[quant][fx] Follow up fixes for qconfig validations for fixedqparams ops (#81010)
Summary:
This adds a few things on top of https://github.com/pytorch/pytorch/pull/80184,
1). node.target was assumed to be "tanh", torch.nn.Tanh etc. this PR handles that properly
2). adds FixedQParamsFakeQuantize support
3). extends the comparison function _partial_wrapper_equals to work with FakeQuantize.with_args(observer=...)
Test Plan:
python test/test_quantization.py TestQuantizeFx
python test/test_quantization.py TestQuantizeFxOps
Reviewers:
Subscribers:
Tasks:
Tags:
Differential Revision: [D37735193](https://our.internmc.facebook.com/intern/diff/D37735193)
Pull Request resolved: https://github.com/pytorch/pytorch/pull/81010
Approved by: https://github.com/andrewor14