training/flax/run_pseudo_labelling_pt.py [489:541]:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
            )

    if data_args.audio_column_name not in next(iter(raw_datasets.values())).column_names:
        raise ValueError(
            f"--audio_column_name '{data_args.audio_column_name}' not found in dataset"
            f" '{data_args.dataset_name}'. Make sure to set `--audio_column_name` to"
            " the correct audio column - one of"
            f" {', '.join(next(iter(raw_datasets.values())).column_names)}."
        )

    if data_args.text_column_name not in next(iter(raw_datasets.values())).column_names:
        raise ValueError(
            f"--text_column_name {data_args.text_column_name} not found in dataset"
            f" '{data_args.dataset_name}'. Make sure to set `--text_column_name` to the"
            " correct text column - one of"
            f" {', '.join(next(iter(raw_datasets.values())).column_names)}."
        )

    # 7. Load pretrained model, tokenizer, and feature extractor
    config = WhisperConfig.from_pretrained(
        (model_args.config_name if model_args.config_name else model_args.model_name_or_path),
        cache_dir=model_args.cache_dir,
        revision=model_args.model_revision,
        token=token,
    )
    feature_extractor = WhisperFeatureExtractor.from_pretrained(
        (model_args.feature_extractor_name if model_args.feature_extractor_name else model_args.model_name_or_path),
        cache_dir=model_args.cache_dir,
        revision=model_args.model_revision,
        token=token,
    )
    tokenizer = WhisperTokenizerFast.from_pretrained(
        (model_args.tokenizer_name if model_args.tokenizer_name else model_args.model_name_or_path),
        cache_dir=model_args.cache_dir,
        use_fast=model_args.use_fast_tokenizer,
        revision=model_args.model_revision,
        token=token,
    )
    processor = WhisperProcessor.from_pretrained(
        (model_args.processor_name if model_args.processor_name else model_args.model_name_or_path),
        cache_dir=model_args.cache_dir,
        revision=model_args.model_revision,
        token=token,
    )
    model = WhisperForConditionalGeneration.from_pretrained(
        model_args.model_name_or_path,
        config=config,
        cache_dir=model_args.cache_dir,
        revision=model_args.model_revision,
        subfolder=model_args.subfolder,
        token=token,
        low_cpu_mem_usage=True,
        torch_dtype=torch_dtype,
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -



training/run_pseudo_labelling.py [511:564]:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
            )

    if data_args.audio_column_name not in next(iter(raw_datasets.values())).column_names:
        raise ValueError(
            f"--audio_column_name '{data_args.audio_column_name}' not found in dataset"
            f" '{data_args.dataset_name}'. Make sure to set `--audio_column_name` to"
            " the correct audio column - one of"
            f" {', '.join(next(iter(raw_datasets.values())).column_names)}."
        )

    if data_args.text_column_name not in next(iter(raw_datasets.values())).column_names:
        raise ValueError(
            f"--text_column_name {data_args.text_column_name} not found in dataset"
            f" '{data_args.dataset_name}'. Make sure to set `--text_column_name` to the"
            " correct text column - one of"
            f" {', '.join(next(iter(raw_datasets.values())).column_names)}."
        )
    
    # 7. Load pretrained model, tokenizer, and feature extractor
    config = WhisperConfig.from_pretrained(
        (model_args.config_name if model_args.config_name else model_args.model_name_or_path),
        cache_dir=model_args.cache_dir,
        revision=model_args.model_revision,
        token=token,
    )
    feature_extractor = WhisperFeatureExtractor.from_pretrained(
        (model_args.feature_extractor_name if model_args.feature_extractor_name else model_args.model_name_or_path),
        cache_dir=model_args.cache_dir,
        revision=model_args.model_revision,
        token=token,
    )
    tokenizer = WhisperTokenizerFast.from_pretrained(
        (model_args.tokenizer_name if model_args.tokenizer_name else model_args.model_name_or_path),
        cache_dir=model_args.cache_dir,
        use_fast=model_args.use_fast_tokenizer,
        revision=model_args.model_revision,
        token=token,
    )
    processor = WhisperProcessor.from_pretrained(
        (model_args.processor_name if model_args.processor_name else model_args.model_name_or_path),
        cache_dir=model_args.cache_dir,
        revision=model_args.model_revision,
        token=token,
    )

    model = WhisperForConditionalGeneration.from_pretrained(
        model_args.model_name_or_path,
        config=config,
        cache_dir=model_args.cache_dir,
        revision=model_args.model_revision,
        subfolder=model_args.subfolder,
        token=token,
        low_cpu_mem_usage=True,
        torch_dtype=torch_dtype,
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -