tutorial113_what_are_autoencoders.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "tutorial113_what_are_autoencoders.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "mount_file_id": "1qRaljciUK0ld8tiPgovZkoA6HwbckI4F",
      "authorship_tag": "ABX9TyMhvbyqEmxLQy8xPB7goIaX",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/bnsreenu/python_for_image_processing_APEER/blob/master/tutorial113_what_are_autoencoders.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DENbQ6cPgfxC"
      },
      "source": [
        "Single image reconstruction using autoencoders.\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GKioYAey1ykd"
      },
      "source": [
        "https://youtu.be/kpnGB6Y89OY"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EyUha4-qf7Rk"
      },
      "source": [
        "from matplotlib import pyplot as plt\n",
        "import numpy as np\n",
        "import cv2\n",
        "from tensorflow.keras.preprocessing.image import img_to_array\n",
        "from tensorflow.keras.layers import Conv2D, MaxPooling2D, UpSampling2D\n",
        "from tensorflow.keras.models import Sequential, Model\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "W7O7AQhPg_v4"
      },
      "source": [
        "SIZE=256"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "StM1xAEqg0gZ",
        "outputId": "dea86e1c-ccb8-40be-aec8-c836768ab7c9"
      },
      "source": [
        "img=cv2.imread('/content/drive/MyDrive/Colab Notebooks/data/monalisa.jpg', 1)   #Change 1 to 0 for grey images\n",
        "img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB) #Changing BGR to RGB to show images in true colors\n",
        "img=cv2.resize(img,(SIZE, SIZE))\n",
        "print(img.shape)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(256, 256, 3)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gnGvoeuogyJw",
        "outputId": "e0f52ef9-7f61-4115-c3d0-6cc75f602eb0"
      },
      "source": [
        "img_array = np.expand_dims(img, axis=0)\n",
        "print(img_array.shape)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(1, 256, 256, 3)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "F9ntktIriLCT"
      },
      "source": [
        "img_array = img_array.astype('float32') / 255.\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mcNkWUD6iLi5",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "3ba990f4-1d92-4398-98bc-ed409200adcd"
      },
      "source": [
        "model = Sequential()\n",
        "model.add(Conv2D(64, (3, 3), activation='relu', padding='same', input_shape=(SIZE, SIZE, 3)))\n",
        "model.add(MaxPooling2D((2, 2), padding='same'))\n",
        "model.add(Conv2D(128, (3, 3), activation='relu', padding='same'))\n",
        "model.add(MaxPooling2D((2, 2), padding='same'))\n",
        "model.add(Conv2D(256, (3, 3), activation='relu', padding='same'))\n",
        " \n",
        "\n",
        "model.add(MaxPooling2D((2, 2), padding='same'))\n",
        "     \n",
        "model.add(Conv2D(256, (3, 3), activation='relu', padding='same'))\n",
        "model.add(UpSampling2D((2, 2)))\n",
        "model.add(Conv2D(128, (3, 3), activation='relu', padding='same'))\n",
        "model.add(UpSampling2D((2, 2)))\n",
        "model.add(Conv2D(64, (3, 3), activation='relu', padding='same'))\n",
        "model.add(UpSampling2D((2, 2)))\n",
        "\n",
        "model.add(Conv2D(3, (3, 3), activation='sigmoid', padding='same'))\n",
        "\n",
        "model.compile(optimizer='adam', loss='mean_squared_error', metrics=['accuracy'])\n",
        "model.summary()\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d (Conv2D)              (None, 256, 256, 64)      1792      \n",
            "_________________________________________________________________\n",
            "max_pooling2d (MaxPooling2D) (None, 128, 128, 64)      0         \n",
            "_________________________________________________________________\n",
            "conv2d_1 (Conv2D)            (None, 128, 128, 128)     73856     \n",
            "_________________________________________________________________\n",
            "max_pooling2d_1 (MaxPooling2 (None, 64, 64, 128)       0         \n",
            "_________________________________________________________________\n",
            "conv2d_2 (Conv2D)            (None, 64, 64, 256)       295168    \n",
            "_________________________________________________________________\n",
            "max_pooling2d_2 (MaxPooling2 (None, 32, 32, 256)       0         \n",
            "_________________________________________________________________\n",
            "conv2d_3 (Conv2D)            (None, 32, 32, 256)       590080    \n",
            "_________________________________________________________________\n",
            "up_sampling2d (UpSampling2D) (None, 64, 64, 256)       0         \n",
            "_________________________________________________________________\n",
            "conv2d_4 (Conv2D)            (None, 64, 64, 128)       295040    \n",
            "_________________________________________________________________\n",
            "up_sampling2d_1 (UpSampling2 (None, 128, 128, 128)     0         \n",
            "_________________________________________________________________\n",
            "conv2d_5 (Conv2D)            (None, 128, 128, 64)      73792     \n",
            "_________________________________________________________________\n",
            "up_sampling2d_2 (UpSampling2 (None, 256, 256, 64)      0         \n",
            "_________________________________________________________________\n",
            "conv2d_6 (Conv2D)            (None, 256, 256, 3)       1731      \n",
            "=================================================================\n",
            "Total params: 1,331,459\n",
            "Trainable params: 1,331,459\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Q3GOnXTpisQx",
        "outputId": "cfd4b894-1c5b-481f-e7bf-3712e9186735"
      },
      "source": [
        "model.fit(img_array, img_array, epochs=1000)\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/1000\n",
            "1/1 [==============================] - 32s 32s/step - loss: 0.0694 - accuracy: 0.2915\n",
            "Epoch 2/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0592 - accuracy: 0.2481\n",
            "Epoch 3/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0519 - accuracy: 0.2500\n",
            "Epoch 4/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0450 - accuracy: 0.5941\n",
            "Epoch 5/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0445 - accuracy: 0.5955\n",
            "Epoch 6/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0403 - accuracy: 0.5997\n",
            "Epoch 7/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0370 - accuracy: 0.5960\n",
            "Epoch 8/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0315 - accuracy: 0.5985\n",
            "Epoch 9/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0243 - accuracy: 0.6073\n",
            "Epoch 10/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0180 - accuracy: 0.5965\n",
            "Epoch 11/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0154 - accuracy: 0.6042\n",
            "Epoch 12/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0150 - accuracy: 0.6057\n",
            "Epoch 13/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0154 - accuracy: 0.5993\n",
            "Epoch 14/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0149 - accuracy: 0.6390\n",
            "Epoch 15/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0137 - accuracy: 0.6374\n",
            "Epoch 16/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0140 - accuracy: 0.6357\n",
            "Epoch 17/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0162 - accuracy: 0.5368\n",
            "Epoch 18/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0127 - accuracy: 0.6319\n",
            "Epoch 19/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0110 - accuracy: 0.6397\n",
            "Epoch 20/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0112 - accuracy: 0.6161\n",
            "Epoch 21/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0096 - accuracy: 0.6068\n",
            "Epoch 22/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0097 - accuracy: 0.6072\n",
            "Epoch 23/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0096 - accuracy: 0.6084\n",
            "Epoch 24/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0073 - accuracy: 0.6340\n",
            "Epoch 25/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0100 - accuracy: 0.6542\n",
            "Epoch 26/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0175 - accuracy: 0.6140\n",
            "Epoch 27/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0070 - accuracy: 0.5953\n",
            "Epoch 28/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0145 - accuracy: 0.4370\n",
            "Epoch 29/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0079 - accuracy: 0.6161\n",
            "Epoch 30/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0120 - accuracy: 0.6035\n",
            "Epoch 31/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0090 - accuracy: 0.6034\n",
            "Epoch 32/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0075 - accuracy: 0.6032\n",
            "Epoch 33/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0104 - accuracy: 0.6032\n",
            "Epoch 34/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0076 - accuracy: 0.6032\n",
            "Epoch 35/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0084 - accuracy: 0.6032\n",
            "Epoch 36/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0092 - accuracy: 0.6031\n",
            "Epoch 37/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0071 - accuracy: 0.6031\n",
            "Epoch 38/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 0.0072 - accuracy: 0.6032\n",
            "Epoch 39/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0078 - accuracy: 0.6035\n",
            "Epoch 40/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0063 - accuracy: 0.6050\n",
            "Epoch 41/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0067 - accuracy: 0.6046\n",
            "Epoch 42/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0069 - accuracy: 0.6045\n",
            "Epoch 43/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0058 - accuracy: 0.6046\n",
            "Epoch 44/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0062 - accuracy: 0.6042\n",
            "Epoch 45/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0060 - accuracy: 0.6045\n",
            "Epoch 46/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0056 - accuracy: 0.6047\n",
            "Epoch 47/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0060 - accuracy: 0.6049\n",
            "Epoch 48/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0049 - accuracy: 0.6047\n",
            "Epoch 49/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0055 - accuracy: 0.6042\n",
            "Epoch 50/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0046 - accuracy: 0.6050\n",
            "Epoch 51/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0049 - accuracy: 0.6060\n",
            "Epoch 52/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0044 - accuracy: 0.6071\n",
            "Epoch 53/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0047 - accuracy: 0.6120\n",
            "Epoch 54/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0041 - accuracy: 0.6116\n",
            "Epoch 55/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0043 - accuracy: 0.6112\n",
            "Epoch 56/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0038 - accuracy: 0.6151\n",
            "Epoch 57/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0042 - accuracy: 0.6376\n",
            "Epoch 58/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 0.0036 - accuracy: 0.6364\n",
            "Epoch 59/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0038 - accuracy: 0.6446\n",
            "Epoch 60/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0035 - accuracy: 0.6702\n",
            "Epoch 61/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0034 - accuracy: 0.6718\n",
            "Epoch 62/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0035 - accuracy: 0.6756\n",
            "Epoch 63/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0032 - accuracy: 0.6777\n",
            "Epoch 64/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0033 - accuracy: 0.6798\n",
            "Epoch 65/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0032 - accuracy: 0.6687\n",
            "Epoch 66/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0030 - accuracy: 0.6935\n",
            "Epoch 67/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0031 - accuracy: 0.7423\n",
            "Epoch 68/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0031 - accuracy: 0.8014\n",
            "Epoch 69/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0028 - accuracy: 0.8248\n",
            "Epoch 70/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0028 - accuracy: 0.8293\n",
            "Epoch 71/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0029 - accuracy: 0.8474\n",
            "Epoch 72/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0028 - accuracy: 0.8556\n",
            "Epoch 73/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0026 - accuracy: 0.8572\n",
            "Epoch 74/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0026 - accuracy: 0.8576\n",
            "Epoch 75/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0027 - accuracy: 0.8564\n",
            "Epoch 76/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0027 - accuracy: 0.8602\n",
            "Epoch 77/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0026 - accuracy: 0.8533\n",
            "Epoch 78/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 0.0025 - accuracy: 0.8598\n",
            "Epoch 79/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0025 - accuracy: 0.8597\n",
            "Epoch 80/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0025 - accuracy: 0.8616\n",
            "Epoch 81/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0025 - accuracy: 0.8535\n",
            "Epoch 82/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0024 - accuracy: 0.8575\n",
            "Epoch 83/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0023 - accuracy: 0.8519\n",
