Gesture_recognition_final.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "FsCWJ6klvu_d",
        "papermill": {
          "duration": 0.030998,
          "end_time": "2020-12-30T11:02:47.928954",
          "exception": false,
          "start_time": "2020-12-30T11:02:47.897956",
          "status": "completed"
        },
        "tags": []
      },
      "source": [
        "# Gesture Recognition\n",
        "\n",
        "### Team members:\n",
        "\n",
        "*   Dhruv Ajmeri\n",
        "*   Umesh Pandey\n",
        "*   Mariam Abbas Zaidi\n",
        "\n",
        "**Problem Statement**\n",
        "\n",
        "Imagine you are working as a data scientist at a home electronics company which manufactures state of the art smart televisions. You want to develop a cool feature in the smart-TV that can recognise five different gestures performed by the user which will help users control the TV without using a remote. \n",
        "\n",
        "In this group project, you are going to build a 3D Conv model that will be able to predict the 5 gestures correctly. \n",
        "\n",
        "Each gesture corresponds to a specific command:\n",
        " \n",
        "| Gesture | Corresponding Action |\n",
        "| --- | --- | \n",
        "| Thumbs Up | Increase the volume. |\n",
        "| Thumbs Down | Decrease the volume. |\n",
        "| Left Swipe | 'Jump' backwards 10 seconds. |\n",
        "| Right Swipe | 'Jump' forward 10 seconds. |\n",
        "| Stop | Pause the movie. |\n",
        "\n",
        "\n",
        "\n",
        "### Objectives:\n",
        "1. **Generator**: Ensure the generator can process a batch of videos without encountering any errors. Perform necessary steps such as cropping, resizing, and normalization successfully.\n",
        "\n",
        "2. **Model**: Create a model that can be trained without any errors. Evaluate the model based on the total number of parameters (considering faster inference/prediction time) and the achieved accuracy. It is recommended to start training with a small amount of data and gradually proceed.\n",
        "\n",
        "3. **Write up**: Include a detailed procedure describing the process followed to select the final model. Begin with an explanation of the rationale behind choosing the base model. Then, discuss the reasons and metrics considered when modifying and experimenting to reach the final model."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "G3a6hHnsH-Iv",
        "papermill": {
          "duration": 0.910687,
          "end_time": "2020-12-30T11:02:48.869231",
          "exception": false,
          "start_time": "2020-12-30T11:02:47.958544",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [],
      "source": [
        "import numpy as np\n",
        "import os\n",
        "from skimage.transform import resize\n",
        "from imageio import imread\n",
        "import datetime\n",
        "import os\n",
        "import matplotlib.pyplot as plt\n",
        "import warnings\n",
        "warnings.filterwarnings('ignore')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8NaUH5pqH-Iw",
        "papermill": {
          "duration": 0.029764,
          "end_time": "2020-12-30T11:02:48.929746",
          "exception": false,
          "start_time": "2020-12-30T11:02:48.899982",
          "status": "completed"
        },
        "tags": []
      },
      "source": [
        "Setting up random seed to generate same results in every run."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "TR67p8LKH-Iw",
        "papermill": {
          "duration": 4.225466,
          "end_time": "2020-12-30T11:02:53.185961",
          "exception": false,
          "start_time": "2020-12-30T11:02:48.960495",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [],
      "source": [
        "np.random.seed(30)\n",
        "import random as rn\n",
        "rn.seed(30)\n",
        "from keras import backend as K\n",
        "import tensorflow as tf\n",
        "tf.random.set_seed(30)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "15br6waEH-Iw",
        "outputId": "5f30c11f-bf31-4e71-915a-ed51590047cd",
        "papermill": {
          "duration": 2.397957,
          "end_time": "2020-12-30T11:02:55.613996",
          "exception": false,
          "start_time": "2020-12-30T11:02:53.216039",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Found GPU at: /device:GPU:0\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "2023-06-05 23:16:52.544424: W tensorflow/core/common_runtime/gpu/gpu_bfc_allocator.cc:39] Overriding allow_growth setting because the TF_FORCE_GPU_ALLOW_GROWTH environment variable is set. Original config value was 0.\n",
            "2023-06-05 23:16:52.544489: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1525] Created device /device:GPU:0 with 14802 MB memory:  -> device: 0, name: Quadro RTX 5000, pci bus id: 0000:1d:00.0, compute capability: 7.5\n"
          ]
        }
      ],
      "source": [
        "device_name = tf.test.gpu_device_name()\n",
        "if \"GPU\" not in device_name:\n",
        "    print(\"GPU device not found\")\n",
        "else:\n",
        "    print('Found GPU at: {}'.format(device_name))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JxvYTOni3PBG",
        "outputId": "dcba0df5-2d33-460e-ebed-953117d7c10f"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "\"from google.colab import drive\\ndrive.mount('/content/drive')\""
            ]
          },
          "execution_count": 5,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "\"\"\"from google.colab import drive\n",
        "drive.mount('/content/drive')\"\"\""
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "C0BcT8H7H-Ix",
        "outputId": "a3d65f34-4c78-4399-a576-4ffa1c5acf54",
        "papermill": {
          "duration": 0.719285,
          "end_time": "2020-12-30T11:02:56.538002",
          "exception": false,
          "start_time": "2020-12-30T11:02:55.818717",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Mon Jun  5 23:16:52 2023       \n",
            "+-----------------------------------------------------------------------------+\n",
            "| NVIDIA-SMI 520.61.05    Driver Version: 520.61.05    CUDA Version: 11.8     |\n",
            "|-------------------------------+----------------------+----------------------+\n",
            "| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |\n",
            "| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |\n",
            "|                               |                      |               MIG M. |\n",
            "|===============================+======================+======================|\n",
            "|   0  Quadro RTX 5000     Off  | 00000000:1D:00.0 Off |                  Off |\n",
            "| 33%   41C    P2    47W / 230W |    259MiB / 16384MiB |      1%      Default |\n",
            "|                               |                      |                  N/A |\n",
            "+-------------------------------+----------------------+----------------------+\n",
            "                                                                               \n",
            "+-----------------------------------------------------------------------------+\n",
            "| Processes:                                                                  |\n",
            "|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |\n",
            "|        ID   ID                                                   Usage      |\n",
            "|=============================================================================|\n",
            "+-----------------------------------------------------------------------------+\n"
          ]
        }
      ],
      "source": [
        "## Checking the GPU configuration\n",
        "!nvidia-smi"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "B8GVODh2H-Iw",
        "papermill": {
          "duration": 0.029947,
          "end_time": "2020-12-30T11:02:55.675119",
          "exception": false,
          "start_time": "2020-12-30T11:02:55.645172",
          "status": "completed"
        },
        "tags": []
      },
      "source": [
        "\n",
        "Within this section, you extract the names of the folders designated for training and validation. Additionally, the `size_of_a_batch` is defined at this point. It's important to ensure that the batch size is optimized to fully utilize the GPU's capabilities. Consequently, you gradually increase the batch size until an error is encountered by the machine."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "ljtBmy75H-Ix",
        "papermill": {
          "duration": 0.082331,
          "end_time": "2020-12-30T11:02:55.787924",
          "exception": false,
          "start_time": "2020-12-30T11:02:55.705593",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [],
      "source": [
        "train_doc = np.random.permutation(open('datasets/Project_data/train.csv').readlines())\n",
        "val_doc = np.random.permutation(open('datasets/Project_data/val.csv').readlines())\n",
        "size_of_a_batch = 64"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DGY4lrhhvu_w",
        "papermill": {
          "duration": 0.030965,
          "end_time": "2020-12-30T11:02:56.600557",
          "exception": false,
          "start_time": "2020-12-30T11:02:56.569592",
          "status": "completed"
        },
        "tags": []
      },
      "source": [
        "## Generator\n",
        "This section holds significant importance within the code. The generator's overall structure has already been provided. In this portion, you will be performing preprocessing on the images, considering the presence of two different image dimensions, and generating a batch of video frames. You are encouraged to experiment with `img_idx`, `y`, `z`, and normalization techniques in order to achieve optimal accuracy."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "KhGrETufvu_y",
        "papermill": {
          "duration": 0.037747,
          "end_time": "2020-12-30T11:02:56.740923",
          "exception": false,
          "start_time": "2020-12-30T11:02:56.703176",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [],
      "source": [
        "def normalizeImage(image):\n",
        "    # applying normalization\n",
        "    return image/255.0"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "1xrANziuvu_x",
        "papermill": {
          "duration": 0.040091,
          "end_time": "2020-12-30T11:02:56.671717",
          "exception": false,
          "start_time": "2020-12-30T11:02:56.631626",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [],
      "source": [
        "def cropResize(image, y, z):\n",
        "    h, w = image.shape\n",
        "    \n",
        "    # if image is smaller then crop at center for 120x120\n",
        "    if w == 160:\n",
        "        image = image[:120, 20:140]\n",
        "    # resizing every image\n",
        "    return resize(image, (y,z))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "-qd1bxIZvu_y",
        "papermill": {
          "duration": 0.038431,
          "end_time": "2020-12-30T11:02:56.810160",
          "exception": false,
          "start_time": "2020-12-30T11:02:56.771729",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [],
      "source": [
        "def preprocess_Image(image, y, z):\n",
        "    return normalizeImage(cropResize(image, y, z))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "LNgZq2e-vu_z",
        "papermill": {
          "duration": 0.038917,
          "end_time": "2020-12-30T11:02:56.880323",
          "exception": false,
          "start_time": "2020-12-30T11:02:56.841406",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [],
      "source": [
        "def make3dFilter(x):\n",
        "    return tuple([x]*3)\n",
        "\n",
        "def make2dFilter(x):\n",
        "    return tuple([x]*2)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "eA6r1IMmvu_0",
        "papermill": {
          "duration": 0.041295,
          "end_time": "2020-12-30T11:02:57.030781",
          "exception": false,
          "start_time": "2020-12-30T11:02:56.989486",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [],
      "source": [
        "def generator(source_path, list_of_folders, size_of_a_batch, Image_tensor):\n",
        "    print( 'Source path = ', source_path, '; batch size =', size_of_a_batch)\n",
        "    while True:\n",
        "        t = np.random.permutation(list_of_folders)\n",
        "        num_batches = int(len(list_of_folders)/size_of_a_batch)\n",
        "        for batch in range(num_batches): # we iterate over the number of batches\n",
        "            yield getBatchData(source_path, t, batch, size_of_a_batch, Image_tensor)\n",
        "        \n",
        "        # Implement the code to handle the remaining data points that are not part of complete batches.\n",
        "        # checking for remaining batches\n",
        "        if len(list_of_folders)%size_of_a_batch != 0:\n",
        "            # updating the batch yield and size \n",
        "            size_of_a_batch = len(list_of_folders)%size_of_a_batch\n",
        "            yield getBatchData(source_path, t, batch, size_of_a_batch, Image_tensor)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "-yPBgXECvu_z",
        "papermill": {
          "duration": 0.046129,
          "end_time": "2020-12-30T11:02:56.958354",
          "exception": false,
          "start_time": "2020-12-30T11:02:56.912225",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [],
      "source": [
        "def getBatchData(source_path, t, batch, size_of_a_batch, Image_tensor):\n",
        "    [x,y,z] = [len(Image_tensor[0]),Image_tensor[1], Image_tensor[2]]\n",
        "    img_idx = Image_tensor[0]\n",
        "    Batch_data = np.zeros((size_of_a_batch,x,y,z,3)) # For each video, you utilize 'x' images, where (y, z) represents the desired final size of the input images. Additionally, the number '3' signifies the RGB channels.\n",
        "    batch_label = np.zeros((size_of_a_batch,5)) # The variable 'batch_label' corresponds to the one-hot representation of the output.\n",
        "    for folder in range(size_of_a_batch): # iterating over the size_of_a_batch\n",
        "        imgs = os.listdir(source_path+'/'+ t[folder + (batch*size_of_a_batch)].split(';')[0]) # read all the images in the folder\n",
        "        for idx,item in enumerate(img_idx): #  Iterate iver the frames/images of a folder to read them in\n",
        "            image = imread(source_path+'/'+ t[folder + (batch*size_of_a_batch)].strip().split(';')[0]+'/'+imgs[item]).astype(np.float32)\n",
        "\n",
        "            # A separate function called 'preprocess_Image' is defined to handle the tasks of cropping, resizing, and normalizing images.\n",
        "            Batch_data[folder,idx,:,:,0] = preprocess_Image(image[:, :, 0], y, z)\n",
        "            Batch_data[folder,idx,:,:,1] = preprocess_Image(image[:, :, 1], y, z)\n",
        "            Batch_data[folder,idx,:,:,2] = preprocess_Image(image[:, :, 2], y, z)\n",
        "\n",
        "        batch_label[folder, int(t[folder + (batch*size_of_a_batch)].strip().split(';')[2])] = 1\n",
        "    return Batch_data, batch_label"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "tzNpErbFvu_1",
        "papermill": {
          "duration": 0.031438,
          "end_time": "2020-12-30T11:02:57.093854",
          "exception": false,
          "start_time": "2020-12-30T11:02:57.062416",
          "status": "completed"
        },
        "tags": []
      },
      "source": [
        "When designing the model architecture, it is crucial to consider that a video is represented within the generator as (number of images, height, width, number of channels). Please keep this representation in mind while creating the model's architecture."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "CiGThqNVvu_2",
        "outputId": "adca4a29-0f76-4a06-b9fe-54010abe3453",
        "papermill": {
          "duration": 0.042645,
          "end_time": "2020-12-30T11:02:57.168049",
          "exception": false,
          "start_time": "2020-12-30T11:02:57.125404",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "# training sequences = 663\n",
            "# validation sequences = 100\n",
            "# epochs = 10\n"
          ]
        }
      ],
      "source": [
        "curr_dt_time = datetime.datetime.now()\n",
        "train_path = 'datasets/Project_data/train'\n",
        "val_path = 'datasets/Project_data/val'\n",
        "num_of_train_seq = len(train_doc)\n",
        "print('# training sequences =', num_of_train_seq)\n",
        "num_of_value_seq = len(val_doc)\n",
        "print('# validation sequences =', num_of_value_seq)\n",
        "no_of_epochs = 10\n",
        "print ('# epochs =', no_of_epochs)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "w2bDaqWXvu_4",
        "outputId": "9870f821-4d23-4a59-db67-92ddb1ba795f",
        "papermill": {
          "duration": 0.045644,
          "end_time": "2020-12-30T11:02:57.245648",
          "exception": false,
          "start_time": "2020-12-30T11:02:57.200004",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "# Image_tensor = [array([ 0,  2,  3,  5,  6,  8,  9, 11, 12, 14, 15, 17, 18, 20, 21, 23, 24,\n",
            "       26, 27, 29]), 100, 100, 3]\n"
          ]
        }
      ],
      "source": [
        "def GetImageTensor(no_of_frames):\n",
        "    img_idx = np.round(np.linspace(0, 29, no_of_frames)).astype(int)\n",
        "    return [img_idx, 100, 100, 3]\n",
        "\n",
        "# define image tensor size\n",
        "Image_tensor = GetImageTensor(20)\n",
        "print ('# Image_tensor =', Image_tensor)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TEzkB8pzvu_5",
        "papermill": {
          "duration": 0.032074,
          "end_time": "2020-12-30T11:02:57.310202",
          "exception": false,
          "start_time": "2020-12-30T11:02:57.278128",
          "status": "completed"
        },
        "tags": []
      },
      "source": [
        "### Generator Validation"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "PKXoahGbvu_7",
        "outputId": "0808e197-f67b-4d29-ce40-da94a7500240",
        "papermill": {
          "duration": 40.636083,
          "end_time": "2020-12-30T11:03:37.978298",
          "exception": false,
          "start_time": "2020-12-30T11:02:57.342215",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Source path =  datasets/Project_data/train ; batch size = 64\n",
            "(64, 20, 100, 100, 3)\n"
          ]
        }
      ],
      "source": [
        "# check complete batch shape\n",
        "sample_generator = generator(train_path, train_doc, size_of_a_batch, Image_tensor)\n",
        "Batch_data_sample, sample_batch_label = next(sample_generator)\n",
        "print(Batch_data_sample.shape)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "k_7-dpHkHbwW",
        "outputId": "75998350-9176-43c8-e567-7ca06955070c"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Source path =  datasets/Project_data/val ; batch size = 64\n",
            "(64, 20, 100, 100, 3)\n"
          ]
        }
      ],
      "source": [
        "# validation batch sample\n",
        "sample_val_generator = generator(val_path, val_doc, size_of_a_batch, Image_tensor)\n",
        "sample_val_Batch_data, sample_val_batch_label = next(sample_val_generator)\n",
        "print(sample_val_Batch_data.shape)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 290
        },
        "id": "Xl6e7d0Ovu_8",
        "outputId": "e4f8a2b8-352c-4251-d5e6-f942988922af",
        "papermill": {
          "duration": 0.308784,
          "end_time": "2020-12-30T11:03:38.322256",
          "exception": false,
          "start_time": "2020-12-30T11:03:38.013472",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ],
      "source": [
        "# plot generated sample images\n",
        "fig, ax = plt.subplots(1,2)\n",
        "ax[0].imshow(sample_val_Batch_data[16,10,:,:,:])   \n",
        "ax[1].imshow(sample_val_Batch_data[25,10,:,:,:])\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "k3HvPgOPvu_-",
        "papermill": {
          "duration": 0.064517,
          "end_time": "2020-12-30T11:03:38.423183",
          "exception": false,
          "start_time": "2020-12-30T11:03:38.358666",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [],
      "source": [
        "def plotModelHistory(h):\n",
        "    fig, ax = plt.subplots(1, 2, figsize=(15,4))\n",
        "    ax[0].plot(h.history['loss'])   \n",
        "    ax[0].plot(h.history['val_loss'])\n",
        "    ax[0].legend(['loss','val_loss'])\n",
        "    ax[0].title.set_text(\"Train loss vs Validation loss\")\n",
        "\n",
        "    ax[1].plot(h.history['categorical_accuracy'])   \n",
        "    ax[1].plot(h.history['val_categorical_accuracy'])\n",
        "    ax[1].legend(['categorical_accuracy','val_categorical_accuracy'])\n",
        "    ax[1].title.set_text(\" Training accuracy vs Validation accuracy\")\n",
        "    plt.show()\n",
        "    \n",
        "    print(\"Max. Training Accuracy\", max(h.history['categorical_accuracy']))\n",
        "    print(\"Max. Validaiton Accuracy\", max(h.history['val_categorical_accuracy']))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dcGAuU41vu__",
        "papermill": {
          "duration": 0.061965,
          "end_time": "2020-12-30T11:03:38.551365",
          "exception": false,
          "start_time": "2020-12-30T11:03:38.489400",
          "status": "completed"
        },
        "tags": []
      },
      "source": [
        "## Model Building\n",
        "In this section, you construct the model utilizing various functionalities provided by Keras. It is essential to utilize `Conv3D` and `MaxPooling3D`, rather than `Conv2D` and `MaxPooling2D`, when building a 3D convolution model. If you are building a Conv2D + RNN model, you should incorporate `TimeDistributed`. Additionally, keep in mind that the last layer should be a softmax layer. Design the network with the objective of achieving high accuracy while utilizing the fewest number of parameters possible, enabling it to fit within the memory constraints of the webcam."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "BlVvi0S9vvAA",
        "papermill": {
          "duration": 0.073376,
          "end_time": "2020-12-30T11:03:38.687892",
          "exception": false,
          "start_time": "2020-12-30T11:03:38.614516",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [],
      "source": [
        "from keras.models import Sequential, Model\n",
        "from keras.callbacks import ModelCheckpoint, ReduceLROnPlateau\n",
        "from keras import optimizers\n",
        "from keras.layers import Dense, GRU, Flatten, TimeDistributed, Flatten, BatchNormalization, Activation, Dropout, LSTM\n",
        "from keras.layers.convolutional import Conv3D, MaxPooling3D, Conv2D, MaxPooling2D\n",
        "from tensorflow.keras.optimizers import Adam"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9A0Y13yhvvAA",
        "papermill": {
          "duration": 0.058747,
          "end_time": "2020-12-30T11:03:38.805775",
          "exception": false,
          "start_time": "2020-12-30T11:03:38.747028",
          "status": "completed"
        },
        "tags": []
      },
      "source": [
        "### Base Model"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "jD0uh84WvvAB",
        "outputId": "41d9e13d-9d78-45c6-8a6a-b91a1b8904d5",
        "papermill": {
          "duration": 0.987539,
          "end_time": "2020-12-30T11:03:39.867110",
          "exception": false,
          "start_time": "2020-12-30T11:03:38.879571",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " conv3d (Conv3D)             (None, 16, 96, 96, 16)    6016      \n",
            "                                                                 \n",
            " max_pooling3d (MaxPooling3D  (None, 8, 48, 48, 16)    0         \n",
            " )                                                               \n",
            "                                                                 \n",
            " batch_normalization (BatchN  (None, 8, 48, 48, 16)    64        \n",
            " ormalization)                                                   \n",
            "                                                                 \n",
            " conv3d_1 (Conv3D)           (None, 6, 46, 46, 32)     13856     \n",
            "                                                                 \n",
            " max_pooling3d_1 (MaxPooling  (None, 6, 23, 23, 32)    0         \n",
            " 3D)                                                             \n",
            "                                                                 \n",
            " batch_normalization_1 (Batc  (None, 6, 23, 23, 32)    128       \n",
            " hNormalization)                                                 \n",
            "                                                                 \n",
            " conv3d_2 (Conv3D)           (None, 4, 21, 21, 64)     55360     \n",
            "                                                                 \n",
            " max_pooling3d_2 (MaxPooling  (None, 4, 11, 11, 64)    0         \n",
            " 3D)                                                             \n",
            "                                                                 \n",
            " batch_normalization_2 (Batc  (None, 4, 11, 11, 64)    256       \n",
            " hNormalization)                                                 \n",
            "                                                                 \n",
            " flatten (Flatten)           (None, 30976)             0         \n",
            "                                                                 \n",
            " dense (Dense)               (None, 128)               3965056   \n",
            "                                                                 \n",
            " batch_normalization_3 (Batc  (None, 128)              512       \n",
            " hNormalization)                                                 \n",
            "                                                                 \n",
            " dropout (Dropout)           (None, 128)               0         \n",
            "                                                                 \n",
            " dense_1 (Dense)             (None, 64)                8256      \n",
            "                                                                 \n",
            " batch_normalization_4 (Batc  (None, 64)               256       \n",
            " hNormalization)                                                 \n",
            "                                                                 \n",
            " dropout_1 (Dropout)         (None, 64)                0         \n",
            "                                                                 \n",
            " dense_2 (Dense)             (None, 5)                 325       \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 4,050,085\n",
            "Trainable params: 4,049,477\n",
            "Non-trainable params: 608\n",
            "_________________________________________________________________\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "2023-06-05 23:17:14.191882: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1525] Created device /job:localhost/replica:0/task:0/device:GPU:0 with 14802 MB memory:  -> device: 0, name: Quadro RTX 5000, pci bus id: 0000:1d:00.0, compute capability: 7.5\n"
          ]
        }
      ],
      "source": [
        "def defineModel(Image_tensor):\n",
        "    inputShape = (len(Image_tensor[0]), Image_tensor[1], Image_tensor[2], Image_tensor[3])\n",
        "\n",
        "    model = Sequential([\n",
        "        Conv3D(16, make3dFilter(5), activation='relu', input_shape=inputShape),\n",
        "        MaxPooling3D(make3dFilter(2), padding='same'),\n",
        "        BatchNormalization(),\n",
        "\n",
        "        Conv3D(32, make3dFilter(3), activation='relu'),\n",
        "        MaxPooling3D(pool_size=(1,2,2), padding='same'),\n",
        "        BatchNormalization(),\n",
        "\n",
        "        Conv3D(64, make3dFilter(3), activation='relu'),\n",
        "        MaxPooling3D(pool_size=(1,2,2), padding='same'),\n",
        "        BatchNormalization(),\n",
        "\n",
        "        Flatten(),\n",
        "        Dense(128, activation='relu'),\n",
        "        BatchNormalization(),\n",
        "        Dropout(0.25),\n",
        "\n",
        "        Dense(64, activation='relu'),\n",
        "        BatchNormalization(),\n",
        "        Dropout(0.25),\n",
        "\n",
        "        Dense(5, activation='softmax')\n",
        "    ])\n",
        "    model.compile(optimizer=tf.keras.optimizers.Adam(), loss='categorical_crossentropy', metrics=['categorical_accuracy'])\n",
        "    return model\n",
        "\n",
        "model = defineModel(Image_tensor)\n",
        "model.summary()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "1SFrztx4vvAE",
        "papermill": {
          "duration": 0.081805,
          "end_time": "2020-12-30T11:03:40.017795",
          "exception": false,
          "start_time": "2020-12-30T11:03:39.935990",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [],
      "source": [
        "training_generator = generator(train_path, train_doc, size_of_a_batch, Image_tensor)\n",
        "val_generator = generator(val_path, val_doc, size_of_a_batch, Image_tensor)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TshGKBpyvvAF",
        "papermill": {
          "duration": 0.037982,
          "end_time": "2020-12-30T11:03:40.093995",
          "exception": false,
          "start_time": "2020-12-30T11:03:40.056013",
          "status": "completed"
        },
        "tags": []
      },
      "source": [
        "The 'steps_per_epoch' and 'validation_steps' parameters are utilized by the 'fit_generator' function to determine the number of 'next()' calls it needs to make. These values determine how many iterations or batches of data will be processed during each epoch for training and validation, respectively."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "hNW5OuAsvvAF",
        "papermill": {
          "duration": 0.047948,
          "end_time": "2020-12-30T11:03:40.179966",
          "exception": false,
          "start_time": "2020-12-30T11:03:40.132018",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [],
      "source": [
        "if (num_of_train_seq%size_of_a_batch) == 0:\n",
        "    steps_per_epoch = int(num_of_train_seq/size_of_a_batch)\n",
        "else:\n",
        "    steps_per_epoch = (num_of_train_seq//size_of_a_batch) + 1\n",
        "\n",
        "if (num_of_value_seq%size_of_a_batch) == 0:\n",
        "    validation_steps = int(num_of_value_seq/size_of_a_batch)\n",
        "else:\n",
        "    validation_steps = (num_of_value_seq//size_of_a_batch) + 1"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "YnIglXSuRxNs",
        "papermill": {
          "duration": 0.037752,
          "end_time": "2020-12-30T11:03:40.255666",
          "exception": false,
          "start_time": "2020-12-30T11:03:40.217914",
          "status": "completed"
        },
        "tags": []
      },
      "source": [
        "### Model Callbacks Setup"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "hWzVNhUhRqpL",
        "outputId": "613bd290-f966-4834-9f1c-d3e7dd2c9c3e",
        "papermill": {
          "duration": 0.048466,
          "end_time": "2020-12-30T11:03:40.343097",
          "exception": false,
          "start_time": "2020-12-30T11:03:40.294631",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:`period` argument is deprecated. Please use `save_freq` to specify the frequency in number of batches seen.\n"
          ]
        }
      ],
      "source": [
        "model_name = 'model_init' + '_' + str(curr_dt_time).replace(' ','').replace(':','_') + '/'\n",
        "    \n",
        "if not os.path.exists(model_name):\n",
        "    os.mkdir(model_name)\n",
        "\n",
        "filepath = model_name + 'model-{epoch:05d}-{loss:.5f}-{categorical_accuracy:.5f}-{val_loss:.5f}-{val_categorical_accuracy:.5f}.h5'\n",
        "\n",
        "checkpoint = ModelCheckpoint(filepath, monitor='val_loss', verbose=1, save_best_only=False, save_weights_only=False, mode='auto', period=1)\n",
        "\n",
        "LR = ReduceLROnPlateau(monitor='val_loss', factor=0.2, verbose=1, patience=4)\n",
        "\n",
        "# callbacks_list = [checkpoint, LR]\n",
        "callbacks_list = [LR]"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-hem7-SrSZme",
        "papermill": {
          "duration": 0.037862,
          "end_time": "2020-12-30T11:03:40.419628",
          "exception": false,
          "start_time": "2020-12-30T11:03:40.381766",
          "status": "completed"
        },
        "tags": []
      },
      "source": [
        "### Test Batch Size & Frames"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "ez2ESV-vHPKA",
        "papermill": {
          "duration": 0.050449,
          "end_time": "2020-12-30T11:03:40.508453",
          "exception": false,
          "start_time": "2020-12-30T11:03:40.458004",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [],
      "source": [
        "def testBatchSize(size_of_a_batch, no_of_frames):\n",
        "    Image_tensor = GetImageTensor(no_of_frames)\n",
        "    print(\"Using frames as\", Image_tensor[0])\n",
        "    global callbacks_list\n",
        "    no_of_epochs = 3\n",
        "    training_generator = generator(train_path, train_doc, size_of_a_batch, Image_tensor)\n",
        "    val_generator = generator(val_path, val_doc, size_of_a_batch, Image_tensor)\n",
        "    if (num_of_train_seq%size_of_a_batch) == 0:\n",
        "        steps_per_epoch = int(num_of_train_seq/size_of_a_batch)\n",
        "    else:\n",
        "        steps_per_epoch = (num_of_train_seq//size_of_a_batch) + 1\n",
        "\n",
        "    if (num_of_value_seq%size_of_a_batch) == 0:\n",
        "        validation_steps = int(num_of_value_seq/size_of_a_batch)\n",
        "    else:\n",
        "        validation_steps = (num_of_value_seq//size_of_a_batch) + 1\n",
        "\n",
        "    model = defineModel(Image_tensor)\n",
        "    model.fit_generator(training_generator, steps_per_epoch=steps_per_epoch, epochs=no_of_epochs, verbose=1, \n",
        "                callbacks=callbacks_list, validation_data=val_generator, \n",
        "                validation_steps=validation_steps, class_weight=None, workers=1, initial_epoch=0)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "vkbb9IUFH-Iz",
        "papermill": {
          "duration": 0.044889,
          "end_time": "2020-12-30T11:03:40.593141",
          "exception": false,
          "start_time": "2020-12-30T11:03:40.548252",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [],
      "source": [
        "# testBatchSize(size_of_a_batch=64, no_of_frames=30)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "0_Ss-pLcH-Iz",
        "papermill": {
          "duration": 0.045619,
          "end_time": "2020-12-30T11:03:40.677942",
          "exception": false,
          "start_time": "2020-12-30T11:03:40.632323",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [],
      "source": [
        "# testBatchSize(size_of_a_batch=64, no_of_frames=16)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "zh6jkT82H-Iz",
        "papermill": {
          "duration": 0.046181,
          "end_time": "2020-12-30T11:03:40.763182",
          "exception": false,
          "start_time": "2020-12-30T11:03:40.717001",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [],
      "source": [
        "# testBatchSize(size_of_a_batch=32, no_of_frames=30)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "0ybqy5yYH-Iz",
        "papermill": {
          "duration": 0.044218,
          "end_time": "2020-12-30T11:03:40.846457",
          "exception": false,
          "start_time": "2020-12-30T11:03:40.802239",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [],
      "source": [
        "# testBatchSize(size_of_a_batch=32, no_of_frames=16)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Vh1IZvwiajxA",
        "papermill": {
          "duration": 0.03831,
          "end_time": "2020-12-30T11:03:40.923787",
          "exception": false,
          "start_time": "2020-12-30T11:03:40.885477",
          "status": "completed"
        },
        "tags": []
      },
      "source": [
        "Based on the aforementioned observations, it is apparent that the batch size does not significantly impact the training time; rather, it is the number of frames that has a greater impact. Considering this, a batch size of 64 has been determined to be optimal. However, it is commented out in the code since running it consumes a substantial amount of time, which may hinder the notebook from completing its execution."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TRxSBVQgbPHo",
        "papermill": {
          "duration": 0.038838,
          "end_time": "2020-12-30T11:03:41.003654",
          "exception": false,
          "start_time": "2020-12-30T11:03:40.964816",
          "status": "completed"
        },
        "tags": []
      },
      "source": [
        "### Model 1 - Frames-16, Epoc-20"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "e9Z5A_HNZgyP",
        "outputId": "ca0b34b0-4f42-4d88-97b1-0643f3165e3c",
        "papermill": {
          "duration": 797.799786,
          "end_time": "2020-12-30T11:16:58.842002",
          "exception": false,
          "start_time": "2020-12-30T11:03:41.042216",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Model: \"conv_3d1\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " conv3d_3 (Conv3D)           (None, 12, 96, 96, 16)    6016      \n",
            "                                                                 \n",
            " max_pooling3d_3 (MaxPooling  (None, 6, 48, 48, 16)    0         \n",
            " 3D)                                                             \n",
            "                                                                 \n",
            " batch_normalization_5 (Batc  (None, 6, 48, 48, 16)    64        \n",
            " hNormalization)                                                 \n",
            "                                                                 \n",
            " conv3d_4 (Conv3D)           (None, 4, 46, 46, 32)     13856     \n",
            "                                                                 \n",
            " max_pooling3d_4 (MaxPooling  (None, 4, 23, 23, 32)    0         \n",
            " 3D)                                                             \n",
            "                                                                 \n",
            " batch_normalization_6 (Batc  (None, 4, 23, 23, 32)    128       \n",
            " hNormalization)                                                 \n",
            "                                                                 \n",
            " conv3d_5 (Conv3D)           (None, 2, 21, 21, 64)     55360     \n",
            "                                                                 \n",
            " max_pooling3d_5 (MaxPooling  (None, 2, 11, 11, 64)    0         \n",
            " 3D)                                                             \n",
            "                                                                 \n",
            " batch_normalization_7 (Batc  (None, 2, 11, 11, 64)    256       \n",
            " hNormalization)                                                 \n",
            "                                                                 \n",
            " flatten_1 (Flatten)         (None, 15488)             0         \n",
            "                                                                 \n",
            " dense_3 (Dense)             (None, 128)               1982592   \n",
            "                                                                 \n",
            " batch_normalization_8 (Batc  (None, 128)              512       \n",
            " hNormalization)                                                 \n",
            "                                                                 \n",
            " dropout_2 (Dropout)         (None, 128)               0         \n",
            "                                                                 \n",
            " dense_4 (Dense)             (None, 64)                8256      \n",
            "                                                                 \n",
            " batch_normalization_9 (Batc  (None, 64)               256       \n",
            " hNormalization)                                                 \n",
            "                                                                 \n",
            " dropout_3 (Dropout)         (None, 64)                0         \n",
            "                                                                 \n",
            " dense_5 (Dense)             (None, 5)                 325       \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 2,067,621\n",
            "Trainable params: 2,067,013\n",
            "Non-trainable params: 608\n",
            "_________________________________________________________________\n",
            "None\n",
            "Source path =  datasets/Project_data/train ; batch size = 64\n",
            "Epoch 1/20\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "2023-06-05 23:17:24.806264: I tensorflow/stream_executor/cuda/cuda_dnn.cc:377] Loaded cuDNN version 8302\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "11/11 [==============================] - ETA: 0s - loss: 1.4840 - categorical_accuracy: 0.4434Source path =  datasets/Project_data/val ; batch size = 64\n",
            "11/11 [==============================] - 102s 10s/step - loss: 1.4840 - categorical_accuracy: 0.4434 - val_loss: 4.3440 - val_categorical_accuracy: 0.2400 - lr: 0.0010\n",
            "Epoch 2/20\n",
            "11/11 [==============================] - 44s 4s/step - loss: 0.9455 - categorical_accuracy: 0.6443 - val_loss: 4.9210 - val_categorical_accuracy: 0.2188 - lr: 0.0010\n",
            "Epoch 3/20\n",
            "11/11 [==============================] - 42s 4s/step - loss: 0.9136 - categorical_accuracy: 0.6877 - val_loss: 4.3319 - val_categorical_accuracy: 0.2321 - lr: 0.0010\n",
            "Epoch 4/20\n",
            "11/11 [==============================] - 38s 4s/step - loss: 0.8816 - categorical_accuracy: 0.6481 - val_loss: 3.6328 - val_categorical_accuracy: 0.1875 - lr: 0.0010\n",
            "Epoch 5/20\n",
            "11/11 [==============================] - 32s 3s/step - loss: 0.6669 - categorical_accuracy: 0.7368 - val_loss: 2.4764 - val_categorical_accuracy: 0.2188 - lr: 0.0010\n",
            "Epoch 6/20\n",
            "11/11 [==============================] - 27s 3s/step - loss: 0.7047 - categorical_accuracy: 0.7273 - val_loss: 3.1981 - val_categorical_accuracy: 0.2500 - lr: 0.0010\n",
            "Epoch 7/20\n",
            "11/11 [==============================] - 28s 3s/step - loss: 0.7207 - categorical_accuracy: 0.7488 - val_loss: 3.7933 - val_categorical_accuracy: 0.1250 - lr: 0.0010\n",
            "Epoch 8/20\n",
            "11/11 [==============================] - 24s 2s/step - loss: 0.4949 - categorical_accuracy: 0.8289 - val_loss: 3.9920 - val_categorical_accuracy: 0.0000e+00 - lr: 0.0010\n",
            "Epoch 9/20\n",
            "11/11 [==============================] - 24s 2s/step - loss: 0.7902 - categorical_accuracy: 0.7433 - val_loss: 2.4265 - val_categorical_accuracy: 0.2500 - lr: 0.0010\n",
            "Epoch 10/20\n",
            "11/11 [==============================] - 24s 2s/step - loss: 0.5106 - categorical_accuracy: 0.8128 - val_loss: 3.3760 - val_categorical_accuracy: 0.0000e+00 - lr: 0.0010\n",
            "Epoch 11/20\n",
            "11/11 [==============================] - 23s 2s/step - loss: 0.5083 - categorical_accuracy: 0.8075 - val_loss: 1.2423 - val_categorical_accuracy: 0.3750 - lr: 0.0010\n",
            "Epoch 12/20\n",
            "11/11 [==============================] - 25s 2s/step - loss: 0.4517 - categorical_accuracy: 0.8503 - val_loss: 1.4185 - val_categorical_accuracy: 0.5000 - lr: 0.0010\n",
            "Epoch 13/20\n",
            "11/11 [==============================] - 25s 2s/step - loss: 0.5265 - categorical_accuracy: 0.7914 - val_loss: 1.4370 - val_categorical_accuracy: 0.5000 - lr: 0.0010\n",
            "Epoch 14/20\n",
            "11/11 [==============================] - 25s 2s/step - loss: 0.4697 - categorical_accuracy: 0.8342 - val_loss: 1.4858 - val_categorical_accuracy: 0.6250 - lr: 0.0010\n",
            "Epoch 15/20\n",
            "11/11 [==============================] - 23s 2s/step - loss: 0.3822 - categorical_accuracy: 0.8717 - val_loss: 0.7827 - val_categorical_accuracy: 0.7500 - lr: 0.0010\n",
            "Epoch 16/20\n",
            "11/11 [==============================] - 24s 2s/step - loss: 0.4255 - categorical_accuracy: 0.8610 - val_loss: 2.9474 - val_categorical_accuracy: 0.2500 - lr: 0.0010\n",
            "Epoch 17/20\n",
            "11/11 [==============================] - 25s 2s/step - loss: 0.3956 - categorical_accuracy: 0.8717 - val_loss: 3.0403 - val_categorical_accuracy: 0.2500 - lr: 0.0010\n",
            "Epoch 18/20\n",
            "11/11 [==============================] - 23s 2s/step - loss: 0.4031 - categorical_accuracy: 0.8556 - val_loss: 2.3063 - val_categorical_accuracy: 0.5000 - lr: 0.0010\n",
            "Epoch 19/20\n",
            "11/11 [==============================] - 24s 2s/step - loss: 0.3240 - categorical_accuracy: 0.8717 - val_loss: 0.5437 - val_categorical_accuracy: 0.8750 - lr: 0.0010\n",
            "Epoch 20/20\n",
            "11/11 [==============================] - 25s 2s/step - loss: 0.3393 - categorical_accuracy: 0.8930 - val_loss: 1.2243 - val_categorical_accuracy: 0.6250 - lr: 0.0010\n"
          ]
        }
      ],
      "source": [
        "no_of_frames = 16\n",
        "no_of_epochs = 20\n",
        "size_of_a_batch = 64\n",
        "\n",
        "Image_tensor = GetImageTensor(no_of_frames)\n",
        "training_generator = generator(train_path, train_doc, size_of_a_batch, Image_tensor)\n",
        "val_generator = generator(val_path, val_doc, size_of_a_batch, Image_tensor)\n",
        "\n",
        "if (num_of_train_seq%size_of_a_batch) == 0:\n",
        "    steps_per_epoch = int(num_of_train_seq/size_of_a_batch)\n",
        "else:\n",
        "    steps_per_epoch = (num_of_train_seq//size_of_a_batch) + 1\n",
        "\n",
        "if (num_of_value_seq%size_of_a_batch) == 0:\n",
        "    validation_steps = int(num_of_value_seq/size_of_a_batch)\n",
        "else:\n",
        "    validation_steps = (num_of_value_seq//size_of_a_batch) + 1\n",
        "    \n",
        "inputShape = (len(Image_tensor[0]), Image_tensor[1], Image_tensor[2], Image_tensor[3])\n",
        "\n",
        "model_1 = Sequential([\n",
        "    Conv3D(16, make3dFilter(5), activation='relu', input_shape=inputShape),\n",
        "    MaxPooling3D(make3dFilter(2), padding='same'),\n",
        "    BatchNormalization(),\n",
        "\n",
        "    Conv3D(32, make3dFilter(3), activation='relu'),\n",
        "    MaxPooling3D(pool_size=(1,2,2), padding='same'),\n",
        "    BatchNormalization(),\n",
        "\n",
        "    Conv3D(64, make3dFilter(3), activation='relu'),\n",
        "    MaxPooling3D(pool_size=(1,2,2), padding='same'),\n",
        "    BatchNormalization(),\n",
        "\n",
        "    Flatten(),\n",
        "    Dense(128, activation='relu'),\n",
        "    BatchNormalization(),\n",
        "    Dropout(0.25),\n",
        "\n",
        "    Dense(64, activation='relu'),\n",
        "    BatchNormalization(),\n",
        "    Dropout(0.25),\n",
        "\n",
        "    Dense(5, activation='softmax')\n",
        "], name=\"conv_3d1\")\n",
        "model_1.compile(optimizer=tf.keras.optimizers.Adam(), loss='categorical_crossentropy', metrics=['categorical_accuracy'])\n",
        "print(model_1.summary())\n",
        "\n",
        "model_1_history = model_1.fit_generator(training_generator, steps_per_epoch=steps_per_epoch, epochs=no_of_epochs, verbose=1, \n",
        "            callbacks=callbacks_list, validation_data=val_generator, \n",
        "            validation_steps=validation_steps, class_weight=None, workers=1, initial_epoch=0)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 376
        },
        "id": "YfeBXCls1r7R",
        "outputId": "aa750d90-c66b-4487-e3e2-e4d4def994d4",
        "papermill": {
          "duration": 0.397617,
          "end_time": "2020-12-30T11:16:59.349239",
          "exception": false,
          "start_time": "2020-12-30T11:16:58.951622",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Max. Training Accuracy 0.893048107624054\n",
            "Max. Validaiton Accuracy 0.875\n"
          ]
        }
      ],
      "source": [
        "plotModelHistory(model_1_history)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dsjCzIj40suJ",
        "papermill": {
          "duration": 0.111361,
          "end_time": "2020-12-30T11:16:59.574028",
          "exception": false,
          "start_time": "2020-12-30T11:16:59.462667",
          "status": "completed"
        },
        "tags": []
      },
      "source": [
        "### Model 2: Frames-30, Epocs-30"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "I82f2MK3bNyZ",
        "outputId": "a212bc29-09fb-4b10-b435-f09aeaf424d7",
        "papermill": {
          "duration": 1957.682254,
          "end_time": "2020-12-30T11:49:37.368875",
          "exception": false,
          "start_time": "2020-12-30T11:16:59.686621",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Model: \"conv_3d2\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " conv3d_6 (Conv3D)           (None, 26, 96, 96, 16)    6016      \n",
            "                                                                 \n",
            " max_pooling3d_6 (MaxPooling  (None, 13, 48, 48, 16)   0         \n",
            " 3D)                                                             \n",
            "                                                                 \n",
            " batch_normalization_10 (Bat  (None, 13, 48, 48, 16)   64        \n",
            " chNormalization)                                                \n",
            "                                                                 \n",
            " conv3d_7 (Conv3D)           (None, 11, 46, 46, 32)    13856     \n",
            "                                                                 \n",
            " max_pooling3d_7 (MaxPooling  (None, 11, 23, 23, 32)   0         \n",
            " 3D)                                                             \n",
            "                                                                 \n",
            " batch_normalization_11 (Bat  (None, 11, 23, 23, 32)   128       \n",
            " chNormalization)                                                \n",
            "                                                                 \n",
            " conv3d_8 (Conv3D)           (None, 9, 21, 21, 64)     55360     \n",
            "                                                                 \n",
            " max_pooling3d_8 (MaxPooling  (None, 9, 11, 11, 64)    0         \n",
            " 3D)                                                             \n",
            "                                                                 \n",
            " batch_normalization_12 (Bat  (None, 9, 11, 11, 64)    256       \n",
            " chNormalization)                                                \n",
            "                                                                 \n",
            " flatten_2 (Flatten)         (None, 69696)             0         \n",
            "                                                                 \n",
            " dense_6 (Dense)             (None, 128)               8921216   \n",
            "                                                                 \n",
            " batch_normalization_13 (Bat  (None, 128)              512       \n",
            " chNormalization)                                                \n",
            "                                                                 \n",
            " dropout_4 (Dropout)         (None, 128)               0         \n",
            "                                                                 \n",
            " dense_7 (Dense)             (None, 64)                8256      \n",
            "                                                                 \n",
            " batch_normalization_14 (Bat  (None, 64)               256       \n",
            " chNormalization)                                                \n",
            "                                                                 \n",
            " dropout_5 (Dropout)         (None, 64)                0         \n",
            "                                                                 \n",
            " dense_8 (Dense)             (None, 5)                 325       \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 9,006,245\n",
            "Trainable params: 9,005,637\n",
            "Non-trainable params: 608\n",
            "_________________________________________________________________\n",
            "None\n",
            "Source path =  datasets/Project_data/train ; batch size = 64\n",
            "Epoch 1/30\n",
            "11/11 [==============================] - ETA: 0s - loss: 1.5127 - categorical_accuracy: 0.4600 Source path =  datasets/Project_data/val ; batch size = 64\n",
            "11/11 [==============================] - 184s 18s/step - loss: 1.5127 - categorical_accuracy: 0.4600 - val_loss: 6.7900 - val_categorical_accuracy: 0.2600 - lr: 0.0010\n",
            "Epoch 2/30\n",
            "11/11 [==============================] - 80s 8s/step - loss: 0.8185 - categorical_accuracy: 0.6759 - val_loss: 5.1806 - val_categorical_accuracy: 0.2656 - lr: 0.0010\n",
            "Epoch 3/30\n",
            "11/11 [==============================] - 73s 7s/step - loss: 1.0348 - categorical_accuracy: 0.6087 - val_loss: 4.4132 - val_categorical_accuracy: 0.3214 - lr: 0.0010\n",
            "Epoch 4/30\n",
            "11/11 [==============================] - 65s 7s/step - loss: 0.9281 - categorical_accuracy: 0.6481 - val_loss: 3.7406 - val_categorical_accuracy: 0.1875 - lr: 0.0010\n",
            "Epoch 5/30\n",
            "11/11 [==============================] - 58s 6s/step - loss: 0.7849 - categorical_accuracy: 0.7033 - val_loss: 4.2431 - val_categorical_accuracy: 0.2188 - lr: 0.0010\n",
            "Epoch 6/30\n",
            "11/11 [==============================] - 52s 5s/step - loss: 0.7598 - categorical_accuracy: 0.6842 - val_loss: 4.9313 - val_categorical_accuracy: 0.1250 - lr: 0.0010\n",
            "Epoch 7/30\n",
            "11/11 [==============================] - 53s 5s/step - loss: 0.6218 - categorical_accuracy: 0.7685 - val_loss: 1.6363 - val_categorical_accuracy: 0.3750 - lr: 0.0010\n",
            "Epoch 8/30\n",
            "11/11 [==============================] - 46s 5s/step - loss: 0.6344 - categorical_accuracy: 0.7807 - val_loss: 1.7523 - val_categorical_accuracy: 0.3750 - lr: 0.0010\n",
            "Epoch 9/30\n",
            "11/11 [==============================] - 48s 5s/step - loss: 0.8309 - categorical_accuracy: 0.6791 - val_loss: 0.9118 - val_categorical_accuracy: 0.5000 - lr: 0.0010\n",
            "Epoch 10/30\n",
            "11/11 [==============================] - 45s 4s/step - loss: 0.5945 - categorical_accuracy: 0.7754 - val_loss: 2.2479 - val_categorical_accuracy: 0.2500 - lr: 0.0010\n",
            "Epoch 11/30\n",
            "11/11 [==============================] - 47s 5s/step - loss: 0.6809 - categorical_accuracy: 0.7594 - val_loss: 2.2030 - val_categorical_accuracy: 0.2500 - lr: 0.0010\n",
            "Epoch 12/30\n",
            "11/11 [==============================] - 49s 5s/step - loss: 0.5040 - categorical_accuracy: 0.8128 - val_loss: 1.5934 - val_categorical_accuracy: 0.3750 - lr: 0.0010\n",
            "Epoch 13/30\n",
            "11/11 [==============================] - ETA: 0s - loss: 0.4699 - categorical_accuracy: 0.8396\n",
            "Epoch 00013: ReduceLROnPlateau reducing learning rate to 0.00020000000949949026.\n",
            "11/11 [==============================] - 45s 5s/step - loss: 0.4699 - categorical_accuracy: 0.8396 - val_loss: 1.0547 - val_categorical_accuracy: 0.6250 - lr: 0.0010\n",
            "Epoch 14/30\n",
            "11/11 [==============================] - 47s 5s/step - loss: 0.4699 - categorical_accuracy: 0.8289 - val_loss: 1.0000 - val_categorical_accuracy: 0.5000 - lr: 2.0000e-04\n",
            "Epoch 15/30\n",
            "11/11 [==============================] - 46s 4s/step - loss: 0.3191 - categorical_accuracy: 0.9091 - val_loss: 2.2756 - val_categorical_accuracy: 0.2500 - lr: 2.0000e-04\n",
            "Epoch 16/30\n",
            "11/11 [==============================] - 47s 5s/step - loss: 0.3038 - categorical_accuracy: 0.8877 - val_loss: 2.3775 - val_categorical_accuracy: 0.3750 - lr: 2.0000e-04\n",
            "Epoch 17/30\n",
            "11/11 [==============================] - ETA: 0s - loss: 0.3140 - categorical_accuracy: 0.9037\n",
            "Epoch 00017: ReduceLROnPlateau reducing learning rate to 4.0000001899898055e-05.\n",
            "11/11 [==============================] - 44s 4s/step - loss: 0.3140 - categorical_accuracy: 0.9037 - val_loss: 2.5241 - val_categorical_accuracy: 0.3750 - lr: 2.0000e-04\n",
            "Epoch 18/30\n",
            "11/11 [==============================] - 44s 4s/step - loss: 0.3606 - categorical_accuracy: 0.8663 - val_loss: 1.3459 - val_categorical_accuracy: 0.6250 - lr: 4.0000e-05\n",
            "Epoch 19/30\n",
            "11/11 [==============================] - 45s 4s/step - loss: 0.3592 - categorical_accuracy: 0.8556 - val_loss: 1.4734 - val_categorical_accuracy: 0.5000 - lr: 4.0000e-05\n",
            "Epoch 20/30\n",
            "11/11 [==============================] - 44s 4s/step - loss: 0.2410 - categorical_accuracy: 0.9358 - val_loss: 1.2564 - val_categorical_accuracy: 0.5000 - lr: 4.0000e-05\n",
            "Epoch 21/30\n",
            "11/11 [==============================] - 47s 5s/step - loss: 0.2870 - categorical_accuracy: 0.9144 - val_loss: 0.6558 - val_categorical_accuracy: 0.6250 - lr: 4.0000e-05\n",
            "Epoch 22/30\n",
            "11/11 [==============================] - 44s 4s/step - loss: 0.2711 - categorical_accuracy: 0.8984 - val_loss: 0.8578 - val_categorical_accuracy: 0.6250 - lr: 4.0000e-05\n",
            "Epoch 23/30\n",
            "11/11 [==============================] - 44s 4s/step - loss: 0.3176 - categorical_accuracy: 0.8717 - val_loss: 0.7893 - val_categorical_accuracy: 0.6250 - lr: 4.0000e-05\n",
