An epoch is one complete pass through every example in a machine-learning training dataset. Training usually divides that pass into batches, computes predictions and loss, backpropagates gradients, and updates parameters batch by batch. section 8.2.3
Listing 8.7 Train the Model
def train_model(
model, train_loader, val_loader, criterion, optimizer,
scheduler, device, num_epochs=10
):
model.train() #A
train_losses = [] #B
val_losses = []
for epoch in range(num_epochs): #C
running_loss = 0.0 #D
for inputs, targets in train_loader:
inputs, targets = inputs.to(device), targets.to(device)
optimizer.zero_grad() #E
outputs = model(inputs) #F
loss = criterion(outputs, targets) #G
loss.backward() #H
optimizer.step() #I
running_loss += loss.item()
train_loss = running_loss / len(train_loader)
train_losses.append(train_loss)
model.eval() #J
val_loss = 0.0
with torch.no_grad(): #K
for inputs, targets in val_loader:
inputs, targets = inputs.to(device), targets.to(device)
outputs = model(inputs)
loss = criterion(outputs, targets)
val_loss += loss.item()
val_loss /= len(val_loader)
val_losses.append(val_loss)
print(f'Epoch {epoch+1}/{num_epochs}',
f'Train Loss: {train_loss:.4f}',
f'Val Loss: {val_loss:.4f}') #L
scheduler.step(val_loss) #M
model.train() #NA PyTorch training loop makes epochs explicit with range(num_epochs); each epoch accumulates batch losses, then evaluates validation loss. section 8.2.3
Reading epoch logs
A 50-epoch run can show training loss falling from 3.7013 to 0.2256 while validation loss improves early, then levels off near 0.66. This suggests reducing epochs or using early stopping, while also considering the learning-rate scheduler.













