| { | |
| "best_metric": 0.3606263447286636, | |
| "best_model_checkpoint": "w2v-bert-2.0-CV_Fleurs-lg-10hrs-v4/checkpoint-25137", | |
| "epoch": 55.0, | |
| "eval_steps": 500, | |
| "global_step": 28215, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 30.792112350463867, | |
| "learning_rate": 9.998031496062993e-05, | |
| "loss": 1.4106, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_cer": 0.1264185884550246, | |
| "eval_loss": 0.5293365716934204, | |
| "eval_runtime": 31.0205, | |
| "eval_samples_per_second": 26.628, | |
| "eval_steps_per_second": 6.673, | |
| "eval_wer": 0.5665790102797036, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 10.16055679321289, | |
| "learning_rate": 9.897047244094488e-05, | |
| "loss": 0.4775, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_cer": 0.11182513176846737, | |
| "eval_loss": 0.4882170259952545, | |
| "eval_runtime": 30.4398, | |
| "eval_samples_per_second": 27.135, | |
| "eval_steps_per_second": 6.8, | |
| "eval_wer": 0.5251016017212526, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 24.60751724243164, | |
| "learning_rate": 9.796062992125985e-05, | |
| "loss": 0.3696, | |
| "step": 1539 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_cer": 0.09981164627486981, | |
| "eval_loss": 0.397329181432724, | |
| "eval_runtime": 30.5019, | |
| "eval_samples_per_second": 27.08, | |
| "eval_steps_per_second": 6.786, | |
| "eval_wer": 0.4564905570164953, | |
| "step": 1539 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 8.973501205444336, | |
| "learning_rate": 9.695078740157481e-05, | |
| "loss": 0.3007, | |
| "step": 2052 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_cer": 0.09322717991737761, | |
| "eval_loss": 0.3895888030529022, | |
| "eval_runtime": 29.8209, | |
| "eval_samples_per_second": 27.699, | |
| "eval_steps_per_second": 6.941, | |
| "eval_wer": 0.4311498924217069, | |
| "step": 2052 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 3.461097240447998, | |
| "learning_rate": 9.594094488188978e-05, | |
| "loss": 0.2564, | |
| "step": 2565 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_cer": 0.09659855331676664, | |
| "eval_loss": 0.3743041455745697, | |
| "eval_runtime": 29.9644, | |
| "eval_samples_per_second": 27.566, | |
| "eval_steps_per_second": 6.908, | |
| "eval_wer": 0.4372459956968683, | |
| "step": 2565 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 9.60593032836914, | |
| "learning_rate": 9.493110236220472e-05, | |
| "loss": 0.2273, | |
| "step": 3078 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_cer": 0.0879880973108153, | |
| "eval_loss": 0.3549309968948364, | |
| "eval_runtime": 30.2119, | |
| "eval_samples_per_second": 27.34, | |
| "eval_steps_per_second": 6.852, | |
| "eval_wer": 0.41836002868754485, | |
| "step": 3078 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 2.7023348808288574, | |
| "learning_rate": 9.392125984251969e-05, | |
| "loss": 0.203, | |
| "step": 3591 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_cer": 0.08468003608794061, | |
| "eval_loss": 0.3835783302783966, | |
| "eval_runtime": 29.6315, | |
| "eval_samples_per_second": 27.876, | |
| "eval_steps_per_second": 6.986, | |
| "eval_wer": 0.4000717188620607, | |
| "step": 3591 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 2.196669340133667, | |
| "learning_rate": 9.291141732283465e-05, | |
| "loss": 0.1775, | |
| "step": 4104 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_cer": 0.08550309438262714, | |
| "eval_loss": 0.3693934977054596, | |
| "eval_runtime": 30.2095, | |
| "eval_samples_per_second": 27.342, | |
| "eval_steps_per_second": 6.852, | |
