{ "best_metric": 13.225122198653509, "best_model_checkpoint": "whisper-small-id/checkpoint-1500", "epoch": 49.504950495049506, "global_step": 5000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.25, "learning_rate": 4.4e-07, "loss": 1.4583, "step": 25 }, { "epoch": 0.5, "learning_rate": 9.400000000000001e-07, "loss": 1.2387, "step": 50 }, { "epoch": 0.74, "learning_rate": 1.44e-06, "loss": 0.8884, "step": 75 }, { "epoch": 0.99, "learning_rate": 1.94e-06, "loss": 0.5379, "step": 100 }, { "epoch": 1.24, "learning_rate": 2.4400000000000004e-06, "loss": 0.4345, "step": 125 }, { "epoch": 1.49, "learning_rate": 2.9400000000000002e-06, "loss": 0.3926, "step": 150 }, { "epoch": 1.73, "learning_rate": 3.44e-06, "loss": 0.3744, "step": 175 }, { "epoch": 1.98, "learning_rate": 3.94e-06, "loss": 0.3637, "step": 200 }, { "epoch": 2.23, "learning_rate": 4.440000000000001e-06, "loss": 0.2892, "step": 225 }, { "epoch": 2.48, "learning_rate": 4.94e-06, "loss": 0.2744, "step": 250 }, { "epoch": 2.72, "learning_rate": 5.4400000000000004e-06, "loss": 0.2732, "step": 275 }, { "epoch": 2.97, "learning_rate": 5.94e-06, "loss": 0.2546, "step": 300 }, { "epoch": 3.22, "learning_rate": 6.440000000000001e-06, "loss": 0.1907, "step": 325 }, { "epoch": 3.47, "learning_rate": 6.9400000000000005e-06, "loss": 0.1771, "step": 350 }, { "epoch": 3.71, "learning_rate": 7.440000000000001e-06, "loss": 0.1757, "step": 375 }, { "epoch": 3.96, "learning_rate": 7.94e-06, "loss": 0.1726, "step": 400 }, { "epoch": 4.21, "learning_rate": 8.44e-06, "loss": 0.1089, "step": 425 }, { "epoch": 4.46, "learning_rate": 8.94e-06, "loss": 0.0957, "step": 450 }, { "epoch": 4.7, "learning_rate": 9.440000000000001e-06, "loss": 0.1016, "step": 475 }, { "epoch": 4.95, "learning_rate": 9.940000000000001e-06, "loss": 0.1014, "step": 500 }, { "epoch": 4.95, "eval_loss": 0.2583056092262268, "eval_runtime": 689.8203, "eval_samples_per_second": 5.245, "eval_steps_per_second": 0.165, "eval_wer": 13.63552522364659, "step": 500 }, { "epoch": 5.2, "learning_rate": 9.951111111111111e-06, "loss": 0.0641, "step": 525 }, { "epoch": 5.45, "learning_rate": 9.895555555555557e-06, "loss": 0.0477, "step": 550 }, { "epoch": 5.69, "learning_rate": 9.84e-06, "loss": 0.0528, "step": 575 }, { "epoch": 5.94, "learning_rate": 9.784444444444445e-06, "loss": 0.05, "step": 600 }, { "epoch": 6.19, "learning_rate": 9.72888888888889e-06, "loss": 0.0353, "step": 625 }, { "epoch": 6.44, "learning_rate": 9.673333333333334e-06, "loss": 0.0259, "step": 650 }, { "epoch": 6.68, "learning_rate": 9.617777777777778e-06, "loss": 0.0299, "step": 675 }, { "epoch": 6.93, "learning_rate": 9.562222222222223e-06, "loss": 0.0255, "step": 700 }, { "epoch": 7.18, "learning_rate": 9.506666666666667e-06, "loss": 0.0163, "step": 725 }, { "epoch": 7.43, "learning_rate": 9.451111111111112e-06, "loss": 0.0136, "step": 750 }, { "epoch": 