11. Computer Vision/tutorial/05. custom-convnets.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "b6a84e61",
   "metadata": {
    "papermill": {
     "duration": 0.008688,
     "end_time": "2022-05-05T17:53:41.004487",
     "exception": false,
     "start_time": "2022-05-05T17:53:40.995799",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "<!--TITLE:Custom Convnets-->\n",
    "# Introduction #\n",
    "\n",
    "Now that you've seen the layers a convnet uses to extract features, it's time to put them together and build a network of your own!\n",
    "\n",
    "# Simple to Refined #\n",
    "\n",
    "In the last three lessons, we saw how convolutional networks perform **feature extraction** through three operations: **filter**, **detect**, and **condense**. A single round of feature extraction can only extract relatively simple features from an image, things like simple lines or contrasts. These are too simple to solve most classification problems. Instead, convnets will repeat this extraction over and over, so that the features become more complex and refined as they travel deeper into the network.\n",
    "\n",
    "<figure>\n",
    "<img src=\"https://i.imgur.com/VqmC1rm.png\" alt=\"Features extracted from an image of a car, from simple to refined.\" width=800>\n",
    "</figure>\n",
    "\n",
    "# Convolutional Blocks #\n",
    "\n",
    "It does this by passing them through long chains of **convolutional blocks** which perform this extraction.\n",
    "\n",
    "<figure>\n",
    "<img src=\"https://i.imgur.com/pr8VwCZ.png\" width=\"400\" alt=\"Extraction as a sequence of blocks.\">\n",
    "</figure>\n",
    "\n",
    "These convolutional blocks are stacks of `Conv2D` and `MaxPool2D` layers, whose role in feature extraction we learned about in the last few lessons.\n",
    "\n",
    "<figure>\n",
    "<!-- <img src=\"./images/2-block-crp.png\" width=\"400\" alt=\"A kind of extraction block: convolution, ReLU, pooling.\"> -->\n",
    "<img src=\"https://i.imgur.com/8D6IhEw.png\" width=\"400\" alt=\"A kind of extraction block: convolution, ReLU, pooling.\">\n",
    "</figure>\n",
    "\n",
    "Each block represents a round of extraction, and by composing these blocks the convnet can combine and recombine the features produced, growing them and shaping them to better fit the problem at hand. The deep structure of modern convnets is what allows this sophisticated feature engineering and has been largely responsible for their superior performance.\n",
    "\n",
    "# Example - Design a Convnet #\n",
    "\n",
    "Let's see how to define a deep convolutional network capable of engineering complex features. In this example, we'll create a Keras `Sequence` model and then train it on our Cars dataset.\n",
    "\n",
    "## Step 1 - Load Data ##\n",
    "\n",
    "This hidden cell loads the data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "d07910bf",
   "metadata": {
    "_kg_hide-input": true,
    "execution": {
     "iopub.execute_input": "2022-05-05T17:53:41.021741Z",
     "iopub.status.busy": "2022-05-05T17:53:41.021084Z",
     "iopub.status.idle": "2022-05-05T17:53:57.050477Z",
     "shell.execute_reply": "2022-05-05T17:53:57.048991Z"
    },
    "papermill": {
     "duration": 16.040339,
     "end_time": "2022-05-05T17:53:57.052575",
     "exception": false,
     "start_time": "2022-05-05T17:53:41.012236",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 5117 files belonging to 2 classes.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-05-05 17:53:50.416360: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2022-05-05 17:53:50.524879: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2022-05-05 17:53:50.525783: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2022-05-05 17:53:50.534996: I tensorflow/core/platform/cpu_feature_guard.cc:142] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 AVX512F FMA\n",
      "To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.\n",
      "2022-05-05 17:53:50.535316: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2022-05-05 17:53:50.536140: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2022-05-05 17:53:50.536781: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2022-05-05 17:53:52.489077: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2022-05-05 17:53:52.490450: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2022-05-05 17:53:52.491565: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2022-05-05 17:53:52.493369: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1510] Created device /job:localhost/replica:0/task:0/device:GPU:0 with 15403 MB memory:  -> device: 0, name: Tesla P100-PCIE-16GB, pci bus id: 0000:00:04.0, compute capability: 6.0\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 5051 files belonging to 2 classes.\n"
     ]
    }
   ],
   "source": [
    "\n",
    "# Imports\n",
    "import os, warnings\n",
    "import matplotlib.pyplot as plt\n",
    "from matplotlib import gridspec\n",
    "\n",
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras.preprocessing import image_dataset_from_directory\n",
    "\n",
    "# Reproducability\n",
    "def set_seed(seed=31415):\n",
    "    np.random.seed(seed)\n",
    "    tf.random.set_seed(seed)\n",
    "    os.environ['PYTHONHASHSEED'] = str(seed)\n",
    "    os.environ['TF_DETERMINISTIC_OPS'] = '1'\n",
    "set_seed()\n",
    "\n",
    "# Set Matplotlib defaults\n",
    "plt.rc('figure', autolayout=True)\n",
    "plt.rc('axes', labelweight='bold', labelsize='large',\n",
    "       titleweight='bold', titlesize=18, titlepad=10)\n",
    "plt.rc('image', cmap='magma')\n",
    "warnings.filterwarnings(\"ignore\") # to clean up output cells\n",
    "\n",
    "\n",
    "# Load training and validation sets\n",
    "ds_train_ = image_dataset_from_directory(\n",
    "    '../input/car-or-truck/train',\n",
    "    labels='inferred',\n",
    "    label_mode='binary',\n",
    "    image_size=[128, 128],\n",
    "    interpolation='nearest',\n",
    "    batch_size=64,\n",
    "    shuffle=True,\n",
    ")\n",
    "ds_valid_ = image_dataset_from_directory(\n",
    "    '../input/car-or-truck/valid',\n",
    "    labels='inferred',\n",
    "    label_mode='binary',\n",
    "    image_size=[128, 128],\n",
    "    interpolation='nearest',\n",
    "    batch_size=64,\n",
    "    shuffle=False,\n",
    ")\n",
    "\n",
    "# Data Pipeline\n",
    "def convert_to_float(image, label):\n",
    "    image = tf.image.convert_image_dtype(image, dtype=tf.float32)\n",
    "    return image, label\n",
    "\n",
    "AUTOTUNE = tf.data.experimental.AUTOTUNE\n",
    "ds_train = (\n",
    "    ds_train_\n",
    "    .map(convert_to_float)\n",
    "    .cache()\n",
    "    .prefetch(buffer_size=AUTOTUNE)\n",
    ")\n",
    "ds_valid = (\n",
    "    ds_valid_\n",
    "    .map(convert_to_float)\n",
    "    .cache()\n",
    "    .prefetch(buffer_size=AUTOTUNE)\n",
    ")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f441ecce",
   "metadata": {
    "papermill": {
     "duration": 0.008475,
     "end_time": "2022-05-05T17:53:57.070288",
     "exception": false,
     "start_time": "2022-05-05T17:53:57.061813",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "## Step 2 - Define Model ##\n",
    "\n",
    "Here is a diagram of the model we'll use:\n",
    "\n",
    "<figure>\n",
    "<!-- <img src=\"./images/2-convmodel-1.png\" width=\"200\" alt=\"Diagram of a convolutional model.\"> -->\n",
    "<img src=\"https://i.imgur.com/U1VdoDJ.png\" width=\"250\" alt=\"Diagram of a convolutional model.\">\n",
    "</figure>\n",
    "\n",
    "Now we'll define the model. See how our model consists of three blocks of `Conv2D` and `MaxPool2D` layers (the base) followed by a head of `Dense` layers. We can translate this diagram more or less directly into a Keras `Sequential` model just by filling in the appropriate parameters."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "f6be0fd1",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-05T17:53:57.088721Z",
     "iopub.status.busy": "2022-05-05T17:53:57.088456Z",
     "iopub.status.idle": "2022-05-05T17:53:57.190628Z",
     "shell.execute_reply": "2022-05-05T17:53:57.189832Z"
    },
    "papermill": {
     "duration": 0.113689,
     "end_time": "2022-05-05T17:53:57.192642",
     "exception": false,
     "start_time": "2022-05-05T17:53:57.078953",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d (Conv2D)              (None, 128, 128, 32)      2432      \n",
      "_________________________________________________________________\n",
      "max_pooling2d (MaxPooling2D) (None, 64, 64, 32)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_1 (Conv2D)            (None, 64, 64, 64)        18496     \n",
      "_________________________________________________________________\n",
      "max_pooling2d_1 (MaxPooling2 (None, 32, 32, 64)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_2 (Conv2D)            (None, 32, 32, 128)       73856     \n",
      "_________________________________________________________________\n",
      "max_pooling2d_2 (MaxPooling2 (None, 16, 16, 128)       0         \n",
      "_________________________________________________________________\n",
      "flatten (Flatten)            (None, 32768)             0         \n",
      "_________________________________________________________________\n",
      "dense (Dense)                (None, 6)                 196614    \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 1)                 7         \n",
      "=================================================================\n",
      "Total params: 291,405\n",
      "Trainable params: 291,405\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "from tensorflow import keras\n",
    "from tensorflow.keras import layers\n",
    "\n",
    "model = keras.Sequential([\n",
    "\n",
    "    # First Convolutional Block\n",
    "    layers.Conv2D(filters=32, kernel_size=5, activation=\"relu\", padding='same',\n",
    "                  # give the input dimensions in the first layer\n",
    "                  # [height, width, color channels(RGB)]\n",
    "                  input_shape=[128, 128, 3]),\n",
    "    layers.MaxPool2D(),\n",
    "\n",
    "    # Second Convolutional Block\n",
    "    layers.Conv2D(filters=64, kernel_size=3, activation=\"relu\", padding='same'),\n",
    "    layers.MaxPool2D(),\n",
    "\n",
    "    # Third Convolutional Block\n",
    "    layers.Conv2D(filters=128, kernel_size=3, activation=\"relu\", padding='same'),\n",
    "    layers.MaxPool2D(),\n",
    "\n",
    "    # Classifier Head\n",
    "    layers.Flatten(),\n",
    "    layers.Dense(units=6, activation=\"relu\"),\n",
    "    layers.Dense(units=1, activation=\"sigmoid\"),\n",
    "])\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b89a9842",
   "metadata": {
    "papermill": {
     "duration": 0.009247,
     "end_time": "2022-05-05T17:53:57.211033",
     "exception": false,
     "start_time": "2022-05-05T17:53:57.201786",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Notice in this definition is how the number of filters doubled block-by-block: 64, 128, 256. This is a common pattern. Since the `MaxPool2D` layer is reducing the *size* of the feature maps, we can afford to increase the *quantity* we create.\n",
    "\n",
    "## Step 3 - Train ##\n",
    "\n",
    "We can train this model just like the model from Lesson 1: compile it with an optimizer along with a loss and metric appropriate for binary classification."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "7801c361",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-05T17:53:57.230719Z",
     "iopub.status.busy": "2022-05-05T17:53:57.230478Z",
     "iopub.status.idle": "2022-05-05T17:57:12.908213Z",
     "shell.execute_reply": "2022-05-05T17:57:12.907387Z"
    },
    "papermill": {
     "duration": 195.690487,
     "end_time": "2022-05-05T17:57:12.910799",
     "exception": false,
     "start_time": "2022-05-05T17:53:57.220312",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-05-05 17:53:57.276219: I tensorflow/compiler/mlir/mlir_graph_optimization_pass.cc:185] None of the MLIR Optimization Passes are enabled (registered 2)\n",
      "2022-05-05 17:54:00.494428: I tensorflow/stream_executor/cuda/cuda_dnn.cc:369] Loaded cuDNN version 8005\n"
     ]
    }
   ],
   "source": [
    "model.compile(\n",
    "    optimizer=tf.keras.optimizers.Adam(epsilon=0.01),\n",
    "    loss='binary_crossentropy',\n",
    "    metrics=['binary_accuracy']\n",
    ")\n",
    "\n",
    "history = model.fit(\n",
    "    ds_train,\n",
    "    validation_data=ds_valid,\n",
    "    epochs=40,\n",
    "    verbose=0,\n",
    ")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "2f91fe13",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-05T17:57:12.934494Z",
     "iopub.status.busy": "2022-05-05T17:57:12.933853Z",
     "iopub.status.idle": "2022-05-05T17:57:13.540058Z",
     "shell.execute_reply": "2022-05-05T17:57:13.539319Z"
    },
    "lines_to_next_cell": 2,
    "papermill": {
     "duration": 0.618644,
     "end_time": "2022-05-05T17:57:13.542080",
     "exception": false,
     "start_time": "2022-05-05T17:57:12.923436",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAagAAAEYCAYAAAAJeGK1AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAABAeElEQVR4nO3dd3hVVdb48e9OJ51UCGmUUEILvSOoDKAM2BFBBXWcseuoM86Mvq86Oo7O73XsMhYsg4LYsYFKB2kBAqGXACEBkhBKAiGk7d8fO4EIhCTk3nvOTdbnee5zc9s5K0e5K2fvddZWWmuEEEIIu/GwOgAhhBDifCRBCSGEsCVJUEIIIWxJEpQQQghbkgQlhBDClrys2nFERIROTEy0avdCCCFsYs2aNYe01pFnP29ZgkpMTCQ1NdWq3QshhLAJpdTe8z0vQ3xCCCFsSRKUEEIIW5IEJYQQwpYsm4M6n9LSUrKysiguLrY6FFvz8/MjNjYWb29vq0MRQginsVWCysrKIigoiMTERJRSVodjS1pr8vPzycrKonXr1laHI4QQTmOrIb7i4mLCw8MlOV2AUorw8HA5yxRCNHq2SlCAJKc6kGMkhGgKbJeghBBCCJAEdY7AwECrQxBCCIEkKCGEsFZ5GUwdAktetDoS25EEVQOtNY8++ihdunSha9eufPLJJwAcOHCAoUOHkpKSQpcuXViyZAnl5eVMnjz59Hv//e9/Wxy9EMJt7PgRDm6A5a9DWYnV0diKrcrMq3vqm01s3l/g0G0mxwTzv7/tXKf3fvHFF6SlpbF+/XoOHTpEnz59GDp0KB9//DEjR47kb3/7G+Xl5RQVFZGWlkZ2djYbN24E4OjRow6NWwjRiKW+C54+UHQItn4DXa61OiLbkDOoGixdupQJEybg6elJdHQ0l1xyCatXr6ZPnz689957PPnkk6SnpxMUFESbNm3