DIGITS requires a build of Caffe. We officially only support recent releases from NVIDIA/caffe (NVcaffe), but any recent build of BVLC/caffe will probably work too.
For best performance, you'll want:
- One or more NVIDIA GPUs (details)
- An NVIDIA driver (details and installation instructions)
- A CUDA toolkit (details and installation instructions)
- cuDNN (download page)
Install some dependencies with Deb packages:
sudo apt-get install --no-install-recommends build-essential cmake git gfortran libatlas-base-dev libboost-all-dev libgflags-dev libgoogle-glog-dev libhdf5-serial-dev libleveldb-dev liblmdb-dev libopencv-dev libprotobuf-dev libsnappy-dev protobuf-compiler python-all-dev python-dev python-h5py python-matplotlib python-numpy python-opencv python-pil python-pip python-protobuf python-scipy python-skimage python-sklearn# example location - can be customized
export CAFFE_HOME=~/caffe
git clone https://github.com/NVIDIA/caffe.git $CAFFE_HOMESetting the CAFFE_HOME environment variable will help DIGITS automatically detect your Caffe installation, but this is optional.
Several PyPI packages need to be installed:
sudo pip install -r $CAFFE_HOME/python/requirements.txtIf you hit some errors about missing imports, then use this command to install the packages in order (see discussion here):
cat $CAFFE_HOME/python/requirements.txt | xargs -n1 sudo pip installWe recommend using CMake to configure Caffe rather than the raw Makefile build for automatic dependency detection:
cd $CAFFE_HOME
mkdir build
cd build
cmake ..
make --jobs=4