Can I use whispering with speaches local transcription in Windows 11

[salilsurendran]

Can I use whispering in Windows 11 with speaches as local transcription? I looked at the docker compose files and it seems to be made for Ubuntu.

[braden-w]

Oh darn, you might be right!

That's strange because it works on macOS and they have CUDA support, and I would presume that a large share of CUDA-supporting computers are Windows.

Could you try it anyway, or maybe try running Windows Subsystem for Linux? Sorry I don't have a Windows machine to test it out right now. Let me know!

[salilsurendran]

Yes, I installed Windows Subsystem for Linux and then downloaded Docker and the corresponding Docker image and model and I was able to use speaches on Windows.

[salilsurendran]

while The system does work, it is very slow, often taking up to 30 seconds to do a single transcription. It was a lot faster when I used Open AI or 11 Labs APIs, even though the darker image I have is installed locally on my system. What can I do to make it faster? should I download a larger model.