EXEMPLO_04- MNIST Vision Transformer.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "9ab02b2d-f2a7-4540-9ac5-5d012b66dba3",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true,
    "tags": [],
    "pycharm": {
     "name": "#%% md\n"
    }
   },
   "source": [
    "# Exemplo 3: MNIST Vision Transformer (VT)\n",
    "\n",
    "**Nível de dificuldade:** <font color=red>avançado</font>\n",
    "\n",
    ":::{warning}\n",
    "\n",
    "Este modelo é significativamente maior que os demais. Deste modo, treina-lo demora horas, ao invés de minutos.\n",
    "Para comparação, a tabela a seguir mostra o número de parâmetros treináveis de cada modelo criado:\n",
    "\n",
    "|    **Modelo**    | **Parametros** | **Parametros treináveis** | **Tamanho relativo** |\n",
    "|:----------------:|:--------------:|:-------------------------:|:--------------------:|\n",
    "|      **VST**     |   21.662.157   |         21.662.154        |         <b><font color=red>622x</font></b>         |\n",
    "|    **Conv2D**    |     34.826     |           34.826          |          1x          |\n",
    "| **Feed-forward** |     409.405    |          409.405          |          11x         |\n",
    "\n",
    ":::"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f628c7fa-6943-4b9e-943f-89ce292b9616",
   "metadata": {
    "tags": [],
    "pycharm": {
     "name": "#%% md\n"
    }
   },
   "source": [
    "# Importando as bibliotecas necessárias\n",
    "\n",
    "Para o exemplo em questão, serão utilizadas as seguintes bibliotecas:\n",
    "\n",
    "- [Matplotlib](https://matplotlib.org/stable/index.html#installation)\n",
    "- [Numpy](https://numpy.org/install/)\n",
    "- [Tensorflow](https://www.tensorflow.org/install)\n",
    "- [tensorflow-addons](https://www.tensorflow.org/addons/overview)\n",
    "\n",
    "\n",
    "```{tip}\n",
    "Acesse os links listados acima, para maiores informações de como instalar as bibliotecas que serão utilizadas.\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "2dbdcdf3-2442-469e-ba1e-bee5d6c49653",
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    },
    "tags": [],
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: numpy in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (1.22.3)\n",
      "Requirement already satisfied: tensorflow in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (2.8.0)\n",
      "Requirement already satisfied: matplotlib in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (3.5.1)\n",
      "Requirement already satisfied: tensorflow-addons in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (0.16.1)\n",
      "Requirement already satisfied: setuptools in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from tensorflow) (60.9.3)\n",
      "Requirement already satisfied: keras<2.9,>=2.8.0rc0 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from tensorflow) (2.8.0)\n",
      "Requirement already satisfied: tensorboard<2.9,>=2.8 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from tensorflow) (2.8.0)\n",
      "Requirement already satisfied: tensorflow-io-gcs-filesystem>=0.23.1 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from tensorflow) (0.24.0)\n",
      "Requirement already satisfied: protobuf>=3.9.2 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from tensorflow) (3.19.4)\n",
      "Requirement already satisfied: grpcio<2.0,>=1.24.3 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from tensorflow) (1.44.0)\n",
      "Requirement already satisfied: typing-extensions>=3.6.6 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from tensorflow) (4.1.1)\n",
      "Requirement already satisfied: absl-py>=0.4.0 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from tensorflow) (1.0.0)\n",
      "Requirement already satisfied: tf-estimator-nightly==2.8.0.dev2021122109 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from tensorflow) (2.8.0.dev2021122109)\n",
      "Requirement already satisfied: opt-einsum>=2.3.2 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from tensorflow) (3.3.0)\n",
      "Requirement already satisfied: termcolor>=1.1.0 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from tensorflow) (1.1.0)\n",
      "Requirement already satisfied: astunparse>=1.6.0 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from tensorflow) (1.6.3)\n",
      "Requirement already satisfied: google-pasta>=0.1.1 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from tensorflow) (0.2.0)\n",
      "Requirement already satisfied: wrapt>=1.11.0 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from tensorflow) (1.13.3)\n",
      "Requirement already satisfied: libclang>=9.0.1 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from tensorflow) (13.0.0)\n",
      "Requirement already satisfied: flatbuffers>=1.12 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from tensorflow) (2.0)\n",
      "Requirement already satisfied: gast>=0.2.1 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from tensorflow) (0.5.3)\n",
      "Requirement already satisfied: h5py>=2.9.0 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from tensorflow) (3.6.0)\n",
      "Requirement already satisfied: six>=1.12.0 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from tensorflow) (1.16.0)\n",
      "Requirement already satisfied: keras-preprocessing>=1.1.1 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from tensorflow) (1.1.2)\n",
      "Requirement already satisfied: pillow>=6.2.0 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from matplotlib) (9.0.1)\n",
      "Requirement already satisfied: kiwisolver>=1.0.1 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from matplotlib) (1.3.2)\n",
      "Requirement already satisfied: cycler>=0.10 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from matplotlib) (0.11.0)\n",
      "Requirement already satisfied: python-dateutil>=2.7 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from matplotlib) (2.8.2)\n",
      "Requirement already satisfied: packaging>=20.0 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from matplotlib) (21.3)\n",
      "Requirement already satisfied: pyparsing>=2.2.1 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from matplotlib) (2.4.2)\n",
      "Requirement already satisfied: fonttools>=4.22.0 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from matplotlib) (4.29.1)\n",
      "Requirement already satisfied: typeguard>=2.7 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from tensorflow-addons) (2.13.3)\n",
      "Requirement already satisfied: wheel<1.0,>=0.23.0 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from astunparse>=1.6.0->tensorflow) (0.37.1)\n",
      "Requirement already satisfied: tensorboard-data-server<0.7.0,>=0.6.0 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from tensorboard<2.9,>=2.8->tensorflow) (0.6.1)\n",
      "Requirement already satisfied: markdown>=2.6.8 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from tensorboard<2.9,>=2.8->tensorflow) (3.3.6)\n",
      "Requirement already satisfied: google-auth<3,>=1.6.3 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from tensorboard<2.9,>=2.8->tensorflow) (1.35.0)\n",
      "Requirement already satisfied: google-auth-oauthlib<0.5,>=0.4.1 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from tensorboard<2.9,>=2.8->tensorflow) (0.4.6)\n",
      "Requirement already satisfied: requests<3,>=2.21.0 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from tensorboard<2.9,>=2.8->tensorflow) (2.25.1)\n",
      "Requirement already satisfied: tensorboard-plugin-wit>=1.6.0 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from tensorboard<2.9,>=2.8->tensorflow) (1.8.1)\n",
      "Requirement already satisfied: werkzeug>=0.11.15 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from tensorboard<2.9,>=2.8->tensorflow) (2.0.3)\n",
      "Requirement already satisfied: cachetools<5.0,>=2.0.0 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from google-auth<3,>=1.6.3->tensorboard<2.9,>=2.8->tensorflow) (4.2.4)\n",
      "Requirement already satisfied: pyasn1-modules>=0.2.1 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from google-auth<3,>=1.6.3->tensorboard<2.9,>=2.8->tensorflow) (0.2.8)\n",
      "Requirement already satisfied: rsa<5,>=3.1.4 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from google-auth<3,>=1.6.3->tensorboard<2.9,>=2.8->tensorflow) (4.8)\n",
      "Requirement already satisfied: requests-oauthlib>=0.7.0 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from google-auth-oauthlib<0.5,>=0.4.1->tensorboard<2.9,>=2.8->tensorflow) (1.3.1)\n",
      "Requirement already satisfied: importlib-metadata>=4.4 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from markdown>=2.6.8->tensorboard<2.9,>=2.8->tensorflow) (4.11.1)\n",
      "Requirement already satisfied: zipp>=0.5 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from importlib-metadata>=4.4->markdown>=2.6.8->tensorboard<2.9,>=2.8->tensorflow) (3.7.0)\n",
      "Requirement already satisfied: pyasn1<0.5.0,>=0.4.6 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from pyasn1-modules>=0.2.1->google-auth<3,>=1.6.3->tensorboard<2.9,>=2.8->tensorflow) (0.4.8)\n",
      "Requirement already satisfied: chardet<5,>=3.0.2 in /Users/erikingwersen/.local/lib/python3.8/site-packages (from requests<3,>=2.21.0->tensorboard<2.9,>=2.8->tensorflow) (4.0.0)\n",
      "Requirement already satisfied: certifi>=2017.4.17 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from requests<3,>=2.21.0->tensorboard<2.9,>=2.8->tensorflow) (2021.10.8)\n",
      "Requirement already satisfied: urllib3<1.27,>=1.21.1 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from requests<3,>=2.21.0->tensorboard<2.9,>=2.8->tensorflow) (1.26.8)\n",
      "Requirement already satisfied: idna<3,>=2.5 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from requests<3,>=2.21.0->tensorboard<2.9,>=2.8->tensorflow) (2.10)\n",
      "Requirement already satisfied: oauthlib>=3.0.0 in /opt/anaconda3/envs/pyml/lib/python3.8/site-packages (from requests-oauthlib>=0.7.0->google-auth-oauthlib<0.5,>=0.4.1->tensorboard<2.9,>=2.8->tensorflow) (3.2.0)\n"
     ]
    }
   ],
   "source": [
    "!pip install numpy tensorflow matplotlib tensorflow-addons"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6f79bea1-2cfa-4d83-a965-cc0ac5a82a40",
   "metadata": {
    "pycharm": {
     "name": "#%% md\n"
    }
   },
   "source": [
    "## Importando pacotes necessários"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "22b3904c-42ae-48ac-aa4f-eaee27cff6fe",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "from __future__ import annotations\n",
    "from typing import List, Any, Tuple, Iterable\n",
    "\n",
    "import os\n",
    "from pathlib import Path\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "from tensorflow import keras\n",
    "from tensorflow.keras import layers\n",
    "import tensorflow_addons as tfa\n",
    "import tensorflow as tf"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b865ca9e-f793-4190-a23c-0ed69b5bd057",
   "metadata": {
    "tags": [],
    "pycharm": {
     "name": "#%% md\n"
    }
   },
   "source": [
    "# Obtendo os dados do problema\n",
    "\n",
    "A próxima linha de código realiza o download dos dados a serem utilizados.\n",
    "A base será salva dentro do diretório `data/MNIST`.\n",
    "\n",
    ":::{note}\n",
    "As bases serão baixadas apenas uma vez. Execuções subsequentes carregarão o arquivo salvo em: `data/MNIST/mnist.npz`.\n",
    ":::\n",
    "\n",
    "**Fonte dos dados:** [MNIST homepage](http://yann.lecun.com/exdb/mnist/)\n",
    "\n",
    "**Função:** keras.datasets.mnist.load_data(path='mnist.npz')\n",
    "\n",
    "**Descrição:** Baixa e carrega o dataset MNIST.\n",
    "\n",
    "**Retorna:** tupla de tuplas de arrays [NumPy](https://numpy.org/doc/stable/user/quickstart.html), com os seguintes objetos: `(x_train, y_train), (x_test, y_test)`. Onde:\n",
    "\n",
    "\n",
    "- `x_train (np.ndarray[60000, 28, 28])`: 60.000 arrays de tamanho 28x28, representando o conjunto de imagens a serem utilizadas para treino do modelo.\n",
    "- `y_train (np.ndarray[60000])`: 60.000 labels, com valores entre 0 e 9, representando cada um dos dígitos de `x_train`.\n",
    "- `x_test (np.ndarray[10000, 28, 28])`: 10.000 arrays de tamanho 28x28, representando o conjunto de imagens a serem utilizadas para teste do modelo.\n",
    "- `y_test (np.ndarray[10000])`: 10.000 labels, com valores entre 0 e 9, representando cada um dos dígitos de `x_test`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "b5ec65cf-82ac-4d61-ade8-284e7477b90b",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "data_dir = Path.cwd().joinpath(r'data/MNIST')\n",
    "data_dir.mkdir(parents=True, exist_ok=True)\n",
    "mnist_fpath = data_dir.joinpath('mnist').with_suffix('.npz')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "6c32c6ce-970c-4da6-a7d2-0d5d4f33422d",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "# the data, split between train and test sets\n",
    "(x_train, y_train), (x_test, y_test) = keras.datasets.mnist.load_data(path=str(mnist_fpath))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "7c547788-cd08-48bc-9e5b-49b6381c5738",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "x_train shape: \u001B[94m\u001B[1m(60000, 28, 28, 1)\u001B[0m\n",
      "Quantidade de imagens a serem utilizadas durante o treino: \u001B[94m\u001B[1m60,000 imagens\u001B[0m\n",
      "Quantidade de imagens a serem utilizadas durante o teste: \u001B[94m\u001B[1m10,000 imagens\u001B[0m\n"
     ]
    }
   ],
   "source": [
    "# Convertendo dimensão das imagens para: (784, 1)\n",
    "x_train_norm = x_train.reshape(-1, 28, 28, 1)\n",
    "x_test_norm = x_test.reshape(-1, 28, 28, 1)\n",
    "y_train = y_train.reshape(-1, 1)\n",
    "\n",
    "print(f\"x_train shape: \\033[94m\\033[1m{x_train_norm.shape}\\033[0m\")\n",
    "print(f\"Quantidade de imagens a serem utilizadas durante o treino: \\033[94m\\033[1m{x_train_norm.shape[0]:,} imagens\\033[0m\")\n",
    "print(f\"Quantidade de imagens a serem utilizadas durante o teste: \\033[94m\\033[1m{x_test_norm.shape[0]:,} imagens\\033[0m\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ecfa49de-908c-4027-a2b2-669572fd4e13",
   "metadata": {
    "tags": [],
    "pycharm": {
     "name": "#%% md\n"
    }
   },
   "source": [
    "# Criando o modelo"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8c6b25ca-7705-4628-bba8-eecdd62ac66f",
   "metadata": {
    "pycharm": {
     "name": "#%% md\n"
    }
   },
   "source": [
    "## Configurando Híperparametros"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "9ef21735-f9a5-4e1e-b4f4-35c18b8e4335",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "input_shape = tuple(x_train_norm.shape[1:])\n",
    "\n",
    "learning_rate = 0.001\n",
    "weight_decay = 0.0001\n",
    "\n",
    "batch_size = 256\n",
    "num_epochs = 100\n",
    "\n",
    "image_size = 72  # We'll resize input images to this size\n",
    "patch_size = 6  # Size of the patches to be extract from the input images\n",
    "\n",
    "num_patches = (image_size // patch_size) ** 2\n",
    "\n",
    "projection_dim = 64\n",
    "num_heads = 4\n",
    "\n",
    "# Size of the transformer layers\n",
    "transformer_units = [projection_dim * 2, projection_dim]\n",
    "\n",
    "transformer_layers = 8\n",
    "mlp_head_units = [2048, 1024]  # Size of the dense layers of the final classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "f5206ad8-8ca2-4a12-bb0f-3787d285ebfb",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "data_augmentation = keras.Sequential(\n",
    "    [\n",
    "        layers.Normalization(),\n",
    "        layers.Resizing(image_size, image_size),\n",
    "        layers.RandomFlip(\"horizontal\"),\n",
    "        layers.RandomRotation(factor=0.02),\n",
    "        layers.RandomZoom(\n",
    "            height_factor=0.2, width_factor=0.2\n",
    "        ),\n",
    "    ],\n",
    "    name=\"data_augmentation\",\n",
    ")\n",
    "# Compute the mean and the variance of the training data for normalization.\n",
    "data_augmentation.layers[0].adapt(x_train_norm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "89fd4973-120b-4244-8d42-4bc96cfa30eb",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "\n",
    "def mlp(x, hidden_units, dropout_rate):\n",
    "    for units in hidden_units:\n",
    "        x = layers.Dense(units, activation=tf.nn.gelu)(x)\n",
    "        x = layers.Dropout(dropout_rate)(x)\n",
    "    return x\n",
    "\n",
    "\n",
    "\n",
    "class Patches(layers.Layer):\n",
    "    def __init__(self, patch_size):\n",
    "        super(Patches, self).__init__()\n",
    "        self.patch_size = patch_size\n",
    "\n",
    "    def call(self, images):\n",
    "        batch_size = tf.shape(images)[0]\n",
    "        patches = tf.image.extract_patches(\n",
    "            images=images,\n",
    "            sizes=[1, self.patch_size, self.patch_size, 1],\n",
    "            strides=[1, self.patch_size, self.patch_size, 1],\n",
    "            rates=[1, 1, 1, 1],\n",
    "            padding=\"VALID\",\n",
    "        )\n",
    "        patch_dims = patches.shape[-1]\n",
    "        patches = tf.reshape(patches, [batch_size, -1, patch_dims])\n",
    "        return patches\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "2496fe51-ac71-4e97-9aa6-25f8b7698e90",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Image size: 72 X 72\n",
      "Patch size: 6 X 6\n",
      "Patches per image: 144\n",
      "Elements per patch: 36\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 288x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 288x288 with 144 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(4, 4))\n",
    "\n",
    "image = x_train_norm[np.random.choice(range(x_train_norm.shape[0]))]\n",
    "\n",
    "plt.imshow(image.astype(\"uint8\"))\n",
    "plt.axis(\"off\")\n",
    "\n",
    "resized_image = tf.image.resize(\n",
    "    tf.convert_to_tensor([image]), size=(image_size, image_size)\n",
    ")\n",
    "patches = Patches(patch_size)(resized_image)\n",
    "print(f\"Image size: {image_size} X {image_size}\")\n",
    "print(f\"Patch size: {patch_size} X {patch_size}\")\n",
    "print(f\"Patches per image: {patches.shape[1]}\")\n",
    "print(f\"Elements per patch: {patches.shape[-1]}\")\n",
    "\n",
    "n = int(np.sqrt(patches.shape[1]))\n",
    "plt.figure(figsize=(4, 4))\n",
    "\n",
    "for i, patch in enumerate(patches[0]):\n",
    "    ax = plt.subplot(n, n, i + 1)\n",
    "    patch_img = tf.reshape(patch, (patch_size, patch_size, 1))\n",
    "    plt.imshow(patch_img.numpy().astype(\"uint8\"))\n",
    "    plt.axis(\"off\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "a9b5aeca-b6c6-4d51-8f88-c1fac1eaeb0e",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "class PatchEncoder(layers.Layer):\n",
    "    def __init__(self, num_patches, projection_dim):\n",
    "        super(PatchEncoder, self).__init__()\n",
    "        self.num_patches = num_patches\n",
    "        self.projection = layers.Dense(units=projection_dim)\n",
    "        self.position_embedding = layers.Embedding(\n",
    "            input_dim=num_patches, output_dim=projection_dim\n",
    "        )\n",
    "\n",
    "    def call(self, patch):\n",
    "        positions = tf.range(start=0, limit=self.num_patches, delta=1)\n",
    "        encoded = self.projection(patch) + self.position_embedding(positions)\n",
    "        return encoded"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "2fe3c05f-0122-4eb6-a88d-23c71a3f5803",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "input_shape = tuple(x_train_norm.shape[1:])\n",
    "\n",
    "learning_rate = 0.001\n",
    "weight_decay = 0.0001\n",
    "\n",
    "batch_size = 256\n",
    "num_epochs = 100\n",
    "\n",
    "image_size = 72  # We'll resize input images to this size\n",
    "patch_size = 6  # Size of the patches to be extract from the input images\n",
    "\n",
    "num_patches = (image_size // patch_size) ** 2\n",
    "\n",
    "projection_dim = 64\n",
    "num_heads = 4\n",
    "\n",
    "# Size of the transformer layers\n",
    "transformer_units = [projection_dim * 2, projection_dim]\n",
    "\n",
    "transformer_layers = 8\n",
    "mlp_head_units = [2048, 1024]  # Size of the dense layers of the final classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "e0169950-6db0-4699-bec7-a4815d2d947a",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "def create_vit_classifier(\n",
    "    input_shape: tuple | None = None,\n",
    "    image_size: int = 72,\n",
    "    patch_size: int = 6,\n",
    "    projection_dim: int = 64,\n",
    "    num_heads: int = 4,\n",
    "    transformer_layers: int = 8,\n",
    "    transformer_units: List[int] | None = None,\n",
    "    mlp_head_units: List[int] | None = None,\n",
    "    num_classes : int = 10,\n",
    "    \n",
    "):\n",
    "    \"\"\"Create Vision Transformer (VIT) classifier model.\n",
    "\n",
    "    Parameters\n",
    "    ----------\n",
    "    input_shape : tuple, optional\n",
    "        The shape of the input data. Defaults to (28, 28, 1).\n",
    "    image_size : int, optional\n",
    "        Resize input images to this size. Defaults to 72 (72 by 72).\n",
    "    patch_size : int, optional\n",
    "        Size of the patches to be extract from the input images.\n",
    "        Defaults to 6 (6 by 6).\n",
    "    projection_dim : int, optional\n",
    "        The dimension of the projection. Defaults to 64.\n",
    "    num_heads : int, optional\n",
    "        The number of heads in the multi-head attention layer.\n",
    "        Defaults to 4.\n",
    "    transformer_layers : int, optional\n",
    "        The number of transformer layers. Defaults to 8.\n",
    "    transformer_units : List[int] | None, optional\n",
    "        Size of the transformer layers. \n",
    "        Defaults to [`projection_dim` * 2, `projection_dim`].\n",
    "    mlp_head_units : List[int] | None, optional\n",
    "        Size of the dense layers of the final classifier.\n",
    "        Defaults to [2048, 1024].\n",
    "    num_classes : int, optional\n",
    "        The number of classes from the target layer. Defaults to 10.\n",
    "\n",
    "    Returns\n",
    "    -------\n",
    "    keras.Model\n",
    "        The Vision Transformer classifier.\n",
    "    \"\"\"\n",
    "    num_patches = (image_size // patch_size) ** 2\n",
    "    \n",
    "    if input_shape is None:\n",
    "        input_shape = (28, 28, 1)\n",
    "    if transformer_units is None:\n",
    "        transformer_units = [projection_dim * 2, projection_dim]\n",
    "    if mlp_head_units is None:\n",
    "        mlp_head_units = [2048, 1024]\n",
    "    inputs = layers.Input(shape=input_shape)\n",
    "    \n",
    "    # Augment data.\n",
    "    augmented = data_augmentation(inputs)\n",
    "    \n",
    "    # Create patches.\n",
    "    patches = Patches(patch_size)(augmented)\n",
    "    \n",
    "    # Encode patches.\n",
    "    encoded_patches = PatchEncoder(num_patches, projection_dim)(patches)\n",
    "\n",
    "    # Create multiple layers of the Transformer block.\n",
    "    for _ in range(transformer_layers):\n",
    "        \n",
    "        # Layer normalization 1.\n",
    "        x1 = layers.LayerNormalization(epsilon=1e-6)(encoded_patches)\n",
    "        \n",
    "        # Create a multi-head attention layer.\n",
    "        attention_output = layers.MultiHeadAttention(\n",
    "            num_heads=num_heads, key_dim=projection_dim, dropout=0.1\n",
    "        )(x1, x1)\n",
    "        \n",
    "        # Skip connection 1.\n",
    "        x2 = layers.Add()([attention_output, encoded_patches])\n",
    "        \n",
    "        # Layer normalization 2.\n",
    "        x3 = layers.LayerNormalization(epsilon=1e-6)(x2)\n",
    "        \n",
    "        # MLP.\n",
    "        x3 = mlp(x3, hidden_units=transformer_units, dropout_rate=0.1)\n",
    "        \n",
    "        # Skip connection 2.\n",
    "        encoded_patches = layers.Add()([x3, x2])\n",
    "\n",
    "    # Create a [batch_size, projection_dim] tensor.\n",
    "    representation = layers.LayerNormalization(epsilon=1e-6)(encoded_patches)\n",
    "    representation = layers.Flatten()(representation)\n",
    "    representation = layers.Dropout(0.5)(representation)\n",
    "    \n",
    "    # Add MLP.\n",
    "    features = mlp(representation, hidden_units=mlp_head_units, dropout_rate=0.5)\n",
    "    \n",
    "    # Classify outputs.\n",
    "    logits = layers.Dense(num_classes)(features)\n",
    "    \n",
    "    # Create the Keras model.\n",
    "    model = keras.Model(inputs=inputs, outputs=logits)\n",
    "    \n",
    "    return model\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "8ca1be5a-5ba1-4d53-aa4f-bfefc02eb5a2",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model\"\n",
      "__________________________________________________________________________________________________\n",
      " Layer (type)                   Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      " input_1 (InputLayer)           [(None, 28, 28, 1)]  0           []                               \n",
      "                                                                                                  \n",
      " data_augmentation (Sequential)  (None, 72, 72, 1)   3           ['input_1[0][0]']                \n",
      "                                                                                                  \n",
      " patches_1 (Patches)            (None, None, 36)     0           ['data_augmentation[0][0]']      \n",
      "                                                                                                  \n",
      " patch_encoder (PatchEncoder)   (None, 144, 64)      11584       ['patches_1[0][0]']              \n",
      "                                                                                                  \n",
      " layer_normalization (LayerNorm  (None, 144, 64)     128         ['patch_encoder[0][0]']          \n",
      " alization)                                                                                       \n",
      "                                                                                                  \n",
      " multi_head_attention (MultiHea  (None, 144, 64)     66368       ['layer_normalization[0][0]',    \n",
      " dAttention)                                                      'layer_normalization[0][0]']    \n",
      "                                                                                                  \n",
      " add (Add)                      (None, 144, 64)      0           ['multi_head_attention[0][0]',   \n",
      "                                                                  'patch_encoder[0][0]']          \n",
      "                                                                                                  \n",
      " layer_normalization_1 (LayerNo  (None, 144, 64)     128         ['add[0][0]']                    \n",
      " rmalization)                                                                                     \n",
      "                                                                                                  \n",
      " dense_1 (Dense)                (None, 144, 128)     8320        ['layer_normalization_1[0][0]']  \n",
      "                                                                                                  \n",
      " dropout (Dropout)              (None, 144, 128)     0           ['dense_1[0][0]']                \n",
      "                                                                                                  \n",
      " dense_2 (Dense)                (None, 144, 64)      8256        ['dropout[0][0]']                \n",
      "                                                                                                  \n",
      " dropout_1 (Dropout)            (None, 144, 64)      0           ['dense_2[0][0]']                \n",
      "                                                                                                  \n",
      " add_1 (Add)                    (None, 144, 64)      0           ['dropout_1[0][0]',              \n",
      "                                                                  'add[0][0]']                    \n",
      "                                                                                                  \n",
      " layer_normalization_2 (LayerNo  (None, 144, 64)     128         ['add_1[0][0]']                  \n",
      " rmalization)                                                                                     \n",
      "                                                                                                  \n",
      " multi_head_attention_1 (MultiH  (None, 144, 64)     66368       ['layer_normalization_2[0][0]',  \n",
      " eadAttention)                                                    'layer_normalization_2[0][0]']  \n",
      "                                                                                                  \n",
      " add_2 (Add)                    (None, 144, 64)      0           ['multi_head_attention_1[0][0]', \n",
      "                                                                  'add_1[0][0]']                  \n",
      "                                                                                                  \n",
      " layer_normalization_3 (LayerNo  (None, 144, 64)     128         ['add_2[0][0]']                  \n",
      " rmalization)                                                                                     \n",
      "                                                                                                  \n",
      " dense_3 (Dense)                (None, 144, 128)     8320        ['layer_normalization_3[0][0]']  \n",
      "                                                                                                  \n",
      " dropout_2 (Dropout)            (None, 144, 128)     0           ['dense_3[0][0]']                \n",
      "                                                                                                  \n",
      " dense_4 (Dense)                (None, 144, 64)      8256        ['dropout_2[0][0]']              \n",
      "                                                                                                  \n",
      " dropout_3 (Dropout)            (None, 144, 64)      0           ['dense_4[0][0]']                \n",
      "                                                                                                  \n",
      " add_3 (Add)                    (None, 144, 64)      0           ['dropout_3[0][0]',              \n",
      "                                                                  'add_2[0][0]']                  \n",
      "                                                                                                  \n",
      " layer_normalization_4 (LayerNo  (None, 144, 64)     128         ['add_3[0][0]']                  \n",
      " rmalization)                                                                                     \n",
      "                                                                                                  \n",
      " multi_head_attention_2 (MultiH  (None, 144, 64)     66368       ['layer_normalization_4[0][0]',  \n",
      " eadAttention)                                                    'layer_normalization_4[0][0]']  \n",
      "                                                                                                  \n",
      " add_4 (Add)                    (None, 144, 64)      0           ['multi_head_attention_2[0][0]', \n",
      "                                                                  'add_3[0][0]']                  \n",
      "                                                                                                  \n",
      " layer_normalization_5 (LayerNo  (None, 144, 64)     128         ['add_4[0][0]']                  \n",
      " rmalization)                                                                                     \n",
      "                                                                                                  \n",
      " dense_5 (Dense)                (None, 144, 128)     8320        ['layer_normalization_5[0][0]']  \n",
      "                                                                                                  \n",
      " dropout_4 (Dropout)            (None, 144, 128)     0           ['dense_5[0][0]']                \n",
      "                                                                                                  \n",
      " dense_6 (Dense)                (None, 144, 64)      8256        ['dropout_4[0][0]']              \n",
      "                                                                                                  \n",
      " dropout_5 (Dropout)            (None, 144, 64)      0           ['dense_6[0][0]']                \n",
      "                                                                                                  \n",
      " add_5 (Add)                    (None, 144, 64)      0           ['dropout_5[0][0]',              \n",
      "                                                                  'add_4[0][0]']                  \n",
      "                                                                                                  \n",
      " layer_normalization_6 (LayerNo  (None, 144, 64)     128         ['add_5[0][0]']                  \n",
      " rmalization)                                                                                     \n",
      "                                                                                                  \n",
      " multi_head_attention_3 (MultiH  (None, 144, 64)     66368       ['layer_normalization_6[0][0]',  \n",
      " eadAttention)                                                    'layer_normalization_6[0][0]']  \n",
      "                                                                                                  \n",
      " add_6 (Add)                    (None, 144, 64)      0           ['multi_head_attention_3[0][0]', \n",
      "                                                                  'add_5[0][0]']                  \n",
      "                                                                                                  \n",
      " layer_normalization_7 (LayerNo  (None, 144, 64)     128         ['add_6[0][0]']                  \n",
      " rmalization)                                                                                     \n",
      "                                                                                                  \n",
      " dense_7 (Dense)                (None, 144, 128)     8320        ['layer_normalization_7[0][0]']  \n",
      "                                                                                                  \n",
      " dropout_6 (Dropout)            (None, 144, 128)     0           ['dense_7[0][0]']                \n",
      "                                                                                                  \n",
      " dense_8 (Dense)                (None, 144, 64)      8256        ['dropout_6[0][0]']              \n",
      "                                                                                                  \n",
      " dropout_7 (Dropout)            (None, 144, 64)      0           ['dense_8[0][0]']                \n",
      "                                                                                                  \n",
      " add_7 (Add)                    (None, 144, 64)      0           ['dropout_7[0][0]',              \n",
      "                                                                  'add_6[0][0]']                  \n",
      "                                                                                                  \n",
      " layer_normalization_8 (LayerNo  (None, 144, 64)     128         ['add_7[0][0]']                  \n",
      " rmalization)                                                                                     \n",
      "                                                                                                  \n",
      " multi_head_attention_4 (MultiH  (None, 144, 64)     66368       ['layer_normalization_8[0][0]',  \n",
      " eadAttention)                                                    'layer_normalization_8[0][0]']  \n",
      "                                                                                                  \n",
      " add_8 (Add)                    (None, 144, 64)      0           ['multi_head_attention_4[0][0]', \n",
      "                                                                  'add_7[0][0]']                  \n",
      "                                                                                                  \n",
      " layer_normalization_9 (LayerNo  (None, 144, 64)     128         ['add_8[0][0]']                  \n",
      " rmalization)                                                                                     \n",
      "                                                                                                  \n",
      " dense_9 (Dense)                (None, 144, 128)     8320        ['layer_normalization_9[0][0]']  \n",
      "                                                                                                  \n",
      " dropout_8 (Dropout)            (None, 144, 128)     0           ['dense_9[0][0]']                \n",
      "                                                                                                  \n",
      " dense_10 (Dense)               (None, 144, 64)      8256        ['dropout_8[0][0]']              \n",
      "                                                                                                  \n",
      " dropout_9 (Dropout)            (None, 144, 64)      0           ['dense_10[0][0]']               \n",
      "                                                                                                  \n",
      " add_9 (Add)                    (None, 144, 64)      0           ['dropout_9[0][0]',              \n",
      "                                                                  'add_8[0][0]']                  \n",
      "                                                                                                  \n",
      " layer_normalization_10 (LayerN  (None, 144, 64)     128         ['add_9[0][0]']                  \n",
      " ormalization)                                                                                    \n",
      "                                                                                                  \n",
      " multi_head_attention_5 (MultiH  (None, 144, 64)     66368       ['layer_normalization_10[0][0]', \n",
      " eadAttention)                                                    'layer_normalization_10[0][0]'] \n",
      "                                                                                                  \n",
      " add_10 (Add)                   (None, 144, 64)      0           ['multi_head_attention_5[0][0]', \n",
      "                                                                  'add_9[0][0]']                  \n",
      "                                                                                                  \n",
      " layer_normalization_11 (LayerN  (None, 144, 64)     128         ['add_10[0][0]']                 \n",
      " ormalization)                                                                                    \n",
      "                                                                                                  \n",
      " dense_11 (Dense)               (None, 144, 128)     8320        ['layer_normalization_11[0][0]'] \n",
      "                                                                                                  \n",
      " dropout_10 (Dropout)           (None, 144, 128)     0           ['dense_11[0][0]']               \n",
      "                                                                                                  \n",
      " dense_12 (Dense)               (None, 144, 64)      8256        ['dropout_10[0][0]']             \n",
      "                                                                                                  \n",
      " dropout_11 (Dropout)           (None, 144, 64)      0           ['dense_12[0][0]']               \n",
      "                                                                                                  \n",
      " add_11 (Add)                   (None, 144, 64)      0           ['dropout_11[0][0]',             \n",
      "                                                                  'add_10[0][0]']                 \n",
      "                                                                                                  \n",
      " layer_normalization_12 (LayerN  (None, 144, 64)     128         ['add_11[0][0]']                 \n",
      " ormalization)                                                                                    \n",
      "                                                                                                  \n",
      " multi_head_attention_6 (MultiH  (None, 144, 64)     66368       ['layer_normalization_12[0][0]', \n",
      " eadAttention)                                                    'layer_normalization_12[0][0]'] \n",
      "                                                                                                  \n",
      " add_12 (Add)                   (None, 144, 64)      0           ['multi_head_attention_6[0][0]', \n",
      "                                                                  'add_11[0][0]']                 \n",
      "                                                                                                  \n",
      " layer_normalization_13 (LayerN  (None, 144, 64)     128         ['add_12[0][0]']                 \n",
      " ormalization)                                                                                    \n",
      "                                                                                                  \n",
      " dense_13 (Dense)               (None, 144, 128)     8320        ['layer_normalization_13[0][0]'] \n",
      "                                                                                                  \n",
      " dropout_12 (Dropout)           (None, 144, 128)     0           ['dense_13[0][0]']               \n",
      "                                                                                                  \n",
      " dense_14 (Dense)               (None, 144, 64)      8256        ['dropout_12[0][0]']             \n",
      "                                                                                                  \n",
      " dropout_13 (Dropout)           (None, 144, 64)      0           ['dense_14[0][0]']               \n",
      "                                                                                                  \n",
      " add_13 (Add)                   (None, 144, 64)      0           ['dropout_13[0][0]',             \n",
      "                                                                  'add_12[0][0]']                 \n",
      "                                                                                                  \n",
      " layer_normalization_14 (LayerN  (None, 144, 64)     128         ['add_13[0][0]']                 \n",
      " ormalization)                                                                                    \n",
      "                                                                                                  \n",
      " multi_head_attention_7 (MultiH  (None, 144, 64)     66368       ['layer_normalization_14[0][0]', \n",
      " eadAttention)                                                    'layer_normalization_14[0][0]'] \n",
      "                                                                                                  \n",
      " add_14 (Add)                   (None, 144, 64)      0           ['multi_head_attention_7[0][0]', \n",
      "                                                                  'add_13[0][0]']                 \n",
      "                                                                                                  \n",
      " layer_normalization_15 (LayerN  (None, 144, 64)     128         ['add_14[0][0]']                 \n",
      " ormalization)                                                                                    \n",
      "                                                                                                  \n",
      " dense_15 (Dense)               (None, 144, 128)     8320        ['layer_normalization_15[0][0]'] \n",
      "                                                                                                  \n",
      " dropout_14 (Dropout)           (None, 144, 128)     0           ['dense_15[0][0]']               \n",
      "                                                                                                  \n",
      " dense_16 (Dense)               (None, 144, 64)      8256        ['dropout_14[0][0]']             \n",
      "                                                                                                  \n",
      " dropout_15 (Dropout)           (None, 144, 64)      0           ['dense_16[0][0]']               \n",
      "                                                                                                  \n",
      " add_15 (Add)                   (None, 144, 64)      0           ['dropout_15[0][0]',             \n",
      "                                                                  'add_14[0][0]']                 \n",
      "                                                                                                  \n",
      " layer_normalization_16 (LayerN  (None, 144, 64)     128         ['add_15[0][0]']                 \n",
      " ormalization)                                                                                    \n",
      "                                                                                                  \n",
      " flatten (Flatten)              (None, 9216)         0           ['layer_normalization_16[0][0]'] \n",
      "                                                                                                  \n",
      " dropout_16 (Dropout)           (None, 9216)         0           ['flatten[0][0]']                \n",
      "                                                                                                  \n",
      " dense_17 (Dense)               (None, 2048)         18876416    ['dropout_16[0][0]']             \n",
      "                                                                                                  \n",
      " dropout_17 (Dropout)           (None, 2048)         0           ['dense_17[0][0]']               \n",
      "                                                                                                  \n",
      " dense_18 (Dense)               (None, 1024)         2098176     ['dropout_17[0][0]']             \n",
      "                                                                                                  \n",
      " dropout_18 (Dropout)           (None, 1024)         0           ['dense_18[0][0]']               \n",
      "                                                                                                  \n",
      " dense_19 (Dense)               (None, 10)           10250       ['dropout_18[0][0]']             \n",
      "                                                                                                  \n",
      "==================================================================================================\n",
      "Total params: 21,662,157\n",
      "Trainable params: 21,662,154\n",
      "Non-trainable params: 3\n",
      "__________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "vit_classifier = create_vit_classifier()\n",
    "vit_classifier.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "0d53ddf3-50e3-4ad9-a3a8-85de8eb1d931",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "211/211 [==============================] - ETA: 0s - loss: 1.1037 - accuracy: 0.7046 - top-5-accuracy: 0.9536"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "KeyboardInterrupt\n",
      "\n"
     ]
    }
   ],
   "source": [
    "\n",
    "def run_experiment(model):\n",
    "\n",
    "    optimizer = tfa.optimizers.AdamW(\n",
    "        learning_rate=learning_rate, weight_decay=weight_decay\n",
    "    )\n",
    "\n",
    "    model.compile(\n",
    "        optimizer=optimizer,\n",
    "        loss=keras.losses.SparseCategoricalCrossentropy(from_logits=True),\n",
    "        metrics=[\n",
    "            keras.metrics.SparseCategoricalAccuracy(name=\"accuracy\"),\n",
    "            keras.metrics.SparseTopKCategoricalAccuracy(5, name=\"top-5-accuracy\"),\n",
    "        ],\n",
    "    )\n",
    "\n",
    "    checkpoint_filepath = \"/tmp/checkpoint\"\n",
    "    checkpoint_callback = keras.callbacks.ModelCheckpoint(\n",
    "        checkpoint_filepath,\n",
    "        monitor=\"val_accuracy\",\n",
    "        save_best_only=True,\n",
    "        save_weights_only=True,\n",
    "    )\n",
    "\n",
    "    history = model.fit(\n",
    "        x=x_train_norm,\n",
    "        y=y_train,\n",
    "        batch_size=batch_size,\n",
    "        epochs=num_epochs,\n",
    "        validation_split=0.1,\n",
    "        callbacks=[checkpoint_callback],\n",
    "    )\n",
    "\n",
    "    model.load_weights(checkpoint_filepath)\n",
    "    _, accuracy, top_5_accuracy = model.evaluate(x_test, y_test)\n",
    "    \n",
    "    print(f\"Test accuracy: {round(accuracy * 100, 2)}%\")\n",
    "    print(f\"Test top 5 accuracy: {round(top_5_accuracy * 100, 2)}%\")\n",
    "\n",
    "    return history\n",
    "\n",
    "\n",
    "vit_classifier = create_vit_classifier()\n",
    "history = run_experiment(vit_classifier)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "c63f1f1c-2b0c-4d54-9ba7-72f330113e37",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "%load_ext tensorboard"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "id": "6c612ea9-918b-4940-a65f-ed803f084db7",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "log_dir = Path(f\"./logs/mnist-fit/{datetime.datetime.now().strftime('%Y%m%d-%H%M%S')}\")\n",
    "log_dir.mkdir(parents=True, exist_ok=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "id": "ae4f8ade-7fc9-4a47-a14e-7f041bb50798",
   "metadata": {
    "tags": [],
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/15\n",
      "469/469 [==============================] - 3s 6ms/step - loss: 0.0101 - accuracy: 0.9971 - val_loss: 0.0120 - val_accuracy: 0.9973\n",
      "Epoch 2/15\n",
      "469/469 [==============================] - 3s 7ms/step - loss: 0.0107 - accuracy: 0.9969 - val_loss: 0.0024 - val_accuracy: 0.9993\n",
      "Epoch 3/15\n",
      "469/469 [==============================] - 3s 7ms/step - loss: 0.0087 - accuracy: 0.9974 - val_loss: 0.0032 - val_accuracy: 0.9990\n",
      "Epoch 4/15\n",
      "469/469 [==============================] - 3s 7ms/step - loss: 0.0087 - accuracy: 0.9975 - val_loss: 0.0044 - val_accuracy: 0.9987\n",
      "Epoch 5/15\n",
      "469/469 [==============================] - 4s 8ms/step - loss: 0.0088 - accuracy: 0.9973 - val_loss: 0.0026 - val_accuracy: 0.9993\n",
      "Epoch 6/15\n",
      "469/469 [==============================] - 4s 8ms/step - loss: 0.0089 - accuracy: 0.9975 - val_loss: 0.0019 - val_accuracy: 0.9993\n",
      "Epoch 7/15\n",
      "469/469 [==============================] - 4s 8ms/step - loss: 0.0079 - accuracy: 0.9979 - val_loss: 0.0017 - val_accuracy: 0.9997\n",
      "Epoch 8/15\n",
      "469/469 [==============================] - 5s 10ms/step - loss: 0.0074 - accuracy: 0.9980 - val_loss: 0.0012 - val_accuracy: 0.9997\n",
      "Epoch 9/15\n",
      "469/469 [==============================] - 4s 9ms/step - loss: 0.0094 - accuracy: 0.9973 - val_loss: 0.0041 - val_accuracy: 0.9988\n",
      "Epoch 10/15\n",
      "469/469 [==============================] - 5s 10ms/step - loss: 0.0095 - accuracy: 0.9971 - val_loss: 0.0014 - val_accuracy: 0.9997\n",
      "Epoch 11/15\n",
      "469/469 [==============================] - 6s 12ms/step - loss: 0.0076 - accuracy: 0.9979 - val_loss: 0.0026 - val_accuracy: 0.9992\n",
      "Epoch 12/15\n",
      "469/469 [==============================] - 6s 13ms/step - loss: 0.0081 - accuracy: 0.9978 - val_loss: 0.0010 - val_accuracy: 0.9997\n",
      "Epoch 13/15\n",
      "469/469 [==============================] - 6s 12ms/step - loss: 0.0063 - accuracy: 0.9983 - val_loss: 0.0038 - val_accuracy: 0.9990\n",
      "Epoch 14/15\n",
      "469/469 [==============================] - 5s 11ms/step - loss: 0.0059 - accuracy: 0.9982 - val_loss: 0.0046 - val_accuracy: 0.9990\n",
      "Epoch 15/15\n",
      "469/469 [==============================] - 6s 12ms/step - loss: 0.0093 - accuracy: 0.9970 - val_loss: 0.0011 - val_accuracy: 0.9998\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x7f89a961a3a0>"
      ]
     },
     "execution_count": 109,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAT4AAAEGCAYAAAD8EfnwAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAAA8+ElEQVR4nO2deXxU5fX/32dmsgcSQthBdqFxQ5Stboi4W1FbWlut1KpUq9bWVtT2W/3qz65ftaVatVS0uCuK+wKIWlxRUJFFgbAvYQmQfZ85vz/uDQQIyZ3MfZIZ8rxfr/vKzJ17zz157p0zz3o+oqpYLBZLeyLQ1g5YLBZLa2MDn8ViaXfYwGexWNodNvBZLJZ2hw18Foul3RFqawca0iEnpF16pfhut3Cp/zYtlkSkinJqtFpisXHmqRm6c1fY07GLvqqerapnxXI9E8RV4OvSK4W7Zh3hu91Hh/T13abFkogs0Hkx2yjcFWbB7N6ejk3qsTo35gsaIK4Cn8ViSQSUsEba2omYsIHPYrFEhQIREnvhgw18FoslaiLYGp/FYmlHKEqtbepaLJb2hAJh29Q1w7IZHVg5MxMUDp9YxhE/KeXdX+ZSsjYJgJrSAMkdIkx4uWDPOWVbgrx4bk+GXVfMUVeURHW948eWcPX/20IwoLz5dA7P3d8tJv+79KzhpqkbyO5SBwpvPNGZl6Z3icmmKV8Bbrx3A6PGl1JUGOJn44b44KW12xAT98yk3eZI9D4+oxOYReQsEVkhIvkicovX83avTGLlzEy+M3MrE14uYON7aZSsD3Hq3wuZ8HIBE14uoO8ZFfQ9vWKf8z79cyd6n1QZtZ+BgHLtHzfzP5f056qxQzh1QhGHDa6K2k5DwnXCtDt7MnnsUG44bzDf+UlhzDZN+Qow59kcfndJ/5jtWLsHYuqembLbHAqEVT1t8YqxwCciQeCfwNlAHvBDEcnzcm7R6iS6HF1DKE0JhKD7iGrWz0nf87kqrH0znf7nle/Zt/7tNDr0qiN7cG3Uvg45toIt65LZuiGFutoA772czZgzi6O205Bd25PIX+L4XFkeZGN+Krk9ovetNXwFWLogk9Ld/jcArF1z98yUXS9EPG7xiska30ggX1XXqGoN8AwwwcuJnQ6vYduiFKp2B6irFDbNT6N8a3DP59sWppDWOUxWvzoAasuFJf/OYth1LbvpnbvXsmNL8p73hQVJvgSperr1rmHgkZV883l68wc3g2lfLf5j6p611bOgKGGPW7xiso+vF7CxwftNwKj9DxKRycBkgNyezk3MHljHUVeWMOeKroTSlJyhNUiDEL3mtQwGNKjtfXF/FkdMKiEpI/4KOjU9zO8fXsdDt/WkoizY/AkWS5yjCrXx91WLijYf3FDVacA0gAFH7Y1ch08s4/CJZQAsujeb9G5O7S5SB+vnpnP+rL2DGoWLU1g/O4OFd3eipiQAASWYouRdWurJh51bk+jSs2bP+9wetRQWJMX8vwVDyu8fXsc7szrx4ZvZMdsDc75azGHqnrXdsyCEiWm5b5tjsqm7GejT4H1vd58nKnc6rpVtCbJ+TjoDvuPU8LZ8lErWgFoyuu9dJH3OU9uY+M5mJr6zmbxJJRz9sxLPQQ9gxZfp9OpfQ7c+1YSSIoydUMQnc7I8n984yo33bGTjqlRmTfNnNBdM+Woxial71lbPggIR9bbFKyZrfJ8Bg0WkP07Auxj4kdeT372+C1VFAQIhGH37LlI6OqW49o0MBpxb3szZ0REJC//8XS/++NQaAkGY80wO61emxmTziJHljJ+4mzXLU3lg7goAHv1TDz57p2Pc+QpwywPrOXpMGVk5dTyxcDmP39ON2U93tnZ9sGvqnpmy64VEr/GJSbEhETkH+DsQBB5R1T80dfyAozLUZmexWMyxQOdRortiilpHHJ2sz7ze1dOxRx+2eZGqHh/L9UxgtI9PVd8A3jB5DYvF0rooUKuJncO4zQc3LBZLYqEI4QRP3m4Dn8ViiZqIJnYfnw18FoslKpwkBTbwWSyWdoUQTvA+vsT23mKxtDpOBuaAp605ROQREdkuIksb7MsRkbkissr928ndLyLyDzfpyVciMrzBOZPc41eJyKTmrhtXNb7CpSlGpp7M3vKl7zYBzuw5zIhdiyWeURVq1Lfll/8B7gcea7DvFmCeqv7Zzep0C3AzTsKTwe42CngQGCUiOcDtwPE4cXmRiLyiqrsPdlFb47NYLFETQTxtzaGq84Fd++2eAMxwX88ALmiw/zF1+ATIFpEewJnAXFXd5Qa7uUCTkpZxVeOzWCzxjzO44bnOlCsiCxu8n+auz2+Kbqpavxh/K1CfXbWxxCe9mth/UGzgs1gsURLV4EZhLCs3VFVFxPflZbapa7FYosLPwY2DsM1twuL+3e7uP1jik6gTotjAZ7FYoias4mlrIa8A9SOzk4CXG+y/zB3dHQ0Uu03i2cAZItLJHQE+w913UGxT12KxRIUi1Ko/oUNEngbG4vQFbsIZnf0z8JyIXAGsB77vHv4GcA6QD1QAlwOo6i4R+X84GaEA7lTV/QdM9iHhAl9LVKW+f9QRZOfWMe1dJz3U/FezePye7mxclco/3ljJ4cc4AkWL/pvJI3/sSV2tEEpSrvr9FoadWLaPrdsn9adgQ/IeWwfDpMpaIimBmfI1KSXCPbPySUpWgiHl/dezefzu7jHZNHXPEk1xrzmiHNxo2pbqDw/y0WmNHKvAtQex8wjwiNfrmhQbOmBiYqy0VFXqD0+u2ed9v6FV3PbwOo4avW9ev6ycMHfOWMO/3lnBTVM38NdfHLbP5x+8kUVqhjcJFVMqa5BYSmCm1NBqq4UpEwdyzelDuOb0IRw/tpShw2PL02jqniWa4l5zKN6auTE0dY1jso/vPzQzlyZaWqoq1aFTeJ/3hw2ups+g6gOOG3RUJZ27Oynu+w6poroqQE21c/MqywPM+lcXfvTLrZ58NaWyBomlBGZKDQ2EqgpnEm0oSQkmKbGmljR1zxJNcc8Lhgc3jGPMs4NMTIyJ1lSV+uD1LAYdWUlyivNtmvHX7nz36h2kpEX/7fJTZc0UiajeFggoD8xdwbNfLeOL+Zms+CLDN9um7tmhoLinCmENeNrilTb3TEQmi8hCEVlYy4G1sLZg3YpUpv+hJzf81ZkTuXppGgXrUjjh7Oh/Ta3KmjkiEeHnpw/hkuPyGDKsgr5DoheTbwxT9+xQeRacwY2gpy1eafPBjYYqax0lp8nqVGuoSu3YksSdV/Tjpqkb6NnPudbyRems/Cqdy0bmEQ5DUWGIm747qFlbJlTWTJHI6m3lJUEWf5TJiFNLWb8iLSZbpu7Zoaa4l+iJSBPKe9OqUmXFQX5/2QB++tsCjhi5t6P8O5N28vQXy3js0+Xc81I+vQZU838v5DdjzYzKmikSTb0tK6eOjI5O321yaoThJ5exMT9WoR1T9+zQUtxThIh62+KVNq/xRUNLVaV+9Z3BFO8Kcclxefz411vp0CnMA//Ti+KdIX7/4wEMPKKSPz69hlcezWXL2mSevLc7T97rTI340zOryc6ti9pXUyprkFhKYKbU0HK61fKbqRsIBCAQcKYoLXg7trI1dc8STXHPC4le4zOmstZwYiKwDbhdVac3dU5HydFRcsD0nZixaaksFgc/VNb6HNlRb5w52tOxN+bNbV8qa01MTLRYLAmN2NTzFoulfeHIS8bviK0XbOCzWCxRoSpE4niOnhds4LNYLFETz5OTvWADn8ViiQonH5/t47NYLO2KxJeXbBeBz9S0k9+t+dJ3m38YeKzvNgFiXsF/qCAGairtrGwV4npyshfaReCzWCz+Ub9WN5Gxgc9isURNPKec8oINfBaLJSqctFS2qWuxWNoZto/PYrG0K5zsLLapa7FY2hHOkjUb+FoNU4pd0DK1qmlnDWHyW06aocqiIC9e34+iTclk967hwvvXkZYVpnB1Cq9NOYyty9IY++sCRl+1Y8/595+UR3JGGAlCIKhc8cpKT77O+GQZlWVBIhFHyOb6c2Ivi3gr2+Yw5a+Jsq0nEFDue2slOwuSuG3SAF9stoXKGrbGd3BEpA/wGNAN50dimqpOjcXmnGdzeOXRXG6autEPF/dQr1Z168UDKCxI4r43VvHJ7Cw2rPKe2+yjh7rS79ulfPua7Xz0YFc+frAr424pIC0rzBm3bWLF3MYTRF76VD7pOeFGP2uKKRMHUeKjiE88l21jmPIX/C/bei64spCNq1JJz4z+fjeGqbL1QqKv3DAZtuuAX6tqHjAauFZE8mIxaEqxyw+1qpVzszjqu4620lHf3bUn0GXk1tHzmEqCcV63jueybQxz6m1myO1Rw8jTSnjzqRzfbLaVylr9qK6Vl2wEVS1Q1c/d16XA10AvU9eLBT/UqsoLk+jQ1cnUnNmljvJCD9oHojw1aSDTzz+cz6PJSqzCH59ezf1vruDsSwqj8rO1STj1NkNle/UdW3j4rh5oxL9g0JZlG9GApy1eaZWfTBHpBxwLLGjks8nAZIBU4ld+MRpEQKT5ZUyXPZdPx+61lBeGeOqygeQO9CYGfeOFg9i5NZmszrX8+ZnVbMxPZemCzFjdtmCmbEeNL6GoMET+knSOHlPmk6dtR73mRiJjPCSLSCbwAvBLVS3Z/3NVnaaqx6vq8UmkmHanUfxQq8rIraV0u/M7Uro9RHrn5nU6Onavdc+tY8gZxWxZ7C3w79zq/MoX70ziwzezGDqsIipfW5NEU28zUbZ5I8oZfUYJMxYs59YH13PMiWVMuW99zHbbqmwVqNOApy1eMeqZiCThBL0nVXWWyWvFgh9qVYePL2HJC07/zZIXcjj89Kb7WmoqAlSXBfa8XvNBB7oc3nyNLyUtTFpGeM/r404pZd2K1hGYaQmJpN5mqmwf/VMPLj0+j0mj8vjTNX1Z/EEmf72+b8x227JsbVP3IIiIANOBr1X1Xj9smlLsaqla1c41qfzj23mcfMNWxly9jRev68eXz3Umq1cNF92/DoCyHSEemXA41WVBRODTR7vws9nfULE7xPNX93evD0ecX8TAU0qbvWanLnXcPn0tAMEgvPtSNgvfi0/lNkgs9TZTZWuKNlNZi3PpSC+YVFk7EXgfWAJE3N2/VdU3DnaOKZU1U9i0VAlIO09L