1%|█▍                                                                                                                                                                                                                                              | 98/17061 [03:46<2:43:13,  1.73it/s]












































































  1%|██▊                                                                                                                                                                                                                                            | 199/17061 [07:35<2:29:26,  1.88it/s]














































































  2%|████▏                                                                                                                                                                                                                                          | 299/17061 [11:25<2:38:08,  1.77it/s]













































































  2%|█████▌                                                                                                                                                                                                                                         | 400/17061 [15:18<2:18:27,  2.01it/s]












































































  3%|██████▉                                                                                                                                                                                                                                        | 496/17061 [19:08<3:11:27,  1.44it/s]
  3%|███████                                                                                                                                                                                                                                        | 500/17061 [19:10<2:19:41,  1.98it/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 = 5437
  Batch size = 16

























































































































































































Configuration saved in ./checkpoint-500/config.json
{'eval_loss': 0.17190051078796387, 'eval_wer': 0.10716071656191475, 'eval_runtime': 383.4174, 'eval_samples_per_second': 14.18, 'eval_steps_per_second': 0.887, 'epoch': 0.32}
Model weights saved in ./checkpoint-500/pytorch_model.bin
Configuration saved in ./checkpoint-500/preprocessor_config.json
Configuration saved in ./preprocessor_config.json
Deleting older checkpoint [checkpoint-10500] due to args.save_total_limit
Deleting older checkpoint [checkpoint-11000] due to args.save_total_limit










































  3%|███████▋                                                                                                                                                                                                                                      | 554/17061 [29:32<22:44:51,  4.96s/it]Traceback (most recent call last):
  File "/mnt/lv_ai_1_dante/javierr/wav2vec2-xls-r-300m-npsc-bokmaal/run_speech_recognition_ctc.py", line 792, in <module>
    main()
  File "/mnt/lv_ai_1_dante/javierr/wav2vec2-xls-r-300m-npsc-bokmaal/run_speech_recognition_ctc.py", line 743, in main
    train_result = trainer.train(resume_from_checkpoint=checkpoint)
  File "/mnt/lv_ai_1_dante/javierr/audio/lib/python3.9/site-packages/transformers/trainer.py", line 1398, in train
    tr_loss_step = self.training_step(model, inputs)
  File "/mnt/lv_ai_1_dante/javierr/audio/lib/python3.9/site-packages/transformers/trainer.py", line 1990, in training_step
    self.scaler.scale(loss).backward()
  File "/mnt/lv_ai_1_dante/javierr/audio/lib/python3.9/site-packages/torch/_tensor.py", line 307, in backward
    torch.autograd.backward(self, gradient, retain_graph, create_graph, inputs=inputs)
  File "/mnt/lv_ai_1_dante/javierr/audio/lib/python3.9/site-packages/torch/autograd/__init__.py", line 154, in backward
    Variable._execution_engine.run_backward(
KeyboardInterrupt