(tested on Windows 7 SP1 64-bit)
First, install Python 3.5 and TensorFlow following these instructions: How to install TensorFlow on Windows?
Then, from the command prompt (make sure pip is connected to Python 3.5. You can verify it by running pip -V):
pip install -U networkx matplotlib scikit-learn scipy pycorenlp
You need to install SpaCy. To do so, two solutions.
Solution 1: Installing Visual Studio Express 2015 (https://www.visualstudio.com/vs/visual-studio-express, free but takes 12 GB of space on the hard drive), then run:
pip install -U spacy
python -m spacy.en.download
Solution 2: Use Anaconda, in which case there is no need to install Visual Studio Express 2015:
conda config --add channels conda-forge
conda install spacy
python -m spacy.en.download
python -m spacy download en
You also need to install Perl (because the official CoNLL-2003 evaluation script is written in Perl): http://strawberryperl.com
Make sure the perl.exe binary is in your Path system environment variable:
The tensorboard.exe binary should also be in your Path system environment variable, if you plan to use TensorBoard (optional).