数据类型dtype CPU张量GPU张量32位浮点torch.float32 或torch.float torch. ... 3, 3, 3]]) >>> x.expand(-1, 4) # -1 means not changing the size of that dimension ...
確定! 回上一頁