cap11_1_simple_autoencoder.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "cap11_1_simple_autoencoder.ipynb",
      "provenance": [],
      "authorship_tag": "ABX9TyOxB4skBla05A7i2FfsUqys",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/viniciusrpb/cic0193_machinelearning/blob/main/cap11_1_simple_autoencoder.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Kvs0Fv3bjzsW"
      },
      "source": [
        "from matplotlib.pyplot import imshow\n",
        "import numpy as np\n",
        "import cv2\n",
        "from keras.preprocessing.image import img_to_array\n",
        "from tensorflow.keras.layers import Conv2D, MaxPooling2D, UpSampling2D\n",
        "from tensorflow.keras.models import Sequential\n",
        "\n"
      ],
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JvuUosBBkskl"
      },
      "source": [
        "np.random.seed(42)\n",
        "\n",
        "dim=256\n",
        "img_data=[]\n",
        "\n",
        "img=cv2.imread('cicunb.png', 1)  \n",
        "img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)\n",
        "\n",
        "img=cv2.resize(img,(dim, dim))\n",
        "img_data.append(img_to_array(img))\n",
        "\n",
        "img_array = np.reshape(img_data, (len(img_data), dim, dim, 3))\n",
        "img_array = img_array.astype('float32') / 255.\n",
        "\n"
      ],
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "RRfkd-Vlku5s",
        "outputId": "49118227-e362-477a-f0a9-28a8adcbc75a"
      },
      "source": [
        "model = Sequential()\n",
        "\n",
        "model.add(Conv2D(32, (3, 3), activation='relu', padding='same', input_shape=(dim, dim, 3)))\n",
        "model.add(MaxPooling2D((2, 2), padding='same'))\n",
        "model.add(Conv2D(8, (3, 3), activation='relu', padding='same'))\n",
        "model.add(MaxPooling2D((2, 2), padding='same'))\n",
        "model.add(Conv2D(8, (3, 3), activation='relu', padding='same'))\n",
        " \n",
        "\n",
        "model.add(MaxPooling2D((2, 2), padding='same'))\n",
        "     \n",
        "model.add(Conv2D(8, (3, 3), activation='relu', padding='same'))\n",
        "model.add(UpSampling2D((2, 2)))\n",
        "model.add(Conv2D(8, (3, 3), activation='relu', padding='same'))\n",
        "model.add(UpSampling2D((2, 2)))\n",
        "model.add(Conv2D(32, (3, 3), activation='relu', padding='same'))\n",
        "model.add(UpSampling2D((2, 2)))\n",
        "model.add(Conv2D(3, (3, 3), activation='relu', padding='same'))\n",
        "model.summary()\n"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential_1\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_7 (Conv2D)            (None, 256, 256, 32)      896       \n",
            "_________________________________________________________________\n",
            "max_pooling2d_3 (MaxPooling2 (None, 128, 128, 32)      0         \n",
            "_________________________________________________________________\n",
            "conv2d_8 (Conv2D)            (None, 128, 128, 8)       2312      \n",
            "_________________________________________________________________\n",
            "max_pooling2d_4 (MaxPooling2 (None, 64, 64, 8)         0         \n",
            "_________________________________________________________________\n",
            "conv2d_9 (Conv2D)            (None, 64, 64, 8)         584       \n",
            "_________________________________________________________________\n",
            "max_pooling2d_5 (MaxPooling2 (None, 32, 32, 8)         0         \n",
            "_________________________________________________________________\n",
            "conv2d_10 (Conv2D)           (None, 32, 32, 8)         584       \n",
            "_________________________________________________________________\n",
            "up_sampling2d_3 (UpSampling2 (None, 64, 64, 8)         0         \n",
            "_________________________________________________________________\n",
            "conv2d_11 (Conv2D)           (None, 64, 64, 8)         584       \n",
            "_________________________________________________________________\n",
            "up_sampling2d_4 (UpSampling2 (None, 128, 128, 8)       0         \n",
            "_________________________________________________________________\n",
            "conv2d_12 (Conv2D)           (None, 128, 128, 32)      2336      \n",
            "_________________________________________________________________\n",
            "up_sampling2d_5 (UpSampling2 (None, 256, 256, 32)      0         \n",
            "_________________________________________________________________\n",
            "conv2d_13 (Conv2D)           (None, 256, 256, 3)       867       \n",
            "=================================================================\n",
            "Total params: 8,163\n",
            "Trainable params: 8,163\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ll9FM_1soz8V"
      },
      "source": [
        "model.compile(optimizer='adam', loss='mean_squared_error', metrics=['accuracy'])\n"
      ],
      "execution_count": 9,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "VJcIm9jFo0eM",
        "outputId": "de0f49d9-22fb-4a4c-8e9a-f084b1cfdd77"
      },
      "source": [
        "model.fit(img_array, img_array, epochs=5000,shuffle=True)"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\u001b[1;30;43mStreaming output truncated to the last 5000 lines.\u001b[0m\n",
            "1/1 [==============================] - 0s 261ms/step - loss: 0.0076 - accuracy: 0.6100\n",
            "Epoch 2502/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0076 - accuracy: 0.6046\n",
            "Epoch 2503/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0076 - accuracy: 0.6113\n",
            "Epoch 2504/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0075 - accuracy: 0.6077\n",
            "Epoch 2505/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0074 - accuracy: 0.6089\n",
            "Epoch 2506/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0074 - accuracy: 0.6059\n",
            "Epoch 2507/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0074 - accuracy: 0.6045\n",
            "Epoch 2508/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0074 - accuracy: 0.6047\n",
            "Epoch 2509/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0075 - accuracy: 0.6024\n",
            "Epoch 2510/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0075 - accuracy: 0.6078\n",
            "Epoch 2511/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0075 - accuracy: 0.6013\n",
            "Epoch 2512/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0075 - accuracy: 0.6115\n",
            "Epoch 2513/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0075 - accuracy: 0.6015\n",
            "Epoch 2514/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0075 - accuracy: 0.6094\n",
            "Epoch 2515/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0074 - accuracy: 0.6024\n",
            "Epoch 2516/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0074 - accuracy: 0.6078\n",
            "Epoch 2517/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0074 - accuracy: 0.6066\n",
            "Epoch 2518/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0074 - accuracy: 0.6051\n",
            "Epoch 2519/5000\n",
            "1/1 [==============================] - 0s 260ms/step - loss: 0.0074 - accuracy: 0.6079\n",
            "Epoch 2520/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0074 - accuracy: 0.6031\n",
            "Epoch 2521/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0074 - accuracy: 0.6070\n",
            "Epoch 2522/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0074 - accuracy: 0.6024\n",
            "Epoch 2523/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0074 - accuracy: 0.6080\n",
            "Epoch 2524/5000\n",
            "1/1 [==============================] - 0s 253ms/step - loss: 0.0074 - accuracy: 0.6007\n",
            "Epoch 2525/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0075 - accuracy: 0.6107\n",
            "Epoch 2526/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0074 - accuracy: 0.6040\n",
            "Epoch 2527/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0074 - accuracy: 0.6110\n",
            "Epoch 2528/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0074 - accuracy: 0.6063\n",
            "Epoch 2529/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0074 - accuracy: 0.6068\n",
            "Epoch 2530/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0074 - accuracy: 0.6064\n",
            "Epoch 2531/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0074 - accuracy: 0.6022\n",
            "Epoch 2532/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0074 - accuracy: 0.6055\n",
            "Epoch 2533/5000\n",
            "1/1 [==============================] - 0s 254ms/step - loss: 0.0074 - accuracy: 0.6017\n",
            "Epoch 2534/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0074 - accuracy: 0.6068\n",
            "Epoch 2535/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0074 - accuracy: 0.6056\n",
            "Epoch 2536/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0074 - accuracy: 0.6103\n",
            "Epoch 2537/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0074 - accuracy: 0.6094\n",
            "Epoch 2538/5000\n",
            "1/1 [==============================] - 0s 259ms/step - loss: 0.0074 - accuracy: 0.6120\n",
            "Epoch 2539/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0074 - accuracy: 0.6047\n",
            "Epoch 2540/5000\n",
            "1/1 [==============================] - 0s 260ms/step - loss: 0.0075 - accuracy: 0.6079\n",
            "Epoch 2541/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0075 - accuracy: 0.5951\n",
            "Epoch 2542/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0075 - accuracy: 0.6046\n",
            "Epoch 2543/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0075 - accuracy: 0.5976\n",
            "Epoch 2544/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0075 - accuracy: 0.6080\n",
            "Epoch 2545/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0075 - accuracy: 0.6069\n",
            "Epoch 2546/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0074 - accuracy: 0.6134\n",
            "Epoch 2547/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0074 - accuracy: 0.6116\n",
            "Epoch 2548/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0074 - accuracy: 0.6111\n",
            "Epoch 2549/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0074 - accuracy: 0.6079\n",
            "Epoch 2550/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0074 - accuracy: 0.6013\n",
            "Epoch 2551/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0074 - accuracy: 0.6022\n",
            "Epoch 2552/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0074 - accuracy: 0.5991\n",
            "Epoch 2553/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0074 - accuracy: 0.6073\n",
            "Epoch 2554/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0074 - accuracy: 0.6078\n",
            "Epoch 2555/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0075 - accuracy: 0.6141\n",
            "Epoch 2556/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0075 - accuracy: 0.6041\n",
            "Epoch 2557/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0075 - accuracy: 0.6126\n",
            "Epoch 2558/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0075 - accuracy: 0.5971\n",
            "Epoch 2559/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0075 - accuracy: 0.6067\n",
            "Epoch 2560/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0075 - accuracy: 0.6005\n",
            "Epoch 2561/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0074 - accuracy: 0.6048\n",
            "Epoch 2562/5000\n",
            "1/1 [==============================] - 0s 260ms/step - loss: 0.0074 - accuracy: 0.6108\n",
            "Epoch 2563/5000\n",
            "1/1 [==============================] - 0s 257ms/step - loss: 0.0074 - accuracy: 0.6063\n",
            "Epoch 2564/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0074 - accuracy: 0.6119\n",
            "Epoch 2565/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0074 - accuracy: 0.6087\n",
            "Epoch 2566/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0074 - accuracy: 0.6071\n",
            "Epoch 2567/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0074 - accuracy: 0.6061\n",
            "Epoch 2568/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0074 - accuracy: 0.6034\n",
            "Epoch 2569/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0074 - accuracy: 0.6041\n",
            "Epoch 2570/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0074 - accuracy: 0.6053\n",
            "Epoch 2571/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0074 - accuracy: 0.6046\n",
            "Epoch 2572/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0074 - accuracy: 0.6095\n",
            "Epoch 2573/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0074 - accuracy: 0.5971\n",
            "Epoch 2574/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0074 - accuracy: 0.6105\n",
            "Epoch 2575/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0074 - accuracy: 0.6008\n",
            "Epoch 2576/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0074 - accuracy: 0.6052\n",
            "Epoch 2577/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0074 - accuracy: 0.6053\n",
            "Epoch 2578/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0074 - accuracy: 0.6053\n",
            "Epoch 2579/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0074 - accuracy: 0.6104\n",
            "Epoch 2580/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0074 - accuracy: 0.6058\n",
            "Epoch 2581/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0074 - accuracy: 0.6089\n",
            "Epoch 2582/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0074 - accuracy: 0.6044\n",
            "Epoch 2583/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0074 - accuracy: 0.6049\n",
            "Epoch 2584/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0074 - accuracy: 0.6034\n",
            "Epoch 2585/5000\n",
            "1/1 [==============================] - 0s 297ms/step - loss: 0.0074 - accuracy: 0.6068\n",
            "Epoch 2586/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0074 - accuracy: 0.6046\n",
            "Epoch 2587/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0074 - accuracy: 0.6117\n",
            "Epoch 2588/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0075 - accuracy: 0.5995\n",
            "Epoch 2589/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0075 - accuracy: 0.6115\n",
            "Epoch 2590/5000\n",
            "1/1 [==============================] - 0s 290ms/step - loss: 0.0076 - accuracy: 0.5987\n",
            "Epoch 2591/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0076 - accuracy: 0.6083\n",
            "Epoch 2592/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0075 - accuracy: 0.6022\n",
            "Epoch 2593/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0075 - accuracy: 0.6067\n",
            "Epoch 2594/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0075 - accuracy: 0.6091\n",
            "Epoch 2595/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0074 - accuracy: 0.6114\n",
            "Epoch 2596/5000\n",
            "1/1 [==============================] - 0s 259ms/step - loss: 0.0074 - accuracy: 0.6115\n",
            "Epoch 2597/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0073 - accuracy: 0.6098\n",
            "Epoch 2598/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0074 - accuracy: 0.6067\n",
            "Epoch 2599/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0074 - accuracy: 0.6052\n",
            "Epoch 2600/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0074 - accuracy: 0.6012\n",
            "Epoch 2601/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0074 - accuracy: 0.6053\n",
            "Epoch 2602/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0074 - accuracy: 0.6068\n",
            "Epoch 2603/5000\n",
            "1/1 [==============================] - 0s 260ms/step - loss: 0.0074 - accuracy: 0.6032\n",
            "Epoch 2604/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0074 - accuracy: 0.6130\n",
            "Epoch 2605/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0074 - accuracy: 0.5978\n",
            "Epoch 2606/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0074 - accuracy: 0.6078\n",
            "Epoch 2607/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0074 - accuracy: 0.6026\n",
            "Epoch 2608/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0073 - accuracy: 0.6054\n",
            "Epoch 2609/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0073 - accuracy: 0.6087\n",
            "Epoch 2610/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0073 - accuracy: 0.6068\n",
            "Epoch 2611/5000\n",
            "1/1 [==============================] - 0s 294ms/step - loss: 0.0073 - accuracy: 0.6119\n",
            "Epoch 2612/5000\n",
            "1/1 [==============================] - 0s 306ms/step - loss: 0.0073 - accuracy: 0.6051\n",
            "Epoch 2613/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0073 - accuracy: 0.6070\n",
            "Epoch 2614/5000\n",
            "1/1 [==============================] - 0s 304ms/step - loss: 0.0073 - accuracy: 0.6010\n",
            "Epoch 2615/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0074 - accuracy: 0.6047\n",
            "Epoch 2616/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0074 - accuracy: 0.6052\n",
            "Epoch 2617/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0074 - accuracy: 0.6107\n",
            "Epoch 2618/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0074 - accuracy: 0.6103\n",
            "Epoch 2619/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0074 - accuracy: 0.6109\n",
            "Epoch 2620/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0074 - accuracy: 0.6022\n",
            "Epoch 2621/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0074 - accuracy: 0.6074\n",
            "Epoch 2622/5000\n",
            "1/1 [==============================] - 0s 260ms/step - loss: 0.0074 - accuracy: 0.5936\n",
            "Epoch 2623/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0074 - accuracy: 0.6053\n",
            "Epoch 2624/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0074 - accuracy: 0.6034\n",
            "Epoch 2625/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0074 - accuracy: 0.6121\n",
            "Epoch 2626/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0074 - accuracy: 0.6160\n",
            "Epoch 2627/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0074 - accuracy: 0.6146\n",
            "Epoch 2628/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0073 - accuracy: 0.6133\n",
            "Epoch 2629/5000\n",
            "1/1 [==============================] - 0s 258ms/step - loss: 0.0073 - accuracy: 0.6034\n",
            "Epoch 2630/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0073 - accuracy: 0.6027\n",
            "Epoch 2631/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0073 - accuracy: 0.5980\n",
            "Epoch 2632/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0073 - accuracy: 0.6031\n",
            "Epoch 2633/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0074 - accuracy: 0.6048\n",
            "Epoch 2634/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0074 - accuracy: 0.6123\n",
            "Epoch 2635/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0075 - accuracy: 0.6059\n",
            "Epoch 2636/5000\n",
            "1/1 [==============================] - 0s 258ms/step - loss: 0.0075 - accuracy: 0.6136\n",
            "Epoch 2637/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0076 - accuracy: 0.5999\n",
            "Epoch 2638/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0076 - accuracy: 0.6095\n",
            "Epoch 2639/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0076 - accuracy: 0.5999\n",
            "Epoch 2640/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0075 - accuracy: 0.6075\n",
            "Epoch 2641/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0075 - accuracy: 0.6069\n",
            "Epoch 2642/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0074 - accuracy: 0.6103\n",
            "Epoch 2643/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0073 - accuracy: 0.6107\n",
            "Epoch 2644/5000\n",
            "1/1 [==============================] - 0s 259ms/step - loss: 0.0073 - accuracy: 0.6090\n",
            "Epoch 2645/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0073 - accuracy: 0.6081\n",
            "Epoch 2646/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0074 - accuracy: 0.6038\n",
            "Epoch 2647/5000\n",
            "1/1 [==============================] - 0s 256ms/step - loss: 0.0074 - accuracy: 0.6091\n",
            "Epoch 2648/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0075 - accuracy: 0.6013\n",
            "Epoch 2649/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0075 - accuracy: 0.6116\n",
            "Epoch 2650/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0075 - accuracy: 0.6017\n",
            "Epoch 2651/5000\n",
            "1/1 [==============================] - 0s 261ms/step - loss: 0.0074 - accuracy: 0.6110\n",
            "Epoch 2652/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0074 - accuracy: 0.6059\n",
            "Epoch 2653/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0073 - accuracy: 0.6065\n",
            "Epoch 2654/5000\n",
            "1/1 [==============================] - 0s 291ms/step - loss: 0.0073 - accuracy: 0.6096\n",
            "Epoch 2655/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0073 - accuracy: 0.6032\n",
            "Epoch 2656/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0074 - accuracy: 0.6106\n",
            "Epoch 2657/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0074 - accuracy: 0.6033\n",
            "Epoch 2658/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0074 - accuracy: 0.6103\n",
            "Epoch 2659/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0074 - accuracy: 0.6056\n",
            "Epoch 2660/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0074 - accuracy: 0.6100\n",
            "Epoch 2661/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0074 - accuracy: 0.6089\n",
            "Epoch 2662/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0073 - accuracy: 0.6089\n",
            "Epoch 2663/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0073 - accuracy: 0.6095\n",
            "Epoch 2664/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0073 - accuracy: 0.6041\n",
            "Epoch 2665/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0073 - accuracy: 0.6079\n",
            "Epoch 2666/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0073 - accuracy: 0.6030\n",
            "Epoch 2667/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0073 - accuracy: 0.6077\n",
            "Epoch 2668/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0073 - accuracy: 0.6045\n",
            "Epoch 2669/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0073 - accuracy: 0.6091\n",
            "Epoch 2670/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0074 - accuracy: 0.6055\n",
            "Epoch 2671/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0074 - accuracy: 0.6111\n",
            "Epoch 2672/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0074 - accuracy: 0.6030\n",
            "Epoch 2673/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0074 - accuracy: 0.6130\n",
            "Epoch 2674/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0074 - accuracy: 0.6018\n",
            "Epoch 2675/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0073 - accuracy: 0.6096\n",
            "Epoch 2676/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0073 - accuracy: 0.6051\n",
            "Epoch 2677/5000\n",
            "1/1 [==============================] - 0s 301ms/step - loss: 0.0073 - accuracy: 0.6073\n",
            "Epoch 2678/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0073 - accuracy: 0.6107\n",
            "Epoch 2679/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0073 - accuracy: 0.6067\n",
            "Epoch 2680/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0073 - accuracy: 0.6121\n",
            "Epoch 2681/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0073 - accuracy: 0.6057\n",
            "Epoch 2682/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0073 - accuracy: 0.6091\n",
            "Epoch 2683/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0073 - accuracy: 0.6027\n",
            "Epoch 2684/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0074 - accuracy: 0.6080\n",
            "Epoch 2685/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0074 - accuracy: 0.6014\n",
            "Epoch 2686/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0074 - accuracy: 0.6112\n",
            "Epoch 2687/5000\n",
            "1/1 [==============================] - 0s 294ms/step - loss: 0.0074 - accuracy: 0.6048\n",
            "Epoch 2688/5000\n",
            "1/1 [==============================] - 0s 304ms/step - loss: 0.0074 - accuracy: 0.6134\n",
            "Epoch 2689/5000\n",
            "1/1 [==============================] - 0s 294ms/step - loss: 0.0074 - accuracy: 0.6059\n",
            "Epoch 2690/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0074 - accuracy: 0.6107\n",
            "Epoch 2691/5000\n",
            "1/1 [==============================] - 0s 301ms/step - loss: 0.0074 - accuracy: 0.6038\n",
            "Epoch 2692/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0074 - accuracy: 0.6091\n",
            "Epoch 2693/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0073 - accuracy: 0.6071\n",
            "Epoch 2694/5000\n",
            "1/1 [==============================] - 0s 300ms/step - loss: 0.0073 - accuracy: 0.6096\n",
            "Epoch 2695/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0073 - accuracy: 0.6102\n",
            "Epoch 2696/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0073 - accuracy: 0.6065\n",
            "Epoch 2697/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0073 - accuracy: 0.6109\n",
            "Epoch 2698/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0073 - accuracy: 0.6042\n",
            "Epoch 2699/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0074 - accuracy: 0.6115\n",
            "Epoch 2700/5000\n",
            "1/1 [==============================] - 0s 298ms/step - loss: 0.0074 - accuracy: 0.6034\n",
            "Epoch 2701/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0074 - accuracy: 0.6136\n",
            "Epoch 2702/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0074 - accuracy: 0.6028\n",
            "Epoch 2703/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0074 - accuracy: 0.6134\n",
            "Epoch 2704/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0074 - accuracy: 0.6012\n",
            "Epoch 2705/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0074 - accuracy: 0.6077\n",
            "Epoch 2706/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0073 - accuracy: 0.6075\n",
            "Epoch 2707/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0073 - accuracy: 0.6062\n",
            "Epoch 2708/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0073 - accuracy: 0.6163\n",
            "Epoch 2709/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0073 - accuracy: 0.6076\n",
            "Epoch 2710/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0073 - accuracy: 0.6119\n",
            "Epoch 2711/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0073 - accuracy: 0.6071\n",
            "Epoch 2712/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0073 - accuracy: 0.6047\n",
            "Epoch 2713/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0073 - accuracy: 0.6073\n",
            "Epoch 2714/5000\n",
            "1/1 [==============================] - 0s 254ms/step - loss: 0.0073 - accuracy: 0.6026\n",
            "Epoch 2715/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0073 - accuracy: 0.6085\n",
            "Epoch 2716/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0073 - accuracy: 0.6097\n",
            "Epoch 2717/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0073 - accuracy: 0.6092\n",
            "Epoch 2718/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0073 - accuracy: 0.6112\n",
            "Epoch 2719/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0073 - accuracy: 0.6021\n",
            "Epoch 2720/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0073 - accuracy: 0.6093\n",
            "Epoch 2721/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0073 - accuracy: 0.5974\n",
            "Epoch 2722/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0073 - accuracy: 0.6082\n",
            "Epoch 2723/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0073 - accuracy: 0.6048\n",
            "Epoch 2724/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0073 - accuracy: 0.6120\n",
            "Epoch 2725/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0073 - accuracy: 0.6141\n",
            "Epoch 2726/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0073 - accuracy: 0.6105\n",
            "Epoch 2727/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0073 - accuracy: 0.6129\n",
            "Epoch 2728/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0073 - accuracy: 0.6043\n",
            "Epoch 2729/5000\n",
            "1/1 [==============================] - 0s 304ms/step - loss: 0.0073 - accuracy: 0.6055\n",
            "Epoch 2730/5000\n",
            "1/1 [==============================] - 0s 255ms/step - loss: 0.0073 - accuracy: 0.6019\n",
            "Epoch 2731/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0073 - accuracy: 0.6034\n",
            "Epoch 2732/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0073 - accuracy: 0.6086\n",
            "Epoch 2733/5000\n",
            "1/1 [==============================] - 0s 261ms/step - loss: 0.0073 - accuracy: 0.6086\n",
            "Epoch 2734/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0073 - accuracy: 0.6165\n",
            "Epoch 2735/5000\n",
            "1/1 [==============================] - 0s 292ms/step - loss: 0.0073 - accuracy: 0.6083\n",
            "Epoch 2736/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0073 - accuracy: 0.6154\n",
            "Epoch 2737/5000\n",
            "1/1 [==============================] - 0s 259ms/step - loss: 0.0073 - accuracy: 0.6007\n",
            "Epoch 2738/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0073 - accuracy: 0.6070\n",
            "Epoch 2739/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0074 - accuracy: 0.5998\n",
            "Epoch 2740/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0074 - accuracy: 0.6082\n",
            "Epoch 2741/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0075 - accuracy: 0.6090\n",
            "Epoch 2742/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0076 - accuracy: 0.6175\n",
            "Epoch 2743/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0076 - accuracy: 0.6104\n",
            "Epoch 2744/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0077 - accuracy: 0.6154\n",
            "Epoch 2745/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0077 - accuracy: 0.5933\n",
            "Epoch 2746/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0075 - accuracy: 0.6104\n",
            "Epoch 2747/5000\n",
            "1/1 [==============================] - 0s 258ms/step - loss: 0.0074 - accuracy: 0.5954\n",
            "Epoch 2748/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0073 - accuracy: 0.6068\n",
            "Epoch 2749/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0073 - accuracy: 0.6105\n",
            "Epoch 2750/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0073 - accuracy: 0.6072\n",
            "Epoch 2751/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0074 - accuracy: 0.6202\n",
            "Epoch 2752/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0074 - accuracy: 0.6090\n",
            "Epoch 2753/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0074 - accuracy: 0.6158\n",
            "Epoch 2754/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0074 - accuracy: 0.6093\n",
            "Epoch 2755/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0074 - accuracy: 0.6057\n",
            "Epoch 2756/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0073 - accuracy: 0.6092\n",
            "Epoch 2757/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0073 - accuracy: 0.6023\n",
            "Epoch 2758/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0073 - accuracy: 0.6049\n",
            "Epoch 2759/5000\n",
            "1/1 [==============================] - 0s 249ms/step - loss: 0.0073 - accuracy: 0.6062\n",
            "Epoch 2760/5000\n",
            "1/1 [==============================] - 0s 258ms/step - loss: 0.0073 - accuracy: 0.6065\n",
            "Epoch 2761/5000\n",
            "1/1 [==============================] - 0s 302ms/step - loss: 0.0073 - accuracy: 0.6125\n",
            "Epoch 2762/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0073 - accuracy: 0.6127\n",
            "Epoch 2763/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0073 - accuracy: 0.6121\n",
            "Epoch 2764/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0073 - accuracy: 0.6127\n",
            "Epoch 2765/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0073 - accuracy: 0.5974\n",
            "Epoch 2766/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0073 - accuracy: 0.6061\n",
            "Epoch 2767/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0073 - accuracy: 0.5931\n",
            "Epoch 2768/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0073 - accuracy: 0.6072\n",
            "Epoch 2769/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0073 - accuracy: 0.6105\n",
            "Epoch 2770/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0073 - accuracy: 0.6127\n",
            "Epoch 2771/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0073 - accuracy: 0.6214\n",
            "Epoch 2772/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0073 - accuracy: 0.6085\n",
            "Epoch 2773/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0073 - accuracy: 0.6108\n",
