{ "best_metric": 0.3162029981613159, "best_model_checkpoint": "ai-light-dance_drums_ft_pretrain_wav2vec2-base-new_onset-idmt-2/checkpoint-459", "epoch": 100.0, "global_step": 900, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 1.0, "eval_loss": 97.93186950683594, "eval_runtime": 2.7466, "eval_samples_per_second": 6.19, "eval_steps_per_second": 1.82, "eval_wer": 1.0, "step": 9 }, { "epoch": 1.11, "learning_rate": 2.9999999999999997e-05, "loss": 17.1836, "step": 10 }, { "epoch": 2.0, "eval_loss": 45.722869873046875, "eval_runtime": 2.9189, "eval_samples_per_second": 5.824, "eval_steps_per_second": 1.713, "eval_wer": 1.0, "step": 18 }, { "epoch": 2.22, "learning_rate": 0.00013, "loss": 13.2869, "step": 20 }, { "epoch": 3.0, "eval_loss": 2.757852554321289, "eval_runtime": 2.8561, "eval_samples_per_second": 5.952, "eval_steps_per_second": 1.751, "eval_wer": 1.0, "step": 27 }, { "epoch": 3.33, "learning_rate": 0.00023, "loss": 2.6495, "step": 30 }, { "epoch": 4.0, "eval_loss": 2.7427165508270264, "eval_runtime": 2.8408, "eval_samples_per_second": 5.984, "eval_steps_per_second": 1.76, "eval_wer": 1.0, "step": 36 }, { "epoch": 4.44, "learning_rate": 0.0002989655172413793, "loss": 1.7135, "step": 40 }, { "epoch": 5.0, "eval_loss": 2.5476696491241455, "eval_runtime": 2.8174, "eval_samples_per_second": 6.034, "eval_steps_per_second": 1.775, "eval_wer": 1.0, "step": 45 }, { "epoch": 5.56, "learning_rate": 0.00029551724137931033, "loss": 1.4609, "step": 50 }, { "epoch": 6.0, "eval_loss": 1.7126435041427612, "eval_runtime": 2.9262, "eval_samples_per_second": 5.81, "eval_steps_per_second": 1.709, "eval_wer": 1.0, "step": 54 }, { "epoch": 6.67, "learning_rate": 0.00029206896551724134, "loss": 1.374, "step": 60 }, { "epoch": 7.0, "eval_loss": 1.3668222427368164, "eval_runtime": 2.8538, "eval_samples_per_second": 5.957, "eval_steps_per_second": 1.752, "eval_wer": 0.9966666666666667, "step": 63 }, { "epoch": 7.78, "learning_rate": 0.0002886206896551724, "loss": 1.2951, "step": 70 }, { "epoch": 8.0, "eval_loss": 1.127436637878418, "eval_runtime": 2.8558, "eval_samples_per_second": 5.953, "eval_steps_per_second": 1.751, "eval_wer": 0.9866666666666667, "step": 72 }, { "epoch": 8.89, "learning_rate": 0.00028517241379310345, "loss": 1.0493, "step": 80 }, { "epoch": 9.0, "eval_loss": 0.7346072196960449, "eval_runtime": 3.0416, "eval_samples_per_second": 5.589, "eval_steps_per_second": 1.644, "eval_wer": 0.5177777777777778, "step": 81 }, { "epoch": 10.0, "learning_rate": 0.00028172413793103445, "loss": 0.8835, "step": 90 }, { "epoch": 10.0, "eval_loss": 0.7663962841033936, "eval_runtime": 2.8583, "eval_samples_per_second": 5.947, "eval_steps_per_second": 1.749, "eval_wer": 0.4122222222222222, "step": 90 }, { "epoch": 11.0, "eval_loss": 0.5438075065612793, "eval_runtime": 2.9205, "eval_samples_per_second": 5.821, "eval_steps_per_second": 1.712, "eval_wer": 0.38666666666666666, "step": 99 }, { "epoch": 11.11, "learning_rate": 0.0002782758620689655, "loss": 0.7019, "step": 100 }, { "epoch": 12.0, "eval_loss": 0.4876076281070709, "eval_runtime": 2.8398, "eval_samples_per_second": 5.986, "eval_steps_per_second": 1.761, "eval_wer": 0.3711111111111111, "step": 108 }, { "epoch": 12.22, "learning_rate": 0.0002748275862068965, "loss": 0.6906, "step": 110 }, { "epoch": 13.0, "eval_loss": 0.519428551197052, "eval_runtime": 2.7785, "eval_samples_per_second": 