[EDIT: I’ve updated the site to include information about setting up Caffe on Ubuntu 14.04 for laptop (no GPU), on Ubuntu 14.04 that has a GPU, and on Ubuntu 14.04 Server]
Below is the installation procedure I used to set up my NVIDIA Titan Z and Caffe. You can request a GPU by filling out an Academic Hardware Grant Request Form.
- I started with a fresh install of Ubuntu 14.04 which installed Third-party software (install with internet connection). (on my laptop I did not start with a fresh system)
- Install Titan Z in the tower. It requires a lot of power, so needs the power from the motherboard and 2 8-pin PEG connectors. (clearly not needed for non-GPU system)
- Install the python suite (a standard list): sudo apt-get install python-numpy python-scipy python-dev python-pip python-nose
- Install the pre-reqs on Caffe’s site:
- Download CUDA 6.5 and install (including adding necessary paths to your .bashrc file) – this is not needed if you are not using a GPU
- Install OpenBLAS: sudo apt-get install libopenblas-dev
- Install OpenCV 2.4.9
- fossies took 2.4.9 off their server, but you can use this link: http://sourceforge.net/projects/opencvlibrary/files/opencv-unix/2.4.9/opencv-2.4.9.zip/download
- If you are installing opencv to use with a GPU, during step 5 (before ‘make’), you will run into issues. Use this fix to resolve the issue. Specifically, in /path/to/opencv-2.4.9/modules/gpu/src/nvidia/core, remove the ‘{static}’ characters from NCVPixelOperations.hpp that are identified in the code on the fix site. (Alternatively, you can use the code here, which has already removed those characters. Note: I tried this and it seemed to fail.)
- On Ubuntu 14.04 Server, I had to fix an issue due to an error during the ‘make’ step (Unsupported gpu architecture ‘compute_11’). I used the following fix: http://stackoverflow.com/questions/28010399/build-opencv-with-cuda-support
- After the above steps, it was impossible to sign into Ubuntu Server until the video was set to the Titan Z. Once that was done, it was possible to continue.
- Unzip Boost 1.55 in /usr/local/ (or unzip it elsewhere and move it there)
- More pre-reqs: sudo apt-get install libprotobuf-dev libleveldb-dev libsnappy-dev libopencv-dev libboost-all-dev libhdf5-serial-dev
- More pre-reqs: sudo apt-get install libgflags-dev libgoogle-glog-dev liblmdb-dev protobuf-compiler git
- Download Caffe
- sudo pip install -r /path/to/caffe/python/requirements.txt
- Compile Caffe
- On Ubuntu Server I received an error while loading shared libraries. This command solved the problem: sudo ldconfig /usr/local/cuda/lib64 (for 64-bit Ubuntu)
- Add python folder to your .bashrc file (from Caffe’s site)
- Other necessary packages: sudo pip install pandas sklearn scikit-image
Advertisements