{ "best_metric": 0.2741861045360565, "best_model_checkpoint": "nrshoudi/wav2vec_arabic_mdd/checkpoint-16170", "epoch": 20.0, "eval_steps": 500, "global_step": 32340, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 1.0, "learning_rate": 5e-05, "loss": 8.6256, "step": 1617 }, { "epoch": 1.0, "eval_loss": 3.215786933898926, "eval_per": 1.0, "eval_runtime": 116.0176, "eval_samples_per_second": 7.008, "eval_steps_per_second": 3.508, "eval_wer": 1.0, "step": 1617 }, { "epoch": 2.0, "learning_rate": 0.0001, "loss": 1.3467, "step": 3234 }, { "epoch": 2.0, "eval_loss": 0.37750244140625, "eval_per": 0.060468293602185995, "eval_runtime": 116.9742, "eval_samples_per_second": 6.95, "eval_steps_per_second": 3.479, "eval_wer": 0.07531146470066725, "step": 3234 }, { "epoch": 3.0, "learning_rate": 9.444444444444444e-05, "loss": 0.1996, "step": 4851 }, { "epoch": 3.0, "eval_loss": 0.30538317561149597, "eval_per": 0.042094600961085464, "eval_runtime": 117.0156, "eval_samples_per_second": 6.948, "eval_steps_per_second": 3.478, "eval_wer": 0.056133638187672065, "step": 4851 }, { "epoch": 4.0, "learning_rate": 8.888888888888889e-05, "loss": 0.1311, "step": 6468 }, { "epoch": 4.0, "eval_loss": 0.31199753284454346, "eval_per": 0.038655422594930745, "eval_runtime": 116.6027, "eval_samples_per_second": 6.972, "eval_steps_per_second": 3.49, "eval_wer": 0.05067425691754934, "step": 6468 }, { "epoch": 5.0, "learning_rate": 8.333333333333334e-05, "loss": 0.1044, "step": 8085 }, { "epoch": 5.0, "eval_loss": 0.31690752506256104, "eval_per": 0.038608310562517664, "eval_runtime": 118.0869, "eval_samples_per_second": 6.885, "eval_steps_per_second": 3.447, "eval_wer": 0.05072091829592646, "step": 8085 }, { "epoch": 6.0, "learning_rate": 7.777777777777778e-05, "loss": 0.0891, "step": 9702 }, { "epoch": 6.0, "eval_loss": 0.30304601788520813, "eval_per": 0.03721850560633186, "eval_runtime": 118.602, "eval_samples_per_second": 6.855, "eval_steps_per_second": 3.432, "eval_wer": 0.04960104521487565, "step": 9702 }, { "epoch": 7.0, "learning_rate": 7.222222222222222e-05, "loss": 0.0788, "step": 11319 }, { "epoch": 7.0, "eval_loss": 0.31736186146736145, "eval_per": 0.03910298690285499, "eval_runtime": 118.2233, "eval_samples_per_second": 6.877, "eval_steps_per_second": 3.443, "eval_wer": 0.052027436890485744, "step": 11319 }, { "epoch": 8.0, "learning_rate": 6.666666666666667e-05, "loss": 0.0619, "step": 12936 }, { "epoch": 8.0, "eval_loss": 0.33120226860046387, "eval_per": 0.0416941486855743, "eval_runtime": 118.3706, "eval_samples_per_second": 6.868, "eval_steps_per_second": 3.438, "eval_wer": 0.05459381270122719, "step": 12936 }, { "epoch": 9.0, "learning_rate": 6.111111111111112e-05, "loss": 0.0552, "step": 14553 }, { "epoch": 9.0, "eval_loss": 0.3352554142475128, "eval_per": 0.03575803260152643, "eval_runtime": 119.3222, "eval_samples_per_second": 6.813, "eval_steps_per_second": 3.411, "eval_wer": 0.04796789697167654, "step": 14553 }, { "epoch": 10.0, "learning_rate": 5.555555555555556e-05, "loss": 0.0493, "step": 16170 }, { "epoch": 10.0, "eval_loss": 0.2741861045360565, "eval_per": 0.033567323094318285, "eval_runtime": 119.9093, "eval_samples_per_second": 6.78, "eval_steps_per_second": 3.394, "eval_wer": 0.04521487564742663, "step": 16170 }, { "epoch": 11.0, "learning_rate": 5e-05, "loss": 0.0417, "step": 17787 }, { "epoch": 11.0, "eval_loss": 