1

我已经使用“Tensorflow”建立了一个 MLP 神经网络,如下所示:

model_mlp=Sequential()
model_mlp.add(Dense(units=35, input_dim=train_X.shape[1], kernel_initializer='normal', activation='relu'))
model_mlp.add(Dense(units=86, kernel_initializer='normal', activation='relu'))
model_mlp.add(Dense(units=86, kernel_initializer='normal', activation='relu'))
model_mlp.add(Dense(units=10, kernel_initializer='normal', activation='relu'))
model_mlp.add(Dense(units=1))

我想使用 pytorch 转换上面的 MLP 代码。怎么做?我尝试按如下方式进行:

    class MLP(nn.Module):
    def __init__(self):
        super(MLP, self).__init__()
        self.fc1 = nn.Linear(train_X.shape[1],35)
        self.fc2 = nn.Linear(35, 86)
        self.fc3 = nn.Linear(86, 86)
        self.fc4 = nn.Linear(86, 10)
        self.fc5 = nn.Linear(10, 1)
    def forward(self, x):
        x = F.relu(self.fc1(x))
        x = F.relu(self.fc2(x))
        x = F.relu(self.fc3(x))
        x = F.relu(self.fc4(x))
        x = self.fc5(x)
        return x
    def predict(self, x_test):
        x_test = torch.from_numpy(x_test).float()
        x_test = self.forward(x_test)
        return x_test.view(-1).data.numpy()
model = MLP()

我使用相同的数据集,但两个代码给出了两个不同的答案。使用 Tensorflow 编写的代码总是比使用 Pytorch 编写的代码产生更好的结果。我想知道我在 pytorch 中的代码是否不正确。如果我在 PyTorch 中编写的代码是正确的,我想知道如何解释这些差异。我期待着任何答复。

4

1 回答 1

1

欢迎来到 pytorch!

我想问题出在网络的初始化上。我就是这样做的:

def init_weights(m):
    if type(m) == nn.Linear:
        torch.nn.init.xavier_normal(m.weight)  # initialize with xaver normal (called gorot in tensorflow)
        m.bias.data.fill_(0.01) # initialize bias with a constant

class MLP(nn.Module):
    def __init__(self, input_dim):
        super(MLP, self).__init__()
        self.mlp = nn.Sequential(nn.Linear(input_dim ,35), nn.ReLU(),
                                 nn.Linear(35, 86), nn.ReLU(),
                                 nn.Linear(86, 86), nn.ReLU(), 
                                 nn.Linear(86, 10), nn.ReLU(),
                                 nn.Linear(10, 1), nn.ReLU())

    def forward(self, x):
        y =self.mlp(x)
        return y

model = MLP(input_dim)
model.apply(init_weights)

optimizer = Adam(model.parameters())
loss_func = BCEWithLogistLoss()

# training loop

for data, label in dataloader:
    optimizer.zero_grad()
    
    pred = model(data)
    loss = loss_func(pred, lable)
    loss.backward()
    optimizer.step()
    

请注意,在 pytorch 中,我们不调用model.forward(x),而是调用model(x)。那是因为应用了后向传递中使用的nn.Module钩子。.__call__()

您可以在此处查看权重初始化的文档:https ://pytorch.org/docs/stable/nn.init.html

于 2020-07-28T07:28:13.613 回答