AO migration: replace torch internal callsites (#94170)
Summary:
Do the following renames:
`torch.quantization` -> `torch.ao.quantization`
`torch.nn.quantized` -> `torch.ao.nn.quantized`
`torch.nn.quantizable` -> `torch.ao.nn.quantizable`
`torch.nn.qat` -> `torch.ao.nn.qat`
`torch.nn.intrinsic` -> `torch.ao.nn.intrinsic`
And then, do
`torch.ao.nn.quantized._reference` -> `torch.ao.nn.quantized.reference` to clean up the aftermath of https://github.com/pytorch/pytorch/pull/84974
Then, manually update `test/test_module_init.py` to fix hanging whitespace due to the replace.
Run this script to do the replacements: https://gist.github.com/vkuzo/7f7afebf8c31b9ba48306223e68a1c82
This is for https://github.com/pytorch/pytorch/issues/81667
Test plan: CI
Pull Request resolved: https://github.com/pytorch/pytorch/pull/94170
Approved by: https://github.com/jerryzh168