Function cg_optimization_mnist

code/logistic_cg.py:145–298

Demonstrate conjugate gradient optimization of a log-linear model This is demonstrated on MNIST. :type n_epochs: int :param n_epochs: number of epochs to run the optimizer :type mnist_pkl_gz: string :param mnist_pkl_gz: the path of the mnist training file from

(n_epochs=50, mnist_pkl_gz='mnist.pkl.gz')

def cg_optimization_mnist(n_epochs=50, mnist_pkl_gz='mnist.pkl.gz'):
    """Demonstrate conjugate gradient optimization of a log-linear model

    This is demonstrated on MNIST.

    :type n_epochs: int
    :param n_epochs: number of epochs to run the optimizer

    :type mnist_pkl_gz: string
    :param mnist_pkl_gz: the path of the mnist training file from
                 http://www.iro.umontreal.ca/~lisa/deep/data/mnist/mnist.pkl.gz

    """
    #############
    # LOAD DATA #
    #############
    datasets = load_data(mnist_pkl_gz)

    train_set_x, train_set_y = datasets[0]
    valid_set_x, valid_set_y = datasets[1]
    test_set_x, test_set_y = datasets[2]

    batch_size = 600    # size of the minibatch

    n_train_batches = train_set_x.get_value(borrow=True).shape[0] // batch_size
    n_valid_batches = valid_set_x.get_value(borrow=True).shape[0] // batch_size
    n_test_batches = test_set_x.get_value(borrow=True).shape[0] // batch_size

    n_in = 28 * 28  # number of input units
    n_out = 10  # number of output units

    ######################
    # BUILD ACTUAL MODEL #
    ######################
    print('... building the model')

    # allocate symbolic variables for the data
    minibatch_offset = T.lscalar()  # offset to the start of a [mini]batch
    x = T.matrix()   # the data is presented as rasterized images
    y = T.ivector()  # the labels are presented as 1D vector of
                     # [int] labels

    # construct the logistic regression class
    classifier = LogisticRegression(input=x, n_in=28 * 28, n_out=10)

    # the cost we minimize during training is the negative log likelihood of
    # the model in symbolic format
    cost = classifier.negative_log_likelihood(y).mean()

    # compile a theano function that computes the mistakes that are made by
    # the model on a minibatch
    test_model = theano.function(
        [minibatch_offset],
        classifier.errors(y),
        givens={
            x: test_set_x[minibatch_offset:minibatch_offset + batch_size],
            y: test_set_y[minibatch_offset:minibatch_offset + batch_size]
        },