            "Epoch 84/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0024 - accuracy: 0.8404\n",
            "Epoch 85/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0024 - accuracy: 0.8545\n",
            "Epoch 86/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0024 - accuracy: 0.8460\n",
            "Epoch 87/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0023 - accuracy: 0.8520\n",
            "Epoch 88/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0022 - accuracy: 0.8487\n",
            "Epoch 89/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0022 - accuracy: 0.8469\n",
            "Epoch 90/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0023 - accuracy: 0.8548\n",
            "Epoch 91/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 0.0023 - accuracy: 0.8487\n",
            "Epoch 92/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0024 - accuracy: 0.8583\n",
            "Epoch 93/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0024 - accuracy: 0.8456\n",
            "Epoch 94/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0024 - accuracy: 0.8590\n",
            "Epoch 95/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0024 - accuracy: 0.8497\n",
            "Epoch 96/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0022 - accuracy: 0.8589\n",
            "Epoch 97/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0021 - accuracy: 0.8561\n",
            "Epoch 98/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0021 - accuracy: 0.8584\n",
            "Epoch 99/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0021 - accuracy: 0.8610\n",
            "Epoch 100/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0022 - accuracy: 0.8568\n",
            "Epoch 101/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0022 - accuracy: 0.8591\n",
            "Epoch 102/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0021 - accuracy: 0.8570\n",
            "Epoch 103/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0020 - accuracy: 0.8598\n",
            "Epoch 104/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0020 - accuracy: 0.8592\n",
            "Epoch 105/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0020 - accuracy: 0.8593\n",
            "Epoch 106/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0020 - accuracy: 0.8592\n",
            "Epoch 107/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0020 - accuracy: 0.8578\n",
            "Epoch 108/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0022 - accuracy: 0.8608\n",
            "Epoch 109/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0024 - accuracy: 0.8525\n",
            "Epoch 110/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0028 - accuracy: 0.8580\n",
            "Epoch 111/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0028 - accuracy: 0.8463\n",
            "Epoch 112/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0021 - accuracy: 0.8590\n",
            "Epoch 113/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0020 - accuracy: 0.8596\n",
            "Epoch 114/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0024 - accuracy: 0.8308\n",
            "Epoch 115/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0021 - accuracy: 0.8602\n",
            "Epoch 116/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0020 - accuracy: 0.8589\n",
            "Epoch 117/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0022 - accuracy: 0.8443\n",
            "Epoch 118/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0019 - accuracy: 0.8570\n",
            "Epoch 119/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0020 - accuracy: 0.8579\n",
            "Epoch 120/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0021 - accuracy: 0.8530\n",
            "Epoch 121/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0019 - accuracy: 0.8593\n",
            "Epoch 122/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0020 - accuracy: 0.8596\n",
            "Epoch 123/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0020 - accuracy: 0.8589\n",
            "Epoch 124/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0018 - accuracy: 0.8646\n",
            "Epoch 125/1000\n",
            "1/1 [==============================] - 0s 24ms/step - loss: 0.0020 - accuracy: 0.8655\n",
            "Epoch 126/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 0.0019 - accuracy: 0.8660\n",
            "Epoch 127/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 0.0018 - accuracy: 0.8686\n",
            "Epoch 128/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0019 - accuracy: 0.8671\n",
            "Epoch 129/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0019 - accuracy: 0.8685\n",
            "Epoch 130/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0017 - accuracy: 0.8684\n",
            "Epoch 131/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0018 - accuracy: 0.8684\n",
            "Epoch 132/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0018 - accuracy: 0.8701\n",
            "Epoch 133/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0018 - accuracy: 0.8672\n",
            "Epoch 134/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0017 - accuracy: 0.8707\n",
            "Epoch 135/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0017 - accuracy: 0.8724\n",
            "Epoch 136/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0017 - accuracy: 0.8719\n",
            "Epoch 137/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0018 - accuracy: 0.8722\n",
            "Epoch 138/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0018 - accuracy: 0.8727\n",
            "Epoch 139/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0017 - accuracy: 0.8745\n",
            "Epoch 140/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 0.0017 - accuracy: 0.8725\n",
            "Epoch 141/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0016 - accuracy: 0.8730\n",
            "Epoch 142/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0016 - accuracy: 0.8741\n",
            "Epoch 143/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0017 - accuracy: 0.8735\n",
            "Epoch 144/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0017 - accuracy: 0.8748\n",
            "Epoch 145/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0018 - accuracy: 0.8737\n",
            "Epoch 146/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0020 - accuracy: 0.8745\n",
            "Epoch 147/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0021 - accuracy: 0.8712\n",
            "Epoch 148/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0022 - accuracy: 0.8726\n",
            "Epoch 149/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0019 - accuracy: 0.8750\n",
            "Epoch 150/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0016 - accuracy: 0.8769\n",
            "Epoch 151/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0017 - accuracy: 0.8745\n",
            "Epoch 152/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0018 - accuracy: 0.8675\n",
            "Epoch 153/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0016 - accuracy: 0.8747\n",
            "Epoch 154/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0016 - accuracy: 0.8750\n",
            "Epoch 155/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0018 - accuracy: 0.8724\n",
            "Epoch 156/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0016 - accuracy: 0.8748\n",
            "Epoch 157/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0016 - accuracy: 0.8769\n",
            "Epoch 158/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0017 - accuracy: 0.8789\n",
            "Epoch 159/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0016 - accuracy: 0.8776\n",
            "Epoch 160/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0015 - accuracy: 0.8786\n",
            "Epoch 161/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0016 - accuracy: 0.8822\n",
            "Epoch 162/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0016 - accuracy: 0.8818\n",
            "Epoch 163/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0015 - accuracy: 0.8764\n",
            "Epoch 164/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0015 - accuracy: 0.8818\n",
            "Epoch 165/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0016 - accuracy: 0.8826\n",
            "Epoch 166/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 0.0016 - accuracy: 0.8759\n",
            "Epoch 167/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0015 - accuracy: 0.8821\n",
            "Epoch 168/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0015 - accuracy: 0.8847\n",
            "Epoch 169/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0015 - accuracy: 0.8820\n",
            "Epoch 170/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0015 - accuracy: 0.8816\n",
            "Epoch 171/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0016 - accuracy: 0.8856\n",
            "Epoch 172/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0016 - accuracy: 0.8852\n",
            "Epoch 173/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0015 - accuracy: 0.8802\n",
            "Epoch 174/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0015 - accuracy: 0.8856\n",
            "Epoch 175/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0014 - accuracy: 0.8854\n",
            "Epoch 176/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0014 - accuracy: 0.8815\n",
            "Epoch 177/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0014 - accuracy: 0.8858\n",
            "Epoch 178/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0015 - accuracy: 0.8859\n",
            "Epoch 179/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0015 - accuracy: 0.8839\n",
            "Epoch 180/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0015 - accuracy: 0.8869\n",
            "Epoch 181/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0016 - accuracy: 0.8860\n",
            "Epoch 182/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0017 - accuracy: 0.8819\n",
            "Epoch 183/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0018 - accuracy: 0.8820\n",
            "Epoch 184/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0017 - accuracy: 0.8835\n",
            "Epoch 185/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0015 - accuracy: 0.8840\n",
            "Epoch 186/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 0.0014 - accuracy: 0.8858\n",
            "Epoch 187/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0015 - accuracy: 0.8810\n",
            "Epoch 188/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0016 - accuracy: 0.8840\n",
            "Epoch 189/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0014 - accuracy: 0.8858\n",
            "Epoch 190/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0014 - accuracy: 0.8875\n",
            "Epoch 191/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0015 - accuracy: 0.8858\n",
            "Epoch 192/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0015 - accuracy: 0.8845\n",
            "Epoch 193/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0014 - accuracy: 0.8892\n",
            "Epoch 194/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0014 - accuracy: 0.8875\n",
            "Epoch 195/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 0.0015 - accuracy: 0.8867\n",
            "Epoch 196/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0014 - accuracy: 0.8892\n",
            "Epoch 197/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0014 - accuracy: 0.8865\n",
            "Epoch 198/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0014 - accuracy: 0.8885\n",
            "Epoch 199/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0014 - accuracy: 0.8901\n",
            "Epoch 200/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0014 - accuracy: 0.8846\n",
            "Epoch 201/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0014 - accuracy: 0.8904\n",
            "Epoch 202/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0014 - accuracy: 0.8906\n",
            "Epoch 203/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0013 - accuracy: 0.8870\n",
            "Epoch 204/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0013 - accuracy: 0.8910\n",
            "Epoch 205/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 0.0013 - accuracy: 0.8900\n",
            "Epoch 206/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0014 - accuracy: 0.8894\n",
            "Epoch 207/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0014 - accuracy: 0.8936\n",
            "Epoch 208/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0014 - accuracy: 0.8891\n",
            "Epoch 209/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0015 - accuracy: 0.8907\n",
            "Epoch 210/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0015 - accuracy: 0.8920\n",
            "Epoch 211/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 0.0015 - accuracy: 0.8870\n",
            "Epoch 212/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0015 - accuracy: 0.8912\n",
            "Epoch 213/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0014 - accuracy: 0.8894\n",
            "Epoch 214/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0013 - accuracy: 0.8898\n",
            "Epoch 215/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0013 - accuracy: 0.8928\n",
            "Epoch 216/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0014 - accuracy: 0.8888\n",
            "Epoch 217/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0014 - accuracy: 0.8924\n",
            "Epoch 218/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 0.0013 - accuracy: 0.8920\n",
            "Epoch 219/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0013 - accuracy: 0.8912\n",
            "Epoch 220/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0013 - accuracy: 0.8945\n",
            "Epoch 221/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0013 - accuracy: 0.8922\n",
            "Epoch 222/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0013 - accuracy: 0.8937\n",
            "Epoch 223/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 0.0013 - accuracy: 0.8954\n",
            "Epoch 224/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0013 - accuracy: 0.8939\n",
            "Epoch 225/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0013 - accuracy: 0.8955\n",
            "Epoch 226/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0013 - accuracy: 0.8950\n",
            "Epoch 227/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0013 - accuracy: 0.8962\n",
            "Epoch 228/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0013 - accuracy: 0.8958\n",
            "Epoch 229/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0013 - accuracy: 0.8958\n",
            "Epoch 230/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0013 - accuracy: 0.8956\n",
            "Epoch 231/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0013 - accuracy: 0.8962\n",
            "Epoch 232/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0013 - accuracy: 0.8945\n",
            "Epoch 233/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0013 - accuracy: 0.8971\n",
            "Epoch 234/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0013 - accuracy: 0.8935\n",
            "Epoch 235/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0013 - accuracy: 0.8975\n",
            "Epoch 236/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0012 - accuracy: 0.8953\n",
            "Epoch 237/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0012 - accuracy: 0.8985\n",
            "Epoch 238/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0012 - accuracy: 0.8977\n",
            "Epoch 239/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0012 - accuracy: 0.8979\n",
            "Epoch 240/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0013 - accuracy: 0.8983\n",
            "Epoch 241/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0013 - accuracy: 0.8978\n",
            "Epoch 242/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0013 - accuracy: 0.8980\n",
            "Epoch 243/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0014 - accuracy: 0.8991\n",
            "Epoch 244/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 0.0014 - accuracy: 0.8952\n",
            "Epoch 245/1000\n",
            "1/1 [==============================] - 0s 27ms/step - loss: 0.0015 - accuracy: 0.8991\n",
            "Epoch 246/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0015 - accuracy: 0.8911\n",
            "Epoch 247/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0013 - accuracy: 0.8984\n",
            "Epoch 248/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0012 - accuracy: 0.8971\n",
            "Epoch 249/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0013 - accuracy: 0.8907\n",
            "Epoch 250/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0013 - accuracy: 0.8985\n",
            "Epoch 251/1000\n",
            "1/1 [==============================] - 0s 23ms/step - loss: 0.0013 - accuracy: 0.8958\n",
            "Epoch 252/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0012 - accuracy: 0.8905\n",
            "Epoch 253/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 0.0013 - accuracy: 0.8807\n",
            "Epoch 254/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 0.0013 - accuracy: 0.8829\n",
            "Epoch 255/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0012 - accuracy: 0.8980\n",
            "Epoch 256/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0012 - accuracy: 0.8942\n",
            "Epoch 257/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0012 - accuracy: 0.8866\n",
            "Epoch 258/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0012 - accuracy: 0.8986\n",
            "Epoch 259/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0012 - accuracy: 0.9002\n",
            "Epoch 260/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0012 - accuracy: 0.8958\n",
            "Epoch 261/1000\n",
            "1/1 [==============================] - 0s 23ms/step - loss: 0.0012 - accuracy: 0.8974\n",
            "Epoch 262/1000\n",
            "1/1 [==============================] - 0s 22ms/step - loss: 0.0012 - accuracy: 0.8943\n",
            "Epoch 263/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0013 - accuracy: 0.8852\n",
            "Epoch 264/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0013 - accuracy: 0.8970\n",
            "Epoch 265/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0013 - accuracy: 0.9004\n",
            "Epoch 266/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0013 - accuracy: 0.8994\n",
            "Epoch 267/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0013 - accuracy: 0.8988\n",
            "Epoch 268/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0013 - accuracy: 0.8966\n",
            "Epoch 269/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0013 - accuracy: 0.8864\n",
            "Epoch 270/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0012 - accuracy: 0.8996\n",
            "Epoch 271/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0012 - accuracy: 0.9001\n",
            "Epoch 272/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0012 - accuracy: 0.8933\n",
            "Epoch 273/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0012 - accuracy: 0.9000\n",
            "Epoch 274/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0012 - accuracy: 0.9014\n",
            "Epoch 275/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0012 - accuracy: 0.8984\n",
            "Epoch 276/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0012 - accuracy: 0.9008\n",
            "Epoch 277/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0012 - accuracy: 0.9011\n",
            "Epoch 278/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0012 - accuracy: 0.8984\n",
            "Epoch 279/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0012 - accuracy: 0.9009\n",
            "Epoch 280/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 0.0012 - accuracy: 0.9007\n",
            "Epoch 281/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 0.0012 - accuracy: 0.8980\n",
            "Epoch 282/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0012 - accuracy: 0.9008\n",
            "Epoch 283/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0012 - accuracy: 0.9023\n",
            "Epoch 284/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0012 - accuracy: 0.9008\n",
            "Epoch 285/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0012 - accuracy: 0.9025\n",
            "Epoch 286/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0012 - accuracy: 0.9010\n",
            "Epoch 287/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0012 - accuracy: 0.8999\n",
            "Epoch 288/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0012 - accuracy: 0.9005\n",
            "Epoch 289/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0012 - accuracy: 0.9015\n",
            "Epoch 290/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0012 - accuracy: 0.9018\n",
            "Epoch 291/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0011 - accuracy: 0.9025\n",
            "Epoch 292/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 0.0011 - accuracy: 0.9021\n",
            "Epoch 293/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0011 - accuracy: 0.9032\n",
            "Epoch 294/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0011 - accuracy: 0.9027\n",
            "Epoch 295/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0011 - accuracy: 0.9027\n",
            "Epoch 296/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0011 - accuracy: 0.9035\n",
            "Epoch 297/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0011 - accuracy: 0.9036\n",
            "Epoch 298/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0011 - accuracy: 0.9036\n",
            "Epoch 299/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0011 - accuracy: 0.9022\n",
            "Epoch 300/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0012 - accuracy: 0.9042\n",
            "Epoch 301/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0012 - accuracy: 0.9023\n",
            "Epoch 302/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0013 - accuracy: 0.9041\n",
            "Epoch 303/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0016 - accuracy: 0.9009\n",
            "Epoch 304/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0018 - accuracy: 0.8983\n",
            "Epoch 305/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0018 - accuracy: 0.8953\n",
            "Epoch 306/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 0.0013 - accuracy: 0.8995\n",
            "Epoch 307/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0012 - accuracy: 0.9018\n",
            "Epoch 308/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0014 - accuracy: 0.9005\n",
            "Epoch 309/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0014 - accuracy: 0.8947\n",
            "Epoch 310/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0012 - accuracy: 0.8975\n",
            "Epoch 311/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0012 - accuracy: 0.8976\n",
            "Epoch 312/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0013 - accuracy: 0.8916\n",
            "Epoch 313/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0012 - accuracy: 0.8932\n",
            "Epoch 314/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 0.0012 - accuracy: 0.9031\n",
            "Epoch 315/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0013 - accuracy: 0.8952\n",
            "Epoch 316/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0012 - accuracy: 0.8970\n",
            "Epoch 317/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0011 - accuracy: 0.9030\n",
            "Epoch 318/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0012 - accuracy: 0.8960\n",
            "Epoch 319/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0012 - accuracy: 0.9037\n",
            "Epoch 320/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0011 - accuracy: 0.8932\n",
            "Epoch 321/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0011 - accuracy: 0.8993\n",
            "Epoch 322/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0012 - accuracy: 0.9022\n",
            "Epoch 323/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0011 - accuracy: 0.9004\n",
            "Epoch 324/1000\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.0011 - accuracy: 0.9044\n",
            "Epoch 325/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0011 - accuracy: 0.9023\n",
            "Epoch 326/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0011 - accuracy: 0.9061\n",
            "Epoch 327/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0011 - accuracy: 0.9025\n",
            "Epoch 328/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0011 - accuracy: 0.9029\n",
            "Epoch 329/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0011 - accuracy: 0.8976\n",
            "Epoch 330/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0011 - accuracy: 0.9050\n",
            "Epoch 331/1000\n",
            "1/1 [==============================] - 0s 23ms/step - loss: 0.0011 - accuracy: 0.8948\n",
            "Epoch 332/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0011 - accuracy: 0.8983\n",
            "Epoch 333/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0011 - accuracy: 0.9056\n",
            "Epoch 334/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0011 - accuracy: 0.8995\n",
            "Epoch 335/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0011 - accuracy: 0.9020\n",
            "Epoch 336/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 0.0011 - accuracy: 0.9041\n",
            "Epoch 337/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0011 - accuracy: 0.9065\n",
            "Epoch 338/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0011 - accuracy: 0.9043\n",
            "Epoch 339/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0011 - accuracy: 0.9051\n",
            "Epoch 340/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0011 - accuracy: 0.9045\n",
            "Epoch 341/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0011 - accuracy: 0.9059\n",
            "Epoch 342/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0011 - accuracy: 0.9036\n",
            "Epoch 343/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0011 - accuracy: 0.9038\n",
            "Epoch 344/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0010 - accuracy: 0.9073\n",
            "Epoch 345/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0011 - accuracy: 0.9037\n",
            "Epoch 346/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0011 - accuracy: 0.9046\n",
            "Epoch 347/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0010 - accuracy: 0.9070\n",
            "Epoch 348/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0010 - accuracy: 0.9068\n",
            "Epoch 349/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0010 - accuracy: 0.9069\n",
            "Epoch 350/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0010 - accuracy: 0.9073\n",
            "Epoch 351/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0010 - accuracy: 0.9072\n",
            "Epoch 352/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 0.0010 - accuracy: 0.9072\n",
            "Epoch 353/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 0.0010 - accuracy: 0.9072\n",
            "Epoch 354/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0010 - accuracy: 0.9071\n",
            "Epoch 355/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0010 - accuracy: 0.9064\n",
            "Epoch 356/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 0.0010 - accuracy: 0.9083\n",
            "Epoch 357/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0010 - accuracy: 0.9070\n",
            "Epoch 358/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0011 - accuracy: 0.9067\n",
            "Epoch 359/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0011 - accuracy: 0.9066\n",
            "Epoch 360/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 0.0011 - accuracy: 0.9082\n",
            "Epoch 361/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 0.0011 - accuracy: 0.9081\n",
            "Epoch 362/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 0.0011 - accuracy: 0.9082\n",
            "Epoch 363/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0012 - accuracy: 0.9068\n",
            "Epoch 364/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0012 - accuracy: 0.9077\n",
            "Epoch 365/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0011 - accuracy: 0.9068\n",
            "Epoch 366/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 0.0011 - accuracy: 0.9086\n",
            "Epoch 367/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0010 - accuracy: 0.9079\n",
            "Epoch 368/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0010 - accuracy: 0.9076\n",
            "Epoch 369/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 0.0011 - accuracy: 0.9082\n",
            "Epoch 370/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 0.0011 - accuracy: 0.9078\n",
            "Epoch 371/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0011 - accuracy: 0.9079\n",
            "Epoch 372/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0010 - accuracy: 0.9082\n",
            "Epoch 373/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0010 - accuracy: 0.9086\n",
            "Epoch 374/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0010 - accuracy: 0.9080\n",
            "Epoch 375/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0010 - accuracy: 0.9080\n",
            "Epoch 376/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 0.0010 - accuracy: 0.9082\n",
            "Epoch 377/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 0.0010 - accuracy: 0.9081\n",
            "Epoch 378/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0010 - accuracy: 0.9088\n",
            "Epoch 379/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0010 - accuracy: 0.9087\n",
            "Epoch 380/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0010 - accuracy: 0.9089\n",
            "Epoch 381/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0010 - accuracy: 0.9085\n",
            "Epoch 382/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 0.0010 - accuracy: 0.9087\n",
            "Epoch 383/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0010 - accuracy: 0.9083\n",
            "Epoch 384/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 0.0010 - accuracy: 0.9078\n",
            "Epoch 385/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0010 - accuracy: 0.9079\n",
            "Epoch 386/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 9.9424e-04 - accuracy: 0.9083\n",
            "Epoch 387/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 9.8897e-04 - accuracy: 0.9087\n",
            "Epoch 388/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 9.8762e-04 - accuracy: 0.9082\n",
            "Epoch 389/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 9.8938e-04 - accuracy: 0.9091\n",
            "Epoch 390/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 9.9294e-04 - accuracy: 0.9090\n",
            "Epoch 391/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 9.9775e-04 - accuracy: 0.9091\n",
            "Epoch 392/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0010 - accuracy: 0.9083\n",
            "Epoch 393/1000\n",
            "1/1 [==============================] - 0s 23ms/step - loss: 0.0010 - accuracy: 0.9087\n",
            "Epoch 394/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 0.0010 - accuracy: 0.9083\n",
            "Epoch 395/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 0.0010 - accuracy: 0.9090\n",
            "Epoch 396/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0010 - accuracy: 0.9082\n",
            "Epoch 397/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 0.0011 - accuracy: 0.9088\n",
            "Epoch 398/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 0.0011 - accuracy: 0.9087\n",
            "Epoch 399/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0011 - accuracy: 0.9092\n",
            "Epoch 400/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0010 - accuracy: 0.9090\n",
            "Epoch 401/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0010 - accuracy: 0.9089\n",
            "Epoch 402/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 9.8082e-04 - accuracy: 0.9089\n",
            "Epoch 403/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 9.7136e-04 - accuracy: 0.9078\n",
            "Epoch 404/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 9.7872e-04 - accuracy: 0.9088\n",
            "Epoch 405/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 9.9546e-04 - accuracy: 0.9064\n",
            "Epoch 406/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0010 - accuracy: 0.9058\n",
            "Epoch 407/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0010 - accuracy: 0.9054\n",
            "Epoch 408/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0010 - accuracy: 0.9056\n",
            "Epoch 409/1000\n",
            "1/1 [==============================] - 0s 22ms/step - loss: 9.9999e-04 - accuracy: 0.9069\n",
            "Epoch 410/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 9.7611e-04 - accuracy: 0.9092\n",
            "Epoch 411/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 9.5961e-04 - accuracy: 0.9093\n",
            "Epoch 412/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 9.5715e-04 - accuracy: 0.9099\n",
            "Epoch 413/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 9.6565e-04 - accuracy: 0.9104\n",
            "Epoch 414/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 9.7865e-04 - accuracy: 0.9080\n",
            "Epoch 415/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 9.9042e-04 - accuracy: 0.9093\n",
            "Epoch 416/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0010 - accuracy: 0.9082\n",
            "Epoch 417/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0010 - accuracy: 0.9103\n",
            "Epoch 418/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0010 - accuracy: 0.9079\n",
            "Epoch 419/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0011 - accuracy: 0.9027\n",
            "Epoch 420/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0011 - accuracy: 0.8929\n",
            "Epoch 421/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0011 - accuracy: 0.8941\n",
            "Epoch 422/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0010 - accuracy: 0.9005\n",
            "Epoch 423/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 9.6172e-04 - accuracy: 0.9073\n",
            "Epoch 424/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 9.5915e-04 - accuracy: 0.9041\n",
            "Epoch 425/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 9.9290e-04 - accuracy: 0.9014\n",
            "Epoch 426/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 9.8640e-04 - accuracy: 0.8961\n",
            "Epoch 427/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 9.5697e-04 - accuracy: 0.9103\n",
            "Epoch 428/1000\n",
            "1/1 [==============================] - 0s 22ms/step - loss: 9.6478e-04 - accuracy: 0.9078\n",
            "Epoch 429/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 9.9115e-04 - accuracy: 0.8943\n",
            "Epoch 430/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 9.7485e-04 - accuracy: 0.9012\n",
            "Epoch 431/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 9.5657e-04 - accuracy: 0.9091\n",
            "Epoch 432/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 9.6589e-04 - accuracy: 0.9067\n",
            "Epoch 433/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 9.6448e-04 - accuracy: 0.9081\n",
            "Epoch 434/1000\n",
            "1/1 [==============================] - 0s 25ms/step - loss: 9.4285e-04 - accuracy: 0.9092\n",
            "Epoch 435/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 9.3692e-04 - accuracy: 0.9066\n",
            "Epoch 436/1000\n",
            "1/1 [==============================] - 0s 25ms/step - loss: 9.4435e-04 - accuracy: 0.9082\n",
            "Epoch 437/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 9.3770e-04 - accuracy: 0.9055\n",
            "Epoch 438/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 9.2654e-04 - accuracy: 0.9108\n",
            "Epoch 439/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 9.3034e-04 - accuracy: 0.9097\n",
            "Epoch 440/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 9.3469e-04 - accuracy: 0.9055\n",
            "Epoch 441/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 9.2821e-04 - accuracy: 0.9109\n",
            "Epoch 442/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 9.2453e-04 - accuracy: 0.9113\n",
            "Epoch 443/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 9.3091e-04 - accuracy: 0.9092\n",
            "Epoch 444/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 9.3712e-04 - accuracy: 0.9116\n",
            "Epoch 445/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 9.4185e-04 - accuracy: 0.9112\n",
            "Epoch 446/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 9.6294e-04 - accuracy: 0.9071\n",
            "Epoch 447/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 0.0010 - accuracy: 0.9083\n",
            "Epoch 448/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 0.0011 - accuracy: 0.9012\n",
            "Epoch 449/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0013 - accuracy: 0.9020\n",
            "Epoch 450/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0015 - accuracy: 0.9049\n",
            "Epoch 451/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 0.0015 - accuracy: 0.9090\n",
            "Epoch 452/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.0012 - accuracy: 0.9086\n",
            "Epoch 453/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 9.5535e-04 - accuracy: 0.9083\n",
            "Epoch 454/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0010 - accuracy: 0.9112\n",
            "Epoch 455/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0012 - accuracy: 0.9072\n",
            "Epoch 456/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 0.0011 - accuracy: 0.9072\n",
            "Epoch 457/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 9.3268e-04 - accuracy: 0.9105\n",
            "Epoch 458/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 0.0010 - accuracy: 0.9096\n",
            "Epoch 459/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 0.0011 - accuracy: 0.9103\n",
            "Epoch 460/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 9.9301e-04 - accuracy: 0.9066\n",
            "Epoch 461/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 9.5545e-04 - accuracy: 0.9059\n",
            "Epoch 462/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 0.0010 - accuracy: 0.9090\n",
            "Epoch 463/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 9.9998e-04 - accuracy: 0.9059\n",
            "Epoch 464/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 9.6543e-04 - accuracy: 0.8941\n",
            "Epoch 465/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 9.5758e-04 - accuracy: 0.9078\n",
            "Epoch 466/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 9.6725e-04 - accuracy: 0.9092\n",
            "Epoch 467/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 9.5175e-04 - accuracy: 0.9070\n",
            "Epoch 468/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 9.4995e-04 - accuracy: 0.9024\n",
            "Epoch 469/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 9.2179e-04 - accuracy: 0.9112\n",
            "Epoch 470/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 9.3660e-04 - accuracy: 0.9113\n",
            "Epoch 471/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 9.3799e-04 - accuracy: 0.9013\n",
            "Epoch 472/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 9.2134e-04 - accuracy: 0.9085\n",
            "Epoch 473/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 9.0320e-04 - accuracy: 0.9124\n",
            "Epoch 474/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 9.3000e-04 - accuracy: 0.9054\n",
            "Epoch 475/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 9.1526e-04 - accuracy: 0.9070\n",
            "Epoch 476/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 9.0113e-04 - accuracy: 0.9113\n",
            "Epoch 477/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 9.0169e-04 - accuracy: 0.9081\n",
            "Epoch 478/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 9.1097e-04 - accuracy: 0.9110\n",
            "Epoch 479/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 8.9815e-04 - accuracy: 0.9088\n",
            "Epoch 480/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 8.9834e-04 - accuracy: 0.9112\n",
            "Epoch 481/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 8.9589e-04 - accuracy: 0.9109\n",
            "Epoch 482/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 8.9273e-04 - accuracy: 0.9089\n",
            "Epoch 483/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 8.8950e-04 - accuracy: 0.9120\n",
            "Epoch 484/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 8.9571e-04 - accuracy: 0.9118\n",
            "Epoch 485/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 8.8870e-04 - accuracy: 0.9084\n",
            "Epoch 486/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 8.8080e-04 - accuracy: 0.9122\n",
            "Epoch 487/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 8.8361e-04 - accuracy: 0.9125\n",
            "Epoch 488/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 8.8476e-04 - accuracy: 0.9093\n",
            "Epoch 489/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 8.8009e-04 - accuracy: 0.9127\n",
            "Epoch 490/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 8.7795e-04 - accuracy: 0.9128\n",
            "Epoch 491/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 8.7894e-04 - accuracy: 0.9095\n",
            "Epoch 492/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 8.7431e-04 - accuracy: 0.9130\n",
            "Epoch 493/1000\n",
            "1/1 [==============================] - 0s 23ms/step - loss: 8.7200e-04 - accuracy: 0.9122\n",
            "Epoch 494/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 8.7156e-04 - accuracy: 0.9112\n",
            "Epoch 495/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 8.7307e-04 - accuracy: 0.9133\n",
            "Epoch 496/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 8.6954e-04 - accuracy: 0.9119\n",
            "Epoch 497/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 8.6745e-04 - accuracy: 0.9128\n",
            "Epoch 498/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 8.6661e-04 - accuracy: 0.9135\n",
            "Epoch 499/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 8.6549e-04 - accuracy: 0.9124\n",
            "Epoch 500/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 8.6273e-04 - accuracy: 0.9132\n",
            "Epoch 501/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 8.6057e-04 - accuracy: 0.9132\n",
            "Epoch 502/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 8.6098e-04 - accuracy: 0.9129\n",
            "Epoch 503/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 8.5973e-04 - accuracy: 0.9135\n",
            "Epoch 504/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 8.5820e-04 - accuracy: 0.9130\n",
            "Epoch 505/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 8.5679e-04 - accuracy: 0.9133\n",
            "Epoch 506/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 8.5657e-04 - accuracy: 0.9136\n",
            "Epoch 507/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 8.5647e-04 - accuracy: 0.9119\n",
            "Epoch 508/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 8.5576e-04 - accuracy: 0.9136\n",
            "Epoch 509/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 8.5604e-04 - accuracy: 0.9128\n",
            "Epoch 510/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 8.5727e-04 - accuracy: 0.9135\n",
            "Epoch 511/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 8.6158e-04 - accuracy: 0.9133\n",
            "Epoch 512/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 8.6900e-04 - accuracy: 0.9137\n",
            "Epoch 513/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 8.8122e-04 - accuracy: 0.9134\n",
            "Epoch 514/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 9.0109e-04 - accuracy: 0.9140\n",
            "Epoch 515/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 9.2932e-04 - accuracy: 0.9115\n",
            "Epoch 516/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 9.6100e-04 - accuracy: 0.9141\n",
            "Epoch 517/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 9.7712e-04 - accuracy: 0.9109\n",
            "Epoch 518/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 9.6612e-04 - accuracy: 0.9141\n",
            "Epoch 519/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 9.2273e-04 - accuracy: 0.9131\n",
            "Epoch 520/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 8.6985e-04 - accuracy: 0.9135\n",
            "Epoch 521/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 8.4263e-04 - accuracy: 0.9139\n",
            "Epoch 522/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 8.5294e-04 - accuracy: 0.9128\n",
            "Epoch 523/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 8.8233e-04 - accuracy: 0.9145\n",
            "Epoch 524/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 9.0001e-04 - accuracy: 0.9130\n",
            "Epoch 525/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 8.9548e-04 - accuracy: 0.9144\n",
            "Epoch 526/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 8.6853e-04 - accuracy: 0.9136\n",
            "Epoch 527/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 8.4148e-04 - accuracy: 0.9142\n",
            "Epoch 528/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 8.3336e-04 - accuracy: 0.9144\n",
            "Epoch 529/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 8.4413e-04 - accuracy: 0.9147\n",
            "Epoch 530/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 8.5950e-04 - accuracy: 0.9144\n",
            "Epoch 531/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 8.6659e-04 - accuracy: 0.9137\n",
            "Epoch 532/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 8.6226e-04 - accuracy: 0.9142\n",
            "Epoch 533/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 8.4899e-04 - accuracy: 0.9132\n",
            "Epoch 534/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 8.3449e-04 - accuracy: 0.9143\n",
            "Epoch 535/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 8.2620e-04 - accuracy: 0.9130\n",
            "Epoch 536/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 8.2647e-04 - accuracy: 0.9147\n",
            "Epoch 537/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 8.3161e-04 - accuracy: 0.9143\n",
            "Epoch 538/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 8.3768e-04 - accuracy: 0.9143\n",
            "Epoch 539/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 8.4212e-04 - accuracy: 0.9149\n",
            "Epoch 540/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 8.4273e-04 - accuracy: 0.9137\n",
            "Epoch 541/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 8.4012e-04 - accuracy: 0.9151\n",
            "Epoch 542/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 8.3555e-04 - accuracy: 0.9127\n",
            "Epoch 543/1000\n",
            "1/1 [==============================] - 0s 24ms/step - loss: 8.2887e-04 - accuracy: 0.9151\n",
            "Epoch 544/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 8.2259e-04 - accuracy: 0.9128\n",
            "Epoch 545/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 8.1705e-04 - accuracy: 0.9153\n",
            "Epoch 546/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 8.1329e-04 - accuracy: 0.9136\n",
            "Epoch 547/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 8.1183e-04 - accuracy: 0.9147\n",
            "Epoch 548/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 8.1221e-04 - accuracy: 0.9152\n",
            "Epoch 549/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 8.1418e-04 - accuracy: 0.9137\n",
            "Epoch 550/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 8.1700e-04 - accuracy: 0.9154\n",
            "Epoch 551/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 8.2092e-04 - accuracy: 0.9133\n",
            "Epoch 552/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 8.2629e-04 - accuracy: 0.9158\n",
            "Epoch 553/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 8.3431e-04 - accuracy: 0.9135\n",
            "Epoch 554/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 8.4523e-04 - accuracy: 0.9156\n",
            "Epoch 555/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 8.6232e-04 - accuracy: 0.9144\n",
            "Epoch 556/1000\n",
            "1/1 [==============================] - 0s 25ms/step - loss: 8.8760e-04 - accuracy: 0.9154\n",
            "Epoch 557/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 9.2165e-04 - accuracy: 0.9145\n",
            "Epoch 558/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 9.6548e-04 - accuracy: 0.9118\n",
            "Epoch 559/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 9.8613e-04 - accuracy: 0.9139\n",
            "Epoch 560/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 9.7687e-04 - accuracy: 0.9102\n",
            "Epoch 561/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 8.9707e-04 - accuracy: 0.9152\n",
            "Epoch 562/1000\n",
            "1/1 [==============================] - 0s 23ms/step - loss: 8.2068e-04 - accuracy: 0.9153\n",
            "Epoch 563/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 8.1102e-04 - accuracy: 0.9116\n",
            "Epoch 564/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 8.5322e-04 - accuracy: 0.9129\n",
            "Epoch 565/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 8.7064e-04 - accuracy: 0.9105\n",
            "Epoch 566/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 8.4147e-04 - accuracy: 0.9162\n",
            "Epoch 567/1000\n",
            "1/1 [==============================] - 0s 23ms/step - loss: 8.2354e-04 - accuracy: 0.9156\n",
            "Epoch 568/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 8.4850e-04 - accuracy: 0.9005\n",
            "Epoch 569/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 8.7130e-04 - accuracy: 0.8969\n",
            "Epoch 570/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 8.4320e-04 - accuracy: 0.9006\n",
            "Epoch 571/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 8.0891e-04 - accuracy: 0.9150\n",
            "Epoch 572/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 8.1938e-04 - accuracy: 0.9159\n",
            "Epoch 573/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 8.2903e-04 - accuracy: 0.9060\n",
            "Epoch 574/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 8.0431e-04 - accuracy: 0.9153\n",
            "Epoch 575/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 7.9066e-04 - accuracy: 0.9148\n",
            "Epoch 576/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 8.0536e-04 - accuracy: 0.9109\n",
            "Epoch 577/1000\n",
            "1/1 [==============================] - 0s 22ms/step - loss: 8.1110e-04 - accuracy: 0.9149\n",
            "Epoch 578/1000\n",
            "1/1 [==============================] - 0s 23ms/step - loss: 7.9870e-04 - accuracy: 0.9130\n",
            "Epoch 579/1000\n",
            "1/1 [==============================] - 0s 25ms/step - loss: 7.9551e-04 - accuracy: 0.9146\n",
            "Epoch 580/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 8.0622e-04 - accuracy: 0.9114\n",
            "Epoch 581/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 8.0810e-04 - accuracy: 0.9070\n",
            "Epoch 582/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 7.9726e-04 - accuracy: 0.9158\n",
            "Epoch 583/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 7.9704e-04 - accuracy: 0.9153\n",
            "Epoch 584/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 8.0547e-04 - accuracy: 0.9140\n",
            "Epoch 585/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 8.0613e-04 - accuracy: 0.9163\n",
            "Epoch 586/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 8.0136e-04 - accuracy: 0.9158\n",
            "Epoch 587/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 8.0318e-04 - accuracy: 0.9129\n",
            "Epoch 588/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 8.1129e-04 - accuracy: 0.9132\n",
            "Epoch 589/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 8.1737e-04 - accuracy: 0.9067\n",
            "Epoch 590/1000\n",
            "1/1 [==============================] - 0s 22ms/step - loss: 8.1490e-04 - accuracy: 0.9152\n",
            "Epoch 591/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 8.1961e-04 - accuracy: 0.9154\n",
            "Epoch 592/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 8.2844e-04 - accuracy: 0.9158\n",
            "Epoch 593/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 8.2922e-04 - accuracy: 0.9158\n",
            "Epoch 594/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 8.2126e-04 - accuracy: 0.9168\n",
            "Epoch 595/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 8.1520e-04 - accuracy: 0.9125\n",
            "Epoch 596/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 8.0944e-04 - accuracy: 0.9139\n",
            "Epoch 597/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 7.9910e-04 - accuracy: 0.9099\n",
            "Epoch 598/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 7.8127e-04 - accuracy: 0.9171\n",
            "Epoch 599/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 7.7274e-04 - accuracy: 0.9170\n",
            "Epoch 600/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 7.7174e-04 - accuracy: 0.9131\n",
            "Epoch 601/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 7.7085e-04 - accuracy: 0.9176\n",
            "Epoch 602/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 7.6819e-04 - accuracy: 0.9153\n",
            "Epoch 603/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 7.7206e-04 - accuracy: 0.9172\n",
            "Epoch 604/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 7.7937e-04 - accuracy: 0.9173\n",
            "Epoch 605/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 7.8670e-04 - accuracy: 0.9170\n",
            "Epoch 606/1000\n",
            "1/1 [==============================] - 0s 22ms/step - loss: 7.9180e-04 - accuracy: 0.9169\n",
            "Epoch 607/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 8.0127e-04 - accuracy: 0.9176\n",
            "Epoch 608/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 8.1517e-04 - accuracy: 0.9120\n",
            "Epoch 609/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 8.3118e-04 - accuracy: 0.9156\n",
            "Epoch 610/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 8.4767e-04 - accuracy: 0.9110\n",
            "Epoch 611/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 8.5520e-04 - accuracy: 0.9163\n",
            "Epoch 612/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 8.5778e-04 - accuracy: 0.9160\n",
            "Epoch 613/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 8.4499e-04 - accuracy: 0.9165\n",
            "Epoch 614/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 8.1917e-04 - accuracy: 0.9174\n",
            "Epoch 615/1000\n",
            "1/1 [==============================] - 0s 25ms/step - loss: 7.8791e-04 - accuracy: 0.9179\n",
            "Epoch 616/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 7.6332e-04 - accuracy: 0.9165\n",
            "Epoch 617/1000\n",
            "1/1 [==============================] - 0s 26ms/step - loss: 7.5550e-04 - accuracy: 0.9176\n",
            "Epoch 618/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 7.6570e-04 - accuracy: 0.9172\n",
            "Epoch 619/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 7.8007e-04 - accuracy: 0.9175\n",
            "Epoch 620/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 7.9463e-04 - accuracy: 0.9180\n",
            "Epoch 621/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 7.9925e-04 - accuracy: 0.9152\n",
            "Epoch 622/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 7.9904e-04 - accuracy: 0.9173\n",
            "Epoch 623/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 7.9255e-04 - accuracy: 0.9138\n",
            "Epoch 624/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 7.7790e-04 - accuracy: 0.9169\n",
            "Epoch 625/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 7.5887e-04 - accuracy: 0.9171\n",
            "Epoch 626/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 7.4753e-04 - accuracy: 0.9173\n",
            "Epoch 627/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 7.4861e-04 - accuracy: 0.9178\n",
            "Epoch 628/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 7.5577e-04 - accuracy: 0.9165\n",
            "Epoch 629/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 7.6065e-04 - accuracy: 0.9182\n",
            "Epoch 630/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 7.6549e-04 - accuracy: 0.9171\n",
            "Epoch 631/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 7.7373e-04 - accuracy: 0.9188\n",
            "Epoch 632/1000\n",
            "1/1 [==============================] - 0s 28ms/step - loss: 7.8116e-04 - accuracy: 0.9183\n",
            "Epoch 633/1000\n",
            "1/1 [==============================] - 0s 24ms/step - loss: 7.8448e-04 - accuracy: 0.9186\n",
            "Epoch 634/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 7.8039e-04 - accuracy: 0.9187\n",
            "Epoch 635/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 7.7457e-04 - accuracy: 0.9183\n",
            "Epoch 636/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 7.6797e-04 - accuracy: 0.9181\n",
            "Epoch 637/1000\n",
            "1/1 [==============================] - 0s 24ms/step - loss: 7.5851e-04 - accuracy: 0.9178\n",
            "Epoch 638/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 7.4819e-04 - accuracy: 0.9172\n",
            "Epoch 639/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 7.4106e-04 - accuracy: 0.9189\n",
            "Epoch 640/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 7.3773e-04 - accuracy: 0.9170\n",
            "Epoch 641/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 7.3541e-04 - accuracy: 0.9190\n",
            "Epoch 642/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 7.3376e-04 - accuracy: 0.9187\n",
            "Epoch 643/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 7.3510e-04 - accuracy: 0.9188\n",
            "Epoch 644/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 7.3894e-04 - accuracy: 0.9191\n",
            "Epoch 645/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 7.4392e-04 - accuracy: 0.9165\n",
            "Epoch 646/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 7.4844e-04 - accuracy: 0.9191\n",
            "Epoch 647/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 7.5477e-04 - accuracy: 0.9171\n",
            "Epoch 648/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 7.6603e-04 - accuracy: 0.9188\n",
            "Epoch 649/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 7.8432e-04 - accuracy: 0.9178\n",
            "Epoch 650/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 8.0877e-04 - accuracy: 0.9189\n",
            "Epoch 651/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 8.4256e-04 - accuracy: 0.9174\n",
            "Epoch 652/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 8.8513e-04 - accuracy: 0.9193\n",
            "Epoch 653/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 9.1495e-04 - accuracy: 0.9179\n",
            "Epoch 654/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 9.2659e-04 - accuracy: 0.9192\n",
            "Epoch 655/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 8.8129e-04 - accuracy: 0.9180\n",
            "Epoch 656/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 8.0584e-04 - accuracy: 0.9194\n",
            "Epoch 657/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 7.4083e-04 - accuracy: 0.9185\n",
            "Epoch 658/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 7.3340e-04 - accuracy: 0.9184\n",
            "Epoch 659/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 7.7223e-04 - accuracy: 0.9194\n",
            "Epoch 660/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 8.0668e-04 - accuracy: 0.9183\n",
            "Epoch 661/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 8.0776e-04 - accuracy: 0.9191\n",
            "Epoch 662/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 7.7459e-04 - accuracy: 0.9164\n",
            "Epoch 663/1000\n",
            "1/1 [==============================] - 0s 25ms/step - loss: 7.3799e-04 - accuracy: 0.9188\n",
            "Epoch 664/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 7.2835e-04 - accuracy: 0.9165\n",
            "Epoch 665/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 7.4355e-04 - accuracy: 0.9193\n",
            "Epoch 666/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 7.5981e-04 - accuracy: 0.9182\n",
            "Epoch 667/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 7.5938e-04 - accuracy: 0.9189\n",
            "Epoch 668/1000\n",
            "1/1 [==============================] - 0s 25ms/step - loss: 7.4536e-04 - accuracy: 0.9196\n",
            "Epoch 669/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 7.3111e-04 - accuracy: 0.9138\n",
            "Epoch 670/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 7.2538e-04 - accuracy: 0.9185\n",
            "Epoch 671/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 7.2737e-04 - accuracy: 0.9153\n",
            "Epoch 672/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 7.3164e-04 - accuracy: 0.9206\n",
            "Epoch 673/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 7.3411e-04 - accuracy: 0.9205\n",
            "Epoch 674/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 7.3166e-04 - accuracy: 0.9182\n",
            "Epoch 675/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 7.2675e-04 - accuracy: 0.9203\n",
            "Epoch 676/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 7.2209e-04 - accuracy: 0.9146\n",
            "Epoch 677/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 7.1942e-04 - accuracy: 0.9194\n",
            "Epoch 678/1000\n",
            "1/1 [==============================] - 0s 22ms/step - loss: 7.1947e-04 - accuracy: 0.9177\n",
            "Epoch 679/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 7.1975e-04 - accuracy: 0.9202\n",
            "Epoch 680/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 7.1774e-04 - accuracy: 0.9202\n",
            "Epoch 681/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 7.1583e-04 - accuracy: 0.9203\n",
            "Epoch 682/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 7.1563e-04 - accuracy: 0.9207\n",
            "Epoch 683/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 7.1630e-04 - accuracy: 0.9182\n",
            "Epoch 684/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 7.1614e-04 - accuracy: 0.9196\n",
            "Epoch 685/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 7.1475e-04 - accuracy: 0.9161\n",
            "Epoch 686/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 7.1248e-04 - accuracy: 0.9196\n",
            "Epoch 687/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 7.0993e-04 - accuracy: 0.9155\n",
            "Epoch 688/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 7.0696e-04 - accuracy: 0.9208\n",
            "Epoch 689/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 7.0397e-04 - accuracy: 0.9181\n",
            "Epoch 690/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 7.0165e-04 - accuracy: 0.9213\n",
            "Epoch 691/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 7.0115e-04 - accuracy: 0.9209\n",
            "Epoch 692/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 7.0230e-04 - accuracy: 0.9195\n",
            "Epoch 693/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 7.0433e-04 - accuracy: 0.9214\n",
            "Epoch 694/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 7.0674e-04 - accuracy: 0.9169\n",
            "Epoch 695/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 7.0901e-04 - accuracy: 0.9205\n",
            "Epoch 696/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 7.1191e-04 - accuracy: 0.9161\n",
            "Epoch 697/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 7.1529e-04 - accuracy: 0.9209\n",
            "Epoch 698/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 7.2160e-04 - accuracy: 0.9170\n",
            "Epoch 699/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 7.3124e-04 - accuracy: 0.9209\n",
            "Epoch 700/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 7.4911e-04 - accuracy: 0.9188\n",
            "Epoch 701/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 7.7778e-04 - accuracy: 0.9213\n",
            "Epoch 702/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 8.2666e-04 - accuracy: 0.9209\n",
            "Epoch 703/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 9.0508e-04 - accuracy: 0.9205\n",
            "Epoch 704/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 9.9046e-04 - accuracy: 0.9197\n",
            "Epoch 705/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.0011 - accuracy: 0.9167\n",
            "Epoch 706/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 9.7426e-04 - accuracy: 0.9184\n",
            "Epoch 707/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 8.1377e-04 - accuracy: 0.9187\n",
            "Epoch 708/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 7.2268e-04 - accuracy: 0.9156\n",
            "Epoch 709/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 7.8400e-04 - accuracy: 0.9114\n",
            "Epoch 710/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 8.3434e-04 - accuracy: 0.9091\n",
            "Epoch 711/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 7.7823e-04 - accuracy: 0.9203\n",
            "Epoch 712/1000\n",
            "1/1 [==============================] - 0s 24ms/step - loss: 7.4444e-04 - accuracy: 0.9163\n",
            "Epoch 713/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 7.8778e-04 - accuracy: 0.8913\n",
            "Epoch 714/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 7.8289e-04 - accuracy: 0.8949\n",
            "Epoch 715/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 7.1256e-04 - accuracy: 0.9113\n",
            "Epoch 716/1000\n",
            "1/1 [==============================] - 0s 28ms/step - loss: 7.1419e-04 - accuracy: 0.9173\n",
            "Epoch 717/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 7.5121e-04 - accuracy: 0.9081\n",
            "Epoch 718/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 7.2336e-04 - accuracy: 0.9043\n",
            "Epoch 719/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 7.0629e-04 - accuracy: 0.9215\n",
            "Epoch 720/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 7.3354e-04 - accuracy: 0.9192\n",
            "Epoch 721/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 7.2335e-04 - accuracy: 0.9124\n",
            "Epoch 722/1000\n",
            "1/1 [==============================] - 0s 23ms/step - loss: 6.9619e-04 - accuracy: 0.9223\n",
            "Epoch 723/1000\n",
            "1/1 [==============================] - 0s 23ms/step - loss: 7.0707e-04 - accuracy: 0.9191\n",
            "Epoch 724/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 7.1207e-04 - accuracy: 0.9130\n",
            "Epoch 725/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 6.9398e-04 - accuracy: 0.9227\n",
            "Epoch 726/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 6.9477e-04 - accuracy: 0.9215\n",
            "Epoch 727/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 7.0746e-04 - accuracy: 0.9105\n",
            "Epoch 728/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 6.9660e-04 - accuracy: 0.9196\n",
            "Epoch 729/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 6.8518e-04 - accuracy: 0.9219\n",
            "Epoch 730/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 6.9397e-04 - accuracy: 0.9180\n",
            "Epoch 731/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 6.9320e-04 - accuracy: 0.9216\n",
            "Epoch 732/1000\n",
            "1/1 [==============================] - 0s 22ms/step - loss: 6.8051e-04 - accuracy: 0.9222\n",
            "Epoch 733/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.8219e-04 - accuracy: 0.9206\n",
            "Epoch 734/1000\n",
            "1/1 [==============================] - 0s 22ms/step - loss: 6.8832e-04 - accuracy: 0.9216\n",
            "Epoch 735/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.8156e-04 - accuracy: 0.9209\n",
            "Epoch 736/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 6.7772e-04 - accuracy: 0.9210\n",
            "Epoch 737/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 6.8214e-04 - accuracy: 0.9213\n",
            "Epoch 738/1000\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 6.8217e-04 - accuracy: 0.9180\n",
            "Epoch 739/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 6.7805e-04 - accuracy: 0.9232\n",
            "Epoch 740/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 6.7866e-04 - accuracy: 0.9235\n",
            "Epoch 741/1000\n",
            "1/1 [==============================] - 0s 26ms/step - loss: 6.8046e-04 - accuracy: 0.9217\n",
            "Epoch 742/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.8017e-04 - accuracy: 0.9233\n",
            "Epoch 743/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.8018e-04 - accuracy: 0.9233\n",
            "Epoch 744/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 6.8240e-04 - accuracy: 0.9178\n",
            "Epoch 745/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 6.8756e-04 - accuracy: 0.9205\n",
            "Epoch 746/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 6.9581e-04 - accuracy: 0.9164\n",
            "Epoch 747/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 7.0673e-04 - accuracy: 0.9213\n",
            "Epoch 748/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 7.2633e-04 - accuracy: 0.9170\n",
            "Epoch 749/1000\n",
            "1/1 [==============================] - 0s 24ms/step - loss: 7.5556e-04 - accuracy: 0.9208\n",
            "Epoch 750/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 7.9109e-04 - accuracy: 0.9141\n",
            "Epoch 751/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 8.2202e-04 - accuracy: 0.9160\n",
            "Epoch 752/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 8.3985e-04 - accuracy: 0.9111\n",
            "Epoch 753/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 8.0899e-04 - accuracy: 0.9172\n",
            "Epoch 754/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 7.5076e-04 - accuracy: 0.9173\n",
            "Epoch 755/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 6.8905e-04 - accuracy: 0.9232\n",
            "Epoch 756/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 6.7450e-04 - accuracy: 0.9232\n",
            "Epoch 757/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 6.9711e-04 - accuracy: 0.9173\n",
            "Epoch 758/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 7.3180e-04 - accuracy: 0.9209\n",
            "Epoch 759/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 7.4446e-04 - accuracy: 0.9190\n",
            "Epoch 760/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 7.2421e-04 - accuracy: 0.9233\n",
            "Epoch 761/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 6.9068e-04 - accuracy: 0.9232\n",
            "Epoch 762/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 6.6820e-04 - accuracy: 0.9216\n",
            "Epoch 763/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 6.7117e-04 - accuracy: 0.9235\n",
            "Epoch 764/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.9451e-04 - accuracy: 0.9230\n",
            "Epoch 765/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 7.0631e-04 - accuracy: 0.9226\n",
            "Epoch 766/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 6.9659e-04 - accuracy: 0.9231\n",
            "Epoch 767/1000\n",
            "1/1 [==============================] - 0s 23ms/step - loss: 6.7992e-04 - accuracy: 0.9224\n",
            "Epoch 768/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 6.6460e-04 - accuracy: 0.9222\n",
            "Epoch 769/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 6.6247e-04 - accuracy: 0.9230\n",
            "Epoch 770/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 6.7101e-04 - accuracy: 0.9233\n",
            "Epoch 771/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 6.7799e-04 - accuracy: 0.9234\n",
            "Epoch 772/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 6.7900e-04 - accuracy: 0.9240\n",
            "Epoch 773/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 6.7746e-04 - accuracy: 0.9221\n",
            "Epoch 774/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 6.6763e-04 - accuracy: 0.9234\n",
            "Epoch 775/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 6.5934e-04 - accuracy: 0.9227\n",
            "Epoch 776/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 6.5570e-04 - accuracy: 0.9234\n",
            "Epoch 777/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 6.5708e-04 - accuracy: 0.9238\n",
            "Epoch 778/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.6041e-04 - accuracy: 0.9225\n",
            "Epoch 779/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 6.6351e-04 - accuracy: 0.9235\n",
            "Epoch 780/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.6231e-04 - accuracy: 0.9223\n",
            "Epoch 781/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 6.5886e-04 - accuracy: 0.9232\n",
            "Epoch 782/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 6.5594e-04 - accuracy: 0.9238\n",
            "Epoch 783/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.5270e-04 - accuracy: 0.9233\n",
            "Epoch 784/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.5073e-04 - accuracy: 0.9238\n",
            "Epoch 785/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 6.4939e-04 - accuracy: 0.9237\n",
            "Epoch 786/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.4941e-04 - accuracy: 0.9233\n",
            "Epoch 787/1000\n",
            "1/1 [==============================] - 0s 22ms/step - loss: 6.4936e-04 - accuracy: 0.9240\n",
            "Epoch 788/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.4990e-04 - accuracy: 0.9233\n",
            "Epoch 789/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.5088e-04 - accuracy: 0.9237\n",
            "Epoch 790/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 6.5094e-04 - accuracy: 0.9239\n",
            "Epoch 791/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 6.5137e-04 - accuracy: 0.9232\n",
            "Epoch 792/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 6.5210e-04 - accuracy: 0.9244\n",
            "Epoch 793/1000\n",
            "1/1 [==============================] - 0s 22ms/step - loss: 6.5321e-04 - accuracy: 0.9223\n",
            "Epoch 794/1000\n",
            "1/1 [==============================] - 0s 29ms/step - loss: 6.5445e-04 - accuracy: 0.9245\n",
            "Epoch 795/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 6.5625e-04 - accuracy: 0.9211\n",
            "Epoch 796/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 6.5908e-04 - accuracy: 0.9245\n",
            "Epoch 797/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 6.6207e-04 - accuracy: 0.9218\n",
            "Epoch 798/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.6764e-04 - accuracy: 0.9247\n",
            "Epoch 799/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 6.7661e-04 - accuracy: 0.9229\n",
            "Epoch 800/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 6.8894e-04 - accuracy: 0.9242\n",
            "Epoch 801/1000\n",
            "1/1 [==============================] - 0s 23ms/step - loss: 7.0713e-04 - accuracy: 0.9227\n",
            "Epoch 802/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 7.2994e-04 - accuracy: 0.9230\n",
            "Epoch 803/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 7.5470e-04 - accuracy: 0.9216\n",
            "Epoch 804/1000\n",
            "1/1 [==============================] - 0s 22ms/step - loss: 7.7254e-04 - accuracy: 0.9231\n",
            "Epoch 805/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 7.7739e-04 - accuracy: 0.9219\n",
            "Epoch 806/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 7.6451e-04 - accuracy: 0.9237\n",
            "Epoch 807/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 7.3296e-04 - accuracy: 0.9172\n",
            "Epoch 808/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.9320e-04 - accuracy: 0.9180\n",
            "Epoch 809/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 6.5682e-04 - accuracy: 0.9182\n",
            "Epoch 810/1000\n",
            "1/1 [==============================] - 0s 23ms/step - loss: 6.4413e-04 - accuracy: 0.9237\n",
            "Epoch 811/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 6.6327e-04 - accuracy: 0.9241\n",
            "Epoch 812/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 6.9244e-04 - accuracy: 0.9170\n",
            "Epoch 813/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 7.0283e-04 - accuracy: 0.9202\n",
            "Epoch 814/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.7909e-04 - accuracy: 0.9200\n",
            "Epoch 815/1000\n",
            "1/1 [==============================] - 0s 23ms/step - loss: 6.5379e-04 - accuracy: 0.9238\n",
            "Epoch 816/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 6.4954e-04 - accuracy: 0.9214\n",
            "Epoch 817/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 6.5212e-04 - accuracy: 0.9174\n",
            "Epoch 818/1000\n",
            "1/1 [==============================] - 0s 22ms/step - loss: 6.4702e-04 - accuracy: 0.9239\n",
            "Epoch 819/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 6.5070e-04 - accuracy: 0.9238\n",
            "Epoch 820/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 6.6329e-04 - accuracy: 0.9202\n",
            "Epoch 821/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 6.6932e-04 - accuracy: 0.9207\n",
            "Epoch 822/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 6.5632e-04 - accuracy: 0.9216\n",
            "Epoch 823/1000\n",
            "1/1 [==============================] - 0s 22ms/step - loss: 6.4401e-04 - accuracy: 0.9239\n",
            "Epoch 824/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 6.4028e-04 - accuracy: 0.9249\n",
            "Epoch 825/1000\n",
            "1/1 [==============================] - 0s 22ms/step - loss: 6.3834e-04 - accuracy: 0.9206\n",
            "Epoch 826/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 6.3359e-04 - accuracy: 0.9243\n",
            "Epoch 827/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 6.3347e-04 - accuracy: 0.9249\n",
            "Epoch 828/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 6.4056e-04 - accuracy: 0.9217\n",
            "Epoch 829/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 6.4490e-04 - accuracy: 0.9245\n",
            "Epoch 830/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 6.4280e-04 - accuracy: 0.9237\n",
            "Epoch 831/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 6.3904e-04 - accuracy: 0.9250\n",
            "Epoch 832/1000\n",
            "1/1 [==============================] - 0s 23ms/step - loss: 6.3783e-04 - accuracy: 0.9244\n",
            "Epoch 833/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 6.3455e-04 - accuracy: 0.9221\n",
            "Epoch 834/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 6.2902e-04 - accuracy: 0.9250\n",
            "Epoch 835/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 6.2571e-04 - accuracy: 0.9246\n",
            "Epoch 836/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 6.2547e-04 - accuracy: 0.9244\n",
            "Epoch 837/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 6.2685e-04 - accuracy: 0.9248\n",
            "Epoch 838/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 6.2568e-04 - accuracy: 0.9243\n",
            "Epoch 839/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.2505e-04 - accuracy: 0.9248\n",
            "Epoch 840/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 6.2673e-04 - accuracy: 0.9258\n",
            "Epoch 841/1000\n",
            "1/1 [==============================] - 0s 25ms/step - loss: 6.2941e-04 - accuracy: 0.9230\n",
            "Epoch 842/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 6.3120e-04 - accuracy: 0.9261\n",
            "Epoch 843/1000\n",
            "1/1 [==============================] - 0s 22ms/step - loss: 6.3199e-04 - accuracy: 0.9240\n",
            "Epoch 844/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.3490e-04 - accuracy: 0.9254\n",
            "Epoch 845/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 6.4114e-04 - accuracy: 0.9248\n",
            "Epoch 846/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 6.4976e-04 - accuracy: 0.9239\n",
            "Epoch 847/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 6.6058e-04 - accuracy: 0.9249\n",
            "Epoch 848/1000\n",
            "1/1 [==============================] - 0s 22ms/step - loss: 6.7631e-04 - accuracy: 0.9243\n",
            "Epoch 849/1000\n",
            "1/1 [==============================] - 0s 25ms/step - loss: 7.0016e-04 - accuracy: 0.9255\n",
            "Epoch 850/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 7.3469e-04 - accuracy: 0.9250\n",
            "Epoch 851/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 7.7080e-04 - accuracy: 0.9257\n",
            "Epoch 852/1000\n",
            "1/1 [==============================] - 0s 22ms/step - loss: 8.0481e-04 - accuracy: 0.9250\n",
            "Epoch 853/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 7.9701e-04 - accuracy: 0.9250\n",
            "Epoch 854/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 7.5928e-04 - accuracy: 0.9241\n",
            "Epoch 855/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 7.0234e-04 - accuracy: 0.9214\n",
            "Epoch 856/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.6651e-04 - accuracy: 0.9182\n",
            "Epoch 857/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 6.6405e-04 - accuracy: 0.9124\n",
            "Epoch 858/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 6.8019e-04 - accuracy: 0.9108\n",
            "Epoch 859/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 6.7488e-04 - accuracy: 0.9082\n",
            "Epoch 860/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.4784e-04 - accuracy: 0.9236\n",
            "Epoch 861/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 6.3712e-04 - accuracy: 0.9245\n",
            "Epoch 862/1000\n",
            "1/1 [==============================] - 0s 23ms/step - loss: 6.5967e-04 - accuracy: 0.9042\n",
            "Epoch 863/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 6.6360e-04 - accuracy: 0.9045\n",
            "Epoch 864/1000\n",
            "1/1 [==============================] - 0s 22ms/step - loss: 6.3814e-04 - accuracy: 0.9132\n",
            "Epoch 865/1000\n",
            "1/1 [==============================] - 0s 22ms/step - loss: 6.1653e-04 - accuracy: 0.9259\n",
            "Epoch 866/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 6.2826e-04 - accuracy: 0.9209\n",
            "Epoch 867/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 6.4324e-04 - accuracy: 0.9085\n",
            "Epoch 868/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.3466e-04 - accuracy: 0.9204\n",
            "Epoch 869/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 6.2665e-04 - accuracy: 0.9238\n",
            "Epoch 870/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.2275e-04 - accuracy: 0.9226\n",
            "Epoch 871/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 6.2083e-04 - accuracy: 0.9264\n",
            "Epoch 872/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 6.2209e-04 - accuracy: 0.9243\n",
            "Epoch 873/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 6.2373e-04 - accuracy: 0.9238\n",
            "Epoch 874/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.1484e-04 - accuracy: 0.9261\n",
            "Epoch 875/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 6.0784e-04 - accuracy: 0.9259\n",
            "Epoch 876/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 6.1237e-04 - accuracy: 0.9260\n",
            "Epoch 877/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 6.1748e-04 - accuracy: 0.9254\n",
            "Epoch 878/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.1285e-04 - accuracy: 0.9257\n",
            "Epoch 879/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 6.0687e-04 - accuracy: 0.9258\n",
            "Epoch 880/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 6.0892e-04 - accuracy: 0.9262\n",
            "Epoch 881/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 6.1133e-04 - accuracy: 0.9259\n",
            "Epoch 882/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 6.0939e-04 - accuracy: 0.9249\n",
            "Epoch 883/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 6.0826e-04 - accuracy: 0.9264\n",
            "Epoch 884/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.0917e-04 - accuracy: 0.9219\n",
            "Epoch 885/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 6.0868e-04 - accuracy: 0.9264\n",
            "Epoch 886/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 6.0723e-04 - accuracy: 0.9261\n",
            "Epoch 887/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.1065e-04 - accuracy: 0.9257\n",
            "Epoch 888/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.1578e-04 - accuracy: 0.9274\n",
            "Epoch 889/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.1904e-04 - accuracy: 0.9257\n",
            "Epoch 890/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 6.2456e-04 - accuracy: 0.9258\n",
            "Epoch 891/1000\n",
            "1/1 [==============================] - 0s 22ms/step - loss: 6.3853e-04 - accuracy: 0.9264\n",
            "Epoch 892/1000\n",
            "1/1 [==============================] - 0s 23ms/step - loss: 6.6220e-04 - accuracy: 0.9190\n",
            "Epoch 893/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.9574e-04 - accuracy: 0.9207\n",
            "Epoch 894/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 7.4540e-04 - accuracy: 0.9146\n",
            "Epoch 895/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 8.0399e-04 - accuracy: 0.9216\n",
            "Epoch 896/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 8.7921e-04 - accuracy: 0.9177\n",
            "Epoch 897/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 8.8833e-04 - accuracy: 0.9231\n",
            "Epoch 898/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 8.4313e-04 - accuracy: 0.9215\n",
            "Epoch 899/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 7.6283e-04 - accuracy: 0.9260\n",
            "Epoch 900/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 6.5521e-04 - accuracy: 0.9244\n",
            "Epoch 901/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 6.2374e-04 - accuracy: 0.9200\n",
            "Epoch 902/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.7350e-04 - accuracy: 0.9261\n",
            "Epoch 903/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 7.1237e-04 - accuracy: 0.9252\n",
            "Epoch 904/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 7.0989e-04 - accuracy: 0.9236\n",
            "Epoch 905/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 6.6887e-04 - accuracy: 0.9254\n",
            "Epoch 906/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 6.1826e-04 - accuracy: 0.9241\n",
            "Epoch 907/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.1993e-04 - accuracy: 0.9266\n",
            "Epoch 908/1000\n",
            "1/1 [==============================] - 0s 22ms/step - loss: 6.6249e-04 - accuracy: 0.9253\n",
            "Epoch 909/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 6.5403e-04 - accuracy: 0.9232\n",
            "Epoch 910/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 6.2191e-04 - accuracy: 0.9269\n",
            "Epoch 911/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 6.1213e-04 - accuracy: 0.9253\n",
            "Epoch 912/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.0925e-04 - accuracy: 0.9238\n",
            "Epoch 913/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 6.2134e-04 - accuracy: 0.9272\n",
            "Epoch 914/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 6.2705e-04 - accuracy: 0.9249\n",
            "Epoch 915/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 6.0810e-04 - accuracy: 0.9269\n",
            "Epoch 916/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 6.0230e-04 - accuracy: 0.9273\n",
            "Epoch 917/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.0511e-04 - accuracy: 0.9244\n",
            "Epoch 918/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 6.0427e-04 - accuracy: 0.9279\n",
            "Epoch 919/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.0696e-04 - accuracy: 0.9258\n",
            "Epoch 920/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.0666e-04 - accuracy: 0.9259\n",
            "Epoch 921/1000\n",
            "1/1 [==============================] - 0s 24ms/step - loss: 5.9895e-04 - accuracy: 0.9266\n",
            "Epoch 922/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 5.9941e-04 - accuracy: 0.9227\n",
            "Epoch 923/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 5.9471e-04 - accuracy: 0.9274\n",
            "Epoch 924/1000\n",
            "1/1 [==============================] - 0s 22ms/step - loss: 5.9212e-04 - accuracy: 0.9277\n",
            "Epoch 925/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 5.9394e-04 - accuracy: 0.9261\n",
            "Epoch 926/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 5.9508e-04 - accuracy: 0.9276\n",
            "Epoch 927/1000\n",
            "1/1 [==============================] - 0s 22ms/step - loss: 5.9435e-04 - accuracy: 0.9247\n",
            "Epoch 928/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 5.9405e-04 - accuracy: 0.9282\n",
            "Epoch 929/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 5.8853e-04 - accuracy: 0.9261\n",
            "Epoch 930/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 5.8527e-04 - accuracy: 0.9283\n",
            "Epoch 931/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 5.8652e-04 - accuracy: 0.9279\n",
            "Epoch 932/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 5.8679e-04 - accuracy: 0.9265\n",
            "Epoch 933/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 5.8786e-04 - accuracy: 0.9284\n",
            "Epoch 934/1000\n",
            "1/1 [==============================] - 0s 28ms/step - loss: 5.8768e-04 - accuracy: 0.9253\n",
            "Epoch 935/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 5.8583e-04 - accuracy: 0.9291\n",
            "Epoch 936/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 5.8366e-04 - accuracy: 0.9272\n",
            "Epoch 937/1000\n",
            "1/1 [==============================] - 0s 22ms/step - loss: 5.8271e-04 - accuracy: 0.9287\n",
            "Epoch 938/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 5.8152e-04 - accuracy: 0.9290\n",
            "Epoch 939/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 5.8067e-04 - accuracy: 0.9272\n",
            "Epoch 940/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 5.8085e-04 - accuracy: 0.9291\n",
            "Epoch 941/1000\n",
            "1/1 [==============================] - 0s 25ms/step - loss: 5.7997e-04 - accuracy: 0.9267\n",
            "Epoch 942/1000\n",
            "1/1 [==============================] - 0s 22ms/step - loss: 5.7892e-04 - accuracy: 0.9290\n",
            "Epoch 943/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 5.7870e-04 - accuracy: 0.9279\n",
            "Epoch 944/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 5.7873e-04 - accuracy: 0.9286\n",
            "Epoch 945/1000\n",
            "1/1 [==============================] - 0s 22ms/step - loss: 5.7899e-04 - accuracy: 0.9288\n",
            "Epoch 946/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 5.7948e-04 - accuracy: 0.9285\n",
            "Epoch 947/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 5.8037e-04 - accuracy: 0.9295\n",
            "Epoch 948/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 5.8134e-04 - accuracy: 0.9285\n",
            "Epoch 949/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 5.8280e-04 - accuracy: 0.9290\n",
            "Epoch 950/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 5.8590e-04 - accuracy: 0.9290\n",
            "Epoch 951/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 5.9060e-04 - accuracy: 0.9274\n",
            "Epoch 952/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 5.9840e-04 - accuracy: 0.9290\n",
            "Epoch 953/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 6.1023e-04 - accuracy: 0.9261\n",
            "Epoch 954/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 6.2809e-04 - accuracy: 0.9276\n",
            "Epoch 955/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 6.5486e-04 - accuracy: 0.9239\n",
            "Epoch 956/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 6.8861e-04 - accuracy: 0.9252\n",
            "Epoch 957/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 7.2677e-04 - accuracy: 0.9193\n",
            "Epoch 958/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 7.4954e-04 - accuracy: 0.9245\n",
            "Epoch 959/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 7.4860e-04 - accuracy: 0.9229\n",
            "Epoch 960/1000\n",
            "1/1 [==============================] - 0s 26ms/step - loss: 7.1730e-04 - accuracy: 0.9282\n",
            "Epoch 961/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 6.6297e-04 - accuracy: 0.9275\n",
            "Epoch 962/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 6.1780e-04 - accuracy: 0.9185\n",
            "Epoch 963/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 5.9832e-04 - accuracy: 0.9210\n",
            "Epoch 964/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 6.0583e-04 - accuracy: 0.9180\n",
            "Epoch 965/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 6.2500e-04 - accuracy: 0.9283\n",
            "Epoch 966/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 6.3911e-04 - accuracy: 0.9289\n",
            "Epoch 967/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 6.3540e-04 - accuracy: 0.9231\n",
            "Epoch 968/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 6.1201e-04 - accuracy: 0.9236\n",
            "Epoch 969/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 5.8635e-04 - accuracy: 0.9193\n",
            "Epoch 970/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 5.7742e-04 - accuracy: 0.9288\n",
            "Epoch 971/1000\n",
            "1/1 [==============================] - 0s 27ms/step - loss: 5.9084e-04 - accuracy: 0.9295\n",
            "Epoch 972/1000\n",
            "1/1 [==============================] - 0s 22ms/step - loss: 6.0935e-04 - accuracy: 0.9251\n",
            "Epoch 973/1000\n",
            "1/1 [==============================] - 0s 23ms/step - loss: 6.1133e-04 - accuracy: 0.9266\n",
            "Epoch 974/1000\n",
            "1/1 [==============================] - 0s 26ms/step - loss: 5.9416e-04 - accuracy: 0.9227\n",
            "Epoch 975/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 5.7298e-04 - accuracy: 0.9302\n",
            "Epoch 976/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 5.6686e-04 - accuracy: 0.9297\n",
            "Epoch 977/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 5.7602e-04 - accuracy: 0.9256\n",
            "Epoch 978/1000\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 5.8691e-04 - accuracy: 0.9286\n",
            "Epoch 979/1000\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 5.8754e-04 - accuracy: 0.9258\n",
            "Epoch 980/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 5.7904e-04 - accuracy: 0.9300\n",
            "Epoch 981/1000\n",
            "1/1 [==============================] - 0s 23ms/step - loss: 5.7159e-04 - accuracy: 0.9299\n",
            "Epoch 982/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 5.6943e-04 - accuracy: 0.9248\n",
            "Epoch 983/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 5.7043e-04 - accuracy: 0.9277\n",
            "Epoch 984/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 5.7091e-04 - accuracy: 0.9253\n",
            "Epoch 985/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 5.6976e-04 - accuracy: 0.9293\n",
            "Epoch 986/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 5.6746e-04 - accuracy: 0.9289\n",
            "Epoch 987/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 5.6567e-04 - accuracy: 0.9276\n",
            "Epoch 988/1000\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 5.6548e-04 - accuracy: 0.9296\n",
            "Epoch 989/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 5.6672e-04 - accuracy: 0.9258\n",
            "Epoch 990/1000\n",
            "1/1 [==============================] - 0s 25ms/step - loss: 5.6854e-04 - accuracy: 0.9282\n",
            "Epoch 991/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 5.6838e-04 - accuracy: 0.9266\n",
            "Epoch 992/1000\n",
            "1/1 [==============================] - 0s 21ms/step - loss: 5.6585e-04 - accuracy: 0.9288\n",
            "Epoch 993/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 5.6188e-04 - accuracy: 0.9286\n",
            "Epoch 994/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 5.5918e-04 - accuracy: 0.9299\n",
            "Epoch 995/1000\n",
            "1/1 [==============================] - 0s 17ms/step - loss: 5.5882e-04 - accuracy: 0.9302\n",
            "Epoch 996/1000\n",
            "1/1 [==============================] - 0s 22ms/step - loss: 5.6000e-04 - accuracy: 0.9290\n",
            "Epoch 997/1000\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 5.6060e-04 - accuracy: 0.9298\n",
            "Epoch 998/1000\n",
            "1/1 [==============================] - 0s 18ms/step - loss: 5.5939e-04 - accuracy: 0.9291\n",
            "Epoch 999/1000\n",
            "1/1 [==============================] - 0s 20ms/step - loss: 5.5737e-04 - accuracy: 0.9304\n",
            "Epoch 1000/1000\n",
            "1/1 [==============================] - 0s 23ms/step - loss: 5.5550e-04 - accuracy: 0.9297\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tensorflow.python.keras.callbacks.History at 0x7f6fa028c450>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "D0eAdFCki5kC"
      },
      "source": [
        "pred = model.predict(img_array)\n",
        "\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 216
        },
        "id": "WMb2XOH0i_si",
        "outputId": "61b39adc-873d-4404-ccc1-cd22f97c1a5a"
      },
      "source": [
        "plt.subplot(1,2,1)\n",
        "plt.imshow(img)\n",
        "plt.title('Original')\n",
        "plt.subplot(1,2,2)\n",
        "plt.imshow(pred[0].reshape(SIZE,SIZE,3))\n",
        "plt.title('Reconstructed')\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    }
  ]
}