Running Featured 67 QED-Nano: Teaching a Tiny Model to Prove Hard Theorems ๐ 67 Who needs 1T parameters? Olympiad proofs with a 4B model
Running 79 Maintain the unmaintainable ๐ 79 Explore the complex relationships between 400+ machine learning models
Running 219 FineVision: Open Data is All You Need ๐ 219 A new open-source dataset for training VLMs
Running 90 Scaling FineWeb to 1000+ languages: Step 1: finding signal in 100s of evaluation tasks ๐ 90 Evaluate multilingual models using FineTasks
Running on CPU Upgrade 187 The Synthetic Data Playbook: Generating Trillions of the Finest Tokens ๐ 187 Visualize synthetic data experiments as an interactive bookshelf
Running on CPU Upgrade 13.9k Open LLM Leaderboard ๐ 13.9k Track, rank and evaluate open LLMs and chatbots
Running on CPU Upgrade Featured 3.05k The Smol Training Playbook ๐ 3.05k The secrets to building world-class LLMs
Running 3.74k The Ultra-Scale Playbook ๐ 3.74k The ultimate guide to training LLM on large GPU Clusters
Running 593 Scaling test-time compute ๐ 593 Boost LLM answers with searchโguided testโtime compute
Running Featured 1.31k FineWeb: decanting the web for the finest text data at scale ๐ท 1.31k Generate a curated webโtext dataset for LLM training