| "eval_wer": 0.4132201769065264, | |
| "step": 4104 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 3.045036792755127, | |
| "learning_rate": 9.190354330708662e-05, | |
| "loss": 0.159, | |
| "step": 4617 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_cer": 0.0889694360467877, | |
| "eval_loss": 0.3487217128276825, | |
| "eval_runtime": 29.9136, | |
| "eval_samples_per_second": 27.613, | |
| "eval_steps_per_second": 6.92, | |
| "eval_wer": 0.4136983026535979, | |
| "step": 4617 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 3.3439998626708984, | |
| "learning_rate": 9.089370078740158e-05, | |
| "loss": 0.1444, | |
| "step": 5130 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_cer": 0.08915937257633075, | |
| "eval_loss": 0.3705226480960846, | |
| "eval_runtime": 30.2455, | |
| "eval_samples_per_second": 27.31, | |
| "eval_steps_per_second": 6.844, | |
| "eval_wer": 0.42768348075543866, | |
| "step": 5130 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 3.339082956314087, | |
| "learning_rate": 8.988385826771653e-05, | |
| "loss": 0.1262, | |
| "step": 5643 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_cer": 0.08284398296902452, | |
| "eval_loss": 0.3619905114173889, | |
| "eval_runtime": 30.7416, | |
| "eval_samples_per_second": 26.869, | |
| "eval_steps_per_second": 6.734, | |
| "eval_wer": 0.3981592158737748, | |
| "step": 5643 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.1660577654838562, | |
| "learning_rate": 8.88740157480315e-05, | |
| "loss": 0.1139, | |
| "step": 6156 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_cer": 0.08422102280821159, | |
| "eval_loss": 0.3665634095668793, | |
| "eval_runtime": 30.3996, | |
| "eval_samples_per_second": 27.171, | |
| "eval_steps_per_second": 6.809, | |
| "eval_wer": 0.40545063351661487, | |
| "step": 6156 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 0.19767990708351135, | |
| "learning_rate": 8.786417322834646e-05, | |
| "loss": 0.1015, | |
| "step": 6669 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_cer": 0.08355624495481094, | |
| "eval_loss": 0.3656393885612488, | |
| "eval_runtime": 30.0647, | |
| "eval_samples_per_second": 27.474, | |
| "eval_steps_per_second": 6.885, | |
| "eval_wer": 0.40114750179297154, | |
| "step": 6669 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 2.6208112239837646, | |
| "learning_rate": 8.685433070866143e-05, | |
| "loss": 0.0914, | |
| "step": 7182 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_cer": 0.08339796451352506, | |
| "eval_loss": 0.3572923541069031, | |
| "eval_runtime": 30.3516, | |
| "eval_samples_per_second": 27.214, | |
| "eval_steps_per_second": 6.82, | |
| "eval_wer": 0.39278030121922064, | |
| "step": 7182 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 3.129692316055298, | |
| "learning_rate": 8.584448818897639e-05, | |
| "loss": 0.0836, | |
| "step": 7695 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_cer": 0.07895028411339211, | |
| "eval_loss": 0.379753053188324, | |
| "eval_runtime": 30.1224, | |
| "eval_samples_per_second": 27.421, | |
| "eval_steps_per_second": 6.872, | |
| "eval_wer": 0.3795123117379871, | |
| "step": 7695 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 4.029957294464111, | |
| "learning_rate": 8.483464566929134e-05, | |
| "loss": 0.0729, | |
| "step": 8208 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_cer": 0.08219503315975245, | |
| "eval_loss": 0.4288513660430908, | |
| "eval_runtime": 30.1728, | |
| "eval_samples_per_second": 27.376, | |
| "eval_steps_per_second": 6.86, | |
| "eval_wer": 0.3895529524264882, | |
