Model evaluation

labels = rng.integers(low=0, high=2, size=n_samples, dtype=int)

logits = torch.tensor(rng.uniform(low=-10, high=10, size=n_samples))

outputs = torch.sigmoid(logits)

threshold = 0.5

predictions = (outputs >= threshold).long()

# long() transforms boolean to 0 for False and 1 for True

from sklearn.metrics import confusion_matrix

cm = confusion_matrix(y_true=labels, y_pred=predictions)

# you can use different predefined evaluation metrics,

# such like accuracy_score() (accuracy), balanced_accuracy_score()

# (balanced accuracy), etc.