Stars
Claude Code 通过 MCP 跑异步长任务（比如指挥 Codex 干活）时，无法订阅服务端的事件，也拿不到任务的最新进展。本项目基于 MCP 标准协议实现了通知订阅，替客户端补足缺口，让 Claude Code 或任意其它客户端都能订阅服务端消息。
[🔥updating ...] AI 自动量化交易机器人(完全本地部署) AI-powered Quantitative Investment Research Platform. 📃 online docs: https://ufund-me.github.io/Qbot ✨ :news: qbot-mini: https://github.com/Charmve/iQuant
API, CLI, and Web App for analyzing and finding a person's profile in 1000 social media \ websites
微舆：人人可用的多Agent舆情分析助手，打破信息茧房，还原舆情原貌，预测未来走向，辅助决策！从0实现，不依赖任何框架。
🧠「大模型」2小时完全从0训练64M的小参数LLM！Train a 64M-parameter LLM from scratch in just 2h!
The official implementation for [NeurIPS2025 Oral] Gated Attention for Large Language Models: Non-linearity, Sparsity, and Attention-Sink-Free
[ICLR2025, ICML2025, NeurIPS2025 Spotlight] Quantized Attention achieves speedup of 2-5x compared to FlashAttention, without losing end-to-end metrics across language, image, and video models.
Janus-Series: Unified Multimodal Understanding and Generation Models
百聆 是一个类似GPT-4o的语音对话机器人，通过ASR+LLM+TTS实现，集成DeepSeek R1等优秀大模型，接入openClaw，真正的个人语音助手，时延低至800ms，Mac等低配置也可运行，支持打断
CosyVoice_DPO_NOTES: Supercharge Your Cosyvoice model with Cutting-Edge DPO Fine-Tuning!
Python library to easily handle JSON contained within strings.
A guidance language for controlling large language models.
The simplest, fastest repository for training/finetuning medium-sized GPTs.
LLM model quantization (compression) toolkit with HW acceleration support for Nvidia, AMD, Intel GPU and Intel/AMD/Apple CPU via HF, vLLM, and SGLang.
An Industrial-Level Controllable and Efficient Zero-Shot Text-To-Speech System
An optimized quantization and inference library for running LLMs locally on modern consumer-class GPUs
Programmer's guide about how to cook at home.
The source code for the blog post The 37 Implementation Details of Proximal Policy Optimization
AI Infra / AI Orchestration / AI Control Plane
Train transformer language models with reinforcement learning.
The local UI to run and train text and diffusion models, including Kimi K3, Gemma 4, Qwen3.6, DeepSeek-V4, FLUX and more.
A Flexible Framework for Experiencing Heterogeneous LLM Inference/Fine-tune Optimizations
SGLang is a high-performance serving framework for large language models and multimodal models.
LiberSonora，寓意“自由的声音”，是一个 AI 赋能的、强大的、开源有声书工具集，包含智能字幕提取、AI标题生成、多语言翻译等功能，支持 GPU 加速、批量离线处理。LiberSonora, meaning "The Voice of Freedom," is an AI-powered robust open-source audiobook toolkit.
🦄🦄🦄AI赋能股票分析：AI加持的股票分析/选股工具。股票行情获取，AI热点资讯分析，AI资金/财务分析，涨跌报警推送。支持A股，港股，美股。支持市场整体/个股情绪分析，AI辅助选股等。数据全部保留在本地。支持DeepSeek，OpenAI， Ollama，LMStudio，AnythingLLM，硅基流动，火山方舟，阿里云百炼等平台或模型。
Fully open reproduction of DeepSeek-R1
DeepSeek Coder: Let the Code Write Itself