Function Calling Dataset Generation 🔧

This template replicates the methodology from the APIGen paper to generate high-quality synthetic datasets for training function-calling AI models.

What This Does

Generate customized API contract data for function calls - perfect for training models to understand when and how to call specific functions to improve specific tool agentic usage.

Sample Run

Check out sample_run.ipynb for a complete example you can run right away.

Source Implementation

The actual template code is located at:

src/starfish/data_gen_template/templates/starfish/function_calling/

Try it out! If you have any questions, let us know - we'd be happy to help. If you like this template, consider starring the repo and building your own! We welcome community contributions and are always happy to chat about new ideas. ⭐