DeepSpeed is a deep learning optimization library that makes distributed training and inference easy, efficient, and effective.

MLX: An array framework for Apple silicon

LXC - Linux Containers

[NeurIPS'23 Oral] Visual Instruction Tuning (LLaVA) built towards GPT-4V level capabilities and beyond.

Tiny data-over-sound library

🎤⌨️ Acoustic keyboard eavesdropping

Port of OpenAI's Whisper model in C/C++

Tensor library for machine learning

LLM inference in C/C++

OpenAssistant is a chat-based assistant that understands tasks, can interact with third-party systems, and retrieve information dynamically to do so.

An example pipeline that builds, tests, and packages a Python library using Bazel.

QICK: Quantum Instrumentation Control Kit

A General-purpose Task-parallel Programming System using Modern C++

Virtual machines for iOS and macOS

Tetris™ clone created in c++ and SDL2

A framework for distributed applications

Robust Speech Recognition via Large-Scale Weak Supervision

DSPy: The framework for programming—not prompting—language models

Optax is a gradient processing and optimization library for JAX.

Composable transformations of Python+NumPy programs: differentiate, vectorize, JIT to GPU/TPU, and more

Open-source, Google Sheets Custom Functions built using Apps Script.

Minimal reproduction of DeepSeek R1-Zero

Examples and guides for using the Gemini API

A developer guide for Gemini's Multimodal Live API

A free and strong UCI chess engine

Code for the manim-generated scenes used in 3blue1brown videos

A speed reading Google Chrome extension (via Rapid Serial Visual Presentation).

jupyter widgets made easy

LLM training in simple, raw C/CUDA

A collection of examples using emscripten for audio and graphics for ADC2018