Pytorch autoencoder print accuracy positional argument error

Question

I got this autoencoder

class AE(nn.Module):
    def __init__(self, **kwargs):
        super(AE,self).__init__()
        self.encoder_hidden_layer = nn.Linear(