Skip to content
Additions and patches to Caffe framework for use with Synopsys DesignWare EV6x Processors
Branch: master
Clone or download
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
.github
cmake
data
densecrf
docker
docs
evaluation
examples
include
installation
matlab
models
python
scripts
src
tools
.Doxyfile
.gitattributes
.gitignore
.travis.yml
CMakeLists.txt
CONTRIBUTING.md
CONTRIBUTORS.md
FEATURES.md
INSTALL.md
LICENSE
Makefile
Makefile.config.example
README.md
appveyor.yml
caffe.cloc

README.md

Synopsys Caffe

Build Status

Synopsys Caffe is a modified version of the popular Caffe Deep Learning framework adapted for use with DesignWare EV6x Processors. It combines multiple customized branches and includes a large range of patches to support diverse models. See FEATURES.md for a short overview.

Installation

Please check out the prerequisites and read the detailed notes at the BVLC Caffe Installation if this is your first time to install Caffe.

Linux

A simple guide:

  1. Ensure that you have all the dependencies mentioned at the BVLC Caffe Installation for your OS installed (protobuf, leveldb, snappy, opencv, hdf5-serial, protobuf-compiler, BLAS, Python, CUDA etc.)
  2. Also Install matio in your environment. After that, add [your installed matio path]/lib to the LD_LIBRARY_PATH.
export LD_LIBRARY_PATH=[your installed matio path]/lib:${LD_LIBRARY_PATH}
  1. Checkout the Synopsys Caffe master branch. Configure the build by copying and modifying the example Makefile.config for your setup.
git clone https://github.com/foss-for-synopsys-dwc-arc-processors/synopsys-caffe.git
cd synopsys-caffe
cp Makefile.config.example Makefile.config
# Modify Makefile.config to suit your needs, e.g. enable/disable the CPU-ONLY, CUDNN, NCCL and set the path for CUDA, Python and BLAS.
# If needed, add [your installed matio path]/include to INCLUDE_DIRS and [your installed matio path]/lib to LIBRARY_DIRS.
  1. Build Caffe and run the tests.
make all
make pycaffe
make test
make runtest
# If no error occurs, you can add the caffe path to the environment for easy use.
export SYNOPSYS_CAFFE_HOME=[your synopsys-caffe root folder path]
export PATH=${SYNOPSYS_CAFFE_HOME}/build/tools:${PATH}
export PYTHONPATH=${SYNOPSYS_CAFFE_HOME}/python:${PYTHONPATH}

Windows

A simple guide:

  1. Download the Visual Studio 2015 Update 3 (Do not use the VS 2017, it is not supported!). Choose to install the support for visual C++ instead of applying the default settings.
  2. Install the CMake 3.4 or higher. Install Python 2.7 or 3.5. Add cmake.exe and python.exe to your PATH.
  3. After installing the Python, please open a cmd prompt and use pip install numpy to install the numpy package.
  4. Checkout the Synopsys Caffe master branch for build. The windows branch is deprecated, please do not use it. We use C:\Projects as the current folder for the following instructions.
C:\Projects> git clone https://github.com/foss-for-synopsys-dwc-arc-processors/synopsys-caffe.git
C:\Projects> cd synopsys-caffe
:: Edit any of the options inside build_win.cmd to suit your needs
C:\Projects\synopsys-caffe> scripts\build_win.cmd
:: If no error occurs, the caffe.exe will be created at C:\Projects\synopsys-caffe\build\tools\Release after a successful build.

Other detailed installation instructions can be found here.

Reference for different Distributions

Synopsys Caffe support the features introduced in following customized branches. Here are some links to the original demos, tutorials and models usage:

You can’t perform that action at this time.
You signed in with another tab or window. Reload to refresh your session. You signed out in another tab or window. Reload to refresh your session.