6-image-augmentation/0-ImageAugmentation.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "6ImageAugmentation.ipynb",
      "version": "0.3.2",
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "mmRejYvwAQU-",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# author: Asmaa ~ 2019\n",
        "# ----------------------\n",
        "\n",
        "# import libraries\n",
        "import os\n",
        "import zipfile\n",
        "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
        "import tensorflow as tf\n",
        "from tensorflow.keras.layers import Conv2D, Dense, MaxPooling2D, Flatten\n",
        "from tensorflow.keras.models import Sequential\n",
        "from tensorflow.keras.optimizers import RMSprop\n",
        "import matplotlib.pyplot as plt"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "yTiFNDigAHU7",
        "colab_type": "text"
      },
      "source": [
        "# Data Prepaeration"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "zj4_Lsj3AXVk",
        "colab_type": "text"
      },
      "source": [
        "### Downloading Data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "x_-902ne_Q4A",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 391
        },
        "outputId": "485c54fa-81fa-4cdd-8e95-ec20b4f6fc4d"
      },
      "source": [
        "!wget --no-check-certificate \\\n",
        "    https://storage.googleapis.com/laurencemoroney-blog.appspot.com/horse-or-human.zip \\\n",
        "    -O /tmp/horse-or-human.zip\n",
        "\n",
        "!wget --no-check-certificate \\\n",
        "    https://storage.googleapis.com/laurencemoroney-blog.appspot.com/validation-horse-or-human.zip \\\n",
        "    -O /tmp/validation-horse-or-human.zip"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "--2019-08-15 07:43:06--  https://storage.googleapis.com/laurencemoroney-blog.appspot.com/horse-or-human.zip\n",
            "Resolving storage.googleapis.com (storage.googleapis.com)... 66.102.1.128, 2a00:1450:400c:c08::80\n",
            "Connecting to storage.googleapis.com (storage.googleapis.com)|66.102.1.128|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 149574867 (143M) [application/zip]\n",
            "Saving to: ‘/tmp/horse-or-human.zip’\n",
            "\n",
            "/tmp/horse-or-human 100%[===================>] 142.65M  66.6MB/s    in 2.1s    \n",
            "\n",
            "2019-08-15 07:43:09 (66.6 MB/s) - ‘/tmp/horse-or-human.zip’ saved [149574867/149574867]\n",
            "\n",
            "--2019-08-15 07:43:10--  https://storage.googleapis.com/laurencemoroney-blog.appspot.com/validation-horse-or-human.zip\n",
            "Resolving storage.googleapis.com (storage.googleapis.com)... 173.194.76.128, 2a00:1450:400c:c00::80\n",
            "Connecting to storage.googleapis.com (storage.googleapis.com)|173.194.76.128|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 11480187 (11M) [application/zip]\n",
            "Saving to: ‘/tmp/validation-horse-or-human.zip’\n",
            "\n",
            "/tmp/validation-hor 100%[===================>]  10.95M  36.2MB/s    in 0.3s    \n",
            "\n",
            "2019-08-15 07:43:11 (36.2 MB/s) - ‘/tmp/validation-horse-or-human.zip’ saved [11480187/11480187]\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MqnTPgEKAZ9m",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# extracting zip files\n",
        "local_zip = '/tmp/horse-or-human.zip'\n",
        "zip_ref = zipfile.ZipFile(local_zip, 'r')\n",
        "zip_ref.extractall('/tmp/horse-or-human')\n",
        "local_zip = '/tmp/validation-horse-or-human.zip'\n",
        "zip_ref = zipfile.ZipFile(local_zip, 'r')\n",
        "zip_ref.extractall('/tmp/validation-horse-or-human')\n",
        "zip_ref.close()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-J2cL5pVAtNF",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# setting image directories to variables\n",
        "train_horse_dir = os.path.join('/tmp/horse-or-human/horses')\n",
        "\n",
        "train_human_dir = os.path.join('/tmp/horse-or-human/humans')\n",
        "\n",
        "validation_horse_dir = os.path.join('/tmp/validation-horse-or-human/horses')\n",
        "\n",