0.28943416476249695, "eval_per": 0.03547536040704796, "eval_runtime": 120.1333, "eval_samples_per_second": 6.767, "eval_steps_per_second": 3.388, "eval_wer": 0.04586813494470627, "step": 17787 }, { "epoch": 12.0, "learning_rate": 4.4444444444444447e-05, "loss": 0.0382, "step": 19404 }, { "epoch": 12.0, "eval_loss": 0.29486820101737976, "eval_per": 0.03500424008291718, "eval_runtime": 119.6619, "eval_samples_per_second": 6.794, "eval_steps_per_second": 3.401, "eval_wer": 0.04633474872847744, "step": 19404 }, { "epoch": 13.0, "learning_rate": 3.888888888888889e-05, "loss": 0.0306, "step": 21021 }, { "epoch": 13.0, "eval_loss": 0.30409368872642517, "eval_per": 0.034792235937058325, "eval_runtime": 120.2788, "eval_samples_per_second": 6.759, "eval_steps_per_second": 3.384, "eval_wer": 0.04717465353926555, "step": 21021 }, { "epoch": 14.0, "learning_rate": 3.3333333333333335e-05, "loss": 0.03, "step": 22638 }, { "epoch": 14.0, "eval_loss": 0.3109176456928253, "eval_per": 0.033920663337416376, "eval_runtime": 118.5483, "eval_samples_per_second": 6.858, "eval_steps_per_second": 3.433, "eval_wer": 0.045261537025803746, "step": 22638 }, { "epoch": 15.0, "learning_rate": 2.777777777777778e-05, "loss": 0.0244, "step": 24255 }, { "epoch": 15.0, "eval_loss": 0.3234339952468872, "eval_per": 0.03347309902949213, "eval_runtime": 118.5882, "eval_samples_per_second": 6.856, "eval_steps_per_second": 3.432, "eval_wer": 0.044841584620409686, "step": 24255 }, { "epoch": 16.0, "learning_rate": 2.2222222222222223e-05, "loss": 0.0215, "step": 25872 }, { "epoch": 16.0, "eval_loss": 0.33212560415267944, "eval_per": 0.03316687081880712, "eval_runtime": 118.6022, "eval_samples_per_second": 6.855, "eval_steps_per_second": 3.432, "eval_wer": 0.04362838878260464, "step": 25872 }, { "epoch": 17.0, "learning_rate": 1.6666666666666667e-05, "loss": 0.0182, "step": 27489 }, { "epoch": 17.0, "eval_loss": 0.3367431163787842, "eval_per": 0.03154150570055592, "eval_runtime": 118.0708, "eval_samples_per_second": 6.886, "eval_steps_per_second": 3.447, "eval_wer": 0.04204190191778265, "step": 27489 }, { "epoch": 18.0, "learning_rate": 1.1111111111111112e-05, "loss": 0.0171, "step": 29106 }, { "epoch": 18.0, "eval_loss": 0.3317033350467682, "eval_per": 0.032766418543295955, "eval_runtime": 117.8323, "eval_samples_per_second": 6.9, "eval_steps_per_second": 3.454, "eval_wer": 0.043021790863702114, "step": 29106 }, { "epoch": 19.0, "learning_rate": 5.555555555555556e-06, "loss": 0.0155, "step": 30723 }, { "epoch": 19.0, "eval_loss": 0.3248533606529236, "eval_per": 0.03210685008951286, "eval_runtime": 118.4543, "eval_samples_per_second": 6.863, "eval_steps_per_second": 3.436, "eval_wer": 0.04227520880966824, "step": 30723 }, { "epoch": 20.0, "learning_rate": 0.0, "loss": 0.0132, "step": 32340 }, { "epoch": 20.0, "eval_loss": 0.31921908259391785, "eval_per": 0.032577970413643645, "eval_runtime": 118.1674, "eval_samples_per_second": 6.88, "eval_steps_per_second": 3.444, "eval_wer": 0.04292846810694788, "step": 32340 }, { "epoch": 20.0, "step": 32340, "total_flos": 1.0355998175044045e+19, "train_loss": 0.54961431122298, "train_runtime": 19355.5909, "train_samples_per_second": 3.342, "train_steps_per_second": 1.671 } ], "logging_steps": 500, "max_steps": 32340, "num_train_epochs": 20, "save_steps": 500, "total_flos": 1.0355998175044045e+19, "trial_name": null, "trial_params": null }