image_colorization.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "image_colourization(my).ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "ajYBLuKhXRAt"
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "import cv2\n",
        "import numpy as np \n",
        "import pandas as pd\n",
        "from tensorflow.keras.models import Model, Sequential\n",
        "from tensorflow.keras.layers import Conv2D, Conv2DTranspose, Reshape,LeakyReLU, Dropout\n",
        "import tensorflow as tf\n",
        "from tensorflow.keras.layers import AveragePooling2D,UpSampling2D\n",
        "from tensorflow import keras"
      ],
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4ehVdUEbXVhP"
      },
      "source": [
        "ab = np.load('/content/drive/MyDrive/colorization data/ab1.npy')\n",
        "gray = np.load('/content/drive/MyDrive/colorization data/gray_scale.npy')"
      ],
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JMXnAZpdXcQ2"
      },
      "source": [
        "def batch_prep(gray_img , batch_size = 100):\n",
        "  img = np.zeros((batch_size , 224 , 224 , 3))\n",
        "  for i in range(0,3):\n",
        "    img[:batch_size , : , : , i] = gray_img[:batch_size]\n",
        "    return img\n",
        "img_in = batch_prep(gray , batch_size = 300)"
      ],
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 288
        },
        "id": "69MfcNzmcNFH",
        "outputId": "f0243067-c18c-41bb-89d9-9a33d86a9ee1"
      },
      "source": [
        "plt.imshow(gray[29] , cmap = plt.cm.gray)"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7f6ddd58c048>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 4
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "__-lY1KjcZvQ",
        "outputId": "ea216eb5-543e-40c3-b269-c2240beca6bf"
      },
      "source": [
        "gray.shape"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(25000, 224, 224)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "2EkKlp7yckqR",
        "outputId": "fe6f1f5f-8d21-49c6-f136-7ca95551fd4a"
      },
      "source": [
        "img_in.shape"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(300, 224, 224, 3)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_GqNM-h2cm5x"
      },
      "source": [
        "def get_rbg(gray_imgs , ab_imgs , n=10):\n",
        "  img1 = np.zeros((n , 224 , 224 , 3))\n",
        "  img1[:,:,:,0] = gray_imgs[0:n:]\n",
        "  img1[:,:,:,1:] = ab_imgs[0:n]\n",
        "  img1 = img1.astype(np.uint8)\n",
        "  imgs = []\n",
        "  for i in range(0,n):\n",
        "    imgs.append(cv2.cvtColor(img1[i] , cv2.COLOR_LAB2RGB))\n",
        "  imgs = np.array(imgs)\n",
        "  return imgs\n",
        "\n",
        "  "
      ],
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qph3v4wqelBt"
      },
      "source": [
        "img_out = get_rbg(gray_imgs = gray , ab_imgs = ab , n=300)"
      ],
      "execution_count": 8,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_4MwYh7Eeqmk"
      },
      "source": [
        "model = Sequential()\n",
        "model.add(Conv2D(strides = 1 , kernel_size = 3, filters = 12, use_bias = True , bias_initializer = tf.keras.initializers.RandomUniform(minval = -0.05 , maxval = 0.05), padding = \"valid\" , activation = tf.nn.relu))\n",
        "model.add(Conv2DTranspose(strides = 1 , kernel_size = 3, filters = 12, use_bias = True , bias_initializer = tf.keras.initializers.RandomUniform(minval = -0.05 , maxval = 0.05), padding = \"valid\" , activation = tf.nn.relu))\n",
        "model.add(LeakyReLU(0.6))\n",
        "model.add(Dropout(0.4))\n",
        "\n",
        "model = Sequential()\n",
        "model.add(Conv2D(strides = 1 , kernel_size = 3, filters = 12, use_bias = True , bias_initializer = tf.keras.initializers.RandomUniform(minval = -0.05 , maxval = 0.05), padding = \"valid\" , activation = tf.nn.relu))\n",
        "model.add(Conv2DTranspose(strides = 1 , kernel_size = 3, filters = 3, use_bias = True , bias_initializer = tf.keras.initializers.RandomUniform(minval = -0.05 , maxval = 0.05), padding = \"valid\" , activation = tf.nn.relu))\n",
        "model.add(LeakyReLU(0.6))\n",
        "model.add(Dropout(0.4))\n",
        "\n",
        "model = Sequential()\n",
        "model.add(Conv2D(strides = 1 , kernel_size = 3, filters = 12, use_bias = True , bias_initializer = tf.keras.initializers.RandomUniform(minval = -0.05 , maxval = 0.05), padding = \"valid\" , activation = tf.nn.relu))\n",
        "model.add(Conv2DTranspose(strides = 1 , kernel_size = 3, filters = 3, use_bias = True , bias_initializer = tf.keras.initializers.RandomUniform(minval = -0.05 , maxval = 0.05), padding = \"valid\" , activation = tf.nn.relu))\n",
        "model.add(LeakyReLU(0.6))\n",
        "model.add(Dropout(0.4))\n",
        "\n",
        "model = Sequential()\n",
        "model.add(Conv2D(strides = 1 , kernel_size = 3, filters = 12, use_bias = True , bias_initializer = tf.keras.initializers.RandomUniform(minval = -0.05 , maxval = 0.05), padding = \"valid\" , activation = tf.nn.relu))\n",
        "model.add(Conv2DTranspose(strides = 1 , kernel_size = 3, filters = 3, use_bias = True , bias_initializer = tf.keras.initializers.RandomUniform(minval = -0.05 , maxval = 0.05), padding = \"valid\" , activation = tf.nn.relu))\n",
        "model.add(AveragePooling2D(pool_size = (2,2)))\n",
        "model.add(UpSampling2D((2,2)))\n",
        "model.add(LeakyReLU(0.6))\n",
        "model.add(Dropout(0.4))\n",
        "\n",
        "model = Sequential()\n",
        "model.add(Conv2D(strides = 1 , kernel_size = 3, filters = 12, use_bias = True , bias_initializer = tf.keras.initializers.RandomUniform(minval = -0.05 , maxval = 0.05), padding = \"valid\" , activation = tf.nn.relu))\n",
        "model.add(Conv2DTranspose(strides = 1 , kernel_size = 3, filters = 3, use_bias = True , bias_initializer = tf.keras.initializers.RandomUniform(minval = -0.05 , maxval = 0.05), padding = \"valid\" , activation = tf.nn.relu))\n",
        "model.add(LeakyReLU(0.6))\n",
        "model.add(Dropout(0.4))"
      ],
      "execution_count": 9,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lux6Nvyhqz7i"
      },
      "source": [
        "model.compile(optimizer = tf.keras.optimizers.Adam() , loss='mape' , metrics = tf.keras.metrics.Accuracy())"
      ],
      "execution_count": 10,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "MlSGpra_syA6",
        "outputId": "21eb7831-9921-4e94-ebde-7cfec3b499af"
      },
      "source": [
        "model.fit(img_in , img_out , epochs=4000, batch_size=16)"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\u001b[1;30;43mStreaming output truncated to the last 5000 lines.\u001b[0m\n",
            "Epoch 1501/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1502/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1503/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1504/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1505/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1506/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1507/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1508/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1509/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1510/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1511/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1512/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1513/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1514/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1515/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1516/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1517/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1518/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1519/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1520/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1521/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1522/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1523/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1524/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1525/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1526/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1527/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1528/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1529/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1530/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1531/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1532/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1533/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1534/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1535/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1536/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1537/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1538/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1539/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1540/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1541/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1542/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1543/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1544/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1545/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1546/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1547/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1548/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1549/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1550/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1551/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1552/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1553/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1554/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1555/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1556/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1557/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1558/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1559/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1560/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1561/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1562/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1563/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1564/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1565/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1566/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1567/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1568/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1569/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1570/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1571/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1572/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1573/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1574/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1575/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1576/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1577/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1578/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1579/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1580/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1581/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1582/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1583/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1584/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1585/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1586/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1587/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1588/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1589/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1590/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1591/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1592/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1593/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1594/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1595/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1596/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1597/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1598/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1599/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1600/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1601/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1602/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1603/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1604/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1605/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1606/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1607/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1608/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1609/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1610/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1611/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1612/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1613/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1614/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1615/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1616/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1617/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1618/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1619/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1620/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1621/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1622/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1623/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1624/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1625/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1626/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1627/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1628/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1629/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1630/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1631/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1632/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1633/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1634/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1635/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1636/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1637/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1638/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1639/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1640/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1641/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1642/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1643/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1644/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1645/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1646/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1647/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1648/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1649/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1650/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1651/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1652/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1653/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1654/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1655/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1656/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1657/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1658/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1659/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1660/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1661/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1662/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1663/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1664/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1665/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1666/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1667/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1668/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1669/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1670/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1671/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1672/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1673/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1674/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1675/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1676/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1677/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1678/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1679/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1680/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1681/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1682/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1683/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1684/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1685/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1686/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1687/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1688/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1689/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1690/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1691/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1692/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1693/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1694/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1695/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1696/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1697/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1698/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1699/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1700/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1701/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1702/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1703/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1704/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1705/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1706/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1707/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1708/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1709/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1710/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1711/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1712/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1713/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1714/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1715/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1716/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1717/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1718/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1719/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1720/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1721/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1722/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1723/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1724/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1725/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1726/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1727/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1728/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1729/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1730/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1731/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1732/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1733/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1734/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1735/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1736/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1737/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1738/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1739/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1740/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1741/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1742/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1743/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1744/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1745/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1746/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1747/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1748/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1749/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1750/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1751/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1752/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1753/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1754/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1755/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1756/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1757/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1758/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1759/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1760/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1761/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1762/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1763/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1764/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1765/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1766/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1767/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1768/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1769/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1770/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1771/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1772/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1773/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1774/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1775/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1776/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1777/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1778/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1779/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1780/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1781/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1782/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1783/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1784/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1785/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1786/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1787/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1788/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1789/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1790/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1791/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1792/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1793/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1794/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1795/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1796/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1797/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1798/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1799/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1800/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1801/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1802/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1803/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1804/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1805/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1806/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1807/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1808/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1809/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1810/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1811/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1812/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1813/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1814/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1815/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1816/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1817/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1818/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1819/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1820/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1821/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1822/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1823/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1824/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1825/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1826/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1827/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1828/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1829/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1830/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1831/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1832/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1833/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1834/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1835/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1836/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1837/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1838/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1839/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1840/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1841/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1842/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1843/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1844/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1845/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1846/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1847/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1848/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1849/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1850/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1851/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1852/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1853/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1854/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1855/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1856/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1857/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1858/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1859/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1860/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1861/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1862/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1863/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1864/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1865/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1866/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1867/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1868/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1869/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1870/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1871/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1872/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1873/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1874/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1875/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1876/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1877/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1878/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1879/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1880/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1881/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1882/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1883/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1884/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1885/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1886/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1887/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1888/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1889/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1890/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1891/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1892/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1893/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1894/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1895/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1896/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1897/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1898/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1899/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1900/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1901/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1902/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1903/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1904/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1905/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1906/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1907/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1908/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1909/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1910/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1911/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1912/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1913/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1914/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1915/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1916/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1917/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1918/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1919/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1920/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1921/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1922/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1923/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1924/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1925/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1926/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1927/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1928/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1929/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1930/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1931/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1932/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1933/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1934/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1935/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1936/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1937/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1938/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1939/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1940/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1941/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1942/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1943/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1944/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1945/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1946/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1947/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1948/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1949/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1950/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1951/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1952/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1953/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1954/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1955/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1956/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1957/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1958/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1959/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1960/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1961/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1962/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1963/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1964/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1965/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1966/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1967/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1968/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1969/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1970/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1971/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1972/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1973/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1974/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1975/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1976/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1977/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1978/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1979/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1980/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1981/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1982/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1983/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1984/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1985/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1986/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1987/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1988/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1989/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1990/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1991/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1992/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1993/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1994/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1995/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1996/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1997/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1998/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 1999/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2000/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2001/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2002/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2003/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2004/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2005/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2006/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2007/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2008/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2009/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2010/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2011/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2012/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2013/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2014/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2015/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2016/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2017/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2018/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2019/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2020/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2021/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2022/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2023/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2024/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2025/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2026/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2027/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2028/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2029/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2030/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2031/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2032/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2033/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2034/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2035/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2036/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2037/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2038/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2039/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2040/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2041/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2042/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2043/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2044/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2045/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2046/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2047/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2048/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2049/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2050/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2051/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2052/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2053/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2054/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2055/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2056/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2057/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2058/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2059/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2060/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2061/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2062/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2063/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2064/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2065/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2066/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2067/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