        "validation_human_dir = os.path.join('/tmp/validation-horse-or-human/humans')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qQEZ6f66BQBz",
        "colab_type": "text"
      },
      "source": [
        "# Pre-processing"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EBWYsuzKA5s4",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# setting scaling and augmentation parameters for trainig set\n",
        "train_datagenerator = ImageDataGenerator( rescale=1./255, rotation_range=40, width_shift_range=0.2,\n",
        "      height_shift_range=0.2, shear_range=0.2, zoom_range=0.2, horizontal_flip=True, fill_mode='nearest')\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "isA9iNxrBhFB",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# rescaling test set\n",
        "val_datagenerator = ImageDataGenerator(rescale=1/255)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3ipJvuOoBufb",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        },
        "outputId": "c6a4d547-1b08-4035-beea-3acdf90e7efa"
      },
      "source": [
        "# flow tarining images from directory and resize images to 300x300 and set classification mode\n",
        "train_generator = train_datagenerator.flow_from_directory(\n",
        "        '/tmp/horse-or-human/', target_size=(300, 300), batch_size=128, class_mode='binary')\n",
        "\n",
        "validation_generator = val_datagenerator.flow_from_directory(\n",
        "        '/tmp/validation-horse-or-human/',  target_size=(300, 300), batch_size=32, class_mode='binary')"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Found 1027 images belonging to 2 classes.\n",
            "Found 256 images belonging to 2 classes.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jN1SwoMpCkvI",
        "colab_type": "text"
      },
      "source": [
        "# Building the Model"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TNKGXNayEa7K",
        "colab_type": "text"
      },
      "source": [
        "### Setting the structure"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "W88wc2W0CKPl",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# initialize the model\n",
        "model = Sequential()\n",
        "\n",
        "# add first convolution \n",
        "# images are 300x300px with 3 color channels\n",
        "model.add(Conv2D(16, (3,3), activation='relu', input_shape=(300, 300, 3)))\n",
        "model.add(MaxPooling2D(2,2))\n",
        "\n",
        "# add second convolution\n",
        "model.add(Conv2D(32, (3,3), activation='relu'))\n",
        "model.add(MaxPooling2D(2,2))\n",
        "\n",
        "# add third convolution\n",
        "model.add(Conv2D(64, (3,3), activation='relu'))\n",
        "model.add(MaxPooling2D(2,2))\n",
        "\n",
        "# add fourth convolution\n",
        "model.add(Conv2D(64, (3,3), activation='relu'))\n",
        "model.add(MaxPooling2D(2,2))\n",
        "\n",
        "# add fifth convolution\n",
        "model.add(Conv2D(64, (3,3), activation='relu'))\n",
        "model.add(MaxPooling2D(2,2))\n",
        "\n",
        "# flatten last output and feed it to DNN\n",
        "model.add(Flatten())\n",
        "\n",
        "# add a hidden layer\n",
        "model.add(Dense(512, activation='relu'))\n",
        "\n",
        "# add output layer, it is a binary classification problem \n",
        "model.add(Dense(1, activation='sigmoid'))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "oXlU4vftEeGK",
        "colab_type": "text"
      },
      "source": [
        "### Compiling the Model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8_51dI8fDBKZ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 88
        },
        "outputId": "f81026d1-59d0-4ba3-d521-c059269d32fe"
      },
      "source": [
        "# compile the model after setting functional properties\n",
        "model.compile(loss='binary_crossentropy', optimizer=RMSprop(lr=1e-4), metrics=['acc'])"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "W0815 07:43:32.834729 140181656041344 deprecation.py:323] From /usr/local/lib/python3.6/dist-packages/tensorflow/python/ops/nn_impl.py:180: add_dispatch_support.<locals>.wrapper (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"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Y6t3czy9EngE",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "c99615da-1a18-478a-d3de-c102b5cc7120"