| "step": 8208 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 0.5638870596885681, | |
| "learning_rate": 8.38248031496063e-05, | |
| "loss": 0.0652, | |
| "step": 8721 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_cer": 0.087386631633929, | |
| "eval_loss": 0.4778367280960083, | |
| "eval_runtime": 30.6189, | |
| "eval_samples_per_second": 26.977, | |
| "eval_steps_per_second": 6.761, | |
| "eval_wer": 0.4157303370786517, | |
| "step": 8721 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 0.010239621624350548, | |
| "learning_rate": 8.281496062992127e-05, | |
| "loss": 0.0599, | |
| "step": 9234 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_cer": 0.07918770477532092, | |
| "eval_loss": 0.4570063352584839, | |
| "eval_runtime": 30.2821, | |
| "eval_samples_per_second": 27.277, | |
| "eval_steps_per_second": 6.836, | |
| "eval_wer": 0.37963184317475496, | |
| "step": 9234 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 3.2985517978668213, | |
| "learning_rate": 8.180511811023623e-05, | |
| "loss": 0.0504, | |
| "step": 9747 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_cer": 0.08564554677978442, | |
| "eval_loss": 0.44442683458328247, | |
| "eval_runtime": 30.5951, | |
| "eval_samples_per_second": 26.998, | |
| "eval_steps_per_second": 6.766, | |
| "eval_wer": 0.4125029882859192, | |
| "step": 9747 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 0.014229847118258476, | |
| "learning_rate": 8.079724409448819e-05, | |
| "loss": 0.0498, | |
| "step": 10260 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_cer": 0.08422102280821159, | |
| "eval_loss": 0.4611632823944092, | |
| "eval_runtime": 30.4634, | |
| "eval_samples_per_second": 27.115, | |
| "eval_steps_per_second": 6.795, | |
| "eval_wer": 0.3944537413339708, | |
| "step": 10260 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "grad_norm": 0.029411843046545982, | |
| "learning_rate": 7.978740157480316e-05, | |
| "loss": 0.0434, | |
| "step": 10773 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "eval_cer": 0.08431599107298311, | |
| "eval_loss": 0.48810598254203796, | |
| "eval_runtime": 30.3298, | |
| "eval_samples_per_second": 27.234, | |
| "eval_steps_per_second": 6.825, | |
| "eval_wer": 0.40019125029882857, | |
| "step": 10773 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "grad_norm": 0.19160379469394684, | |
| "learning_rate": 7.877755905511811e-05, | |
| "loss": 0.0386, | |
| "step": 11286 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "eval_cer": 0.08035898004083636, | |
| "eval_loss": 0.509917676448822, | |
| "eval_runtime": 31.249, | |
| "eval_samples_per_second": 26.433, | |
| "eval_steps_per_second": 6.624, | |
| "eval_wer": 0.377719340186469, | |
| "step": 11286 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "grad_norm": 0.18706804513931274, | |
| "learning_rate": 7.776771653543308e-05, | |
| "loss": 0.0381, | |
| "step": 11799 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "eval_cer": 0.08241662577755267, | |
| "eval_loss": 0.49042966961860657, | |
| "eval_runtime": 30.3145, | |
| "eval_samples_per_second": 27.248, | |
| "eval_steps_per_second": 6.828, | |
| "eval_wer": 0.38656466650729143, | |
| "step": 11799 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "grad_norm": 0.048221975564956665, | |
| "learning_rate": 7.675984251968504e-05, | |
| "loss": 0.0344, | |
| "step": 12312 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_cer": 0.08344544864591083, | |
| "eval_loss": 0.4621882736682892, | |
| "eval_runtime": 30.4774, | |
| "eval_samples_per_second": 27.102, | |
| "eval_steps_per_second": 6.792, | |
