| { | |
| "best_metric": 0.17176076769828796, | |
| "best_model_checkpoint": "wavlm-korean-phoneme-v3/checkpoint-7500", | |
| "epoch": 11.980047885075818, | |
| "global_step": 7500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 2.9999999999999997e-05, | |
| "loss": 5.4024, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 5.9999999999999995e-05, | |
| "loss": 2.7142, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 8.999999999999999e-05, | |
| "loss": 2.6325, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 0.00011999999999999999, | |
| "loss": 2.4302, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 0.00015, | |
| "loss": 1.5708, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "eval_cer": 0.5738695561575328, | |
| "eval_loss": 1.0533815622329712, | |
| "eval_runtime": 257.0727, | |
| "eval_samples_per_second": 8.663, | |
| "eval_steps_per_second": 1.085, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 0.00017999999999999998, | |
| "loss": 0.9274, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "learning_rate": 0.00020999999999999998, | |
| "loss": 0.7045, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "learning_rate": 0.00023999999999999998, | |
| "loss": 0.5879, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "learning_rate": 0.00027, | |
| "loss": 0.5315, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "learning_rate": 0.0003, | |
| "loss": 0.4842, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "eval_cer": 0.2984314260873314, | |
| "eval_loss": 0.4019908010959625, | |
| "eval_runtime": 213.6605, | |
| "eval_samples_per_second": 10.423, | |
| "eval_steps_per_second": 1.306, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "learning_rate": 0.00029900990099009897, | |
| "loss": 0.4486, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "learning_rate": 0.000298019801980198, | |
| "loss": 0.4091, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "learning_rate": 0.000297029702970297, | |
| "loss": 0.3663, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "learning_rate": 0.000296039603960396, | |
| "loss": 0.3333, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "learning_rate": 0.000295049504950495, | |
| "loss": 0.3181, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "eval_cer": 0.20734923223947518, | |
| "eval_loss": 0.2830340266227722, | |
| "eval_runtime": 213.9837, | |
| "eval_samples_per_second": 10.407, | |
| "eval_steps_per_second": 1.304, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "learning_rate": 0.000294059405940594, | |
| "loss": 0.3238, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "learning_rate": 0.00029306930693069307, | |
| "loss": 0.3005, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "learning_rate": 0.00029207920792079207, | |
| "loss": 0.3013, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "learning_rate": 0.00029108910891089107, | |
| "loss": 0.2795, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 3.19, | |
| "learning_rate": 0.00029009900990099006, | |
| "loss": 0.2421, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.19, | |
| "eval_cer": 0.17599212479611415, | |
| "eval_loss": 0.2522282004356384, | |
| "eval_runtime": 210.0211, | |
| "eval_samples_per_second": 10.604, | |
| "eval_steps_per_second": 1.328, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.35, | |
| "learning_rate": 0.00028910891089108906, | |
| "loss": 0.2436, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.51, | |
| "learning_rate": 0.00028811881188118806, | |
| "loss": 0.234, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.67, | |
| "learning_rate": 0.0002871287128712871, | |
| "loss": 0.2347, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.83, | |
| "learning_rate": 0.0002861386138613861, | |
| "loss": 0.2356, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.99, | |
| "learning_rate": 0.0002851485148514851, | |
| "loss": 0.2255, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.99, | |
| "eval_cer": 0.1599614862291713, | |
| "eval_loss": 0.22431106865406036, | |
| "eval_runtime": 203.7474, | |
| "eval_samples_per_second": 10.93, | |
| "eval_steps_per_second": 1.369, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 4.15, | |
| "learning_rate": 0.00028415841584158416, | |
| "loss": 0.1892, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 4.31, | |
| "learning_rate": 0.00028316831683168316, | |