6.119, "eval_steps_per_second": 1.8, "eval_wer": 0.36, "step": 117 }, { "epoch": 13.33, "learning_rate": 0.00027137931034482756, "loss": 0.6535, "step": 120 }, { "epoch": 14.0, "eval_loss": 0.4489333927631378, "eval_runtime": 2.8252, "eval_samples_per_second": 6.017, "eval_steps_per_second": 1.77, "eval_wer": 0.35555555555555557, "step": 126 }, { "epoch": 14.44, "learning_rate": 0.0002679310344827586, "loss": 0.6225, "step": 130 }, { "epoch": 15.0, "eval_loss": 0.4382663071155548, "eval_runtime": 2.7906, "eval_samples_per_second": 6.092, "eval_steps_per_second": 1.792, "eval_wer": 0.3333333333333333, "step": 135 }, { "epoch": 15.56, "learning_rate": 0.0002644827586206896, "loss": 0.547, "step": 140 }, { "epoch": 16.0, "eval_loss": 0.4521160423755646, "eval_runtime": 2.8251, "eval_samples_per_second": 6.017, "eval_steps_per_second": 1.77, "eval_wer": 0.35555555555555557, "step": 144 }, { "epoch": 16.67, "learning_rate": 0.00026103448275862067, "loss": 0.5525, "step": 150 }, { "epoch": 17.0, "eval_loss": 0.5476210117340088, "eval_runtime": 2.8122, "eval_samples_per_second": 6.045, "eval_steps_per_second": 1.778, "eval_wer": 0.33444444444444443, "step": 153 }, { "epoch": 17.78, "learning_rate": 0.0002575862068965517, "loss": 0.6152, "step": 160 }, { "epoch": 18.0, "eval_loss": 0.4465982913970947, "eval_runtime": 2.8461, "eval_samples_per_second": 5.973, "eval_steps_per_second": 1.757, "eval_wer": 0.36, "step": 162 }, { "epoch": 18.89, "learning_rate": 0.00025413793103448273, "loss": 0.5055, "step": 170 }, { "epoch": 19.0, "eval_loss": 0.39812201261520386, "eval_runtime": 2.805, "eval_samples_per_second": 6.061, "eval_steps_per_second": 1.783, "eval_wer": 0.32555555555555554, "step": 171 }, { "epoch": 20.0, "learning_rate": 0.0002506896551724138, "loss": 0.5204, "step": 180 }, { "epoch": 20.0, "eval_loss": 0.4923681318759918, "eval_runtime": 2.8028, "eval_samples_per_second": 6.065, "eval_steps_per_second": 1.784, "eval_wer": 0.30777777777777776, "step": 180 }, { "epoch": 21.0, "eval_loss": 0.4085298776626587, "eval_runtime": 2.8109, "eval_samples_per_second": 6.048, "eval_steps_per_second": 1.779, "eval_wer": 0.32, "step": 189 }, { "epoch": 21.11, "learning_rate": 0.0002472413793103448, "loss": 0.4742, "step": 190 }, { "epoch": 22.0, "eval_loss": 0.4254695177078247, "eval_runtime": 2.7979, "eval_samples_per_second": 6.076, "eval_steps_per_second": 1.787, "eval_wer": 0.3233333333333333, "step": 198 }, { "epoch": 22.22, "learning_rate": 0.00024379310344827584, "loss": 0.4774, "step": 200 }, { "epoch": 23.0, "eval_loss": 0.43211016058921814, "eval_runtime": 2.8119, "eval_samples_per_second": 6.046, "eval_steps_per_second": 1.778, "eval_wer": 0.28888888888888886, "step": 207 }, { "epoch": 23.33, "learning_rate": 0.00024034482758620687, "loss": 0.5029, "step": 210 }, { "epoch": 24.0, "eval_loss": 0.44116583466529846, "eval_runtime": 2.7953, "eval_samples_per_second": 6.082, "eval_steps_per_second": 1.789, "eval_wer": 0.31666666666666665, "step": 216 }, { "epoch": 24.44, "learning_rate": 0.00023689655172413792, "loss": 0.4889, "step": 220 }, { "epoch": 25.0, "eval_loss": 0.40511757135391235, "eval_runtime": 2.8082, "eval_samples_per_second": 6.054, "eval_steps_per_second": 1.781, "eval_wer": 0.30444444444444446, "step": 225 }, { "epoch": 25.56, "learning_rate": 0.00023344827586206895, "loss": 0.4446, "step": 