[Quant][core][gpu][improvement] Refactored implementation for conv2d_cudnn to use packed parameters (Reland PR#73510) (#74220)
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/74220
This a reland of https://github.com/pytorch/pytorch/pull/73510 --
please look at this PR directly for a summary and test plan.
The only change in this PR is we add the ops to check_forward_backward_compatibility.py
to get around the backwards compatibility issue introduced in the previous PR that
changes the name of the cudnn conv operations.
Test Plan: Imported from OSS
Reviewed By: jerryzh168
Differential Revision: D34886988
Pulled By: dzdang
fbshipit-source-id: 9bdedcd6686956070cd1da8e76b69c0c76f34403
(cherry picked from commit 2a047a669b7fc9cbb4cbdde91dc775594ace4280)