| "loss": 0.1927, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 4.47, | |
| "learning_rate": 0.00028217821782178216, | |
| "loss": 0.1832, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 4.63, | |
| "learning_rate": 0.00028118811881188116, | |
| "loss": 0.1914, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 4.79, | |
| "learning_rate": 0.00028019801980198015, | |
| "loss": 0.1879, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.79, | |
| "eval_cer": 0.14389492063720172, | |
| "eval_loss": 0.21251684427261353, | |
| "eval_runtime": 199.7095, | |
| "eval_samples_per_second": 11.151, | |
| "eval_steps_per_second": 1.397, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.95, | |
| "learning_rate": 0.0002792079207920792, | |
| "loss": 0.186, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 5.11, | |
| "learning_rate": 0.0002782178217821782, | |
| "loss": 0.1716, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 5.27, | |
| "learning_rate": 0.0002772277227722772, | |
| "loss": 0.1588, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 5.43, | |
| "learning_rate": 0.0002762376237623762, | |
| "loss": 0.1555, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 5.59, | |
| "learning_rate": 0.0002752475247524752, | |
| "loss": 0.1604, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 5.59, | |
| "eval_cer": 0.13375631417464845, | |
| "eval_loss": 0.2036324292421341, | |
| "eval_runtime": 200.1074, | |
| "eval_samples_per_second": 11.129, | |
| "eval_steps_per_second": 1.394, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 5.75, | |
| "learning_rate": 0.0002742574257425742, | |
| "loss": 0.158, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 5.91, | |
| "learning_rate": 0.00027326732673267325, | |
| "loss": 0.1619, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 6.07, | |
| "learning_rate": 0.00027227722772277225, | |
| "loss": 0.1497, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 6.23, | |
| "learning_rate": 0.00027128712871287125, | |
| "loss": 0.1397, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 6.39, | |
| "learning_rate": 0.0002702970297029703, | |
| "loss": 0.1382, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 6.39, | |
| "eval_cer": 0.127418786959927, | |
| "eval_loss": 0.189874529838562, | |
| "eval_runtime": 198.5164, | |
| "eval_samples_per_second": 11.218, | |
| "eval_steps_per_second": 1.405, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 6.55, | |
| "learning_rate": 0.0002693069306930693, | |
| "loss": 0.1371, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 6.71, | |
| "learning_rate": 0.0002683168316831683, | |
| "loss": 0.1385, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 6.87, | |
| "learning_rate": 0.0002673267326732673, | |
| "loss": 0.1382, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 7.03, | |
| "learning_rate": 0.0002663366336633663, | |
| "loss": 0.138, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 7.19, | |
| "learning_rate": 0.00026534653465346534, | |
| "loss": 0.1228, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 7.19, | |
| "eval_cer": 0.12003937601942934, | |
| "eval_loss": 0.1979004442691803, | |
| "eval_runtime": 201.8805, | |
| "eval_samples_per_second": 11.031, | |
| "eval_steps_per_second": 1.382, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 7.35, | |
| "learning_rate": 0.00026435643564356434, | |
| "loss": 0.128, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 7.51, | |
| "learning_rate": 0.00026336633663366334, | |
| "loss": 0.1248, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 7.67, | |
| "learning_rate": 0.00026237623762376234, | |
| "loss": 0.1191, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 7.83, | |
| "learning_rate": 0.00026138613861386134, | |
| "loss": 0.1241, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 7.99, | |
| "learning_rate": 0.00026039603960396033, | |
| "loss": 0.1275, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 7.99, | |
| "eval_cer": 0.11670534809694548, | |
| "eval_loss": 0.18830785155296326, | |
| "eval_runtime": 215.0184, | |
| "eval_samples_per_second": 10.357, | |
| "eval_steps_per_second": 1.298, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 8.15, | |
| "learning_rate": 0.0002594059405940594, | |
| "loss": 0.1133, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 8.31, | |
| "learning_rate": 0.0002584158415841584, | |
