8 ms·
If there is a better mode I can easily integrate using python I’m all ears. For what I’m building quality is most important
by dontreact 3y ago
If there is a better mode I can easily integrate using python I’m all ears. For what I’m building quality is most important
- abraxas 3y agoOh, Azure's speech recognition API beats it handily on English language. Both in accuracy and speed. Another is Deepgram. Even this obscure vendor seems to be able to handle the samples I tried better than Whisper: https://picovoice.ai/platform/cat/ https://picovoice.ai/platform/cat/ But yeah, go with Azure as your starting point. It is good and the price is likely acceptable unless you're transcribing all of youtube.
- dontreact 3y agoUmm I want to pay zero and run locally
- generalizations 3y agoIf you use the large_v2 version of whisper, and give it a prompt to indicate what it's transcribing, it can do extremely well. But do use the prompt feature.
- dontreact 3y agoYeah exactly this is why there’s hype. It’s the best model that you can use for free easily