Implementing the validation step
Once we trained a neural network model we need to monitor its performance during training. Using PyTorch Lightning, implement the validation_step() method to calculate and log the validation loss at each epoch.
Deze oefening maakt deel uit van de cursus
Scalable AI Models with PyTorch Lightning
Oefeninstructies
- Compute predictions using the model on input batch.
- Calculate validation loss using
F.cross_entropy(). - Log the validation loss with
self.log()asval_loss.
Praktische interactieve oefening
Probeer deze oefening eens door deze voorbeeldcode in te vullen.
import torch.nn.functional as F
def validation_step(self, batch, batch_idx):
x, y = batch
# Compute predictions using the model
preds = ____(x)
# Calculate validation loss
loss = F.____(preds, y)
# Log the validation loss
self.____('val_loss', loss)