      },
      "source": [
        "fitting = model.fit_generator( train_generator, steps_per_epoch=8, epochs=100, verbose=1,\n",
        "      validation_data = validation_generator, validation_steps=8)"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.6983 - acc: 0.5150 - val_loss: 0.6766 - val_acc: 0.5000\n",
            "Epoch 2/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.6800 - acc: 0.5506 - val_loss: 0.6836 - val_acc: 0.5000\n",
            "Epoch 3/100\n",
            "8/8 [==============================] - 34s 4s/step - loss: 0.6688 - acc: 0.5996 - val_loss: 0.6436 - val_acc: 0.5977\n",
            "Epoch 4/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.6577 - acc: 0.6274 - val_loss: 0.6103 - val_acc: 0.7148\n",
            "Epoch 5/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.6360 - acc: 0.6674 - val_loss: 0.5950 - val_acc: 0.6914\n",
            "Epoch 6/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.6336 - acc: 0.7186 - val_loss: 0.5721 - val_acc: 0.6992\n",
            "Epoch 7/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.6215 - acc: 0.7075 - val_loss: 0.6111 - val_acc: 0.5781\n",
            "Epoch 8/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.6501 - acc: 0.6785 - val_loss: 0.5869 - val_acc: 0.6172\n",
            "Epoch 9/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.5731 - acc: 0.7742 - val_loss: 0.4807 - val_acc: 0.8047\n",
            "Epoch 10/100\n",
            "8/8 [==============================] - 34s 4s/step - loss: 0.5656 - acc: 0.7324 - val_loss: 0.5268 - val_acc: 0.7031\n",
            "Epoch 11/100\n",
            "8/8 [==============================] - 27s 3s/step - loss: 0.5020 - acc: 0.7506 - val_loss: 0.4840 - val_acc: 0.7461\n",
            "Epoch 12/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.5564 - acc: 0.7164 - val_loss: 0.5285 - val_acc: 0.7070\n",
            "Epoch 13/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.4914 - acc: 0.7653 - val_loss: 0.4154 - val_acc: 0.8125\n",
            "Epoch 14/100\n",
            "8/8 [==============================] - 29s 4s/step - loss: 0.5366 - acc: 0.7253 - val_loss: 0.6449 - val_acc: 0.6406\n",
            "Epoch 15/100\n",
            "8/8 [==============================] - 34s 4s/step - loss: 0.4866 - acc: 0.7812 - val_loss: 0.4650 - val_acc: 0.7734\n",
            "Epoch 16/100\n",
            "8/8 [==============================] - 29s 4s/step - loss: 0.4796 - acc: 0.7353 - val_loss: 0.5680 - val_acc: 0.7227\n",
            "Epoch 17/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.4512 - acc: 0.7820 - val_loss: 0.5697 - val_acc: 0.7227\n",
            "Epoch 18/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.4617 - acc: 0.7831 - val_loss: 0.7287 - val_acc: 0.6602\n",
            "Epoch 19/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.4089 - acc: 0.7875 - val_loss: 1.1353 - val_acc: 0.5820\n",
            "Epoch 20/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.4475 - acc: 0.7953 - val_loss: 0.8031 - val_acc: 0.6797\n",
            "Epoch 21/100\n",
            "8/8 [==============================] - 33s 4s/step - loss: 0.4469 - acc: 0.7900 - val_loss: 1.1005 - val_acc: 0.5977\n",
            "Epoch 22/100\n",
            "8/8 [==============================] - 26s 3s/step - loss: 0.5004 - acc: 0.7687 - val_loss: 0.9708 - val_acc: 0.6250\n",
            "Epoch 23/100\n",
            "8/8 [==============================] - 34s 4s/step - loss: 0.4044 - acc: 0.8320 - val_loss: 1.0375 - val_acc: 0.6328\n",
            "Epoch 24/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.3589 - acc: 0.8498 - val_loss: 1.0406 - val_acc: 0.6562\n",
            "Epoch 25/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.4567 - acc: 0.7731 - val_loss: 0.9346 - val_acc: 0.6836\n",
            "Epoch 26/100\n",
            "8/8 [==============================] - 25s 3s/step - loss: 0.3453 - acc: 0.8295 - val_loss: 0.7795 - val_acc: 0.7422\n",
            "Epoch 27/100\n",
            "8/8 [==============================] - 34s 4s/step - loss: 0.3842 - acc: 0.8252 - val_loss: 1.2616 - val_acc: 0.6250\n",
            "Epoch 28/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.4474 - acc: 0.8198 - val_loss: 1.1525 - val_acc: 0.6562\n",
            "Epoch 29/100\n",
            "8/8 [==============================] - 34s 4s/step - loss: 0.3746 - acc: 0.8350 - val_loss: 1.1677 - val_acc: 0.6484\n",
            "Epoch 30/100\n",
            "8/8 [==============================] - 26s 3s/step - loss: 0.3825 - acc: 0.8140 - val_loss: 0.9279 - val_acc: 0.7031\n",
            "Epoch 31/100\n",
            "8/8 [==============================] - 34s 4s/step - loss: 0.3624 - acc: 0.8447 - val_loss: 1.1542 - val_acc: 0.6641\n",
            "Epoch 32/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.3408 - acc: 0.8543 - val_loss: 1.5895 - val_acc: 0.5820\n",
            "Epoch 33/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.3475 - acc: 0.8309 - val_loss: 1.1497 - val_acc: 0.6641\n",
            "Epoch 34/100\n",
            "8/8 [==============================] - 29s 4s/step - loss: 0.2778 - acc: 0.8699 - val_loss: 1.4714 - val_acc: 0.6289\n",
            "Epoch 35/100\n",
            "8/8 [==============================] - 29s 4s/step - loss: 0.3825 - acc: 0.8231 - val_loss: 1.1184 - val_acc: 0.6758\n",
            "Epoch 36/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.3257 - acc: 0.8576 - val_loss: 1.3454 - val_acc: 0.6523\n",
            "Epoch 37/100\n",
            "8/8 [==============================] - 33s 4s/step - loss: 0.3528 - acc: 0.8457 - val_loss: 1.1830 - val_acc: 0.6562\n",
            "Epoch 38/100\n",
            "8/8 [==============================] - 31s 4s/step - loss: 0.3159 - acc: 0.8699 - val_loss: 1.5487 - val_acc: 0.6289\n",
            "Epoch 39/100\n",
            "8/8 [==============================] - 25s 3s/step - loss: 0.3381 - acc: 0.8501 - val_loss: 1.7136 - val_acc: 0.6406\n",
            "Epoch 40/100\n",
            "8/8 [==============================] - 34s 4s/step - loss: 0.3513 - acc: 0.8486 - val_loss: 1.2834 - val_acc: 0.6406\n",
            "Epoch 41/100\n",
            "8/8 [==============================] - 25s 3s/step - loss: 0.2545 - acc: 0.8773 - val_loss: 1.2080 - val_acc: 0.6758\n",
            "Epoch 42/100\n",
            "8/8 [==============================] - 33s 4s/step - loss: 0.2958 - acc: 0.8701 - val_loss: 2.0754 - val_acc: 0.5898\n",
            "Epoch 43/100\n",
            "8/8 [==============================] - 29s 4s/step - loss: 0.2675 - acc: 0.8843 - val_loss: 0.8204 - val_acc: 0.7266\n",
            "Epoch 44/100\n",
            "8/8 [==============================] - 29s 4s/step - loss: 0.2750 - acc: 0.8754 - val_loss: 1.1127 - val_acc: 0.7148\n",
            "Epoch 45/100\n",
            "8/8 [==============================] - 29s 4s/step - loss: 0.2790 - acc: 0.8754 - val_loss: 1.1755 - val_acc: 0.7148\n",
            "Epoch 46/100\n",
            "8/8 [==============================] - 29s 4s/step - loss: 0.2767 - acc: 0.8765 - val_loss: 1.7212 - val_acc: 0.6094\n",
            "Epoch 47/100\n",
            "8/8 [==============================] - 29s 4s/step - loss: 0.2369 - acc: 0.8932 - val_loss: 1.3710 - val_acc: 0.6719\n",
            "Epoch 48/100\n",
            "8/8 [==============================] - 33s 4s/step - loss: 0.3003 - acc: 0.8848 - val_loss: 1.3033 - val_acc: 0.6797\n",
            "Epoch 49/100\n",
            "8/8 [==============================] - 29s 4s/step - loss: 0.3021 - acc: 0.8721 - val_loss: 1.6522 - val_acc: 0.6484\n",
            "Epoch 50/100\n",
            "8/8 [==============================] - 29s 4s/step - loss: 0.2310 - acc: 0.9010 - val_loss: 2.1034 - val_acc: 0.6133\n",
            "Epoch 51/100\n",
            "8/8 [==============================] - 29s 4s/step - loss: 0.2516 - acc: 0.8843 - val_loss: 1.4030 - val_acc: 0.6719\n",
            "Epoch 52/100\n",
            "8/8 [==============================] - 25s 3s/step - loss: 0.2461 - acc: 0.8669 - val_loss: 2.4327 - val_acc: 0.6055\n",
            "Epoch 53/100\n",
            "8/8 [==============================] - 34s 4s/step - loss: 0.2323 - acc: 0.9141 - val_loss: 2.0122 - val_acc: 0.6406\n",
            "Epoch 54/100\n",
            "8/8 [==============================] - 29s 4s/step - loss: 0.2466 - acc: 0.8921 - val_loss: 1.3740 - val_acc: 0.6914\n",