            "Epoch 2774/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0073 - accuracy: 0.5986\n",
            "Epoch 2775/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0073 - accuracy: 0.6010\n",
            "Epoch 2776/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0073 - accuracy: 0.6058\n",
            "Epoch 2777/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0073 - accuracy: 0.6059\n",
            "Epoch 2778/5000\n",
            "1/1 [==============================] - 0s 260ms/step - loss: 0.0073 - accuracy: 0.6140\n",
            "Epoch 2779/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0073 - accuracy: 0.6109\n",
            "Epoch 2780/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0072 - accuracy: 0.6117\n",
            "Epoch 2781/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0072 - accuracy: 0.6099\n",
            "Epoch 2782/5000\n",
            "1/1 [==============================] - 0s 247ms/step - loss: 0.0072 - accuracy: 0.6063\n",
            "Epoch 2783/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0073 - accuracy: 0.6099\n",
            "Epoch 2784/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0073 - accuracy: 0.6047\n",
            "Epoch 2785/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0073 - accuracy: 0.6106\n",
            "Epoch 2786/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0073 - accuracy: 0.6060\n",
            "Epoch 2787/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0072 - accuracy: 0.6102\n",
            "Epoch 2788/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0072 - accuracy: 0.6083\n",
            "Epoch 2789/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0072 - accuracy: 0.6108\n",
            "Epoch 2790/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0072 - accuracy: 0.6101\n",
            "Epoch 2791/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0072 - accuracy: 0.6098\n",
            "Epoch 2792/5000\n",
            "1/1 [==============================] - 0s 260ms/step - loss: 0.0072 - accuracy: 0.6109\n",
            "Epoch 2793/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0073 - accuracy: 0.6055\n",
            "Epoch 2794/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0073 - accuracy: 0.6125\n",
            "Epoch 2795/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0073 - accuracy: 0.6003\n",
            "Epoch 2796/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0073 - accuracy: 0.6124\n",
            "Epoch 2797/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0073 - accuracy: 0.6039\n",
            "Epoch 2798/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0074 - accuracy: 0.6122\n",
            "Epoch 2799/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0074 - accuracy: 0.6089\n",
            "Epoch 2800/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0075 - accuracy: 0.6136\n",
            "Epoch 2801/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0075 - accuracy: 0.6077\n",
            "Epoch 2802/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0076 - accuracy: 0.6178\n",
            "Epoch 2803/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0076 - accuracy: 0.5956\n",
            "Epoch 2804/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0075 - accuracy: 0.6116\n",
            "Epoch 2805/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0074 - accuracy: 0.5948\n",
            "Epoch 2806/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0073 - accuracy: 0.6040\n",
            "Epoch 2807/5000\n",
            "1/1 [==============================] - 0s 260ms/step - loss: 0.0073 - accuracy: 0.6113\n",
            "Epoch 2808/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0073 - accuracy: 0.6093\n",
            "Epoch 2809/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0073 - accuracy: 0.6243\n",
            "Epoch 2810/5000\n",
            "1/1 [==============================] - 0s 252ms/step - loss: 0.0074 - accuracy: 0.6166\n",
            "Epoch 2811/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0074 - accuracy: 0.6180\n",
            "Epoch 2812/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0074 - accuracy: 0.6083\n",
            "Epoch 2813/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0074 - accuracy: 0.6041\n",
            "Epoch 2814/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0073 - accuracy: 0.6018\n",
            "Epoch 2815/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0073 - accuracy: 0.6027\n",
            "Epoch 2816/5000\n",
            "1/1 [==============================] - 0s 290ms/step - loss: 0.0072 - accuracy: 0.6071\n",
            "Epoch 2817/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0072 - accuracy: 0.6102\n",
            "Epoch 2818/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0072 - accuracy: 0.6148\n",
            "Epoch 2819/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0073 - accuracy: 0.6154\n",
            "Epoch 2820/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0073 - accuracy: 0.6140\n",
            "Epoch 2821/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0073 - accuracy: 0.6113\n",
            "Epoch 2822/5000\n",
            "1/1 [==============================] - 0s 261ms/step - loss: 0.0073 - accuracy: 0.6089\n",
            "Epoch 2823/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0073 - accuracy: 0.5965\n",
            "Epoch 2824/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0073 - accuracy: 0.6084\n",
            "Epoch 2825/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0073 - accuracy: 0.5944\n",
            "Epoch 2826/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0073 - accuracy: 0.6124\n",
            "Epoch 2827/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0073 - accuracy: 0.6123\n",
            "Epoch 2828/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0072 - accuracy: 0.6148\n",
            "Epoch 2829/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0072 - accuracy: 0.6195\n",
            "Epoch 2830/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0073 - accuracy: 0.6091\n",
            "Epoch 2831/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0073 - accuracy: 0.6123\n",
            "Epoch 2832/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0073 - accuracy: 0.5989\n",
            "Epoch 2833/5000\n",
            "1/1 [==============================] - 0s 260ms/step - loss: 0.0073 - accuracy: 0.6061\n",
            "Epoch 2834/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0073 - accuracy: 0.6006\n",
            "Epoch 2835/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0073 - accuracy: 0.6097\n",
            "Epoch 2836/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0073 - accuracy: 0.6111\n",
            "Epoch 2837/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0073 - accuracy: 0.6139\n",
            "Epoch 2838/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0073 - accuracy: 0.6139\n",
            "Epoch 2839/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0073 - accuracy: 0.6145\n",
            "Epoch 2840/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0072 - accuracy: 0.6077\n",
            "Epoch 2841/5000\n",
            "1/1 [==============================] - 0s 292ms/step - loss: 0.0073 - accuracy: 0.6132\n",
            "Epoch 2842/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0073 - accuracy: 0.5995\n",
            "Epoch 2843/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0073 - accuracy: 0.6096\n",
            "Epoch 2844/5000\n",
            "1/1 [==============================] - 0s 294ms/step - loss: 0.0072 - accuracy: 0.6000\n",
            "Epoch 2845/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0072 - accuracy: 0.6105\n",
            "Epoch 2846/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0072 - accuracy: 0.6133\n",
            "Epoch 2847/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0072 - accuracy: 0.6176\n",
            "Epoch 2848/5000\n",
            "1/1 [==============================] - 0s 301ms/step - loss: 0.0072 - accuracy: 0.6183\n",
            "Epoch 2849/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0072 - accuracy: 0.6124\n",
            "Epoch 2850/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0072 - accuracy: 0.6084\n",
            "Epoch 2851/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0072 - accuracy: 0.5994\n",
            "Epoch 2852/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0072 - accuracy: 0.6055\n",
            "Epoch 2853/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0072 - accuracy: 0.6003\n",
            "Epoch 2854/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0072 - accuracy: 0.6138\n",
            "Epoch 2855/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0072 - accuracy: 0.6092\n",
            "Epoch 2856/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0073 - accuracy: 0.6179\n",
            "Epoch 2857/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0073 - accuracy: 0.6126\n",
            "Epoch 2858/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0073 - accuracy: 0.6163\n",
            "Epoch 2859/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0074 - accuracy: 0.6053\n",
            "Epoch 2860/5000\n",
            "1/1 [==============================] - 0s 293ms/step - loss: 0.0074 - accuracy: 0.6147\n",
            "Epoch 2861/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0075 - accuracy: 0.5952\n",
            "Epoch 2862/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0075 - accuracy: 0.6120\n",
            "Epoch 2863/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0074 - accuracy: 0.6017\n",
            "Epoch 2864/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0074 - accuracy: 0.6088\n",
            "Epoch 2865/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0073 - accuracy: 0.6162\n",
            "Epoch 2866/5000\n",
            "1/1 [==============================] - 0s 298ms/step - loss: 0.0072 - accuracy: 0.6140\n",
            "Epoch 2867/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0072 - accuracy: 0.6192\n",
            "Epoch 2868/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0072 - accuracy: 0.6133\n",
            "Epoch 2869/5000\n",
            "1/1 [==============================] - 0s 295ms/step - loss: 0.0073 - accuracy: 0.6099\n",
            "Epoch 2870/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0073 - accuracy: 0.6041\n",
            "Epoch 2871/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0074 - accuracy: 0.6075\n",
            "Epoch 2872/5000\n",
            "1/1 [==============================] - 0s 301ms/step - loss: 0.0073 - accuracy: 0.6022\n",
            "Epoch 2873/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0073 - accuracy: 0.6114\n",
            "Epoch 2874/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0073 - accuracy: 0.6047\n",
            "Epoch 2875/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0073 - accuracy: 0.6193\n",
            "Epoch 2876/5000\n",
            "1/1 [==============================] - 0s 300ms/step - loss: 0.0073 - accuracy: 0.6046\n",
            "Epoch 2877/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0072 - accuracy: 0.6163\n",
            "Epoch 2878/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0072 - accuracy: 0.6071\n",
            "Epoch 2879/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0072 - accuracy: 0.6084\n",
            "Epoch 2880/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0072 - accuracy: 0.6119\n",
            "Epoch 2881/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0072 - accuracy: 0.6030\n",
            "Epoch 2882/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0072 - accuracy: 0.6144\n",
            "Epoch 2883/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0073 - accuracy: 0.6054\n",
            "Epoch 2884/5000\n",
            "1/1 [==============================] - 0s 291ms/step - loss: 0.0073 - accuracy: 0.6153\n",
            "Epoch 2885/5000\n",
            "1/1 [==============================] - 0s 261ms/step - loss: 0.0073 - accuracy: 0.6072\n",
            "Epoch 2886/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0073 - accuracy: 0.6140\n",
            "Epoch 2887/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0073 - accuracy: 0.6086\n",
            "Epoch 2888/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0073 - accuracy: 0.6101\n",
            "Epoch 2889/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0072 - accuracy: 0.6106\n",
            "Epoch 2890/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0072 - accuracy: 0.6085\n",
            "Epoch 2891/5000\n",
            "1/1 [==============================] - 0s 295ms/step - loss: 0.0072 - accuracy: 0.6116\n",
            "Epoch 2892/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0072 - accuracy: 0.6081\n",
            "Epoch 2893/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0072 - accuracy: 0.6117\n",
            "Epoch 2894/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0072 - accuracy: 0.6090\n",
            "Epoch 2895/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0072 - accuracy: 0.6111\n",
            "Epoch 2896/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0072 - accuracy: 0.6088\n",
            "Epoch 2897/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0072 - accuracy: 0.6121\n",
            "Epoch 2898/5000\n",
            "1/1 [==============================] - 0s 291ms/step - loss: 0.0073 - accuracy: 0.6091\n",
            "Epoch 2899/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0073 - accuracy: 0.6171\n",
            "Epoch 2900/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0073 - accuracy: 0.6033\n",
            "Epoch 2901/5000\n",
            "1/1 [==============================] - 0s 291ms/step - loss: 0.0074 - accuracy: 0.6215\n",
            "Epoch 2902/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0074 - accuracy: 0.6024\n",
            "Epoch 2903/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0073 - accuracy: 0.6136\n",
            "Epoch 2904/5000\n",
            "1/1 [==============================] - 0s 259ms/step - loss: 0.0073 - accuracy: 0.6035\n",
            "Epoch 2905/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0073 - accuracy: 0.6064\n",
            "Epoch 2906/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0072 - accuracy: 0.6108\n",
            "Epoch 2907/5000\n",
            "1/1 [==============================] - 0s 260ms/step - loss: 0.0072 - accuracy: 0.6105\n",
            "Epoch 2908/5000\n",
            "1/1 [==============================] - 0s 259ms/step - loss: 0.0072 - accuracy: 0.6172\n",
            "Epoch 2909/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0072 - accuracy: 0.6131\n",
            "Epoch 2910/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0072 - accuracy: 0.6156\n",
            "Epoch 2911/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0072 - accuracy: 0.6086\n",
            "Epoch 2912/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0073 - accuracy: 0.6100\n",
            "Epoch 2913/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0073 - accuracy: 0.6036\n",
            "Epoch 2914/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0073 - accuracy: 0.6085\n",
            "Epoch 2915/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0073 - accuracy: 0.6070\n",
            "Epoch 2916/5000\n",
            "1/1 [==============================] - 0s 293ms/step - loss: 0.0073 - accuracy: 0.6149\n",
            "Epoch 2917/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0073 - accuracy: 0.6060\n",
            "Epoch 2918/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0073 - accuracy: 0.6230\n",
            "Epoch 2919/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0073 - accuracy: 0.6033\n",
            "Epoch 2920/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0072 - accuracy: 0.6111\n",
            "Epoch 2921/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0072 - accuracy: 0.6047\n",
            "Epoch 2922/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0072 - accuracy: 0.6060\n",
            "Epoch 2923/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0072 - accuracy: 0.6162\n",
            "Epoch 2924/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0072 - accuracy: 0.6106\n",
            "Epoch 2925/5000\n",
            "1/1 [==============================] - 0s 295ms/step - loss: 0.0072 - accuracy: 0.6230\n",
            "Epoch 2926/5000\n",
            "1/1 [==============================] - 0s 301ms/step - loss: 0.0072 - accuracy: 0.6108\n",
            "Epoch 2927/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0072 - accuracy: 0.6142\n",
            "Epoch 2928/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0072 - accuracy: 0.6018\n",
            "Epoch 2929/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0073 - accuracy: 0.6051\n",
            "Epoch 2930/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0073 - accuracy: 0.6038\n",
            "Epoch 2931/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0073 - accuracy: 0.6064\n",
            "Epoch 2932/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0072 - accuracy: 0.6150\n",
            "Epoch 2933/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0072 - accuracy: 0.6134\n",
            "Epoch 2934/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0072 - accuracy: 0.6180\n",
            "Epoch 2935/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0072 - accuracy: 0.6139\n",
            "Epoch 2936/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0072 - accuracy: 0.6111\n",
            "Epoch 2937/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0072 - accuracy: 0.6042\n",
            "Epoch 2938/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0073 - accuracy: 0.6063\n",
            "Epoch 2939/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0073 - accuracy: 0.6036\n",
            "Epoch 2940/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0073 - accuracy: 0.6150\n",
            "Epoch 2941/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0074 - accuracy: 0.6049\n",
            "Epoch 2942/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0074 - accuracy: 0.6274\n",
            "Epoch 2943/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0073 - accuracy: 0.6070\n",
            "Epoch 2944/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0073 - accuracy: 0.6153\n",
            "Epoch 2945/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0072 - accuracy: 0.6048\n",
            "Epoch 2946/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0072 - accuracy: 0.6048\n",
            "Epoch 2947/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0072 - accuracy: 0.6110\n",
            "Epoch 2948/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0072 - accuracy: 0.6045\n",
            "Epoch 2949/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0072 - accuracy: 0.6179\n",
            "Epoch 2950/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0072 - accuracy: 0.6118\n",
            "Epoch 2951/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0072 - accuracy: 0.6194\n",
            "Epoch 2952/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0073 - accuracy: 0.6117\n",
            "Epoch 2953/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0073 - accuracy: 0.6120\n",
            "Epoch 2954/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0073 - accuracy: 0.6044\n",
            "Epoch 2955/5000\n",
            "1/1 [==============================] - 0s 290ms/step - loss: 0.0073 - accuracy: 0.6062\n",
            "Epoch 2956/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0072 - accuracy: 0.6040\n",
            "Epoch 2957/5000\n",
            "1/1 [==============================] - 0s 261ms/step - loss: 0.0072 - accuracy: 0.6119\n",
            "Epoch 2958/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0072 - accuracy: 0.6125\n",
            "Epoch 2959/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0072 - accuracy: 0.6177\n",
            "Epoch 2960/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0072 - accuracy: 0.6115\n",
            "Epoch 2961/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0072 - accuracy: 0.6130\n",
            "Epoch 2962/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0072 - accuracy: 0.6028\n",
            "Epoch 2963/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0072 - accuracy: 0.6108\n",
            "Epoch 2964/5000\n",
            "1/1 [==============================] - 0s 292ms/step - loss: 0.0072 - accuracy: 0.6058\n",
            "Epoch 2965/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0072 - accuracy: 0.6174\n",
            "Epoch 2966/5000\n",
            "1/1 [==============================] - 0s 292ms/step - loss: 0.0072 - accuracy: 0.6150\n",
            "Epoch 2967/5000\n",
            "1/1 [==============================] - 0s 293ms/step - loss: 0.0072 - accuracy: 0.6194\n",
            "Epoch 2968/5000\n",
            "1/1 [==============================] - 0s 291ms/step - loss: 0.0072 - accuracy: 0.6119\n",
            "Epoch 2969/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0072 - accuracy: 0.6082\n",
            "Epoch 2970/5000\n",
            "1/1 [==============================] - 0s 301ms/step - loss: 0.0072 - accuracy: 0.6017\n",
            "Epoch 2971/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0072 - accuracy: 0.6051\n",
            "Epoch 2972/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0072 - accuracy: 0.6023\n",
            "Epoch 2973/5000\n",
            "1/1 [==============================] - 0s 293ms/step - loss: 0.0073 - accuracy: 0.6187\n",
            "Epoch 2974/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0073 - accuracy: 0.6030\n",
            "Epoch 2975/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0073 - accuracy: 0.6247\n",
            "Epoch 2976/5000\n",
            "1/1 [==============================] - 0s 292ms/step - loss: 0.0072 - accuracy: 0.6095\n",
            "Epoch 2977/5000\n",
            "1/1 [==============================] - 0s 296ms/step - loss: 0.0072 - accuracy: 0.6094\n",
            "Epoch 2978/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0072 - accuracy: 0.6127\n",
            "Epoch 2979/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0072 - accuracy: 0.6034\n",
            "Epoch 2980/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0072 - accuracy: 0.6148\n",
            "Epoch 2981/5000\n",
            "1/1 [==============================] - 0s 296ms/step - loss: 0.0072 - accuracy: 0.6087\n",
            "Epoch 2982/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0072 - accuracy: 0.6151\n",
            "Epoch 2983/5000\n",
            "1/1 [==============================] - 0s 302ms/step - loss: 0.0072 - accuracy: 0.6114\n",
            "Epoch 2984/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0073 - accuracy: 0.6125\n",
            "Epoch 2985/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0073 - accuracy: 0.6075\n",
            "Epoch 2986/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0073 - accuracy: 0.6087\n",
            "Epoch 2987/5000\n",
            "1/1 [==============================] - 0s 292ms/step - loss: 0.0072 - accuracy: 0.6063\n",
            "Epoch 2988/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0072 - accuracy: 0.6117\n",
            "Epoch 2989/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0072 - accuracy: 0.6080\n",
            "Epoch 2990/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0072 - accuracy: 0.6205\n",
            "Epoch 2991/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0072 - accuracy: 0.6032\n",
            "Epoch 2992/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0072 - accuracy: 0.6172\n",
            "Epoch 2993/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0072 - accuracy: 0.6053\n",
            "Epoch 2994/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0072 - accuracy: 0.6053\n",
            "Epoch 2995/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0072 - accuracy: 0.6125\n",
            "Epoch 2996/5000\n",
            "1/1 [==============================] - 0s 258ms/step - loss: 0.0072 - accuracy: 0.6051\n",
            "Epoch 2997/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0072 - accuracy: 0.6183\n",
            "Epoch 2998/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0072 - accuracy: 0.6143\n",
            "Epoch 2999/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0072 - accuracy: 0.6185\n",
            "Epoch 3000/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0072 - accuracy: 0.6118\n",
            "Epoch 3001/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0072 - accuracy: 0.6110\n",
            "Epoch 3002/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0072 - accuracy: 0.6011\n",
            "Epoch 3003/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0073 - accuracy: 0.6108\n",
            "Epoch 3004/5000\n",
            "1/1 [==============================] - 0s 299ms/step - loss: 0.0073 - accuracy: 0.5990\n",
            "Epoch 3005/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0073 - accuracy: 0.6184\n",
            "Epoch 3006/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0073 - accuracy: 0.6050\n",
            "Epoch 3007/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0073 - accuracy: 0.6205\n",
            "Epoch 3008/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0072 - accuracy: 0.6102\n",
            "Epoch 3009/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0072 - accuracy: 0.6153\n",
            "Epoch 3010/5000\n",
            "1/1 [==============================] - 0s 294ms/step - loss: 0.0072 - accuracy: 0.6140\n",
            "Epoch 3011/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0072 - accuracy: 0.6040\n",
            "Epoch 3012/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0072 - accuracy: 0.6151\n",
            "Epoch 3013/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0072 - accuracy: 0.6065\n",
            "Epoch 3014/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0072 - accuracy: 0.6177\n",
            "Epoch 3015/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0072 - accuracy: 0.6154\n",
            "Epoch 3016/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0073 - accuracy: 0.6152\n",
            "Epoch 3017/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0073 - accuracy: 0.6119\n",
            "Epoch 3018/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0073 - accuracy: 0.6097\n",
            "Epoch 3019/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0072 - accuracy: 0.5999\n",
            "Epoch 3020/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0072 - accuracy: 0.6155\n",
            "Epoch 3021/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0072 - accuracy: 0.6029\n",
            "Epoch 3022/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0072 - accuracy: 0.6161\n",
            "Epoch 3023/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0072 - accuracy: 0.6098\n",
            "Epoch 3024/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0071 - accuracy: 0.6128\n",
            "Epoch 3025/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0071 - accuracy: 0.6135\n",
            "Epoch 3026/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0071 - accuracy: 0.6131\n",
            "Epoch 3027/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0072 - accuracy: 0.6168\n",
            "Epoch 3028/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0072 - accuracy: 0.6096\n",
            "Epoch 3029/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0072 - accuracy: 0.6168\n",
            "Epoch 3030/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0072 - accuracy: 0.6034\n",
            "Epoch 3031/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0072 - accuracy: 0.6161\n",
            "Epoch 3032/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0072 - accuracy: 0.6038\n",
            "Epoch 3033/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0072 - accuracy: 0.6147\n",
            "Epoch 3034/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0072 - accuracy: 0.6093\n",
            "Epoch 3035/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0072 - accuracy: 0.6166\n",
            "Epoch 3036/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0072 - accuracy: 0.6140\n",
            "Epoch 3037/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0072 - accuracy: 0.6179\n",
            "Epoch 3038/5000\n",
            "1/1 [==============================] - 0s 297ms/step - loss: 0.0072 - accuracy: 0.6122\n",
            "Epoch 3039/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0071 - accuracy: 0.6119\n",
            "Epoch 3040/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0071 - accuracy: 0.6060\n",
            "Epoch 3041/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0071 - accuracy: 0.6066\n",
            "Epoch 3042/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0071 - accuracy: 0.6072\n",
            "Epoch 3043/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0071 - accuracy: 0.6145\n",
            "Epoch 3044/5000\n",
            "1/1 [==============================] - 0s 261ms/step - loss: 0.0071 - accuracy: 0.6157\n",
            "Epoch 3045/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0072 - accuracy: 0.6243\n",
            "Epoch 3046/5000\n",
            "1/1 [==============================] - 0s 258ms/step - loss: 0.0072 - accuracy: 0.6136\n",
            "Epoch 3047/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0072 - accuracy: 0.6175\n",
            "Epoch 3048/5000\n",
            "1/1 [==============================] - 0s 258ms/step - loss: 0.0072 - accuracy: 0.6025\n",
            "Epoch 3049/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0072 - accuracy: 0.6089\n",
            "Epoch 3050/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0072 - accuracy: 0.6013\n",
            "Epoch 3051/5000\n",
            "1/1 [==============================] - 0s 255ms/step - loss: 0.0072 - accuracy: 0.6145\n",
            "Epoch 3052/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0073 - accuracy: 0.6086\n",
            "Epoch 3053/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0074 - accuracy: 0.6217\n",
            "Epoch 3054/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0074 - accuracy: 0.6083\n",
            "Epoch 3055/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0075 - accuracy: 0.6204\n",
            "Epoch 3056/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0074 - accuracy: 0.5990\n",
            "Epoch 3057/5000\n",
            "1/1 [==============================] - 0s 256ms/step - loss: 0.0074 - accuracy: 0.6167\n",
            "Epoch 3058/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0072 - accuracy: 0.6013\n",
            "Epoch 3059/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0072 - accuracy: 0.6076\n",
            "Epoch 3060/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0071 - accuracy: 0.6169\n",
            "Epoch 3061/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0072 - accuracy: 0.6086\n",
            "Epoch 3062/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0072 - accuracy: 0.6269\n",
            "Epoch 3063/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0073 - accuracy: 0.6144\n",
            "Epoch 3064/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0073 - accuracy: 0.6191\n",
            "Epoch 3065/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0073 - accuracy: 0.6125\n",
            "Epoch 3066/5000\n",
            "1/1 [==============================] - 0s 296ms/step - loss: 0.0072 - accuracy: 0.6062\n",
            "Epoch 3067/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0072 - accuracy: 0.6055\n",
            "Epoch 3068/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0071 - accuracy: 0.6054\n",
            "Epoch 3069/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0071 - accuracy: 0.6084\n",
            "Epoch 3070/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0071 - accuracy: 0.6142\n",
            "Epoch 3071/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0072 - accuracy: 0.6183\n",
            "Epoch 3072/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0072 - accuracy: 0.6199\n",
            "Epoch 3073/5000\n",
            "1/1 [==============================] - 0s 261ms/step - loss: 0.0072 - accuracy: 0.6179\n",
            "Epoch 3074/5000\n",
            "1/1 [==============================] - 0s 291ms/step - loss: 0.0072 - accuracy: 0.6110\n",
            "Epoch 3075/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0071 - accuracy: 0.6092\n",
            "Epoch 3076/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0072 - accuracy: 0.6008\n",
            "Epoch 3077/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0072 - accuracy: 0.6110\n",
            "Epoch 3078/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0072 - accuracy: 0.6034\n",
            "Epoch 3079/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0071 - accuracy: 0.6208\n",
            "Epoch 3080/5000\n",
            "1/1 [==============================] - 0s 260ms/step - loss: 0.0071 - accuracy: 0.6150\n",
            "Epoch 3081/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0071 - accuracy: 0.6160\n",
            "Epoch 3082/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0071 - accuracy: 0.6154\n",
            "Epoch 3083/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0071 - accuracy: 0.6024\n",
            "Epoch 3084/5000\n",
            "1/1 [==============================] - 0s 261ms/step - loss: 0.0072 - accuracy: 0.6116\n",
            "Epoch 3085/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0071 - accuracy: 0.6012\n",
            "Epoch 3086/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0071 - accuracy: 0.6116\n",
            "Epoch 3087/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0071 - accuracy: 0.6131\n",
            "Epoch 3088/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0071 - accuracy: 0.6156\n",
            "Epoch 3089/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0071 - accuracy: 0.6136\n",
            "Epoch 3090/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0071 - accuracy: 0.6142\n",
            "Epoch 3091/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0071 - accuracy: 0.6077\n",
            "Epoch 3092/5000\n",
            "1/1 [==============================] - 0s 295ms/step - loss: 0.0071 - accuracy: 0.6123\n",
            "Epoch 3093/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0071 - accuracy: 0.6053\n",
            "Epoch 3094/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0072 - accuracy: 0.6167\n",
            "Epoch 3095/5000\n",
            "1/1 [==============================] - 0s 290ms/step - loss: 0.0072 - accuracy: 0.6073\n",
            "Epoch 3096/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0072 - accuracy: 0.6192\n",
            "Epoch 3097/5000\n",
            "1/1 [==============================] - 0s 290ms/step - loss: 0.0071 - accuracy: 0.6084\n",
            "Epoch 3098/5000\n",
            "1/1 [==============================] - 0s 296ms/step - loss: 0.0071 - accuracy: 0.6186\n",
            "Epoch 3099/5000\n",
            "1/1 [==============================] - 0s 305ms/step - loss: 0.0071 - accuracy: 0.6101\n",
            "Epoch 3100/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0071 - accuracy: 0.6148\n",
            "Epoch 3101/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0071 - accuracy: 0.6113\n",
            "Epoch 3102/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0071 - accuracy: 0.6117\n",
            "Epoch 3103/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0071 - accuracy: 0.6100\n",
            "Epoch 3104/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0071 - accuracy: 0.6135\n",
            "Epoch 3105/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0071 - accuracy: 0.6098\n",
            "Epoch 3106/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0071 - accuracy: 0.6206\n",
            "Epoch 3107/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0071 - accuracy: 0.6091\n",
            "Epoch 3108/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0071 - accuracy: 0.6188\n",
            "Epoch 3109/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0071 - accuracy: 0.6058\n",
            "Epoch 3110/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0071 - accuracy: 0.6141\n",
            "Epoch 3111/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0071 - accuracy: 0.6071\n",
            "Epoch 3112/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0071 - accuracy: 0.6150\n",
            "Epoch 3113/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0071 - accuracy: 0.6129\n",
            "Epoch 3114/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0071 - accuracy: 0.6174\n",
            "Epoch 3115/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0071 - accuracy: 0.6131\n",
            "Epoch 3116/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0071 - accuracy: 0.6117\n",
            "Epoch 3117/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0071 - accuracy: 0.6083\n",
            "Epoch 3118/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0071 - accuracy: 0.6105\n",
            "Epoch 3119/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0071 - accuracy: 0.6100\n",
            "Epoch 3120/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0071 - accuracy: 0.6173\n",
            "Epoch 3121/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0071 - accuracy: 0.6114\n",
            "Epoch 3122/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0072 - accuracy: 0.6229\n",
            "Epoch 3123/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0071 - accuracy: 0.6079\n",
            "Epoch 3124/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0071 - accuracy: 0.6175\n",
            "Epoch 3125/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0071 - accuracy: 0.6047\n",
            "Epoch 3126/5000\n",
            "1/1 [==============================] - 0s 259ms/step - loss: 0.0071 - accuracy: 0.6126\n",
            "Epoch 3127/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0071 - accuracy: 0.6100\n",
            "Epoch 3128/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0071 - accuracy: 0.6128\n",
            "Epoch 3129/5000\n",
            "1/1 [==============================] - 0s 259ms/step - loss: 0.0071 - accuracy: 0.6156\n",
            "Epoch 3130/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0071 - accuracy: 0.6134\n",
            "Epoch 3131/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0071 - accuracy: 0.6153\n",
            "Epoch 3132/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0071 - accuracy: 0.6149\n",
            "Epoch 3133/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0071 - accuracy: 0.6109\n",
            "Epoch 3134/5000\n",
            "1/1 [==============================] - 0s 290ms/step - loss: 0.0071 - accuracy: 0.6158\n",
            "Epoch 3135/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0071 - accuracy: 0.6052\n",
            "Epoch 3136/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0071 - accuracy: 0.6185\n",
            "Epoch 3137/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0071 - accuracy: 0.6063\n",
            "Epoch 3138/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0071 - accuracy: 0.6168\n",
            "Epoch 3139/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0071 - accuracy: 0.6101\n",
            "Epoch 3140/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0071 - accuracy: 0.6154\n",
            "Epoch 3141/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0071 - accuracy: 0.6122\n",
            "Epoch 3142/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0071 - accuracy: 0.6145\n",
            "Epoch 3143/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0071 - accuracy: 0.6096\n",
            "Epoch 3144/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0071 - accuracy: 0.6107\n",
            "Epoch 3145/5000\n",
            "1/1 [==============================] - 0s 257ms/step - loss: 0.0071 - accuracy: 0.6089\n",
            "Epoch 3146/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0071 - accuracy: 0.6143\n",
            "Epoch 3147/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0072 - accuracy: 0.6104\n",
            "Epoch 3148/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0072 - accuracy: 0.6246\n",
            "Epoch 3149/5000\n",
            "1/1 [==============================] - 0s 258ms/step - loss: 0.0072 - accuracy: 0.6081\n",
            "Epoch 3150/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0072 - accuracy: 0.6224\n",
            "Epoch 3151/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0072 - accuracy: 0.6054\n",
            "Epoch 3152/5000\n",
            "1/1 [==============================] - 0s 261ms/step - loss: 0.0072 - accuracy: 0.6141\n",
            "Epoch 3153/5000\n",
            "1/1 [==============================] - 0s 294ms/step - loss: 0.0072 - accuracy: 0.6081\n",
            "Epoch 3154/5000\n",
            "1/1 [==============================] - 0s 257ms/step - loss: 0.0072 - accuracy: 0.6105\n",
            "Epoch 3155/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0072 - accuracy: 0.6140\n",
            "Epoch 3156/5000\n",
            "1/1 [==============================] - 0s 258ms/step - loss: 0.0071 - accuracy: 0.6128\n",
            "Epoch 3157/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0071 - accuracy: 0.6155\n",
            "Epoch 3158/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0071 - accuracy: 0.6148\n",
            "Epoch 3159/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0071 - accuracy: 0.6109\n",
            "Epoch 3160/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0071 - accuracy: 0.6100\n",
            "Epoch 3161/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0071 - accuracy: 0.6118\n",
            "Epoch 3162/5000\n",
            "1/1 [==============================] - 0s 290ms/step - loss: 0.0072 - accuracy: 0.6097\n",
            "Epoch 3163/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0072 - accuracy: 0.6231\n",
            "Epoch 3164/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0073 - accuracy: 0.6068\n",
            "Epoch 3165/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0074 - accuracy: 0.6261\n",
            "Epoch 3166/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0073 - accuracy: 0.5996\n",
            "Epoch 3167/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0073 - accuracy: 0.6080\n",
            "Epoch 3168/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0072 - accuracy: 0.6042\n",
            "Epoch 3169/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0072 - accuracy: 0.6089\n",
            "Epoch 3170/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0071 - accuracy: 0.6210\n",
            "Epoch 3171/5000\n",
            "1/1 [==============================] - 0s 293ms/step - loss: 0.0071 - accuracy: 0.6174\n",
            "Epoch 3172/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0071 - accuracy: 0.6264\n",
            "Epoch 3173/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0071 - accuracy: 0.6149\n",
            "Epoch 3174/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0072 - accuracy: 0.6147\n",
            "Epoch 3175/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0072 - accuracy: 0.5999\n",
            "Epoch 3176/5000\n",
            "1/1 [==============================] - 0s 309ms/step - loss: 0.0072 - accuracy: 0.6060\n",
            "Epoch 3177/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0072 - accuracy: 0.6018\n",
            "Epoch 3178/5000\n",
            "1/1 [==============================] - 0s 296ms/step - loss: 0.0072 - accuracy: 0.6169\n",
            "Epoch 3179/5000\n",
            "1/1 [==============================] - 0s 295ms/step - loss: 0.0072 - accuracy: 0.6155\n",
            "Epoch 3180/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0071 - accuracy: 0.6256\n",
            "Epoch 3181/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0071 - accuracy: 0.6123\n",
            "Epoch 3182/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0071 - accuracy: 0.6159\n",
            "Epoch 3183/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0071 - accuracy: 0.6015\n",
            "Epoch 3184/5000\n",
            "1/1 [==============================] - 0s 261ms/step - loss: 0.0071 - accuracy: 0.6050\n",
            "Epoch 3185/5000\n",
            "1/1 [==============================] - 0s 295ms/step - loss: 0.0071 - accuracy: 0.6097\n",
            "Epoch 3186/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0071 - accuracy: 0.6099\n",
            "Epoch 3187/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0071 - accuracy: 0.6297\n",
            "Epoch 3188/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0071 - accuracy: 0.6165\n",
            "Epoch 3189/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0071 - accuracy: 0.6243\n",
            "Epoch 3190/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0071 - accuracy: 0.6117\n",
            "Epoch 3191/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0071 - accuracy: 0.6079\n",
            "Epoch 3192/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0071 - accuracy: 0.6043\n",
            "Epoch 3193/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0071 - accuracy: 0.6035\n",
            "Epoch 3194/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0071 - accuracy: 0.6117\n",
            "Epoch 3195/5000\n",
            "1/1 [==============================] - 0s 258ms/step - loss: 0.0071 - accuracy: 0.6117\n",
            "Epoch 3196/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0071 - accuracy: 0.6230\n",
            "Epoch 3197/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0071 - accuracy: 0.6110\n",
            "Epoch 3198/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0071 - accuracy: 0.6210\n",
            "Epoch 3199/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0071 - accuracy: 0.6052\n",
            "Epoch 3200/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0071 - accuracy: 0.6115\n",
            "Epoch 3201/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0072 - accuracy: 0.6057\n",
            "Epoch 3202/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0072 - accuracy: 0.6159\n",
            "Epoch 3203/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0072 - accuracy: 0.6167\n",
            "Epoch 3204/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0071 - accuracy: 0.6250\n",
            "Epoch 3205/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0071 - accuracy: 0.6157\n",
            "Epoch 3206/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0071 - accuracy: 0.6195\n",
            "Epoch 3207/5000\n",
            "1/1 [==============================] - 0s 294ms/step - loss: 0.0071 - accuracy: 0.6030\n",
            "Epoch 3208/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0071 - accuracy: 0.6079\n",
            "Epoch 3209/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0071 - accuracy: 0.6038\n",
            "Epoch 3210/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0071 - accuracy: 0.6154\n",
            "Epoch 3211/5000\n",
            "1/1 [==============================] - 0s 290ms/step - loss: 0.0071 - accuracy: 0.6200\n",
            "Epoch 3212/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0071 - accuracy: 0.6249\n",
            "Epoch 3213/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0071 - accuracy: 0.6189\n",
            "Epoch 3214/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0071 - accuracy: 0.6154\n",
            "Epoch 3215/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0071 - accuracy: 0.6060\n",
            "Epoch 3216/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0071 - accuracy: 0.6055\n",
            "Epoch 3217/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0071 - accuracy: 0.6051\n",
            "Epoch 3218/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0071 - accuracy: 0.6159\n",
            "Epoch 3219/5000\n",
            "1/1 [==============================] - 0s 293ms/step - loss: 0.0071 - accuracy: 0.6120\n",
            "Epoch 3220/5000\n",
            "1/1 [==============================] - 0s 298ms/step - loss: 0.0071 - accuracy: 0.6270\n",
            "Epoch 3221/5000\n",
            "1/1 [==============================] - 0s 294ms/step - loss: 0.0071 - accuracy: 0.6105\n",
            "Epoch 3222/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0071 - accuracy: 0.6193\n",
            "Epoch 3223/5000\n",
            "1/1 [==============================] - 0s 290ms/step - loss: 0.0071 - accuracy: 0.6065\n",
            "Epoch 3224/5000\n",
            "1/1 [==============================] - 0s 290ms/step - loss: 0.0072 - accuracy: 0.6142\n",
            "Epoch 3225/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0072 - accuracy: 0.6092\n",
            "Epoch 3226/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0072 - accuracy: 0.6171\n",
            "Epoch 3227/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0072 - accuracy: 0.6116\n",
            "Epoch 3228/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0072 - accuracy: 0.6194\n",
            "Epoch 3229/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0072 - accuracy: 0.6084\n",
            "Epoch 3230/5000\n",
            "1/1 [==============================] - 0s 291ms/step - loss: 0.0072 - accuracy: 0.6183\n",
            "Epoch 3231/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0071 - accuracy: 0.6021\n",
            "Epoch 3232/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0071 - accuracy: 0.6168\n",
            "Epoch 3233/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0071 - accuracy: 0.6076\n",
            "Epoch 3234/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0070 - accuracy: 0.6170\n",
            "Epoch 3235/5000\n",
            "1/1 [==============================] - 0s 294ms/step - loss: 0.0071 - accuracy: 0.6231\n",
            "Epoch 3236/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0071 - accuracy: 0.6161\n",
            "Epoch 3237/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0071 - accuracy: 0.6251\n",
            "Epoch 3238/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0072 - accuracy: 0.6065\n",
            "Epoch 3239/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0072 - accuracy: 0.6141\n",
            "Epoch 3240/5000\n",
            "1/1 [==============================] - 0s 252ms/step - loss: 0.0072 - accuracy: 0.5974\n",
            "Epoch 3241/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0071 - accuracy: 0.6090\n",
            "Epoch 3242/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0071 - accuracy: 0.6085\n",
            "Epoch 3243/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0071 - accuracy: 0.6189\n",
            "Epoch 3244/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0071 - accuracy: 0.6244\n",
            "Epoch 3245/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0071 - accuracy: 0.6185\n",
            "Epoch 3246/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0071 - accuracy: 0.6239\n",
            "Epoch 3247/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0071 - accuracy: 0.6100\n",
            "Epoch 3248/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0071 - accuracy: 0.6115\n",
            "Epoch 3249/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0071 - accuracy: 0.6035\n",
            "Epoch 3250/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0071 - accuracy: 0.6081\n",
            "Epoch 3251/5000\n",
            "1/1 [==============================] - 0s 259ms/step - loss: 0.0071 - accuracy: 0.6085\n",
            "Epoch 3252/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0071 - accuracy: 0.6168\n",
            "Epoch 3253/5000\n",
            "1/1 [==============================] - 0s 256ms/step - loss: 0.0071 - accuracy: 0.6159\n",
            "Epoch 3254/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0071 - accuracy: 0.6258\n",
            "Epoch 3255/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0071 - accuracy: 0.6091\n",
            "Epoch 3256/5000\n",
            "1/1 [==============================] - 0s 256ms/step - loss: 0.0071 - accuracy: 0.6222\n",
            "Epoch 3257/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0071 - accuracy: 0.6007\n",
            "Epoch 3258/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0071 - accuracy: 0.6074\n",
            "Epoch 3259/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0071 - accuracy: 0.6090\n",
            "Epoch 3260/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0070 - accuracy: 0.6114\n",
            "Epoch 3261/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0070 - accuracy: 0.6235\n",
            "Epoch 3262/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0071 - accuracy: 0.6186\n",
            "Epoch 3263/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0071 - accuracy: 0.6269\n",
            "Epoch 3264/5000\n",
            "1/1 [==============================] - 0s 297ms/step - loss: 0.0071 - accuracy: 0.6096\n",
            "Epoch 3265/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0071 - accuracy: 0.6132\n",
            "Epoch 3266/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0071 - accuracy: 0.5982\n",
            "Epoch 3267/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0071 - accuracy: 0.6075\n",
            "Epoch 3268/5000\n",
            "1/1 [==============================] - 0s 261ms/step - loss: 0.0071 - accuracy: 0.6078\n",
            "Epoch 3269/5000\n",
            "1/1 [==============================] - 0s 259ms/step - loss: 0.0071 - accuracy: 0.6143\n",
            "Epoch 3270/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0071 - accuracy: 0.6243\n",
            "Epoch 3271/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0071 - accuracy: 0.6170\n",
            "Epoch 3272/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0071 - accuracy: 0.6238\n",
            "Epoch 3273/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0070 - accuracy: 0.6105\n",
            "Epoch 3274/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0070 - accuracy: 0.6104\n",
            "Epoch 3275/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0071 - accuracy: 0.6057\n",
            "Epoch 3276/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0071 - accuracy: 0.6081\n",
            "Epoch 3277/5000\n",
            "1/1 [==============================] - 0s 258ms/step - loss: 0.0071 - accuracy: 0.6100\n",
            "Epoch 3278/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0071 - accuracy: 0.6213\n",
            "Epoch 3279/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0071 - accuracy: 0.6134\n",
            "Epoch 3280/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0072 - accuracy: 0.6273\n",
            "Epoch 3281/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0072 - accuracy: 0.6020\n",
            "Epoch 3282/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0072 - accuracy: 0.6144\n",
            "Epoch 3283/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0072 - accuracy: 0.6008\n",
            "Epoch 3284/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0071 - accuracy: 0.6113\n",
            "Epoch 3285/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0071 - accuracy: 0.6187\n",
            "Epoch 3286/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0071 - accuracy: 0.6158\n",
            "Epoch 3287/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0071 - accuracy: 0.6250\n",
            "Epoch 3288/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0071 - accuracy: 0.6126\n",
            "Epoch 3289/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0070 - accuracy: 0.6152\n",
            "Epoch 3290/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0071 - accuracy: 0.6064\n",
            "Epoch 3291/5000\n",
            "1/1 [==============================] - 0s 295ms/step - loss: 0.0071 - accuracy: 0.6098\n",
            "Epoch 3292/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0071 - accuracy: 0.6084\n",
            "Epoch 3293/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0071 - accuracy: 0.6117\n",
            "Epoch 3294/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0071 - accuracy: 0.6148\n",
            "Epoch 3295/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0071 - accuracy: 0.6178\n",
            "Epoch 3296/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0071 - accuracy: 0.6167\n",
            "Epoch 3297/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0070 - accuracy: 0.6195\n",
            "Epoch 3298/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0071 - accuracy: 0.6056\n",
            "Epoch 3299/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0071 - accuracy: 0.6171\n",
            "Epoch 3300/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0071 - accuracy: 0.5993\n",
            "Epoch 3301/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0070 - accuracy: 0.6108\n",
            "Epoch 3302/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0070 - accuracy: 0.6112\n",
            "Epoch 3303/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0070 - accuracy: 0.6142\n",
            "Epoch 3304/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0070 - accuracy: 0.6238\n",
            "Epoch 3305/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0071 - accuracy: 0.6133\n",
            "Epoch 3306/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0071 - accuracy: 0.6221\n",
            "Epoch 3307/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0070 - accuracy: 0.6094\n",
            "Epoch 3308/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0070 - accuracy: 0.6091\n",
            "Epoch 3309/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0070 - accuracy: 0.6067\n",
            "Epoch 3310/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0070 - accuracy: 0.6084\n",
            "Epoch 3311/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0071 - accuracy: 0.6114\n",
            "Epoch 3312/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0071 - accuracy: 0.6125\n",
            "Epoch 3313/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0070 - accuracy: 0.6215\n",
            "Epoch 3314/5000\n",
            "1/1 [==============================] - 0s 256ms/step - loss: 0.0070 - accuracy: 0.6150\n",
            "Epoch 3315/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0070 - accuracy: 0.6184\n",
            "Epoch 3316/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0070 - accuracy: 0.6103\n",
            "Epoch 3317/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0070 - accuracy: 0.6089\n",
            "Epoch 3318/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0070 - accuracy: 0.6056\n",
            "Epoch 3319/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0070 - accuracy: 0.6097\n",
            "Epoch 3320/5000\n",
            "1/1 [==============================] - 0s 290ms/step - loss: 0.0070 - accuracy: 0.6131\n",
            "Epoch 3321/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0071 - accuracy: 0.6219\n",
            "Epoch 3322/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0072 - accuracy: 0.6124\n",
            "Epoch 3323/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0073 - accuracy: 0.6290\n",
            "Epoch 3324/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0073 - accuracy: 0.5991\n",
            "Epoch 3325/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0074 - accuracy: 0.6165\n",
            "Epoch 3326/5000\n",
            "1/1 [==============================] - 0s 300ms/step - loss: 0.0074 - accuracy: 0.6002\n",
            "Epoch 3327/5000\n",
            "1/1 [==============================] - 0s 261ms/step - loss: 0.0074 - accuracy: 0.6125\n",
            "Epoch 3328/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0073 - accuracy: 0.6148\n",
            "Epoch 3329/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0071 - accuracy: 0.6183\n",
            "Epoch 3330/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0070 - accuracy: 0.6192\n",
            "Epoch 3331/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0070 - accuracy: 0.6141\n",
            "Epoch 3332/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0071 - accuracy: 0.6108\n",
            "Epoch 3333/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0071 - accuracy: 0.6122\n",
            "Epoch 3334/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0072 - accuracy: 0.6082\n",
            "Epoch 3335/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0071 - accuracy: 0.6114\n",
            "Epoch 3336/5000\n",
            "1/1 [==============================] - 0s 260ms/step - loss: 0.0071 - accuracy: 0.6171\n",
            "Epoch 3337/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0071 - accuracy: 0.6083\n",
            "Epoch 3338/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0071 - accuracy: 0.6255\n",
            "Epoch 3339/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0071 - accuracy: 0.6050\n",
            "Epoch 3340/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0070 - accuracy: 0.6105\n",
            "Epoch 3341/5000\n",
            "1/1 [==============================] - 0s 260ms/step - loss: 0.0070 - accuracy: 0.6082\n",
            "Epoch 3342/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0070 - accuracy: 0.6028\n",
            "Epoch 3343/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0070 - accuracy: 0.6155\n",
            "Epoch 3344/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0071 - accuracy: 0.6093\n",
            "Epoch 3345/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0071 - accuracy: 0.6232\n",
            "Epoch 3346/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0070 - accuracy: 0.6164\n",
            "Epoch 3347/5000\n",
            "1/1 [==============================] - 0s 293ms/step - loss: 0.0071 - accuracy: 0.6176\n",
            "Epoch 3348/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0071 - accuracy: 0.6110\n",
            "Epoch 3349/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0071 - accuracy: 0.6075\n",
            "Epoch 3350/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0070 - accuracy: 0.6092\n",
            "Epoch 3351/5000\n",
            "1/1 [==============================] - 0s 247ms/step - loss: 0.0070 - accuracy: 0.6102\n",
            "Epoch 3352/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0070 - accuracy: 0.6161\n",
            "Epoch 3353/5000\n",
            "1/1 [==============================] - 0s 259ms/step - loss: 0.0070 - accuracy: 0.6132\n",
            "Epoch 3354/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0070 - accuracy: 0.6172\n",
            "Epoch 3355/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0071 - accuracy: 0.6077\n",
            "Epoch 3356/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0072 - accuracy: 0.6189\n",
            "Epoch 3357/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0072 - accuracy: 0.6001\n",
            "Epoch 3358/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0072 - accuracy: 0.6206\n",
            "Epoch 3359/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0072 - accuracy: 0.6057\n",
            "Epoch 3360/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0071 - accuracy: 0.6180\n",
            "Epoch 3361/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0071 - accuracy: 0.6144\n",
            "Epoch 3362/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0070 - accuracy: 0.6157\n",
            "Epoch 3363/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0070 - accuracy: 0.6183\n",
            "Epoch 3364/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0070 - accuracy: 0.6133\n",
            "Epoch 3365/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0071 - accuracy: 0.6218\n",
            "Epoch 3366/5000\n",
            "1/1 [==============================] - 0s 293ms/step - loss: 0.0071 - accuracy: 0.6111\n",
            "Epoch 3367/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0072 - accuracy: 0.6210\n",
            "Epoch 3368/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0072 - accuracy: 0.6034\n",
            "Epoch 3369/5000\n",
            "1/1 [==============================] - 0s 294ms/step - loss: 0.0072 - accuracy: 0.6123\n",
            "Epoch 3370/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0071 - accuracy: 0.5999\n",
            "Epoch 3371/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0071 - accuracy: 0.6138\n",
            "Epoch 3372/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0070 - accuracy: 0.6155\n",
            "Epoch 3373/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0070 - accuracy: 0.6213\n",
            "Epoch 3374/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0070 - accuracy: 0.6238\n",
            "Epoch 3375/5000\n",
            "1/1 [==============================] - 0s 305ms/step - loss: 0.0070 - accuracy: 0.6118\n",
            "Epoch 3376/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0071 - accuracy: 0.6173\n",
            "Epoch 3377/5000\n",
            "1/1 [==============================] - 0s 308ms/step - loss: 0.0071 - accuracy: 0.5999\n",
            "Epoch 3378/5000\n",
            "1/1 [==============================] - 0s 258ms/step - loss: 0.0071 - accuracy: 0.6157\n",
            "Epoch 3379/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0071 - accuracy: 0.6094\n",
            "Epoch 3380/5000\n",
            "1/1 [==============================] - 0s 305ms/step - loss: 0.0071 - accuracy: 0.6220\n",
            "Epoch 3381/5000\n",
            "1/1 [==============================] - 0s 297ms/step - loss: 0.0070 - accuracy: 0.6197\n",
            "Epoch 3382/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0070 - accuracy: 0.6213\n",
            "Epoch 3383/5000\n",
            "1/1 [==============================] - 0s 295ms/step - loss: 0.0070 - accuracy: 0.6137\n",
            "Epoch 3384/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0070 - accuracy: 0.6135\n",
            "Epoch 3385/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0070 - accuracy: 0.6076\n",
            "Epoch 3386/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0070 - accuracy: 0.6081\n",
            "Epoch 3387/5000\n",
            "1/1 [==============================] - 0s 306ms/step - loss: 0.0070 - accuracy: 0.6135\n",
            "Epoch 3388/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0070 - accuracy: 0.6165\n",
            "Epoch 3389/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0070 - accuracy: 0.6168\n",
            "Epoch 3390/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0070 - accuracy: 0.6223\n",
            "Epoch 3391/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0070 - accuracy: 0.6126\n",
            "Epoch 3392/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0070 - accuracy: 0.6194\n",
            "Epoch 3393/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0070 - accuracy: 0.6050\n",
            "Epoch 3394/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0071 - accuracy: 0.6142\n",
            "Epoch 3395/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0071 - accuracy: 0.6039\n",
            "Epoch 3396/5000\n",
            "1/1 [==============================] - 0s 302ms/step - loss: 0.0071 - accuracy: 0.6170\n",
            "Epoch 3397/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0071 - accuracy: 0.6104\n",
            "Epoch 3398/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0071 - accuracy: 0.6200\n",
            "Epoch 3399/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0070 - accuracy: 0.6116\n",
            "Epoch 3400/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0070 - accuracy: 0.6166\n",
            "Epoch 3401/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0070 - accuracy: 0.6097\n",
            "Epoch 3402/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0070 - accuracy: 0.6139\n",
            "Epoch 3403/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0070 - accuracy: 0.6115\n",
            "Epoch 3404/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0070 - accuracy: 0.6224\n",
            "Epoch 3405/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0070 - accuracy: 0.6130\n",
            "Epoch 3406/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0070 - accuracy: 0.6241\n",
            "Epoch 3407/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0070 - accuracy: 0.6123\n",
            "Epoch 3408/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0070 - accuracy: 0.6120\n",
            "Epoch 3409/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0070 - accuracy: 0.6074\n",
            "Epoch 3410/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0070 - accuracy: 0.6059\n",
            "Epoch 3411/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0070 - accuracy: 0.6125\n",
            "Epoch 3412/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0070 - accuracy: 0.6119\n",
