univariate_time_series_multi_step_enhanced.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.8.3"
    },
    "nav_menu": {},
    "toc": {
      "navigate_menu": true,
      "number_sections": true,
      "sideBar": true,
      "threshold": 6,
      "toc_cell": false,
      "toc_section_display": "block",
      "toc_window_display": false
    },
    "colab": {
      "name": "univariate-time-series-multi-step-enhanced.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true,
      "include_colab_link": true
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/mehrotrasan16/Keras-Deep-Learning/blob/05-RNN-LSTM-Seq2Seq-Networks/univariate_time_series_multi_step_enhanced.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zYLjiPLMZPT0"
      },
      "source": [
        "import sys\n",
        "import sklearn\n",
        "import tensorflow as tf\n",
        "from tensorflow import keras\n",
        "\n",
        "import numpy as np\n",
        "from numpy import load\n",
        "import os\n",
        "\n",
        "# to make this notebook's output stable across runs\n",
        "np.random.seed(42)\n",
        "tf.random.set_seed(42)\n",
        "\n",
        "# To plot pretty figures\n",
        "%matplotlib inline\n",
        "import matplotlib as mpl\n",
        "import matplotlib.pyplot as plt\n",
        "mpl.rc('axes', labelsize=14)\n",
        "mpl.rc('xtick', labelsize=12)\n",
        "mpl.rc('ytick', labelsize=12)\n",
        "\n",
        "# Where to save the figures\n",
        "PROJECT_ROOT_DIR = \".\"\n",
        "IMAGES_PATH = os.path.join(PROJECT_ROOT_DIR, \"images\")\n",
        "os.makedirs(IMAGES_PATH, exist_ok=True)\n",
        "\n",
        "def save_fig(fig_id, tight_layout=True, fig_extension=\"png\", resolution=300):\n",
        "    path = os.path.join(IMAGES_PATH, fig_id + \".\" + fig_extension)\n",
        "    print(\"Saving figure\", fig_id)\n",
        "    if tight_layout:\n",
        "        plt.tight_layout()\n",
        "    plt.savefig(path, format=fig_extension, dpi=resolution)"
      ],
      "execution_count": 34,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KrMUJUGlbwPx",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "492d1c05-f7e5-4dd5-b434-9ccbce5646e7"
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": 35,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Drive already mounted at /content/drive; to attempt to forcibly remount, call drive.mount(\"/content/drive\", force_remount=True).\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cbNhwKz8ZPT-"
      },
      "source": [
        "### Load Dataset"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ry9ZNFm1ZPT_"
      },
      "source": [
        "def plot_learning_curves(loss, val_loss):\n",
        "    plt.plot(np.arange(len(loss)) + 0.5, loss, \"b.-\", label=\"Training loss\")\n",
        "    plt.plot(np.arange(len(val_loss)) + 1, val_loss, \"r.-\", label=\"Validation loss\")\n",
        "    plt.gca().xaxis.set_major_locator(mpl.ticker.MaxNLocator(integer=True))\n",
        "    plt.axis([1, 20, 0, 0.2])\n",
        "    plt.legend(fontsize=14)\n",
        "    plt.xlabel(\"Epochs\")\n",
        "    plt.ylabel(\"Loss\")\n",
        "    plt.grid(True)"
      ],
      "execution_count": 36,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xuZtjCm-ZPUD"
      },
      "source": [
        "def plot_series(series, y=None, y_pred=None, x_label=\"$t$\", y_label=\"$x(t)$\"):\n",
        "    plt.plot(series, \".-\")\n",
        "    if y is not None:\n",
        "        plt.plot(n_steps, y, \"bx\", markersize=10)\n",
        "    if y_pred is not None:\n",
        "        plt.plot(n_steps, y_pred, \"ro\")\n",
        "    plt.grid(True)\n",
        "    if x_label:\n",
        "        plt.xlabel(x_label, fontsize=16)\n",
        "    if y_label:\n",
        "        plt.ylabel(y_label, fontsize=16, rotation=0)\n",
        "    plt.hlines(0, 0, 100, linewidth=1)\n",
        "    plt.axis([0, n_steps + 1, -1, 1])"
      ],
      "execution_count": 37,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "J7ccHOnvZPUG"
      },
      "source": [
        "def plot_multiple_forecasts(X, Y, Y_pred):\n",
        "    n_steps = X.shape[1]\n",
        "    ahead = Y.shape[1]\n",
        "    plot_series(X[0, :, 0])\n",
        "    plt.plot(np.arange(n_steps, n_steps + ahead), Y[0, :, 0], \"ro-\", label=\"Actual\")\n",
        "    plt.plot(np.arange(n_steps, n_steps + ahead), Y_pred[0, :, 0], \"bx-\", label=\"Forecast\", markersize=10)\n",
        "    plt.axis([0, n_steps + ahead, -1, 1])\n",
        "    plt.legend(fontsize=14)"
      ],
      "execution_count": 38,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_mr0vXF6ZPUJ"
      },
      "source": [
        "n_steps = 50\n",
        "series = load('/content/drive/MyDrive/Colab Notebooks/CS581-TestResults/assignment5/HW5-supp-files/univariate-time-series-multi-step.npy')\n",
        "\n",
        "X_train, Y_train = series[:7000, :n_steps], series[:7000, -10:, 0]\n",
        "X_valid, Y_valid = series[7000:9000, :n_steps], series[7000:9000, -10:, 0]\n",
        "X_test, Y_test = series[9000:, :n_steps], series[9000:, -10:, 0]"
      ],
      "execution_count": 39,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "sX4v4vtgZPUN"
      },
      "source": [
        "## Improving multi-step prediction model\n",
        "\n",
        "Instead of training the model to forecast the next 10 values only at the very last time step, we can train it to forecast the next 10 values at each and every time step. In other words, we can turn this sequence-to-vector RNN into a sequence-to-sequence RNN. The advantage of this technique is that the loss will contain a term for the output of the RNN at each and every time step, not just the output at the last time step. This means there will be many more error gradients flowing through the model, and they will also flow from the output of each time step. This will both stabilize and speed up training. To be clear, at time step 0 the model will output a vector containing the forecasts for time steps 1 to 10, then at time step 1 the model will forecast time steps 2 to 11, and so on. So each target must be a sequence of the same length as the input sequence, containing a 10-dimensional vector at each step. Let’s prepare these target sequences."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "M4KWn25DZPUN"
      },
      "source": [
        "Y = np.empty((10000, n_steps, 10)) "
      ],
      "execution_count": 40,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "a4AHBv67ZPUQ"
      },
      "source": [
        "# each target is a sequence of 10D vectors, because in each of the 50 time steps the model will predict 10 time steps\n",
        "for step_ahead in range(1, 10 + 1):\n",
        "    Y[..., step_ahead - 1] = series[..., step_ahead:step_ahead + n_steps, 0]\n",
        "    # same as     Y[:, :, step_ahead - 1] = series[:, step_ahead:step_ahead + n_steps, 0]\n",
        "    # broadcasting 1-51 to 0th element of z axis of each of the 10k rows\n",
        "    # then 2-52 to 1st element of z-axis ... all the way till 10-60 to 9th element of z axis of the 10k rows \n",
        "Y_train = Y[:7000]\n",
        "Y_valid = Y[7000:9000]\n",
        "Y_test = Y[9000:]"
      ],
      "execution_count": 41,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "k6GHtozRZPUV",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "0c6b3d0a-9f92-489f-85ea-f3bbc445621a"
      },
      "source": [
        "X_train.shape, Y_train.shape"
      ],
      "execution_count": 42,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((7000, 50, 1), (7000, 50, 10))"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 42
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JgdkeKYJhmHD",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "ac0a1b1d-e21e-4318-a2cc-e8cdda0e546a"
      },
      "source": [
        "X_valid.shape,Y_valid.shape"
      ],
      "execution_count": 43,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((2000, 50, 1), (2000, 50, 10))"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 43
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RB3eWHESc3cv",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "b1f97788-77bf-4ec6-ac25-9f9248e00644"
      },
      "source": [
        "Y_train[0].shape"
      ],
      "execution_count": 44,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(50, 10)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 44
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZNgFktMNZPUa"
      },
      "source": [
        "model = keras.models.Sequential([\n",
        "    keras.layers.LSTM(256,return_sequences=True,input_shape=[50, 1]),\n",
        "    keras.layers.TimeDistributed(keras.layers.Dense(100)),\n",
        "    keras.layers.TimeDistributed(keras.layers.Dense(10))\n",
        "])\n",
        "\n",
        "def last_time_step_mse(Y_true, Y_pred):\n",
        "    return keras.metrics.mean_squared_error(Y_true[:, -1], Y_pred[:, -1])\n",
        "\n",
        "model.compile(loss=\"mse\", optimizer=keras.optimizers.Adam(lr=0.01), metrics=[last_time_step_mse])"
      ],
      "execution_count": 48,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qULy2WU_nBoy",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "08bf7d82-b897-4174-8221-28b82a9d637b"
      },
      "source": [
        "history = model.fit(X_train, Y_train, epochs=20, validation_data=(X_valid, Y_valid))"
      ],
      "execution_count": 49,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/20\n",
            "219/219 [==============================] - 2s 8ms/step - loss: 0.0784 - last_time_step_mse: 0.0763 - val_loss: 0.0577 - val_last_time_step_mse: 0.0453\n",
            "Epoch 2/20\n",
            "219/219 [==============================] - 1s 6ms/step - loss: 0.0500 - last_time_step_mse: 0.0336 - val_loss: 0.0442 - val_last_time_step_mse: 0.0220\n",
            "Epoch 3/20\n",
            "219/219 [==============================] - 1s 6ms/step - loss: 0.0421 - last_time_step_mse: 0.0237 - val_loss: 0.0408 - val_last_time_step_mse: 0.0223\n",
            "Epoch 4/20\n",
            "219/219 [==============================] - 1s 6ms/step - loss: 0.0384 - last_time_step_mse: 0.0188 - val_loss: 0.0391 - val_last_time_step_mse: 0.0179\n",
            "Epoch 5/20\n",
            "219/219 [==============================] - 1s 6ms/step - loss: 0.0359 - last_time_step_mse: 0.0157 - val_loss: 0.0360 - val_last_time_step_mse: 0.0138\n",
            "Epoch 6/20\n",
            "219/219 [==============================] - 1s 6ms/step - loss: 0.0338 - last_time_step_mse: 0.0125 - val_loss: 0.0322 - val_last_time_step_mse: 0.0101\n",
            "Epoch 7/20\n",
            "219/219 [==============================] - 1s 6ms/step - loss: 0.0324 - last_time_step_mse: 0.0108 - val_loss: 0.0320 - val_last_time_step_mse: 0.0105\n",
            "Epoch 8/20\n",
            "219/219 [==============================] - 1s 6ms/step - loss: 0.0313 - last_time_step_mse: 0.0099 - val_loss: 0.0306 - val_last_time_step_mse: 0.0091\n",
            "Epoch 9/20\n",