| "loss": 0.1121, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 8.47, | |
| "learning_rate": 0.0002574257425742574, | |
| "loss": 0.112, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 8.63, | |
| "learning_rate": 0.00025643564356435644, | |
| "loss": 0.108, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 8.79, | |
| "learning_rate": 0.00025544554455445543, | |
| "loss": 0.1093, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 8.79, | |
| "eval_cer": 0.11560598113112645, | |
| "eval_loss": 0.1854041963815689, | |
| "eval_runtime": 221.0978, | |
| "eval_samples_per_second": 10.072, | |
| "eval_steps_per_second": 1.262, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 8.94, | |
| "learning_rate": 0.00025445544554455443, | |
| "loss": 0.1094, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 9.11, | |
| "learning_rate": 0.00025346534653465343, | |
| "loss": 0.1139, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 9.26, | |
| "learning_rate": 0.00025247524752475243, | |
| "loss": 0.1015, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 9.42, | |
| "learning_rate": 0.0002514851485148515, | |
| "loss": 0.1062, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 9.58, | |
| "learning_rate": 0.0002504950495049505, | |
| "loss": 0.1089, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 9.58, | |
| "eval_cer": 0.1084421323407894, | |
| "eval_loss": 0.1830228716135025, | |
| "eval_runtime": 220.3658, | |
| "eval_samples_per_second": 10.106, | |
| "eval_steps_per_second": 1.266, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 9.74, | |
| "learning_rate": 0.0002495148514851485, | |
| "loss": 0.1076, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 9.9, | |
| "learning_rate": 0.0002485247524752475, | |
| "loss": 0.1063, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 10.06, | |
| "learning_rate": 0.0002475346534653465, | |
| "loss": 0.1009, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 10.22, | |
| "learning_rate": 0.0002465445544554455, | |
| "loss": 0.0987, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 10.38, | |
| "learning_rate": 0.0002455643564356435, | |
| "loss": 0.0968, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 10.38, | |
| "eval_cer": 0.10436082229774882, | |
| "eval_loss": 0.17644909024238586, | |
| "eval_runtime": 220.8492, | |
| "eval_samples_per_second": 10.084, | |
| "eval_steps_per_second": 1.263, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 10.54, | |
| "learning_rate": 0.0002445742574257425, | |
| "loss": 0.0937, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 10.7, | |
| "learning_rate": 0.00024358415841584157, | |
| "loss": 0.0964, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 10.86, | |
| "learning_rate": 0.00024259405940594056, | |
| "loss": 0.0956, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 11.02, | |
| "learning_rate": 0.0002416039603960396, | |
| "loss": 0.0971, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 11.18, | |
| "learning_rate": 0.0002406138613861386, | |
| "loss": 0.0862, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 11.18, | |
| "eval_cer": 0.10214771755610005, | |
| "eval_loss": 0.19613920152187347, | |
| "eval_runtime": 296.3031, | |
| "eval_samples_per_second": 7.516, | |
| "eval_steps_per_second": 0.942, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 11.34, | |
| "learning_rate": 0.00023962376237623759, | |
| "loss": 0.0901, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 11.5, | |
| "learning_rate": 0.00023863366336633664, | |
| "loss": 0.0957, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 11.66, | |
| "learning_rate": 0.00023764356435643564, | |
| "loss": 0.0929, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 11.82, | |
| "learning_rate": 0.00023665346534653463, | |
| "loss": 0.0918, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 11.98, | |
| "learning_rate": 0.00023566336633663366, | |
| "loss": 0.0933, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 11.98, | |
| "eval_cer": 0.10178126190082704, | |
| "eval_loss": 0.17176076769828796, | |
| "eval_runtime": 220.1598, | |
| "eval_samples_per_second": 10.115, | |
| "eval_steps_per_second": 1.267, | |
| "step": 7500 | |
| } | |
| ], | |
| "max_steps": 31300, | |
| "num_train_epochs": 50, | |
| "total_flos": 6.187118202284854e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |