Stars
A high-throughput and memory-efficient inference and serving engine for LLMs
Fine-tuning & Reinforcement Learning for LLMs. 🦥 Train OpenAI gpt-oss, DeepSeek-R1, Qwen3, Gemma 3, TTS 2x faster with 70% less VRAM.
A toolkit for developing and comparing reinforcement learning algorithms.
Run your own AI cluster at home with everyday devices 📱💻 🖥️⌚
🤗 Diffusers: State-of-the-art diffusion models for image, video, and audio generation in PyTorch.
A generative world for general-purpose robotics & embodied AI learning.
Fully open reproduction of DeepSeek-R1
Code for the paper "Language Models are Unsupervised Multitask Learners"
Distribute and run LLMs with a single file.
Build resilient language agents as graphs.
Educational framework exploring ergonomic, lightweight multi-agent orchestration. Managed by OpenAI Solution team.
SGLang is a fast serving framework for large language models and vision language models.
gpt-oss-120b and gpt-oss-20b are two open-weight language models by OpenAI
✨✨Latest Advances on Multimodal Large Language Models
Train transformer language models with reinforcement learning.
A Flexible Framework for Experiencing Cutting-edge LLM Inference Optimizations
Ongoing research training transformer models at scale
A tiny scalar-valued autograd engine and a neural net library on top of it with PyTorch-like API
Code for loralib, an implementation of "LoRA: Low-Rank Adaptation of Large Language Models"
Minimal reproduction of DeepSeek R1-Zero
FlashMLA: Efficient Multi-head Latent Attention Kernels
A minimal GPU design in Verilog to learn how GPUs work from the ground up
DeepEP: an efficient expert-parallel communication library
[NeurIPS 2024 Best Paper Award][GPT beats diffusion🔥] [scaling laws in visual generation📈] Official impl. of "Visual Autoregressive Modeling: Scalable Image Generation via Next-Scale Prediction". A…
An Easy-to-use, Scalable and High-performance RLHF Framework based on Ray (PPO & GRPO & REINFORCE++ & vLLM & Ray & Dynamic Sampling & Async Agentic RL)
Production-tested AI infrastructure tools for efficient AGI development and community-driven innovation
Agent Reinforcement Trainer: train multi-step agents for real-world tasks using GRPO. Give your agents on-the-job training. Reinforcement learning for Qwen2.5, Qwen3, Llama, and more!