            "Epoch 3413/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0070 - accuracy: 0.6204\n",
            "Epoch 3414/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0070 - accuracy: 0.6153\n",
            "Epoch 3415/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0070 - accuracy: 0.6178\n",
            "Epoch 3416/5000\n",
            "1/1 [==============================] - 0s 260ms/step - loss: 0.0070 - accuracy: 0.6069\n",
            "Epoch 3417/5000\n",
            "1/1 [==============================] - 0s 259ms/step - loss: 0.0070 - accuracy: 0.6147\n",
            "Epoch 3418/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0071 - accuracy: 0.6042\n",
            "Epoch 3419/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0071 - accuracy: 0.6215\n",
            "Epoch 3420/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0072 - accuracy: 0.6084\n",
            "Epoch 3421/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0072 - accuracy: 0.6236\n",
            "Epoch 3422/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0072 - accuracy: 0.6058\n",
            "Epoch 3423/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0072 - accuracy: 0.6164\n",
            "Epoch 3424/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0071 - accuracy: 0.6050\n",
            "Epoch 3425/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0071 - accuracy: 0.6131\n",
            "Epoch 3426/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0070 - accuracy: 0.6135\n",
            "Epoch 3427/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0070 - accuracy: 0.6167\n",
            "Epoch 3428/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0070 - accuracy: 0.6204\n",
            "Epoch 3429/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0071 - accuracy: 0.6130\n",
            "Epoch 3430/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0071 - accuracy: 0.6221\n",
            "Epoch 3431/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0071 - accuracy: 0.6052\n",
            "Epoch 3432/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0071 - accuracy: 0.6196\n",
            "Epoch 3433/5000\n",
            "1/1 [==============================] - 0s 290ms/step - loss: 0.0071 - accuracy: 0.6046\n",
            "Epoch 3434/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0070 - accuracy: 0.6169\n",
            "Epoch 3435/5000\n",
            "1/1 [==============================] - 0s 292ms/step - loss: 0.0070 - accuracy: 0.6105\n",
            "Epoch 3436/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0070 - accuracy: 0.6165\n",
            "Epoch 3437/5000\n",
            "1/1 [==============================] - 0s 297ms/step - loss: 0.0070 - accuracy: 0.6169\n",
            "Epoch 3438/5000\n",
            "1/1 [==============================] - 0s 298ms/step - loss: 0.0070 - accuracy: 0.6128\n",
            "Epoch 3439/5000\n",
            "1/1 [==============================] - 0s 304ms/step - loss: 0.0071 - accuracy: 0.6203\n",
            "Epoch 3440/5000\n",
            "1/1 [==============================] - 0s 291ms/step - loss: 0.0071 - accuracy: 0.6058\n",
            "Epoch 3441/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0071 - accuracy: 0.6237\n",
            "Epoch 3442/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0071 - accuracy: 0.6078\n",
            "Epoch 3443/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0070 - accuracy: 0.6203\n",
            "Epoch 3444/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0070 - accuracy: 0.6147\n",
            "Epoch 3445/5000\n",
            "1/1 [==============================] - 0s 297ms/step - loss: 0.0070 - accuracy: 0.6098\n",
            "Epoch 3446/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0070 - accuracy: 0.6175\n",
            "Epoch 3447/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0070 - accuracy: 0.6055\n",
            "Epoch 3448/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0070 - accuracy: 0.6193\n",
            "Epoch 3449/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0070 - accuracy: 0.6120\n",
            "Epoch 3450/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0070 - accuracy: 0.6178\n",
            "Epoch 3451/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0070 - accuracy: 0.6130\n",
            "Epoch 3452/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0070 - accuracy: 0.6102\n",
            "Epoch 3453/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0070 - accuracy: 0.6091\n",
            "Epoch 3454/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0070 - accuracy: 0.6130\n",
            "Epoch 3455/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0070 - accuracy: 0.6145\n",
            "Epoch 3456/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0070 - accuracy: 0.6213\n",
            "Epoch 3457/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0070 - accuracy: 0.6181\n",
            "Epoch 3458/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0070 - accuracy: 0.6175\n",
            "Epoch 3459/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0070 - accuracy: 0.6109\n",
            "Epoch 3460/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0070 - accuracy: 0.6121\n",
            "Epoch 3461/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0070 - accuracy: 0.6048\n",
            "Epoch 3462/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0070 - accuracy: 0.6160\n",
            "Epoch 3463/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0070 - accuracy: 0.6102\n",
            "Epoch 3464/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0070 - accuracy: 0.6199\n",
            "Epoch 3465/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0070 - accuracy: 0.6119\n",
            "Epoch 3466/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0070 - accuracy: 0.6186\n",
            "Epoch 3467/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0070 - accuracy: 0.6064\n",
            "Epoch 3468/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0071 - accuracy: 0.6172\n",
            "Epoch 3469/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0071 - accuracy: 0.6058\n",
            "Epoch 3470/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0071 - accuracy: 0.6196\n",
            "Epoch 3471/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0071 - accuracy: 0.6058\n",
            "Epoch 3472/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0070 - accuracy: 0.6195\n",
            "Epoch 3473/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0070 - accuracy: 0.6092\n",
            "Epoch 3474/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0070 - accuracy: 0.6156\n",
            "Epoch 3475/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0070 - accuracy: 0.6152\n",
            "Epoch 3476/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0070 - accuracy: 0.6113\n",
            "Epoch 3477/5000\n",
            "1/1 [==============================] - 0s 291ms/step - loss: 0.0070 - accuracy: 0.6208\n",
            "Epoch 3478/5000\n",
            "1/1 [==============================] - 0s 302ms/step - loss: 0.0070 - accuracy: 0.6101\n",
            "Epoch 3479/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0070 - accuracy: 0.6237\n",
            "Epoch 3480/5000\n",
            "1/1 [==============================] - 0s 299ms/step - loss: 0.0070 - accuracy: 0.6099\n",
            "Epoch 3481/5000\n",
            "1/1 [==============================] - 0s 260ms/step - loss: 0.0070 - accuracy: 0.6173\n",
            "Epoch 3482/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0070 - accuracy: 0.6081\n",
            "Epoch 3483/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0070 - accuracy: 0.6087\n",
            "Epoch 3484/5000\n",
            "1/1 [==============================] - 0s 290ms/step - loss: 0.0070 - accuracy: 0.6124\n",
            "Epoch 3485/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0070 - accuracy: 0.6123\n",
            "Epoch 3486/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0070 - accuracy: 0.6196\n",
            "Epoch 3487/5000\n",
            "1/1 [==============================] - 0s 293ms/step - loss: 0.0070 - accuracy: 0.6168\n",
            "Epoch 3488/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0070 - accuracy: 0.6211\n",
            "Epoch 3489/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0070 - accuracy: 0.6111\n",
            "Epoch 3490/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0070 - accuracy: 0.6172\n",
            "Epoch 3491/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0070 - accuracy: 0.6039\n",
            "Epoch 3492/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0070 - accuracy: 0.6165\n",
            "Epoch 3493/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0070 - accuracy: 0.6071\n",
            "Epoch 3494/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0070 - accuracy: 0.6184\n",
            "Epoch 3495/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0070 - accuracy: 0.6115\n",
            "Epoch 3496/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0070 - accuracy: 0.6164\n",
            "Epoch 3497/5000\n",
            "1/1 [==============================] - 0s 307ms/step - loss: 0.0070 - accuracy: 0.6140\n",
            "Epoch 3498/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0070 - accuracy: 0.6129\n",
            "Epoch 3499/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0070 - accuracy: 0.6151\n",
            "Epoch 3500/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0070 - accuracy: 0.6109\n",
            "Epoch 3501/5000\n",
            "1/1 [==============================] - 0s 292ms/step - loss: 0.0069 - accuracy: 0.6156\n",
            "Epoch 3502/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0069 - accuracy: 0.6125\n",
            "Epoch 3503/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0069 - accuracy: 0.6174\n",
            "Epoch 3504/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0069 - accuracy: 0.6140\n",
            "Epoch 3505/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0070 - accuracy: 0.6190\n",
            "Epoch 3506/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0070 - accuracy: 0.6069\n",
            "Epoch 3507/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0070 - accuracy: 0.6203\n",
            "Epoch 3508/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0071 - accuracy: 0.6029\n",
            "Epoch 3509/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0071 - accuracy: 0.6201\n",
            "Epoch 3510/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0070 - accuracy: 0.6060\n",
            "Epoch 3511/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0070 - accuracy: 0.6165\n",
            "Epoch 3512/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0070 - accuracy: 0.6142\n",
            "Epoch 3513/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0071 - accuracy: 0.6103\n",
            "Epoch 3514/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0070 - accuracy: 0.6186\n",
            "Epoch 3515/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0070 - accuracy: 0.6141\n",
            "Epoch 3516/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0069 - accuracy: 0.6178\n",
            "Epoch 3517/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0069 - accuracy: 0.6145\n",
            "Epoch 3518/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0069 - accuracy: 0.6099\n",
            "Epoch 3519/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0069 - accuracy: 0.6097\n",
            "Epoch 3520/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0070 - accuracy: 0.6058\n",
            "Epoch 3521/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0070 - accuracy: 0.6128\n",
            "Epoch 3522/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0070 - accuracy: 0.6085\n",
            "Epoch 3523/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0070 - accuracy: 0.6179\n",
            "Epoch 3524/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0069 - accuracy: 0.6125\n",
            "Epoch 3525/5000\n",
            "1/1 [==============================] - 0s 296ms/step - loss: 0.0069 - accuracy: 0.6170\n",
            "Epoch 3526/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0070 - accuracy: 0.6104\n",
            "Epoch 3527/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0070 - accuracy: 0.6147\n",
            "Epoch 3528/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0070 - accuracy: 0.6050\n",
            "Epoch 3529/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0071 - accuracy: 0.6194\n",
            "Epoch 3530/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0072 - accuracy: 0.6042\n",
            "Epoch 3531/5000\n",
            "1/1 [==============================] - 0s 261ms/step - loss: 0.0073 - accuracy: 0.6241\n",
            "Epoch 3532/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0073 - accuracy: 0.6029\n",
            "Epoch 3533/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0072 - accuracy: 0.6201\n",
            "Epoch 3534/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0071 - accuracy: 0.6072\n",
            "Epoch 3535/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0070 - accuracy: 0.6124\n",
            "Epoch 3536/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0069 - accuracy: 0.6139\n",
            "Epoch 3537/5000\n",
            "1/1 [==============================] - 0s 296ms/step - loss: 0.0070 - accuracy: 0.6088\n",
            "Epoch 3538/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0070 - accuracy: 0.6249\n",
            "Epoch 3539/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0071 - accuracy: 0.6137\n",
            "Epoch 3540/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0071 - accuracy: 0.6266\n",
            "Epoch 3541/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0071 - accuracy: 0.6123\n",
            "Epoch 3542/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0070 - accuracy: 0.6115\n",
            "Epoch 3543/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0070 - accuracy: 0.6012\n",
            "Epoch 3544/5000\n",
            "1/1 [==============================] - 0s 294ms/step - loss: 0.0069 - accuracy: 0.6036\n",
            "Epoch 3545/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0069 - accuracy: 0.6101\n",
            "Epoch 3546/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0069 - accuracy: 0.6159\n",
            "Epoch 3547/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0069 - accuracy: 0.6239\n",
            "Epoch 3548/5000\n",
            "1/1 [==============================] - 0s 295ms/step - loss: 0.0070 - accuracy: 0.6204\n",
            "Epoch 3549/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0070 - accuracy: 0.6206\n",
            "Epoch 3550/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0070 - accuracy: 0.6017\n",
            "Epoch 3551/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0070 - accuracy: 0.6095\n",
            "Epoch 3552/5000\n",
            "1/1 [==============================] - 0s 292ms/step - loss: 0.0070 - accuracy: 0.5981\n",
            "Epoch 3553/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0070 - accuracy: 0.6179\n",
            "Epoch 3554/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0069 - accuracy: 0.6137\n",
            "Epoch 3555/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0069 - accuracy: 0.6218\n",
            "Epoch 3556/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0069 - accuracy: 0.6212\n",
            "Epoch 3557/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0069 - accuracy: 0.6135\n",
            "Epoch 3558/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0069 - accuracy: 0.6187\n",
            "Epoch 3559/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0070 - accuracy: 0.6058\n",
            "Epoch 3560/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0070 - accuracy: 0.6176\n",
            "Epoch 3561/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0070 - accuracy: 0.6039\n",
            "Epoch 3562/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0070 - accuracy: 0.6157\n",
            "Epoch 3563/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0069 - accuracy: 0.6083\n",
            "Epoch 3564/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0069 - accuracy: 0.6153\n",
            "Epoch 3565/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0069 - accuracy: 0.6170\n",
            "Epoch 3566/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0069 - accuracy: 0.6139\n",
            "Epoch 3567/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0069 - accuracy: 0.6216\n",
            "Epoch 3568/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0070 - accuracy: 0.6115\n",
            "Epoch 3569/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0070 - accuracy: 0.6212\n",
            "Epoch 3570/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0069 - accuracy: 0.6069\n",
            "Epoch 3571/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0069 - accuracy: 0.6119\n",
            "Epoch 3572/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0069 - accuracy: 0.6061\n",
            "Epoch 3573/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0069 - accuracy: 0.6102\n",
            "Epoch 3574/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0069 - accuracy: 0.6151\n",
            "Epoch 3575/5000\n",
            "1/1 [==============================] - 0s 260ms/step - loss: 0.0069 - accuracy: 0.6166\n",
            "Epoch 3576/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0069 - accuracy: 0.6235\n",
            "Epoch 3577/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0069 - accuracy: 0.6152\n",
            "Epoch 3578/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0069 - accuracy: 0.6175\n",
            "Epoch 3579/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0069 - accuracy: 0.6041\n",
            "Epoch 3580/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0069 - accuracy: 0.6086\n",
            "Epoch 3581/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0069 - accuracy: 0.6034\n",
            "Epoch 3582/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0069 - accuracy: 0.6150\n",
            "Epoch 3583/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0069 - accuracy: 0.6151\n",
            "Epoch 3584/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0069 - accuracy: 0.6236\n",
            "Epoch 3585/5000\n",
            "1/1 [==============================] - 0s 252ms/step - loss: 0.0069 - accuracy: 0.6166\n",
            "Epoch 3586/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0070 - accuracy: 0.6203\n",
            "Epoch 3587/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0070 - accuracy: 0.6033\n",
            "Epoch 3588/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0070 - accuracy: 0.6127\n",
            "Epoch 3589/5000\n",
            "1/1 [==============================] - 0s 296ms/step - loss: 0.0070 - accuracy: 0.5987\n",
            "Epoch 3590/5000\n",
            "1/1 [==============================] - 0s 291ms/step - loss: 0.0070 - accuracy: 0.6168\n",
            "Epoch 3591/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0070 - accuracy: 0.6087\n",
            "Epoch 3592/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0070 - accuracy: 0.6227\n",
            "Epoch 3593/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0069 - accuracy: 0.6192\n",
            "Epoch 3594/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0069 - accuracy: 0.6168\n",
            "Epoch 3595/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0069 - accuracy: 0.6142\n",
            "Epoch 3596/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0069 - accuracy: 0.6057\n",
            "Epoch 3597/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0069 - accuracy: 0.6166\n",
            "Epoch 3598/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0069 - accuracy: 0.6072\n",
            "Epoch 3599/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0070 - accuracy: 0.6235\n",
            "Epoch 3600/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0069 - accuracy: 0.6115\n",
            "Epoch 3601/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0070 - accuracy: 0.6142\n",
            "Epoch 3602/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0070 - accuracy: 0.6066\n",
            "Epoch 3603/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0070 - accuracy: 0.6054\n",
            "Epoch 3604/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0070 - accuracy: 0.6133\n",
            "Epoch 3605/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0070 - accuracy: 0.6140\n",
            "Epoch 3606/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0070 - accuracy: 0.6210\n",
            "Epoch 3607/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0069 - accuracy: 0.6161\n",
            "Epoch 3608/5000\n",
            "1/1 [==============================] - 0s 257ms/step - loss: 0.0069 - accuracy: 0.6103\n",
            "Epoch 3609/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0069 - accuracy: 0.6072\n",
            "Epoch 3610/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0069 - accuracy: 0.6010\n",
            "Epoch 3611/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0069 - accuracy: 0.6125\n",
            "Epoch 3612/5000\n",
            "1/1 [==============================] - 0s 261ms/step - loss: 0.0069 - accuracy: 0.6116\n",
            "Epoch 3613/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0069 - accuracy: 0.6224\n",
            "Epoch 3614/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0069 - accuracy: 0.6175\n",
            "Epoch 3615/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0069 - accuracy: 0.6177\n",
            "Epoch 3616/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0069 - accuracy: 0.6053\n",
            "Epoch 3617/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0069 - accuracy: 0.6087\n",
            "Epoch 3618/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0069 - accuracy: 0.6013\n",
            "Epoch 3619/5000\n",
            "1/1 [==============================] - 0s 255ms/step - loss: 0.0069 - accuracy: 0.6153\n",
            "Epoch 3620/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0070 - accuracy: 0.6084\n",
            "Epoch 3621/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0070 - accuracy: 0.6197\n",
            "Epoch 3622/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0070 - accuracy: 0.6089\n",
            "Epoch 3623/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0071 - accuracy: 0.6204\n",
            "Epoch 3624/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0071 - accuracy: 0.6052\n",
            "Epoch 3625/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0072 - accuracy: 0.6217\n",
            "Epoch 3626/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0072 - accuracy: 0.6000\n",
            "Epoch 3627/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0071 - accuracy: 0.6194\n",
            "Epoch 3628/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0070 - accuracy: 0.6022\n",
            "Epoch 3629/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0069 - accuracy: 0.6153\n",
            "Epoch 3630/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0069 - accuracy: 0.6188\n",
            "Epoch 3631/5000\n",
            "1/1 [==============================] - 0s 260ms/step - loss: 0.0069 - accuracy: 0.6117\n",
            "Epoch 3632/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0069 - accuracy: 0.6260\n",
            "Epoch 3633/5000\n",
            "1/1 [==============================] - 0s 253ms/step - loss: 0.0070 - accuracy: 0.6072\n",
            "Epoch 3634/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0070 - accuracy: 0.6180\n",
            "Epoch 3635/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0070 - accuracy: 0.6069\n",
            "Epoch 3636/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0070 - accuracy: 0.6086\n",
            "Epoch 3637/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0070 - accuracy: 0.6063\n",
            "Epoch 3638/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0070 - accuracy: 0.6091\n",
            "Epoch 3639/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0069 - accuracy: 0.6169\n",
            "Epoch 3640/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0069 - accuracy: 0.6133\n",
            "Epoch 3641/5000\n",
            "1/1 [==============================] - 0s 261ms/step - loss: 0.0069 - accuracy: 0.6198\n",
            "Epoch 3642/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0069 - accuracy: 0.6088\n",
            "Epoch 3643/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0069 - accuracy: 0.6120\n",
            "Epoch 3644/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0069 - accuracy: 0.6034\n",
            "Epoch 3645/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0070 - accuracy: 0.6116\n",
            "Epoch 3646/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0070 - accuracy: 0.6103\n",
            "Epoch 3647/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0070 - accuracy: 0.6206\n",
            "Epoch 3648/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0069 - accuracy: 0.6115\n",
            "Epoch 3649/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0069 - accuracy: 0.6208\n",
            "Epoch 3650/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0069 - accuracy: 0.6047\n",
            "Epoch 3651/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0069 - accuracy: 0.6130\n",
            "Epoch 3652/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0069 - accuracy: 0.6047\n",
            "Epoch 3653/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0069 - accuracy: 0.6105\n",
            "Epoch 3654/5000\n",
            "1/1 [==============================] - 0s 261ms/step - loss: 0.0069 - accuracy: 0.6163\n",
            "Epoch 3655/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0069 - accuracy: 0.6138\n",
            "Epoch 3656/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0069 - accuracy: 0.6214\n",
            "Epoch 3657/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0069 - accuracy: 0.6094\n",
            "Epoch 3658/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0069 - accuracy: 0.6180\n",
            "Epoch 3659/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0069 - accuracy: 0.6048\n",
            "Epoch 3660/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0069 - accuracy: 0.6142\n",
            "Epoch 3661/5000\n",
            "1/1 [==============================] - 0s 297ms/step - loss: 0.0069 - accuracy: 0.6022\n",
            "Epoch 3662/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0069 - accuracy: 0.6141\n",
            "Epoch 3663/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0069 - accuracy: 0.6105\n",
            "Epoch 3664/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0069 - accuracy: 0.6147\n",
            "Epoch 3665/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0069 - accuracy: 0.6149\n",
            "Epoch 3666/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0069 - accuracy: 0.6088\n",
            "Epoch 3667/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0069 - accuracy: 0.6210\n",
            "Epoch 3668/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0069 - accuracy: 0.6064\n",
            "Epoch 3669/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0069 - accuracy: 0.6217\n",
            "Epoch 3670/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0069 - accuracy: 0.6046\n",
            "Epoch 3671/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0069 - accuracy: 0.6092\n",
            "Epoch 3672/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0069 - accuracy: 0.6059\n",
            "Epoch 3673/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0070 - accuracy: 0.6061\n",
            "Epoch 3674/5000\n",
            "1/1 [==============================] - 0s 253ms/step - loss: 0.0070 - accuracy: 0.6158\n",
            "Epoch 3675/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0069 - accuracy: 0.6144\n",
            "Epoch 3676/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0069 - accuracy: 0.6171\n",
            "Epoch 3677/5000\n",
            "1/1 [==============================] - 0s 251ms/step - loss: 0.0069 - accuracy: 0.6171\n",
            "Epoch 3678/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0069 - accuracy: 0.6058\n",
            "Epoch 3679/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0069 - accuracy: 0.6125\n",
            "Epoch 3680/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0069 - accuracy: 0.5970\n",
            "Epoch 3681/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0069 - accuracy: 0.6151\n",
            "Epoch 3682/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0069 - accuracy: 0.6085\n",
            "Epoch 3683/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0069 - accuracy: 0.6214\n",
            "Epoch 3684/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0069 - accuracy: 0.6181\n",
            "Epoch 3685/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0069 - accuracy: 0.6131\n",
            "Epoch 3686/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0069 - accuracy: 0.6110\n",
            "Epoch 3687/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0069 - accuracy: 0.6018\n",
            "Epoch 3688/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0069 - accuracy: 0.6088\n",
            "Epoch 3689/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0069 - accuracy: 0.6079\n",
            "Epoch 3690/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0069 - accuracy: 0.6174\n",
            "Epoch 3691/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0069 - accuracy: 0.6168\n",
            "Epoch 3692/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0069 - accuracy: 0.6138\n",
            "Epoch 3693/5000\n",
            "1/1 [==============================] - 0s 297ms/step - loss: 0.0069 - accuracy: 0.6115\n",
            "Epoch 3694/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0069 - accuracy: 0.6058\n",
            "Epoch 3695/5000\n",
            "1/1 [==============================] - 0s 290ms/step - loss: 0.0069 - accuracy: 0.6104\n",
            "Epoch 3696/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0069 - accuracy: 0.6100\n",
            "Epoch 3697/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0069 - accuracy: 0.6157\n",
            "Epoch 3698/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0069 - accuracy: 0.6153\n",
            "Epoch 3699/5000\n",
            "1/1 [==============================] - 0s 293ms/step - loss: 0.0069 - accuracy: 0.6114\n",
            "Epoch 3700/5000\n",
            "1/1 [==============================] - 0s 304ms/step - loss: 0.0069 - accuracy: 0.6150\n",
            "Epoch 3701/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0069 - accuracy: 0.6012\n",
            "Epoch 3702/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0070 - accuracy: 0.6170\n",
            "Epoch 3703/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0071 - accuracy: 0.5961\n",
            "Epoch 3704/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0071 - accuracy: 0.6221\n",
            "Epoch 3705/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0071 - accuracy: 0.6067\n",
            "Epoch 3706/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0071 - accuracy: 0.6202\n",
            "Epoch 3707/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0072 - accuracy: 0.6101\n",
            "Epoch 3708/5000\n",
            "1/1 [==============================] - 0s 299ms/step - loss: 0.0072 - accuracy: 0.6127\n",
            "Epoch 3709/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0071 - accuracy: 0.6049\n",
            "Epoch 3710/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0070 - accuracy: 0.6107\n",
            "Epoch 3711/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0069 - accuracy: 0.6014\n",
            "Epoch 3712/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0069 - accuracy: 0.6198\n",
            "Epoch 3713/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0069 - accuracy: 0.6019\n",
            "Epoch 3714/5000\n",
            "1/1 [==============================] - 0s 306ms/step - loss: 0.0069 - accuracy: 0.6122\n",
            "Epoch 3715/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0069 - accuracy: 0.6093\n",
            "Epoch 3716/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0069 - accuracy: 0.6004\n",
            "Epoch 3717/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0069 - accuracy: 0.6211\n",
            "Epoch 3718/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0069 - accuracy: 0.6042\n",
            "Epoch 3719/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0069 - accuracy: 0.6201\n",
            "Epoch 3720/5000\n",
            "1/1 [==============================] - 0s 301ms/step - loss: 0.0069 - accuracy: 0.6110\n",
            "Epoch 3721/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0069 - accuracy: 0.6039\n",
            "Epoch 3722/5000\n",
            "1/1 [==============================] - 0s 305ms/step - loss: 0.0069 - accuracy: 0.6065\n",
            "Epoch 3723/5000\n",
            "1/1 [==============================] - 0s 298ms/step - loss: 0.0069 - accuracy: 0.5983\n",
            "Epoch 3724/5000\n",
            "1/1 [==============================] - 0s 293ms/step - loss: 0.0069 - accuracy: 0.6105\n",
            "Epoch 3725/5000\n",
            "1/1 [==============================] - 0s 301ms/step - loss: 0.0069 - accuracy: 0.6174\n",
            "Epoch 3726/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0068 - accuracy: 0.6196\n",
            "Epoch 3727/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0069 - accuracy: 0.6246\n",
            "Epoch 3728/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0069 - accuracy: 0.6131\n",
            "Epoch 3729/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0069 - accuracy: 0.6097\n",
            "Epoch 3730/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0069 - accuracy: 0.6023\n",
            "Epoch 3731/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0068 - accuracy: 0.6015\n",
            "Epoch 3732/5000\n",
            "1/1 [==============================] - 0s 290ms/step - loss: 0.0068 - accuracy: 0.6056\n",
            "Epoch 3733/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0069 - accuracy: 0.6129\n",
