How to install Tensorflow GPU Cuda Toolkit and Keras in Anaconda Windows by Mahesh Huddar

[@sabarish1610]

while testing GPU it shows like TypeError: Unable to convert function return value to a Python type! The signature was () -> handle

[@atharvalaad8926]

Have you managed to solve this error? I am also encountering the same . please help!

[@SIVAJI_333]

but what about Microsoft visual c++ redistributable installation you didn't do that step.It's the second step in tf-gpu setup

[@TheerthaShetty-pg3zc]

Sir what if we give directly give "Pip install tensorflow" in the jupyter notebook cell and install it, instead of the anconda promt

[@MaheshHuddar]

Yes you can But for tensorflow installation better to create a different environment and then install

[@vishvajeetsutar7532]

sir i am using google colab for tensorflow experiments, but recently gpu of colab is restricted to use as the limited access is ended, my laptop has 4gpu and i have successfully installed the content you in this video , i want to just ask that the file that are executed on colab are able to execute on jupiter or there may be problem for that.

[@MaheshHuddar]

You can execute on local machine without any problem

[@2023Trip]

works 100%

[@MaheshHuddar]

Do like share and subscribe

[@pmatos0071]

Excelent video , thank you so much for this

[@MaheshHuddar]

Welcome please do like share and subscribe

[@Madara_Uchiha320]

It works perfectly and shows GPU, but when I run code and then my CPU usage goes upto 100% but GPU usage is still at 0%

[@princekumar-bd6vo]

will this work on integrated gpu?

[@MaheshHuddar]

Yes it will work But training time will be more as you are using the CPU version of tensorflow

[@MaheshHuddar]

While installing tensorflow 2.10.0 don't install cudatoolkit and cudnn It will install CPU version of tensorflow