So in order to get started learning Machine Learning and the sort, I need to learn a framework called TensorFlow.
Since my computer setup is such that it runs a cuda enabled graphics card the GTX 1060 and knowing the fact that I can do python, my life is made simple thanks to a video tutorial I found at YouTube worth sharing.
The steps are as follows:
Install Visual Studio 2019. I really like doing things using the latest programs. It’s up to you really if you have the previous versions already or if you are you are using other environments. For me personally, I really like using visual studio.
Download cuDNN. It is a CUDA Deep Neural Network library. You do have to sign up to be able to download the library by the way.
Extract the cuDNN zip folder to your desktop. Open a new windows explorer and navigate to
C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\V[cuda_version]\ .
Now copy the contents of the downloaded and extracted cuDNN folder into the appropriate folders (files from cuDNN bin go into the new windows explorer bin folder etc.).
Navigate to your system environment variables and edit the path. Add the following two directories into your path:
- C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v[cuda_version]\bin
- C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v [cuda_version]\libnvvp
For the environment variables, it is up to you but the above is the needed definitely. For example, my personal environment variables are as follows:
Download and Install Anaconda
Set up a virtual environment using python [python_version]
- conda create -n [name] python=[python_version].
Activate the virtual environment
- activate [name]
Install the necessary packages.
- pip install --ignore-installed --upgrade tensorflow-gpu
- pip install keras
- pip install pandas (optional)
- pip install numpy (optional)
- pip install matplotlib (optional)