Replies: 2 comments 1 reply
-
Hi, yes, that's correct. Whisper (and, therefore, Buzz) only supports X-to-X and X-to-English transcriptions, but not English-to-X. English-to-X models might be available sometime in the future, but they aren't for now. |
Beta Was this translation helpful? Give feedback.
-
However, I've just come across this discussion on the Whisper project which indicates the opposite: openai/whisper#649 (reply in thread) I'm attaching the audio source file (ZIP) and the resulting text file to this message. I was pleasantly surprised! |
Beta Was this translation helpful? Give feedback.
-
Hello,
First of all, thank you for this great audio transcription project, which is really promising. Very good work!
The "transcription" part doesn't pose any problems (apart from a few bugs / crashes, perhaps linked to my ageing PC running Win10 22H2), but I don't understand how the "translation" function works: I can transcribe "FR" audio to a "EN" text translation, but I can't do it the other way round, starting from an "EN" audio transcription to obtain a "FR" text translation. The text result is always in English, whatever the settings (Whipser / Medium / Translate / French).
Is this feature not yet available in this direction, or am I doing it wrong?
Thank you in advance for your help.
Beta Was this translation helpful? Give feedback.
All reactions