in svhn_based_exp/utils.py [0:0]
def get_output_from_model(inputs, net, repre, model, device): x = get_repre(torch.flatten(net(inputs), 1), repre, device) return x