| "eval_wer": 0.4028209419077217, | |
| "step": 12312 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "grad_norm": 1.851014494895935, | |
| "learning_rate": 7.575e-05, | |
| "loss": 0.0302, | |
| "step": 12825 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "eval_cer": 0.08198926858608081, | |
| "eval_loss": 0.4985555112361908, | |
| "eval_runtime": 30.5557, | |
| "eval_samples_per_second": 27.033, | |
| "eval_steps_per_second": 6.775, | |
| "eval_wer": 0.39182404972507767, | |
| "step": 12825 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "grad_norm": 0.03737611323595047, | |
| "learning_rate": 7.474015748031497e-05, | |
| "loss": 0.0268, | |
| "step": 13338 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "eval_cer": 0.08118203833552289, | |
| "eval_loss": 0.5162001252174377, | |
| "eval_runtime": 30.3567, | |
| "eval_samples_per_second": 27.21, | |
| "eval_steps_per_second": 6.819, | |
| "eval_wer": 0.3954099928281138, | |
| "step": 13338 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "grad_norm": 0.08066482096910477, | |
| "learning_rate": 7.373031496062992e-05, | |
| "loss": 0.0272, | |
| "step": 13851 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "eval_cer": 0.07914022064293515, | |
| "eval_loss": 0.4747571647167206, | |
| "eval_runtime": 30.1451, | |
| "eval_samples_per_second": 27.401, | |
| "eval_steps_per_second": 6.867, | |
| "eval_wer": 0.3773607458761654, | |
| "step": 13851 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "grad_norm": 1.6533007621765137, | |
| "learning_rate": 7.27224409448819e-05, | |
| "loss": 0.0235, | |
| "step": 14364 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "eval_cer": 0.07853875496604884, | |
| "eval_loss": 0.47181904315948486, | |
| "eval_runtime": 30.7257, | |
| "eval_samples_per_second": 26.883, | |
| "eval_steps_per_second": 6.737, | |
| "eval_wer": 0.3822615347836481, | |
| "step": 14364 | |
| }, | |
| { | |
| "epoch": 29.0, | |
| "grad_norm": 0.020692087709903717, | |
| "learning_rate": 7.171259842519685e-05, | |
| "loss": 0.0219, | |
| "step": 14877 | |
| }, | |
| { | |
| "epoch": 29.0, | |
| "eval_cer": 0.0796942021874357, | |
| "eval_loss": 0.5317575335502625, | |
| "eval_runtime": 29.8983, | |
| "eval_samples_per_second": 27.627, | |
| "eval_steps_per_second": 6.923, | |
| "eval_wer": 0.37377480277312936, | |
| "step": 14877 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "grad_norm": 0.030614351853728294, | |
| "learning_rate": 7.070275590551182e-05, | |
| "loss": 0.0205, | |
| "step": 15390 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "eval_cer": 0.07830133430412003, | |
| "eval_loss": 0.5195885300636292, | |
| "eval_runtime": 30.3135, | |
| "eval_samples_per_second": 27.249, | |
| "eval_steps_per_second": 6.829, | |
| "eval_wer": 0.37688262012909396, | |
| "step": 15390 | |
| }, | |
| { | |
| "epoch": 31.0, | |
| "grad_norm": 1.4625964164733887, | |
| "learning_rate": 6.969291338582678e-05, | |
| "loss": 0.0203, | |
| "step": 15903 | |
| }, | |
| { | |
| "epoch": 31.0, | |
| "eval_cer": 0.07879200367210623, | |
| "eval_loss": 0.5203195214271545, | |
| "eval_runtime": 29.9013, | |
| "eval_samples_per_second": 27.624, | |
| "eval_steps_per_second": 6.923, | |
| "eval_wer": 0.37413339708343296, | |
| "step": 15903 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "grad_norm": 0.2324291467666626, | |
| "learning_rate": 6.868307086614174e-05, | |
| "loss": 0.019, | |
| "step": 16416 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "eval_cer": 0.08089713354120831, | |
| "eval_loss": 0.5031083822250366, | |
| "eval_runtime": 30.2, | |
| "eval_samples_per_second": 27.351, | |
