Installing Tensorflow for CPU and GPU on a Windows machine (nvidia gpus).
I've recently been working quite a bit with Tensorflow . Below are my notes on how to install Tensorflow on a Windows machine for both CPU and GPU. The process was extremely painful for me my first time so these notes should be pretty useful for anyone starting out for the first time. You can find more details in my github repo: https://github.com/luweizhang/deeplearning/tree/master/tensorflow-setup Step 1: Install the correct NVidia CUDA and CUDNN drivers: CUDA Drivers - http://docs.nvidia.com/cuda/cuda-installation-guide-microsoft-windows CUDNN - CUDA for Deep Neural Networks - https://developer.nvidia.com/cudnn As of 1/4/2018, I recommend pairing tensorflow 1.4, with CUDA 8.0 and CUDNN version 6.X Step 2: Set up virtual environment for CPU Tensorflow and GPU Tensorflow and install the necessary packages. You will need to have anaconda installed to run conda commands. conda create --name tensorflow python=3.5 activate tensorflow==1.4.0 pi...