TF household_power EXAM EDITION(seq2seq: multi future tx).ipynb

tf-household_power-exam-edition-seq2seq-multi-future-tx.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "TF household_power EXAM EDITION(seq2seq: multi future tx).ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "machine_shape": "hm",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/gist/XinyueZ/b4cd1fbdea78274abb291f90b0c50ade/tf-household_power-exam-edition-seq2seq-multi-future-tx.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fWWjotKqmwwW"
      },
      "source": [
        "import urllib\n",
        "import os\n",
        "import zipfile\n",
        "import pandas as pd\n",
        "import tensorflow as tf"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6QPJo8NDm-7W"
      },
      "source": [
        "def download_and_extract_data():\n",
        "    url = 'https://dl.dropbox.com/s/lisgoq82kff6sx3/household_power.zip'\n",
        "    urllib.request.urlretrieve(url, 'household_power.zip')\n",
        "    with zipfile.ZipFile('household_power.zip', 'r') as zip_ref:\n",
        "        zip_ref.extractall()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ScNbqaTqnAQ4"
      },
      "source": [
        "def normalize_series(data, min, max):\n",
        "    data = data - min\n",
        "    data = data / max\n",
        "    return data"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yaeORAt8zbXc"
      },
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "def plot_series(time, series, format=\"-\", start=0, end=168*2):\n",
        "    plt.plot(time[start:end], series[start:end], format)\n",
        "    plt.xlabel(\"Time\")\n",
        "    plt.ylabel(\"Value\")\n",
        "    plt.grid(True) "
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "S34QNahhEedB"
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "\n",
        "\n",
        "def plot_graphs(history, string):\n",
        "  plt.plot(history.history[string])\n",
        "  plt.plot(history.history['val_'+string])\n",
        "  plt.xlabel(\"Epochs\")\n",
        "  plt.ylabel(string)\n",
        "  plt.legend([string, 'val_'+string])\n",
        "  plt.show()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "g0OWK10a4LPo"
      },
      "source": [
        "def process_in_duration(message,  proc):\n",
        "  import datetime\n",
        "  \n",
        "  tm_start = datetime.datetime.now()\n",
        "\n",
        "  result = proc()\n",
        "\n",
        "  tm_end = datetime.datetime.now()  \n",
        "  delta = tm_end - tm_start\n",
        " \n",
        "  print(f\"\\033[92m{message} used {delta}\")\n",
        "  return result"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vkutKobJzKcP"
      },
      "source": [
        "# Downloads and extracts the dataset to the directory that\n",
        "# contains this file.\n",
        "download_and_extract_data()\n",
        "# Reads the dataset from the csv.\n",
        "df = pd.read_csv('household_power_consumption.csv', sep=',',\n",
        "                 infer_datetime_format=True, \n",
        "                 index_col='datetime', \n",
        "                 header=0)\n",
        "\n",
        "# Number of features in the dataset. We use all features as predictors to\n",
        "# predict all features at future time steps.\n",
        "N_FEATURES = len(df.columns)\n",
        "\n",
        "# Normalizes the data\n",
        "data = df.values\n",
        "split_time = int(len(data) * 0.5)\n",
        "data = normalize_series(data, data.min(axis=0), data.max(axis=0))\n",
        "\n",
        "# Splits the data into training and validation sets.\n",
        "x_train = data[:split_time]\n",
        "x_valid = data[split_time:]"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "UIW38ssPzYk2",
        "outputId": "11534d98-8057-46bc-c83b-26c6929454c6"
      },
      "source": [
        "print(data.shape)\n",
        "\n",
        "time_steps = np.array([i for i in range(len(data))])\n",
        "\n",
        "for i in range(7):\n",
        "  plt.figure(figsize=(10, 6))\n",
        "  plot_series(time_steps, data[:,i]) "
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(86400, 7)\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RZ_DeciRS2Uy"
      },
      "source": [
        "# Architecture\n",
        "\n",
        "***Sequence -> Vector -> Sequence***"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nY7BgrJxVOxz"
      },
      "source": [
        "### Base line\n",
        "![image.png](data:image/png;base64,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)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bHBqtfbeSe0h"
      },
      "source": [
        "![image.png](data:image/png;base64,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)![image.png](data:image/png;base64,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)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_zl5uB-bnHHd"
      },
      "source": [
        "def windowed_dataset(series, batch_size, n_past=24, n_future=24, shift=1):\n",
        "    print(series.shape)\n",
        "    ds = tf.data.Dataset.from_tensor_slices(series)\n",
        "    ds = ds.window(size=(n_past + n_future),\n",
        "                   shift = shift,\n",
        "                   drop_remainder = True)\n",
        "    ds = ds.flat_map(lambda w: w.batch(n_past + n_future))\n",
        "    ds = ds.map(lambda w: (w[:n_past], w[-n_past:])) \n",
        "    return ds.batch(batch_size).prefetch(1)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ul4liK5jIx6o"
      },
      "source": [
        "def model_forecast(model, series, batch_size, window_size_forecast, shift=1):\n",
        "    ds = tf.data.Dataset.from_tensor_slices(series) \n",
        "    ds = ds.window(size=window_size_forecast,\n",
        "                   shift = shift,\n",
        "                   drop_remainder = True) \n",
        "    ds = ds.flat_map(lambda w: w.batch(window_size_forecast)) \n",
        "    ds = ds.batch(batch_size).prefetch(1)\n",
        "    forecast = model.predict(ds)\n",
        "    return forecast"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TkfOd1TanRGR",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "3e9d9a23-055f-49e4-dc51-7b0297cad90a"
      },
      "source": [
        "tf.keras.backend.clear_session()\n",
        "tf.random.set_seed(51)\n",
        "np.random.seed(51)\n",
        "\n",
        "BATCH_SIZE = 32  \n",
        "N_PAST = 24 \n",
        "N_FUTURE = 24  \n",
        "SHIFT = 1 \n",
        "\n",
        "train_set = windowed_dataset(series=x_train, batch_size=BATCH_SIZE,\n",
        "                  n_past=N_PAST, n_future=N_FUTURE,\n",
        "                  shift=SHIFT)\n",
        "valid_set = windowed_dataset(series=x_valid, batch_size=BATCH_SIZE,\n",
        "                  n_past=N_PAST, n_future=N_FUTURE,\n",
        "                  shift=SHIFT)\n",
        "\n",
        "print(train_set)\n",
        "print(valid_set)\n",
        " \n",
        "inputs = tf.keras.layers.Input(shape = [N_PAST, N_FEATURES])\n",
        "outputs = []\n",
        "rnn_units = 32\n",
        "\n",
        "rnn = tf.keras.layers.LSTM(rnn_units, return_state=True, return_sequences=False)\n",
        "rnn_cell = tf.keras.layers.LSTMCell(rnn_units)\n",
        "densor = tf.keras.layers.Dense(N_FEATURES)\n",
        "\n",
        "# First tx to predict, and the output of this tx will be used for (N_FEATURES-1) predictions.\n",
        "x, *state = rnn(inputs)\n",
        "x = densor(x)\n",
        "outputs.append(x)\n",
        "\n",
        "# Use first prediction to predict others.\n",
        "for Tx in range(1, N_FUTURE):\n",
        "  x, state = rnn_cell(x, states=state)\n",
        "  x = densor(x)\n",
        "  outputs.append(x)\n",
        "\n",
        "outputs = tf.stack(outputs)\n",
        "outputs = tf.transpose(outputs, [1, 0, 2])\n",
        "\n",
        "model = tf.keras.Model(inputs, outputs)\n",
        "\n",
        "base_lr = 1e-8\n",
        "\n",
        "lr_schedule = tf.keras.callbacks.LearningRateScheduler(\n",
        "    lambda epoch: base_lr * 10**(epoch / 20)\n",
        ")\n",
        "\n",
        "optimizer = tf.keras.optimizers.SGD(learning_rate=base_lr, momentum=0.9)\n",
        "\n",
        "model.compile(loss=tf.keras.losses.Huber(),\n",
        "  optimizer=optimizer,\n",
        "  metrics=[\"mae\", \"mse\"]\n",
        ")\n",
        "\n",
        "model.summary()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(43200, 7)\n",
            "(43200, 7)\n",
            "<PrefetchDataset shapes: ((None, None, 7), (None, None, 7)), types: (tf.float64, tf.float64)>\n",
            "<PrefetchDataset shapes: ((None, None, 7), (None, None, 7)), types: (tf.float64, tf.float64)>\n",
            "Model: \"model\"\n",
            "__________________________________________________________________________________________________\n",
            "Layer (type)                    Output Shape         Param #     Connected to                     \n",
            "==================================================================================================\n",
            "input_1 (InputLayer)            [(None, 24, 7)]      0                                            \n",
            "__________________________________________________________________________________________________\n",
            "lstm (LSTM)                     [(None, 32), (None,  5120        input_1[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "dense (Dense)                   (None, 7)            231         lstm[0][0]                       \n",
            "                                                                 lstm_cell_1[0][0]                \n",
            "                                                                 lstm_cell_1[1][0]                \n",
            "                                                                 lstm_cell_1[2][0]                \n",
            "                                                                 lstm_cell_1[3][0]                \n",
            "                                                                 lstm_cell_1[4][0]                \n",
            "                                                                 lstm_cell_1[5][0]                \n",
            "                                                                 lstm_cell_1[6][0]                \n",
            "                                                                 lstm_cell_1[7][0]                \n",
            "                                                                 lstm_cell_1[8][0]                \n",
            "                                                                 lstm_cell_1[9][0]                \n",
            "                                                                 lstm_cell_1[10][0]               \n",
            "                                                                 lstm_cell_1[11][0]               \n",
            "                                                                 lstm_cell_1[12][0]               \n",
            "                                                                 lstm_cell_1[13][0]               \n",
            "                                                                 lstm_cell_1[14][0]               \n",
            "                                                                 lstm_cell_1[15][0]               \n",
            "                                                                 lstm_cell_1[16][0]               \n",
            "                                                                 lstm_cell_1[17][0]               \n",
            "                                                                 lstm_cell_1[18][0]               \n",
            "                                                                 lstm_cell_1[19][0]               \n",
            "                                                                 lstm_cell_1[20][0]               \n",
            "                                                                 lstm_cell_1[21][0]               \n",
            "                                                                 lstm_cell_1[22][0]               \n",
            "__________________________________________________________________________________________________\n",
            "lstm_cell_1 (LSTMCell)          ((None, 32), [(None, 5120        dense[0][0]                      \n",
            "                                                                 lstm[0][1]                       \n",
            "                                                                 lstm[0][2]                       \n",
            "                                                                 dense[1][0]                      \n",
            "                                                                 lstm_cell_1[0][1]                \n",
            "                                                                 lstm_cell_1[0][2]                \n",
            "                                                                 dense[2][0]                      \n",
            "                                                                 lstm_cell_1[1][1]                \n",
            "                                                                 lstm_cell_1[1][2]                \n",
            "                                                                 dense[3][0]                      \n",
            "                                                                 lstm_cell_1[2][1]                \n",
            "                                                                 lstm_cell_1[2][2]                \n",
            "                                                                 dense[4][0]                      \n",
            "                                                                 lstm_cell_1[3][1]                \n",
            "                                                                 lstm_cell_1[3][2]                \n",
            "                                                                 dense[5][0]                      \n",
            "                                                                 lstm_cell_1[4][1]                \n",
            "                                                                 lstm_cell_1[4][2]                \n",
            "                                                                 dense[6][0]                      \n",
            "                                                                 lstm_cell_1[5][1]                \n",
            "                                                                 lstm_cell_1[5][2]                \n",
            "                                                                 dense[7][0]                      \n",
            "                                                                 lstm_cell_1[6][1]                \n",
            "                                                                 lstm_cell_1[6][2]                \n",
            "                                                                 dense[8][0]                      \n",
            "                                                                 lstm_cell_1[7][1]                \n",
            "                                                                 lstm_cell_1[7][2]                \n",
            "                                                                 dense[9][0]                      \n",
            "                                                                 lstm_cell_1[8][1]                \n",
            "                                                                 lstm_cell_1[8][2]                \n",
            "                                                                 dense[10][0]                     \n",
            "                                                                 lstm_cell_1[9][1]                \n",
            "                                                                 lstm_cell_1[9][2]                \n",
            "                                                                 dense[11][0]                     \n",
            "                                                                 lstm_cell_1[10][1]               \n",
            "                                                                 lstm_cell_1[10][2]               \n",
            "                                                                 dense[12][0]                     \n",
            "                                                                 lstm_cell_1[11][1]               \n",
            "                                                                 lstm_cell_1[11][2]               \n",
            "                                                                 dense[13][0]                     \n",
            "                                                                 lstm_cell_1[12][1]               \n",
            "                                                                 lstm_cell_1[12][2]               \n",
            "                                                                 dense[14][0]                     \n",
            "                                                                 lstm_cell_1[13][1]               \n",
            "                                                                 lstm_cell_1[13][2]               \n",
            "                                                                 dense[15][0]                     \n",
            "                                                                 lstm_cell_1[14][1]               \n",
            "                                                                 lstm_cell_1[14][2]               \n",
            "                                                                 dense[16][0]                     \n",
            "                                                                 lstm_cell_1[15][1]               \n",
            "                                                                 lstm_cell_1[15][2]               \n",
            "                                                                 dense[17][0]                     \n",
            "                                                                 lstm_cell_1[16][1]               \n",
            "                                                                 lstm_cell_1[16][2]               \n",
            "                                                                 dense[18][0]                     \n",
            "                                                                 lstm_cell_1[17][1]               \n",
            "                                                                 lstm_cell_1[17][2]               \n",
            "                                                                 dense[19][0]                     \n",
            "                                                                 lstm_cell_1[18][1]               \n",
            "                                                                 lstm_cell_1[18][2]               \n",
            "                                                                 dense[20][0]                     \n",
            "                                                                 lstm_cell_1[19][1]               \n",
            "                                                                 lstm_cell_1[19][2]               \n",
            "                                                                 dense[21][0]                     \n",
            "                                                                 lstm_cell_1[20][1]               \n",
            "                                                                 lstm_cell_1[20][2]               \n",
            "                                                                 dense[22][0]                     \n",
            "                                                                 lstm_cell_1[21][1]               \n",
            "                                                                 lstm_cell_1[21][2]               \n",
            "__________________________________________________________________________________________________\n",
            "tf.stack (TFOpLambda)           (24, None, 7)        0           dense[0][0]                      \n",
            "                                                                 dense[1][0]                      \n",
            "                                                                 dense[2][0]                      \n",
            "                                                                 dense[3][0]                      \n",
            "                                                                 dense[4][0]                      \n",
            "                                                                 dense[5][0]                      \n",
            "                                                                 dense[6][0]                      \n",
            "                                                                 dense[7][0]                      \n",
            "                                                                 dense[8][0]                      \n",
            "                                                                 dense[9][0]                      \n",
            "                                                                 dense[10][0]                     \n",
            "                                                                 dense[11][0]                     \n",
            "                                                                 dense[12][0]                     \n",
            "                                                                 dense[13][0]                     \n",
            "                                                                 dense[14][0]                     \n",
            "                                                                 dense[15][0]                     \n",
            "                                                                 dense[16][0]                     \n",
            "                                                                 dense[17][0]                     \n",
            "                                                                 dense[18][0]                     \n",
            "                                                                 dense[19][0]                     \n",
            "                                                                 dense[20][0]                     \n",
            "                                                                 dense[21][0]                     \n",
            "                                                                 dense[22][0]                     \n",
            "                                                                 dense[23][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "tf.compat.v1.transpose (TFOpLam (None, 24, 7)        0           tf.stack[0][0]                   \n",
            "==================================================================================================\n",
            "Total params: 10,471\n",
            "Trainable params: 10,471\n",
            "Non-trainable params: 0\n",
            "__________________________________________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ALU40CmP2fhc",
        "outputId": "d7642348-8a01-49bc-86ff-1d5688bfad33"
      },
      "source": [
        "lr_history = process_in_duration(\n",
        "    \"Training for learning-rate end,\", \n",
        "    lambda:\n",
        "          model.fit(train_set, \n",
        "                    validation_data=valid_set,\n",
        "                    epochs=100, callbacks=[lr_schedule]))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "1349/1349 [==============================] - 28s 15ms/step - loss: 0.0346 - mae: 0.1400 - mse: 0.0692 - val_loss: 0.0349 - val_mae: 0.1365 - val_mse: 0.0698\n",
            "Epoch 2/100\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0346 - mae: 0.1399 - mse: 0.0692 - val_loss: 0.0349 - val_mae: 0.1365 - val_mse: 0.0698\n",
            "Epoch 3/100\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0346 - mae: 0.1399 - mse: 0.0692 - val_loss: 0.0349 - val_mae: 0.1365 - val_mse: 0.0698\n",
            "Epoch 4/100\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0346 - mae: 0.1399 - mse: 0.0692 - val_loss: 0.0349 - val_mae: 0.1365 - val_mse: 0.0698\n",
            "Epoch 5/100\n",
            "1349/1349 [==============================] - 20s 14ms/step - loss: 0.0346 - mae: 0.1399 - mse: 0.0692 - val_loss: 0.0349 - val_mae: 0.1365 - val_mse: 0.0698\n",
            "Epoch 6/100\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0346 - mae: 0.1399 - mse: 0.0692 - val_loss: 0.0349 - val_mae: 0.1365 - val_mse: 0.0698\n",
            "Epoch 7/100\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0346 - mae: 0.1399 - mse: 0.0692 - val_loss: 0.0349 - val_mae: 0.1365 - val_mse: 0.0698\n",
            "Epoch 8/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0346 - mae: 0.1399 - mse: 0.0692 - val_loss: 0.0349 - val_mae: 0.1365 - val_mse: 0.0698\n",
            "Epoch 9/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0346 - mae: 0.1399 - mse: 0.0692 - val_loss: 0.0349 - val_mae: 0.1365 - val_mse: 0.0698\n",
            "Epoch 10/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0346 - mae: 0.1399 - mse: 0.0692 - val_loss: 0.0349 - val_mae: 0.1364 - val_mse: 0.0698\n",
            "Epoch 11/100\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0346 - mae: 0.1399 - mse: 0.0692 - val_loss: 0.0349 - val_mae: 0.1364 - val_mse: 0.0698\n",
            "Epoch 12/100\n",
            "1349/1349 [==============================] - 20s 14ms/step - loss: 0.0346 - mae: 0.1399 - mse: 0.0692 - val_loss: 0.0349 - val_mae: 0.1364 - val_mse: 0.0698\n",
            "Epoch 13/100\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0346 - mae: 0.1398 - mse: 0.0691 - val_loss: 0.0349 - val_mae: 0.1364 - val_mse: 0.0698\n",
            "Epoch 14/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0346 - mae: 0.1398 - mse: 0.0691 - val_loss: 0.0349 - val_mae: 0.1364 - val_mse: 0.0698\n",
            "Epoch 15/100\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0346 - mae: 0.1398 - mse: 0.0691 - val_loss: 0.0349 - val_mae: 0.1364 - val_mse: 0.0698\n",
            "Epoch 16/100\n",
            "1349/1349 [==============================] - 20s 14ms/step - loss: 0.0346 - mae: 0.1398 - mse: 0.0691 - val_loss: 0.0349 - val_mae: 0.1363 - val_mse: 0.0698\n",
            "Epoch 17/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0346 - mae: 0.1398 - mse: 0.0691 - val_loss: 0.0349 - val_mae: 0.1363 - val_mse: 0.0697\n",
            "Epoch 18/100\n",
            "1349/1349 [==============================] - 20s 14ms/step - loss: 0.0345 - mae: 0.1397 - mse: 0.0691 - val_loss: 0.0349 - val_mae: 0.1363 - val_mse: 0.0697\n",
            "Epoch 19/100\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0345 - mae: 0.1397 - mse: 0.0691 - val_loss: 0.0349 - val_mae: 0.1362 - val_mse: 0.0697\n",
            "Epoch 20/100\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0345 - mae: 0.1397 - mse: 0.0691 - val_loss: 0.0349 - val_mae: 0.1362 - val_mse: 0.0697\n",
            "Epoch 21/100\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0345 - mae: 0.1396 - mse: 0.0690 - val_loss: 0.0348 - val_mae: 0.1362 - val_mse: 0.0697\n",
            "Epoch 22/100\n",
            "1349/1349 [==============================] - 20s 14ms/step - loss: 0.0345 - mae: 0.1396 - mse: 0.0690 - val_loss: 0.0348 - val_mae: 0.1361 - val_mse: 0.0697\n",
            "Epoch 23/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0345 - mae: 0.1395 - mse: 0.0690 - val_loss: 0.0348 - val_mae: 0.1361 - val_mse: 0.0696\n",
            "Epoch 24/100\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0345 - mae: 0.1395 - mse: 0.0690 - val_loss: 0.0348 - val_mae: 0.1360 - val_mse: 0.0696\n",
            "Epoch 25/100\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0345 - mae: 0.1394 - mse: 0.0689 - val_loss: 0.0348 - val_mae: 0.1360 - val_mse: 0.0696\n",
            "Epoch 26/100\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0345 - mae: 0.1393 - mse: 0.0689 - val_loss: 0.0348 - val_mae: 0.1359 - val_mse: 0.0695\n",
            "Epoch 27/100\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0344 - mae: 0.1393 - mse: 0.0689 - val_loss: 0.0348 - val_mae: 0.1358 - val_mse: 0.0695\n",
            "Epoch 28/100\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0344 - mae: 0.1392 - mse: 0.0688 - val_loss: 0.0347 - val_mae: 0.1357 - val_mse: 0.0695\n",
            "Epoch 29/100\n",
            "1349/1349 [==============================] - 20s 14ms/step - loss: 0.0344 - mae: 0.1391 - mse: 0.0688 - val_loss: 0.0347 - val_mae: 0.1356 - val_mse: 0.0694\n",
            "Epoch 30/100\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0344 - mae: 0.1390 - mse: 0.0687 - val_loss: 0.0347 - val_mae: 0.1355 - val_mse: 0.0694\n",
            "Epoch 31/100\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0343 - mae: 0.1389 - mse: 0.0687 - val_loss: 0.0346 - val_mae: 0.1354 - val_mse: 0.0693\n",
            "Epoch 32/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0343 - mae: 0.1388 - mse: 0.0686 - val_loss: 0.0346 - val_mae: 0.1353 - val_mse: 0.0692\n",
            "Epoch 33/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0343 - mae: 0.1386 - mse: 0.0685 - val_loss: 0.0346 - val_mae: 0.1352 - val_mse: 0.0691\n",
            "Epoch 34/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0342 - mae: 0.1385 - mse: 0.0684 - val_loss: 0.0345 - val_mae: 0.1350 - val_mse: 0.0691\n",
            "Epoch 35/100\n",
            "1349/1349 [==============================] - 20s 14ms/step - loss: 0.0342 - mae: 0.1383 - mse: 0.0683 - val_loss: 0.0345 - val_mae: 0.1348 - val_mse: 0.0690\n",
            "Epoch 36/100\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0341 - mae: 0.1381 - mse: 0.0682 - val_loss: 0.0344 - val_mae: 0.1347 - val_mse: 0.0689\n",
            "Epoch 37/100\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0341 - mae: 0.1379 - mse: 0.0681 - val_loss: 0.0344 - val_mae: 0.1345 - val_mse: 0.0687\n",
            "Epoch 38/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0340 - mae: 0.1377 - mse: 0.0680 - val_loss: 0.0343 - val_mae: 0.1343 - val_mse: 0.0686\n",
            "Epoch 39/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0339 - mae: 0.1375 - mse: 0.0678 - val_loss: 0.0342 - val_mae: 0.1340 - val_mse: 0.0684\n",
            "Epoch 40/100\n",
            "1349/1349 [==============================] - 20s 14ms/step - loss: 0.0338 - mae: 0.1372 - mse: 0.0677 - val_loss: 0.0341 - val_mae: 0.1338 - val_mse: 0.0683\n",
            "Epoch 41/100\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0337 - mae: 0.1369 - mse: 0.0675 - val_loss: 0.0340 - val_mae: 0.1335 - val_mse: 0.0681\n",
            "Epoch 42/100\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0336 - mae: 0.1366 - mse: 0.0673 - val_loss: 0.0339 - val_mae: 0.1332 - val_mse: 0.0679\n",
            "Epoch 43/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0335 - mae: 0.1363 - mse: 0.0671 - val_loss: 0.0338 - val_mae: 0.1329 - val_mse: 0.0676\n",
            "Epoch 44/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0334 - mae: 0.1359 - mse: 0.0668 - val_loss: 0.0337 - val_mae: 0.1326 - val_mse: 0.0674\n",
            "Epoch 45/100\n",
            "1349/1349 [==============================] - 20s 14ms/step - loss: 0.0333 - mae: 0.1355 - mse: 0.0665 - val_loss: 0.0336 - val_mae: 0.1322 - val_mse: 0.0671\n",
            "Epoch 46/100\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0331 - mae: 0.1351 - mse: 0.0662 - val_loss: 0.0334 - val_mae: 0.1318 - val_mse: 0.0668\n",
            "Epoch 47/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0329 - mae: 0.1347 - mse: 0.0659 - val_loss: 0.0332 - val_mae: 0.1314 - val_mse: 0.0664\n",
            "Epoch 48/100\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0327 - mae: 0.1342 - mse: 0.0655 - val_loss: 0.0330 - val_mae: 0.1309 - val_mse: 0.0661\n",
            "Epoch 49/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0325 - mae: 0.1337 - mse: 0.0651 - val_loss: 0.0328 - val_mae: 0.1305 - val_mse: 0.0656\n",
            "Epoch 50/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0323 - mae: 0.1331 - mse: 0.0646 - val_loss: 0.0326 - val_mae: 0.1300 - val_mse: 0.0652\n",
            "Epoch 51/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0321 - mae: 0.1326 - mse: 0.0641 - val_loss: 0.0323 - val_mae: 0.1294 - val_mse: 0.0647\n",
            "Epoch 52/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0318 - mae: 0.1320 - mse: 0.0636 - val_loss: 0.0320 - val_mae: 0.1289 - val_mse: 0.0641\n",
            "Epoch 53/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0315 - mae: 0.1313 - mse: 0.0630 - val_loss: 0.0317 - val_mae: 0.1283 - val_mse: 0.0635\n",
            "Epoch 54/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0312 - mae: 0.1307 - mse: 0.0623 - val_loss: 0.0314 - val_mae: 0.1277 - val_mse: 0.0628\n",
            "Epoch 55/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0308 - mae: 0.1300 - mse: 0.0616 - val_loss: 0.0311 - val_mae: 0.1271 - val_mse: 0.0621\n",
            "Epoch 56/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0304 - mae: 0.1292 - mse: 0.0609 - val_loss: 0.0307 - val_mae: 0.1264 - val_mse: 0.0613\n",
            "Epoch 57/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0300 - mae: 0.1285 - mse: 0.0600 - val_loss: 0.0302 - val_mae: 0.1258 - val_mse: 0.0605\n",
            "Epoch 58/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0296 - mae: 0.1277 - mse: 0.0591 - val_loss: 0.0298 - val_mae: 0.1251 - val_mse: 0.0596\n",
            "Epoch 59/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0291 - mae: 0.1269 - mse: 0.0582 - val_loss: 0.0293 - val_mae: 0.1243 - val_mse: 0.0586\n",
            "Epoch 60/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0286 - mae: 0.1261 - mse: 0.0572 - val_loss: 0.0288 - val_mae: 0.1236 - val_mse: 0.0576\n",
            "Epoch 61/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0281 - mae: 0.1253 - mse: 0.0561 - val_loss: 0.0283 - val_mae: 0.1229 - val_mse: 0.0566\n",
            "Epoch 62/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0275 - mae: 0.1245 - mse: 0.0550 - val_loss: 0.0277 - val_mae: 0.1221 - val_mse: 0.0554\n",
            "Epoch 63/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0269 - mae: 0.1237 - mse: 0.0539 - val_loss: 0.0271 - val_mae: 0.1214 - val_mse: 0.0543\n",
            "Epoch 64/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0263 - mae: 0.1230 - mse: 0.0527 - val_loss: 0.0265 - val_mae: 0.1207 - val_mse: 0.0530\n",
            "Epoch 65/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0257 - mae: 0.1223 - mse: 0.0514 - val_loss: 0.0259 - val_mae: 0.1201 - val_mse: 0.0518\n",
            "Epoch 66/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0251 - mae: 0.1216 - mse: 0.0501 - val_loss: 0.0252 - val_mae: 0.1195 - val_mse: 0.0505\n",
            "Epoch 67/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0244 - mae: 0.1209 - mse: 0.0488 - val_loss: 0.0246 - val_mae: 0.1190 - val_mse: 0.0491\n",
            "Epoch 68/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0238 - mae: 0.1204 - mse: 0.0476 - val_loss: 0.0239 - val_mae: 0.1185 - val_mse: 0.0478\n",
            "Epoch 69/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0231 - mae: 0.1198 - mse: 0.0463 - val_loss: 0.0233 - val_mae: 0.1180 - val_mse: 0.0465\n",
            "Epoch 70/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0225 - mae: 0.1193 - mse: 0.0450 - val_loss: 0.0226 - val_mae: 0.1176 - val_mse: 0.0452\n",
            "Epoch 71/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0219 - mae: 0.1189 - mse: 0.0438 - val_loss: 0.0220 - val_mae: 0.1172 - val_mse: 0.0440\n",
            "Epoch 72/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0213 - mae: 0.1185 - mse: 0.0426 - val_loss: 0.0214 - val_mae: 0.1169 - val_mse: 0.0429\n",
            "Epoch 73/100\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0208 - mae: 0.1181 - mse: 0.0416 - val_loss: 0.0209 - val_mae: 0.1167 - val_mse: 0.0418\n",
            "Epoch 74/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0203 - mae: 0.1179 - mse: 0.0406 - val_loss: 0.0204 - val_mae: 0.1166 - val_mse: 0.0408\n",
            "Epoch 75/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0199 - mae: 0.1178 - mse: 0.0398 - val_loss: 0.0200 - val_mae: 0.1167 - val_mse: 0.0399\n",
            "Epoch 76/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0195 - mae: 0.1178 - mse: 0.0391 - val_loss: 0.0196 - val_mae: 0.1168 - val_mse: 0.0392\n",
            "Epoch 77/100\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0192 - mae: 0.1178 - mse: 0.0384 - val_loss: 0.0193 - val_mae: 0.1169 - val_mse: 0.0386\n",
            "Epoch 78/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0190 - mae: 0.1179 - mse: 0.0380 - val_loss: 0.0190 - val_mae: 0.1172 - val_mse: 0.0381\n",
            "Epoch 79/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0188 - mae: 0.1181 - mse: 0.0376 - val_loss: 0.0188 - val_mae: 0.1175 - val_mse: 0.0377\n",
            "Epoch 80/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0186 - mae: 0.1184 - mse: 0.0373 - val_loss: 0.0187 - val_mae: 0.1179 - val_mse: 0.0373\n",
            "Epoch 81/100\n",
            "1349/1349 [==============================] - 20s 14ms/step - loss: 0.0185 - mae: 0.1188 - mse: 0.0370 - val_loss: 0.0185 - val_mae: 0.1184 - val_mse: 0.0371\n",
            "Epoch 82/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0184 - mae: 0.1192 - mse: 0.0368 - val_loss: 0.0184 - val_mae: 0.1188 - val_mse: 0.0368\n",
            "Epoch 83/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0183 - mae: 0.1196 - mse: 0.0366 - val_loss: 0.0183 - val_mae: 0.1190 - val_mse: 0.0366\n",
            "Epoch 84/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0182 - mae: 0.1197 - mse: 0.0364 - val_loss: 0.0182 - val_mae: 0.1191 - val_mse: 0.0364\n",
            "Epoch 85/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0181 - mae: 0.1198 - mse: 0.0362 - val_loss: 0.0181 - val_mae: 0.1191 - val_mse: 0.0362\n",
            "Epoch 86/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0180 - mae: 0.1197 - mse: 0.0360 - val_loss: 0.0180 - val_mae: 0.1189 - val_mse: 0.0359\n",
            "Epoch 87/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0179 - mae: 0.1194 - mse: 0.0358 - val_loss: 0.0178 - val_mae: 0.1186 - val_mse: 0.0357\n",
            "Epoch 88/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0178 - mae: 0.1191 - mse: 0.0355 - val_loss: 0.0177 - val_mae: 0.1181 - val_mse: 0.0354\n",
            "Epoch 89/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0176 - mae: 0.1186 - mse: 0.0352 - val_loss: 0.0175 - val_mae: 0.1176 - val_mse: 0.0350\n",
            "Epoch 90/100\n",
            "1349/1349 [==============================] - 20s 14ms/step - loss: 0.0174 - mae: 0.1181 - mse: 0.0349 - val_loss: 0.0173 - val_mae: 0.1169 - val_mse: 0.0346\n",
            "Epoch 91/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0172 - mae: 0.1174 - mse: 0.0344 - val_loss: 0.0170 - val_mae: 0.1161 - val_mse: 0.0341\n",
            "Epoch 92/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0169 - mae: 0.1165 - mse: 0.0339 - val_loss: 0.0167 - val_mae: 0.1151 - val_mse: 0.0334\n",
            "Epoch 93/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0166 - mae: 0.1154 - mse: 0.0331 - val_loss: 0.0162 - val_mae: 0.1137 - val_mse: 0.0325\n",
            "Epoch 94/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0160 - mae: 0.1138 - mse: 0.0320 - val_loss: 0.0155 - val_mae: 0.1116 - val_mse: 0.0310\n",
            "Epoch 95/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0150 - mae: 0.1112 - mse: 0.0300 - val_loss: 0.0141 - val_mae: 0.1085 - val_mse: 0.0282\n",
            "Epoch 96/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0130 - mae: 0.1065 - mse: 0.0261 - val_loss: 0.0114 - val_mae: 0.1024 - val_mse: 0.0227\n",
            "Epoch 97/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0105 - mae: 0.0964 - mse: 0.0210 - val_loss: 0.0095 - val_mae: 0.0902 - val_mse: 0.0189\n",
            "Epoch 98/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0094 - mae: 0.0853 - mse: 0.0189 - val_loss: 0.0088 - val_mae: 0.0813 - val_mse: 0.0176\n",
            "Epoch 99/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0090 - mae: 0.0790 - mse: 0.0180 - val_loss: 0.0085 - val_mae: 0.0771 - val_mse: 0.0170\n",
            "Epoch 100/100\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0088 - mae: 0.0761 - mse: 0.0176 - val_loss: 0.0083 - val_mae: 0.0752 - val_mse: 0.0166\n",
            "\u001b[92mTraining for learning-rate end, used 0:32:33.380458\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vlPb2fYWo5RL",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 288
        },
        "outputId": "2d70ef02-27d7-4334-f3f0-4f981df94215"
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "plt.semilogx(lr_history.history[\"lr\"], lr_history.history[\"loss\"])\n",
        "plt.axis([1e-4, 10e-2, 0, .02])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(0.0001, 0.1, 0.0, 0.02)"
            ]
          },
          "metadata": {},
          "execution_count": 13
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZMAAAEACAYAAAB27puMAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3de3hV9Z3v8fcnCQEJECAECgkIhVhFLl4ittqrjhY7VWxrW2xPi6dOqa2cnrYzcwbPnPbp48w81d6ccbR1qNqinYoOvZiOWmq99DLeCOpwFQ2IJYCAiIAol8D3/LEXuI07yYaVsLOTz+t59sNev/Vbv/1dLNkf12WvpYjAzMwsjZJCF2BmZsXPYWJmZqk5TMzMLDWHiZmZpeYwMTOz1BwmZmaWWl5hImm6pNWSmiTNzTG/r6Q7k/mPSxqbtJ8naYmkZcmf52Qtc3rS3iTpeklK2odKul/Sc8mfQzpnVc3MrKt0GCaSSoEbgQuAicClkia26nY5sD0iJgDXAdcm7S8BF0bEZGAWcHvWMj8EPg/UJa/pSftc4IGIqAMeSKbNzKwby2fPZBrQFBFrI2IfsACY0arPDGB+8n4hcK4kRcRTEbExaV8BHJfsxYwEBkXEY5H51eRtwMU5xpqf1W5mZt1UPmFSA6zPmm5O2nL2iYgWYAdQ1arPx4AnI2Jv0r+5jTFHRMSm5P2LwIg8ajQzswIqOxYfIulkMoe+zj+S5SIiJOW834uk2cBsgIqKitNPPPHE1HWamfUmS5YseSkiqjtjrHzCZAMwOmu6NmnL1adZUhlQCWwDkFQL/BL4bESsyepf28aYmyWNjIhNyeGwLbmKioh5wDyA+vr6aGxszGNVzMzsEEkvdNZY+RzmWgzUSRonqRyYCTS06tNA5gQ7wCXAg8lexWDgHmBuRPzXoc7JYaydkt6ZXMX1WeDuHGPNymo3M7NuqsMwSc6BzAEWAauAuyJihaSrJV2UdLsFqJLUBHyNN67AmgNMAL4h6enkNTyZ9yXgZqAJWAPcl7RfA5wn6TngL5JpMzPrxtQTbkHvw1xmZkdO0pKIqO+MsfwLeDMzS81hYmZmqTlMzMwsNYeJmZml5jAxM7PUHCZmZpaaw8TMzFJzmJiZWWoOEzMzS81hYmZmqR2TW9B3tWde3MUXf7qEybWVTK0dzKSaSiqP61PosszMeo0eESb9y0tZuWkn9y1/8XDbuGEVTBw1iIkjM6+TRg5ixKC+JI+aNzOzTtQjwmTM0P78/m8/wCuv7WPZhh0sbd7B0uZXWNr8Cvcs3XS439CKck4aOZCT3jaIE0cO4qSRA5kwfAB9y0oLWL2ZWfHrEWFyyOD+5bynrpr31L3x4LCde/bzzKZdrNy4g1WbdrFy005uf+wF9rYcBKCsRIyvHvDGXsyozF7M0IryQq2GmVnR6VFhksugfn2YNm4o08YNPdzWcuAg67a9xqpNO3nmxZ2s2rSLx9Zu45dPvfEAyZGV/Zg4chCTaiqZOrqSSTWVDB/YrxCrYGbW7fX4MMmlrLSECcMHMGH4AC6cOupw+8u797Fq005WbtzJyk07Wb5hBw+u3sKhR76MrOzH5JpKpo4ezGljhjB1dCX9y3vlX6GZ2Zv4mzDL0Ipyzp4wjLMnDDvctntvCys27mRp8yuHz8f8duVmAEpLxMSRgzhtzGBOO34I73x7FSMGee/FzHofP2nxKGzfvY+n1m9nyQvbefKFV3h6/Su8vv8AAG+vruCs8VW86+3DeOfbh1I1oO8xq8vM7Eh05pMW8woTSdOBfwFKgZsj4ppW8/sCtwGnA9uAT0bEOklVwELgDOAnETEn6T8Q+GPWELXATyPiK5IuA74DHDqBcUNE3NxefYV+bG/LgYOs3LSTx9Zu49E123ji+ZfZvS8TLieNHMQ5J1ZzzokjOGX0YEpLfGmymXUPxzRMJJUCzwLnAc3AYuDSiFiZ1edLwJSIuELSTOAjEfFJSRXAqcAkYNKhMMm1QsBXI+IPSZjUt9U3l0KHSWv7Dxxk2YYdPLpmG394diuNL2znwMFgaEU57z+hmnNOGs773zGcAX19lNHMCqczwySfb7NpQFNErE0+fAEwA1iZ1WcG8M3k/ULgBkmKiN3AnyRNaGtwSScAw3nznkpR61NawmljhnDamCFc+YEJ7HhtP394bisPPrOFB1dv4RdPbaBfnxLOPWkEF00dxfvfUe3fuphZUcsnTGqA9VnTzcCZbfWJiBZJO4Aq4KU8xp8J3Blv3kX6mKT3ktkj+mpErM+9aHGo7N+HC6eO4sKpozhwMHjyz9v59X9v5J6lm7hn6SYG9ivjgklv4+JTanjX+Cr/St/Mik53OM4yE/hM1vSvgTsiYq+kLwDzgXNaLyRpNjAbYMyYMceizk5RWiLOGDuUM8YO5Rsfnsh/rdnG3U9v4N5lL3JXYzMThg9g1ruO56On1VLhw2BmViTy+bbaAIzOmq7ljZPjrfs0SyoDKsmciG+XpKlAWUQsOdQWEdnL3Qx8O9eyETEPmAeZcyYdr0b3U1ZawvtOqOZ9J1SzZ/8B7l22iZ88so6v372Cb/9mNR+vH81n33U8Y4dVFLpUM7N25RMmi4E6SePIhMZM4FOt+jQAs4BHgUuAByO/a44vBe7IbpA0MiIO3VDrImBVHuMUvX59SvnoabV85NQanlr/CvMfWcdtj67jx488zwWT3sbffvBExjlUzKybyvfS4A8B/0zm0uBbI+KfJF0NNEZEg6R+wO1krtx6GZiZdcJ+HTAIKAdeAc4/dCWYpLXAhyLimazP+haZEGlJxvpi9vxcutvVXJ1ly8493P7YC9zyp+fZ13KQT505hi+fW8cw/3bFzDrBMf+dSXfXU8PkkK279vIvDzzLHU+sp19ZCV9433j+6j3jfCsXM0ulM8PET1osAtUD+/KPF0/mt199L++pq+b79z/L+7/zMA+v3lLo0szMAIdJURlfPYCbPnM6P//iWQytKOeyHy/mW/etYv+Bg4Uuzcx6OYdJETr9+CH86sqz+fSZY/i336/lE//2KOtffq3QZZlZL+YwKVL9+pTyTx+ZzI2fOo2mza/yl9f/kd8s39TxgmZmXcBhUuT+cspI7vnyexg7rIIrfvok37pvFT3hogozKy4Okx5gTFV/Fl5x1uHDXv/6YFOhSzKzXsbXlvYQ5WUl/MOMSby+/wDfv/9Zhg3oy6fOLJ7bzJhZcXOY9CAlJeLaj01h++59/L9fLWNoRR+mTxpZ6LLMrBfwYa4epk9pCTd++jSmjh7Mlxc8zWNrO7xFmplZag6THqh/eRm3zjqDMUP78/n5jazcuLPQJZlZD+cw6aGGVJRz2+emMaBfGbN+/AQbX3m90CWZWQ/mMOnBRg0+jts+N41de/bz7d+0e69MM7NUHCY9XN2IgVx21jju/u+NPPOiD3eZWddwmPQCX3zfeAb0LeO7i1YXuhQz66EcJr1AZf8+XPG+8fxu1RaWvLC90OWYWQ/kMOkl/ufZYxk2oJzvLHrGt1sxs07nMOkl+peXMecDE3hs7cv88bmXCl2OmfUwDpNe5NIzx1Az+Di+s2i1907MrFM5THqRvmWlfPW8E1i2YQf3LX+x0OWYWQ+SV5hImi5ptaQmSXNzzO8r6c5k/uOSxibtVZIekvSqpBtaLfNwMubTyWt4e2NZ5/jIqTXUDR/A9367mhY/odHMOkmHYSKpFLgRuACYCFwqaWKrbpcD2yNiAnAdcG3Svgf4OvA3bQz/6Yg4JXkdeqB5W2NZJygtEX99/jtYs3U3v3hqQ6HLMbMeIp89k2lAU0SsjYh9wAJgRqs+M4D5yfuFwLmSFBG7I+JPZEIlXznHOoLlrQMfPHkEU2sruf6B53zuxMw6RT5hUgOsz5puTtpy9omIFmAHUJXH2D9ODnF9PSsw8hpL0mxJjZIat27dmsdH2SGSuHTaGJq3v86ara8Wuhwz6wEKeQL+0xExGXhP8vrMkSwcEfMioj4i6qurq7ukwJ7srPHDAHhkjW9Rb2bp5RMmG4DRWdO1SVvOPpLKgEqg3W+piNiQ/LkL+BmZw2lHNZYdudFDj6Nm8HE80uS/WjNLL58wWQzUSRonqRyYCTS06tMAzEreXwI8GO0cjJdUJmlY8r4P8GFg+dGMZUdHEmdPqOLRtds4eNB/vWaWTodhkpy3mAMsAlYBd0XECklXS7oo6XYLUCWpCfgacPjyYUnrgO8Dl0lqTq4E6wsskrQUeJrM3siPOhrLOtdZ44ex4/X9rNzkuwmbWTp5PQM+Iu4F7m3V9o2s93uAj7ex7Ng2hj29jf5tjmWd613jM9c1PLpmG5NqKgtcjZkVM/8CvhcbMagf46sreGSN79VlZuk4THq5s8YP44nnX2a/fw1vZik4THq5s8ZXsXvfAZY27yh0KWZWxBwmvdw73545b/JIkw91mdnRc5j0ckMqypk4cpB/vGhmqThMjLPGV7Hkz9vZs/9AoUsxsyLlMDHOmlDFvpaDPOnnw5vZUXKYGGeMHUppiXyoy8yOmsPEGNivD1NqK/17EzM7ag4TAzLnTf67eQev7m0pdClmVoQcJgZkfrx44GCw+PmXC12KmRUhh4kBcPrxQygvLfGhLjM7Kg4TA6Bfn1JOO36wT8Kb2VFxmNhhZ40fxspNO9m+e1+hSzGzIuMwscPOGl9FBDz+vPdOzOzIOEzssCm1g+nXp4TH1vokvJkdGYeJHVZeVsLJoypZvsF3EDazI+MwsTeZXFPJio07afHzTczsCOQVJpKmS1otqUnSW57JLqmvpDuT+Y9LGpu0V0l6SNKrkm7I6t9f0j2SnpG0QtI1WfMuk7RV0tPJ66/Sr6bla0ptJa/vP8CarbsLXYqZFZEOw0RSKXAjcAEwEbhU0sRW3S4HtkfEBOA64NqkfQ/wdeBvcgz93Yg4ETgVOFvSBVnz7oyIU5LXzUe0RpbKlNrMs+CXNr9S4ErMrJjks2cyDWiKiLURsQ9YAMxo1WcGMD95vxA4V5IiYndE/IlMqBwWEa9FxEPJ+33Ak0BtivWwTjJu2AAqykt93sTMjkg+YVIDrM+abk7acvaJiBZgB1CVTwGSBgMXAg9kNX9M0lJJCyWNbmO52ZIaJTVu3bo1n4+yPJSWiJNrKlnqMDGzI1DQE/CSyoA7gOsjYm3S/GtgbERMAe7njT2eN4mIeRFRHxH11dXVx6bgXmJKTSUrN+5kv0/Cm1me8gmTDUD23kFt0pazTxIQlUA+v3ybBzwXEf98qCEitkXE3mTyZuD0PMaxTjS5tpK9LQd5bvOrhS7FzIpEPmGyGKiTNE5SOTATaGjVpwGYlby/BHgwIqK9QSX9I5nQ+Uqr9pFZkxcBq/Ko0TrR5JrMSfhlG3wS3szyU9ZRh4hokTQHWASUArdGxApJVwONEdEA3ALcLqkJeJlM4AAgaR0wCCiXdDFwPrAT+HvgGeBJSQA3JFdufVnSRUBLMtZlnbSulqexVRUM7FvGsg07+OQZha7GzIpBh2ECEBH3Ave2avtG1vs9wMfbWHZsG8Oqjf5XAVflU5d1jZISMammkmXNPglvZvnxL+Atpym1lazatIt9LT4Jb2Ydc5hYTpNrK9l34CDPbt5V6FLMrAg4TCynKTWDAVjqQ11mlgeHieU0euhxVB7Xx1d0mVleHCaWkyQm11SyzL+EN7M8OEysTZNrK1n94i727D9Q6FLMrJtzmFibptRUsv9AsPpFn4Q3s/Y5TKxNkw/djt6HusysAw4Ta1PN4OMYWlHOMj/bxMw64DCxNr1xEn5noUsxs27OYWLtmlJbybObfRLezNrnMLF2Taqp5MDBYOUm752YWdscJtauQ8+E900fzaw9DhNr19sG9WPYgL6+rYqZtcthYu2SxJTaSpb78mAza4fDxDo0uaaS57bsYuee/YUuxcy6KYeJdeicE4dzMOCOx/9c6FLMrJtymFiHpo4ezLsnDONHf3zelwibWU55hYmk6ZJWS2qSNDfH/L6S7kzmPy5pbNJeJekhSa9KuqHVMqdLWpYsc72SB8FLGirpfknPJX8OSb+altaccybw0qt7uatxfaFLMbNuqMMwkVQK3AhcAEwELpU0sVW3y4HtETEBuA64NmnfA3wd+JscQ/8Q+DxQl7ymJ+1zgQciog54IJm2Ajtz3FDqjx/CTQ+v8aN8zewt8tkzmQY0RcTaiNgHLABmtOozA5ifvF8InCtJEbE7Iv5EJlQOkzQSGBQRj0VEALcBF+cYa35WuxWQJK48ZwIbd+zhV09tKHQ5ZtbN5BMmNUD2sY3mpC1nn4hoAXYAVR2M2dzGmCMiYlPy/kVgRK4BJM2W1CipcevWrXmshqX1/hOqmVQziB883MSBg1HocsysG+nWJ+CTvZac31oRMS8i6iOivrq6+hhX1jtJYs4HJrBu22v859KNhS7HzLqRfMJkAzA6a7o2acvZR1IZUAls62DM2jbG3JwcBjt0OGxLHjXaMXL+xLdRN3wAP3hoDQe9d2JmiXzCZDFQJ2mcpHJgJtDQqk8DMCt5fwnwYLJXkVNyGGunpHcmV3F9Frg7x1izstqtGygpEV/6wHhWb97F71ZtLnQ5ZtZNdBgmyTmQOcAiYBVwV0SskHS1pIuSbrcAVZKagK+RdQWWpHXA94HLJDVnXQn2JeBmoAlYA9yXtF8DnCfpOeAvkmnrRi6cMooxQ/tzw0NNtPP/DGbWi6gnfBnU19dHY2NjocvoVe544s9c9Ytl3Pa5abz3BJ+zMitGkpZERH1njNWtT8Bb9/XR02oYVdmP/7NwqW8CaWYOEzs6fctKuXnWGUjw8ZseZdGKFwtdkpkVkMPEjtrEUYO4+8qzOeFtA7nip0u46fdrfA7FrJdymFgqwwf1487Z7+RDk0dyzX3P8LcLl/p2K2a9UFmhC7Di169PKf8681TGVw/g+gee48/bXuMbF05kUk1loUszs2PEYWKdoqREfO28ExhfXcHcny/jw//6J04dM5jPvut4Lpg0kn59Sgtdopl1IV8abJ1ux2v7WfhkMz997AWef2k3QyvK+UT9aD595hhGD+1f6PLMLNGZlwY7TKzLHDwYPLJmG7c/to77V25m+MB+PDL3HEpKVOjSzIzODRMf5rIuU1Ii3l03jHfXDWPjK6+z7qXdDhKzHsphYsfEqMHHMWrwcYUuw8y6iC8NNjOz1BwmZmaWmsPEzMxSc5iYmVlqDhMzM0vNYWJmZqk5TMzMLDWHiZmZpZZXmEiaLmm1pCZJc3PM7yvpzmT+45LGZs27KmlfLemDSds7JD2d9dop6SvJvG9K2pA170Ods6pmZtZVOvwFvKRS4EbgPKAZWCypISJWZnW7HNgeERMkzQSuBT4paSIwEzgZGAX8TtIJEbEaOCVr/A3AL7PGuy4ivpt+9czM7FjIZ89kGtAUEWsjYh+wAJjRqs8MYH7yfiFwriQl7QsiYm9EPA80JeNlOxdYExEvHO1KmJlZYeUTJjXA+qzp5qQtZ5+IaAF2AFV5LjsTuKNV2xxJSyXdKmlIrqIkzZbUKKlx69ateayGmZl1lYKegJdUDlwE/EdW8w+B8WQOg20Cvpdr2YiYFxH1EVFfXV3d5bWamVnb8gmTDcDorOnapC1nH0llQCWwLY9lLwCejIjNhxoiYnNEHIiIg8CPeOthMTMz62byCZPFQJ2kccmexEygoVWfBmBW8v4S4MHIPHWrAZiZXO01DqgDnsha7lJaHeKSNDJr8iPA8nxXxszMCqPDq7kiokXSHGARUArcGhErJF0NNEZEA3ALcLukJuBlMoFD0u8uYCXQAlwZEQcAJFWQuULsC60+8tuSTgECWJdjvpmZdTN+bK+ZWS/VmY/t9S/gzcwsNYeJmZml5jAxM7PUHCZmZpaaw8TMzFJzmJiZWWoOEzMzS81hYmZmqTlMzMwsNYeJmZml5jAxM7PUHCZmZpaaw8TMzFJzmJiZWWoOEzMzS81hYmZmqTlMzMwsNYeJmZmllleYSJouabWkJklzc8zvK+nOZP7jksZmzbsqaV8t6YNZ7eskLZP0tKTGrPahku6X9Fzy55B0q2hmZl2twzCRVArcCFwATAQulTSxVbfLge0RMQG4Drg2WXYiMBM4GZgO/CAZ75APRMQprZ5BPBd4ICLqgAeSaTMz68by2TOZBjRFxNqI2AcsAGa06jMDmJ+8XwicK0lJ+4KI2BsRzwNNyXjtyR5rPnBxHjWamVkB5RMmNcD6rOnmpC1nn4hoAXYAVR0sG8BvJS2RNDurz4iI2JS8fxEYkUeNZmZWQGUF/Ox3R8QGScOB+yU9ExF/yO4QESEpci2cBNBsgDFjxnR9tWZm1qZ89kw2AKOzpmuTtpx9JJUBlcC29paNiEN/bgF+yRuHvzZLGpmMNRLYkquoiJgXEfURUV9dXZ3HapiZWVfJJ0wWA3WSxkkqJ3NCvaFVnwZgVvL+EuDBiIikfWZytdc4oA54QlKFpIEAkiqA84HlOcaaBdx9dKtmZmbHSoeHuSKiRdIcYBFQCtwaESskXQ00RkQDcAtwu6Qm4GUygUPS7y5gJdACXBkRBySNAH6ZOUdPGfCziPhN8pHXAHdJuhx4AfhEJ66vmZl1AWV2IIpbfX19NDY2dtzRzMwOk7Sk1U8zjpp/AW9mZqk5TMzMLDWHiZmZpeYwMTOz1BwmZmaWmsPEzMxSc5iYmVlqDhMzM0vNYWJmZqk5TMzMLDWHiZmZpeYwMTOz1BwmZmaWmsPEzMxSc5iYmVlqDhMzM0vNYWJmZqk5TMzMLLW8wkTSdEmrJTVJmptjfl9JdybzH5c0NmveVUn7akkfTNpGS3pI0kpJKyT976z+35S0QdLTyetD6VfTzMy6UllHHSSVAjcC5wHNwGJJDRGxMqvb5cD2iJggaSZwLfBJSROBmcDJwCjgd5JOAFqAv46IJyUNBJZIuj9rzOsi4rudtZJmZta18tkzmQY0RcTaiNgHLABmtOozA5ifvF8InCtJSfuCiNgbEc8DTcC0iNgUEU8CRMQuYBVQk351zMysEPIJkxpgfdZ0M2/94j/cJyJagB1AVT7LJofETgUez2qeI2mppFslDcmjRjMzK6CCnoCXNAD4OfCViNiZNP8QGA+cAmwCvtfGsrMlNUpq3Lp16zGp18zMcssnTDYAo7Oma5O2nH0klQGVwLb2lpXUh0yQ/HtE/OJQh4jYHBEHIuIg8CMyh9neIiLmRUR9RNRXV1fnsRpmZtZV8gmTxUCdpHGSysmcUG9o1acBmJW8vwR4MCIiaZ+ZXO01DqgDnkjOp9wCrIqI72cPJGlk1uRHgOVHulJmZnZsdXg1V0S0SJoDLAJKgVsjYoWkq4HGiGggEwy3S2oCXiYTOCT97gJWkrmC68qIOCDp3cBngGWSnk4+6v9GxL3AtyWdAgSwDvhCJ66vmZl1AWV2IIpbfX19NDY2FroMM7OiImlJRNR3xlj+BbyZmaXmMDEzs9QcJmZmlprDxMzMUnOYmJlZag4TMzNLzWFiZmapOUzMzCw1h4mZmaXmMDEzs9QcJmZmlprDxMzMUnOYmJlZag4TMzNLzWFiZmapOUzMzCw1h4mZmaXmMDEzs9QcJmZmllpeYSJpuqTVkpokzc0xv6+kO5P5j0samzXvqqR9taQPdjSmpHHJGE3JmOXpVtHMzLpah2EiqRS4EbgAmAhcKmliq26XA9sjYgJwHXBtsuxEYCZwMjAd+IGk0g7GvBa4LhlrezK2mZl1Y/nsmUwDmiJibUTsAxYAM1r1mQHMT94vBM6VpKR9QUTsjYjngaZkvJxjJsuck4xBMubFR796ZmZ2LJTl0acGWJ813Qyc2VafiGiRtAOoStofa7VsTfI+15hVwCsR0ZKj/5tImg3MTib3Slqex7p0tUpgR4HHOpLl8unbXp+jmZerfRjwUgd1dLXO3HZpxusO26/Yth10j+13LLdde/OPpP0dHdSQv4ho9wVcAtycNf0Z4IZWfZYDtVnTa8j8R3YD8D+y2m9Jxss5ZrJMU1b7aGB5HjU2dtTnWLyAeYUe60iWy6dve32OZl6u9u6w/Tpz2xX79iu2bdddtt+x3Hbdcfvlc5hrQ/Klfkht0pazj6QyMgm4rZ1l22rfBgxOxmjrs7qzX3eDsY5kuXz6ttfnaOZ15t9RZ+rsuop5+xXbtoPusf2O5bZrb35Btp+SdGq7Q+aL/VngXDJf7IuBT0XEiqw+VwKTI+IKSTOBj0bEJySdDPyMzDmSUcADQB2gtsaU9B/AzyNigaSbgKUR8YMOamyMiPqjWH/rBrz9ipe3XXHrzO3X4TmTyJwDmQMsAkqBW5Mv/avJ7CI1kDl8dbukJuBlMldwkfS7C1gJtABXRsSBZCXeMmbykX8HLJD0j8BTydgdmZf3Glt35O1XvLztilunbb8O90zMzMw64l/Am5lZag4TMzNLzWFiZmap9YowkVQhqVHShwtdi+VP0kmSbpK0UNIXC12PHRlJF0v6UXKPvfMLXY8dGUlvl3SLpIUd9+7mYSLpVklbWv+6vaMbT+bwd8BdXVOl5dIZ2y4iVkXEFcAngLO7sl57s07afr+KiM8DVwCf7Mp67c06afutjYi8743Yra/mkvRe4FXgtoiYlLSVkvmNynlkbreyGLiUzCXG32o1xOeAqWRu09IPeCki/vPYVN+7dca2i4gtki4CvgjcHhE/O1b193adtf2S5b4H/HtEPHmMyu/1Onn7LYyISzr6zHzuzVUwEfGH7NvZJw7fJBJA0gJgRkR8C3jLYSxJ7wcqyNyd+HVJ90bEwa6s2zpn2yXjNAANku4h8wNYOwY66d+egGuA+xwkx1Zn/fs7Et06TNqQz40nD4uIvweQdBmZPRMHSeEc0bZL/kfgo0Bf4N4urczycUTbD/hfwF8AlZImRMRNXVmcdehI//1VAf8EnCrpqiR02lSMYXJUIuInha7BjkxEPAw8XOAy7ChFxPXA9YWuw45ORGwjc74rL936BHwb8rnxpHVP3nbFzduvuHXp9ivGMFkM1CWP9y0ncx+whgLXZPnxtitu3n7FrUu3X7cOE0l3AI8C75DULOnyyDw469BNIlcBd2Xfwdi6B2+74ubtV9wKsf269aXBZmZWHLr1nomZmf0h/f8AAAA1SURBVBUHh4mZmaXmMDEzs9QcJmZmlprDxMzMUnOYmJlZag4TMzNLzWFiZmapOUzMzCy1/w/mciyENcQaGQAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "L0vc10arAqFs",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 288
        },
        "outputId": "6af4c2bc-05aa-4c84-df77-c9ffa9fadef3"
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "plt.semilogx(lr_history.history[\"lr\"], lr_history.history[\"mae\"])\n",
        "plt.axis([1e-4, 1e-1, 0, 0.2])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(0.0001, 0.1, 0.0, 0.2)"
            ]
          },
          "metadata": {},
          "execution_count": 14
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-0jlZjYjAqne",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 288
        },
        "outputId": "1760b176-bf1a-49e4-e358-3f5436fa78e8"
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "plt.semilogx(lr_history.history[\"lr\"], lr_history.history[\"mse\"])\n",
        "plt.axis([1e-4, 1e-1, 0, 0.06])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(0.0001, 0.1, 0.0, 0.06)"
            ]
          },
          "metadata": {},
          "execution_count": 15
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ylSRbCVUJb5l",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 802
        },
        "outputId": "6ce15fa9-6580-4726-cac4-7dbc6fafaf27"
      },
      "source": [
        "plot_graphs(lr_history, \"loss\")\n",
        "plot_graphs(lr_history, \"mae\")\n",
        "plot_graphs(lr_history, \"mse\") "
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DVMGLGbgggA6"
      },
      "source": [
        "learning_rate=9e-4"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "2IhlGFYw0Lq1",
        "outputId": "5dffe30f-e9d4-4ff2-e658-64a03b9be363"
      },
      "source": [
        "tf.keras.backend.clear_session()\n",
        "tf.random.set_seed(51)\n",
        "np.random.seed(51)\n",
        "\n",
        "\n",
        "BATCH_SIZE = 32\n",
        "N_PAST = 24  \n",
        "N_FUTURE = 24\n",
        "SHIFT = 1  \n",
        "\n",
        "train_set = windowed_dataset(series=x_train, batch_size=BATCH_SIZE,\n",
        "                  n_past=N_PAST, n_future=N_FUTURE,\n",
        "                  shift=SHIFT)\n",
        "\n",
        "valid_set = windowed_dataset(series=x_valid, batch_size=BATCH_SIZE,\n",
        "                  n_past=N_PAST, n_future=N_FUTURE,\n",
        "                  shift=SHIFT)\n",
        "\n",
        "# Code to define your model.\n",
        "print(train_set)\n",
        "print(valid_set)\n",
        "\n",
        "inputs = tf.keras.layers.Input(shape = [N_PAST, N_FEATURES])\n",
        "outputs = []\n",
        "rnn_units = 32\n",
        "\n",
        "rnn = tf.keras.layers.LSTM(rnn_units, return_state=True, return_sequences=False)\n",
        "rnn_cell = tf.keras.layers.LSTMCell(rnn_units)\n",
        "densor = tf.keras.layers.Dense(N_FEATURES)\n",
        "\n",
        "# First tx to predict, and the output of this tx will be used for (N_FEATURES-1) predictions.\n",
        "x, *state = rnn(inputs)\n",
        "x = densor(x)\n",
        "outputs.append(x)\n",
        "\n",
        "# Use first prediction to predict others.\n",
        "for Tx in range(1, N_FUTURE):\n",
        "  x, state = rnn_cell(x, states=state)\n",
        "  x = densor(x)\n",
        "  outputs.append(x)\n",
        "\n",
        "outputs = tf.stack(outputs)\n",
        "outputs = tf.transpose(outputs, [1, 0, 2])\n",
        "\n",
        "train_model = tf.keras.Model(inputs, outputs)\n",
        "\n",
        "train_model.compile(\n",
        "    loss=tf.keras.losses.Huber(),\n",
        "    optimizer=tf.keras.optimizers.SGD(learning_rate=learning_rate, momentum=0.9),\n",
        "    metrics=[\"mae\", \"mse\"]\n",
        ")\n",
        "\n",
        "train_model.summary()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(43200, 7)\n",
            "(43200, 7)\n",
            "<PrefetchDataset shapes: ((None, None, 7), (None, None, 7)), types: (tf.float64, tf.float64)>\n",
            "<PrefetchDataset shapes: ((None, None, 7), (None, None, 7)), types: (tf.float64, tf.float64)>\n",
            "Model: \"model\"\n",
            "__________________________________________________________________________________________________\n",
            "Layer (type)                    Output Shape         Param #     Connected to                     \n",
            "==================================================================================================\n",
            "input_1 (InputLayer)            [(None, 24, 7)]      0                                            \n",
            "__________________________________________________________________________________________________\n",
            "lstm (LSTM)                     [(None, 32), (None,  5120        input_1[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "dense (Dense)                   (None, 7)            231         lstm[0][0]                       \n",
            "                                                                 lstm_cell_1[0][0]                \n",
            "                                                                 lstm_cell_1[1][0]                \n",
            "                                                                 lstm_cell_1[2][0]                \n",
            "                                                                 lstm_cell_1[3][0]                \n",
            "                                                                 lstm_cell_1[4][0]                \n",
            "                                                                 lstm_cell_1[5][0]                \n",
            "                                                                 lstm_cell_1[6][0]                \n",
            "                                                                 lstm_cell_1[7][0]                \n",
            "                                                                 lstm_cell_1[8][0]                \n",
            "                                                                 lstm_cell_1[9][0]                \n",
            "                                                                 lstm_cell_1[10][0]               \n",
            "                                                                 lstm_cell_1[11][0]               \n",
            "                                                                 lstm_cell_1[12][0]               \n",
            "                                                                 lstm_cell_1[13][0]               \n",
            "                                                                 lstm_cell_1[14][0]               \n",
            "                                                                 lstm_cell_1[15][0]               \n",
            "                                                                 lstm_cell_1[16][0]               \n",
            "                                                                 lstm_cell_1[17][0]               \n",
            "                                                                 lstm_cell_1[18][0]               \n",
            "                                                                 lstm_cell_1[19][0]               \n",
            "                                                                 lstm_cell_1[20][0]               \n",
            "                                                                 lstm_cell_1[21][0]               \n",
            "                                                                 lstm_cell_1[22][0]               \n",
            "__________________________________________________________________________________________________\n",
            "lstm_cell_1 (LSTMCell)          ((None, 32), [(None, 5120        dense[0][0]                      \n",
            "                                                                 lstm[0][1]                       \n",
            "                                                                 lstm[0][2]                       \n",
            "                                                                 dense[1][0]                      \n",
            "                                                                 lstm_cell_1[0][1]                \n",
            "                                                                 lstm_cell_1[0][2]                \n",
            "                                                                 dense[2][0]                      \n",
            "                                                                 lstm_cell_1[1][1]                \n",
            "                                                                 lstm_cell_1[1][2]                \n",
            "                                                                 dense[3][0]                      \n",
            "                                                                 lstm_cell_1[2][1]                \n",
            "                                                                 lstm_cell_1[2][2]                \n",
            "                                                                 dense[4][0]                      \n",
            "                                                                 lstm_cell_1[3][1]                \n",
            "                                                                 lstm_cell_1[3][2]                \n",
            "                                                                 dense[5][0]                      \n",
            "                                                                 lstm_cell_1[4][1]                \n",
            "                                                                 lstm_cell_1[4][2]                \n",
            "                                                                 dense[6][0]                      \n",
            "                                                                 lstm_cell_1[5][1]                \n",
            "                                                                 lstm_cell_1[5][2]                \n",
            "                                                                 dense[7][0]                      \n",
            "                                                                 lstm_cell_1[6][1]                \n",
            "                                                                 lstm_cell_1[6][2]                \n",
            "                                                                 dense[8][0]                      \n",
            "                                                                 lstm_cell_1[7][1]                \n",
            "                                                                 lstm_cell_1[7][2]                \n",
            "                                                                 dense[9][0]                      \n",
