submission_3_(1).ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "submission 3 (1).ipynb",
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "EHTydy1EKgPI",
        "outputId": "2a9a9912-903b-42c1-ea03-2437a9d3962a",
        "colab": {
          "resources": {
            "http://localhost:8080/nbextensions/google.colab/files.js": {
              "data": "Ly8gQ29weXJpZ2h0IDIwMTcgR29vZ2xlIExMQwovLwovLyBMaWNlbnNlZCB1bmRlciB0aGUgQXBhY2hlIExpY2Vuc2UsIFZlcnNpb24gMi4wICh0aGUgIkxpY2Vuc2UiKTsKLy8geW91IG1heSBub3QgdXNlIHRoaXMgZmlsZSBleGNlcHQgaW4gY29tcGxpYW5jZSB3aXRoIHRoZSBMaWNlbnNlLgovLyBZb3UgbWF5IG9idGFpbiBhIGNvcHkgb2YgdGhlIExpY2Vuc2UgYXQKLy8KLy8gICAgICBodHRwOi8vd3d3LmFwYWNoZS5vcmcvbGljZW5zZXMvTElDRU5TRS0yLjAKLy8KLy8gVW5sZXNzIHJlcXVpcmVkIGJ5IGFwcGxpY2FibGUgbGF3IG9yIGFncmVlZCB0byBpbiB3cml0aW5nLCBzb2Z0d2FyZQovLyBkaXN0cmlidXRlZCB1bmRlciB0aGUgTGljZW5zZSBpcyBkaXN0cmlidXRlZCBvbiBhbiAiQVMgSVMiIEJBU0lTLAovLyBXSVRIT1VUIFdBUlJBTlRJRVMgT1IgQ09ORElUSU9OUyBPRiBBTlkgS0lORCwgZWl0aGVyIGV4cHJlc3Mgb3IgaW1wbGllZC4KLy8gU2VlIHRoZSBMaWNlbnNlIGZvciB0aGUgc3BlY2lmaWMgbGFuZ3VhZ2UgZ292ZXJuaW5nIHBlcm1pc3Npb25zIGFuZAovLyBsaW1pdGF0aW9ucyB1bmRlciB0aGUgTGljZW5zZS4KCi8qKgogKiBAZmlsZW92ZXJ2aWV3IEhlbHBlcnMgZm9yIGdvb2dsZS5jb2xhYiBQeXRob24gbW9kdWxlLgogKi8KKGZ1bmN0aW9uKHNjb3BlKSB7CmZ1bmN0aW9uIHNwYW4odGV4dCwgc3R5bGVBdHRyaWJ1dGVzID0ge30pIHsKICBjb25zdCBlbGVtZW50ID0gZG9jdW1lbnQuY3JlYXRlRWxlbWVudCgnc3BhbicpOwogIGVsZW1lbnQudGV4dENvbnRlbnQgPSB0ZXh0OwogIGZvciAoY29uc3Qga2V5IG9mIE9iamVjdC5rZXlzKHN0eWxlQXR0cmlidXRlcykpIHsKICAgIGVsZW1lbnQuc3R5bGVba2V5XSA9IHN0eWxlQXR0cmlidXRlc1trZXldOwogIH0KICByZXR1cm4gZWxlbWVudDsKfQoKLy8gTWF4IG51bWJlciBvZiBieXRlcyB3aGljaCB3aWxsIGJlIHVwbG9hZGVkIGF0IGEgdGltZS4KY29uc3QgTUFYX1BBWUxPQURfU0laRSA9IDEwMCAqIDEwMjQ7CgpmdW5jdGlvbiBfdXBsb2FkRmlsZXMoaW5wdXRJZCwgb3V0cHV0SWQpIHsKICBjb25zdCBzdGVwcyA9IHVwbG9hZEZpbGVzU3RlcChpbnB1dElkLCBvdXRwdXRJZCk7CiAgY29uc3Qgb3V0cHV0RWxlbWVudCA9IGRvY3VtZW50LmdldEVsZW1lbnRCeUlkKG91dHB1dElkKTsKICAvLyBDYWNoZSBzdGVwcyBvbiB0aGUgb3V0cHV0RWxlbWVudCB0byBtYWtlIGl0IGF2YWlsYWJsZSBmb3IgdGhlIG5leHQgY2FsbAogIC8vIHRvIHVwbG9hZEZpbGVzQ29udGludWUgZnJvbSBQeXRob24uCiAgb3V0cHV0RWxlbWVudC5zdGVwcyA9IHN0ZXBzOwoKICByZXR1cm4gX3VwbG9hZEZpbGVzQ29udGludWUob3V0cHV0SWQpOwp9CgovLyBUaGlzIGlzIHJvdWdobHkgYW4gYXN5bmMgZ2VuZXJhdG9yIChub3Qgc3VwcG9ydGVkIGluIHRoZSBicm93c2VyIHlldCksCi8vIHdoZXJlIHRoZXJlIGFyZSBtdWx0aXBsZSBhc3luY2hyb25vdXMgc3RlcHMgYW5kIHRoZSBQeXRob24gc2lkZSBpcyBnb2luZwovLyB0byBwb2xsIGZvciBjb21wbGV0aW9uIG9mIGVhY2ggc3RlcC4KLy8gVGhpcyB1c2VzIGEgUHJvbWlzZSB0byBibG9jayB0aGUgcHl0aG9uIHNpZGUgb24gY29tcGxldGlvbiBvZiBlYWNoIHN0ZXAsCi8vIHRoZW4gcGFzc2VzIHRoZSByZXN1bHQgb2YgdGhlIHByZXZpb3VzIHN0ZXAgYXMgdGhlIGlucHV0IHRvIHRoZSBuZXh0IHN0ZXAuCmZ1bmN0aW9uIF91cGxvYWRGaWxlc0NvbnRpbnVlKG91dHB1dElkKSB7CiAgY29uc3Qgb3V0cHV0RWxlbWVudCA9IGRvY3VtZW50LmdldEVsZW1lbnRCeUlkKG91dHB1dElkKTsKICBjb25zdCBzdGVwcyA9IG91dHB1dEVsZW1lbnQuc3RlcHM7CgogIGNvbnN0IG5leHQgPSBzdGVwcy5uZXh0KG91dHB1dEVsZW1lbnQubGFzdFByb21pc2VWYWx1ZSk7CiAgcmV0dXJuIFByb21pc2UucmVzb2x2ZShuZXh0LnZhbHVlLnByb21pc2UpLnRoZW4oKHZhbHVlKSA9PiB7CiAgICAvLyBDYWNoZSB0aGUgbGFzdCBwcm9taXNlIHZhbHVlIHRvIG1ha2UgaXQgYXZhaWxhYmxlIHRvIHRoZSBuZXh0CiAgICAvLyBzdGVwIG9mIHRoZSBnZW5lcmF0b3IuCiAgICBvdXRwdXRFbGVtZW50Lmxhc3RQcm9taXNlVmFsdWUgPSB2YWx1ZTsKICAgIHJldHVybiBuZXh0LnZhbHVlLnJlc3BvbnNlOwogIH0pOwp9CgovKioKICogR2VuZXJhdG9yIGZ1bmN0aW9uIHdoaWNoIGlzIGNhbGxlZCBiZXR3ZWVuIGVhY2ggYXN5bmMgc3RlcCBvZiB0aGUgdXBsb2FkCiAqIHByb2Nlc3MuCiAqIEBwYXJhbSB7c3RyaW5nfSBpbnB1dElkIEVsZW1lbnQgSUQgb2YgdGhlIGlucHV0IGZpbGUgcGlja2VyIGVsZW1lbnQuCiAqIEBwYXJhbSB7c3RyaW5nfSBvdXRwdXRJZCBFbGVtZW50IElEIG9mIHRoZSBvdXRwdXQgZGlzcGxheS4KICogQHJldHVybiB7IUl0ZXJhYmxlPCFPYmplY3Q+fSBJdGVyYWJsZSBvZiBuZXh0IHN0ZXBzLgogKi8KZnVuY3Rpb24qIHVwbG9hZEZpbGVzU3RlcChpbnB1dElkLCBvdXRwdXRJZCkgewogIGNvbnN0IGlucHV0RWxlbWVudCA9IGRvY3VtZW50LmdldEVsZW1lbnRCeUlkKGlucHV0SWQpOwogIGlucHV0RWxlbWVudC5kaXNhYmxlZCA9IGZhbHNlOwoKICBjb25zdCBvdXRwdXRFbGVtZW50ID0gZG9jdW1lbnQuZ2V0RWxlbWVudEJ5SWQob3V0cHV0SWQpOwogIG91dHB1dEVsZW1lbnQuaW5uZXJIVE1MID0gJyc7CgogIGNvbnN0IHBpY2tlZFByb21pc2UgPSBuZXcgUHJvbWlzZSgocmVzb2x2ZSkgPT4gewogICAgaW5wdXRFbGVtZW50LmFkZEV2ZW50TGlzdGVuZXIoJ2NoYW5nZScsIChlKSA9PiB7CiAgICAgIHJlc29sdmUoZS50YXJnZXQuZmlsZXMpOwogICAgfSk7CiAgfSk7CgogIGNvbnN0IGNhbmNlbCA9IGRvY3VtZW50LmNyZWF0ZUVsZW1lbnQoJ2J1dHRvbicpOwogIGlucHV0RWxlbWVudC5wYXJlbnRFbGVtZW50LmFwcGVuZENoaWxkKGNhbmNlbCk7CiAgY2FuY2VsLnRleHRDb250ZW50ID0gJ0NhbmNlbCB1cGxvYWQnOwogIGNvbnN0IGNhbmNlbFByb21pc2UgPSBuZXcgUHJvbWlzZSgocmVzb2x2ZSkgPT4gewogICAgY2FuY2VsLm9uY2xpY2sgPSAoKSA9PiB7CiAgICAgIHJlc29sdmUobnVsbCk7CiAgICB9OwogIH0pOwoKICAvLyBXYWl0IGZvciB0aGUgdXNlciB0byBwaWNrIHRoZSBmaWxlcy4KICBjb25zdCBmaWxlcyA9IHlpZWxkIHsKICAgIHByb21pc2U6IFByb21pc2UucmFjZShbcGlja2VkUHJvbWlzZSwgY2FuY2VsUHJvbWlzZV0pLAogICAgcmVzcG9uc2U6IHsKICAgICAgYWN0aW9uOiAnc3RhcnRpbmcnLAogICAgfQogIH07CgogIGNhbmNlbC5yZW1vdmUoKTsKCiAgLy8gRGlzYWJsZSB0aGUgaW5wdXQgZWxlbWVudCBzaW5jZSBmdXJ0aGVyIHBpY2tzIGFyZSBub3QgYWxsb3dlZC4KICBpbnB1dEVsZW1lbnQuZGlzYWJsZWQgPSB0cnVlOwoKICBpZiAoIWZpbGVzKSB7CiAgICByZXR1cm4gewogICAgICByZXNwb25zZTogewogICAgICAgIGFjdGlvbjogJ2NvbXBsZXRlJywKICAgICAgfQogICAgfTsKICB9CgogIGZvciAoY29uc3QgZmlsZSBvZiBmaWxlcykgewogICAgY29uc3QgbGkgPSBkb2N1bWVudC5jcmVhdGVFbGVtZW50KCdsaScpOwogICAgbGkuYXBwZW5kKHNwYW4oZmlsZS5uYW1lLCB7Zm9udFdlaWdodDogJ2JvbGQnfSkpOwogICAgbGkuYXBwZW5kKHNwYW4oCiAgICAgICAgYCgke2ZpbGUudHlwZSB8fCAnbi9hJ30pIC0gJHtmaWxlLnNpemV9IGJ5dGVzLCBgICsKICAgICAgICBgbGFzdCBtb2RpZmllZDogJHsKICAgICAgICAgICAgZmlsZS5sYXN0TW9kaWZpZWREYXRlID8gZmlsZS5sYXN0TW9kaWZpZWREYXRlLnRvTG9jYWxlRGF0ZVN0cmluZygpIDoKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgJ24vYSd9IC0gYCkpOwogICAgY29uc3QgcGVyY2VudCA9IHNwYW4oJzAlIGRvbmUnKTsKICAgIGxpLmFwcGVuZENoaWxkKHBlcmNlbnQpOwoKICAgIG91dHB1dEVsZW1lbnQuYXBwZW5kQ2hpbGQobGkpOwoKICAgIGNvbnN0IGZpbGVEYXRhUHJvbWlzZSA9IG5ldyBQcm9taXNlKChyZXNvbHZlKSA9PiB7CiAgICAgIGNvbnN0IHJlYWRlciA9IG5ldyBGaWxlUmVhZGVyKCk7CiAgICAgIHJlYWRlci5vbmxvYWQgPSAoZSkgPT4gewogICAgICAgIHJlc29sdmUoZS50YXJnZXQucmVzdWx0KTsKICAgICAgfTsKICAgICAgcmVhZGVyLnJlYWRBc0FycmF5QnVmZmVyKGZpbGUpOwogICAgfSk7CiAgICAvLyBXYWl0IGZvciB0aGUgZGF0YSB0byBiZSByZWFkeS4KICAgIGxldCBmaWxlRGF0YSA9IHlpZWxkIHsKICAgICAgcHJvbWlzZTogZmlsZURhdGFQcm9taXNlLAogICAgICByZXNwb25zZTogewogICAgICAgIGFjdGlvbjogJ2NvbnRpbnVlJywKICAgICAgfQogICAgfTsKCiAgICAvLyBVc2UgYSBjaHVua2VkIHNlbmRpbmcgdG8gYXZvaWQgbWVzc2FnZSBzaXplIGxpbWl0cy4gU2VlIGIvNjIxMTU2NjAuCiAgICBsZXQgcG9zaXRpb24gPSAwOwogICAgd2hpbGUgKHBvc2l0aW9uIDwgZmlsZURhdGEuYnl0ZUxlbmd0aCkgewogICAgICBjb25zdCBsZW5ndGggPSBNYXRoLm1pbihmaWxlRGF0YS5ieXRlTGVuZ3RoIC0gcG9zaXRpb24sIE1BWF9QQVlMT0FEX1NJWkUpOwogICAgICBjb25zdCBjaHVuayA9IG5ldyBVaW50OEFycmF5KGZpbGVEYXRhLCBwb3NpdGlvbiwgbGVuZ3RoKTsKICAgICAgcG9zaXRpb24gKz0gbGVuZ3RoOwoKICAgICAgY29uc3QgYmFzZTY0ID0gYnRvYShTdHJpbmcuZnJvbUNoYXJDb2RlLmFwcGx5KG51bGwsIGNodW5rKSk7CiAgICAgIHlpZWxkIHsKICAgICAgICByZXNwb25zZTogewogICAgICAgICAgYWN0aW9uOiAnYXBwZW5kJywKICAgICAgICAgIGZpbGU6IGZpbGUubmFtZSwKICAgICAgICAgIGRhdGE6IGJhc2U2NCwKICAgICAgICB9LAogICAgICB9OwogICAgICBwZXJjZW50LnRleHRDb250ZW50ID0KICAgICAgICAgIGAke01hdGgucm91bmQoKHBvc2l0aW9uIC8gZmlsZURhdGEuYnl0ZUxlbmd0aCkgKiAxMDApfSUgZG9uZWA7CiAgICB9CiAgfQoKICAvLyBBbGwgZG9uZS4KICB5aWVsZCB7CiAgICByZXNwb25zZTogewogICAgICBhY3Rpb246ICdjb21wbGV0ZScsCiAgICB9CiAgfTsKfQoKc2NvcGUuZ29vZ2xlID0gc2NvcGUuZ29vZ2xlIHx8IHt9OwpzY29wZS5nb29nbGUuY29sYWIgPSBzY29wZS5nb29nbGUuY29sYWIgfHwge307CnNjb3BlLmdvb2dsZS5jb2xhYi5fZmlsZXMgPSB7CiAgX3VwbG9hZEZpbGVzLAogIF91cGxvYWRGaWxlc0NvbnRpbnVlLAp9Owp9KShzZWxmKTsK",
              "ok": true,
              "headers": [
                [
                  "content-type",
                  "application/javascript"
                ]
              ],
              "status": 200,
              "status_text": ""
            }
          },
          "base_uri": "https://localhost:8080/",
          "height": 73
        }
      },
      "source": [
        "from google.colab import files\n",
        "!pip install -q kaggle\n",
        "uploaded = files.upload()"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "\n",
              "     <input type=\"file\" id=\"files-46131a70-baf9-4edb-8c9f-35a895b38e9b\" name=\"files[]\" multiple disabled\n",
              "        style=\"border:none\" />\n",
              "     <output id=\"result-46131a70-baf9-4edb-8c9f-35a895b38e9b\">\n",
              "      Upload widget is only available when the cell has been executed in the\n",
              "      current browser session. Please rerun this cell to enable.\n",
              "      </output>\n",
              "      <script src=\"/nbextensions/google.colab/files.js\"></script> "
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Saving kaggle.json to kaggle.json\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SoVU9p3yKjOk",
        "outputId": "3985a881-9213-4ced-cde3-ef15c19214fc",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 70
        }
      },
      "source": [
        "!mkdir -p ~/.kaggle\n",
        "!cp kaggle.json ~/.kaggle/\n",
        "!chmod 600 ~/.kaggle/kaggle.json\n",
        "!kaggle datasets download -d ashishsaxena2209/animal-image-datasetdog-cat-and-panda"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Downloading animal-image-datasetdog-cat-and-panda.zip to /content\n",
            " 99% 373M/376M [00:03<00:00, 117MB/s]\n",
            "100% 376M/376M [00:03<00:00, 106MB/s]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9vKt9JBdKkZP"
      },
      "source": [
        "import tensorflow as tf\n",
        "import keras_preprocessing\n",
        "from keras_preprocessing import image\n",
        "from keras_preprocessing.image import ImageDataGenerator"
      ],
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4ZZVQ0wLKnCm"
      },
      "source": [
        "import os\n",
        "import zipfile\n",
        "import shutil\n",
        "\n",
        "local_zip = '/content/animal-image-datasetdog-cat-and-panda.zip'\n",
        "zip_ref = zipfile.ZipFile(local_zip, 'r')\n",
        "zip_ref.extractall('/content/')\n",
        "zip_ref.close()"
      ],
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1FiuNqFsKovt",
        "outputId": "92f7dc36-bf03-4918-a276-3f05c8ca8eac",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "dir_utama = os.path.join('/content/animals')\n",
        "\n",
        "print(os.listdir(dir_utama))"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "['images', 'cats', 'panda', 'animals', 'dogs']\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ON7c0QbYKsSz",
        "outputId": "67491c69-0355-48a7-8e85-50876c331061",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "ignore_dir = ['images', 'animals']\n",
        "\n",
        "for dir in ignore_dir:\n",
        "  path = os.path.join(dir_utama, dir)\n",
        "  shutil.rmtree(path)\n",
        "\n",
        "print(os.listdir(dir_utama))"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "['cats', 'panda', 'dogs']\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NG8sxe3pKuA7",
        "outputId": "47bac238-8b92-49ac-83cb-f271e8dcb5a9",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 421
        }
      },
      "source": [
        "from PIL import Image\n",
        "jumlah_sampel = 0\n",
        "\n",
        "for x in os.listdir(dir_utama):\n",
        "  dir = os.path.join('/content/animals/', x)\n",
        "  y = len(os.listdir(dir))\n",
        "  print(x+':', y)\n",
        "  jumlah_sampel = jumlah_sampel + y\n",
        "  \n",
        "  img_name = os.listdir(dir)\n",
        "  for z in range(5):\n",
        "    img_path = os.path.join(dir, img_name[z])\n",
        "    img = Image.open(img_path)\n",
        "    print('-',img.size)\n",
        "  print('---------------')\n",
        "\n",
        "\n",
        "print('\\nJumlah total sampel:', jumlah_sampel)"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "cats: 1000\n",
            "- (249, 250)\n",
            "- (150, 139)\n",
            "- (500, 497)\n",
            "- (399, 286)\n",
            "- (499, 375)\n",
            "---------------\n",
            "panda: 1000\n",
            "- (595, 424)\n",
            "- (497, 500)\n",
            "- (500, 375)\n",
            "- (500, 378)\n",
            "- (500, 375)\n",
            "---------------\n",
            "dogs: 1000\n",
            "- (369, 500)\n",
            "- (269, 500)\n",
            "- (362, 399)\n",
            "- (500, 374)\n",
            "- (500, 312)\n",
            "---------------\n",
            "\n",
            "Jumlah total sampel: 3000\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lcigiyCMKzE4"
      },
      "source": [
        "train_datagen = ImageDataGenerator(\n",
        "    rescale = 1./255,\n",
        "    rotation_range = 20,\n",
        "    horizontal_flip = True,\n",
        "    shear_range = 0.2,\n",
        "    zoom_range = 0.2,\n",
        "    fill_mode = 'nearest',\n",
        "    validation_split = 0.2 # Membagi Data Validasi 20% dan Training 80%\n",
        ")"
      ],
      "execution_count": 8,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YelO_2S8KzqC"
      },
      "source": [
        "batch_size = 128 \n",
        "\n",
        "generator_latih = train_datagen.flow_from_directory(\n",
        "    dir_utama,\n",
        "    target_size = (150,150), # Besar Input harus sama besar dan dijadikan 150x150\n",
        "    class_mode = 'categorical', # Menggunakan lebih dari 2 kelas\n",
        "    batch_size = batch_size,\n",
        "    subset = 'training' # Penentuan sebagai data latih\n",
        ")\n",
        "\n",
        "generator_validasi = train_datagen.flow_from_directory(\n",
        "    dir_utama, # Direktori sama dengan generator latih karena pembagian dilakukan oleh Generator\n",
        "    target_size = (150,150),\n",
        "    class_mode = 'categorical',\n",
        "    batch_size = batch_size,\n",
        "    subset = 'validation' # Penentuan sebagai data latih\n",
        ")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DJbBsfrdK3Za"
      },
      "source": [
        "model = tf.keras.models.Sequential([\n",
        "    # Besar masukan dibuat sebesar 150x150 dengan warna 3 bytes\n",
        "    # Konvulasi pertama\n",
        "    tf.keras.layers.Conv2D(64, (3,3), activation='relu', input_shape=(150, 150, 3)),\n",
        "    tf.keras.layers.MaxPooling2D(2, 2),\n",
        "    # Konvolusi Kedua\n",
        "    tf.keras.layers.Conv2D(64, (3,3), activation='relu'),\n",
        "    tf.keras.layers.MaxPooling2D(2,2),\n",
        "    # Konvolusi Ketiga\n",
        "    tf.keras.layers.Conv2D(128, (3,3), activation='relu'),\n",
        "    tf.keras.layers.MaxPooling2D(2,2),\n",
        "    # Konvolusi Keempat\n",
        "    tf.keras.layers.Conv2D(128, (3,3), activation='relu'),\n",
        "    tf.keras.layers.MaxPooling2D(2,2),\n",
        "    # Meratakan input menjadi 1 dimensi\n",
        "    tf.keras.layers.Flatten(),\n",
        "    # Menggunakan Dropout sebesar 50%\n",
        "    tf.keras.layers.Dropout(0.5), \n",
        "    # 512 neuron terhadap hidden layer\n",
        "    tf.keras.layers.Dense(512, activation='relu'),\n",
        "    # 3 neuron/kelas pada output layer\n",
        "    tf.keras.layers.Dense(3, activation='softmax')\n",
        "])\n",
        "\n",
        "model.summary()\n",
        "\n",
        "model.compile(loss = 'categorical_crossentropy', \n",
        "              optimizer = 'rmsprop', # Opmitizer RMSprop\n",
        "              metrics = ['accuracy'])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VSkSbnOOK37m"
      },
      "source": [
        "class myCallback(tf.keras.callbacks.Callback):\n",
        "  def on_epoch_end(self, epoch, logs={}):\n",
        "    if (logs.get('accuracy')>0.80 and logs.get('val_accuracy')>0.80) :\n",
        "      print(\"\\nAkurasi di atas 80%, hentikan pelatihan!\")\n",
        "      self.model.stop_training = True\n",
        "\n",
        "callbacks = myCallback()"
      ],
      "execution_count": 16,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "l8KL2sfiK_GM"
      },
      "source": [
        "history = model.fit(generator_latih, \n",
        "                    epochs = 40, \n",
        "                    steps_per_epoch = 2400//batch_size, \n",
        "                    validation_data = generator_validasi, \n",
        "                    verbose = 1, \n",
        "                    validation_steps = 600//batch_size,\n",
        "                    callbacks = [callbacks])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "naPlxdm4qQeu",
        "outputId": "a2ca34d4-9e4c-45a1-bbac-a1c406415a54",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 541
        }
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "\n",
        "def plot_graphs(history, string):\n",
        "  plt.plot(history.history[string])\n",
        "  plt.plot(history.history['val_'+string])\n",
        "  plt.xlabel(\"Epochs\")\n",
        "  plt.ylabel(string)\n",
        "  plt.legend([string, 'val_'+string])\n",
        "  plt.show()\n",
        "  \n",
        "plot_graphs(history, \"accuracy\")\n",
        "plot_graphs(history, \"loss\")"
      ],
      "execution_count": 18,
      "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"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Oy89u82RqUMg"
      },
      "source": [
        "# Convert the model.\n",
        "converter = tf.lite.TFLiteConverter.from_keras_model(model)\n",
        "tflite_model = converter.convert()\n",
        "\n",
        "# Save the model.\n",
        "with open('model.tflite', 'wb') as f:\n",
        "  f.write(tflite_model)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pxFLcLdpqWkK"
      },
      "source": [
        "!ls -a"
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}