| "eval_steps_per_second": 6.854, | |
| "eval_wer": 0.3858474778866842, | |
| "step": 16416 | |
| }, | |
| { | |
| "epoch": 33.0, | |
| "grad_norm": 0.04213400557637215, | |
| "learning_rate": 6.76732283464567e-05, | |
| "loss": 0.0179, | |
| "step": 16929 | |
| }, | |
| { | |
| "epoch": 33.0, | |
| "eval_cer": 0.08099210180597984, | |
| "eval_loss": 0.5772231221199036, | |
| "eval_runtime": 30.1355, | |
| "eval_samples_per_second": 27.41, | |
| "eval_steps_per_second": 6.869, | |
| "eval_wer": 0.37449199139373657, | |
| "step": 16929 | |
| }, | |
| { | |
| "epoch": 34.0, | |
| "grad_norm": 1.1665908098220825, | |
| "learning_rate": 6.666338582677165e-05, | |
| "loss": 0.0175, | |
| "step": 17442 | |
| }, | |
| { | |
| "epoch": 34.0, | |
| "eval_cer": 0.07630700074391808, | |
| "eval_loss": 0.4905574917793274, | |
| "eval_runtime": 30.2826, | |
| "eval_samples_per_second": 27.276, | |
| "eval_steps_per_second": 6.836, | |
| "eval_wer": 0.3675591680612001, | |
| "step": 17442 | |
| }, | |
| { | |
| "epoch": 35.0, | |
| "grad_norm": 0.009321126155555248, | |
| "learning_rate": 6.565354330708662e-05, | |
| "loss": 0.0166, | |
| "step": 17955 | |
| }, | |
| { | |
| "epoch": 35.0, | |
| "eval_cer": 0.0786020671425632, | |
| "eval_loss": 0.5370538830757141, | |
| "eval_runtime": 30.0507, | |
| "eval_samples_per_second": 27.487, | |
| "eval_steps_per_second": 6.888, | |
| "eval_wer": 0.3693521396127181, | |
| "step": 17955 | |
| }, | |
| { | |
| "epoch": 36.0, | |
| "grad_norm": 0.08575043827295303, | |
| "learning_rate": 6.464370078740158e-05, | |
| "loss": 0.0138, | |
| "step": 18468 | |
| }, | |
| { | |
| "epoch": 36.0, | |
| "eval_cer": 0.07882365976036342, | |
| "eval_loss": 0.5748464465141296, | |
| "eval_runtime": 30.2194, | |
| "eval_samples_per_second": 27.333, | |
| "eval_steps_per_second": 6.85, | |
| "eval_wer": 0.37437245995696866, | |
| "step": 18468 | |
| }, | |
| { | |
| "epoch": 37.0, | |
| "grad_norm": 0.006642826832830906, | |
| "learning_rate": 6.363582677165355e-05, | |
| "loss": 0.0134, | |
| "step": 18981 | |
| }, | |
| { | |
| "epoch": 37.0, | |
| "eval_cer": 0.07782649298026242, | |
| "eval_loss": 0.5342956781387329, | |
| "eval_runtime": 30.2477, | |
| "eval_samples_per_second": 27.308, | |
| "eval_steps_per_second": 6.843, | |
| "eval_wer": 0.36971073392302173, | |
| "step": 18981 | |
| }, | |
| { | |
| "epoch": 38.0, | |
| "grad_norm": 0.032926104962825775, | |
| "learning_rate": 6.26259842519685e-05, | |
| "loss": 0.0135, | |
| "step": 19494 | |
| }, | |
| { | |
| "epoch": 38.0, | |
| "eval_cer": 0.08035898004083636, | |
| "eval_loss": 0.5406966805458069, | |
| "eval_runtime": 30.4073, | |
| "eval_samples_per_second": 27.164, | |
| "eval_steps_per_second": 6.808, | |
| "eval_wer": 0.3839349748983983, | |
| "step": 19494 | |
| }, | |
| { | |
| "epoch": 39.0, | |
| "grad_norm": 0.4430481791496277, | |
| "learning_rate": 6.161614173228348e-05, | |
| "loss": 0.0123, | |
| "step": 20007 | |
| }, | |
| { | |
| "epoch": 39.0, | |
| "eval_cer": 0.07668687380300417, | |
| "eval_loss": 0.5342694520950317, | |
| "eval_runtime": 29.877, | |
| "eval_samples_per_second": 27.647, | |
| "eval_steps_per_second": 6.928, | |
| "eval_wer": 0.36612479081998567, | |
| "step": 20007 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "grad_norm": 0.005585499573498964, | |
| "learning_rate": 6.0606299212598426e-05, | |
| "loss": 0.0124, | |
| "step": 20520 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "eval_cer": 0.08170436379176625, | |
| "eval_loss": 0.5632651448249817, | |
| "eval_runtime": 30.3597, | |
| "eval_samples_per_second": 27.207, | |
