Transfer Learning.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "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": "3.7.1"
    },
    "colab": {
      "name": "Transfer Learning.ipynb",
      "provenance": []
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "dpXznQQ8qhIY",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "34961549-7366-4bac-a629-b2084f53d2f6"
      },
      "source": [
        "import tensorflow as tf\n",
        "print(tf.__version__)"
      ],
      "execution_count": 93,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "2.2.0\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qguXIo9_r_cf",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "from skimage.transform import resize\n",
        "import matplotlib.pyplot as plt\n",
        "import pickle\n",
        "\n",
        "from keras.datasets import cifar10\n",
        "from keras.utils import to_categorical\n",
        "from sklearn.model_selection import train_test_split\n",
        "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
        "from keras.engine import Model\n",
        "from keras.applications import vgg16 as vgg\n",
        "from keras.layers import Dropout, Flatten, Dense, GlobalAveragePooling2D,BatchNormalization\n",
        "from keras import Sequential\n",
        "from keras.optimizers import Adam\n",
        "from tensorflow.keras.models import load_model"
      ],
      "execution_count": 94,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Lt-cceTzOQvf",
        "colab_type": "text"
      },
      "source": [
        "## Loading the dataset"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WDDmz99UImu3",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 104
        },
        "outputId": "e520afe7-2994-48cb-8478-cf7cea3015a3"
      },
      "source": [
        "#load dataset\n",
        "(x_train, y_train), (x_test, y_test) = cifar10.load_data()\n",
        "x_train, x_val, y_train, y_val = train_test_split(x_train, y_train, test_size=0.15, stratify = np.array(y_train), random_state=42)\n",
        "#one hot encoding target variables\n",
        "y_train = to_categorical(y_train)\n",
        "y_val = to_categorical(y_val)\n",
        "y_test = to_categorical(y_test)\n",
        "print('x_train: ', x_train.shape, 'y_train: ', y_train.shape)\n",
        "print('x_val: ', x_val.shape, 'y_val: ', y_val.shape)\n",
        "print('x_test: ', x_test.shape, 'y_test: ', y_test.shape) "
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Downloading data from https://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz\n",
            "170500096/170498071 [==============================] - 2s 0us/step\n",
            "x_train:  (42500, 32, 32, 3) y_train:  (42500, 10)\n",
            "x_val:  (7500, 32, 32, 3) y_val:  (7500, 10)\n",
            "x_test:  (10000, 32, 32, 3) y_test:  (10000, 10)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "D9tjO8BT4YDM",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 373
        },
        "outputId": "7539cdbd-c86e-4679-9b49-c6cd3bd6cae2"
      },
      "source": [
        "# plot first few images\n",
        "fig = plt.figure(figsize=(12,7))\n",
        "for i in range(10):\n",
        "  # define subplot\n",
        "\tplt.subplot(2, 5, i+1)\n",
        "\t# plot raw pixel data\n",
        "\tplt.imshow(x_train[i])\n",
        "# show the figure\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 864x504 with 10 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "24EHhLZ-iXs2",
        "colab_type": "text"
      },
      "source": [
        "## Preprocessing data"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pp9ufTHXhPyo",
        "colab_type": "text"
      },
      "source": [
        "VGG16 takes input images having minimum of 48 pixels of height and width. Therefore resizing the images"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "L5hZiqdcbMkZ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 69
        },
        "outputId": "68416c51-d449-4fd7-cda9-3a78078d46c2"
      },
      "source": [
        "x_train = np.array([resize(x, (48, 48)) for x in x_train])\n",
        "x_val = np.array([resize(x,(48, 48)) for x in x_val])\n",
        "x_test = np.array([resize(x,(48, 48)) for x in x_test])\n",
        "print('x_train: ', x_train.shape, 'y_train: ', y_train.shape)\n",
        "print('x_val: ', x_val.shape, 'y_val: ', y_val.shape)\n",
        "print('x_test: ', x_test.shape, 'y_test: ', y_test.shape)"
      ],
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "x_train:  (42500, 48, 48, 3) y_train:  (42500, 10)\n",
            "x_val:  (7500, 48, 48, 3) y_val:  (7500, 10)\n",
            "x_test:  (10000, 48, 48, 3) y_test:  (10000, 10)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "eeAfeZ0-itpF",
        "colab_type": "text"
      },
      "source": [
        "## Preparing Model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tiHUU_L2kKoj",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "batch_size= 32\n",
        "epochs= 40\n",
        "nb_classes= 10\n",
        "learning_rate= 0.0001"
      ],
      "execution_count": 45,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0ojD7ll3fENC",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "base_model = vgg.VGG16(weights= 'imagenet', include_top= False, input_shape=(48,48,3))"
      ],
      "execution_count": 46,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pGJhQu-8f2il",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 815
        },
        "outputId": "e7cdeb27-b1f5-4a75-a4ad-38808e244213"
      },
      "source": [
        "base_model.summary()"
      ],
      "execution_count": 47,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"vgg16\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "input_2 (InputLayer)         (None, 48, 48, 3)         0         \n",
            "_________________________________________________________________\n",
            "block1_conv1 (Conv2D)        (None, 48, 48, 64)        1792      \n",
            "_________________________________________________________________\n",
            "block1_conv2 (Conv2D)        (None, 48, 48, 64)        36928     \n",
            "_________________________________________________________________\n",
            "block1_pool (MaxPooling2D)   (None, 24, 24, 64)        0         \n",
            "_________________________________________________________________\n",
            "block2_conv1 (Conv2D)        (None, 24, 24, 128)       73856     \n",
            "_________________________________________________________________\n",
            "block2_conv2 (Conv2D)        (None, 24, 24, 128)       147584    \n",
            "_________________________________________________________________\n",
            "block2_pool (MaxPooling2D)   (None, 12, 12, 128)       0         \n",
            "_________________________________________________________________\n",
            "block3_conv1 (Conv2D)        (None, 12, 12, 256)       295168    \n",
            "_________________________________________________________________\n",
            "block3_conv2 (Conv2D)        (None, 12, 12, 256)       590080    \n",
            "_________________________________________________________________\n",
            "block3_conv3 (Conv2D)        (None, 12, 12, 256)       590080    \n",
            "_________________________________________________________________\n",
            "block3_pool (MaxPooling2D)   (None, 6, 6, 256)         0         \n",
            "_________________________________________________________________\n",
            "block4_conv1 (Conv2D)        (None, 6, 6, 512)         1180160   \n",
            "_________________________________________________________________\n",
            "block4_conv2 (Conv2D)        (None, 6, 6, 512)         2359808   \n",
            "_________________________________________________________________\n",
            "block4_conv3 (Conv2D)        (None, 6, 6, 512)         2359808   \n",
            "_________________________________________________________________\n",
            "block4_pool (MaxPooling2D)   (None, 3, 3, 512)         0         \n",
            "_________________________________________________________________\n",
            "block5_conv1 (Conv2D)        (None, 3, 3, 512)         2359808   \n",
            "_________________________________________________________________\n",
            "block5_conv2 (Conv2D)        (None, 3, 3, 512)         2359808   \n",
            "_________________________________________________________________\n",
            "block5_conv3 (Conv2D)        (None, 3, 3, 512)         2359808   \n",
            "_________________________________________________________________\n",
            "block5_pool (MaxPooling2D)   (None, 1, 1, 512)         0         \n",
            "=================================================================\n",
            "Total params: 14,714,688\n",
            "Trainable params: 14,714,688\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YriNnqqJgpyz",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Extract the last layer from third block of vgg16 model\n",
        "last = base_model.get_layer('block3_pool').output"
      ],
      "execution_count": 48,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8HQvV-SbogjX",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "x = GlobalAveragePooling2D()(last)\n",
        "x= BatchNormalization()(x)\n",
        "x = Dense(256, activation='relu')(x)\n",
        "x = Dense(256, activation='relu')(x)\n",
        "x = Dropout(0.6)(x)\n",
        "pred = Dense(nb_classes, activation='softmax')(x)\n",
        "model = Model(base_model.input, pred)"
      ],
      "execution_count": 49,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SJcFrX795k0i",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "for layer in base_model.layers:\n",
        "     layer.trainable = False"
      ],
      "execution_count": 50,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GQlI6Z2Kpy0w",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model.compile(loss='categorical_crossentropy', optimizer=Adam(learning_rate= learning_rate), metrics=['accuracy'])"
      ],
      "execution_count": 51,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "21adg0KnvmgA",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 745
        },
        "outputId": "cc7ee3bc-231f-47d8-b1ab-edf5ed27dc19"
      },
      "source": [
        "model.summary()"
      ],
      "execution_count": 52,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"model_2\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "input_2 (InputLayer)         (None, 48, 48, 3)         0         \n",
            "_________________________________________________________________\n",
            "block1_conv1 (Conv2D)        (None, 48, 48, 64)        1792      \n",
            "_________________________________________________________________\n",
            "block1_conv2 (Conv2D)        (None, 48, 48, 64)        36928     \n",
            "_________________________________________________________________\n",
            "block1_pool (MaxPooling2D)   (None, 24, 24, 64)        0         \n",
            "_________________________________________________________________\n",
            "block2_conv1 (Conv2D)        (None, 24, 24, 128)       73856     \n",
            "_________________________________________________________________\n",
            "block2_conv2 (Conv2D)        (None, 24, 24, 128)       147584    \n",
            "_________________________________________________________________\n",
            "block2_pool (MaxPooling2D)   (None, 12, 12, 128)       0         \n",
            "_________________________________________________________________\n",
            "block3_conv1 (Conv2D)        (None, 12, 12, 256)       295168    \n",
            "_________________________________________________________________\n",
            "block3_conv2 (Conv2D)        (None, 12, 12, 256)       590080    \n",
            "_________________________________________________________________\n",
            "block3_conv3 (Conv2D)        (None, 12, 12, 256)       590080    \n",
            "_________________________________________________________________\n",
            "block3_pool (MaxPooling2D)   (None, 6, 6, 256)         0         \n",
            "_________________________________________________________________\n",
            "global_average_pooling2d_2 ( (None, 256)               0         \n",
            "_________________________________________________________________\n",
            "batch_normalization_2 (Batch (None, 256)               1024      \n",
            "_________________________________________________________________\n",
            "dense_4 (Dense)              (None, 256)               65792     \n",
            "_________________________________________________________________\n",
            "dense_5 (Dense)              (None, 256)               65792     \n",
            "_________________________________________________________________\n",
            "dropout_2 (Dropout)          (None, 256)               0         \n",
            "_________________________________________________________________\n",
            "dense_6 (Dense)              (None, 10)                2570      \n",
            "=================================================================\n",
            "Total params: 1,870,666\n",
            "Trainable params: 134,666\n",
            "Non-trainable params: 1,736,000\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "FpO52pdCx7UF",
        "colab_type": "text"
      },
      "source": [
        "## Data Augmentation"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "I7yyOzv2w5JH",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "train_data_gen = ImageDataGenerator(rescale = 1./255)\n",
        "train_data_gen.fit(x_train)\n",
        "train_generator = train_data_gen.flow(x_train, y_train, batch_size=batch_size)"
      ],
      "execution_count": 53,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eYRjPLvAZUM8",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "validation_data_gen = ImageDataGenerator(rescale = 1./255)\n",
        "validation_data_gen.fit(x_val)\n",
        "validation_generator = validation_data_gen.flow(x_val, y_val, batch_size=batch_size)"
      ],
      "execution_count": 54,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fPbcWfuZCwaX",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "test_batch_size=100\n",
        "test_data_gen = ImageDataGenerator(rescale = 1./255)\n",
        "test_data_gen.fit(x_test)\n",
        "test_generator = test_data_gen.flow(x_test, y_test, batch_size=test_batch_size)"
