Utilities
The utils module provides various utility functions and classes.
Training Engine
- fit.utils.engine.train_epoch(model, dataloader, loss_fn, optimizer, device=None)[source]
Train for a single epoch.
- Parameters:
model – Model to train
dataloader – DataLoader for training data
loss_fn – Loss function
optimizer – Optimizer
device – Device to use (not used in this version, but kept for PyTorch compatibility)
- Returns:
Dict with epoch metrics (loss, accuracy)
- fit.utils.engine.evaluate(model, dataloader, loss_fn, device=None)[source]
Evaluate model on a dataset.
- Parameters:
model – Model to evaluate
dataloader – DataLoader for evaluation data
loss_fn – Loss function
device – Device to use (not used in this version, but kept for PyTorch compatibility)
- Returns:
Dict with evaluation metrics (loss, accuracy)
- fit.utils.engine.train(model, train_loader, val_loader, loss_fn, optimizer, epochs=10, device=None, scheduler=None, early_stopping=None, tracker=None)[source]
Complete training loop.
- Parameters:
model – Model to train
train_loader – DataLoader for training data
val_loader – DataLoader for validation data (optional)
loss_fn – Loss function
optimizer – Optimizer
epochs – Number of epochs to train
device – Device to use (not used in this version, but kept for PyTorch compatibility)
scheduler – Learning rate scheduler (optional)
early_stopping – Early stopping settings (optional)
tracker – TrainingTracker for logging (optional)
- Returns:
TrainingTracker with training history