Why predict function in encoding?

here we are just encoding images why are we using predict function?

hey @Par1hsharma ,
we picked a sub-model from the actual Resnet model , which returns the features of particular input image.
Hence to get them , we used a predict function.