Minimum requirements:

• Python 2.7+ (v2.7.9 is recommended)
• NumPy

Requirements to enable CUDA:

• CUDA 6.5+
• PyCUDA
• scikits.cuda (pip install 'scikits.cuda>=0.5.0b1,!=0.042')
• Mako (depending through PyCUDA)

Recommended:

• CuDNN v2
• scikit-learn (to run some examples)
• OpenCV 2.4 (to run some examples)

Install Chainer via PyPI:

pip install chainer

You can also install Chainer from the source code:

python setup.py install

If you want to enable CUDA, first you have to install CUDA and set the environment variable CUDA_DIR to the installed path. You also have to set binary and library paths by appropriate environment variables like PATH and LD_LIBRARY_PATH. Then install CUDA-related packages by:

pip install chainer-cuda-deps

or, from the source:

python cuda_deps/setup.py install

MIT License (see LICENSE file).