Target and input must have the same number of elements.

出现这个报错的原因是，交叉熵损失函数torch.nn.BCELoss()=F.binary_cross_entropy()，要求输入input和目标target 必须形状一致，并且都是浮点数 ：

Input: (N, *)(N,∗) where *∗ means, any number of additional dimensions

Target: (N, *)(N,∗), same shape as the input

Output: scalar. If reduction is 'none', then (N, *)(N,∗), same shape as input.

上方的报错可见，目标是一维的，含元素个数655360。而输入是一维的，含元素个数327680。所以只要在网络中把二者元素个数改成一致即可。