5YfKWqehXXXcI9/zdOysEx5sdyprH0CCT/axWCyN4leNT0R+BVyJ03W4BLgc6AE8A3QGFgE/VtUaEUnBmRt8HLAT+IGqrmvJdeO3EW6xWOKS+kSkXramEJFewC+A41X1SCAIXAz8Bfibqg4CdgNXuKdcAex29//NPa5F2MBnsViiQhHqIgFPmwdCQJqIhIB0oAAYBzzvfj4DuMB9PcF9j/v5ae5YQtTYwGexWKImgnjagFwRWdhgm1xvQ1U3A3cDG3ACXjFO07ZIVevng21i78KHXsBG99w69/gWjWglzpofi8USH2hUfXyFBxvcEJFOOLW4/kARMBM4yw8Xm8MGvhj4w4Bhvtv8Vf5y320C/G3Qt4zYtbQ/fBQbGg+sVdUdACIyCzgByBaRkFur6w1sdo/fDPQBNrlN4yycQY6osU1di8USNX4MbuA0cUeLSLrbV3casBx4F6ifLzMJeNl9/Yr7Hvfzd7SF8/Fsjc9isUSFIoS9DVw0bUd1gYg8D3yOk83pC2Aa8DrwjIjc5e6b7p4yHXhcRPKBXTgjwC3CBj6LxRI1fuXjU9Xbgdv3270GGNnIsVXARD+uawOfxWKJCo1ucCMusYHPYrFEjdrAZ7FY2heJn6Qg4QKfKXGVeBHEeWjkYNI713HZm06WkKqiAK/f0IuSTcl07F3Duf/YTGpWBFV47/91Y+17mSSlRTjjLwV0O9JJa7VsVhaf/jMXgJHXFnLERd7SkcdLGXjB1HNgQmwoKSXCPbPySUpWgiHl/dezefzu7j5421ZiQ4lf4zM2nUVEUkXkUxFZLCLLROSOWG3Wi6v8zyX9uWrsEE6dUMRhg71lLW4Lu3OezeF3l/SP6pwLH9lXPOfTf+XSZ0wFl89bTZ8xFXz2L2ei+rr/ZlC0LpnL561m/F0FvHO780WqKgrwyX25/PCFtfxw1lo+uS+XquLmb3M8lUFzmPK1nikTB/HzM4b6prBWWy1MmTiQa04fwjWnD+H4saUMHV4es13T5XAwVCEcEU9bvGJyHl81ME5VjwGGAWeJyOhYDJoSV4knQZzU7H0VuNa8nUmeW2PLu6iY1XM7ALD67Q5868JiRKDHsVVUlwQo2x5i3fuZ9D2hnNTsCKlZEfqeUM66+c2nTo+nMmiOthLZaTlCVUUQgFCSEkxSXzJZtWU5RLFkLS4xKTakqlovMJDkbjHdblPiKvEsiFNRGCLTFTHK6FJHRaETRMq2hejQwMfM7nWUbQsddH9zxHMZ7I9RXw2JDQUCygNzV/DsV8v4Yn4mK77IiNlmW90zxWnqetniFaN9fCISxFl0PAj4p6q2C7EhU4hgMxwaxpSQUyQi/Pz0IWR0DHP79LX0HVLJ+hVpPnjcFiT+4IbRJWuqGlbVYTjr7UaKyJGNHONZbMiUuEo8C+Kk59ZR5ooYlTUQMcrsVkdpAx/LtobI7FZ30P3NEc9lsD8mfTUt5FReEmTxR5mMOLV5mYHmaMt7pupti1daZa2uqhbhrL+LKfOCKXGVeBbEGXBaGctnOb4sn5XFgPFle/Z//WIWqlDwRSrJHSJkdq2j30llrP8gg6riAFXFAdZ/kEG/k5qXNIznMtgfU76aEhvKyqkjo6NjNzk1wvCTy9iYH7vdtrxntql7EESkC1CrqkUikgacTgwZU8GcuEo8CeI8M7EfVbuD/PuEQYy5YQcjfraT13/Ri2Uzs+nQq5bz/rEJgP5jy1j3XgaPjhtIyJ3OApCaHWHUtYU8daEzkjr6ukJSsyMHvV498VQGbeWrKbGhnG61/GbqBgIBCARg/qtZLHg7drttJTbkjOomdn4Tk2JDR+NkSw3i1CyfU9U7mzon0cSGTPCr/K+N2LVpqVys2FDMYkNpg3pq/7snN38g8PWFd7Q7saGvAEOSYRaLpS2J52asFxJu5YbFYmlblPjuv/OCDXwWiyVqEqdx3zg28FksluhQ0DhejuYFG/gsFkvU2KauxWJpdyTQQHajHDTwich9NNGUV9VfGPGonWNq2skl32wyYvfJob2N2DVGIn1jjUy98cfEoVzjW9hqXlgslsRBgUM18KnqjIbvRSRdVf1duGixWBKSRKo4N0az605EZIyILAe+cd8fIyIPGPfMYrHEKYJGvG3xipcFd38HzsRVLFfVxcDJBn2yWCzxjnrc4hRPo7qqulH27WgNH+xYi8VyiKOH9uBGPRtF5NuAikgScANgZiW9xWJJDOK4NucFL4HvamAq0AvYAswGrjXpVFMkksqaKXWtWHz95rFM8mdmgMKgieUMneTk6lvxeAYrn8pEgtDzlCqG31RMwYcpfHlPFuFaIZikHDulmO6jq1vV36YIBJT73lrJzoIkbps0wBebJux26VnDTVM3kN2lDhTeeKIzL03vErPd3gOr+O2D6/a8735YDY/f3Z0XH+4as+3mOcRrfKpaCFzS0gu46ecXAptV9byW2oG9qlK3XjyAwoIk7ntjFZ/MzmLDqthykJmyW6+uVVURJBhS7n0pn8/e6cA3n7dcbyEWX4tWhsifmcFZz20nkKS8e1UuvcZWUl4QYtM7aZzz8jaCyVC10+n6TekU4ZQHC0nvFqFoZYh3ruzCRfMLWs3f5rjgykI2rkolPdPfnhe/7YbrhGl39iR/STppGWHuf2sln8/vEHMZbFqdys/PGAo45fzkomV8+Ga2Dx57oPkUj3GNl1HdASLyqojsEJHtIvKyiETzM+hb0zjRVNZMqGvF4mvxmiRyj64hlKYEQtB1RDUb5qax6pkM8q4qJejq1qR2dp7qnLxa0rs5r7MG1xGuFsI1B7Puv79NkdujhpGnlfDmUzkx2zJtd9f2JPKXOHoyleVBNuan+i4KNOzEUgrWp7B9c3LzB8dK/Tw+L1uc4mVU9yngOaAH0BOYCTztxbiI9AbOBR5uqYMNSUSVNb/VtWLxNXtwLdsXJlO9O0BdpbDlv6lUFIQoXRdix8IU3vp+V+Ze2oWdSw7Ubdg4O42cvJo9wbE1/G2Kq+/YwsN39fB9yoQpu/V0613DwCMr+eZzf4W1xk4o4r2Xsn212RTtQXMjXVUfV9U6d3sC8FpH/zswhSYqxiIyWUQWisjCWqLvP4p36tW1LjkujyHDKug7pLLNfMkaWEfeVaW8c0Uu71yVS6dv1SJBJRIWqosDnPnsdo6dUsT7v+y8z0NbtCrEF/dkMfKO3W3me0NGjS+hqDC0pxYV73brSU0P8/uH1/HQbT2pKAv6ZjeUFGH0GcXMfy3bN5vNkuDTWQ4a+EQkR0RygDdF5BYR6ScifUVkCvBGc4ZF5Dxgu6ouauq49qKy5pe6Vqy+DvpeBWfP2s4ZT+wguWOEjv3qSO8Wps/plYhA7tG1SACqdzuPRsXWIPOv68yYv+yiw2HR93mZKNu8EeWMPqOEGQuWc+uD6znmxDKm3Lc+Jpsm7QIEQ8rvH17HO7M6+d4PN+LUUvKXpFNU2IqqeD41dUUkW0SeF5FvRORrd8FEjojMFZFV7t9O7rEiIv8QkXwR+UpEhrfU/aZqfItwBiW+D/wMRyXtPeAa4AcebJ8AnC8i64BngHEi8kRLHYXEU1kzoa4Vq6/1AxflW4JsnJtGv/Mq6D2+km2fOj86JWtDRGqdgY2aEuHdn3Vm2K+L6To8ys49n/xtjEf/1INLj89j0qg8/nRNXxZ/kMlfr+8bk02TdkG58Z6NbFyVyqxpsY/m7s/YC3a3ajMXQNTb5oGpwFuqOhQ4Bmc84BZgnqoOBua57wHOBga722TgwZb639Ra3f4tNeqefytwK4CIjAV+o6qXxmIz0VTWTKhrxerr/F90prooQCCkjLitiOSOysCLyvnkd5147TvdCCQpY/68GxFY8WQmpRtCLH2gI0sfcPweN71wz+BHa/h7KHDEyHLGT9zNmuWpPDB3BeAE2c/eiV1pLSUtzPCTS5l6c5+YbXlGBXzoAxWRLJxVYD8BUNUaoEZEJgBj3cNm4FS4bgYmAI+po5D2iVtb7KGq0U01wKPKmisEnkeDvj1VfczzRfYGvians1iVNXPYtFQJiIG0VAsib8esspbSt4/2uPUGT8euv+am9UBhg13TVHUagIgMA6YBy3Fqe4twZoFsVtVs9xgBdqtqtoi8BvxZVT9wP5sH3KyqUWeSanYen4jcjhN983D69s4GPgA8Bz5VfQ8nalsslkMB7wMXhU3IS4aA4cD1qrpARKayt1nrXEZVRTw2mqPAy6ju94DTgK2qejlOZG4duXaLxRKf+DOquwnYpKoL3PfP4wTCbSLSA8D9u939fDPQsE3f290XNV4CX6WqRoA6EenoOtGKHQoWiyWu8GkCs6puxckFMMTddRpOs/cVYJK7bxLwsvv6FeAyd3R3NFDckv498LZWd6GIZAP/xmmDlwEft+RiFovl0MDHxuf1wJMikgysAS7HqZA9JyJXAOtxZpaA09V2DpAPVLjHtggva3V/7r58SETeAjqq6lctvaDFYjkE8CnwqeqXQGN9gAeMcrqjub4kSGlKbOigkwNFZLiqfu6HAxaLJfHwf7ihdWmqxndPE58pMM5nXyxgRlkLc9NOvrXIjELp18fVGbGbUMTzYtc4TkDghaYmMJ/amo5YLJYEIc7X4XrBCopbLJbosYHPYrG0NyTBE5HawGexWKInwWt8XjIwi4hcKiK3ue8PE5GR5l2zWCzxiNfMLPE88utl5cYDwBjgh+77UuCfxjyyWCzxT4KnnvfS1B2lqsNF5AsAVd3tzrJudUypVYEZJTBT6mIzPllGZVmQSMQRsrn+nCHNn9QMN967gVHjSykqDPGzcdHb2/lkhKKXIohAyiChx+0BNvw8TKTC+Ty8C1KPEPrcG2TnYxGK33Q7icJQvRYOfztIMCu6L4oJlTUT98zUcxvrPYuJOK7NecFL4Kt1ldIUQES64FFjyU1CWoojQF7XRJYGT5hSqzKhBGZSXQxgysRBlOz2r4t2zrM5vPJoLjdN3Rj1ubXbld3PRBgwM0ggVdh0c5iS2Uq/6Xv923RTmMxTnMDW+bIAnS9zGhul8yPselKjDnrgvxqaqXtm6rmN5Z7FSjw3Y73gpan7D+BFoKuI/AEnJdUfo7jGqao6LNagB+bUqkwogZlTbjPD0gWZlMYQSDUMWg1ap2gVhBpUaMJlSvlnSoexBwa3kreUjmdGH/RMqKGZumemnttY71mLUWdU18sWr3hZq/ukiCzCWTsnwAWq6otcZCz4qVbVmBLY0OEVcWdzDyr88enVoPD6E51588lcf+y2kKSuQudLA6w6N0wgBTJGC5lj9v6mlr6nZIwUgpn7BrhIpVL2sdL9Zi+/v/tSr4aWnunft8voPXMxpbLW6iR4jc9LItLDcDIhvNpwn6pu8GBfgTluIsF/1Wde3c/+ZJz8+aTi7WEwpVaVKNx44SB2bk0mq3Mtf35mNRvzU1m6ILPN/AmXKKX/VQa9GiSYCZtujlD8RoSsc5yAVjJbyb7gwFpd2ftK+jESdTO3oRra0WPKfPkfWoND6rk91AMf8DrOvyk4qef7AyuAIzyce6KqbhaRrsBcEflGVec3PMANhtPAST3fnEETalUmlMBMKrft3OrUSop3JvHhm1kMHVbRpoGvfIGS1AtCnZwA1mGcULFYyToH6nYrVcuUzLsPrNUVz25ZM7deDW3EactJTlHSO4SZct/6mIWBTN4zkyprbcEh38enqkep6tHu38HASDzm41PVze7f7Tj9hDHO/zOjVmVCCcyUcltKWpi0jPCe18edUsq6FW0r3pPUXahcokQqFVWl4lMlpb8T0ErnKZknCoGUfQNcuFSp+Lzxfr/mMKWGZuqemVZZs0RP1D2jqvq5iIxq7jgRyQACqlrqvj4DuLMFPu7BlFqVCSUwU+pinbrUcfv0tQAEg/DuS9ksfC92ta5bHljP0WPKyMqp44mFy3n8nm7Mfrqzp3PTjhI6niasvSSMhCBliJB9kRPQSuZE6PyTA39fS99VMkcLgbT4metl6p6Zem5juWcxk+A1vmZV1kTkxgZvAzg58Tur6pnNnDcAp5YHToB9SlX/0NQ5VmUNY2mpTKU4smmpEosFOi9mlbXUnn203+Qbmz8QWHHHjYv8mNHhN16e2g4NXtfh9Pm90NxJqroGR5jIYrEcaiR4ja/JwOdOXO6gqr9pJX8sFkucIyT+4EZTqedDqlonIie0pkMWiyUBOFQDH/ApTn/elyLyCjATKK//UFVnGfbNYrHEI3GeecULXvr4UoGdOBob9fP5FLCBz2Jpr8TxcjQvNBX4urojukvZG/DqSfB4b7FYYuFQrvEFgUz2DXj1JPi/HcfEs7JWI5iadnL+8p1G7L6S10rz3A51EusxPYCmAl+BqsY04dhisRyCHOIqa/Ezpd5iscQVh3JTt50vobBYLAflUA18qrqrNR2xWCyJQzwnGfWClZe0WCzRcQj08UWf+tZisbRrJIrNkz2RoIh8ISKvue/7i8gCEckXkWfrxc1EJMV9n+9+3q+l/0NC1vj8VtcypYJlSmXNhN2klAj3zMonKVkJhpT3X8/m8bu7x2w31rJd83gq62emgMJhE6sZeFnVns/yH01l+f9lcOaHu0jppBTMS+Kb+9IRAQnBkbeU0znK6Tam7hkkjiqcJ/yt8d0AfA3U5+n6C/A3VX1GRB4CrgAedP/uVtVBInKxe9wPWnJBo4FPRLKBh4EjcYrqp6rqKYlpU/itrmVCBcuUYpcpu7XVwpSJA6mqCBIMKfe+lM9n73Tgm88zYrIbS9mWrAqyfmYKJz1bTCAJPpncgW6n1JDZN0JlQYAdHyWR1mPvM9BldC3dxxUjAsUrgiy6sQPjXi/y7KtpZbxEUYXzgl+juiLSGzgX+ANwo4gIziqxH7mHzAD+FyfwTXBfAzwP3C8ios3l1msE003dqcBbqjoUJ0VVzCJFJtS1TKhgmVLsMqfeJlRVODoQoSQlmKS+zKWOpWzLVgfpdHQdoTQIhKDziDoK3nbS7i/9Szp5v67Ypz0VytibzjBcKVF/O00q4yWSKpwn1OMGuSKysME2eT9LfwemsHcRXGegSFXrq+qbgF7u617ARgD382L3+KgxVuMTkSzgZOAnAKpaA9Q0dY4XTKhrNcQvFSxTil0mlcACAeX+2Svp2a+GV//TmRVfxFbb259oy7bD4DBfT02npkgIpCjb5yeRfUQdBfOSSO0aIWvogTWngreT+fpv6VTvFEY9VBqVfybLNlFV4RpFoxrVLTxYIlIROQ/YrqqLRGSsP855w2SNrz+wA3jU7bh82E1Bvw8iMrn+16CW6iYNNlTXMsEhpYLVAiIR4eenD+GS4/IYMqyCvkMqfbPdkrLtMDDMoCsr+fjKjnwyuSMdh4aJ1AqrpqUx9PrGfesxvoZxrxcx8v5SvvlHmm/+x4Lp57ZN8F7ja4oTgPNFZB3wDE4TdyqQLSL1lbLewGb39WagDzhp84AsnAQqUWMy8IVw0lo9qKrH4qS0umX/g1R1mqoer6rHJ5HSpMF6da0ZC5Zz64PrOebEMqbct94XZ/1WwTKl2GVSCaye8pIgiz/KZMSp0dWYDkYsZdv3u9Wc8nwxJz5eQnLHCB0GhanYHOS9C7OYOz6bqm0B5n83i6od+44hdj6+jopNQap3e1+AZKpsTT23rfEsHAxRb1tTqOqtqtpbVfsBFwPvqOolwLvA99zDJgEvu69fcd/jfv5OS/r3wGzg2wRsUtUF7vvncQJhizGlrmVCBcuUYpcpu1k5dWR0dJqOyakRhp9cxsZ8PzrJYyvb6p1O4KrYEqDg7RT6TKjmrA92c/rbRZz+dhGp3SKc/EIxqV2UsvWBPf2SRcuDRGqE5Gzv3wtTZZt4qnAe8KfGdzBuxhnoyMfpw5vu7p8OdHb330gjFSmvGOvjU9WtIrJRRIao6gqcJXDLTV0vFkyoYJlS7DJlN6dbLb+ZuoFAAAIBmP9qFgvejl29Lday/eyGDk4fXxIc9T9lJHU8+LepYG4ym15OQUIQTFWOu6c0Ku0mU2Vrirb01++1uqr6HvCe+3oNjUjRqmoVMNGP6zWrshaTcZFhONNZkoE1wOWquvtgx1uVNUs9Ni2VGfxQWUvv0keHXuRNZe2LaYmrstZiVPVLIO7+aYvF0nIOabEhi8ViOSg28FkslvaGJFim8P2xgc9isUTHIZCdxQY+i8USNbaPz2KxtDtsItL2TDSTxLyS4H0nfmFq2smtq7/y3eafBh7tu03A0PMVZ3baCBv4LBZLdHhYjhbv2MBnsViixwY+i8XSnrATmC0WS7tEIokd+Wzgs1gs0WHn8bUuN967gVHjSykqDPGzcUPi3m5Gxzp+dfdG+g2pQhXu/fVhfL0o9qzGiSRiZMpuS23++6zDueqtlQBUFgV56ReHUbwpmazeNVxw3wbSssLsXJ3Cazf3ZtuyNE65cSujrioEoK5aeOLigYRrhEhYGHJWMSf/cluz1zQlZgUw45NlVJYFiUQcfZPrz/Hv+W0KO53lIIjIEODZBrsGALep6t9banPOszm88mguN03dGKt7rWL3mjs3s/Ddjtw1uT+hpAgpabE/LYkmYmTCrl82P36oC/2+XcaYq3fw8UNd+OShLpx681ZSs+o4/bYtrJqzb+qsYLLyoyfWkJwRIVwLj/9gEANPaT5Zqwkxq4ZMmTiIkt2tXIdJ8BqfsUSkqrpCVYep6jDgOKACeDEWm0sXZFJq4AabsJveIcxRo8p562lHXKauNkB5SezXSDQRIxN2/bK56u0sjrrIyZJ21EW7WTnXSeKZkRum59GVBPZLZiwCyRnOj1ekTojUeRM0MiFm1db4kYG5LWktQfHTgNWq6k+e+ASg+2HVFO8M8eu/beCfs1fwy//bQEpa7LKCjQnM+PElSiS7ftksLwyR2dUR88roUkd5YfM/TJEwTD9vMFNH5tH/hFJ6DYtOl8QvMas9qPDHp1dz/5srOPuSQn9sNntNnIn2XrY4pbUC38XA0419EI3YUCIRDMKgoyp47bFcrj1zCFUVAX5w3fa2dstyEERAPFRRAkG44rVVXPfh12xZnM6OFU3rxDTEhJjVjRcO4rqzhvC7Swdw/k8KOXJUmS92m0Mi3rZ4xXjgE5Fk4HxgZmOfRyM2lEgUFiSxoyBpj0TjB69nM+io2FXLEk3EyIRdv2xm5NZRtt2p5ZVtD5He2XuNPLVjhL5jylgzv4On4/0Ws6pn51an5lu8M4kP38xi6DDz8pL18/hsU7dpzgY+V9Xmh78OIXbvSKJwSzK9B1YBMOzEUjasjD2wJ5qIkQm7ftkcfFoJS2Z1AmDJrE4MHt90P2HFziBVJc5XprZKWPtBB3IGemml+C9mBZCSFiYtI7zn9XGnlLJuRStobnht5sZxU7c1hoJ+yEGaudFyywPrOXpMGVk5dTyxcDmP39ON2U/HvpjdlN1//r4XN9+3nlCSsnVDMvfceFjMNhNNxMiE3Vhs7lqbwv0nDOWkG7Yx+urtvHT9YSx+LoesXs50FoCyHSH+c8EgqsuCiMBn/8nlqrdWUrYjiddu6kMkDBoRvnVuEYPHNT+qa0LMCqBTlzpun74WcLpW3n0pm4XvxS4Q5YV4rs15wbTYUAawARigqs0OuyWc2JDNzpJwtPfsLAsib8csNtQhu7cee/INno59/9Up7VJsqBxHF9NisRxCJHqNL6FWblgsljhAgXBiRz4b+CwWS9TYGp/FYml/JHhftA18FoslamyNz2KxtC9sWqp2ToJX99sjfxp0jO82v/91ge82AZ77VncjdmNFALGDGxaLpb0hCf6j31pJCiwWy6GCRrE1gYj0EZF3RWS5iCwTkRvc/TkiMldEVrl/O7n7RUT+ISL5IvKViAxv6b9gA5/FYokS39bq1gG/VtU8YDRwrYjkAbcA81R1MDDPfQ/Ouv/B7jYZeLCl/4ENfBaLJWr8yM6iqgWq+rn7uhT4GugFTABmuIfNAC5wX08AHlOHT4BsEenREv9tH5/FYoke7318uSKysMH7aao6bf+DRKQfcCywAOimqvUjRluBekGVXkBDfYhN7r6oR5ds4LNYLNGhUY3qFjaXpEBEMoEXgF+qaok0SM6gqipeMsRGSUIFPpNqVaYUxgIB5b63VrKzIInbJg2I2V5SSoR7ZuWTlKwEQ8r7r2fz+N2xT3tItLI1db9iVS1b+Vg6a2amgcKAiZUcPqmCpfdnsnZmGik5Tkrio35ZSo9T9iZSLd8SYPZ3csm7toyhP40ukajfz5dnfApFIpKEE/SeVNVZ7u5tItJDVQvcpmx96vLNQJ8Gp/d290WN0cAnIr8CrsQppiXA5apa1VJ7ptSqTCmMAVxwZSEbV6WSnhm73gZAbbUwZeJAqiqCBEPKvS/l89k7Hfjm89hkKxOpbE3eL2i5alnxyhBrZqYx/rmdBJJg/lWd6DHWSVQ6eFL5QYPa4r90pPtJNY1+1hx+P19e8WM6izhVu+nA16p6b4OPXgEmAX92/77cYP91IvIMMAoobtAkjgpjgxsi0gv4BXC8qh4JBHG0N1qMKbUqUwpjuT1qGHlaCW8+lROzrb0IVRWOXkMoSQkmqS/zqBOpbE3dr1gpWROk89G1hNIgEIIuI2rYPLfpYLz57RQyeofpOKgu6uuZeb484s+o7gnAj4FxIvKlu52DE/BOF5FVwHj3PcAbwBogH/g38POWum+6qRsC0kSkFkgHtvhl2E+1qsZUu4YOj1274Oo7tvDwXT1Iz/RXdSUQUO6fvZKe/Wp49T+d9+h6+EW8l62p+wXsUS1D4fUnOvPmk7meT80aXMeSv3egercQTFW2zk+h05G1JGcr+U9msP7lNDodWcuwKaUkZym15cI3D2dw8vTdrHg0+nto6vlqFgV8uKSqfoCzEKQxDshIrE7W5Gtjv7JZXd3NwN04GZgLcKqlc/Y/riUqaybUqvxm1PgSigpDe2pRfhKJCD8/fQiXHJfHkGEV9B0Su4hRPYlQtiaJRbWs48AwQ68sZ/6VOcy/KofsoXVIAAZdXME5c3Zwxos7SesS4cu/OgJFy/6ZyeGTKkjKiL7KbvL5ag5BEfW2xSvGanzubOsJQH+gCJgpIpeq6hMNj3OHtqeBk3q+Obsm1KpMKIHljShn9BkljDhtOckpSnqHMFPuW89fr+8bq7t7KC8JsvijTEacWsr6FWkx20uUsjWlCOfYPlC1bOmCTM/nD/heJQO+5/wQffW3TNK7hUnN3Vs9GjCxkvevzgZg11dJbJqdyuK7O1BbKkgAgikw+JLma6+t8Xw1SSSOtSM9YHIC83hgraruUNVaYBbw7dhMmlGrMqEE9uifenDp8XlMGpXHn67py+IPMn15KLNy6sjo6HRkJ6dGGH5yGRvz/ejUT5yyNaUI54dqWdVO5ytVviXA5rmpHHZeFZXb937NNs1NIWuw05837oldnDdvB+fN28HgyyoYOrnMU9ADc8+XJ+qbul62OMVkH98GYLSIpAOVOG32hU2f0jSm1KpMKYyZIKdbLb+ZuoFAAAIBmP9qFgvejl1ZK5HK1tT98kO17KMbsqkpCiAhZfjvS0juqCy4qyNF34RAIKNXmOP+tyRmX9uaeG7GesG0ytodwA9w1uR9AVypqgftyEs4lTVL4mFAuez7yxMnLdUCnRezylpWek8dM/gKT8fO/uqudqmydjtwu8lrWCyW1ia+xcK9kFArNywWSxxgVdYsFkt7JNH7+Gzgs1gs0WMDn8ViaVcoELGBz2KxtCvs4IYlUTAwjQNIvC+AAX9NqaFdu2ql7zbXXtDi5Ej7kmj3fT9s4LNYLNGhQDiOl2V4wAY+i8USJQpqA5/FYmlv2KauxWJpV9hRXYvF0i6xNT6LxdLusIGv9TClBGZKucyEEpipMug9sIrfPrhuz/vuh9Xw+N3defHhrjHZTaSyrceEclks/i7+TzbLn8tCFY74fjHHXF7EjuUp/Pe2rtRVC4EQnPK/2+l2TBXVpQHe/nV3SguSiNTBsVfs5lvf8zkNliqEW1fcyG9Mq6zdAFyFk1f/36r691jsmVICM6FcZkoJzFQZbFqdys/PGLrH9ycXLfMlC3MilW09fiuXxeLvzpXJLH8ui++9sIFgkvLqFb3oO66cj/+ay4jrd9L3lArWvZfBR3/N5cInN7HkiWw6Darh3GlbqNwZ5Mkz+3H4+SUEk5u9VHQkeI3PpMrakThBbyRwDHCeiAyKxaYpJTATymWmlMDMlcFehp1YSsH6FLZv9uPbkjhlC2aUy2Lxd/fqZLodU0VSmhIIQc8RlayZnQkCNWXO17emNEBGVyers4hSWx5AFWorhJSsMAET1Rt/VNbaDJOp578FLFDVClWtA/4LXOSXcT+VwMD5VX5g7gqe/WoZX8zPjFm5rDElML8DlN9lUM/YCUW891K2b/YSqWzrlcs04t9Kl1j8zRlcw5aFaVTtDlBbKaz/bwZlW0Oc+LsdfPSXLsw4qT8f/aULo39TCMBRlxaxe3Uy/zlhAE+f14+T/mcH4vu3XJ1RXS9bnGIy8C0FThKRzm76+XPYVwUdiB+VNZPKZSYwpYYWSoow+oxi5r+W7ZvNRCnbtlQuOxg5g2oYPnkXr1zem1d/2ovcb1UjAVj6VBYn/nYHk95fywm/3c67v3X6DDe8n0Hut6r5yYdr+MEr65l/Z1dqSn3+miuoRjxt8YpJecmvgb8Ac4C3gC+BAzpNVHWaqh6vqscnkdKsXRNKYA1pqFwWCyaVwEyWwYhTS8lfkk5RoT++NiTey7ZeuWzGguXc+uB6jjmxjCn3rY/Zbqz+5k0s4fsvbeCipzeRkhUmu38NK17syIAzHenLQWeXsW2x01/4zQsdGXBGGSKQ3beWjr1r2b3G7w4+nCVrXrY4xWSND1WdrqrHqerJwG4gxlXXZpTATCiXmVICM1UG9Yy9YLevzdxEKltTymWx+lux06nRl24JsWZOBw7/TikZXevY8qkjKbrp4zSy+zlN58yedWz62KmxVhQGKVqbTMc+NY0bbimqjrykly1OMT2q21VVt4vIYTj9e6NjsWdKCcyEcpkpJTBTZQCOpOLwk0uZevMBPRItJpHK1hSx+vvWdT2o2h0kkAQn376NlI4Rxv5hGx/c1ZVIWAgmRxh71zYARly7k3k3d+fpc/uCwpibdpCWYyAAxfHAhRdMq6y9D3QGaoEbVXVeU8dblTWD2LRUCYeJtFS/uWAV+UsqYlNZC+bq6LRzPR07p/yxdqmydpJJ+xaLpS2I76kqXkiolRsWiyUOsEkKLBZLe0MBTfAla0ZHdS0WyyGIuolIvWzNICJnicgKEckXkVtawXvA1vgsFksLUB+auiISBP4JnA5sAj4TkVdUdXnMxpvB1vgsFkv0+FPjGwnkq+oaVa0BngEmGPcdw9NZokVEdgBepsrnAoUGXLB2E8vXRLMbD772VdWYZr6LyFvuNb2QCjSUdpumqtNcO98DzlLVK933PwZGqep1sfjnhbhq6nq9ISKy0MTcIGs3sXxNNLuJ5GtTqOpZrXUtU9imrsViaSs2s2/ikt7uPuPYwGexWNqKz4DBItJfRJKBi4FXWuPCcdXUjYJp1q4xu4nka6LZTSRfjaOqdSJyHTAbCAKPqOqy1rh2XA1uWCwWS2tgm7oWi6XdYQOfxWJpdyRc4DOxxEVEHhGR7SKy1A97rs0+IvKuiCwXkWWu4pwfdlNF5FMRWezavcMPuw3sB0XkCxF5zUeb60RkiYh8KSILfbKZLSLPi8g3IvK1iIzxweYQ18f6rUREfumDu4jIr9z7tVREnhYRXxIIisgNrs1lfvnaLlDVhNlwOkBXAwOAZGAxkOeD3ZOB4cBSH33tAQx3X3fAyT7th68CZLqvk4AFwGgf/b4ReAp4zUeb64Bcn5+FGcCV7utkINvAs7YVZ8JvrLZ6AWuBNPf9c8BPfLB7JI62TTrOQOXbwCA/y+FQ3RKtxmdkiYuqzgd2xWpnP5sFqvq5+7oU+BrnCxCrXVXVMvdtkrv5MkIlIr2Bc4GH/bBnChHJwvmxmg6gqjWqWuTzZU4DVqtq7KIbDiEgTURCOIFqiw82jSoZHsokWuDrBWxs8H4TPgQT04hIP+BYnNqZH/aCIvIlsB2Yq6q+2AX+DkwB/M5VrsAcEVkkIpN9sNcf2AE86jbLHxaR2DQrD+Ri4Gk/DKnqZuBuYANQABSr6hwfTHtSMrQcSKIFvoRDRDKBF4BfqmqJHzZVNayqw3Bmuo90xdtjQkTOA7ar6qJYbTXCiao6HDgbuFZETo7RXgina+JBVT0WKAd8S2nkTqY9H5jpk71OOC2T/kBPIENELo3VrnpUMrQcSKIFvjZb4tISRCQJJ+g9qaqz/LbvNu/eBfxYO3kCcL6IrMPpQhgnIk/4YLe+xoOqbgdexOmyiIVNwKYGNd3ncQKhX5wNfK6q23yyNx5Yq6o7VLUWmAV82w/D6ruSYfsg0QJfmy1xiRYREZw+qK9V9V4f7XYRkWz3dRpOLrNvYrWrqreqam9V7YdTru+oasy1EhHJEJEO9a+BM3CaaLH4uhXYKCJD3F2nAX7mcPshPjVzXTYAo0Uk3X0uTsPp840ZEenq/q1XMnzKD7uHOgm1ZE0NLXERkaeBsUCuiGwCblfV6TGaPQH4MbDE7Y8D+K2qvhGj3R7ADDeJYwB4TlV9m3pigG7Ai873nRDwlKq+5YPd64En3R/ANcDlPtisD86nAz/zwx6Aqi4QkeeBz4E64Av8W2b2gojUKxlea2CQ55DELlmzWCztjkRr6losFkvM2MBnsVjaHTbwWSyWdocNfBaLpd1hA5/FYml32MCXQIhI2M0aslREZrrLlFpq6z+uyhXukq+8Jo4dKyJRT7h1s7IcoMZ1sP37HVPW1OeNHP+/IvKbaH20tE9s4EssKlV1mKoeCdQAVzf80F0AHzWqeqU2LeI8Fp9WGlgs8YANfInL+8Agtzb2voi8Aix3Exj8n4h8JiJficjPwFlJIiL3u7kM3wa61hsSkfdE5Hj39Vki8rmb72+em2DhauBXbm3zJHf1yAvuNT4TkRPcczuLyBw3N9zDOCm0mkREXnKTFyzbP4GBiPzN3T9PRLq4+waKyFvuOe+LyFBfStPSrkiolRsWB7dmdzbOwnRw1qkeqapr3eBRrKojRCQF+FBE5uBkhxkC5OGsplgOPLKf3S7Av4GTXVs5qrpLRB4CylT1bve4p4C/qeoH7lKp2Tgpkm4HPlDVO0XkXOAKD//OT91rpAGficgLqroTyAAWquqvROQ21/Z1OCserlbVVSIyCngAGNeCYrS0Y2zgSyzSGix/ex9nLfC3gU9Vda27/wzg6Pr+OyALGIyTv+5pVQ0DW0TknUbsjwbm19tS1YPlKBwP5LnL0AA6ulloTsbNB6eqr4vIbg//0y9E5EL3dR/X1504qbGedfc/Acxyr/FtYGaDa6d4uIbFsg828CUWlW46qj24AaC84S7gelWdvd9x5/joRwAn63NVI754RkTG4gTRMapaISLvAQdLya7udYv2LwOLJVpsH9+hx2zgGjclFiJyuLvwfj7wA7cPsAdwaiPnfgKcLCL93XNz3P2lOOnz65mDkyQA97hh7sv5wI/cfWcDnZrxNQvY7Qa9oTg1znoCQH2t9Uc4TegSYK2ITHSvISJyTDPXsFgOwAa+Q4+HcfrvPhdHPOlfODX7F4FV7mePAR/vf6Kq7gAm4zQrF7O3qfkqcGH94AbwC+B4d/BkOXtHl+/ACZzLcJq8G5rx9S0gJCJfA3/GCbz1lOMkWV2K04d3p7v/EuAK179l+CA9YGl/2OwsFoul3WFrfBaLpd1hA5/FYml32MBnsVjaHTbwWSyWdocNfBaLpd1hA5/FYml32MBnsVjaHf8fIhXanux/LN4AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAT4AAAEGCAYAAAD8EfnwAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAAA+KElEQVR4nO2deXxU5fX/32dmsq+EsIcdCkVEpChYlyIo7nVpbW21pa1KtWqttrVulZ92sZu2VquWL1atdVe0qKxuRdtKRVwQZAlb2EJIIGRfZub8/rg3gAjJncx9khnyvF+v+8rcO/eee+aZOyfPej6iqlgsFktXItDZDlgsFktHYwOfxWLpctjAZ7FYuhw28Fksli6HDXwWi6XLEepsB/Ynp1uKFvZL891uxYpU321aLMlIA7U0aaPEY+O0k7O0YlfE07nvfdS4QFVPj+d+JkiowFfYL40Zs4/03e7jI4t8t2mxJCNL9LW4bZTvirBkgbffVEqfdYVx39AACRX4LBZLMqBENNrZTsSFDXwWiyUmFIiS3AsfbOCzWCwxE8XW+CwWSxdCUZptU9disXQlFIjYpq4ZVv09m+Jns0Bh2IW1jJxWw1vXFVC9wXG5qSpAam6UM18sA+Djv+aw7vksJKCMv6WSvic2er7X9XeXMOGUairLQ3x/8gjfPoMpu+MnVXHFL7YRDCjznizgmft6xW0z2coAIBBQ7p2/hortKdw2bYgvNk34a7IMTDwLXkj2Pj6jE5hF5HQRWS0ixSJyo9frKteEKH42i9OfKePMF3ew9c10qjcFOfGPuzjzxTLOfLGM/lPr6X9qPQB7ikNsmpvB2S+XMnlWOe/e0Y2ot2lGACx8uoBbLh4c8+frDLuBgHLVr7dy68WDuXzSCE4+t5IBwxvitptMZdDCeZeVs3ltuq82TfhrqgxMPQttoUBE1dOWqBgLfCISBP4CnAGMAr4hIqO8XLtnfQqFY5oIZSiBEPQ8ppGSRRl731eFkvkZDDzLCXybX8tg4Jn1BFMhuyhCzoAwFR95n7T88ZJsqnf7X/k1YXfE0XVs25hKaUka4eYAb/4zn+NO2xO33WQqA4DCPk0cO6WKeU8U+GrXhL+mysDUs+CFqMctUTFZ4zsWKFbV9araBDwFnOvlwvzhzZQtTaVxd4BwvbDtX+nUbd/34JQtTSW9e5TcQWEA6ncEyeqzr4qX2TtC/Y6gn58lYejeu5md2/YF9fLtKRT2ae5EjzqHK27fxqxf9kGjcS1CSGo661lQlIjHLVEx2cfXD9i83/4WYMKBJ4nIdGA6QPe+zpeYNzTMqMuref3SQoKZSrfPNyPBfYW46ZVMBp1VZ9B1SyIz4ZQqKstDFC/PZMxxNZ3tTpdDFZoTN6Z5otMHN1R1JjATYPDo7L3FOeyrdQz7qhPcPrg7l8zeTo0uGobNizI44/myvTYyekWo3b6vhldXGiSjVwydfElERWkKPfo27d0v7NNM+faUTvSo4xl1TC0Tp1ZxzJSVpKYpmTkRbrh3E7+7ZmBnu9ahdN6zIERI7pq2yabuVqD/fvtF7jFPNFQ4rtVuC7J5UQaDznaCYOl/08gdHN4bCAGKJtezaW4GkSao2RKkelOI7mOaDmo32Vn9QSb9BjfRq38joZQok86t5J2FeZ3tVofy8J19uGT8KKZNGMWdVw7kw7ezu1zQg857FhSIqrctUTFZ43sXGC4ig3EC3kXAN71evPiH3WmsDBAIKcfcVklqrlOKm17JZODZn27m5g8PM/CMel4+qzcSVMbfVkkghi6+G+/fxJjjasgrCPOPpSt57K5eLHiyu3cDHWg3GhH+cks/fv3EegJBWPhUAZvWxD+ymUxlYBIT/poqA1PPgheSvcYnJsWGRORM4E9AEPibqv6qtfMHj85Wm53FYjHHEn2NKt0VV9Q6YkyqPvVKT0/njhmw9T1VHR/P/UxgtI9PVecCc03ew2KxdCwKNGty5zDu9MENi8WSXChCJMmTt9vAZ7FYYiaqyd3HZwOfxWKJCSdJgQ18FoulSyFEbB+fxWLpSjgZmG3g842KFalGpp4s2PaB7zYBTus71ohdiyWRURWa1J+18CLyN+BsoExVR7vHCoCngUHARuBrqrpbRAS4BzgTqAO+o6rL3GumAbe6Zn+pqo+2dt/kDtsWi6VTiCKeNg88AhwoP3kj8JqqDgdec/fByfQ03N2mAw/A3kA5AycXwLHADBHp1tpNbeCzWCwx4QxuBDxtbdpSXQzsOuDwuUBLje1R4Lz9jv9dHd4B8kWkD3AasEhVd6nqbmARnw2mnyKhmroWiyUZiGlwo1BElu63P9NNTNIavVR1u/u6FGhJK32wjE/9Wjl+SGzgs1gsMRHj4EZ5PEvWVFVFxPd1tbapa7FYYiai4mlrJzvcJizu35YcdIfK+BRzJigb+CwWS0woQrOGPG3tZA4wzX09Dfjnfse/LQ4TgT1uk3gBMFVEurmDGlPdY4ck6Zq67VGV+tqRR5BfGGbmG6sBWPxSHo/d1ZvNa9P589w1fO4oR7vjvX9l87df9yXcLIRSlMt/vo2xJzgZft/8Zz5P/bkXkYiTAfiyW7cf8n4APfo28dN7SsjvEQaFuf/ozosP9Yjz0zuYUtZKJvW2ZCqDlLQod80uJiVVCYaUt17J57E/9PbB285RWWsZ3PADEXkSmITTF7gFZ3T2N8AzInIpsAn4mnv6XJypLMU401m+C6Cqu0TkFzip8ADuUNUDB0w+hbHAd7D5OfHSoip100VDKN+ewr1z1/LOgjxK2lDa+tXj6/n9tQP27g8a2cBtszby55/1/9R5eQUR7nh0Pd17h9m4Kp2bvzmEJ5atpGpXkFm/6Mt9C1aT3z3C768dwPtvZbd6z0hYmHlHX4qXZ5KRFeG++WtYtjinTV/bor1l0Fl2Fz5dwJyHC/npPZvbPtkjyVYGzY3CDRcOpaEuSDCk3P1iMe++nsOqZVkJ6W9bKHE1Yz9tS/Ubh3hrykHOVeCqQ9j5G/A3r/c12dR9hDaGlGOlvapSOd0+nYZ+wPBG+g/7rO7usCPr6d7bETAaOKKBxoYATY3C9pJU+g1pJL+7Y+foE6t5e25+q/fcVZZC8fJMAOprg2wuTvdFCMaUslYyqbclWxmA0FDnTPgNpSjBFMWPNJidq7IW8LQlKsY8O8T8nLjoSFWpt1/JY9joelLTlL6DmtiyLo3SzalEwvCf+Xns3Opd26BXURNDR9ezallm3H6ZKoNkUm9LxjIIBJT7F63m6Y9W8P7ibFa/H19tDzpRZU0hogFPW6LS6X18+6uspRN/YPCDjavTeehXffn1k+sAyMmPcM2dW/j1FQMJBODz42vZvjHNk630zAg/n7WRB2/rS13N4Sl5aWmbaFT4wakjyMqNMOOhDQwcUc+m1RltX5iAOIMbyf0sd3rg219lLVcKWm0AdISq1M5tKdxx6SB+ek8JfQftu9fEqVVMnFoFOAMVwYACrQu7BEPKz2dt5PXZ3fj3vHxf/DNVBsmk3pbMZVBbFeTD/2RzzMnVcQe+zvzOkj0RaVJ5b1pVqmZPkJ9/ewjfu3k7Rxxb+6n3Ksud/xHVlUFeeqSQ07/ZViteuf6uzWxem87smf6M5oK5Mkgm9bZkK4O8gjBZuU7/cGp6lHEn1bC5OP4BiM5TWROi6m1LVDq9xhcL7VWVuu6c4ezZFeLiL4ziWz8uJadbhPtv7ceeihA//9YQhh5Rz6+fXM+chwvZtiGVx+/uzeN3O9MN7nxqHfmFYR74eT/Wr3T+Q198XSlFQz87OLI/RxxbyykX7mb9ynTuX+RMo3n4zj68+3pup5RBZ9lNJqU5U3YLejXzk3tKCAQgEHCmUy15Nb7nwKS/Xkj2Gp8xlbX95+cAO4AZqvpQa9fkSoFOkM+MYseNTUtlsTj4obLWf3SuXv/sRE/nXj9qUddSWWtlfo7FYklqxKaet1gsXQtHXtKO6losli6EqhBN4Dl6XrCBz2KxxEwiT072gg18FoslJpx8fLaPz2KxdCmsvGRSYGrayU3rPvLd5p1Dx/hu02IYMVT7MTTVLF4UEnpyshe6ROCzWCz+YdfqWiyWLkkip5zygg18FoslJpy0VLapa7FYuhi2j89isXQpnOwstqlrsVi6EM6SNRv4OgxTil3QPrWq/zv9c1w+fw0A9ZVBXvzhAPZsSSWvqInz7i0hIy9Cxbo0Xv5ZETtWZPCl60uZcHn53usbqgLMvamInWvSEYEzf7PFk68myiHZVOFMPguBgHLv/DVUbE/htmlD4rZXNLSBmx/YuHe/94AmHvtDb16Y1TNu252hssZhUOMz5r2I9BeRN0RkpYisEJFr47W58OkCbrl4sB/ufYoWtapbLx7M5ZNGcPK5lQwY3hCTjf8+2INBX6zhitdXM+iLNbzzoBM00vPCnHrbNiZcuvMz1yy6oy9DTqrh+4vWcOnLaykc5u2eJsqhRRVu+qSRXHv2cM75TnnMZXAw/Cjbg2HqWQA477JyNvuoVLZlXTo/mDqSH0wdydWnj6CxPuBLRm5TZeuFKOJpS1RMhu0w8GNVHQVMBK4SkVHxGDSh2AX+qFWtfTWPIy/YDcCRF+xmzSInE25WYYS+Y+oJHJARvKE6wOZ3sznqa04m52Cqkp4b9XQvE+WQbKpwpp6Fwj5NHDulinlPFPhuG2DsCdVs35RG2dbUtk9ug85SWWsZ1fWyJSomVda2q+oy93U18AnQz9T94sEPtara8hDZPR1pyqweYWrLW/9R7tmcSmZBmFduKOJv5wxn7k1FNNUlxoOSDKpwprji9m3M+mUfNGrmu5h0biVvvpjvi63OLNuoBjxtiUqHeCYig4CjgSUHeW+6iCwVkaXNtJ7OPVkQAZHWlxtFw0LpigyOvriC7720lpSMKP99MP4+n3jpyqpwE06porI8tLfm6zehlCgTp+5h8cv5Rux3FFZzwwMikg08D/xIVasOfD8WlTVT+KFWlVUYpqbMqfXVlIXI7B5p9fycPs3k9m6m39h6AEaeUdnpgS+ZVOFMMOqYWiZOreKYKStJTVMycyLccO8mfnfNQF/sH3NyNcXLM6ks9+fzd1bZKhBO4NqcF4x6LyIpOEHvcVWdbfJe8eCHWtXwKVUsn90NgOWzuzH8lNb7WrJ7hMnp00zFekefd+N/cigc1pk13uRShTPBw3f24ZLxo5g2YRR3XjmQD9/O9i3oAUw6b7dvzVzo3LJN9qausRqfiAjwEPCJqt7th00Til3QfrWqXRvSuO/4kZx47Q4mXlHGi9cM4MNnCsjr50xnAajZGeKR84bRWBNEBN59pJDL568hLSfK1BlbmXNdfyLNQn7/Js763RaWzGo76Jgoh2RThTP1LJgiLSPCuJOquedn/X2z2WkqawnejPWCSZW1E4C3gOVAy3Dlzao691DXmFJZM4VNS2UBkiotlR8qa91G9tTJf/uqp3NnH/9Al1NZexsSeCKPxWJpN8le40vcRrjFYklIWhKR+jGqKyLXuQscPhaRJ0UkXUQGi8gSESkWkadFJNU9N83dL3bfH9Tez2ADn8ViiQlFCEcDnrbWEJF+wA+B8ao6GggCFwG/Bf6oqsOA3cCl7iWXArvd4390z2sXNvBZLJaY8XHJWgjIEJEQkAlsByYDz7nvPwqc574+193HfX+KO4gaMzbwWSyW2NCYmrqFLQsU3G36XjOqW4E/ACU4AW8P8B5Qqaph97Qt7Fvx1Q/Y7F4bds9v11B+UmVnSTRMjMBetXaN7zYB/jL8c0bsWroeMYoNlR9qVFdEuuHU4gYDlcCzwOk+uNgmNvBZLJaY8WlU9xRgg6ruBBCR2cDxQL6IhNxaXRGw1T1/K9Af2OI2jfOAivbc2DZ1LRZLTChCJBrwtLVBCTBRRDLdvropwErgDaBlouA04J/u6znuPu77r2s7JyLbGp/FYokZP3LtqeoSEXkOWIaTxu59nHX7rwBPicgv3WMPuZc8BDwmIsXALpwR4HZhA5/FYokJVf8mMKvqDGDGAYfXA8ce5NwG4EI/7msDn8ViiRlN8pUbNvBZLJYYSf4kBUkX+EyJq/ht1w/xng8fyWflM3mowhFf28NR362k/JNU3rytF811AXL7NXPqXaWk5kSp2hLiidMHkT/Yyc/We2wDk35RFtP9TJRtSlqUu2YXk5KqBEPKW6/k89gfesdt14SvJgWXHn1nBfU1QaJRR9/kmjPjF0gyVbZesDW+QyAi6cBiIM29z3Nue77dtIir3HTREMq3p3Dv3LW8syCPkjiFYUzYbRHvKV6eSUZWhPvmr2HZ4hzPNivWpLLymTy++nwJwRTlpUv7MXByLW/c0psv/mwn/SbUs/LZXN6f1Y0J1zkj+nkDmrnopZJ2+WuqbJsbhRsuHEpDXZBgSLn7xWLefT2HVcuy2m3TlK/xfmdtccOFw6jyUSfERNl6QRUihlLzdxQmp7M0ApNV9ShgLHC6iEyMx6ApcRUTduMV79m9LpVeRzWQkqEEQtD3mHrWL8imckMKfY91sjb3P6GOdQuy4/KzBXPCNUJDnZPCPpSiBFM07mxLpnw1JbhkDv/L1itWZe0QqEONu5vibnF9LabEVUyLtrRHvKdgeBPblmbQsDtAc72w6V9Z1JSGKBjexIZXnf/o6+ZlU1O6L9V41ZYUnv7yAF74ZhHb3s2IyUeTZRAIKPcvWs3TH63g/cXZrH4/vhpJR4js+Cm4BIAKv35yHffNW80ZF5e3fb5H/C5bLyhOU9fLlqgY7eMTkSDO2rthwF9U9aBiQ8B0gHTMiLx0Ju0V7ykY1sS46buY890iQhlRCj/fiARg8p2lvPWLniz9S3cGTakhkOL8L8nqEWHav9aT3i1K2cdpzLuyL9+Yu4nUHG+SlSaJRoUfnDqCrNwIMx7awMAR9WxaHVtg7khMCC5df/4wKkpTyevezG+eWsfm4nQ+XhJ/bb1zyjb5BzeMrtxQ1YiqjsVZdnKsiIw+yDkzVXW8qo5PIa1Ve6bEVUzZjVe8Z9SFVXztxRIueHILaXkR8gc30W1oM19+ZCtfe7GEz51dTd4Ap6YTTFPSuzlBrufoRnIHNFO50ftn6AjhmtqqIB/+J5tjTq6Oy45JX00ILgFUlDo11D0VKfx7Xh4jx9b5Zhv8K1uvqHrbEpUOWbKmqpU4y1DiWoBsSlzFjN34xXvqKpzaRvW2EOsX5vC5c6r3HtMoLL2/O0dcVAlAfUWQqCvstqckhT2bUsnt7735Z6ps8wrCZOU6jqWmRxl3Ug2bi+MbLDAnsmNGcCktI0JGVmTv6y98qZqNq+MfMDFRtl6xTd1DICI9gGZVrRSRDOBU4kgcCObEVUzY9UO8Z/7VfWjYHSSQAifN2EFabpQPH8ln+eP5AAydWsPnv+oodm57N4Ml93QnEFIkAF+6fQfp+d6buabKtqBXMz+5p4RAAAIBWPxSHkteTUwBI1OCS916hJnx0AYAgkF448V8lr4Zn00wU7ZecEZ1k3uZv0mxoTE4SQODODXLZ1T1jtauSTaxIRPYtFRJSBcTG8oY1lcH/2F62ycCn5x/e5cTG/oIONqUfYvF0nkkcjPWC0m3csNisXQuSmL333nBBj6LxRIzCTxg6wkb+CwWS2woaJIvWbOBz2KxxIxt6losli5HIk9O9sIhA5+I3EsrTXlV/aERj7o4pqadXLxqixG7j48sMmI3qTAVBUxMk/HB1Za1uslMazW+pR3mhcViSR4UOFwDn6o+uv++iGSqqr8LDC0WS1KS7E3dNtediMhxIrISWOXuHyUi9xv3zGKxJCiCRr1tiYqXBXd/Ak7DFe5V1Q+Bkwz6ZLFYEh31uCUonkZ1VXWzfLqjNWLGHYvFkvDo4T240cJmEfkioCKSAlwLfGLWLYvFktAkcG3OC14C3xXAPUA/YBuwALjKpFOH4vq7S5hwSjWV5SG+Pzl+lSrTdhNVEW7V37MpfjYLFIZdWMvIaTW8dV0B1Rucx6GpKkBqbpQzX3RU2j7+aw7rns9CAsr4Wyrpe2Jjh/p7KAIB5d75a6jYnsJt04b4YtOEr6aeLzCj3uaNw7zGp6rlwMXtvYGbfn4psFVVz26vHYCFTxcw5+FCfnrP5njMdIjdRFWEq1wTovjZLE5/poxAivLG5YX0m1TPiX/ctfec936Ttzdl/Z7iEJvmZnD2y6XUlwV57bs9OGd+KQGPGdlNlQPAeZeVs3ltOpnZ/vS8mPLV1HPbgt/qbZ7ofEWDuPAyqjtERF4SkZ0iUiYi/xSRWP69+tY0/nhJNtUGvmATdhNVEW7P+hQKxzQRctXbeh7TSMmifRoNqlAyP4OBZzlKbptfy2DgmfUEUyG7KELOgDAVH6Ueyrzv/h6Kwj5NHDulinlPFMRtqwVTvpp6bjuNlnl8XrYExcuo7hPAM0AfoC/wLPCkF+MiUgScBcxqr4PJSqIqwuUPb6ZsaSqNuwOE64Vt/0qnbvu+H2XZ0lTSu0fJHRQGoH5HkKw++2pUmb0j1O/wLsBjqhyuuH0bs37Zx9cpEx2h3uY7htTb2rxtkmtuePk3lKmqj+23/w8R+alH+38CbgByDnXC4a6ylmjkDQ0z6vJqXr+0kGCm0u3zzUhw3xO66ZVMBp2V2PPUJ5xSRWV5iOLlmYw5rqbtCw5jTKm3tUkCBzUvHLLGJyIFIlIAzBORG0VkkIgMFJEbgLltGRaRs4EyVX2vtfNiUVlLJhJZEW7YV+s4Y3YZU/+xk9TcfbW7aBg2L3Kati1k9IpQu31fDa+uNEhGL+99aibKYdQxtUycWsWjS1Zy0wObOOqEGm64d1NcNqFjlOb8xrR62yHxqakrIvki8pyIrBKRT9wFEwUiskhE1rp/u7nnioj8WUSKReQjERnXXvdba+q+hzMo8TXg+zgqaW8CVwJf92D7eODLIrIReAqYLCL/aK+jyUYiK8I1VDhfe+22IJsXZTDobOfHUvrfNHIHh8nsvS+wFU2uZ9PcDCJNULMlSPWmEN3HNB3Uril/D+ThO/twyfhRTJswijuvHMiHb2fzu2sGxmXTlK8mMaXe5gVRb5sH7gHmq+pI4Cic8YAbgddUdTjwmrsPcAYw3N2mAw+01//W1uoObq9R9/qbgJsARGQS8BNVvSQemzfev4kxx9WQVxDmH0tX8thdvVjwZPd4TBqzm8iKcIt/2J3GygCBkHLMbZWk5jpP6KZXMhl49qdrDPnDwww8o56Xz+qNBJXxt1V6HtH1y9+OwpSvpp5bU+ptbaICPvStikgeziqw7wCoahPQJCLnApPc0x7FqXD9DDgX+Ls6CmnvuLXFPqq6PeZ7e1FZc4XARwF7nwJV/bvnm+wLfK1OZ7Eqa+awaamSEANpqZZEX41bZS1tYH/tc9O1ns7ddOVPNwH7j7rMVNWZACIyFpgJrMSp7b2HMwtkq6rmu+cIsFtV80XkZeA3qvq2+95rwM9UNeZMUm0ObojIDJzoOwqnb+8M4G3Ac+BT1TdxorbFYjkc8D64Ud6KvGQIGAdco6pLROQe9jVrnduoqojHRnMMeJnO8lVgClCqqt/FicyJ2/FhsVjM40+Sgi3AFlVd4u4/hxMId4hIHwD3b5n7/lag/37XF7nHYsZL4KtX1SgQFpFc14n+bVxjsVgOV3yawKyqpTi5AFrW2U3BafbOAaa5x6YB/3RfzwG+7Y7uTgT2tKd/D7zN41sqIvnA/+G0wWuA/7bnZhaL5fDAx8bnNcDjIpIKrAe+i1Mhe0ZELgU24cwsAaer7UygGKhzz20XXtbq/sB9+aCIzAdyVfWj9t7QYrEcBvgU+FT1A+BgfYCfGeV0R3N9SZDSmtjQIScHisg4VV3mhwMWiyX58H+4oWNprcZ3VyvvKTDZZ18sYEZZC3PTTkYsNbOyYfX4BF8j2xEk8mLXBE5A4IXWJjCf3JGOWCyWJCHB08p74TDKlWOxWDoMG/gsFktXQ5I8EakNfBaLJXaSvMbnJQOziMglInKbuz9ARI4175rFYklEvGZmSeSRXy8rN+4HjgO+4e5XA38x5pHFYkl8kjz1vJem7gRVHSci7wOo6m53lnWH06NvEz+9p4T8HmFQmPuP7rz4UA9fbJtQ1zKlLpaVG+a6P2xm0IgGVOHuHw/gk/ey4rIZrxLYrscj7Pmn0/GTNkzoPSPIlqsiROucf/vhXZBxhNDvLueRq1sapezuCBqGYL4wYGbsvS7JorJm6rk1qd7WJglcm/OCl6et2VVKUwAR6YFHjSU3CWk1jgB5uJUsDZ6IhIWZd/SleHkmGVkR7pu/hmWLczpduayjbLZw5R1bWfpGLr+cPphQSpS0jPh7muNRAmsuUyqfjjLomRCBdGHbjWGqFyoDZu17vLb+NEz2l5wGRqRa2fHbCEX3hkjpLYR3te9XlCwqa6aeW9Pqba2RyM1YL3hp6v4ZeAHoKSK/wklJ9esY7nGyqo6NN+gB7CpLoXi5o8tRXxtkc3G6L2IwJtS1TCl2ZeZEOHJCLfOfdNTFws0BaqviH6OKVwlMI6CNoGEl2gCh/So0kRqlbqmSPclp+lTNj5J9coCU3s5+qCD2JlEyqayZem47Tb1NnVFdL1ui4mWt7uMi8h7O2jkBzlNVX+Qi46FXURNDR9ezaln8AkUHU9caOS4+7QITNgF6D2hkT0WIH/+xhCGjGlj7UQYP3NaPxvoY0iL7TEpPoeCSAOvODhNIg8yJQtbEff9Ta95UMo8RgtlOgGsuAQ1DyfQw0Tql20VB8s728j94Hy0qa5nZ/v26TH1n++Pnc9upHO41PhEZgJMJ4SWctDC17jEvKLBQRN5z1dQOZn+6iCwVkaXNNHoymp4Z4eezNvLgbX2pq+m8H3xnEAzCsCPrePnvhVx12gga6gJ8/eqyti80SKRKqfmXMmROiKHzQ2g97Jm7LyBVL4ySe9q+R03D0PCJUnRPkKL7QlQ8FKFpk/df0v4qa8nEYfXc+pOPr9PwUk9+BecjCE7q+cHAauAID9eeoKpbRaQnsEhEVqnq4v1PcNNQzwQn9XxbBoMh5eezNvL67G78e16+BxfaxoS6linFrvLtKezcnsLq953BjLdfyedrnRz46v6npPSFUDenRpd9coCGj5S8MyFcqdSvUPr+fl9zNtQLsvKFQIYQyIDMo4XGtUrqQG9N3haVtWOmrCQ1TcnMiXDDvZviFhwyqbJm4rntTA77Pj5VPVJVx7h/hwPH4jEfn6pudf+W4fQTxjn/T7n+rs1sXpvO7Jn+jOaCGXUtU4pdu3emUL4tlaKhDQCMPaGakjWdK8sZ6g31HyvRBkVVqXs3Suog572aV6NknyAE0vYFtewvBaj/QN3+QKX+YyV1kPd+vuRTWTPz3FraT8w9o6q6TEQmtHWeiGQBAVWtdl9PBe5oh497OeLYWk65cDfrV6Zz/6LVgPMjePf1+JSlTKhrmVQX+8vP+/GzezcRSlFKS1K563qvPQ+HJh4lsIzRAXKmKJsuDkMQ0kcIeRc4/1OrFirdv/Pp/69pg4Ws44SN3wiDQN55AdKGdf6cL1Pfmann1pR6myeSvMbXpsqaiFy/324AJyd+d1U9rY3rhuDU8sAJsE+o6q9au8aqrGEsLZWpFEc2LVVysURfi1tlLb1vfx00/fq2TwRW3379e37M6PAbLzW+nP1eh3H6/J5v6yJVXY8jTGSxWA43krzG12rgcycu56jqTzrIH4vFkuAIyT+40Vrq+ZCqhkXk+I50yGKxJAGHa+AD/ofTn/eBiMwBngVqW95U1dmGfbNYLIlIgmde8YKXPr50oAJHY6NlPp8CNvBZLF2VBF6O5oXWAl9Pd0T3Y/YFvBaSPN5bLJZ4OJxrfEEgm08HvBaS/GMnMImsrHUQTE07OX/lTiN2XxhlJxD7QnI9pp+htcC3XVXjmnBssVgOQxJ8Ha4XWgt8nT+V3mKxJCSHc1O3iy+hsFgsh+RwDXyquqsjHbFYLMlDIicZ9YKVl7RYLLFxmPfxWSwWy2cQkn8AIKkCnym1KlN2TamsmbBrSrErJS3KXbOLSUlVgiHlrVfyeewPvT1fX/xYBhufTQeFQRc2MOzb9XxyXyYbn0snrZtT7Rj1o1p6f8lJILp6Zgabns9AgsqYm2vodUJs022S6TszabdNfKzxuTkBlgJbVfVsERkMPAV0B94DvqWqTSKSBvwd+ALOooqvq+rG9twzNqGDGBGRfBF5TkRWicgnInJcPPZa1KqmTxrJtWcP55zvlDNgeEPcfpqw26LYdevFg7l80ghOPrfSF19N2V34dAG3XDw4bjsH0two3HDhUK48dQRXnjqC8ZOqGTmutu0Lgaq1QTY+m86kp3cz+YXdlL6ZSs0m55Ed9u16Jr/gHG8JelXFQbbMS2fKS7v44sw9fPiLHDQGAbZk+85M2fWCz4Li1wL76/j8Fvijqg4DdgOXuscvBXa7x//ontcujAY+4B5gvqqOxElRFZdIkSm1KhN2TSl2mbJrTrFLaKhz9CVCKUowRT3P0a5eF6RgTDOhDAiEoPCYZra9euhs09tfT6XojAaCqZBVFCVrQIRdy71/pmT7zkzZ9YRPmhsiUgScBcxy9wVneexz7imPAue5r89193Hfn+KeHzPGAp+I5AEnAQ8BqGqTqlb6Zd+UWpVfdg+m2OVHkDZl1ySBgHL/otU8/dEK3l+cvVcvpC1yhkcofy+FxkohXA+li1Op3+4E0fVPZPDaed1475ZsmvY4z35DWZCM3vuGGzN6RWjY4f0RT7bvrNOehdjkJQtbxMTc7UDRsT8BN7Bv9W93oFJVw+7+FqCf+7ofsBnAfX+Pe37MmOzjGwzsBB4WkaNw2urXquqn2jluQUwHSMdbsDGlVnVYqWAlENGo8INTR5CVG2HGQxsYOKKeTasz2rwud2iEz11Wz38uyyOYoeSPDCNBZfBF9Yy8sg4EVv45k+W/y+ILv6rpgE9i2Yv3Zmz5oTIwi8jZQJmqvicik/xxzBsmm7ohnLRWD6jq0TgprW488CRVnamq41V1fApti+aYUqvy264pxS6TSmCmqa0K8uF/sjnm5GrP1wz6SgMnP1fJSY/tISVXyR4UIb1QkSBIwBnw2L3c+fzpPSPUl+57pOt3BEnv5X3CWbJ9Z535LPjUx3c88GUR2YgzmDEZp3ssX0RaKmVFwFb39VagPzj5QoE8nEGOmDEZ+LYAW1R1ibv/HE4gjANTalX+2zWl2GVOCcwMeQVhsnKdEYbU9CjjTqphc7F3AZ/GCqcZW7ctwLZXUyk6q5GGnfse2+2vppE73GkV9Tm5iS3z0ok0Qe2WADWbghQcGT6o3YORbN9Zpz4LPvTxqepNqlqkqoOAi4DXVfVi4A3gq+5p04B/uq/nuPu477+ubYkGHQJjTV1VLRWRzSIyQlVX4yyBWxmPTVNqVSbsmlLsMmXXlGJXQa9mfnJPCYEABAKw+KU8lrzqvVyXXJtHU6UgKXDUrTWk5ipLf5bNnlUhEMjsF+Ho/+c0c3OHRyg6rZHXzilAgspRt9YgMfRYJNt3ZlLJry0Mr9X9GfCUiPwSeB93nMD9+5iIFAO7cIJlu2hTZS0eRGQszmhNKrAe+K6q7j7U+VZlzdKCTUtlBj9U1jJ79NeRF3hTWXt/ZvKqrLUbVf0ASLgPbbFY2s9hLTZksVgsh8QGPovF0tWQJMsUfiA28Fksltiw2VksFktXxPbxWSyWLodNRJoMtG8dc+eQ5H0nfmFq2smPi