Stars
Soku CLI — call Soku ads/GA4/PostHog data capabilities from any AI agent or shell.
Fast, accurate & comprehensive text measurement & layout
👻 Ghostty is a fast, feature-rich, and cross-platform terminal emulator that uses platform-native UI and GPU acceleration.
Framework for evaluating and improving agents
Google Workspace CLI — one command-line tool for Drive, Gmail, Calendar, Sheets, Docs, Chat, Admin, and more. Dynamically built from Google Discovery Service. Includes AI agent skills.
🚀 Pytorch Distributed native training library for LLMs/VLMs with OOTB Hugging Face support
DFlash: Block Diffusion for Flash Speculative Decoding
Algorithm powering the For You feed on X
Source code for the X Recommendation Algorithm
Pocket Flow: 100-line LLM framework. Let Agents build Agents!
Domain-specific language designed to streamline the development of high-performance GPU/CPU/Accelerators kernels
Accelerating MoE with IO and Tile-aware Optimizations
An extremely fast Python type checker and language server, written in Rust.
An extremely fast Python linter and code formatter, written in Rust.
An extremely fast Python package and project manager, written in Rust.
The simplest, fastest repository for training/finetuning medium-sized GPTs.
A toolchain for web projects, aimed to provide functionalities to maintain them. Biome offers formatter and linter, usable via CLI and LSP.
Virtual whiteboard for sketching hand-drawn like diagrams
🧠「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!
Fast and memory-efficient exact attention
Genai-bench is a powerful benchmark tool designed for comprehensive token-level performance evaluation of large language model (LLM) serving systems.
slime is an LLM post-training framework for RL Scaling.
learn LLM inference on Apple Silicon for systems engineers: build a tiny vLLM + Qwen
🚀 Efficient implementations for emerging model architectures
🙌 OpenHands: AI-Driven Development