            "Epoch 55/100\n",
            "8/8 [==============================] - 29s 4s/step - loss: 0.2437 - acc: 0.8877 - val_loss: 1.5956 - val_acc: 0.6719\n",
            "Epoch 56/100\n",
            "8/8 [==============================] - 29s 4s/step - loss: 0.2510 - acc: 0.9255 - val_loss: 0.1802 - val_acc: 0.9336\n",
            "Epoch 57/100\n",
            "8/8 [==============================] - 29s 4s/step - loss: 0.4347 - acc: 0.8532 - val_loss: 2.0109 - val_acc: 0.6328\n",
            "Epoch 58/100\n",
            "8/8 [==============================] - 33s 4s/step - loss: 0.2069 - acc: 0.9141 - val_loss: 2.1071 - val_acc: 0.6328\n",
            "Epoch 59/100\n",
            "8/8 [==============================] - 29s 4s/step - loss: 0.1893 - acc: 0.9266 - val_loss: 2.5602 - val_acc: 0.6172\n",
            "Epoch 60/100\n",
            "8/8 [==============================] - 24s 3s/step - loss: 0.3203 - acc: 0.8734 - val_loss: 1.9233 - val_acc: 0.6172\n",
            "Epoch 61/100\n",
            "8/8 [==============================] - 31s 4s/step - loss: 0.2294 - acc: 0.9032 - val_loss: 2.3812 - val_acc: 0.6172\n",
            "Epoch 62/100\n",
            "8/8 [==============================] - 34s 4s/step - loss: 0.1979 - acc: 0.9248 - val_loss: 2.0474 - val_acc: 0.6445\n",
            "Epoch 63/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.2288 - acc: 0.8988 - val_loss: 2.9821 - val_acc: 0.5547\n",
            "Epoch 64/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.1918 - acc: 0.9188 - val_loss: 2.6001 - val_acc: 0.6172\n",
            "Epoch 65/100\n",
            "8/8 [==============================] - 33s 4s/step - loss: 0.2461 - acc: 0.8926 - val_loss: 2.5051 - val_acc: 0.6055\n",
            "Epoch 66/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.1611 - acc: 0.9366 - val_loss: 2.3387 - val_acc: 0.6367\n",
            "Epoch 67/100\n",
            "8/8 [==============================] - 26s 3s/step - loss: 0.2912 - acc: 0.8863 - val_loss: 0.9264 - val_acc: 0.7539\n",
            "Epoch 68/100\n",
            "8/8 [==============================] - 34s 4s/step - loss: 0.2028 - acc: 0.9229 - val_loss: 2.0178 - val_acc: 0.6680\n",
            "Epoch 69/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.2008 - acc: 0.9121 - val_loss: 2.1749 - val_acc: 0.6484\n",
            "Epoch 70/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.1941 - acc: 0.9121 - val_loss: 3.1189 - val_acc: 0.5469\n",
            "Epoch 71/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.1666 - acc: 0.9333 - val_loss: 2.3321 - val_acc: 0.6445\n",
            "Epoch 72/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.1752 - acc: 0.9143 - val_loss: 0.8197 - val_acc: 0.7695\n",
            "Epoch 73/100\n",
            "8/8 [==============================] - 33s 4s/step - loss: 0.1998 - acc: 0.9180 - val_loss: 2.2436 - val_acc: 0.6445\n",
            "Epoch 74/100\n",
            "8/8 [==============================] - 31s 4s/step - loss: 0.2844 - acc: 0.9043 - val_loss: 2.2537 - val_acc: 0.6406\n",
            "Epoch 75/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.1751 - acc: 0.9422 - val_loss: 2.4086 - val_acc: 0.6523\n",
            "Epoch 76/100\n",
            "8/8 [==============================] - 26s 3s/step - loss: 0.1721 - acc: 0.9238 - val_loss: 1.2011 - val_acc: 0.7227\n",
            "Epoch 77/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.1612 - acc: 0.9399 - val_loss: 2.3848 - val_acc: 0.6445\n",
            "Epoch 78/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.1936 - acc: 0.9143 - val_loss: 2.7548 - val_acc: 0.6211\n",
            "Epoch 79/100\n",
            "8/8 [==============================] - 34s 4s/step - loss: 0.1646 - acc: 0.9365 - val_loss: 3.1867 - val_acc: 0.5938\n",
            "Epoch 80/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.1912 - acc: 0.9199 - val_loss: 1.9004 - val_acc: 0.6719\n",
            "Epoch 81/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.1176 - acc: 0.9611 - val_loss: 1.9241 - val_acc: 0.6836\n",
            "Epoch 82/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.2233 - acc: 0.9266 - val_loss: 2.6071 - val_acc: 0.6289\n",
