Local model

Open kennydd0 opened this issue 1 year ago • 3 comments

Running the server with --model ollama/xxx does not seems to work. It's still using gpt, on open interpreter it works.

Mar 20 '24 09:03 kennydd0

Same. Would love to see a list of recommended models

Mar 23 '24 04:03 beamercola

I pushed a fix to this yesterday, you can run the server with --local, select ollama, then select your ollama model. Please let me know if it works for you @kennydd0 or @beamercola !

Mar 23 '24 20:03 tyfiero

@tyfiero thanks for the fix! it worked for me :)

Mar 24 '24 01:03 keldenl

Yay! I'm going to close this for now, but anyone can open up a new issue if they have more --local issues.

Mar 24 '24 01:03 tyfiero