parameter-golf

by openai · Developer Tools

Train the smallest LM you can that fits in 16MB. Best model wins!

New0 ratings5,178 starsActive
Developer ToolsDeveloper ToolPythonTeX
GitHub⚑ Report
parameter-golf — image 1

About this project

OpenAI Model Craft Challenge: Parameter Golf is a challenge to train the best language model that fits in a 16MB artifact and trains in under 10 minutes on 8xH100s, evaluated by compression on the FineWeb validation set (tokenizer-agnostic, bits per byte). This challenge is heavily inspired by the NanoGPT Speedrunning challenge, where participants compete to train a model that reaches 3.28 FineWeb validation loss as quickly as possible. We're excited to see how optimizing for a parameter-constrained setting pushes people toward unique architectures (test-time compute, aggressive parameter tying, depth recurrence, low-rank training, ...), compression schemes (low precision, QAT, bitnets,…

Technologies

ShellPythonC++TeX

Project health

Low recent activity
Last update4 months ago
Contributors45
Latest release
Open issues & PRs1,434
LicenseMIT
On GitHubsince 2026

GitHub

5,178
stars
3,274
forks
45
contributors
1,434
open issues & PRs
Python
language
4 months ago
last commit
View on GitHub ↗

Reviews

out of 5 · 0 ratings
★★★★★
0%
★★★★
0%
★★★
0%
★★
0%
0%
Sign in to write a review
No reviews yet
Be the first to review parameter-golf.

Built by

openai
Imported from GitHub · not yet claimed on GitPalace
View developer pageSign in with GitHub to claim

Maintain openai/parameter-golf? Claiming verifies admin access through your GitHub account and gives you control of this listing.

You might also like