Closed snowmint closed 9 months ago
This project uses OpenAI's Whisper model, which is only trained on any-to-english translation and on transcription (output in the same language as audio input) tasks. Any-to-any translation is not possible with this model. Something you could do to get output in other languages would be to run the output through some translation API. This fork implements something like that (I haven't tested it).
Is there a possible argument for setting the translated language instead of English? I was looking into the code but I am not sure where I can modify it to match my thought. Thank you for reading my question, hope there is a way to do this.