Feature extraction in torchvision.models.vit_b_16

Hi @ptrblck and @diegoaichele, can you please help me on this: Using nn.Sequential to use models.video.mvit_v2_s for feature extraction