| "eval_steps_per_second": 6.818, | |
| "eval_wer": 0.3801099689218264, | |
| "step": 20520 | |
| }, | |
| { | |
| "epoch": 41.0, | |
| "grad_norm": 0.006700966972857714, | |
| "learning_rate": 5.9596456692913396e-05, | |
| "loss": 0.0131, | |
| "step": 21033 | |
| }, | |
| { | |
| "epoch": 41.0, | |
| "eval_cer": 0.07735165165640481, | |
| "eval_loss": 0.5580958724021912, | |
| "eval_runtime": 30.4235, | |
| "eval_samples_per_second": 27.15, | |
| "eval_steps_per_second": 6.804, | |
| "eval_wer": 0.36325603633755676, | |
| "step": 21033 | |
| }, | |
| { | |
| "epoch": 42.0, | |
| "grad_norm": 0.0002161410520784557, | |
| "learning_rate": 5.858661417322835e-05, | |
| "loss": 0.0094, | |
| "step": 21546 | |
| }, | |
| { | |
| "epoch": 42.0, | |
| "eval_cer": 0.07888697193687776, | |
| "eval_loss": 0.5862228870391846, | |
| "eval_runtime": 30.2348, | |
| "eval_samples_per_second": 27.319, | |
| "eval_steps_per_second": 6.846, | |
| "eval_wer": 0.3683958881185752, | |
| "step": 21546 | |
| }, | |
| { | |
| "epoch": 43.0, | |
| "grad_norm": 0.005379769951105118, | |
| "learning_rate": 5.75767716535433e-05, | |
| "loss": 0.0101, | |
| "step": 22059 | |
| }, | |
| { | |
| "epoch": 43.0, | |
| "eval_cer": 0.07608540812611786, | |
| "eval_loss": 0.5479421615600586, | |
| "eval_runtime": 30.4733, | |
| "eval_samples_per_second": 27.106, | |
| "eval_steps_per_second": 6.793, | |
| "eval_wer": 0.36457088214200334, | |
| "step": 22059 | |
| }, | |
| { | |
| "epoch": 44.0, | |
| "grad_norm": 0.008168138563632965, | |
| "learning_rate": 5.656692913385827e-05, | |
| "loss": 0.0094, | |
| "step": 22572 | |
| }, | |
| { | |
| "epoch": 44.0, | |
| "eval_cer": 0.07605375203786068, | |
| "eval_loss": 0.5737643241882324, | |
| "eval_runtime": 30.1141, | |
| "eval_samples_per_second": 27.429, | |
| "eval_steps_per_second": 6.874, | |
| "eval_wer": 0.3620607219698781, | |
| "step": 22572 | |
| }, | |
| { | |
| "epoch": 45.0, | |
| "grad_norm": 0.001390150049701333, | |
| "learning_rate": 5.555708661417323e-05, | |
| "loss": 0.0078, | |
| "step": 23085 | |
| }, | |
| { | |
| "epoch": 45.0, | |
| "eval_cer": 0.0777473527596195, | |
| "eval_loss": 0.5284022688865662, | |
| "eval_runtime": 30.2759, | |
| "eval_samples_per_second": 27.282, | |
| "eval_steps_per_second": 6.837, | |
| "eval_wer": 0.37819746593354053, | |
| "step": 23085 | |
| }, | |
| { | |
| "epoch": 46.0, | |
| "grad_norm": NaN, | |
| "learning_rate": 5.4551181102362206e-05, | |
| "loss": 0.0074, | |
| "step": 23598 | |
| }, | |
| { | |
| "epoch": 46.0, | |
| "eval_cer": 0.07898194020164928, | |
| "eval_loss": 0.627711832523346, | |
| "eval_runtime": 30.4473, | |
| "eval_samples_per_second": 27.129, | |
| "eval_steps_per_second": 6.799, | |
| "eval_wer": 0.3724599569686828, | |
| "step": 23598 | |
| }, | |
| { | |
| "epoch": 47.0, | |
| "grad_norm": 0.0076849376782774925, | |
| "learning_rate": 5.354133858267717e-05, | |
| "loss": 0.01, | |
| "step": 24111 | |
| }, | |
| { | |
| "epoch": 47.0, | |
| "eval_cer": 0.0765127653175897, | |
| "eval_loss": 0.5825507640838623, | |
| "eval_runtime": 30.4404, | |
| "eval_samples_per_second": 27.135, | |
| "eval_steps_per_second": 6.8, | |
| "eval_wer": 0.3686349509921109, | |
| "step": 24111 | |
| }, | |
| { | |
| "epoch": 48.0, | |
| "grad_norm": 0.015365912578999996, | |
| "learning_rate": 5.2531496062992125e-05, | |
| "loss": 0.0088, | |
| "step": 24624 | |
| }, | |
| { | |
| "epoch": 48.0, | |
| "eval_cer": 0.07611706421437503, | |
| "eval_loss": 0.5601234436035156, | |
| "eval_runtime": 30.4389, | |
| "eval_samples_per_second": 27.136, | |
| "eval_steps_per_second": 6.801, | |
