Installing TensorFlow 2.1 with CUDA and cuDNN on Windows 10

M. Haroon Shakeel
1 min readFeb 5, 2020

--

Hello World!
Today we will learn how to setup TensorFlow 2.1 with CUDA and cuDNN on Windows 10 in three simple steps.

Step 1: Download and install Anaconda with Python 3.x.

Step 2: Download CUDA 10.1 and cuDNN 7.6.5 for CUDA 10.1. Make sure the versions are compatible with TensorFlow by visiting the TensorFlow official page. You will also need Microsoft Visual C++ Redistributables, which can be downloaded here. Install CUDA and VCredist with default settings. Extract cuDNN zip file. Copy all the files inside cuda folder and replace them in

 C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v10.1

Step 3: Open Anaconda Prompt as Administrator from the start menu. Run command:

pip install tensorflow-gpu==2.1.0 

More details on setting up TF GPU and OpenCV could be found in this video.

--

--

M. Haroon Shakeel
M. Haroon Shakeel

Written by M. Haroon Shakeel

I am PhD scholar, specializing in deep learning for natural language processing and computer vision, specifically interested in learning with minimal data.

Responses (2)