[Quant] Add qint32 quantization support (#72472)
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/72472
Add dtype=int32 support for observer
Test Plan:
python3 test/test_quantization.py TestObserver.test_per_tensor_observers
Imported from OSS
Reviewed By: jerryzh168
Differential Revision: D34056640
fbshipit-source-id: 4fa15a7274cfbb6a7dd4e698e3989cc0c0626e7b
(cherry picked from commit bf4351de45812a9143b179b7049cc1e93611daf6)