Codes/ACP-740-ROC.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "ACP-740-ROC.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "authorship_tag": "ABX9TyMer3u9MikPd/blfrAGJP71",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/mrzResearchArena/Anticancer-Peptides-CNN/blob/master/ACP-740-ROC.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0llsRkjoGzFk",
        "colab_type": "text"
      },
      "source": [
        "### [1]. Google Drive Initialization:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8EazYO0pGwdP",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 128
        },
        "outputId": "a51c9379-9925-4518-a643-e0f06e501a86"
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Go to this URL in a browser: https://accounts.google.com/o/oauth2/auth?client_id=947318989803-6bn6qk8qdgf4n4g3pfee6491hc0brc4i.apps.googleusercontent.com&redirect_uri=urn%3aietf%3awg%3aoauth%3a2.0%3aoob&scope=email%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdocs.test%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive.photos.readonly%20https%3a%2f%2fwww.googleapis.com%2fauth%2fpeopleapi.readonly&response_type=code\n",
            "\n",
            "Enter your authorization code:\n",
            "··········\n",
            "Mounted at /content/drive\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wV07T6tHKHQm",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "02fb53ad-f82b-417c-d5e2-5c5dc3646fad"
      },
      "source": [
        "ls"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\u001b[0m\u001b[01;34mdrive\u001b[0m/  \u001b[01;34msample_data\u001b[0m/\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cypUjLtbJ10d",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "613cc2ac-760b-478d-872f-4c9e53d361a6"
      },
      "source": [
        "cd 'drive/My Drive/ACP-PyFeat'"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/content/drive/My Drive/ACP-PyFeat\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2H7nRPtkKNAq",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 127
        },
        "outputId": "2957b0ae-dda2-4a9d-ff37-6b7d3dce4e3f"
      },
      "source": [
        "ls"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "bits-164.npy    blosum-500.npy  model-240.png      PyFeat-500-20.npy\n",
            "bits-240.npy    blosum-740.npy  model-740.png      PyFeat-500.npy\n",
            "bits-500.npy    bpf-164.npy     PyFeat-164-20.npy  PyFeat-740-20.npy\n",
            "bits-740.npy    bpf-240.npy     PyFeat-164.npy     PyFeat-740.npy\n",
            "blosum-164.npy  bpf-500.npy     PyFeat-240-20.npy  ROC-240.png\n",
            "blosum-240.npy  bpf-740.npy     PyFeat-240.npy     ROC.png\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0MiRDXCXHQmr",
        "colab_type": "text"
      },
      "source": [
        "### [2]. Load Deep Learning API (Keras):"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "60Y5muxdCr4I",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# # Initialize TF-2.x:\n",
        "# try:\n",
        "#     %tensorflow_version 2.x  # Colab only.\n",
        "# except Exception:\n",
        "#     pass\n",
        "\n",
        "# import tensorflow as tf\n",
        "# print('We\\'re using TF-{}.'.format(tf.__version__))\n",
        "\n",
        "# # Colab Auto Click: [https://paste.ubuntu.com/p/R78fGBJbjb/]"
      ],
      "execution_count": 5,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tPXoYOkCEn_Y",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "91368f70-429b-4fac-ac32-2ab11d06f16f"
      },
      "source": [
        "# Deep Neural Networks:\n",
        "import tensorflow as tf; print('We\\'re using TF-{}.'.format(tf.__version__))\n",
        "# import keras; print('We\\'re using Keras-{}.'.format(keras.__version__))\n",
        "from tensorflow.keras.layers import (Input, Dense, Dropout, Flatten, BatchNormalization,\n",
        "                                     Conv1D, Conv2D, MaxPooling1D, MaxPooling2D,\n",
        "                                     LSTM, GRU, Embedding, Bidirectional, Concatenate)\n",
        "from tensorflow.keras.regularizers import (l1, l2, l1_l2)\n",
        "from tensorflow.keras.optimizers import (RMSprop, Adam, SGD)\n",
        "from tensorflow.keras.models import (Sequential, Model)\n",
        "\n",
        "# Core:\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "from scipy import interp\n",
        "import matplotlib.patches as patches\n",
        "\n",
        "# Performance:\n",
        "from sklearn.metrics import (confusion_matrix, classification_report, matthews_corrcoef, precision_score, roc_curve, auc)\n",
        "from sklearn.model_selection import (StratifiedKFold, KFold, train_test_split)\n",
        "\n",
        "#Utilities:\n",
        "from tensorflow.keras.callbacks import EarlyStopping, ModelCheckpoint\n",
        "from tensorflow.keras.utils import to_categorical as labelEncoding   # Usages: Y = labelEncoding(Y, dtype=int)\n",
        "from tensorflow.keras.utils import plot_model                        # Usages: plot_model(model, to_file='model.png', show_shapes=True, show_layer_names=False, expand_nested=True)\n",
        "\n",
        "#end-import"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "We're using TF-2.3.0.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bxP0codaeeIm",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def lossPlot(results):\n",
        "    plt.title(label='Loss: Training and Validation')\n",
        "    plt.plot(results.history['loss'], label='Training Loss')\n",
        "    plt.plot(results.history['val_loss'], label='Validation Loss')\n",
        "    plt.xlabel('Epoch')\n",
        "    plt.ylabel('Loss')\n",
        "    plt.legend()\n",
        "    plt.show()\n",
        "#end-def\n",
        "\n",
        "def accuracyPlot(results):\n",
        "    plt.title(label='Accuracy: Training and Validation')\n",
        "    plt.plot(results.history['accuracy'], label='Training Accuracy')\n",
        "    plt.plot(results.history['val_accuracy'], label='Validation Accuracy')\n",
        "    plt.xlabel('Epoch')\n",
        "    plt.ylabel('Accuracy')\n",
        "    plt.legend()\n",
        "    plt.show()\n",
        "#end-def"
      ],
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "npYy04WNcxLG",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def rocPlot(TPR, meanFPR):\n",
        "    plt.plot([0,1], [0,1],linestyle = '--',lw = 2,color = 'black')\n",
        "    meanTPR = np.mean(TPR, axis=0)\n",
        "    meanAUC = auc(meanFPR, meanTPR)\n",
        "    plt.plot(meanFPR, meanTPR, color='blue',\n",
        "            label=r'Mean ROC (AUC = %0.2f )' % (meanAUC),lw=2, alpha=1)\n",
        "\n",
        "    plt.xlabel('False Positive Rate (FPR)')\n",
        "    plt.ylabel('True Positive Rate (TPR)')\n",
        "    plt.title('Receiver Operating Characteristic Curve (ROC Curve)')\n",
        "    plt.legend(loc=\"lower right\")\n",
        "    plt.savefig('ROC-740.png')\n",
        "    plt.show()\n",
        "#end-def"
      ],
      "execution_count": 8,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "j3ztKoZ2y98R",
        "colab_type": "text"
      },
      "source": [
        "### [3]. Load Datasets:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lsWfth72frPL",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 90
        },
        "outputId": "128b4cca-ef7e-4d2c-8d21-bdca6dabeb60"
      },
      "source": [
        "T = 15 # terminus_length\n",
        "\n",
        "X1 = np.load('bpf-740.npy')\n",
        "X2 = np.load('bits-740.npy')\n",
        "X3 = np.load('blosum-740.npy')\n",
        "\n",
        "\n",
        "X1 = X1[:,0:T,:]\n",
        "X2 = X2[:,0:T,:]\n",
        "X3 = X3[:,0:T,:]\n",
        "\n",
        "\n",
        "Y  = [1 for _ in range(376)]\n",
        "Y += [0 for _ in range(364)]\n",
        "\n",
        "Y = labelEncoding(Y, dtype=int)\n",
        "\n",
        "\n",
        "print(X1.shape)\n",
        "print(X2.shape)\n",
        "print(X3.shape)\n",
        "print(Y.shape)"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(740, 15, 20)\n",
            "(740, 15, 31)\n",
            "(740, 15, 20)\n",
            "(740, 2)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "a6-4jleCHmXd",
        "colab_type": "text"
      },
      "source": [
        "### [4]. Generate Model:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "z_nLr1AOfvG1",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "### Model-740\n",
        "\n",
        "def Network():\n",
        "    ### Head-1:\n",
        "    input1 = Input(shape=X1[0].shape)\n",
        "\n",
        "    x = Conv1D(filters=10, kernel_size=4, padding='same', activation='relu', kernel_regularizer=l2(l=0.01))(input1)\n",
        "    x = BatchNormalization()(x)\n",
        "    x = Dropout(rate=0.80)(x)\n",
        "\n",
        "    x = Conv1D(filters=8, kernel_size=3, padding='same', activation='relu', kernel_regularizer=l2(l=0.01))(x)\n",
        "    x = BatchNormalization()(x)\n",
        "    x = Dropout(rate=0.70)(x)\n",
        "\n",
        "    head1 = Flatten()(x)\n",
        "\n",
        "\n",
        "    ### Head-2:\n",
        "    input2 = Input(shape=X2[0].shape)\n",
        "\n",
        "    x = Conv1D(filters=10, kernel_size=4, padding='same', activation='relu', kernel_regularizer=l2(l=0.01))(input2)\n",
        "    x = BatchNormalization()(x)\n",
        "    x = Dropout(rate=0.70)(x)\n",
        "\n",
        "    x = Conv1D(filters=8, kernel_size=3, padding='same', activation='relu', kernel_regularizer=l2(l=0.01))(x)\n",
        "    x = BatchNormalization()(x)\n",
        "    x = Dropout(rate=0.70)(x)\n",
        "\n",
        "    head2 = Flatten()(x)\n",
        "\n",
        "\n",
        "    ### Head-3:\n",
        "    input3 = Input(shape=X3[0].shape)\n",
        "    \n",
        "    x = Conv1D(filters=10, kernel_size=4, padding='same', activation='relu',)(input3)\n",
        "    x = BatchNormalization()(x)\n",
        "    x = Dropout(rate=0.70)(x)\n",
        "\n",
        "    x = Conv1D(filters=8, kernel_size=3, padding='same', activation='relu', kernel_regularizer=l2(l=0.01))(x)\n",
        "    x = BatchNormalization()(x)\n",
        "    x = Dropout(rate=0.70)(x)\n",
        "\n",
        "    head3 = Flatten()(x)\n",
        "\n",
        "\n",
        "    # merge\n",
        "    merge = Concatenate()([head1, head2, head3])\n",
        "\n",
        "    output = Dense(units=8, activation='relu', kernel_regularizer=l2(l=0.01))(merge)\n",
        "    output = BatchNormalization()(output)\n",
        "    output = Dropout(rate=0.70)(output)\n",
        "\n",
        "    output = Dense(units=2, activation='softmax')(output)\n",
        "\n",
        "    return Model(inputs=[input1, input2, input3], outputs=[output])\n",
        "#end-def"
      ],
      "execution_count": 10,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YwGc-yzTfyBE",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "e2577352-9e15-4abc-cbcd-c102d1f9a612"
      },
      "source": [
        "model = Network()\n",
        "model.summary()\n",
        "plot_model(model, to_file='model-740.png', show_shapes=True, show_layer_names=False, expand_nested=True)"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"functional_1\"\n",
            "__________________________________________________________________________________________________\n",
            "Layer (type)                    Output Shape         Param #     Connected to                     \n",
            "==================================================================================================\n",
            "input_1 (InputLayer)            [(None, 15, 20)]     0                                            \n",
            "__________________________________________________________________________________________________\n",
            "input_2 (InputLayer)            [(None, 15, 31)]     0                                            \n",
            "__________________________________________________________________________________________________\n",
            "input_3 (InputLayer)            [(None, 15, 20)]     0                                            \n",
            "__________________________________________________________________________________________________\n",
            "conv1d (Conv1D)                 (None, 15, 10)       810         input_1[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "conv1d_2 (Conv1D)               (None, 15, 10)       1250        input_2[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "conv1d_4 (Conv1D)               (None, 15, 10)       810         input_3[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization (BatchNorma (None, 15, 10)       40          conv1d[0][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_2 (BatchNor (None, 15, 10)       40          conv1d_2[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_4 (BatchNor (None, 15, 10)       40          conv1d_4[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "dropout (Dropout)               (None, 15, 10)       0           batch_normalization[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "dropout_2 (Dropout)             (None, 15, 10)       0           batch_normalization_2[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "dropout_4 (Dropout)             (None, 15, 10)       0           batch_normalization_4[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "conv1d_1 (Conv1D)               (None, 15, 8)        248         dropout[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "conv1d_3 (Conv1D)               (None, 15, 8)        248         dropout_2[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "conv1d_5 (Conv1D)               (None, 15, 8)        248         dropout_4[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_1 (BatchNor (None, 15, 8)        32          conv1d_1[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_3 (BatchNor (None, 15, 8)        32          conv1d_3[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_5 (BatchNor (None, 15, 8)        32          conv1d_5[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "dropout_1 (Dropout)             (None, 15, 8)        0           batch_normalization_1[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "dropout_3 (Dropout)             (None, 15, 8)        0           batch_normalization_3[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "dropout_5 (Dropout)             (None, 15, 8)        0           batch_normalization_5[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "flatten (Flatten)               (None, 120)          0           dropout_1[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "flatten_1 (Flatten)             (None, 120)          0           dropout_3[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "flatten_2 (Flatten)             (None, 120)          0           dropout_5[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "concatenate (Concatenate)       (None, 360)          0           flatten[0][0]                    \n",
            "                                                                 flatten_1[0][0]                  \n",
            "                                                                 flatten_2[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "dense (Dense)                   (None, 8)            2888        concatenate[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_6 (BatchNor (None, 8)            32          dense[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "dropout_6 (Dropout)             (None, 8)            0           batch_normalization_6[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "dense_1 (Dense)                 (None, 2)            18          dropout_6[0][0]                  \n",
            "==================================================================================================\n",
            "Total params: 6,768\n",
            "Trainable params: 6,644\n",
            "Non-trainable params: 124\n",
            "__________________________________________________________________________________________________\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<IPython.core.display.Image object>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 11
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Y3-ANHVKH-RJ",
        "colab_type": "text"
      },
      "source": [
        "### [5]. Run Model:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_1wfc1k6LQVT",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "f36055b3-34db-4798-9cfe-37a05b14e81a"
      },
      "source": [
        "setEpochNumber     = 500     # Performed-welled in epoch 600.\n",
        "setBatchSizeNumber = 8\n",
        "####################################################\n",
        "\n",
        "cv = KFold(n_splits=5, shuffle=True, random_state=101)\n",
        "\n",
        "Accuracy = []\n",
        "Sensitivity = []\n",
        "Specificity = []\n",
        "Precision = []\n",
        "MCC = []\n",
        "\n",
        "# ROC Curve:\n",
        "fig1 = plt.figure(figsize=[12,12])\n",
        "\n",
        "TPR = []\n",
        "meanFPR = np.linspace(0, 1, 100)\n",
        "\n",
        "i = 1\n",
        "\n",
        "# CM = np.array([\n",
        "#      [0, 0],\n",
        "#      [0, 0],\n",
        "# ], dtype=int)\n",
        "\n",
        "for train, test in cv.split(Y):\n",
        "\n",
        "    # Compile Model:\n",
        "    model = Network()\n",
        "    model.compile(optimizer=Adam(lr=0.005),\n",
        "                  loss='categorical_crossentropy',\n",
        "                  metrics=['accuracy'])\n",
        "\n",
        "    # Run Model:\n",
        "    results = model.fit(x=[X1[train,:,:], X2[train,:,:], X3[train,:,:]],\n",
        "                        y=Y[train,:],\n",
        "                        validation_data=([X1[test,:,:], X2[test,:,:], X3[test,:,:],], Y[test,:]),\n",
        "                        batch_size=setBatchSizeNumber, epochs=setEpochNumber,\n",
        "                        verbose=1,\n",
        "                        callbacks=[])\n",
        "\n",
        "    # Evaluate the Model:\n",
        "    accuracy = model.evaluate(x=[X1[test,:,:], X2[test,:,:], X3[test,:,:]], y=Y[test,:])\n",
        "    Accuracy.append(accuracy[1])\n",
        "    \n",
        "    # Performance Metices:\n",
        "    Yactual = Y[test,:].argmax(axis=1)\n",
        "    Yp = model.predict([X1[test,:,:], X2[test,:,:], X3[test,:,:]])\n",
        "    v = Yp\n",
        "    Yp = Yp.argmax(axis=1)\n",
        "\n",
        "    CM = confusion_matrix(y_pred=Yp, y_true=Yactual)\n",
        "    TN, FP, FN, TP = CM.ravel()\n",
        "\n",
        "    MCC.append(matthews_corrcoef(y_true=Yactual, y_pred=Yp))\n",
        "    Sensitivity.append( TP / (TP + FN) )\n",
        "    Specificity.append( TN / (TN + FP) )\n",
        "    Precision.append(precision_score(y_true=Yactual, y_pred=Yp))\n",
        "    \n",
        "    # ROC Curve\n",
        "    fpr, tpr, _ = roc_curve(Yactual, v[:,1])\n",
        "    TPR.append(interp(meanFPR, fpr, tpr))\n",
        "    rocauc = auc(fpr, tpr)\n",
        "    plt.plot(fpr, tpr, lw=2, alpha=0.3, label='ROC fold %d (AUC = %0.2f)' % (i, rocauc))\n",
        "    i= i+1\n",
        "    \n",
        "    # # Performance Plot\n",
        "    # print('#################################################')\n",
        "    # print('Fold\\'s Accuracy: {:.2f}'.format(accuracy[1]*100.0))\n",
        "    # lossPlot(results)\n",
        "    # accuracyPlot(results)\n",
        "    # print('#################################################')\n",
        "\n",
        "#end-for\n",
        "\n",
        "rocPlot(TPR, meanFPR)\n",
        "\n",
        "print(Accuracy)\n",
        "print('Accuracy: {:.2f}'.format(np.sum(Accuracy)/5.0))\n",
        "print('Sensitivity: {0:.4f}'.format(np.sum(Sensitivity)/5.00))\n",
        "print('Specificity: {0:.4f}'.format(np.sum(Specificity)/5.00))\n",
        "print('MCC: {0:.4f}'.format(np.sum(MCC)/5.00))\n",
        "print('Precision: {0:.4f}'.format(np.sum(Precision)/5.00))"
      ],
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/500\n",
            "74/74 [==============================] - 1s 11ms/step - loss: 1.7053 - accuracy: 0.5169 - val_loss: 1.3822 - val_accuracy: 0.5000\n",
            "Epoch 2/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 1.4142 - accuracy: 0.4949 - val_loss: 1.3127 - val_accuracy: 0.4932\n",
            "Epoch 3/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 1.2738 - accuracy: 0.5236 - val_loss: 1.2199 - val_accuracy: 0.4932\n",
            "Epoch 4/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 1.1947 - accuracy: 0.5253 - val_loss: 1.1480 - val_accuracy: 0.4932\n",
            "Epoch 5/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 1.1080 - accuracy: 0.5405 - val_loss: 1.0701 - val_accuracy: 0.4797\n",
            "Epoch 6/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 1.0259 - accuracy: 0.5895 - val_loss: 1.0161 - val_accuracy: 0.5338\n",
            "Epoch 7/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.9908 - accuracy: 0.6098 - val_loss: 0.9584 - val_accuracy: 0.6014\n",
            "Epoch 8/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.9692 - accuracy: 0.5760 - val_loss: 0.9454 - val_accuracy: 0.6014\n",
            "Epoch 9/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.9105 - accuracy: 0.6419 - val_loss: 0.8896 - val_accuracy: 0.6284\n",
            "Epoch 10/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8866 - accuracy: 0.6216 - val_loss: 0.9081 - val_accuracy: 0.5405\n",
            "Epoch 11/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.8919 - accuracy: 0.6351 - val_loss: 0.8538 - val_accuracy: 0.7095\n",
            "Epoch 12/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.8874 - accuracy: 0.6064 - val_loss: 0.8118 - val_accuracy: 0.7838\n",
            "Epoch 13/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.8400 - accuracy: 0.6672 - val_loss: 0.8139 - val_accuracy: 0.7297\n",
            "Epoch 14/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8531 - accuracy: 0.6318 - val_loss: 0.8078 - val_accuracy: 0.7297\n",
            "Epoch 15/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8556 - accuracy: 0.6132 - val_loss: 0.7958 - val_accuracy: 0.8041\n",
            "Epoch 16/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.8429 - accuracy: 0.6520 - val_loss: 0.7588 - val_accuracy: 0.7703\n",
            "Epoch 17/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7932 - accuracy: 0.6791 - val_loss: 0.7198 - val_accuracy: 0.7365\n",
            "Epoch 18/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.8278 - accuracy: 0.6672 - val_loss: 0.7415 - val_accuracy: 0.7770\n",
            "Epoch 19/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.8166 - accuracy: 0.6791 - val_loss: 0.7198 - val_accuracy: 0.7838\n",
            "Epoch 20/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7935 - accuracy: 0.6926 - val_loss: 0.7031 - val_accuracy: 0.8041\n",
            "Epoch 21/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.8078 - accuracy: 0.6757 - val_loss: 0.7395 - val_accuracy: 0.7703\n",
            "Epoch 22/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.8339 - accuracy: 0.6757 - val_loss: 0.7744 - val_accuracy: 0.7770\n",
            "Epoch 23/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.8078 - accuracy: 0.6892 - val_loss: 0.7333 - val_accuracy: 0.7973\n",
            "Epoch 24/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.8053 - accuracy: 0.6774 - val_loss: 0.7410 - val_accuracy: 0.8041\n",
            "Epoch 25/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.8384 - accuracy: 0.6672 - val_loss: 0.7747 - val_accuracy: 0.7905\n",
            "Epoch 26/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.8346 - accuracy: 0.6520 - val_loss: 0.7235 - val_accuracy: 0.7770\n",
            "Epoch 27/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.8363 - accuracy: 0.7044 - val_loss: 0.7292 - val_accuracy: 0.7905\n",
            "Epoch 28/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.8115 - accuracy: 0.6976 - val_loss: 0.7361 - val_accuracy: 0.7838\n",
            "Epoch 29/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8037 - accuracy: 0.6791 - val_loss: 0.7301 - val_accuracy: 0.7973\n",
            "Epoch 30/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.8002 - accuracy: 0.6824 - val_loss: 0.7094 - val_accuracy: 0.8243\n",
            "Epoch 31/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7821 - accuracy: 0.7044 - val_loss: 0.7260 - val_accuracy: 0.7703\n",
            "Epoch 32/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.8047 - accuracy: 0.6909 - val_loss: 0.7410 - val_accuracy: 0.7973\n",
            "Epoch 33/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7971 - accuracy: 0.6959 - val_loss: 0.7276 - val_accuracy: 0.7838\n",
            "Epoch 34/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7631 - accuracy: 0.7078 - val_loss: 0.6927 - val_accuracy: 0.7973\n",
            "Epoch 35/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.8047 - accuracy: 0.6959 - val_loss: 0.7163 - val_accuracy: 0.7973\n",
            "Epoch 36/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7911 - accuracy: 0.7027 - val_loss: 0.6914 - val_accuracy: 0.8041\n",
            "Epoch 37/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7730 - accuracy: 0.6841 - val_loss: 0.6866 - val_accuracy: 0.7973\n",
            "Epoch 38/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8009 - accuracy: 0.6926 - val_loss: 0.7109 - val_accuracy: 0.7838\n",
            "Epoch 39/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7506 - accuracy: 0.7264 - val_loss: 0.7112 - val_accuracy: 0.7230\n",
            "Epoch 40/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.8181 - accuracy: 0.6740 - val_loss: 0.7123 - val_accuracy: 0.8108\n",
            "Epoch 41/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7925 - accuracy: 0.7111 - val_loss: 0.6864 - val_accuracy: 0.8176\n",
            "Epoch 42/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7947 - accuracy: 0.6943 - val_loss: 0.7043 - val_accuracy: 0.8108\n",
            "Epoch 43/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7820 - accuracy: 0.6993 - val_loss: 0.6836 - val_accuracy: 0.8041\n",
            "Epoch 44/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7834 - accuracy: 0.6858 - val_loss: 0.6854 - val_accuracy: 0.7703\n",
            "Epoch 45/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7766 - accuracy: 0.7145 - val_loss: 0.7168 - val_accuracy: 0.7635\n",
            "Epoch 46/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7901 - accuracy: 0.6909 - val_loss: 0.6987 - val_accuracy: 0.7973\n",
            "Epoch 47/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7842 - accuracy: 0.6807 - val_loss: 0.6559 - val_accuracy: 0.8041\n",
            "Epoch 48/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7208 - accuracy: 0.7230 - val_loss: 0.6553 - val_accuracy: 0.7973\n",