230 }, { "epoch": 26.0, "eval_loss": 0.39176392555236816, "eval_runtime": 2.8027, "eval_samples_per_second": 6.066, "eval_steps_per_second": 1.784, "eval_wer": 0.3088888888888889, "step": 234 }, { "epoch": 26.67, "learning_rate": 0.00023, "loss": 0.4255, "step": 240 }, { "epoch": 27.0, "eval_loss": 0.403872549533844, "eval_runtime": 2.786, "eval_samples_per_second": 6.102, "eval_steps_per_second": 1.795, "eval_wer": 0.29555555555555557, "step": 243 }, { "epoch": 27.78, "learning_rate": 0.000226551724137931, "loss": 0.4396, "step": 250 }, { "epoch": 28.0, "eval_loss": 0.4112667143344879, "eval_runtime": 2.8335, "eval_samples_per_second": 6.0, "eval_steps_per_second": 1.765, "eval_wer": 0.29555555555555557, "step": 252 }, { "epoch": 28.89, "learning_rate": 0.00022310344827586204, "loss": 0.4265, "step": 260 }, { "epoch": 29.0, "eval_loss": 0.5575971603393555, "eval_runtime": 2.7693, "eval_samples_per_second": 6.139, "eval_steps_per_second": 1.806, "eval_wer": 0.3022222222222222, "step": 261 }, { "epoch": 30.0, "learning_rate": 0.0002196551724137931, "loss": 0.4289, "step": 270 }, { "epoch": 30.0, "eval_loss": 0.35578060150146484, "eval_runtime": 2.7632, "eval_samples_per_second": 6.152, "eval_steps_per_second": 1.81, "eval_wer": 0.30777777777777776, "step": 270 }, { "epoch": 31.0, "eval_loss": 0.33904996514320374, "eval_runtime": 2.7428, "eval_samples_per_second": 6.198, "eval_steps_per_second": 1.823, "eval_wer": 0.31666666666666665, "step": 279 }, { "epoch": 31.11, "learning_rate": 0.00021620689655172412, "loss": 0.3817, "step": 280 }, { "epoch": 32.0, "eval_loss": 0.3738517165184021, "eval_runtime": 2.7731, "eval_samples_per_second": 6.13, "eval_steps_per_second": 1.803, "eval_wer": 0.3422222222222222, "step": 288 }, { "epoch": 32.22, "learning_rate": 0.00021275862068965515, "loss": 0.4192, "step": 290 }, { "epoch": 33.0, "eval_loss": 0.31792330741882324, "eval_runtime": 2.7573, "eval_samples_per_second": 6.166, "eval_steps_per_second": 1.813, "eval_wer": 0.3055555555555556, "step": 297 }, { "epoch": 33.33, "learning_rate": 0.0002093103448275862, "loss": 0.3719, "step": 300 }, { "epoch": 34.0, "eval_loss": 0.3621741235256195, "eval_runtime": 2.7351, "eval_samples_per_second": 6.215, "eval_steps_per_second": 1.828, "eval_wer": 0.30333333333333334, "step": 306 }, { "epoch": 34.44, "learning_rate": 0.00020586206896551723, "loss": 0.3685, "step": 310 }, { "epoch": 35.0, "eval_loss": 0.40566930174827576, "eval_runtime": 2.7612, "eval_samples_per_second": 6.157, "eval_steps_per_second": 1.811, "eval_wer": 0.32555555555555554, "step": 315 }, { "epoch": 35.56, "learning_rate": 0.0002024137931034483, "loss": 0.3752, "step": 320 }, { "epoch": 36.0, "eval_loss": 0.39497387409210205, "eval_runtime": 2.7507, "eval_samples_per_second": 6.18, "eval_steps_per_second": 1.818, "eval_wer": 0.31, "step": 324 }, { "epoch": 36.67, "learning_rate": 0.0001989655172413793, "loss": 0.378, "step": 330 }, { "epoch": 37.0, "eval_loss": 0.39072442054748535, "eval_runtime": 2.7439, "eval_samples_per_second": 6.196, "eval_steps_per_second": 1.822, "eval_wer": 0.3566666666666667, "step": 333 }, { "epoch": 37.78, "learning_rate": 0.00019551724137931032, "loss": 0.4438, "step": 340 }, { "epoch": 38.0, "eval_loss": 0.33762815594673157, "eval_runtime": 2.7815, "eval_samples_per_second": 6.112, "eval_steps_per_second": 1.798, "eval_wer": 