| { | |
| "best_metric": 0.1746283918619156, | |
| "best_model_checkpoint": "wavlm-english-phoneme/checkpoint-6500", | |
| "epoch": 50.0, | |
| "global_step": 6500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 2.97e-05, | |
| "loss": 4.896, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "learning_rate": 5.97e-05, | |
| "loss": 2.5812, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "learning_rate": 8.969999999999998e-05, | |
| "loss": 2.4853, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.08, | |
| "learning_rate": 0.0001197, | |
| "loss": 2.2031, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.85, | |
| "learning_rate": 0.00014969999999999998, | |
| "loss": 1.8587, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.85, | |
| "eval_cer": 0.7105220883534137, | |
| "eval_loss": 1.3614039421081543, | |
| "eval_runtime": 21.402, | |
| "eval_samples_per_second": 21.587, | |
| "eval_steps_per_second": 2.71, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.62, | |
| "learning_rate": 0.00017969999999999998, | |
| "loss": 1.1164, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 5.38, | |
| "learning_rate": 0.00020969999999999997, | |
| "loss": 0.7271, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 6.15, | |
| "learning_rate": 0.0002397, | |
| "loss": 0.5784, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 6.92, | |
| "learning_rate": 0.0002697, | |
| "loss": 0.4487, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 7.69, | |
| "learning_rate": 0.00029969999999999997, | |
| "loss": 0.3517, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 7.69, | |
| "eval_cer": 0.16610441767068274, | |
| "eval_loss": 0.3352901339530945, | |
| "eval_runtime": 21.5161, | |
| "eval_samples_per_second": 21.472, | |
| "eval_steps_per_second": 2.696, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 8.46, | |
| "learning_rate": 0.00029459999999999995, | |
| "loss": 0.3089, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 9.23, | |
| "learning_rate": 0.00028914545454545454, | |
| "loss": 0.2465, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 0.00028369090909090907, | |
| "loss": 0.2099, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 10.77, | |
| "learning_rate": 0.0002782363636363636, | |
| "loss": 0.1649, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 11.54, | |
| "learning_rate": 0.00027278181818181813, | |
| "loss": 0.1473, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 11.54, | |
| "eval_cer": 0.11325301204819277, | |
| "eval_loss": 0.23726579546928406, | |
| "eval_runtime": 34.3746, | |
| "eval_samples_per_second": 13.44, | |
| "eval_steps_per_second": 1.687, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 12.31, | |
| "learning_rate": 0.0002673272727272727, | |
| "loss": 0.1335, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 13.08, | |
| "learning_rate": 0.00026187272727272725, | |
| "loss": 0.118, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 13.85, | |
| "learning_rate": 0.00025641818181818183, | |
| "loss": 0.1083, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 14.62, | |
| "learning_rate": 0.00025096363636363636, | |
| "loss": 0.106, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 15.38, | |
| "learning_rate": 0.0002455090909090909, | |
| "loss": 0.0799, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 15.38, | |
| "eval_cer": 0.09397590361445783, | |
| "eval_loss": 0.20781992375850677, | |
| "eval_runtime": 21.3113, | |
| "eval_samples_per_second": 21.679, | |
| "eval_steps_per_second": 2.722, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 16.15, | |
| "learning_rate": 0.00024005454545454545, | |
| "loss": 0.0767, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 16.92, | |
| "learning_rate": 0.00023459999999999998, | |
| "loss": 0.0761, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 17.69, | |
| "learning_rate": 0.00022914545454545454, | |
| "loss": 0.0705, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 18.46, | |
| "learning_rate": 0.00022369090909090907, | |
| "loss": 0.0619, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 19.23, | |
| "learning_rate": 0.00021823636363636363, | |
| "loss": 0.0634, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 19.23, | |
| "eval_cer": 0.08538152610441767, | |
| "eval_loss": 0.20878814160823822, | |
| "eval_runtime": 20.9398, | |
| "eval_samples_per_second": 22.063, | |
| "eval_steps_per_second": 2.77, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "learning_rate": 0.00021278181818181816, | |
| "loss": 0.0556, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 20.77, | |
| "learning_rate": 0.00020732727272727272, | |
| "loss": 0.0562, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 21.54, | |
| "learning_rate": 0.00020187272727272725, | |
| "loss": 0.0556, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 22.31, | |
| "learning_rate": 0.0001964181818181818, | |
| "loss": 0.0483, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 23.08, | |
| "learning_rate": 0.00019096363636363634, | |
| "loss": 0.0472, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 23.08, | |
| "eval_cer": 0.0767871485943775, | |
| "eval_loss": 0.18513090908527374, | |
| "eval_runtime": 21.0154, | |
| "eval_samples_per_second": 21.984, | |
| "eval_steps_per_second": 2.76, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 23.85, | |
| "learning_rate": 0.00018550909090909087, | |
| "loss": 0.044, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 24.62, | |
| "learning_rate": 0.00018005454545454543, | |
| "loss": 0.0409, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 25.38, | |
| "learning_rate": 0.00017459999999999996, | |
| "loss": 0.0386, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 26.15, | |
| "learning_rate": 0.00016914545454545452, | |