            "Epoch 3734/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0069 - accuracy: 0.6173\n",
            "Epoch 3735/5000\n",
            "1/1 [==============================] - 0s 293ms/step - loss: 0.0069 - accuracy: 0.6160\n",
            "Epoch 3736/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0069 - accuracy: 0.6153\n",
            "Epoch 3737/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0068 - accuracy: 0.6055\n",
            "Epoch 3738/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0069 - accuracy: 0.6065\n",
            "Epoch 3739/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0068 - accuracy: 0.6048\n",
            "Epoch 3740/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0068 - accuracy: 0.6101\n",
            "Epoch 3741/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0068 - accuracy: 0.6164\n",
            "Epoch 3742/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0068 - accuracy: 0.6183\n",
            "Epoch 3743/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0068 - accuracy: 0.6181\n",
            "Epoch 3744/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0068 - accuracy: 0.6113\n",
            "Epoch 3745/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0068 - accuracy: 0.6077\n",
            "Epoch 3746/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0068 - accuracy: 0.6064\n",
            "Epoch 3747/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0068 - accuracy: 0.6057\n",
            "Epoch 3748/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0068 - accuracy: 0.6133\n",
            "Epoch 3749/5000\n",
            "1/1 [==============================] - 0s 292ms/step - loss: 0.0069 - accuracy: 0.6108\n",
            "Epoch 3750/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0069 - accuracy: 0.6230\n",
            "Epoch 3751/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0070 - accuracy: 0.6038\n",
            "Epoch 3752/5000\n",
            "1/1 [==============================] - 0s 300ms/step - loss: 0.0071 - accuracy: 0.6237\n",
            "Epoch 3753/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0071 - accuracy: 0.5967\n",
            "Epoch 3754/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0071 - accuracy: 0.6111\n",
            "Epoch 3755/5000\n",
            "1/1 [==============================] - 0s 297ms/step - loss: 0.0072 - accuracy: 0.6036\n",
            "Epoch 3756/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0073 - accuracy: 0.6146\n",
            "Epoch 3757/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0071 - accuracy: 0.6121\n",
            "Epoch 3758/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0070 - accuracy: 0.6208\n",
            "Epoch 3759/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0070 - accuracy: 0.6042\n",
            "Epoch 3760/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0070 - accuracy: 0.6209\n",
            "Epoch 3761/5000\n",
            "1/1 [==============================] - 0s 261ms/step - loss: 0.0069 - accuracy: 0.5972\n",
            "Epoch 3762/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0069 - accuracy: 0.6025\n",
            "Epoch 3763/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0069 - accuracy: 0.6075\n",
            "Epoch 3764/5000\n",
            "1/1 [==============================] - 0s 294ms/step - loss: 0.0070 - accuracy: 0.5948\n",
            "Epoch 3765/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0070 - accuracy: 0.6259\n",
            "Epoch 3766/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0069 - accuracy: 0.6129\n",
            "Epoch 3767/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0069 - accuracy: 0.6177\n",
            "Epoch 3768/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0069 - accuracy: 0.6142\n",
            "Epoch 3769/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0069 - accuracy: 0.5993\n",
            "Epoch 3770/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0069 - accuracy: 0.6055\n",
            "Epoch 3771/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0068 - accuracy: 0.6032\n",
            "Epoch 3772/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0069 - accuracy: 0.6107\n",
            "Epoch 3773/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0069 - accuracy: 0.6234\n",
            "Epoch 3774/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0069 - accuracy: 0.6183\n",
            "Epoch 3775/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0069 - accuracy: 0.6205\n",
            "Epoch 3776/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0069 - accuracy: 0.6135\n",
            "Epoch 3777/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0069 - accuracy: 0.5969\n",
            "Epoch 3778/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0069 - accuracy: 0.6090\n",
            "Epoch 3779/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0069 - accuracy: 0.5947\n",
            "Epoch 3780/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0069 - accuracy: 0.6143\n",
            "Epoch 3781/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0069 - accuracy: 0.6146\n",
            "Epoch 3782/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0069 - accuracy: 0.6161\n",
            "Epoch 3783/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0069 - accuracy: 0.6172\n",
            "Epoch 3784/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0069 - accuracy: 0.6064\n",
            "Epoch 3785/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0068 - accuracy: 0.6055\n",
            "Epoch 3786/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0068 - accuracy: 0.6096\n",
            "Epoch 3787/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0068 - accuracy: 0.6105\n",
            "Epoch 3788/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0068 - accuracy: 0.6226\n",
            "Epoch 3789/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0068 - accuracy: 0.6103\n",
            "Epoch 3790/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0068 - accuracy: 0.6109\n",
            "Epoch 3791/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0068 - accuracy: 0.6043\n",
            "Epoch 3792/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0068 - accuracy: 0.6053\n",
            "Epoch 3793/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0068 - accuracy: 0.6083\n",
            "Epoch 3794/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0068 - accuracy: 0.6172\n",
            "Epoch 3795/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0069 - accuracy: 0.6146\n",
            "Epoch 3796/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0069 - accuracy: 0.6201\n",
            "Epoch 3797/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0069 - accuracy: 0.5985\n",
            "Epoch 3798/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0070 - accuracy: 0.6137\n",
            "Epoch 3799/5000\n",
            "1/1 [==============================] - 0s 293ms/step - loss: 0.0071 - accuracy: 0.5933\n",
            "Epoch 3800/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0071 - accuracy: 0.6192\n",
            "Epoch 3801/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0071 - accuracy: 0.6086\n",
            "Epoch 3802/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0071 - accuracy: 0.6247\n",
            "Epoch 3803/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0070 - accuracy: 0.6152\n",
            "Epoch 3804/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0069 - accuracy: 0.6184\n",
            "Epoch 3805/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0069 - accuracy: 0.6039\n",
            "Epoch 3806/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0068 - accuracy: 0.6098\n",
            "Epoch 3807/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0068 - accuracy: 0.6028\n",
            "Epoch 3808/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0068 - accuracy: 0.6125\n",
            "Epoch 3809/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0069 - accuracy: 0.6169\n",
            "Epoch 3810/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0069 - accuracy: 0.6100\n",
            "Epoch 3811/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0070 - accuracy: 0.6231\n",
            "Epoch 3812/5000\n",
            "1/1 [==============================] - 0s 291ms/step - loss: 0.0070 - accuracy: 0.5980\n",
            "Epoch 3813/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0069 - accuracy: 0.6136\n",
            "Epoch 3814/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0069 - accuracy: 0.6008\n",
            "Epoch 3815/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0069 - accuracy: 0.6058\n",
            "Epoch 3816/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0069 - accuracy: 0.6121\n",
            "Epoch 3817/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0068 - accuracy: 0.6058\n",
            "Epoch 3818/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0068 - accuracy: 0.6192\n",
            "Epoch 3819/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0068 - accuracy: 0.6136\n",
            "Epoch 3820/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0069 - accuracy: 0.6186\n",
            "Epoch 3821/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0069 - accuracy: 0.6155\n",
            "Epoch 3822/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0069 - accuracy: 0.6117\n",
            "Epoch 3823/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0069 - accuracy: 0.6049\n",
            "Epoch 3824/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0069 - accuracy: 0.6082\n",
            "Epoch 3825/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0069 - accuracy: 0.5995\n",
            "Epoch 3826/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0069 - accuracy: 0.6173\n",
            "Epoch 3827/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0068 - accuracy: 0.6064\n",
            "Epoch 3828/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0068 - accuracy: 0.6178\n",
            "Epoch 3829/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0068 - accuracy: 0.6121\n",
            "Epoch 3830/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0068 - accuracy: 0.6082\n",
            "Epoch 3831/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0068 - accuracy: 0.6141\n",
            "Epoch 3832/5000\n",
            "1/1 [==============================] - 0s 290ms/step - loss: 0.0068 - accuracy: 0.6056\n",
            "Epoch 3833/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0068 - accuracy: 0.6145\n",
            "Epoch 3834/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0068 - accuracy: 0.6097\n",
            "Epoch 3835/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0068 - accuracy: 0.6130\n",
            "Epoch 3836/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0068 - accuracy: 0.6059\n",
            "Epoch 3837/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0069 - accuracy: 0.6111\n",
            "Epoch 3838/5000\n",
            "1/1 [==============================] - 0s 254ms/step - loss: 0.0069 - accuracy: 0.6042\n",
            "Epoch 3839/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0069 - accuracy: 0.6183\n",
            "Epoch 3840/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0069 - accuracy: 0.6052\n",
            "Epoch 3841/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0070 - accuracy: 0.6217\n",
            "Epoch 3842/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0070 - accuracy: 0.6011\n",
            "Epoch 3843/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0069 - accuracy: 0.6159\n",
            "Epoch 3844/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0069 - accuracy: 0.6043\n",
            "Epoch 3845/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0069 - accuracy: 0.6163\n",
            "Epoch 3846/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0068 - accuracy: 0.6141\n",
            "Epoch 3847/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0068 - accuracy: 0.6153\n",
            "Epoch 3848/5000\n",
            "1/1 [==============================] - 0s 291ms/step - loss: 0.0068 - accuracy: 0.6138\n",
            "Epoch 3849/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0068 - accuracy: 0.6120\n",
            "Epoch 3850/5000\n",
            "1/1 [==============================] - 0s 295ms/step - loss: 0.0068 - accuracy: 0.6143\n",
            "Epoch 3851/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0068 - accuracy: 0.6110\n",
            "Epoch 3852/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0068 - accuracy: 0.6145\n",
            "Epoch 3853/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0069 - accuracy: 0.6055\n",
            "Epoch 3854/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0069 - accuracy: 0.6128\n",
            "Epoch 3855/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0069 - accuracy: 0.5975\n",
            "Epoch 3856/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0069 - accuracy: 0.6154\n",
            "Epoch 3857/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0069 - accuracy: 0.6019\n",
            "Epoch 3858/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0069 - accuracy: 0.6194\n",
            "Epoch 3859/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0069 - accuracy: 0.6101\n",
            "Epoch 3860/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0068 - accuracy: 0.6158\n",
            "Epoch 3861/5000\n",
            "1/1 [==============================] - 0s 293ms/step - loss: 0.0068 - accuracy: 0.6134\n",
            "Epoch 3862/5000\n",
            "1/1 [==============================] - 0s 296ms/step - loss: 0.0068 - accuracy: 0.6120\n",
            "Epoch 3863/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0068 - accuracy: 0.6146\n",
            "Epoch 3864/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0068 - accuracy: 0.6108\n",
            "Epoch 3865/5000\n",
            "1/1 [==============================] - 0s 292ms/step - loss: 0.0068 - accuracy: 0.6118\n",
            "Epoch 3866/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0068 - accuracy: 0.6090\n",
            "Epoch 3867/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0068 - accuracy: 0.6117\n",
            "Epoch 3868/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0068 - accuracy: 0.6081\n",
            "Epoch 3869/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0068 - accuracy: 0.6163\n",
            "Epoch 3870/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0069 - accuracy: 0.6051\n",
            "Epoch 3871/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0069 - accuracy: 0.6216\n",
            "Epoch 3872/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0069 - accuracy: 0.6023\n",
            "Epoch 3873/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0069 - accuracy: 0.6177\n",
            "Epoch 3874/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0069 - accuracy: 0.6002\n",
            "Epoch 3875/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0069 - accuracy: 0.6128\n",
            "Epoch 3876/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0069 - accuracy: 0.6098\n",
            "Epoch 3877/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0069 - accuracy: 0.6157\n",
            "Epoch 3878/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0068 - accuracy: 0.6150\n",
            "Epoch 3879/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0068 - accuracy: 0.6178\n",
            "Epoch 3880/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0068 - accuracy: 0.6123\n",
            "Epoch 3881/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0068 - accuracy: 0.6149\n",
            "Epoch 3882/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0068 - accuracy: 0.6061\n",
            "Epoch 3883/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0068 - accuracy: 0.6086\n",
            "Epoch 3884/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0068 - accuracy: 0.6081\n",
            "Epoch 3885/5000\n",
            "1/1 [==============================] - 0s 256ms/step - loss: 0.0068 - accuracy: 0.6116\n",
            "Epoch 3886/5000\n",
            "1/1 [==============================] - 0s 301ms/step - loss: 0.0068 - accuracy: 0.6148\n",
            "Epoch 3887/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0068 - accuracy: 0.6147\n",
            "Epoch 3888/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0068 - accuracy: 0.6150\n",
            "Epoch 3889/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0068 - accuracy: 0.6100\n",
            "Epoch 3890/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0068 - accuracy: 0.6107\n",
            "Epoch 3891/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0068 - accuracy: 0.6019\n",
            "Epoch 3892/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0068 - accuracy: 0.6136\n",
            "Epoch 3893/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0068 - accuracy: 0.6032\n",
            "Epoch 3894/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0069 - accuracy: 0.6209\n",
            "Epoch 3895/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0069 - accuracy: 0.6071\n",
            "Epoch 3896/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0069 - accuracy: 0.6186\n",
            "Epoch 3897/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0069 - accuracy: 0.6073\n",
            "Epoch 3898/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0070 - accuracy: 0.6126\n",
            "Epoch 3899/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0070 - accuracy: 0.6068\n",
            "Epoch 3900/5000\n",
            "1/1 [==============================] - 0s 291ms/step - loss: 0.0070 - accuracy: 0.6184\n",
            "Epoch 3901/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0069 - accuracy: 0.6060\n",
            "Epoch 3902/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0069 - accuracy: 0.6234\n",
            "Epoch 3903/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0069 - accuracy: 0.5972\n",
            "Epoch 3904/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0069 - accuracy: 0.6130\n",
            "Epoch 3905/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0068 - accuracy: 0.6009\n",
            "Epoch 3906/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0068 - accuracy: 0.6106\n",
            "Epoch 3907/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0068 - accuracy: 0.6210\n",
            "Epoch 3908/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0068 - accuracy: 0.6138\n",
            "Epoch 3909/5000\n",
            "1/1 [==============================] - 0s 261ms/step - loss: 0.0069 - accuracy: 0.6260\n",
            "Epoch 3910/5000\n",
            "1/1 [==============================] - 0s 261ms/step - loss: 0.0069 - accuracy: 0.6083\n",
            "Epoch 3911/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0068 - accuracy: 0.6081\n",
            "Epoch 3912/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0068 - accuracy: 0.5967\n",
            "Epoch 3913/5000\n",
            "1/1 [==============================] - 0s 261ms/step - loss: 0.0069 - accuracy: 0.6008\n",
            "Epoch 3914/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0068 - accuracy: 0.6052\n",
            "Epoch 3915/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0068 - accuracy: 0.6154\n",
            "Epoch 3916/5000\n",
            "1/1 [==============================] - 0s 257ms/step - loss: 0.0068 - accuracy: 0.6216\n",
            "Epoch 3917/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0068 - accuracy: 0.6237\n",
            "Epoch 3918/5000\n",
            "1/1 [==============================] - 0s 296ms/step - loss: 0.0068 - accuracy: 0.6121\n",
            "Epoch 3919/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0068 - accuracy: 0.6114\n",
            "Epoch 3920/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0068 - accuracy: 0.5961\n",
            "Epoch 3921/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0068 - accuracy: 0.6074\n",
            "Epoch 3922/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0068 - accuracy: 0.6056\n",
            "Epoch 3923/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0068 - accuracy: 0.6186\n",
            "Epoch 3924/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0068 - accuracy: 0.6207\n",
            "Epoch 3925/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0068 - accuracy: 0.6153\n",
            "Epoch 3926/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0068 - accuracy: 0.6151\n",
            "Epoch 3927/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0068 - accuracy: 0.6047\n",
            "Epoch 3928/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0068 - accuracy: 0.6060\n",
            "Epoch 3929/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0068 - accuracy: 0.6065\n",
            "Epoch 3930/5000\n",
            "1/1 [==============================] - 0s 258ms/step - loss: 0.0068 - accuracy: 0.6089\n",
            "Epoch 3931/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0068 - accuracy: 0.6156\n",
            "Epoch 3932/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0068 - accuracy: 0.6100\n",
            "Epoch 3933/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0068 - accuracy: 0.6128\n",
            "Epoch 3934/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0068 - accuracy: 0.6084\n",
            "Epoch 3935/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0068 - accuracy: 0.6103\n",
            "Epoch 3936/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0068 - accuracy: 0.6119\n",
            "Epoch 3937/5000\n",
            "1/1 [==============================] - 0s 260ms/step - loss: 0.0068 - accuracy: 0.6129\n",
            "Epoch 3938/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0068 - accuracy: 0.6132\n",
            "Epoch 3939/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0068 - accuracy: 0.6130\n",
            "Epoch 3940/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0068 - accuracy: 0.6108\n",
            "Epoch 3941/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0068 - accuracy: 0.6122\n",
            "Epoch 3942/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0068 - accuracy: 0.6061\n",
            "Epoch 3943/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0069 - accuracy: 0.6162\n",
            "Epoch 3944/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0070 - accuracy: 0.5989\n",
            "Epoch 3945/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0072 - accuracy: 0.6244\n",
            "Epoch 3946/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0073 - accuracy: 0.5959\n",
            "Epoch 3947/5000\n",
            "1/1 [==============================] - 0s 290ms/step - loss: 0.0073 - accuracy: 0.6262\n",
            "Epoch 3948/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0073 - accuracy: 0.6017\n",
            "Epoch 3949/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0072 - accuracy: 0.6169\n",
            "Epoch 3950/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0069 - accuracy: 0.6104\n",
            "Epoch 3951/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0068 - accuracy: 0.6078\n",
            "Epoch 3952/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0068 - accuracy: 0.6131\n",
            "Epoch 3953/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0069 - accuracy: 0.6064\n",
            "Epoch 3954/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0070 - accuracy: 0.6198\n",
            "Epoch 3955/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0071 - accuracy: 0.6080\n",
            "Epoch 3956/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0070 - accuracy: 0.6235\n",
            "Epoch 3957/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0069 - accuracy: 0.5945\n",
            "Epoch 3958/5000\n",
            "1/1 [==============================] - 0s 290ms/step - loss: 0.0068 - accuracy: 0.6093\n",
            "Epoch 3959/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0068 - accuracy: 0.5981\n",
            "Epoch 3960/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0068 - accuracy: 0.5982\n",
            "Epoch 3961/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0069 - accuracy: 0.6213\n",
            "Epoch 3962/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0069 - accuracy: 0.6109\n",
            "Epoch 3963/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0069 - accuracy: 0.6306\n",
            "Epoch 3964/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0068 - accuracy: 0.6121\n",
            "Epoch 3965/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0068 - accuracy: 0.6084\n",
            "Epoch 3966/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0068 - accuracy: 0.6068\n",
            "Epoch 3967/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0068 - accuracy: 0.5989\n",
            "Epoch 3968/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0068 - accuracy: 0.6169\n",
            "Epoch 3969/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0068 - accuracy: 0.6143\n",
            "Epoch 3970/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0068 - accuracy: 0.6236\n",
            "Epoch 3971/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0068 - accuracy: 0.6164\n",
            "Epoch 3972/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0068 - accuracy: 0.6092\n",
            "Epoch 3973/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0068 - accuracy: 0.5995\n",
            "Epoch 3974/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0068 - accuracy: 0.5999\n",
            "Epoch 3975/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0068 - accuracy: 0.6038\n",
            "Epoch 3976/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0068 - accuracy: 0.6143\n",
            "Epoch 3977/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0068 - accuracy: 0.6188\n",
            "Epoch 3978/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0068 - accuracy: 0.6195\n",
            "Epoch 3979/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0068 - accuracy: 0.6117\n",
            "Epoch 3980/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0068 - accuracy: 0.6062\n",
            "Epoch 3981/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0068 - accuracy: 0.6035\n",
            "Epoch 3982/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0067 - accuracy: 0.6087\n",
            "Epoch 3983/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0067 - accuracy: 0.6116\n",
            "Epoch 3984/5000\n",
            "1/1 [==============================] - 0s 299ms/step - loss: 0.0068 - accuracy: 0.6201\n",
            "Epoch 3985/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0068 - accuracy: 0.6153\n",
            "Epoch 3986/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0068 - accuracy: 0.6193\n",
            "Epoch 3987/5000\n",
            "1/1 [==============================] - 0s 301ms/step - loss: 0.0069 - accuracy: 0.5988\n",
            "Epoch 3988/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0069 - accuracy: 0.6114\n",
            "Epoch 3989/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0069 - accuracy: 0.5914\n",
            "Epoch 3990/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0069 - accuracy: 0.6160\n",
            "Epoch 3991/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0068 - accuracy: 0.6103\n",
            "Epoch 3992/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0068 - accuracy: 0.6175\n",
            "Epoch 3993/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0068 - accuracy: 0.6199\n",
            "Epoch 3994/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0068 - accuracy: 0.6098\n",
            "Epoch 3995/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0068 - accuracy: 0.6199\n",
            "Epoch 3996/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0068 - accuracy: 0.6075\n",
            "Epoch 3997/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0068 - accuracy: 0.6158\n",
            "Epoch 3998/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0069 - accuracy: 0.6071\n",
            "Epoch 3999/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0069 - accuracy: 0.6090\n",
            "Epoch 4000/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0069 - accuracy: 0.6044\n",
            "Epoch 4001/5000\n",
            "1/1 [==============================] - 0s 293ms/step - loss: 0.0068 - accuracy: 0.6111\n",
            "Epoch 4002/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0068 - accuracy: 0.6057\n",
            "Epoch 4003/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0068 - accuracy: 0.6204\n",
            "Epoch 4004/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0068 - accuracy: 0.6076\n",
            "Epoch 4005/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0068 - accuracy: 0.6178\n",
            "Epoch 4006/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0068 - accuracy: 0.6063\n",
            "Epoch 4007/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0067 - accuracy: 0.6058\n",
            "Epoch 4008/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0067 - accuracy: 0.6099\n",
            "Epoch 4009/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0068 - accuracy: 0.6059\n",
            "Epoch 4010/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0068 - accuracy: 0.6208\n",
            "Epoch 4011/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0068 - accuracy: 0.6120\n",
            "Epoch 4012/5000\n",
            "1/1 [==============================] - 0s 296ms/step - loss: 0.0068 - accuracy: 0.6197\n",
            "Epoch 4013/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0068 - accuracy: 0.6095\n",
            "Epoch 4014/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0068 - accuracy: 0.6074\n",
            "Epoch 4015/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0068 - accuracy: 0.6055\n",
            "Epoch 4016/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0068 - accuracy: 0.6057\n",
            "Epoch 4017/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0068 - accuracy: 0.6082\n",
            "Epoch 4018/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0068 - accuracy: 0.6173\n",
            "Epoch 4019/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0068 - accuracy: 0.6102\n",
            "Epoch 4020/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0068 - accuracy: 0.6196\n",
            "Epoch 4021/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0068 - accuracy: 0.6023\n",
            "Epoch 4022/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0068 - accuracy: 0.6135\n",
            "Epoch 4023/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0068 - accuracy: 0.6017\n",
            "Epoch 4024/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0068 - accuracy: 0.6146\n",
            "Epoch 4025/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0068 - accuracy: 0.6114\n",
            "Epoch 4026/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0068 - accuracy: 0.6174\n",
            "Epoch 4027/5000\n",
            "1/1 [==============================] - 0s 256ms/step - loss: 0.0068 - accuracy: 0.6151\n",
            "Epoch 4028/5000\n",
            "1/1 [==============================] - 0s 257ms/step - loss: 0.0068 - accuracy: 0.6105\n",
            "Epoch 4029/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0068 - accuracy: 0.6072\n",
            "Epoch 4030/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0067 - accuracy: 0.6057\n",
            "Epoch 4031/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0067 - accuracy: 0.6067\n",
            "Epoch 4032/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0067 - accuracy: 0.6150\n",
            "Epoch 4033/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0068 - accuracy: 0.6113\n",
            "Epoch 4034/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0068 - accuracy: 0.6177\n",
            "Epoch 4035/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0068 - accuracy: 0.6069\n",
            "Epoch 4036/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0068 - accuracy: 0.6133\n",
            "Epoch 4037/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0068 - accuracy: 0.6051\n",
            "Epoch 4038/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0068 - accuracy: 0.6156\n",
            "Epoch 4039/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0068 - accuracy: 0.6061\n",
            "Epoch 4040/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0068 - accuracy: 0.6144\n",
            "Epoch 4041/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0068 - accuracy: 0.6060\n",
            "Epoch 4042/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0069 - accuracy: 0.6167\n",
            "Epoch 4043/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0069 - accuracy: 0.6082\n",
            "Epoch 4044/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0069 - accuracy: 0.6203\n",
            "Epoch 4045/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0069 - accuracy: 0.6022\n",
            "Epoch 4046/5000\n",
            "1/1 [==============================] - 0s 258ms/step - loss: 0.0069 - accuracy: 0.6181\n",
            "Epoch 4047/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0069 - accuracy: 0.5954\n",
            "Epoch 4048/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0068 - accuracy: 0.6181\n",
            "Epoch 4049/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0068 - accuracy: 0.6077\n",
            "Epoch 4050/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0067 - accuracy: 0.6201\n",
            "Epoch 4051/5000\n",
            "1/1 [==============================] - 0s 293ms/step - loss: 0.0068 - accuracy: 0.6196\n",
            "Epoch 4052/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0067 - accuracy: 0.6088\n",
            "Epoch 4053/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0067 - accuracy: 0.6134\n",
            "Epoch 4054/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0068 - accuracy: 0.6021\n",
            "Epoch 4055/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0068 - accuracy: 0.6126\n",
            "Epoch 4056/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0068 - accuracy: 0.6078\n",
            "Epoch 4057/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0068 - accuracy: 0.6129\n",
            "Epoch 4058/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0068 - accuracy: 0.6109\n",
            "Epoch 4059/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0068 - accuracy: 0.6100\n",
            "Epoch 4060/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0068 - accuracy: 0.6104\n",
            "Epoch 4061/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0067 - accuracy: 0.6110\n",
            "Epoch 4062/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0067 - accuracy: 0.6097\n",
            "Epoch 4063/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0067 - accuracy: 0.6132\n",
            "Epoch 4064/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0068 - accuracy: 0.6072\n",
            "Epoch 4065/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0067 - accuracy: 0.6125\n",
            "Epoch 4066/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0067 - accuracy: 0.6048\n",
            "Epoch 4067/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0067 - accuracy: 0.6122\n",
            "Epoch 4068/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0067 - accuracy: 0.6076\n",
            "Epoch 4069/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0067 - accuracy: 0.6135\n",
            "Epoch 4070/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0067 - accuracy: 0.6133\n",
            "Epoch 4071/5000\n",
            "1/1 [==============================] - 0s 250ms/step - loss: 0.0067 - accuracy: 0.6164\n",
            "Epoch 4072/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0067 - accuracy: 0.6169\n",
            "Epoch 4073/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0067 - accuracy: 0.6146\n",
            "Epoch 4074/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0067 - accuracy: 0.6080\n",
            "Epoch 4075/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0067 - accuracy: 0.6084\n",
            "Epoch 4076/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0068 - accuracy: 0.5990\n",
            "Epoch 4077/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0068 - accuracy: 0.6133\n",
            "Epoch 4078/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0068 - accuracy: 0.6037\n",
            "Epoch 4079/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0068 - accuracy: 0.6236\n",
            "Epoch 4080/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0069 - accuracy: 0.6037\n",
            "Epoch 4081/5000\n",
            "1/1 [==============================] - 0s 257ms/step - loss: 0.0069 - accuracy: 0.6237\n",
            "Epoch 4082/5000\n",
            "1/1 [==============================] - 0s 257ms/step - loss: 0.0069 - accuracy: 0.6007\n",
            "Epoch 4083/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0069 - accuracy: 0.6161\n",
            "Epoch 4084/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0070 - accuracy: 0.6061\n",
            "Epoch 4085/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0069 - accuracy: 0.6120\n",
            "Epoch 4086/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0069 - accuracy: 0.6106\n",
            "Epoch 4087/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0068 - accuracy: 0.6134\n",
            "Epoch 4088/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0067 - accuracy: 0.6066\n",
            "Epoch 4089/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0068 - accuracy: 0.6145\n",
            "Epoch 4090/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0068 - accuracy: 0.6042\n",
            "Epoch 4091/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0068 - accuracy: 0.6138\n",
            "Epoch 4092/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0068 - accuracy: 0.6133\n",
            "Epoch 4093/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0068 - accuracy: 0.6086\n",
            "Epoch 4094/5000\n",
            "1/1 [==============================] - 0s 258ms/step - loss: 0.0068 - accuracy: 0.6170\n",
            "Epoch 4095/5000\n",
            "1/1 [==============================] - 0s 294ms/step - loss: 0.0068 - accuracy: 0.6000\n",
            "Epoch 4096/5000\n",
            "1/1 [==============================] - 0s 260ms/step - loss: 0.0068 - accuracy: 0.6134\n",
            "Epoch 4097/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0068 - accuracy: 0.6041\n",
            "Epoch 4098/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0067 - accuracy: 0.6097\n",
            "Epoch 4099/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0067 - accuracy: 0.6120\n",
            "Epoch 4100/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0068 - accuracy: 0.6095\n",
            "Epoch 4101/5000\n",
            "1/1 [==============================] - 0s 290ms/step - loss: 0.0068 - accuracy: 0.6180\n",
            "Epoch 4102/5000\n",
            "1/1 [==============================] - 0s 302ms/step - loss: 0.0067 - accuracy: 0.6133\n",
            "Epoch 4103/5000\n",
            "1/1 [==============================] - 0s 291ms/step - loss: 0.0067 - accuracy: 0.6158\n",
            "Epoch 4104/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0067 - accuracy: 0.6097\n",
            "Epoch 4105/5000\n",
            "1/1 [==============================] - 0s 290ms/step - loss: 0.0067 - accuracy: 0.6024\n",
            "Epoch 4106/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0067 - accuracy: 0.6038\n",
            "Epoch 4107/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0067 - accuracy: 0.6029\n",
            "Epoch 4108/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0067 - accuracy: 0.6083\n",
            "Epoch 4109/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0067 - accuracy: 0.6142\n",
            "Epoch 4110/5000\n",
            "1/1 [==============================] - 0s 260ms/step - loss: 0.0067 - accuracy: 0.6110\n",
            "Epoch 4111/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0068 - accuracy: 0.6200\n",
            "Epoch 4112/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0068 - accuracy: 0.6042\n",
            "Epoch 4113/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0068 - accuracy: 0.6176\n",
            "Epoch 4114/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0068 - accuracy: 0.6028\n",
            "Epoch 4115/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0068 - accuracy: 0.6175\n",
            "Epoch 4116/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0068 - accuracy: 0.6069\n",
            "Epoch 4117/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0069 - accuracy: 0.6130\n",
            "Epoch 4118/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0069 - accuracy: 0.6052\n",
            "Epoch 4119/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0069 - accuracy: 0.6167\n",
            "Epoch 4120/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0068 - accuracy: 0.6046\n",
            "Epoch 4121/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0069 - accuracy: 0.6232\n",
            "Epoch 4122/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0069 - accuracy: 0.5989\n",
            "Epoch 4123/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0069 - accuracy: 0.6215\n",
            "Epoch 4124/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0068 - accuracy: 0.5980\n",
            "Epoch 4125/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0067 - accuracy: 0.6057\n",
            "Epoch 4126/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0068 - accuracy: 0.6112\n",
            "Epoch 4127/5000\n",
            "1/1 [==============================] - 0s 291ms/step - loss: 0.0068 - accuracy: 0.5999\n",
            "Epoch 4128/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0067 - accuracy: 0.6214\n",
            "Epoch 4129/5000\n",
            "1/1 [==============================] - 0s 292ms/step - loss: 0.0067 - accuracy: 0.6190\n",
            "Epoch 4130/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0068 - accuracy: 0.6180\n",
            "Epoch 4131/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0068 - accuracy: 0.6131\n",
            "Epoch 4132/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0068 - accuracy: 0.6053\n",
            "Epoch 4133/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0068 - accuracy: 0.5933\n",
            "Epoch 4134/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0068 - accuracy: 0.6085\n",
            "Epoch 4135/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0068 - accuracy: 0.5990\n",
            "Epoch 4136/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0068 - accuracy: 0.6245\n",
            "Epoch 4137/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0068 - accuracy: 0.6107\n",
            "Epoch 4138/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0067 - accuracy: 0.6197\n",
            "Epoch 4139/5000\n",
            "1/1 [==============================] - 0s 292ms/step - loss: 0.0067 - accuracy: 0.6081\n",
            "Epoch 4140/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0067 - accuracy: 0.5994\n",
            "Epoch 4141/5000\n",
            "1/1 [==============================] - 0s 294ms/step - loss: 0.0067 - accuracy: 0.6045\n",
            "Epoch 4142/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0067 - accuracy: 0.6085\n",
            "Epoch 4143/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0067 - accuracy: 0.6216\n",
            "Epoch 4144/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0067 - accuracy: 0.6256\n",
            "Epoch 4145/5000\n",
            "1/1 [==============================] - 0s 294ms/step - loss: 0.0067 - accuracy: 0.6212\n",
            "Epoch 4146/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0067 - accuracy: 0.6072\n",
            "Epoch 4147/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0067 - accuracy: 0.6034\n",
            "Epoch 4148/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0068 - accuracy: 0.5891\n",
            "Epoch 4149/5000\n",
            "1/1 [==============================] - 0s 299ms/step - loss: 0.0068 - accuracy: 0.6087\n",
            "Epoch 4150/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0068 - accuracy: 0.6051\n",
            "Epoch 4151/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0068 - accuracy: 0.6289\n",
            "Epoch 4152/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0068 - accuracy: 0.6162\n",
            "Epoch 4153/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0068 - accuracy: 0.6187\n",
            "Epoch 4154/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0068 - accuracy: 0.5998\n",
            "Epoch 4155/5000\n",
            "1/1 [==============================] - 0s 305ms/step - loss: 0.0068 - accuracy: 0.6030\n",
            "Epoch 4156/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0067 - accuracy: 0.5993\n",
            "Epoch 4157/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0067 - accuracy: 0.6120\n",
            "Epoch 4158/5000\n",
            "1/1 [==============================] - 0s 260ms/step - loss: 0.0067 - accuracy: 0.6193\n",
            "Epoch 4159/5000\n",
            "1/1 [==============================] - 0s 296ms/step - loss: 0.0067 - accuracy: 0.6236\n",
            "Epoch 4160/5000\n",
            "1/1 [==============================] - 0s 258ms/step - loss: 0.0067 - accuracy: 0.6194\n",
            "Epoch 4161/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0067 - accuracy: 0.6106\n",
            "Epoch 4162/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0067 - accuracy: 0.6028\n",
            "Epoch 4163/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0067 - accuracy: 0.5994\n",
            "Epoch 4164/5000\n",
            "1/1 [==============================] - 0s 306ms/step - loss: 0.0067 - accuracy: 0.6055\n",
            "Epoch 4165/5000\n",
            "1/1 [==============================] - 0s 313ms/step - loss: 0.0067 - accuracy: 0.6117\n",
            "Epoch 4166/5000\n",
            "1/1 [==============================] - 0s 306ms/step - loss: 0.0067 - accuracy: 0.6179\n",
            "Epoch 4167/5000\n",
            "1/1 [==============================] - 0s 300ms/step - loss: 0.0067 - accuracy: 0.6133\n",
            "Epoch 4168/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0067 - accuracy: 0.6187\n",
            "Epoch 4169/5000\n",
            "1/1 [==============================] - 0s 296ms/step - loss: 0.0068 - accuracy: 0.6001\n",
            "Epoch 4170/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0068 - accuracy: 0.6181\n",
            "Epoch 4171/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0069 - accuracy: 0.5985\n",
            "Epoch 4172/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0069 - accuracy: 0.6227\n",
            "Epoch 4173/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0069 - accuracy: 0.6046\n",
            "Epoch 4174/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0069 - accuracy: 0.6138\n",
            "Epoch 4175/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0069 - accuracy: 0.6059\n",
            "Epoch 4176/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0068 - accuracy: 0.6057\n",
            "Epoch 4177/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0067 - accuracy: 0.6090\n",
            "Epoch 4178/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0067 - accuracy: 0.6158\n",
            "Epoch 4179/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0067 - accuracy: 0.6161\n",
            "Epoch 4180/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0067 - accuracy: 0.6177\n",
            "Epoch 4181/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0068 - accuracy: 0.6069\n",
            "Epoch 4182/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0067 - accuracy: 0.6027\n",
            "Epoch 4183/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0067 - accuracy: 0.6052\n",
            "Epoch 4184/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0067 - accuracy: 0.6014\n",
            "Epoch 4185/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0067 - accuracy: 0.6194\n",
            "Epoch 4186/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0067 - accuracy: 0.6110\n",
            "Epoch 4187/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0067 - accuracy: 0.6253\n",
            "Epoch 4188/5000\n",
            "1/1 [==============================] - 0s 260ms/step - loss: 0.0067 - accuracy: 0.6075\n",
            "Epoch 4189/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0067 - accuracy: 0.6107\n",
            "Epoch 4190/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0067 - accuracy: 0.5992\n",
            "Epoch 4191/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0067 - accuracy: 0.6057\n",
            "Epoch 4192/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0067 - accuracy: 0.6077\n",
            "Epoch 4193/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0067 - accuracy: 0.6163\n",
            "Epoch 4194/5000\n",
            "1/1 [==============================] - 0s 291ms/step - loss: 0.0067 - accuracy: 0.6171\n",
            "Epoch 4195/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0067 - accuracy: 0.6188\n",
            "Epoch 4196/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0067 - accuracy: 0.6079\n",
            "Epoch 4197/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0067 - accuracy: 0.6070\n",
            "Epoch 4198/5000\n",
            "1/1 [==============================] - 0s 303ms/step - loss: 0.0067 - accuracy: 0.5994\n",
            "Epoch 4199/5000\n",
            "1/1 [==============================] - 0s 307ms/step - loss: 0.0067 - accuracy: 0.6095\n",
            "Epoch 4200/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0067 - accuracy: 0.6056\n",
            "Epoch 4201/5000\n",
            "1/1 [==============================] - 0s 303ms/step - loss: 0.0067 - accuracy: 0.6200\n",
            "Epoch 4202/5000\n",
            "1/1 [==============================] - 0s 301ms/step - loss: 0.0067 - accuracy: 0.6121\n",
            "Epoch 4203/5000\n",
            "1/1 [==============================] - 0s 295ms/step - loss: 0.0067 - accuracy: 0.6175\n",
            "Epoch 4204/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0067 - accuracy: 0.6052\n",
            "Epoch 4205/5000\n",
            "1/1 [==============================] - 0s 296ms/step - loss: 0.0068 - accuracy: 0.6084\n",
            "Epoch 4206/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0068 - accuracy: 0.6006\n",
            "Epoch 4207/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0068 - accuracy: 0.6167\n",
            "Epoch 4208/5000\n",
            "1/1 [==============================] - 0s 296ms/step - loss: 0.0068 - accuracy: 0.6111\n",
            "Epoch 4209/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0069 - accuracy: 0.6306\n",
            "Epoch 4210/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0069 - accuracy: 0.5991\n",
            "Epoch 4211/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0069 - accuracy: 0.6210\n",
            "Epoch 4212/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0069 - accuracy: 0.5857\n",
            "Epoch 4213/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0068 - accuracy: 0.6034\n",
            "Epoch 4214/5000\n",
            "1/1 [==============================] - 0s 298ms/step - loss: 0.0067 - accuracy: 0.6038\n",
            "Epoch 4215/5000\n",
            "1/1 [==============================] - 0s 300ms/step - loss: 0.0067 - accuracy: 0.6077\n",
            "Epoch 4216/5000\n",
            "1/1 [==============================] - 0s 299ms/step - loss: 0.0068 - accuracy: 0.6327\n",
            "Epoch 4217/5000\n",
            "1/1 [==============================] - 0s 292ms/step - loss: 0.0067 - accuracy: 0.6223\n",
            "Epoch 4218/5000\n",
            "1/1 [==============================] - 0s 291ms/step - loss: 0.0067 - accuracy: 0.6244\n",
            "Epoch 4219/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0068 - accuracy: 0.6109\n",
            "Epoch 4220/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0068 - accuracy: 0.6005\n",
            "Epoch 4221/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0068 - accuracy: 0.5948\n",
            "Epoch 4222/5000\n",
            "1/1 [==============================] - 0s 309ms/step - loss: 0.0068 - accuracy: 0.6023\n",
            "Epoch 4223/5000\n",
            "1/1 [==============================] - 0s 294ms/step - loss: 0.0067 - accuracy: 0.5986\n",
            "Epoch 4224/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0068 - accuracy: 0.6240\n",
            "Epoch 4225/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0067 - accuracy: 0.6095\n",
            "Epoch 4226/5000\n",
            "1/1 [==============================] - 0s 297ms/step - loss: 0.0067 - accuracy: 0.6223\n",
            "Epoch 4227/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0067 - accuracy: 0.6108\n",
            "Epoch 4228/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0067 - accuracy: 0.6018\n",
            "Epoch 4229/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0067 - accuracy: 0.6085\n",
            "Epoch 4230/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0067 - accuracy: 0.6021\n",
            "Epoch 4231/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0067 - accuracy: 0.6189\n",
            "Epoch 4232/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0067 - accuracy: 0.6151\n",
            "Epoch 4233/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0067 - accuracy: 0.6208\n",
            "Epoch 4234/5000\n",
            "1/1 [==============================] - 0s 253ms/step - loss: 0.0067 - accuracy: 0.6063\n",
            "Epoch 4235/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0067 - accuracy: 0.6074\n",
            "Epoch 4236/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0067 - accuracy: 0.5946\n",
            "Epoch 4237/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0067 - accuracy: 0.6120\n",
            "Epoch 4238/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0067 - accuracy: 0.6086\n",
            "Epoch 4239/5000\n",
            "1/1 [==============================] - 0s 259ms/step - loss: 0.0067 - accuracy: 0.6254\n",
            "Epoch 4240/5000\n",
            "1/1 [==============================] - 0s 295ms/step - loss: 0.0067 - accuracy: 0.6130\n",
            "Epoch 4241/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0067 - accuracy: 0.6193\n",
            "Epoch 4242/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0067 - accuracy: 0.5977\n",
            "Epoch 4243/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0067 - accuracy: 0.6082\n",
            "Epoch 4244/5000\n",
            "1/1 [==============================] - 0s 299ms/step - loss: 0.0067 - accuracy: 0.6008\n",
            "Epoch 4245/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0067 - accuracy: 0.6142\n",
            "Epoch 4246/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0067 - accuracy: 0.6174\n",
            "Epoch 4247/5000\n",
            "1/1 [==============================] - 0s 299ms/step - loss: 0.0067 - accuracy: 0.6156\n",
            "Epoch 4248/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0067 - accuracy: 0.6226\n",
            "Epoch 4249/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0067 - accuracy: 0.6080\n",
            "Epoch 4250/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0067 - accuracy: 0.6145\n",
            "Epoch 4251/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0067 - accuracy: 0.5989\n",
            "Epoch 4252/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0067 - accuracy: 0.6063\n",
            "Epoch 4253/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0068 - accuracy: 0.5993\n",
            "Epoch 4254/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0068 - accuracy: 0.6103\n",
            "Epoch 4255/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0068 - accuracy: 0.6089\n",
            "Epoch 4256/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0067 - accuracy: 0.6209\n",
            "Epoch 4257/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0067 - accuracy: 0.6110\n",
            "Epoch 4258/5000\n",
            "1/1 [==============================] - 0s 292ms/step - loss: 0.0068 - accuracy: 0.6244\n",
            "Epoch 4259/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0068 - accuracy: 0.5968\n",
            "Epoch 4260/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0068 - accuracy: 0.6167\n",
            "Epoch 4261/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0067 - accuracy: 0.5960\n",
            "Epoch 4262/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0067 - accuracy: 0.6107\n",
            "Epoch 4263/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0067 - accuracy: 0.6155\n",
            "Epoch 4264/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0067 - accuracy: 0.6075\n",
            "Epoch 4265/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0067 - accuracy: 0.6196\n",
            "Epoch 4266/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0067 - accuracy: 0.6115\n",
            "Epoch 4267/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0067 - accuracy: 0.6122\n",
            "Epoch 4268/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0067 - accuracy: 0.6130\n",
            "Epoch 4269/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0067 - accuracy: 0.6057\n",
            "Epoch 4270/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0067 - accuracy: 0.6046\n",
            "Epoch 4271/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0067 - accuracy: 0.6089\n",
            "Epoch 4272/5000\n",
            "1/1 [==============================] - 0s 291ms/step - loss: 0.0067 - accuracy: 0.5981\n",
            "Epoch 4273/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0068 - accuracy: 0.6201\n",
            "Epoch 4274/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0068 - accuracy: 0.6025\n",
            "Epoch 4275/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0068 - accuracy: 0.6235\n",
            "Epoch 4276/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0067 - accuracy: 0.6068\n",
            "Epoch 4277/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0067 - accuracy: 0.6101\n",
            "Epoch 4278/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0067 - accuracy: 0.6047\n",
            "Epoch 4279/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0067 - accuracy: 0.6057\n",
            "Epoch 4280/5000\n",
            "1/1 [==============================] - 0s 291ms/step - loss: 0.0067 - accuracy: 0.6150\n",
            "Epoch 4281/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0067 - accuracy: 0.6231\n",
            "Epoch 4282/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0067 - accuracy: 0.6176\n",
            "Epoch 4283/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0067 - accuracy: 0.6191\n",
            "Epoch 4284/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0067 - accuracy: 0.5955\n",
            "Epoch 4285/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0067 - accuracy: 0.5959\n",
            "Epoch 4286/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0067 - accuracy: 0.6012\n",
            "Epoch 4287/5000\n",
            "1/1 [==============================] - 0s 294ms/step - loss: 0.0067 - accuracy: 0.6077\n",
            "Epoch 4288/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0067 - accuracy: 0.6242\n",
            "Epoch 4289/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0067 - accuracy: 0.6179\n",
            "Epoch 4290/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0067 - accuracy: 0.6191\n",
            "Epoch 4291/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0067 - accuracy: 0.6066\n",
            "Epoch 4292/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0067 - accuracy: 0.6022\n",
            "Epoch 4293/5000\n",
            "1/1 [==============================] - 0s 260ms/step - loss: 0.0067 - accuracy: 0.5968\n",
            "Epoch 4294/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0067 - accuracy: 0.6107\n",
            "Epoch 4295/5000\n",
            "1/1 [==============================] - 0s 261ms/step - loss: 0.0067 - accuracy: 0.6107\n",
            "Epoch 4296/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0068 - accuracy: 0.6320\n",
            "Epoch 4297/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0068 - accuracy: 0.6124\n",
            "Epoch 4298/5000\n",
            "1/1 [==============================] - 0s 293ms/step - loss: 0.0068 - accuracy: 0.6236\n",
            "Epoch 4299/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0069 - accuracy: 0.5943\n",
            "Epoch 4300/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0069 - accuracy: 0.6103\n",
            "Epoch 4301/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0069 - accuracy: 0.5924\n",
            "Epoch 4302/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0068 - accuracy: 0.6149\n",
            "Epoch 4303/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0068 - accuracy: 0.6108\n",
            "Epoch 4304/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0067 - accuracy: 0.6183\n",
            "Epoch 4305/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0067 - accuracy: 0.6160\n",
            "Epoch 4306/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0066 - accuracy: 0.6097\n",
            "Epoch 4307/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0067 - accuracy: 0.6133\n",
            "Epoch 4308/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0067 - accuracy: 0.6046\n",
            "Epoch 4309/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0067 - accuracy: 0.6123\n",
            "Epoch 4310/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0068 - accuracy: 0.6038\n",
            "Epoch 4311/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0067 - accuracy: 0.6127\n",
            "Epoch 4312/5000\n",
            "1/1 [==============================] - 0s 290ms/step - loss: 0.0067 - accuracy: 0.6048\n",
            "Epoch 4313/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0067 - accuracy: 0.6184\n",
            "Epoch 4314/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0067 - accuracy: 0.6044\n",
            "Epoch 4315/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0067 - accuracy: 0.6214\n",
            "Epoch 4316/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0067 - accuracy: 0.6036\n",
            "Epoch 4317/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0067 - accuracy: 0.6148\n",
            "Epoch 4318/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0066 - accuracy: 0.6074\n",
            "Epoch 4319/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0066 - accuracy: 0.6057\n",
            "Epoch 4320/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0067 - accuracy: 0.6188\n",
            "Epoch 4321/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0067 - accuracy: 0.6103\n",
            "Epoch 4322/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0067 - accuracy: 0.6243\n",
            "Epoch 4323/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0067 - accuracy: 0.6099\n",
            "Epoch 4324/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0067 - accuracy: 0.6117\n",
            "Epoch 4325/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0067 - accuracy: 0.5960\n",
            "Epoch 4326/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0068 - accuracy: 0.6016\n",
            "Epoch 4327/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0067 - accuracy: 0.5963\n",
            "Epoch 4328/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0067 - accuracy: 0.6149\n",
            "Epoch 4329/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0067 - accuracy: 0.6145\n",
            "Epoch 4330/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0067 - accuracy: 0.6294\n",
            "Epoch 4331/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0067 - accuracy: 0.6125\n",
            "Epoch 4332/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0067 - accuracy: 0.6153\n",
            "Epoch 4333/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0067 - accuracy: 0.5981\n",
            "Epoch 4334/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0067 - accuracy: 0.6034\n",
            "Epoch 4335/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0066 - accuracy: 0.6052\n",
            "Epoch 4336/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0066 - accuracy: 0.6164\n",
            "Epoch 4337/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0066 - accuracy: 0.6195\n",
            "Epoch 4338/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0066 - accuracy: 0.6193\n",
            "Epoch 4339/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0066 - accuracy: 0.6113\n",
            "Epoch 4340/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0066 - accuracy: 0.6030\n",
            "Epoch 4341/5000\n",
            "1/1 [==============================] - 0s 300ms/step - loss: 0.0066 - accuracy: 0.6033\n",
            "Epoch 4342/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0066 - accuracy: 0.6030\n",
            "Epoch 4343/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0066 - accuracy: 0.6152\n",
            "Epoch 4344/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0067 - accuracy: 0.6118\n",
            "Epoch 4345/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0067 - accuracy: 0.6220\n",
            "Epoch 4346/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0067 - accuracy: 0.6050\n",
            "Epoch 4347/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0067 - accuracy: 0.6141\n",
            "Epoch 4348/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0067 - accuracy: 0.5980\n",
            "Epoch 4349/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0068 - accuracy: 0.6155\n",
            "Epoch 4350/5000\n",
            "1/1 [==============================] - 0s 257ms/step - loss: 0.0069 - accuracy: 0.6005\n",
            "Epoch 4351/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0069 - accuracy: 0.6236\n",
            "Epoch 4352/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0069 - accuracy: 0.6036\n",
            "Epoch 4353/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0069 - accuracy: 0.6243\n",
            "Epoch 4354/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0068 - accuracy: 0.6012\n",
            "Epoch 4355/5000\n",
            "1/1 [==============================] - 0s 260ms/step - loss: 0.0067 - accuracy: 0.6133\n",
            "Epoch 4356/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0067 - accuracy: 0.6029\n",
            "Epoch 4357/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0066 - accuracy: 0.6060\n",
            "Epoch 4358/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0066 - accuracy: 0.6152\n",
            "Epoch 4359/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0067 - accuracy: 0.6113\n",
            "Epoch 4360/5000\n",
            "1/1 [==============================] - 0s 261ms/step - loss: 0.0068 - accuracy: 0.6297\n",
            "Epoch 4361/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0068 - accuracy: 0.6083\n",
            "Epoch 4362/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0068 - accuracy: 0.6236\n",
            "Epoch 4363/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0068 - accuracy: 0.5935\n",
            "Epoch 4364/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0068 - accuracy: 0.6061\n",
            "Epoch 4365/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0067 - accuracy: 0.5988\n",
            "Epoch 4366/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0067 - accuracy: 0.6052\n",
            "Epoch 4367/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0066 - accuracy: 0.6209\n",
            "Epoch 4368/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0067 - accuracy: 0.6177\n",
            "Epoch 4369/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0067 - accuracy: 0.6288\n",
            "Epoch 4370/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0068 - accuracy: 0.6086\n",
            "Epoch 4371/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0068 - accuracy: 0.6111\n",
            "Epoch 4372/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0067 - accuracy: 0.5917\n",
            "Epoch 4373/5000\n",
            "1/1 [==============================] - 0s 259ms/step - loss: 0.0067 - accuracy: 0.5998\n",
            "Epoch 4374/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0067 - accuracy: 0.6025\n",
            "Epoch 4375/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0066 - accuracy: 0.6122\n",
            "Epoch 4376/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0066 - accuracy: 0.6223\n",
            "Epoch 4377/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0067 - accuracy: 0.6176\n",
