Illinois_COVID_19_Curve_ML_Model.ipynb

{
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "2BRlzpAW5-28",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Run on TensorFlow 2.x\n",
        "%tensorflow_version 2.x\n",
        "from __future__ import absolute_import, division, print_function, unicode_literals"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "q2Fvm5d_6GmW",
        "colab_type": "code",
        "cellView": "both",
        "outputId": "de9ac1ca-4bfe-4d58-fbb6-be59f70cd9c0",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "# Import relevant modules\n",
        "import json, requests\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "from matplotlib import pyplot as plt\n",
        "import tensorflow as tf\n",
        "from tensorflow import keras\n",
        "from sklearn.preprocessing import PolynomialFeatures\n",
        "\n",
        "# The following lines adjust the granularity of reporting. \n",
        "pd.options.display.max_rows = 10\n",
        "pd.options.display.float_format = \"{:.1f}\".format\n",
        "\n",
        "print(\"Imported modules.\")"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Imported modules.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "liPuSwpC6f8X",
        "colab_type": "code",
        "outputId": "852e74b6-1a6b-4cdc-c475-14f8bb78d3ca",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "covid_data = json.loads(requests.get('https://www.dph.illinois.gov/sitefiles/COVIDHistoricalTestResults.json?nocache=1').text)\n",
        "stateTestingResults = covid_data[\"state_testing_results\"][\"values\"]\n",
        "\n",
        "indices = np.arange(1, len(stateTestingResults))\n",
        "changeInCasesByDay = []\n",
        "\n",
        "for i in range(1, len(stateTestingResults)):\n",
        "  changeInCases = stateTestingResults[i][\"confirmed_cases\"] - stateTestingResults[i-1][\"confirmed_cases\"]\n",
        "  changeInCasesByDay.append(changeInCases)\n",
        "\n",
        "train_df = pd.DataFrame({\"Day\": indices, \"Change in Cases\": changeInCasesByDay})\n",
        "\n",
        "print(\"Imported data.\")"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Imported data.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RBId0kGVmzss",
        "colab_type": "code",
        "outputId": "e2462991-95bb-41d9-ceb2-ba11592d74ba",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 282
        }
      },
      "source": [
        "# Creating dataset\n",
        "x = list(train_df[\"Day\"])\n",
        "y = list(train_df[\"Change in Cases\"])\n",
        "\n",
        "# Converting to numpy arrays\n",
        "x = np.array(x)\n",
        "y = np.array(y)\n",
        "\n",
        "# Scaling dataset\n",
        "x = x/max(x)\n",
        "y = y/max(y)\n",
        "\n",
        "# Display dataset\n",
        "plt.scatter(x,y)\n",
        "plt.show()\n",
        "\n",
        "print('Created dataset.')"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Created dataset.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sx9oJsCnY7T8",
        "colab_type": "code",
        "outputId": "64dd549b-ca6f-4765-e3e8-e2a6d65382db",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "def build_model(learning_rate):\n",
        "  model = tf.keras.Sequential()\n",
        "  model.add(keras.layers.Dense(units=1, input_shape=[5]))\n",
        "  optimizer = tf.keras.optimizers.Adam(learning_rate = learning_rate)\n",
        "  model.compile(optimizer=optimizer, loss='mean_squared_error')\n",
        "  return model\n",
        "\n",
        "def train_model(model, feature, label, epochs, batch_size):\n",
        "  history = model.fit(x=feature, y=label, batch_size=batch_size, epochs=epochs, verbose=True)\n",
        "  trained_weight = model.get_weights()[0]\n",
        "  trained_bias = model.get_weights()[1]\n",
        "  hist_loss = history.history['loss']\n",
        "  return hist_loss, trained_weight, trained_bias\n",
        "\n",
        "print('Created functions that build and train model.')"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Created functions that build and train model.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wT5cbPEWog2l",
        "colab_type": "code",
        "outputId": "72cf341e-d6cd-49b1-d02f-2fe3d53f1123",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "def plot_model_loss(hist_loss):\n",
        "  plt.plot(hist_loss)\n",
        "  plt.xlabel('Epochs')\n",
        "  plt.ylabel('MSE Loss')\n",
        "  plt.title('MSE Loss vs Epochs Run')\n",
        "  plt.show()\n",
        "\n",
        "def plot_model(hist_loss, x_poly):\n",
        "  mse = hist_loss[-1]\n",
        "  y_hat = model.predict(x_poly)\n",
        "\n",
        "  plt.figure(figsize=(12,7))\n",
        "  plt.title('ML Curve Fit')\n",
        "  plt.scatter(x, y, label='Actual Data')\n",
        "  plt.plot(x, y_hat, color='red', label='Predicted Curve', linewidth=4.0)\n",
        "  plt.xlabel('$Days$', fontsize=20)\n",
        "  plt.ylabel('$Change in Cases$', fontsize=20)\n",
        "  plt.text(0, 0.70, 'MSE = {:.3f}'.format(mse), fontsize=20)\n",
        "  plt.grid(True)\n",
        "  plt.legend(fontsize=20)\n",
        "  plt.show()\n",
        "\n",
        "print('Created functions that plots model and model loss.')"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Created functions that plots model and model loss.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "b4FH5oIOadgb",
        "colab_type": "code",
        "outputId": "8f095f2e-6526-4c08-fba9-2b7067ee1f57",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "poly = PolynomialFeatures(degree=4)\n",
        "x_poly = poly.fit_transform(x.reshape(-1,1))\n",
        "\n",
        "learning_rate = 0.005\n",
        "epochs = 750\n",
        "batch_size = 10\n",
        "\n",
        "model = build_model(learning_rate)\n",
        "\n",
        "hist_loss, trained_weight, trained_bias = train_model(model, x_poly, y, epochs, batch_size)\n",
        "\n",
        "plot_model_loss(hist_loss)\n",
        "plot_model(hist_loss, x_poly)"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.8935\n",
            "Epoch 2/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.6509\n",
            "Epoch 3/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.4442\n",
            "Epoch 4/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.3054\n",
            "Epoch 5/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.2066\n",
            "Epoch 6/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.1411\n",
            "Epoch 7/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.1065\n",
            "Epoch 8/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0858\n",
            "Epoch 9/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0745\n",
            "Epoch 10/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0707\n",
            "Epoch 11/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0691\n",
            "Epoch 12/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0678\n",
            "Epoch 13/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0672\n",
            "Epoch 14/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0666\n",
            "Epoch 15/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0661\n",
            "Epoch 16/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0656\n",
            "Epoch 17/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0650\n",
            "Epoch 18/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0645\n",
            "Epoch 19/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0639\n",
            "Epoch 20/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0634\n",
            "Epoch 21/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0630\n",
            "Epoch 22/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0625\n",
            "Epoch 23/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0622\n",
            "Epoch 24/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0616\n",
            "Epoch 25/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0611\n",
            "Epoch 26/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0607\n",
            "Epoch 27/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0602\n",
            "Epoch 28/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0599\n",
            "Epoch 29/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0594\n",
            "Epoch 30/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0589\n",
            "Epoch 31/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0585\n",
            "Epoch 32/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0580\n",
            "Epoch 33/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0576\n",
            "Epoch 34/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0573\n",
            "Epoch 35/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0568\n",
            "Epoch 36/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0564\n",
            "Epoch 37/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0559\n",
            "Epoch 38/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0555\n",
            "Epoch 39/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0554\n",
            "Epoch 40/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0547\n",
            "Epoch 41/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0542\n",
            "Epoch 42/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0538\n",
            "Epoch 43/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0533\n",
            "Epoch 44/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0530\n",
            "Epoch 45/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0525\n",
            "Epoch 46/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0522\n",
            "Epoch 47/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0516\n",
            "Epoch 48/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0512\n",
            "Epoch 49/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0508\n",
            "Epoch 50/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0503\n",
            "Epoch 51/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0503\n",
            "Epoch 52/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0496\n",
            "Epoch 53/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0491\n",
            "Epoch 54/750\n",
            "9/9 [==============================] - 0s 2ms/step - loss: 0.0487\n",
            "Epoch 55/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0482\n",
            "Epoch 56/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0477\n",
            "Epoch 57/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0473\n",
            "Epoch 58/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0468\n",
            "Epoch 59/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0465\n",
            "Epoch 60/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0461\n",
            "Epoch 61/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0456\n",
            "Epoch 62/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0452\n",
            "Epoch 63/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0447\n",
            "Epoch 64/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0443\n",
            "Epoch 65/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0439\n",
            "Epoch 66/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0434\n",
            "Epoch 67/750\n",
            "9/9 [==============================] - 0s 2ms/step - loss: 0.0430\n",
            "Epoch 68/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0426\n",
            "Epoch 69/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0424\n",
            "Epoch 70/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0418\n",
            "Epoch 71/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0418\n",
            "Epoch 72/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0410\n",
            "Epoch 73/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0407\n",
            "Epoch 74/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0403\n",
            "Epoch 75/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0397\n",
            "Epoch 76/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0394\n",
            "Epoch 77/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0390\n",
            "Epoch 78/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0386\n",
            "Epoch 79/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0383\n",
            "Epoch 80/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0378\n",
            "Epoch 81/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0377\n",
            "Epoch 82/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0370\n",
            "Epoch 83/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0367\n",
            "Epoch 84/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0363\n",
            "Epoch 85/750\n",
            "9/9 [==============================] - 0s 2ms/step - loss: 0.0360\n",
            "Epoch 86/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0354\n",
            "Epoch 87/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0351\n",
            "Epoch 88/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0347\n",
            "Epoch 89/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0343\n",
            "Epoch 90/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0340\n",
            "Epoch 91/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0336\n",
            "Epoch 92/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0333\n",
            "Epoch 93/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0330\n",
            "Epoch 94/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0326\n",
            "Epoch 95/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0322\n",
            "Epoch 96/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0319\n",
            "Epoch 97/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0315\n",
            "Epoch 98/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0314\n",
            "Epoch 99/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0311\n",
            "Epoch 100/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0306\n",
            "Epoch 101/750\n",
            "9/9 [==============================] - 0s 2ms/step - loss: 0.0305\n",
            "Epoch 102/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0299\n",
            "Epoch 103/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0296\n",
            "Epoch 104/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0292\n",
            "Epoch 105/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0289\n",
            "Epoch 106/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0286\n",
            "Epoch 107/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0283\n",
            "Epoch 108/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0280\n",
            "Epoch 109/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0278\n",
            "Epoch 110/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0274\n",
            "Epoch 111/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0273\n",
            "Epoch 112/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0267\n",
            "Epoch 113/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0265\n",
            "Epoch 114/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0263\n",
            "Epoch 115/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0260\n",
            "Epoch 116/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0257\n",
            "Epoch 117/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0255\n",
            "Epoch 118/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0251\n",
            "Epoch 119/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0249\n",
            "Epoch 120/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0246\n",
            "Epoch 121/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0244\n",
            "Epoch 122/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0240\n",
            "Epoch 123/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0238\n",
            "Epoch 124/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0235\n",
            "Epoch 125/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0233\n",
            "Epoch 126/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0230\n",
            "Epoch 127/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0228\n",
            "Epoch 128/750\n",
            "9/9 [==============================] - 0s 972us/step - loss: 0.0225\n",
            "Epoch 129/750\n",
            "9/9 [==============================] - 0s 961us/step - loss: 0.0224\n",
            "Epoch 130/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0220\n",
            "Epoch 131/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0219\n",
            "Epoch 132/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0217\n",
            "Epoch 133/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0213\n",
            "Epoch 134/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0213\n",
            "Epoch 135/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0209\n",
            "Epoch 136/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0211\n",
            "Epoch 137/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0205\n",
            "Epoch 138/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0202\n",
            "Epoch 139/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0201\n",
            "Epoch 140/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0198\n",
            "Epoch 141/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0197\n",
            "Epoch 142/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0195\n",
            "Epoch 143/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0193\n",
            "Epoch 144/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0191\n",
            "Epoch 145/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0190\n",
            "Epoch 146/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0188\n",
            "Epoch 147/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0185\n",
            "Epoch 148/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0183\n",
            "Epoch 149/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0183\n",
            "Epoch 150/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0180\n",
            "Epoch 151/750\n",
            "9/9 [==============================] - 0s 2ms/step - loss: 0.0178\n",
            "Epoch 152/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0176\n",
            "Epoch 153/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0175\n",
            "Epoch 154/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0174\n",
            "Epoch 155/750\n",
            "9/9 [==============================] - 0s 2ms/step - loss: 0.0173\n",
            "Epoch 156/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0170\n",
            "Epoch 157/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0169\n",
            "Epoch 158/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0167\n",
            "Epoch 159/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0165\n",
            "Epoch 160/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0164\n",
            "Epoch 161/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0162\n",
            "Epoch 162/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0161\n",
            "Epoch 163/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0161\n",
            "Epoch 164/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0160\n",
            "Epoch 165/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0156\n",
            "Epoch 166/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0155\n",
            "Epoch 167/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0155\n",
            "Epoch 168/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0153\n",
            "Epoch 169/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0151\n",
            "Epoch 170/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0151\n",
            "Epoch 171/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0150\n",
            "Epoch 172/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0148\n",
            "Epoch 173/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0147\n",
            "Epoch 174/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0147\n",
            "Epoch 175/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0145\n",
            "Epoch 176/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0144\n",
            "Epoch 177/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0144\n",
            "Epoch 178/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0142\n",
            "Epoch 179/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0140\n",
            "Epoch 180/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0139\n",
            "Epoch 181/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0139\n",
            "Epoch 182/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0138\n",
            "Epoch 183/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0136\n",
            "Epoch 184/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0135\n",
            "Epoch 185/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0135\n",
            "Epoch 186/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0135\n",
            "Epoch 187/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0134\n",
            "Epoch 188/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0132\n",
            "Epoch 189/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0131\n",
            "Epoch 190/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0130\n",
            "Epoch 191/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0129\n",
            "Epoch 192/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0129\n",
            "Epoch 193/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0128\n",
            "Epoch 194/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0127\n",
            "Epoch 195/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0126\n",
            "Epoch 196/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0126\n",
            "Epoch 197/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0124\n",
            "Epoch 198/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0126\n",
            "Epoch 199/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0123\n",
            "Epoch 200/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0124\n",
            "Epoch 201/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0122\n",
            "Epoch 202/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0121\n",
            "Epoch 203/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0121\n",
            "Epoch 204/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0121\n",
            "Epoch 205/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0122\n",
            "Epoch 206/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0120\n",
            "Epoch 207/750\n",
            "9/9 [==============================] - 0s 2ms/step - loss: 0.0118\n",
            "Epoch 208/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0118\n",
            "Epoch 209/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0117\n",
            "Epoch 210/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0118\n",
            "Epoch 211/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0116\n",
            "Epoch 212/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0116\n",
            "Epoch 213/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0115\n",
            "Epoch 214/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0115\n",
            "Epoch 215/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0114\n",
            "Epoch 216/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0114\n",
            "Epoch 217/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0113\n",
            "Epoch 218/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0113\n",
            "Epoch 219/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0112\n",
            "Epoch 220/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0113\n",
            "Epoch 221/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0111\n",
            "Epoch 222/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0111\n",
            "Epoch 223/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0111\n",
            "Epoch 224/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0110\n",
            "Epoch 225/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0110\n",
            "Epoch 226/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0110\n",
            "Epoch 227/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0109\n",
            "Epoch 228/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0110\n",
            "Epoch 229/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0110\n",
            "Epoch 230/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0108\n",
            "Epoch 231/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0109\n",
            "Epoch 232/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0108\n",
            "Epoch 233/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0107\n",
            "Epoch 234/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0107\n",
            "Epoch 235/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0107\n",
            "Epoch 236/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0106\n",
            "Epoch 237/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0107\n",
            "Epoch 238/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0107\n",
            "Epoch 239/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0108\n",
            "Epoch 240/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0106\n",
            "Epoch 241/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0105\n",
            "Epoch 242/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0106\n",
            "Epoch 243/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0104\n",
            "Epoch 244/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0104\n",
            "Epoch 245/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0105\n",
            "Epoch 246/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0106\n",
            "Epoch 247/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0106\n",
            "Epoch 248/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0104\n",
            "Epoch 249/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0104\n",
            "Epoch 250/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0104\n",
            "Epoch 251/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0103\n",
            "Epoch 252/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0104\n",
            "Epoch 253/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0104\n",
            "Epoch 254/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0103\n",
            "Epoch 255/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0103\n",
            "Epoch 256/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0103\n",
            "Epoch 257/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0102\n",
            "Epoch 258/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0102\n",
            "Epoch 259/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0102\n",
            "Epoch 260/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0102\n",
            "Epoch 261/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0103\n",
            "Epoch 262/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0105\n",
            "Epoch 263/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0103\n",
            "Epoch 264/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0102\n",
            "Epoch 265/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 266/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 267/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 268/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 269/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 270/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 271/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 272/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 273/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 274/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 275/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 276/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 277/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 278/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 279/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 280/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 281/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 282/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 283/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 284/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 285/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 286/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 287/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 288/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 289/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 290/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 291/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 292/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 293/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 294/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 295/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 296/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 297/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 298/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 299/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 300/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 301/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 302/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 303/750\n",
            "9/9 [==============================] - 0s 2ms/step - loss: 0.0101\n",
            "Epoch 304/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 305/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 306/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 307/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 308/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 309/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 310/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 311/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 312/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 313/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 314/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 315/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 316/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 317/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 318/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 319/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 320/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 321/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 322/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 323/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 324/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 325/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 326/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 327/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 328/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 329/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 330/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 331/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 332/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 333/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 334/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 335/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 336/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 337/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 338/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 339/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 340/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 341/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 342/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 343/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 344/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 345/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 346/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 347/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 348/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 349/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 350/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 351/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 352/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 353/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 354/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 355/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 356/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 357/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 358/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 359/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 360/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 361/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 362/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 363/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 364/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0102\n",
            "Epoch 365/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 366/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 367/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 368/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 369/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 370/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 371/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 372/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 373/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 374/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 375/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 376/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 377/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 378/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 379/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 380/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 381/750\n",
            "9/9 [==============================] - 0s 2ms/step - loss: 0.0098\n",
            "Epoch 382/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 383/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 384/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 385/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 386/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 387/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 388/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0103\n",
            "Epoch 389/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0105\n",
            "Epoch 390/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 391/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 392/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 393/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 394/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 395/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 396/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 397/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 398/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 399/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 400/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 401/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 402/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 403/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 404/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 405/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 406/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 407/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 408/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 409/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0097\n",
            "Epoch 410/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 411/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 412/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 413/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 414/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 415/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 416/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 417/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 418/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 419/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 420/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 421/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 422/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 423/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 424/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 425/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0103\n",
            "Epoch 426/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 427/750\n",
            "9/9 [==============================] - 0s 2ms/step - loss: 0.0099\n",
            "Epoch 428/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 429/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 430/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 431/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 432/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 433/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 434/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0102\n",
            "Epoch 435/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 436/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 437/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 438/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 439/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 440/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0102\n",
            "Epoch 441/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 442/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 443/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 444/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 445/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 446/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0104\n",
            "Epoch 447/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 448/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 449/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 450/750\n",
            "9/9 [==============================] - 0s 2ms/step - loss: 0.0101\n",
            "Epoch 451/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 452/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 453/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 454/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 455/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 456/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 457/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 458/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 459/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 460/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 461/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 462/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 463/750\n",
            "9/9 [==============================] - 0s 2ms/step - loss: 0.0101\n",
            "Epoch 464/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0097\n",
            "Epoch 465/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 466/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 467/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 468/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0102\n",
            "Epoch 469/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 470/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 471/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 472/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 473/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 474/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 475/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 476/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 477/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 478/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 479/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 480/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 481/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 482/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 483/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 484/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 485/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 486/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 487/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 488/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 489/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 490/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 491/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 492/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 493/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 494/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 495/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 496/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 497/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 498/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 499/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 500/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0102\n",
            "Epoch 501/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 502/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 503/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 504/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 505/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 506/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 507/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 508/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 509/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 510/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 511/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0102\n",
            "Epoch 512/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 513/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 514/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0102\n",
            "Epoch 515/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0097\n",
            "Epoch 516/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0102\n",
            "Epoch 517/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 518/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 519/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 520/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 521/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 522/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 523/750\n",
            "9/9 [==============================] - 0s 2ms/step - loss: 0.0099\n",
            "Epoch 524/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 525/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 526/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 527/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 528/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 529/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 530/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 531/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 532/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 533/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 534/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 535/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 536/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 537/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 538/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 539/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 540/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 541/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 542/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 543/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 544/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 545/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 546/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 547/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 548/750\n",
            "9/9 [==============================] - 0s 965us/step - loss: 0.0098\n",
            "Epoch 549/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 550/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 551/750\n",
            "9/9 [==============================] - 0s 979us/step - loss: 0.0098\n",
            "Epoch 552/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 553/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 554/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 555/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0103\n",
            "Epoch 556/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0097\n",
            "Epoch 557/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 558/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0102\n",
            "Epoch 559/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0104\n",
            "Epoch 560/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0102\n",
            "Epoch 561/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0102\n",
            "Epoch 562/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 563/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 564/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 565/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 566/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 567/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 568/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 569/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 570/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 571/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 572/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 573/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 574/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 575/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 576/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 577/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 578/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 579/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 580/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 581/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 582/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 583/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 584/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 585/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 586/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 587/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 588/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 589/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 590/750\n",
            "9/9 [==============================] - 0s 944us/step - loss: 0.0101\n",
            "Epoch 591/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 592/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 593/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 594/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 595/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 596/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 597/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 598/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 599/750\n",
            "9/9 [==============================] - 0s 2ms/step - loss: 0.0100\n",
            "Epoch 600/750\n",
            "9/9 [==============================] - 0s 2ms/step - loss: 0.0101\n",
            "Epoch 601/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0097\n",
            "Epoch 602/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 603/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 604/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 605/750\n",
            "9/9 [==============================] - 0s 942us/step - loss: 0.0102\n",
            "Epoch 606/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 607/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 608/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 609/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 610/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 611/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 612/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 613/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0102\n",
            "Epoch 614/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0097\n",
            "Epoch 615/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 616/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 617/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 618/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 619/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 620/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 621/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 622/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 623/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 624/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 625/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 626/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 627/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 628/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 629/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 630/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 631/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 632/750\n",
            "9/9 [==============================] - 0s 3ms/step - loss: 0.0099\n",
            "Epoch 633/750\n",
            "9/9 [==============================] - 0s 2ms/step - loss: 0.0099\n",
            "Epoch 634/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 635/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 636/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 637/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0102\n",
            "Epoch 638/750\n",
            "9/9 [==============================] - 0s 2ms/step - loss: 0.0098\n",
            "Epoch 639/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 640/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 641/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 642/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 643/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 644/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 645/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 646/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 647/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 648/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0102\n",
            "Epoch 649/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 650/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 651/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 652/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 653/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 654/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 655/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0097\n",
            "Epoch 656/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 657/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 658/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 659/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 660/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 661/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 662/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 663/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 664/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0102\n",
            "Epoch 665/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 666/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 667/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 668/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 669/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 670/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 671/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 672/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 673/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 674/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 675/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 676/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 677/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 678/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 679/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 680/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 681/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 682/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 683/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 684/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 685/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 686/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 687/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 688/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 689/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 690/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0097\n",
            "Epoch 691/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 692/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 693/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 694/750\n",
            "9/9 [==============================] - 0s 2ms/step - loss: 0.0098\n",
            "Epoch 695/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 696/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 697/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 698/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 699/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 700/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 701/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 702/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 703/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0102\n",
            "Epoch 704/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 705/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 706/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 707/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 708/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 709/750\n",
            "9/9 [==============================] - 0s 2ms/step - loss: 0.0102\n",
            "Epoch 710/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 711/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 712/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 713/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 714/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0106\n",
            "Epoch 715/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 716/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 717/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 718/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 719/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 720/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 721/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 722/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 723/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0105\n",
            "Epoch 724/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 725/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0102\n",
            "Epoch 726/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 727/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 728/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 729/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 730/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 731/750\n",
            "9/9 [==============================] - 0s 2ms/step - loss: 0.0099\n",
            "Epoch 732/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 733/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 734/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 735/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 736/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 737/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 738/750\n",
            "9/9 [==============================] - 0s 2ms/step - loss: 0.0100\n",
            "Epoch 739/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 740/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 741/750\n",
            "9/9 [==============================] - 0s 2ms/step - loss: 0.0098\n",
            "Epoch 742/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n",
            "Epoch 743/750\n",
            "9/9 [==============================] - 0s 2ms/step - loss: 0.0102\n",
            "Epoch 744/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0102\n",
            "Epoch 745/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0097\n",
            "Epoch 746/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0101\n",
            "Epoch 747/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 748/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0098\n",
            "Epoch 749/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0100\n",
            "Epoch 750/750\n",
            "9/9 [==============================] - 0s 1ms/step - loss: 0.0099\n"
          ],
          "name": "stdout"
        },
        {
          "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": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 864x504 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    }
  ],
  "metadata": {
    "colab": {
      "name": "Illinois COVID-19 Curve ML Model.ipynb",
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}