7.67, "learning_rate": 9.395555555555556e-06, "loss": 0.0142, "step": 775 }, { "epoch": 7.92, "learning_rate": 9.340000000000002e-06, "loss": 0.0154, "step": 800 }, { "epoch": 8.17, "learning_rate": 9.284444444444444e-06, "loss": 0.0109, "step": 825 }, { "epoch": 8.42, "learning_rate": 9.22888888888889e-06, "loss": 0.0105, "step": 850 }, { "epoch": 8.66, "learning_rate": 9.173333333333334e-06, "loss": 0.0085, "step": 875 }, { "epoch": 8.91, "learning_rate": 9.117777777777778e-06, "loss": 0.0087, "step": 900 }, { "epoch": 9.16, "learning_rate": 9.062222222222224e-06, "loss": 0.0078, "step": 925 }, { "epoch": 9.41, "learning_rate": 9.006666666666666e-06, "loss": 0.0062, "step": 950 }, { "epoch": 9.65, "learning_rate": 8.951111111111112e-06, "loss": 0.0058, "step": 975 }, { "epoch": 9.9, "learning_rate": 8.895555555555556e-06, "loss": 0.0058, "step": 1000 }, { "epoch": 9.9, "eval_loss": 0.31691908836364746, "eval_runtime": 691.2286, "eval_samples_per_second": 5.234, "eval_steps_per_second": 0.165, "eval_wer": 13.285068707922163, "step": 1000 }, { "epoch": 10.15, "learning_rate": 8.84e-06, "loss": 0.0056, "step": 1025 }, { "epoch": 10.4, "learning_rate": 8.784444444444446e-06, "loss": 0.0041, "step": 1050 }, { "epoch": 10.64, "learning_rate": 8.72888888888889e-06, "loss": 0.0041, "step": 1075 }, { "epoch": 10.89, "learning_rate": 8.673333333333334e-06, "loss": 0.0043, "step": 1100 }, { "epoch": 11.14, "learning_rate": 8.617777777777778e-06, "loss": 0.0038, "step": 1125 }, { "epoch": 11.39, "learning_rate": 8.562222222222224e-06, "loss": 0.0031, "step": 1150 }, { "epoch": 11.63, "learning_rate": 8.506666666666668e-06, "loss": 0.0038, "step": 1175 }, { "epoch": 11.88, "learning_rate": 8.451111111111112e-06, "loss": 0.0031, "step": 1200 }, { "epoch": 12.13, "learning_rate": 8.395555555555557e-06, "loss": 0.003, "step": 1225 }, { "epoch": 12.38, "learning_rate": 8.34e-06, "loss": 0.0026, "step": 1250 }, { "epoch": 12.62, "learning_rate": 8.284444444444446e-06, "loss": 0.0024, "step": 1275 }, { "epoch": 12.87, "learning_rate": 8.22888888888889e-06, "loss": 0.0023, "step": 1300 }, { "epoch": 13.12, "learning_rate": 8.173333333333334e-06, "loss": 0.0022, "step": 1325 }, { "epoch": 13.37, "learning_rate": 8.11777777777778e-06, "loss": 0.0019, "step": 1350 }, { "epoch": 13.61, "learning_rate": 8.062222222222222e-06, "loss": 0.0019, "step": 1375 }, { "epoch": 13.86, "learning_rate": 8.006666666666667e-06, "loss": 0.0022, "step": 1400 }, { "epoch": 14.11, "learning_rate": 7.951111111111111e-06, "loss": 0.0018, "step": 1425 }, { "epoch": 14.36, "learning_rate": 7.895555555555557e-06, "loss": 0.0016, "step": 1450 }, { "epoch": 14.6, "learning_rate": 7.840000000000001e-06, "loss": 0.0017, "step": 1475 }, { "epoch": 14.85, "learning_rate": 7.784444444444445e-06, "loss": 0.0017, "step": 1500 }, { "epoch": 14.85, "eval_loss": 0.34876465797424316, "eval_runtime": 690.0853, "eval_samples_per_second": 5.243, "eval_steps_per_second": 0.165, "eval_wer": 13.225122198653509, "step": 1500 }, { "epoch": 15.1, "learning_rate": 7.72888888888889e-06, "loss": 0.0016, "step": 1525 }, { "epoch": 15.35, "learning_rate": 7.673333333333333e-06, "loss": 0.0017, "step": 1550 }, { "epoch": 15.59, "learning_rate": 7.617777777777778e-06, "loss": 0.0015, "step": 1575 }, { "epoch": 15.84, "learning_rate": 7.562222222222223e-06, "loss": 0.0014, "step": 1600 }, { "epoch": 16.09, "learning_rate": 7.506666666666668e-06, "loss": 0.0014, "step": 1625 }, { "epoch": 16.34, "learning_rate": 7.451111111111111e-06, "loss": 0.0013, "step": 1650 }, { "epoch": 16.58, "learning_rate": 7.395555555555556e-06, "loss": 0.0013, "step": 1675 }, { "epoch": 16.83, "learning_rate": 7.340000000000001e-06, "loss": 0.0017, "step": 1700 }, { "epoch": 17.08, "learning_rate": 7.284444444444445e-06, "loss": 0.0013, "step": 1725 }, { "epoch": 17.33, "learning_rate": 7.22888888888889e-06, "loss": 0.0012, "step": 1750 }, { "epoch": 17.57, "learning_rate": 7.173333333333335e-06, "loss": 0.0012, "step": 1775 }, { "epoch": 17.82, "learning_rate": 7.117777777777778e-06, "loss": 0.0012, "step": 1800 }, { "epoch": 18.07, "learning_rate": 7.062222222222223e-06, "loss": 0.0013, "step": 1825 }, { "epoch": 18.32, "learning_rate": 7.006666666666667e-06, "loss": 0.0012, "step": 1850 }, { "epoch": 18.56, "learning_rate": 6.951111111111112e-06, "loss": 0.0011, "step": 1875 }, { "epoch": 18.81, "learning_rate": 6.8955555555555565e-06, "loss": 0.0011, "step": 1900 }, { "epoch": 19.06, "learning_rate": 6.8400000000000014e-06, "loss": 0.0011, "step": 1925 }, { "epoch": 19.31, "learning_rate": 6.784444444444445e-06, "loss": 0.001, "step": 1950 }, { "epoch": 19.55, "learning_rate": 6.7288888888888895e-06, "loss": 0.001, "step": 1975 }, { "epoch": 19.8, "learning_rate": 6.6733333333333335e-06, "loss": 0.001, "step": 2000 }, { "epoch": 19.8, "eval_loss": 0.3639397919178009, "eval_runtime": 690.6306, "eval_samples_per_second": 5.239, "eval_steps_per_second": 0.165, "eval_wer": 13.3542377570783, "step": 2000 }, { "epoch": 20.05, "learning_rate": 6.617777777777778e-06, "loss": 0.0012, "step": 2025 }, { "epoch": 20.3, "learning_rate": 6.562222222222223e-06, "loss": 0.0009, "step": 2050 }, { "epoch": 20.54, "learning_rate": 6.5066666666666665e-06, "loss": 0.0009, "step": 2075 }, { "epoch": 20.79, "learning_rate": 6.451111111111111e-06, "loss": 0.0009, "step": 2100 }, { "epoch": 21.04, "learning_rate": 6.395555555555556e-06, "loss": 0.0009, "step": 2125 }, { "epoch": 21.29, "learning_rate": 6.34e-06, "loss": 0.0009, "step": 2150 }, { "epoch": 21.53, "learning_rate": 6.284444444444445e-06, "loss": 0.0008, "step": 2175 }, { "epoch": 21.78, "learning_rate": 6.22888888888889e-06, "loss": 0.0008, "step": 2200 }, { "epoch": 22.03, "learning_rate": 6.173333333333333e-06, "loss": 0.0009, "step": 2225 }, { "epoch": 22.28, "learning_rate": 6.117777777777778e-06, "loss": 0.0008, "step": 2250 }, { "epoch": 22.52, "learning_rate": 6.062222222222223e-06, "loss": 0.0008, "step": 2275 }, { "epoch": 22.77, "learning_rate": 6.006666666666667e-06, "loss": 0.0008, "step": 2300 }, { "epoch": 23.02, "learning_rate": 5.951111111111112e-06, "loss": 0.0008, "step": 2325 }, { "epoch": 23.27, "learning_rate": 5.895555555555557e-06, "loss": 0.0007, "step": 2350 }, { "epoch": 23.51, "learning_rate": 5.84e-06, "loss": 0.0007, "step": 2375 }, { "epoch": 23.76, "learning_rate": 5.784444444444445e-06, "loss": 0.0007, "step": 2400 }, { "epoch": 24.01, "learning_rate": 5.72888888888889e-06, "loss": 0.0007, "step": 2425 }, { "epoch": 24.26, "learning_rate": 5.673333333333334e-06, "loss": 0.0007, "step": 2450 }, { "epoch": 24.5, "learning_rate": 5.617777777777779e-06, "loss": 0.0007, "step": 2475 }, { "epoch": 24.75, "learning_rate": 5.562222222222222e-06, "loss": 0.0007, "step": 2500 }, { "epoch": 24.75, "eval_loss": 0.3756162226200104, "eval_runtime": 691.1293, "eval_samples_per_second": 5.235, "eval_steps_per_second": 0.165, "eval_wer": 13.501798395278058, "step": 2500 }, { "epoch": 25.0, "learning_rate": 5.506666666666667e-06, "loss": 0.0007, "step": 2525 }, { "epoch": 25.25, "learning_rate": 5.451111111111112e-06, "loss": 0.0006, "step": 2550 }, { "epoch": 25.5, "learning_rate": 5.3955555555555565e-06, "loss": 0.0007, "step": 2575 }, { "epoch": 25.74, "learning_rate": 5.3400000000000005e-06, "loss": 0.0006, "step": 2600 }, { "epoch": 25.99, "learning_rate": 5.2844444444444454e-06, "loss": 0.0006, "step": 2625 }, { "epoch": 26.24, "learning_rate": 5.228888888888889e-06, "loss": 0.0006, "step": 2650 }, { "epoch": 26.49, "learning_rate": 5.1733333333333335e-06, "loss": 0.0006, "step": 2675 }, { "epoch": 26.73, "learning_rate": 5.117777777777778e-06, "loss": 0.0006, "step": 2700 }, { "epoch": 26.98, "learning_rate": 5.062222222222222e-06, "loss": 0.0006, "step": 2725 }, { "epoch": 27.23, "learning_rate": 5.006666666666667e-06, "loss": 0.0006, "step": 2750 }, { "epoch": 27.48, "learning_rate": 4.951111111111111e-06, "loss": 0.0006, "step": 2775 }, { "epoch": 27.72, "learning_rate": 4.895555555555556e-06, "loss": 0.0006, "step": 2800 }, { "epoch": 27.97, "learning_rate": 4.84e-06, "loss": 0.0006, "step": 2825 }, { "epoch": 28.22, "learning_rate": 4.784444444444445e-06, "loss": 0.0006, "step": 2850 }, { "epoch": 28.47, "learning_rate": 4.728888888888889e-06, "loss": 0.0005, "step": 2875 }, { "epoch": 28.71, "learning_rate": 4.673333333333333e-06, "loss": 0.0005, "step": 2900 }, { "epoch": 28.96, "learning_rate": 4.617777777777778e-06, "loss": 0.0008, "step": 2925 }, { "epoch": 29.21, "learning_rate": 4.562222222222222e-06, "loss": 0.0008, "step": 2950 }, { "epoch": 29.46, "learning_rate": 