sean-m
/
nncase
mirror of https://github.com/kendryte/nncase.git
1
0
Fork 0
Code Issues Releases Wiki Activity
1,170 Commits
37 Branches
52 Tags
123 MiB
C# 56.8%
C++ 31%
Python 9.2%
CMake 1.4%
C 1.2%
Other 0.4%
 
 
 
 
 
 
Go to file
sunnycase cd1c64824c Move stackvm 2022-05-13 17:00:36 +08:00
.github restore ci 2022-04-14 15:11:19 +08:00
benchmark Split runtime function from runtime module (#337) 2021-08-18 18:56:39 +08:00
cmake Move stackvm 2022-05-13 17:00:36 +08:00
csharp Move stackvm 2022-05-13 17:00:36 +08:00
docs add k510 codegen 2022-01-17 15:03:46 +08:00
examples fix issue #554 2022-03-28 19:35:40 +08:00
modules Move stackvm 2022-05-13 17:00:36 +08:00
python Move stackvm 2022-05-13 17:00:36 +08:00
src Move stackvm 2022-05-13 17:00:36 +08:00
targets add conv2d glb buffer offset calc 2022-04-16 19:23:49 +08:00
tests Replace Torch with onnxruntime in Evaluator (#551) 2022-03-29 19:04:56 +08:00
third_party add test stage1 2021-12-17 10:50:50 +08:00
toolchains Add benchmark tool (#316) 2021-08-07 23:29:44 +08:00
tools Move stackvm 2022-05-13 17:00:36 +08:00
.clang-format Move stackvm 2022-05-13 17:00:36 +08:00
.gitattributes initial commit 2018-04-04 20:43:17 +08:00
.gitignore add nuget config 2022-02-15 17:21:36 +08:00
.gitmodules Remove onnx submodule 2021-01-18 19:42:41 +08:00
.pep8 Refactor test compare module (#283) 2021-07-14 11:56:59 +08:00
CMakeLists.txt Move stackvm 2022-05-13 17:00:36 +08:00
LICENSE Update license and readme 2019-03-01 11:51:02 +08:00
NuGet.Config Replace Torch with onnxruntime in Evaluator (#551) 2022-03-29 19:04:56 +08:00
README.md Add cibuildwheel (#330) 2021-08-15 18:52:40 +08:00
conanfile.py Move stackvm 2022-05-13 17:00:36 +08:00
nncase.sln Move stackvm 2022-05-13 17:00:36 +08:00
pyproject.toml Add cibuildwheel (#330) 2021-08-15 18:52:40 +08:00
setup.py Add cibuildwheel (#330) 2021-08-15 18:52:40 +08:00

README.md

nncase

License compiler-build

nncase is a neural network compiler for AI accelerators.

nncase 是一个为 AI 加速器设计的神经网络编译器。

技术交流 QQ 群：790699378

Telegram: nncase community

Install from binaries

从二进制安装

Download prebuilt binaries from Release.

下载预编译的二进制文件 Release。

Build from source

从源码编译

Build from source

Supported operators

支持的算子

Usage

使用方法

Resources

资源

Architecture

架构

nncase arch

Features

  • Supports multiple inputs and outputs and multi-branch structure
  • Static memory allocation, no heap memory acquired
  • Operators fusion and optimizations
  • Support float and quantized uint8 inference
  • Support post quantization from float model with calibration dataset
  • Flat model with zero copy loading

功能

  • 支持多输入输出网络，支持多分支结构
  • 静态内存分配，不需要堆内存
  • 算子合并和优化
  • 支持 float 和量化 uint8 推理
  • 支持训练后量化，使用浮点模型和量化校准集
  • 平坦模型，支持零拷贝加载