            "219/219 [==============================] - 1s 6ms/step - loss: 0.0303 - last_time_step_mse: 0.0087 - val_loss: 0.0299 - val_last_time_step_mse: 0.0080\n",
            "Epoch 10/20\n",
            "219/219 [==============================] - 1s 6ms/step - loss: 0.0294 - last_time_step_mse: 0.0081 - val_loss: 0.0296 - val_last_time_step_mse: 0.0083\n",
            "Epoch 11/20\n",
            "219/219 [==============================] - 1s 6ms/step - loss: 0.0288 - last_time_step_mse: 0.0074 - val_loss: 0.0292 - val_last_time_step_mse: 0.0072\n",
            "Epoch 12/20\n",
            "219/219 [==============================] - 1s 6ms/step - loss: 0.0284 - last_time_step_mse: 0.0071 - val_loss: 0.0280 - val_last_time_step_mse: 0.0061\n",
            "Epoch 13/20\n",
            "219/219 [==============================] - 1s 6ms/step - loss: 0.0280 - last_time_step_mse: 0.0068 - val_loss: 0.0288 - val_last_time_step_mse: 0.0064\n",
            "Epoch 14/20\n",
            "219/219 [==============================] - 1s 6ms/step - loss: 0.0278 - last_time_step_mse: 0.0067 - val_loss: 0.0280 - val_last_time_step_mse: 0.0062\n",
            "Epoch 15/20\n",
            "219/219 [==============================] - 1s 6ms/step - loss: 0.0274 - last_time_step_mse: 0.0063 - val_loss: 0.0282 - val_last_time_step_mse: 0.0070\n",
            "Epoch 16/20\n",
            "219/219 [==============================] - 1s 6ms/step - loss: 0.0269 - last_time_step_mse: 0.0060 - val_loss: 0.0279 - val_last_time_step_mse: 0.0062\n",
            "Epoch 17/20\n",
            "219/219 [==============================] - 1s 6ms/step - loss: 0.0268 - last_time_step_mse: 0.0060 - val_loss: 0.0284 - val_last_time_step_mse: 0.0069\n",
            "Epoch 18/20\n",
            "219/219 [==============================] - 1s 6ms/step - loss: 0.0263 - last_time_step_mse: 0.0056 - val_loss: 0.0270 - val_last_time_step_mse: 0.0049\n",
            "Epoch 19/20\n",
            "219/219 [==============================] - 1s 6ms/step - loss: 0.0264 - last_time_step_mse: 0.0056 - val_loss: 0.0269 - val_last_time_step_mse: 0.0056\n",
            "Epoch 20/20\n",
            "219/219 [==============================] - 1s 6ms/step - loss: 0.0263 - last_time_step_mse: 0.0055 - val_loss: 0.0269 - val_last_time_step_mse: 0.0054\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YynbFDQ8ZPUe",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 291
        },
        "outputId": "80b6a41e-8db0-4446-bbbf-ab41397140aa"
      },
      "source": [
        "plot_learning_curves(history.history[\"loss\"], history.history[\"val_loss\"])\n",
        "plt.show()"
      ],
      "execution_count": 50,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2jqgAvpIZPUh"
      },
      "source": [
        "series = load('/content/drive/MyDrive/Colab Notebooks/CS581-TestResults/assignment5/HW5-supp-files/univariate-time-series-multi-step-test.npy')\n",
        "X_new, Y_new = series[:, :50, :], series[:, 50:, :]\n",
        "Y_pred = model.predict(X_new)[:, -1][..., np.newaxis]"
      ],
      "execution_count": 51,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8MpOhoGfZPUl",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 293
        },
        "outputId": "8e923250-3a2e-4ec1-8a8f-a9ad06709ebc"
      },
      "source": [
        "plot_multiple_forecasts(X_new, Y_new, Y_pred)\n",
        "plt.show()"
      ],
      "execution_count": 52,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAaIAAAEUCAYAAABpmDIHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOydeXhU1fnHP+9k34CQBJBAEsK+yBYCQQQyoNZWrVYxWqO2daHFWtu6trW21tbWou2vm1VRW5fGJS6Ida+QAMqWhDVhCRAS9pCdhOyZ8/vjzsQQEsgyy51wPs8zz2TuOffc98xk5nvPOe95X1FKodFoNBqNp7B42gCNRqPRnN9oIdJoNBqNR9FCpNFoNBqPooVIo9FoNB5FC5FGo9FoPIoWIo1Go9F4FC1EGo1Go/EophQiEblbRLJFpEFEXjpH3Z+KyHEROSki/xKRgDZlcSKSISK1IrJbRC5xufEajUaj6RamFCLgKPA74F9nqyQiXwN+BiwEYoF44DdtqrwObAEigIeBt0UkyhUGazQajaZniJkjK4jI74BhSqnvdlL+GlColPqF/fVCIE0pNURExgA7gEilVLW9fK29/Fm3dECj0Wg058TX0wb0konAijavtwGDRSTCXlbgEKE25RM7akhEFgOLAQIDAxNiYmJcY7EJsNlsWCxmHQz3Ht0/76Uv9w36fv/y8/NLlVLdnnXydiEKBaravHb8HdZBmaM8uqOGlFLLgGUAY8eOVXv27HGupSYiMzOT5ORkT5vhMnT/vJe+3Dfo+/0TkaKenOft0lwD9Gvz2vF3dQdljvJqNBqNRmMavF2I8oApbV5PAYqVUmX2sngRCWtXnudG+zQajUZzDkwpRCLiKyKBgA/gIyKBItLRNOIrwO0iMkFEBgC/BF4CUErlA1uBX9vP/xYwGXjHLZ3QaDQaTZcwpRBhCEodhmv2zfa/fykiMSJSIyIxAEqpT4ClQAZwECgCft2mnRuBGUAF8ASwSClV4rZeaDQajeacmNJZQSn1KPBoJ8Wh7er+GfhzJ+0UAsnOs0yj0Wg0zsasIyKNRqPRnCdoIdJoNBqNR9FCpNFoNBqPooVIo9FoNB5FC5FGo9FoPIoWIo1Go9F4FC1EGo1Go/EoWog0Go1G41G0EGk0Go3Go2gh0mg0Go1H0UKk0Wg0Go+ihUij0Wg0HkULkUaj0Wg8ihYijUaj0XgULUQajUbjYZYuhYyMs9fJyDDq9UW0EGk0Go2HSUyElJTOxSgjwyhPTHSvXe5CC5FGo9F4GKsV0tM7FiOHCKWnG/X6IqYVIhEZKCLLReSUiBSJyE2d1PvYnj7c8WgUkR1tygtFpK5N+Wfu64VGo9F0jY7E6HwQITBpqnA7TwONwGBgKvChiGxTSuW1raSU+nrb1yKSCaxq19ZVSqnPXWirRqPR9Jq2YrRgAaxa1fdFCEw6IhKREOA64BGlVI1S6gvgfeCWc5wXB8wFXnG1jRqNRuMKrEfTGHFyK+npYCkvofGTlWet3xecGEwpRMAYoFkpld/m2DZg4jnOuxVYq5QqbHc8TURKROQzEZniRDs1Go3GeaSl8fL3MshqnMIgjnPCFsWVS+fzwX0dezH0FScGUUp52oYzEJG5wFtKqSFtjt0JpCqlks9y3j7gd0qpl9ocmwNsBgT4sf0xTilV2e7cxcBigKioqIT09HSn9cds1NTUEBoa6mkzXIbun/fSl/sG5+5fw7V/55sVr+FDM4WM4P/4CU/wcyzY+NWju5g/v7S17pYtA/jNbybw61/vZNq0yk7bdCdWqzVHKTWj2ycqpUz3AKYBte2O3Qf89yznXAzUAKHnaHs3xppRp3XGjBmj+jIZGRmeNsGl6P55L325b0qdvX+rVinVnwoFSv2enykFSoF6ggcU2JTFotSrr35VNzLSeDYTQLbqwW++WZ0V8gFfERmtlNprPzYFyDvLOd8B3lVK1ZyjbYUxOtJoNBpT4JhiG+RbTv/mKn7CX1rLHuJJwgb48cPKx7n1VlizBpYv71tODKZcI1JKnQLeBR4TkRD79NrVwKsd1ReRICAFeKnd8RgRmSMi/iISKCIPAJHAly7tgEaj0XRCR1EUsrLgjjtgb3M8v+cXbCCJpTzQWn5XzAe89EITfn7w/POwZEnfESEwqRDZuQsIAk4ArwNLlFJ5IjJXRNqPeq4BKoH2K3phwDNABXAEuBz4ulKqzKWWa3pETlEFT2fsI6eowtOmaDQuo6MoCnffDa++qpgRlMuQoJOkWN4mkWyIjTUUavt2Ypb+CL+mWgCefryCjIf7zo4Us07NoZQqxxCY9sfXAqHtjr2OIVbt6+YBk11lo8Z55BRVcNPzG2hqseHvayHtjiQSYsM9bZZG43Ta7hVyTK/9+c9w5IjwAMu4sd/bpH8YgNX61XbIjMIRpHx+J89yJ7eQxg2210n5/fWk8znWxy/xYG+cg5lHRJrziDX5JTQ027ApaGy2saFAD1o1fZdWMbq6nneifsATj9RwMWv4ne9jpL8XcNq0W0YGpKz6PumkcDOvMZUtbGUq6aSQ8sS0cwZL9Qa0EGlMwc5jJ1v/timobWj2oDUajeuxHk0jvfFb3Fz6f9QSRB6TSLfciPVoWmud1hA/tuuxkgnAIt5mPRcxmr2k264/a7BUb0ELkcbjrM4v4X87i7l66lB+eslopgzrz9OZ+/n3lwc8bZpG4zoefphpDetpwB+FD3fzD6yNn8LDD7dWycqyT9/FFrQeW8TbALzLtVhjC0hPN+p5M1qINB6lqraJh97ezuhBofzxusn8+JIxpP9gNl+bOJjf/Hcn/1i117H/S6PpWxw8yFIeQOHDbbzIMywhg2Q4eLC1yoMP2r3jHn8cgoMBGEs+k9jBO1wHl1yC1WrU82a0EGk8ym/+m0dJTQN/TplKoJ8PAAG+Pjx903SunRbNU5/l85M3t/J0xl7tTafpU2QMuoE/cT8RlPA8dxprPqSTMeiGMyunpsKyZYYXnQiL+v+Ptczl+Isfwo9+BHFxYLEYz2lpZ55vcrQQaTzGJ7nHeXfLEe62juLCYf1PK/P1sfDU9VO4bMJgVmw9ylOf5pP6wgYtRpo+QUYGpJz6N4KNG0jHgsJKJukBt5JS+++O13xSU6GwEGw2Fn15LwoLywd9H/7xDygqMuIwFBXB4sVeJ0ZaiDQeoaymgYeX72Di0H7cvWBUh3UsFmHKcEOgFNCkvek0fQCHA8KPHwqkgSC+ZXkfRCA2FuuLN5O+IvCcDggTJsDYsfB2VQeu27W1p60zeQNaiDRuRynFXWmbqaxt4o6L4/Hz6fzfMCk+El+LEZHJ18dCUnyEu8zUaFyCwwFh3z4It1Qy/+oBYLMZo53U1FbX7rM5IIjAokWQ2TCbEiLPrNBmnckb0EKkcTuvrC9i44FybErx8+XbzzrdlhAbzvPfmYFF4NIJg/UmV43X8+CDcPHF8P4KG1fa3sdv9pnBqrvigLBoEdjw4b0z9/1DTIyTrHUPWog0bufdzYeBrk+3WccO4vJJQ1i3v4zGZpsbLNRoXMuaNVBRaeFbLIekpG6fv3QplJfDyEHVvG1p59wQHAyPP+5VCfO0EGncSotNUVh2CouAj4Cfb9em266fMZzyU42s3FXsBis1GteyfDkE+TbyNcvnkJDQ7fMTE+GGGyBxQRgrWUDZwNFGweDBsGwZGUNTvSphnhYijVvZdKCcqrpm7r10DPdeNrbLMeXmjY5iSL9A0rMPucFKjcZ12Gzw3nvwtfBNBE8Z3bo/qDs41pE++QRabBbef3SzsXB0992tIuRNaSK0EGncykc7jhHoZ+G2i0fwQ+uoLq/5+FiE6xKiWZ1fwvGqehdbqdG4juxsOHIErq1+pUfTcg6sVnjnHWP70HNpoTB5MhnvVnidCIEWIo0babEpPsk7zoJxgwj2737g9+sThmNT8I59jUmj8UaWLwdfX8WV9W/BrFm9amvBArj2Wti4ER70/RMpW35G+n8avUqEQAuRxo1kF5ZTUt3A1ydd0KPz4yJDmDViIG9lH9JhfzRey/LlkDzmKOFU9mpE5OAnPzGen8xZyBKewRq4vtdtuhstRBq38XHucQJ8LSwYN6jHbaTMGE5hWS2bDpQ70TKNxvl0lIm1qCiYPXvgW+GZMGAAGYdH99qzrcaeJnT2zGYjXt2/vC9YsBYijVuw2RQf5x4jeWwUIQE9z8f49QuHEBrgS3q2np7TmJuOMrGuXWtsPr26/CUyRt1Jyo2WXnm2ZWTAzTfDmDEQFOpL+phfkvL6NV6XFkILkcYtbD5YQfHJBr5xYc+m5RwE+/ty1ZQL+GjHMarrm5xknUbjfNpmYnUIw9q1Ucya0UL+bhspO3/dK6eC1lxF6cZaUXY2zL+qP+kqhZQU5VViZFohEpGBIrJcRE6JSJGI3NRJvUdFpElEato84tuUTxWRHBGptT9PdV8vNA4+3HEMf18LC8cP7nVb188YTl1TC/e/tU0HQdWYmrZi9MYbkJ8fxuQhx0lRb5D+SK5TRMhqNUZfJ0/C3jFXYG3+H+k/2+JVCfNMK0TA00AjMBhIBZ4RkYmd1H1TKRXa5lEAICL+wArgP0A48DKwwn5c4yZsNsUnuceZPyaK0F5MyzlQNoUAn+YV64jcGtPjEKM77jBev71yIOmkYL2z42C/XaE1YZ5dyBzTe1kyE3x8sFa861UJ80wpRCISAlwHPKKUqlFKfQG8D9zSzaaSAV/gL0qpBqXU3wABFjjTXs3Z2XKokmNV9VzRy2k5BxvaOCroiNwab8BqhUsvNf6+O+a/WEcfgYieB/BtTZhnZ/x4Y19sVm4QzJgBGRlelTCv97enrmEM0KyUym9zbBswv5P6V4lIOXAM+IdS6hn78YnAdnW6r+92+/FP2jYgIouBxQBRUVFkZmb2uhNmpaamxq39e31XA74C/qX5ZGbu7XV7AZUt+Fig2WbcVQRUFpGZ+ZXzgrv75276cv/6at+2bBlARsYEbrihkGfSFzBl+tVEOLmfI0dOZeVKoWjKSIa/+SZffvwxLUFBTr2Gy1BKme4BzAWOtzt2J5DZQd0JwFDAB7gIQ4y+bS97BHijXf004NGzXX/MmDGqL5ORkeG2a9lsNjX795+r21/a5NR2NxaUqtG/+Ejd+XLWGWXu7J8n6Mv964t9W7VKqchI43ndG2+oVSSryNBatWqVc6/z058qFRioVOOHnykFSn3yiXMv0AWAbNWD33xTTs0BNUC/dsf6AdXtKyqldiqljiqlWpRS64C/Aou6247GNbyRdYijVfWMv6D9x9A7Zo6IYM6oCPaV1Di1XY3GmbR3Kui3c6eRifWpg053JkhMhPp6yAu/GPz8vMdTAZOuEQH5gK+IjG5zbAqQ14VzFcaMDfb6k0VE2pRP7mI7ml6SU1TBL9/LBWDZmgKnOxUkxUdQUHKKEyd17DmN+WgvQgD9du2CwECst8Wf4drdW1odFnKDYOZMLUS9RSl1CngXeExEQkRkDnA18Gr7uiJytYiEi8FM4B4MTzmATKAFuEdEAkTkbvvxVS7vhIZ1+0tpsRnLc80tzncqcKSP2KCjLGhMSHvPNjBGRCQkgJ9flzKxdoeRI2HAAHt7Vivk5Bg+3V6AKYXIzl1AEHACeB1YopTKE5G5ItJ2PuZGYB/GdNsrwB+VUi8DKKUagWuAW4FK4DbgGvtxjYtxDEMt3cg71B0mDu1HWICv9prTmJLTPNvS0iA2ln55ebB9u/GarmVi7SoihsNcVhbQ0mI8BgyAuLjW65kVs3rNoZQqhzNz4Cql1gKhbV5/+xztbAG6n3lK02vW5JcyOCyAW2bHMntkpNPTfPv6WEgcMVALkcbcpKXB4sVQW2vcnFVXG68BUlOdeqnERFj6Rxt1u58lCEApKCpy2fWchZlHRBovJvdIFZsKy7lzXjx3LxjtdBFykBQ/kIKSUxTrdSKNWXn4YaitPf1Yba1x3MkkJhqJ8rbWj3XL9ZyFFiKNS3h5XSHB/j5cP2O4S6/Tuk6kR0Uas3LwYPeO9wKHw0I2M9xyPWehhUjjdMpqGlix7SjXTR9G/yA/l15rwgWOdSLtsKAxKTEx3TveC6KjYYjPCbLoIKS3C67nLLQQaZzOG1mHaGy28Z2LYl1+LV8fCzNHDGSjHhFpzMrjjxvxd9oSHGwcdzIikDi50Yg554brOQstRBqn0tRi49X1RcwdHcmoQWFuuWZSfAQFpefHOlFOUQUf7G/UgV69idRUWLYMYmNRIhAba7x2keNA4reGsYexnBw2wTgQFubS6zkDLUQap/Jp3nGOn6zne3Pi3HbN82WdaN2+Uq5/dh1v723ipud11HGvIjUVCgtZvWoVFBa6VBQSE0EpIeeVPGPPUlKSqUUItBBpnMxLXxYSGxFM8piepwPvLhPOk/1Er6wvwr4/mIZmG4++n0fukSpyiip4OmOfFiYNYOwlAvt+osmTYds2j9rTFUy7j0jjfew4XEV2UQW/unICFouc+wQn4WMRZo4Y2OcdFgpKa77aJGwR9hyv5sq/f4EjgFWAr4W0O5Jc5iqv8Q4iI2HECLsQXTwF/v1vOH4chgzxtGmdokdEGqeQU1TBz97dTqCvhUUzhrn9+rNHRnCg9BTHq/rmOlF+cTX5xTXcOjuW60b7kf792WT98hKs4wZhhFrWuZnOd5Yu/Sq8XGKiXYimTDEO2EdFGRlGPbOhhUjTa3KKKrjp+Q3kHT1Jk02xt9j9EbH7+jrRm1mH8PMRfrRwNFeO9CchNpz+QX7cbR1FgK/xNRYRp4dR0ngPiYlfBVFNTDQCKpQMtQvR9u2tQVgTO/Ds9jRaiPooOUUVPPnpbpevGxyvqmfpJ7tpaLYBRn4rT4jB+Av6ERbYN9eJGppbWL7lCJdOGExkaMBpZQmx4bx2ZxKDwgKIjQjW03LnMY4gqikp4GtfdMnaFw7DhpHxaeMZkcDNhF4j6oPkFFVw47L1NLUonltdwJuLk0iIG+jU9ldsPUJh6SnW7S+jxaZwLAn5uyC4aVfwsQiz7HHnLo9w3/qUO/h85wnKTzVyQ2LHGxITYsP5/vyR/PaDnewtrmb0YPe4zWvMh0OMrr/eeJ2VBUHRN5OSuYT0/5lThECPiPoka/NLaGqxp1+wKZ74ZLcjO22vWbW7mOufXccr64tYs7eUyycOYc2DVt76wUXcd9lYjy6WJ8VHUFhWS/qehj7lQfZG1kGiBwRx8ajITut8c8pQfCzCu1uOuNEyjRmxWuGtt8DHB/7zH0jZ/kvSVQrWixo8bVqnaCHqgxytqgOM9As+FiGrsIKfvbOjNTdQT9lfUsNP39za6kLsIzB+aD+GDzSmhH5oHeXRqSFHOKGPDzST+kLf2GdzqLyWL/aVsihhGD5n8USMCgtg/pgo3ttypNefs8b7sVph4kTYtw9+8PVCrLaVsGuXp83qFC1EfYwT1fV8sP0YSfEDue+ysaQvTuKeBaN4M/sQ97y+hUb7Wk532VhQxrX/XIdSxvSbj4tyDPUGR2QFRd/xIHsr5zAA13fBE/G66cM4VlXfJ/qt6R0ZGVBQYPz9z5XjyCDZ1PuJ9BpRH+P//reXxmYbT1w7mbjIEAAS4gYSFujH4x/t4khlHXEBjYSNqOjy6GXF1iM88NZ2hg0M4qXvzqSkpoENBWUkxUeYanF89shILJKPTZlPJHtCi03xVvYh5o6OYlh48DnrLxw/iLBAX97JOcycs0zjafo2Du+4J5+EJUvgR/dAym/TSf/gLazf8bR1HaNHRH2IvcXVvJl1kJuTYltFyMGd8+JZMj+erYcqeW9f10LE5BSWc8uLG/nxG1uZFjOAd5dcREyEOabhOiIhNpzFc+MBWHrdZNPZ113W7C3hWFU9NyZ2LZVGoJ8PV04eyse5xznV0Oxi6zRmxCFC6elw223g7w+19T6kj