AC3_AllAge_Unfreeze_500.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/Wanita-8943/Main_Project/blob/main/AC3_AllAge_Unfreeze_500.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ow7eWoNw6U-c"
      },
      "source": [
        "#เรียกใช้ CSV"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "id": "z8o_VVNXzcL8"
      },
      "outputs": [],
      "source": [
        "import pandas as pd\n",
        "import shutil"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {
        "id": "1_2Fe8u81d5r",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "5f9b4dd0-4917-4d99-c99d-946354d36a5d"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ],
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {
        "id": "mbLFqTO1ze9O",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 424
        },
        "outputId": "c374e703-005d-468a-ccce-7e26171f5231"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "      Fig_Age  Fig_Person_Sex  Age(year) Class  Class_Re       Filename  \\\n",
              "0           1               1          7   Y07         0         V1.jpg   \n",
              "1           2               1          7   Y07         0    Flip_V1.jpg   \n",
              "2           3               2          7   Y07         0         V2.jpg   \n",
              "3           4               2          7   Y07         0    Flip_V2.jpg   \n",
              "4           5               3          7   Y07         0         V3.jpg   \n",
              "...       ...             ...        ...   ...       ...            ...   \n",
              "4745      121              77         25   Y25        18  Flip_J463.jpg   \n",
              "4746      122              78         25   Y25        18       J464.jpg   \n",
              "4747      123              78         25   Y25        18  Flip_J464.jpg   \n",
              "4748      124              79         25   Y25        18       J465.jpg   \n",
              "4749      125              79         25   Y25        18  Flip_J465.jpg   \n",
              "\n",
              "                                          Path_filename     Sex Floder  \n",
              "0     /content/drive/My Drive/TVT_All_Age/train/Y07/...  Female   Both  \n",
              "1     /content/drive/My Drive/TVT_All_Age/train/Y07/...  Female   Both  \n",
              "2     /content/drive/My Drive/TVT_All_Age/train/Y07/...  Female   Both  \n",
              "3     /content/drive/My Drive/TVT_All_Age/train/Y07/...  Female   Both  \n",
              "4     /content/drive/My Drive/TVT_All_Age/train/Y07/...  Female   Both  \n",
              "...                                                 ...     ...    ...  \n",
              "4745  /content/drive/My Drive/TVT_All_Age/test/Y25/F...    Male   Both  \n",
              "4746  /content/drive/My Drive/TVT_All_Age/test/Y25/J...    Male   Both  \n",
              "4747  /content/drive/My Drive/TVT_All_Age/test/Y25/F...    Male   Both  \n",
              "4748  /content/drive/My Drive/TVT_All_Age/test/Y25/J...    Male   Both  \n",
              "4749  /content/drive/My Drive/TVT_All_Age/test/Y25/F...    Male   Both  \n",
              "\n",
              "[4750 rows x 9 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-6707796d-64ce-475a-99b7-d74faaf49dda\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Fig_Age</th>\n",
              "      <th>Fig_Person_Sex</th>\n",
              "      <th>Age(year)</th>\n",
              "      <th>Class</th>\n",
              "      <th>Class_Re</th>\n",
              "      <th>Filename</th>\n",
              "      <th>Path_filename</th>\n",
              "      <th>Sex</th>\n",
              "      <th>Floder</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>7</td>\n",
              "      <td>Y07</td>\n",
              "      <td>0</td>\n",
              "      <td>V1.jpg</td>\n",
              "      <td>/content/drive/My Drive/TVT_All_Age/train/Y07/...</td>\n",
              "      <td>Female</td>\n",
              "      <td>Both</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>7</td>\n",
              "      <td>Y07</td>\n",
              "      <td>0</td>\n",
              "      <td>Flip_V1.jpg</td>\n",
              "      <td>/content/drive/My Drive/TVT_All_Age/train/Y07/...</td>\n",
              "      <td>Female</td>\n",
              "      <td>Both</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>7</td>\n",
              "      <td>Y07</td>\n",
              "      <td>0</td>\n",
              "      <td>V2.jpg</td>\n",
              "      <td>/content/drive/My Drive/TVT_All_Age/train/Y07/...</td>\n",
              "      <td>Female</td>\n",
              "      <td>Both</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>7</td>\n",
              "      <td>Y07</td>\n",
              "      <td>0</td>\n",
              "      <td>Flip_V2.jpg</td>\n",
              "      <td>/content/drive/My Drive/TVT_All_Age/train/Y07/...</td>\n",
              "      <td>Female</td>\n",
              "      <td>Both</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>5</td>\n",
              "      <td>3</td>\n",
              "      <td>7</td>\n",
              "      <td>Y07</td>\n",
              "      <td>0</td>\n",
              "      <td>V3.jpg</td>\n",
              "      <td>/content/drive/My Drive/TVT_All_Age/train/Y07/...</td>\n",
              "      <td>Female</td>\n",
              "      <td>Both</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4745</th>\n",
              "      <td>121</td>\n",
              "      <td>77</td>\n",
              "      <td>25</td>\n",
              "      <td>Y25</td>\n",
              "      <td>18</td>\n",
              "      <td>Flip_J463.jpg</td>\n",
              "      <td>/content/drive/My Drive/TVT_All_Age/test/Y25/F...</td>\n",
              "      <td>Male</td>\n",
              "      <td>Both</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4746</th>\n",
              "      <td>122</td>\n",
              "      <td>78</td>\n",
              "      <td>25</td>\n",
              "      <td>Y25</td>\n",
              "      <td>18</td>\n",
              "      <td>J464.jpg</td>\n",
              "      <td>/content/drive/My Drive/TVT_All_Age/test/Y25/J...</td>\n",
              "      <td>Male</td>\n",
              "      <td>Both</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4747</th>\n",
              "      <td>123</td>\n",
              "      <td>78</td>\n",
              "      <td>25</td>\n",
              "      <td>Y25</td>\n",
              "      <td>18</td>\n",
              "      <td>Flip_J464.jpg</td>\n",
              "      <td>/content/drive/My Drive/TVT_All_Age/test/Y25/F...</td>\n",
              "      <td>Male</td>\n",
              "      <td>Both</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4748</th>\n",
              "      <td>124</td>\n",
              "      <td>79</td>\n",
              "      <td>25</td>\n",
              "      <td>Y25</td>\n",
              "      <td>18</td>\n",
              "      <td>J465.jpg</td>\n",
              "      <td>/content/drive/My Drive/TVT_All_Age/test/Y25/J...</td>\n",
              "      <td>Male</td>\n",
              "      <td>Both</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4749</th>\n",
              "      <td>125</td>\n",
              "      <td>79</td>\n",
              "      <td>25</td>\n",
              "      <td>Y25</td>\n",
              "      <td>18</td>\n",
              "      <td>Flip_J465.jpg</td>\n",
              "      <td>/content/drive/My Drive/TVT_All_Age/test/Y25/F...</td>\n",
              "      <td>Male</td>\n",
              "      <td>Both</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>4750 rows × 9 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-6707796d-64ce-475a-99b7-d74faaf49dda')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-6707796d-64ce-475a-99b7-d74faaf49dda button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-6707796d-64ce-475a-99b7-d74faaf49dda');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 3
        }
      ],
      "source": [
        "df = pd.read_csv (r'/content/drive/MyDrive/cut_panoramic/Data/New_All_Age_0.csv')\n",
        "df"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5qxePnnn7TGW"
      },
      "source": [
        "#Imports"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {
        "id": "D-hCRloc3t39"
      },
      "outputs": [],
      "source": [
        "from tensorflow.keras import models\n",
        "from tensorflow.keras import layers\n",
        "from tensorflow.keras import optimizers\n",
        "import os\n",
        "import glob\n",
        "import shutil\n",
        "import sys\n",
        "import numpy as np\n",
        "from skimage.io import imread\n",
        "import matplotlib.pyplot as plt\n",
        "from IPython.display import Image\n",
        "%matplotlib inline"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RooqSdBc7QHC"
      },
      "source": [
        "#กำหนดค่าพารามิเตอร์\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {
        "id": "thDb7U9B3xOo"
      },
      "outputs": [],
      "source": [
        "batch_size = 16\n",
        "width = 150\n",
        "height = 150\n",
        "epochs = 500\n",
        "NUM_TRAIN = 2850\n",
        "NUM_TEST = 950\n",
        "dropout_rate = 0.2\n",
        "input_shape = (height, width, 3)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pumGmy6f3eSW"
      },
      "source": [
        "#Clone efficientnet repo\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "P7iy2f8n16p0",
        "outputId": "de2d263a-484f-4328-9924-a2cb41681782"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "/content\n",
            "Cloning into 'efficientnet_keras_transfer_learning'...\n",
            "remote: Enumerating objects: 1079, done.\u001b[K\n",
            "remote: Counting objects: 100% (242/242), done.\u001b[K\n",
            "remote: Compressing objects: 100% (121/121), done.\u001b[K\n",
            "remote: Total 1079 (delta 121), reused 241 (delta 121), pack-reused 837\u001b[K\n",
            "Receiving objects: 100% (1079/1079), 13.94 MiB | 33.36 MiB/s, done.\n",
            "Resolving deltas: 100% (618/618), done.\n",
            "/content/efficientnet_keras_transfer_learning\n"
          ]
        }
      ],
      "source": [
        "#ดึงข้อมูลใน Github มาใช้\n",
        "import os\n",
        "%cd /content\n",
        "if not os.path.isdir(\"efficientnet_keras_transfer_learning\"):\n",
        " !git clone https://github.com/Wanita-8943/efficientnet_keras_transfer_learning\n",
        "%cd efficientnet_keras_transfer_learning/\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {
        "id": "Gqg_EUxrKkcK"
      },
      "outputs": [],
      "source": [
        "# Options: EfficientNetB0, EfficientNetB1, EfficientNetB2, EfficientNetB3\n",
        "# Higher the number, the more complex the model is.\n",
        "from efficientnet import EfficientNetB0 as Net\n",
        "from efficientnet import center_crop_and_resize, preprocess_input"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# loading pretrained conv base model\n",
        "# โหลดโมเดล มาโดยตัด output ของโมเดลออก เเต่ยังใช้ input อันเดิม\n",
        "# เเละโหลด weight ของโมเดล มาด้วยที่ชื่อว่า imagenet\n",
        "conv_base = Net(weights='imagenet', include_top=False, input_shape=input_shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "sjq4AxnfPMPF",
        "outputId": "a9a1ca9b-48fb-47e6-ac7e-f5280824078e"
      },
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "WARNING:tensorflow:From /usr/local/lib/python3.9/dist-packages/tensorflow/python/autograph/pyct/static_analysis/liveness.py:83: Analyzer.lamba_check (from tensorflow.python.autograph.pyct.static_analysis.liveness) is deprecated and will be removed after 2023-09-23.\n",
            "Instructions for updating:\n",
            "Lambda fuctions will be no more assumed to be used in the statement where they are used, or at least in the same block. https://github.com/tensorflow/tensorflow/issues/56089\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading data from https://github.com/qubvel/efficientnet/releases/download/v0.0.1/efficientnet-b0_imagenet_1000_notop.h5\n",
            "16717576/16717576 [==============================] - 0s 0us/step\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "#load model \n"
      ],
      "metadata": {
        "id": "Jqlg5Z7X5MlQ"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "import sys\n",
        "sys.path.append('/content/drive/MyDrive/cut_panoramic/Model/Classification/All_Age/AC2_AllAge_Freeze_250.h5')\n",
        "\n",
        "from efficientnet.layers import Swish, DropConnect\n",
        "from efficientnet.model import ConvKernalInitializer\n",
        "from tensorflow.keras.utils import get_custom_objects\n",
        "\n",
        "get_custom_objects().update({\n",
        "    'ConvKernalInitializer': ConvKernalInitializer,\n",
        "    'Swish': Swish,\n",
        "    'DropConnect':DropConnect\n",
        "})\n",
        "    "
      ],
      "metadata": {
        "id": "CS9-uI0G5JDO"
      },
      "execution_count": 9,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#load model \n",
        "from tensorflow.keras.models import load_model\n",
        "model = load_model('/content/drive/MyDrive/cut_panoramic/Model/Classification/All_Age/AC2_AllAge_Freeze_250.h5')\n",
        "height = width = model.input_shape[1]"
      ],
      "metadata": {
        "id": "X8ZC22Qe5J4Q"
      },
      "execution_count": 10,
      "outputs": []
    },
    {
      "cell_type": "code",
      "execution_count": 11,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "NadBB12251jh",
        "outputId": "1c08ffb0-a39c-42bf-f46b-97ac48f2db3b"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " efficientnet-b0 (Functional  (None, 5, 5, 1280)       4049564   \n",
            " )                                                               \n",
            "                                                                 \n",
            " gap (GlobalMaxPooling2D)    (None, 1280)              0         \n",
            "                                                                 \n",
            " dropout_out (Dropout)       (None, 1280)              0         \n",
            "                                                                 \n",
            " fc_out (Dense)              (None, 19)                24339     \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 4,073,903\n",
            "Trainable params: 24,339\n",
            "Non-trainable params: 4,049,564\n",
            "_________________________________________________________________\n"
          ]
        }
      ],
      "source": [
        "model.summary()"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# multiply_16\n",
        "# set 'multiply_16' and following layers trainable\n",
        "model.trainable = True\n",
        "\n",
        "set_trainable = False\n",
        "for layer in conv_base.layers:\n",
        "    if layer.name == 'multiply_16':\n",
        "        set_trainable = True\n",
        "    if set_trainable:\n",
        "        layer.trainable = True\n",
        "    else:\n",
        "        layer.trainable = False  "
      ],
      "metadata": {
        "id": "lBYpecYxPfq5"
      },
      "execution_count": 12,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "QpOX8Bk0Pjt1",
        "outputId": "4e5825f9-248b-4391-df55-4f6f1f673906"
      },
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " efficientnet-b0 (Functional  (None, 5, 5, 1280)       4049564   \n",
            " )                                                               \n",
            "                                                                 \n",
            " gap (GlobalMaxPooling2D)    (None, 1280)              0         \n",
            "                                                                 \n",
            " dropout_out (Dropout)       (None, 1280)              0         \n",
            "                                                                 \n",
            " fc_out (Dense)              (None, 19)                24339     \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 4,073,903\n",
            "Trainable params: 4,031,887\n",
            "Non-trainable params: 42,016\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "J36J9EAE7qSB"
      },
      "source": [
        "#สร้างโฟลเดอร์ Train Valodation และ Test"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 14,
      "metadata": {
        "id": "nqCFbjRQ3okB"
      },
      "outputs": [],
      "source": [
        "base_dir = '/content/drive/MyDrive/TVT_All_Age'\n",
        "os.makedirs(base_dir, exist_ok=True)\n",
        "\n",
        "# Directories for our training,\n",
        "# validation and test splits\n",
        "train_dir = os.path.join(base_dir, 'train')\n",
        "os.makedirs(train_dir, exist_ok=True)\n",
        "validation_dir = os.path.join(base_dir, 'validation')\n",
        "os.makedirs(validation_dir, exist_ok=True)\n",
        "test_dir = os.path.join(base_dir, 'test')\n",
        "os.makedirs(test_dir, exist_ok=True)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bWEnlTSwazL5"
      },
      "source": [
        "\n",
        "#Train"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 15,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "xGPrsn9no_pa",
        "outputId": "04ea8c56-c13b-466d-b8d2-f20495ffb59a"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Found 2850 images belonging to 19 classes.\n",
            "Found 950 images belonging to 19 classes.\n"
          ]
        }
      ],
      "source": [
        "# Train ด้วย ImageDataGenerator ของ Keras ซึ่งจะเพิ่มข้อมูลเสริมระหว่างการฝึกเพื่อลดโอกาสเกิด overfitting\n",
        "#overfitting เกิดจากข้อมูลที่ซับซ้อนกันเกินไป\n",
        "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
        "\n",
        "train_datagen = ImageDataGenerator(\n",
        "      rescale=1./255, #โมเดลส่วนใหญ่ต้องใช้ RGB ในช่วง 0–1\n",
        "      rotation_range=40,\n",
        "      width_shift_range=0.2,\n",
        "      height_shift_range=0.2,\n",
        "      shear_range=0.2,\n",
        "      zoom_range=0.2,\n",
        "      horizontal_flip=True,\n",
        "      fill_mode='nearest')\n",
        "\n",
        "# Note that the validation data should not be augmented!\n",
        "test_datagen = ImageDataGenerator(rescale=1./255)\n",
        "\n",
        "train_generator = train_datagen.flow_from_directory(\n",
        "        # This is the target directory #ไดเรกเป้าหมาย\n",
        "        train_dir,\n",
        "        # รูปภาพทั้งหมดจะถูกปรับขนาดตามความสูงและความกว้างของเป้าหมาย\n",
        "        target_size=(height, width),\n",
        "        batch_size=batch_size,\n",
        "        # Since we use categorical_crossentropy loss, we need categorical labels\n",
        "        #เนื่องจากเราใช้ categorical_crossentropy loss เราจึงต้องมีป้ายกำกับตามหมวดหมู่\n",
        "        class_mode='categorical')\n",
        "\n",
        "validation_generator = test_datagen.flow_from_directory( #การดึงภาพจาก Directory มาเข้าโมเดล \n",
        "        validation_dir,\n",
        "        target_size=(height, width),\n",
        "        batch_size=batch_size,\n",
        "        class_mode='categorical')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 16,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "N6qUmmF856ZE",
        "outputId": "09859a31-c7d9-408b-c9c1-821ebb80376c"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "<ipython-input-16-caa7b37242a8>:1: UserWarning: `Model.fit_generator` is deprecated and will be removed in a future version. Please use `Model.fit`, which supports generators.\n",
            "  history = model.fit_generator(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/250\n",
            "178/178 [==============================] - 371s 2s/step - loss: 2.5983 - acc: 0.2417 - val_loss: 4.1094 - val_acc: 0.1917\n",
            "Epoch 2/250\n",
            "178/178 [==============================] - 25s 135ms/step - loss: 2.6036 - acc: 0.2459 - val_loss: 4.1441 - val_acc: 0.1949\n",
            "Epoch 3/250\n",
            "178/178 [==============================] - 25s 138ms/step - loss: 2.5874 - acc: 0.2463 - val_loss: 4.3000 - val_acc: 0.1674\n",
            "Epoch 4/250\n",
            "178/178 [==============================] - 24s 135ms/step - loss: 2.5934 - acc: 0.2392 - val_loss: 4.3266 - val_acc: 0.1822\n",
            "Epoch 5/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.6586 - acc: 0.2371 - val_loss: 4.1449 - val_acc: 0.1875\n",
            "Epoch 6/250\n",
            "178/178 [==============================] - 19s 107ms/step - loss: 2.5995 - acc: 0.2410 - val_loss: 4.1604 - val_acc: 0.1886\n",
            "Epoch 7/250\n",
            "178/178 [==============================] - 25s 136ms/step - loss: 2.6414 - acc: 0.2329 - val_loss: 4.1480 - val_acc: 0.1907\n",
            "Epoch 8/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.6312 - acc: 0.2530 - val_loss: 4.1328 - val_acc: 0.1949\n",
            "Epoch 9/250\n",
            "178/178 [==============================] - 19s 107ms/step - loss: 2.6274 - acc: 0.2332 - val_loss: 4.1348 - val_acc: 0.1864\n",
            "Epoch 10/250\n",
            "178/178 [==============================] - 25s 139ms/step - loss: 2.5959 - acc: 0.2329 - val_loss: 4.1850 - val_acc: 0.1758\n",
            "Epoch 11/250\n",
            "178/178 [==============================] - 24s 135ms/step - loss: 2.5986 - acc: 0.2421 - val_loss: 4.2859 - val_acc: 0.1896\n",
            "Epoch 12/250\n",
            "178/178 [==============================] - 25s 136ms/step - loss: 2.6776 - acc: 0.2347 - val_loss: 4.2101 - val_acc: 0.1843\n",
            "Epoch 13/250\n",
            "178/178 [==============================] - 24s 135ms/step - loss: 2.6741 - acc: 0.2272 - val_loss: 4.1938 - val_acc: 0.1939\n",
            "Epoch 14/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.6201 - acc: 0.2350 - val_loss: 4.1917 - val_acc: 0.1758\n",
            "Epoch 15/250\n",
            "178/178 [==============================] - 25s 139ms/step - loss: 2.5648 - acc: 0.2442 - val_loss: 4.2037 - val_acc: 0.1833\n",
            "Epoch 16/250\n",
            "178/178 [==============================] - 25s 139ms/step - loss: 2.5976 - acc: 0.2385 - val_loss: 4.3472 - val_acc: 0.1864\n",
            "Epoch 17/250\n",
            "178/178 [==============================] - 24s 135ms/step - loss: 2.6461 - acc: 0.2378 - val_loss: 4.1817 - val_acc: 0.1992\n",
            "Epoch 18/250\n",
            "178/178 [==============================] - 24s 135ms/step - loss: 2.6751 - acc: 0.2428 - val_loss: 4.2505 - val_acc: 0.2002\n",
            "Epoch 19/250\n",
            "178/178 [==============================] - 25s 138ms/step - loss: 2.6925 - acc: 0.2385 - val_loss: 4.2910 - val_acc: 0.1843\n",
            "Epoch 20/250\n",
            "178/178 [==============================] - 25s 136ms/step - loss: 2.6365 - acc: 0.2417 - val_loss: 4.4337 - val_acc: 0.1801\n",
            "Epoch 21/250\n",
            "178/178 [==============================] - 25s 139ms/step - loss: 2.6414 - acc: 0.2325 - val_loss: 4.1922 - val_acc: 0.1811\n",
            "Epoch 22/250\n",
            "178/178 [==============================] - 25s 136ms/step - loss: 2.6569 - acc: 0.2428 - val_loss: 4.2220 - val_acc: 0.2023\n",
            "Epoch 23/250\n",
            "178/178 [==============================] - 25s 136ms/step - loss: 2.7612 - acc: 0.2227 - val_loss: 4.4432 - val_acc: 0.1907\n",
            "Epoch 24/250\n",
            "178/178 [==============================] - 25s 139ms/step - loss: 2.7032 - acc: 0.2354 - val_loss: 4.2059 - val_acc: 0.1790\n",
            "Epoch 25/250\n",
            "178/178 [==============================] - 25s 140ms/step - loss: 2.5791 - acc: 0.2445 - val_loss: 4.4713 - val_acc: 0.1748\n",
            "Epoch 26/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.6318 - acc: 0.2410 - val_loss: 4.4537 - val_acc: 0.1854\n",
            "Epoch 27/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.6757 - acc: 0.2410 - val_loss: 4.2289 - val_acc: 0.1833\n",
            "Epoch 28/250\n",
            "178/178 [==============================] - 24s 135ms/step - loss: 2.6640 - acc: 0.2414 - val_loss: 4.0620 - val_acc: 0.1833\n",
            "Epoch 29/250\n",
            "178/178 [==============================] - 25s 136ms/step - loss: 2.6514 - acc: 0.2470 - val_loss: 4.1479 - val_acc: 0.1674\n",
            "Epoch 30/250\n",
            "178/178 [==============================] - 26s 142ms/step - loss: 2.6705 - acc: 0.2177 - val_loss: 4.1902 - val_acc: 0.1716\n",
            "Epoch 31/250\n",
            "178/178 [==============================] - 25s 141ms/step - loss: 2.6566 - acc: 0.2385 - val_loss: 4.3134 - val_acc: 0.1801\n",
            "Epoch 32/250\n",
            "178/178 [==============================] - 25s 136ms/step - loss: 2.6530 - acc: 0.2484 - val_loss: 4.2814 - val_acc: 0.1716\n",
            "Epoch 33/250\n",
            "178/178 [==============================] - 25s 136ms/step - loss: 2.5814 - acc: 0.2452 - val_loss: 4.2421 - val_acc: 0.1610\n",
            "Epoch 34/250\n",
            "178/178 [==============================] - 25s 136ms/step - loss: 2.6800 - acc: 0.2318 - val_loss: 4.0206 - val_acc: 0.1833\n",
            "Epoch 35/250\n",
            "178/178 [==============================] - 25s 138ms/step - loss: 2.7295 - acc: 0.2410 - val_loss: 4.0478 - val_acc: 0.1833\n",
            "Epoch 36/250\n",
            "178/178 [==============================] - 25s 137ms/step - loss: 2.6875 - acc: 0.2308 - val_loss: 4.1520 - val_acc: 0.1886\n",
            "Epoch 37/250\n",
            "178/178 [==============================] - 26s 142ms/step - loss: 2.6491 - acc: 0.2477 - val_loss: 4.2267 - val_acc: 0.1642\n",
            "Epoch 38/250\n",
            "178/178 [==============================] - 25s 141ms/step - loss: 2.6802 - acc: 0.2347 - val_loss: 4.0806 - val_acc: 0.1907\n",
            "Epoch 39/250\n",
            "178/178 [==============================] - 24s 135ms/step - loss: 2.6304 - acc: 0.2452 - val_loss: 4.1564 - val_acc: 0.1928\n",
            "Epoch 40/250\n",
            "178/178 [==============================] - 25s 137ms/step - loss: 2.6860 - acc: 0.2347 - val_loss: 4.1289 - val_acc: 0.1663\n",
            "Epoch 41/250\n",
            "178/178 [==============================] - 25s 139ms/step - loss: 2.6948 - acc: 0.2294 - val_loss: 4.2296 - val_acc: 0.1695\n",
            "Epoch 42/250\n",
            "178/178 [==============================] - 25s 138ms/step - loss: 2.6357 - acc: 0.2371 - val_loss: 4.3629 - val_acc: 0.1801\n",
            "Epoch 43/250\n",
            "178/178 [==============================] - 25s 140ms/step - loss: 2.6804 - acc: 0.2339 - val_loss: 4.1663 - val_acc: 0.1758\n",
            "Epoch 44/250\n",
            "178/178 [==============================] - 26s 143ms/step - loss: 2.6879 - acc: 0.2350 - val_loss: 4.3347 - val_acc: 0.1896\n",
            "Epoch 45/250\n",
            "178/178 [==============================] - 26s 142ms/step - loss: 2.6425 - acc: 0.2537 - val_loss: 4.4456 - val_acc: 0.1621\n",
            "Epoch 46/250\n",
            "178/178 [==============================] - 25s 138ms/step - loss: 2.6991 - acc: 0.2318 - val_loss: 4.1263 - val_acc: 0.1843\n",
            "Epoch 47/250\n",
            "178/178 [==============================] - 25s 139ms/step - loss: 2.6952 - acc: 0.2322 - val_loss: 4.2611 - val_acc: 0.1748\n",
            "Epoch 48/250\n",
            "178/178 [==============================] - 25s 139ms/step - loss: 2.6912 - acc: 0.2399 - val_loss: 4.2432 - val_acc: 0.1843\n",
            "Epoch 49/250\n",
            "178/178 [==============================] - 25s 136ms/step - loss: 2.6824 - acc: 0.2459 - val_loss: 4.0864 - val_acc: 0.1970\n",
            "Epoch 50/250\n",
            "178/178 [==============================] - 25s 136ms/step - loss: 2.6883 - acc: 0.2297 - val_loss: 4.2118 - val_acc: 0.1822\n",
            "Epoch 51/250\n",
            "178/178 [==============================] - 25s 140ms/step - loss: 2.6284 - acc: 0.2368 - val_loss: 4.5588 - val_acc: 0.1790\n",
            "Epoch 52/250\n",
            "178/178 [==============================] - 25s 139ms/step - loss: 2.6962 - acc: 0.2304 - val_loss: 4.2487 - val_acc: 0.1716\n",
            "Epoch 53/250\n",
            "178/178 [==============================] - 24s 135ms/step - loss: 2.7253 - acc: 0.2371 - val_loss: 4.1266 - val_acc: 0.1864\n",
            "Epoch 54/250\n",
            "178/178 [==============================] - 24s 135ms/step - loss: 2.6709 - acc: 0.2318 - val_loss: 4.3739 - val_acc: 0.1928\n",
            "Epoch 55/250\n",
            "178/178 [==============================] - 24s 135ms/step - loss: 2.7403 - acc: 0.2428 - val_loss: 4.5700 - val_acc: 0.1631\n",
            "Epoch 56/250\n",
            "178/178 [==============================] - 25s 136ms/step - loss: 2.6866 - acc: 0.2435 - val_loss: 4.1875 - val_acc: 0.1811\n",
            "Epoch 57/250\n",
            "178/178 [==============================] - 24s 135ms/step - loss: 2.7127 - acc: 0.2361 - val_loss: 4.2510 - val_acc: 0.1960\n",
            "Epoch 58/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.6970 - acc: 0.2315 - val_loss: 4.0919 - val_acc: 0.1833\n",
            "Epoch 59/250\n",
            "178/178 [==============================] - 25s 136ms/step - loss: 2.6832 - acc: 0.2258 - val_loss: 4.3770 - val_acc: 0.1833\n",
            "Epoch 60/250\n",
            "178/178 [==============================] - 25s 139ms/step - loss: 2.7725 - acc: 0.2347 - val_loss: 4.1681 - val_acc: 0.1706\n",
            "Epoch 61/250\n",
            "178/178 [==============================] - 25s 140ms/step - loss: 2.6630 - acc: 0.2311 - val_loss: 4.2797 - val_acc: 0.1843\n",
            "Epoch 62/250\n",
            "178/178 [==============================] - 20s 109ms/step - loss: 2.7192 - acc: 0.2209 - val_loss: 4.3611 - val_acc: 0.1780\n",
            "Epoch 63/250\n",
            "178/178 [==============================] - 25s 137ms/step - loss: 2.8514 - acc: 0.2336 - val_loss: 4.3227 - val_acc: 0.1727\n",
            "Epoch 64/250\n",
            "178/178 [==============================] - 21s 113ms/step - loss: 2.6308 - acc: 0.2339 - val_loss: 4.1677 - val_acc: 0.1907\n",
            "Epoch 65/250\n",
            "178/178 [==============================] - 25s 139ms/step - loss: 2.6843 - acc: 0.2304 - val_loss: 4.4109 - val_acc: 0.1748\n",
            "Epoch 66/250\n",
            "178/178 [==============================] - 25s 140ms/step - loss: 2.7022 - acc: 0.2442 - val_loss: 4.2541 - val_acc: 0.1727\n",
            "Epoch 67/250\n",
            "178/178 [==============================] - 24s 133ms/step - loss: 2.6185 - acc: 0.2414 - val_loss: 4.4927 - val_acc: 0.1737\n",
            "Epoch 68/250\n",
            "178/178 [==============================] - 21s 114ms/step - loss: 2.6843 - acc: 0.2414 - val_loss: 4.3053 - val_acc: 0.1843\n",
            "Epoch 69/250\n",
            "178/178 [==============================] - 26s 145ms/step - loss: 2.7567 - acc: 0.2248 - val_loss: 4.1155 - val_acc: 0.1790\n",
            "Epoch 70/250\n",
            "178/178 [==============================] - 26s 141ms/step - loss: 2.7362 - acc: 0.2392 - val_loss: 4.3200 - val_acc: 0.1822\n",
            "Epoch 71/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.6735 - acc: 0.2350 - val_loss: 4.2188 - val_acc: 0.1716\n",
            "Epoch 72/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.7022 - acc: 0.2459 - val_loss: 4.3546 - val_acc: 0.1822\n",
            "Epoch 73/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.6999 - acc: 0.2495 - val_loss: 4.1990 - val_acc: 0.1811\n",
            "Epoch 74/250\n",
            "178/178 [==============================] - 21s 113ms/step - loss: 2.6617 - acc: 0.2445 - val_loss: 4.4922 - val_acc: 0.1833\n",
            "Epoch 75/250\n",
            "178/178 [==============================] - 25s 138ms/step - loss: 2.6716 - acc: 0.2586 - val_loss: 4.2536 - val_acc: 0.1621\n",
            "Epoch 76/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.6757 - acc: 0.2318 - val_loss: 4.2714 - val_acc: 0.1843\n",
            "Epoch 77/250\n",
            "178/178 [==============================] - 24s 133ms/step - loss: 2.7051 - acc: 0.2354 - val_loss: 4.3949 - val_acc: 0.1748\n",
            "Epoch 78/250\n",
            "178/178 [==============================] - 24s 133ms/step - loss: 2.7607 - acc: 0.2375 - val_loss: 4.4151 - val_acc: 0.1811\n",
            "Epoch 79/250\n",
            "178/178 [==============================] - 25s 136ms/step - loss: 2.7614 - acc: 0.2470 - val_loss: 4.3155 - val_acc: 0.1716\n",
            "Epoch 80/250\n",
            "178/178 [==============================] - 25s 140ms/step - loss: 2.7428 - acc: 0.2248 - val_loss: 4.3544 - val_acc: 0.1811\n",
            "Epoch 81/250\n",
            "178/178 [==============================] - 25s 136ms/step - loss: 2.7375 - acc: 0.2329 - val_loss: 4.3822 - val_acc: 0.1737\n",
            "Epoch 82/250\n",
            "178/178 [==============================] - 25s 136ms/step - loss: 2.6527 - acc: 0.2399 - val_loss: 4.2149 - val_acc: 0.1769\n",
            "Epoch 83/250\n",
            "178/178 [==============================] - 20s 111ms/step - loss: 2.6850 - acc: 0.2399 - val_loss: 4.4667 - val_acc: 0.1674\n",
            "Epoch 84/250\n",
            "178/178 [==============================] - 25s 141ms/step - loss: 2.7195 - acc: 0.2287 - val_loss: 4.3570 - val_acc: 0.1706\n",
            "Epoch 85/250\n",
            "178/178 [==============================] - 25s 137ms/step - loss: 2.6940 - acc: 0.2428 - val_loss: 4.5084 - val_acc: 0.1769\n",
            "Epoch 86/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.7111 - acc: 0.2378 - val_loss: 4.4599 - val_acc: 0.1790\n",
            "Epoch 87/250\n",
            "178/178 [==============================] - 24s 135ms/step - loss: 2.7919 - acc: 0.2463 - val_loss: 4.4404 - val_acc: 0.1727\n",
            "Epoch 88/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.6656 - acc: 0.2470 - val_loss: 4.5093 - val_acc: 0.1748\n",
            "Epoch 89/250\n",
            "178/178 [==============================] - 25s 137ms/step - loss: 2.7357 - acc: 0.2361 - val_loss: 4.5347 - val_acc: 0.1642\n",
            "Epoch 90/250\n",
            "178/178 [==============================] - 25s 140ms/step - loss: 2.7813 - acc: 0.2456 - val_loss: 4.4046 - val_acc: 0.1748\n",
            "Epoch 91/250\n",
            "178/178 [==============================] - 25s 137ms/step - loss: 2.7131 - acc: 0.2484 - val_loss: 4.3493 - val_acc: 0.1833\n",
            "Epoch 92/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.6792 - acc: 0.2449 - val_loss: 4.6612 - val_acc: 0.1864\n",
            "Epoch 93/250\n",
            "178/178 [==============================] - 25s 136ms/step - loss: 2.7207 - acc: 0.2459 - val_loss: 4.4398 - val_acc: 0.1790\n",
            "Epoch 94/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.6312 - acc: 0.2544 - val_loss: 4.5523 - val_acc: 0.1706\n",
            "Epoch 95/250\n",
            "178/178 [==============================] - 25s 137ms/step - loss: 2.7243 - acc: 0.2431 - val_loss: 4.6639 - val_acc: 0.1610\n",
            "Epoch 96/250\n",
            "178/178 [==============================] - 25s 138ms/step - loss: 2.8105 - acc: 0.2290 - val_loss: 4.7165 - val_acc: 0.1811\n",
            "Epoch 97/250\n",
            "178/178 [==============================] - 25s 138ms/step - loss: 2.7059 - acc: 0.2361 - val_loss: 4.6161 - val_acc: 0.1811\n",
            "Epoch 98/250\n",
            "178/178 [==============================] - 25s 137ms/step - loss: 2.7603 - acc: 0.2262 - val_loss: 4.3734 - val_acc: 0.1907\n",
            "Epoch 99/250\n",
            "178/178 [==============================] - 25s 137ms/step - loss: 2.8164 - acc: 0.2396 - val_loss: 4.2761 - val_acc: 0.1706\n",
            "Epoch 100/250\n",
            "178/178 [==============================] - 24s 135ms/step - loss: 2.7633 - acc: 0.2361 - val_loss: 4.3530 - val_acc: 0.1981\n",
            "Epoch 101/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.7479 - acc: 0.2339 - val_loss: 4.4350 - val_acc: 0.1875\n",
            "Epoch 102/250\n",
            "178/178 [==============================] - 25s 140ms/step - loss: 2.6539 - acc: 0.2364 - val_loss: 4.3521 - val_acc: 0.1706\n",
            "Epoch 103/250\n",
            "178/178 [==============================] - 26s 142ms/step - loss: 2.6960 - acc: 0.2449 - val_loss: 4.5417 - val_acc: 0.1716\n",
            "Epoch 104/250\n",
            "178/178 [==============================] - 25s 139ms/step - loss: 2.7274 - acc: 0.2431 - val_loss: 4.2263 - val_acc: 0.1780\n",
            "Epoch 105/250\n",
            "178/178 [==============================] - 25s 138ms/step - loss: 2.8284 - acc: 0.2375 - val_loss: 4.4781 - val_acc: 0.1653\n",
            "Epoch 106/250\n",
            "178/178 [==============================] - 25s 136ms/step - loss: 2.7257 - acc: 0.2368 - val_loss: 4.7533 - val_acc: 0.1716\n",
            "Epoch 107/250\n",
            "178/178 [==============================] - 25s 136ms/step - loss: 2.7838 - acc: 0.2311 - val_loss: 4.3071 - val_acc: 0.1843\n",
            "Epoch 108/250\n",
            "178/178 [==============================] - 25s 137ms/step - loss: 2.7608 - acc: 0.2336 - val_loss: 4.3438 - val_acc: 0.1727\n",
            "Epoch 109/250\n",
            "178/178 [==============================] - 24s 135ms/step - loss: 2.6814 - acc: 0.2392 - val_loss: 4.3332 - val_acc: 0.1896\n",
            "Epoch 110/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.7400 - acc: 0.2371 - val_loss: 4.3381 - val_acc: 0.1748\n",
            "Epoch 111/250\n",
            "178/178 [==============================] - 25s 139ms/step - loss: 2.6562 - acc: 0.2452 - val_loss: 4.5321 - val_acc: 0.1769\n",
            "Epoch 112/250\n",
            "178/178 [==============================] - 25s 141ms/step - loss: 2.7078 - acc: 0.2435 - val_loss: 4.4758 - val_acc: 0.1684\n",
            "Epoch 113/250\n",
            "178/178 [==============================] - 24s 135ms/step - loss: 2.7465 - acc: 0.2279 - val_loss: 4.5809 - val_acc: 0.1716\n",
            "Epoch 114/250\n",
            "178/178 [==============================] - 24s 135ms/step - loss: 2.7949 - acc: 0.2417 - val_loss: 4.4397 - val_acc: 0.1695\n",
            "Epoch 115/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.8481 - acc: 0.2339 - val_loss: 4.4313 - val_acc: 0.1748\n",
            "Epoch 116/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.7637 - acc: 0.2368 - val_loss: 4.6091 - val_acc: 0.1727\n",
            "Epoch 117/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.6942 - acc: 0.2296 - val_loss: 4.8464 - val_acc: 0.1674\n",
            "Epoch 118/250\n",
            "178/178 [==============================] - 24s 133ms/step - loss: 2.6985 - acc: 0.2343 - val_loss: 4.6626 - val_acc: 0.1684\n",
            "Epoch 119/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.7569 - acc: 0.2385 - val_loss: 4.7674 - val_acc: 0.1748\n",
            "Epoch 120/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.6994 - acc: 0.2470 - val_loss: 4.5899 - val_acc: 0.1716\n",
            "Epoch 121/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.7217 - acc: 0.2438 - val_loss: 4.5323 - val_acc: 0.1621\n",
            "Epoch 122/250\n",
            "178/178 [==============================] - 25s 137ms/step - loss: 2.6412 - acc: 0.2428 - val_loss: 4.6951 - val_acc: 0.1653\n",
            "Epoch 123/250\n",
            "178/178 [==============================] - 26s 141ms/step - loss: 2.7774 - acc: 0.2364 - val_loss: 4.5435 - val_acc: 0.1674\n",
            "Epoch 124/250\n",
            "178/178 [==============================] - 25s 138ms/step - loss: 2.8166 - acc: 0.2347 - val_loss: 4.4233 - val_acc: 0.1610\n",
            "Epoch 125/250\n",
            "178/178 [==============================] - 25s 137ms/step - loss: 2.6704 - acc: 0.2385 - val_loss: 4.6498 - val_acc: 0.1822\n",
            "Epoch 126/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.7366 - acc: 0.2435 - val_loss: 4.6374 - val_acc: 0.1663\n",
            "Epoch 127/250\n",
            "178/178 [==============================] - 24s 133ms/step - loss: 2.7768 - acc: 0.2452 - val_loss: 4.6908 - val_acc: 0.1653\n",
            "Epoch 128/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.7337 - acc: 0.2322 - val_loss: 4.6887 - val_acc: 0.1706\n",
            "Epoch 129/250\n",
            "178/178 [==============================] - 25s 139ms/step - loss: 2.7891 - acc: 0.2350 - val_loss: 4.5712 - val_acc: 0.1758\n",
            "Epoch 130/250\n",
            "178/178 [==============================] - 25s 140ms/step - loss: 2.7932 - acc: 0.2315 - val_loss: 4.5100 - val_acc: 0.1684\n",
            "Epoch 131/250\n",
            "178/178 [==============================] - 25s 137ms/step - loss: 2.7373 - acc: 0.2347 - val_loss: 4.5268 - val_acc: 0.1621\n",
            "Epoch 132/250\n",
            "178/178 [==============================] - 24s 135ms/step - loss: 2.7356 - acc: 0.2410 - val_loss: 4.6297 - val_acc: 0.1822\n",
            "Epoch 133/250\n",
            "178/178 [==============================] - 20s 109ms/step - loss: 2.7665 - acc: 0.2297 - val_loss: 4.9070 - val_acc: 0.1748\n",
            "Epoch 134/250\n",
            "178/178 [==============================] - 25s 139ms/step - loss: 2.7934 - acc: 0.2375 - val_loss: 4.6932 - val_acc: 0.1600\n",
            "Epoch 135/250\n",
            "178/178 [==============================] - 24s 135ms/step - loss: 2.7664 - acc: 0.2431 - val_loss: 4.7686 - val_acc: 0.1706\n",
            "Epoch 136/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.6696 - acc: 0.2428 - val_loss: 4.7023 - val_acc: 0.1589\n",
            "Epoch 137/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.7211 - acc: 0.2414 - val_loss: 4.8189 - val_acc: 0.1642\n",
            "Epoch 138/250\n",
            "178/178 [==============================] - 25s 134ms/step - loss: 2.8210 - acc: 0.2399 - val_loss: 4.6328 - val_acc: 0.1748\n",
            "Epoch 139/250\n",
            "178/178 [==============================] - 26s 140ms/step - loss: 2.7441 - acc: 0.2410 - val_loss: 4.6430 - val_acc: 0.1610\n",
            "Epoch 140/250\n",
            "178/178 [==============================] - 25s 139ms/step - loss: 2.7463 - acc: 0.2438 - val_loss: 4.7166 - val_acc: 0.1642\n",
            "Epoch 141/250\n",
            "178/178 [==============================] - 25s 135ms/step - loss: 2.8455 - acc: 0.2378 - val_loss: 4.7508 - val_acc: 0.1547\n",
            "Epoch 142/250\n",
            "178/178 [==============================] - 24s 135ms/step - loss: 2.7979 - acc: 0.2237 - val_loss: 4.3926 - val_acc: 0.1917\n",
            "Epoch 143/250\n",
            "178/178 [==============================] - 25s 135ms/step - loss: 2.7131 - acc: 0.2364 - val_loss: 4.5685 - val_acc: 0.1631\n",
            "Epoch 144/250\n",
            "178/178 [==============================] - 24s 135ms/step - loss: 2.8114 - acc: 0.2474 - val_loss: 4.6062 - val_acc: 0.1642\n",
            "Epoch 145/250\n",
            "178/178 [==============================] - 25s 136ms/step - loss: 2.7499 - acc: 0.2466 - val_loss: 4.4054 - val_acc: 0.1748\n",
            "Epoch 146/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.7791 - acc: 0.2382 - val_loss: 4.5656 - val_acc: 0.1706\n",
            "Epoch 147/250\n",
            "178/178 [==============================] - 25s 135ms/step - loss: 2.8266 - acc: 0.2350 - val_loss: 4.4012 - val_acc: 0.1907\n",
            "Epoch 148/250\n",
            "178/178 [==============================] - 25s 140ms/step - loss: 2.7499 - acc: 0.2428 - val_loss: 4.6333 - val_acc: 0.1843\n",
            "Epoch 149/250\n",
            "178/178 [==============================] - 25s 138ms/step - loss: 2.7715 - acc: 0.2364 - val_loss: 4.6033 - val_acc: 0.1758\n",
            "Epoch 150/250\n",
            "178/178 [==============================] - 24s 135ms/step - loss: 2.8316 - acc: 0.2361 - val_loss: 4.4433 - val_acc: 0.1928\n",
            "Epoch 151/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.7756 - acc: 0.2262 - val_loss: 4.5346 - val_acc: 0.1917\n",
            "Epoch 152/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.7972 - acc: 0.2294 - val_loss: 4.6056 - val_acc: 0.1769\n",
            "Epoch 153/250\n",
            "178/178 [==============================] - 25s 135ms/step - loss: 2.8019 - acc: 0.2304 - val_loss: 4.4485 - val_acc: 0.1864\n",
            "Epoch 154/250\n",
            "178/178 [==============================] - 25s 138ms/step - loss: 2.8117 - acc: 0.2339 - val_loss: 4.4543 - val_acc: 0.1992\n",
            "Epoch 155/250\n",
            "178/178 [==============================] - 21s 113ms/step - loss: 2.8083 - acc: 0.2406 - val_loss: 4.6032 - val_acc: 0.1928\n",
            "Epoch 156/250\n",
            "178/178 [==============================] - 25s 139ms/step - loss: 2.7951 - acc: 0.2389 - val_loss: 4.7130 - val_acc: 0.1939\n",
            "Epoch 157/250\n",
            "178/178 [==============================] - 25s 136ms/step - loss: 2.8002 - acc: 0.2414 - val_loss: 4.5303 - val_acc: 0.1875\n",
            "Epoch 158/250\n",
            "178/178 [==============================] - 20s 111ms/step - loss: 2.7986 - acc: 0.2442 - val_loss: 4.5241 - val_acc: 0.1896\n",
            "Epoch 159/250\n",
            "178/178 [==============================] - 25s 140ms/step - loss: 2.7974 - acc: 0.2269 - val_loss: 4.5023 - val_acc: 0.1811\n",
            "Epoch 160/250\n",
            "178/178 [==============================] - 25s 141ms/step - loss: 2.7536 - acc: 0.2417 - val_loss: 4.6121 - val_acc: 0.1917\n",
            "Epoch 161/250\n",
            "178/178 [==============================] - 25s 135ms/step - loss: 2.7891 - acc: 0.2421 - val_loss: 4.6730 - val_acc: 0.1928\n",
            "Epoch 162/250\n",
            "178/178 [==============================] - 25s 136ms/step - loss: 2.7393 - acc: 0.2519 - val_loss: 4.4344 - val_acc: 0.1928\n",
            "Epoch 163/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.7929 - acc: 0.2410 - val_loss: 4.3506 - val_acc: 0.1822\n",
            "Epoch 164/250\n",
            "178/178 [==============================] - 25s 137ms/step - loss: 2.7868 - acc: 0.2421 - val_loss: 4.5744 - val_acc: 0.1769\n",
            "Epoch 165/250\n",
            "178/178 [==============================] - 25s 137ms/step - loss: 2.8302 - acc: 0.2389 - val_loss: 4.8001 - val_acc: 0.1578\n",
            "Epoch 166/250\n",
            "178/178 [==============================] - 25s 140ms/step - loss: 2.7468 - acc: 0.2470 - val_loss: 4.7674 - val_acc: 0.1780\n",
            "Epoch 167/250\n",
            "178/178 [==============================] - 25s 137ms/step - loss: 2.8604 - acc: 0.2311 - val_loss: 4.5454 - val_acc: 0.2013\n",
            "Epoch 168/250\n",
            "178/178 [==============================] - 25s 136ms/step - loss: 2.8454 - acc: 0.2350 - val_loss: 4.8448 - val_acc: 0.1748\n",
            "Epoch 169/250\n",
            "178/178 [==============================] - 21s 114ms/step - loss: 2.7804 - acc: 0.2364 - val_loss: 4.6157 - val_acc: 0.1811\n",
            "Epoch 170/250\n",
            "178/178 [==============================] - 26s 141ms/step - loss: 2.8966 - acc: 0.2310 - val_loss: 4.4727 - val_acc: 0.1801\n",
            "Epoch 171/250\n",
            "178/178 [==============================] - 25s 139ms/step - loss: 2.8305 - acc: 0.2445 - val_loss: 4.6332 - val_acc: 0.1928\n",
            "Epoch 172/250\n",
            "178/178 [==============================] - 25s 136ms/step - loss: 2.8463 - acc: 0.2272 - val_loss: 4.5783 - val_acc: 0.1843\n",
            "Epoch 173/250\n",
            "178/178 [==============================] - 25s 136ms/step - loss: 2.8581 - acc: 0.2385 - val_loss: 4.6815 - val_acc: 0.1684\n",
            "Epoch 174/250\n",
            "178/178 [==============================] - 20s 110ms/step - loss: 2.8596 - acc: 0.2255 - val_loss: 4.8173 - val_acc: 0.1684\n",
            "Epoch 175/250\n",
            "178/178 [==============================] - 25s 139ms/step - loss: 2.8046 - acc: 0.2315 - val_loss: 4.6777 - val_acc: 0.1843\n",
            "Epoch 176/250\n",
            "178/178 [==============================] - 24s 133ms/step - loss: 2.7932 - acc: 0.2474 - val_loss: 4.5711 - val_acc: 0.1706\n",
            "Epoch 177/250\n",
            "178/178 [==============================] - 24s 135ms/step - loss: 2.8101 - acc: 0.2399 - val_loss: 4.9538 - val_acc: 0.1663\n",
            "Epoch 178/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.7667 - acc: 0.2470 - val_loss: 4.8440 - val_acc: 0.1653\n",
            "Epoch 179/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.7759 - acc: 0.2466 - val_loss: 4.6880 - val_acc: 0.1727\n",
            "Epoch 180/250\n",
            "178/178 [==============================] - 25s 139ms/step - loss: 2.8530 - acc: 0.2410 - val_loss: 4.8329 - val_acc: 0.1780\n",
            "Epoch 181/250\n",
            "178/178 [==============================] - 25s 139ms/step - loss: 2.8787 - acc: 0.2428 - val_loss: 4.7160 - val_acc: 0.1875\n",
            "Epoch 182/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.7649 - acc: 0.2481 - val_loss: 4.7208 - val_acc: 0.1886\n",
            "Epoch 183/250\n",
            "178/178 [==============================] - 24s 135ms/step - loss: 2.6971 - acc: 0.2488 - val_loss: 4.6237 - val_acc: 0.1907\n",
            "Epoch 184/250\n",
            "178/178 [==============================] - 25s 137ms/step - loss: 2.8131 - acc: 0.2565 - val_loss: 4.9002 - val_acc: 0.1642\n",
            "Epoch 185/250\n",
            "178/178 [==============================] - 25s 136ms/step - loss: 2.8453 - acc: 0.2431 - val_loss: 4.7377 - val_acc: 0.1600\n",
            "Epoch 186/250\n",
            "178/178 [==============================] - 25s 139ms/step - loss: 2.8522 - acc: 0.2414 - val_loss: 4.9316 - val_acc: 0.1642\n",
            "Epoch 187/250\n",
            "178/178 [==============================] - 25s 141ms/step - loss: 2.8105 - acc: 0.2276 - val_loss: 4.7523 - val_acc: 0.1790\n",
            "Epoch 188/250\n",
            "178/178 [==============================] - 25s 138ms/step - loss: 2.7899 - acc: 0.2396 - val_loss: 4.8218 - val_acc: 0.1843\n",
            "Epoch 189/250\n",
            "178/178 [==============================] - 25s 135ms/step - loss: 2.8110 - acc: 0.2290 - val_loss: 4.8184 - val_acc: 0.1748\n",
            "Epoch 190/250\n",
            "178/178 [==============================] - 25s 137ms/step - loss: 2.7916 - acc: 0.2332 - val_loss: 5.0154 - val_acc: 0.1896\n",
            "Epoch 191/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.8197 - acc: 0.2406 - val_loss: 4.8542 - val_acc: 0.1780\n",
            "Epoch 192/250\n",
            "178/178 [==============================] - 25s 135ms/step - loss: 2.8668 - acc: 0.2343 - val_loss: 4.7499 - val_acc: 0.1907\n",
            "Epoch 193/250\n",
            "178/178 [==============================] - 26s 142ms/step - loss: 2.8008 - acc: 0.2371 - val_loss: 4.8542 - val_acc: 0.1896\n",
            "Epoch 194/250\n",
            "178/178 [==============================] - 26s 141ms/step - loss: 2.8272 - acc: 0.2339 - val_loss: 5.2144 - val_acc: 0.1695\n",
            "Epoch 195/250\n",
            "178/178 [==============================] - 25s 138ms/step - loss: 2.8024 - acc: 0.2357 - val_loss: 5.0917 - val_acc: 0.1684\n",
            "Epoch 196/250\n",
            "178/178 [==============================] - 25s 140ms/step - loss: 2.8581 - acc: 0.2484 - val_loss: 4.7891 - val_acc: 0.1706\n",
            "Epoch 197/250\n",
            "178/178 [==============================] - 25s 139ms/step - loss: 2.8207 - acc: 0.2526 - val_loss: 4.9827 - val_acc: 0.1674\n",
            "Epoch 198/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.8029 - acc: 0.2541 - val_loss: 4.9418 - val_acc: 0.1684\n",
            "Epoch 199/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.7288 - acc: 0.2498 - val_loss: 4.8652 - val_acc: 0.1547\n",
            "Epoch 200/250\n",
            "178/178 [==============================] - 25s 138ms/step - loss: 2.8477 - acc: 0.2385 - val_loss: 4.6929 - val_acc: 0.1917\n",
            "Epoch 201/250\n",
            "178/178 [==============================] - 25s 140ms/step - loss: 2.8613 - acc: 0.2438 - val_loss: 4.7184 - val_acc: 0.1822\n",
            "Epoch 202/250\n",
            "178/178 [==============================] - 20s 110ms/step - loss: 2.8308 - acc: 0.2498 - val_loss: 4.6986 - val_acc: 0.1631\n",
            "Epoch 203/250\n",
            "178/178 [==============================] - 25s 136ms/step - loss: 2.8819 - acc: 0.2361 - val_loss: 4.9885 - val_acc: 0.1769\n",
            "Epoch 204/250\n",
            "178/178 [==============================] - 25s 138ms/step - loss: 2.7892 - acc: 0.2368 - val_loss: 4.9069 - val_acc: 0.1631\n",
            "Epoch 205/250\n",
            "178/178 [==============================] - 25s 139ms/step - loss: 2.7928 - acc: 0.2481 - val_loss: 4.6587 - val_acc: 0.1886\n",
            "Epoch 206/250\n",
            "178/178 [==============================] - 24s 135ms/step - loss: 2.8340 - acc: 0.2431 - val_loss: 4.8970 - val_acc: 0.1780\n",
            "Epoch 207/250\n",
            "178/178 [==============================] - 24s 135ms/step - loss: 2.8476 - acc: 0.2177 - val_loss: 4.8996 - val_acc: 0.1780\n",
            "Epoch 208/250\n",
            "178/178 [==============================] - 24s 133ms/step - loss: 2.8027 - acc: 0.2452 - val_loss: 4.8346 - val_acc: 0.1907\n",
            "Epoch 209/250\n",
            "178/178 [==============================] - 25s 136ms/step - loss: 2.8437 - acc: 0.2470 - val_loss: 4.9095 - val_acc: 0.1917\n",
            "Epoch 210/250\n",
            "178/178 [==============================] - 25s 138ms/step - loss: 2.9084 - acc: 0.2354 - val_loss: 4.6672 - val_acc: 0.1833\n",
            "Epoch 211/250\n",
            "178/178 [==============================] - 25s 137ms/step - loss: 2.8333 - acc: 0.2392 - val_loss: 4.8088 - val_acc: 0.1653\n",
            "Epoch 212/250\n",
            "178/178 [==============================] - 20s 109ms/step - loss: 2.8379 - acc: 0.2354 - val_loss: 4.9110 - val_acc: 0.1769\n",
            "Epoch 213/250\n",
            "178/178 [==============================] - 25s 137ms/step - loss: 2.8923 - acc: 0.2417 - val_loss: 4.8142 - val_acc: 0.1992\n",
            "Epoch 214/250\n",
            "178/178 [==============================] - 25s 139ms/step - loss: 2.8516 - acc: 0.2537 - val_loss: 4.8352 - val_acc: 0.1769\n",
            "Epoch 215/250\n",
            "178/178 [==============================] - 20s 111ms/step - loss: 2.8185 - acc: 0.2445 - val_loss: 4.9399 - val_acc: 0.1780\n",
            "Epoch 216/250\n",
            "178/178 [==============================] - 25s 134ms/step - loss: 2.8287 - acc: 0.2361 - val_loss: 4.7794 - val_acc: 0.1854\n",
            "Epoch 217/250\n",
            "178/178 [==============================] - 25s 137ms/step - loss: 2.9047 - acc: 0.2361 - val_loss: 4.9690 - val_acc: 0.1653\n",
            "Epoch 218/250\n",
            "178/178 [==============================] - 25s 139ms/step - loss: 2.7579 - acc: 0.2456 - val_loss: 5.1558 - val_acc: 0.1737\n",
            "Epoch 219/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.8257 - acc: 0.2428 - val_loss: 5.0626 - val_acc: 0.1811\n",
            "Epoch 220/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.8061 - acc: 0.2445 - val_loss: 4.9506 - val_acc: 0.1695\n",
            "Epoch 221/250\n",
            "178/178 [==============================] - 24s 133ms/step - loss: 2.8014 - acc: 0.2350 - val_loss: 4.9787 - val_acc: 0.1801\n",
            "Epoch 222/250\n",
            "178/178 [==============================] - 25s 137ms/step - loss: 2.7735 - acc: 0.2392 - val_loss: 4.9436 - val_acc: 0.1716\n",
            "Epoch 223/250\n",
            "178/178 [==============================] - 25s 137ms/step - loss: 2.8267 - acc: 0.2470 - val_loss: 4.9657 - val_acc: 0.1727\n",
            "Epoch 224/250\n",
            "178/178 [==============================] - 25s 137ms/step - loss: 2.9320 - acc: 0.2368 - val_loss: 4.7055 - val_acc: 0.1928\n",
            "Epoch 225/250\n",
            "178/178 [==============================] - 24s 135ms/step - loss: 2.7941 - acc: 0.2378 - val_loss: 5.1272 - val_acc: 0.1642\n",
            "Epoch 226/250\n",
            "178/178 [==============================] - 24s 135ms/step - loss: 2.8871 - acc: 0.2325 - val_loss: 4.6404 - val_acc: 0.1822\n",
            "Epoch 227/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.8416 - acc: 0.2392 - val_loss: 4.8922 - val_acc: 0.1716\n",
            "Epoch 228/250\n",
            "178/178 [==============================] - 24s 133ms/step - loss: 2.8056 - acc: 0.2350 - val_loss: 4.8687 - val_acc: 0.1801\n",
            "Epoch 229/250\n",
            "178/178 [==============================] - 24s 135ms/step - loss: 2.8094 - acc: 0.2343 - val_loss: 4.7780 - val_acc: 0.1758\n",
            "Epoch 230/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.8430 - acc: 0.2477 - val_loss: 4.9094 - val_acc: 0.1875\n",
            "Epoch 231/250\n",
            "178/178 [==============================] - 24s 135ms/step - loss: 2.8473 - acc: 0.2396 - val_loss: 5.0365 - val_acc: 0.1631\n",
            "Epoch 232/250\n",
            "178/178 [==============================] - 25s 135ms/step - loss: 2.9024 - acc: 0.2290 - val_loss: 4.8241 - val_acc: 0.1674\n",
            "Epoch 233/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.9074 - acc: 0.2414 - val_loss: 4.7580 - val_acc: 0.1758\n",
            "Epoch 234/250\n",
            "178/178 [==============================] - 25s 138ms/step - loss: 2.8824 - acc: 0.2428 - val_loss: 4.7397 - val_acc: 0.1790\n",
            "Epoch 235/250\n",
            "178/178 [==============================] - 25s 140ms/step - loss: 2.9316 - acc: 0.2357 - val_loss: 4.8504 - val_acc: 0.1769\n",
            "Epoch 236/250\n",
            "178/178 [==============================] - 25s 136ms/step - loss: 2.8402 - acc: 0.2442 - val_loss: 4.9806 - val_acc: 0.1854\n",
            "Epoch 237/250\n",
            "178/178 [==============================] - 20s 112ms/step - loss: 2.8589 - acc: 0.2445 - val_loss: 4.6873 - val_acc: 0.1833\n",
            "Epoch 238/250\n",
            "178/178 [==============================] - 25s 139ms/step - loss: 2.9231 - acc: 0.2354 - val_loss: 5.0563 - val_acc: 0.1864\n",
            "Epoch 239/250\n",
            "178/178 [==============================] - 25s 139ms/step - loss: 2.8711 - acc: 0.2474 - val_loss: 4.9258 - val_acc: 0.1748\n",
            "Epoch 240/250\n",
            "178/178 [==============================] - 25s 136ms/step - loss: 2.7936 - acc: 0.2442 - val_loss: 4.9254 - val_acc: 0.1780\n",
            "Epoch 241/250\n",
            "178/178 [==============================] - 25s 136ms/step - loss: 2.9449 - acc: 0.2304 - val_loss: 4.7760 - val_acc: 0.1875\n",
            "Epoch 242/250\n",
            "178/178 [==============================] - 25s 135ms/step - loss: 2.8326 - acc: 0.2399 - val_loss: 5.0461 - val_acc: 0.1748\n",
            "Epoch 243/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.9608 - acc: 0.2392 - val_loss: 4.7636 - val_acc: 0.1631\n",
            "Epoch 244/250\n",
            "178/178 [==============================] - 25s 138ms/step - loss: 2.8383 - acc: 0.2431 - val_loss: 4.6405 - val_acc: 0.1642\n",
            "Epoch 245/250\n",
            "178/178 [==============================] - 25s 139ms/step - loss: 2.9826 - acc: 0.2470 - val_loss: 4.7001 - val_acc: 0.1939\n",
            "Epoch 246/250\n",
            "178/178 [==============================] - 25s 138ms/step - loss: 2.8590 - acc: 0.2431 - val_loss: 4.9786 - val_acc: 0.1875\n",
            "Epoch 247/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.9179 - acc: 0.2435 - val_loss: 4.7190 - val_acc: 0.1886\n",
            "Epoch 248/250\n",
            "178/178 [==============================] - 24s 135ms/step - loss: 2.8847 - acc: 0.2294 - val_loss: 4.8479 - val_acc: 0.1801\n",
            "Epoch 249/250\n",
            "178/178 [==============================] - 24s 134ms/step - loss: 2.8717 - acc: 0.2438 - val_loss: 5.1508 - val_acc: 0.1684\n",
            "Epoch 250/250\n",
            "178/178 [==============================] - 25s 135ms/step - loss: 2.8970 - acc: 0.2406 - val_loss: 4.9209 - val_acc: 0.1854\n"
          ]
        }
      ],
      "source": [
        "history = model.fit_generator(\n",
        "      train_generator,\n",
        "      steps_per_epoch= NUM_TRAIN //batch_size,\n",
        "      epochs=epochs,\n",
        "      validation_data=validation_generator,\n",
        "      validation_steps= NUM_TEST //batch_size,\n",
        "      verbose=1,\n",
        "      use_multiprocessing=True,\n",
        "      workers=4)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 17,
      "metadata": {
        "id": "Y3K89-CM-dfg",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 545
        },
        "outputId": "00110f9f-18e6-4b49-b786-f6fa08cf713c"
      },
      "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"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ],
      "source": [
        "acc = history.history['acc']\n",
        "val_acc = history.history['val_acc']\n",
        "loss = history.history['loss']\n",
        "val_loss = history.history['val_loss']\n",
        "\n",
        "epochs_x = range(len(acc))\n",
        "\n",
        "plt.plot(epochs_x, acc, 'ro', label='Training acc')\n",
        "plt.plot(epochs_x, val_acc, 'k', label='Validation acc')\n",
        "plt.title('Training and validation accuracy')\n",
        "plt.legend()\n",
        "\n",
        "plt.figure()\n",
        "\n",
        "plt.plot(epochs_x, loss, 'ro', label='Training loss')\n",
        "plt.plot(epochs_x, val_loss, 'k', label='Validation loss')\n",
        "plt.title('Training and validation loss')\n",
        "plt.legend()\n",
        "\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "R19IJQSYoW7J"
      },
      "source": [
        "#Download the model\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 18,
      "metadata": {
        "id": "Zed4TdFcG2iJ"
      },
      "outputs": [],
      "source": [
        "os.makedirs('/content/drive/My Drive/cut_panoramic/Model', exist_ok=True)\n",
        "model.save('/content/drive/MyDrive/cut_panoramic/Model/Classification/All_Age/AC3_AllAge_Unfreeze_500.h5')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 19,
      "metadata": {
        "id": "P5eMxm1NV-oY",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 17
        },
        "outputId": "aa2fbaf6-41bd-4bd5-d11e-e5723963ac87"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ],
            "application/javascript": [
              "\n",
              "    async function download(id, filename, size) {\n",
              "      if (!google.colab.kernel.accessAllowed) {\n",
              "        return;\n",
              "      }\n",
              "      const div = document.createElement('div');\n",
              "      const label = document.createElement('label');\n",
              "      label.textContent = `Downloading \"${filename}\": `;\n",
              "      div.appendChild(label);\n",
              "      const progress = document.createElement('progress');\n",
              "      progress.max = size;\n",
              "      div.appendChild(progress);\n",
              "      document.body.appendChild(div);\n",
              "\n",
              "      const buffers = [];\n",
              "      let downloaded = 0;\n",
              "\n",
              "      const channel = await google.colab.kernel.comms.open(id);\n",
              "      // Send a message to notify the kernel that we're ready.\n",
              "      channel.send({})\n",
              "\n",
              "      for await (const message of channel.messages) {\n",
              "        // Send a message to notify the kernel that we're ready.\n",
              "        channel.send({})\n",
              "        if (message.buffers) {\n",
              "          for (const buffer of message.buffers) {\n",
              "            buffers.push(buffer);\n",
              "            downloaded += buffer.byteLength;\n",
              "            progress.value = downloaded;\n",
              "          }\n",
              "        }\n",
              "      }\n",
              "      const blob = new Blob(buffers, {type: 'application/binary'});\n",
              "      const a = document.createElement('a');\n",
              "      a.href = window.URL.createObjectURL(blob);\n",
              "      a.download = filename;\n",
              "      div.appendChild(a);\n",
              "      a.click();\n",
              "      div.remove();\n",
              "    }\n",
              "  "
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ],
            "application/javascript": [
              "download(\"download_54e42003-50d2-44ec-93b5-654ba2294e28\", \"AC3_AllAge_Unfreeze_500.h5\", 16778608)"
            ]
          },
          "metadata": {}
        }
      ],
      "source": [
        "from google.colab import files\n",
        "files.download('/content/drive/MyDrive/cut_panoramic/Model/Classification/All_Age/AC3_AllAge_Unfreeze_500.h5')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "xlsuaFIUVriv"
      },
      "outputs": [],
      "source": []
    }
  ],
  "metadata": {
    "colab": {
      "provenance": [],
      "machine_shape": "hm",
      "include_colab_link": true
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU",
    "gpuClass": "standard"
  },
  "nbformat": 4,
  "nbformat_minor": 0
}