A reactive notebook for Python — run reproducible experiments, query with SQL, execute as a script, deploy as an app, and version with git. Stored as pure Python. All in a modern, AI-native editor.

Pie: Programmable LLM Serving

Open-source implementation of AlphaEvolve

Tempo is a system for declarative, efficient, end-to-end compiled dynamic deep learning

LLMServingSim: A HW/SW Co-Simulation Infrastructure for LLM Inference Serving at Scale

🌎💪 BrowserGym, a Gym environment for web task automation

verl: Volcano Engine Reinforcement Learning for LLMs

gpt-oss-120b and gpt-oss-20b are two open-weight language models by OpenAI

A library for advanced large language model reasoning

[DEPRECATED] Moved to ROCm/rocm-systems repo

Causal depthwise conv1d in CUDA, with a PyTorch interface

Deep learning at the speed of light.

OpenR: An Open Source Framework for Advanced Reasoning with Large Language Models

Qwen3 is the large language model series developed by Qwen team, Alibaba Cloud.

The Lean version manager

Mooncake is the serving platform for Kimi, a leading LLM service provided by Moonshot AI.

Machine Learning Engineering Open Book

Tile primitives for speedy kernels

A curated list of resources for learning and exploring Triton, OpenAI's programming language for writing efficient GPU code.

DeepEP: an efficient expert-parallel communication library

FlashMLA: Efficient Multi-head Latent Attention Kernels

Minimalistic large language model 3D-parallelism training

MoBA: Mixture of Block Attention for Long-Context LLMs

Minimalistic 4D-parallelism distributed training framework for education purpose

CVPR and NeurIPS poster examples and templates. May we have in-person poster session soon!

A PyTorch native platform for training generative AI models

Tips for Writing a Research Paper using LaTeX

Code for "LayerSkip: Enabling Early Exit Inference and Self-Speculative Decoding", ACL 2024