gender_prediction.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "BHdx90Z6pB6Z"
   },
   "source": [
    "In the following notebook, I have developed models for predicting the gender of a person based on an image.\n",
    "\n",
    "The initial sections are dedicated to configuring the environment. The sections labeled *Loading the data* and *Preparing the first model* represent the initial approaches developed over the past few weeks. The subsequent sections (*Second approach*) utilize a different training and validation set, where the training set was split to increase the number of samples in the validation set. This was done to enhance the representativeness of the validation set, as I found the initially provided 1000 samples to be insufficient for validating the models. This adjustment resulted in improved results."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "lEf75b1opB6b"
   },
   "source": [
    "# GPU Acceleration"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Checks if the computer has any GPU avaliable."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "id": "qtoUr3hQpB6d",
    "outputId": "3275c412-dbb8-4f09-e59e-352f4274db93"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-03-16 01:21:49.061380: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:939] 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",
      "2023-03-16 01:21:49.097467: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:939] 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",
      "2023-03-16 01:21:49.098251: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:939] 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"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[PhysicalDevice(name='/physical_device:GPU:0', device_type='GPU')]"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "tf.config.list_physical_devices('GPU')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "lr0wmkEud4CX"
   },
   "source": [
    "# If using collab"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "t0lXEqzqd4CZ"
   },
   "source": [
    "Following cell mounts your google drive into the collab enviroment."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "Ji-wB1YUd4Cb",
    "outputId": "267902c5-8701-4292-e558-bbf60aff660e"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mounted at /content/drive/\n"
     ]
    }
   ],
   "source": [
    "from google.colab import drive\n",
    "drive.mount('/content/drive/')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "j8ynQv7Od4Cc"
   },
   "source": [
    "For this notebook, I upload the dataset zip files into a folder in my drive called *datasets* . "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "zdoCJq-Ed4Cf"
   },
   "outputs": [],
   "source": [
    "%cd /content/drive/MyDrive/datasets/\n",
    "!ls"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "mqUio2g9d4Ch"
   },
   "source": [
    "Once uploaded, you can unzip the files using the following cells."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "G3gRZZU9d4Ck"
   },
   "outputs": [],
   "source": [
    "# For test set \n",
    "!unzip faces_dataset_test.zip\n",
    "# For train set (it takes around 15 mins but only needs to be executed once)\n",
    "!unzip faces_dataset_train.zip"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "XzY2JNhkd4Cm"
   },
   "source": [
    "You can use the following cells to check that everything has been decompresed correctly. I just check that the *metadata.json* file exists in the decompressed folders."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "zQx_QC85d4Co",
    "outputId": "f485cec6-ffbc-4850-a2bc-da3f166c8949"
   },
   "outputs": [],
   "source": [
    "%cd /content/drive/MyDrive/datasets/faces_dataset_train/\n",
    "%ls | grep metadata\n",
    "%cd /content/drive/MyDrive/datasets/faces_dataset_test/\n",
    "%ls | grep metadata"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "g41W77kDe-e2",
    "outputId": "1e6604c3-09a2-4776-aefd-26d035feb07f"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "faces_dataset_test\tfaces_dataset_train\n",
      "faces_dataset_test.zip\tfaces_dataset_train.zip\n"
     ]
    }
   ],
   "source": [
    "!ls"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "nop4n5qhd4Cq"
   },
   "source": [
    "# Setting paths"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "_msZivU_d4Cs"
   },
   "source": [
    "### Execute this cell if you are using collab with the corresponding paths:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "91PL-LC0d4Cs"
   },
   "outputs": [],
   "source": [
    "# Dataset path\n",
    "from os import listdir\n",
    "import os.path\n",
    "\n",
    "PATH = '/content/drive/MyDrive/Datasets/faces_dataset_train/'\n",
    "PATH_TEST = '/content/drive/MyDrive/Datasets/faces_dataset_test/'\n",
    "\n",
    "assert os.path.exists(PATH), 'Verify the above commands'\n",
    "assert os.path.exists(PATH_TEST), 'Verify the above commands'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "ZtpqZ34Hd4Ct"
   },
   "source": [
    "### Or if you are using a local enviroment, use your local paths:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "id": "oZSAcVZq5AC9"
   },
   "outputs": [],
   "source": [
    "# Dataset path\n",
    "from os import listdir\n",
    "import os.path\n",
    "\n",
    "PATH = 'faces_dataset_train/'\n",
    "PATH_TEST = 'faces_dataset_test/'\n",
    "\n",
    "assert os.path.exists(PATH), 'Verify the above commands'\n",
    "assert os.path.exists(PATH_TEST), 'Verify the above commands'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "z_M7eE7Yd4Cy"
   },
   "source": [
    "# Loading the data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "7NjzTnkrzNGs"
   },
   "source": [
    "### Metadata access"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "lgQyEdGXd4C1"
   },
   "source": [
    "Accesing the metadata.json file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "SciKNCwQzLoy",
    "outputId": "46c0c792-d336-4407-89e3-53b6a72ab953"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "i07/nm0322407_rm852269056_1955-3-29_2005.jpg ----> {'age': 50, 'gender': 'M'}\n"
     ]
    }
   ],
   "source": [
    "import json\n",
    "\n",
    "with open(PATH + 'metadata.json') as fIn:\n",
    "    metadata = json.loads(fIn.read())\n",
    "with open(PATH_TEST + 'metadata.json') as fIn:\n",
    "    metadata_test = json.loads(fIn.read())\n",
    "\n",
    "metadata_example_key = next(iter(metadata.keys()))\n",
    "metadata_example_value = metadata[metadata_example_key]\n",
    "\n",
    "print(metadata_example_key, end=' ----> ')\n",
    "print(metadata_example_value)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "7m5BGd5BboRf"
   },
   "source": [
    "### Image reading"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "obSyck4wd4C6"
   },
   "source": [
    "The following enviroment variable is just to skip some errors that occurs in a Mac enviroment"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "q79TewH1d4C6"
   },
   "outputs": [],
   "source": [
    "import os    \n",
    "os.environ['KMP_DUPLICATE_LIB_OK'] = 'True'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "T0PLWSCUd4C7"
   },
   "source": [
    "Functions to process the data (image size scaled from 200x200 to 64x64)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "C11_Qk_RWYoz"
   },
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "from matplotlib import pyplot\n",
    "\n",
    "# Size at which images will be processed\n",
    "img_height = 64\n",
    "img_width = 64\n",
    "\n",
    "def decode_img(img):\n",
    "    '''Transforms a string that represents the path of an image into a matrix\n",
    "    (img_height, img_width, 3) of uint8 representing the pixel values\n",
    "    '''\n",
    "    # Convert the compressed string to a 3D uint8 tensor\n",
    "    img = tf.io.decode_jpeg(img, channels=3)\n",
    "    # Resize the image to the desired size.\n",
    "    return tf.cast(tf.image.resize(img, [img_height, img_width]),tf.uint8)\n",
    "\n",
    "def process_path(file_path):\n",
    "    '''Transforms strings of the form 'path|age|gender' to tuples (image, label)\n",
    "    '''\n",
    "    parts = tf.strings.split(file_path, '|')\n",
    "    print(parts)\n",
    "    gender = tf.cast(0 if parts[2] == 'M' else 1, tf.dtypes.int32)\n",
    "    # Load the raw data from the file as a string\n",
    "    img = tf.io.read_file(parts[0])\n",
    "    img = decode_img(img)\n",
    "    return img, gender\n",
    "\n",
    "def format_metadata(m, path):\n",
    "    '''Transforms the metadata dictionary m into a list of strings of the form 'path to the image|age|gender'.\n",
    "    '''\n",
    "    return list(map(lambda x: '{0}{1}|{2}|{3}'.format(path,x,m[x]['age'],m[x]['gender']),m.keys()))\n",
    "\n",
    "# We build a dataset of text strings from the metadata\n",
    "train_metadata = format_metadata(metadata, PATH)\n",
    "train_ds = tf.data.Dataset.from_tensor_slices(train_metadata)\n",
    "print(train_ds.take(1))\n",
    "\n",
    "# Mapping to an image dataset with tags\n",
    "train_ds = train_ds.shuffle(len(train_ds)).map(process_path)\n",
    "\n",
    "# Same process for test\n",
    "test_ds = tf.data.Dataset.from_tensor_slices(format_metadata(metadata_test, PATH_TEST))\n",
    "test_ds = test_ds.map(process_path)\n",
    "\n",
    "# Checking\n",
    "for image, label in train_ds.take(5):\n",
    "    print(\"Image shape: \", image.numpy().shape)\n",
    "    print(\"Label: \", label.numpy())\n",
    "    pyplot.imshow(image.numpy())\n",
    "    pyplot.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "hYx17xd8jDp-",
    "outputId": "08905d64-483b-4a86-82a7-8341db7ee2e0"
   },
   "outputs": [],
   "source": [
    "ds_train_transform = train_ds\n",
    "ds_test_transform = test_ds\n",
    "print(train_metadata[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "l73h5Y3fd4C_"
   },
   "source": [
    "Rescaling input"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "EtsmgEMAil8H"
   },
   "outputs": [],
   "source": [
    "# OPTIONAL: rescaling input to [-1, 1]\n",
    "ds_train_transform = ds_train_transform.map(lambda image, label: (tf.cast(image, tf.float32)/127.5 - 1., label))\n",
    "ds_test_transform = ds_test_transform.map(lambda image, label: (tf.cast(image, tf.float32)/127.5 - 1., label))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "90iw7HLJd4DA"
   },
   "source": [
    "Applying stratification"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "2mdFmqMopB6v"
   },
   "source": [
    "Following code counts the number of samples of each class"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "cY7mRJcmd4DA",
    "outputId": "9dc391bf-6ff1-4b5d-80fc-fa0eac75d699"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Class 0 --> 74253\n",
      "Class 1 --> 52612\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "\n",
    "labels = [0 if x.split('|')[2] == 'M' else 1 for x in train_metadata]\n",
    "class_counts = np.unique(labels, return_counts=True)\n",
    "total = 0\n",
    "for c, count in zip(class_counts[0], class_counts[1]):\n",
    "    total += count\n",
    "    print(f\"Class {c} --> {count}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "lgiNRV9MhIPQ",
    "outputId": "7a492d14-1fd3-4015-8377-d1ac0a924f8b"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tf.Tensor([0.8294171 1.1705829], shape=(2,), dtype=float64)\n"
     ]
    }
   ],
   "source": [
    "# OPTIONAL: sample gender weights\n",
    "import numpy as np\n",
    "labels = [0 if x.split('|')[2] == 'M' else 1 for x in train_metadata]\n",
    "freqs = np.array([1. - np.mean(labels), np.mean(labels)])\n",
    "i_freqs = 1. / freqs\n",
    "sample_weights = tf.convert_to_tensor(len(i_freqs) * i_freqs / i_freqs.sum())\n",
    "print(sample_weights)\n",
    "ds_train_transform = ds_train_transform.map(lambda image, label: (image, label, sample_weights[label]))\n",
    "ds_test_transform = ds_test_transform.map(lambda image, label: (image, label, sample_weights[label]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "wn6Ja7qlpB6w",
    "outputId": "fb0c85ae-6cf3-4b4b-fec2-a2f5de277c5f"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000\n"
     ]
    }
   ],
   "source": [
    "print(len(ds_test_transform))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "m15xf03dd4DM"
   },
   "source": [
    "Define batches"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "x0vHwUzTAjJ9"
   },
   "outputs": [],
   "source": [
    "batch_size = 128\n",
    "ds_train_batch = ds_train_transform.batch(batch_size)\n",
    "ds_test_batch = ds_test_transform.batch(batch_size)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "B7yZ2d9Yd4DO"
   },
   "source": [
    "# Preparing the first model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "8OnLpWm-d4DP"
   },
   "source": [
    "A first simple model is constructed now to have an starting point in terms of complexity and performance. The model is constructed using the Functional API to build the Model class. \n",
    "\n",
    "- Input feature map has a shape of 32x32 with 3 color channels (depth). \n",
    "- Input is processed with a combination of Conv2D layers with increasing number of filters and MaxPooling layers to reduce dimensionality. \n",
    "- For Conv2D layers, a fixed kernel size of 3 was used and ReLu as activation function (only positive values).\n",
    "- At last, a Flatten() layer output is sended to an output Dense layer with 1 neuron to learn the threshold to classify on Male or Female following a sigmoid activation function mapping.\n",
    "- For this first approach, no regularization techniques had been applied."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "KQPRRm70d4DP"
   },
   "outputs": [],
   "source": [
    "from tensorflow import keras\n",
    "from tensorflow.keras import layers\n",
    "\n",
    "inputs = keras.Input(shape=(64,64,3))\n",
    "x = layers.Conv2D(filters=32, kernel_size = 3, activation=\"relu\")(inputs)\n",
    "x = layers.MaxPooling2D(pool_size=2)(x)\n",
    "x = layers.Conv2D(filters=64, kernel_size = 3, activation=\"relu\")(x)\n",
    "x = layers.MaxPooling2D(pool_size=2)(x)\n",
    "x = layers.Conv2D(filters=128, kernel_size = 3, activation=\"relu\")(x)\n",
    "x = layers.Flatten()(x)\n",
    "\n",
    "outputs = layers.Dense(1, activation=\"sigmoid\")(x)\n",
    "model = keras.Model(inputs=inputs, outputs=outputs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "iomUKijnd4DQ",
    "outputId": "c5360259-af4c-4ada-c7f5-ba99be75ee3f"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model_3\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " input_4 (InputLayer)        [(None, 64, 64, 3)]       0         \n",
      "                                                                 \n",
      " conv2d_9 (Conv2D)           (None, 62, 62, 32)        896       \n",
      "                                                                 \n",
      " max_pooling2d_6 (MaxPooling  (None, 31, 31, 32)       0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " conv2d_10 (Conv2D)          (None, 29, 29, 64)        18496     \n",
      "                                                                 \n",
      " max_pooling2d_7 (MaxPooling  (None, 14, 14, 64)       0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " conv2d_11 (Conv2D)          (None, 12, 12, 128)       73856     \n",
      "                                                                 \n",
      " flatten_3 (Flatten)         (None, 18432)             0         \n",
      "                                                                 \n",
      " dense_3 (Dense)             (None, 1)                 18433     \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 111,681\n",
      "Trainable params: 111,681\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "740o4jdVd4DR"
   },
   "source": [
    "For a first approach, 5 epochs are used just to study the evolution of the model during training."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "waZBnIa-d4DR"
   },
   "outputs": [],
   "source": [
    "num_epochs = 10"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "xphK1aTbd4DV"
   },
   "source": [
    "The Adam optimizer was chosen as a default approach for his adaptative nature. For loss function, BinaryCrossentropy loss is the selected one for being a binary classification task. For this first approach, no *weighted_metrics* were used.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "FVJc-aCad4DV"
   },
   "outputs": [],
   "source": [
    "import tensorflow\n",
    "from tensorflow import keras\n",
    "from tensorflow.keras.optimizers import Adam\n",
    "\n",
    "model.compile(optimizer=Adam(),\n",
    "              loss=tf.keras.losses.BinaryCrossentropy(),\n",
    "              metrics=['accuracy'],\n",
    "              weighted_metrics = [])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "vDhbO-a3d4DZ",
    "outputId": "6478e252-5916-41cc-9bae-284141ef1b99"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "992/992 [==============================] - 6s 6ms/step - loss: 0.3625 - accuracy: 0.8446 - val_loss: 0.3462 - val_accuracy: 0.8640\n",
      "Epoch 2/10\n",
      "992/992 [==============================] - 6s 6ms/step - loss: 0.2921 - accuracy: 0.8893 - val_loss: 0.3589 - val_accuracy: 0.8520\n",
      "Epoch 3/10\n",
      "992/992 [==============================] - 6s 6ms/step - loss: 0.2751 - accuracy: 0.8970 - val_loss: 0.3214 - val_accuracy: 0.8670\n",
      "Epoch 4/10\n",
      "992/992 [==============================] - 6s 6ms/step - loss: 0.2628 - accuracy: 0.9026 - val_loss: 0.3264 - val_accuracy: 0.8660\n",
      "Epoch 5/10\n",
      "992/992 [==============================] - 6s 6ms/step - loss: 0.2516 - accuracy: 0.9064 - val_loss: 0.3368 - val_accuracy: 0.8600\n",
      "Epoch 6/10\n",
      "992/992 [==============================] - 6s 6ms/step - loss: 0.2400 - accuracy: 0.9104 - val_loss: 0.3334 - val_accuracy: 0.8750\n",
      "Epoch 7/10\n",
      "992/992 [==============================] - 6s 6ms/step - loss: 0.2285 - accuracy: 0.9140 - val_loss: 0.3360 - val_accuracy: 0.8620\n",
      "Epoch 8/10\n",
      "992/992 [==============================] - 6s 6ms/step - loss: 0.2154 - accuracy: 0.9168 - val_loss: 0.3622 - val_accuracy: 0.8580\n",
      "Epoch 9/10\n",
      "992/992 [==============================] - 6s 6ms/step - loss: 0.2050 - accuracy: 0.9190 - val_loss: 0.3566 - val_accuracy: 0.8640\n",
      "Epoch 10/10\n",
      "992/992 [==============================] - 6s 6ms/step - loss: 0.1947 - accuracy: 0.9218 - val_loss: 0.3823 - val_accuracy: 0.8740\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(ds_train_batch,\n",
    "                    epochs=num_epochs,\n",
    "                    validation_data=ds_test_batch)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "RyKzGBT-d4Db"
   },
   "source": [
    "We are gonna create a models folder and save the models there."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "jbFIfW5Ud4Dc",
    "outputId": "98d4b9b9-6376-4a55-ab13-a5eef49c135c"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/bin/bash: /home/mia1/anaconda3/envs/tf/lib/libtinfo.so.6: no version information available (required by /bin/bash)\n",
      "mkdir: no se puede crear el directorio «models»: El archivo ya existe\n"
     ]
    }
   ],
   "source": [
    "!mkdir models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "ENAWlHIad4Dd"
   },
   "outputs": [],
   "source": [
    "# Save the model\n",
    "model.save('models/first_model.h5')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "gzmxHNdRd4Dd"
   },
   "source": [
    "The following code loads the model from the created folder:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "kN8rV6hAd4De"
   },
   "outputs": [],
   "source": [
    "# Load the model\n",
    "from tensorflow.keras.models import load_model\n",
    "loaded_model = load_model('models/first_model.h5')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "wjf7DHj0d4De"
   },
   "source": [
    "Plotting the results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "AfMgKqild4Df",
    "outputId": "dadddff6-3402-45ea-9bf1-b8ba8ad61b96"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkkAAAGwCAYAAAC99fF4AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjYuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8o6BhiAAAACXBIWXMAAA9hAAAPYQGoP6dpAABsj0lEQVR4nO3deVzVVf7H8dcFZBVQREBlEbfccAEU0TQrNc0sl3LJTNudsslxmt/ktIw5ltW0L1o2lVmJVmpZWaaViplLiPuSuQAiiKACorLd+/vjm9dQVES4X5b38/G4D7nnfvl+P1fU+/ac8z3HYrPZbIiIiIhICU5mFyAiIiJSFSkkiYiIiJRCIUlERESkFApJIiIiIqVQSBIREREphUKSiIiISCkUkkRERERK4WJ2AdWV1Wrl0KFDeHt7Y7FYzC5HREREysBms5Gbm0vjxo1xcrp4X5FCUjkdOnSIkJAQs8sQERGRckhJSSE4OPiixygklZO3tzdg/Cb7+PiYXI2IiIiURU5ODiEhIfbP8YtRSCqnM0NsPj4+CkkiIiLVTFmmymjitoiIiEgpFJJERERESqGQJCIiIlIKzUkSERH5k+LiYgoLC80uQ8qpTp06ODs7V8i5FJJEREQw1s9JT0/n+PHjZpciV6hevXoEBQVd8TqGCkkiIiJgD0gBAQF4enpqoeBqyGazcfLkSTIyMgBo1KjRFZ1PIUlERGq94uJie0Bq0KCB2eXIFfDw8AAgIyODgICAKxp608RtERGp9c7MQfL09DS5EqkIZ36OVzq3TCFJRETkDxpiqxkq6ueokCQiIiJSCoUkERERkVIoJImIiAgATZs25dVXX62Qc61YsQKLxVKtl1TQ3W0iIiLVWO/evenUqVOFhJsNGzbg5eV15UVVhNPZ4OYNFvP6cxSSREREajCbzUZxcTEuLpf+yG/YsKEDKroEmw1OHIbcNPDwg3qhYNKEeg23iYiIlMJms3GyoMjhD5vNVuYax40bx8qVK3nttdewWCxYLBZmz56NxWJh6dKlREdH4+bmRnx8PHv37uWWW24hMDCQunXr0qVLF5YvX17ifOcOt1ksFv73v/8xZMgQPD09admyJYsXLy737+mCBQto164dbm5uNG3alJdeeqnE6zPeeouWLZrh3jCMwI59uHXcg/bXPv/8cyIiIvDw8KBBgwb06dOHvLy8ctdSFupJEhERKcWpwmLaPrXU4dfdMfUGPF3L9vH82muv8dtvv9G+fXumTp0KwPbt2wH4v//7P1588UWaNWtGvXr1OHjwIDfeeCPTpk3D3d2dDz/8kEGDBrF7925CQ0MveI2nn36aF154gf/+97+88cYbjB49mqSkJPz8/C7rfSUkJDB8+HCmTJnCiBEjWLNmDQ8++CANGjRg3Lhx/LphPX995BE+ev0/dI/uwNFCd+ITtoPFQlpaGqNGjeKFF15gyJAh5ObmEh8ff1mBsjwUkkRERKopX19fXF1d8fT0JCgoCIBdu3YBMHXqVPr27Ws/tkGDBnTs2NH+fNq0aSxatIjFixczYcKEC15j3LhxjBo1CoBnn32WN954g/Xr19O/f//LqvXll1/m+uuv58knnwSgVatW7Nixg//+97+Mu3MMyTt+xcvTg5v69MQ7uC1hnn507nE9AGlpaRQVFTF06FDCwsIAiIiIuKzrl4dCkoiISCk86jizY+oNply3IkRHR5d4npeXx9NPP83XX3/NoUOHKCoq4tSpUyQnJ1/0PB06dLB/7eXlhbe3t31vtMuxc+dObrnllhJtPXr04NVXX6U44zf6xnYgLLgRzXoMpn//AfTv398+zNexY0euv/56IiIiuOGGG+jXrx+33nor9evXv+w6LofmJImIiJTCYrHg6eri8EdFrRZ97l1q//jHP1iwYAHPPPMM8fHxbNq0iYiICAoKCi56njp16pz3+2K1Wi+7HpvNdt57sxUXAjYozMPb24eNG9YTFzePRo0a8dRTT9GxY0eOHz+Os7Mzy5Yt49tvv6Vt27a88cYbXHXVVezfv/+y67gcCkkiIiLVmKurK8XFxZc8Lj4+nnHjxjFkyBAiIiIICgriwIEDlV/gH9q2bcvq1avPNhTls+aHJbRqFoZzHTfwb4mLVz369OnDCy+8wJYtWzhw4AA//vgjYISzHj168PTTT5OYmIirqyuLFi2q1Jo13CYiIlKNNW3alHXr1nHgwAHq1q17wV6eFi1asHDhQgYNGoTFYuHJJ58sV49Qef3973+nS5cu/Oc//2HEkJv55cevePP9OGY89wT4t+Lr75axb98+evXqRf369VmyZAlWq5WrrrqKdevW8cMPP9CvXz8CAgJYt24dR44coU2bNpVas3qSREREqrFHH30UZ2dn2rZtS8OGDS84x+iVV16hfv36dO/enUGDBnHDDTcQGRnpsDojIyP59NNPmRc3l/aR0Tz1wltM/efDjJvwf+DiRr169Vi4cCHXXXcdbdq04e233yYuLo527drh4+PDqlWruPHGG2nVqhVPPPEEL730EgMGDKjUmi22yr5/robKycnB19eX7OxsfHx8zC5HRESuwOnTp9m/fz/h4eG4u7ubXU7NdTobju4HbODqBX7NwKniB7Uu9vO8nM9v03uSZsyYYX8TUVFRxMfHX/DY1atX06NHDxo0aICHhwetW7fmlVdeKXFM79697Qtq/fkxcOBA+zFTpkw57/Uzt06KiIhIJTiZBUf3ATZw8wG/FpUSkCqSqdXNnz+fiRMnMmPGDHr06ME777zDgAED2LFjR6kLW3l5eTFhwgQ6dOiAl5cXq1ev5oEHHsDLy4v7778fgIULF5aYqZ+VlUXHjh257bbbSpyrXbt2JVYadXaumFsuRUREaoPx48fz8ccfl/raHXfcwdtvv3224UQG5KQaX3v4Qb0QU/dkKytTh9tiYmKIjIxk5syZ9rY2bdowePBgpk+fXqZzDB06FC8vLz766KNSX3/11Vd56qmnSEtLs98OOWXKFL744gs2bdpU7to13CYiUnNouO3yZWRkkJOTU+prPj4+BAQEGPuw5aYZe7EBeAWAT+NK34utoobbTOtJKigoICEhgccee6xEe79+/VizZk2ZzpGYmMiaNWuYNm3aBY957733GDly5HnrRezZs4fGjRvj5uZGTEwMzz77LM2aNbvgefLz88nPz7c/v9AfDBERkdogICDACEIXYrNBdooxzAbg3QjqBpq2WW15mNbXlZmZSXFxMYGBgSXaAwMDSU9Pv+j3BgcH4+bmRnR0NA899BD33ntvqcetX7+ebdu2nfd6TEwMc+bMYenSpbz77rukp6fTvXt3srKyLnjN6dOn4+vra3+EhISU8Z2KiIjUMlYrHNt/NiD5hoB3ULUKSFAFJm6ft/pmKStynis+Pp5ff/2Vt99+m1dffZW4uLhSj3vvvfdo3749Xbt2LdE+YMAAhg0bRkREBH369OGbb74B4MMPP7zgNSdPnkx2drb9kZKSUpa3JyIiUrtYi+HoXuNONixQPxy8/M2uqlxMG27z9/fH2dn5vF6jjIyM83qXzhUeHg4Ym9sdPnyYKVOm2DffO+PkyZPMmzfPvivyxXh5eREREcGePXsueIybmxtubm6XPJeIiEitVVxoBKTCU8bEbL9m4OZtdlXlZlpPkqurK1FRUSxbtqxE+7Jly+jevXuZz2Oz2UrMFTrj008/JT8/nzvuuOOS58jPz2fnzp00atSozNcVERGRPykqgMw9RkBycoEGLap1QAKTlwCYNGkSY8aMITo6mtjYWGbNmkVycjLjx48HjCGu1NRU5syZA8Bbb71FaGgorVu3Box1k1588UUefvjh88793nvvMXjwYBo0aHDea48++iiDBg0iNDSUjIwMpk2bRk5ODmPHjq3EdysiIlJDFZ6CrL1gLQSnOkZAqlP97xI0NSSNGDGCrKwspk6dSlpaGu3bt2fJkiWEhYUBkJaWVmJ5davVyuTJk9m/fz8uLi40b96c5557jgceeKDEeX/77TdWr17N999/X+p1Dx48yKhRo8jMzKRhw4Z069aNtWvX2q8rIiJSWzRt2pSJEycyceLESx5rsVhYtGgRgwcPPttYkGcEJFsxuLiDX3Nwca20eh3J9KUuH3zwQR588MFSX5s9e3aJ5w8//HCpvUbnatWqFRdb/mnevHmXVaOIiIiU4nSOcRebzQp1PI2A5Gx6tKgwNeediIiIiOOcOgbHkjD2YfMGv3Bwqlm7V5i+BICIiEiVZLMZQ0mOflzGRhjvvPMOTZo0wWq1lmi/+eabGTt2LHv37uWWW24hMDCQunXr0qVLlxJbcpVb3hE4doCtO3/juhEP4RESQYOGAdx///2cOHHCftiKFSvo2rUrXl5e1KtXjx49epCUlATA5s2bufbaa/H29sbHx4eoqCh+/fXXK6+tAqknSUREpDSFJ+HZxo6/7r8OgavXpY8DbrvtNv7617/y008/cf311wNw7Ngxli5dyldffcWJEye48cYbmTZtGu7u7nz44YcMGjSI3bt3l7pHapmcOgbZBzl56hT9x/yVbrE92LBhFhkZGdx7771MmDCB2bNnU1RUxODBg7nvvvuIi4ujoKCA9evX29dCHD16NJ07d2bmzJk4OzuzadMm6tSpU76aKolCkoiISDXl5+dH//79mTt3rj0kffbZZ/j5+XH99dfj7OxMx44d7cdPmzaNRYsWsXjxYiZMmHB5FzvTw3XqGACffLOGU6cLmDNnjn3rrzfffJNBgwbx/PPPU6dOHbKzs7npppto3rw5YOzPekZycjL/+Mc/7Hest2zZsly/B5VJIUlERKQ0dTyNXh0zrnsZRo8ezf3338+MGTNwc3Pjk08+YeTIkTg7O5OXl8fTTz/N119/zaFDhygqKuLUqVMl7hwvE5sVjv/pe3yasPPAITp27Fhib9QePXpgtVrZvXs3vXr1Yty4cdxwww307duXPn36MHz4cPuahJMmTeLee+/lo48+ok+fPtx22232MFVVaE6SiIhIaSwWY9jL0Y/L3N9s0KBBWK1WvvnmG1JSUoiPj7cvpPyPf/yDBQsW8MwzzxAfH8+mTZuIiIigoKCg7BewFsPR/fYeJLwaQt2Ai24jdqb9gw8+4JdffqF79+7Mnz+fVq1asXbtWgCmTJnC9u3bGThwID/++CNt27Zl0aJFl/XeK5tCkoiISDXm4eHB0KFD+eSTT4iLi6NVq1ZERUUBxl6n48aNY8iQIURERBAUFMSBAwfKfvLiImMNpPwc7JHhj1W027Zty6ZNm8jLy7Mf/vPPP+Pk5ESrVq3sbZ07d2by5MmsWbOG9u3bM3fuXPtrrVq14m9/+xvff/89Q4cO5YMPPij370NlUEgSERGp5kaPHs0333zD+++/X2I7rhYtWrBw4UI2bdrE5s2buf3228+7E+6Cigsgaw8U5oHFGRqUHAobPXo07u7ujB07lm3btvHTTz/x8MMPM2bMGAIDA9m/fz+TJ0/ml19+ISkpie+//57ffvuNNm3acOrUKSZMmMCKFStISkri559/ZsOGDSXmLFUFmpMkIiJSzV133XX4+fmxe/dubr/9dnv7K6+8wt1330337t3x9/fnn//8Jzk5OZc+YdFpowepuOCPbUaaQx2PEod4enqydOlSHnnkEbp06YKnpyfDhg3j5Zdftr++a9cuPvzwQ7KysmjUqBETJkzggQceoKioiKysLO68804OHz6Mv78/Q4cO5emnn67Q35crZbFdbGlquaCcnBx8fX3Jzs7Gx8fH7HJEROQKnD59mv379xMeHo67e/Xfc+yKFJyEo3vBWgTObkZAcnEzu6rLcrGf5+V8fqsnSURERAz5ucYkbVsxuHgYAcm5aq1d5EgKSSIiIsInH8zigQkT/3hmKXGXXVhYGNu3bzelLjMpJImIiNR2J7O4uUdbYr6PM/Zh8w0Gp7P3dlW1lbAdRSGpirrY+hMiIiIV5sRhyDmEd10vvBuGQL3Qy16rqabSEgBVzIYDR7ln9gZmrtxrdikiIrWH1QrFhViLi8yuxHFsNshJhZw/VhX3CqgxAanMyxxcgnqSqpjkrJP8sCuDXem5jO/VHCen6v+HVUSkSis4ieuiu3FqdDOHik7Q0M8XV3dPLHXcjcnLLm41IjiUYLNBbhqcPm489woAtwaQn29qWVfKZrNRUFDAkSNHcHJywtXV9YrOp5BUxQzs0Iinv9pO6vFTxP+eyTWtGppdkohIzXU6G+aOxCl5DeEpG0hrfTeH8jqB058+Hi0WcHYt+XByqb7ByWaDk1lQeBKwgGd9yMsFcs2urMJ4enoSGhqKk9OVDZgpJFUx7nWcGRoZzOw1B5i3PlkhSUSksuRlwsdDIW0zuPngOnIOoXUDKTq0heLDW+DILjiy848wcQ4XT2h4FQS0hYA2xsM3pMRk5yop/wQseRRSfwUnV7jhGWgeYXZVFcrZ2RkXF5cKmderkFQFjewawuw1B1i24zAZuacJ8K7lC5uJiFS07FT4aDBk/gaeDeCOhdC4Exagjn8L7PdyWa2Q9TscSoS0TX/8uhlOpMDx3bBn8dlzuvlAo47QuBM07mw86odXnR6nE0dg3jCjfldvGDUXwnuZXVWVppBUBbUO8qFzaD0Sk4/zecJBHuzdwuySRERqjqy9MGcwZCeDTxMY8wU0bFX6sU5OxmsNW0HHEUabtdgIV4cS4dAm49f0LcYmsAfijccZ7r7QqNPZ0NS4szmTo48lwUdDjJW0Pf3hjgVGmJOL0rYk5VTZ25J8+msK//f5FkL9PFnxaO/qOYE7/wS4uIOzsriIVBHp24ywkJcBfs3gzi+N0HKliouM4bk/9zilbzX2PjuXR/2SoalRJ2NdosoKThk7jfecmwa+oTBmEfjX3v98X87nt0JSOVV2SDpZUETMMz+Qm1/EJ/fG0KOFf4Vfo9Ic2gRr3oDtiyCsu/EXshYvay8iVUTKevjkVmOydmB7Y4jNO7DyrldUYMxp+nOP0+HtYC08/1hP/5LBqXFn8Gl05TWkrIdPbjPuYmvY2vj32KfxlZ+3GlNIcgBHbHD75Bfb+GhtEgM7NOKt2yMr5RoVxmaD35fDmtdh/6qSr3X/K/T7jzl1iYgA7P0J5t1uTMIO7gqjPzV6dBytKN8ISn/ucTq8w9gr7Vx1A8/vcbqcULdnOXw65o/33AVu/xQ8/SrqnVRbCkkO4IiQtP1QNgNfX00dZwu/TL4e/7pVcBfmonzY+hmsedP4HxOAxRnaDzMmMH7/uNF2+6fQ6gbz6hSR2mvnV/D53cbQV7NrYeQn4OpldlVnFZ46G5zO9Dgd2Qm2UhZE9G58To9TJ/AqZaRh6+ew6AGwFkHz62HER1XrPZvocj6/NVmkCmvX2JeOwb5sPpjNgoSDPHBNc7NLOuvUMfj1A1j3DpxIN9pcvSFqLHT7izG+DnA8Gda/A4vGw/jV4NvEvJpFpPbZNBe+fMgIHG0GwbD3jMUhq5I6HhAcbTzOKDhpzGn6c4/Tkd2Qewh2H4Ld35w91jek5B11Gbtg6b8Am/Ef1sFvg8uVLapYW6knqZwc0ZMEMG99Mo8t3Eq4vxc//v0a8/dzO5YEa2fCxjlQmGe0eTeGbuMhapxxJ8efFeXDe/2Mv+ShsTD2a03kFhHHWPs2fPdP4+tOo2HQ69X735/8E8ZddH/uccrac+Hju9wHA16o+ms3OZiG2xzAUSEpL7+Irs8sJ6+gmLj7uhHbvEGlXeuiDiX+MRn7i7Nj5wHtoMdfod3Qi/8v5eg+eLsXFOTC1ZOgz78dUrKI1FI2G6x8AVY8azyP+Qvc8GzNDAunsyFtS8kep2NJcM0/4Zr/qzprNFUhGm6rQbzcXLi5UxPi1icTtz7ZsSHJaj07GfvP6340uxa6PwzNryvbX0C/ZnDz6/D5XbD6ZWjaA1r0qby6RaT2stlg6eOw9i3jee/JRmCoqWHB3RfCexqPM6zWmhkITaCQVA3c3jWUuPXJfLctnaN5Bfh5VfLYclE+bPkUfnnTWPcDjH2K2g8zwlFQOZawbz/UCFq/vg8LHzDmJ1XE7a0iImdYi+Grv0Lix8bz/s8ZcyRrGwWkCqOQVA1EBPvSvokP21JzWLjxIPf2bFY5Fzp1zAgx696BE4eNNldviB4HMePPTsYurxumQ8oGOLwVFtxrLOJWnecHiEjVUZRv/LuyczFYnODmN6HzaLOrkmpOcbOaGNXVWBE2bn0yFT6N7FgSfPtPeLkd/DDVCEjejaHvf2DSdug37coDEkAdd7htNrjWhaTVsOqFKz+niEhBHsSNNAKSsyvc9qECklQIhaRq4uaOjfGo48zeI3lsOHCsYk6auhE+uwte7wTr3jbuVguMgCGz4JHNxqTsc+9Wu1L+LeCmV42vV74A+1ZU7PlFpHY5ddzYcmPvj1DHE26fD21vNrsqqSEUkqoJb/c63NzRWEp+3vrk8p/IaoXflsIHA+Hda2H7QmP9kObXGcvVj483NnGszDU1OtwGkXcCNlhwH+QerrxriUjNdSIDZt8EKeuM/9CN+cL4t0ykgigkVSOjYowht6+3pnH8ZCmbJl5M4WljbaMZ3WDucGO4y8kFOow0JlGPWVT2u9UqQv/nIaCtscnkwvuMCZciImV1PAXe72/McfRqCOO+gdAYs6uSGsb0kDRjxgzCw8Nxd3cnKiqK+Pj4Cx67evVqevToQYMGDfDw8KB169a88sorJY6ZPXs2FovlvMfp06fLfd2qomOwL20a+VBQZGVRYmrZvunkUVj1IrwaAYsfhszd4OZj7Kf2yBYY+k757la7Uq6exvykOp6wfyXEv+T4GkSkesrcYwSko3uN1abvXmrOv2NS45kakubPn8/EiRN5/PHHSUxMpGfPngwYMIDk5NKHk7y8vJgwYQKrVq1i586dPPHEEzzxxBPMmjWrxHE+Pj6kpaWVeLi7u5f7ulWFxWJhVNcQoAwTuI8dgCX/B6+0gx//Y/TY+DQxJmH/bZux4azZW4Q0vAoGvmx8vWI6HFhtbj0iUvWlbTYCUs5BaNAS7v4OGlShLZukRjF1xe2YmBgiIyOZOXOmva1NmzYMHjyY6dOnl+kcQ4cOxcvLi48++ggwepImTpzI8ePHK/W6jlpx+1zZpwqJeXY5pwutLPhLd6LCztnFOjXBWBl7x5dnN0cMjPhjZewh4FzHYbWW2RcPwqZPoG6QMfRXt6HZFYlIVZT0izFdID8HgjoY0wRK29xV5CIu5/PbtJ6kgoICEhIS6NevX4n2fv36sWbNmjKdIzExkTVr1nDNNdeUaD9x4gRhYWEEBwdz0003kZiYeMXXzc/PJycnp8TDDL4edbipgzGBO+7MBG6rFXZ/Bx/cCO9eB9sX/TEZ+3pjIuP4eOgwvGoGJIAb/wsNWxsb5S6633g/IiJ/tme5cRdbfo6xD+S4rxWQpNKZFpIyMzMpLi4mMDCwRHtgYCDp6ekX/d7g4GDc3NyIjo7moYce4t5777W/1rp1a2bPns3ixYuJi4vD3d2dHj16sGfPniu67vTp0/H19bU/QkJCLvctV5gzayZ9v+UAJ9e+DzNiIG4EJP1sTMbuOArG/wxjFkLza6v+cvyuXsb8JBcP4zben1+55LeISC2yfZGxDlLRKWjRF+5YWPHLk4iUwvTljs/d1d5ms11yp/v4+HhOnDjB2rVreeyxx2jRogWjRo0CoFu3bnTr1s1+bI8ePYiMjOSNN97g9ddfL/d1J0+ezKRJk+zPc3JyTAtKkf5W/u27hJtOf4Xnd9lGo5sPRN8FXR8wf65ReQS0MXqUFk+AH5+B0O4QFmt2VSJito1z4KtHjN7xdkOMddwqc4kSkT8xLST5+/vj7Ox8Xu9NRkbGeb085woPDwcgIiKCw4cPM2XKFHtIOpeTkxNdunSx9ySV97pubm64ubld8n1VqqP7Ye0MLIkfc1fhSbBAhsWfhn0nYokcC+6OmxtVKTrfYezvtmU+fH63MT/Jy4Eb+opI1bLmTfj+cePryDuNhWidnE0tSWoX04bbXF1diYqKYtmyZSXaly1bRvfu3ct8HpvNRn5+/kVf37RpE40aNarQ6zrUwQT4dCy8EQnrZ0HhSYoC2vP34gl0P/Uym0PGVP+ABMaw4MCXjTtWcg/BF+M1P0mkNrLZ4MdpZwNS94dh0OsKSOJwpg63TZo0iTFjxhAdHU1sbCyzZs0iOTmZ8ePHA8YQV2pqKnPmzAHgrbfeIjQ0lNatWwPGukkvvvgiDz/8sP2cTz/9NN26daNly5bk5OTw+uuvs2nTJt56660yX7dKsFrht++MO9WS/zShvEUf6P4wLuHXYP10M0WJqcStS6ZTSD3TSq1QbnWN+UnvXgd7vodf3oAej5hdlYg4itUK3/3T+A8hwHVPQs+/V/25lVIjmRqSRowYQVZWFlOnTiUtLY327duzZMkSwsLCAEhLSyuxdpHVamXy5Mns378fFxcXmjdvznPPPccDDzxgP+b48ePcf//9pKen4+vrS+fOnVm1ahVdu3Yt83VNVXgatswzupmzjCFCnOpAxG3QfQIEtrMfOqprKIsSU1m8+RBP3NQGb/cqevfa5QpqDwOeg6//BsufNu5kCel66e8TkeqtuAi+fMj4NxDgxheh633m1iS1mqnrJFVnlbZO0ornYcWzxtduvsZk7JgHwKfxeYfabDb6vLySvUfyeGZIe0bHVIGQV1FsNmNe0vaFxoq6D6wCTz+zqxKRylJ4GhbcA7u+BoszDJ5p7CMpUsGqxTpJcgFRY8GvOdzwLEzaDn2fLjUgwZkVuI3lAOKuZNPbqshigUGvgV8zyE4x/nepPC9SM+WfgLm3GQHJ2Q1GfKyAJFWCQlJV4x0EDydA7EPg5n3Jw4dGBuPq7MS21By2Hsx2QIEO5O5jzE9ydoXdS2DtzEt+i4hUMyePwpxbYP8qqOMFoz+D1jeaXZUIoJBUNV3GBEU/L1f6tw8CIG5DDetNAmjU0ehVA1j2lLHtiojUDLnpMHsgpP4KHvVh7FfQ7JpLf5+Igygk1QBnhty+TEwlL7/I5GoqQZd7oe0tYC2Ez8bBqeNmVyQiV+pYkrFRbcYOY9/GcUsgOMrsqkRKUEiqAbo18yPc34u8gmK+2nzI7HIqnsUCN78B9cLgeLKxKrfmJ4lUXxm74P0b4Nh+4+/13d9CYFuzqxI5j0JSDWCxWBjZxdgipcZN4D7D3deYn+RUB3Z+BevfNbsiESmP1I3wwQDITTM2tr57qXGDhkgVpJBUQwyLCqaOs4XNB7PZfqiGTeA+o0kk9PuP8fX3j8OhTaaWIyKX6cBq+PBmOHUUGkfCXd+CTyOzqxK5IIWkGsK/rhv92hkTuOetTzG5mkoUMx5a3wTFBcb8pNM5ZlckImXx21L4eBgU5ELTnjB2sdY+kypPIakGuf2PCdxfJKZysqAGTuAGY37SLW+Cb6gxn+Grv2p+kkhVt/VzmHc7FJ2GVgOM2/zLsMSJiNkUkmqQ2GYNCPXzJDe/iK+3pJldTuXxqA+3fQBOLrB9Efz6vtkViciF/Po+LLgXrEXG9kojPoI6HmZXJVImCkk1iJOThZFdjQnc82rqBO4zgqOhzxTj6+8mQ9oWU8sRkVKsfsXYgxEbRN8DQ2aBcw3ZY1JqBYWkGubWqGBcnCxsTD7OrvQaPl8ndgK06g/F+cb8pPxcsysSETCGwJdPMR4AV0+CgS+Bkz5ypHrRn9gaJsDbnb5tA4EaPoEbjPlJg2eCTxM4utf4H6vmJ4mYy2qFbyYZvUgAfZ6GPv++rJ0ERKoKhaQaaOQfE7gXbjzI6cJik6upZJ5+cOv7xq7hWz+DjXPMrkik9iouhEX3/zFP0AI3vQpXTzS5KJHyU0iqgXq28KdJPQ9yThexZGsNnsB9Rmg3uO4J4+tv/w8Obze3HpHaqPAUzL/D+M+Kkwvc+h5E32V2VSJXRCGpBnJysjCqaw1fgftcPSZCiz7GLcafjYP8E2ZXJJUh+yDEvwzf/N34devnkLwOcg4ZwzxijtM58PGt8Nt34OIOI+Og/TCzqxK5Yi5mFyCV47boEF5ZvocNB46x53AuLQNr+JokTk4w5B14+2rI/A2WPApD3ja7KqkIBXnGVjSb5sL+VcAF5p051QHfJuAbAvVC//g15OyvPsHg4urQ0mu0ogLIO2JsL7LkUTiUCK7ecPt8aNrD7OpEKoRCUg0V6OPOda0DWLbjMPM2pPDkTbVg80gvfxj2Hnx4E2yOg/Be0Ol2s6uS8rBaIeln4+e440so+FPPYNjVENIFctIgOwWOp0BOKlgL4dgB41EqC3gHnR+efEPBN9j4urYvcGi1wqljcOLwH4+MUr7+49dTR0t+r2cDuGMBNO5sTu0ilcBis+l2oPLIycnB19eX7OxsfHx8zC6nVD/tyuCu2Ruo51mHtZOvx72Os9klOcbK/8JP06COJ9z3EwS0NrsiKausvbB5HmyZB8f/NFRcvyl0vB06jjC+PldxkdGjcSY0ZSf/8WvK2V+LTl/6+u71zganEkHqj94pzwbV8y6tgrzSg05uesm2vAxj0ceycnIBrwDwbwk3/hcaXlV570GkglzO57d6kmqwXq0a0tjXnUPZp1m6PZ1bOjUxuyTH6DkJklbDvhXG/KT7fgRXT7Orkgs5nW2snL4pDlLWnm1384F2g41wFNrt4uHE2cUIM/VCIKyU1202yMssPTydCVWns+H0cUg/DulbS79OHU+j1+nc8HTmuXcjcHLQf0aKC43hrkv1+JzIKNkTVxYeflA3EOoGnPPrOV971NfaR1KjqSepnKpDTxLAa8v38Mry34gJ92P+A7Fml+M4JzKM+UknDkPnMcZ+b1J1WIth309GMNr19dleHosTNOsNnUbDVTc6NtyezjEmhmenGL1Y5wapE+mXPofF2Vi367xeqD8N69Vxv/D322xlH+46mXV578/FA7xLCTrnBiGvAM3dkhrtcj6/FZLKqbqEpLTsU/R47kesNvjx79fQrGFds0tynH0rYc4tgA2GvgsdhptdkWTsgs1zYcunxvDYGf5XQadR0GEE+DQ2r76LKcr/U4g6tzcq+Y95UWUYqvIKOBue3LyNHq4/hx9rYdlrsjj/EW4u0eNTNwBc61bPoUKRCqbhNrFr5OvBtVcF8MOuDOZtSOFfN7YxuyTHaXYNXPNPWPkcfDXRmFDq39Lsqmqfk0eNW/U3zzXugDrDoz60v9UIR40jq/4HuIsbNGhuPEpjLTbm+Px5CC/7YMlAVZhnzPvJy4DUhAtfy6P+pXt86gYaw2Ia7hKpNOpJKqfq0pMEsHzHYe6d8yt+Xq78Mvk63FxqyQRuMD645twCB+IhsD3cu1w7kDtCcSHsWWYEo93fne0dcXKBFn2Nuw5b3WAEj9rizFDan4fyCk6AV8NzglBA7fp9EXEw9SRJCb2vakigjxuHc/L5fvthBnWsosMZlcHJGYb9z5ifdHgbfDcZBr1qdlU1k80G6VuMeUZbP4OTmWdfC4owJmBH3AZ1G5pXo5ksFmMbHU8/aNzJ7GpEpAwUkmoBF2cnRkSH8PqPvzNvQ3LtCklgrI0zdBZ8NBQSPoDwnloNuCLlHoatnxrhKONPW8J4BRjzwDqOgqD25tUnIlJOCkm1xPAuIbzx0+/8/HsWBzLzaOrvZXZJjtX8Ouj5d4h/ERY/Ao06XXhuiVxa4WnYvcRY7PH3H8D2x0bKzq7GXWmdbofm1xu35ouIVFP6F6yWCK7vyTWtGrJi9xHmbUjhsQG1cIHF3pMhaQ0krzHWT7pn2cVvx5aSbDY4+Ksxz2jbAmNdoTOCuxg9Ru2HGpOORURqAIWkWmRkl1BW7D7C5wkpTOrbCleXWnZXjPMfO5PP7GHMnfn+CRj4otlVVX3ZB41VsDfHQdbvZ9t9mhi37He6XXcNikiNpJBUi1zfJoCG3m4cyc3nh52HGRDRyOySHM+nsTE/6ZNbYcO70PRqY1VnKelCm8q6eEDbm41eo/BejltdWkTEBApJtUgdZyeGRwfz1k97mbs+uXaGJICWfaHHI/Dza7D4YWjUEfzCza7KfFarMRS5KQ52fHH+prKdRkHbW7QJrIjUGgpJtczILqG89dNe4vdkknL0JCF+tXRPs+uehOS1kLIOPr8L7l5ae9emObrv7HDaeZvKjoKOI0vfVFZEpIZTSKplQvw86dnSn/g9mczfkMKjN9TSXbud68Cw94z1kw4lwrJ/w4DnzK7KcU5nw/YvjGCU/MvZdldvY/ix0+0QGlv1V8EWEalECkm10KiuocTvyeTTX1N4pE9L6jjXsgncZ9QLgSFvQ9xIWDfTmJ/U5iazq6o8F9pUFgs0v9ZY7LH1QMduKisiUoWZ/uk4Y8YMwsPDcXd3Jyoqivj4+Aseu3r1anr06EGDBg3w8PCgdevWvPLKKyWOeffdd+nZsyf169enfv369OnTh/Xr15c4ZsqUKVgslhKPoKCgSnl/VVGfNoH413UlIzefH3dlmF2Oua4aALETjK+/fLDkcFN1lp8LKevh1/fh60nw3g3wXBh8PAy2fW4EJP+roM8UmLQDxiyCDrcpIImI/ImpPUnz589n4sSJzJgxgx49evDOO+8wYMAAduzYQWho6HnHe3l5MWHCBDp06ICXlxerV6/mgQcewMvLi/vvvx+AFStWMGrUKLp37467uzsvvPAC/fr1Y/v27TRp0sR+rnbt2rF8+XL7c2fn2nOXjquLE7dGhfD2yr3ErU/mhna1JyCW6vp/G0NOqQnw+d1w17fGcFx1YLXC8QOQvg0Obze2Xjm8HY7tL/1493oQcasxnFYdNpUVETGRqRvcxsTEEBkZycyZM+1tbdq0YfDgwUyfPr1M5xg6dCheXl589NFHpb5eXFxM/fr1efPNN7nzzjsBoyfpiy++YNOmTeWuvTptcFuaA5l59H5xBRYLrP7ndTSpV8s3fT2WBO/0NObqdH8Y+k0zu6Lznc4pGYQOb4PDO4yd5UtTN8jYDiSwnbG5b2B7Yz2j6hIARUQqQbXY4LagoICEhAQee+yxEu39+vVjzZo1ZTpHYmIia9asYdq0C3+gnTx5ksLCQvz8/Eq079mzh8aNG+Pm5kZMTAzPPvsszZo1u+B58vPzyc/Ptz/PyckpU41VVVN/L7o3b8CavVnM32AsLlmr1Q+DW2bA/NGw5g1o2tPYpd4M1mI4uv+PEPSnQHShoUBnNwhofTYIBbYzHl7+jq1bRKSGMS0kZWZmUlxcTGBgYIn2wMBA0tPTL/q9wcHBHDlyhKKiIqZMmcK99957wWMfe+wxmjRpQp8+fextMTExzJkzh1atWnH48GGmTZtG9+7d2b59Ow0aNCj1PNOnT+fpp5++jHdY9Y3qGsqavVl8uiGFv17XApfaOoH7jDY3Qcx4WPc2LHoAxq8G3+DKveapY0Zv0J8DUcZOKDxZ+vE+TUoGocD20KCF9kgTEakEpv/LajlnToTNZjuv7Vzx8fGcOHGCtWvX8thjj9GiRQtGjRp13nEvvPACcXFxrFixAnf3s3t0DRgwwP51REQEsbGxNG/enA8//JBJkyaVes3JkyeXeC0nJ4eQkJAyvceqql+7QPy8XEnPOc2K3Ufo0zbw0t9U0/WdaqyflLYJPr8Hxn1dMcNTxUVwdO/ZIHRmDlHOwdKPd/GAgDZng1BQewhoC55+pR8vIiIVzrSQ5O/vj7Oz83m9RhkZGef1Lp0rPNxYHTkiIoLDhw8zZcqU80LSiy++yLPPPsvy5cvp0KHDRc/n5eVFREQEe/bsueAxbm5uuLnVrMUG3VycGRbZhHfj9zNvQ7JCEhgLSt42G97pBSlr4adnjDvALsfJo+eEoW1wZNefbrk/h2+oEYb+PH/Ir5m2/BARMZlpIcnV1ZWoqCiWLVvGkCFD7O3Lli3jlltuKfN5bDZbiblCAP/973+ZNm0aS5cuJTo6+pLnyM/PZ+fOnfTs2bPsb6CGGNk1lHfj9/PjrgzSsk/RyLeWT+AGY4uSm9+Az8bC6leMLTla9jn/uOJCY8PX9HPmDuWmlX7eOl4Q2LbkROqANuBRr1LfjoiIlI+pw22TJk1izJgxREdHExsby6xZs0hOTmb8+PGAMcSVmprKnDlzAHjrrbcIDQ2ldevWgLFu0osvvsjDDz9sP+cLL7zAk08+ydy5c2natKm9p6pu3brUrVsXgEcffZRBgwYRGhpKRkYG06ZNIycnh7Fjxzry7VcJzRvWJSbcj3X7j/LphoM80ke7uQPGqtMH7oUN/4NF98OdiyHvyB9BaDsc3gpHdkNxQenfX7/p+ROp64eDUy2f9yUiUo2YGpJGjBhBVlYWU6dOJS0tjfbt27NkyRLCwsIASEtLIzn57B09VquVyZMns3//flxcXGjevDnPPfccDzzwgP2YGTNmUFBQwK233lriWv/+97+ZMmUKAAcPHmTUqFFkZmbSsGFDunXrxtq1a+3XrW1ujwk1QtKvKUy4rgXOTlo7B4B+zxh7u6Vvhbd7lH6Mq/efJlG3g6AIo3dIm8CKiFR7pq6TVJ1V93WS/ux0YTHdpv/A8ZOFfHBXF669KsDskqqOrL3wv+vh1HFjntCZIHQmFPmGqndIRKQaqRbrJEnV4V7HmaGdg3n/5/3ErUtWSPqzBs3hbzsAG7h6mV2NiIg4kP4LLACM6mosZ/DDrgwyci5wF1Zt5eqpgCQiUgspJAkALQO9iQ6rT7HVxmcJF1i7R0REpBZRSBK7UV2NTYXj1idjtWqqmoiI1G4KSWI3sEMjfNxdOHjsFKt/zzS7HBEREVMpJImdex1nhkYae5XFrb/AZqoiIiK1hEKSlDDyjwncy3Yc5khu/iWOFhERqbkUkqSE1kE+dA6tR5HVxueawC0iIrWYQpKc58wE7nkbNIFbRERqL4UkOc9NHRrh7eZCUtZJftmXZXY5IiIiplBIkvN4urpwS+fGgCZwi4hI7aWQJKU6M+S2dHs6WSc0gVtERGofhSQpVbvGvnQM9qWw2MaCjZrALSIitY9CklyQfQL3+hRsNk3gFhGR2kUhSS5oUMfGeLk6sy8zj3X7j5pdjoiIiEMpJMkFebm5cHOnJoAmcIuISO2jkCQXdfsfQ27fbk3nWF6BydWIiIg4jkKSXFREsC/tm/hQUGxlYWKq2eWIiIg4jEKSXNLILkZvUtz6ZE3gFhGRWkMhSS7plk6N8ajjzO8ZJ/g16ZjZ5YiIiDiEQpJckrd7HW7u+McK3Os0gVtERGoHhSQpk1ExxpDbN1vTyD5ZaHI1IiIilU8hScqkY7AvrYO8yS+ysihRK3CLiEjNp5AkZWKxWLg95swEbq3ALSIiNZ9CkpTZLZ2a4F7Hid2Hc9mYfNzsckRERCqVQpKUma9HHW7qYEzgnqcVuEVEpIZTSJLLMqprCABfbTlEzmlN4BYRkZpLIUkuS2RofVoF1uV0oZUvtQK3iIjUYApJclksFguj/tjPba4mcIuISA2mkCSXbUjnJri6OLEzLYctB7PNLkdERKRSKCTJZavn6crAiEaAsZ+biIhITaSQJOVyZsht8eZDnMgvMrkaERGRiqeQJOXSpWl9mjf04mRBMYs3HTK7HBERkQqnkCTl8ucJ3BpyExGRmsj0kDRjxgzCw8Nxd3cnKiqK+Pj4Cx67evVqevToQYMGDfDw8KB169a88sor5x23YMEC2rZti5ubG23btmXRokVXdF0p3dDIYFydndiams1WTeAWEZEaxtSQNH/+fCZOnMjjjz9OYmIiPXv2ZMCAASQnl94z4eXlxYQJE1i1ahU7d+7kiSee4IknnmDWrFn2Y3755RdGjBjBmDFj2Lx5M2PGjGH48OGsW7eu3NeV0vl5udK/fRAAcRv0eyciIjWLxWbiQjcxMTFERkYyc+ZMe1ubNm0YPHgw06dPL9M5hg4dipeXFx999BEAI0aMICcnh2+//dZ+TP/+/alfvz5xcXHlvm5+fj75+fn25zk5OYSEhJCdnY2Pj0/Z33QN88veLEa9u5a6bi6s+9f1eLm5mF2SiIjIBeXk5ODr61umz2/TepIKCgpISEigX79+Jdr79evHmjVrynSOxMRE1qxZwzXXXGNv++WXX8475w033GA/Z3mvO336dHx9fe2PkJCQMtVY03Vr5ke4vxcn8ov4eosmcIuISM1hWkjKzMykuLiYwMDAEu2BgYGkp6df9HuDg4Nxc3MjOjqahx56iHvvvdf+Wnp6+kXPWd7rTp48mezsbPsjJSWlTO+zprNYLIzsYgTGuev1eyIiIjWH6RO3LRZLiec2m+28tnPFx8fz66+/8vbbb/Pqq6/ah9Eu55yXe103Nzd8fHxKPMQwLCqYOs4WNqccZ8ehHLPLERERqRCmhSR/f3+cnZ3P673JyMg4r5fnXOHh4URERHDffffxt7/9jSlTpthfCwoKuug5r+S6Ujr/um70a2dM4J6nCdwiIlJDmBaSXF1diYqKYtmyZSXaly1bRvfu3ct8HpvNVmJCdWxs7Hnn/P777+3nrKjrSkmjuhhrJi3amMqpgmKTqxEREblypt6KNGnSJMaMGUN0dDSxsbHMmjWL5ORkxo8fDxjzgFJTU5kzZw4Ab731FqGhobRu3Row1k168cUXefjhh+3nfOSRR+jVqxfPP/88t9xyC19++SXLly9n9erVZb6uXL7uzRsQ6udJ8tGTfL3lELdFa2K7iIhUb6aGpBEjRpCVlcXUqVNJS0ujffv2LFmyhLCwMADS0tJKrF1ktVqZPHky+/fvx8XFhebNm/Pcc8/xwAMP2I/p3r078+bN44knnuDJJ5+kefPmzJ8/n5iYmDJfVy6fk5OFkV1DeOG73czbkKKQJCIi1Z6p6yRVZ5ezzkJtkZF7mu7Tf6TIamPpxF5cFeRtdkkiIiIlVIt1kqTmCfB2p08bY/K79nMTEZHqTiFJKtSoGGMC98KNBzldqAncIiJSfZUrJKWkpHDw4EH78/Xr1zNx4sQSe6hJ7dSzhT9N6nmQc7qIb7elmV2OiIhIuZUrJN1+++389NNPgLHCdd++fVm/fj3/+te/mDp1aoUWKNWLk5OFUV2NSdtx67QCt4iIVF/lCknbtm2ja9euAHz66ae0b9+eNWvWMHfuXGbPnl2R9Uk1dFt0CM5OFtYfOMrvGblmlyMiIlIu5QpJhYWFuLm5AbB8+XJuvvlmAFq3bk1amoZYartAH3euax0AwP/i92O16gZKERGpfsoVktq1a8fbb79NfHw8y5Yto3///gAcOnSIBg0aVGiBUj2N/mMC97wNKQx/5xd2pmlPNxERqV7KFZKef/553nnnHXr37s2oUaPo2LEjAIsXL7YPw0nt1vuqAJ66qS2ers78mnSMm95YzbSvd3Aiv8js0kRERMqk3ItJFhcXk5OTQ/369e1tBw4cwNPTk4CAgAorsKrSYpJlk5Z9iv98vYMlW40NhQN93HjqpnbcGBGExWIxuToREaltKn0xyVOnTpGfn28PSElJSbz66qvs3r27VgQkKbtGvh7MGB3F7Lu6ENbAk8M5+Tw0dyN3vr+e/Zl5ZpcnIiJyQeUKSbfccot909njx48TExPDSy+9xODBg5k5c2aFFig1Q++rAlg6sRePXN8SVxcn4vdkcsMrq3h52W9adFJERKqkcoWkjRs30rNnTwA+//xzAgMDSUpKYs6cObz++usVWqDUHO51nPlb31YsndiLXq0aUlBs5fUf9nDDq6v4aXeG2eWJiIiUUK6QdPLkSby9jc1Lv//+e4YOHYqTkxPdunUjKSmpQguUmifc34sP7+rCjNGRBPm4k5R1krs+2MBfPk7g0PFTZpcnIiIClDMktWjRgi+++IKUlBSWLl1Kv379AMjIyNAkZikTi8XCjRGNWP73a7ivZzjOTha+3ZZOn5dX8s7KvRQWW80uUUREarlyhaSnnnqKRx99lKZNm9K1a1diY2MBo1epc+fOFVqg1Gx13Vx4fGBbvvnr1USH1edkQTHTv93FwNfjWbcvy+zyRESkFiv3EgDp6emkpaXRsWNHnJyMrLV+/Xp8fHxo3bp1hRZZFWkJgIpntdpYsPEg07/dxdG8AgCGRjbhXze2wb+um8nViYhITXA5n9/lDklnHDx4EIvFQpMmTa7kNNWOQlLlOX6ygBeW7iZufTI2G/i4u/CP/q25vWsozk5aW0lERMqv0tdJslqtTJ06FV9fX8LCwggNDaVevXr85z//wWrVXBK5MvU8XXl2SAQL/9Kddo19yDldxJNfbGPIjJ/ZcvC42eWJiEgtUa6Q9Pjjj/Pmm2/y3HPPkZiYyMaNG3n22Wd54403ePLJJyu6RqmlOofWZ/GEq5l6Szu83VzYcjCbW976mSe/2Eb2qUKzyxMRkRquXMNtjRs35u233+bmm28u0f7ll1/y4IMPkpqaWmEFVlUabnOsjNzTTF+yi0WJxp8t/7qu/OvGNgzp3ETbm4iISJlV+nDb0aNHS52c3bp1a44ePVqeU4pcVIC3O6+M6MTc+2Jo3tCLzBMFTPp0MyNnrWXP4VyzyxMRkRqoXCGpY8eOvPnmm+e1v/nmm3To0OGKixK5kO7N/fn2kV78X/+rcK/jxLr9RxnwWjzTv91JXn6R2eWJiEgNUq7htpUrVzJw4EBCQ0OJjY3FYrGwZs0aUlJSWLJkiX3LkppMw23mO3jsJE9/tYNlOw4D0NjXnacGteOGdoEaghMRkVJV+nDbNddcw2+//caQIUM4fvw4R48eZejQoWzfvp0PPvigXEWLXK7g+p68e2c0/7szmuD6HhzKPs34jxO4e/YGkrNOml2eiIhUc1e8TtKfbd68mcjISIqLa/6u7upJqlpOFRTz1k+/886qvRQW23BzceKha1vwwDXNcHNxNrs8ERGpIiq9J0mkqvFwdebRG67iu4m96NGiAflFVl5e9hv9X40nfs8Rs8sTEZFqSCFJapTmDevy8T0xvD6qMw293difmceY99bz0NyNpGefNrs8ERGpRhSSpMaxWCzc3LExP/z9Gu7q0RQnC3yzJY3rX1rB/+L3UVSsVeFFROTSLmtO0tChQy/6+vHjx1m5cqXmJEmVsv1QNk98sY3E5OMAtA7yZtrg9kQ39TO3MBERcbhK2+D2rrvuKtNxteEON4Wk6sVqtfFZQgrTv93F8ZPGlibDo4N5bEAb/LxcTa5OREQcpdJCkpylkFQ9Hc0r4PlvdzH/1xQA6nnW4Z/9WzMiOgQnJ62tJCJS0ykkOYBCUvWWkHSUxxdtY1e6saVJp5B6TBvcnvZNfE2uTEREKpNCkgMoJFV/RcVWPvwliZe/301eQTFOFrgztimT+rXCx72O2eWJiEgl0DpJImXg4uzEPVeH88Pfe3NTh0ZYbTB7zQGuf2klX25KRf9/EBGp3UwPSTNmzCA8PBx3d3eioqKIj4+/4LELFy6kb9++NGzYEB8fH2JjY1m6dGmJY3r37o3FYjnvMXDgQPsxU6ZMOe/1oKCgSnuPUrUF+brz5u2RfHRPV8L9vTiSm88j8zYx+n/r+D3jhNnliYiISUwNSfPnz2fixIk8/vjjJCYm0rNnTwYMGEBycnKpx69atYq+ffuyZMkSEhISuPbaaxk0aBCJiYn2YxYuXEhaWpr9sW3bNpydnbnttttKnKtdu3Yljtu6dWulvlep+nq2bMh3E3vy976tcHNxYs3eLAa8tor/Lt3FqYKav6yFiIiUZOqcpJiYGCIjI5k5c6a9rU2bNgwePJjp06eX6Rzt2rVjxIgRPPXUU6W+/uqrr/LUU0+RlpaGl5cXYPQkffHFF2zatKnMtebn55Ofn29/npOTQ0hIiOYk1VDJWSf59+Jt/LTb2NLEv64bY7qFcUe3UBrUdTO5OhERKa9qMSepoKCAhIQE+vXrV6K9X79+rFmzpkznsFqt5Obm4ud34UUB33vvPUaOHGkPSGfs2bOHxo0bEx4ezsiRI9m3b99FrzV9+nR8fX3tj5CQkDLVKNVTaANP3h/XhXfGRNGkngeZJ/J5ZflvxD73I//8fAu7/7grTkREai7TQlJmZibFxcUEBgaWaA8MDCQ9Pb1M53jppZfIy8tj+PDhpb6+fv16tm3bxr333luiPSYmhjlz5rB06VLeffdd0tPT6d69O1lZWRe81uTJk8nOzrY/UlJSylSjVF8Wi4Ub2gWx4h+9eW1kJzoG+1JQZGX+rync8Ooqxry3jp92ZWC1aoK3iEhN5GJ2ARZLyQX8bDbbeW2liYuLY8qUKXz55ZcEBASUesx7771H+/bt6dq1a4n2AQMG2L+OiIggNjaW5s2b8+GHHzJp0qRSz+Xm5oabm4ZZaqM6zk7c0qkJN3dsTELSMd5bvZ+l29OJ35NJ/J5Mmjf04q4e4QyLDMbD1dnsckVEpIKYFpL8/f1xdnY+r9coIyPjvN6lc82fP5977rmHzz77jD59+pR6zMmTJ5k3bx5Tp069ZC1eXl5ERESwZ8+esr8BqXUsFgvRTf2IbupHytGTfLjmAPM3pLD3SB5PfLGNF7/fze1dQ7kztilBvu5mlysiIlfItOE2V1dXoqKiWLZsWYn2ZcuW0b179wt+X1xcHOPGjWPu3Lklbus/16effkp+fj533HHHJWvJz89n586dNGrUqOxvQGq1ED9PnripLWsmX8dTN7UlxM+D4ycLmbFiL1c//yMT5yWy5eBxs8sUEZErYOrdbfPnz2fMmDG8/fbbxMbGMmvWLN599122b99OWFgYkydPJjU1lTlz5gBGQLrzzjt57bXXGDp0qP08Hh4e+PqW3E6iZ8+eNGnShHnz5p133UcffZRBgwYRGhpKRkYG06ZNY+XKlWzdupWwsLAy1a4Vt+XPiq02lu04zPs/72f9/qP29i5N63PP1eH0bRuEs/aGExEx3eV8fps6J2nEiBFkZWUxdepU0tLSaN++PUuWLLEHlbS0tBJrJr3zzjsUFRXx0EMP8dBDD9nbx44dy+zZs+3Pf/vtN1avXs33339f6nUPHjzIqFGjyMzMpGHDhnTr1o21a9eWOSCJnMvZyUL/9kH0bx/E1oPZvP/zfr7afIgNB46x4cAxQvw8GNc9nOHRwXhryxMRkWpBe7eVk3qS5FIO55xmzi8H+GRdMsdPFgJQ182F4dEh3NWjKSF+niZXKCJS+2iDWwdQSJKyOlVQzMLEg7y/ej97j+QB4GSBfm2DuKdnONFh9ct0R6eIiFw5hSQHUEiSy2W12li15wjvrd5P/J5Me3uHYF/u7hHOjRGNcHUxfTtFEZEaTSHJARSS5Er8djiX91fvZ2FiKgVFVgACfdy4M7Ypo2NCqefpanKFIiI1k0KSAygkSUXIOpHP3HXJzFmbxJFcY29A9zpODIsM5q4e4bQIqGtyhSIiNYtCkgMoJElFyi8q5uvNaby3ej870nLs7b2vasg9V4dzdQt/zVsSEakACkkOoJAklcFms7F231HeW72fH3Yd5szfzqsCvbn76qbc0qkJ7nW09YmISHkpJDmAQpJUtgOZecxec4BPf03hZEExAA28XBkdE8odsWEEeGvrExGRy6WQ5AAKSeIo2acKmb8hmQ/XJJF6/BQArs5ODOrYmHuuDqdtY/35ExEpK4UkB1BIEkcrKrby3fZ03lu9n8Tk4/b22GYNuPvqcK5vHYCTtj4REbkohSQHUEgSM21MPsb7q/fz7bZ0iq3GX+GmDTy5q0c4t0YF4+Vm6o5DIiJVlkKSAygkSVWQevwUc9YcYO76ZHJPFwHg4+7CqK6hjO3elMb1PEyuUESkalFIcgCFJKlK8vKLWLDR2PrkQNZJwNh0d0D7IO6+OpzI0PomVygiUjUoJDmAQpJURVarjR93ZfDe6v38si/L3t45tB73XB1O/3ZBuDhr6xMRqb0UkhxAIUmquu2Hsnl/9QG+2nyIgmJj65NGvu4Mjw5hRJcQDcWJSK2kkOQACklSXWTknubjtcl8sjaJrLwCAJwscO1VAYzqGkrvqxqqd0lEag2FJAdQSJLq5nRhMUu3pxO3Ppm1+47a24N83BnexehdaqLeJRGp4RSSHEAhSaqzvUdOMG99Mp8nHOTYyULA6F3q/Ufv0rXqXRKRGkohyQEUkqQmyC8qZun2w8StSy4x0TvQx40R0SEM7xJCcH1PEysUEalYCkkOoJAkNc2+IyeYvyGFzxIOcvSPuUsWC1zTqiG3dw3lutYB6l0SkWpPIckBFJKkpsovKub77YeJW5/Mmr0le5eGR4cwPDqEED/1LolI9aSQ5AAKSVIb7M/MY96GZD7/9aD9zjiLBXq1bMjtMUbvUh31LolINaKQ5AAKSVKbFBRZWbbjMHPXJ/Hz72d7lwK83ezrLql3SUSqA4UkB1BIktrqQGYe8zak8HlCCpknzvYu9WzZkNu7hnB9m0D1LolIlaWQ5AAKSVLbFRRZWb7zMHPXJbP690x7e0NvN4ZHBzOyS6h6l0SkylFIcgCFJJGzkrKM3qXPfi3Zu3R1C39u7xpKn7bqXRKRqkEhyQEUkkTOV1Bk5Yedh5m7Ppn4PWd7l/zrnu1dCm2g3iURMY9CkgMoJIlcXHLWSeZtSObTXw+SeSLf3t6zpT+juobSp00gri7qXRIRx1JIcgCFJJGyKSw+07uUQvyeI5z5F8e/riu3RoUwqmsIYQ28zC1SRGoNhSQHUEgSuXwpR8/2Lh3JPdu7dHULo3epb1v1LolI5VJIcgCFJJHyM3qXMohbn8yqP/UuNfBy5dboYEZ1CaWpv3qXRKTiKSQ5gEKSSMVIOXqST39NYf6GFDL+1LvUo0UDRnUNpV/bIPUuiUiFUUhyAIUkkYpVWGzlx11G79LK387pXYoKZmTXUMLVuyQiV0ghyQEUkkQqz8FjJ/l0Qwrzf03hcM7Z3qXYZg24PSaUfu0CcXNxNrFCEamuLufz2/Q+7BkzZhAeHo67uztRUVHEx8df8NiFCxfSt29fGjZsiI+PD7GxsSxdurTEMbNnz8ZisZz3OH36dLmvKyKOFVzfk0n9ruLnf17Hu3dGc+1VDbFY4Jd9WTwcl0j36T/y/He7SDl60uxSRaQGMzUkzZ8/n4kTJ/L444+TmJhIz549GTBgAMnJyaUev2rVKvr27cuSJUtISEjg2muvZdCgQSQmJpY4zsfHh7S0tBIPd3f3cl9XRMzh4uxE37aBfHBXV1b/8zr+en1LgnzcycorYOaKvfT670/cM3sDP+3OwGpVp7iIVCxTh9tiYmKIjIxk5syZ9rY2bdowePBgpk+fXqZztGvXjhEjRvDUU08BRk/SxIkTOX78eKVeV8NtIuYoKrbyw64MPl6bVGJV71A/T0bHhHJbdAh+Xq4mVigiVVm1GG4rKCggISGBfv36lWjv168fa9asKdM5rFYrubm5+Pn5lWg/ceIEYWFhBAcHc9NNN5XoaSrvdfPz88nJySnxEBHHc3F24oZ2QXx0Tww//v0a7rk6HB93F5KPnmT6t7voNv0HJn26icTkY2jKpYhcCdNCUmZmJsXFxQQGBpZoDwwMJD09vUzneOmll8jLy2P48OH2ttatWzN79mwWL15MXFwc7u7u9OjRgz179lzRdadPn46vr6/9ERISUta3KiKVpFnDujx5U1vW/asPzw+LoF1jHwqKrCzcmMqQGWsY9OZq5m9I5lRBsdmlikg1ZPrEbYvFUuK5zWY7r600cXFxTJkyhfnz5xMQEGBv79atG3fccQcdO3akZ8+efPrpp7Rq1Yo33njjiq47efJksrOz7Y+UlJSyvD0RcQAPV2dGdAnl64evZtGD3Rka2QRXFye2pebwzwVbiXl2OVO/2sG+IyfMLlVEqhEXsy7s7++Ps7Pzeb03GRkZ5/XynGv+/Pncc889fPbZZ/Tp0+eixzo5OdGlSxd7T1J5r+vm5oabm9tFryUi5rJYLHQOrU/n0Po8MbAtn/2awsfrkkg5eor3f97P+z/v5+oW/oyJDeP61gG4OJv+/0QRqcJM+xfC1dWVqKgoli1bVqJ92bJldO/e/YLfFxcXx7hx45g7dy4DBw685HVsNhubNm2iUaNGV3RdEale/LxceeCa5qx89Fo+uKsL17cOwGKB1b9n8sBHCfR84Sde/2EPGTmnL30yEamVTOtJApg0aRJjxowhOjqa2NhYZs2aRXJyMuPHjweMIa7U1FTmzJkDGAHpzjvv5LXXXqNbt2723iAPDw98fX0BePrpp+nWrRstW7YkJyeH119/nU2bNvHWW2+V+boiUnM4OVm49qoArr0qgJSjJ5m7Ppn5G1JIyz7Ny8t+4/Uf9nBD+yDGdAsjJtyvTMP9IlI7mBqSRowYQVZWFlOnTiUtLY327duzZMkSwsLCAEhLSyuxdtE777xDUVERDz30EA899JC9fezYscyePRuA48ePc//995Oeno6vry+dO3dm1apVdO3atczXFZGaKcTPk3/2b83EPi35dms6H61NIiHpGN9sSeObLWm0DKjLmNgwhnRugrd7HbPLFRGTaVuSctI6SSI1w/ZD2Xy8NpkvElM5VWjcBefl6syQyCbc0S2M1kH6+y1Sk2jvNgdQSBKpWXJOF7Iw4SAfrU1i75E8e3vXpn7cERtG/3ZBuLpoordIdaeQ5AAKSSI1k81m45d9WXy8Noml2w9T/Md2J/51XRnZJZRRMaE0qedhcpUiUl4KSQ6gkCRS86Vnn2behmTi1idzOCcfACcLXN8mkDHdwri6hT9OTproLVKdKCQ5gEKSSO1RWGxl+Y7DfLQ2iTV7s+ztTRt4cke3MG6NCqaep/aLE6kOFJIcQCFJpHb6PSOXj9cmsyDhILn5RQC4uThxc8fGjIkNo0NwPXMLFJGLUkhyAIUkkdrtZEERX246xJxfktiZdnbD647BvtzRLYxBHRvjXsfZxApFpDQKSQ6gkCQiYEz03ph8nI/XJvHNljQKiq0A+HrUYXh0MKNjwmjq72VylSJyhkKSAygkici5sk7k8+mvB/l4bRKpx0/Z23u1asiYbmFc1zoAZ030FjGVQpIDKCSJyIUUW22s/C2Dj35JYsVvRzjzr2yTeh7cHhPK8OgQGnprw2wRMygkOYBCkoiURXLWST5Zn8SnG1I4drIQgDrOFga0b8SY2DCiw+prvzgRB1JIcgCFJBG5HKcLi1myNY2P1iaRmHzc3n5VoDd3/LFfXF03U7fTFKkVFJIcQCFJRMprW2o2H/2SxJebUzldaEz01n5xIo6hkOQACkkicqWyTxayYONBPl6XxL4/7RfXpWl97ugWRv/2Qbi5aBkBkYqkkOQACkkiUlFsNhu/7M3io7VJfL/j7H5xDbxcGd4lhNu7hhLi52lylSI1g0KSAygkiUhlKG2/OIsFrr0qgDHdwujVqqGWERC5AgpJDqCQJCKVqbDYyg87D/Px2mRW/55pbw+u78HomDCGRwfToK6WERC5XApJDqCQJCKOsu/ICT5Zl8xnv6aQc9rYL87V2YkbI4IYExtGZKiWERApK4UkB1BIEhFHO1VQzFdbDvHx2iS2HMy2t7cO8mZMbBiDOzXBS8sIiFyUQpIDKCSJiJk2pxj7xS3efIj8ImMZgbpuLgz9YxmBVoHeJlcoUjUpJDmAQpKIVAXHTxbwecJBPlmXzP7Ms8sIdA33M5YRaBeEq4uTiRWKVC0KSQ6gkCQiVYnVamPN3iw+WnuA5Tsz7MsI+Nd1ZUSXEG6PCaNJPQ+TqxQxn0KSAygkiUhVlZZ9irj1Kcxbn0xGrrGMgJMFrmsdwB3dwujVsiFOWkZAaimFJAdQSBKRqq6w2MqyHYf5eG0Sa/Zm2dtD/TwZHRPKbdEh+Hm5mlihiOMpJDmAQpKIVCe/Z5zgk3VJfJ5wkNwzywi4OHFTRCPuiA2jc0g9LSMgtYJCkgMoJIlIdXSyoIivNh9izi9JbD+UY29v28iHMbFh3NKpMZ6uWkZAai6FJAdQSBKR6sxms7Ep5Tgfr03mqy2HKPhjGQFvNxeGRQVzR7dQWgRoGQGpeRSSHEAhSURqimN5xjICH69LIinrpL29WzNjGYF+bbWMgNQcCkkOoJAkIjWN1Wpj9e+ZfLQ2iR92HuaPVQRo6O3GyC4hjOoaSmMtIyDVnEKSAygkiUhNduj4KeLWJxO3PoXME2eXEbi+TSBjuoVxdQt/LSMg1ZJCkgMoJIlIbVBQZOX7Hel8vDaJtfuO2tubNvBkdEwYt0YFU1/LCEg1opDkAApJIlLb7DmcyyfrklmQcJDc/LPLCAzq0Jj7ezXjqiBN9JaqTyHJARSSRKS2yssv4stNh/h4bRI70s4uI3B96wAevLY5UWF+JlYncnEKSQ6gkCQitZ3NZmNj8nHeW72Pb7elc+bTpGtTP/7Suzm9r2qoBSqlylFIcgCFJBGRs/YdOcE7K/exMPEghcXGx0rrIG/+0rs5AyMa4eKsJQSkaricz2/T/9TOmDGD8PBw3N3diYqKIj4+/oLHLly4kL59+9KwYUN8fHyIjY1l6dKlJY5599136dmzJ/Xr16d+/fr06dOH9evXlzhmypQpWCyWEo+goKBKeX8iIrVBs4Z1ef7WDsT/33Xc1zMcT1dndqXn8si8TVz70go+WpvE6cJis8sUuSymhqT58+czceJEHn/8cRITE+nZsycDBgwgOTm51ONXrVpF3759WbJkCQkJCVx77bUMGjSIxMRE+zErVqxg1KhR/PTTT/zyyy+EhobSr18/UlNTS5yrXbt2pKWl2R9bt26t1PcqIlIbBPm68/jAtqx57Dr+3rcVfl6upBw9xZNfbOPq539ixorfyTldaHaZImVi6nBbTEwMkZGRzJw5097Wpk0bBg8ezPTp08t0jnbt2jFixAieeuqpUl8vLi6mfv36vPnmm9x5552A0ZP0xRdfsGnTpnLXruE2EZFLO1VQzPwNybwbv5/U46cAY+uT0d3CuPvqpgR4u5tcodQ21WK4raCggISEBPr161eivV+/fqxZs6ZM57BareTm5uLnd+E7KU6ePElhYeF5x+zZs4fGjRsTHh7OyJEj2bdv30WvlZ+fT05OTomHiIhcnIerM+N6hLPiH715eXhHWgbUJTe/iLdX7uXq53/iX4u2kpSVZ3aZIqUyLSRlZmZSXFxMYGBgifbAwEDS09PLdI6XXnqJvLw8hg8ffsFjHnvsMZo0aUKfPn3sbTExMcyZM4elS5fy7rvvkp6eTvfu3cnKyrrgeaZPn46vr6/9ERISUqYaRUQE6jg7MTQymKUTe/HundF0Dq1HQZGVueuSufbFFTwcl8j2Q9lmlylSgukTt8+9PdRms5XpltG4uDimTJnC/PnzCQgIKPWYF154gbi4OBYuXIi7+9ku3QEDBjBs2DAiIiLo06cP33zzDQAffvjhBa83efJksrOz7Y+UlJSyvD0REfkTJycLfdsGsvAv3Zl/fzd6X9UQqw2+2nyIga+vZuz761m7LwvdeC1VgYtZF/b398fZ2fm8XqOMjIzzepfONX/+fO655x4+++yzEj1Ef/biiy/y7LPPsnz5cjp06HDR83l5eREREcGePXsueIybmxtubm4XPY+IiJSNxWIhplkDYpo1YPuhbN5euY9vthxi5W9HWPnbESJD6/GX3i24vnWA9ogT05jWk+Tq6kpUVBTLli0r0b5s2TK6d+9+we+Li4tj3LhxzJ07l4EDB5Z6zH//+1/+85//8N133xEdHX3JWvLz89m5cyeNGjW6vDchIiJXrF1jX94Y1ZmfHu3N6JhQXF2c2Jh8nPvm/MoNr65iQcJBCoutZpcptZCpd7fNnz+fMWPG8PbbbxMbG8usWbN499132b59O2FhYUyePJnU1FTmzJkDGAHpzjvv5LXXXmPo0KH283h4eODr6wsYQ2xPPvkkc+fOpUePHvZj6tatS926dQF49NFHGTRoEKGhoWRkZDBt2jRWrlzJ1q1bCQsLK1PturtNRKRyZOSe5oOfD/DxL0n2PeKa1PPgvp7hjOgSioers8kVSnVWrVbcnjFjBi+88AJpaWm0b9+eV155hV69egEwbtw4Dhw4wIoVKwDo3bs3K1euPO8cY8eOZfbs2QA0bdqUpKSk847597//zZQpUwAYOXIkq1atIjMzk4YNG9KtWzf+85//0LZt2zLXrZAkIlK5ck4X8vHaJN5fvZ/MEwUA+Hm5Mq57U8bGNsXXs47JFUp1VK1CUnWlkCQi4hinC4v5POEg76zaS8pRY60lL1dnbo8J5Z6rmxHkq7WWpOwUkhxAIUlExLGKiq18szWNmSv2sis9F4A6zhaGdg7mgWua0axhXZMrlOpAIckBFJJERMxhs9lY8dsRZv60l/UHjgJgscCA9kH85ZoWRAT7mlyhVGUKSQ6gkCQiYr5fDxzl7ZV7Wb4zw952dQt/HuzdnNjmDcq07p7ULgpJDqCQJCJSdexOz+XtlXtZvPkQxVbjY61jsC9/6d2cfm2DtNaS2CkkOYBCkohI1ZNy9CT/i9/HvA0p5BcZays1a+jF+GuaM7hTE1xdTN9oQkymkOQACkkiIlVX5ol8PlxzgA/XHCDntLHWUpCPO/f2DGdU11C83EzbcEJMppDkAApJIiJV34n8IuauS+J/8fvJyM0HwNejDmO7N2Vc96b4ebmaXKE4mkKSAygkiYhUH/lFxSzamMo7q/axPzMPAI86zozsGsK9PZvRpJ6HyRWKoygkOYBCkohI9VNstfHdtnRmrvydbak5ALg4WbilUxPGX9OMloHeJlcolU0hyQEUkkREqi+bzcbq3zOZuWIva/Zm2du7hvtxa2QwN3ZoRF3NW6qRFJIcQCFJRKRm2JRynJkrfuf7HYc584noXseJ/u2CGBYVTPfm/jhrCYEaQyHJARSSRERqlkPHT7EoMZUFGw+y70ievb2RrztDOjdhWFQwzbX1SbWnkOQACkkiIjWTzWZjU8pxFmw8yFeb08g+VWh/rVNIPYZFBXNzh8b4etYxsUopL4UkB1BIEhGp+fKLivlhZwYLEg6y4rcj9tW8XZ2d6NM2gGGRwVzTqiEuzlqksrpQSHIAhSQRkdrlSG4+X25K5fOEg+xKz7W3+9d1ZXAnYziuTSN9HlR1CkkOoJAkIlJ7bT+UzYKEVL7clEpWXoG9vW0jH4ZFBXNLp8b413UzsUK5EIUkB1BIEhGRwmIrK3YfYUHCQX7YdZjCYuMj1cXJQu+rGjIsMpjr2gTg5uJscqVyhkKSAygkiYjInx3LK+CrLYdYkHCQzQez7e31POtwc8fGDIsMpkOwLxaLlhMwk0KSAygkiYjIhew5nMuCjaksSjzI4Zx8e3uLgLoMiwxmSOcmBPm6m1hh7aWQ5AAKSSIicinFVmNl7wUJB1m6PZ38IisAThbo0cKfW6OC6dc2CA9XDcc5ikKSAygkiYjI5cg5XciSLWks2HiQDQeO2dvrurkwMKIRw6KC6dK0vobjKplCkgMoJImISHklZeWxYGMqCzce5OCxU/b2UD9PhkY2YVhkMCF+niZWWHMpJDmAQpKIiFwpq9XG+gNH+TzhIN9uTSOvoNj+Wky4H8OigrkxQpvtViSFJAdQSBIRkYp0sqCI77als2DjQdbszbJvtutRx5n+7YMYFhlMbPMG2mz3CikkOYBCkoiIVJbU46f4IjGVBQkH2ZepzXYrkkKSAygkiYhIZbPZbCSmHOfzhIN8vfkQOaeL7K91Dq3HsMhgBmmz3cuikOQACkkiIuJIpwuLWb7zMAsSDrJqT+bZzXZdnOjbJpBhUU3o1VKb7V6KQpIDKCSJiIhZMnJP82XiIRZsPHezXTcGd2qszXYvQiHJARSSRETEbDabje2Hcliw8SBfbjrE0T9tttumkQ9DOzfhlk6NCfDR6t5nKCQ5gEKSiIhUJQVFVlbszmDBxoP8uCvDvtmukwV6tmzI0MgmWt0bhSSHUEgSEZGq6lheAV9vTWPRxoNsTD5ub6/r5sKA9kEMjQwmJtwPp1q4nIBCkgMoJImISHWwPzOPRRsPsjAxtcTq3k3qeTCkcxOGRDapVcsJKCQ5gEKSiIhUJ1arjQ0HjrIoMZVvtqSRm392OYGOIfUYFtmEQR0aU9/L1cQqK59CkgMoJImISHV1urCYZTsOs3BjyeUE6jhbuPaqAIZGNuHa1gG4udS8+UuX8/lt+mIKM2bMIDw8HHd3d6KiooiPj7/gsQsXLqRv3740bNgQHx8fYmNjWbp06XnHLViwgLZt2+Lm5kbbtm1ZtGjRFV1XRESkJnGv48ygjo354K6urJ18PU/e1JZ2jX0oLLbx/Y7DjP94I12f+YEnvtjKxuRj1Nb+FFND0vz585k4cSKPP/44iYmJ9OzZkwEDBpCcnFzq8atWraJv374sWbKEhIQErr32WgYNGkRiYqL9mF9++YURI0YwZswYNm/ezJgxYxg+fDjr1q0r93VFRERqqobebtxzdTjf/LUnSyf24oFrmhHo40b2qUI+XpvM0BlruO6llbz+wx5Sjp40u1yHMnW4LSYmhsjISGbOnGlva9OmDYMHD2b69OllOke7du0YMWIETz31FAAjRowgJyeHb7/91n5M//79qV+/PnFxceW+bn5+Pvn5+fbnOTk5hISEaLhNRERqnGKrjTV7M1m4MZXvtqVzqrDY/lrXcD+Gdm7CjR0a4eNe/bZDqRbDbQUFBSQkJNCvX78S7f369WPNmjVlOofVaiU3Nxc/Pz972y+//HLeOW+44Qb7Oct73enTp+Pr62t/hISElKlGERGR6sbZyULPlg15ZUQnNjzRh5du60iPFg2wWGD9/qM8tnArXaYtZ8Lcjfy0K4OiYqvZJVcKF7MunJmZSXFxMYGBgSXaAwMDSU9PL9M5XnrpJfLy8hg+fLi9LT09/aLnLO91J0+ezKRJk+zPz/QkiYiI1GR13VwYFhXMsKhgDh0/xRebUlm4MZXfM07w9ZY0vt6Shn9dV27p1IQhnZvQrrEPFkvNWH/JtJB0xrm/kTabrUy/uXFxcUyZMoUvv/ySgICAyz7n5V7Xzc0NNze3S9YlIiJSUzWu58GDvVvwl2uasy3V2A5l8eZDZJ4o4L3V+3lv9X6uCvRmaGQTBnduQmA13w7FtJDk7++Ps7Pzeb03GRkZ5/XynGv+/Pncc889fPbZZ/Tp06fEa0FBQRc955VcV0RERIyOhohgXyKCfXl8YBtW7j7CosRUlu04zO7DuUz/dhfPf7eLHi38GRrZhBvaBeHpanq/zGUzbU6Sq6srUVFRLFu2rET7smXL6N69+wW/Ly4ujnHjxjF37lwGDhx43uuxsbHnnfP777+3n7O81xUREZHz1XF2ok/bQN4aHcmGx/vw7JAIosPqY7VB/J5M/jZ/M9HTlvP3Tzfz8+9n12SqDkyNdZMmTWLMmDFER0cTGxvLrFmzSE5OZvz48YAxDyg1NZU5c+YARkC68847ee211+jWrZu9N8jDwwNfX18AHnnkEXr16sXzzz/PLbfcwpdffsny5ctZvXp1ma8rIiIil8/Xsw63x4Rye0woSVl5LEo05i8lHz3Jgo0HWbDxII183RncuQlDOzehZaC32SVflOkrbs+YMYMXXniBtLQ02rdvzyuvvEKvXr0AGDduHAcOHGDFihUA9O7dm5UrV553jrFjxzJ79mz7888//5wnnniCffv20bx5c5555hmGDh1a5uuWhVbcFhERuTSbzUZC0jEWJqby9eZD5Jw+ux1KRBNfhkY24eaOjWlQ1zHzfrUtiQMoJImIiFye04XF/Lgrg4UbD7Ji9xGK/hh6c3Gy0PuqhgyNDOa61gG416m87VAUkhxAIUlERKT8sk7k89XmQyxMTGXLwWx7u4+7CwM7NGZYZBOiwupX+HICCkkOoJAkIiJSMfYczmVhYipfJKaSln3a3t6jRQM+ubdbhV7rcj6/q9/9eCIiIlKjtAz05p/9W/Nov6tYty+LBRtT+XZbGlFhfpf+5kqknqRyUk+SiIhI5TlZUERhsQ1fj4rdH049SSIiIlKtVYXFJ01bTFJERESkKlNIEhERESmFQpKIiIhIKRSSREREREqhkCQiIiJSCoUkERERkVIoJImIiIiUQiFJREREpBQKSSIiIiKlUEgSERERKYVCkoiIiEgpFJJERERESqGQJCIiIlIK87fYraZsNhsAOTk5JlciIiIiZXXmc/vM5/jFKCSVU25uLgAhISEmVyIiIiKXKzc3F19f34seY7GVJUrJeaxWK4cOHcLb2xuLxVKh587JySEkJISUlBR8fHwq9Nxy+fTzqFr086ha9POoevQzuTibzUZubi6NGzfGyenis47Uk1ROTk5OBAcHV+o1fHx89Ae8CtHPo2rRz6Nq0c+j6tHP5MIu1YN0hiZui4iIiJRCIUlERESkFApJVZCbmxv//ve/cXNzM7sUQT+PqkY/j6pFP4+qRz+TiqOJ2yIiIiKlUE+SiIiISCkUkkRERERKoZAkIiIiUgqFJBEREZFSKCRVMTNmzCA8PBx3d3eioqKIj483u6Raafr06XTp0gVvb28CAgIYPHgwu3fvNrss+cP06dOxWCxMnDjR7FJqtdTUVO644w4aNGiAp6cnnTp1IiEhweyyaqWioiKeeOIJwsPD8fDwoFmzZkydOhWr1Wp2adWaQlIVMn/+fCZOnMjjjz9OYmIiPXv2ZMCAASQnJ5tdWq2zcuVKHnroIdauXcuyZcsoKiqiX79+5OXlmV1arbdhwwZmzZpFhw4dzC6lVjt27Bg9evSgTp06fPvtt+zYsYOXXnqJevXqmV1arfT888/z9ttv8+abb7Jz505eeOEF/vvf//LGG2+YXVq1piUAqpCYmBgiIyOZOXOmva1NmzYMHjyY6dOnm1iZHDlyhICAAFauXEmvXr3MLqfWOnHiBJGRkcyYMYNp06bRqVMnXn31VbPLqpUee+wxfv75Z/V2VxE33XQTgYGBvPfee/a2YcOG4enpyUcffWRiZdWbepKqiIKCAhISEujXr1+J9n79+rFmzRqTqpIzsrOzAfDz8zO5ktrtoYceYuDAgfTp08fsUmq9xYsXEx0dzW233UZAQACdO3fm3XffNbusWuvqq6/mhx9+4LfffgNg8+bNrF69mhtvvNHkyqo3bXBbRWRmZlJcXExgYGCJ9sDAQNLT002qSsDYMXrSpElcffXVtG/f3uxyaq158+axceNGNmzYYHYpAuzbt4+ZM2cyadIk/vWvf7F+/Xr++te/4ubmxp133ml2ebXOP//5T7Kzs2ndujXOzs4UFxfzzDPPMGrUKLNLq9YUkqoYi8VS4rnNZjuvTRxrwoQJbNmyhdWrV5tdSq2VkpLCI488wvfff4+7u7vZ5QhgtVqJjo7m2WefBaBz585s376dmTNnKiSZYP78+Xz88cfMnTuXdu3asWnTJiZOnEjjxo0ZO3as2eVVWwpJVYS/vz/Ozs7n9RplZGSc17skjvPwww+zePFiVq1aRXBwsNnl1FoJCQlkZGQQFRVlbysuLmbVqlW8+eab5Ofn4+zsbGKFtU+jRo1o27ZtibY2bdqwYMECkyqq3f7xj3/w2GOPMXLkSAAiIiJISkpi+vTpCklXQHOSqghXV1eioqJYtmxZifZly5bRvXt3k6qqvWw2GxMmTGDhwoX8+OOPhIeHm11SrXb99dezdetWNm3aZH9ER0czevRoNm3apIBkgh49epy3LMZvv/1GWFiYSRXVbidPnsTJqeRHurOzs5YAuELqSapCJk2axJgxY4iOjiY2NpZZs2aRnJzM+PHjzS6t1nnooYeYO3cuX375Jd7e3vYePl9fXzw8PEyurvbx9vY+bz6Yl5cXDRo00Dwxk/ztb3+je/fuPPvsswwfPpz169cza9YsZs2aZXZptdKgQYN45plnCA0NpV27diQmJvLyyy9z9913m11ataYlAKqYGTNm8MILL5CWlkb79u155ZVXdMu5CS40D+yDDz5g3Lhxji1GStW7d28tAWCyr7/+msmTJ7Nnzx7Cw8OZNGkS9913n9ll1Uq5ubk8+eSTLFq0iIyMDBo3bsyoUaN46qmncHV1Nbu8akshSURERKQUmpMkIiIiUgqFJBEREZFSKCSJiIiIlEIhSURERKQUCkkiIiIipVBIEhERESmFQpKIiIhIKRSSREREREqhkCQiUkEsFgtffPGF2WWISAVRSBKRGmHcuHFYLJbzHv379ze7NBGpprTBrYjUGP379+eDDz4o0ebm5mZSNSJS3aknSURqDDc3N4KCgko86tevDxhDYTNnzmTAgAF4eHgQHh7OZ599VuL7t27dynXXXYeHhwcNGjTg/vvv58SJEyWOef/992nXrh1ubm40atSICRMmlHg9MzOTIUOG4OnpScuWLVm8eHHlvmkRqTQKSSJSazz55JMMGzaMzZs3c8cddzBq1Ch27twJwMmTJ+nfvz/169dnw4YNfPbZZyxfvrxECJo5cyYPPfQQ999/P1u3bmXx4sW0aNGixDWefvpphg8fzpYtW7jxxhsZPXo0R48edej7FJEKYhMRqQHGjh1rc3Z2tnl5eZV4TJ061Waz2WyAbfz48SW+JyYmxvaXv/zFZrPZbLNmzbLVr1/fduLECfvr33zzjc3JycmWnp5us9lstsaNG9sef/zxC9YA2J544gn78xMnTtgsFovt22+/rbD3KSKOozlJIlJjXHvttcycObNEm5+fn/3r2NjYEq/FxsayadMmAHbu3EnHjh3x8vKyv96jRw+sViu7d+/GYrFw6NAhrr/++ovW0KFDB/vXXl5eeHt7k5GRUd63JCImUkgSkRrDy8vrvOGvS7FYLADYbDb716Ud4+HhUabz1alT57zvtVqtl1WTiFQNmpMkIrXG2rVrz3veunVrANq2bcumTZvIy8uzv/7zzz/j5OREq1at8Pb2pmnTpvzwww8OrVlEzKOeJBGpMfLz80lPTy/R5uLigr+/PwCfffYZ0dHRXH311XzyySesX7+e9957D4DRo0fz73//m7FjxzJlyhSOHDnCww8/zJgxYwgMDARgypQpjB8/noCAAAYMGEBubi4///wzDz/8sGPfqIg4hEKSiNQY3333HY0aNSrRdtVVV7Fr1y7AuPNs3rx5PPjggwQFBfHJJ5/Qtm1bADw9PVm6dCmPPPIIXbp0wdPTk2HDhvHyyy/bzzV27FhOnz7NK6+8wqOPPoq/vz+33nqr496giDiUxWaz2cwuQkSkslksFhYtWsTgwYPNLkVEqgnNSRIREREphUKSiIiISCk0J0lEagXNLBCRy6WeJBEREZFSKCSJiIiIlEIhSURERKQUCkkiIiIipVBIEhERESmFQpKIiIhIKRSSREREREqhkCQiIiJSiv8HY39LTqnJ+9sAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the results\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "plt.plot(history.history['accuracy'], label='train_accuracy')\n",
    "plt.plot(history.history['val_accuracy'], label='val_accuracy')\n",
    "plt.xlabel('Epoch')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.legend(loc='lower right')\n",
    "plt.show()\n",
    "\n",
    "plt.plot(history.history['loss'], label='train_loss')\n",
    "plt.plot(history.history['val_loss'], label='val_loss')\n",
    "plt.xlabel('Epoch')\n",
    "plt.ylabel('Loss')\n",
    "plt.legend(loc='upper right')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "NNrAxPu9d4Dg"
   },
   "source": [
    "Looking through the results, it seems that while the accuracy keeps converging the validation accuracy gets stucket growing and decreasing drawing peacks. On the other hand, the validation loss grows while the train loss slowly decreases, so it seems that the model is overfitting and not performing well on new cases."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "9C07i1I_d4Dg"
   },
   "source": [
    "## Testing other architectures"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "xy32T3v3d4Dh"
   },
   "source": [
    "Along this section I have tried different model architectures in order to try to improve the model's performance and capacity to generalize. The idea was to first start with a very simple architecture, and see if more complex ones are worth in terms of performance. However, none of this architectures provides good results, so in the next section another approach was taken, studying first the data and making models with better results."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "FEcOuiazd4Dh"
   },
   "source": [
    "### Testing architecture 1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "OYHYIaQtd4Di"
   },
   "source": [
    "On this second approach I had tried to increment a little bit the complexity of the model and using regularization techniques to reduce overfitting. After each convolutional layer a batch normalization one is added before the pooling layer, and also dropout layers with 0.25 dropout rate were added in some of them. Also, another Conv2D block with 256 filters was added before flattening the output. Another changes comparing with the previous model were that two dense layers with corresponding dropouts were added at the end of the model before the output layer."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "e62N4t5Td4Dj"
   },
   "outputs": [],
   "source": [
    "from tensorflow import keras\n",
    "from tensorflow.keras import *\n",
    "from keras.layers import MaxPooling2D, Flatten, Dense, Dropout\n",
    "\n",
    "\n",
    "inputs = keras.Input(shape=(64,64,3))\n",
    "x = layers.Conv2D(filters=64, kernel_size = 3, activation=\"relu\")(inputs)\n",
    "x = layers.BatchNormalization()(x)\n",
    "x = layers.MaxPooling2D(pool_size=2)(x)\n",
    "x = layers.Dropout(0.25)(x)\n",
    "x = layers.Conv2D(filters=128, kernel_size = 3, activation=\"relu\")(x)\n",
    "x = layers.BatchNormalization()(x)\n",
    "x = layers.MaxPooling2D(pool_size=2)(x)\n",
    "x = layers.Dropout(0.25)(x)\n",
    "x = layers.Conv2D(filters=256, kernel_size = 3, activation=\"relu\")(x)\n",
    "x = layers.BatchNormalization()(x)\n",
    "x = layers.MaxPooling2D(pool_size=2)(x)\n",
    "\n",
    "x = layers.Flatten()(x)\n",
    "x = Dense(128,activation=\"relu\")(x)\n",
    "x = Dropout(0.5)(x)\n",
    "x = Dense(64,activation=\"relu\")(x)\n",
    "x = Dropout(0.5)(x)\n",
    "\n",
    "outputs = layers.Dense(1, activation=\"sigmoid\")(x)\n",
    "model = keras.Model(inputs=inputs, outputs=outputs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "SOfQgA9Vd4Dj",
    "outputId": "83d5fd0a-c036-442f-dccf-a6ee4868fc4a"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n",
      "992/992 [==============================] - 12s 12ms/step - loss: 0.4631 - accuracy: 0.7977 - val_loss: 0.3644 - val_accuracy: 0.8570\n",
      "Epoch 2/20\n",
      "992/992 [==============================] - 12s 12ms/step - loss: 0.3401 - accuracy: 0.8691 - val_loss: 0.3442 - val_accuracy: 0.8660\n",
      "Epoch 3/20\n",
      "992/992 [==============================] - 12s 12ms/step - loss: 0.3150 - accuracy: 0.8814 - val_loss: 0.3464 - val_accuracy: 0.8690\n",
      "Epoch 4/20\n",
      "992/992 [==============================] - 12s 12ms/step - loss: 0.3040 - accuracy: 0.8885 - val_loss: 0.3249 - val_accuracy: 0.8810\n",
      "Epoch 5/20\n",
      "992/992 [==============================] - 12s 12ms/step - loss: 0.2954 - accuracy: 0.8913 - val_loss: 0.3562 - val_accuracy: 0.8680\n",
      "Epoch 6/20\n",
      "992/992 [==============================] - 12s 12ms/step - loss: 0.2877 - accuracy: 0.8959 - val_loss: 0.3176 - val_accuracy: 0.8740\n",
      "Epoch 7/20\n",
      "992/992 [==============================] - 12s 12ms/step - loss: 0.2802 - accuracy: 0.8998 - val_loss: 0.3540 - val_accuracy: 0.8750\n",
      "Epoch 8/20\n",
      "992/992 [==============================] - 12s 12ms/step - loss: 0.2772 - accuracy: 0.8994 - val_loss: 0.3361 - val_accuracy: 0.8820\n",
      "Epoch 9/20\n",
      "992/992 [==============================] - 12s 12ms/step - loss: 0.2784 - accuracy: 0.8994 - val_loss: 0.3187 - val_accuracy: 0.8750\n",
      "Epoch 10/20\n",
      "992/992 [==============================] - 12s 12ms/step - loss: 0.2681 - accuracy: 0.9039 - val_loss: 0.3051 - val_accuracy: 0.8950\n",
      "Epoch 11/20\n",
      "992/992 [==============================] - 12s 12ms/step - loss: 0.2679 - accuracy: 0.9044 - val_loss: 0.3266 - val_accuracy: 0.8800\n",
      "Epoch 12/20\n",
      "992/992 [==============================] - 13s 13ms/step - loss: 0.2616 - accuracy: 0.9074 - val_loss: 0.2999 - val_accuracy: 0.8910\n",
      "Epoch 13/20\n",
      "992/992 [==============================] - 12s 12ms/step - loss: 0.2605 - accuracy: 0.9066 - val_loss: 0.3186 - val_accuracy: 0.8840\n",
      "Epoch 14/20\n",
      "992/992 [==============================] - 12s 12ms/step - loss: 0.2563 - accuracy: 0.9101 - val_loss: 0.3144 - val_accuracy: 0.8840\n",
      "Epoch 15/20\n",
      "992/992 [==============================] - 12s 12ms/step - loss: 0.2534 - accuracy: 0.9093 - val_loss: 0.3199 - val_accuracy: 0.8770\n",
      "Epoch 16/20\n",
      "992/992 [==============================] - 12s 12ms/step - loss: 0.2511 - accuracy: 0.9109 - val_loss: 0.3090 - val_accuracy: 0.8820\n",
      "Epoch 17/20\n",
      "992/992 [==============================] - 12s 12ms/step - loss: 0.2473 - accuracy: 0.9116 - val_loss: 0.2935 - val_accuracy: 0.8850\n",
      "Epoch 18/20\n",
      "992/992 [==============================] - 12s 12ms/step - loss: 0.2450 - accuracy: 0.9131 - val_loss: 0.3127 - val_accuracy: 0.8840\n",
      "Epoch 19/20\n",
      "992/992 [==============================] - 12s 12ms/step - loss: 0.2459 - accuracy: 0.9119 - val_loss: 0.2940 - val_accuracy: 0.8830\n",
      "Epoch 20/20\n",
      "992/992 [==============================] - 12s 12ms/step - loss: 0.2403 - accuracy: 0.9141 - val_loss: 0.2949 - val_accuracy: 0.8740\n"
     ]
    }
   ],
   "source": [
    "from tensorflow.keras.optimizers import Adam\n",
    "from tensorflow.keras.losses import BinaryCrossentropy\n",
    "\n",
    "loss = BinaryCrossentropy()\n",
    "num_epochs = 20\n",
    "\n",
    "model.compile(optimizer=Adam(),\n",
    "              loss=loss,\n",
    "              metrics =['accuracy'])\n",
    "\n",
    "# Train the model\n",
    "history = model.fit(ds_train_batch,\n",
    "                    epochs=num_epochs,\n",
    "                    validation_data=ds_test_batch)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "kwU8MLC1d4Dm"
   },
   "outputs": [],
   "source": [
    "# Save the model\n",
    "model.save('models/second_model.h5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "aLFyIGOod4Dn",
    "outputId": "acf87679-79be-406f-befd-60472587ef12"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the results\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "plt.plot(history.history['accuracy'], label='train_accuracy')\n",
    "plt.plot(history.history['val_accuracy'], label='val_accuracy')\n",
    "plt.xlabel('Epoch')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.legend(loc='lower right')\n",
    "plt.show()\n",
    "\n",
    "plt.plot(history.history['loss'], label='train_loss')\n",
    "plt.plot(history.history['val_loss'], label='val_loss')\n",
    "plt.xlabel('Epoch')\n",
    "plt.ylabel('Loss')\n",
    "plt.legend(loc='upper right')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "QDup823ad4Do"
   },
   "source": [
    "With a more complex model and with the double of epochs, we achieve little better results. Validation error seems to be a little less stable with regularization techniques, but it seems to be stucked on almost that previous range of values."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "7SNfeKlqd4EE"
   },
   "source": [
    "### Testing architecture 2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "GOeqq2TRd4EF"
   },
   "source": [
    "On this last approach, I just tried to copy a complex model. In this case I took AlexNet as reference and tried to build a similar model (a little simpler but keeping main concept)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "MU9iakDrd4EH"
   },
   "source": [
    "Read all the data again (now using 128x128 rescaled images):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "P3XC3oIxd4EH",
    "outputId": "7b22668b-ac02-4552-e737-2680ea2e84be"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<TakeDataset shapes: (), types: tf.string>\n",
      "Tensor(\"StringSplit/RaggedGetItem/strided_slice_5:0\", shape=(None,), dtype=string)\n",
      "Tensor(\"StringSplit/RaggedGetItem/strided_slice_5:0\", shape=(None,), dtype=string)\n",
      "/home/mia1/Documentos/DL/faces_dataset_train/i07/nm0322407_rm852269056_1955-3-29_2005.jpg|50|M\n",
      "tf.Tensor([0.8294171 1.1705829], shape=(2,), dtype=float64)\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "from matplotlib import pyplot\n",
    "\n",
