A visuailzation tool to make deep understaning and easier debugging for RLHF training.

The Arcade Learning Environment (ALE) -- a platform for AI research.

A Flexible Framework for Experiencing Cutting-edge LLM Inference Optimizations

Clean, minimal, accessible reproduction of DeepSeek R1-Zero

Sky-T1: Train your own O1 preview model within $450

🚀 Efficient implementations of state-of-the-art linear attention models in Torch and Triton

Super-Efficient RLHF Training of LLMs with Parameter Reallocation

A Massively Parallel Large Scale Self-Play Framework

USP: Unified (a.k.a. Hybrid, 2D) Sequence Parallel Attention for Long Context Transformers Model Training and Inference

A generative world for general-purpose robotics & embodied AI learning.

A curated list of awesome self-hosted GitHub Action runners in a large comparison matrix

(JAIR'2022) A mini-scale reproduction code of the AlphaStar program. Note: the original AlphaStar is the AI proposed by DeepMind to play StarCraft II. JAIR = Journal of Artificial Intelligence Rese…

A flexible and efficient training framework for large-scale alignment tasks

Ray is an AI compute engine. Ray consists of a core distributed runtime and a set of AI Libraries for accelerating ML workloads.

AI demo for playing ARPG/Soul-like game with RL frame

A novel parallel UCT algorithm with linear speedup and negligible performance loss.

verl: Volcano Engine Reinforcement Learning for LLMs

An elegant PyTorch deep reinforcement learning library.

Awesome-LLM-KV-Cache: A curated list of 📙Awesome LLM KV Cache Papers with Codes.

Mooncake is the serving platform for Kimi, a leading LLM service provided by Moonshot AI.

A collection of LLM papers, blogs, and projects, with a focus on OpenAI o1 🍓 and reasoning techniques.

Efficient Triton Kernels for LLM Training

arXiv LaTeX Cleaner: Easily clean the LaTeX code of your paper to submit to arXiv

Code for the paper "Training Diffusion Models with Reinforcement Learning"

SGLang is a fast serving framework for large language models and vision language models.

A lightweight library for portable low-level GPU computation using WebGPU.