Training and evaluating
In this exercise, we'll bring together al that we've practiced so far, by training and evaluating a neural network on the real-world dataset of handwritten Ethiopian MNIST characters.
ImageClassifier is a predefined neural network model implemented using PyTorch Lightning. It consists of convolutional layers for feature extraction, activation functions for introducing non-linearity, and fully connected layers for classification.
The Ethiopic MNIST dataset was pre-imported for you.
This exercise is part of the course
Scalable AI Models with PyTorch Lightning
Exercise instructions
- Import the
Trainer. - Define
ImageClassifiermodel and trainer. - Train the model.
- Evaluate the model on the validation set.
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
# Import the Trainer
from lightning.pytorch import ____
# Define ImageClassifier model & trainer and set epoch parameter
model = ____()
trainer = ____(max_epochs=5)
# Train the model
trainer.fit(____, train_loader, val_loader)
# Evaluate the model
val_results = trainer.____(____, val_loader)
print("Validation Accuracy:", val_results[0]["val_acc"])