bug: calling o1 or o3 causes NeMo to use deprecated chat completion API

[Headless-Sourceman]

Did you check docs and existing issues?

• I have read all the NeMo-Guardrails docs
• I have updated the package to the latest version before submitting this issue
• (optional) I have used the develop branch
• I have searched the existing issues of NeMo-Guardrails

Python version (python --version)

Python 3.10.11

Operating system/version

Microsoft Windows 11 Enterprise

NeMo-Guardrails version (if you must use a specific version and not the latest

0.11.1

Describe the bug

Trying to use Azure OpenAI o3-mini (or any o1 model) causes a bad request error.

Steps To Reproduce

1. Put this model into config.yml

type: main
    engine: azure
    model: o3-mini
    api_version: 2024-12-01-preview
    ...

2. Start a chat with
   nemoguardrails chat --config your/path --verbose --debug-level=DEBUG

3. Ask anything that uses the main model.

Expected Behavior

Receive a chat completion without error.

Actual Behavior

HTTPStatusError: Client error '400 Bad Request' for url
                             'https://...openai.azure.com/openai/.../o3-mini/completions?api-version=...'

Stack trace ends with
nemoguardrails.actions.llm.utils.LLMCallException: LLM Call Exception: Error code: 400 - {'error': {'code': 'OperationNotSupported', 'message': 'The completion operation does not work with the specified model, o3-mini. Please choose different model and try again. You can learn more about which models can be used with each operation here: https://go.microsoft.com/fwlink/?linkid=2197993.'}}

[Pouyanpi]

Thank you @Headless-Sourceman for reporting this bug.

Once #1085 is merged you can use these models, but you will face an issue. temperature is fixed to 1 for both o1 and o3 and we try to modify it in the code base. One option is to set lowest_temperature in the config.

I update this issue once the temperature problem gets resolved.

[Pouyanpi]

Thank you @Headless-Sourceman for reporting this bug.

Once #1085 is merged you can use these models, but you will face an issue. temperature is fixed to 1 for both o1 and o3 and we try to modify it in the code base. One option is to set lowest_temperature in the config.

I update this issue once the temperature problem gets resolved.

The fix is added by #1085 and will be available in v0.14.0 version of nemoguardrails.

If you'd like you can try it on develop branch

Looking for an easy way to find the correct provider name, see:

https://github.com/NVIDIA/NeMo-Guardrails/blob/develop/docs/user-guides/cli.md#providers
https://github.com/NVIDIA/NeMo-Guardrails/blob/develop/docs/user-guides/configuration-guide.md#exploring-available-providers

But my comments about lowest temperature still applies. If you want feel free to open an issue with that, you are very welcome to open a PR to fix this as well 👍🏻