Provide option to save quantized data for DNNLOWP without layout optimization (#19681)
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/19681
For accelerator, we need to lower just the quantized weights data without layout transformation. This diff attempts to provide this option.
Reviewed By: jerryzh168, zrphercule
Differential Revision: D15066568
fbshipit-source-id: 133d749e087c2ad4a899bee5e96f597f70b2443c