    "# Size at which images will be processed\n",
    "img_height = 128\n",
    "img_width = 128\n",
    "\n",
    "# We build a dataset of text strings from the metadata\n",
    "train_metadata = format_metadata(metadata, PATH)\n",
    "train_ds = tf.data.Dataset.from_tensor_slices(train_metadata)\n",
    "print(train_ds.take(1))\n",
    "\n",
    "# Mapping to an image dataset with tags\n",
    "train_ds = train_ds.shuffle(len(train_ds)).map(process_path)\n",
    "\n",
    "# Same process for test\n",
    "test_ds = tf.data.Dataset.from_tensor_slices(format_metadata(metadata_test, PATH_TEST))\n",
    "test_ds = test_ds.map(process_path)\n",
    "\n",
    "    \n",
    "ds_train_transform = train_ds\n",
    "ds_test_transform = test_ds\n",
    "print(train_metadata[0])\n",
    "\n",
    "# OPTIONAL: rescaling input to [-1, 1]\n",
    "ds_train_transform = ds_train_transform.map(lambda image, label: (tf.cast(image, tf.float32)/127.5 - 1., label))\n",
    "ds_test_transform = ds_test_transform.map(lambda image, label: (tf.cast(image, tf.float32)/127.5 - 1., label))\n",
    "\n",
    "\n",
    "# OPTIONAL: sample gender weights\n",
    "import numpy as np\n",
    "labels = [0 if x.split('|')[2] == 'M' else 1 for x in train_metadata]\n",
    "freqs = np.array([1. - np.mean(labels), np.mean(labels)])\n",
    "i_freqs = 1. / freqs\n",
    "sample_weights = tf.convert_to_tensor(len(i_freqs) * i_freqs / i_freqs.sum())\n",
    "print(sample_weights)\n",
    "ds_train_transform = ds_train_transform.map(lambda image, label: (image, label, sample_weights[label]))\n",
    "ds_test_transform = ds_test_transform.map(lambda image, label: (image, label, sample_weights[label]))\n",
    "\n",
    "\n",
    "batch_size = 32\n",
    "ds_train_batch = ds_train_transform.batch(batch_size)\n",
    "ds_test_batch = ds_test_transform.batch(batch_size)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "X4pcuX94d4EI"
   },
   "source": [
    "Building the AlexNet based model:\n",
    "\n",
    "- Input size 128x128\n",
    "- First convolutional layers with higher kernel sizes and strides in order to detect larger features\n",
    "- Padding added to most of the conv layers\n",
    "- Dense layers at the end with dropout\n",
    "- Conv2D and maxpooling combinations, batch normalization after each convolutional layer and dropout after dense ones."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "8Nq3uy7md4EI"
   },
   "outputs": [],
   "source": [
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import Dense, Conv2D, MaxPooling2D, Flatten, Dropout, BatchNormalization\n",
    "\n",
    "model = Sequential([\n",
    "    Conv2D(96, (11,11), strides=(4,4), activation='relu', input_shape=(128,128,3)),\n",
    "    BatchNormalization(),\n",
    "    MaxPooling2D(pool_size=(3,3), strides=(2,2)),\n",
    "    Conv2D(256, (5,5), strides=(1,1), activation='relu', padding=\"same\"),\n",
    "    BatchNormalization(),\n",
    "    MaxPooling2D(pool_size=(3,3), strides=(2,2)),\n",
    "    Conv2D(384, (3,3), strides=(1,1), activation='relu', padding=\"same\"),\n",
    "    Conv2D(384, (3,3), strides=(1,1), activation='relu', padding=\"same\"),\n",
    "    Conv2D(256, (3,3), strides=(1,1), activation='relu', padding=\"same\"),\n",
    "    BatchNormalization(),\n",
    "    MaxPooling2D(pool_size=(3,3), strides=(2,2)),\n",
    "    Flatten(),\n",
    "    Dense(512, activation='relu'),\n",
    "    Dropout(0.5),\n",
    "    Dense(512, activation='relu'),\n",
    "    Dropout(0.5),\n",
    "    Dense(1, activation='sigmoid')\n",
    "])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "ce_HcuNPpB7E",
    "outputId": "de84b418-166b-4330-dcb3-e8c0f9fdd81d"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_1\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " conv2d_23 (Conv2D)          (None, 30, 30, 96)        34944     \n",
      "                                                                 \n",
      " batch_normalization_9 (Batc  (None, 30, 30, 96)       384       \n",
      " hNormalization)                                                 \n",
      "                                                                 \n",
      " max_pooling2d_17 (MaxPoolin  (None, 14, 14, 96)       0         \n",
      " g2D)                                                            \n",
      "                                                                 \n",
      " conv2d_24 (Conv2D)          (None, 14, 14, 256)       614656    \n",
      "                                                                 \n",
      " batch_normalization_10 (Bat  (None, 14, 14, 256)      1024      \n",
      " chNormalization)                                                \n",
      "                                                                 \n",
      " max_pooling2d_18 (MaxPoolin  (None, 6, 6, 256)        0         \n",
      " g2D)                                                            \n",
      "                                                                 \n",
      " conv2d_25 (Conv2D)          (None, 6, 6, 384)         885120    \n",
      "                                                                 \n",
      " conv2d_26 (Conv2D)          (None, 6, 6, 384)         1327488   \n",
      "                                                                 \n",
      " conv2d_27 (Conv2D)          (None, 6, 6, 256)         884992    \n",
      "                                                                 \n",
      " batch_normalization_11 (Bat  (None, 6, 6, 256)        1024      \n",
      " chNormalization)                                                \n",
      "                                                                 \n",
      " max_pooling2d_19 (MaxPoolin  (None, 2, 2, 256)        0         \n",
      " g2D)                                                            \n",
      "                                                                 \n",
      " flatten_7 (Flatten)         (None, 1024)              0         \n",
      "                                                                 \n",
      " dense_13 (Dense)            (None, 512)               524800    \n",
      "                                                                 \n",
      " dropout_10 (Dropout)        (None, 512)               0         \n",
      "                                                                 \n",
      " dense_14 (Dense)            (None, 512)               262656    \n",
      "                                                                 \n",
      " dropout_11 (Dropout)        (None, 512)               0         \n",
      "                                                                 \n",
      " dense_15 (Dense)            (None, 1)                 513       \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 4,537,601\n",
      "Trainable params: 4,536,385\n",
      "Non-trainable params: 1,216\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "n70-0cMGd4EJ"
   },
   "source": [
    "Training:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "tZsLFP9td4EJ",
    "outputId": "5ac523e8-0447-420c-edbb-0d055b54d7e2"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "3965/3965 [==============================] - 25s 6ms/step - loss: 0.4157 - accuracy: 0.8163 - val_loss: 0.3763 - val_accuracy: 0.8600\n",
      "Epoch 2/10\n",
      "3965/3965 [==============================] - 24s 6ms/step - loss: 0.3160 - accuracy: 0.8779 - val_loss: 0.3553 - val_accuracy: 0.8690\n",
      "Epoch 3/10\n",
      "3965/3965 [==============================] - 24s 6ms/step - loss: 0.2950 - accuracy: 0.8879 - val_loss: 0.3299 - val_accuracy: 0.8670\n",
      "Epoch 4/10\n",
      "3965/3965 [==============================] - 24s 6ms/step - loss: 0.2790 - accuracy: 0.8952 - val_loss: 0.3046 - val_accuracy: 0.8850\n",
      "Epoch 5/10\n",
      "3965/3965 [==============================] - 24s 6ms/step - loss: 0.2685 - accuracy: 0.9003 - val_loss: 0.3420 - val_accuracy: 0.8690\n",
      "Epoch 6/10\n",
      "3965/3965 [==============================] - 23s 6ms/step - loss: 0.2573 - accuracy: 0.9060 - val_loss: 0.3322 - val_accuracy: 0.8800\n",
      "Epoch 7/10\n",
      "3965/3965 [==============================] - 23s 6ms/step - loss: 0.2479 - accuracy: 0.9101 - val_loss: 0.3325 - val_accuracy: 0.8770\n",
      "Epoch 8/10\n",
      "3965/3965 [==============================] - 24s 6ms/step - loss: 0.2392 - accuracy: 0.9128 - val_loss: 0.3159 - val_accuracy: 0.8740\n",
      "Epoch 9/10\n",
      "3965/3965 [==============================] - 23s 6ms/step - loss: 0.2305 - accuracy: 0.9167 - val_loss: 0.3217 - val_accuracy: 0.8620\n",
      "Epoch 10/10\n",
      "3965/3965 [==============================] - 23s 6ms/step - loss: 0.2186 - accuracy: 0.9206 - val_loss: 0.3174 - val_accuracy: 0.8740\n"
     ]
    }
   ],
   "source": [
    "from tensorflow.keras.optimizers import Adam\n",
    "from tensorflow.keras.losses import BinaryCrossentropy\n",
    "from tensorflow.keras.callbacks import EarlyStopping\n",
    "\n",
    "loss = BinaryCrossentropy()\n",
    "num_epochs = 10\n",
    "\n",
    "model.compile(optimizer=Adam(),\n",
    "              loss=loss,\n",
    "              metrics =['accuracy'])\n",
    "\n",
    "early_stopping = EarlyStopping(monitor='val_loss', patience=3, restore_best_weights=True)\n",
    "# Train the model\n",
    "history = model.fit(ds_train_batch,\n",
    "                    epochs=num_epochs,\n",
    "                    validation_data=ds_test_batch\n",
    "                    callbacks=[early_stopping])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "uapdSS5xpB7G"
   },
   "outputs": [],
   "source": [
    "# Save the model\n",
    "model.save('models/third_model.h5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "3MLXr3sMpB7G",
    "outputId": "ebfa29a2-05e1-465d-adbc-a49123e759c8"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the results\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "plt.plot(history.history['accuracy'], label='train_accuracy')\n",
    "plt.plot(history.history['val_accuracy'], label='val_accuracy')\n",
    "plt.xlabel('Epoch')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.legend(loc='lower right')\n",
    "plt.show()\n",
    "\n",
    "plt.plot(history.history['loss'], label='train_loss')\n",
    "plt.plot(history.history['val_loss'], label='val_loss')\n",
    "plt.xlabel('Epoch')\n",
    "plt.ylabel('Loss')\n",
    "plt.legend(loc='upper right')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "astZaWerp8TT"
   },
   "source": [
    "It seems that the results for this much complex model are equal or even worse than with simpler models. Here, the difference between training and validation error is even more pronunciated. So in next section I tried a different approach."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "47hq7E1vpB7H"
   },
   "source": [
    "# Second approach"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Qa9KmLPSpB7H"
   },
   "source": [
    "For this second approach, I decided to take a look to the data. \n",
    "- There are around 120k images on the training set\n",
    "- And only 1k on the new validation set\n",
    "\n",
    "As models seemed to be stucked on that validation performance, I have suppossed that the data on the given test set is not sufficient or at least not enought representative. However, the training set is quite large, so I use a 20% split of the training set to create a new validation set. Also, I work here with 128x128 images (I have checked it also with the old data, but results remain similar. Its true that 64x64 and specially 32x32 are very small image sizes to fed the model with to try to detect gender on a face)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "FY_CjG_zpB7H"
   },
   "source": [
    "## Modified preprocessing"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "arOvqwHHpB7O"
   },
   "source": [
    "Read metadata again:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "id": "xhu8oie7pB7P",
    "outputId": "3af11cc3-e8fc-4687-9308-78795afa7f64"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "i07/nm0322407_rm852269056_1955-3-29_2005.jpg ----> {'age': 50, 'gender': 'M'}\n"
     ]
    }
   ],
   "source": [
    "import json\n",
    "\n",
    "with open(PATH + 'metadata.json') as fIn:\n",
    "    metadata = json.loads(fIn.read())\n",
    "with open(PATH_TEST + 'metadata.json') as fIn:\n",
    "    metadata_test = json.loads(fIn.read())\n",
    "\n",
    "metadata_example_key = next(iter(metadata.keys()))\n",
    "metadata_example_value = metadata[metadata_example_key]\n",
    "\n",
    "print(metadata_example_key, end=' ----> ')\n",
    "print(metadata_example_value)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "2LTHwhuPpB7P"
   },
   "source": [
    "Defining height and width"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "id": "_F7U7XplpB7P"
   },
   "outputs": [],
   "source": [
    "# Size at which images will be processed\n",
    "img_height = 128\n",
    "img_width = 128"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "tIgEdStWpB7Q"
   },
   "source": [
    "Splitting training set into train and validation (Remember that above cell to define functions must be executed)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "id": "EsIyMQfipB7Q"
   },
   "outputs": [],
   "source": [
    "def decode_img(img):\n",
    "    '''Transforms a string that represents the path of an image into a matrix\n",
    "    (img_height, img_width, 3) of uint8 representing the pixel values\n",
    "    '''\n",
    "    # Convert the compressed string to a 3D uint8 tensor\n",
    "    img = tf.io.decode_jpeg(img, channels=3)\n",
    "    # Resize the image to the desired size.\n",
    "    return tf.cast(tf.image.resize(img, [img_height, img_width]),tf.uint8)\n",
    "\n",
    "def process_path(file_path):\n",
    "    '''Transforms strings of the form 'path|age|gender' to tuples (image, label)\n",
    "    '''\n",
    "    parts = tf.strings.split(file_path, '|')\n",
    "    print(parts)\n",
    "    gender = tf.cast(0 if parts[2] == 'M' else 1, tf.dtypes.int32)\n",
    "    # Load the raw data from the file as a string\n",
    "    img = tf.io.read_file(parts[0])\n",
    "    img = decode_img(img)\n",
    "    return img, gender\n",
    "\n",
    "def format_metadata(m, path):\n",
    "    '''Transforms the metadata dictionary m into a list of strings of the form 'path to the image|age|gender'.\n",
    "    '''\n",
    "    return list(map(lambda x: '{0}{1}|{2}|{3}'.format(path,x,m[x]['age'],m[x]['gender']),m.keys()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "id": "gTOe4dWtpB7Q",
    "outputId": "db66700f-4ee2-4158-badb-e6b73c0dd47e"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Tensor(\"StringSplit/RaggedGetItem/strided_slice_5:0\", shape=(None,), dtype=string)\n",
      "Tensor(\"StringSplit/RaggedGetItem/strided_slice_5:0\", shape=(None,), dtype=string)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-03-16 01:24:07.386402: I tensorflow/core/platform/cpu_feature_guard.cc:151] 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",
      "2023-03-16 01:24:07.388084: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:939] 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",
      "2023-03-16 01:24:07.388818: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:939] 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",
      "2023-03-16 01:24:07.389503: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:939] 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",
      "2023-03-16 01:24:08.374344: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:939] 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",
      "2023-03-16 01:24:08.375460: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:939] 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",
      "2023-03-16 01:24:08.376510: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:939] 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",
      "2023-03-16 01:24:08.377517: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1525] Created device /job:localhost/replica:0/task:0/device:GPU:0 with 22027 MB memory:  -> device: 0, name: NVIDIA GeForce RTX 3090, pci bus id: 0000:01:00.0, compute capability: 8.6\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "from matplotlib import pyplot\n",
    "\n",
    "# We build a dataset of text strings from the metadata\n",
    "train_metadata = format_metadata(metadata, PATH)\n",
    "train_ds = tf.data.Dataset.from_tensor_slices(train_metadata)\n",
    "\n",
    "# Split in training and validation from training set\n",
    "validation_percentage = 20\n",
    "\n",
    "train_size = int((100 - validation_percentage) / 100.0 * len(train_ds))\n",
    "validation_size = len(train_ds) - train_size\n",
    "\n",
    "train_ds = train_ds.shuffle(len(train_ds)) # shuffle data\n",
    "validation_ds = train_ds.take(validation_size).map(process_path) # take images of 20% of the dataset\n",
    "train_ds = train_ds.skip(validation_size).map(process_path) # take images unless the other 20%"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "C4T5POOUpB7Q",
    "outputId": "7954289a-e4a9-4d4d-d021-09d80a4a044d"
   },
   "outputs": [],
   "source": [
    "ds_train_transform = train_ds\n",
    "ds_test_transform = validation_ds\n",
    "print(train_metadata[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "tDOIzCOMpB7R"
   },
   "source": [
    "Rescaling input"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "id": "UiLy6Bh8pB7R"
   },
   "outputs": [],
   "source": [
    "# OPTIONAL: rescaling input to [-1, 1]\n",
    "ds_train_transform = ds_train_transform.map(lambda image, label: (tf.cast(image, tf.float32)/127.5 - 1., label))\n",
    "ds_test_transform = ds_test_transform.map(lambda image, label: (tf.cast(image, tf.float32)/127.5 - 1., label))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "9Nk-MBFKpB7R"
   },
   "source": [
    "Stratification of the data:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "id": "a0lwezGtpB7T",
    "outputId": "4f901175-e23d-4250-ea65-f9a53bfcfa19"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tf.Tensor([0.8294171 1.1705829], shape=(2,), dtype=float64)\n"
     ]
    }
   ],
   "source": [
    "# OPTIONAL: sample gender weights\n",
    "import numpy as np\n",
    "labels = [0 if x.split('|')[2] == 'M' else 1 for x in train_metadata]\n",
    "freqs = np.array([1. - np.mean(labels), np.mean(labels)])\n",
    "i_freqs = 1. / freqs\n",
    "sample_weights = tf.convert_to_tensor(len(i_freqs) * i_freqs / i_freqs.sum())\n",
    "print(sample_weights)\n",
    "ds_train_transform = ds_train_transform.map(lambda image, label: (image, label, sample_weights[label]))\n",
    "ds_test_transform = ds_test_transform.map(lambda image, label: (image, label, sample_weights[label]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "KOpl--mHpB7T"
   },
   "source": [
    "Checking train and validation sets:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "id": "ynnn9I9LpB7T",
    "outputId": "e3c24058-a749-4059-cde4-38e22f548237"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test set size: 25373\n",
      "Train set size: 101492\n"
     ]
    }
   ],
   "source": [
    "print('Test set size: ' + str(len(ds_test_transform)))\n",
    "print('Train set size: ' + str(len(ds_train_transform)))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "cSvV0IWEpB7T"
   },
   "source": [
    "Defyning batches:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "id": "KNFIKDBApB7T"
   },
   "outputs": [],
   "source": [
    "batch_size = 128\n",
    "ds_train_batch = ds_train_transform.batch(batch_size)\n",
    "ds_test_batch = ds_test_transform.batch(batch_size)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "x-UNe8u3pB7T"
   },
   "source": [
    "# Training with modified data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "1KicxBSzpB7T"
   },
   "source": [
    "Lets take a look on how this modifications affect the previously designed models"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "FFjFzhg9pB7U"
   },
   "source": [
    "## TestingModel class"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "8-uo4CUhpB7U"
   },
   "source": [
    "For convenience, I define a class (**TestingModel**) to define the seen architectures, compile, train, save and plot the model with a little less of repeated code.\n",
    "\n",
    "Three architectures defined: \n",
    "- **basic** : first one seen, consisting on Conv2D-MaxPooling layers combinations\n",
    "- **regularized** : second seen before, with regularization techniques\n",
    "- **alexnet** : last seen, based on AlexNet"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "id": "vdCbKF-opB7U"
   },
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "from tensorflow import keras\n",
    "from tensorflow.keras import layers\n",
    "from tensorflow.keras.optimizers import Adam\n",
    "from tensorflow.keras.models import load_model\n",
    "from tensorflow.keras.losses import BinaryCrossentropy\n",
    "from tensorflow.keras.layers import Dense, Dropout, Flatten, Conv2D, MaxPooling2D, BatchNormalization, Input, Add\n",
    "from tensorflow.keras.callbacks import EarlyStopping\n",
    "from tensorflow.keras import regularizers\n",
    "from tensorflow.keras import Sequential\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "class TestingModel:\n",
    "    def __init__(self, input_shape):\n",
    "        self.input_shape = input_shape\n",
    "        self.model = None\n",
    "    \n",
    "    def build(self, model_name):\n",
    "        if model_name == 'basic':\n",
    "            inputs = keras.Input(shape=self.input_shape)\n",
    "            x = layers.Conv2D(filters=32, kernel_size = 3, activation=\"relu\")(inputs)\n",
    "            x = layers.MaxPooling2D(pool_size=2)(x)\n",
    "            x = layers.Conv2D(filters=64, kernel_size = 3, activation=\"relu\")(x)\n",
    "            x = layers.MaxPooling2D(pool_size=2)(x)\n",
    "            x = layers.Conv2D(filters=128, kernel_size = 3, activation=\"relu\")(x)\n",
    "            x = layers.Flatten()(x)\n",
    "\n",
    "            outputs = layers.Dense(1, activation=\"sigmoid\")(x)\n",
    "            model = keras.Model(inputs=inputs, outputs=outputs)\n",
    "        elif model_name == 'regularized':\n",
    "            inputs = keras.Input(shape=self.input_shape)\n",
    "            x = layers.Conv2D(filters=64, kernel_size = 3, activation=\"relu\")(inputs)\n",
    "            x = layers.BatchNormalization()(x)\n",
    "            x = layers.MaxPooling2D(pool_size=2)(x)\n",
    "            x = layers.Dropout(0.25)(x)\n",
    "            x = layers.Conv2D(filters=128, kernel_size = 3, activation=\"relu\")(x)\n",
    "            x = layers.BatchNormalization()(x)\n",
    "            x = layers.MaxPooling2D(pool_size=2)(x)\n",
    "            x = layers.Dropout(0.25)(x)\n",
    "            x = layers.Conv2D(filters=256, kernel_size = 3, activation=\"relu\")(x)\n",
    "            x = layers.BatchNormalization()(x)\n",
    "            x = layers.MaxPooling2D(pool_size=2)(x)\n",
    "\n",
    "            x = layers.Flatten()(x)\n",
    "            x = Dense(128,activation=\"relu\")(x)\n",
    "            x = Dropout(0.5)(x)\n",
    "            x = Dense(64,activation=\"relu\")(x)\n",
    "            x = Dropout(0.5)(x)\n",
    "\n",
    "            outputs = layers.Dense(1, activation=\"sigmoid\")(x)\n",
    "            model = keras.Model(inputs=inputs, outputs=outputs)\n",
    "        elif model_name == 'alexnet':\n",
    "            model = Sequential([\n",
    "            Conv2D(96, (11,11), strides=(4,4), activation='relu', input_shape=(128,128,3)),\n",
    "            BatchNormalization(),\n",
    "            MaxPooling2D(pool_size=(3,3), strides=(2,2)),\n",
    "            Conv2D(256, (5,5), strides=(1,1), activation='relu', padding=\"same\"),\n",
    "            BatchNormalization(),\n",
    "            MaxPooling2D(pool_size=(3,3), strides=(2,2)),\n",
    "            Conv2D(384, (3,3), strides=(1,1), activation='relu', padding=\"same\"),\n",
    "            Conv2D(384, (3,3), strides=(1,1), activation='relu', padding=\"same\"),\n",
    "            Conv2D(256, (3,3), strides=(1,1), activation='relu', padding=\"same\"),\n",
    "            BatchNormalization(),\n",
    "            MaxPooling2D(pool_size=(3,3), strides=(2,2)),\n",
    "            Flatten(),\n",
    "            Dense(512, activation='relu'),\n",
    "            Dropout(0.5),\n",
    "            Dense(512, activation='relu'),\n",
    "            Dropout(0.5),\n",
    "            Dense(1, activation='sigmoid')\n",
    "        ])\n",
    "        else:\n",
    "            raise ValueError('Invalid model name')\n",
    "\n",
    "        self.model = model\n",
    "        return model\n",
    "\n",
    "    def train(self, train_ds, val_ds, epochs, optimizer, loss_fn, metrics, early_stopping=None):\n",
    "            self.model.compile(optimizer=optimizer, loss=loss_fn, metrics=metrics)\n",
    "            if early_stopping is not None:\n",
    "                history = self.model.fit(train_ds, validation_data=val_ds, epochs=epochs, callbacks=early_stopping)\n",
    "            else: \n",
    "                history = self.model.fit(train_ds, validation_data=val_ds, epochs=epochs)\n",
    "\n",
    "            return history\n",
    "\n",
    "    def summary(self):\n",
    "        self.model.summary()\n",
    "\n",
    "    def save(self,full_path):\n",
    "        self.model.save(full_path)\n",
    "\n",
    "    def plot(self,history):\n",
    "        # Plot the results\n",
    "        plt.plot(history.history['accuracy'], label='train_accuracy')\n",
    "        plt.plot(history.history['val_accuracy'], label='val_accuracy')\n",
    "        plt.xlabel('Epoch')\n",
    "        plt.ylabel('Accuracy')\n",
    "        plt.legend(loc='lower right')\n",
    "        plt.show()\n",
    "\n",
    "        plt.plot(history.history['loss'], label='train_loss')\n",
    "        plt.plot(history.history['val_loss'], label='val_loss')\n",
    "        plt.xlabel('Epoch')\n",
    "        plt.ylabel('Loss')\n",
    "        plt.legend(loc='upper right')\n",
    "        plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "BOWmRqPzpB7V"
   },
   "source": [
    "## Models training"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The training of the models is showed below. \n",
    "\n",
    "- Adam optimizer is used. I test others but Adam is the one that gives better results\n",
    "- Early Stopping with a patience of 5 epochs without increasing the validation loss is implemented on each model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "id": "x9Y2_pSppB7W"
   },
   "outputs": [],
   "source": [
    "tm = TestingModel((128,128,3))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "id": "DuwR1k-0pB7W"
   },
   "outputs": [],
   "source": [
    "epochs = 50\n",
    "optimizer = Adam() # I tried other optimizers but Adam is the one that gives better results\n",
    "loss = BinaryCrossentropy()\n",
    "metrics = ['accuracy']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "q8nLXpUXpB7W"
   },
   "source": [
    "Training the <span style=\"color:tomato\">**first model**</span> again with 50 epochs:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "id": "jaNWrSJspB7W",
    "outputId": "82f9cd87-6eb7-4176-b693-bcc56ca46bd6"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " input_1 (InputLayer)        [(None, 128, 128, 3)]     0         \n",
      "                                                                 \n",
      " conv2d (Conv2D)             (None, 126, 126, 32)      896       \n",
      "                                                                 \n",
      " max_pooling2d (MaxPooling2D  (None, 63, 63, 32)       0         \n",
      " )                                                               \n",
      "                                                                 \n",
      " conv2d_1 (Conv2D)           (None, 61, 61, 64)        18496     \n",
      "                                                                 \n",
      " max_pooling2d_1 (MaxPooling  (None, 30, 30, 64)       0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " conv2d_2 (Conv2D)           (None, 28, 28, 128)       73856     \n",
      "                                                                 \n",
      " flatten (Flatten)           (None, 100352)            0         \n",
      "                                                                 \n",
      " dense (Dense)               (None, 1)                 100353    \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 193,601\n",
      "Trainable params: 193,601\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/50\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-03-16 01:27:06.167917: I tensorflow/stream_executor/cuda/cuda_dnn.cc:366] Loaded cuDNN version 8100\n",
      "2023-03-16 01:27:07.591353: I tensorflow/core/platform/default/subprocess.cc:304] Start cannot spawn child process: No such file or directory\n",
      "2023-03-16 01:27:07.592171: I tensorflow/core/platform/default/subprocess.cc:304] Start cannot spawn child process: No such file or directory\n",
      "2023-03-16 01:27:07.592194: W tensorflow/stream_executor/gpu/asm_compiler.cc:80] Couldn't get ptxas version string: INTERNAL: Couldn't invoke ptxas --version\n",
      "2023-03-16 01:27:07.593034: I tensorflow/core/platform/default/subprocess.cc:304] Start cannot spawn child process: No such file or directory\n",
      "2023-03-16 01:27:07.593094: W tensorflow/stream_executor/gpu/redzone_allocator.cc:314] INTERNAL: Failed to launch ptxas\n",
      "Relying on driver to perform ptx compilation. \n",
      "Modify $PATH to customize ptxas location.\n",
      "This message will be only logged once.\n",
      "2023-03-16 01:27:08.530642: I tensorflow/stream_executor/cuda/cuda_blas.cc:1774] TensorFloat-32 will be used for the matrix multiplication. This will only be logged once.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "793/793 [==============================] - 18s 17ms/step - loss: 0.4074 - accuracy: 0.8147 - val_loss: 0.3202 - val_accuracy: 0.8715\n",
      "Epoch 2/50\n",
      "793/793 [==============================] - 14s 17ms/step - loss: 0.3132 - accuracy: 0.8767 - val_loss: 0.2948 - val_accuracy: 0.8909\n",
      "Epoch 3/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.2871 - accuracy: 0.8913 - val_loss: 0.2859 - val_accuracy: 0.8794\n",
      "Epoch 4/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.2735 - accuracy: 0.8980 - val_loss: 0.2571 - val_accuracy: 0.9042\n",
      "Epoch 5/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.2631 - accuracy: 0.9018 - val_loss: 0.2539 - val_accuracy: 0.9025\n",
      "Epoch 6/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.2524 - accuracy: 0.9048 - val_loss: 0.2424 - val_accuracy: 0.9137\n",
      "Epoch 7/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.2436 - accuracy: 0.9062 - val_loss: 0.2235 - val_accuracy: 0.9204\n",
      "Epoch 8/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.2347 - accuracy: 0.9091 - val_loss: 0.2203 - val_accuracy: 0.9176\n",
      "Epoch 9/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.2247 - accuracy: 0.9121 - val_loss: 0.2005 - val_accuracy: 0.9233\n",
      "Epoch 10/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.2129 - accuracy: 0.9169 - val_loss: 0.1896 - val_accuracy: 0.9296\n",
      "Epoch 11/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.2037 - accuracy: 0.9197 - val_loss: 0.1772 - val_accuracy: 0.9321\n",
      "Epoch 12/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.1926 - accuracy: 0.9236 - val_loss: 0.1756 - val_accuracy: 0.9320\n",
      "Epoch 13/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.1854 - accuracy: 0.9247 - val_loss: 0.1699 - val_accuracy: 0.9336\n",
      "Epoch 14/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.1787 - accuracy: 0.9273 - val_loss: 0.1551 - val_accuracy: 0.9386\n",
      "Epoch 15/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.1704 - accuracy: 0.9288 - val_loss: 0.1447 - val_accuracy: 0.9419\n",
      "Epoch 16/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.1622 - accuracy: 0.9328 - val_loss: 0.1376 - val_accuracy: 0.9466\n",
      "Epoch 17/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.1612 - accuracy: 0.9324 - val_loss: 0.1385 - val_accuracy: 0.9441\n",
      "Epoch 18/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.1552 - accuracy: 0.9346 - val_loss: 0.1284 - val_accuracy: 0.9475\n",
      "Epoch 19/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.1499 - accuracy: 0.9374 - val_loss: 0.1237 - val_accuracy: 0.9497\n",
      "Epoch 20/50\n",
      "793/793 [==============================] - 14s 17ms/step - loss: 0.1470 - accuracy: 0.9387 - val_loss: 0.1368 - val_accuracy: 0.9453\n",
      "Epoch 21/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.1431 - accuracy: 0.9394 - val_loss: 0.1224 - val_accuracy: 0.9493\n",
      "Epoch 22/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.1402 - accuracy: 0.9403 - val_loss: 0.1188 - val_accuracy: 0.9511\n",
      "Epoch 23/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.1372 - accuracy: 0.9425 - val_loss: 0.1176 - val_accuracy: 0.9504\n",
      "Epoch 24/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.1369 - accuracy: 0.9416 - val_loss: 0.1165 - val_accuracy: 0.9515\n",
      "Epoch 25/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.1364 - accuracy: 0.9411 - val_loss: 0.1145 - val_accuracy: 0.9534\n",
