 0.098, "step": 4225 }, { "epoch": 1.1, "learning_rate": 5.298232960692391e-06, "loss": 0.0801, "step": 4250 }, { "epoch": 1.11, "learning_rate": 5.280201947349441e-06, "loss": 0.0575, "step": 4275 }, { "epoch": 1.11, "learning_rate": 5.262170934006491e-06, "loss": 0.0558, "step": 4300 }, { "epoch": 1.12, "learning_rate": 5.244139920663541e-06, "loss": 0.0637, "step": 4325 }, { "epoch": 1.13, "learning_rate": 5.226108907320591e-06, "loss": 0.0697, "step": 4350 }, { "epoch": 1.13, "learning_rate": 5.2080778939776415e-06, "loss": 0.0576, "step": 4375 }, { "epoch": 1.14, "learning_rate": 5.190046880634692e-06, "loss": 0.0714, "step": 4400 }, { "epoch": 1.15, "learning_rate": 5.172015867291742e-06, "loss": 0.0675, "step": 4425 }, { "epoch": 1.15, "learning_rate": 5.153984853948791e-06, "loss": 0.0555, "step": 4450 }, { "epoch": 1.16, "learning_rate": 5.135953840605842e-06, "loss": 0.0568, "step": 4475 }, { "epoch": 1.16, "learning_rate": 5.1179228272628916e-06, "loss": 0.0698, "step": 4500 }, { "epoch": 1.16, "eval_loss": 0.08182203769683838, "eval_runtime": 4681.2354, "eval_samples_per_second": 2.33, "eval_steps_per_second": 0.291, "eval_wer": 23.162003903708523, "step": 4500 }, { "epoch": 1.17, "learning_rate": 5.099891813919942e-06, "loss": 0.06, "step": 4525 }, { "epoch": 1.18, "learning_rate": 5.081860800576992e-06, "loss": 0.0637, "step": 4550 }, { "epoch": 1.18, "learning_rate": 5.063829787234042e-06, "loss": 0.0519, "step": 4575 }, { "epoch": 1.19, "learning_rate": 5.045798773891093e-06, "loss": 0.04, "step": 4600 }, { "epoch": 1.2, "learning_rate": 5.0277677605481424e-06, "loss": 0.0714, "step": 4625 }, { "epoch": 1.2, "learning_rate": 5.009736747205193e-06, "loss": 0.049, "step": 4650 }, { "epoch": 1.21, "learning_rate": 4.9917057338622435e-06, "loss": 0.0687, "step": 4675 }, { "epoch": 1.22, "learning_rate": 4.973674720519293e-06, "loss": 0.0469, "step": 4700 }, { "epoch": 1.22, "learning_rate": 4.955643707176344e-06, "loss": 0.073, "step": 4725 }, { "epoch": 1.23, "learning_rate": 4.9376126938333925e-06, "loss": 0.0583, "step": 4750 }, { "epoch": 1.24, "learning_rate": 4.919581680490443e-06, "loss": 0.0631, "step": 4775 }, { "epoch": 1.24, "learning_rate": 4.9015506671474935e-06, "loss": 0.0599, "step": 4800 }, { "epoch": 1.25, "learning_rate": 4.883519653804543e-06, "loss": 0.0597, "step": 4825 }, { "epoch": 1.26, "learning_rate": 4.865488640461594e-06, "loss": 0.0697, "step": 4850 }, { "epoch": 1.26, "learning_rate": 4.847457627118644e-06, "loss": 0.0638, "step": 4875 }, { "epoch": 1.27, "learning_rate": 4.829426613775694e-06, "loss": 0.0492, "step": 4900 }, { "epoch": 1.27, "learning_rate": 4.811395600432744e-06, "loss": 0.0576, "step": 4925 }, { "epoch": 1.28, "learning_rate": 4.793364587089794e-06, "loss": 0.0631, "step": 4950 }, { "epoch": 1.29, "learning_rate": 4.775333573746845e-06, "loss": 0.0616, "step": 4975 }, { "epoch": 1.29, "learning_rate": 4.757302560403895e-06, "loss": 0.0434, "step": 5000 }, { "epoch": 1.29, "eval_loss": 0.07633961737155914, "eval_runtime": 4650.315, "eval_samples_per_second": 2.345, "eval_steps_per_second": 0.293, "eval_wer": 22.19027430898063, "step": 5000 }, { "epoch": 1.3, "learning_rate": 4.739271547060944e-06, "loss": 0.0745, "step": 5025 }, { "epoch": 1.31, "learning_rate": 4.7212405337179944e-06, "loss": 0.0392, "step": 5050 }, { "epoch": 1.31, "learning_rate": 