Stars
Emacs Plus formulae for the Homebrew package manager
A scalable, end-to-end training pipeline for general-purpose agents
A benchmark for LLMs on complicated tasks in the terminal
SkyRL: A Modular Full-stack RL Library for LLMs
Production-Grade Container Scheduling and Management
[COLM 2025] Code for Paper: Learning Adaptive Parallel Reasoning with Language Models
A multi-player tournament benchmark that tests LLMs in social reasoning, strategy, and deception. Players engage in public and private conversations, form alliances, and vote to eliminate each other
Moatless Testbeds allows you to create isolated testbed environments in a Kubernetes cluster where you can apply code changes through git patches and run tests or SWE-Bench evaluations.
YuE: Open Full-song Music Generation Foundation Model, something similar to Suno.ai but open
Official Repo for Fine-Tuning Large Vision-Language Models as Decision-Making Agents via Reinforcement Learning
Official Code for ACL 2023 Outstanding Paper: World-to-Words: Grounded Open Vocabulary Acquisition through Fast Mapping in Vision-Language Models
What would you do with 1000 H100s...
Minimal reproduction of DeepSeek R1-Zero
Legible, Scalable, Reproducible Foundation Models with Named Tensors and Jax
Scalable RL solution for advanced reasoning of language models
Fast and memory-efficient exact attention
Ring attention implementation with flash attention
Code for Paper: Training Software Engineering Agents and Verifiers with SWE-Gym [ICML 2025]
Open Overleaf/ShareLaTex projects in vscode, with full collaboration support.
Run, manage, and scale AI workloads on any AI infrastructure. Use one system to access & manage all AI compute (Kubernetes, 17+ clouds, or on-prem).
Sandboxed code execution for AI agents, locally or on the cloud. Massively parallel, easy to extend. Powering SWE-agent and more.
Qwen3-Coder is the code version of Qwen3, the large language model series developed by Qwen team, Alibaba Cloud.
Efficient Triton Kernels for LLM Training