            "Epoch 4378/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0067 - accuracy: 0.6223\n",
            "Epoch 4379/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0067 - accuracy: 0.5997\n",
            "Epoch 4380/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0068 - accuracy: 0.6134\n",
            "Epoch 4381/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0068 - accuracy: 0.5931\n",
            "Epoch 4382/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0067 - accuracy: 0.6175\n",
            "Epoch 4383/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0067 - accuracy: 0.6064\n",
            "Epoch 4384/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0066 - accuracy: 0.6147\n",
            "Epoch 4385/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0066 - accuracy: 0.6149\n",
            "Epoch 4386/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0067 - accuracy: 0.6038\n",
            "Epoch 4387/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0067 - accuracy: 0.6181\n",
            "Epoch 4388/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0067 - accuracy: 0.6076\n",
            "Epoch 4389/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0067 - accuracy: 0.6273\n",
            "Epoch 4390/5000\n",
            "1/1 [==============================] - 0s 257ms/step - loss: 0.0068 - accuracy: 0.6069\n",
            "Epoch 4391/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0067 - accuracy: 0.6127\n",
            "Epoch 4392/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0067 - accuracy: 0.5946\n",
            "Epoch 4393/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0066 - accuracy: 0.5952\n",
            "Epoch 4394/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0067 - accuracy: 0.6049\n",
            "Epoch 4395/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0067 - accuracy: 0.6084\n",
            "Epoch 4396/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0067 - accuracy: 0.6281\n",
            "Epoch 4397/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0067 - accuracy: 0.6218\n",
            "Epoch 4398/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0067 - accuracy: 0.6224\n",
            "Epoch 4399/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0067 - accuracy: 0.6040\n",
            "Epoch 4400/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0067 - accuracy: 0.5980\n",
            "Epoch 4401/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0067 - accuracy: 0.5946\n",
            "Epoch 4402/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0066 - accuracy: 0.6036\n",
            "Epoch 4403/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0066 - accuracy: 0.6175\n",
            "Epoch 4404/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0066 - accuracy: 0.6192\n",
            "Epoch 4405/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0066 - accuracy: 0.6228\n",
            "Epoch 4406/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0066 - accuracy: 0.6086\n",
            "Epoch 4407/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0066 - accuracy: 0.6080\n",
            "Epoch 4408/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0067 - accuracy: 0.5984\n",
            "Epoch 4409/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0067 - accuracy: 0.6123\n",
            "Epoch 4410/5000\n",
            "1/1 [==============================] - 0s 292ms/step - loss: 0.0067 - accuracy: 0.6043\n",
            "Epoch 4411/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0067 - accuracy: 0.6246\n",
            "Epoch 4412/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0067 - accuracy: 0.6066\n",
            "Epoch 4413/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0067 - accuracy: 0.6171\n",
            "Epoch 4414/5000\n",
            "1/1 [==============================] - 0s 293ms/step - loss: 0.0066 - accuracy: 0.6031\n",
            "Epoch 4415/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0066 - accuracy: 0.6070\n",
            "Epoch 4416/5000\n",
            "1/1 [==============================] - 0s 296ms/step - loss: 0.0066 - accuracy: 0.6116\n",
            "Epoch 4417/5000\n",
            "1/1 [==============================] - 0s 292ms/step - loss: 0.0066 - accuracy: 0.6126\n",
            "Epoch 4418/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0066 - accuracy: 0.6242\n",
            "Epoch 4419/5000\n",
            "1/1 [==============================] - 0s 306ms/step - loss: 0.0067 - accuracy: 0.6089\n",
            "Epoch 4420/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0066 - accuracy: 0.6155\n",
            "Epoch 4421/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0066 - accuracy: 0.5977\n",
            "Epoch 4422/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0066 - accuracy: 0.6046\n",
            "Epoch 4423/5000\n",
            "1/1 [==============================] - 0s 261ms/step - loss: 0.0066 - accuracy: 0.6003\n",
            "Epoch 4424/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0066 - accuracy: 0.6127\n",
            "Epoch 4425/5000\n",
            "1/1 [==============================] - 0s 312ms/step - loss: 0.0066 - accuracy: 0.6169\n",
            "Epoch 4426/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0066 - accuracy: 0.6206\n",
            "Epoch 4427/5000\n",
            "1/1 [==============================] - 0s 293ms/step - loss: 0.0066 - accuracy: 0.6191\n",
            "Epoch 4428/5000\n",
            "1/1 [==============================] - 0s 293ms/step - loss: 0.0066 - accuracy: 0.6118\n",
            "Epoch 4429/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0066 - accuracy: 0.6067\n",
            "Epoch 4430/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0066 - accuracy: 0.6015\n",
            "Epoch 4431/5000\n",
            "1/1 [==============================] - 0s 290ms/step - loss: 0.0066 - accuracy: 0.6053\n",
            "Epoch 4432/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0066 - accuracy: 0.6052\n",
            "Epoch 4433/5000\n",
            "1/1 [==============================] - 0s 309ms/step - loss: 0.0066 - accuracy: 0.6171\n",
            "Epoch 4434/5000\n",
            "1/1 [==============================] - 0s 310ms/step - loss: 0.0067 - accuracy: 0.6070\n",
            "Epoch 4435/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0067 - accuracy: 0.6239\n",
            "Epoch 4436/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0067 - accuracy: 0.6030\n",
            "Epoch 4437/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0068 - accuracy: 0.6216\n",
            "Epoch 4438/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0068 - accuracy: 0.5986\n",
            "Epoch 4439/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0068 - accuracy: 0.6185\n",
            "Epoch 4440/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0068 - accuracy: 0.6002\n",
            "Epoch 4441/5000\n",
            "1/1 [==============================] - 0s 296ms/step - loss: 0.0067 - accuracy: 0.6116\n",
            "Epoch 4442/5000\n",
            "1/1 [==============================] - 0s 292ms/step - loss: 0.0067 - accuracy: 0.6048\n",
            "Epoch 4443/5000\n",
            "1/1 [==============================] - 0s 290ms/step - loss: 0.0067 - accuracy: 0.6072\n",
            "Epoch 4444/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0066 - accuracy: 0.6122\n",
            "Epoch 4445/5000\n",
            "1/1 [==============================] - 0s 292ms/step - loss: 0.0066 - accuracy: 0.6135\n",
            "Epoch 4446/5000\n",
            "1/1 [==============================] - 0s 295ms/step - loss: 0.0066 - accuracy: 0.6194\n",
            "Epoch 4447/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0066 - accuracy: 0.6099\n",
            "Epoch 4448/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0067 - accuracy: 0.6152\n",
            "Epoch 4449/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0067 - accuracy: 0.5985\n",
            "Epoch 4450/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0067 - accuracy: 0.6081\n",
            "Epoch 4451/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0067 - accuracy: 0.5966\n",
            "Epoch 4452/5000\n",
            "1/1 [==============================] - 0s 302ms/step - loss: 0.0067 - accuracy: 0.6139\n",
            "Epoch 4453/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0067 - accuracy: 0.6094\n",
            "Epoch 4454/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0067 - accuracy: 0.6232\n",
            "Epoch 4455/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0067 - accuracy: 0.6112\n",
            "Epoch 4456/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0067 - accuracy: 0.6212\n",
            "Epoch 4457/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0066 - accuracy: 0.6044\n",
            "Epoch 4458/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0066 - accuracy: 0.6157\n",
            "Epoch 4459/5000\n",
            "1/1 [==============================] - 0s 306ms/step - loss: 0.0066 - accuracy: 0.6098\n",
            "Epoch 4460/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0066 - accuracy: 0.6126\n",
            "Epoch 4461/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0066 - accuracy: 0.6148\n",
            "Epoch 4462/5000\n",
            "1/1 [==============================] - 0s 293ms/step - loss: 0.0066 - accuracy: 0.6043\n",
            "Epoch 4463/5000\n",
            "1/1 [==============================] - 0s 301ms/step - loss: 0.0067 - accuracy: 0.6136\n",
            "Epoch 4464/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0067 - accuracy: 0.5977\n",
            "Epoch 4465/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0067 - accuracy: 0.6197\n",
            "Epoch 4466/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0067 - accuracy: 0.6041\n",
            "Epoch 4467/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0067 - accuracy: 0.6171\n",
            "Epoch 4468/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0066 - accuracy: 0.6046\n",
            "Epoch 4469/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0067 - accuracy: 0.6028\n",
            "Epoch 4470/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0067 - accuracy: 0.6110\n",
            "Epoch 4471/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0066 - accuracy: 0.6085\n",
            "Epoch 4472/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0066 - accuracy: 0.6234\n",
            "Epoch 4473/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0066 - accuracy: 0.6182\n",
            "Epoch 4474/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0066 - accuracy: 0.6168\n",
            "Epoch 4475/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0066 - accuracy: 0.6101\n",
            "Epoch 4476/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0066 - accuracy: 0.5999\n",
            "Epoch 4477/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0066 - accuracy: 0.6044\n",
            "Epoch 4478/5000\n",
            "1/1 [==============================] - 0s 259ms/step - loss: 0.0066 - accuracy: 0.6048\n",
            "Epoch 4479/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0066 - accuracy: 0.6126\n",
            "Epoch 4480/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0066 - accuracy: 0.6168\n",
            "Epoch 4481/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0066 - accuracy: 0.6112\n",
            "Epoch 4482/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0066 - accuracy: 0.6118\n",
            "Epoch 4483/5000\n",
            "1/1 [==============================] - 0s 260ms/step - loss: 0.0066 - accuracy: 0.6010\n",
            "Epoch 4484/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0066 - accuracy: 0.6133\n",
            "Epoch 4485/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0066 - accuracy: 0.6100\n",
            "Epoch 4486/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0066 - accuracy: 0.6223\n",
            "Epoch 4487/5000\n",
            "1/1 [==============================] - 0s 293ms/step - loss: 0.0066 - accuracy: 0.6106\n",
            "Epoch 4488/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0066 - accuracy: 0.6109\n",
            "Epoch 4489/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0066 - accuracy: 0.5971\n",
            "Epoch 4490/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0067 - accuracy: 0.6068\n",
            "Epoch 4491/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0067 - accuracy: 0.6059\n",
            "Epoch 4492/5000\n",
            "1/1 [==============================] - 0s 257ms/step - loss: 0.0067 - accuracy: 0.6223\n",
            "Epoch 4493/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0068 - accuracy: 0.6116\n",
            "Epoch 4494/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0068 - accuracy: 0.6216\n",
            "Epoch 4495/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0069 - accuracy: 0.5864\n",
            "Epoch 4496/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0069 - accuracy: 0.6147\n",
            "Epoch 4497/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0068 - accuracy: 0.5901\n",
            "Epoch 4498/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0067 - accuracy: 0.6177\n",
            "Epoch 4499/5000\n",
            "1/1 [==============================] - 0s 315ms/step - loss: 0.0067 - accuracy: 0.6161\n",
            "Epoch 4500/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0066 - accuracy: 0.6107\n",
            "Epoch 4501/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0066 - accuracy: 0.6223\n",
            "Epoch 4502/5000\n",
            "1/1 [==============================] - 0s 302ms/step - loss: 0.0067 - accuracy: 0.6073\n",
            "Epoch 4503/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0067 - accuracy: 0.6219\n",
            "Epoch 4504/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0067 - accuracy: 0.6098\n",
            "Epoch 4505/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0067 - accuracy: 0.6120\n",
            "Epoch 4506/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0068 - accuracy: 0.5974\n",
            "Epoch 4507/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0067 - accuracy: 0.6001\n",
            "Epoch 4508/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0067 - accuracy: 0.5946\n",
            "Epoch 4509/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0066 - accuracy: 0.6100\n",
            "Epoch 4510/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0066 - accuracy: 0.6133\n",
            "Epoch 4511/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0067 - accuracy: 0.6295\n",
            "Epoch 4512/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0066 - accuracy: 0.6169\n",
            "Epoch 4513/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0066 - accuracy: 0.6106\n",
            "Epoch 4514/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0066 - accuracy: 0.6048\n",
            "Epoch 4515/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0066 - accuracy: 0.5962\n",
            "Epoch 4516/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0066 - accuracy: 0.6105\n",
            "Epoch 4517/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0066 - accuracy: 0.6054\n",
            "Epoch 4518/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0067 - accuracy: 0.6273\n",
            "Epoch 4519/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0067 - accuracy: 0.6127\n",
            "Epoch 4520/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0066 - accuracy: 0.6162\n",
            "Epoch 4521/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0066 - accuracy: 0.6019\n",
            "Epoch 4522/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0066 - accuracy: 0.6004\n",
            "Epoch 4523/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0066 - accuracy: 0.6095\n",
            "Epoch 4524/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0066 - accuracy: 0.6098\n",
            "Epoch 4525/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0066 - accuracy: 0.6248\n",
            "Epoch 4526/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0066 - accuracy: 0.6152\n",
            "Epoch 4527/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0066 - accuracy: 0.6156\n",
            "Epoch 4528/5000\n",
            "1/1 [==============================] - 0s 260ms/step - loss: 0.0066 - accuracy: 0.5985\n",
            "Epoch 4529/5000\n",
            "1/1 [==============================] - 0s 298ms/step - loss: 0.0066 - accuracy: 0.5992\n",
            "Epoch 4530/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0066 - accuracy: 0.5960\n",
            "Epoch 4531/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0066 - accuracy: 0.6107\n",
            "Epoch 4532/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0066 - accuracy: 0.6144\n",
            "Epoch 4533/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0066 - accuracy: 0.6261\n",
            "Epoch 4534/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0066 - accuracy: 0.6153\n",
            "Epoch 4535/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0066 - accuracy: 0.6175\n",
            "Epoch 4536/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0067 - accuracy: 0.5916\n",
            "Epoch 4537/5000\n",
            "1/1 [==============================] - 0s 293ms/step - loss: 0.0067 - accuracy: 0.6114\n",
            "Epoch 4538/5000\n",
            "1/1 [==============================] - 0s 295ms/step - loss: 0.0067 - accuracy: 0.6006\n",
            "Epoch 4539/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0066 - accuracy: 0.6245\n",
            "Epoch 4540/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0066 - accuracy: 0.6184\n",
            "Epoch 4541/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0066 - accuracy: 0.6155\n",
            "Epoch 4542/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0066 - accuracy: 0.6068\n",
            "Epoch 4543/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0066 - accuracy: 0.5993\n",
            "Epoch 4544/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0066 - accuracy: 0.6042\n",
            "Epoch 4545/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0066 - accuracy: 0.6091\n",
            "Epoch 4546/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0066 - accuracy: 0.6180\n",
            "Epoch 4547/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0066 - accuracy: 0.6167\n",
            "Epoch 4548/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0066 - accuracy: 0.6115\n",
            "Epoch 4549/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0066 - accuracy: 0.6008\n",
            "Epoch 4550/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0066 - accuracy: 0.6002\n",
            "Epoch 4551/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0066 - accuracy: 0.6023\n",
            "Epoch 4552/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0066 - accuracy: 0.6169\n",
            "Epoch 4553/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0066 - accuracy: 0.6183\n",
            "Epoch 4554/5000\n",
            "1/1 [==============================] - 0s 297ms/step - loss: 0.0066 - accuracy: 0.6254\n",
            "Epoch 4555/5000\n",
            "1/1 [==============================] - 0s 303ms/step - loss: 0.0066 - accuracy: 0.6047\n",
            "Epoch 4556/5000\n",
            "1/1 [==============================] - 0s 316ms/step - loss: 0.0066 - accuracy: 0.6104\n",
            "Epoch 4557/5000\n",
            "1/1 [==============================] - 0s 311ms/step - loss: 0.0066 - accuracy: 0.5932\n",
            "Epoch 4558/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0066 - accuracy: 0.6154\n",
            "Epoch 4559/5000\n",
            "1/1 [==============================] - 0s 296ms/step - loss: 0.0067 - accuracy: 0.6075\n",
            "Epoch 4560/5000\n",
            "1/1 [==============================] - 0s 297ms/step - loss: 0.0067 - accuracy: 0.6247\n",
            "Epoch 4561/5000\n",
            "1/1 [==============================] - 0s 300ms/step - loss: 0.0067 - accuracy: 0.6080\n",
            "Epoch 4562/5000\n",
            "1/1 [==============================] - 0s 295ms/step - loss: 0.0067 - accuracy: 0.6139\n",
            "Epoch 4563/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0067 - accuracy: 0.5981\n",
            "Epoch 4564/5000\n",
            "1/1 [==============================] - 0s 294ms/step - loss: 0.0067 - accuracy: 0.6109\n",
            "Epoch 4565/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0067 - accuracy: 0.6024\n",
            "Epoch 4566/5000\n",
            "1/1 [==============================] - 0s 293ms/step - loss: 0.0066 - accuracy: 0.6205\n",
            "Epoch 4567/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0066 - accuracy: 0.6074\n",
            "Epoch 4568/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0066 - accuracy: 0.6177\n",
            "Epoch 4569/5000\n",
            "1/1 [==============================] - 0s 301ms/step - loss: 0.0066 - accuracy: 0.5981\n",
            "Epoch 4570/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0066 - accuracy: 0.6083\n",
            "Epoch 4571/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0066 - accuracy: 0.6077\n",
            "Epoch 4572/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0066 - accuracy: 0.6134\n",
            "Epoch 4573/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0066 - accuracy: 0.6241\n",
            "Epoch 4574/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0066 - accuracy: 0.6133\n",
            "Epoch 4575/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0066 - accuracy: 0.6175\n",
            "Epoch 4576/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0067 - accuracy: 0.5953\n",
            "Epoch 4577/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0067 - accuracy: 0.6123\n",
            "Epoch 4578/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0067 - accuracy: 0.5987\n",
            "Epoch 4579/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0066 - accuracy: 0.6189\n",
            "Epoch 4580/5000\n",
            "1/1 [==============================] - 0s 291ms/step - loss: 0.0066 - accuracy: 0.6105\n",
            "Epoch 4581/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0066 - accuracy: 0.6136\n",
            "Epoch 4582/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0066 - accuracy: 0.6124\n",
            "Epoch 4583/5000\n",
            "1/1 [==============================] - 0s 291ms/step - loss: 0.0066 - accuracy: 0.6056\n",
            "Epoch 4584/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0066 - accuracy: 0.6136\n",
            "Epoch 4585/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0066 - accuracy: 0.6097\n",
            "Epoch 4586/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0066 - accuracy: 0.6138\n",
            "Epoch 4587/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0066 - accuracy: 0.6115\n",
            "Epoch 4588/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0066 - accuracy: 0.6062\n",
            "Epoch 4589/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0066 - accuracy: 0.6057\n",
            "Epoch 4590/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0066 - accuracy: 0.6039\n",
            "Epoch 4591/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0066 - accuracy: 0.6102\n",
            "Epoch 4592/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0066 - accuracy: 0.6162\n",
            "Epoch 4593/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0066 - accuracy: 0.6143\n",
            "Epoch 4594/5000\n",
            "1/1 [==============================] - 0s 298ms/step - loss: 0.0066 - accuracy: 0.6228\n",
            "Epoch 4595/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0066 - accuracy: 0.6016\n",
            "Epoch 4596/5000\n",
            "1/1 [==============================] - 0s 260ms/step - loss: 0.0066 - accuracy: 0.6177\n",
            "Epoch 4597/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0066 - accuracy: 0.6013\n",
            "Epoch 4598/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0066 - accuracy: 0.6173\n",
            "Epoch 4599/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0067 - accuracy: 0.6072\n",
            "Epoch 4600/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0067 - accuracy: 0.6104\n",
            "Epoch 4601/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0067 - accuracy: 0.6016\n",
            "Epoch 4602/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0067 - accuracy: 0.6106\n",
            "Epoch 4603/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0067 - accuracy: 0.6039\n",
            "Epoch 4604/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0066 - accuracy: 0.6230\n",
            "Epoch 4605/5000\n",
            "1/1 [==============================] - 0s 291ms/step - loss: 0.0067 - accuracy: 0.6045\n",
            "Epoch 4606/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0067 - accuracy: 0.6220\n",
            "Epoch 4607/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0066 - accuracy: 0.5973\n",
            "Epoch 4608/5000\n",
            "1/1 [==============================] - 0s 297ms/step - loss: 0.0066 - accuracy: 0.6043\n",
            "Epoch 4609/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0066 - accuracy: 0.5991\n",
            "Epoch 4610/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0066 - accuracy: 0.6041\n",
            "Epoch 4611/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0066 - accuracy: 0.6218\n",
            "Epoch 4612/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0066 - accuracy: 0.6153\n",
            "Epoch 4613/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0066 - accuracy: 0.6290\n",
            "Epoch 4614/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0066 - accuracy: 0.6103\n",
            "Epoch 4615/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0066 - accuracy: 0.6081\n",
            "Epoch 4616/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0066 - accuracy: 0.5934\n",
            "Epoch 4617/5000\n",
            "1/1 [==============================] - 0s 257ms/step - loss: 0.0067 - accuracy: 0.5993\n",
            "Epoch 4618/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0066 - accuracy: 0.6015\n",
            "Epoch 4619/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0066 - accuracy: 0.6174\n",
            "Epoch 4620/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0066 - accuracy: 0.6172\n",
            "Epoch 4621/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0066 - accuracy: 0.6252\n",
            "Epoch 4622/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0066 - accuracy: 0.6016\n",
            "Epoch 4623/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0066 - accuracy: 0.6143\n",
            "Epoch 4624/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0066 - accuracy: 0.5943\n",
            "Epoch 4625/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0066 - accuracy: 0.6139\n",
            "Epoch 4626/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0066 - accuracy: 0.6096\n",
            "Epoch 4627/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0066 - accuracy: 0.6123\n",
            "Epoch 4628/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0066 - accuracy: 0.6198\n",
            "Epoch 4629/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0066 - accuracy: 0.6076\n",
            "Epoch 4630/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0066 - accuracy: 0.6167\n",
            "Epoch 4631/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0066 - accuracy: 0.6075\n",
            "Epoch 4632/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0066 - accuracy: 0.6087\n",
            "Epoch 4633/5000\n",
            "1/1 [==============================] - 0s 290ms/step - loss: 0.0066 - accuracy: 0.6093\n",
            "Epoch 4634/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0066 - accuracy: 0.6051\n",
            "Epoch 4635/5000\n",
            "1/1 [==============================] - 0s 261ms/step - loss: 0.0066 - accuracy: 0.6070\n",
            "Epoch 4636/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0066 - accuracy: 0.6082\n",
            "Epoch 4637/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0066 - accuracy: 0.6072\n",
            "Epoch 4638/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0066 - accuracy: 0.6187\n",
            "Epoch 4639/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0066 - accuracy: 0.6048\n",
            "Epoch 4640/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0066 - accuracy: 0.6212\n",
            "Epoch 4641/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0066 - accuracy: 0.6032\n",
            "Epoch 4642/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0066 - accuracy: 0.6182\n",
            "Epoch 4643/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0067 - accuracy: 0.6047\n",
            "Epoch 4644/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0067 - accuracy: 0.6154\n",
            "Epoch 4645/5000\n",
            "1/1 [==============================] - 0s 305ms/step - loss: 0.0068 - accuracy: 0.6045\n",
            "Epoch 4646/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0068 - accuracy: 0.6138\n",
            "Epoch 4647/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0067 - accuracy: 0.5996\n",
            "Epoch 4648/5000\n",
            "1/1 [==============================] - 0s 260ms/step - loss: 0.0068 - accuracy: 0.6226\n",
            "Epoch 4649/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0068 - accuracy: 0.5970\n",
            "Epoch 4650/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0068 - accuracy: 0.6292\n",
            "Epoch 4651/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0067 - accuracy: 0.5960\n",
            "Epoch 4652/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0066 - accuracy: 0.6065\n",
            "Epoch 4653/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0066 - accuracy: 0.6056\n",
            "Epoch 4654/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0066 - accuracy: 0.5956\n",
            "Epoch 4655/5000\n",
            "1/1 [==============================] - 0s 293ms/step - loss: 0.0066 - accuracy: 0.6190\n",
            "Epoch 4656/5000\n",
            "1/1 [==============================] - 0s 292ms/step - loss: 0.0066 - accuracy: 0.6179\n",
            "Epoch 4657/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0066 - accuracy: 0.6208\n",
            "Epoch 4658/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0067 - accuracy: 0.6158\n",
            "Epoch 4659/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0067 - accuracy: 0.6064\n",
            "Epoch 4660/5000\n",
            "1/1 [==============================] - 0s 290ms/step - loss: 0.0066 - accuracy: 0.5935\n",
            "Epoch 4661/5000\n",
            "1/1 [==============================] - 0s 304ms/step - loss: 0.0066 - accuracy: 0.6051\n",
            "Epoch 4662/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0066 - accuracy: 0.6013\n",
            "Epoch 4663/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0066 - accuracy: 0.6237\n",
            "Epoch 4664/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0066 - accuracy: 0.6153\n",
            "Epoch 4665/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0065 - accuracy: 0.6160\n",
            "Epoch 4666/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0065 - accuracy: 0.6064\n",
            "Epoch 4667/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0066 - accuracy: 0.5974\n",
            "Epoch 4668/5000\n",
            "1/1 [==============================] - 0s 304ms/step - loss: 0.0066 - accuracy: 0.6087\n",
            "Epoch 4669/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0066 - accuracy: 0.6059\n",
            "Epoch 4670/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0066 - accuracy: 0.6224\n",
            "Epoch 4671/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0066 - accuracy: 0.6154\n",
            "Epoch 4672/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0066 - accuracy: 0.6183\n",
            "Epoch 4673/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0066 - accuracy: 0.6065\n",
            "Epoch 4674/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0066 - accuracy: 0.6092\n",
            "Epoch 4675/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0066 - accuracy: 0.6008\n",
            "Epoch 4676/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0066 - accuracy: 0.6141\n",
            "Epoch 4677/5000\n",
            "1/1 [==============================] - 0s 256ms/step - loss: 0.0066 - accuracy: 0.6062\n",
            "Epoch 4678/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0066 - accuracy: 0.6216\n",
            "Epoch 4679/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0066 - accuracy: 0.6021\n",
            "Epoch 4680/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0066 - accuracy: 0.6159\n",
            "Epoch 4681/5000\n",
            "1/1 [==============================] - 0s 306ms/step - loss: 0.0066 - accuracy: 0.5967\n",
            "Epoch 4682/5000\n",
            "1/1 [==============================] - 0s 296ms/step - loss: 0.0066 - accuracy: 0.6136\n",
            "Epoch 4683/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0066 - accuracy: 0.6044\n",
            "Epoch 4684/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0066 - accuracy: 0.6153\n",
            "Epoch 4685/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0066 - accuracy: 0.6128\n",
            "Epoch 4686/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0066 - accuracy: 0.6145\n",
            "Epoch 4687/5000\n",
            "1/1 [==============================] - 0s 293ms/step - loss: 0.0066 - accuracy: 0.6105\n",
            "Epoch 4688/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0065 - accuracy: 0.6102\n",
            "Epoch 4689/5000\n",
            "1/1 [==============================] - 0s 253ms/step - loss: 0.0065 - accuracy: 0.6060\n",
            "Epoch 4690/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0065 - accuracy: 0.6112\n",
            "Epoch 4691/5000\n",
            "1/1 [==============================] - 0s 259ms/step - loss: 0.0065 - accuracy: 0.6067\n",
            "Epoch 4692/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0066 - accuracy: 0.6134\n",
            "Epoch 4693/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0066 - accuracy: 0.6056\n",
            "Epoch 4694/5000\n",
