Starred repositories
Reusable AI agent skills for Codex, including a PowerShell skill for safe Windows command invocation and optional MCP utilities.
Hy3 (295B A21B), a leading reasoning and agent model in its size, with great cost efficiency.
Minimal Eagle3 draft model trainer for speculative decoding
DFlash: Block Diffusion for Flash Speculative Decoding
Architecture wiki for the open-sourced X "For You" recommendation algorithm (xai-org/x-algorithm) — 21 source-anchored pages
🐹 Clean, uninstall, analyze, optimize, and monitor your Mac from the terminal.
Official code implementation of Context Cascade Compression: Exploring the Upper Limits of Text Compression
LLM model quantization (compression) toolkit with HW acceleration support for Nvidia, AMD, Intel GPU and Intel/AMD/Apple CPU via HF, vLLM, and SGLang.
The agent that grows with you
[EMNLP'23, ACL'24] To speed up LLMs' inference and enhance LLM's perceive of key information, compress the prompt and KV-Cache, which achieves up to 20x compression with minimal performance loss.
Provide with pre-build flash-attention 2 and 3 package wheels on Linux and Windows using GitHub Actions
Former GUI client for gRPC services. No longer maintained.
Give your AI agent eyes to see the entire internet. Read & search Twitter, Reddit, YouTube, GitHub, Bilibili, XiaoHongShu — one CLI, zero API fees.
AI agents running research on single-GPU nanochat training automatically
本文旨在整理文本生成领域国内外工业界和企业家的研究者和研究机构。排名不分先后。更新中，欢迎大家补充
Your own personal AI assistant. Any OS. Any Platform. The lobster way. 🦞
omo/lazycodex: The coding agent for tokenmaxxers;the one and only agent harness for complex codebases. For your Codex, for your OpenCode
An Open Foundation Model and Benchmark to Accelerate Generative Recommendation
LLM 驱动的多市场股票智能分析系统：多源行情、实时新闻、决策看板与自动推送，支持零成本定时运行。 LLM-powered multi-market stock analysis system with multi-source market data, real-time news, decision dashboard, automated notifications, and cost…
An AI prompt optimizer for writing better prompts and getting better AI results.
仍然可用的微信RPA！支持4.0系列微信pywechat是一个基于pywinauto实现的windows桌面微信自动化操作工具，基本实现了PC微信内置的各项操作。
使用hook实现微信机器人，可用于公众号更新和历史采集，自动同意好友请求，自动邀请进群等
Awesome papers in Large Language Models (LLM). They focus on state-of-the-art LLM methods, such as algorithms, system, SFT, RL, Multi-modal LLMs, MOE, Quantization, and Applications (RAG, agent, co…
Spec-driven development (SDD) for AI coding assistants.
Masked Depth Modeling for Spatial Perception