How to Install TensorFlow GPU on Windows 11 | GPU Setup on Windows | TensorFlow GPU Setup

hi everyone let's go ahead and see how you can install tensorflow GPU on a Windows computer you need to come here at tensorflow.org install pip Windows otherwise you can just simply search here install tensorflow on Windows then you need to open here the first link you just need to open the first link thereafter you need to hide this one and then just scroll it little down and thereafter you need to select here Windows native so in this Windows native it says that currently tensorflow 2.10 or after that I mean 2.10 or before that is only supported with Windows native so with tensorflow 2.11 it is not supported with Windows as of now so you cannot install so basically we have to make sure that we install our tensorflow less than 2.11 apart from that you also need to make sure that you install python 10 or less than python 3.10 so whenever you go with python 3.10 up more than the python 3.10 then uh it would again start creating a problem so whenever if you have installed a latest version of the Anaconda then probably you might have a higher version of the Python so you have to create a new environment for your Python 3.10 and then you have to install your tensorflow and then Cuda toolkit for GPU use so here I have shown how you can install Anaconda the link is given in this video description from there you can watch this video just see how you can install that here is the process which I'm gonna show you how you can install anaconda and how you can install Python 3.10 and then how you can install Cuda toolkit and the tensorflow in fact the Anaconda installation is given in the video and a python and Cuda apart from that the tensorflow installation is given in this in this node here so first of all you need to create here a new python environment so I'm gonna give a name to my python environment name as pye310 thereafter you need to open here Anaconda prompt you just need to search here Anaconda prompt just open this Anaconda prone thereafter it will open your anaconda prompt here do not worry you need to create here first of all a new environment so I'm just gonna paste that here conda create minus n p y 310 that's the python 3.10 and python equal to 3.10 so it's going to install Python 3.10 and it's going to create a new environment it will also install all the necessary libraries needed for your python 3.10 all right so here it is installing it will take a while to install all these necessary packages here we have to wait for some time so seems like it has installed all these necessary packages now we can activate our kunda package here so you just need to copy conda activate py at 310 so I just write here conda activate python 310 so python 310 is now activated once python 310 is activated thereafter you need to install Cuda toolkit 11.2 and qdnn 8.1.0 so I'm just going to copy this whole thing from here and then I'm gonna paste it here all right so this thing is going to install Cuda toolkit and qdnn in your computer under py310 environment so do remember whenever you need to use your GPU and tensorflow you have to make sure that you activate your you you activate your environment py310 so that you can activate with Konda activate py310 it will take a while to install all these necessary libraries thereafter we will follow here our other steps like python install tensorflow so this will also take a while to install so let's wait for some time and let it finish first it has asked me whether you want to proceed or not yes definitely I want to install all the Nvidia necessary packages in my PC so it has successfully installed this so in video packages are installed now we are ready to install here tensorflow you just need to copy this and then paste it here depending on your internet speed and computer speed it might take some time to install Nvidia packages and tensorflow because it downloads these softwares from the internet so obviously depending on internet speed it may take some time at your computer so installation is getting done here it has downloaded all the necessary packages which is really quite fast because my internet is speed is really super fast so it to it took like 10 to 20 seconds for me to only install all these packages here once all these packages are installed then we are going to run our Python and then we will be testing whether our GPU is working or not all right so in background you can see that it has successfully installed all the necessary packages now we are ready to run this code to test if our if our GPU is working with tensorflow or not so you need to just write here python and thereafter you just need to paste that here import tensorflow as TF so this is first time so it will take some time to import tensorflow all right we are still waiting here now you can clearly see here as soon as I run here this command tf.config.list physical devices it is saying that there is one GPU available to my computer otherwise you can simply write here TF DOT test dot is GPU available you just paste it here and then if you run this it says that true that's me Nvidia GE Force RTX 2060 GPU is available to my PC so congratulations you have successfully installed here tensorflow and you have also installed Nvidia to toolkit and you have enabled your tensorflow to use your GPU now you can use GPU for a better performance on your deep learning algorithms all right this is all about in this lesson thanks a lot for watching this if you have liked this video please do subscribe this Channel and share it with your friends thank you

Channel: KGP Talkie

Views: 33,384

Keywords: kgp talkie, kgp talkie videos, machine learning tutorials, kgp talkie data science, kgp talkie ml, kgp talkie for machine learning, tensorflow gpu install windows 11, install tensorflow gpu, tensorflow gpu, install tensorflow gpu windows 11, install tensorflow in anaconda, how to install tensorflow in anaconda, how to use gpu in jupyter notebook, how to install tensorflow in windows 11, how to install tensorflow, how to install cuda on windows 11, install tensorflow, gpu setup

Id: QUjtDIalh0k

Length: 7min 31sec (451 seconds)

Published: Sat Sep 16 2023