All computers should have python-virtualenv installed. If you need other version than the one provided with the system please use it to install it manually,

Computers in room 527

  • CUDA: 10.1
  • OpenCL: 1x OpenCL 1.2 CUDA/1x Intel OpenCL 1.2 (build 8)
  • Python 2.7 and Python 3.6
    • h5py (2.7.1)
    • Keras (2.1.3)
    • numpy (1.14.0)
    • Pillow (5.0.0)
    • pycuda (2017.1.1)
    • pyopencl (2018.1)
    • scipy (0.19.1)
    • tensorflow (2.1.0) - WITH GPU SUPPORT
    • Theano (1.0.1)

APL09 and APL10

  • CUDA: 9.1
  • OpenCL: 2x OpenCL 1.1 CUDA
  • Python 2.7 i Python 3.4
    • h5py (2.7.0)
    • Keras (2.0.5)
    • numpy (1.14.0)
    • Pillow (4.1.1)
    • pycuda (2017.1)
    • pyopencl (2017.1.1)
    • scipy (0.19.1)
    • tensorflow (1.5.0) - CPU ONLY
    • Theano (0.9.0)

APL11

  • scl enable devtoolset-7 bash - dla GCC 7
  • CUDA: 9.1 (z libcudnn.so w wersji 7)
  • OpenCL: 2x OpenCL 1.2 CUDA/1x OpenCL 1.1 (Build 31360.31426)
  • Python 2.7 i Python 3.4
    • Keras (2.0.5)
    • numba (0.33.0)
    • h5py (2.7.0)
    • Keras (2.0.5)
    • numpy (1.14.0)
    • Pillow (4.1.1)
    • pycuda (2017.1)
    • pyopencl (2017.1.1)
    • scipy (0.19.1)
    • tensorflow (1.5.0) - WITH GPU SUPPORT
    • Theano (0.9.0)

APL12

  • scl enable devtoolset-3 bash/scl enable devtoolset-4 bash
  • OpenCL: 3x OpenCL 1.2 (Build 8)
  • Python 2.7 - scl enable python27 bash
    • h5py (2.7.0)
    • Keras (2.0.5)
    • numpy (1.14.0)
    • pyopencl (2017.1.1)
    • scipy (0.19.1)
    • tensorflow (1.5.0) - CPU ONLY
    • Theano (0.9.0)
  • Python 3.4 - available directly
    • h5py (2.7.0)
    • Keras (2.0.5)
    • numpy (1.14.0)
    • pyopencl (2017.1.1)
    • scipy (0.19.1)
    • tensorflow (1.5.0) - CPU ONLY
    • Theano (0.9.0)