      "Epoch 26/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.1341 - accuracy: 0.9426 - val_loss: 0.1134 - val_accuracy: 0.9535\n",
      "Epoch 27/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.1331 - accuracy: 0.9433 - val_loss: 0.1138 - val_accuracy: 0.9511\n",
      "Epoch 28/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.1296 - accuracy: 0.9440 - val_loss: 0.1104 - val_accuracy: 0.9527\n",
      "Epoch 29/50\n",
      "793/793 [==============================] - 14s 17ms/step - loss: 0.1291 - accuracy: 0.9444 - val_loss: 0.1170 - val_accuracy: 0.9507\n",
      "Epoch 30/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.1281 - accuracy: 0.9453 - val_loss: 0.1015 - val_accuracy: 0.9571\n",
      "Epoch 31/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.1277 - accuracy: 0.9447 - val_loss: 0.1072 - val_accuracy: 0.9541\n",
      "Epoch 32/50\n",
      "793/793 [==============================] - 14s 17ms/step - loss: 0.1250 - accuracy: 0.9453 - val_loss: 0.1041 - val_accuracy: 0.9564\n",
      "Epoch 33/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.1250 - accuracy: 0.9467 - val_loss: 0.1051 - val_accuracy: 0.9554\n",
      "Epoch 34/50\n",
      "793/793 [==============================] - 14s 17ms/step - loss: 0.1238 - accuracy: 0.9465 - val_loss: 0.1075 - val_accuracy: 0.9556\n",
      "Epoch 35/50\n",
      "793/793 [==============================] - 14s 17ms/step - loss: 0.1246 - accuracy: 0.9457 - val_loss: 0.1056 - val_accuracy: 0.9558\n",
      "Epoch 36/50\n",
      "793/793 [==============================] - 14s 17ms/step - loss: 0.1232 - accuracy: 0.9468 - val_loss: 0.1041 - val_accuracy: 0.9566\n",
      "Epoch 37/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.1214 - accuracy: 0.9475 - val_loss: 0.1004 - val_accuracy: 0.9568\n",
      "Epoch 38/50\n",
      "793/793 [==============================] - 14s 17ms/step - loss: 0.1215 - accuracy: 0.9477 - val_loss: 0.1078 - val_accuracy: 0.9549\n",
      "Epoch 39/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.1226 - accuracy: 0.9461 - val_loss: 0.1003 - val_accuracy: 0.9572\n",
      "Epoch 40/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.1216 - accuracy: 0.9470 - val_loss: 0.0970 - val_accuracy: 0.9577\n",
      "Epoch 41/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.1199 - accuracy: 0.9479 - val_loss: 0.1012 - val_accuracy: 0.9561\n",
      "Epoch 42/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.1204 - accuracy: 0.9474 - val_loss: 0.0943 - val_accuracy: 0.9586\n",
      "Epoch 43/50\n",
      "793/793 [==============================] - 14s 17ms/step - loss: 0.1194 - accuracy: 0.9476 - val_loss: 0.1035 - val_accuracy: 0.9560\n",
      "Epoch 44/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.1185 - accuracy: 0.9481 - val_loss: 0.0941 - val_accuracy: 0.9586\n",
      "Epoch 45/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.1176 - accuracy: 0.9480 - val_loss: 0.0984 - val_accuracy: 0.9564\n",
      "Epoch 46/50\n",
      "793/793 [==============================] - 14s 17ms/step - loss: 0.1185 - accuracy: 0.9484 - val_loss: 0.1028 - val_accuracy: 0.9545\n",
      "Epoch 47/50\n",
      "793/793 [==============================] - 14s 17ms/step - loss: 0.1174 - accuracy: 0.9484 - val_loss: 0.1020 - val_accuracy: 0.9565\n",
      "Epoch 48/50\n",
      "793/793 [==============================] - 13s 17ms/step - loss: 0.1176 - accuracy: 0.9483 - val_loss: 0.0987 - val_accuracy: 0.9577\n",
      "Epoch 49/50\n",
      "793/793 [==============================] - 14s 17ms/step - loss: 0.1163 - accuracy: 0.9490 - val_loss: 0.0973 - val_accuracy: 0.9565\n",
      "Epoch 50/50\n",
      "793/793 [==============================] - 14s 17ms/step - loss: 0.1167 - accuracy: 0.9481 - val_loss: 0.0966 - val_accuracy: 0.9574\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "tm.build(model_name= 'basic')\n",
    "tm.summary()\n",
    "history = tm.train(ds_train_batch, ds_test_batch, epochs, optimizer, loss, metrics)\n",
    "tm.plot(history)\n",
    "tm.save('models/first_fixed.h5')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "7wLcEthQpB7X"
   },
   "source": [
    "Training the <span style=\"color:tomato\">*second model*</span> again (for this model i have used 20 epochs because of problems with cuda) :"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "tU3VIpd-pB7X",
    "outputId": "1729eed4-2cfe-4d9f-c208-f6b69f529a0b"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model_4\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " input_6 (InputLayer)        [(None, 128, 128, 3)]     0         \n",
      "                                                                 \n",
      " conv2d_12 (Conv2D)          (None, 126, 126, 64)      1792      \n",
      "                                                                 \n",
      " batch_normalization_9 (Batc  (None, 126, 126, 64)     256       \n",
      " hNormalization)                                                 \n",
      "                                                                 \n",
      " max_pooling2d_11 (MaxPoolin  (None, 63, 63, 64)       0         \n",
      " g2D)                                                            \n",
      "                                                                 \n",
      " dropout_12 (Dropout)        (None, 63, 63, 64)        0         \n",
      "                                                                 \n",
      " conv2d_13 (Conv2D)          (None, 61, 61, 128)       73856     \n",
      "                                                                 \n",
      " batch_normalization_10 (Bat  (None, 61, 61, 128)      512       \n",
      " chNormalization)                                                \n",
      "                                                                 \n",
      " max_pooling2d_12 (MaxPoolin  (None, 30, 30, 128)      0         \n",
      " g2D)                                                            \n",
      "                                                                 \n",
      " dropout_13 (Dropout)        (None, 30, 30, 128)       0         \n",
      "                                                                 \n",
      " conv2d_14 (Conv2D)          (None, 28, 28, 256)       295168    \n",
      "                                                                 \n",
      " batch_normalization_11 (Bat  (None, 28, 28, 256)      1024      \n",
      " chNormalization)                                                \n",
      "                                                                 \n",
      " max_pooling2d_13 (MaxPoolin  (None, 14, 14, 256)      0         \n",
      " g2D)                                                            \n",
      "                                                                 \n",
      " flatten_4 (Flatten)         (None, 50176)             0         \n",
      "                                                                 \n",
      " dense_10 (Dense)            (None, 128)               6422656   \n",
      "                                                                 \n",
      " dropout_14 (Dropout)        (None, 128)               0         \n",
      "                                                                 \n",
      " dense_11 (Dense)            (None, 64)                8256      \n",
      "                                                                 \n",
      " dropout_15 (Dropout)        (None, 64)                0         \n",
      "                                                                 \n",
      " dense_12 (Dense)            (None, 1)                 65        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 6,803,585\n",
      "Trainable params: 6,802,689\n",
      "Non-trainable params: 896\n",
      "_________________________________________________________________\n",
      "Epoch 1/20\n",
      "793/793 [==============================] - 43s 53ms/step - loss: 1.0541 - accuracy: 0.6324 - val_loss: 0.5633 - val_accuracy: 0.7456\n",
      "Epoch 2/20\n",
      "793/793 [==============================] - 42s 52ms/step - loss: 0.5586 - accuracy: 0.7251 - val_loss: 0.4895 - val_accuracy: 0.7767\n",
      "Epoch 3/20\n",
      "793/793 [==============================] - 41s 52ms/step - loss: 0.4849 - accuracy: 0.7813 - val_loss: 0.4549 - val_accuracy: 0.8224\n",
      "Epoch 4/20\n",
      "793/793 [==============================] - 42s 52ms/step - loss: 0.4260 - accuracy: 0.8208 - val_loss: 0.3717 - val_accuracy: 0.8446\n",
      "Epoch 5/20\n",
      "793/793 [==============================] - 41s 52ms/step - loss: 0.3646 - accuracy: 0.8565 - val_loss: 0.3279 - val_accuracy: 0.8837\n",
      "Epoch 6/20\n",
      "793/793 [==============================] - 42s 52ms/step - loss: 0.3367 - accuracy: 0.8728 - val_loss: 0.3157 - val_accuracy: 0.8838\n",
      "Epoch 7/20\n",
      "793/793 [==============================] - 41s 52ms/step - loss: 0.3187 - accuracy: 0.8829 - val_loss: 0.2870 - val_accuracy: 0.9006\n",
      "Epoch 8/20\n",
      "793/793 [==============================] - 42s 52ms/step - loss: 0.3088 - accuracy: 0.8883 - val_loss: 0.2807 - val_accuracy: 0.9004\n",
      "Epoch 9/20\n",
      "793/793 [==============================] - 42s 52ms/step - loss: 0.2975 - accuracy: 0.8930 - val_loss: 0.2964 - val_accuracy: 0.8955\n",
      "Epoch 10/20\n",
      "793/793 [==============================] - 41s 52ms/step - loss: 0.2900 - accuracy: 0.8966 - val_loss: 0.2850 - val_accuracy: 0.9015\n",
      "Epoch 11/20\n",
      "793/793 [==============================] - 41s 52ms/step - loss: 0.2827 - accuracy: 0.8997 - val_loss: 0.2781 - val_accuracy: 0.9122\n",
      "Epoch 12/20\n",
      "793/793 [==============================] - 41s 52ms/step - loss: 0.2764 - accuracy: 0.9019 - val_loss: 0.2661 - val_accuracy: 0.9046\n",
      "Epoch 13/20\n",
      "793/793 [==============================] - 41s 52ms/step - loss: 0.2679 - accuracy: 0.9052 - val_loss: 0.2573 - val_accuracy: 0.9131\n",
      "Epoch 14/20\n",
      "793/793 [==============================] - 41s 52ms/step - loss: 0.2608 - accuracy: 0.9080 - val_loss: 0.2458 - val_accuracy: 0.9112\n",
      "Epoch 15/20\n",
      "793/793 [==============================] - 41s 52ms/step - loss: 0.2574 - accuracy: 0.9098 - val_loss: 0.2313 - val_accuracy: 0.9215\n",
      "Epoch 16/20\n",
      "793/793 [==============================] - 41s 52ms/step - loss: 0.2528 - accuracy: 0.9103 - val_loss: 0.2545 - val_accuracy: 0.9175\n",
      "Epoch 17/20\n",
      "793/793 [==============================] - 41s 52ms/step - loss: 0.2495 - accuracy: 0.9115 - val_loss: 0.2465 - val_accuracy: 0.9265\n",
      "Epoch 18/20\n",
      "793/793 [==============================] - 41s 52ms/step - loss: 0.2411 - accuracy: 0.9140 - val_loss: 0.2313 - val_accuracy: 0.9218\n",
      "Epoch 19/20\n",
      "793/793 [==============================] - 41s 52ms/step - loss: 0.2396 - accuracy: 0.9152 - val_loss: 0.2449 - val_accuracy: 0.9157\n",
      "Epoch 20/20\n",
      "793/793 [==============================] - 41s 52ms/step - loss: 0.2339 - accuracy: 0.9172 - val_loss: 0.2138 - val_accuracy: 0.9273\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "tm.build(model_name= 'regularized')\n",
    "tm.summary()\n",
    "early_stopping = EarlyStopping(monitor='val_loss', patience=5, restore_best_weights=True)\n",
    "\n",
    "history = tm.train(ds_train_batch, ds_test_batch, epochs, optimizer, loss, metrics, early_stopping=early_stopping)\n",
    "tm.plot(history)\n",
    "tm.save('models/second_fixed.h5')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "kf_4XGKMpB7X"
   },
   "source": [
    "Training the <span style=\"color:tomato\">*third model*</span> again:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "With 20 epochs:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "mnWtP5XypB7X",
    "outputId": "dfe70648-1cee-4709-a2fb-2e82757a001e"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " conv2d (Conv2D)             (None, 30, 30, 96)        34944     \n",
      "                                                                 \n",
      " batch_normalization (BatchN  (None, 30, 30, 96)       384       \n",
      " ormalization)                                                   \n",
      "                                                                 \n",
      " max_pooling2d (MaxPooling2D  (None, 14, 14, 96)       0         \n",
      " )                                                               \n",
      "                                                                 \n",
      " conv2d_1 (Conv2D)           (None, 14, 14, 256)       614656    \n",
      "                                                                 \n",
      " batch_normalization_1 (Batc  (None, 14, 14, 256)      1024      \n",
      " hNormalization)                                                 \n",
      "                                                                 \n",
      " max_pooling2d_1 (MaxPooling  (None, 6, 6, 256)        0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " conv2d_2 (Conv2D)           (None, 6, 6, 384)         885120    \n",
      "                                                                 \n",
      " conv2d_3 (Conv2D)           (None, 6, 6, 384)         1327488   \n",
      "                                                                 \n",
      " conv2d_4 (Conv2D)           (None, 6, 6, 256)         884992    \n",
      "                                                                 \n",
      " batch_normalization_2 (Batc  (None, 6, 6, 256)        1024      \n",
      " hNormalization)                                                 \n",
      "                                                                 \n",
      " max_pooling2d_2 (MaxPooling  (None, 2, 2, 256)        0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " flatten (Flatten)           (None, 1024)              0         \n",
      "                                                                 \n",
      " dense (Dense)               (None, 512)               524800    \n",
      "                                                                 \n",
      " dropout (Dropout)           (None, 512)               0         \n",
      "                                                                 \n",
      " dense_1 (Dense)             (None, 512)               262656    \n",
      "                                                                 \n",
      " dropout_1 (Dropout)         (None, 512)               0         \n",
      "                                                                 \n",
      " dense_2 (Dense)             (None, 1)                 513       \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 4,537,601\n",
      "Trainable params: 4,536,385\n",
      "Non-trainable params: 1,216\n",
      "_________________________________________________________________\n",
      "Epoch 1/20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-03-15 05:29:11.382114: I tensorflow/stream_executor/cuda/cuda_dnn.cc:366] Loaded cuDNN version 8100\n",
      "2023-03-15 05:29:11.818657: I tensorflow/core/platform/default/subprocess.cc:304] Start cannot spawn child process: No such file or directory\n",
      "2023-03-15 05:29:11.819160: I tensorflow/core/platform/default/subprocess.cc:304] Start cannot spawn child process: No such file or directory\n",
      "2023-03-15 05:29:11.819168: W tensorflow/stream_executor/gpu/asm_compiler.cc:80] Couldn't get ptxas version string: INTERNAL: Couldn't invoke ptxas --version\n",
      "2023-03-15 05:29:11.819602: I tensorflow/core/platform/default/subprocess.cc:304] Start cannot spawn child process: No such file or directory\n",
      "2023-03-15 05:29:11.819632: W tensorflow/stream_executor/gpu/redzone_allocator.cc:314] INTERNAL: Failed to launch ptxas\n",
      "Relying on driver to perform ptx compilation. \n",
      "Modify $PATH to customize ptxas location.\n",
      "This message will be only logged once.\n",
      "2023-03-15 05:29:12.461219: I tensorflow/stream_executor/cuda/cuda_blas.cc:1774] TensorFloat-32 will be used for the matrix multiplication. This will only be logged once.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "793/793 [==============================] - 14s 15ms/step - loss: 0.4403 - accuracy: 0.7999 - val_loss: 0.3565 - val_accuracy: 0.8523\n",
      "Epoch 2/20\n",
      "793/793 [==============================] - 12s 15ms/step - loss: 0.3188 - accuracy: 0.8744 - val_loss: 0.3460 - val_accuracy: 0.8856\n",
      "Epoch 3/20\n",
      "793/793 [==============================] - 12s 15ms/step - loss: 0.2990 - accuracy: 0.8839 - val_loss: 0.2894 - val_accuracy: 0.8965\n",
      "Epoch 4/20\n",
      "793/793 [==============================] - 12s 15ms/step - loss: 0.2878 - accuracy: 0.8913 - val_loss: 0.3049 - val_accuracy: 0.8948\n",
