Can you guide a bit as to how I can train this on my dataset which may have different sized images and not 512x512
Can you guide a bit as to how I can train this on my dataset which may have different sized images and not 512x512