0.31, "step": 342 }, { "epoch": 38.89, "learning_rate": 0.00019206896551724134, "loss": 0.3978, "step": 350 }, { "epoch": 39.0, "eval_loss": 0.3395032286643982, "eval_runtime": 2.8043, "eval_samples_per_second": 6.062, "eval_steps_per_second": 1.783, "eval_wer": 0.2833333333333333, "step": 351 }, { "epoch": 40.0, "learning_rate": 0.0001886206896551724, "loss": 0.3639, "step": 360 }, { "epoch": 40.0, "eval_loss": 0.36456623673439026, "eval_runtime": 2.7816, "eval_samples_per_second": 6.112, "eval_steps_per_second": 1.798, "eval_wer": 0.28555555555555556, "step": 360 }, { "epoch": 41.0, "eval_loss": 0.3546362519264221, "eval_runtime": 2.7679, "eval_samples_per_second": 6.142, "eval_steps_per_second": 1.806, "eval_wer": 0.30444444444444446, "step": 369 }, { "epoch": 41.11, "learning_rate": 0.00018517241379310343, "loss": 0.3535, "step": 370 }, { "epoch": 42.0, "eval_loss": 0.36989736557006836, "eval_runtime": 2.8116, "eval_samples_per_second": 6.046, "eval_steps_per_second": 1.778, "eval_wer": 0.28888888888888886, "step": 378 }, { "epoch": 42.22, "learning_rate": 0.00018172413793103448, "loss": 0.3311, "step": 380 }, { "epoch": 43.0, "eval_loss": 0.38820379972457886, "eval_runtime": 2.7493, "eval_samples_per_second": 6.183, "eval_steps_per_second": 1.819, "eval_wer": 0.3022222222222222, "step": 387 }, { "epoch": 43.33, "learning_rate": 0.0001782758620689655, "loss": 0.3475, "step": 390 }, { "epoch": 44.0, "eval_loss": 0.4749428331851959, "eval_runtime": 2.8377, "eval_samples_per_second": 5.991, "eval_steps_per_second": 1.762, "eval_wer": 0.28888888888888886, "step": 396 }, { "epoch": 44.44, "learning_rate": 0.00017482758620689654, "loss": 0.4048, "step": 400 }, { "epoch": 45.0, "eval_loss": 0.34369951486587524, "eval_runtime": 2.794, "eval_samples_per_second": 6.085, "eval_steps_per_second": 1.79, "eval_wer": 0.2911111111111111, "step": 405 }, { "epoch": 45.56, "learning_rate": 0.00017137931034482757, "loss": 0.2984, "step": 410 }, { "epoch": 46.0, "eval_loss": 0.36636969447135925, "eval_runtime": 2.7731, "eval_samples_per_second": 6.13, "eval_steps_per_second": 1.803, "eval_wer": 0.27, "step": 414 }, { "epoch": 46.67, "learning_rate": 0.0001679310344827586, "loss": 0.3535, "step": 420 }, { "epoch": 47.0, "eval_loss": 0.3290660083293915, "eval_runtime": 2.767, "eval_samples_per_second": 6.144, "eval_steps_per_second": 1.807, "eval_wer": 0.28888888888888886, "step": 423 }, { "epoch": 47.78, "learning_rate": 0.00016448275862068962, "loss": 0.3015, "step": 430 }, { "epoch": 48.0, "eval_loss": 0.35376498103141785, "eval_runtime": 2.7572, "eval_samples_per_second": 6.166, "eval_steps_per_second": 1.813, "eval_wer": 0.27666666666666667, "step": 432 }, { "epoch": 48.89, "learning_rate": 0.00016103448275862068, "loss": 0.3628, "step": 440 }, { "epoch": 49.0, "eval_loss": 0.44109877943992615, "eval_runtime": 2.7694, "eval_samples_per_second": 6.139, "eval_steps_per_second": 1.805, "eval_wer": 0.2733333333333333, "step": 441 }, { "epoch": 50.0, "learning_rate": 0.0001575862068965517, "loss": 0.3303, "step": 450 }, { "epoch": 50.0, "eval_loss": 0.34246134757995605, "eval_runtime": 2.7763, "eval_samples_per_second": 6.123, "eval_steps_per_second": 1.801, "eval_wer": 0.29, "step": 450 }, { "epoch": 51.0, "eval_loss": 0.3162029981613159, "eval_runtime": 2.7842, "eval_samples_per_second": 6.106, "eval_steps_per_second": 1.796, "eval_wer": 