Pytorch自定义创建BP神经网络

class BPNet(nn.Module):
    def __init__(self, in_dim, n_hidden_1, n_hidden_2,\
                 n_hidden_3, n_hidden_4, n_hidden_5, out_dim):
        super(BPNet, self).__init__()
        self.layer1 = nn.Sequential(nn.Linear(in_dim, n_hidden_1))
        self.layer2 = nn.Sequential(nn.Linear(n_hidden_1, n_hidden_2), nn.BatchNorm1d(n_hidden_2))
        self.layer3 = nn.Sequential(nn.Linear(n_hidden_2, n_hidden_3), nn.BatchNorm1d(n_hidden_3), nn.ReLU(True))
        self.layer4 = nn.Sequential(nn.Linear(n_hidden_3, n_hidden_4),  nn.BatchNorm1d(n_hidden_4), nn.ReLU(True), nn.Dropout(0.1))
        self.layer5 = nn.Sequential(nn.Linear(n_hidden_4, n_hidden_5), nn.BatchNorm1d(n_hidden_5))
        self.layer6 = nn.Sequential(nn.Linear(n_hidden_5, out_dim))

    def forward(self, x):
        x = self.layer1(x)
        x = self.layer2(x)
        x = self.layer3(x)
        x = self.layer4(x)
        x = self.layer5(x)
        x = self.layer6(x)
        return x

net = BPNet(in_dim=5, n_hidden_1=20, n_hidden_2=250, n_hidden_3=500, n_hidden_4=250, n_hidden_5=50, out_dim=2)  # 实例化网络

简洁写法

cfg = {
    '': [20, 200, 500, 200, 50],

}

class BPNet(nn.Module):
    def __init__(self, name):
        super(BPNet, self).__init__()
        self.features = self._make_layers(cfg[name])
        self.classifier = nn.Sequential(
            nn.Linear(cfg[name][-1], 2)
        )

    def forward(self, x):
        out = self.features(x)
        out = out.view(out.size(0), -1)
        out = self.classifier(out)
        return out

    def _make_layers(self, cfg):
        layers = []
        in_dim = 5
        for x in cfg:
            layers += [nn.Linear(in_dim, x),
                       nn.BatchNorm1d(x),
                       nn.ReLU(inplace=True)]
            in_dim = x
        return nn.Sequential(*layers)

net = BPNet('1')