Skip to content

[Bug]: Duplicated model property in llm_config in conversible agent #1025

@yashwanth-namora

Description

@yashwanth-namora

Describe the bug

@afourney, I and Soham were debugging the default model with some logs interestingly the conversible agent logs gave the below JSON(PFA). Containing the 2 models in the llm_config. Mostly when the agent config contains the function calls.

Steps to reproduce

No response

Expected Behavior

Expect to be only one model defined in the config.

Screenshots and logs

Screenshot 2023-12-19 at 4 01 29 PM

Additional Information

I think this is happening coz of this piece of code in ConversableAgent line number 114. I might be wrong as well

if llm_config is False:
    print("LLM_CONFIG FALSE CASE")
    self.llm_config = False
    self.client = None
else:
    self.llm_config = self.DEFAULT_CONFIG.copy()
    if isinstance(llm_config, dict):
        self.llm_config.update(llm_config)
        print("LLM_CONFIG NOT FALSE CASE")
    self.client = OpenAIWrapper(**self.llm_config)

Metadata

Metadata

Assignees

Labels

No labels
No labels

Type

No type
No fields configured for issues without a type.

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions