[PT2][Quant][BE] Refactor qnnpack_quantizer.py (#102701)
This diff refactors annotate functions so as to couple annotate functions with
corresponding quantization configs that they support. This will help in dynamic
quantization which is only supported for linear layers
Differential Revision: [D46235071](https://our.internmc.facebook.com/intern/diff/D46235071/)
Pull Request resolved: https://github.com/pytorch/pytorch/pull/102701
Approved by: https://github.com/jerryzh168
