{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jAFeIqsrGwqM"
      },
      "source": [
        "## Food Vision 101 Object Classification Project - In this notebook, we're going be running a series of machine learning experiments (5) to see if it is possible to beat the Food101 paper: https://data.vision.ee.ethz.ch/cvl/datasets_extra/food-101/static/bossard_eccv14_food-101.pdf\n",
        "\n",
        "The dataset that we're going to be working with is the Food101 dataset from TensorFlow datasets - https://www.tensorflow.org/datasets/catalog/food101"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JyMDkPshN02E"
      },
      "source": [
        "## GPU access"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "pBpV2iPPGjsu",
        "outputId": "7f57f728-63b6-4108-965f-9a2090d5cf25"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Tue Jul  8 09:52:59 2025       \n",
            "+-----------------------------------------------------------------------------------------+\n",
            "| NVIDIA-SMI 550.54.15              Driver Version: 550.54.15      CUDA Version: 12.4     |\n",
            "|-----------------------------------------+------------------------+----------------------+\n",
            "| GPU  Name                 Persistence-M | Bus-Id          Disp.A | Volatile Uncorr. ECC |\n",
            "| Fan  Temp   Perf          Pwr:Usage/Cap |           Memory-Usage | GPU-Util  Compute M. |\n",
            "|                                         |                        |               MIG M. |\n",
            "|=========================================+========================+======================|\n",
            "|   0  Tesla T4                       Off |   00000000:00:04.0 Off |                    0 |\n",
            "| N/A   47C    P8              9W /   70W |       0MiB /  15360MiB |      0%      Default |\n",
            "|                                         |                        |                  N/A |\n",
            "+-----------------------------------------+------------------------+----------------------+\n",
            "                                                                                         \n",
            "+-----------------------------------------------------------------------------------------+\n",
            "| Processes:                                                                              |\n",
            "|  GPU   GI   CI        PID   Type   Process name                              GPU Memory |\n",
            "|        ID   ID                                                               Usage      |\n",
            "|=========================================================================================|\n",
            "|  No running processes found                                                             |\n",
            "+-----------------------------------------------------------------------------------------+\n"
          ]
        }
      ],
      "source": [
        "# Confirm access to a GPU\n",
        "!nvidia-smi"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mvqjrNxpOhTs"
      },
      "source": [
        "## Donwloading helper functions (`helper_functions.py`) for project"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "SFDxd2TwOLlQ",
        "outputId": "5a9a76d4-6c0f-4aca-a254-268c7c327f2c"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "--2025-07-08 09:52:59--  https://raw.githubusercontent.com/mrdbourke/tensorflow-deep-learning/refs/heads/main/extras/helper_functions.py\n",
            "Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 185.199.108.133, 185.199.110.133, 185.199.109.133, ...\n",
            "Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|185.199.108.133|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 10246 (10K) [text/plain]\n",
            "Saving to: ‘helper_functions.py’\n",
            "\n",
            "helper_functions.py 100%[===================>]  10.01K  --.-KB/s    in 0.001s  \n",
            "\n",
            "2025-07-08 09:52:59 (17.1 MB/s) - ‘helper_functions.py’ saved [10246/10246]\n",
            "\n"
          ]
        }
      ],
      "source": [
        "# Download helper_functions.py script\n",
        "!wget https://raw.githubusercontent.com/mrdbourke/tensorflow-deep-learning/refs/heads/main/extras/helper_functions.py"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "INX7ipwuO6TX"
      },
      "outputs": [],
      "source": [
        "# Import necessary helper functions from script\n",
        "from helper_functions import plot_loss_curves, compare_historys, create_tensorboard_callback"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "sGW7Rl5JPzuF"
      },
      "source": [
        "## Downloading Food101 dataset from TensorFlow"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5qofGDoKPs9_",
        "outputId": "2448e441-71ed-44b9-e232-1bee98584f80"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "True\n"
          ]
        }
      ],
      "source": [
        "import tensorflow_datasets as tfds, tensorflow as tf\n",
        "all_datasets = tfds.list_builders() # List all of the avaibale datasets in TensorFlow\n",
        "print(f'food101' in all_datasets) # Verify that food101 is the dataset we're looking for"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 165,
          "referenced_widgets": [
            "d2c469c406b34d908bce2005d83fc71a",
            "070c153b47a5446ead4cfd33e445a353",
            "d9306a816f214abdb8cf574ebf709b46",
            "fe8de247b5cb425583c5bf27c9af8ba6",
            "c5755b7ad51d42cca9e7e988eeddf5f3",
            "672b3d9d101c4f12aca1150a5dadcbfa",
            "123b186b38ca41d79d80911dbe903fa2",
            "2f2d4c9c52f04650ada8fa9fb73dc88e",
            "a24bc721a5604b2897a95dbb9a939a7b",
            "eea00e87224d40058841cae29bc31422",
            "50198740d8754fd0b5f981d2f5d18216",
            "cefc3ab00ef54545b999e9496a3cdafc",
            "538a975e0f8a459693beef53cc29bdaf",
            "d1135c2c009e4855a943961e72d2a27f",
            "f0440086fff94d1884427fcbe04891a5",
            "bd2cf466ea694aa89570b3fbd138b898",
            "cebed4658448496f8e55d3802793a1f4",
            "14a4526363c4466e869fa443f97f6e6b",
            "36846b004e114d7b944a613467e681f8",
            "84ff016157be4925a04b58d95f082789",
            "d6a592cd794c4ee690ba191ed79e1e7c",
            "de4ec5033d534ffbb02d69042306a150",
            "a95ac3001e114027a9f1977b3e0174b7",
            "b9acf517ac3b4fe58765be83a1d04bef",
            "09267f5a51214c1793335127a9c24735",
            "1a6de41681eb4ebcbfa7ae2221b21ccf",
            "b46ee0d834be43448a06c5807e69bf70",
            "dce7c2b873bf41868f1f3b0e5fdf338e",
            "7db0bc2f67b741ae97af7109246ea9dc",
            "c1de916f690a4c669c1289260898deaa",
            "5c8f1397fe64470c8772d69eceef0f98",
            "b6ab17db1dca4d1aa0b357aebb3f2bc6",
            "08ff321340884cc59e7b6980b4f7eaed",
            "328c21f61fc14bdaa1b9696ad401a4d1",
            "74ef17eca5494ef8888f6c0368f240bb",
            "6a066e89ce244a218ed69d07614eea96",
            "e4ec96b177e5463b833d93c6160a83d2",
            "dcee912aca734fefa5960c02a103812a",
            "c044f15da41d422b897789d3a23fff1f",
            "68563007c4ec4f17a55b4564c477c89f",
            "a1f9cc1bfd9c453aaa2f13063efba2b9",
            "016ab7c6c8a94e9e8b649b6f9282ac50",
            "dae95b3ad15b49458c9cc97db6c6e416",
            "9c0b49eb1eed43d0b3883a9ec0fb7687",
            "88d9d16dba9f4b3696a8e6bdb84342f7",
            "a32e6069c86f49369b885f1779b77b69",
            "182b4d8b21a24c02a5f406cb80afadb9",
            "6ccd2305901b4c159dc8d94b7eee56d1",
            "a1fa11f7ffcf4d5baa25029746a85359",
            "2548881164074759b5821b383a0868ce",
            "bb611b2e49c54beeaa3fd8179a02d02f",
            "db3ee37a37394fe4958f20955df03678",
            "39ad3e2e877b4ee387803eee4205471f",
            "582f767e6ae34cc3b4d4609b659188d3",
            "7e4f4bd09072432eb46c10af0049691a",
            "29244d4e3cce4cb6986375a06c5fb989",
            "99aaa93baa204138a8c99a15767b2255",
            "c5c41b46de524f6f9b9d108d0268962e",
            "3c4d25ceec3f4751862e7b214c3c7a00",
            "be590df5cf50448ea48b256e49cf03e5",
            "0427d2980c1b4fd9bdd8b8d8e39106e1",
            "1e10c9596e0440f6a6ce8355d2c6b839",
            "a96afb54c73a4d3cb9dec2f740e5d267",
            "124f26e4a08740ba86716949ff9a9b63",
            "1a2cdf5865f546e5b8dcec5b53411295",
            "bdbe8af8fd074e9b95ee694361bdc741",
            "d3fa957626fd4429a23ab4d5703c8c2b",
            "28933e23e5a842dcb65a3ffb35830eb1",
            "f4cbbe4f314d497c8fd8e1fa1eb7fe2a",
            "32611d03af3a4e8ba239c55222071295",
            "c101958924e84cf49d7736f0c1d38e61",
            "56693237f9494c61b06c67b647002153",
            "70f18fa7f150471a947f02aefc63490f",
            "f9cb6e7146a5408dba4df4cb9b234d75",
            "f2f53d28190b472c9948f1f514fda1aa",
            "d115a283aaad4c9285e235d8c3885da1",
            "8a1f8bf626d7477283149c8fe00174d8",
            "98b497980f1d4c42b3f236a2d5fe9d95",
            "806c8729bdea4e16957e7fc5c02e6a9c",
            "b68187ae5cfc413a90160a2e3acc7490",
            "4c1b26542bbe451bb17e74b9806c0778",
            "013b4b8a73a248daa2f990e88b7507e5",
            "5345a752413f466da4ad7e26209b563c",
            "a8fc4228c6b745479eacb2e84175b0e3",
            "4317dc663eb641ed82450c65aa318057",
            "adbf1bd6d7ce49a797c1e9f97be51aaf",
            "020c5acfa0b944798b25a4718e4a20fa",
            "84f6e58b755f4377bba18414d91daf11"
          ]
        },
        "id": "OCIRXKW5QYCl",
        "outputId": "4fef2622-66b0-446a-b32d-70015c4ad17b"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "WARNING:absl:Variant folder /root/tensorflow_datasets/food101/2.0.0 has no dataset_info.json\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading and preparing dataset Unknown size (download: Unknown size, generated: Unknown size, total: Unknown size) to /root/tensorflow_datasets/food101/2.0.0...\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "Dl Completed...: 0 url [00:00, ? url/s]"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "d2c469c406b34d908bce2005d83fc71a"
            }
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "Dl Size...: 0 MiB [00:00, ? MiB/s]"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "cefc3ab00ef54545b999e9496a3cdafc"
            }
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "Extraction completed...: 0 file [00:00, ? file/s]"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "a95ac3001e114027a9f1977b3e0174b7"
            }
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "Generating splits...:   0%|          | 0/2 [00:00<?, ? splits/s]"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "328c21f61fc14bdaa1b9696ad401a4d1"
            }
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "Generating train examples...: 0 examples [00:00, ? examples/s]"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "88d9d16dba9f4b3696a8e6bdb84342f7"
            }
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "Shuffling /root/tensorflow_datasets/food101/incomplete.F7OCNN_2.0.0/food101-train.tfrecord*...:   0%|         …"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "29244d4e3cce4cb6986375a06c5fb989"
            }
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "Generating validation examples...: 0 examples [00:00, ? examples/s]"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "d3fa957626fd4429a23ab4d5703c8c2b"
            }
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "Shuffling /root/tensorflow_datasets/food101/incomplete.F7OCNN_2.0.0/food101-validation.tfrecord*...:   0%|    …"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "98b497980f1d4c42b3f236a2d5fe9d95"
            }
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Dataset food101 downloaded and prepared to /root/tensorflow_datasets/food101/2.0.0. Subsequent calls will reuse this data.\n"
          ]
        }
      ],
      "source": [
        "# Load in Food101 data\n",
        "(train_data, test_data), ds_info = tfds.load(name = 'food101', split = ['train', 'validation'], as_supervised = True, with_info = True)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ocSo65Y0RbY-"
      },
      "source": [
        "## Create preprocessing functions for our data"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "BDryi5Gbcs0X"
      },
      "outputs": [],
      "source": [
        "from tensorflow.keras import layers\n",
        "\n",
        "# Define data augmentation layer\n",
        "data_augmentation = tf.keras.Sequential([\n",
        "    layers.RandomFlip('horizontal'),\n",
        "    layers.RandomRotation(0.1),\n",
        "    layers.RandomZoom(0.1),\n",
        "], name = 'data_augmentation')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "eL2OD9BDRCDX"
      },
      "outputs": [],
      "source": [
        "def preprocess_data(image, label, image_size = 224):\n",
        "  image = tf.image.resize(image, [image_size, image_size]) # Reshape target image to 224x224\n",
        "  image = tf.cast(image, dtype = tf.float32)\n",
        "  image = data_augmentation(image) # Pass image through data augmentation layer\n",
        "  return image, label"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mTEtimeOUi5w",
        "outputId": "a398292b-5e7a-42c1-b8c0-e26e58377475"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Original image before preprocessing: \n",
            "[[[233 253 251]\n",
            "  [231 253 250]\n",
            "  [228 254 251]\n",
            "  ...\n",
            "  [ 85  82  89]\n",
            "  [ 68  67  75]\n",
            "  [ 57  57  67]]\n",
            "\n",
            " [[232 254 252]\n",
            "  [229 254 251]\n",
            "  [226 255 251]\n",
            "  ...\n",
            "  [121 116 120]\n",
            "  [100  99 104]\n",
            "  [ 86  85  91]]]\n",
            "Original image shape: \n",
            "(512, 512, 3)\n",
            "Original image datatype: <dtype: 'uint8'>\n",
            "Preprocessed image: \n",
            "[[[ 86.03539   95.39833  111.06804 ]\n",
            "  [ 81.26731   91.58968  112.24811 ]\n",
            "  [ 73.00026   85.95603  109.79344 ]\n",
            "  ...\n",
            "  [236.12347  197.25575  102.01086 ]\n",
            "  [232.18433  194.47562   98.82944 ]\n",
            "  [231.21236  194.17421   98.00889 ]]\n",
            "\n",
            " [[150.72676  148.40372  134.82703 ]\n",
            "  [153.93556  154.36476  142.41916 ]\n",
            "  [151.76395  154.18515  143.17636 ]\n",
            "  ...\n",
            "  [238.04805  198.88608  106.111496]\n",
            "  [229.41438  191.63219   98.988785]\n",
            "  [216.67569  180.16093   88.57987 ]]]\n",
            "Preprocessed image shape: \n",
            "(224, 224, 3)\n",
            "Preprocessed image datatype: \n",
            "<dtype: 'float32'>\n"
          ]
        }
      ],
      "source": [
        "# Get a sample image and label from dataset\n",
        "for image, label in train_data.take(1):\n",
        "  transformed_image = preprocess_data(image, label)\n",
        "\n",
        "# Access the preprocessed image and label from tuple\n",
        "preprocessed_image, label = transformed_image\n",
        "\n",
        "# Print out image before and after prepcrocessing & shapes\n",
        "print(f'Original image before preprocessing: \\n{image[:2]}')\n",
        "print(f'Original image shape: \\n{image.shape}')\n",
        "print(f'Original image datatype: {image.dtype}')\n",
        "print(f'Preprocessed image: \\n{preprocessed_image[:2]}')\n",
        "print(f'Preprocessed image shape: \\n{preprocessed_image.shape}')\n",
        "print(f'Preprocessed image datatype: \\n{preprocessed_image.dtype}')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "UX54aZPJ8MhT",
        "outputId": "3e17ff6e-bbb0-4bb3-c58a-ead5eecdadbc"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Label: 56\n"
          ]
        }
      ],
      "source": [
        "for image, label in train_data.take(1):\n",
        "  print(f'Label: {label.numpy()}')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6sCEqdhEXuUr"
      },
      "source": [
        "## Turning our datasets into batches and prefetched datasets (so they run as fast as possible)\n",
        "\n",
        "We're going to be turning our datasets into efficient data pipelines so that upon execution, they are running as fast as possbile: https://www.tensorflow.org/guide/data"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "gCKj_ASXdGzw"
      },
      "outputs": [],
      "source": [
        "# Map preprocessing function, shuffle and parallelise it to training data\n",
        "train_data = train_data.map(map_func = preprocess_data, num_parallel_calls = tf.data.AUTOTUNE)\n",
        "train_data = train_data.shuffle(1000).batch(32).prefetch(tf.data.AUTOTUNE)\n",
        "\n",
        "# Validation/test pipeline\n",
        "test_data = test_data.map(map_func = preprocess_data, num_parallel_calls = tf.data.AUTOTUNE)\n",
        "test_data = test_data.batch(32).prefetch(tf.data.AUTOTUNE)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "j2Ft5JRydrDs"
      },
      "source": [
        "## Create a checkpoint callback for our model"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "wqRHV_V1bx50"
      },
      "outputs": [],
      "source": [
        "# Import tensorboard callback from script\n",
        "from helper_functions import create_tensorboard_callback\n",
        "\n",
        "# Setup model checkpoint path to save all of models' progress during training\n",
        "checkpoint_path = 'model_checkpoint.weights.h5'\n",
        "model_checkpoint = tf.keras.callbacks.ModelCheckpoint(filepath = checkpoint_path, save_weights_only = True, monitor = 'val_accuracy', save_best_only = True)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "EjOx2dY_fDeC"
      },
      "source": [
        "## Mixed precision training (for speeding up model performance)\n",
        "\n",
        "Mixed training uses both `tf.float32` and `tf.float16` datatypes for speeding up a model's performance: https://www.tensorflow.org/guide/mixed_precision"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "zdxOfcGee5p5",
        "outputId": "ea7f53a4-64a6-4b43-8773-48591a78c384"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<DTypePolicy \"mixed_float16\">"
            ]
          },
          "metadata": {},
          "execution_count": 12
        }
      ],
      "source": [
        "from tensorflow.keras import mixed_precision\n",
        "mixed_precision.set_global_policy('mixed_float16') # Set global data policy for mixed precision\n",
        "mixed_precision.global_policy()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7xqhbyG8gdwM"
      },
      "source": [
        "## Model 1: Pretrained feature-extactor model (EfficientNetB0)\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 420
        },
        "id": "jihCPRersGZz",
        "outputId": "1fb704bd-9d57-48dc-d245-4617b1c86d61"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading data from https://storage.googleapis.com/keras-applications/efficientnetb0_notop.h5\n",
            "\u001b[1m16705208/16705208\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 0us/step\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "\u001b[1mModel: \"functional_1\"\u001b[0m\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\">Model: \"functional_1\"</span>\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
              "┃\u001b[1m \u001b[0m\u001b[1mLayer (type)                   \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mOutput Shape          \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m      Param #\u001b[0m\u001b[1m \u001b[0m┃\n",
              "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
              "│ input_layer (\u001b[38;5;33mInputLayer\u001b[0m)        │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m3\u001b[0m)    │             \u001b[38;5;34m0\u001b[0m │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ efficientnetb0 (\u001b[38;5;33mFunctional\u001b[0m)     │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m7\u001b[0m, \u001b[38;5;34m7\u001b[0m, \u001b[38;5;34m1280\u001b[0m)     │     \u001b[38;5;34m4,049,571\u001b[0m │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ global_average_pooling2d        │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m1280\u001b[0m)           │             \u001b[38;5;34m0\u001b[0m │\n",
              "│ (\u001b[38;5;33mGlobalAveragePooling2D\u001b[0m)        │                        │               │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ batch_normalization             │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m1280\u001b[0m)           │         \u001b[38;5;34m5,120\u001b[0m │\n",
              "│ (\u001b[38;5;33mBatchNormalization\u001b[0m)            │                        │               │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ dense (\u001b[38;5;33mDense\u001b[0m)                   │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m256\u001b[0m)            │       \u001b[38;5;34m327,936\u001b[0m │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ dropout (\u001b[38;5;33mDropout\u001b[0m)               │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m256\u001b[0m)            │             \u001b[38;5;34m0\u001b[0m │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ output_layer (\u001b[38;5;33mDense\u001b[0m)            │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m101\u001b[0m)            │        \u001b[38;5;34m25,957\u001b[0m │\n",
              "└─────────────────────────────────┴────────────────────────┴───────────────┘\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\">┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
              "┃<span style=\"font-weight: bold\"> Layer (type)                    </span>┃<span style=\"font-weight: bold\"> Output Shape           </span>┃<span style=\"font-weight: bold\">       Param # </span>┃\n",
              "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
              "│ input_layer (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">InputLayer</span>)        │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">224</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">224</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">3</span>)    │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ efficientnetb0 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Functional</span>)     │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">7</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">7</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">1280</span>)     │     <span style=\"color: #00af00; text-decoration-color: #00af00\">4,049,571</span> │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ global_average_pooling2d        │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">1280</span>)           │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
              "│ (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">GlobalAveragePooling2D</span>)        │                        │               │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ batch_normalization             │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">1280</span>)           │         <span style=\"color: #00af00; text-decoration-color: #00af00\">5,120</span> │\n",
              "│ (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">BatchNormalization</span>)            │                        │               │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ dense (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)                   │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">256</span>)            │       <span style=\"color: #00af00; text-decoration-color: #00af00\">327,936</span> │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ dropout (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dropout</span>)               │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">256</span>)            │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ output_layer (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)            │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">101</span>)            │        <span style=\"color: #00af00; text-decoration-color: #00af00\">25,957</span> │\n",
              "└─────────────────────────────────┴────────────────────────┴───────────────┘\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "\u001b[1m Total params: \u001b[0m\u001b[38;5;34m4,408,584\u001b[0m (16.82 MB)\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Total params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">4,408,584</span> (16.82 MB)\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m356,453\u001b[0m (1.36 MB)\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">356,453</span> (1.36 MB)\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m4,052,131\u001b[0m (15.46 MB)\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Non-trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">4,052,131</span> (15.46 MB)\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        }
      ],
      "source": [
        "from tensorflow.keras import layers\n",
        "\n",
        "# Setup data inputs\n",
        "input_size = (224, 224, 3)\n",
        "\n",
        "# Get number of classes\n",
        "num_classes = ds_info.features['label'].num_classes\n",
        "\n",
        "# Create feature extractor model with Functional API\n",
        "base_model = tf.keras.applications.EfficientNetB0(include_top = False, input_shape = (224, 224, 3), weights = 'imagenet')\n",
        "base_model.trainable = False\n",
        "\n",
        "inputs = layers.Input(shape = (224, 224, 3), name = 'input_layer')\n",
        "x = base_model(inputs, training = False)\n",
        "x = layers.GlobalAveragePooling2D()(x)\n",
        "x = layers.BatchNormalization()(x)\n",
        "x = layers.Dense(256, activation = 'relu')(x)\n",
        "x = layers.Dropout(0.5)(x)\n",
        "outputs = layers.Dense(num_classes, activation = 'softmax', name = 'output_layer')(x)\n",
        "model_1 = tf.keras.Model(inputs, outputs)\n",
        "\n",
        "# Conmpile model\n",
        "model_1.compile(loss = tf.keras.losses.SparseCategoricalCrossentropy(from_logits = False),\n",
        "                optimizer = tf.keras.optimizers.Adam(),\n",
        "                metrics = ['accuracy'])\n",
        "\n",
        "# Get model summary\n",
        "model_1.summary()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ELOqch4WvE2Y",
        "outputId": "0d347fd8-4e22-4f2b-fa0d-9f91db8989a9"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Saving TensorBoard log files to: ./tensorflow_logs/food_101_all_modelling_experiments/model_1/20250708-100054\n",
            "Epoch 1/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m244s\u001b[0m 91ms/step - accuracy: 0.2960 - loss: 3.1697 - val_accuracy: 0.6120 - val_loss: 1.4577\n",
            "Epoch 2/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m200s\u001b[0m 84ms/step - accuracy: 0.4623 - loss: 2.1737 - val_accuracy: 0.6253 - val_loss: 1.4035\n",
            "Epoch 3/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m208s\u001b[0m 87ms/step - accuracy: 0.4867 - loss: 2.0628 - val_accuracy: 0.6419 - val_loss: 1.3437\n",
            "Epoch 4/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m201s\u001b[0m 84ms/step - accuracy: 0.4965 - loss: 1.9946 - val_accuracy: 0.6472 - val_loss: 1.3284\n",
            "Epoch 5/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m199s\u001b[0m 83ms/step - accuracy: 0.5080 - loss: 1.9510 - val_accuracy: 0.6425 - val_loss: 1.3027\n",
            "Epoch 6/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m201s\u001b[0m 84ms/step - accuracy: 0.5174 - loss: 1.9139 - val_accuracy: 0.6510 - val_loss: 1.2882\n",
            "Epoch 7/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m200s\u001b[0m 83ms/step - accuracy: 0.5204 - loss: 1.8912 - val_accuracy: 0.6650 - val_loss: 1.2645\n",
            "Epoch 8/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m199s\u001b[0m 83ms/step - accuracy: 0.5282 - loss: 1.8518 - val_accuracy: 0.6584 - val_loss: 1.2690\n",
            "Epoch 9/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m200s\u001b[0m 83ms/step - accuracy: 0.5380 - loss: 1.8233 - val_accuracy: 0.6684 - val_loss: 1.2400\n",
            "Epoch 10/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m199s\u001b[0m 83ms/step - accuracy: 0.5388 - loss: 1.7956 - val_accuracy: 0.6615 - val_loss: 1.2476\n"
          ]
        }
      ],
      "source": [
        "# Fit\n",
        "initial_epochs = 10\n",
        "history_1 = model_1.fit(train_data, epochs = initial_epochs, steps_per_epoch = len(train_data), validation_data = test_data, validation_steps = int(0.15 *len(test_data)), callbacks = [create_tensorboard_callback(dir_name = './tensorflow_logs', experiment_name = 'food_101_all_modelling_experiments/model_1'), model_checkpoint])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 927
        },
        "id": "GKW4lDynv2-Y",
        "outputId": "3647cfe4-a7fc-40e6-80d8-68e4c261590d"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "# Plot model loss curves\n",
        "plot_loss_curves(history = history_1)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "600wEUVO4TrW",
        "outputId": "2848a00c-aee3-4701-922f-10e05d298ad5"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\u001b[1m790/790\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m68s\u001b[0m 86ms/step - accuracy: 0.6648 - loss: 1.2459\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[1.2509543895721436, 0.6629306674003601]"
            ]
          },
          "metadata": {},
          "execution_count": 16
        }
      ],
      "source": [
        "# Evaluate model on test data\n",
        "model_1.evaluate(test_data)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "em8uk8qc4e79",
        "outputId": "2c7bbb36-ee81-46f4-fbae-353258077a10"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\u001b[1m790/790\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m74s\u001b[0m 85ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[4.753e-03, 8.345e-07, 8.053e-05, 1.013e-06, 1.824e-05, 1.717e-05,\n",
              "        4.470e-06, 0.000e+00, 5.939e-03, 4.768e-07, 1.639e-04, 5.960e-08,\n",
              "        4.508e-04, 6.285e-04, 8.575e-03, 4.113e-05, 2.313e-02, 9.567e-05,\n",
              "        5.960e-08, 4.768e-07, 7.749e-07, 1.599e-03, 1.059e-02, 4.554e-05,\n",
              "        2.384e-07, 2.205e-06, 8.345e-07, 8.315e-05, 8.643e-06, 7.056e-01,\n",
              "        2.186e-03, 8.118e-05, 6.590e-04, 0.000e+00, 1.267e-04, 1.788e-06,\n",
              "        3.493e-05, 1.252e-06, 0.000e+00, 2.038e-05, 2.921e-06, 2.980e-07,\n",
              "        2.260e-04, 2.146e-06, 7.749e-07, 6.252e-03, 6.258e-06, 9.537e-06,\n",
              "        4.172e-07, 8.047e-06, 2.384e-07, 0.000e+00, 1.788e-06, 9.239e-06,\n",
              "        0.000e+00, 1.317e-05, 5.364e-06, 6.855e-06, 2.454e-02, 1.609e-06,\n",
              "        2.980e-07, 3.576e-07, 4.768e-06, 7.683e-05, 0.000e+00, 4.172e-07,\n",
              "        8.643e-06, 1.788e-07, 1.073e-06, 1.562e-05, 0.000e+00, 2.384e-07,\n",
              "        4.745e-04, 4.925e-03, 1.073e-06, 0.000e+00, 5.126e-06, 6.199e-06,\n",
              "        9.358e-06, 6.557e-07, 2.086e-06, 0.000e+00, 1.442e-05, 5.411e-02,\n",
              "        1.192e-06, 1.490e-06, 2.551e-05, 2.433e-04, 0.000e+00, 3.576e-07,\n",
              "        5.960e-08, 0.000e+00, 1.788e-07, 5.245e-06, 1.032e-01, 1.947e-04,\n",
              "        9.537e-07, 1.854e-05, 3.807e-03, 2.873e-05, 3.653e-02]],\n",
              "      dtype=float16)"
            ]
          },
          "metadata": {},
          "execution_count": 17
        }
      ],
      "source": [
        "# Make predictions with model\n",
        "model_1_preds = model_1.predict(test_data)\n",
        "model_1_preds[:1]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "oARM1r3g9JYn",
        "outputId": "4c1e35c4-9f3b-45ca-f93b-c686ce8e56aa"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([29, 81, 91, 53, 97, 97, 10, 31,  3,  8])"
            ]
          },
          "metadata": {},
          "execution_count": 18
        }
      ],
      "source": [
        "# Convert model preds to labels\n",
        "labels = model_1_preds.argmax(axis = 1)\n",
        "labels[:10]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "diL_tNEn50FF"
      },
      "outputs": [],
      "source": [
        "# Load back in model weights\n",
        "loaded_model = model_1.load_weights('/content/model_checkpoint.weights.h5')\n",
        "loaded_model"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4XTxgZEHhM18"
      },
      "source": [
        "## Fine-tuning model 1\n",
        "\n",
        "> 🔑 **Note:** When fine-tuning a model, you typically lower the learning so that the model slowly updates it weights when training on a similar dataset (Food101 for example for this project)."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "HBfRKeVlhl4-",
        "outputId": "c2f10ad9-6f5a-4e5f-ae36-3187b4a70e79"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Saving TensorBoard log files to: ./tensorflow_logs/food_101_all_modelling_experiments/model_1/20250708-103729\n",
            "Epoch 11/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m251s\u001b[0m 95ms/step - accuracy: 0.5466 - loss: 1.7559 - val_accuracy: 0.6827 - val_loss: 1.1683\n",
            "Epoch 12/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m204s\u001b[0m 85ms/step - accuracy: 0.5681 - loss: 1.6711 - val_accuracy: 0.6854 - val_loss: 1.1479\n",
            "Epoch 13/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m204s\u001b[0m 85ms/step - accuracy: 0.5800 - loss: 1.6314 - val_accuracy: 0.7031 - val_loss: 1.1158\n",
            "Epoch 14/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m204s\u001b[0m 85ms/step - accuracy: 0.5863 - loss: 1.5857 - val_accuracy: 0.6954 - val_loss: 1.1099\n",
            "Epoch 15/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m205s\u001b[0m 85ms/step - accuracy: 0.5930 - loss: 1.5642 - val_accuracy: 0.6960 - val_loss: 1.1214\n",
            "Epoch 16/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m263s\u001b[0m 86ms/step - accuracy: 0.5954 - loss: 1.5500 - val_accuracy: 0.7063 - val_loss: 1.0975\n",
            "Epoch 17/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m205s\u001b[0m 86ms/step - accuracy: 0.5990 - loss: 1.5298 - val_accuracy: 0.7042 - val_loss: 1.0803\n",
            "Epoch 18/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m208s\u001b[0m 87ms/step - accuracy: 0.6056 - loss: 1.5017 - val_accuracy: 0.7108 - val_loss: 1.0673\n",
            "Epoch 19/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m208s\u001b[0m 87ms/step - accuracy: 0.6099 - loss: 1.4893 - val_accuracy: 0.7084 - val_loss: 1.0679\n",
            "Epoch 20/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m261s\u001b[0m 86ms/step - accuracy: 0.6154 - loss: 1.4640 - val_accuracy: 0.7203 - val_loss: 1.0419\n"
          ]
        }
      ],
      "source": [
        "# Unfreeze last 20 layers of feature-extractor model\n",
        "base_model.trainable = True\n",
        "for layer in base_model.layers[:-20]:\n",
        "  layer.trainable = False\n",
        "\n",
        "# Freeze all BatchNormalization layers within the last 20 layers\n",
        "for layer in base_model.layers[-20:]:\n",
        "  if isinstance(layer, layers.BatchNormalization):\n",
        "    layer.trainable = False\n",
        "\n",
        "# Compile\n",
        "model_1.compile(loss = tf.keras.losses.SparseCategoricalCrossentropy(from_logits = False),\n",
        "                optimizer = tf.keras.optimizers.Adam(learning_rate = 1e-5),\n",
        "                metrics = ['accuracy'])\n",
        "\n",
        "fine_tune_epochs = 10\n",
        "total_epochs = initial_epochs + fine_tune_epochs\n",
        "\n",
        "# Fit fine-tuned model\n",
        "history_1_fine_tune = model_1.fit(train_data, epochs = total_epochs, initial_epoch = initial_epochs, steps_per_epoch = len(train_data), validation_data = test_data, validation_steps = int(0.15 * len(test_data)), callbacks = [create_tensorboard_callback(dir_name = './tensorflow_logs', experiment_name = 'food_101_all_modelling_experiments/model_1_fine-tuned'), model_checkpoint])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 927
        },
        "id": "ohFGB63gsMgr",
        "outputId": "53d656f1-52c7-433a-a7c2-1e81abb66e42"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "# Plot loss curves for fine-tuned model\n",
        "plot_loss_curves(history = history_1_fine_tune)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "sqW3z5f-5FBt"
      },
      "source": [
        "These loss curves look very impressive, and from the model's `val_accuracy` metric, we have passed the accuracy score the authors of the Food101 paper obtained!\n",
        "\n",
        "However, we're going to be continuing some more modelling experiments to see the maximum accuracy score we can achieve, most likely from using pretrained ImageNet models from TensorFlow Hub: https://www.tensorflow.org/hub\n",
        "\n",
        "> 🔑 **Note:** **Fine-tuning** a model leverages the power of transfer learning, drastically improving accuracy metrics (both training and validation) and hence, giving the ideal position for the loss curves (i.e. accuracy up and loss down)."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4Kc1OfS-D7yi"
      },
      "source": [
        "## Model 2: Tiny VGG architecture\n",
        "\n",
        "Model 2's architecture will be represented as the one from the Tiny VGG model architecture: https://poloclub.github.io/cnn-explainer/"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 545
        },
        "id": "tW1QD5g-CTBD",
        "outputId": "937e86ed-c4ee-46ea-8a40-8699096963ac"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "\u001b[1mModel: \"sequential\"\u001b[0m\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\">Model: \"sequential\"</span>\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
              "┃\u001b[1m \u001b[0m\u001b[1mLayer (type)                   \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mOutput Shape          \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m      Param #\u001b[0m\u001b[1m \u001b[0m┃\n",
              "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
              "│ rescaling_2 (\u001b[38;5;33mRescaling\u001b[0m)         │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m3\u001b[0m)    │             \u001b[38;5;34m0\u001b[0m │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ conv2d (\u001b[38;5;33mConv2D\u001b[0m)                 │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m64\u001b[0m)   │         \u001b[38;5;34m1,792\u001b[0m │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ batch_normalization_1           │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m64\u001b[0m)   │           \u001b[38;5;34m256\u001b[0m │\n",
              "│ (\u001b[38;5;33mBatchNormalization\u001b[0m)            │                        │               │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ conv2d_1 (\u001b[38;5;33mConv2D\u001b[0m)               │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m64\u001b[0m)   │        \u001b[38;5;34m36,928\u001b[0m │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ max_pooling2d (\u001b[38;5;33mMaxPooling2D\u001b[0m)    │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m112\u001b[0m, \u001b[38;5;34m112\u001b[0m, \u001b[38;5;34m64\u001b[0m)   │             \u001b[38;5;34m0\u001b[0m │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ conv2d_2 (\u001b[38;5;33mConv2D\u001b[0m)               │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m112\u001b[0m, \u001b[38;5;34m112\u001b[0m, \u001b[38;5;34m128\u001b[0m)  │        \u001b[38;5;34m73,856\u001b[0m │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ batch_normalization_2           │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m112\u001b[0m, \u001b[38;5;34m112\u001b[0m, \u001b[38;5;34m128\u001b[0m)  │           \u001b[38;5;34m512\u001b[0m │\n",
              "│ (\u001b[38;5;33mBatchNormalization\u001b[0m)            │                        │               │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ conv2d_3 (\u001b[38;5;33mConv2D\u001b[0m)               │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m112\u001b[0m, \u001b[38;5;34m112\u001b[0m, \u001b[38;5;34m128\u001b[0m)  │       \u001b[38;5;34m147,584\u001b[0m │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ max_pooling2d_1 (\u001b[38;5;33mMaxPooling2D\u001b[0m)  │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m56\u001b[0m, \u001b[38;5;34m56\u001b[0m, \u001b[38;5;34m128\u001b[0m)    │             \u001b[38;5;34m0\u001b[0m │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ flatten (\u001b[38;5;33mFlatten\u001b[0m)               │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m401408\u001b[0m)         │             \u001b[38;5;34m0\u001b[0m │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ dropout_1 (\u001b[38;5;33mDropout\u001b[0m)             │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m401408\u001b[0m)         │             \u001b[38;5;34m0\u001b[0m │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ dense_1 (\u001b[38;5;33mDense\u001b[0m)                 │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m101\u001b[0m)            │    \u001b[38;5;34m40,542,309\u001b[0m │\n",
              "└─────────────────────────────────┴────────────────────────┴───────────────┘\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\">┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
              "┃<span style=\"font-weight: bold\"> Layer (type)                    </span>┃<span style=\"font-weight: bold\"> Output Shape           </span>┃<span style=\"font-weight: bold\">       Param # </span>┃\n",
              "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
              "│ rescaling_2 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Rescaling</span>)         │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">224</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">224</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">3</span>)    │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ conv2d (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Conv2D</span>)                 │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">224</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">224</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">64</span>)   │         <span style=\"color: #00af00; text-decoration-color: #00af00\">1,792</span> │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ batch_normalization_1           │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">224</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">224</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">64</span>)   │           <span style=\"color: #00af00; text-decoration-color: #00af00\">256</span> │\n",
              "│ (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">BatchNormalization</span>)            │                        │               │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ conv2d_1 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Conv2D</span>)               │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">224</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">224</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">64</span>)   │        <span style=\"color: #00af00; text-decoration-color: #00af00\">36,928</span> │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ max_pooling2d (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">MaxPooling2D</span>)    │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">112</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">112</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">64</span>)   │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ conv2d_2 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Conv2D</span>)               │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">112</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">112</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">128</span>)  │        <span style=\"color: #00af00; text-decoration-color: #00af00\">73,856</span> │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ batch_normalization_2           │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">112</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">112</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">128</span>)  │           <span style=\"color: #00af00; text-decoration-color: #00af00\">512</span> │\n",
              "│ (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">BatchNormalization</span>)            │                        │               │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ conv2d_3 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Conv2D</span>)               │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">112</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">112</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">128</span>)  │       <span style=\"color: #00af00; text-decoration-color: #00af00\">147,584</span> │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ max_pooling2d_1 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">MaxPooling2D</span>)  │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">56</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">56</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">128</span>)    │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ flatten (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Flatten</span>)               │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">401408</span>)         │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ dropout_1 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dropout</span>)             │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">401408</span>)         │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ dense_1 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)                 │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">101</span>)            │    <span style=\"color: #00af00; text-decoration-color: #00af00\">40,542,309</span> │\n",
              "└─────────────────────────────────┴────────────────────────┴───────────────┘\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "\u001b[1m Total params: \u001b[0m\u001b[38;5;34m40,803,237\u001b[0m (155.65 MB)\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Total params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">40,803,237</span> (155.65 MB)\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m40,802,853\u001b[0m (155.65 MB)\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">40,802,853</span> (155.65 MB)\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m384\u001b[0m (1.50 KB)\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Non-trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">384</span> (1.50 KB)\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        }
      ],
      "source": [
        "# Create model 2 (TinyVGG architecture) with rescaling layer\n",
        "model_2 = tf.keras.Sequential([\n",
        "    layers.Input(shape = input_size),\n",
        "    layers.Rescaling(1./255),\n",
        "    layers.Conv2D(64, 3, padding = 'same', activation = 'relu'),\n",
        "    layers.BatchNormalization(),\n",
        "    layers.Conv2D(64, 3, padding = 'same', activation = 'relu'),\n",
        "    layers.MaxPooling2D(),\n",
        "    layers.Conv2D(128, 3, padding = 'same', activation = 'relu'),\n",
        "    layers.BatchNormalization(),\n",
        "    layers.Conv2D(128, 3, padding = 'same', activation = 'relu'),\n",
        "    layers.MaxPooling2D(),\n",
        "    layers.Flatten(),\n",
        "    layers.Dropout(0.5),\n",
        "    layers.Dense(num_classes, activation = 'softmax')\n",
        "])\n",
        "\n",
        "# Compile\n",
        "model_2.compile(loss = tf.keras.losses.SparseCategoricalCrossentropy(from_logits = False),\n",
        "                optimizer = tf.keras.optimizers.Adam(),\n",
        "                metrics = ['accuracy'])\n",
        "\n",
        "# Get model summary\n",
        "model_2.summary()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ROhc4ERLWLNe",
        "outputId": "006e8cd4-4fd9-4cd6-e293-9ed8a32dbc91"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Saving TensorBoard log files to: ./tensorflow_logs/food_101_all_modelling_experiments/model_1/20250708-111425\n",
            "Epoch 1/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m645s\u001b[0m 145ms/step - accuracy: 0.0096 - loss: 15.7179 - val_accuracy: 0.0114 - val_loss: 15.9346\n",
            "Epoch 2/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m317s\u001b[0m 133ms/step - accuracy: 0.0094 - loss: 15.9661 - val_accuracy: 0.0114 - val_loss: 15.9346\n",
            "Epoch 3/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m318s\u001b[0m 133ms/step - accuracy: 0.0094 - loss: 15.9662 - val_accuracy: 0.0114 - val_loss: 15.9346\n",
            "Epoch 4/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m320s\u001b[0m 133ms/step - accuracy: 0.0094 - loss: 15.9672 - val_accuracy: 0.0114 - val_loss: 15.9346\n",
            "Epoch 5/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m321s\u001b[0m 133ms/step - accuracy: 0.0094 - loss: 15.9658 - val_accuracy: 0.0114 - val_loss: 15.9346\n",
            "Epoch 6/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m319s\u001b[0m 133ms/step - accuracy: 0.0094 - loss: 15.9662 - val_accuracy: 0.0114 - val_loss: 15.9346\n",
            "Epoch 7/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m324s\u001b[0m 134ms/step - accuracy: 0.0095 - loss: 15.9654 - val_accuracy: 0.0114 - val_loss: 15.9346\n",
            "Epoch 8/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m320s\u001b[0m 133ms/step - accuracy: 0.0095 - loss: 15.9654 - val_accuracy: 0.0114 - val_loss: 15.9346\n",
            "Epoch 9/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m317s\u001b[0m 133ms/step - accuracy: 0.0095 - loss: 15.9650 - val_accuracy: 0.0114 - val_loss: 15.9346\n",
