Documentation of use_cache in llm_config

[rickyloynd-microsoft]

@qingyun-wu informed me that llm_config has been used in the past to set use_cache, but this is not currently documented anywhere. Clear documentation is needed since our ChatCompletion.create method carries a use_cache argument, and users will hit a runtime error if they pass use_cache values through llm_config and directly to ChatCompletion.create at the same time.

[sonichi]

when will a user write both llm_config and ChatCompletion.create?

[rickyloynd-microsoft]

They could do this by accident.

[sonichi]

The current doc is:

llm_config dict or False - llm inference configuration. Please refer to Completion.create for available options. To disable llm-based auto reply, set to False.

https://microsoft.github.io/autogen/docs/reference/agentchat/conversable_agent#__init__

If any clarification is desired, this is the place.

[thinkall]

We are closing this issue due to inactivity; please reopen if the problem persists.