            "Epoch 24/30\n",
            "11/11 [==============================] - 46s 4s/step - loss: 0.2679 - categorical_accuracy: 0.9091 - val_loss: 1.4249 - val_categorical_accuracy: 0.5000 - lr: 4.0000e-05\n",
            "Epoch 25/30\n",
            "11/11 [==============================] - ETA: 0s - loss: 0.2164 - categorical_accuracy: 0.9465\n",
            "Epoch 00025: ReduceLROnPlateau reducing learning rate to 8.000000525498762e-06.\n",
            "11/11 [==============================] - 47s 5s/step - loss: 0.2164 - categorical_accuracy: 0.9465 - val_loss: 0.7869 - val_categorical_accuracy: 0.7500 - lr: 4.0000e-05\n",
            "Epoch 26/30\n",
            "11/11 [==============================] - 46s 4s/step - loss: 0.2733 - categorical_accuracy: 0.9091 - val_loss: 1.0721 - val_categorical_accuracy: 0.3750 - lr: 8.0000e-06\n",
            "Epoch 27/30\n",
            "11/11 [==============================] - 43s 4s/step - loss: 0.2560 - categorical_accuracy: 0.9144 - val_loss: 0.3572 - val_categorical_accuracy: 0.8750 - lr: 8.0000e-06\n",
            "Epoch 28/30\n",
            "11/11 [==============================] - 47s 5s/step - loss: 0.1865 - categorical_accuracy: 0.9465 - val_loss: 1.9474 - val_categorical_accuracy: 0.2500 - lr: 8.0000e-06\n",
            "Epoch 29/30\n",
            "11/11 [==============================] - 42s 4s/step - loss: 0.2017 - categorical_accuracy: 0.9358 - val_loss: 0.9380 - val_categorical_accuracy: 0.6250 - lr: 8.0000e-06\n",
            "Epoch 30/30\n",
            "11/11 [==============================] - 48s 5s/step - loss: 0.2338 - categorical_accuracy: 0.9519 - val_loss: 0.9426 - val_categorical_accuracy: 0.6250 - lr: 8.0000e-06\n"
          ]
        }
      ],
      "source": [
        "no_of_frames = 30\n",
        "no_of_epochs = 30\n",
        "size_of_a_batch = 64\n",
        "\n",
        "Image_tensor = GetImageTensor(no_of_frames)\n",
        "training_generator = generator(train_path, train_doc, size_of_a_batch, Image_tensor)\n",
        "val_generator = generator(val_path, val_doc, size_of_a_batch, Image_tensor)\n",
        "\n",
        "if (num_of_train_seq%size_of_a_batch) == 0:\n",
        "    steps_per_epoch = int(num_of_train_seq/size_of_a_batch)\n",
        "else:\n",
        "    steps_per_epoch = (num_of_train_seq//size_of_a_batch) + 1\n",
        "\n",
        "if (num_of_value_seq%size_of_a_batch) == 0:\n",
        "    validation_steps = int(num_of_value_seq/size_of_a_batch)\n",
        "else:\n",
        "    validation_steps = (num_of_value_seq//size_of_a_batch) + 1\n",
        "    \n",
        "inputShape = (len(Image_tensor[0]), Image_tensor[1], Image_tensor[2], Image_tensor[3])\n",
        "\n",
        "model_2 = Sequential([\n",
        "    Conv3D(16, make3dFilter(5), activation='relu', input_shape=inputShape),\n",
        "    MaxPooling3D(make3dFilter(2), padding='same'),\n",
        "    BatchNormalization(),\n",
        "\n",
        "    Conv3D(32, make3dFilter(3), activation='relu'),\n",
        "    MaxPooling3D(pool_size=(1,2,2), padding='same'),\n",
        "    BatchNormalization(),\n",
        "\n",
        "    Conv3D(64, make3dFilter(3), activation='relu'),\n",
        "    MaxPooling3D(pool_size=(1,2,2), padding='same'),\n",
        "    BatchNormalization(),\n",
        "\n",
        "    Flatten(),\n",
        "    Dense(128, activation='relu'),\n",
        "    BatchNormalization(),\n",
        "    Dropout(0.25),\n",
        "\n",
        "    Dense(64, activation='relu'),\n",
        "    BatchNormalization(),\n",
        "    Dropout(0.25),\n",
        "\n",
        "    Dense(5, activation='softmax')\n",
        "], name=\"conv_3d2\")\n",
        "model_2.compile(optimizer=tf.keras.optimizers.Adam(), loss='categorical_crossentropy', metrics=['categorical_accuracy'])\n",
        "print(model_2.summary())\n",
        "\n",
        "model_2_history = model_2.fit_generator(training_generator, steps_per_epoch=steps_per_epoch, epochs=no_of_epochs, verbose=1, \n",
        "            callbacks=callbacks_list, validation_data=val_generator, \n",
        "            validation_steps=validation_steps, class_weight=None, workers=1, initial_epoch=0)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 378
        },
        "id": "dGXMP96h1zDT",
        "outputId": "e2ddde42-7393-4680-c1cc-442fa5e01b3f",
        "papermill": {
          "duration": 0.489963,
          "end_time": "2020-12-30T11:49:38.077303",
          "exception": false,
          "start_time": "2020-12-30T11:49:37.587340",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [
        {
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAA2EAAAEICAYAAADWaPyEAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8/fFQqAAAACXBIWXMAAAsTAAALEwEAmpwYAACYBElEQVR4nOzdd3iUVfbA8e9N7x1IQgKhdwgdpKigAoqiKGJBxbprRXd117W7q/uzr70XxN67giIiUqQEQu8QWhIIkArpub8/7kwYQspMMi3hfJ4nzyQz77zvnWQyM+c9556rtNYIIYQQQgghhHAPH08PQAghhBBCCCFOJhKECSGEEEIIIYQbSRAmhBBCCCGEEG4kQZgQQgghhBBCuJEEYUIIIYQQQgjhRhKECSGEEEIIIYQbSRAmnEYp9ZNS6qpG3jdDKXWGs8fkjZRS05VSC21+LlJKdbRn20Ycq9F/kwb2O1Mp9Yiz9yuEEE2llFqvlDrN2dsK76SU0kqpzpbvX1VK3W/Pto04zuVKqZ8bO04havLz9ACEZymlimx+DAFKgUrLz3/RWn9g77601hOcOTZvpZRqC+wCummtt9e47Stgu9b6Tnv3p7UOc9K4HgI6a62n2ez7pPibCCGaL6VUO2CDzVWhwFHAupDpBK31H/buT2vdyxXbCtdQSs0GlmmtH6hx/STgNSBJa11hz7601n910phSgJ2Av/XYls9Ddn8mEqIhkgk7yWmtw6xfwG7gXJvrql9slFISsFtorfcBvwJX2F6vlIoBzgbe9cS4hBCiOdJa767xXgTQz+a66gBM3ovs08x+T+8C05RSqsb1VwAf2BuAicZpZs+VFkWCMFErpdRpSqm9Sql/KqWygXeUUtFKqe+VUjlKqVzL90k295mvlLrO8v10pdRCpdRTlm13KqXsysoopQKVUs8qpTItX88qpQItt8VZjpunlDqslPpDKeVjue2fSql9SqlCpdRmpdTYWvY9VCmVrZTytbnuAqXUGsv3Q5RSK5RSBUqp/UqpZ+oY5rvUCMKAS4ANWuu1Sqm7lVLbLWPZoJS6oJ7Ha1tKEauU+tZy/GVApxrbPqeU2mO5PU0pNcpy/XjgHmCqpbxxteV627+Jj1LqPqXULqXUAaXULKVUpOW2FMs4rlJK7VZKHVRK3VvXmGt5DNcrpbZZ/ibfKqUSLdcrpdT/LMcrUEqtVUr1ttx2tuV3U2j5u9mdPRRCnBws7yWLLK8jh4CHlFKdlFLzlFKHLK9VHyilomzuU13erpR6SCn1qeX1rlCZ8sNBjdx2gFJqleW2z5RSn6g6yrLtGGOyUupLZd5PDymlXrS57Xql1Eab948BluuPK6VTNmXhqnHv2TFKqXeUeZ/NVUp9bbl+nVLqXJvt/C2PoX8tj3OjUmqizc9+luMNUEoFKaXetzy+PKXUcqVUm1p+XV8DscAom/1EAxOBWcq8Ly+x7CNLKfWiUiqgjt/7caXySqm7LPfJVEpdU2Pbcyx/zwJl3lcfsrl5geUyT5n31OHqxKkEp1geU77l8hSb2+Yrpf5jee4WKqV+VkrF1THmRv2dLLdNUkqlWx7DdmU+C5wwxcPy3H7f8r31/f5apdRuYJ7l+s+U+XyUr5RaoJTqZXP/YKXU08p8fshX5vNdsFLqB6XUrTUezxpVz2cecYwEYaI+8UAM0B64AfN8ecfyczugGHixznvDUGAzEAc8Abyl1AlnumpzLzAMSAX6AUOA+yy3/R3YC7QC2mACD62U6gbcAgzWWocD44CMmjvWWi8FjgBjbK6+DPjQ8v1zwHNa6whMAPRpHWP8CohTSo20ue4KjmXBtmPeUCKBh4H3lVIJDT90XgJKgATgGsuXreWY30uMZcyfKaWCtNazgf8Cn1jOHPerZd/TLV+nAx2BME78+40EugFjgQeUUj0aGrBSagzwf8DFlnHvAj623HwWMBroivldXAwcstz2FqbkNRzojeWNQAghahgK7MC85j8KKMxrTiLQA0gGHqrn/udhXpOigG+p/32r1m0tH/q/AmZiXn8/Aur7oFnnGJU5Cfg95rUyBWhrOSZKqSmW7a4EIizjOYR9HH3Pfg8zDaEX0Br4n+X6WcA0m+3OBrK01qtqOeZHwKU2P48DDmqtVwJXYV73kzFB1l8tYziO1roY8157pc3VFwObtNarMVMk7sB8lhiOeX+6qe5fg2EJSO4EzgS6ADXnnR+xHDMKOAe4USl1vuW20ZbLKMt76pIa+44BfgCetzy2Z4AflFKxNptdBlyN+d0GWMZSm0b9nZRSQzB/q7ssj2E0tXzuqcepmOfmOMvPP2F+T62BlRxfevkUMBA4BfMc+wdQhSWLad1IKdUP83z+wYFxnLy01vIlX2itwfzznmH5/jSgDAiqZ/tUINfm5/nAdZbvpwPbbG4LwdT3x9tx7O3A2Ta3jQMyLN//G/gGM/fJ9v6dgQOYF1n/Bh7nI8Dblu/DMS/E7S0/L8AETXF2/L7eBF63fN/F8vtqXce26cAkm9/NQpvbtGX8vkA50N3mtv/ablvLfnMxZTtg3rjfr3G77d/kV+Amm9u6WY7nh/kgoDG199bblwGX1HHcmcAjlu/fAp6wuS3Mst8UTLC7BRNU+9TYx27gL0CEp5/78iVf8uU9X9bXRMv304HdDWx/PrDK5mfb95OHgLk2t/UEih3dFvMBdx+gbG5faH0dtOMxVY8RE0jkAH61bDcHmNHQ78Xys+3r8Gk48J6NOWFWBUTXsl0iUGh9bQY+B/5Rxz47W7YNsfz8AfCA5ftrgMVAXzt+PyOBPOv4gUXAHXVsezvwVR3PF9vfydvAYzbbda35O6yx32eB/1m+T7Fs62dz+3Qs78eYk67Latx/CTDd8v184D6b224CZtv5XLH37/Sadby13JaB5Xlt89x+v8Zj61jPGKIs20RigsRiLJ81amwXhPkc0sXy81PAy/Y8TvnSkgkT9crRWpdYf1BKhSilXrOkowswAUuUsintqyHb+o3W+qjlW3uaUCRizhBa7bJcB/AksA34WSm1Qyl1t2X/2zAvzA8BB5RSHytLSVwtPgQmK1PiOBlYqbW2Hu9azAv1Jkt5wcQ69gHmDNAUpVQQ5gV5jtb6AIBS6kpLiUCeUioPk+mptRTBRitMQLSnxmOvppS601L+kW/Zb6Qd+7Wq7ffqhzm7bJVt8/1RGvH30loXYc7cttVaz8Oc0XsJ83d5XSkVYdn0QswZ1l1Kqd+VUsPtfBxCiJOL7WsiSqk2ltf4fZb3ovep/3Ww5utakKp7Hkxd2yYC+7Tlk2Zt43JgjMnALl37XKdkzInIxnDkPTsZOKy1zq25E611JiYIulCZEsoJ1NGQwvLeuxE4VykVgsncWStL3sMElR9bSumeUEr517GfhcBB4HylVCdMBcyHlsfR1VKil215HP/Fvve9ROp/Px2qlPrNUgaYj8nUNfb91Lr/tjY/2/V+2ti/E017roDN70Yp5auUesxS0ljAsYxanOUrqLZjWZ5vn2Dm9PlgsqLvNWFMJxUJwkR9dI2f/47JngzVplzPmq63p8TQEZmYtLxVO8t1aK0LtdZ/11p3xLzY/01Z5n5prT/UWo+03FcDj9e2c631BsyL5QSOL0VEa71Va30pJh3/OPC5Uiq0jnEuBA4DkzDp+HcBlFLtgTcw5ZGxWusoYB0N/55ygArMC6vtY8ey31GYEoCLMWfFooB8m/3W/HvVVNvvtQLY38D9GnLcfi2/r1jMWWO01s9rrQdizip3xZROoLVerrWehPldf03dpZ9CiJNbzde2/1qu62N5L5qG89+HasoC2tYoqU+ua2PqH+MeoF0dgeAeaswFtnEUU1ViFV/jdkfes/cAMcpmnloN1jKzKcASbRpS1cVakjgJMy96G4DWulxr/bDWuiemjG0ix5cc1jTLcvs0zElN63vTK8AmTLYlAjMNwZ6/dxZ1vJ9afIgpOU3WWkcCr9L491Pr/uv7PdWlsX+n+p4rR6j/uQLHP8bLMH+/MzAnd1NsxnAQM02irmO9C1yOKRM9qmuUboq6SRAmHBGOSUnnWeqhH3TRcT4C7lNKtVJmIusDmLOIKKUmKqU6W94I8zG14lVKqW5KqTGW7FaJZZxV9RzjQ2AG5sXuM+uVSqlpSqlWWusqTGkEde3HckZ0FiZYiwK+s9wUinlxy7Hs82pMJqxeWutK4EvMxPMQpVRPTE29VTgmaMoB/JRSD2DmDFjtB1IsZ6Nq8xFwh1Kqg1IqjGNzyJraeeoj4GqlVKrl9/9fYKnWOkMpNdhyttEf86ZQgvl7BSiz5kqk1rocKKD+v5cQQliFA0VAvjJLhtzlhmMuwbzf3KJM84lJmGxNY8a4DBMgPKaUClWmgcUIy21vAncqpQYqo7PlxB6YsvbLLFmL8Zg5PfWp8z1ba52FmQP0sjKNIfyVUqNt7vs1MADzPjmrgeN8jJn/eyM2JzWVUqcrpfpYMjoFmDL1+l7nZ2GCgOs5vstwuOX+RUqp7pbj2ONTYLpSqqclS1fzM0s4JstUosz8qstsbsuxjLXWNTyBH4GuSqnLLM+HqZgTjd/bObaa42jM3+ktzHvvWGUab7W1/H7APFcusWw/CLjIjjGUYqpYQjDv49YxVGFKO59RSiVann/DLe/3WIKuKuBpJAvmEAnChCOeBYIxZ0X+BGa76DiPACuANcBazARRa7ejLsBczJvbEkzt8W9AIPCYZWzZmOzKv+o5xkeYN7B5WuuDNtePB9Yrs37ac5g5USdMJLYxC3P26xOtdSlUZ9qetoxvP9AHU9phj1swJQvZmNr2d2xum4P5nW/BZPJKOL7UwhpMHlJKraxl329jXiAXYNY/KQFurWU7h2it5wL3A19gPlh0wnSKBBMkvoGpGd+FeYF/0nLbFUCGpfThr5gzaUII0ZCHMQFCPqYBwJeuPqDWugxTvn4t5gTdNMwH7lJHx2g54XYuZj7VbkyzqamW2z7DNB/5EDPX6mtMIwQwAdG5luNfbrmtPs9S/3v2FZjAaBNmTvXtNmMsxrymd6CB368lUFiCyXZ9YnNTPGY+WQGmZPF36vmQrrXOwMwhC8VkqKzuxARIhZj3k09OuHPt+/sJ8zuYh5nGULP5003Av5VShZiTvZ/a3Pco5u+wSJlpBcNq7PsQJrP3d8z72j+AiTU+T9jrWRrxd9JaL8M0/vgf5nn2O8eyc/dj3otzMc/FD6nfLMx79D7Mmn1/1rj9TsznseWYCqDHOT6GmIX5rPN+A8cRNtTx5c1CCCGEEKIhSqmlwKta63ca3LgZslRbdNVaT2twY3FSU0pdCdxgmRIi7CSZMCGEEEKIBiilTlVKxVvKz64C+uK6ihCPspTFXQu87umxCO9mKfW8CXmuOEyCMCGEEEKIhnUDVmPKAf8OXGQpxWtRlFLXY0rdf9JaL2hoe3HyUkqNw8yf20/DJY+iBilHFEIIIYQQQgg3kkyYEEIIIYQQQrhRXYsVNklcXJxOSUlxxa6FEEJ4kbS0tINa61aeHkdzIe+PQghx8qjvPbLBIEwp1Y3j24F2BB7QWj9b131SUlJYsWKFo+MUQgjRzCildnl6DM2JvD8KIcTJo773yAaDMK31ZiDVsiNfzBoCXzlrcEIIIYQQQghxMnF0TthYYLvWWs58CiGEEEIIIUQjOBqEXQJ8VNsNSqkblFIrlFIrcnJymj4yIYQQQgghhGiB7G7MoZQKAM4D/lXb7Vrr17Es1DZo0CDpey+EaBbKy8vZu3cvJSUlnh6KVwsKCiIpKQl/f39PD6XFkeeg8DT5/xbC/RzpjjgBWKm13u+qwQghhLvt3buX8PBwUlJSUEp5ejheSWvNoUOH2Lt3Lx06dPD0cFoceQ4KT5L/byE8w5FyxEupoxRRCCGaq5KSEmJjY+XDbz2UUsTGxkqmxkXkOSg8Sf6/hfAMu4IwpVQocCbwpWuHI4QQ7icffhsmvyPXkt+v8CR5/gnhfnYFYVrrI1rrWK11vqsHhNaw5GXY+L3LDyWEEEIIIYQQAKUVlfy6cT9PzN7k8mM5MifMPZSCFW9DTAfoMdHToxFCCJcLCwujqKjI08MQQggharX70FESo4Lw83W0sbrzVVVptucUkRQdQnCAb5P3d6S0gvmbc5i9PpvfNh2gqLSC8CA/rh7RgVbhgU4Yce28LwgDSBkJ676AqkrwafovVwghhBDOMX/+fAICAjjllFNcfqyzzz6bDz/8kKioKIfuN3PmTFasWMGLL77omoEJcZIoq6ji8dmbeGvhTganRPP8pf1JiAz22Hi2HSjkn1+sJW1XLr4+iu7x4aQmR5GaHEX/dlF0jAvDx6fh8tr8o+XM3bif2euzWbAlh9KKKmJDAzi3XwLjesVzSqc4AvxcG3B6bxCW9g5kr4HE/p4ejRBCuIXWmn/84x/89NNPKKW47777mDp1KllZWUydOpWCggIqKip45ZVXOOWUU7j22mtZsWIFSimuueYa7rjjDk8/BHESmD9/PmFhYS4NwrTWaK358ccfXXYMd7A+Dh8fz2cPhHDUnsNHueWjVazek8c5fRL4bfMBznl+Ic9c3I/TurV261jKKqp4Zf52XvptGyGBvtx3Tg9yj5aRviePb9Mz+WDpbgDCg/zolxRVHZj1S46qzmblFJby84ZsZq/LZsn2Q1RUaRIig7h0SDvG945ncEoMvnYEcM7inUFY+xHmMmOhBGFCCLd5+Lv1bMgscOo+eyZG8OC5veza9ssvvyQ9PZ3Vq1dz8OBBBg8ezOjRo/nwww8ZN24c9957L5WVlRw9epT09HT27dvHunXrAMjLy3PquIVnePI5OGvWLJ566imUUvTt25eLL76YRx55hLKyMmJjY/nggw8oLi7m1VdfxdfXl/fff58XXniB7t2789e//pXdu82HoGeffZYRI0aQk5PDZZddRmZmJsOHD+eXX34hLS2NuLg4nnnmGd5++20ArrvuOm6//XYyMjIYN24cQ4cOJS0tjR9//JFTTz2VFStWEBcXd8L43nvvPb777rsTxtimTZsGH2td9ysqKuLWW2+tPrnx4IMPcuGFFzJ79mzuueceKisriYuL49dff+Whhx4iLCyMO++8E4DevXvz/fdmPnvNx/HYY4+xfPlyiouLueiii3j44YcBWL58OTNmzODIkSMEBgby66+/cs455/D888+TmpoKwMiRI3nppZfo16+fw3974TpfrtzLzoNHuGRIO9pGeS4z5Co/r8/mzs9WozW8cvkAJvRJYHtOETd/sJLp7yznxtM68fczu7qlPHHV7lzu/mItm/cXcm6/RB48tydxYcfKBK3liel78qq/Xvl9O5VVZtnitlHBxIUHsmZvHlpDSmwI143qyPje8fRLivRYYxrvDMIiEiCmE2QsglNu9fRohBDCLRYuXMill16Kr68vbdq04dRTT2X58uUMHjyYa665hvLycs4//3xSU1Pp2LEjO3bs4NZbb+Wcc87hrLPO8vTwRTO2fv16HnnkERYvXkxcXByHDx9GKcWff/6JUoo333yTJ554gqeffpq//vWvxwUfl112GXfccQcjR45k9+7djBs3jo0bN/Lwww8zZswY/vWvfzF79mzeeustANLS0njnnXdYunQpWmuGDh3KqaeeSnR0NFu3buXdd99l2LBhDY4PTIBS2xgbUtf9/vOf/xAZGcnatWsByM3NJScnh+uvv54FCxbQoUOH6mPXp+bjePTRR4mJiaGyspKxY8eyZs0aunfvztSpU/nkk08YPHgwBQUFBAcHc+211zJz5kyeffZZtmzZQklJiQRgXsaUxK2hvFLz0m/bOKtnPNNHpDC0Q0yz7zRpW37Yp20kL102gHaxIQB0ahXG1zeP4OHvNvDK/O2syDjs0vLEI6UVPPXzZmYuziA+Ioi3rhrE2B4nnmTx8VF0aRNOlzbhTBmUDEBxWSXrMvNJ322CsuyCEmaM7cL43vF0axPuFX8n7wzCAFJGwIZvZF6YEMJt7M1Yudvo0aNZsGABP/zwA9OnT+dvf/sbV155JatXr2bOnDm8+uqrfPrpp9WZBdF8eeo5OG/ePKZMmUJcXBwAMTExrF27troctqysrM6FfOfOncuGDRuqfy4oKKCoqIiFCxfy1VdfATB+/Hiio6MBc7LhggsuIDQ0FIDJkyfzxx9/cN5559G+ffsTArC6xgdmoWt7xlhTXfebO3cuH3/8cfV20dHRfPfdd4wePbp6G+ux61PzcXz66ae8/vrrVFRUkJWVxYYNG1BKkZCQwODBgwGIiIgAYMqUKfznP//hySef5O2332b69Ol2PSbhHlpr7v1qHSEBfrx37RB+XJvNx8t3M3t9Nt3jw7l6RAqTUtsS5N/8Prvalh9OPyWFf53dnUC/4x9HkL8v/ze5D0M7xHDPV2tdVp74+5Yc7vlyLfvyirliWHv+Mb4b4UH+dt8/OMCXwSkxDE5p+P/VU7y3SLn9SCjJh/3rPD0SIYRwi1GjRvHJJ59QWVlJTk4OCxYsYMiQIezatYs2bdpw/fXXc91117Fy5UoOHjxIVVUVF154IY888ggrV6709PBFC3Prrbdyyy23sHbtWl577bU6F/Otqqrizz//JD09vbpMNiwsrFHHtAZmzh6js+5ny8/Pj6qqquqfbfdh+zh27tzJU089xa+//sqaNWs455xz6j1eSEgIZ555Jt988w2ffvopl19+ucNjE67zedpelu48zN0TutM3KYq7J3Rnyd1jeWxyHwD++cVahv3frzz20yb25RV7eLT2+3l9Nuc8/wc7DhTxyuUDeOi8XicEYLbO79+W724dSevwQKa/s5wnZm+iorKqzu3tlXukjL99ks5Vby8j0N+Hz/46nP+c39uhAKy58N4gLMU6L2yRZ8chhBBucsEFF9C3b1/69evHmDFjeOKJJ4iPj2f+/Pn069eP/v3788knnzBjxgz27dvHaaedRmpqKtOmTeP//u//PD180YyNGTOGzz77jEOHDgFw+PBh8vPzadu2LQDvvvtu9bbh4eEUFhZW/3zWWWfxwgsvVP+cnp4OwIgRI/j0008B+Pnnn8nNzQXMyYavv/6ao0ePcuTIEb766itGjRrl8PiAOsfYkLrud+aZZ/LSSy9V/5ybm8uwYcNYsGABO3fuPO7YKSkp1Sc/Vq5cWX17TQUFBYSGhhIZGcn+/fv56aefAOjWrRtZWVksX74cgMLCQioqKgAzT+62225j8ODB1RlE4XmHj5Tx3x83Mqh9NFMtZW9gsi6XDGnHTzNG8fENwxjWIZbXF2xn1OPz+Ot7afy54xBaaw+OvG5lFVX85/sN3PBeGu1jQ/n+tpFM6JNg132t5YmXDknm5fnbufSNP8nOd/yEBpgM47erMznjmd/5dnUmt47pzI+3jfLqTFZTeW85YmQSRKfArkUw/CZPj0YIIVzGukaYUoonn3ySJ5988rjbr7rqKq666qoT7ifZL+EsvXr14t577+XUU0/F19eX/v3789BDDzFlyhSio6MZM2ZMdZBx7rnnctFFF/HNN9/wwgsv8Pzzz3PzzTfTt29fKioqGD16NK+++ioPPvggl156Ke+99x7Dhw8nPj6e8PBwBgwYwPTp0xkyZAhgAo7+/fuTkZHh0PhmzpxZ5xgbUtf97rvvPm6++WZ69+6Nr68vDz74IJMnT+b1119n8uTJVFVV0bp1a3755RcuvPBCZs2aRa9evRg6dChdu3at9VjWEyjdu3cnOTmZESPMSeaAgAA++eQTbr31VoqLiwkODmbu3LmEhYUxcOBAIiIiuPrqq+39Ewo3+O+PGyksqeDRC/rU2gZdKcWwjrEM6xjL3tyjvP/n7uNKFS8d0o5xveKJjwzywOhPtDf3KLd8uIr0esoPG2LKE/sytEMs93y1lrOf/6PB8sSS8kr2F5SQnV9CtuVy8fZD/L4lh35Jkbx/3VB6JEQ09eF5PeWKyHzQoEF6xYoVTd/R1zfD5h/grh0g7V2FEC6wceNGevTo4elhNAu1/a6UUmla60EeGlKzU9v7Y0t9DpaWluLr64ufnx9LlizhxhtvrM6SifplZmZy2mmnsWnTJre1t2/q87CySrMhs4BeiRF2rdPU3Py54xCXvP4nN57WiX+O7273/YrLKvkmfR/vLtnFxizT+bR/uygm9I5nfK+E6qYX7vbLhv38/dN0tIYnLuprd/arPtbuiZuyC7lhdEc6twojqzrQKia7oJTs/GJyj5afcN+oEH9uOb0zV4/o4NY28a5W33uk92bCwKwXlv4+HNgA8b09PRohhBBC2Gn37t1cfPHFVFVVERAQwBtvvOHpITULs2bN4t577+WZZ55pVuuLPfPLZl76bTvtY0O4Ylh7pgxKJjK4ZczjKa2o5N6v1pIcE8xtY7o4dF9rqeIlQ9qx7UAhs9dlM3t9Nv/9cRP//XETPRIiTEDWO54urcNc3rWvrKKKJ2Zv4k1L98MXL+tP+1jH5mLW5Vj3xPW8vmBH9fVxYQHERwbRNiqIge2jiI8Iok1EEAmRwcRHBtImIqhFzvlqiJcHYZZ5YbsWSRAmhBBCNCNdunRh1apVHh3Do48+ymeffXbcdVOmTOHee+/10IgaduWVV3LllVd6ehgO2bK/kNd+38HIznGUlFfyyA8beeaXLVw4IImrTmlP59bhjd73tgOF/LTWBC5HSiv47+Q+nNIpzomjb9hrv+9ge84R3rl6MMEBje962Ll1OLeMCeeWMV3Yc/goc9abhYP/N3cLz/yyhY5xoYy3BGR92jp//SpnlB82xFqeeP2ojvj7+tAmIogAv+ZzMsGdvDsIi2pnvjL+gKF/8fRohBBCCNGM3HvvvV4dcLUEVVWae79aS1iQH89f2p+Y0ADW7ctn5uIMPlm+h/f+3MWoLnFMPyWF07u1brBUUWvNun0FzF6fxex12WzPOQLAgHZRFJcppr25lBlju3LLmM5uKVvbefAIL/62jXP6JHC6E9uwJ8eYBYOvG9WRAwUlzNmwnznrsnltwQ5enr+dtlHBXDI4mWtGdiA0sOkf123LD62LL7tSx1aN65B6MvHuIAxMq/qtc6CqSuaFCSGEEEJ4kc/S9rA8I5cnLuxLTGgAAL3bRvLUlH7cPaE7Hy/bzft/7ubad1fQPjaEK4enMGVQEhE25WdVVZq03bmmVG9dNvvyivH1UQzrGMP0U1I4q1c8bSKCOFJawX1fr+N/c7ewLOMQz07tT6vwQJc9Nq0193+9jkBfHx44t6fLjtM6IogrhrXnimHtyT1SxtyN+/luTRZP/7KFd5fsYsbYzlwypB3+vo5/DnZl+aFoGu8PwlJGwOoPIWcTtHHdP4AQQgghhLDfoaJS/u+nTQxJiWHKoKQTbo8LC+SWMV34y6mdmLM+m5mLMvjP9xt4+ufNXDggiZFd4liwJYefN+wnp7CUAF8fRnWJY8YZXTijR5vqoM4qNNCPZy7ux/COsdz/zTrOfv4Pnrsk1WXlid+uzmThtoP8e1Iv2kS4p6NhdGgAUwYlM2VQMmm7cnl89ibu/2Y9by7cyd/P6sbEPgl2Nz6xLT+8anh77jmnh9PLD0XjNYMgbKS53LVIgjAhhBBCCC/x6I8bOVJawaMX9K53/pK/rw8T+yYysW8ia/ceX6oYEuDL6d1aM653PKd3a9VggwalFBcPTqZvciQ3f7CSaW8u5fYzunLz6c4tT8w/Ws5/vt9Av6RILh/a3mn7dcTA9tF8csMw5m/O4fHZm7jto1W89vt2/jm+O6O6xNX7O7ctP3z58gGc7eLyQ+E47w/CotpDRBJkLIQh13t6NEII4VFhYWHV64rVlJGRwcSJE1m3bp2bRyWEONks3naQL1fu45bTO9Oljf2NN/okRfL0xf3419nd2ZJdyID20QT5O56d6R4fwbe3jOS+r9fxzC9bWLbzMP+bmuq08sTHZm/i8JEyZl49xKMt05VSnN69NaO7tuKb9H08/fMWrnx7Gad0iuWf47vTLznquO1tyw97t43gpcsGSPmhl/L+SVZKmZLEXYvAS1cbF0IIIU4WYWHOm3D/9ddfs2HDBqftrz6nnHJKo+730EMP8dRTTzl5NM1baUUl9329jvaxIdwypnOj9hEXFsgpneMaFYBZWcsTH7+wD8szDnP283+wZPuhRu/PKm3XYT5atptrRnSgd9vIJu/PGXx9FJMHJDHvzlN5YGJPNmUXMumlRdz0QRo7csyJub25R7n4tSW8uXAnVw1vzxc3ntJ8ArDyEk+PwO28PxMGpiRxzSdwcAu06ubp0QghWqqf7obstc7dZ3wfmPBYnTfffffdJCcnc/PNNwPmA5+fnx+//fYbubm5lJeX88gjjzBp0iSHDltSUsKNN97IihUr8PPz45lnnuH0009n/fr1XH311ZSVlVFVVcUXX3xBYmIiF198MXv37qWyspL777+fqVOnNulhC2GPr7/+mokTJ9Kzp+umG1RUVODn58fixYtddgx3sD4Ob/DK/O3sOHiEWdcMaVIQ5QxKKaYObke/5Chu/mAll7/5J3ec0ZWbGlmeWF5ZxT1friMxMog7zuzqghE3TaCfL9eM7MCUQUm8sWAHby7cyZz1+zm3bwK/bc6hqko3v/LDAxvh1ZHw14XQuuUtXF8X7/hvbkh7y3phGQslCBNCtChTp07l9ttvrw7CPv30U+bMmcNtt91GREQEBw8eZNiwYZx33nkOrRnz0ksvoZRi7dq1bNq0ibPOOostW7bw6quvMmPGDC6//HLKysqorKzkxx9/JDExkR9++AGA/Px8lzxWYYcWcCLg8ccf5/3338fHx4cJEybw2GOP8cYbb/D6669TVlZG586dee+990hPT+fbb7/l999/55FHHuGLL74A4OabbyYnJ4eQkBDeeOMNunfvzvbt27n88ss5cuQIkyZN4tlnn6WoqAitNf/4xz/46aefUEpx3333MXXqVObPn8/9999PdHQ0mzZtYsuWLceV8to7xpCQkAYfb133279/P3/961/ZscMsWvvKK69wyimnMGvWLJ566imUUvTt25f33nuP6dOnM3HiRC666CLgWNlxbY/j/PPPZ8+ePZSUlDBjxgxuuOEGAGbPns0999xDZWUlcXFx/PLLL3Tr1o3FixfTqlUrqqqq6Nq1K0uWLKFVq1Z2/S1rsyOniJd/2855/RIZ3bXx+3E22/LEp3/ZwtJGlie+tXAnm/cX8voVA53SGt5VwoP8+dtZ3bhieAovzNvKh0t30z0hvHmWHx7YAFUVJtkiQZiXiekI4QkmCBt8radHI4Roqer5oOoq/fv358CBA2RmZpKTk0N0dDTx8fHccccdLFiwAB8fH/bt28f+/fuJj4+3e78LFy7k1ltvBaB79+60b9+eLVu2MHz4cB599FH27t3L5MmT6dKlC3369OHvf/87//znP5k4cSKjRo1y1cMVXsiZJwJ++uknvvnmG5YuXUpISAiHDx8GYPLkyVx/vZnXfd999/HWW29x6623ct555x0XfIwdO5ZXX32VLl26sHTpUm666SbmzZvHjBkzmDFjBpdeeimvvvpq9fG+/PJL0tPTWb16NQcPHmTw4MGMHj0agJUrV7Ju3To6dOjQpDE2pK773XbbbZx66ql89dVXVFZWUlRUxPr163nkkUdYvHgxcXFx1ceuT83H8fbbbxMTE0NxcTGDBw/mwgsvpKqqiuuvv54FCxbQoUMHDh8+jI+PD9OmTeODDz7g9ttvZ+7cufTr169JAZjWmvu+Xkegvw/3TfS+D8vW8sRhHWN44Jv1jHhsHj0SI+ifHEWq5at9bEidz+M9h4/y7NwtnNmzDWf1sv/11pNahQfy70m9mTG2CxHB/o1qY+9xhdnm8mjTS0mbE7uCMKVUFPAm0BvQwDVa6yUuHFfNAZiSxJ0LzLwwJ68gLoQQnjRlyhQ+//xzsrOzmTp1Kh988AE5OTmkpaXh7+9PSkoKJSXOqZe/7LLLGDp0KD/88ANnn302r732GmPGjGHlypX8+OOP3HfffYwdO5YHHnjAKccTDmrmJwLmzp3L1VdfXZ1BiomJAWDdunXcd9995OXlUVRUxLhx4064b1FREYsXL2bKlCnV15WWlgKwZMkSvv76a8A8h++8807AnGy49NJL8fX1pU2bNpx66qksX76ciIgIhgwZckIA1tQx1qau+82bN49Zs2YB4OvrS2RkJLNmzWLKlCnExcUdd+z61Hwczz//PF999RUAe/bsYevWreTk5DB69Ojq7az7veaaa5g0aRK33347b7/9NldffbVdj6kuX6fvY/H2Qzxyfm9ah7unZbujrOWJA9pF8/nKvaTvzuPTFXuYuTgDgKgQf/olWYKydlGkJkURHRqA1poHvlmHj1I8fF4vzz6IRogNc916aS5nDcKOSBBWm+eA2Vrri5RSAUDD+Xlnaz8C1n4Gh7ZDXOMmgQohhDeaOnUq119/PQcPHuT333/n008/pXXr1vj7+/Pbb7+xa9cuh/c5atQoPvjgA8aMGcOWLVvYvXs33bp1Y8eOHXTs2JHbbruN3bt3s2bNGrp3705MTAzTpk0jKiqKN9980wWPUngzV58ImD59Ol9//TX9+vVj5syZzJ8//4RtqqqqiIqKIj09vfEPxEZoqGMlWfaM0Zn3s+Xn50dVVRVgfg9lZWXVt9k+jvnz5zN37lyWLFlCSEgIp512Wr1/l+TkZNq0acO8efNYtmwZH3zwgcNjs8o7WsYj32+kf7soLhvSrtH7cZcubcL51wSTrauorGLrgSLS9+SRvjuP9D15PL91a3W/t5TYEDrEhfLb5hzuO6cHiVHBHhz5SegkzYQ1mLNUSkUCo4G3ALTWZVrrPBeP60QplvKYjD/cfmghhHClXr16UVhYSNu2bUlISODyyy9nxYoV9OnTh1mzZtG9e3eH93nTTTdRVVVFnz59mDp1KjNnziQwMJBPP/2U3r17k5qayrp167jyyitZu3YtQ4YMITU1lYcffpj77rvPBY9SeLOpU6fy8ccf8/nnnzNlyhTy8/MbdSLgzDPP5J133uHo0aMA1eV2hYWFJCQkUF5eflwgEB4eTmFhIQARERF06NCBzz77DDClb6tXrwZg2LBh1XPGPv744+r7jxo1ik8++YTKykpycnJYsGABQ4YMceoYG1LX/caOHcsrr7wCQGVlJfn5+YwZM4bPPvuMQ4cOHXfslJQU0tLSAPj2228pLy+v9Vj5+flER0cTEhLCpk2b+PPPP6t/PwsWLGDnzp3H7RfguuuuY9q0aUyZMgVf38Y30Xjsp03kFZfz3wv62L1YsLfw8/WhR0IElw5px+MX9WXOHaNZ+9A4Prp+GP8c351u8eFsyCpgcEo0009J8fRwTz5F1iDsoGfH4Wb2ZMI6ADnAO0qpfkAaMENrfcR2I6XUDcANAO3aueAMSWwnCGtjWtUPalo6XQghvM3atceaMcTFxbFkSe0V33WtEQbmg5x1jbCgoCDeeeedE7a5++67ufvuu4+7bty4cXaXXomWqbYTAeeeey59+vRh0KBBdp8IGD9+POnp6QwaNIiAgADOPvts/vvf//Kf//yHoUOH0qpVK4YOHVodeF1yySVcf/31PP/883z++ed88MEH3HjjjTzyyCOUl5dzySWX0K9fP5599lmmTZvGo48+yvjx44mMNG3DL7jgApYsWUK/fv1QSvHEE08QHx/Ppk2bnDbGhtR1v+eee44bbriBt956C19fX1555RWGDx/Ovffey6mnnoqvry/9+/dn5syZXH/99UyaNIl+/foxfvz4OrN448eP59VXX6VHjx5069aNYcOGAdCqVStef/11Jk+eTFVVFa1bt+aXX34B4LzzzuPqq69uUini8ozDfLx8D38Z3ZEeCRGN3o83CQv0Y3inWIZ3ivX0UMRJmglTuoG1t5RSg4A/gRFa66VKqeeAAq31/XXdZ9CgQXrFihXOHSnAZ1fD7j/hbxtkXpgQwik2btxIjx7eN8HcG9X2u1JKpWmtB3loSM1Obe+P8hxs2NGjRwkODkYpxccff8xHH33EN9984+lhNQsrVqzgjjvu4I8/6q8kqut5WFZRxcQX/uBIaSW//G00IQHNo6ebaEb+rx2U5kNCP/jLAk+Pxqnqe4+05z9pL7BXa73U8vPnwN31bO86KSNh/ZdweIfJjAkhxElo7dq1XHHFFcddFxgYyNKlS+u4hxDNW1paGrfccgtaa6Kionj77bc9PaRm4bHHHuOVV17hgw8+4GhpBYePlhHg60NIgC/BAb74+jTcSe/NhTvYsr+It64aJAGYcL6yoyYAA2nMUZPWOlsptUcp1U1rvRkYC7hnefuaUkaay12LJAgTQjiN1tqhNbg8rU+fPk5rXmCvhqomxMnF3ScCRo0aVT0/zFNuvvlmFi1adNx1M2bMaHLHQVe6++67uf3vd7G/oIRtOUX4KEWVzf9ykL8vIf4mICuvrKKisgo/mxbnew4f5flftzK+Vzxje7TxxEMQLZ11PlhI3ElXjmjvKY1bgQ8snRF3AJ55xYnrCqGtzHphA670yBCEEC1LUFAQhw4dIjY2tlkFYu6ktebQoUMEBXlnS+qWQE4EeL+XXnrJ00NwSHlFFfsLS8g9UoZSijYRQcSFBaK1pri8kqNl5qugpJxDR0rZX1DC1Id+pk/bSNO6PTmKj5fvwVcpHjyvp6cfjmiprPPB2vSCnb9D2REIaGaLTTeSXUGY1jod8HzNv1KmVX3GIlkvTAjhFElJSezdu5ecnBxPD8WrBQUFkZSU5OlhtEhyIkA4U0VlFTlFpRwqKkNj1o9qFR5os4ivItzXh/Agf8C0xN+fc5CjoSFMHZzM6r15zFyUQVmlaZn/4Lk9SYiUlu3CRQqzzGWb3iYIO3pIgjCvlTISNnwNebsgOsXToxFCNHP+/v61LugqhLvIiQDhDFprikorKCypQGsIDvAlItiP/AIf8rPqv29QUBCDenZmeD8TmJVVVLExq4DMvGLO6lX/At1CNEnhfnPZxrJA9pGDEOX969A5Q/MMwsCUJEoQJoQQopmTEwGiKSoqq/g8bS/Pzt1KdkEJp3VrxT/GdadnYuNbyQf4+dAvOYp+yVHOG6gQtSnMAt9AiO1sfj56uP7tW5DmF4S16g4hsaYksf80T49GCCGEEMIjfl6fzeOzN7E95wipyVE8e0kqwzrKuleiGSnaD+HxEBpnfj6JmnM03JvU2ygF7U+BXQs9PRIhhBAtmFJqvFJqs1Jqm1LqhKVZlFLtlFK/KaVWKaXWKKXO9sQ4xcnpzT92cMN7aWjg1WkD+eqmUyQAE81PYZYJwkJizM9HD3p2PG7U/IIwgJRRkLfbfAkhhBBOppTyBV4CJgA9gUuVUjVbxN0HfKq17g9cArzs3lGKk9Wny/fwyA8bmdA7np9vH8343vHS1EU0T4WWTFhgJChfyYR5vfYjzGXGovq3E0IIIRpnCLBNa71Da10GfAxMqrGNBqwTbyKBTDeOT5ykflqbxd1frmFUlzievST1uHW9hGh2CrMhLB58fMx0oyOSCfNurXtCcLSUJAohhHCVtsAem5/3Wq6z9RAwTSm1F/gRs6amEC7zx9YcZnycTmpyFK9dMZBAP19PD0mIxis7CqX5JhMGJgiTTJiX8/GxrBcmQZgQQgiPuRSYqbVOAs4G3lNKnfC+qpS6QSm1Qim1QtrQi8ZauTuXv7yXRsdWobwzfQghAc2vt5oQxymyLNRsDcJC4yQIaxbaj4DcDMjf5+mRCCGEaHn2Ack2PydZrrN1LfApgNZ6CRAExNXckdb6da31IK31oFatWrlouKIl25RdwPS3l9EqPJBZ1w4hMsTf00MSoukKawRhITEShDULKZZ5YbtkXpgQQginWw50UUp1UEoFYBpvfFtjm93AWAClVA9MECapLuFUGQePcMVbywgO8OX9a4fSOjzI00MSwjmsQViYNQiTTFjz0KY3BEVCxh+eHokQQogWRmtdAdwCzAE2YrogrldK/VspdZ5ls78D1yulVgMfAdO11tozIxaOKCqt4PctOTzz82YWb/feRgDZ+SVMe2spFZVVvH/tUJJjQjw9JCGc54RMWKxZrLmq0nNjcqPmW1Ds4wvtTpEOiUIIIVxCa/0jpuGG7XUP2Hy/ARjh7nEJx+UXl5O26zBLdxzmz52HWbcvn8oqEy9/lraX+Xed5pImFxsyC4gI9iMp2vHgKfdIGVe8tZTcI2V8eP0wurQJd/r4hPCoomzwDTTN9sAEYWgozoPQlr/mXfMNwgBSRsKWn6AgCyISPD0aIYQQQniBvKNlLN1pgq6lOw+xMauAKg0Bvj6kJkdx02mdGNohlqLSCv76fhpfpO3jsqHtnDqGHTlFnPviQiqrNH3aRjK+dzzje8fTqVVYg/ctKq1g+jvL2HX4KO9ePYR+yVFOHZsQXqEwG8LbgHWNu1DLlNqjByUI83q288L6XOTZsQghhBDCoz5dvoe3F+1kU3YhAIF+PgxoF81tY7swtEMs/dtFEeR/LOOltaZfchQvz9/GlEFJ+Dtxza1n524lwNeHm8/oxNyNB3hyzmaenLOZLq3DmNA7nnG94+mZEHHCIssl5ZVc/+4K1mUW8Nq0gQzv1PI/jIqTVGE2hNskUUJizOVJMi+seQdh8X0hMMK0qpcgTAghhDhpbdlfyL++WkuPhHDuPKsrQzvG0jcpst4yQ6UUM8Z25pqZK/hq1T4uHpRc57aO2JRdwHdrMrnx1E7cMqYLt4zpQlZ+MXPWZTN7fTYv/raN5+dto11MCON7xzOuVzz9k6Oo0ppbPlzFkh2H+N/UfpzRs41TxiOEVyrMhlbdjv0cYsmEnSQLNjfvIMzHF9oNl/XChBBCiJOY1pqHvl1PWKAfs64ZSkxogN33Pb1ba3q3jeCl37YxuX9b/JyQDXv65y2EBfrxl9Gdqq9LiAxm+ogOTB/RgUNFpfyyYT+z12fzzqKdvL5gB63DA0mKDmbl7jz+PakXF/RPavI4hPBqRdnQ8bRjP4dYsr4nSSas+XZHtEoZAYe2QuF+T49ECCGEEB7w49psFm8/xJ1ndXUoAAOTDbt1TBd2HTrKd2symzyW1Xvy+GXDfq4f1bHO9bxiwwK5ZEg7Zl49hLT7z+S5S1IZ0C6azdmF3DWuG1cOT2nyOITwauXFUJJv5oRZVQdhkglrHtqPNJe7FkHvyZ4dixBCCCHc6mhZBY/+sIGeCRFcNrR9o/ZxZo82dI8P54V52zivX1t8fVTDd6rDUz9vJiY0gGtGdrBr+4ggfyaltmVSattGH1MIl8vdBZFJpgrNGarb09vMCfMPgoAw06b+JND8M2EJ/cx6Yeu+8PRIhBBCCOFmL/+2ncz8Eh6e1KvRwZOPj8mG7cg5wg9rsxo9lqU7DvHH1oPceGonwgKb/3luIQAoyIQXBsDaz5y3z+qFmmvMewyJkXLEZsPXD4bdDJu+h30rPT0aIYQQQrhJxsEjvL5gBxf0b8vglJgm7WtC73i6tA7jxXlbqapyfM1trTVP/byZ1uGBXDG8cRk5IbzSvjSoqoD96523z0LLyY7wGktMhcSdNI05mn8QBjDsRgiOgXmPeHokQgghhHCT/3y/AX9fxb8mdG/yvnx8FLeM6cyW/UXMWZ/t8P0XbD3I8oxcbh3T+bg2+EI0e5mrzGVuhvP2WWTp5RAef/z1IbGSCbOllMpQSq1VSqUrpVa4elAOC4qAkXfA9l9h12JPj0YIIYQQLjZv035+3XSAGWd0oXVEkFP2ObFvIh3jQnl+3ja0tj8bprXm6Z830zYqmKmDnbvosxAel5luLp0ZhBVmgW8ABEcff31onARhtThda52qtR7kstE0xeDrTF3pr/8BB144hRBCCNG8lJRX8vB3G+jUKpTpp9jXAMMevj6Km0/vzMasAuZuPGD3/eas38+avfnMOKMLAX4to8hICMB8ps5KN9/nZjjvM3bhfgiLhxqLlUsmrDkKCIHRd8HuxSYjJoQQQogW6a2FO9l16CgPndfL6UHPpNRE2sWE8PyvW+3KhlVWaZ75ZTMdW4Uyub90OBQtTP5eExRFd4DSAijOdc5+C7NOLEUEE4SVH4Wyo845jhez95VLAz8rpdKUUjfUtoFS6gal1Aql1IqcnBznjdARA66CyHZmbphkw4QQQogWJzOvmBfnbWN8r3hGdWnl9P37+fpw8+mdWLsvn/mbG/488/2aTLbsL+KOM7o6ZaFnIbyKNQvW63xz6aySxKL9dQdhcFJkw+x9tRiptR4ATABuVkqNrrmB1vp1rfUgrfWgVq2c/6JoF78AOO2fZgLhph88MwZHHdwKb4837T+FEEIIUa9Hf9xIldbce04Plx3jgv5JtI0K5rkGsmHllVX875ctdI8P55w+CXVuJ0SzlZkOyhe6TzQ/OysIqysTFhpnLiUIM7TW+yyXB4CvgCGuHFST9L0EYrvAb49CVaWnR1M/reHHu2D3Etiz1NOjEUIIIbza4u0H+WFNFjed1pnkmBCXHSfAz4ebTu9E+p48Fm6ru132lyv3knHoKH8/qxs+TVjgWQivlZUOrXtAq27mZ2cEYeXFUJLfQCas5bepbzAIU0qFKqXCrd8DZwHrXD2wRvP1g9P/BQc2wLovPT2a+m3+CXb8Zr7P2+PZsQghhBBerLyyioe+XU9SdDB/ObWjy4930cAkEiKD6pwbVlpRyfO/bqNfchRn9Gjt8vEI4XZam0xYQioEhps1vJwRhFUv1FxfEHa46cfxcvZkwtoAC5VSq4FlwA9a69muHVYT9bwA2vSB+f+FynJPj6Z2FaUw5x6I6waBEZAvQZgQQoiWrbJKU1rRuCqV95bsYsv+Ih6Y2NMt63AF+vny11M7sTwjlz93nPiB8KOlu9mXV8ydZ3VF1ezwJkRLULDPZKQSU83P0SnOCcLqWiMMjgVhJ8GCzQ0GYVrrHVrrfpavXlrrR90xsCbx8YEx98LhHZD+oadHU7s/X4HcnTD+vxCZLJkwIYQQLVpZRRXT31lG/3//wl2frWZ5xmG71+LKKSzlf79sYXTXVpzZs42LR3rM1MHJtA4P5Plftx53fXFZJS/+tp2hHWIY2TnObeMRwq2s64MlpJpLZwVhhVnmsrYgLCjKzEGTOWHNWNfx0HYQ/P6EyTp5k8L9sOBJ6DoBOp8BUcmSCRNCCNFiVVVp7vxsNX9sPciIznH8tC6bKa8uYczTv/Py/G0cKCip9/6Pz95ESUUlD57b061ZpyB/X/5yaieW7DjE8oxj2bB3l2RwsKiUO8d1kyyYaLmy0k1AFN/b/BydYlrWN7XKzFqOGF5LMxsfHwiJkTlhzZpSMPZ+KNgLK97x9GiO9+vDJjAcZ0kqSiZMCCFEC6W15pEfNvLt6kz+Ob47b1w5iGX3juWpKf1oFR7IE7M3M/yxeVw7czmz12VTVlF13P1X7s7l87S9XDOyA51ahbl9/JcNaUdcWEB1NqygpJxXf9/OqV1bMTglxu3jEcJtMldBq+7gH2x+jk4BXWkCsaYozAbfAAiOrv32k2TBZj9PD8ClOp4GKaPgj6dgwBUQEOrpEcG+NEj/AEbMgNhO5rqoZCjNN51igiI9Oz4hhBDCiV5bsIO3F+3k6hEp/NXSUCMkwI+LBiZx0cAkdh48wudpe/g8bS+/vn+A2NAALujflosHJ9OpVRgPfrOeNhGB3Dqmi0fGHxzgy/WjOvJ/P21i5e5cFmzJIe9oOXee1c0j4xHCLaxNObqOO3ZddIq5zM2AmA6N33dhtmnKUVcWOSROGnO0CGMfgCM5sPQ1T4/EPKF/+ieEtoZRdx67PjLZXEo2TAghRAvyRdpeHvtpE+f2S+T+c2ovJewQF8pd47qz6J9jeGf6YIZ0iOHdJRmc9b8FjH16Pmv35XPP2T0IC/TceeNpw9oTHeLPYz9t4s0/djK+Vzx9kuSkqWjBrE05rPPB4PggrCmKsiG8nrmdITHSmKNFSB4CXcbBouegOM+zY1n7GexdDmc8CEERx66Pamcum5reFUIIIbzEb5sP8I8v1jCicyxPTenb4Dpafr4+nN69Na9MG8if/xrL/ZYuiGf0aMN5/RLdNOrahQb6cd2ojizbeZgjZRX87ayuHh2PEC5nbcqR2P/YdRGJpowwd2fT9l2YXXtTDquTpByx5QdhAGPug5I8WPKS58ZQWgS/PGCezP0uO/42ayZMmnMIIYRwopzCUt77cxeVVfZ1IXSWVbtzuen9lXSPD+fVaQMJ9HOspXxsWCDXjuzA7NtH8+ZVg7yi+cWVw9sTYymV7Nom3NPDEcK1ajblAPDxNYmDpmbCCrNrb8phFRoHxYehqqrubVqAlj0nzCqhL/Q8H/58GYb+xfxx3W3h/0xLzotnmc4vtkJbgW8g5O12/7iEEEK0WB8s3cWzc7eSf7SMW9w0p2p7ThHXzFxOq/BAZl49hPAgf7cc19XCg/yZ+7dTCQ10/RplQnhcZvrxTTmsmtqmvrzYJEbC6itHjAVdZbYLabnNb06OTBjA6fdA+VETDLlbbgYsfgH6XGzKI2vy8YHIJMmECSGEcKoVGbkAPPPLFpbtdP1E9/0FJVz51jJ8fRSzrhlCq/BAlx/TnWJCAxzO6gnR7GhtMmHWRZptNTUIq16ouZ5MWIglWdLC54WdPEFYq27QdyosfxMKstx77J/vNyncMx6qe5soaVMvhBDCeSoqq1i1O5fJ/dvSLiaE2z5axeEjZS47XkFJOVe9vYy8o2W8M30IKXFe0JFYCOG4gkzT1M62KYdVdIrp5l2c27h9V68R1kBjDmjx88JOniAM4NR/QlWFWSjZXXYugI3fwqi/QWTbureLlAWbhRBCOM+m7EKOlFVyardWvHjZAA4fKePOz1ZT5YL5YSXllVz/7gq25xTx6hUDpXOgEM1ZVrq5rCsTBo3PhhVaEiENzQkDCcJalJgOMOBKWPlu0ycV2qOyAmb/y0xiHH5L/dtGtTMp2vIS149LCCFEi5e2y5ypHtg+mt5tI7lvYg/mbTrAWwub2NmshsoqzR2fpLN052GemtKPUV1aOXX/Qgg3y0wH5QNtep94W5ODMEs5YlgD3RHBtMhvwU6uIAxg9F3g4wdzH3L9sVa+C/vXwVmPnDixsSZrh8SCfa4flxBCiBZvxa5c4iOCaBtl3n+uGNae8b3ieXz2JlbtbmQpUQ1aax76dj0/rcvmvnN6MCm1nooPIUTzkLnKNOUICDnxtqj25rIpmTAf//obblQHYZIJa1kiEmHkHbD+K1Mq6CrFuTDvEUgZBT3Oa3j7KOuCzdIhUQghRNOlZRxmYEp0dXt3pRSPX9SX+MggbvlwFflHy5u0/4rKKv7z/Ube+3MXfxndketGdXTGsIUQnmRtylHbfDAw69yGxDY+CCvab9YIq2/ZCf9g8A+FIxKEtTwjZpjyvx//AZVNexOq0/zHTGvN8Y/V/0SzkrXChBBCOMm+vGIy80sY1D76uOsjg/158bIB7C8o4R9frEbrxs0PO1RUypVvL+PtRTuZfkoK/xzf3RnDFkJ4mrUpR23zwaya0iGxMKv+hZqtToIFm0/OIMw/2ARHORtNt0RnO7AJlr0BA68+fpG7+kQkmvpb6ZAohBCiiVZkmHb0g1NOLPlJTY7i7gndmbN+P7OW7HJ436v35HHuCwtZsSuXJy/qy0Pn9cLHx/OLKQshnMDalKOuTBg0MQjbX/8aYVahsTInrMXqdjZ0Ggu//ReKDjhvv1rD7LshMAxOv9f++/n6Q3iiZMKEEEI0WdquXEICfOkeH17r7deO7MDY7q159IeNrNuXb/d+P1m+mymvLkEpxZc3nsKUQcnOGrIQwhtYm3LE96l7m+gUkzSorHB8/4VZ9XdGtJJMWAumFEx43KzcPfdh5+03bSbs+A1Ou8dE8Y6ITJJMmBBCiCZbkZFL/3ZR+PnW/javlOKpKf2IDQvglg9XUlhSf2l+aUUl93y1ln9+sZYhHWL47taR9G4rbeiFaHGy0utuymEVnQK6Egr2Orbv8hIzVceucsS4Fh+E+Xl6AB4V1wWG3wyLnoWB0yF5cNP2t2c5/HgXdD4Dhlzv+P2jkmHP0qaNQQghxEmtqLSCTdkF3DKmS73bRYcG8Pyl/bnk9T+556t1PH9JanUTD1tZ+cXc+P5K0vfkceNpnbjzrG74Svlh86O1aUiWMgp8ToJz8EU5pgmEvdNCXG3/Bshe48IDKOg89tgaW42htcmEdT6j/u1s29Rbv7dHkXWhZjvnhDmrMUfZUdj8o1kr2BF9poCPr3PGUIuTOwgD07J+zSfw451w/bzG/7IL98OnV5gFmS98s3H7iUw2XRurKl36RxdCCNFyrdqdS5XmhKYctRmcEsPfzuzKk3M2M6JTLJcMaXfc7X/uOMQtH66kuKySV6cNYHxvO8qIhHfatRhmnQeTXoL+0zw9Gtf79WFY9wX8fRMEeThrW1kO708+tlCxq/SZYj6DNlZhFhw5UH9TDmj8WmHWNcLsCsJioPyIqVhraJmnhqz5GL6/w/H79TxfgjCXCgwz63h9cS2ses9kxBxVWQ6fTYeSfLj2Fwhu+I2vVlHJJkovzDKliUIIIYSDlmfk4qOgf7sou7a/8dRO/LnjEA9+u57UdlF0j49Aa83bizL4748baR8bwsc3DKNz69rnl4lmYt8Kc7ninZMjCNuXBuVHYc2njatOcqYtc8xnu0kvQftTXHOMhf+D1R/D+Mcdnw5jlZluLutrygEQ0dasuetwEGYJQutbqNnKmtE7eqjpn4kPbQe/YLhxkX0dy618A5p23AZIEAbQ+0JY8baZG9bjvPoXkKvNnHth92K48K2mpb0jLWcg8/ZIECaEEKJR0nYdpnt8BOFB/nZt7+OjeObiVCY89wc3f7CST/4ynH9/t4FvV2dyVs82PH1xP7v3JbxY5ipzuW8FZK+tv/FCc1d2BHI2me/TZsLg6xz78O1saTNNM4q+l4Cviz56D70RVs6C1R/BKbc0bh9Z6Q035QCTHYpq14ggzFqOaGdjDnBOEGYtm4zt1LT9OJndRcFKKV+l1Cql1PeuHJBHKAUTnjCTBX971LH7rv4Ylr0Gw2+BPhc1bRxRslaYEEJ4C6XUeKXUZqXUNqXU3XVsc7FSaoNSar1S6kN3j7GmisoqVu3OY1CKYxUZrcIDee6SVHYcPMKox3/juzWZ3DWuG69OGygBWEuRmW7mg/kGQtq7nh6Na2WvA10Fnc+E/etg30rPjSVvN2ybC/2vcF0ABtCmJyQPNQFfI9f/IzMd4rrV35TDKrqD40FYUTb4+NuX7AixZMKOOKFNvaNz19zEkZmZM4CNrhqIx8X3hsHXm4xYlp0TJzPT4bsZ5kXtDCd0WLRG+nm7m74vIYQQjaaU8gVeAiYAPYFLlVI9a2zTBfgXMEJr3Qu43d3jrGlTdiFHyyoZaMd8sJpGdI7jzrO6ERbkx7tXD+Hm0zvL+l8tRXEe5O6EjqdBr/PNXPiyox4elAtZ17oa9yj4h0LaO54by8r3zOWAK1x/rIHT4dBWM//PUVqbbGlD88GsGrNWWGG2mQ9mT1ayOhN22LFj1KR18w7ClFJJwDmAC1Y29iKn/8vM5/rpHw2fRThyCD65wkTqU2Y65+xGQKh50kkmTAghPG0IsE1rvUNrXQZ8DEyqsc31wEta61wArbUTF51sHOsizYNqWaTZHjef3pll94xldNdWzhyW8LSs1eYyMRUGXg2lBbD+S48OyaUy0yG0FcR1NVVK676AkgL3j6OywvQb6HyGKd9ztZ7nQ2CkyYY5ytqUo6H5YFbRKVCcawJ8u4+Rbd9CzXD8nLCmOHoIyoqabxAGPAv8A6iqawOl1A1KqRVKqRU5OTnOGJv7BUfDGQ/B7iWw9rO6t6usgC+uMa1Pp85qWjvQmiKTZa0wIYTwvLaA7YvxXst1troCXZVSi5RSfyqlxte2I3e+P67YlUtCZBBtoxrfTay2NvWimbNmhhL6Q7thpuSsMR/Um4usdBNMKGWyQ+VHYe2n7h/HVktDjsY0fWuMgBDoNxU2fON4BsnalMORTBg4lg2zZsLsERRl5qcdbWI5onV8zTEIU0pNBA5ordPq205r/brWepDWelCrVs34DFrqNEgcAD/fV/dZk3n/hh3z4Zynoe1A5x4/KlkyYUII0Tz4AV2A04BLgTeUUlE1N3LX+6PWmhUZuY0qRRQtXGa6OckbGnssMNm73MydamnKjpqmHNZgIrE/xPeFFTMbP1eqsdJmmk6AXWs9P+MaA6dDZalp0OEIe5tyWDUmCCtyIAjz8YHgmKZnwppzEAaMAM5TSmVgyjHGKKXed+moPMnHB85+ymS5Fjxx4u3rv4JFz8Gga1xT3xvZDvL3uv+FQgghhK19QLLNz0mW62ztBb7VWpdrrXcCWzBBmUfsyysmu6CEwY0sRRQtWFY6JPQ79nO/S0yDjpUtsEHHfktTDmtZnTXo3L8WMt3YoCNvD2z9xXxWdGVDjpra9IKkIY436MhMN+WbAaH2bR/d3lzaG4SVl5jyRXuDMLAs2NzUTNhOc+mOclAHNRiEaa3/pbVO0lqnAJcA87TWLXuBiaSBpovNn69AzpZj1+/fAF/fbJ7c4x93zbGjkk3avKkTEYUQQjTFcqCLUqqDUioA8/73bY1tvsZkwVBKxWHKE3e4cYzHSduVCyCZMHG8knw4vMNkhKxCYqDnJFjdAht0VJfV2TzePlPAP8S9JZirLA05+ruhIUdNA6fDwS1meo29stKP/501JCjSZKrsDcKKLAs127NGmFVIbNM/D+dmmGPa0/HRzRzpjnhyGfug6ahjbdJRnAefXG4Wd754Fvi5aAG3SGubeumQKIQQnqK1rgBuAeZgOgN/qrVer5T6t1LqPMtmc4BDSqkNwG/AXVrrJtbONN6KjFxCA3zpHi+LKgsbtk05bA2cDqX5psKnJclKN005IhKPXRcUYdaEXeumBh2VFaYrYuexxzJG7tTrAscadBRkmSDJ3qYcVo50SHRkjTCr0Nimzwk7nOGVpYjgYBCmtZ6vtZ7oqsF4lbBWMOZe2PEbbPwWvvqLaR0/5V2IcOAJ5CjrWmHSnEMIITxKa/2j1rqr1rqT1vpRy3UPaK2/tXyvtdZ/01r31Fr30Vp/7MnxrtiVS/920fj5yvlVYcOaGUqokeVof4opP2tpDToy04815bA16GooPwLrPnf9GLb9AoWZ7mvIUVNACPS9GNZ/bV8mydq4xd6mHFaOBGFF1iDM0UyYE+aEtYQg7KQz6Fpo3Qu+uA62zIbxj0H74a49ZqQs2CyEEMIxhSXlbM4ukFJEcaKs9GNNOWxVN+hYBvvXe2JkzlezKYetxAGm6cSKd1w/737FO6YVuzsbctRU3aDDjnNDmemONeWwik4xn1crKxretrAxQVicCSKr6mzOXr+KUijYJ0FYs+TrB2c/CZVl0O8yGHyd648ZHA0BYZIJE0IIYbdVu/Oo0jAoRYIwUUNm+vFNOWz1uxR8AyCthTTo2L8OdGXtZXXWoDN7jVmU2FXy9phMWP8rwNffdcdpSHxvSBpsX4OOrHTHmnJYRadAVYUJdBpSmAU+/mYemb1CYs3fsyTPsXFZ5e0BNMR0aNz9XUyCsIakjIDbVsGkF+1b4buplDJnrCQTJoQQwk4rMg7jo6B/OwnChI2SfDi8ve4yM2uDjjUft4wGHQ2tdeWOBh2r3jdBjys6aDtq4HQ4uBl2/1n/dpmrHJ8PBo61qS/cb7KDPg6EHtULNjeyOYcXt6cHCcLsE9MRfHzdd7yoZDP/TAghhLDDil259EiIICzQja2whfezNuWoOR/M1sDpJljb8LU7RuRaWemmhC2i5rrqFkGR0HsyrP0cSgudf/zKCtMVsdMY7/jg3+sCCIyoP+i0NuVwdD4YOBiEZTlWigjmJAE0vjmHtT29N/wtaiFBmDeSTJgQQgg7VVRWkb4nj0EyH0zU1FBmCKD9CIjt3DIadGSmm8daX+XSQEuDjrUuaNCxba4pzfNUQ46aAkItDTq+qjubZG3K0ZhMWERb8PGzLwgr2t+IIMwyj7GxzTlyM8AvyGTgvJAEYd4oKtksaFda5OmRCCGE8HIbswo5WlbJQFmkWdSUlQ4RScfKumpjnSu1Z6lZD7W5Ki82TTkaCibaDoQ2fVwTdKbNNB/4u01w/r4by9qgY82ntd+emQ4ox5tygOmdEJnswkyY5Xnb2AWbrZ0R3TGdqBEkCPNG0iFRCCGEnVbsMme4JRMmTmDNDDWk32WmQcfKZtygI9vSlKOhx6sUDLzKBKjObNCRvxe2zoH+0zzbkKOm+D7QdhCk1dEV0tqUIzCscfu3p019RalJLjiyUDM4IRO2y2tLEUGCMO8UKWuFCSGEsM+KXbkkRgaRGBXs6aEIb9JQUw5bobHQ4zxY/ZHJKDVH1Wtd1TP/zarvxeAX7Nxs2Kr3QVfBgCudt09nGTjdZAn3LD3xNnsD9brEdGg4CGtMe3ow6535hzQuCNPaq9cIAwnCvJN1weZ8ac4hhBCiblprVmQcZpCUIoqastaYy/qactiyNuhY/7WrRuRamen1N+WwFRQJvS90XoOOqkpYOct7GnLU1HsyBISfGHQWZptFlO0JXOsSnQLFh81zpy5F+82lo0EYNH7B5qOHoazQO/8eFhKEeaOweLOWgmTChBBC1GNvbjH7C0plfTBxourMUKp926eMhJhOzbdBR1Z6w005bA2cDmVFsO6Lph/b2xpy1GTboKM499j11sYtjWnKYVXdIXFX3dsUZplLdwZhXt6eHiQI804+PhDZVuaECSGEqFfaLvOBaqDMBxM1ZaY33JTDVnWDjj/hwEZXjsz5yovNmB0JJpIGQetezgk602ZCaGvodnbT9+UqA6dDRcnxDTqy0ml0Uw4re9rUV5cjJji+/5DYxjXm8PL29CBBmPeKTJZMmBBCiHqt2HWYsEA/usdHeHoowttYM0OOSL3MVOKkNbMGHfY25bClFAy62jTnsGaEGiN/H2yZDf0v966GHDUl9IXEAbDCpkFHZnrTmnKA/UGYjx8EN6JsuqmZsKj2jt/XTSQI81ZR7SQTJoQQol4rMnLp3y4KXx/vbMEsPKSkAA5tc7zMLDQOepzb/Bp0NHatqz5Tmt6go7ohx1WN34e7DLoacjbCnmXm58YE6jUFRUJwdMNBWFi8qfRyVGhc44OwsDamuYeXkiDMW0UmmydtRZmnRyKEEMILFZSUs3l/oZQiihNlrTaXjfmAPXA6lOTBhm+cOCAXy0w3GZPIJMfuFxxlmlas/axxa7NaG3J0PN10CfR2vWwadBRmm7laTZkPZhWdcqz8rzZF2RDeyAWTQ2LM3L3yEsfu5+WdEUGCMO8VlQxoKNjr6ZEIIYTwQqt256E1DJbOiKKmxmaGAFJGQUzH5tWgIyvdPNbGLMrblAYd2341n9O8tSFHTYFh0HcKrP8SdvxurmtqJgwaXiusMLtx88Hg2ILNxYcdu58EYaLRZK0wIYQQ9ViRcRhfH0VqcpSnhyK8TWa6adUe1srx+/r4mKBi9xI4sMnZI3M+a1OOxgYTSYOhdc/GBZ1pMyG0lXc35KjJ2qBj3n8wTTn6Nn2f0SmQt9tkBmtTmG1KAxvDumCzI805KsrM4tkShIlGqV4rTIIwIYQQJ1qRkUuPhHBCA/08PRThbayZocbqZ2nQsbIZNOjYv9405Wjs47V2hcxceayM0x4FmaYhR+rl4BfQuGN7QkI/sy5Y/h6I69K0phxW0SlQVWHa9NdUUWqyWI3NhFm7ezoyLyx/D6C9PgiTV25vFZEEKMmECSGEOEF5ZRXpe/KYOjjZ00M5uWydC7GdvHv+j7UpR9+pjd9HWCvoMRHSP4SxD4B/sPPG52yZq8xlU8rq+l4MvzwAP98HHUbbedx0E/wNuLLxx/WUgdPN780Z88Hg+A6JUe2Ov616oeYmZsIcCcKaQXt6kCDMe/kFmEXt8mVOmBBCiONtzCqguLxSmnK4U2E2fDQVel8Ek1/z9Gjqlr3GXCb2b9p++l1mFvfdvQQ6jWn6uFwlK93SlKMJJySCo013w2Wvwc4F9t+v+0QTlDc3vS+CRc9B13HO2Z9tEFYziG3KGmFwbE6YQ0FYhmVcXnyyBAnCvFtkMuTv9vQohBBCeJkVGWaR5kEpEoS5zar3TcmVNfPiraxrXjU1y5E06Nj+vDkIy1zd+KYcts5+AsY96th9fJrpx+jAMLjNic/jiCRQvrU357AGYY2dExYcBSjH5oTlZoBfUOOP6SbN9NlzkohKhn0rPT0KIYQQXiZtVy5to4JJiPTiMrGWpKrq2Pyog1tMO3NnzKVxhaz0xjflsBUSYxa6tXZa9EblJWbdq65nOWd/3rzYsjfz9TOfWesLwhqbCfPxNZlKRzNhUe0bty6ZGzU4OqVUkFJqmVJqtVJqvVLqYXcMTGAyYQX7zIu/EEIIAWitWZ5xWLJg7rTjN9P9rc/FgD5W8ueNMtOdN9cnMdW7M3/715vspLMer2i8utrUF2aZjKF1bldjhMbBUQczYV4+Hwzs645YCozRWvcDUoHxSqlhLh2VMKKSobLs2KRGIYQQJ729ucUcKCxlkMwHc5+0meZD5Nj7zc/Wkj9vU1pomnI4Y+0nMMFN3m446uAaTe6SaakWctbjFY1XVxBWtN+UBTYlKxUSa/9zUGvI3dUygjBtWJcR97d8aZeOShiRlg4z0qZeCCGExYpd5sPIwPaySLNbFO6HzT9C6mWm81tYvPeW6GWtAbRzM2HgxY83HYJjmtaUQzhHdIopGSwpOP76wizTaK4pQmLtL0cszoXSgpYRhAEopXyVUunAAeAXrfXSWra5QSm1Qim1Iicnx8nDPElZ1wrLk+YcQgghjBUZuYQH+tEtPtzTQzk5pFsacgyYbn5OTPXeTJgz2rXbsgZzXvt4V5vH2tSmHKLprJ0I83Ydf33hfnPioilCYu1vzNFM2tODnUGY1rpSa50KJAFDlFK9a9nmda31IK31oFatmjgZVBiRSeZSMmFCCCEs0nblktouCl8f+eDpclVVkPYupIyCuM7muoTUY805vE1WOoQnQlhr5+wvJMZk/7wxE2ZtyiHzwbyDbZt6W87IhIXGmUyYtqMQr7o9fUrTjukGDhVoaq3zgN+A8S4ZjTheYDgERcmCzUIIIQDILy5n8/5CBqdIKaJb7JxvzuwPnH7susRUTHOOtZ4ZU30y050/Pyqxv3dmwqxNOWQ+mHeoLQirKIXiw84pR9SVUJLX8LbVQVj7ph3TDezpjthKKRVl+T4YOBPY5OJxCauoZMmECSGEAGDV7ly0RppyuEvaTDPnqPvEY9dZMy/elh2yNuVwdmYoIdUEot7WnCPLUnopmTDvEBxlEge2QZi1sZwzgjCw7zmYmwGhrSEgtGnHdAN7MmEJwG9KqTXAcsycsO9dOyxRLbKdZMKEEEIAEBsayMWDkuiXHOXpobR8RQdg0w+mIYd/0LHrIxLMHBdvyw5Zm3Ik9nfufqubc6x27n6bKjPdBMhR7Tw9EmFVs0NioSUIa/KcsDhzac+8sMM7m0UpItixWLPWeg3g5P9oYbeoZNj5u6mDlYmnQghxUuuTFMkTF/Xz9DBODukfmHI321JEq8RU78uEWcfj7PI828xfp9Odu++myEqXphzeJjoF9q879nNhlrlscibMUn5tT4fE3F3QrnmspOXdS0kL03a1rMi03BRCCCGE61VVmVLE9iMhrsuJt1ubc5QdcffI6paZ7tymHFbW5hzelPkrL4ED0pTD60SnmI7eVZXmZ2eVI4ZaMmENBWEVZVCwt9lkwiQI83bWNvUyL0wIIYRwj52/m7Kq2rJgYDIwusq7mnNYM0OukJDqXZm/A9KUwytFp0Bl2bEMWGEWKN9j5YSNVT0nrIFyxPw95v9SgjDhFNYFCGVemBBCCOEeaTMhOBp6nFv77d62flZpIRzc6rrMUGKqCUq9pSonU5pyeKWaHRILs00WzKeJ4UZAKPgFN5wJa0bt6UGCMO9nnXAqmTAhhBDC9YoOwKbvoV+Nhhy2IhIgrI33ZIey12KacqS6Zv/V88K8pDlHZroJkqUph3exBj+HLQsmF2ab/xNnCImFIxKECXcKiTXRv2TChBBCCNdL/9DSkOOq+rdLSD2WkfE0a0bOZZkwS382b3m8WenmsUpTDu8SmWTKD4/LhCU4Z98hMfZlwnwDnHdMF5MgzNspZVkrbLenRyKEEEK0bFVVsPJdaD8CWnWrf9vEVO9pzpG5ynzwDHdS1qGmkBizZI43lF9am3LIfDDv4+tvAjFrEFaU7bznZGhcw3PCcjMgqn3Tyx/dpHmM8mQXmSyZMCGEEMLVMhbA4R11N+SwlZDqPc05rJkhV/KWtvzWphwyH8w7WdcKqygzmSunZcJi7cuExXRwzvHcQIKw5iAqWeaEtXRHDsHS1+G9C2D3n54ejRBCnJzSZkJQFPQ4r+FtrZkYT2eHrE05XJ0Z8pbmHNbft2TCvJM1CLO2p3fanLA4OHq47tu1NsdtJvPBwI7FmoUXiEw20X/ZEdMhRrQMFaWw9WdY/TFsmQNV5aB8ICCs2Sw0KIQQLUZRDmz8HoZcX3dDDlvhCRDa2vPZIWtTDldnhmybc3Q8zbXHqk9WugmUo9p7bgyibtEppmzw0DbzszMzYaUF5rOTX+CJtxfnmtubURAmmbDmoLpD4l7PjkM0ndawdwX88Hd4uht8Mg32Lodhf4W/LoJB15rArLTQ0yMV4qSnlBqvlNqslNqmlLq7nu0uVEpppdQgd45PONnqD83JsAENNOSwUspkYzydCXNXZqi6OUe6a4/TkMx0MxZpyuGdrEHQnqXm0mlzwqxrhdWRDWtmnRFBMmHNQ6TNgs0NTRQW3ilvD6z5xGS9Dm01HS97TIR+l0CH08DX8q/YezIsf8Nkxvpc5MkRC3FSU0r5Ai8BZwJ7geVKqW+11htqbBcOzACWun+Uwmm0NqWI7YZD6+723y8hFbbN9WylSla6pSlHvGuPY23O4cnMX0Wpacpxyi2eG4Oon3VO1u4l5tKZmTAwWbaIWvYpQZhwiShZsLlZytsN23+DtZ9Bxh/muvYjYcQM6DkJgiJOvE/yMPOCte5LCcKE8KwhwDat9Q4ApdTHwCRgQ43t/gM8Dtzl3uEJp9ppachx6j8du19if0tzjnXQbqhrxtaQzHT3NalI7OfZTNj+9SZbKU05vFd1Jmy5aVcfEuec/Vr3U1dzDmsQ1ozKVCUIaw7CE8DHT5pzeLuCLBNs7fwddv4BebvM9TGd4PT7oO/FEN3Ai4OPD/Q8H1a8BSUFtQdqQgh3aAvYvujuBY77lK2UGgAka61/UErVGYQppW4AbgBo104Wl/VK1oYcPSc5dj9rCWBWumeCsNIi0ya/92T3HC8hFTZ+B8V5EBzlnmPasmbhpCmH9wqOhqBIKMmH8ETntYu3ZsKO1NGmPjcDQltBYJhzjucGEoQ1Bz6+EJEomTBvc+SgJej6w5xFPbTVXB8UBSkjYfjNkDIKWvdwrHa992RY+gps/tGUKwohvI5Sygd4Bpje0LZa69eB1wEGDRqkXTsy4bAjB01gMfg68A927L7W5hyeyg65qymHVXXQuRo6nuqeY9rKXCVNOZqD6BTzHHFmiWxIQ3PCdjarUkSQIKz5iGwnmTBvkLkKVn9igq4D6811AWFmYc+BV0GH0dCmtwmcG6vtIIhIMiWJEoQJ4Sn7gGSbn5Ms11mFA72B+cqcZIkHvlVKnae1XuG2UYqmS7c05BhoZ0MOW9bmHJ6aJ+XuzFBC/2PH9UgQlm4eqzTl8G6uCMKCowFVfzlisodKghtJgrDmIirZZFy8Tc4WWPeF6X4TkQSRbSGirUlFt7QXyfJimHmuWSSy3TDocyGkjDZvCL7+zjuOjw/0Oh+WvmZargZHO2/fQgh7LQe6KKU6YIKvS4DLrDdqrfOB6skOSqn5wJ0SgDUz1oYcycNM1UJjVDfnOAoBIc4cXcMy0yEs3vVNOaxCY02zME9k/qxNOYbf7P5jC8dYM1LOfF76+pkS2KO1lCNWlpsO4n2nOu94biBBWHMRmQyFmeaJ5swP/E3126Ow4esTr/cPPRaQRba1CdASzZk0a6vR5mTH71BWCNO+gM5nuPZYvSbDkhdh0w/Qf5prjyWEOIHWukIpdQswB/AF3tZar1dK/RtYobX+1rMjFE6RsRAOb4fRTeirkphqac6x1v3zwjJXuX9+VEI/z2T+rE05ZD6Y97MGYWFOPjkQEld7Jix/j/kflHJE4RKRSeYJVpDZcHMHd6kog+3zIHUanP4vM7b8vVCwD/L3mcuCfbB1o2XldMtUiNBWcNOfEOqkjjnusul7CIww2S9XazvArA+3/isJwoTwEK31j8CPNa57oI5tT3PHmEQDNnwL23+1f/u9aaZyo9f5jT9m9SLG6e4NwqxNOXpd4L5jgukIuen7pjfnOLQdlrwEutK+7Q/vMJfSGdH7uSITBmZeWG2NOZphe3qQIKz5iLJZK8xbgrDdi83q5N3PMUFiZBIkD6l924oyKMo2Z7I+uQJ++idc9JZ7x9sUVZWw+Sfochb4Bbj+eEqZN9YlL5lJqCExrj+mEEI0d7/cD4X7HessO+J2xxty2IpINCcX3V2it2U2oKH9cPce11nNOX5/HNZ+7tgJ2ZRRze6D9kkpsT8kDTHr7jlTaBwc3nni9RKECZeKtLQ19qYOiVvmgG+gfS/CfgEmsxPVzpR9zP8v9L4Qup/t+nE6w55lpg65+znuO2avybDoOdO1qzETxoUQ4mRSnGs+jJ3xEIy8w33HVcpkZ9xdopc203QJdEd1hi1nNOc4ehjWfw2DroZznnbWyIS3CI6G635x/n5DYmDv8hOvz80A3wDnLQztJg0271dKJSulflNKbVBKrVdKzXDHwEQNkUnm0ls6JGptMkMdRkNAqGP3HXkHtO4F399hyhmag03fm39wV88Fs5XQD2I6wvov3XdMIYRorrJWm0tPlKslpkLOJtOcwx0ObjNLpAy8ynnrMNnLGc05Vn8MlaUwcLqzRiVOBiGxZk6YrrHSR26GOcnflM7UHmDPf24F8HetdU9gGHCzUqqna4clTuAfZNYiydvt6ZEYB7eaNRm6jXf8vn4BcP5LcCQHfr7P+WNzNq1Ng4wOp7p38WRrSeLOBVCU477jCiFEc5S5ylwm9HP/sRNSzbzt/evcc7yVM8HHz8zJ9oSmNOewdqRsOwji+zhzVKKlC4kzHapLC46/Pjej2ZUigh1BmNY6S2u90vJ9IbARaOvqgYlaRCV7TyZsy2xz2WVc4+6f2B9OuRVWvWeae3izAxtNwOnOUkSrXpPNG/tGacQmhBD1ykw35XmemENrnSfljnlhFaVmbbNuE8zyMJ6QmGoaZZTkO37f3X/Cwc2SBROOsy7YXLM5R0sNwmwppVKA/sBSl4xG1C8y2XvmhG2ZYxYljkpueNu6nHY3xHaGb2eYLk/eatMPgIJuHpi/1qYXxHYxXRKFEELULSvdc+3LI9qas/TumBe26XtTkuXJIKZ6Xthqx++bNhMCwqH3ZKcOSZwErE1cjh4+dl1xrjkZEN3BM2NqAruDMKVUGPAFcLvWuqCW229QSq1QSq3IyZHSKZeISjYt4KuqPDuO4lzYvQS6NqIU0ZZ/MEx6yWT3fv23c8bmCpu+h6TBnjnjqJR5o8pYaDp+iZPHph+PlVcJIepnbcrhqfblSpkKD3dkwtJmmvkvHce4/lh1aWzm7+hhc1Kx78WOzycXwprltl2wuZl2RgQ7gzCllD8mAPtAa11rlwCt9eta60Fa60GtWrVy5hiFVWQ7M5H1iIeD3G2/mnU9mhqEAbQbBkNugGWvmxIFb5O/15zZ9EQpolWvCwANG77x3BiEe1WWw5fXw7e3njgBWQhxImtGxpML+Vqbc5QXu+4Yh7abecIDPNCQw1ZoHEQkOZ75W/OJ+Rwz6GqXDEu0cCHWTJjNgs0tOQhTSingLWCj1voZ1w9J1Ml2rTBP2jLb/CO0HeCc/Y19wDy2b2527ZtXY2yyrNPafaLnxtC6B7TqIV0STyZZq6GsCLLXSjZMCHtYMzKeXMg3IdWcoMx2YXOOle+C8oX+HmrIYSsx1bFMWHVDjoHSkEM0Tm1zwqqDMC9ZQ9cB9pxGGQFcAYxRSqVbvprJ4k4tTKQlCPNkh8TKCtj6C3Qd57xWoIFhcO5zcGgbzH/MOft0lk3fQ1w3iOvs2XH0nmxKQAsyPTsO4R4Zf5hLvyDzoUsIUb+sdFOi58mF7asXMU53zf4rymDVB5aGHPGuOYYjElLh8Hb7m3PsWWoyhdKQQzRWQKhZn7ZmJiwkDgLDPTasxrKnO+JCrbXSWvfVWqdavn50x+BEDd6QCduzFEryTBDmTJ3GQP8rYPEL3nPmvzjXzMXyZCmiVa8LzOX6rz06DOEmO/8w2c9ek2Ht597duEYIb5CZ7tksGBxrzuGqeWGbvjdzYQZ6SSlforU5xxr7trc25OglDTlEIyllSmFtG3Mc3tksSxHBwe6IwsOCIiEw0rMdErfMBh9/6Hi68/d91iMQ1hq+ucWc8fO0LT+b0hJPliJaxXWBNn2kS+LJoLLczI9MGWkWYi0rklJUIepTnGeWEfHkfDCwNOdIdd2JxLSZZm54Jxe8/zZGdXMOOx5vca6lIccUU/0iRGOFxJzYmEOCMOEW9q4VVlVlXhjnPw7fzTDrijjDltnmw6ErFi0OjoJznjGLXS561vn7d9Sm7yEs/tjZPk/rdT7sXeY9yxQI18hMh/Ij5v8seagph02TkkThJpt+gG9vg6pKT4/EftamHJ7OhIEZgyuacxzaDjt/h4FXOm8qQFM50pxj9SdQUSKliKLpQuKOlSNWlpsGahKECbeItLSpr01Jvumg9/XN8Ex3eP00mP9fc/Ys/cOmH/vQdji4xTldEevS/WzofRH8/gTs3+C64zSkvMR0gex+tmc7UNmyliRu+NqjwxAuZp0P1n6EObM+8CrYtwL2r/fsuETLt+N3+PQqMw9x5++eHo39rEGAN5wwS0x1TXOOlbNMQ45UL2jIYcue5hzWhhyJAyChnxsGJVq0kNhjjTny95r/NwnChFtE2SzYrDXkbIZFz8PMifBER/j0Stj0nfkAd/6rcOc288a06FnTVKMptswxl86eD1bThMdNpu3bWzx3Nnbn7yYb4Q3zwaxiO5k3sHVSmtaiZSw088HCLEt99L0EfAMkGyZcK3stfDINYjub0vf0jzw9Ivtlpnu+KYeVNRvnzOYcFWWQbmnIEZHgvP06gz3NOfYsg5yNkgUTzmE7J6wZt6cHCcKan8hkKM03JYbP9YWXhsAv95sn5PBb4Oqf4K4dMOUdSL3UfJAb9XfzRG1qBmXLbGjVHWJcvCp5aBxMeAL2pcGfL7v2WHXZ9D0ERkDKaM8cvy69JkPmSjMRVbQ81vlgHUYduy40FnqcC2s+9r4lHETLkLcb3r/IdBeb9gX0vhA2fgclBZ4emX0yV3lHKSJAZJI5U+/M5hybfzTrg3pjEFPdEbKe5hxpMyEgzDyvhGiqkFjzObiiTIIw4WatupnLNZ9C614w8X9w+zq4aTGc+TC0PwV8/Y6/T7dzzLySP55p/MKvJQWwa5Hrs2BWvS804573iCmDdKeqStj8E3Q5E/wC3HvshvQ631xKSWLLZDsfzNaAqyzlxt96ZFiiBTt6GN6/ECqKTQAW2Rb6XWZ+bg6vM97SlMNKKRMQOjMTlvaOOQHbaYzz9uksDWX+inNNY6E+0pBDOIl1rbDiwyYI8/GHiESPDqmxJAhrbjqfCX9ZAP/YCZd9DIOuOda6vi4+PjDyDjiw/lhJoaO2/wpVFa6dD2ZLKTjnabMexDc3N72U0hF7l5uzjt5UimgVnWIWupSSxJbJdj6YrZRREN1B1gwTzlVeDB9dArm74JKPzMLwAEmDILZL8yhJ9KamHFaJqXBgo3My14d3wI75MMCLGnLYCmtlWvPXlflb85k05BDOZQ3Cjh4yQVhUO+/837CDBGHNjY+PmRfkH+TY/fpcZFrbLmxkNmzLHAiOhqQhjt+3sSIS4JynzCLFvz/uvuNu+t6cWel8pvuO6YhekyF7jfszhML1Mv6A1j1NSa4tHx/zIWzXIji41TNjEy1LZQV8fq2Zr3PhG5BiE/grZcrZdy82QYA386amHFYJqaZZgDOa6VgbcvT3soYcturK/GltsngJqd6TqRTNnzUIO3KwWbenBy8Mwioqq3hr4U5+3bjf00NpWXz94ZRbzWLLuxY7dt+qStj6swlKapY6ulrfiyH1cljwpOnc5Wpaw8bvoeOprmnD7wzWkkRZO6plsV0frDapl4OPn2TDRNNpDT/eCZt/MPNve046cZu+lwAKVn/s9uE5JDPdnGD0hqYcVtaAsKnrhVWUwar3TQWKN5dbJabCoW0nziHcuxwObIBBXrK4tGgZrCcprZkwCcKc69Ple3jou/WUlDejdUqagwFXQGgr+ONpx+63d4V5sndzUyliTWc/aRYr/vJ6KDrg2mPlbDLzC7yxFNEqMsmsH7X+a8fud+QQbJ0LZUdcMizRRJmroPxo3UFYeBvzYSz9I+9YzFw0XwueMhmKkXfA0Btq3yayLXQ8DVZ/ZNad9FZZ6ZDoZW3Prc05mjovbMtP3tuQw5a1FDS7RnMOacghXMGaCTu8HUryJAhzJj9fH+6f2JM9h4t5e5F0gHMq/2AYdqOZ3+VI56Yts005RKexLhtavQJCYcpM05jgq7+49gPBpu/NZbezXXcMZ+h1gVnUOmdL/dvl74Olr5slDJ7qDB9cCM8PgBXvuHeenWhY9XywOoIwMB/Gjh40GQwhGmPle/DbI9DvUhj7YP3bpl5mOifudrB6wl1K8k25pDfNB4NjzTkyVzdtP2kzzWLInT303msva6mh7eeK4jwzd7mPpeumEM4SbMl671tpLiUIc66RXeI4o0cbXpq3jQMFJZ4eTssy+DrTen3h/+y/z5Y5puticJTLhtWgNr1g/P/B9nmw+DnXHWfTD5A0GMLjXXcMZ+g5CVC1lyQe2g4Ln4U3xsL/esJPd5kM4si/wcXvQXR7+P52eGW4ebyN7Zjp7Za/BR9d2nyyRhkLTcfT0Ni6t+k0xnRJkzXDRGNsmWOWN+k0Fs57wQQL9ek+EQLCIf1D94zPUdamHN443ygx1ayNVd7IzzC5Geb9zlsbctgKa22ac9hm/tZ+ZjpsensWTzQ/vn4QFGWqtMD1yya5kFcGYQD3ntODssoqnpyz2dNDaVmCIk0gtuEb+yb45+02XRXd1RWxPgOvNhmgX/8Du5c6f//5e01JmDeXIlpFJEK74bD+KxNEZa+F3/4LLw+HFwbA3AfNxPCxD8DNy+GWZTD2fuh5HlwzB6Z+YO738WXw9ngzOb8lObgNZt9t1tdZ8qKnR9OwhuaDWflYJujv+O3Y+ihC2GNvGnw2HeL7wMWzzDzhhgSEmDmoG77xzjJma+YlwYuaclglpJqOwo1tzpH2Ligf727IYSsh9djfQ2tTbZHQz7sapoiWIzQOjlimp0S19+xYmsBrg7AOcaFcPaIDn6/cy9q99azELhw37CbwC4RFzza8rbWlvTcEYUrBuc+ZlvxfXHtsxXRn2fyTuew+0bn7dZXek80ctmf7wqsj4fcnTAfL8Y/B7Wvhhvlmoe5WXY+/n1LQYyLc9KdZZy53J7x1JnwyzQQvzZ3WJtPnHwwdTze/F29f3Lqh+WC2+k8zH85Wvuf6cYmW4dB2+HCKyVhc/plj6zWlXgZlRWbxZm+TlW4yw/Vljz2lukRvpeP3rSw/1pAjsq1Th+UyialwaKtpzrF3hTl5O1AacggXsc4LC4n13iZqdvDaIAzgljGdiQkJ4N/fr0e31JIpTwhrBf2vgNWfmOxPfbbMhphOENfZPWNrSFAkXPQ2FGbDt7c6t5Ru0/cQ19U0AWkOel1g/jatuprg9M4tcPWPZt5fVLuG7+/rZ9aZu3UlnHYPbP8NXhoC3//N9Q1QXGn1R2Z+1RkPw6SXTPboh797d9nlzgXmsub6YLWJTDKdStM/kHl9onblJSbw2vmH6W74/mRz/bQvTSDmiHbDzZyL9A+cPswmy0z3zlJEMMFhcEzjmnNs/smc5W9OpXy2zTnSZoJ/qJkPJoQrWIOwZjwfDMDN/cYdExHkz53juvGvL9fyw9osJvb14hatzc2I20x3rMUvwoTHat+mtMh8OBxSR/csT2k7EM54CH6+F5a9DkP/0vR9FueaOTmn3Nr0fblLaBzc1oizrDUFhsFp/zRthH9/wjwvVn9sniPDb3HsrLmnHTkEc+6F5GEw4CqzvtaY+2H2P838OW/t0mXPfDBbA68ypaRbf4buXt5ERjhXZTkUZELBPtN4p2Cv5TLz2PdHDx5/n8BIuOIriO3k+PGUgn6Xwfz/M+Xp9pzgcYeSfNMdLfUyT4+kdkqZALExzTnSZpo5Vp3PcPaoXMcaDO9cAOu+MMvLSEMO4SotJAjz6kwYwMWDkumREMH//bhJWtY7U1Q76DPFrDl05FDt2+yYD5Vl0HWcW4dml+E3m1KNn+9zrNNjXbb+Yur3m0spoiuEtTaLY9+8DLqcYT50vTCgeZUo/nwflBbAuc+aAAxgyPXmLO1Pd5uOXd6mosys32dPKaJVl3EQFi9rhp2M1n0Jz/WFdybAl9fB3IdgzaemrDi0tSk1Pv0+OP8VuPIbuCXNZMmTBjb+mP0uAbSpnvAW3tyUwyoh1fHmHM2pIYetsNYQnmhO7EpDDuFqEoS5h6+P4sFze7Ivr5g3Fuzw9HBalhG3m3koS1+t/fYts80Z1HbD3TosuygFk16GkDj4/GooLWza/jZ9bz7UJg5wzvias9hOZuL+tXOhohS+ucks2O3tdvwOqz+EETOgdY9j1/v4mnLNowfh14c9N766ODIfzMrXD/pfbjJh+ftcNzbhfdoNNZ0Np30JNy2Ff+2Ff+2Gm5bAtM/Nc/3Uu0yGqONpppTcP6hpx4xuDymjTKmvt5T1enNTDqvEVMebc6x8z7y/NZeGHLYSU6H8CMT3lYYcwrWsCzY38yDMq8sRrYZ1jGVC73henr+dKYOSiY9s4huKMFp3N5mfZa+ZMjzbyY1VVaYpR+ex9nXR8oTQWLjoLZh5Dnx/B0x+o+GWy7UpLzGLGPebeix7IiB5MEx4Ar66wQTqw2/29IjqVl5ingPRHWD0XSfenpgKQ2+EP18y6yMlD3H7EOtkXR/MkSAMzLzOP542c3VO/YfzxyW8U3SKZz549LvUnJDZs8wEgp7mzU05rKzzpL6bYRZbt8ee5dDlLDP3s7lJSDUdaQdOb9x7sRD2kkyYe91zdg8qteaJ2Zs8PZSWZdTfTG192jvHX5+1ykwM9oauiPVpf4ppKrH2M9NNqjF2/m7O3jWH1vTu1vdi8xz49d9mor+3+uNpMz9k4v9MV8TanH6PWfj0uxlmXo23yFgIbXpDSIxj94vpYDIdK99z7QLmQoBZ3sI/xHsadGSmmxbo3iyqHfS9BPwCzLxje75ad4dRd3p65I3Te7JpGNX3Yk+PRLR0HUZDr8nNvnqpWWTCAJJjQrhuZAdenr+dK4a3p3+7aE8PqWVoOxA6nApLXoIhfzlWtrJljmmD3eVMz47PHqP+BhkL4Me7zELLrbs7dv9N35sFrFNGu2Z8zZlSMPFZeHkofHMzTP/R+7KFOZvN4uN9p0Kn0+veLjAMzn4SPr7UPN9H3u62IdbJOh9swJWNu/+Aq0w57o55zWsSv2h+AsPNIvHrv4IJj9d9ssMdqptyXOq5MdhDKZj8mqdH4T5xXWDKTE+PQpwMIpNgyjsNb+flGvw0pZR6Wyl1QCm1zh0Dqs9Np3emVXgg//5+g7Ssd6ZRf4ei/cef4dz8EyQPdfzsvCf4+JpSxIBQsxhp2VH771tVaR5rlzPN2UpxoogEs/bY7iWmG6U3qaqC7243AdZZjza8ffezTQnu/MeavthxbkbT1x/LXOn4fDBb3c8xZRlp0qBDuEG/S03jm00/eHYcWWvMpTfPBxNCiAbYc0p7JuAVNWlhgX7cNa4bq3bn8U16ZqP3s+fwUfKLvagcydM6jDYZsUXPmXWHCjLNWh/e2BWxLuHxMPl104nqzTPg65vg9ydh7eewN80s7Fxb4L53ORzJgW7S5rte/S418xTmPuRdZYnp78PuxXDmf8z6d/aY8ETT1g7TGpa+Bi8OgbfHQ9kRx/dhZZ0PZs/6YLXxCzR/m80/Nu+13UTzkDLKzMNK/9Cz47CuveXNnRGFEKIBDZYjaq0XKKVS3DAWu1w0IIn3luzisZ82cVavNoQE2F9RuSOniCfnbOanddn4KOibFMWoLnGM6BzHgHbRBPh5WZmVuygFI/8Gn1xu1lKyfqj09vlgNXUeazqDrf7EtPgtzDr+9sBIiEkxzRtiOpjLjD/Ax795lF16klLmd/vSMLNI9lXfe74ssSgHfr7fBDCOdBKLbAtj7oPZd5vSqt6THTvmNzeZroTJw2DPn7D4BTjtbsfHD42fD2ZrwFWw5EXzwdgbSixFy+XjY9rV//E0FGSZLLknZKab+Z3WDmlCCNEMNZs5YVY+PooHzu3JlFeX8OrvO/jbmV0bvE9OYSnP/bqFj5btIdDPh1vHdEYBC7cd5OX523lh3jaC/X0Z2jGGkZ3jGNWlFV3bhKFOpu4+3c6GVt3N3JrIZIhqb35ubgZOP7Y+SdlRyNtlSsZyd5rLwztMlm/T96Z1MEDnMyEo0lMjbj4iEmHco/DtLbDiLbP+lifNucecMJj4rOOduIbcYBaknn03dBoDwVEN32frXPj6RjMfZcKT5vF/Nt1kkAdc5fgH0ooy2L3ULLzcFK26QrtTYOUs057/ZHrdEu7X71JY8CSs+cRzQX9WumTBhBDNntOCMKXUDcANAO3atXPWbms1OCWGiX0TeO337UwdnEzbqNonCBeVVvDGgh288ccOyiqquGxIO24b24VW4YEA/O2sbhSUlLNk+yEWbTvIwq0HeWTzRmAjrcIDGdnZZMlGdo5r+W3xfXxg5B3w1V/gwAbTpKO5f5gLCDHrRdmuGWVVWQEFe828ntY93T60Zqv/NJM9+uVB0wgipoNnxrF9Hqz9FE79pwlCHOXjaxZ0fmOM6fw48Zm6ty0vMeuL/fmyea5c+Q20sTxnznjIlAL+9ihMetGxMWSuNIuapoxyfPw1DbzK/O8uetZkCHSlme943GXVide3PwXaDWv68cXJI7aTyQKnf9i4oL8k3/xP2duy/YT7F8ChbabroBBCNGNOC8K01q8DrwMMGjTI5V0z/nV2D37ZsJ/HftrEC5cePzm3vLKKj5ft5rlft3KwqIyz+8Rz17judIgLPWE/EUH+jOsVz7he8QDsyytm0daD/LHtIAu25PDVKrMQav92UVzQvy0T+yYSE9pCGzj0vtB8mMzb3bzmgzWGr5/n1ttpzpSC854/VpZ45bfuL0ssLzZrgsV2NmW0jZXYH4b+Ff58xZRY1bZ22IFN8MW1sH+dOTFx5sPHd4WL6WCyakteMvuK723/8TP+AJQJhJqq5ySYc6+Zs+eI0++VIEw4LvVSs9RD5kozn9geWpvFnufca06C3LbKdFx0VLalKYcsBiyEaOaaXTmiVduoYP4yuiPPz9vGVcPbMyglBq01P63L5sk5m9l58AhDOsTwxpXdHWpn3zYqmIsHJ3Px4GSqqjQbswuYvzmHb9MzeeCb9fz7uw2c1q0V5/dvyxk92hDk7+vCR+lmvv4w9kFY/Hzju7WJli8yyZQlfncbpL0Ng69z7/F/f8JkMK/67tiSCo11+j2w4RvTYfEvvx9bmFxrU3I5514ICIPLPoOuZ9W+j9F3ms6iP98HV3xlf2bAGfPBrPyD4eZlpsmMj69ZXsLHF5Rvjcsa1/t46ULswrv1ugB++qfJhtkThB3cBt/fbk48xPc1gdSSlxo3lzJzlbmUckQhRDPXYBCmlPoIOA2IU0rtBR7UWr/l6oHZ46+ndeLTFXt5+LsN3HtODx6fvYlVu/Po2iaMt6cP4vRurZs0r8vHR9ErMZJeiZHcfHpnNmYV8PWqfXyTnsncjQcIC/RjQu94LujflqEdY/H1aeble0BVrwup6DH55G1SIuwz4EqbssQzIbq9Y/fP223mlWz/zWSTWve0lI72NHMRgyJqv9/+DeYkQerlpqtnUwWGW9YOu8yUG46YAUcOmizf5h9NyeWkl+svnQqONmWRs++GbXPta/JSUWqZDza96Y/BKjTWfAmnUUqNB54DfIE3tdaP1bj9b8B1QAWQA1yjtd7l9oG6W1CkWeph7ecw7r+mS2dtKkph4bPwx1PgF2zmbw64Cj6fbhraDL7O8eYa0pRDCNFCKFestzVo0CC9YsUKp++3Nl+t2ssdn6wGID4iiL+d2ZULBya5NCCqrNIs3XGIr1btY/a6bApLK4iPCGJSaiLn929Lj4Q6PkB6scoqzU/rsnjh123sPHiE60Z14JYxnR3qPilOMnl74OXh0HaAmSdlzwmP/L2ms9rK98z2Xc4yXSwPbIJym1bvkcnH5vNZA7TYLvDe+XBwK9yywrkBx8eXw7Zf4ZynzByx4lw489+mBNGecsuKMrOgtW8A/HWRKXetz64l8M54mPoB9JjonMfgIUqpNK31IE+Pw9mUUr7AFuBMYC+wHLhUa73BZpvTgaVa66NKqRuB07TWU+vbrzvfH11q21x4/0K4eJYph60pY5HJfh3cYkrdx/3fsZMZB7fCS0NNKe+Ex068b31eGGhO1FzyQcPbCiGEh9X3HtnsP2FP6teW1XvyaRMRxNUjUtxSHujrozilcxyndI7jP+f35teNB/hq1T7eWriT1xbsoEvrMNrHhhIR7EdEkD8Rwf5EBvsTEeRHRLC/5TpzW2SIP2EBfvh4KItWWaX5fk0mL8zbxrYDRXRuHcaZvdrw8vztfLVqH/ec3YOJfRNOrk6Rwj5RyXDWv838rLSZMOjqurctyIQ/noGV75pSvwFXwqi/mdJGME0j8nfDgY2mMcyBjeZrx3yoLDt+X+e/6vyMz4THzYfCb242H/CmfenY/C6/ADjjYfj0Clj1Xv2/CzCliM6aDyZcZQiwTWu9A0Ap9TEwCagOwrTWv9ls/yfgwFoJzVzH0yE8wZQk2gZhRw/DLw+Y/4OodnD55ydmh+O6QP/LTcnvsBvtz6RLUw4hRAvS7IMwHx/FQ+f18tjxg/x9OadvAuf0TeDwkTJ+WJvFz+uzycwrZlN2OfnF5RSWVNS7jwBfH0Z3jePcfomc0aMNoYGu/7NUVFbx7epMXpy3jR0Hj9CtTTgvXtafCb0T8PVRXDPiMA98s55bP1rFB0t38fB5vekW34hJ1KJlG3i1KUv8+X5TuheVfPzthfvNsgcr3jYd+VIvN3Ooomp0UPXxOdYopduEY9dXVphlBayBmY+faaLhbJFJcMGrkLUaRv39+OYb9upxLrQbbprb9Lmo/qYDGX84bz6YcJW2wB6bn/cCQ+vZ/lrgp9pucGf3YLfx8YW+U01ZYdEBCG0Faz41S0cU58KI202ZbkBI7fc/9W6zpuP8x+CCV+w7ZnVTjlRnPAIhhPCoZl+O2BxUVmmKSisoKC6noKScguIK8qu/L2dvbjGz12WTXVBCsL8vY3u05rx+iZzarRWBfs7N7JVXVvH1qn289Ns2Mg4dpXt8ODPGdmFcr/gTsnGVVZqPlu3myTmbKSqt4KrhKdx+ZhcigmQyv7CRu8uUJbYbajJISplFjRc9C8vfMpms1Eth9F0tvxvl3jR4cwyMuhPG3l/7NhWl8Fh7ky0b/3/uHZ8LtOByxIuA8Vrr6yw/XwEM1VrfUsu204BbgFO11qX17bdFvT/mbIaXhsCwm8yJkh3zIWmwmftlTyb55/tg8Ytw4+Jjyz7UZ/GL8PO9cOc2CGvV1NELIYTLtehyxObA10cRaSlJrMsDE3uyPOMw363J5Me12Xy/JovwID/G94rnvNREhneMxc+38c0yyiqq+GrVXl76bTu7Dx+ld9sIXr9iIGf0aFNnKaSvj2LasPac3SeBp37ezDuLd/Lt6n3cPaEHk/u3daiE8mBRKem781i1JxcfpbhkSLs613cTzUx0e1OW+MPfzVnxowdh2RtQUWLKhkbfadYWOhkkDYTeF8GSF2HQNRDZ9sRt9lnXB5MOpF5uH2Cb2k2yXHccpdQZwL3YEYC1OK26me6If74MgRFwztMw8Br7l60Y+TdImwXz/gOXftTw9lnpENFWAjAhRIsgmTAvVF5ZxaJtB/l2dSY/r99PUWkFcWEBnN0ngfP6JTKgXfRxAZDWmrLKKkorqigtr6K0ovK479dlFvDq/O3syyumX1Ikt43twpjujneOXLs3nwe+Xceq3XkMaBfFvyf1pnfbyBO2K62oZH1mgSXoyiN9Ty57DhcDJrDTWqOUYkLveK4b1ZHU5Kgm/b6EF6iqglnnHVv7qs8UU4oU19nTI3O/3F3w4mDoPdmUONb0+5OmZPEfO1pEOWILzoT5YRpzjMUEX8uBy7TW62226Q98jsmYbbVnvy3u/XHH77DxO1PGG5Hg+P0XPGWCsGt+Ntn0+rwwCOK6wqUfNm6sQgjhZvW9R0oQ5uVKyiuZv/kA367O5NeNByitqCImNAA/H2UCLUvA1dCfsX+7KGaM7cKpXVs1qclGVZXmi5V7eXz2Jg4dKePSIe24Ylh7tuwvZJUl6NqYWUBZZRUACZFBpCZH0b9dFKnJ0fRpG8mhI6W8uziDj5ftobC0goHto7luZAfO6hXfItr8n7Ty98Gy1yH1MnOG/GT2ywOw6Dm44fcT56+8ex4UH4a/LvTI0JytpQZhAEqps4FnMS3q39ZaP6qU+jewQmv9rVJqLtAHyLLcZbfW+rz69invjzWUHYHnUs3i61f/WHeX1dJC+L9ks7bfqf9w6xCFEKKxJAhrIQpLypm7cT9Lth/C10cR6OdLoJ+P+fL3Pf7Sz8fc7u9DXGggvdtGOLXDYX5xOc/O3cKsJbuorDLPoWB/X/okRdK/XRT9k03QFR9Z92K6RaUVfLp8D+8s3smew8UkxwQz/ZQOXDwoiXCZdyaas5J8eL6/aa9/1XfHPlhWlMJj7UypYguYDwYtOwhzBXl/rMWyN+DHO2vvpGiVsQhmnl3/wulCCOFlJAgTLrNlfyHpe/LolRhBtzbhjZq3Vlml+WVDNm/+sZMVu3IJD/TjkiHJXHVKCknRdXTWEsLbWT9YXvrxsY6PuxbDOxPgkg+h+zmeHZ+TSBDmGHl/rEVFGbw0GALC4S8Lap9TtuQl03nxzq0Q1tr9YxRCiEaQxhzCZbq2Cadrm6a1rvf1UYzvncD43gmk78njrYU7eXtRBm8vymB873jO7ZtI64hA4kIDiQ0LICTAV9YtE95v4HRY+tqx9v2+/rI+mBC18QuAMffDF9fCui+g75QTt8lMh/BECcCEEC2GBGHCq6QmR/HCpf25e0J3Zi3O4MNlu/lhTdZx2wT5+xBrCchiQwOIDTPf2wZpYH+Q1rVNGB1bhTn5kYiTnq8/nPlv+PhSs5j1kOtN45L4PhAc7enRCeFdek2Ghc/Cb4+YxZ/9Ao6/PXMVJPb3yNCEEMIVJAgTXqltVDD/OrsHM87owrYDRRwqKuPQkTIOFZVy6EgZB4tKOVRURk5RKZuyCzlUVFbdDKQxBrSLYsqgZCb2TZD5aMJ5uk2A9iNh/v9Bz/NhzzIYdK2nRyWE9/HxgTMehA8ugpXvmpMWVqWFcGgb9L3Yc+MTQggnkyBMeLWQAD/6JkU1uJ3WmsLSCg4VlVFcVmn3/qu0ZtG2g3yWtpd/fbmWh79bz4TeCUwZmMSwjrEOrYUmxAmUgnGPwOunwWdXmbXTZH0wIWrX+QxoPwJ+fwL6XQqBlgqFrDWAhoRUT45OCCGcSoIw0SIopYgI8ieiEVms3m0juWF0R9L35PFZ2l6+W53JV6v2kRQdzIUDkrhoYBLJMd7ZIKSkvJLKKo1SoFDHdXe2vU5hfkc+CplP526J/c2i1Ws+xswHG+7pEQnhnZSCMx6Ct86Epa/A6LvM9Vnp5rLmcg9CCNGMSRAmBCYw6d8umv7tonlgYk/mrM/msxV7eX7eVp77dSvDO8YyZVASE3onEBzg65YxVVVpcopK2ZdXTGb1V8lxP+ceLXdon74+ioHtoxnbvTVje7SmU6swCcrcYez9sOFriOsi88GEqE/yEOh2Dix63pTuhsRIUw4hRIskLeqFqMe+vGK+TNvL5yv3suvQUcIC/ejcOoxgf1+CA3wJ9vclyN+X4ACf6u+D/H2rbw/y96GqCsorqyirrKKswlyWV2jKKispq6iivFJTWlFFeWUVpRVVHCgoITO/mOz8Esorj///DA/0o210MIlRwSRGBREfEYS/rw8a0Bo0Znvrv7XW2nK9ue5IWQV/bD3IxqwCANrFhDDGEpAN6RBDoJ97Asz6VFVpDh4pJSuvhMy8YkvQab4/UlZBYmQwSdHBJMUEkxQdQlJ0MK3Dg7x/oe/NsyEoosV1RpQW9Y6R90c7HNgILw+HU26Bsx6BFwebxZwv/cjTIxNCCIdIi3ohGqltVDC3ju3CLWM6s2znYb5O38e+vBJKyirJKSyluLyS4rJKSsorzfflldh7XsPXRxHg64O/ryLAz5cAX0WAnw+twgMZ0C7aEmgF0zYqqPr7xpRb1iYzr5h5mw4wb9MBPlq2m5mLMwgN8GVUl1aM6dGa07u1plV4YK33LSwpJyvfZOSsgVJmvsnMHT5SRoCfDyH+fgQF+BLs72MTsPodF6wGB/iiUGTnF7PPZj9ZeSUnNFkJCfAlMSqY0ABfNmYVcrCo9Ljb/X0ViVGW4Cwq5LggLTk6hNbhgU6Z36e15vCRMjIOHSHj4FGOlFUQ7O9LSIAfIQG+li8/gqu/N48zwNcH1W18k48vxEmhdQ8zJ2zp69D/Cji4FXpf5OlRCSGEU0kmTAgn0lpTVllFSVlVdVDmoyDAzwd/Xx8C/HwsgZeP12RuissqWbLjIL9uNEFZVn4JAP2SoxjWMYaikgoy84qrA6/Ckorj7u/ro2gTHkhiVDBxYYGUVVZRXGYeuzU4PVpWSUlZJUctc9hs+SiIjzCBZoIlw9c2KpjESGsQGkxEsN9xZZMl5ZXsyytmb24xe3OPWi6PfZ9TeHyQFuDnQ1JUMEkxISRHB5McY4Kz5JhgkqNDiArxr97/sUDrKBkHj5iAy+b7mo/fHn4+iuAAX+IjgugaH073NuF0izdfydEhbmsAU1WlyS8u5/DRMg4fKSM+IqjJ8x0lE+YYeX+0U+4ueHEQxHWF/evgsk+h6zhPj0oIIRwimTAh3OT/27vb2CrvMo7j3x+nzw9ACx1jPIs4gkYnITPGxSyYmc03bIlZRmIyX80XLs7MFy6+cZqYGKNGXywYjCQzUXFxU3mniy7ZfIOwydgDDyJhaQkUtkIsLVDaXr44d1lBTmnLOec+/O/fJ2l6zn3a0+vqP+d/9er/vv9HEq1NJVqbSizi9tjqvr2lxNaNy9i6cRkRwaFTw/z98CB/O3yGna8eZ3F7c7bK1MFn1vVebZZWLG5j+aJ27uhupam0YNY/78rE5NUVxMkI+rrm9v0Abc0l1vd1sb7C+7tNNWn9Q6P0n7vIwNAo/edG6R+6yMGB85y/7lq6rtYmVva001xa8H+N1gLByp4O1izp4JHVK1izpJN1SztYs6SThW3NXBybYPTKOKNj5ZxGLo9fbTzLx8YZyR4bODfKwYHz17z3XUdLiQ3Lurl7WRd337mQjXeW3wD9RiuR4xOTXBqfvGb19VL2u7x4ZaLcYI2MXfPxwcgY57Lb50bHmN4DP/3Ax/jGFzbM6XdvVhc9a8rXhO3dUb7vnRHNLDFeCTOziiYmo2FW7Kpp+NIV+ocuZo1ZefWsf2iUsYlJ1i7pZO3SDxutVT0dtDTNrUm8mQuXxzk6OMzR08McPj3M0cFhjpwe5oORsatfs6Szha62pmtWFa+/RrASCXo6WujtbKE3+9zTWX5z895pH+vv6GLF4vZbysUrYXPj+jgHI+/Dzz8Frd3wrcN5R2NmNmdeCTOzeUmxAQPobmtm013NbLprYS4/v6u1ic2re9i8+tqdEs8OX+boYNaYnR7m8vhEtsHLtA1fmssbvkxdV9fW9OEmMIvam+ntbGVRe3OyY2cF0rkUtj0HVy7mHYmZWdW5CTMzaxB93a30dbfyuY8uzTsUs8bw8YfzjsDMrCZmdY6NpAclHZF0TNIztQ7KzMzMzMwsVTdtwiSVgOeAh4BNwHZJm2odmJmZmZmZWYpmsxJ2L3AsIo5HxBiwG9hW27DMzMzMzMzSNJsmbAXQP+3+QHbsGpKekLRf0v6zZ89WKz4zMzMzM7OkVG3f5YjYGRFbImJLX19ftZ7WzMzMzMwsKbNpwk4Cq6bdX5kdMzMzMzMzszmaTRO2D9ggaZ2kFuAxYE9twzIzMzMzM0vTTd8nLCLGJT0J/AUoAbsi4p2aR2ZmZmZmZpYgRUT1n1Q6C7x3i0+zFHi/CuHcLoqUb5FyhWLlW6RcwfkCrIkIXwg8S66P8+J801WkXMH5pqxSrhVrZE2asGqQtD8ituQdR70UKd8i5QrFyrdIuYLztXwUbRycb7qKlCs435TNJ9eq7Y5oZmZmZmZmN+cmzMzMzMzMrI4auQnbmXcAdVakfIuUKxQr3yLlCs7X8lG0cXC+6SpSruB8UzbnXBv2mjAzMzMzM7MUNfJKmJmZmZmZWXLchJmZmZmZmdVRwzVhkh6UdETSMUnP5B1PrUk6IektSQck7c87nmqTtEvSGUlvTzvWK+llSf/OPvfkGWM1Vcj3WUknszE+IOlLecZYLZJWSXpF0ruS3pH0VHY8ufGdIddUx7ZN0j8lvZnl+73s+DpJe7P5+feSWvKOtWhcI9Ph+pj0HFqY+giukfOtkQ11TZikEnAUeAAYAPYB2yPi3VwDqyFJJ4AtEZHkm9lJ+jxwAfh1RHwiO/YjYCgifpj9EdETEd/OM85qqZDvs8CFiPhxnrFVm6TlwPKIeENSN/A68DDwVRIb3xlyfZQ0x1ZAZ0RckNQM/AN4CngaeCkidkv6BfBmROzIM9YicY1Mi+uj6yOJjK9r5PxqZKOthN0LHIuI4xExBuwGtuUck92CiHgVGLru8Dbg+ez285RfqEmokG+SIuJURLyR3R4GDgErSHB8Z8g1SVF2IbvbnH0EsBX4Q3Y8ibG9zbhGJsT1MV1Fqo/gGsk8a2SjNWErgP5p9wdIeBAzAfxV0uuSnsg7mDpZFhGnstungWV5BlMnT0o6mJ2OkcTpB9NJWgt8GthL4uN7Xa6Q6NhKKkk6AJwBXgb+A5yPiPHsS4owPzca18j0JT1/VpDkHDqlSPURXCPnUiMbrQkrovsiYjPwEPD1bLm+MKJ8PmzjnBNbGzuA9cA9wCngJ7lGU2WSuoAXgW9GxH+nP5ba+N4g12THNiImIuIeYCXlFZiN+UZkBVXYGpna/FlBsnMoFKs+gmvkXJ+j0Zqwk8CqafdXZseSFREns89ngD9SHsjUDWbnD0+dR3wm53hqKiIGsxfrJPBLEhrj7FzoF4HfRMRL2eEkx/dGuaY8tlMi4jzwCvBZYLGkpuyh5OfnBuQameBr7DpJzp+VpDyHFqk+gmsk86iRjdaE7QM2ZLuLtACPAXtyjqlmJHVmFzAiqRP4IvD2zN+VhD3A49ntx4E/5xhLzU1NuJlHSGSMswtTfwUcioifTnsoufGtlGvCY9snaXF2u53yRhCHKBeaL2dflsTY3mZcIxN5jc0guflzJgnPoYWpj+AayTxrZEPtjgiQbV/5M6AE7IqIH+QbUe1I+gjl/+wBNAG/TS1fSb8D7geWAoPAd4E/AS8Aq4H3gEcjIomLdSvkez/lpfgATgBfm3ZO+G1L0n3Aa8BbwGR2+DuUzwNPanxnyHU7aY7tJylfVFyi/M+6FyLi+9mctRvoBf4FfCUiLucXafG4RqaTr+uj6yOJjK9r5PxqZMM1YWZmZmZmZilrtNMRzczMzMzMkuYmzMzMzMzMrI7chJmZmZmZmdWRmzAzMzMzM7M6chNmZmZmZmZWR27CzMzMzMzM6shNmJmZmZmZWR39D8JXKjia3BWCAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 1080x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Max. Training Accuracy 0.9518716335296631\n",
            "Max. Validaiton Accuracy 0.875\n"
          ]
        }
      ],
      "source": [
        "plotModelHistory(model_2_history)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "m_-TyE8YDCnP",
        "papermill": {
          "duration": 0.227485,
          "end_time": "2020-12-30T11:49:38.532430",
          "exception": false,
          "start_time": "2020-12-30T11:49:38.304945",
          "status": "completed"
        },
        "tags": []
      },
      "source": [
        "Reduce parameters, with padding in Conv3D layers and filter of (3,3,3)."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "n_Fqznw7DwLr",
        "papermill": {
          "duration": 0.223765,
          "end_time": "2020-12-30T11:49:38.994860",
          "exception": false,
          "start_time": "2020-12-30T11:49:38.771095",
          "status": "completed"
        },
        "tags": []
      },
      "source": [
        "### Model 3: Frames-30, Epocs-20, Reduced parameters"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "UlmHHFrL8HYu",
        "outputId": "387218c1-3a45-43fd-f53a-087ce6f7856b",
        "papermill": {
          "duration": 1345.022772,
          "end_time": "2020-12-30T12:12:04.236613",
          "exception": false,
          "start_time": "2020-12-30T11:49:39.213841",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Model: \"conv_3d3\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " conv3d_12 (Conv3D)          (None, 30, 100, 100, 16)  1312      \n",
            "                                                                 \n",
            " max_pooling3d_12 (MaxPoolin  (None, 15, 50, 50, 16)   0         \n",
            " g3D)                                                            \n",
            "                                                                 \n",
            " batch_normalization_20 (Bat  (None, 15, 50, 50, 16)   64        \n",
            " chNormalization)                                                \n",
            "                                                                 \n",
            " conv3d_13 (Conv3D)          (None, 15, 50, 50, 32)    13856     \n",
            "                                                                 \n",
            " max_pooling3d_13 (MaxPoolin  (None, 8, 25, 25, 32)    0         \n",
            " g3D)                                                            \n",
            "                                                                 \n",
            " batch_normalization_21 (Bat  (None, 8, 25, 25, 32)    128       \n",
            " chNormalization)                                                \n",
            "                                                                 \n",
            " conv3d_14 (Conv3D)          (None, 8, 25, 25, 64)     55360     \n",
            "                                                                 \n",
            " max_pooling3d_14 (MaxPoolin  (None, 4, 13, 13, 64)    0         \n",
            " g3D)                                                            \n",
            "                                                                 \n",
            " batch_normalization_22 (Bat  (None, 4, 13, 13, 64)    256       \n",
            " chNormalization)                                                \n",
            "                                                                 \n",
            " flatten_4 (Flatten)         (None, 43264)             0         \n",
            "                                                                 \n",
            " dense_12 (Dense)            (None, 128)               5537920   \n",
            "                                                                 \n",
            " batch_normalization_23 (Bat  (None, 128)              512       \n",
            " chNormalization)                                                \n",
            "                                                                 \n",
            " dropout_8 (Dropout)         (None, 128)               0         \n",
            "                                                                 \n",
            " dense_13 (Dense)            (None, 64)                8256      \n",
            "                                                                 \n",
            " batch_normalization_24 (Bat  (None, 64)               256       \n",
            " chNormalization)                                                \n",
            "                                                                 \n",
            " dropout_9 (Dropout)         (None, 64)                0         \n",
            "                                                                 \n",
            " dense_14 (Dense)            (None, 5)                 325       \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 5,618,245\n",
            "Trainable params: 5,617,637\n",
            "Non-trainable params: 608\n",
            "_________________________________________________________________\n",
            "None\n",
            "Source path =  datasets/Project_data/train ; batch size = 64\n",
            "Epoch 1/20\n",
            "11/11 [==============================] - ETA: 0s - loss: 1.2683 - categorical_accuracy: 0.5234 Source path =  datasets/Project_data/val ; batch size = 64\n",
            "11/11 [==============================] - 213s 21s/step - loss: 1.2683 - categorical_accuracy: 0.5234 - val_loss: 1.5636 - val_categorical_accuracy: 0.3300 - lr: 0.0010\n",
            "Epoch 2/20\n",
            "11/11 [==============================] - 82s 8s/step - loss: 0.6673 - categorical_accuracy: 0.7628 - val_loss: 1.8763 - val_categorical_accuracy: 0.4531 - lr: 0.0010\n",
            "Epoch 3/20\n",
            "11/11 [==============================] - 82s 8s/step - loss: 0.6990 - categorical_accuracy: 0.7431 - val_loss: 3.2086 - val_categorical_accuracy: 0.0893 - lr: 0.0010\n",
            "Epoch 4/20\n",
            "11/11 [==============================] - 66s 7s/step - loss: 0.5545 - categorical_accuracy: 0.7725 - val_loss: 3.2881 - val_categorical_accuracy: 0.2188 - lr: 0.0010\n",
            "Epoch 5/20\n",
            "11/11 [==============================] - ETA: 0s - loss: 0.4703 - categorical_accuracy: 0.8182\n",
            "Epoch 00005: ReduceLROnPlateau reducing learning rate to 0.00020000000949949026.\n",
            "11/11 [==============================] - 63s 6s/step - loss: 0.4703 - categorical_accuracy: 0.8182 - val_loss: 3.4089 - val_categorical_accuracy: 0.1562 - lr: 0.0010\n",
            "Epoch 6/20\n",
            "11/11 [==============================] - 52s 5s/step - loss: 0.4963 - categorical_accuracy: 0.8373 - val_loss: 3.6529 - val_categorical_accuracy: 0.1250 - lr: 2.0000e-04\n",
            "Epoch 7/20\n",
            "11/11 [==============================] - 54s 5s/step - loss: 0.4394 - categorical_accuracy: 0.8621 - val_loss: 3.8376 - val_categorical_accuracy: 0.1250 - lr: 2.0000e-04\n",
            "Epoch 8/20\n",
            "11/11 [==============================] - 48s 5s/step - loss: 0.2864 - categorical_accuracy: 0.8877 - val_loss: 4.4846 - val_categorical_accuracy: 0.1250 - lr: 2.0000e-04\n",
            "Epoch 9/20\n",
            "11/11 [==============================] - ETA: 0s - loss: 0.3144 - categorical_accuracy: 0.9037\n",
            "Epoch 00009: ReduceLROnPlateau reducing learning rate to 4.0000001899898055e-05.\n",
            "11/11 [==============================] - 46s 5s/step - loss: 0.3144 - categorical_accuracy: 0.9037 - val_loss: 6.0127 - val_categorical_accuracy: 0.0000e+00 - lr: 2.0000e-04\n",
            "Epoch 10/20\n",
            "11/11 [==============================] - 45s 4s/step - loss: 0.2506 - categorical_accuracy: 0.9144 - val_loss: 4.9484 - val_categorical_accuracy: 0.2500 - lr: 4.0000e-05\n",
            "Epoch 11/20\n",
            "11/11 [==============================] - 48s 5s/step - loss: 0.2258 - categorical_accuracy: 0.9519 - val_loss: 4.2175 - val_categorical_accuracy: 0.1250 - lr: 4.0000e-05\n",
            "Epoch 12/20\n",
            "11/11 [==============================] - 47s 5s/step - loss: 0.2346 - categorical_accuracy: 0.9144 - val_loss: 4.1745 - val_categorical_accuracy: 0.2500 - lr: 4.0000e-05\n",
            "Epoch 13/20\n",
            "11/11 [==============================] - ETA: 0s - loss: 0.3064 - categorical_accuracy: 0.8984\n",
            "Epoch 00013: ReduceLROnPlateau reducing learning rate to 8.000000525498762e-06.\n",
            "11/11 [==============================] - 49s 5s/step - loss: 0.3064 - categorical_accuracy: 0.8984 - val_loss: 4.5713 - val_categorical_accuracy: 0.2500 - lr: 4.0000e-05\n",
            "Epoch 14/20\n",
            "11/11 [==============================] - 43s 4s/step - loss: 0.2087 - categorical_accuracy: 0.9305 - val_loss: 4.0297 - val_categorical_accuracy: 0.2500 - lr: 8.0000e-06\n",
            "Epoch 15/20\n",
            "11/11 [==============================] - 46s 4s/step - loss: 0.2320 - categorical_accuracy: 0.9412 - val_loss: 5.4352 - val_categorical_accuracy: 0.1250 - lr: 8.0000e-06\n",
            "Epoch 16/20\n",
            "11/11 [==============================] - 44s 4s/step - loss: 0.2038 - categorical_accuracy: 0.9358 - val_loss: 3.6151 - val_categorical_accuracy: 0.2500 - lr: 8.0000e-06\n",
            "Epoch 17/20\n",
            "11/11 [==============================] - ETA: 0s - loss: 0.2926 - categorical_accuracy: 0.9091\n",
            "Epoch 00017: ReduceLROnPlateau reducing learning rate to 1.6000001778593287e-06.\n",
            "11/11 [==============================] - 46s 5s/step - loss: 0.2926 - categorical_accuracy: 0.9091 - val_loss: 3.8433 - val_categorical_accuracy: 0.2500 - lr: 8.0000e-06\n",
            "Epoch 18/20\n",
            "11/11 [==============================] - 44s 4s/step - loss: 0.2062 - categorical_accuracy: 0.9412 - val_loss: 3.8465 - val_categorical_accuracy: 0.3750 - lr: 1.6000e-06\n",
            "Epoch 19/20\n",
            "11/11 [==============================] - 48s 5s/step - loss: 0.2579 - categorical_accuracy: 0.9305 - val_loss: 4.6961 - val_categorical_accuracy: 0.1250 - lr: 1.6000e-06\n",
            "Epoch 20/20\n",
            "11/11 [==============================] - 45s 4s/step - loss: 0.1815 - categorical_accuracy: 0.9626 - val_loss: 4.2136 - val_categorical_accuracy: 0.1250 - lr: 1.6000e-06\n"
          ]
        }
      ],
      "source": [
        "no_of_frames = 30\n",
        "no_of_epochs = 20\n",
        "size_of_a_batch = 64\n",
        "\n",
        "Image_tensor = GetImageTensor(no_of_frames)\n",
        "training_generator = generator(train_path, train_doc, size_of_a_batch, Image_tensor)\n",
        "val_generator = generator(val_path, val_doc, size_of_a_batch, Image_tensor)\n",
        "\n",
        "if (num_of_train_seq%size_of_a_batch) == 0:\n",
        "    steps_per_epoch = int(num_of_train_seq/size_of_a_batch)\n",
        "else:\n",
        "    steps_per_epoch = (num_of_train_seq//size_of_a_batch) + 1\n",
        "\n",
        "if (num_of_value_seq%size_of_a_batch) == 0:\n",
        "    validation_steps = int(num_of_value_seq/size_of_a_batch)\n",
        "else:\n",
        "    validation_steps = (num_of_value_seq//size_of_a_batch) + 1\n",
        "    \n",
        "inputShape = (len(Image_tensor[0]), Image_tensor[1], Image_tensor[2], Image_tensor[3])\n",
        "\n",
        "model_3 = Sequential([\n",
        "    Conv3D(16, make3dFilter(3), padding='same', activation='relu', input_shape=inputShape),\n",
        "    MaxPooling3D(make3dFilter(2), padding='same'),\n",
        "    BatchNormalization(),\n",
        "\n",
        "    Conv3D(32, make3dFilter(3), padding='same', activation='relu'),\n",
        "    MaxPooling3D(pool_size=(2), padding='same'),\n",
        "    BatchNormalization(),\n",
        "\n",
        "    Conv3D(64, make3dFilter(3), padding='same', activation='relu'),\n",
        "    MaxPooling3D(pool_size=(2), padding='same'),\n",
        "    BatchNormalization(),\n",
        "\n",
        "    Flatten(),\n",
        "    Dense(128, activation='relu'),\n",
        "    BatchNormalization(),\n",
        "    Dropout(0.25),\n",
        "\n",
        "    Dense(64, activation='relu'),\n",
        "    BatchNormalization(),\n",
        "    Dropout(0.25),\n",
        "\n",
        "    Dense(5, activation='softmax')\n",
        "], name=\"conv_3d3\")\n",
        "model_3.compile(optimizer=tf.keras.optimizers.Adam(), loss='categorical_crossentropy', metrics=['categorical_accuracy'])\n",
        "print(model_3.summary())\n",
        "\n",
        "model_3_history = model_3.fit_generator(training_generator, steps_per_epoch=steps_per_epoch, epochs=no_of_epochs, verbose=1, \n",
        "            callbacks=callbacks_list, validation_data=val_generator, \n",
        "            validation_steps=validation_steps, class_weight=None, workers=1, initial_epoch=0)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 378
        },
        "id": "Y0iIvYzUH-I8",
        "outputId": "27e0b786-9c46-4fbf-983a-aeddaba8317c",
        "papermill": {
          "duration": 0.585342,
          "end_time": "2020-12-30T12:12:05.113531",
          "exception": false,
          "start_time": "2020-12-30T12:12:04.528189",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Max. Training Accuracy 0.9625668525695801\n",
            "Max. Validaiton Accuracy 0.453125\n"
          ]
        }
      ],
      "source": [
        "plotModelHistory(model_3_history)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "50alkFqTbYVq",
        "papermill": {
          "duration": 0.295298,
          "end_time": "2020-12-30T12:12:05.702374",
          "exception": false,
          "start_time": "2020-12-30T12:12:05.407076",
          "status": "completed"
        },
        "tags": []
      },
      "source": [
        "### Model 4: Frames-20, Same pooling in Conv3D layer"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "kR92LCmTBQCD",
        "outputId": "0d3fa1c9-1bff-42a9-e178-46a80ccf1c37",
        "papermill": {
          "duration": 904.604753,
          "end_time": "2020-12-30T12:27:10.607081",
          "exception": false,
          "start_time": "2020-12-30T12:12:06.002328",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Model: \"conv_3d4\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " conv3d_15 (Conv3D)          (None, 20, 100, 100, 16)  1312      \n",
            "                                                                 \n",
            " max_pooling3d_15 (MaxPoolin  (None, 10, 50, 50, 16)   0         \n",
            " g3D)                                                            \n",
            "                                                                 \n",
            " batch_normalization_25 (Bat  (None, 10, 50, 50, 16)   64        \n",
            " chNormalization)                                                \n",
            "                                                                 \n",
            " conv3d_16 (Conv3D)          (None, 10, 50, 50, 32)    13856     \n",
            "                                                                 \n",
            " max_pooling3d_16 (MaxPoolin  (None, 5, 25, 25, 32)    0         \n",
            " g3D)                                                            \n",
            "                                                                 \n",
            " batch_normalization_26 (Bat  (None, 5, 25, 25, 32)    128       \n",
            " chNormalization)                                                \n",
            "                                                                 \n",
            " conv3d_17 (Conv3D)          (None, 5, 25, 25, 64)     55360     \n",
            "                                                                 \n",
            " max_pooling3d_17 (MaxPoolin  (None, 3, 13, 13, 64)    0         \n",
            " g3D)                                                            \n",
            "                                                                 \n",
            " batch_normalization_27 (Bat  (None, 3, 13, 13, 64)    256       \n",
            " chNormalization)                                                \n",
            "                                                                 \n",
            " conv3d_18 (Conv3D)          (None, 3, 13, 13, 128)    221312    \n",
            "                                                                 \n",
            " max_pooling3d_18 (MaxPoolin  (None, 2, 7, 7, 128)     0         \n",
            " g3D)                                                            \n",
            "                                                                 \n",
            " batch_normalization_28 (Bat  (None, 2, 7, 7, 128)     512       \n",
            " chNormalization)                                                \n",
            "                                                                 \n",
            " flatten_5 (Flatten)         (None, 12544)             0         \n",
            "                                                                 \n",
            " dense_15 (Dense)            (None, 128)               1605760   \n",
            "                                                                 \n",
            " batch_normalization_29 (Bat  (None, 128)              512       \n",
            " chNormalization)                                                \n",
            "                                                                 \n",
            " dropout_10 (Dropout)        (None, 128)               0         \n",
            "                                                                 \n",
            " dense_16 (Dense)            (None, 64)                8256      \n",
            "                                                                 \n",
            " batch_normalization_30 (Bat  (None, 64)               256       \n",
            " chNormalization)                                                \n",
            "                                                                 \n",
            " dropout_11 (Dropout)        (None, 64)                0         \n",
            "                                                                 \n",
            " dense_17 (Dense)            (None, 5)                 325       \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 1,907,909\n",
            "Trainable params: 1,907,045\n",
            "Non-trainable params: 864\n",
            "_________________________________________________________________\n",
            "None\n",
            "Source path =  datasets/Project_data/train ; batch size = 64\n",
            "Epoch 1/20\n",
            "11/11 [==============================] - ETA: 0s - loss: 1.3902 - categorical_accuracy: 0.4947Source path =  datasets/Project_data/val ; batch size = 64\n",
            "11/11 [==============================] - 120s 12s/step - loss: 1.3902 - categorical_accuracy: 0.4947 - val_loss: 1.5973 - val_categorical_accuracy: 0.2100 - lr: 0.0010\n",
            "Epoch 2/20\n",
            "11/11 [==============================] - 54s 5s/step - loss: 0.8400 - categorical_accuracy: 0.6482 - val_loss: 2.2372 - val_categorical_accuracy: 0.1719 - lr: 0.0010\n",
            "Epoch 3/20\n",
            "11/11 [==============================] - 49s 5s/step - loss: 0.8279 - categorical_accuracy: 0.6838 - val_loss: 2.3253 - val_categorical_accuracy: 0.1964 - lr: 0.0010\n",
            "Epoch 4/20\n",
            "11/11 [==============================] - 44s 4s/step - loss: 0.7304 - categorical_accuracy: 0.7253 - val_loss: 2.6060 - val_categorical_accuracy: 0.1562 - lr: 0.0010\n",
            "Epoch 5/20\n",
            "11/11 [==============================] - ETA: 0s - loss: 0.5429 - categorical_accuracy: 0.8182\n",
            "Epoch 00005: ReduceLROnPlateau reducing learning rate to 0.00020000000949949026.\n",
            "11/11 [==============================] - 39s 4s/step - loss: 0.5429 - categorical_accuracy: 0.8182 - val_loss: 2.0162 - val_categorical_accuracy: 0.3438 - lr: 0.0010\n",
            "Epoch 6/20\n",
            "11/11 [==============================] - 34s 3s/step - loss: 0.5348 - categorical_accuracy: 0.8278 - val_loss: 1.7464 - val_categorical_accuracy: 0.2500 - lr: 2.0000e-04\n",
            "Epoch 7/20\n",
            "11/11 [==============================] - 32s 3s/step - loss: 0.4886 - categorical_accuracy: 0.8079 - val_loss: 3.4597 - val_categorical_accuracy: 0.0000e+00 - lr: 2.0000e-04\n",
            "Epoch 8/20\n",
            "11/11 [==============================] - 30s 3s/step - loss: 0.3342 - categorical_accuracy: 0.9091 - val_loss: 3.0141 - val_categorical_accuracy: 0.2500 - lr: 2.0000e-04\n",
            "Epoch 9/20\n",
            "11/11 [==============================] - ETA: 0s - loss: 0.3079 - categorical_accuracy: 0.8984\n",
            "Epoch 00009: ReduceLROnPlateau reducing learning rate to 4.0000001899898055e-05.\n",
            "11/11 [==============================] - 31s 3s/step - loss: 0.3079 - categorical_accuracy: 0.8984 - val_loss: 3.3114 - val_categorical_accuracy: 0.2500 - lr: 2.0000e-04\n",
            "Epoch 10/20\n",
            "11/11 [==============================] - 31s 3s/step - loss: 0.4389 - categorical_accuracy: 0.8342 - val_loss: 2.5775 - val_categorical_accuracy: 0.2500 - lr: 4.0000e-05\n",
            "Epoch 11/20\n",
            "11/11 [==============================] - 29s 3s/step - loss: 0.4076 - categorical_accuracy: 0.8503 - val_loss: 2.9652 - val_categorical_accuracy: 0.2500 - lr: 4.0000e-05\n",
            "Epoch 12/20\n",
            "11/11 [==============================] - 32s 3s/step - loss: 0.2756 - categorical_accuracy: 0.9251 - val_loss: 3.4282 - val_categorical_accuracy: 0.2500 - lr: 4.0000e-05\n",
            "Epoch 13/20\n",
            "11/11 [==============================] - ETA: 0s - loss: 0.3978 - categorical_accuracy: 0.8717\n",
            "Epoch 00013: ReduceLROnPlateau reducing learning rate to 8.000000525498762e-06.\n",
            "11/11 [==============================] - 29s 3s/step - loss: 0.3978 - categorical_accuracy: 0.8717 - val_loss: 4.0668 - val_categorical_accuracy: 0.1250 - lr: 4.0000e-05\n",
            "Epoch 14/20\n",
            "11/11 [==============================] - 31s 3s/step - loss: 0.3485 - categorical_accuracy: 0.9091 - val_loss: 4.2742 - val_categorical_accuracy: 0.0000e+00 - lr: 8.0000e-06\n",
            "Epoch 15/20\n",
            "11/11 [==============================] - 30s 3s/step - loss: 0.3151 - categorical_accuracy: 0.8984 - val_loss: 2.1549 - val_categorical_accuracy: 0.5000 - lr: 8.0000e-06\n",
            "Epoch 16/20\n",
            "11/11 [==============================] - 30s 3s/step - loss: 0.3260 - categorical_accuracy: 0.8877 - val_loss: 4.0116 - val_categorical_accuracy: 0.1250 - lr: 8.0000e-06\n",
            "Epoch 17/20\n",
            "11/11 [==============================] - ETA: 0s - loss: 0.3228 - categorical_accuracy: 0.8877\n",
            "Epoch 00017: ReduceLROnPlateau reducing learning rate to 1.6000001778593287e-06.\n",
            "11/11 [==============================] - 29s 3s/step - loss: 0.3228 - categorical_accuracy: 0.8877 - val_loss: 3.4662 - val_categorical_accuracy: 0.1250 - lr: 8.0000e-06\n",
            "Epoch 18/20\n",
            "11/11 [==============================] - 32s 3s/step - loss: 0.3650 - categorical_accuracy: 0.8877 - val_loss: 3.1169 - val_categorical_accuracy: 0.2500 - lr: 1.6000e-06\n",
            "Epoch 19/20\n",
            "11/11 [==============================] - 31s 3s/step - loss: 0.2236 - categorical_accuracy: 0.9519 - val_loss: 2.8249 - val_categorical_accuracy: 0.3750 - lr: 1.6000e-06\n",
            "Epoch 20/20\n",
            "11/11 [==============================] - 29s 3s/step - loss: 0.3247 - categorical_accuracy: 0.9037 - val_loss: 2.8857 - val_categorical_accuracy: 0.2500 - lr: 1.6000e-06\n"
          ]
        }
      ],
      "source": [
        "no_of_frames = 20\n",
        "no_of_epochs = 20\n",
        "size_of_a_batch = 64\n",
        "\n",
        "Image_tensor = GetImageTensor(no_of_frames)\n",
        "training_generator = generator(train_path, train_doc, size_of_a_batch, Image_tensor)\n",
        "val_generator = generator(val_path, val_doc, size_of_a_batch, Image_tensor)\n",
        "\n",
        "if (num_of_train_seq%size_of_a_batch) == 0:\n",
        "    steps_per_epoch = int(num_of_train_seq/size_of_a_batch)\n",
        "else:\n",
        "    steps_per_epoch = (num_of_train_seq//size_of_a_batch) + 1\n",
        "\n",
        "if (num_of_value_seq%size_of_a_batch) == 0:\n",
        "    validation_steps = int(num_of_value_seq/size_of_a_batch)\n",
        "else:\n",
        "    validation_steps = (num_of_value_seq//size_of_a_batch) + 1\n",
        "    \n",
        "inputShape = (len(Image_tensor[0]), Image_tensor[1], Image_tensor[2], Image_tensor[3])\n",
        "\n",
        "model_4 = Sequential([\n",
        "    Conv3D(16, make3dFilter(3), padding='same', activation='relu', input_shape=inputShape),\n",
        "    MaxPooling3D(make3dFilter(2), padding='same'),\n",
        "    BatchNormalization(),\n",
        "\n",
        "    Conv3D(32, make3dFilter(3), padding='same', activation='relu'),\n",
        "    MaxPooling3D(pool_size=(2), padding='same'),\n",
        "    BatchNormalization(),\n",
        "\n",
        "    Conv3D(64, make3dFilter(3), padding='same', activation='relu'),\n",
        "    MaxPooling3D(pool_size=(2), padding='same'),\n",
        "    BatchNormalization(),\n",
        "\n",
        "    Conv3D(128, make3dFilter(3), padding='same', activation='relu'),\n",
        "    MaxPooling3D(pool_size=(2), padding='same'),\n",
        "    BatchNormalization(),\n",
        "\n",
        "    Flatten(),\n",
        "    Dense(128, activation='relu'),\n",
        "    BatchNormalization(),\n",
        "    Dropout(0.25),\n",
        "\n",
        "    Dense(64, activation='relu'),\n",
        "    BatchNormalization(),\n",
        "    Dropout(0.25),\n",
        "\n",
        "    Dense(5, activation='softmax')\n",
        "], name=\"conv_3d4\")\n",
        "model_4.compile(optimizer=tf.keras.optimizers.Adam(), loss='categorical_crossentropy', metrics=['categorical_accuracy'])\n",
        "print(model_4.summary())\n",
        "\n",
        "model_4_history = model_4.fit_generator(training_generator, steps_per_epoch=steps_per_epoch, epochs=no_of_epochs, verbose=1, \n",
        "            callbacks=callbacks_list, validation_data=val_generator, \n",
        "            validation_steps=validation_steps, class_weight=None, workers=1, initial_epoch=0)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 374
        },
        "id": "2igfIb3mH-I8",
        "outputId": "0f51e349-5687-43c9-ec59-8599eca7c75b",
        "papermill": {
          "duration": 0.951406,
          "end_time": "2020-12-30T12:27:12.141064",
          "exception": false,
          "start_time": "2020-12-30T12:27:11.189658",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Max. Training Accuracy 0.9518716335296631\n",
            "Max. Validaiton Accuracy 0.5\n"
          ]
        }
      ],
      "source": [
        "plotModelHistory(model_4_history)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jd5Nr0nUaR8X",
        "papermill": {
          "duration": 0.37571,
          "end_time": "2020-12-30T12:27:12.918785",
          "exception": false,
          "start_time": "2020-12-30T12:27:12.543075",
          "status": "completed"
        },
        "tags": []
      },
      "source": [
        "### Model 5: Frames:20, Epocs-20, BS-64, reduced kernel to (2,2,2), switching BatchNormalization before MaxPooling."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 241
        },
        "id": "wuKTh7-yHs13",
        "outputId": "95ba4562-3bdc-4803-b7dd-be8bd03be221",
        "papermill": {
          "duration": 904.859345,
          "end_time": "2020-12-30T12:42:18.145385",
          "exception": false,
          "start_time": "2020-12-30T12:27:13.286040",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Model: \"conv_3d5\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " conv3d_19 (Conv3D)          (None, 20, 100, 100, 16)  400       \n",
            "                                                                 \n",
            " batch_normalization_31 (Bat  (None, 20, 100, 100, 16)  64       \n",
            " chNormalization)                                                \n",
            "                                                                 \n",
            " max_pooling3d_19 (MaxPoolin  (None, 10, 50, 50, 16)   0         \n",
            " g3D)                                                            \n",
            "                                                                 \n",
            " conv3d_20 (Conv3D)          (None, 10, 50, 50, 32)    4128      \n",
            "                                                                 \n",
            " batch_normalization_32 (Bat  (None, 10, 50, 50, 32)   128       \n",
            " chNormalization)                                                \n",
            "                                                                 \n",
            " max_pooling3d_20 (MaxPoolin  (None, 5, 25, 25, 32)    0         \n",
            " g3D)                                                            \n",
            "                                                                 \n",
            " conv3d_21 (Conv3D)          (None, 5, 25, 25, 64)     16448     \n",
            "                                                                 \n",
            " batch_normalization_33 (Bat  (None, 5, 25, 25, 64)    256       \n",
            " chNormalization)                                                \n",
            "                                                                 \n",
            " max_pooling3d_21 (MaxPoolin  (None, 2, 12, 12, 64)    0         \n",
            " g3D)                                                            \n",
            "                                                                 \n",
            " conv3d_22 (Conv3D)          (None, 2, 12, 12, 128)    65664     \n",
            "                                                                 \n",
            " batch_normalization_34 (Bat  (None, 2, 12, 12, 128)   512       \n",
            " chNormalization)                                                \n",
            "                                                                 \n",
            " max_pooling3d_22 (MaxPoolin  (None, 1, 6, 6, 128)     0         \n",
            " g3D)                                                            \n",
            "                                                                 \n",
            " flatten_6 (Flatten)         (None, 4608)              0         \n",
            "                                                                 \n",
            " dense_18 (Dense)            (None, 256)               1179904   \n",
            "                                                                 \n",
            " dense_19 (Dense)            (None, 128)               32896     \n",
            "                                                                 \n",
            " dense_20 (Dense)            (None, 5)                 645       \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 1,301,045\n",
            "Trainable params: 1,300,565\n",
            "Non-trainable params: 480\n",
            "_________________________________________________________________\n",
            "None\n",
            "Source path =  datasets/Project_data/train ; batch size = 64\n",
            "Epoch 1/20\n",
            "11/11 [==============================] - ETA: 0s - loss: 3.6410 - categorical_accuracy: 0.3137Source path =  datasets/Project_data/val ; batch size = 64\n",
            "11/11 [==============================] - 123s 12s/step - loss: 3.6410 - categorical_accuracy: 0.3137 - val_loss: 1.5462 - val_categorical_accuracy: 0.2400 - lr: 0.0010\n",
            "Epoch 2/20\n",
            "11/11 [==============================] - 55s 5s/step - loss: 1.0029 - categorical_accuracy: 0.5850 - val_loss: 2.7107 - val_categorical_accuracy: 0.1406 - lr: 0.0010\n",
            "Epoch 3/20\n",
            "11/11 [==============================] - 50s 5s/step - loss: 0.7972 - categorical_accuracy: 0.7075 - val_loss: 5.5231 - val_categorical_accuracy: 0.1071 - lr: 0.0010\n",
            "Epoch 4/20\n",
            "11/11 [==============================] - 41s 4s/step - loss: 0.5918 - categorical_accuracy: 0.7554 - val_loss: 8.1858 - val_categorical_accuracy: 0.2188 - lr: 0.0010\n",
            "Epoch 5/20\n",
            "11/11 [==============================] - ETA: 0s - loss: 0.4332 - categorical_accuracy: 0.8469\n",
            "Epoch 00005: ReduceLROnPlateau reducing learning rate to 0.00020000000949949026.\n",
            "11/11 [==============================] - 40s 4s/step - loss: 0.4332 - categorical_accuracy: 0.8469 - val_loss: 7.6545 - val_categorical_accuracy: 0.1875 - lr: 0.0010\n",
            "Epoch 6/20\n",
            "11/11 [==============================] - 31s 3s/step - loss: 0.4029 - categorical_accuracy: 0.8325 - val_loss: 8.3188 - val_categorical_accuracy: 0.3750 - lr: 2.0000e-04\n",
            "Epoch 7/20\n",
            "11/11 [==============================] - 33s 3s/step - loss: 0.2853 - categorical_accuracy: 0.8966 - val_loss: 16.3852 - val_categorical_accuracy: 0.0000e+00 - lr: 2.0000e-04\n",
            "Epoch 8/20\n",
            "11/11 [==============================] - 33s 3s/step - loss: 0.2370 - categorical_accuracy: 0.9305 - val_loss: 16.5965 - val_categorical_accuracy: 0.1250 - lr: 2.0000e-04\n",
            "Epoch 9/20\n",
            "11/11 [==============================] - ETA: 0s - loss: 0.1810 - categorical_accuracy: 0.9465\n",
            "Epoch 00009: ReduceLROnPlateau reducing learning rate to 4.0000001899898055e-05.\n",
            "11/11 [==============================] - 27s 3s/step - loss: 0.1810 - categorical_accuracy: 0.9465 - val_loss: 14.0104 - val_categorical_accuracy: 0.0000e+00 - lr: 2.0000e-04\n",
            "Epoch 10/20\n",
            "11/11 [==============================] - 29s 3s/step - loss: 0.1230 - categorical_accuracy: 0.9733 - val_loss: 12.4750 - val_categorical_accuracy: 0.2500 - lr: 4.0000e-05\n",
            "Epoch 11/20\n",
            "11/11 [==============================] - 29s 3s/step - loss: 0.0853 - categorical_accuracy: 0.9947 - val_loss: 15.2570 - val_categorical_accuracy: 0.1250 - lr: 4.0000e-05\n",
            "Epoch 12/20\n",
            "11/11 [==============================] - 31s 3s/step - loss: 0.1194 - categorical_accuracy: 0.9786 - val_loss: 13.4687 - val_categorical_accuracy: 0.0000e+00 - lr: 4.0000e-05\n",
            "Epoch 13/20\n",
            "11/11 [==============================] - ETA: 0s - loss: 0.1402 - categorical_accuracy: 0.9626\n",
            "Epoch 00013: ReduceLROnPlateau reducing learning rate to 8.000000525498762e-06.\n",
            "11/11 [==============================] - 30s 3s/step - loss: 0.1402 - categorical_accuracy: 0.9626 - val_loss: 18.5966 - val_categorical_accuracy: 0.2500 - lr: 4.0000e-05\n",
            "Epoch 14/20\n",
            "11/11 [==============================] - 31s 3s/step - loss: 0.1181 - categorical_accuracy: 0.9840 - val_loss: 12.4250 - val_categorical_accuracy: 0.2500 - lr: 8.0000e-06\n",
            "Epoch 15/20\n",
            "11/11 [==============================] - 29s 3s/step - loss: 0.1286 - categorical_accuracy: 0.9786 - val_loss: 12.0093 - val_categorical_accuracy: 0.1250 - lr: 8.0000e-06\n",
            "Epoch 16/20\n",
            "11/11 [==============================] - 30s 3s/step - loss: 0.1053 - categorical_accuracy: 0.9786 - val_loss: 11.7629 - val_categorical_accuracy: 0.2500 - lr: 8.0000e-06\n",
            "Epoch 17/20\n",
            "11/11 [==============================] - ETA: 0s - loss: 0.0697 - categorical_accuracy: 0.9893\n",
            "Epoch 00017: ReduceLROnPlateau reducing learning rate to 1.6000001778593287e-06.\n",
            "11/11 [==============================] - 29s 3s/step - loss: 0.0697 - categorical_accuracy: 0.9893 - val_loss: 16.9639 - val_categorical_accuracy: 0.0000e+00 - lr: 8.0000e-06\n",
            "Epoch 18/20\n",
            "11/11 [==============================] - 31s 3s/step - loss: 0.0873 - categorical_accuracy: 0.9893 - val_loss: 7.5047 - val_categorical_accuracy: 0.5000 - lr: 1.6000e-06\n",
            "Epoch 19/20\n",
            "11/11 [==============================] - 31s 3s/step - loss: 0.1211 - categorical_accuracy: 0.9679 - val_loss: 6.7127 - val_categorical_accuracy: 0.2500 - lr: 1.6000e-06\n",
            "Epoch 20/20\n",
            "11/11 [==============================] - 31s 3s/step - loss: 0.0794 - categorical_accuracy: 0.9893 - val_loss: 17.6388 - val_categorical_accuracy: 0.1250 - lr: 1.6000e-06\n"
          ]
        }
      ],
      "source": [
        "no_of_frames = 20\n",
        "no_of_epochs = 20\n",
        "size_of_a_batch = 64\n",
        "\n",
        "Image_tensor = GetImageTensor(no_of_frames)\n",
        "training_generator = generator(train_path, train_doc, size_of_a_batch, Image_tensor)\n",
        "val_generator = generator(val_path, val_doc, size_of_a_batch, Image_tensor)\n",
        "\n",
        "if (num_of_train_seq%size_of_a_batch) == 0:\n",
        "    steps_per_epoch = int(num_of_train_seq/size_of_a_batch)\n",
        "else:\n",
        "    steps_per_epoch = (num_of_train_seq//size_of_a_batch) + 1\n",
        "\n",
        "if (num_of_value_seq%size_of_a_batch) == 0:\n",
        "    validation_steps = int(num_of_value_seq/size_of_a_batch)\n",
        "else:\n",
        "    validation_steps = (num_of_value_seq//size_of_a_batch) + 1\n",
        "    \n",
        "inputShape = (len(Image_tensor[0]), Image_tensor[1], Image_tensor[2], Image_tensor[3])\n",
        "\n",
        "model_5 = Sequential([\n",
        "    Conv3D(16, make3dFilter(2), padding='same', activation='relu', input_shape=inputShape),\n",
        "    BatchNormalization(),\n",
        "    MaxPooling3D(make3dFilter(2)),\n",
        "\n",
        "    Conv3D(32, make3dFilter(2), padding='same', activation='relu'),\n",
        "    BatchNormalization(),\n",
        "    MaxPooling3D(pool_size=(2)),\n",
        "\n",
        "    Conv3D(64, make3dFilter(2), padding='same', activation='relu'),\n",
        "    BatchNormalization(),\n",
        "    MaxPooling3D(pool_size=(2)),\n",
        "\n",
        "    Conv3D(128, make3dFilter(2), padding='same', activation='relu'),\n",
        "    BatchNormalization(),\n",
        "    MaxPooling3D(pool_size=(2)),\n",
        "\n",
        "    Flatten(),\n",
        "    Dense(256, activation='relu'),\n",
        "    # BatchNormalization(),\n",
        "    # Dropout(0.3),\n",
        "\n",
        "    Dense(128, activation='relu'),\n",
        "    # BatchNormalization(),\n",
        "    # Dropout(0.2),\n",
        "\n",
        "    Dense(5, activation='softmax')\n",
        "], name=\"conv_3d5\")\n",
        "model_5.compile(optimizer=tf.keras.optimizers.Adam(), loss='categorical_crossentropy', metrics=['categorical_accuracy'])\n",
        "print(model_5.summary())\n",
        "\n",
        "model_5_history = model_5.fit_generator(training_generator, steps_per_epoch=steps_per_epoch, epochs=no_of_epochs, verbose=1, \n",
        "            callbacks=callbacks_list, validation_data=val_generator, \n",
        "            validation_steps=validation_steps, class_weight=None, workers=1, initial_epoch=0)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "VEhWSkczH-I8",
        "papermill": {
          "duration": 0.768627,
          "end_time": "2020-12-30T12:42:19.380375",
          "exception": false,
          "start_time": "2020-12-30T12:42:18.611748",
          "status": "completed"
        },
        "tags": [],
        "outputId": "92955648-5d35-423e-ed50-3eec39f63aea"
      },
      "outputs": [
        {
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAA3AAAAEICAYAAAAeBBZSAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8/fFQqAAAACXBIWXMAAAsTAAALEwEAmpwYAACOrklEQVR4nOzdd3hUVfrA8e+bDiQhhISEHnondBQEFVQQESsilrW7rmtv666uuqvu+rOtdS1rxUIRRFCwgyKCSpHQe02AkFATIHXO748zE4aQkDY1eT/Pk2dm7r1z7zuTSe6895zzHjHGoJRSSimllFIq8IX4OwCllFJKKaWUUpWjCZxSSimllFJKBQlN4JRSSimllFIqSGgCp5RSSimllFJBQhM4pZRSSimllAoSmsAppZRSSimlVJDQBE4FDBH5UkSuqeZzt4rIWZ6OKRCJyLUiMt/tca6ItK3MttU4VrV/JxXs9z0RecLT+1VKqZoSkVUicoant1WBSUSMiLR33n9dRP5emW2rcZwrReSb6saplLswfweggpuI5Lo9rA/kA8XOx380xnxU2X0ZY871ZGyBSkSaA9uATsaYTaXWTQc2GWPuq+z+jDHRHorrMaC9MeYqt33Xid+JUip4iUgrYLXbogbAEcA10e25xpifKrs/Y0w3b2yrvENEvgJ+M8Y8Umr5BcAbQAtjTFFl9mWMucVDMaUAW4Bw17Gd34cq/Z1IqZPRFjhVI8aYaNcPsB04321ZyT8qEdGLBU7GmAzge+Bq9+UiEg+MAt73R1xKKRWMjDHbS52LAFLdlpUkb3ouqpwge5/eB64SESm1/Grgo8omb6p6guyzUmtoAqe8QkTOEJF0EfmLiOwG3hWRRiLyhYhkich+5/0Wbs/5QURudN6/VkTmi8izzm23iEilWoNEJFJEXhCRnc6fF0Qk0rkuwXncAyKyT0R+EpEQ57q/iEiGiOSIyDoRGV7GvgeKyG4RCXVbdpGILHfeHyAii0XkkIhkisjz5YT5PqUSOOByYLUxZoWIPCgim5yxrBaRi07yet27fzQWkZnO4/8GtCu17YsissO5fomIDHEuHwn8DRjn7JKZ5lzu/jsJEZGHRWSbiOwRkQki0tC5LsUZxzUisl1EskXkofJiLuM13CQiG52/k5ki0sy5XETkP87jHRKRFSLS3blulPO9yXH+3irdaqmUqhuc55Kfnf9H9gKPiUg7EZkjInud/6s+EpE4t+eUdMkXkcdEZIrz/12O2C6T/aq5bR8R+d257hMRmSzldCWvRIwtReRTsefTvSLyitu6m0Rkjdv5o49z+XHd/8StK7tU75wdLyLvij3P7heRz5zLV4rI+W7bhTtfQ+8yXucaERnt9jjMebw+IhIlIh86X98BEVkkIkllvF2fAY2BIW77aQSMBiaIPS8vdO5jl4i8IiIR5bzvx3XvF5H7nc/ZKSLXl9r2POfv85DY8+pjbqvnOW8PiD2nnionDn8Y5HxNB523g9zW/SAijzs/uzki8o2IJJQTc7V+T851F4jIMudr2CT2u8AJw1Kcn+0Pnfdd5/sbRGQ7MMe5/BOx348Oisg8Eenm9vx6IvKc2O8PB8V+v6snIrNE5PZSr2e5nOQ7j7I0gVPelAzEA62Bm7Gft3edj1sBR4FXyn02DATWAQnA08DbIidcYSvLQ8ApQC8gFRgAPOxcdy+QDiQCSdikxYhIJ+A2oL8xJgYYAWwtvWNjzK/AYWCY2+IrgI+d918EXjTGxGKTpynlxDgdSBCR09yWXc2x1rdN2JNRQ+AfwIci0rTil86rQB7QFLje+eNuEfZ9iXfG/ImIRBljvgL+BUx2XrFOLWPf1zp/zgTaAtGc+Ps7DegEDAceEZEuFQUsIsOAfwOXOePeBkxyrj4HGAp0xL4XlwF7nevexnbTjQG64zyJKKVUKQOBzdj/+U8Cgv2f0wzoArQEHjvJ88dg/yfFATM5+XmrzG2dCcN04D3s/9+JwMm+pJYbo9gLiF9g/1emAM2dx0RExjq3+wMQ64xnL5VT1XP2B9ihE92AJsB/nMsnAFe5bTcK2GWM+b2MY04Exrs9HgFkG2OWAtdg/++3xCZotzhjOI4x5ij2XPsHt8WXAWuNMWnYYR13Y79LnIo9P91a/ttgOZOZ+4CzgQ5A6XH2h53HjAPOA/4kIhc61w113sY5z6kLS+07HpgFvOR8bc8Ds0SksdtmVwDXYd/bCGcsZanW70lEBmB/V/c7X8NQyvjecxKnYz+bI5yPv8S+T02ApRzfXfRZoC8wCPsZewBw4Gw9dW0kIqnYz/OsKsRRNxlj9Ed/PPKD/cM/y3n/DKAAiDrJ9r2A/W6PfwBudN6/Ftjotq4+djxDciWOvQkY5bZuBLDVef+fwAzsWC/357cH9mD/QYdX8DqfAN5x3o/B/hNv7Xw8D5twJVTi/XoLeNN5v4Pz/WpSzrbLgAvc3pv5buuMM/5QoBDo7LbuX+7blrHf/diuRmBP+h+WWu/+O/keuNVtXSfn8cKwXyIMdqyBa/1vwOXlHPc94Ann/beBp93WRTv3m4JNlNdjE/KQUvvYDvwRiPX3Z19/9Ed/AufH9T/Ref9aYHsF218I/O722P188hjwndu6rsDRqm6L/XKcAYjb+vmu/4OVeE0lMWKTkCwgrIztvgburOh9cT52/z98BlU4Z2MvtjmARmVs1wzIcf1vBqYCD5Szz/bObes7H38EPOK8fz2wAOhZiffnNOCAK37gZ+Ducra9C5hezufF/T15B3jKbbuOpd/DUvt9AfiP836Kc9swt/XX4jwfYy/Y/lbq+QuBa533fwAedlt3K/BVJT8rlf09veGKt4x1W3F+rt0+2x+Wem1tTxJDnHObhtgE8yjO7xqltovCfg/p4Hz8LPDfyrzOuv6jLXDKm7KMMXmuByJSX0TecDahH8ImO3Hi1h2xlN2uO8aYI867lSnY0Qx7ZdJlm3MZwDPARuAbEdksIg86978R+0/9MWCPiEwSZze+MnwMXCy2W+bFwFJjjOt4N2D/ya91dokYXc4+wF55GisiUdh/5l8bY/YAiMgfnN0aDojIAWwLU5ndJ9wkYpOpHaVeewkRuc/ZZeWgc78NK7Ffl7Le1zDsVW2X3W73j1CN35cxJhd7xbi5MWYO9kriq9jfy5siEuvc9BLsld1tIvKjiJxaydehlKpb3P8nIiJJzv/xGc5z0Yec/P9g6f9rUVL+uJ/ytm0GZBjnt9Sy4qpCjC2BbabssV0tsRcxq6Mq5+yWwD5jzP7SOzHG7MQmUJeI7fZ5LuUU73Cee9cA54tIfWyLoatHywfYhHSSs/vf0yISXs5+5gPZwIUi0g7b8+Zj5+vo6OxWuNv5Ov5F5c57zTj5+XSgiMx1dl08iG0hrO751LX/5m6PK3U+re7viZp9VsDtvRGRUBF5ytkN8xDHWvISnD9RZR3L+XmbjB3DGIJtjf2gBjHVGZrAKW8ypR7fi221GWhsF0NXF4PKdIusip3YrgQurZzLMMbkGGPuNca0xZ4o7hHnWDdjzMfGmNOczzXA/5W1c2PMauw/2nM5vvskxpgNxpjx2C4E/wdMFZEG5cQ5H9gHXIDtQvA+gIi0Bv6H7dLZ2BgTB6yk4vcpCyjC/lN2f+049zsE223hMuzVuDjgoNt+S/++SivrfS0CMit4XkWO26/z/WqMvVqNMeYlY0xf7NXsjtjuHhhjFhljLsC+159RfndVpVTdVvp/27+cy3o4z0VX4fnzUGm7gOalhgG0LG9jTh7jDqBVOUnkDkqNfXZzBNubxSW51PqqnLN3APHiNi6vFFfXuLHAQmOLd5XH1Y3yAuw48I0AxphCY8w/jDFdsV3vRnN8N8nSJjjXX4W9IOo6N70GrMW28sRih05U5ve9i3LOp04fY7vJtjTGNARep/rnU9f+T/Y+lae6v6eTfVYOc/LPChz/Gq/A/v7Owl4YTnGLIRs7tKO8Y70PXInt2nrElOpuqsqmCZzypRhsM/oBZ//vR710nInAwyKSKHbQ7yPYq5eIyGgRae88iR7E9o13iEgnERnmbFXLc8bpOMkxPgbuxP6j/MS1UESuEpFEY4wD252D8vbjvBI7AZvoxQGfO1c1wP5jzHLu8zpsC9xJGWOKgU+xg/Tri0hX7BgClxhswpUFhInII9gxEi6ZQIrzKlhZJgJ3i0gbEYnm2Ji5mlb4mghcJyK9nO//v4BfjTFbRaS/8ypnOPaEkof9fUWInVOnoTGmEDjEyX9fSinlEgPkAgfFTutyvw+OuRB7vrlNbKGOC7CtRNWJ8TdscvGUiDQQW+xjsHPdW8B9ItJXrPbOi4Jgu+Jf4WwtGYkdw3Qy5Z6zjTG7sGOe/iu2iEa4iAx1e+5nQB/seXJCBceZhB3v/CfcLoiKyJki0sPZknQI27X+ZP/nJ2ATiJs4vppzjPP5uSLS2XmcypgCXCsiXZ2tg6W/s8RgW7fyxI4nu8JtXZYz1jLnaAVmAx1F5Arn52Ec9iLlF5WMrXQc1fk9vY099w4XW6SsufP9AftZudy5fT/g0krEkI/tPVMfex53xeDAdkd9XkSaOT9/pzrP9zgTNgfwHNr6VmmawClfegGoh70a8wvwlZeO8wSwGFgOrMAOpnVVleoAfIc9MS7E9rWeC0QCTzlj241t1fnrSY4xEXvym2OMyXZbPhJYJXZ+vBexY8BOGHTtZgL2qttkY0w+lLTwPeeMLxPoge2OUhm3YbtZ7Mb25X/Xbd3X2Pd8PbYFMY/ju4e4EtG9IrK0jH2/g/3nOg87v00ecHsZ21WJMeY74O/ANOyXknbYipxgE8z/YfvIb8OeHJ5xrrsa2OrsrnEL9gqeUkpV5B/Y5OIgtljCp94+oDGmANvl/gbsxb2rsF/W86sao/Ni3fnY8WPbsYW5xjnXfYIt1PIxdmzZZ9iiEWCTqfOdx7/Sue5kXuDk5+yrsUnVWuwY8rvcYjyK/Z/ehgreX2eSsRDbyjbZbVUydvzcIWw3yx85yRd8Y8xW7Ji5BtiWMZf7sMlVDvZ8MvmEJ5e9vy+x78Ec7NCL0oWybgX+KSI52AvFU9yeewT7e/hZ7FCIU0rtey+2RfFe7HntAWB0qe8TlfUC1fg9GWN+wxZJ+Q/2c/Yjx1oF/449F+/HfhY/5uQmYM/RGdg5GX8ptf4+7PexRdieR//H8TnIBOx3nQ8rOI5ykuO7YyullFJKKW8TkV+B140x71a4cRBy9vLoaIy5qsKNVZ0mIn8AbnYOY1GVoC1wSimllFJeJiKni0iys8vcNUBPvNcTxa+cXfluAN70dywqsDm7p96KflaqRBM4pZRSSinv6wSkYbsw3gtc6uw+WKuIyE3Y7vlfGmPmVbS9qrtEZAR2vGAmFXfTVG60C6VSSimllFJKBQltgVNKKaWUUkqpIFHeRJR+lZCQYFJSUvwdhlJKKS9bsmRJtjEm0d9xBAs9PyqlVN1R3jkyIBO4lJQUFi9e7O8wlFJKeZmIbPN3DMFEz49KKVV3lHeO1C6USimllFJKKRUkNIFTSimllFJKqSChCZxSSimllFJKBYmAHANXlsLCQtLT08nLy/N3KAEtKiqKFi1aEB4e7u9QlFJKKaWUUh4WNAlceno6MTExpKSkICL+DicgGWPYu3cv6enptGnTxt/hKKVUnSUi7wCjgT3GmO5lrBfgRWAUcAS41hiz1LdRKqWUCkZB04UyLy+Pxo0ba/J2EiJC48aNtZVSKaX87z1g5EnWnwt0cP7cDLzmg5iUUkrVAkGTwAGavFWCvkdKKeV/xph5wL6TbHIBMMFYvwBxItLUN9EppZQKZkHThVIpFWBys2DTHEgd5+9IlApGzYEdbo/Tnct2ld5QRG7GttLRqlUrnwSnVLDYd7iAxVv3sWFPLq3i69M5OYY2CQ0ICw2qNgqlqkQTuCqIjo4mNzfX32EoFRjmPQO/vQHNekFiJ39Ho1StZYx5E3gToF+/fsbP4SjlN8YYtu09wqKt+1i8dT+Lt+1jU9bhE7aLCA2hXZNoOifH0Ck5hk5J9rZpwyjtqeRHxhj2Hylk54GjZBw4yu6DeRQWOzyy7xaN6tG3dTyJMZEe2V91HTxayNLt+1m3O4dbTm/nteNoAqeUqjqHA9bMtPc3fq8JnFJVlwG0dHvcwrlMKeVUWOxg9c5DbgnbfrJz8wFoWC+cfq0bcUnfFvRPiadTcgw79h1h3e4c1u3OYe3uHBZu2sv034/9WcVGhdmELjmGTsmxJYldw3qBWbk7N7+I37fb1xwTGU5MVBgxUfY2Niqc6KgwQkMCJyHNKyxm18G8kgRtZ8mPXbbz4FHyCj2TsJUnpXF9+qXE0z+lEf1S4mmb0MCrSXvGgaMsdn4+F23dx7rMHIyB0BDhsn4tiW8Q4ZXjagJXDcYYHnjgAb788ktEhIcffphx48axa9cuxo0bx6FDhygqKuK1115j0KBB3HDDDSxevBgR4frrr+fuu+/290tQqmZ2/Ao5uwCBTd/Dqbf6OyKlgs1M4DYRmQQMBA4aY07oPqlUXZKTV8jv2w+weNt+Fm/dx+/bD3C0sBiAlvH1GNohgb4pjeifEk/7xGhCSiUv3Zo1pFuzhsctO3ikkHWZOazbfYi1zuRuxu87ycnfXrJN04ZRxxI7Z1LXvkk0kWGh3n/RbnYfzGPxtmPJwJpdh3BU0ObeICK0JKlzJXjRUWHEupK9yGPLG0SGAjVPZgqLHWQeynNL0vLYdfAo2bkFx20nAk1iImkWV48uzWIZ3qUJzeLq0SyuHs3j6pHcMIrIsJp3dXUY2JSVy+Kt+1i0dT/fr8lk6pJ0AOIbRNCvdSP6ORO67s0aElHNYxY7DOszc0qOs3jrPnYetIUDG0SE0qd1I87t3pT+KY1IbRlHg0jvpVlBmcD94/NVrN55yKP77NoslkfP71apbT/99FOWLVtGWloa2dnZ9O/fn6FDh/Lxxx8zYsQIHnroIYqLizly5AjLli0jIyODlStXAnDgwAGPxq2UX6z+DEIjocelsPJTKMyD8Ch/R6VUwBCRicAZQIKIpAOPAuEAxpjXgdnYKQQ2YqcRuM4/kSrlP3tz81m4ee8JCUuI2O9l4/q3tF+8W8eT3LB655iG9cMZ0CaeAW3iS5YZY9h5MK8kqVvvbLH7eWM2hcU2YwoNEdokNKBTcgydnUld5+RYWjSqd0LiWB0Oh2HDntzjErb0/UcBqBceSp/Wcdw2rAP9UxrRLK4euXlF5OQVkZtfyCHn/Zy8wlK3RRw4UsCOfUec2xSSX+S9Fq8GEaE0b1SPpg3r0b15Q5rHRR2XoCXFRlU7WaqqPq0a0adVI24ean+/m7IOs2TbsUTrm9WZAESGhdCrZRz9U+Lpl9KIPq0bERtVdgvs0YJi0tIPlCRsS7ftJye/CICk2Ej6pcRzc2ubGHZOjvHpuMugTOD8bf78+YwfP57Q0FCSkpI4/fTTWbRoEf379+f666+nsLCQCy+8kF69etG2bVs2b97M7bffznnnncc555zj7/CVqhmHA1bPgPZnQdcLYNlHsH0htDvT35EpFTCMMeMrWG+AP/soHKUCgjGGVTsPMWftHuas3UNa+gGMOZaw3D6sA/1SGtG7VSOivdh6ISI0dyYZwzonlSwvLHawJfuws6XuEOt257I8/QCzlh9rHK8fEUqHJPekzt42jj752Ku8wmJWZBw81h106z4O5dlkICE6kv4pjbhucBv6pzSiS9NYwj2UDBQUOUoSvCMFxR7ZZ1iokBQbRWxUWECOKRQR2jeJpn2TaMb1t4Wf9uTksWTrfhZt3c+Sbft47cdNFM81iECnpJiShC4yLLQk8Vu182BJQt8xKZrzezWzXTNbx9OiUT2/vvagTOAq21Lma0OHDmXevHnMmjWLa6+9lnvuuYc//OEPpKWl8fXXX/P6668zZcoU3nnnHX+HqlT1ubpPdrsQUk6D0AjbjVITOKWUUqXk5hcxf0MWc9bu4Yd1WezJyUcEUlvEcdfwjpzeKZFuzTyXsNREeGgIHZNi6JgUA6nNSpbn5hexPjOnZHzdut05fLsmk8mLjxWSTYiOpFNyNJ2SYumcHEPH5Biyc/JZ5GxhW5F+kAJnwY72TaIZ1aNpyVitVvH1vZYMRISF0Dg6ssIEs7ZrEhPFuT2acm4PO1vLkYIilm0/YFvotu3j06XpfPDLNsAWwUlt2ZAbh7Slf4pt2Yur752xbNUVlAmcvw0ZMoQ33niDa665hn379jFv3jyeeeYZtm3bRosWLbjpppvIz89n6dKljBo1ioiICC655BI6derEVVdd5e/wlaoZV/fJjiMhogG0OgU2zgFtXFZKqTrPGMPm7MPMdbayLdq6j8JiQ0xUGEM7JjKsUxNO75RIQhAlFNGRYSVd9FyMMWTl5h9XNGV9Zg4f/7btuEId4aFCj+YNuW5wCv1S4unbupHXCluoyqsfEcag9gkMap8AQFGxg7W7c8gvKqZbs4ZEhft2/GNVaQJXDRdddBELFy4kNTUVEeHpp58mOTmZ999/n2eeeYbw8HCio6OZMGECGRkZXHfddTgc9o/53//+t5+jV6oG3LtPRsXaZe2GwXePwaFdEKvzECulVF2TV1jMr1v2MXftHuau28O2vUcA2+3s+tPaMKxTE/q0bhQQrWyeIiI0iYmiSUwUQzokliwvdhi27zvC+swc4uqFk9oyLuCTAQVhoSF0b96w4g0DhCZwVeCaA05EeOaZZ3jmmWeOW3/NNddwzTXXnPC8pUuX+iQ+pbwu/bdj3Sdd2g23CdzmudDrCn9FppRSyod2HjjK3HV7mLs2i583ZnO0sJio8BAGtUvgxiFtObNTIi0a1fd3mD7nKn7SJqGBv0NRtViFCZyIvAOMBvYYY7o7l00GXBM/xQEHjDG9ynjuViAHKAaKjDH9PBK1Uso/Vn12rPukS1J3aNDEzgenCVzdsXY2JHWDRq39HYlSyouKih1s3Xu4pAT/ut05rMvMKWlla9GoHmP7teDMzk04tW1jbW1Sygcq0wL3HvAKMMG1wBgzznVfRJ4DDp7k+WcaY7KrG6BSKkA4HHb8W/vhx7pPAoSE2G6UG7+124TUni4yqhwFR2DSeDjzITj9AX9Ho5TyAGMMuw/lHZ+o7c5hY1YuBc5S9CECbRIa0L1ZQ64c2IozOzWhfZPogKxEqFRtVmECZ4yZJyIpZa0T+xd7GTDMw3EppQKNq/tk13+cuK7dMFg+CXYtg+Z9fB6a8rG9G+1tQgf/xqGUqpaDRwtZn5lTUi5//e5c1u4+VFLWHiA51k5uPaRDQskk1+0So7WFTakAUNMxcEOATGPMhnLWG+AbETHAG8aYN8vbkYjcDNwM0KpVqxqGpZTyOFf3yU7nnriunfMazqY5msDVBdnr7W1CR//GoZSqlIIiB3PX7WHGsgyWbT/AzoN5JetiIsPolBzD+anNnHOaxdIpKYaG9cue3Fgp5X81TeDGAxNPsv40Y0yGiDQBvhWRtcaYeWVt6Ezu3gTo16+fqWFcSilPKqk+War7pEt0IiT3tAnc0Pt8H5/yrewNgEB8O39HopQqhzGGlRmHmLY0nZlpO9l3uICE6EhOa9+YTsnH5ipr1jBKu0AqFWSqncCJSBhwMdC3vG2MMRnO2z0iMh0YAJSZwCmlAlj6IsjZCV0fK3+b9sNhwcuQd6jsJE/VHtnrbfGS8Ch/R6KUKiXzUB7Tf8/g06XprM/MJSIshLO7JnFpnxYM6ZBAWC0q5a9UXVWTFrizgLXGmPSyVopIAyDEGJPjvH8O8M8aHE8p5S+rP4PQCOg0svxt2g2D+f+BrT9B5/N8Fpryg+wN2n1SqQBytKCYb1bvZtrSDOZvyMJhoE+rOJ68qDujezTT7pBK1TKVmUZgInAGkCAi6cCjxpi3gcsp1X1SRJoBbxljRgFJwHRns3wY8LEx5ivPhh+4oqOjS+aNK23r1q2MHj2alStX+jgqparB4bDj39oNh6iTTHLZ8hQIb2CnE9AEznLYym21qjKnwwF7N0Db0/0diVJ1mjGGRVv3M21JOrNX7CInv4jmcfX485ntuah3c9omRvs7RKWUl1SmCuX4cpZfW8ayncAo5/3NQGoN41NK+Zur++RZj518u7AIaDPEjoNT1jcPw45f4abv/R2J5xzcAUV5WoFSKT/Zse8I05am8+nSDLbvO0L9iFDO7d6US/o255Q2jQkJ0fFsStV2NS1i4h9fPgi7V3h2n8k94Nynyl394IMP0rJlS/785z8D8NhjjxEWFsbcuXPZv38/hYWFPPHEE1xwwQVVOmxeXh5/+tOfWLx4MWFhYTz//POceeaZrFq1iuuuu46CggIcDgfTpk2jWbNmXHbZZaSnp1NcXMzf//53xo0bV/FBlKqJynSfdGk3HNZ/Bfs2Q3xbr4cW0IoLIe1jOHoA8nMhspZcDc92Fh3WLpRK+UxOXiGzV+xi2pIMftu6DxEY1K4xdw7vwMjuyTSIDM6vc0qp6tG/+EoaN24cd911V0kCN2XKFL7++mvuuOMOYmNjyc7O5pRTTmHMmDFVqub06quvIiKsWLGCtWvXcs4557B+/Xpef/117rzzTq688koKCgooLi5m9uzZNGvWjFmzZgFw8ODJ5k9XygNc1Scr6j7p0n64vd34PQyo4wnc1vlwdL+9n7UWWvTzbzyeolMIKOUzxhimLN7B41+sITe/iLYJDbh/RCcu7N2c5nH1/B2eUspPgjOBO0lLmbf07t2bPXv2sHPnTrKysmjUqBHJycncfffdzJs3j5CQEDIyMsjMzCQ5ObnS+50/fz633347AJ07d6Z169asX7+eU089lSeffJL09HQuvvhiOnToQI8ePbj33nv5y1/+wujRoxkyZIi3Xm5wMsZ+YXYUQXQTf0dTO2QshkMZMPyRym0f3xbiWtlulANu8m5sgW71DAgJs5/H3StqVwJXrxHUb+zvSJSq1fbm5vPXT1fwzepMTm3bmPtHdqJ3yzgt+a+UCtIEzk/Gjh3L1KlT2b17N+PGjeOjjz4iKyuLJUuWEB4eTkpKCnl5eRXvqBKuuOIKBg4cyKxZsxg1ahRvvPEGw4YNY+nSpcyePZuHH36Y4cOH88gjlfxiXRsYA4ez4cB2OLjd3h7YDgd2HLtfeBjCouDmH6FJZ39HHPxWfebsPlnG5N1lEbGtdSs+sV0IQ+to5TNHMaz5HLqcDxu+g8xV/o7Ic1wVKPVLpFJeM2dtJg9MXcGho4U8fF4Xrh/cRse2KaVKaAJXBePGjeOmm24iOzubH3/8kSlTptCkSRPCw8OZO3cu27Ztq/I+hwwZwkcffcSwYcNYv34927dvp1OnTmzevJm2bdtyxx13sH37dpYvX07nzp2Jj4/nqquuIi4ujrfeessLr9KPjIHcPc5kbJstlnCgVKJWdPT450Q1tC0+jdtB2zOgYQv46TmY8We44RsICfXLS6kVHA47/q3dsMp1n3RpPxyWvAs7foOUwV4LL6BtWwBHsqHrhXBoVy1L4NZDx3P8HYVStdKRgiKenLWGj37dTufkGD68cQCdk3VeTaXU8TSBq4Ju3bqRk5ND8+bNadq0KVdeeSXnn38+PXr0oF+/fnTuXPUWn1tvvZU//elP9OjRg7CwMN577z0iIyOZMmUKH3zwAeHh4SQnJ/O3v/2NRYsWcf/99xMSEkJ4eDivvfaaF16ln/z0PPzwFBTnH7+8XrxN0BI7QYdz7P2GLe1tXMuyE4uYZJh2Ayx8FQbf4Zv4a6Oqdp90aTMUJBQ2fV93E7jVMyCsHnQ4G7bMgxVT7QWKYG+1OrofDu/R8W9KecGyHQe4e/Iytu49zM1D23LvOR2JDNOLkEqpE2kCV0UrVhyrfpmQkMDChQvL3K68OeAAUlJSSuaAi4qK4t133z1hmwcffJAHH3zwuGUjRoxgxIgR1Qk78K2YCvFtoP+NzuTMmahVp3Jf90tg5acw90noNAoS2ns+3rqgqt0nXaIaQov+tpBJVZO/2sDhgDUzbfIW0QCSusHit+Fgur3oEMyyN9pbTeCU8piiYgevzt3ES3M2kBQTycc3nsKp7XSMqVKqfLVodlkVtPJzIWuN7W424CboOAKadKl+2XURGP28HQs34892PJKqmpLqk1XsPunSfjjsSrNjFuuaHb9CbiZ0dU4pktTd3taGbpRagVIpj9qSfZhLX1/If75bz/k9m/LlXUM1eVNKVUgTOC9asWIFvXr1Ou5n4MCB/g4r8OxKA+OA5n09t8+YZBj5FOz4BX5703P7rSsylsChdJtUV0e74YCBzT94MKggsXoGhEbaCxEASV3tbaaH5670h+z1EBIOca39HYlSQc0Yw8e/bmfUiz+xOSuXl8f35oXLe9OwXh0t/KSUqpKg6kJpjAmq8rk9evRg2bJlPj2mMcanx/OIjCX2tnkfz+439XJY9Sl89w/7ZbquTyxdFas/s1/Uq9p90qVZL1tqfuP30ONST0YW2FzdJ9ufBZExdllkDDRKqSUtcBtswaDQoDp1KBVQsnPzeXDacr5bs4fB7Rvz7NhUmjbUOd2UUpUXNC1wUVFR7N27NzgTFB8xxrB3716ioqL8HUrVZCyxV/QbJHh2vyIw+gVbyn7G7fbLtaqYw2HHv7UfDvXiqrePkFBbFXTTHFu8o67IWGILv7i6T7okda8lCdx6SOjg7yiU8ipjDHtz873yfeO71ZmM+M885m3I5pHRXfng+oGavCmlqixoLqO2aNGC9PR0srKy/B1KQIuKiqJFixb+DqNqMpZ6b5Ljhs1hxJMw83ZbSKKuTy5dGa7uk8Mertl+2g2HVdNhz2pbyKMuKGm5HHn88qRusG42FB6F8CD9slZcCPu3QNcx/o5EKa/JLyrmzx8t5bs1e2gQEUrbxGjaJTagXWI07ZtE065JNK0b169ydcjD+UU8MWs1E3/bQdemsUy8vBcdk2K89CqUUrVd0CRw4eHhtGnTxt9hKE/L3WMn5R74R+8do/fVtirlt4/aqQga6fidk6pp90mXdsPs7cbv60YCZwysnll24ZekbnacZ9ZaaNbbP/HV1L4t4CjSAiaq1sovKuZPHy5lzto93HhaG4ochk1ZuSzaup/Plu0s2S5EoFV8fdol2oSuXWIDm9wlRhNXP+KE/S7dvp97Ji9j274j3HJ6O+4+u4NOD6CUqpGgSeBULeUa/+atFjiwXSnHvAT/PRU+vwOu/iz45+PyFmOOVZ+sbvdJl4bNIbGLnQ+uLszHt/N3ezHijAdPXOeqRLl7ZfAmcCUVKLULpap93JO3Jy/qzpUDj7/Qdzi/iC3Zh9mUlcumPblsyrL3f9qYTUHRse75jRtEOBM722q393ABb87bTHJsFJNuOoWBbbXCpFKq5jSBU/6VscRO+pzc07vHiWsFZ/8DZt0LSydA32u8e7xglbEEDu6AM//mmf21Hw6//Q8KjkBEfc/sM1CtngEhYWW3XDZqA+H1g3scnCuBa6wJnKpdKkreABpEhtG9eUO6Nz++db3YYcjYf9Qmdlm5bNxjb79elcm+wzsAuLhPcx4b043YKK0wqZTyDE3glH9lLLFl1n3x5b7v9bY4xzcP2yqBDZt7/5jBZtV0Z/fJUZ7ZX7szYeErsO1nO7F1beVquWxzOtSPP3F9SAg06QqZK30fm6dkb4DoZIiK9XckSnlMZZK3kwkNEVo1rk+rxvU5s3OT49btO1zAoaOFpCQ08GTISikVPFUoVS1kjE3gPDn/28mEhMCYl+04ns/vrFvVESvDk90nXVoPthOqb5rjmf0Fqt0rnAU+Lih/m6RutgUuWD93WoFS1TL5RcXc8sES5qzdw78u6lHl5K0i8Q0iNHlTSnlFhQmciLwjIntEZKXbssdEJENEljl/yrxcLyIjRWSdiGwUkTIGhqg6bd9myDvouwQOIL4NDH8UNn4LaRN9d9yaMgY++zN8/0/vJQCu7pPdLvTcPsPrQetBtpBJbbZ6hu0K3Hl0+dskdYej+yBnt+/i8hRjbAucFjBRtYQreZu7Lot/XdSDKwa28ndISilVaZVpgXsPGFnG8v8YY3o5f2aXXikiocCrwLlAV2C8iHStSbCqlimZwNuHCRzAgJuh1anw1YNwaJdvj11dy6fAsg/hp+fgu0e9k8R5uvukS7vhkL0ODqZ7dr+BwhhbuTPlNGhwkgIFrkqcwTgOLncP5B/UBK6KKrqIKSKtRGSuiPwuIsvLuxiqPEuTN6VUsKswgTPGzAP2VWPfA4CNxpjNxpgCYBJwkv5Fqs5JXwzhDSCxs2+PGxICF7wKRfkw657A79J2dD988xA07wf9b4SfX4R5z3j2GCUl8M/0XPdJF/fpBGqjPWtg78aTd58EO9YTIHOF92PytL0b7K12oay0Sl7EfBiYYozpDVwO/Ne3UdY9mrwppWqDmoyBu815xfAdEWlUxvrmwA63x+nOZUpZGUugWS8I8cN8OI3b2Ymq182GFVN9f/yq+P5xOLIXRj8P5z4DqVfA3CdhwSueO0bGUlsCv+uFntunS5MuENOs9o6DWz0DEOhy/sm3q9cIGrYMzha4kikEtAWuCipzEdMArqowDYGdKK/JK9TkTSlVO1Q3gXsNaAf0AnYBz9U0EBG5WUQWi8jirKysmu5OBbqiAti93PfdJ92dciu06A9f3m+7iAWi9CWw+B0Y8EdomnqsEEvXC22r3OJ3PXOcVZ/a7pOdvdCDS8S2wm3+ARzFnt+/v62eYYu1RDepeFtXIZNgk73BToMQq9fgqqAyFzEfA64SkXRgNnB7WTvS82PN5RUW86cPNXlTStUO1UrgjDGZxphiY4wD+B/2SmNpGUBLt8ctnMvK2+ebxph+xph+iYmJ1QlLBZPMlVBc4N8ELiTUdqUsOGLnhws0xUXwxV0Qk3z8vGyhYXDx/6DDCPjibkibXLPjHNd9sqzGdA9oPwzyDtiWvtokax1kram4+6RLUjfbmlWU7924PC17PTRuby8gKE8aD7xnjGkBjAI+EJET3mQ9P9aMe/L274s1eVNKBb9qnY1FpKnbw4uAsiY3WgR0EJE2IhKB7d8/szrHU7WQvwqYlJbYCc54ENbMtEU8Asmit2wr5Yh/nTj3VlgEXDYB2gyBz/5kE7Dq8mb3SZe2ZwICm2rZODjX+15R90mXpG52GgtXl8Rgkb1eu09WXWUuYt4ATAEwxiwEooAEn0RXR5RO3sYP0ORNKRX8KjONwERgIdBJRNJF5AbgaRFZISLLgTOBu53bNhOR2QDGmCLgNuBrYA12oHYQ9h1SXpGxFBo0gYYt/B0JDLoDmvaCWffB4b3+jsY6tAvmPGG7Hna7qOxtwqPg8onQoh9MvR42fFu9Y62e7r3uky7146FZ79o3Dm71DGh5CsQ2rXhbsFMJQHB1oyw4Agd2aAJXdZW5iLkdGA4gIl2wCZz2kfSQvMJibtHkTSlVC1WmCuV4Y0xTY0y4MaaFMeZtY8zVxpgexpiexpgxxphdzm13GmNGuT13tjGmozGmnTHmSW++EBVkXBN4i/g7Etsl8cL/2jnpvnzA39FYX//NdjEd9ezJ36PIaLhiiq1wOPkq2PJT1Y5jDKyaAW3P8F73SZf2w23l0aMHvHscX9m7yVaUrGz3SYD4dhAaaSf+Dhb7NgFGK1BWUXkXMUXknyIyxrnZvcBNIpIGTASuNSbQy+IGB1fy9oMmb0qpWkgHNCjfyztou2T5u/uku6RuMPR+WDkV1s7ybyyb5tiiIkPusdUyK1IvDq6aDo3awMfjYMeiyh9rp7P7pCcn7y5Pu+FgimHLj94/li+snmFvK9t9EuzFgiZdgqsFTitQVltZFzGNMY8YY2Y67682xgw2xqQ651T9xr8R1w7uydtTmrwppWohTeCU7+38HTDQvI+/IznekHsgqYctDHKkOlMfekBhni2oEt8WBt9V+ec1aAx/+AxikuDDS2BXWuWe55q8u/N51Ym2alr0g8jY2jMf3OoZdm6+uJYVb+suqXuQJXAbAKncxQSl/Kx08na5Jm9KqVpIEzjleyUFTAIsgQsNhwtftXOuff23irf3hp9fgH2b4bzn7Bi3qohJhj/MtAVPPrjIVkg8GV92nwT7/rYZalsYg72X2P6tsGtZ1bpPuiR1g8N7AnfqitKy10NcKwiv5+9IlDopTd6UUnWFJnDK9zKW2pLkvkgaqqppKpx2N6RNhPVf+/bYezfBT89Dt4tt8ZLqiGsJf5gBIWHw/hibDJbHl90nXdoNg4M7YO9G3x3TG1zVJ7uOOfl2ZUnqZm+DpRVOK1CqILBgUzaXvbFQkzelVJ2gCZzyPVcBk0A19H5I7AKf3wWHs31zTGNg9n0QGmGnDaiJxu1sEldcAO9fAAfTy95u1Wc20evkxeqTpbUfbm+DvRvl6hm2cmmjlKo/N5gSOIcDsjdqAqcC1u/b93PlW79wxf9+Zc+hfF65orcmb0qpWk8TOOVbh3ZCzq7ATuDCIm1VyqP74O1zbHc5b1s13XYtHPZw5UvSn0yTLnD1dDt59vtjICfz+PXGwOrPbPfJ+vE1P15lNUqxlRiDeT64AzsgY3H1uk8CNEiA6GQ7mX2gO5QORUe1AqUKOGt3H+KmCYu56L8LWLMrh7+P7soP95/B6J7N/B2aUkp5nSZwyrcCZQLvijTvY8eTHdlrk7hdy713rLxD8NVfIbkn9L/Rc/tt1guunGoT5g8uPL4wy87f4YCXJ+8uT7thsHU+FOX7/tiesOZze1vdBA4guXtwJHBagVIFmK3Zh7lz0u+c++JP/LJpL/ee3ZF5D5zJDae1ISo81N/hKaWUT2gCp3wrfbGteuia0DiQtRoIN3xj4313FGz2Uvn7uf+C3EwY/YItM+9JrQbC+Il2fN2HF9tkEZzVJ8N8U32ytPbDofAIbP/F98f2hNUzbLXSmlRlTOpmi8wUF3ouLm/I3mBvNYFTfrbr4FH++ukKhj//I1+v2s0tp7fjp7+cye3DOxAd6eH/m0opFeA0gVO+lbHEtj5UtcKivyR2sklcwxa2PP/KaZ7d/640+O0N6HcdtPBSq2TbM2DcB3by6I8vg4LD/uk+6ZIyxCbFwdiN8tBO2PFLzVrfwF7AKC4I/GIu2eshKs52+1TKD/bm5vP4F6s5/ZkfmLpkB1cNbMW8+8/kLyM7E1c/wt/hKaWUX2gCp3zHUQw7l9m5s4JJw+Zw/ZfQoj9MvQF+ed0z+3U44It7oH5jGP6IZ/ZZno4j4JK3YMevtkuov7pPAkRGQ6tTYOMc/xy/JtZ8YW9rnMAFSSGT7A229U3E35GoOuZQXiHPf7OOoU/P5d2ft3BBajPm3HsG/7igO01ig+QCoFJKeYn2O1C+k70BCnICf/xbWeo1gqs/hWk3wld/sePKznqsZl9sl75ni2Fc9IZvplTodhEUHoXP/uS/7pMu7YbB9/+wxVVikvwXR1WtnmErlCbWsEth4w62FTJzJfS41DOxeUP2emh/tr+jUHXI0YJi3luwldd/3MTBo4Wc16Mpd5/dkfZNov0dmlJKBQxN4JTvBEsBk/KE14PLJthy/z+/YMetjXnZTlBdVblZ8N1jtjthz3GejrR8va6wUxUc2euf7pMurgRu81xIvdx/cVRF7h7Y9jOc/pea7ysswnbP3R3AhUyOHrCfca1AqXygoMjBpEXbeXnORrJy8jmzUyL3ntOJ7s0b+js0pZQKOJrAKd/JWAKRsXYS72AVEgrnPQ8xTWHuk3A4C8a+b7sFVsW3f4eCI3Dec77vnhYILT7JPaF+gp0PLlgSuDWfA6bm3SddkrrDlnme2Zc3uMbnaQET5WWzlu/i31+uIX3/UQakxPPfK/vQP8WPF5iUUirA6Rg45TsZS6BZbwgJ8o+dCJz+AJz/kp277f3zqzbh99b5kDYRBt1uW2HqopAQ2wq3aY4dCxgMVs+wXR+bdPHM/pK6Qc7O46d3CCQ6hYDygXnrs7ht4lIa1gvn/esHMPmPp2jyppRSFQjyb9IqaBQeteN9grX7ZFn6XgPjPoI9qys/4XdRgS1cEtcKht7v9RADWvvhcCQbdntxjj1POZxtE++uF3iuxTTQC5lkr7fj9Bq19nckqpbKPJTH3ZOX0aFJNFNvGcTpHRMRLZijlFIV0gRO+cbuFeAoql0JHEDnUVWb8HvhK5C9Ds59BiLq+ybGQNX2THu7KQiqUa6dBabYc90n4dhciAGbwG2A+LbVG+OpVAWKih3cMfF3jhQU8+oVfagXoZNwK6VUZWkCp3zDVcCkRZBNIVAZlZ3we/82+PFp6DwaOo30bYyBKCbJTogdDAnc6hnQqA0k9/DcPqOb2HGAmQFayCR7vRYwUV7z0vcb+HXLPh6/sDsdkmL8HY5SSgUVTeCUb2QsgdjmEJPs70i8ozITfn/5F9v9buRTvo8vULUfBtt/gfxcf0dSvqP7YcuPnu0+CXZfSd0CM4ErLoR9m3X8m/KK+RuyeXnuRi7p04JL+7bwdzhKKRV0KkzgROQdEdkjIivdlj0jImtFZLmITBeRuHKeu1VEVojIMhFZ7MG4VbDJWALN+/g7Cu862YTfa2fB+i/hjAchrqX/Ygw07YaDoxC2/uTvSMq37kvb/deT3SddknvAnjV2kvtAsn+rfc2awCkP25OTx12Tl9EuMZrHL+zm73CUUiooVaYF7j2gdH+vb4HuxpiewHrgryd5/pnGmF7GmFrYd05VypF99mp+bRv/VhbXhN+dz7MTfn/7qG1d+vIv0KQrnHKrvyMMLK1OgfD6djqBQLV6BjRsZSuoelpSNyjKs38fgUQrUCovKHYY7py4jNz8Qv57ZR/qR+hMRkopVR0VJnDGmHnAvlLLvjHGFDkf/gJoHwhVvp1L7W1dSODg2ITf/a63E36/PhgO7rDzx2lBiOOFRULKaYE7Di7voI2t6xjvzNdXUokywLpRliRwQTxnowo4L8/ZwMLNe/nnmO501HFvSilVbZ4YA3c98GU56wzwjYgsEZGbT7YTEblZRBaLyOKsrCwPhKUCRsZSQKBpL39H4juuCb/PfMh2R+t1FbQ+1d9RBaZ2w2HfJpj/H9j5e2B1J1z/NRQXQNcLvbP/hE4goYFXiTJ7A0QnQ1RDf0eiaokFG7N58fsNXNy7OWP76TVfpZSqiRr1XxCRh4Ai4KNyNjnNGJMhIk2Ab0VkrbNF7wTGmDeBNwH69etnahKXCjDpi22Rj6hYf0fiW64Jvzuda7+oq7J1uwiWToDvHrOPI2Oh5UBoPci2zjXtBWER/olt9QxbfMdbrcfhUbbSY8AlcFqBUnlOVk4+d05eRtuEBjx+YXed600ppWqo2gmciFwLjAaGG2PKTLiMMRnO2z0iMh0YAJSZwKlayhhbwKTjCH9H4j+eLD1fG8Ukwa0L4NBO2LYAtv0MW3+Gjd/a9WH1oOUAaD3YJnUt+tluqt6WnwMbvrVdYUO8WLA3qRvsWOS9/VeVMTaB636pvyNRtUCxw3D35GUcOlrIBzcMoEGkjntTSqmaqtZ/UhEZCTwAnG6MOVLONg2AEGNMjvP+OcA/qx2pCk4HtsOR7Loz/k1VX2wz6HGp/QHIzYLtC21Ct+1n+OHfgIHQCPt5aj3IJnUtB0CkF8bTbPgGivO9U33SXVJ3O+1E3sHA6LJ4OMvGogVMlAf8d+5G5m/M5qmLe9A5uY71wlBKKS+pMIETkYnAGUCCiKQDj2KrTkZiu0UC/GKMuUVEmgFvGWNGAUnAdOf6MOBjY8xXXnkVKnC5JvDWBE5VVXSiLR7SdYx9fPSAnTNu28+2pW7+C/DTc3YMWdNUSBlsE7qkbhDbouatZqtn2HFgLQfW9JWcXFJ3e5u5OjDGSZYUMNEulKpmftm8l/98t54LejVjXH+dPkUppTylwgTOGDO+jMVvl7PtTmCU8/5mILVG0angl7EEQiOPVdtTqrrqxUGnkfYH7PQM6b/ZZG7rz/DrG7DgZbsuNBLi20B8O2jc1nnbzt7GNK04uSs4bLtP9rrSu90n4fhKlAGRwG2wt9oCp2ogOzefOyb+TkrjBjx5UQ8d96aUUh6kndGVd2Usta0jWj5feVpkNLQbZn8ACvPslBVZ62xVy72b7e3G72xXSJewehDf9sTErnE7iE6yxWc2fgeFR7zffRJs19GouMApZJK9wc7NF9vc35GoIOVwjns7cLSQ964bQLSOe1NKKY/S/6rKe4qLYNcy6HONvyNRdUF4lHNc3KDjlzsccCgd9m46PrHbsxbWfQWOwmPbRkTblru8Q1A/4cR9eYOI7UYZMAncemjc3vstj6rWeu3HTfy0IZt/XdSDrs103JtSSnmaJnDKe7LW2FYMHf+m/CkkBOJa2Z92Zx6/rrjILbnbfCzJKzwKp/zJzufnC0nd4PcPbbLp78Qpez206O/fGFTQ+m3LPp77Zh3npzZj/AAd96aUUt6gCZzynpICJn38G4dS5QkNg0Yp9ofh/osjuTsUHoYDW233Tn8pPGorx/a60n8xqKC1Nzef2ycupVV8ff51kc73ppRS3qJ9ZJT3ZCyBeo38+4VUqWBQUsjEz90o924CDCS0928cKug4HIZ7pqSx/0ghr1zRh5goHfeslFLeogmc8p6Mpbb7pF6FVerkErsA4v8ErmQKAa1A6QkiMlJE1onIRhF5sJxtLhOR1SKySkQ+9nWMnvLGvM38uD6Lv4/uSvfmATCfoVJK1WLahVJ5R8Fh2LMaOp/n70iUCnwR9W0VzMyV/o0jewMgtiqnqhERCQVeBc4G0oFFIjLTGLPabZsO2HlVBxtj9otIE/9EWzOLtu7j2W/WcV7Pplw1sJW/w1FKqVpPW+CUd+xKA+PQAiZKVVZSt8BogYtraRNKVVMDgI3GmM3GmAJgElB6XoqbgFeNMfsBjDF7fBxjje0/XMAdE3+nRaN6PHWxzvemlFK+oAmc8g5XAZNmWsBEqUpJ6gH7ttgJyv0le712n/Sc5sAOt8fpzmXuOgIdReRnEflFREaWtSMRuVlEFovI4qysLC+FW3UOh+HeT9LYm1vAqzruTSmlfEYTOOUd6Ytt2fboRH9HolRwSOoGGNizxj/Hdzhg70ZN4HwrDOgAnAGMB/4nInGlNzLGvGmM6WeM6ZeYGDj/U//302bmrN3Dw6O76Lg3pZTyIU3glHe4CpgopSqnpBKln8bBHcqw8zYmdPDP8WufDMB9IrQWzmXu0oGZxphCY8wWYD02oQt46fuP8MzX6xjVI5mrT2nt73CUUqpO0SImyvNy98DB7TDwj/6ORKngEdcKImL8Nw5OK1B62iKgg4i0wSZulwNXlNrmM2zL27sikoDtUrnZl0FW1xfLd1HkMPz13C467q0SCgsLSU9PJy8vz9+hqDooKiqKFi1aEB6u3ZxrC03glOdlLLW32gKnVOWJ+LeQSfYGe6sJnEcYY4pE5DbgayAUeMcYs0pE/gksNsbMdK47R0RWA8XA/caYvf6LuvJmLd9Fass4WsZrwZvKSE9PJyYmhpSUFE14lU8ZY9i7dy/p6em0adPG3+EoD9EETnlexhKQUGja09+RKBVckrrBiqlgjO/nT8xeD1ENoUHgjLEKdsaY2cDsUssecbtvgHucP0Fja/ZhVmQc5KFRXfwdStDIy8vT5E35hYjQuHFjAqkAkqo5HQOnPC9jCTTpChEN/B2JUsEluTvkH4SDOyre1tNcFSj1C6aqwKwVuwAY1bOpnyMJLpq8KX/Rz17towmc8ixjbALXXKcPUKrKkrrbW390o8zeoN0nVaV8sXwXfVrF0Tyunr9DUUqpOkkTOOVZ+zZD3gEd/6ZUdTRxdknzdSXKvIOQu1srUKoKbcrKZc2uQ4zu2czfoSgv+eGHH1iwYIFPjjVq1CgOHDhQ5ee999573HbbbZ4PSKkgoWPglGe5JvDWBE6pqouMgUYpvm+By95ob7UFTlVg1vJdiMCoHtp9srb64YcfiI6OZtCgQV47hjEGYwyzZ8+ueOMA5nodISHaHqJ8q1IJnIi8A4wG9hhjujuXxQOTgRRgK3CZMWZ/Gc+9BnjY+fAJY8z7NQ9bBayMJRBeHxI7+zsSpYJTUnc/JHA6hYCqnFnLd9G/dTzJDaP8HUrQ+sfnq1i985BH99m1WSyPnt/tpNtMmDCBZ599FhGhZ8+eXHbZZTzxxBMUFBTQuHFjPvroI44ePcrrr79OaGgoH374IS+//DKdO3fmlltuYfv27QC88MILDB48mKysLK644gp27tzJqaeeyrfffsuSJUtISEjg+eef55133gHgxhtv5K677mLr1q2MGDGCgQMHsmTJEmbPns3pp5/O4sWLSUhIOCG+Dz74gM8///yEGJOSkip8P8p7Xm5uLrfffjuLFy9GRHj00Ue55JJL+Oqrr/jb3/5GcXExCQkJfP/99zz22GNER0dz3333AdC9e3e++OILgBNex1NPPcWiRYs4evQol156Kf/4xz8AWLRoEXfeeSeHDx8mMjKS77//nvPOO4+XXnqJXr16AXDaaafx6quvkpqaWq3fvaqbKtsC9x7wCjDBbdmDwPfGmKdE5EHn47+4P8mZ5D0K9AMMsEREZpaV6KlaImMJNOsNodq4q1S1JHWDdbOh8CiE+2iMUfZ6CAmzrX9KlWNDZg7rMnP4x5iTJwoq8KxatYonnniCBQsWkJCQwL59+xARfvnlF0SEt956i6effprnnnuOW2655bjE5YorruDuu+/mtNNOY/v27YwYMYI1a9bwj3/8g2HDhvHXv/6Vr776irfffhuAJUuW8O677/Lrr79ijGHgwIGcfvrpNGrUiA0bNvD+++9zyimnVBgf2OSmrBgrUt7zHn/8cRo2bMiKFSsA2L9/P1lZWdx0003MmzePNm3alBz7ZEq/jieffJL4+HiKi4sZPnw4y5cvp3PnzowbN47JkyfTv39/Dh06RL169bjhhht47733eOGFF1i/fj15eXmavKkqq9S3bGPMPBFJKbX4AuAM5/33gR8olcABI4BvjTH7AETkW2AkMLF64aqAVlQAu5bDwJv9HYlSwSupOxgH7Fnju2JA2eshvi2E6iSvqnxfOLtPntsj2d+hBLWKWsq8Yc6cOYwdO5aEhAQA4uPjWbFiBePGjWPXrl0UFBSUO0fYd999x+rVq0seHzp0iNzcXObPn8/06dMBGDlyJI0aNQJg/vz5XHTRRTRoYCtRX3zxxfz000+MGTOG1q1bn5C8lRcf2PnzKhNjaeU977vvvmPSpEkl2zVq1IjPP/+coUOHlmzjOvbJlH4dU6ZM4c0336SoqIhdu3axevVqRISmTZvSv39/AGJjYwEYO3Ysjz/+OM888wzvvPMO1157baVek1LuatJpN8kYs8t5fzdQVpt2c8C9Hna6c9kJRORmEVksIot1roogtWcVFOfr+DelaiLJ+eXOl90otQKlqoAxhlkrdjGwTTxNYrT7ZG1w++23c9ttt7FixQreeOMN8vLyytzO4XDwyy+/sGzZMpYtW0ZGRgbR0dHVOqYrqfN0jJ56nruwsDAcDkfJY/d9uL+OLVu28Oyzz/L999+zfPlyzjvvvJMer379+px99tnMmDGDKVOmcOWVV1Y5NqU8MurSORmpqeE+3jTG9DPG9EtM1Ilkg5IWMFGq5hq1seNIfZXAFRfa6rFagVKdxLrMHDbuyeU8rT4ZlIYNG8Ynn3zC3r17Adi3bx8HDx6keXN7Tf3994+VJ4iJiSEnJ6fk8TnnnMPLL79c8njZsmUADB48mClTpgDwzTffsH+/HR0zZMgQPvvsM44cOcLhw4eZPn06Q4YMqXJ8QLkxVqS855199tm8+uqrJY/379/PKaecwrx589iyZctxx05JSWHp0qUALF26tGR9aYcOHaJBgwY0bNiQzMxMvvzySwA6derErl27WLRoEQA5OTkUFRUBdlzgHXfcQf/+/UtaLpWqipokcJki0hTAebunjG0ygJZuj1s4l6naKH0JNEiEhi0r3lYpVbaQEGjS1XdTCezfBo5CbYFTJzVr+S5CBM7trt0ng1G3bt146KGHOP3000lNTeWee+7hscceY+zYsfTt27ek6yLA+eefz/Tp0+nVqxc//fQTL730EosXL6Znz5507dqV119/HYBHH32Ub775hu7du/PJJ5+QnJxMTEwMffr04dprr2XAgAEMHDiQG2+8kd69e1c5PqDcGCtS3vMefvhh9u/fT/fu3UlNTWXu3LkkJiby5ptvcvHFF5Oamsq4ceMAuOSSS9i3bx/dunXjlVdeoWPHsv9Hpqam0rt3bzp37swVV1zB4MGDAYiIiGDy5MncfvvtpKamcvbZZ5e0zPXt25fY2Fiuu+66Sr8mpdyJbTyrxIZ2DNwXblUonwH2uhUxiTfGPFDqOfHAEsA1kGMp0Nc1Jq48/fr1M4sXL67SC1EB4JUBEN8Grpjs70iUCm4z74A1n8MDm0HEu8daOxsmjYcbv4cW/bx7rDKIyBJjjO8PHKT8cX40xjDsuR9pFhfFRzeeOH5JVWzNmjV06dLF32F4VH5+PqGhoYSFhbFw4UL+9Kc/lbTOqZPbuXMnZ5xxBmvXrvXZFAS18TNYF5R3jqzUp0ZEJgILgU4iki4iNwBPAWeLyAbgLOdjRKSfiLwF4EzUHgcWOX/+WVHypoJU3kFbCEG7TypVc0nd4eg+yNlV8bY15ZpCoHF77x9LBaXVuw6xJfuwTt6tjrN9+3b69+9Pamoqd9xxB//73//8HVJQmDBhAgMHDuTJJ5/U+eNUtVW2CuX4clYNL2PbxcCNbo/fAd6pVnQqeOxcBhhN4JTyhOTu9jZzFcR6+Utz9gaIToJ6cd49jgpas5bvIjREGNFNu0+qYzp06MDvv//u1xiefPJJPvnkk+OWjR07loceeshPEVXsD3/4A3/4wx/8HUbtcnQ/hEX5buqdAKCTdSnPcBUwaXbyfu5KqUpo0tXeZq6EDmd791jZ63X8myqXMYYvlu9icPsE4htE+DscpY7z0EMPBXSypnzk7RF22p2LXvd3JD6jbbfKMzKWQHw7qF/x/ClKqQrUi7PFgLxdidIYZwKnFShV2VZmHGL7viOM7tHU36EopdSJjuyD7HWwarodzlNHaAKnPCNjqXafVMqTkrp5P4E7nA15B7QFTpXri+U7CQ/V7pNKqQDlOk8W5cHqGf6NxYc0gVM1d2gn5OzUBE4pT0rqZlvHivK9dwxXARNtgVNlcHWfPK19Ag3rh/s7HKWUOpFryp3oZEib5N9YfEgTOFVzOoG3Up6X1A0cRZC1znvHKEngtAVOnWjZjgNkHDiqk3crpQLX7pVQPwEG3AjbfrZzm9YBmsCpmstYAiHhkNzD35EoVXskOf+evNmNMnsDhNWD2BbeO4YKWrOW7yIiNISzuyb5OxTlY9HR0R7b12effcbq1as9tr+TGTRoULWe99hjj/Hss896OBrlE5krbeXmnnYCdpZP8W88PqIJnKq5jCX2jyc8yt+RKFV7xLe1ZZFd3UO8Ye8GSGgPOheRKsXhMMxasYuhHRNoWE+7T6rq80UCV1RUBMCCBQu8ehxvc70OVUnFRbBnjZ07Na4VpAyBtIm2QFctp2dtVTMOB2T8rt0nlfK00DBI7OzlFjidQkCV7fcd+9l1ME8n7/aGLx+Ed8/z7M+XD570kA8++CCvvvpqyePHHnuMJ554guHDh9OnTx969OjBjBmVLwDxf//3f/To0YPU1FQefNAe+3//+1/JxN6XXHIJR44cYcGCBcycOZP777+fXr16sWnTJjZt2sTIkSPp27cvQ4YMYe3atQBs2rSJU045hR49evDwww+XtAIaY7j//vvp3r07PXr0YPLkyQD88MMPDBkyhDFjxtC1q516xb3lsLIxVkZ5z8vMzOSiiy4iNTWV1NTUkgRywoQJ9OzZk9TUVK6++moArr32WqZOnVqyT1esZb2OCy+8kL59+9KtWzfefPPNkud89dVX9OnTh9TUVIYPH47D4aBDhw5kZWUB4HA4aN++fcnjWm/vRijOP9YDLPVy2LcJ0hf7Ny4f0ARO1czeDVCQowmcUt6Q1N17CVxhnh0roAmcKsMXy3cRERbC8C5N/B2K8oBx48YxZcqxrmVTpkzhmmuuYfr06SxdupS5c+dy7733YirRcvHll18yY8YMfv31V9LS0njggQcAuPjii1m0aBFpaWl06dKFt99+m0GDBjFmzBieeeYZli1bRrt27bj55pt5+eWXWbJkCc8++yy33norAHfeeSd33nknK1asoEWLY926P/30U5YtW0ZaWhrfffcd999/P7t27QJg6dKlvPjii6xfv77aMVZGec+74447OP3000lLS2Pp0qV069aNVatW8cQTTzBnzhzS0tJ48cUXK9x/6dfxzjvvsGTJEhYvXsxLL73E3r17ycrK4qabbmLatGmkpaXxySefEBISwlVXXcVHH30EwHfffUdqaiqJiYmVel1Bz9VDJam7ve0yxg4LSJvov5h8RCfyVjWjBUyU8p6kbrDsQ8jdA9Ee/iK9bxNgtAKlOoHDYZi9YhdndkokJkq7T3rcuU/5/JC9e/dmz5497Ny5k6ysLBo1akRycjJ333038+bNIyQkhIyMDDIzM0lOPvmUEd999x3XXXcd9evXByA+3s7/unLlSh5++GEOHDhAbm4uI0aMOOG5ubm5LFiwgLFjx5Ysy8+3lXYXLlzIZ599BsAVV1zBfffdB8D8+fMZP348oaGhJCUlcfrpp7No0SJiY2MZMGAAbdq08WiMZSnveXPmzGHChAkAhIaG0rBhQyZMmMDYsWNJSEg47tgnU/p1vPTSS0yfPh2AHTt2sGHDBrKyshg6dGjJdq79Xn/99VxwwQXcddddvPPOO1x33XWVek21wu4VtgaD60JkVCx0Pg9WToOR/4awSP/G50WawKmayVgCETHQWL8EKuVxSd3sbeZKiB7m2X1rBUpVjsXb9pN5KF+rT9YyY8eOZerUqezevZtx48bx0UcfkZWVxZIlSwgPDyclJYW8vLxq7//aa6/ls88+IzU1lffee48ffvjhhG0cDgdxcXEsW7as+i/ETYMGDTweoyef5y4sLAyHwwHY96GgoKBknfvr+OGHH/juu+9YuHAh9evX54wzzjjp76Vly5YkJSUxZ84cfvvtt5LWuDohcyUkdoKwiGPLUsfDyqmw4Rvocr7/YvMy7UKpqqfgiJ1vY+0saN5biyAo5Q2ubiHe6EaZvQEQiG/n+X2roPbF8p1EhYcwvLN2n6xNxo0bx6RJk5g6dSpjx47l4MGDNGnShPDwcObOncu2bZUrv3722Wfz7rvvlowD27dvHwA5OTk0bdqUwsLC45KImJgYcnJyAIiNjaVNmzZ88skngB3flpaWBsApp5zCtGnTAJg06dh8XkOGDGHy5MkUFxeTlZXFvHnzGDBggEdjrEh5zxs+fDivvfYaAMXFxRw8eJBhw4bxySefsHfv3uOOnZKSwpIlttfSzJkzKSwsLPNYBw8epFGjRtSvX5+1a9fyyy+/lLw/8+bNY8uWLcftF+DGG2/kqquuYuzYsYSGhlb6dQW93SuPnSdd2p4B0Um1fk44/dYdTIryIW0yZG/0Xww7l8EX98BznWD6HyG8Hgy933/xKFWbNWgMMU29lMCth7iWEFHf8/tWQavYYZi9YjfDOjehQaR20qlNunXrRk5ODs2bN6dp06ZceeWVLF68mB49ejBhwgQ6d+5cqf2MHDmSMWPG0K9fP3r16lVSfv/xxx9n4MCBDB48+Lh9XX755TzzzDP07t2bTZs28dFHH/H222+TmppKt27dSoqnvPDCCzz//PP07NmTjRs30rBhQwAuuuiikoIgw4YN4+mnn66wm2dVY6xIec978cUXmTt3Lj169KBv376sXr2abt268dBDD3H66aeTmprKPffcA8BNN93Ejz/+SGpqKgsXLiy39XDkyJEUFRXRpUsXHnzwQU455RQAEhMTefPNN7n44otJTU1l3LhxJc8ZM2YMubm5dav75OFsyN1tq6C7Cw2DHmNh/ddweK9/YvMBqcyAVV/r16+fWby49leQqbLvHoP5/7H3EzpB51HQeTQ06+PdFrCj+2HFVFj6vu1vHBYFXS+EPldD68Eg4r1jK1XXfXgJ5GbCLfM9u983hkKDRLhqmmf3W0UissQY08+vQQQRb58fF2zK5or//cqrV/ThvJ5NvXacumbNmjV06dLF32EEtCNHjlCvXj1EhEmTJjFx4sQqVcasyxYvXszdd9/NTz/9VO42te4zuGkufHAhXP0ZtDvz+HW7V8Lrg2HUszDgJn9E5zHlnSP18lqw2LUcfn4Jul8KLQfC2i/s4/n/gehkm8x1Og/aDPHMoE1jYOt8WDoB1syEojxI7mn/GHqMhXpxNT+GUqpiSd1gyzwoLoRQDxWUcDhsF8rWgz2zP1VrzFq+i3rhoQzT7pPKx5YsWcJtt92GMYa4uDjeeecdf4cUFJ566ilee+21ujX2DY5VoHRNIeAuubvtWpk2MegTuPJoAhcMiotg5u1QvzGc9yzUawQDb4Yj+2DDtzaZS5sMi9+xBUU6nG2r8HQ4G6IaVu1Yh3ZB2sew9APYvwUiG0Lvq6D31dCsl1denlLqJJK6Q3GBne+miYeunubshMIjWoFSHaeo2MFXK3czvEsT6kXUoXE0qkwrVqwomcPMJTIykl9//dUrxxsyZEjJeDh/+fOf/8zPP/983LI777wzoLsmPvjggyXz3NUpmatsA0aDhLLXp14O3zxsL1bWwnOdJnDB4NfXYNcyGPu+Td5c6sdD6jj7U5gHW360ydy6L2HVp7a0apsh0GmUTehiy6koVlxkq/UsnWBvTTG0Pg3O+Ct0HWPHuSml/MNViXL3Ss8lcFqBUpXhl8372Hu4QCfv9hJjDBJEQw569OjhsWqRwcJ9svPaJBCHS9XY7pUnjn9z12MsfPuILWYy/O++i8tHqp3AiUgnYLLborbAI8aYF9y2OQOYAWxxLvrUGPPP6h6zTtq3BeY8abtHdr2g/O3Co6DjCPvjKLaz0K/9wlaJnH2f/WnWxyZync+DxM6wbzP8/gEsm2gHgkYnweA7bGtbY61Mp1RASOhoL8ZkrgTGVrh5pWRvOLZv5TUiMhJ4EQgF3jLGlDkBmIhcAkwF+htj/DYAfNaKnTSICOWMTnVkEmAfioqKYu/evTRu3DiokjgV/Iwx7N27l6ioKH+H4jlFBZC1FtoPL3+bmGRoNwyWT4YzH6p11dKrncAZY9YBvQBEJBTIAKaXselPxpjR1T1OnWYMfHGXHfdy3rOVLxYSEgqtBtqfs/9pr7a7krk5j9uf6GSbtEkodDgH+vzBdrn01BgbpZRnhIbbCy6erESZvd52j26gX9S9xXlefBU4G0gHFonITGPM6lLbxQB3At7pl1ZJhcUOvly5m7O7JhEVrt0nPa1Fixakp6eTlZXl71BUHRQVFUWLFi38HYbnZK8HR2HZ49/cpY6HaTfAtp9tj7RaxFNdKIcDm4wxlZtERFVO2kTY/AOc93z53R8rImInOUzsBEPutWPc1s223S2bpkLqFRCrlcaUCmiuQiaekr3ejgnQlgBvGgBsNMZsBhCRScAFwOpS2z0O/B/g1/lYFmzay4EjhTp5t5eEh4fTpk0bf4ehVO3gKmBSeg640jqNsrUh0ibVugTOU+2JlwMTy1l3qoikiciXItLNQ8er/XL3wFd/hVanQl8PDp6NbQr9b4DLJtiETpM3pQJfs9628MjHl8P2X2q+v+wN2n3S+5oDO9wepzuXlRCRPkBLY8ysk+1IRG4WkcUisthbLThfpO0kJjKMoR3LKQiglFKBYvcKCI2Exu1Pvl1EfTv8aPUMKDjim9h8pMYJnIhEAGOAT8pYvRRobYxJBV4GPjvJfrx+ggoqXz1oq8Sd/1Kt67erlKqiftfZokI7foV3RsDbI2yxIoej6vvKOwQ5u2plVa5gIiIhwPPAvRVta4x50xjTzxjTLzHR891eC4ocfL1qN2d3SyIyTLtPKqUCXOZKaNLZTtpdkdTLoSDH9j6rRTyRGZwLLDXGZJZeYYw5ZIzJdd6fDYSLSJmX97x9ggoq676CldNg6AOQqFfJlarzwiLhjAfh7pVw7tNwaCdMvBxeOxV+/8gO6K6svVrAxEcygJZuj1s4l7nEAN2BH0RkK3AKMFNEfD6p+fyNWRzKK2K0TtytlAp0xtgKlEkVjH9zaT0YGra0w5JqEU8kcOMpp/ukiCSLs9ySiAxwHm+vB45Ze+XnwKx7oElXGHynv6NRSgWSiAYw8I9wx1K4+C0ICYMZt8JLvWDBK/b/R0W0AqWvLAI6iEgbZ0+Vy4GZrpXGmIPGmARjTIoxJgX4BRjjjyqUXyzfRWxUGKe1r+MXT5VSgS83E45kn3wKAXchIdBzHGyaAzm7vRubD9UogRORBtgKW5+6LbtFRG5xPrwUWCkiacBLwOWmVk5G4UHf/9NeXR/zMoRF+DsapVQgCg2HnmPhlvlw5TSIbwvfPAT/6QbfP27H0JYne71N/OK1oII3GWOKgNuAr4E1wBRjzCoR+aeIjPFvdMfkFxXz7apMRnRLJiJMu+t7VHERbPnJ31EoVbtUtoCJu9TLwThgRVmjvYJTjapQGmMOA41LLXvd7f4rwCs1OUadsv1X+O1/MPAWaOHzXjRKqWAjAh3Osj/pS+DnF+Cn52DhK9DrChh0u03u3GWvh0ZtdMoQH3AOHZhdatkj5Wx7hi9iKm3e+mxy8osYnarVJz3ul1ftRMI3fAct+/s7GqVqh93OBK6yLXBgx3w37wtpk+15sRbQy22BoigfZt4ODVvAsIf9HY1SKti06AvjPoDbFtvuIr9/CC/3hU+uhZ3Ljm2nFSiVm1nLd9KofjiD2jWueGNVeQVHYMHL9v5WD04BolRdl7kSYltAvUZVe17qeMhcYStY1gKawAWK+f+B7HUw+j8QGe3vaJRSwSqhPYx5Ce5aAYPugI3fw5unw4QLYON3sHeTVqBUAOQVFvPt6kxGdk8mPFS/DnjUkvfgcBZENoStP/s7GqVqj90rq9b65tLtYggJt3PC1QL6HzsQ7FkL856FHpdBh7P9HY1SqjaISYaz/wF3r4Kz/2n/z3x4CTgKtQVOAfDDuiwOFxRzXg/tPulRhXnw84uQMsSOVd3xqx0Pp5SqmcI8OwygKuPfXBo0ho4j7Di4WvD3qAmcvzkctutkZAyM/Le/o1FK1TZRsbai7V3LbXGkTqOg/XB/R6UCwBfLd9K4QQSntI33dyi1y+8fQO5uGHo/tB4EBbmwO83fUSkV/LLWgimGpG7Ve37PcbaK5ZYfPBqWP2gC52+L34b032DkU9CgzCnylFKq5sIioc8fYPxE2zqn6rSjBcV8v2YPI7snE6bdJz2nqADmvwAtB0KboXYOKoBtC/wallK1gqsCZXIl54ArreMIiIqrFd0o9b+2Px1Mh+8eg3bDoedl/o5GKaVUHTF33R6OFhZznk7e7VlpE+FQOgx9wFaJjUmG+HY6Dk4pT9i9EsLqnVhdubLCIqH7JbDmC8g75NnYfEwTOH8xBmbda+elGP0f+49eKaWU8oEvlu8kITqSgW20+qTHFBfB/OehWe/juymnDIbtC+yQCaVU9WWuhKSuEBJa/X2kjoeio7Bmpufi8gNN4Pxl1aew/isY9ndo1Nrf0SillKojDucXMWftHkb1SCY0RC8eesyKT2D/1mOtby6tB0PeQdizym+hKRX0jLFTAFSngIm7Fv1sq3iQd6PUBM4fjuyD2Q9Asz4w8I/+jkYppVQd8v3aPeQVOhjdU6tPeoyjGH56FpJ6QKdzj1/nGgen3SiVqr5DOyHvQPXHv7mIQOrlsPUnOLDdI6H5gyZw/vDNw/ZDOOblmjUDK6WUUlU0a/lOkmIj6de6ihPhqvKtmg57N8LQ+04cEhHXEhq2gm2awClVba4CJjVtgYNjdSeWT6n5vvxEEzhf2zQXln0Eg++q3kSESimlVDXl5BUyd10Wo3o0JUS7T3qGw2Hnck3sDF3GlL1NymBbidIY38amVG2xe4W9re4UAu4apdiW8bRJQfs3qQmcLxUcgc/vhMbt7fwwSimllA99v2YPBUUORmv1Sc9Z+zlkrYEh90FIOV+rWg+CI9l2EmKlVNVlroS41nZuU09IvRz2boCMpZ7Zn49pAudLP/wLDmyD81+C8Ch/R6OUUqqOOaNTIs+NTaV3S+0+6RHGwLxnbFGE7heXv13JOLj5volLqdpm98qaj39z1/UCCIuyU38EIU3gfGXn77DwVeh7ne1KoZRSSvlYXP0ILunbQrtPesr6r2zXriH3nnxMe3xbiE7WCb2Vqo6CI7Bvk2e6T7pENYROo2DlNCgq8Nx+fUQTOF8oLoSZt0ODJnD2P/wdjVJKKaVqyhj48WmIa3WsKEJ5RJzj4H4O2jE3SvnNnjV23mRPFDBxlzoeju6Djd96dr8+oAmcLyx81V6hO+85m/ErpZRSKrht+h52LoXT7oHQ8Iq3bz0IcnbB/i3ej02p2iTTWcDE08X/2g2DBolB2Y1SEzhvO7ADfvw/6HQedBnt72iUUkopVVPGwI/PQGxz6HVF5Z7T+jR7q/PBKVU1u1dCRDTEpXh2v6Fh0OMyWPeVnaM5iGgC521f/9X+oz/3KX9HopRSSilP2PoT7PgFTrsbwiIr95zETlC/sY6DU6qqMlfa8W/lVXmtidTLwVEIqz71/L69qMbvhIhsFZEVIrJMRBaXsV5E5CUR2Sgiy0WkT02PGTQ2fAtrPofT77d95JVSSikV/H582hYl6X115Z8jYrtRbtNKlEpVmjGQucrz499ckntAk66QNtk7+/cST6WyZxpjehlj+pWx7lygg/PnZuA1Dx0zsBXmwez7IaEjnHq7v6NRSimllCds/8W2wA2+o+pTArU+DQ5st8MrlFIVO7Ad8g95fvybi4hthUv/DfZu8s4xvMAXXSgvACYY6xcgTkRq/wyiP79gByqPehbCIvwdjVJKKaU84cenoX6CnRaoqloPsrfajVKpyslcaW+TPDgHXGk9LgMJgbRJ3juGh3kigTPANyKyRERuLmN9c8D9UlO6c9lxRORmEVksIouzsrI8EJYf7dsMPz0P3S+Btqf7OxqllFJKeUL6Elt9ctBtEFG/6s9P6marUW/TQiZKVcrulYBAUlfvHSO2KbQ9A5ZPAofDe8fxIE8kcKcZY/pgu0r+WUSGVmcnxpg3jTH9jDH9EhMTPRCWnxgDsx+A0Ag450l/R6OUUkopT5n3DNRrBP1vrN7zQ0Kh1amawClVWZkrIL4tRDTw7nFSx9vumtsXevc4HhJW0x0YYzKct3tEZDowAJjntkkG0NLtcQvnstpp7Rd2QsAR/7YZvVJKKaWC367lsP5LOPMhiIyp/n5aD4b1X0FOJsQkeS6+YJSz2yaz3S62Y5Fqk9Uz7GfGE1IG2znLAkF+ji3Q12Ns5eY/rKndK22hEW/rfB6EN4C0j+37HeBqlMCJSAMgxBiT47x/DvDPUpvNBG4TkUnAQOCgMWZXTY4bsAoOw5cP2ko5A8rqTaqUUkqpoDTvGYiMrfn5vbXzy+G2n6H7xTWPK5h9/TdYOQ2K8is/n14wOLIPpt4AjiI7tqomjMPWVbjhG2je1yPhVT8WAzNug9Wf2deVerl3j5efY+tJ+OKzEdEAel4GSydAn2ug5QDvH7MGatoClwRMF3vVJAz42BjzlYjcAmCMeR2YDYwCNgJHgGqM+g0S856BQ+lw6dt2ckCllFIqkOTstkU4Rvyr6hUU67LM1bBmJgy9H+rF1WxfTVPtlf5tC+p2Apd3ENbOgpAwmHUvtOgPCR38HZVnrPrUzi32x5+gac+a7evofnh9CEy93u4vKtYzMVbH0vdt8hYSBmkTvZ/AZa62t96aQqC0s/9hx7hOvQFu+anmf+teVKPLAsaYzcaYVOdPN2PMk87lrzuTN5zVJ/9sjGlnjOlhjDlhrrhaIWsdLHgZel0JrU7xdzRKKaXUiXb8Covfhmk3QHGRv6MJHj89CxHRcMqtNd9XaBi0Gqjj4FbPgKI8GPsehEXB1OvsFEy1QdokaNLNM13/6jWCS96yU098cZdtBfOHPWtsL7O2Z8Bp98DmH+Ggl0dEZa6wt96aQqC0qIZw6buQsxNm3u6/97oSfDGNQO1njL16FBENZ5fuQaqUUkoFiK4XwMj/s+O1v7gzoL+gBIzsDbDyU+h/A9SP98w+Ww+GPattV7u6Km0yNG4PnUfDha/B7hXw7SP+jqrm9m6C9EWQOs5z4/panQJn/tV2N/39Q8/ssyoKj8In10FkNFz0prPlzcCKT7x73MxVNqlq2LLibT2lRT8Y9nfb4r7kXd8dt4o0gfOEldPspJ7DH4EGCf6ORimllCrfKbfA6X+xXwS/e9Tf0QS+n56zLUSn3u65fZaMg6uj88Ht3wbb5ttEQAQ6jbStm7+9AWtn+zu6mkmbZMeH9bjMs/s97R5oMxRm3297ffnS13+DrDVw0eu28E7jdtByoO1G6c2LQLtX2u6Tvi5wM+gOWzTmq78e68YZYDSBq6m8g/aD3aw39L3W39EopZRSFTvjr7YU/s8vwvwX/B2NZxzMsFfsPfmFct8WWD4F+l0H0R6c4qh5H5sU1tVulMun2Nue444tO+sxSO4JM26Fg+l+CavGHA47l1jbMzxfiTwk1LZ+RdS34+EKj3p2/+VZ9RksfgcG3wntzzq2PPVyyFoLu9K8c1yHw/49+2r8m7uQELjoDVu0aOp1UHDE9zFUQBO4mpr7b8jdA+c9b/+4lFJKqUAnAuc+A90vsa1wSyf4O6KaObQL3hoOrw2CV/rB3H/BnrU13+/8523BhkF31Hxf7sIibdGOupjAGWNbblKGQFyrY8vDIu14uOJCmHZTcI7R3L7QziWWOt47+49tChe+Dpkr4ZuHvXMMd/u3wcw7bPXLYX8/fl23i+ycx2mTvHTsLVB42Hfj30qLbgIXv2GT1K//6p8YTkITuJrYtdw29/e73l5NU0oppZxEZKSIrBORjSLyYBnr7xGR1SKyXES+F5HWPg0wJMR+GWw3HD6/087tFIwK82DylZB3yI5Dj21mq0L/dyD891T48Rk7LqmqDuyAZROhz9Xemde19WA77ivvoOf3HcjSF8O+TWVXMGzczl4Q377A/g6DTdpEWw+h83neO0bHc+DU22DRW979my0uhGk3AgYuefvEOd/qNYKOI+04uOJCzx8/c6W99UcLnEu7YXDa3bDkPTsONoBoAlddDoctXFIvHob/veLtlVJK1RkiEgq8CpwLdAXGi0jXUpv9DvQzxvQEpgJP+zZKICwCxn1gr7BPvQG2zPN5CDViDHx+B2QsgYvftN28rvkc7llrWxijGsLcJ+DlPvDGUNtd9MD2yu375xfs7eC7vBN760F2jq/tv3pn/4Fq+STbfbTLmLLXp46zLVjznoat830bW00UHrWVNbuMsXOKedPwR+3QnRl/rvznuarm/gvSf4PzX4D4NmVvkzoejmTDpjmeP/7ulXYsYZMunt93VZz5kG0t//xO2L/Vv7G40QSuupZ9ZD/YZ//TXoVQSimljhkAbHROt1MATAIucN/AGDPXGOMaXPEL0MLHMVoRDeCKKRDfFiZeATt/90sY1bLgJVg+Gc58GLqMPrY8JgkG3gzXfwV3r4JznrRdIb97FF7oAW+dBQv/C4d2lr3fQ7tg6QfQazzEeakCXov+EBJui3nUFUX5tvBb59Enn89s1LPQqI3tSnl4r+/iq4l1syH/kPfnRgN74eWSt21jgje6m26aC/P/A33+YLtZl6f9WVC/sW159LTMlbZKaXg9z++7KkLD7TQOiL3I5Y3WxmqolQlcQZGDvMJi7x3gyD5b6rbVqd7r56yUUiqYNQd2uD1Ody4rzw3Al2WtEJGbRWSxiCzOysryYIhu6sfD1Z9C/Ubw4SW2dH6gW/8NfPuoHYsz9L7yt2vYAgbdBjfNgTuW2daLojw7ruX5rvDuKPjtf5Dr9t4ueAkcRbbyn7dE1Lctn3WpEuWGb+zE1BV9d4qMhrHv2tadGX8Ojuku0iZBbAs7ts8XGrezrWM7foEfn/LcfnP3wPQ/QkJHO+XIyYRFQPdLbeXQowc8FwMcq0AZCBqlwJgXIWMxzHnC39EAtTCByyssZswr83n6Ky+WWP3+H7bP+nnP2TEESimlVDWJyFVAP6DMQT/GmDeNMf2MMf0SEz1YCbG02GZw9We229IHF3l/kt6ayFpnJyNP7gEX/LfyZcbj28CQe+CW+XDbYluN83A2zL4PnusIEy6wydzid6HnZeV3HfOU1oNsi2fBYe8eJ1CkTYLoJFulsSJNU+Hsx2H9l/DrG14PrUZyMmHj9/Yz48vvhT0uhV5Xwbxn7cTaNeVwwPRb7Hfcse/aiwwVSb0civNh9Wc1P77L0QNwcLv/CpiUpdtFttr8zy/Y37Wf1brsIyo8lP4p8by7YAtLt+/3/AHSl8CS92HgLZDUzfP7V0opVRtkAO5971o4lx1HRM4CHgLGGGPyfRRb+Rq3g6um2S9wH1wUmBNNH90PEy+3VQvHT6zcl8yyJHSAM/4Cf/4V/rTAtrYd2G6TuaI8GHKvZ+MuS8pg29K34zfvH8vfjuyD9V9Dj7EQGla55wz8I3Q8F779O+xc5tXwamTlVDDFvuk+Wdqop21Xw09vthcjamLhK7Dpexjxr8p/x23W27bWebIaZeYqe5vUw3P79IQR/4bELraFMifTr6HUugQO4IGRnUiOjeIvU5eTX+TBrpSOYph1t716dMYJBcWUUkopl0VABxFpIyIRwOXATPcNRKQ38AY2edvjhxjL1jTVJkb7t8JHl0J+rr8jOqa4CD65zlaIHPeR7R5ZUyL2y+rwv8PtS+HmH2whlIQONd93RVoOBAmtG9MJrJwGjsKqJTkicMGrUD/Bzn2Wn+O9+GoibZJNZBI7+f7YEQ1sa9nR/bb1zOGo3n7Sl9geZl3G2OrqlSVif6fbF9p5Ez3BlcAFUgsc2ItFY9+1n8Ppf6z+e+0BtTKBi4kK58mLurNhTy7/nVuN0sHlWfyOnbBw5L9OPvhWKaVUnWaMKQJuA74G1gBTjDGrROSfIuIqv/cMEA18IiLLRGRmObvzvZTT7JxcO5fB5Kts8YlA8O3fYfNcGP0faDXQ8/sXsV/E2/hoHFNkjE2Y68I4uLRJdkxTchVbVRo0hkv+Z+cFm32/d2KricxVsHu5f2siJPeAEU/Cxm/hl/9W/fl5B2Ha9RDTFMa8VPkuyS49LgPk2ATtNZW5wlZ5j/HC9B011aQLjHzK/h9a8JLfwqiVCRzAsM5JXNirGf/9YSPrdnvgik3uHvj+cdtvu9vFNd+fUkqpWs0YM9sY09EY084Y86Rz2SPGmJnO+2cZY5KMMb2cP+XUVfeTzqPgglfsF5VPb7a9UPxp6Qf2y+kpt9q52WqL1oPs3GiFef6OxHuyN9gCENXtYphyGgx9wFY79NbE0dWVNslWOD1ZtUZf6H+jre753WOQsbTyzzMGvrjbtmpf8nb1KqvHtbQXPdImeqbgzO6VtvWtqomkr/S9FrpeCHMet3+7flBrEziAR87vRkxUOH+ZtpxiRw0/UN8+AoVHbGnbQP1AKaWUUp7U6wpbgn/1Z3buU39VA9z+i/2S2fZMW9iiNkk5zRaByFji70i8J22SLY7TY2z19zH0fjv5+Rf3QPZGz8VWE45i2+rU4RxokODfWERgzMt2mM/U6+3E9pXx+4e2e+uZf6tZq3bPy20raU3HczqKYc+awBv/5k4Ezn8RYprB1Os8X4GzEmp1AhffIIJHz+/Ksh0HeG/B1urvaOvP9qrCoNt90ydeKaWUChSDbrMFPpa8658S2gd22G6ccS3t+JPKFsAIFq1OAaT2joNzOOxcfe2GQUxy9fcTGgYX/8+Wrp96XWB06938A+Tu9k/xkrLUj7dzlh3YBrPuqfiCS9Y62y21zelw2t01O3bXMRBWr+Zzwu3dBEVHA2/8W2n14uDSd2y13i/u8vnFrVqdwAGMSW3GsM5NePbrdWzfe6TiJ5RWXGivOjZsZa/+KKWUUnXN8EegzzXw07Ow8FXfHbfgMEwab7+sj59Uve5dga5eIzs2rLYmcNsXwMEdtoWmpho2t9NG7F5uuwr62/LJENUQOo70dyTHtD7VTo+x4hNY9nH52xUetQWBIhrAxW9CSGjNjhsZA13Oh1Wf1iy5zlxhb4Oh0nvL/jDsYVg1HZZO8Omha30CJyI8cWF3QkOEv01fgalqhvzr65C1Bs59qvqlipVSSqlgJmILh3QZA1//DZbV8Cp7ZRgDn91qx8Nc8rZ/Kvz5SutBtutZcaG/I/G8tIkQEQ2dz/PM/jqPggF/tOMh133pmX1WR34OrPnc1kUIi/RfHGUZcq+dUHz2fZC1vuxtvnkY9qyCi16vWcuou9TLbUGU9V9Vfx+7V9oxhYmdPROTtw2+y9bH+PIvtuunj9T6BA6gWVw9/nJuZ+ZvzOaTJemVf+LBDPjhKegwAjqN8l6ASimlVKALCbXds9qcDjP+7P0vz/OetWPvzv4HdDzHu8fyt5TBdpx9IM91Vh0FR2DVDFvwwZMXwc/+p628+NmtcGin5/ZbFWs+t78zf1afLE9IqO1uGl7PdjctXSBn9UxY9JYdGtThbM8dt+0ZEJ1cs0IzmSvtvHKBlhSXJyQELnoTIqNti2bhUd8ctrpPFJGWIjJXRFaLyCoRubOMbc4QkYPO8sjLROSRmoVbfVcOaMWAlHie+GI1ew5VotLT3k22z72jCM79Py1copRSSoVFwuUf2dL3n1wLc56050tPW/M5zH0Ceo6DQXd4fv+BptUge7ttvn/j8LR1s6Egx/NjxMKj4NJ3bVe9aTf5p0Jq2kRo1AZaDvD9sSsjtilc+JpNiL79+7HlB7bDzNugWR8Y5uGv5SGh0HMsbPim+pOK715puxQHk5gkuOgN22Pv67/55JA1aYErAu41xnQFTgH+LCJdy9juJ7cSyf+swfFqJCREeOqSHuQVOXh05qryN3Q44Lf/weunwb5N9gpGfBvfBaqUUkoFssgYuHKqbYn76Vl4uQ+8dZY9dx7ZV/P9714Jn/4RmveF86sxJ1Uwik6EhE61bz64tInQsKWtHulpCR3gvGdt0jvvWc/v/2QOpsOWn2xiGsifz44j4JQ/w29vwpovoLgIpt1ov+te+rYtCONpqeNt48fKaVV/7pF9kLMz8AuYlKX9cBh8p50zevUMrx+u2gmcMWaXMWap834OdqLS5p4KzBvaJkZz11kd+HLlbr5auevEDQ6mw4cX2T7DrU6FW3+xVXWUUkopdUyDxnDlFLh7tS3rX3DYnjuf7QiTrrQtaNUpZHA4GyaOh6hYGPeRbWmpK1oPstMl+Hu+PU/JyYRNc6DnZbabmTekjreTSP/4lG+T3+VTAGNbiAPdWY/aFvMZf7aVKXf8Cue/APFtvXO8pG62e2t1ulFmrnTuIwgTOIBhf4fm/WDG7bB/m1cP5ZG/KBFJAXoDv5ax+lQRSRORL0Wk3JIyInKziCwWkcVZWVmeCKtMNw1pS7dmsfx9xioOHnEOFjbGVur576mwY5EdqH3VNIht5rU4lFJKqaAX2xQG3wF/WgB//AkG/tEW45h8FTzXyc7ZtWNR5UpsFxXAlD9Abqbtphnb1PvxB5KU0yD/EOxe4Z39GwM/PQ+b5npn/6Wt+ASMwzPVJ8sjAqOfh0YptmXJEy3AFTHGJietTg2OHlphkba7qaMIlr4Pva+CHpd695ip42HnUjtNQVXsdiZwyQE8B9zJhIbblk2M/Tx6sShRjRM4EYkGpgF3GWNKzxq4FGhtjEkFXgY+K28/xpg3jTH9jDH9EhMTaxpWucJDQ/i/S3qy73ABT85eDbl77NXCz/5krxr8aT70uz6wm8SVUkqpQCICTXvCiCfhnjVw5TRoN9xeHH37LHi5L/z4NOzfWv4+vvqLLaV/wSu2+2Rd09o1Ds5L0wkseRe+/wdMvhr2bfbOMdylTbK/x8SO3j1OZIydjyt3D8y4zfvzce38HbLXBc7cb5XRuJ0dEtT1Qjj3ae8fr/ulduL2qrbCZa6EBk0guol34vKFRim2hbPwKBzZ67XD1CiBE5FwbPL2kTHm09LrjTGHjDG5zvuzgXAR8fNU9dC9eUNuHtqWQ0s/pfDlgbDxOzjnCbh2lvealJVSSqm6IDQMOpxlr0Tftx4ueNX2aJn7JLyYCu+cC0veh6MHjj1n0Vt27Mjgu2yXu7ootpktirHVCwlc5ir46q92LFpICEy93rZ4esvuFXY+L19VaGzW21YrXTfLjsX0prRJEBppk6Fg0nkUXPa+nffN22KS7AWc5ZPteLvK2r0iOMe/ldb9Erj5B89Nz1CGmlShFOBtYI0x5vlytkl2boeIDHAez3vpaGUd3c99uc/yesQLbC6I4+j1c2wp1ZpOYqiUUkqpY6JibZeta7+Au1bYMSKHs+DzO+x4uSnXwML/2jmUOoywE4bXZa0H24mvq/KltyIFR2x588hYGPueTah3/m5b47wlbRKEhNs50nzllFvtZ+ibh2DXcu8co7gQVk6FTudCvTjvHKO2SL0cDmXA1p8qt31xIWStDd7xb6WFhnl19zVpgRsMXA0Mc5smYJSI3CIitzi3uRRYKSJpwEvA5abKM2l72Ibv4L+nErrqU9J73sF5Rx/j2d/rxHR4SimllP/EtYKh98Fti+CmOdD3GtgyD77+K8S3s3PM1fULqSmD4eh+W47cU756ELLXw8Vv2q5pXc6H/jfCwldgw7eeO45LcZEd/9bhHFvsxldE4ML/Qr1428KYn+v5Y2z8znaLC8S53wJN5/PsRYPlkyu3ffYGKC6oPQmcl1U7PTTGzAdOOlDMGPMK8Ep1j+FR+bl21vkl79rZ3cdPpEWz3oyTFbz78xZG92xK71aN/B2lUkopVbuJ2LFRzfvCOU/aK/RJ3WxrXV3nKre/bYF9T2pq5TRbuOK0e6DdmceWn/MEbFsI02+BP/3s2a5eW36whWj8MUasQQJc8j94fwx8+YBN6DwpbSLUT7Al49XJhdeDrhfAqukw6pmKu266KlDWhi6UPlA3mp62LYDXBsGS9+DU2+DmH21/aeDBczuTFBvFg9NWUFDkwS4LSimllDq5sAj7ZdiLY0WCSlwriG0BWz0woff+rfD5XdCiP5xZanLh8How9l0oPAKfengi7LRJEBVn5yDzhzZDYej9sOwjZ7l/Dzm6H9Z9CT3G2mqDqmKpl0NBLqydVfG2u1dAaAQkeLnoTS1RuxO4wjz4+iF4d5S94nfdbFshy21emZiocJ64sDvrMnN47YdNfgxWKaWUUnWaiO1GuW1BzaopFhfaboQIXPJ22QlHYidbkXDLPJj/n+ofy13eITthdPdLbPl6fzn9L7bM/xd3w14PfbdbNd128Qum6pP+1moQNGxlWy4rkrnSfiY1Oa6U2pvAZSyFN4baPt79roNbfj5WoreU4V2SGJPajFfmbmB9Zo6PA1VKKaWUcmo9CA7vgb0bq7+POY9DxhIY8xI0al3+dr2vssnW3H/B9rKm8q2iNTOh6Kj/k5zQMFs2PyTMcxU30ybZIThNU2u+r7oiJARSx8HmH+DQrpNvm7kKkoJ0/jc/qH0JXHEhzP03vHWWnRDzqml2Yu7I6JM+7dHzuxIdGcZfpi2n2OHfOitKKaWUqqNan2Zvqzsf3Mbv4ecXoe910O3Ck28rYr8jxbWEaTfYboI1kTbJTsfUon/N9uMJcS1txc1dy2pecXPfZtjxq01MdZ7gqul5uZ3QfcUn5W+Tm2XHTer4t0qrfQkcwPqv7BWlWxdC+7Mq9ZTG0ZE8en43ft9+gPcXbPVufEoppZRSZWnczk5mXJ354HIyYfofoUlXGPnvyj0nqiFc8g7k7IKZt1e/6+aB7bYgTer4wElyuoyG/jfZ3ljrv6n+ftImAwI96ugchTWR0B6a97PdKMv7bGWusLdagbLSal8CFxpuJ+S+5H9Qr2pVJS/o1YwzOyXyzNfr2LHviJcCVEoppZQqR8k4uJ+rlkw5HDD9Zlt1+9J3bKGSymrRF4Y/Cms+txOqV4erYEigTcR+zhM2Mfjsloq78ZXFGJt8tD0dGjb3fHx1QerlsGe1LVRSlt2uCpTahbKyal8CBxV2lyyPiPDERT0IEfjb9BX4e8o6pZRSStVBrQfbSZAPbKv8cxa8aMcanfsUNOlS9WOeehu0Gw5f/dWOR6oKY2z3ydaDoVFK1Y/tTeFRcOm7UHi0ehU3t/9ifw8691v1db/ETuyeNqns9ZkrIaYZ1I/3bVxBrHYmcDXQPK4eD57bmZ82ZDN1Sbq/w1FKKaVUXeM+H1xl7FgE3z8O3S6CPtdU75ghIXDRG1AvDj65DgoOV/65GUth7wboOa56x/a2xI52LrKtP8H856v23LSJEF4fOo/2Tmx1Qf14O63Eik/sRO+l7V7pmXkP6xBN4Mpw5cDW9E9pxBOz1rAnJ8/f4SillFKqLknsbIeBVGYc3NEDMO16273v/BdrNv4sOtEmcdnr4asHK/+85ZMgNLLioin+1OtK6H6pLXS3bWHlnlOYB6s+gy5jqt27SzmlXm6rq26ee/zyonzIXqcFTKpIE7gyhIQIT13Sk6OFxYx/8xfunZLG89+sY/Ki7czfkM2W7MPkF3lw0kullFJKKZeQENsKV1ElSmPg8zvh0E5biCSqYc2P3e5MOO1uWDoBVk6rePuiAlgxFTqf55nje8txFTdvhCP7Kn7O+i8h/6D/p0WoDTqcYy9KlJ4TLmsdOIq0gEkVhfk7gEDVLjGaZy7tyQcLt7FgUzaZh/IoPbtAYkwkzePq0bxRPXvr+mlkf2KjdDJCpZRSSlVD60Gw9gubnMU2K3ubJe/B6s/grMegpQdL95/5N9g6Hz6/C5r1gfg25W+78Vs4ui84xohFxdrxcG+fYytujvvw5C2WaZPs2Kw2Q30XY20VFmnHwv3+IeQdPJbsZ2oBk+rQBO4kLujVnAt62YpDhcUOdh/MI+PAUTL2Hz3udvXOQ3y7OpOCIsdxz4+JCjs+qSt1m9AgkpCQACm1q5RSSqnA4T4OrselJ67fs8Z2c2x7Jgy607PHDg2HS96C14fY+eGu/9ouK0vaRGiQCO2GeTYGb2neB856FL55GBa9BQNuKnu73CzY8C0Muh1CQn0bY22VOt6+56tnQJ8/2GW7V0JYFMS3829sQUYTuEoKDw2hZXx9WsbXL3O9w2HIPpzPzgN5zsTuSEmCl77/KL9t3UdO3vEDNyPCQmjWMMqtBa/+ca15yQ2jiAjTXq5KKaVUnZPcAyJjbUtY6QSu4IgtNBIZCxe/abtcelqj1jDmJfjkGpjzOJz9zxO3ObIP1n8N/W+E0CD6SnnKn2Hzj/D1Q9Dq1LLHX62cBqZYu096UvO+0Li9bdl0JXCZK23V1GD6/AQAfbc8JCREaBITRZOYKHq1jCtzm0N5hewsowUv48BRfliXxZ6c/OO2F4GkmGMJXjNny12TmEgSYyJJjLa3UeF6ZUgppZSqVUJCodUpZVei/PpvkLUGrvoUopt4L4ZuF8KW6+HnF203wvZnHb9+1XQoLgi+JCckBC58DV4fDFOvg5t/gIgGx2+TNhGaplZvSgZVNhHoeTnMfQL2b4O4VjaB6zTK35EFHU3gfCg2KpzY5HA6J8eWuT6/qJhdB/LYeeAo6aUSvWU7DvDlyl0UFp84N11MZBiJMZEkuCV1rgQvISaCxOgoEmMiaRwdQXiotugppZRSQaH1INjwje3OF51ol62aDkvehcF3Qfvh3o9hxL/sXGjTb4FbfoaYpGPr0iZBk66Q3NP7cXhadKJtvZxwIXz5AFzw6rF1e9bArmUw8il/RVd79bzMJnDLp0Dvq+DIXh3/Vg2awAWQyLBQUhIakJLQoMz1DochOzefPTn5ZOXmk5WTT7bz1vWzZvch5m3IP6G7pkt8gwgSoiNIiI6kXngooSFCeGgIoSFCWKgQHhJCaKgQHiKEhoQQHirOdSF2mWubEHGuC6FBZCgN64UTVz/C3tYLJ7ZeOKE6vk8ppZSqvtan2dvtC6DrBbbVYuad0LwfDHvYNzGE14NL34E3z4TpN8NV020L1t5NkP6b7VpZk6kL/KntGTDkXvjpWTuW0NVVNW0SSKiddkB5VqPW9nOdNhGa9bLLtAJllWkCF0RCQoQmsVE0iY2qcNu8wuLjkzu3+67lB48WUuwwFBY7nLeGYoehyOEoue9aV1S6BGclxESFEVc/nLh6NrFrWN8mdzbZs7cN60WU3I+ODCMyLISIsBAiw0KJCAvRJFAppVTd1ayXnUR668+2m9m0GwADl75dflERb2jSBc59yk5Z8PMLMOQeZ5ITAj0u810c3nDGX+0E35/fZQucxLW2rUMdzj7W6qk8K/VymHkbLH7XPtZJvKtME7haKio8lBaN6tOiUdlFV6rKGFOSyBU5DEXFNskrcjg4nF/MwaMFHDxayIEj9ufg0ULnY+fyo4XsPHiUg0fs/eJKJoShIUJEaAiR4SFEhNrkzj3Biwx1PXZfZ9dHhdvbyLAQosJDiQwPISrM3kaGhRAZ7rau1HNct+GhggTrlUWllF+JyEjgRSAUeMsY81Sp9ZHABKAvsBcYZ4zZ6us4VQALDYeWA+x8cHOfhPRFMPY9aJTi+1j6XAObf4A5T9gKmcsnQZvTIbap72PxpNAwZ8XN02DqDXYKhZydMOJJf0dWe3W9AGbfB+tmQcOWUC/O3xEFnRolcHpyqjtEbBfLMA/USzHGcLigmANHCjhwpJBDzgQvN7+IgiIH+UUOClw/xcXkFzooKD62LN+1TbGDgqJiDhcUsf+Ia3sH+YUO8ouKyS9ykFdYfML8fVURIpQkdFHuCZ/b/ShnMhjleuyWBLqeExEWQlhICOFhIUSESsn98FDbhdX+VHzfGJtAFxY7KCo2FDpbS90T6qJi53r37UqtL3IYHA5DsXG77zA4nIl6sXEtg2KHg2Jj75esd27rMIZQEUJChLAQexsqx993rQsNEUKcn6MQsY9DnduEhggGg8MBBnscjL11PXYYwHlrXLfOz5Nx2zZEIDQkhFCB0NCQ4+I57laOxVXWT1iIlLz/Jb8vV3di5+/E9XxvJvnG7XdSVGxKfu/Hfq/H/66LSj4Tx7YrKrZ/G0XFdj9hIUJEmP1MuS6MuH/WIl2PnZ/RyNBQwsOOvWa9qFExEQkFXgXOBtKBRSIy0xiz2m2zG4D9xpj2InI58H/AON9HqwJa69PsmKHMlTaJ6naRf+IQgfNfhIwl8NFYO8H1mT7qxultca1gzCsw5WqbxEU2hE7n+juq2isq1k78vnKadp+spmoncHpyUtUlIkRHhhEdGUaLRt49livhySssLkn88gptUphXdPxtvtttXqHD7TnHP84rLCbPeXvwaCF7CouPX+fcp6lB4hhoQkuSMY5LykTEJnOlksHqdLkNZhGhIfYChzMxCguxjyNCQxABY3AmwTbZdCVkxi0hdm3jMDaRdd0PtM+RCCWJ3+3D2vPH03XunnIMADYaYzYDiMgk4ALA/Rx5AfCY8/5U4BUREWMC7beu/Kr1IHub2Nn/RTWiGtqJsN8ZAeENoMto/8bjSV3HQL8bYPHbNlEOr+fviGq31PE2gStrCgdVoZq0wOnJSQU8ESlpVYjx4XGNsWMK84pscufeWlJYwf2iYkOBq2XFuayg2EFhkUGEkmIzYaHHtxCFubUKhZckFG6tRW7rw0JCCAmh5NbVChbi1iIW4tZyVR2upM7VWufe0lfsts71EyKCCIQ4W3dCQgTBPg4RwLnOtUzE/n7FbbkIOFwJUuljFbuO6aDYAUUOBw7XrbN1yxVTkcNuX+RwUOBswXL9XoqcLV6uFi3bumVKLT+2vTH2tYTIseTXthKK87WJ231KWvRCnb8XkeNbBcNK/c4jSv2uXQWHwly/85AQwsOOrQ8NkeM+VwVFxz57xx7bH9uqbSgsOva40Pl6C4ocdG5adkVdBUBzYIfb43RgYHnbGGOKROQg0BjIdt9IRG4GbgZo1aqVt+JVgarlADj1Nuh7LUR4ZlhEjbToZ7txFuadWHo/2I14Euo1gr7X+DuS2q/tmbaATOp4f0cSlGqSwHns5AR6glK1i4gQEWZbY2KjfDjQPICEhAghCDpNoVI1Y4x5E3gToF+/fnoBtK4JDQ+88Vhdzvd3BN4RXg+G/93fUdQNoWEw/BF/RxG0AmZSMGPMm8aYfsaYfomJWvVHKaVUUMsAWro9buFcVuY2IhIGNMSOF1dKKaXKVZMETk9OSimlVNkWAR1EpI2IRACXAzNLbTMTcPXVuhSYo0MMlFJKVaQmCZyenJRSSqkyGGOKgNuAr4E1wBRjzCoR+aeIjHFu9jbQWEQ2AvcAD/onWqWUUsGk2mPgnGPaXCenUOAd18kJWGyMmYk9OX3gPDntwyZ5SimlVK1njJkNzC617BG3+3nAWF/HpZRSKrjVaB44PTkppZRSSimllO8ETBETpZRSSimllFInpwmcUkoppZRSSgUJTeCUUkoppZRSKkhIIBaFFJEsYFsNd5NAGROGB7hgjBmCM+5gjBmCM+5gjBmCM+5gjLm1MUYn/6ykOnx+hOCMOxhjhuCMOxhjhuCMOxhjhuCMu8xzZEAmcJ4gIouNMf38HUdVBGPMEJxxB2PMEJxxB2PMEJxxB2PMyveC9XMSjHEHY8wQnHEHY8wQnHEHY8wQvHGXRbtQKqWUUkoppVSQ0AROKaWUUkoppYJEbU7g3vR3ANUQjDFDcMYdjDFDcMYdjDFDcMYdjDEr3wvWz0kwxh2MMUNwxh2MMUNwxh2MMUPwxn2CWjsGTimllFJKKaVqm9rcAqeUUkoppZRStYomcEoppZRSSikVJII+gRORkSKyTkQ2isiDZayPFJHJzvW/ikiKH8J0j6eliMwVkdUiskpE7ixjmzNE5KCILHP+POKPWEsTka0issIZ0+Iy1ouIvOR8r5eLSB9/xOkWTye393CZiBwSkbtKbRMQ77WIvCMie0RkpduyeBH5VkQ2OG8blfPca5zbbBCRa/wc8zMistb5+58uInHlPPeknyVvKifux0Qkw+1zMKqc5570/42PY57sFu9WEVlWznP99l4r/wq286MzpqA8Rwbb+dEZk54jfR9zQJ8jg/H86Dx23TtHGmOC9gcIBTYBbYEIIA3oWmqbW4HXnfcvByb7OeamQB/n/RhgfRkxnwF84e/3t4zYtwIJJ1k/CvgSEOAU4Fd/x1zqs7IbOyFiwL3XwFCgD7DSbdnTwIPO+w8C/1fG8+KBzc7bRs77jfwY8zlAmPP+/5UVc2U+S36I+zHgvkp8hk76/8aXMZda/xzwSKC91//f3v28WFXGcRx/f0Ej0LBfoKYtMlq16AciFdbGMI1wKiKMoB8KIeSiVRt3/QFtIlr0Ay2kot+zMNJq0WoqGtIKI6cgVMYRMrRoUda3xfPcOJ45Z+YGzjnP9/p5wWHuPecZ+M5zn3M+PPc+94y2/raI+ZjrCJmRkfOxMl6UkQtfc9EZGTEf2+quHR+5jIz+Cdw6YMrdf3L3P4E3gLFamzFgT378NrDBzKzDGs/h7tPuPpkf/wYcBlb1Vc95Nga86skEcKmZrey7qGwD8KO7/9x3IU3c/TPgVG13dezuAe5t+NW7gAPufsrdfwUOAJsWqs6qpprdfb+7n81PJ4DVXdTyf7T09TCGud4siLlqztezB4HXu6hFwgiXjzDSGVlyPoIy8ryLmJER8xEuzIyMPoFbBRytPD/G7Av9f23ySXMauKKT6uaRl6vcBHzecPhWMztoZh+a2fXdVtbKgf1m9pWZPdFwfJjXoy9baT95S+xrgOXuPp0fnwCWN7Qpuc+3kd5xbjLfWOrDzrys5ZWWpTil9vXtwIy7H2k5XmJfy8ILnY8QLiMj5yMoI/sQKSOj5iOMaEZGn8CFZWZLgXeAp9z9TO3wJGkZww3Ac8D7HZfXZr273wxsBp40szv6LmgYZnYRsAV4q+FwqX19Dk+f84f5nx9mtgs4C+xtaVLaWHoBuBa4EZgmLbeI4iHmfmextL4WmVfAjAx7nikjuxcsIyPnI4xoRkafwB0Hrq48X533NbYxs0XAMuCXTqprYWaLScG0193frR939zPu/nt+vA9YbGZXdlzmLO5+PP88CbxH+si8apjXow+bgUl3n6kfKLWvs5nBEpv882RDm+L63MweA+4BHs6hOssQY6lT7j7j7n+7+z/Aiy31lNjXi4D7gTfb2pTW19KZkPmYawmXkYHzEZSRnYqWkVHzEUY7I6NP4L4ErjOza/I7SFuB8VqbcWBw16EHgE/bTpgu5LW4LwOH3f3ZljYrBt9DMLN1pNep70nnEjO7ZPCY9EXcb2vNxoFHLLkFOF1Z3tCn1ndfSuzriurYfRT4oKHNR8BGM7ssL2vYmPf1wsw2AU8DW9z9j5Y2w4ylTtW+i3IfzfUMc73p2p3A9+5+rOlgiX0tnQmXjxAzI4PnIygjOxMxIwPnI4xyRg57t5NSN9KdnX4g3f1mV973DOnkALiYtCxgCvgCWNNzvetJH/MfAr7O293ADmBHbrMT+I50F58J4LYC+nlNrudgrm3Q19W6DXg+vxbfAGsLqHsJKWyWVfYV19ek8JwG/iKtHd9O+i7KJ8AR4GPg8tx2LfBS5Xe35fE9BTzec81TpHXwg7E9uMPdVcC+ucZSz3W/lsfsIVLorKzXnZ/Put70VXPev3swlitti+lrbf1uTeOVgvMx1xQuI9vOMwrPx1yXMrLbmovOyJaai87Htrrz/t2MaEZa/gNERERERESkcNGXUIqIiIiIiFwwNIETEREREREJQhM4ERERERGRIDSBExERERERCUITOBERERERkSA0gRMREREREQlCEzgREREREZEg/gUbbu9eifDAvgAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 1080x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Max. Training Accuracy 0.9946523904800415\n",
            "Max. Validaiton Accuracy 0.5\n"
          ]
        }
      ],
      "source": [
        "plotModelHistory(model_5_history)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6OFOJQZFcjQW",
        "papermill": {
          "duration": 0.484716,
          "end_time": "2020-12-30T12:42:20.344397",
          "exception": false,
          "start_time": "2020-12-30T12:42:19.859681",
          "status": "completed"
        },
        "tags": []
      },
      "source": [
        "### Model 6: Switching Model architecture to Conv2D+LSTM"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "QDuEGEx3TTyp",
        "papermill": {
          "duration": 2133.535445,
          "end_time": "2020-12-30T13:17:54.330950",
          "exception": false,
          "start_time": "2020-12-30T12:42:20.795505",
          "status": "completed"
        },
        "tags": [],
        "outputId": "6e2b427e-b091-4a0d-e706-3eb3353552b4"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Model: \"conv_2d_lstm\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " time_distributed (TimeDistr  (None, 20, 100, 100, 16)  448      \n",
            " ibuted)                                                         \n",
            "                                                                 \n",
            " time_distributed_1 (TimeDis  (None, 20, 100, 100, 16)  64       \n",
            " tributed)                                                       \n",
            "                                                                 \n",
            " time_distributed_2 (TimeDis  (None, 20, 50, 50, 16)   0         \n",
            " tributed)                                                       \n",
            "                                                                 \n",
            " time_distributed_3 (TimeDis  (None, 20, 50, 50, 32)   4640      \n",
            " tributed)                                                       \n",
            "                                                                 \n",
            " time_distributed_4 (TimeDis  (None, 20, 50, 50, 32)   128       \n",
            " tributed)                                                       \n",
            "                                                                 \n",
            " time_distributed_5 (TimeDis  (None, 20, 25, 25, 32)   0         \n",
            " tributed)                                                       \n",
            "                                                                 \n",
            " time_distributed_6 (TimeDis  (None, 20, 25, 25, 64)   18496     \n",
            " tributed)                                                       \n",
            "                                                                 \n",
            " time_distributed_7 (TimeDis  (None, 20, 25, 25, 64)   256       \n",