0.3011111111111111, "step": 459 }, { "epoch": 51.11, "learning_rate": 0.00015413793103448276, "loss": 0.271, "step": 460 }, { "epoch": 52.0, "eval_loss": 0.36851903796195984, "eval_runtime": 2.7813, "eval_samples_per_second": 6.112, "eval_steps_per_second": 1.798, "eval_wer": 0.29333333333333333, "step": 468 }, { "epoch": 52.22, "learning_rate": 0.0001506896551724138, "loss": 0.3299, "step": 470 }, { "epoch": 53.0, "eval_loss": 0.42159539461135864, "eval_runtime": 2.767, "eval_samples_per_second": 6.144, "eval_steps_per_second": 1.807, "eval_wer": 0.29333333333333333, "step": 477 }, { "epoch": 53.33, "learning_rate": 0.00014724137931034482, "loss": 0.2782, "step": 480 }, { "epoch": 54.0, "eval_loss": 0.4713245928287506, "eval_runtime": 2.8739, "eval_samples_per_second": 5.915, "eval_steps_per_second": 1.74, "eval_wer": 0.30444444444444446, "step": 486 }, { "epoch": 54.44, "learning_rate": 0.00014379310344827585, "loss": 0.348, "step": 490 }, { "epoch": 55.0, "eval_loss": 0.4310116767883301, "eval_runtime": 2.7991, "eval_samples_per_second": 6.073, "eval_steps_per_second": 1.786, "eval_wer": 0.30777777777777776, "step": 495 }, { "epoch": 55.56, "learning_rate": 0.00014034482758620688, "loss": 0.2969, "step": 500 }, { "epoch": 56.0, "eval_loss": 0.48976629972457886, "eval_runtime": 2.8118, "eval_samples_per_second": 6.046, "eval_steps_per_second": 1.778, "eval_wer": 0.27666666666666667, "step": 504 }, { "epoch": 56.67, "learning_rate": 0.0001368965517241379, "loss": 0.2757, "step": 510 }, { "epoch": 57.0, "eval_loss": 0.5194886326789856, "eval_runtime": 2.7939, "eval_samples_per_second": 6.085, "eval_steps_per_second": 1.79, "eval_wer": 0.2788888888888889, "step": 513 }, { "epoch": 57.78, "learning_rate": 0.00013344827586206896, "loss": 0.2662, "step": 520 }, { "epoch": 58.0, "eval_loss": 0.46313199400901794, "eval_runtime": 2.7984, "eval_samples_per_second": 6.075, "eval_steps_per_second": 1.787, "eval_wer": 0.2911111111111111, "step": 522 }, { "epoch": 58.89, "learning_rate": 0.00013, "loss": 0.2706, "step": 530 }, { "epoch": 59.0, "eval_loss": 0.427517294883728, "eval_runtime": 2.7564, "eval_samples_per_second": 6.168, "eval_steps_per_second": 1.814, "eval_wer": 0.2833333333333333, "step": 531 }, { "epoch": 60.0, "learning_rate": 0.00012655172413793102, "loss": 0.2684, "step": 540 }, { "epoch": 60.0, "eval_loss": 0.553533136844635, "eval_runtime": 2.7742, "eval_samples_per_second": 6.128, "eval_steps_per_second": 1.802, "eval_wer": 0.2788888888888889, "step": 540 }, { "epoch": 61.0, "eval_loss": 0.4733206331729889, "eval_runtime": 2.7693, "eval_samples_per_second": 6.139, "eval_steps_per_second": 1.806, "eval_wer": 0.29777777777777775, "step": 549 }, { "epoch": 61.11, "learning_rate": 0.00012310344827586205, "loss": 0.2819, "step": 550 }, { "epoch": 62.0, "eval_loss": 0.49692198634147644, "eval_runtime": 2.7566, "eval_samples_per_second": 6.167, "eval_steps_per_second": 1.814, "eval_wer": 0.2833333333333333, "step": 558 }, { "epoch": 62.22, "learning_rate": 0.00011965517241379309, "loss": 0.2819, "step": 560 }, { "epoch": 63.0, "eval_loss": 0.6202179789543152, "eval_runtime": 2.7859, "eval_samples_per_second": 6.102, "eval_steps_per_second": 1.795, "eval_wer": 0.2788888888888889, "step": 567 }, { "epoch": 63.33, "learning_rate": 0.00011620689655172413, "loss": 0.2889, "step": 570 }, { "epoch": 64.0, "eval_loss": 