Day 17 煉丹爐開始煉丹啦 - 訓練神經網路 介紹整個訓練的流程，但是準確率大概在 75% 左右就到極限了，因此今天加入了卷積神經網路。

定義網路時會寫兩個 module，分別卷積和全連階層，整個運作的流程是 Conv (卷積) → Flatten (平坦) → FC (Fully Connect, 全連接層)，程式碼如下：

class CustomConvNeuralNetwork(nn.Module):
    def __init__(self):
        super().__init__()
        self.flatten = nn.Flatten()
        self.cnn_module = nn.Sequential(
            nn.Conv2d(in_channels=3, out_channels=6, kernel_size=5),
            nn.ReLU(),
            nn.MaxPool2d(2, 2),
            nn.Conv2d(in_channels=6, out_channels=16, kernel_size=5),
            nn.MaxPool2d(2, 2),
            nn.ReLU(),
        )
        self.fc_modeul = nn.Sequential(
            nn.Linear(16 * 53 * 53, 120),
            nn.ReLU(),
            nn.Linear(120, 84),
            nn.ReLU(),
            nn.Linear(84, 2)
        )


    def forward(self, x):
        x = self.cnn_module(x)
        x = self.flatten(x)
        x = self.fc_modeul(x)
        
        return x

cnn_model = CustomConvNeuralNetwork().to(device)
print(cnn_model)

訓練和測試的程式碼與 Day17 相同：

def train_loop(dataloader, model, loss_fn, optimizer):
    size = len(dataloader.dataset)
    model.train()
    for batch, (X, y) in enumerate(dataloader):
        # 將資料讀取到GPU中
        X, y = X.to(device), y.to(device)
        # 運算出結果並計算loss
        pred = model(X)
        loss = loss_fn(pred, y)

        # 反向傳播
        loss.backward()
        optimizer.step()
        optimizer.zero_grad()

        if batch % 100 == 0:
            loss, current = loss.item(), (batch + 1) * len(X)
            print(f"loss: {loss:>7f}  [{current:>5d}/{size:>5d}]")

def test_loop(dataloader, model, loss_fn):
    model.eval()
    size = len(dataloader.dataset)
    num_batches = len(dataloader)
    test_loss, correct = 0, 0
    # 驗證或測試時記得加入 torch.no_grad() 讓神經網路不要更新
    with torch.no_grad():
        for X, y in dataloader:
            X, y = X.to(device), y.to(device)
            pred = model(X)
            test_loss += loss_fn(pred, y).item()
            correct += (pred.argmax(1) == y).type(torch.float).sum().item()

    test_loss /= num_batches
    correct /= size
    print(f"Test Error: \n Accuracy: {(100*correct):>0.1f}%, Avg loss: {test_loss:>8f} \n")

最後我們一樣訓練十個 epochs，看看結果如何：

loss_fn = nn.CrossEntropyLoss()
optimizer = torch.optim.SGD(cnn_model.parameters(), lr=learning_rate)

epochs = 10
for t in range(epochs):
    print(f"Epoch {t+1}\n-------------------------------")
    train_loop(train_dataloader, cnn_model, loss_fn, optimizer)
    test_loop(val_dataloader, cnn_model, loss_fn)
print("Done!")

最後可以看到準確率大約為 80%：

結語

加入了卷積的運算雖然有增加一些準確率，但應該還能做得更好，接下來會介紹卷積的運作和嘗試訓練出更準的模型。