CIFAR-10 on keras MUCH deeper NN.ipynb

{
    "metadata": {
        "language_info": {
            "codemirror_mode": {
                "name": "ipython", 
                "version": 2
            }, 
            "pygments_lexer": "ipython2", 
            "file_extension": ".py", 
            "name": "python", 
            "version": "2.7.11", 
            "mimetype": "text/x-python", 
            "nbconvert_exporter": "python"
        }, 
        "kernelspec": {
            "language": "python", 
            "display_name": "Python 2 with Spark 2.1", 
            "name": "python2-spark21"
        }
    }, 
    "nbformat": 4, 
    "cells": [
        {
            "metadata": {}, 
            "execution_count": 2, 
            "cell_type": "code", 
            "source": "# Large CNN model for the CIFAR-10 Dataset\nimport numpy\nfrom keras.datasets import cifar10\nfrom keras.models import Sequential\nfrom keras.layers import Dense\nfrom keras.layers import Dropout\nfrom keras.layers import Flatten\nfrom keras.constraints import maxnorm\nfrom keras.optimizers import SGD\nfrom keras.layers.convolutional import Conv2D\nfrom keras.layers.convolutional import MaxPooling2D\nfrom keras.utils import np_utils\nfrom keras import backend as K\nK.set_image_dim_ordering('th')\n# fix random seed for reproducibility\nseed = 7\nnumpy.random.seed(seed)\n# load data\n(X_train, y_train), (X_test, y_test) = cifar10.load_data()\n# normalize inputs from 0-255 to 0.0-1.0\nX_train = X_train.astype('float32')\nX_test = X_test.astype('float32')\nX_train = X_train / 255.0\nX_test = X_test / 255.0\n# one hot encode outputs\ny_train = np_utils.to_categorical(y_train)\ny_test = np_utils.to_categorical(y_test)\nnum_classes = y_test.shape[1]", 
            "outputs": [
                {
                    "name": "stderr", 
                    "output_type": "stream", 
                    "text": "Using TensorFlow backend.\n"
                }
            ]
        }, 
        {
            "metadata": {}, 
            "execution_count": 4, 
            "cell_type": "code", 
            "source": "# Create the model\nmodel = Sequential()\nmodel.add(Conv2D(32, (3, 3), input_shape=(3, 32, 32), activation='relu', padding='same'))\nmodel.add(Dropout(0.2))\nmodel.add(Conv2D(32, (3, 3), activation='relu', padding='same'))\nmodel.add(MaxPooling2D(pool_size=(2, 2)))\nmodel.add(Conv2D(64, (3, 3), activation='relu', padding='same'))\nmodel.add(Dropout(0.2))\nmodel.add(Conv2D(64, (3, 3), activation='relu', padding='same'))\nmodel.add(MaxPooling2D(pool_size=(2, 2)))\nmodel.add(Conv2D(128, (3, 3), activation='relu', padding='same'))\nmodel.add(Dropout(0.2))\nmodel.add(Conv2D(128, (3, 3), activation='relu', padding='same'))\nmodel.add(MaxPooling2D(pool_size=(2, 2)))\nmodel.add(Flatten())\nmodel.add(Dropout(0.2))\nmodel.add(Dense(1024, activation='relu', kernel_constraint=maxnorm(3)))\nmodel.add(Dropout(0.2))\nmodel.add(Dense(512, activation='relu', kernel_constraint=maxnorm(3)))\nmodel.add(Dropout(0.2))\nmodel.add(Dense(num_classes, activation='softmax'))", 
            "outputs": []
        }, 
        {
            "metadata": {}, 
            "execution_count": 6, 
            "cell_type": "code", 
            "source": "# Compile model\nepochs = 25\nlrate = 0.01\ndecay = lrate/epochs\nsgd = SGD(lr=lrate, momentum=0.9, decay=decay, nesterov=False)\nmodel.compile(loss='categorical_crossentropy', optimizer=sgd, metrics=['accuracy'])\nprint(model.summary())", 
            "outputs": [
                {
                    "name": "stdout", 
                    "output_type": "stream", 
                    "text": "_________________________________________________________________\nLayer (type)                 Output Shape              Param #   \n=================================================================\nconv2d_1 (Conv2D)            (None, 32, 32, 32)        896       \n_________________________________________________________________\ndropout_1 (Dropout)          (None, 32, 32, 32)        0         \n_________________________________________________________________\nconv2d_2 (Conv2D)            (None, 32, 32, 32)        9248      \n_________________________________________________________________\nmax_pooling2d_1 (MaxPooling2 (None, 32, 16, 16)        0         \n_________________________________________________________________\nconv2d_3 (Conv2D)            (None, 64, 16, 16)        18496     \n_________________________________________________________________\ndropout_2 (Dropout)          (None, 64, 16, 16)        0         \n_________________________________________________________________\nconv2d_4 (Conv2D)            (None, 64, 16, 16)        36928     \n_________________________________________________________________\nmax_pooling2d_2 (MaxPooling2 (None, 64, 8, 8)          0         \n_________________________________________________________________\nconv2d_5 (Conv2D)            (None, 128, 8, 8)         73856     \n_________________________________________________________________\ndropout_3 (Dropout)          (None, 128, 8, 8)         0         \n_________________________________________________________________\nconv2d_6 (Conv2D)            (None, 128, 8, 8)         147584    \n_________________________________________________________________\nmax_pooling2d_3 (MaxPooling2 (None, 128, 4, 4)         0         \n_________________________________________________________________\nflatten_1 (Flatten)          (None, 2048)              0         \n_________________________________________________________________\ndropout_4 (Dropout)          (None, 2048)              0         \n_________________________________________________________________\ndense_1 (Dense)              (None, 1024)              2098176   \n_________________________________________________________________\ndropout_5 (Dropout)          (None, 1024)              0         \n_________________________________________________________________\ndense_2 (Dense)              (None, 512)               524800    \n_________________________________________________________________\ndropout_6 (Dropout)          (None, 512)               0         \n_________________________________________________________________\ndense_3 (Dense)              (None, 10)                5130      \n=================================================================\nTotal params: 2,915,114\nTrainable params: 2,915,114\nNon-trainable params: 0\n_________________________________________________________________\nNone\n"
                }
            ]
        }, 
        {
            "metadata": {}, 
            "execution_count": null, 
            "cell_type": "code", 
            "source": "# Fit the model\nmodel.fit(X_train, y_train, validation_data=(X_test, y_test), epochs=epochs, batch_size=64)\n", 
            "outputs": []
        }, 
        {
            "metadata": {}, 
            "execution_count": null, 
            "cell_type": "code", 
            "source": "# Final evaluation of the model\nscores = model.evaluate(X_test, y_test, verbose=0)\nprint(\"Accuracy: %.2f%%\" % (scores[1]*100))", 
            "outputs": []
        }
    ], 
    "nbformat_minor": 1
}