After an intensive code review, we’re thrilled to announce the final results of the MPEG-G Microbiome Classification Challenge!

This round wasn’t just about raw accuracy — we looked at the full picture:

✅ On-time submissions

✅ Model consistency across Federated Learning (FL) and Central models

✅ Carbon efficiency scores

And now, the moment you’ve all been waiting for… 🏆

🥇 1st Place — Tiny Margins, Big Margins

🥈 2nd Place — Kmers

🥉 3rd Place — Ever_learners

These teams didn’t just deliver top-tier results — they produced clean, well-documented, and sustainable solutions that truly embodied the spirit of reproducible modelling.

To everyone who participated: you’ve pushed the boundaries of microbiome classification and made this an exceptional competition. Even if you didn’t make the podium this time, your submissions have added tremendous value to our collective learning.

💬 We’d love to see your approaches, code snippets, and thought processes shared in the discussion forums — that’s how the whole community grows stronger together.

Thank you all for your effort, creativity, and engagement. Keep your eyes peeled — more challenges (and chances to shine) are just around the corner.

Keep Zinding! 🚀