Stars
AI 基础知识 - GPU 架构、CUDA 编程、大模型基础及AI Agent 相关知识。
DeepSeek 4 Flash and PRO local inference engine for Metal, CUDA and ROCm
MoonEP: A Perfectly Balanced Expert Parallelism Library via Dynamic Redundant Experts
SGLang is a high-performance serving framework for large language models and multimodal models.
verl/HybridFlow: A Flexible and Efficient RL Post-Training Framework
End-to-End NPU: RTL systolic array → AXI4 SoC → TVM compiler → FPGA inference
ESP32 desk dashboard that shows Claude Code usage
See your claude code logs in clear details in your dashboard
Generate Excalidraw diagrams from Mermaid
Agentic Chat Interface directly in Neovim with ACP providers from Claude-Code, Gemini, Codex, OpenCode, and Cursor-agent
Tmux on Windows Powershell - tmux for PowerShell, Windows Terminal, cmd.exe. Includes psmux, pmux, and tmux commands. This is native High-Performance Tmux designed for Windows in Rust 🦀
Oh my tmux! My self-contained, pretty & versatile tmux configuration made with 💛🩷💙🖤❤️🤍
A next.js web application that integrates AI capabilities with draw.io diagrams. This app allows you to create, modify, and enhance diagrams through natural language commands and AI-assisted visual…
A from-scratch Prefill/Decode disaggregation inference engine for LLMs
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)
Ray is an AI compute engine. Ray consists of a core distributed runtime and a set of AI Libraries for accelerating ML workloads.
Home for "How To Scale Your Model", a short blog-style textbook about scaling LLMs on TPUs
Ultra-lightweight, open-source, self-hosted personal AI agent framework in Python with WebUI, tools, memory, MCP, multi-agent workflows, automation, and chat apps
Domain-specific language designed to streamline the development of high-performance GPU/CPU/Accelerators kernels
The RL Bridge for LLM-based Agent Applications. Made Simple & Flexible.
LMCache: Supercharge Your LLM with the Fastest KV Cache Layer
😼 优雅地使用基于 clash/mihomo 的代理环境