            "Epoch 83/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.1553 - acc: 0.9388 - val_loss: 4.6904 - val_acc: 0.5391\n",
            "Epoch 84/100\n",
            "8/8 [==============================] - 34s 4s/step - loss: 0.1702 - acc: 0.9307 - val_loss: 2.7973 - val_acc: 0.6211\n",
            "Epoch 85/100\n",
            "8/8 [==============================] - 26s 3s/step - loss: 0.1277 - acc: 0.9328 - val_loss: 1.1174 - val_acc: 0.7578\n",
            "Epoch 86/100\n",
            "8/8 [==============================] - 34s 4s/step - loss: 0.1871 - acc: 0.9160 - val_loss: 2.7663 - val_acc: 0.6250\n",
            "Epoch 87/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.1379 - acc: 0.9444 - val_loss: 4.4805 - val_acc: 0.5352\n",
            "Epoch 88/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.1833 - acc: 0.9277 - val_loss: 2.3220 - val_acc: 0.6484\n",
            "Epoch 89/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.1579 - acc: 0.9410 - val_loss: 2.3672 - val_acc: 0.6562\n",
            "Epoch 90/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.1536 - acc: 0.9333 - val_loss: 1.6966 - val_acc: 0.6875\n",
            "Epoch 91/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.1373 - acc: 0.9388 - val_loss: 2.7687 - val_acc: 0.6367\n",
            "Epoch 92/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.1288 - acc: 0.9444 - val_loss: 5.4026 - val_acc: 0.5430\n",
            "Epoch 93/100\n",
            "8/8 [==============================] - 34s 4s/step - loss: 0.2157 - acc: 0.9160 - val_loss: 2.8899 - val_acc: 0.6367\n",
            "Epoch 94/100\n",
            "8/8 [==============================] - 30s 4s/step - loss: 0.1307 - acc: 0.9444 - val_loss: 2.9055 - val_acc: 0.6367\n",
            "Epoch 95/100\n",
            "8/8 [==============================] - 29s 4s/step - loss: 0.1140 - acc: 0.9511 - val_loss: 3.4474 - val_acc: 0.6055\n",
            "Epoch 96/100\n",
            "8/8 [==============================] - 29s 4s/step - loss: 0.1587 - acc: 0.9221 - val_loss: 2.7509 - val_acc: 0.6406\n",
            "Epoch 97/100\n",
            "8/8 [==============================] - 29s 4s/step - loss: 0.0929 - acc: 0.9588 - val_loss: 3.3346 - val_acc: 0.6289\n",
            "Epoch 98/100\n",
            "8/8 [==============================] - 29s 4s/step - loss: 0.1279 - acc: 0.9499 - val_loss: 3.7029 - val_acc: 0.5938\n",
            "Epoch 99/100\n",
            "8/8 [==============================] - 25s 3s/step - loss: 0.1875 - acc: 0.9488 - val_loss: 4.9262 - val_acc: 0.5664\n",
            "Epoch 100/100\n",
            "8/8 [==============================] - 29s 4s/step - loss: 0.1111 - acc: 0.9511 - val_loss: 3.0126 - val_acc: 0.6445\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-vy9WOPaTSX4",
        "colab_type": "text"
      },
      "source": [
        "# Metrics Visualization"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2Em_wdc7E4jL",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 545
        },
        "outputId": "772afd9d-50a3-4540-cf78-e8e7c0b41924"
      },
      "source": [
        "# get metrics from the model\n",
        "acc = fitting.history['acc']\n",
        "val_acc = fitting.history['val_acc']\n",
        "loss = fitting.history['loss']\n",
        "val_loss = fitting.history['val_loss']\n",
        "\n",
        "epochs = range(len(acc))\n",
        "\n",
        "# plot accuracy\n",
        "plt.plot(epochs, acc, 'r', label='Training accuracy')\n",
        "plt.plot(epochs, val_acc, 'b', label='Validation accuracy')\n",
        "plt.title('Training and validation accuracy')\n",
        "plt.legend()\n",
        "\n",
        "plt.figure()\n",
        "\n",
        "# plot loss\n",
        "plt.plot(epochs, loss, 'r', label='Training Loss')\n",
        "plt.plot(epochs, val_loss, 'b', label='Validation Loss')\n",
        "plt.title('Training and validation loss')\n",
        "plt.legend()\n",
        "\n",
        "plt.show()"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAEICAYAAACktLTqAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAW4AAAEICAYAAAB/Dx7IAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    }
  ]
}