optimize dim reduce performance on norm, argmax and argmin (#72083)
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/72083
Pull Request resolved: https://github.com/pytorch/pytorch/pull/64479
Test Plan: Imported from OSS
Reviewed By: VitalyFedyunin
Differential Revision: D33862408
Pulled By: frank-wei
fbshipit-source-id: eb291d59144e2ddc566d8c1491fe09b5b3f53fb0
(cherry picked from commit 11c384049d12ca67edd2b5ef5e6c3a7a7fefb835)
