cifar_10_with_Alexnet.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "cifar-10-with-Alexnet.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "machine_shape": "hm",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/pravinkr/alexnet-cifar10-using-keras/blob/master/cifar_10_with_Alexnet.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0uoD7EUdjgXP",
        "colab_type": "code",
        "outputId": "585d5484-d21c-4226-dad9-e1918d512ab4",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 80
        }
      },
      "source": [
        "#Simple CNN model for CIFAR-10 dataset\n",
        "import numpy as np\n",
        "\n",
        "# Simple CNN model for CIFAR-10\n",
        "import numpy as np\n",
        "import os\n",
        "from keras.datasets import cifar10, cifar100\n",
        "from keras.models import Sequential\n",
        "from keras.layers import Dense, Conv2D\n",
        "from keras.layers import Dropout\n",
        "from keras.layers import Flatten\n",
        "from keras.constraints import maxnorm\n",
        "from keras.optimizers import SGD, Adam\n",
        "from keras.layers.convolutional import Convolution2D\n",
        "from keras.layers.convolutional import MaxPooling2D\n",
        "from keras.callbacks import ModelCheckpoint, LearningRateScheduler\n",
        "from keras.callbacks import ReduceLROnPlateau\n",
        "from keras.callbacks import EarlyStopping\n",
        "from keras.regularizers import l2\n",
        "from keras import backend as K\n",
        "\n",
        "from keras.utils import np_utils\n",
        "import matplotlib.pyplot as plt\n",
        "from keras.preprocessing.image import ImageDataGenerator\n",
        "%matplotlib inline\n",
        "\n",
        "import skimage\n",
        "from skimage.util import img_as_ubyte\n"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Using TensorFlow backend.\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<p style=\"color: red;\">\n",
              "The default version of TensorFlow in Colab will soon switch to TensorFlow 2.x.<br>\n",
              "We recommend you <a href=\"https://www.tensorflow.org/guide/migrate\" target=\"_blank\">upgrade</a> now \n",
              "or ensure your notebook will continue to use TensorFlow 1.x via the <code>%tensorflow_version 1.x</code> magic:\n",
              "<a href=\"https://colab.research.google.com/notebooks/tensorflow_version.ipynb\" target=\"_blank\">more info</a>.</p>\n"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "k34XYTn_gbVF",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        },
        "outputId": "f7eb15db-1aa0-43b3-b44e-e7d194041c59"
      },
      "source": [
        "#Defining Variables\n",
        "\n",
        "#Data set information\n",
        "DATASET = 'cifar-10'\n",
        "#DATASET = 'cifar-100'\n",
        "input_shape=(32,32,3)\n",
        "\n",
        "if DATASET == 'cifar-10':\n",
        "  num_classes = 10\n",
        "elif DATASET == 'cifar-100':\n",
        "  num_classes = 100\n",
        "\n",
        "#Model Parameters\n",
        "\n",
        "model_type = 'Alexnet'\n",
        "#model_type = 'Resnet'\n",
        "\n",
        "epochs = 25\n",
        "lrate = 0.01\n",
        "decay = lrate/epochs\n",
        "batch_size = 32\n",
        "\n",
        "data_augmentation = True\n",
        "flow_from_dir = False\n",
        "\n",
        "subtract_mean = True\n",
        "seed = 7\n",
        "np.random.seed(seed)\n",
        "\n",
        "validate_dir = os.path.join(\"data\",\"test\")\n",
        "train_dir = os.path.join(\"data\",\"train\")\n",
        "\n",
        "# Prepare model model saving directory.\n",
        "save_dir = os.path.join(os.getcwd(), 'saved_models')\n",
        "\n",
        "#model_name = 'cifar10_%s_model.{epoch:03d}.h5' % model_type\n",
        "model_name = DATASET+'_'+'%s_model.{epoch:03d}.h5' % model_type\n",
        "\n",
        "if not os.path.isdir(save_dir):\n",
        "    os.makedirs(save_dir)\n",
        "filepath = os.path.join(save_dir, model_name)\n",
        "print(\"Model Type: \",model_type)\n",
        "print(\"Model File: \",filepath)"
      ],
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model Type:  Alexnet\n",
            "Model File:  /content/saved_models/cifar-10_Alexnet_model.{epoch:03d}.h5\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BUmkHwlR4Bex",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "if flow_from_dir:\n",
        "  if DATASET == 'cifar-10':\n",
        "    %pip install cifar2png\n",
        "    %cifar2png cifar10 \"data\""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pJ39ZC-BXg37",
        "colab_type": "code",
        "outputId": "d163ee06-7f91-4c49-af8c-73f782d995ba",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 268
        }
      },
      "source": [
        "if not flow_from_dir:\n",
        "  #load data\n",
        "  if DATASET == 'cifar-10':\n",
        "    (X_train, y_train),(X_test,y_test) = cifar10.load_data()\n",
        "    X_train.shape, X_test.shape, X_train.shape[1:],X_train.dtype\n",
        "  elif DATASET == 'cifar-100':\n",
        "    (X_train, y_train),(X_test,y_test) = cifar100.load_data()\n",
        "    X_train.shape, X_test.shape, X_train.shape[1:],X_train.dtype\n",
        "\n",
        "  #Plot data\n",
        "  for i in range(0,9):\n",
        "    plt.subplot(330+1+i)\n",
        "    plt.imshow(X_train[i])\n",
        "  plt.show() \n",
        "  \n"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAU4AAAD7CAYAAAAFI30bAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 432x288 with 9 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "R_4A1KRAhfbe",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def convert_to_imagenet_size(images):\n",
        "  tmp_images = []\n",
        "  for image in images:\n",
        "    image = skimage.transform.resize(image,(224,224),mode='constant')\n",
        "    image = img_as_ubyte(image)\n",
        "    tmp_images.append(image)\n",
        "  return np.array(tmp_images)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QVJcPP50W2ri",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "e1b49a99-0375-4c75-8104-189aed693d39"
      },
      "source": [
        "#Data Normalization\n",
        "if not flow_from_dir:\n",
        "  if not data_augmentation:\n",
        "    X_train = X_train.astype('float32')\n",
        "    X_test = X_test.astype('float32')\n",
        "    X_train = X_train/255.\n",
        "    X_test = X_test/255.\n",
        "\n",
        "    if subtract_mean:\n",
        "      X_train_mean = np.mean(X_train)\n",
        "      X_train -= X_train_mean\n",
        "      X_test_mean = np.mean(X_test)\n",
        "      X_test -= X_test_mean\n",
        "\n",
        "  print(y_train.shape, y_test.shape)"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(50000, 1) (10000, 1)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XerWb6T3YK-x",
        "colab_type": "code",
        "outputId": "e638818d-01d9-44a7-d156-21277c037389",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "if not flow_from_dir:\n",
        "  y_train = np_utils.to_categorical(y_train)\n",
        "  y_test = np_utils.to_categorical(y_test)\n",
        "  print(y_train.shape, y_test.shape)"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(50000, 10) (10000, 10)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PS86-ZpEZMUY",
        "colab_type": "code",
        "outputId": "ae1a217d-da9a-4a4b-8fc3-a368fd98f46f",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "#num_classes = y_train.shape[1]\n",
        "#input_shape=X_train.shape[1:]\n",
        "\n",
        "input_shape,num_classes"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((32, 32, 3), 10)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hXFuLmd6bjpY",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#Define Alexnet Model\n",
        "def AlexnetModel(input_shape,num_classes):\n",
        "  model = Sequential()\n",
        "  model.add(Conv2D(filters=96,kernel_size=(3,3),strides=(4,4),input_shape=input_shape, activation='relu'))\n",
        "  model.add(MaxPooling2D(pool_size=(2,2),strides=(2,2)))\n",
        "  model.add(Conv2D(256,(5,5),padding='same',activation='relu'))\n",
        "  model.add(MaxPooling2D(pool_size=(2,2),strides=(2,2)))\n",
        "  model.add(Conv2D(384,(3,3),padding='same',activation='relu'))\n",
        "  model.add(Conv2D(384,(3,3),padding='same',activation='relu'))\n",
        "  model.add(Conv2D(256,(3,3),padding='same',activation='relu'))\n",
        "  model.add(MaxPooling2D(pool_size=(2,2),strides=(2,2)))\n",
        "\n",
        "  model.add(Flatten())\n",
        "  model.add(Dense(4096, activation='relu'))\n",
        "  model.add(Dropout(0.4))\n",
        "  model.add(Dense(4096, activation='relu'))\n",
        "  model.add(Dropout(0.4))\n",
        "  model.add(Dense(num_classes,activation='softmax'))\n",
        "\n",
        "  #model.compile(optimizer='adam',loss='categorical_crossentropy',metrics=['accuracy'])\n",
        "\n",
        "  #model.summary()\n",
        "  return model"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eRNzvi-qZXvl",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def get_model(input_shape,num_classes):\n",
        "  # Create the model - Normal model with CNN, dropouts and MaxPooling.\n",
        "  # This model gives accuracy of 77% on the test set after 25 epochs\n",
        "\n",
        "  model = Sequential()\n",
        "\n",
        "  #model.add(Conv2D(32,(3,3),input_shape=(32,32,3),padding='same',activation='relu'))\n",
        "  model.add(Conv2D(32,(3,3),input_shape=input_shape,padding='same',activation='relu'))\n",
        "  model.add(Dropout(0.2))\n",
        "\n",
        "  model.add(Conv2D(32,(3,3),activation='relu',padding='same'))\n",
        "  model.add(MaxPooling2D(pool_size=(2,2)))\n",
        "\n",
        "  model.add(Conv2D(64,(3,3),activation='relu',padding='same'))\n",
        "  model.add(Dropout(0.2))\n",
        "\n",
        "  model.add(Conv2D(64,(3,3),activation='relu',padding='same'))\n",
        "  model.add(MaxPooling2D(pool_size=(2,2)))\n",
        "\n",
        "  model.add(Conv2D(128,(3,3),activation='relu',padding='same'))\n",
        "  model.add(Dropout(0.2))\n",
        "\n",
        "  model.add(Conv2D(128,(3,3),activation='relu',padding='same'))\n",
        "  model.add(MaxPooling2D(pool_size=(2,2)))\n",
        "\n",
        "  model.add(Flatten())\n",
        "  model.add(Dropout(0.2))\n",
        "\n",
        "  model.add(Dense(1024,activation='relu'))\n",
        "  model.add(Dropout(0.2))\n",
        "\n",
        "  model.add(Dense(512,activation='relu'))\n",
        "  model.add(Dropout(0.2))\n",
        "\n",
        "  model.add(Dense(num_classes,activation='softmax'))\n",
        "\n",
        "  return model\n",
        "\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "p2cGLs6n9l0L",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def lr_schedule(epoch):\n",
        "    \"\"\"Learning Rate Schedule\n",
        "\n",
        "    Learning rate is scheduled to be reduced after 80, 120, 160, 180 epochs.\n",
        "    Called automatically every epoch as part of callbacks during training.\n",
        "\n",
        "    # Arguments\n",
        "        epoch (int): The number of epochs\n",
        "\n",
        "    # Returns\n",
        "        lr (float32): learning rate\n",
        "    \"\"\"\n",
        "    lr = 1e-3\n",
        "    if epoch > 180:\n",
        "        lr *= 0.5e-3\n",
        "    elif epoch > 160:\n",
        "        lr *= 1e-3\n",
        "    elif epoch > 120:\n",
        "        lr *= 1e-2\n",
        "    elif epoch > 80:\n",
        "        lr *= 1e-1\n",
        "    print('Learning rate: ', lr)\n",
        "    return lr\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sdnE9Sd0-NbF",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Prepare callbacks for model saving and for learning rate adjustment.\n",
        "\n",
        "early_stopping = EarlyStopping(monitor='val_loss', min_delta=0.001, patience=5, verbose=0, mode='auto', baseline=None, restore_best_weights=True)\n",
        "\n",
        "checkpoint = ModelCheckpoint(filepath=filepath,\n",
        "                             monitor='val_acc',\n",
        "                             verbose=1,\n",
        "                             save_best_only=True)\n",
        "\n",
        "lr_scheduler = LearningRateScheduler(lr_schedule)\n",
        "\n",
        "lr_reducer = ReduceLROnPlateau(factor=np.sqrt(0.1),\n",
        "                               cooldown=0,\n",
        "                               patience=5,\n",
        "                               min_lr=0.5e-6)\n",
        "\n",
        "callbacks = [checkpoint, lr_reducer, lr_scheduler,early_stopping]\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AbLu2lefac_h",
        "colab_type": "code",
        "outputId": "7b1e97aa-92ac-457f-e905-ad55b736cedd",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 972
        }
      },
      "source": [
        "#model = get_model(input_shape,num_classes)\n",
        "if model_type == 'Alexnet':\n",
        "  #input_shape = (224,224,3)\n",
        "  model = AlexnetModel(input_shape,num_classes)\n",
        "else:\n",
        "  model = get_model(input_shape,num_classes)\n",
        "\n",
        "# Compile model\n",
        "\n",
        "#optimizer = SGD(lr=lrate, momentum=0.9, decay=decay, nesterov=False)\n",
        "optimizer = Adam(lr=0.001, beta_1=0.9, beta_2=0.999, amsgrad=False)\n",
        "model.compile(loss= 'categorical_crossentropy' , optimizer=optimizer, metrics=[ 'accuracy' ])\n",
        "print(\"Model Summary of \",model_type)\n",
        "print(model.summary())\n",
        "\n"
      ],
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:66: The name tf.get_default_graph is deprecated. Please use tf.compat.v1.get_default_graph instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:541: The name tf.placeholder is deprecated. Please use tf.compat.v1.placeholder instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:4432: The name tf.random_uniform is deprecated. Please use tf.random.uniform instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:4267: The name tf.nn.max_pool is deprecated. Please use tf.nn.max_pool2d instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:148: The name tf.placeholder_with_default is deprecated. Please use tf.compat.v1.placeholder_with_default instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:3733: calling dropout (from tensorflow.python.ops.nn_ops) with keep_prob is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Please use `rate` instead of `keep_prob`. Rate should be set to `rate = 1 - keep_prob`.\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/optimizers.py:793: The name tf.train.Optimizer is deprecated. Please use tf.compat.v1.train.Optimizer instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:3576: The name tf.log is deprecated. Please use tf.math.log instead.\n",
            "\n",
            "Model Summary of  Alexnet\n",
            "Model: \"sequential_1\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_1 (Conv2D)            (None, 8, 8, 96)          2688      \n",
            "_________________________________________________________________\n",
            "max_pooling2d_1 (MaxPooling2 (None, 4, 4, 96)          0         \n",
            "_________________________________________________________________\n",
            "conv2d_2 (Conv2D)            (None, 4, 4, 256)         614656    \n",
            "_________________________________________________________________\n",
            "max_pooling2d_2 (MaxPooling2 (None, 2, 2, 256)         0         \n",
            "_________________________________________________________________\n",
            "conv2d_3 (Conv2D)            (None, 2, 2, 384)         885120    \n",
            "_________________________________________________________________\n",
            "conv2d_4 (Conv2D)            (None, 2, 2, 384)         1327488   \n",
            "_________________________________________________________________\n",
            "conv2d_5 (Conv2D)            (None, 2, 2, 256)         884992    \n",
            "_________________________________________________________________\n",
            "max_pooling2d_3 (MaxPooling2 (None, 1, 1, 256)         0         \n",
            "_________________________________________________________________\n",
            "flatten_1 (Flatten)          (None, 256)               0         \n",
            "_________________________________________________________________\n",
            "dense_1 (Dense)              (None, 4096)              1052672   \n",
            "_________________________________________________________________\n",
            "dropout_1 (Dropout)          (None, 4096)              0         \n",
            "_________________________________________________________________\n",
            "dense_2 (Dense)              (None, 4096)              16781312  \n",
            "_________________________________________________________________\n",
            "dropout_2 (Dropout)          (None, 4096)              0         \n",
            "_________________________________________________________________\n",
            "dense_3 (Dense)              (None, 10)                40970     \n",
            "=================================================================\n",
            "Total params: 21,589,898\n",
            "Trainable params: 21,589,898\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "None\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AIEd9V4KMYR7",
        "colab_type": "code",
        "outputId": "58a18034-df34-4209-f617-410ca42103e5",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "if data_augmentation:\n",
        "\tprint('Using real-time data augmentation.')\n",
        "\t# This will do preprocessing and realtime data augmentation:\n",
        "\tdatagen = ImageDataGenerator(\n",
        "        featurewise_center=True,  # set input mean to 0 over the dataset\n",
        "        samplewise_center=False,  # set each sample mean to 0\n",
        "        featurewise_std_normalization=True,  # divide inputs by std of the dataset\n",
        "        samplewise_std_normalization=False,  # divide each input by its std\n",
        "        zca_whitening=False,  # apply ZCA whitening\n",
        "        zca_epsilon=1e-06,  # epsilon for ZCA whitening\n",
        "        rotation_range=20,  # randomly rotate images in the range (degrees, 0 to 180)\n",
        "        # randomly shift images horizontally (fraction of total width)\n",
        "        width_shift_range=0.2,\n",
        "        # randomly shift images vertically (fraction of total height)\n",
        "        height_shift_range=0.2,\n",
        "        shear_range=0.,  # set range for random shear\n",
        "        zoom_range=0.,  # set range for random zoom\n",
        "        channel_shift_range=0.,  # set range for random channel shifts\n",
        "        # set mode for filling points outside the input boundaries\n",
        "        fill_mode='nearest',\n",
        "        cval=0.,  # value used for fill_mode = \"constant\"\n",
        "        horizontal_flip=True,  # randomly flip images\n",
        "        vertical_flip=False,  # randomly flip images\n",
        "        # set rescaling factor (applied before any other transformation)\n",
        "        rescale=None,\n",
        "        # set function that will be applied on each input\n",
        "        preprocessing_function=None,\n",
        "        # image data format, either \"channels_first\" or \"channels_last\"\n",
        "        data_format=None,\n",
        "        # fraction of images reserved for validation (strictly between 0 and 1)\n",
        "        validation_split=0.0)\n",
        "\n",
        "    # Compute quantities required for feature-wise normalization\n",
        "    # (std, mean, and principal components if ZCA whitening is applied).\n",
        "\n",
        "\n",
        "if not flow_from_dir:\n",
        "\tif not data_augmentation:\n",
        "\t\tprint('Not using data augmentation.')\n",
        "\t\thistory = model.fit(X_train, y_train,\n",
        "          batch_size=batch_size,\n",
        "          epochs=epochs,\n",
        "          validation_data=(X_test, y_test),\n",
        "          shuffle=True,\n",
        "          callbacks=callbacks)\n",
        "\telse:   \n",
        "\t\tprint(\"model using data from downloaded dataset in memoory\")\n",
        "\t\tdatagen.fit(X_train)\n",
        "\t\t# Fit the model on the batches generated by datagen.flow().\n",
        "\t\thistory = model.fit_generator(datagen.flow(X_train, y_train,\n",
        "                                 batch_size=batch_size),\n",
        "                                 epochs=epochs,\n",
        "                                 validation_data=datagen.flow((X_test, y_test),batch_size=batch_size),\n",
        "                                 workers=4,\n",
        "                                 callbacks=callbacks)\n",
        "  \n",
        "\tscores = model.evaluate(X_test, y_test, verbose=1)\n",
        "\n",
        "else:\n",
        "\t# generator for reading train data from folder\n",
        "\n",
        "\tprint('Model fit using flow from directory')\n",
        "\n",
        "\n",
        "\ttrain_generator = datagen.flow_from_directory(\n",
        "      train_dir,\n",
        "      target_size = (224, 224),\n",
        "      color_mode = 'rgb',\n",
        "      batch_size = batch_size,\n",
        "      class_mode = 'categorical',\n",
        "      shuffle=False,\n",
        "      seed=42\n",
        "      )\n",
        "\t#\n",
        "\t# generator for reading validation data from folder\n",
        "\n",
        "\tvalidation_generator = datagen.flow_from_directory(\n",
        "      validate_dir,\n",
        "      target_size = (224, 224),\n",
        "      color_mode = 'rgb',\n",
        "      batch_size = batch_size,\n",
        "      class_mode = 'categorical',\n",
        "      shuffle=False,\n",
        "      seed=42)\n",
        "  \n",
        "\tSTEP_SIZE_TRAIN=train_generator.n//train_generator.batch_size\n",
        "\tSTEP_SIZE_VALID=validation_generator.n//validation_generator.batch_size\n",
        "  \n",
        "\thistory = model.fit_generator(generator=train_generator,\n",
        "                steps_per_epoch=STEP_SIZE_TRAIN,\n",
        "                validation_data=validation_generator,\n",
        "                validation_steps=STEP_SIZE_VALID,\n",
        "                epochs=10,\n",
        "                callbacks=callbacks\n",
        "                )\n",
        "  \n",
        "\tscores = model.evaluate_generator(generator=validation_generator,\n",
        "                           steps=STEP_SIZE_VALID, verbose=1)\n",
        "\n",
        "\t#Predict the output\n",
        "\t''' Use when to Predict the output\n",
        "\tSTEP_SIZE_TEST=test_generator.n//test_generator.batch_size\n",
        "\ttest_generator.reset()\n",
        "\tpred=model.predict_generator(test_generator,\n",
        "                               steps=STEP_SIZE_TEST,\n",
        "                               verbose=1)\n",
        "\n",
        "\t'''\n",
        "# Save model and weights\n",
        "\n",
        "\n",
        "\n",
        "# Score trained model.\n",
        "\n",
        "print('Test loss:', scores[0])\n",
        "print('Test accuracy:', scores[1])"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Using real-time data augmentation.\n",
            "model using data from downloaded dataset in memoory\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/ops/math_grad.py:1424: where (from tensorflow.python.ops.array_ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Use tf.where in 2.0, which has the same broadcast rule as np.where\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:1033: The name tf.assign_add is deprecated. Please use tf.compat.v1.assign_add instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:1020: The name tf.assign is deprecated. Please use tf.compat.v1.assign instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:3005: The name tf.Session is deprecated. Please use tf.compat.v1.Session instead.\n",
            "\n",
            "Epoch 1/25\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:190: The name tf.get_default_session is deprecated. Please use tf.compat.v1.get_default_session instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:197: The name tf.ConfigProto is deprecated. Please use tf.compat.v1.ConfigProto instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:207: The name tf.global_variables is deprecated. Please use tf.compat.v1.global_variables instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:216: The name tf.is_variable_initialized is deprecated. Please use tf.compat.v1.is_variable_initialized instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:223: The name tf.variables_initializer is deprecated. Please use tf.compat.v1.variables_initializer instead.\n",
            "\n",
            "Learning rate:  0.001\n",
            "1563/1563 [==============================] - 43s 27ms/step - loss: 1.9232 - acc: 0.2341 - val_loss: 1.7385 - val_acc: 0.3150\n",
            "\n",
            "Epoch 00001: val_acc improved from -inf to 0.31500, saving model to /content/saved_models/cifar10_Alexnet_model.001.h5\n",
            "Epoch 2/25\n",
            "Learning rate:  0.001\n",
            "1563/1563 [==============================] - 36s 23ms/step - loss: 1.6967 - acc: 0.3445 - val_loss: 1.6254 - val_acc: 0.3954\n",
            "\n",
            "Epoch 00002: val_acc improved from 0.31500 to 0.39540, saving model to /content/saved_models/cifar10_Alexnet_model.002.h5\n",
            "Epoch 3/25\n",
            "Learning rate:  0.001\n",
            "1563/1563 [==============================] - 35s 22ms/step - loss: 1.5809 - acc: 0.4172 - val_loss: 1.5175 - val_acc: 0.4419\n",
            "\n",
            "Epoch 00003: val_acc improved from 0.39540 to 0.44190, saving model to /content/saved_models/cifar10_Alexnet_model.003.h5\n",
            "Epoch 4/25\n",
            "Learning rate:  0.001\n",
            "1563/1563 [==============================] - 35s 22ms/step - loss: 1.5134 - acc: 0.4498 - val_loss: 1.4829 - val_acc: 0.4512\n",
            "\n",
            "Epoch 00004: val_acc improved from 0.44190 to 0.45120, saving model to /content/saved_models/cifar10_Alexnet_model.004.h5\n",
            "Epoch 5/25\n",
            "Learning rate:  0.001\n",
            "1563/1563 [==============================] - 35s 22ms/step - loss: 1.4614 - acc: 0.4726 - val_loss: 1.4205 - val_acc: 0.4824\n",
            "\n",
            "Epoch 00005: val_acc improved from 0.45120 to 0.48240, saving model to /content/saved_models/cifar10_Alexnet_model.005.h5\n",
            "Epoch 6/25\n",
            "Learning rate:  0.001\n",
            "1563/1563 [==============================] - 35s 22ms/step - loss: 1.4288 - acc: 0.4861 - val_loss: 1.4115 - val_acc: 0.4947\n",
            "\n",
            "Epoch 00006: val_acc improved from 0.48240 to 0.49470, saving model to /content/saved_models/cifar10_Alexnet_model.006.h5\n",
            "Epoch 7/25\n",
            "Learning rate:  0.001\n",
            "1563/1563 [==============================] - 35s 22ms/step - loss: 1.3980 - acc: 0.4978 - val_loss: 1.3864 - val_acc: 0.4967\n",
            "\n",
            "Epoch 00007: val_acc improved from 0.49470 to 0.49670, saving model to /content/saved_models/cifar10_Alexnet_model.007.h5\n",
            "Epoch 8/25\n",
            "Learning rate:  0.001\n",
            "1563/1563 [==============================] - 35s 22ms/step - loss: 1.3752 - acc: 0.5095 - val_loss: 1.3971 - val_acc: 0.4990\n",
            "\n",
            "Epoch 00008: val_acc improved from 0.49670 to 0.49900, saving model to /content/saved_models/cifar10_Alexnet_model.008.h5\n",
            "Epoch 9/25\n",
            "Learning rate:  0.001\n",
            "1563/1563 [==============================] - 35s 22ms/step - loss: 1.3646 - acc: 0.5119 - val_loss: 1.3314 - val_acc: 0.5177\n",
            "\n",
            "Epoch 00009: val_acc improved from 0.49900 to 0.51770, saving model to /content/saved_models/cifar10_Alexnet_model.009.h5\n",
            "Epoch 10/25\n",
            "Learning rate:  0.001\n",
            "1563/1563 [==============================] - 35s 23ms/step - loss: 1.3438 - acc: 0.5203 - val_loss: 1.3243 - val_acc: 0.5259\n",
            "\n",
            "Epoch 00010: val_acc improved from 0.51770 to 0.52590, saving model to /content/saved_models/cifar10_Alexnet_model.010.h5\n",
            "Epoch 11/25\n",
            "Learning rate:  0.001\n",
            "1563/1563 [==============================] - 36s 23ms/step - loss: 1.3347 - acc: 0.5248 - val_loss: 1.3071 - val_acc: 0.5345\n",
            "\n",
            "Epoch 00011: val_acc improved from 0.52590 to 0.53450, saving model to /content/saved_models/cifar10_Alexnet_model.011.h5\n",
            "Epoch 12/25\n",
            "Learning rate:  0.001\n",
            "1563/1563 [==============================] - 35s 22ms/step - loss: 1.3257 - acc: 0.5291 - val_loss: 1.2970 - val_acc: 0.5357\n",
            "\n",
            "Epoch 00012: val_acc improved from 0.53450 to 0.53570, saving model to /content/saved_models/cifar10_Alexnet_model.012.h5\n",
            "Epoch 13/25\n",
            "Learning rate:  0.001\n",
            "1563/1563 [==============================] - 35s 23ms/step - loss: 1.3148 - acc: 0.5349 - val_loss: 1.2879 - val_acc: 0.5405\n",
            "\n",
            "Epoch 00013: val_acc improved from 0.53570 to 0.54050, saving model to /content/saved_models/cifar10_Alexnet_model.013.h5\n",
            "Epoch 14/25\n",
            "Learning rate:  0.001\n",
            "1563/1563 [==============================] - 35s 23ms/step - loss: 1.3086 - acc: 0.5336 - val_loss: 1.2900 - val_acc: 0.5338\n",
            "\n",
            "Epoch 00014: val_acc did not improve from 0.54050\n",
            "Epoch 15/25\n",
            "Learning rate:  0.001\n",
            "1563/1563 [==============================] - 35s 23ms/step - loss: 1.2954 - acc: 0.5390 - val_loss: 1.2864 - val_acc: 0.5401\n",
            "\n",
            "Epoch 00015: val_acc did not improve from 0.54050\n",
            "Epoch 16/25\n",
            "Learning rate:  0.001\n",
            "1563/1563 [==============================] - 35s 23ms/step - loss: 1.2935 - acc: 0.5412 - val_loss: 1.3087 - val_acc: 0.5377\n",
            "\n",
            "Epoch 00016: val_acc did not improve from 0.54050\n",
            "Epoch 17/25\n",
            "Learning rate:  0.001\n",
            "1563/1563 [==============================] - 35s 23ms/step - loss: 1.2844 - acc: 0.5475 - val_loss: 1.2555 - val_acc: 0.5554\n",
            "\n",
            "Epoch 00017: val_acc improved from 0.54050 to 0.55540, saving model to /content/saved_models/cifar10_Alexnet_model.017.h5\n",
            "Epoch 18/25\n",
            "Learning rate:  0.001\n",
            "1563/1563 [==============================] - 35s 23ms/step - loss: 1.2733 - acc: 0.5510 - val_loss: 1.2510 - val_acc: 0.5557\n",
            "\n",
            "Epoch 00018: val_acc improved from 0.55540 to 0.55570, saving model to /content/saved_models/cifar10_Alexnet_model.018.h5\n",
            "Epoch 19/25\n",
            "Learning rate:  0.001\n",
            "1563/1563 [==============================] - 35s 23ms/step - loss: 1.2709 - acc: 0.5504 - val_loss: 1.2733 - val_acc: 0.5479\n",
            "\n",
            "Epoch 00019: val_acc did not improve from 0.55570\n",
            "Epoch 20/25\n",
            "Learning rate:  0.001\n",
            "1563/1563 [==============================] - 36s 23ms/step - loss: 1.2584 - acc: 0.5561 - val_loss: 1.2760 - val_acc: 0.5458\n",
            "\n",
            "Epoch 00020: val_acc did not improve from 0.55570\n",
            "Epoch 21/25\n",
            "Learning rate:  0.001\n",
            "1563/1563 [==============================] - 35s 22ms/step - loss: 1.2604 - acc: 0.5557 - val_loss: 1.2461 - val_acc: 0.5533\n",
            "\n",
            "Epoch 00021: val_acc did not improve from 0.55570\n",
            "Epoch 22/25\n",
            "Learning rate:  0.001\n",
            "1563/1563 [==============================] - 35s 22ms/step - loss: 1.2483 - acc: 0.5611 - val_loss: 1.2210 - val_acc: 0.5641\n",
            "\n",
            "Epoch 00022: val_acc improved from 0.55570 to 0.56410, saving model to /content/saved_models/cifar10_Alexnet_model.022.h5\n",
            "Epoch 23/25\n",
            "Learning rate:  0.001\n",
            "1563/1563 [==============================] - 35s 23ms/step - loss: 1.2502 - acc: 0.5591 - val_loss: 1.2775 - val_acc: 0.5486\n",
            "\n",
            "Epoch 00023: val_acc did not improve from 0.56410\n",
            "Epoch 24/25\n",
            "Learning rate:  0.001\n",
            "1563/1563 [==============================] - 35s 22ms/step - loss: 1.2446 - acc: 0.5605 - val_loss: 1.2226 - val_acc: 0.5598\n",
            "\n",
            "Epoch 00024: val_acc did not improve from 0.56410\n",
            "Epoch 25/25\n",
            "Learning rate:  0.001\n",
            "1563/1563 [==============================] - 35s 23ms/step - loss: 1.2336 - acc: 0.5667 - val_loss: 1.2210 - val_acc: 0.5670\n",
            "\n",
            "Epoch 00025: val_acc improved from 0.56410 to 0.56700, saving model to /content/saved_models/cifar10_Alexnet_model.025.h5\n",
            "10000/10000 [==============================] - 1s 100us/step\n",
            "Test loss: 12.261363667297363\n",
            "Test accuracy: 0.2066\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "T95kj0_iP54_",
        "colab_type": "code",
        "outputId": "4fa00520-075c-464a-99f6-ee5b16f1d061",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 590
        }
      },
      "source": [
        "# list all data in history\n",
        "print(history.history.keys())\n",
        "# summarize history for accuracy\n",
        "plt.plot(history.history['acc'])\n",
        "plt.plot(history.history['val_acc'])\n",
        "plt.title('model accuracy')\n",
        "plt.ylabel('accuracy')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train', 'test'], loc='upper left')\n",
        "plt.show()\n",
        "# summarize history for loss\n",
        "plt.plot(history.history['loss'])\n",
        "plt.plot(history.history['val_loss'])\n",
        "plt.title('model loss')\n",
        "plt.ylabel('loss')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train', 'test'], loc='upper left')\n",
        "plt.show()"
      ],
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "dict_keys(['val_loss', 'val_acc', 'loss', 'acc', 'lr'])\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYwAAAEWCAYAAAB1xKBvAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nDTN9AAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYYAAAEWCAYAAABi5jCmAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    }
  ]
}