Recognize that the returned tensor shares the identical details. For those who create a transform during the "view" that you are switching the initial tensor's info: I feel a superb example (widespread circumstance early on in pytorch prior to the flatten layer was official extra was this popular code): Definitely https://insurance-claims-assistan99788.educationalimpactblog.com/60913048/the-5-second-trick-for-find-out-more