6. Step 4: Evaluating the Model

Evaluate the trained model on the test set to measure its performance.

from sklearn.metrics import classification_report # Predict on the evaluation setpredictions = trainer.predict(eval_dataset) # Convert predictions to labelspredicted_labels = predictions.predictions.argmax(-1) # Print classification reportprint(classification_report(eval_dataset['label'], predicted_labels))

Code breakdown:

1. Import and Setup

from sklearn.metrics import classification_report

This imports scikit-learn's classification report tool for evaluating model performance.

2. Making Predictions

predictions = trainer.predict(eval_dataset)predicted_labels = predictions.predictions.argmax(-1)

This code:

• Uses the trained model to make predictions on the evaluation dataset

• Converts the raw predictions into class labels using argmax (choosing the highest probability class)

3. Evaluation

print(classification_report(eval_dataset['label'], predicted_labels))

This generates a report comparing the true labels with predicted labels, showing metrics like precision, recall, and F1-score for each sentiment class. This evaluation step is crucial for understanding how well the model performs on unseen data before deploying it for real customer feedback analysis.