[quant][graphmode] FP16 quant support - Insert cast operators (#40709)
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/40709
Cast to kHalf and back to kFloat before the linear operator to mimic FP16 quant support
Test Plan:
python test/test_quantization.py test_convert_dynamic_fp16
Imported from OSS
Differential Revision: D22335977
fbshipit-source-id: f964128ec733469672a1ed4cb0d757d0a6c22c3a
