0%| | 99/499680 [03:57<142:58:51, 1.03s/it] 0%| | 200/499680 [07:55<131:47:55, 1.05it/s] 0%| | 298/499680 [11:52<147:05:38, 1.06s/it] 0%| | 399/499680 [15:57<146:33:14, 1.06s/it] 0%| | 498/499680 [20:03<149:47:46, 1.08s/it] 0%| | 500/499680 [20:05<135:25:31, 1.02it/s]The following columns in the evaluation set don't have a corresponding argument in `Wav2Vec2ForCTC.forward` and have been ignored: input_length. If input_length are not expected by `Wav2Vec2ForCTC.forward`, you can safely ignore this message. ***** Running Evaluation ***** Num examples = 40740 Batch size = 12 Configuration saved in ./checkpoint-500/config.json {'eval_loss': 2.797088623046875, 'eval_wer': 1.0, 'eval_runtime': 2782.5245, 'eval_samples_per_second': 14.641, 'eval_steps_per_second': 1.22, 'epoch': 0.04} Model weights saved in ./checkpoint-500/pytorch_model.bin Feature extractor saved in ./checkpoint-500/preprocessor_config.json Feature extractor saved in ./preprocessor_config.json 0%|▏ | 599/499680 [1:11:53<148:35:07, 1.07s/it] 0%|▏ | 699/499680 [1:15:54<143:01:10, 1.03s/it] 0%|▏ | 800/499680 [1:19:57<134:05:24, 1.03it/s] 0%|▏ | 898/499680 [1:23:57<154:25:15, 1.11s/it] 0%|▏ | 1000/499680 [1:28:03<139:02:50, 1.00s/it]The following columns in the evaluation set don't have a corresponding argument in `Wav2Vec2ForCTC.forward` and have been ignored: input_length. If input_length are not expected by `Wav2Vec2ForCTC.forward`, you can safely ignore this message. ***** Running Evaluation ***** Num examples = 40740 Batch size = 12 0%| | 0/3395 [00:00