0.39546066522598267, "eval_runtime": 2.7637, "eval_samples_per_second": 6.151, "eval_steps_per_second": 1.809, "eval_wer": 0.2733333333333333, "step": 576 }, { "epoch": 64.44, "learning_rate": 0.00011275862068965516, "loss": 0.2515, "step": 580 }, { "epoch": 65.0, "eval_loss": 0.38055384159088135, "eval_runtime": 2.7537, "eval_samples_per_second": 6.173, "eval_steps_per_second": 1.816, "eval_wer": 0.26555555555555554, "step": 585 }, { "epoch": 65.56, "learning_rate": 0.0001093103448275862, "loss": 0.2468, "step": 590 }, { "epoch": 66.0, "eval_loss": 0.3472989499568939, "eval_runtime": 2.8011, "eval_samples_per_second": 6.069, "eval_steps_per_second": 1.785, "eval_wer": 0.2722222222222222, "step": 594 }, { "epoch": 66.67, "learning_rate": 0.00010586206896551723, "loss": 0.2557, "step": 600 }, { "epoch": 67.0, "eval_loss": 0.417023628950119, "eval_runtime": 2.7867, "eval_samples_per_second": 6.1, "eval_steps_per_second": 1.794, "eval_wer": 0.2722222222222222, "step": 603 }, { "epoch": 67.78, "learning_rate": 0.00010241379310344827, "loss": 0.2477, "step": 610 }, { "epoch": 68.0, "eval_loss": 0.4748758375644684, "eval_runtime": 2.7749, "eval_samples_per_second": 6.126, "eval_steps_per_second": 1.802, "eval_wer": 0.2677777777777778, "step": 612 }, { "epoch": 68.89, "learning_rate": 9.89655172413793e-05, "loss": 0.2965, "step": 620 }, { "epoch": 69.0, "eval_loss": 0.438679039478302, "eval_runtime": 2.7924, "eval_samples_per_second": 6.088, "eval_steps_per_second": 1.791, "eval_wer": 0.2611111111111111, "step": 621 }, { "epoch": 70.0, "learning_rate": 9.551724137931034e-05, "loss": 0.2606, "step": 630 }, { "epoch": 70.0, "eval_loss": 0.45858699083328247, "eval_runtime": 2.7888, "eval_samples_per_second": 6.096, "eval_steps_per_second": 1.793, "eval_wer": 0.26555555555555554, "step": 630 }, { "epoch": 71.0, "eval_loss": 0.5755282044410706, "eval_runtime": 2.7561, "eval_samples_per_second": 6.168, "eval_steps_per_second": 1.814, "eval_wer": 0.2733333333333333, "step": 639 }, { "epoch": 71.11, "learning_rate": 9.206896551724137e-05, "loss": 0.2442, "step": 640 }, { "epoch": 72.0, "eval_loss": 0.5582060217857361, "eval_runtime": 2.7794, "eval_samples_per_second": 6.116, "eval_steps_per_second": 1.799, "eval_wer": 0.26555555555555554, "step": 648 }, { "epoch": 72.22, "learning_rate": 8.862068965517241e-05, "loss": 0.347, "step": 650 }, { "epoch": 73.0, "eval_loss": 0.38965609669685364, "eval_runtime": 2.7716, "eval_samples_per_second": 6.134, "eval_steps_per_second": 1.804, "eval_wer": 0.27111111111111114, "step": 657 }, { "epoch": 73.33, "learning_rate": 8.517241379310344e-05, "loss": 0.2444, "step": 660 }, { "epoch": 74.0, "eval_loss": 0.33690622448921204, "eval_runtime": 2.7566, "eval_samples_per_second": 6.167, "eval_steps_per_second": 1.814, "eval_wer": 0.25333333333333335, "step": 666 }, { "epoch": 74.44, "learning_rate": 8.172413793103448e-05, "loss": 0.2811, "step": 670 }, { "epoch": 75.0, "eval_loss": 0.34874993562698364, "eval_runtime": 2.7479, "eval_samples_per_second": 6.186, "eval_steps_per_second": 1.82, "eval_wer": 0.2577777777777778, "step": 675 }, { "epoch": 75.56, "learning_rate": 7.827586206896551e-05, "loss": 0.24, "step": 680 }, { "epoch": 76.0, "eval_loss": 0.3692302107810974, "eval_runtime": 2.785, "eval_samples_per_second": 6.104, "eval_steps_per_second": 1.795, "eval_wer": 0.2588888888888889, "step": 