Fixes “The model produced invalid content” error when calling functions

[davorrunje]

Why are these changes needed?

This fixes the issue related to the more strict checking of JSON parameters in function calling as described here:

https://community.openai.com/t/error-the-model-produced-invalid-content/747511

This PR removes the 'name' parameter from messages in OpenAI client. It also introduces an additional parameter in the tool JSON specification that was previously missing.

Related issue number

Closes #3247

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.
• I've added tests (if relevant) corresponding to the changes introduced in this PR.
• I've made sure all auto checks have passed.

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 88.88889% with 1 line in your changes missing coverage. Please review.

Project coverage is 42.41%. Comparing base (6279247) to head (96658f9).
Report is 130 commits behind head on main.

Files with missing lines	Patch %	Lines
autogen/oai/client.py	87.50%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #3429      +/-   ##
==========================================
+ Coverage   32.90%   42.41%   +9.51%     
==========================================
  Files          94       96       +2     
  Lines       10235    10988     +753     
  Branches     2193     2538     +345     
==========================================
+ Hits         3368     4661    +1293     
+ Misses       6580     5919     -661     
- Partials      287      408     +121     

Flag	Coverage Δ
unittest	11.95% <33.33%> (?)
unittests	42.36% <88.88%> (+9.45%)	⬆️

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.

[marklysze]

@davorrunje, thanks for creating this - interesting that OpenAI are also removing name. This is likely to affect group chat with speaker selection. May need to incorporate the recent name transforms as a simpler integration for that (but that's for another discussion/time :) ).

I'll give it a test...

[davorrunje]

More about this error: https://community.openai.com/t/bizarre-issue-preventing-response-from-gpt-4o-mini-the-model-produces-invalid-content/875432

[marklysze]

@davorrunje, thanks again for working on a fix for the exception.

Would you have an example that I could use to replicate the exception?

[davorrunje]

@marklysze here is an example of failing completion call

[marklysze]

Thanks @davorrunje, is it possible to get the AutoGen code that would have generated this?

As an update: I ran the params through OpenAI's API (response = completions.create(**params)) and it ran through okay and returned a function call.

With my agent's termination check it failed on checking the content for the termination keyword as content is None:
is_termination_msg=lambda x: True if "FINISH" in x["content"] else False

The exception: argument of type 'NoneType' is not iterable

Updating my termination expression corrected that:
is_termination_msg=lambda x: True if x["content"] and "FINISH" in x["content"] else False

[davorrunje]

@marklysze I marked the PR as a draft as I am not sure yet what kind of workaround is the best one. The issue seems to be on the OpenAPI side and it is more common with GPT-4o-mini than older models. Maybe we should remove names only if we get an exception? Even changing the system message a little bit helps sometimes. The list of possible workarounds is (https://community.openai.com/t/bizarre-issue-preventing-response-from-gpt-4o-mini-the-model-produces-invalid-content/875432):

• Remove tools and tool_choice args.
• Slightly reduce the length of the prompt, even by a couple of words.
• Add at least a second user message or an assistant message.
• Change the name or remove the name parameter (but that didn’t always work, depending on the prompt). Some names work, others don’t. ‘Gregory’ worked, for example.
• Change the message. Some work and some don’t. ‘Hello.’ didn’t work, but ‘Hello, friend.’ worked.
• Change to gpt-4o or gpt-3.5-turbo.

[marklysze]

Thanks @davorrunje, appreciate the detail and the note on the draft.

It is definitely tricky because removing the name is a considerable change and could affect people's existing code - at the very least I'd suggest it is made as an option and defaults to existing behaviour of leaving it in.

From the list you provided, I think the third point would be the safest to do to minimise the impact on the validity of the messages. In the following linked LinkedIn post they added a message "DO NOT PRODUCE INVALID CONTENT" and that seemed to fix it! :).

LinkedIn Post on this being intermittent behaviour.

I would love to be able to replicate it, have you got any other examples that you can get to throw an exception?

I wonder if rather than change the messages to start with, we run inference and if it throws that specific exception then it adjusts the messages (perhaps with 3rd option above) and tries again up to x times.

[davorrunje]

Yes, I think we need to add a simple hack first and see how it goes. I see this error quite often when working with function calls and GPT-4o and GPT-4o-mini. I'll try adding a message as suggested first.

[marklysze]

Okay, not sure if it's possible but if you do get the exception and are able to get the params["messages"] it would be interesting to see if it is replicated.

See how you go with the additional message.

[lmcmahi]

Hi ,
I have the same error "The model produced invalid content” error when calling functions", how I can apply this fix, any idea ?

[marklysze]

Hi , I have the same error "The model produced invalid content” error when calling functions", how I can apply this fix, any idea ?

Hi @lmcmahi, would you be able to provide a code sample that produces this error? It would help in testing out viable fixes.

[zhwuwuwu]

Hi, I've encountered exactly the same error, having you find a good solution?