private-gpt icon indicating copy to clipboard operation
private-gpt copied to clipboard
verified publisher icon zylon-ai

Better support of openailike LLM tools

Open jcbonnet-fwd opened this issue 1 year ago • 2 comments

Add timeout parameter for better support of openailike LLM tools on local computer (like LM Studio). Reuse other parameters to improve settings of the OpenAILike object.

jcbonnet-fwd avatar Apr 18 '24 14:04 jcbonnet-fwd

@jcbonnet-fwd could you please pull the changes from main? There are conflicts. Thanks!

imartinez avatar Apr 30 '24 08:04 imartinez

@jcbonnet-fwd could you please pull the changes from main? There are conflicts. Thanks!

Is it OK now?

jcbonnet-fwd avatar May 02 '24 18:05 jcbonnet-fwd

I think we could add timeout=openai_settings.request_timeout to openai LLM too, apart fro openailike

Sure, it makes sense. As there's also an "ollama_settings.request_timeout", it could be an idea to move the "request_timeout" parameter to the more generic "llm" yaml object (and add it to llamacpp, sagemaker, azopeonai, if available there). What do you think?

jcbonnet-fwd avatar May 13 '24 13:05 jcbonnet-fwd