            "Epoch 10/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m322s\u001b[0m 133ms/step - accuracy: 0.0094 - loss: 15.9660 - val_accuracy: 0.0114 - val_loss: 15.9346\n"
          ]
        }
      ],
      "source": [
        "# Fit\n",
        "history_2 = model_2.fit(train_data, epochs = initial_epochs, steps_per_epoch = len(train_data), validation_data = test_data, validation_steps = int(0.15 * len(test_data)), callbacks = [create_tensorboard_callback(dir_name = './tensorflow_logs', experiment_name = 'food_101_all_modelling_experiments/model_2'), model_checkpoint])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 927
        },
        "id": "JMZg0R3kXV22",
        "outputId": "9acbac0c-95b6-4409-ee56-eee8daab807d"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "# Plot loss curves\n",
        "plot_loss_curves(history = history_2)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "UDGHxY5Qatnj",
        "outputId": "3d09e9ba-e477-4ac9-ce3d-7794a9496bf4"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\u001b[1m790/790\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m72s\u001b[0m 91ms/step - accuracy: 0.0102 - loss: 15.9541\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[15.958532333374023, 0.009900989942252636]"
            ]
          },
          "metadata": {},
          "execution_count": 25
        }
      ],
      "source": [
        "# Evaluate model\n",
        "model_2.evaluate(test_data)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "21j5-DbHazQ_",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "a3b09183-9ee6-4b15-9cc8-e5aee4250dc1"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\u001b[1m790/790\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m70s\u001b[0m 87ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
              "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
              "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0.,\n",
              "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
              "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
              "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
              "        0., 0., 0., 0., 0.]], dtype=float16)"
            ]
          },
          "metadata": {},
          "execution_count": 26
        }
      ],
      "source": [
        "# Make predictions with model\n",
        "model_2_preds = model_2.predict(test_data)\n",
        "model_2_preds[:1]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "bNPW6ztva2cg",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "cd91e379-e16f-4708-c152-e8c92fc6c92f"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([43, 43, 43, 43, 43, 43, 43, 43, 43, 43])"
            ]
          },
          "metadata": {},
          "execution_count": 27
        }
      ],
      "source": [
        "# Convert model preds to label format\n",
        "labels = model_2_preds.argmax(axis = 1)\n",
        "labels[:10]"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Here, we can see that the modek us predicing only 1 class for each train food101 image we fed it.\n",
        "\n",
        "We see that the TinyVGG model architecture performs very poorly on the dataset fed into it, hinting that for larger datasets, it is simply not learning anything from the train dataset."
      ],
      "metadata": {
        "id": "wLm4X_Owsokg"
      }
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "sYua_7Ouc6ec"
      },
      "source": [
        "## Model 3: Pretrained Feature-extraction ResNet101 architecture"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 420
        },
        "id": "GG3tCcZXbGLh",
        "outputId": "cbd75a2a-be13-4d80-ff25-69506cdadeb3"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/resnet/resnet101_weights_tf_dim_ordering_tf_kernels_notop.h5\n",
            "\u001b[1m171446536/171446536\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 0us/step\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "\u001b[1mModel: \"functional_3\"\u001b[0m\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\">Model: \"functional_3\"</span>\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
              "┃\u001b[1m \u001b[0m\u001b[1mLayer (type)                   \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mOutput Shape          \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m      Param #\u001b[0m\u001b[1m \u001b[0m┃\n",
              "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
              "│ input_layer (\u001b[38;5;33mInputLayer\u001b[0m)        │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m3\u001b[0m)    │             \u001b[38;5;34m0\u001b[0m │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ resnet101 (\u001b[38;5;33mFunctional\u001b[0m)          │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m7\u001b[0m, \u001b[38;5;34m7\u001b[0m, \u001b[38;5;34m2048\u001b[0m)     │    \u001b[38;5;34m42,658,176\u001b[0m │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ global_average_pooling2d_1      │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m2048\u001b[0m)           │             \u001b[38;5;34m0\u001b[0m │\n",
              "│ (\u001b[38;5;33mGlobalAveragePooling2D\u001b[0m)        │                        │               │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ batch_normalization_3           │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m2048\u001b[0m)           │         \u001b[38;5;34m8,192\u001b[0m │\n",
              "│ (\u001b[38;5;33mBatchNormalization\u001b[0m)            │                        │               │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ dense_2 (\u001b[38;5;33mDense\u001b[0m)                 │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m256\u001b[0m)            │       \u001b[38;5;34m524,544\u001b[0m │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ dropout_2 (\u001b[38;5;33mDropout\u001b[0m)             │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m256\u001b[0m)            │             \u001b[38;5;34m0\u001b[0m │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ output_layer (\u001b[38;5;33mDense\u001b[0m)            │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m101\u001b[0m)            │        \u001b[38;5;34m25,957\u001b[0m │\n",
              "└─────────────────────────────────┴────────────────────────┴───────────────┘\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\">┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
              "┃<span style=\"font-weight: bold\"> Layer (type)                    </span>┃<span style=\"font-weight: bold\"> Output Shape           </span>┃<span style=\"font-weight: bold\">       Param # </span>┃\n",
              "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
              "│ input_layer (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">InputLayer</span>)        │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">224</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">224</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">3</span>)    │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ resnet101 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Functional</span>)          │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">7</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">7</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">2048</span>)     │    <span style=\"color: #00af00; text-decoration-color: #00af00\">42,658,176</span> │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ global_average_pooling2d_1      │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">2048</span>)           │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
              "│ (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">GlobalAveragePooling2D</span>)        │                        │               │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ batch_normalization_3           │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">2048</span>)           │         <span style=\"color: #00af00; text-decoration-color: #00af00\">8,192</span> │\n",
              "│ (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">BatchNormalization</span>)            │                        │               │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ dense_2 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)                 │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">256</span>)            │       <span style=\"color: #00af00; text-decoration-color: #00af00\">524,544</span> │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ dropout_2 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dropout</span>)             │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">256</span>)            │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ output_layer (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)            │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">101</span>)            │        <span style=\"color: #00af00; text-decoration-color: #00af00\">25,957</span> │\n",
              "└─────────────────────────────────┴────────────────────────┴───────────────┘\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "\u001b[1m Total params: \u001b[0m\u001b[38;5;34m43,216,869\u001b[0m (164.86 MB)\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Total params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">43,216,869</span> (164.86 MB)\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m554,597\u001b[0m (2.12 MB)\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">554,597</span> (2.12 MB)\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m42,662,272\u001b[0m (162.74 MB)\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Non-trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">42,662,272</span> (162.74 MB)\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        }
      ],
      "source": [
        "from tensorflow.keras import layers\n",
        "\n",
        "input_shape = (224, 224, 3)\n",
        "num_classes = ds_info.features['label'].num_classes\n",
        "\n",
        "# Create feature extractor model\n",
        "base_model = tf.keras.applications.ResNet101(include_top = False, input_shape = (224, 224, 3), weights = 'imagenet')\n",
        "base_model.trainable = False\n",
        "\n",
        "inputs = layers.Input(shape = input_size, name = 'input_layer')\n",
        "x = base_model(inputs, training = False)\n",
        "x = layers.GlobalAveragePooling2D()(x)\n",
        "x = layers.BatchNormalization()(x)\n",
        "x = layers.Dense(256, activation = 'relu')(x)\n",
        "x = layers.Dropout(0.5)(x)\n",
        "outputs = layers.Dense(num_classes, activation = 'softmax', name = 'output_layer')(x)\n",
        "model_3 = tf.keras.Model(inputs, outputs)\n",
        "\n",
        "# Compile the model\n",
        "model_3.compile(\n",
        "    loss = tf.keras.losses.SparseCategoricalCrossentropy(from_logits = False),\n",
        "    optimizer = tf.keras.optimizers.Adam(),\n",
        "    metrics = ['accuracy'])\n",
        "\n",
        "# Get model summary\n",
        "model_3.summary()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ZHY55k3__nqI",
        "outputId": "059f1aef-f836-4ba6-c138-065884947463"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Saving TensorBoard log files to: ./tensorflow_logs/food_101_all_modelling_experiments/model_3/20250708-121534\n",
            "Epoch 1/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m267s\u001b[0m 102ms/step - accuracy: 0.2000 - loss: 3.7883 - val_accuracy: 0.4711 - val_loss: 2.1220\n",
            "Epoch 2/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m239s\u001b[0m 96ms/step - accuracy: 0.3360 - loss: 2.7839 - val_accuracy: 0.4942 - val_loss: 2.0245\n",
            "Epoch 3/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m266s\u001b[0m 98ms/step - accuracy: 0.3612 - loss: 2.6618 - val_accuracy: 0.5037 - val_loss: 1.9330\n",
            "Epoch 4/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m238s\u001b[0m 99ms/step - accuracy: 0.3764 - loss: 2.5751 - val_accuracy: 0.5177 - val_loss: 1.9208\n",
            "Epoch 5/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m259s\u001b[0m 98ms/step - accuracy: 0.3930 - loss: 2.5064 - val_accuracy: 0.5355 - val_loss: 1.8492\n",
            "Epoch 6/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m263s\u001b[0m 98ms/step - accuracy: 0.4007 - loss: 2.4401 - val_accuracy: 0.5519 - val_loss: 1.8019\n",
            "Epoch 7/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m265s\u001b[0m 99ms/step - accuracy: 0.4127 - loss: 2.3881 - val_accuracy: 0.5493 - val_loss: 1.8019\n",
            "Epoch 8/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m240s\u001b[0m 100ms/step - accuracy: 0.4242 - loss: 2.3430 - val_accuracy: 0.5458 - val_loss: 1.7789\n",
            "Epoch 9/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m235s\u001b[0m 98ms/step - accuracy: 0.4237 - loss: 2.3205 - val_accuracy: 0.5442 - val_loss: 1.7689\n",
            "Epoch 10/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m260s\u001b[0m 98ms/step - accuracy: 0.4340 - loss: 2.2823 - val_accuracy: 0.5583 - val_loss: 1.7619\n"
          ]
        }
      ],
      "source": [
        "# Fit feature-extration model\n",
        "history_3 = model_3.fit(train_data, epochs = initial_epochs, steps_per_epoch = len(train_data), validation_data = test_data, validation_steps = int(0.15 * len(test_data)), callbacks = [create_tensorboard_callback(dir_name = './tensorflow_logs', experiment_name = 'food_101_all_modelling_experiments/model_3'), model_checkpoint])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 927
        },
        "id": "VTw_1Xr3tFOM",
        "outputId": "4f592128-ee6f-446e-f7ec-f436ed90f459"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "# Plot loss curves\n",
        "plot_loss_curves(history = history_3)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "f-21vitKuJUD",
        "outputId": "7f4f23a4-69ae-45b8-8582-ce2cbe673ea4"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\u001b[1m790/790\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m76s\u001b[0m 96ms/step - accuracy: 0.5555 - loss: 1.7587\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[1.76921808719635, 0.5519207715988159]"
            ]
          },
          "metadata": {},
          "execution_count": 31
        }
      ],
      "source": [
        "# Evaluate model on test data\n",
        "model_3.evaluate(test_data)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Ijooq3b6uLEq",
        "outputId": "ecf03720-4646-4bf5-d6a4-536431e4225e"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\u001b[1m790/790\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m85s\u001b[0m 99ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[2.7180e-03, 4.4703e-06, 1.2696e-05, 2.6226e-05, 2.4855e-05,\n",
              "        1.3819e-03, 1.4341e-04, 6.1989e-06, 1.1086e-02, 1.2150e-03,\n",
              "        5.6648e-03, 4.8280e-06, 1.5545e-03, 1.1490e-02, 7.7820e-04,\n",
              "        4.9686e-04, 1.8494e-02, 3.6931e-04, 1.4305e-06, 4.9353e-05,\n",
              "        1.6093e-06, 1.1778e-03, 3.8929e-03, 4.9353e-05, 9.9540e-06,\n",
              "        4.3571e-05, 5.0962e-05, 6.2406e-05, 4.9744e-03, 1.8481e-01,\n",
              "        4.2725e-03, 3.7503e-04, 7.9513e-05, 1.1921e-07, 1.4923e-02,\n",
              "        2.9206e-06, 3.3569e-03, 6.5446e-05, 3.2187e-06, 3.1090e-04,\n",
              "        9.0599e-06, 1.6093e-06, 4.8447e-03, 8.9407e-07, 7.7486e-07,\n",
              "        1.0254e-02, 8.4043e-06, 2.7895e-05, 3.1829e-04, 3.9041e-05,\n",
              "        2.8610e-06, 8.6427e-06, 3.3915e-05, 1.5478e-03, 0.0000e+00,\n",
              "        1.4748e-02, 4.0016e-03, 3.5477e-04, 1.6052e-01, 1.1086e-05,\n",
              "        1.7285e-06, 4.3213e-05, 5.7220e-06, 1.0133e-04, 0.0000e+00,\n",
              "        1.8477e-05, 1.9951e-03, 9.1493e-05, 8.2254e-06, 1.2660e-04,\n",
              "        0.0000e+00, 1.1921e-07, 1.5388e-02, 7.1297e-03, 3.3879e-04,\n",
              "        0.0000e+00, 1.3173e-05, 7.1287e-05, 5.7840e-04, 5.0068e-06,\n",
              "        1.0133e-04, 6.5565e-07, 2.0719e-04, 1.2413e-02, 6.5565e-07,\n",
              "        2.3842e-06, 8.0764e-05, 5.0831e-04, 5.9605e-08, 5.3644e-06,\n",
              "        0.0000e+00, 5.3644e-07, 2.3246e-06, 1.9014e-05, 3.0688e-01,\n",
              "        1.5998e-04, 6.6376e-04, 3.8385e-04, 1.6804e-03, 3.0828e-04,\n",
              "        1.8018e-01]], dtype=float16)"
            ]
          },
          "metadata": {},
          "execution_count": 32
        }
      ],
      "source": [
        "# Make predictions with model\n",
        "model_3_preds = model_3.predict(test_data)\n",
        "model_3_preds[:1]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "djvIbfBGuOH_",
        "outputId": "d8f1d055-1a7f-41bc-c4ca-c14fcca8ef3e"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([94, 81, 91, 53, 97,  6, 66, 31,  3,  6])"
            ]
          },
          "metadata": {},
          "execution_count": 33
        }
      ],
      "source": [
        "# Convert model preds to labels\n",
        "labels = model_3_preds.argmax(axis = 1)\n",
        "labels[:10]"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "H3PhS9I4lXuk"
      },
      "source": [
        "## Fine-tuning model 3"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "I6rogdyDuQb5",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "c7fd2e62-0b92-4811-b4c1-e463c33b63dd"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Saving TensorBoard log files to: ./tensorflow_logs/food_101_all_modelling_experiments/model_3/20250708-130037\n",
            "Epoch 11/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m280s\u001b[0m 107ms/step - accuracy: 0.4525 - loss: 2.1782 - val_accuracy: 0.5742 - val_loss: 1.6431\n",
            "Epoch 12/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m300s\u001b[0m 101ms/step - accuracy: 0.4748 - loss: 2.0875 - val_accuracy: 0.5903 - val_loss: 1.5975\n",
            "Epoch 13/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m261s\u001b[0m 101ms/step - accuracy: 0.4857 - loss: 2.0263 - val_accuracy: 0.5985 - val_loss: 1.5621\n",
            "Epoch 14/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m242s\u001b[0m 101ms/step - accuracy: 0.5041 - loss: 1.9555 - val_accuracy: 0.5964 - val_loss: 1.5348\n",
            "Epoch 15/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m262s\u001b[0m 101ms/step - accuracy: 0.5090 - loss: 1.9240 - val_accuracy: 0.6054 - val_loss: 1.5153\n",
            "Epoch 16/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m263s\u001b[0m 102ms/step - accuracy: 0.5193 - loss: 1.8690 - val_accuracy: 0.6112 - val_loss: 1.4953\n",
            "Epoch 17/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m261s\u001b[0m 101ms/step - accuracy: 0.5299 - loss: 1.8291 - val_accuracy: 0.6112 - val_loss: 1.5058\n",
            "Epoch 18/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m264s\u001b[0m 102ms/step - accuracy: 0.5378 - loss: 1.7861 - val_accuracy: 0.6184 - val_loss: 1.4582\n",
            "Epoch 19/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m243s\u001b[0m 101ms/step - accuracy: 0.5493 - loss: 1.7381 - val_accuracy: 0.6192 - val_loss: 1.4471\n",
            "Epoch 20/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m261s\u001b[0m 101ms/step - accuracy: 0.5548 - loss: 1.7089 - val_accuracy: 0.6292 - val_loss: 1.4314\n"
          ]
        }
      ],
      "source": [
        "# Unfreeze last 20 layers in base model (feature-extractor model)\n",
        "base_model.trainable = True\n",
        "for layer in base_model.layers[:-20]:\n",
        "  layer.trainable = False\n",
        "\n",
        "for layer in base_model.layers[-20:]:\n",
        "  if isinstance(layer, layers.BatchNormalization):\n",
        "    layer.trainable = False\n",
        "\n",
        "fine_tune_epochs = 10\n",
        "total_epochs = initial_epochs + fine_tune_epochs\n",
        "\n",
        "# Recompile\n",
        "model_3.compile(loss = tf.keras.losses.SparseCategoricalCrossentropy(from_logits = False),\n",
        "                optimizer = tf.keras.optimizers.Adam(learning_rate = 1e-5),\n",
        "                metrics = ['accuracy'])\n",
        "\n",
        "# Fit fine-tuned model\n",
        "history_3_fine_tune = model_3.fit(train_data, epochs = total_epochs, initial_epoch = initial_epochs, steps_per_epoch = len(train_data), validation_data = test_data, validation_steps = int(0.15 * len(test_data)), callbacks = [create_tensorboard_callback(dir_name = './tensorflow_logs', experiment_name = 'food_101_all_modelling_experiments/model_3_fine-tuned'), model_checkpoint])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "zax8hisBrwee",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 927
        },
        "outputId": "577f80c9-1d4d-473e-bead-ab9042b806c9"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAiwAAAHHCAYAAACcHAM1AAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjAsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvlHJYcgAAAAlwSFlzAAAPYQAAD2EBqD+naQAAaQVJREFUeJzt3Xd4lFXexvHvpBdIKCE9JEDonQRCQlEURVEEbKAIAQVXBRTzqsCqKLoa27K4oqKuyKog2FBUFktElI406T0JLSEB0iFl5nn/GBgcCJAgMJPk/lzXXDDnKfkNQXJ7znnOMRmGYSAiIiLixFwcXYCIiIjIhSiwiIiIiNNTYBERERGnp8AiIiIiTk+BRURERJyeAouIiIg4PQUWERERcXoKLCIiIuL0FFhERETE6SmwiIiIiNNTYBGR83rrrbcwmUzExcU5uhQRqcFM2ktIRM6nW7duHDx4kNTUVHbu3El0dLSjSxKRGkg9LCJyTnv37mXZsmVMmTKFBg0aMGvWLEeXVK7CwkJHlyAil5kCi4ic06xZs6hbty433XQTt99+e7mBJScnh0cffZSoqCg8PT0JDw9n2LBhZGdn2845ceIEzz77LM2aNcPLy4uQkBBuvfVWdu/eDcAvv/yCyWTil19+sbt3amoqJpOJmTNn2tqGDx9OrVq12L17N3379qV27doMGTIEgN9++4077riDhg0b4unpSUREBI8++ijHjx8/q+5t27Zx55130qBBA7y9vWnevDlPPvkkAIsWLcJkMjFv3ryzrps9ezYmk4nly5dX+s9TRC6em6MLEBHnNWvWLG699VY8PDy46667ePvtt1m9ejWdO3cGoKCggB49erB161buvfdeOnXqRHZ2NvPnz2f//v0EBARgNpu5+eabSUlJYfDgwTzyyCPk5+fz448/smnTJpo0aVLpusrKyujTpw/du3fntddew8fHB4DPPvuMoqIiHnzwQerXr8+qVat444032L9/P5999pnt+j/++IMePXrg7u7O/fffT1RUFLt37+abb77hhRde4OqrryYiIoJZs2YxcODAs/5MmjRpQnx8/F/4kxWRSjNERMrx+++/G4Dx448/GoZhGBaLxQgPDzceeeQR2zmTJk0yAOPLL78863qLxWIYhmHMmDHDAIwpU6ac85xFixYZgLFo0SK743v37jUA44MPPrC1JSYmGoAxYcKEs+5XVFR0VltycrJhMpmMtLQ0W1vPnj2N2rVr27X9uR7DMIyJEycanp6eRk5Ojq3t8OHDhpubm/HMM8+c9XVE5PLSkJCIlGvWrFkEBQXRq1cvAEwmE4MGDWLOnDmYzWYAvvjiC9q3b39WL8Sp80+dExAQwNixY895zsV48MEHz2rz9va2/b6wsJDs7GwSEhIwDIN169YBkJWVxa+//sq9995Lw4YNz1nPsGHDKC4u5vPPP7e1zZ07l7KyMu65556LrltELo4Ci4icxWw2M2fOHHr16sXevXvZtWsXu3btIi4ujszMTFJSUgDYvXs3bdq0Oe+9du/eTfPmzXFzu3Qj0G5uboSHh5/Vnp6ezvDhw6lXrx61atWiQYMGXHXVVQDk5uYCsGfPHoAL1t2iRQs6d+5sN29n1qxZdO3aVU9KiTiA5rCIyFl+/vlnDh06xJw5c5gzZ85Zx2fNmsX1119/yb7euXpaTvXknMnT0xMXF5ezzr3uuus4evQo48ePp0WLFvj6+nLgwAGGDx+OxWKpdF3Dhg3jkUceYf/+/RQXF7NixQqmTZtW6fuIyF+nwCIiZ5k1axaBgYG8+eabZx378ssvmTdvHtOnT6dJkyZs2rTpvPdq0qQJK1eupLS0FHd393LPqVu3LmB94ujP0tLSKlzzxo0b2bFjB//9738ZNmyYrf3HH3+0O69x48YAF6wbYPDgwSQlJfHJJ59w/Phx3N3dGTRoUIVrEpFLR0NCImLn+PHjfPnll9x8883cfvvtZ73GjBlDfn4+8+fP57bbbmPDhg3lPv5rnFyT8rbbbiM7O7vcnolT50RGRuLq6sqvv/5qd/ytt96qcN2urq529zz1+9dff93uvAYNGtCzZ09mzJhBenp6ufWcEhAQwI033sjHH3/MrFmzuOGGGwgICKhwTSJy6aiHRUTszJ8/n/z8fG655ZZyj3ft2tW2iNzs2bP5/PPPueOOO7j33nuJiYnh6NGjzJ8/n+nTp9O+fXuGDRvGhx9+SFJSEqtWraJHjx4UFhby008/8dBDD9G/f3/8/f254447eOONNzCZTDRp0oRvv/2Ww4cPV7juFi1a0KRJEx577DEOHDiAn58fX3zxBceOHTvr3H//+990796dTp06cf/999OoUSNSU1P57rvvWL9+vd25w4YN4/bbbwfg+eefr/gfpIhcWo58RElEnE+/fv0MLy8vo7Cw8JznDB8+3HB3dzeys7ONI0eOGGPGjDHCwsIMDw8PIzw83EhMTDSys7Nt5xcVFRlPPvmk0ahRI8Pd3d0IDg42br/9dmP37t22c7KysozbbrvN8PHxMerWrWv87W9/MzZt2lTuY82+vr7l1rVlyxajd+/eRq1atYyAgABj1KhRxoYNG866h2EYxqZNm4yBAwcaderUMby8vIzmzZsbTz/99Fn3LC4uNurWrWv4+/sbx48fr+CfoohcatpLSETkPMrKyggNDaVfv368//77ji5HpMbSHBYRkfP46quvyMrKspvIKyJXnnpYRETKsXLlSv744w+ef/55AgICWLt2raNLEqnR1MMiIlKOt99+mwcffJDAwEA+/PBDR5cjUuOph0VEREScnnpYRERExOkpsIiIiIjTqxYLx1ksFg4ePEjt2rX/0u6vIiIicuUYhkF+fj6hoaFn7Q92pmoRWA4ePEhERISjyxAREZGLsG/fvnJ3YP+zahFYateuDVg/sJ+fn4OrERERkYrIy8sjIiLC9nP8fKpFYDk1DOTn56fAIiIiUsVUZDqHJt2KiIiI01NgEREREaenwCIiIiJOr1rMYakos9lMaWmpo8sQuSAPD48LPuInIlKT1IjAYhgGGRkZ5OTkOLoUkQpxcXGhUaNGeHh4OLoUERGnUCMCy6mwEhgYiI+PjxaXE6d2aiHEQ4cO0bBhQ/19FRGhBgQWs9lsCyv169d3dDkiFdKgQQMOHjxIWVkZ7u7uji5HRMThqv0g+ak5Kz4+Pg6uRKTiTg0Fmc1mB1ciIuIcqn1gOUXd6lKV6O+riIi9GhNYREREpOpSYKkhoqKimDp1aoXP/+WXXzCZTHqySkREnEK1n3RblV199dV06NChUkHjXFavXo2vr2+Fz09ISODQoUP4+/v/5a8tIiLyV6mHpQozDIOysrIKndugQYNKTTz28PAgODi4xs+l0EKDIiJA0VE4sMahJSiwOKnhw4ezePFiXn/9dUwmEyaTiZkzZ2Iymfjf//5HTEwMnp6eLFmyhN27d9O/f3+CgoKoVasWnTt35qeffrK735lDQiaTif/85z8MHDgQHx8fmjZtyvz5823HzxwSmjlzJnXq1OH777+nZcuW1KpVixtuuIFDhw7ZrikrK+Phhx+mTp061K9fn/Hjx5OYmMiAAQMq9JkXLlxI9+7dbdfffPPN7N692+6c/fv3c9ddd1GvXj18fX2JjY1l5cqVtuPffPMNnTt3xsvLi4CAAAYOHGj3mb/66iu7+9WpU4eZM2cCkJqaislkYu7cuVx11VV4eXkxa9Ysjhw5wl133UVYWBg+Pj60bduWTz75xO4+FouFV155hejoaDw9PWnYsCEvvPACANdccw1jxoyxOz8rKwsPDw9SUlIq9GcjIuIQB9fB16NhSkv4bDhYHPfkYo0MLIZhUFRS5pCXYRgVqvH1118nPj6eUaNGcejQIQ4dOkRERAQAEyZM4KWXXmLr1q20a9eOgoIC+vbtS0pKCuvWreOGG26gX79+pKenn/drTJ48mTvvvJM//viDvn37MmTIEI4ePXrO84uKinjttdf46KOP+PXXX0lPT+exxx6zHX/55ZeZNWsWH3zwAUuXLiUvL++sgHA+hYWFJCUl8fvvv5OSkoKLiwsDBw7EYrEAUFBQwFVXXcWBAweYP38+GzZs4IknnrAd/+677xg4cCB9+/Zl3bp1pKSk0KVLlwp//VMmTJjAI488wtatW+nTpw8nTpwgJiaG7777jk2bNnH//fczdOhQVq1aZbtm4sSJvPTSSzz99NNs2bKF2bNnExQUBMDIkSOZPXs2xcXFtvM//vhjwsLCuOaaaypdn4jIZVV6AtbPhveugXevhnUfQ9kJ8PSH/AyHlVUj57AcLzXTatL3DvnaW57rg4/Hhf/Y/f398fDwwMfHh+DgYAC2bdsGwHPPPcd1111nO7devXq0b9/e9v75559n3rx5zJ8//6z/s/+z4cOHc9dddwHw4osv8u9//5tVq1Zxww03lHt+aWkp06dPp0mTJgCMGTOG5557znb8jTfeYOLEibZejWnTprFgwYILftZTbrvtNrv3M2bMoEGDBmzZsoU2bdowe/ZssrKyWL16NfXq1QMgOjradv4LL7zA4MGDmTx5sq3tz38uFTVu3DhuvfVWu7Y/B7OxY8fy/fff8+mnn9KlSxfy8/N5/fXXmTZtGomJiQA0adKE7t27A3DrrbcyZswYvv76a+68807A2mM1fPjwGj/kJiJO5Ohe+H2GNaAcP/k/r64e0GoAdB4JEV3Agf9m1cjAUtXFxsbavS8oKODZZ5/lu+++49ChQ5SVlXH8+PEL9rC0a9fO9ntfX1/8/Pw4fPjwOc/38fGxhRWAkJAQ2/m5ublkZmba9Wi4uroSExNj6wG5kJ07dzJp0iRWrlxJdna27br09HTatGnD+vXr6dixoy2snGn9+vWMGjWqQl/rfM788zWbzbz44ot8+umnHDhwgJKSEoqLi21zgrZu3UpxcTHXXnttuffz8vJi6NChzJgxgzvvvJO1a9eyadMmuyE4ERGHsJhh10+w+j+w80fg5CiAfwTEjoCOw6BWA4eWeEqNDCze7q5sea6Pw772X3Xm0z6PPfYYP/74I6+99hrR0dF4e3tz++23U1JSct77nLnku8lkOm+4KO/8ig5xVUS/fv2IjIzkvffeIzQ0FIvFQps2bWyfw9vb+7zXX+h4efWWN6n2zD/fV199lddff52pU6fStm1bfH19GTduXIXrAuuwUIcOHdi/fz8ffPAB11xzDZGRkRe8TkTksig8Aus+svao5KSdbm9yLXQZBU2vB5e//vPqUqqRgcVkMlVoWMbRPDw8KrQ0+9KlSxk+fLhtKKagoIDU1NTLXJ09f39/goKCWL16NT179gSsPRNr166lQ4cOF7z+yJEjbN++nffee48ePXoAsGTJErtz2rVrx3/+8x+OHj1abi9Lu3btSElJYcSIEeV+jQYNGthNEt65cydFRUUXrG3p0qX079+fe+65B7BOsN2xYwetWrUCoGnTpnh7e5OSksLIkSPLvUfbtm2JjY3lvffeY/bs2UybNu2CX1dE5JIyDOuTPqveg83zwHxyXp1XHeh4D8TeC/WbnPcWjuT8P7VrsKioKFauXElqaiq1atU6Z+9H06ZN+fLLL+nXrx8mk4mnn366wsMwl9LYsWNJTk4mOjqaFi1a8MYbb3Ds2LEKzdOoW7cu9evX59133yUkJIT09HQmTJhgd85dd93Fiy++yIABA0hOTiYkJIR169YRGhpKfHw8zzzzDNdeey1NmjRh8ODBlJWVsWDBAsaPHw9Yn9aZNm0a8fHxmM1mxo8fX6GNBZs2bcrnn3/OsmXLqFu3LlOmTCEzM9MWWLy8vBg/fjxPPPEEHh4edOvWjaysLDZv3sx9991nu8/IkSMZM2YMvr6+dk8viYhcViVFsOlz67DPoQ2n20M6WHtTWt8KHs6/316NfEqoqnjsscdwdXWlVatWNGjQ4JxzUqZMmULdunVJSEigX79+9OnTh06dOl3hamH8+PHcddddDBs2jPj4eGrVqkWfPn3w8vK64LUuLi7MmTOHNWvW0KZNGx599FFeffVVu3M8PDz44YcfCAwMpG/fvrRt25aXXnoJV1drt+XVV1/NZ599xvz58+nQoQPXXHON3ZM8//znP4mIiKBHjx7cfffdPPbYYxVam+app56iU6dO9OnTh6uvvprg4OCzHtV++umn+b//+z8mTZpEy5YtGTRo0Fnzge666y7c3Ny46667KvRnIiLyl2TvgoV/hyktYP5Ya1hx9YT2d8PIn+Fvi609K1UgrACYjEs5CcFB8vLy8Pf3Jzc3Fz8/P7tjJ06cYO/evTRq1Eg/JK4wi8VCy5YtufPOO3n++ecdXY7Dpaam0qRJE1avXn3BQKm/tyJyUcxlsPN767DPnkWn2+tEQuf7oMM94FvfcfWd4Xw/v8+kISG5ZNLS0vjhhx+46qqrKC4uZtq0aezdu5e7777b0aU5VGlpKUeOHOGpp56ia9euDun9EpFqruAwrP0v/D4T8vafbDRZJ892GWWdTOtStQdVFFjkknFxcWHmzJk89thjGIZBmzZt+Omnn2jZsiXp6em2OR/l2bJlCw0bNryC1V45S5cupVevXjRr1ozPP//c0eWISHVhGJC+wjo3ZcvXYDn51KN3Peg0zPpYct0oh5Z4KSmwyCUTERHB0qVLyz0WGhrK+vXrz3ltaGjoZarK8a6++upL+vi3iNRwxQWw8VNY/T5kbjrdHhZr7U1pNQDcq99QsgKLXBFubm52q9KKiEglZW23hpQNn0BxnrXNzRva3m6dnxLa0bH1XWYKLCIiIs7KXArbF1gn0ab+drq9XmPrcvkd7gbvuo6r7wpSYBEREXE2eYesk2jXzIT8kwtemlyg2Y3W3pTGvar8JNrKUmARERFxBoYBqUusk2i3fQuWMmu7TwDEJELMCKgT4dgaHeii4tmbb75JVFQUXl5exMXF2S3OVZ6cnBxGjx5NSEgInp6eNGvWzG4X3+TkZDp37kzt2rUJDAxkwIABbN++/WJKExERqVpO5FmHfN7qCv+9GbZ8ZQ0rDePhtvchaQtcO6lGhxW4iB6WuXPnkpSUxPTp04mLi2Pq1Kn06dOH7du3ExgYeNb5JSUlXHfddQQGBvL5558TFhZGWloaderUsZ2zePFiRo8eTefOnSkrK+Pvf/87119/PVu2bDlrIzoREZFqIXOztTdlw1woLbS2uftCuzutwz7BbR1bn5Op9Eq3cXFxdO7c2bZ5m8ViISIigrFjx5619wvA9OnTefXVV9m2bVuF9m0ByMrKIjAwkMWLF9s20jsfrXRbvqioKMaNG8e4ceMcXYpUUk3+eytSrZWVwNb51qd90pedbg9oZp1E234wePk7rr4r7LKtdFtSUsKaNWuYOHGirc3FxYXevXuzfPnycq+ZP38+8fHxjB49mq+//poGDRpw9913M378eNseMGfKzc0FKHdHXoDi4mKKi4tt7/Py8irzMURERK6s3P3WCbRr/guFJ/cZM7lCi5usQaVRT6jARrE1WaUCS3Z2NmazmaCgILv2oKAgtm3bVu41e/bs4eeff2bIkCEsWLCAXbt28dBDD1FaWsozzzxz1vkWi4Vx48bRrVs32rRpU+49k5OTmTx5cmVKlyrGbDZjMplwqWGz4EWkCrGYoewElBVbfy09fvL3x0+3Hc+BTV9YH002LNbragVBzHDry6/6Lpp5qV32p4QsFguBgYG8++67uLq6EhMTw4EDB3j11VfLDSyjR49m06ZNLFmy5Jz3nDhxIklJSbb3eXl5RERUr8lI7777Ls8++yz79++3+6Hdv39/6tevz5NPPklSUhIrVqygsLCQli1bkpycTO/evS/q602ZMoUPPviAPXv2UK9ePfr168crr7xCrVq1bOcsXbqUJ598klWrVuHp6UmXLl2YM2cOdevWxWKx8Nprr/Huu++yb98+goKC+Nvf/saTTz7JL7/8Qq9evTh27Jht7tL69evp2LEje/fuJSoqipkzZzJu3Dg+/PBDJkyYwI4dO9i1axdZWVn8/e9/Z926dZSWltKhQwf+9a9/2e3Hk5OTw/jx4/nqq6/Izc0lOjqal156iV69ehESEsKMGTO4/fbbbed/9dVXDBkyhIyMDGrXrn1Rf14i4iQMw7pWyamQUPqnsHDqVXriT8GiIuf96fzznXdqKfyKiuwOXUZCi5vBtWJTJOS0SgWWgIAAXF1dyczMtGvPzMwkODi43GtCQkJwd3e3G/5p2bIlGRkZlJSU4OHhYWsfM2YM3377Lb/++ivh4eHnrMPT0xNPT8/KlG7PMKC06OKv/yvcfSrU7XfHHXcwduxYFi1axLXXXgvA0aNHWbhwIQsWLKCgoIC+ffvywgsv4OnpyYcffki/fv3Yvn37Re3J4+Liwr///W8aNWrEnj17eOihh3jiiSd46623AGvAuPbaa7n33nt5/fXXcXNzY9GiRZjNZsAaIt977z3+9a9/0b17dw4dOnTOXrdzKSoq4uWXX+Y///kP9evXJzAwkD179pCYmMgbb7yBYRj885//pG/fvuzcuZPatWtjsVi48cYbyc/P5+OPP6ZJkyZs2bIFV1dXfH19GTx4MB988IFdYDn1XmFFxAkVHYVdP8GeX6y/r0jAwAm2vnBxAzev0y/3U7/3hNBO1km0gS0dXWWVVqnA4uHhQUxMDCkpKQwYMACw9qCkpKQwZsyYcq/p1q0bs2fPxmKx2HoKduzYQUhIiC2sGIbB2LFjmTdvHr/88guNGjX6Cx+pAkqL4EUHdcP9/SB4XPjJp7p163LjjTcye/ZsW2D5/PPPCQgIoFevXri4uNC+fXvb+c8//zzz5s1j/vz55/xenM+fJ+ZGRUXxj3/8gwceeMAWWF555RViY2Nt7wFat24NQH5+Pq+//jrTpk0jMTERgCZNmtC9e/dK1VBaWspbb71l97muueYau3Peffdd6tSpw+LFi7n55pv56aefWLVqFVu3bqVZs2YANG7c2Hb+yJEjSUhI4NChQ4SEhHD48GEWLFjATz/9VKnaROQyMQw4vBV2LIQd38P+VaeHTi7GuUKDm7f1V/eTv/75PLv2Px139zrjvHPc080LXLWs2eVW6T/hpKQkEhMTiY2NpUuXLkydOpXCwkJGjBgBwLBhwwgLCyM5ORmABx98kGnTpvHII48wduxYdu7cyYsvvsjDDz9su+fo0aOZPXs2X3/9NbVr1yYjIwMAf39/vL29L8XnrJKGDBnCqFGjeOutt/D09GTWrFkMHjwYFxcXCgoKePbZZ/nuu+84dOgQZWVlHD9+nPT09Iv6Wj/99BPJycls27aNvLw8ysrKOHHiBEVFRfj4+LB+/XruuOOOcq/dunUrxcXFtmB1sTw8PGjXrp1dW2ZmJk899RS//PILhw8fxmw2U1RUZPuc69evJzw83BZWztSlSxdat27Nf//7XyZMmMDHH39MZGRkhZ4+E5HLpPQ47P0Ndn5vDSm5++yPB7aGZtdD3UYVDBInA4erhyauVmOVDiyDBg0iKyuLSZMmkZGRQYcOHVi4cKFtIm56errdnIuIiAi+//57Hn30Udq1a0dYWBiPPPII48ePt53z9ttvA9Zdbf/sgw8+YPjw4RfxsS7A3cfa0+EI7j4VPrVfv34YhsF3331H586d+e233/jXv/4FwGOPPcaPP/7Ia6+9RnR0NN7e3tx+++2UlJRUuqTU1FRuvvlmHnzwQV544QXq1avHkiVLuO+++ygpKcHHx+e8wfFCofLU34c/P0FfWnr22K+3tzemM/6xSUxM5MiRI7z++utERkbi6elJfHy87XNWJNCOHDmSN998kwkTJvDBBx8wYsSIs76OiFxmuQdOB5Q9i61DPae4eUGjq6whpWmfGr9AmpTvovqwxowZc85hh19++eWstvj4eFasWHHO+1VyKZi/zmSq0LCMo3l5eXHrrbcya9Ysdu3aRfPmzW2TTZcuXcrw4cMZOHAgAAUFBaSmpl7U11mzZg0Wi4V//vOftnDx6aef2p3Trl07UlJSyn06q2nTpnh7e5OSksLIkSPPOt6gQQMADh06RN261k261q9fX6Hali5dyltvvUXfvn0B2LdvH9nZ2XZ17d+/nx07dpyzl+Wee+7hiSee4N///jdbtmyxDVuJyGVkMcOBNSeHen6AzI32x/3CoFkfa0Bp1BM8Kv4/c1IzadDNyQ0ZMoSbb76ZzZs3c88999jamzZtypdffkm/fv0wmUw8/fTTWCwXN+4bHR1NaWkpb7zxBv369WPp0qVMnz7d7pyJEyfStm1bHnroIR544AE8PDxYtGgRd9xxBwEBAYwfP54nnngCDw8PunXrRlZWFps3b+a+++4jOjqaiIgInn32WV544QV27NjBP//5zwrV1rRpUz766CNiY2PJy8vj8ccft+tVueqqq+jZsye33XYbU6ZMITo6mm3btmEymbjhhhsA63ygW2+9lccff5zrr7/+vBO6ReQvOJ4Du3+29qLs+hGKjvzpoAkiukDT66HZDRDUWsM3Uila5MLJXXPNNdSrV4/t27dz991329qnTJlC3bp1SUhIoF+/fvTp08fuUd/KaN++PVOmTOHll1+mTZs2zJo1yzYH6ZRmzZrxww8/sGHDBrp06UJ8fDxff/01bm7WzPv000/zf//3f0yaNImWLVsyaNAgDh+2Lo7k7u7OJ598wrZt22jXrh0vv/wy//jHPypU2/vvv8+xY8fo1KkTQ4cO5eGHHz5rC4gvvviCzp07c9ddd9GqVSueeOIJ29NLp5wa3rr33nsv6s9IRMphGJC1A5b+G2beDK80hs9HwB9zrGHF0x9a3woD34XHd8N9P0DPxyC4jcKKVFqll+Z3RlqaXy7ko48+4tFHH+XgwYN2j9I7K/29FadVVmzdUXjnD9bhnmOp9scDmlvnojS7ASLitN6InNdlW5pfpKopKiri0KFDvPTSS/ztb3+rEmFFxOnkZ5wMKN/D7kWnN+oD65M5Ud2tAaXp9VDvMi9LITWWAksNMGvWLP72t7+VeywyMpLNmzdf4YqunFdeeYUXXniBnj172u2BJSLnYbHAoXXWybI7FsKh9fbHawWdnovS+GrwrFXeXUQuKQ0J1QD5+flnrU58iru7O5GRkVe4IrkQ/b2VK64439p7suN7a2/KqQ36TgntZA0oza6H4Pagfb7kEtCQkNipXbu2lqEXkbMd2X16LkrqUvu9cTxqQZNrrI8eR18HtYPOfR+RK0CBRUSkpjCXQvpyay/Kju/hyE774/Uan56LEtkN3DTnS5xHjQksF7tGiYgjVIORWnEWBVnWNVF2fG9dI6U47/QxFzeITLAu3tbsBgiIdlydIhdQ7QOLh4cHLi4uHDx4kAYNGuDh4aFl2cWpGYZBVlYWJpMJd3c9EiqVZBiQ8cfpCbMH1mC3m7FPwMkJs32gSS/w8ndYqSKVUe0Di4uLC40aNeLQoUMcPOig/YNEKslkMhEeHo6rq6ujS6m+LGbI3GwdIklbCoc2gLnM0VX9daVFcPyofVtwO2tAaXaDdfKsJsxKFVTtAwtYe1kaNmxIWVnZWSugijgjd3d3hZVLrawEDq6D9GWQtgzSV0JxrqOrujzcfayPGze93vryD3N0RSJ/WY0ILICte11d7CI1REkh7F9tDSdpy2D/7/Y7BAN41LbubxOZYP3VoxqsJ2JygQYtwF2Pw0v1UmMCi4hUc0VHYd9K6/BO2nLrYmeWM4Z4fOpDw3jrEzCR8RDUFlz1z6BIVaD/UkWkaso7dHJ4Z7m1B+VwOSs2+4Vbe08iT4aUgGbadE+kilJgERHnZxhwbO/J4Z2Tk2SP7T37vPpNTwaUk686Da98rSJyWSiwiIjzsVgga+vp+Sdpy6Ag44yTTBDc9nQ4aRgPtQIdUq6IXH4KLCLieOZS62PFp8JJ+nI4kWN/jos7hMWcHt6J6KI1RERqEAUWEbnySo9bn9pJW2Yd3tm/2rp+yJ+5+0JE55MTZBOsYcXd2zH1iojDKbCIyOV3Ite67knaUmvvyYG19hvtAXjXPfkETwI0TICQduCqZQhExEqBRUQuvYLDp4d20pZCxibslocHqB1yeu5JZDfr2iFagVVEzkGBRUT+GsOAnPSTAeXkHJQju84+r17j070nkQlQN0qPGItIhSmwiEjlGYZ1kuyGT2Drt5C3/4wTTBDU+vQQT2QC1A52SKkiUj0osIhIxeVnwsZPYf1sOLzldLuLG4R2PD280zDOOidFROQSUWARkfMrPQHbF1h7U3algHFyA1FXD2jeF9rfBY16gIevY+sUkWpNgUVEzmYY1keN18+GzV9an/I5JbyzNaS0uVW9KCJyxSiwiMhpufthwxxrb8qfJ876hUH7wdagEtDUcfWJSI2lwCJS05UUwtZvrL0pe3/F9vixuw+07HdyyKcnuLg6tEwRqdkUWERqIovF+gjy+k9gy1dQUnD6WGR36HAXtOoPnrUdVqKIyJ8psIjUJEf3nB7yyUk/3V43CtrfDe0HWX8vIuJkLmpZyTfffJOoqCi8vLyIi4tj1apV5z0/JyeH0aNHExISgqenJ82aNWPBggV/6Z4iUkEncmHNf2HGDfDvjrD4ZWtY8agNHYfCiP/Bw+vh6vEKKyLitCrdwzJ37lySkpKYPn06cXFxTJ06lT59+rB9+3YCA8/e2r2kpITrrruOwMBAPv/8c8LCwkhLS6NOnToXfU8RuQCLGfb8cnJht2+g7MTJAyZo0svam9LiJvDwcWSVIiIVZjIMw7jwaafFxcXRuXNnpk2bBoDFYiEiIoKxY8cyYcKEs86fPn06r776Ktu2bcPdvfyNzCp7zzPl5eXh7+9Pbm4ufn5+lfk4ItVL1nbr5Nk/PoX8g6fbA5pb56W0GwR+oY6rT0TkTyrz87tSPSwlJSWsWbOGiRMn2tpcXFzo3bs3y5cvL/ea+fPnEx8fz+jRo/n6669p0KABd999N+PHj8fV1fWi7llcXExxcbHdBxapsYqOwqYvrEHl4NrT7V51oO3t0OFuCO2kfXtEpEqrVGDJzs7GbDYTFBRk1x4UFMS2bdvKvWbPnj38/PPPDBkyhAULFrBr1y4eeughSktLeeaZZy7qnsnJyUyePLkypYtUL+ZS2PWTNaTsWAjmEmu7yRWaXm/tTWl2A7h5OrZOEZFL5LI/JWSxWAgMDOTdd9/F1dWVmJgYDhw4wKuvvsozzzxzUfecOHEiSUlJtvd5eXlERERcqpJFnFfGRmtI2fgZFGadbg9qaw0pbe+AWpr3JSLVT6UCS0BAAK6urmRmZtq1Z2ZmEhxc/k6sISEhuLu74+p6etGpli1bkpGRQUlJyUXd09PTE09P/Z9jjVOQBVu/hqwd4FMffAOsP5x9G5x+edaufkMfBYetAWX9J5C58XS7bwNoe6c1qAS3dVx9IiJXQKUCi4eHBzExMaSkpDBgwADA2oOSkpLCmDFjyr2mW7duzJ49G4vFgouL9SnqHTt2EBISgoeHB0Cl7yk1SOER2PYNbPoSUn8Dw3L+8928ToaXAPANPP37M4ONbwNr6HF10qWIyoph+/+sT/ns/PGMDQdvtD7lE30tuJY/kV1EpLqp9L/WSUlJJCYmEhsbS5cuXZg6dSqFhYWMGDECgGHDhhEWFkZycjIADz74INOmTeORRx5h7Nix7Ny5kxdffJGHH364wveUGub4Mdj2nTWk7Pnl9A9rsE4ejeoGJ/KsQyKFWdYeiMJsKC20Pr6bu8/6uiAT+NQ7GWwCrCGmVuAZYacB1Dr56+Xejdgw4MBaWD/LOon2RM7pY2Ex1smzrW+11iwiUsNUOrAMGjSIrKwsJk2aREZGBh06dGDhwoW2SbPp6em2nhSAiIgIvv/+ex599FHatWtHWFgYjzzyCOPHj6/wPaUGOJEH2xdYQ8run8FSevpYcDtoPdD6qtfo3PcoKTwZYrL/FGSy7F8FJ38tOgIY1l+LjkDWuW9r4+57nmDz516cQOsuxi4VXJcx7+Dp1Wezd5xurx1qXXm2/d3QoFnF7iUiUk1Veh0WZ6R1WKqo4gLrEy6bvrQ+8WI+/ag6ga1Ph5SA6Ev/tS1ma1D5cw9NYRYUHj4demzth/+08FoFmVytQ07nCzbHc6whZc8v2DYcdPOGljdbe1MaXaUNB0WkWrts67CI/GUlRbDze9g8D3b8AGXHTx8LaGYd8mg9EAJbXN46XFytwaFWIAS1Pv+5hmHdHPDPPTR2vTZnBJ7jx6zDWIWHra+KaJhwcsPBAeCl0C0iciYFFrn8Sk9Ye1A2fwnbF1rnmpxSr/HpkBLU2jmf8DGZrE8feda21nsh5lJr703Bn3prTvXc/DnwWMzW5fHbDz7/UJeIiCiwyGVSVgy7F1lDyrYFUJJ/+lidhqdDSkh75wwpf4WrO9QOtr5EROSSUGCRS8dcCnsWW0PK1m+hOPf0Mb+wk3NSboUwLRMvIiKVo8Aif425zLo+yuYvrbsCHz92+litYGg9wBpSwjtX/KkZERGRMyiwSOVZzJC2zBpStsyHouzTx3wbQKv+1pDSMF4hRURELgkFFqkYiwX2rbQ+3bPlKyj401YK3vWg1S3WkBLZzXlXjxURkSpLP1nk3AwDDqyxrpOy5SvIO3D6mJc/tOxnnZfS6CotES8iIpeVAovYMww4tN4aUjZ/Bbnpp495+lkfw209EBr3AjcPR1UpIiI1jAKLWENK5qaTIWUeHNt7+pi7r3WzvTa3QpNrwd3LcXWKiEiNpcBSkx3eejqkHNl5ut3NG5r1sYaUpteDu7fjahQREUGBpebJ3mkNKJu+hKytp9tdPaHpdSdDSh/wrOW4GkVERM6gwFITWCyw9r+w+n3I3Hi63cUdontbQ0qzG7SHjYiIOC0Fluru6F74egykLbG+d3GzTphtPdA6gda7jkPLExERqQgFlurKYoHV/4GfnoHSIuvk2V4TocMQ8Knn6OpEREQqRYGlOjqzVyWqB9zyhnYEFhGRKkuBpTopr1fluskQe5+WyBcRkSpNgaW6UK+KiIhUYwosVZ16VUREpAZQYKnK1KsiIiI1hAJLVaReFRERqWEUWKoa9aqIiEgNpMBSVZzVq+ID1z2nXhUREakRFFiqgjN7VSK7Q/9p6lUREZEaQ4HFmVks8Pv78OMzUFqoXhUREamxFFic1dG9MH8spP5mfa9eFRERqcEUWJyNelVERETOosDiTNSrIiIiUi4FFmegXhUREZHzUmBxNPWqiIiIXNBF/e/7m2++SVRUFF5eXsTFxbFq1apznjtz5kxMJpPdy8vLy+6cgoICxowZQ3h4ON7e3rRq1Yrp06dfTGlVh8UCq96Dt7tZw4q7D/R9DRK/UVgRERE5Q6V7WObOnUtSUhLTp08nLi6OqVOn0qdPH7Zv305gYGC51/j5+bF9+3bbe5PJZHc8KSmJn3/+mY8//pioqCh++OEHHnroIUJDQ7nlllsqW6LzU6+KiIhIpVS6h2XKlCmMGjWKESNG2HpCfHx8mDFjxjmvMZlMBAcH215BQUF2x5ctW0ZiYiJXX301UVFR3H///bRv3/68PTdVknpVRERELkqlAktJSQlr1qyhd+/ep2/g4kLv3r1Zvnz5Oa8rKCggMjKSiIgI+vfvz+bNm+2OJyQkMH/+fA4cOIBhGCxatIgdO3Zw/fXXV/LjOLGje+HDW2DBY9aJtZHd4cFl0GWUJtaKiIhcQKWGhLKzszGbzWf1kAQFBbFt27Zyr2nevDkzZsygXbt25Obm8tprr5GQkMDmzZsJDw8H4I033uD+++8nPDwcNzc3XFxceO+99+jZs2e59ywuLqa4uNj2Pi8vrzIf48rSE0AiIiJ/2WV/Sig+Pp74+Hjb+4SEBFq2bMk777zD888/D1gDy4oVK5g/fz6RkZH8+uuvjB49mtDQULvenFOSk5OZPHny5S79r9NcFRERkUuiUoElICAAV1dXMjMz7dozMzMJDg6u0D3c3d3p2LEju3btAuD48eP8/e9/Z968edx0000AtGvXjvXr1/Paa6+VG1gmTpxIUlKS7X1eXh4RERGV+SiXl3pVRERELqlK/fT08PAgJiaGlJQUW5vFYiElJcWuF+V8zGYzGzduJCQkBIDS0lJKS0txOeMHuaurKxaLpdx7eHp64ufnZ/dyGsdSNVdFRETkEqv0kFBSUhKJiYnExsbSpUsXpk6dSmFhISNGjABg2LBhhIWFkZycDMBzzz1H165diY6OJicnh1dffZW0tDRGjhwJWB95vuqqq3j88cfx9vYmMjKSxYsX8+GHHzJlypRL+FEvM/WqiIiIXDaVDiyDBg0iKyuLSZMmkZGRQYcOHVi4cKFtIm56erpdb8mxY8cYNWoUGRkZ1K1bl5iYGJYtW0arVq1s58yZM4eJEycyZMgQjh49SmRkJC+88AIPPPDAJfiIV8CxVPh6jOaqiIiIXCYmwzAMRxfxV+Xl5eHv709ubu6VHR4qr1el92ToPFK9KiIiIhdQmZ/f2kvoYpXbq/IG1Gvs0LJERESqIwWWylKvioiIyBWnwFIZ6lURERFxCAWWilCvioiIiEMpsFyIelVEREQcToHlfA5tgBk3qldFRETEwRRYzieoDQS1AldP9aqIiIg4kALL+bi4wt2fglcd9aqIiIg4kALLhfjUc3QFIiIiNZ66DURERMTpKbCIiIiI01NgEREREaenwCIiIiJOT4FFREREnJ4Ci4iIiDg9BRYRERFxegosIiIi4vQUWERERMTpKbCIiIiI01NgEREREaenwCIiIiJOT4FFREREnJ4Ci4iIiDg9BRYRERFxegosIiIi4vQUWERERMTpKbCIiIiI01NgEREREaenwCIiIiLnVFxmZtnubOat2+/QOtwc+tVFRETEqVgsBlsz8li6K5vfdmazOvUoJ0ot1PZy45b2Ybi6mBxSlwKLiIhIDbf/WBFLd2WzZNcRlu3K5khhid3xgFqedI+uT/6JUur4eDikxosKLG+++SavvvoqGRkZtG/fnjfeeIMuXbqUe+7MmTMZMWKEXZunpycnTpywa9u6dSvjx49n8eLFlJWV0apVK7744gsaNmx4MSWKiIjIOeQeL2X57iMs2ZXF0l1H2JtdaHfcx8OVuEb16BYdQI+mDWgWVAuTyTE9K6dUOrDMnTuXpKQkpk+fTlxcHFOnTqVPnz5s376dwMDAcq/x8/Nj+/bttvdnfujdu3fTvXt37rvvPiZPnoyfnx+bN2/Gy8ursuWJiIjIGYrLzKxNy7EO8+zKZuP+HCzG6eOuLibah/vTPTqAbtEBdGxYFw8355rmajIMw7jwaafFxcXRuXNnpk2bBoDFYiEiIoKxY8cyYcKEs86fOXMm48aNIycn55z3HDx4MO7u7nz00UeVq/6kvLw8/P39yc3Nxc/P76LuISIiUl1YLAbbMvJPDvNks2rvUY6Xmu3OadzAlx4nA0rXJvXx83K/4nVW5ud3pXpYSkpKWLNmDRMnTrS1ubi40Lt3b5YvX37O6woKCoiMjMRisdCpUydefPFFWrduDVgDz3fffccTTzxBnz59WLduHY0aNWLixIkMGDCg3PsVFxdTXFxse5+Xl1eZjyEiIlLtHMw5zpKd1oCybHc22QVnzkPxoFt0gK0XJbSOt4MqvTiVCizZ2dmYzWaCgoLs2oOCgti2bVu51zRv3pwZM2bQrl07cnNzee2110hISGDz5s2Eh4dz+PBhCgoKeOmll/jHP/7Byy+/zMKFC7n11ltZtGgRV1111Vn3TE5OZvLkyZUpXUREpFrJPV7Kij1HWLIzm6W7stlzxjwUb3dX4hrXo3t0AN2bBtA8qLbD56H8FZf9KaH4+Hji4+Nt7xMSEmjZsiXvvPMOzz//PBaLBYD+/fvz6KOPAtChQweWLVvG9OnTyw0sEydOJCkpyfY+Ly+PiIiIy/xJREREHKekzMLa9GO2x43/OGMeiosJ2kfUsfWgdHLCeSh/RaUCS0BAAK6urmRmZtq1Z2ZmEhwcXKF7uLu707FjR3bt2mW7p5ubG61atbI7r2XLlixZsqTce3h6euLp6VmZ0kVERKoUwzDYnplvG+ZZuaeceSgBvnRvenIeSuP6+Htf+XkoV0qlAouHhwcxMTGkpKTY5pdYLBZSUlIYM2ZMhe5hNpvZuHEjffv2td2zc+fOdk8RAezYsYPIyMjKlCciIlKlHco9PQ9l6a4jZBcU2x2v73tyHsrJkBJWxeah/BWVHhJKSkoiMTGR2NhYunTpwtSpUyksLLSttTJs2DDCwsJITk4G4LnnnqNr165ER0eTk5PDq6++SlpaGiNHjrTd8/HHH2fQoEH07NmTXr16sXDhQr755ht++eWXS/MpRUREnFDeiVJW7D5ie9x4T5b9PBQvdxfiGtW3m4fi4qCVZh2t0oFl0KBBZGVlMWnSJDIyMujQoQMLFy60TcRNT0/HxeX0mNmxY8cYNWoUGRkZ1K1bl5iYGJYtW2Y3BDRw4ECmT59OcnIyDz/8MM2bN+eLL76ge/ful+AjioiIOIeSMgvrTs5DWbIrmw37czH/aSKKiwnahf9pHkpkHTzdXB1YsfOo9DoszkjrsIiIiDMyDIMdmQX8tjOLpbuyWbn3KEUl9vNQGgX42gJKfOP6+PtU33koZ7ps67CIiIjIhR3OO8Gc1fuYsyqdg7n2W9HU9/UgITqAHtEBJETXJ7yuj4OqrFoUWERERC4BwzBYtfcoH65I4/tNGZSdHOrxcnehS6P6dI+uT/foBrQIrrnzUP4KBRYREZG/oKC4jHnrDvDx8jS2Z+bb2mMj6zI0PpI+rYPxctc8lL9KgUVEROQi7MjM56PlaXy5dj+FJ+eleLu7MqBjGPd0bUjrUH8HV1i9KLCIiIhUUKnZwg+bM/lweSor9x61tTdu4MvQrpHcFhPukE0EawIFFhERkQvIyD3BJ6vS+WRVOofzrYu5ubqYuK5lEEPjI0loUr9K79NTFSiwiIiIlMMwDJbvOcLHK9L4fnOmbb2UgFqe3N0lgrviGhLiX3NWmnU0BRYREZE/yTtRyry1B/hoRRq7DhfY2rs0qsfQrtZJtNVpU8GqQoFFREQE2JaRx4fL0/hq3QHb4m6+Hq4M7BTGPV0jaRGshUkdSYFFRERqrJIyCws3Z/DR8lRWpx6ztUcH1mJYfCQDO4ZRW5NonYICi4iI1DgHc44ze2U6c1ank11QAlgn0fZpHcTQrlF0bVxPk2idjAKLiIjUCIZhsHTXET5cnspPWzM5tedgYG1P7o5ryF1dGhLk5+XYIuWcFFhERKRayz1eyhdr9vPxijT2ZBfa2rs2rsew+CiuaxWEu6sm0To7BRYREamWNh/M5eMVaXy17iDHS62TaGt5unFrpzCGdo2kaVBtB1colaHAIiIi1UZxmZn/bczgw+WprE3PsbU3D6rN0PhIBnQMo5anfvRVRfquiYhIlbf/WBGzVqbz6ep9HCm0TqJ1czFxQ5tghsVH0TmqribRVnEKLCIiUiVZLAa/7crmo+Wp/LztsG0SbbCfF3fHNWRwlwgCa2sSbXWhwCIiIlVKTlEJn/2+n49XppF2pMjW3i26PkO7RtK7ZRBumkRb7SiwiIhIlbBxfy4fLk9l/oaDFJdZAKjt6cZtMeHc0zWS6MBaDq5QLicFFhERcVonSs1898chPlyRxoZ9Obb2FsG1GRYfxYCOofh46EdZTaDvsoiIOJ30I0XMWpXGp6v3cayoFAB3VxN924YwLD6STg01ibamUWARERGnkFtUyncbD/HV+gOs2nvU1h7q78WQrpHcGRtBg9qeDqxQHEmBRUREHOZEqZlF2w4zb90BftmeRYnZYjvWo2kAQ7tGck2LQE2iFQUWERG5siwWgxV7j/D1uoMs2HSI/BNltmMtgmszoGMYt7QPJbSOtwOrFGejwCIiIlfE1kN5fLXuAPM3HORQ7glbe4i/F7d0CGVAhzBahvg5sEJxZgosIiJy2RzIOc789Qf5at0Btmfm29pre7lxU9sQBnQMo0tUPVxcNIFWzk+BRURELqncolIWbDrEV+sOsPJPk2c9XF3o1aIBAzuGcXXzQLzcXR1YpVQ1CiwiIvKXnSg188t26+TZRdvsJ8/GNarHwI5h3NgmBH8fdwdWKVWZAouIiFwUi8Vg5d6jfL3+AN9ttJ882zzo5OTZDqGEafKsXAIKLCIiUinbMvKYt+4A36w/yME/TZ4N9vOif0dNnpXL46IebH/zzTeJiorCy8uLuLg4Vq1adc5zZ86ciclksnt5eZ1798wHHngAk8nE1KlTL6Y0ERG5DA7mHGf64t3cMPVXbpj6G+8s3sPB3BPU9nJjUGwEn4zqyrIJ1zDxxpYKK3JZVLqHZe7cuSQlJTF9+nTi4uKYOnUqffr0Yfv27QQGBpZ7jZ+fH9u3b7e9P9dyyvPmzWPFihWEhoZWtiwREbnEco+X8r+TK8+u3HsUw7C2n5o8O6BDGL1aaPKsXBmVDixTpkxh1KhRjBgxAoDp06fz3XffMWPGDCZMmFDuNSaTieDg4PPe98CBA4wdO5bvv/+em266qbJliYjIJVBcZmbRtiy+WneAn7cdtps82+Xk5Nm+mjwrDlCpwFJSUsKaNWuYOHGirc3FxYXevXuzfPnyc15XUFBAZGQkFouFTp068eKLL9K6dWvbcYvFwtChQ3n88cft2s+luLiY4uJi2/u8vLzKfAwREfkTi8VgVerJybN/HCLvT5NnmwXVsq08G17Xx4FVSk1XqcCSnZ2N2WwmKCjIrj0oKIht27aVe03z5s2ZMWMG7dq1Izc3l9dee42EhAQ2b95MeHg4AC+//DJubm48/PDDFaojOTmZyZMnV6Z0ERE5w/aMfOatO8D89QfOnjzbIZT+HcJoGVJbuyKLU7jsTwnFx8cTHx9ve5+QkEDLli155513eP7551mzZg2vv/46a9eurfB/FBMnTiQpKcn2Pi8vj4iIiEteu4hIdXMo17ry7Lx1B9iW8aeVZz3duLFtMAM6hhHXqD6uWnlWnEylAktAQACurq5kZmbatWdmZl5wjsop7u7udOzYkV27dgHw22+/cfjwYRo2bGg7x2w283//939MnTqV1NTUs+7h6emJp6e2GBcRqYjc46Us3HSIr9YdZMXeI7bJs+6uJno1D2RAxzCu0eRZcXKVCiweHh7ExMSQkpLCgAEDAOv8k5SUFMaMGVOhe5jNZjZu3Ejfvn0BGDp0KL1797Y7p0+fPgwdOtQ2sVdERCqnuMzML9utk2dTth2mpOxPk2ej6jGgYxh92wZTx8fDgVWKVFylh4SSkpJITEwkNjaWLl26MHXqVAoLC23hYtiwYYSFhZGcnAzAc889R9euXYmOjiYnJ4dXX32VtLQ0Ro4cCUD9+vWpX7++3ddwd3cnODiY5s2b/9XPJyJSYxiGwerUY8xbd4AFGw+Re7zUdkyTZ6Wqq3RgGTRoEFlZWUyaNImMjAw6dOjAwoULbRNx09PTcXE5vR7dsWPHGDVqFBkZGdStW5eYmBiWLVtGq1atLt2nEBGpwQ7lHueLNfv5bM1+0o4U2dqD/Dzp3yHs5MqzmjwrVZvJME6NZlZdeXl5+Pv7k5ubi5+fVlgUkeqvuMzMT1sO8+nv+/htZxaWk/+S1/J0o2/bYAZ0CCOusSbPinOrzM9v7SUkIlKFbDmYx6e/7+Or9QfIKTo95BPXqB53xkZwY9tgfDz0T7tUP/pbLSLi5HKLSpm/4QCf/r6fjQdybe3Bfl7cHhPO7THhRAX4OrBCkctPgUVExAlZLAbLdh/h09/3sXBzhu0pH3dXE9e3CuaO2HB6NG2gIR+pMRRYREScyL6jRXy+Zj+fr9nPgZzjtvYWwbUZ1DmC/h3CqOerR5Gl5lFgERFxsBOlZr7fnMGnv+9j6a4jtnY/Lzf6dwjjztgI2oT56SkfqdEUWEREHMAwDDYeyOXT3/fx9fqD5J/ccNBkgm5NArgjNpw+rYO1+qzISQosIiJX0NHCEuatO8Bnv++z28snrI43d8SGc1uncCLqaWE3kTMpsIiIXGZmi8GvO7P47Pd9/Lglk1KzddEUDzcXbmwTzJ2xEcQ3ro+LJtCKnJMCi4jIZZKaXchna/bxxZoDZOSdsLW3C/fnjtgIbmkXir+PuwMrFKk6FFhERC6hopIyFmy0TqBdtfeorb2ujzsDO4ZzR2w4LUO0IrdIZSmwiIj8RYZhsDY9h89+38c3Gw5SWGIGwMUEPZs1YFBsBNe0DMTTTRNoRS6WAouIyEU6nH+CeWsP8Onv+9idVWhrj6zvw52xEdzaKYwQf28HVihSfSiwiIhUQqnZwqJth/n09/0s2n4Y88ldB73dXenbNoQ7Y8Pp0qie1kwRucQUWEREKmDX4Xw+/X0/X67dT3ZBia29U8M63BkbwU3tQqjtpQm0IpeLAouIyDnknyjluz8O8env+1ibnmNrD6jlyW2dwrgjNpzowNqOK1CkBlFgERH5E8MwWLX3KJ/+vp8FGw9xvNQ6gdbVxcQ1LQK5MzaCq5s3wN3VxcGVitQsCiwiIkBG7gm+WLufz37fR+qRIlt7kwa+DOocwYCOYQTW9nJghSI1mwKLiNRo69KP8d5ve1i4KYOT82ep5elGv/Yh3B4TQaeGdTSBVsQJKLCISI1jthj8tDWT937dw+9px2ztXRrV487YCPq2DcbHQ/88ijgT/RcpIjXG8RIzn6/dz/u/7bEN+3i4utC/QygjezSmebAm0Io4KwUWEan2svKL+Wh5Kh+tSONYUSkA/t7u3NO1IYnxUQT6aW6KiLNTYBGRamvX4Xz+89tevlx3gJIyCwAR9by5r1sj7oiNwNdT/wSKVBX6r1VEqhXDMFi59yjv/bqHlG2Hbe0dIupwf8/G9GkdjKuLJtGKVDUKLCJSLZSZLSzYlMF7v+5h44FcAEwmuK5lEPf3bExMZF097SNShSmwiEiVVlBcxpxV6XywNJUDOccB8HRz4Y7YcO7r3phGAb4OrlBELgUFFhGpkg7lHmfm0lRmr0wnv7gMgPq+HiQmRHFP10jq+Xo4uEIRuZQUWESkStl8MJf//LaXbzYcpOzkSm9NGvgyqkdjBnQMw8vd1cEVisjloMAiIk7PMAwW78jiP7/tZcmubFt718b1GNWjMb2aB+KiibQi1ZoCi4g4reIyM/PXH+Q/v+1le2Y+YN2EsG/bEEb1aES78DqOLVBErhgFFhFxOjlFJcxamc7MZalk5RcD4OvhyuAuDRnRLYrwuj4OrlBErjQFFhFxGulHipixdC9zV+/jeKkZgGA/L0Z0i2Jwl4b4e7s7uEIRcRSXi7nozTffJCoqCi8vL+Li4li1atU5z505cyYmk8nu5eV1ehns0tJSxo8fT9u2bfH19SU0NJRhw4Zx8ODBiylNRKqgdenHeGjWGq5+bREzl6VyvNRMyxA//jWoPb8+0Yu/XdVEYUWkhqt0D8vcuXNJSkpi+vTpxMXFMXXqVPr06cP27dsJDAws9xo/Pz+2b99ue//nxZuKiopYu3YtTz/9NO3bt+fYsWM88sgj3HLLLfz+++8X8ZFEpCo4147JVzVrwKgejekWXV8LvYmIjckwDKMyF8TFxdG5c2emTZsGgMViISIigrFjxzJhwoSzzp85cybjxo0jJyenwl9j9erVdOnShbS0NBo2bHjB8/Py8vD39yc3Nxc/P78Kfx0RufJO7Zg8Y8le9mYXAuDuaqJ/hzBG9mhEi2D9NyxSU1Tm53elelhKSkpYs2YNEydOtLW5uLjQu3dvli9ffs7rCgoKiIyMxGKx0KlTJ1588UVat259zvNzc3MxmUzUqVOn3OPFxcUUFxfb3ufl5VXmY4iIA2QXFPPh8jQ+Wp5qt2PykLiGJCZEEaQdk0XkPCoVWLKzszGbzQQFBdm1BwUFsW3btnKvad68OTNmzKBdu3bk5uby2muvkZCQwObNmwkPDz/r/BMnTjB+/Hjuuuuuc6at5ORkJk+eXJnSRcRBdh0u4P0le/hirXZMFpGLd9n/pYiPjyc+Pt72PiEhgZYtW/LOO+/w/PPP251bWlrKnXfeiWEYvP322+e858SJE0lKSrK9z8vLIyIi4tIXLyIX5Vw7JrePqMP9PRrTp3UQbq4XNedfRGqoSgWWgIAAXF1dyczMtGvPzMwkODi4Qvdwd3enY8eO7Nq1y679VFhJS0vj559/Pu9YlqenJ56enpUpXUSugPPtmDyqZ2NitWOyiFykSgUWDw8PYmJiSElJYcCAAYB10m1KSgpjxoyp0D3MZjMbN26kb9++trZTYWXnzp0sWrSI+vXrV6YsEXGwc+2YfHtMOPd1b0TjBrUcXKGIVHWVHhJKSkoiMTGR2NhYunTpwtSpUyksLGTEiBEADBs2jLCwMJKTkwF47rnn6Nq1K9HR0eTk5PDqq6+SlpbGyJEjAWtYuf3221m7di3ffvstZrOZjIwMAOrVq4eHh3ZcFXFGFovBlkN5fLPhILNXpZN/4vSOycPio7ina0Pq11JPqIhcGpUOLIMGDSIrK4tJkyaRkZFBhw4dWLhwoW0ibnp6Oi4up8emjx07xqhRo8jIyKBu3brExMSwbNkyWrVqBcCBAweYP38+AB06dLD7WosWLeLqq6++yI8mIpda+pEiluzKZumubJbtzrY97QPQ+OSOyQO1Y7KIXAaVXofFGWkdFpHLI7ugmGW7j7BsVzZLdmWz/9hxu+O+Hq50bVyfu7o05JoW2jFZRCrnsq3DIiLVW2FxGatSj7J0ZzZLdx9h6yH7NY7cXU10jKhLt+gAukXXp31EHdz1tI+IXAEKLCI1WKnZwh/7c1iy8whLd2Wzbt8xSs32na4tQ/zo1qQ+3ZoG0CWqntZNERGH0L88IjWIYRjsyCxg6cl5KCv2HKGwxGx3Tlgdb3o0DSAhOoCEJvUJ0MRZEXECCiwi1dzBnOMs2ZXNsl3WYZ6s/GK743V93EloEmAb5mlYz0drpYiI01FgEalmcotKWb4n+2RIOcKekxsMnuLl7kKXRvWtwzzRAbQK8dNkWRFxegosIlXciVIzv6ceY+lu6zDPxgO5/PnZPxeTdUn87tEBJDQJoFNkHTzd9NixiFQtCiwiVYzZYrDpQK4toKxOPWbbVPCUpoG1Tg7xBBDXuB5+Xu4OqlZE5NJQYBFxcoZhsDe78ORE2SMs251N3slVZU8J9vOyzUHpFh1AkJ+Xg6oVEbk8FFhEnNDh/BMs23XEtqrsodwTdsdre7kR37g+3Ztah3maNPDVRFkRqdYUWEScQP6JUlbuOWob5tmRWWB33MPVhdiourZhnjahfrhpwTYRqUEUWEQcoMxsYU3aMZbuti7Ytn5fDmbL6ZmyJhO0CfW3DfPERtbD20MTZUWk5lJgEbnCVu45wqSvN7M9M9+uPaq+D92iA+geHUDXxvWp66udykVETlFgEblCDuefIHnBNuatOwCAn5cbVzcPtD5uHF2f8Lo+Dq5QRMR5KbCIXGZlZgsfrUhjyg87yC8uw2SCu7s05PE+zanjo14UEZGKUGARuYzWpB3lqa8223Y9bhfuz/P929A+oo5jCxMRqWIUWEQug+yCYl7+3zY+W7MfAH9vd564oTmDOzfEVcvgi4hUmgKLyCVkthjMXpnGq99vty3uNig2gvE3tqCeJtGKiFw0BRaRS2Rd+jGe/noTmw5Yh39ah/rx/IA2dGpY18GViYhUfQosIn/R0cISXlm4jTmr9wHWVWgf79OcIXGRGv4REblEFFhELpLFYjBn9T5e+X4bOUWlANzWKZyJfVsQUMvTwdWJiFQvCiwiF+GP/Tk8/fVmNuzLAaBFcG2eH9CGzlH1HFuYiEg1pcAiUgk5RSW8+v12Zq9KxzCglqcbSdc1Y1h8pPb2ERG5jBRYRCrAYjH4fM1+Xlq4jaOFJQAM6BDK3/u2JNDPy8HViYhUfwosIhew+WAuT3+1ibXpOQA0C6rFc/3b0LVxfccWJiJSgyiwiJxD7vFS/vXjDj5cnorFAF8PV8b1bsbwblG4a/hHROSKUmAROYNhGMxbd4AXF2wju6AYgJvbhfDUTa0I9tfwj4iIIyiwiPzJtow8nv5qE6tTjwHQpIEvz/VvQ7foAAdXJiJSsymwiAD5J0qZ+tNOZi5LxWwx8HZ35eFrm3Jf90Z4uGn4R0TE0RRYpEYzDIP5Gw7yj++2kpVvHf65sU0wT93cirA63g6uTkRETlFgkRprZ2Y+k77ezPI9RwBoFODLs7e05qpmDRxcmYiInOmi+rrffPNNoqKi8PLyIi4ujlWrVp3z3JkzZ2IymexeXl72ExcNw2DSpEmEhITg7e1N79692blz58WUJnJBhcVlJC/Yyo2v/8byPUfwdHPh/65rxsJxPRRWREScVKUDy9y5c0lKSuKZZ55h7dq1tG/fnj59+nD48OFzXuPn58ehQ4dsr7S0NLvjr7zyCv/+97+ZPn06K1euxNfXlz59+nDixInKfyKRczAMg+/+OMS1/1zMO7/uocxicF2rIH5Kuoqx1zbF083V0SWKiMg5mAzDMCpzQVxcHJ07d2batGkAWCwWIiIiGDt2LBMmTDjr/JkzZzJu3DhycnLKvZ9hGISGhvJ///d/PPbYYwDk5uYSFBTEzJkzGTx48AVrysvLw9/fn9zcXPz8/CrzcaSG2J1VwDNfb2bJrmwAIup5M/mW1lzTIsjBlYmI1FyV+fldqR6WkpIS1qxZQ+/evU/fwMWF3r17s3z58nNeV1BQQGRkJBEREfTv35/Nmzfbju3du5eMjAy7e/r7+xMXF3fOexYXF5OXl2f3EilPUUkZryzcxg1Tf2XJrmw83FwY17spPz56lcKKiEgVUqnAkp2djdlsJijI/h/6oKAgMjIyyr2mefPmzJgxg6+//pqPP/4Yi8VCQkIC+/fvB7BdV5l7Jicn4+/vb3tFRERU5mNIDWAYBgs3ZXDdlF9565fdlJoNejVvwI+P9mRc72Z4uWv4R0SkKrnsTwnFx8cTHx9ve5+QkEDLli155513eP755y/qnhMnTiQpKcn2Pi8vT6FFbFKzC3n2m838sj0LgLA63jzTrxXXtQrCZDI5uDoREbkYlQosAQEBuLq6kpmZadeemZlJcHBwhe7h7u5Ox44d2bVrF4DtuszMTEJCQuzu2aFDh3Lv4enpiaenZ2VKlxrgRKmZtxbtYvriPZSYLXi4unB/z8aM7hWNt4d6VEREqrJKDQl5eHgQExNDSkqKrc1isZCSkmLXi3I+ZrOZjRs32sJJo0aNCA4OtrtnXl4eK1eurPA9RX7akknvKYv598+7KDFb6NE0gIXjevBYn+YKKyIi1UClh4SSkpJITEwkNjaWLl26MHXqVAoLCxkxYgQAw4YNIywsjOTkZACee+45unbtSnR0NDk5Obz66qukpaUxcuRIAEwmE+PGjeMf//gHTZs2pVGjRjz99NOEhoYyYMCAS/dJpVrad7SIZ+dvJmWb9bH6EH8vJt3cihvaBGv4R0SkGql0YBk0aBBZWVlMmjSJjIwMOnTowMKFC22TZtPT03FxOd1xc+zYMUaNGkVGRgZ169YlJiaGZcuW0apVK9s5TzzxBIWFhdx///3k5OTQvXt3Fi5ceNYCcyKnnCg1887iPbz1yy6Kyyy4uZgY2aMxD18bjY+HFnAWEaluKr0OizPSOiw1y6Lth3l2/mbSjhQBkNCkPs/1b010YG0HVyYiIpVRmZ/f+l9RqRIsFoPFO7KYsXQvv+20Lv4W5OfJUze14uZ2IRr+ERGp5hRYxKkVFJfx+e/7+O/yNPZmFwLg6mLi3m5RPNK7GbU89VdYRKQm0L/24pTSjxQxc1kqn/2+j/ziMgBqe7kxuHMEw+KjiKjn4+AKRUTkSlJgEadhGAbLdx9hxtJUUrZlcmp2VeMGvoxIiOLWTuH4qkdFRKRG0r/+4nAnSs3MW3eAmUtT2Z6Zb2u/qlkDRnSLomfTBri4aI6KiEhNpsAiDnMo9zgfLk/jk1Xp5BSVAuDj4cptncJJTIgiOrCWgysUERFnocAiV5RhGKxNP8aMpaks3JSB2WId9wmv683whCjuiI3A39vdwVWKiIizUWCRK6KkzMJ3Gw/ywdJU/tifa2vv2rgeI7o1onfLIFw17CMiIuegwCKXVVZ+MbNWpjFrZTpZ+cUAeLi5MKBDKMMTGtEqVAv9iYjIhSmwyGWx6UAuM5bu5dsNhygxWwDrQm9Du0ZyV5eG1K+l3bZFRKTiFFjkkikzW/hhSyYfLN3L6tRjtvaODeswolsjbmwTjLtrpTYIFxERARRY5BLIKSrhk1X7+Gh5KgdzTwDg5mLipnYhjOjWiA4RdRxboIiIVHkKLHLRdmTm88HSVOat28+JUuuwT31fD+6Oa8g9XSMJ8tNu2yIicmkosEilWCwGi7Yf5oOlqSzZlW1rbxnix4huUdzSPhQvd1cHVigiItWRAotUSP6JUj77fT//XZ5K2pEiAFxMcH2rYEZ0i6JLo3raMVlERC4bBRY5r9TsQmYuS+XzNfspOLkJoZ+XG4O7NGRo10htQigiIleEAoucxTAMlu46wgdL9/Lz9sO2TQibNPBleLdG3NYpDB8P/dUREZErRz91xOZ4yclNCJftZUdmga29V/MGDO/WiB7RAdqEUEREHEKBRTiQc5wPl6cyZ9U+co+f3oTwjhjrJoSNG2gTQhERcSwFlhrKMAx+TzvGB0v38v3mTNsmhBH1vEmMj+LOzhH4eWkTQhERcQ4KLDVMcZmZbzcc4oNle9l0IM/WHt+4PiO6RXGtNiEUEREnpMBSQ2QXFPPRcusmhNkF1k0IPd1cGNgxjMSEKFqGaBNCERFxXgosNcBvO7N4+JN1HCuyzk8J9vNiaLx1E8J6vh4Ork5EROTCFFiqMYvF4O3Fu3nth+0YBrQIrs3oXtHcoE0IRUSkilFgqaZyj5fyf59u4KetmQAMio1gcv/WWjZfRESqJAWWamhbRh4PfLSG1CNFeLi58NwtrRncpaGjyxIREbloCizVzFfrDjDhyz84UWohrI43b9/TiXbhdRxdloiIyF+iwFJNlJRZeOG7Lfx3eRoAPZoG8PrgjppUKyIi1YICSzWQkXuCh2atYW16DgBjr4lmXO9mWk9FRESqDQWWKm757iOM/WQt2QUl1PZy4193dqB3qyBHlyUiInJJXdSzrW+++SZRUVF4eXkRFxfHqlWrKnTdnDlzMJlMDBgwwK69oKCAMWPGEB4ejre3N61atWL69OkXU1qNYRgG7/26h3veX0l2QQktgmvzzZjuCisiIlItVTqwzJ07l6SkJJ555hnWrl1L+/bt6dOnD4cPHz7vdampqTz22GP06NHjrGNJSUksXLiQjz/+mK1btzJu3DjGjBnD/PnzK1tejVBQXMbo2Wt5YcFWzBaDgR3DmPdQN6ICfB1dmoiIyGVR6cAyZcoURo0axYgRI2w9IT4+PsyYMeOc15jNZoYMGcLkyZNp3LjxWceXLVtGYmIiV199NVFRUdx///20b9++wj03Ncmuw/n0n7aEBRszcHc18Xz/1ky5sz3eHlpfRUREqq9KBZaSkhLWrFlD7969T9/AxYXevXuzfPnyc1733HPPERgYyH333Vfu8YSEBObPn8+BAwcwDINFixaxY8cOrr/++sqUV+0t2HiI/tOWsjurkGA/L+b+LZ6h8VGYTJpcKyIi1VulJt1mZ2djNpsJCrKfJxEUFMS2bdvKvWbJkiW8//77rF+//pz3feONN7j//vsJDw/Hzc0NFxcX3nvvPXr27Fnu+cXFxRQXF9ve5+XllXtedVFmtvDywm2899tewLqz8ht3dySglqeDKxMREbkyLutTQvn5+QwdOpT33nuPgICAc573xhtvsGLFCubPn09kZCS//voro0ePJjQ01K4355Tk5GQmT558OUt3Gln5xYyZvZaVe48C8LerGvP49c1x015AIiJSg5gMwzAqenJJSQk+Pj58/vnndk/6JCYmkpOTw9dff213/vr16+nYsSOurqfnV1gsFsA6lLR9+3ZCQ0Px9/dn3rx53HTTTbbzRo4cyf79+1m4cOFZdZTXwxIREUFubi5+fn4V/ThOb03aUR6atZbMvGJ8PVx57Y723Ng2xNFliYiIXBJ5eXn4+/tX6Od3pXpYPDw8iImJISUlxRZYLBYLKSkpjBkz5qzzW7RowcaNG+3annrqKfLz83n99deJiIjgxIkTlJaW4uJi32Pg6upqCzdn8vT0xNOz+g6HGIbBh8vTeP7bLZRZDKIDazH9nhiiA2s5ujQRERGHqPSQUFJSEomJicTGxtKlSxemTp1KYWEhI0aMAGDYsGGEhYWRnJyMl5cXbdq0sbu+Tp06ALZ2Dw8PrrrqKh5//HG8vb2JjIxk8eLFfPjhh0yZMuUvfryqp6ikjL9/uZGv1h8E4KZ2IbxyWzt8PbXGn4iI1FyV/ik4aNAgsrKymDRpEhkZGXTo0IGFCxfaJuKmp6ef1VtyIXPmzGHixIkMGTKEo0ePEhkZyQsvvMADDzxQ2fKqtL3ZhTzw0Rq2Z+bj6mJi4o0tuK97Iz0FJCIiNV6l5rA4q8qMgTmrH7dkkjR3PfnFZQTU8uTNuzsS17i+o8sSERG5bC7bHBa59MwWg3/9uINpi3YBEBtZlzeHdCLIz8vBlYmIiDgPBRYHOlpYwiNz1vHbzmwARnSL4u99W+KuR5ZFRETsKLA4yIZ9OTw0ay0Hco7j7e7KS7e1pX+HMEeXJSIi4pQUWK4wwzCYs3ofz3y9mRKzhUYBvky/J4bmwbUdXZqIiIjTUmC5gk6Umpn09SY+/X0/ANe3CuK1O9vj5+Xu4MpEREScmwLLFbLvaBEPfLyGzQfzcDHB431a8MBVjfXIsoiISAUosFwBi7YfZtyc9eQeL6Werwdv3NWRbtHn3ltJRERE7CmwXEYWi8EbP+9iasoODAPaR9Th7SGdCK3j7ejSREREqhQFlsskt6iUcXPXsWh7FgBD4hoyqV8rPN1cL3CliIiInEmB5TLYfDCXBz9eS/rRIjzdXPjHgDbcERvh6LJERESqLAWWS+yLNfv5+7yNFJdZiKjnzfR7Ymgd6u/oskRERKo0BZZLpLjMzPPfbuHjFekA9GregKmDOuLvo0eWRURE/ioFlkvgYM5xHpy1lg37cjCZYNy1zRh7TTQuLnpkWURE5FJQYPmLlu7KZuwn6zhaWIK/tztTB3egV/NAR5clIiJSrSiwXCTDMJi+eA+vfr8NiwGtQ/2Yfk8MEfV8HF2aiIhItaPAchHyTpTy+Gcb+H5zJgB3xITz/IA2eLnrkWUREZHLQYGlknZk5vPAR2vYk12Ih6sLk/u3ZnDnCC2xLyIichkpsFTC/A0HGf/5HxwvNRPq78Xb98TQPqKOo8sSERGp9hRYKqDUbCF5wTZmLN0LQPfoAP59V0fq+Xo4uDIREZGaQYHlAg7nnWD07LWsTj0GwOheTUi6rjmuemRZRETkilFgOY8dmfkM+c9KsvKLqe3pxj/vbM/1rYMdXZaIiEiNo8ByHhF1fajv60E9Hw+mD42hUYCvo0sSERGpkRRYzsPbw5UZwztTx8cdHw/9UYmIiDiKfgpfQGgdb0eXICIiUuO5OLoAERERkQtRYBERERGnp8AiIiIiTk+BRURERJyeAouIiIg4PQUWERERcXoKLCIiIuL0LiqwvPnmm0RFReHl5UVcXByrVq2q0HVz5szBZDIxYMCAs45t3bqVW265BX9/f3x9fencuTPp6ekXU56IiIhUM5UOLHPnziUpKYlnnnmGtWvX0r59e/r06cPhw4fPe11qaiqPPfYYPXr0OOvY7t276d69Oy1atOCXX37hjz/+4Omnn8bLy6uy5YmIiEg1ZDIMw6jMBXFxcXTu3Jlp06YBYLFYiIiIYOzYsUyYMKHca8xmMz179uTee+/lt99+Iycnh6+++sp2fPDgwbi7u/PRRx9d1IfIy8vD39+f3Nxc/Pz8LuoeIiIicmVV5ud3pXpYSkpKWLNmDb179z59AxcXevfuzfLly8953XPPPUdgYCD33XffWccsFgvfffcdzZo1o0+fPgQGBhIXF2cXaM5UXFxMXl6e3UtERESqr0oFluzsbMxmM0FBQXbtQUFBZGRklHvNkiVLeP/993nvvffKPX748GEKCgp46aWXuOGGG/jhhx8YOHAgt956K4sXLy73muTkZPz9/W2viIiIynwMERERqWIu61NC+fn5DB06lPfee4+AgIByz7FYLAD079+fRx99lA4dOjBhwgRuvvlmpk+fXu41EydOJDc31/bat2/fZfsMIiIi4niV2q05ICAAV1dXMjMz7dozMzMJDg4+6/zdu3eTmppKv379bG2nAoqbmxvbt28nIiICNzc3WrVqZXdty5YtWbJkSbl1eHp64unpaXt/ahqOhoZERESqjlM/tysynbZSgcXDw4OYmBhSUlJsjyZbLBZSUlIYM2bMWee3aNGCjRs32rU99dRT5Ofn8/rrrxMREYGHhwedO3dm+/btduft2LGDyMjICtWVn58PoKEhERGRKig/Px9/f//znlOpwAKQlJREYmIisbGxdOnShalTp1JYWMiIESMAGDZsGGFhYSQnJ+Pl5UWbNm3srq9Tpw6AXfvjjz/OoEGD6NmzJ7169WLhwoV88803/PLLLxWqKTQ0lH379lG7dm1MJlNlP9J55eXlERERwb59+/QEkhPQ98O56PvhfPQ9cS76fpyfYRjk5+cTGhp6wXMrHVgGDRpEVlYWkyZNIiMjgw4dOrBw4ULbRNz09HRcXCo3NWbgwIFMnz6d5ORkHn74YZo3b84XX3xB9+7dK3S9i4sL4eHhlf0oleLn56e/bE5E3w/nou+H89H3xLno+3FuF+pZOaXS67DUNFrjxbno++Fc9P1wPvqeOBd9Py4d7SUkIiIiTk+B5QI8PT155pln7J5KEsfR98O56PvhfPQ9cS76flw6GhISERERp6ceFhEREXF6CiwiIiLi9BRYRERExOkpsIiIiIjTU2C5gDfffJOoqCi8vLyIi4tj1apVji6pRkpOTqZz587Url2bwMBABgwYcNZ2DuI4L730EiaTiXHjxjm6lBrrwIED3HPPPdSvXx9vb2/atm3L77//7uiyaiSz2czTTz9No0aN8Pb2pkmTJjz//PMV2i9Hzk2B5Tzmzp1LUlISzzzzDGvXrqV9+/b06dOHw4cPO7q0Gmfx4sWMHj2aFStW8OOPP1JaWsr1119PYWGho0ur8VavXs0777xDu3btHF1KjXXs2DG6deuGu7s7//vf/9iyZQv//Oc/qVu3rqNLq5Fefvll3n77baZNm8bWrVt5+eWXeeWVV3jjjTccXVqVpseazyMuLo7OnTszbdo0wLrRY0REBGPHjmXChAkOrq5my8rKIjAwkMWLF9OzZ09Hl1NjFRQU0KlTJ9566y3+8Y9/0KFDB6ZOnerosmqcCRMmsHTpUn777TdHlyLAzTffTFBQEO+//76t7bbbbsPb25uPP/7YgZVVbephOYeSkhLWrFlD7969bW0uLi707t2b5cuXO7AyAcjNzQWgXr16Dq6kZhs9ejQ33XST3X8ncuXNnz+f2NhY7rjjDgIDA+nYsSPvvfeeo8uqsRISEkhJSWHHjh0AbNiwgSVLlnDjjTc6uLKqrdKbH9YU2dnZmM1m26aOpwQFBbFt2zYHVSVg7ekaN24c3bp1O2s3cLly5syZw9q1a1m9erWjS6nx9uzZw9tvv01SUhJ///vfWb16NQ8//DAeHh4kJiY6urwaZ8KECeTl5dGiRQtcXV0xm8288MILDBkyxNGlVWkKLFLljB49mk2bNrFkyRJHl1Jj7du3j0ceeYQff/wRLy8vR5dT41ksFmJjY3nxxRcB6NixI5s2bWL69OkKLA7w6aefMmvWLGbPnk3r1q1Zv34948aNIzQ0VN+Pv0CB5RwCAgJwdXUlMzPTrj0zM5Pg4GAHVSVjxozh22+/5ddffyU8PNzR5dRYa9as4fDhw3Tq1MnWZjab+fXXX5k2bRrFxcW4uro6sMKaJSQkhFatWtm1tWzZki+++MJBFdVsjz/+OBMmTGDw4MEAtG3blrS0NJKTkxVY/gLNYTkHDw8PYmJiSElJsbVZLBZSUlKIj493YGU1k2EYjBkzhnnz5vHzzz/TqFEjR5dUo1177bVs3LiR9evX216xsbEMGTKE9evXK6xcYd26dTvrMf8dO3YQGRnpoIpqtqKiIlxc7H+8urq6YrFYHFRR9aAelvNISkoiMTGR2NhYunTpwtSpUyksLGTEiBGOLq3GGT16NLNnz+brr7+mdu3aZGRkAODv74+3t7eDq6t5ateufdb8IV9fX+rXr695RQ7w6KOPkpCQwIsvvsidd97JqlWrePfdd3n33XcdXVqN1K9fP1544QUaNmxI69atWbduHVOmTOHee+91dGlVmyHn9cYbbxgNGzY0PDw8jC5duhgrVqxwdEk1ElDu64MPPnB0aXLSVVddZTzyyCOOLqPG+uabb4w2bdoYnp6eRosWLYx3333X0SXVWHl5ecYjjzxiNGzY0PDy8jIaN25sPPnkk0ZxcbGjS6vStA6LiIiIOD3NYRERERGnp8AiIiIiTk+BRURERJyeAouIiIg4PQUWERERcXoKLCIiIuL0FFhERETE6SmwiEi1YTKZ+OqrrxxdhohcBgosInJJDB8+HJPJdNbrhhtucHRpIlINaC8hEblkbrjhBj744AO7Nk9PTwdVIyLViXpYROSS8fT0JDg42O5Vt25dwDpc8/bbb3PjjTfi7e1N48aN+fzzz+2u37hxI9dccw3e3t7Ur1+f+++/n4KCArtzZsyYQevWrfH09CQkJIQxY8bYHc/OzmbgwIH4+PjQtGlT5s+fbzt27NgxhgwZQoMGDfD29qZp06ZnBSwRcU4KLCJyxTz99NPcdtttbNiwgSFDhjB48GC2bt0KQGFhIX369KFu3bqsXr2azz77jJ9++skukLz99tuMHj2a+++/n40bNzJ//nyio6PtvsbkyZO58847+eOPP+jbty9Dhgzh6NGjtq+/ZcsW/ve//7F161befvttAgICrtwfgIhcPEfvvigi1UNiYqLh6upq+Pr62r1eeOEFwzCsO24/8MADdtfExcUZDz74oGEYhvHuu+8adevWNQoKCmzHv/vuO8PFxcXIyMgwDMMwQkNDjSeffPKcNQDGU089ZXtfUFBgAMb//vc/wzAMo1+/fsaIESMuzQcWkStKc1hE5JLp1asXb7/9tl1bvXr1bL+Pj4+3OxYfH8/69esB2Lp1K+3bt8fX19d2vFu3blgsFrZv347JZOLgwYNce+21562hXbt2tt/7+vri5+fH4cOHAXjwwQe57bbbWLt2Lddffz0DBgwgISHhoj6riFxZCiwicsn4+vqeNURzqXh7e1foPHd3d7v3JpMJi8UCwI033khaWhoLFizgxx9/5Nprr2X06NG89tprl7xeEbm0NIdFRK6YFStWnPW+ZcuWALRs2ZINGzZQWFhoO7506VJcXFxo3rw5tWvXJioqipSUlL9UQ4MGDUhMTOTjjz9m6tSpvPvuu3/pfiJyZaiHRUQumeLiYjIyMuza3NzcbBNbP/vsM2JjY+nevTuzZs1i1apVvP/++wAMGTKEZ555hsTERJ599lmysrIYO3YsQ4cOJSgoCIBnn32WBx54gMDAQG688Uby8/NZunQpY8eOrVB9kyZNIiYmhtatW1NcXMy3335rC0wi4twUWETkklm4cCEhISF2bc2bN2fbtm2A9QmeOXPm8NBDDxESEsInn3xCq1atAPDx8eH777/nkUceoXPnzvj4+HDbbbcxZcoU270SExM5ceIE//rXv3jssccICAjg9ttvr3B9Hh4eTJw4kdTUVLy9venRowdz5sy5BJ9cRC43k2EYhqOLEJHqz2QyMW/ePAYMGODoUkSkCtIcFhEREXF6CiwiIiLi9DSHRUSuCI0+i8hfoR4WERERcXoKLCIiIuL0FFhERETE6SmwiIiIiNNTYBERERGnp8AiIiIiTk+BRURERJyeAouIiIg4PQUWERERcXr/D7MQy4FmYos1AAAAAElFTkSuQmCC\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "# Plot loss curves for fine-tuned model\n",
        "plot_loss_curves(history = history_3_fine_tune)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_0ToxaTd3MhB"
      },
      "source": [
        "## Model 4: Pretrained Feature-extraction ConvNetXt-Tiny\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "MaUkkua56LuQ",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 484
        },
        "outputId": "2bd41e0d-68b6-48ce-d7fc-ded16cc842e9"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/convnext/convnext_base_notop.h5\n",
            "\u001b[1m350926856/350926856\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 0us/step\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "\u001b[1mModel: \"functional_8\"\u001b[0m\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\">Model: \"functional_8\"</span>\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
              "┃\u001b[1m \u001b[0m\u001b[1mLayer (type)                   \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mOutput Shape          \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m      Param #\u001b[0m\u001b[1m \u001b[0m┃\n",
              "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
              "│ input_layer (\u001b[38;5;33mInputLayer\u001b[0m)        │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m3\u001b[0m)    │             \u001b[38;5;34m0\u001b[0m │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ cast_4 (\u001b[38;5;33mCast\u001b[0m)                   │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m3\u001b[0m)    │             \u001b[38;5;34m0\u001b[0m │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ rescaling_3 (\u001b[38;5;33mRescaling\u001b[0m)         │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m3\u001b[0m)    │             \u001b[38;5;34m0\u001b[0m │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ convnext_base (\u001b[38;5;33mFunctional\u001b[0m)      │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m7\u001b[0m, \u001b[38;5;34m7\u001b[0m, \u001b[38;5;34m1024\u001b[0m)     │    \u001b[38;5;34m87,566,464\u001b[0m │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ global_average_pooling2d_2      │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m1024\u001b[0m)           │             \u001b[38;5;34m0\u001b[0m │\n",
              "│ (\u001b[38;5;33mGlobalAveragePooling2D\u001b[0m)        │                        │               │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ batch_normalization_4           │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m1024\u001b[0m)           │         \u001b[38;5;34m4,096\u001b[0m │\n",
              "│ (\u001b[38;5;33mBatchNormalization\u001b[0m)            │                        │               │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ dense_3 (\u001b[38;5;33mDense\u001b[0m)                 │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m256\u001b[0m)            │       \u001b[38;5;34m262,400\u001b[0m │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ dropout_3 (\u001b[38;5;33mDropout\u001b[0m)             │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m256\u001b[0m)            │             \u001b[38;5;34m0\u001b[0m │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ output_layer (\u001b[38;5;33mDense\u001b[0m)            │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m101\u001b[0m)            │        \u001b[38;5;34m25,957\u001b[0m │\n",
              "└─────────────────────────────────┴────────────────────────┴───────────────┘\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\">┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
              "┃<span style=\"font-weight: bold\"> Layer (type)                    </span>┃<span style=\"font-weight: bold\"> Output Shape           </span>┃<span style=\"font-weight: bold\">       Param # </span>┃\n",
              "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
              "│ input_layer (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">InputLayer</span>)        │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">224</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">224</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">3</span>)    │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ cast_4 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Cast</span>)                   │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">224</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">224</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">3</span>)    │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ rescaling_3 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Rescaling</span>)         │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">224</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">224</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">3</span>)    │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ convnext_base (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Functional</span>)      │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">7</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">7</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">1024</span>)     │    <span style=\"color: #00af00; text-decoration-color: #00af00\">87,566,464</span> │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ global_average_pooling2d_2      │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">1024</span>)           │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
              "│ (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">GlobalAveragePooling2D</span>)        │                        │               │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ batch_normalization_4           │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">1024</span>)           │         <span style=\"color: #00af00; text-decoration-color: #00af00\">4,096</span> │\n",
              "│ (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">BatchNormalization</span>)            │                        │               │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ dense_3 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)                 │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">256</span>)            │       <span style=\"color: #00af00; text-decoration-color: #00af00\">262,400</span> │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ dropout_3 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dropout</span>)             │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">256</span>)            │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ output_layer (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)            │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">101</span>)            │        <span style=\"color: #00af00; text-decoration-color: #00af00\">25,957</span> │\n",
              "└─────────────────────────────────┴────────────────────────┴───────────────┘\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "\u001b[1m Total params: \u001b[0m\u001b[38;5;34m87,858,917\u001b[0m (335.16 MB)\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Total params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">87,858,917</span> (335.16 MB)\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m290,405\u001b[0m (1.11 MB)\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">290,405</span> (1.11 MB)\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m87,568,512\u001b[0m (334.05 MB)\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Non-trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">87,568,512</span> (334.05 MB)\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        }
      ],
      "source": [
        "import tensorflow as tf\n",
        "import tensorflow_hub as hub\n",
        "\n",
        "input_size = (224, 224, 3)\n",
        "\n",
        "num_classes = ds_info.features['label'].num_classes\n",
        "\n",
        "# Create base model\n",
        "base_model = tf.keras.applications.ConvNeXtBase(include_top = False, input_shape = input_size, weights = 'imagenet')\n",
        "base_model.trainable = False\n",
        "\n",
        "# Input and preprocessing\n",
        "inputs = layers.Input(shape = input_size, name = 'input_layer')\n",
        "x = layers.Rescaling(1./255)(inputs) # Rescaling layer needed because ConvNetXtBase does not have rescaling built-in\n",
        "x = base_model(x, training = False)\n",
        "x = layers.GlobalAveragePooling2D()(x)\n",
        "x = layers.BatchNormalization()(x)\n",
        "x = layers.Dense(256, activation = 'relu')(x)\n",
        "x = layers.Dropout(0.5)(x)\n",
        "outputs = layers.Dense(num_classes, activation = 'softmax', name = 'output_layer')(x)\n",
        "\n",
        "# Define model\n",
        "model_4 = tf.keras.Model(inputs, outputs)\n",
        "\n",
        "# Compile\n",
        "model_4.compile(loss = tf.keras.losses.SparseCategoricalCrossentropy(from_logits = False),\n",
        "                optimizer = tf.keras.optimizers.Adam(),\n",
        "                metrics = ['accuracy'])\n",
        "\n",
        "# Summary\n",
        "model_4.summary()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "VvJepGgn7MKg",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "0ca2549b-acb2-4df7-f495-dee4d95f0f3a"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Saving TensorBoard log files to: ./tensorflow_logs/food_101_all_modelling_experiments/model_4/20250708-134449\n",
            "Epoch 1/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m357s\u001b[0m 139ms/step - accuracy: 0.1523 - loss: 3.9948 - val_accuracy: 0.3350 - val_loss: 2.7943\n",
            "Epoch 2/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m361s\u001b[0m 134ms/step - accuracy: 0.2536 - loss: 3.2363 - val_accuracy: 0.3525 - val_loss: 2.6893\n",
            "Epoch 3/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m320s\u001b[0m 134ms/step - accuracy: 0.2724 - loss: 3.1292 - val_accuracy: 0.3565 - val_loss: 2.6583\n",
            "Epoch 4/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m322s\u001b[0m 134ms/step - accuracy: 0.2792 - loss: 3.0740 - val_accuracy: 0.3620 - val_loss: 2.6502\n",
            "Epoch 5/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m322s\u001b[0m 134ms/step - accuracy: 0.2938 - loss: 3.0169 - val_accuracy: 0.3851 - val_loss: 2.5787\n",
            "Epoch 6/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m322s\u001b[0m 134ms/step - accuracy: 0.2940 - loss: 2.9931 - val_accuracy: 0.3837 - val_loss: 2.5781\n",
            "Epoch 7/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m323s\u001b[0m 135ms/step - accuracy: 0.3027 - loss: 2.9500 - val_accuracy: 0.3835 - val_loss: 2.5543\n",
            "Epoch 8/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m321s\u001b[0m 134ms/step - accuracy: 0.3085 - loss: 2.9233 - val_accuracy: 0.3750 - val_loss: 2.5359\n",
            "Epoch 9/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m321s\u001b[0m 134ms/step - accuracy: 0.3134 - loss: 2.8881 - val_accuracy: 0.3957 - val_loss: 2.5024\n",
            "Epoch 10/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m324s\u001b[0m 135ms/step - accuracy: 0.3165 - loss: 2.8755 - val_accuracy: 0.3859 - val_loss: 2.5149\n"
          ]
        }
      ],
      "source": [
        "# Fit feature-extractor\n",
        "history_4 = model_4.fit(train_data, epochs = initial_epochs, steps_per_epoch = len(train_data), validation_data = test_data, validation_steps = int(0.15 * len(test_data)), callbacks = [create_tensorboard_callback(dir_name = './tensorflow_logs', experiment_name = 'food_101_all_modelling_experiments/model_4'), model_checkpoint])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "-91A7NkB7Uuc",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 927
        },
        "outputId": "72e664c9-658a-4fbb-a812-d63710fee354"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "# Plot loss curves\n",
        "plot_loss_curves(history = history_4)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "6pU76uhWAt2H",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "269cc656-9438-404e-887b-8796086669e4"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\u001b[1m790/790\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m103s\u001b[0m 130ms/step - accuracy: 0.3916 - loss: 2.5191\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[2.527322769165039, 0.38934653997421265]"
            ]
          },
          "metadata": {},
          "execution_count": 39
        }
      ],
      "source": [
        "# Evaluate model\n",
        "model_4.evaluate(test_data)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "B-s_ZGn1AvP-",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "6a646af9-a1fd-42fc-d443-b52acc730ae5"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\u001b[1m790/790\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m109s\u001b[0m 131ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[2.1561e-02, 1.7662e-03, 1.2875e-03, 6.7854e-04, 1.1854e-03,\n",
              "        2.1057e-03, 6.1750e-04, 4.6730e-03, 1.0632e-01, 8.1787e-03,\n",
              "        8.3008e-03, 2.1267e-04, 2.5360e-02, 1.5930e-02, 1.3626e-02,\n",
              "        1.8082e-03, 1.5297e-02, 6.1913e-03, 1.2331e-03, 4.6921e-03,\n",
              "        1.8764e-04, 1.8326e-02, 3.9185e-02, 3.8433e-03, 2.9635e-04,\n",
              "        1.2379e-03, 3.4847e-03, 8.7128e-03, 2.0444e-04, 1.7365e-02,\n",
              "        1.5650e-03, 3.3512e-03, 8.8072e-04, 1.7285e-05, 1.8435e-03,\n",
              "        1.0920e-03, 1.7914e-02, 1.4410e-03, 7.8344e-04, 1.9407e-03,\n",
              "        3.6454e-04, 5.4312e-04, 6.7711e-03, 3.1490e-03, 3.8815e-04,\n",
              "        1.3626e-02, 4.3440e-04, 8.2111e-04, 4.6539e-03, 1.6203e-03,\n",
              "        2.4271e-04, 1.2970e-03, 3.5453e-04, 6.9466e-03, 8.1658e-05,\n",
              "        4.1695e-03, 2.6302e-03, 1.1902e-03, 4.5929e-02, 2.0611e-04,\n",
              "        4.7565e-04, 3.4561e-03, 1.8225e-03, 8.6737e-04, 1.6236e-04,\n",
              "        4.3130e-04, 3.2410e-02, 1.4811e-03, 1.9102e-03, 7.8049e-03,\n",
              "        3.8385e-05, 1.9321e-03, 1.7593e-02, 7.4615e-03, 1.6203e-03,\n",
              "        1.0365e-04, 3.0270e-03, 1.3542e-03, 3.7689e-03, 2.1553e-03,\n",
              "        8.1177e-03, 2.1553e-03, 1.7462e-03, 7.4806e-03, 4.7922e-04,\n",
              "        4.2129e-04, 2.4378e-04, 3.5801e-03, 5.7220e-05, 6.8617e-04,\n",
              "        8.1658e-05, 9.1267e-04, 9.6977e-05, 3.0041e-03, 3.2654e-02,\n",
              "        2.9011e-03, 1.0920e-03, 8.8440e-02, 8.4717e-02, 1.8225e-03,\n",
              "        2.0898e-01]], dtype=float16)"
            ]
          },
          "metadata": {},
          "execution_count": 40
        }
      ],
      "source": [
        "# Make predictions with feature-extractor\n",
        "model_4_preds = model_4.predict(test_data)\n",
        "model_4_preds[:1]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "-VtxV0DTAxrt",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "c31baf7f-7d49-4496-bde3-175323b357e7"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([100,  81,  44,  80,  97,  36,  10,  31,   3,   6])"
            ]
          },
          "metadata": {},
          "execution_count": 41
        }
      ],
      "source": [
        "# Convert preds to labels\n",
        "labels = model_4_preds.argmax(axis = 1)\n",
        "labels[:10]"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "WOBDykOYCqQL"
      },
      "source": [
        "## Fine-tuning model 4"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "vswsJ9QxAzbe",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "3c32c637-2506-4db1-ba70-6fe4b971ff16"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Saving TensorBoard log files to: ./tensorflow_logs/food_101_all_modelling_experiments/model_4/20250708-144321\n",
            "Epoch 11/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m654s\u001b[0m 261ms/step - accuracy: 0.3268 - loss: 2.8137 - val_accuracy: 0.4025 - val_loss: 2.4461\n",
            "Epoch 12/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m654s\u001b[0m 255ms/step - accuracy: 0.3379 - loss: 2.7621 - val_accuracy: 0.4145 - val_loss: 2.4353\n",
            "Epoch 13/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m606s\u001b[0m 255ms/step - accuracy: 0.3475 - loss: 2.7196 - val_accuracy: 0.4142 - val_loss: 2.4025\n",
            "Epoch 14/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m624s\u001b[0m 256ms/step - accuracy: 0.3494 - loss: 2.6939 - val_accuracy: 0.4076 - val_loss: 2.4065\n",
            "Epoch 15/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m619s\u001b[0m 254ms/step - accuracy: 0.3556 - loss: 2.6722 - val_accuracy: 0.4102 - val_loss: 2.3917\n",
            "Epoch 16/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m621s\u001b[0m 254ms/step - accuracy: 0.3573 - loss: 2.6626 - val_accuracy: 0.4206 - val_loss: 2.3924\n",
            "Epoch 17/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m622s\u001b[0m 254ms/step - accuracy: 0.3612 - loss: 2.6455 - val_accuracy: 0.4243 - val_loss: 2.3710\n",
            "Epoch 18/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m606s\u001b[0m 255ms/step - accuracy: 0.3638 - loss: 2.6263 - val_accuracy: 0.4150 - val_loss: 2.3870\n",
            "Epoch 19/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m606s\u001b[0m 255ms/step - accuracy: 0.3681 - loss: 2.5999 - val_accuracy: 0.4253 - val_loss: 2.3625\n",
            "Epoch 20/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m621s\u001b[0m 255ms/step - accuracy: 0.3730 - loss: 2.5918 - val_accuracy: 0.4285 - val_loss: 2.3614\n"
          ]
        }
      ],
      "source": [
        "# Unfreeze last 20 layers of base model\n",
        "base_model.trainable = True\n",
        "for layer in base_model.layers[:-20]:\n",
        "  layer.trainable = False\n",
        "\n",
        "for layer in base_model.layers[-20:]:\n",
        "  if isinstance(layer, layers.BatchNormalization):\n",
        "    layer.trainable = False\n",
        "\n",
        "# Recompile fine-tuned model\n",
        "model_4.compile(loss = tf.keras.losses.SparseCategoricalCrossentropy(from_logits = False),\n",
        "                optimizer = tf.keras.optimizers.Adam(learning_rate = 1e-5),\n",
        "                metrics = ['accuracy'])\n",
        "\n",
        "# Fit fine-tuned model\n",
        "history_4_fine_tune = model_4.fit(train_data, epochs = total_epochs, initial_epoch = initial_epochs, steps_per_epoch = len(train_data), validation_data = test_data, validation_steps = int(0.15 * len(test_data)), callbacks = [create_tensorboard_callback(dir_name = './tensorflow_logs', experiment_name = 'food_101_all_modelling_experiments/model_4_fine-tuned'), model_checkpoint])"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Though fine-tuning model 4 did improve model metrics, it still wasn't enough to beat the Food101 paper accuracy score (~56.75%)."
      ],
      "metadata": {
        "id": "KSa5ba8Cb7Mp"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "pgZ0Zc-ozBCZ",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 927
        },
        "outputId": "37707730-7971-4060-b5e9-9629421d2ecc"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAiwAAAHHCAYAAACcHAM1AAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjAsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvlHJYcgAAAAlwSFlzAAAPYQAAD2EBqD+naQAAXlhJREFUeJzt3Xd4VFX+x/H3ZNJ7QiokQAhJ6D0gBBEFKboILuqirICNFRMUXXQFK5bNWtbFtWBZBVdFLCvlh4DSEghSVZpA6CQICSGQCqkzvz8mDERqQpKZJJ/X89yHmXvPnfkOCc7Hc84912A2m82IiIiI2DEHWxcgIiIicjkKLCIiImL3FFhERETE7imwiIiIiN1TYBERERG7p8AiIiIidk+BRUREROyeAouIiIjYPQUWERERsXsKLCIiImL3FFhEpNbNmjULg8HApk2bbF2KiNRTCiwiIiJi9xRYRERExO4psIiIXfjll18YOnQo3t7eeHp6MmDAANatW1epTWlpKdOmTSMqKgpXV1eaNGlC3759Wbp0qbVNRkYG99xzD2FhYbi4uBAaGsrw4cM5ePBgHX8iEalJjrYuQETk119/5dprr8Xb25snnngCJycn3n//ffr3709ycjK9evUC4PnnnycxMZH777+fnj17kpeXx6ZNm/j555+58cYbARg5ciS//vorEydOpGXLlhw7doylS5eSlpZGy5YtbfgpReRqGMxms9nWRYhIwzZr1izuueceNm7cSI8ePc47fuutt7Jo0SJ27txJq1atADh69CgxMTF07dqV5ORkALp06UJYWBgLFy684Pvk5OTg5+fHa6+9xuTJk2vvA4lIndOQkIjYVHl5OT/88AMjRoywhhWA0NBQ7rrrLlJSUsjLywPA19eXX3/9lT179lzwtdzc3HB2diYpKYmTJ0/WSf0iUjcUWETEprKysjh16hQxMTHnHWvbti0mk4n09HQAXnjhBXJycoiOjqZjx448/vjjbN261drexcWFV155hcWLFxMcHEy/fv149dVXycjIqLPPIyK1Q4FFROqNfv36sW/fPj7++GM6dOjAf/7zH7p168Z//vMfa5tJkyaxe/duEhMTcXV15ZlnnqFt27b88ssvNqxcRK6WAouI2FRgYCDu7u6kpqaed2zXrl04ODgQHh5u3efv788999zDF198QXp6Op06deL555+vdF5kZCR//etf+eGHH9i+fTslJSX885//rO2PIiK1SIFFRGzKaDQyaNAg5s+fX+nS48zMTGbPnk3fvn3x9vYGIDs7u9K5np6etG7dmuLiYgBOnTpFUVFRpTaRkZF4eXlZ24hI/aTLmkWkznz88ccsWbLkvP3PP/88S5cupW/fvjz00EM4Ojry/vvvU1xczKuvvmpt165dO/r370/37t3x9/dn06ZNfPPNNyQkJACwe/duBgwYwB133EG7du1wdHRk7ty5ZGZmMmrUqDr7nCJS83RZs4jUujOXNV9Meno6WVlZTJkyhTVr1mAymejVqxcvv/wyvXv3trZ7+eWXWbBgAbt376a4uJgWLVpw99138/jjj+Pk5ER2djbPPfccy5cvJz09HUdHR9q0acNf//pXbr/99rr4qCJSSxRYRERExO5pDouIiIjYPQUWERERsXsKLCIiImL3FFhERETE7imwiIiIiN1TYBERERG71yAWjjOZTBw5cgQvLy8MBoOtyxEREZErYDabyc/Pp2nTpjg4XLoPpUEEliNHjlS614iIiIjUH+np6YSFhV2yTYMILF5eXoDlA5+554iIiIjYt7y8PMLDw63f45fSIALLmWEgb29vBRYREZF65kqmc2jSrYiIiNg9BRYRERGxewosIiIiYvcaxBwWERGxvfLyckpLS21dhtgZJycnjEbjVb+OAouIiFwVs9lMRkYGOTk5ti5F7JSvry8hISFXtVaaAouIiFyVM2ElKCgId3d3LeApVmazmVOnTnHs2DEAQkNDq/1aCiwiIlJt5eXl1rDSpEkTW5cjdsjNzQ2AY8eOERQUVO3hIU26FRGRajszZ8Xd3d3GlYg9O/P7cTVznBRYRETkqmkYSC6lJn4/FFhERETE7imwiIiIXKWWLVsyffr0K26flJSEwWCo0yur+vfvz6RJk+rs/WqaJt2KiEij1L9/f7p06VKloHExGzduxMPD44rb9+nTh6NHj+Lj43PV791YqIflMrak53Asv8jWZYiISB0zm82UlZVdUdvAwMAqTTx2dna+6nVJGhsFlkvIOVXC+E83Mfhfq1iy/aityxERkRoybtw4kpOTefPNNzEYDBgMBmbNmoXBYGDx4sV0794dFxcXUlJS2LdvH8OHDyc4OBhPT09iY2NZtmxZpdf7/ZCQwWDgP//5D7feeivu7u5ERUWxYMEC6/HfDwnNmjULX19fvv/+e9q2bYunpydDhgzh6NGz3z1lZWU8/PDD+Pr60qRJE/72t78xduxYRowYUa2/g5MnTzJmzBj8/Pxwd3dn6NCh7Nmzx3r80KFDDBs2DD8/Pzw8PGjfvj2LFi2ynjt69GgCAwNxc3MjKiqKmTNnVquOK1WlwJKYmEhsbCxeXl4EBQUxYsQIUlNTL3lO//79rb8M524333yztc24cePOOz5kyJDqfaIalHu6FH8PF06eKuXBz37mr19tIa9Iy06LiFyM2WzmVEmZTTaz2XzFdb755pv07t2bBx54gKNHj3L06FHCw8MBePLJJ/nHP/7Bzp076dSpEwUFBdx0000sX76cX375hSFDhjBs2DDS0tIu+R7Tpk3jjjvuYOvWrdx0002MHj2aEydOXLT9qVOneP311/n0009ZtWoVaWlpTJ482Xr8lVde4fPPP2fmzJmsWbOGvLw85s2bd8Wf+ffGjRvHpk2bWLBgAWvXrsVsNnPTTTdZLz2Oj4+nuLiYVatWsW3bNl555RU8PT0BeOaZZ9ixYweLFy9m586dzJgxg4CAgGrXciWqNIclOTmZ+Ph4YmNjKSsrY+rUqQwaNIgdO3ZcdOzu22+/paSkxPo8Ozubzp07c/vtt1dqN2TIkErpzMXFpSql1YoWTTyYHx/H9GW7eS95H//7+TDr9mfzzzs6c00rLZAkIvJ7p0vLaffs9zZ57x0vDMbd+cq+1nx8fHB2dsbd3Z2QkBAAdu3aBcALL7zAjTfeaG3r7+9P586drc9ffPFF5s6dy4IFC0hISLjoe4wbN44777wTgL///e/8+9//ZsOGDRf9H/LS0lLee+89IiMjAUhISOCFF16wHn/rrbeYMmUKt956KwBvv/22tcejqvbs2cOCBQtYs2YNffr0AeDzzz8nPDycefPmcfvtt5OWlsbIkSPp2LEjAK1atbKen5aWRteuXenRowdg6WGqbVUKLEuWLKn0fNasWQQFBfHTTz/Rr1+/C57j7+9f6fmcOXNwd3c/L7C4uLhYf2nsibOjA08MacMNbYJ47KstpJ04xZ0fruOBa1vx2I3RuDpd/Q2dRETEfpz5Ej6joKCA559/nu+++46jR49SVlbG6dOnL9vD0qlTJ+tjDw8PvL29rUvUX4i7u7s1rIBlGfsz7XNzc8nMzKRnz57W40ajke7du2Mymar0+QB27tyJo6MjvXr1su5r0qQJMTEx7Ny5E4CHH36YCRMm8MMPPzBw4EBGjhxp/UwTJkxg5MiR/PzzzwwaNIgRI0ZYg09tuaqrhHJzc4HzQ8mlfPTRR4waNeq8HpmkpCSCgoLw8/Pjhhtu4KWXXrroMs/FxcUUFxdbn+fl5VWj+qrp0dKfRY9cy8vf7eCLDel8sGo/yalZ/OtPXWjX1LvW319EpD5wczKy44XBNnvvmvD776fJkyezdOlSXn/9dVq3bo2bmxu33XZbpdGDC3Fycqr03GAwXDJcXKh9VYa5atr999/P4MGD+e677/jhhx9ITEzkn//8JxMnTmTo0KEcOnSIRYsWsXTpUgYMGEB8fDyvv/56rdVT7Um3JpOJSZMmERcXR4cOHa7onA0bNrB9+3buv//+SvuHDBnCf//7X5YvX84rr7xCcnIyQ4cOpby8/IKvk5iYiI+Pj3U7M+5Y2zxdHEn8Yyf+M6YHAZ7OpGbmM/ydFGYk7aPcZLtfKhERe2EwGHB3drTJVtUrbpydnS/6PXOuNWvWMG7cOG699VY6duxISEgIBw8erObfUPX4+PgQHBzMxo0brfvKy8v5+eefq/V6bdu2paysjPXr11v3ZWdnk5qaSrt27az7wsPDefDBB/n222/561//yocffmg9FhgYyNixY/nss8+YPn06H3zwQbVquVLV7mGJj49n+/btpKSkXPE5H330ER07dqzUpQUwatQo6+OOHTvSqVMnIiMjSUpKYsCAAee9zpQpU3jsscesz/Py8uostAAMbBfM9837MeXbbfywI5NXluxixa5M/nl7F5o30f00RETqg5YtW7J+/XoOHjyIp6fnRXs/oqKi+Pbbbxk2bBgGg4FnnnmmWsMwV2vixIkkJibSunVr2rRpw1tvvcXJkyerdWl0VFQUw4cP54EHHuD999/Hy8uLJ598kmbNmjF8+HAAJk2axNChQ4mOjubkyZOsXLmStm3bAvDss8/SvXt32rdvT3FxMQsXLrQeqy3V6mFJSEhg4cKFrFy5krCwsCs6p7CwkDlz5nDfffddtm2rVq0ICAhg7969Fzzu4uKCt7d3pa2uNfF04f27u/PqbZ3wdHFk48GTDH1zFV9uTLNpF56IiFyZyZMnYzQaadeuHYGBgRedk/LGG2/g5+dHnz59GDZsGIMHD6Zbt251XC387W9/484772TMmDH07t0bT09PBg8ejKura7Veb+bMmXTv3p0//OEP9O7dG7PZzKJFi6xDU+Xl5cTHx9O2bVuGDBlCdHQ07777LmDpnZoyZQqdOnWiX79+GI1G5syZU2Of9UIM5ip8u5rNZiZOnMjcuXNJSkoiKirqit9o1qxZPPjgg/z222+XvQX54cOHad68OfPmzeOWW2657Gvn5eXh4+NDbm6uTcJL+olT/PWrLWw4aLlcbWDbIBL/2IlAL9tf6SQiUpuKioo4cOAAERER1f7ilOoxmUy0bduWO+64gxdffNHW5VzSxX5PqvL9XaUelvj4eD777DNmz56Nl5cXGRkZZGRkcPr0aWubMWPGMGXKlPPO/eijjxgxYsR5YaWgoIDHH3+cdevWcfDgQZYvX87w4cNp3bo1gwfbZuJWVYX7u/PF+GuYMrQNzkYHlu08xpDpq/jh1wxblyYiIg3EoUOH+PDDD9m9ezfbtm1jwoQJHDhwgLvuusvWpdWJKgWWGTNmkJubS//+/QkNDbVuX375pbVNWlpapZX5AFJTU0lJSbngcJDRaGTr1q3ccsstREdHc99999G9e3dWr15tF2uxXCmjg4G/XBfJ/IQ42oR4kV1YwvhPf+KJb7aQr8XmRETkKjk4ODBr1ixiY2OJi4tj27ZtLFu2jLZt25KWloanp+dFt8tdgl0fVGlIyF7Zekjo94rLynlj6W4+WLUfsxnC/Nx4444u9Iy48su/RUTqAw0J2YeysrJLXrnUsmVLHB1td7/jmhgS0t2aa4GLo5EpQ9tyQ0wQf/16C4dPnuZPH6xlfD/LYnMujlpsTkREao6joyOtW7e2dRm1Sjc/rEW9WjVh8SPXcnv3MMxmeD95P8PfXsOujNpf6E5ERKQhUWCpZV6uTrx2e2c+uLs7TTyc2ZWRzy1vreH9ZC02JyIicqUUWOrIoPYhLJnUj4FtgygpN5G4eBd3friO9BOnbF2aiIiI3VNgqUOBXi58OKYH//hjR9ydjWw4cIKhb67m603pWmxORETkEhRY6pjBYGBUz+YsfuRaerTwo6C4jMe/2cpfPv2J7ILiy7+AiIhII6TAYiMtmnjw5V9688SQGJyMBn7Ykcng6atYvjPT1qWJiMgVaNmyJdOnT7+itgaDgXnz5tVqPQ2dAosNGR0MPNS/NfPi44gO9uR4QQn3fbKJKd9upbC4zNbliYiI2A0FFjvQvqkPCxL68sC1ERgM8MWGdIa+uZpNFfcmEhERaewUWOyEq5ORp25ux+z7r6GZrxtpJ05xx/treXXJLkrK6v425iIiDdkHH3xA06ZNMZkq//d1+PDh3Hvvvezbt4/hw4cTHByMp6cnsbGxLFu2rMbef9u2bdxwww24ubnRpEkTxo8fT0FBgfV4UlISPXv2xMPDA19fX+Li4jh06BAAW7Zs4frrr8fLywtvb2+6d+/Opk2baqw2e6XAYmd6RzZh8aRr+WO3ZpjM8G7SPka8s4bdmfm2Lk1E5PLMZigptM1Whastb7/9drKzs1m5cqV134kTJ1iyZAmjR4+moKCAm266ieXLl/PLL78wZMgQhg0bViP35CksLGTw4MH4+fmxceNGvv76a5YtW0ZCQgJgWWZ/xIgRXHfddWzdupW1a9cyfvx4DAYDAKNHjyYsLIyNGzfy008/8eSTT+Lk5HTVddk7Lc1vh7xdnXjjji7c2DaYqXO3seNoHn94K4UnBsdwb1wEDg4GW5coInJhpafg701t895Tj4CzxxU19fPzY+jQocyePZsBAwYA8M033xAQEMD111+Pg4MDnTt3trZ/8cUXmTt3LgsWLLAGi+qaPXs2RUVF/Pe//8XDw1Lv22+/zbBhw3jllVdwcnIiNzeXP/zhD0RGRgLQtm1b6/lpaWk8/vjjtGnTBoCoqKirqqe+UA+LHRvaMZTvJ/Xj+phASspMvPTdTkb/Zz2/5Zy2dWkiIvXe6NGj+d///kdxsWVJic8//5xRo0bh4OBAQUEBkydPpm3btvj6+uLp6cnOnTtrpIdl586ddO7c2RpWAOLi4jCZTKSmpuLv78+4ceMYPHgww4YN48033+To0aPWto899hj3338/AwcO5B//+Af79u276prqA/Ww2Lkgb1c+HhfLFxvSeXHhDtbuz2bIv1YxbXh7bu3azNpFKCJiF5zcLT0dtnrvKhg2bBhms5nvvvuO2NhYVq9ezb/+9S8AJk+ezNKlS3n99ddp3bo1bm5u3HbbbZSUlNRG5eeZOXMmDz/8MEuWLOHLL7/k6aefZunSpVxzzTU8//zz3HXXXXz33XcsXryY5557jjlz5nDrrbfWSW22osBSDxgMBu7q1Zw+kU149KvN/JKWw2NfbWHpjkxevrUj/h7Oti5RRMTCYLjiYRlbc3V15Y9//COff/45e/fuJSYmhm7dugGwZs0axo0bZw0BBQUFHDx4sEbet23btsyaNYvCwkJrL8uaNWtwcHAgJibG2q5r16507dqVKVOm0Lt3b2bPns0111wDQHR0NNHR0Tz66KPceeedzJw5s8EHFg0J1SMtAzz4+i+9mTwoGkcHA4u3ZzB4+ipWph6zdWkiIvXS6NGj+e677/j4448ZPXq0dX9UVBTffvstmzdvZsuWLdx1113nXVF0Ne/p6urK2LFj2b59OytXrmTixIncfffdBAcHc+DAAaZMmcLatWs5dOgQP/zwA3v27KFt27acPn2ahIQEkpKSOHToEGvWrGHjxo2V5rg0VAos9Yyj0YGEG6KYFx9H6yBPsvKLuWfmRp6au41TJVpsTkSkKm644Qb8/f1JTU3lrrvusu5/44038PPzo0+fPgwbNozBgwdbe1+ulru7O99//z0nTpwgNjaW2267jQEDBvD2229bj+/atYuRI0cSHR3N+PHjiY+P5y9/+QtGo5Hs7GzGjBlDdHQ0d9xxB0OHDmXatGk1Ups9M5gbwF338vLy8PHxITc3F29vb1uXU2eKSst5dUkqH685AEDLJu688acudGvuZ+PKRKSxKCoq4sCBA0RERODq6mrrcsROXez3pCrf3+phqcdcnYw8O6wdn9/fi1AfVw5mn+K2GT/yzx9SKS3XYnMiItJwKLA0AHGtA1gyqR+3drUsNvfWir3c+u4a9h7TYnMiIrXt888/x9PT84Jb+/btbV1eg6GrhBoIHzcn/vWnLgxoG8RTc7ez/bc8bv53Ck8ObcPY3i212JyISC255ZZb6NWr1wWPNYYVaOuKAksD84dOTYlt6c/j32xl1e4spv3fDpbtzOS12zrT1NfN1uWJiDQ4Xl5eeHl52bqMBk9DQg1QsLcrn9wTy4vD2+Pq5MCavdkMnr6Kb38+TAOYYy0iIo2QAksDZTAYuLt3SxY9fC2dw33JLyrjsa+28McZP7Lp4AlblyciDUxNrVEiDVNN/H7osuZGoKzcxHvJ+3hn5T5Ol5YDMLh9MH8b0oZWgZ42rk5E6jOTycSePXswGo0EBgbi7OysW4aIldlspqSkhKysLMrLy4mKisLB4WxfSVW+vxVYGpFjeUX8a9luvtyYjskMjg6WJf8fHhBFgKeLrcsTkXqqpKSEo0ePcurUKVuXInbK3d2d0NBQnJ0r30pGgUUuaU9mPv9YvIvluyxL+nu6OPLgda24r28r3JyNNq5OROojs9lMWVkZ5eXlti5F7IzRaMTR0fGCPW8KLHJF1u7L5u+LdrLtt1wAQrxdeWxQNCO7hWHUZdAiIlLLFFjkiplMZv5v6xFe+z6VwydPA9AmxIspN7XluuhAG1cnIiINmQKLVFlxWTn//fEQb63YQ16R5SaK10YF8OTQNrRv6mPj6kREpCFSYJFqyzlVwtsr9vLftYcoKTdhMMCtXZsxeVCMFp4TEZEapcAiVy39xCle+z6VBVuOAODi6MC9fSOY0D8Sb1ctNS0iIlev1u7WnJiYSGxsLF5eXgQFBTFixAhSU1MveU7//v0xGAznbTfffLO1jdls5tlnnyU0NBQ3NzcGDhzInj17qlKa1LBwf3f+fWdX5sfH0SvCn+IyEzOS9nHdqyuZueYAJWVaJEpEROpOlQJLcnIy8fHxrFu3jqVLl1JaWsqgQYMoLCy86DnffvstR48etW7bt2/HaDRy++23W9u8+uqr/Pvf/+a9995j/fr1eHh4MHjwYIqKiqr/yaRGdA73Zc74a/jPmB5EBnpw8lQp0/5vB4P+lcyibUe11L+IiNSJqxoSysrKIigoiOTkZPr163dF50yfPp1nn32Wo0eP4uHhgdlspmnTpvz1r39l8uTJAOTm5hIcHMysWbMYNWrUZV9TQ0J1o6zcxJeb0vnX0j0cLygGoFtzX6be1JYeLf1tXJ2IiNQ3tTYk9Hu5uZb1O/z9r/zL6qOPPmLUqFF4eHgAcODAATIyMhg4cKC1jY+PD7169WLt2rUXfI3i4mLy8vIqbVL7HI0OjO7VguTH+/PIgCjcnIz8nJbDbe+t5S+fbmJ/VoGtSxQRkQaq2oHFZDIxadIk4uLi6NChwxWds2HDBrZv3879999v3ZeRkQFAcHBwpbbBwcHWY7+XmJiIj4+PdQsPD6/mp5Dq8HBx5NEbo0l+vD939gzHwQDf/5rJoH+t4tn528mu6H0RERGpKdUOLPHx8Wzfvp05c+Zc8TkfffQRHTt2pGfPntV9WwCmTJlCbm6udUtPT7+q15PqCfJ2JfGPnVgyqR8D2gRRZjLz37WHuO61JN5ZuZfTJVqiW0REaka1AktCQgILFy5k5cqVhIWFXdE5hYWFzJkzh/vuu6/S/pCQEAAyMzMr7c/MzLQe+z0XFxe8vb0rbWI70cFefDQultkP9KJjMx8Kist47ftUrn89ia82pVNu0sRcERG5OlUKLGazmYSEBObOncuKFSuIiIi44nO//vpriouL+fOf/1xpf0REBCEhISxfvty6Ly8vj/Xr19O7d++qlCc21icygPnxcbw5qgvNfN3IyCviiW+2cvO/V5O8O8vW5YmISD1WpauEHnroIWbPns38+fOJiYmx7vfx8cHNzbIK6pgxY2jWrBmJiYmVzr322mtp1qzZBYeQXnnlFf7xj3/wySefEBERwTPPPMPWrVvZsWMHrq6ul61LVwnZn6LScv679iBvr9irpf5FROSCqvL97ViVF54xYwZgWQzuXDNnzmTcuHEApKWl4eBQueMmNTWVlJQUfvjhhwu+7hNPPEFhYSHjx48nJyeHvn37smTJkisKK2KfXJ2MjO8XyR09wq1L/a/ec5yUvSla6l9ERKpMS/NLnUg/cYpXv0/l/7TUv4iIVNC9hMRubU7P4e+LdrLhwAkA/D2cefiG1tzVqwXOjle1LJCIiNQzCixi18xmM8t2HuMfi3eyL8tyW4eWTdx5YkgbhnYIwWAw2LhCERGpCwosUi9cbKn/p25uS/cWWupfRKShU2CReqWguIwPVu3nw1X7OV1qWWxuSPsQnhgSQ6tATxtXJyIitUWBReqlzLwi/rV0N19tSsdkBkcHA3f1as4jA6Jo4uli6/JERKSGKbBIvbY7M59/LN7Fil3HAPB0cWRC/0jujYvAzdlo4+pERKSmKLBIg/DjvuP8fdFOtv9muRt3iLcrjw2KZmS3MIwOmpgrIlLfKbBIg2EymVmw5QivfZ/KbzmnAWgT4sWjN0YzqF2wrigSEanHFFikwbnQUv/tm3ozaWA0A9sGKbiIiNRDCizSYOWcKuHD1fuZteYghSWWK4o6hfkwaWAU18couIiI1CcKLNLgnSgs4YNV+/nv2oOcqgguncN9mTQwiv7RgQouIiL1gAKLNBrZBcUVweWQdQ2Xrs19eXRgNNdGBSi4iIjYMQUWaXSy8ot5P3kfn60/RFGpCYDuLfx4dGA0ca2bKLiIiNghBRZptI7lF/Fe0n4+X3+I4jJLcOnZ0p9JN0bRJzLAxtWJiMi5FFik0TuWV8S7SfuYvSGNkorg0ivCn0dvjOaaVk1sXJ2IiIACi63LETuSkVvEu0l7mbMhnZJyS3Dp3aoJj94YTc8I3WBRRMSWFFhEfudIzmneTdrLlxvTKS23/Mr3bR3AozdG6c7QIiI2osAichGHT57inZX7+HpTOmUmy6/+tVEBPHpjNN2a+9m4OhGRxkWBReQy0k+c4p2Ve/n6p8OUVwSX/jGBPDowms7hvrYtTkSkkVBgEblCadmneGvFHr795TdrcLmhTRCPDoymY5iPjasTEWnYFFhEqujg8ULeWrGXub8cpiK3MLBtMJMGRtGhmYKLiEhtUGARqab9WQW8tWIv8zf/Zg0ug9oFM2lgNO2a6ndLRKQmKbCIXKW9xwp4a8UeFmw5wpl/IUM7hPDIwCjahOh3TESkJiiwiNSQPZn5vLl8D99tO2oNLjd3DOWRgVFEB3vZtjgRkXpOgUWkhu3OzOfNZZbgAmAwwB86NeWRAa1pHaTgIiJSHQosIrVk59E83ly2hyW/ZgCW4HJL56Y8PCCKyEBPG1cnIlK/KLCI1LJfj+Ty5rI9/LAjEwAHA4zo0oyJA6KICPCwcXUiIvWDAotIHdn+Wy7Tl+1m2c5jABgdDIzo0oyHB7SmRRMFFxGRS1FgEaljWw/nMH3ZHlbsOhtcRnZrxsQbogj3d7dxdSIi9kmBRcRGNqfn8K+lu0nenQWAo4OB27qHEX99awUXEZHfUWARsbGfDp1k+rLdrN5zHAAno4Hbe4QTf31rmvm62bg6ERH7UJXvb4eqvHBiYiKxsbF4eXkRFBTEiBEjSE1Nvex5OTk5xMfHExoaiouLC9HR0SxatMh6/Pnnn8dgMFTa2rRpU5XSROxK9xZ+fHpfL755sDdxrZtQWm5m9vo0+r+2kqfnbeNo7mlblygiUq84VqVxcnIy8fHxxMbGUlZWxtSpUxk0aBA7duzAw+PCEwxLSkq48cYbCQoK4ptvvqFZs2YcOnQIX1/fSu3at2/PsmXLzhbmWKXSROxSj5b+fH7/NWw4cIJ/Ld3N2v3ZfLYuja82HuaO2DBu7tiU7i38cHas0v87iIg0OlVKBUuWLKn0fNasWQQFBfHTTz/Rr1+/C57z8ccfc+LECX788UecnJwAaNmy5fmFODoSEhJSlXJE6o2eEf58Mf4a1u3P5o2lu9lw4ASfrUvjs3VpeDgb6dM6gOuiA7kuOlBzXURELuCqujFyc3MB8Pf3v2ibBQsW0Lt3b+Lj45k/fz6BgYHcdddd/O1vf8NoNFrb7dmzh6ZNm+Lq6krv3r1JTEykefPmV1OeiN25plUTvhx/DWv3ZfPVpnRW7zlOdmEJS3dksrRiTZdWgR7W8HJNqya4Ohkv86oiIg1ftSfdmkwmbrnlFnJyckhJSblouzZt2nDw4EFGjx7NQw89xN69e3nooYd4+OGHee655wBYvHgxBQUFxMTEcPToUaZNm8Zvv/3G9u3b8fI6f9nz4uJiiouLrc/z8vIIDw/XpFupd0wmM78eySN59zGSd2fxc1oO5aaz/yRdHB3o1aqJNcBEBnpgMBhsWLGISM2pk6uEJkyYwOLFi0lJSSEsLOyi7aKjoykqKuLAgQPWHpU33niD1157jaNHj17wnJycHFq0aMEbb7zBfffdd97x559/nmnTpp23X4FF6rvc06X8uPc4ybuzSN6dxdHcokrHm/m6cV2MJbz0iWyCl6uTjSoVEbl6tR5YEhISmD9/PqtWrSIiIuKSba+77jqcnJwqTahdvHgxN910E8XFxTg7O1/wvNjYWAYOHEhiYuJ5x9TDIo2B2Wxmz7ECklMt4WXDgROUlJusxx0dDPRo6cd10UFcFx1I21Av9b6ISL1SlcBSpTksZrOZiRMnMnfuXJKSki4bVgDi4uKYPXs2JpMJBwfLlRC7d+8mNDT0omGloKCAffv2cffdd1/wuIuLCy4uLlUpXaTeMRgMRAd7ER3sxQP9WnGqpIx1+7OtAeZg9inW7T/Buv0neGXJLgK9XKxDR9dGBeDrfuF/XyIi9VGVelgeeughZs+ezfz584mJibHu9/Hxwc3NshjWmDFjaNasmbVnJD09nfbt2zN27FgmTpzInj17uPfee3n44Yd56qmnAJg8eTLDhg2jRYsWHDlyhOeee47NmzezY8cOAgMDL1uXFo6Txujg8UJW7ckiOTWLH/dlc7q03HrMwQCdw32tAaZTmC9GB/W+iIh9qbUhoYt1N8+cOZNx48YB0L9/f1q2bMmsWbOsx9euXcujjz7K5s2badasGffdd1+lq4RGjRrFqlWryM7OJjAwkL59+/Lyyy8TGRl5RXUpsEhjV1xWzqaDJy1zX1KzSM3Mr3Tc192Ja6Ms4aVfdABBXq42qlRE5CwtzS/SyB3NPc2qiom7q/ccJ7+orNLxdqHe1sm73Vv44WTUwnUiUvcUWETEqqzcxOb0HOuVR1sP51Y67uniSJ/IJtYAE+anhetEpG4osIjIRR0vKCZlj+XS6VW7s8guLKl0PDLQw3LlUUwgvSL8tXCdiNQaBRYRuSJXsnDdNWcWrosJpFWAFq4TkZqjwCIi1XK5hevC/NysVx71aR2Ap4tuUioi1afAIiJX7XIL1zkZDXRv4Ue/6EC6N/ejY5gP7s4KMCJy5RRYRKTGXWjhunM5GCA62IuuzX3pHOZLl+a+RAV5af0XEbkoBRYRqXVnFq5bs/c4W9JzycgrOq+Nu7ORDs186BruS+dwX7qE+xLq46p5MCICKLDYuhyRRikjt4jN6TlsOZzD5rQcth7OobCk/Lx2gV4udKkIL53DfOkU7oO3buIo0igpsIiIzZWbzOzLKmBzeo4lyKTnsCsjv9JVSGdEBnrQJdyPLuE+dA73pU2IN86OWsxOpKFTYBERu3S6pJxfj+SeDTGHc0g/cfq8ds6ODrRv6m3tiekS7ktzf3cNJYk0MAosIlJvHC8oZuvhHDan51p7YnJPl57XztfdyTKZ98xwUrgv/h66I7VIfabAIiL1ltls5mD2KbZU9MJsTs9hx5G8SpdUn9Hc390aXrqE+9K+qbdW5hWpRxRYRKRBKSkzsfNonnVC7+bDOezPKjyvnaODgTahXtYJvV2b+9IqwBMHXVotYpcUWESkwcs9VcrW33Iq9cQcLyg5r52XiyOdwn3oHGbpieka7kuQt6sNKhaR31NgEZFGx2w281vOabak57I5/SRb0nPZ9lsup0vPv7Q61MfVOpTUOcyXTmE+eOg2AyJ1ToFFRAQoKzexO7PAOpl3y+Ecdmfm8/srq8+s0hvXOoDrY4KIjfDDxVFzYURqmwKLiMhFFBSXsf23s1ckbU7POe8mj+7ORmt46R8TSFNfNxtVK9KwKbCIiFRBZl4Rmw6eJCn1GEm7s8jKL650vE2IF/0rwkv3Fn44GbWonUhNUGAREakmk8nMjqN5JKUeY2VqFr+knaw0hOTl4si10QGWABMdqAm8IldBgUVEpIacLCxh1Z4skiruUn2isPKVSB2aeVcMHQXRJdxXd6cWqQIFFhGRWlBuMrP1cA4rU7NISj3G1sO5lY77ujtxXXQg18cE0S86UCvxilyGAouISB3Iyi8mebclvKzanUVeUZn1mMEAncN8uT4miOvbBNKhqY8WsBP5HQUWEZE6VlZu4pf0HFbussx92Xk0r9LxAE9nrou2hJdrWwfi4+5ko0pF7IcCi4iIjWXkFlVM3D3Gmr3ZFBSf7X0xOhjo3tyP/m0sw0dtQrx0J2pplBRYRETsSEmZiU2HTpCUmsXKXcfYc6yg0vEQb1f6xwTSPyaIvlEBeGrVXWkkFFhEROxY+olTJO3OImnXMdbsO05R6dk7UTsZDcS29LfOfYkM9FTvizRYCiwiIvVEUWk56w+cYOWuYySlHuNg9qlKx8P83KzhpXerANycdcsAaTgUWERE6qkDxwuti9at259NSdnZ3hdnRwd6t2rC9RXDRy0DPGxYqcjVU2AREWkATpWUsXZfNitTj7FyVxa/5ZyudDwiwIP+MZaJuz0j/HF1Uu+L1C8KLCIiDYzZbGbvsQJreNl48ARl59wzwM3JSFzrJvSPCeL6NkE00w0bpR5QYBERaeDyi0pZs/e45cqj1GNk5p1/w8Yb2gRxQ5sgujb30y0DxC4psIiINCJms5mdR/Mrel+O8fPvbtjod+aWAW2C6B8dpEXrxG5U5fu7SvdIT0xMJDY2Fi8vL4KCghgxYgSpqamXPS8nJ4f4+HhCQ0NxcXEhOjqaRYsWVWrzzjvv0LJlS1xdXenVqxcbNmyoSmkiIo2WwWCgXVNv4q9vzTcT+vDT0zcy/U9duKVzU7xdHTl5qpR5m4/wyJzNdHtpKXe8t5b3kvexOzOfBvD/rNJIVKmHZciQIYwaNYrY2FjKysqYOnUq27dvZ8eOHXh4XHi2eklJCXFxcQQFBTF16lSaNWvGoUOH8PX1pXPnzgB8+eWXjBkzhvfee49evXoxffp0vv76a1JTUwkKCrpsXephERG5sLJyEz+n5bB8VyYrdx1jd2blRevC/NysQ0fXtGqiibtSp+psSCgrK4ugoCCSk5Pp16/fBdu89957vPbaa+zatQsnpwt3Q/bq1YvY2FjefvttAEwmE+Hh4UycOJEnn3zysnUosIiIXJn0E6dYmXqM5TuPsfZ3l01bJu4GWANMiI+rDSuVxqDOAsvevXuJiopi27ZtdOjQ4YJtbrrpJvz9/XF3d2f+/PkEBgZy11138be//Q2j0UhJSQnu7u588803jBgxwnre2LFjycnJYf78+ee9ZnFxMcXFZyeY5eXlER4ersAiIlIFp0rKWLM3mxW7jrFiV+Z5E3fbhXozoK3lqqPOYb6auCs1riqBpdo3rDCZTEyaNIm4uLiLhhWA/fv3s2LFCkaPHs2iRYvYu3cvDz30EKWlpTz33HMcP36c8vJygoODK50XHBzMrl27LviaiYmJTJs2rbqli4gI4O7syI3tgrmxXTBmcwd2HM1jxc5jrEg9xub0HHYczWPH0TzeWrGXJh7OXBcTyIA2wVwbHYC3qybuSt2qdg/LhAkTWLx4MSkpKYSFhV20XXR0NEVFRRw4cACj0TI2+sYbb/Daa69x9OhRjhw5QrNmzfjxxx/p3bu39bwnnniC5ORk1q9ff95rqodFRKR2ZRcUk5SaxYpdx1i1O4v8c+427ehgud/RDW2CuKFtEK0CPHS/I6mWWu9hSUhIYOHChaxateqSYQUgNDQUJycna1gBaNu2LRkZGZSUlBAQEIDRaCQzM7PSeZmZmYSEhFzwNV1cXHBxcalO6SIicgWaeLowsnsYI7uHUVpuYuNBy/2OVuw6xr6sQtbuz2bt/mxeXrSTFk3crfNeekb44+KoibtS86oUWMxmMxMnTmTu3LkkJSURERFx2XPi4uKYPXs2JpMJBwfLVdS7d+8mNDQUZ2dnALp3787y5cutc1hMJhPLly8nISGhih9HRERqmpPRgT6RAfSJDOCpm9txKLuwYt7LMdbvP8Gh7FPMXHOQmWsO4uFspG+UZeLu9TFBBHlr4q7UjCoNCT300EPMnj2b+fPnExMTY93v4+ODm5tlGegxY8bQrFkzEhMTAUhPT6d9+/aMHTuWiRMnsmfPHu69914efvhhnnrqKcByWfPYsWN5//336dmzJ9OnT+err75i165d581tuRBdJSQiYhsFxWWk7Dlu6X1JPUZWfuWJux2b+Vh7Xzo288FBE3flHLV2ldDFxihnzpzJuHHjAOjfvz8tW7Zk1qxZ1uNr167l0UcfZfPmzTRr1oz77rvPepXQGW+//TavvfYaGRkZdOnShX//+9/06tXriupSYBERsT2TycyvR/Ksa75sOZxb6XiApwvXxwQyoG0QfaMC8XSp9nUf0kBoaX4REbG5Y/lFlom7O4+xek8WhSXl1mNORgO9IppwfZsgBrQJomXAhRcflYZNgUVEROxKcVk5Gw+ctK75cjD7VKXjrQI8rENHPVr64+xYpTvHSD2lwCIiInZtf1aBdeLuhgMnKDvnbo1eLo5cGx3AyG5h3NAmSJdMN2AKLCIiUm/kFZWSsuc4y3ceIyn1GNmFJdZj7UK9SbihNUPah2jCbgOkwCIiIvWSyWRmy+Ecvtt6lNkb0jhVMe+ldZAn8ddHMqxTUxyNGi5qKBRYRESk3jtZWMLMHw8yc80B8ossK+0293fnof6R/LFbmOa5NAAKLCIi0mDkFZXy6dpDfJRygBMVw0WhPq48eF0kf4oNx9VJK+vWVwosIiLS4JwqKWP2+jQ+WLWfYxUL1AV4uvDAtRGMvqaF1nWphxRYRESkwSoqLefrnw7zXtI+fss5DYCvuxP3xkUwtk9LfNx0J+n6QoFFREQavNJyE/N++Y13k/Zx4HghYLkkekyfFtwbF0ETT90k194psIiISKNRbjLz3bajvLNiL6mZ+QC4ORkZ3as5D/RrRbBuwGi3FFhERKTRMZnMLN2Zydsr9rLtN8t9jJwdHbijRxgPXhdJmJ+7jSuU31NgERGRRstsNpO8O4u3V+xl06GTADg6GLi1azMeur41Ebpvkd1QYBERkUbPbDaz/sAJ3l6xl5S9xwFwMMAfOjUl/vrWxIR42bhCUWARERE5x89pJ3lnxV6W7zpm3Te4fTAJ10fRMczHhpU1bgosIiIiF/DrkVzeXbmPRduPcubb77roQCbe0JoeLf1tW1wjpMAiIiJyCXuP5fPuyn3M33KE8oo7RV/Typ+JN0TRJ7KJ7hBdRxRYRERErsCh7ELeS97HNz8dprTc8nXYtbkvE29ozfUxQQoutUyBRUREpAqO5Jzmg1X7+WJDGsVlJgDahXoz8YbWDG4fgoODgkttUGARERGphqz8Yv6Tsp9P1x7iVEk5AK2DPIm/PpJhnZriaNQdomuSAouIiMhVOFlYwswfDzJzzQHyi8oAaO7vzkP9I/ljtzCcHRVcaoICi4iISA3IKyrl07WH+CjlACcKSwBo6uPKX66L5E+x4bg6GW1cYf2mwCIiIlKDTpWUMXt9Gh+s2s+x/GIAAjxdGN8vgtG9WuDh4mjjCusnBRYREZFaUFRazjc/HWZG0j5+yzkNgK+7E/fFRTCmT0t83JxsXGH9osAiIiJSi0rLTcz75TfeTdrHgeOFAHi5ODKmTwvujYugiaeLjSusHxRYRERE6kC5ycx3247yzoq9pGbmA+DmZGR0r+aM79eKIG9XG1do3xRYRERE6pDJZGbZzkzeXrmXrYdzAXB2dOCOHmEMbh+Cn7szPm5O+Lo74eniqAXpKiiwiIiI2IDZbGbVnuO8tXwPmw6dvGAbRweDNbz4ujvj6+aEj7sTvm7OFfvO7vet2O/j7oSXi2ODW8CuKt/fmtYsIiJSQwwGA9dFB9IvKoD1B07wccoB0k6cIudUKTmnSygqNVFmMpNdWEJ2YQlQeMWv7WAAHzcnS2+Nu1NFoDnbc+Pr5oSfx5nnZwOPl6sTxgYQdBRYREREapjBYOCaVk24plWTSvuLSsut4SXnVCk5p0rJrXh88pzHljal5J4q4eSpUk6XlmMyw8mKdlWrBbxdL9Rz44TPOc9/H4S8XR3tamVfBRYREZE64upkJMTHSIhP1SbjFpWWk3faEmIsgaak4nHJOeGmchDKOVVCYUk5ZjPkni4l93Qph7JPVel9vVwdz4YZNydm3dPTZr01CiwiIiJ2ztXJiKuTscpXHZWUmSrCSuWem7NBp6Sil6f07PPCUvKLLbcjyC8qI7+ojPQTp3F3Ntp0aEmBRUREpIFydnQg0MuFQK+qrQtTWm6q1KOTWzH/xpaqNDiVmJhIbGwsXl5eBAUFMWLECFJTUy95zqxZszAYDJU2V9fKCXHcuHHntRkyZEjVP42IiIhcNSejA008XYgM9KR7Cz9uaBPMTR1DbVpTlXpYkpOTiY+PJzY2lrKyMqZOncqgQYPYsWMHHh4eFz3P29u7UrC50PXnQ4YMYebMmdbnLi5aJVBEREQsqhRYlixZUun5rFmzCAoK4qeffqJfv34XPc9gMBASEnLJ13ZxcblsGxEREWmcrup6pdxcy2p+/v7+l2xXUFBAixYtCA8PZ/jw4fz666/ntUlKSiIoKIiYmBgmTJhAdnb2RV+vuLiYvLy8SpuIiIg0XNVe6dZkMnHLLbeQk5NDSkrKRdutXbuWPXv20KlTJ3Jzc3n99ddZtWoVv/76K2FhYQDMmTMHd3d3IiIi2LdvH1OnTsXT05O1a9diNBrPe83nn3+eadOmnbdfK92KiIjUH3WyNP+ECRNYvHgxKSkp1uBxJUpLS2nbti133nknL7744gXb7N+/n8jISJYtW8aAAQPOO15cXExxcbH1eV5eHuHh4QosIiIi9UhVAku1hoQSEhJYuHAhK1eurFJYAXBycqJr167s3bv3om1atWpFQEDARdu4uLjg7e1daRMREZGGq0qBxWw2k5CQwNy5c1mxYgURERFVfsPy8nK2bdtGaOjFL486fPgw2dnZl2wjIiIijUeVAkt8fDyfffYZs2fPxsvLi4yMDDIyMjh9+rS1zZgxY5gyZYr1+QsvvMAPP/zA/v37+fnnn/nzn//MoUOHuP/++wHLhNzHH3+cdevWcfDgQZYvX87w4cNp3bo1gwcPrqGPKSIiIvVZlS5rnjFjBgD9+/evtH/mzJmMGzcOgLS0NBwczuagkydP8sADD5CRkYGfnx/du3fnxx9/pF27dgAYjUa2bt3KJ598Qk5ODk2bNmXQoEG8+OKLWotFREREgKuYdGtPqjJpR0REROxDrU+6FREREalLCiwiIiJi9xRYRERExO4psIiIiIjdU2ARERERu6fAIiIiInZPgUVERETsngKLiIiI2D0FFhEREbF7CiwiIiJi9xRYRERExO4psIiIiIjdU2ARERERu6fAIiIiInZPgUVERETsngKLiIiI2D0FFhEREbF7CiwiIiJi9xRYRERExO4psIiIiIjdU2ARERERu6fAIiIiInZPgUVERETsngKLiIiI2D0FFhEREbF7CiwiIiJi9xRYRERExO4psIiIiIjdU2ARERERu6fAIiIiInZPgUVERETsXpUCS2JiIrGxsXh5eREUFMSIESNITU295DmzZs3CYDBU2lxdXSu1MZvNPPvss4SGhuLm5sbAgQPZs2dP1T+NiIiINEhVCizJycnEx8ezbt06li5dSmlpKYMGDaKwsPCS53l7e3P06FHrdujQoUrHX331Vf7973/z3nvvsX79ejw8PBg8eDBFRUVV/0QiIiLS4DhWpfGSJUsqPZ81axZBQUH89NNP9OvX76LnGQwGQkJCLnjMbDYzffp0nn76aYYPHw7Af//7X4KDg5k3bx6jRo2qSokiIiLSAF3VHJbc3FwA/P39L9muoKCAFi1aEB4ezvDhw/n111+txw4cOEBGRgYDBw607vPx8aFXr16sXbv2asoTERGRBqLagcVkMjFp0iTi4uLo0KHDRdvFxMTw8ccfM3/+fD777DNMJhN9+vTh8OHDAGRkZAAQHBxc6bzg4GDrsd8rLi4mLy+v0iYiIiINV5WGhM4VHx/P9u3bSUlJuWS73r1707t3b+vzPn360LZtW95//31efPHFar13YmIi06ZNq9a5IiIiUv9Uq4clISGBhQsXsnLlSsLCwqp0rpOTE127dmXv3r0A1rktmZmZldplZmZedN7LlClTyM3NtW7p6enV+BQiIiJSX1QpsJjNZhISEpg7dy4rVqwgIiKiym9YXl7Otm3bCA0NBSAiIoKQkBCWL19ubZOXl8f69esr9cycy8XFBW9v70qbiIiINFxVGhKKj49n9uzZzJ8/Hy8vL+scEx8fH9zc3AAYM2YMzZo1IzExEYAXXniBa665htatW5OTk8Nrr73GoUOHuP/++wHLFUSTJk3ipZdeIioqioiICJ555hmaNm3KiBEjavCjioiISH1VpcAyY8YMAPr3719p/8yZMxk3bhwAaWlpODic7bg5efIkDzzwABkZGfj5+dG9e3d+/PFH2rVrZ23zxBNPUFhYyPjx48nJyaFv374sWbLkvAXmREREpHEymM1ms62LuFp5eXn4+PiQm5ur4SEREZF6oirf37qXkIiIiNg9BRYRERGxewosIiIiYvcUWERERMTuKbCIiIiI3VNgEREREbunwCIiIiJ2T4FFRERE7J4Ci4iIiNg9BRYRERGxewosIiIiYvcUWERERMTuKbCIiIiI3VNgEREREbunwCIiIiJ2T4FFRERE7J4Ci4iIiNg9BRYRERGxewosIiIiYvcUWERERMTuKbCIiIiI3VNgEREREbunwCIiIiJ2T4FFRERE7J4Ci4iIiNg9BRYRERGxewosIiIiYvcUWERERMTuKbCIiIiI3VNgEREREbunwCIiIiJ2T4FFRERE7F6VAktiYiKxsbF4eXkRFBTEiBEjSE1NveLz58yZg8FgYMSIEZX2jxs3DoPBUGkbMmRIVUoTERGRBqxKgSU5OZn4+HjWrVvH0qVLKS0tZdCgQRQWFl723IMHDzJ58mSuvfbaCx4fMmQIR48etW5ffPFFVUoTERGRBsyxKo2XLFlS6fmsWbMICgrip59+ol+/fhc9r7y8nNGjRzNt2jRWr15NTk7OeW1cXFwICQmpSjkiIiLSSFzVHJbc3FwA/P39L9nuhRdeICgoiPvuu++ibZKSkggKCiImJoYJEyaQnZ190bbFxcXk5eVV2kRERKThqlIPy7lMJhOTJk0iLi6ODh06XLRdSkoKH330EZs3b75omyFDhvDHP/6RiIgI9u3bx9SpUxk6dChr167FaDSe1z4xMZFp06ZVt3QRERGpZwxms9lcnRMnTJjA4sWLSUlJISws7IJt8vPz6dSpE++++y5Dhw4FLBNsc3JymDdv3kVfe//+/URGRrJs2TIGDBhw3vHi4mKKi4utz/Py8ggPDyc3Nxdvb+/qfBwRERGpY3l5efj4+FzR93e1elgSEhJYuHAhq1atumhYAdi3bx8HDx5k2LBh1n0mk8nyxo6OpKamEhkZed55rVq1IiAggL17914wsLi4uODi4lKd0qtu7bvg3RRCOoJfBDjoSnAREZG6VqXAYjabmThxInPnziUpKYmIiIhLtm/Tpg3btm2rtO/pp58mPz+fN998k/Dw8Aued/jwYbKzswkNDa1KeTWv9DT88DSYyy3PnTwguD2EdLAEmOCOENwOnD1sW6eIiEgDV6XAEh8fz+zZs5k/fz5eXl5kZGQA4OPjg5ubGwBjxoyhWbNmJCYm4urqet78Fl9fXwDr/oKCAqZNm8bIkSMJCQlh3759PPHEE7Ru3ZrBgwdf7ee7OiWnoMtdkLkdju2E0kI4vMGyWRmgSSQEd6gIMp0sj72bgsFgs9JFREQakioFlhkzZgDQv3//SvtnzpzJuHHjAEhLS8OhCsMmRqORrVu38sknn5CTk0PTpk0ZNGgQL774Yt0N+1yMRxMY/rblcXkZZO+1hJeMrZCx3fK4INOyP3sv7Jh39lw3f0uACe5o6Y0J6QABMeDobJOPIiIiUp9Ve9KtPanKpJ0aV3AMMrZVBJltliBzfPfZYaRzOThBYJuKIFMxrBTSEdwvfVm4iIhIQ1SV728FltpQWgRZOy3hxRpmtkNx7oXbezc7Z0ipYm6MfytN8BURkQZNgcUemc2Qk3Y2vGRstTw+efDC7Z08LBN6z+2JCWoHLp51WraIiEhtUWCpT4ryIPPXc4aUtsGxHVBWdIHGBkvPy+/nxng30wRfERGpdxRY6rvyMjix7/y5MQUZF27v5le5Jya4g2WujCb4ioiIHVNgaagKsiBz29krlDK2QVbqJSb4xpwTZDpAaBdw863rqkVERC5IgaUxKSuGrF1ne2HOXHZddIEJvo5u0CcB4iZpLoyIiNicAktjZzZD7uHK82KOboGcQ5bjHkFww9PQ9c/gcP7NJUVEROqCAoucz2yGnf8Hy56DE/st+4LawaAXofVA29YmIiKNUlW+v7XQR2NhMEC7W+Ch9TA4EVx9LVcjfTYSPv0jZO6wdYUiIiIXpcDS2Dg6Q++H4OFf4Jp4y+TcfcvhvThY8DDkZ9q6QhERkfMosDRW7v4w5O8Qvx7a3gJmE/z8CbzVDZJfs9z4UURExE4osDR2TSLhT5/CPUugWXcoKYCVL8HbPWDzF2Ay2bpCERERBRap0KI33LcMRn4EPs0h7zeY9yB82B8OrLZ1dSIi0sgpsMhZDg7Q8TZI2AgDp4GLt+Vy6E/+AF/cCcf32LpCERFppBRY5HxOrtB3kmVibuz9YDBC6iJ49xpY9DgUZtu6QhERaWQUWOTiPALg5n/CQ2sheiiYymDDB/DvrrDmTSi90A0aRUREap4Ci1xeYAzcNQfGLLDcl6g4F5Y+C+/Ewvb/WRalExERqUUKLHLlWl0H45Nh+LvgFQo5afDNvfDRjZC+wdbViYhIA6bAIlXjYISuo2HiT3D9U+DkAYc3WkLLV2PhxAFbVygiIg2QAotUj7MHXPcEPPwzdBsDBgfYMQ/e6QnfPwWnT9q6QhERaUAUWOTqeIXALW/BX1ZDq+uhvATWvm2ZmLvuPSgvtXWFIiLSACiwSM0I6QB3z4XR/4PAtpYeliV/g3d6wc6FmpgrIiJXRYFFao7BAFED4cEU+MN08AiEE/vgy9Ew62b47WdbVygiIvWUAovUPKMj9LjHsvDctZPB0RUOrYEPr4dvx0PuYVtXKCIi9YwCi9QeFy8Y8IzliqJOoyz7tn4Jb3WH5S9Acb5t6xMRkXpDgUVqn08Y/PF9GJ8ELfpCWRGs/qdlYu6mj6G8zNYVioiInVNgkbrTtCuMWwijZoN/JBRmwcJH4b042LNUE3NFROSiFFikbhkM0OZmiF8PQ18FN3/I2gWf3wafjoCM7bauUERE7JACi9iG0Ql6/cUyMbfPRDA6w/4keK8vzE+A/AxbVygiInbEYDbX/374vLw8fHx8yM3Nxdvb29blSHWcOADLp8Gvcy3PnTwg7hHok2BZVbc+KiuBohw4nVPx58nKj80mCOsJzXtZJiiLiDQyVfn+VmAR+5K+Ab6fark/EVhusnjD09D5Tst9jOqaqRyKci8cOE5X/HkmlFiPVewvLbyy9zAYLfN7Iq6Fln0h/Bpw8ayVjyMiYk8UWKR+M5stPS3LnoecQ5Z9IR1h0MuWO0ZX5/VKCi4SMi70+JxgUpQHXOU/EVcfcPUFN19w86t47AdlxZD2I5w8WLm9g6MlwLQ8E2B6KcCISINUa4ElMTGRb7/9ll27duHm5kafPn145ZVXiImJuaLz58yZw5133snw4cOZN2+edb/ZbOa5557jww8/JCcnh7i4OGbMmEFUVNQVva4CSwNVVgzr34dVr0NxrmVf9BCImwSYryxwnD5p6SExXeWl004elpDh5ntO+PA9J4Bc5LGrz+V7hnLS4OAaOJgCB1efDWlnODhC027n9MD0qr/DZCIi56i1wDJkyBBGjRpFbGwsZWVlTJ06le3bt7Njxw48PC79H9CDBw/St29fWrVqhb+/f6XA8sorr5CYmMgnn3xCREQEzzzzDNu2bWPHjh24urpeti4FlgauMBuSX4FNH11d8DC6XFnIqBRM/Cyhw9H56j/HlTp5yLIy8MEUOLAactMqH3dwhGbdK/fAOLvXXX0iIjWkzoaEsrKyCAoKIjk5mX79+l20XXl5Of369ePee+9l9erV5OTkWAOL2WymadOm/PWvf2Xy5MkA5ObmEhwczKxZsxg1atRl61BgaSSO74Xlz0Paessk1d8Hi0s9dvMDJzfb1X41Th6q6H2p6IHJTa983MHJEmDO9MCE9VSAEZF6oSrf345X80a5uZZuen9//0u2e+GFFwgKCuK+++5j9erVlY4dOHCAjIwMBg4caN3n4+NDr169WLt27QUDS3FxMcXFxdbneXl5V/MxpL4IaA1/+szWVdQ9vxaWretoy3ycnENne18Oroa83yB9nWVb9ZolwIT1OKcHpmf9DWsiIhWqHVhMJhOTJk0iLi6ODh06XLRdSkoKH330EZs3b77g8YwMy3obwcHBlfYHBwdbj/1eYmIi06ZNq17hIvWZwQB+LS1b1z9bAszJg2d7Xw6shvwjkLbWsq161bLGTbMelvAScS2ExSrAiEi9U+3AEh8fz/bt20lJSblom/z8fO6++24+/PBDAgICqvtW55kyZQqPPfaY9XleXh7h4eE19voi9YbBAP4Rlq3b3RUB5sA5PTApFQHmR8t2JsCExVoCTMszAebyc8VERGypWoElISGBhQsXsmrVKsLCwi7abt++fRw8eJBhw4ZZ95lMJssbOzqSmppKSEgIAJmZmYSGhlrbZWZm0qVLlwu+rouLCy4uLtUpXaRhMxjAv5Vl6zbGEmBO7D/bA3MwBfKPWib1HlpjmcxsdDkbYCKutfTGKMCIiJ2pUmAxm81MnDiRuXPnkpSURERExCXbt2nThm3btlXa9/TTT5Ofn8+bb75JeHg4Tk5OhISEsHz5cmtAycvLY/369UyYMKFqn0ZEKjMYoEmkZes+9pwAs/psD0xBBhxKsWzJ/7AEmPCe5/TA9ABH/Q+CiNhWlQJLfHw8s2fPZv78+Xh5eVnnmPj4+ODmZhkTHzNmDM2aNSMxMRFXV9fz5rf4+voCVNo/adIkXnrpJaKioqyXNTdt2pQRI0ZcxUcTkfNUCjDjLAEme19F78uZAJN59jmJ4Oha0QNzbUUPTHcFGBGpc1UKLDNmzACgf//+lfbPnDmTcePGAZCWloaDQ9XuqfjEE09QWFjI+PHjycnJoW/fvixZsuSK1mARkatgMFiuvgpoDT3uqQgweyv3wBQeOxtgkrAEmPCeZ69COjfAmExgLrfc0sD8+8emise/P17x5wWPV/Uc8zmPzz1uusC+M69T8VoGI7Tqb/lsBoMNfygiciFaml9ELs5shuN7KvfAFGZVbmOo+B8Us6nu66sNfhHQeRR0usMyF0hEao3uJSQitcNshuO7K/fAnDpehRcwWAKOg9HSo3HmT4Phd/scKh5X/FnpHIdzjv/+nHPaWR87/O41zz1+zr5TJ2D395VvWhneCzr9CdrfCu6XXm9KRKpOgUVE6obZDPkV6yVdNnwY7X+opaQQdi6ErXNgf9LZXiOjM0QPhk6jIGpQ3d6qQaQBU2AREblaeUdh29ew9UvI3H52v5sfdBhpCS9hPew/hInYMQUWEZGalLHd0uuy9WvLZeBn+Edahow63WFZvE+uTk4a7Fth2dI3QmAMxN5vuUu78aruJCN2SoFFRKQ2mMotQ0Vbv4Sd/welp84ea967Yr7LCEsvjFxecb5lHtSZkJK998LtvMMsV7F1GwuegXVbo9QqBRYRkdpWXAC7FsKWL2B/MlDxn1Kjs6VHoPOd0Hqg5rucy1QORzdXBJSVkL4eTGVnjxuMlsvKI2+w/LlvJfz8Xzh9wnLc6AztRkDPByxrA2k4rt5TYBERqUt5RyzzXbbMgWM7zu5387fMd+k8yrJeTWP8gs1Jh/0rLSFlfxKcPln5uH8rS0CJvMGyro+rT+XjpUXw61zY+B/4bdPZ/SGdLMNFHW8HZ/da/xhSOxRYRERswWyGjG2WIaNtX1tWDT6jSWvLRN1Od4BfC9vVWNuKCyz3qTozzHN8d+XjLj7Qqp8loLS6vmpzf478Ahv+A9u/gbIiyz5XH+jyZ4i9z7KCs9QrCiwiIrZWXgYHkmBLxXyXstNnjzXvA53/ZBnecPO1UYE1xGSCjC1nh3nS1oGp9Oxxg9FyNdWZXpSm3a5+Au2pE/DLZ7DpIzh58Oz+yAGW4aKoQZbL6MXuKbCIiNiT4nxLaNkyBw6s4ux8FxeIGWoZMmo9EIxONi3ziuX+VjHEs9ISUs7MMTnDtwW0HlAxzHNt7YUykwn2LoONH8KepVj/Xn2aV0zSHQMeAbXz3lIjFFhEROxV7m+w7StLz0vWzrP73QMq5rv8ydILYU/zXUoK4eC5wzyplY87e0Gr6yDyektIscUtDU4cgE0fwy+fnp0nY3SxrFLc84HGO4fIzimwiIjYO7MZMrZagsu2ry03mTyjSZQluHT6E/g2r/vaTCZLbdY1UdZDecnZ4wYHSwA4M8zTrLv99A6Vnobt31p6XY78cnZ/aBdLcOkwEpzcbFaeVKbAIiJSn5SXWYZXtsyBXd9Vnu/Som/FfJfh519BU5PyjliGd84M9ZzKrnzcpzm0rggoEf3qx1ozh3+yBJft30J5sWWfmx90/TP0uFc3t7QDCiwiIvVVUV7FfJcvLIuqnZmX4egKMTdZ5rtE3nD1PRolp+DQj2d7Uc4dngJw9rQEkzO9KP6t6u+QSmE2/PJf2Pgx5KZV7DRY5g31fMDypybp2oQCi4hIQ5B7GLZ+ZblMOmvX2f3uAdDxNsuQUdOuVxYkTCbLPZHOBJS0tZWHeTBAs25nA0pYrP0M89QUU7llcu7GDy2Tdc/wbWG5LLrr3bordx1TYBERaUjMZssKsVu+tKxBUph19lhAjGXIqOMd4Bte+bz8jMrDPOeeB5Yl763DPNc1ri/r7H1nJ+kW5Vr2Obpa5rjE3m8Jb1LrFFhERBqq8lJLCNl6Zr5LxQJqGCwrxbYbblmbZN9KOPZr5XOdPCDi2rO9KE1a199hnppScsoSAjd8aJlofEaz7pbg0v6P4ORqu/oaOAUWEZHGoCgXdiywDBkdXH2BBgZo2uWcYZ6eurfRxZjNcHiTZbjo17lnh8vc/KHb3ZZJun4tbVpiQ6TAIiLS2OSkWea77FthWe4+8gaI6A8eTWxdWf1TkGWZpLtpJuSmV+w0QPRgiH3A8nfr4GDTEhsKBRYREZGrZSqH3Ussw0X7V57d7xdhmaTbZXTjmvdTCxRYREREatLxvZY7Rm+eDcVnJum6QceRll6Xpl1sWl59pcAiIiJSG0oKLSsTb/gPZG47uz8s1hJc2o8ARxeblVffKLCIiIjUJrPZcsuCDR/Cjvln71DtHmC56WKPe2xzW4V6RoFFRESkruRnws//hZ9mQt5vln0GB4geYrk0unlvyyJ8Do66jPx3FFhERETqWnkZ7F5s6XU5kHzhNgZjRXhxAqOjJcRYH1eEGuPv/qz0+Pdtz+y7mra/O36ptj5hNfpXVpXvb8cafWcREZHGyugIbYdZtqzdlkm6W76A4ryzbczlUFYOFF30ZeyWgxM8e9xmb6/AIiIiUtMCo+GmV2FIIpSesqxQbCqzbGceW/eVWnpnrI+vpO3v99d023NrKbc8tvG9pRRYREREaouDEVy8bF1Fg6Cl+kRERMTuKbCIiIiI3VNgEREREbtXpcCSmJhIbGwsXl5eBAUFMWLECFJTUy95zrfffkuPHj3w9fXFw8ODLl268Omnn1ZqM27cOAwGQ6VtyJAhVf80IiIi0iBVadJtcnIy8fHxxMbGUlZWxtSpUxk0aBA7duzAw8Pjguf4+/vz1FNP0aZNG5ydnVm4cCH33HMPQUFBDB482NpuyJAhzJw50/rcxUVLG4uIiIjFVS0cl5WVRVBQEMnJyfTr1++Kz+vWrRs333wzL774ImDpYcnJyWHevHnVqkMLx4mIiNQ/Vfn+vqo5LLm5ljtW+vtf2e21zWYzy5cvJzU19byAk5SURFBQEDExMUyYMIHs7OyrKU1EREQakGr3sJhMJm655RZycnJISUm5ZNvc3FyaNWtGcXExRqORd999l3vvvdd6fM6cObi7uxMREcG+ffuYOnUqnp6erF27FqPReN7rFRcXU1xcbH2el5dHeHi4elhERETqkTpZmj8+Pp7t27dfNqwAeHl5sXnzZgoKCli+fDmPPfYYrVq1on///gCMGjXK2rZjx4506tSJyMhIkpKSGDBgwHmvl5iYyLRp06pbuoiIiNQz1ephSUhIYP78+axatYqIiIgqv+n9999Peno633///UXbBAYG8tJLL/GXv/zlvGPqYREREan/aq2HxWw2M3HiRObOnUtSUlK1wgpYhpPODRy/d/jwYbKzswkNDb3gcRcXF11FJCIi0ohUKbDEx8cze/Zs5s+fj5eXFxkZGQD4+Pjg5uYGwJgxY2jWrBmJiYmAZfimR48eREZGUlxczKJFi/j000+ZMWMGAAUFBUybNo2RI0cSEhLCvn37eOKJJ2jdunWly55FRESk8apSYDkTMs7MPTlj5syZjBs3DoC0tDQcHM5efFRYWMhDDz3E4cOHcXNzo02bNnz22Wf86U9/AsBoNLJ161Y++eQTcnJyaNq0KYMGDeLFF19UL4qIiIgAV7kOi73QOiwiIiL1T51cJWRPzmSuvLw8G1ciIiIiV+rM9/aV9J00iMCSn58PQHh4uI0rERERkarKz8/Hx8fnkm0axJCQyWTiyJEjeHl5YTAYavS1z1wynZ6eruEmO6Cfh33Rz8P+6GdiX/TzuDSz2Ux+fj5NmzatNP/1QhpED4uDgwNhYWG1+h7e3t76ZbMj+nnYF/087I9+JvZFP4+Lu1zPyhlXdS8hERERkbqgwCIiIiJ2T4HlMlxcXHjuuee0Joyd0M/DvujnYX/0M7Ev+nnUnAYx6VZEREQaNvWwiIiIiN1TYBERERG7p8AiIiIidk+BRUREROyeAstlvPPOO7Rs2RJXV1d69erFhg0bbF1So5SYmEhsbCxeXl4EBQUxYsQIUlNTbV2WVPjHP/6BwWBg0qRJti6l0frtt9/485//TJMmTXBzc6Njx45s2rTJ1mU1SuXl5TzzzDNERETg5uZGZGQkL7744hXdL0cuToHlEr788ksee+wxnnvuOX7++Wc6d+7M4MGDOXbsmK1La3SSk5OJj49n3bp1LF26lNLSUgYNGkRhYaGtS2v0Nm7cyPvvv0+nTp1sXUqjdfLkSeLi4nBycmLx4sXs2LGDf/7zn/j5+dm6tEbplVdeYcaMGbz99tvs3LmTV155hVdffZW33nrL1qXVa7qs+RJ69epFbGwsb7/9NmC5Z1F4eDgTJ07kySeftHF1jVtWVhZBQUEkJyfTr18/W5fTaBUUFNCtWzfeffddXnrpJbp06cL06dNtXVaj8+STT7JmzRpWr15t61IE+MMf/kBwcDAfffSRdd/IkSNxc3Pjs88+s2Fl9Zt6WC6ipKSEn376iYEDB1r3OTg4MHDgQNauXWvDygQgNzcXAH9/fxtX0rjFx8dz8803V/p3InVvwYIF9OjRg9tvv52goCC6du3Khx9+aOuyGq0+ffqwfPlydu/eDcCWLVtISUlh6NChNq6sfmsQNz+sDcePH6e8vJzg4OBK+4ODg9m1a5eNqhKw9HRNmjSJuLg4OnToYOtyGq05c+bw888/s3HjRluX0ujt37+fGTNm8NhjjzF16lQ2btzIww8/jLOzM2PHjrV1eY3Ok08+SV5eHm3atMFoNFJeXs7LL7/M6NGjbV1avabAIvVOfHw827dvJyUlxdalNFrp6ek88sgjLF26FFdXV1uX0+iZTCZ69OjB3//+dwC6du3K9u3bee+99xRYbOCrr77i888/Z/bs2bRv357NmzczadIkmjZtqp/HVVBguYiAgACMRiOZmZmV9mdmZhISEmKjqiQhIYGFCxeyatUqwsLCbF1Oo/XTTz9x7NgxunXrZt1XXl7OqlWrePvttykuLsZoNNqwwsYlNDSUdu3aVdrXtm1b/ve//9moosbt8ccf58knn2TUqFEAdOzYkUOHDpGYmKjAchU0h+UinJ2d6d69O8uXL7fuM5lMLF++nN69e9uwssbJbDaTkJDA3LlzWbFiBREREbYuqVEbMGAA27ZtY/PmzdatR48ejB49ms2bNyus1LG4uLjzLvPfvXs3LVq0sFFFjdupU6dwcKj89Wo0GjGZTDaqqGFQD8slPPbYY4wdO5YePXrQs2dPpk+fTmFhIffcc4+tS2t04uPjmT17NvPnz8fLy4uMjAwAfHx8cHNzs3F1jY+Xl9d584c8PDxo0qSJ5hXZwKOPPkqfPn34+9//zh133MGGDRv44IMP+OCDD2xdWqM0bNgwXn75ZZo3b0779u355ZdfeOONN7j33nttXVr9ZpZLeuutt8zNmzc3Ozs7m3v27Glet26drUtqlIALbjNnzrR1aVLhuuuuMz/yyCO2LqPR+r//+z9zhw4dzC4uLuY2bdqYP/jgA1uX1Gjl5eWZH3nkEXPz5s3Nrq6u5latWpmfeuopc3Fxsa1Lq9e0DouIiIjYPc1hEREREbunwCIiIiJ2T4FFRERE7J4Ci4iIiNg9BRYRERGxewosIiIiYvcUWERERMTuKbCISINhMBiYN2+ercsQkVqgwCIiNWLcuHEYDIbztiFDhti6NBFpAHQvIRGpMUOGDGHmzJmV9rm4uNioGhFpSNTDIiI1xsXFhZCQkEqbn58fYBmumTFjBkOHDsXNzY1WrVrxzTffVDp/27Zt3HDDDbi5udGkSRPGjx9PQUFBpTYff/wx7du3x8XFhdDQUBISEiodP378OLfeeivu7u5ERUWxYMEC67GTJ08yevRoAgMDcXNzIyoq6ryAJSL2SYFFROrMM888w8iRI9myZQujR49m1KhR7Ny5E4DCwkIGDx6Mn58fGzdu5Ouvv2bZsmWVAsmMGTOIj49n/PjxbNu2jQULFtC6detK7zFt2jTuuOMOtm7dyk033cTo0aM5ceKE9f137NjB4sWL2blzJzNmzCAgIKDu/gJEpPpsffdFEWkYxo4dazYajWYPD49K28svv2w2my133H7wwQcrndOrVy/zhAkTzGaz2fzBBx+Y/fz8zAUFBdbj3333ndnBwcGckZFhNpvN5qZNm5qfeuqpi9YAmJ9++mnr84KCAjNgXrx4sdlsNpuHDRtmvueee2rmA4tIndIcFhGpMddffz0zZsyotM/f39/6uHfv3pWO9e7dm82bNwOwc+dOOnfujIeHh/V4XFwcJpOJ1NRUDAYDR44cYcCAAZesoVOnTtbHHh4eeHt7c+zYMQAmTJjAyJEj+fnnnxk0aBAjRoygT58+1fqsIlK3FFhEpMZ4eHicN0RTU9zc3K6onZOTU6XnBoMBk8kEwNChQzl06BCLFi1i6dKlDBgwgPj4eF5//fUar1dEapbmsIhInVm3bt15z9u2bQtA27Zt2bJlC4WFhdbja9aswcHBgZiYGLy8vGjZsiXLly+/qhoCAwMZO3Ysn332GdOnT+eDDz64qtcTkbqhHhYRqTHFxcVkZGRU2ufo6Gid2Pr111/To0cP+vbty+eff86GDRv46KOPABg9ejTPPfccY8eO5fnnnycrK4uJEydy9913ExwcDMDzzz/Pgw8+SFBQEEOHDiU/P581a9YwceLEK6rv2WefpXv37rRv357i4mIWLlxoDUwiYt8UWESkxixZsoTQ0NBK+2JiYti1axdguYJnzpw5PPTQQ4SGhvLFF1/Qrl07ANzd3fn+++955JFHiI2Nxd3dnZEjR/LGG29YX2vs2LEUFRXxr3/9i8mTJxMQEMBtt912xfU5OzszZcoUDh48iJubG9deey1z5sypgU8uIrXNYDabzbYuQkQaPoPBwNy5cxkxYoStSxGRekhzWERERMTuKbCIiIiI3dMcFhGpExp9FpGroR4WERERsXsKLCIiImL3FFhERETE7imwiIiIiN1TYBERERG7p8AiIiIidk+BRUREROyeAouIiIjYPQUWERERsXv/D5+6FmoYVpghAAAAAElFTkSuQmCC\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "# Plot loss curves for fine-tuned model\n",
        "plot_loss_curves(history = history_4_fine_tune)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "y6Dl-XFs5XwG"
      },
      "source": [
        "## Model 5: Feature-extraction (EfficientNetB7)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "stGHUGdIzDeI",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 385
        },
        "outputId": "a54248cd-574f-4150-d8b5-bc7e73b01043"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "\u001b[1mModel: \"functional_11\"\u001b[0m\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\">Model: \"functional_11\"</span>\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
              "┃\u001b[1m \u001b[0m\u001b[1mLayer (type)                   \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mOutput Shape          \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m      Param #\u001b[0m\u001b[1m \u001b[0m┃\n",
              "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
              "│ input_layer (\u001b[38;5;33mInputLayer\u001b[0m)        │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m3\u001b[0m)    │             \u001b[38;5;34m0\u001b[0m │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ efficientnetb7 (\u001b[38;5;33mFunctional\u001b[0m)     │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m7\u001b[0m, \u001b[38;5;34m7\u001b[0m, \u001b[38;5;34m2560\u001b[0m)     │    \u001b[38;5;34m64,097,687\u001b[0m │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ global_average_pooling2d_5      │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m2560\u001b[0m)           │             \u001b[38;5;34m0\u001b[0m │\n",
              "│ (\u001b[38;5;33mGlobalAveragePooling2D\u001b[0m)        │                        │               │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ batch_normalization_7           │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m2560\u001b[0m)           │        \u001b[38;5;34m10,240\u001b[0m │\n",
              "│ (\u001b[38;5;33mBatchNormalization\u001b[0m)            │                        │               │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ dense_6 (\u001b[38;5;33mDense\u001b[0m)                 │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m256\u001b[0m)            │       \u001b[38;5;34m655,616\u001b[0m │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ dropout_6 (\u001b[38;5;33mDropout\u001b[0m)             │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m256\u001b[0m)            │             \u001b[38;5;34m0\u001b[0m │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ output_layer (\u001b[38;5;33mDense\u001b[0m)            │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m101\u001b[0m)            │        \u001b[38;5;34m25,957\u001b[0m │\n",
              "└─────────────────────────────────┴────────────────────────┴───────────────┘\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\">┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
              "┃<span style=\"font-weight: bold\"> Layer (type)                    </span>┃<span style=\"font-weight: bold\"> Output Shape           </span>┃<span style=\"font-weight: bold\">       Param # </span>┃\n",
              "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
              "│ input_layer (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">InputLayer</span>)        │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">224</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">224</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">3</span>)    │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ efficientnetb7 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Functional</span>)     │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">7</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">7</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">2560</span>)     │    <span style=\"color: #00af00; text-decoration-color: #00af00\">64,097,687</span> │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ global_average_pooling2d_5      │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">2560</span>)           │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
              "│ (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">GlobalAveragePooling2D</span>)        │                        │               │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ batch_normalization_7           │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">2560</span>)           │        <span style=\"color: #00af00; text-decoration-color: #00af00\">10,240</span> │\n",
              "│ (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">BatchNormalization</span>)            │                        │               │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ dense_6 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)                 │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">256</span>)            │       <span style=\"color: #00af00; text-decoration-color: #00af00\">655,616</span> │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ dropout_6 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dropout</span>)             │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">256</span>)            │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
              "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
              "│ output_layer (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)            │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">101</span>)            │        <span style=\"color: #00af00; text-decoration-color: #00af00\">25,957</span> │\n",
              "└─────────────────────────────────┴────────────────────────┴───────────────┘\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "\u001b[1m Total params: \u001b[0m\u001b[38;5;34m64,789,500\u001b[0m (247.15 MB)\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Total params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">64,789,500</span> (247.15 MB)\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m686,693\u001b[0m (2.62 MB)\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">686,693</span> (2.62 MB)\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m64,102,807\u001b[0m (244.53 MB)\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Non-trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">64,102,807</span> (244.53 MB)\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        }
      ],
      "source": [
        "# Define input size\n",
        "input_size = (224, 224, 3)\n",
        "\n",
        "num_classes = ds_info.features['label'].num_classes\n",
        "\n",
        "# Create base model\n",
        "base_model = tf.keras.applications.EfficientNetB7(include_top = False, input_shape = input_size, weights = 'imagenet')\n",
        "base_model.trainable = False\n",
        "\n",
        "# Input and preprocessing\n",
        "inputs = layers.Input(shape = input_size, name = 'input_layer')\n",
        "x = base_model(inputs, training = False)\n",
        "x = layers.GlobalAveragePooling2D()(x)\n",
        "x = layers.BatchNormalization()(x)\n",
        "x = layers.Dense(256, activation = 'relu')(x)\n",
        "x = layers.Dropout(0.5)(x)\n",
        "outputs = layers.Dense(num_classes, activation = 'softmax', name = 'output_layer')(x)\n",
        "model_5 = tf.keras.Model(inputs, outputs)\n",
        "\n",
        "# Compile\n",
        "model_5.compile(loss = tf.keras.losses.SparseCategoricalCrossentropy(from_logits = False),\n",
        "                optimizer = tf.keras.optimizers.Adam(),\n",
        "                metrics = ['accuracy'])\n",
        "\n",
        "# Summary\n",
        "model_5.summary()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "GzibWOJS89Dg",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "73cc98df-9efe-4e86-9b87-806d1a151262"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Saving TensorBoard log files to: ./tensorflow_logs/food_101_all_modelling_experiments/model_5/20250708-170400\n",
            "Epoch 1/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m443s\u001b[0m 154ms/step - accuracy: 0.2722 - loss: 3.4361 - val_accuracy: 0.5467 - val_loss: 1.7473\n",
            "Epoch 2/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m321s\u001b[0m 134ms/step - accuracy: 0.4042 - loss: 2.4719 - val_accuracy: 0.5593 - val_loss: 1.6861\n",
            "Epoch 3/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m321s\u001b[0m 134ms/step - accuracy: 0.4250 - loss: 2.3635 - val_accuracy: 0.5771 - val_loss: 1.6401\n",
            "Epoch 4/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m323s\u001b[0m 135ms/step - accuracy: 0.4439 - loss: 2.2745 - val_accuracy: 0.5953 - val_loss: 1.5505\n",
            "Epoch 5/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m382s\u001b[0m 135ms/step - accuracy: 0.4542 - loss: 2.2085 - val_accuracy: 0.6013 - val_loss: 1.5169\n",
            "Epoch 6/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m380s\u001b[0m 134ms/step - accuracy: 0.4684 - loss: 2.1481 - val_accuracy: 0.6135 - val_loss: 1.5016\n",
            "Epoch 7/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m322s\u001b[0m 135ms/step - accuracy: 0.4807 - loss: 2.0936 - val_accuracy: 0.6187 - val_loss: 1.4546\n",
            "Epoch 8/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m323s\u001b[0m 135ms/step - accuracy: 0.4880 - loss: 2.0544 - val_accuracy: 0.6325 - val_loss: 1.4126\n",
            "Epoch 9/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m381s\u001b[0m 135ms/step - accuracy: 0.4910 - loss: 2.0194 - val_accuracy: 0.6270 - val_loss: 1.4056\n",
            "Epoch 10/10\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m382s\u001b[0m 135ms/step - accuracy: 0.5057 - loss: 1.9763 - val_accuracy: 0.6183 - val_loss: 1.4192\n"
          ]
        }
      ],
      "source": [
        "# Fit\n",
        "history_5 = model_5.fit(train_data, epochs = initial_epochs, steps_per_epoch = len(train_data), validation_data = test_data, validation_steps = int(0.1 * len(test_data)), callbacks = [create_tensorboard_callback(dir_name = './tensorflow_logs', experiment_name = 'food_101_all_modelling_experiments/model_5'), model_checkpoint])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "gNUyuuPHIemU",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 927
        },
        "outputId": "d17cadce-fa99-417b-bb69-af987e874100"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "# Plot loss curves\n",
        "plot_loss_curves(history = history_5)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "MhLHU6Vq9Ab1",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "6c4e1a40-2476-42cc-bec8-d9e4d12815db"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\u001b[1m790/790\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m110s\u001b[0m 139ms/step - accuracy: 0.6320 - loss: 1.3923\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[1.4002346992492676, 0.6305742859840393]"
            ]
          },
          "metadata": {},
          "execution_count": 50
        }
      ],
      "source": [
        "# Evaluate model\n",
        "model_5.evaluate(test_data)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "sV7bd9w89FHy",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "fea45ec8-25c3-4451-e4ed-bb07875e31d8"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\u001b[1m790/790\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m132s\u001b[0m 146ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[6.9695e-03, 8.2731e-05, 1.4172e-03, 9.4175e-05, 2.1343e-03,\n",
              "        1.6241e-03, 3.4714e-04, 1.5271e-04, 2.8046e-02, 4.4918e-04,\n",
              "        8.5211e-04, 2.1935e-05, 6.7711e-03, 4.3793e-03, 1.1536e-02,\n",
              "        1.5078e-03, 8.7891e-02, 5.7487e-03, 9.5367e-07, 1.3328e-04,\n",
              "        1.2398e-05, 9.4681e-03, 3.0762e-02, 7.9880e-03, 6.8545e-05,\n",
              "        4.3035e-04, 3.9339e-04, 6.1150e-03, 2.8610e-03, 2.7734e-01,\n",
              "        1.2732e-01, 2.0447e-03, 3.3116e-04, 1.6689e-06, 3.0346e-03,\n",
              "        1.1718e-04, 2.5797e-04, 3.1114e-04, 2.5630e-05, 7.6103e-04,\n",
              "        1.8787e-04, 3.0935e-05, 2.5539e-03, 1.4293e-04, 1.5199e-05,\n",
              "        1.4557e-02, 1.0836e-04, 5.2118e-04, 9.2328e-05, 3.0160e-04,\n",
              "        1.9848e-04, 1.0848e-05, 2.9111e-04, 1.3649e-04, 2.3842e-07,\n",
              "        8.4209e-04, 5.6362e-04, 2.4643e-03, 3.4515e-02, 1.1313e-04,\n",
              "        8.6010e-05, 6.2943e-05, 7.3767e-04, 2.9392e-03, 2.6226e-06,\n",
              "        3.5048e-05, 4.2844e-04, 3.0994e-04, 4.7612e-04, 1.7047e-04,\n",
              "        3.9339e-06, 1.0729e-06, 1.0818e-02, 7.9346e-02, 6.5088e-04,\n",
              "        4.7684e-07, 6.2370e-04, 2.2471e-05, 1.6332e-04, 5.1260e-06,\n",
              "        9.0837e-05, 7.9870e-06, 6.5327e-04, 4.3060e-02, 4.4286e-05,\n",
              "        8.2135e-05, 2.4986e-04, 1.9360e-03, 1.9670e-06, 6.5446e-05,\n",
              "        5.1379e-05, 2.6608e-04, 1.5676e-05, 4.5717e-05, 9.8755e-02,\n",
              "        5.9052e-03, 2.4319e-04, 1.0118e-03, 1.5495e-02, 3.2043e-03,\n",
              "        4.4586e-02]], dtype=float16)"
            ]
          },
          "metadata": {},
          "execution_count": 51
        }
      ],
      "source": [
        "# Make predictions\n",
        "model_5_preds = model_5.predict(test_data)\n",
        "model_5_preds[:1]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "7q0pXhrh9HJP",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "b0fb7d12-ab0f-4c87-f844-c1e486526232"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([29, 81, 57, 53, 97, 97, 67, 31,  3, 94])"
            ]
          },
          "metadata": {},
          "execution_count": 52
        }
      ],
      "source": [
        "# Convert predictions to labels\n",
        "labels = model_5_preds.argmax(axis = 1)\n",
        "labels[:10]"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Eb_tIK469K1q"
      },
      "source": [
        "## Fine-tuning model 5"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "2m9EZAER9KgB",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "c813b3b0-0888-4de3-fdef-5b987bdd5461"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Saving TensorBoard log files to: ./tensorflow_logs/food_101_all_modelling_experiments/model_5/20250708-180741\n",
            "Epoch 11/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m456s\u001b[0m 161ms/step - accuracy: 0.5264 - loss: 1.8688 - val_accuracy: 0.6499 - val_loss: 1.3159\n",
            "Epoch 12/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m356s\u001b[0m 139ms/step - accuracy: 0.5420 - loss: 1.7817 - val_accuracy: 0.6539 - val_loss: 1.2953\n",
            "Epoch 13/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m383s\u001b[0m 140ms/step - accuracy: 0.5518 - loss: 1.7408 - val_accuracy: 0.6630 - val_loss: 1.2825\n",
            "Epoch 14/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m381s\u001b[0m 139ms/step - accuracy: 0.5616 - loss: 1.7025 - val_accuracy: 0.6665 - val_loss: 1.2616\n",
            "Epoch 15/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m382s\u001b[0m 140ms/step - accuracy: 0.5673 - loss: 1.6628 - val_accuracy: 0.6693 - val_loss: 1.2224\n",
            "Epoch 16/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m334s\u001b[0m 140ms/step - accuracy: 0.5755 - loss: 1.6369 - val_accuracy: 0.6697 - val_loss: 1.2290\n",
            "Epoch 17/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m333s\u001b[0m 140ms/step - accuracy: 0.5774 - loss: 1.6176 - val_accuracy: 0.6748 - val_loss: 1.2072\n",
            "Epoch 18/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m332s\u001b[0m 139ms/step - accuracy: 0.5850 - loss: 1.5909 - val_accuracy: 0.6816 - val_loss: 1.1988\n",
            "Epoch 19/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m333s\u001b[0m 139ms/step - accuracy: 0.5895 - loss: 1.5706 - val_accuracy: 0.6744 - val_loss: 1.1847\n",
            "Epoch 20/20\n",
            "\u001b[1m2368/2368\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m334s\u001b[0m 140ms/step - accuracy: 0.5953 - loss: 1.5454 - val_accuracy: 0.6875 - val_loss: 1.1628\n"
          ]
        }
      ],
      "source": [
        "# Unfreeze first 20 layers\n",
        "base_model.trainable = True\n",
        "for layer in base_model.layers[:-20]:\n",
        "  layer.trainable = False\n",
        "\n",
        "for layer in base_model.layers[-20:]:\n",
        "  if isinstance(layer, layers.BatchNormalization):\n",
        "    layer.trainable = False\n",
        "\n",
        "# Recompile\n",
        "model_5.compile(loss = tf.keras.losses.SparseCategoricalCrossentropy(from_logits = False),\n",
        "                optimizer = tf.keras.optimizers.Adam(learning_rate = 1e-5),\n",
        "                metrics = ['accuracy'])\n",
        "\n",
        "# Fit\n",
        "history_5_fine_tune = model_5.fit(train_data, epochs = total_epochs, initial_epoch = initial_epochs, steps_per_epoch = len(train_data), validation_data = test_data, validation_steps = int(0.1 * len(test_data)), callbacks = [create_tensorboard_callback(dir_name = './tensorflow_logs', experiment_name = 'food_101_all_modelling_experiments/model_5_fine-tuned'), model_checkpoint])"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        " Though model 5 was able to beat the accuracy score of the Food101 paper, model 1 still outperformed model 5 by ~4%.\n",
        "\n",
        " > 🔑 **Note:** Not all pretrained models from TensorFlow Hub will be able to beat the Food101 accuracy metric, despite their weights trained from ImageNet.\n",
        "\n",
        " For example, in our modelling experiments, model 1 (EfficientNetB0) has a simple pretrained model that outperformed model 5 (EfficientNetB2) which has a complex architecture. So, there are other factors to consider when picking a pretrained model, such as model size and its number of parameters. Using a larger model does not necessarily mean that you will get a better accuracy metric."
      ],
      "metadata": {
        "id": "vY-JTcaso3FN"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "SJLcvZK_9k4j",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 927
        },
        "outputId": "42f31a16-cb8e-4dd5-b3d2-fb4393bd3f14"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "# Plot fine-tuned results\n",
        "plot_loss_curves(history = history_5_fine_tune)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "FhB1OoNQWAQq"
      },
      "source": [
        "## Uploading our model results to TensorBoard\n",
        "\n",
        "We're now going to be uploading our modelling experiments to TensorBoard, so we can view each of our model's training logs: https://www.tensorflow.org/tensorboard"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Upload models' results to TensorBoard\n",
        "!pip install -q tensorboard\n",
        "%load_ext tensorboard\n",
        "%tensorboard --logdir=/content/tensorflow_logs"
      ],
      "metadata": {
        "id": "fzAfTOD9nh_C"
      },
      "execution_count": null,
      "outputs": []
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "gpuType": "T4",
      "machine_shape": "hm",
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}