Skip to content

@xlite-dev xlite-dev

All

    Repositories list

    75 repositories

    • LeetCUDA

      Public
      An Open-source book with Modern CUDA learning Notes for Beginners - BF16/FP8/FP4, HGEMM, CuTe, Flash-Attention, etc.
      cudacuda-kernelscuda-demo
      cudacuda-kernelscuda-democuda-toolkitcuda-librarycuda-kernellearn-cudacuda-cpphgemmflash-attention
      Cuda
      •
      GNU General Public License v3.0
      •1.3k•12k•3•0•Updated Oct 18, 2026Oct 18, 2026

    • ffpa-attn

      Public
      Kernel Library for Large Headdim Attention (64~1024, BF16/FP8/FP4), 1.5x~15x↑ vs PyTorch SDPA.
      Python
      •
      Apache License 2.0
      •28•0•0•0•Updated Oct 12, 2026Oct 12, 2026

    • sglang

      Public
      SGLang is a fast serving framework for large language models and vision language models.
      Python
      •
      Apache License 2.0
      •9.7k•3•0•0•Updated Oct 5, 2026Oct 5, 2026

    • GCMP

      Public
      通过集成国内主流原生大模型提供商，为开发者提供更加丰富、更适合本土需求的 AI 编程助手选择。 目前已内置支持 智谱AI、MiniMax、MoonshotAI、DeepSeek、阿里云百炼、快手万擎、火山方舟、腾讯云、Xiaomi MiMo 等原生大模型提供商。 此外，扩展插件已适配支持 OpenAI 与 Anthro…
      TypeScript
      •
      MIT License
      •60•0•0•0•Updated Sep 27, 2026Sep 27, 2026

    • .github

      Public
      0•1•0•0•Updated Sep 25, 2026Sep 25, 2026

    • lihang-notes

      Public
      📚《统计学习方法-李航: 笔记》 200页PDF，包含详细的公式细节讲解以及R语言实现 🎉
      svmkd-treeml
      svmkd-treemlstatistics-learninglihangli-hanglihang-codelihang-notesli-hang-notes
      TeX
      •
      GNU General Public License v3.0
      •62•512•2•0•Updated Sep 22, 2026Sep 22, 2026

    • DeepGEMM

      Public
      DeepGEMM: clean and efficient FP8 GEMM kernels with fine-grained scaling
      Cuda
      •
      MIT License
      •1.4k•0•0•0•Updated Sep 22, 2026Sep 22, 2026

    • diffusers

      Public
      🤗 Diffusers: State-of-the-art diffusion models for image, video, and audio generation in PyTorch and FLAX.
      Python
      •
      Apache License 2.0
      •7.4k•0•0•0•Updated Sep 18, 2026Sep 18, 2026

    • lite.ai.toolkit

      Public
      A lite C++ AI toolkit: 100+ models with MNN, ORT and TRT, including Det, Seg, Stable-Diffusion, Face-Fusion.
      tensorrtmnnncnn
      tensorrtmnnncnnonnxonnxruntimeyolov5tnnmnn-modelyoloxrobustvideomatting
      C++
      •
      GNU General Public License v3.0
      •783•4.4k•0•1•Updated Sep 5, 2026Sep 5, 2026

    • flashinfer

      Public
      FlashInfer: Kernel Library for LLM Serving
      Python
      •
      Apache License 2.0
      •1.6k•0•0•0•Updated Sep 3, 2026Sep 3, 2026

    • flash-linear-attention

      Public
      🚀 Efficient implementations for emerging model architectures
      Python
      •
      MIT License
      •779•1•0•0•Updated Aug 27, 2026Aug 27, 2026

    • Awesome-LLM-Inference

      Public
      📚A curated list of Awesome LLM/VLM Inference Papers with Codes: Flash-Attention, Paged-Attention, WINT8/4, Parallelism, etc.🎉
      mlavllmllm-inference
      mlavllmllm-inferenceawesome-llmflash-attentiontensorrt-llmpaged-attentiondeepseekflash-attention-3deepseek-v3
      Python
      •
      GNU General Public License v3.0
      •438•5.5k•0•10•Updated Aug 14, 2026Aug 14, 2026

    • quack

      Public
      A Quirky Assortment of CuTe Kernels
      Python
      •
      Apache License 2.0
      •166•1•0•0•Updated Aug 5, 2026Aug 5, 2026

