We represent this as a 2-D tensor of floating-point numbers, with a shape [None, 784] . (Here None means that a dimension can be of any ...
確定! 回上一頁