[quant] fix int16 quantization scale in conv weight (#74665)
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/74665
fix int16 quantization scale in conv weight
Test Plan:
python3 test/test_quantization.py TestQuantizeEagerOps.test_int16_reference_module
Imported from OSS
Reviewed By: mrshenli
Differential Revision: D35106497
fbshipit-source-id: 61030786d20d845ef36ea40cdacdd7dcccf12ae9
(cherry picked from commit 187b6bd02c9ccc17416b1620c3b164c84352ef81)