A high-performance RL training-inference weight synchronization framework, designed to enable second-level parameter updates from training to inference in RL workflows

IEEE 754-style floating-point converter

Tile primitives for speedy kernels

Financial data platform for analysts, quants and AI agents.

[ICLR2025, ICML2025, NeurIPS2025 Spotlight] Quantized Attention achieves speedup of 2-5x compared to FlashAttention, without losing end-to-end metrics across language, image, and video models.

This is a series of GPU optimization topics. Here we will introduce how to optimize the CUDA kernel in detail. I will introduce several basic kernel optimizations, including: elementwise, reduce, s…

CUDA 算子手撕与面试指南

Efficent platform for inference and serving local LLMs including an OpenAI compatible API server.

[🔥updating ...] AI 自动量化交易机器人(完全本地部署) AI-powered Quantitative Investment Research Platform. 📃 online docs: https://ufund-me.github.io/Qbot ✨ :news: qbot-mini: https://github.com/Charmve/iQuant

Material for gpu-mode lectures

基于Python的开源量化交易平台开发框架

IPTV直播源抓取 自动整合hao趣网直播源+TVBox直播源+其他网上直播源 择取分辨率、速度最佳视频流 定期更新

[MLSys'25] QServe: W4A8KV4 Quantization and System Co-design for Efficient LLM Serving; [MLSys'25] LServe: Efficient Long-sequence LLM Serving with Unified Sparse Attention

Code base and slides for ECE408：Applied Parallel Programming On GPU.

📚A curated list of Awesome LLM/VLM Inference Papers with Codes: Flash-Attention, Paged-Attention, WINT8/4, Parallelism, etc.🎉

CUDA/Metal accelerated language model inference

LLM training in simple, raw C/CUDA

Large Language Model Text Generation Inference

Lightning fast C++/CUDA neural network framework

This project aim to reproduce Sora (Open AI T2V model), we wish the open source community contribute to this project.

Fast Multimodal LLM on Mobile Devices

FlashInfer: Kernel Library for LLM Serving

Full and up-to-date source code of the chapters of the "SFML Game Development" book

LLM Inference benchmark

how to learn PyTorch and OneFlow