Skip to main content

Replit x Weights & Biases ML Hackathon

We're thrilled to partner with Replit for an ML Hackathon in early February. Here's why you should care and how you can join
Created on January 12|Last edited on January 12

Get Started

We’re extremely excited to announce that we're partnering with Replit for their very first Machine Learning Hackathon!
If you’re interested in joining the hackathon, head over to the official registration site.
We have all the details below. Let's get into it:


Machine Learning Hackathon

For one week in February, selected participants will get early access to Replits' new GPUs service to train ML models to power their repls and write up their results in Weights & Biases' Reports. To get started with your ML journey on Replit check out the Weights & Biases profile on Replit for our pinned repls.
Register for the hackathon


Prize Pool

Yes, there are prizes. What kid you ask? Compute. Specifically "cycles."
What are Cycles?
Cycles are Replit's virtual tokens that can be used to purchase extra compute power and functionality for your Repls. On Replit, cycles can be purchased for $0.01 USD each.
The total prize pool for this hackathon is over 500,000 cycles and there are multiple opportunities to win. We hav prizes for Best Weights & Biases Report, Best Repl, an Honorable Mention award, and of course, a Grand Prize. Here's a little more info:
Grand Prize: 300,000 cycles
The Grand Prize is given to an individual with the best project who has clearly demonstrated their machine learning prowess. All of their practice and expertise have accumulated into this one outstanding project.
Best Weights & Biases Report: 100,000 cycles
There’s nothing better than cozying up with a blanket, some hot cocoa, and a good research paper. The winner of the Best Weights & Biases Report is someone who has meticulous documentation, engaging writing, and enlightening insights!
Best Replit Project: 100,000 cycles
No minimum viable product will win this award. The Best Replit Project will be awarded to someone whose work is polished and refined. They’ve taken advantage of Replit’s numerous features and have even let everyone enjoy their work by sharing it to Community using the #wandb tag!
Honorable Mention: 50,000 cycles
Honorable mention goes to the project that made a real splash with the judges! This prize will go to one lucky participant who made something we did not expect to see!

How to Enter

To participate, just head over to the Official Hackathon site to submit the registration form! If selected, you will receive an email with all of the details. At the end of the Hackathon, you must submit your projects using the submission form and select which prizes you wish to compete for.

Dates

This hackathon runs for one week, from February 4th to the 11th. Valid entries must be submitted by 11:59 p.m. PT on the 11th. Winners will be announced soon thereafter!

Conclusion

That's all you need to know for now! We'll share more about the contest as we get closer to it but we're looking forward to checking out your research. And if you need that link one last time, you can sign up here.
Thanks!
Iterate on AI agents and models faster. Try Weights & Biases today.