nzw0301/shallow_CNN_swish_cifar10.ipynb

## shallow_CNN_swish_cifar10.ipynb
{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.models import Sequential\n",
    "from keras.datasets import cifar10\n",
    "from keras.layers import Dense, Activation, Flatten, Lambda, BatchNormalization\n",
    "from keras.utils import np_utils\n",
    "from keras.layers import Conv2D, MaxPool2D\n",
    "from keras import backend as K\n",
    "\n",
    "from matplotlib import pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "import numpy as np\n",
    "np.random.seed(13)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X_train shape: (50000, 32, 32, 3)\n",
      "50000 train samples\n",
      "10000 test samples\n"
     ]
    }
   ],
   "source": [
    "batch_size = 128\n",
    "num_classes = 10\n",
    "epochs = 10\n",
    "num_filters = 10\n",
    "\n",
    "img_rows, img_cols = 32, 32\n",
    "img_channels = 3\n",
    "\n",
    "(x_train, y_train), (x_test, y_test) = cifar10.load_data()\n",
    "print('X_train shape:', x_train.shape)\n",
    "print(x_train.shape[0], 'train samples')\n",
    "print(x_test.shape[0], 'test samples')\n",
    "x_train = x_train.astype('float32')\n",
    "x_test = x_test.astype('float32')\n",
    "x_train /= 255\n",
    "x_test /= 255\n",
    "\n",
    "y_train = np_utils.to_categorical(y_train, num_classes)\n",
    "y_test = np_utils.to_categorical(y_test, num_classes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = Sequential()\n",
    "model.add(Conv2D(num_filters, (3, 3), input_shape=(img_rows, img_cols, img_channels)))\n",
    "model.add(BatchNormalization())\n",
    "model.add(Activation('relu'))\n",
    "model.add(Conv2D(num_filters, (3, 3)))\n",
    "model.add(BatchNormalization())\n",
    "model.add(Activation('relu'))\n",
    "model.add(MaxPool2D(pool_size=(2, 2)))\n",
    "model.add(Flatten())\n",
    "model.add(Dense(512))\n",
    "model.add(BatchNormalization())\n",
    "model.add(Activation('relu'))          \n",
    "model.add(Dense(num_classes, activation='softmax'))\n",
    "\n",
    "model.compile(loss='categorical_crossentropy',\n",
    "              optimizer='adadelta',\n",
    "              metrics=['accuracy'])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 50000 samples, validate on 10000 samples\n",
      "Epoch 1/10\n",
      "50000/50000 [==============================] - 157s - loss: 1.4029 - acc: 0.5035 - val_loss: 1.7708 - val_acc: 0.3682\n",
      "Epoch 2/10\n",
      "50000/50000 [==============================] - 154s - loss: 1.0331 - acc: 0.6409 - val_loss: 1.2078 - val_acc: 0.5777\n",
      "Epoch 3/10\n",
      "50000/50000 [==============================] - 151s - loss: 0.8502 - acc: 0.7088 - val_loss: 1.0877 - val_acc: 0.6213\n",
      "Epoch 4/10\n",
      "50000/50000 [==============================] - 149s - loss: 0.7025 - acc: 0.7659 - val_loss: 1.1030 - val_acc: 0.6198\n",
      "Epoch 5/10\n",
      "50000/50000 [==============================] - 150s - loss: 0.5818 - acc: 0.8128 - val_loss: 1.0825 - val_acc: 0.6333\n",
      "Epoch 6/10\n",
      "50000/50000 [==============================] - 149s - loss: 0.4718 - acc: 0.8567 - val_loss: 1.0877 - val_acc: 0.6355\n",
      "Epoch 7/10\n",
      "50000/50000 [==============================] - 140s - loss: 0.3754 - acc: 0.8962 - val_loss: 1.1202 - val_acc: 0.6293\n",
      "Epoch 8/10\n",
      "50000/50000 [==============================] - 148s - loss: 0.2917 - acc: 0.9273 - val_loss: 1.1751 - val_acc: 0.6332\n",
      "Epoch 9/10\n",
      "50000/50000 [==============================] - 154s - loss: 0.2219 - acc: 0.9517 - val_loss: 1.2387 - val_acc: 0.6271\n",
      "Epoch 10/10\n",
      "50000/50000 [==============================] - 148s - loss: 0.1663 - acc: 0.9693 - val_loss: 1.2991 - val_acc: 0.6231\n"
     ]
    }
   ],
   "source": [
    "relu_res = model.fit(x_train, y_train, batch_size=batch_size, epochs=epochs, validation_data=(x_test, y_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 50000 samples, validate on 10000 samples\n",
      "Epoch 1/10\n",
      "50000/50000 [==============================] - 157s - loss: 1.4018 - acc: 0.5049 - val_loss: 1.6820 - val_acc: 0.4100\n",
      "Epoch 2/10\n",
      "50000/50000 [==============================] - 148s - loss: 1.0685 - acc: 0.6240 - val_loss: 1.1553 - val_acc: 0.5933\n",
      "Epoch 3/10\n",
      "50000/50000 [==============================] - 149s - loss: 0.8963 - acc: 0.6917 - val_loss: 1.1347 - val_acc: 0.6011\n",
      "Epoch 4/10\n",
      "50000/50000 [==============================] - 152s - loss: 0.7695 - acc: 0.7377 - val_loss: 1.0776 - val_acc: 0.6287\n",
      "Epoch 5/10\n",
      "50000/50000 [==============================] - 149s - loss: 0.6602 - acc: 0.7804 - val_loss: 1.1297 - val_acc: 0.6189\n",
      "Epoch 6/10\n",
      "50000/50000 [==============================] - 171s - loss: 0.5649 - acc: 0.8156 - val_loss: 1.0652 - val_acc: 0.6482\n",
      "Epoch 7/10\n",
      "50000/50000 [==============================] - 164s - loss: 0.4732 - acc: 0.8515 - val_loss: 1.0692 - val_acc: 0.6561\n",
      "Epoch 8/10\n",
      "50000/50000 [==============================] - 167s - loss: 0.3933 - acc: 0.8818 - val_loss: 1.1042 - val_acc: 0.6501\n",
      "Epoch 9/10\n",
      "50000/50000 [==============================] - 162s - loss: 0.3205 - acc: 0.9085 - val_loss: 1.1198 - val_acc: 0.6429\n",
      "Epoch 10/10\n",
      "50000/50000 [==============================] - 151s - loss: 0.2596 - acc: 0.9313 - val_loss: 1.2499 - val_acc: 0.6320\n"
     ]
    }
   ],
   "source": [
    "model = Sequential()\n",
    "model.add(Conv2D(num_filters, (3, 3), input_shape=(img_rows, img_cols, img_channels)))\n",
    "model.add(BatchNormalization())\n",
    "model.add(Lambda(lambda x: x*K.sigmoid(x)))\n",
    "model.add(Conv2D(num_filters, (3, 3)))\n",
    "model.add(BatchNormalization())\n",
    "model.add(Lambda(lambda x: x*K.sigmoid(x)))\n",
    "model.add(MaxPool2D(pool_size=(2, 2)))\n",
    "model.add(Flatten())\n",
    "model.add(Dense(512))\n",
    "model.add(BatchNormalization())\n",
    "model.add(Lambda(lambda x: x*K.sigmoid(x)))\n",
    "model.add(Dense(num_classes, activation='softmax'))\n",
    "\n",
    "model.compile(loss='categorical_crossentropy',\n",
    "              optimizer='adadelta',\n",
    "              metrics=['accuracy'])\n",
    "\n",
    "swish_res = model.fit(x_train, y_train, batch_size=batch_size, epochs=epochs, validation_data=(x_test, y_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAD9CAYAAABHnDf0AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAf0X6euhixOVDuIAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x18857f7cc0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAD8CAYAAACMwORRAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nxsQhC3djjIlDFu7GGBOHLNyNMSYO/R9gJNSLOgzKJAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1885df6c18>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "x = range(epochs)\n",
    "plt.plot(x, relu_res.history['acc'], label='relu train')\n",
    "plt.plot(x, relu_res.history['val_acc'], label='relu val')\n",
    "plt.plot(x, swish_res.history['acc'], label='swish train')\n",
    "plt.plot(x, swish_res.history['val_acc'], label='swish val')\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "plt.plot(x, relu_res.history['loss'], label='relu train')\n",
    "plt.plot(x, relu_res.history['val_loss'], label='relu val')\n",
    "plt.plot(x, swish_res.history['loss'], label='swish train')\n",
    "plt.plot(x, swish_res.history['val_loss'], label='swish val')\n",
    "plt.legend()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}