Function loss

03_auto_gradient.py:13–14

(y_pred, y_val)

# Loss function
def loss(y_pred, y_val):
    return (y_pred - y_val) ** 2

# Before training
print("Prediction (before training)",  4, forward(4).item())