The model for inference can be first yielded through the following operation:
To obtain the inference time consumption, please download the PyTorch C++ api and extract it in a libtorch directory, and then run the following operations.
The returned accuracy should be about 94.32%. On a single V100, the average running time is 0.682 milliseconds for each image. This was run on PyTorch torch-1.0.1.post2-cp35-cp35m-linux_x86_64.whl.