crapromer

rssbox-android

It is a rss reader for android. Based on Rust and Slint-ui.

cudnn-memo

Example code of cuDNN.

CUDA

CUDA and cuDNN examples

gpu-tensor-core

A set of programs testing CUDA Tensor Core performance

TiledCUDA

TiledCUDA is an efficient kernel template library written in CuTe, which provides a wrapper for cutlass CuTe and enables more efficient fusion.

CLImage

A C++ GPGPU OpenCL library for Android and Unix systems.

operators

算子库

llama3-from-scratch

llama3 implementation one matrix multiplication at a time

mnistCUDNN

instance_norm

CUDA-Learn-Notes

🎉CUDA 笔记 / 大模型手撕CUDA / C++笔记，更新随缘: flash_attn、sgemm、sgemv、warp reduce、block reduce、dot product、elementwise、softmax、layernorm、rmsnorm、hist etc.

mlops-coding-course

Learn how to create, develop, and maintain a state-of-the-art MLOps code base

cucumat

private implementation of cudamat

MapEditor

a simple ui pack based on python and pyqt5 target to simulate the appearance of imgui

cuda-repo

From zero to hero CUDA for accelerating maths and machine learning on GPU.

CompPhys

CompPhys - a Computational Physics repository

PyTorch-GAN

PyTorch implementations of Generative Adversarial Networks.

How_to_optimize_in_GPU

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, sgemv, sgemm, etc. The performance of these kernels is basically at or near the theoretical limit.

mlmodelscope

MLModelScope is an open source, extensible, and customizable platform to facilitate evaluation and measurement of ML models within AI pipelines.

onnx2X

ONNX2Pytorch

CudaGeMM

Benchmarking different cuda GeMM kernels

cuda_gemm_benchmark

Base on gtest/benchmark, refer to https://github.com/Liu-xiandong/How_to_optimize_in_GPU

llm.c

LLM training in simple, raw C/CUDA

CUDA_Bench

CUDA GPU Benchmark

CUDA_Freshman

ABigSurveyOfLLMs

A collection of 150+ surveys on LLMs

Fast-TransX

An Efficient implementation of TransE and its extended models for Knowledge Representation Learning

OpenKE

An Open-Source Package for Knowledge Embedding (KE)

tqdm.cpp

C++ port of tqdm

Chinese-LLaMA-Alpaca-2

中文LLaMA-2 & Alpaca-2大模型二期项目 + 64K超长上下文模型 (Chinese LLaMA-2 & Alpaca-2 LLMs with 64K long context models)