4.703209520375045e-06, "loss": 0.0609, "step": 5075 }, { "epoch": 1.32, "learning_rate": 4.685178507032095e-06, "loss": 0.0632, "step": 5100 }, { "epoch": 1.33, "learning_rate": 4.667147493689145e-06, "loss": 0.0586, "step": 5125 }, { "epoch": 1.33, "learning_rate": 4.649116480346195e-06, "loss": 0.0514, "step": 5150 }, { "epoch": 1.34, "learning_rate": 4.631085467003245e-06, "loss": 0.0732, "step": 5175 }, { "epoch": 1.35, "learning_rate": 4.613054453660296e-06, "loss": 0.0731, "step": 5200 }, { "epoch": 1.35, "learning_rate": 4.5950234403173455e-06, "loss": 0.0773, "step": 5225 }, { "epoch": 1.36, "learning_rate": 4.576992426974396e-06, "loss": 0.0475, "step": 5250 }, { "epoch": 1.37, "learning_rate": 4.5589614136314465e-06, "loss": 0.0695, "step": 5275 }, { "epoch": 1.37, "learning_rate": 4.540930400288496e-06, "loss": 0.0624, "step": 5300 }, { "epoch": 1.38, "learning_rate": 4.522899386945546e-06, "loss": 0.0516, "step": 5325 }, { "epoch": 1.38, "learning_rate": 4.5048683736025955e-06, "loss": 0.0582, "step": 5350 }, { "epoch": 1.39, "learning_rate": 4.486837360259646e-06, "loss": 0.0495, "step": 5375 }, { "epoch": 1.4, "learning_rate": 4.4688063469166966e-06, "loss": 0.0689, "step": 5400 }, { "epoch": 1.4, "learning_rate": 4.450775333573746e-06, "loss": 0.0634, "step": 5425 }, { "epoch": 1.41, "learning_rate": 4.432744320230797e-06, "loss": 0.0379, "step": 5450 }, { "epoch": 1.42, "learning_rate": 4.414713306887847e-06, "loss": 0.0701, "step": 5475 }, { "epoch": 1.42, "learning_rate": 4.396682293544897e-06, "loss": 0.0536, "step": 5500 }, { "epoch": 1.42, "eval_loss": 0.07404134422540665, "eval_runtime": 4631.2871, "eval_samples_per_second": 2.355, "eval_steps_per_second": 0.295, "eval_wer": 22.3413856066489, "step": 5500 }, { "epoch": 1.43, "learning_rate": 4.3786512802019475e-06, "loss": 0.0549, "step": 5525 }, { "epoch": 1.44, "learning_rate": 4.360620266858997e-06, "loss": 0.0664, "step": 5550 }, { "epoch": 1.44, "learning_rate": 4.342589253516048e-06, "loss": 0.0526, "step": 5575 }, { "epoch": 1.45, "learning_rate": 4.324558240173098e-06, "loss": 0.0581, "step": 5600 }, { "epoch": 1.46, "learning_rate": 4.306527226830147e-06, "loss": 0.0379, "step": 5625 }, { "epoch": 1.46, "learning_rate": 4.2884962134871975e-06, "loss": 0.079, "step": 5650 }, { "epoch": 1.47, "learning_rate": 4.270465200144248e-06, "loss": 0.061, "step": 5675 }, { "epoch": 1.48, "learning_rate": 4.252434186801298e-06, "loss": 0.0581, "step": 5700 }, { "epoch": 1.48, "learning_rate": 4.234403173458348e-06, "loss": 0.0695, "step": 5725 }, { "epoch": 1.49, "learning_rate": 4.216372160115398e-06, "loss": 0.047, "step": 5750 }, { "epoch": 1.49, "learning_rate": 4.198341146772448e-06, "loss": 0.0497, "step": 5775 }, { "epoch": 1.5, "learning_rate": 4.180310133429499e-06, "loss": 0.0632, "step": 5800 }, { "epoch": 1.51, "learning_rate": 4.1622791200865486e-06, "loss": 0.0566, "step": 5825 }, { "epoch": 1.51, "learning_rate": 4.144248106743599e-06, "loss": 0.0582, "step": 5850 }, { "epoch": 1.52, "learning_rate": 4.12621709340065e-06, "loss": 0.0624, "step": 5875 }, { "epoch": 1.53, "learning_rate": 4.108186080057699e-06, "loss": 0.0523, "step": 5900 }, { "epoch": 1.53, "learning_rate": 4.090155066714749e-06, "loss": 0.0498, "step": 5925 }, { "epoch": 1.54, "learning_rate": 4.072124053371799e-06, "loss": 0.0561, "step": 5950 }, { "epoch": 1.55, "learning_rate": 4.054093040028849e-06, "loss": 0.0548, "step": 5975 }, { "epoch": 1.55, "learning_rate": 4.0360620266859e-06, "loss": 0.0403, "step": 6000 }, { "epoch": 1.55, "eval_loss": 0.07143695652484894, "eval_runtime": 4576.9308, "eval_samples_per_second": 2.383, "eval_steps_per_second": 0.298, "eval_wer": 20.996075303796673, "step": 6000 }, { "epoch": 1.56, "learning_rate": 4.018031013342949e-06, "loss": 0.047, "step": 6025 }, { "epoch": 1.57, "learning_rate": 4e-06, "loss": 0.0445, "step": 6050 }, { "epoch": 1.57, "learning_rate": 3.9819689866570495e-06, "loss": 0.0459, "step": 6075 }, { "epoch": 1.58, "learning_rate": 3.9639379733141e-06, "loss": 0.042, "step": 6100 }, { "epoch": 1.59, "learning_rate": 3.9459069599711505e-06, "loss": 0.0513, "step": 6125 }, { "epoch": 1.59, "learning_rate": 3.9278759466282e-06, "loss": 0.055, "step": 6150 }, { "epoch": 1.6, "learning_rate": 3.909844933285251e-06, "loss": 0.0557, "step": 6175 }, { "epoch": 1.6, "learning_rate": 3.8918139199423e-06, "loss": 0.0646, "step": 6200 }, { "epoch": 1.61, "learning_rate": 3.873782906599351e-06, "loss": 0.0493, "step": 6225 }, { "epoch": 1.62, "learning_rate": 3.8557518932564006e-06, "loss": 0.0523, "step": 6250 }, { "epoch": 1.62, "learning_rate": 3.837720879913451e-06, "loss": 0.0515, "step": 6275 }, { "epoch": 1.63, "learning_rate": 3.819689866570502e-06, "loss": 0.054, "step": 6300 }, { "epoch": 1.64, "learning_rate": 3.8016588532275513e-06, "loss": 0.0643, "step": 6325 }, { "epoch": 1.64, "learning_rate": 3.7836278398846013e-06, "loss": 0.0699, "step": 6350 }, { "epoch": 1.65, "learning_rate": 3.7655968265416514e-06, "loss": 0.0474, "step": 6375 }, { "epoch": 1.66, "learning_rate": 3.747565813198702e-06, "loss": 0.0359, "step": 6400 }, { "epoch": 1.66, "learning_rate": 3.7295347998557516e-06, "loss": 0.0587, "step": 6425 }, { "epoch": 1.67, "learning_rate": 3.7115037865128017e-06, "loss": 0.0615, "step": 6450 }, { "epoch": 1.68, "learning_rate": 3.693472773169852e-06, "loss": 0.0482, "step": 6475 }, { "epoch": 1.68, "learning_rate": 3.6754417598269023e-06, "loss": 0.0526, "step": 6500 }, { "epoch": 1.68, "eval_loss": 0.06842584162950516, "eval_runtime": 4579.0646, "eval_samples_per_second": 2.382, "eval_steps_per_second": 0.298, "eval_wer": 20.52385249858333, "step": 6500 }, { "epoch": 1.69, "learning_rate": 3.6574107464839524e-06, "loss": 0.0636, "step": 6525 }, { "epoch": 1.7, "learning_rate": 3.6393797331410025e-06, "loss": 0.0671, "step": 6550 }, { "epoch": 1.7, "learning_rate": 3.621348719798052e-06, "loss": 0.0443, "step": 6575 }, { "epoch": 1.71, "learning_rate": 3.6033177064551027e-06, "loss": 0.0484, "step": 6600 }, { "epoch": 1.71, "learning_rate": 3.5852866931121528e-06, "loss": 0.0734, "step": 6625 }, { "epoch": 1.72, "learning_rate": 3.567255679769203e-06, "loss": 0.0705, "step": 6650 }, { "epoch": 1.73, "learning_rate": 3.549224666426253e-06, "loss": 0.0559, "step": 6675 }, { "epoch": 1.73, "learning_rate": 3.5311936530833035e-06, "loss": 0.0569, "step": 6700 }, { "epoch": 1.74, "learning_rate": 3.513162639740353e-06, "loss": 0.0457, "step": 6725 }, { "epoch": 1.75, "learning_rate": 3.4951316263974032e-06, "loss": 0.0558, "step": 6750 }, { "epoch": 1.75, "learning_rate": 3.4771006130544533e-06, "loss": 0.0449, "step": 6775 }, { "epoch": 1.76, "learning_rate": 3.459069599711504e-06, "loss": 0.0493, "step": 6800 }, { "epoch": 