            " tributed)                                                       \n",
            "                                                                 \n",
            " time_distributed_8 (TimeDis  (None, 20, 12, 12, 64)   0         \n",
            " tributed)                                                       \n",
            "                                                                 \n",
            " time_distributed_9 (TimeDis  (None, 20, 12, 12, 128)  73856     \n",
            " tributed)                                                       \n",
            "                                                                 \n",
            " time_distributed_10 (TimeDi  (None, 20, 12, 12, 128)  512       \n",
            " stributed)                                                      \n",
            "                                                                 \n",
            " time_distributed_11 (TimeDi  (None, 20, 6, 6, 128)    0         \n",
            " stributed)                                                      \n",
            "                                                                 \n",
            " time_distributed_12 (TimeDi  (None, 20, 6, 6, 256)    295168    \n",
            " stributed)                                                      \n",
            "                                                                 \n",
            " time_distributed_13 (TimeDi  (None, 20, 6, 6, 256)    1024      \n",
            " stributed)                                                      \n",
            "                                                                 \n",
            " time_distributed_14 (TimeDi  (None, 20, 3, 3, 256)    0         \n",
            " stributed)                                                      \n",
            "                                                                 \n",
            " time_distributed_15 (TimeDi  (None, 20, 2304)         0         \n",
            " stributed)                                                      \n",
            "                                                                 \n",
            " lstm (LSTM)                 (None, 256)               2622464   \n",
            "                                                                 \n",
            " dropout_12 (Dropout)        (None, 256)               0         \n",
            "                                                                 \n",
            " dense_21 (Dense)            (None, 256)               65792     \n",
            "                                                                 \n",
            " dropout_13 (Dropout)        (None, 256)               0         \n",
            "                                                                 \n",
            " dense_22 (Dense)            (None, 5)                 1285      \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 3,084,133\n",
            "Trainable params: 3,083,141\n",
            "Non-trainable params: 992\n",
            "_________________________________________________________________\n",
            "None\n",
            "Source path =  datasets/Project_data/train ; batch size = 32\n",
            "Epoch 1/30\n",
            "21/21 [==============================] - ETA: 0s - loss: 1.2714 - categorical_accuracy: 0.4736Source path =  datasets/Project_data/val ; batch size = 32\n",
            "21/21 [==============================] - 121s 6s/step - loss: 1.2714 - categorical_accuracy: 0.4736 - val_loss: 1.5895 - val_categorical_accuracy: 0.2600 - lr: 0.0010\n",
            "Epoch 2/30\n",
            "21/21 [==============================] - 77s 4s/step - loss: 0.9527 - categorical_accuracy: 0.6480 - val_loss: 1.7266 - val_categorical_accuracy: 0.3125 - lr: 0.0010\n",
            "Epoch 3/30\n",
            "21/21 [==============================] - 69s 3s/step - loss: 0.8843 - categorical_accuracy: 0.6651 - val_loss: 2.3217 - val_categorical_accuracy: 0.1875 - lr: 0.0010\n",
            "Epoch 4/30\n",
            "21/21 [==============================] - 66s 3s/step - loss: 0.8207 - categorical_accuracy: 0.6792 - val_loss: 3.5548 - val_categorical_accuracy: 0.1250 - lr: 0.0010\n",
            "Epoch 5/30\n",
            "21/21 [==============================] - ETA: 0s - loss: 0.6904 - categorical_accuracy: 0.7535\n",
            "Epoch 00005: ReduceLROnPlateau reducing learning rate to 0.00020000000949949026.\n",
            "21/21 [==============================] - 57s 3s/step - loss: 0.6904 - categorical_accuracy: 0.7535 - val_loss: 4.0435 - val_categorical_accuracy: 0.1875 - lr: 0.0010\n",
            "Epoch 6/30\n",
            "21/21 [==============================] - 58s 3s/step - loss: 0.6602 - categorical_accuracy: 0.7311 - val_loss: 3.9672 - val_categorical_accuracy: 0.1875 - lr: 2.0000e-04\n",
            "Epoch 7/30\n",
            "21/21 [==============================] - 56s 3s/step - loss: 0.4113 - categorical_accuracy: 0.8599 - val_loss: 4.1711 - val_categorical_accuracy: 0.1250 - lr: 2.0000e-04\n",
            "Epoch 8/30\n",
            "21/21 [==============================] - 58s 3s/step - loss: 0.4165 - categorical_accuracy: 0.8459 - val_loss: 3.5683 - val_categorical_accuracy: 0.2500 - lr: 2.0000e-04\n",
            "Epoch 9/30\n",
            "21/21 [==============================] - ETA: 0s - loss: 0.3002 - categorical_accuracy: 0.9020\n",
            "Epoch 00009: ReduceLROnPlateau reducing learning rate to 4.0000001899898055e-05.\n",
            "21/21 [==============================] - 59s 3s/step - loss: 0.3002 - categorical_accuracy: 0.9020 - val_loss: 2.8437 - val_categorical_accuracy: 0.3125 - lr: 2.0000e-04\n",
            "Epoch 10/30\n",
            "21/21 [==============================] - 59s 3s/step - loss: 0.2927 - categorical_accuracy: 0.9048 - val_loss: 2.7872 - val_categorical_accuracy: 0.0625 - lr: 4.0000e-05\n",
            "Epoch 11/30\n",
            "21/21 [==============================] - 58s 3s/step - loss: 0.2140 - categorical_accuracy: 0.9384 - val_loss: 3.3995 - val_categorical_accuracy: 0.0625 - lr: 4.0000e-05\n",
            "Epoch 12/30\n",
            "21/21 [==============================] - 56s 3s/step - loss: 0.2373 - categorical_accuracy: 0.9272 - val_loss: 2.1246 - val_categorical_accuracy: 0.3750 - lr: 4.0000e-05\n",
            "Epoch 13/30\n",
            "21/21 [==============================] - ETA: 0s - loss: 0.2212 - categorical_accuracy: 0.9412\n",
            "Epoch 00013: ReduceLROnPlateau reducing learning rate to 8.000000525498762e-06.\n",
            "21/21 [==============================] - 58s 3s/step - loss: 0.2212 - categorical_accuracy: 0.9412 - val_loss: 3.1133 - val_categorical_accuracy: 0.2500 - lr: 4.0000e-05\n",
            "Epoch 14/30\n",
            "21/21 [==============================] - 57s 3s/step - loss: 0.2482 - categorical_accuracy: 0.9272 - val_loss: 3.5280 - val_categorical_accuracy: 0.1250 - lr: 8.0000e-06\n",
            "Epoch 15/30\n",
            "21/21 [==============================] - 59s 3s/step - loss: 0.1992 - categorical_accuracy: 0.9412 - val_loss: 2.1892 - val_categorical_accuracy: 0.4375 - lr: 8.0000e-06\n",
            "Epoch 16/30\n",
            "21/21 [==============================] - 60s 3s/step - loss: 0.2179 - categorical_accuracy: 0.9440 - val_loss: 3.2736 - val_categorical_accuracy: 0.2500 - lr: 8.0000e-06\n",
            "Epoch 17/30\n",
            "21/21 [==============================] - ETA: 0s - loss: 0.1950 - categorical_accuracy: 0.9496\n",
            "Epoch 00017: ReduceLROnPlateau reducing learning rate to 1.6000001778593287e-06.\n",
            "21/21 [==============================] - 57s 3s/step - loss: 0.1950 - categorical_accuracy: 0.9496 - val_loss: 2.4737 - val_categorical_accuracy: 0.2500 - lr: 8.0000e-06\n",
            "Epoch 18/30\n",
            "21/21 [==============================] - 56s 3s/step - loss: 0.2022 - categorical_accuracy: 0.9300 - val_loss: 1.7236 - val_categorical_accuracy: 0.6250 - lr: 1.6000e-06\n",
            "Epoch 19/30\n",
            "21/21 [==============================] - 59s 3s/step - loss: 0.2108 - categorical_accuracy: 0.9384 - val_loss: 2.6983 - val_categorical_accuracy: 0.3750 - lr: 1.6000e-06\n",
            "Epoch 20/30\n",
            "21/21 [==============================] - 56s 3s/step - loss: 0.2115 - categorical_accuracy: 0.9356 - val_loss: 2.5831 - val_categorical_accuracy: 0.4375 - lr: 1.6000e-06\n",
            "Epoch 21/30\n",
            "21/21 [==============================] - ETA: 0s - loss: 0.2063 - categorical_accuracy: 0.9328\n",
            "Epoch 00021: ReduceLROnPlateau reducing learning rate to 3.200000264769187e-07.\n",
            "21/21 [==============================] - 56s 3s/step - loss: 0.2063 - categorical_accuracy: 0.9328 - val_loss: 2.1767 - val_categorical_accuracy: 0.3750 - lr: 1.6000e-06\n",
            "Epoch 22/30\n",
            "21/21 [==============================] - 60s 3s/step - loss: 0.1654 - categorical_accuracy: 0.9664 - val_loss: 1.2446 - val_categorical_accuracy: 0.4375 - lr: 3.2000e-07\n",
            "Epoch 23/30\n",
            "21/21 [==============================] - 58s 3s/step - loss: 0.2429 - categorical_accuracy: 0.9356 - val_loss: 1.4203 - val_categorical_accuracy: 0.5625 - lr: 3.2000e-07\n",
            "Epoch 24/30\n",
            "21/21 [==============================] - 58s 3s/step - loss: 0.2210 - categorical_accuracy: 0.9300 - val_loss: 1.6344 - val_categorical_accuracy: 0.5000 - lr: 3.2000e-07\n",
            "Epoch 25/30\n",
            "21/21 [==============================] - 56s 3s/step - loss: 0.2122 - categorical_accuracy: 0.9412 - val_loss: 1.6576 - val_categorical_accuracy: 0.4375 - lr: 3.2000e-07\n",
            "Epoch 26/30\n",
            "21/21 [==============================] - ETA: 0s - loss: 0.2227 - categorical_accuracy: 0.9300\n",
            "Epoch 00026: ReduceLROnPlateau reducing learning rate to 6.400000529538374e-08.\n",
            "21/21 [==============================] - 58s 3s/step - loss: 0.2227 - categorical_accuracy: 0.9300 - val_loss: 1.2502 - val_categorical_accuracy: 0.5000 - lr: 3.2000e-07\n",
            "Epoch 27/30\n",
            "21/21 [==============================] - 58s 3s/step - loss: 0.2074 - categorical_accuracy: 0.9440 - val_loss: 1.1775 - val_categorical_accuracy: 0.6875 - lr: 6.4000e-08\n",
            "Epoch 28/30\n",
            "21/21 [==============================] - 56s 3s/step - loss: 0.1825 - categorical_accuracy: 0.9636 - val_loss: 1.4374 - val_categorical_accuracy: 0.5625 - lr: 6.4000e-08\n",
            "Epoch 29/30\n",
            "21/21 [==============================] - 58s 3s/step - loss: 0.2025 - categorical_accuracy: 0.9440 - val_loss: 0.3442 - val_categorical_accuracy: 0.8125 - lr: 6.4000e-08\n",
            "Epoch 30/30\n",
            "21/21 [==============================] - 59s 3s/step - loss: 0.1915 - categorical_accuracy: 0.9384 - val_loss: 0.6619 - val_categorical_accuracy: 0.6875 - lr: 6.4000e-08\n"
          ]
        }
      ],
      "source": [
        "no_of_frames = 20\n",
        "no_of_epochs = 30\n",
        "size_of_a_batch = 32\n",
        "\n",
        "Image_tensor = GetImageTensor(no_of_frames)\n",
        "training_generator = generator(train_path, train_doc, size_of_a_batch, Image_tensor)\n",
        "val_generator = generator(val_path, val_doc, size_of_a_batch, Image_tensor)\n",
        "\n",
        "if (num_of_train_seq%size_of_a_batch) == 0:\n",
        "    steps_per_epoch = int(num_of_train_seq/size_of_a_batch)\n",
        "else:\n",
        "    steps_per_epoch = (num_of_train_seq//size_of_a_batch) + 1\n",
        "\n",
        "if (num_of_value_seq%size_of_a_batch) == 0:\n",
        "    validation_steps = int(num_of_value_seq/size_of_a_batch)\n",
        "else:\n",
        "    validation_steps = (num_of_value_seq//size_of_a_batch) + 1\n",
        "    \n",
        "inputShape = (len(Image_tensor[0]), Image_tensor[1], Image_tensor[2], Image_tensor[3])\n",
        "\n",
        "model_6 = Sequential([\n",
        "    TimeDistributed(Conv2D(16, make2dFilter(3), padding='same', activation='relu'), input_shape=inputShape),\n",
        "    TimeDistributed(BatchNormalization()),\n",
        "    TimeDistributed(MaxPooling2D(make2dFilter(2))),\n",
        "\n",
        "    TimeDistributed(Conv2D(32, make2dFilter(3), padding='same', activation='relu')),\n",
        "    TimeDistributed(BatchNormalization()),\n",
        "    TimeDistributed(MaxPooling2D(make2dFilter(2))),\n",
        "\n",
        "    TimeDistributed(Conv2D(64, make2dFilter(3), padding='same', activation='relu')),\n",
        "    TimeDistributed(BatchNormalization()),\n",
        "    TimeDistributed(MaxPooling2D(make2dFilter(2))),\n",
        "\n",
        "    TimeDistributed(Conv2D(128, make2dFilter(3), padding='same', activation='relu')),\n",
        "    TimeDistributed(BatchNormalization()),\n",
        "    TimeDistributed(MaxPooling2D(make2dFilter(2))),\n",
        "\n",
        "    TimeDistributed(Conv2D(256, make2dFilter(3), padding='same', activation='relu')),\n",
        "    TimeDistributed(BatchNormalization()),\n",
        "    TimeDistributed(MaxPooling2D(make2dFilter(2))),\n",
        "\n",
        "    TimeDistributed(Flatten()),\n",
        "    LSTM(256),\n",
        "    Dropout(0.2),\n",
        "\n",
        "    Dense(256, activation='relu'),\n",
        "    Dropout(0.2),\n",
        "\n",
        "    Dense(5, activation='softmax')\n",
        "], name=\"conv_2d_lstm\")\n",
        "model_6.compile(optimizer=tf.keras.optimizers.Adam(), loss='categorical_crossentropy', metrics=['categorical_accuracy'])\n",
        "print(model_6.summary())\n",
        "\n",
        "model_6_history = model_6.fit_generator(training_generator, steps_per_epoch=steps_per_epoch, epochs=no_of_epochs, verbose=1, \n",
        "            callbacks=callbacks_list, validation_data=val_generator, \n",
        "            validation_steps=validation_steps, class_weight=None, workers=1, initial_epoch=0)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "hTaR1Cv0H-I8",
        "papermill": {
          "duration": 0.926799,
          "end_time": "2020-12-30T13:17:55.889071",
          "exception": false,
          "start_time": "2020-12-30T13:17:54.962272",
          "status": "completed"
        },
        "tags": [],
        "outputId": "c3809d46-ceaf-4a0b-b154-7eb953df45d7"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Max. Training Accuracy 0.9663865566253662\n",
            "Max. Validaiton Accuracy 0.8125\n"
          ]
        }
      ],
      "source": [
        "plotModelHistory(model_6_history)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "skhLjh_vyXfR",
        "papermill": {
          "duration": 0.635548,
          "end_time": "2020-12-30T13:17:57.159194",
          "exception": false,
          "start_time": "2020-12-30T13:17:56.523646",
          "status": "completed"
        },
        "tags": []
      },
      "source": [
        "### Model 7: Transfer Learning (MobileNet) with LSTM"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "6o_v7A10vBKd",
        "papermill": {
          "duration": 0.635537,
          "end_time": "2020-12-30T13:17:58.435917",
          "exception": false,
          "start_time": "2020-12-30T13:17:57.800380",
          "status": "completed"
        },
        "tags": []
      },
      "outputs": [],
      "source": [
        "from tensorflow.keras.applications import mobilenet"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "1o8quf9byzns",
        "papermill": {
          "duration": 1255.865789,
          "end_time": "2020-12-30T13:38:54.928343",
          "exception": false,
          "start_time": "2020-12-30T13:17:59.062554",
          "status": "completed"
        },
        "tags": [],
        "outputId": "4972f229-23f9-4ef4-9bb1-40c425665311"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:`input_shape` is undefined or non-square, or `rows` is not in [128, 160, 192, 224]. Weights for input shape (224, 224) will be loaded as the default.\n",
            "Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/mobilenet/mobilenet_1_0_224_tf_no_top.h5\n",
            "17227776/17225924 [==============================] - 2s 0us/step\n",
            "17235968/17225924 [==============================] - 2s 0us/step\n",
            "Model: \"mobilenet_lstm\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " time_distributed_16 (TimeDi  (None, 20, 3, 3, 1024)   3228864   \n",
            " stributed)                                                      \n",
            "                                                                 \n",
            " time_distributed_17 (TimeDi  (None, 20, 3, 3, 1024)   4096      \n",
            " stributed)                                                      \n",
            "                                                                 \n",
            " time_distributed_18 (TimeDi  (None, 20, 1, 1, 1024)   0         \n",
            " stributed)                                                      \n",
            "                                                                 \n",
            " time_distributed_19 (TimeDi  (None, 20, 1024)         0         \n",
            " stributed)                                                      \n",
            "                                                                 \n",
            " lstm_1 (LSTM)               (None, 256)               1311744   \n",
            "                                                                 \n",
            " dropout_14 (Dropout)        (None, 256)               0         \n",
            "                                                                 \n",
            " dense_23 (Dense)            (None, 256)               65792     \n",
            "                                                                 \n",
            " dropout_15 (Dropout)        (None, 256)               0         \n",
            "                                                                 \n",
            " dense_24 (Dense)            (None, 5)                 1285      \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 4,611,781\n",
            "Trainable params: 1,380,869\n",
            "Non-trainable params: 3,230,912\n",
            "_________________________________________________________________\n",
            "None\n",
            "Source path =  datasets/Project_data/train ; batch size = 64\n",
            "Epoch 1/30\n",
            "11/11 [==============================] - ETA: 0s - loss: 1.4139 - categorical_accuracy: 0.3952 Source path =  datasets/Project_data/val ; batch size = 64\n",
            "11/11 [==============================] - 125s 12s/step - loss: 1.4139 - categorical_accuracy: 0.3952 - val_loss: 1.2384 - val_categorical_accuracy: 0.4700 - lr: 0.0010\n",
            "Epoch 2/30\n",
            "11/11 [==============================] - 52s 5s/step - loss: 0.6860 - categorical_accuracy: 0.7787 - val_loss: 0.9332 - val_categorical_accuracy: 0.5938 - lr: 0.0010\n",
            "Epoch 3/30\n",
            "11/11 [==============================] - 50s 5s/step - loss: 0.5699 - categorical_accuracy: 0.7826 - val_loss: 0.8814 - val_categorical_accuracy: 0.5714 - lr: 0.0010\n",
            "Epoch 4/30\n",
            "11/11 [==============================] - 43s 4s/step - loss: 0.4523 - categorical_accuracy: 0.8369 - val_loss: 0.6066 - val_categorical_accuracy: 0.8125 - lr: 0.0010\n",
            "Epoch 5/30\n",
            "11/11 [==============================] - 39s 4s/step - loss: 0.1593 - categorical_accuracy: 0.9617 - val_loss: 0.9380 - val_categorical_accuracy: 0.5625 - lr: 0.0010\n",
            "Epoch 6/30\n",
            "11/11 [==============================] - 31s 3s/step - loss: 0.1917 - categorical_accuracy: 0.9474 - val_loss: 0.4318 - val_categorical_accuracy: 0.8750 - lr: 0.0010\n",
            "Epoch 7/30\n",
            "11/11 [==============================] - 34s 3s/step - loss: 0.2196 - categorical_accuracy: 0.9360 - val_loss: 1.1285 - val_categorical_accuracy: 0.2500 - lr: 0.0010\n",
            "Epoch 8/30\n",
            "11/11 [==============================] - 30s 3s/step - loss: 0.0739 - categorical_accuracy: 0.9893 - val_loss: 0.6980 - val_categorical_accuracy: 0.8750 - lr: 0.0010\n",
            "Epoch 9/30\n",
            "11/11 [==============================] - 29s 3s/step - loss: 0.0956 - categorical_accuracy: 0.9679 - val_loss: 0.9376 - val_categorical_accuracy: 0.7500 - lr: 0.0010\n",
            "Epoch 10/30\n",
            "11/11 [==============================] - 31s 3s/step - loss: 0.1003 - categorical_accuracy: 0.9679 - val_loss: 0.3216 - val_categorical_accuracy: 0.8750 - lr: 0.0010\n",
            "Epoch 11/30\n",
            "11/11 [==============================] - 30s 3s/step - loss: 0.0508 - categorical_accuracy: 0.9947 - val_loss: 1.4255 - val_categorical_accuracy: 0.6250 - lr: 0.0010\n",
            "Epoch 12/30\n",
            "11/11 [==============================] - 31s 3s/step - loss: 0.0456 - categorical_accuracy: 0.9786 - val_loss: 1.9937 - val_categorical_accuracy: 0.5000 - lr: 0.0010\n",
            "Epoch 13/30\n",
            "11/11 [==============================] - 29s 3s/step - loss: 0.0814 - categorical_accuracy: 0.9679 - val_loss: 0.0272 - val_categorical_accuracy: 1.0000 - lr: 0.0010\n",
            "Epoch 14/30\n",
            "11/11 [==============================] - 32s 3s/step - loss: 0.0498 - categorical_accuracy: 0.9840 - val_loss: 0.9396 - val_categorical_accuracy: 0.8750 - lr: 0.0010\n",
            "Epoch 15/30\n",
            "11/11 [==============================] - 30s 3s/step - loss: 0.0207 - categorical_accuracy: 1.0000 - val_loss: 0.3097 - val_categorical_accuracy: 0.8750 - lr: 0.0010\n",
            "Epoch 16/30\n",
            "11/11 [==============================] - 31s 3s/step - loss: 0.0440 - categorical_accuracy: 0.9893 - val_loss: 0.6139 - val_categorical_accuracy: 0.7500 - lr: 0.0010\n",
            "Epoch 17/30\n",
            "11/11 [==============================] - ETA: 0s - loss: 0.0218 - categorical_accuracy: 1.0000\n",
            "Epoch 00017: ReduceLROnPlateau reducing learning rate to 0.00020000000949949026.\n",
            "11/11 [==============================] - 28s 3s/step - loss: 0.0218 - categorical_accuracy: 1.0000 - val_loss: 0.6801 - val_categorical_accuracy: 0.7500 - lr: 0.0010\n",
            "Epoch 18/30\n",
            "11/11 [==============================] - 31s 3s/step - loss: 0.0387 - categorical_accuracy: 0.9893 - val_loss: 1.3158 - val_categorical_accuracy: 0.6250 - lr: 2.0000e-04\n",
            "Epoch 19/30\n",
            "11/11 [==============================] - 30s 3s/step - loss: 0.0080 - categorical_accuracy: 1.0000 - val_loss: 1.1689 - val_categorical_accuracy: 0.7500 - lr: 2.0000e-04\n",
            "Epoch 20/30\n",
            "11/11 [==============================] - 30s 3s/step - loss: 0.0177 - categorical_accuracy: 0.9947 - val_loss: 0.9555 - val_categorical_accuracy: 0.7500 - lr: 2.0000e-04\n",
            "Epoch 21/30\n",
            "11/11 [==============================] - ETA: 0s - loss: 0.0156 - categorical_accuracy: 0.9947\n",
            "Epoch 00021: ReduceLROnPlateau reducing learning rate to 4.0000001899898055e-05.\n",
            "11/11 [==============================] - 30s 3s/step - loss: 0.0156 - categorical_accuracy: 0.9947 - val_loss: 0.3854 - val_categorical_accuracy: 0.8750 - lr: 2.0000e-04\n",
            "Epoch 22/30\n",
            "11/11 [==============================] - 28s 3s/step - loss: 0.0252 - categorical_accuracy: 0.9947 - val_loss: 0.4879 - val_categorical_accuracy: 0.8750 - lr: 4.0000e-05\n",
            "Epoch 23/30\n",
            "11/11 [==============================] - 31s 3s/step - loss: 0.0112 - categorical_accuracy: 1.0000 - val_loss: 0.5882 - val_categorical_accuracy: 0.8750 - lr: 4.0000e-05\n",
            "Epoch 24/30\n",
            "11/11 [==============================] - 31s 3s/step - loss: 0.0100 - categorical_accuracy: 1.0000 - val_loss: 0.6989 - val_categorical_accuracy: 0.7500 - lr: 4.0000e-05\n",
            "Epoch 25/30\n",
            "11/11 [==============================] - ETA: 0s - loss: 0.0066 - categorical_accuracy: 1.0000\n",
            "Epoch 00025: ReduceLROnPlateau reducing learning rate to 8.000000525498762e-06.\n",
            "11/11 [==============================] - 29s 3s/step - loss: 0.0066 - categorical_accuracy: 1.0000 - val_loss: 0.9829 - val_categorical_accuracy: 0.6250 - lr: 4.0000e-05\n",
            "Epoch 26/30\n",
            "11/11 [==============================] - 32s 3s/step - loss: 0.0107 - categorical_accuracy: 0.9947 - val_loss: 1.5296 - val_categorical_accuracy: 0.6250 - lr: 8.0000e-06\n",
            "Epoch 27/30\n",
            "11/11 [==============================] - 28s 3s/step - loss: 0.0085 - categorical_accuracy: 1.0000 - val_loss: 0.3340 - val_categorical_accuracy: 0.8750 - lr: 8.0000e-06\n",
            "Epoch 28/30\n",
            "11/11 [==============================] - 30s 3s/step - loss: 0.0082 - categorical_accuracy: 1.0000 - val_loss: 1.0224 - val_categorical_accuracy: 0.8750 - lr: 8.0000e-06\n",
            "Epoch 29/30\n",
            "11/11 [==============================] - ETA: 0s - loss: 0.0056 - categorical_accuracy: 1.0000\n",
            "Epoch 00029: ReduceLROnPlateau reducing learning rate to 1.6000001778593287e-06.\n",
            "11/11 [==============================] - 29s 3s/step - loss: 0.0056 - categorical_accuracy: 1.0000 - val_loss: 0.3825 - val_categorical_accuracy: 0.7500 - lr: 8.0000e-06\n",
            "Epoch 30/30\n",
            "11/11 [==============================] - 31s 3s/step - loss: 0.0085 - categorical_accuracy: 1.0000 - val_loss: 0.5709 - val_categorical_accuracy: 0.7500 - lr: 1.6000e-06\n"
          ]
        }
      ],
      "source": [
        "mobilenet = mobilenet.MobileNet(weights='imagenet', include_top=False)\n",
        "\n",
        "no_of_frames = 20\n",
        "no_of_epochs = 30\n",
        "size_of_a_batch = 64\n",
        "\n",
        "Image_tensor = GetImageTensor(no_of_frames)\n",
        "training_generator = generator(train_path, train_doc, size_of_a_batch, Image_tensor)\n",
        "val_generator = generator(val_path, val_doc, size_of_a_batch, Image_tensor)\n",
        "\n",
        "if (num_of_train_seq%size_of_a_batch) == 0:\n",
        "    steps_per_epoch = int(num_of_train_seq/size_of_a_batch)\n",
        "else:\n",
        "    steps_per_epoch = (num_of_train_seq//size_of_a_batch) + 1\n",
        "\n",
        "if (num_of_value_seq%size_of_a_batch) == 0:\n",
        "    validation_steps = int(num_of_value_seq/size_of_a_batch)\n",
        "else:\n",
        "    validation_steps = (num_of_value_seq//size_of_a_batch) + 1\n",
        "    \n",
        "inputShape = (len(Image_tensor[0]), Image_tensor[1], Image_tensor[2], Image_tensor[3])\n",
        "\n",
        "model_7 = Sequential([\n",
        "    TimeDistributed(mobilenet, input_shape=inputShape)\n",
        "], name=\"mobilenet_lstm\")\n",
        "\n",
        "for layer in model_7.layers:\n",
        "    layer.trainable = False\n",
        "\n",
        "model_7.add(TimeDistributed(BatchNormalization()))\n",
        "model_7.add(TimeDistributed(MaxPooling2D(make2dFilter(2))))\n",
        "model_7.add(TimeDistributed(Flatten()))\n",
        "\n",
        "model_7.add(LSTM(256))\n",
        "model_7.add(Dropout(0.2))\n",
        "\n",
        "model_7.add(Dense(256,activation='relu'))\n",
        "model_7.add(Dropout(0.2))\n",
        "\n",
        "model_7.add(Dense(5, activation='softmax'))\n",
        "\n",
        "model_7.compile(optimizer=tf.keras.optimizers.Adam(), loss='categorical_crossentropy', metrics=['categorical_accuracy'])\n",
        "print(model_7.summary())\n",
        "\n",
        "model_7_history = model_7.fit_generator(training_generator, steps_per_epoch=steps_per_epoch, epochs=no_of_epochs, verbose=1, \n",
        "            callbacks=callbacks_list, validation_data=val_generator, \n",
        "            validation_steps=validation_steps, class_weight=None, workers=1, initial_epoch=0)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "cfJz7NMiH-I8",
        "papermill": {
          "duration": 1.153986,
          "end_time": "2020-12-30T13:38:56.988440",
          "exception": false,
          "start_time": "2020-12-30T13:38:55.834454",
          "status": "completed"
        },
        "tags": [],
        "outputId": "2e12ec5a-1033-42b4-976d-5eab170ea4ae"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Max. Training Accuracy 1.0\n",
            "Max. Validaiton Accuracy 1.0\n"
          ]
        }
      ],
      "source": [
        "plotModelHistory(model_7_history)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-kJPNQeINPtG"
      },
      "source": [
        "# ----------------------------------------------------Thank you---------------------------------------\n"
      ]
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "gpuType": "T4",
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3 (ipykernel)",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.8.10"
    },
    "papermill": {
      "duration": 9375.075711,
      "end_time": "2020-12-30T13:38:59.003426",
      "environment_variables": {},
      "exception": null,
      "input_path": "__notebook__.ipynb",
      "output_path": "__notebook__.ipynb",
      "parameters": {},
      "start_time": "2020-12-30T11:02:43.927715",
      "version": "2.1.0"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}