    • cutlass

      Public
      CUDA Templates and Python DSLs for High-Performance Linear Algebra
      C++
      •
      Other
      •2.1k•1•0•0•Updated Aug 5, 2026Aug 5, 2026

    • MSA

      Public
      Python
      •
      MIT License
      •66•0•0•0•Updated Jul 30, 2026Jul 30, 2026

    • CuTe

      Public
      Reference implementation and examples of the CuTe Layout representation and algebra.
      Python
      •
      Apache License 2.0
      •41•0•0•0•Updated Jul 28, 2026Jul 28, 2026

    • cudnn-frontend

      Public
      cuDNN Frontend is NVIDIA's modern, open-source entry point to the cuDNN library and a growing collection of high-performance open-source kernels.
      Python
      •
      MIT License
      •312•0•0•0•Updated Jul 14, 2026Jul 14, 2026

    • TurboDiffusion

      Public
      TurboDiffusion: 100–200× Acceleration for Video Diffusion Models
      Python
      •
      Apache License 2.0
      •277•0•0•0•Updated Jul 14, 2026Jul 14, 2026

    • flash-attention

      Public
      Fast and memory-efficient exact attention
      Python
      •
      BSD 3-Clause "New" or "Revised" License
      •3.2k•0•0•0•Updated Jul 13, 2026Jul 13, 2026

    • modern-gpu-programming-for-mlsys

      Public
      modern gpu programming
      HTML
      •146•2•0•0•Updated Jun 28, 2026Jun 28, 2026

    • vllm-omni

      Public
      A framework for efficient model inference with omni-modality models
      Python
      •
      Apache License 2.0
      •1.9k•0•0•0•Updated Jun 26, 2026Jun 26, 2026

    • vllm

      Public
      A high-throughput and memory-efficient inference and serving engine for LLMs
      Python
      •
      Apache License 2.0
      •24k•0•0•0•Updated Jun 26, 2026Jun 26, 2026

    • deepseek-v4-for-copilot

      Public
      Pick DeepSeek V4 from the Copilot Chat model picker — and keep everything else Copilot already gives you.
      TypeScript
      •
      MIT License
      •146•0•0•0•Updated Jun 20, 2026Jun 20, 2026

    • DeepEP

      Public
      DeepEP: an efficient expert-parallel communication library
      Cuda
      •
      MIT License
      •1.5k•0•0•0•Updated Jun 15, 2026Jun 15, 2026

    • Awesome-DiT-Inference

      Public
      📚A curated list of Awesome Diffusion Inference Papers with Codes: Sampling, Cache, Quantization, Parallelism, etc.🎉
      fluxwandiffusion
      fluxwandiffusionditsorastable-diffusionsdxlsd15deepcacheopen-sora-plan
      Python
      •
      GNU General Public License v3.0
      •29•591•1•1•Updated Jun 13, 2026Jun 13, 2026

    • cache-dit

      Public
      A PyTorch-native inference engine with cache, parallelism, quantization for Diffusion Transformers.
      Python
      •
      Apache License 2.0
      •92•4•0•0•Updated Jun 13, 2026Jun 13, 2026

    • cllms-for-copilot

      Public
      Pick Qwen, GLM, MiniMax, Xiaomi MiMo, Moonshot Kimi & Tencent Hunyuan models from the Copilot Chat model picker. Vision, thinking, BYOK.
      TypeScript
      •
      MIT License
      •2•1•0•0•Updated Jun 13, 2026Jun 13, 2026

    • svdquant-kernels

      Public
      Cross-architecture CUDA kernels for SVDQuant (W4A4 with low-rank correction)
      Python
      •
      Apache License 2.0
      •4•3•0•0•Updated May 28, 2026May 28, 2026

    • FlashMLA

      Public
      FlashMLA: Efficient Multi-head Latent Attention Kernels
      C++
      •
      MIT License
      •1.2k•1•0•0•Updated Apr 30, 2026Apr 30, 2026

    • TileKernels

      Public
      A kernel library written in tilelang
      Python
      •
      MIT License
      •204•0•0•0•Updated Apr 23, 2026Apr 23, 2026
    ProTip! When viewing an organization's repositories, you can use the props. filter to filter by custom property.