Reshape : target shape must have the same size as current shape when recording with autograd. ... torch.reshape(a, (2, 2)) tensor([ [0., 1.] ...
確定! 回上一頁