Hindi Transcription Quality

[seemantsingh-code]

Encountering an issue with the Python Whisper library while transcribing Hindi audio to Hindi text. The quality of the transcript is poor. Interestingly, when passing the parameter 'task=translate', it yields a good English translation.

Does this suggests that the library can identify words, but struggles with accurate Hindi transcription. Any suggestions on how to enhance the transcription quality for Hindi text?