[PT2][Quant] Move embedding quantization to osss (#103088)
Summary:
This is in preperation to enable embeddign quantization on models with
embeddings.
Test Plan: test_embedding_quantizer
Reviewed By: jerryzh168
Differential Revision: D46267689
Pull Request resolved: https://github.com/pytorch/pytorch/pull/103088
Approved by: https://github.com/andrewor14