| "eval_wer": 0.36600525938321776, | |
| "step": 24624 | |
| }, | |
| { | |
| "epoch": 49.0, | |
| "grad_norm": 0.0013850124087184668, | |
| "learning_rate": 5.152165354330709e-05, | |
| "loss": 0.0083, | |
| "step": 25137 | |
| }, | |
| { | |
| "epoch": 49.0, | |
| "eval_cer": 0.07686098228841862, | |
| "eval_loss": 0.5409994721412659, | |
| "eval_runtime": 30.7659, | |
| "eval_samples_per_second": 26.848, | |
| "eval_steps_per_second": 6.728, | |
| "eval_wer": 0.3606263447286636, | |
| "step": 25137 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "grad_norm": 0.0006066646310500801, | |
| "learning_rate": 5.0513779527559055e-05, | |
| "loss": 0.0074, | |
| "step": 25650 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "eval_cer": 0.07804808559806264, | |
| "eval_loss": 0.559197723865509, | |
| "eval_runtime": 30.2523, | |
| "eval_samples_per_second": 27.304, | |
| "eval_steps_per_second": 6.842, | |
| "eval_wer": 0.3613435333492709, | |
| "step": 25650 | |
| }, | |
| { | |
| "epoch": 51.0, | |
| "grad_norm": 0.04679730907082558, | |
| "learning_rate": 4.950393700787402e-05, | |
| "loss": 0.007, | |
| "step": 26163 | |
| }, | |
| { | |
| "epoch": 51.0, | |
| "eval_cer": 0.07793728928916253, | |
| "eval_loss": 0.5890862941741943, | |
| "eval_runtime": 30.6371, | |
| "eval_samples_per_second": 26.961, | |
| "eval_steps_per_second": 6.757, | |
| "eval_wer": 0.3689935453024145, | |
| "step": 26163 | |
| }, | |
| { | |
| "epoch": 52.0, | |
| "grad_norm": 0.012076578103005886, | |
| "learning_rate": 4.849409448818898e-05, | |
| "loss": 0.0067, | |
| "step": 26676 | |
| }, | |
| { | |
| "epoch": 52.0, | |
| "eval_cer": 0.07788980515677678, | |
| "eval_loss": 0.5806664228439331, | |
| "eval_runtime": 30.6637, | |
| "eval_samples_per_second": 26.937, | |
| "eval_steps_per_second": 6.751, | |
| "eval_wer": 0.3662443222567535, | |
| "step": 26676 | |
| }, | |
| { | |
| "epoch": 53.0, | |
| "grad_norm": 0.0049696676433086395, | |
| "learning_rate": 4.748425196850394e-05, | |
| "loss": 0.0067, | |
| "step": 27189 | |
| }, | |
| { | |
| "epoch": 53.0, | |
| "eval_cer": 0.07728833947989047, | |
| "eval_loss": 0.5851139426231384, | |
| "eval_runtime": 30.4226, | |
| "eval_samples_per_second": 27.151, | |
| "eval_steps_per_second": 6.804, | |
| "eval_wer": 0.363973224958164, | |
| "step": 27189 | |
| }, | |
| { | |
| "epoch": 54.0, | |
| "grad_norm": 0.0018264329992234707, | |
| "learning_rate": 4.64744094488189e-05, | |
| "loss": 0.0065, | |
| "step": 27702 | |
| }, | |
| { | |
| "epoch": 54.0, | |
| "eval_cer": 0.07667104575887558, | |
| "eval_loss": 0.5988566875457764, | |
| "eval_runtime": 29.9692, | |
| "eval_samples_per_second": 27.562, | |
| "eval_steps_per_second": 6.907, | |
| "eval_wer": 0.36672244800382503, | |
| "step": 27702 | |
| }, | |
| { | |
| "epoch": 55.0, | |
| "grad_norm": 5.14693021774292, | |
| "learning_rate": 4.5464566929133864e-05, | |
| "loss": 0.005, | |
| "step": 28215 | |
| }, | |
| { | |
| "epoch": 55.0, | |
| "eval_cer": 0.07852292692192026, | |
| "eval_loss": 0.5746394395828247, | |
| "eval_runtime": 30.2216, | |
| "eval_samples_per_second": 27.331, | |
| "eval_steps_per_second": 6.849, | |
| "eval_wer": 0.37568730576141524, | |
| "step": 28215 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 51300, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 100, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 20, | |
| "early_stopping_threshold": 0.001 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 6 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 6.001178563865598e+19, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |