Fashion_dataset_mnist_train-up.ipynb

{
  "cells": [
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "5EcOBgC2iorW"
      },
      "outputs": [],
      "source": [
        "# import packages that are needed\n",
        "\n",
        "import tensorflow as tf\n",
        "from tensorflow import keras\n",
        "from tensorflow.keras.datasets import fashion_mnist\n",
        "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from tensorflow.keras.models import Sequential\n",
        "from tensorflow.keras.layers import Dense, Flatten\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "-undeM8DoOyV",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "86b81a46-193d-45ae-838f-cfdd9b683fc3"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/train-labels-idx1-ubyte.gz\n",
            "29515/29515 [==============================] - 0s 0us/step\n",
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/train-images-idx3-ubyte.gz\n",
            "26421880/26421880 [==============================] - 0s 0us/step\n",
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/t10k-labels-idx1-ubyte.gz\n",
            "5148/5148 [==============================] - 0s 0us/step\n",
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/t10k-images-idx3-ubyte.gz\n",
            "4422102/4422102 [==============================] - 0s 0us/step\n"
          ]
        }
      ],
      "source": [
        "# load Fashion-MNIST dataset\n",
        "(x_train, y_train), (x_test, y_test) = fashion_mnist.load_data()\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "2K7g5M2CoMYI"
      },
      "outputs": [],
      "source": [
        "# Reshape loaded model\n",
        "x_train = x_train.reshape((60000, 28, 28, 1)) / 255.0\n",
        "x_test = x_test.reshape((10000, 28, 28, 1)) / 255.0\n",
        "\n",
        "y_train = keras.utils.to_categorical(y_train, 10)\n",
        "y_test = keras.utils.to_categorical(y_test, 10)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "6-QWHZxYoKW3"
      },
      "outputs": [],
      "source": [
        "# use imageData generator for augmentation\n",
        "datagen = ImageDataGenerator(\n",
        "    rotation_range=20,\n",
        "    width_shift_range=0.1,\n",
        "    height_shift_range=0.1,\n",
        "    shear_range=0.1,\n",
        "    zoom_range=0.1,\n",
        "    fill_mode='nearest'\n",
        ")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "yU3OUoB0oE8B"
      },
      "outputs": [],
      "source": [
        "# create model structure\n",
        "model = Sequential([\n",
        "    Flatten(input_shape=(28, 28)),\n",
        "    Dense(128, activation='relu'),\n",
        "    Dense(64, activation='relu'),\n",
        "    Dense(32, activation='relu'),\n",
        "    Dense(10, activation='softmax')\n",
        "])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "FkB0r8ccoDgs"
      },
      "outputs": [],
      "source": [
        "# compile the model\n",
        "model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "ije6Dfljn_5_"
      },
      "outputs": [],
      "source": [
        "# augment images for training\n",
        "datagen.fit(x_train)\n",
        "augmented_images = []\n",
        "for batch in datagen.flow(x_train, batch_size=len(x_train), shuffle=False):\n",
        "    augmented_images.append(batch)\n",
        "    if len(augmented_images) * len(batch) >= len(x_train) * 2:\n",
        "        break\n",
        "augmented_images = np.concatenate(augmented_images)\n",
        "y_train_augmented = np.concatenate((y_train, y_train), axis=0)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 131
        },
        "id": "nrDUSVhqn9sG",
        "outputId": "984acd47-2c13-4819-a12b-39412fc7c219"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Number of images with augmentation:  120000\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 2500x200 with 10 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "#print augmentation images\n",
        "fig = plt.figure(figsize=(25, 2))\n",
        "print(\"Number of images with augmentation: \",augmented_images.shape[0])\n",
        "for i, image in enumerate(augmented_images[:10]):\n",
        "    # Create a new subplot for each image\n",
        "    plt.subplot(1, 10, i+1)\n",
        "    plt.xticks([])\n",
        "    plt.yticks([])\n",
        "    plt.grid(False)\n",
        "    plt.imshow(image, cmap='gray')\n",
        "    plt.xlabel(np.argmax(y_train_augmented[i]))\n",
        "\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 131
        },
        "id": "Gkf1LL3Sn7cS",
        "outputId": "7623bca4-5421-4055-cf1b-001925046b56"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Number of images without augmentation:  60000\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 2500x200 with 10 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "#print original images\n",
        "fig = plt.figure(figsize=(25, 2))\n",
        "print(\"Number of images without augmentation: \",x_train.shape[0])\n",
        "for i, image in enumerate(x_train[:10]):\n",
        "    # Create a new subplot for each image\n",
        "    plt.subplot(1, 10, i+1)\n",
        "    plt.xticks([])\n",
        "    plt.yticks([])\n",
        "    plt.grid(False)\n",
        "    plt.imshow(image, cmap='gray')\n",
        "    plt.xlabel(np.argmax(y_train[i]))\n",
        "\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1JGjgNQJoXQ_",
        "outputId": "2cd8b542-f4cf-4620-eecb-188e5db765de"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/50\n",
            "469/469 - 4s - loss: 0.5673 - accuracy: 0.8029 - 4s/epoch - 8ms/step\n",
            "Epoch 2/50\n",
            "469/469 - 2s - loss: 0.3963 - accuracy: 0.8593 - 2s/epoch - 4ms/step\n",
            "Epoch 3/50\n",
            "469/469 - 2s - loss: 0.3509 - accuracy: 0.8726 - 2s/epoch - 4ms/step\n",
            "Epoch 4/50\n",
            "469/469 - 2s - loss: 0.3256 - accuracy: 0.8813 - 2s/epoch - 4ms/step\n",
            "Epoch 5/50\n",
            "469/469 - 2s - loss: 0.3100 - accuracy: 0.8863 - 2s/epoch - 4ms/step\n",
            "Epoch 6/50\n",
            "469/469 - 3s - loss: 0.2936 - accuracy: 0.8907 - 3s/epoch - 5ms/step\n",
            "Epoch 7/50\n",
            "469/469 - 2s - loss: 0.2772 - accuracy: 0.8967 - 2s/epoch - 5ms/step\n",
            "Epoch 8/50\n",
            "469/469 - 2s - loss: 0.2701 - accuracy: 0.9000 - 2s/epoch - 4ms/step\n",
            "Epoch 9/50\n",
            "469/469 - 2s - loss: 0.2597 - accuracy: 0.9029 - 2s/epoch - 4ms/step\n",
            "Epoch 10/50\n",
            "469/469 - 2s - loss: 0.2514 - accuracy: 0.9066 - 2s/epoch - 4ms/step\n",
            "Epoch 11/50\n",
            "469/469 - 2s - loss: 0.2462 - accuracy: 0.9079 - 2s/epoch - 4ms/step\n",
            "Epoch 12/50\n",
            "469/469 - 3s - loss: 0.2365 - accuracy: 0.9117 - 3s/epoch - 5ms/step\n",
            "Epoch 13/50\n",
            "469/469 - 3s - loss: 0.2323 - accuracy: 0.9126 - 3s/epoch - 5ms/step\n",
            "Epoch 14/50\n",
            "469/469 - 2s - loss: 0.2234 - accuracy: 0.9157 - 2s/epoch - 4ms/step\n",
            "Epoch 15/50\n",
            "469/469 - 2s - loss: 0.2147 - accuracy: 0.9200 - 2s/epoch - 4ms/step\n",
            "Epoch 16/50\n",
            "469/469 - 2s - loss: 0.2134 - accuracy: 0.9209 - 2s/epoch - 4ms/step\n",
            "Epoch 17/50\n",
            "469/469 - 2s - loss: 0.2052 - accuracy: 0.9220 - 2s/epoch - 4ms/step\n",
            "Epoch 18/50\n",
            "469/469 - 3s - loss: 0.2005 - accuracy: 0.9240 - 3s/epoch - 6ms/step\n",
            "Epoch 19/50\n",
            "469/469 - 2s - loss: 0.1963 - accuracy: 0.9255 - 2s/epoch - 5ms/step\n",
            "Epoch 20/50\n",
            "469/469 - 2s - loss: 0.1897 - accuracy: 0.9286 - 2s/epoch - 4ms/step\n",
            "Epoch 21/50\n",
            "469/469 - 2s - loss: 0.1885 - accuracy: 0.9278 - 2s/epoch - 5ms/step\n",
            "Epoch 22/50\n",
            "469/469 - 2s - loss: 0.1839 - accuracy: 0.9302 - 2s/epoch - 4ms/step\n",
            "Epoch 23/50\n",
            "469/469 - 2s - loss: 0.1773 - accuracy: 0.9324 - 2s/epoch - 4ms/step\n",
            "Epoch 24/50\n",
            "469/469 - 3s - loss: 0.1715 - accuracy: 0.9368 - 3s/epoch - 7ms/step\n",
            "Epoch 25/50\n",
            "469/469 - 2s - loss: 0.1688 - accuracy: 0.9371 - 2s/epoch - 4ms/step\n",
            "Epoch 26/50\n",
            "469/469 - 2s - loss: 0.1659 - accuracy: 0.9370 - 2s/epoch - 4ms/step\n",
            "Epoch 27/50\n",
            "469/469 - 2s - loss: 0.1587 - accuracy: 0.9402 - 2s/epoch - 4ms/step\n",
            "Epoch 28/50\n",
            "469/469 - 2s - loss: 0.1630 - accuracy: 0.9383 - 2s/epoch - 4ms/step\n",
            "Epoch 29/50\n",
            "469/469 - 2s - loss: 0.1547 - accuracy: 0.9406 - 2s/epoch - 4ms/step\n",
            "Epoch 30/50\n",
            "469/469 - 3s - loss: 0.1494 - accuracy: 0.9435 - 3s/epoch - 7ms/step\n",
            "Epoch 31/50\n",
            "469/469 - 2s - loss: 0.1458 - accuracy: 0.9441 - 2s/epoch - 4ms/step\n",
            "Epoch 32/50\n",
            "469/469 - 2s - loss: 0.1464 - accuracy: 0.9445 - 2s/epoch - 4ms/step\n",
            "Epoch 33/50\n",
            "469/469 - 2s - loss: 0.1405 - accuracy: 0.9462 - 2s/epoch - 4ms/step\n",
            "Epoch 34/50\n",
            "469/469 - 2s - loss: 0.1368 - accuracy: 0.9484 - 2s/epoch - 4ms/step\n",
            "Epoch 35/50\n",
            "469/469 - 2s - loss: 0.1355 - accuracy: 0.9491 - 2s/epoch - 4ms/step\n",
            "Epoch 36/50\n",
            "469/469 - 3s - loss: 0.1338 - accuracy: 0.9490 - 3s/epoch - 6ms/step\n",
            "Epoch 37/50\n",
            "469/469 - 2s - loss: 0.1329 - accuracy: 0.9496 - 2s/epoch - 4ms/step\n",
            "Epoch 38/50\n",
            "469/469 - 2s - loss: 0.1278 - accuracy: 0.9514 - 2s/epoch - 4ms/step\n",
            "Epoch 39/50\n",
            "469/469 - 2s - loss: 0.1236 - accuracy: 0.9529 - 2s/epoch - 4ms/step\n",
            "Epoch 40/50\n",
            "469/469 - 2s - loss: 0.1229 - accuracy: 0.9537 - 2s/epoch - 4ms/step\n",
            "Epoch 41/50\n",
            "469/469 - 2s - loss: 0.1191 - accuracy: 0.9551 - 2s/epoch - 5ms/step\n",
            "Epoch 42/50\n",
            "469/469 - 3s - loss: 0.1191 - accuracy: 0.9553 - 3s/epoch - 6ms/step\n",
            "Epoch 43/50\n",
            "469/469 - 2s - loss: 0.1142 - accuracy: 0.9569 - 2s/epoch - 4ms/step\n",
            "Epoch 44/50\n",
            "469/469 - 2s - loss: 0.1110 - accuracy: 0.9582 - 2s/epoch - 4ms/step\n",
            "Epoch 45/50\n",
            "469/469 - 2s - loss: 0.1097 - accuracy: 0.9578 - 2s/epoch - 4ms/step\n",
            "Epoch 46/50\n",
            "469/469 - 2s - loss: 0.1077 - accuracy: 0.9597 - 2s/epoch - 4ms/step\n",
            "Epoch 47/50\n",
            "469/469 - 2s - loss: 0.1062 - accuracy: 0.9593 - 2s/epoch - 5ms/step\n",
            "Epoch 48/50\n",
            "469/469 - 3s - loss: 0.1077 - accuracy: 0.9597 - 3s/epoch - 5ms/step\n",
            "Epoch 49/50\n",
            "469/469 - 2s - loss: 0.1008 - accuracy: 0.9618 - 2s/epoch - 4ms/step\n",
            "Epoch 50/50\n",
            "469/469 - 2s - loss: 0.0973 - accuracy: 0.9635 - 2s/epoch - 4ms/step\n",
            "313/313 [==============================] - 1s 2ms/step - loss: 0.4934 - accuracy: 0.8881\n",
            "(60000, 28, 28, 1)\n",
            "Test accuracy before augmentation : 0.8881000280380249\n"
          ]
        }
      ],
      "source": [
        "# fit normal test data and show accuracy before augmented data\n",
        "model.fit(x_train, y_train, epochs=50, batch_size=128, verbose = 2)\n",
        "test_loss, test_acc = model.evaluate(x_test, y_test)\n",
        "print(x_train.shape)\n",
        "print('Test accuracy before augmentation :', test_acc)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "O1yTNKcan5Fj",
        "outputId": "717de39d-295f-482f-8c51-807bf26eed73"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/50\n",
            "938/938 [==============================] - 7s 8ms/step - loss: 0.3642 - accuracy: 0.8591 - val_loss: 0.4449 - val_accuracy: 0.8513\n",
            "Epoch 2/50\n",
            "938/938 [==============================] - 6s 6ms/step - loss: 0.3583 - accuracy: 0.8605 - val_loss: 0.4538 - val_accuracy: 0.8457\n",
            "Epoch 3/50\n",
            "938/938 [==============================] - 5s 5ms/step - loss: 0.3536 - accuracy: 0.8628 - val_loss: 0.4531 - val_accuracy: 0.8484\n",
            "Epoch 4/50\n",
            "938/938 [==============================] - 5s 6ms/step - loss: 0.3484 - accuracy: 0.8641 - val_loss: 0.4295 - val_accuracy: 0.8462\n",
            "Epoch 5/50\n",
            "938/938 [==============================] - 5s 6ms/step - loss: 0.3450 - accuracy: 0.8660 - val_loss: 0.4584 - val_accuracy: 0.8443\n",
            "Epoch 6/50\n",
            "938/938 [==============================] - 5s 5ms/step - loss: 0.3402 - accuracy: 0.8672 - val_loss: 0.4707 - val_accuracy: 0.8441\n",
            "Epoch 7/50\n",
            "938/938 [==============================] - 6s 6ms/step - loss: 0.3361 - accuracy: 0.8691 - val_loss: 0.4753 - val_accuracy: 0.8411\n",
            "Epoch 8/50\n",
            "938/938 [==============================] - 5s 5ms/step - loss: 0.3308 - accuracy: 0.8714 - val_loss: 0.4600 - val_accuracy: 0.8489\n",
            "Epoch 9/50\n",
            "938/938 [==============================] - 5s 6ms/step - loss: 0.3272 - accuracy: 0.8727 - val_loss: 0.4596 - val_accuracy: 0.8514\n",
            "Epoch 10/50\n",
            "938/938 [==============================] - 5s 6ms/step - loss: 0.3229 - accuracy: 0.8744 - val_loss: 0.5203 - val_accuracy: 0.8325\n",
            "Epoch 11/50\n",
            "938/938 [==============================] - 5s 5ms/step - loss: 0.3190 - accuracy: 0.8762 - val_loss: 0.4742 - val_accuracy: 0.8467\n",
            "Epoch 12/50\n",
            "938/938 [==============================] - 6s 6ms/step - loss: 0.3154 - accuracy: 0.8773 - val_loss: 0.4987 - val_accuracy: 0.8429\n",
            "Epoch 13/50\n",
            "938/938 [==============================] - 5s 5ms/step - loss: 0.3123 - accuracy: 0.8783 - val_loss: 0.4912 - val_accuracy: 0.8399\n",
            "Epoch 14/50\n",
            "938/938 [==============================] - 5s 5ms/step - loss: 0.3109 - accuracy: 0.8790 - val_loss: 0.5046 - val_accuracy: 0.8361\n",
            "Epoch 15/50\n",
            "938/938 [==============================] - 5s 6ms/step - loss: 0.3067 - accuracy: 0.8797 - val_loss: 0.5081 - val_accuracy: 0.8434\n",
            "Epoch 16/50\n",
            "938/938 [==============================] - 5s 5ms/step - loss: 0.2998 - accuracy: 0.8831 - val_loss: 0.5163 - val_accuracy: 0.8343\n",
            "Epoch 17/50\n",
            "938/938 [==============================] - 6s 6ms/step - loss: 0.2984 - accuracy: 0.8839 - val_loss: 0.5614 - val_accuracy: 0.8326\n",
            "Epoch 18/50\n",
            "938/938 [==============================] - 5s 5ms/step - loss: 0.2957 - accuracy: 0.8842 - val_loss: 0.5561 - val_accuracy: 0.8439\n",
            "Epoch 19/50\n",
            "938/938 [==============================] - 5s 5ms/step - loss: 0.2944 - accuracy: 0.8849 - val_loss: 0.5676 - val_accuracy: 0.8318\n",
            "Epoch 20/50\n",
            "938/938 [==============================] - 6s 6ms/step - loss: 0.2900 - accuracy: 0.8871 - val_loss: 0.5500 - val_accuracy: 0.8407\n",
            "Epoch 21/50\n",
            "938/938 [==============================] - 5s 6ms/step - loss: 0.2883 - accuracy: 0.8870 - val_loss: 0.5708 - val_accuracy: 0.8281\n",
            "Epoch 22/50\n",
            "938/938 [==============================] - 6s 6ms/step - loss: 0.2841 - accuracy: 0.8890 - val_loss: 0.5512 - val_accuracy: 0.8399\n",
            "Epoch 23/50\n",
            "938/938 [==============================] - 5s 5ms/step - loss: 0.2810 - accuracy: 0.8913 - val_loss: 0.5949 - val_accuracy: 0.8271\n",
            "Epoch 24/50\n",
            "938/938 [==============================] - 6s 6ms/step - loss: 0.2807 - accuracy: 0.8909 - val_loss: 0.5729 - val_accuracy: 0.8340\n",
            "Epoch 25/50\n",
            "938/938 [==============================] - 5s 5ms/step - loss: 0.2775 - accuracy: 0.8924 - val_loss: 0.6136 - val_accuracy: 0.8262\n",
            "Epoch 26/50\n",
            "938/938 [==============================] - 5s 5ms/step - loss: 0.2754 - accuracy: 0.8931 - val_loss: 0.5681 - val_accuracy: 0.8447\n",
            "Epoch 27/50\n",
            "938/938 [==============================] - 6s 6ms/step - loss: 0.2727 - accuracy: 0.8937 - val_loss: 0.6357 - val_accuracy: 0.8292\n",
            "Epoch 28/50\n",
            "938/938 [==============================] - 5s 5ms/step - loss: 0.2706 - accuracy: 0.8942 - val_loss: 0.6015 - val_accuracy: 0.8338\n",
            "Epoch 29/50\n",
            "938/938 [==============================] - 6s 6ms/step - loss: 0.2677 - accuracy: 0.8967 - val_loss: 0.5803 - val_accuracy: 0.8381\n",
            "Epoch 30/50\n",
            "938/938 [==============================] - 5s 5ms/step - loss: 0.2662 - accuracy: 0.8960 - val_loss: 0.6045 - val_accuracy: 0.8307\n",
            "Epoch 31/50\n",
            "938/938 [==============================] - 5s 5ms/step - loss: 0.2625 - accuracy: 0.8981 - val_loss: 0.6316 - val_accuracy: 0.8245\n",
            "Epoch 32/50\n",
            "938/938 [==============================] - 6s 6ms/step - loss: 0.2629 - accuracy: 0.8983 - val_loss: 0.6200 - val_accuracy: 0.8276\n",
            "Epoch 33/50\n",
            "938/938 [==============================] - 5s 5ms/step - loss: 0.2588 - accuracy: 0.9000 - val_loss: 0.5929 - val_accuracy: 0.8335\n",
            "Epoch 34/50\n",
            "938/938 [==============================] - 6s 6ms/step - loss: 0.2580 - accuracy: 0.8993 - val_loss: 0.6157 - val_accuracy: 0.8398\n",
            "Epoch 35/50\n",
            "938/938 [==============================] - 5s 5ms/step - loss: 0.2548 - accuracy: 0.9006 - val_loss: 0.6417 - val_accuracy: 0.8270\n",
            "Epoch 36/50\n",
            "938/938 [==============================] - 5s 5ms/step - loss: 0.2531 - accuracy: 0.9017 - val_loss: 0.6507 - val_accuracy: 0.8351\n",
            "Epoch 37/50\n",
            "938/938 [==============================] - 6s 7ms/step - loss: 0.2544 - accuracy: 0.9002 - val_loss: 0.6616 - val_accuracy: 0.8307\n",
            "Epoch 38/50\n",
            "938/938 [==============================] - 5s 5ms/step - loss: 0.2497 - accuracy: 0.9028 - val_loss: 0.6706 - val_accuracy: 0.8324\n",
            "Epoch 39/50\n",
            "938/938 [==============================] - 6s 6ms/step - loss: 0.2481 - accuracy: 0.9040 - val_loss: 0.6641 - val_accuracy: 0.8328\n",
            "Epoch 40/50\n",
            "938/938 [==============================] - 5s 5ms/step - loss: 0.2484 - accuracy: 0.9034 - val_loss: 0.6786 - val_accuracy: 0.8372\n",
            "Epoch 41/50\n",
            "938/938 [==============================] - 5s 5ms/step - loss: 0.2451 - accuracy: 0.9043 - val_loss: 0.6877 - val_accuracy: 0.8342\n",
            "Epoch 42/50\n",
            "938/938 [==============================] - 6s 6ms/step - loss: 0.2422 - accuracy: 0.9059 - val_loss: 0.6859 - val_accuracy: 0.8381\n",
            "Epoch 43/50\n",
            "938/938 [==============================] - 5s 5ms/step - loss: 0.2426 - accuracy: 0.9056 - val_loss: 0.6901 - val_accuracy: 0.8327\n",
            "Epoch 44/50\n",
            "938/938 [==============================] - 6s 7ms/step - loss: 0.2384 - accuracy: 0.9075 - val_loss: 0.7373 - val_accuracy: 0.8366\n",
            "Epoch 45/50\n",
            "938/938 [==============================] - 5s 5ms/step - loss: 0.2382 - accuracy: 0.9068 - val_loss: 0.6926 - val_accuracy: 0.8380\n",
            "Epoch 46/50\n",
            "938/938 [==============================] - 6s 6ms/step - loss: 0.2384 - accuracy: 0.9065 - val_loss: 0.7573 - val_accuracy: 0.8156\n",
            "Epoch 47/50\n",
            "938/938 [==============================] - 5s 5ms/step - loss: 0.2333 - accuracy: 0.9087 - val_loss: 0.7249 - val_accuracy: 0.8331\n",
            "Epoch 48/50\n",
            "938/938 [==============================] - 5s 5ms/step - loss: 0.2316 - accuracy: 0.9106 - val_loss: 0.7325 - val_accuracy: 0.8311\n",
            "Epoch 49/50\n",
            "938/938 [==============================] - 6s 6ms/step - loss: 0.2322 - accuracy: 0.9098 - val_loss: 0.7510 - val_accuracy: 0.8289\n",
            "Epoch 50/50\n",
            "938/938 [==============================] - 5s 5ms/step - loss: 0.2320 - accuracy: 0.9095 - val_loss: 0.7397 - val_accuracy: 0.8336\n"
          ]
        }
      ],
      "source": [
        "# fit the model\n",
        "history = model.fit(\n",
        "    augmented_images, y_train_augmented,\n",
        "    validation_data=(x_test, y_test),\n",
        "    epochs=50, batch_size=128\n",
        ")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "NvOvZ693n2wC",
        "outputId": "1f192a06-eca1-4a53-e022-684ca328a6b4"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "313/313 - 0s - loss: 0.7397 - accuracy: 0.8336 - 490ms/epoch - 2ms/step\n",
            "(120000, 28, 28, 1)\n",
            "Test accuracy after augmentation: 0.8335999846458435\n"
          ]
        }
      ],
      "source": [
        "# evaluate the model\n",
        "test_loss, test_acc = model.evaluate(x_test, y_test, verbose=2)\n",
        "print(augmented_images.shape)\n",
        "print('Test accuracy after augmentation:', test_acc)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 430
        },
        "id": "eWE5QaYSn09w",
        "outputId": "9a90bfc1-7123-4982-8156-86cba9b53210"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "# plot accuracy and loss\n",
        "plt.plot(history.history['accuracy'], label='train_accuracy')\n",
        "plt.plot(history.history['val_accuracy'], label='val_accuracy')\n",
        "plt.plot(history.history['loss'], label='train_loss')\n",
        "plt.plot(history.history['val_loss'], label='val_loss')\n",
        "plt.legend()\n",
        "plt.show()"
      ]
    }
  ],
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}