4.506666666666667e-06, "loss": 0.0005, "step": 2975 }, { "epoch": 29.7, "learning_rate": 4.451111111111112e-06, "loss": 0.0005, "step": 3000 }, { "epoch": 29.7, "eval_loss": 0.3843989372253418, "eval_runtime": 689.531, "eval_samples_per_second": 5.247, "eval_steps_per_second": 0.165, "eval_wer": 13.561744904546714, "step": 3000 }, { "epoch": 29.95, "learning_rate": 4.395555555555556e-06, "loss": 0.0005, "step": 3025 }, { "epoch": 30.2, "learning_rate": 4.34e-06, "loss": 0.0005, "step": 3050 }, { "epoch": 30.45, "learning_rate": 4.284444444444445e-06, "loss": 0.0005, "step": 3075 }, { "epoch": 30.69, "learning_rate": 4.228888888888889e-06, "loss": 0.0006, "step": 3100 }, { "epoch": 30.94, "learning_rate": 4.173333333333334e-06, "loss": 0.0005, "step": 3125 }, { "epoch": 31.19, "learning_rate": 4.117777777777779e-06, "loss": 0.0005, "step": 3150 }, { "epoch": 31.44, "learning_rate": 4.062222222222223e-06, "loss": 0.0005, "step": 3175 }, { "epoch": 31.68, "learning_rate": 4.006666666666667e-06, "loss": 0.0005, "step": 3200 }, { "epoch": 31.93, "learning_rate": 3.951111111111112e-06, "loss": 0.0005, "step": 3225 }, { "epoch": 32.18, "learning_rate": 3.895555555555556e-06, "loss": 0.0005, "step": 3250 }, { "epoch": 32.43, "learning_rate": 3.8400000000000005e-06, "loss": 0.0005, "step": 3275 }, { "epoch": 32.67, "learning_rate": 3.784444444444445e-06, "loss": 0.0004, "step": 3300 }, { "epoch": 32.92, "learning_rate": 3.728888888888889e-06, "loss": 0.0004, "step": 3325 }, { "epoch": 33.17, "learning_rate": 3.673333333333334e-06, "loss": 0.0004, "step": 3350 }, { "epoch": 33.42, "learning_rate": 3.617777777777778e-06, "loss": 0.0004, "step": 3375 }, { "epoch": 33.66, "learning_rate": 3.5622222222222224e-06, "loss": 0.0005, "step": 3400 }, { "epoch": 33.91, "learning_rate": 3.5066666666666673e-06, "loss": 0.0004, "step": 3425 }, { "epoch": 34.16, "learning_rate": 3.4511111111111113e-06, "loss": 0.0004, "step": 3450 }, { "epoch": 34.41, "learning_rate": 3.3955555555555558e-06, "loss": 0.0004, "step": 3475 }, { "epoch": 34.65, "learning_rate": 3.3400000000000006e-06, "loss": 0.0005, "step": 3500 }, { "epoch": 34.65, "eval_loss": 0.39218270778656006, "eval_runtime": 689.0125, "eval_samples_per_second": 5.251, "eval_steps_per_second": 0.165, "eval_wer": 13.640136493590335, "step": 3500 }, { "epoch": 34.9, "learning_rate": 3.2844444444444447e-06, "loss": 0.0004, "step": 3525 }, { "epoch": 35.15, "learning_rate": 3.228888888888889e-06, "loss": 0.0004, "step": 3550 }, { "epoch": 35.4, "learning_rate": 3.173333333333334e-06, "loss": 0.0004, "step": 3575 }, { "epoch": 35.64, "learning_rate": 3.117777777777778e-06, "loss": 0.0004, "step": 3600 }, { "epoch": 35.89, "learning_rate": 3.0622222222222225e-06, "loss": 0.0004, "step": 3625 }, { "epoch": 36.14, "learning_rate": 3.0066666666666674e-06, "loss": 0.0004, "step": 3650 }, { "epoch": 