684 }, { "epoch": 76.67, "learning_rate": 7.482758620689654e-05, "loss": 0.2466, "step": 690 }, { "epoch": 77.0, "eval_loss": 0.45672333240509033, "eval_runtime": 2.7533, "eval_samples_per_second": 6.174, "eval_steps_per_second": 1.816, "eval_wer": 0.2577777777777778, "step": 693 }, { "epoch": 77.78, "learning_rate": 7.137931034482758e-05, "loss": 0.2769, "step": 700 }, { "epoch": 78.0, "eval_loss": 0.40405967831611633, "eval_runtime": 2.7964, "eval_samples_per_second": 6.079, "eval_steps_per_second": 1.788, "eval_wer": 0.2633333333333333, "step": 702 }, { "epoch": 78.89, "learning_rate": 6.79310344827586e-05, "loss": 0.2464, "step": 710 }, { "epoch": 79.0, "eval_loss": 0.38128018379211426, "eval_runtime": 2.7812, "eval_samples_per_second": 6.113, "eval_steps_per_second": 1.798, "eval_wer": 0.26222222222222225, "step": 711 }, { "epoch": 80.0, "learning_rate": 6.448275862068965e-05, "loss": 0.2791, "step": 720 }, { "epoch": 80.0, "eval_loss": 0.3990322947502136, "eval_runtime": 2.7516, "eval_samples_per_second": 6.178, "eval_steps_per_second": 1.817, "eval_wer": 0.25555555555555554, "step": 720 }, { "epoch": 81.0, "eval_loss": 0.39965325593948364, "eval_runtime": 2.7649, "eval_samples_per_second": 6.149, "eval_steps_per_second": 1.808, "eval_wer": 0.24888888888888888, "step": 729 }, { "epoch": 81.11, "learning_rate": 6.103448275862068e-05, "loss": 0.2365, "step": 730 }, { "epoch": 82.0, "eval_loss": 0.4537028968334198, "eval_runtime": 2.7885, "eval_samples_per_second": 6.097, "eval_steps_per_second": 1.793, "eval_wer": 0.25333333333333335, "step": 738 }, { "epoch": 82.22, "learning_rate": 5.758620689655172e-05, "loss": 0.2693, "step": 740 }, { "epoch": 83.0, "eval_loss": 0.5943002104759216, "eval_runtime": 2.8023, "eval_samples_per_second": 6.067, "eval_steps_per_second": 1.784, "eval_wer": 0.2611111111111111, "step": 747 }, { "epoch": 83.33, "learning_rate": 5.413793103448275e-05, "loss": 0.2285, "step": 750 }, { "epoch": 84.0, "eval_loss": 0.5804929733276367, "eval_runtime": 2.7598, "eval_samples_per_second": 6.16, "eval_steps_per_second": 1.812, "eval_wer": 0.26555555555555554, "step": 756 }, { "epoch": 84.44, "learning_rate": 5.068965517241379e-05, "loss": 0.2468, "step": 760 }, { "epoch": 85.0, "eval_loss": 0.5608753561973572, "eval_runtime": 2.7731, "eval_samples_per_second": 6.13, "eval_steps_per_second": 1.803, "eval_wer": 0.26555555555555554, "step": 765 }, { "epoch": 85.56, "learning_rate": 4.724137931034482e-05, "loss": 0.2226, "step": 770 }, { "epoch": 86.0, "eval_loss": 0.59482342004776, "eval_runtime": 2.7726, "eval_samples_per_second": 6.131, "eval_steps_per_second": 1.803, "eval_wer": 0.26666666666666666, "step": 774 }, { "epoch": 86.67, "learning_rate": 4.379310344827586e-05, "loss": 0.2419, "step": 780 }, { "epoch": 87.0, "eval_loss": 0.5909682512283325, "eval_runtime": 2.7879, "eval_samples_per_second": 6.098, "eval_steps_per_second": 1.793, "eval_wer": 0.2544444444444444, "step": 783 }, { "epoch": 87.78, "learning_rate": 4.034482758620689e-05, "loss": 0.2254, "step": 790 }, { "epoch": 88.0, "eval_loss": 0.5740683078765869, "eval_runtime": 2.7572, "eval_samples_per_second": 6.166, "eval_steps_per_second": 1.813, "eval_wer": 0.26, "step": 792 }, { "epoch": 88.89, "learning_rate": 3.689655172413793e-05, "loss": 0.2083, "step": 800 }, { "epoch": 89.0, "eval_loss": 0.498391717672348, "eval_runtime": 