            "                                                                 lstm_cell_1[8][1]                \n",
            "                                                                 lstm_cell_1[8][2]                \n",
            "                                                                 dense[10][0]                     \n",
            "                                                                 lstm_cell_1[9][1]                \n",
            "                                                                 lstm_cell_1[9][2]                \n",
            "                                                                 dense[11][0]                     \n",
            "                                                                 lstm_cell_1[10][1]               \n",
            "                                                                 lstm_cell_1[10][2]               \n",
            "                                                                 dense[12][0]                     \n",
            "                                                                 lstm_cell_1[11][1]               \n",
            "                                                                 lstm_cell_1[11][2]               \n",
            "                                                                 dense[13][0]                     \n",
            "                                                                 lstm_cell_1[12][1]               \n",
            "                                                                 lstm_cell_1[12][2]               \n",
            "                                                                 dense[14][0]                     \n",
            "                                                                 lstm_cell_1[13][1]               \n",
            "                                                                 lstm_cell_1[13][2]               \n",
            "                                                                 dense[15][0]                     \n",
            "                                                                 lstm_cell_1[14][1]               \n",
            "                                                                 lstm_cell_1[14][2]               \n",
            "                                                                 dense[16][0]                     \n",
            "                                                                 lstm_cell_1[15][1]               \n",
            "                                                                 lstm_cell_1[15][2]               \n",
            "                                                                 dense[17][0]                     \n",
            "                                                                 lstm_cell_1[16][1]               \n",
            "                                                                 lstm_cell_1[16][2]               \n",
            "                                                                 dense[18][0]                     \n",
            "                                                                 lstm_cell_1[17][1]               \n",
            "                                                                 lstm_cell_1[17][2]               \n",
            "                                                                 dense[19][0]                     \n",
            "                                                                 lstm_cell_1[18][1]               \n",
            "                                                                 lstm_cell_1[18][2]               \n",
            "                                                                 dense[20][0]                     \n",
            "                                                                 lstm_cell_1[19][1]               \n",
            "                                                                 lstm_cell_1[19][2]               \n",
            "                                                                 dense[21][0]                     \n",
            "                                                                 lstm_cell_1[20][1]               \n",
            "                                                                 lstm_cell_1[20][2]               \n",
            "                                                                 dense[22][0]                     \n",
            "                                                                 lstm_cell_1[21][1]               \n",
            "                                                                 lstm_cell_1[21][2]               \n",
            "__________________________________________________________________________________________________\n",
            "tf.stack (TFOpLambda)           (24, None, 7)        0           dense[0][0]                      \n",
            "                                                                 dense[1][0]                      \n",
            "                                                                 dense[2][0]                      \n",
            "                                                                 dense[3][0]                      \n",
            "                                                                 dense[4][0]                      \n",
            "                                                                 dense[5][0]                      \n",
            "                                                                 dense[6][0]                      \n",
            "                                                                 dense[7][0]                      \n",
            "                                                                 dense[8][0]                      \n",
            "                                                                 dense[9][0]                      \n",
            "                                                                 dense[10][0]                     \n",
            "                                                                 dense[11][0]                     \n",
            "                                                                 dense[12][0]                     \n",
            "                                                                 dense[13][0]                     \n",
            "                                                                 dense[14][0]                     \n",
            "                                                                 dense[15][0]                     \n",
            "                                                                 dense[16][0]                     \n",
            "                                                                 dense[17][0]                     \n",
            "                                                                 dense[18][0]                     \n",
            "                                                                 dense[19][0]                     \n",
            "                                                                 dense[20][0]                     \n",
            "                                                                 dense[21][0]                     \n",
            "                                                                 dense[22][0]                     \n",
            "                                                                 dense[23][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "tf.compat.v1.transpose (TFOpLam (None, 24, 7)        0           tf.stack[0][0]                   \n",
            "==================================================================================================\n",
            "Total params: 10,471\n",
            "Trainable params: 10,471\n",
            "Non-trainable params: 0\n",
            "__________________________________________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LtWeIEMDNlxe"
      },
      "source": [
        "save_model = \"ts-win2win.h5\"\n",
        "early_stopping = tf.keras.callbacks.EarlyStopping(patience=50)\n",
        "model_checkpoint = tf.keras.callbacks.ModelCheckpoint(save_model, save_best_only=True)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "SlDEwJHT0SvR",
        "outputId": "d85685e5-4d0a-4d0c-f7fc-0b06d0eed616"
      },
      "source": [
        "train_history = process_in_duration(\n",
        "    \"Training end,\", \n",
        "    lambda:\n",
        "          train_model.fit(train_set, \n",
        "                    validation_data=valid_set,\n",
        "                    epochs=500, \n",
        "                    callbacks=[early_stopping, model_checkpoint]\n",
        "                    ))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/500\n",
            "1349/1349 [==============================] - 24s 15ms/step - loss: 0.0219 - mae: 0.1210 - mse: 0.0437 - val_loss: 0.0185 - val_mae: 0.1181 - val_mse: 0.0369\n",
            "Epoch 2/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0182 - mae: 0.1198 - mse: 0.0365 - val_loss: 0.0178 - val_mae: 0.1183 - val_mse: 0.0356\n",
            "Epoch 3/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0176 - mae: 0.1186 - mse: 0.0353 - val_loss: 0.0172 - val_mae: 0.1164 - val_mse: 0.0344\n",
            "Epoch 4/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0170 - mae: 0.1165 - mse: 0.0339 - val_loss: 0.0164 - val_mae: 0.1138 - val_mse: 0.0327\n",
            "Epoch 5/500\n",
            "1349/1349 [==============================] - 20s 14ms/step - loss: 0.0158 - mae: 0.1133 - mse: 0.0317 - val_loss: 0.0148 - val_mae: 0.1099 - val_mse: 0.0295\n",
            "Epoch 6/500\n",
            "1349/1349 [==============================] - 20s 14ms/step - loss: 0.0132 - mae: 0.1066 - mse: 0.0264 - val_loss: 0.0109 - val_mae: 0.1004 - val_mse: 0.0217\n",
            "Epoch 7/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0100 - mae: 0.0917 - mse: 0.0200 - val_loss: 0.0091 - val_mae: 0.0845 - val_mse: 0.0181\n",
            "Epoch 8/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0091 - mae: 0.0807 - mse: 0.0183 - val_loss: 0.0086 - val_mae: 0.0779 - val_mse: 0.0171\n",
            "Epoch 9/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0088 - mae: 0.0766 - mse: 0.0177 - val_loss: 0.0084 - val_mae: 0.0755 - val_mse: 0.0167\n",
            "Epoch 10/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0087 - mae: 0.0749 - mse: 0.0174 - val_loss: 0.0082 - val_mae: 0.0744 - val_mse: 0.0165\n",
            "Epoch 11/500\n",
            "1349/1349 [==============================] - 20s 14ms/step - loss: 0.0086 - mae: 0.0739 - mse: 0.0172 - val_loss: 0.0082 - val_mae: 0.0738 - val_mse: 0.0163\n",
            "Epoch 12/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0085 - mae: 0.0733 - mse: 0.0171 - val_loss: 0.0081 - val_mae: 0.0734 - val_mse: 0.0162\n",
            "Epoch 13/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0085 - mae: 0.0729 - mse: 0.0170 - val_loss: 0.0081 - val_mae: 0.0730 - val_mse: 0.0162\n",
            "Epoch 14/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0084 - mae: 0.0726 - mse: 0.0169 - val_loss: 0.0080 - val_mae: 0.0728 - val_mse: 0.0161\n",
            "Epoch 15/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0084 - mae: 0.0724 - mse: 0.0168 - val_loss: 0.0080 - val_mae: 0.0726 - val_mse: 0.0160\n",
            "Epoch 16/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0084 - mae: 0.0721 - mse: 0.0167 - val_loss: 0.0080 - val_mae: 0.0724 - val_mse: 0.0160\n",
            "Epoch 17/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0083 - mae: 0.0719 - mse: 0.0167 - val_loss: 0.0080 - val_mae: 0.0722 - val_mse: 0.0159\n",
            "Epoch 18/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0083 - mae: 0.0718 - mse: 0.0166 - val_loss: 0.0079 - val_mae: 0.0721 - val_mse: 0.0158\n",
            "Epoch 19/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0083 - mae: 0.0716 - mse: 0.0166 - val_loss: 0.0079 - val_mae: 0.0719 - val_mse: 0.0158\n",
            "Epoch 20/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0083 - mae: 0.0715 - mse: 0.0165 - val_loss: 0.0079 - val_mae: 0.0718 - val_mse: 0.0158\n",
            "Epoch 21/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0082 - mae: 0.0713 - mse: 0.0165 - val_loss: 0.0079 - val_mae: 0.0717 - val_mse: 0.0157\n",
            "Epoch 22/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0082 - mae: 0.0712 - mse: 0.0164 - val_loss: 0.0078 - val_mae: 0.0716 - val_mse: 0.0157\n",
            "Epoch 23/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0082 - mae: 0.0711 - mse: 0.0164 - val_loss: 0.0078 - val_mae: 0.0715 - val_mse: 0.0156\n",
            "Epoch 24/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0082 - mae: 0.0709 - mse: 0.0163 - val_loss: 0.0078 - val_mae: 0.0714 - val_mse: 0.0156\n",
            "Epoch 25/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0081 - mae: 0.0708 - mse: 0.0163 - val_loss: 0.0078 - val_mae: 0.0713 - val_mse: 0.0155\n",
            "Epoch 26/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0081 - mae: 0.0707 - mse: 0.0162 - val_loss: 0.0078 - val_mae: 0.0712 - val_mse: 0.0155\n",
            "Epoch 27/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0081 - mae: 0.0706 - mse: 0.0162 - val_loss: 0.0077 - val_mae: 0.0711 - val_mse: 0.0155\n",
            "Epoch 28/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0081 - mae: 0.0705 - mse: 0.0161 - val_loss: 0.0077 - val_mae: 0.0710 - val_mse: 0.0154\n",
            "Epoch 29/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0080 - mae: 0.0704 - mse: 0.0161 - val_loss: 0.0077 - val_mae: 0.0709 - val_mse: 0.0154\n",
            "Epoch 30/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0080 - mae: 0.0702 - mse: 0.0160 - val_loss: 0.0077 - val_mae: 0.0708 - val_mse: 0.0154\n",
            "Epoch 31/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0080 - mae: 0.0701 - mse: 0.0160 - val_loss: 0.0077 - val_mae: 0.0707 - val_mse: 0.0153\n",
            "Epoch 32/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0080 - mae: 0.0700 - mse: 0.0159 - val_loss: 0.0076 - val_mae: 0.0707 - val_mse: 0.0153\n",
            "Epoch 33/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0079 - mae: 0.0699 - mse: 0.0159 - val_loss: 0.0076 - val_mae: 0.0706 - val_mse: 0.0153\n",
            "Epoch 34/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0079 - mae: 0.0698 - mse: 0.0159 - val_loss: 0.0076 - val_mae: 0.0705 - val_mse: 0.0152\n",
            "Epoch 35/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0079 - mae: 0.0696 - mse: 0.0158 - val_loss: 0.0076 - val_mae: 0.0704 - val_mse: 0.0152\n",
            "Epoch 36/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0079 - mae: 0.0695 - mse: 0.0158 - val_loss: 0.0076 - val_mae: 0.0703 - val_mse: 0.0152\n",
            "Epoch 37/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0079 - mae: 0.0694 - mse: 0.0157 - val_loss: 0.0076 - val_mae: 0.0702 - val_mse: 0.0151\n",
            "Epoch 38/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0078 - mae: 0.0692 - mse: 0.0157 - val_loss: 0.0075 - val_mae: 0.0701 - val_mse: 0.0151\n",
            "Epoch 39/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0078 - mae: 0.0691 - mse: 0.0156 - val_loss: 0.0075 - val_mae: 0.0699 - val_mse: 0.0150\n",
            "Epoch 40/500\n",
            "1349/1349 [==============================] - 20s 14ms/step - loss: 0.0078 - mae: 0.0689 - mse: 0.0156 - val_loss: 0.0075 - val_mae: 0.0698 - val_mse: 0.0150\n",
            "Epoch 41/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0078 - mae: 0.0688 - mse: 0.0155 - val_loss: 0.0075 - val_mae: 0.0697 - val_mse: 0.0149\n",
            "Epoch 42/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0077 - mae: 0.0686 - mse: 0.0155 - val_loss: 0.0075 - val_mae: 0.0696 - val_mse: 0.0149\n",
            "Epoch 43/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0077 - mae: 0.0684 - mse: 0.0154 - val_loss: 0.0074 - val_mae: 0.0694 - val_mse: 0.0149\n",
            "Epoch 44/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0077 - mae: 0.0682 - mse: 0.0153 - val_loss: 0.0074 - val_mae: 0.0692 - val_mse: 0.0148\n",
            "Epoch 45/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0076 - mae: 0.0680 - mse: 0.0153 - val_loss: 0.0074 - val_mae: 0.0690 - val_mse: 0.0147\n",
            "Epoch 46/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0076 - mae: 0.0678 - mse: 0.0152 - val_loss: 0.0073 - val_mae: 0.0688 - val_mse: 0.0147\n",
            "Epoch 47/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0076 - mae: 0.0675 - mse: 0.0151 - val_loss: 0.0073 - val_mae: 0.0686 - val_mse: 0.0146\n",
            "Epoch 48/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0075 - mae: 0.0672 - mse: 0.0151 - val_loss: 0.0073 - val_mae: 0.0683 - val_mse: 0.0146\n",
            "Epoch 49/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0075 - mae: 0.0669 - mse: 0.0150 - val_loss: 0.0072 - val_mae: 0.0680 - val_mse: 0.0145\n",
            "Epoch 50/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0075 - mae: 0.0666 - mse: 0.0149 - val_loss: 0.0072 - val_mae: 0.0677 - val_mse: 0.0144\n",
            "Epoch 51/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0074 - mae: 0.0662 - mse: 0.0148 - val_loss: 0.0072 - val_mae: 0.0674 - val_mse: 0.0144\n",
            "Epoch 52/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0074 - mae: 0.0658 - mse: 0.0147 - val_loss: 0.0072 - val_mae: 0.0670 - val_mse: 0.0143\n",
            "Epoch 53/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0073 - mae: 0.0654 - mse: 0.0147 - val_loss: 0.0071 - val_mae: 0.0666 - val_mse: 0.0143\n",
            "Epoch 54/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0073 - mae: 0.0650 - mse: 0.0146 - val_loss: 0.0071 - val_mae: 0.0662 - val_mse: 0.0142\n",
            "Epoch 55/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0073 - mae: 0.0646 - mse: 0.0145 - val_loss: 0.0071 - val_mae: 0.0658 - val_mse: 0.0142\n",
            "Epoch 56/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0072 - mae: 0.0642 - mse: 0.0145 - val_loss: 0.0071 - val_mae: 0.0653 - val_mse: 0.0141\n",
            "Epoch 57/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0072 - mae: 0.0637 - mse: 0.0144 - val_loss: 0.0070 - val_mae: 0.0649 - val_mse: 0.0141\n",
            "Epoch 58/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0072 - mae: 0.0633 - mse: 0.0144 - val_loss: 0.0070 - val_mae: 0.0645 - val_mse: 0.0140\n",
            "Epoch 59/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0072 - mae: 0.0629 - mse: 0.0143 - val_loss: 0.0070 - val_mae: 0.0641 - val_mse: 0.0140\n",
            "Epoch 60/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0071 - mae: 0.0625 - mse: 0.0143 - val_loss: 0.0070 - val_mae: 0.0637 - val_mse: 0.0140\n",
            "Epoch 61/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0071 - mae: 0.0621 - mse: 0.0142 - val_loss: 0.0070 - val_mae: 0.0634 - val_mse: 0.0139\n",
            "Epoch 62/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0071 - mae: 0.0618 - mse: 0.0142 - val_loss: 0.0070 - val_mae: 0.0630 - val_mse: 0.0139\n",
            "Epoch 63/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0071 - mae: 0.0614 - mse: 0.0141 - val_loss: 0.0069 - val_mae: 0.0627 - val_mse: 0.0139\n",
            "Epoch 64/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0070 - mae: 0.0611 - mse: 0.0141 - val_loss: 0.0069 - val_mae: 0.0624 - val_mse: 0.0139\n",
            "Epoch 65/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0070 - mae: 0.0608 - mse: 0.0141 - val_loss: 0.0069 - val_mae: 0.0621 - val_mse: 0.0138\n",
            "Epoch 66/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0070 - mae: 0.0605 - mse: 0.0140 - val_loss: 0.0069 - val_mae: 0.0618 - val_mse: 0.0138\n",
            "Epoch 67/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0070 - mae: 0.0603 - mse: 0.0140 - val_loss: 0.0069 - val_mae: 0.0616 - val_mse: 0.0138\n",
            "Epoch 68/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0070 - mae: 0.0600 - mse: 0.0140 - val_loss: 0.0069 - val_mae: 0.0613 - val_mse: 0.0138\n",
            "Epoch 69/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0070 - mae: 0.0598 - mse: 0.0139 - val_loss: 0.0069 - val_mae: 0.0611 - val_mse: 0.0138\n",
            "Epoch 70/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0070 - mae: 0.0595 - mse: 0.0139 - val_loss: 0.0069 - val_mae: 0.0609 - val_mse: 0.0137\n",
            "Epoch 71/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0069 - mae: 0.0593 - mse: 0.0139 - val_loss: 0.0069 - val_mae: 0.0607 - val_mse: 0.0137\n",
            "Epoch 72/500\n",
            "1349/1349 [==============================] - 20s 14ms/step - loss: 0.0069 - mae: 0.0591 - mse: 0.0139 - val_loss: 0.0069 - val_mae: 0.0605 - val_mse: 0.0137\n",
            "Epoch 73/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0069 - mae: 0.0590 - mse: 0.0138 - val_loss: 0.0069 - val_mae: 0.0603 - val_mse: 0.0137\n",
            "Epoch 74/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0069 - mae: 0.0588 - mse: 0.0138 - val_loss: 0.0068 - val_mae: 0.0602 - val_mse: 0.0137\n",
            "Epoch 75/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0069 - mae: 0.0586 - mse: 0.0138 - val_loss: 0.0068 - val_mae: 0.0600 - val_mse: 0.0137\n",
            "Epoch 76/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0069 - mae: 0.0585 - mse: 0.0138 - val_loss: 0.0068 - val_mae: 0.0599 - val_mse: 0.0137\n",
            "Epoch 77/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0069 - mae: 0.0583 - mse: 0.0138 - val_loss: 0.0068 - val_mae: 0.0597 - val_mse: 0.0137\n",
            "Epoch 78/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0069 - mae: 0.0582 - mse: 0.0138 - val_loss: 0.0068 - val_mae: 0.0596 - val_mse: 0.0137\n",
            "Epoch 79/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0069 - mae: 0.0581 - mse: 0.0137 - val_loss: 0.0068 - val_mae: 0.0595 - val_mse: 0.0136\n",
            "Epoch 80/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0069 - mae: 0.0580 - mse: 0.0137 - val_loss: 0.0068 - val_mae: 0.0594 - val_mse: 0.0136\n",
            "Epoch 81/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0069 - mae: 0.0579 - mse: 0.0137 - val_loss: 0.0068 - val_mae: 0.0593 - val_mse: 0.0136\n",
            "Epoch 82/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0069 - mae: 0.0578 - mse: 0.0137 - val_loss: 0.0068 - val_mae: 0.0592 - val_mse: 0.0136\n",
            "Epoch 83/500\n",
            "1349/1349 [==============================] - 20s 14ms/step - loss: 0.0068 - mae: 0.0577 - mse: 0.0137 - val_loss: 0.0068 - val_mae: 0.0592 - val_mse: 0.0136\n",
            "Epoch 84/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0068 - mae: 0.0577 - mse: 0.0137 - val_loss: 0.0068 - val_mae: 0.0591 - val_mse: 0.0136\n",
            "Epoch 85/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0068 - mae: 0.0576 - mse: 0.0137 - val_loss: 0.0068 - val_mae: 0.0590 - val_mse: 0.0136\n",
            "Epoch 86/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0068 - mae: 0.0575 - mse: 0.0137 - val_loss: 0.0068 - val_mae: 0.0589 - val_mse: 0.0136\n",
            "Epoch 87/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0068 - mae: 0.0574 - mse: 0.0137 - val_loss: 0.0068 - val_mae: 0.0589 - val_mse: 0.0136\n",
            "Epoch 88/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0068 - mae: 0.0574 - mse: 0.0136 - val_loss: 0.0068 - val_mae: 0.0588 - val_mse: 0.0136\n",
            "Epoch 89/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0068 - mae: 0.0573 - mse: 0.0136 - val_loss: 0.0068 - val_mae: 0.0588 - val_mse: 0.0136\n",
            "Epoch 90/500\n",
            "1349/1349 [==============================] - 20s 14ms/step - loss: 0.0068 - mae: 0.0573 - mse: 0.0136 - val_loss: 0.0068 - val_mae: 0.0587 - val_mse: 0.0136\n",
            "Epoch 91/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0068 - mae: 0.0572 - mse: 0.0136 - val_loss: 0.0068 - val_mae: 0.0587 - val_mse: 0.0136\n",
            "Epoch 92/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0068 - mae: 0.0572 - mse: 0.0136 - val_loss: 0.0068 - val_mae: 0.0586 - val_mse: 0.0136\n",
            "Epoch 93/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0068 - mae: 0.0571 - mse: 0.0136 - val_loss: 0.0068 - val_mae: 0.0586 - val_mse: 0.0135\n",
            "Epoch 94/500\n",
            "1349/1349 [==============================] - 20s 14ms/step - loss: 0.0068 - mae: 0.0571 - mse: 0.0136 - val_loss: 0.0068 - val_mae: 0.0585 - val_mse: 0.0135\n",
            "Epoch 95/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0068 - mae: 0.0570 - mse: 0.0136 - val_loss: 0.0068 - val_mae: 0.0585 - val_mse: 0.0135\n",
            "Epoch 96/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0068 - mae: 0.0570 - mse: 0.0136 - val_loss: 0.0068 - val_mae: 0.0584 - val_mse: 0.0135\n",
            "Epoch 97/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0068 - mae: 0.0569 - mse: 0.0136 - val_loss: 0.0068 - val_mae: 0.0584 - val_mse: 0.0135\n",
            "Epoch 98/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0068 - mae: 0.0569 - mse: 0.0136 - val_loss: 0.0068 - val_mae: 0.0583 - val_mse: 0.0135\n",
            "Epoch 99/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0068 - mae: 0.0568 - mse: 0.0136 - val_loss: 0.0068 - val_mae: 0.0583 - val_mse: 0.0135\n",
            "Epoch 100/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0068 - mae: 0.0568 - mse: 0.0135 - val_loss: 0.0068 - val_mae: 0.0583 - val_mse: 0.0135\n",
            "Epoch 101/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0068 - mae: 0.0568 - mse: 0.0135 - val_loss: 0.0068 - val_mae: 0.0582 - val_mse: 0.0135\n",
            "Epoch 102/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0068 - mae: 0.0567 - mse: 0.0135 - val_loss: 0.0068 - val_mae: 0.0582 - val_mse: 0.0135\n",
            "Epoch 103/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0068 - mae: 0.0567 - mse: 0.0135 - val_loss: 0.0067 - val_mae: 0.0582 - val_mse: 0.0135\n",
            "Epoch 104/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0068 - mae: 0.0567 - mse: 0.0135 - val_loss: 0.0067 - val_mae: 0.0581 - val_mse: 0.0135\n",
            "Epoch 105/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0068 - mae: 0.0566 - mse: 0.0135 - val_loss: 0.0067 - val_mae: 0.0581 - val_mse: 0.0135\n",
            "Epoch 106/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0068 - mae: 0.0566 - mse: 0.0135 - val_loss: 0.0067 - val_mae: 0.0581 - val_mse: 0.0135\n",
            "Epoch 107/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0566 - mse: 0.0135 - val_loss: 0.0067 - val_mae: 0.0580 - val_mse: 0.0135\n",
            "Epoch 108/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0565 - mse: 0.0135 - val_loss: 0.0067 - val_mae: 0.0580 - val_mse: 0.0135\n",
            "Epoch 109/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0565 - mse: 0.0135 - val_loss: 0.0067 - val_mae: 0.0580 - val_mse: 0.0135\n",
            "Epoch 110/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0565 - mse: 0.0135 - val_loss: 0.0067 - val_mae: 0.0579 - val_mse: 0.0135\n",
            "Epoch 111/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0564 - mse: 0.0135 - val_loss: 0.0067 - val_mae: 0.0579 - val_mse: 0.0135\n",
            "Epoch 112/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0564 - mse: 0.0135 - val_loss: 0.0067 - val_mae: 0.0579 - val_mse: 0.0135\n",
            "Epoch 113/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0564 - mse: 0.0135 - val_loss: 0.0067 - val_mae: 0.0578 - val_mse: 0.0134\n",
            "Epoch 114/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0563 - mse: 0.0135 - val_loss: 0.0067 - val_mae: 0.0578 - val_mse: 0.0134\n",
            "Epoch 115/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0563 - mse: 0.0135 - val_loss: 0.0067 - val_mae: 0.0578 - val_mse: 0.0134\n",
            "Epoch 116/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0563 - mse: 0.0134 - val_loss: 0.0067 - val_mae: 0.0578 - val_mse: 0.0134\n",
            "Epoch 117/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0562 - mse: 0.0134 - val_loss: 0.0067 - val_mae: 0.0577 - val_mse: 0.0134\n",
            "Epoch 118/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0562 - mse: 0.0134 - val_loss: 0.0067 - val_mae: 0.0577 - val_mse: 0.0134\n",
            "Epoch 119/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0562 - mse: 0.0134 - val_loss: 0.0067 - val_mae: 0.0577 - val_mse: 0.0134\n",
            "Epoch 120/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0562 - mse: 0.0134 - val_loss: 0.0067 - val_mae: 0.0576 - val_mse: 0.0134\n",
            "Epoch 121/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0561 - mse: 0.0134 - val_loss: 0.0067 - val_mae: 0.0576 - val_mse: 0.0134\n",
            "Epoch 122/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0561 - mse: 0.0134 - val_loss: 0.0067 - val_mae: 0.0576 - val_mse: 0.0134\n",
            "Epoch 123/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0561 - mse: 0.0134 - val_loss: 0.0067 - val_mae: 0.0576 - val_mse: 0.0134\n",
            "Epoch 124/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0561 - mse: 0.0134 - val_loss: 0.0067 - val_mae: 0.0575 - val_mse: 0.0134\n",
            "Epoch 125/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0560 - mse: 0.0134 - val_loss: 0.0067 - val_mae: 0.0575 - val_mse: 0.0134\n",
            "Epoch 126/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0560 - mse: 0.0134 - val_loss: 0.0067 - val_mae: 0.0575 - val_mse: 0.0134\n",
            "Epoch 127/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0560 - mse: 0.0134 - val_loss: 0.0067 - val_mae: 0.0575 - val_mse: 0.0134\n",
            "Epoch 128/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0560 - mse: 0.0134 - val_loss: 0.0067 - val_mae: 0.0575 - val_mse: 0.0134\n",
            "Epoch 129/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0559 - mse: 0.0134 - val_loss: 0.0067 - val_mae: 0.0574 - val_mse: 0.0134\n",
            "Epoch 130/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0559 - mse: 0.0134 - val_loss: 0.0067 - val_mae: 0.0574 - val_mse: 0.0134\n",
            "Epoch 131/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0559 - mse: 0.0134 - val_loss: 0.0067 - val_mae: 0.0574 - val_mse: 0.0134\n",
            "Epoch 132/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0559 - mse: 0.0134 - val_loss: 0.0067 - val_mae: 0.0574 - val_mse: 0.0134\n",
            "Epoch 133/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0558 - mse: 0.0134 - val_loss: 0.0067 - val_mae: 0.0573 - val_mse: 0.0134\n",
            "Epoch 134/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0558 - mse: 0.0134 - val_loss: 0.0067 - val_mae: 0.0573 - val_mse: 0.0134\n",
            "Epoch 135/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0558 - mse: 0.0133 - val_loss: 0.0067 - val_mae: 0.0573 - val_mse: 0.0134\n",
            "Epoch 136/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0558 - mse: 0.0133 - val_loss: 0.0067 - val_mae: 0.0573 - val_mse: 0.0133\n",
            "Epoch 137/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0557 - mse: 0.0133 - val_loss: 0.0067 - val_mae: 0.0573 - val_mse: 0.0133\n",
            "Epoch 138/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0557 - mse: 0.0133 - val_loss: 0.0067 - val_mae: 0.0572 - val_mse: 0.0133\n",
            "Epoch 139/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0557 - mse: 0.0133 - val_loss: 0.0067 - val_mae: 0.0572 - val_mse: 0.0133\n",
            "Epoch 140/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0557 - mse: 0.0133 - val_loss: 0.0067 - val_mae: 0.0572 - val_mse: 0.0133\n",
            "Epoch 141/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0557 - mse: 0.0133 - val_loss: 0.0067 - val_mae: 0.0572 - val_mse: 0.0133\n",
            "Epoch 142/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0556 - mse: 0.0133 - val_loss: 0.0067 - val_mae: 0.0572 - val_mse: 0.0133\n",
            "Epoch 143/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0556 - mse: 0.0133 - val_loss: 0.0067 - val_mae: 0.0571 - val_mse: 0.0133\n",
            "Epoch 144/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0556 - mse: 0.0133 - val_loss: 0.0067 - val_mae: 0.0571 - val_mse: 0.0133\n",
            "Epoch 145/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0067 - mae: 0.0556 - mse: 0.0133 - val_loss: 0.0067 - val_mae: 0.0571 - val_mse: 0.0133\n",
            "Epoch 146/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0556 - mse: 0.0133 - val_loss: 0.0067 - val_mae: 0.0571 - val_mse: 0.0133\n",
            "Epoch 147/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0555 - mse: 0.0133 - val_loss: 0.0067 - val_mae: 0.0571 - val_mse: 0.0133\n",
            "Epoch 148/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0555 - mse: 0.0133 - val_loss: 0.0066 - val_mae: 0.0570 - val_mse: 0.0133\n",
            "Epoch 149/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0555 - mse: 0.0133 - val_loss: 0.0066 - val_mae: 0.0570 - val_mse: 0.0133\n",
            "Epoch 150/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0555 - mse: 0.0133 - val_loss: 0.0066 - val_mae: 0.0570 - val_mse: 0.0133\n",
            "Epoch 151/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0555 - mse: 0.0133 - val_loss: 0.0066 - val_mae: 0.0570 - val_mse: 0.0133\n",
            "Epoch 152/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0554 - mse: 0.0133 - val_loss: 0.0066 - val_mae: 0.0570 - val_mse: 0.0133\n",
            "Epoch 153/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0554 - mse: 0.0133 - val_loss: 0.0066 - val_mae: 0.0570 - val_mse: 0.0133\n",
            "Epoch 154/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0554 - mse: 0.0133 - val_loss: 0.0066 - val_mae: 0.0569 - val_mse: 0.0133\n",
            "Epoch 155/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0554 - mse: 0.0133 - val_loss: 0.0066 - val_mae: 0.0569 - val_mse: 0.0133\n",
            "Epoch 156/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0554 - mse: 0.0133 - val_loss: 0.0066 - val_mae: 0.0569 - val_mse: 0.0133\n",
            "Epoch 157/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0553 - mse: 0.0133 - val_loss: 0.0066 - val_mae: 0.0569 - val_mse: 0.0133\n",
            "Epoch 158/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0553 - mse: 0.0132 - val_loss: 0.0066 - val_mae: 0.0569 - val_mse: 0.0133\n",
            "Epoch 159/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0553 - mse: 0.0132 - val_loss: 0.0066 - val_mae: 0.0569 - val_mse: 0.0133\n",
            "Epoch 160/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0066 - mae: 0.0553 - mse: 0.0132 - val_loss: 0.0066 - val_mae: 0.0568 - val_mse: 0.0132\n",
            "Epoch 161/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0553 - mse: 0.0132 - val_loss: 0.0066 - val_mae: 0.0568 - val_mse: 0.0132\n",
            "Epoch 162/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0552 - mse: 0.0132 - val_loss: 0.0066 - val_mae: 0.0568 - val_mse: 0.0132\n",
            "Epoch 163/500\n",
            "1349/1349 [==============================] - 20s 14ms/step - loss: 0.0066 - mae: 0.0552 - mse: 0.0132 - val_loss: 0.0066 - val_mae: 0.0568 - val_mse: 0.0132\n",
            "Epoch 164/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0552 - mse: 0.0132 - val_loss: 0.0066 - val_mae: 0.0568 - val_mse: 0.0132\n",
            "Epoch 165/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0552 - mse: 0.0132 - val_loss: 0.0066 - val_mae: 0.0568 - val_mse: 0.0132\n",
            "Epoch 166/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0066 - mae: 0.0552 - mse: 0.0132 - val_loss: 0.0066 - val_mae: 0.0567 - val_mse: 0.0132\n",
            "Epoch 167/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0552 - mse: 0.0132 - val_loss: 0.0066 - val_mae: 0.0567 - val_mse: 0.0132\n",
            "Epoch 168/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0551 - mse: 0.0132 - val_loss: 0.0066 - val_mae: 0.0567 - val_mse: 0.0132\n",
            "Epoch 169/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0551 - mse: 0.0132 - val_loss: 0.0066 - val_mae: 0.0567 - val_mse: 0.0132\n",
            "Epoch 170/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0066 - mae: 0.0551 - mse: 0.0132 - val_loss: 0.0066 - val_mae: 0.0567 - val_mse: 0.0132\n",
            "Epoch 171/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0551 - mse: 0.0132 - val_loss: 0.0066 - val_mae: 0.0567 - val_mse: 0.0132\n",
            "Epoch 172/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0551 - mse: 0.0132 - val_loss: 0.0066 - val_mae: 0.0567 - val_mse: 0.0132\n",
            "Epoch 173/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0551 - mse: 0.0132 - val_loss: 0.0066 - val_mae: 0.0566 - val_mse: 0.0132\n",
            "Epoch 174/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0550 - mse: 0.0132 - val_loss: 0.0066 - val_mae: 0.0566 - val_mse: 0.0132\n",
            "Epoch 175/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0550 - mse: 0.0132 - val_loss: 0.0066 - val_mae: 0.0566 - val_mse: 0.0132\n",
            "Epoch 176/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0550 - mse: 0.0132 - val_loss: 0.0066 - val_mae: 0.0566 - val_mse: 0.0132\n",
            "Epoch 177/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0550 - mse: 0.0132 - val_loss: 0.0066 - val_mae: 0.0566 - val_mse: 0.0132\n",
            "Epoch 178/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0550 - mse: 0.0132 - val_loss: 0.0066 - val_mae: 0.0566 - val_mse: 0.0132\n",
            "Epoch 179/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0549 - mse: 0.0132 - val_loss: 0.0066 - val_mae: 0.0565 - val_mse: 0.0132\n",
            "Epoch 180/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0549 - mse: 0.0132 - val_loss: 0.0066 - val_mae: 0.0565 - val_mse: 0.0132\n",
            "Epoch 181/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0549 - mse: 0.0131 - val_loss: 0.0066 - val_mae: 0.0565 - val_mse: 0.0132\n",
            "Epoch 182/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0549 - mse: 0.0131 - val_loss: 0.0066 - val_mae: 0.0565 - val_mse: 0.0132\n",
            "Epoch 183/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0549 - mse: 0.0131 - val_loss: 0.0066 - val_mae: 0.0565 - val_mse: 0.0132\n",
            "Epoch 184/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0549 - mse: 0.0131 - val_loss: 0.0066 - val_mae: 0.0565 - val_mse: 0.0132\n",
            "Epoch 185/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0066 - mae: 0.0548 - mse: 0.0131 - val_loss: 0.0066 - val_mae: 0.0565 - val_mse: 0.0131\n",
            "Epoch 186/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0548 - mse: 0.0131 - val_loss: 0.0066 - val_mae: 0.0564 - val_mse: 0.0131\n",
            "Epoch 187/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0548 - mse: 0.0131 - val_loss: 0.0066 - val_mae: 0.0564 - val_mse: 0.0131\n",
            "Epoch 188/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0548 - mse: 0.0131 - val_loss: 0.0066 - val_mae: 0.0564 - val_mse: 0.0131\n",
            "Epoch 189/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0548 - mse: 0.0131 - val_loss: 0.0066 - val_mae: 0.0564 - val_mse: 0.0131\n",
            "Epoch 190/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0548 - mse: 0.0131 - val_loss: 0.0066 - val_mae: 0.0564 - val_mse: 0.0131\n",
            "Epoch 191/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0547 - mse: 0.0131 - val_loss: 0.0066 - val_mae: 0.0564 - val_mse: 0.0131\n",
            "Epoch 192/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0066 - mae: 0.0547 - mse: 0.0131 - val_loss: 0.0066 - val_mae: 0.0564 - val_mse: 0.0131\n",
            "Epoch 193/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0065 - mae: 0.0547 - mse: 0.0131 - val_loss: 0.0066 - val_mae: 0.0563 - val_mse: 0.0131\n",
            "Epoch 194/500\n",
            "1349/1349 [==============================] - 20s 14ms/step - loss: 0.0065 - mae: 0.0547 - mse: 0.0131 - val_loss: 0.0066 - val_mae: 0.0563 - val_mse: 0.0131\n",
            "Epoch 195/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0065 - mae: 0.0547 - mse: 0.0131 - val_loss: 0.0066 - val_mae: 0.0563 - val_mse: 0.0131\n",
            "Epoch 196/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0065 - mae: 0.0547 - mse: 0.0131 - val_loss: 0.0066 - val_mae: 0.0563 - val_mse: 0.0131\n",
            "Epoch 197/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0065 - mae: 0.0546 - mse: 0.0131 - val_loss: 0.0066 - val_mae: 0.0563 - val_mse: 0.0131\n",
            "Epoch 198/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0065 - mae: 0.0546 - mse: 0.0131 - val_loss: 0.0065 - val_mae: 0.0563 - val_mse: 0.0131\n",
            "Epoch 199/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0065 - mae: 0.0546 - mse: 0.0131 - val_loss: 0.0065 - val_mae: 0.0563 - val_mse: 0.0131\n",
            "Epoch 200/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0065 - mae: 0.0546 - mse: 0.0131 - val_loss: 0.0065 - val_mae: 0.0562 - val_mse: 0.0131\n",
            "Epoch 201/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0065 - mae: 0.0546 - mse: 0.0131 - val_loss: 0.0065 - val_mae: 0.0562 - val_mse: 0.0131\n",
            "Epoch 202/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0065 - mae: 0.0546 - mse: 0.0131 - val_loss: 0.0065 - val_mae: 0.0562 - val_mse: 0.0131\n",
            "Epoch 203/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0065 - mae: 0.0545 - mse: 0.0131 - val_loss: 0.0065 - val_mae: 0.0562 - val_mse: 0.0131\n",
            "Epoch 204/500\n",
            "1349/1349 [==============================] - 20s 14ms/step - loss: 0.0065 - mae: 0.0545 - mse: 0.0131 - val_loss: 0.0065 - val_mae: 0.0562 - val_mse: 0.0131\n",
            "Epoch 205/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0065 - mae: 0.0545 - mse: 0.0130 - val_loss: 0.0065 - val_mae: 0.0562 - val_mse: 0.0131\n",
            "Epoch 206/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0065 - mae: 0.0545 - mse: 0.0130 - val_loss: 0.0065 - val_mae: 0.0561 - val_mse: 0.0131\n",
            "Epoch 207/500\n",
            "1349/1349 [==============================] - 20s 14ms/step - loss: 0.0065 - mae: 0.0545 - mse: 0.0130 - val_loss: 0.0065 - val_mae: 0.0561 - val_mse: 0.0131\n",
            "Epoch 208/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0065 - mae: 0.0545 - mse: 0.0130 - val_loss: 0.0065 - val_mae: 0.0561 - val_mse: 0.0131\n",
            "Epoch 209/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0065 - mae: 0.0544 - mse: 0.0130 - val_loss: 0.0065 - val_mae: 0.0561 - val_mse: 0.0131\n",
            "Epoch 210/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0065 - mae: 0.0544 - mse: 0.0130 - val_loss: 0.0065 - val_mae: 0.0561 - val_mse: 0.0130\n",
            "Epoch 211/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0065 - mae: 0.0544 - mse: 0.0130 - val_loss: 0.0065 - val_mae: 0.0561 - val_mse: 0.0130\n",
            "Epoch 212/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0065 - mae: 0.0544 - mse: 0.0130 - val_loss: 0.0065 - val_mae: 0.0561 - val_mse: 0.0130\n",
            "Epoch 213/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0065 - mae: 0.0544 - mse: 0.0130 - val_loss: 0.0065 - val_mae: 0.0560 - val_mse: 0.0130\n",
            "Epoch 214/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0065 - mae: 0.0543 - mse: 0.0130 - val_loss: 0.0065 - val_mae: 0.0560 - val_mse: 0.0130\n",
            "Epoch 215/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0065 - mae: 0.0543 - mse: 0.0130 - val_loss: 0.0065 - val_mae: 0.0560 - val_mse: 0.0130\n",
            "Epoch 216/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0065 - mae: 0.0543 - mse: 0.0130 - val_loss: 0.0065 - val_mae: 0.0560 - val_mse: 0.0130\n",
            "Epoch 217/500\n",
            "1349/1349 [==============================] - 20s 14ms/step - loss: 0.0065 - mae: 0.0543 - mse: 0.0130 - val_loss: 0.0065 - val_mae: 0.0560 - val_mse: 0.0130\n",
            "Epoch 218/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0065 - mae: 0.0543 - mse: 0.0130 - val_loss: 0.0065 - val_mae: 0.0560 - val_mse: 0.0130\n",
            "Epoch 219/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0065 - mae: 0.0543 - mse: 0.0130 - val_loss: 0.0065 - val_mae: 0.0559 - val_mse: 0.0130\n",
            "Epoch 220/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0065 - mae: 0.0542 - mse: 0.0130 - val_loss: 0.0065 - val_mae: 0.0559 - val_mse: 0.0130\n",
            "Epoch 221/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0065 - mae: 0.0542 - mse: 0.0130 - val_loss: 0.0065 - val_mae: 0.0559 - val_mse: 0.0130\n",
            "Epoch 222/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0065 - mae: 0.0542 - mse: 0.0130 - val_loss: 0.0065 - val_mae: 0.0559 - val_mse: 0.0130\n",
            "Epoch 223/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0065 - mae: 0.0542 - mse: 0.0130 - val_loss: 0.0065 - val_mae: 0.0559 - val_mse: 0.0130\n",
            "Epoch 224/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0065 - mae: 0.0542 - mse: 0.0130 - val_loss: 0.0065 - val_mae: 0.0559 - val_mse: 0.0130\n",
            "Epoch 225/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0065 - mae: 0.0541 - mse: 0.0130 - val_loss: 0.0065 - val_mae: 0.0558 - val_mse: 0.0130\n",
            "Epoch 226/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0065 - mae: 0.0541 - mse: 0.0129 - val_loss: 0.0065 - val_mae: 0.0558 - val_mse: 0.0130\n",
            "Epoch 227/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0065 - mae: 0.0541 - mse: 0.0129 - val_loss: 0.0065 - val_mae: 0.0558 - val_mse: 0.0130\n",
            "Epoch 228/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0065 - mae: 0.0541 - mse: 0.0129 - val_loss: 0.0065 - val_mae: 0.0558 - val_mse: 0.0130\n",
            "Epoch 229/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0065 - mae: 0.0541 - mse: 0.0129 - val_loss: 0.0065 - val_mae: 0.0558 - val_mse: 0.0130\n",
            "Epoch 230/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0065 - mae: 0.0541 - mse: 0.0129 - val_loss: 0.0065 - val_mae: 0.0558 - val_mse: 0.0130\n",
            "Epoch 231/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0065 - mae: 0.0540 - mse: 0.0129 - val_loss: 0.0065 - val_mae: 0.0557 - val_mse: 0.0130\n",
            "Epoch 232/500\n",
            "1349/1349 [==============================] - 20s 14ms/step - loss: 0.0065 - mae: 0.0540 - mse: 0.0129 - val_loss: 0.0065 - val_mae: 0.0557 - val_mse: 0.0129\n",
            "Epoch 233/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0065 - mae: 0.0540 - mse: 0.0129 - val_loss: 0.0065 - val_mae: 0.0557 - val_mse: 0.0129\n",
            "Epoch 234/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0065 - mae: 0.0540 - mse: 0.0129 - val_loss: 0.0065 - val_mae: 0.0557 - val_mse: 0.0129\n",
            "Epoch 235/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0065 - mae: 0.0540 - mse: 0.0129 - val_loss: 0.0065 - val_mae: 0.0557 - val_mse: 0.0129\n",
            "Epoch 236/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0064 - mae: 0.0539 - mse: 0.0129 - val_loss: 0.0065 - val_mae: 0.0556 - val_mse: 0.0129\n",
            "Epoch 237/500\n",
            "1349/1349 [==============================] - 20s 14ms/step - loss: 0.0064 - mae: 0.0539 - mse: 0.0129 - val_loss: 0.0065 - val_mae: 0.0556 - val_mse: 0.0129\n",
            "Epoch 238/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0064 - mae: 0.0539 - mse: 0.0129 - val_loss: 0.0065 - val_mae: 0.0556 - val_mse: 0.0129\n",
            "Epoch 239/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0064 - mae: 0.0539 - mse: 0.0129 - val_loss: 0.0065 - val_mae: 0.0556 - val_mse: 0.0129\n",
            "Epoch 240/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0064 - mae: 0.0539 - mse: 0.0129 - val_loss: 0.0065 - val_mae: 0.0556 - val_mse: 0.0129\n",
            "Epoch 241/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0064 - mae: 0.0538 - mse: 0.0129 - val_loss: 0.0065 - val_mae: 0.0556 - val_mse: 0.0129\n",
            "Epoch 242/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0064 - mae: 0.0538 - mse: 0.0129 - val_loss: 0.0064 - val_mae: 0.0555 - val_mse: 0.0129\n",
            "Epoch 243/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0064 - mae: 0.0538 - mse: 0.0129 - val_loss: 0.0064 - val_mae: 0.0555 - val_mse: 0.0129\n",
            "Epoch 244/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0064 - mae: 0.0538 - mse: 0.0129 - val_loss: 0.0064 - val_mae: 0.0555 - val_mse: 0.0129\n",
            "Epoch 245/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0064 - mae: 0.0538 - mse: 0.0128 - val_loss: 0.0064 - val_mae: 0.0555 - val_mse: 0.0129\n",
            "Epoch 246/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0064 - mae: 0.0537 - mse: 0.0128 - val_loss: 0.0064 - val_mae: 0.0555 - val_mse: 0.0129\n",
            "Epoch 247/500\n",
            "1349/1349 [==============================] - 20s 14ms/step - loss: 0.0064 - mae: 0.0537 - mse: 0.0128 - val_loss: 0.0064 - val_mae: 0.0554 - val_mse: 0.0129\n",
            "Epoch 248/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0064 - mae: 0.0537 - mse: 0.0128 - val_loss: 0.0064 - val_mae: 0.0554 - val_mse: 0.0129\n",
            "Epoch 249/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0064 - mae: 0.0537 - mse: 0.0128 - val_loss: 0.0064 - val_mae: 0.0554 - val_mse: 0.0129\n",
            "Epoch 250/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0064 - mae: 0.0537 - mse: 0.0128 - val_loss: 0.0064 - val_mae: 0.0554 - val_mse: 0.0129\n",
            "Epoch 251/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0064 - mae: 0.0536 - mse: 0.0128 - val_loss: 0.0064 - val_mae: 0.0554 - val_mse: 0.0129\n",
            "Epoch 252/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0064 - mae: 0.0536 - mse: 0.0128 - val_loss: 0.0064 - val_mae: 0.0554 - val_mse: 0.0128\n",
            "Epoch 253/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0064 - mae: 0.0536 - mse: 0.0128 - val_loss: 0.0064 - val_mae: 0.0554 - val_mse: 0.0128\n",
            "Epoch 254/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0064 - mae: 0.0536 - mse: 0.0128 - val_loss: 0.0064 - val_mae: 0.0553 - val_mse: 0.0128\n",
            "Epoch 255/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0064 - mae: 0.0536 - mse: 0.0128 - val_loss: 0.0064 - val_mae: 0.0553 - val_mse: 0.0128\n",
            "Epoch 256/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0064 - mae: 0.0535 - mse: 0.0128 - val_loss: 0.0064 - val_mae: 0.0553 - val_mse: 0.0128\n",
            "Epoch 257/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0064 - mae: 0.0535 - mse: 0.0128 - val_loss: 0.0064 - val_mae: 0.0553 - val_mse: 0.0128\n",
            "Epoch 258/500\n",
            "1349/1349 [==============================] - 20s 14ms/step - loss: 0.0064 - mae: 0.0535 - mse: 0.0128 - val_loss: 0.0064 - val_mae: 0.0553 - val_mse: 0.0128\n",
            "Epoch 259/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0064 - mae: 0.0535 - mse: 0.0128 - val_loss: 0.0064 - val_mae: 0.0553 - val_mse: 0.0128\n",
            "Epoch 260/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0064 - mae: 0.0535 - mse: 0.0128 - val_loss: 0.0064 - val_mae: 0.0553 - val_mse: 0.0128\n",
            "Epoch 261/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0064 - mae: 0.0534 - mse: 0.0128 - val_loss: 0.0064 - val_mae: 0.0552 - val_mse: 0.0128\n",
            "Epoch 262/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0064 - mae: 0.0534 - mse: 0.0128 - val_loss: 0.0064 - val_mae: 0.0552 - val_mse: 0.0128\n",
            "Epoch 263/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0064 - mae: 0.0534 - mse: 0.0128 - val_loss: 0.0064 - val_mae: 0.0552 - val_mse: 0.0128\n",
            "Epoch 264/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0064 - mae: 0.0534 - mse: 0.0128 - val_loss: 0.0064 - val_mae: 0.0552 - val_mse: 0.0128\n",
            "Epoch 265/500\n",
            "1349/1349 [==============================] - 20s 14ms/step - loss: 0.0064 - mae: 0.0534 - mse: 0.0128 - val_loss: 0.0064 - val_mae: 0.0552 - val_mse: 0.0128\n",
            "Epoch 266/500\n",
            "1349/1349 [==============================] - 20s 14ms/step - loss: 0.0064 - mae: 0.0534 - mse: 0.0127 - val_loss: 0.0064 - val_mae: 0.0552 - val_mse: 0.0128\n",
            "Epoch 267/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0064 - mae: 0.0533 - mse: 0.0127 - val_loss: 0.0064 - val_mae: 0.0552 - val_mse: 0.0128\n",
            "Epoch 268/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0064 - mae: 0.0533 - mse: 0.0127 - val_loss: 0.0064 - val_mae: 0.0552 - val_mse: 0.0128\n",
            "Epoch 269/500\n",
            "1349/1349 [==============================] - 19s 14ms/step - loss: 0.0064 - mae: 0.0533 - mse: 0.0127 - val_loss: 0.0064 - val_mae: 0.0552 - val_mse: 0.0128\n",
            "Epoch 270/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0064 - mae: 0.0533 - mse: 0.0127 - val_loss: 0.0064 - val_mae: 0.0551 - val_mse: 0.0128\n",
            "Epoch 271/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0064 - mae: 0.0533 - mse: 0.0127 - val_loss: 0.0064 - val_mae: 0.0551 - val_mse: 0.0128\n",
            "Epoch 272/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0064 - mae: 0.0533 - mse: 0.0127 - val_loss: 0.0064 - val_mae: 0.0551 - val_mse: 0.0128\n",
            "Epoch 273/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0064 - mae: 0.0532 - mse: 0.0127 - val_loss: 0.0064 - val_mae: 0.0551 - val_mse: 0.0128\n",
            "Epoch 274/500\n",
            "1349/1349 [==============================] - 21s 15ms/step - loss: 0.0064 - mae: 0.0532 - mse: 0.0127 - val_loss: 0.0064 - val_mae: 0.0551 - val_mse: 0.0128\n",
            "Epoch 275/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0064 - mae: 0.0532 - mse: 0.0127 - val_loss: 0.0064 - val_mae: 0.0551 - val_mse: 0.0128\n",
            "Epoch 276/500\n",
            "1349/1349 [==============================] - 21s 15ms/step - loss: 0.0064 - mae: 0.0532 - mse: 0.0127 - val_loss: 0.0064 - val_mae: 0.0551 - val_mse: 0.0128\n",
            "Epoch 277/500\n",
            "1349/1349 [==============================] - 21s 15ms/step - loss: 0.0064 - mae: 0.0532 - mse: 0.0127 - val_loss: 0.0064 - val_mae: 0.0551 - val_mse: 0.0128\n",
            "Epoch 278/500\n",
            "1349/1349 [==============================] - 21s 15ms/step - loss: 0.0064 - mae: 0.0532 - mse: 0.0127 - val_loss: 0.0064 - val_mae: 0.0551 - val_mse: 0.0128\n",
            "Epoch 279/500\n",
            "1349/1349 [==============================] - 21s 15ms/step - loss: 0.0064 - mae: 0.0532 - mse: 0.0127 - val_loss: 0.0064 - val_mae: 0.0551 - val_mse: 0.0127\n",
            "Epoch 280/500\n",
            "1349/1349 [==============================] - 21s 15ms/step - loss: 0.0063 - mae: 0.0532 - mse: 0.0127 - val_loss: 0.0064 - val_mae: 0.0551 - val_mse: 0.0127\n",
            "Epoch 281/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0531 - mse: 0.0127 - val_loss: 0.0064 - val_mae: 0.0550 - val_mse: 0.0127\n",
            "Epoch 282/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0531 - mse: 0.0127 - val_loss: 0.0064 - val_mae: 0.0550 - val_mse: 0.0127\n",
            "Epoch 283/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0531 - mse: 0.0127 - val_loss: 0.0064 - val_mae: 0.0550 - val_mse: 0.0127\n",
            "Epoch 284/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0531 - mse: 0.0127 - val_loss: 0.0064 - val_mae: 0.0550 - val_mse: 0.0127\n",
            "Epoch 285/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0531 - mse: 0.0127 - val_loss: 0.0064 - val_mae: 0.0550 - val_mse: 0.0127\n",
            "Epoch 286/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0531 - mse: 0.0127 - val_loss: 0.0064 - val_mae: 0.0550 - val_mse: 0.0127\n",
            "Epoch 287/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0531 - mse: 0.0127 - val_loss: 0.0064 - val_mae: 0.0550 - val_mse: 0.0127\n",
            "Epoch 288/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0531 - mse: 0.0127 - val_loss: 0.0064 - val_mae: 0.0550 - val_mse: 0.0127\n",
            "Epoch 289/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0530 - mse: 0.0127 - val_loss: 0.0064 - val_mae: 0.0550 - val_mse: 0.0127\n",
            "Epoch 290/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0530 - mse: 0.0127 - val_loss: 0.0064 - val_mae: 0.0550 - val_mse: 0.0127\n",
            "Epoch 291/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0530 - mse: 0.0127 - val_loss: 0.0064 - val_mae: 0.0549 - val_mse: 0.0127\n",
            "Epoch 292/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0530 - mse: 0.0127 - val_loss: 0.0064 - val_mae: 0.0549 - val_mse: 0.0127\n",
            "Epoch 293/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0530 - mse: 0.0127 - val_loss: 0.0064 - val_mae: 0.0549 - val_mse: 0.0127\n",
            "Epoch 294/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0530 - mse: 0.0127 - val_loss: 0.0064 - val_mae: 0.0549 - val_mse: 0.0127\n",
            "Epoch 295/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0530 - mse: 0.0127 - val_loss: 0.0064 - val_mae: 0.0549 - val_mse: 0.0127\n",
            "Epoch 296/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0530 - mse: 0.0127 - val_loss: 0.0064 - val_mae: 0.0549 - val_mse: 0.0127\n",
            "Epoch 297/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0530 - mse: 0.0127 - val_loss: 0.0064 - val_mae: 0.0549 - val_mse: 0.0127\n",
            "Epoch 298/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0530 - mse: 0.0127 - val_loss: 0.0064 - val_mae: 0.0549 - val_mse: 0.0127\n",
            "Epoch 299/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0529 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0549 - val_mse: 0.0127\n",
            "Epoch 300/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0529 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0549 - val_mse: 0.0127\n",
            "Epoch 301/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0529 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0548 - val_mse: 0.0127\n",
            "Epoch 302/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0529 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0548 - val_mse: 0.0127\n",
            "Epoch 303/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0529 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0548 - val_mse: 0.0127\n",
            "Epoch 304/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0529 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0548 - val_mse: 0.0127\n",
            "Epoch 305/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0529 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0548 - val_mse: 0.0127\n",
            "Epoch 306/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0529 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0548 - val_mse: 0.0127\n",
            "Epoch 307/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0529 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0548 - val_mse: 0.0127\n",
            "Epoch 308/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0529 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0548 - val_mse: 0.0127\n",
            "Epoch 309/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0528 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0548 - val_mse: 0.0127\n",
            "Epoch 310/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0528 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0548 - val_mse: 0.0127\n",
            "Epoch 311/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0528 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0548 - val_mse: 0.0127\n",
            "Epoch 312/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0528 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0547 - val_mse: 0.0127\n",
            "Epoch 313/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0528 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0547 - val_mse: 0.0127\n",
            "Epoch 314/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0528 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0547 - val_mse: 0.0127\n",
            "Epoch 315/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0528 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0547 - val_mse: 0.0127\n",
            "Epoch 316/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0528 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0547 - val_mse: 0.0127\n",
            "Epoch 317/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0528 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0547 - val_mse: 0.0127\n",
            "Epoch 318/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0528 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0547 - val_mse: 0.0127\n",
            "Epoch 319/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0528 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0547 - val_mse: 0.0127\n",
            "Epoch 320/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0527 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0547 - val_mse: 0.0127\n",
            "Epoch 321/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0527 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0547 - val_mse: 0.0127\n",
            "Epoch 322/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0527 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0547 - val_mse: 0.0127\n",
            "Epoch 323/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0527 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0547 - val_mse: 0.0127\n",
            "Epoch 324/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0527 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0547 - val_mse: 0.0127\n",
            "Epoch 325/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0527 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0546 - val_mse: 0.0126\n",
            "Epoch 326/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0527 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0546 - val_mse: 0.0126\n",
            "Epoch 327/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0527 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0546 - val_mse: 0.0126\n",
            "Epoch 328/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0527 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0546 - val_mse: 0.0126\n",
            "Epoch 329/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0527 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0546 - val_mse: 0.0126\n",
            "Epoch 330/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0527 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0546 - val_mse: 0.0126\n",
            "Epoch 331/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0527 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0546 - val_mse: 0.0126\n",
            "Epoch 332/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0527 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0546 - val_mse: 0.0126\n",
            "Epoch 333/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0526 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0546 - val_mse: 0.0126\n",
            "Epoch 334/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0526 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0546 - val_mse: 0.0126\n",
            "Epoch 335/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0526 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0546 - val_mse: 0.0126\n",
            "Epoch 336/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0526 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0546 - val_mse: 0.0126\n",
            "Epoch 337/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0526 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0546 - val_mse: 0.0126\n",
            "Epoch 338/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0526 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0546 - val_mse: 0.0126\n",
            "Epoch 339/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0526 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0545 - val_mse: 0.0126\n",
            "Epoch 340/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0526 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0545 - val_mse: 0.0126\n",
            "Epoch 341/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0526 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0545 - val_mse: 0.0126\n",
            "Epoch 342/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0526 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0545 - val_mse: 0.0126\n",
            "Epoch 343/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0526 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0545 - val_mse: 0.0126\n",
            "Epoch 344/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0526 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0545 - val_mse: 0.0126\n",
            "Epoch 345/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0526 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0545 - val_mse: 0.0126\n",
            "Epoch 346/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0526 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0545 - val_mse: 0.0126\n",
            "Epoch 347/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0525 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0545 - val_mse: 0.0126\n",
            "Epoch 348/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0525 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0545 - val_mse: 0.0126\n",
            "Epoch 349/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0525 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0545 - val_mse: 0.0126\n",
            "Epoch 350/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0525 - mse: 0.0126 - val_loss: 0.0063 - val_mae: 0.0545 - val_mse: 0.0126\n",
            "Epoch 351/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0525 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0545 - val_mse: 0.0126\n",
            "Epoch 352/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0525 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0545 - val_mse: 0.0126\n",
            "Epoch 353/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0525 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0545 - val_mse: 0.0126\n",
            "Epoch 354/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0525 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0545 - val_mse: 0.0126\n",
            "Epoch 355/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0525 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0544 - val_mse: 0.0126\n",
            "Epoch 356/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0525 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0544 - val_mse: 0.0126\n",
            "Epoch 357/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0525 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0544 - val_mse: 0.0126\n",
            "Epoch 358/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0525 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0544 - val_mse: 0.0126\n",
            "Epoch 359/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0525 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0544 - val_mse: 0.0126\n",
            "Epoch 360/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0525 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0544 - val_mse: 0.0126\n",
            "Epoch 361/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0525 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0544 - val_mse: 0.0126\n",
            "Epoch 362/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0524 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0544 - val_mse: 0.0126\n",
            "Epoch 363/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0524 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0544 - val_mse: 0.0126\n",
            "Epoch 364/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0524 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0544 - val_mse: 0.0126\n",
            "Epoch 365/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0524 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0544 - val_mse: 0.0126\n",
            "Epoch 366/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0524 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0544 - val_mse: 0.0126\n",
            "Epoch 367/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0524 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0544 - val_mse: 0.0126\n",
            "Epoch 368/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0524 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0544 - val_mse: 0.0126\n",
            "Epoch 369/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0524 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0544 - val_mse: 0.0126\n",
            "Epoch 370/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0524 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0544 - val_mse: 0.0126\n",
            "Epoch 371/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0524 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0544 - val_mse: 0.0126\n",
            "Epoch 372/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0524 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0544 - val_mse: 0.0126\n",
            "Epoch 373/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0524 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0544 - val_mse: 0.0126\n",
            "Epoch 374/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0524 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0543 - val_mse: 0.0126\n",
            "Epoch 375/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0524 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0543 - val_mse: 0.0126\n",
            "Epoch 376/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0524 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0543 - val_mse: 0.0126\n",
            "Epoch 377/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0524 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0543 - val_mse: 0.0126\n",
            "Epoch 378/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0523 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0543 - val_mse: 0.0126\n",
            "Epoch 379/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0523 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0543 - val_mse: 0.0126\n",
            "Epoch 380/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0523 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0543 - val_mse: 0.0126\n",
            "Epoch 381/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0523 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0543 - val_mse: 0.0126\n",
            "Epoch 382/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0523 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0543 - val_mse: 0.0126\n",
            "Epoch 383/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0063 - mae: 0.0523 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0543 - val_mse: 0.0126\n",
            "Epoch 384/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0063 - mae: 0.0523 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0543 - val_mse: 0.0126\n",
            "Epoch 385/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0523 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0543 - val_mse: 0.0126\n",
            "Epoch 386/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0523 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0543 - val_mse: 0.0126\n",
            "Epoch 387/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0523 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0543 - val_mse: 0.0126\n",
            "Epoch 388/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0523 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0543 - val_mse: 0.0125\n",
            "Epoch 389/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0523 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0543 - val_mse: 0.0125\n",
            "Epoch 390/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0523 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0543 - val_mse: 0.0125\n",
            "Epoch 391/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0523 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0543 - val_mse: 0.0125\n",
            "Epoch 392/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0523 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0543 - val_mse: 0.0125\n",
            "Epoch 393/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0523 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0543 - val_mse: 0.0125\n",
            "Epoch 394/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0523 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0543 - val_mse: 0.0125\n",
            "Epoch 395/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0523 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0542 - val_mse: 0.0125\n",
            "Epoch 396/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0522 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0542 - val_mse: 0.0125\n",
            "Epoch 397/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0522 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0542 - val_mse: 0.0125\n",
            "Epoch 398/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0522 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0542 - val_mse: 0.0125\n",
            "Epoch 399/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0522 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0542 - val_mse: 0.0125\n",
            "Epoch 400/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0522 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0542 - val_mse: 0.0125\n",
            "Epoch 401/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0522 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0542 - val_mse: 0.0125\n",
            "Epoch 402/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0522 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0542 - val_mse: 0.0125\n",
            "Epoch 403/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0522 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0542 - val_mse: 0.0125\n",
            "Epoch 404/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0522 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0542 - val_mse: 0.0125\n",
            "Epoch 405/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0522 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0542 - val_mse: 0.0125\n",
            "Epoch 406/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0522 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0542 - val_mse: 0.0125\n",
            "Epoch 407/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0522 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0542 - val_mse: 0.0125\n",
            "Epoch 408/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0522 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0542 - val_mse: 0.0125\n",
            "Epoch 409/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0522 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0542 - val_mse: 0.0125\n",
            "Epoch 410/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0522 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0542 - val_mse: 0.0125\n",
            "Epoch 411/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0522 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0542 - val_mse: 0.0125\n",
            "Epoch 412/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0062 - mae: 0.0522 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0542 - val_mse: 0.0125\n",
            "Epoch 413/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0522 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0542 - val_mse: 0.0125\n",
            "Epoch 414/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0521 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0542 - val_mse: 0.0125\n",
            "Epoch 415/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0521 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0542 - val_mse: 0.0125\n",
            "Epoch 416/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0521 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0542 - val_mse: 0.0125\n",
            "Epoch 417/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0521 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0542 - val_mse: 0.0125\n",
            "Epoch 418/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0521 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0541 - val_mse: 0.0125\n",
            "Epoch 419/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0521 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0541 - val_mse: 0.0125\n",
            "Epoch 420/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0521 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0541 - val_mse: 0.0125\n",
            "Epoch 421/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0521 - mse: 0.0125 - val_loss: 0.0063 - val_mae: 0.0541 - val_mse: 0.0125\n",
            "Epoch 422/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0521 - mse: 0.0124 - val_loss: 0.0063 - val_mae: 0.0541 - val_mse: 0.0125\n",
            "Epoch 423/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0521 - mse: 0.0124 - val_loss: 0.0063 - val_mae: 0.0541 - val_mse: 0.0125\n",
            "Epoch 424/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0521 - mse: 0.0124 - val_loss: 0.0063 - val_mae: 0.0541 - val_mse: 0.0125\n",
            "Epoch 425/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0521 - mse: 0.0124 - val_loss: 0.0063 - val_mae: 0.0541 - val_mse: 0.0125\n",
            "Epoch 426/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0062 - mae: 0.0521 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0541 - val_mse: 0.0125\n",
            "Epoch 427/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0521 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0541 - val_mse: 0.0125\n",
            "Epoch 428/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0521 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0541 - val_mse: 0.0125\n",
            "Epoch 429/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0521 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0541 - val_mse: 0.0125\n",
            "Epoch 430/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0521 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0541 - val_mse: 0.0125\n",
            "Epoch 431/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0521 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0541 - val_mse: 0.0125\n",
            "Epoch 432/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0521 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0541 - val_mse: 0.0125\n",
            "Epoch 433/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0520 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0541 - val_mse: 0.0125\n",
            "Epoch 434/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0520 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0541 - val_mse: 0.0125\n",
            "Epoch 435/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0520 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0541 - val_mse: 0.0125\n",
            "Epoch 436/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0520 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0541 - val_mse: 0.0125\n",
            "Epoch 437/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0520 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0541 - val_mse: 0.0125\n",
            "Epoch 438/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0062 - mae: 0.0520 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0541 - val_mse: 0.0125\n",
            "Epoch 439/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0520 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0541 - val_mse: 0.0125\n",
            "Epoch 440/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0520 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0541 - val_mse: 0.0125\n",
            "Epoch 441/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0520 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0541 - val_mse: 0.0125\n",
            "Epoch 442/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0520 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0541 - val_mse: 0.0125\n",
            "Epoch 443/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0062 - mae: 0.0520 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0541 - val_mse: 0.0125\n",
            "Epoch 444/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0520 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0541 - val_mse: 0.0125\n",
            "Epoch 445/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0520 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0541 - val_mse: 0.0125\n",
            "Epoch 446/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0520 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0540 - val_mse: 0.0125\n",
            "Epoch 447/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0520 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0540 - val_mse: 0.0125\n",
            "Epoch 448/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0520 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0540 - val_mse: 0.0125\n",
            "Epoch 449/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0062 - mae: 0.0520 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0540 - val_mse: 0.0125\n",
            "Epoch 450/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0520 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0540 - val_mse: 0.0125\n",
            "Epoch 451/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0520 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0540 - val_mse: 0.0125\n",
            "Epoch 452/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0062 - mae: 0.0519 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0540 - val_mse: 0.0125\n",
            "Epoch 453/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0519 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0540 - val_mse: 0.0125\n",
            "Epoch 454/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0519 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0540 - val_mse: 0.0125\n",
            "Epoch 455/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0519 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0540 - val_mse: 0.0125\n",
            "Epoch 456/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0519 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0540 - val_mse: 0.0125\n",
            "Epoch 457/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0519 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0540 - val_mse: 0.0125\n",
            "Epoch 458/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0519 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0540 - val_mse: 0.0125\n",
            "Epoch 459/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0062 - mae: 0.0519 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0540 - val_mse: 0.0125\n",
            "Epoch 460/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0519 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0540 - val_mse: 0.0125\n",
            "Epoch 461/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0519 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0540 - val_mse: 0.0125\n",
            "Epoch 462/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0519 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0540 - val_mse: 0.0125\n",
            "Epoch 463/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0062 - mae: 0.0519 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0540 - val_mse: 0.0125\n",
            "Epoch 464/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0062 - mae: 0.0519 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0540 - val_mse: 0.0125\n",
            "Epoch 465/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0519 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0540 - val_mse: 0.0125\n",
            "Epoch 466/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0519 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0540 - val_mse: 0.0125\n",
            "Epoch 467/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0519 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0540 - val_mse: 0.0125\n",
            "Epoch 468/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0062 - mae: 0.0519 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0540 - val_mse: 0.0124\n",
            "Epoch 469/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0519 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0540 - val_mse: 0.0124\n",
            "Epoch 470/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0519 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0540 - val_mse: 0.0124\n",
            "Epoch 471/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0519 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0540 - val_mse: 0.0124\n",
            "Epoch 472/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0519 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0540 - val_mse: 0.0124\n",
            "Epoch 473/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0518 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0540 - val_mse: 0.0124\n",
            "Epoch 474/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0518 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0540 - val_mse: 0.0124\n",
            "Epoch 475/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0518 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0540 - val_mse: 0.0124\n",
            "Epoch 476/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0518 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0540 - val_mse: 0.0124\n",
            "Epoch 477/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0518 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0540 - val_mse: 0.0124\n",
            "Epoch 478/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0518 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0539 - val_mse: 0.0124\n",
            "Epoch 479/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0518 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0539 - val_mse: 0.0124\n",
            "Epoch 480/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0518 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0539 - val_mse: 0.0124\n",
            "Epoch 481/500\n",
            "1349/1349 [==============================] - 21s 15ms/step - loss: 0.0062 - mae: 0.0518 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0539 - val_mse: 0.0124\n",
            "Epoch 482/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0062 - mae: 0.0518 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0539 - val_mse: 0.0124\n",
            "Epoch 483/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0062 - mae: 0.0518 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0539 - val_mse: 0.0124\n",
            "Epoch 484/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0062 - mae: 0.0518 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0539 - val_mse: 0.0124\n",
            "Epoch 485/500\n",
            "1349/1349 [==============================] - 21s 15ms/step - loss: 0.0062 - mae: 0.0518 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0539 - val_mse: 0.0124\n",
            "Epoch 486/500\n",
            "1349/1349 [==============================] - 21s 15ms/step - loss: 0.0062 - mae: 0.0518 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0539 - val_mse: 0.0124\n",
            "Epoch 487/500\n",
            "1349/1349 [==============================] - 21s 15ms/step - loss: 0.0062 - mae: 0.0518 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0539 - val_mse: 0.0124\n",
            "Epoch 488/500\n",
            "1349/1349 [==============================] - 21s 15ms/step - loss: 0.0062 - mae: 0.0518 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0539 - val_mse: 0.0124\n",
            "Epoch 489/500\n",
            "1349/1349 [==============================] - 21s 15ms/step - loss: 0.0062 - mae: 0.0518 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0539 - val_mse: 0.0124\n",
            "Epoch 490/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0062 - mae: 0.0518 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0539 - val_mse: 0.0124\n",
            "Epoch 491/500\n",
            "1349/1349 [==============================] - 20s 15ms/step - loss: 0.0062 - mae: 0.0518 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0539 - val_mse: 0.0124\n",
            "Epoch 492/500\n",
            "1349/1349 [==============================] - 21s 15ms/step - loss: 0.0062 - mae: 0.0518 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0539 - val_mse: 0.0124\n",
            "Epoch 493/500\n",
            "1349/1349 [==============================] - 21s 15ms/step - loss: 0.0062 - mae: 0.0518 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0539 - val_mse: 0.0124\n",
            "Epoch 494/500\n",
            "1349/1349 [==============================] - 21s 15ms/step - loss: 0.0062 - mae: 0.0517 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0539 - val_mse: 0.0124\n",
            "Epoch 495/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0062 - mae: 0.0517 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0539 - val_mse: 0.0124\n",
            "Epoch 496/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0062 - mae: 0.0517 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0539 - val_mse: 0.0124\n",
            "Epoch 497/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0062 - mae: 0.0517 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0539 - val_mse: 0.0124\n",
            "Epoch 498/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0517 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0539 - val_mse: 0.0124\n",
            "Epoch 499/500\n",
            "1349/1349 [==============================] - 21s 16ms/step - loss: 0.0062 - mae: 0.0517 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0539 - val_mse: 0.0124\n",
            "Epoch 500/500\n",
            "1349/1349 [==============================] - 22s 16ms/step - loss: 0.0062 - mae: 0.0517 - mse: 0.0124 - val_loss: 0.0062 - val_mae: 0.0539 - val_mse: 0.0124\n",
            "\u001b[92mTraining end, used 2:50:06.521910\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DnlAjypE0ZLu",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 806
        },
        "outputId": "d99b7a3f-eee6-4188-b826-49d43e66e636"
      },
      "source": [
        "plot_graphs(train_history, \"loss\")\n",
        "plot_graphs(train_history, \"mae\")\n",
        "plot_graphs(train_history, \"mse\") "
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-06iUL2AIhha"
      },
      "source": [
        "val_model = tf.keras.models.load_model(save_model)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "l02VwQ8TJv2t",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "3381ac47-98a9-47bb-e51f-5a5a2dd2348e"
      },
      "source": [
        "window_size_forecast = N_PAST\n",
        "batch_size_forecast = BATCH_SIZE\n",
        "\n",
        " \n",
        "\n",
        "forcast_val = model_forecast(val_model, \n",
        "                             data[split_time-window_size_forecast:], \n",
        "                             batch_size = batch_size_forecast, \n",
        "                             window_size_forecast = window_size_forecast,\n",
        "                             shift = window_size_forecast)\n",
        "forcast_val = forcast_val[:-1]\n",
        "forcast_val = forcast_val.reshape((forcast_val.shape[0]*forcast_val.shape[1]), forcast_val.shape[2])\n",
        "print(\"forcast_val.shape:\")\n",
        "print(forcast_val.shape)\n",
        "print()\n",
        "\n",
        "print(\"x_valid.shape:\")\n",
        "print(x_valid.shape)\n",
        "print()\n",
        "\n",
        "print(\"split_time\")\n",
        "print(split_time)\n",
        "print()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "forcast_val.shape:\n",
            "(43200, 7)\n",
            "\n",
            "x_valid.shape:\n",
            "(43200, 7)\n",
            "\n",
            "split_time\n",
            "43200\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "c6QXShFPKMO2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "6d68f902-d7e1-45f6-92b0-f40e84aff087"
      },
      "source": [
        "time_steps_valid = np.array([i for i in range(split_time)])\n",
        "\n",
        "for i in range(7):\n",
        "  print(f\"\\n\\033[92mfeature: {i}:\\n\")\n",
        "  feature_forecast = forcast_val[:, i]\n",
        "\n",
        "  print(\"feature_forecast.shape:\")\n",
        "  print(feature_forecast.shape)\n",
        "  \n",
        "  feature_valid = x_valid[:, i]\n",
        "\n",
        "  print(\"feature_valid.shape:\")\n",
        "  print(feature_valid.shape)\n",
        "\n",
        "  print(f\"\\033[92mMAE: {tf.keras.metrics.mean_absolute_error(feature_valid, feature_forecast).numpy()}\")\n",
        "  print(f\"\\033[92mMSE: {tf.keras.metrics.mean_squared_error(feature_valid, feature_forecast).numpy()}\")\n",
        "\n",
        "  plt.figure(figsize=(10, 6))\n",
        "\n",
        "  plot_series(time_steps_valid, feature_valid)\n",
        "  plot_series(time_steps_valid, feature_forecast) \n",
        "\n",
        "  plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "\u001b[92mfeature: 0:\n",
            "\n",
            "feature_forecast.shape:\n",
            "(43200,)\n",
            "feature_valid.shape:\n",
            "(43200,)\n",
            "\u001b[92mMAE: 0.0535602867603302\n",
            "\u001b[92mMSE: 0.0071175540797412395\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "\u001b[92mfeature: 1:\n",
            "\n",
            "feature_forecast.shape:\n",
            "(43200,)\n",
            "feature_valid.shape:\n",
            "(43200,)\n",
            "\u001b[92mMAE: 0.09775446355342865\n",
            "\u001b[92mMSE: 0.015018971636891365\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "\u001b[92mfeature: 2:\n",
            "\n",
            "feature_forecast.shape:\n",
            "(43200,)\n",
            "feature_valid.shape:\n",
            "(43200,)\n",
            "\u001b[92mMAE: 0.009037098847329617\n",
            "\u001b[92mMSE: 0.00013285366003401577\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAm4AAAFzCAYAAACHCIXLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOzdd3zbV7n48c/RtLxnnB1nrzZdadIdly7KaOmFlhYKhfZS5g8u/IDLhcu+/C7lMi6jjEIppVA6KS2ldNdtOtJmNXs0w0nsxPEekrV1fn8cSbZs2ZZsy5bt5/165RVb+ko6/tqSHj3nOc9RWmuEEEIIIUT2s4z3AIQQQgghRGokcBNCCCGEmCAkcBNCCCGEmCAkcBNCCCGEmCAkcBNCCCGEmCAkcBNCCCGEmCBs4z2A0VJeXq6rqqoy/jgej4e8vLyMP85kIOcqPXK+0iPnKz1yvlIn5yo9cr7SEztfmzdvbtZaV6R7+0kTuFVVVbFp06aMP05NTQ3V1dUZf5zJQM5VeuR8pUfOV3rkfKVOzlV65HylJ3a+lFJHhnN7mSoVQgghhJggJHATQgghhJggJHATQgghhJggJHATQgghhJggJHATQgghhJggJHATQgghhJggJHATQgghhJggJHATQgghhJggJHATQgghhJggMhq4KaXerpTap5Q6oJT6SpLrnUqp+6PXv66Uqope7lBK3aWU2qGU2qaUqs7kOIUQQgghJoKMBW5KKStwO3AlsAK4QSm1os9htwBtWutFwE+A26KXfwxAa30qcBnwI6WUZAeFEEIIMaVlMhhaAxzQWh/SWgeA+4Cr+xxzNXB39OuHgEuUUgoT6D0PoLVuBNqB1RkcqxBCCCEmufp2L52+4HgPY0QyGbjNAo71+r4uelnSY7TWIaADKAO2AVcppWxKqfnAWcCcDI5VCCGEEJPcB367gc/f9+Z4D2NEbOM9gAH8HlgObAKOAK8C4b4HKaVuBW4FqKyspKamJuMDc7vdY/I4k4Gcq/TI+UqPnK/0yPlKnZyr9EyU8+ULaY60dHOkpZv7/vE80/PGpwJrpOcrk4FbPYlZstnRy5IdU6eUsgFFQIvWWgOfjx2klHoV2N/3AbTWdwB3AKxevVpXV1eP5viTqqmpYSweZzKQc5UeOV/pkfOVHjlfqZNzlZ6Jcr521nfAsy8DsCc0jeurTxmXcYz0fGUy3NwILFZKzVdKOYDrgcf6HPMYcFP06/cBz2uttVIqVymVB6CUugwIaa13Z3CsQgghhJjEDjd7ADh1VhEPbq7D7Q+N84iGJ2OBW7Rm7TPAU8Ae4AGt9S6l1HeUUldFD7sTKFNKHQC+AMRahkwDtiil9gD/DnwoU+MUQgghxOR3qMkEbjdfUEV3IMyRFs84j2h4MlrjprV+Aniiz2Xf6PW1D7g2ye1qgaWZHJsQQgghpo7DzW5mFbuYXZILQIs7MM4jGh7pjSaEEEKISe9ws4cFFXmU5TkAaPVI4CaEEEIIkXW01hxq8jC/PI+yPCcAzW7/OI9qeCRwE0IIIcSk1uwO0OUPsaA8j0KXDZtF0TJBM27Z2sdNCCGEEGLEth5t45UDzQDMr8hHKUVZvoPWCVrjJoGbEEIIISalR7bW8YUHtqE1KAVLKvMBKMtz0uKZmFOlErgJIYQQYtJ543Ar//eBbZwzv4yvXLkMpWBGkQuAsnwHzZJxE0IIIYTIDi/tb0IpxW9vWk2+MzHcKctzUDtB+7jJ4gQhhBBCTDqHWzzMLnH1C9oAyvKdE7bGTQI3IYQQQkw6tc0eqsrykl5Xlu/AEwjjDYTHeFQjJ4GbEEIIISYVrTW1zaZvWzLl0V5uE3GBggRuQgghhJhUmtx+PIEwVWW5Sa8vncC7J0jgJoQQQohJpba5G4CqATJuZfkmcJuI+5VK4CaEEEKISaW22awYHXCqNH/ibnslgZsQQgghJpXDLR5sFsWsYlfS6+MZN5kqFUIIIYQYX7XNHuaW5mKzJg9zch02cuwWqXETQgghhBhvh5s9A9a3xZTlOWnukqlSIYQQQohx0+z2c7DJzeLovqQDmVXsoq7NO0ajGj0SuAkhhBBi0rj39aMEw5rrVs8Z9Lh5ZbkcaZ14215J4CaEEEKISSEQinDPhiOsW1LBworBM25V5Xmc7PTTHQiN0ehGhwRuQgghhJgU/rqljqYuPx89v2rIY+eWmua8R1u7Mzyq0SWBmxBCCCEmvGOt3XzvH3tYPa+EixZXDHl8bB/TWLPeicI23gMQQgghhBiuSETz7J6T/OTZtwD4yftPx2JRQ95ublks4zax6twkcBNCCCHEhPXLmgP88On9VBQ4+dF1pzGnNPn+pH0VueyU5NqpbZGMmxBCCCFExvmCYe56pZZ1Syq486bVAzbcHci8sjyOtEysjJvUuAkhhBBiQvr7tuO0eAJ87MIFaQdtYFqCTLQaNwnchBBCCDEh/eHVWpZU5nP+orJh3X5eWR4nOrz4Q+FRHlnmSOAmhBBCiAnHFwyz63gn7zx1JkoNvRghmaqyXCLarEidKCRwE0IIIcSE0+w2+4xWFjqHfR+LpxUA8NZJ96iMaSxI4CaEEEKICafZHQCgPH/4gduiafkoBXsbukZrWBkngZsQQgghJpzmLpNxKy8YfuDmcliZV5rL/pMSuAkhhBBCZExsqrQ83zGi+1k6vYB9ErgJIYQQQmROT+A2/IwbwNLKAmqbPfiCE2NlqQRuQgghhJhwmt0BCpw2cuzWEd3P0umFRDQcaJwYCxQkcBNCCCHEhNPk9o+ovi1m6fR8gHid24v7m9h8pG3E95spErgJIYQQYsJp7vKPuL4NoKosD4fVwr7oytLv/3Mvt79wYMT3mykSuAkhhBBiwml2+0dc3wZgs1pYOC2ffSe7iEQ0h5rcLKzIG4URZoYEbkIIIYSYcJrdgVEJ3ACWVuazr6GL+nYv/lCEhRX5o3K/mSCBmxBCCCEmlEAoQoc3OHqB2/RCTnT42HqsHYCF0yRwE0IIIYQYFS2eWPPdkde4Qc8ChSd3ngCQjJsQQgghxGhp7hr5dle9LZ1eCMDzexspybVTmjc6AWEmZDRwU0q9XSm1Tyl1QCn1lSTXO5VS90evf10pVRW93K6UulsptUMptUcp9R+ZHKcQQgghJo7Rar4bM7MohwKnDV8wu+vbIIOBm1LKCtwOXAmsAG5QSq3oc9gtQJvWehHwE+C26OXXAk6t9anAWcDHY0GdEEIIIaa2pmjgVjFKgZtSiiXTC4DsniaFzGbc1gAHtNaHtNYB4D7g6j7HXA3cHf36IeASpZQCNJCnlLIBLiAAdGZwrEIIIYSYIOIZt1GqcQNYUmkCtwVZ3AoEMhu4zQKO9fq+LnpZ0mO01iGgAyjDBHEe4ARwFPih1ro1g2MVQgghxASx7Vg7s4pd5Dpso3afSytNpi3bM26j9xOPrjVAGJgJlADrlVLPaq0P9T5IKXUrcCtAZWUlNTU1GR+Y2+0ek8eZDORcpUfOV3rkfKVHzlfq5FylZ6zPVziieWlfN2dPt43q4+Z2R1heasFXt5uaxj2jdr99jfR8ZTJwqwfm9Pp+dvSyZMfURadFi4AW4APAk1rrINColHoFWA0kBG5a6zuAOwBWr16tq6urM/BjJKqpqWEsHmcykHOVHjlf6ZHzlR45X6mTc5WesT5fm4+04X36Va676FSqV80c1fu+7h2jendJjfR8ZXKqdCOwWCk1XynlAK4HHutzzGPATdGv3wc8r7XWmOnRtwEopfKAc4C9GRyrEEIIISaAl99qRik4f2H5eA9lXGQscIvWrH0GeArYAzygtd6llPqOUuqq6GF3AmVKqQPAF4BYy5DbgXyl1C5MAHiX1np7psYqhBBCiInh5QNNnDKziJIs7rWWSRmtcdNaPwE80eeyb/T62odp/dH3du5klwshhBBi6vKHwmw92s4tF84f76GMG9k5QQghhBATQl2bl1BEszTaumMqksBNCCGEEBPC0ZZuAOaV5Y7zSMaPBG5CCCGEmBCOtprAbU6pBG5CCCGEEFntaGs3Lrt11La6mogkcBNCCCHEqOv0BfnC/W/SEt2eajQcaelmbmkuZnfMqUkCNyGEEEKMuhf3NfHXrfW8cXj0dqw81trN3Clc3wYSuAkhhBAiA7YdawegrTs44vt6aX8THd4gR1tNxm0qy9a9SoUQQggxgW2riwVugRHdj9sf4qa73uCSZZV4g+EpH7hJxk0IIYQQKQmFIykft6O+A4D2EQZuHd4gWsOze04CyFTpeA9ACCGEENnvxf1NrPzmU5zs9A157P6TbnxBE+SNdKrU7QslfC8ZNyGEEEKIJI60ePi/D2zD7Q9x7+tH8IciHIk2wR1MbJq0MMdGm2dkGbcunwn88p02lILZJa4R3d9EJzVuQgghhEjqb1uP8/CWOopz7bywtwmAVs/Q7T22HWunONfOihmFI65x64pm3L511UoAnDbriO5vopPATQghhBBJbTpiWnnc+fLh+GUtKWTQDjS6WVpZQEmegz3HO0c0hi6/CdxOn1PEomlTd4/SGJkqFUIIMSkda+0mHNHjPYwJKxzRbD3aztr5pSgFc0rNFGUqU58tngDTCnMozXWMQsbNTJUW5NhHdD+ThQRuQgghJp1dxztY9z8v8Mzuk+M9lAlrX0MXbn+I69fM4dtXreTbV60k32lLKePW7PZTluegJNdOuzc4ogA6NlVakCOThCBTpUIIISahP204QkSbAEIMz+ajbQCcNbc03oKjJM8+ZMbNFwzT5QtRnu8g12FDa+j0BinJcwxrHF2+IFaLwmWf2rVtMZJxE0IIMal0+oL8betxwAQRYng217ZSUeCMT5EClOY5h8y4tUavL893UpJnpjdHMl3q9oWiK0qn7v6kvUnGTQghxKTyyJZ6vNGATQK34dFa8/rhVlbPK0kImMryHEP2cYtlOcvyndit5rYj6eXW5QvJNGkvknETQggxqWysbWVOqQurRcUDOJGeLUfbOdHh47IVlQmXl+Q6hpwqbXHHMm4OSnLN9OhIdk/o9IVkYUIvEsIKIYSYVLoDYYpdDlrdgXj3fpGex7cfx2Gz9AvcyvIdtHgCaK0HnLpsimbcyvOd6OiahNYRNOF1+4MUOCVciZGMmxBCiEnF4w+R67DiclhlqnQYIhHNEztOUL2kol+mqzTPgT8UGTSTGcu4leU7KI7WuLXLVOmokcBNCCHEpNIdCJPrsOK0WWWqdBg2HWnjZKefd502s991pdGpz1hwlkyz20+uw0quw0aB04bNoka0OEECt0QSuAkhhJhUPIEQuU4bOXYLfpkqTduWaBuQdUsq+l1XGm3pMdjUZ4vbT1m+OU4pRXGuY8DFCf5QmFcPNhPceBc8+umkx7j9IfIlcIuTMyGEEGJS6faHyYtOlUrGLX2HmzyU5zspcvVfEFAaDchaB8mgNbsDlOc749+X5Cbv/fbkzhN85a87aO8O8tC0f3BWaCt9q+a01nT5grI4oRfJuAkhhJhUugMhch02cmxS4zYch5s9zC/PTXpdbKq0dYip0rK83oFb8m2vfvPSIYpddm5YM5cTbV2EfV39jvGHIgTDWqZKe5HATQghxKShtY7XuMnihOE53OJhfnle0utiGbfegVibL8LNf9hIfbsXMBm3ioKeXRIWTstne11HQtbNHwqzq76TK1ZO53vvOYXKXAu2iB/CoYTH69nuSjJuMRK4CSGEmDQC4QihiCbPaYsuTpAat3R0+YI0dfmZX56f9PoCpw27VSXsnrCrJczzexv59mO7iEQ0rZ7EjNtHzqvCGwzzpw1H4pftrO8kEI5wxtwSLBZFmctMkob97n7jiT2uMCRwE0IIMWl0+8MU4aYi0hRdnCAZt3TUNncDDJhxU0pRkutImCpt8JhmbU/vPslft9YT0cQXJwAsnV7AxUsr+MOrtbywt5HGTh9bowsgzpxXDIDLan5Pbe2tCY8nG8z3J4GbEEKISaM7GObLtvu5bNvncNllcUK6DjWbjNdAgRvAjKIcttW1E4mYgO2EJ8Lc0lwWVuTx5Ye2ASQsTgD4xLqFtHgCfPQPG3n7T9fz7J6TzC5xMa0gB4AcFQvc2hJu5/bLVGlfErgJIYSYNLr9IYpVF7n+ZnLsUuOWju5AiNrmbpSCeWXJFycA3HzBfPY2dPHI1nrABG5Lpxdw78fO4dLlZqeFBRWJgd/aBWU88/mL+N2HV9PlC7LhUCtnzi2JX++IBm4d7e0Jt4tNlebLVGmcBG5CCCEmJG8gzLHWboLhnjo2TyCMnTC2UDc5dotseZWiP204wunfeYbHtx9nZpGLHLt1wGPfvWomp80u4n+e2ofbH6LRo1lQkUdlYQ53fHg1W79+GStnFvW73eLKAi5dUcmnL14EwJlzi+PXxQK3ro7EwK1Tpkr7kTMhhBCDObENvG2woHq8RyL6+OJD2/jH9hM4rBbu/dhaVleV0u0PYSOMNewl1wbeYHjQfTUFhCOaX794kEAowluNbi5YVD7o8RaL4ktXLOPGO1/n7ldrCWlY2GsxQ0meY5Bbw6eqF1HksnPNmbPjl9kwAVq3uyPh2FiNW6FMlcZJxk0IIQaz/sfwz6+M9yhEEluPtHHm3GJy7Jb4isXuQBh7NAjIs5gCen9Ism6DeXbPSeravHz9XSuYWZTD6XOKh7zNuQvLKM938vuXDwP9p0YH47BZ+Oj58xMa/FoiZkq0b+DmjgZuec6BM4BTjWTchBBiMOEghHzjPQrRR3t3gOMdPj58XhXLZnTz1y11dPmCeAIh8qPTboXK/N58wfCgU39T3R9eqWVWsYubzp3Hh86Zh906dHbSalFcsbKSP79+FIAFFcnbh6QsbILsQHdnwsWdviC5Dis2q+SZYuRMCCHEYCIhE7yJrPDLmgPc9uRe9jaYLvvLZxTyvrNm4wtG+OeOhoSMW348cJOM20BC4Qgba1t516oZ2KwWHDZLytPK7zh1BgB59p49TIct+hwLehN3TzjR4WV6Uc7I7nuSkYybEEIMRochIoFbtnh610n2NnTGa56WTy+gosDJgvI8/vZmPW9bNg0bJuOW2yvjJpI72eUnFNHMK0t9qjNm7fxSSnLtlDlG4fxGM24Rf2LgdqzVy5ySgVe4TkWScRNCiMFEQvE3FTH+2roD+IIR7nmtlrI8BxUFTpRSnF1VyoFGN97oqlKAPG22YJoKvdy6fMF4X7V01LWahrtzSl1p39ZmtfDDa0/jvYtHmG2DeMZN+908t+ckP3p6HwDH2rqZXZL+2CYzCdyEEGIwkXC//RPF+GmNbrV0vMPHshkF8Wm9aYVOmt1+uvwhHMr8vlxMjYxbly/Iuf/9PO/99avsOdE59A16OdZmgtvZw8xqXbK8kuVlo1A/GA3cnBEv33xsF7+sOUiz2097d5A5pZJx600CNyGEGEwkLBm3LBEMR+LtIQCWTy+Mfz2twElEw7HW7nhPsJwpknE72tqN2x9ie10HV/50Pdf95jWOtHhSum1dm8m4zSwe5zqy6HMsT/moa/MSjmhq9jUByFRpHxK4CSHEYKTGLWu0dZs399Oi7SqWzegVuBWawONwswdHdKo0J2KCEv8kX5xwot1kFn9302q+/Pal7Kjr4Fc1B1O6bV2bl8pCJ07bOK661Tr+HMujZwX3s7tPAsObxp3MMhq4KaXerpTap5Q6oJTq1whJKeVUSt0fvf51pVRV9PIPKqXe7PUvopQ6PZNjFUKIpCIh0BGTeRPjqr3bvLl/cM1cbr1oAZdFt1cCk3EDqG3xYI9m3JwRk3Gb7FOlxzvMz7lyZiGfql7EpSsqeXbPScIp1LzVtXWPf0ar16rtXHxcecp0HDYLL71lMm7DncadrDIWuCmlrMDtwJXACuAGpdSKPofdArRprRcBPwFuA9Ba/1lrfbrW+nTgQ8BhrfWbmRqrEEIMKBKdmpPp0nEXq2+bVeLiq+9YTlFuTwPXWMbNF4zEu/A7wybjNtmnSuvbvTisFsrzTPB62YpKmt0Bth5tG+KWZtXmuBf/93puzcoN86nqRSyelk93IEyew0pJruya0FsmM25rgANa60Na6wBwH3B1n2OuBu6Ofv0QcInq30DmhuhthRBi7EWi02zSy23ctUUDt5Lc/qsYK/Kd8a9jfdzs8Yzb5J4qPd7uY0ZxDhaLefusXlqB3ap4OjrVOJBQOEJDp2/8M1q9ShHmF8Kps4tYFq1fnFOaK9uV9ZHJPm6zgGO9vq8D1g50jNY6pJTqAMqA5l7HvJ/+AR8ASqlbgVsBKisrqampGZWBD8btdo/J40wGcq7SI+crPWN1vs7u6iAPeOWlGoKOwiGPz1aT4e9rw1HzBr9v+yYa9/fPOxTYoSsIVm0Ct+a6g8CF7NizjxrvoZQfZ6Kdqz1HvORaSBjz0hILj246zHm5AwdvTd0RwhGNp/EoNTUnhv34Iz1f9kA750e/9nW2sKGmBpvH/K5dke4J9btIxUjPV1Y34FVKrQW6tdY7k12vtb4DuANg9erVurq6OuNjqqmpYSweZzKQc5UeOV/pGbPztTMHuuH8c9dAwfTMP16GTIa/r10vHIDd+7jyknVJt7Ca9eZL7G3oik+Vzi4vgCMwe958qqsXpfw4E+1c/cdrz3FeVTnV1afFL3vLcojvPbGHlWedS0WBM+ntXjvYAi9t4NJzz+D8ITaWH8yIz1dHHbwK2FzkqCDV1dVYZzVx/743OG3RHKqrVw7/vrPQSM9XJqdK64E5vb6fHb0s6TFKKRtQBLT0uv564C8ZHKMQQgxOatyyRqsnQK7DOuC+oxUFThQRrJipUUvQg1Lgn8Q1bqFwhJOdvn7tPE6ZVQTAruMdyW4GmOa2QPbUuLlKIGDamCyfUYjdqlhSWTCOA8tOmQzcNgKLlVLzlVIOTBD2WJ9jHgNuin79PuB5rbUGUEpZgOuQ+jYhxHiKrSaVGrdx1+YJJK1vi6kszInvmgCgAh5ybNZJvTjhZJefiIaZxYnB14qZZlp/1/GBG/IeavJgtypmFI134BZ9brlKTL1bKEB5vpOnP7+Oa1fPHt+xZaGMTZVGa9Y+AzwFWIHfa613KaW+A2zSWj8G3Anco5Q6ALRigruYi4BjWuvUCxOEEGK0SeCWNVq7A4NuZj6twBnfpxSAgJscu2VSL0443m4WYPQN3IpcduaV5bKzfuCM254TnSysyMdhG+eWrr0zbgABN9hKmV+e/v6pU0FGa9y01k8AT/S57Bu9vvYB1w5w2xrgnEyOTwghhqRjgZtMlY63Nk+AkiEDt17bk/nd5Ngnd8YtFrjNSrLzwSkzi9he3z7gbfc2dHL+wuHXto2aeMbNNFYm4Ibc0tRu63eDPRcsU2c/ganzkwohxHDEatxk94Rx19YdHLSn17TCnPiuCQAE3Ljs1kndgLc+Grglm+5cOauQY61eOrr7/+22egKc7PSzbEYW1JDFArdYsBZIYbuuE9vgD++C/54Nv7kIal/J3PiyjARuQggxGJkqzRpD17j1yrjZ8yDgwWm3Tuqp0qMt3ZTk2slz9p9AO2Vm4gKFe16r5UBjFwB7o5vRL5+RBS1u+k6V+t2DH+9tg/s+CM374dxPg68D7n4X1L6c2XFmCQnchBBiMBK4ZYVAKEKXPzRojduSygLOnBWti3KVmBo3m5rUGbc3als5Y25J0utiK0u3HG3jQGMXX390F3e8ZMrG9zSYAC7W6HZcJatxG8zjn4euE3DDX+CK78EnX4HShfDgR6GrIbNjzQISuAkhxGCkxi0rtEc3mB+sxq0gx84vb1hlvnGVgI5QbAtN2sDtZKePQ00ezl1QlvT60jwHq+eV8NDmOh7ZarpxbTpitsHac6KT8nzHgD3exlTvVaUw+FTp8Tdh1yNw0Zdh1lnmspxCuO6P4O+CJ76U2bFmAQnchBBiMPEat9Dgx4mMao0FbkPtW9mn0L3IFpjQixOa3X6u+/Vr7D/Z1e+6DYdM29NzBgjcAG48Zx61Ld3c+fJhlDItQFo9AfY2dGbHNCmkl3Hb+DuwuWDtxxMvr1wBF/wb7HkMjm3MzDizhARuQggxmIhk3LLBnmhN1vTC/qsnE/QJAoot/qzPuP11Sx33vn406XX3vXGUN2pb2XKk/4bxGw61UJBji/dsS+bKU6dTmufAF4xw3VmmJ/7TuxrY19DFymgN3LiLLfzJm2b+725Jfpy3DXY8BKuu7VmB2tu5nzH38fR/gmkJOylJ4CaEEIORnRPGndaa360/zIKKPM4coJ4rLvb7igZu+Vbf+C9O8HfB+h9D0Jf06p899xbffXw3Xb7EOspwRMcDuljGsbfXDrawdn4pVsvAm7A7bVZuXDuXwhwbX7xiKXar4ntP7GGpPsQNp+aP4IcaRbEsaf40cORD25Hkx227H0JeOPtfk1/vzId1X4ZjG+DwS5kZaxaQwE1MWduOtfPVR3agJ/EnMzFCkQgQ/fsIy1TpeHn1YAu7jndy64ULsAwSpAD9Mm5ltkC8Pm7w24Wg5SCEMhCgH6qB574NOx/ud1WL209tSzfeYJi/b0vc6P35vY0c7zDBXnuflh6x262dP/A0acznLl3C+i+/jYoCJytnFtHlC/Kg83vM23n78H+m0RT7ndmcUDwP2gcI3HY9ApWnwozTkl8PcMaHIK8CXvnp6I8zS0jgJqasJ3ac4N7Xj9KWpMeREEDPwgSQjNs4uvvVWsrznbznjFlDH9yn0L3CEcQTCA8+XfrqL0w/sJ+fCb86j+K27aMw6l4CZk9QtvXfenvrUdMgN9dh5f6NidOlj75ZT3m+k2kFTlo9iX9/J6IB3dyy3CEf3mpRFEVrA8+uKsGFH5fuhoYdaf8oGRF7blkdUDIvecat6yQcex2Wv3vw+7LnwNpPwMHnsufnG2USuIkpqy7auLLF7R/nkYisFen1Zi8NeMfNgSY3a+aXDLi5fIJIYjPXErsJClo8AwTeb/4Fnv4aVF0AV/4AIiFO2/ZNkyUbLcHoKsna9dBWm3DVlqNt2CyKz7xtEdvqOtjb0LO36MEmD6tmF1Ge7+yXNWzqMq9b6a4K/diFC/jZ1VXmm8Y9ad02Y2LBtsXWk3HrOxOy7x+AHjpwAzj7FjPl+uovRn2o2UACNw8KBmEAACAASURBVDFlxbaKGfAFXYjeK0mlj9u4OdnhY1rBEIsSYmK/p7wKAEqVWaGY9ANa2xF47P/A/HVw/b1mpeIn1tOdOxse/Ei/IGvYgt6er7fdl3DVlqNtrJhZyHvPNJupv3LAFOZrrTnS4mFeWS6leY5+GbfGLpNxq8hPL3CbVpjDZVXRlbndzeBuSuv2GZGQcauCYDd4mhOP2fN3KF0A05YPfX+uEjj9A7Drr+BuHPXhjjcJ3MSUVd9mXkz7viAKESeB27hz+0N4AmGmF6UZuBXOAquTkoCpG0v6Ae3ln4BScM2vwRbtD+csYMepXzX1jY99dnRWJ8amSmevgd2PxS8OhSNsO9bBmXNLqCzMYXaJi81HWgFocvvpDoSpKsujONfer6RjuBk3wKzOjGncnf7tR1vsdxabKoXEOjdvm1lssPzd5veVijW3moBw8x9GdajZQAI3MSX5Q2Eaoy98U3KqtKMenvoaHHhuvEeS3XSv1YhS4zYuGqK1XEO2AYlJKHSfS4HXNJ5tcff5/XXUw5t/NsXshTMTrvK5ZsAlX4fDL8LuR0c0fsBMlVpssOwd0Lgr3t1/b0MX3mCYM+aa1hZnzSthU20bWmuOtphgb24049aWZKq0MMeW2vRxX70Dt6a9w/uZRlM8cLObqVJIzHbuf9p8iFqWwjRpTPliWHgJbLwzMwtOxpEEbmJKOtHesyx/Sk2VhkOw/kemCPu1XyQtlha9SMZt3J3sNM/VaYUpZpZivzOrHUqqyHEfA5J8QNvwSxOYX/Bvye9n9c1mBeNTXxuwjUfKgl6zd+rCt5nvo/Vzb0X3DV0RbYS7el4JjV1+6lq68O55ilus/2DV0T9xanAHbq+PcKQn+9fY5R/+rgexwE1ZsyTjFgCL3WTTiueay3pn3PY8BgUzenZKSNXaT4C7wdx+EpHATUxJsfo2mEJTpW1H4M7L4LnvwOLLoGCmBCNDkcUJ4y4WuKWecYsVutuhZB6WjiM4bZbE53k4BNvvh6VX9gQKfVmscMV/QWcdbL1nBD8BZgsnu8sEgrnlcPB5wOxiYLWo+MrQs+aVcpllEyV3nc+Fr3+Cr9v/TNkr3+banZ/gefsX8L5+V7RFjcm4pVz315fXTMcy47TsWKAQDphpUjC92HLLe1aWBrrNzMCyd4IlzZBl0aWmLu7134zueMeZBG5iSoqtKHXaLFMj43Z4PdyxzvSpuvYP8P4/mVV3so3T4BIyblPg7yQLNUQDt8p0p0qj027K10FVboDm3lOlB58HTxOcdsPg9zV/Hcw5B17+35FNtwW94Mg1gcfCi+HgCxCJcKjJw5wSF06bFSIRlu3+Kb91/Bh32M7vZn6bd7rugS8dZOPZP6aFIvKf+jzcczV01NPkHmHGzeaCmWdA497x32UgHASrref7kl693A4+Z5ruprKatC+LBdZ8HOregPrNozPWLCCBm5iS6tu8KAXLZhRO/hq3PY/Dn/7FbAVz6wuw8hpzucUmwchQEvq4SZA7Hho7/RQ4beQ5bUMfDL2mSqMrFIHlrjZaPL2e59v+Aq5SWHTZ4PelFKz7ksm6bbs3/cHHBLvBHu23tvBt4GmEI69wsMnNgop887f16KexvPxDXsx7O9fr/+Zh71mUlldCXjnexVdxTeDb7FvzPby1Gwn+uprpnTtHFri5SmD6KeDvgLbDw//ZRkPvjBuYOrfmAya7uOn3kFsG884f3n2f/gHTGuT1O0ZnrFlAAjcxJdW3e5lW4GR6Yf/GlpPKvn/CAx82UyI3PwllC3uuszpkqnQoEWnAO94aOnxUprqiFHp+TxZbfIXiIntLz+IEXwfs/Qec+r6elaSDWXiJqa1a/6PhP18Cnp7Abfm7oWgu+u+fpaG5hRXFIbj/gyYwvPhr6Hf/jNqOEHtOdDIvOoVamucAFPcGq7nK9226wjb+aPk253ueHd54vO0mcJu9xnx/9PXh3c9oCQcTA7elV5pg+aGPmuzoBV8wGdThyCk0wdvOhydNaxAJ3MSUdLzdy8xiF6V5kzhwO/IqPHCTCdo+9Ei8IWmc1S6B21Ckxm3cNXT6Uq9vgz6tJaoAmGdp7Hme734Uwn5YdX2/m3oDYSKRPtOGSsFFX4b2o7D9gWH8BPRMlQI4C+A9t6NaD/Gw5T/49J4bTQ3XO34I677MuqXTOLvK7PpQVZYHQEmeCWrWv9XMW3o27498jy2Rxbxtz9fhue+mP9XZ3WoCt2nLwVlk9vYcT5FgYmB26rWw5O2w+29QNBfWfGxk97/mVvMYk6Q1iARuYkqqb/cyq9hFeb5pbNnvxXqiazsC999oCq9vfNi8WfRltUswMhRZVTruGjt9qa8ohcRVpTlFkFPMTBppdvvNvsTb7oOyxTDrzISbBUIRLrjtee7fdKz/fS65AqavgvU/HN6Uee+pUoD5F/HW6m9xQpfiq1gFtzwVD06UUnz57cuwWhQrZxYBUBLdrupQs9mB4S23kw8Fv0LDwuvMmB7+VwilUfLhbQNXsVmAMefsLMi49ZkqVQre/VOYtRre8QPT2mUkyhebhQqTpDWIBG4TkdaJfXhEWrTWnGj3MavYRWmeg4iGdu8kelMO+uC+D5o3mBvu659pi7HYZfpvKLJX6bj5ysPb+eKD22js8qeZcQsAygQlACVVVIQa8IcidDcegiOvwGnX92vkerzdS4snwKEmd//7VAou+hK0HjJZoHT1DdyAl0vew43BrxF6/3392lycXVXKlv+8jHMXmg3kXXYrTpt5u85zmJ8rhI2OS38El3wTdj4Ef7zaZNJS4W3reV2Ycw407Rnf95S+U6UABdPhY8+ZadPRMIlag0jgNlG0HYFXfw5/+QD8YAHcVgX1W8Z7VBOS2x8iEI5Qlu+I1o5Aq2cSLVB45utwcge897dQvmjg46wOKbgfimTcxs1rh1p4aHMdoYhOfUUp9A8CSuZR5j2MhQihjXeby1a9v9/NjrWZhrcdA32IW/YuKF9qdltId2oy0N0zVRp1sMlNQY6N8vzkdXaxTeHBZOFKcs1x7z1rdvzyisIcuPAL8L7fm/eD311qVo4PJvbB32WmY5m71vx/bGN6P9NoCgdMTWImLbwEShfC67/O7OOMAQncslnQC1v+CHe9A366Cp7+T9PleuHF5vrjW8wLwlNfy4795iaI9ujWMcW5DsryTAq+X1f1iWr/U/DGHXDOp8z0zmCssqp0SJHeOydI4DaWmrt6PkylH7j1qpda8R7yvCf4uu0eCrb+Cla8B4rn9LvZsVbTIqi9e4Dfs8UCF3weTu6Et55OfTwQbcCbGLgdavKwoCIfleIWTrE6t4uXTmNuaS52q6LYFf05T3kv3PSYCch++zaz08CAY+k2NX6xwG3WWSZoivaW6yccNPWy2+4zi5367iE6GvpOlWaCxWJq3eo2TvjWIBK4ZaNw0Cxd/ulpZgNk90l429fh33bAZ7fAe+80XbibD0DtetMB/6X/Ge9RTxixrWNKch2URT/tTopebt52+PvnYNoKuPRbQx9vdUiN21B6Z9wm2bny+EPccMcGDjQmmRocZ57o/qQXLamgMMfG8hlJajQHEgkmZm9WXkPn7Go+anuKsCUHrvxB0psdbR0i4wZmJWrRXHjph+ll3YKepIHbwoq8lO+iNM8EaStnFXL+onLmleVhsfQK+uaeY6YWi+bAvdfCs99OnlGPTYnGAjdHnmkRtPkuswAjJhQwszz/uwruuhIe+Tj85Xr44WL426ehPUkt4HAlmyrNhEnSGkQCt2xzbCP8+kL455egfAnc9Dh8ZhNc9MWeDt9KmbYOLW+ZT38AW+6WrFuKYqvLSnLtlOVNosDtma+bIP/qX6RWzGuRVaVDmsQ1boebPbx2qIUtRzJX23THSwd5fPvxtG/XHO2teNVpM9n2zcuZV5Z6gJOs0N172Q84GJnBhuVfhYLKpDcbcqoUTCbv/M+ahq5HXkltPKGA+QDQK3Dz+EM0dPpYWJGf2n0A0wpymF6Yw7SCHL7+ruXcf+s5/Q8qXQD/+gyc+WF4+cfwx6ugtU+Ptnjg1qv29dJvAQqe/A9z/e5H4dcXmFme8sVw3R/N+9BH/2ka2u54EH51/ujs4wr9s6SZklMIp3/QtAbpOpn5x8sQCdyyRSQML/w/+P3l4O+C9/8Zbvo7zL+wXxEtYJ5MzW/ByV1mOXfIDxtuH/txT0CxqZCSPEd8+qF1LKdKtYZdj5jpk9FybKOZVj/306nv5yftQIaWUOM2ueoB3X7z8wwaqIxAIBThtif38Zl7t/LrF4eou+qjKTpNWlHgTHkqMS4c6hcETJu7hA/m3M593rUD3qwumnHrHOp8nHEj5FXA+h+nNp6gud/eNW6Ho6tDF5SnHpB+6Yql3PXRswHIddgoyx/gw5ndBVf9HN7zazixHX51Hrz2y57WNn0zbgBFs8008N7HTf30Ax82mcsPPGCmYFdcbd5z5p0HV34fPv26qZ994MOpn4fBjMVUacwkaA0igVs28LbBn98HL95megt96lVY/q7kAVtM2WKT1q7fYp5MK6+BN34LnpaxG/cE1Xuq1G61UJBji182Jhp3w4MfMVMZoyESgSf/HfKnw7p/T/12VllVOqTYm53NNenOldtnArdOX2YCt6Ot3YQjmhlFOXz/n3tp7Ex9o/Z44DZQcDKYvj3BMMX9Fy0p5+UDzYTCkaQ3O9ZmPkgNGcjaXaaG9OBzcHzr0OOJBW69Mm4HoytXF6SRcZtZ7GJ5dDP6lJx+gwmwqi6Ep/7D7FN89PXkgRvAui+bGZ63/afZFu8zmwauky2dDx990vRbe+7bLDh498i2zRqrjBuYgHPRpbBp4rYGkcBtvHWeMIsPal+Gd/8MrvmV6T00lPLFgDZblVSuNG/YAQ+8+tOMD3mia+sOohQURQt78502PP4xzKb4Osz/b/wGGnaM/P6232+KbS/9VvJ+bQOxOmSv0qHEAjd7zqQL3DyBaOCWoYxbrK3GDWtMiUdtS3fKt21y92Tc0hYOmDKAPi5aUkGHN8i2uo5+17n9IVo9AfNaEAgT6tXXUWvNNb98hfs39qr/OvsWcBaaFaZDiWXWEwI3D0oR3xkhY4pmwQfuh3/5LXTUmxmdZ75hrusbuCllZngu+pJJBMTaqQzE5oBr7oDVNzP32F/h8c8nLuZJx1hm3MAE3u6TsPWesXvMUSSBW5pUJAy+ztG5s5aD8PsrTObsgw/CWTelftveWxdVroBpy8ynnzd+O2m29ciUNk+Awhw71mhhb77TFp82GhMBT8/XT311ZPfld8Oz3zLTo0laHAxK9iodmu6VcZtkQW5XPOOWmZ8rNh24bkkF0FP8n4rmLj8WRbxdT1oGyN5csKgci4KX9vevBT4WHduKmSaj1XthaV2bl61H23n1YK/ZjJwi0zB392PQtH/w8cSe772mSg81uZlTkkuOfYjgaDQoBauuMwvbLv6aWRVqdQzc3zEdFgu888ccmftes7jh0U8l7jaSqnBg7DJuYPaLnbPWLDIJpp4JzhYSuKVpzrFHTA3A/R8aul/OYBp2wO/fburZbnoMFlSnd/uyXv25Kk8x/1d/xdS6pfIpcApr6w4kvCHk54x14BZdxbfyX+DwSyNbnfXyj01Tybd/37yIpkP2Kh1aLFibhBm32N985jJuHsrzHSyfUYhS6QVuTW4/pXnO+IertAwQuBXnOjhtTjEvDhK4nTrLzHZ4gj0Zt531HQnHxK39JNhyzM4Fg4lPlbriF5lWIGksuBgNjjwzHfrZrXDLMwnjGRGlOLzgwyYo3PYXs/o03XrQsZwqBRPMXvw16Do+IWvdJHBLk8tbb/7ADr0Id14Ox99M/06OvAZ3vdPcz81Ppl5M3puzAApmgNVpmgqCycKddoPZ1qMz/ZVcU0V7d5DiXs0txz7jFn0hP/ND5v9djwzvfloPw6u/MJm2OWvSv73VbjJKw53emApigZvNlT1B7snd8Is1qXfJH0Cma9wON3uYX56Hw2ZhZpGrf+AziKYu//CmSSHaDiR5EHDOgjJ21nfgDyVmhY72DdxCPYHbjmjgdrS1z2Ki/ApYe6vZv3Swkod44GYCtUhEc7jZw4Ly1OvbRlX+NJh5+ujf77ovm10cdjwID9+S3vMlMkbtQHpbsM7U/63/Uc9r8gQhgVua7EG3yXZ97HlTs/CHd5msSar2PwX3XGOePDc/BRVLhz+YylNgxirTSDVm3ZfMm/H6Hw3/fie5tu5AvAs5RAO3DE0XJRWbOqk8xQTtOx8a3v0883VTh3Lpt4Z3+9gn3EnWn2xUJdS4Zcl5atwNzftGlvGnd8ZtFP7273oHbLs/4aJDze54cDKn1JVexm0kgdsgPcFWziwkFNH9ete9drCFaQVO5kZrznpn3GKBW7PbjzfQZxrwgs+badNnvzXweAKJq0pPdPrwBsNjn3EbCxd+AS7/L7Mt2IMfSb34f6xr3GLe9p/gaYSNvx37xx4BCdzSZA92maLO8kVmY+CiWfCn96bWz2brn+C+D5hg7eYnk3bvTsvVt8O1dydeVlJlevhsvjuxmaKI65txyxvrxQmxqVJHnul4fmKbaaacjsMvwZ6/wwVfgMKZwxtHLCsxyaYAR1XvVaXZEuDGNhP39y+yT0fXaGXcImHT0+zNP8Uv6vAGaXYHmB8NTuaW5qadcRtoK6ghhYOJH2Z7ia3K3H28p065rq2b5/c1ct3qOfGdCDzRU6K1Zkd9B4U5tvixCVwlpsfmgWfNLEwyfRYn1A6jFciEct7/MaUbex8373f+rqFvM9ZTpTFzzzErTF/+39GrXR8DKQduSqkML3+ZGGyhrp7VOIUzTUPCGafDAzfBa7cnXxId9JrGho9+2qRmb/o75JWPfDAFlSZw7OvCL4KywIvJO4RPda2eAKV9Mm5dY704QVlMfczKa8zXb/459duHQ+bvqWgunPeZ4Y8j9gk3WzJJ2Uhn4arScDRwG+Ebjdtvfu8jrnELRYu7j26IZ5P79imbW5pLY1eSjFUSWmua3YGMTJVWleXhslvZfaLn3P3ljaMo4Ia1c+MrzbujGbe6Ni/t3UEuXznd/IjJgs+zP2Z2K3jmG8nLDoLRDHs0cDvRYc7XrJJRqjHLRud8Et71E9My5fdXmhWtgxmvjBuYrJu3DV6aOO+XQwZuSqnzlFK7gb3R709TSv0y4yPLUvZgV+JqnNxSs7hg2TvNCsF73mN65UQiJmDb/qDpQL3hl7D2E/DBh0z35kwqmgWrb4Y37x3xdMpk4wuG8QbD8ca7AAU5JuOmR9KHKB0Bj9l2RSkT/C99B2z5IyrVjM6Wu82OGZd/d2QFxrGshARuA4vXuGXRVGk84zbSwM38bJ5AmOAAvc1SEluVFw6YPS2Bw82xPmUmcJtTaoKWfhmrJDq9IQLhyPB6uMXGMUAQYLUols0oiGfcwhHN/RvreNuySmYVuyiMZ9zMa0FsYcI7TjWBW9KsoT3HFLqfeBN2/bX/9YHExQkno/3s0tp/dSJafTN84EFoqzX7pw7U8y4SMc+zAYLtjJt5hmmqvOFXQ68QzhKpZNx+AlwBtABorbcBF2VyUNlIa02HJ4At5E7cKgTME/K6e+CdP4a6zaZXzn9Ng+9Nh7/+K+gIfOgRuPK2AVP4o+6Cz5sXrxdvG5vHmyB6NphPnCqNaPAGh7GMfTiCHjNNGnP2LdDdTEXTq0PftrsVnv8vmHeB6WY+ErE3t2yZAsxG8Rq33OwL3EaacetV19k1khrPUK92CgeeA6Chw4xxZrEJVuZGA7dU6tya3Ob+hl/jFhr0dXb5jEL2nOhEa83R1m6a3X4uX2m2wbJbLeQ5rPGM296GLpSC8xaW47Jb4016Ex4uojk8850w/VSTdfP32fs1vnOCec43dPgoctnHphXIeFt8KdzytJkGvfNyMyvVNysZe/0Zj6nSmEu+aZ7jT3xxZI2Ex0hKU6Va6779CsboHS57tHUHOf+7j2LR4f6NC8G0Yjj7FvjCLrMJ/DmfhOqvwocfhc9sNn1jxlJBZc+Kp8a9Y/vYWaz3rgkx+U7zIj9mCxQCfTacnl8NpQuZXff40C8aT33VZFre8YPBd9ZIhdS4DS0L24FERjnjBiOcLo0HbgoOPh+97yBWi8IVDU7SCdwaR7JrAgy5QnHFjEI6fSHq273xRQqLp/Ws8CzOdeCOno76di+VBTnk2K0DLrD4+7bjXPq/L9O87v9BZz28+P3EA4LdoKzxMZ3s9DF9smfbeqtcAbe+aGrJnvoq/Pm90NXQc33sA9F4TZWCWSF86Tfh8Ivpla2Mk1QCt2NKqfMArZSyK6W+COzJ8LiyTpHLTomKfpIarHFhThGc+j4zjVX976Y/W7r9tUbLeZ8zn/Je+N74PH4WavMMEriNVZ1boE/GzWKB8/4PhV37zYKDgRx4zvRJOv/fzG4ZIxX7hDvJ9uAcVTqaHbC5AJ3YXDTQbdoBjXE5wuEG0wakva15RPfj9ocojwZHI1qgEAvc5qyNr3b1+MPkOazxfUZL8xzkOawc69tSI4mDTaYmLDa9mrYBdk6I6b1AIRa4LewVuBW67HRH24HUt3njtWhzSpIvsHirsYtwRPOmXmoWhr32y8Q2UUGv+aAWPRcnO31MKxxmUDpR5ZXB9feaurcjr5l2Nht+ZYK22Aei8QzcAM66GeaeZ4LLLG+nlUpE8Qng08AsoB44Pfr9lGK1KObkRF+gkmXcslFeGZz3WdjzmFn1JGiLbzCf2McNxjpw69PD6YwP4cmdA89+M/kS+o56+OutUL7EbEkzGqyScRtS74wbJJ6rv30S7lgHPz9zTNvvtHaYYMPTOfI+brGgZEQtQWKB26rrAAU7H6bLF6Igp+c5ppSiNN9Bewp7Au+oa6ck187s4RbvDzlVWoDVothW186BRjfTCpwU9hprkcsWr3Grb/fGp3vnlOZS1+btVwtbH50+3XeyCy77jtmA/pFP9ExpBzwJuyac7PRPrYxbjFKm7u0TL8Ps1fDkV0z99/6nzPXjOVUK5gP01b8wweQjH8/q/pZDBm5a62at9Qe11pVa62la6xu11lNyJ/PZ8cBtFLYKGSsX/JvZkP7xzydutTQawkEzDbv7UfPpaQIshEg6VZoz1lOl7sSMG4DVxsGFN0PrIXj6a4lTpr4OeODD5g3y/X/qCSJGSmrchta7HQj0TOs07DS9qs78MCy5Emq+D037xmRIXd3meRzwDL8dSDii8QTCzCo2f0sjy7hFA5TSBTDvPNjxEB5fkDxnYg1XnsMW3x91MNvrOlg1uzierUvbECsUcx02TplVxBuHWznQ5GbRtMQPUUUuO91BTSSiOdHhZVY0cJtd4sLtD8XrZGPq26OBW0O048BVP4emPaYWFaIZN3Mf4Yimye1netEUDNxiyhfBjQ/DDfeZ17S/fcJcPt4ZNzBN7K+8zbRbevVn4z2aAQ1ZKa+UugvoV3ijtb45hdu+HfgpYAV+p7X+fp/rncAfgbMwix/er7WujV63CvgNUAhEgLO11uO6qdhMhxe8jM4eb2PF5oSrfgZ3XQmPfRbe+7vh10Z11Jt+TbUvQ90maN6f+KbfuMc8VhZrjU6V9t05AcYy49YNxf17OLWWnQnnfgZe+4V5ETv/c2ZF1mOfhZa34No/jKxhc1/xGjcJ3AbUL+MWPVcv3mY2Gb/sOybDc/vZ5sPRR5/I+JA8HjNdp73DD9xiAdSMoljGbQR/A/E+ZS7Tl/AfX6DMup+TznkJh5l+iYOXR3sDYfaf7OKyFZXDH88g7UBizplfyl2v1GK3Kt571uyE64pdDjxBs+1WMKzjWclY5u1Ehy9hVXo849YQ7Ve25HI46yPmjX/+OlPjFt01odntJxzRTJuKGbfelIKlV5ra7633wJt/Mas7s8EZH4K3njEforUeeS1xBqSyxPHxXl/nANcAQ04AK6WswO3AZUAdsFEp9ZjWenevw24B2rTWi5RS1wO3Ae9XStmAPwEf0lpvU0qVAeP+7jLNFq1vmChTpTHzzoNLvgHPfcfs+lD9ldT+GLU2Xdr3/N1spty4y1zuLDJbLC25HKatMNN399/Y8wKexfY1dDG7xIXT1pMNyIqp0pjLvmuKzl/7hfkHpm7yxr+aLVpGkzXLAreuBrON17xzx3skPXTfjFsAOupM+cGFX+x5LVj372bq5/ibmdlOKMofChPwd4MVLIEUGpsOIJZdjgUjo5JxsznNSucnvsRpXTUcLbsl4bBch3XI59iu4x1ENKyaXTz88YRDQ067rZlfym9eOkQgTP+MW64dT1BTFw3IZkfP0YxoluxEhze+GX0wHKGh04fNojjY5CYQiuCwWUwD2mMb4ZFbIX96fKo01gpkSk6VJmNzwtn/av5lC6XMAkNbFmQABzBk4Ka1frj390qpvwAvp3Dfa4ADWutD0dvdB1wN9A7crga+Ff36IeAXyuTHLwe2R1uPkC1TsxXW6FTjRAvcwHTYb37LrHhyN8Dbb0s+5RaJQP0m2PsPE7C1HgQUzD0XLv8ezL/QbNVk6bOU3e6aELVSO+o7WDW7KOGy2FTpmO2eEHAn1LwksFjMVMvaT5rO46ULzKfSTGR5s6nG7a1nTA1fyAdfOzHeo+kRbwcSDdwiQdNoFmDle3qOO+16ePZbdLzyO1zX/My8eWfAoSYPjuhnWHtoBIFb9G99WoHZyH1kNW7RD2w2l2ksPvMMljXs4AVn4ttLnsNGY6d/0LvaXmeyiKf1eY6mJRwYMnBbXVWKUuaz6aKKxMBtdomLQAQ2HDJvO8kybjENHT4iGs5dWMorB1o43Oxh6fSCaIuou03/ssZdMP+i+PEggVvWy+KgDVLLuPW1GJiWwnGzgN5tROqAtQMdo7UOKaU6gDJgCWYV61NABXCf1rpfW2Ol1K3ArQCVlZXU1NSk95OkyepuoEu72PTSy8OvvxhPxdcyf26QeZv/QGD73zgx4xLcSilXMwAAIABJREFU+QvRyoIj0EpB1wHKWrbgCLajsdBWcirNiz9Bc/k5BJwlEAD2tcG+9f3uerU3iO/kcXb2+h243e6M/07S4Qmavk1ryoMJ4wqETSXAtj37meOvzfg4LvK7OdbQwuE+56b/+VpjCghatmdkHAWd+zgL2P7mFlqPjd/f86y6v7PowJ2AAjQvpvg3MxZ/X1WHD1IF7HnrMMuB1199mdl1j1JpzeXlPY2wt+fxF5WcQ9HOh/iB5wouqCrIyHg2nAhxBibIckU8PPf8C1gtqf3uep+vA20mID28fzcuq2b3wVpqaoYXMFc2bGM5sGHzm/hcDSxUs1ka/ge+1hMJv5+uNj/NHeFBf2fPbPNR4lTs3rIh4RN+yrSmOhKk9lg9tUP8bczJt3C0K8LJA9upqesJtK1uU5R+z3rTjPXgjk0c36OIaI1VwYbt+5jtOwzAnhZzHufaungFuO+ZDcwrtFBVZD7Yli75PKfu+C4tnX521tSw/qgJug/u3EzLgcmz42S2vdZnu5Ger1Rq3LowNW4q+n8D8O/DfsTUx3UBcDbQDTynlNqstX6u90Fa6zuAOwBWr16tq6urMzqot/bcQbs7n7POvSBhFdKEcvElUPtRHOt/zLxDj/S0OwCz6GLJxbD0najFl1LqKqEUE0UPaX8J+bmF9P4d1NTUkOnfSTpeOdAMvM7VF57BhYsr4pdrrbE990+mzZxLdfWyzA4iFICaEPMWLWfeRdUJV435+TpeDFtg1cplsGwMH7e3F/4bDvwOlr3LFAa/8lOq161LaSp/TM5XeD0ctbL8lFWwF9auPhNqfw5V51J98SUJh+51+Zn9zxpWsZfq6lFa+dvH5qf34VQmcCukmyWnrWFOWWp7XvY+X5b9TfD6G5y/5kwert1GQWkx1dXDrDHadAj2wjnnr4PCGTDdA/f9jQvKu6iuflf8sOc7drK7/figv7Pvv/kSZ8x3UV199vDGEg7Ci1C1YAlV6wZ+HICr/Pt4eEsd77ni4oQP4lprfrDxnzR0a4pcdq689OL4ddNffx5HcSnV1WY6vHlzHWzcxs1XnsMD+9dz1y6TvV7/5Yuj7Uyq4dSVlOeVUz3rLDY9tQ/r3oO8+/KLUw64J4Jse63PdiM9X6lMlQ73o2M90HsX9dnRy5IdUxetayvC5BjqgJe01s0ASqkngDOB5xhHBbqTRvKJeAITN3ADqLrA/At6oeWAmS/ILTPbLw03k2h1ZMeU2yBi0zCnzkqchlFKjd1G87F9CweqcRtL471X6cs/MVP3p99oFrWs/7G5PBIeux1GhqLDpiwgNvXmaTKLcFZe0+/Q7Zbl5EUqWND4DJCZwO2tk24utkcgDDYVoa6pJeXArbfYVGm+005hjn3IxQnBcARvMJz8dS+25VW09ELPWYsCFvt2ADfED8t1DL04odkd4Iy5I6lvizVzHfrv53OXLubWdQv6zZ4opVheamHDiXB8RWnM9KIcTrT3TJXWtXWjFMwty+Udp85gX0Mn+0+6qW/39vShW3JF/PiGTh8V+c5JFbSJsTdgrlYpdeZg/1K4743AYqXUfKWUA7geeKzPMY8BN0W/fh/wvDZNcp4CTlVK5UYDunUwvMz5aHKFO2nX+fGViROe3WW2aZmxyuxvOpLpX5sjef+xLLKzvoM5pS6Kc/vXL4zZRvOxlix924GMh/FcnLDjIXj2W2YV4lU/M8FRrFF1ZIxqDVMRCYHF1hPkHnkF0KbZbB+1Ld08EVnD0u7N4G0ftSFsPdrG0ZZufMEwm4+2UWjvyZI3nGwc1n3GFifk59godNnoHKIVzu0vHODK/+1fIgH09HGzmcDNay/mrcgs5rm3JRyW77QSCEcIhJL3x4pENG3dAUrzRlBfFFvlnsK+l3arZcAP4CvKzFRn343gZxTlcKKjZxFWfZuXaQVOnDYrP7/hDH7xAfPW2OxOXst3stNH5VRuBSJGxWAfSwbrKKmBQfdwitasfQYThFmB32utdymlvgNs0lo/BtwJ3KOUOgC0YoI7tNZtSqkfY4I/DTyhtf5Hqj9UpuQEO2hnBqEUmkhOOVZH/z36ssyO+o5+2baYghxbyn3ctte1U5hjp6p8GMFXLHCzD7Mr/GiKBW5j3cft+FZ49NNmwct7ft2z0EVF/9dZtKNeJBLdrih6rg6vN9/PXt3v0MPNHk6E1/Jx2z9g/5NmwcIIaa352B834bBaeOeqGTR1+Zk+S0E0VmpuaRrW/XbFM242ilx2TnYO/tzdfbyT+nYvHd4gRa4+wU58VakJSNz+EBsjS3hf5yZz/qIBea7DvN10B0I4khR/d/lChCM6ocdi2kZp+6R44NYn4zaz2MUzu0+itUYpRX27N+GY2DZdTV3JA7emLv/wd4QQImrAwE1rffFA16VKa/0E8ESfy77R62sfcO0At/0TpiVI1rAFOmjTiwl4sqR9QjaxOrN6qtQXDHO0tZtr+/RsislzptYcFOBTf97C0soC7vzIMOpwAtE3yGyYKh2PvUr9XfDgRyC3HK67J3H1liX6chTJpsAtFM0GRs/V0ddg1llJM6aHmz3s1Qs5rsuYvutvWEYhcDvR4aPZbX4/v11/mMtWVJLvDpsgKeSjqWl4gVs84+a0UZHvZP3+5ngwkkysNcbxdm+SwM1rnv/R27p9ITZGlvGB0AumndD0U+KPBeAJhClOEru0eEywU5Y/GoHbyKbay10WvnTFUqqXViRcPr0wB38oQlt3kNI8B0dbuzljbk+XgSKXHZtFDRi4tXhGOBUsBCluMq+UOkUpdZ1S6sOxf5keWNaJhLH42mnn/7N33uFx1Wfavs/0Ko16sST3bgO2wTYYgzE9JEAIm4QQkk0jhWyym2SzSTbJZlM2Pbshm7qE8CWBVBIgMRBCMZhubAy4ybZsWVavI02v5/vjd85oRjOaGWk00tic+7q4ZDRFR2XmPOd53/d5HQz7so+0vy7RG0tauKn5SQ0TrqBVHOb8HDdPMELnSIDDvdOMYiipUukc9Lg9+Glwd4ggaEfqSTHhvJVcqXR8QThyDFa+Kf1ucZkTgz6cZiMPxTYitT0OwcKWwIMo7wO8c3MLjeUWPnv1CuFw2cXP7kRXN8HI1IWuNxTBZtKLVX6VNjyhKKNZ+tw6R0SGpRo2m0I0lHDbAHyhGLtlJSi647nE523KJgX/JC0JmbaaTJkZ3Ht52yVLWN2Y6tA3usaz3A71jNE5EmBDkhDT6SSqHeaEcEsuC8fjMiO+AkvBGhrkIdwkSfoP4AfKf5cA3wKuLfJxlR7BUSRkxnAw7ItwoHsUf54OzesCQ2k7bn1KflTdJMudHWZDXgG8R5Wl1F3uwPRCS0tKuCmuxGwJt9aH4JV7RHBtppDdRKm0hHYEyjGlxy3JwVmV/vbXOxYkFI2zaVEVD8Y2IsVC4zsYC2B/9xg6Cf79Dat45jPbWVTjEK8zRbiZoj6ePDJ1180biiYcMLV0N9kC+NFAJNEDp653SiESSMmE9IQidMo1hGz1KcLNbsoedD2kOItV9gIWsIeUC6o8etymQ72yaaLHHeTePZ0Y9RLXnjMv5T41TjMD3hAvHB9izZf+lshuGwtGiMZlKgv5/jQ0yM9xuxG4FOiVZfk9wNmI6c/XF+Yy+OhLPK6/kAPdo7zpB09z9/Mdc31UpYPeWJrDCS/fDWPd9CqOW90kwZf5CrcjSU7b0T7x71c73Vz63Z30e/LYyBYuwanS2ehxC3lhx6egZiVcNMnEZcJxK7FSqZTkuNWfBRUL0u52YlD8Xi9YXMVeeSlBS43YZVogB7tHWVzjwGrSj5cxkxy3OnOIh/f3Tvl5x4JJwq1CCLeOYX/G+6puG0wi3KIhceGmICZHJQL158HJ5xJ7d+1mtcct8+834bjZpyG6omExlfyLqwFpZlfDJdGoDBZ0jvi5b18321fUpjloNU4zg94QezpGCEfjHB8QF3tDPlWYao6bRmHkI9yCsizHgagkSWVAP6kxH68P9AaoXkrUVM6uo4PEZegeLf0VT7NGKcaBhDxw/0fg5bvpzyXc8hxOONLnTQzfquXS2x87StuAjxMDvtzHVEqO22z2uO38Oox1wpu+P3kqeUmWSuOK46YIkwxuG8BxVbgtqUJGR1v1djj2aGJgR5bT1j3nxf6uMdZMHKiJjQu39XU6Hj3YRyg6NbHbPuijSXHamiuFi3RKEWiBcIwP/WoPn/3Ta+w5OZLob5OkyYRbcHwlGOMbSKJNm8HTLUrjiJVXkMVx803TceveBz/dCo/9p4g5+vCzRVs7Vu0wY9BJfP+xowx6Q7xlfXrPbLXDxIAnxHHl/WBAmTBV0wgK6uHT0CB7HMgPJUm6EHhRkiQX8H/AHmAv8NxkjzvTcSRdDKpNwxqU5nCCWjbxD9I3FsRi1FFmydy0LIYTYsTj2U+wR/o8rGksx27Sc6TXw7F+D48eEpEMkzkJKSSEWwlMliXiQIoslIba4IWfiuXNLekxGglKcqpU6XGrXir2T573gYx3ax/0YTXqWVbrxGrU85L9IiFojj7CvlNutnzjce57eWKMZXYGPCF6x4KsVvZiJoiGwVYBko5VFRKeUDSRUZgPsbjM0X4vy+uE6+u0GHHZjJxSHLcHXuni4QO93L+vi1t+/gJtimO0qqFskh63YIrjpk6sSguUcrhSLnWYx6dKMzHsDWM16rGa9BlvT0OW4fmfwB2XiX7Cd/webvoN1K3K7/HTQKeT+MglSzi72cU/bGjikhXpS4SE4xZO/NzUfrchRcBpPW4ahZJt9OYI8G2gEfABv0EsjC+TZbk4O3hOA5ym8amroUmyel6XlOJwghpP4h+mNxyirswy6dScUzmpeELR9Km5JI70edi6tAaDXuJwr4cf7WxL3JbXVGoiDqQEHDdJEm5SLCzKXb37wT8IS68oLNNvIo98QZzYt38h+/1KcapUDeCVJNj84UnvNhoQU4Y6nUSDy8Iz4aXcaKyg9eG7uHnURiAS42DPGNevmzfpc0zkQLcQYxkdN4MFzE5qzeI96Eifh/MW5LfPtn3IRzgaZ3n9uCBsqbRxShFlv36+g2V1Dj77hpW85xe7uXdPJ3aTntWNZTzRmqGfLhqcMJygrORqXAOWcmh/Gs5+e2I4wTtJCO/wVDLcYhHY8QnY+0tYdjVc/6Pi7PPNwCcuz75LpsZhJhaXOdAlhlP6VeE2XUdRQ2MCkzpusix/X5bl84GLENsM7gQeBt4sSdLSWTq+kkMVbqsayhLNtBqU5nBCWHXchkTwpXPy4Eu1XNQxNHmfz9E+D/2eEMvrHayod7K3Y4Q/7e3ijWc1AHkuqQ97RVm5VJYY602ix+3uG+GO7XDPW8WJdqY4+Ry07oAL/wWcddnvqyvB4QQ1gDcHvlA0UQpsKLfwyKFB/hxYz0rv82xdYMNlMzLqT+8ljMbik7q8L54YRq+TUoVbLCp+PnozWMqxyz7sJj1H+/LPUFT7NJfXjS/Faa6wcWrYzyun3LzWNco7N8/n/EVVWIw62gZ8NFXYmOeyMeAJpU+xRkMpwwneYBSdBFazSSxXP74TZDkxnDDZVOlwvhOXkSD87hYh2rZ+Ct5+z6yJtnyoUd5nwjHxd6w6bsPK+UJz3DQKJWePmyzLJ2VZ/qYsy+sQ+0uuBw4X/chKlG3NRr5y/RrObnYlcodA5IQd7C58/P+0RW8SJ5Ril92mgupuBYbpz5FYvrhGlI2ODaTHfMiyzD/85Dmu+J+nAFha52RZnZNITGZdi4svvkmUZnKt8wEg4i+N/jYVnVG4F0NtsOgSIQgOz1DWtSzD418BRx1s/kju+0uluDkhNl7CzYIvHMOmuLYNyuRhbMV1WAnxs/Pd1DrNuAOpFzayLLP9u0/y4yfb0p4P4Oljg6xrdiVKjEDSlgITWFxIvkGW1Dk52p9/PM3hXg+SBEvrxgdkmiqtdI0E+PnTJ7CZ9Lx53TwsRj0XLK4Wt1dYE1sEekYnDOFEAimOmzcUxW42CHd70SUwegqG2rAa9UiS+FllYsQXpiKXqImG4bfvgCMPwTXfhUu/ML5xo0SontDDpg4tDfnCOC0GTIbSOl6N04984kAMkiS9SZKku4GHgFbghqIfWYnS7NRxy+b5VDtMDPvCxJSr5f959ChvuH0XP3zi2LQbkU9rEplgJeS6qY3h/iF6x4LUOScvUcyvsqPXSRzrT3cuDvd66BkNclaTi4XVds5pcnHpijquXF3Hj25ej8sqvve84mHCvtKYKFXRK8It5IGaFbD4EiHcZuJv+PgTYkXU1k/l19NXklOlsfHjyoI/FMWhlAKvOauBt53bzE1vvQmslXDwflxWE+4JjtuwL0zHsJ+/HUifCh3xhXmta5QLl1an3pDIKTOL31f/IZbWOjgyFcetz8OCKjsW4/j31VJpIxyL88Ar3dyyeT5OZRXUJUoAbXOlLbEhoHvigEJajtv4xCqLlQU7bY+j00nYjPpJnekhXzj7xGU8LoaN2h6DN90O570/7+95NqlJep9ZVucY73HL9f1paORJtuGEyyVJuhOx8P0DwA5gsSzLb5dl+f7ZOsBSpcpuIi6Pj7A/eqgPk0HHt//Wyu92n5rjo5sDEsKthPr+1C0F/iGCkfikE6UAJoOO+VW2jMLt6aODAPzknet54lPbqLCbaKmy8dNbzqWh3IrJoMOk103au5N2TKXkuOmN4ncW8oDZCcvfAKMd0Le/sOeVZXji61DeDBvenfv+MF6SLKXhBDk/4eYLxxIrnS5ZXss3bzwLk8kEK98IRx6m0iynBdy2K2X517pG08qoz7YNIcuwdaJwS6yXMomNBGNdrK2MMuAJ4c5zFV9rr4dldakXD2okiMNs4EMXL058ftvyWnQSLKy206Q4bmkDCtF0xy0h3CoXQsVCaHscEENAk13gjPjC2cN3//4FeO0Polcy37+pOUAVbgadxPqWivFSqS+klUk1ZoRsjttngWeBlbIsXyvL8j2yLOeRd/D6oFp5cQ55w5wa9nOs38unr1xOU4WVXccG5/jo5oC5XFg+GcpUqRT2YSJC7SThuypLahy0ZYj02HVskMU19kQJLBM2sz4/xy04KgRSqaA3imNCVoSbkoN1+MFcj8xO+9PQ+SJs+XjKxGFWpFKMA8m/x82eaRpy1XUQ9rIx/nKacOsYFn9rsgzPHR8CxMTnwe4xHj3Uh9Ns4OymCeuR1AsjvRnqxCqpswydwHg4dDaCkRjtQ76UwQSARTXiYuL9WxemlCubK2088NELedt5zdSXW9BJ47EhCSbkuKml0gSLt0P7LoiGsZsNaRc433uklbueOYEvHJs8KuPZ/4Xn/hc2fhC2fjLn9zmXOMwGLEYdLZU2GsqtjPgjhKNxhrxhqhzaYIJG4WQbTtguy/IdsiyPzOYBnS6ok0FD3hA7W0UcxPYVtaxqKONwz+uw10194y6lUml4/ETmwkt9FscNYHGtg/ZBH25/OLFIOhSN8eKJIbYurcn6WLvJkF+Pm29I7OksFXRG8A+Lf5ud4KiFpvNED1Eh7PoO2GtFBEjex6IKt1IaTsivx80fjiZ63FJYeDFYXJzrfyrdcRv0I0lgNep5rk1c7P30qTbecPsu/vxyF5sXV2HQT3iLVkOuDWaoXyu+RPQ4QF4DCscHfMRl0hy3pgobOz52If+0PX3ubM28cixGPUa9jkaXNT2oNxoE4/hFjTcUxZkcu7P0CvFaPL4Tm0mfMpwgyzK/eKadrz14CJikcf/Yo/DI54UIvuobMzvxXAQkSaK+zMLiWkfiYnHQG9JKpRozhtYlOU3UBtRBX5gnWgeYX2VjYbWdFfVOTgz6prU/8LRGLZVGS6lUOu6eVUqerKVSEI5bNC5z6y/38IFfvsQ9L3aw68ggwUicC5dkF1t28+S9Oyn4B8Feldfhzwp6U6pwA3Gi7X4ZvNNbYE7nHjFJeMFHU6YNc6IrxRy3WJ6OWyx1iEBFb4QV17Dc/TTRcDBld+XJIR+N5VbOW1jJM21DBCMx7nz6BOfOr+DL163mc29Ymf58quNmMAuR7aijfPQwNpOeI325BxRUt2x+ZXq5fnVjOXpddlHUUmlLF26R4ITNCdHEBCkgHDdLOey/V3Hcxl8n/Z4QnlCUSEz0VKaVSoePwx/fB3Wr4fofl9wgwmR8723n8NmrV1CjOGz9npC2p1Rjxjg9XgUliGp5d7sDPNs2yLZlNUiSxIqGMuIyGXulzmjmYmF5Fk4N++no6Uv8f4XkyV0qrRUuxIvtw1iNer7y14N87Lcv01xp5YIl2cWWzWTIneMmy+AbLC3HTW8AvyjTYVbKZ0svEx/bHpvec+76LlhccO57p/a4UiyVyrGcYiEWlwlEYok4kDRWXYc55mWLbn+K69Y+5Gd+lY2Ll9VwrN/Le+/azaA3zCcuX8a7zl/AwuoMvZDRpOEEgLo1SH0HWFzjSGxvyIa6BUHtV5sq86tsqZE5spxhc0IstVRqMMHKN8HhHZQbYylB1er7pMsmWi1SSqVhH/z2neLfb/t1afWG5mB9SwWLahyJfre2fi/RuKyVSjVmBE24TROX1YheJ/H44X6CkThbFEdmeb1wLQ735j+ef0ZQYsMJP3vqOM8dOpn4/xZLINE8Phlqn4/VqOfeD1+Aw2xgfpWdez90Qc7H2s363JsTQmMiM81eSsLNBIEJjlv92WKl0tG/T/35+g6I3LbNH556L19JTpXm7nFTexvtk/2NLNpG2FjGm/VPM5oUCdIx7Gd+lZ1bNs/n+nMaebZtiLXzyjl/cZaLhOQ4EBDl0oHDNDr1ibVu2egc8WM36RNCaao0V9oY8oXHXbNYBJDBYKZzxM+XHjjAgCeUWioFWH0DhD2cF9mTcoGjCrcvX7eGeS4ri1SxKstw/20wcAhuvFMMOZyGqBeLh3tF+4xWKtWYCXLXADQyotNJVNpN7G4XJz01tXxBlR2zQff663MrMcetyx1gIwFiRjv6iI8V5bldHKfFyEXLarhwSRWrGst47JPbsJlEb08u7CYDQ97M4b0JfMrQSik5bjrjuMNlVvqedDpYchkceTjvOIwEu74n4k423jqNYynBqdI8etxUwa5uBkjDYKZ/wXVceeS3HHIPQK2T0UCEYV+YBVU2TAYd33vrOWxaVMW6Ftek2z2A1OEEEMItHmGVsZc93twxM50jAZoqbNm/RhbUEmvHkJ9VjWViohTAYOEPL3Vy17PtrGoo49KVE1ZBLbwY7DVc5HmQn4dXJz7dNuDFaTbwprMauPbsxvH7P/VtOPBnuOw/Ycml0zrWUkDthVYv5LVSqcZMoDluBVBlNyHLIoFcncTS6ySW1TlpzaPf5IzCMDc5bsFIjB2v9vBfDx5K7FoEUcK2EcJjFlsNFtpyuxEAv3zvRm69SMQhlFuNeYk2UHed5hCHakmypBy3JOcl2SFbejkERqBzd/7PNXAE9t8r8rWmk2RfiqXSPBw3tbcxY4+bep/VN2GWIlgP3QuMb+iYXyViOHQ6iZs2trBiwrRnGonhBOX1VrkIgAX6QYZ8YaKx7IMdQrhNr0wKoscNGO9zU3tajRb6PUGqHWYe/PjW9GEevQE2fZAV3udpCR1LfPpYv5fFtY5UIXnwfnjia3DW28VU8mmMyaCjwmbk6WODiVgVDY1C0YRbAVQr/QqbFqWepJbXOznU8zoTbnM0nPDlvx7ktnv28rOnjvPpP76aCD/uGQ3ikAL0hG14ZCtN5hxuWIHYzXr8uaZKfUqzv62UhhOShVuSaFhyufidHnwg/+d66ttiuvCCf5resZTiVKkcz+k4Jhy3LOV0U9M5vBZfQF3bH0CWOalEgcyvmuKJfKLj5poPQCP9yLII9c1G54i/MOFWpQo3pZ8uMu649Y2FqM0Scs3GWwnqHbw3fm/idXqs35voLQWg5xX40wfFZPObvl/yE6T5sKDaTrnVyM//8TyaK/MIotbQyIEm3ApAnSzduDBVuC2ucTDoDaVMT53xqCeSWSiVPvhaD79+/iSxuMxDr/VwzdoGvvSmVTx3fIi/vNqDPxxlNBDBTpDugJ4R2UG1rsjCzWTI/ftWS6Ul5bgllW6SNzpYymDxpcL9yGeLwuAx2P9HOO990//+SnKqNJpTuKm/94w5bgouq5HfxC7FNdYKHc9zcoLjljfJcSAgnE2jnZqoGMRRF5pnwheR8QSjNFVMXzyUW42UW43pjptBOG512QaALOXsb3o7V+p2E259lLFghH5PaFy49R2AX98oLmzedvfUJpJLmJ/esoEnPrmNS5bX5r6zhkYeaMKtANQJoYnCrUJp/B0LlEa/16yQCOCdGcctFM188v7RzmN85O69fPH+/fzllW5G/BGuXlvPLecvYM28Mr718GG63aIs6tQF8WLFjZMyeVTEVISKM+1rMxkIRePZS1X+UuxxU1wigyV98f2q62CsE7r25n6eXd8R4v2Cj03/WE7TUqk6nJAxx02hzGrkz7EtBA1l8PwPaR/0Ues05xx6SSPhuCm/K0kCVwvl4R4ABryTv/4GA+JvsxDHDUS5VBWe48MSquOWXWwdXfI+WuUmjPe9n86jrwDKnuDWh+Gua4RIvuVP4Kwr6BhLiVqnJfcOVg2NKaAJtwK4aWMzX7l+TdqbVblViJiJgZtnNDMYwLu/a5S1//EIL3ekZj93DPn51sOtbF9RiyRJfP6+/Rh0Ehctq0Gvk3jzuiY6RwK82ukGoEIfwiebCZlc6Lr3wh3b4S/F6ZmxK43p/mz5fb4hMNry29s5W6gCINME6PKrxPDCwfuyP8dQG7z6e+G2OQpwFUpyqjT3cIIavOyYbDgB0ftqsDjYXXU9HPorgb62qbttkOJwJXA1Y/d3AzAwlk24Cee0EMcNRLlU7Sc92iXK/zG9mSFvKGfkjr3Mxa2RTxBHYvmfr+LHxv9m61M3wW/eBo56eM9DULO8oOPT0DjT0YRbASypdXLL5vlpny+zvp7d8erjAAAgAElEQVQdt8K/57tfOEk4Fmdvhzvl892jop/mvVsWctWaeryhKJsWVVKmLMRe3Sh6tB47JDZZ2Ajiw4pkqxKN9iCa5weOFHyME1Fzq7KG8PpLLMMNxn9vmYSbtUJMl77yGxGyOhlPfks8TyFuG5TuVGmewwm53DOXzchjzmtBp2fb8G+n3t8GqbtKE0/cgtEr9iNnd9xU4VaY4za/0kbnSIBILM5Th8S6rVNjceIy1OYIuZ7nsnBKruPFK/7MgbrrOVvXhkmvE9OjH3zqtI390NCYTTThVgSK7bjJssyTRwaIx/PoPZotZmg4wReK8sA+4R60DXiRZZmnjgwQjcUTjdeVdhPv3SLe4C9fOV5SWaUItyePDCARxxALEDPYcVYq97niq6J5/qlvF3SMmVDDV7OuvfKV2NYEyC7cQOSx+Qbg1d9lvr3jeXj1t7DpQ4WXt6QSdNzyCOD1KcMJk+a4KZRbjZyMlBM562aujT3KWvvo1I9n4nACgKsFKThKgyWcWGieicFAvKAMN5WldWLDSPugj4ER8T280iuEfdbhBEjs+22PVHFP9ce41vh/6D7wKFz4z+mleg0NjYxowq0IqA7QWLA4vTovnBjm3Xe+yPPKYuqSQD8zpdIdr/XgC8cosxho6/fy8ik377rzRR491M+QItyqHSY2zK/gvtu2cHOS41lmMdJSacMbitJiF6L2fZeuZdn2d4lJx823iTT//X8cX/M0Q6hREFkXzZei46ZTTuKmSYTbwotEVthzP0wfUohFYMcnoawJLv70DByL8nZUSsItnx43xXGzZhlOAHBZTYwGIpxc8xFkJLb13TX145k4nADgagFgtW00h3CTC8pwU1laK/5WjvZ7GXKLvMq93fkJt1qnGb1OotsdoMsdYF6B7p+GxusRTbgVgWI7bm0DosG+N4+k9FkjUSotULi92sPCajtXrq6nbcDH3pOixNnlDjDsFc+tNvqe0+xKy1lb1SBct4VlohHbaHUitWwSbptOB6uuFxEPx58o6DgnopbJsk6W+oZKa6IUsve4gWh+v+BjMNgKL/18/POyDA/9G/Tth6u/MTPriEq1VJqrxy0cw6TXYTJkfzsttxpxByIcC7q4J3YpLR33iUnKqRALieNJnnRVhNsy83Aewq1wobS4xoEkwUvtI4SCotftpS7xMdc+YINeR53TTPeoItxcZ8bkqIbGbKIJtyLgtBiQpMJ63PacHOZgd+btC2p4Z67MplllhoYTjvV7ObupnCW1IlLlySOi+bnfE2TIF6LMYsgaiqv2ubU4lOnOiU7SvPVij+axae7hnITEcMJkpVJZVhy306xUCrDmRrF4/qHPiJ9bJACPf1UIuQs+JvZQzgSn6VSpLxRN/P6zUW4zMhaI0DHs4/vRG5At5bDjU/nFrahEQ6luGySy3BYahnJOlc6EcLOa9DRX2Hh4fw8WSbzeR0LiNVmdxy7ORpeVbneAbneAxnLNcdPQmCqacCsCOp2Ew2woyHH7t3tf45N/eCXjbeoo/lApCbcZWHkVjMToHg2wsNohIgKAp4+JCI2BsRBDvnDOJc2r5wnh1mRThJt5whognR4WbxcCZConzBwkhhMmK5WGfSI6oQQct1PDYqdkNBbPT7jpdHDDz6C8CX59A3xjvoj/WPsPcNmXZu7ASnGqNI8AXl84mlesh8tqxO2PiHBuWyW6y74EHc/CvnvyP55YODV7D8TFgNFGkzQwqeM2GogQiBY+UaqyrM5B92gQM+L1HsREpd2U03UEaHBZOdg9RjAS10qlGhrTQBNuRaLcapy24ybLMp0jfg71jHFi0Jd2+0llFH+klISbTi8ckwKGEzqG/cgyLKi2sVgJ5VS1VZ8nyLA3nHNJ85rGciQJmhOOW4b9jUsuBW/v1MtUWVAb0ycdTlAz3Ow1mW8vAk8dGeD6Hz6Tlom347Ue7nq2nbYB33iPW66F8NYKuPUJeNPtcM5N8O6/wlvumNoe01wkSqUltDkhjwBefyiWl+N26cpaonGZ+/Z1iYnSdbdAywWi5Dx8Ir/jyeS4KVludfF+vKFoxj7LzhHxnjETjhuIiXoAu058rRDGnP1tKo0uS6L/t9GlCTcNjamiCbciUWYxMhacnnAb9oUJRsTJ68HXelJuk2WZjiEh5pIdty53INH7NmfoTQWVSo8PiO9rUbWD5gorRr1ooq4rM9M/FmLIF8q5pLm2zMK9H76A7QuVE0Km3qvFytLqY49O+1gnoi4YzzicEAnAk8oka9m8Gfuaubh3byf7TrlpH0zdGtHWn9QjmavHLRlrBWx4t1hFtHDrTB8uSOpwQimVSvPpccvPcdswv5Ib1s8TFydVNsXJ/Kn4vu99X34XPbFw6kSpiquFinAvAIOe9Ndg54iI0plJxw2gxTbuuOWKAlFJLo/O04SbhsaU0YRbkSi3GqddKu1yizdZg05ix6upwm3IF07EDyT3uH3xvv18/LcvT/NoZ4gChZvqLi6otmHQ61hQZceol9i2rJZ+T4hhX5gqR+7IgPUtFZjjyg7FTIKkrAFqVsCJp6Z9rBOZ1HHz9MEvroZ9v4Yt/wwLL56xr5kNWZZ5tk1MHR+fIOiPKz/nvrGgWP4NqXtK54oSK5UGIzHisSjBeI44kFA064L5ZD579Uqq7CbWzisXn3C1wHU/gK49cP9tufe0RoPpjpvyPI6giNFJrKNKYly4zYxQWlrrZI10nLeE7ydQt4EY+ik4btaM/9bQ0MgPTbgViTKrgbHA9JyDLuVN9tpzGjk4oVyq9rc5zIYU4XZswEuPe46nTA15CLfDD8Ivr4Pf3gxP/zd4+xM3nRj0UuM041TiVM5bWMkFi6tprrQyGogI4WbP7+RAWBErmUqlAAu2igyyGdqtqtdJWIy61B63oTa44zIYaIW33wOX/2fOTLCZ4li/N9HvdHxCuV0Vcn2jU3Tcik2JTZX+9ZVudMRoHw5kvZ8/HEvk+OWixmnmmc9s530XJgXNrroOLv0ivPYH+NvnsvdeRsOZhVt5M8bwKBX6ALuODaTd3Dnix6Kn4Aw3lSU2H3eavkvIVIH+prtxmA3CRcyDhnLhzFmN+sR6QA0NjfzRhFuRmAnH7f0XLgJSy6Udw+IkfHZzeUK4RWJxOkcCDPvD2XdlFptcjttDn4Hf3iT6eQaPwKNfgu+tgj+8B9qfpn3Ax8KkNPmvXb+Gu95zXqIEE5fJWSpNoO4knTicoLJwK0R8+e3hzBO7yTC+OWHwKPziDeJr/OMOWHHNjH2dfHhGGeqwGHUpwn/EF2bEL/4ue8eC+fe4zQYlFsD7pz0dAASi2XPPfOFoYjglHyxGfXqW2oWfgE0fhhd+DH/52OQXFLFQ+nACJCJBrpgXYefhTMItQLVVKjjDDYB4DOsDt1JtDGB+1x8wuRp46ONbef/WRXk9XHXZGl2WmTkeDY3XGZpwKxL59Ljt7RjhSw8c4KkjqW+0Xe4AdpOelQ1O1rW4EuVSXyjKySE/kgRnNbkYDUSIxOJ0uwPE4jKyDMP+ORxY0JvGA0In0n9InJTW3QIffQk+uhtu2w0bPwBtj8Fd1/DNvvfzTh4Er/h5SJI40SSXYCYtlUbDIlR3tFOstBpRmr2Nk+SLzb9QfGyfwXKp2YA/HAP3Kfh/1wrn6N1/FREks8wzbUM0V1pZ11yRUio9Pjj+776xUNJU6SQCdzYpoVJp54if3SeE+PXlMM79ofwdt0mRJLjq67D1U7D3l3DXG2G0K/1+mYYTIBEJsr0+SGufJ3Hxp9I5EqDKOkNv97u+B+270F3zPczz1gLQXGnDYszvZ1BhM2Ix6rQyqYbGNMn/MlFjSpRbjfjDMSKxeMbcsUcP9vH+X74EwIAnxEXLxqcNu0YCNLqsSJLENWsb+OqOQ3zr4cP8+Mk2nGYDDWUWGpVyw4g/TPvQeE/LoCectvR+1sjmuO36rhBRl395fLVNzTJxsrr0iwT2/RH3X/6Ha3tvh+/cDs4GMYFpdbFetnGHsQ8rYdY+bYLnoxD2Q8SvfPRlbmi3VU1emrRXQd0aOLELLvrXGfn2bSY9cf8w/PoDIv7jPQ9C3aoZee58GfSG+Pyf9/P44X7eem4zkgQPJTm2bcoAyMJqu9LjVoKOWwmUSv+8tws94jj8OYxz7xR63LIiSXDpF6B2JTzwMfjhJtj+72LbR3JOoiHD61tx3NaXjQG17Gzt5+ZN41tFOkf8bKydAXer7wA8+U1Y8xZYd/O0nkKSJC5eVsO6lorCj0dD43WIJtyKRPKi+UzZY+oV8TyXNc2ZS14Fc7Ui3H60s43VjWWcGvaztqmcSqXXa9gX5uTQeClsMEsA587WfoKROFetqS/sm5uMyYTb8HGx3P38j4KtMv12o5VjDdfy5nAld7/RxhbdfnGC8A9D0I0tMECdFCCICZ25EhxOsXPUaBNTo0YbmGxCGCZ/rF6W/XgXbIU9v5jcxZgii6osvLXtU8i6E0jvuh/q1xT8nFPll8+d5G8He3nfloXcdskS7t3byYg/wogvTIXdRNuAF5Nex/qWCp46OgDWSkACe+2sH2saOh0glYTjtvvkCKvr7eAGbxbhFo3FCUXjeU2V5s3aG4VLu+OT8PBn4NkfiF2wZ98k/lYt5emPsVeDwUp1tI+mirXsbB1ICLfRQARPMEq1tcBdoPEY3PcRsLrg6sL2/f70lnMLOxYNjdcxmnArEslrrzIJN7WJfV6FFc+EnaZd7gDnNLvE7S4rFy2rwe0Pc88HNmPQiavmlzvcAAx7w4mBBcgs3GJxmU/+fh/37evGZtJz5eori9NbMtlwQuvDIptr462TPnTIJ47b0nw2zL8k5TZdXObNn3+IaFzmhbddiiPP2IGcLNomyrcdz4l/F8hXXX+hklf4Y92nuHHBloKfbzp0DvtpLLfy+TcKp29RjSgVHx/0scFu4viAj/lVNua5LAx6Q0QXX4Hhw8+Cq3lOjjcNnb4k4kBG/WEaHQZwgy88+bCAP6IsmM8jx21KVC6Cd/4J2h4XbvXfvyB6QiWdCEKeiJLlJo12cP6iKv5+qA9ZlpEkKXFhV20t8DW/5xfQsw9uvFM41hoaGnOC1uNWJMqsQhNPtmjeH4qh10nUOMx4khw3XyiK2x9J6f+4413ncv9tW3CYDViMeixGfaJJf9gvHDc1DymTcHu10819+7pZVufAH44Vb8fpZI7bqedFKSeLOFDFqzpRmoxOJ1Gj9LlV2Ap0DZJZcKFozp+J9VeH/krlntvZV/MmPnViHYd7M68rKzadI6mLuxdWi941dUChrd/Loho7deUWZBkG/NFZL+dmRdKXRKnUHYjgMou3R08kVbgNekO85xcvcqhnjFGljlqW4e+2YCRJhEW/50H4yAuw9RNQv3bySBlXC7g7OHdBBW5/JFEWV7ePLHEV8HbvHxZrzhZshdU3TP95NDQ0CkYTbkUi16J5byiKzaTHaTGkOG7d7vS8JZNBl+aQJYSbT/S4rW4sw2TQMehNF06vnBLu3AcvWgxAW3/6NoYZIdNwgizDqReheVPWh44Lt8wmcG2ZhTKLIa+VOnljdkDLZmgrcOH8UBv8+UPQuI7mm3+IJEk8vL93Zo5xinSO+GlKEv3NFVYMOonDPWPs7xrl+KCPzYuqqFP6IHtH5zhCZiI6Q0mUSkcDESqUZn5vkuMmyzKf/uOrPNE6wJNHBhKZaU2VRW60r10B2z8vtldsmsS5drWA+xQb5ot2hD0nhwF47FA/a+aVUWEp4LXz5LcgOApXfUMISg0NjTlDE25FQr0Cn0y4+cOiodlpMaT0uKm9b7kmrtT8o0FPiI5hPwuq7dQ4zAxm2FX4SucoNU4zFy4VezKLtmEhk+M2ego8PTmFmzckfgaZHDeAeS4LdTNVIk1m8Xboe00E5U6HWATufb8o8b31V1S5ylnX7OKxQ/25HzvDRGJxeseCKaLfoNexbXktv9t9ih88fhSLUccN65uoV4Zb+samv6KsKOj0cy7c4nGZ0UCEcrMQKJ4k4fb7l07x+GHxu20f9CWEW0vlzGwkKAhXCwSGWVwWx2Uz8lL7CEPeEHs7Rrh0Rd30n9d9Cl76OZxz85z0bWpoaKSiCbciUZ40nJAJnxLa6bQYCUbiRJT8tSHFMavJsUzdoNfhshk50D1GOBpnfpWNaoeJgQyl0ldOuTm7yUWt04zTbCiecDOY04XbqRfFx+aNWR/qCUaRJLBPEqvw2atX8v23r5uJo0xlibL+qu3x6T3+yW9C916xBkopBV+6so7XukbF1OYs0jsaJC6nrzX6zNXL8Udi/O1AH9edPY9yqzEhgmf7GHMi6ea8VOoJRZFlKFdKpcEYBJRtJU8eGaC50sr6FhftQz5ODvkx6iUayksg2kKZLJVGT7GhpYI9HSM80TqALMNlKwsQbk9+U3zc9pkZOEgNDY1CKapwkyTpKkmSWiVJOiZJUtqrXpIksyRJv1Nuf0GSpAXK5xdIkhSQJGmf8t9PinmcxaAsR6nUHxKhnWppUC0VupX755NwXmkz8XhrP5IE5zS7qHaY00qlo/4Ixwd9rGtxIUkSi2odRXTcjOnCreN5MeVZuzrrQz1B4UBONjTRXGljVWMR1jLVrQVHHbTumPpjTz4nGsfPuRlWX5/4tHqSnGnXbcQX5tm2wUlDlk8pi8TnTVhrtKTWyU0bhah852YxaVhlN2HQScXrd5wuJVAqTfStKY5bVNYzouQjjgWi1DjMLKx20D7o59Swn6YKG3pdCZQPKxaIj8Mn2LCgguMDPr7/2BHqysysmTfN185IO+y7B859X+ahCA0NjVmnaMJNkiQ98EPgamAVcJMkSRO7oN8HjMiyvAT4b+CbSbe1ybJ8jvLfh4p1nMXCYtRjNugSa4cm4guNO25AYkBhNBBBkiYvGSZTaTchy/DFN65idWO5ItxSv96rXaK/7ewmMaW6uMaeWOYejMS48r+f4pEDM9SPpTelL8ru3A1NG8Z3Yk7CWDBSnAbvXOh0sOp6OPKI6OHJl+AY/PlW4XJc/c2Um5bVOWiqsLKzdWaF2/cfO8o7/u8Fzv/G4+xuH067Pds+ys9fs4rf3bqZtU0iSkKnk6grs9Djzr7OadYpgalS9WKr3CLc3xi6ceEWjFBmNbKgykbvmAi7LYkyKUCV6GFl6BgXLa1BksBhNvKNt5w1/Sny534oXNAtH5+549TQ0CiIYjpuG4Fjsiwfl2U5DPwWuG7Cfa4D/p/y7z8Cl0pn0A6ULUuq+eur3YSi6Q6CL6nHDcYdt1F/GKfZkNcV/Ds3z+ezV6/gPVvE3sNqp4lhX5h4fLwnRx1MUE/Yi2sc9IwG8YaidAz7ae3zcMeuE4V9oyp6U+qqHlkWGW41K3I+1BOMTjqYUHTW3ihWCR1+ML/7yzLs+IRItn/zz9LCayVJYmmt+DnPJAe6R1lYbWfEF+YJpc/qjl3HOdrnAYRwkyQylu0sRj2bFqVGOCystqftMZ1zSmCq1B0QIs1pEq/BGHpGfOLv2hOMUmYxMr9axKwc6/eWjnCzlIs8vqFjrJlXziv/cQUPfXwrlyyfZkaffxhe/jWc9VYoa5jZY9XQ0Jg2xTxTzgNOJf1/JzCxQz1xH1mWo5IkjQLq2WWhJEkvA2PA52VZ3jXxC0iSdCtwK0BdXR07d+6c0W8gE16vN++vs94R5XFvmO/87nG2zEt1kwbdfpyyj+OHRWzErudfYrBKT2t7ELMUz+truJT/du4UP2Z3T4RYXOavj+6kTDnpPPlqkFqbxMsvPANAsF8IxD8+/CTukBB4L7YP8/sHH6fWVpiOX9o3SE3Qx7PKsYdG+yE0xrGhCJ05vp9TvQHiMrPyO0xDltlkqcX/1M94zZ37BFXf8xgrWv/AiQU3c/J4AI7vTLtPcCxIrzu/36NKtr8tWZbZ3+nn/EYDHh+8fKSdHfoevvqYnz8828pnN1nZcyhEhVni2afzW+NljYTY3RPl8SeeQFci10ubw1Hc3d0czuPnFhgdZP8fvo4h6idoqWW0fAWyrjDX1hge49SpE7xDfxLHrjZAOG7PvLSPaJeBwTEfHmuI4STHMzLSw86dgwV93ZniHEMNHN/Dvgw/v6m8dwHMb/8dCyN+XjRuxj8Xr8s5ZKo/q9c72s9rahT68yrVAN4eoEWW5SFJkjYA90mStFqW5ZRwLFmWfwb8DODcc8+Vt23bVvQD27lzJ/l+nYtlmfs6nuK5YR2fe8eFKeUK+ZlHWdhcy9ZN8/nm7qdZtGI121bXc9eJF2nQh9m27cIpH5vnlW7uPvwyy886l+X1wgX63v6nWT7PyLZtQjM39Xv4wctPUd6yHCkQhb0HAOgyNvHWbTk2DeQi+DcYeibx83npL3cCsGTDdpas2pb1od95bRe1Tgvbtp1X2DFMl9jNWJ+5nW3rlmTv5Rk8Cs/cAQu2svBdP2ChLvMwxROj+zm4rzvvv5VfP3+Sh187zK8+enHGslaXO0Dgb4+zff0KPPu6iEsSC1evhsd20ToSx9qylmjrERbVyWzbdkFeX7PLepK/n9zP8nWbEzmAc84rNuprq6nP9XPrfInArz+INZhU5jfaYP4FIkx5/haoXQXGDJPI0ZDYaevugMEjMHAYBlrFf/5BtgAYIe6xEWk8l/0nFnLBgiVcvHk+wb8/xIrF83nzxYv50nOPAHDJxrVsW12kbSRTZWwDtD6U8e9uKu9dRALw4vtg6RVsvOZdM3qIpwNT+llpaD+vKVLoz6uYwq0LSE5cbVI+l+k+nZIkGYByYEiWZRkIAciyvEeSpDZgGfBSEY93xpEkibef18xXdxyidyyYUsLyhaLYTBmGE/yRxETqVKlWJlEHvSGWI4TbqWE/a9aOu0gtlXaMeokjfV6isTgWo44N8yu4b18X/3J5gcJtwnCCJahEbCjTbtnwBKMsqp7D64hz3wvP/q9YoP3G72W+TzQEf3yP2BV5w8/Gl6JnoNxqZCwYIR6X0eVR9v79S6d4tSvKk0cG2JahtHW4R1yzrKh38mK5lVc63YnoGINO4msPHqJrJMBWJfIlH5bUiHDeo32e0hFu+ZRKfUPwy+uRJAu84w9QvQT6Dwvn8/gT8Mjnx5/L2SBWNMWjEA2K3ba+ASApVNfiEuX8FW+AmhU80O3kG7tlHv/czRj0Ojr//SGGfWFl+lumzGKk3Gqk0i5aE+ZXlUipFKBqifj+Am7xfU+XffeAf1DrbdPQKEGKeabcDSyVJGkhQqC9HXjHhPs8ALwbeA64EXhclmVZkqQaYFiW5ZgkSYuApcDxIh5r0VjZIKa5Tgz6EsItHpfxR2LYJxlOyNRcng8LqsUJZN8pN1uWVOMNRRnxR2hOiocwGXQsqnZwpNeDUa+jqcLGRUtr+PpDh0V21TRFIwB6sxA3sgyShCWoNOdXzM/+OOa4xw2EuFz3Ttj7S5FQP9F1i8fh/tug9zW46XdQ1pj16cqsRmRZfF/lOSaEg5EYB7uFMPvOI61cvKwmzXU73Cv62JbVO2kot/DwgSCnlAyxT16xnG//7TBxWdyeL0vrxH2P9XszisU5IZ+p0me/D2Evr573X2xcdoX4XOUiIbwAxrrFGrP+Q6IPMegWFxV6s9hxWzZPRLeUN4t9to7alFDZ/Q8eYsjQjkXZP1pmMeD2RxKvUXUryoIqG8O+cMrra86pWiI+DrfBvA3Te454TAwlNK4XzqWGhkZJUbQzpdKz9lHgb4AeuFOW5QOSJH0ZeEmW5QeAnwO/kiTpGDCMEHcAFwFfliQpAsSBD8mynD5GdxqwUGliPjHo44LFwg0JRmPIMhnjQEYDkbyiQDLRUG7lvAUV3Lu3k49sW5w4sTdPSHVfVu9k78kRXDYjTRVWFivOy/EBL+taKhL3i8VlXul00+MOcsHiKirsOdZN6U2ALN749QbhuJnLhKORBVmW8QQjeU3SFpWtnxDN2Ds+BW+/e9xRi8fhb5+D1/4Al34Rll+V86mSN2fkEm4HukeJxmU21OnZ0zXGztYBLlmRKqQO9wpXrMxipKHcQjgaZ3/XKBajjg9dvIj3bFlA/1iIRlf+IcWVdhNVdhPH+osUDzMdck2Vevvhxf+Dtf+A3z6Jk1vWCGveMu1DGPWnvgYrFGdNDcpW/06X1TnpHQ1iN5dQx4kq3IYKEG7HHhPC78Y7tS0JGholSFFz3GRZflCW5WWyLC+WZflryue+qIg2ZFkOyrL8D7IsL5FleaMsy8eVz98ry/JqJQpkvSzLfynmcRaT+jILZoOOEwPj03vekDgx2cwGjHodFqMOj1JWc/vDuKzT38d544Ymjg/42HfKnRQPkeoIrKh30uUOcHzAR3OFjcW1Qri1DaROGP7imRPc8KNnue2evfzimTwmT/XKyU4pl1qC/cLJyvHmH4qKEtScOm4gjvXKr8GRh+AvHwPvgCjB/ebtYhn9pg/DhZ/I76mUnaqT5fgl83KHmPx9xwoTZoMusVsymdbeMVY2CIesXnFud58cptFlRZIkLEY9LVU2DPqpvaQX1zo4WkrCTdKDnDmnDhDCOuKHi/+taIfgDoRTnOcKm4kRf5jRgHjdlil/p5++agW/en/2jSCzTsVCEd8xeHT6z7H7DpFtuPLamTsuDQ2NGUPbnFBkdDqJhdV22ofGRZE/JEpB6pYAp8WIJxjFG44Sl/ML352MN6xtwGLUce/eznHHbULpdZlSIgtEYjRXWmmusGLUS2nBvKeG/TjNBsqtRoZ8GZbHT8SgbHuIiSw3IdzyK5PC+AlxTtn0Qdj6SSEQvrMEfrRJbFV4w3fgqq/n7UDk2lWbzMun3MxzWamy6jirqZy9HSMptwfCMdoGfImBE9VVOzUcKLg3bWmtg2P9XkRbaQmQa+VV+9NQs1L0tRWJ0UAk5eKpcoLjpoZrV9pNCbe6ZDCYxGtu6Nj0Hj/SDkcfgfXvHr8Q09DQKClK4Ex55rOw2k6rkrUFIsMNwKb00KM3d/4AACAASURBVKiL5hOJ7QX0mTktRq5YVc+OV3u49uxGbCZ9YiG9yoqkPqimCuHSLKiy0zbBeRn0hakpMxOLywlxlZWE4xYhHotjDvSB6w05H6b2DjlKQbiBKIeuvgFaHxIN3iuuydnTNpGpCLd9HW7OaXEBY6xvqeAXz7QTisYwG4Swf+CVLmJxma1LawASe0Yh3U2dKktrHYwGIgx4Q9Q6i7ALdqpkK5XGonDqBTjrbUU9BLc/kvJzrXGY2d81mnSBUeKCpma56O+bDnvuEo7dhn+cySPS0NCYQTTHbRZYUG2nY8ifWFXkV/YeOsyqcBMTiOpJ3lXIgABw1Zp6RvwRdrzWS1OFNa3RfZ7Lik1x+9TG6kU19jTHbcgbotpuVoRlbgGCXnHcokHufeZVjPEgA4bcOxLVE6LTXEInxPo1cPG/wsYPTFm0Qf7CbcATossdYF2z6ANc11JBOBZnf5cYVpBlmTufbmdFvZNNCysBqLabMerF73S6gywqapCs6s7OOdmmSntfhbAXFhS3YX5sQp9pfbmFAW+IYWUrSUk4w9moPwsGW8UE7VSIhsRwzvKroXxecY5NQ0OjYDThNgssrLYTjcuJ+IbxHjchnsoUx83tV/eUTr/HDeCiZTWYDDoGvaGME286nZSYKFQHFxbXODg55E8suwex8L7KYcJpNubnuFnEBK0cHGXX7r0AHPBXZHsEkCTcSv2EOAXyFW4dimBS+wzXzxcC7mWlXPpc2xCtfR7eu2VhQoCr66qAgkulahld7Yecc7JNlZ58VnxsyS+nbrq4J0xXN5RbkGUSvYCFOOKzQsPZok+w/+DUHnfwfvAPwXnvK85xaWhozAiacJsFFimuhrpeaLzHLblUOu64FRTJgXDyLlwiJlibJ1nHs6axjEq7KfG1Ftc4iMblhJAAGPIpws1iSIjNZD70qz38bnfH+CeswhE60dFJZKgdgN1ue87j9UyY1jsTsBh1mPS6nMKtX1nyXqeUKWudFpoqrOw5KYTbffu6KLMYuPacVNevURlQmLhQfqrMc4m/j9IRblmGE04+I2I/irh+KRyN4w/HUlxvtTR9pM+DSa/DbCjxt82Gs8THnlem9rjdP4fKxbBw24wfkoaGxsxR4u9AZwYL1EgQZWpzvMdNGU5QHC11R2IhwwkqV6wSJcrJSmmfumI5v/nA5oSLk5gsVVyFaCzOiD9Mld2MQ3EEk5FlmUcP9fHMsaHxT9qEcHv+wFHq9aLU93RvbhfNEzrzHDdJkiizGnMKtz5VuJWZE5/bML+Cl06OIMsyezvcnLegEosxNexXFROFlkqtJj1VdhOdI6VSKtVl7nGTZeh4XmxGKCKJiydbsuMmfsZH+ryUWQ3TX9g+W5Q3g7ViasKt51U49bwIotZppwUNjVJGe4XOAlV24VodHxSiyK8IlfEet9RSaaGOG8CVq+s5p9nF+YurMt5eYTclphRB9LiByJsDGPFHkGWodpgoU3rwkvGGokTjMoNK3w+QcNw6Ojs5t0a4JvtH9PR7si9bPxNLpQDlVgNjuYSbJ4RRL1GRVB7fvKiKAU+Il0+5Odbv5Zzm9By8RTV2nGbDjAwUNFVYS8hxm6RUGhiBwLBYY1VEMrneqkgeDURKfzABxORzw9lTE24v/FSsDFt3c/GOS0NDY0bQhNssIEkSy+qcHOkTws2nDCeoPW5Oi5FAJMaQN4zFqEtzV6ZDhd3EfbdtYXVjeV73d5oNmA06hpXYjyGfEGRVDnOiVBqPj0dGjPjECW7AkyzcRD+bLjhCs9lPUO8ghp497anxFhNJTJWWUpDpDFCep+NW67SkrMW6QBHbP94plpwnhyKr3HrRIv76sQvR57FOKxdNFTa6Ska4TTKcMKLkCFYsKOqXd/vF33+ycCuzGJKie06Tv9H6s0SPWyyPoSLfoAiXPvvtidewhoZG6aIJt1lieb2T1l4PsizjC0Ux6CRMSliqejLoHPHPiNs2HSRJuD4J4eYVH1W3UJbBHxk/oY4oJ7gUx81oIW6wUiF5qdJ5iJnKMRt0/OSp49z59ImUwYdkPMEoNpN+yuGxpU651Zgof09G/1iI2qQyKUBLpY15Lit/P9iHJMFZzeni22YyML8qd/9gPsyrsNLpDqQI8zlDmiQOZKRdfKxYWNQvr/79V9nHfyeSJCVct5IfTFBpOFsEYecTC7LnLpG9uPGDRT8sDQ2NwjmzzpQlzPI6J6OBCP2eEP5wDJtJn+iVUU8GrX2egrYmFEqF3ZQmyITjlrpPFWBYud+IP5IiyEJGFxWSl/K4m4ipnHedP58ed4Av//UgP32yLePXFeuuThMnYwrk67jVTSh3SpLEliXCdVtc4yh6ea6pwko4Gk8V4XOFTi9WjE1kWHXccgc6F4Iq3Codqa9Dtc/ttCiVAjQrGx3ad2W/XywihhIWXQK1K4p/XBoaGgWjCbdZQu0nO9zrwRuKppQFL1paTYXNyMkhf869lsWk0m5Mc9yqHabEsSYPKIwkbVJQ7wvg05fhwoM14iZiLOPfr1nFi/9+GdesbeD2x45xNCmIWEUsmD9NTohToNxqTIQqT0bfWDBlMEFF3Wu7LkN/20yjDjh0ukugXDppqbQd7LVgmhmXcTLUDSGVEyJ5xh230+QCw9UMNSvEFoRsHHoAPN2w6UOzc1waGhoFowm3WWK5kpt2pNeDPxzFliTcasss3H7TOiQJKuZQuFXYTIkBiSFfCINOosxiTLhhyY7bSJIgSe5zc+OgRu9DHxgmYixLfP5L167GatJz++Ppq3jGgpEzrr8NhHDzTOgNTCYQjjEWjFJblj5gsGVJNTaTnguXVhf7MBNbAkpiQCFbqbSyuGVSEBckVqMeqym1z7RBEW6n1QXGkstE9l0oyy7aF34qys9Lr5i949LQ0CgITbjNEhV2E7VOM4d7PfhCsUSzs8rWpTX87JZz+eglS+foCIVwU0ugQ94wlXYTOp2UOFmNTeK4JZfYBmMOqvVe8A+lCLcap5mzm12cHEpdZO/2h3mpfYRVjWWcaZTbTMgyk4YXq9O2dRmEW43TzIv/fhnXnj31rQ1TRQ3xLYlIkMl2lY60F30wAUSpdOKKOEhy3E6nkv6Sy0Sf22Tl0o7nxQqxjbdqESAaGqcR2qt1Flle7+RIn+K4mdJPAJevqmNtU35ToMWgwm5iNBAhGosz6A1T5RAlPPVk5U0Wbv5x4ZbsuPWErdTFB0COETGmfi91TnMit0zlj3s6CUXj3LypZca/n7km1/aEfuXnlqlUCmLKdjYyw+xmAxU2Y2k4bjpDeqk0GobRztkRbn4ROj0RNfD4tBlOAJF5Z7TD0b+n3ybL8OiXwFEHG94964emoaExfTThNouISBAPI/4I9hIsDVbajMiyEBpDvhDVyglsfDghVbipTs2A4riFo3G6QhYMshAqYVOqi1ZXZmHAEyKmlA5lWeaeFzpY3+LKO7bkdCKXcBsP35375e7q72bOkTI4bu4OQC76RCkIx60iw8o5tQ8w020li8EMSy6F/fdCcDT1tqN/h47n4OJPF71vUENDY2bRhNsscuGSakLROMf6vdjNhWe1zTQVSoloxB8Re0rtqnBL73Eb9oVpdFlwmg2JE37niJ8R2ZG4T5rjVm4hLovl9QB7O9wcH/Rx86biTgrOFZV2IdwGvJkDiPvGFMdtBkJ0C6XMYkz5/U6XQW8orRw+JXS6dOGWiAJZMP3nzZPkv/tkltY5+b93ncsVq+uKfgwzytZPQNANz/94/HOBEXjwk2J92HrNbdPQON3QhNsscsmKWn79vk2sqHfOyrTgVKlMCLcwQ95QolRqM+nRSamOm9sfocJmotppTvS4nRzKIdyc4vl6FafpUI9YizXZdofTnaXKQMqhnvFJ2vv3dfGr59oBsafUbNCVxKSi02JgLJC5F28qfG3HId54+9OJDRxTJlOpVA3fnY3hBH84cQEzkctX1WE2lN4FV1Ya18GKN8JzP8QS6INIEP78YRjrgRvuAP1pVPrV0NAANOE261y4tJqH//ki/nFL8U9CU0UtA50Y8OELxxK9V5Ik4TCnLppXS0o1DnPCcXvm2CBj0vgareThBBgvCapOU9uAF5tJT30JlAqLQZnFyIIqG/u7xstU333kCD958jggBGxtmbkkdl+WWY14QumO29NHB3nxxHDezzPoDeEJRfnQr/bgD09DCGaaKvX0iM/ba6f+fFMgGInhD8cyDiec1mz/PMSjnLf7n+D7Z8ORh+DK/4KmDXN9ZBoaGtNAE24aCVSn4fkTYnG86hiB6HNT95XKsiwcN7uJaqeJQW+IcDTOn17uYtnC8bJn2DTBcUsIN+G4tQ34WFRjT1n3dKaxel45+7uFcGsf9NEx7KdnNEAkFufUsJ8ml22Oj1AwmeP21R0H+c7fWvN+Hl8oitNioLXPw+OH+6d+IJkCeP1DYKss+uTj+NaEM0y41a6E215gsHoj1K+Bd90Pm26d66PS0NCYJppw00igho6+cFw4LMtThJshUSr1hWOEY3EqbMaE4/booT6GfWEuOme5eIDJiaxLLcNUO0zoJFEiBGjr97K4xsGZzJrGck4NBxj1R3jq6AAAcRl63EE6hv3MryoN4ab2uMlyauZc71iQnrH8p019oRiLqkWzu3eSGJSsZCqV+gbBVvw8O1W4TVYqPa1xtXBo1afgnffCom1zfTQaGhoFoAk3jQRWkx6LUUeXO4DTbEiEjkJq8/pI0gmu2mFmLBjle38/QkO5hfNWLBYPsKf3rRn0OqodZnrHggTCMbrcgTNfuM0T5eID3aM82TqAWhU93DvGoDdMS4kIN6fFQFwWolwlGInh9kfoGw3lvcfUG4pSrfRGBiIZ8thyIenSS6X+IbAXX7gNnamOm4aGxhmFJtw0UlBdt6V1jpTeK4dlvMdNzXCrtJlYM68co17CG4zymatXoLe5AAlsmQcO6sos9I2FEs3rZ7pwU2NO9naM8NzxIS5aWgPAs22iHD2/sjSiGNR8suTJ0n6lFzEciyeCmXPhCxco3DIF8PoGRam0yKgXJGdcj5uGhsYZxdyPs2mUFC6bie7RIMuSyqQgHJm2ASHcxktKRjbMr6T1K1en9qlZXZOWturKLHSO+GkbEGt4FteWhnApFpV2E/NcVr779yPIMrz7gvnsOjrAM8cGAUqmVKpGvowFojQorYm9SWHJvaPBhCCbDFmW8YWiVDpMSJJY6TVlMpVK/UOzUiod0oSbhobGaYDmuGmkoJ60lmYQbmqPm7rPVJ1CTRsuaNoITedmfP66MjP9nhBtA14kCRZUndnCDeCasxpYO6+cn92yge0r6mgot3K0XwjXUimVllnSHbdk4dadxwL6UDROJCbjMBuwGvXTE26SHuS4SPYH4b4FRmalVDrsC6FX9vNqaGholCqa46aRgtqYvTxNuI03r6vxH5M6Ezf/XnzcuTPtproyC8O+MAe6x2iusGExnma5WNPgc29YmfL/zZVWutwBKmzGkhEJaql0LKVUmuS4jWUOEU7Gp5TS7SY9NpN++qVSEIJNbwD/MCDP0nCCyCY8k6ecNTQ0Tn80x00jhUqbOIEvq0vtPSu3GonEZDyhKEf6PFQ7TLimsf5HzWz7+8E+1reUXgjxbNBcIVy2lhJyG8e3Y4wPBvSOBrEYdRh0Ej2j+Qg3IdTsZgMWY4HCTS2X+kUv4Gz0uA37QoltFxoaGhqliua4aaSwdWkNA94QNc7UfibVgTvYPcah3jFW1JdlenhOVs8rw2LUccvm+fzL5csKPt7TkeZKIdzmV5ZGmRTGS6VjgdRSaX2ZhUhMpjcP4aYOr6il0uC0pkpVxy0KmMEvegFno1Q6Gojgsmr9bRoaGqWNJtw0UrhsVR2XrUrfx3hWk+hYf7nDzZE+L+8+f3r7RVc3lnP4K1cXdIynO82VYmF5qQwmQNJwQpLj1j8Woq7MQlyW6RnN3ePmUzYl2M0GrKZp9rgll0pBTJRCQaVSWZbz2k7hC8WodmjCTUNDo7TRSqUaeVHlMDPPZeX+fV2Eo/FpO24aSaXSEnLcLEY9JoMupcetdyxIXZmF+nLrlBy3wkqlyrVkolRamOO2t2OEFV94mH5Pfj16drN2LauhoVHaaMJNI2/OairncK9YmL6yQRNu0+WcZhf/euVyrlpTP9eHkkJZ0uSwLMuiVFpuoaHcQs9oMG2rwkR8E0qlgUg86/0zkiiVKo/1K3tSrdPrcTvc4yEUjdM3Gsp5X28oikMTbhoaGiWOJtw08uasJjFMYPj/7d17kFzlmd/x79PX6Zmei0YjRlekQRJ3bDACLF+wbMoYk2S1pOQsTrKmXM7iG8kmWVcZJ6kNa8dJTCXLVnY39pKAl7U3AZbdtRUWjL0GGbxVxshegS4YLCOB7tJImku3Zvr65o9zeqan1T3T3WKm+0i/T9WUus853fPOU6e7Hj3vLWTn/fpr8ykSDvH5D66ju01mlJb0dESnxriNTuTI5otc1B1naU8HmXxxahmYWqZmlcbD3hi3prpK/a+k0u4J6WGI90KkuS7M0mLR2cLcbVHFTUSCQN9SUrd3+uPc1l2UJB45/5fxuNCU1upzznHwtDembWlvB2F/fNihkYlZ9/FM+bNKk6Uxbm9XV2mV7dPqVdoNIZufvVpYLDrS2YISNxFpe/qWkrpdtcJL3C5f2j3HlRJEPQlvrb5/8ch2fviL4wAs6+3gom5vCZcf7x3mav8eqCZdMcbtTLML8MLMyQnnMDHh1FTFbfZu2zO5UtKp/5CISHtT4iZ1601E+fe3X8ENQ/O/ppYsvO6OCG+cSPPKwVHev36AD185yLWrFhEOGe9c1ceTrxzmMx9YW/P16UyeeCRENBxqfjmQUPlyIHhj3HpXNvHXeKYrbrMnbuVJp4hIO9MYN2nIb918CdeuujAXzj3f9XREOTQyQb7o+OR71/CJjWsI+7sI/KN3LGPXoTH2D6drvr58cH8iFmIiV5hzQsNZShU3V5qccI5dpf64vLkSt/I16ERE2pkSNxEBptdyM4PrV8+sqt5+zTIAnnzlcM3Xlw/u74xFKBQduUKDiVv5Om7O+RvMn0viVt/khOntupS4iUh7U+ImIsD07gmXDXbTm5g543V5X4LrLu6bGvtWTSozPbi/tAdtwxMUyrtK8xkoZKGj+QrvKb+rNDfH5ISUukpFJCCUuIkIMF1xu2FN9TGM11+8iD2Hx8jVGOifzuSnBvcn/MSt4XFuVrZXacZbM5B4c5NhcoXi1Lp0mTkmJ6TLZsSKiLQzJW4iAnizSoGak0+uWdlLJl/kl8dSVc+ns9NdpYmY99XS8LZXpeVAigXIjHmP480t9ly+7lz9kxM0q1RE2tu8Jm5mdpuZvWZme83s3irn42b2mH/+RTNbU3H+YjNLmdkX5rOdIgLXXbyIm4b6ef+66stvlJYC2XVotOr5VNkYt8Q5d5UWGBvxdk14ZbiJ2alMj28DTU4QkfPHvCVuZhYG/hj4KHAl8HEzu7Lisk8Bp51z64AHgK9VnP994On5aqOITBsa6OKxT2+sucju0OIukvEIO2skbulMnmTsHMe4+V2lqckM/+U7PwVgx7Emts5ienwbULN7t0TLgYhIUMxnxe1GYK9z7g3nXBZ4FNhccc1m4BH/8RPALWbeMu1m9uvAPmD3PLZRROoUChlXLe+ZJXErnFVxa3jbK7/i9v1dhxg+6W0wf7rQ5HZX6forbulMHjPojKmrVETa23wmbiuAA2XPD/rHql7jnMsDo8BiM0sCXwR+bx7bJyINumZFL68eOXuCgnOOdLZscoKfADW8e4KfuI2lJxmIeYnXqXxHU209XT7GbY6KWypToCsWwf9/o4hI22rXfoH7gAecc6nZvkjN7G7gboDBwUG2bds27w1LpVIL8nvOB4pVY4IQr/BYnky+yLeffI6h3unq1GTe4RwcPfQm27Yd4VDKS5R+9vJOIsdfrfv9e0d2cx1w+OgxevEW+/3liTNV4zJXvH72Ky/xi4Xgjf1vsW3bsZrX7n0zQ4RC28f/XATh/moXilVjFK/GnGu85jNxOwSsKnu+0j9W7ZqDZhYBeoGTwE3AFjO7H+gDimY26Zz7o/IXO+ceBB4E2LBhg9u0adN8/B0zbNu2jYX4PecDxaoxQYjX1akM3/7Fc7xwuodPbr5h6vjxsUn42x/yjisuY9O7V3Pg1Bn48XMMrb+MTRtWzfKOFd5KwA7o6EzSnyvAJGQ7B6vGZa54vZDaQ9ebb5GIRViydJBNm66pee1fHP45/ZNjbR//cxGE+6tdKFaNUbwac67xms+u0peA9WY2ZGYx4E5ga8U1W4G7/MdbgGed5/3OuTXOuTXAHwD/uTJpE5GFN5CM828+fCnP/uI4f7tnuoJVOSuz1FXa8DpufldpJpuhLzJJzqKcyjTXfXk6naWvM0YsbHVNTtDEBBEJgnlL3Pwxa/cAzwCvAo8753ab2ZfN7Nf8yx7CG9O2F/i3wFlLhohIe7nrPWu4ZEkX3/jRr6aOlRawnd7yyp9V2ugYN/O+kiazOXpDk2RCnVOL6Dbq9Jks/V0xYpFQXZMTtIabiATBvP4X0zn3FPBUxbHfLXs8CXxsjve4b14aJyJNiYZD3DS0mGd2H506lqpYwLYj0uw6bt5XUiaTpadrglwkyXjZJINGnDqTY1FXjEy+UMc6bgVW9DU3CUJEZCFp5wQRadjqxZ2cSmcZn/SSqnRFV2koZMQjoaYX4M3kciSZIB/pYjJXnLOrs5pT6Qz9nVGv4qauUhE5TyhxE5GGXdzfCcBbp84A3nZXMHMB20Qs3Pg6bv4CvLlcji53hkLU26e00e5S5xwnxjMMJOPEwvV2lSpxE5H2p8RNRBpWStwO+IlbtS2jEtFw012lIYok3BmK8STAVGXvh68e49YHfkQmP/v7prMFJnNFBrrjRMNzV9xSmby2uxKRQFDiJiINu3ixl7i9edKvuFXZMioRDXMyleW3/mw7rx8br++NQ95XUoQC8UIa4jMrbk/vOsrrx1IcG83M+jbD4975Jcn4nJMT8oUimXyRrpgSNxFpf/qmEpGG9XRE6euMTnWVpvxZpZ3R6ZmZHdEwL/xymGyhyM2XLuHSwe6539jvKg1bkVghTaijB4Axv+K248AIACdScyRu/vmB7jjxSIiTsyRu0zNiNatURNqfKm4i0pTV/Z3TY9wyebpiYUKh6TXXErHwVBfl2ESdM0OnukodkVyKsJ+4jU/mGZ3Isfd4CphOzGo54VfcBpKxOScnpLJnd/OKiLQrJW4i0pRVlYlbReKTKKu+lSpmc/JnlSbIECpmiXT2AV7i9srBkanLTqayVV9eUkrsltQxOaFaN6+ISLtS4iYiTbm4v5NDpyfIF4qMVxnc31GWuNU9K9TvKi3tUxrr6vVfn+Pv35pO3OasuKWymEF/V4xoOFRzOZFfnUjxzb/bD6jiJiLBoG8qEWnK6sWd5IuOI6OTVStuyXiYzliYRZ2xBrpKvcStz7wu0XiXV3FLTebZcWCE9RclOT6e8RK33tpvM5zK0N8ZIxIO1ZycMHImy2/8yU8YOZPlxqF+rl4xyxuKiLQJVdxEpCmrytZyq7Zl1Gc3rePr//x6BpKx+ituU4mb1wUbSfTQEQ0xnvESt2tX9TGQjNU1xm0gGQeombh9+ck9jJzJ8t173svjn97Iku54fW0UEWkhJW4i0pTBHm+LqBPjGVKZwlldjZct7eYDly6hJxGtf4yb31W6OOx1lRLvprsjyuvHxjmVznLNyl4GknGGx+ce4zbQHQO8xC1T1lX6F9sPcNsfPM9f/fwQn920lquWq9ImIsGhxE1EmrK4y0uMTqazs+480N0RabirdFHIq7h5iVuE7ftPA3Dlsh4GuuNzVtyGUxmW+BW3uD/GzTkHwEM/3sfYRI4vffRy/uWH1tfXLhGRNqHETUSa0tMRJRwyTqUzsyZuPR3RBrpKvfcojXEj3kN3R3RqZ4bLl/WwJBmfdR238u2uwKu4OQf5ope4DaeyfOCyi/j0B9YSi+grUESCRd9aItKUUMhY1BnjZCo765ZRDXWVhiIcD13EqsJB73m8m54O731XL+4kGY9MjZnLFlzVtyjf7gogGva+5rL5IsWi41Q6w5JkrIG/VESkfShxE5GmDSRjHB/PzLplVHc8wmSuOOdG7wCY8ZWOL5CzqPfc7yoFr5vU+51eQjaePTtx2/ryYZ7aeWTGdaWqWjZf5PSZLEUHi5OaiCAiwaTlQESkaf1dsalFeGttGdWT8JKw8clcXQnTT3JreWLll/h4188hmqA77r3+iorEbbRK4vbVv9nDsTF/8d3uisStUGTUH2u3WBU3EQkoVdxEpGn9XTEO+Ilbra7SUsVszB/n9v9ePsy9f/kKxWL1rs6xiRz7lt0Od/45mJ1dcfMTsrHMzNc752bsqDDgJ2exsq7S0qSGxV2quIlIMKniJiJNG0jGyfhdoLNNTgCv4jY2meM/fGcXoxM5Nq5dzOZrV8y4dnwyRyZfZFHndEWs23/9FctLFTfvXGXFbWwiT77o+Gc3XUxHNDy1qX15xa2U2A2o4iYiAaXETUSa1t81nQDNNjkBvMTqf7+wj9GJHCsXJbj/e6/xkauWztgaa9+wt37b0EDn1LHN1y4nEQuxvNdbN67UVVpZcTuZ9qppG9Ys4o7rVk4dL6+4nSxV3DTGTUQCSl2lItK08sRttnXcAI6OTfLQC29w+zVLuX/LOzg0MsF9W3fP6DItJW6XLElOHVsz0MXdN6/FzABvD9TueISxiorbqXTWb9PMpKx8csJwKks4ZPT5yaSISNAocRORppV3Oc41OeGn+06Szha447qVvGftAPd8cB2PvnSAr/zNnqlr3ziRxszbwH42PYkoZypWGBn2u0EXd83sBp3RVZrO0N8VIxSy+v5AEZE2o8RNRJpWXt2aa3LC9je93Q8uX+qNg2GsjgAADNRJREFUPfudWy/lN9+9mm/+3X5eOzoOwBvDaVYuSszoPq0mGY8wWahecaucMVrqKs35FbfKxE5EJEiUuIlI0+rpKk3GIph51bRENMyKvgQAZsbv3HopiWiYB59/A4B9wymGBpJV32fGe3ZEmMhXJm6Zs9oE0xW3TMEb4zag8W0iEmBK3ESkaeVdpbUqbqGQ0e2fu3QwOaObsq8zxm/csIrv7jjE4ZEJ9p1Ic8lA15y/NxmPMFGxi9ZwKkt3PEI8MrNaV75zwsl0Vmu4iUigKXETkaaV9isNh4z4LPt+lpb0WO8v0VHuU+8bwgFfeXIP6WyBS5bMnbh1V624ZemvkpTFyycnjGe0hpuIBJoSNxFpWmm/0q5YeGrWZzWlCQqXVUncVvV38rHrV/L0rqMADNVRcevuiFC5b/2pdPasblKY7iodm8yRzhYY6FbFTUSCS4mbiJyTgWSsZjdpSWmj+PWD1cev/atb1k9NIihfCqSWZDzCmXzlOm7ZqtW0UuJ2dHTSa68qbiISYErcROSc9HfFak5MKCl1lV629OyKG8DyvgSffN8alnTHWdbTMefvTMajZAuQL0xvXH8ylak6Y7SUEB4e8RI3jXETkSDTzgkick4+sXH11ObttSzqjNLdEWHpLEnZFz9yOb99y/q61lhL+hW8dKZAb2cI5xynz1Qf41aquB087e2pqlmlIhJkStxE5JzcdvWyOa/53AfXccd1K2YdBxcKGZ2x+r6SSrNUxzM5ejujjE3myRVc1YpbaVbpniNjAAzVMflBRKRdKXETkXk3NNBV16SDepUW9U1lvBkK03uQ1u4qHZ/Ms7SnY2rTexGRINIYNxEJnFJXacqfWlprn1LwKnnRsFfpqzU5QkQkKJS4iUjglGaxjvuJ28l09X1KS0pVt/UXVZ8cISISFErcRCRwSl2l435X6fAsXaUwPUHhUlXcRCTglLiJSOAk4944tVJX6ZGRScIh46Lu6rNWSxMU1FUqIkGnxE1EAmdqjFvGW4bk8OgEg91xwjWWEilV3Napq1REAk6Jm4gETlcsjDGz4rasL1Hz+lgkxGBPnN6EZpSKSLApcRORwDEzOiLTY9yOjE6wrLf24r7d8QiXL+1ZqOaJiMwbreMmIoGUiBjjk3mccxwZneTWq5bWvPb+Le8kEQ0vYOtEROaHEjcRCaRExOsqPZXOkskXZ6241dojVUQkaOa1q9TMbjOz18xsr5ndW+V83Mwe88+/aGZr/OM3mtkO/+dlM7tjPtspIsGTiBipTJ4jo97m8ct6a49xExE5X8xb4mZmYeCPgY8CVwIfN7MrKy77FHDaObcOeAD4mn98F7DBOXctcBvwJ2am6qCITOmIGONlidvyvtoVNxGR88V8VtxuBPY6595wzmWBR4HNFddsBh7xHz8B3GJm5pw745zL+8c7ADeP7RSRAPK6SnMcGZ0AVHETkQvDfCZuK4ADZc8P+seqXuMnaqPAYgAzu8nMdgM7gc+UJXIiIlNdpYdHJomFQzW3uxIROZ+0bfejc+5F4CozuwJ4xMyeds5Nll9jZncDdwMMDg6ybdu2eW9XKpVakN9zPlCsGqN4NSZSzDGSzrPj9f30xhzPP/+jVjepren+qp9i1RjFqzHnGq/5TNwOAavKnq/0j1W75qA/hq0XOFl+gXPuVTNLAVcD2yvOPQg8CLBhwwa3adOmt7P9VW3bto2F+D3nA8WqMYpXY76z9/tkCjlysR4uWWps2rSx1U1qa7q/6qdYNUbxasy5xms+u0pfAtab2ZCZxYA7ga0V12wF7vIfbwGedc45/zURADNbDVwO7J/HtopIwCQi3vZWOw6MsHyWXRNERM4n81Zxc87lzewe4BkgDDzsnNttZl8GtjvntgIPAd8ys73AKbzkDuB9wL1mlgOKwOecc8Pz1VYRCZ5rBsKMXL2UgWScf3rTxa1ujojIgpjXMW7OuaeApyqO/W7Z40ngY1Ve9y3gW/PZNhEJtuXJEF//h9e3uhkiIgtKe5WKiIiIBIQSNxEREZGAUOImIiIiEhBK3EREREQCQombiIiISEAocRMREREJCCVuIiIiIgGhxE1EREQkIJS4iYiIiASEEjcRERGRgFDiJiIiIhIQStxEREREAkKJm4iIiEhAmHOu1W14W5jZCeDNBfhVA8DwAvye84Fi1RjFqzGKV2MUr/opVo1RvBpTitdq59ySRl983iRuC8XMtjvnNrS6HUGgWDVG8WqM4tUYxat+ilVjFK/GnGu81FUqIiIiEhBK3EREREQCQolb4x5sdQMCRLFqjOLVGMWrMYpX/RSrxihejTmneGmMm4iIiEhAqOImIiIiEhBK3OpkZreZ2WtmttfM7m11e9qRme03s51mtsPMtvvH+s3sB2b2S//fRa1uZ6uY2cNmdtzMdpUdqxof8/wP/357xcze1bqWL7wasbrPzA7599cOM7u97NyX/Fi9ZmYfaU2rW8fMVpnZc2a2x8x2m9lv+8d1f1UxS7x0j1Uwsw4z+6mZvezH6vf840Nm9qIfk8fMLOYfj/vP9/rn17Sy/Qttlnj9qZntK7u3rvWPN/5ZdM7pZ44fIAz8CrgEiAEvA1e2ul3t9gPsBwYqjt0P3Os/vhf4Wqvb2cL43Ay8C9g1V3yA24GnAQPeDbzY6va3QazuA75Q5dor/c9kHBjyP6vhVv8NCxyvZcC7/MfdwOt+XHR/NRYv3WNn/+0GJP3HUeBF/555HLjTP/4N4LP+488B3/Af3wk81uq/oU3i9afAlirXN/xZVMWtPjcCe51zbzjnssCjwOYWtykoNgOP+I8fAX69hW1pKefc88CpisO14rMZ+DPn+QnQZ2bLFqalrVcjVrVsBh51zmWcc/uAvXif2QuGc+6Ic+7n/uNx4FVgBbq/qpolXrVcsPeYf4+k/KdR/8cBHwKe8I9X3lule+4J4BYzswVqbsvNEq9aGv4sKnGrzwrgQNnzg8z+Ib9QOeD7ZvYzM7vbPzbonDviPz4KDLamaW2rVnx0z1V3j9+d8HBZt7tiVcbvmroO73/6ur/mUBEv0D12FjMLm9kO4DjwA7yK44hzLu9fUh6PqVj550eBxQvb4taqjJdzrnRvfdW/tx4ws7h/rOF7S4mbvJ3e55x7F/BR4PNmdnP5SefVhTWNuQbFZ05fB9YC1wJHgP/e2ua0HzNLAn8J/Gvn3Fj5Od1fZ6sSL91jVTjnCs65a4GVeJXGy1vcpLZWGS8zuxr4El7cbgD6gS82+/5K3OpzCFhV9nylf0zKOOcO+f8eB/4a7wN+rFT29f893roWtqVa8dE9V8E5d8z/QiwC/4vprirFCjCzKF4S8ufOub/yD+v+qqFavHSPzc45NwI8B2zE69KL+KfK4zEVK/98L3BygZvaFsridZvfPe+ccxngm5zDvaXErT4vAev9WTQxvAGXW1vcprZiZl1m1l16DNwK7MKL013+ZXcB321NC9tWrfhsBT7hzzh6NzBa1uV1QaoY93EH3v0FXqzu9GezDQHrgZ8udPtayR9D9BDwqnPu98tO6f6qola8dI+dzcyWmFmf/zgBfBhvTOBzwBb/ssp7q3TPbQGe9au9F4Qa8fpF2X+gDG88YPm91dBnMTLbSfE45/Jmdg/wDN4M04edc7tb3Kx2Mwj8tT8GNQL8H+fc98zsJeBxM/sU8CbwT1rYxpYys/8LbAIGzOwg8B+B/0r1+DyFN9toL3AG+OSCN7iFasRqkz+F3uHNYP40gHNut5k9DuwB8sDnnXOFVrS7hd4L/Caw0x9bA/Dv0P1VS614fVz32FmWAY+YWRiv2PO4c+5JM9sDPGpm/wn4e7xEGP/fb5nZXrwJRne2otEtVCtez5rZErzZozuAz/jXN/xZ1M4JIiIiIgGhrlIRERGRgFDiJiIiIhIQStxEREREAkKJm4iIiEhAKHETERERCQglbiJyQTGzxWa2w/85amaH/McpM/ufrW6fiMhstByIiFywzOw+IOWc+2+tbouISD1UcRMRAcxsk5k96T++z8weMbMXzOxNM/vHZna/me00s+/52yVhZteb2Y/M7Gdm9kzFyvsiIm87JW4iItWtBT4E/BrwbeA559w1wATwD/zk7Q+BLc6564GHga+2qrEicmHQllciItU97ZzLmdlOvK3uvucf3wmsAS4DrgZ+4G/1FgYumP0+RaQ1lLiJiFSXAXDOFc0sV7ZRdhHvu9OA3c65ja1qoIhceNRVKiLSnNeAJWa2EcDMomZ2VYvbJCLnOSVuIiJNcM5lgS3A18zsZWAH8J7WtkpEzndaDkREREQkIFRxExEREQkIJW4iIiIiAaHETURERCQglLiJiIiIBIQSNxEREZGAUOImIiIiEhBK3EREREQCQombiIiISED8f1vM8vh8x0BAAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 720x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "\u001b[92mfeature: 3:\n",
            "\n",
            "feature_forecast.shape:\n",
            "(43200,)\n",
            "feature_valid.shape:\n",
            "(43200,)\n",
            "\u001b[92mMAE: 0.05011060833930969\n",
            "\u001b[92mMSE: 0.006676157005131245\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "\u001b[92mfeature: 4:\n",
            "\n",
            "feature_forecast.shape:\n",
            "(43200,)\n",
            "feature_valid.shape:\n",
            "(43200,)\n",
            "\u001b[92mMAE: 0.029436832293868065\n",
            "\u001b[92mMSE: 0.0072934296913445\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAnYAAAFzCAYAAACzcUTXAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOydd5xcdbn/39/ZTe+9k5AGSQihhtAkgAhIFemCqCii4v3dq94rV8V2r14Liigi0hTpEERBelsIJZUQ0gjpvW/vs7vf3x/PnN3Z2ZnZKafN5nm/Xnmd3Zkz5zw5c/acz3mqsdaiKIqiKIqiFD6RoA1QFEVRFEVR3EGFnaIoiqIoShdBhZ2iKIqiKEoXQYWdoiiKoihKF0GFnaIoiqIoShdBhZ2iKIqiKEoXoThoA/xk6NChdsKECZ7uo6amhj59+ni6j66EHq/M0WOVHXq8skOPV3bo8cocPVbZEX+8li5dut9aOyybzx9Uwm7ChAksWbLE032UlJQwd+5cT/fRldDjlTl6rLJDj1d26PHKDj1emaPHKjvij5cxZku2n9dQrKIoiqIoShdBhZ2iKIqiKEoXQYWdoiiKoihKF0GFnaIoiqIoShdBhZ2iKIqiKEoXQYWdoiiKoihKF0GFnaIoiqIoShdBhZ2iKIqiKEoXQYWdoiiKoihKF0GFnaIoiqIoShdBhZ2iKIqiKEoXQYWdEiimJQr7Pg7aDEVRFEXpEqiwUwJl4sYH4K6ToaEqaFMURVEUpeBRYacER9VuRu98CZoboXRT0NYoiqIoSsETqLAzxpxjjFlrjFlvjLk5yfs9jDGPx95faIyZkPD+IcaYamPMd/yyWXGRd35PUUuj/Fy6MVhbFEVRFKULEJiwM8YUAX8EzgWmA1cZY6YnrHY9UGatnQzcBvwy4f3fAi94baviEe//jf1DjpefVdgpiqIoSt4E6bGbDay31m601jYCjwEXJaxzEfBA7Od5wJnGGANgjLkY2ASs8slexU2am6Cxiqp+U6DPcBV2iqIoiuICxQHuewywLe737cAJqdax1jYZYyqAIcaYeuC7wFlA2jCsMeYG4AaAESNGUFJS4orxqaiurvZ8H12BoqZaTgXqmqCiaAgtm5axXI9bWvTcyg49Xtmhxys79Hhljh6r7Mj3eAUp7PLhx8Bt1trqmAMvJdbau4G7AY477jg7d+5cTw0rKSnB6310Car3wttQ3Ks/A4YcBZve0uPWCXpuZYcer+zQ45UderwyR49VduR7vIIUdjuAcXG/j429lmyd7caYYmAAcADx7F1qjPkVMBBoMcbUW2vv8N5sxRWitQA0F/WAwRNh+aMQrYNuvQI2TFEURVEKlyCF3WJgijHmUETAXQlcnbDOM8B1wHvApcDr1loLnOqsYIz5MVCtoq7AiNYB0BKJCTuAss0wfFpwNimKoihKgRNY8YS1tgm4CXgJWAM8Ya1dZYz5qTHmwthq9yE5deuBbwEdWqIoBUo7j92h8pr2slMURVGUvAg0x85a+zzwfMJrP4z7uR64rJNt/NgT4xRvafXYdYdBjrDTylhFURRFyQedPKEEQ0zYNRf1gN6DoecACcUqiqIoipIzKuyUYIiFYlsiPeT3/mOhcmeABimKoihK4aPCTgmG+OIJgP6joTKxKFpRFEVRlGxQYacEQ2vxRHf5XYWdoiiKouSNCjslGDp47MZAzT5oagjQKEVRFEUpbFTYKcEQ3+4EYMAYWVbtCsggRVEURSl8VNgpwRCtA1OENbGOO/1Hy1ILKBRFURQlZ1TYKcEQrYNuvcGZ9ds/5rGr0Dw7RVEURckVFXZKMERr28+FbfXYqbBTFEVRlFxRYacEQ7SuvbDr0Q96DNBQrKIoiqLkgQo7JRiitRKKjUdbniiKoihKXqiwU4IhWt/eYwdSGavCTlEURVFyRoWdEgxO8UQ8/UdrKFZRFEVR8kCFnRIMicUTIJWx1XuhqTEYmxRFURSlwFFhpwRDtA669Wz/2oCxgIXK7YGYpBQo1sKBDdDSHLQliqIogaPCTgmGZMUTA8fLsmyL//YohUnpRvjr+fCHY2D5o0FboyiKEjgq7JRgSGx3AjAoJuzKVdgpGfL2bbBjiTwkbF/i335bWmDdq/CPr8OKef7tV1EUpROKgzZAOUhJVjzRbzREiqF8azA2KYVHbSkMngi9h8CeVd7vz1pY+RS8+UvY/3GbDTMv9X7fiqIoGaAeO8V/rE1ePFFULAUUGopVMqWhSppbj5gBe1eLJ80ryrbA3y6Ep66HSDf47H0wfAZgvdunoihKlqiwU/ynOQq2uaOwAwnHaihWyZSGSujRX4RdY7V35876V+HPn4CdH8D5t8GNb4uXrqibPKgoiqKEBBV2iv9Ea2WZGIoFKaBQj52SKfEeO/AmHLtiHjx8uVRt31ACx30JIrFLp4mA9dBLqCiKkiUq7BT/idbJMpXHrmYvNNZ6s+/aUvjgEXj6RrjvbGmToRQu9ZXQsz8MmwYY94XdyqfgqS/DISfCl16EIZPav28MGopVFCVMaPGE4j9pPXYTZFm+FYYf7t4+d6+Ad26HVU9DS5OE7xoqYfeHHW/WSuHgeOy695bvcc9K97a9dSE8/TU4ZA587knZRyLqsVMUJWSosFP8pzOPHUiulBvCrr4SXvkhLP0rdO8Ls2+AIy+H4l5w5wna1LaQaY5CUx30GCC/j5ghAt4NqvfC45+T+cVXPpJc1AFgNMdOUZRQocJO8Z9WYZcixw7cybPbtRyeuE68f3O+Bqd9F3oNlPf2r5elelsKl4YqWfboJ8th02DNsxCt7zjVJBtaWuAfX5Ptf/4Z6D049brG6DmkKEqoUGGn+E9rKLYX0NT+vb7DxZtWtjm/fWwsgUevFiH3xeclnBaPk/yuHrvCpb5Clj37y3LYVBFZB9bDyCNy3+7S+6UK9tO3wojp6dc1ETTHTlGUMKHFE4r/OB674iShWGNg6BTYtyb37W94Ax6+TMK6X3m9o6gDaYQM0nZFKUwSPXZDp8py/9rct1m1B179KRx6Ghz/5Qw+oKFYRVHChQo7xX/aeeySMHIm7M4xCX7Xcnj8GhgyBb7wHPQbmXw9UyTLlqbk7yvhp6FSlj1iHrshkwED+9flvs2XfyB5e+f9Nlbx2gkmosJOUZRQocJO8Z90xRMgSfA1eyWBPRtqDsCjV0HPgXDNvPS5URFH2KnHrmBJ9Nh16yVe2n05euy2L4UVT8BJ/wZDJ2f2GW13oihKyFBhp/hPU5riCYARsfyobCocW1rg71+Bmv1w5cPQf3T69R2PnSa+Fy71MY9dzwFtrw2d2jbDNRushVd/BL2Hwin/nvnntHhCUZSQocJO8Z/OPHYjZ8oym55kC++CDa/Bub+A0Ud1vr567Aqf1lBsv7bXhk6VUGy23+uG12DzfKmcjt9ep4Qox66pEV7/mRQNeTkzV1GUUKNVsYr/NFTLMpXHrvdg6Dc68ykCpZvg9f+BKWfDsV/M7DMm9kyjxROFS2KOHcCww6C5QVrcZMNbt8KAcXDsF7L7XFgaFFfugseuhp3vy+/RWujRN1ibFEUJBPXYKf5TuR36joCiNM8VI4/IrIDCWnj2/0lo9fwME96hrSpWPXaFS0MVRLpBcY+214YeJstswrFb3oOt78FJ34Ti7tnZEIYcu70fwb2flP/zlE/Ja1oUpCgHLSrsFP8p2wIDD0m/zogZ0raiqSH9essegk1vwlk/kSHtmRLRqtiCx5kTGy/mh8Vanuxdnfl23vkd9B4CR1+bgxEB59iVboK/XQgtUfjiCzD5LHldz2tFOWhRYaf4T/nWtgkTqRg7W25OW95NvU7lLnjp+zD+5MxDsA6txRPqsStYnDmx8fQaJOfWzmWZbWPPKvj4RTjhxjRjw9IQZLuT2lJ48GJoboTP/xNGHakPLIqiqLBTfKa5CSp3dO6xmzhXGhh/9Fzy962F578j+VQX/qFtkkSmtN4AQ5AfpeRGQ2X7/DqHMcfAjgyF3Tu3Q7c+GTYjTkJQoVhn7FnFDrj6CRg+TV4v6hZ7X4WdohysqLBT/KVqp9x0OhN23XvD5DNF2CXziKz+B3z0Lzj9ezBkUvZ2aPFE4dNQlVzYjT4GKrbSrbEi/efLNsOKeXDcF9P3PExHUMUTC+8ST+PZP4Nxs9ted3JHm6P+26QoSigIVNgZY84xxqw1xqw3xtyc5P0expjHY+8vNMZMiL0+2xjzQezfcmPMZ/y2XckRp1pxUCehWIDDzxMhmBhWq9wFz30bRh0Fc76Rmx3GyE1ZiycKl4bKtjmx8Yw5BoB+VZ1MoHj3DjkH5nw9DyMCaHdSvlWqwKeeA7NvaP9exPHY6XmtKAcrgQk7Y0wR8EfgXGA6cJUxJnHi9vVAmbV2MnAb8MvY6yuB46y1RwHnAH82xmjrlkLAEXad5diB3LhMEXz4RNtrLc3SiDhaB5fck76ytjMixeqxK2TqK5P3nBs1CzD0q1qf+rPV+2DZgzDrChgwJncbgsixe/4/AQOfvrVjFXhrioF67BTlYCVIj91sYL21dqO1thF4DLgoYZ2LgAdiP88DzjTGGGttrbXWSSLpSeD9BpSMKd8KmMwqWHsPhllXSthpY4k0YH36Rmkk++lb2yogc8UUaS5SIdLcBI01qUOxPfrBsMPoX5nGY7fwLqm4PjmLKRPJ8DvHbmOJhGDnfhcGjuv4fmsbHz2vFeVgJUgv1xhgW9zv24ETUq1jrW0yxlQAQ4D9xpgTgPuB8cC1cUKvHcaYG4AbAEaMGEFJSYmb/4cOVFdXe76PQubwNQsZ1H0w7739HtD58SrqewHH9H6Tng9dQVNxL3o0lrHx0GvZWj4a8jzOp1jLrq1b2FAg35eeW8LkdfcwdP8iujdWsG1PGZuSHJPDI6MZVLaYkjdea6uAjlEcrWbOgj9RNnQOq1buAHbkbMuM/QfoVVfFEj++F2s55v3/onuPISyqn0ZLkn0O2f8RM4GlixZQ1f9AVpvX8ys79Hhljh6r7Mj3eBVs+NJauxCYYYyZBjxgjHnBWlufZL27gbsBjjvuODt37lxP7SopKcHrfRQ0m34NI6a0HqOMjtdRU2D+rRQ1R2H6hUycfhET3bDlvR6MGzOacQXyfem5FWPrbdCwF4DxU2Yw/tS5HdcZWgrzXmfu+GKYeFr7917/GTTXMOyzv2LuqCPzs2XPfbC/3J/vZe2L8ObHcMHv+cSxn0q+zroorIRjjz4Kxh2f1eb1/MoOPV6Zo8cqO/I9XkEKux1AfCxhLB0fnZ11tsdy6AYA7R5DrbVrjDHVwBHAEu/MVVyhfCscMie7zwydDJ+5y31bIlo8UZBU7237OVnxBMBh59JU1IviD59oL+xqDsCCO2H6xdL3LV+Mjw2K375NclOPujr1OhqKVZSDniBz7BYDU4wxhxpjugNXAs8krPMMcF3s50uB1621NvaZYgBjzHjgcGCzP2YrOdPUKOPEMqmI9QNTpMUT2VK1W3LbgqR6DxzxWfk34dTk63Trxb5hJ8KaZyAa58h/7SeSnzf3v92xxUTwJcdux/uwbYE0UnZ61SWjVdhp8YSiHKwEJuxiOXE3AS8Ba4AnrLWrjDE/NcZcGFvtPmCIMWY98C3AaYlyCrDcGPMB8DTwdWvtfn//B0rWlG8R78bgHPrOeUGkSD122XL36XD7LHj/b8Hsv6kRag/A0Klw6f0w7LCUq+4dfpq0RPnwcXlhzbPw/gNw8v+D4Ye7ZJBP7U4W3gXd+8HR16RfTxsUK8pBT6A5dtba54HnE177YdzP9cBlST73IPCg5wYq7nIg1n5iyORg7XDQdifZ0dIifQWLesAz/wZTz4W+w/y1oWafLPsO73TVskEzZTTdv/4Dti6A1f+UViinf989e/xoUFy9D1b+HY6/PnXo2aG1QbEKO0U5WNHJE4p/HNggy1wmRXiBCZHHzlpYdA9sfDNoS1ITrZXlmGMACxVb/behercs+47sfF1TBNf+HcafBMsfhSlnwRUPQ3F39+zxo93JynkSWj3mus7X1Rw7RTnoUWGn+MeB9TKkPdfxTW4TluKJpgZ4+qsy+3bxvUFbk5rGGlkOi4UxK7b7b4NTONF3RGbr9+gH1/4Dvr0WLn8gee+3fPDDY/fBI+JpHJHYvz0JYcux27kM7jlTcgQVRfEFFXaKfxxYH54wLISjeKKpAR6/VvLAinoEM3c0UxqrZRmksKuKeez6ZSjsQKaTZLN+VnicY7dnFez+EGalqYSNJ0weu49fhr98GnYsgb2rg7ZGUQ4aVNgp/nFgQ7iEXdDFE81N8MTnYd1LcP5tUggQhhtyKhyPXf/R0L1vsB67Pp3n2PmC1yPFlj8mYm3mpZmt7xRPBJ1jt+ZZeOxq6BcLmYfBM64oBwkq7BR/aKiWxPuw5NdBsB47a+G5/5DxUOf9Bo77UvBCszMcYdejr4yEq9iWfn0vqN4NvQa7myeXD17m2Fkr7VomzoU+QzP7TOus2ACF3bpX4MkvwOij4KpYRXLQnnFFOYhQYaf4Q+lGWYbKY1cslZ5BsOBP0jLk1G/D8V+W18IQGk6HE4rt7gi7gDx2mebX+YKHDYr3rIKyzTDtgsw/Ewm43cmOpeKFHj4Nrvk79Boor4c5xUBRuhgq7BR/CFurE4gVTwRwA9y6EF65BQ4/H864Jc6e4pCHYkMg7Kp2e5gvlwPGwxy7Nc8CBg77dOafCbJ4omoPPHq1eBc/95S0ZnHm9Ab1AKUoByEq7BTvidbDqr/Lz4NdmfLqDkF4yOorYN6XRBhd9MdYKC9GpCjcN0AnFNu9j9hfsw+idf7aEDaPnZeh2I/+BYecmFHPvlZaGxT7fF43RyX8Wl8BVz7aJr4jsVuMeuwUxTdU2CneUrkL7vukeB9Ou1lEQVgIIqft5Vsk1/CSe9vCVA4mEvJQrCPs+sKAWNuQisTxzh5ireTYhUrYedTupGwL7FkJh5+X3eecHLtmnz12b90KW9+FC38PI49oe904wi7E57WidDFU2HVFGqrhnd/DtkXB2lG6Ee7/FJRugqufgNNdms/pFn577DaWyEirE2+Cccd3fD9SVCCh2JjHDvwtoKgvh+bGcAk7r9qdrH9VllM+ld3ngsix274U3vo1HHkFHHl5+/daQ7Eq7BTFLwIdKaZ4wLpX4JlvQtUumHk5jJsdjB3V++DBz4jIvO4ZGHNsMHakw8/QZ3MUXvguDJoAp38vhT3F4b4BNtbIjbq4R5yw8ynPrnyrnNcgc2LDglceuw2vi1d06JTsPud3jl1zFP75Deg3Cs79Vcf3TQhDsdF6sSssldWK4jLqsesqWAuv/y88fKm0g+g9JDjvT1MDPHqFJFN/7slwijrwN/S55H7Y9xGc/XPo1iuFPWGviq2RMKwx0G80YPwRdutfg7tOhe1L4LzfymiwsOBFjl1zVEbLTT6zfQ5mJrQKO5/OowV3wr41cN6tHVMLoC00HJbzeu8a+MOxIkYVpYuiwq4r0NIig87f+jUcfQ185TXoPTS4i+mrP5G2B5fcDWOPC8aGTPCrCrW+At74ufQjS1fhGPriiWrpYQfi7eg30nth9/6D8rDSfwzcOB+Ovz57seMpHrQ72b4YGqtg0pnZfzYSkQcWP87rih1Q8gs5pw87N/k6Tig2DB677UvgvrOhcjvU7A3aGkXxDBV2XYGXvw9L/wKn/AdceId4hCJFwVxM178KC/4Is2+A6Rf6v/9s8Kt4YsFdkh/2yZ+kFyWhz7GraV/80m+UFIJ4xaJ74JmbYNIZ8OVXwlVR7eDF5IkNr4sgmnhabp+PFHcsnthYAn+cI6kRblHyczlfz/m/1Os4odigH1h2r4CHLoHeg6TlUphTHhQlT1TYFTpL/yrhkNlfhTN/1CYcTADen2gd/OtbkgN11k/93Xcu+BH6rCuD9/4Ih50nnfiDticfEoVd/9FS9ewFK5+C578jx+3KR8JVTR2PF6HYTW/BmGOg54DcPh/p1vEBYe2LEjLdsSR/+wD2fgQfPALHf0XyRlPaEoKq2Mpd8NCl0L0ffP4Z6DsyHB5ERfEIFXaFzJ5V8Px/Ssjm7J+39wYZ4//F9J3fQ/kWGZGVKo8sTPgR+lx0DzRUwNybM7Qn7MKub9vv/cdApQceu22L4ekb4ZCT4NL7pVgjrLjtsWusgR3vw4RTct9GshSDXctlud0lYff6/8i5cOq3O1/XqwKTTGhqgCeuhYYqyfcdND7WmDzEf2eKkicq7AqVpka5+fXoL7lsRQkFzn6LhMqd8PZvYcYlcOgn/NtvPnhdPBGth0V3w+SzYNSRna8f9skTDVUJHrtRIlrdDO/Vlkqj234j4cqHoVtP97btCS63O9m2SCpax+ch7IoSzqOWFglFguS+5sueVdI8ec7Xoc+Qztc3AT6wvP6/krP4mT/BiOlt9oTZM64oeaLCrlB57w7Y/SFccHvyAeHG5xy7t38nN5NP/si/feaL1+J3xRMyneGkmzJb3+/vLFNe+j5seTdJKHaMLKtcCsdaC/+8SRLbL3sAeg92Z7teYlwuntjyjpwHh5yQ+zYSc+zKNkkxRnEv8djlK0Tn/xa69YETvprZ+kF57LYugHf/AMd+EaZf1PZ62D3jipInKuwKkao9MP83kn807fzk60R8fCqt3CW5frOuSp9vEzYixd4dI2vhvTthxEw4NMMk+DCGiCq2y0PEh08kL54AqHRp+sSqv8Pa52R+7phj3Nmm17idY7f5HRg1C3r0y30bkW7tz6PdH8py5mdFNOfTVLp0o3xPx38pc+Ht57XIoalRWpoMPAQ+9b/t31OPndLFUWFXiLzxM2iqh0/9T+p1/Ax/vPsHuVBmkm8TJoyHVahb3pVk9Tk3Zt6eI4yhWGd6ScW2mLCLExz9R8vSjQKK2lJ4/r9g9DES4isU3PRGReukuGHCyfltJ1LUvkHxrg/l3Dr6Wvk9nzy7RffI/3lOFn3ggijkWngXHFgvfQ979G3/XtjbCilKnqiwKzRKN8GyB+H4L8OQSanX8yv80VANyx6C6RfD4EO935+beHmBX/pX6DFAcg4zJYyehO2LZVm+VfrYJVbFgjseuzd/CXWlcOEfOuaLhhoXc+x2fiAj0w45Kb/tFCVUxe5aDsOmiWgu6pF7nl1Dlfytz/iM5Fdmit+h2Oq98OavYMrZMOWTKewJ2d+ZoriICrtC453b5en75H9Pv55fYb0VT0gC/ewbvN+X23glpGpLYfU/YdYV0L135p8LY+7PtoWyLN0E2PbCrlsv6DUo/8rY/etg8b1wzHXtB8gXAm7m2DkiemySOcLZkOj53btajmtxdxh2mExAyYXlj0FDJZxwY5b2+Cyk3r4NorVw9s9S2BPCvzNFcREVdoVE1W744GE46nOdPzH7kYhvLSy6F0bODG4mbT54JX4/fByaG+DYL2T3ubAVT0TrJIzXo39baC+xn1z/MfkXT7z6Y0nsTzVDN8yYCK7l2O1YIjlhfYflt51IMTTHCbv6ShkxCNJjcv/H2W/TWhHfo4/JfpqMnx67qj0yvm/Wlann7IbRM64oLqLCrpBYdLc8iZ/8b52v60fC8o6lsHeVhIVDNeYpQ7y6wK94EkYeCSNmZPe5sE2e2PmBCLppF7S91j0hX6nfqPxCsbs+lNYZJ90EfYfnvp3AcDEUu31p/t466Oixa6qH4ljbmKFToXwbNNZmt80dS8XTl+3DCvic7/t7CWeny/dVj53SxVFhVyg0NcrczKnnZDZayY+L6YePyw1jxme83Y9XRIrdP0alG+UmOPPSHOwJ0Q0nWg8L/yQ/x+cJdvDY5Tl94q1fi0cw2/BeWHDLG1W5S2aYjnFhtnKkuM3D2hyVh5dWYTcFsFC6IbttLnsQuvXO7W/dr6rY+krJbT3is53kH6vHTunaqLArFNY+J60KjvtSZut7nSDcHJWxT4edm/voo6DxQkitfEqW2RRNOITlhlNbCn89T/IET/8BjD667b3ECsP+o+W8bGrMfj97P4I1z0g/tF4D87M5KNxqd+KM+so2zJmM+OKJpnpZdosXdkheY6Y01sKKp6QXXM/+2dvjVyh2+aNS4NNZVbVWxSpdnEIqPzu4WXK/5N9MOiOz9b2+eG14HWoPwJFXeLcPr/FCSK14CsbNgYHjsv9sGNqd1OyHv10seViXPwjTL5RQY7c+EK3pGIodMFaWldsz8yTHs/heqdI84Wvu2B4EzpB7a/NLR9i+WPrPjcxgQklnxOfYNTXI0vHYDZ4EmOyE3ZpnpMHx0dfkZo8f7U5aWiRVZcxxnfdA1KpYpYujHrtCoHyrDAY/5vMi2DLB66fkFU9KReSkM73bh9ekKp7INWdqzyrpXZdLGBbavtugvAnROnj0SjiwDq56VEQdiGBxhGpiKHZQrMVN6abs9tVQLVWWMy7ObCxVaImJuXzz7LYvlSIkN0aoxT8gROtk6czb7d5bvstsCiiWPSTf8/gc++u5PZ0jGZtKpG9dJtX5YUp5UBQPUGFXCKx4UpYzL8/8M17mtTQ1wscvw+HnSQuFQiWZx66uDH49Cdb8K/vtrXxKtjn94tztgWC8Cc44r+2L4ZJ7YHKCYB+QQtg5XrrSjdntb+U88QIdd31u9oaFVo9dHsKluQl2LnMnDAvthV2rx65X2/vZVMaWboTN8+Hoz+XukfQjx+6DR6DnQHlQ6IywpDwoikeosAs71so4p3FzYND4zD9nPOxjt3m+9K47PMU4s0Ih2ZP73jUSYn77t9lty1oRdhNPy71dRavHLoBw7NK/itg645Y2T108rR67xKrYkSIasvHYWQuL74PhMwqzTU48rVonD4/dvjUS5najIhZiOXax4gknx87x2AEMmSLerUw8wx88AhiYdXXu9njdxqe+Uh7Ejvhs+/9nKtRjp3RxVNiFnd0rpM3AkZdl9zkvn0o/ek5yribO9Wb7fuEco/gwmiNQdizNbvTSjqVQthmOyDEMC3HCzuebzv718NL35Ps85VvJ1xk+XcaJJc4wNUa8dtl47Ha8L/NLj/9SYbbJaYcTis1DuDjn2Zhj8zcH2guXVmEXF+IdOkUa+FZuT7+dlmb44FHJ6x0wJnd7vHzIBMkBbKqTWdUZ2ROyfpGK4jIq7MLOqpNeLVEAACAASURBVKdj4b0s2wx4VTzR0gJrn5dQXbdena8fZiKx2qH4i3zpRjnePfrDgj9lvq0V86CoO0zLw4sZRCjWWvjXv4uX5+I/Sd5hMo79InxzaXKPyOBDoSwLj92S++TBIJvUgrASXzyRK9uXQK/B2RefpCJSLFXr0LEqFmDULFnuWp5+OxtLRPzlWjTRao/Hoc/lj0lRSKahbPXYKV0cFXZhZ+0LMhQ82wRzr55Kd30gkwYOP8/9bfuNI2LiL/Jlm6T6eOal4plsjib/bDwtzbDq7zDlU/m1fgnCY7f6HxJaP/OHbbNfk1FUDP1GJH9v8KHi6czkQaK2VELWR16eW+uMsNHqccxD2O1YIqLELe9lJEm7k3iP3YgZIv52Lku/nQ8elgKpfP/WTcS9Js6JVO+FzW/DzMsyP365VMU21sJrP4Xnvi2pMUr4sVau4XvXeHf+hRQVdmGmdJPk30w9N/vPejWfccNrsizkaliHZB6y0o0iVCacIuGdPSs7386Wd6B6j+T45IPjQfRL2EXr4aUfwIiZ4pHLlcETZYRaVQYzY5c/JmLj+AIvmnDIt3iivhL2rXWnMbFDu6rYJDl23XrBsGnphV1dmeStzbwss7y1dHgZiv3oX4BNnheailw8dq/+GOb/RgrZ/v4VKPnlQScWCo5Fd8NjV8Odc+Avn26rED8IUGEXZj5+UZaHnZP9Z726mK5/DUYdlf88yzCQzENWulGEythYUv+2RZ1vZ8U8KSqYmsP3FE+rSPBJ2L3/gITazv5Z5m10kpFpZay10o9x7PHS2qNLkGe7k13LAdt577VsKCpO4rFLSJsYfZSMjEtl94p5ItbzDcNCLBTrUU7bmmfl/Bs+PfPPZJt/vLEEFv1Z+i3+1yaZ1V3yc0bvfDFrcxWf2PkBvPwDiaKc9T+w9V144b+Ctso3VNiFmbUvwNDDcsu98aJ4or5ChE5iK4xCJdFjV1sq/8dBh0rj3X6jOxd2TY0yoeGwT0uPsHzwMxQbrYe3b5PeZId+Ir9tZSrsNs+XHnmF3uIknnw9dk6em5P35gbtcuycdicJXrfRR0FdqfTITMayh8ST64ZdXjUEriuT/p7TLsgujO0IzUzF+Fu3wsDx8MkfyWcvvAMmncmkDX+BA1mOZlP8oeQX0v7mM3+W2eqnfAve/xus+kfQlvmCCruwUl8hIb7DcgjDgjdPyRvflAt0VwjDQkch5VTEDp4oN4pxszsXdhteh/ry3JsSt7PHCcX60O5k2YOSK3nad/PP7eo/RgpHOrvJLb4v815jhUK+OXa7lssDRN/hrpnUPsfOaVCc0PjYGROXLBy7Z5Xk0rrhrQPv5lZ//LL8Pw+/IHt7ILPrY2MNbF0goV6nWCwSgYvuoCVSDP/4mo4nCxvNUcm7PPw86D1YXjv9+/Kg8soP2x52ujCBCjtjzDnGmLXGmPXGmJuTvN/DGPN47P2FxpgJsdfPMsYsNcasiC0znLNVQKx/VS5auQo7L8b4bHhdWl4Ueu8xh8ScNqeyc3BsmsK42VCxNf2Q+5XzJMF84un52+NXVWxLi1T8jjkuf28diEAeepi05klF1R7Jhzr6msKvpm5Hnu1Odi1311sHsRw7p91J7CaWONFixBEiAJMJu2UPy3szs2yxlAqvpuCsewn6DM++TUyyoqlUbHlXegImjnLsP5r1k78M2xZKSoMSHna8L83PJ57W9lpRMZz1EyjfIg+YXZzAhJ0xpgj4I3AuMB24yhiTmChxPVBmrZ0M3Ab8Mvb6fuACa+1M4DrgQX+s9pG1L0LvIbk3LfWixcDmt2H8SdIaoyuQmNPmhBIHTZDluBNkuW1h8s831sJHz8O0C92ZwOHXSLENr0PpBjjhRvcqMcfNlrYdqW6Wy/4mDyr5FGmEkXzanTTWyAQIt4VdUXGSBsUJwq64h1Tirn2+ve115fDBQ+LtcGvUmxfRg5ZmOY8nfzJ1i55UZPMAteF1OXaHnNjhrT0jTofxp0hhRc3+7GxQvGPTm4CBCQkPrZPPlAfwt34lRUtdmCA9drOB9dbajdbaRuAx4KKEdS4CnMehecCZxhhjrV1mrXVK8FYBvYwxeZZuhYjmJlj3Mkw5O/ekdmPcDX9U75X8qPEnubfNoEkMxZZtgb4j2zxKI48UD+XGN5J//uMXZWKAG2HYdvZ4HIpddLd4OqYn/rnlwSFz5Cl5z6qO77U0w9IH4NDTYOhk9/YZBkwexRO7VwLWI49dfFWskVB5IrOuEmEZ34h74Z8lDeTUFI2qc8ELj92OpZJjN+WT2X82m1zWDa+LqEvmZTYGzvsNNFZLiE8JBxvflOKsZA8mZ94i582iP/tvl48UB7jvMcC2uN+3AyekWsda22SMqQCGIB47h88C71trkwbOjTE3ADcAjBgxgpKSEleMT0V1dXXe+xhQvpKj68tZ2TSO/Tlua+K2HYxtjvKWS//fYXvfYQbw/oFeVLp4DN04Xrkyctd6Dgfee+8dGnoOZ/qOTfRpKWZxnD0z+h9Bv5XPsqDvxR28W0cuv53ePYawYHMUtpSQL8P2fsQMYPGiBdT03d3hfTeOVY/6vcxZ9zJbxl/O5rffzWtb7bcLJwLrXnuQHWPb9z0bsn8hMyu2sXLs1Tmfz7ngx7k1esd6pgLvvPM20e7Z9TAcs/05pgDvbqmjcXeJazZN2LqdCS1NlLzxBhM3rWNMpBvz33yzw3pFTcM4KdKdPc//mo8P+zr15XuJrrydiiEnsHJtGax1x6ZZ5RVEWqIsc/G7mLDpYcYT4Z1d3Wnan912x27bzGRg/vw3aS7uk3K97g0HOGnfR2zoN4dtSWyvrq6mZPVuDh17EeM/eJhlzKBi4Izs/iMHCX5d5yPNDZyydQHbx57PxhT7O2LI8Qx46zYWNE5P+/0HSd7Hy1obyD/gUuDeuN+vBe5IWGclMDbu9w3A0LjfZ8Rem5TJPo899ljrNW+88Ub+G3nxe9b+dKi19ZW5b+PVn1j7k8H52+Lw3H9a+78jrY02uLdN69LxypVlD1v7o/7WHtgovz98hbV/Orn9OksfkHV2r2z/+v718nrJL92zZ/Uzss2dy5O+7cqxevNXso/SzflvK56WFmtvPdzaJ7/Y8b17P2Xtb4+wtqnR3X12gi/n1sK75XhW7c3+s09/3dpfTZJj5yYlse+4KSp/t/93SOp1//5Va38+1tpti23pbadY++OB1u78wF17/naxtfec6e427/qEtfeeldtn37tTjk/NgfTrffyKrLf5naRvt55fDTVyft8x29pofW42dXF8u85vWyzf2epnU6+zY5ms88Yv/LEpB+KPF7DEZqmvggzF7gDGxf0+NvZa0nWMMcXAAOBA7PexwNPA5621XavmfO0LMOHUjnM5s8HtSrQt70q+nxu5ZGEhsTquqa5jv6/JZ8ny45fav77kfgl5HfN59+zxuirWWmkQPP5kGDTe3W0bA4ecAFsT8hG3LoBtC+Ckm7pObmY8+bQ7cQon3J6X2xpqjMbO6Z6p153zddn/vWcysHyFjJVzOzTsdii2tlSOXa7V+ZlWxR5YJ8shU9Kv1703nP9bmeld8ovcbFLcwelsMCRNysfoo+Dw8+G9P0pOaRckSGG3GJhijDnUGNMduBJ4JmGdZ5DiCBAP3+vWWmuMGQg8B9xsrX3HN4v9YP86SWzPtRrWwUQA60539LpymcAw/uT8txUmEnNtmho69vvqP0py7eKFXV2Z9PmadgH0G+mePdm0YciFHUvhwHqYdaU32z/kJGl4vOtD+d1aeOvXMgfVrdYZYSPXdifRepkq47aIgjYB3dIk53RiRWw8o46E/7cc5v43q6f/pzfnhusPme8AFg49NbfPZ1oVe2C9jAjsM7TzbU45S87xd34H2xbnZpeSP6UbAdNWAJeKuTdDQwUsuNMPq3wnMGFnrW0CbgJeAtYAT1hrVxljfmqMcebD3AcMMcasB74FOC1RbgImAz80xnwQ++diI6gAWfuCLKeend923Gx2u20hYLtW4QR0LFZoqk/u3TjiEvE6rX9Vfn/5FmioglP+w2V7smjDkAsfPg5FPdwtmojnyMuk9cvLPxBRt+whOWanfhu6hzOXJW9y9djtXS3nnRfCLt7zG+3EYwfync29mX3DPXpwc9tjt/lt8axn2+ak1Z4Mq2L3rxPPT6Ye1bN/Lo3Nn/yCVskGRelG6auZ7mEGpLhi2oXw3p1Qvc8f23wk0D521trnrbVTrbWTrLU/i732Q2vtM7Gf6621l1lrJ1trZ1trN8Ze/19rbR9r7VFx//YG+X9JSksLvPN7qOqYCJ+Sj1+URooDD8lv326Op9ryjvS1GuviPMswkHiBT+axAwlXDZ0Kz/6HTGtY9qCEFl0PWXnYx66lRcYvTTlLvBBe0GsQnHaztBv4503wwnelT96cr3uzv1CQY1WsFxMnHBxh19yU+pz2E7fbnWyaL2H/XP9fmT70HljfeRg2np4D4PIHoWYfzPuiTKVR/MWZ9Z0JZ9wC0Voo+T9vbQoAnTzhJXtWwCu3wCNXSM+zzqgtha3v5TYbNpGIi2G9Le/K03GXaixLklBsffL/Y3EPuPAPULFNelaNPkYEjOv2eJhjt/N9mTQxLYth6blw/PUwfAaseAKGHQYX35V9n7FCwuTYoHjXchECA13OdYT251FTfce8Ub9xc251zX7Yu0pykHO2J4MHqMYaqNyRPlcrGaOPggtul1Fnf/+yiGvFP5xZ35kwbKpcr5b+Bfau8dYunwmy3UnXp2yzLHd9AM9/By7uJJ6/7hW5QUzNM78O2i5e+V5QG2ukO/1J38zfprCReIGP1qf2AhwyB258G3r2hwHj3E94B29nxa55Rm74+Yb4O6OoG9w4X37OtQdjIeF4xrPNsfOqcALicuyiqdML/MTNUOzmt2WZz8SUTP7OnPF4ufRdPOoqmcP70veg+fNwyd3Qo2/221Gyo74SavdnN1v9tJslReVf34IvPNdlHkK7xv8irDjCbtZVcvJ01u364xeg74i2OY754FYodvtiefLvaoUT0HHSQ2c3wZFHSIjci5sxeBeKtVbCsIeeBr0GurvtZESKDg5RB+QUim2OSiNnL8KwkMRjF7Cwc3MKzub50K1PftfITIqUDqyXZTah2HhO/Aac+yu5pt93lhQuKd7SOhIyC2HXZwh86mew9V3x3HURVNh5SdkWyTs6+hq5yG56K/W6TY2w7lXxqLjx1OCW92fLuyISu8p82HgSxW9TQ7A3wcTZtW6xb62EKKad7+52ldyKJ/atheYGGOmxsGvOoCrWD9z02G2aD+NPzK91TiZFSo6wy0YkJHLCV+Fz86SK/p4zYd6XYMt73hVHHew4IyEzzbFzOPoaeeh95YdSMNMF0FCsl5RtlhyasbNlNNX6V1PfXLe8IyOZ3AjDgnutM7a8KxVEXiXcB0liTltTmlCsL/Z4VBW77mVZTvE4DHswkku7kz0rZTlypuvmANlXxXqNW+1OqvbA/rVw1NX52wPpvYgH1kvKRffe+e1r8pnwjUXS9mfpX2HlU3ItHT5DWilFiqXCvucAuOjOLhMKDITWWd9ZCjtjJE3qrlPh8WvhK68VfBW/nkVeUrZZ+ukUd4eJp8H611KHbD56Ti7AE+e6s+9Ijm0Y4mlqlFBsVwzDQnuvZnOTXOiDTDT3KhS7/hUYPh0GjHF3u0qcxy4LYbd7hbSdyTYxP1NahV20a1XFbnHy6/IonHDsgfRis3xb/p0JHHr2h0/9D3xrNVx6P0y/WF7f+YE08N7xPix/VKppldwp3SipTLnkMw4YC5feJ02mn7hO/m4KGBV2XtHSLFWUTof/yWdCxVYZup1IcxRWPQ1Tz8n/CdHBjeKJncvEi9XV+tc5xAuppjr5OVCPnQdVsQ1VEv6ZnMOwdCUDcsix27MShk+DIo8CJu0aFIehKta4I+w2zYce/fMPYWfyAFW9W0SCm/ToB0d8Fi78PXzpBfi39+HfP4Szfybv13fNKQi+UblTBFquTDoDLvidPAg/cZ0UDhYoKuy8omoXNDe2dcB2bqxrn++47sYSqeY58nL39u9G8cSW2FCPQ07M354wEl884TyhBZpj50FV7MY3xXMz5Sz3tqm0kW27E2th90opxPGK+PMo6PQCcC8Uu3m+XIvyFcSJRVPJqN7r7lSZdDgFTV10vJVvVO+DPnnOKTj2C3Deb6Sf7L2fLNg2KCrsvKJsiywdYTfwEBhzHKx4quO6Hz4BPQe2zSV1AzdEwpZ3YdjhmY3UKUTaeezq5ecgb4JejBRb/yp07wvj5ri3TaWNbHPsqvfIQ9wIj/LrQJqJgzxYhqEq1o3iicpdkveWbxgWOvfYNVRDY7X7HrtU9IwJO/XY5UfNXug7LP/tHP9luOYpcc786WR49t8lfaKA0OIJr3BancQ3IJ15Kbx4s1TFDTtMXqsrl/y6Iy+TXDy3yFcktDRL/sfMS92zKWzEFys4HrsgmzC32uNiKHbTWzDhFHfPLSWOLD12u53CCS89drHLejSWXhB0Vawb7U6c6EE+jYlb7emkSKl6jyz98tj1VI9d3rS0SPPqfD12DpPPhJuWylSKpX+VVij9Rsts5T5D5TkuWiuN60PYo1A9dl5RtlmeVAeMa3ttxmfktRXz2l577w6I1sBx17u7/3wnT+xeIVW6XbVwAjr2+4KQ5Ni5FIqt3AmlG9y5GSrJybZ4wqmIHTHDG3ugLceuoUqWXcFjt3k+9BjgTiVxZx47ZwSkXx67Xuqxy5u6Mvk++7jgsXPoMwTOuxW+8zGc91uYcDJUbJciyI1vyD3SuW+EDPXYeUX5VlH48Z6SfiPlJvvBwzDnayIo3rsTZlwiTwJuYvJsnbHlXVmO76L5ddD+Ah91hF2AN0G3q2I3xSZAuBG+UpKTbSh2z0p52Os1yDOTWh/qGqtlGYocu3yF3dtSxOVG4+vO0lSqY8LON49drJVUfYU/++uK1MRGxbsRik2k92AZPXa8y84XD1GPnVc0VCW/eJ/xAylrf/QqmSHbVAenf8/9/edbPLH1XQkj51NlFHYSk8whYI+dy8UTm9+SMI+X+VwHO9k2KN690ltvHbTl2DU4wi7gqth8Q7FOft2EU9yxp1OPXSwU65fHrqib5MFqKDZ3qmPCzq1QbIGjws4rmuqS57aMmw0X3gHbFkio7JJ7YGiOY2vSkY9IsFY8dl05DAvt8xCbQuCxczsUu2m+3Ay16amHZNHuJFov7Y5GeJhfB23nUWg8dnm2O2nNr3NJ2HVWFVu9G4q6e+tVTaTnwPCFYrcugAV3BW1FZjg9APuqsAMNxXpHuo7vs66AEdNlXI1XHa7zKZ7Y/zHUHui6/esckhVPBBqKdWm+L0gqQPkWmPP1/LelpCYbj92+j+S79bJwApIIu6Bz7PJsd+Jmfp1jD6T32PUd4d1M6GT0Ghg+j92yh2DFkzIazc9jkQutHjsPQrEFiD7Ke0W0Ln2F5ciZ3o4tieSRr+U8IXd1YZe03UkX6WOn+XX+kE2OXWvhhMehcafPmxOKDboqNt/iCTfz6yCDqlgPmhN3Rs8B4fPYNVTJdTFsdiWjZq+kIPjpZQ0xKuy8Iuj+Ua3FEzlcULe8C31H5jcAuxBoFVJhq4p1od3J5vnQewgMm5b/tpTUZOOx270SuvXOfkh5toTNY5dPjp3b+XWQmcfOr8IJh54h9Ng5VdVOlXCYqd4n3rqwexZ9QoWdV3TmsfOaXCssW1pkEsahp3b9P5L4nLYweOzcqoq1VvPrfMPJsctg1T0rZWavW56nVHQongjaYxebFZvN2DUHt/PrILOqWL89dr1CmGPXUCnLql3B2pEJNXu7biP9HNCrvlcE7bHrLNyQij0rJRF10pnu2xQ22hVPOA2KwxCKzbM1RNkmqNyu/ev8INORYtZK3yuv8+sgzmMXoj52kJuwczu/DtI/QDU1SE+0IDx2YWt3konHLloPHz0Pb/9OmqHn8h27QfVeLZyIQ4snvCJaL2GXoMi1eGLDa7KcdLq79oSRpO1OQuCxyzcU25pf94n8tqN0TqY5dpU7xSPjdUUsdMyxC1rYReKLgrL0JWya725+HaR/gKr2udWJQ6+BEjpvjrY1mA6aVmGXwmNXXwkPXyYdHhxGHwOX3u99ukEiNfvEG64A6rHzjmhtsN6fXCssN7wuNx+/n1iDIP4YOR67oiBz7CKAyT8Uu3m+3JiGTnXFLCUdGbY72RcbJu7HzafVY1cjy8DbnWTZ68+hdKNMTpl4mkf2JPk7c6orfS+ecKZPhMhrl85j19wkom7HErj4T/CfG+CC38t3ds8ZbQ3u/cBaEXZeNCcuUFTYeUFLM7REg20MmkuFZWON9C46GLx10P4YReukd1XQOWmRPFtDxOfXdfUcyTCQqWjZv16WfohtJ8fOKZ4IMtcX4jzRWZ7Xa1+U5dRz3LUn3bWxrkyWvYe4u8/OcKZPhKWAoqUlvcfug4fFU3fhHXDU1ZLfdux18JXX5dg9eIk4CfygvgKaG7U5cRwq7LwgDMO3c0nE3/C6/IFM/qQ3NoWNdu1OGoIPWUEs0TwPYXdgvSR/a36dP2Qait3/sdy8/Ujwdjx2lTuhR3/5FyS5tl5a+7x4ON0O66W7NtaWytLvthlhmxcbraH1nE702DXWQsn/wdjjYdaV7d8bMgm++IIsH7kC1r7gva3anLgDKuy8oDVfKwweuyzCHx8+ISXj412sQAsz8e1Fgi52cYgU5+ex2/SWLDW/zh8y9dgdWAdDpvjjRW3NR7Mw4zPt51UHQS6h2LoyCecddq779mTksRvs/n7T4YRiw+Kxc7x1mI7CbvG94sX75E+Sn899h8F1z0pKz+PXwKqn87enqVHyD5PhiHG/v7MQo8LOC0LhscvyYlpfAR+/BEd8ti35uqsTL37D4rHLNxS7eT70G931exCGhgxz7Pav92Z0YDKMaXtoOfoaf/aZjlxCseteEY/aYZ/2zp5kHru6mEhwQqN+ETaPnSPsBo4TYec4CFqaYdE9EhGYkGbkZO/B8Pl/ildv3pcYsTvHsGx9BZT8Em6dDP/4Wop1YsfMEceKCjtPcIRdkB67bIsn1jwLzQ0w83LvbAob7Yon6oNPModYl/4chZ210qVf8+v8I5N2Jw1VULUThkz2xyYQYTdkitxYgyaXdidL/woDDpEqS7dJF82oKxNR53WvwUR6hlTYDZ0q+eKO4F33MlRsheO/3Pk2evaHa56CCacw7aPb4cX/bitS64zybfDa/8DvZkLJz6XLRPm25Os6BScq7FpRYecFTY7HLgyh2AxEgrWw9AEYdCiM8eBCGlaMkZuO0+4kFB674tzbnexdI/kmOkbMPxzRki7H7oBTOOGTxw7g8PPgtO+GQ+BH0jxkNkdlakA8O96XxsRzbvSmmCndQ29dGfQKIKQXtuIJRyw5xT5OAcXie6HfKDm/MqF7H7j6SbaPOQ8W3Al3HC8zaJ1WPPE0VMHqZyQ37/YjYf5vxDP41bfEO9iSIhTrHDO/vawh5iCJuflMNJZjVyjFEx/9C7YvgvNvC8eNwE+cYoWm+uBnakJ+odiNb8hy4ly3rFE6JQOPnVMRO8RHYXfp/f7tqzPShWKf+TdY+RRc9lc4PBZ2fe+PUvBx9LXe2JPuobe2NJh5o916yoNl6Dx2sXO2ag9gYP2rMPd72fXa69aT9VNuYOwZX4FXfgj//Ab861sw6kgRic1RKN8iBUYtTVLdesq3pMp24CGyjUi31Dl2zjHrpR47BxV2XtAUglBsayVaJzl2TY3yxzbscDj6897bFTYcIRWWHLt8qmI3vC7iwbkYKt6TSZjxwDpZ72DNe0yV77trOSx/BLr3kyT7U78thR4rn4KTbpJQnif2pMuxKwsuCb/XYKgtC2bficSHYkEmEi2+V6aAnHBDbtucdDoc+iZsfRc+ek62uW+tfOcDxkmhzKQzYNwJHYVjUbfUkYz6CujWJzyNnUOACjs32buGfpVrITpefg+Dxy5dVWxTAzz5RWkq+bl5B0/RRDxOFWpTfTBP6qnsieflW6QnWZ8LUn8uWg+b34FjDkJxHiSZtDvZv07Edhg8wkGQqt3JKz+Sv7kb34FXboG3fiWvH/FZOP373tuTqip2yCTv9p2OPkPaWncEjSPshk+HsbPhtZ+IMD/jlvyuk5GI5ABnO/s3UpzaY1dXrmHYBA7CO7mHvPZTDtu+Gib9WH4PtHjCCRGl8P60NMMTn4ePX4Rzfw1TzvLPtjDRGooNiccuEul4w1n1D6jYysRxZXD6Gck/t22BeIonpXhf8YZMPHblW2DQBF/MCSXJPHY1ByR1YO73YMAYCR3P+YaE42Zd6W1KSGdVsUE94PUZBrX7g9l3Io6w6zkArn0anroe9q6GE24Mxp6ibqlz7OrLNQybgAo7N+k5gOKmmnC0O+mseOKVH4qo+/StMPsr/tkVNhwhFa0LSVVsQii2qQEqtkGfYRyy7WnY/JXkbQY2vC55KNk+CSt5kkG7k8qdMOlMf8wJI8ly7Gpio7viC0rGHiv/vCZVVWxLs4T1giieAOg9FA5sCGbfiTRUSngzUgQ9+sLVjwc7xzbSTcaYJaO+QitiE9CqWDfpObC9sAvUY5fmqXTrAnjvDph9w8Et6iDBYxfw6CXoWBVbtgWwcMYPaI70hBVPdPyMtZKzMv5EuQgr/tFZu5PmqPQBGzDGP5vCRlKPXSzk2CeA+Z6pqmKd6srAPHZDofZAMPtOpKEKevRr/1qQOWxFxemrYjUU2w4Vdm7ScwDFzbVtw7fD0O4k2Q1nY4lc3M64xVeTQklr8URI+tglVsWWbpTliCPYP3S2tANIzDXZtVxaahzxWf/sVITOcuyqdst7/Uf7ZVH4iIRN2MW1OYonqKkTDr2HSC6t4xgIkmTCLkiKuqepiq3QUGwCGQs7Y0xvLw3pEjgnV/UeWQYp7Jyn0mSh+jvlDwAAIABJREFU2K0LYPgM76rOComw5diZovY3QEfYDTqUvcNPkRygjW+2/8zKp8TTN+1C/+xUYnTisavcIcv+Y/0xJ4wkDcXGcsn8mJ2bjGTV546wCzLHDtqOTTIaa2DbIlj3amqh4wZhE3aRdFWx6rFLpFNhZ4w5yRizGvgo9vssY8ydnltWiDgnV3Ws509RgDMaTQqPXUszbF8C42b7b1MYCZ3HLtLRY9djAPQeTOngY+TnFU+2vd/SAiv/LjlcOivRfzornmgVdgexxy5VKNZEghNRyfpFOtMVggzFQuoCirItcMdsuO8sePizcMdxsPYFb2xpqAyXsCtKURXb0iy2ao5dOzLx2N0GnA0cALDWLgdcmTBujDnHGLPWGLPeGHNzkvd7GGMej72/0BgzIfb6EGPMG8aYamPMHW7Y4gqOsKvaLd66IJv9piqe2LsGGqvgkDn+2xRGIkXQ3Cj5G0F6WFvtScixK90Igw8FY7CRbnDk5bByHpRukvc/fgEqt2sYNig6C8VWxITdwZxjl6zdSc1+CT36PbrLIdEzDsF77HrHhF1Nkjy7qj3wwAUSqr3sr3D5g9C9Lzx6lcxudZtQeuySCDtnQoaGYtuRUSjWWps4pC2PKeWCMaYI+CNwLjAduMoYMz1hteuBMmvtZERg/jL2ej1wC/CdfO1wFeepoWpX8GG9VAnC2xbKUj12gimCxlr5OQweu8QQUenG9o1tT/22iL83fylhmRe+C8OmwRGX+G+rkrr5rkPlTrkB9ziI0x5SeeyCyK9zSOaxqw2Jxy5ZL7s3fyHn0rV/hxmfgekXwvWvwNRz4PnvwII/uWtLQ1W4zlmnQXGiZ7xex4klIxNht80YcxJgjTHdjDHfAda4sO/ZwHpr7UZrbSPwGHBRwjoXAQ/Efp4HnGmMMdbaGmvt24jACw+tHrs9wXt/UhVPbFsEfUfAwPH+2xRGIkUQjRW7BC3Gof0NpzkK5VvbC7v+o6SS+cPH4c+fkFYo5/9Wu64HRiftTiq3Q/8xB9+ovniSNUuv2R9cfh2I2EyaY2eCC+ulCsVWbIf3H4RjroUxce1guveGKx6Cw8+HF2+WGaxuEbZQbCR2fUvMs3M8dhqKbUcmwu5G4BvAGGAHcFTs93wZA8R7ArfHXku6jrW2CagAhriwb29w3MGNVcGLhFTFE3tWwqijDu4bTTxh89jFT56o2CY3n8RRVKd+B465TvK25v43jD/JfzsVIROP3cGcXwdxVbHxodh9baHHIEiaY1cm1/BIQM0ievQXAZNYPPH272R5yrc6fqaoWJo7TzwdnvmmNDPPF2vDF4p1piIl5tnV6ZzYZHTaoNhaux/4nA+2eIIx5gbgBoARI0ZQUlLi2b6Kmmo5NfZzdUMLSzzcV2dEmhv5BLBxw3q2NrXZcWLZTkrNSNYGaFs81dXVnn4nnXFcbR2mpoY+wJp1m9hTFZwtALPKK4i0NLKspIRBpR8wC1i2pZyKipL2x6rfxeBcd0PyXYYNP86t3jXbmA2sWr2Kffs7Fq+cuG8jpYOPDc3fWzq8Ol6DSlcyC3j//aVUbpCHqJMrd7On1+GsD+i4nBRtZv+O7Xwct/9pWz6iHz1ZlKFNXhyvE4v7UbphJWu7yXaLmmo5aenf2Dv8E6z9YAOQvIFxZMyNzNq/m37zrmf16jXsH3ZizjZEmuv5hG1hw459bHPp/5fvsRq7bSuTgflvvUFzcZ/W14ftfZcZwOIVH1OzqTFvO8NCvserU2FnjPkLSTKDrbVfynmvwg5gXNzvY2OvJVtnuzGmGBhArIgjU6y1dwN3Axx33HF27ty5udqbyc6wb0cwtNB30FA83VdnNEdhPkycMJ6Jp81ttY/5NYyaOJ1RQdoWR0lJSbDH6aP+ULYZgGnHfYJpkwO0BWDbUKivlGOyugI+hKNPnAsjZgR/rAoMX47X/nWwGGZMmwYzE/bVHIWSckYddmxo/t7S4dnx2mjgQzhm1pEyNaWpEUpqGDv1aMY61ya/WdqL0SOHMzr+/7vtdug2OuNj4MnxWjOGUf2L286XZQ9BSwOjzruZUZ3lRZ80Bx6+jCNW/wrO/x0ce112+7YW1r0M838LwKSjP8GkxHM6R/I+VgvXwgY49cQ5MlPXYelmWA3Hn3pWlypQyvd4ZeJz/hfwXOzfa0B/oDrnPbaxGJhijDnUGNMduBJ4JmGdZwDn7LwUeN3adLN7AsYYmpynicBDsUkq0aJ10NygbTHiiRRLPgmEI+8wviq2dYJJCHL/lBSkybGr2oU0J+46N5ycSAxX1wbcww6SV8WGYTRVnyHtQ7EfPAKDJ8HY4zv/bK+B8Pl/SFj22X+Df3y9rdI3HfWVsOR+uOtUeORySR/49K0wI0QFWZGYDyqxMlZDsUnJJBT7VPzvxphHgbfz3bG1tskYcxPwElAE3G+tXWWM+SmwxFr7DHAf8KAxZj1Qiog/x47NiMjsboy5GPiUtXZ1vnblS7RbX7o1VYWgeCJJ7k/QfZrCSHy7hYHjUq/nF/FVsa0zh7U3eGhJN1KscpcsD/ocu4SHzNapE0Hm2CWZPFFfCQMPCcYehz7DWiMIlG2GLe/AGT/IPCe6ex/43JNQ8gt469cyavCEG6WSduhU+X9bK71Wty2EtS/C6n9AtFaa1l90p7RUClsxlmNPYo5dfYWIPr1GtqNTYZeEKcBwN3ZurX0eeD7htR/G/VwPXJbisxPcsMFtQuOxg45jc1rL+dVj14rj2ew3KiTFE0Vt1YOtwi4E55KSnHR97JwHqd7hrffyhcTJE61TJwJsd5Js8kR9RfBtM3oPbetjt+Qvcg0/8sr0n0kkUgRnfB+mXQCv/VRapbz5CyjqIZ6txlop8AMp2Jh5mRRjjTkmvEV1rVWxicKuXLysYbU7IDLJsatCrlomttwNfNdjuwqWVmEXtMcOOl68Wm80KuxacbwJYQjDQvs2DE3qsQs/aUKxrQ1vD/IwUeJ0jjAIu2RVsfUVwfdu6zNERFfNflj6FxFnuUYSRh0J18yT0Oq6V+DAOvFKdusllfajZkn7lLB555LR6rFLaHdSp+PEkpFJKDZENc/hp6m4r/wQBo9d4sUr6M7qYcS56QQdgnHokGMX8Gg6JT3p2p3o35uQ2O4kDKHYxIfeaL3kHwctEkbOkuU9p4vQPPGb+W+z/+jsCynCRqocu4bK4L+zEJJS2Bljjkn3QWvt++6bU/i0eexC4GVJTBDWUGxHWj12YRF2cWI8Whf8aDolPWlDseWAkfm+BzOJ4rdmnzysBOkdS3zodQqoghYJUz8FZ/8cXvoejDsBxmVQNHEwkCrHrqEKeoZoQkZISOex+02a9yxwhsu2dAlaPXZhyIsykeTFExqKbcPJ/xkUllBsQvFEGEL6Smo689j1HBBcw9uwkJhj5+SyBfnAkvjQWx8SYQdw4jdgxAwYdGjQloSHVJMnGqqgrysp/12KlMLOWnu6n4Z0FdqKJ0JwQ06s/Kotg259wlEkEBYcF39oPHZxkyea6sNxHilpSJNjV1+u+XXQcW51c2PwqSqJ18bW0VQhEHYAE+cGbUG4SDV5ImwzbUNCRlWxxpgjgOlA61+jtfZvXhlVyLSFYsPgsUssnijTfJ9EwlY8EX/DUY9d+OnMY6d/bx3nVjfVB583mnhtbIgJOxUJ4SRVVWzYZtqGhEyqYn8EzEWE3fPAuUgfOxV2SQiXxy6xeKIUeuuNph0mIv8GjA3aEqFDKDYEDwhKajrLsVNh1zEU29QQAo9dUbg9dkp7kuXYhXGmbUjIJPnjUuBMYLe19ovALGS0l5KEaLcw5dglKZ7Qwon2FHWXyQBhKfmPr4ptqgtHEY6ShjQNiuvKgp9kEAYS2500N0JxyDx2rcJOPXahJFmOXbRW/u5U2HUgk1BsvbW2xRjTZIzpD+yl/YxXJY5Qeezie6KB3GgO9i74iZzy7+1H+ARNYoPioD0bSnoSRUs89eqxAzq2O2mql2a5QRL/dwbhKp5QOpIsx67BabKswi6RdO1O/gg8CiwyxgwE7gGWInNi3/PHvMKjtvchMPWccJSpJ1686kq1IjaRUbOCtqA9iaFY9bCGm1SVnS0tsRw79di1it/WUGxj8AVcJtLe+1NfIa917xucTUpqkuXYNcRNz1Dakc5j9zHwa2A0UIOIvLOA/tbaD32wrSBpLu4FVz8etBlCvMeu9UajQiHUdOhjpx67UJOqeKKxSl5Tj11bjp1zjJobgg95Rook18+hoVIEgvaMDCfJJk84vQfVY9eBlDl21trbrbUnAp8ADgD3Ay8CnzHGTPHJPiUf4kVCQ4VcWNVjF24iRXE5dvWaY1coJIZi68plqTl2cVWxccUTYauKra8IXmwqqUk2eUJDsSnptHjCWrvFWvtLa+3RwFXAxcBHnlum5E988YSONyoM2oViazXHLuyk8tjp31sbiccorFWxml8XXpJVxaqwS0mnws4YU2yMucAY8zDwArAWuMRzy5T8iQ/F1jo3GvXYhZpITIxbK/MrtY9duEnV7qQ+5rHTHLuO7U6aw5Bjl+ixq1TvaphJm2Onwi6RdMUTZyEeuk8Di4DHgBustTU+2abkS/xTqY4TKwxaQw7N4rFTYRdyUrQ7UY9dGx08diFoUNyhKrYCBk0IzBylE5Lm2GnxRCrSFU/8N/AI8G1rbZlP9ihuEh+Kdf4ItOor3Dg3wWgtYIMPWSnpSdXuxMmxU2EX1+7EEXYhGCmW2AqqoVJDsWEmaY5drHhC72kdSDcr9gw/DVE8IH48VXOjLIMOgSjpcRLNG6tlqcUT4cZ04rHT8F6SUGxD8A2Kk+bYqecntCTLsauvlAeEoM+lEJLJ5AmlUDGR9uEPCP5JWUmP82Ta4Ag7/b5CjeOxS8yxqyuTJrwaSm8firU2HA2K43PsWprVYxd2kk2e0HFiKVFh15WJv3g5PZtU2IUbox67wsLx2CUpnug1UPuiQft2J47HJejIQbtWUJqrFXqccyixKlaFXVJU2HVl4i9ercJO3dahJjEUq0I83KRrd6L5dUL8MWqNHITBYxf7zlrnxKrHLrQYI167xKpYFXZJUWHXlYm/eKnHrjBwhF2DeuwKglTtTurKNb/OoTXHrqUt1zfoUGx8/rGThK85duGmqFsSj51+Z8lQYdeVicQJu+YGeXKOpCuEVgKnQyhWhXioSVUVG62F7n38tyeMxIdiWx8ww+Cxiwk79dgVBpFummOXISrsujLGxIViYwnLmvMTbrQqtsBIkWMXrdPCCYf4yuGwhGLb9fjU8W8FQVFxgseuUoVdClTYdWUSiyeCvpgqnWMSQrEaOg83qdqdRGtVlMdjYkIqLG2X2nnsdEpIQaA5dhmjwq4rk1g8EfTFVOkcJ1Te6rFTr0+oSZVjpx679jitl5xQbOA5dnGTJ9RjVxgUdWubPGGt5tilQYVdV0Y9doVHh+IJFQfhxyTx2NWpxy6eSFFCjl3A1fkdPHZGRULYiRS3eeyaGuRn9dglRYVdVya+eKKpXsN6hYCTjN8Y662lwi78mIjm2HWGiVWhNoekOj++KrauXCpiI3o7DDXxVbGtvQdV2CVDz+SujIm0hRuaG4MPfyid41RSVu+VZbGKg9BjDO1Csc1R8SaosGvDFMWmToSk3Umix07DsOEnvirWaVGjXtakqLDrysQPum6q11BsITB0iix3fQgY/c4KgfjRfSDeOlBhF08kdi1qrYoN0azYunItnCgE4qtiWz12fYOzJ8SosOvKtCueaAw+/KF0zsDx8j1V7RRhoO1pCgDTPhSrwq4jjvgNSyg2sY+deuzCT3xVrHrs0qLCrivTrniiPvinZKVzIkUwJOa1C/rmp2RGB49drSy1eKINp91Ja1VsCDx2IKkq9eqxKwjic+zqdVpIOlTYdWXibzhNDSoUCoVhh8lShUFhkJhjpx67jjhpIWGaPAFik45/KwwixZpjlyEq7Loy8b2amrXdScEw7HBZ6jixAiFVKFaFeStOhX5rg+IQVMWCeBHVY1cYJPXY6Ri4ZKiw68okhmKDrkRTMqPVY6cen4Igsd1JayhWv79WTOwh0ymeCDoU2zqTuUbEpnrswk98jp0z31fbnSRFhV1XJr5XkzYoLhwcj522OikMEkOxjnhRj10brZMnQjJSzMmxqz0gS/XYhZ/4yRMNlfL3VdQtWJtCigq7rowpSsixU2FXEAw+VJ5O1eNTGBiTonhCv79WnHYnzQ0i8pzReUFhEoSdeuzCT/zkifoKza9Lgwq7rky7PnYq7AqGom4wfBr0Hhy0JUpGpMixCzqPLEw4kyeclJCg2/iox67waDd5olLz69IQqLAzxpxjjFlrjFlvjLk5yfs9jDGPx95faIyZEPfef8deX2uMOdtPuwsGp4+dtTpSrNC44iE45xdBW6FkgrY76RwnetDUGI62S87oPvXYFQ7xkyfqK7XVSRoCE3bGmCLgj8C5wHTgKmPM9ITVrgfKrLWTgduAX8Y+Ox24EpgBnAPcGdueEo9zMW1pAqwWTxQSg8ZDv5FBW6FkgrY76ZxIUVsoNgwPmOqxKzzaTZ6o1FBsGoJMdJgNrLfWbgQwxjwGXASsjlvnIuDHsZ/nAXcYY0zs9cestQ3AJmPM+tj23vPJ9qT85NlVvLu6jj+tDdSMVq6t3MMnGxv56t1v8QDw4JLd/OujcNjmUF4enuMVdvRYZYdfx+vu+iYWrtrFfdtlX5+pWsuVwNUPfEizKZzkbi+P16/L6tlVtZ96U8O0Rss3/xzseXxG7Wa+Cjy7YCUXAF96fD01kT1ZbUP/HjPHjWN1fUUpc+rr+Mqf3+O3e/ewpVsfbg/wPJo+uj8/umBGYPtPR5DCbgywLe737cAJqdax1jYZYyqAIbHXFyR8dkyynRhjbgBuABgxYgQlJSVu2J6U7dsbaG5upry83LN9ZENdtBFsC7UV+wGorG+ivCkctjmE6XiFHT1W2eHX8WpugWhDQ+u+bLSCZiIcKK8OPpcsC7w8XtFmS3NLA9a00NBSFPh5XB1rlNyjfh8AuyqitJjsbNK/x8xx41jVNDYRaYlSXl5Or//f3v0H2VXWdxz/fHYTgmT5FYkxEFooxh+IM0C2KJU6G4HwazS0pR38o007dbAWprUzjkYdC1XooIO1tWNlUk1NrTVSW2tqlRiRLe0fyK/GQlBMqjIQESzh1ypks9lv/zjnbM7evXe5N2fvfe6efb9mdu45557kPvvNuTff+3zP8zwHn9PTWpQ0/o9OPqvR0Z925e8eGxurlKskHprUfRGxSdImSRoeHo6RkZGuvdbIiDQ6OqpuvkZHdtwp3fkN/dM7z5M+Ll198Vm6es0lqVs1TV/Fq88Rq870LF43HanLXnmSLntr/t669Q7pvqXavvHS7r/2HOpqvG7+c512zAnZyManntP2dyb+HNr5tPSv0rrTjpIePkZf33hZx38F78f2zUmstv+XdM8d2v7eS6Tr9+vSc16rS9f11/9nc6VqvFIOntgr6eTS/qr8WNNzbC+SdKykJ9v8sygGTxTzavXDvS1A3XhAM+6x4/666YqVJyZeSD85cdEeKbvHjoET80MxKnZiPLuOljAqtpWUid3dklbbPtX2EcoGQ2xrOGebpA359hWSvhURkR+/Mh81e6qk1ZLu6lG7549i5YmDfTIpKFBLTaY7IbGbbmq6kz4ZPFEeFfsSEoR5oVh5olgnllGxLSUrxeb3zF0jabukQUmbI2KX7Q9Juicitkn6jKTP5YMj9ilL/pSfd4uygRYTkq6OKCZsw5TiW2kx/QKjYoG512xJMRK76aamO9kvHdEH08AUn43P/lg65by0bUF7ilUmnn8qe2RUbEtJ77GLiK9J+lrDsT8tbb8g6Tdb/NkbJN3Q1QbOd8W30qkJU0nsgDnXbLoTErvpytOdDB6fujWHVp6YeEE65VfTtgXtKVYrmZp7kMSuFVaeqLMZiV0flECAupmxpNjzTE7cqOjVnBjvjy+YA6VpT1dfmK4daF/RYzeV2FFCb4XErs4aS7H98IEK1E7jPXaUYmcoLynWD59DRY/d0SdKL2ucFx99aSBP7H6WTd9FKbY1Ers6Kz68xknsgK5p2mNHYjdNsW71wfH+uNe3+NL7ivPn1VyDC9ogpdh2kdjV2VSP3c+yx374QAXqZsZ0Jz+nFNtooDR4oh++YBZteMUFaduB9g00lGLpsWuJxK7OGDwB9EBDj93EC/TYNfJgabqTPvgc+oVzpctvll7zltQtQbsa77EjsWup9itPLGhFKZbBE0D3zJjuhMETM3ggS34P7u+PCYoHF0tnvi11K9CJco/d4qWHSrOYgR67OhvI/3nH81Lsoj74QAXqZsZ0JwyemKFYBefgOF8wcXjK99hxf92sSOzqjB47oPuK3igpW/JockJaRGI3jQekieJziC+YOAzlHjvKsLMisauzqXvsWHkC6J7SdCfFe40eu+k8cOgLJp9DOBzFPXbP/lhaujxtW/ociV2dleexG1h8qDQLYO6Ue+yK5IXEbrqBQWn/WLbdD4MnMP8UK08cHJfO+5O0belz/E9fZ+VSLGVYoDvK86BN9dgxeGIaD0j7n8m2TzwrbVswPxWDbs74DWk109TMhmEldVb02I3/jG/JQNeYHrsXU3zJPPZk6aQ1aduC+enEs6Rzr6G3rg0kdnV2xNLscexxEjugW1y+x+6F7JEeu+mK+31PX89KDzg8S4aki25I3Yp5gVJsnb38ddnjTx8isQO6pbyk2FQpllsfpimqB6dfnrYdwAJAYldnx54sHfVSScFINKBbykuKTezPHrmndbpjTpSWv1paNZy6JUDtkdjVmS2tPDPbpscO6JJSKXbyQPY4wF0u06z9gPSOOyjDAj1AYld3xQg0ehCA7pg2QfF49tgPy2b1E5svl0CPkNjV3YlFjx3/0QBdUV5S7GDeY1dMpgoAPUZiV3f02AHdVe6xm5zIHknsACRCYld3x5yULb9CYgd0Sekeu6IUO0BiByAN7vCtO1ta/0npqBNStwSop/J0J5RiASRGYrcQvPKi1C0A6sulwsdUKZZ7WgGkQSkWACppVorlOzOANEjsAKAKSrEA+giJHQBU0XS6E0qxANIgsQOAKqZNd3JAkg+tjQoAPUZiBwCVlO+xO0BvHYCkSOwAoIrGe+y4vw5AQiR2AFCFBzR1j93kAUbEAkiKxA4AKmnssaMUCyAdEjsAqMIDDffYUYoFkA6JHQBUUZ7uhFIsgMRI7ACgivJ0J5RiASRGYgcAlXiqw04HxynFAkiKxA4AqihPdzI5QWIHICkSOwCoYtqSYuPSAIkdgHRI7ACgEiYoBtA/kiR2tpfZ3mF7d/54fIvzNuTn7La9oXT8BtuP2B7rXasBoInydCeUYgEklqrHbqOk2yJitaTb8v1pbC+TdK2k10s6R9K1pQTw3/JjAJDWtCXFKMUCSCtVYrde0pZ8e4uky5ucc5GkHRGxLyKekrRD0sWSFBF3RsRjPWkpAMymvKQYpVgAiTmKEkIvX9R+OiKOy7ct6aliv3TOuyUdGRHX5/sflPR8RNxUOmcsIoZe5LWuknSVJK1YsWLN1q1b5/aXaTA2NqahoVmbhBLi1T5i1Zlexes1D35MRz+3R3e9/lMavvuP9PxLTtSuM2YUIfoe11dniFf7iFVnyvFau3btvREx3Mmf79oU6ba/KenlTZ76QHknIsJ217LLiNgkaZMkDQ8Px8jISLdeSpI0Ojqqbr9GnRCv9hGrzvQsXk9+XjrwSPZa9y/W0IqV8/LfieurM8SrfcSqM1Xj1bXELiIuaPWc7cdtr4yIx2yvlPREk9P2Shop7a+SNDqnjQSAqqZNd8LKEwDSSnWP3TZJxSjXDZK+0uSc7ZLW2T4+HzSxLj8GAP2jvKTY5IQ0yFqxANJJldjdKOlC27slXZDvy/aw7U9LUkTsk/RhSXfnPx/Kj8n2R20/Kuko24/avi7B7wAAmrGkGKNiASSU5KtlRDwp6fwmx++R9PbS/mZJm5uc9x5J7+lmGwGgLW6coJhSLIB0WHkCAKqYcY8dPXYA0iGxA4BKSj12kwekAe6xA5AOiR0AVFEsKRZBKRZAciR2AFBFUYqdPJg9UooFkBCJHQBUUUx3Mnkg26cUCyAhEjsAqMSHyrASpVgASZHYAUAVxXQnU4kdpVgA6ZDYAUAVHlB2jx2JHYD0SOwAoJKGHjtWngCQEIkdAFThgWx+4oPj2T49dgASIrEDgCqmpjuZyPZJ7AAkRGIHAFUU050UPXaUYgEkRGIHAFVNm+6ExA5AOiR2AFDF1ATFlGIBpEdiBwBVFPfYUYoF0AdI7ACgksYJill5AkA6JHYAUIUHsnvsKMUC6AMkdgBQxdSSYkUpdlHa9gBY0EjsAKCKYkkxSrEA+gCJHQBUYkqxAPoGiR0AVDHVY0cpFkB6JHYAUIWdPU6tFUspFkA6JHYAUIXzj9GJIrGjFAsgHRI7AKik6LHbnz2S2AFIiMQOAKrI8zpN5IkdK08ASIjEDgCqKEqxBynFAkiPxA4AKmkYPEGPHYCESOwAoIry4AkPSgN8rAJIh08gAKjCpcETlGEBJEZiBwBVlO+xYw47AImR2AFAJXmP3cQ4q04ASI7EDgCqmOqx20+PHYDkSOwAoAqXeuy4xw5AYiR2AFBJafAEpVgAiZHYAUAVU9OdUIoFkB6JHQBU4dIExZRiASRGYgcAVZQTO0qxABJLktjZXmZ7h+3d+ePxLc7bkJ+z2/aG/NhRtv/d9vds77J9Y29bDwBl5cETlGIBpJWqx26jpNsiYrWk2/L9aWwvk3StpNdLOkfStaUE8KaIeLWksyS90fYlvWk2ADSYNt0JpVgAaaVK7NZL2pJvb5F0eZNzLpK0IyL2RcRTknZIujgifh4Rt0tSRIxLuk/Sqh60GQBmYroTAH3EEdH7F7Wfjojj8m1LeqrYL53zbklHRsTzi5/9AAAJBUlEQVT1+f4HJT0fETeVzjlOWWJ3QUT8oMVrXSXpKklasWLFmq1bt3bjV5oyNjamoaGhrr5GnRCv9hGrzvQqXit//A296vuf1IFFQ3rm2Nfqgde9v+uv2Q1cX50hXu0jVp0px2vt2rX3RsRwJ3++a3f62v6mpJc3eeoD5Z2ICNsdZ5e2F0n6gqRPtErq8r9/k6RNkjQ8PBwjIyOdvlRHRkdH1e3XqBPi1T5i1Zmexeu+R6TvS4snxnTCKa+Zt/9GXF+dIV7tI1adqRqvriV2EXFBq+dsP257ZUQ8ZnulpCeanLZX0khpf5Wk0dL+Jkm7I+Iv56C5AHB4ilKsJC19Wbp2AIDS3WO3TdKGfHuDpK80OWe7pHW2j88HTazLj8n29ZKOlfSuHrQVAFpz6WN0iMQOQFqpErsbJV1oe7ekC/J92R62/WlJioh9kj4s6e7850MRsc/2KmXl3NMl3Wd7p+23p/glAGBquhNJWro8XTMAQF0sxc4mIp6UdH6T4/dIentpf7OkzQ3nPKppn6QAkBA9dgD6CCtPAEAV3GMHoI+Q2AFAFdN67CjFAkiLxA4A5sLgEmnJMalbAWCBI7EDgCqKUuzQy6aXZQEgARI7AKiiKMUyIhZAHyCxA4BKSj12AJAYiR0AVEGPHYA+QmIHAFWYHjsA/YPEDgCqmOqxI7EDkB6JHQBUUvTYUYoFkB6JHQBUQY8dgD5CYgcAVZx8jrTmd6WT1qRuCQBoUeoGAMC8dtQy6S1/lboVACCJHjsAAIDaILEDAACoCRI7AACAmiCxAwAAqAkSOwAAgJogsQMAAKgJEjsAAICaILEDAACoCRI7AACAmiCxAwAAqAkSOwAAgJogsQMAAKgJEjsAAICacESkbkPP2P6ppIe7/DInSPq/Lr9GnRCv9hGrzhCvzhCvzhCv9hGrzpTj9YsRsbyTP7ygErtesH1PRAynbsd8QbzaR6w6Q7w6Q7w6Q7zaR6w6UzVelGIBAABqgsQOAACgJkjs5t6m1A2YZ4hX+4hVZ4hXZ4hXZ4hX+4hVZyrFi3vsAAAAaoIeOwAAgJogsZtDti+2/ZDtPbY3pm5Pv7H9I9v3295p+5782DLbO2zvzh+PT93OVGxvtv2E7QdKx5rGx5lP5Nfa/9g+O13L02gRr+ts782vsZ22Ly099748Xg/ZvihNq9OwfbLt220/aHuX7T/Oj3N9NTFLvLi+mrB9pO27bH8nj9ef5cdPtf3tPC5ftH1EfnxJvr8nf/6UlO3vpVli9VnbPyxdW2fmxzt/L0YEP3PwI2lQ0v9K+iVJR0j6jqTTU7ern34k/UjSCQ3HPippY769UdJHUrczYXzeJOlsSQ+8WHwkXSrp65Is6Q2Svp26/X0Sr+skvbvJuafn78klkk7N36uDqX+HHsZqpaSz8+2jJX0/jwnXV2fx4vpqHi9LGsq3F0v6dn7d3CLpyvz4zZLemW//oaSb8+0rJX0x9e/QB7H6rKQrmpzf8XuRHru5c46kPRHxg4gYl7RV0vrEbZoP1kvakm9vkXR5wrYkFRF3SNrXcLhVfNZL+vvI3CnpONsre9PS/tAiXq2sl7Q1IvZHxA8l7VH2nl0QIuKxiLgv335O0nclnSSur6ZmiVcrC/36iogYy3cX5z8h6c2SvpQfb7y+iuvuS5LOt+0eNTepWWLVSsfvRRK7uXOSpEdK+49q9g+ChSgkfcP2vbavyo+tiIjH8u2fSFqRpml9q1V8uN5auyYvWWwulfaJVy4ve52lrKeA6+tFNMRL4vpqyvag7Z2SnpC0Q1mv5dMRMZGfUo7JVLzy55+R9NLetjidxlhFRHFt3ZBfWx+3vSQ/1vG1RWKHXjovIs6WdImkq22/qfxkZP3ODNNugfi05VOSTpN0pqTHJH0sbXP6i+0hSf8s6V0R8Wz5Oa6vmZrEi+urhYg4GBFnSlqlrLfy1Ymb1LcaY2X7DEnvUxazX5a0TNJ7D/fvJ7GbO3slnVzaX5UfQy4i9uaPT0j6srI3/+NFt3L++ES6FvalVvHhemsiIh7PPzQnJf2tDpXDFny8bC9WlqR8PiL+JT/M9dVCs3hxfb24iHha0u2SzlVWNlyUP1WOyVS88uePlfRkj5uaXClWF+fl/4iI/ZL+ThWuLRK7uXO3pNX5KKAjlN0Qui1xm/qG7aW2jy62Ja2T9ICyGG3IT9sg6StpWti3WsVnm6TfyUdMvUHSM6WS2oLVcO/Jrym7xqQsXlfmo/FOlbRa0l29bl8q+f1Ln5H03Yj4i9JTXF9NtIoX11dztpfbPi7ffomkC5Xdl3i7pCvy0xqvr+K6u0LSt/Ie49prEavvlb5gWdm9iOVrq6P34qLZnkT7ImLC9jWStisbIbs5InYlblY/WSHpy/n9sYsk/WNE3Gr7bkm32P59SQ9L+q2EbUzK9hckjUg6wfajkq6VdKOax+drykZL7ZH0c0m/1/MGJ9YiXiP5NAGhbBT2OyQpInbZvkXSg5ImJF0dEQdTtDuRN0r6bUn35/f2SNL7xfXVSqt4vY3rq6mVkrbYHlTWYXRLRHzV9oOSttq+XtJ/K0uWlT9+zvYeZQOgrkzR6ERaxepbtpcrG/26U9If5Od3/F5k5QkAAICaoBQLAABQEyR2AAAANUFiBwAAUBMkdgAAADVBYgcAAFATJHYAUGL7pbZ35j8/sb033x6z/Tep2wcAs2G6EwBowfZ1ksYi4qbUbQGAdtBjBwBtsD1i+6v59nW2t9j+T9sP2/512x+1fb/tW/PlqGR7je3/sH2v7e0NKxcAwJwjsQOAw3OapDdLequkf5B0e0S8TtLzki7Lk7u/lnRFRKyRtFnSDakaC2BhYEkxADg8X4+IA7bvV7aM4K358fslnSLpVZLOkLQjX0pvUNKCWW8VQBokdgBwePZLUkRM2j5QWsR8UtlnqyXtiohzUzUQwMJDKRYAuuMhScttnytJthfbfm3iNgGoORI7AOiCiBiXdIWkj9j+jqSdkn4lbasA1B3TnQAAANQEPXYAAAA1QWIHAABQEyR2AAAANUFiBwAAUBMkdgAAADVBYgcAAFATJHYAAAA1QWIHAABQE/8PfetNXhh8Ws4AAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 720x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "\u001b[92mfeature: 5:\n",
            "\n",
            "feature_forecast.shape:\n",
            "(43200,)\n",
            "feature_valid.shape:\n",
            "(43200,)\n",
            "\u001b[92mMAE: 0.022585954517126083\n",
            "\u001b[92mMSE: 0.0052195661701262\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "\u001b[92mfeature: 6:\n",
            "\n",
            "feature_forecast.shape:\n",
            "(43200,)\n",
            "feature_valid.shape:\n",
            "(43200,)\n",
            "\u001b[92mMAE: 0.11454086005687714\n",
            "\u001b[92mMSE: 0.045187368988990784\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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