:grey_question: Dealing with Structured Outputs (`pydantic`) :memo:

[adriens]

:grey_question: About

ollama-python is a very very convenient way to deal with local LLMs. Actually we can do a lot of stuff from the sdk facade, and I do now wonder if there are some code snippets to achieve Structured Outputs for SDK, or eventually around, see:

• Llama Index
• Query Engine with Pydantic Outputs
• LLM Pydantic Program
• LLM Structured Output for Function Calling with Ollama
• Langchain Ollama Functions
• OllamaFunctions python

:dart: The question

Would there be a way to achieve this from sdk ?