KernelBench: Can LLMs Write GPU Kernels? - Benchmark + Toolkit with Torch -> CUDA (+ more DSLs)

Autonomous GPU Kernel Generation via Deep Agents

Building the Virtuous Cycle for AI-driven LLM Systems

Complete Claude Code configuration collection - agents, skills, hooks, commands, rules, MCPs. Battle-tested configs from an Anthropic hackathon winner.

A tool analyzing unused GPU code by machine learning workloads

LLaDA2.0 is the diffusion language model series developed by InclusionAI team, Ant Group.

Official implementation of our NeurIPS 2025 paper: "FlyLoRA: Boosting Task Decoupling and Parameter Efficiency via Implicit Rank-Wise Mixture-of-Experts."

DLSlime: Flexible & Efficient Heterogeneous Transfer Toolkit

Research and development (R&D) is crucial for the enhancement of industrial productivity, especially in the AI era, where the core aspects of R&D are mainly focused on data and models. We are commi…

交易模块

FactorHub是一个自研的现代化量化因子分析平台，专为量化投资研究者设计。平台完全自主研发，集成了「数据获取-因子管理-因子分析-策略回测-因子挖掘」的完整工作流程，通过直观的Web界面和强大的计算引擎，大幅降低量化分析的门槛，提高研究效率。 FactorHub = Factor（因子） + Hub（中心），意为因子分析和管理的核心枢纽。

Domain-specific language designed to streamline the development of high-performance GPU/CPU/Accelerators kernels

Lightning-Fast RL for LLM Reasoning and Agents. Made Simple & Flexible.

Agent Reinforcement Trainer: train multi-step agents for real-world tasks using GRPO. Give your agents on-the-job training. Reinforcement learning for Qwen2.5, Qwen3, Llama, and more!

Qwen-Image is a powerful image generation foundation model capable of complex text rendering and precise image editing.

Hierarchical Reasoning Model Official Release

Efficient GPU communication over multiple NICs.

Claude Code is an agentic coding tool that lives in your terminal, understands your codebase, and helps you code faster by executing routine tasks, explaining complex code, and handling git workflo…

Code for Enabling Realtime Reinforcement Learning at Scale with Staggered Asynchronous Inference

Mirage Persistent Kernel: Compiling LLMs into a MegaKernel

Softlearning is a reinforcement learning framework for training maximum entropy policies in continuous domains. Includes the official implementation of the Soft Actor-Critic algorithm.

UCCL is an efficient communication library for GPUs, covering collectives, P2P (e.g., KV cache transfer, RL weight transfer), and EP (e.g., GPU-driven)

Official Repository of Absolute Zero Reasoner

ACI.dev is the open source tool-calling platform that hooks up 600+ tools into any agentic IDE or custom AI agent through direct function calling or a unified MCP server. The birthplace of VibeOps.

Minimalistic 4D-parallelism distributed training framework for education purpose

Official Repository of "Learning to Reason under Off-Policy Guidance"