Perfecting the forward method
After setting up layers in the __init__ method, the forward method dictates how data flows through them. In PyTorch Lightning, this separation keeps your code clean and easy to maintain. You've already seen how to structure the constructor-now it's time to focus on the forward pass, ensuring your classification logic is clear and optimized for training. Here, the layers in __init__ are already defined for you, so you can concentrate purely on the forward flow.
The lightning.pytorch and torch.nn have already been imported as pl and nn.
この演習はコースの一部です
Scalable AI Models with PyTorch Lightning
演習の手順
- Implement the
forwardmethod insideClassifierModel. - Apply a ReLU activation after the hidden layer.
実践的なインタラクティブ演習
このサンプルコードを完成させて、この演習に挑戦してみましょう。
class ClassifierModel(pl.LightningModule):
def __init__(self, input_dim, hidden_dim, output_dim):
super().__init__()
self.hidden = nn.Linear(input_dim, hidden_dim)
self.output = nn.Linear(hidden_dim, output_dim)
# Define forward method
def ____(self, ____):
# Complete the forward pass
x = self.hidden(x)
x = ____(x)
x = self.output(x)
return x