P6_02_CNN_from_scratch.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "name": "IML_P6_CNN_from_scratch_v3.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "CbDxrkupTYEp",
        "outputId": "9a5e7c4f-92e8-42ad-aa84-d7535c74151e"
      },
      "source": [
        "import os\n",
        "import time\n",
        "import numpy as np\n",
        "import tensorflow as tf\n",
        "import tensorflow.keras as keras \n",
        "\n",
        "!pip install -U keras-tuner\n",
        "import kerastuner as kt"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Collecting keras-tuner\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/20/ec/1ef246787174b1e2bb591c95f29d3c1310070cad877824f907faba3dade9/keras-tuner-1.0.2.tar.gz (62kB)\n",
            "\r\u001b[K     |█████▏                          | 10kB 23.0MB/s eta 0:00:01\r\u001b[K     |██████████▍                     | 20kB 10.1MB/s eta 0:00:01\r\u001b[K     |███████████████▋                | 30kB 6.8MB/s eta 0:00:01\r\u001b[K     |████████████████████▉           | 40kB 3.2MB/s eta 0:00:01\r\u001b[K     |██████████████████████████      | 51kB 3.9MB/s eta 0:00:01\r\u001b[K     |███████████████████████████████▎| 61kB 4.3MB/s eta 0:00:01\r\u001b[K     |████████████████████████████████| 71kB 3.3MB/s \n",
            "\u001b[?25hRequirement already satisfied, skipping upgrade: packaging in /usr/local/lib/python3.7/dist-packages (from keras-tuner) (20.9)\n",
            "Requirement already satisfied, skipping upgrade: future in /usr/local/lib/python3.7/dist-packages (from keras-tuner) (0.16.0)\n",
            "Requirement already satisfied, skipping upgrade: numpy in /usr/local/lib/python3.7/dist-packages (from keras-tuner) (1.19.5)\n",
            "Requirement already satisfied, skipping upgrade: tabulate in /usr/local/lib/python3.7/dist-packages (from keras-tuner) (0.8.9)\n",
            "Collecting terminaltables\n",
            "  Downloading https://files.pythonhosted.org/packages/9b/c4/4a21174f32f8a7e1104798c445dacdc1d4df86f2f26722767034e4de4bff/terminaltables-3.1.0.tar.gz\n",
            "Collecting colorama\n",
            "  Downloading https://files.pythonhosted.org/packages/44/98/5b86278fbbf250d239ae0ecb724f8572af1c91f4a11edf4d36a206189440/colorama-0.4.4-py2.py3-none-any.whl\n",
            "Requirement already satisfied, skipping upgrade: tqdm in /usr/local/lib/python3.7/dist-packages (from keras-tuner) (4.41.1)\n",
            "Requirement already satisfied, skipping upgrade: requests in /usr/local/lib/python3.7/dist-packages (from keras-tuner) (2.23.0)\n",
            "Requirement already satisfied, skipping upgrade: scipy in /usr/local/lib/python3.7/dist-packages (from keras-tuner) (1.4.1)\n",
            "Requirement already satisfied, skipping upgrade: scikit-learn in /usr/local/lib/python3.7/dist-packages (from keras-tuner) (0.22.2.post1)\n",
            "Requirement already satisfied, skipping upgrade: pyparsing>=2.0.2 in /usr/local/lib/python3.7/dist-packages (from packaging->keras-tuner) (2.4.7)\n",
            "Requirement already satisfied, skipping upgrade: chardet<4,>=3.0.2 in /usr/local/lib/python3.7/dist-packages (from requests->keras-tuner) (3.0.4)\n",
            "Requirement already satisfied, skipping upgrade: urllib3!=1.25.0,!=1.25.1,<1.26,>=1.21.1 in /usr/local/lib/python3.7/dist-packages (from requests->keras-tuner) (1.24.3)\n",
            "Requirement already satisfied, skipping upgrade: idna<3,>=2.5 in /usr/local/lib/python3.7/dist-packages (from requests->keras-tuner) (2.10)\n",
            "Requirement already satisfied, skipping upgrade: certifi>=2017.4.17 in /usr/local/lib/python3.7/dist-packages (from requests->keras-tuner) (2020.12.5)\n",
            "Requirement already satisfied, skipping upgrade: joblib>=0.11 in /usr/local/lib/python3.7/dist-packages (from scikit-learn->keras-tuner) (1.0.1)\n",
            "Building wheels for collected packages: keras-tuner, terminaltables\n",
            "  Building wheel for keras-tuner (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for keras-tuner: filename=keras_tuner-1.0.2-cp37-none-any.whl size=78938 sha256=99e8b683dbe52b55c322edf85092a8b3884949239b6408ad40f7ec6d45bf2eb2\n",
            "  Stored in directory: /root/.cache/pip/wheels/bb/a1/8a/7c3de0efb3707a1701b36ebbfdbc4e67aedf6d4943a1f463d6\n",
            "  Building wheel for terminaltables (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for terminaltables: filename=terminaltables-3.1.0-cp37-none-any.whl size=15356 sha256=f05cb574a020d23ecde56e43d58ae7ab38239f4e39f8dfedd479ede8b3ada5b2\n",
            "  Stored in directory: /root/.cache/pip/wheels/30/6b/50/6c75775b681fb36cdfac7f19799888ef9d8813aff9e379663e\n",
            "Successfully built keras-tuner terminaltables\n",
            "Installing collected packages: terminaltables, colorama, keras-tuner\n",
            "Successfully installed colorama-0.4.4 keras-tuner-1.0.2 terminaltables-3.1.0\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4npkOCEyWIJ6"
      },
      "source": [
        "#----------------------#"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "cmQNT6kCWLct",
        "outputId": "1e5791b3-103f-4a23-e6d0-3c833f230b5a"
      },
      "source": [
        "# Using images stored on Google Drive\n",
        "\n",
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Mounted at /content/drive\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9qX3ZRn-WjcJ",
        "outputId": "ccc46ecf-4340-4253-c578-080d1daa9c03"
      },
      "source": [
        "!ls \"/content/drive/MyDrive/Gaetan_Travail/ML/Formation_OC_IML/P6/Projet/Data\""
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Data_augmentation  Images_few_10  Images_few_50  Images_split  README.txt\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 350
        },
        "id": "rAmOeNllYaAJ",
        "outputId": "6b06585b-a429-48e8-e814-cd6b7712ed77"
      },
      "source": [
        "# Test -> show an image from the dataset\n",
        "# Format = PIL\n",
        "keras.preprocessing.image.load_img(\"/content/drive/MyDrive/Gaetan_Travail/ML/Formation_OC_IML/P6/Projet/Data/Images_few_10/train/German_shepherd/n02106662_590.jpg\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<PIL.JpegImagePlugin.JpegImageFile image mode=RGB size=500x333 at 0x7FD017C6C810>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "o_LJDLYdWLlL"
      },
      "source": [
        "#----------------------#"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4BFkNgZYV7ZQ",
        "outputId": "e80e8e45-83e4-4055-9460-2eaf5abce76e"
      },
      "source": [
        "# Dataset generation (train, validation et test)\n",
        "\n",
        "# In order to train the model faster,\n",
        "# we will use 10 dog breeds\n",
        "# and resize the image -> 150x150\n",
        "\n",
        "path_google_drive = \"/content/drive/MyDrive/Gaetan_Travail/ML/Formation_OC_IML/P6/Projet/Data/\"\n",
        "\n",
        "batch_size_ = 10\n",
        "IMG_SIZE_ = 150\n",
        "\n",
        "# Data augmentation:\n",
        "# only for the train set\n",
        "datagen_train = tf.keras.preprocessing.image.ImageDataGenerator(\n",
        "    rescale=1/255,  # normalization\n",
        "    rotation_range=20,\n",
        "    width_shift_range=0.2,\n",
        "    height_shift_range=0.2,\n",
        "    horizontal_flip=True,\n",
        "    validation_split=0\n",
        ")\n",
        "\n",
        "test_datagen = tf.keras.preprocessing.image.ImageDataGenerator(\n",
        "    rescale=1/255,  # normalization\n",
        "    validation_split=0\n",
        ")\n",
        "\n",
        "# Train set:\n",
        "path = path_google_drive + \"Images_few_10/train/\"\n",
        "\n",
        "x_train = datagen_train.flow_from_directory(\n",
        "    path,\n",
        "    target_size=(IMG_SIZE_, IMG_SIZE_), # resizing\n",
        "    batch_size=batch_size_,\n",
        "    class_mode=\"categorical\",\n",
        "    shuffle=True,\n",
        "    seed=0,\n",
        "    subset='training'\n",
        ")\n",
        "\n",
        "# Validation set:\n",
        "path = path_google_drive + \"Images_few_10/val/\"\n",
        "\n",
        "x_val = test_datagen.flow_from_directory(\n",
        "    path,\n",
        "    batch_size=batch_size_,\n",
        "    class_mode=\"categorical\",\n",
        "    target_size=(IMG_SIZE_, IMG_SIZE_), # resizing\n",
        "    seed=0,\n",
        "    subset='training'\n",
        ")\n",
        "\n",
        "# Test set:\n",
        "path = path_google_drive + \"Images_few_10/test/\"\n",
        "\n",
        "x_test = test_datagen.flow_from_directory(\n",
        "    path,\n",
        "    batch_size=batch_size_,\n",
        "    class_mode=\"categorical\",\n",
        "    target_size=(IMG_SIZE_, IMG_SIZE_), # resizing\n",
        "    seed=0,\n",
        "    subset='training'\n",
        ")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Found 1025 images belonging to 10 classes.\n",
            "Found 338 images belonging to 10 classes.\n",
            "Found 350 images belonging to 10 classes.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pwfCoqRrZfB4"
      },
      "source": [
        "#----------------------#"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mRe3WXtj3UsV"
      },
      "source": [
        "# In order to get the first parts of our CNN,\n",
        "# we will compare two architectures :\n",
        "# one inspired by AlexNet,\n",
        "# the other one inspired by VGG16"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8CPR5eqlpcdq"
      },
      "source": [
        "nbr_class = 10\n",
        "IMG_SIZE_ = 150\n",
        "\n",
        "def build_model_AlexNet():\n",
        "  input_ = keras.layers.Input(shape=(IMG_SIZE_, IMG_SIZE_, 3))\n",
        "\n",
        "  x = keras.layers.Conv2D(96, 11, activation=\"relu\", padding=\"same\")(input_)\n",
        "  x = keras.layers.MaxPool2D(2)(x)\n",
        "\n",
        "  x = keras.layers.Conv2D(256, 5, activation=\"relu\", padding=\"same\")(x)\n",
        "  x = keras.layers.MaxPool2D(2)(x)\n",
        "\n",
        "  x = keras.layers.Conv2D(384, 3, activation=\"relu\", padding=\"same\")(x)\n",
        "  x = keras.layers.Conv2D(384, 3, activation=\"relu\", padding=\"same\")(x)\n",
        "  x = keras.layers.Conv2D(256, 3, activation=\"relu\", padding=\"same\")(x)\n",
        "  x = keras.layers.MaxPool2D(2)(x)\n",
        "\n",
        "  x = keras.layers.Flatten()(x)\n",
        "\n",
        "  x = keras.layers.Dense(4096, activation=\"relu\")(x)\n",
        "  x = keras.layers.Dense(4096, activation=\"relu\")(x)\n",
        "\n",
        "  output_ = keras.layers.Dense(nbr_class, activation=\"softmax\")(x)\n",
        "\n",
        "  model = keras.Model(inputs=[input_], outputs=[output_])\n",
        "\n",
        "  model.compile(\n",
        "        loss=\"categorical_crossentropy\",\n",
        "        optimizer=tf.keras.optimizers.Adam(learning_rate=1e-4),\n",
        "        metrics=[\"accuracy\"]\n",
        "  )\n",
        "\n",
        "  return(model)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fm69JVVapcj9"
      },
      "source": [
        "nbr_class = 10\n",
        "IMG_SIZE_ = 150\n",
        "\n",
        "def build_model_VGG():\n",
        "  input_ = keras.layers.Input(shape=(IMG_SIZE_, IMG_SIZE_, 3))\n",
        "\n",
        "  x = keras.layers.Conv2D(64, 3, padding=\"same\", activation=\"relu\")(input_)\n",
        "  x = keras.layers.Conv2D(64, 3, padding=\"same\", activation=\"relu\")(x)\n",
        "  x = keras.layers.MaxPool2D(pool_size=(2,2), strides=(2,2))(x)\n",
        "\n",
        "  x = keras.layers.Conv2D(128, 3, padding=\"same\", activation=\"relu\")(x)\n",
        "  x = keras.layers.Conv2D(128, 3, padding=\"same\", activation=\"relu\")(x)\n",
        "  x = keras.layers.MaxPool2D(pool_size=(2,2),strides=(2,2))(x)\n",
        "\n",
        "  x = keras.layers.Conv2D(256, 3, padding=\"same\", activation=\"relu\")(x)\n",
        "  x = keras.layers.Conv2D(256, 3, padding=\"same\", activation=\"relu\")(x)\n",
        "  x = keras.layers.Conv2D(256, 3, padding=\"same\", activation=\"relu\")(x)\n",
        "  x = keras.layers.MaxPool2D(pool_size=(2,2),strides=(2,2))(x)\n",
        "\n",
        "  x = keras.layers.Conv2D(512, 3, padding=\"same\", activation=\"relu\")(x)\n",
        "  x = keras.layers.Conv2D(512, 3, padding=\"same\", activation=\"relu\")(x)\n",
        "  x = keras.layers.Conv2D(512, 3, padding=\"same\", activation=\"relu\")(x)\n",
        "  x = keras.layers.MaxPool2D(pool_size=(2,2),strides=(2,2))(x)\n",
        "\n",
        "  x = keras.layers.Conv2D(512, 3, padding=\"same\", activation=\"relu\")(x)\n",
        "  x = keras.layers.Conv2D(512, 3, padding=\"same\", activation=\"relu\")(x)\n",
        "  x = keras.layers.Conv2D(512, 3, padding=\"same\", activation=\"relu\")(x)\n",
        "  x = keras.layers.MaxPool2D(pool_size=(2,2),strides=(2,2))(x)\n",
        "\n",
        "  x = keras.layers.Flatten()(x)\n",
        "\n",
        "  x = keras.layers.Dense(4096, activation=\"relu\")(x)\n",
        "  x = keras.layers.Dense(4096, activation=\"relu\")(x)\n",
        "\n",
        "  output_ = keras.layers.Dense(nbr_class, activation=\"softmax\")(x)\n",
        "\n",
        "  model = keras.Model(inputs=[input_], outputs=[output_])\n",
        "\n",
        "  model.compile(\n",
        "        loss=\"categorical_crossentropy\",\n",
        "        optimizer=tf.keras.optimizers.Adam(learning_rate=1e-4),\n",
        "        metrics=[\"accuracy\"]\n",
        "  )\n",
        "\n",
        "  return(model)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "efFwOc4_9u-l"
      },
      "source": [
        "#----------------------#"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "YSdTQ4cBquLg",
        "outputId": "a5f98a58-7c48-4cff-98d0-e902c1572c7c"
      },
      "source": [
        "model_test_1 = build_model_AlexNet()\n",
        "\n",
        "model_test_1.fit(\n",
        "      x_train,\n",
        "      epochs=1000,\n",
        "      validation_data=x_val,\n",
        "      callbacks=[keras.callbacks.EarlyStopping(patience=3)]\n",
        "  )"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/1000\n",
            "103/103 [==============================] - 15s 131ms/step - loss: 2.3077 - accuracy: 0.1166 - val_loss: 2.2990 - val_accuracy: 0.1391\n",
            "Epoch 2/1000\n",
            "103/103 [==============================] - 13s 122ms/step - loss: 2.2993 - accuracy: 0.1397 - val_loss: 2.2987 - val_accuracy: 0.1391\n",
            "Epoch 3/1000\n",
            "103/103 [==============================] - 13s 121ms/step - loss: 2.2966 - accuracy: 0.1537 - val_loss: 2.2904 - val_accuracy: 0.1391\n",
            "Epoch 4/1000\n",
            "103/103 [==============================] - 12s 120ms/step - loss: 2.2971 - accuracy: 0.1402 - val_loss: 2.2950 - val_accuracy: 0.1391\n",
            "Epoch 5/1000\n",
            "103/103 [==============================] - 13s 122ms/step - loss: 2.2941 - accuracy: 0.1435 - val_loss: 2.2892 - val_accuracy: 0.1391\n",
            "Epoch 6/1000\n",
            "103/103 [==============================] - 12s 120ms/step - loss: 2.2885 - accuracy: 0.1597 - val_loss: 2.2922 - val_accuracy: 0.1391\n",
            "Epoch 7/1000\n",
            "103/103 [==============================] - 12s 120ms/step - loss: 2.2933 - accuracy: 0.1391 - val_loss: 2.2871 - val_accuracy: 0.1391\n",
            "Epoch 8/1000\n",
            "103/103 [==============================] - 12s 120ms/step - loss: 2.2898 - accuracy: 0.1297 - val_loss: 2.2578 - val_accuracy: 0.1864\n",
            "Epoch 9/1000\n",
            "103/103 [==============================] - 12s 119ms/step - loss: 2.2567 - accuracy: 0.1762 - val_loss: 2.2709 - val_accuracy: 0.1627\n",
            "Epoch 10/1000\n",
            "103/103 [==============================] - 12s 119ms/step - loss: 2.2605 - accuracy: 0.1697 - val_loss: 2.1801 - val_accuracy: 0.1893\n",
            "Epoch 11/1000\n",
            "103/103 [==============================] - 12s 119ms/step - loss: 2.2322 - accuracy: 0.1341 - val_loss: 2.1966 - val_accuracy: 0.1598\n",
            "Epoch 12/1000\n",
            "103/103 [==============================] - 12s 120ms/step - loss: 2.2124 - accuracy: 0.1877 - val_loss: 2.1827 - val_accuracy: 0.1509\n",
            "Epoch 13/1000\n",
            "103/103 [==============================] - 12s 121ms/step - loss: 2.1716 - accuracy: 0.2133 - val_loss: 2.1625 - val_accuracy: 0.1716\n",
            "Epoch 14/1000\n",
            "103/103 [==============================] - 12s 119ms/step - loss: 2.1615 - accuracy: 0.2031 - val_loss: 2.1428 - val_accuracy: 0.1716\n",
            "Epoch 15/1000\n",
            "103/103 [==============================] - 12s 119ms/step - loss: 2.1749 - accuracy: 0.1757 - val_loss: 2.1362 - val_accuracy: 0.1805\n",
            "Epoch 16/1000\n",
            "103/103 [==============================] - 12s 120ms/step - loss: 2.1496 - accuracy: 0.2040 - val_loss: 2.1320 - val_accuracy: 0.2012\n",
            "Epoch 17/1000\n",
            "103/103 [==============================] - 12s 120ms/step - loss: 2.1723 - accuracy: 0.2102 - val_loss: 2.1521 - val_accuracy: 0.1923\n",
            "Epoch 18/1000\n",
            "103/103 [==============================] - 12s 120ms/step - loss: 2.1724 - accuracy: 0.1933 - val_loss: 2.1313 - val_accuracy: 0.2012\n",
            "Epoch 19/1000\n",
            "103/103 [==============================] - 12s 121ms/step - loss: 2.1082 - accuracy: 0.2117 - val_loss: 2.1238 - val_accuracy: 0.2012\n",
            "Epoch 20/1000\n",
            "103/103 [==============================] - 12s 120ms/step - loss: 2.1384 - accuracy: 0.2093 - val_loss: 2.1357 - val_accuracy: 0.1953\n",
            "Epoch 21/1000\n",
            "103/103 [==============================] - 12s 120ms/step - loss: 2.1361 - accuracy: 0.2257 - val_loss: 2.1389 - val_accuracy: 0.1982\n",
            "Epoch 22/1000\n",
            "103/103 [==============================] - 12s 120ms/step - loss: 2.1329 - accuracy: 0.2265 - val_loss: 2.1532 - val_accuracy: 0.1953\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tensorflow.python.keras.callbacks.History at 0x7fd0004b7110>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "i7ybSUazquTV",
        "outputId": "dffc3b42-b539-4c26-ddf6-0a699a288f42"
      },
      "source": [
        "model_test_2 = build_model_VGG()\n",
        "\n",
        "model_test_2.fit(\n",
        "      x_train,\n",
        "      epochs=1000,\n",
        "      validation_data=x_val,\n",
        "      callbacks=[keras.callbacks.EarlyStopping(patience=3)]\n",
        "  )"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/1000\n",
            "103/103 [==============================] - 15s 126ms/step - loss: 2.3006 - accuracy: 0.1212 - val_loss: 2.2924 - val_accuracy: 0.1391\n",
            "Epoch 2/1000\n",
            "103/103 [==============================] - 12s 114ms/step - loss: 2.2913 - accuracy: 0.1426 - val_loss: 2.2941 - val_accuracy: 0.1391\n",
            "Epoch 3/1000\n",
            "103/103 [==============================] - 12s 114ms/step - loss: 2.2986 - accuracy: 0.1345 - val_loss: 2.2924 - val_accuracy: 0.1391\n",
            "Epoch 4/1000\n",
            "103/103 [==============================] - 12s 117ms/step - loss: 2.2998 - accuracy: 0.1241 - val_loss: 2.2916 - val_accuracy: 0.1391\n",
            "Epoch 5/1000\n",
            "103/103 [==============================] - 12s 115ms/step - loss: 2.2916 - accuracy: 0.1436 - val_loss: 2.2918 - val_accuracy: 0.1391\n",
            "Epoch 6/1000\n",
            "103/103 [==============================] - 12s 115ms/step - loss: 2.2937 - accuracy: 0.1389 - val_loss: 2.2915 - val_accuracy: 0.1391\n",
            "Epoch 7/1000\n",
            "103/103 [==============================] - 12s 117ms/step - loss: 2.2933 - accuracy: 0.1276 - val_loss: 2.2914 - val_accuracy: 0.1391\n",
            "Epoch 8/1000\n",
            "103/103 [==============================] - 12s 114ms/step - loss: 2.2956 - accuracy: 0.1418 - val_loss: 2.2913 - val_accuracy: 0.1391\n",
            "Epoch 9/1000\n",
            "103/103 [==============================] - 12s 115ms/step - loss: 2.2851 - accuracy: 0.1522 - val_loss: 2.2916 - val_accuracy: 0.1391\n",
            "Epoch 10/1000\n",
            "103/103 [==============================] - 12s 115ms/step - loss: 2.2892 - accuracy: 0.1563 - val_loss: 2.2915 - val_accuracy: 0.1391\n",
            "Epoch 11/1000\n",
            "103/103 [==============================] - 12s 114ms/step - loss: 2.2931 - accuracy: 0.1437 - val_loss: 2.2913 - val_accuracy: 0.1391\n",
            "Epoch 12/1000\n",
            "103/103 [==============================] - 12s 116ms/step - loss: 2.2918 - accuracy: 0.1359 - val_loss: 2.2914 - val_accuracy: 0.1391\n",
            "Epoch 13/1000\n",
            "103/103 [==============================] - 12s 115ms/step - loss: 2.2978 - accuracy: 0.1281 - val_loss: 2.2914 - val_accuracy: 0.1391\n",
            "Epoch 14/1000\n",
            "103/103 [==============================] - 12s 116ms/step - loss: 2.2923 - accuracy: 0.1401 - val_loss: 2.2913 - val_accuracy: 0.1391\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tensorflow.python.keras.callbacks.History at 0x7fd0157614d0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5KLQuXRjqupx",
        "outputId": "47b6e2cd-ee95-4f27-900a-423be137da09"
      },
      "source": [
        "model_test_1.evaluate(x_test, verbose=1)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "35/35 [==============================] - 172s 5s/step - loss: 2.1334 - accuracy: 0.2086\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "aLi34BcLquwZ",
        "outputId": "63bb9724-d96e-4d74-fe0a-eecda91fd29e"
      },
      "source": [
        "model_test_2.evaluate(x_test, verbose=1)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "35/35 [==============================] - 2s 47ms/step - loss: 2.2906 - accuracy: 0.1400\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fG64jFKJtWMB"
      },
      "source": [
        "'''\n",
        "AlexNet seems to be a good start in order to build our own CNN.\n",
        "'''"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "boLbH2SPtmAY"
      },
      "source": [
        "#----------------------#"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "03gOUBuY4GeZ"
      },
      "source": [
        "# In order to get better regularization, we are adding dropout layers.\n",
        "# Besides, l2 regularization is used with the fully connected layers.\n",
        "\n",
        "# We are using keras tuner to find interesting hyperparameters values."
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PO9bKRBYogll"
      },
      "source": [
        "nbr_class = 10\n",
        "IMG_SIZE_ = 150\n",
        "\n",
        "def build_model_10_dogs(hp):\n",
        "  input_ = keras.layers.Input(shape=(IMG_SIZE_, IMG_SIZE_, 3))\n",
        "\n",
        "  #----------------------#\n",
        "\n",
        "  x = keras.layers.experimental.preprocessing.RandomCrop(IMG_SIZE_, IMG_SIZE_, seed=0)(input_)\n",
        "\n",
        "  #----------------------#\n",
        "\n",
        "  x = keras.layers.Conv2D(\n",
        "        filters=hp.Choice(\n",
        "                    'filter1_value',\n",
        "                    values=[64, 96, 128],\n",
        "                    default=96\n",
        "                ),\n",
        "        kernel_size=hp.Choice(\n",
        "                      'kernel1_value',\n",
        "                      values=[7, 11, 15],\n",
        "                      default=11\n",
        "                    ),\n",
        "        padding=\"same\",\n",
        "        use_bias=False\n",
        "      )(x)\n",
        "  x = keras.layers.BatchNormalization()(x)\n",
        "  x = keras.layers.Activation(\"relu\")(x)\n",
        "  x = keras.layers.MaxPool2D(2)(x)\n",
        "\n",
        "  #----------------------#\n",
        "\n",
        "  x = keras.layers.Conv2D(\n",
        "        filters=hp.Choice(\n",
        "                    'filter2_value',\n",
        "                    values=[64, 128, 256],\n",
        "                    default=256\n",
        "                ),\n",
        "        kernel_size=3,\n",
        "        padding=\"same\", \n",
        "        use_bias=False\n",
        "      )(x)\n",
        "  x = keras.layers.BatchNormalization()(x)\n",
        "  x = keras.layers.Activation(\"relu\")(x)\n",
        "  x = keras.layers.MaxPool2D(2)(x)\n",
        "\n",
        "  #----------------------#\n",
        "\n",
        "  x = keras.layers.Conv2D(\n",
        "        filters=hp.Choice(\n",
        "                    'filter3_value',\n",
        "                    values=[64, 128, 256],\n",
        "                    default=256\n",
        "                ),\n",
        "        kernel_size=3,\n",
        "        padding=\"same\", \n",
        "        use_bias=False\n",
        "      )(x)\n",
        "  x = keras.layers.BatchNormalization()(x)\n",
        "  x = keras.layers.Activation(\"relu\")(x)\n",
        "\n",
        "  x = keras.layers.Conv2D(\n",
        "        filters=hp.Choice(\n",
        "                    'filter4_value',\n",
        "                    values=[128, 256, 384],\n",
        "                    default=256\n",
        "                ),\n",
        "        kernel_size=3,\n",
        "        padding=\"same\", \n",
        "        use_bias=False\n",
        "      )(x)\n",
        "  x = keras.layers.BatchNormalization()(x)\n",
        "  x = keras.layers.Activation(\"relu\")(x)\n",
        "\n",
        "  x = keras.layers.Conv2D(\n",
        "        filters=hp.Choice(\n",
        "                    'filter5_value',\n",
        "                    values=[64, 128, 256],\n",
        "                    default=256\n",
        "                ),\n",
        "        kernel_size=3,\n",
        "        padding=\"same\", \n",
        "        use_bias=False\n",
        "      )(x)\n",
        "  x = keras.layers.BatchNormalization()(x)\n",
        "  x = keras.layers.Activation(\"relu\")(x)\n",
        "  x = keras.layers.MaxPool2D(2)(x)\n",
        "\n",
        "  #----------------------#\n",
        "\n",
        "  x = keras.layers.Flatten()(x)\n",
        "\n",
        "  hp_units1 = hp.Int('units1', min_value=1024, max_value=4096, step=1024)\n",
        "  x = keras.layers.Dense(units=hp_units1, use_bias=False, kernel_regularizer=keras.regularizers.l2(0.01))(x)\n",
        "  x = keras.layers.BatchNormalization()(x)\n",
        "  x = keras.layers.Activation(\"relu\")(x)\n",
        "\n",
        "  #----------------------#\n",
        "\n",
        "  x = keras.layers.Dropout(\n",
        "        rate=hp.Choice(\n",
        "                  'rate1_value',\n",
        "                  values=[0.3, 0.4, 0.5],\n",
        "                  default=0.4\n",
        "        )\n",
        "      )(x)\n",
        "\n",
        "  #----------------------#\n",
        "\n",
        "  hp_units2 = hp.Int('units2', min_value=1024, max_value=4096, step=1024)\n",
        "  x = keras.layers.Dense(units=hp_units2, use_bias=False, kernel_regularizer=keras.regularizers.l2(0.01))(x)\n",
        "  x = keras.layers.BatchNormalization()(x)\n",
        "  x = keras.layers.Activation(\"relu\")(x)\n",
        "\n",
        "  #----------------------#\n",
        "\n",
        "  x = keras.layers.Dropout(\n",
        "        rate=hp.Choice(\n",
        "                  'rate2_value',\n",
        "                  values=[0.3, 0.4, 0.5],\n",
        "                  default=0.4\n",
        "        )\n",
        "      )(x)\n",
        "  \n",
        "  #----------------------#\n",
        "\n",
        "  output_ = keras.layers.Dense(nbr_class, activation=\"softmax\", kernel_regularizer=keras.regularizers.l2(0.1))(x)\n",
        "\n",
        "  #----------------------#\n",
        "\n",
        "  model = keras.Model(inputs=[input_], outputs=[output_])\n",
        "\n",
        "  #----------------------#\n",
        "\n",
        "  hp_learning_rate = hp.Choice('learning_rate', values=[1e-4, 1e-3])\n",
        "  model.compile(\n",
        "        loss=\"categorical_crossentropy\",\n",
        "        optimizer=tf.keras.optimizers.Adam(learning_rate=hp_learning_rate),\n",
        "        metrics=[\"accuracy\"]\n",
        "  )\n",
        "\n",
        "  #----------------------#\n",
        "\n",
        "  return(model)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FW54NNluvdpr"
      },
      "source": [
        "# We don't want to use a grid search, because it is too slow for our CNN.\n",
        "# Instead, we will use a bayesian optimization.\n",
        "\n",
        "tuner_bayesian = kt.BayesianOptimization(\n",
        "    build_model_10_dogs,\n",
        "    objective='val_accuracy',\n",
        "    max_trials=2,\n",
        "    executions_per_trial=2,\n",
        "    seed=0,\n",
        "    overwrite=True\n",
        ")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gzRXdlLwvdxG",
        "outputId": "ff7171d8-a44d-48d5-e8f8-aceb87fbbedf"
      },
      "source": [
        "# We are using few epochs.\n",
        "# The goal is to find the trends of these hyparameters.\n",
        "n_epochs = 30\n",
        "\n",
        "tuner_bayesian.search(\n",
        "    x_train,\n",
        "    validation_data=x_val,\n",
        "    epochs=n_epochs,\n",
        "    callbacks=[keras.callbacks.EarlyStopping(patience=5)]\n",
        ")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Trial 2 Complete [00h 12m 07s]\n",
            "val_accuracy: 0.3535503000020981\n",
            "\n",
            "Best val_accuracy So Far: 0.3535503000020981\n",
            "Total elapsed time: 00h 17m 41s\n",
            "INFO:tensorflow:Oracle triggered exit\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "V5zt5udT9NjT",
        "outputId": "2aab8523-4431-49c2-f841-d4c937f1d151"
      },
      "source": [
        "# These are the hyperparameters values:\n",
        "best_hyperparameters = tuner_bayesian.get_best_hyperparameters()[0]\n",
        "print(best_hyperparameters.values)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "{'filter1_value': 64, 'kernel1_value': 7, 'filter2_value': 256, 'filter3_value': 64, 'filter4_value': 256, 'filter5_value': 64, 'units1': 2048, 'rate1_value': 0.4, 'units2': 2048, 'rate2_value': 0.4, 'learning_rate': 0.0001}\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "CyPg2gMT9XGU",
        "outputId": "b2f0d016-a7fc-4acf-a810-d0eefbe6a15e"
      },
      "source": [
        "# It will give us this CNN:\n",
        "best_model = tuner_bayesian.get_best_models(num_models=1)[0]\n",
        "print(best_model.summary())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"model\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "input_1 (InputLayer)         [(None, 150, 150, 3)]     0         \n",
            "_________________________________________________________________\n",
            "random_crop (RandomCrop)     (None, 150, 150, 3)       0         \n",
            "_________________________________________________________________\n",
            "conv2d (Conv2D)              (None, 150, 150, 64)      9408      \n",
            "_________________________________________________________________\n",
            "batch_normalization (BatchNo (None, 150, 150, 64)      256       \n",
            "_________________________________________________________________\n",
            "activation (Activation)      (None, 150, 150, 64)      0         \n",
            "_________________________________________________________________\n",
            "max_pooling2d (MaxPooling2D) (None, 75, 75, 64)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_1 (Conv2D)            (None, 75, 75, 256)       147456    \n",
            "_________________________________________________________________\n",
            "batch_normalization_1 (Batch (None, 75, 75, 256)       1024      \n",
            "_________________________________________________________________\n",
            "activation_1 (Activation)    (None, 75, 75, 256)       0         \n",
            "_________________________________________________________________\n",
            "max_pooling2d_1 (MaxPooling2 (None, 37, 37, 256)       0         \n",
            "_________________________________________________________________\n",
            "conv2d_2 (Conv2D)            (None, 37, 37, 64)        147456    \n",
            "_________________________________________________________________\n",
            "batch_normalization_2 (Batch (None, 37, 37, 64)        256       \n",
            "_________________________________________________________________\n",
            "activation_2 (Activation)    (None, 37, 37, 64)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_3 (Conv2D)            (None, 37, 37, 256)       147456    \n",
            "_________________________________________________________________\n",
            "batch_normalization_3 (Batch (None, 37, 37, 256)       1024      \n",
            "_________________________________________________________________\n",
            "activation_3 (Activation)    (None, 37, 37, 256)       0         \n",
            "_________________________________________________________________\n",
            "conv2d_4 (Conv2D)            (None, 37, 37, 64)        147456    \n",
            "_________________________________________________________________\n",
            "batch_normalization_4 (Batch (None, 37, 37, 64)        256       \n",
            "_________________________________________________________________\n",
            "activation_4 (Activation)    (None, 37, 37, 64)        0         \n",
            "_________________________________________________________________\n",
            "max_pooling2d_2 (MaxPooling2 (None, 18, 18, 64)        0         \n",
            "_________________________________________________________________\n",
            "flatten (Flatten)            (None, 20736)             0         \n",