      ],
      "execution_count": 66,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "aVYFQPcnzQp2",
        "colab_type": "text"
      },
      "source": [
        "## Training model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CBk7uGn8yrIz",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "97d8a469-853b-47c0-ef21-5b1db2985560"
      },
      "source": [
        "\n",
        "train_steps_per_epoch = x_train.shape[0] // batch_size\n",
        "val_steps_per_epoch = x_val.shape[0] // batch_size\n",
        "\n",
        "history = model.fit_generator(train_generator,\n",
        "                              steps_per_epoch=train_steps_per_epoch,\n",
        "                              validation_data=validation_generator,\n",
        "                              validation_steps=val_steps_per_epoch,\n",
        "                              epochs=epochs,\n",
        "                              verbose=1)"
      ],
      "execution_count": 55,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/40\n",
            "1328/1328 [==============================] - 29s 22ms/step - loss: 1.8496 - accuracy: 0.3339 - val_loss: 1.6692 - val_accuracy: 0.4774\n",
            "Epoch 2/40\n",
            "1328/1328 [==============================] - 28s 21ms/step - loss: 1.5176 - accuracy: 0.4600 - val_loss: 1.3690 - val_accuracy: 0.5276\n",
            "Epoch 3/40\n",
            "1328/1328 [==============================] - 28s 21ms/step - loss: 1.4045 - accuracy: 0.5022 - val_loss: 1.3322 - val_accuracy: 0.5541\n",
            "Epoch 4/40\n",
            "1328/1328 [==============================] - 28s 21ms/step - loss: 1.3322 - accuracy: 0.5295 - val_loss: 1.3999 - val_accuracy: 0.5778\n",
            "Epoch 5/40\n",
            "1328/1328 [==============================] - 29s 21ms/step - loss: 1.2853 - accuracy: 0.5443 - val_loss: 1.3872 - val_accuracy: 0.5848\n",
            "Epoch 6/40\n",
            "1328/1328 [==============================] - 28s 21ms/step - loss: 1.2491 - accuracy: 0.5577 - val_loss: 1.1652 - val_accuracy: 0.5992\n",
            "Epoch 7/40\n",
            "1328/1328 [==============================] - 28s 21ms/step - loss: 1.2123 - accuracy: 0.5750 - val_loss: 1.0299 - val_accuracy: 0.6051\n",
            "Epoch 8/40\n",
            "1328/1328 [==============================] - 28s 21ms/step - loss: 1.1883 - accuracy: 0.5803 - val_loss: 1.0578 - val_accuracy: 0.6158\n",
            "Epoch 9/40\n",
            "1328/1328 [==============================] - 29s 21ms/step - loss: 1.1623 - accuracy: 0.5894 - val_loss: 1.1272 - val_accuracy: 0.6217\n",
            "Epoch 10/40\n",
            "1328/1328 [==============================] - 29s 22ms/step - loss: 1.1458 - accuracy: 0.5960 - val_loss: 1.0074 - val_accuracy: 0.6251\n",
            "Epoch 11/40\n",
            "1328/1328 [==============================] - 29s 21ms/step - loss: 1.1249 - accuracy: 0.6042 - val_loss: 1.1840 - val_accuracy: 0.6268\n",
            "Epoch 12/40\n",
            "1328/1328 [==============================] - 28s 21ms/step - loss: 1.1107 - accuracy: 0.6057 - val_loss: 0.9298 - val_accuracy: 0.6320\n",
            "Epoch 13/40\n",
            "1328/1328 [==============================] - 28s 21ms/step - loss: 1.0982 - accuracy: 0.6153 - val_loss: 1.0696 - val_accuracy: 0.6381\n",
            "Epoch 14/40\n",
            "1328/1328 [==============================] - 28s 21ms/step - loss: 1.0792 - accuracy: 0.6195 - val_loss: 0.9626 - val_accuracy: 0.6375\n",
            "Epoch 15/40\n",
            "1328/1328 [==============================] - 28s 21ms/step - loss: 1.0690 - accuracy: 0.6238 - val_loss: 0.7974 - val_accuracy: 0.6390\n",
            "Epoch 16/40\n",
            "1328/1328 [==============================] - 28s 21ms/step - loss: 1.0568 - accuracy: 0.6299 - val_loss: 0.8387 - val_accuracy: 0.6381\n",
            "Epoch 17/40\n",
            "1328/1328 [==============================] - 28s 21ms/step - loss: 1.0434 - accuracy: 0.6312 - val_loss: 1.1771 - val_accuracy: 0.6484\n",
            "Epoch 18/40\n",
            "1328/1328 [==============================] - 28s 21ms/step - loss: 1.0349 - accuracy: 0.6345 - val_loss: 1.0244 - val_accuracy: 0.6457\n",
            "Epoch 19/40\n",
            "1328/1328 [==============================] - 28s 21ms/step - loss: 1.0228 - accuracy: 0.6413 - val_loss: 0.9086 - val_accuracy: 0.6486\n",
            "Epoch 20/40\n",
            "1328/1328 [==============================] - 29s 21ms/step - loss: 1.0113 - accuracy: 0.6437 - val_loss: 1.2320 - val_accuracy: 0.6571\n",
            "Epoch 21/40\n",
            "1328/1328 [==============================] - 29s 22ms/step - loss: 1.0078 - accuracy: 0.6466 - val_loss: 0.9177 - val_accuracy: 0.6539\n",
            "Epoch 22/40\n",
            "1328/1328 [==============================] - 28s 21ms/step - loss: 0.9954 - accuracy: 0.6510 - val_loss: 0.8345 - val_accuracy: 0.6577\n",
            "Epoch 23/40\n",
            "1328/1328 [==============================] - 28s 21ms/step - loss: 0.9911 - accuracy: 0.6493 - val_loss: 0.8051 - val_accuracy: 0.6610\n",
            "Epoch 24/40\n",
            "1328/1328 [==============================] - 28s 21ms/step - loss: 0.9816 - accuracy: 0.6557 - val_loss: 0.6638 - val_accuracy: 0.6591\n",
            "Epoch 25/40\n",
            "1328/1328 [==============================] - 28s 21ms/step - loss: 0.9696 - accuracy: 0.6601 - val_loss: 0.8616 - val_accuracy: 0.6642\n",
            "Epoch 26/40\n",
            "1328/1328 [==============================] - 28s 21ms/step - loss: 0.9677 - accuracy: 0.6615 - val_loss: 0.9878 - val_accuracy: 0.6606\n",
            "Epoch 27/40\n",
            "1328/1328 [==============================] - 28s 21ms/step - loss: 0.9608 - accuracy: 0.6630 - val_loss: 0.9251 - val_accuracy: 0.6583\n",
            "Epoch 28/40\n",
            "1328/1328 [==============================] - 29s 21ms/step - loss: 0.9532 - accuracy: 0.6646 - val_loss: 0.8594 - val_accuracy: 0.6683\n",
            "Epoch 29/40\n",
            "1328/1328 [==============================] - 29s 22ms/step - loss: 0.9461 - accuracy: 0.6695 - val_loss: 1.0583 - val_accuracy: 0.6639\n",
            "Epoch 30/40\n",
            "1328/1328 [==============================] - 29s 22ms/step - loss: 0.9394 - accuracy: 0.6691 - val_loss: 0.9864 - val_accuracy: 0.6635\n",
            "Epoch 31/40\n",
            "1328/1328 [==============================] - 29s 22ms/step - loss: 0.9302 - accuracy: 0.6733 - val_loss: 1.0564 - val_accuracy: 0.6647\n",
            "Epoch 32/40\n",
            "1328/1328 [==============================] - 29s 22ms/step - loss: 0.9256 - accuracy: 0.6735 - val_loss: 0.8121 - val_accuracy: 0.6656\n",
            "Epoch 33/40\n",
            "1328/1328 [==============================] - 28s 21ms/step - loss: 0.9216 - accuracy: 0.6767 - val_loss: 0.8054 - val_accuracy: 0.6730\n",
            "Epoch 34/40\n",
            "1328/1328 [==============================] - 28s 21ms/step - loss: 0.9104 - accuracy: 0.6801 - val_loss: 0.8620 - val_accuracy: 0.6693\n",
            "Epoch 35/40\n",
            "1328/1328 [==============================] - 28s 21ms/step - loss: 0.9058 - accuracy: 0.6827 - val_loss: 0.7987 - val_accuracy: 0.6738\n",
            "Epoch 36/40\n",
            "1328/1328 [==============================] - 28s 21ms/step - loss: 0.8962 - accuracy: 0.6844 - val_loss: 0.8971 - val_accuracy: 0.6742\n",
            "Epoch 37/40\n",
            "1328/1328 [==============================] - 28s 21ms/step - loss: 0.8974 - accuracy: 0.6856 - val_loss: 0.9158 - val_accuracy: 0.6710\n",
            "Epoch 38/40\n",
            "1328/1328 [==============================] - 28s 21ms/step - loss: 0.8894 - accuracy: 0.6858 - val_loss: 0.7463 - val_accuracy: 0.6757\n",
            "Epoch 39/40\n",
            "1328/1328 [==============================] - 28s 21ms/step - loss: 0.8773 - accuracy: 0.6919 - val_loss: 1.0770 - val_accuracy: 0.6731\n",
            "Epoch 40/40\n",
            "1328/1328 [==============================] - 28s 21ms/step - loss: 0.8764 - accuracy: 0.6910 - val_loss: 0.7537 - val_accuracy: 0.6747\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zFnX9KAzCa3L",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def save_history(history, fn):\n",
        "    with open(fn, 'wb') as fw:\n",
        "        pickle.dump(history.history, fw, protocol=2)\n",
        "\n",
        "save_history(history, 'history_augmented_1.bin') # categorical_crossentropy\n",
        "model.save('pre_trained_vgg16_1.h5')"
      ],
      "execution_count": 95,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QjIdO_q4s-Fh",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# load model\n",
        "model = load_model('/content/drive/My Drive/00_transfer_learning/pre_trained_vgg16_1.h5')"
      ],
      "execution_count": 96,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "AJ1CNYn-ykFd",
        "colab_type": "text"
      },
      "source": [
        "## Evaluate Model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bEWXrtMw24Re",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "7af08691-5ba1-48fe-a5c3-a3ed4d9852b3"
      },
      "source": [
        "total_validation = 7500\n",
        "score = model.evaluate_generator(generator=validation_generator, steps=total_validation // batch_size)\n",
        "score"
      ],
      "execution_count": 57,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[1.1409269571304321, 0.6746794581413269]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 57
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GAC84TlDzwFm",
        "colab_type": "text"
      },
      "source": [
        "## Model Performance"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1-xtnk66vHVc",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def load_history(fn):\n",
        "    class Temp():\n",
        "        pass\n",
        "    history = Temp()\n",
        "    with open(fn, 'rb') as fr:\n",
        "        history.history = pickle.load(fr)\n",
        "    return history"
      ],
      "execution_count": 23,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5C4qYgzFvQ0S",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "history = load_history('/content/drive/My Drive/00_transfer_learning/history_augmented.bin')"
      ],
      "execution_count": 37,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jE9ndiQy9AsY",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 281
        },
        "outputId": "a603ae60-283d-4428-d9ee-c3ae85dcbe6c"
      },
      "source": [
        "acc = history.history['accuracy']\n",
        "val_acc = history.history['val_accuracy']\n",
        "\n",
        "loss=history.history['loss']\n",
        "val_loss=history.history['val_loss']\n",
        "\n",
        "epochs_range = range(epochs)\n",
        "\n",
        "plt.figure(figsize=(12, 4))\n",
        "plt.subplot(1, 2, 1)\n",
        "plt.plot(epochs_range, acc, label='Training Accuracy')\n",
        "plt.plot(epochs_range, val_acc, label='Validation Accuracy')\n",
        "plt.legend(loc='lower right')\n",
        "plt.title('Training and Validation Accuracy')\n",
        "\n",
        "plt.subplot(1, 2, 2)\n",
        "plt.plot(epochs_range, loss, label='Training Loss')\n",
        "plt.plot(epochs_range, val_loss, label='Validation Loss')\n",
        "plt.legend(loc='upper right')\n",
        "plt.title('Training and Validation Loss')\n",
        "plt.show()"
      ],
      "execution_count": 60,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 864x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Yr9hMaYi1XUm",
        "colab_type": "text"
      },
      "source": [
        "## Predictions"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wzIOPp7w1mtk",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "label_dict = {0:'airplane',\n",
        "             1:'automobile',\n",
        "             2:'bird',\n",
        "             3:'cat',\n",
        "             4:'deer',\n",
        "             5:'dog',\n",
        "             6:'frog',\n",
        "             7:'horse',\n",
        "             8:'ship',\n",
        "             9:'truck'}"
      ],
      "execution_count": 103,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oo5a_IwsCbyu",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 107
        },
        "outputId": "eeeed8f0-0a03-4239-9c0b-1a8d0e7d94a0"
      },
      "source": [
        "test_generator.reset()\n",
        "total_test = 10000\n",
        "test_batch_size = 100\n",
        "prediction=model.predict_generator(test_generator,\n",
        "                                   steps=total_test // test_batch_size,\n",
        "                                   verbose=1)"
      ],
      "execution_count": 104,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:From <ipython-input-104-12ae9454e57c>:6: Model.predict_generator (from tensorflow.python.keras.engine.training) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Please use Model.predict, which supports generators.\n",
            "100/100 [==============================] - 4s 43ms/step\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YUKdiVy1EsRh",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "test_class_indices=np.argmax(y_test,axis=1)\n",
        "labels = label_dict\n",
        "test_labels = [labels[k] for k in test_class_indices]"
      ],
      "execution_count": 105,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "at-kigFaDFOl",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "predicted_class_indices=np.argmax(prediction,axis=1)\n",
        "labels = label_dict\n",
        "predictions = [labels[k] for k in predicted_class_indices]"
      ],
      "execution_count": 106,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ApjnGpuuDPx8",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "results=pd.DataFrame({\"Test Labels\":test_labels,\n",
        "                      \"Predictions\":predictions})\n",
        "results.to_csv(\"results_with_augmentation.csv\",index=False)"
      ],
      "execution_count": 115,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SSChhCl8cucq",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "82171260-1fc3-456c-846f-6bd14611cb2a"
      },
      "source": [
        "for img in range(5):\n",
        "  img_resized = resize(x_test[img],(48, 48),3)\n",
        "  plt.imshow(img_resized)\n",
        "  plt.show()"
      ],
      "execution_count": 111,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zJFKZCgzdc9p",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}