2.7579, "eval_samples_per_second": 6.164, "eval_steps_per_second": 1.813, "eval_wer": 0.2611111111111111, "step": 801 }, { "epoch": 90.0, "learning_rate": 3.344827586206896e-05, "loss": 0.2318, "step": 810 }, { "epoch": 90.0, "eval_loss": 0.5093002915382385, "eval_runtime": 2.7728, "eval_samples_per_second": 6.131, "eval_steps_per_second": 1.803, "eval_wer": 0.26, "step": 810 }, { "epoch": 91.0, "eval_loss": 0.5284357070922852, "eval_runtime": 2.7447, "eval_samples_per_second": 6.194, "eval_steps_per_second": 1.822, "eval_wer": 0.2633333333333333, "step": 819 }, { "epoch": 91.11, "learning_rate": 2.9999999999999997e-05, "loss": 0.2458, "step": 820 }, { "epoch": 92.0, "eval_loss": 0.48846206068992615, "eval_runtime": 2.7457, "eval_samples_per_second": 6.192, "eval_steps_per_second": 1.821, "eval_wer": 0.26555555555555554, "step": 828 }, { "epoch": 92.22, "learning_rate": 2.6551724137931032e-05, "loss": 0.2394, "step": 830 }, { "epoch": 93.0, "eval_loss": 0.4817904829978943, "eval_runtime": 2.7858, "eval_samples_per_second": 6.102, "eval_steps_per_second": 1.795, "eval_wer": 0.26222222222222225, "step": 837 }, { "epoch": 93.33, "learning_rate": 2.3103448275862067e-05, "loss": 0.2018, "step": 840 }, { "epoch": 94.0, "eval_loss": 0.5037254691123962, "eval_runtime": 2.7345, "eval_samples_per_second": 6.217, "eval_steps_per_second": 1.829, "eval_wer": 0.26, "step": 846 }, { "epoch": 94.44, "learning_rate": 1.9655172413793102e-05, "loss": 0.235, "step": 850 }, { "epoch": 95.0, "eval_loss": 0.5010991096496582, "eval_runtime": 2.8024, "eval_samples_per_second": 6.066, "eval_steps_per_second": 1.784, "eval_wer": 0.2577777777777778, "step": 855 }, { "epoch": 95.56, "learning_rate": 1.6206896551724137e-05, "loss": 0.2252, "step": 860 }, { "epoch": 96.0, "eval_loss": 0.49311307072639465, "eval_runtime": 2.7705, "eval_samples_per_second": 6.136, "eval_steps_per_second": 1.805, "eval_wer": 0.2611111111111111, "step": 864 }, { "epoch": 96.67, "learning_rate": 1.275862068965517e-05, "loss": 0.2147, "step": 870 }, { "epoch": 97.0, "eval_loss": 0.4880674481391907, "eval_runtime": 2.7947, "eval_samples_per_second": 6.083, "eval_steps_per_second": 1.789, "eval_wer": 0.2588888888888889, "step": 873 }, { "epoch": 97.78, "learning_rate": 9.310344827586206e-06, "loss": 0.2227, "step": 880 }, { "epoch": 98.0, "eval_loss": 0.49562954902648926, "eval_runtime": 2.7533, "eval_samples_per_second": 6.174, "eval_steps_per_second": 1.816, "eval_wer": 0.2588888888888889, "step": 882 }, { "epoch": 98.89, "learning_rate": 5.862068965517241e-06, "loss": 0.2168, "step": 890 }, { "epoch": 99.0, "eval_loss": 0.5096976161003113, "eval_runtime": 2.7653, "eval_samples_per_second": 6.148, "eval_steps_per_second": 1.808, "eval_wer": 0.2588888888888889, "step": 891 }, { "epoch": 100.0, "learning_rate": 2.413793103448276e-06, "loss": 0.2282, "step": 900 }, { "epoch": 100.0, "eval_loss": 0.5173905491828918, "eval_runtime": 2.7832, "eval_samples_per_second": 6.108, "eval_steps_per_second": 1.797, "eval_wer": 0.26, "step": 900 }, { "epoch": 100.0, "step": 900, "total_flos": 1.389987159899058e+18, "train_loss": 0.7617763585514492, "train_runtime": 3519.9031, "train_samples_per_second": 4.006, "train_steps_per_second": 0.256 } ], "max_steps": 900, "num_train_epochs": 100, "total_flos": 1.389987159899058e+18, "trial_name": null, "trial_params": null }