            "Epoch 49/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7679 - accuracy: 0.7027 - val_loss: 0.6899 - val_accuracy: 0.7905\n",
            "Epoch 50/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7332 - accuracy: 0.6926 - val_loss: 0.6651 - val_accuracy: 0.7770\n",
            "Epoch 51/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7526 - accuracy: 0.6926 - val_loss: 0.6819 - val_accuracy: 0.7568\n",
            "Epoch 52/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7447 - accuracy: 0.7027 - val_loss: 0.6704 - val_accuracy: 0.7905\n",
            "Epoch 53/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7612 - accuracy: 0.7061 - val_loss: 0.6817 - val_accuracy: 0.7838\n",
            "Epoch 54/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7683 - accuracy: 0.7196 - val_loss: 0.6611 - val_accuracy: 0.7905\n",
            "Epoch 55/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7420 - accuracy: 0.7264 - val_loss: 0.6394 - val_accuracy: 0.8243\n",
            "Epoch 56/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7778 - accuracy: 0.6740 - val_loss: 0.6593 - val_accuracy: 0.8041\n",
            "Epoch 57/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7551 - accuracy: 0.7095 - val_loss: 0.6737 - val_accuracy: 0.7973\n",
            "Epoch 58/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7571 - accuracy: 0.7111 - val_loss: 0.6972 - val_accuracy: 0.7973\n",
            "Epoch 59/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7462 - accuracy: 0.7399 - val_loss: 0.6593 - val_accuracy: 0.7770\n",
            "Epoch 60/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7458 - accuracy: 0.7128 - val_loss: 0.6404 - val_accuracy: 0.7973\n",
            "Epoch 61/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7182 - accuracy: 0.7331 - val_loss: 0.6501 - val_accuracy: 0.7973\n",
            "Epoch 62/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7784 - accuracy: 0.6858 - val_loss: 0.6947 - val_accuracy: 0.7568\n",
            "Epoch 63/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7192 - accuracy: 0.7179 - val_loss: 0.6263 - val_accuracy: 0.8108\n",
            "Epoch 64/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7324 - accuracy: 0.7145 - val_loss: 0.6578 - val_accuracy: 0.7770\n",
            "Epoch 65/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7560 - accuracy: 0.7145 - val_loss: 0.6719 - val_accuracy: 0.7973\n",
            "Epoch 66/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7373 - accuracy: 0.7095 - val_loss: 0.6513 - val_accuracy: 0.7905\n",
            "Epoch 67/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7250 - accuracy: 0.7128 - val_loss: 0.6397 - val_accuracy: 0.7905\n",
            "Epoch 68/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7696 - accuracy: 0.6723 - val_loss: 0.6652 - val_accuracy: 0.7973\n",
            "Epoch 69/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7207 - accuracy: 0.7196 - val_loss: 0.6421 - val_accuracy: 0.7973\n",
            "Epoch 70/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7636 - accuracy: 0.6993 - val_loss: 0.6594 - val_accuracy: 0.7770\n",
            "Epoch 71/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7481 - accuracy: 0.7213 - val_loss: 0.6634 - val_accuracy: 0.7905\n",
            "Epoch 72/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7046 - accuracy: 0.7432 - val_loss: 0.6217 - val_accuracy: 0.8041\n",
            "Epoch 73/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7222 - accuracy: 0.7399 - val_loss: 0.6263 - val_accuracy: 0.8108\n",
            "Epoch 74/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7264 - accuracy: 0.7517 - val_loss: 0.6274 - val_accuracy: 0.8446\n",
            "Epoch 75/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7476 - accuracy: 0.7247 - val_loss: 0.6379 - val_accuracy: 0.8176\n",
            "Epoch 76/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7227 - accuracy: 0.7061 - val_loss: 0.6188 - val_accuracy: 0.8041\n",
            "Epoch 77/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7133 - accuracy: 0.7416 - val_loss: 0.6183 - val_accuracy: 0.8311\n",
            "Epoch 78/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6826 - accuracy: 0.7449 - val_loss: 0.6121 - val_accuracy: 0.7973\n",
            "Epoch 79/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6899 - accuracy: 0.7196 - val_loss: 0.6231 - val_accuracy: 0.7973\n",
            "Epoch 80/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7434 - accuracy: 0.7128 - val_loss: 0.6549 - val_accuracy: 0.7770\n",
            "Epoch 81/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7135 - accuracy: 0.7027 - val_loss: 0.6154 - val_accuracy: 0.8041\n",
            "Epoch 82/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6953 - accuracy: 0.7568 - val_loss: 0.6035 - val_accuracy: 0.7973\n",
            "Epoch 83/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7254 - accuracy: 0.7162 - val_loss: 0.6225 - val_accuracy: 0.8041\n",
            "Epoch 84/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7279 - accuracy: 0.7230 - val_loss: 0.6350 - val_accuracy: 0.8108\n",
            "Epoch 85/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6871 - accuracy: 0.7399 - val_loss: 0.6389 - val_accuracy: 0.8041\n",
            "Epoch 86/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7194 - accuracy: 0.7128 - val_loss: 0.6375 - val_accuracy: 0.8041\n",
            "Epoch 87/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7438 - accuracy: 0.7145 - val_loss: 0.6115 - val_accuracy: 0.8243\n",
            "Epoch 88/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7203 - accuracy: 0.7061 - val_loss: 0.6245 - val_accuracy: 0.8176\n",
            "Epoch 89/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6855 - accuracy: 0.7264 - val_loss: 0.5965 - val_accuracy: 0.7973\n",
            "Epoch 90/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6897 - accuracy: 0.7449 - val_loss: 0.5929 - val_accuracy: 0.8243\n",
            "Epoch 91/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6643 - accuracy: 0.7584 - val_loss: 0.5813 - val_accuracy: 0.7905\n",
            "Epoch 92/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6932 - accuracy: 0.7264 - val_loss: 0.6010 - val_accuracy: 0.8311\n",
            "Epoch 93/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7366 - accuracy: 0.7027 - val_loss: 0.6484 - val_accuracy: 0.7905\n",
            "Epoch 94/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6919 - accuracy: 0.7365 - val_loss: 0.6324 - val_accuracy: 0.7838\n",
            "Epoch 95/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7350 - accuracy: 0.7365 - val_loss: 0.6332 - val_accuracy: 0.7838\n",
            "Epoch 96/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7115 - accuracy: 0.7230 - val_loss: 0.6278 - val_accuracy: 0.8041\n",
            "Epoch 97/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7023 - accuracy: 0.7095 - val_loss: 0.6266 - val_accuracy: 0.8243\n",
            "Epoch 98/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7155 - accuracy: 0.7297 - val_loss: 0.6415 - val_accuracy: 0.7770\n",
            "Epoch 99/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6864 - accuracy: 0.7382 - val_loss: 0.6161 - val_accuracy: 0.8243\n",
            "Epoch 100/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7014 - accuracy: 0.7432 - val_loss: 0.6245 - val_accuracy: 0.7905\n",
            "Epoch 101/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6944 - accuracy: 0.7382 - val_loss: 0.6219 - val_accuracy: 0.7905\n",
            "Epoch 102/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7150 - accuracy: 0.7128 - val_loss: 0.5971 - val_accuracy: 0.8041\n",
            "Epoch 103/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6916 - accuracy: 0.7348 - val_loss: 0.6426 - val_accuracy: 0.7635\n",
            "Epoch 104/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6730 - accuracy: 0.7534 - val_loss: 0.6223 - val_accuracy: 0.7568\n",
            "Epoch 105/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7233 - accuracy: 0.7264 - val_loss: 0.6404 - val_accuracy: 0.8041\n",
            "Epoch 106/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7050 - accuracy: 0.7179 - val_loss: 0.6117 - val_accuracy: 0.7703\n",
            "Epoch 107/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6904 - accuracy: 0.7314 - val_loss: 0.6029 - val_accuracy: 0.7973\n",
            "Epoch 108/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6904 - accuracy: 0.7483 - val_loss: 0.6069 - val_accuracy: 0.8176\n",
            "Epoch 109/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6883 - accuracy: 0.7280 - val_loss: 0.6271 - val_accuracy: 0.8041\n",
            "Epoch 110/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6860 - accuracy: 0.7297 - val_loss: 0.5947 - val_accuracy: 0.7973\n",
            "Epoch 111/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6908 - accuracy: 0.7601 - val_loss: 0.6038 - val_accuracy: 0.7973\n",
            "Epoch 112/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6698 - accuracy: 0.7365 - val_loss: 0.6007 - val_accuracy: 0.7770\n",
            "Epoch 113/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6572 - accuracy: 0.7584 - val_loss: 0.6072 - val_accuracy: 0.7568\n",
            "Epoch 114/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6367 - accuracy: 0.7449 - val_loss: 0.5779 - val_accuracy: 0.7973\n",
            "Epoch 115/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6843 - accuracy: 0.7517 - val_loss: 0.5906 - val_accuracy: 0.8041\n",
            "Epoch 116/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6650 - accuracy: 0.7618 - val_loss: 0.5831 - val_accuracy: 0.7703\n",
            "Epoch 117/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6636 - accuracy: 0.7652 - val_loss: 0.5853 - val_accuracy: 0.8041\n",
            "Epoch 118/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6553 - accuracy: 0.7449 - val_loss: 0.5927 - val_accuracy: 0.8378\n",
            "Epoch 119/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6945 - accuracy: 0.7449 - val_loss: 0.6101 - val_accuracy: 0.8041\n",
            "Epoch 120/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7091 - accuracy: 0.7331 - val_loss: 0.5973 - val_accuracy: 0.8041\n",
            "Epoch 121/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6670 - accuracy: 0.7416 - val_loss: 0.6052 - val_accuracy: 0.7973\n",
            "Epoch 122/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6550 - accuracy: 0.7517 - val_loss: 0.5750 - val_accuracy: 0.8176\n",
            "Epoch 123/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6465 - accuracy: 0.7483 - val_loss: 0.5651 - val_accuracy: 0.8041\n",
            "Epoch 124/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6616 - accuracy: 0.7365 - val_loss: 0.5842 - val_accuracy: 0.7973\n",
            "Epoch 125/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6946 - accuracy: 0.7280 - val_loss: 0.6052 - val_accuracy: 0.7973\n",
            "Epoch 126/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6607 - accuracy: 0.7483 - val_loss: 0.5696 - val_accuracy: 0.8108\n",
            "Epoch 127/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6544 - accuracy: 0.7584 - val_loss: 0.5866 - val_accuracy: 0.7838\n",
            "Epoch 128/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6316 - accuracy: 0.7736 - val_loss: 0.5687 - val_accuracy: 0.7973\n",
            "Epoch 129/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6659 - accuracy: 0.7399 - val_loss: 0.5870 - val_accuracy: 0.8243\n",
            "Epoch 130/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6956 - accuracy: 0.7382 - val_loss: 0.6058 - val_accuracy: 0.7973\n",
            "Epoch 131/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6314 - accuracy: 0.7584 - val_loss: 0.5820 - val_accuracy: 0.7770\n",
            "Epoch 132/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6737 - accuracy: 0.7601 - val_loss: 0.6037 - val_accuracy: 0.7905\n",
            "Epoch 133/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6736 - accuracy: 0.7534 - val_loss: 0.6022 - val_accuracy: 0.7703\n",
            "Epoch 134/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7024 - accuracy: 0.7331 - val_loss: 0.6083 - val_accuracy: 0.8041\n",
            "Epoch 135/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6875 - accuracy: 0.7517 - val_loss: 0.5989 - val_accuracy: 0.8108\n",
            "Epoch 136/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6878 - accuracy: 0.7179 - val_loss: 0.6093 - val_accuracy: 0.8108\n",
            "Epoch 137/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6616 - accuracy: 0.7703 - val_loss: 0.5874 - val_accuracy: 0.7973\n",
            "Epoch 138/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6876 - accuracy: 0.7534 - val_loss: 0.6237 - val_accuracy: 0.7973\n",
            "Epoch 139/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6701 - accuracy: 0.7584 - val_loss: 0.5959 - val_accuracy: 0.7770\n",
            "Epoch 140/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6816 - accuracy: 0.7466 - val_loss: 0.5863 - val_accuracy: 0.8176\n",
            "Epoch 141/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7066 - accuracy: 0.7568 - val_loss: 0.5995 - val_accuracy: 0.8041\n",
            "Epoch 142/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6919 - accuracy: 0.7196 - val_loss: 0.5881 - val_accuracy: 0.8041\n",
            "Epoch 143/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6605 - accuracy: 0.7432 - val_loss: 0.5725 - val_accuracy: 0.8176\n",
            "Epoch 144/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6546 - accuracy: 0.7517 - val_loss: 0.5645 - val_accuracy: 0.8108\n",
            "Epoch 145/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6760 - accuracy: 0.7449 - val_loss: 0.5520 - val_accuracy: 0.8108\n",
            "Epoch 146/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6198 - accuracy: 0.7838 - val_loss: 0.5675 - val_accuracy: 0.8041\n",
            "Epoch 147/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6737 - accuracy: 0.7449 - val_loss: 0.5997 - val_accuracy: 0.7973\n",
            "Epoch 148/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6504 - accuracy: 0.7534 - val_loss: 0.5771 - val_accuracy: 0.7905\n",
            "Epoch 149/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6969 - accuracy: 0.7230 - val_loss: 0.5755 - val_accuracy: 0.8041\n",
            "Epoch 150/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6765 - accuracy: 0.7331 - val_loss: 0.6030 - val_accuracy: 0.7905\n",
            "Epoch 151/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6406 - accuracy: 0.7618 - val_loss: 0.5500 - val_accuracy: 0.8041\n",
            "Epoch 152/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6413 - accuracy: 0.7331 - val_loss: 0.5869 - val_accuracy: 0.7973\n",
            "Epoch 153/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6712 - accuracy: 0.7449 - val_loss: 0.5748 - val_accuracy: 0.8378\n",
            "Epoch 154/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6733 - accuracy: 0.7517 - val_loss: 0.5870 - val_accuracy: 0.8378\n",
            "Epoch 155/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6643 - accuracy: 0.7399 - val_loss: 0.5748 - val_accuracy: 0.8311\n",
            "Epoch 156/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6386 - accuracy: 0.7534 - val_loss: 0.5827 - val_accuracy: 0.8176\n",
            "Epoch 157/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6440 - accuracy: 0.7449 - val_loss: 0.5613 - val_accuracy: 0.8243\n",
            "Epoch 158/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6609 - accuracy: 0.7416 - val_loss: 0.5697 - val_accuracy: 0.8243\n",
            "Epoch 159/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6228 - accuracy: 0.7787 - val_loss: 0.5557 - val_accuracy: 0.8176\n",
            "Epoch 160/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6375 - accuracy: 0.7804 - val_loss: 0.5751 - val_accuracy: 0.8176\n",
            "Epoch 161/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6474 - accuracy: 0.7382 - val_loss: 0.5943 - val_accuracy: 0.8108\n",
            "Epoch 162/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6383 - accuracy: 0.7770 - val_loss: 0.5819 - val_accuracy: 0.7973\n",
            "Epoch 163/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6640 - accuracy: 0.7500 - val_loss: 0.5873 - val_accuracy: 0.7905\n",
            "Epoch 164/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6423 - accuracy: 0.7720 - val_loss: 0.5682 - val_accuracy: 0.8041\n",
            "Epoch 165/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6571 - accuracy: 0.7534 - val_loss: 0.5655 - val_accuracy: 0.8176\n",
            "Epoch 166/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6377 - accuracy: 0.7534 - val_loss: 0.5715 - val_accuracy: 0.8176\n",
            "Epoch 167/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6396 - accuracy: 0.7432 - val_loss: 0.6095 - val_accuracy: 0.7838\n",
            "Epoch 168/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6365 - accuracy: 0.7534 - val_loss: 0.5534 - val_accuracy: 0.7905\n",
            "Epoch 169/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6514 - accuracy: 0.7517 - val_loss: 0.5661 - val_accuracy: 0.7973\n",
            "Epoch 170/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6338 - accuracy: 0.7601 - val_loss: 0.5488 - val_accuracy: 0.8108\n",
            "Epoch 171/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6423 - accuracy: 0.7601 - val_loss: 0.5535 - val_accuracy: 0.8041\n",
            "Epoch 172/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6711 - accuracy: 0.7432 - val_loss: 0.5600 - val_accuracy: 0.8243\n",
            "Epoch 173/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6444 - accuracy: 0.7500 - val_loss: 0.5699 - val_accuracy: 0.7905\n",
            "Epoch 174/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6809 - accuracy: 0.7534 - val_loss: 0.5868 - val_accuracy: 0.7838\n",
            "Epoch 175/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6510 - accuracy: 0.7500 - val_loss: 0.5724 - val_accuracy: 0.7770\n",
            "Epoch 176/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6114 - accuracy: 0.7872 - val_loss: 0.5691 - val_accuracy: 0.7838\n",
            "Epoch 177/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7068 - accuracy: 0.7280 - val_loss: 0.5913 - val_accuracy: 0.8176\n",
            "Epoch 178/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6333 - accuracy: 0.7838 - val_loss: 0.5585 - val_accuracy: 0.7905\n",
            "Epoch 179/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6071 - accuracy: 0.7872 - val_loss: 0.5814 - val_accuracy: 0.7973\n",
            "Epoch 180/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6677 - accuracy: 0.7483 - val_loss: 0.5967 - val_accuracy: 0.8446\n",
            "Epoch 181/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6989 - accuracy: 0.7449 - val_loss: 0.5810 - val_accuracy: 0.8108\n",
            "Epoch 182/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6480 - accuracy: 0.7635 - val_loss: 0.5598 - val_accuracy: 0.8176\n",
            "Epoch 183/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6570 - accuracy: 0.7483 - val_loss: 0.5499 - val_accuracy: 0.8311\n",
            "Epoch 184/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6300 - accuracy: 0.7534 - val_loss: 0.5462 - val_accuracy: 0.8311\n",
            "Epoch 185/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6185 - accuracy: 0.7787 - val_loss: 0.5772 - val_accuracy: 0.7838\n",
            "Epoch 186/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6317 - accuracy: 0.7686 - val_loss: 0.5572 - val_accuracy: 0.8041\n",
            "Epoch 187/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6247 - accuracy: 0.7466 - val_loss: 0.5748 - val_accuracy: 0.8108\n",
            "Epoch 188/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6348 - accuracy: 0.7534 - val_loss: 0.5780 - val_accuracy: 0.7973\n",
            "Epoch 189/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6769 - accuracy: 0.7280 - val_loss: 0.6002 - val_accuracy: 0.7973\n",
            "Epoch 190/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6459 - accuracy: 0.7787 - val_loss: 0.5848 - val_accuracy: 0.8108\n",
            "Epoch 191/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6853 - accuracy: 0.7635 - val_loss: 0.6355 - val_accuracy: 0.7635\n",
            "Epoch 192/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6437 - accuracy: 0.7635 - val_loss: 0.5717 - val_accuracy: 0.7905\n",
            "Epoch 193/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6350 - accuracy: 0.7500 - val_loss: 0.5612 - val_accuracy: 0.7973\n",
            "Epoch 194/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6139 - accuracy: 0.7753 - val_loss: 0.5600 - val_accuracy: 0.7973\n",
            "Epoch 195/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6312 - accuracy: 0.7872 - val_loss: 0.5576 - val_accuracy: 0.8176\n",
            "Epoch 196/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6449 - accuracy: 0.7584 - val_loss: 0.5607 - val_accuracy: 0.8311\n",
            "Epoch 197/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6441 - accuracy: 0.7416 - val_loss: 0.5520 - val_accuracy: 0.8176\n",
            "Epoch 198/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6631 - accuracy: 0.7314 - val_loss: 0.5665 - val_accuracy: 0.7973\n",
            "Epoch 199/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6494 - accuracy: 0.7517 - val_loss: 0.5435 - val_accuracy: 0.8176\n",
            "Epoch 200/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6176 - accuracy: 0.7669 - val_loss: 0.5565 - val_accuracy: 0.8176\n",
            "Epoch 201/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6191 - accuracy: 0.7618 - val_loss: 0.5489 - val_accuracy: 0.8243\n",
            "Epoch 202/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6536 - accuracy: 0.7669 - val_loss: 0.5526 - val_accuracy: 0.8243\n",
            "Epoch 203/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6207 - accuracy: 0.7584 - val_loss: 0.5442 - val_accuracy: 0.8446\n",
            "Epoch 204/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6426 - accuracy: 0.7686 - val_loss: 0.5521 - val_accuracy: 0.8378\n",
            "Epoch 205/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6025 - accuracy: 0.7534 - val_loss: 0.5500 - val_accuracy: 0.8243\n",
            "Epoch 206/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6324 - accuracy: 0.7416 - val_loss: 0.5657 - val_accuracy: 0.8041\n",
            "Epoch 207/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6261 - accuracy: 0.7787 - val_loss: 0.5757 - val_accuracy: 0.8108\n",
            "Epoch 208/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6470 - accuracy: 0.7568 - val_loss: 0.5544 - val_accuracy: 0.8108\n",
            "Epoch 209/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6709 - accuracy: 0.7483 - val_loss: 0.5670 - val_accuracy: 0.8108\n",
            "Epoch 210/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6382 - accuracy: 0.7500 - val_loss: 0.5606 - val_accuracy: 0.7973\n",
            "Epoch 211/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5844 - accuracy: 0.7905 - val_loss: 0.5285 - val_accuracy: 0.8176\n",
            "Epoch 212/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6815 - accuracy: 0.7584 - val_loss: 0.5572 - val_accuracy: 0.8378\n",
            "Epoch 213/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6550 - accuracy: 0.7416 - val_loss: 0.5458 - val_accuracy: 0.8176\n",
            "Epoch 214/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6036 - accuracy: 0.7770 - val_loss: 0.5263 - val_accuracy: 0.8243\n",
            "Epoch 215/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6264 - accuracy: 0.7601 - val_loss: 0.5254 - val_accuracy: 0.8446\n",
            "Epoch 216/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6070 - accuracy: 0.7787 - val_loss: 0.5209 - val_accuracy: 0.8311\n",
            "Epoch 217/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6117 - accuracy: 0.7517 - val_loss: 0.5557 - val_accuracy: 0.8041\n",
            "Epoch 218/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6324 - accuracy: 0.7568 - val_loss: 0.5536 - val_accuracy: 0.7905\n",
            "Epoch 219/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6358 - accuracy: 0.7382 - val_loss: 0.5454 - val_accuracy: 0.8108\n",
            "Epoch 220/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5750 - accuracy: 0.7804 - val_loss: 0.5282 - val_accuracy: 0.8378\n",
            "Epoch 221/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6043 - accuracy: 0.7568 - val_loss: 0.5309 - val_accuracy: 0.8243\n",
            "Epoch 222/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5963 - accuracy: 0.7787 - val_loss: 0.5285 - val_accuracy: 0.8378\n",
            "Epoch 223/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6812 - accuracy: 0.7162 - val_loss: 0.5748 - val_accuracy: 0.8041\n",
            "Epoch 224/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6171 - accuracy: 0.7416 - val_loss: 0.5378 - val_accuracy: 0.8243\n",
            "Epoch 225/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6188 - accuracy: 0.7568 - val_loss: 0.5426 - val_accuracy: 0.8378\n",
            "Epoch 226/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6591 - accuracy: 0.7618 - val_loss: 0.5599 - val_accuracy: 0.8378\n",
            "Epoch 227/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6230 - accuracy: 0.7787 - val_loss: 0.5359 - val_accuracy: 0.8378\n",
            "Epoch 228/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6142 - accuracy: 0.7568 - val_loss: 0.5285 - val_accuracy: 0.8176\n",
            "Epoch 229/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6191 - accuracy: 0.7466 - val_loss: 0.5444 - val_accuracy: 0.8176\n",
            "Epoch 230/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5975 - accuracy: 0.7686 - val_loss: 0.5409 - val_accuracy: 0.8446\n",
            "Epoch 231/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6306 - accuracy: 0.7584 - val_loss: 0.5433 - val_accuracy: 0.8243\n",
            "Epoch 232/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6316 - accuracy: 0.7635 - val_loss: 0.5736 - val_accuracy: 0.7905\n",
            "Epoch 233/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6102 - accuracy: 0.7720 - val_loss: 0.5444 - val_accuracy: 0.7973\n",
            "Epoch 234/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6096 - accuracy: 0.7652 - val_loss: 0.5602 - val_accuracy: 0.8041\n",
            "Epoch 235/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6153 - accuracy: 0.7669 - val_loss: 0.5672 - val_accuracy: 0.7905\n",
            "Epoch 236/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6423 - accuracy: 0.7534 - val_loss: 0.5509 - val_accuracy: 0.7973\n",
            "Epoch 237/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6350 - accuracy: 0.7449 - val_loss: 0.5542 - val_accuracy: 0.8176\n",
            "Epoch 238/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6046 - accuracy: 0.7669 - val_loss: 0.5393 - val_accuracy: 0.8041\n",
            "Epoch 239/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6253 - accuracy: 0.7517 - val_loss: 0.5300 - val_accuracy: 0.8176\n",
            "Epoch 240/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6393 - accuracy: 0.7838 - val_loss: 0.5480 - val_accuracy: 0.8041\n",
            "Epoch 241/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6206 - accuracy: 0.7703 - val_loss: 0.5268 - val_accuracy: 0.8108\n",
            "Epoch 242/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6271 - accuracy: 0.7348 - val_loss: 0.5400 - val_accuracy: 0.8243\n",
            "Epoch 243/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5990 - accuracy: 0.7686 - val_loss: 0.5448 - val_accuracy: 0.8108\n",
            "Epoch 244/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6329 - accuracy: 0.7500 - val_loss: 0.5426 - val_accuracy: 0.8378\n",
            "Epoch 245/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6436 - accuracy: 0.7466 - val_loss: 0.5404 - val_accuracy: 0.8041\n",
            "Epoch 246/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6477 - accuracy: 0.7280 - val_loss: 0.5465 - val_accuracy: 0.8176\n",
            "Epoch 247/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6052 - accuracy: 0.7483 - val_loss: 0.5302 - val_accuracy: 0.8176\n",
            "Epoch 248/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6174 - accuracy: 0.7365 - val_loss: 0.5542 - val_accuracy: 0.8041\n",
            "Epoch 249/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6334 - accuracy: 0.7568 - val_loss: 0.5566 - val_accuracy: 0.8041\n",
            "Epoch 250/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6322 - accuracy: 0.7652 - val_loss: 0.5696 - val_accuracy: 0.7973\n",
            "Epoch 251/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6328 - accuracy: 0.7534 - val_loss: 0.5735 - val_accuracy: 0.8311\n",
            "Epoch 252/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6274 - accuracy: 0.7568 - val_loss: 0.5678 - val_accuracy: 0.8176\n",
            "Epoch 253/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6949 - accuracy: 0.7230 - val_loss: 0.5742 - val_accuracy: 0.8176\n",
            "Epoch 254/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6273 - accuracy: 0.7652 - val_loss: 0.5731 - val_accuracy: 0.7905\n",
            "Epoch 255/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6059 - accuracy: 0.7753 - val_loss: 0.5615 - val_accuracy: 0.8108\n",
            "Epoch 256/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6752 - accuracy: 0.7483 - val_loss: 0.5714 - val_accuracy: 0.8176\n",
            "Epoch 257/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6354 - accuracy: 0.7551 - val_loss: 0.5580 - val_accuracy: 0.8311\n",
            "Epoch 258/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6009 - accuracy: 0.7787 - val_loss: 0.5331 - val_accuracy: 0.8243\n",
            "Epoch 259/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6713 - accuracy: 0.7669 - val_loss: 0.5633 - val_accuracy: 0.8041\n",
            "Epoch 260/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6345 - accuracy: 0.7517 - val_loss: 0.5486 - val_accuracy: 0.8108\n",
            "Epoch 261/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6064 - accuracy: 0.7618 - val_loss: 0.5625 - val_accuracy: 0.7838\n",
            "Epoch 262/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6306 - accuracy: 0.7736 - val_loss: 0.5924 - val_accuracy: 0.7838\n",
            "Epoch 263/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5863 - accuracy: 0.8007 - val_loss: 0.5605 - val_accuracy: 0.7770\n",
            "Epoch 264/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6480 - accuracy: 0.7686 - val_loss: 0.5431 - val_accuracy: 0.7838\n",
            "Epoch 265/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6575 - accuracy: 0.7280 - val_loss: 0.5436 - val_accuracy: 0.8108\n",
            "Epoch 266/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6029 - accuracy: 0.7534 - val_loss: 0.5326 - val_accuracy: 0.8108\n",
            "Epoch 267/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6389 - accuracy: 0.7618 - val_loss: 0.5382 - val_accuracy: 0.8176\n",
            "Epoch 268/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6247 - accuracy: 0.7551 - val_loss: 0.5403 - val_accuracy: 0.8108\n",
            "Epoch 269/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6068 - accuracy: 0.7483 - val_loss: 0.5500 - val_accuracy: 0.8108\n",
            "Epoch 270/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6072 - accuracy: 0.7534 - val_loss: 0.5438 - val_accuracy: 0.8108\n",
            "Epoch 271/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6399 - accuracy: 0.7517 - val_loss: 0.5474 - val_accuracy: 0.7905\n",
            "Epoch 272/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6227 - accuracy: 0.7331 - val_loss: 0.5714 - val_accuracy: 0.7905\n",
            "Epoch 273/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6124 - accuracy: 0.7872 - val_loss: 0.5396 - val_accuracy: 0.8243\n",
            "Epoch 274/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6101 - accuracy: 0.7483 - val_loss: 0.5559 - val_accuracy: 0.8108\n",
            "Epoch 275/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5878 - accuracy: 0.7905 - val_loss: 0.5316 - val_accuracy: 0.8041\n",
            "Epoch 276/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6159 - accuracy: 0.7686 - val_loss: 0.5366 - val_accuracy: 0.8108\n",
            "Epoch 277/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6441 - accuracy: 0.7449 - val_loss: 0.5421 - val_accuracy: 0.8108\n",
            "Epoch 278/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5989 - accuracy: 0.7821 - val_loss: 0.5480 - val_accuracy: 0.8041\n",
            "Epoch 279/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6256 - accuracy: 0.7584 - val_loss: 0.5333 - val_accuracy: 0.8108\n",
            "Epoch 280/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6465 - accuracy: 0.7500 - val_loss: 0.5482 - val_accuracy: 0.8041\n",
            "Epoch 281/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6180 - accuracy: 0.7601 - val_loss: 0.5380 - val_accuracy: 0.8176\n",
            "Epoch 282/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5921 - accuracy: 0.7669 - val_loss: 0.5441 - val_accuracy: 0.7905\n",
            "Epoch 283/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5964 - accuracy: 0.7990 - val_loss: 0.5330 - val_accuracy: 0.8108\n",
            "Epoch 284/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6135 - accuracy: 0.7517 - val_loss: 0.5432 - val_accuracy: 0.7973\n",
            "Epoch 285/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6374 - accuracy: 0.7449 - val_loss: 0.5298 - val_accuracy: 0.8176\n",
            "Epoch 286/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6013 - accuracy: 0.7652 - val_loss: 0.5235 - val_accuracy: 0.8176\n",
            "Epoch 287/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6257 - accuracy: 0.7686 - val_loss: 0.5341 - val_accuracy: 0.7905\n",
            "Epoch 288/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5544 - accuracy: 0.7821 - val_loss: 0.5200 - val_accuracy: 0.7973\n",
            "Epoch 289/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6662 - accuracy: 0.7280 - val_loss: 0.5616 - val_accuracy: 0.8243\n",
            "Epoch 290/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6051 - accuracy: 0.7618 - val_loss: 0.5570 - val_accuracy: 0.8243\n",
            "Epoch 291/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5800 - accuracy: 0.7770 - val_loss: 0.5374 - val_accuracy: 0.8041\n",
            "Epoch 292/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6645 - accuracy: 0.7382 - val_loss: 0.5422 - val_accuracy: 0.8108\n",
            "Epoch 293/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6340 - accuracy: 0.7500 - val_loss: 0.5448 - val_accuracy: 0.8243\n",
            "Epoch 294/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6036 - accuracy: 0.7551 - val_loss: 0.5451 - val_accuracy: 0.8041\n",
            "Epoch 295/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5746 - accuracy: 0.7703 - val_loss: 0.5484 - val_accuracy: 0.8176\n",
            "Epoch 296/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5930 - accuracy: 0.7534 - val_loss: 0.5532 - val_accuracy: 0.8108\n",
            "Epoch 297/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6397 - accuracy: 0.7314 - val_loss: 0.5590 - val_accuracy: 0.7770\n",
            "Epoch 298/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6202 - accuracy: 0.7720 - val_loss: 0.5377 - val_accuracy: 0.7905\n",
            "Epoch 299/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5926 - accuracy: 0.7821 - val_loss: 0.5325 - val_accuracy: 0.8108\n",
            "Epoch 300/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6142 - accuracy: 0.7534 - val_loss: 0.5362 - val_accuracy: 0.7905\n",
            "Epoch 301/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5962 - accuracy: 0.7720 - val_loss: 0.5381 - val_accuracy: 0.8041\n",
            "Epoch 302/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6480 - accuracy: 0.7280 - val_loss: 0.5577 - val_accuracy: 0.7905\n",
            "Epoch 303/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6520 - accuracy: 0.7280 - val_loss: 0.5427 - val_accuracy: 0.8176\n",
            "Epoch 304/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6117 - accuracy: 0.7669 - val_loss: 0.5480 - val_accuracy: 0.8041\n",
            "Epoch 305/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6261 - accuracy: 0.7517 - val_loss: 0.5501 - val_accuracy: 0.8041\n",
            "Epoch 306/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5990 - accuracy: 0.7601 - val_loss: 0.5606 - val_accuracy: 0.8243\n",
            "Epoch 307/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6356 - accuracy: 0.7517 - val_loss: 0.5700 - val_accuracy: 0.7905\n",
            "Epoch 308/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6469 - accuracy: 0.7399 - val_loss: 0.5453 - val_accuracy: 0.7838\n",
            "Epoch 309/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6325 - accuracy: 0.7416 - val_loss: 0.5350 - val_accuracy: 0.8176\n",
            "Epoch 310/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6461 - accuracy: 0.7162 - val_loss: 0.5547 - val_accuracy: 0.8243\n",
            "Epoch 311/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6292 - accuracy: 0.7517 - val_loss: 0.5377 - val_accuracy: 0.8041\n",
            "Epoch 312/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6112 - accuracy: 0.7416 - val_loss: 0.5364 - val_accuracy: 0.8108\n",
            "Epoch 313/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5983 - accuracy: 0.7787 - val_loss: 0.5295 - val_accuracy: 0.8041\n",
            "Epoch 314/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6054 - accuracy: 0.7601 - val_loss: 0.5413 - val_accuracy: 0.7973\n",
            "Epoch 315/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6146 - accuracy: 0.7804 - val_loss: 0.5455 - val_accuracy: 0.8176\n",
            "Epoch 316/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6002 - accuracy: 0.7635 - val_loss: 0.5551 - val_accuracy: 0.8041\n",
            "Epoch 317/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5783 - accuracy: 0.7753 - val_loss: 0.5414 - val_accuracy: 0.7973\n",
            "Epoch 318/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6133 - accuracy: 0.7889 - val_loss: 0.5561 - val_accuracy: 0.7973\n",
            "Epoch 319/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5969 - accuracy: 0.7855 - val_loss: 0.5671 - val_accuracy: 0.7973\n",
            "Epoch 320/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6137 - accuracy: 0.7466 - val_loss: 0.5417 - val_accuracy: 0.8041\n",
            "Epoch 321/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6046 - accuracy: 0.7686 - val_loss: 0.5421 - val_accuracy: 0.7905\n",
            "Epoch 322/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5918 - accuracy: 0.7584 - val_loss: 0.5269 - val_accuracy: 0.8041\n",
            "Epoch 323/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5811 - accuracy: 0.7500 - val_loss: 0.5474 - val_accuracy: 0.8041\n",
            "Epoch 324/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5730 - accuracy: 0.7905 - val_loss: 0.5501 - val_accuracy: 0.8176\n",
            "Epoch 325/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6385 - accuracy: 0.7601 - val_loss: 0.5487 - val_accuracy: 0.8176\n",
            "Epoch 326/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6242 - accuracy: 0.7669 - val_loss: 0.5571 - val_accuracy: 0.8176\n",
            "Epoch 327/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6085 - accuracy: 0.7652 - val_loss: 0.5560 - val_accuracy: 0.8243\n",
            "Epoch 328/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6532 - accuracy: 0.7584 - val_loss: 0.5912 - val_accuracy: 0.7770\n",
            "Epoch 329/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5911 - accuracy: 0.7601 - val_loss: 0.5455 - val_accuracy: 0.7973\n",
            "Epoch 330/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5942 - accuracy: 0.7686 - val_loss: 0.5574 - val_accuracy: 0.7770\n",
            "Epoch 331/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6682 - accuracy: 0.7382 - val_loss: 0.5646 - val_accuracy: 0.7635\n",
            "Epoch 332/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6321 - accuracy: 0.7686 - val_loss: 0.5528 - val_accuracy: 0.7770\n",
            "Epoch 333/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6468 - accuracy: 0.7382 - val_loss: 0.5664 - val_accuracy: 0.7703\n",
            "Epoch 334/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5858 - accuracy: 0.7517 - val_loss: 0.5559 - val_accuracy: 0.7838\n",
            "Epoch 335/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6313 - accuracy: 0.7703 - val_loss: 0.5577 - val_accuracy: 0.8041\n",
            "Epoch 336/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6608 - accuracy: 0.7618 - val_loss: 0.5556 - val_accuracy: 0.7973\n",
            "Epoch 337/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6165 - accuracy: 0.7703 - val_loss: 0.5447 - val_accuracy: 0.8041\n",
            "Epoch 338/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6339 - accuracy: 0.7568 - val_loss: 0.5511 - val_accuracy: 0.7973\n",
            "Epoch 339/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5798 - accuracy: 0.7787 - val_loss: 0.5301 - val_accuracy: 0.8176\n",
            "Epoch 340/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5858 - accuracy: 0.7652 - val_loss: 0.5507 - val_accuracy: 0.7973\n",
            "Epoch 341/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6161 - accuracy: 0.7652 - val_loss: 0.5593 - val_accuracy: 0.8041\n",
            "Epoch 342/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5932 - accuracy: 0.7770 - val_loss: 0.5444 - val_accuracy: 0.7973\n",
            "Epoch 343/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6045 - accuracy: 0.7753 - val_loss: 0.5600 - val_accuracy: 0.7905\n",
            "Epoch 344/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6246 - accuracy: 0.7416 - val_loss: 0.5872 - val_accuracy: 0.7838\n",
            "Epoch 345/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6398 - accuracy: 0.7314 - val_loss: 0.5489 - val_accuracy: 0.7770\n",
            "Epoch 346/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6332 - accuracy: 0.7601 - val_loss: 0.5398 - val_accuracy: 0.8176\n",
            "Epoch 347/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5847 - accuracy: 0.7686 - val_loss: 0.5624 - val_accuracy: 0.7905\n",
            "Epoch 348/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6436 - accuracy: 0.7399 - val_loss: 0.5545 - val_accuracy: 0.7973\n",
            "Epoch 349/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6188 - accuracy: 0.7500 - val_loss: 0.5348 - val_accuracy: 0.7838\n",
            "Epoch 350/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6291 - accuracy: 0.7517 - val_loss: 0.5502 - val_accuracy: 0.7770\n",
            "Epoch 351/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5889 - accuracy: 0.7770 - val_loss: 0.5271 - val_accuracy: 0.8176\n",
            "Epoch 352/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6509 - accuracy: 0.7314 - val_loss: 0.5561 - val_accuracy: 0.8176\n",
            "Epoch 353/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5853 - accuracy: 0.7703 - val_loss: 0.5227 - val_accuracy: 0.8041\n",
            "Epoch 354/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5739 - accuracy: 0.7669 - val_loss: 0.5267 - val_accuracy: 0.7838\n",
            "Epoch 355/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5982 - accuracy: 0.7821 - val_loss: 0.5541 - val_accuracy: 0.7770\n",
            "Epoch 356/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6265 - accuracy: 0.7635 - val_loss: 0.5552 - val_accuracy: 0.7838\n",
            "Epoch 357/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6189 - accuracy: 0.7601 - val_loss: 0.5424 - val_accuracy: 0.7973\n",
            "Epoch 358/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5906 - accuracy: 0.7686 - val_loss: 0.5422 - val_accuracy: 0.7838\n",
            "Epoch 359/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6085 - accuracy: 0.7534 - val_loss: 0.5355 - val_accuracy: 0.8041\n",
            "Epoch 360/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6081 - accuracy: 0.7652 - val_loss: 0.5336 - val_accuracy: 0.7770\n",
            "Epoch 361/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6127 - accuracy: 0.7703 - val_loss: 0.5177 - val_accuracy: 0.8041\n",
            "Epoch 362/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6422 - accuracy: 0.7449 - val_loss: 0.5472 - val_accuracy: 0.8041\n",
            "Epoch 363/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6022 - accuracy: 0.7686 - val_loss: 0.5599 - val_accuracy: 0.7703\n",
            "Epoch 364/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6460 - accuracy: 0.7551 - val_loss: 0.5490 - val_accuracy: 0.7973\n",
            "Epoch 365/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5950 - accuracy: 0.7703 - val_loss: 0.5174 - val_accuracy: 0.7973\n",
            "Epoch 366/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6092 - accuracy: 0.7449 - val_loss: 0.5308 - val_accuracy: 0.8108\n",
            "Epoch 367/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6018 - accuracy: 0.7838 - val_loss: 0.5457 - val_accuracy: 0.8108\n",
            "Epoch 368/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6135 - accuracy: 0.7449 - val_loss: 0.5502 - val_accuracy: 0.7973\n",
            "Epoch 369/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6143 - accuracy: 0.7584 - val_loss: 0.5268 - val_accuracy: 0.8108\n",
            "Epoch 370/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5545 - accuracy: 0.7872 - val_loss: 0.5301 - val_accuracy: 0.8041\n",
            "Epoch 371/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6116 - accuracy: 0.7703 - val_loss: 0.5548 - val_accuracy: 0.8108\n",
            "Epoch 372/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6198 - accuracy: 0.7568 - val_loss: 0.5635 - val_accuracy: 0.7838\n",
            "Epoch 373/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6180 - accuracy: 0.7551 - val_loss: 0.5346 - val_accuracy: 0.8108\n",
            "Epoch 374/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6105 - accuracy: 0.7703 - val_loss: 0.5237 - val_accuracy: 0.8243\n",
            "Epoch 375/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6052 - accuracy: 0.7703 - val_loss: 0.5222 - val_accuracy: 0.8378\n",
            "Epoch 376/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5897 - accuracy: 0.7753 - val_loss: 0.5090 - val_accuracy: 0.8243\n",
            "Epoch 377/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5672 - accuracy: 0.7804 - val_loss: 0.5054 - val_accuracy: 0.7905\n",
            "Epoch 378/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5767 - accuracy: 0.7973 - val_loss: 0.4937 - val_accuracy: 0.8176\n",
            "Epoch 379/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5849 - accuracy: 0.7584 - val_loss: 0.5290 - val_accuracy: 0.8041\n",
            "Epoch 380/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5783 - accuracy: 0.7703 - val_loss: 0.5138 - val_accuracy: 0.8243\n",
            "Epoch 381/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5842 - accuracy: 0.7618 - val_loss: 0.5351 - val_accuracy: 0.8108\n",
            "Epoch 382/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6136 - accuracy: 0.7770 - val_loss: 0.5348 - val_accuracy: 0.8108\n",
            "Epoch 383/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5359 - accuracy: 0.8074 - val_loss: 0.5087 - val_accuracy: 0.8108\n",
            "Epoch 384/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5769 - accuracy: 0.7838 - val_loss: 0.5485 - val_accuracy: 0.8041\n",
            "Epoch 385/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6539 - accuracy: 0.7416 - val_loss: 0.5535 - val_accuracy: 0.8243\n",
            "Epoch 386/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5758 - accuracy: 0.7753 - val_loss: 0.5306 - val_accuracy: 0.8041\n",
            "Epoch 387/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5879 - accuracy: 0.7770 - val_loss: 0.5365 - val_accuracy: 0.8041\n",
            "Epoch 388/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5696 - accuracy: 0.7905 - val_loss: 0.5279 - val_accuracy: 0.8108\n",
            "Epoch 389/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5727 - accuracy: 0.7804 - val_loss: 0.5470 - val_accuracy: 0.7973\n",
            "Epoch 390/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6383 - accuracy: 0.7635 - val_loss: 0.5588 - val_accuracy: 0.7973\n",
            "Epoch 391/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5722 - accuracy: 0.8024 - val_loss: 0.5278 - val_accuracy: 0.7770\n",
            "Epoch 392/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6218 - accuracy: 0.7804 - val_loss: 0.5251 - val_accuracy: 0.7838\n",
            "Epoch 393/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6007 - accuracy: 0.7686 - val_loss: 0.5454 - val_accuracy: 0.7838\n",
            "Epoch 394/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5831 - accuracy: 0.7855 - val_loss: 0.5203 - val_accuracy: 0.7770\n",
            "Epoch 395/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6020 - accuracy: 0.7669 - val_loss: 0.5241 - val_accuracy: 0.8041\n",
            "Epoch 396/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5943 - accuracy: 0.7736 - val_loss: 0.5341 - val_accuracy: 0.7770\n",
            "Epoch 397/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5761 - accuracy: 0.7720 - val_loss: 0.5275 - val_accuracy: 0.7838\n",
            "Epoch 398/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6196 - accuracy: 0.7720 - val_loss: 0.5459 - val_accuracy: 0.7770\n",
            "Epoch 399/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6149 - accuracy: 0.7618 - val_loss: 0.5307 - val_accuracy: 0.7905\n",
            "Epoch 400/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5784 - accuracy: 0.7669 - val_loss: 0.5323 - val_accuracy: 0.8108\n",
            "Epoch 401/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6368 - accuracy: 0.7534 - val_loss: 0.5512 - val_accuracy: 0.8041\n",
            "Epoch 402/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6295 - accuracy: 0.7568 - val_loss: 0.5223 - val_accuracy: 0.8243\n",
            "Epoch 403/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5883 - accuracy: 0.7720 - val_loss: 0.5112 - val_accuracy: 0.7973\n",
            "Epoch 404/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5726 - accuracy: 0.7601 - val_loss: 0.5095 - val_accuracy: 0.8108\n",
            "Epoch 405/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5656 - accuracy: 0.8091 - val_loss: 0.5130 - val_accuracy: 0.8378\n",
            "Epoch 406/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5837 - accuracy: 0.7821 - val_loss: 0.5127 - val_accuracy: 0.8378\n",
            "Epoch 407/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5967 - accuracy: 0.7449 - val_loss: 0.5339 - val_accuracy: 0.7838\n",
            "Epoch 408/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5903 - accuracy: 0.7466 - val_loss: 0.5219 - val_accuracy: 0.7905\n",
            "Epoch 409/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6001 - accuracy: 0.7720 - val_loss: 0.5163 - val_accuracy: 0.7905\n",
            "Epoch 410/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5564 - accuracy: 0.7905 - val_loss: 0.5402 - val_accuracy: 0.8041\n",
            "Epoch 411/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6113 - accuracy: 0.7568 - val_loss: 0.5554 - val_accuracy: 0.8311\n",
            "Epoch 412/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6072 - accuracy: 0.7635 - val_loss: 0.5435 - val_accuracy: 0.7905\n",
            "Epoch 413/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5957 - accuracy: 0.7669 - val_loss: 0.5226 - val_accuracy: 0.7973\n",
            "Epoch 414/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6394 - accuracy: 0.7466 - val_loss: 0.5378 - val_accuracy: 0.8041\n",
            "Epoch 415/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5919 - accuracy: 0.7703 - val_loss: 0.5190 - val_accuracy: 0.7973\n",
            "Epoch 416/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6270 - accuracy: 0.7280 - val_loss: 0.5231 - val_accuracy: 0.7973\n",
            "Epoch 417/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6299 - accuracy: 0.7483 - val_loss: 0.5345 - val_accuracy: 0.8041\n",
            "Epoch 418/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6125 - accuracy: 0.7703 - val_loss: 0.5283 - val_accuracy: 0.8041\n",
            "Epoch 419/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5770 - accuracy: 0.7568 - val_loss: 0.5097 - val_accuracy: 0.8176\n",
            "Epoch 420/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5856 - accuracy: 0.7584 - val_loss: 0.5176 - val_accuracy: 0.8041\n",
            "Epoch 421/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5763 - accuracy: 0.7889 - val_loss: 0.5206 - val_accuracy: 0.8176\n",
            "Epoch 422/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5542 - accuracy: 0.7855 - val_loss: 0.5112 - val_accuracy: 0.8108\n",
            "Epoch 423/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5971 - accuracy: 0.7568 - val_loss: 0.5324 - val_accuracy: 0.8108\n",
            "Epoch 424/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6042 - accuracy: 0.7703 - val_loss: 0.5495 - val_accuracy: 0.7770\n",
            "Epoch 425/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6045 - accuracy: 0.7686 - val_loss: 0.5404 - val_accuracy: 0.7635\n",
            "Epoch 426/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5886 - accuracy: 0.7720 - val_loss: 0.5118 - val_accuracy: 0.8041\n",
            "Epoch 427/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5818 - accuracy: 0.7466 - val_loss: 0.5228 - val_accuracy: 0.8041\n",
            "Epoch 428/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5848 - accuracy: 0.7720 - val_loss: 0.5166 - val_accuracy: 0.8311\n",
            "Epoch 429/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6342 - accuracy: 0.7365 - val_loss: 0.5337 - val_accuracy: 0.8108\n",
            "Epoch 430/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5803 - accuracy: 0.7770 - val_loss: 0.5248 - val_accuracy: 0.7770\n",
            "Epoch 431/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6048 - accuracy: 0.7416 - val_loss: 0.5342 - val_accuracy: 0.8176\n",
            "Epoch 432/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5796 - accuracy: 0.7889 - val_loss: 0.5212 - val_accuracy: 0.8041\n",
            "Epoch 433/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6106 - accuracy: 0.7686 - val_loss: 0.5307 - val_accuracy: 0.8041\n",
            "Epoch 434/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6144 - accuracy: 0.7652 - val_loss: 0.5399 - val_accuracy: 0.7973\n",
            "Epoch 435/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6199 - accuracy: 0.7534 - val_loss: 0.5509 - val_accuracy: 0.7905\n",
            "Epoch 436/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6471 - accuracy: 0.7365 - val_loss: 0.5423 - val_accuracy: 0.7973\n",
            "Epoch 437/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6154 - accuracy: 0.7500 - val_loss: 0.5392 - val_accuracy: 0.7905\n",
            "Epoch 438/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6091 - accuracy: 0.7517 - val_loss: 0.5632 - val_accuracy: 0.8041\n",
            "Epoch 439/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6062 - accuracy: 0.7584 - val_loss: 0.5338 - val_accuracy: 0.7838\n",
            "Epoch 440/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6156 - accuracy: 0.7551 - val_loss: 0.5456 - val_accuracy: 0.7905\n",
            "Epoch 441/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5651 - accuracy: 0.7753 - val_loss: 0.5241 - val_accuracy: 0.7635\n",
            "Epoch 442/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5723 - accuracy: 0.7669 - val_loss: 0.5261 - val_accuracy: 0.7905\n",
            "Epoch 443/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6373 - accuracy: 0.7365 - val_loss: 0.5448 - val_accuracy: 0.8041\n",
            "Epoch 444/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6016 - accuracy: 0.7736 - val_loss: 0.5286 - val_accuracy: 0.7905\n",
            "Epoch 445/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5660 - accuracy: 0.7652 - val_loss: 0.5317 - val_accuracy: 0.8108\n",
            "Epoch 446/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6101 - accuracy: 0.7551 - val_loss: 0.5291 - val_accuracy: 0.7973\n",
            "Epoch 447/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5755 - accuracy: 0.7686 - val_loss: 0.5309 - val_accuracy: 0.7973\n",
            "Epoch 448/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5903 - accuracy: 0.7517 - val_loss: 0.5136 - val_accuracy: 0.8108\n",
            "Epoch 449/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5488 - accuracy: 0.7956 - val_loss: 0.5292 - val_accuracy: 0.8041\n",
            "Epoch 450/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5599 - accuracy: 0.7872 - val_loss: 0.5198 - val_accuracy: 0.7905\n",
            "Epoch 451/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5902 - accuracy: 0.7686 - val_loss: 0.5220 - val_accuracy: 0.7973\n",
            "Epoch 452/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5727 - accuracy: 0.7804 - val_loss: 0.5226 - val_accuracy: 0.7703\n",
            "Epoch 453/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5991 - accuracy: 0.7753 - val_loss: 0.5380 - val_accuracy: 0.7838\n",
            "Epoch 454/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6305 - accuracy: 0.7483 - val_loss: 0.5415 - val_accuracy: 0.7838\n",
            "Epoch 455/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6001 - accuracy: 0.7483 - val_loss: 0.5153 - val_accuracy: 0.7703\n",
            "Epoch 456/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5720 - accuracy: 0.7922 - val_loss: 0.5272 - val_accuracy: 0.7770\n",
            "Epoch 457/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6499 - accuracy: 0.7584 - val_loss: 0.5475 - val_accuracy: 0.7973\n",
            "Epoch 458/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6164 - accuracy: 0.7635 - val_loss: 0.5341 - val_accuracy: 0.7905\n",
            "Epoch 459/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5789 - accuracy: 0.7601 - val_loss: 0.5325 - val_accuracy: 0.7770\n",
            "Epoch 460/500\n",
            "74/74 [==============================] - 1s 8ms/step - loss: 0.6379 - accuracy: 0.7382 - val_loss: 0.5563 - val_accuracy: 0.7635\n",
            "Epoch 461/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6040 - accuracy: 0.7770 - val_loss: 0.5649 - val_accuracy: 0.7635\n",
            "Epoch 462/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6224 - accuracy: 0.7635 - val_loss: 0.5392 - val_accuracy: 0.7905\n",
            "Epoch 463/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6150 - accuracy: 0.7247 - val_loss: 0.5246 - val_accuracy: 0.7838\n",
            "Epoch 464/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6391 - accuracy: 0.7382 - val_loss: 0.5478 - val_accuracy: 0.7838\n",
            "Epoch 465/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5992 - accuracy: 0.7449 - val_loss: 0.5347 - val_accuracy: 0.7905\n",
            "Epoch 466/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5749 - accuracy: 0.8024 - val_loss: 0.5272 - val_accuracy: 0.7905\n",
            "Epoch 467/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5830 - accuracy: 0.7787 - val_loss: 0.5458 - val_accuracy: 0.7703\n",
            "Epoch 468/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6088 - accuracy: 0.7618 - val_loss: 0.5607 - val_accuracy: 0.7703\n",
            "Epoch 469/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6040 - accuracy: 0.7568 - val_loss: 0.5477 - val_accuracy: 0.7905\n",
            "Epoch 470/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5977 - accuracy: 0.7618 - val_loss: 0.5292 - val_accuracy: 0.8108\n",
            "Epoch 471/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5804 - accuracy: 0.7551 - val_loss: 0.5498 - val_accuracy: 0.8243\n",
            "Epoch 472/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6407 - accuracy: 0.7331 - val_loss: 0.5464 - val_accuracy: 0.7905\n",
            "Epoch 473/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6083 - accuracy: 0.7517 - val_loss: 0.5441 - val_accuracy: 0.7770\n",
            "Epoch 474/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5833 - accuracy: 0.7517 - val_loss: 0.5102 - val_accuracy: 0.8108\n",
            "Epoch 475/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5698 - accuracy: 0.7635 - val_loss: 0.5215 - val_accuracy: 0.7973\n",
            "Epoch 476/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6038 - accuracy: 0.7652 - val_loss: 0.5265 - val_accuracy: 0.8243\n",
            "Epoch 477/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6137 - accuracy: 0.7584 - val_loss: 0.5388 - val_accuracy: 0.7973\n",
            "Epoch 478/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5688 - accuracy: 0.7720 - val_loss: 0.5281 - val_accuracy: 0.7905\n",
            "Epoch 479/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5737 - accuracy: 0.7990 - val_loss: 0.5188 - val_accuracy: 0.7973\n",
            "Epoch 480/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6075 - accuracy: 0.7686 - val_loss: 0.5391 - val_accuracy: 0.8041\n",
            "Epoch 481/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5978 - accuracy: 0.7568 - val_loss: 0.5435 - val_accuracy: 0.8108\n",
            "Epoch 482/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5716 - accuracy: 0.7568 - val_loss: 0.5307 - val_accuracy: 0.8108\n",
            "Epoch 483/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5723 - accuracy: 0.7720 - val_loss: 0.5367 - val_accuracy: 0.7905\n",
            "Epoch 484/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5596 - accuracy: 0.7635 - val_loss: 0.5473 - val_accuracy: 0.7973\n",
            "Epoch 485/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5600 - accuracy: 0.7601 - val_loss: 0.5420 - val_accuracy: 0.7973\n",
            "Epoch 486/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5984 - accuracy: 0.7466 - val_loss: 0.5353 - val_accuracy: 0.8176\n",
            "Epoch 487/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5728 - accuracy: 0.7736 - val_loss: 0.5410 - val_accuracy: 0.7973\n",
            "Epoch 488/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5783 - accuracy: 0.7787 - val_loss: 0.5414 - val_accuracy: 0.7973\n",
            "Epoch 489/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5813 - accuracy: 0.7905 - val_loss: 0.5421 - val_accuracy: 0.8041\n",
            "Epoch 490/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6284 - accuracy: 0.7584 - val_loss: 0.5386 - val_accuracy: 0.7973\n",
            "Epoch 491/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6252 - accuracy: 0.7365 - val_loss: 0.5423 - val_accuracy: 0.7905\n",
            "Epoch 492/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6035 - accuracy: 0.7551 - val_loss: 0.5499 - val_accuracy: 0.7973\n",
            "Epoch 493/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6125 - accuracy: 0.7635 - val_loss: 0.5440 - val_accuracy: 0.8243\n",
            "Epoch 494/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5618 - accuracy: 0.7821 - val_loss: 0.5283 - val_accuracy: 0.8108\n",
            "Epoch 495/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5950 - accuracy: 0.7855 - val_loss: 0.5544 - val_accuracy: 0.7973\n",
            "Epoch 496/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6200 - accuracy: 0.7770 - val_loss: 0.5545 - val_accuracy: 0.7905\n",
            "Epoch 497/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5887 - accuracy: 0.7736 - val_loss: 0.5428 - val_accuracy: 0.7838\n",
            "Epoch 498/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5668 - accuracy: 0.7939 - val_loss: 0.5380 - val_accuracy: 0.7973\n",
            "Epoch 499/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5950 - accuracy: 0.7855 - val_loss: 0.5392 - val_accuracy: 0.8041\n",
            "Epoch 500/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5968 - accuracy: 0.7686 - val_loss: 0.5452 - val_accuracy: 0.8108\n",
            "5/5 [==============================] - 0s 3ms/step - loss: 0.5452 - accuracy: 0.8108\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:62: DeprecationWarning: scipy.interp is deprecated and will be removed in SciPy 2.0.0, use numpy.interp instead\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/500\n",
            "74/74 [==============================] - 1s 11ms/step - loss: 1.6459 - accuracy: 0.5203 - val_loss: 1.3424 - val_accuracy: 0.5203\n",
            "Epoch 2/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 1.3370 - accuracy: 0.5389 - val_loss: 1.2454 - val_accuracy: 0.6351\n",
            "Epoch 3/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 1.2228 - accuracy: 0.5439 - val_loss: 1.1495 - val_accuracy: 0.5338\n",
            "Epoch 4/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 1.1196 - accuracy: 0.4966 - val_loss: 1.0695 - val_accuracy: 0.5541\n",
            "Epoch 5/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 1.0265 - accuracy: 0.5389 - val_loss: 1.0120 - val_accuracy: 0.4865\n",
            "Epoch 6/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.9807 - accuracy: 0.5355 - val_loss: 0.9729 - val_accuracy: 0.5608\n",
            "Epoch 7/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.9505 - accuracy: 0.5541 - val_loss: 0.9181 - val_accuracy: 0.6351\n",
            "Epoch 8/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.9168 - accuracy: 0.5794 - val_loss: 0.8845 - val_accuracy: 0.6419\n",
            "Epoch 9/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.8989 - accuracy: 0.6030 - val_loss: 0.8747 - val_accuracy: 0.6689\n",
            "Epoch 10/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.8661 - accuracy: 0.6182 - val_loss: 0.8401 - val_accuracy: 0.7230\n",
            "Epoch 11/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8672 - accuracy: 0.6064 - val_loss: 0.8278 - val_accuracy: 0.7500\n",
            "Epoch 12/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8577 - accuracy: 0.6233 - val_loss: 0.8154 - val_accuracy: 0.7095\n",
            "Epoch 13/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8383 - accuracy: 0.6351 - val_loss: 0.8095 - val_accuracy: 0.6892\n",
            "Epoch 14/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7956 - accuracy: 0.6757 - val_loss: 0.7596 - val_accuracy: 0.7297\n",
            "Epoch 15/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.8228 - accuracy: 0.6689 - val_loss: 0.7673 - val_accuracy: 0.6892\n",
            "Epoch 16/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.8390 - accuracy: 0.6486 - val_loss: 0.7693 - val_accuracy: 0.7297\n",
            "Epoch 17/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8049 - accuracy: 0.6841 - val_loss: 0.7610 - val_accuracy: 0.7500\n",
            "Epoch 18/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8421 - accuracy: 0.6622 - val_loss: 0.7732 - val_accuracy: 0.7230\n",
            "Epoch 19/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7974 - accuracy: 0.7280 - val_loss: 0.7623 - val_accuracy: 0.7432\n",
            "Epoch 20/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8551 - accuracy: 0.6655 - val_loss: 0.8175 - val_accuracy: 0.7365\n",
            "Epoch 21/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8234 - accuracy: 0.6909 - val_loss: 0.7873 - val_accuracy: 0.7635\n",
            "Epoch 22/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.8113 - accuracy: 0.6672 - val_loss: 0.8004 - val_accuracy: 0.7297\n",
            "Epoch 23/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.8334 - accuracy: 0.6706 - val_loss: 0.8274 - val_accuracy: 0.7095\n",
            "Epoch 24/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.8151 - accuracy: 0.6892 - val_loss: 0.7821 - val_accuracy: 0.7095\n",
            "Epoch 25/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8548 - accuracy: 0.6419 - val_loss: 0.8056 - val_accuracy: 0.7365\n",
            "Epoch 26/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8035 - accuracy: 0.6824 - val_loss: 0.7429 - val_accuracy: 0.7230\n",
            "Epoch 27/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7768 - accuracy: 0.7230 - val_loss: 0.7659 - val_accuracy: 0.7297\n",
            "Epoch 28/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8313 - accuracy: 0.6774 - val_loss: 0.7429 - val_accuracy: 0.7162\n",
            "Epoch 29/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7711 - accuracy: 0.7382 - val_loss: 0.7360 - val_accuracy: 0.7635\n",
            "Epoch 30/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7648 - accuracy: 0.7027 - val_loss: 0.7430 - val_accuracy: 0.7432\n",
            "Epoch 31/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.8043 - accuracy: 0.7061 - val_loss: 0.7625 - val_accuracy: 0.7297\n",
            "Epoch 32/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7895 - accuracy: 0.7128 - val_loss: 0.7345 - val_accuracy: 0.7297\n",
            "Epoch 33/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7764 - accuracy: 0.7111 - val_loss: 0.7432 - val_accuracy: 0.7500\n",
            "Epoch 34/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7857 - accuracy: 0.6959 - val_loss: 0.7055 - val_accuracy: 0.7635\n",
            "Epoch 35/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7772 - accuracy: 0.7078 - val_loss: 0.6831 - val_accuracy: 0.7703\n",
            "Epoch 36/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7999 - accuracy: 0.6943 - val_loss: 0.7314 - val_accuracy: 0.7432\n",
            "Epoch 37/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7606 - accuracy: 0.7264 - val_loss: 0.7182 - val_accuracy: 0.7500\n",
            "Epoch 38/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8063 - accuracy: 0.7027 - val_loss: 0.7214 - val_accuracy: 0.7973\n",
            "Epoch 39/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7620 - accuracy: 0.7128 - val_loss: 0.6928 - val_accuracy: 0.7770\n",
            "Epoch 40/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7486 - accuracy: 0.7348 - val_loss: 0.7074 - val_accuracy: 0.7568\n",
            "Epoch 41/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7640 - accuracy: 0.7111 - val_loss: 0.7422 - val_accuracy: 0.7568\n",
            "Epoch 42/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8063 - accuracy: 0.6993 - val_loss: 0.7306 - val_accuracy: 0.7568\n",
            "Epoch 43/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7863 - accuracy: 0.7196 - val_loss: 0.6998 - val_accuracy: 0.7500\n",
            "Epoch 44/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7471 - accuracy: 0.7331 - val_loss: 0.7107 - val_accuracy: 0.7230\n",
            "Epoch 45/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7613 - accuracy: 0.7061 - val_loss: 0.7105 - val_accuracy: 0.7568\n",
            "Epoch 46/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7551 - accuracy: 0.7297 - val_loss: 0.7208 - val_accuracy: 0.7162\n",
            "Epoch 47/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7472 - accuracy: 0.7264 - val_loss: 0.7118 - val_accuracy: 0.7095\n",
            "Epoch 48/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7243 - accuracy: 0.7331 - val_loss: 0.7073 - val_accuracy: 0.7365\n",
            "Epoch 49/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7293 - accuracy: 0.7416 - val_loss: 0.7369 - val_accuracy: 0.7297\n",
            "Epoch 50/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7384 - accuracy: 0.7399 - val_loss: 0.6781 - val_accuracy: 0.7703\n",
            "Epoch 51/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7580 - accuracy: 0.7297 - val_loss: 0.7065 - val_accuracy: 0.7432\n",
            "Epoch 52/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7822 - accuracy: 0.6791 - val_loss: 0.7142 - val_accuracy: 0.7432\n",
            "Epoch 53/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7478 - accuracy: 0.7095 - val_loss: 0.7125 - val_accuracy: 0.7297\n",
            "Epoch 54/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7004 - accuracy: 0.7466 - val_loss: 0.7027 - val_accuracy: 0.7500\n",
            "Epoch 55/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7289 - accuracy: 0.7061 - val_loss: 0.7477 - val_accuracy: 0.6892\n",
            "Epoch 56/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7239 - accuracy: 0.7179 - val_loss: 0.6900 - val_accuracy: 0.7500\n",
            "Epoch 57/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6997 - accuracy: 0.7568 - val_loss: 0.6661 - val_accuracy: 0.7297\n",
            "Epoch 58/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7701 - accuracy: 0.6740 - val_loss: 0.6913 - val_accuracy: 0.7432\n",
            "Epoch 59/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7056 - accuracy: 0.7331 - val_loss: 0.6524 - val_accuracy: 0.7568\n",
            "Epoch 60/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7407 - accuracy: 0.7264 - val_loss: 0.6696 - val_accuracy: 0.7297\n",
            "Epoch 61/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6989 - accuracy: 0.7348 - val_loss: 0.6608 - val_accuracy: 0.7432\n",
            "Epoch 62/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7401 - accuracy: 0.7111 - val_loss: 0.6690 - val_accuracy: 0.7432\n",
            "Epoch 63/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7413 - accuracy: 0.7044 - val_loss: 0.6804 - val_accuracy: 0.7568\n",
            "Epoch 64/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6815 - accuracy: 0.7280 - val_loss: 0.6591 - val_accuracy: 0.7297\n",
            "Epoch 65/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7345 - accuracy: 0.7196 - val_loss: 0.6925 - val_accuracy: 0.7568\n",
            "Epoch 66/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7563 - accuracy: 0.7247 - val_loss: 0.6920 - val_accuracy: 0.7568\n",
            "Epoch 67/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7044 - accuracy: 0.7264 - val_loss: 0.7136 - val_accuracy: 0.7230\n",
            "Epoch 68/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7114 - accuracy: 0.7500 - val_loss: 0.6762 - val_accuracy: 0.7568\n",
            "Epoch 69/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7402 - accuracy: 0.7196 - val_loss: 0.6546 - val_accuracy: 0.7703\n",
            "Epoch 70/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7533 - accuracy: 0.7010 - val_loss: 0.6766 - val_accuracy: 0.7500\n",
            "Epoch 71/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7099 - accuracy: 0.7010 - val_loss: 0.6384 - val_accuracy: 0.7365\n",
            "Epoch 72/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6956 - accuracy: 0.7095 - val_loss: 0.6291 - val_accuracy: 0.7500\n",
            "Epoch 73/500\n",
            "74/74 [==============================] - 1s 8ms/step - loss: 0.6679 - accuracy: 0.7432 - val_loss: 0.6241 - val_accuracy: 0.7770\n",
            "Epoch 74/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7203 - accuracy: 0.7061 - val_loss: 0.6445 - val_accuracy: 0.7770\n",
            "Epoch 75/500\n",
            "74/74 [==============================] - 1s 8ms/step - loss: 0.7067 - accuracy: 0.7230 - val_loss: 0.6292 - val_accuracy: 0.7770\n",
            "Epoch 76/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6859 - accuracy: 0.7416 - val_loss: 0.6284 - val_accuracy: 0.7770\n",
            "Epoch 77/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7308 - accuracy: 0.7348 - val_loss: 0.6695 - val_accuracy: 0.7568\n",
            "Epoch 78/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7092 - accuracy: 0.7331 - val_loss: 0.6362 - val_accuracy: 0.7770\n",
            "Epoch 79/500\n",
            "74/74 [==============================] - 1s 8ms/step - loss: 0.6989 - accuracy: 0.7044 - val_loss: 0.6203 - val_accuracy: 0.7905\n",
            "Epoch 80/500\n",
            "74/74 [==============================] - 1s 8ms/step - loss: 0.7193 - accuracy: 0.7078 - val_loss: 0.6100 - val_accuracy: 0.7973\n",
            "Epoch 81/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6573 - accuracy: 0.7466 - val_loss: 0.6090 - val_accuracy: 0.7770\n",
            "Epoch 82/500\n",
            "74/74 [==============================] - 1s 8ms/step - loss: 0.6422 - accuracy: 0.7466 - val_loss: 0.5909 - val_accuracy: 0.7838\n",
            "Epoch 83/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7281 - accuracy: 0.7213 - val_loss: 0.6210 - val_accuracy: 0.7973\n",
            "Epoch 84/500\n",
            "74/74 [==============================] - 1s 8ms/step - loss: 0.7202 - accuracy: 0.7314 - val_loss: 0.6171 - val_accuracy: 0.7905\n",
            "Epoch 85/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6656 - accuracy: 0.7534 - val_loss: 0.5828 - val_accuracy: 0.7905\n",
            "Epoch 86/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6662 - accuracy: 0.7618 - val_loss: 0.5803 - val_accuracy: 0.8041\n",
            "Epoch 87/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6804 - accuracy: 0.7297 - val_loss: 0.6078 - val_accuracy: 0.7905\n",
            "Epoch 88/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7014 - accuracy: 0.7264 - val_loss: 0.6119 - val_accuracy: 0.7703\n",
            "Epoch 89/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6858 - accuracy: 0.7365 - val_loss: 0.5920 - val_accuracy: 0.8108\n",
            "Epoch 90/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6973 - accuracy: 0.7095 - val_loss: 0.6098 - val_accuracy: 0.7635\n",
            "Epoch 91/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6754 - accuracy: 0.7500 - val_loss: 0.6329 - val_accuracy: 0.7568\n",
            "Epoch 92/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6711 - accuracy: 0.7517 - val_loss: 0.6170 - val_accuracy: 0.7905\n",
            "Epoch 93/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6839 - accuracy: 0.7128 - val_loss: 0.5997 - val_accuracy: 0.7905\n",
            "Epoch 94/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6999 - accuracy: 0.7365 - val_loss: 0.5917 - val_accuracy: 0.8041\n",
            "Epoch 95/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7067 - accuracy: 0.7297 - val_loss: 0.5898 - val_accuracy: 0.7905\n",
            "Epoch 96/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6715 - accuracy: 0.7483 - val_loss: 0.5981 - val_accuracy: 0.7703\n",
            "Epoch 97/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6948 - accuracy: 0.7196 - val_loss: 0.6021 - val_accuracy: 0.8041\n",
            "Epoch 98/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6413 - accuracy: 0.7399 - val_loss: 0.5789 - val_accuracy: 0.7973\n",
            "Epoch 99/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6495 - accuracy: 0.7449 - val_loss: 0.5751 - val_accuracy: 0.8176\n",
            "Epoch 100/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6656 - accuracy: 0.7399 - val_loss: 0.5934 - val_accuracy: 0.7568\n",
            "Epoch 101/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6463 - accuracy: 0.7635 - val_loss: 0.6086 - val_accuracy: 0.7703\n",
            "Epoch 102/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6396 - accuracy: 0.7449 - val_loss: 0.5830 - val_accuracy: 0.7905\n",
            "Epoch 103/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6985 - accuracy: 0.7432 - val_loss: 0.6046 - val_accuracy: 0.7838\n",
            "Epoch 104/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6769 - accuracy: 0.7635 - val_loss: 0.5949 - val_accuracy: 0.7905\n",
            "Epoch 105/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6865 - accuracy: 0.7348 - val_loss: 0.6109 - val_accuracy: 0.7703\n",
            "Epoch 106/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6697 - accuracy: 0.7348 - val_loss: 0.6035 - val_accuracy: 0.7838\n",
            "Epoch 107/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6464 - accuracy: 0.7449 - val_loss: 0.5974 - val_accuracy: 0.7905\n",
            "Epoch 108/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7035 - accuracy: 0.7399 - val_loss: 0.6088 - val_accuracy: 0.7973\n",
            "Epoch 109/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6785 - accuracy: 0.7314 - val_loss: 0.5992 - val_accuracy: 0.7905\n",
            "Epoch 110/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6832 - accuracy: 0.7280 - val_loss: 0.6001 - val_accuracy: 0.7973\n",
            "Epoch 111/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6374 - accuracy: 0.7466 - val_loss: 0.5570 - val_accuracy: 0.8041\n",
            "Epoch 112/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6392 - accuracy: 0.7635 - val_loss: 0.5679 - val_accuracy: 0.7838\n",
            "Epoch 113/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6307 - accuracy: 0.7601 - val_loss: 0.6301 - val_accuracy: 0.7635\n",
            "Epoch 114/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7094 - accuracy: 0.7365 - val_loss: 0.5905 - val_accuracy: 0.7905\n",
            "Epoch 115/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6397 - accuracy: 0.7568 - val_loss: 0.5582 - val_accuracy: 0.7838\n",
            "Epoch 116/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6504 - accuracy: 0.7753 - val_loss: 0.6044 - val_accuracy: 0.7770\n",
            "Epoch 117/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6701 - accuracy: 0.7669 - val_loss: 0.5814 - val_accuracy: 0.8041\n",
            "Epoch 118/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6775 - accuracy: 0.7601 - val_loss: 0.5721 - val_accuracy: 0.8581\n",
            "Epoch 119/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6839 - accuracy: 0.7517 - val_loss: 0.5715 - val_accuracy: 0.7905\n",
            "Epoch 120/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6713 - accuracy: 0.7314 - val_loss: 0.5791 - val_accuracy: 0.7703\n",
            "Epoch 121/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6449 - accuracy: 0.7568 - val_loss: 0.5448 - val_accuracy: 0.8243\n",
            "Epoch 122/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5846 - accuracy: 0.7635 - val_loss: 0.5382 - val_accuracy: 0.8243\n",
            "Epoch 123/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6858 - accuracy: 0.7382 - val_loss: 0.5745 - val_accuracy: 0.8378\n",
            "Epoch 124/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6610 - accuracy: 0.7517 - val_loss: 0.5579 - val_accuracy: 0.8446\n",
            "Epoch 125/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6621 - accuracy: 0.7517 - val_loss: 0.5627 - val_accuracy: 0.8108\n",
            "Epoch 126/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6668 - accuracy: 0.7399 - val_loss: 0.5821 - val_accuracy: 0.7838\n",
            "Epoch 127/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6466 - accuracy: 0.7432 - val_loss: 0.5549 - val_accuracy: 0.8176\n",
            "Epoch 128/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6607 - accuracy: 0.7416 - val_loss: 0.5376 - val_accuracy: 0.8378\n",
            "Epoch 129/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6175 - accuracy: 0.7720 - val_loss: 0.5949 - val_accuracy: 0.7635\n",
            "Epoch 130/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6478 - accuracy: 0.7466 - val_loss: 0.5509 - val_accuracy: 0.7838\n",
            "Epoch 131/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6322 - accuracy: 0.7568 - val_loss: 0.5314 - val_accuracy: 0.8311\n",
            "Epoch 132/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6459 - accuracy: 0.7534 - val_loss: 0.5216 - val_accuracy: 0.8108\n",
            "Epoch 133/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6465 - accuracy: 0.7618 - val_loss: 0.5417 - val_accuracy: 0.7973\n",
            "Epoch 134/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6228 - accuracy: 0.7770 - val_loss: 0.5246 - val_accuracy: 0.8108\n",
            "Epoch 135/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6787 - accuracy: 0.7230 - val_loss: 0.5366 - val_accuracy: 0.8243\n",
            "Epoch 136/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6395 - accuracy: 0.7466 - val_loss: 0.5521 - val_accuracy: 0.7838\n",
            "Epoch 137/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6221 - accuracy: 0.7365 - val_loss: 0.5131 - val_accuracy: 0.7905\n",
            "Epoch 138/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6251 - accuracy: 0.7432 - val_loss: 0.5196 - val_accuracy: 0.7838\n",
            "Epoch 139/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6423 - accuracy: 0.7365 - val_loss: 0.5333 - val_accuracy: 0.8041\n",
            "Epoch 140/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6201 - accuracy: 0.7534 - val_loss: 0.5304 - val_accuracy: 0.7973\n",
            "Epoch 141/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5982 - accuracy: 0.7821 - val_loss: 0.4975 - val_accuracy: 0.8243\n",
            "Epoch 142/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6519 - accuracy: 0.7314 - val_loss: 0.5354 - val_accuracy: 0.8243\n",
            "Epoch 143/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6108 - accuracy: 0.7568 - val_loss: 0.5325 - val_accuracy: 0.7838\n",
            "Epoch 144/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5864 - accuracy: 0.7753 - val_loss: 0.5177 - val_accuracy: 0.8041\n",
            "Epoch 145/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6400 - accuracy: 0.7382 - val_loss: 0.5428 - val_accuracy: 0.7838\n",
            "Epoch 146/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6351 - accuracy: 0.7483 - val_loss: 0.5501 - val_accuracy: 0.7703\n",
            "Epoch 147/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6203 - accuracy: 0.7736 - val_loss: 0.5129 - val_accuracy: 0.8176\n",
            "Epoch 148/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6584 - accuracy: 0.7196 - val_loss: 0.5360 - val_accuracy: 0.8108\n",
            "Epoch 149/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6268 - accuracy: 0.7264 - val_loss: 0.5360 - val_accuracy: 0.8108\n",
            "Epoch 150/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6283 - accuracy: 0.7466 - val_loss: 0.5250 - val_accuracy: 0.8378\n",
            "Epoch 151/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6447 - accuracy: 0.7331 - val_loss: 0.5440 - val_accuracy: 0.8041\n",
            "Epoch 152/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6548 - accuracy: 0.7551 - val_loss: 0.5455 - val_accuracy: 0.8243\n",
            "Epoch 153/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6316 - accuracy: 0.7348 - val_loss: 0.5137 - val_accuracy: 0.8243\n",
            "Epoch 154/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5960 - accuracy: 0.7686 - val_loss: 0.5302 - val_accuracy: 0.7838\n",
            "Epoch 155/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6192 - accuracy: 0.7534 - val_loss: 0.5528 - val_accuracy: 0.8041\n",
            "Epoch 156/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6235 - accuracy: 0.7669 - val_loss: 0.5573 - val_accuracy: 0.8041\n",
            "Epoch 157/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6247 - accuracy: 0.7804 - val_loss: 0.5596 - val_accuracy: 0.7838\n",
            "Epoch 158/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6359 - accuracy: 0.7568 - val_loss: 0.5538 - val_accuracy: 0.8176\n",
            "Epoch 159/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6088 - accuracy: 0.7348 - val_loss: 0.5365 - val_accuracy: 0.8108\n",
            "Epoch 160/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6338 - accuracy: 0.7449 - val_loss: 0.5615 - val_accuracy: 0.7770\n",
            "Epoch 161/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5961 - accuracy: 0.7669 - val_loss: 0.5530 - val_accuracy: 0.7770\n",
            "Epoch 162/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6292 - accuracy: 0.7669 - val_loss: 0.5200 - val_accuracy: 0.8514\n",
            "Epoch 163/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6597 - accuracy: 0.7399 - val_loss: 0.5408 - val_accuracy: 0.8243\n",
            "Epoch 164/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6163 - accuracy: 0.7517 - val_loss: 0.5358 - val_accuracy: 0.7905\n",
            "Epoch 165/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6091 - accuracy: 0.7753 - val_loss: 0.5881 - val_accuracy: 0.7500\n",
            "Epoch 166/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6320 - accuracy: 0.7652 - val_loss: 0.5450 - val_accuracy: 0.8446\n",
            "Epoch 167/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6597 - accuracy: 0.7348 - val_loss: 0.5613 - val_accuracy: 0.8243\n",
            "Epoch 168/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6678 - accuracy: 0.7230 - val_loss: 0.5476 - val_accuracy: 0.8108\n",
            "Epoch 169/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6424 - accuracy: 0.7500 - val_loss: 0.5882 - val_accuracy: 0.7635\n",
            "Epoch 170/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6263 - accuracy: 0.7551 - val_loss: 0.5574 - val_accuracy: 0.7973\n",
            "Epoch 171/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5938 - accuracy: 0.7686 - val_loss: 0.5606 - val_accuracy: 0.7703\n",
            "Epoch 172/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5970 - accuracy: 0.7855 - val_loss: 0.5102 - val_accuracy: 0.8041\n",
            "Epoch 173/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6237 - accuracy: 0.7196 - val_loss: 0.5160 - val_accuracy: 0.8378\n",
            "Epoch 174/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6190 - accuracy: 0.7416 - val_loss: 0.5221 - val_accuracy: 0.7973\n",
            "Epoch 175/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6181 - accuracy: 0.7601 - val_loss: 0.5316 - val_accuracy: 0.8041\n",
            "Epoch 176/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6412 - accuracy: 0.7500 - val_loss: 0.5357 - val_accuracy: 0.7973\n",
            "Epoch 177/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5917 - accuracy: 0.7703 - val_loss: 0.5389 - val_accuracy: 0.7905\n",
            "Epoch 178/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5907 - accuracy: 0.7703 - val_loss: 0.5105 - val_accuracy: 0.7973\n",
            "Epoch 179/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6394 - accuracy: 0.7432 - val_loss: 0.5305 - val_accuracy: 0.7905\n",
            "Epoch 180/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6099 - accuracy: 0.7348 - val_loss: 0.5400 - val_accuracy: 0.7973\n",
            "Epoch 181/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6145 - accuracy: 0.7770 - val_loss: 0.5380 - val_accuracy: 0.7770\n",
            "Epoch 182/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6020 - accuracy: 0.7416 - val_loss: 0.5485 - val_accuracy: 0.7703\n",
            "Epoch 183/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5938 - accuracy: 0.7635 - val_loss: 0.5239 - val_accuracy: 0.8041\n",
            "Epoch 184/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6147 - accuracy: 0.7669 - val_loss: 0.5282 - val_accuracy: 0.8041\n",
            "Epoch 185/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6039 - accuracy: 0.7804 - val_loss: 0.5135 - val_accuracy: 0.8176\n",
            "Epoch 186/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6026 - accuracy: 0.7720 - val_loss: 0.5244 - val_accuracy: 0.8311\n",
            "Epoch 187/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6381 - accuracy: 0.7432 - val_loss: 0.5442 - val_accuracy: 0.7905\n",
            "Epoch 188/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6304 - accuracy: 0.7584 - val_loss: 0.5361 - val_accuracy: 0.7770\n",
            "Epoch 189/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6375 - accuracy: 0.7196 - val_loss: 0.5487 - val_accuracy: 0.7838\n",
            "Epoch 190/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6107 - accuracy: 0.7416 - val_loss: 0.5346 - val_accuracy: 0.8108\n",
            "Epoch 191/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6341 - accuracy: 0.7635 - val_loss: 0.5258 - val_accuracy: 0.7905\n",
            "Epoch 192/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6221 - accuracy: 0.7584 - val_loss: 0.5365 - val_accuracy: 0.7973\n",
            "Epoch 193/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5997 - accuracy: 0.7601 - val_loss: 0.5339 - val_accuracy: 0.7770\n",
            "Epoch 194/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5875 - accuracy: 0.7787 - val_loss: 0.5422 - val_accuracy: 0.7770\n",
            "Epoch 195/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5906 - accuracy: 0.7686 - val_loss: 0.5157 - val_accuracy: 0.8041\n",
            "Epoch 196/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6031 - accuracy: 0.7753 - val_loss: 0.5122 - val_accuracy: 0.8243\n",
            "Epoch 197/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6011 - accuracy: 0.7584 - val_loss: 0.5247 - val_accuracy: 0.7770\n",
            "Epoch 198/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6004 - accuracy: 0.7618 - val_loss: 0.5251 - val_accuracy: 0.7905\n",
            "Epoch 199/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6322 - accuracy: 0.7399 - val_loss: 0.5516 - val_accuracy: 0.7838\n",
            "Epoch 200/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5799 - accuracy: 0.7449 - val_loss: 0.5381 - val_accuracy: 0.7770\n",
            "Epoch 201/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6593 - accuracy: 0.7399 - val_loss: 0.5250 - val_accuracy: 0.8176\n",
            "Epoch 202/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6175 - accuracy: 0.7601 - val_loss: 0.5246 - val_accuracy: 0.8378\n",
            "Epoch 203/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6022 - accuracy: 0.7787 - val_loss: 0.5111 - val_accuracy: 0.8108\n",
            "Epoch 204/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6045 - accuracy: 0.7517 - val_loss: 0.5135 - val_accuracy: 0.8243\n",
            "Epoch 205/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5959 - accuracy: 0.7669 - val_loss: 0.5212 - val_accuracy: 0.8176\n",
            "Epoch 206/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5855 - accuracy: 0.7973 - val_loss: 0.5456 - val_accuracy: 0.7703\n",
            "Epoch 207/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6059 - accuracy: 0.7584 - val_loss: 0.4992 - val_accuracy: 0.8311\n",
            "Epoch 208/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6363 - accuracy: 0.7753 - val_loss: 0.5378 - val_accuracy: 0.8041\n",
            "Epoch 209/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6240 - accuracy: 0.7568 - val_loss: 0.5270 - val_accuracy: 0.7905\n",
            "Epoch 210/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6241 - accuracy: 0.7449 - val_loss: 0.5152 - val_accuracy: 0.8176\n",
            "Epoch 211/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6118 - accuracy: 0.7736 - val_loss: 0.5537 - val_accuracy: 0.8041\n",
            "Epoch 212/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6031 - accuracy: 0.7551 - val_loss: 0.5579 - val_accuracy: 0.7973\n",
            "Epoch 213/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6597 - accuracy: 0.7534 - val_loss: 0.5680 - val_accuracy: 0.7838\n",
            "Epoch 214/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6301 - accuracy: 0.7551 - val_loss: 0.5521 - val_accuracy: 0.8108\n",
            "Epoch 215/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6050 - accuracy: 0.7720 - val_loss: 0.5445 - val_accuracy: 0.7973\n",
            "Epoch 216/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6447 - accuracy: 0.7416 - val_loss: 0.5186 - val_accuracy: 0.8108\n",
            "Epoch 217/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5807 - accuracy: 0.7568 - val_loss: 0.5019 - val_accuracy: 0.8108\n",
            "Epoch 218/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6082 - accuracy: 0.7483 - val_loss: 0.5109 - val_accuracy: 0.8108\n",
            "Epoch 219/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6338 - accuracy: 0.7365 - val_loss: 0.5191 - val_accuracy: 0.8176\n",
            "Epoch 220/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6085 - accuracy: 0.7635 - val_loss: 0.5181 - val_accuracy: 0.7973\n",
            "Epoch 221/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6002 - accuracy: 0.7500 - val_loss: 0.5231 - val_accuracy: 0.7973\n",
            "Epoch 222/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6170 - accuracy: 0.7432 - val_loss: 0.5338 - val_accuracy: 0.8041\n",
            "Epoch 223/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6513 - accuracy: 0.7331 - val_loss: 0.5673 - val_accuracy: 0.7770\n",
            "Epoch 224/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6606 - accuracy: 0.7551 - val_loss: 0.5448 - val_accuracy: 0.7905\n",
            "Epoch 225/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6342 - accuracy: 0.7720 - val_loss: 0.5422 - val_accuracy: 0.7703\n",
            "Epoch 226/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6001 - accuracy: 0.7551 - val_loss: 0.5166 - val_accuracy: 0.7973\n",
            "Epoch 227/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6128 - accuracy: 0.7297 - val_loss: 0.5321 - val_accuracy: 0.7838\n",
            "Epoch 228/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6249 - accuracy: 0.7466 - val_loss: 0.5287 - val_accuracy: 0.7973\n",
            "Epoch 229/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6190 - accuracy: 0.7584 - val_loss: 0.5223 - val_accuracy: 0.8041\n",
            "Epoch 230/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6131 - accuracy: 0.7753 - val_loss: 0.5256 - val_accuracy: 0.7973\n",
            "Epoch 231/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6079 - accuracy: 0.7635 - val_loss: 0.5136 - val_accuracy: 0.7838\n",
            "Epoch 232/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5981 - accuracy: 0.7466 - val_loss: 0.5091 - val_accuracy: 0.7905\n",
            "Epoch 233/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5828 - accuracy: 0.7517 - val_loss: 0.5200 - val_accuracy: 0.7838\n",
            "Epoch 234/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5844 - accuracy: 0.7821 - val_loss: 0.5027 - val_accuracy: 0.8041\n",
            "Epoch 235/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5861 - accuracy: 0.7770 - val_loss: 0.5181 - val_accuracy: 0.7635\n",
            "Epoch 236/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6067 - accuracy: 0.7686 - val_loss: 0.5194 - val_accuracy: 0.7838\n",
            "Epoch 237/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5883 - accuracy: 0.7703 - val_loss: 0.5124 - val_accuracy: 0.8041\n",
            "Epoch 238/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5967 - accuracy: 0.7568 - val_loss: 0.5040 - val_accuracy: 0.8378\n",
            "Epoch 239/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5838 - accuracy: 0.7618 - val_loss: 0.5101 - val_accuracy: 0.8041\n",
            "Epoch 240/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5798 - accuracy: 0.7720 - val_loss: 0.5135 - val_accuracy: 0.8041\n",
            "Epoch 241/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6260 - accuracy: 0.7483 - val_loss: 0.5087 - val_accuracy: 0.8108\n",
            "Epoch 242/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5941 - accuracy: 0.7652 - val_loss: 0.5086 - val_accuracy: 0.8108\n",
            "Epoch 243/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5776 - accuracy: 0.7703 - val_loss: 0.5314 - val_accuracy: 0.7770\n",
            "Epoch 244/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5517 - accuracy: 0.7821 - val_loss: 0.5007 - val_accuracy: 0.7770\n",
            "Epoch 245/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5829 - accuracy: 0.7635 - val_loss: 0.4976 - val_accuracy: 0.8243\n",
            "Epoch 246/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5846 - accuracy: 0.7534 - val_loss: 0.4943 - val_accuracy: 0.8176\n",
            "Epoch 247/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6096 - accuracy: 0.7601 - val_loss: 0.5268 - val_accuracy: 0.7770\n",
            "Epoch 248/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6024 - accuracy: 0.7449 - val_loss: 0.4954 - val_accuracy: 0.8041\n",
            "Epoch 249/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5767 - accuracy: 0.7703 - val_loss: 0.5225 - val_accuracy: 0.7973\n",
            "Epoch 250/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6473 - accuracy: 0.7264 - val_loss: 0.5140 - val_accuracy: 0.8176\n",
            "Epoch 251/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6073 - accuracy: 0.7500 - val_loss: 0.4960 - val_accuracy: 0.8176\n",
            "Epoch 252/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6053 - accuracy: 0.7534 - val_loss: 0.5222 - val_accuracy: 0.8041\n",
            "Epoch 253/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5816 - accuracy: 0.7669 - val_loss: 0.4960 - val_accuracy: 0.8041\n",
            "Epoch 254/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6323 - accuracy: 0.7483 - val_loss: 0.5192 - val_accuracy: 0.8311\n",
            "Epoch 255/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5832 - accuracy: 0.7601 - val_loss: 0.5242 - val_accuracy: 0.7703\n",
            "Epoch 256/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6261 - accuracy: 0.7365 - val_loss: 0.5213 - val_accuracy: 0.7905\n",
            "Epoch 257/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6085 - accuracy: 0.7736 - val_loss: 0.5217 - val_accuracy: 0.8041\n",
            "Epoch 258/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6033 - accuracy: 0.7601 - val_loss: 0.4988 - val_accuracy: 0.8041\n",
            "Epoch 259/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5683 - accuracy: 0.7770 - val_loss: 0.4982 - val_accuracy: 0.8176\n",
            "Epoch 260/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5913 - accuracy: 0.7568 - val_loss: 0.4985 - val_accuracy: 0.8243\n",
            "Epoch 261/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5542 - accuracy: 0.7889 - val_loss: 0.5025 - val_accuracy: 0.7905\n",
            "Epoch 262/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6153 - accuracy: 0.7635 - val_loss: 0.5342 - val_accuracy: 0.8041\n",
            "Epoch 263/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5772 - accuracy: 0.7753 - val_loss: 0.5152 - val_accuracy: 0.8176\n",
            "Epoch 264/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5955 - accuracy: 0.7618 - val_loss: 0.5076 - val_accuracy: 0.8176\n",
            "Epoch 265/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6156 - accuracy: 0.7551 - val_loss: 0.5192 - val_accuracy: 0.8176\n",
            "Epoch 266/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5780 - accuracy: 0.7500 - val_loss: 0.4994 - val_accuracy: 0.8378\n",
            "Epoch 267/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6081 - accuracy: 0.7669 - val_loss: 0.5223 - val_accuracy: 0.8041\n",
            "Epoch 268/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5713 - accuracy: 0.7703 - val_loss: 0.5389 - val_accuracy: 0.7905\n",
            "Epoch 269/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5755 - accuracy: 0.7804 - val_loss: 0.4960 - val_accuracy: 0.8311\n",
            "Epoch 270/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5980 - accuracy: 0.7635 - val_loss: 0.4895 - val_accuracy: 0.8243\n",
            "Epoch 271/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6036 - accuracy: 0.7686 - val_loss: 0.5049 - val_accuracy: 0.8176\n",
            "Epoch 272/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5939 - accuracy: 0.7618 - val_loss: 0.5118 - val_accuracy: 0.8176\n",
            "Epoch 273/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5566 - accuracy: 0.7601 - val_loss: 0.4982 - val_accuracy: 0.8176\n",
            "Epoch 274/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6266 - accuracy: 0.7500 - val_loss: 0.5190 - val_accuracy: 0.7973\n",
            "Epoch 275/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5976 - accuracy: 0.7618 - val_loss: 0.4887 - val_accuracy: 0.8446\n",
            "Epoch 276/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5773 - accuracy: 0.7855 - val_loss: 0.4778 - val_accuracy: 0.8378\n",
            "Epoch 277/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6122 - accuracy: 0.7517 - val_loss: 0.5075 - val_accuracy: 0.8243\n",
            "Epoch 278/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6316 - accuracy: 0.7601 - val_loss: 0.5049 - val_accuracy: 0.8176\n",
            "Epoch 279/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5601 - accuracy: 0.7821 - val_loss: 0.5192 - val_accuracy: 0.8041\n",
            "Epoch 280/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5768 - accuracy: 0.7601 - val_loss: 0.5025 - val_accuracy: 0.8243\n",
            "Epoch 281/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5915 - accuracy: 0.7669 - val_loss: 0.4781 - val_accuracy: 0.8243\n",
            "Epoch 282/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6088 - accuracy: 0.7584 - val_loss: 0.4841 - val_accuracy: 0.8176\n",
            "Epoch 283/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6321 - accuracy: 0.7331 - val_loss: 0.5172 - val_accuracy: 0.8176\n",
            "Epoch 284/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5808 - accuracy: 0.7635 - val_loss: 0.4928 - val_accuracy: 0.7973\n",
            "Epoch 285/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5770 - accuracy: 0.7584 - val_loss: 0.4863 - val_accuracy: 0.8243\n",
            "Epoch 286/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5728 - accuracy: 0.7821 - val_loss: 0.4986 - val_accuracy: 0.8378\n",
            "Epoch 287/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5947 - accuracy: 0.7703 - val_loss: 0.4834 - val_accuracy: 0.8514\n",
            "Epoch 288/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5969 - accuracy: 0.7686 - val_loss: 0.5107 - val_accuracy: 0.7905\n",
            "Epoch 289/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6103 - accuracy: 0.7517 - val_loss: 0.5321 - val_accuracy: 0.7838\n",
            "Epoch 290/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6185 - accuracy: 0.7584 - val_loss: 0.5359 - val_accuracy: 0.7703\n",
            "Epoch 291/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5954 - accuracy: 0.7635 - val_loss: 0.5083 - val_accuracy: 0.7973\n",
            "Epoch 292/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5626 - accuracy: 0.7736 - val_loss: 0.4892 - val_accuracy: 0.7838\n",
            "Epoch 293/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5904 - accuracy: 0.7821 - val_loss: 0.4919 - val_accuracy: 0.8176\n",
            "Epoch 294/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5985 - accuracy: 0.7534 - val_loss: 0.4964 - val_accuracy: 0.8041\n",
            "Epoch 295/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6359 - accuracy: 0.7517 - val_loss: 0.5026 - val_accuracy: 0.8243\n",
            "Epoch 296/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5739 - accuracy: 0.7855 - val_loss: 0.4939 - val_accuracy: 0.8176\n",
            "Epoch 297/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5939 - accuracy: 0.7449 - val_loss: 0.5115 - val_accuracy: 0.8041\n",
            "Epoch 298/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6304 - accuracy: 0.7500 - val_loss: 0.5400 - val_accuracy: 0.8041\n",
            "Epoch 299/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6191 - accuracy: 0.7669 - val_loss: 0.5152 - val_accuracy: 0.8108\n",
            "Epoch 300/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5811 - accuracy: 0.7855 - val_loss: 0.4995 - val_accuracy: 0.8243\n",
            "Epoch 301/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6054 - accuracy: 0.7584 - val_loss: 0.5009 - val_accuracy: 0.8378\n",
            "Epoch 302/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5626 - accuracy: 0.8074 - val_loss: 0.5089 - val_accuracy: 0.8041\n",
            "Epoch 303/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5775 - accuracy: 0.7686 - val_loss: 0.5130 - val_accuracy: 0.8108\n",
            "Epoch 304/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5985 - accuracy: 0.7720 - val_loss: 0.5155 - val_accuracy: 0.7973\n",
            "Epoch 305/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5867 - accuracy: 0.7500 - val_loss: 0.4885 - val_accuracy: 0.7973\n",
            "Epoch 306/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5639 - accuracy: 0.7770 - val_loss: 0.4942 - val_accuracy: 0.8378\n",
            "Epoch 307/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5762 - accuracy: 0.7551 - val_loss: 0.4950 - val_accuracy: 0.8176\n",
            "Epoch 308/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5984 - accuracy: 0.7534 - val_loss: 0.4949 - val_accuracy: 0.8243\n",
            "Epoch 309/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5867 - accuracy: 0.7753 - val_loss: 0.4870 - val_accuracy: 0.8311\n",
            "Epoch 310/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5926 - accuracy: 0.7601 - val_loss: 0.5018 - val_accuracy: 0.8311\n",
            "Epoch 311/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6319 - accuracy: 0.7196 - val_loss: 0.5171 - val_accuracy: 0.8176\n",
            "Epoch 312/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5800 - accuracy: 0.7821 - val_loss: 0.5279 - val_accuracy: 0.7973\n",
            "Epoch 313/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5485 - accuracy: 0.7703 - val_loss: 0.5125 - val_accuracy: 0.7973\n",
            "Epoch 314/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5811 - accuracy: 0.7855 - val_loss: 0.5315 - val_accuracy: 0.7703\n",
            "Epoch 315/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6228 - accuracy: 0.7432 - val_loss: 0.5060 - val_accuracy: 0.7838\n",
            "Epoch 316/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5485 - accuracy: 0.7753 - val_loss: 0.4985 - val_accuracy: 0.7838\n",
            "Epoch 317/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5682 - accuracy: 0.7601 - val_loss: 0.4909 - val_accuracy: 0.7905\n",
            "Epoch 318/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5706 - accuracy: 0.7669 - val_loss: 0.5178 - val_accuracy: 0.7973\n",
            "Epoch 319/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5645 - accuracy: 0.7855 - val_loss: 0.5276 - val_accuracy: 0.7568\n",
            "Epoch 320/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6267 - accuracy: 0.7416 - val_loss: 0.5062 - val_accuracy: 0.7973\n",
            "Epoch 321/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5911 - accuracy: 0.7551 - val_loss: 0.4856 - val_accuracy: 0.8176\n",
            "Epoch 322/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6189 - accuracy: 0.7568 - val_loss: 0.5024 - val_accuracy: 0.8176\n",
            "Epoch 323/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5817 - accuracy: 0.7703 - val_loss: 0.4915 - val_accuracy: 0.8243\n",
            "Epoch 324/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5617 - accuracy: 0.7584 - val_loss: 0.4821 - val_accuracy: 0.8243\n",
            "Epoch 325/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6012 - accuracy: 0.7551 - val_loss: 0.4928 - val_accuracy: 0.8176\n",
            "Epoch 326/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5829 - accuracy: 0.7551 - val_loss: 0.4915 - val_accuracy: 0.8311\n",
            "Epoch 327/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6077 - accuracy: 0.7618 - val_loss: 0.5143 - val_accuracy: 0.8243\n",
            "Epoch 328/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5861 - accuracy: 0.7652 - val_loss: 0.4988 - val_accuracy: 0.8041\n",
            "Epoch 329/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5805 - accuracy: 0.7703 - val_loss: 0.5371 - val_accuracy: 0.7635\n",
            "Epoch 330/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5758 - accuracy: 0.7753 - val_loss: 0.5039 - val_accuracy: 0.7905\n",
            "Epoch 331/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5827 - accuracy: 0.7889 - val_loss: 0.4800 - val_accuracy: 0.8176\n",
            "Epoch 332/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5781 - accuracy: 0.7686 - val_loss: 0.5143 - val_accuracy: 0.7973\n",
            "Epoch 333/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5896 - accuracy: 0.7804 - val_loss: 0.4942 - val_accuracy: 0.7838\n",
            "Epoch 334/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5793 - accuracy: 0.7787 - val_loss: 0.4911 - val_accuracy: 0.7973\n",
            "Epoch 335/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5927 - accuracy: 0.7584 - val_loss: 0.4918 - val_accuracy: 0.8176\n",
            "Epoch 336/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5817 - accuracy: 0.7889 - val_loss: 0.4933 - val_accuracy: 0.8108\n",
            "Epoch 337/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6059 - accuracy: 0.7483 - val_loss: 0.5061 - val_accuracy: 0.7770\n",
            "Epoch 338/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5461 - accuracy: 0.7720 - val_loss: 0.5144 - val_accuracy: 0.7838\n",
            "Epoch 339/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5632 - accuracy: 0.7804 - val_loss: 0.4846 - val_accuracy: 0.7973\n",
            "Epoch 340/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5864 - accuracy: 0.7669 - val_loss: 0.4766 - val_accuracy: 0.8041\n",
            "Epoch 341/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6002 - accuracy: 0.7517 - val_loss: 0.5145 - val_accuracy: 0.8041\n",
            "Epoch 342/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5435 - accuracy: 0.7905 - val_loss: 0.4830 - val_accuracy: 0.8176\n",
            "Epoch 343/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5629 - accuracy: 0.7736 - val_loss: 0.4837 - val_accuracy: 0.8243\n",
            "Epoch 344/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5982 - accuracy: 0.7618 - val_loss: 0.5031 - val_accuracy: 0.8243\n",
            "Epoch 345/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5496 - accuracy: 0.7787 - val_loss: 0.4805 - val_accuracy: 0.8041\n",
            "Epoch 346/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5735 - accuracy: 0.7753 - val_loss: 0.5058 - val_accuracy: 0.7905\n",
            "Epoch 347/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5884 - accuracy: 0.7922 - val_loss: 0.4967 - val_accuracy: 0.8311\n",
            "Epoch 348/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6171 - accuracy: 0.7534 - val_loss: 0.4711 - val_accuracy: 0.8243\n",
            "Epoch 349/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6174 - accuracy: 0.7500 - val_loss: 0.4843 - val_accuracy: 0.8108\n",
            "Epoch 350/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5973 - accuracy: 0.7669 - val_loss: 0.4981 - val_accuracy: 0.7973\n",
            "Epoch 351/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5520 - accuracy: 0.7855 - val_loss: 0.4752 - val_accuracy: 0.8176\n",
            "Epoch 352/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5993 - accuracy: 0.7517 - val_loss: 0.4888 - val_accuracy: 0.7973\n",
            "Epoch 353/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5262 - accuracy: 0.8125 - val_loss: 0.5014 - val_accuracy: 0.7838\n",
            "Epoch 354/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5914 - accuracy: 0.7584 - val_loss: 0.5039 - val_accuracy: 0.7973\n",
            "Epoch 355/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6109 - accuracy: 0.7601 - val_loss: 0.5238 - val_accuracy: 0.7500\n",
            "Epoch 356/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5517 - accuracy: 0.7973 - val_loss: 0.4936 - val_accuracy: 0.8041\n",
            "Epoch 357/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6026 - accuracy: 0.7703 - val_loss: 0.4844 - val_accuracy: 0.8108\n",
            "Epoch 358/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5766 - accuracy: 0.7669 - val_loss: 0.4831 - val_accuracy: 0.8041\n",
            "Epoch 359/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5556 - accuracy: 0.7821 - val_loss: 0.5032 - val_accuracy: 0.7905\n",
            "Epoch 360/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6075 - accuracy: 0.7635 - val_loss: 0.4745 - val_accuracy: 0.8176\n",
            "Epoch 361/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5751 - accuracy: 0.7703 - val_loss: 0.4882 - val_accuracy: 0.8176\n",
            "Epoch 362/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5796 - accuracy: 0.7466 - val_loss: 0.4767 - val_accuracy: 0.8108\n",
            "Epoch 363/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5366 - accuracy: 0.8074 - val_loss: 0.4727 - val_accuracy: 0.7973\n",
            "Epoch 364/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5330 - accuracy: 0.7990 - val_loss: 0.4718 - val_accuracy: 0.8378\n",
            "Epoch 365/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5888 - accuracy: 0.7787 - val_loss: 0.5473 - val_accuracy: 0.7703\n",
            "Epoch 366/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5849 - accuracy: 0.7618 - val_loss: 0.5063 - val_accuracy: 0.8108\n",
            "Epoch 367/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5728 - accuracy: 0.7804 - val_loss: 0.4693 - val_accuracy: 0.8378\n",
            "Epoch 368/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5617 - accuracy: 0.7838 - val_loss: 0.4791 - val_accuracy: 0.8243\n",
            "Epoch 369/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5826 - accuracy: 0.7601 - val_loss: 0.4904 - val_accuracy: 0.8108\n",
            "Epoch 370/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5299 - accuracy: 0.7753 - val_loss: 0.4810 - val_accuracy: 0.8176\n",
            "Epoch 371/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5636 - accuracy: 0.7720 - val_loss: 0.5075 - val_accuracy: 0.8176\n",
            "Epoch 372/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6058 - accuracy: 0.7686 - val_loss: 0.5219 - val_accuracy: 0.7973\n",
            "Epoch 373/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5709 - accuracy: 0.7956 - val_loss: 0.4936 - val_accuracy: 0.8041\n",
            "Epoch 374/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5586 - accuracy: 0.7889 - val_loss: 0.5014 - val_accuracy: 0.8041\n",
            "Epoch 375/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5794 - accuracy: 0.7720 - val_loss: 0.4973 - val_accuracy: 0.7973\n",
            "Epoch 376/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5417 - accuracy: 0.7703 - val_loss: 0.4858 - val_accuracy: 0.8041\n",
            "Epoch 377/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5649 - accuracy: 0.7821 - val_loss: 0.4622 - val_accuracy: 0.8243\n",
            "Epoch 378/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5791 - accuracy: 0.7753 - val_loss: 0.4953 - val_accuracy: 0.8243\n",
            "Epoch 379/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5357 - accuracy: 0.7939 - val_loss: 0.4884 - val_accuracy: 0.7905\n",
            "Epoch 380/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6048 - accuracy: 0.7601 - val_loss: 0.5039 - val_accuracy: 0.8041\n",
            "Epoch 381/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5830 - accuracy: 0.7838 - val_loss: 0.4898 - val_accuracy: 0.8176\n",
            "Epoch 382/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5460 - accuracy: 0.7736 - val_loss: 0.4874 - val_accuracy: 0.8108\n",
            "Epoch 383/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5855 - accuracy: 0.7483 - val_loss: 0.5222 - val_accuracy: 0.7838\n",
            "Epoch 384/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5862 - accuracy: 0.7635 - val_loss: 0.4969 - val_accuracy: 0.7770\n",
            "Epoch 385/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5508 - accuracy: 0.7618 - val_loss: 0.4837 - val_accuracy: 0.7905\n",
            "Epoch 386/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5774 - accuracy: 0.7568 - val_loss: 0.5032 - val_accuracy: 0.7838\n",
            "Epoch 387/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5809 - accuracy: 0.7601 - val_loss: 0.5005 - val_accuracy: 0.8311\n",
            "Epoch 388/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5518 - accuracy: 0.7990 - val_loss: 0.4861 - val_accuracy: 0.8176\n",
            "Epoch 389/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5734 - accuracy: 0.7753 - val_loss: 0.4929 - val_accuracy: 0.8176\n",
            "Epoch 390/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5788 - accuracy: 0.7787 - val_loss: 0.4978 - val_accuracy: 0.8041\n",
            "Epoch 391/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5136 - accuracy: 0.7905 - val_loss: 0.4686 - val_accuracy: 0.8311\n",
            "Epoch 392/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6217 - accuracy: 0.7584 - val_loss: 0.5042 - val_accuracy: 0.7973\n",
            "Epoch 393/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5694 - accuracy: 0.7736 - val_loss: 0.4875 - val_accuracy: 0.8041\n",
            "Epoch 394/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5470 - accuracy: 0.7838 - val_loss: 0.4783 - val_accuracy: 0.8041\n",
            "Epoch 395/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5914 - accuracy: 0.7703 - val_loss: 0.4759 - val_accuracy: 0.8378\n",
            "Epoch 396/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5922 - accuracy: 0.7736 - val_loss: 0.4960 - val_accuracy: 0.8378\n",
            "Epoch 397/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5913 - accuracy: 0.7736 - val_loss: 0.5036 - val_accuracy: 0.8176\n",
            "Epoch 398/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5512 - accuracy: 0.7889 - val_loss: 0.4972 - val_accuracy: 0.8243\n",
            "Epoch 399/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5831 - accuracy: 0.7568 - val_loss: 0.4810 - val_accuracy: 0.8176\n",
            "Epoch 400/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5404 - accuracy: 0.7905 - val_loss: 0.4774 - val_accuracy: 0.8108\n",
            "Epoch 401/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5688 - accuracy: 0.7787 - val_loss: 0.5132 - val_accuracy: 0.8041\n",
            "Epoch 402/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5726 - accuracy: 0.7855 - val_loss: 0.4874 - val_accuracy: 0.8311\n",
            "Epoch 403/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6168 - accuracy: 0.7517 - val_loss: 0.5085 - val_accuracy: 0.8243\n",
            "Epoch 404/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5659 - accuracy: 0.7804 - val_loss: 0.4962 - val_accuracy: 0.7905\n",
            "Epoch 405/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5547 - accuracy: 0.7855 - val_loss: 0.4904 - val_accuracy: 0.7973\n",
            "Epoch 406/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5441 - accuracy: 0.7838 - val_loss: 0.4833 - val_accuracy: 0.7973\n",
            "Epoch 407/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5958 - accuracy: 0.7534 - val_loss: 0.4918 - val_accuracy: 0.8108\n",
            "Epoch 408/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5709 - accuracy: 0.7601 - val_loss: 0.4885 - val_accuracy: 0.8041\n",
            "Epoch 409/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5493 - accuracy: 0.7939 - val_loss: 0.4816 - val_accuracy: 0.7973\n",
            "Epoch 410/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5800 - accuracy: 0.7736 - val_loss: 0.4885 - val_accuracy: 0.8243\n",
            "Epoch 411/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5798 - accuracy: 0.7568 - val_loss: 0.4787 - val_accuracy: 0.8176\n",
            "Epoch 412/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5871 - accuracy: 0.7686 - val_loss: 0.4914 - val_accuracy: 0.8243\n",
            "Epoch 413/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5940 - accuracy: 0.7618 - val_loss: 0.4874 - val_accuracy: 0.8243\n",
            "Epoch 414/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5850 - accuracy: 0.7686 - val_loss: 0.4736 - val_accuracy: 0.8041\n",
            "Epoch 415/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5770 - accuracy: 0.7669 - val_loss: 0.4982 - val_accuracy: 0.8108\n",
            "Epoch 416/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5700 - accuracy: 0.7753 - val_loss: 0.4946 - val_accuracy: 0.8108\n",
            "Epoch 417/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5779 - accuracy: 0.7720 - val_loss: 0.4919 - val_accuracy: 0.8108\n",
            "Epoch 418/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5824 - accuracy: 0.7432 - val_loss: 0.5067 - val_accuracy: 0.7905\n",
            "Epoch 419/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5832 - accuracy: 0.7652 - val_loss: 0.4916 - val_accuracy: 0.7973\n",
            "Epoch 420/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5888 - accuracy: 0.7517 - val_loss: 0.4959 - val_accuracy: 0.7770\n",
            "Epoch 421/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5747 - accuracy: 0.7669 - val_loss: 0.4820 - val_accuracy: 0.8041\n",
            "Epoch 422/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5558 - accuracy: 0.7652 - val_loss: 0.4976 - val_accuracy: 0.7905\n",
            "Epoch 423/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5433 - accuracy: 0.7703 - val_loss: 0.4863 - val_accuracy: 0.8108\n",
            "Epoch 424/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5290 - accuracy: 0.7872 - val_loss: 0.4553 - val_accuracy: 0.8176\n",
            "Epoch 425/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5927 - accuracy: 0.7584 - val_loss: 0.4913 - val_accuracy: 0.8446\n",
            "Epoch 426/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6249 - accuracy: 0.7365 - val_loss: 0.4911 - val_accuracy: 0.8311\n",
            "Epoch 427/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5624 - accuracy: 0.7686 - val_loss: 0.4702 - val_accuracy: 0.8378\n",
            "Epoch 428/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5876 - accuracy: 0.7889 - val_loss: 0.4776 - val_accuracy: 0.8311\n",
            "Epoch 429/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5475 - accuracy: 0.7618 - val_loss: 0.4986 - val_accuracy: 0.8041\n",
            "Epoch 430/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6018 - accuracy: 0.7466 - val_loss: 0.4955 - val_accuracy: 0.7973\n",
            "Epoch 431/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5553 - accuracy: 0.7855 - val_loss: 0.5023 - val_accuracy: 0.7973\n",
            "Epoch 432/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5547 - accuracy: 0.7770 - val_loss: 0.5159 - val_accuracy: 0.8041\n",
            "Epoch 433/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5820 - accuracy: 0.7872 - val_loss: 0.4976 - val_accuracy: 0.8311\n",
            "Epoch 434/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5883 - accuracy: 0.7686 - val_loss: 0.5024 - val_accuracy: 0.8108\n",
            "Epoch 435/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5885 - accuracy: 0.7905 - val_loss: 0.5065 - val_accuracy: 0.8176\n",
            "Epoch 436/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5889 - accuracy: 0.7686 - val_loss: 0.4892 - val_accuracy: 0.8378\n",
            "Epoch 437/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5133 - accuracy: 0.8142 - val_loss: 0.5117 - val_accuracy: 0.7838\n",
            "Epoch 438/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6193 - accuracy: 0.7720 - val_loss: 0.5471 - val_accuracy: 0.7770\n",
            "Epoch 439/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5894 - accuracy: 0.7889 - val_loss: 0.4987 - val_accuracy: 0.8243\n",
            "Epoch 440/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6216 - accuracy: 0.7466 - val_loss: 0.5155 - val_accuracy: 0.8108\n",
            "Epoch 441/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5954 - accuracy: 0.7601 - val_loss: 0.5198 - val_accuracy: 0.8041\n",
            "Epoch 442/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5676 - accuracy: 0.7720 - val_loss: 0.5002 - val_accuracy: 0.8041\n",
            "Epoch 443/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5615 - accuracy: 0.7872 - val_loss: 0.4798 - val_accuracy: 0.8243\n",
            "Epoch 444/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5802 - accuracy: 0.7703 - val_loss: 0.4802 - val_accuracy: 0.8176\n",
            "Epoch 445/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5464 - accuracy: 0.7872 - val_loss: 0.4935 - val_accuracy: 0.8041\n",
            "Epoch 446/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5373 - accuracy: 0.8074 - val_loss: 0.5001 - val_accuracy: 0.7973\n",
            "Epoch 447/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5792 - accuracy: 0.7652 - val_loss: 0.4955 - val_accuracy: 0.7905\n",
            "Epoch 448/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5663 - accuracy: 0.7872 - val_loss: 0.4905 - val_accuracy: 0.7905\n",
            "Epoch 449/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5449 - accuracy: 0.7889 - val_loss: 0.4768 - val_accuracy: 0.8243\n",
            "Epoch 450/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5704 - accuracy: 0.7905 - val_loss: 0.4783 - val_accuracy: 0.8243\n",
            "Epoch 451/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5479 - accuracy: 0.7770 - val_loss: 0.4731 - val_accuracy: 0.8176\n",
            "Epoch 452/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5581 - accuracy: 0.7939 - val_loss: 0.4863 - val_accuracy: 0.8176\n",
            "Epoch 453/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5834 - accuracy: 0.7500 - val_loss: 0.4925 - val_accuracy: 0.8108\n",
            "Epoch 454/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5783 - accuracy: 0.7635 - val_loss: 0.4933 - val_accuracy: 0.8041\n",
            "Epoch 455/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5320 - accuracy: 0.7703 - val_loss: 0.4741 - val_accuracy: 0.8176\n",
            "Epoch 456/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5435 - accuracy: 0.7635 - val_loss: 0.4901 - val_accuracy: 0.8041\n",
            "Epoch 457/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5708 - accuracy: 0.7635 - val_loss: 0.5127 - val_accuracy: 0.8041\n",
            "Epoch 458/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5617 - accuracy: 0.7787 - val_loss: 0.4957 - val_accuracy: 0.7973\n",
            "Epoch 459/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5786 - accuracy: 0.7720 - val_loss: 0.5104 - val_accuracy: 0.7838\n",
            "Epoch 460/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5985 - accuracy: 0.7534 - val_loss: 0.4863 - val_accuracy: 0.8108\n",
            "Epoch 461/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5397 - accuracy: 0.8024 - val_loss: 0.4886 - val_accuracy: 0.8176\n",
            "Epoch 462/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5650 - accuracy: 0.7838 - val_loss: 0.4991 - val_accuracy: 0.7973\n",
            "Epoch 463/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5918 - accuracy: 0.7449 - val_loss: 0.4880 - val_accuracy: 0.8514\n",
            "Epoch 464/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5934 - accuracy: 0.7432 - val_loss: 0.5032 - val_accuracy: 0.8311\n",
            "Epoch 465/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6115 - accuracy: 0.7297 - val_loss: 0.5010 - val_accuracy: 0.8243\n",
            "Epoch 466/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5423 - accuracy: 0.8024 - val_loss: 0.4842 - val_accuracy: 0.8243\n",
            "Epoch 467/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5670 - accuracy: 0.7618 - val_loss: 0.4846 - val_accuracy: 0.8378\n",
            "Epoch 468/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5391 - accuracy: 0.7821 - val_loss: 0.4881 - val_accuracy: 0.8311\n",
            "Epoch 469/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5773 - accuracy: 0.7838 - val_loss: 0.5142 - val_accuracy: 0.8041\n",
            "Epoch 470/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5821 - accuracy: 0.7703 - val_loss: 0.4929 - val_accuracy: 0.8176\n",
            "Epoch 471/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5771 - accuracy: 0.7736 - val_loss: 0.4856 - val_accuracy: 0.8176\n",
            "Epoch 472/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5780 - accuracy: 0.7821 - val_loss: 0.4966 - val_accuracy: 0.8311\n",
            "Epoch 473/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5554 - accuracy: 0.7855 - val_loss: 0.5020 - val_accuracy: 0.8108\n",
            "Epoch 474/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5730 - accuracy: 0.7669 - val_loss: 0.5116 - val_accuracy: 0.8176\n",
            "Epoch 475/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5539 - accuracy: 0.7872 - val_loss: 0.4994 - val_accuracy: 0.8041\n",
            "Epoch 476/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5812 - accuracy: 0.7720 - val_loss: 0.4821 - val_accuracy: 0.8243\n",
            "Epoch 477/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5858 - accuracy: 0.7889 - val_loss: 0.5092 - val_accuracy: 0.7973\n",
            "Epoch 478/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6008 - accuracy: 0.7753 - val_loss: 0.5047 - val_accuracy: 0.8041\n",
            "Epoch 479/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6007 - accuracy: 0.7686 - val_loss: 0.5019 - val_accuracy: 0.8041\n",
            "Epoch 480/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5667 - accuracy: 0.7720 - val_loss: 0.4855 - val_accuracy: 0.8041\n",
            "Epoch 481/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5362 - accuracy: 0.7838 - val_loss: 0.4819 - val_accuracy: 0.8176\n",
            "Epoch 482/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5594 - accuracy: 0.7770 - val_loss: 0.5261 - val_accuracy: 0.7973\n",
            "Epoch 483/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6002 - accuracy: 0.7804 - val_loss: 0.5010 - val_accuracy: 0.8581\n",
            "Epoch 484/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5747 - accuracy: 0.7753 - val_loss: 0.4957 - val_accuracy: 0.7905\n",
