Fruit_Classification_implement_TF_lite.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "authorship_tag": "ABX9TyN31fsioiWCeFCVq5kSqH28",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/Awangnugrawan/Fruit-Classification-implement-TF-lite/blob/main/Fruit_Classification_implement_TF_lite.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# SUBMISSION PROJECT DICODING \"Belajar Pengemabangan Machine Learning\"\n",
        "## DATA DIRI\n",
        "\n",
        "*  Nama= Awang Mulya Nugrawan\n",
        "*  username DICODING= awang_nugrawan\n",
        "*  email = awangmulyanugrawan@gmail.com\n",
        "*  No.hp = 082191862002\n",
        "*  Asal= Makassar, Sulawesi Selatan\n",
        "*  Status = Mahasiswa Sisfor Universitas Hasanuddin\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n"
      ],
      "metadata": {
        "id": "d-O7EBWjn7hx"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "#LIBRARY"
      ],
      "metadata": {
        "id": "nMTi7SqcoOlq"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import tensorflow as tf\n",
        "import zipfile,os\n",
        "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
        "from tensorflow.keras.layers import Dense, Conv2D, Flatten, MaxPooling2D\n",
        "from tensorflow.keras.models import Sequential\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "from google.colab import files\n",
        "import matplotlib.pyplot as plt\n",
        "import matplotlib.image as mpimg\n",
        "import pathlib"
      ],
      "metadata": {
        "id": "AD-EyZUNc18K"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "#Import Dataset dengan API KAGGLE"
      ],
      "metadata": {
        "id": "JLeRPFf5qXEo"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "jYiLM6w2ayxP"
      },
      "outputs": [],
      "source": [
        "! chmod 600 /content/kaggle.json"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "! KAGGLE_CONFIG_DIR=/content/  kaggle datasets download -d sshikamaru/fruit-recognition"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "hgHwWmgJbG01",
        "outputId": "258189bd-5253-4f01-d774-4e746e618b7f"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading fruit-recognition.zip to /content\n",
            " 99% 101M/102M [00:04<00:00, 21.4MB/s] \n",
            "100% 102M/102M [00:04<00:00, 21.6MB/s]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "zip_file = zipfile.ZipFile('/content/fruit-recognition.zip')\n",
        "zip_file.extractall('/tmp/')\n",
        "\n",
        "dir_path = '/tmp/data/train'\n",
        "\n",
        "os.listdir(dir_path)\n",
        "\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "h3Nde-H9b7IC",
        "outputId": "d10d8969-b802-4074-f1f7-34355c7b73d0"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['Banana',\n",
              " 'Passion Fruit',\n",
              " 'Lemon',\n",
              " 'Grape Blue',\n",
              " 'Onion White',\n",
              " 'Plum',\n",
              " 'Apple Braeburn',\n",
              " 'Pepper Green',\n",
              " 'Limes',\n",
              " 'Avocado',\n",
              " 'Papaya',\n",
              " 'Cactus fruit',\n",
              " 'Orange',\n",
              " 'Potato Red',\n",
              " 'Blueberry',\n",
              " 'Cantaloupe',\n",
              " 'Strawberry',\n",
              " 'Apricot',\n",
              " 'Pepper Red',\n",
              " 'Cucumber Ripe',\n",
              " 'Mango',\n",
              " 'Watermelon',\n",
              " 'Kiwi',\n",
              " 'Clementine',\n",
              " 'Cherry',\n",
              " 'Apple Granny Smith',\n",
              " 'Pomegranate',\n",
              " 'Raspberry',\n",
              " 'Pear',\n",
              " 'Tomato',\n",
              " 'Corn',\n",
              " 'Pineapple',\n",
              " 'Peach']"
            ]
          },
          "metadata": {},
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "#Pre-processing data dengan image augmentation."
      ],
      "metadata": {
        "id": "iBxqeWRlqiLj"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "train_datagen = ImageDataGenerator(\n",
        "                    rescale=1.0/255,\n",
        "                    rotation_range=30,\n",
        "                    width_shift_range=0.2,\n",
        "                    height_shift_range=0.2,\n",
        "                    shear_range=0.2,\n",
        "                    zoom_range=0.2,\n",
        "                    vertical_flip=True,\n",
        "                    fill_mode='nearest',   \n",
        "                    validation_split = 0.2)\n",
        "\n",
        "validation_datagen = ImageDataGenerator(\n",
        "    rescale=1.0/255\n",
        ")\n"
      ],
      "metadata": {
        "id": "jSioqQ1wdrtn"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "#Mempersiapkan data latih yang akan dipelajari oleh model."
      ],
      "metadata": {
        "id": "dCVIH2S3qmLL"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "train_generator = train_datagen.flow_from_directory(\n",
        "        dir_path,  # direktori data latih\n",
        "        class_mode='categorical',\n",
        "        target_size=(100, 100), \n",
        "        shuffle =True,\n",
        "        subset= 'training'\n",
        "        )\n",
        "\n",
        "validation_generator = train_datagen.flow_from_directory(\n",
        "        dir_path, # direktori data validasi\n",
        "        target_size=(100, 100), \n",
        "        class_mode='categorical',\n",
        "        subset= 'validation'\n",
        "        )  "
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "j4Jv7y5werz_",
        "outputId": "ab8ac86f-109d-439a-e67a-30b357ce8200"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Found 13493 images belonging to 33 classes.\n",
            "Found 3361 images belonging to 33 classes.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "#Membangun arsitektur model dengan Convolutional Neural Network (CNN)."
      ],
      "metadata": {
        "id": "R6qhZjBQqzZM"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "model = Sequential([\n",
        "    Conv2D(16, (3,3), activation='relu', input_shape=(100, 100, 3)),\n",
        "    MaxPooling2D(2, 2),\n",
        "    Conv2D(32, (3,3), activation='relu'),\n",
        "    MaxPooling2D(2,2),\n",
        "    Conv2D(64, (3,3), activation='relu'),\n",
        "    MaxPooling2D(2,2),\n",
        "    Conv2D(128, (3,3), activation='relu'),\n",
        "    MaxPooling2D(2,2),\n",
        "    Flatten(),\n",
        "    Dense(512, activation='relu'),\n",
        "    Dense(33, activation='softmax') \n",
        "])\n",
        "\n"
      ],
      "metadata": {
        "id": "Xiv1BbPreyh7"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "#Compile dan latih model dengan model.compile dan model.fit hingga mendapatkan akurasi yang diinginkan."
      ],
      "metadata": {
        "id": "uXDdDwhwrCqj"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "model.compile(loss='categorical_crossentropy', \n",
        "              optimizer='adam', \n",
        "              metrics=['accuracy'])"
      ],
      "metadata": {
        "id": "xZpxpPR9j6A8"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "#Callback"
      ],
      "metadata": {
        "id": "xnuTXp-ArD88"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "class myCallback(tf.keras.callbacks.Callback):\n",
        "  def on_epoch_end(self, epoch, logs={}):\n",
        "    if(logs.get('accuracy')>0.92):\n",
        "      print(\"\\nAkurasi telah mencapai >92%!\")\n",
        "      self.model.stop_training = True\n",
        "callbacks = myCallback()"
      ],
      "metadata": {
        "id": "xmdBpZnDfN8b"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "hist = model.fit(\n",
        "      train_generator,\n",
        "      steps_per_epoch=32,  # berapa batch yang akan dieksekusi pada setiap epoch\n",
        "      epochs=20,\n",
        "      validation_data=validation_generator, # menampilkan akurasi pengujian data validasi\n",
        "      validation_steps=8,  # berapa batch yang akan dieksekusi pada setiap epoch\n",
        "      verbose=2,\n",
        "      batch_size=64,\n",
        "      callbacks=[callbacks])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_6sVauIDfQXF",
        "outputId": "8d98c30a-9bf4-4c54-9f59-7aae37b2e7f4"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/20\n",
            "32/32 - 15s - loss: 3.3598 - accuracy: 0.0811 - val_loss: 2.6104 - val_accuracy: 0.2891 - 15s/epoch - 480ms/step\n",
            "Epoch 2/20\n",
            "32/32 - 14s - loss: 2.1307 - accuracy: 0.2891 - val_loss: 1.8820 - val_accuracy: 0.4062 - 14s/epoch - 445ms/step\n",
            "Epoch 3/20\n",
            "32/32 - 15s - loss: 1.5389 - accuracy: 0.4590 - val_loss: 1.2555 - val_accuracy: 0.5000 - 15s/epoch - 455ms/step\n",
            "Epoch 4/20\n",
            "32/32 - 15s - loss: 1.1785 - accuracy: 0.5801 - val_loss: 0.9988 - val_accuracy: 0.6445 - 15s/epoch - 467ms/step\n",
            "Epoch 5/20\n",
            "32/32 - 15s - loss: 0.9321 - accuracy: 0.6514 - val_loss: 0.7367 - val_accuracy: 0.6992 - 15s/epoch - 456ms/step\n",
            "Epoch 6/20\n",
            "32/32 - 15s - loss: 0.8223 - accuracy: 0.7002 - val_loss: 0.7424 - val_accuracy: 0.7344 - 15s/epoch - 462ms/step\n",
            "Epoch 7/20\n",
            "32/32 - 14s - loss: 0.6801 - accuracy: 0.7500 - val_loss: 0.5985 - val_accuracy: 0.8008 - 14s/epoch - 453ms/step\n",
            "Epoch 8/20\n",
            "32/32 - 15s - loss: 0.6044 - accuracy: 0.7920 - val_loss: 0.5188 - val_accuracy: 0.8242 - 15s/epoch - 465ms/step\n",
            "Epoch 9/20\n",
            "32/32 - 14s - loss: 0.5445 - accuracy: 0.8135 - val_loss: 0.5755 - val_accuracy: 0.8164 - 14s/epoch - 450ms/step\n",
            "Epoch 10/20\n",
            "32/32 - 14s - loss: 0.5370 - accuracy: 0.7920 - val_loss: 0.9393 - val_accuracy: 0.6875 - 14s/epoch - 450ms/step\n",
            "Epoch 11/20\n",
            "32/32 - 14s - loss: 0.6693 - accuracy: 0.7734 - val_loss: 0.3709 - val_accuracy: 0.8789 - 14s/epoch - 447ms/step\n",
            "Epoch 12/20\n",
            "32/32 - 14s - loss: 0.3911 - accuracy: 0.8564 - val_loss: 0.3299 - val_accuracy: 0.8828 - 14s/epoch - 443ms/step\n",
            "Epoch 13/20\n",
            "32/32 - 15s - loss: 0.3634 - accuracy: 0.8691 - val_loss: 0.3680 - val_accuracy: 0.8320 - 15s/epoch - 456ms/step\n",
            "Epoch 14/20\n",
            "32/32 - 15s - loss: 0.3163 - accuracy: 0.8838 - val_loss: 0.3562 - val_accuracy: 0.8711 - 15s/epoch - 454ms/step\n",
            "Epoch 15/20\n",
            "32/32 - 14s - loss: 0.3319 - accuracy: 0.8730 - val_loss: 0.3096 - val_accuracy: 0.8984 - 14s/epoch - 447ms/step\n",
            "Epoch 16/20\n",
            "32/32 - 14s - loss: 0.2964 - accuracy: 0.8945 - val_loss: 0.1870 - val_accuracy: 0.9180 - 14s/epoch - 444ms/step\n",
            "Epoch 17/20\n",
            "32/32 - 14s - loss: 0.2459 - accuracy: 0.9180 - val_loss: 0.2682 - val_accuracy: 0.8945 - 14s/epoch - 442ms/step\n",
            "Epoch 18/20\n",
            "\n",
            "Akurasi telah mencapai >92%!\n",
            "32/32 - 14s - loss: 0.2053 - accuracy: 0.9287 - val_loss: 0.3066 - val_accuracy: 0.8867 - 14s/epoch - 442ms/step\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "#Plot Akurasi dan Loss"
      ],
      "metadata": {
        "id": "cdDG7TMerShu"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Menampilkan plot akurasi pelatihan\n",
        "plt.plot(hist.history['accuracy'])\n",
        "plt.plot(hist.history['val_accuracy'])\n",
        "plt.title('Model Accuracy')\n",
        "plt.ylabel('Accuracy')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['Train', 'Validation'], loc='upper left')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "3fvH5cNzmQ1L",
        "outputId": "f48dfca0-4e75-4f13-96a9-c95806284555"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(hist.history['loss'])\n",
        "plt.plot(hist.history['val_loss'])\n",
        "plt.title('Model Loss')\n",
        "plt.ylabel('Loss')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['Train', 'Validation'], loc='upper right')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "GayTwwb_mSet",
        "outputId": "87d7caa4-2bae-4dda-9cac-d7e728111fbd"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "#Menyimpan model ke dalam format TF-Lite."
      ],
      "metadata": {
        "id": "lOKR-yEBrW3C"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Menyimpan model dalam format SavedModel\n",
        "export_dir = 'saved_model/'\n",
        "tf.saved_model.save(model, export_dir)\n",
        " \n",
        "# Convert SavedModel menjadi vegs.tflite\n",
        "converter = tf.lite.TFLiteConverter.from_saved_model(export_dir)\n",
        "tflite_model = converter.convert()\n",
        " \n",
        "tflite_model_file = pathlib.Path('vegs.tflite')\n",
        "tflite_model_file.write_bytes(tflite_model)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "fS3Y2iOAtkTn",
        "outputId": "4ce7f45c-9752-4937-9fca-3be5f2ad0f12"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "WARNING:absl:Found untraced functions such as _jit_compiled_convolution_op, _jit_compiled_convolution_op, _jit_compiled_convolution_op, _jit_compiled_convolution_op while saving (showing 4 of 4). These functions will not be directly callable after loading.\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "4658644"
            ]
          },
          "metadata": {},
          "execution_count": 64
        }
      ]
    }
  ]
}