Welcome to the UNICORN Challenge!
To help you get up and running, we provide three public repositories:

1) unicorn_baseline_template

A minimal working template that runs on all tasks.
✅ Perfect for understanding the submission flow.
🔍 Includes example data for each task so you can debug locally and verify that your output format is correct.

2) unicorn_baseline

Our official baseline implementations with more complete logic for input preprocessing and foundation model inference.
✅ Ideal if you want to study a working pipeline that produces meaningful outputs.
🔍 Also includes example data for local testing and output validation.

⚠️ Our baseline currently supports all tasks, except the radiology vision tasks (Task 2, 6, 7, 10 and 11).
These are being actively worked on and should be included soon.

3) unicorn_eval

The official evaluation toolkit for the challenge:
🧩 Contains adaptors used in vision tasks to map features → predictions
🛠️ Participants can submit custom adaptors via pull request — you'll get a unique name including your team name to ensure you receive credit!

For more details on adaptors and how to contribute, check out this README.

💬 Questions? Drop them here in the forum — we’re happy to help.
Good luck and happy programming!