1.77, "learning_rate": 3.441038586368554e-06, "loss": 0.0608, "step": 6825 }, { "epoch": 1.77, "learning_rate": 3.423007573025604e-06, "loss": 0.0444, "step": 6850 }, { "epoch": 1.78, "learning_rate": 3.4049765596826537e-06, "loss": 0.0569, "step": 6875 }, { "epoch": 1.79, "learning_rate": 3.3869455463397042e-06, "loss": 0.0657, "step": 6900 }, { "epoch": 1.79, "learning_rate": 3.3689145329967543e-06, "loss": 0.0662, "step": 6925 }, { "epoch": 1.8, "learning_rate": 3.3508835196538044e-06, "loss": 0.0561, "step": 6950 }, { "epoch": 1.81, "learning_rate": 3.3328525063108545e-06, "loss": 0.0518, "step": 6975 }, { "epoch": 1.81, "learning_rate": 3.314821492967905e-06, "loss": 0.0505, "step": 7000 }, { "epoch": 1.81, "eval_loss": 0.06526884436607361, "eval_runtime": 4621.9984, "eval_samples_per_second": 2.36, "eval_steps_per_second": 0.295, "eval_wer": 19.052616114340882, "step": 7000 }, { "epoch": 1.82, "learning_rate": 3.2967904796249547e-06, "loss": 0.0534, "step": 7025 }, { "epoch": 1.82, "learning_rate": 3.2787594662820048e-06, "loss": 0.0494, "step": 7050 }, { "epoch": 1.83, "learning_rate": 3.260728452939055e-06, "loss": 0.036, "step": 7075 }, { "epoch": 1.84, "learning_rate": 3.2426974395961054e-06, "loss": 0.0541, "step": 7100 }, { "epoch": 1.84, "learning_rate": 3.2246664262531555e-06, "loss": 0.0473, "step": 7125 }, { "epoch": 1.85, "learning_rate": 3.206635412910205e-06, "loss": 0.0508, "step": 7150 }, { "epoch": 1.86, "learning_rate": 3.1886043995672552e-06, "loss": 0.0481, "step": 7175 }, { "epoch": 1.86, "learning_rate": 3.1705733862243058e-06, "loss": 0.057, "step": 7200 }, { "epoch": 1.87, "learning_rate": 3.152542372881356e-06, "loss": 0.0283, "step": 7225 }, { "epoch": 1.88, "learning_rate": 3.134511359538406e-06, "loss": 0.0436, "step": 7250 }, { "epoch": 1.88, "learning_rate": 3.116480346195456e-06, "loss": 0.0687, "step": 7275 }, { "epoch": 1.89, "learning_rate": 3.098449332852506e-06, "loss": 0.0441, "step": 7300 }, { "epoch": 1.9, "learning_rate": 3.0804183195095562e-06, "loss": 0.0504, "step": 7325 }, { "epoch": 1.9, "learning_rate": 3.0623873061666063e-06, "loss": 0.0384, "step": 7350 }, { "epoch": 1.91, "learning_rate": 3.0443562928236564e-06, "loss": 0.04, "step": 7375 }, { "epoch": 1.92, "learning_rate": 3.026325279480707e-06, "loss": 0.0482, "step": 7400 }, { "epoch": 1.92, "learning_rate": 3.008294266137757e-06, "loss": 0.0553, "step": 7425 }, { "epoch": 1.93, "learning_rate": 2.9902632527948067e-06, "loss": 0.0712, "step": 7450 }, { "epoch": 1.93, "learning_rate": 2.9722322394518568e-06, "loss": 0.0648, "step": 7475 }, { "epoch": 1.94, "learning_rate": 2.9542012261089073e-06, "loss": 0.053, "step": 7500 }, { "epoch": 1.94, "eval_loss": 0.06240825355052948, "eval_runtime": 4653.2425, "eval_samples_per_second": 2.344, "eval_steps_per_second": 0.293, "eval_wer": 18.697924318425084, "step": 7500 }, { "epoch": 1.95, "learning_rate": 2.9361702127659574e-06, "loss": 0.0613, "step": 7525 }, { "epoch": 1.95, "learning_rate": 2.9181391994230075e-06, "loss": 0.0492, "step": 7550 }, { "epoch": 1.96, "learning_rate": 2.9001081860800576e-06, "loss": 0.0372, "step": 7575 }, { "epoch": 1.97, "learning_rate": 2.8820771727371077e-06, "loss": 0.0474, "step": 7600 }, { "epoch": 1.97, "learning_rate": 2.8640461593941577e-06, "loss": 0.0472, "step": 7625 }, { "epoch": 1.98, "learning_rate": 