In the anaconda Promat interface, import torch is passed, and the error is reported in the jupyter notebook (only provide ideas and understanding!)

This article is only for personal understanding, I hope for advice, it will be beneficial and better!

In order to complete a project one day, I started to install tensorflow. I found countless blogs on the Internet, thinking about downloading it quickly.

but! but! two thousand hours latter~

After trying to download tensorflow from countless bloggers, I typed import tensorflow in Jupyter Notebook and got an error!Not only can't import the tensorflow package, but import torch will report an error!I think it should be the result of my own random operations...

Two thousand hours latter~

Some ideas were born, hoping to solve some of the problems of our friends!

—————————————————————————————————————————————

Through n operations, I understand it like this:

1. Create a virtual environment (name as needed)

2. Jupyter Notebook should be installed in the virtual environment

The default base (root), click install to download jupyter, Launch indicates that it has been downloaded!

3. Pytorch can be installed in the base environment.

—————————————————————————————————————————————

My problem is finally solved: modify the kernel (kernel)

Click on Jupyter Notebook(anaconda3) to open

Click New, you can see that I have randomly created several virtual environments of TF3, py36, pytorch, tensorflow, and finally I used root.

It seems that it is because I loaded the kernel in the base, my impression is this~

How to say, it is better to read and try more about configuring the environment. I believe that it will always succeed. Everyone is in a different situation. Maybe the first blog you found was successful, or maybe the 99th blog was successful.Harvest inspiration, but there is bound to be a harvest!

2022/8/3