Compact and Agent-Native MoE Training System

AISystem 主要是指AI系统，包括AI芯片、AI编译器、AI推理和训练框架等AI全栈底层技术

Standalone Flash Attention v2 kernel without libtorch dependency

Mirage Persistent Kernel: Compiling LLMs into a MegaKernel

arXiv LaTeX Cleaner: Easily clean the LaTeX code of your paper to submit to arXiv

My personal frontpage app

Papers and resources related to the security and privacy of LLMs 🤖

SpotServe: Serving Generative Large Language Models on Preemptible Instances

Automatically Discovering Fast Parallelization Strategies for Distributed Deep Neural Network Training

Wechat Chat History Exporter 微信聊天记录导出备份程序

The AI Compute Platform for frontier teams. SkyPilot turns fragmented AI compute into one AI supercomputer, so frontier AI teams build custom intelligence faster.

[ICCV 2023] PointCLIP V2: Prompting CLIP and GPT for Powerful 3D Open-world Learning

[ICCV 2023] The first DETR model for monocular 3D object detection with depth-guided transformer

A list of awesome compiler projects and papers for tensor computation and deep learning.

Reduce end to end training time from days to hours (or hours to minutes), and energy requirements/costs by an order of magnitude using coresets and data selection.

Minkowski Engine is an auto-diff neural network library for high-dimensional sparse tensors

A curated reading list of research in Mixture-of-Experts(MoE).

我的各类笔记（~ 至 2016.12）

A fully adaptive, zero-tuning parameter manager that enables efficient distributed machine learning training

MLNLP社区用来帮助缩短参考文献的工具。A tool for simplifying bibtex with official info

A high-performance distributed deep learning system targeting large-scale and automated distributed training.

Towards Generalized and Efficient Blackbox Optimization System/Package (KDD 2021 & JMLR 2024)

[CVPR 2022] PointCLIP: Point Cloud Understanding by CLIP

PKU一键出入校备案小工具

Submarine is Cloud Native Machine Learning Platform.

OneFlow is a deep learning framework designed to be user-friendly, scalable and efficient.