Reputation: 303
I recently found an article that indicates that the conventional methods for downloading python machine learning modules such as tensorflow and keras are not optimized for computers with a cpu. How can I configure tensorflow and keras to make it most compatible with my processor on MacOSX in python 2.7?
If it helps, I use pycharm to download most of my libraries and for my coding interface.
Upvotes: 1
Views: 848
Reputation: 130
For any environment if you want to install tensorflow, you can simply run this command :
pip install tensorflow (for CPU, python2.7)
pip3 install tensorflow (for CPU, python3)
You need to mention externally if you want to install tensorflow with GPU like this:-
pip install --upgrade tensorflow-gpu
but for GPU you will need CUDA (NVDIA graphics) to run.
and very same way, you can install keras Where you dont have to pass keras-gpu externally while using command:-
pip install keras
Upvotes: 1
Reputation: 50
I think what you read meant that tensorflow programs work much faster if your computer has a GPU. You need a Nvidia GPU in your computer to install tensorflow with GPU support on your Mac and as far as I know, after version 1.2 tensorflow no longer provides GPU support for MacOS
Upvotes: 0