3mElBU3te4zMhumFSIRGSgiy0XklIgUichNndR7QERyRaRaRA6IyAPtygtFpE5EauyPz9zTA/fzxMe7CfH35Z6FozssDw30a92Z39BsIz278/wkn+Yd5/rn1rN2bykWgZ9eOpoBwf4usNq5pCYZEb9Larw/G/ybmw4xMMSfS8YP7vI5102Ppq6phU9yj7vQMs258ETYpbYiZLUaIjR1KmzaBNa5zaQHfZeUFGVKMTKtEAFPA43AYCAVeEZEJnZQT4BbgXDgcuBuEbmxXZ2rlFKh9sdlrjTaU6zbX8rK3Sf44YJRDAzpWDCS4iMI8LMgGG9aetZhnl29/zSPuobmFp7O2McP0za3OiUIkFNU6fI+OIPhA4MZEiys3VviaVN6xardxXy28zgXj4rA37frX9OE2HBiBgbz7pbDLrROczbW7Stl0bPreOrTPW51msnKOnOf0MyZkJ0NLZOmYD25gvRnyo2ICybDlEIkIiHAdcAjSqkapdQXwPvALe3rKqWWKqU2K6WalVJ7gBXAHPda7FlsNsXvP9pF9IAgvntRXKf1EmLDSbsjietG+/HqHbP4xoUX8MTHu1nyn82s3VvC/W9tJfnJTJ78dA8JseEEmNQp4VxMjPRhQ0EZDc0tnjalR+QUVbD4lRxsCj7NK+7WD5mIcO30aNbtL+NoZZ0LrdR0xgfbjxlhl3Cv08yDD565T2jmTDh1Cnb1TwLAGprFgw+6xZxuYVZnhTFAs1Iqv82xbcD8s50kIgLMBZ5rV5QmIhZgC/CAUuoM9xERWQwsBoiKiiIzM7Pn1ruZ5XsbyT3SxNUj/djw5dpz1k8e3Ejz4VwWDVWENfrzZt5xPsn7airnpnF+XBbXwL5B/uwub2HcQB+qD2wj84Are+E8RoU2sbJJeHFFJhMifDxtTrf5YH8jzfbhaFOzjdc/z6J65Fej3JqamrP+fw5ttKEU/PmdtVw50vzTqW05V9+8gZJiY7+OBWOLQ0BlEZmZxgjV3f1TKgiYRdpWP/4A7F++nEOBgW67flcxqxCFAifbHasCzrVl/FGMz//fbY6lApsxZph+DHwqIuOUUqfNNSmllgHLAMaOHauSk5N7artbeXfzYVbsN3T104Mt3HrZlHMu0mdmZuLonxXgne28kXUIML440bHxJCePItl1ZruU+s8z8NtTR1XwUJKTx3vanG7jP6yUt/duRAB/PwvfviTxtM+07efXGW8dXMeWykaenD8fEe+JNNGVvpmdd45tIbKslO/NGXGGZ6m7+2ezwT33QGXjVBg+nJHV1Yw04ftryqk5oAbo1+5YP6C6sxNE5G6MtaIrlFKtW4iVUl8qpeqUUrVKqT8AlRijJq8np6iCn7+7o/V1T6cBrp8xnEA/75yG64hAX2F6TDhr80s9bUqPcKTQ+Nb06B5Hqrh2+jD2l5ziVyvy+sTGXm9i66EKZo2IMIVnqcViBDndtAkjEvf27R61pzPMKkT5gK+ItHX/mgLkdVRZRG4DfgYsVEqda5VWAd5zi9gJa/JLuPmFjYQH+/V6LcexdnSvh0P0OJN5Y6LYeewkJdXmDWvSGdmFRl6lX105ocefxbDwIAD+s6Goz0SZ8AbKaho4VF7HlOH9PW1KKzNnGvpTNyEBdu+GevOlSjGlECmlTgHvAo+JSIiIzAGuBl5tX1dEUoHfA5cqpQralcWIyBwR8ReRQLtrdyTwpet74To+2H6U21/OYkRkCP/90Vxeu7P3ImLWvUE9Zd7oKAC+3Od9o6LsogrGDA7tlbv89sNVQN+KMuENON73KcMGeNiSr5g5E5qbYWvYXGhpgZ07PW3SGZhSiOzcBQQBJ4DXgSVKqTwRmSsibRPe/A6IALLa7BV61l4WBjwDVABHMNy7v66U8spvpeFNlc3dr21h6vABvL44iaiwgD4nIs5g4tB+hAf7scbL3LhbbIqcogpm9DJaelJ8BL72KT5fH++fbvUWthyqxCIwKdo8IyJH/qFNdZOMP0w4PWdWZwWUUuXANR0cX4vhzOB4PeIsbeQBk11ioJtpm9rBIvDTS8a0BvnUnInFIlw8Ooq1e0tRSnnNgn1+cTXV9c3M6OVNRUJsOH//9jSWpG3m2zOH65sUN7HtUCVjBocREmCen9ahQyE6GjYdiIKgIFPGnDPziEjThg0FZa2pHQTjzktzduaOjqSkuoHdxzv1cTEdjvWhRCfkj/r6hRcwdnAYe467P2Pu+YhSim2HK5k63DzTcg5mzoRNWRaYNEkLkabnJMVHtHpY9AXPNnfgWCfypigLWYUVDO4X0Ops0Fus4waRVVjOyfomp7Sn6ZyD5bVU1jYxxaRCtG8flPsPgdWrDXe6uDhIS/O0aYAWIq9heHgQCrCOjeoznm2uZkj/QMYMDmXtXu9xWMguLGdG3ECnTSUuHD+IZpvyWld2b2KrfZbCTI4KDmbONJ6zNzQbm4uUgqIiWLzYFGKkhchL2HzQcL/90cLRWoS6wdzRUWw8UE59k/nD/RyprONoVT2JTvx8pw0fwIBgP1buLnZam5qO2XqokiA/H8YMDj13ZTeTkACCjU0t008vqK2Fhx/2jFFt0ELkJeQUVeDva2Hi0Pb7fDVnY96YKBqbbWw8UO5pU86JY32otx5zbfH1sTB/TBSr95TozK0uZtuhSi6M7o+vj/l+Vvv3h3HsZhMzzyw82HkUfnfh9HdMRP4mIh90cLyfiDwqIuPbHPuJiOywx4HTnIWcogomR/cnwNf7Yqd5kplxA/H3tbA23/zrRNmFFYQG+DJuyLkiWXWPBeMGUXaqkW2HtYOLq2hqsZF79KSpNrK2Z2bITjYxkzNuR2JiPGHOaThVAERkJPADjJhv7ZkB/Bpo63P8HBAFmDRvoDmob2oh98hJPSXXA4L8fRg3OIz3th4xfXSBrMJypsUMcPod9fwxUVgEMnafcGq7mq/YfayaxmabKR0VHMxcFEMxQzhEm0SLwcHw+OOeM8qOs0ciPwG2KaWyOyibBjQArdt6lVJ1wCvA/U62o0+Rd7SKxhYb07UQdZucogp2HjtJaU1jl7LSeoqquib2FFczI9Z503IOBgT7MyN2ICt3aSFyFVvto00zum47SPyhMS23KdieK2LYMFi2DFJTPWiVQZeESERGiUiTiDzW7vgz9syoM0QkALgZeK2D83cBTwEBQJOIKBF5x178BjBBRC7qVSpYRxkAACAASURBVE/6MI4fz+kxWoi6y4aCMmz2xH9NLeYNdbPlYAVKQWKcaz7jBeMHsfPYSY5V6RxFrmDboUoiQ/2JHuAct3tnsnSpkb118mQja2vWFY8aBS+/3CpCGRlGPU/RJSFSSu0DXgB+IiIRACLyK+A24Fv2EVASMADoKCHOrUAB8F9gtv1xr71sK0ZU7ct73o2+TU5RBbERwUSFBXjaFK8jKT4CP/tUl0XEtPuvsgsr8LEIU2Ncc0e9YNwgADJ2m3+tzBvZeqiSKcMGmDKCR2KikUJ83Tp76vCj0UZBTg7wVYpxRyggT9CdqbnHAB/gZyJyB8Z6zy1Kqc/t5UkYMRY7CmS0DRgGrFJKbbA/igCUUjZ7eVIP+9CnUUqRU1RJgh4N9YiE2HBeu2MW/j4WLp0w2LTrbFmF5Uwa2o9gf9eEhhk9KJRh4UGs0m7cTudkfRP7S2pMuz5ktRopxFNSjHA/2dv8aRkeB5s3t4pQ+xTj7qbLQqSUOgb8BfgR8Cxwj1IqvU2VocBJpVRjB6dPBPwxEtR1RIn9fE07DpXXUVrToNeHekFC3EDGXRBGTUOzp03pkMZmG1sPVTrVbbs9IsLCcYP4Yl+pV+yp8iZyD1ehlLnXhxxitHIl1NTA7tFXkfGFnylECLrvrLAXY51nvVLq6XZlgRjOCB0xHWO0tLWT8jqMSNuaduQcNPaWmPVO3luIjwxh/wlzxlzLPVpFQ7PNZetDDqzjBlHfZGO9SdfJvBWHo8LkYeZ13QZDbP7+d+Pvnx+5m5TDfyL936c8LkLQDSESkYUY7tbrgTki0j6qdRnGGlFHTAP2K6Xap/92MBDQMUg6IKeogrAAX8YMdu7ekvON+KhQjlbVU9tovlGRYyNrggs85tqSFB+Bv4+Fv36+17Teg97I1oOVjIgM6VX+KHdxyy0QGAj/3TOGJTyDtV+Op00Cuu41Nx1YjuGwkAwcBP7QrtpuwF9EhnXQxATauG13wAhgT1dsOd/ILqxgaswAfCzmWwT1JuKjQgA4UHrKw5acSXZhBXFucEbJO3qSZpsxDaiztjqPbYcrmWLy0ZCD1auNUHPhA2w8wxIy3jSHS/85hUhERgEfA58BP7KvAf0G+IaIzGtTdY39uYMYElQCU0TkayKS5PC8s7c/ABjT5nyNnep6Y2+JnpbrPfGRRvyvghJzCVFOYTlr95YyIjLE5dfaUFCG3ZOdRp211Skcr6qn+GSDqdeHHDgcE26/HSoqLfwz/GFSXryMjAxPW3YOIRKRIRgCtAtItXu4gbEJdTfwhKOuUqoQ2ARc1UFTvwKKgfcwpvbGtym7AmjEGHFp2rD1UCVK6fUhZ+D4oTeTEOUUVXDTCxupa2rhi32lLh+hJMVH4OdrfOV9LOZ1Zfcm3t18CIAAP3OH3mrrHXf77caxlrhRpA/5MSkpeFyMzipESqnjSql4pVSyUqqhzfEWpdR4pVT7TajPANeKSHC7dnKVUrOUUkFKKVFKfdGm+Gbgrfbpu0VkoIgsF5FTIlIkIjd1ZKMY/FFEyuyPP0obZ34RmSoiOSJSa3+eeva3xDzkFFUgYm5vHG8hyN+H6AFBFJSax2FhQ0EZjc3GvV2LTbl8hJIQG86rt83ER4RvXHiBvsHpJTlFFfzpf3sB+M37eaad6mzvoj1lCoSFwRqfZKyHXiH95TqPi5GzQ/z8BzgK3NWVynZRWIAx1deepzFGSoOBVOAZEZnYQb3FGCnFp2CkBb8K+L69fX9ghd2ucOBlYIX9uOnJKapg7OAwwgJ1SnBnEB8VYqoRUVJ8BBb7PZO/m5IdzoqPYMLQfpTVdLTLQtMdNhSUtUY0N3PUjqys0120fX3h4othdfE4sNmwhm8lPd2o5ymcKkRKqWbge0BtF08ZAnzXHrmhFREJAa4DHlFK1dhHUO8Dt3TQxneAPymlDiuljgB/Ar5rL0sGfIG/KKUalFJ/w8i0vaBbHfMALTbF1oOV+q7VicRHhlBQUoNS5kiHkBAbTniIH2MHh7k12eGk6H7kHq0yzfvgrUy3R8EQzJ01+cEHz9wnNG8e7DzUjxIiIScHq9Wo5zGUUqZ7YLh717Y7dj/w3w7qVgGz2ryeAVTb//4p8HG7+h8A93XQzmIgG8gODg5WGPuePPbwi4xVsQ99oEImWj1uS195hE2/UsU+9IHyCR3ocVsA5dNvkIp96AMVOu0Kt143dOrXjfehX5TH3wNvfgTEXKhiH/pARXzjJ8p/6DiP29O9x2wFSv2La9SLzm07uye/+WbNAxQKtN9zVAV0tJkm1F7Wtl6ofZ2ofVmn7SillimlZiilZgwbNszjYvzvD1YDkLf6v05vOyMjw+P9c+Wjs/59+Ma/AFizdY/HbVRK8eYqYw9H1kevufXzW/u+EZf4gy+3efw9cHbf3Pn4/QtvYxEoeGcpDUd2eVX