2068/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2069/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2070/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2071/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2072/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2073/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2074/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2075/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2076/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2077/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2078/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2079/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2080/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2081/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2082/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2083/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2084/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2085/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2086/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2087/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2088/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2089/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2090/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2091/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2092/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2093/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2094/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2095/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2096/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2097/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2098/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2099/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2100/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2101/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2102/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2103/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2104/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2105/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2106/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2107/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2108/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2109/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2110/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2111/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2112/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2113/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2114/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2115/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2116/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2117/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2118/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2119/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2120/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2121/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2122/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2123/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2124/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2125/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2126/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2127/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2128/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2129/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2130/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2131/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2132/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2133/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2134/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2135/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2136/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2137/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2138/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2139/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2140/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2141/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2142/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2143/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2144/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2145/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2146/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2147/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2148/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2149/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2150/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2151/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2152/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2153/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2154/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2155/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2156/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2157/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2158/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2159/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2160/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2161/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2162/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2163/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2164/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2165/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2166/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2167/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2168/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2169/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2170/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2171/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2172/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2173/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2174/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2175/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2176/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2177/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2178/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2179/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2180/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2181/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2182/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2183/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2184/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2185/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2186/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2187/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2188/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2189/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2190/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2191/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2192/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2193/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2194/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2195/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2196/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2197/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2198/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2199/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2200/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2201/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2202/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2203/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2204/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2205/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2206/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2207/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2208/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2209/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2210/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2211/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2212/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2213/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2214/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2215/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2216/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2217/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2218/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2219/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2220/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2221/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2222/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2223/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2224/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2225/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2226/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2227/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2228/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2229/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2230/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2231/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2232/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2233/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2234/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2235/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2236/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2237/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2238/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2239/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2240/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2241/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2242/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2243/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2244/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2245/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2246/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2247/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2248/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2249/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2250/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2251/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2252/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2253/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2254/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2255/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2256/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2257/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2258/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2259/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2260/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2261/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2262/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2263/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2264/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2265/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2266/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2267/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2268/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2269/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2270/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2271/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2272/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2273/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2274/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2275/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2276/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2277/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2278/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2279/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2280/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2281/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2282/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2283/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2284/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2285/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2286/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2287/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2288/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2289/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2290/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2291/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2292/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2293/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2294/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2295/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2296/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2297/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2298/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2299/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2300/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2301/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2302/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2303/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2304/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2305/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2306/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2307/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2308/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2309/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2310/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2311/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2312/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2313/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2314/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2315/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2316/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2317/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2318/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2319/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2320/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2321/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2322/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2323/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2324/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2325/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2326/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2327/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2328/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2329/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2330/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2331/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2332/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2333/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2334/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2335/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2336/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2337/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2338/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2339/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2340/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2341/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2342/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2343/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2344/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2345/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2346/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2347/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2348/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2349/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2350/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2351/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2352/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2353/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2354/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2355/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2356/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2357/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2358/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2359/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2360/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2361/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2362/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2363/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2364/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2365/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2366/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2367/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2368/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2369/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2370/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2371/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2372/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2373/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2374/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2375/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2376/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2377/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2378/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2379/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2380/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2381/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2382/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2383/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2384/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2385/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2386/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2387/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2388/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2389/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2390/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2391/4000\n",
            "19/19 [==============================] - 0s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2392/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2393/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2394/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2395/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2396/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2397/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2398/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2399/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2400/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2401/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2402/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2403/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2404/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2405/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2406/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2407/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2408/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2409/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2410/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2411/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2412/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2413/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2414/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2415/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2416/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2417/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2418/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2419/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2420/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2421/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2422/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2423/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2424/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2425/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2426/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2427/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2428/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2429/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2430/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2431/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2432/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2433/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2434/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2435/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2436/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2437/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2438/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2439/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2440/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2441/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2442/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2443/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2444/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2445/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2446/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2447/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2448/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2449/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2450/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2451/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2452/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2453/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2454/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2455/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2456/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2457/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2458/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2459/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2460/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2461/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2462/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2463/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2464/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2465/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2466/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2467/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2468/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2469/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2470/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2471/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2472/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2473/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2474/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2475/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2476/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2477/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2478/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2479/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2480/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2481/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2482/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2483/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2484/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2485/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2486/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2487/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2488/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2489/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2490/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2491/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2492/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2493/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2494/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2495/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2496/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2497/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2498/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2499/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2500/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2501/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2502/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2503/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2504/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2505/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2506/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2507/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2508/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2509/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2510/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2511/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2512/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2513/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2514/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2515/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2516/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2517/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2518/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2519/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2520/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2521/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2522/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2523/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2524/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2525/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2526/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2527/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2528/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2529/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2530/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2531/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2532/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2533/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2534/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2535/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2536/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2537/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2538/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2539/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2540/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2541/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2542/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2543/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2544/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2545/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2546/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2547/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2548/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2549/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2550/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2551/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2552/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2553/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2554/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2555/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2556/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2557/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2558/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2559/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2560/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2561/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2562/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2563/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2564/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2565/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2566/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2567/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2568/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2569/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2570/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2571/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2572/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2573/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2574/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2575/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2576/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2577/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2578/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2579/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2580/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2581/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2582/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2583/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2584/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2585/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2586/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2587/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2588/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2589/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2590/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2591/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2592/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2593/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2594/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2595/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2596/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2597/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2598/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2599/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2600/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2601/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2602/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2603/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2604/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2605/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2606/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2607/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2608/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2609/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2610/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2611/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2612/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2613/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2614/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2615/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2616/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2617/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2618/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2619/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2620/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2621/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2622/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2623/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2624/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2625/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2626/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2627/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2628/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2629/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2630/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2631/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2632/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2633/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2634/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2635/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2636/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2637/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2638/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2639/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2640/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2641/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2642/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2643/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2644/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2645/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2646/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2647/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2648/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2649/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2650/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2651/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2652/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2653/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2654/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2655/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2656/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2657/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2658/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2659/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2660/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2661/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2662/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2663/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2664/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2665/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2666/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2667/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2668/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2669/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2670/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2671/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2672/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2673/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2674/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2675/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2676/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2677/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2678/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2679/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2680/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2681/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2682/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2683/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2684/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2685/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2686/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2687/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2688/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2689/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2690/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2691/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2692/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2693/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2694/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2695/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2696/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2697/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2698/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2699/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2700/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2701/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2702/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2703/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2704/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2705/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2706/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2707/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2708/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2709/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2710/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2711/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2712/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2713/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2714/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2715/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2716/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2717/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2718/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2719/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2720/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2721/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2722/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2723/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2724/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2725/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2726/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2727/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2728/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2729/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2730/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2731/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2732/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2733/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2734/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2735/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2736/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2737/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2738/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2739/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2740/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2741/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2742/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2743/4000\n",
            "19/19 [==============================] - 1s 31ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2744/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2745/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2746/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2747/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2748/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2749/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2750/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2751/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2752/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2753/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2754/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2755/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2756/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2757/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2758/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2759/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2760/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2761/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2762/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2763/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2764/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2765/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2766/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2767/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2768/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2769/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2770/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2771/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2772/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2773/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2774/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2775/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2776/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2777/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2778/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2779/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2780/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2781/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2782/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2783/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2784/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2785/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2786/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2787/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2788/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2789/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2790/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2791/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2792/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2793/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2794/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2795/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2796/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2797/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2798/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2799/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2800/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2801/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2802/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2803/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2804/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2805/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2806/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2807/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2808/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2809/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2810/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2811/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2812/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2813/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2814/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2815/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2816/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2817/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2818/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2819/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2820/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2821/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2822/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2823/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2824/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2825/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2826/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2827/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2828/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2829/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2830/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2831/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2832/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2833/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2834/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2835/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2836/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2837/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2838/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2839/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2840/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2841/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2842/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2843/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2844/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2845/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2846/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2847/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2848/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2849/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2850/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2851/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2852/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2853/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2854/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2855/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2856/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2857/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2858/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2859/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2860/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2861/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2862/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2863/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2864/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2865/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2866/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2867/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2868/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2869/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2870/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2871/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2872/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2873/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2874/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2875/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2876/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2877/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2878/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2879/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2880/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2881/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2882/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2883/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2884/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2885/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2886/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2887/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2888/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2889/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2890/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2891/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2892/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2893/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2894/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2895/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2896/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2897/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2898/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2899/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2900/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2901/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2902/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2903/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2904/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2905/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2906/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2907/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2908/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2909/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2910/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2911/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2912/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2913/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2914/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2915/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2916/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2917/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2918/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2919/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2920/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2921/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2922/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2923/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2924/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2925/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2926/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2927/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2928/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2929/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2930/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2931/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2932/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2933/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2934/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2935/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2936/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2937/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2938/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2939/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2940/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2941/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2942/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2943/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2944/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2945/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2946/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2947/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2948/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2949/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2950/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2951/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2952/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2953/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2954/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2955/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2956/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2957/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2958/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2959/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2960/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2961/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2962/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2963/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2964/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2965/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2966/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2967/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2968/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2969/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2970/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2971/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2972/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2973/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2974/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2975/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2976/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2977/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2978/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2979/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2980/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2981/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2982/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2983/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2984/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2985/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2986/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2987/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2988/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2989/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2990/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2991/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2992/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2993/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2994/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2995/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2996/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2997/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2998/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 2999/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3000/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3001/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3002/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3003/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3004/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3005/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3006/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3007/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3008/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3009/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3010/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3011/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3012/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3013/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3014/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3015/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3016/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3017/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3018/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3019/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3020/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3021/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3022/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3023/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3024/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3025/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3026/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3027/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3028/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3029/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3030/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3031/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3032/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3033/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3034/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3035/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3036/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3037/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3038/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3039/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3040/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3041/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3042/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3043/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3044/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3045/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3046/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3047/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3048/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3049/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3050/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3051/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3052/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3053/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3054/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3055/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3056/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3057/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3058/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3059/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3060/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3061/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3062/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3063/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3064/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3065/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3066/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3067/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3068/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3069/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3070/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3071/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3072/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3073/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3074/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3075/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3076/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3077/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3078/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3079/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3080/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3081/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3082/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3083/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3084/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3085/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3086/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3087/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3088/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3089/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3090/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3091/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3092/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3093/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3094/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3095/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3096/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3097/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3098/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3099/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3100/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3101/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3102/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3103/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3104/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3105/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3106/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3107/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3108/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3109/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3110/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3111/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3112/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3113/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3114/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3115/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3116/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3117/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3118/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3119/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3120/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3121/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3122/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3123/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3124/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3125/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3126/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3127/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3128/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3129/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3130/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3131/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3132/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3133/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3134/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3135/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3136/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3137/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3138/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3139/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3140/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3141/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3142/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3143/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3144/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3145/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3146/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3147/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3148/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3149/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3150/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3151/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3152/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3153/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3154/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3155/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3156/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3157/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3158/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3159/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3160/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3161/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3162/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3163/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3164/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3165/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3166/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3167/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3168/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3169/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3170/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3171/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3172/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3173/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3174/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3175/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3176/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3177/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3178/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3179/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3180/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3181/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3182/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3183/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3184/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3185/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3186/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3187/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3188/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3189/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3190/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3191/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3192/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3193/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3194/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3195/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3196/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3197/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3198/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3199/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3200/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3201/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3202/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3203/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3204/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3205/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3206/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3207/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3208/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3209/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3210/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3211/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3212/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3213/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3214/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3215/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3216/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3217/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3218/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3219/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3220/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3221/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3222/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3223/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3224/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3225/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3226/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3227/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3228/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3229/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3230/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3231/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3232/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3233/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3234/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3235/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3236/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3237/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3238/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3239/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3240/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3241/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3242/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3243/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3244/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3245/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3246/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3247/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3248/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3249/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3250/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3251/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3252/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3253/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3254/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3255/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3256/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3257/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3258/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3259/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3260/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3261/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3262/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3263/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3264/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3265/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3266/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3267/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3268/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3269/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3270/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3271/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3272/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3273/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3274/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3275/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3276/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3277/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3278/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3279/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3280/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3281/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3282/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3283/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3284/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3285/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3286/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3287/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3288/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3289/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3290/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3291/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3292/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3293/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3294/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3295/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3296/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3297/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3298/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3299/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3300/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3301/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3302/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3303/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3304/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3305/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3306/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3307/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3308/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3309/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3310/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3311/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3312/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3313/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3314/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3315/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3316/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3317/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3318/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3319/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3320/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3321/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3322/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3323/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3324/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3325/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3326/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3327/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3328/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3329/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3330/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3331/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3332/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3333/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3334/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3335/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3336/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3337/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3338/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3339/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3340/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3341/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3342/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3343/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3344/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3345/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3346/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3347/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3348/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3349/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3350/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3351/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3352/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3353/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3354/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3355/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3356/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3357/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3358/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3359/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3360/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3361/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3362/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3363/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3364/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3365/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3366/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3367/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3368/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3369/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3370/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3371/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3372/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3373/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3374/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3375/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3376/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3377/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3378/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3379/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3380/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3381/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3382/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3383/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3384/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3385/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3386/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3387/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3388/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3389/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3390/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3391/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3392/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3393/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3394/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3395/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3396/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3397/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3398/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3399/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3400/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3401/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3402/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3403/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3404/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3405/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3406/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3407/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3408/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3409/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3410/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3411/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3412/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3413/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3414/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3415/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3416/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3417/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3418/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3419/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3420/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3421/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3422/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3423/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3424/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3425/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3426/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3427/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3428/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3429/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3430/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3431/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3432/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3433/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3434/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3435/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3436/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3437/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3438/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3439/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3440/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3441/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3442/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3443/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3444/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3445/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3446/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3447/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3448/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3449/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3450/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3451/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3452/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3453/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3454/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3455/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3456/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3457/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3458/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3459/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3460/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3461/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3462/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3463/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3464/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3465/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3466/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3467/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3468/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3469/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3470/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3471/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3472/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3473/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3474/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3475/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3476/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3477/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3478/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3479/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3480/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3481/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3482/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3483/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3484/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3485/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3486/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3487/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3488/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3489/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3490/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3491/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3492/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3493/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3494/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3495/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3496/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3497/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3498/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3499/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3500/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3501/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3502/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3503/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3504/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3505/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3506/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3507/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3508/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3509/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3510/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3511/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3512/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3513/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3514/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3515/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3516/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3517/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3518/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3519/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3520/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3521/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3522/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3523/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3524/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3525/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3526/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3527/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3528/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3529/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3530/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3531/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3532/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3533/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3534/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3535/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3536/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3537/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3538/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3539/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3540/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3541/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3542/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3543/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3544/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3545/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3546/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3547/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3548/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3549/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3550/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3551/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3552/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3553/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3554/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3555/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3556/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3557/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3558/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3559/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3560/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3561/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3562/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3563/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3564/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3565/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3566/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3567/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3568/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3569/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3570/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3571/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3572/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3573/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3574/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3575/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3576/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3577/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3578/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3579/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3580/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3581/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3582/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3583/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3584/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3585/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3586/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3587/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3588/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3589/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3590/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3591/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3592/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3593/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3594/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3595/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3596/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3597/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3598/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3599/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3600/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3601/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3602/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3603/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3604/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3605/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3606/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3607/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3608/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3609/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3610/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3611/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3612/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3613/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3614/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3615/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3616/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3617/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3618/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3619/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3620/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3621/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3622/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3623/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3624/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3625/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3626/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3627/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3628/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3629/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3630/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3631/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3632/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3633/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3634/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3635/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3636/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3637/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3638/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3639/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3640/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3641/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3642/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3643/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3644/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3645/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3646/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3647/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3648/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3649/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3650/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3651/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3652/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3653/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3654/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3655/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3656/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3657/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3658/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3659/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3660/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3661/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3662/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3663/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3664/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3665/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3666/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3667/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3668/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3669/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3670/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3671/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3672/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3673/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3674/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3675/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3676/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3677/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3678/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3679/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3680/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3681/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3682/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3683/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3684/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3685/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3686/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3687/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3688/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3689/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3690/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3691/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3692/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3693/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3694/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3695/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3696/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3697/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3698/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3699/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3700/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3701/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3702/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3703/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3704/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3705/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3706/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3707/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3708/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3709/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3710/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3711/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3712/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3713/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3714/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3715/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3716/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3717/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3718/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3719/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3720/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3721/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3722/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3723/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3724/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3725/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3726/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3727/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3728/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3729/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3730/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3731/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3732/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3733/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3734/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3735/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3736/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3737/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3738/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3739/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3740/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3741/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3742/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3743/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3744/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3745/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3746/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3747/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3748/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3749/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3750/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3751/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3752/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3753/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3754/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3755/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3756/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3757/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3758/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3759/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3760/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3761/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3762/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3763/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3764/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3765/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3766/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3767/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3768/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3769/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3770/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3771/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3772/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3773/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3774/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3775/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3776/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3777/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3778/4000\n",
            "19/19 [==============================] - 1s 31ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3779/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3780/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3781/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3782/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3783/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3784/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3785/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3786/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3787/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3788/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3789/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3790/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3791/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3792/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3793/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3794/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3795/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3796/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3797/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3798/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3799/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3800/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3801/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3802/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3803/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3804/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3805/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3806/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3807/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3808/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3809/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3810/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3811/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3812/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3813/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3814/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3815/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3816/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3817/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3818/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3819/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3820/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3821/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3822/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3823/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3824/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3825/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3826/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3827/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3828/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3829/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3830/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3831/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3832/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3833/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3834/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3835/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3836/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3837/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3838/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3839/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3840/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3841/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3842/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3843/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3844/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3845/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3846/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3847/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3848/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3849/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3850/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3851/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3852/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3853/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3854/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3855/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3856/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3857/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3858/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3859/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3860/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3861/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3862/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3863/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3864/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3865/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3866/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3867/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3868/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3869/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3870/4000\n",
            "19/19 [==============================] - 1s 26ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3871/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3872/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3873/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3874/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3875/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3876/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3877/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3878/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3879/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3880/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3881/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3882/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3883/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3884/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3885/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3886/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3887/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3888/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3889/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3890/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3891/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3892/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3893/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3894/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3895/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3896/4000\n",
            "19/19 [==============================] - 1s 31ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3897/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3898/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3899/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3900/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3901/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3902/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3903/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3904/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3905/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3906/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3907/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3908/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3909/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3910/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3911/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3912/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3913/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3914/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3915/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3916/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3917/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3918/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3919/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3920/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3921/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3922/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3923/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3924/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3925/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3926/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3927/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3928/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3929/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3930/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3931/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3932/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3933/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3934/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3935/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3936/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3937/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3938/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3939/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3940/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3941/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3942/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3943/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3944/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3945/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3946/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3947/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3948/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3949/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3950/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3951/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3952/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3953/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3954/4000\n",
            "19/19 [==============================] - 1s 31ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3955/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3956/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3957/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3958/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3959/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3960/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3961/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3962/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3963/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3964/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3965/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3966/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3967/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3968/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3969/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3970/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3971/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3972/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3973/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3974/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3975/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3976/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3977/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3978/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3979/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3980/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3981/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3982/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3983/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3984/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3985/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3986/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3987/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3988/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3989/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3990/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3991/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3992/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3993/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3994/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3995/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3996/4000\n",
            "19/19 [==============================] - 1s 28ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3997/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3998/4000\n",
            "19/19 [==============================] - 1s 29ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 3999/4000\n",
            "19/19 [==============================] - 1s 30ms/step - loss: 93.1914 - accuracy: 0.0681\n",
            "Epoch 4000/4000\n",
            "19/19 [==============================] - 1s 27ms/step - loss: 93.1914 - accuracy: 0.0681\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tensorflow.python.keras.callbacks.History at 0x7f6d8246a668>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "a0T5yRw5t-e7"
      },
      "source": [
        "prediction=model.predict(img_in)"
      ],
      "execution_count": 14,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 288
        },
        "id": "VoDxy9WZ3G2Y",
        "outputId": "da5f6091-9554-46c3-8276-74583b4a495e"
      },
      "source": [
        "plt.imshow(prediction[29])"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7f6d78cdef60>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 17
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAQEAAAD8CAYAAAB3lxGOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAOV0lEQVR4nO3db8yddX3H8fdnKDxAE0BZQ0pdC6kmaJaCBEmGxP1RgSwWfMBKltk5smoCiSYuS4FkI3s2J5oYHaZGYlkc6KZIY3RaG6J7MJSitfwTKFhCm9JOWIBNowLfPTi/W45373rfvc85nHPn934lJ+c6v+s65/qeXO0n13WdO79vqgpJ/fqdaRcgaboMAalzhoDUOUNA6pwhIHXOEJA6N7EQSHJJkoeT7EuydVL7kTSaTOLvBJKcADwCvBM4ANwDXFVVD459Z5JGMqkzgQuAfVX1eFX9Ergd2DihfUkawasm9LmrgSeHXh8A3nasjZP4Z4vS5P20qk6fPzipEFhUki3AlmntX+rQEwsNTioEDgJrhl6f2cZ+raq2AdvAMwFpmiZ1T+AeYH2SdUlOBDYBOya0L0kjmMiZQFW9kORa4JvACcAtVfXAJPYlaTQT+YnwuIvwckB6JdxbVefPH/QvBqXOGQJS5wwBqXOGgNQ5Q0DqnCEgdc4QkDpnCEidMwSkzhkCUucMAalzhoDUOUNA6pwhIHXOEJA6t+wQSLImyV1JHkzyQJIPtfEbkxxMsqc9LhtfuZLGbZSZhV4APlJVP0jyWuDeJDvbuk9U1cdGL0/SpC07BKrqEHCoLT+f5CEGU41LWkHGck8gyVrgXOB7bejaJHuT3JLk1HHsQ9JkjBwCSV4DfBn4cFU9B9wMnA1sYHCmcNMx3rclye4ku0etQdLyjTTRaJJXA18DvllVH19g/Vrga1X1lkU+x4lGpckb70SjSQJ8DnhoOACSnDG02RXA/cvdh6TJG+XXgT8A/gK4L8meNnY9cFWSDUAB+4EPjFShpImy74DUD/sOSDqaISB1zhCQOmcISJ0zBKTOGQJS5wwBqXOGgNQ5Q0DqnCEgdc4QkDpnCEidMwSkzhkCUucMAalzhoDUuVFmFgIgyX7geeBF4IWqOj/JacAXgbUMZhe6sqr+Z9R9SRq/cZ0J/GFVbRiatWQrsKuq1gO72mtJM2hSlwMbge1teTtw+YT2I2lE4wiBAr6V5N4kW9rYqtahCOApYNX8N9l3QJoNI98TAC6qqoNJfhfYmeTHwyurqhaaSLSqtgHbwIlGpWka+Uygqg625yPAHcAFwOG5/gPt+cio+5E0GSOFQJKTW0dikpwMvItBs5EdwOa22WbgzlH2I2lyRr0cWAXcMWhGxKuAf62q/0hyD/ClJFcDTwBXjrgfSRNi8xGpHzYfkXQ0Q0DqnCEgdc4QkDpnCEidMwSkzhkCUucMAalzhoDUOUNA6pwhIHXOEJA6ZwhInTMEpM4ZAlLnlj2pSJI3MegtMOcs4O+AU4C/Bv67jV9fVV9fdoWSJmosk4okOQE4CLwNeD/wv1X1seN4v5OKSJM30UlF/hh4rKqeGNPnSXqFjCsENgG3Db2+NsneJLckOXVM+5A0ASOHQJITgfcA/9aGbgbOBjYAh4CbjvE+m49IM2DkewJJNgLXVNW7Fli3FvhaVb1lkc/wnoA0eRO7J3AVQ5cCc01HmisY9CGQNKNG6jvQGo68E/jA0PBHk2xg0KNw/7x1kmaMfQekfth3QNLRDAGpc4aA1DlDQOqcISB1zhCQOmcISJ0zBKTOGQJS5wwBqXOGgNQ5Q0DqnCEgdc4QkDpnCEidW1IItAlDjyS5f2jstCQ7kzzank9t40nyyST72mSj502qeEmjW+qZwOeBS+aNbQV2VdV6YFd7DXApsL49tjCYeFTSjFpSCFTVd4Fn5g1vBLa35e3A5UPjt9bA3cAp8+YdlDRDRrknsKqqDrXlp4BVbXk18OTQdgfamKQZNNJEo3Oqqo53nsAkWxhcLkiaolHOBA7Pnea35yNt/CCwZmi7M9vYb6iqbVV1/kITH0p65YwSAjuAzW15M3Dn0Pj72q8EFwLPDl02SJo1VbXog0FzkUPArxhc418NvI7BrwKPAt8GTmvbBvg08BhwH3D+Ej6/fPjwMfHH7oX+/9l3QOqHfQckHc0QkDpnCEidMwSkzhkCUucMAalzhoDUOUNA6pwhIHXOEJA6ZwhInTMEpM4ZAlLnDAGpc4aA1DlDQOrcoiFwjMYj/5Tkx625yB1JTmnja5P8PMme9vjMJIuXNLqlnAl8nqMbj+wE3lJVvw88Alw3tO6xqtrQHh8cT5mSJmXREFio8UhVfauqXmgv72Ywo7CkFWgc9wT+CvjG0Ot1SX6Y5DtJ3n6sNyXZkmR3kt1jqEHSMo3UfCTJDcALwBfa0CHgDVX1dJK3Al9N8uaqem7+e6tqG7CtfY4TjUpTsuwzgSR/Cfwp8Oc1N2941S+q6um2fC+DacffOIY6JU3IskIgySXA3wLvqaqfDY2fnuSEtnwWg87Ej4+jUEmTsejlQJLbgHcAr09yAPh7Br8GnATsTAJwd/sl4GLgH5L8CngJ+GBVze9mLGmG2HxE6ofNRyQdzRCQOmcISJ0zBKTOGQJS5wwBqXOGgNQ5Q0DqnCEgdc4QkDpnCEidMwSkzhkCUucMAalzhoDUueX2HbgxycGh/gKXDa27Lsm+JA8nefekCpc0HsvtOwDwiaH+Al8HSHIOsAl4c3vPP89NNyZpNi2r78BvsRG4vU04+hNgH3DBCPVJmrBR7glc29qQ3ZLk1Da2GnhyaJsDbewo9h2QZsNyQ+Bm4GxgA4NeAzcd7wdU1baqOn+hOc8kvXKWFQJVdbiqXqyql4DP8vIp/0FgzdCmZ7YxSTNquX0Hzhh6eQUw98vBDmBTkpOSrGPQd+D7o5UoaZKW23fgHUk2AAXsBz4AUFUPJPkS8CCD9mTXVNWLkyld0jjYd0Dqh30HJB3NEJA6ZwhInTMEpM4ZAlLnDAGpc4aA1DlDQOqcISB1zhCQOmcISJ0zBKTOGQJS5wwBqXOGgNS55fYd+OJQz4H9Sfa08bVJfj607jOTLF7S6BadWYhB34FPAbfODVTVn80tJ7kJeHZo+8eqasO4CpQ0WYuGQFV9N8nahdYlCXAl8EfjLUvSK2XUewJvBw5X1aNDY+uS/DDJd5K8fcTPlzRhS7kc+G2uAm4ben0IeENVPZ3krcBXk7y5qp6b/8YkW4AtI+5f0oiWfSaQ5FXAe4Evzo219mNPt+V7gceANy70fpuPSLNhlMuBPwF+XFUH5gaSnD7XgDTJWQz6Djw+WomSJmkpPxHeBvwX8KYkB5Jc3VZt4jcvBQAuBva2nwz/HfhgVS21mamkKbDvgNQP+w5IOpohIHXOEJA6ZwhInTMEpM4ZAlLnDAGpc4aA1DlDQOqcISB1zhCQOmcISJ0zBKTOGQJS5wwBqXNLmVRkTZK7kjyY5IEkH2rjpyXZmeTR9nxqG0+STybZl2RvkvMm/SUkLd9SzgReAD5SVecAFwLXJDkH2Arsqqr1wK72GuBSBtOKrWcwkejNY69a0tgsGgJVdaiqftCWnwceAlYDG4HtbbPtwOVteSNwaw3cDZyS5IyxVy5pLI7rnkBrQnIu8D1gVVUdaqueAla15dXAk0NvO9DGJM2gJfcdSPIa4MvAh6vquUHzoYGqquOdJ9C+A9JsWNKZQJJXMwiAL1TVV9rw4bnT/PZ8pI0fBNYMvf3MNvYb7DsgzYal/DoQ4HPAQ1X18aFVO4DNbXkzcOfQ+PvarwQXAs8OXTZImjGLTjme5CLgP4H7gJfa8PUM7gt8CXgD8ARwZVU900LjU8AlwM+A91fV7kX24ZTj0uQtOOW4fQekfth3QNLRDAGpc4aA1DlDQOqcISB1zhCQOmcISJ0zBKTOGQJS5wwBqXOGgNQ5Q0DqnCEgdc4QkDpnCEidMwSkzhkCUucMAalzS55yfMJ+Cvxfe16pXs/Krh9W/ndY6fXDZL/D7y00OBNzDAIk2b2Spx9f6fXDyv8OK71+mM538HJA6pwhIHVulkJg27QLGNFKrx9W/ndY6fXDFL7DzNwTkDQds3QmIGkKph4CSS5J8nCSfUm2TruepUqyP8l9SfYk2d3GTkuyM8mj7fnUadc5LMktSY4kuX9obMGaWy/JT7bjsjfJedOr/Ne1LlT/jUkOtuOwJ8llQ+uua/U/nOTd06n6ZUnWJLkryYNJHkjyoTY+3WNQVVN7ACcAjwFnAScCPwLOmWZNx1H7fuD188Y+Cmxty1uBf5x2nfPquxg4D7h/sZqBy4BvAAEuBL43o/XfCPzNAtue0/49nQSsa//OTphy/WcA57Xl1wKPtDqnegymfSZwAbCvqh6vql8CtwMbp1zTKDYC29vyduDyKdZylKr6LvDMvOFj1bwRuLUG7gZOmWtFPy3HqP9YNgK3V9UvquonwD4G/96mpqoOVdUP2vLzwEPAaqZ8DKYdAquBJ4deH2hjK0EB30pyb5ItbWxVvdyG/Slg1XRKOy7HqnklHZtr2+nyLUOXYDNdf5K1wLkMuntP9RhMOwRWsouq6jzgUuCaJBcPr6zB+dyK+ullJdYM3AycDWwADgE3TbecxSV5DfBl4MNV9dzwumkcg2mHwEFgzdDrM9vYzKuqg+35CHAHg1PNw3Ona+35yPQqXLJj1bwijk1VHa6qF6vqJeCzvHzKP5P1J3k1gwD4QlV9pQ1P9RhMOwTuAdYnWZfkRGATsGPKNS0qyclJXju3DLwLuJ9B7ZvbZpuBO6dT4XE5Vs07gPe1O9QXAs8OnbLOjHnXyFcwOA4wqH9TkpOSrAPWA99/pesbliTA54CHqurjQ6umewymebd06A7oIwzu3t4w7XqWWPNZDO48/wh4YK5u4HXALuBR4NvAadOudV7dtzE4Zf4Vg+vLq49VM4M70p9ux+U+4PwZrf9fWn1723+aM4a2v6HV/zBw6QzUfxGDU/29wJ72uGzax8C/GJQ6N+3LAUlTZghInTMEpM4ZAlLnDAGpc4aA1DlDQOqcISB17v8BkdUzvFyee40AAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jNtrM--e3MlK"
      },
      "source": [
        "model.save('model.h5')"
      ],
      "execution_count": 18,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FnU58UFu4jT9"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}