AlgoPerf

We just released AlgoPerf v1.0.1! 📚 New workload: AlgoPerf now includes a decoder-only 150M Language Model. ⚙️ Hardware switch: moving from 8xV100 to 4xA100. 🐞 Bug fixes and improved API. Coming soon: updated leaderboard with new optimizers! 👀

We're all about acceleration! 😉 Watch @priya-kasimbeg.bsky.social & @fsschneider.bsky.social speedrun an explanation of the AlgoPerf benchmark, rules, and results all within a tight 5 minutes for our #ICLR2025 paper video on "Accelerating Neural Network Training". See you in Singapore!

Hi there! This account will post about the AlgoPerf benchmark and leaderboard updates for faster neural network training via better training algorithms. But let's start with what AlgoPerf is, what we have done so far, and how you can train neural nets ~30% faster.

This is just the beginning! With the help of the community, we want to test even more training recipes, push the SOTA for neural network training methods even further, and improve the AlgoPerf benchmark. Follow us & join the working group (mlcommons.org/working-grou...).

Hi there! This account will post about the AlgoPerf benchmark and leaderboard updates for faster neural network training via better training algorithms. But let's start with what AlgoPerf is, what we have done so far, and how you can train neural nets ~30% faster.

Then, we asked the community to submit training algorithms. The results? The winner of the external tuning ruleset, using Distributed Shampoo, reduced training time by ~30% over our well-tuned baseline—showing that non-diagonal methods can beat Adam, even in wall-clock time!

Hi there! This account will post about the AlgoPerf benchmark and leaderboard updates for faster neural network training via better training algorithms. But let's start with what AlgoPerf is, what we have done so far, and how you can train neural nets ~30% faster.

Over the past years, we've built the AlgoPerf: Training Algorithms benchmark. The core ideas: (1) Only the training algorithm changes; everything else (hardware, model, data) stays fixed. (2) Submissions compete directly; no more weak baselines!

Hi there! This account will post about the AlgoPerf benchmark and leaderboard updates for faster neural network training via better training algorithms. But let's start with what AlgoPerf is, what we have done so far, and how you can train neural nets ~30% faster.

Currently, training neural nets is a complicated & fragile process with many important choices: How should I set/tune the learning rate? Using what schedule? Should I use SGD or Adam (or maybe Nadam/Amos/Shampoo/SOAP/Muon/... the list is virtually endless)?