一、tf.losses.mean_squared_error函数函数
tf.losses.mean_squared_error( labels, predictions, weights=1.0, scope=None, loss_collection=tf.GraphKeys.LOSSES, reduction=Reduction.SUM_BY_NONZERO_WEIGHTS )
返回值:spa
加权损失浮动Tensor.若是reduction是NONE,则它的形状与labels相同;不然,它是标量.code