tool

WhisperX transcription tool.

Classes

WhisperXTool

WhisperX transcription and translation tool.
class WhisperXTool(model_name='distil-large-v3', task='transcribe', device='cpu', overwrite=False, verbose=True)[source]

Bases: TranscriptionTool

WhisperX transcription and translation tool.

Parameters:

  • model_name (str) – The name of the model to use.

  • device (str) – The device where the computation should be executed.

  • overwrite (bool) – Whether to overwrite existing files, otherwise raise an error.

  • verbose (Union[bool, int]) – Whether to execute the computation verbosely.

  • task (str)

inference(audio_path, transcription_path, batch_size=16, language=None)[source]

WhisperX’s backed transcription method.

Parameters:

  • audio_path (Union[str, Path]) – Path to the audio track.

  • transcription_path (Union[str, Path]) – Path to the transcription file.

  • batch_size (int) – Batch size, reduce if low on GPU memory.

  • language (Optional[str]) – Country-code string of the spoken language.

Returns: