Copy_of_S+P_Week_4_Lesson_1.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Copy of S+P Week 4 Lesson 1.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true,
      "include_colab_link": true
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/RhuthuHegde/Tensorflow-coursera/blob/master/Copy_of_S%2BP_Week_4_Lesson_1.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zX4Kg8DUTKWO",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#@title Licensed under the Apache License, Version 2.0 (the \"License\");\n",
        "# you may not use this file except in compliance with the License.\n",
        "# You may obtain a copy of the License at\n",
        "#\n",
        "# https://www.apache.org/licenses/LICENSE-2.0\n",
        "#\n",
        "# Unless required by applicable law or agreed to in writing, software\n",
        "# distributed under the License is distributed on an \"AS IS\" BASIS,\n",
        "# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.\n",
        "# See the License for the specific language governing permissions and\n",
        "# limitations under the License."
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "view-in-github"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/lmoroney/dlaicourse/blob/master/TensorFlow%20In%20Practice/Course%204%20-%20S%2BP/S%2BP%20Week%204%20Lesson%201.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "cellView": "both",
        "colab_type": "code",
        "id": "D1J15Vh_1Jih",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "176e60da-1a33-4101-b4e1-3020809b8dc7"
      },
      "source": [
        "!pip install tensorflow==2.0.0\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Collecting tensorflow==2.0.0\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/46/0f/7bd55361168bb32796b360ad15a25de6966c9c1beb58a8e30c01c8279862/tensorflow-2.0.0-cp36-cp36m-manylinux2010_x86_64.whl (86.3MB)\n",
            "\u001b[K     |████████████████████████████████| 86.3MB 95kB/s \n",
            "\u001b[?25hCollecting tensorboard<2.1.0,>=2.0.0\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/76/54/99b9d5d52d5cb732f099baaaf7740403e83fe6b0cedde940fabd2b13d75a/tensorboard-2.0.2-py3-none-any.whl (3.8MB)\n",
            "\u001b[K     |████████████████████████████████| 3.8MB 42.0MB/s \n",
            "\u001b[?25hCollecting gast==0.2.2\n",
            "  Downloading https://files.pythonhosted.org/packages/4e/35/11749bf99b2d4e3cceb4d55ca22590b0d7c2c62b9de38ac4a4a7f4687421/gast-0.2.2.tar.gz\n",
            "Requirement already satisfied: absl-py>=0.7.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0.0) (0.8.1)\n",
            "Requirement already satisfied: protobuf>=3.6.1 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0.0) (3.12.4)\n",
            "Requirement already satisfied: grpcio>=1.8.6 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0.0) (1.31.0)\n",
            "Requirement already satisfied: wheel>=0.26 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0.0) (0.35.1)\n",
            "Collecting tensorflow-estimator<2.1.0,>=2.0.0\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/fc/08/8b927337b7019c374719145d1dceba21a8bb909b93b1ad6f8fb7d22c1ca1/tensorflow_estimator-2.0.1-py2.py3-none-any.whl (449kB)\n",
            "\u001b[K     |████████████████████████████████| 450kB 38.6MB/s \n",
            "\u001b[?25hRequirement already satisfied: google-pasta>=0.1.6 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0.0) (0.2.0)\n",
            "Requirement already satisfied: six>=1.10.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0.0) (1.15.0)\n",
            "Collecting keras-applications>=1.0.8\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/71/e3/19762fdfc62877ae9102edf6342d71b28fbfd9dea3d2f96a882ce099b03f/Keras_Applications-1.0.8-py3-none-any.whl (50kB)\n",
            "\u001b[K     |████████████████████████████████| 51kB 6.3MB/s \n",
            "\u001b[?25hRequirement already satisfied: keras-preprocessing>=1.0.5 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0.0) (1.1.2)\n",
            "Requirement already satisfied: numpy<2.0,>=1.16.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0.0) (1.18.5)\n",
            "Requirement already satisfied: opt-einsum>=2.3.2 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0.0) (3.3.0)\n",
            "Requirement already satisfied: wrapt>=1.11.1 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0.0) (1.12.1)\n",
            "Requirement already satisfied: termcolor>=1.1.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0.0) (1.1.0)\n",
            "Requirement already satisfied: astor>=0.6.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0.0) (0.8.1)\n",
            "Requirement already satisfied: requests<3,>=2.21.0 in /usr/local/lib/python3.6/dist-packages (from tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.0) (2.23.0)\n",
            "Requirement already satisfied: setuptools>=41.0.0 in /usr/local/lib/python3.6/dist-packages (from tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.0) (49.6.0)\n",
            "Requirement already satisfied: markdown>=2.6.8 in /usr/local/lib/python3.6/dist-packages (from tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.0) (3.2.2)\n",
            "Requirement already satisfied: google-auth-oauthlib<0.5,>=0.4.1 in /usr/local/lib/python3.6/dist-packages (from tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.0) (0.4.1)\n",
            "Requirement already satisfied: google-auth<2,>=1.6.3 in /usr/local/lib/python3.6/dist-packages (from tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.0) (1.17.2)\n",
            "Requirement already satisfied: werkzeug>=0.11.15 in /usr/local/lib/python3.6/dist-packages (from tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.0) (1.0.1)\n",
            "Requirement already satisfied: h5py in /usr/local/lib/python3.6/dist-packages (from keras-applications>=1.0.8->tensorflow==2.0.0) (2.10.0)\n",
            "Requirement already satisfied: chardet<4,>=3.0.2 in /usr/local/lib/python3.6/dist-packages (from requests<3,>=2.21.0->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.0) (3.0.4)\n",
            "Requirement already satisfied: idna<3,>=2.5 in /usr/local/lib/python3.6/dist-packages (from requests<3,>=2.21.0->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.0) (2.10)\n",
            "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.6/dist-packages (from requests<3,>=2.21.0->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.0) (2020.6.20)\n",
            "Requirement already satisfied: urllib3!=1.25.0,!=1.25.1,<1.26,>=1.21.1 in /usr/local/lib/python3.6/dist-packages (from requests<3,>=2.21.0->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.0) (1.24.3)\n",
            "Requirement already satisfied: importlib-metadata; python_version < \"3.8\" in /usr/local/lib/python3.6/dist-packages (from markdown>=2.6.8->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.0) (1.7.0)\n",
            "Requirement already satisfied: requests-oauthlib>=0.7.0 in /usr/local/lib/python3.6/dist-packages (from google-auth-oauthlib<0.5,>=0.4.1->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.0) (1.3.0)\n",
            "Requirement already satisfied: rsa<5,>=3.1.4; python_version >= \"3\" in /usr/local/lib/python3.6/dist-packages (from google-auth<2,>=1.6.3->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.0) (4.6)\n",
            "Requirement already satisfied: cachetools<5.0,>=2.0.0 in /usr/local/lib/python3.6/dist-packages (from google-auth<2,>=1.6.3->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.0) (4.1.1)\n",
            "Requirement already satisfied: pyasn1-modules>=0.2.1 in /usr/local/lib/python3.6/dist-packages (from google-auth<2,>=1.6.3->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.0) (0.2.8)\n",
            "Requirement already satisfied: zipp>=0.5 in /usr/local/lib/python3.6/dist-packages (from importlib-metadata; python_version < \"3.8\"->markdown>=2.6.8->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.0) (3.1.0)\n",
            "Requirement already satisfied: oauthlib>=3.0.0 in /usr/local/lib/python3.6/dist-packages (from requests-oauthlib>=0.7.0->google-auth-oauthlib<0.5,>=0.4.1->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.0) (3.1.0)\n",
            "Requirement already satisfied: pyasn1>=0.1.3 in /usr/local/lib/python3.6/dist-packages (from rsa<5,>=3.1.4; python_version >= \"3\"->google-auth<2,>=1.6.3->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.0) (0.4.8)\n",
            "Building wheels for collected packages: gast\n",
            "  Building wheel for gast (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for gast: filename=gast-0.2.2-cp36-none-any.whl size=7542 sha256=eacc60859e13bd29691d2ad434a277c6bf90e204aaabfc615e93bb8942e9e6b7\n",
            "  Stored in directory: /root/.cache/pip/wheels/5c/2e/7e/a1d4d4fcebe6c381f378ce7743a3ced3699feb89bcfbdadadd\n",
            "Successfully built gast\n",
            "\u001b[31mERROR: tensorflow-probability 0.11.0 has requirement gast>=0.3.2, but you'll have gast 0.2.2 which is incompatible.\u001b[0m\n",
            "Installing collected packages: tensorboard, gast, tensorflow-estimator, keras-applications, tensorflow\n",
            "  Found existing installation: tensorboard 2.3.0\n",
            "    Uninstalling tensorboard-2.3.0:\n",
            "      Successfully uninstalled tensorboard-2.3.0\n",
            "  Found existing installation: gast 0.3.3\n",
            "    Uninstalling gast-0.3.3:\n",
            "      Successfully uninstalled gast-0.3.3\n",
            "  Found existing installation: tensorflow-estimator 2.3.0\n",
            "    Uninstalling tensorflow-estimator-2.3.0:\n",
            "      Successfully uninstalled tensorflow-estimator-2.3.0\n",
            "  Found existing installation: tensorflow 2.3.0\n",
            "    Uninstalling tensorflow-2.3.0:\n",
            "      Successfully uninstalled tensorflow-2.3.0\n",
            "Successfully installed gast-0.2.2 keras-applications-1.0.8 tensorboard-2.0.2 tensorflow-2.0.0 tensorflow-estimator-2.0.1\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "BOjujz601HcS",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "a04f1be9-652b-446f-d542-954fbb788a04"
      },
      "source": [
        "import tensorflow as tf\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "print(tf.__version__)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "2.0.0\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "Zswl7jRtGzkk",
        "colab": {}
      },
      "source": [
        "def plot_series(time, series, format=\"-\", start=0, end=None):\n",
        "    plt.plot(time[start:end], series[start:end], format)\n",
        "    plt.xlabel(\"Time\")\n",
        "    plt.ylabel(\"Value\")\n",
        "    plt.grid(True)\n",
        "\n",
        "def trend(time, slope=0):\n",
        "    return slope * time\n",
        "\n",
        "def seasonal_pattern(season_time):\n",
        "    \"\"\"Just an arbitrary pattern, you can change it if you wish\"\"\"\n",
        "    return np.where(season_time < 0.4,\n",
        "                    np.cos(season_time * 2 * np.pi),\n",
        "                    1 / np.exp(3 * season_time))\n",
        "\n",
        "def seasonality(time, period, amplitude=1, phase=0):\n",
        "    \"\"\"Repeats the same pattern at each period\"\"\"\n",
        "    season_time = ((time + phase) % period) / period\n",
        "    return amplitude * seasonal_pattern(season_time)\n",
        "\n",
        "def noise(time, noise_level=1, seed=None):\n",
        "    rnd = np.random.RandomState(seed)\n",
        "    return rnd.randn(len(time)) * noise_level\n",
        "\n",
        "time = np.arange(4 * 365 + 1, dtype=\"float32\")\n",
        "baseline = 10\n",
        "series = trend(time, 0.1)  \n",
        "baseline = 10\n",
        "amplitude = 40\n",
        "slope = 0.05\n",
        "noise_level = 5\n",
        "\n",
        "# Create the series\n",
        "series = baseline + trend(time, slope) + seasonality(time, period=365, amplitude=amplitude)\n",
        "# Update with noise\n",
        "series += noise(time, noise_level, seed=42)\n",
        "\n",
        "split_time = 1000\n",
        "time_train = time[:split_time]\n",
        "x_train = series[:split_time]\n",
        "time_valid = time[split_time:]\n",
        "x_valid = series[split_time:]\n",
        "\n",
        "window_size = 20\n",
        "batch_size = 32\n",
        "shuffle_buffer_size = 1000"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "4sTTIOCbyShY",
        "colab": {}
      },
      "source": [
        "def windowed_dataset(series, window_size, batch_size, shuffle_buffer):\n",
        "    series = tf.expand_dims(series, axis=-1)\n",
        "    ds = tf.data.Dataset.from_tensor_slices(series)\n",
        "    ds = ds.window(window_size + 1, shift=1, drop_remainder=True)\n",
        "    ds = ds.flat_map(lambda w: w.batch(window_size + 1))\n",
        "    ds = ds.shuffle(shuffle_buffer)\n",
        "    ds = ds.map(lambda w: (w[:-1], w[1:]))\n",
        "    return ds.batch(batch_size).prefetch(1)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "_eaAX9g_jS5W",
        "colab": {}
      },
      "source": [
        "def model_forecast(model, series, window_size):\n",
        "    ds = tf.data.Dataset.from_tensor_slices(series)\n",
        "    ds = ds.window(window_size, shift=1, drop_remainder=True)\n",
        "    ds = ds.flat_map(lambda w: w.batch(window_size))\n",
        "    ds = ds.batch(32).prefetch(1)\n",
        "    forecast = model.predict(ds)\n",
        "    return forecast"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "Yqc2GTsps0qf",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "6a9d3c4b-e392-4779-b334-e5d1c5ce6abe"
      },
      "source": [
        "tf.keras.backend.clear_session()\n",
        "tf.random.set_seed(51)\n",
        "np.random.seed(51)\n",
        "\n",
        "window_size = 30\n",
        "train_set = windowed_dataset(x_train, window_size, batch_size=128, shuffle_buffer=shuffle_buffer_size)\n",
        "\n",
        "model = tf.keras.models.Sequential([\n",
        "  tf.keras.layers.Conv1D(filters=32, kernel_size=5,\n",
        "                      strides=1, padding=\"causal\",\n",
        "                      activation=\"relu\",\n",
        "                      input_shape=[None, 1]),\n",
        "  tf.keras.layers.Bidirectional(tf.keras.layers.LSTM(32, return_sequences=True)),\n",
        "  tf.keras.layers.Bidirectional(tf.keras.layers.LSTM(32, return_sequences=True)),\n",
        "  tf.keras.layers.Dense(1),\n",
        "  tf.keras.layers.Lambda(lambda x: x * 200)\n",
        "])\n",
        "lr_schedule = tf.keras.callbacks.LearningRateScheduler(\n",
        "    lambda epoch: 1e-8 * 10**(epoch / 20))\n",
        "optimizer = tf.keras.optimizers.SGD(lr=1e-8, momentum=0.9)\n",
        "model.compile(loss=tf.keras.losses.Huber(),\n",
        "              optimizer=optimizer,\n",
        "              metrics=[\"mae\"])\n",
        "history = model.fit(train_set, epochs=100, callbacks=[lr_schedule])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/100\n",
            "8/8 [==============================] - 7s 917ms/step - loss: 73.2800 - mae: 73.6896\n",
            "Epoch 2/100\n",
            "8/8 [==============================] - 1s 110ms/step - loss: 72.6322 - mae: 72.9788\n",
            "Epoch 3/100\n",
            "8/8 [==============================] - 1s 107ms/step - loss: 71.4605 - mae: 71.8377\n",
            "Epoch 4/100\n",
            "8/8 [==============================] - 1s 104ms/step - loss: 70.0359 - mae: 70.4184\n",
            "Epoch 5/100\n",
            "8/8 [==============================] - 1s 103ms/step - loss: 68.2689 - mae: 68.7583\n",
            "Epoch 6/100\n",
            "8/8 [==============================] - 1s 103ms/step - loss: 66.4883 - mae: 66.8782\n",
            "Epoch 7/100\n",
            "8/8 [==============================] - 1s 104ms/step - loss: 64.2616 - mae: 64.7354\n",
            "Epoch 8/100\n",
            "8/8 [==============================] - 1s 103ms/step - loss: 62.0457 - mae: 62.3713\n",
            "Epoch 9/100\n",
            "8/8 [==============================] - 1s 104ms/step - loss: 59.3746 - mae: 59.7458\n",
            "Epoch 10/100\n",
            "8/8 [==============================] - 1s 103ms/step - loss: 56.3313 - mae: 56.8317\n",
            "Epoch 11/100\n",
            "8/8 [==============================] - 1s 104ms/step - loss: 53.3244 - mae: 53.5844\n",
            "Epoch 12/100\n",
            "8/8 [==============================] - 1s 104ms/step - loss: 49.4466 - mae: 49.9128\n",
            "Epoch 13/100\n",
            "8/8 [==============================] - 1s 104ms/step - loss: 45.2945 - mae: 45.5964\n",
            "Epoch 14/100\n",
            "8/8 [==============================] - 1s 103ms/step - loss: 41.5921 - mae: 41.9902\n",
            "Epoch 15/100\n",
            "8/8 [==============================] - 1s 104ms/step - loss: 40.6574 - mae: 41.1555\n",
            "Epoch 16/100\n",
            "8/8 [==============================] - 1s 111ms/step - loss: 39.5009 - mae: 40.0524\n",
            "Epoch 17/100\n",
            "8/8 [==============================] - 1s 104ms/step - loss: 38.1321 - mae: 38.6531\n",
            "Epoch 18/100\n",
            "8/8 [==============================] - 1s 106ms/step - loss: 36.6134 - mae: 37.1057\n",
            "Epoch 19/100\n",
            "8/8 [==============================] - 1s 105ms/step - loss: 35.0407 - mae: 35.5333\n",
            "Epoch 20/100\n",
            "8/8 [==============================] - 1s 105ms/step - loss: 33.6130 - mae: 34.0769\n",
            "Epoch 21/100\n",
            "8/8 [==============================] - 1s 109ms/step - loss: 32.2603 - mae: 32.7742\n",
            "Epoch 22/100\n",
            "8/8 [==============================] - 1s 106ms/step - loss: 31.0060 - mae: 31.5719\n",
            "Epoch 23/100\n",
            "8/8 [==============================] - 1s 105ms/step - loss: 29.9019 - mae: 30.4469\n",
            "Epoch 24/100\n",
            "8/8 [==============================] - 1s 110ms/step - loss: 28.9141 - mae: 29.3412\n",
            "Epoch 25/100\n",
            "8/8 [==============================] - 1s 104ms/step - loss: 27.8442 - mae: 28.2367\n",
            "Epoch 26/100\n",
            "8/8 [==============================] - 1s 106ms/step - loss: 26.6497 - mae: 27.1097\n",
            "Epoch 27/100\n",
            "8/8 [==============================] - 1s 103ms/step - loss: 25.4563 - mae: 25.9551\n",
            "Epoch 28/100\n",
            "8/8 [==============================] - 1s 109ms/step - loss: 24.2907 - mae: 24.7682\n",
            "Epoch 29/100\n",
            "8/8 [==============================] - 1s 104ms/step - loss: 23.0599 - mae: 23.5445\n",
            "Epoch 30/100\n",
            "8/8 [==============================] - 1s 102ms/step - loss: 21.6999 - mae: 22.2905\n",
            "Epoch 31/100\n",
            "8/8 [==============================] - 1s 103ms/step - loss: 20.6535 - mae: 21.0224\n",
            "Epoch 32/100\n",
            "8/8 [==============================] - 1s 104ms/step - loss: 19.3034 - mae: 19.7113\n",
            "Epoch 33/100\n",
            "8/8 [==============================] - 1s 104ms/step - loss: 17.8995 - mae: 18.4525\n",
            "Epoch 34/100\n",
            "8/8 [==============================] - 1s 102ms/step - loss: 16.9921 - mae: 17.4434\n",
            "Epoch 35/100\n",
            "8/8 [==============================] - 1s 106ms/step - loss: 16.2409 - mae: 16.7384\n",
            "Epoch 36/100\n",
            "8/8 [==============================] - 1s 104ms/step - loss: 15.6820 - mae: 16.1243\n",
            "Epoch 37/100\n",
            "8/8 [==============================] - 1s 110ms/step - loss: 15.0608 - mae: 15.5439\n",
            "Epoch 38/100\n",
            "8/8 [==============================] - 1s 104ms/step - loss: 14.4595 - mae: 15.0026\n",
            "Epoch 39/100\n",
            "8/8 [==============================] - 1s 103ms/step - loss: 14.0791 - mae: 14.4808\n",
            "Epoch 40/100\n",
            "8/8 [==============================] - 1s 104ms/step - loss: 13.4649 - mae: 13.9302\n",
            "Epoch 41/100\n",
            "8/8 [==============================] - 1s 105ms/step - loss: 12.9145 - mae: 13.3710\n",
            "Epoch 42/100\n",
            "8/8 [==============================] - 1s 106ms/step - loss: 12.3464 - mae: 12.7804\n",
            "Epoch 43/100\n",
            "8/8 [==============================] - 1s 104ms/step - loss: 11.7356 - mae: 12.1553\n",
            "Epoch 44/100\n",
            "8/8 [==============================] - 1s 105ms/step - loss: 11.0543 - mae: 11.5075\n",
            "Epoch 45/100\n",
            "8/8 [==============================] - 1s 103ms/step - loss: 10.3970 - mae: 10.8390\n",
            "Epoch 46/100\n",
            "8/8 [==============================] - 1s 103ms/step - loss: 9.6753 - mae: 10.1738\n",
            "Epoch 47/100\n",
            "8/8 [==============================] - 1s 103ms/step - loss: 9.0906 - mae: 9.5766\n",
            "Epoch 48/100\n",
            "8/8 [==============================] - 1s 102ms/step - loss: 8.6122 - mae: 9.0605\n",
            "Epoch 49/100\n",
            "8/8 [==============================] - 1s 109ms/step - loss: 8.1542 - mae: 8.6017\n",
            "Epoch 50/100\n",
            "8/8 [==============================] - 1s 105ms/step - loss: 7.7260 - mae: 8.1924\n",
            "Epoch 51/100\n",
            "8/8 [==============================] - 1s 102ms/step - loss: 7.3652 - mae: 7.8068\n",
            "Epoch 52/100\n",
            "8/8 [==============================] - 1s 105ms/step - loss: 6.9616 - mae: 7.4467\n",
            "Epoch 53/100\n",
            "8/8 [==============================] - 1s 107ms/step - loss: 6.6863 - mae: 7.1167\n",
            "Epoch 54/100\n",
            "8/8 [==============================] - 1s 108ms/step - loss: 6.3220 - mae: 6.7790\n",
            "Epoch 55/100\n",
            "8/8 [==============================] - 1s 104ms/step - loss: 5.9906 - mae: 6.4550\n",
            "Epoch 56/100\n",
            "8/8 [==============================] - 1s 108ms/step - loss: 5.7089 - mae: 6.1556\n",
            "Epoch 57/100\n",
            "8/8 [==============================] - 1s 104ms/step - loss: 5.4460 - mae: 5.9035\n",
            "Epoch 58/100\n",
            "8/8 [==============================] - 1s 104ms/step - loss: 5.2157 - mae: 5.6731\n",
            "Epoch 59/100\n",
            "8/8 [==============================] - 1s 104ms/step - loss: 4.9922 - mae: 5.4522\n",
            "Epoch 60/100\n",
            "8/8 [==============================] - 1s 105ms/step - loss: 4.8104 - mae: 5.3024\n",
            "Epoch 61/100\n",
            "8/8 [==============================] - 1s 104ms/step - loss: 4.7175 - mae: 5.2023\n",
            "Epoch 62/100\n",
            "8/8 [==============================] - 1s 105ms/step - loss: 4.9295 - mae: 5.3703\n",
            "Epoch 63/100\n",
            "8/8 [==============================] - 1s 103ms/step - loss: 4.5738 - mae: 5.0650\n",
            "Epoch 64/100\n",
            "8/8 [==============================] - 1s 103ms/step - loss: 4.7220 - mae: 5.1538\n",
            "Epoch 65/100\n",
            "8/8 [==============================] - 1s 105ms/step - loss: 4.7053 - mae: 5.3071\n",
            "Epoch 66/100\n",
            "8/8 [==============================] - 1s 105ms/step - loss: 5.2452 - mae: 5.8564\n",
            "Epoch 67/100\n",
            "8/8 [==============================] - 1s 105ms/step - loss: 6.3944 - mae: 7.0185\n",
            "Epoch 68/100\n",
            "8/8 [==============================] - 1s 103ms/step - loss: 5.8145 - mae: 6.2172\n",
            "Epoch 69/100\n",
            "8/8 [==============================] - 1s 104ms/step - loss: 5.7929 - mae: 6.2083\n",
            "Epoch 70/100\n",
            "8/8 [==============================] - 1s 105ms/step - loss: 6.1573 - mae: 6.6167\n",
            "Epoch 71/100\n",
            "8/8 [==============================] - 1s 107ms/step - loss: 5.8503 - mae: 6.3295\n",
            "Epoch 72/100\n",
            "8/8 [==============================] - 1s 109ms/step - loss: 6.4572 - mae: 7.0011\n",
            "Epoch 73/100\n",
            "8/8 [==============================] - 1s 107ms/step - loss: 7.6252 - mae: 7.8872\n",
            "Epoch 74/100\n",
            "8/8 [==============================] - 1s 102ms/step - loss: 15.5506 - mae: 15.6197\n",
            "Epoch 75/100\n",
            "8/8 [==============================] - 1s 102ms/step - loss: 14.8549 - mae: 14.8218\n",
            "Epoch 76/100\n",
            "8/8 [==============================] - 1s 108ms/step - loss: 15.6386 - mae: 15.4366\n",
            "Epoch 77/100\n",
            "8/8 [==============================] - 1s 102ms/step - loss: 17.5484 - mae: 18.4639\n",
            "Epoch 78/100\n",
            "8/8 [==============================] - 1s 104ms/step - loss: 17.5276 - mae: 18.2544\n",
            "Epoch 79/100\n",
            "8/8 [==============================] - 1s 104ms/step - loss: 17.7635 - mae: 18.0921\n",
            "Epoch 80/100\n",
            "8/8 [==============================] - 1s 106ms/step - loss: 9.7347 - mae: 10.0811\n",
            "Epoch 81/100\n",
            "8/8 [==============================] - 1s 109ms/step - loss: 9.2012 - mae: 10.1283\n",
            "Epoch 82/100\n",
            "8/8 [==============================] - 1s 102ms/step - loss: 9.4738 - mae: 9.6584\n",
            "Epoch 83/100\n",
            "8/8 [==============================] - 1s 103ms/step - loss: 10.7859 - mae: 12.0523\n",
            "Epoch 84/100\n",
            "8/8 [==============================] - 1s 103ms/step - loss: 25.7430 - mae: 26.5311\n",
            "Epoch 85/100\n",
            "8/8 [==============================] - 1s 105ms/step - loss: 19.7741 - mae: 19.4900\n",
            "Epoch 86/100\n",
            "8/8 [==============================] - 1s 102ms/step - loss: 11.0158 - mae: 11.8060\n",
            "Epoch 87/100\n",
            "8/8 [==============================] - 1s 101ms/step - loss: 10.9255 - mae: 11.6367\n",
            "Epoch 88/100\n",
            "8/8 [==============================] - 1s 111ms/step - loss: 11.3700 - mae: 12.0015\n",
            "Epoch 89/100\n",
            "8/8 [==============================] - 1s 105ms/step - loss: 11.7735 - mae: 13.4100\n",
            "Epoch 90/100\n",
            "8/8 [==============================] - 1s 106ms/step - loss: 28.7528 - mae: 29.0607\n",
            "Epoch 91/100\n",
            "8/8 [==============================] - 1s 106ms/step - loss: 15.4962 - mae: 15.5951\n",
            "Epoch 92/100\n",
            "8/8 [==============================] - 1s 103ms/step - loss: 22.9931 - mae: 22.6299\n",
            "Epoch 93/100\n",
            "8/8 [==============================] - 1s 106ms/step - loss: 26.1823 - mae: 25.7465\n",
            "Epoch 94/100\n",
            "8/8 [==============================] - 1s 104ms/step - loss: 17.5895 - mae: 17.6440\n",
            "Epoch 95/100\n",
            "8/8 [==============================] - 1s 106ms/step - loss: 14.2346 - mae: 14.4800\n",
            "Epoch 96/100\n",
            "8/8 [==============================] - 1s 111ms/step - loss: 20.0196 - mae: 20.1561\n",
            "Epoch 97/100\n",
            "8/8 [==============================] - 1s 105ms/step - loss: 14.4974 - mae: 14.6962\n",
            "Epoch 98/100\n",
            "8/8 [==============================] - 1s 103ms/step - loss: 19.3200 - mae: 19.7313\n",
            "Epoch 99/100\n",
            "8/8 [==============================] - 1s 103ms/step - loss: 17.3532 - mae: 18.1741\n",
            "Epoch 100/100\n",
            "8/8 [==============================] - 1s 108ms/step - loss: 18.6312 - mae: 19.4826\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "MjTvASUns0qh",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 290
        },
        "outputId": "d46854ab-1e6f-45f9-a12b-76582cb315da"
      },
      "source": [
        "plt.semilogx(history.history[\"lr\"], history.history[\"loss\"])\n",
        "plt.axis([1e-8, 1e-4, 0, 30])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(1e-08, 0.0001, 0.0, 30.0)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 9
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "4uh-97bpLZCA",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "cff7c825-1b7a-45ae-fc7f-43ec7094c2fe"
      },
      "source": [
        "tf.keras.backend.clear_session()\n",
        "tf.random.set_seed(51)\n",
        "np.random.seed(51)\n",
        "# batch_size = 16\n",
        "dataset = windowed_dataset(x_train, window_size, batch_size, shuffle_buffer_size)\n",
        "\n",
        "model = tf.keras.models.Sequential([\n",
        "  tf.keras.layers.Conv1D(filters=32, kernel_size=3,\n",
        "                      strides=1, padding=\"causal\",\n",
        "                      activation=\"relu\",\n",
        "                      input_shape=[None, 1]),\n",
        "  tf.keras.layers.LSTM(32, return_sequences=True),\n",
        "  tf.keras.layers.LSTM(32, return_sequences=True),\n",
        "  tf.keras.layers.Dense(1),\n",
        "  tf.keras.layers.Lambda(lambda x: x * 200)\n",
        "])\n",
        "\n",
        "optimizer = tf.keras.optimizers.SGD(lr=1e-5, momentum=0.9)\n",
        "model.compile(loss=tf.keras.losses.Huber(),\n",
        "              optimizer=optimizer,\n",
        "              metrics=[\"mae\"])\n",
        "history = model.fit(dataset,epochs=500)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/500\n",
            "61/61 [==============================] - 6s 95ms/step - loss: 14.3561 - mae: 14.8810\n",
            "Epoch 2/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 6.7435 - mae: 7.2288\n",
            "Epoch 3/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 5.8958 - mae: 6.3724\n",
            "Epoch 4/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 5.4559 - mae: 5.9363\n",
            "Epoch 5/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 5.3584 - mae: 5.8322\n",
            "Epoch 6/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 5.1366 - mae: 5.6142\n",
            "Epoch 7/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.9855 - mae: 5.4666\n",
            "Epoch 8/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.9234 - mae: 5.4021\n",
            "Epoch 9/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.8415 - mae: 5.3153\n",
            "Epoch 10/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.8826 - mae: 5.3576\n",
            "Epoch 11/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.9908 - mae: 5.4644\n",
            "Epoch 12/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.6935 - mae: 5.1685\n",
            "Epoch 13/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.6833 - mae: 5.1609\n",
            "Epoch 14/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.6865 - mae: 5.1716\n",
            "Epoch 15/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.6954 - mae: 5.1758\n",
            "Epoch 16/500\n",
            "61/61 [==============================] - 2s 32ms/step - loss: 4.7161 - mae: 5.1930\n",
            "Epoch 17/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.6769 - mae: 5.1562\n",
            "Epoch 18/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.5281 - mae: 5.0043\n",
            "Epoch 19/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.5824 - mae: 5.0662\n",
            "Epoch 20/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.6598 - mae: 5.1339\n",
            "Epoch 21/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.5390 - mae: 5.0210\n",
            "Epoch 22/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.5337 - mae: 5.0102\n",
            "Epoch 23/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.4309 - mae: 4.9089\n",
            "Epoch 24/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.4680 - mae: 4.9403\n",
            "Epoch 25/500\n",
            "61/61 [==============================] - 2s 31ms/step - loss: 4.5240 - mae: 5.0004\n",
            "Epoch 26/500\n",
            "61/61 [==============================] - 2s 32ms/step - loss: 4.4126 - mae: 4.8826\n",
            "Epoch 27/500\n",
            "61/61 [==============================] - 2s 31ms/step - loss: 4.5570 - mae: 5.0350\n",
            "Epoch 28/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.3756 - mae: 4.8560\n",
            "Epoch 29/500\n",
            "61/61 [==============================] - 2s 31ms/step - loss: 4.3422 - mae: 4.8229\n",
            "Epoch 30/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.4253 - mae: 4.9046\n",
            "Epoch 31/500\n",
            "61/61 [==============================] - 2s 31ms/step - loss: 4.6133 - mae: 5.0905\n",
            "Epoch 32/500\n",
            "61/61 [==============================] - 2s 31ms/step - loss: 4.3991 - mae: 4.8723\n",
            "Epoch 33/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.3710 - mae: 4.8485\n",
            "Epoch 34/500\n",
            "61/61 [==============================] - 2s 32ms/step - loss: 4.3745 - mae: 4.8596\n",
            "Epoch 35/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.3688 - mae: 4.8432\n",
            "Epoch 36/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.3614 - mae: 4.8379\n",
            "Epoch 37/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.3043 - mae: 4.7800\n",
            "Epoch 38/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.3351 - mae: 4.8196\n",
            "Epoch 39/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.3412 - mae: 4.8155\n",
            "Epoch 40/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.3244 - mae: 4.7998\n",
            "Epoch 41/500\n",
            "61/61 [==============================] - 2s 28ms/step - loss: 4.3962 - mae: 4.8741\n",
            "Epoch 42/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.4198 - mae: 4.8899\n",
            "Epoch 43/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.2795 - mae: 4.7519\n",
            "Epoch 44/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.2623 - mae: 4.7399\n",
            "Epoch 45/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.2863 - mae: 4.7639\n",
            "Epoch 46/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.2783 - mae: 4.7510\n",
            "Epoch 47/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.2994 - mae: 4.7721\n",
            "Epoch 48/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.2807 - mae: 4.7600\n",
            "Epoch 49/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.3284 - mae: 4.8020\n",
            "Epoch 50/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.2712 - mae: 4.7511\n",
            "Epoch 51/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.2292 - mae: 4.7033\n",
            "Epoch 52/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.2101 - mae: 4.6884\n",
            "Epoch 53/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.2183 - mae: 4.6881\n",
            "Epoch 54/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.2312 - mae: 4.7082\n",
            "Epoch 55/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.3697 - mae: 4.8366\n",
            "Epoch 56/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.2860 - mae: 4.7607\n",
            "Epoch 57/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.1911 - mae: 4.6638\n",
            "Epoch 58/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.1807 - mae: 4.6508\n",
            "Epoch 59/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.2479 - mae: 4.7331\n",
            "Epoch 60/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.5068 - mae: 4.9786\n",
            "Epoch 61/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.2578 - mae: 4.7320\n",
            "Epoch 62/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.2252 - mae: 4.6966\n",
            "Epoch 63/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.2083 - mae: 4.6859\n",
            "Epoch 64/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.2712 - mae: 4.7513\n",
            "Epoch 65/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.1786 - mae: 4.6539\n",
            "Epoch 66/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.1911 - mae: 4.6734\n",
            "Epoch 67/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.3655 - mae: 4.8374\n",
            "Epoch 68/500\n",
            "61/61 [==============================] - 2s 28ms/step - loss: 4.2037 - mae: 4.6777\n",
            "Epoch 69/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.2213 - mae: 4.6916\n",
            "Epoch 70/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.1888 - mae: 4.6608\n",
            "Epoch 71/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.2453 - mae: 4.7168\n",
            "Epoch 72/500\n",
            "61/61 [==============================] - 2s 28ms/step - loss: 4.2804 - mae: 4.7591\n",
            "Epoch 73/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.2418 - mae: 4.7168\n",
            "Epoch 74/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.1990 - mae: 4.6702\n",
            "Epoch 75/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.1793 - mae: 4.6517\n",
            "Epoch 76/500\n",
            "61/61 [==============================] - 2s 28ms/step - loss: 4.3465 - mae: 4.8213\n",
            "Epoch 77/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.1883 - mae: 4.6612\n",
            "Epoch 78/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.1738 - mae: 4.6464\n",
            "Epoch 79/500\n",
            "61/61 [==============================] - 2s 31ms/step - loss: 4.2149 - mae: 4.6896\n",
            "Epoch 80/500\n",
            "61/61 [==============================] - 2s 32ms/step - loss: 4.1670 - mae: 4.6487\n",
            "Epoch 81/500\n",
            "61/61 [==============================] - 2s 31ms/step - loss: 4.1696 - mae: 4.6433\n",
            "Epoch 82/500\n",
            "61/61 [==============================] - 2s 31ms/step - loss: 4.2155 - mae: 4.6899\n",
            "Epoch 83/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.1552 - mae: 4.6276\n",
            "Epoch 84/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.1436 - mae: 4.6217\n",
            "Epoch 85/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.1972 - mae: 4.6700\n",
            "Epoch 86/500\n",
            "61/61 [==============================] - 2s 28ms/step - loss: 4.2397 - mae: 4.7148\n",
            "Epoch 87/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.1811 - mae: 4.6538\n",
            "Epoch 88/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.1816 - mae: 4.6567\n",
            "Epoch 89/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.1119 - mae: 4.5802\n",
            "Epoch 90/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.3039 - mae: 4.7793\n",
            "Epoch 91/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.1579 - mae: 4.6301\n",
            "Epoch 92/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.1257 - mae: 4.5966\n",
            "Epoch 93/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.1426 - mae: 4.6184\n",
            "Epoch 94/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.1367 - mae: 4.6124\n",
            "Epoch 95/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.1394 - mae: 4.6145\n",
            "Epoch 96/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.1763 - mae: 4.6508\n",
            "Epoch 97/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0899 - mae: 4.5681\n",
            "Epoch 98/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.1140 - mae: 4.5880\n",
            "Epoch 99/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.2150 - mae: 4.6918\n",
            "Epoch 100/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.1460 - mae: 4.6159\n",
            "Epoch 101/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.1671 - mae: 4.6380\n",
            "Epoch 102/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.1451 - mae: 4.6181\n",
            "Epoch 103/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.1427 - mae: 4.6137\n",
            "Epoch 104/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.1142 - mae: 4.5908\n",
            "Epoch 105/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.1535 - mae: 4.6255\n",
            "Epoch 106/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0814 - mae: 4.5566\n",
            "Epoch 107/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0768 - mae: 4.5560\n",
            "Epoch 108/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0739 - mae: 4.5500\n",
            "Epoch 109/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.1365 - mae: 4.6111\n",
            "Epoch 110/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.1288 - mae: 4.6048\n",
            "Epoch 111/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.1327 - mae: 4.6053\n",
            "Epoch 112/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.2866 - mae: 4.7621\n",
            "Epoch 113/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0869 - mae: 4.5610\n",
            "Epoch 114/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0621 - mae: 4.5362\n",
            "Epoch 115/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0603 - mae: 4.5403\n",
            "Epoch 116/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0730 - mae: 4.5459\n",
            "Epoch 117/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0925 - mae: 4.5672\n",
            "Epoch 118/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0527 - mae: 4.5213\n",
            "Epoch 119/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0910 - mae: 4.5646\n",
            "Epoch 120/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.1406 - mae: 4.6155\n",
            "Epoch 121/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.1028 - mae: 4.5748\n",
            "Epoch 122/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0382 - mae: 4.5126\n",
            "Epoch 123/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0530 - mae: 4.5252\n",
            "Epoch 124/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0639 - mae: 4.5368\n",
            "Epoch 125/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.2128 - mae: 4.6911\n",
            "Epoch 126/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0268 - mae: 4.5000\n",
            "Epoch 127/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.1085 - mae: 4.5798\n",
            "Epoch 128/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.1250 - mae: 4.6026\n",
            "Epoch 129/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0565 - mae: 4.5286\n",
            "Epoch 130/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0519 - mae: 4.5230\n",
            "Epoch 131/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.1111 - mae: 4.5859\n",
            "Epoch 132/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.1857 - mae: 4.6611\n",
            "Epoch 133/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0718 - mae: 4.5426\n",
            "Epoch 134/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.1107 - mae: 4.5892\n",
            "Epoch 135/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.1687 - mae: 4.6391\n",
            "Epoch 136/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0273 - mae: 4.4983\n",
            "Epoch 137/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0517 - mae: 4.5269\n",
            "Epoch 138/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.2101 - mae: 4.6872\n",
            "Epoch 139/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0221 - mae: 4.4975\n",
            "Epoch 140/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0660 - mae: 4.5407\n",
            "Epoch 141/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.1649 - mae: 4.6377\n",
            "Epoch 142/500\n",
            "61/61 [==============================] - 2s 33ms/step - loss: 4.0906 - mae: 4.5627\n",
            "Epoch 143/500\n",
            "61/61 [==============================] - 3s 46ms/step - loss: 4.1045 - mae: 4.5761\n",
            "Epoch 144/500\n",
            "61/61 [==============================] - 3s 46ms/step - loss: 4.0611 - mae: 4.5354\n",
            "Epoch 145/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.1638 - mae: 4.6404\n",
            "Epoch 146/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0974 - mae: 4.5710\n",
            "Epoch 147/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0346 - mae: 4.5049\n",
            "Epoch 148/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.1142 - mae: 4.5889\n",
            "Epoch 149/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0651 - mae: 4.5358\n",
            "Epoch 150/500\n",
            "61/61 [==============================] - 2s 31ms/step - loss: 4.0917 - mae: 4.5659\n",
            "Epoch 151/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0697 - mae: 4.5462\n",
            "Epoch 152/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0982 - mae: 4.5713\n",
            "Epoch 153/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0807 - mae: 4.5586\n",
            "Epoch 154/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0256 - mae: 4.4990\n",
            "Epoch 155/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0766 - mae: 4.5446\n",
            "Epoch 156/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0925 - mae: 4.5642\n",
            "Epoch 157/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0371 - mae: 4.5079\n",
            "Epoch 158/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0685 - mae: 4.5431\n",
            "Epoch 159/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0169 - mae: 4.4923\n",
            "Epoch 160/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.1050 - mae: 4.5812\n",
            "Epoch 161/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0571 - mae: 4.5296\n",
            "Epoch 162/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0150 - mae: 4.4895\n",
            "Epoch 163/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0093 - mae: 4.4828\n",
            "Epoch 164/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0219 - mae: 4.4994\n",
            "Epoch 165/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.2132 - mae: 4.6905\n",
            "Epoch 166/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0519 - mae: 4.5294\n",
            "Epoch 167/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0008 - mae: 4.4779\n",
            "Epoch 168/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0378 - mae: 4.5091\n",
            "Epoch 169/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0719 - mae: 4.5440\n",
            "Epoch 170/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0691 - mae: 4.5453\n",
            "Epoch 171/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0016 - mae: 4.4790\n",
            "Epoch 172/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0032 - mae: 4.4729\n",
            "Epoch 173/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0987 - mae: 4.5729\n",
            "Epoch 174/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0328 - mae: 4.5070\n",
            "Epoch 175/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0024 - mae: 4.4750\n",
            "Epoch 176/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9953 - mae: 4.4689\n",
            "Epoch 177/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0188 - mae: 4.4922\n",
            "Epoch 178/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0011 - mae: 4.4769\n",
            "Epoch 179/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0401 - mae: 4.5135\n",
            "Epoch 180/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9876 - mae: 4.4597\n",
            "Epoch 181/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9867 - mae: 4.4566\n",
            "Epoch 182/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9914 - mae: 4.4628\n",
            "Epoch 183/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0291 - mae: 4.5036\n",
            "Epoch 184/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9840 - mae: 4.4573\n",
            "Epoch 185/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0285 - mae: 4.5014\n",
            "Epoch 186/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0462 - mae: 4.5199\n",
            "Epoch 187/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0093 - mae: 4.4823\n",
            "Epoch 188/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9849 - mae: 4.4551\n",
            "Epoch 189/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0158 - mae: 4.4890\n",
            "Epoch 190/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0038 - mae: 4.4729\n",
            "Epoch 191/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9935 - mae: 4.4693\n",
            "Epoch 192/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0341 - mae: 4.5070\n",
            "Epoch 193/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0361 - mae: 4.5076\n",
            "Epoch 194/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.1194 - mae: 4.5954\n",
            "Epoch 195/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0504 - mae: 4.5236\n",
            "Epoch 196/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.1203 - mae: 4.5926\n",
            "Epoch 197/500\n",
            "61/61 [==============================] - 2s 31ms/step - loss: 4.0265 - mae: 4.5020\n",
            "Epoch 198/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0022 - mae: 4.4746\n",
            "Epoch 199/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0190 - mae: 4.4944\n",
            "Epoch 200/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0157 - mae: 4.4894\n",
            "Epoch 201/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0564 - mae: 4.5251\n",
            "Epoch 202/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0141 - mae: 4.4881\n",
            "Epoch 203/500\n",
            "61/61 [==============================] - 2s 31ms/step - loss: 4.0049 - mae: 4.4777\n",
            "Epoch 204/500\n",
            "61/61 [==============================] - 2s 31ms/step - loss: 3.9670 - mae: 4.4402\n",
            "Epoch 205/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9631 - mae: 4.4343\n",
            "Epoch 206/500\n",
            "61/61 [==============================] - 2s 32ms/step - loss: 3.9902 - mae: 4.4570\n",
            "Epoch 207/500\n",
            "61/61 [==============================] - 2s 31ms/step - loss: 4.0198 - mae: 4.4900\n",
            "Epoch 208/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9548 - mae: 4.4273\n",
            "Epoch 209/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0494 - mae: 4.5225\n",
            "Epoch 210/500\n",
            "61/61 [==============================] - 2s 31ms/step - loss: 4.0459 - mae: 4.5180\n",
            "Epoch 211/500\n",
            "61/61 [==============================] - 2s 32ms/step - loss: 4.0318 - mae: 4.5079\n",
            "Epoch 212/500\n",
            "61/61 [==============================] - 2s 31ms/step - loss: 4.0195 - mae: 4.4929\n",
            "Epoch 213/500\n",
            "61/61 [==============================] - 2s 31ms/step - loss: 3.9585 - mae: 4.4302\n",
            "Epoch 214/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0154 - mae: 4.4899\n",
            "Epoch 215/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9714 - mae: 4.4414\n",
            "Epoch 216/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0604 - mae: 4.5349\n",
            "Epoch 217/500\n",
            "61/61 [==============================] - 2s 28ms/step - loss: 3.9714 - mae: 4.4452\n",
            "Epoch 218/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0487 - mae: 4.5223\n",
            "Epoch 219/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9940 - mae: 4.4644\n",
            "Epoch 220/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9706 - mae: 4.4407\n",
            "Epoch 221/500\n",
            "61/61 [==============================] - 2s 28ms/step - loss: 3.9712 - mae: 4.4470\n",
            "Epoch 222/500\n",
            "61/61 [==============================] - 2s 28ms/step - loss: 4.0372 - mae: 4.5068\n",
            "Epoch 223/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9657 - mae: 4.4308\n",
            "Epoch 224/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0090 - mae: 4.4811\n",
            "Epoch 225/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9664 - mae: 4.4388\n",
            "Epoch 226/500\n",
            "61/61 [==============================] - 2s 31ms/step - loss: 3.9643 - mae: 4.4389\n",
            "Epoch 227/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9561 - mae: 4.4283\n",
            "Epoch 228/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0029 - mae: 4.4735\n",
            "Epoch 229/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9799 - mae: 4.4527\n",
            "Epoch 230/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9983 - mae: 4.4706\n",
            "Epoch 231/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9758 - mae: 4.4468\n",
            "Epoch 232/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9701 - mae: 4.4435\n",
            "Epoch 233/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0085 - mae: 4.4801\n",
            "Epoch 234/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0193 - mae: 4.4908\n",
            "Epoch 235/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9679 - mae: 4.4375\n",
            "Epoch 236/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9792 - mae: 4.4475\n",
            "Epoch 237/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9566 - mae: 4.4311\n",
            "Epoch 238/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9819 - mae: 4.4516\n",
            "Epoch 239/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9613 - mae: 4.4376\n",
            "Epoch 240/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9911 - mae: 4.4638\n",
            "Epoch 241/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0177 - mae: 4.4889\n",
            "Epoch 242/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0419 - mae: 4.5128\n",
            "Epoch 243/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9399 - mae: 4.4119\n",
            "Epoch 244/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0052 - mae: 4.4760\n",
            "Epoch 245/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9940 - mae: 4.4652\n",
            "Epoch 246/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9796 - mae: 4.4510\n",
            "Epoch 247/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9339 - mae: 4.4056\n",
            "Epoch 248/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0539 - mae: 4.5233\n",
            "Epoch 249/500\n",
            "61/61 [==============================] - 2s 32ms/step - loss: 4.0277 - mae: 4.5022\n",
            "Epoch 250/500\n",
            "61/61 [==============================] - 2s 32ms/step - loss: 3.9429 - mae: 4.4140\n",
            "Epoch 251/500\n",
            "61/61 [==============================] - 2s 31ms/step - loss: 3.9612 - mae: 4.4303\n",
            "Epoch 252/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9641 - mae: 4.4375\n",
            "Epoch 253/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9506 - mae: 4.4241\n",
            "Epoch 254/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0189 - mae: 4.4953\n",
            "Epoch 255/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9752 - mae: 4.4480\n",
            "Epoch 256/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0551 - mae: 4.5293\n",
            "Epoch 257/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0913 - mae: 4.5622\n",
            "Epoch 258/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9695 - mae: 4.4423\n",
            "Epoch 259/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9315 - mae: 4.4043\n",
            "Epoch 260/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9974 - mae: 4.4686\n",
            "Epoch 261/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0003 - mae: 4.4747\n",
            "Epoch 262/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9476 - mae: 4.4174\n",
            "Epoch 263/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9911 - mae: 4.4647\n",
            "Epoch 264/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9835 - mae: 4.4556\n",
            "Epoch 265/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9394 - mae: 4.4098\n",
            "Epoch 266/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9336 - mae: 4.4073\n",
            "Epoch 267/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9412 - mae: 4.4136\n",
            "Epoch 268/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9493 - mae: 4.4228\n",
            "Epoch 269/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9343 - mae: 4.4047\n",
            "Epoch 270/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9775 - mae: 4.4516\n",
            "Epoch 271/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9424 - mae: 4.4131\n",
            "Epoch 272/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0188 - mae: 4.4924\n",
            "Epoch 273/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9406 - mae: 4.4111\n",
            "Epoch 274/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9342 - mae: 4.4068\n",
            "Epoch 275/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9700 - mae: 4.4375\n",
            "Epoch 276/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9641 - mae: 4.4374\n",
            "Epoch 277/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9310 - mae: 4.4045\n",
            "Epoch 278/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9694 - mae: 4.4445\n",
            "Epoch 279/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9517 - mae: 4.4285\n",
            "Epoch 280/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9519 - mae: 4.4254\n",
            "Epoch 281/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9801 - mae: 4.4539\n",
            "Epoch 282/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9457 - mae: 4.4191\n",
            "Epoch 283/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0543 - mae: 4.5285\n",
            "Epoch 284/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9305 - mae: 4.3992\n",
            "Epoch 285/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9540 - mae: 4.4220\n",
            "Epoch 286/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9095 - mae: 4.3779\n",
            "Epoch 287/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9852 - mae: 4.4570\n",
            "Epoch 288/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9460 - mae: 4.4201\n",
            "Epoch 289/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9369 - mae: 4.4102\n",
            "Epoch 290/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9325 - mae: 4.4058\n",
            "Epoch 291/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9841 - mae: 4.4537\n",
            "Epoch 292/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0126 - mae: 4.4831\n",
            "Epoch 293/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9843 - mae: 4.4582\n",
            "Epoch 294/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9222 - mae: 4.3922\n",
            "Epoch 295/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0090 - mae: 4.4841\n",
            "Epoch 296/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9538 - mae: 4.4282\n",
            "Epoch 297/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9613 - mae: 4.4339\n",
            "Epoch 298/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9576 - mae: 4.4302\n",
            "Epoch 299/500\n",
            "61/61 [==============================] - 2s 31ms/step - loss: 3.9894 - mae: 4.4654\n",
            "Epoch 300/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9153 - mae: 4.3876\n",
            "Epoch 301/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9609 - mae: 4.4331\n",
            "Epoch 302/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9426 - mae: 4.4155\n",
            "Epoch 303/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9348 - mae: 4.4055\n",
            "Epoch 304/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9839 - mae: 4.4602\n",
            "Epoch 305/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9148 - mae: 4.3858\n",
            "Epoch 306/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9375 - mae: 4.4095\n",
            "Epoch 307/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9161 - mae: 4.3892\n",
            "Epoch 308/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9324 - mae: 4.4021\n",
            "Epoch 309/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9334 - mae: 4.4049\n",
            "Epoch 310/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9139 - mae: 4.3835\n",
            "Epoch 311/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9553 - mae: 4.4251\n",
            "Epoch 312/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0109 - mae: 4.4838\n",
            "Epoch 313/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9470 - mae: 4.4156\n",
            "Epoch 314/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 4.0083 - mae: 4.4794\n",
            "Epoch 315/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9835 - mae: 4.4568\n",
            "Epoch 316/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9741 - mae: 4.4442\n",
            "Epoch 317/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9338 - mae: 4.4048\n",
            "Epoch 318/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9065 - mae: 4.3796\n",
            "Epoch 319/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9588 - mae: 4.4275\n",
            "Epoch 320/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9487 - mae: 4.4183\n",
            "Epoch 321/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9240 - mae: 4.3979\n",
            "Epoch 322/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9266 - mae: 4.3988\n",
            "Epoch 323/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9104 - mae: 4.3792\n",
            "Epoch 324/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8888 - mae: 4.3596\n",
            "Epoch 325/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9164 - mae: 4.3865\n",
            "Epoch 326/500\n",
            "61/61 [==============================] - 2s 31ms/step - loss: 3.9154 - mae: 4.3876\n",
            "Epoch 327/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9151 - mae: 4.3879\n",
            "Epoch 328/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9396 - mae: 4.4129\n",
            "Epoch 329/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9033 - mae: 4.3795\n",
            "Epoch 330/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9973 - mae: 4.4682\n",
            "Epoch 331/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9552 - mae: 4.4285\n",
            "Epoch 332/500\n",
            "61/61 [==============================] - 2s 31ms/step - loss: 3.9384 - mae: 4.4150\n",
            "Epoch 333/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9282 - mae: 4.3983\n",
            "Epoch 334/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9085 - mae: 4.3815\n",
            "Epoch 335/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9285 - mae: 4.3980\n",
            "Epoch 336/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9251 - mae: 4.4000\n",
            "Epoch 337/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9734 - mae: 4.4446\n",
            "Epoch 338/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9081 - mae: 4.3824\n",
            "Epoch 339/500\n",
            "61/61 [==============================] - 2s 31ms/step - loss: 3.9194 - mae: 4.3894\n",
            "Epoch 340/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9149 - mae: 4.3860\n",
            "Epoch 341/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9314 - mae: 4.4034\n",
            "Epoch 342/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9088 - mae: 4.3777\n",
            "Epoch 343/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9116 - mae: 4.3835\n",
            "Epoch 344/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9247 - mae: 4.3969\n",
            "Epoch 345/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9130 - mae: 4.3843\n",
            "Epoch 346/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9343 - mae: 4.4066\n",
            "Epoch 347/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8891 - mae: 4.3610\n",
            "Epoch 348/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9296 - mae: 4.4000\n",
            "Epoch 349/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9092 - mae: 4.3781\n",
            "Epoch 350/500\n",
            "61/61 [==============================] - 2s 31ms/step - loss: 3.8810 - mae: 4.3498\n",
            "Epoch 351/500\n",
            "61/61 [==============================] - 2s 28ms/step - loss: 3.8822 - mae: 4.3545\n",
            "Epoch 352/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9151 - mae: 4.3884\n",
            "Epoch 353/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.8878 - mae: 4.3574\n",
            "Epoch 354/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.8798 - mae: 4.3565\n",
            "Epoch 355/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9445 - mae: 4.4141\n",
            "Epoch 356/500\n",
            "61/61 [==============================] - 2s 31ms/step - loss: 3.9099 - mae: 4.3836\n",
            "Epoch 357/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9687 - mae: 4.4380\n",
            "Epoch 358/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9769 - mae: 4.4511\n",
            "Epoch 359/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9661 - mae: 4.4344\n",
            "Epoch 360/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9041 - mae: 4.3735\n",
            "Epoch 361/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9111 - mae: 4.3826\n",
            "Epoch 362/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9040 - mae: 4.3745\n",
            "Epoch 363/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9431 - mae: 4.4116\n",
            "Epoch 364/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9501 - mae: 4.4236\n",
            "Epoch 365/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9660 - mae: 4.4367\n",
            "Epoch 366/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.8896 - mae: 4.3635\n",
            "Epoch 367/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8896 - mae: 4.3631\n",
            "Epoch 368/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9162 - mae: 4.3879\n",
            "Epoch 369/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9069 - mae: 4.3742\n",
            "Epoch 370/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.8937 - mae: 4.3663\n",
            "Epoch 371/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9257 - mae: 4.3974\n",
            "Epoch 372/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9671 - mae: 4.4395\n",
            "Epoch 373/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.8845 - mae: 4.3546\n",
            "Epoch 374/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9036 - mae: 4.3732\n",
            "Epoch 375/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.8816 - mae: 4.3564\n",
            "Epoch 376/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9456 - mae: 4.4198\n",
            "Epoch 377/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9241 - mae: 4.3955\n",
            "Epoch 378/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9456 - mae: 4.4176\n",
            "Epoch 379/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8986 - mae: 4.3666\n",
            "Epoch 380/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.8803 - mae: 4.3537\n",
            "Epoch 381/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9101 - mae: 4.3845\n",
            "Epoch 382/500\n",
            "61/61 [==============================] - 2s 31ms/step - loss: 3.9547 - mae: 4.4280\n",
            "Epoch 383/500\n",
            "61/61 [==============================] - 2s 32ms/step - loss: 3.9000 - mae: 4.3713\n",
            "Epoch 384/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9121 - mae: 4.3836\n",
            "Epoch 385/500\n",
            "61/61 [==============================] - 2s 31ms/step - loss: 3.9141 - mae: 4.3860\n",
            "Epoch 386/500\n",
            "61/61 [==============================] - 2s 31ms/step - loss: 3.8650 - mae: 4.3357\n",
            "Epoch 387/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9194 - mae: 4.3899\n",
            "Epoch 388/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9555 - mae: 4.4244\n",
            "Epoch 389/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9507 - mae: 4.4245\n",
            "Epoch 390/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.8952 - mae: 4.3689\n",
            "Epoch 391/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8837 - mae: 4.3537\n",
            "Epoch 392/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8839 - mae: 4.3541\n",
            "Epoch 393/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.8846 - mae: 4.3567\n",
            "Epoch 394/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0075 - mae: 4.4801\n",
            "Epoch 395/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9239 - mae: 4.3930\n",
            "Epoch 396/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9012 - mae: 4.3727\n",
            "Epoch 397/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9085 - mae: 4.3797\n",
            "Epoch 398/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.8910 - mae: 4.3631\n",
            "Epoch 399/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8981 - mae: 4.3698\n",
            "Epoch 400/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9301 - mae: 4.4037\n",
            "Epoch 401/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9174 - mae: 4.3902\n",
            "Epoch 402/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9410 - mae: 4.4161\n",
            "Epoch 403/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9057 - mae: 4.3741\n",
            "Epoch 404/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9140 - mae: 4.3885\n",
            "Epoch 405/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9712 - mae: 4.4446\n",
            "Epoch 406/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0038 - mae: 4.4750\n",
            "Epoch 407/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.8866 - mae: 4.3571\n",
            "Epoch 408/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9481 - mae: 4.4200\n",
            "Epoch 409/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8816 - mae: 4.3556\n",
            "Epoch 410/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9306 - mae: 4.4003\n",
            "Epoch 411/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9030 - mae: 4.3782\n",
            "Epoch 412/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9266 - mae: 4.4041\n",
            "Epoch 413/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9021 - mae: 4.3728\n",
            "Epoch 414/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9057 - mae: 4.3776\n",
            "Epoch 415/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9149 - mae: 4.3840\n",
            "Epoch 416/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8819 - mae: 4.3527\n",
            "Epoch 417/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8946 - mae: 4.3648\n",
            "Epoch 418/500\n",
            "61/61 [==============================] - 2s 31ms/step - loss: 3.8713 - mae: 4.3438\n",
            "Epoch 419/500\n",
            "61/61 [==============================] - 2s 33ms/step - loss: 3.8938 - mae: 4.3616\n",
            "Epoch 420/500\n",
            "61/61 [==============================] - 2s 32ms/step - loss: 3.9404 - mae: 4.4106\n",
            "Epoch 421/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9102 - mae: 4.3795\n",
            "Epoch 422/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8751 - mae: 4.3474\n",
            "Epoch 423/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9293 - mae: 4.4010\n",
            "Epoch 424/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.8904 - mae: 4.3611\n",
            "Epoch 425/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.8829 - mae: 4.3554\n",
            "Epoch 426/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9394 - mae: 4.4077\n",
            "Epoch 427/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.8845 - mae: 4.3548\n",
            "Epoch 428/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9609 - mae: 4.4306\n",
            "Epoch 429/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8635 - mae: 4.3323\n",
            "Epoch 430/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.8759 - mae: 4.3480\n",
            "Epoch 431/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9524 - mae: 4.4288\n",
            "Epoch 432/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 4.0878 - mae: 4.5573\n",
            "Epoch 433/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.8810 - mae: 4.3521\n",
            "Epoch 434/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.8658 - mae: 4.3379\n",
            "Epoch 435/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9146 - mae: 4.3864\n",
            "Epoch 436/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8851 - mae: 4.3554\n",
            "Epoch 437/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9330 - mae: 4.4035\n",
            "Epoch 438/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8701 - mae: 4.3418\n",
            "Epoch 439/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9008 - mae: 4.3695\n",
            "Epoch 440/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.8829 - mae: 4.3543\n",
            "Epoch 441/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8725 - mae: 4.3418\n",
            "Epoch 442/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.8960 - mae: 4.3672\n",
            "Epoch 443/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8666 - mae: 4.3388\n",
            "Epoch 444/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.8766 - mae: 4.3474\n",
            "Epoch 445/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8946 - mae: 4.3672\n",
            "Epoch 446/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.8634 - mae: 4.3352\n",
            "Epoch 447/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.8712 - mae: 4.3406\n",
            "Epoch 448/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9400 - mae: 4.4109\n",
            "Epoch 449/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9837 - mae: 4.4574\n",
            "Epoch 450/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.8682 - mae: 4.3354\n",
            "Epoch 451/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8750 - mae: 4.3456\n",
            "Epoch 452/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9136 - mae: 4.3841\n",
            "Epoch 453/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.8806 - mae: 4.3547\n",
            "Epoch 454/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8587 - mae: 4.3329\n",
            "Epoch 455/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9007 - mae: 4.3722\n",
            "Epoch 456/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9453 - mae: 4.4184\n",
            "Epoch 457/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8751 - mae: 4.3459\n",
            "Epoch 458/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.8801 - mae: 4.3536\n",
            "Epoch 459/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.8994 - mae: 4.3719\n",
            "Epoch 460/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8733 - mae: 4.3433\n",
            "Epoch 461/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9061 - mae: 4.3767\n",
            "Epoch 462/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.8899 - mae: 4.3607\n",
            "Epoch 463/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8527 - mae: 4.3248\n",
            "Epoch 464/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8802 - mae: 4.3539\n",
            "Epoch 465/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9294 - mae: 4.3987\n",
            "Epoch 466/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8804 - mae: 4.3522\n",
            "Epoch 467/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9281 - mae: 4.4010\n",
            "Epoch 468/500\n",
            "61/61 [==============================] - 2s 31ms/step - loss: 3.8808 - mae: 4.3512\n",
            "Epoch 469/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8992 - mae: 4.3687\n",
            "Epoch 470/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8774 - mae: 4.3506\n",
            "Epoch 471/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9053 - mae: 4.3756\n",
            "Epoch 472/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.8538 - mae: 4.3214\n",
            "Epoch 473/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.8703 - mae: 4.3384\n",
            "Epoch 474/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9116 - mae: 4.3844\n",
            "Epoch 475/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8572 - mae: 4.3285\n",
            "Epoch 476/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8631 - mae: 4.3336\n",
            "Epoch 477/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8649 - mae: 4.3374\n",
            "Epoch 478/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8880 - mae: 4.3589\n",
            "Epoch 479/500\n",
            "61/61 [==============================] - 3s 44ms/step - loss: 3.9562 - mae: 4.4264\n",
            "Epoch 480/500\n",
            "61/61 [==============================] - 3s 46ms/step - loss: 3.8598 - mae: 4.3284\n",
            "Epoch 481/500\n",
            "61/61 [==============================] - 2s 34ms/step - loss: 3.8951 - mae: 4.3667\n",
            "Epoch 482/500\n",
            "61/61 [==============================] - 2s 31ms/step - loss: 3.8834 - mae: 4.3504\n",
            "Epoch 483/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8928 - mae: 4.3644\n",
            "Epoch 484/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9140 - mae: 4.3860\n",
            "Epoch 485/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8558 - mae: 4.3226\n",
            "Epoch 486/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8530 - mae: 4.3224\n",
            "Epoch 487/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.8675 - mae: 4.3384\n",
            "Epoch 488/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.9240 - mae: 4.3946\n",
            "Epoch 489/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.8794 - mae: 4.3508\n",
            "Epoch 490/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8708 - mae: 4.3426\n",
            "Epoch 491/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.8721 - mae: 4.3459\n",
            "Epoch 492/500\n",
            "61/61 [==============================] - 2s 31ms/step - loss: 3.8761 - mae: 4.3485\n",
            "Epoch 493/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8620 - mae: 4.3332\n",
            "Epoch 494/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8665 - mae: 4.3378\n",
            "Epoch 495/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8882 - mae: 4.3623\n",
            "Epoch 496/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8975 - mae: 4.3703\n",
            "Epoch 497/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.8723 - mae: 4.3471\n",
            "Epoch 498/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9084 - mae: 4.3845\n",
            "Epoch 499/500\n",
            "61/61 [==============================] - 2s 29ms/step - loss: 3.8720 - mae: 4.3423\n",
            "Epoch 500/500\n",
            "61/61 [==============================] - 2s 30ms/step - loss: 3.9098 - mae: 4.3803\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "MKkic-mLdkRZ",
        "colab": {}
      },
      "source": [
        "rnn_forecast = model_forecast(model, series[..., np.newaxis], window_size)\n",
        "rnn_forecast = rnn_forecast[split_time - window_size:-1, -1, 0]"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "4N1toSetdnQq",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 388
        },
        "outputId": "a8851ade-9a97-4701-a9f1-ad25b2a55923"
      },
      "source": [
        "plt.figure(figsize=(10, 6))\n",
        "plot_series(time_valid, x_valid)\n",
        "plot_series(time_valid, rnn_forecast)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "ULKO3JINdqkp",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "efb9d77f-d8d6-4c89-a8b5-a5850e8f2967"
      },
      "source": [
        "tf.keras.metrics.mean_absolute_error(x_valid, rnn_forecast).numpy()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "5.5321884"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 25
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "ok8LjNbbkig4",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 607
        },
        "outputId": "38266046-52af-4dfd-c53a-ee4dc25efe42"
      },
      "source": [
        "import matplotlib.image  as mpimg\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "#-----------------------------------------------------------\n",
        "# Retrieve a list of list results on training and test data\n",
        "# sets for each training epoch\n",
        "#-----------------------------------------------------------\n",
        "mae=history.history['mae']\n",
        "loss=history.history['loss']\n",
        "\n",
        "epochs=range(len(loss)) # Get number of epochs\n",
        "\n",
        "#------------------------------------------------\n",
        "# Plot MAE and Loss\n",
        "#------------------------------------------------\n",
        "plt.plot(epochs, mae, 'r')\n",
        "plt.plot(epochs, loss, 'b')\n",
        "plt.title('MAE and Loss')\n",
        "plt.xlabel(\"Epochs\")\n",
        "plt.ylabel(\"Accuracy\")\n",
        "plt.legend([\"MAE\", \"Loss\"])\n",
        "\n",
        "plt.figure()\n",
        "\n",
        "epochs_zoom = epochs[200:]\n",
        "mae_zoom = mae[200:]\n",
        "loss_zoom = loss[200:]\n",
        "\n",
        "#------------------------------------------------\n",
        "# Plot Zoomed MAE and Loss\n",
        "#------------------------------------------------\n",
        "plt.plot(epochs_zoom, mae_zoom, 'r')\n",
        "plt.plot(epochs_zoom, loss_zoom, 'b')\n",
        "plt.title('MAE and Loss')\n",
        "plt.xlabel(\"Epochs\")\n",
        "plt.ylabel(\"Accuracy\")\n",
        "plt.legend([\"MAE\", \"Loss\"])\n",
        "\n",
        "plt.figure()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 0 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 26
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX4AAAEWCAYAAABhffzLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3deXgUVfY38O8JCSQsQoCwRg2MIi4gIAMoMOA2AwiKuMEg4PIO44KguDsjqDP+1HFB0VEHV8YFXEZRQRFEVpUlLCIIsgYJsu8hQLbz/nGq0p10h3SWTifV38/z9NNdt6q7bnU6p26dunVLVBVERBQ9YiJdASIiqlgM/EREUYaBn4goyjDwExFFGQZ+IqIow8BPRBRlGPiJwkxE3haRf0a6HkQuBn6q1EQkTUSyRKRhofLlIqIiklKo/BGnvHOh8htEJFdEMgo9moV/K4rm1GtBJOtA0YeBn6qCzQAGuRMi0gZAzcILiYgAGApgn/Nc2A+qWrvQ47dwVZqosmLgp6rgHRQM5MMA/DfIct0BNAUwEsBAEale2hWKyAsislVEDonIUhHp7jfvERH5UET+KyKHRWS1iHT0m99eRJY58z4AEF/KOlwgIktE5KDzfIHfvBtEZJOzjs0iMtgpP01E5jrv2eOsn6gABn6qChYCOElEzhSRagAGAng3yHLDAHwB4ENnul8Z1rkEQDsA9QG8D+AjEfEP4JcDmAygHoDPAbwEAM7OZgpsZ1UfwEcArirpykWkPoBpAMYDaADgOQDTRKSBiNRyynurah0AFwBY4bz1HwBmAEgEkAzgxZKum7yPgZ+qCrfVfymANQC2+c8UkZoArgHwvqpmA/gYgemeLiJywO+xsaiVqeq7qrpXVXNU9VkANQCc4bfIAlX9UlVznbqd664DQByA51U1W1U/hu1ESuoyAOtV9R2nDpMArIVvZ5YH4BwRSVDV7aq62inPBnAqgGaqekxVef6AAjDwU1XxDoA/A7gBwdM8VwLIAfClM/0egN4ikuS3zEJVref3+F1RKxORe0RkjZMyOQCgLgD/E8w7/F5nAogXkVgAzQBs04KjH24JbRMLaBbkfVsANFfVIwCuA3ALgO0iMk1EWjvL3AdAACx2UlA3lWLd5HEM/FQlqOoW2EnePgA+CbLIMAC1AfwqIjtgKZY42M6iRJx8/n0ArgWQqKr1AByEBdTibAfQ3DnR7DqlpHUA8Bus5e7vFDhHOqr6tapeCjunsRbAa075DlX9i6o2A/BXAC+LyGmlWD95GAM/VSU3A7jIafHmE5HmAC4G0BeWl28HS708heC9e4pTB3b0sBtArIiMAXBSiO/9wXnvSBGJE5EBADoV8x4RkXj/B+zIpZWI/FlEYkXkOgBnAZgqIo1F5Aon138cQAYs9QMRuUZEkp3P3Q9A3XlELgZ+qjJUdaOqpgaZNQTAClWd4bR4d6jqDtgJ0LYico6z3PlB+vH/PsjnfQ1gOoB1sPTKMQBbQ6xjFoABsJTUPlhKJtgRir8LABwt9DgI25HdDWAv7Aikr6rugf3fjoYdFewD0APArc5n/R7AIhHJgJ10HqWqm0KpO0UP4Y1YiIiiC1v8RERRhoGfiCjKMPATEUUZBn4ioigTG+kKhKJhw4aakpIS6WoQEVUpS5cu3aOqSYXLq0TgT0lJQWpqsF58RERUFBEJetU4Uz1ERFGGgZ+IKMow8BMRRZkqkeMnIiqt7OxspKen49ixY5GuStjEx8cjOTkZcXFxIS3PwE9Enpaeno46deogJSUFBQdN9QZVxd69e5Geno4WLVqE9B6meojI044dO4YGDRp4MugDgIigQYMGJTqiYeAnIs/zatB3lXT7vB34p04Fnnoq0rUgIqpUvB34v/wSeOaZSNeCiKKciOD666/Pn87JyUFSUhL69u1bYLn+/fujS5cuBcoeeeQRNG/eHO3atct/HDhwoEz18fbJXRGA9xsgogirVasWVq1ahaNHjyIhIQEzZ85E8+bNCyxz4MABLF26FLVr18amTZvQsmXL/Hl33XUX7rnnnnKrj7db/Az8RFRJ9OnTB9OmTQMATJo0CYMGDSow/5NPPkG/fv0wcOBATJ48Oax18XaLPyaGgZ+IfO68E1ixonw/s1074Pnni11s4MCBeOyxx9C3b1+sXLkSN910E+bPn58/f9KkSRgzZgwaN26Mq666Cg899FD+vHHjxuHdd98FACQmJmL27NllqrK3A78IkMf7TBNR5LVt2xZpaWmYNGkS+vTpU2Dezp07sX79enTr1g0igri4OKxatQrnnGO3iy7vVI/3Az9b/ETkCqFlHk6XX3457rnnHsyZMwd79+7NL//www+xf//+/AuwDh06hEmTJuHxxx8PSz2Y4yciqiA33XQTxo4dizZt2hQonzRpEqZPn460tDSkpaVh6dKlYc3zM/ATEVWQ5ORkjBw5skBZWloatmzZUqAbZ4sWLVC3bl0sWrQIgOX4/btzpqWllakeolUgMHbs2FFLcyOWFy6cgrnzY/BJzuVhqBURVQVr1qzBmWeeGelqhF2w7RSRparasfCyns7x/7K/Eebnnh7pahARVSqeTvXExCgU3h6jg4iopDwd+EWAPG9vIhFRiXk6KoqALX4iokLCFvhF5E0R2SUiq4LMu1tEVEQahmv9ACBg4CciKiycLf63AfQqXCgiJwP4I4Bfw7huZ10M/EREhYUt8KvqPAD7gswaB+A+AGHvR8rAT0SVQe3atSNdhQIqNMcvIlcA2KaqP4aw7HARSRWR1N27d5dyfQz8RESFVVjgF5GaAB4CMCaU5VV1gqp2VNWOSUlJpVpnTAwDPxFVTitWrECXLl3Qtm1bXHnlldi/fz8AYPz48TjrrLPQtm1bDBw4EAAwd+7c/Kt227dvj8OHD5dp3RV5AdfvALQA8KNzf8hkAMtEpJOq7gjHCtmdk4j8RXBU5gBDhw7Fiy++iB49emDMmDF49NFH8fzzz+PJJ5/E5s2bUaNGjfw7bT3zzDP497//ja5duyIjIwPx8fFlqnOFRUVV/UlVG6lqiqqmAEgH0CFcQR/wS/VUgWEpiCh6HDx4EAcOHECPHj0AAMOGDcO8efMA2PDNgwcPxrvvvovYWGubd+3aFaNHj8b48eNx4MCB/PLSCluLX0QmAegJoKGIpAMYq6pvhGt9wevgF/hLeBd6IvKeCI/KHJJp06Zh3rx5+OKLL/D444/jp59+wgMPPIDLLrsMX375Jbp27Yqvv/4arVu3LvU6whb4VXVQMfNTwrVuF1v8RFQZ1a1bF4mJiZg/fz66d++Od955Bz169EBeXh62bt2KCy+8EN26dcPkyZORkZGBvXv3ok2bNmjTpg2WLFmCtWvXVs7AXxkw8BNRZZCZmYnk5OT86dGjR2PixIm45ZZbkJmZiZYtW+Ktt95Cbm4urr/+ehw8eBCqipEjR6JevXp4+OGHMXv2bMTExODss89G7969y1QfBn4iojDLK+IWsAsXLgwoW7BgQUDZiy++WK718XSXl/zunAz8RET5PB3487tz8obrRET5PB/4FTFs8RNFuapwp8GyKOn2eT7wA2DgJ4pi8fHx2Lt3r2eDv6pi7969Jbqoy/MndwFA8zhwA1G0Sk5ORnp6Oko75ldVEB8fX6DXUHEY+InI0+Li4tCiRYtIV6NSiYpUj+Z58xCPiKg0PB34Y5ytY+AnIvLxdOB3W/x5OezOSUTkiorAzxY/EZEPAz8RUZRh4CciijIM/EREUcbbgZ+9eoiIAng68Me4LX7GfSKifJ4O/OJE/rxcRn4iIpe3A7/b4s9lP34iIld0BH7m+ImI8jHwExFFGW8HfvbqISIK4O3A7zT52auHiMjH04E/JsYiPlv8REQ+ng78bouf3TmJiHy8HfjdHD+7cxIR5fN24GeOn4goQNgCv4i8KSK7RGSVX9nTIrJWRFaKyKciUi9c67f12TNz/EREPuFs8b8NoFehspkAzlHVtgDWAXgwjOtn4CciCiJsgV9V5wHYV6hshqrmOJMLASSHa/0AAz8RUTCRzPHfBOCromaKyHARSRWR1N27d5dqBfkndxn3iYjyRSTwi8jfAOQAeK+oZVR1gqp2VNWOSUlJpVpPjDM6J1v8REQ+sRW9QhG5AUBfABerhrct7qZ68nLYnZOIyFWhgV9EegG4D0APVc0M+/pi2J2TiKiwcHbnnATgBwBniEi6iNwM4CUAdQDMFJEVIvJquNZvdbBnpnqIiHzC1uJX1UFBit8I1/qCYeAnIgrk7St32auHiCiAtwO/sFcPEVFhng78MbwRCxFRAE8HfrdXD4dlJiLy8Xbg58ldIqIA0RH4GfeJiPJ5O/BzyAYiogDeDvxM9RARBfB24Gc/fiKiAJ4O/Bydk4gokKcDP7tzEhEF8nbgZ46fiCiAtwM/c/xERAG8Hfg5Vg8RUQCPB357ZoufiMjH24GfvXqIiAJ4OvDHMMdPRBTA04Gf3TmJiAJFReBnqoeIyMfbgT//5C4DPxGRy9uBP/8OXJGtBxFRZVJs4BeRfiJSJXcQ+f342eInIsoXSkC/DsB6EfmXiLQOd4XKE1v8RESBig38qno9gPYANgJ4W0R+EJHhIlIn7LUro/zROdniJyLKF1IKR1UPAfgYwGQATQFcCWCZiNwRxrqVma87Z4QrQkRUiYSS479cRD4FMAdAHIBOqtobwLkA7g5v9cqG3TmJiALFhrDMVQDGqeo8/0JVzRSRm8NTrfLBsXqIiAKFkup5BMBid0JEEkQkBQBUdVZRbxKRN0Vkl4is8iurLyIzRWS985xY6pqHgC1+IqJAoQT+jwD494vJdcqK8zaAXoXKHgAwS1VPBzDLmQ4btviJiAKFEvhjVTXLnXBeVy/uTU5qaF+h4isATHReTwTQP8R6lgpb/EREgUIJ/LtF5HJ3QkSuALCnlOtrrKrbndc7ADQuakGny2iqiKTu3r27VCvj6JxERIFCCfy3AHhIRH4Vka0A7gfw17KuWK1zfZEhWVUnqGpHVe2YlJRUqnXkd+fkBVxERPmK7dWjqhsBdBGR2s50RhnWt1NEmqrqdhFpCmBXGT6rWEz1EBEFCqU7J0TkMgBnA4j3G//msVKs73MAwwA86Tx/VorPCBkDPxFRoFAu4HoVNl7PHQAEwDUATg3hfZMA/ADgDBFJd/r8PwngUhFZD+ASZzps2KuHiChQKC3+C1S1rYisVNVHReRZAF8V9yZVHVTErItLVMMyEI7VQ0QUIJSTu8ec50wRaQYgGzZeT6WX3+LnyV0ionyhtPi/EJF6AJ4GsAzWE+e1sNaqnPha/BGuCBFRJXLCwO/cgGWWqh4A8D8RmQogXlUPVkjtyiimGrtzEhEVdsJUj6rmAfi33/TxqhL0AfbqISIKJpQc/ywRuUrcfpxVCAM/EVGgUAL/X2GDsh0XkUMiclhEDoW5XuWC3TmJiAKFcuVupb/FYlF4cpeIKFCxgV9E/hCsvPCNWSoj9uMnIgoUSnfOe/1exwPoBGApgIvCUqNyxH78RESBQkn19POfFpGTATwfthqVI7c7Jxv8REQ+oZzcLSwdwJnlXZFwELcffy4jPxGRK5Qc/4vwjZsfA6Ad7AreSs83kmiEK0JEVImEkuNP9XudA2CSqn4XpvqUK/bqISIKFErg/xjAMVXNBQARqSYiNVU1M7xVKzv26iEiChTSlbsAEvymEwB8E57qlC/26iEiChRK4I/3v92i87pm+KpUfpjqISIKFErgPyIiHdwJETkPwNHwVan8sDsnEVGgUHL8dwL4SER+g916sQnsVoyVntvi57DMREQ+oVzAtUREWgM4wyn6RVWzw1ut8sHROYmIAoVys/XbAdRS1VWqugpAbRG5LfxVKzvm+ImIAoWS4/+LcwcuAICq7gfwl/BVqfww8BMRBQol8FfzvwmLiFQDUD18VSo/DPxERIFCObk7HcAHIvIfZ/qvAL4KX5XKj+9GLIz8RESuUAL//QCGA7jFmV4J69lT6eV352SvHiKifMWmepwbri8CkAYbi/8iAGvCW63yIdVs8/K0yt0umIgobIps8YtIKwCDnMceAB8AgKpeWDFVKzt25yQiCnSiFv9aWOu+r6p2U9UXAeSWx0pF5C4RWS0iq0RkkojEl8fnBqyHJ3eJiAKcKPAPALAdwGwReU1ELoZduVsmItIcwEgAHVX1HADVAAws6+cGXRcDPxFRgCIDv6pOUdWBAFoDmA0buqGRiLwiIn8s43pjASSISCxswLffyvh5QbFXDxFRoFBO7h5R1fede+8mA1gO6+lTKqq6DcAzAH6FHVEcVNUZpf28E/Hl+MPx6UREVVOJ7rmrqvtVdYKqXlzaFYpIIoArALQA0AxALRG5Pshyw0UkVURSd+/eXap1cXROIqJApbnZelldAmCzqu52Bnv7BMAFhRdydjAdVbVjUlJSqVbE7pxERIEiEfh/BdBFRGo6Q0FcjDBdF8DunEREgSo88KvqIth9fJcB+Mmpw4RwrIu9eoiIAoUyZEO5U9WxAMaGez0M/EREgSKR6qkwDPxERIEY+ImIooynAz+7cxIRBfJ04OfN1omIAnk78OcP2RDZehARVSYM/EREUYaBn4goyjDwExFFmegI/ByygYgoX3QE/uycyFaEiKgS8XTgj3G2Li+LgZ+IyOXpwO9r8WdHtiJERJWIpwO/S9niJyLK5/nAL8hji5+IyE8UBH7lyV0iIj/eD/zCXj1ERP68H/jZ4iciKsDzgb96tVxkZUW6FkRElYfnA3/N2GxkZsdFuhpERJWG5wN/rerZyMyuHulqEBFVGp4P/DVr5CAzrwaQmxvpqhARVQpREPhzcQS1gKNHI10VIqJKIQoCfx4yURPIzIx0VYiIKgXPB/5aNZ3An5ER6aoQEVUKng/8NWvHWODfvz/SVSEiqhS8H/hPirXAv2dPpKtCRFQpeD/w16tuJ3cZ+ImIAEQo8ItIPRH5WETWisgaETk/XOuqVT+eLX4iIj+xEVrvCwCmq+rVIlIdQM1wrahmYg1kIha6ew8kXCshIqpCKjzwi0hdAH8AcAMAqGoWgLCNplOzliAXscjetR+8fpeIKDKpnhYAdgN4S0SWi8jrIlKr8EIiMlxEUkUkdffu3aVeWU3nWCJz5+FSfwYRkZdEIvDHAugA4BVVbQ/gCIAHCi+kqhNUtaOqdkxKSir1ymo5u5QjaaXfeRAReUkkAn86gHRVXeRMfwzbEYRF7dr2fHjlJmDbtnCthoioyqjwwK+qOwBsFZEznKKLAfwcrvU1aGDPe9EAmDkzXKshIqoyItWr5w4A7zk9ejYBuDFcK2rY0J73xDYF1qwJ12qIiKqMiAR+VV0BoGNFrCs/8Dc6C/hlZUWskoioUvP8lbv5gb9+K2Dt2shWhoioEvB84K9ZE0hIAPbUSQE2bOAonUQU9Twf+AFr9e+plWJ34Vq8ONLVISKKqKgJ/LtjGgMiwNy5FvxHjQJUI101IqIKFxWB//TTgaUr45B30SXAK68AnTsD48cDe/dGumpERBUuKgL/lVcCO3YA3w94BvAf/mHr1shViogoQqIi8F9yiT0vPtoGqFPHN+PXXyNTISKiCIqKwN+woT3W/iLA8uXAjBk2gy1+IopCURH4AaB1a6cb/+9+B1x8MVCjhuX5r70WOP98tv6JKGpEasiGCte6NTBlinXkkZgYoEMH4IcfgEOHgCNHgJEjgfh4YNgwoHfvSFeXiChsoqbF37693X1xyxanYM4cYN064LffgAEDgM8+Az74wI4CCvvuO+sKyrF+iMgDoibwd+1qz999B+TlAYeOVbd+njExwAUX+BacPh2YOhWYPRto0wbo3h14802bx9E9icgDoibwn3MOULeupXtGjLDX2dnOzD/+EWjWDHjnHRvHuV8/4KKLgFWrgAULfMM8lLXf/6hRQBluKkNEVB6iJsdfrZoF/Mcf95Vt2wakpABo0cJ3kxZVYOjQgm/+6it7DjbIW2qqnSPo0aP4SrhppGPH7HwCEVEERE2LHwDuvLPgdNCOPFddBdxxB3D77b6yw879ej/8EDj3XOCRR6wr6NSpwO9/D/TsacE8mFWrgG++KVi2ebM9Z2YCr79uuSd/+/bZzoSIKAyiKvA3bGidd1q0sOmg3fhr1rSW+Usv2YnfM8+0E7sLFthJ4Lw84NFHgVNOsZSQa84cYP164F//sh3Bhg1W3qkTcOmlwM6dvmU3brTnsWOBv/zFdiAAkJVlzw0a2PuKkpUFvP9+4A6jOEeO+OW3iChaRU2qx/XCC8ATT9hN2H/91YJ/erp15Q/QtCmwYoUN89C8ue8M8d13W/qnTx/rJzp8eGAX0CFD7EKxo0dteuxY37yvvgLuugs4eNCm16wBDhwAbrkF+M9/rOznIHej3LXLgv077wD33Wdlf/5z6Btfu7bVedq00N9DRJ4jWgVGqOzYsaOmpqaW62c2bFjwXO1bb1kHnmbNbPz+YqnakQAATJxovYFatbKTBnv3Avfea3n8wikgkdBHBW3VCrjmGmDRIuCBB+zIQdXOJ8ydCzz4IPB//2c7g6NHbWeQlGSpqMK2b7eNc+sezHXXWarr2mttOj3d0lGtWtk6hg2zI5Q//CG0+hf20UfAp5/a0QoRhZ2ILFXVwLsdqmqlf5x33nla3v7zH9V+/VSffFK1XTvVlBTV6tVV778/cNmsLNV33lHNzQ3xw/PyVCdOVL3uOtWXX1Z9+21VQLVzZ9XfflO9/XbVGjWsrHVr1Zo1VU86SXXuXCsL9REbq3rPPaotW6o2aOArb9nSyocOVd20SXX5ctX+/X3z3Q1x67l+verWrb75S5eqZmernnuufTF/+pPq4MG++eeco7pkierw4baOvDx7FMd9/7ZtIf+dgjp4UDUjQ/Xo0bJ9DpHHAUjVIDE1alv8/h56yNI/gGV00tKA2FhgyRJLic+da8tMmgQMHFjKlaSmWuqoeXObPnLEehK1amXhMCPDBpA7cMDyT1lZdsXZr7/acp9/Dtx2m6VrBg068boSE4H9++119+52TuG333zzJ02yi9cWL7a0j3te4//9P98ynTvbkUaoEhNtKIz4eEtzLVgA9OplX978+XaBXPXqtuyUKUDLltbVaskS+8L9U2EnkpsL1K9vV1y3aGHb5h55nYiqvTc2jNnNX36x7f3LX0KrE0XOjz/a/6D/NTwexBb/CXzxhTVEzzjDnq+5RrVPH18DdehQe37uubKtZ8MG1VmzyqHCb7xhLe7UVNVDh+yQ5LrrVJ9+2rfM4sW+iouovvuu6g8/qMbHFzxqiIuz5+rVA48omjVTrVZNtWHDkh2JBHv4HzH4H7G4r+fPV509W/WWW1QXLlRdtUr11VdVhw1THT1adeNG1e3bVQcNKvgZzz6resMNqpdeakdTqrbs3/6m2qWL6r33qq5bp/rww7b8mWfaZ3/6qer06cGPGrZvV/3xR990Xp7qf/9b8PsdPFh1xAg7MsrNVZ061VenFSuK/tvl5qp+/33wI6Rnn1Xt2TNw3t69VrZzZ3G/DJORYb+JnJzQlq+swrkN7t/K41BEiz/iQT2UR7gDf1aW6vjxFkOvuSYwPv3hD/Y8YkTZ1uPG1lCyIuXmwAHVHTt80+vWqX7wgeqCBZbX+u031b//3Tb8iSesgl9/rTpypOqcOaqrV1sgq1PHAimgumiR6pEjNv+VV1Q7dlT97DNLGX3zje0obrvN9wUmJ/ten3mmPf/+96XfiZx2muqtt6omJBQsT0iwP5I73bRp8PcnJvpe16ljO4emTS1d1rOnakyMzbv+etUHH1Tt0aPg+9u3971++GHVAQMCd5i1atnryy6zHVF2turKlaotWlj5G2+ofvyx6owZqpMnqz7zjO/9P/5oqbqffrIdTrVqvm294w7VIUNUH3/cWhIvvaT65pu+v+/Bg5Y2dLf/lVdUjx1T3b078LeRl2c7sKee8v0of/lFdd680H9fhw/bDnHiRN9nbNpkdXdlZdmOPVhK8MCBgvXZscO3TPPmqtdeG7jOnBzb2XfuXIL8q2PLFtV9+3zfdWbmiZffsEH1r3+1nWlZHTpU8vqWUVGBn6meQpYvt/Hbgjn5ZODVV32De2ZmAjfeaOmgTz4p/rPdo/9duyrxBbyqwdMUbnl2NhAXd+LPyMuzoTAA68aalGQpkEaNLD0zfTpw+eWWavr+e6B/fxsWIyHB0kHTptk6jh61iy+2bLGTwh062Ill9+Tzli3WQ6pbN1vnn/8MrF7tq8esWTY9cqSvLDnZ0mDBrpNo3tzqm5Jin+12fY2Jses6XnzxxNt91VWWpit8X+f4ePvu3B5eJREbC+TkFL9c69ZW7+nTi16mZ0/giivsh7x2LfD3v/vmdepk16i89ppNjxhhvdnOOw/YtMnScdWrA4MHWxe4RYuAlSuBf/zD9xlXXGHbOGOGbfMHHwDHj9tnzpxpd797+237frp1s9TisGH297zuOjvp/7//AbfeCowebUOqAFbWvLnlY5cvtw4Tu3b51nvJJfb9du8OnH22XXUfG2t52e7d7S5M558P3H+/bccFF9jvDgBeftmum4mNtZ52cXH2fcfE2G+qUSNLv776KnDaafYd9+kDPPaYfZ+HDwNNmljvv/fes9/w3/9udUxMtFRmx472Gc2bA//8p3XKmD7dyvfssbrcfLPv/y4z0z539Wr7HQ4ZUuoUJVM9JbBli53rvOoqX8OgZ0/f6xo1VC+80NeoA+w9qta4mTfPt2N/7DFr9OXl+Zb9/nub98MPqnXrqqanF1z/hAmWOaASOnZMdc0aXwtT1b74detUjx/3tSSPHrUW+MaN1gpbvNheq1p5Xp6Vb96smpZmrVpV1XHjVJctU/3yS0s/zZplLe7PPvP9ADZtshTbsmXWEp8zx35It91mR1W33676r39Zau655yx19dprqlOmWA+CK6+0H0nt2qp3323Lbd+ueskldpTz9deq771nRyedO9vRlvvDatHCjlZGj7Z1HDlirVVAtUMH1VNPDTz6qVnTejicdpqlBJs0KfkR2KOPqnbv7pu+9dbSH801bmzPxaUXO3cu/TpO9Khd244CizparFu35J95/vmq3br5pkXs+YwzbPGlQDIAABAhSURBVH1ugOnVy1KbbscP9/HBB6X+lwBb/KXzySfWWB03zs4HrV1bsOt8gwbWe7N9eztfOWWKnUN8/nkbmsfdiU+ZYg1bwBo9NWpYF9IZM4AJE+z2kK+/bl35ExNtuaL+ND/9ZD07Fy0CTj019G2ZONEaUR4/n1X1HTliP6KTTip6Gf+jqpwcX4swI8M6APjbvt06FgD2g9uwwToI1KljJ8nr17fPyMgA6tWzFnBMjH1mWpq1dNPSgIULrZPAihX2nsREayU3amT/BKtXWz/ps86yrrvbtvkugMzJsSO55GRreU+ebD/gO+6w5caNs3q/8Qbw7LN2HcuZZ9oRwZIl1vLt188unFy2zFrLr79uLf9TT7UjiK1b7fHZZ7YN119vJ9z377eWfJ06tg0rVthRyNSpVqdevWy5hg2tLtu22WBeu3fbtTstW9qyKSn2zzdypHVYyM627dqwwVr6PXva3yUry67y91e3LtC2rb2ve3f7Gy9bZvU64wz7e+fl2Wf16WPvWbLEjooee6zUnQWKavEz8JfCTz/Z3+fCC+1/o1Mn34jNzZvb7yY+HrjnHjuyA+z3WtS9XlJSbAeydKkdqbrZioUL7be/Zo39zp97zso3b7aj5scft95Grp9/tg46KSmB68jO9nWqKepP/uOPlgWoXz/4/F9+se0I6TqHIPbvt+/mnHNK936iiNMgqdCMDOuh5v+PoWr/MK1aWZonMdHel5Xl+0dcudLSRI0anfjzy4CpnjA6csQyDG4mYft2O3IGVE8+2a4XAKxDzWOPFX1E2KpV0fP8z4+6jxtvtIzE559besgt//RTy1hs3+6rY2qqb74/t87Hjtm8Tp188559VvW+++w8sPv+wYMtK3H8uJUtXhz69+QenR8/XrrvmYhKBpUt1SMi1QCkAtimqn1PtGxla/GHwj0CTEmxo9ERI+yk8KhRdrQ8b54dPd56q7XsMzKAhx8GnnzSpgHgiy98qSTAWv9ZWb6hfkLx6qvWsPjvf32jSx8+bEeWe/ZYymjgQDvX6g5bceWVNgpF//62nHuey/8cY+fONjRRbKwdibZrZ42bffvsyNgdgHTPHjv67tDB15BJTbVzhoA1ho4e9WUiQjFunB05tGtnR/OViX8GhijSKl2qR0RGA+gI4CQvBv7ydPy4dVDp1MnSvgkJdp7hyBHrzHH8uKWB4uMtKIrY9V5ffukL2hdeaEF2+XK7v8ymTQU7tlxySeAgoqVVrZo95+bac2ysnSdxhzoaM8bOlcTFWdps5UrryJCQYAG9SRP7DFXfzqVlS3vvunWWEnXt2WNp6i1bLMXq2rTJPrtHD0tbb94MfPyxpZ//9CffEXWonSWOHbPvq0EDm16+3I7e3bSaql1f9/nnlqqrV6/k39vGjdZY+NOfSv5eKpmhQ+238+WXka5JeFWqVA+AZACzAFwEYGpxy1f2VE9Fczue+Fu9umCXaFXrlDJ6tHX9PnzY3nfrrTYSQ7NmlnaZMCGws0e7dr5U0g03qN55Z8H5Q4aojhqlevHFqo0aWVn9+tbx5Lzzik5XnejRo4dqmzb2ul8/60LvzqtRQ7Vt24K9rNyHe/0ZYF34d+60+vm/t3Cd7rnHypKSrPNLdraNrPH003bt2KhR1rkmI8O6eS9aZKm72FjVMWOsw5D7WevXq65dW/C6uGefVf3nPy0lNneu72+Vl2dd9C+6KHDUiowM33UeGRm2HXl51v3+9tutY8fw4Vb+88/W60zVru3at6/gZz31lKXoXnih+G7qCxeq/vprCX58QbidpMrb66/bNX3BLFtml0GURm6u729V3DU1eXn2N66qUJku4ALwMYDzAPQsKvADGA5LBaWecsopYfxqopd70Wp6uv24Dx2yHUFOjur+/b6LJo8ftwvcVqywoOb/T/7bb3bNlisjwy643bXLzj3cdZfqt99asBs82Lqptm+v+t13dt1TvXpWHhsbeD0WYL37/IchOussu9L6yBHVsWPtHEr//sF3OAMHWk+6svTucx916oS2XLVqgWVDh9qFxElJvrK6dW1IJff8jf8OzN2JXHZZwWvFCj969/a9njLFguQLLxRcplEju9brppusF+rBg76/6Usv+Zb76CP7W15/ve3Qjx2zHV5Wll1TNX++7SD8dySrV9tOKSHBGhJr1tj5oKuvtp2026V59mzVCy6wHq8ff6z60EPWA3bnTmuIpKX5lhsyxH4bGRkFg3NentXn5Zftd+XOGzvWeuuuXWu/WX+zZtk6tmyx7Z0xw7bv/fd979+61RpF7kXfhY0aZcstX27Tubk2zlfh5XNzreHl9uotLCfHGhzLlhUsz8s78ZBTR44UPS8URQX+Ck/1iEhfAH1U9TYR6QngHmWqJ+plZ/vSLgsXWs++OnUs7ZOVZQOFnn66LVNUp4f//c/Om5x/vh3Cv/WWnc8YO9ZSRaefbtePffedfc7PP1tKadcu68nXr5/dN+eccyzltHChrft3v7Nutu3bA1dfbb3wUlPtuqPvvrOeXJ06WQ/JBQush+PVV1uvwZNOsuWDadrUelrWrWtpuqNHLW1V2Esv2TkiwK5N+vbb4r/PxES7juiJJyyl4UpIsPRfqPf5KTyY7GmnWQ+z777z3c/i5JML3juodm07n1Srlv0Nd+wI/tn169v7CktKslTmCy/Y9Cmn2HdV3NBRNWrYcu3b29/TvRYtVEOHWupO1b6nvXutZylg061a2XmliROtju61bf37W1rRvVHf2Wfb/F69rBv3vffa39n9G955p12/NWKEpS7ff9/qmpBgv/c337TPWr/elp81y/7upVFpcvwi8gSAIQByAMQDOAnAJ6pa5Gk6Bn6qDHJybEeRm2tdU4u6+nrXLl8PPVULDk2aWJBLTray9HQLmP6jZQOW4//mGwtcfftaV/KTT7ag17ixBabsbLsotndv21FMm2Y7qFNPtZ3ZkCG+cw8bN9oJ/ltusfMSb75p67jmGptfv74Fos8/t3MtPXvazisz0wLfsWMWeL/4wroVi/hGGq9Rw3aWTz9tO8nJk62O//iH3bNo61bb4boXw86ebTukMWOsHps32zYePOgLmjNmWP3dC75vvNE+Oy3NOhR06GDbMWKEdXFft87e17Wr7cwPHrQdLmDT7gXghQ0YYN/b8eMn/pvHxdl35H8fJbfLdkVo1Mi6WTdpUrr3V5rAX2DlbPETVSlHj9oOoVkz2wEWN3qHy71ZnIjvKCIvz9cR4OefrdztubZvnwVct8t7MG7Cxv1M1/HjtnM74ww78oiNtc+Ji7NraebNsyOynTttumtX62zQpIntpN2u9ps22WfUqGGf17Wr7SwGDLAdapMmdkS5YIF99oABtv7sbOts8Mordl3Z2rV21Nizp73/6qvte3z0Udsx3nijdQhwr9/q3duOynJz7WiwYcMS/5nyMfATEUWZogJ/RG+9qKpzAMyJZB2IiKINLzUhIooyDPxERFGGgZ+IKMow8BMRRRkGfiKiKMPAT0QUZRj4iYiiTJW4A5eI7AawpZRvbwhgT7FLeQu3OTpwm6NDWbb5VFUNGFykSgT+shCR1GBXrnkZtzk6cJujQzi2makeIqIow8BPRBRloiHwT4h0BSKA2xwduM3Rody32fM5fiIiKigaWvxEROSHgZ+IKMp4OvCLSC8R+UVENojIA5GuT3kRkTdFZJeIrPIrqy8iM0VkvfOc6JSLiIx3voOVItIhcjUvHRE5WURmi8jPIrJaREY55Z7dZgAQkXgRWSwiPzrb/ahT3kJEFjnb94GIVHfKazjTG5z5KZGsf2mJSDURWS4iU51pT28vAIhImoj8JCIrRCTVKQvb79uzgV9EqgH4N4DeAM4CMEhEzopsrcrN2wB6FSp7AMAsVT0dwCxnGrDtP915DAfwSgXVsTzlALhbVc8C0AXA7c7f0svbDADHAVykqucCaAegl4h0AfAUgHGqehqA/QBudpa/GcB+p3ycs1xVNArAGr9pr2+v60JVbefXZz98v29V9eQDwPkAvvabfhDAg5GuVzluXwqAVX7TvwBo6rxuCuAX5/V/AAwKtlxVfQD4DMClUbbNNQEsA9AZdhVnrFOe/zsH8DWA853Xsc5yEum6l3A7k50gdxGAqQDEy9vrt91pABoWKgvb79uzLX4AzQFs9ZtOd8q8qrGqbnde7wDQ2Hntqe/BOZxvD2ARomCbnbTHCgC7AMwEsBHAAVXNcRbx37b87XbmHwTQoGJrXGbPA7gPgHN7djSAt7fXpQBmiMhSERnulIXt9x3Re+5SeKiqiojn+umKSG0A/wNwp6oeEpH8eV7dZlXNBdBOROoB+BRA6whXKWxEpC+AXaq6VER6Rro+Faybqm4TkUYAZorIWv+Z5f379nKLfxuAk/2mk50yr9opIk0BwHne5ZR74nsQkThY0H9PVT9xij29zf5U9QCA2bBURz0RcRtt/tuWv93O/LoA9lZwVcuiK4DLRSQNwGRYuucFeHd786nqNud5F2wH3wlh/H17OfAvAXC60yOgOoCBAD6PcJ3C6XMAw5zXw2B5cLd8qNMToAuAg36Hj1WCWNP+DQBrVPU5v1me3WYAEJEkp6UPEUmAnddYA9sBXO0sVni73e/jagDfqpMErgpU9UFVTVbVFNj/67eqOhge3V6XiNQSkTruawB/BLAK4fx9R/qkRphPmPQBsA6WF/1bpOtTjts1CcB2ANmw/N7NsNzmLADrAXwDoL6zrMB6N20E8BOAjpGufym2txssB7oSwArn0cfL2+xsR1sAy53tXgVgjFPeEsBiABsAfASghlMe70xvcOa3jPQ2lGHbewKYGg3b62zfj85jtRurwvn75pANRERRxsupHiIiCoKBn4goyjDwExFFGQZ+IqIow8BPRBRlGPgpqolIrjMiovsot1FcRSRF/EZQJaosOGQDRbujqtou0pUgqkhs8RMF4YyP/i9njPTFInKaU54iIt8646DPEpFTnPLGIvKpM3b+jyJygfNR1UTkNWc8/RnOFbgQkZFi9xdYKSKTI7SZFKUY+CnaJRRK9VznN++gqrYB8BJs1EgAeBHARFVtC+A9AOOd8vEA5qqNnd8BdgUmYGOm/1tVzwZwAMBVTvkDANo7n3NLuDaOKBheuUtRTUQyVLV2kPI02E1QNjkDxO1Q1QYisgc29nm2U75dVRuKyG4Ayap63O8zUgDMVLuRBkTkfgBxqvpPEZkOIAPAFABTVDUjzJtKlI8tfqKiaRGvS+K43+tc+M6rXQYbb6UDgCV+o08ShR0DP1HRrvN7/sF5/T1s5EgAGAxgvvN6FoBbgfybp9Qt6kNFJAbAyao6G8D9sOGEA446iMKFrQyKdgnOHa5c01XV7dKZKCIrYa32QU7ZHQDeEpF7AewGcKNTPgrABBG5GdayvxU2gmow1QC86+wcBMB4tfH2iSoEc/xEQTg5/o6quifSdSEqb0z1EBFFGbb4iYiiDFv8RERRhoGfiCjKMPATEUUZBn4ioijDwE9EFGX+P0W62lxMpIDfAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 0 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VbbVgZrtIEqu",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}