2.846015146051208e-06, "loss": 0.0456, "step": 7650 }, { "epoch": 1.99, "learning_rate": 2.827984132708258e-06, "loss": 0.0496, "step": 7675 }, { "epoch": 1.99, "learning_rate": 2.8099531193653084e-06, "loss": 0.0517, "step": 7700 }, { "epoch": 2.0, "learning_rate": 2.7919221060223585e-06, "loss": 0.048, "step": 7725 }, { "epoch": 2.01, "learning_rate": 2.773891092679408e-06, "loss": 0.0342, "step": 7750 }, { "epoch": 2.01, "learning_rate": 2.7558600793364583e-06, "loss": 0.0259, "step": 7775 }, { "epoch": 2.02, "learning_rate": 2.737829065993509e-06, "loss": 0.0323, "step": 7800 }, { "epoch": 2.03, "learning_rate": 2.719798052650559e-06, "loss": 0.0338, "step": 7825 }, { "epoch": 2.03, "learning_rate": 2.701767039307609e-06, "loss": 0.0356, "step": 7850 }, { "epoch": 2.04, "learning_rate": 2.6837360259646595e-06, "loss": 0.0241, "step": 7875 }, { "epoch": 2.04, "learning_rate": 2.665705012621709e-06, "loss": 0.0359, "step": 7900 }, { "epoch": 2.05, "learning_rate": 2.6476739992787593e-06, "loss": 0.0336, "step": 7925 }, { "epoch": 2.06, "learning_rate": 2.6296429859358094e-06, "loss": 0.0395, "step": 7950 }, { "epoch": 2.06, "learning_rate": 2.61161197259286e-06, "loss": 0.0319, "step": 7975 }, { "epoch": 2.07, "learning_rate": 2.59358095924991e-06, "loss": 0.0234, "step": 8000 }, { "epoch": 2.07, "eval_loss": 0.0625641718506813, "eval_runtime": 4589.637, "eval_samples_per_second": 2.376, "eval_steps_per_second": 0.297, "eval_wer": 18.46076353180683, "step": 8000 }, { "epoch": 2.08, "learning_rate": 2.5755499459069596e-06, "loss": 0.0411, "step": 8025 }, { "epoch": 2.08, "learning_rate": 2.5575189325640097e-06, "loss": 0.0183, "step": 8050 }, { "epoch": 2.09, "learning_rate": 2.5394879192210603e-06, "loss": 0.0457, "step": 8075 }, { "epoch": 2.1, "learning_rate": 2.5214569058781103e-06, "loss": 0.0464, "step": 8100 }, { "epoch": 2.1, "learning_rate": 2.5034258925351604e-06, "loss": 0.0334, "step": 8125 }, { "epoch": 2.11, "learning_rate": 2.4853948791922105e-06, "loss": 0.0239, "step": 8150 }, { "epoch": 2.12, "learning_rate": 2.4673638658492606e-06, "loss": 0.0329, "step": 8175 }, { "epoch": 2.12, "learning_rate": 2.4493328525063107e-06, "loss": 0.0261, "step": 8200 }, { "epoch": 2.13, "learning_rate": 2.431301839163361e-06, "loss": 0.0294, "step": 8225 }, { "epoch": 2.14, "learning_rate": 2.413270825820411e-06, "loss": 0.0277, "step": 8250 }, { "epoch": 2.14, "learning_rate": 2.3952398124774614e-06, "loss": 0.037, "step": 8275 }, { "epoch": 2.15, "learning_rate": 2.3772087991345115e-06, "loss": 0.0258, "step": 8300 }, { "epoch": 2.15, "learning_rate": 2.359177785791561e-06, "loss": 0.0354, "step": 8325 }, { "epoch": 2.16, "learning_rate": 2.3411467724486113e-06, "loss": 0.0234, "step": 8350 }, { "epoch": 2.17, "learning_rate": 2.3231157591056618e-06, "loss": 0.0316, "step": 8375 }, { "epoch": 2.17, "learning_rate": 2.305084745762712e-06, "loss": 0.0358, "step": 8400 }, { "epoch": 2.18, "learning_rate": 2.287053732419762e-06, "loss": 0.0226, "step": 8425 }, { "epoch": 2.19, "learning_rate": 2.269022719076812e-06, "loss": 0.0265, "step": 8450 }, { "epoch": 2.19, "learning_rate": 2.25171294626758e-06, "loss": 0.0345, "step": 8475 }, { "epoch": 2.2, "learning_rate": 2.23368193292463e-06, "loss": 0.0268, "step": 8500 }, { "epoch": 2.2, "eval_loss": 0.0614020936191082, "eval_runtime": 4575.3897, "eval_samples_per_second": 