IyMjgvvvuY86cOQQHB1sdvhDCHRzeDTvnwaAHITQeUt+zOiJbse0ZVF3PdFxt6NChLF68mO+++47Jkyfzxz/+kVtuuYX169czd+5cpk6dyqxZs5g2bZrVoQoh7G7N+6A8oNdk8PaDeU/DoR0QkWR1ZLYgZ1A1GDJkCJ988gnl5eXk5eWxePFi+vbty969e4mOjuZ3v/sdd9xxB2vXruXQoUNUVFRw7bXX8swzz7B27VqrwxdC2F3ZKVj3X+gwGkJaQY+bwcPLJC0B2PgMympXX301y5cvp3v37iileOGFF2jRogUffPAB//rXv/D29iYwMJAPP/yQ7OxspkyZQkVFBQDPPfecxdELIWxvyzdQlA+9p5jHgVHQcQykfQSXPmHOqJo4pbW2ZMe9e/fWZy9YuGXLFjp16mRJPO5GjpUQbm7aaCg8APetBY/KwayMhfDhOLj6Leg+3tLwXEkptUZr3fvs52WITwghXC13C2T+Ys6ePKp9DScOhbC2sEaKJUASlBBCuF7qNFNanjLx1897VBZMZC43SayJkwQlhBCudOo4rJ8JyVdBQMS5r6dMNMlLSs4lQQkhhEtt/BxOFUDv287/ekA4JI8zSaykyLWx2YwkKCGEcBWtTeeIqGSI71/z+3pNgVPHYNMXrovNhiRBCSGEq+xfCwfWm7OnC63rljAQIjqYuaomTBKUEEK4yupp4B0A3WopIVfKVPhlr4EDG1wTmw1JgmqAC60dtWfPHrp06eLCaIQQtnbyiJl/6nY9+NWhX2f3G8HLr0mXnNeaoJRS05RSuUqpjTW8rpRSryildiqlNiilejo+TCGEcHPrZ0LZyZqLI87WrDl0vgY2zIJThc6Nzabq0urofeA14MMaXh8NJFXe+gFvVt43zA+PwcH0Bm/mV1p0hdH/rPHlxx57jLi4OO655x4AnnzySby8vFiwYAFHjhyhtLSUZ555hnHjxtVrt8XFxdx1112kpqbi5eXFiy++yPDhw9m0aRNTpkyhpKSEiooKPv/8c2JiYrjhhhvIysqivLycJ554gvHjm84V5UI0Slqb+aRWvaFl97p/rvcUWP8xpH92piVSE1LrGZTWejFw+AJvGQd8qI0VQKhSqqWjAnSl8ePHM2vWrNOPZ82axa233sqXX37J2rVrWbBgAQ8//DD1bQ/1+uuvo5QiPT2dGTNmcOutt1JcXMzUqVN54IEHSEtLIzU1ldjYWObMmUNMTAzr169n48aNjBo1ytG/phDC1fYshUPboc/t9ftcbB+I7mKSm0Vt6azkiGaxrYB91R5nVT534Ow3KqXuBO4EiI+Pv/BWL3Cm4yw9evQgNzeX/fv3k5eXR/PmzWnRogUPPfQQixcvxsPDg+zsbHJycmjRokWdt7t06VLuu+8+ADp27EhCQgLbt29nwIABPPvss2RlZXHNNdeQlJRE165defjhh/nzn//MmDFjGDJkiLN+XSGEq6S+C36h0Pnq+n1OKdNZ4vtHTAVgq17OiM62XFokobV+S2vdW2vdOzIy0pW7rrPrr7+ezz77jE8++YTx48fz0UcfkZeXx5o1a0hLSyM6Opri4mKH7Oumm25i9uzZNGvWjCuuuIL58+fTvn171q5dS9euXXn88cd5+umnHbIvIYRFCnNM5/KUieDdrP6f7zbeVP41wc4SjkhQ2UBctcexlc+5pfHjxzNz5kw+++wzrr/+eo4dO0ZUVBTe3t4sWLCAvXv31nubQ4YM4aOPPgJg+/btZGZm0qFDBzIyMmjTpg33338/48aNY8OGDezfvx9/f38mTZrEo48+KmtLCeHu1v0XKsoufg7JLxi6XmsqAIuPOTY2m3NEgpoN3FJZzdcfOKa1Pmd4z1107tyZwsJCWrVqRcuWLZk4cSKpqal07dqVDz/8kI4dO9Z7m3fffTcVFRV07dqV8ePH8/777+Pr68usWbPo0qULKSkpbNy4kVtuuYX09HT69u1LSkoKTz31FI8//rgTfkshhEtoDWs/gNZDG7ZKbq8pUFoE66Y7LjY3UOt6UEqpGcAwIALIAf4X8AbQWk9VSilMld8ooAiYorVOPf/WzpD1oBpGjpUQbuDABvjPEBj3OvSY1LBtfTDWdDm/+StIHOSQ8OyipvWgai2S0FpPqOV1DdzTgNiEEKJx2vGjuW83ouHbuv59mDYSZk6A2+ZCVOP/A1U6STRQeno6KSkpv7r169fwy8CEEI3Ajp+gZQoERTd8W/5hMOlz8GoG06+Fgv0N36bNOaLM3KG01qgLNVG0ma5du5KWlubSfdb3OiwhhAWKDkPWKhjyiOO2GRoPEz+F966A6dfBbT+AX4jjtm8ztjqD8vPzIz8/X76AL0BrTX5+Pn5+flaHIoS4kF3zQVdA+5GO3W7LbjD+v3BoG8ycCGWnHLt9G7HVGVRsbCxZWVnk5eVZHYqt+fn5ERsba3UYQogL2fEj+IdDTA/Hb7vtcBj3Bnx5J3x1N1zztlkuvpGxVYLy9vamdevWVochhBANU1EOO3+GdpeDh6dz9tF9PBRkw7ynIDgGfvN35+zHQrZKUEII0Shkr4WifEj6jXP3M/ghUyzxyysQ3Ar6/8G5+3MxSVBCCOFoO34E5QFtL3XufpSC0c9D4QGY8xgEtYDOVzl3ny7U+AYthRDCajt+hNi+pjTc2Tw84dp3IK4vfHEnZK5w/j5dRBKUEEI4UuFBOJAG7Z08vFeddzOYMBOahcLy11y3XyeTBCWEOL+TR2Hu38y9qLudP5t7Z88/nc0/DGJ6Qv4u1+7XiSRBCSHOb9OX5q/xVW9ZHYl72fEjBMWYhQZdLbwtHM6AigrX79sJJEEJIc5v9yJzv+otKHXMGmiNXnkp7FoASSNMAYOrhbeFsmJTft4ISIISQpyrogIyFkF4OziRB+mfWh2Re8hcAacKXD+8VyWsrbk/3DiG+SRBCSHOlbMRTh42feSiu8Ly183aRuLCdvwIHt7Q5hJr9h/eztzn77Rm/w4mCUoIca6Mhea+zSUw8F7I2wI751kaklvY8SMkDATfIGv2H9TSdDvPz7Bm/w4mCUoIca7diyCig2mh0/ka88XXiMqXneLIXsjb6vjmsPXh4QFhbWSITwjRSJWdgr2/nBmm8vKBvndCxgI4uNHa2Oxs50/m3qr5pyrhbWWITwjRSGWthtIiaDPszHO9p4C3v5mLEue34ydonnhmHsgq4W3hyB4oL7M2DgeQBCWE+LWMRaaPXMKgM881aw49JplqvoID1sVmV6UnzXFLGmlNeXl14e2gogyO7rU2DgeQBCWE+LWMhaYjQbPQXz/f/y7zxbf6bSuisrc9y6DspPXDe1Ct1Nz9CyUkQQkhzigugOw1vx7eqxLWBjqNgdXvQskJl4dmazt+NNVziYNqf6+zNaJSc0lQQogz9i4DXV7zdTwD7oPio5D2sUvDsjWtYcdcc8y8m1kdDQREgG9wo+jJJwlKCHFGxiJzJhDb9/yvx/WFVr1hxRtm1VhhzlSO7DHtjexAqUZTai4JSghxRsZCiO8P3n7nf10pc+Hu4QzY9oNLQ7OtHT+aezvMP1UJbydDfEKIRqTwoOkYcb75p+o6/hZC4+XC3Srb50JkJ3NM7CK8LRzdZ65pc2OSoIQQxu7F5r62PnKeXtDvLshcDllrnB+XnZ0qNBc122V4r0p4O0DD4d1WR9IgkqCEEEbGQnO9U4tutb+3583gGyJnURmLoKLUXsN70Gi6mkuCEkKYSrSMRdB6KHh41v5+3yDodSts/hqOZjo/PrvaMddUzMX3tzqSXwtvY+7dfB5KEpQQwpQkF2RB63osE9HvD6ZoYsVU58VlZxUVpr1R2+Hg6W11NL/WrDn4h7t9qbkkKCEE7F5o7msrkKgupBV0vhrWfgjFx5wRlb1lLofCA9BxjNWRnF9YW7fvJiEJSghh5p9C4sz1M/XRawqUFJpWP03N+hngEwgdr7Q6kvNrBKXmkqCEaOoqyk0FX5tL6t/oNKK9uW9q81ClJ2HTV5A8DnwCrI7m/MLbmDO8U8etjuSiSYISoqk7sN4M0bUZXv/PBkSYzhPH9jk+Ljvb+p05c+x+o9WR1KyqJ58bD/NJghKiqata3r310Pp/VikIjWsUSzvUy/qZEBwLCYOtjqRmjaDUXBKUEE3d7kUQ1RkCoy7u86HxTWuIrzAHds2D7uPNEut2VTWf6MaVfHU6ukqpUUqpbUqpnUqpx87zerxSaoFSap1SaoNS6grHhyqEcLjSYshcUXv3iAsJjTdtdZqK9E9BV0A3Gw/vAfgGQlDLxp2glFKewOvAaCAZmKCUSj7rbY8Ds7TWPYAbgTccHagQwgn2rYSy4vqVl58tJA5OHnbryfh6WT8TWvWCyPZWR1K7sLaNfoivL7BTa52htS4BZgLjznqPBoIrfw4B9jsuRCGE02QsBA8vSBh48duoapLaFAolDqZDTjp0n2B1JHUT3tatS83rkqBaAdX/z8uqfK66J4FJSqks4HvgvvNtSCl1p1IqVSmVmpeXdxHhCiEcKmOhWd/JN+jit1GVoJrCPNT6meDhDZ2vsTqSuglvC0X5cPKI1ZFcFEfN8E0A3tdaxwJXAP9VSp2zba31W1rr3lrr3pGRkQ7atRDiopw8AgfSGja8B00nQZWXmfmn9iMhINzqaOqmqpIv3z1LzeuSoLKBuGqPYyufq+52YBaA1no54AdEOCJAIYST7FlqJvsbUiABEBAFnr6NP0FlLITjOfa+9ulsp6+Fcs95qLokqNVAklKqtVLKB1MEMfus92QClwEopTphEpSM4QlhZxmLwDvADPE1hIcHhMQ2/gS1foZpwmq3pTUupHkioNy2kq/WBKW1LgPuBeYCWzDVepuUUk8rpcZWvu1h4HdKqfXADGCy1lo7K2ghhANkLDTFEV4+Dd9WaHzjLpIoLoCt30KXa8HL1+po6s7bz1xI7aaFEl51eZPW+ntM8UP15/6n2s+bgUGODU0I4TTHsiF/B/Sa7JjthcbBth8csy072vy1Kcd3l+q96ty41NzGl0ELIZymqr1RQ+efqoTGw4k8KClyzPbsZv1MM5/TqpfVkdRfeDszxOeGg1qSoIRoijIWQECkaXHkCKEJ5v5YlmO2ZydH9sLepaY4or7d3u0gvC2cKoATh6yOpN4kQQnR1FRUmDOoNsMd10supLLQtzEWSmyYZe67jbc2jot1utTc/eahJEEJ0dTkbjLDcW0vYnmNmpzuJtHIEpTWpnovcciZ39HdhLtvV3NJUEI0NbsWmPuGXqBbXVAL0zKpsZ1BZa02X+zudO3T2UITzH8bNyw1lwQlRFOzaz5EdoTgGMdt08Oz8lqoRlZqvn6GWZCx09ja32tXnl7meigZ4hNC2FppMWQuv7jVc2vT2NaFKjsFGz+HTmPAL7j299tZWFu3XFlXEpQQTUnmcnM9jyPnn6qE2CRBbf7aXFjbUNvnQPEx9x7eqxLe1gzxVVRYHUm9SIISoinJWGC6cSc44br60Hg4ftCceVglbxvMugUWv9Dwba2fCYEtoPWwhm/LauFtoewkFB6wOpJ6kQQlRFOyawHE9TWrrTpaaGWpuZXXQmWtNvdpHzcsURbmwI4fodv1Zg7H3YW5ZyWfJCghmooTh+DgBucM74E9lt3ISjX3Rfmw9buL387aD6CiDHpOdkhYlqvqau5mhRKSoIRoKk63N7rUOdu3Q4LKToXWQ82Fw2s/uLhtlJdC6nvQ9lKIaOfY+KwS3Aq8/Nyu1FwSlBBNxa4F4BcKMSnO2X5QDChP6xJUSRHkbIbYvtDjZpOQL6Zybet3ULgf+t7p8BAt4+EBYW0kQQkhbEhrUyDReqi5ZskZPL3MX+pWLbtxIA10uWno2mMSKA9Y+9/6b2f1O6Yi0Z3WfaqLsDYyByWEsKFDO6Ag23nzT1VC46w7g6qaf4rtDSGtTIJZN90M2dVVzmbYswT63O68RG6V8LZweLdZut5NSIISoinIqGpv5OwEFW9dN4nsNebMJzDKPO41GU7kmuuZ6mr1O2b5+h43OyVES4W3g4pSx53h5u+Ck0ccs60aSIISoinYtcC0uwlr7dz9hMab+Zv6nLU4SvYaiK22XlO7ERDUEtbUsVii+Ji59qnrdRAQ7pwYreTIUvOyUzBzIky/1qnrTEmCEqKxKy81w1ZtnVS9V11IHOgKM5zoSoU55sygVe8zz3l6mTOhnT/Xbdhx/UwoPQF97nBenFY6XWrugAS18DnI2wLD/uLUNbIkQQnR2GWthpLjzh/eA+tKzbOrzT9V17NyqG7d9At/XmtY9bZJcK16Oj4+OwiMAp/Ahieofath2csm+SeNcExsNXDrBPXRyr1kHz1pdRhC2NuuBaairfVQ5+8r1KKFC7NSzZISLbufFU88tLvMVPNdqDggYyHk74C+v3NqmJZSqrLUvAEX65aehK/uMtWaI//huNhq4LYJKvvoSZ74aiNDnp/P7e+vZt6WHMornDcWKoTbylgAMT2hWajz9xUcCyjXF0pkp0J0Z/Budu5rPW8182I7f67586veBv9wSL7KaSHaQni7hs1Bzfu7SeTjXnNJh3e3TVCtQpux+NFh3D2sHRuyj3H7B6kMeX4+r8zbQU5BsdXhCWEPJ4+a4gFnl5dX8fIx60y58gyqohyy1/16/qm6DqMhIArWvH/+14/ug+0/mETm7ee0MG0hvK35b1NWUv/P7lkGK94wc3SOXOzyAtw2QVF2itjpg3mk5E2W3+jDmzd1p21UIC/+tJ2B/5zPH/67hsXb86iQsyrRlO1ZYooWXDH/VCUkzrUX6x7aDiWF584/VfH0hh4TYcdcKNh/7uup08x979ucF6NdhLcz/z8c2VO/z506bob2mifA5U85JbTzcd82vcXHIKYHbPgErzXvMTqwBaM7X82BflfwfmYEn67JZs6mg8SH+TMiOZo+iWH0bR1GWICP1ZEL4Tq7FoB3AMT2cd0+Q+Nh3wrX7a/qAt2azqAAet4CS/9tiiUu+dOZ50uLTc++DlecmT9rzKqXmke2r/vnfvofc+Y15XvndMKvgfsmqMAouG4alJyA7XPNypep02i58k3+EhLPo/2uZqnvJbyzvRnTV+zl3aW7AWgXFUjf1mH0rUxYMaHnGbMWorHIWACJg83Qm6uExpl/j+VlrlmqIjsVfEPOlFGfT1gbaH2JKZYY8ojpTQew+SvT+byxlpafLbwyQdWnUGLXAkh9F/rfAwkDnRNXDdw3QVXxCYAu15hb8THY+j1s/ByvFa8zrOJlhoW3o7zPQLJ9WrP2VCt+OuTLN2n7+XilGSNvFdqMfq3D6B4XSueYYDq1DCbA1/0PixAc2WOapfb9vWv3GxpveuIVHnDNWUnWGlMa7lHLjEWvW+Gz2yBjPrS73Dy36i0IT3LZnIrl/MOgWXNTMNL1BgiKvvD7i4/B1/eaY3TZE66JsZrG9U3sFwIpE8yt6DBsmQ2bv8Zz67fEnzxMPHAVoINbcjy0A3s8ElhT3JK52yP5Zl0kpXihFLQOD6BzqxA6xwRX3kJkaFC4n12V7Y1cVSBRpfq1UM5OUCUnIHcTDHm49vd2HGMq9da8bxJU9hpzG/2CUy82tZ2et8Kyl+DfnaHz1dD/D6bB7vnM/aupgLz9p/NXSDpZ40pQ1fmHmV5cvSabi/AKD5r/kXM2o3I2EZS7ia55v9C1vITJgPb35nhwW/b5tCO9PJ4lGS15Y300BZjx1pYhfvSID6V3Qhh9EsPo1DIIL0/3rTERTUDGArMERkQ95hocIaQyQbmiUGJ/mpn0v9D8UxUvX+g+AVZONZ0nVr1jLlztPsHpYdrKiKfMnNyqt2DdR5A+yyxR0u/3kDzOFJWAmTpZNx0GP1RzAYqTNd4EVZ1SENzS3KpO7cG0gMnfBTkbUTkbCTqYTvLBVJKPf8t4AD8oDmjFwWZJbNEJ/Lg3kRfSEynGF38fT3rGN6d3YnP6JIaREhcqQ4PCPirKIWMRdLzS9WcHIbHm3hWl5jV1kKhJr8mw/DX45RUzT9Zjkkuu57Gd8LYw+nkY/jdI+xhW/Qc+vx1+fNx0ck++GmbfD1HJpp2RRZr2N6qnN0R1NLeu1515/nguHEyHg+n4HUwn8WA6iYcXM1pX8GKAD/nNu7PBqxvfH+vA6/NaUKq98PRQdI4J5pL2kYzpFkOHFkHW/V5CHEiD4qOuLS+v4u0HgdFwdK/z95W9BkITICCibu+PSIKEQSZJQePuHFEXfsFmiK/vnbDzJ1jxJsx/xtw8vGDiLHPmaZGmnaBqEhhl2qO0u+zMcyUnIHM5KmMREbsXcemBaVyK5l9BARyO6EO6T3e+O96eNxYc4dX5O0mKCuS33WMY060lbSJdV5YpBHBm/smqyX9XLbuRtQbi+9XvM70mw95lkDgEojo5JSy34+EB7UeaW+5WWPMeRHc5t3WUi0mCqiufADM8WDVEWHQY9ixBZSwifPcihu1fyDDg+ZBQ9gd1Y9HJdnwxL57XfmpNu5bhp5NVXJi/lb+FaCp2LYDorhAYac3+Q+Jg/zrn7qPwIBRkQat76ve5TmMh6TMYdL9z4nJ3UR3N8J8NSIK6WP5hZkIxeZx5fCwbdi/GY+8yYjNXMPH4Yib6QJmHL1uPJ7Hw57Y8/mNHSlr2Znj3tozs3IKE8ABrfwfROJ06DvtWQv+7rIshNB62fAMVFbWXf1+srHrOP1Xx9jNDV8L2JEE5SkirMyXuAMfzYN8KvDJX0GXvL3Q+8C1Kf015vgcH5zUn7+dQ9vtGEBDeiphWCYS3iEcFRUNgCzPE6BcMXn7g6dO0SmBFw+340ayc6ury8upC400Mxw+a3nzOkJ0KHt7Qoptzti8sJwnKWQIjodNvzQ1QJScgKxXPzBWE5uxE5+yjvOAAAQe2EHGw4MLb8vQ5k6y8fM3N09dcZBfZ0dyiOpl7V3SsFvZVehJ+fhIiOpg5FqucvhZqn/MSVFYqtOjS+Bu8NmGSoFzFJwDaXAJtLiEAqBrcyy0s5uONWazYsJXMzN2E6yO0a3aclGgvOkT4Eh/siZcuNUssl58yXYjLT0FZMRzLgrUfQmnRmf0EtjBjyJGdILLySyriAi1gROOy9CVTPXfL7DPXs1ih+sW69S1iqIuKcjPH1dSuYWpi6pSglFKjgJcBT+AdrfU/z/OeG4AnAQ2s11rf5MA4G62oID9uGtCOmwa041hRKfO35TBn40E+3H6Ikxnl+Pt4MiQpgss6RXNpxygiAs8q+ayoMBdE5m01t9ytZinmtR+cSVwJg8zV48ljLbkaXLjI4QzTELXLteaPISudvhbKSaXmedvMKsE1dUAQjUKtCUop5Qm8DowAsoDVSqnZWuvN1d6TBPwFGKS1PqKUinJWwI1ZiL83V/eI5eoesRSXlrN8Vz4/b8lh3pZc5m7KQSnoERfKZZ2iGZEcTVJUIMrDw7TAb55gSkSrVFSYL4fNX5tk9eWd8MOfoPuNJllFJ1v3iwrH0xp++LM5a/rNM1ZHY0YM/COc102ivhfoCrektL7weklKqQHAk1rrkZWP/wKgtX6u2nteALZrrd+p64579+6tU1NTLyropkZrzab9BaeTVXr2MQDiwppxeadoRnSKpk/rMLxrar1UUWHWBVr7gamsKi8xrU163Wp6cflINaHb2/o9zJxgktPA+6yOxnhruJkTvflLx2979v2mE/mf9jivSlC4jFJqjdb6nL826jLE1wqo/mdQFnD2oHL7yp0swwwDPqm1nnOeIO4E7gSIj4+vW+QCpRRdWoXQpVUID17enoPHipm31SSrj1Zm8t6yPQT5eTG8QxSXJ0dzSftIQppVm3/w8Dg9/8WJfFg/wzTM/PoemPMXs1DbJX+SROWuSorM2VNkJ+j3B6ujOSM0DnI21/6+i5G9xgzvSXJq1BxVJOEFJAHDgFhgsVKqq9b6aPU3aa3fAt4CcwbloH03OS1C/JjYL4GJ/RIoKiljyY5DzKs8u5q9fj9eHoq+rcO4vFM0v+kcTWzzahcHB4TDwHthwD2QuRxWv2M6G2/8HK78v18PEwr3sPTfcCwTJn9nbWHE2ULjTcNRrR17qcSp45C72SwyKBq1uiSobKB6z/zYyueqywJWaq1Lgd1Kqe2YhLXaIVGKGvn7eDGycwtGdm5BeYUmbd9Rft6Sw8+bc3j62808/e1meic0Z1xKDFd0bUl4VZGFUmbxsYSB0Od38O2D8PEN5sLjUc+bxrrC/vJ3mT8wut5gFia0k5B4U216PLf2dYfq40Ca6WAu80+NXl3Oj1cDSUqp1kopH+BGYPZZ7/kKc/aEUioCM+SX4bgwRV14eih6JTTnz6M68tMfL2HhI8N4dGQHCopLeeLrTfT9xzxunbaKL9ZmcfxU2ZkPJgyA3y+BS5+AbXPg9b6w6m1Tyivs63RhhC/85u9WR3OuUCctu3F6iXep4Gvsaj2D0lqXKaXuBeZi5pemaa03KaWeBlK11rMrX/uNUmozUA48qrXOd2bgonaJEQHcM7wd9wxvx9aDBXydtp/Zafv546z1+Hqlc3lyNGO7xzCsQyS+Xj4w9BFTNPHdH+H7R8xc1W9fhhZdrf5VxPls/c50oB75HAS1sDqac52+FmqvY892slOheWLdO5gLt1VrFZ+zSBWfNbTWrM08wtdp+/luwwHyT5QQEejL7YNbM6l/PEF+3uYv8/RPTQHFySNmvmrYY1JEYSclReZM1zcYfr8YPG14zf2pQnguFi5/CgY/6LjtvpgM8QPguncdt01hqZqq+KQEpolRStErIYynx3Vh5V8v470pfejUMojn52xl4D/n88KcrRw6UQLdboB7V0OPiWZxt9f6wIZPTfIS1lvyf2bo7Mr/Z8/kBOAbBM2aO3bhwoIDUJAt809NhCSoJszL04PhHaL47+39mH3vIIYkRfDmol0M+ud8/ufrjewr9oOxr8KUOeAfDl/cAe9cDvuk9sVSh3aaPxq63WiKXOwsJM6xc1BVF+jWZYl34fYkQQkAusWG8sbEXvz8x0u4KqUVM1ZlMuz/LeShT9LY5tsF7lwE494wXzbvXg6f3e6aBenEr2kNPzxqmgePeNrqaGoXGu/YM6isqg7mMi/aFEiCEr/SNjKQ56/rxuI/DWfKwETmbjrIyJcWc/8n6znW8Qa4by0MfRS2fguv9Yb5z5rrUoRrrHgTds2H4X9zbOm2s4QmmATlqKHh7DUmOUkH8yZBEpQ4r5YhzXh8TDLL/nwp913aju/TD3DFy0tYm1MKlz4O96ZCxzGw+AV4tRes+8i0VBLOoTUs/hfM/Qt0uBL63GF1RHUTGmeaFhcdbth2ykrg4EbTwVzmn5oMSVDigpoH+PDwbzrw6R8GoBRcP3U5by7cRUVwrKmiuu1Hs1jj13fDW0PPdA4QjqM1/PQEzH/GzDvd8KF9CyPOVr3UvK4Kc2DnPFj2CnxxJ7w5CP4RA1MHmQ7mVq5zJVzKTf4vF1brEd+c7+4fwl+/SOf5OVv5ZdchXrwhhcj4fnD7z7DxM1jwrOlGEdvHnGW1vkRWA26oinLT5WPth9D3TtPlw536z4VUNqE5tg9a9az5ffm7YO5fzRDeibwzzwfFQHRnSBoB0V3M6rmR7Z0bs7ANSVCizkKaefPaTT0YtCqCp77ZxOiXl/Dv8d0ZkhRpytI7Xw1pH8GiF+DDceYv3eF/M50qRP2VlZhlUjZ9aeb9hv/N/RJ+9YULa5KxEGbdan7uOMaskhvd2SQk/zCnhyjsy43+FBN2oJTipn7xzL53MM39vbll2iqen7OV0vIK06i012RTSDH6BbOo3HujYPq1kL3W6tDdS0kRzLzJJKcRfzdnpO6WnMAst+Ebcv4EpTWsfAv+e43phHHnArjqdeh/F7QeKslJSIISF6dDiyBm3zuYG/vE8ebCXdzwn+XsO1y5gq+3H/T7PTyw3pRCZ6+Ft4fDjJsgY5GZY5B5qpoVHzNJfefPptXUoPutjqhhQuPOvSShrMQMXf7wqBm+u/0nCGtjSXjCvqTVkWiwb9bv569fpIOCZ6/uytjuMb9+Q3EBrJwKv7wKpwrMcz5BEN4GwtpCeNvK+3bm56b8l/OJQzD9GsjZBNe8ZZZvd3czJsCRvXD3L+bxiUPwyc2Q+QsMfsg0KfbwtDZGYamGLFgoxAX9tnsM3WNDeeCTddw/Yx0Lt+Xy1NjOpq8fgF+wWRCx752mE0D+LnM7vMuUDW/+yiyfUCWqM1w9FVp2s+T3sURZiVn1eM5jZjjsxhnQ/jdWR+UYIXGwZ6k5a87ZZBLWiVy45h3odr3V0QkbkwQlHCI+3J9Pfz+AV+bv5LX5O0jdc4SXbkyhZ3zzM29qFgrtLje36spKTBly/i44tB2Wvw7vXGaajPa/yz3nXuri1HHYNQ+2fGvK808dA79QmPQFJA6yOjrHCY03Z87rppvlQfyCYcr3slyGqJUM8QmHW73nMA/OTONgQTEPXJbEPcPb4elRjyRzIh9m3wvbvod2I+CqNyAwynkBu9KJfNj+g0lKGQvMgn7NwszqsB2vhLbDwbuZ1VE61ubZMOtm83OrXjD+I1kQU/xKTUN8kqCEUxw7WcoTX21k9vr99Elszos3pBAX5l/7B6tobZaj//Fx0xX7qqmQdHntn7OTkiI4tM1UM+ZtNU12M38xw5khcSYhdRxjlo5wlwtvL0b+LtMNv+t18NtXpE2ROIckKGGJL9dl8cRXm1DAM1d3YVxKq/ptIGczfH475G6G/nfD5U+Cl68zQr14pcWQu+lMIsrdau6PZgKV/748vCGyI3QYZZJSy+6Nd+jyfIoOm6U3mtLvLOpMEpSwTGZ+EQ9+so61mUcZlxLDk7/tTPMAn7pvoPQk/PS/sOo/EN3VtFiK7OC8gGtTcAD2rYR9qyBrFRxYD+Ul5jVPHwhPgqiOJiFFdjD3YW3MdWJCiHNIghKWKiuv4NX5O3l9wU5C/b15amwXrujaAlWfv6i3zTE9/0qKYMjDENcHIjuZ+Sln/WVeehJyt0DW6jNJqWp9I09f074nrq+ZW4nqbJYib8zDdUI4gSQoYQub9xfw5883kJ59jJGdo/n7uC5EBddjTqLwIHx1l1lyokqz5iZRRXU8cx+VDAERddtmeZmpIjycAfk7q912wbEsTg/TBcVAfD+I7Qtx/cyyD171OBMUQpyXJChhG2XlFbyzdDcv/rQdPy8PHh+TzPW9Yut3NlWYA3lbKud7tpiznNytplS7ire/GVbz8AYPr8qfPc3jqp9LT5qLSCtKz3zON8RcMBzeztwi25sGuCGxjjsIQojTJEEJ28nIO86fP9/A6j1HGJIUwT+u7lq/Sr+zaQ2FB0yyytsKBfuhvBQqykwCqij/9ePyMnMGVNXNoioh+YfLZL4QLiQJSthSRYXmo5V7+ecPW9HAn0Z24JYBiXjU57opIYRbqylBSbNYYSkPD8XNAxKZ+9BQeieG8eQ3mxn/1nKyjhRZHZoQwmKSoIQtxDb354Mpffh/13dny4FCRr+8hG837Lc6LCGEhSRBCdtQSnFdr1i+v38IbSMDuffjdTz66XpOnCqzOjQhhAUkQQnbiQ/359M/DOC+S9vx2dosxry6lPSsY7V/UAjRqEiCErbk7enBw7/pwIzf9ae4tJxr3lzGfxbtoqJCFjoUoqmQBCVsrX+bcH54YAiXdYzmuR+2csu0VeQWFFsdlhDCBSRBCdsL9ffhzUk9ee6arqTuPcyol5fw8+Ycq8MSQjiZJCjhFpRSTOgbz7f3DaFFsB93fJjKs99tprS8ovYPCyHckiQo4VbaRQXyxd0Dubl/Am8v2c34/yxn/9GTVoclhHACSVDC7fh5e/L3q7rw6oQebDtYyJWvLGHBtlyrwxJCOJgkKOG2fts9hm/uG0x0sB9T3lvNv+ZupUyG/IRoNCRBCbfWJjKQr+4ZxI194nh9wS4mvrOSHKnyE6JRkAQl3J6ftyf/vLYbL97QnQ1Zx7jylSUs23nI6rCEEA1UpwSllBqllNqmlNqplHrsAu+7VimllVLndKUVwtmu6RnL7HsHEervw6R3V/LiT9tlyE8IN1ZrglJKeQKvA6OBZGCCUir5PO8LAh4AVjo6SCHqKik6iNn3DuLqHq14Zd4Orpu6nN2HTlgdlhDiItTlDKovsFNrnaG1LgFmAuPO876/A88DMgEgLOXv48WLN6TwyoQeZOQd54qXlzB9xV6sWvtMCHFx6pKgWgH7qj3OqnzuNKVUTyBOa/3dhTaklLpTKZWqlErNy8urd7BC1MfY7jH8+NAl9E5szuNfbWTK+6ulTZIQbqTBRRJKKQ/gReDh2t6rtX5La91ba907MjKyobsWolYtQvz4YEpfnhrbmeW78hn50mJ+SD9gdVhCiDqoS4LKBuKqPY6tfK5KENAFWKiU2gP0B2ZLoYSwCw8Pxa0DE/nu/iHEhflz10dr+eOsNAqKS60OTQhxAXVJUKuBJKVUa6WUD3AjMLvqRa31Ma11hNY6UWudCKwAxmqtU50SsRAXqV1UIJ/fNZD7L0vi67T9jH5pCct35VsdlhCiBrUmKK11GXAvMBfYAszSWm9SSj2tlBrr7ACFcCRvTw/+OKI9n/1hAD5eHtz0zgqe/W4zp8rKrQ5NCHEWZVVlU+/evXVqqpxkCesUlZTxj++3MH1FJh1bBPHv8Sl0ahlsdVhCNDlKqTVa63OmhaSThGiy/H28eOaqrrw3uQ+Hjpcw7jWzam+5rNorhC1IghJN3vCOUfz40FCGd4zkuR+2ctPbK8g6UmR1WEI0eZKghADCAnyYOqkX/7quG5v2FzD6pSV8viZLLu4VwkKSoISopJTi+t5x/PDAEDq2DOLhT9dzz8drOXKixOrQhGiSJEEJcZa4MH9m3jmAP4/qyE+bcxj18mJW7zlsdVhCNDmSoIQ4D08PxV3D2vLl3YNo5u3JjW+t4O3FGTLkJ4QLSYIS4gK6tAph9n2DGdEpmme/38Jd09dKBwohXEQSlBC1CPbz5s1JPXn8yk78tCWHsa8uZcuBAqvDEqLRkwQlRB0opbhjSBtm/K4/RSXlXP3GMj5bk2V1WEI0apKghKiHvq3D+Pb+waTEhfLIp+v5yxfpFJdKmyQhnEESlBD1FBXkx/Tb+3H3sLbMWJXJdVN/Yd9hubBXCEeTBCXERfDy9OBPozryzi29ycwvYsyrS1myQxbhFMKRJEEJ0QCXJ0fz7X1DaBnix63TVvHu0t1Sii6Eg0iCEqKB4sP9+fyugYxIjubv327m0c82yPIdQjiAJCghHCDA14s3J/bigcuS+GxNFje+tYLcgmKrwxLCrUmCEsJBPDwUD41oz5sTe7L1QCFjX1vG+n1HrQ5LCLclCUoIBxvdtSWf3zUQTw/F9f9Zzlfrsq0OSQi3JAlKCCdIjglm9r2D6BEXyoOfpPHcD1tkIUQh6kkSlBBOEh7oy/Q7+jGpfzz/WZTB7R+slj5+QtSDJCghnMjb04NnrurKM1d1YemOQ1z7xi9k5stFvULUhSQoIVxgUv8EPrytL7mFp7jqjWWs2i3rSwlRG0lQQrjIwHYRfHXPIEKbeTPxnRV8mrrP6pCEsDVJUEK4UOuIAL68exB9W4fx6GcbeO6HLVRI8YQQ5yUJSggXC/H35v0pfZnYzxRP/H76Gk6cKrM6LCFsRxKUEBYwxRNdePK3yczbksN1U5eTffSk1WEJYSuSoISwiFKKyYNaM21yH7IOFzHutWWsyzxidVhC2IYkKCEsNqxDFF/cPRB/H0/Gv7WC95ZJR3QhQBKUELaQFB3EV/cMYki7CJ76ZjOT31tNXuEpS2Pad7iIuz9aw87c45bGIZouSVBC2ERYgA/v3Nqbv4/rzIqMfEa9tJj5W3MsiUVrzRNfb+T79IPc8u5KDhyT+THhepKghLARpRQ3D0jkm/sGExnky23vp/I/X2+kuNS160v9tDmHhdvyuKlfPIXFZdz87iqOnChxaQxCSIISwobaRwfx9b2DuH1waz5cvpexry1ly4ECl+z7ZEk5T32zmQ7RQTw1tjNv39qbzMNF3PbBaopKpBxeuI4kKCFsytfLkyfGJPPBbX05UlTKuNeXMW3pbqdf2Pvmwp1kHz3J0+M64+3pQf824bxyYw/W7zvKXdPXUlpe4dT9C1FFEpQQNndJ+0jmPDCEoUkRPP3tZia/v9ppq/XuOXSCqYszGJcSQ7824aefH9WlBf+4uiuLtufx6KfrpfuFcAlJUEK4gfBAX96+pTd/v6oLq3bnM/KlxczZeNCh+9Ba89Q3m/D2UPz1ik7nvH5j33geHdmBr9L288x3W6QUXjidJCgh3IRSipv7J/DtfUOIbe7PH6av4dFP13PcQW2Sft6Sy4JteTw0oj3RwX7nfc/dw9oyZVAi05bt5s1FuxyyXyFqUqcEpZQapZTappTaqZR67Dyv/1EptVkptUEpNU8pleD4UIUQAO2iAvn8roHcO7wdn6/NYvTLi0nd07DlO4pLy3nqm020jw7k1oGJNb5PKcUTVyYzLiWGF+ZsY+aqzAbtV4gLqTVBKaU8gdeB0UAyMEEplXzW29YBvbXW3YDPgBccHagQ4gwfLw8eGdmBWb8fAMAN/1nO/5u77aILGN5YuIusIyd5amwXvD0v/LXg4aH413XdGdo+kr9+mc7cTY4dahSiSl3OoPoCO7XWGVrrEmAmMK76G7TWC7TWVcuErgBiHRumEOJ8eieG8f39Q7i2ZyyvLdjJNW/8Uu/OD3vzTzB10S5+2z2GAW3Da/8AJkFOndSTbrGh3DdjHX//djMZedJxQjhWXRJUK6D6ympZlc/V5Hbgh4YEJYSouyA/b/51fXemTurJviNFjHl1CR/8soeSsrqdTT39zWa8PRR/O09hxIX4+3jx3uQ+jOzcgg9+2cOl/7eIie+s4If0A1KKLhzCy5EbU0pNAnoDl9Tw+p3AnQDx8fGO3LUQTd6oLi3pEd+cRz/bwP/O3sSr83dwXa84buwTR2JEwHk/8/PmHOZtzeWvV3SkRcj5CyMupHmAD69O6EHumE7MWr2PGav2cddHa4kK8uXGvvFM6BtHy5BmDf3VRBOlaisVVUoNAJ7UWo+sfPwXAK31c2e973LgVeASrXVubTvu3bu3Tk1Nvdi4hRA10FqzcHseM1ZmMm9rLuUVmkHtwrmpbwIjkqPx8TIDJ8Wl5Yz49yJ8vTz54YEhtc491UV5hWbB1lymr9zLou15eCjFZR2jmNg/gYFtwx2yD9H4KKXWaK17n/N8HRKUF7AduAzIBlYDN2mtN1V7Tw9MccQorfWOugQkCUoI58spKObTVHNmk330JOEBPlzXO5Yb+8TzdVo2L/28g4/v6MfAdhEO3/e+w0V8tDKTT1P3kX+ihEBfL/q3CWNIUiSDkyJoExGAUsrh+xXu56ITVOWHrwBeAjyBaVrrZ5VSTwOpWuvZSqmfga7AgcqPZGqtx15om5KghHCd8grN0p2HmLEyk5+25FBeofH0UIzq0oLXb+rp1H2fKitnwdY8Fu/IY+mOQ2QeNvVUMSF+DE6KYHBSJIPbRRAW4OPUOIR9NShBOYMkKCGskVtQzKdrsli1+zDPX9vtouaeGiIzv4glO02yWrbzEAXFZSgFXWJCuGNIa8Z2j5EzqyZGEpQQwnbKKzTp2cdYsj2P79IPsPVgISlxoTwxphO9EsKsDk+4iCQoIYStVVRovliXzb/mbiWn4BRXdmvJY6M6Ehfmb3VowslqSlBSUiOEsAUPD8V1vWJZ8MgwHrw8iflbcrns/xbx3PdbKCgutTo8YQFJUEIIW/H38eLBy9uz4JFhjE2J4a0lGQz710L+u3wPZXIBcJMiQ3xCCFvbmH2MZ77bzIqMw7SNDODKbjH0iAule1yoVP41EjIHJYRwW1prft6Sy6vzd7Ax+xhV6yUmhPuTEhd6+pYcE4yvlydgCjAOFhSTmV/EviNF7DtcRGbl7eCxYh4b3ZFxKRfq2iZcpaYE5dBWR0II4QxKKUYkRzMiOZoTp8pIzz5G2r6jpGUeZWXGYb5O2w+Aj6cH7VsEcuJUOVlHiigtP/MHuKeHomWIH/Fh/gT4evHXL9LpEdec+HApwrArOYMSQri9g8eKSdt3hHX7jrJ5fwHBzbyJD/Mnrrk/8WHm1jLU73SrpeyjJxn10mLaRwfxyZ398ZIWTJaSIT4hhKjm67RsHpiZxsMj2nPfZUlWh9OkSZm5EEJUMy6lFeNSYnhp3g7S9h21OhxxHpKghBBN1tPjuhAd5MuDM9dx4lSZ1eGIs0iCEkI0WSHNvPm/G1LYe7iIZ77bYnU44iySoIQQTdqAtuHcObQNM1Zl8uOmg1aHI6qRBCWEaPIeHtGB5JbBPPZFOrmFxVaHIypJghJCNHk+Xh68fGMKJ06V8afPNmBVdbP4NUlQQggBJEUH8dcrOrFwWx7TV+y1OhyBJCghhDjtlgEJXNI+kme+28LO3EKrw7HE7kMnePyrdJ6cvYmMvOOWxiIX6gohRDW5hcWMemkJLUP8+PLuQfh4NY2/43fkFPLagp18s34/3p4eaA0l5RVc1jGK2we3ZkDbcKetdCydJIQQoo7mbjrI7/+7ht4JzXlybGe6tAqxOiSn2bT/GK/N38mcTQdp5u3Jzf0TuGNIGwCmr9jL9BV7yT9RQscWQdw2uDXjUmJON+R1FElQQghRD5+tyeK577dwuKiEG/vE8chvOhAe6Gt1WA6zft9RXp2/g5+35BLk68WtAxO5bXDrc5YwKS4tZ3bafqYt283Wg4VEBPowqX8Ck/onEOGg4yEJSggh6unYyVJe/nkHHy7fQzMfTx68vD23DEg43XTWnWityS08RXrWMT5csZfF2/MIaebN7YNbc+vAREKaedf6+WU783l3aQYLtuXh4+XBTX3jeXJs5wbHJglKCCEu0s7cQp76ZjNLdhyiXVQg/zMmmaHtI60Oq0Zl5RVkHDrB5v0FbD5QwJYDBWzeX0D+iRIAwgN8uGNIG24ekECgb/1XXdqVd5z3lu0mwNeLv4zu1OB4JUEJIUQDVC2a+Mx3m9mbX8TlnaJ5YkwnEsIDLIupuLScPfkn2J13goxDJ9h96ATbDhayLaeQkrIK4MwaWcktg+nUMpjklsF0jwvFz9ux80gNIQlKCCEc4FRZOe8u3c1r83dSVq7p1yaM9tFBdIgOon2LIJKiAgmo41lJeYUm//gpcgpOcbK0nPIKTYXWlFdoyrWmokKffq6sQpNTcIrdh46z+5BJSvuP/brrRVSQL0nRJhklxwST3DKENpEBth+SlAQlhBAOlFNQzBsLdrI28yg7cgspLq04/Vp8mL9JWi0CSYoKOr38fG5BMQcLijlYcIrcgmJyC09RXlG/7+BgPy9aRwbSJiKA1tVuiREBFzVcZwey5LsQQjhQdLAfT43rApgzoX2Hi9iWU8j2yiG27TmFLNyWS1m1BBTs50WLED+ig/1IioqgRbAf0SF+RAf5EuDrhYdSeHooPD3MMveelY+rno8M8qW5v7fTrkeyG0lQQgjRQJ4eisTKs5iRnVucfr6krII9+Sfw9vQgOtgXfx/5yq0POVpCCOEkPl4etI8OsjoMt2XvmTMhhBBNliQoIYQQtiQJSgghhC1JghJCCGFLkqCEEELYkiQoIYQQtiQJSgghhC1JghJCCGFLkqCEEELYkmXNYpVSecBeB2wqAjjkgO24ksTsGhKz67hj3BKza9Ql5gSt9TkLbFmWoBxFKZV6vi64diYxu4bE7DruGLfE7BoNiVmG+IQQQtiSJCghhBC21BgS1FtWB3ARJGbXkJhdxx3jlphd46Jjdvs5KCGEEI1TYziDEkII0QhJghJCCGFLbpuglFKjlFLblFI7lVKPWR1PXSil9iil0pVSaUqpVKvjqYlSappSKlcptbHac2FKqZ+UUjsq75tbGePZaoj5SaVUduXxTlNKXWFljGdTSsUppRYopTYrpTYppR6ofN62x/oCMdv2WCul/JRSq5RS6ytjfqry+dZKqZWV3yGfKKV8rI61ygVifl8ptbvacU6xONRzKKU8lVLrlFLfVj6+6OPslglKKeUJvA6MBpKBCUqpZGujqrPhWusUm1/L8D4w6qznHgPmaa2TgHmVj+3kfc6NGeDflcc7RWv9vYtjqk0Z8LDWOhnoD9xT+f+xnY91TTGDfY/1KeBSrXV3IAUYpZTqDzyPibkdcAS43boQz1FTzACPVjvOaVYFeAEPAFuqPb7o4+yWCQroC+zUWmdorUuAmcA4i2NqNLTWi4HDZz09Dvig8ucPgKtcGVNtaojZ1rTWB7TWayt/LsT8o26FjY/1BWK2LW0cr3zoXXnTwKXAZ5XP2+041xSzrSmlYoErgXcqHysacJzdNUG1AvZVe5yFzf+RVNLAj0qpNUqpO60Opp6itdYHKn8+CERbGUw93KuU2lA5BGibobKzKaUSgR7AStzkWJ8VM9j4WFcOO6UBucBPwC7gqNa6rPIttvsOOTtmrXXVcX628jj/Wynla12E5/US8CegovJxOA04zu6aoNzVYK11T8zQ5D1KqaFWB3QxtLk2wfZ/zQFvAm0xQyQHgP+zNJoaKKUCgc+BB7XWBdVfs+uxPk/Mtj7WWutyrXUKEIsZgelobUS1OztmpVQX4C+Y2PsAYcCfrYvw15RSY4BcrfUaR23TXRNUNhBX7XFs5XO2prXOrrzPBb7E/ENxFzlKqZYAlfe5FsdTK611TuU/8grgbWx4vJVS3pgv+o+01l9UPm3rY32+mN3hWANorY8CC4ABQKhSyqvyJdt+h1SLeVTlEKvWWp8C3sNex3kQMFYptQcz7XIp8DINOM7umqBWA0mV1SE+wI3AbItjuiClVIBSKqjqZ+A3wMYLf8pWZgO3Vv58K/C1hbHUSdWXfKWrsdnxrhyffxfYorV+sdpLtj3WNcVs52OtlIpUSoVW/twMGIGZO1sAXFf5Nrsd5/PFvLXaHy4KM5djm+Ostf6L1jpWa52I+U6er7WeSAOOs9t2kqgsY30J8ASmaa2ftTaiC1NKtcGcNQF4AR/bNWal1AxgGKZNfg7wv8BXwCwgHrNMyg1aa9sUJdQQ8zDMkJMG9gC/rza3Yzml1GBgCZDOmTH7v2LmdGx5rC8Q8wRseqyVUt0wk/OemD/KZ2mtn678NzkTM1S2DphUeWZiuQvEPB+IBBSQBvyhWjGFbSilhgGPaK3HNOQ4u22CEkII0bi56xCfEEKIRk4SlBBCCFuSBCWEEMKWJEEJIYSwJUlQQgghbEkSlBBCCFuSBCWEEMKW/j/M4WoNjx0/UAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "\n",
    "history_frame = pd.DataFrame(history.history)\n",
    "history_frame.loc[:, ['loss', 'val_loss']].plot()\n",
    "history_frame.loc[:, ['binary_accuracy', 'val_binary_accuracy']].plot();"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f70ae4d6",
   "metadata": {
    "papermill": {
     "duration": 0.012133,
     "end_time": "2022-05-05T17:57:13.566696",
     "exception": false,
     "start_time": "2022-05-05T17:57:13.554563",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "This model is much smaller than the VGG16 model from Lesson 1 -- only 3 convolutional layers versus the 16 of VGG16. It was nevertheless able to fit this dataset fairly well. We might still be able to improve this simple model by adding more convolutional layers, hoping to create features better adapted to the dataset. This is what we'll try in the exercises.\n",
    "\n",
    "# Conclusion #\n",
    "\n",
    "In this tutorial, you saw how to build a custom convnet composed of many **convolutional blocks** and capable of complex feature engineering. \n",
    "\n",
    "# Your Turn #\n",
    "\n",
    "In the exercises, you'll create a convnet that performs as well on this problem as VGG16 does -- without pretraining! [**Try it now!**](https://www.kaggle.com/kernels/fork/11989565)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "66c18ac8",
   "metadata": {
    "papermill": {
     "duration": 0.011681,
     "end_time": "2022-05-05T17:57:13.590559",
     "exception": false,
     "start_time": "2022-05-05T17:57:13.578878",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "---\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "*Have questions or comments? Visit the [course discussion forum](https://www.kaggle.com/learn/computer-vision/discussion) to chat with other learners.*"
   ]
  }
 ],
 "metadata": {
  "jupytext": {
   "formats": "md,ipynb"
  },
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.7.12"
  },
  "papermill": {
   "default_parameters": {},
   "duration": 225.476476,
   "end_time": "2022-05-05T17:57:17.291055",
   "environment_variables": {},
   "exception": null,
   "input_path": "__notebook__.ipynb",
   "output_path": "__notebook__.ipynb",
   "parameters": {},
   "start_time": "2022-05-05T17:53:31.814579",
   "version": "2.3.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}