Feature request: adding qwencode

[adrien-barret]

Summary

Is it possible to run Qwencode llm localy? or do you know any tool that I can use to connect holmes to it?
Appendix

No response

[aantn]

Hey, I'm not sure! What would be the motivation for using Qwencode with Holmes and not another LLM?

[adrien-barret]

Qwencode seem to work well on some kind of code and we can run locally without any paid cost

[aantn]

Got it, thanks. How are you running it locally today?

…

On Mon, Oct 7, 2024 at 11:16 PM Adrien Barret ***@***.***> wrote: Qwencode seem to work well on some kind of code and we can run locally without any paid cost — Reply to this email directly, view it on GitHub <#151 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AADYUB2IZJITCXSFU7IXFX3Z2LTZ5AVCNFSM6AAAAABPP4PADGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGOJXHAYDQOJRHA> . You are receiving this because you commented.Message ID: ***@***.***>

[adrien-barret]

I'm using jan but it will probably not work with holmes

[aantn]

As far as I can tell, jan does provide an OpenAI-compatible http endpoint, so that could work. However, I am not sure if Jan or Qwencode support tool-calling. That would likely be the main challenge, but if you're able to confirm they do support it, we'd be happy to help get this working.

[adrien-barret]

will check that ! thanks