Zero mean, unit variance in data preprocessor?

[collinmccarthy]

Hi, a lot of the yolov5 configs have

data_preprocessor=dict(
        type='mmdet.DetDataPreprocessor',
        mean=[0., 0., 0.],
        std=[255., 255., 255.],
        bgr_to_rgb=True
    ),

or

data_preprocessor=dict(
        type='YOLOv5DetDataPreprocessor',
        mean=[0., 0., 0.],
        std=[255., 255., 255.],
        bgr_to_rgb=True)
)

How is this possible - or rather - where is the mean/std stored and used? Is there a mean/std stored in the pre-trained checkpoints that's automatically loaded when mean=0 and std=1, during runner init? Any help here is much appreciated!