      "Epoch 5/20\n",
      "793/793 [==============================] - 12s 15ms/step - loss: 0.2768 - accuracy: 0.8952 - val_loss: 0.3394 - val_accuracy: 0.9014\n",
      "Epoch 6/20\n",
      "793/793 [==============================] - 12s 15ms/step - loss: 0.2710 - accuracy: 0.8987 - val_loss: 0.3088 - val_accuracy: 0.9079\n",
      "Epoch 7/20\n",
      "793/793 [==============================] - 12s 15ms/step - loss: 0.2605 - accuracy: 0.9037 - val_loss: 0.2889 - val_accuracy: 0.9066\n",
      "Epoch 8/20\n",
      "793/793 [==============================] - 12s 15ms/step - loss: 0.2557 - accuracy: 0.9059 - val_loss: 0.2972 - val_accuracy: 0.9052\n",
      "Epoch 9/20\n",
      "793/793 [==============================] - 12s 15ms/step - loss: 0.2470 - accuracy: 0.9098 - val_loss: 0.2801 - val_accuracy: 0.9136\n",
      "Epoch 10/20\n",
      "793/793 [==============================] - 12s 16ms/step - loss: 0.2409 - accuracy: 0.9124 - val_loss: 0.2833 - val_accuracy: 0.9168\n",
      "Epoch 11/20\n",
      "793/793 [==============================] - 12s 15ms/step - loss: 0.2323 - accuracy: 0.9152 - val_loss: 0.2794 - val_accuracy: 0.9208\n",
      "Epoch 12/20\n",
      "793/793 [==============================] - 12s 15ms/step - loss: 0.2246 - accuracy: 0.9189 - val_loss: 0.2756 - val_accuracy: 0.9233\n",
      "Epoch 13/20\n",
      "793/793 [==============================] - 12s 15ms/step - loss: 0.2189 - accuracy: 0.9201 - val_loss: 0.2558 - val_accuracy: 0.9263\n",
      "Epoch 14/20\n",
      "793/793 [==============================] - 12s 16ms/step - loss: 0.2116 - accuracy: 0.9221 - val_loss: 0.2145 - val_accuracy: 0.9343\n",
      "Epoch 15/20\n",
      "793/793 [==============================] - 13s 16ms/step - loss: 0.2013 - accuracy: 0.9255 - val_loss: 0.2091 - val_accuracy: 0.9296\n",
      "Epoch 16/20\n",
      "793/793 [==============================] - 12s 15ms/step - loss: 0.1930 - accuracy: 0.9273 - val_loss: 0.2249 - val_accuracy: 0.9338\n",
      "Epoch 17/20\n",
      "793/793 [==============================] - 12s 16ms/step - loss: 0.1827 - accuracy: 0.9296 - val_loss: 0.2086 - val_accuracy: 0.9421\n",
      "Epoch 18/20\n",
      "793/793 [==============================] - 13s 16ms/step - loss: 0.1706 - accuracy: 0.9316 - val_loss: 0.2178 - val_accuracy: 0.9375\n",
      "Epoch 19/20\n",
      "793/793 [==============================] - 12s 16ms/step - loss: 0.1634 - accuracy: 0.9328 - val_loss: 0.1615 - val_accuracy: 0.9414\n",
      "Epoch 20/20\n",
      "793/793 [==============================] - 12s 16ms/step - loss: 0.1521 - accuracy: 0.9357 - val_loss: 0.1936 - val_accuracy: 0.9441\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "tm.build(model_name= 'alexnet')\n",
    "tm.summary()\n",
    "\n",
    "early_stopping = EarlyStopping(monitor='val_loss', patience=5, restore_best_weights=True)\n",
    "history = tm.train(ds_train_batch, ds_test_batch, 20, optimizer, loss, metrics, early_stopping=early_stopping)\n",
    "tm.plot(history)\n",
    "tm.save('models/third_fixed.h5')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "mjwXCEUGpB7X"
   },
   "source": [
    "With 50 epochs:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "3hG_K92bpB7Y",
    "outputId": "d8a371cd-68ac-4954-ba38-91e79f1546f0"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " conv2d (Conv2D)             (None, 30, 30, 96)        34944     \n",
      "                                                                 \n",
      " batch_normalization (BatchN  (None, 30, 30, 96)       384       \n",
      " ormalization)                                                   \n",
      "                                                                 \n",
      " max_pooling2d (MaxPooling2D  (None, 14, 14, 96)       0         \n",
      " )                                                               \n",
      "                                                                 \n",
      " conv2d_1 (Conv2D)           (None, 14, 14, 256)       614656    \n",
      "                                                                 \n",
      " batch_normalization_1 (Batc  (None, 14, 14, 256)      1024      \n",
      " hNormalization)                                                 \n",
      "                                                                 \n",
      " max_pooling2d_1 (MaxPooling  (None, 6, 6, 256)        0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " conv2d_2 (Conv2D)           (None, 6, 6, 384)         885120    \n",
      "                                                                 \n",
      " conv2d_3 (Conv2D)           (None, 6, 6, 384)         1327488   \n",
      "                                                                 \n",
      " conv2d_4 (Conv2D)           (None, 6, 6, 256)         884992    \n",
      "                                                                 \n",
      " batch_normalization_2 (Batc  (None, 6, 6, 256)        1024      \n",
      " hNormalization)                                                 \n",
      "                                                                 \n",
      " max_pooling2d_2 (MaxPooling  (None, 2, 2, 256)        0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " flatten (Flatten)           (None, 1024)              0         \n",
      "                                                                 \n",
      " dense (Dense)               (None, 512)               524800    \n",
      "                                                                 \n",
      " dropout (Dropout)           (None, 512)               0         \n",
      "                                                                 \n",
      " dense_1 (Dense)             (None, 512)               262656    \n",
      "                                                                 \n",
      " dropout_1 (Dropout)         (None, 512)               0         \n",
      "                                                                 \n",
      " dense_2 (Dense)             (None, 1)                 513       \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 4,537,601\n",
      "Trainable params: 4,536,385\n",
      "Non-trainable params: 1,216\n",
      "_________________________________________________________________\n",
      "Epoch 1/50\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-03-15 20:37:37.973450: I tensorflow/stream_executor/cuda/cuda_dnn.cc:366] Loaded cuDNN version 8100\n",
      "2023-03-15 20:37:38.451568: I tensorflow/core/platform/default/subprocess.cc:304] Start cannot spawn child process: No such file or directory\n",
      "2023-03-15 20:37:38.451955: I tensorflow/core/platform/default/subprocess.cc:304] Start cannot spawn child process: No such file or directory\n",
      "2023-03-15 20:37:38.451964: W tensorflow/stream_executor/gpu/asm_compiler.cc:80] Couldn't get ptxas version string: INTERNAL: Couldn't invoke ptxas --version\n",
      "2023-03-15 20:37:38.452290: I tensorflow/core/platform/default/subprocess.cc:304] Start cannot spawn child process: No such file or directory\n",
      "2023-03-15 20:37:38.452317: W tensorflow/stream_executor/gpu/redzone_allocator.cc:314] INTERNAL: Failed to launch ptxas\n",
      "Relying on driver to perform ptx compilation. \n",
      "Modify $PATH to customize ptxas location.\n",
      "This message will be only logged once.\n",
      "2023-03-15 20:37:39.030907: I tensorflow/stream_executor/cuda/cuda_blas.cc:1774] TensorFloat-32 will be used for the matrix multiplication. This will only be logged once.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "793/793 [==============================] - 13s 14ms/step - loss: 0.4398 - accuracy: 0.7992 - val_loss: 0.3942 - val_accuracy: 0.8147\n",
      "Epoch 2/50\n",
      "793/793 [==============================] - 10s 13ms/step - loss: 0.3213 - accuracy: 0.8728 - val_loss: 0.3808 - val_accuracy: 0.8641\n",
      "Epoch 3/50\n",
      "793/793 [==============================] - 11s 13ms/step - loss: 0.2972 - accuracy: 0.8855 - val_loss: 0.3207 - val_accuracy: 0.8943\n",
      "Epoch 4/50\n",
      "793/793 [==============================] - 11s 13ms/step - loss: 0.2852 - accuracy: 0.8906 - val_loss: 0.3375 - val_accuracy: 0.8653\n",
      "Epoch 5/50\n",
      "793/793 [==============================] - 11s 13ms/step - loss: 0.2776 - accuracy: 0.8941 - val_loss: 0.2926 - val_accuracy: 0.9023\n",
      "Epoch 6/50\n",
      "793/793 [==============================] - 11s 13ms/step - loss: 0.2684 - accuracy: 0.8997 - val_loss: 0.2700 - val_accuracy: 0.9086\n",
      "Epoch 7/50\n",
      "793/793 [==============================] - 11s 13ms/step - loss: 0.2618 - accuracy: 0.9020 - val_loss: 0.2730 - val_accuracy: 0.9110\n",
      "Epoch 8/50\n",
      "793/793 [==============================] - 11s 13ms/step - loss: 0.2552 - accuracy: 0.9061 - val_loss: 0.2848 - val_accuracy: 0.9094\n",
      "Epoch 9/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.2482 - accuracy: 0.9086 - val_loss: 0.2540 - val_accuracy: 0.9193\n",
      "Epoch 10/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.2392 - accuracy: 0.9130 - val_loss: 0.2558 - val_accuracy: 0.9190\n",
      "Epoch 11/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.2332 - accuracy: 0.9148 - val_loss: 0.2440 - val_accuracy: 0.9227\n",
      "Epoch 12/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.2256 - accuracy: 0.9181 - val_loss: 0.2335 - val_accuracy: 0.9124\n",
      "Epoch 13/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.2206 - accuracy: 0.9193 - val_loss: 0.2703 - val_accuracy: 0.9116\n",
      "Epoch 14/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.2091 - accuracy: 0.9218 - val_loss: 0.2180 - val_accuracy: 0.9287\n",
      "Epoch 15/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.1993 - accuracy: 0.9243 - val_loss: 0.2330 - val_accuracy: 0.9163\n",
      "Epoch 16/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.1914 - accuracy: 0.9263 - val_loss: 0.2493 - val_accuracy: 0.9249\n",
      "Epoch 17/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.1824 - accuracy: 0.9274 - val_loss: 0.2522 - val_accuracy: 0.9364\n",
      "Epoch 18/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.1710 - accuracy: 0.9307 - val_loss: 0.1922 - val_accuracy: 0.9389\n",
      "Epoch 19/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.1599 - accuracy: 0.9321 - val_loss: 0.1781 - val_accuracy: 0.9393\n",
      "Epoch 20/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.1516 - accuracy: 0.9341 - val_loss: 0.1895 - val_accuracy: 0.9388\n",
      "Epoch 21/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.1445 - accuracy: 0.9358 - val_loss: 0.1437 - val_accuracy: 0.9432\n",
      "Epoch 22/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.1318 - accuracy: 0.9386 - val_loss: 0.1380 - val_accuracy: 0.9469\n",
      "Epoch 23/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.1278 - accuracy: 0.9396 - val_loss: 0.1318 - val_accuracy: 0.9484\n",
      "Epoch 24/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.1201 - accuracy: 0.9415 - val_loss: 0.1233 - val_accuracy: 0.9496\n",
      "Epoch 25/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.1159 - accuracy: 0.9430 - val_loss: 0.1210 - val_accuracy: 0.9476\n",
      "Epoch 26/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.1122 - accuracy: 0.9436 - val_loss: 0.1173 - val_accuracy: 0.9497\n",
      "Epoch 27/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.1089 - accuracy: 0.9442 - val_loss: 0.1114 - val_accuracy: 0.9514\n",
      "Epoch 28/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.1042 - accuracy: 0.9461 - val_loss: 0.1160 - val_accuracy: 0.9486\n",
      "Epoch 29/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.1026 - accuracy: 0.9463 - val_loss: 0.1149 - val_accuracy: 0.9496\n",
      "Epoch 30/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.1008 - accuracy: 0.9465 - val_loss: 0.1011 - val_accuracy: 0.9483\n",
      "Epoch 31/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.0983 - accuracy: 0.9469 - val_loss: 0.1065 - val_accuracy: 0.9499\n",
      "Epoch 32/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.0952 - accuracy: 0.9477 - val_loss: 0.1009 - val_accuracy: 0.9491\n",
      "Epoch 33/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.0937 - accuracy: 0.9471 - val_loss: 0.1022 - val_accuracy: 0.9487\n",
      "Epoch 34/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.0919 - accuracy: 0.9491 - val_loss: 0.0992 - val_accuracy: 0.9516\n",
      "Epoch 35/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.0908 - accuracy: 0.9483 - val_loss: 0.0893 - val_accuracy: 0.9542\n",
      "Epoch 36/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.0881 - accuracy: 0.9490 - val_loss: 0.1018 - val_accuracy: 0.9427\n",
      "Epoch 37/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.0880 - accuracy: 0.9489 - val_loss: 0.0964 - val_accuracy: 0.9477\n",
      "Epoch 38/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.0859 - accuracy: 0.9499 - val_loss: 0.1076 - val_accuracy: 0.9455\n",
      "Epoch 39/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.0855 - accuracy: 0.9501 - val_loss: 0.0952 - val_accuracy: 0.9521\n",
      "Epoch 40/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.0842 - accuracy: 0.9502 - val_loss: 0.0873 - val_accuracy: 0.9531\n",
      "Epoch 41/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.0826 - accuracy: 0.9510 - val_loss: 0.0813 - val_accuracy: 0.9544\n",
      "Epoch 42/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.0807 - accuracy: 0.9516 - val_loss: 0.0987 - val_accuracy: 0.9481\n",
      "Epoch 43/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.0833 - accuracy: 0.9503 - val_loss: 0.0879 - val_accuracy: 0.9514\n",
      "Epoch 44/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.0804 - accuracy: 0.9513 - val_loss: 0.0813 - val_accuracy: 0.9534\n",
      "Epoch 45/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.0809 - accuracy: 0.9506 - val_loss: 0.0913 - val_accuracy: 0.9504\n",
      "Epoch 46/50\n",
      "793/793 [==============================] - 11s 14ms/step - loss: 0.0807 - accuracy: 0.9506 - val_loss: 0.0832 - val_accuracy: 0.9540\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "tm.build(model_name= 'alexnet')\n",
    "tm.summary()\n",
    "early_stopping = EarlyStopping(monitor='val_loss', patience=5, restore_best_weights=True)\n",
    "history = tm.train(ds_train_batch, ds_test_batch, 50, keras.optimizers.Adam(), loss, metrics, early_stopping=early_stopping)\n",
    "\n",
    "tm.plot(history)\n",
    "tm.save('models/alex50.h5')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "By taking a look to the new trained models, it can be see that there is not a huge difference between them (in this case, I cannot train the second one for 50 epochs without crashing the jupyter kernel, but the accuracy on validation was around 94% on one try around the 47 epoch). The models were all saved for possible future use."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Conclussions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Two aproachs were followed during the present work trying to achieve a good accuracy for a model on the task of predicting gender of a face from an image. While on the second approach splitting the training set to get a new validation set the results were much more accurate, it seems that there is not a huge difference between the most simpler model consisting only on a combination of convolutional layers and max pooling ones, and the most complex one, trying to imitate the AlexNet architecture. Both of them achieve more than 95% of accuracy on validation. Taking a look to the graphics, more epochs seem to not be significative on improving the model accuracy as it seems to be on a convergence process. This fact of more simpler models performing better on this task is common in the two approaches followed during the work, with and without the larger validation set. Another variations of this models were also tested and results remained similar. However, maybe the future work can be focus on amplify the base data. A good starting point can be the second testing architecture with data generators, trying to have more data to fit the model to try to improve generalization capabilities.\n",
    "\n",
    "To summarize; at the very end it seems that the most simpler solution to the problem is, in this case,, the most effective to solve it. \n",
    "\n",
    "\n",
    "\n"
   ]
  }
 ],
 "metadata": {
  "accelerator": "TPU",
  "colab": {
   "provenance": []
  },
  "gpuClass": "standard",
  "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": "3.10.6"
  },
  "vscode": {
   "interpreter": {
    "hash": "30163434fcee4d7d3816babd7a4ea09001dfd0cad8518a6f1cf2fe607c555d48"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}