Closed ultmaster closed 4 years ago
Hi. Thanks for this great repo.
I'm interested in why you did convert to BGR tensor intentionally and keep the values between 0 - 255 instead of normalization (div 255, minus mean, div std) as usual.
Related code:
https://github.com/megvii-model/SinglePathOneShot/blob/36eed6cf083497ffa9cfe7b8da25bb0b6ba5a452/src/Supernet/train.py#L37
Just in order to keep the same as that in our own DL framework.
Thanks for response.
Hi. Thanks for this great repo.
I'm interested in why you did convert to BGR tensor intentionally and keep the values between 0 - 255 instead of normalization (div 255, minus mean, div std) as usual.
Related code:
https://github.com/megvii-model/SinglePathOneShot/blob/36eed6cf083497ffa9cfe7b8da25bb0b6ba5a452/src/Supernet/train.py#L37