Environment: Windows10 [applicable under linux], anaconda
[please read it patiently! It's really useful, there are many blog posts on the Internet, and it's useless to try a few. Then I explored it myself, and I felt that it might be like this, and then I just tried it. Of course, I have tried many kinds before, but I haven't found the right way]
1. Description of the goal to be achieved:
Target description explanation: I have anaconda installed on my computer, and there are multiple virtual environments on anaconda, of which there are two environments about TensorFlow-GPU: tensorflow2GPU and tf23GPU, these two environments, the first one is about TensorFlow2.0.0GPU version Yes, the second one is about TensorFlow 2.3.0 GPU version. as follows:
However, what is interesting is that these two versions cannot coexist in their respective virtual environments by default. My goal is to allow these two versions to exist independently or to be used normally! ! !
2. Origin of the problem:
The origin of the problem: [If you are impatient, you can go directly to the last step]
At the beginning, tf23GPU did not exist, only tensorflow2GPU. But I want to get familiar with TensorFlo