2.384, "eval_steps_per_second": 0.298, "eval_wer": 19.62138224862006, "step": 8500 }, { "epoch": 2.21, "learning_rate": 2.2156509195816806e-06, "loss": 0.0329, "step": 8525 }, { "epoch": 2.21, "learning_rate": 2.1976199062387307e-06, "loss": 0.0471, "step": 8550 }, { "epoch": 2.22, "learning_rate": 2.1795888928957808e-06, "loss": 0.0366, "step": 8575 }, { "epoch": 2.23, "learning_rate": 2.161557879552831e-06, "loss": 0.0247, "step": 8600 }, { "epoch": 2.23, "learning_rate": 2.143526866209881e-06, "loss": 0.0257, "step": 8625 }, { "epoch": 2.24, "learning_rate": 2.125495852866931e-06, "loss": 0.0532, "step": 8650 }, { "epoch": 2.25, "learning_rate": 2.107464839523981e-06, "loss": 0.0343, "step": 8675 }, { "epoch": 2.25, "learning_rate": 2.0894338261810312e-06, "loss": 0.027, "step": 8700 }, { "epoch": 2.26, "learning_rate": 2.0714028128380817e-06, "loss": 0.031, "step": 8725 }, { "epoch": 2.26, "learning_rate": 2.053371799495132e-06, "loss": 0.0467, "step": 8750 }, { "epoch": 2.27, "learning_rate": 2.0353407861521815e-06, "loss": 0.029, "step": 8775 }, { "epoch": 2.28, "learning_rate": 2.0173097728092316e-06, "loss": 0.0422, "step": 8800 }, { "epoch": 2.28, "learning_rate": 1.9992787594662817e-06, "loss": 0.0328, "step": 8825 }, { "epoch": 2.29, "learning_rate": 1.981247746123332e-06, "loss": 0.0363, "step": 8850 }, { "epoch": 2.3, "learning_rate": 1.9632167327803823e-06, "loss": 0.031, "step": 8875 }, { "epoch": 2.3, "learning_rate": 1.9451857194374324e-06, "loss": 0.0302, "step": 8900 }, { "epoch": 2.31, "learning_rate": 1.9271547060944825e-06, "loss": 0.0248, "step": 8925 }, { "epoch": 2.32, "learning_rate": 1.9091236927515326e-06, "loss": 0.0307, "step": 8950 }, { "epoch": 2.32, "learning_rate": 1.8910926794085827e-06, "loss": 0.0344, "step": 8975 }, { "epoch": 2.33, "learning_rate": 1.873061666065633e-06, "loss": 0.0403, "step": 9000 }, { "epoch": 2.33, "eval_loss": 0.05997191742062569, "eval_runtime": 4588.7385, "eval_samples_per_second": 2.377, "eval_steps_per_second": 0.297, "eval_wer": 17.98014565450081, "step": 9000 }, { "epoch": 2.34, "learning_rate": 1.8550306527226828e-06, "loss": 0.0326, "step": 9025 }, { "epoch": 2.34, "learning_rate": 1.8369996393797331e-06, "loss": 0.0307, "step": 9050 }, { "epoch": 2.35, "learning_rate": 1.8189686260367832e-06, "loss": 0.0396, "step": 9075 }, { "epoch": 2.36, "learning_rate": 1.8009376126938333e-06, "loss": 0.044, "step": 9100 }, { "epoch": 2.36, "learning_rate": 1.7829065993508834e-06, "loss": 0.0292, "step": 9125 }, { "epoch": 2.37, "learning_rate": 1.7648755860079337e-06, "loss": 0.0436, "step": 9150 }, { "epoch": 2.37, "learning_rate": 1.7468445726649836e-06, "loss": 0.0318, "step": 9175 }, { "epoch": 2.38, "learning_rate": 1.728813559322034e-06, "loss": 0.0347, "step": 9200 }, { "epoch": 2.39, "learning_rate": 1.7107825459790838e-06, "loss": 0.0169, "step": 9225 }, { "epoch": 2.39, "learning_rate": 1.692751532636134e-06, "loss": 0.0287, "step": 9250 }, { "epoch": 2.4, "learning_rate": 1.6747205192931842e-06, "loss": 0.0552, "step": 9275 }, { "epoch": 2.41, "learning_rate": 1.6566895059502343e-06, "loss": 0.0326, "step": 9300 }, { "epoch": 2.41, "learning_rate": 1.6386584926072844e-06, "loss": 0.0239, "step": 9325 }, { "epoch": 2.42, "learning_rate": 1.6206274792643347e-06, "loss": 0.0329, "step": 9350 }, { "epoch": 2.43, "learning_rate": 1.6025964659213846e-06, "loss": 