[fx2trt] Add dequantize support (#63448)
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/63448
Only available after TensorRT 8.0
Test Plan: buck run mode/opt caffe2/torch/fb/fx2trt:test_dequantize
Reviewed By: 842974287
Differential Revision: D30296863
fbshipit-source-id: 44b9630ef0d210e7f20e650dc81c519f7e41f5f3
