| import numpy as np | |
| from sklearn.metrics import classification_report | |
| def evaluate_model(model, val_generator): | |
| val_generator.reset() | |
| preds = model.predict(val_generator) | |
| y_pred = np.argmax(preds, axis=1) | |
| y_true = val_generator.classes | |
| labels = list(val_generator.class_indices.keys()) | |
| report = classification_report(y_true, y_pred, target_names=labels, output_dict=True) | |
| print(classification_report(y_true, y_pred, target_names=labels)) | |
| return report | |