Validate llm_config passed to ConversableAgent (issue #1522)

[gunnarku]

Based on #1522 (and a follow-up discussion in this PR), this commit implements the additional validation checks in ConversableAgent.

Add the following validation and raise ValueError if:

• The llm_config is None (validated in ConversableAgent).
• The llm_config has no model specified and config_list is empty (validated in OpenAIWrapper).
• The config_list has at least one entry, but not all the entries have the model is specified (validated in OpenAIWrapper).

The rest of the changes are code churn to adjust or add the test cases.

Why are these changes needed?

The root cause is #1522. The intent here is to enforce the contract for input parameters that are passed to ConversableAgent. With these changes the following happens:

from autogen import UserProxyAgent, AssistantAgent

# Throws an error because "config_list" is empty and "llm_config" does not
# have a value for "model" specified.
assistant = AssistantAgent("assistant", llm_config={"config_list": []})

# Throws an error because "config_list" has no "model" specified.
assistant = AssistantAgent("assistant", llm_config={"config_list":
                                                    [{"model":"", "api_key":
                                                      "sk-mysecretkey"}]})

# Works.
assistant = AssistantAgent("assistant", llm_config={"config_list": 
                                                   [{"model":"gpt-4", "api_key": "sk-mysecretkey"}]})

# Throws an error because one of the "config_list" entries has an empty model.
assistant = AssistantAgent("assistant", llm_config={"config_list":
                                                    [{"model":"gpt-4",
                                                      "api_key": "sk-mysecretkey"},
                                                     {"model":"", "api_key":
                                                      "sk-mysecretkey"}]})

# Throws an error because no "llm_config" is specified.
assistant = AssistantAgent("assistant")

# Works.
assistant = AssistantAgent("assistant", llm_config=False)

# Works: "llm_config" specifies the "model" attribute.
assistant = AssistantAgent("assistant", llm_config={"model": "gpt-4"})
proxy = UserProxyAgent("user")

proxy.initiate_chat(assistant)

Related issue number

Hopefully fixes #1522.

Checks

• I've included any doc changes needed for https://microsoft.github.io/autogen/. See https://microsoft.github.io/autogen/docs/Contribute#documentation to build and test documentation locally.
• [ X] I've added tests (if relevant) corresponding to the changes introduced in this PR.
• [ X] I've made sure all auto checks have passed.

[codecov-commenter]

Codecov Report

Attention: 1 lines in your changes are missing coverage. Please review.

Comparison is base (f68c09b) 37.85% compared to head (49d897a) 49.68%.

Files	Patch %	Lines
autogen/oai/client.py	87.50%	0 Missing and 1 partial ⚠️

Additional details and impacted files

@@             Coverage Diff             @@
##             main    #1654       +/-   ##
===========================================
+ Coverage   37.85%   49.68%   +11.83%     
===========================================
  Files          50       50               
  Lines        5733     5742        +9     
  Branches     1297     1410      +113     
===========================================
+ Hits         2170     2853      +683     
+ Misses       3391     2659      -732     
- Partials      172      230       +58     

Flag	Coverage Δ
unittests	49.59% <80.00%> (+11.74%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.