The following five steps which I will explain below worked for me.
1 2 3 4 5 |
conda install python=3.5 conda install mingw libpython pip install git+https://github.com/Theano/Theano.git pip install --ignore-installed --upgrade https://storage.googleapis.com/tensorflow/windows/cpu/tensorflow-1.1.0-cp35-cp35m-win_amd64.whl conda install --upgrade keras |
As of today Tensorflow is not compatible with Python 3.6 therefore you need to make sure you are using Python 3.5.
Step 1: install Python 3.5 according to conda.io
1 |
conda install python=3.5 |
Step 2: install Theano’s dependencies according to https://medium.com/@xenocide15
1 |
conda install mingw libpython |
Step 3: install Theano according to https://medium.com/@xenocide15
1 |
pip install git+https://github.com/Theano/Theano.git |
Step 4: install Tensorflow according to tensorflow.org
1 |
pip install --ignore-installed --upgrade https://storage.googleapis.com/tensorflow/windows/cpu/tensorflow-1.1.0-cp35-cp35m-win_amd64.whl |
Step 5: install Keras
1 |
conda install --upgrade keras |