A light weight neural network optimizer framework for different software/hardware backends. The first version is still under development.

Overview

LightNet is a lightweight neural network optimizer framework for different software/hardware backends.

See Introduction for the detailed introduction. See LightNet for its full documentation.

Installation

Requirements

The following steps have been tested for Ubuntu 16.04 but should work with other distros as well.

Most required packages can be installed using the following commands (sudo permission may be required):

apt-get install build-essential perl git pkg-config check
cpan install JSON Sort::strverscmp

This project also depends on TensorLight, a lightweight tensor operation library. Install it according to its repository before continuing to build LightNet.

• (Optional) Packages for building documents

  Use the following commands to install the packages for building documents:

    apt-get install python3-pip
    pip3 install mkdocs markdown==3.1.1
  

• (Optional) Packages for building python packages

  Use the following commands to install the packages for building python packages (python3 for example):

    apt-get install python3-setuptools
  

• (Optional) CUDA dependency

  You also have to install CUDA 8.0 (or later) according to its website CUDA Toolkit if you want to build with CUDA support. Remember to put nvcc (usually in /usr/local/cuda/bin) in environment variable PATH.

• (Optional) cuDNN dependency

  You also have to install CUDA 8.0 (or later) and cuDNN 7.0 (or later) libraries, according to their websites CUDA Toolkit and cuDNN if you want to build with cuDNN support.

• (Optional) TensorRT dependency

  You also have to install CUDA 8.0 (or later) and TensorRT 3.0 (or later) libraries, according to their websites CUDA Toolkit and TensorRT if you want to build with TensorRT support.

Building and Installation

1. Clone this repository to your local directory.

    cd <my_working_directory>
    git clone https://github.com/zhaozhixu/LightNet.git
    cd LightNet
   

2. Configure and build

   First, configure your installation using:

    chmod +x configure
    ./configure
   

   There are options to custom your building and installation process. You can append them after ./configure. For example, use

    ./configure --install-dir=DIR
   

   to set the installation directory (default is /usr/local). Use

    ./configure --with-cuda=yes
   

   if you want to build with CUDA support.

   Detailed ./configure options can be displayed using ./configure -h.

   After that, use make to compile the binaries, and run the test. Finally, use make install to install the build directory into the installation directory.

3. Other make options

   Use make info to see other make options. Especially, you can use make clean to clean up the build directory and all object files, and make uninstall to remove installed files from the installation directory.

Usage

After compilation, the following components will be installed:

• lightnet: LightNet command tool
• liblightnet.so: LightNet runtime library
• ir2json.pl: LightNet intermediate language (IL) interpreter
• pylightnet (optional): LightNet python wrapper package

LightNet provides 3 interfaces which can be used by developers and users:

• Command line
• C API
• Python API

Command line

From the command line, type lightnet -h can display the program usage, as follows.

Usage: lightnet [OPTION...] SOURCE
Apply compilation procedures to SOURCE according to the options.
If SOURCE is -, read standard input.

Options:
  -h, --help             display this message
  -v, --version          display version information
  -o, --outfile=FILE     specify output file name; if FILE is -, print to
                         standard output; if FILE is !, do not print;
                         (default: out.json)
  -t, --target=TARGET    specify target platform (default: cpu)
  -f, --datafile=FILE    specify tensor data file
  -c, --compile          compile only; do not run
  -r, --run              run only; do not compile; SOURCE should have been
                         memory-planned
  -Wwarn                 display warnings (default)
  -w, -Wno-warn          do not display warnings
  -Winter                display internal warnings (default)
  -Wno-inter             do not display internal warnings
  -debug                 display debug messages (only works with LN_DEBUG
                         defined when compiling)