google
Added power function to backend to new backends
This PR implements the power function specifically to symmetric and PyTorch backends with their respective test.
The changes were implemented in the files:
tensornetwork/backends/pytorch/*
tensornetwork/backends/symmetric/*
This is in regards to the issue #840.
@mganahl @alewis Hello! here is the new PR where you may find the new changes and the tensordot issue. I await your feedback and thank you for the oportunity
Hello @alewis @mganahl! Running the test from pylint seems to cause and issue which I cannot resolve, it states that Python does not exist in module tensorflow, I assume the issue is init.py thought updating it with a small change such as adding a . does not seem to solve the issue, also my pylint returns a 10/10 both locally and here in GitHub's CI, the difference being that I do not get this issue locally.
In my search I found that this issue is not some enclosed issue, I found some people raising this issue as well in other places here are some examples:
- https://github.com/Unity-Technologies/ml-agents/issues/1411
- https://stackoverflow.com/questions/40334643/no-name-in-module-error-from-pylint
- https://pycodequ.al/docs/pylint-messages/e0611-no-name-in-module.html
I will continue my search for a solution, my main problem being that we can't resolve the tensordot issue without resolving this first, I would love some help or feedback. Thank you so much in advance for all the help and support you have given me!