            "Epoch 485/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5444 - accuracy: 0.8057 - val_loss: 0.4880 - val_accuracy: 0.8108\n",
            "Epoch 486/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5567 - accuracy: 0.7872 - val_loss: 0.4879 - val_accuracy: 0.8108\n",
            "Epoch 487/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5744 - accuracy: 0.7787 - val_loss: 0.4937 - val_accuracy: 0.7973\n",
            "Epoch 488/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6055 - accuracy: 0.7399 - val_loss: 0.4996 - val_accuracy: 0.7905\n",
            "Epoch 489/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5955 - accuracy: 0.7568 - val_loss: 0.4983 - val_accuracy: 0.8108\n",
            "Epoch 490/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6066 - accuracy: 0.7551 - val_loss: 0.4932 - val_accuracy: 0.8108\n",
            "Epoch 491/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5897 - accuracy: 0.7618 - val_loss: 0.5076 - val_accuracy: 0.8108\n",
            "Epoch 492/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5764 - accuracy: 0.7517 - val_loss: 0.5213 - val_accuracy: 0.7838\n",
            "Epoch 493/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6030 - accuracy: 0.7804 - val_loss: 0.5178 - val_accuracy: 0.8041\n",
            "Epoch 494/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5522 - accuracy: 0.7787 - val_loss: 0.5019 - val_accuracy: 0.8041\n",
            "Epoch 495/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5670 - accuracy: 0.7703 - val_loss: 0.5073 - val_accuracy: 0.7905\n",
            "Epoch 496/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5503 - accuracy: 0.7889 - val_loss: 0.4916 - val_accuracy: 0.8041\n",
            "Epoch 497/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6162 - accuracy: 0.7534 - val_loss: 0.4971 - val_accuracy: 0.7973\n",
            "Epoch 498/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5460 - accuracy: 0.7973 - val_loss: 0.4909 - val_accuracy: 0.7770\n",
            "Epoch 499/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5987 - accuracy: 0.7500 - val_loss: 0.4967 - val_accuracy: 0.7973\n",
            "Epoch 500/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5665 - accuracy: 0.7838 - val_loss: 0.4810 - val_accuracy: 0.8378\n",
            "5/5 [==============================] - 0s 3ms/step - loss: 0.4810 - accuracy: 0.8378\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:62: DeprecationWarning: scipy.interp is deprecated and will be removed in SciPy 2.0.0, use numpy.interp instead\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/500\n",
            "74/74 [==============================] - 1s 14ms/step - loss: 2.0234 - accuracy: 0.5203 - val_loss: 1.4052 - val_accuracy: 0.5473\n",
            "Epoch 2/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 1.5460 - accuracy: 0.5591 - val_loss: 1.3579 - val_accuracy: 0.5878\n",
            "Epoch 3/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 1.3631 - accuracy: 0.5676 - val_loss: 1.2984 - val_accuracy: 0.5541\n",
            "Epoch 4/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 1.2962 - accuracy: 0.5338 - val_loss: 1.2307 - val_accuracy: 0.5946\n",
            "Epoch 5/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 1.1824 - accuracy: 0.5828 - val_loss: 1.1557 - val_accuracy: 0.5946\n",
            "Epoch 6/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 1.1240 - accuracy: 0.5659 - val_loss: 1.0865 - val_accuracy: 0.6689\n",
            "Epoch 7/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 1.0818 - accuracy: 0.6030 - val_loss: 1.0357 - val_accuracy: 0.6419\n",
            "Epoch 8/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 1.0265 - accuracy: 0.5439 - val_loss: 0.9981 - val_accuracy: 0.6216\n",
            "Epoch 9/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.9670 - accuracy: 0.6014 - val_loss: 0.9281 - val_accuracy: 0.7027\n",
            "Epoch 10/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.9437 - accuracy: 0.5946 - val_loss: 0.8932 - val_accuracy: 0.6689\n",
            "Epoch 11/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.9148 - accuracy: 0.5946 - val_loss: 0.8792 - val_accuracy: 0.7162\n",
            "Epoch 12/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.9155 - accuracy: 0.5929 - val_loss: 0.8572 - val_accuracy: 0.7095\n",
            "Epoch 13/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8927 - accuracy: 0.5963 - val_loss: 0.8386 - val_accuracy: 0.7500\n",
            "Epoch 14/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8634 - accuracy: 0.6014 - val_loss: 0.7977 - val_accuracy: 0.7365\n",
            "Epoch 15/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8476 - accuracy: 0.6250 - val_loss: 0.8119 - val_accuracy: 0.7365\n",
            "Epoch 16/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8326 - accuracy: 0.6622 - val_loss: 0.7827 - val_accuracy: 0.7297\n",
            "Epoch 17/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7966 - accuracy: 0.6723 - val_loss: 0.7767 - val_accuracy: 0.7365\n",
            "Epoch 18/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8550 - accuracy: 0.6368 - val_loss: 0.7816 - val_accuracy: 0.7568\n",
            "Epoch 19/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.8296 - accuracy: 0.6605 - val_loss: 0.7573 - val_accuracy: 0.7500\n",
            "Epoch 20/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7932 - accuracy: 0.6858 - val_loss: 0.7375 - val_accuracy: 0.7162\n",
            "Epoch 21/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.8112 - accuracy: 0.6486 - val_loss: 0.7636 - val_accuracy: 0.7365\n",
            "Epoch 22/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.8091 - accuracy: 0.6757 - val_loss: 0.7915 - val_accuracy: 0.6622\n",
            "Epoch 23/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7653 - accuracy: 0.7010 - val_loss: 0.7162 - val_accuracy: 0.7432\n",
            "Epoch 24/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7824 - accuracy: 0.6858 - val_loss: 0.7238 - val_accuracy: 0.7703\n",
            "Epoch 25/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8144 - accuracy: 0.6520 - val_loss: 0.7743 - val_accuracy: 0.7162\n",
            "Epoch 26/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.8376 - accuracy: 0.6655 - val_loss: 0.7650 - val_accuracy: 0.7568\n",
            "Epoch 27/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.8307 - accuracy: 0.6723 - val_loss: 0.7398 - val_accuracy: 0.7365\n",
            "Epoch 28/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7899 - accuracy: 0.7027 - val_loss: 0.7187 - val_accuracy: 0.7365\n",
            "Epoch 29/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7990 - accuracy: 0.6993 - val_loss: 0.7327 - val_accuracy: 0.7905\n",
            "Epoch 30/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7918 - accuracy: 0.6909 - val_loss: 0.7168 - val_accuracy: 0.7568\n",
            "Epoch 31/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.8037 - accuracy: 0.6807 - val_loss: 0.7320 - val_accuracy: 0.7568\n",
            "Epoch 32/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8017 - accuracy: 0.6841 - val_loss: 0.7379 - val_accuracy: 0.7568\n",
            "Epoch 33/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7923 - accuracy: 0.6605 - val_loss: 0.7149 - val_accuracy: 0.7770\n",
            "Epoch 34/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7790 - accuracy: 0.6943 - val_loss: 0.7047 - val_accuracy: 0.7838\n",
            "Epoch 35/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7760 - accuracy: 0.7128 - val_loss: 0.7234 - val_accuracy: 0.7500\n",
            "Epoch 36/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7928 - accuracy: 0.6943 - val_loss: 0.7386 - val_accuracy: 0.7500\n",
            "Epoch 37/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7988 - accuracy: 0.6976 - val_loss: 0.7255 - val_accuracy: 0.7635\n",
            "Epoch 38/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7697 - accuracy: 0.7111 - val_loss: 0.7058 - val_accuracy: 0.7635\n",
            "Epoch 39/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.8020 - accuracy: 0.6993 - val_loss: 0.7166 - val_accuracy: 0.8108\n",
            "Epoch 40/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7469 - accuracy: 0.7213 - val_loss: 0.6796 - val_accuracy: 0.7703\n",
            "Epoch 41/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.8030 - accuracy: 0.6875 - val_loss: 0.7045 - val_accuracy: 0.7838\n",
            "Epoch 42/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7895 - accuracy: 0.6909 - val_loss: 0.7039 - val_accuracy: 0.7838\n",
            "Epoch 43/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7900 - accuracy: 0.6858 - val_loss: 0.6915 - val_accuracy: 0.7568\n",
            "Epoch 44/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7802 - accuracy: 0.6892 - val_loss: 0.6718 - val_accuracy: 0.7770\n",
            "Epoch 45/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7714 - accuracy: 0.7078 - val_loss: 0.7048 - val_accuracy: 0.7838\n",
            "Epoch 46/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7843 - accuracy: 0.6841 - val_loss: 0.7027 - val_accuracy: 0.7905\n",
            "Epoch 47/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7767 - accuracy: 0.7111 - val_loss: 0.6963 - val_accuracy: 0.7635\n",
            "Epoch 48/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7781 - accuracy: 0.7280 - val_loss: 0.7005 - val_accuracy: 0.7568\n",
            "Epoch 49/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7437 - accuracy: 0.7449 - val_loss: 0.6766 - val_accuracy: 0.7973\n",
            "Epoch 50/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7717 - accuracy: 0.7416 - val_loss: 0.7088 - val_accuracy: 0.7973\n",
            "Epoch 51/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7765 - accuracy: 0.7196 - val_loss: 0.7080 - val_accuracy: 0.7500\n",
            "Epoch 52/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7863 - accuracy: 0.7027 - val_loss: 0.7083 - val_accuracy: 0.7905\n",
            "Epoch 53/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7609 - accuracy: 0.7230 - val_loss: 0.6936 - val_accuracy: 0.7635\n",
            "Epoch 54/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7731 - accuracy: 0.7247 - val_loss: 0.7020 - val_accuracy: 0.7230\n",
            "Epoch 55/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7461 - accuracy: 0.7230 - val_loss: 0.6895 - val_accuracy: 0.7703\n",
            "Epoch 56/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7631 - accuracy: 0.7078 - val_loss: 0.6943 - val_accuracy: 0.7635\n",
            "Epoch 57/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7393 - accuracy: 0.7382 - val_loss: 0.6789 - val_accuracy: 0.7905\n",
            "Epoch 58/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7484 - accuracy: 0.7111 - val_loss: 0.6573 - val_accuracy: 0.8243\n",
            "Epoch 59/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7727 - accuracy: 0.6926 - val_loss: 0.7024 - val_accuracy: 0.7770\n",
            "Epoch 60/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8109 - accuracy: 0.7111 - val_loss: 0.7010 - val_accuracy: 0.7838\n",
            "Epoch 61/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7520 - accuracy: 0.7280 - val_loss: 0.6889 - val_accuracy: 0.7770\n",
            "Epoch 62/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7597 - accuracy: 0.7314 - val_loss: 0.6716 - val_accuracy: 0.8041\n",
            "Epoch 63/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7245 - accuracy: 0.7230 - val_loss: 0.6727 - val_accuracy: 0.7635\n",
            "Epoch 64/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.8042 - accuracy: 0.6926 - val_loss: 0.7147 - val_accuracy: 0.7703\n",
            "Epoch 65/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7667 - accuracy: 0.7264 - val_loss: 0.6597 - val_accuracy: 0.8041\n",
            "Epoch 66/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7287 - accuracy: 0.7399 - val_loss: 0.6421 - val_accuracy: 0.7838\n",
            "Epoch 67/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7176 - accuracy: 0.7399 - val_loss: 0.6283 - val_accuracy: 0.7905\n",
            "Epoch 68/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7481 - accuracy: 0.7213 - val_loss: 0.6798 - val_accuracy: 0.7838\n",
            "Epoch 69/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7863 - accuracy: 0.6757 - val_loss: 0.6773 - val_accuracy: 0.7770\n",
            "Epoch 70/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7074 - accuracy: 0.7213 - val_loss: 0.6315 - val_accuracy: 0.8041\n",
            "Epoch 71/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7313 - accuracy: 0.7365 - val_loss: 0.6732 - val_accuracy: 0.7770\n",
            "Epoch 72/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7638 - accuracy: 0.7348 - val_loss: 0.6637 - val_accuracy: 0.7973\n",
            "Epoch 73/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7442 - accuracy: 0.7061 - val_loss: 0.6501 - val_accuracy: 0.8446\n",
            "Epoch 74/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7433 - accuracy: 0.7230 - val_loss: 0.6429 - val_accuracy: 0.7905\n",
            "Epoch 75/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7246 - accuracy: 0.7179 - val_loss: 0.6487 - val_accuracy: 0.8176\n",
            "Epoch 76/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7739 - accuracy: 0.6875 - val_loss: 0.6802 - val_accuracy: 0.8176\n",
            "Epoch 77/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7665 - accuracy: 0.6926 - val_loss: 0.6695 - val_accuracy: 0.8243\n",
            "Epoch 78/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7361 - accuracy: 0.7280 - val_loss: 0.6339 - val_accuracy: 0.8108\n",
            "Epoch 79/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7594 - accuracy: 0.7280 - val_loss: 0.6513 - val_accuracy: 0.7973\n",
            "Epoch 80/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7368 - accuracy: 0.7247 - val_loss: 0.6427 - val_accuracy: 0.7973\n",
            "Epoch 81/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7048 - accuracy: 0.7230 - val_loss: 0.6236 - val_accuracy: 0.8108\n",
            "Epoch 82/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7242 - accuracy: 0.7213 - val_loss: 0.6631 - val_accuracy: 0.7838\n",
            "Epoch 83/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7345 - accuracy: 0.7500 - val_loss: 0.6568 - val_accuracy: 0.8108\n",
            "Epoch 84/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7219 - accuracy: 0.7280 - val_loss: 0.6293 - val_accuracy: 0.8243\n",
            "Epoch 85/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7143 - accuracy: 0.7162 - val_loss: 0.6336 - val_accuracy: 0.7973\n",
            "Epoch 86/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7593 - accuracy: 0.7230 - val_loss: 0.6491 - val_accuracy: 0.8041\n",
            "Epoch 87/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6944 - accuracy: 0.7416 - val_loss: 0.6298 - val_accuracy: 0.8108\n",
            "Epoch 88/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7321 - accuracy: 0.7264 - val_loss: 0.6573 - val_accuracy: 0.7905\n",
            "Epoch 89/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7657 - accuracy: 0.7078 - val_loss: 0.6643 - val_accuracy: 0.8311\n",
            "Epoch 90/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6970 - accuracy: 0.7432 - val_loss: 0.6110 - val_accuracy: 0.8176\n",
            "Epoch 91/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7306 - accuracy: 0.7416 - val_loss: 0.6580 - val_accuracy: 0.8176\n",
            "Epoch 92/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7384 - accuracy: 0.7213 - val_loss: 0.6730 - val_accuracy: 0.7973\n",
            "Epoch 93/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7416 - accuracy: 0.7280 - val_loss: 0.6406 - val_accuracy: 0.8176\n",
            "Epoch 94/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7767 - accuracy: 0.7213 - val_loss: 0.6626 - val_accuracy: 0.8176\n",
            "Epoch 95/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7387 - accuracy: 0.7382 - val_loss: 0.6504 - val_accuracy: 0.8041\n",
            "Epoch 96/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7056 - accuracy: 0.7568 - val_loss: 0.6386 - val_accuracy: 0.7770\n",
            "Epoch 97/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7221 - accuracy: 0.7331 - val_loss: 0.6401 - val_accuracy: 0.8041\n",
            "Epoch 98/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7098 - accuracy: 0.7111 - val_loss: 0.6445 - val_accuracy: 0.8176\n",
            "Epoch 99/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7510 - accuracy: 0.7145 - val_loss: 0.6410 - val_accuracy: 0.8243\n",
            "Epoch 100/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7411 - accuracy: 0.7179 - val_loss: 0.6286 - val_accuracy: 0.8108\n",
            "Epoch 101/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7315 - accuracy: 0.7348 - val_loss: 0.6261 - val_accuracy: 0.7905\n",
            "Epoch 102/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7207 - accuracy: 0.7399 - val_loss: 0.6292 - val_accuracy: 0.8108\n",
            "Epoch 103/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7176 - accuracy: 0.6993 - val_loss: 0.6427 - val_accuracy: 0.8108\n",
            "Epoch 104/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7386 - accuracy: 0.6926 - val_loss: 0.6546 - val_accuracy: 0.7838\n",
            "Epoch 105/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7263 - accuracy: 0.7230 - val_loss: 0.6219 - val_accuracy: 0.7703\n",
            "Epoch 106/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7321 - accuracy: 0.7230 - val_loss: 0.6558 - val_accuracy: 0.7838\n",
            "Epoch 107/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7131 - accuracy: 0.7432 - val_loss: 0.6400 - val_accuracy: 0.7905\n",
            "Epoch 108/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7127 - accuracy: 0.7432 - val_loss: 0.6411 - val_accuracy: 0.8311\n",
            "Epoch 109/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7485 - accuracy: 0.7449 - val_loss: 0.6422 - val_accuracy: 0.7973\n",
            "Epoch 110/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7352 - accuracy: 0.7449 - val_loss: 0.6216 - val_accuracy: 0.8243\n",
            "Epoch 111/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7127 - accuracy: 0.7365 - val_loss: 0.6137 - val_accuracy: 0.8311\n",
            "Epoch 112/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6830 - accuracy: 0.7365 - val_loss: 0.5846 - val_accuracy: 0.8446\n",
            "Epoch 113/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7337 - accuracy: 0.7027 - val_loss: 0.6353 - val_accuracy: 0.8108\n",
            "Epoch 114/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7193 - accuracy: 0.7247 - val_loss: 0.6157 - val_accuracy: 0.8176\n",
            "Epoch 115/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7086 - accuracy: 0.7162 - val_loss: 0.6189 - val_accuracy: 0.8176\n",
            "Epoch 116/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7084 - accuracy: 0.7145 - val_loss: 0.6161 - val_accuracy: 0.7973\n",
            "Epoch 117/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7049 - accuracy: 0.7264 - val_loss: 0.6312 - val_accuracy: 0.8041\n",
            "Epoch 118/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6923 - accuracy: 0.7517 - val_loss: 0.6249 - val_accuracy: 0.7838\n",
            "Epoch 119/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7435 - accuracy: 0.7280 - val_loss: 0.6374 - val_accuracy: 0.8041\n",
            "Epoch 120/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6975 - accuracy: 0.7483 - val_loss: 0.6094 - val_accuracy: 0.8108\n",
            "Epoch 121/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7214 - accuracy: 0.7230 - val_loss: 0.6161 - val_accuracy: 0.8041\n",
            "Epoch 122/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7126 - accuracy: 0.7382 - val_loss: 0.6209 - val_accuracy: 0.8378\n",
            "Epoch 123/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7113 - accuracy: 0.7601 - val_loss: 0.6311 - val_accuracy: 0.8041\n",
            "Epoch 124/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6903 - accuracy: 0.7601 - val_loss: 0.6128 - val_accuracy: 0.7973\n",
            "Epoch 125/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7143 - accuracy: 0.7382 - val_loss: 0.6223 - val_accuracy: 0.8176\n",
            "Epoch 126/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7101 - accuracy: 0.7264 - val_loss: 0.6299 - val_accuracy: 0.7905\n",
            "Epoch 127/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6804 - accuracy: 0.7584 - val_loss: 0.6029 - val_accuracy: 0.8176\n",
            "Epoch 128/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6787 - accuracy: 0.7399 - val_loss: 0.6254 - val_accuracy: 0.8243\n",
            "Epoch 129/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7140 - accuracy: 0.7061 - val_loss: 0.6329 - val_accuracy: 0.8378\n",
            "Epoch 130/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7103 - accuracy: 0.7449 - val_loss: 0.6393 - val_accuracy: 0.7973\n",
            "Epoch 131/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7043 - accuracy: 0.7635 - val_loss: 0.6486 - val_accuracy: 0.7770\n",
            "Epoch 132/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7119 - accuracy: 0.7432 - val_loss: 0.6256 - val_accuracy: 0.8108\n",
            "Epoch 133/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7533 - accuracy: 0.7247 - val_loss: 0.6362 - val_accuracy: 0.8176\n",
            "Epoch 134/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7202 - accuracy: 0.7416 - val_loss: 0.6115 - val_accuracy: 0.7973\n",
            "Epoch 135/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7111 - accuracy: 0.7449 - val_loss: 0.6104 - val_accuracy: 0.8041\n",
            "Epoch 136/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6838 - accuracy: 0.7230 - val_loss: 0.5878 - val_accuracy: 0.8041\n",
            "Epoch 137/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7094 - accuracy: 0.7297 - val_loss: 0.5907 - val_accuracy: 0.8176\n",
            "Epoch 138/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7122 - accuracy: 0.7264 - val_loss: 0.5921 - val_accuracy: 0.8041\n",
            "Epoch 139/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6729 - accuracy: 0.7365 - val_loss: 0.5995 - val_accuracy: 0.8243\n",
            "Epoch 140/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6671 - accuracy: 0.7736 - val_loss: 0.6016 - val_accuracy: 0.7973\n",
            "Epoch 141/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6996 - accuracy: 0.7500 - val_loss: 0.6139 - val_accuracy: 0.8108\n",
            "Epoch 142/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7131 - accuracy: 0.7280 - val_loss: 0.5956 - val_accuracy: 0.8041\n",
            "Epoch 143/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7147 - accuracy: 0.7686 - val_loss: 0.6186 - val_accuracy: 0.8108\n",
            "Epoch 144/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6994 - accuracy: 0.7196 - val_loss: 0.6248 - val_accuracy: 0.7905\n",
            "Epoch 145/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7085 - accuracy: 0.7078 - val_loss: 0.6114 - val_accuracy: 0.8243\n",
            "Epoch 146/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6725 - accuracy: 0.7399 - val_loss: 0.5976 - val_accuracy: 0.8311\n",
            "Epoch 147/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6702 - accuracy: 0.7500 - val_loss: 0.5876 - val_accuracy: 0.8514\n",
            "Epoch 148/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7068 - accuracy: 0.7314 - val_loss: 0.6177 - val_accuracy: 0.8176\n",
            "Epoch 149/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7145 - accuracy: 0.7568 - val_loss: 0.6266 - val_accuracy: 0.8311\n",
            "Epoch 150/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6513 - accuracy: 0.7618 - val_loss: 0.5911 - val_accuracy: 0.8243\n",
            "Epoch 151/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6930 - accuracy: 0.7314 - val_loss: 0.5942 - val_accuracy: 0.8378\n",
            "Epoch 152/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6775 - accuracy: 0.7601 - val_loss: 0.5869 - val_accuracy: 0.8378\n",
            "Epoch 153/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7081 - accuracy: 0.7264 - val_loss: 0.6171 - val_accuracy: 0.8176\n",
            "Epoch 154/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7389 - accuracy: 0.7027 - val_loss: 0.6204 - val_accuracy: 0.8243\n",
            "Epoch 155/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6951 - accuracy: 0.7432 - val_loss: 0.6052 - val_accuracy: 0.8378\n",
            "Epoch 156/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7057 - accuracy: 0.7399 - val_loss: 0.6107 - val_accuracy: 0.8378\n",
            "Epoch 157/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6741 - accuracy: 0.7365 - val_loss: 0.6018 - val_accuracy: 0.8108\n",
            "Epoch 158/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7023 - accuracy: 0.7297 - val_loss: 0.6035 - val_accuracy: 0.8041\n",
            "Epoch 159/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6578 - accuracy: 0.7669 - val_loss: 0.5846 - val_accuracy: 0.8514\n",
            "Epoch 160/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6390 - accuracy: 0.7517 - val_loss: 0.5958 - val_accuracy: 0.8108\n",
            "Epoch 161/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6942 - accuracy: 0.7669 - val_loss: 0.6003 - val_accuracy: 0.8243\n",
            "Epoch 162/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6891 - accuracy: 0.7584 - val_loss: 0.5991 - val_accuracy: 0.8108\n",
            "Epoch 163/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6904 - accuracy: 0.7247 - val_loss: 0.5983 - val_accuracy: 0.8311\n",
            "Epoch 164/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7047 - accuracy: 0.7230 - val_loss: 0.5971 - val_accuracy: 0.8108\n",
            "Epoch 165/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6974 - accuracy: 0.7264 - val_loss: 0.5911 - val_accuracy: 0.8243\n",
            "Epoch 166/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6602 - accuracy: 0.7466 - val_loss: 0.5803 - val_accuracy: 0.7973\n",
            "Epoch 167/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6902 - accuracy: 0.7365 - val_loss: 0.5775 - val_accuracy: 0.8108\n",
            "Epoch 168/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6762 - accuracy: 0.7466 - val_loss: 0.5947 - val_accuracy: 0.8378\n",
            "Epoch 169/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6933 - accuracy: 0.7635 - val_loss: 0.5995 - val_accuracy: 0.8041\n",
            "Epoch 170/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6824 - accuracy: 0.7399 - val_loss: 0.6058 - val_accuracy: 0.8041\n",
