Efficient Triton Kernels for LLM Training

Fully open reproduction of DeepSeek-R1

An extremely fast Python package and project manager, written in Rust.

No fortress, purely open ground. OpenManus is Coming.

The official implementation of work "REPARO: Compositional 3D Assets Generation with Differentiable 3D Layout Alignment".

Official codebase for "SWE-RL: Advancing LLM Reasoning via Reinforcement Learning on Open Software Evolution"

A modern download manager that supports all platforms. Built with Golang and Flutter.

DeepEP: an efficient expert-parallel communication library

🔥 A minimal training framework for scaling FLA models

Agentless🐱: an agentless approach to automatically solve software development problems

This repo contains the dataset and code for the paper "SWE-Lancer: Can Frontier LLMs Earn $1 Million from Real-World Freelance Software Engineering?"

Production-tested AI infrastructure tools for efficient AGI development and community-driven innovation

[ICLR 2025 Spotlight] Boltzmann-Aligned Inverse Folding Model as a Predictor of Mutational Effects on Protein-Protein Interactions

Official Repo for Open-Reasoner-Zero

Democratizing Reinforcement Learning for LLMs

A cat(1) clone with wings.

s1: Simple test-time scaling

🤗 smolagents: a barebones library for agents that think in python code.

Keep searching, reading webpages, reasoning until it finds the answer (or exceeding the token budget)

ripgrep recursively searches directories for a regex pattern while respecting your gitignore

RAGEN leverages reinforcement learning to train LLM reasoning agents in interactive, stochastic environments.

Simple RL training for reasoning

A plugin manager for Fish

CYaRon: Yet Another Random Olympic-iNformatics test data generator