# Manual training loop with gradient access
optimizer = keras.optimizers.SGD(learning_rate=0.1)

for epoch in range(10):
    with tf.GradientTape() as tape:
        # Forward pass
        predictions = model(X, training=True)
        loss = keras.losses.binary_crossentropy(y_tensor, predictions)
        loss = tf.reduce_mean(loss)  # Average across batch

        # Compute gradients
        gradients = tape.gradient(loss, model.trainable_variables)

        # Apply gradients
        optimizer.apply_gradients(zip(gradients, model.trainable_variables))

        if epoch % 2 == 0:
            print(f"Epoch {epoch}, Loss: {loss.numpy():.4f}")

        __
        __