            "1/1 [==============================] - 0s 291ms/step - loss: 0.0066 - accuracy: 0.6155\n",
            "Epoch 4695/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0065 - accuracy: 0.6092\n",
            "Epoch 4696/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0065 - accuracy: 0.6136\n",
            "Epoch 4697/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0065 - accuracy: 0.6074\n",
            "Epoch 4698/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0065 - accuracy: 0.6104\n",
            "Epoch 4699/5000\n",
            "1/1 [==============================] - 0s 257ms/step - loss: 0.0066 - accuracy: 0.6069\n",
            "Epoch 4700/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0066 - accuracy: 0.6128\n",
            "Epoch 4701/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0066 - accuracy: 0.6079\n",
            "Epoch 4702/5000\n",
            "1/1 [==============================] - 0s 261ms/step - loss: 0.0066 - accuracy: 0.6182\n",
            "Epoch 4703/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0067 - accuracy: 0.6004\n",
            "Epoch 4704/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0068 - accuracy: 0.6210\n",
            "Epoch 4705/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0068 - accuracy: 0.5911\n",
            "Epoch 4706/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0069 - accuracy: 0.6249\n",
            "Epoch 4707/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0068 - accuracy: 0.5963\n",
            "Epoch 4708/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0068 - accuracy: 0.6204\n",
            "Epoch 4709/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0067 - accuracy: 0.6052\n",
            "Epoch 4710/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0066 - accuracy: 0.6063\n",
            "Epoch 4711/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0066 - accuracy: 0.6146\n",
            "Epoch 4712/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0066 - accuracy: 0.6044\n",
            "Epoch 4713/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0066 - accuracy: 0.6232\n",
            "Epoch 4714/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0067 - accuracy: 0.6121\n",
            "Epoch 4715/5000\n",
            "1/1 [==============================] - 0s 261ms/step - loss: 0.0067 - accuracy: 0.6210\n",
            "Epoch 4716/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0068 - accuracy: 0.5983\n",
            "Epoch 4717/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0067 - accuracy: 0.6053\n",
            "Epoch 4718/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0066 - accuracy: 0.5899\n",
            "Epoch 4719/5000\n",
            "1/1 [==============================] - 0s 298ms/step - loss: 0.0066 - accuracy: 0.6088\n",
            "Epoch 4720/5000\n",
            "1/1 [==============================] - 0s 259ms/step - loss: 0.0065 - accuracy: 0.6103\n",
            "Epoch 4721/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0065 - accuracy: 0.6189\n",
            "Epoch 4722/5000\n",
            "1/1 [==============================] - 0s 261ms/step - loss: 0.0066 - accuracy: 0.6215\n",
            "Epoch 4723/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0066 - accuracy: 0.6071\n",
            "Epoch 4724/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0067 - accuracy: 0.6165\n",
            "Epoch 4725/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0067 - accuracy: 0.5920\n",
            "Epoch 4726/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0067 - accuracy: 0.6164\n",
            "Epoch 4727/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0067 - accuracy: 0.5990\n",
            "Epoch 4728/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0066 - accuracy: 0.6217\n",
            "Epoch 4729/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0065 - accuracy: 0.6140\n",
            "Epoch 4730/5000\n",
            "1/1 [==============================] - 0s 292ms/step - loss: 0.0065 - accuracy: 0.6071\n",
            "Epoch 4731/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0066 - accuracy: 0.6132\n",
            "Epoch 4732/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0066 - accuracy: 0.5959\n",
            "Epoch 4733/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0066 - accuracy: 0.6182\n",
            "Epoch 4734/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0066 - accuracy: 0.6062\n",
            "Epoch 4735/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0066 - accuracy: 0.6249\n",
            "Epoch 4736/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0066 - accuracy: 0.6069\n",
            "Epoch 4737/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0066 - accuracy: 0.6026\n",
            "Epoch 4738/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0065 - accuracy: 0.5987\n",
            "Epoch 4739/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0065 - accuracy: 0.5941\n",
            "Epoch 4740/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0066 - accuracy: 0.6160\n",
            "Epoch 4741/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0066 - accuracy: 0.6167\n",
            "Epoch 4742/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0066 - accuracy: 0.6291\n",
            "Epoch 4743/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0066 - accuracy: 0.6183\n",
            "Epoch 4744/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0066 - accuracy: 0.6081\n",
            "Epoch 4745/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0066 - accuracy: 0.5946\n",
            "Epoch 4746/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0065 - accuracy: 0.5965\n",
            "Epoch 4747/5000\n",
            "1/1 [==============================] - 0s 290ms/step - loss: 0.0065 - accuracy: 0.6062\n",
            "Epoch 4748/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0065 - accuracy: 0.6188\n",
            "Epoch 4749/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0065 - accuracy: 0.6223\n",
            "Epoch 4750/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0065 - accuracy: 0.6175\n",
            "Epoch 4751/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0065 - accuracy: 0.6056\n",
            "Epoch 4752/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0065 - accuracy: 0.5998\n",
            "Epoch 4753/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0065 - accuracy: 0.6013\n",
            "Epoch 4754/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0065 - accuracy: 0.6090\n",
            "Epoch 4755/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0065 - accuracy: 0.6129\n",
            "Epoch 4756/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0065 - accuracy: 0.6175\n",
            "Epoch 4757/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0065 - accuracy: 0.6061\n",
            "Epoch 4758/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0066 - accuracy: 0.6123\n",
            "Epoch 4759/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0066 - accuracy: 0.5959\n",
            "Epoch 4760/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0067 - accuracy: 0.6215\n",
            "Epoch 4761/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0068 - accuracy: 0.5989\n",
            "Epoch 4762/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0068 - accuracy: 0.6312\n",
            "Epoch 4763/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0068 - accuracy: 0.5985\n",
            "Epoch 4764/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0067 - accuracy: 0.6148\n",
            "Epoch 4765/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0066 - accuracy: 0.6002\n",
            "Epoch 4766/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0066 - accuracy: 0.5963\n",
            "Epoch 4767/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0066 - accuracy: 0.6135\n",
            "Epoch 4768/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0066 - accuracy: 0.6080\n",
            "Epoch 4769/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0066 - accuracy: 0.6300\n",
            "Epoch 4770/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0067 - accuracy: 0.6185\n",
            "Epoch 4771/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0067 - accuracy: 0.6162\n",
            "Epoch 4772/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0067 - accuracy: 0.5948\n",
            "Epoch 4773/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0066 - accuracy: 0.5951\n",
            "Epoch 4774/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0065 - accuracy: 0.5916\n",
            "Epoch 4775/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0065 - accuracy: 0.6109\n",
            "Epoch 4776/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0065 - accuracy: 0.6195\n",
            "Epoch 4777/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0066 - accuracy: 0.6279\n",
            "Epoch 4778/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0066 - accuracy: 0.6174\n",
            "Epoch 4779/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0066 - accuracy: 0.6033\n",
            "Epoch 4780/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0066 - accuracy: 0.6052\n",
            "Epoch 4781/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0066 - accuracy: 0.5937\n",
            "Epoch 4782/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0066 - accuracy: 0.6167\n",
            "Epoch 4783/5000\n",
            "1/1 [==============================] - 0s 296ms/step - loss: 0.0066 - accuracy: 0.6065\n",
            "Epoch 4784/5000\n",
            "1/1 [==============================] - 0s 298ms/step - loss: 0.0066 - accuracy: 0.6259\n",
            "Epoch 4785/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0065 - accuracy: 0.6135\n",
            "Epoch 4786/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0065 - accuracy: 0.6062\n",
            "Epoch 4787/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0065 - accuracy: 0.6069\n",
            "Epoch 4788/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0065 - accuracy: 0.5962\n",
            "Epoch 4789/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0066 - accuracy: 0.6138\n",
            "Epoch 4790/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0066 - accuracy: 0.6082\n",
            "Epoch 4791/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0065 - accuracy: 0.6185\n",
            "Epoch 4792/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0065 - accuracy: 0.6083\n",
            "Epoch 4793/5000\n",
            "1/1 [==============================] - 0s 300ms/step - loss: 0.0065 - accuracy: 0.6071\n",
            "Epoch 4794/5000\n",
            "1/1 [==============================] - 0s 297ms/step - loss: 0.0065 - accuracy: 0.6011\n",
            "Epoch 4795/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0065 - accuracy: 0.6041\n",
            "Epoch 4796/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0065 - accuracy: 0.6092\n",
            "Epoch 4797/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0065 - accuracy: 0.6141\n",
            "Epoch 4798/5000\n",
            "1/1 [==============================] - 0s 291ms/step - loss: 0.0065 - accuracy: 0.6207\n",
            "Epoch 4799/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0065 - accuracy: 0.6148\n",
            "Epoch 4800/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0065 - accuracy: 0.6140\n",
            "Epoch 4801/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0065 - accuracy: 0.5966\n",
            "Epoch 4802/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0066 - accuracy: 0.6030\n",
            "Epoch 4803/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0066 - accuracy: 0.5920\n",
            "Epoch 4804/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0066 - accuracy: 0.6223\n",
            "Epoch 4805/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0067 - accuracy: 0.6080\n",
            "Epoch 4806/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0066 - accuracy: 0.6340\n",
            "Epoch 4807/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0066 - accuracy: 0.6100\n",
            "Epoch 4808/5000\n",
            "1/1 [==============================] - 0s 293ms/step - loss: 0.0065 - accuracy: 0.6086\n",
            "Epoch 4809/5000\n",
            "1/1 [==============================] - 0s 296ms/step - loss: 0.0065 - accuracy: 0.5966\n",
            "Epoch 4810/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0065 - accuracy: 0.5946\n",
            "Epoch 4811/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0065 - accuracy: 0.6120\n",
            "Epoch 4812/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0065 - accuracy: 0.6121\n",
            "Epoch 4813/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0065 - accuracy: 0.6283\n",
            "Epoch 4814/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0066 - accuracy: 0.6100\n",
            "Epoch 4815/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0066 - accuracy: 0.6079\n",
            "Epoch 4816/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0066 - accuracy: 0.5897\n",
            "Epoch 4817/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0066 - accuracy: 0.6002\n",
            "Epoch 4818/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0066 - accuracy: 0.6056\n",
            "Epoch 4819/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0065 - accuracy: 0.6212\n",
            "Epoch 4820/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0065 - accuracy: 0.6227\n",
            "Epoch 4821/5000\n",
            "1/1 [==============================] - 0s 291ms/step - loss: 0.0065 - accuracy: 0.6214\n",
            "Epoch 4822/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0065 - accuracy: 0.6058\n",
            "Epoch 4823/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0065 - accuracy: 0.5983\n",
            "Epoch 4824/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0065 - accuracy: 0.5935\n",
            "Epoch 4825/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0065 - accuracy: 0.6070\n",
            "Epoch 4826/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0065 - accuracy: 0.6165\n",
            "Epoch 4827/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0065 - accuracy: 0.6251\n",
            "Epoch 4828/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0065 - accuracy: 0.6239\n",
            "Epoch 4829/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0065 - accuracy: 0.6100\n",
            "Epoch 4830/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0065 - accuracy: 0.5997\n",
            "Epoch 4831/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0065 - accuracy: 0.5937\n",
            "Epoch 4832/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0065 - accuracy: 0.6007\n",
            "Epoch 4833/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0065 - accuracy: 0.6111\n",
            "Epoch 4834/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0065 - accuracy: 0.6205\n",
            "Epoch 4835/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0065 - accuracy: 0.6185\n",
            "Epoch 4836/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0065 - accuracy: 0.6123\n",
            "Epoch 4837/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0065 - accuracy: 0.5980\n",
            "Epoch 4838/5000\n",
            "1/1 [==============================] - 0s 296ms/step - loss: 0.0065 - accuracy: 0.6073\n",
            "Epoch 4839/5000\n",
            "1/1 [==============================] - 0s 292ms/step - loss: 0.0066 - accuracy: 0.5959\n",
            "Epoch 4840/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0067 - accuracy: 0.6302\n",
            "Epoch 4841/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0068 - accuracy: 0.6028\n",
            "Epoch 4842/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0069 - accuracy: 0.6321\n",
            "Epoch 4843/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0069 - accuracy: 0.5885\n",
            "Epoch 4844/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0068 - accuracy: 0.6070\n",
            "Epoch 4845/5000\n",
            "1/1 [==============================] - 0s 302ms/step - loss: 0.0067 - accuracy: 0.5951\n",
            "Epoch 4846/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0066 - accuracy: 0.6014\n",
            "Epoch 4847/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0065 - accuracy: 0.6184\n",
            "Epoch 4848/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0065 - accuracy: 0.6172\n",
            "Epoch 4849/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0066 - accuracy: 0.6245\n",
            "Epoch 4850/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0067 - accuracy: 0.6112\n",
            "Epoch 4851/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0068 - accuracy: 0.6115\n",
            "Epoch 4852/5000\n",
            "1/1 [==============================] - 0s 294ms/step - loss: 0.0067 - accuracy: 0.5878\n",
            "Epoch 4853/5000\n",
            "1/1 [==============================] - 0s 291ms/step - loss: 0.0066 - accuracy: 0.6043\n",
            "Epoch 4854/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0066 - accuracy: 0.5935\n",
            "Epoch 4855/5000\n",
            "1/1 [==============================] - 0s 291ms/step - loss: 0.0065 - accuracy: 0.6158\n",
            "Epoch 4856/5000\n",
            "1/1 [==============================] - 0s 292ms/step - loss: 0.0066 - accuracy: 0.6177\n",
            "Epoch 4857/5000\n",
            "1/1 [==============================] - 0s 295ms/step - loss: 0.0066 - accuracy: 0.6112\n",
            "Epoch 4858/5000\n",
            "1/1 [==============================] - 0s 297ms/step - loss: 0.0066 - accuracy: 0.6196\n",
            "Epoch 4859/5000\n",
            "1/1 [==============================] - 0s 292ms/step - loss: 0.0066 - accuracy: 0.5988\n",
            "Epoch 4860/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0066 - accuracy: 0.6157\n",
            "Epoch 4861/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0066 - accuracy: 0.5972\n",
            "Epoch 4862/5000\n",
            "1/1 [==============================] - 0s 291ms/step - loss: 0.0065 - accuracy: 0.6156\n",
            "Epoch 4863/5000\n",
            "1/1 [==============================] - 0s 286ms/step - loss: 0.0065 - accuracy: 0.6097\n",
            "Epoch 4864/5000\n",
            "1/1 [==============================] - 0s 290ms/step - loss: 0.0066 - accuracy: 0.6067\n",
            "Epoch 4865/5000\n",
            "1/1 [==============================] - 0s 292ms/step - loss: 0.0065 - accuracy: 0.6158\n",
            "Epoch 4866/5000\n",
            "1/1 [==============================] - 0s 301ms/step - loss: 0.0065 - accuracy: 0.6009\n",
            "Epoch 4867/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0066 - accuracy: 0.6183\n",
            "Epoch 4868/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0065 - accuracy: 0.6070\n",
            "Epoch 4869/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0065 - accuracy: 0.6160\n",
            "Epoch 4870/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0065 - accuracy: 0.6080\n",
            "Epoch 4871/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0065 - accuracy: 0.6003\n",
            "Epoch 4872/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0065 - accuracy: 0.5990\n",
            "Epoch 4873/5000\n",
            "1/1 [==============================] - 0s 294ms/step - loss: 0.0065 - accuracy: 0.5996\n",
            "Epoch 4874/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0065 - accuracy: 0.6109\n",
            "Epoch 4875/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0065 - accuracy: 0.6199\n",
            "Epoch 4876/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0065 - accuracy: 0.6205\n",
            "Epoch 4877/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0065 - accuracy: 0.6145\n",
            "Epoch 4878/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0065 - accuracy: 0.6105\n",
            "Epoch 4879/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0065 - accuracy: 0.6032\n",
            "Epoch 4880/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0065 - accuracy: 0.6071\n",
            "Epoch 4881/5000\n",
            "1/1 [==============================] - 0s 294ms/step - loss: 0.0065 - accuracy: 0.6052\n",
            "Epoch 4882/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0065 - accuracy: 0.6099\n",
            "Epoch 4883/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0065 - accuracy: 0.6086\n",
            "Epoch 4884/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0065 - accuracy: 0.6114\n",
            "Epoch 4885/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0065 - accuracy: 0.6060\n",
            "Epoch 4886/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0065 - accuracy: 0.6127\n",
            "Epoch 4887/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0065 - accuracy: 0.6056\n",
            "Epoch 4888/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0065 - accuracy: 0.6169\n",
            "Epoch 4889/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0065 - accuracy: 0.6026\n",
            "Epoch 4890/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0066 - accuracy: 0.6169\n",
            "Epoch 4891/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0066 - accuracy: 0.5988\n",
            "Epoch 4892/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0066 - accuracy: 0.6179\n",
            "Epoch 4893/5000\n",
            "1/1 [==============================] - 0s 261ms/step - loss: 0.0066 - accuracy: 0.5987\n",
            "Epoch 4894/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0066 - accuracy: 0.6193\n",
            "Epoch 4895/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0066 - accuracy: 0.6030\n",
            "Epoch 4896/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0065 - accuracy: 0.6139\n",
            "Epoch 4897/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0065 - accuracy: 0.6065\n",
            "Epoch 4898/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0065 - accuracy: 0.6070\n",
            "Epoch 4899/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0065 - accuracy: 0.6152\n",
            "Epoch 4900/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0065 - accuracy: 0.6070\n",
            "Epoch 4901/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0065 - accuracy: 0.6199\n",
            "Epoch 4902/5000\n",
            "1/1 [==============================] - 0s 269ms/step - loss: 0.0066 - accuracy: 0.6028\n",
            "Epoch 4903/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0066 - accuracy: 0.6149\n",
            "Epoch 4904/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0066 - accuracy: 0.5941\n",
            "Epoch 4905/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0066 - accuracy: 0.6143\n",
            "Epoch 4906/5000\n",
            "1/1 [==============================] - 0s 293ms/step - loss: 0.0066 - accuracy: 0.6006\n",
            "Epoch 4907/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0065 - accuracy: 0.6238\n",
            "Epoch 4908/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0065 - accuracy: 0.6098\n",
            "Epoch 4909/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0065 - accuracy: 0.6169\n",
            "Epoch 4910/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0065 - accuracy: 0.6055\n",
            "Epoch 4911/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0065 - accuracy: 0.6011\n",
            "Epoch 4912/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0065 - accuracy: 0.6105\n",
            "Epoch 4913/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0065 - accuracy: 0.6062\n",
            "Epoch 4914/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0065 - accuracy: 0.6241\n",
            "Epoch 4915/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0065 - accuracy: 0.6074\n",
            "Epoch 4916/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0065 - accuracy: 0.6145\n",
            "Epoch 4917/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0065 - accuracy: 0.5950\n",
            "Epoch 4918/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0065 - accuracy: 0.6005\n",
            "Epoch 4919/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0065 - accuracy: 0.6036\n",
            "Epoch 4920/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0065 - accuracy: 0.6134\n",
            "Epoch 4921/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0065 - accuracy: 0.6245\n",
            "Epoch 4922/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0065 - accuracy: 0.6175\n",
            "Epoch 4923/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0065 - accuracy: 0.6165\n",
            "Epoch 4924/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0065 - accuracy: 0.5984\n",
            "Epoch 4925/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0065 - accuracy: 0.6015\n",
            "Epoch 4926/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0065 - accuracy: 0.5958\n",
            "Epoch 4927/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0065 - accuracy: 0.6154\n",
            "Epoch 4928/5000\n",
            "1/1 [==============================] - 0s 263ms/step - loss: 0.0065 - accuracy: 0.6101\n",
            "Epoch 4929/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0065 - accuracy: 0.6297\n",
            "Epoch 4930/5000\n",
            "1/1 [==============================] - 0s 276ms/step - loss: 0.0066 - accuracy: 0.6053\n",
            "Epoch 4931/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0066 - accuracy: 0.6182\n",
            "Epoch 4932/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0066 - accuracy: 0.5915\n",
            "Epoch 4933/5000\n",
            "1/1 [==============================] - 0s 290ms/step - loss: 0.0065 - accuracy: 0.6066\n",
            "Epoch 4934/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0065 - accuracy: 0.6013\n",
            "Epoch 4935/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0065 - accuracy: 0.6077\n",
            "Epoch 4936/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0065 - accuracy: 0.6183\n",
            "Epoch 4937/5000\n",
            "1/1 [==============================] - 0s 275ms/step - loss: 0.0065 - accuracy: 0.6108\n",
            "Epoch 4938/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0065 - accuracy: 0.6201\n",
            "Epoch 4939/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0065 - accuracy: 0.6073\n",
            "Epoch 4940/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0065 - accuracy: 0.6100\n",
            "Epoch 4941/5000\n",
            "1/1 [==============================] - 0s 290ms/step - loss: 0.0065 - accuracy: 0.6032\n",
            "Epoch 4942/5000\n",
            "1/1 [==============================] - 0s 299ms/step - loss: 0.0065 - accuracy: 0.6017\n",
            "Epoch 4943/5000\n",
            "1/1 [==============================] - 0s 293ms/step - loss: 0.0065 - accuracy: 0.6041\n",
            "Epoch 4944/5000\n",
            "1/1 [==============================] - 0s 284ms/step - loss: 0.0065 - accuracy: 0.6078\n",
            "Epoch 4945/5000\n",
            "1/1 [==============================] - 0s 298ms/step - loss: 0.0065 - accuracy: 0.6115\n",
            "Epoch 4946/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0065 - accuracy: 0.6167\n",
            "Epoch 4947/5000\n",
            "1/1 [==============================] - 0s 287ms/step - loss: 0.0065 - accuracy: 0.6093\n",
            "Epoch 4948/5000\n",
            "1/1 [==============================] - 0s 292ms/step - loss: 0.0065 - accuracy: 0.6144\n",
            "Epoch 4949/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0065 - accuracy: 0.6014\n",
            "Epoch 4950/5000\n",
            "1/1 [==============================] - 0s 270ms/step - loss: 0.0065 - accuracy: 0.6120\n",
            "Epoch 4951/5000\n",
            "1/1 [==============================] - 0s 292ms/step - loss: 0.0065 - accuracy: 0.6002\n",
            "Epoch 4952/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0065 - accuracy: 0.6201\n",
            "Epoch 4953/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0066 - accuracy: 0.6038\n",
            "Epoch 4954/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0065 - accuracy: 0.6227\n",
            "Epoch 4955/5000\n",
            "1/1 [==============================] - 0s 293ms/step - loss: 0.0065 - accuracy: 0.6043\n",
            "Epoch 4956/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0065 - accuracy: 0.6096\n",
            "Epoch 4957/5000\n",
            "1/1 [==============================] - 0s 278ms/step - loss: 0.0065 - accuracy: 0.6042\n",
            "Epoch 4958/5000\n",
            "1/1 [==============================] - 0s 264ms/step - loss: 0.0066 - accuracy: 0.6034\n",
            "Epoch 4959/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0065 - accuracy: 0.6092\n",
            "Epoch 4960/5000\n",
            "1/1 [==============================] - 0s 262ms/step - loss: 0.0065 - accuracy: 0.6098\n",
            "Epoch 4961/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0065 - accuracy: 0.6132\n",
            "Epoch 4962/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0065 - accuracy: 0.6154\n",
            "Epoch 4963/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0065 - accuracy: 0.6053\n",
            "Epoch 4964/5000\n",
            "1/1 [==============================] - 0s 273ms/step - loss: 0.0065 - accuracy: 0.6055\n",
            "Epoch 4965/5000\n",
            "1/1 [==============================] - 0s 261ms/step - loss: 0.0065 - accuracy: 0.5976\n",
            "Epoch 4966/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0065 - accuracy: 0.6075\n",
            "Epoch 4967/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0065 - accuracy: 0.6091\n",
            "Epoch 4968/5000\n",
            "1/1 [==============================] - 0s 272ms/step - loss: 0.0065 - accuracy: 0.6186\n",
            "Epoch 4969/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0065 - accuracy: 0.6135\n",
            "Epoch 4970/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0065 - accuracy: 0.6097\n",
            "Epoch 4971/5000\n",
            "1/1 [==============================] - 0s 268ms/step - loss: 0.0065 - accuracy: 0.6044\n",
            "Epoch 4972/5000\n",
            "1/1 [==============================] - 0s 256ms/step - loss: 0.0065 - accuracy: 0.5998\n",
            "Epoch 4973/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0065 - accuracy: 0.6117\n",
            "Epoch 4974/5000\n",
            "1/1 [==============================] - 0s 282ms/step - loss: 0.0065 - accuracy: 0.6118\n",
            "Epoch 4975/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0065 - accuracy: 0.6216\n",
            "Epoch 4976/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0065 - accuracy: 0.6040\n",
            "Epoch 4977/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0066 - accuracy: 0.6142\n",
            "Epoch 4978/5000\n",
            "1/1 [==============================] - 0s 261ms/step - loss: 0.0067 - accuracy: 0.5891\n",
            "Epoch 4979/5000\n",
            "1/1 [==============================] - 0s 253ms/step - loss: 0.0067 - accuracy: 0.6225\n",
            "Epoch 4980/5000\n",
            "1/1 [==============================] - 0s 277ms/step - loss: 0.0068 - accuracy: 0.5957\n",
            "Epoch 4981/5000\n",
            "1/1 [==============================] - 0s 265ms/step - loss: 0.0069 - accuracy: 0.6321\n",
            "Epoch 4982/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0069 - accuracy: 0.6001\n",
            "Epoch 4983/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0068 - accuracy: 0.6103\n",
            "Epoch 4984/5000\n",
            "1/1 [==============================] - 0s 285ms/step - loss: 0.0067 - accuracy: 0.5950\n",
            "Epoch 4985/5000\n",
            "1/1 [==============================] - 0s 266ms/step - loss: 0.0066 - accuracy: 0.5941\n",
            "Epoch 4986/5000\n",
            "1/1 [==============================] - 0s 283ms/step - loss: 0.0065 - accuracy: 0.6048\n",
            "Epoch 4987/5000\n",
            "1/1 [==============================] - 0s 271ms/step - loss: 0.0065 - accuracy: 0.6184\n",
            "Epoch 4988/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0066 - accuracy: 0.6221\n",
            "Epoch 4989/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0066 - accuracy: 0.6197\n",
            "Epoch 4990/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0066 - accuracy: 0.6161\n",
            "Epoch 4991/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0066 - accuracy: 0.5877\n",
            "Epoch 4992/5000\n",
            "1/1 [==============================] - 0s 281ms/step - loss: 0.0066 - accuracy: 0.6048\n",
            "Epoch 4993/5000\n",
            "1/1 [==============================] - 0s 274ms/step - loss: 0.0066 - accuracy: 0.5887\n",
            "Epoch 4994/5000\n",
            "1/1 [==============================] - 0s 297ms/step - loss: 0.0065 - accuracy: 0.6171\n",
            "Epoch 4995/5000\n",
            "1/1 [==============================] - 0s 288ms/step - loss: 0.0065 - accuracy: 0.6138\n",
            "Epoch 4996/5000\n",
            "1/1 [==============================] - 0s 289ms/step - loss: 0.0065 - accuracy: 0.6145\n",
            "Epoch 4997/5000\n",
            "1/1 [==============================] - 0s 279ms/step - loss: 0.0065 - accuracy: 0.6157\n",
            "Epoch 4998/5000\n",
            "1/1 [==============================] - 0s 304ms/step - loss: 0.0065 - accuracy: 0.5990\n",
            "Epoch 4999/5000\n",
            "1/1 [==============================] - 0s 280ms/step - loss: 0.0066 - accuracy: 0.6181\n",
            "Epoch 5000/5000\n",
            "1/1 [==============================] - 0s 267ms/step - loss: 0.0067 - accuracy: 0.6006\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7f8595c45d90>"
            ]
          },
          "metadata": {},
          "execution_count": 11
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 323
        },
        "id": "EH0QY2PE6gps",
        "outputId": "81c7247d-d2ec-43b4-f592-f49a86389f29"
      },
      "source": [
        "output = model.predict(img_array)\n",
        "\n",
        "\n",
        "imshow(output[0].reshape(SIZE,SIZE,3), cmap=\"gray\")"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7f8593fa28d0>"
            ]
          },
          "metadata": {},
          "execution_count": 12
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    }
  ]
}