Python load_mnist_dataset Beispiele

Beispiel #1

Datei: run_conv_net.py Projekt: zc8340311/Deep-Learning-TensorFlow

assert FLAGS.dataset in ['mnist', 'cifar10', 'custom']
assert FLAGS.opt in ['gradient_descent', 'ada_grad', 'momentum', 'adam']
assert FLAGS.loss_func in ['mean_squared', 'softmax_cross_entropy']

if __name__ == '__main__':

    utilities.random_seed_np_tf(FLAGS.seed)

    if FLAGS.dataset == 'mnist':

        # ################# #
        #   MNIST Dataset   #
        # ################# #

        trX, trY, vlX, vlY, teX, teY = datasets.load_mnist_dataset(
            mode='supervised')

    elif FLAGS.dataset == 'cifar10':

        # ################### #
        #   Cifar10 Dataset   #
        # ################### #

        trX, trY, teX, teY = datasets.load_cifar10_dataset(FLAGS.cifar_dir,
                                                           mode='supervised')
        vlX = teX[:5000]  # Validation set is the first half of the test set
        vlY = teY[:5000]

    elif FLAGS.dataset == 'custom':

        # ################## #

Beispiel #2

Datei: run_deep_autoencoder.py Projekt: yyan162/DeepLearning_Summary

# Parameters validation
assert FLAGS.dataset in ['mnist', 'cifar10', 'custom']
assert len(rbm_layers) > 0

if __name__ == '__main__':

    utilities.random_seed_np_tf(FLAGS.seed)

    if FLAGS.dataset == 'mnist':

        # ################# #
        #   MNIST Dataset   #
        # ################# #

        trX, vlX, teX = datasets.load_mnist_dataset(mode='unsupervised')
        trRef = trX
        vlRef = vlX
        teRef = teX

    elif FLAGS.dataset == 'cifar10':

        # ################### #
        #   Cifar10 Dataset   #
        # ################### #

        trX, teX = datasets.load_cifar10_dataset(FLAGS.cifar_dir,
                                                 mode='unsupervised')
        # Validation set is the first half of the test set
        vlX = teX[:5000]
        trRef = trX

Beispiel #3

Datei: run_rbm.py Projekt: CamZHU/Deep-Learning-TensorFlow

assert FLAGS.dataset in ['mnist', 'cifar10', 'custom']
assert FLAGS.cifar_dir != '' if FLAGS.dataset == 'cifar10' else True
assert FLAGS.visible_unit_type in ['bin', 'gauss']

if __name__ == '__main__':

    utilities.random_seed_np_tf(FLAGS.seed)

    if FLAGS.dataset == 'mnist':

        # ################# #
        #   MNIST Dataset   #
        # ################# #

        trX, vlX, teX = datasets.load_mnist_dataset(mode='unsupervised')
        width, height = 28, 28

    elif FLAGS.dataset == 'cifar10':

        # ################### #
        #   Cifar10 Dataset   #
        # ################### #

        trX, teX = datasets.load_cifar10_dataset(FLAGS.cifar_dir, mode='unsupervised')
        vlX = teX[:5000]  # Validation set is the first half of the test set
        width, height = 32, 32

    elif FLAGS.dataset == 'custom':

        # ################## #

Beispiel #4

def generate_encodings():
    """Generates encodings for MNIST dataset."""
    train_images, train_labels, validation_images, validation_labels, test_images, test_labels = datasets.load_mnist_dataset(
        mode='supervised')

    # Convert one-hot to integer
    train_labels = [np.argmax(label) for label in train_labels]
    test_labels = [np.argmax(label) for label in test_labels]

    # Initialize the autoencoder
    autoencoder = Autoencoder([2048, 1024, 256, 128],
                              pretrain=False,
                              pretrain_epochs=0,
                              finetune_epochs=120,
                              finetune_batch_size=64)
    autoencoder.train(train_images)
    autoencoder.generate_encodings(
        train_images,
        train_labels,
        save_to_path='../data/mnist_train_encodings_6')
    autoencoder.generate_encodings(
        test_images,
        test_labels,
        save_to_path='../data/mnist_test_encodings_6')

Beispiel #5

Datei: run_dbn.py Projekt: yy2261/Deep-Learning-TensorFlow

# Parameters validation
assert FLAGS.dataset in ['mnist', 'cifar10', 'custom']
assert FLAGS.finetune_act_func in ['sigmoid', 'tanh', 'relu']
assert len(rbm_layers) > 0

if __name__ == '__main__':

    utilities.random_seed_np_tf(FLAGS.seed)

    if FLAGS.dataset == 'mnist':

        # ################# #
        #   MNIST Dataset   #
        # ################# #

        trX, trY, vlX, vlY, teX, teY = datasets.load_mnist_dataset(mode='supervised')

    elif FLAGS.dataset == 'cifar10':

        # ################### #
        #   Cifar10 Dataset   #
        # ################### #

        trX, trY, teX, teY = datasets.load_cifar10_dataset(FLAGS.cifar_dir, mode='supervised')
        vlX = teX[:5000]  # Validation set is the first half of the test set
        vlY = teY[:5000]

    elif FLAGS.dataset == 'custom':

        # ################## #
        #   Custom Dataset   #

Beispiel #6

Datei: mnist_encoding_generator.py Projekt: Jakobovski/coremind

def generate_encodings():
    """Generates encodings for MNIST dataset."""
    train_images, train_labels, validation_images, validation_labels, test_images, test_labels = datasets.load_mnist_dataset(mode='supervised')

    # Convert one-hot to integer
    train_labels = [np.argmax(label) for label in train_labels]
    test_labels = [np.argmax(label) for label in test_labels]

    # Initialize the autoencoder
    autoencoder = Autoencoder([2048, 1024, 256, 64], pretrain=False, pretrain_epochs=0, finetune_epochs=120, finetune_batch_size=64)
    autoencoder.train(train_images)
    autoencoder.generate_encodings(train_images, train_labels, save_to_path='../data/mnist_train_encodings_3')
    autoencoder.generate_encodings(test_images, test_labels, save_to_path='../data/mnist_test_encodings_3')