DefTruth DefTruth

🎯

#pragma unroll

AI Infra Engineer @vipshop, Owner @xlite-dev, Prev @PaddlePaddle🤖

Achievements

DefTruth/README.md

🤗🎉 Hi! I'm DefTruth, an AI Infra Engineer (DiT / LLM). I built xlite-dev, Cache-DiT, ffpa-attn, LeetCUDA, lite.ai.toolkit, lihang-notes, torchlm, Awesome-LLM-Inference, Awesome-DiT-Inference | 🎯🦅 I also contributed to FastDeploy, SGLang , SGLang-Diffusion , vLLM , vLLM-Omni , Diffusers , GCMP | 🚪🐍 I also built some open-source books at 📖 LeetCUDA, 📖 lihang-notes.

xlite-dev/LeetCUDA xlite-dev/LeetCUDA Public

An Open-source book with Modern CUDA learning Notes for Beginners - BF16/FP8/FP4, HGEMM, CuTe, Flash-Attention, etc.

TeX 12.2k 1.3k
xlite-dev/lite.ai.toolkit xlite-dev/lite.ai.toolkit Public

A lite C++ AI toolkit: 100+ models with MNN, ORT and TRT, including Det, Seg, Stable-Diffusion, Face-Fusion.

C++ 4.4k 784
PaddlePaddle/FastDeploy PaddlePaddle/FastDeploy Public

High-performance Inference and Deployment Toolkit for LLMs and VLMs based on PaddlePaddle

Python 3.7k 757
sgl-project/sglang sgl-project/sglang Public

SGLang is a high-performance serving framework for large language models and multimodal models.

Python 38.8k 10.4k
vipshop/cache-dit vipshop/cache-dit Public

A PyTorch-native inference engine with cache, parallelism, quantization and cpu offload for 100+ DiTs.

Python 1.3k 94
vipshop/ffpa-attn vipshop/ffpa-attn Public

Kernel Library for Large Headdim Attention (64~1024, BF16/FP8/FP4), 1.5x~15x↑ vs PyTorch SDPA.

Python 348 29

Something went wrong, please refresh the page to try again.
If the problem persists, check the GitHub status page or contact support.

There was an error while loading. Please reload this page.