Enabled per channel quantized static linear/conv (#37622)
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/37622
Enable channelwise quantized test on qlinear and qconv.
Dynmaic linear to follow.
Test Plan:
pytest test/quantization/test_quantized.py
pytest test/quantization/test_quantized_module.py
Imported from OSS
Differential Revision: D21339046
fbshipit-source-id: 32377680d3a6424ca1e98d3707b82839eeb349a7