| "loss": 0.0397, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 26.92, | |
| "learning_rate": 0.00016369090909090905, | |
| "loss": 0.0337, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 26.92, | |
| "eval_cer": 0.0757429718875502, | |
| "eval_loss": 0.2069997787475586, | |
| "eval_runtime": 21.3619, | |
| "eval_samples_per_second": 21.627, | |
| "eval_steps_per_second": 2.715, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 27.69, | |
| "learning_rate": 0.00015823636363636364, | |
| "loss": 0.0368, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 28.46, | |
| "learning_rate": 0.0001527818181818182, | |
| "loss": 0.0329, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 29.23, | |
| "learning_rate": 0.0001473272727272727, | |
| "loss": 0.0306, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "learning_rate": 0.00014187272727272726, | |
| "loss": 0.0298, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 30.77, | |
| "learning_rate": 0.0001364181818181818, | |
| "loss": 0.0283, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 30.77, | |
| "eval_cer": 0.07180722891566266, | |
| "eval_loss": 0.19239871203899384, | |
| "eval_runtime": 34.8631, | |
| "eval_samples_per_second": 13.252, | |
| "eval_steps_per_second": 1.664, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 31.54, | |
| "learning_rate": 0.00013096363636363635, | |
| "loss": 0.0256, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 32.31, | |
| "learning_rate": 0.0001255090909090909, | |
| "loss": 0.0257, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 33.08, | |
| "learning_rate": 0.00012005454545454545, | |
| "loss": 0.0238, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 33.85, | |
| "learning_rate": 0.0001146, | |
| "loss": 0.0228, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 34.62, | |
| "learning_rate": 0.00010919999999999998, | |
| "loss": 0.0242, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 34.62, | |
| "eval_cer": 0.0668273092369478, | |
| "eval_loss": 0.18180887401103973, | |
| "eval_runtime": 20.8913, | |
| "eval_samples_per_second": 22.115, | |
| "eval_steps_per_second": 2.776, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 35.38, | |
| "learning_rate": 0.00010374545454545453, | |
| "loss": 0.0209, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 36.15, | |
| "learning_rate": 9.829090909090907e-05, | |
| "loss": 0.0192, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 36.92, | |
| "learning_rate": 9.283636363636363e-05, | |
| "loss": 0.0171, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 37.69, | |
| "learning_rate": 8.738181818181818e-05, | |
| "loss": 0.0205, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 38.46, | |
| "learning_rate": 8.192727272727272e-05, | |
| "loss": 0.0187, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 38.46, | |
| "eval_cer": 0.06730923694779116, | |
| "eval_loss": 0.19057846069335938, | |
| "eval_runtime": 27.3046, | |
| "eval_samples_per_second": 16.92, | |
| "eval_steps_per_second": 2.124, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 39.23, | |
| "learning_rate": 7.647272727272727e-05, | |
| "loss": 0.0159, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "learning_rate": 7.101818181818181e-05, | |
| "loss": 0.0154, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 40.77, | |
| "learning_rate": 6.556363636363636e-05, | |
| "loss": 0.0165, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 41.54, | |
| "learning_rate": 6.01090909090909e-05, | |
| "loss": 0.0135, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 42.31, | |
| "learning_rate": 5.465454545454545e-05, | |
| "loss": 0.0156, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 42.31, | |
| "eval_cer": 0.06409638554216868, | |
| "eval_loss": 0.17488864064216614, | |
| "eval_runtime": 20.9298, | |
| "eval_samples_per_second": 22.074, | |
| "eval_steps_per_second": 2.771, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 43.08, | |
| "learning_rate": 4.9199999999999997e-05, | |
| "loss": 0.0137, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 43.85, | |
| "learning_rate": 4.374545454545454e-05, | |
| "loss": 0.0138, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 44.62, | |
| "learning_rate": 3.8290909090909086e-05, | |
| "loss": 0.0117, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 45.38, | |
| "learning_rate": 3.283636363636364e-05, | |
| "loss": 0.0132, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 46.15, | |
| "learning_rate": 2.738181818181818e-05, | |
| "loss": 0.0107, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 46.15, | |
| "eval_cer": 0.060481927710843375, | |
| "eval_loss": 0.17818936705589294, | |
| "eval_runtime": 20.9013, | |
| "eval_samples_per_second": 22.104, | |
| "eval_steps_per_second": 2.775, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 46.92, | |
| "learning_rate": 2.1927272727272727e-05, | |
| "loss": 0.0124, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 47.69, | |
| "learning_rate": 1.647272727272727e-05, | |
| "loss": 0.0119, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 48.46, | |
| "learning_rate": 1.1018181818181816e-05, | |
| "loss": 0.0097, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 49.23, | |
| "learning_rate": 5.563636363636363e-06, | |
| "loss": 0.0105, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "learning_rate": 1.0909090909090908e-07, | |
| "loss": 0.0099, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "eval_cer": 0.06088353413654619, | |
| "eval_loss": 0.1746283918619156, | |
| "eval_runtime": 27.3261, | |
| "eval_samples_per_second": 16.907, | |
| "eval_steps_per_second": 2.123, | |
| "step": 6500 | |
| } | |
| ], | |
| "max_steps": 6500, | |
| "num_train_epochs": 50, | |
| "total_flos": 1.985767056261625e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |