After Autotrain, Can the LLM do JSON function calling?

[venturaEffect]

I'm planning to use MemGPT with the finetuned model done with Autotrain and the creator MemGPT has told me:

'as long as the model / LLM you generate can do JSON function calling, memgpt should work with it'

So, is it possible to use it with my finetuned model of Mistral done with Autotrain?

Appreciate