0.0286, "step": 9375 }, { "epoch": 2.43, "learning_rate": 1.5845654525784349e-06, "loss": 0.0298, "step": 9400 }, { "epoch": 2.44, "learning_rate": 1.566534439235485e-06, "loss": 0.0391, "step": 9425 }, { "epoch": 2.45, "learning_rate": 1.548503425892535e-06, "loss": 0.0173, "step": 9450 }, { "epoch": 2.45, "learning_rate": 1.5304724125495851e-06, "loss": 0.0329, "step": 9475 }, { "epoch": 2.46, "learning_rate": 1.5124413992066354e-06, "loss": 0.0419, "step": 9500 }, { "epoch": 2.46, "eval_loss": 0.05825132131576538, "eval_runtime": 4572.6194, "eval_samples_per_second": 2.385, "eval_steps_per_second": 0.298, "eval_wer": 17.92767645392155, "step": 9500 }, { "epoch": 2.47, "learning_rate": 1.4944103858636853e-06, "loss": 0.0519, "step": 9525 }, { "epoch": 2.47, "learning_rate": 1.4763793725207356e-06, "loss": 0.0319, "step": 9550 }, { "epoch": 2.48, "learning_rate": 1.4583483591777857e-06, "loss": 0.0206, "step": 9575 }, { "epoch": 2.48, "learning_rate": 1.4403173458348358e-06, "loss": 0.0339, "step": 9600 }, { "epoch": 2.49, "learning_rate": 1.422286332491886e-06, "loss": 0.0218, "step": 9625 }, { "epoch": 2.5, "learning_rate": 1.4042553191489362e-06, "loss": 0.0353, "step": 9650 }, { "epoch": 2.5, "learning_rate": 1.386224305805986e-06, "loss": 0.022, "step": 9675 }, { "epoch": 2.51, "learning_rate": 1.3681932924630364e-06, "loss": 0.0531, "step": 9700 }, { "epoch": 2.52, "learning_rate": 1.3501622791200865e-06, "loss": 0.0302, "step": 9725 }, { "epoch": 2.52, "learning_rate": 1.3321312657771366e-06, "loss": 0.0239, "step": 9750 }, { "epoch": 2.53, "learning_rate": 1.3141002524341867e-06, "loss": 0.0207, "step": 9775 }, { "epoch": 2.54, "learning_rate": 1.296069239091237e-06, "loss": 0.0271, "step": 9800 }, { "epoch": 2.54, "learning_rate": 1.2780382257482868e-06, "loss": 0.0245, "step": 9825 }, { "epoch": 2.55, "learning_rate": 1.2600072124053372e-06, "loss": 0.0217, "step": 9850 }, { "epoch": 2.56, "learning_rate": 1.2419761990623872e-06, "loss": 0.0297, "step": 9875 }, { "epoch": 2.56, "learning_rate": 1.2239451857194373e-06, "loss": 0.0313, "step": 9900 }, { "epoch": 2.57, "learning_rate": 1.2059141723764874e-06, "loss": 0.0269, "step": 9925 }, { "epoch": 2.58, "learning_rate": 1.1878831590335377e-06, "loss": 0.0273, "step": 9950 }, { "epoch": 2.58, "learning_rate": 1.1698521456905876e-06, "loss": 0.033, "step": 9975 }, { "epoch": 2.59, "learning_rate": 1.151821132347638e-06, "loss": 0.0333, "step": 10000 }, { "epoch": 2.59, "eval_loss": 0.057692721486091614, "eval_runtime": 4599.0725, "eval_samples_per_second": 2.372, "eval_steps_per_second": 0.297, "eval_wer": 17.71779965160451, "step": 10000 }, { "epoch": 2.59, "learning_rate": 1.133790119004688e-06, "loss": 0.0359, "step": 10025 }, { "epoch": 2.6, "learning_rate": 1.115759105661738e-06, "loss": 0.0413, "step": 10050 }, { "epoch": 2.61, "learning_rate": 1.0977280923187882e-06, "loss": 0.0304, "step": 10075 }, { "epoch": 2.61, "learning_rate": 1.0796970789758385e-06, "loss": 0.0297, "step": 10100 }, { "epoch": 2.62, "learning_rate": 1.0616660656328886e-06, "loss": 0.0365, "step": 10125 }, { "epoch": 2.63, "learning_rate": 1.0436350522899387e-06, "loss": 0.041, "step": 10150 }, { "epoch": 2.63, "learning_rate": 1.025604038946989e-06, "loss": 0.0273, "step": 10175 }, { "epoch": 2.64, "learning_rate": 1.0075730256040389e-06, "loss": 0.0293, "step": 10200 }, { "epoch": 