Skip to content

Latest commit

 

History

History
60 lines (42 loc) · 1.56 KB

installation.md

File metadata and controls

60 lines (42 loc) · 1.56 KB

Installation

CapsLayer is build on the top of TensorFlow with its python API, so you need python to be installed first.

Step 1: Install dependencies

  • TensorFlow

Follow the Tensorflow installation instructions(for China)

  • NumPy and SciPy

system-level installation:

$ sudo pip install numpy
$ sudo pip install scipy

user-level installation:

$ pip install --user numpy
$ pip install --user scipy

Step 2: Install CapsLayer

2.1 Clone or download this repository:

$ git clone https://github.com/naturomics/CapsLayer.git (In this way you should have git installed)

2.2 Install CapsLayer

$ cd path/to/CapsLayer
# Option 1, temporarily add the path of CapsLayer to your PYTHONPATH environment variable so you don't
# need to install capslayer as system-level library
$ export PYTHONPATH=path/to/CapsLayer:${PYTHONPATH}

# Option 2, install capslayer to a user-local path
$ python setup.py install --user

# Option 3, install capslayer to a system-level path
$ sudo python setup.py install

2.3 Make sure it's installed correctly, or you will fail to import capslayer with such 'Module Not Found' error #27

$ echo 'import capslayer as cl'>/tmp/test.py
$ python /tmp/test.py

if no error comes up, congratulations! And Play it fun!

Step 3: Try your first CapsLayer program

Follow the instructions in the Model Section to run your first CapsLayer model.

TODO

  • Test on Windows