            "_________________________________________________________________\n",
            "dense (Dense)                (None, 2048)              42467328  \n",
            "_________________________________________________________________\n",
            "batch_normalization_5 (Batch (None, 2048)              8192      \n",
            "_________________________________________________________________\n",
            "activation_5 (Activation)    (None, 2048)              0         \n",
            "_________________________________________________________________\n",
            "dropout (Dropout)            (None, 2048)              0         \n",
            "_________________________________________________________________\n",
            "dense_1 (Dense)              (None, 2048)              4194304   \n",
            "_________________________________________________________________\n",
            "batch_normalization_6 (Batch (None, 2048)              8192      \n",
            "_________________________________________________________________\n",
            "activation_6 (Activation)    (None, 2048)              0         \n",
            "_________________________________________________________________\n",
            "dropout_1 (Dropout)          (None, 2048)              0         \n",
            "_________________________________________________________________\n",
            "dense_2 (Dense)              (None, 10)                20490     \n",
            "=================================================================\n",
            "Total params: 47,300,554\n",
            "Trainable params: 47,290,954\n",
            "Non-trainable params: 9,600\n",
            "_________________________________________________________________\n",
            "None\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QmhdGzeg5R47"
      },
      "source": [
        "#----------------------#"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bc36xQtf5Tow"
      },
      "source": [
        "# Now we can use these hyperparameters and train our CNN"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zrWp-_pt9b2w"
      },
      "source": [
        "model = tuner_bayesian.hypermodel.build(best_hyperparameters)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2LgEszi1-Wke"
      },
      "source": [
        "# We will use two callbacks :\n",
        "# early stopping -> in order to avoid overfitting\n",
        "# tensorboard_cb -> it will record some informations inside logs,\n",
        "# it will be used by TensorBoard\n",
        "\n",
        "# Early stopping\n",
        "early_stopping_cb = keras.callbacks.EarlyStopping(patience=10, restore_best_weights=True)\n",
        "\n",
        "# TensorBoard\n",
        "root_logdir = \"/content/drive/MyDrive/Gaetan_Travail/ML/Formation_OC_IML/P6/Projet/logs/\"\n",
        "\n",
        "def get_run_logdir():    \n",
        "    run_id = time.strftime(\"run_%Y_%m_%d-%H_%M_%S\")\n",
        "    return(os.path.join(root_logdir, run_id))\n",
        "\n",
        "tensorboard_cb = keras.callbacks.TensorBoard(get_run_logdir())"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "OHqauYep9dpn",
        "outputId": "19980d6b-cc6d-41c3-b147-370d9c642167"
      },
      "source": [
        "# Training of the model\n",
        "# epochs is very high because we are using an early stopping callback\n",
        "# -> we don't need to focus on the number of epochs, because early stopping will\n",
        "# stop the model before it will overfit.\n",
        "\n",
        "history = model.fit(\n",
        "  x_train,\n",
        "  epochs=1000,\n",
        "  batch_size=batch_size_,\n",
        "  validation_data=x_val,\n",
        "  callbacks=[early_stopping_cb, tensorboard_cb]\n",
        ")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/1000\n",
            "103/103 [==============================] - 14s 123ms/step - loss: 61.9381 - accuracy: 0.1277 - val_loss: 59.5830 - val_accuracy: 0.0976\n",
            "Epoch 2/1000\n",
            "103/103 [==============================] - 12s 114ms/step - loss: 58.5677 - accuracy: 0.2263 - val_loss: 57.4095 - val_accuracy: 0.1006\n",
            "Epoch 3/1000\n",
            "103/103 [==============================] - 12s 115ms/step - loss: 55.2810 - accuracy: 0.2301 - val_loss: 53.4058 - val_accuracy: 0.1450\n",
            "Epoch 4/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 51.8972 - accuracy: 0.2481 - val_loss: 50.2504 - val_accuracy: 0.1450\n",
            "Epoch 5/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 48.6107 - accuracy: 0.2589 - val_loss: 45.9753 - val_accuracy: 0.2692\n",
            "Epoch 6/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 45.1712 - accuracy: 0.2984 - val_loss: 42.9391 - val_accuracy: 0.2663\n",
            "Epoch 7/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 42.0350 - accuracy: 0.2788 - val_loss: 40.0477 - val_accuracy: 0.2189\n",
            "Epoch 8/1000\n",
            "103/103 [==============================] - 12s 115ms/step - loss: 39.0614 - accuracy: 0.2738 - val_loss: 37.4573 - val_accuracy: 0.2574\n",
            "Epoch 9/1000\n",
            "103/103 [==============================] - 12s 114ms/step - loss: 36.0329 - accuracy: 0.3115 - val_loss: 34.1086 - val_accuracy: 0.3225\n",
            "Epoch 10/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 33.4244 - accuracy: 0.3171 - val_loss: 31.9185 - val_accuracy: 0.2663\n",
            "Epoch 11/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 30.7733 - accuracy: 0.3044 - val_loss: 29.1148 - val_accuracy: 0.3077\n",
            "Epoch 12/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 28.3657 - accuracy: 0.3380 - val_loss: 27.5908 - val_accuracy: 0.2337\n",
            "Epoch 13/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 26.0587 - accuracy: 0.3703 - val_loss: 25.1571 - val_accuracy: 0.2840\n",
            "Epoch 14/1000\n",
            "103/103 [==============================] - 12s 114ms/step - loss: 24.1397 - accuracy: 0.3327 - val_loss: 23.5274 - val_accuracy: 0.2870\n",
            "Epoch 15/1000\n",
            "103/103 [==============================] - 12s 114ms/step - loss: 22.2139 - accuracy: 0.3401 - val_loss: 21.1521 - val_accuracy: 0.3817\n",
            "Epoch 16/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 20.5015 - accuracy: 0.3992 - val_loss: 19.5261 - val_accuracy: 0.3136\n",
            "Epoch 17/1000\n",
            "103/103 [==============================] - 12s 114ms/step - loss: 18.8188 - accuracy: 0.3696 - val_loss: 18.3531 - val_accuracy: 0.2515\n",
            "Epoch 18/1000\n",
            "103/103 [==============================] - 12s 114ms/step - loss: 17.4685 - accuracy: 0.3559 - val_loss: 16.9937 - val_accuracy: 0.3136\n",
            "Epoch 19/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 16.1021 - accuracy: 0.4139 - val_loss: 16.2255 - val_accuracy: 0.2633\n",
            "Epoch 20/1000\n",
            "103/103 [==============================] - 12s 116ms/step - loss: 14.9951 - accuracy: 0.4010 - val_loss: 14.7400 - val_accuracy: 0.3107\n",
            "Epoch 21/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 13.8459 - accuracy: 0.4189 - val_loss: 13.9425 - val_accuracy: 0.2574\n",
            "Epoch 22/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 12.9482 - accuracy: 0.3952 - val_loss: 13.9393 - val_accuracy: 0.1982\n",
            "Epoch 23/1000\n",
            "103/103 [==============================] - 12s 114ms/step - loss: 12.1405 - accuracy: 0.3810 - val_loss: 12.2993 - val_accuracy: 0.2751\n",
            "Epoch 24/1000\n",
            "103/103 [==============================] - 12s 114ms/step - loss: 11.2835 - accuracy: 0.4403 - val_loss: 11.3703 - val_accuracy: 0.2929\n",
            "Epoch 25/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 10.5964 - accuracy: 0.4248 - val_loss: 10.6585 - val_accuracy: 0.3018\n",
            "Epoch 26/1000\n",
            "103/103 [==============================] - 12s 114ms/step - loss: 10.0046 - accuracy: 0.4273 - val_loss: 10.1518 - val_accuracy: 0.2751\n",
            "Epoch 27/1000\n",
            "103/103 [==============================] - 12s 112ms/step - loss: 9.4206 - accuracy: 0.4126 - val_loss: 9.7045 - val_accuracy: 0.2870\n",
            "Epoch 28/1000\n",
            "103/103 [==============================] - 12s 114ms/step - loss: 8.8469 - accuracy: 0.4362 - val_loss: 8.8336 - val_accuracy: 0.3994\n",
            "Epoch 29/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 8.3847 - accuracy: 0.4319 - val_loss: 8.9035 - val_accuracy: 0.2515\n",
            "Epoch 30/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 7.9630 - accuracy: 0.4646 - val_loss: 8.5600 - val_accuracy: 0.2663\n",
            "Epoch 31/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 7.5363 - accuracy: 0.4685 - val_loss: 8.0574 - val_accuracy: 0.3136\n",
            "Epoch 32/1000\n",
            "103/103 [==============================] - 12s 112ms/step - loss: 7.1866 - accuracy: 0.4694 - val_loss: 7.5327 - val_accuracy: 0.3284\n",
            "Epoch 33/1000\n",
            "103/103 [==============================] - 12s 114ms/step - loss: 6.8072 - accuracy: 0.4776 - val_loss: 7.5627 - val_accuracy: 0.3018\n",
            "Epoch 34/1000\n",
            "103/103 [==============================] - 12s 114ms/step - loss: 6.5068 - accuracy: 0.4664 - val_loss: 7.0604 - val_accuracy: 0.3047\n",
            "Epoch 35/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 6.3225 - accuracy: 0.4490 - val_loss: 7.4985 - val_accuracy: 0.2485\n",
            "Epoch 36/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 6.0149 - accuracy: 0.4651 - val_loss: 7.2396 - val_accuracy: 0.2485\n",
            "Epoch 37/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 5.7784 - accuracy: 0.4736 - val_loss: 6.1101 - val_accuracy: 0.3462\n",
            "Epoch 38/1000\n",
            "103/103 [==============================] - 12s 114ms/step - loss: 5.4980 - accuracy: 0.5165 - val_loss: 6.3377 - val_accuracy: 0.2692\n",
            "Epoch 39/1000\n",
            "103/103 [==============================] - 12s 115ms/step - loss: 5.3567 - accuracy: 0.4727 - val_loss: 5.8010 - val_accuracy: 0.3373\n",
            "Epoch 40/1000\n",
            "103/103 [==============================] - 12s 116ms/step - loss: 5.0142 - accuracy: 0.5541 - val_loss: 5.5860 - val_accuracy: 0.3728\n",
            "Epoch 41/1000\n",
            "103/103 [==============================] - 12s 116ms/step - loss: 4.9713 - accuracy: 0.5026 - val_loss: 8.2626 - val_accuracy: 0.1953\n",
            "Epoch 42/1000\n",
            "103/103 [==============================] - 12s 116ms/step - loss: 4.7719 - accuracy: 0.5187 - val_loss: 5.4996 - val_accuracy: 0.2929\n",
            "Epoch 43/1000\n",
            "103/103 [==============================] - 12s 115ms/step - loss: 4.5494 - accuracy: 0.5143 - val_loss: 5.1067 - val_accuracy: 0.3757\n",
            "Epoch 44/1000\n",
            "103/103 [==============================] - 12s 114ms/step - loss: 4.4406 - accuracy: 0.5519 - val_loss: 5.2997 - val_accuracy: 0.3432\n",
            "Epoch 45/1000\n",
            "103/103 [==============================] - 12s 112ms/step - loss: 4.2673 - accuracy: 0.5213 - val_loss: 5.2160 - val_accuracy: 0.3225\n",
            "Epoch 46/1000\n",
            "103/103 [==============================] - 12s 115ms/step - loss: 4.2720 - accuracy: 0.4960 - val_loss: 4.9367 - val_accuracy: 0.3373\n",
            "Epoch 47/1000\n",
            "103/103 [==============================] - 12s 115ms/step - loss: 4.1258 - accuracy: 0.5135 - val_loss: 5.5292 - val_accuracy: 0.2811\n",
            "Epoch 48/1000\n",
            "103/103 [==============================] - 12s 114ms/step - loss: 4.0549 - accuracy: 0.4920 - val_loss: 4.9558 - val_accuracy: 0.3728\n",
            "Epoch 49/1000\n",
            "103/103 [==============================] - 12s 115ms/step - loss: 3.8920 - accuracy: 0.5277 - val_loss: 4.6858 - val_accuracy: 0.3639\n",
            "Epoch 50/1000\n",
            "103/103 [==============================] - 12s 114ms/step - loss: 3.7214 - accuracy: 0.5670 - val_loss: 4.3255 - val_accuracy: 0.3905\n",
            "Epoch 51/1000\n",
            "103/103 [==============================] - 12s 114ms/step - loss: 3.6732 - accuracy: 0.5759 - val_loss: 4.1959 - val_accuracy: 0.3964\n",
            "Epoch 52/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 3.5857 - accuracy: 0.5370 - val_loss: 4.1652 - val_accuracy: 0.3757\n",
            "Epoch 53/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 3.5263 - accuracy: 0.5485 - val_loss: 4.6334 - val_accuracy: 0.3343\n",
            "Epoch 54/1000\n",
            "103/103 [==============================] - 12s 114ms/step - loss: 3.4942 - accuracy: 0.5582 - val_loss: 4.3377 - val_accuracy: 0.3669\n",
            "Epoch 55/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 3.3992 - accuracy: 0.5453 - val_loss: 4.1140 - val_accuracy: 0.3580\n",
            "Epoch 56/1000\n",
            "103/103 [==============================] - 12s 114ms/step - loss: 3.3912 - accuracy: 0.5245 - val_loss: 4.0126 - val_accuracy: 0.3639\n",
            "Epoch 57/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 3.2783 - accuracy: 0.5436 - val_loss: 3.9191 - val_accuracy: 0.4053\n",
            "Epoch 58/1000\n",
            "103/103 [==============================] - 12s 112ms/step - loss: 3.1890 - accuracy: 0.5790 - val_loss: 4.2774 - val_accuracy: 0.2751\n",
            "Epoch 59/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 3.1364 - accuracy: 0.5865 - val_loss: 4.0570 - val_accuracy: 0.3432\n",
            "Epoch 60/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 3.1774 - accuracy: 0.5488 - val_loss: 3.9591 - val_accuracy: 0.3343\n",
            "Epoch 61/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 3.0779 - accuracy: 0.5968 - val_loss: 4.3082 - val_accuracy: 0.3047\n",
            "Epoch 62/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 3.0650 - accuracy: 0.5845 - val_loss: 3.8638 - val_accuracy: 0.3935\n",
            "Epoch 63/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 3.0485 - accuracy: 0.5445 - val_loss: 3.9096 - val_accuracy: 0.4024\n",
            "Epoch 64/1000\n",
            "103/103 [==============================] - 12s 114ms/step - loss: 3.0798 - accuracy: 0.5389 - val_loss: 3.8218 - val_accuracy: 0.3905\n",
            "Epoch 65/1000\n",
            "103/103 [==============================] - 12s 114ms/step - loss: 2.9451 - accuracy: 0.5982 - val_loss: 3.7346 - val_accuracy: 0.3757\n",
            "Epoch 66/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 2.9687 - accuracy: 0.5730 - val_loss: 4.0109 - val_accuracy: 0.3166\n",
            "Epoch 67/1000\n",
            "103/103 [==============================] - 12s 115ms/step - loss: 2.8620 - accuracy: 0.6146 - val_loss: 3.5714 - val_accuracy: 0.4053\n",
            "Epoch 68/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 2.9434 - accuracy: 0.5832 - val_loss: 3.8023 - val_accuracy: 0.3550\n",
            "Epoch 69/1000\n",
            "103/103 [==============================] - 12s 114ms/step - loss: 2.8764 - accuracy: 0.5925 - val_loss: 3.6163 - val_accuracy: 0.4172\n",
            "Epoch 70/1000\n",
            "103/103 [==============================] - 12s 114ms/step - loss: 2.8639 - accuracy: 0.5976 - val_loss: 4.4789 - val_accuracy: 0.3373\n",
            "Epoch 71/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 2.7756 - accuracy: 0.6159 - val_loss: 3.4247 - val_accuracy: 0.4053\n",
            "Epoch 72/1000\n",
            "103/103 [==============================] - 12s 114ms/step - loss: 2.8063 - accuracy: 0.5869 - val_loss: 3.3097 - val_accuracy: 0.4497\n",
            "Epoch 73/1000\n",
            "103/103 [==============================] - 12s 115ms/step - loss: 2.8348 - accuracy: 0.6043 - val_loss: 3.7621 - val_accuracy: 0.3728\n",
            "Epoch 74/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 2.8098 - accuracy: 0.6086 - val_loss: 3.5211 - val_accuracy: 0.3935\n",
            "Epoch 75/1000\n",
            "103/103 [==============================] - 12s 114ms/step - loss: 2.8653 - accuracy: 0.5723 - val_loss: 3.5456 - val_accuracy: 0.4379\n",
            "Epoch 76/1000\n",
            "103/103 [==============================] - 12s 114ms/step - loss: 2.7006 - accuracy: 0.6349 - val_loss: 3.7830 - val_accuracy: 0.3166\n",
            "Epoch 77/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 2.7236 - accuracy: 0.6081 - val_loss: 4.2319 - val_accuracy: 0.3136\n",
            "Epoch 78/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 2.8078 - accuracy: 0.6027 - val_loss: 4.4064 - val_accuracy: 0.3343\n",
            "Epoch 79/1000\n",
            "103/103 [==============================] - 12s 112ms/step - loss: 2.7539 - accuracy: 0.6006 - val_loss: 4.1039 - val_accuracy: 0.2988\n",
            "Epoch 80/1000\n",
            "103/103 [==============================] - 12s 112ms/step - loss: 2.7523 - accuracy: 0.5896 - val_loss: 4.0435 - val_accuracy: 0.3343\n",
            "Epoch 81/1000\n",
            "103/103 [==============================] - 12s 113ms/step - loss: 2.8215 - accuracy: 0.5781 - val_loss: 3.6746 - val_accuracy: 0.3669\n",
            "Epoch 82/1000\n",
            "103/103 [==============================] - 12s 114ms/step - loss: 2.7537 - accuracy: 0.5764 - val_loss: 3.6154 - val_accuracy: 0.3787\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bkY4ubVd9dsJ"
      },
      "source": [
        "# Saving the model\n",
        "model.save(\"/content/drive/MyDrive/Gaetan_Travail/ML/Formation_OC_IML/P6/Projet/dogs_classification_10_optimisation_08042021.h5\")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "O16L56DP9du9",
        "outputId": "133b321e-0031-4d1b-c3e4-4873016c1919"
      },
      "source": [
        "loss, acc = model.evaluate(x_train, verbose=0)\n",
        "print(\"Train set: \")\n",
        "print(\"loss: \", loss)\n",
        "print(\"accuracy: \", acc)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Train set: \n",
            "loss:  2.679717540740967\n",
            "accuracy:  0.6370731592178345\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "-shrnG1u9sEK",
        "outputId": "92ecae15-5faf-45b7-a0c5-fde2e4bdf408"
      },
      "source": [
        "loss, acc = model.evaluate(x_val, verbose=0)\n",
        "print(\"Validation set: \")\n",
        "print(\"loss: \", loss)\n",
        "print(\"accuracy: \", acc)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Validation set: \n",
            "loss:  3.3096959590911865\n",
            "accuracy:  0.4497041404247284\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "lLp0LIe39sJS",
        "outputId": "dafdba7e-255d-4132-d3c9-be5dbf44ef79"
      },
      "source": [
        "loss, acc = model.evaluate(x_test, verbose=0)\n",
        "print(\"Test set: \")\n",
        "print(\"loss: \", loss)\n",
        "print(\"accuracy: \", acc)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Test set: \n",
            "loss:  3.362010955810547\n",
            "accuracy:  0.43714284896850586\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_lfIv2ps6roK"
      },
      "source": [
        "'''\n",
        "We build a CNN, which reach 43.71% accuracy on the test set (10 dog breeds).\n",
        "'''"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "r47Vh1CHRxWA"
      },
      "source": [
        "#----------------------#"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SrCqmvF9gZOq"
      },
      "source": [
        "# Visualization using TensorBoard"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JVZnTTzccaa2"
      },
      "source": [
        "# Load the TensorBoard notebook extension\n",
        "# %load_ext tensorboard"
      ],
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LW1FnL2Wb8cu"
      },
      "source": [
        "# %tensorboard --logdir \"/content/drive/MyDrive/Gaetan_Travail/ML/Formation_OC_IML/P6/Projet/logs/\""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "C-RtQKSa4kQ1"
      },
      "source": [
        "'''\n",
        "!tensorboard dev upload \\\n",
        "  --logdir \"/content/drive/MyDrive/Gaetan_Travail/ML/Formation_OC_IML/P6/Projet/logs/\" \\\n",
        "  --name \"CNN from scratch\" \\\n",
        "  --description \"TensorBoard visualization\" \\\n",
        "  --one_shot\n",
        "'''"
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}