Add quantization support to reshape and size for the ONNX exporter (#106629)
Fixes https://github.com/microsoft/onnx-converters-private/issues/175
Add quantization support for Reshape-14, Size-9 and Size-11
For Size operators, we don't requantize outputs because we want the original scalar in the graph
Pull Request resolved: https://github.com/pytorch/pytorch/pull/106629
Approved by: https://github.com/BowenBao