/GhrWERQE22K/w21Tpjit3Z5iViGqAdqnIu2HERz1XHX7ATXKeFe6046p+CzvOMH+PpRWdxasQtNdRkaZy4V744EyBgT7MWaQezcrjxsSho9FyD3S2f5yTVfYUFDGpOj+9PPCNVx/f5g9G9bUz4S8PKiv96g9ZhWifMBXREa3OTYFyOugbp69rKN6ecDktl50GA4NHbVjGnKKKliTX0ptYwupL240rTeOtzGkXyBBfj6mEaJNB8pJjBuIxc2blQP9fBg9KJTco+0nCzRdpb6pha0HK027LtQV5s2DrScuoKo5GHbs8KgtphQipdQp4F3gMREJEZE5wNXAqx1UfwW4V0SiRWQocB/wkr0sE2gB7hGRABG52358lSvt7y2f7yrGMcht0hsPnYbFIoyIDDGFC3fxyXoKy2qZNcK1YX06Y1J0f3KPaIeFnrL5YAWNLTaS4j3z+TmDefNAKeELLobNncWjdg+mFCI7d2EEQj0BvA4sUUrlichcEWn7S/IcRp6jHUAu8KH9GMqIAnENRj6kSoz8SdeojiOEmwbHh2IRc3vjeCNmceHeeMCILzdrhGc+20lD+1Fa08gJPfXbIzYUlGMRXBox3dUkJYGfn2JNwGWtuYk8hWuSnzgBpVQ5hoi0P74WwwnB8VoBD9ofHbWzBUhwkZku4UDZKQYG+3P73DiS4iO1C7cTiY8K5aMdx2hobiHA13O74TcWlBEa4Mv4CzwTzHZStBEbLfdIFYP7BXrEBm/Gm9eHHAQFwcyZwprcS2FzR5NN7sPMI6LzkqYWG2vzS7l0wmB+aB2tRcjJjIwKwaagqKyrW91cw6YD5STEhuPr45mv4PgL+iECO47odaLu0hfWhxzMmwfZ1WM5tX0/NHpuokgLkcnYXFRBdUMzyWOjPG1Kn+Sr4KeeWycqq2lg74kaZnlwfSEkwJf4yBDtOdcD+sL6kIN586DZ5sP6pgTDe85DaCEyGZn5JfhahDmjIz1tSp9khD0dxH4PrhNlFTrWhzz7Q3ZhdH/ytOdct+kL60MO5swBi0WxhnkedVjQQmQyMveUkBAb7tVzz2YmNMCXwf0C2O/BEdHGA+UE+lm4MNqzwWwnRffnWFU9pTXaYaE79IX1IQdhYTA9tpzVzIc77oC4OEhLc7sdWohMRPHJenYdO0ny2EGeNqVPEx8Z6lHPuY0F5UyPCcff17Nfv4lDDYeFvKN6eq6r9JX1oaVL7dG209KYdyiNjcyingAoKoLFiyEtjYwMo5470EJkIlbvKQHQ60MuxnDh9kzw06q6JnYdP8lMD0/LAUwYagQdydUOC12mr6wPJSYaqSEy7vuA+c2f00AgWSQahbW1ZNz3ASkpRj13oIXIRGTmn2BIv0DGDfGMS+/5QnxUKCfrmyk75X4voezCcpTy3P6htvQP8iM2Ithr14lyiip4OmOfWyOP9JX1IavVSA2RUvy31s3zazASbmeQTErx305LHeFqTLuP6HyjqcXG2r2lfGPSBZwekUjjbOKjvsrWGhka4NZrbzpQjr+PhWkx5kh2OGlof7YfqfS0Gd0mY3cxd7ySg82mCPCzuC2NRl9aH7JaIX3wPaQUv8gI9rOa+VzEOlJIJ33wPVitr7vNFj0iMgmbiyqortdu2+5gpAdduDccKGfK8P4EmiS19KTo/hwqr6OqtsnTpnSZspoGHnx7Oy02hQLqm2w8v6aAhuYWl163r6wPtcX6pytJD7iVYwwlk2RDhAJuxfqnK91qhxYik6Ddtt1HdHgQ/r4WCkrd67BwqqGZ3CNVplgfcjAp2lgn8pbpubKaBlJf2EhlXRP+PhYsYiSm+yTvOPOWZvDrFbn83//yXTJd11fWh04jNRXrizeTEvIhTfgzk41Y/3g5pKa61Qw9NWcStNu2+/CxCHERwW4fEeUUVdBiU8w0wfqQA4fnXO7RKi4aZe6bIIcIHSg9xUvfm0mgnw8bCspIGjGQU40tLP1kNy+vLwLguTX7nT5d11fWh9qTMTSVj4LgguAmPi25nJUHPmKhm23QQmQCHG7bD10+ztOmnDfER4aSX+zetFQrth5BxBBCszAwxJ/oAUGmj7BQVtPATc9vpLDsFP/6biJz7KLZVmh2HKki7+hJFNDQZEStd64Q9Z31IQcZGYb3XHo6nCj25cZvC9c+vZD3rnafowLoqTlToN223U98VAgHy2tparG55Xo5RRW8u+UISsEdL2eZKsfUxKH9TJ2bKGN3MV/7yxoKSmtOE6H2JMVHEOBn/KQpICHWeQ4h6/eXklNYwYjIEKe16WnaipDVCtctEuLCShnecoCUFGXsM3ITWohMgHbbdj/xUaE02xQHy90T/HRNfgmObUtmyzE1Kbo/B0pPUdPQ7GlTziCnqILbX86mtKYRQc7q5JEQG07aHUmkzBgGQH7x2adeO3P/zimq4B+r9rJ88xFe3VBE6gsb+fbzG2lRio93HDPVTURPaS9CAL6+8NObislTk/h1ym5jn5GbxEhPzXmYZu227RHaunA7Uoi7EseduhlzTF0Y3R+lYOdRc2y0bcuKLUew2QW8xXbu6baE2HCmxwzgQOkpns7YR8qM4R2KV05RBanPb6Ch2YaPRfjmlKEE+vtwoKSGjQfKW68JEBrw1fktNuX0KT9PkJVFh/uEbntsBI8+V07Gp42kpxv13DFFp0dEHuaNTYeorm8mZmCQp005r3C4cP9nQ5Fb7nAPV9QR6GvhJ5eMcduel64y0e459+zq/aa72y8sMzwbfboh4CLCTy8ZQ/HJBt7YdLDDOl/uK6G+2YYCmm2K97Ye4bO84+wprmkVIQEWz43n5e/NJNDP0i0bzM6DD3YsMKGDglky8jOW77+Q4cONeq3hgM5Cb8MBaSHyIDlFFTz6XyP0+t9XuXeH+PnOPrvH3Or8ElJf2ODS914pxeo9JcwbE8U9C82XY+pQeR0Aq3afcPl70R2OVdWxbn8ZV154AfdeNrZbAj57ZAQzRwzkn5n7qW86fX+RzabYdMDoowCBvhbe+v5ssn95Kc/fOqNVdAL8LHxt0hAS4gaSdkdSt23wRpYuhWnz++FHE//3a2Ojc2s4ILsYtRcdxzRfb8IBmU6IRGSgiCwXkVMiUiQiN52l7gMikisi1SJyQEQeaFdeKCJ1IlJjf3zm+h50nQ0FZTTbb7+aWsy1btDXafteu3rNZn9JDUcq65hvUmcUd74X3eHldUXYlOKhr4/jh9ZR3RIAx6joRHUDaRtPHxU99dkevthXSuqsGO7/2ljS7kwiwe6S7Vhnai86CbHh3bbBG0lMhCXvfY2FrOTfb4VQWtomHFAK/PnPp4tOR2tNPcGMa0RPA43AYGAq8KGIbFNKdZS1SYBbge3ASOAzETmklHqjTZ2rlFKfu9ronjB6kDE9JPSdIb+3kBQfga9FaLYpfH1c+95n2r0i548xpxC5873oKqcamnltYxGXTxrC8IHBPWpj9sgIZsdH8Ezmfm6aGUOQvw9pG4v4Z+Z+bpoVw++umdThumxCbHifF5zOsFoh/W0frrtkDnVNfvzzn/CrXxnHf/5zuP9+eOop47WzRAhMNiISkRDgOuARpVSNUuoL4H3glo7qK6WWKqU2K6WalVJ7gBXAHPdZ3DtKa4ygm9+dE9fnh/xmIyE2nKeunwLA9y6Kc+l7vzq/hFGDQhkW3rMfVFeTEBvOb6+ZBMCPF5rjrv+dzYc5Wd/M7ReP6FU7P710DKU1DaRtLCJj9wkeeS8X69goHvvmRO0c1AlWK7xzbRr+NPDnPynq6mDVKnj8cbjtNnjkEZg4ERYtco4IAYgnQuF3hohMA75USgW3OXY/MF8pddU5zhVgM/CcUupZ+7FCIAhDcLcADyiltnVy/mJgMUBUVFRCenp67zt0Dv66uZ6DJ208NT/IrV+KmpoaQkNd7ynmKbraP6UUD66pIzrUwk8SAl1iS0Oz4ocra1kY68u3xzknwKorPr9mm2LJ57UsGO7Lt8e7NxBsW2pqaggOCeHna+sI8RMeSQrs9XdjaVYdByptNNogKkh49KIgAn09I0Le8t0Lz8lh1f2HuJe/0L9/I1VVfhhzN19xxRVHuf/+/NOOWa3WHKXUjO5ez2xTc6FA+y3eVUBXNtg8iiE4/25zLBVDnAT4MfCpiIxTSp0RblgptQxYBjB27FiVnJzcXdu7RUNzC0tW/o9FCTFYrZNceq32ZGZm4ur+eZLu9O+K6jzeyDpI0py5LglEump3Mc0qm5sXTmfuaOdMzbnq85u2dz3Hm1pITr7Y6W13lczMTBqjxlFcm8M/bpqKdfLQXrd5JLCIh9/LBaCiUYgaPdVjoz6v+e7Nns3kX0XwYsCD5FUMZfp0I4FrczP85jdw113wzDNDUWqoU0ZEbp2aE5FMEVGdPL4AaoB+7U7rB5w1FouI3I2xVnSFUqo177FS6kulVJ1SqlYp9QegEpjr3F71jI0F5dQ1tWAdZ851g/OFBeMGUd9kY93+Upe0v3pPCUF+PiR6QXyyxLhwco+epLbRsxtbX/jiANEDgrh84hCntFdZ19R6L9+snYK6RkAAmdGpFFf48Qi/5eDWMuq+yOGxx+Ctt+Cxx75yYHDGple3CpFSKlkpJZ08LgbyAV8RGd3mtClAR44KAIjIbcDPgIVKqcPnMoH240sPsWr3CQJ8LcyON3egyb7OrPiBBPv7sHLXCZe0n5lfwuyREaZJ+3A2ZsQNpMWm2HrIc/mJCqta2HSgnO/NicPXxzk/T47QP31pH5CryXj4c1L2/o50UniMX/Fz2++4/7Vp/PyynNYRUFtvut6KkamcFZRSp4B3gcdEJERE5gBXA692VF9EUoHfA5cqpQralcWIyBwR8ReRQLtrdyTwpWt70TUy95zgopERBPmb/weqLxPg68Pc0ZGs2n3C6anDC0tPUVRW6zUxBKfHhCMC2YWe2UeUU1TBv3IbCPLzISVxuNPa7cwlW9MxGRmQ8sQ00knBSiYZJPMHfsFT3Mcf3og7TXScJUamEiI7d2E4GJwAXgeWOFy3RWSuiLQNIPU7IALIarNX6Fl7WRjwDFABHAEuB76ulPL4uLygpIbCsloWjBvkaVM0wMJxgzlWVc+uY86Nxr0639xu2+3pH+TH2MFhZBWWu/3aOUUV3PT8Bg5WKxpbbOw9R5y47nK+7APqLa0u2bbrsZIJQBaJpJPCvfyFdNv1Z4iOQ4yysnp+XdMJkVKqXCl1jVIqRCkVo5R6rU3ZWqVUaJvXI5RSfkqp0DaPH9jL8pRSk+3tRCilFiqlsj3Rp/ZktEbb1kJkBpLt63Srdhc7td3MPScYERlCbIT3RGyeERfO5qIKmt0UldzB+v2lNDQb11RK6XUcD9Eagy72qwmmB3myVZSssQUdio7VaoQD6immE6LzgYzdJxg9KLTHG/U0zmVQWCBThvVn5W7nrRPVN7WwvqDMa0ZDDhLjjCRzu4+7N1eTI8yQAP56HcdjtMage/xxCO7g9ykiAut3Y3nwZxaIi4O0NKdcVwuRm6lpaGbjgTI9LWcyFowbzNZDlZTWNJy7chfYdKCc+iabacP6dIYj+6g7482t21dKes4h5o+J5NrRfnodxwykpsKyZRAbCyIQEwPjx8PmzXDwICgFRUWweLFTxEgLkZv5cl8pTS1KT8uZjIXjB6GUMVp1BqvzS4w7exOlBe8K0QOCGNo/0G3rRCeq67nnja2MjArlmZsTuGqkvxYhs5CaCoWFYLMZonPq1Jl1amvh4Yd7fSktRB1QVq9cdkeYsfsEYYG+zIjTXzYzMXFoPwb3C2CVE4UoKd47vSIT4gaSVVjudC/C9rTYFD9+fSs1DU08fdN0gv3Ntr9ecxqHDnV8/GDHqTa6gxaiDqhuVKQ+7/xw+EopMvacYN7oKPyctEdC4xxEhAXjBrMmv4TG5t4t1H+Se4x9J2oYGeU9TgptSYwLp/hkA4cr6lx6nb+u3Mv6gjJ+e/UkxursxOYnJqZ7x7uB/jXshAYXhMPfeewkxScbvGZfyfnGwnGDONVobKjsKTlFFdz92hYAXtt40DS5fbrDjFjXrxP964sC/rZyL9axUVw/w3l7hjQupCMHhuBg43gv0ULUGYLTPXcc6w96fciczBkVSYCvhZW9cOP+cMex1hxT3hpOZuyQMMICfF22TpRVWM5jH+wCYP3+Mq8U6/OS9g4MsbHG69TUXjetJ2U7IMhXUAoC/Zyr0+9vO8qQfgEcLK8lKsxzEY41HRPk78NFIyP4aPsxIkP9SYqP7NbCeXOLjTX2TazeHE7GxyJMiw13WYSF9Oyv1hocCSG1g4KXkJrqFOFpjx4RdUBUkODva+GNTZ0szvWAzD0nyC+uofhkg6nSMWtOZ9TgMIqrG/jTZ/nd/pyeydzPvhM13HfpGK8PJ5MYG86e4mqqapuc3nbJScNF3pvFWuNc9IioAywC37jwAt7bcoRffGO8Uzyf3sgyRE3xVTpmb/2R6sv4WoyYuDbVvc8p90gVf125l29OGcqPFo4+Z32z07qf6GA5C8YNdlq79U0tZBdVsGBcFAmxA0mKj9DfA40eEXXGjYnDqW5o5oPtR53S3vGqegR9F2h2Lhk/GB+7GHU1bXZ9Uwv3pm9lYIg/j1090dUmuoWpwwfgaxGnT899vquYmoZmbr84Xsd+07SihagTZo4YSHxUSOtIpjdU1jaSe6SKb04d6vVTNn2dhNhw/vWdGQT4Wpg4tF+XPqc//y+f/OIali6azIBgfzdY6XqC/H2YGN3f6UL03pajDAoL0DdimtPQQtQJIsK3E2PIKaogv7h3cbc+21lMs01x+8Uj9F2gFzB/7CDuvXQMmw9WnjNh3qvrC1m2poBLJwzuc96QibHhbD5Uwd9W5jtlTbOytpHV+Sf45pShraNOjQa0EJ2Va6dH4+cjvL6pdzuHP9pxjGHhQVwY3d9JlmlczXcuimNo/0Ce+Hg3NlvHEQY+31nMr1YYORvX5pf0OQeUgSH+NLco/vL5Xqc42Hy44xhNLYprpkU7yUJNX0EL0VmICA3gsolDeHfzEeqbWnrURlVtE1/sLeWKCy9ARN8FeguBfj7ce9lYth+u4sMdx84oP1ZVxwNvb8MhUU1eumfobDgiTLR13OgNK7YcZWRUCBOH9nOGeZo+hBaic3DTzBiq6pr4JPd4j87/dOdxmm2Kb1x4gZMt07iab02LZtyQMJ78dM9pYX+OVNZxw3MbqG+y4e/bd1NQzx0T1epF2FXHjc44XFHLpsJyrpkarW/INGeghegczI6PIGZgcI+n5xzTcpOH6Wk5b8PHIvzs6+M4WF5L2sYiAA6V13LDc+upqG3ktTtn8fqdfTcFdUJsOM/enIBFYMG4Qb3q3/vbDO/Tq6fqaTnNmZhOiERkoIgsF5FTIlIkIjedpe6jItLUJk14jYjEtymfKiI5IlJrf57aXXssFuGGxOFsPFDObz/Y2a158qraJr7cp6flvJn5Y6K4aGQEf/5fPr9ekcu3/vkl1fXNpN0xi2kx4X0+BfUlEwZz7fRhZO4poeJUY4/bWbHlKNNjBhAToZNBas7EdEIEPA00AoOBVOAZETnb5ow326UKLwAQEX9gBfAfIBx4GVhhP94tHJGBX/ziADcuW09OF2NwfbbzOE0telrOmxERvjU1mur6Zl5eX0RpTSO/vGI8k4cN8LRpbuPOufHUNbW0jgq7y65jJ9lTXK2dFDSdYiohEpEQ4DrgEaVUjVLqC+B94JYeNJeMETniL0qpBqXU3zAyES/obkN7jlfjGM80tSh++PoW1u4tOWe+Fj0t1zc40SZrq0XgRLVzsrh6C2OHhDF/TBQvrSvqkdPOe1uP4GMRrtA3ZJpOMFuInzFAs1Iqv82xbcD8s5xzlYiUA8eAfyilnrEfnwhsV6erxXb78U/aNyIii4HFAFFRUWRmZraWBVS24GeBZpsRdPZUXT23vLiJmDAhvr8PF0f7Mir89DBAp5oUa/JruTTWj9WrV3ex++6hpqbmtP71NZzdP8fn32IDX4GAyiIyMw87rf3u4onPb1b/FlbnN/DHN1cxf5hfl8/Lr2jmlZwG4vpZ2JG9/pz19f/m+YnZhCgUONnuWBXQWdasdGAZUAzMAt4RkUql1Ov2tqq62pZSapm9LcaOHauSk5Nby5KBadMr2FBQRlJ8BJOi+/HkJ3t44YsDHKxuZu3RFl6/cxYz26SFfiv7EC1qOz+4YiZTh5trGiczM5O2/etrOLt/yZz++Xt6PcgTn998pfjwyBesPWHjkZvmYenChtScwnKe/N8GmlrgYLUibMSUc753+n/z/MStQiQimXQ+uvkS+BHQfpNBP6DD0AZKqZ1tXq4Tkb8Ci4DXgZrutHUuEmLDT/sShYf4YxFjj0WLTXFf+jbevWtOa3qHj3YcI3pAEFP0tFyfoP3nf74hIiyeF8+P39hKZv6JDgOh5hRVsKGglAv6B7G/pIb/bDhIU4sxIWGzqTMCyNpsNkpLS6msrKSlxZjy69+/P7t27XJPpzyAN/fPx8eHAQMGEBkZicXirAcSPAAAFhdJREFU3FUdtwqRUir5bOX2NSJfERmtlNprPzwFyOvqJaB1OScPuE9EpM303GQMZ4hekxQfgb+vhaZmGxaLcPxkPVf+fS3/TJ3OqEFhfLGvlO9eFKe95TR9hm9ceAF//Hg3y9YUnCFE7289wr3p21qTAloELozuz85jJ7HZVIf7rA4fPoyIEBcXh5+fHyJCdXU1YWF9N224t/ZPKUVTUxPFxcUcPnyYGCekB2+LqabmlFKnRORd4DERuQOYClwNXNRRfRG5GlgDVAKJwD3AL+zFmUALcI+IPAvcaT++yhm2JsSGk3ZHUut0TZCfD0vScrjhuQ1MHT6AphZFfFSoMy6l0ZgCPx8Lt108gt99uItH3sslMS6cA6W1fLTjGHvaxGMUYEnySB742jj7KKnjKc1Tp04xduxYp99da5yPiODv7090dDR79uxxevumEiI7dwH/Ak4AZcASpVQegIjMBT5WSjl+4W+01w0ADgN/VEq9DKCUahSRa4AXgCeAXcA1Sqmeb4ZoR/vpmvfvvpjbX8oi277X6Df/zWPM4LDzekpH07dwbGV4dUMRr24w3Llnxg3k9jlx/GfjQZpbbPj5WlpHTOea0tQi5F246vMynRAppcqBazopW4vhhOB4/e1ztLUFSHCqgWehf5AfyWOjyCmq0AnwNH2S7YerEL6aA/9B8kgeunwcAN+YPNQ0Dh0a78J0QuTtzB4ZSYDfPpqabX0y/pjm/CYpPoIAP0vr//cl479aKzrfHTo0PUcLkZNpv3akv5iavoT+/zY3jz76KG+//Ta5ubmeNqVb6AlaF9DX449pzm/0/7fB5s2b8fHxYc6cOd06Lzk5mbvvvttFVnknWog0Go13k5YGcXFgsRjPaWluuewLL7zAXXfdRW5urtfuDTILWog0Go33kpYGixdDUREoZTwvXuxyMaqrq+O1115j8eLFLFq0iBdffPG08g0bNrBgwQJCQkLo378/CxYs4OjRo/zgBz9g9erVPP3004gIIkJhYSGZmZmICKWlX6WmLywsRETIzs4GoKWlhdtvv50RI0YQFBTE6NGjWbp0KTabDW9HrxFpNBrTEPDQQ7Bz57krOtiwARraBaGtrYXbb4fnn+9aG1Onwl/+0vVrAm+//TaxsbFceOH/t3fvwVHUWwLHvycwSQgPBUzCJVcTMUYDSABFVORVvmCjy7oiCmyurkBcQEpUtrQKFS5wgVgIuHdRRHIlyhWkvKirLO4tXkJcH7zXhSASSNAQdtGVEEggIfz2j+4ZJkOAPGbS0+P5VHUl8/t195yTmcmZ6f7Nr28iKyuLESNGMGfOHDweD7t372bw4MFkZWUxf/58YmJi2Lx5M2fPniUnJ4dDhw5x4403Mnv2bMCa27KoqOiy93nu3DmSkpJYtWoV8fHxfPPNN2RnZ9OxY0fGjBnToPjDjRYipZR7BRahy7UHSW5uLllZ1kUBBg4cSFxcHB9//DHDhw/nlVdeoWfPnixZssS3fnp6OmDNrBAdHU1cXBydOnVq0H16PB5mzJjhu52SksKOHTtYsWKFFiKllAqWMzk5RDdkCpyUFOtwXKDkZAjRLNcHDhwgPz+f9957D7BmHRg9ejS5ubkMHz6cnTt38uCDD4bkvhcvXszSpUspLi6msrKS6upqkpOTQ3JfzUkLkVLKvf7wB+ucUEXF+ba4OKs9RJYuXUpNTU2t+da801n+8MMPjdqnd8YC/6vWVFdX11rn/fffZ/LkycybN4877riDdu3asWjRIj788MNG3Wc40UKklHKv0aOtn1OnwuHDcM01VhHytgfZ2bNnycvLY86cOdx///21+rKysnj77bfp1asXGzZcfErL6Oho32zjXvHx8QCUlpb6ft+1a1etdfLz8+nbt2+tod+FhYVNyidcaCFSSrnb6NEhKzyB1qxZw08//cS4cePo2LH2rCmPPvooixcvZvXq1dx+++1kZ2czceJEYmNj2bJlC/feey/t27cnJSWFb775hqKiItq0aUOHDh1ITU3l6quvZvr06cydO5eioiJmzZpVa/9paWksW7aMtWvXkpqaysqVK/n8889p39793+fS4dtKKVVPubm5DB48+IIiBPDwww9TVFTEsWPHWLduHfv27eO2226jb9++rFy5Eo/HurLtlClTiI6OpmvXrsTHx3P48GE8Hg8rV67k4MGDZGRkMG3aNN+oOq8nn3ySESNGMGrUKPr06UNRURHPPfdcs+QdalL7StoKrCu0hmKq83AR6VeJ1PzcoaCgwDeazMut1+upr0jIr67HzUtEthtjbmnoPvUTkVJKKUdpIVJKKeUoLURKKaUcpYVIKaWUo8KuEIlIBxH5UEROiUixiIy6xLprReSk31IlIt/69ReJSKVf/1+bJwullFL1FY7fI1oEVAGJQE9gjYjsNsbsCVzRGDPU/7aIbAICv0n2gDFmXYhiVUop1URh9YlIRFoDDwEvGWNOGmPygX8DsuqxbQrQH3gnlDEqpZQKrrAqREAacNYYs9+vbTfQrR7b/g7YYowpCmj/s4gcE5G/ikhGkOJUSikVJOF2aK4NcCKgrQyozzfAfgfMCmgbDewABHga+A8RudEYczxwYxHJBrLBmvdpU4hm7g0HJ0+e1PxcLFLyu+KKKygvL6/VVlNTc0FbJImE/E6fPh38558xptkWYBNgLrLkA72AioBtngM+ucx+7wROAm0us94+rHNGl4wzLS3NRLKNGzc6HUJIaX7usHfv3gvaTpw4Ua9tc3KM2bDh0uts2GCtF07qm184q+tx8wK2mUbUhmY9NGeMGWSMkYssdwL7gZYicr3fZhnABQMVAjwGrDbGnLxcCFifjpRSLtanD4wYARs31t2/caPV36dPaO7/8ccf913q238JnDHbTUSEDz74wJH7DqtzRMaYU8BqYIaItBaRfsAw4N2LbSMirYARwLKA9mtEpJ+IRItIrIj8M3AV8EXIElBKNYvBg2HVqrqLkbcIrVplrRcqd999N6WlpbWW7t27N3g/VVVVIYjOXcKqENkmAK2A/wVWAOONPXRbRPqLSOCnnr8DjgOB743aAm8AvwAlwBBgqDHm5xDGrpRqJnUVo+YqQgAxMTF06tSp1tKyZUs2b95M3759iY2NJTExkWeeeaZWsRk0aBDjx49nypQpxMfH069fPwD27t1LZmYmbdu2JSEhgZEjR3L06NFa95mXl8dNN91ETEwMiYmJPPbYY76++fPn06NHD1q3bk1SUhJjx47l+PHzp8PLysrIysoiISGB2NhYunTpwsKFCwHrsuNgzSAuIr7bzSXcBitgjPk/rOJSV98WrAEN/m0rsApW4Lp7gB6hiFEpFRrPPx/D3r0N26ZzZ7jvPvjNb6C0FNLT4fe/t5b66NkT7P/HTVZSUsLQoUPJyspi2bJlFBYWMnbsWKKionj11Vd96y1fvpzs7Gy2bNmCMYbS0lIGDBjAmDFjmDdvHtXV1UydOpVhw4bx5ZdfEhUVxZtvvsnTTz/N7NmzyczM5OTJk7UuwBcVFcXChQvp0qULxcXFTJo0iUmTJvHuu9YBpRdffJFvv/2WTz/9lMTERA4dOsSxY8cA2Lp1KwkJCbz11lvcf//9tGjRIjh/kHoKu0KklFIN0b69VYS8F2htruvEffbZZ7Rpc/59cf/+/enduzedO3fm9ddfJyoqivT0dObOncuTTz7JzJkzfetee+21tQrTyy+/TEZGBjk5Ob62d955hw4dOrBt2zZuvfVWZs6cyeTJk3n22Wd969x8882+3ydPnuz7PSUlhVdeeYVhw4aRl5dHVFQUxcXF9O7dm1tvvRWA5ORk3/req8JeeeWVdOrUKRh/ngbRQqSUChs5OWdo2za6Qdt4D8e99BK88QZMmxb6w3IAAwYMYMmSJb7brVq1YtKkSdx2221ERZ0/63HnnXdSVVXFgQMHuPbaa4HaBQRg+/btbN68uVZh8yosLCQlJYWSkhLuuuuui8azYcMG5syZQ0FBAWVlZdTU1FBVVcXRo0fp3Lkz48ePZ/jw4Wzfvp177rmHBx54gIEDBzb1zxAUWoiUUq4VeE5o8ODmO0cUFxdHampqvdcXOT9gt3Xr1rX6zp07R2ZmJvPmzbtgu8TERCorKy+57+LiYjIzMxk3bhwzZsygY8eO7Nixg5EjR/rOTw0dOpTi4mLWrl3L+vXryczM5OGHH+btt9+udw6hEo6DFZRS6rLqGphwqdF0zSE9PZ2vvvqKc+fO+dry8/OJjo7muuuuu+h2vXv3Zs+ePSQnJ5Oamlpr8Q5eSEpKYv369XVuv23bNqqqqliwYAG33347aWlpHDly5IL1rrrqKt/5q9zcXPLy8jhz5gwAHo+HmpqaJv4FGkcLkVLKdS41Os7JYjRhwgSOHDnChAkTKCgoYM2aNbzwwgs89dRTxMXFXXS7iRMnUlZWxiOPPMLXX3/NwYMHWbduHdnZ2b6ZGKZOncrChQtZsGAB+/fvZ9euXb7zTNdffz3nzp1j4cKFHDp0iBUrVvhGxHm9/PLLfPTRR3z//fcUFBSwevVqunTpQkxMDGCdV1q/fj1Hjx7ll19+CdFfqG5aiJRSrrN166UPv3mL0datzRtXUlISa9euZefOnfTs2ZMnnniCkSNHMnv27Etu17lzZ7744guioqIYMmQI3bp1Y+LEicTExPgKxfjx41m0aBFvvfUW3bt3Z8iQIezZY33Xv0ePHrz22mvMnz+frl27snTp0gsO88XExDB16lQyMjLo168f5eXlfPLJJ77+V199lY0bN3L11VfTq1evIP9lLk2sWRmUvxtuuMF89913TocRMps2bWLQoEFOhxEymp87FBQUkJ6eXqutvLyctm3rM7WkO0VCfnU9bl4ist0Yc0tD96mfiJRSSjlKC5FSSilHaSFSSinlKC1ESimlHKWFSCnlGB0s5S6hery0ECmlHOHxeC47Y4AKL5WVlXg8nqDvVwuRUsoRCQkJlJSUUFFRoZ+MwpwxhoqKCkpKSkhISAj6/nWuOaWUI9q1awfAkSNHqK6uBuD06dPExsY6GVZIuTk/j8dDYmKi73ELJi1ESinHtGvXrtY/tk2bNjX7t/qbU6Tn11h6aE4ppZSjtBAppZRyVNgVIhF5SkS2icgZEVlWj/WfEZGjInJCRP4kIjF+fSkislFEKkRkn4jcHdLglVJKNVjYFSLgCDAL+NPlVhSR+4AXgLuAZKAL4H+l+hXATqAjMBX4QETigx2wUkqpxgu7QmSMWW2M+Qj4uR6rPwbkGmP2GGN+AWYCjwOISBrQG5hmjKk0xvwF+BZ4KDSRK6WUagy3j5rrBnzsd3s3kCgiHe2+g8aY8oD+bnXtSESygWz75hkR+e8QxBsurgJ+cjqIENL83CuSc4PIz++Gxmzk9kLUBijzu+39vW0dfd7+pLp2ZIxZAiwBEJFtjbmmhltofu4WyflFcm7w68ivMds166E5EdkkIuYiS34jdnkS8P92lff38jr6vP3lKKWUChvNWoiMMYOMMXKR5c5G7HIPkOF3OwP4H2PMz3ZfFxFpG9C/p/EZKKWUCrawG6wgIi1FJBZoAbQQkVgRudghxHeAMSLSVUSuBF4ElgEYY/YDu4Bp9j4eBHoAf6lHGEuamkeY0/zcLZLzi+TcQPOrk4TbZIMiMh2YFtD8e2PMdBG5BtgLdDXGHLbXfxZ4HmiFVWT+yRhzxu5LwSpMfYHDwERjzLrQZ6GUUqq+wq4QKaWU+nUJu0NzSimlfl20ECmllHKUFiI/ItJBRD4UkVMiUiwio5yOqSkuNW+fiNxlz79XYc/Hl+xQmI0iIjEikms/TuUisktEhvr1uzo/ABFZLiKl9jyK+0VkrF+f6/MDEJHrReS0iCz3axtlP66nROQjEengZIyNZX9d5bSInLSX7/z6XJ+jiDwqIgV2DoUi0t9ub/BzUwtRbYuAKiARGA28ISJ1zsTgEnXO2yciVwGrgZeADsA24P1mj65pWgI/AAOBK7BGTK6yJ7qNhPwA5gApxph2wN8Cs0Tk5gjKD6zX3FbvDfv19iaQhfU6rABedya0oHjKGNPGXm6AyMhRRO4BcoB/xJpAYABwsLHPTR2sYBOR1sAvQHd76Dci8i5QYox5wdHgmkhEZgG/NcY8bt/OBh43xtxh326NNe1IL2PMPscCbSIR+S+sSW87EmH5icgNwCbgaeBKIiA/EXkU+HuskbCpxph/EJHZWMV3lL3OdUAB0DFguq6wJyKbgOXGmKUB7a7PUUT+E2uez9yA9kb9b9FPROelAWe9Rch20bnpXK4bVm4AGGNOAYW4OFcRScR6DPcQQfmJyOsiUgHsA0qBfycC8hORdsAM4NmArsDcCrGOUqQ1X3RBNUdEfhKRL0RkkN3m6hxFpAVwCxAvIgdE5EcR+VcRaUUjn5taiM5rA5wIaCvD+tgZaS42D58rcxURD/BnIM9+1xUx+RljJmDF3R/rkMcZIiO/mVjvqH8MaI+E3Lyex7o0TRLWFz0/sT/9uD3HRMADDMd6XvYEemEdHm9UblqIzvs1zU0XMbmKSBTwLtY7yqfs5ojJD8AYU2OMyQd+C4zH5fmJSE/gbmBBHd2uzs2fMeZrY0y5MeaMMSYP+AL4G9yfY6X984/GmFJjzE/AfJqQmxai8/YDLUXker+2SJ2brtYcffZx3OtwWa4iIkAu1ju0h4wx1XZXRORXh5acz8PN+Q0CUoDDInIUmAI8JCI7uDC3LkAM1uvT7QwguDxH+9pvP2Ll42u2fzbuuWmM0cVegJVYV3VtDfTD+kjZzem4mpBPSyAWa/TVu/bvLYF4O7eH7LYc4Cun421EfouBr4A2Ae2uzw9IAB7FOtTRArgPOIU1es7V+QFxQCe/ZR7wgZ1XN6xD5P3t1+FyYKXTMTcixyvtx8z7mhttP35pkZAj1vm9rfbztD2wBetwa6Oem44nFE4L1nDDj+wnzGFglNMxNTGf6VjvVPyX6Xbf3VgnwCuxRmOlOB1vA3NLtvM5jXU4wLuMjpD84oHPgeP2P61vgXF+/a7OLyDX6Vijy7y3R9mvv1NYF77s4HSMjXz8tmIdkjqO9YbpnkjJEesc0et2bkeBfwFi7b4GPzd1+LZSSilH6TkipZRSjtJCpJRSylFaiJRSSjlKC5FSSilHaSFSSinlKC1ESimlHKWFSCmllKO0ECnlEiLSTkSmi0i607EoFUxaiJRyj1uAaVjfalcqYmghUso9emFdBmKv04EoFUw6xY9SLiAiBcCNAc2rjTEPORGPUsGkhUgpFxCRPlizw+8BZtvNpcaYYueiUio4WjodgFKqXnZjXRjvj8aYr5wORqlg0nNESrlDNyAa2OF0IEoFmxYipdyhN9b1l3Y5HYhSwaaFSCl36AUUGmNOOB2IUsGmhUgpd+iKDttWEUoHKyjlDseB3iJyH1AGfG+M+dnhmJQKCh2+rZQLiEh3IBfoAcQC/Y0x+c5GpVRwaCFSSinlKD1HpJRSylFaiJRSSjlKC5FSSilHaSFSSinlKC1ESimlHKWFSCmllKO0ECmllHKUFiKllFKO+n9Ke+s1X+UhtAAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NjmyXLq3ZPUo",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "67dfdcf8-cdef-4300-9dfe-721e58ddc093"
      },
      "source": [
        "model.evaluate(X_test, Y_test)\n",
        "#outputs [loss (mse), last_time_step_mse]"
      ],
      "execution_count": 53,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "32/32 [==============================] - 0s 3ms/step - loss: 0.0267 - last_time_step_mse: 0.0052\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[0.026704376563429832, 0.005172676872462034]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 53
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "X1L2zNHAtyxM"
      },
      "source": [
        ""
      ],
      "execution_count": 53,
      "outputs": []
    }
  ]
}