2.65, "learning_rate": 9.89542012261089e-07, "loss": 0.0254, "step": 10225 }, { "epoch": 2.65, "learning_rate": 9.71510998918139e-07, "loss": 0.0233, "step": 10250 }, { "epoch": 2.66, "learning_rate": 9.534799855751892e-07, "loss": 0.0252, "step": 10275 }, { "epoch": 2.67, "learning_rate": 9.354489722322394e-07, "loss": 0.0246, "step": 10300 }, { "epoch": 2.67, "learning_rate": 9.174179588892895e-07, "loss": 0.0196, "step": 10325 }, { "epoch": 2.68, "learning_rate": 8.993869455463396e-07, "loss": 0.0354, "step": 10350 }, { "epoch": 2.69, "learning_rate": 8.813559322033897e-07, "loss": 0.0352, "step": 10375 }, { "epoch": 2.69, "learning_rate": 8.633249188604399e-07, "loss": 0.0279, "step": 10400 }, { "epoch": 2.7, "learning_rate": 8.4529390551749e-07, "loss": 0.0329, "step": 10425 }, { "epoch": 2.7, "learning_rate": 8.272628921745401e-07, "loss": 0.0247, "step": 10450 }, { "epoch": 2.71, "learning_rate": 8.092318788315903e-07, "loss": 0.0222, "step": 10475 }, { "epoch": 2.72, "learning_rate": 7.912008654886404e-07, "loss": 0.0264, "step": 10500 }, { "epoch": 2.72, "eval_loss": 0.05672454833984375, "eval_runtime": 4593.3082, "eval_samples_per_second": 2.375, "eval_steps_per_second": 0.297, "eval_wer": 17.501626545217956, "step": 10500 }, { "epoch": 2.72, "learning_rate": 7.731698521456905e-07, "loss": 0.0232, "step": 10525 }, { "epoch": 2.73, "learning_rate": 7.551388388027407e-07, "loss": 0.0227, "step": 10550 }, { "epoch": 2.74, "learning_rate": 7.371078254597908e-07, "loss": 0.0385, "step": 10575 }, { "epoch": 2.74, "learning_rate": 7.190768121168409e-07, "loss": 0.0323, "step": 10600 }, { "epoch": 2.75, "learning_rate": 7.010457987738911e-07, "loss": 0.0201, "step": 10625 }, { "epoch": 2.76, "learning_rate": 6.830147854309412e-07, "loss": 0.0336, "step": 10650 }, { "epoch": 2.76, "learning_rate": 6.649837720879913e-07, "loss": 0.0348, "step": 10675 }, { "epoch": 2.77, "learning_rate": 6.469527587450415e-07, "loss": 0.0238, "step": 10700 }, { "epoch": 2.78, "learning_rate": 6.289217454020915e-07, "loss": 0.0223, "step": 10725 }, { "epoch": 2.78, "learning_rate": 6.108907320591416e-07, "loss": 0.0294, "step": 10750 }, { "epoch": 2.79, "learning_rate": 5.928597187161918e-07, "loss": 0.0273, "step": 10775 }, { "epoch": 2.8, "learning_rate": 5.748287053732419e-07, "loss": 0.035, "step": 10800 }, { "epoch": 2.8, "learning_rate": 5.56797692030292e-07, "loss": 0.0388, "step": 10825 }, { "epoch": 2.81, "learning_rate": 5.387666786873421e-07, "loss": 0.0218, "step": 10850 }, { "epoch": 2.81, "learning_rate": 5.207356653443923e-07, "loss": 0.0347, "step": 10875 }, { "epoch": 2.82, "learning_rate": 5.027046520014424e-07, "loss": 0.0211, "step": 10900 }, { "epoch": 2.83, "learning_rate": 4.846736386584926e-07, "loss": 0.0265, "step": 10925 }, { "epoch": 2.83, "learning_rate": 4.666426253155427e-07, "loss": 0.0362, "step": 10950 }, { "epoch": 2.84, "learning_rate": 4.4861161197259284e-07, "loss": 0.0253, "step": 10975 }, { "epoch": 2.85, "learning_rate": 4.30580598629643e-07, "loss": 0.0274, "step": 11000 }, { "epoch": 2.85, "eval_loss": 0.055780503898859024, "eval_runtime": 4593.0706, "eval_samples_per_second": 2.375, "eval_steps_per_second": 0.297, "eval_wer": 17.449157344638696, "step": 11000 } ], "max_steps": 11592, "num_train_epochs": 3, "total_flos": 5.078872177311744e+19, "trial_name": null, "trial_params": null }