Transformer反归一化

针对部分大佬写的代码，并未对数据进行反归一化，这导致对于我这种小白画图不太友好，因此我通过一上午的检索找到了一种修改方法，经测试有效！

针对改代码错误：ValueError: cannot select an axis to squeeze out which has size not equal to one

初始代码：

if test_data.scale and self.args.inverse:
    shape = outputs.shape
    # outputs = test_data.inverse_transform(outputs.squeeze(0)).reshape(shape)
    # batch_y = test_data.inverse_transform(batch_y.squeeze(0)).reshape(shape)
修改后代码：

    if test_data.scale and self.args.inverse:
        shape = outputs.shape
        outputs = test_data.inverse_transform(
            outputs.reshape((shape[0] * shape[1],) + shape[2:])).reshape(shape)
        batch_y = test_data.inverse_transform(
            batch_y.reshape((shape[0] * shape[1],) + shape[2:])).reshape(shape)

另一处修改，在上面这个修改下面几行处，请自行查找

修改前：
input = test_data.inverse_transform(input.squeeze(0)).reshape(shape)
修改后：
input = test_data.inverse_transform(
    input.reshape((shape[0] * shape[1],) + shape[2:])).reshape(shape)
如果有效，请点赞收藏！