36.39, "learning_rate": 2.9511111111111114e-06, "loss": 0.0004, "step": 3675 }, { "epoch": 36.63, "learning_rate": 2.895555555555556e-06, "loss": 0.0004, "step": 3700 }, { "epoch": 36.88, "learning_rate": 2.84e-06, "loss": 0.0004, "step": 3725 }, { "epoch": 37.13, "learning_rate": 2.784444444444445e-06, "loss": 0.0004, "step": 3750 }, { "epoch": 37.38, "learning_rate": 2.7288888888888893e-06, "loss": 0.0004, "step": 3775 }, { "epoch": 37.62, "learning_rate": 2.6733333333333333e-06, "loss": 0.0004, "step": 3800 }, { "epoch": 37.87, "learning_rate": 2.617777777777778e-06, "loss": 0.0004, "step": 3825 }, { "epoch": 38.12, "learning_rate": 2.5622222222222226e-06, "loss": 0.0004, "step": 3850 }, { "epoch": 38.37, "learning_rate": 2.5066666666666667e-06, "loss": 0.0004, "step": 3875 }, { "epoch": 38.61, "learning_rate": 2.451111111111111e-06, "loss": 0.0004, "step": 3900 }, { "epoch": 38.86, "learning_rate": 2.3955555555555556e-06, "loss": 0.0004, "step": 3925 }, { "epoch": 39.11, "learning_rate": 2.3400000000000005e-06, "loss": 0.0004, "step": 3950 }, { "epoch": 39.36, "learning_rate": 2.2844444444444445e-06, "loss": 0.0004, "step": 3975 }, { "epoch": 39.6, "learning_rate": 2.228888888888889e-06, "loss": 0.0004, "step": 4000 }, { "epoch": 39.6, "eval_loss": 0.3981035053730011, "eval_runtime": 688.9924, "eval_samples_per_second": 5.251, "eval_steps_per_second": 0.165, "eval_wer": 13.603246334040394, "step": 4000 }, { "epoch": 39.85, "learning_rate": 2.1733333333333334e-06, "loss": 0.0003, "step": 4025 }, { "epoch": 40.1, "learning_rate": 2.117777777777778e-06, "loss": 0.0003, "step": 4050 }, { "epoch": 40.35, "learning_rate": 2.0622222222222223e-06, "loss": 0.0003, "step": 4075 }, { "epoch": 40.59, "learning_rate": 2.006666666666667e-06, "loss": 0.0003, "step": 4100 }, { "epoch": 40.84, "learning_rate": 1.9511111111111113e-06, "loss": 0.0003, "step": 4125 }, { "epoch": 41.09, "learning_rate": 1.8955555555555557e-06, "loss": 0.0003, "step": 4150 }, { "epoch": 41.34, "learning_rate": 1.8400000000000002e-06, "loss": 0.0003, "step": 4175 }, { "epoch": 41.58, "learning_rate": 1.7844444444444444e-06, "loss": 0.0003, "step": 4200 }, { "epoch": 41.83, "learning_rate": 1.728888888888889e-06, "loss": 0.0003, "step": 4225 }, { "epoch": 42.08, "learning_rate": 1.6733333333333335e-06, "loss": 0.0003, "step": 4250 }, { "epoch": 42.33, "learning_rate": 1.6177777777777778e-06, "loss": 0.0003, "step": 4275 }, { "epoch": 42.57, "learning_rate": 1.5622222222222225e-06, "loss": 0.0003, "step": 4300 }, { "epoch": 42.82, "learning_rate": 1.506666666666667e-06, "loss": 0.0003, "step": 4325 }, { "epoch": 43.07, "learning_rate": 1.4511111111111112e-06, "loss": 0.0003, "step": 4350 }, { "epoch": 43.32, "learning_rate": 1.3955555555555556e-06, "loss": 0.0003, "step": 4375 }, { "epoch": 43.56, "learning_rate": 1.34e-06, "loss": 0.0003, "step": 4400 }, { "epoch": 43.81, "learning_rate": 1.2844444444444445e-06, "loss": 0.0003, "step": 4425 }, { "epoch": 44.06, "learning_rate": 1.228888888888889e-06, "loss": 0.0003, "step": 4450 }, { "epoch": 44.31, "learning_rate": 1.1733333333333335e-06, "loss": 0.0003, "step": 4475 }, { "epoch": 44.55, "learning_rate": 1.117777777777778e-06, "loss": 0.0003, "step": 4500 }, { "epoch": 44.55, "eval_loss": 0.40189599990844727, "eval_runtime": 688.8345, "eval_samples_per_second": 5.252, "eval_steps_per_second": 0.165, "eval_wer": 13.663192843309046, "step": 4500 }, { "epoch": 44.8, "learning_rate": 1.0622222222222222e-06, "loss": 0.0003, "step": 4525 }, { "epoch": 45.05, "learning_rate": 1.0066666666666668e-06, "loss": 0.0003, "step": 4550 }, { "epoch": 45.3, "learning_rate": 9.511111111111111e-07, "loss": 0.0003, "step": 4575 }, { "epoch": 45.54, "learning_rate": 8.955555555555557e-07, "loss": 0.0003, "step": 4600 }, { "epoch": 45.79, "learning_rate": 8.400000000000001e-07, "loss": 0.0003, "step": 4625 }, { "epoch": 46.04, "learning_rate": 7.844444444444445e-07, "loss": 0.0003, "step": 4650 }, { "epoch": 46.29, "learning_rate": 7.28888888888889e-07, "loss": 0.0003, "step": 4675 }, { "epoch": 46.53, "learning_rate": 6.733333333333334e-07, "loss": 0.0003, "step": 4700 }, { "epoch": 46.78, "learning_rate": 6.177777777777778e-07, "loss": 0.0003, "step": 4725 }, { "epoch": 47.03, "learning_rate": 5.622222222222223e-07, "loss": 0.0003, "step": 4750 }, { "epoch": 47.28, "learning_rate": 5.066666666666667e-07, "loss": 0.0003, "step": 4775 }, { "epoch": 47.52, "learning_rate": 4.511111111111111e-07, "loss": 0.0003, "step": 4800 }, { "epoch": 47.77, "learning_rate": 3.9555555555555557e-07, "loss": 0.0003, "step": 4825 }, { "epoch": 48.02, "learning_rate": 3.4000000000000003e-07, "loss": 0.0003, "step": 4850 }, { "epoch": 48.27, "learning_rate": 2.844444444444445e-07, "loss": 0.0003, "step": 4875 }, { "epoch": 48.51, "learning_rate": 2.2888888888888892e-07, "loss": 0.0003, "step": 4900 }, { "epoch": 48.76, "learning_rate": 1.7333333333333335e-07, "loss": 0.0003, "step": 4925 }, { "epoch": 49.01, "learning_rate": 1.1777777777777778e-07, "loss": 0.0003, "step": 4950 }, { "epoch": 49.26, "learning_rate": 6.222222222222223e-08, "loss": 0.0003, "step": 4975 }, { "epoch": 49.5, "learning_rate": 6.666666666666667e-09, "loss": 0.0003, "step": 5000 }, { "epoch": 49.5, "eval_loss": 0.4034419059753418, "eval_runtime": 687.9591, "eval_samples_per_second": 5.259, "eval_steps_per_second": 0.166, "eval_wer": 13.64935903347782, "step": 5000 }, { "epoch": 49.5, "step": 5000, "total_flos": 9.217763988701184e+19, "train_loss": 0.042498688932601364, "train_runtime": 25239.7616, "train_samples_per_second": 12.678, "train_steps_per_second": 0.198 } ], "max_steps": 5000, "num_train_epochs": 50, "total_flos": 9.217763988701184e+19, "trial_name": null, "trial_params": null }