Caffe

Caffe is a deep learning framework made with expression, speed, and modularity in mind. It is developed by the Berkeley Vision and Learning Center (BVLC) and community contributors.

Check out the project site for all the details like

• DIY Deep Learning for Vision with Caffe
• Tutorial Documentation
• BVLC reference models and the community model zoo
• Installation instructions

and step-by-step examples.

Note: This repo is ported from [email protected]:MSRDL/caffe.git for research development. The master branch follows BVLC/Caffe master, and the WinCaffe branch will merge the latest changes from master and keep it compilable in Windows.

Prerequisite

1. Visual Studio 2015
2. Python 2.7 - Anaconda is recommended: https://www.continuum.io/downloads
3. Cuda 8.0

Windows Setup

1. Clone the repository:

   git clone https://github.com/MSRCCS/Caffe.git
   

2. Download 3rd party dependencies - under the caffe root folder, run:

   python .\scripts\download_prebuilt_dependencies.py
   

3. Download cuDNN v5.0 from nVidia website. Please select v5 of CuDnn, which is the version that verifies to build with this WinCaffe package. Then under the caffe root folder, run:

   python .\scripts\install_cudnn.py $downloadedZipFile
   

   where $downloadedZipFile is the path to your downloaded cuDNN file.

4. Set system environment variable PYTHON_ROOT = path\to\your\python\root, which is needed by Visual Studio to find python libraries.

Now, you should be able to build caffe.sln in Visual Studio 2015.

Common issues when pulling new commits from BVLC's branch

• If linking fails: it's likely that there are new cpp files that need to be added to the caffelib project.

License and Citation

Caffe is released under the BSD 2-Clause license. The BVLC reference models are released for unrestricted use.

Please cite Caffe in your publications if it helps your research:

@article{jia2014caffe,
  Author = {Jia, Yangqing and Shelhamer, Evan and Donahue, Jeff and Karayev, Sergey and Long, Jonathan and Girshick, Ross and Guadarrama, Sergio and Darrell, Trevor},
  Journal = {arXiv preprint arXiv:1408.5093},
  Title = {Caffe: Convolutional Architecture for Fast Feature Embedding},
  Year = {2014}
}