[Feature] Best way to access raw API json results from a Prediction object.

[kalanyuz]

What feature would you like to see?

I'm doing an experiments with dspy Evaluate() to compare different models and there's some information I want to get from the raw response JSON that is slightly different for each model provider. Is there a way for me to get that?

Would you like to contribute?

• Yes, I'd like to help implement this.
• No, I just want to request it.

Additional Context

No response

[TomeHirata]

It depends on how you want to use the raw response JSON. If you just want to inspect for observability purpose, you can run mlflow.autolog to capture raw LLM request/response.