search

MahmoudAshraf97 / whisper-diarization

Automatic Speech Recognition with Speaker Diarization based on OpenAI Whisper
BSD 2-Clause "Simplified" License
2.44k stars 238 forks source link

Please, fix the language bug #148

Closed agershun closed 4 months ago

agershun commented 6 months ago

Please, add the line:

language=language

into the file transcription_helpers.py in the transcribe_batched() function:

https://github.com/MahmoudAshraf97/whisper-diarization/blob/39572386eb4170fc16440b770666f23ccf9bdc80/transcription_helpers.py#L64

    # Faster Whisper batched
    whisper_model = whisperx.load_model(
        model_name,
        device, # <--- PLEASE ADD language=language HERE
        compute_type=compute_dtype,
        asr_options={"suppress_numerals": suppress_numerals},
    )

The problem was: If the audio has very long rings in the very beginning of the file, it miss the language parameter and try to detect the language automatically.

PS. Sorry, that I did not make a PR

agershun commented 6 months ago

Just created the pull request:

https://github.com/MahmoudAshraf97/whisper-diarization/pull/149