Hey all!
So, while I’ve been working on my own stuff with Gemini, I’ve really enjoyed using the function calling beta API.
I also like the different kinds of prompt formatting stuff you can set up in Google AI studio.
The thing is, while I want to use more than just the chat prompt style, I feel like the other formats would be perfect for function calling, something that isn’t available yet in the studio.
If that gets implemented at some point, it would be huge lol. That way, I could help prime the model better, and perhaps even fine tune the model a little bit based on the results of the function call.
Please note that I do not mean actually executing the functions, but working with what the model returns when it does attempt a function call.
Otherwise, keep on building guys!