1f4bvOuYUf4bjPhSfLHtGsEPovF4h+xpZxKSGzgs1gssWMDn8Vi6UrYCcwWi6VLItHkjnw28Fksltiw8/g6FlOiQACPvrOC+pog0aijwXHNmfEL7px/eRlnfGMXqrBhVTp3XT+A5sb4M4Elk9gQmPG3vTYfOWMo35m3DoD6yiAvX1tE1ZYUcouaOefPm0nPi6IKb/yiNxvezCaUoZz+2630Gu1oWVRtS2HhTX2pLk0BlAseKmnznvEKLnW0XS/Y6SyHQERGAE/vd2gIcJuq/qm9NltEgYqXZ5KRFeG++WtYtjiHkrX+pOK54cJhVPmkO9G9dxPnfa+cy08eSVNDgFse3Mikc3ez6Jn4Uj21CMzcdNEQyrencO/ctbyzIC/uMlj4dAFzHi7kp/dsjsvOgZjw1y+b//trIQOOq2XCFeUsebCQ//21ByfdsIMN/8pm98ZUvvdaMds/yODVGX24+PkNAMz7ST8m/GAng06opak2gATarvq0CC411AUJhpS7Xyzm3ddzWLXMWwr+jrbriSSv8RlLRKqqq1V1rKqOxZGDqwNeiMemKbEhUwRDSlp6lEBQScuIUlEaf3bcZBMbMuGvXzbXvZrDERdUAnDEBZUUL8rZe3zU+ZWIQN+j62msClJTFqJibRrRCAw6wVFPSM2KkpLhJQK0X3Cpc+x6uLO/KmsdTkc1dacA61R1k18GfRcbUuHXT64DhVf+0Z15jxfGZa6iNJXnHuzJY/9bSWODsOxfuSxbHF/CVDi4wMzIcXVx2zWFCX/9sllXHiK7p5OhOatHmLpy5+dQsyOFnD77Mjfn9G6mZkeI6tIU0nOj/PMH/dmzOYWBx9dy4k93eLpXIKDct2ANfQc18dIj3T0LLnWW3VZRkn6ivWl5yRYuAp482BsiMr1FgamZRk/GTIgCXX/+MK4+fQS3XDKEL3+nnNET4hOvyc4Lc9xpe5g2cRTfHDea9MwIky/Y5YuvFv8RwZmn0QoaFra8m8mXbizlkhfWs2dzKiuez/dkv0Vw6eIvjGLE2DoGjqiP22eTdtsiBpW1hMR44BORVODLwLMHez9RxIYqSp0axJ6KFP49L4+RY+OrlRx9Yg2lJans2RUiEhb+PS+fUeO9CWm37mdyiQ2Z8Ncvm5mFYWrK3FpeWYjM7k4tL7tXM9Xb9zWGqktTyO4VJrt3Mz0/30D+gGYCIRh2ShU7VrStFrc/7RFc6ky7B6NlHl8yN3U7osZ3BrBMVb21CVrFjNhQWkaEjKzI3tdf+FI1G1fHN1hQtjWFz4+rIy09CihjT6j2ZRAm2cSGTPjrl82hU6pZMTsfgBWz8xl6SvXe4ytfyEcVtr2fQVpOhOyeYXqPqaexOkhdhdPKKHkni+7D2m6lxCu41NF220TV+5agdEQf3zc4RDM3VkyJDXXrEWbGQ86oXTAIb7yYz9I347O5+v0s3nolj78sWE0kLBSvyGDe4/GL9ySb2JAJf+OxuXtDGn89/nN88doyjv1+OS//sIiPn80nt18zZ/95CwCDJ9Ww/s0cHpo8nJSMKKf91lE3DAThpBtLefbbg0Ch1+h6xnx9N6/f3qfVe8YruNTRdr2QyLU5L5gWG8oCSoAhqtrmsJsxsSGbncXi0tWzs/ghNpSTX6RHn3Stp3PfeumGLik2VAvEX22wWCwJRbLX+JJq5YbFYkkAFIgkd+Szgc9iscSMrfFZLJauR5L3RdvAZ7FYYsbW+CwWS9fCpqUygImpJ0leLbf4h4mpJ+ev3Om7TTAnuBQvAogd3LBYLF0NSfLKREclKbBYLIcLGsPWCiLSX0TeEJGVIrJCRK51jxeIyCIRWev+7eYeFxH5s4gUi8hHIjKuvR/BBj6LxRIjvq3VDQM/VtVRwETgKhEZBdwIvKaqw4HX3H1w1v0Pd7fpwAPt/QQ28FkslpjxIzuLqm5X1WXu62rgE6AfcC7wqHvao8B57utzgb+rwztAvoi0vlD6ENg+PovFEjve+/gKRWTpfvszVXXmgSeJyCDgaGAJ0EtVt7tvlQItgir9gP21Eba4x7YTIzbwWSyW2NCYRnXL20pSICLZwPPAj1S1Svab2aGqKuL/rMGkC3wm1NDAjBIYOKnB752/hortKdw2bYgvNpPJrgn1NlNqe36olhU/lsHGZ9NBYdCFDQz7dj2f3JfJxufSSevm/H5H/aiW3l9yEqmunpnBpuczkKAy5uYaep3gXUPGpDJem/gUikQkBSfoPa6qs93DO0Skj6pud5uyZe7xrUD//S4vco/FjNHAJyLXAZfhFNNy4Luq2hCvXT/V0MCcchnAeZeVs3ltOpnZER88TT67JtTbTKntxataVrU2yMZn05n09G4CKfCf6Xn0/pKTqHTYt+sZ/r1Pp4WvKg6yZV46U17aRUNZgH9fms+pc3chHtUUTCnjecGP6SziVO0eAj5R1bv3e2sOMA34jfv3n/sdv1pEngImAHv2axLHhLHBDRHpB/wQGK+qo4EgjvZGwmFKuaywTxPHTqli3hMFPniZnHZNqLeZU9uLT7Wsel2QgjHNhDIgEILCY5rZ9uqh5RS2v55K0RkNBFMhqyhK1oAIu5Z7LytTynie8GdU93jgW8BkEfnA3c7ECXinisha4BR3H2AusB4oBv4P+EF73TddaiEgQ0SagUxgW9wWfVZDA3PKZVfcvo1Zv+xDZra/qivJZtckfqvtxaNaljM8wop7UmisFIJpSuniVLodESY1P8r6JzIomZNO/hHNHHlDLal5SkNZkG5j9gXsjF4RGnYkwUQLBXx4RFT1bQ4t8fSZjMTqZE2+Kv47m9XV3Qr8AScD83acaunCA8+LVWXNbzU0U0w4pYrK8tDemklXtWsSE2p78aiW5Q6N8LnL6vnPZXn8Z3oe+SPDSFAZfFE9UxfsYvLs3aT3iLL8dx0gAWkQQRH1tiUqJpu63XDm3QwG+gJZInLJgefFqrLmtxqaY9N/JbBRx9QycWoVjy5ZyU0PbOKoE2q44d74ZYWTza4pTKnttdBe1bJBX2ng5OcqOemxPaTkKtmDIqQXKhIECTgDHruXO89Wes8I9aX7foL1O4Kk90qS2nY06m1LUEzWq08BNqjqTlVtBmYDX4zHoAk1NDCjBPbwnX24ZPwopk0YxZ1XDuTDt7P53TUD4/Y12eyawYzanh+qZY0VTsutbluAba+mUnRWIw079/3Mtr+aRu5wR8ayz8lNbJmXTqQJarcEqNkUpODI8EHtJhQtTV0vW4Jiso+vBJgoIplAPU6bfWnrl7SOCTU0MKdcZjGj3mZKbc8P1bIl1+bRVClIChx1aw2pucrSn2WzZ1UIBDL7RTj6/zndM7nDIxSd1shr5xQgQeWoW2s8j+iCOWU8LyRyM9YLplXWbge+jrMm733gMlU9ZEderhTohMAp/juS5F+SJbFJprRUfqis5WX21eOGX+rp3AUf/bJLqqzNAGaYvIfFYuloElss3AtJt3LDYrF0MlZlzWKxdEWSvY/PBj6LxRI7NvBZLJYuhQJRG/gsFkuXwg5u+E+SF2jCYkK9Duz3hTk1tOlr1vtuc/35bS8L9USSf++JF/gsFktio0AkgZdleMAGPovFEiMKagOfxWLpatimrsVi6VLYUV2LxdIlsTU+i8XS5bCBr+MwpSrlh7rWwTCl3AZmVNbOv7yMM76xC1XYsCqdu64fQHNj/CkbTfhqomxNPQfxPrfLH81l1TO5oDDya1Uc+Z0qylem8vaMQiKNgoTghBnl9DyqkW1L0llwZW9yi5yU9oOm1vKFqyvj/gyfQhUi/opcdTRGE/yLyLUi8rGIrBCRH8Vrb+HTBdxy8WAfPPs0LepaV546gitPHcH4SdWMHFcbl80W5bZbLx7M5ZNGcPK5lQwYHrfA3F5a1ND8onvvJs77XjlXn/k5vj9lJMEgTDp3ty+2/fbVVNmaeA4gvud215oUVj2Ty/nPbeUrc7ZQ8kYmezaFWPL7AsZdvZuvzNnK+B/uYsnv9wlE9Rlfz1fmbOUrc7b6H/Ra8EdsqNMwmXp+NHA5cCxwFHC2iAyLx6Y5Van41LUOhinlNjCnshYMKWnpUQJBJS0jSkVpfOn3wYyv5srW/+cA4ntuK9el0vOoBkIZSiAEfY5tYMPCLESgucb5+TbVBMjs2cE1MBv4DsnngSWqWqeqYeBfwAUG7xcXgYBy/6LVPP3RCt5fnB2TutbBOJhymz8SiPvU0DTq32qMitJUnnuwJ4/9byVPvv8xtVVBli2OP7u1CV9Nlq3fz0G8dBveROnSdBp2BwjXCyX/yqR2e4jjbq7gnd915/GTBvDOb7pz7I937b1mxwfpPHdOP+Zd2ptda+P/5/VZ1BnV9bIlKCYD38fAiSLS3U0/fyafVkEHYldZM0U86lodiSk1tOy8MMedtodpE0fxzXGjSc+MMPmCXW1f2ArJqNyWaM9Bt2HNHHX5HuZ+rw9zL+1N9883IUFY+WQux91cwcWLSzju5goW3+zIrBYe0cg33yjhqy9t5Yhv7WHhD+Lvo/wMCqpRT1uiYlJe8hPgt8BCYD7wAfCZ+nisKmumaa+61oGYUG4Dc2poR59YQ2lJKnt2hYiEhX/Py2fU+Pj6t0z5aqps98ev58APRl5YzQUvbOXLT2wnLTdC3qBm1ryQw+Cpzvcz5Ixayj5y+lBTs5WULKemNWBSPdEwNOwy8DOPRL1tCYrRwQ1VfUhVv6CqJwG7gTUm79de/FDXOhATym1gTg2tbGsKnx9XR1p6FFDGnlBNSZwDEqZ8NVW2Jp4DP6ivcH6mNduCbFiYxbBzasjqGWb7/xzftv03nbxBTlO/bmdwb9da2YdpaFRI6+ZzAFJNenlJo9NZRKSnqpaJyACc/r2J8dgzpSrlh7rWgSSbctvq97N465U8/rJgNZGwULwig3mPd4xiV6yYKlsTzwHE/9wuuroXDZVBAiHlhBnlpOVGOemXO/nPrwqJhiGYppz4C0fwaP38LD55MhcJKqF0Zcofd5hJzJPAAxdeMK2y9hbQHWgGrlfV11o7P1cKdIJMMeZPl8ampUo6TKSluvH8VaxbXhufylqwUCdmnOXp3IW1f++SKmsnmrRvsVg6g8SequKFpFq5YbFYEgCbpMBisXQ1FFC7ZM1isXQp1E1E6mVrAxE5XURWi0ixiNzYAd4DtsZnsVjagfrQ1BWRIPAX4FRgC/CuiMxR1ZVxG28DW+OzWCyx40+N71igWFXXq2oT8BRwrnHfMTydJVZEZCfgZWp/IVBuwAVrN7l8TTa7ieDrQFWNSxZOROa79/RCOrB/6pyZqjrTtfNV4HRVvczd/xYwQVWvjsc/LyRUU9frFyIiS03MDbJ2k8vXZLObTL62hqqe3lH3MoVt6losls5iK59OXFLkHjOODXwWi6WzeBcYLiKDRSQVuAiY0xE3TqimbgzMtHaN2U0mX5PNbjL5ahxVDYvI1cACIAj8TVVXdMS9E2pww2KxWDoC29S1WCxdDhv4LBZLlyPpAp+JJS4i8jcRKRORj/2w59rsLyJviMhKV2XuWp/spovI/0TkQ9fu7X7Y3c9+UETeF5GXfbS5UUSWi8gHIrLUJ5v5IvKciKwSkU9E5DgfbI5wfWzZqvxQB3RtX+d+Xx+LyJMi4ktyRr+VDLsMqpo0G04H6DpgCJAKfAiM8sHuScA44GMffe0DjHNf5+Bkn/bDVwGy3dcpwBJgoo9+Xw88Abzso82NQKHPz8KjwGXu61Qg38CzVooz4TdeW/2ADUCGu/8M8B0f7I7G0bbJxBmofBUY5mc5HK5bstX4jCxxUdXFQHzKOp+1uV1Vl7mvq4FPcH4A8dpVVa1xd1PczZcRKhEpAs4CZvlhzxQikofzz+ohAFVtUtVKn28zBVinqvGLhDiEgAwRCeEEqm0+2EwqJcNEItkCXz9g8377W/AhmJhGRAYBR+PUzvywFxSRD4AyYJGq+mIX+BNwA+C3WIICC0XkPRGZ7oO9wcBO4GG3WT5LRPzWgbwIeNIPQ6q6FfgDUAJsB/ao6kIfTHtSMrR8lmQLfEmHiGQDzwM/UtUqP2yqakRVx+LMdD/WFW+PCxE5GyhT1ffitXUQTlDVccAZwFUiclKc9kI4XRMPqOrRQC3gW0ojdzLtl4FnfbLXDadlMhjoC2SJyCXx2lWPSoaWz5Jsga/Tlri0BxFJwQl6j6vqbL/tu827NwA/1k4eD3xZRDbidCFMFpF/+GC3pcaDqpYBL+B0WcTDFmDLfjXd53ACoV+cASxT1R0+2TsF2KCqO1W1GZgNfNEPw5okSoaJRrIFvk5b4hIrIiI4fVCfqOrdPtrtISL57usMnFxmq+K1q6o3qWqRqg7CKdfXVTXuWomIZIlITstrYCpOEy0eX0uBzSIywj00BfAzh9s38KmZ61ICTBSRTPe5mILT5xs3ItLT/duiZPiEH3YPd5JqyZoaWuIiIk8Ck4BCEdkCzFDVh+I0ezzwLWC52x8HcLOqzo3Tbh/gUTeJYwB4RlV9m3pigF7AC87vnRDwhKrO98HuNcDj7j/A9cB3fbDZEpxPBb7vhz0AVV0iIs8By4Aw8D7+LTN7XkRalAyvMjDIc1hil6xZLJYuR7I1dS0WiyVubOCzWCxdDhv4LBZLl8MGPovF0uWwgc9isXQ5bOBLIkQk4mYN+VhEnnWXKbXX1iOuyhXukq9RrZw7SURinnDrZmX5jBrXoY4fcE5Na+8f5Pz/JyI/idVHS9fEBr7kol5Vx6rqaKAJuGL/N90F8DGjqpdp6yLOk/BppYHFkgjYwJe8vAUMc2tjb4nIHGClm8Dg9yLyroh8JCLfB2cliYjc5+YyfBXo2WJIRN4UkfHu69NFZJmb7+81N8HCFcB1bm3zRHf1yPPuPd4VkePda7uLyEI3N9wsnBRarSIiL7rJC1YcmMBARP7oHn9NRHq4x4aKyHz3mrdEZKQvpWnpUiTVyg2Lg1uzOwNnYTo461RHq+oGN3jsUdVjRCQN+LeILMTJDjMCGIWzmmIl8LcD7PYA/g84ybVVoKq7RORBoEZV/+Ce9wTwR1V9210qtQAnRdIM4G1VvUNEzgIu9fBxvufeIwN4V0SeV9UKIAtYqqrXichtru2rcVY8XKGqa0VkAnA/MLkdxWjpwtjAl1xk7Lf87S2ctcBfBP6nqhvc41OBMS39d0AeMBwnf92TqhoBtonI6wexPxFY3GJLVQ+Vo/AUYJS7DA0g181CcxJuPjhVfUVEdnv4TD8UkfPd1/1dXytwUmM97R7/BzDbvccXgWf3u3eah3tYLJ/CBr7kot5NR7UXNwDU7n8IuEZVFxxw3pk++hHAyfrccBBfPCMik3CC6HGqWicibwKHSsmu7n0rDywDiyVWbB/f4ccC4Eo3JRYi8jl34f1i4OtuH2Af4OSDXPsOcJKIDHavLXCPV+Okz29hIU6SANzzxrovFwPfdI+dAXRrw9c8YLcb9Ebi1DhbCAAttdZv4jShq4ANInKhew8RkaPauIfF8hls4Dv8mIXTf7dMHPGkv+LU7F8A1rrv/R3474EXqupOYDpOs/JD9jU1XwLObxncAH4IjHcHT1ayb3T5dpzAuQKnyVvShq/zgZCIfAL8BifwtlCLk2T1Y5w+vDvc4xcDl7r+rcAH6QFL18NmZ7FYLF0OW+OzWCxdDhv4LBZLl8MGPovF0uWwgc9isXQ5bOCzWCxdDhv4LBZLl8MGPovF0uX4/9FLhnr7jqkpAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model.fit(\n",
    "    x_train_norm, \n",
    "    y_train_norm,\n",
    "    batch_size=batch_size, \n",
    "    epochs=epochs, \n",
    "    # validation_split=0.1, \n",
    "    callbacks=[tensorboard_callback, cm_callback],\n",
    "    validation_data=(x_test_norm, y_test_norm),\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "id": "6666ba3b-7ac4-4d29-a3ac-1effd58d64ce",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "PosixPath('logs/mnist-fit/20220418-022041')"
      ]
     },
     "execution_count": 110,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "log_dir"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "02b408df-5f38-49aa-8a0b-4f6f15efc8c0",
   "metadata": {
    "pycharm": {
     "name": "#%% md\n"
    }
   },
   "source": [
    "# Avaliando o modelo\n",
    "\n",
    "Agora que treinamos o nosso modelo, podemos avaliar a sua real acurácia, introduzindo dados nunca antes vistos pelo modelo.\n",
    "Essa etapa é extremamente importante, para avaliarmos se o modelo durante o treino foi capaz de inferir as características que fazem\n",
    "o dígito 5, ser 5, ou um 4, ser 4, etc."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "id": "24dcf3d7-b6ba-4db9-93c2-d43e628e6966",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "313/313 [==============================] - 1s 3ms/step - loss: 0.0832 - accuracy: 0.9858\n",
      "Acurácia do modelo: 98.58%\n"
     ]
    }
   ],
   "source": [
    "score = model.evaluate(x_test_norm, y_test_norm, verbose=1)\n",
    "print(f\"Acurácia do modelo: {score[1]:.2%}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e4a88fc5-3937-4180-8814-800e2fa39122",
   "metadata": {
    "pycharm": {
     "name": "#%% md\n"
    }
   },
   "source": [
    "**Opção 2:** Tensorboard"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "b0e956ca-1cbf-49ee-90cc-51db7452dbe6",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "313/313 [==============================] - 1s 3ms/step - loss: 0.0816 - accuracy: 0.9847\n",
      "Acurácia do modelo: 98.47%\n"
     ]
    }
   ],
   "source": [
    "score = model.evaluate(x_test_norm, y_test_norm, verbose=1, callbacks=[tensorboard_callback, cm_callback])\n",
    "print(f\"Acurácia do modelo: {score[1]:.2%}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "430ac52c-9392-41a0-b1b7-24a7ab6041bf",
   "metadata": {
    "tags": [],
    "pycharm": {
     "name": "#%% md\n"
    }
   },
   "source": [
    "## Criando função `predict_img`\n",
    "\n",
    "A função `predict_img` utiliza o modelo treinado para prever uma imagem ou lista de imagens.\n",
    "Ela também utiliza a função `plot_imgs` criada anteriormente para visualizar as imagens."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "c8d35f8c-58b7-42e8-bdee-f1c7e14a9662",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "def plot_imgs(\n",
    "    imgs: List[np.ndarray],\n",
    "    labels: List[Any] | None = None,\n",
    "    labels_prefix: str | None = None,\n",
    "    cols: int = 5,\n",
    "    figsize: Tuple[int, int] | None = None, \n",
    "    maxsize: int = 30,\n",
    "    cmap: str = 'gray',\n",
    ") -> None:\n",
    "    \"\"\"Plot array of images. \n",
    "    \n",
    "    Parameters\n",
    "    ----------\n",
    "    imgs: List[np.ndarray]\n",
    "        List of images.\n",
    "    labels: List[Any] | None, optional\n",
    "        List of labels for each image. If no label is specified,\n",
    "        then each figure is plotted without label.\n",
    "    cols: int, optional\n",
    "        Number of columns to plot. Defaults to 5.\n",
    "    figsize: Tuple[int, int], optional\n",
    "        Figure size. If no value specified, then function\n",
    "        uses `figsize = (12, 3 * rows)`.\n",
    "    maxsize: int, optional\n",
    "        Maximum number of images to plot.\n",
    "    cmap: str, optional\n",
    "        Colormap of each plot. Defaults to \"gray\".\n",
    "    \"\"\"\n",
    "    if imgs.ndim == 2:\n",
    "        imgs = [imgs]\n",
    "    if labels_prefix is None:\n",
    "        labels_prefix = 'Label:'\n",
    "\n",
    "    num_imgs = len(imgs)\n",
    "    maxsize = min(num_imgs, maxsize)\n",
    "    rows = int(np.ceil(maxsize/cols))\n",
    "    cols = min(cols, num_imgs)\n",
    "\n",
    "    if figsize is None:\n",
    "        figsize = (12, 3 * rows)\n",
    "    \n",
    "    fig, axes = plt.subplots(nrows=rows, ncols=cols, figsize=figsize)\n",
    "    \n",
    "    current_col = 0\n",
    "    current_row = -1\n",
    "    \n",
    "    for i, img in enumerate(imgs[:maxsize]):\n",
    "        \n",
    "        _current_col = i // cols\n",
    "        \n",
    "        if current_col != _current_col:\n",
    "            current_row = 0\n",
    "            current_col += 1\n",
    "        else:\n",
    "            current_row += 1\n",
    "        if rows <= 1:\n",
    "            ax = axes[current_row] \n",
    "        else:\n",
    "            ax = axes[current_col, current_row]\n",
    "        \n",
    "        ax.imshow(img, cmap=cmap)\n",
    "        ax.axis('off')\n",
    "        \n",
    "        if labels is not None and len(labels) >= i + 1:\n",
    "            curr_label = f'{labels_prefix.rstrip(\" \")} {labels[i]}'\n",
    "            ax.set_title(\n",
    "                curr_label, \n",
    "                fontsize=int((105*(figsize[0]/cols))/max(14, len(curr_label))),\n",
    "            )\n",
    "\n",
    "    plt.show()\n",
    "\n",
    "\n",
    "def predict_img(\n",
    "    img: Iterable | np.ndarray, \n",
    "    model: keras.engine.functional.Functional,\n",
    "    real_label: Iterable | str | int | None = None,\n",
    "    plot_results: bool = True,\n",
    "    sample_size: int | None = None,\n",
    "):\n",
    "    \"\"\"\n",
    "    Predict the label of an image or array of images.\n",
    "    \n",
    "    Function also normalizes images, if they're not normalized already.\n",
    "    \n",
    "    Parameters\n",
    "    ----------\n",
    "    img : Iterable | np.ndarray\n",
    "        The image to predict.\n",
    "    model : keras.engine.functional.Functional\n",
    "        The model used to predict the image.\n",
    "        If None, a ValueError will be raised.\n",
    "    real_label : Iterable | str | int | None, default: None\n",
    "        The real label of the image.\n",
    "        If not None, the predicted label will be displayed next to the real one.\n",
    "    plot_results: bool, default: True\n",
    "        Whether to plot the image(s).\n",
    "    sample_size: int | None, default: None\n",
    "        If specified, selects this amount of examples randomly and plots them.\n",
    "\n",
    "    Returns\n",
    "    -------\n",
    "    preds : List[int]\n",
    "        The predicted label, or labels of the image(s).\n",
    "\n",
    "    Raises\n",
    "    ------\n",
    "    ValueError\n",
    "        If the model is None.\n",
    "\n",
    "    Examples\n",
    "    --------\n",
    "    You can use this function as follows:\n",
    "\n",
    "    .. code-block:: python\n",
    "\n",
    "        # predicts and plots the first five images from test dataset.\n",
    "        predict_img(x_test[0:5], y_test[0:5], model)\n",
    "        \n",
    "        # predicts the first five images from test dataset.\n",
    "        predict_img(x_test[0:5], y_test[0:5], model, plot_results=False)\n",
    "        \n",
    "        # predicts and plots the first five images from normalized test dataset.\n",
    "        predict_img(x_test_norm[0:5], y_test[0:5], model)\n",
    "\n",
    "    \"\"\"\n",
    "    if model is None:\n",
    "        raise ValueError('You need to pass the trained model to the \\\"model\\\" argument')\n",
    "    if isinstance(sample_size, int):\n",
    "        take_samples = min(len(img), sample_size)\n",
    "        if not take_samples == sample_size:\n",
    "            print(\n",
    "                f'The number of samples you\\'re trying to plot ({sample_size}) is greater than the array of '\n",
    "                f'examples you\\'re trying to plot. Selecting {take_samples} samples instead.'\n",
    "            )\n",
    "        indexes = range(0, len(img)-1)\n",
    "        random_indexes = random.sample(indexes, take_samples)\n",
    "        img = np.array([img[index] for index in random_indexes])\n",
    "        if isinstance(real_label, Iterable) and len(real_label) == len(img):\n",
    "            real_label = [real_label[index] for index in random_indexes]\n",
    "    _img = img\n",
    "    if not img.ndim == 4:\n",
    "        img = img.reshape(-1, 28, 28, 1)\n",
    "    preds = model.predict(img)\n",
    "    preds = [np.argmax(pred) for pred in preds]\n",
    "    if plot_results:\n",
    "        if isinstance(real_label, Iterable) and len(real_label) == len(preds):\n",
    "            labels_prefix = 'Real Label:'\n",
    "            labels = [f'{real_lbl} / Pred: {pred}' for real_lbl, pred in zip(real_label, preds)]\n",
    "        else:\n",
    "            labels = preds\n",
    "            labels_prefix = 'Predicted Label:'\n",
    "        plot_imgs(_img, labels, labels_prefix)\n",
    "\n",
    "    return preds"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "2e0f061c-2d0c-47ca-9606-f9a589f659b3",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "from typing import Iterable\n",
    "import random\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "3504a859-5de8-4596-b896-a5cfdfbfef26",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAArYAAADTCAYAAAB0pVAsAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAAAbUUlEQVR4nO3debAdZZnH8d8TSAKEbIrKIosQpCAMkDAIYmQCA64YZZFFRJNxSpFikGUELRRDOUgh45ApxxWBkJESIY4SRNkyIlIRBUIiEMAJSUSCQZYQFknITd754+1L3tv3Ped0n9udc897v5+qUzn3OW+//XbfPk+e7tuLOecEAAAAdLthnR4AAAAAUAUKWwAAACSBwhYAAABJoLAFAABAEihsAQAAkAQKWwAAACSBwhYAAABJoLBtk5nNNDOXvWY3aeeC126bb4TdycymButrRafHs7mY2Ypguad2ejzNmNlu4Xbd6fEARZCz60HOJmcPNoULWzObnfvC51+vmdkzZrbQzL5rZkeZmdU5eKCI3H9oyX+phxIzG2VmU8zsLDO71sz+aGYbixQwqSNno1uRs9NnZrua2VfM7F4ze9rM1prZcjO72cw+bmYj2+17ywrHOVzSdtlrkqTPSFpoZqc45x6tcD5oIjvCsLz3Z+cc/1EhSWZ2u6TDJW3R6bF0KXL2IEDOxlBjZqdL+ndJW+c+2i17fUDS+Wb2Mefcg2X7b7ewXSvp17nYCElvlbRnEJss6bdmdqhz7pE25wUAMXuKorYocjaAjjOzL0i6JAhtlLRE0mpJEyTtkMX3lfRrM3unc+6xMvNot7B92jn3vtgHZranpFnyFbckjZN0nZkd4Jwbcn9SYO8bqN0rkhZJui97nSN/BBKbkLMLImcD9TCzf5T0tSC0QNIM59wfs8+HSTpB0hWStpU0XtJNZravc+61ovOp/OIx59z/SfqQpFuD8H5ZDACqcrakv5M01jk3xTl3lnPuh5Je7PC4ugo5G0DdsvP3L5PUu+P4mKSjeotaSXLObXTOXSfpmGDSPSWdVmZetdwVwTm3UdK/5sLvr2NeAIYm59xPnXMPOec2dHos3Y6cDaBm71Xfv6R9zjn3t1hD59wdkn4chM4rc2Frbbf7cs49JOnJILRf+LmZTQ+uerwziE82s/80swfN7Lns80WN5pPdauSbZvaH7ArfdWb2lJndaWbnmdn4MuM2s+Fm9s9mNt/MVgVX6t1kZsdlh8rL9Ff61jFmtoeZfSlbhj9nY3jVzP5kZr/IlmuP3DQzs6tHl+fija6IntliDHuZ2UVmtiBbn+vM7NnsCupLzWyvMush6/N9ZnZ9thxrs35/Y2ZnmNmYsv11kpntbGafNn8l/mIzW21m67N/HzOzOWZ27ECuMjezMdm6uTtbV2uzdfdjM3tvm31W+n1BOsjZr/dHzt7UJzm73DzI2Y0dG7xfLum2Fu2/F7zfSdIhhefknCv0kjRbksteKwpOsyCY5tHcZ9ODz+6UP9/3MvkTiV3utSjS99uyFZNvm389J+ljBcc7QdIDLfq7TdIbJc0MYrOb9BlOu1uL+W8j6TuS1hdYro2Sjg2mnVlgmvA1s8EYtpb07QJjWC/pUknDCqzXbeT3vpr197ikAyRNLbudFZh/n3VTQX8/abCdxl6LJE0o0OeKYJqp2bp4vEXf10nauuCYK/u+yF+1Wtn6rPoln09afjdTf4mcTc4mZ/f2R87ucM6W9FQwhu8UaL+lpJeDaS4uOq8qb/cVMyJ43+rE38slnZG9/5ukh7N/d803NLMDJN0i6S1B+BX5K+tekrS9pL3lz+V4g6QfmtlY59x3Gs3czHaWNF/SLkH41Wwcr0raS9KbJR0l6UZJv2qxPKWY2Zsk/VLSgbmPnsheG+WvFpwgv1wmKdxjXip/jtzWkg4L4uF5c6GlkTGMl3SzpHcG4R5Jj0h6RtJY+XMaR8hvdOfJr6+TmyzXcEk/lfSeIOzk1+uzknaWtIek3SXdIelzjfoaRPbXpvOENkpaJulp+SvPx8tve1sHbX9rZpOcc0/mO2rgrZLmyv9nLPlk+Wf52zJNDOZ9oqRxZvYh59z6Rp3V8X0pyvx9ZD/Z+7PjwpzBjpxdEDmbnB0gZzfv803adLcDSfptq2mccz1mdq/8ToPkfy/FlKi2Z6vEXpn8aQ4vBNPckft8evDZi9m/L0s6XdJWubZ7BO/HyyeN3mmXSzpe0pa5aXaVdEPQ7jVJk5uM9/ag7UZJ/yZpdG55Pir/xXbySaO3/ewm/YZ7Vrs1WVfzc22vl7RPpO04+Y3uHknTq9wzk3RTMO0r8ufcjcm12VZ+b3pD0PZfmvT5pdxyzZO0S67NgfJ7yfn12nI7K7hcM9tdJw36e1DSlfLnIPbb+5Y0UtLHJf0lmO8vWvS5Imjbuw4eyG+z2Xb989w6vaBJv5V/X8psY+qbNwa87gv+fu4s8t1M/SVyNjnbkbOz/sjZHczZ8jtu4fK/q+B0Pwimebzw/EoMLFzYlhuv/K1jwgW5OPf59NznPZIOK9Dv1cE0iyWNb9H+yqD97Q3afDg3lnOb9DdZ/qhE2H52k/ZFkuTpuXZfKvg72XYgG3Buuk8G062RNKlE+9WSRkXabC9/5KS33Y1q8Gcw+eT/SG49tNzOCi7bzCq/qLFlbdDubdm66Z13v//0grYrcsv+sPzV/rG2W6hvonxV0lsatK3j+1J4GxOFbcdeImf3tiVn929Pzo63I2fXU9h+Ireudi443YXBND2F51diYOHCNt145Q+f5881ye/FTM99/l8FxrCT/F6Jkz9fqOFGF27Q2rTH7iS9PdLmluDzeyVZiz5n5sY+u0nbpkky2+CXBW2a7iUWWN7CG3BuuoeC6T5dcJpbmk0j6QvB5y9L2r5Ff1Nz66vpdlZi2Wa2s04qmvdXg3l/sUm7Fbllf3eLfneSP0LT2/78Bm3q+L4U3sZEYduxl8jZYfuZubE33C5y7XaLfE7O3tR+am59Nd3OSizbzHbWSUXzJmdXX9iekVtXYwpOd1Zuum2KTFfZXRGyK1PfZv5RaQvlz7/pNdc5t7BFF98vMJuT5B8DKUm3OueWtJrAOfeKpJ8FoSPCz81sW/lzsHp9y2VrtInvyP9ZpwqHyO8l9rqoon4LM7MD5c8Dkvze6tUFJ/1h8P6IyOfhvejmOudWNevMOXenfLJOyT3B+3cUnGaxc+43zRo451ZK+p8gdGykWeXfl7Kcc9Odc9b7GkhfqBY5u23k7Aw5+3Xk7OZG5X5eW3C6V3M/b1tkonYvHts1u01JEQsl/VOLNmvkz4Fp5d3B+zsKzl+5vifnPjtQfW97dkurzpxzT5vZQkkHlRhDI+EyrXTO/a6CPgcyhrtck5PacxquVzMbIX+VaK+W6zXzS/lH6Q162W1h3iXpYPkT+cfJf4HDZPCG4P1OBbsus64+nr3f38yG5353dXxf0J3I2eRsiZxNzu6M4bmfewpOl283Itoqp867Irwkv5f8Fedcq+p8eYE9bslf3dlruplFHxEZEW6cb8p9NiF4/1yrPdTAQ6omSe4dvL+/gv7aEa7XyWZW9Eu6dfA+v153Vt+NsOhefVfs/ZvZKZIuVuQK8CbGFmzXzroaKX+18+NBrI7vC9JFzi6GnN0XOZuc3Ur+QQxbRWIxW+V+frnIzNotbNdK+nUutl5+L/5JSb+XP4n5pYL9FW33xuD9/ipz+4dN8htqeHPj50r0U6ZtM+He4V8r6rOscL3unL3KarZepeLrq6r1Whszm6X2bnEzsmC7dtfVuNzPdXxf0J3I2eTsPHJ2a+TsauQL0q1VrLDdpkU/Ue0Wtk8754ruSRSxsWC7/Hka7cifVxxuuK3u2xhaV8FY8vOvqs+yqliv+XNx8gmh6Lrt1DooxMxOUN8E+Yj8VakLJP1J/nZJr/YezTKzqSp//8x211V+ndfxfUF3ImeTs/PI2eTszeXZ3M87qNjOwPbB+xedc4VOYaj7AQ1VW6NNezQnO+euq6DPF4P3o0tMV6ZtMy8E78dV1GdZa4L333POnVZBny/mfh6tYhtyVeu1Ll8M3t8o6aMtzm9rZ3mKTpNvl1/ndXxfgDLI2fUgZxdHzu68x3I/76Jip2+Ef4l4tOjMBmNl38zTwfs319DnjtkJ9EXs3rpJIeH5YXtW1GdZda9Xqe9VxM1UtV4rZ2ZvVt+LK84qcNHGW9uYVbvrKv9n0Tp+r0AZ5Ox6kLMLIGcPGn9U3wvBDig43aTg/SNFZ9ZthW14G453NmxVzgPB++HquyKjsisr/76i+YfLdICZjWnYspg+fyLMxlpmDIcMcP6SJOfcX+WfDd2r6G1TirbrhHDv8Vnn3IoC0xzaxnzaWVcrs3UequP7ApRBzm6NnF0fcvYgkO1MhMs2pdU0Zra9+l4oelfR+XVbYRte+fkhM3tjw5bFLZV/jF6vEwtMc7j6nvsxEP+rTXsyI+Rvgj4Qr+R+3jraqq87tOkejzuUuBKzlXBDbLlezWy0pA9WNO865G9Z0pSZjVXf+0IW9cHsXp2tnBS8j91DsY7vC1AGObs1cnZ9yNmDx43B+yOzo+nNnBK83yD/+OhCuq2w/an80z4kf5L1NwfaYXbC+Jwg9Bkz26VRezMbJv9M8kpkt6qZG4Rmmlk7fwrp9YL63vx4QoN24Rj+Iik8l+fyCo5CSNI1wftJZnZ8i/YXqJqT5+sS/me6nZnt1aL9pWpveUbLPwGoITM7SdJ+QWh2pFnl3xegJHJ2ay+InF0Xcvbg8SNtunhuuKTzGzXMdhLODEI3O+eeKTynEo9Em60KH5unvo9nvLPEdB9V30eszZE0usU0IyQdJ38ofKvI5zvJn8Td2+ciRZ7jLH+x3Xdz8x/Q4xmzNnup76P2lkqa2GKZ3i9pSoPPfhf09QM1eNZ3bpo95E9c753uPkl7tJjG5P9sc4OkoyKfD8v66e3zBUnvaNDXJ+T/JDeoH88oaXnQ1+2SRjZY7q9EtpOGy6P+j2fcIOmUBm3fmdte7230O67p+7Jb0fUpHqnbsZfI2RI5Oz8NOZuc3XR9qsacLek/gr57JB0baTM82z57222UtF+p+ZQYULiwA9541WaSzKb9eu4Xv1rStyV9TP5PTv8g6cOSPi/p+tyXv98vPevz9Fyfz8g/M/r9WZ+ny5/b5eT3rn8atJ3dZKwtk2TWLp8keuT3yKfLPyJvqqSTsw2j9znl0xv09bncfP8i/4X+WfA6KTLdtGy+vdOtzzawT0t6j/x5MR+UdJr84zSfCNq+r8FYJmXrq7fda5KukP9zz2Hyf26YF3z+oyq3s2wMM3PrY23J16lN1u0S+edgT5V0pPxe5sLg8+8XWR71TZLXBe9vzNbRYfKPYPxB9nsJl2XS5vy+aPAkyVMb/L7C79GGVr/TVF8iZ5OzydmxdUvOLp43mrZt4/c6Vv4isDA//7d8UX54tp3+IbfsXys9nxIDChd2wBuvBpAks+nPy1aKK/mKJsmsz0sLTN8j/7jJmUFsdpM+CyXJrO0p8ofqiy7L9Ab9DJc/D6zZtDMbTHuUpOfbWK/RJJn1OU0+Obbq42r5ZFN3kiz7mh70tYWkmwtOd0nR5VHfJDk1Wxet+n9N0rTN/X3R4EmS06v4nab6Ejm790XOjr/I2f1f5Oya/8omf2eI5QWX5WpJVnYe3XaO7eucc1+Xf/zcj9T3/KSY5fLnqhzkmjwq0jl3vvwe9pMNmjwi6b3OuavKj7g159y1kvZR33NRYl6Qv8H0rxr0s14+2U2X9HNJf5b0asEx3C7p7fL/YbS6h+Hz8nuqH5I/utCoz3nyV+7+vkGTZySd7ZybUWSMneSc2yDpI/J71I3W6VL5eyV+scHnReYzQ9LZ8usm5neSDs7WbZH+Kv++AGWQs8nZnUDOHlycc8vkn6j2XfW/cLLXMvmj7jNcVg2XYW1MM+iY2Vby5w3tLn9z42Hyh+ZXSHrIOfdEyf6GSXq3fMIaK39vuQedc/dVOOxWY9hG/k8Zu8ov03r5e94tkbQw+7LWPYZh8veb21f++dNbyT/SbqX8fxiPOOeKPoGot8+Jkg6W9Bb5JLtM/uhPq3sLDjpmNl7+zydvkz8qsErSkiq3EzMbLn80YA/5R14+Leke59ySAfRZ6fcFKIucXdsYyNlNkLMHFzMbJf/72EXSGG36fTTaoSrWbwqFLQAAANC1pyIAAAAAIQpbAAAAJIHCFgAAAEmgsAUAAEAStuz0AFJiZlyJh0HDOWedHgMwmJGzMZiQs6vBEVsAAAAkgcIWAAAASaCwBQAAQBIobAEAAJAEClsAAAAkgcIWAAAASaCwBQAAQBIobAEAAJAEClsAAAAkgcIWAAAASaCwBQAAQBIobAEAAJAEClsAAAAkgcIWAAAASaCwBQAAQBIobAEAAJAEClsAAAAkgcIWAAAASaCwBQAAQBIobAEAAJAEClsAAAAkYctODwDp22effaLx+fPnR+NXXXVVNH7BBRdUNiYAAJAejtgCAAAgCRS2AAAASAKFLQAAAJJAYQsAAIAkUNgCAAAgCeac6/QYkmFmQ3plTpw4MRq/7bbbovEdd9wxGl+0aFE0PmnSpLbGNVQ556zTYwAGs6Gesxs599xzo/Evf/nL0fiYMWNK9W8WT009PT39Yo3uhjNr1qxo/LXXXis1lsGEnF0NjtgCAAAgCRS2AAAASAKFLQAAAJJAYQsAAIAkUNgCAAAgCVt2egBIx5FHHhmNN7r7AQBg8xgxYkS/2DXXXBNte8IJJ5Tqu+zdlRq1Hzas/7G2Sy65JNr285//fDT+2c9+NhqfO3duwdGh23HEFgAAAEmgsAUAAEASKGwBAACQBApbAAAAJIHCFgAAAEmwslczorGh/tzxnXbaKRpftGhRNL7ddtuVaj9p0qR2hjVk8dxxoLkUc/aoUaOi8W9961v9Yqeeemrdw9nsFixYEI1PmzYtGl+9enWdwymFnF0NjtgCAAAgCRS2AAAASAKFLQAAAJJAYQsAAIAkUNgCAAAgCdwVoUIpXmFbhZUrV0bjO+64YzS+Zs2aaPzoo4+Oxu++++72BpY4rrAFmksxZ1922WXR+DnnnFPbPB988MFofOnSpdH4rFmzovGPfOQj/WJnn312u8Pq48QTT4zG586dW0n/VSBnV4MjtgAAAEgChS0AAACSQGELAACAJFDYAgAAIAkUtgAAAEgCd0WoUIpX2Fah7F0RGvnJT34SjR9//PGlxzQUcIUt0Fw35+xDDjkkGr/hhhui8bL5tozJkydH44sXLy7VzxZbbNEvNmfOnGjbk046qVTfd911VzR++OGHl+qnTuTsanDEFgAAAEmgsAUAAEASKGwBAACQBApbAAAAJIHCFgAAAEnYstMDQPpuvfXWaHzGjBml+jn44IOrGA4AdI399tsvGm90l5jtt9++zuHUasOGDf1i3/jGN6JtjznmmGh85MiR0fiECROi8X333Tcaf+ihh6JxDH4csQUAAEASKGwBAACQBApbAAAAJIHCFgAAAEng4jHUbv78+dF42YvHAGCo2WWXXaLxOi8Sa3TB7w477BCNr1q1qraxLFy4MBpfv359NN7o4rFGjxTeddddo3EuHuteHLEFAABAEihsAQAAkAQKWwAAACSBwhYAAABJoLAFAABAErgrArpGo6taP/WpT0XjV155ZZ3DAYCudtFFF0XjF198cTRuZtF4T09PZWMCBoojtgAAAEgChS0AAACSQGELAACAJFDYAgAAIAkUtgAAAEgCd0VA1xg2LL4fNmrUqM08EgDoLi+//HK/2OWXXx5tu2HDhrqHA9SGI7YAAABIAoUtAAAAkkBhCwAAgCRQ2AIAACAJFLYAAABIAndFQO3mz58fjS9btiwa33333escDgAMORs3buwXe+mllzowEqBeHLEFAABAEihsAQAAkAQKWwAAACSBwhYAAABJoLAFAABAErgrAmr34osvRuPr1q0r1c/zzz8fjd96662lxwQAANLDEVsAAAAkgcIWAAAASaCwBQAAQBIobAEAAJAEClsAAAAkgbsioHbHHHNMNL733nuX6mft2rXR+NKlS0uPCQC6wdixYzs9hCQ8+uij0fgDDzywmUeCunHEFgAAAEmgsAUAAEASKGwBAACQBApbAAAAJIHCFgAAAEngrgjoGqNHj47GDzrooGj8nnvuqXM4AFC7iRMndnoIm42Z9Yt94AMfiLYdOXJkqb6feuqpUnF0L47YAgAAIAkUtgAAAEgChS0AAACSQGELAACAJHDxGLrGmjVronEuEgOQqgULFnR6CJtN7EKxefPmleqj0aNzZ8yY0daY0H04YgsAAIAkUNgCAAAgCRS2AAAASAKFLQAAAJJAYQsAAIAkcFcEAAASF3sk+fXXXx9te/LJJ0fjGzZsqGQsxx13XDR+7bXXDrjvhx9+OBp/8sknB9w3ugNHbAEAAJAEClsAAAAkgcIWAAAASaCwBQAAQBIobAEAAJAE7oqArjFu3LhofMqUKdH43XffXeNoAKB+q1atisafffbZaHy77baLxs2sX6zR3QkOOOCAaPz++++PxhvZZ599ovGrr746Gh8+fHjhvpctWxaNX3jhhYX7QJo4YgsAAIAkUNgCAAAgCRS2AAAASAKFLQAAAJJAYQsAAIAkcFcE1G7NmjXR+Lp166LxkSNHRuOvvPJKNL5kyZL2BgYAg9x9990XjZ9xxhnR+HXXXTfged50003R+I033hiNN7pbwplnnhmNjxo1qvBYnnjiiWj86KOPjsYfe+yxwn0jTRyxBQAAQBIobAEAAJAEClsAAAAkgcIWAAAASaCwBQAAQBLMOdfpMSTDzFiZJSxatCga33///aPxjRs3RuOnnXZaNH7FFVe0Na5UOOf6PxwewOu6OWePHz8+Gp83b140fuihh9Y5nEqsXLmyX+yoo46Ktk3x7gfk7GpwxBYAAABJoLAFAABAEihsAQAAkAQKWwAAACSBwhYAAABJ2LLTA8DQdfPNN0fjje6K8NRTT0XjQ/3uBwCGntWrV0fjF154YTR+1lln9YsdccQR0bbbbLNN2+MK9fT0ROONxv7Vr361XyzFux+gXhyxBQAAQBIobAEAAJAEClsAAAAkgcIWAAAASaCwBQAAQBLMua59VPag083PHUd6eO440NxQz9lTpkyJxs8999xofNq0adH4rFmzovHFixdH43PmzGk9uCGInF0NjtgCAAAgCRS2AAAASAKFLQAAAJJAYQsAAIAkcPFYhYb6hQgYXLgQAWiOnI3BhJxdDY7YAgAAIAkUtgAAAEgChS0AAACSQGELAACAJFDYAgAAIAkUtgAAAEgChS0AAACSQGELAACAJFDYAgAAIAkUtgAAAEgChS0AAACSQGELAACAJFDYAgAAIAkUtgAAAEgChS0AAACSQGELAACAJFDYAgAAIAnmnOv0GAAAAIAB44gtAAAAkkBhCwAAgCRQ2AIAACAJFLYAAABIAoUtAAAAkkBhCwAAgCT8P8IpFyAwIPB1AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 864x216 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "[1, 0]"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "predict_img(x_test, m, y_test, sample_size=2)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0e7b613c-f7af-4753-9859-2c976d55487f",
   "metadata": {
    "pycharm": {
     "name": "#%% md\n"
    }
   },
   "source": [
    "## Erros do modelo"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 332,
   "id": "ff24af2f-87c9-4e4c-884f-f7b3544a38ba",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "predictions = predict_img(x_test, model=model, plot_results=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 333,
   "id": "bca4099a-e032-4a8a-ac8d-a6fc8caf3ad9",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Modelo errou: 76 de 10,000 (0.76%)\n"
     ]
    }
   ],
   "source": [
    "false_predictions = [\n",
    "    idx for idx, (true_label, predicted_label) in enumerate(zip(y_test, predictions))\n",
    "    if true_label != predicted_label\n",
    "]\n",
    "print(\n",
    "    f'Modelo errou: {len(false_predictions):,} de {len(y_test):,} ({len(false_predictions)/len(y_test):.2%})'\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 334,
   "id": "484754c9-bd3c-4874-aef2-7629739a2a08",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x1296 with 30 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "false_preds = predict_img(x_test_norm[false_predictions], y_test[false_predictions], model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "e012148a-0598-4d85-9b35-0f3c1481a43f",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "m = keras.models.load_model('models/mnist-VIT')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2032c2ff-a515-4fd0-bdd2-00e662f4e7fe",
   "metadata": {
    "tags": [],
    "pycharm": {
     "name": "#%% md\n"
    }
   },
   "source": [
    "# Conclusão\n",
    "\n",
    "O modelo criado foi capaz de aprender com alta acurácia, a reconhecer dígitos escritos à mão.\n",
    "\n",
    "\n",
    "Como observado em [Erros do modelo](#Erros-do-modelo), as imagens que o nosso modelo não foi capaz de prever corretamente, seriam extremamente difíceis de serem classificadas até mesmo por um humano treinado. Isso porque, elas se assemelham bastante à outros dígitos. Sendo assim, podemos classificá-las como \"outliers\"."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4536cb98-d5d0-4ec2-b877-81175cfb3b35",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.8"
  },
  "widgets": {
   "application/vnd.jupyter.widget-state+json": {
    "state": {},
    "version_major": 2,
    "version_minor": 0
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}