Minimize Wrapper Method

Functions

minimize_wrapper(optimizer, model, X_train, Y_train, loss, batch_size, epochs, metrics)

A wrapper function to perform multiple epochs of training using the minimize function.

Parameters:

Name	Type	Description	Default
optimizer		The optimizer instance.	required
model	Sequential	The neural network model.	required
X_train	ndarray	Input training data.	required
Y_train	ndarray	True labels for training data.	required
loss	str	The loss function to use.	required
batch_size	int	The size of each mini-batch.	required
epochs	int	The number of epochs to train for.	required
metrics	list	List of metrics to compute during training.	required

Returns:

Name	Type	Description
dict		A history dictionary containing the loss and specified metrics for each epoch.

Example

history = minimize_wrapper(optimizer, model, X_train, Y_train, 'categorical_crossentropy', 32, 10, ['accuracy'])
print(history['loss'])
print(history['accuracy'])

Note

This function manages the training process over multiple epochs, calling the minimize function for each epoch and collecting the training history.

Source code in microkeras/optimizers/minimize_wrapper.py

def minimize_wrapper(optimizer, model, X_train, Y_train, loss, batch_size, epochs, metrics):
    """
    A wrapper function to perform multiple epochs of training using the minimize function.

    Args:
        optimizer: The optimizer instance.
        model (Sequential): The neural network model.
        X_train (numpy.ndarray): Input training data.
        Y_train (numpy.ndarray): True labels for training data.
        loss (str): The loss function to use.
        batch_size (int): The size of each mini-batch.
        epochs (int): The number of epochs to train for.
        metrics (list): List of metrics to compute during training.

    Returns:
        dict: A history dictionary containing the loss and specified metrics for each epoch.

    Example:
        ```python
        history = minimize_wrapper(optimizer, model, X_train, Y_train, 'categorical_crossentropy', 32, 10, ['accuracy'])
        print(history['loss'])
        print(history['accuracy'])
        ```

    Note:
        This function manages the training process over multiple epochs, calling the minimize
        function for each epoch and collecting the training history.
    """
    history = {}

    for metric in metrics:
        if metric in ['accuracy']:
            history[metric] = []
    history['loss'] = []

    for epoch in range(epochs):
        print(f"Epoch {epoch + 1}/{epochs}")
        minimize(optimizer, model, X_train, Y_train, loss, batch_size, metrics)

        # Calculate metrics for the dataset after each epoch
        forward(model, X_train)  # Update the forward pass for entire dataset

        if 'accuracy' in metrics:
            acc = get_accuracy(model, X_train, Y_train)
            history['accuracy'].append(acc)

        if loss == 'categorical_crossentropy':
            loss_value = categorical_crossentropy(Y_train, model.layers[-1].A)
        elif loss == 'mean_squared_error':
            loss_value = mean_squared_error(Y_train, model.layers[-1].A)

        history['loss'].append(loss_value)

    return history