            "Epoch 171/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7261 - accuracy: 0.7162 - val_loss: 0.6439 - val_accuracy: 0.8041\n",
            "Epoch 172/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6872 - accuracy: 0.7247 - val_loss: 0.5966 - val_accuracy: 0.8243\n",
            "Epoch 173/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7106 - accuracy: 0.7399 - val_loss: 0.6113 - val_accuracy: 0.8378\n",
            "Epoch 174/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6906 - accuracy: 0.7432 - val_loss: 0.6128 - val_accuracy: 0.8176\n",
            "Epoch 175/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6932 - accuracy: 0.7230 - val_loss: 0.5756 - val_accuracy: 0.8311\n",
            "Epoch 176/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6916 - accuracy: 0.7365 - val_loss: 0.5732 - val_accuracy: 0.8311\n",
            "Epoch 177/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6868 - accuracy: 0.7297 - val_loss: 0.5875 - val_accuracy: 0.8446\n",
            "Epoch 178/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6573 - accuracy: 0.7534 - val_loss: 0.5933 - val_accuracy: 0.8243\n",
            "Epoch 179/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6957 - accuracy: 0.7213 - val_loss: 0.5897 - val_accuracy: 0.8176\n",
            "Epoch 180/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6727 - accuracy: 0.7568 - val_loss: 0.5811 - val_accuracy: 0.8108\n",
            "Epoch 181/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6923 - accuracy: 0.7382 - val_loss: 0.5868 - val_accuracy: 0.8378\n",
            "Epoch 182/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7053 - accuracy: 0.7179 - val_loss: 0.5885 - val_accuracy: 0.8243\n",
            "Epoch 183/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6793 - accuracy: 0.7720 - val_loss: 0.5726 - val_accuracy: 0.8446\n",
            "Epoch 184/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6769 - accuracy: 0.7500 - val_loss: 0.5756 - val_accuracy: 0.8446\n",
            "Epoch 185/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6781 - accuracy: 0.7568 - val_loss: 0.5648 - val_accuracy: 0.8514\n",
            "Epoch 186/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6650 - accuracy: 0.7568 - val_loss: 0.5777 - val_accuracy: 0.8243\n",
            "Epoch 187/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6689 - accuracy: 0.7584 - val_loss: 0.5980 - val_accuracy: 0.8176\n",
            "Epoch 188/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6692 - accuracy: 0.7568 - val_loss: 0.5863 - val_accuracy: 0.8108\n",
            "Epoch 189/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6969 - accuracy: 0.7247 - val_loss: 0.5812 - val_accuracy: 0.8378\n",
            "Epoch 190/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6632 - accuracy: 0.7584 - val_loss: 0.5834 - val_accuracy: 0.8176\n",
            "Epoch 191/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6659 - accuracy: 0.7534 - val_loss: 0.5860 - val_accuracy: 0.8243\n",
            "Epoch 192/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6704 - accuracy: 0.7534 - val_loss: 0.5677 - val_accuracy: 0.8446\n",
            "Epoch 193/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6561 - accuracy: 0.7635 - val_loss: 0.5744 - val_accuracy: 0.8378\n",
            "Epoch 194/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6343 - accuracy: 0.7720 - val_loss: 0.5748 - val_accuracy: 0.8041\n",
            "Epoch 195/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6515 - accuracy: 0.7399 - val_loss: 0.5739 - val_accuracy: 0.8108\n",
            "Epoch 196/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6665 - accuracy: 0.7365 - val_loss: 0.6054 - val_accuracy: 0.8041\n",
            "Epoch 197/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6705 - accuracy: 0.7416 - val_loss: 0.5795 - val_accuracy: 0.8243\n",
            "Epoch 198/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6916 - accuracy: 0.7297 - val_loss: 0.6139 - val_accuracy: 0.8041\n",
            "Epoch 199/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7163 - accuracy: 0.7230 - val_loss: 0.6035 - val_accuracy: 0.8176\n",
            "Epoch 200/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6610 - accuracy: 0.7416 - val_loss: 0.5582 - val_accuracy: 0.8446\n",
            "Epoch 201/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6499 - accuracy: 0.7686 - val_loss: 0.5580 - val_accuracy: 0.8446\n",
            "Epoch 202/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6602 - accuracy: 0.7669 - val_loss: 0.5733 - val_accuracy: 0.8378\n",
            "Epoch 203/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6917 - accuracy: 0.7551 - val_loss: 0.5920 - val_accuracy: 0.8243\n",
            "Epoch 204/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6571 - accuracy: 0.7483 - val_loss: 0.5632 - val_accuracy: 0.8378\n",
            "Epoch 205/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6984 - accuracy: 0.7416 - val_loss: 0.6092 - val_accuracy: 0.8108\n",
            "Epoch 206/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6473 - accuracy: 0.7568 - val_loss: 0.5936 - val_accuracy: 0.8243\n",
            "Epoch 207/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6849 - accuracy: 0.7669 - val_loss: 0.5870 - val_accuracy: 0.8041\n",
            "Epoch 208/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7024 - accuracy: 0.7365 - val_loss: 0.5996 - val_accuracy: 0.7770\n",
            "Epoch 209/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6652 - accuracy: 0.7466 - val_loss: 0.5720 - val_accuracy: 0.8446\n",
            "Epoch 210/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6442 - accuracy: 0.7601 - val_loss: 0.5701 - val_accuracy: 0.8378\n",
            "Epoch 211/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6902 - accuracy: 0.7551 - val_loss: 0.6113 - val_accuracy: 0.8446\n",
            "Epoch 212/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7135 - accuracy: 0.7264 - val_loss: 0.5963 - val_accuracy: 0.8243\n",
            "Epoch 213/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7169 - accuracy: 0.7247 - val_loss: 0.5840 - val_accuracy: 0.8378\n",
            "Epoch 214/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7067 - accuracy: 0.7179 - val_loss: 0.5808 - val_accuracy: 0.8311\n",
            "Epoch 215/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7017 - accuracy: 0.7264 - val_loss: 0.6098 - val_accuracy: 0.8243\n",
            "Epoch 216/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6993 - accuracy: 0.7010 - val_loss: 0.5808 - val_accuracy: 0.8378\n",
            "Epoch 217/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6873 - accuracy: 0.7314 - val_loss: 0.5761 - val_accuracy: 0.8446\n",
            "Epoch 218/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6731 - accuracy: 0.7213 - val_loss: 0.5820 - val_accuracy: 0.8514\n",
            "Epoch 219/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6851 - accuracy: 0.7247 - val_loss: 0.5698 - val_accuracy: 0.8446\n",
            "Epoch 220/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6667 - accuracy: 0.7466 - val_loss: 0.5717 - val_accuracy: 0.8311\n",
            "Epoch 221/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6718 - accuracy: 0.7449 - val_loss: 0.6131 - val_accuracy: 0.8176\n",
            "Epoch 222/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6806 - accuracy: 0.7314 - val_loss: 0.5726 - val_accuracy: 0.8311\n",
            "Epoch 223/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7262 - accuracy: 0.7280 - val_loss: 0.5984 - val_accuracy: 0.8446\n",
            "Epoch 224/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6654 - accuracy: 0.7365 - val_loss: 0.5784 - val_accuracy: 0.8243\n",
            "Epoch 225/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6769 - accuracy: 0.7314 - val_loss: 0.5660 - val_accuracy: 0.8311\n",
            "Epoch 226/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6302 - accuracy: 0.7551 - val_loss: 0.5811 - val_accuracy: 0.8176\n",
            "Epoch 227/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6691 - accuracy: 0.7179 - val_loss: 0.5833 - val_accuracy: 0.7973\n",
            "Epoch 228/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6591 - accuracy: 0.7449 - val_loss: 0.5923 - val_accuracy: 0.8108\n",
            "Epoch 229/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6959 - accuracy: 0.7247 - val_loss: 0.6109 - val_accuracy: 0.8378\n",
            "Epoch 230/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7007 - accuracy: 0.7331 - val_loss: 0.6110 - val_accuracy: 0.8108\n",
            "Epoch 231/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7162 - accuracy: 0.7382 - val_loss: 0.5796 - val_accuracy: 0.8311\n",
            "Epoch 232/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6653 - accuracy: 0.7365 - val_loss: 0.5772 - val_accuracy: 0.8041\n",
            "Epoch 233/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6695 - accuracy: 0.7264 - val_loss: 0.5636 - val_accuracy: 0.8243\n",
            "Epoch 234/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7330 - accuracy: 0.7044 - val_loss: 0.6037 - val_accuracy: 0.8108\n",
            "Epoch 235/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6627 - accuracy: 0.7432 - val_loss: 0.5714 - val_accuracy: 0.8311\n",
            "Epoch 236/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6539 - accuracy: 0.7399 - val_loss: 0.5857 - val_accuracy: 0.8041\n",
            "Epoch 237/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6855 - accuracy: 0.7314 - val_loss: 0.5968 - val_accuracy: 0.8176\n",
            "Epoch 238/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6797 - accuracy: 0.7297 - val_loss: 0.5868 - val_accuracy: 0.7973\n",
            "Epoch 239/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6471 - accuracy: 0.7652 - val_loss: 0.5908 - val_accuracy: 0.8041\n",
            "Epoch 240/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6897 - accuracy: 0.7466 - val_loss: 0.5801 - val_accuracy: 0.8311\n",
            "Epoch 241/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7107 - accuracy: 0.7449 - val_loss: 0.5860 - val_accuracy: 0.8243\n",
            "Epoch 242/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6589 - accuracy: 0.7466 - val_loss: 0.5699 - val_accuracy: 0.8243\n",
            "Epoch 243/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6722 - accuracy: 0.7500 - val_loss: 0.5846 - val_accuracy: 0.8041\n",
            "Epoch 244/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6688 - accuracy: 0.7365 - val_loss: 0.5698 - val_accuracy: 0.8243\n",
            "Epoch 245/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7069 - accuracy: 0.6976 - val_loss: 0.6051 - val_accuracy: 0.7838\n",
            "Epoch 246/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6502 - accuracy: 0.7179 - val_loss: 0.5862 - val_accuracy: 0.8041\n",
            "Epoch 247/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6361 - accuracy: 0.7838 - val_loss: 0.5874 - val_accuracy: 0.8311\n",
            "Epoch 248/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6669 - accuracy: 0.7483 - val_loss: 0.5628 - val_accuracy: 0.8378\n",
            "Epoch 249/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6496 - accuracy: 0.7601 - val_loss: 0.5575 - val_accuracy: 0.8514\n",
            "Epoch 250/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7162 - accuracy: 0.7078 - val_loss: 0.5901 - val_accuracy: 0.8514\n",
            "Epoch 251/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6528 - accuracy: 0.7568 - val_loss: 0.5760 - val_accuracy: 0.8716\n",
            "Epoch 252/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6897 - accuracy: 0.7314 - val_loss: 0.5932 - val_accuracy: 0.8108\n",
            "Epoch 253/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6271 - accuracy: 0.7584 - val_loss: 0.5787 - val_accuracy: 0.8243\n",
            "Epoch 254/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6881 - accuracy: 0.7416 - val_loss: 0.6052 - val_accuracy: 0.8108\n",
            "Epoch 255/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7209 - accuracy: 0.7331 - val_loss: 0.5997 - val_accuracy: 0.8176\n",
            "Epoch 256/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6416 - accuracy: 0.7720 - val_loss: 0.5673 - val_accuracy: 0.8446\n",
            "Epoch 257/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6793 - accuracy: 0.7551 - val_loss: 0.5738 - val_accuracy: 0.8716\n",
            "Epoch 258/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6313 - accuracy: 0.7720 - val_loss: 0.5688 - val_accuracy: 0.8378\n",
            "Epoch 259/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6619 - accuracy: 0.7331 - val_loss: 0.5703 - val_accuracy: 0.8311\n",
            "Epoch 260/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7023 - accuracy: 0.7669 - val_loss: 0.5753 - val_accuracy: 0.8378\n",
            "Epoch 261/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6755 - accuracy: 0.7618 - val_loss: 0.5992 - val_accuracy: 0.8108\n",
            "Epoch 262/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6762 - accuracy: 0.7365 - val_loss: 0.5905 - val_accuracy: 0.8041\n",
            "Epoch 263/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6713 - accuracy: 0.7449 - val_loss: 0.5743 - val_accuracy: 0.8108\n",
            "Epoch 264/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6505 - accuracy: 0.7584 - val_loss: 0.5873 - val_accuracy: 0.8176\n",
            "Epoch 265/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6768 - accuracy: 0.7264 - val_loss: 0.5943 - val_accuracy: 0.8311\n",
            "Epoch 266/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6280 - accuracy: 0.7686 - val_loss: 0.5458 - val_accuracy: 0.8243\n",
            "Epoch 267/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6679 - accuracy: 0.7601 - val_loss: 0.5704 - val_accuracy: 0.8311\n",
            "Epoch 268/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6863 - accuracy: 0.7331 - val_loss: 0.5828 - val_accuracy: 0.8108\n",
            "Epoch 269/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6642 - accuracy: 0.7196 - val_loss: 0.5768 - val_accuracy: 0.8311\n",
            "Epoch 270/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6756 - accuracy: 0.7348 - val_loss: 0.5891 - val_accuracy: 0.8243\n",
            "Epoch 271/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6704 - accuracy: 0.7534 - val_loss: 0.5676 - val_accuracy: 0.8311\n",
            "Epoch 272/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6519 - accuracy: 0.7635 - val_loss: 0.5535 - val_accuracy: 0.8378\n",
            "Epoch 273/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6662 - accuracy: 0.7449 - val_loss: 0.5603 - val_accuracy: 0.8378\n",
            "Epoch 274/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6488 - accuracy: 0.7584 - val_loss: 0.5636 - val_accuracy: 0.8176\n",
            "Epoch 275/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6174 - accuracy: 0.7838 - val_loss: 0.5931 - val_accuracy: 0.8176\n",
            "Epoch 276/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6921 - accuracy: 0.7297 - val_loss: 0.5931 - val_accuracy: 0.8378\n",
            "Epoch 277/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7276 - accuracy: 0.7128 - val_loss: 0.5946 - val_accuracy: 0.8176\n",
            "Epoch 278/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6477 - accuracy: 0.7601 - val_loss: 0.5680 - val_accuracy: 0.8378\n",
            "Epoch 279/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6350 - accuracy: 0.7753 - val_loss: 0.5490 - val_accuracy: 0.8311\n",
            "Epoch 280/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6489 - accuracy: 0.7466 - val_loss: 0.5568 - val_accuracy: 0.8378\n",
            "Epoch 281/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6104 - accuracy: 0.7770 - val_loss: 0.5672 - val_accuracy: 0.8446\n",
            "Epoch 282/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6438 - accuracy: 0.7652 - val_loss: 0.5560 - val_accuracy: 0.8311\n",
            "Epoch 283/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6607 - accuracy: 0.7669 - val_loss: 0.5718 - val_accuracy: 0.8446\n",
            "Epoch 284/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6640 - accuracy: 0.7584 - val_loss: 0.5708 - val_accuracy: 0.8311\n",
            "Epoch 285/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6237 - accuracy: 0.7669 - val_loss: 0.5423 - val_accuracy: 0.8311\n",
            "Epoch 286/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6796 - accuracy: 0.7466 - val_loss: 0.5668 - val_accuracy: 0.8243\n",
            "Epoch 287/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6691 - accuracy: 0.7568 - val_loss: 0.5607 - val_accuracy: 0.8311\n",
            "Epoch 288/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6771 - accuracy: 0.7652 - val_loss: 0.5750 - val_accuracy: 0.8243\n",
            "Epoch 289/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6494 - accuracy: 0.7669 - val_loss: 0.5577 - val_accuracy: 0.8243\n",
            "Epoch 290/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6476 - accuracy: 0.7348 - val_loss: 0.5651 - val_accuracy: 0.8311\n",
            "Epoch 291/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6469 - accuracy: 0.7601 - val_loss: 0.5568 - val_accuracy: 0.8311\n",
            "Epoch 292/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6597 - accuracy: 0.7365 - val_loss: 0.5592 - val_accuracy: 0.8311\n",
            "Epoch 293/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6463 - accuracy: 0.7314 - val_loss: 0.5445 - val_accuracy: 0.8311\n",
            "Epoch 294/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6216 - accuracy: 0.7635 - val_loss: 0.5572 - val_accuracy: 0.8311\n",
            "Epoch 295/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6774 - accuracy: 0.7213 - val_loss: 0.6066 - val_accuracy: 0.8108\n",
            "Epoch 296/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6886 - accuracy: 0.7432 - val_loss: 0.6092 - val_accuracy: 0.8176\n",
            "Epoch 297/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6987 - accuracy: 0.7280 - val_loss: 0.5867 - val_accuracy: 0.8243\n",
            "Epoch 298/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6333 - accuracy: 0.7770 - val_loss: 0.5583 - val_accuracy: 0.8581\n",
            "Epoch 299/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6609 - accuracy: 0.7483 - val_loss: 0.5726 - val_accuracy: 0.8378\n",
            "Epoch 300/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6820 - accuracy: 0.7449 - val_loss: 0.5600 - val_accuracy: 0.8311\n",
            "Epoch 301/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6772 - accuracy: 0.7382 - val_loss: 0.5788 - val_accuracy: 0.8176\n",
            "Epoch 302/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6123 - accuracy: 0.7686 - val_loss: 0.5523 - val_accuracy: 0.8311\n",
            "Epoch 303/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6578 - accuracy: 0.7517 - val_loss: 0.5653 - val_accuracy: 0.8514\n",
            "Epoch 304/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6731 - accuracy: 0.7584 - val_loss: 0.5775 - val_accuracy: 0.8311\n",
            "Epoch 305/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7563 - accuracy: 0.6993 - val_loss: 0.5968 - val_accuracy: 0.8378\n",
            "Epoch 306/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6817 - accuracy: 0.7449 - val_loss: 0.5732 - val_accuracy: 0.8243\n",
            "Epoch 307/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6363 - accuracy: 0.7669 - val_loss: 0.5524 - val_accuracy: 0.8446\n",
            "Epoch 308/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6956 - accuracy: 0.7500 - val_loss: 0.5784 - val_accuracy: 0.8311\n",
            "Epoch 309/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7116 - accuracy: 0.7111 - val_loss: 0.6023 - val_accuracy: 0.8378\n",
            "Epoch 310/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6585 - accuracy: 0.7753 - val_loss: 0.5777 - val_accuracy: 0.8243\n",
            "Epoch 311/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6588 - accuracy: 0.7314 - val_loss: 0.5965 - val_accuracy: 0.8108\n",
            "Epoch 312/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6605 - accuracy: 0.7551 - val_loss: 0.5688 - val_accuracy: 0.8716\n",
            "Epoch 313/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7068 - accuracy: 0.7399 - val_loss: 0.6038 - val_accuracy: 0.8378\n",
            "Epoch 314/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6967 - accuracy: 0.7382 - val_loss: 0.5866 - val_accuracy: 0.8243\n",
            "Epoch 315/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7111 - accuracy: 0.7128 - val_loss: 0.5961 - val_accuracy: 0.8176\n",
            "Epoch 316/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6549 - accuracy: 0.7635 - val_loss: 0.5589 - val_accuracy: 0.8243\n",
            "Epoch 317/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6535 - accuracy: 0.7466 - val_loss: 0.5435 - val_accuracy: 0.8311\n",
            "Epoch 318/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6761 - accuracy: 0.7314 - val_loss: 0.5657 - val_accuracy: 0.8378\n",
            "Epoch 319/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6321 - accuracy: 0.7534 - val_loss: 0.5652 - val_accuracy: 0.8243\n",
            "Epoch 320/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6859 - accuracy: 0.7466 - val_loss: 0.5855 - val_accuracy: 0.8176\n",
            "Epoch 321/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6756 - accuracy: 0.7348 - val_loss: 0.5688 - val_accuracy: 0.8446\n",
            "Epoch 322/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6316 - accuracy: 0.7720 - val_loss: 0.5328 - val_accuracy: 0.8446\n",
            "Epoch 323/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6389 - accuracy: 0.7483 - val_loss: 0.5401 - val_accuracy: 0.8378\n",
            "Epoch 324/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6669 - accuracy: 0.7348 - val_loss: 0.5694 - val_accuracy: 0.7973\n",
            "Epoch 325/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6306 - accuracy: 0.7534 - val_loss: 0.5622 - val_accuracy: 0.8378\n",
            "Epoch 326/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6492 - accuracy: 0.7517 - val_loss: 0.5642 - val_accuracy: 0.8311\n",
            "Epoch 327/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6655 - accuracy: 0.7382 - val_loss: 0.5602 - val_accuracy: 0.8649\n",
            "Epoch 328/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6764 - accuracy: 0.7348 - val_loss: 0.5677 - val_accuracy: 0.8514\n",
            "Epoch 329/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6028 - accuracy: 0.7635 - val_loss: 0.5358 - val_accuracy: 0.8649\n",
            "Epoch 330/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6334 - accuracy: 0.7280 - val_loss: 0.5244 - val_accuracy: 0.7973\n",
            "Epoch 331/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5991 - accuracy: 0.7770 - val_loss: 0.5164 - val_accuracy: 0.8378\n",
            "Epoch 332/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6319 - accuracy: 0.7584 - val_loss: 0.5556 - val_accuracy: 0.8378\n",
            "Epoch 333/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6612 - accuracy: 0.7500 - val_loss: 0.5433 - val_accuracy: 0.8514\n",
            "Epoch 334/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6651 - accuracy: 0.7399 - val_loss: 0.5639 - val_accuracy: 0.8446\n",
            "Epoch 335/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6552 - accuracy: 0.7365 - val_loss: 0.5729 - val_accuracy: 0.8446\n",
            "Epoch 336/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6571 - accuracy: 0.7500 - val_loss: 0.5512 - val_accuracy: 0.8446\n",
            "Epoch 337/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6217 - accuracy: 0.7669 - val_loss: 0.5444 - val_accuracy: 0.8378\n",
            "Epoch 338/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6463 - accuracy: 0.7399 - val_loss: 0.5700 - val_accuracy: 0.8176\n",
            "Epoch 339/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6492 - accuracy: 0.7500 - val_loss: 0.5594 - val_accuracy: 0.8243\n",
            "Epoch 340/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6869 - accuracy: 0.7095 - val_loss: 0.5819 - val_accuracy: 0.8176\n",
            "Epoch 341/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6515 - accuracy: 0.7416 - val_loss: 0.5794 - val_accuracy: 0.8243\n",
            "Epoch 342/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6627 - accuracy: 0.7416 - val_loss: 0.5613 - val_accuracy: 0.8378\n",
            "Epoch 343/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6411 - accuracy: 0.7382 - val_loss: 0.5524 - val_accuracy: 0.8446\n",
            "Epoch 344/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6247 - accuracy: 0.7584 - val_loss: 0.5315 - val_accuracy: 0.8581\n",
            "Epoch 345/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6755 - accuracy: 0.7196 - val_loss: 0.5516 - val_accuracy: 0.8378\n",
            "Epoch 346/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6621 - accuracy: 0.7551 - val_loss: 0.5839 - val_accuracy: 0.8108\n",
            "Epoch 347/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6497 - accuracy: 0.7432 - val_loss: 0.5572 - val_accuracy: 0.8446\n",
            "Epoch 348/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6567 - accuracy: 0.7382 - val_loss: 0.5516 - val_accuracy: 0.8108\n",
            "Epoch 349/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6382 - accuracy: 0.7213 - val_loss: 0.5514 - val_accuracy: 0.8243\n",
            "Epoch 350/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6297 - accuracy: 0.7348 - val_loss: 0.5342 - val_accuracy: 0.8446\n",
            "Epoch 351/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6549 - accuracy: 0.7314 - val_loss: 0.5689 - val_accuracy: 0.8311\n",
            "Epoch 352/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7060 - accuracy: 0.7264 - val_loss: 0.5645 - val_accuracy: 0.8243\n",
            "Epoch 353/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6728 - accuracy: 0.7247 - val_loss: 0.5540 - val_accuracy: 0.8243\n",
            "Epoch 354/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6289 - accuracy: 0.7399 - val_loss: 0.5466 - val_accuracy: 0.8243\n",
            "Epoch 355/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6386 - accuracy: 0.7314 - val_loss: 0.5544 - val_accuracy: 0.8311\n",
            "Epoch 356/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6375 - accuracy: 0.7416 - val_loss: 0.5621 - val_accuracy: 0.8311\n",
            "Epoch 357/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6605 - accuracy: 0.7517 - val_loss: 0.5509 - val_accuracy: 0.8446\n",
            "Epoch 358/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6572 - accuracy: 0.7551 - val_loss: 0.5530 - val_accuracy: 0.8243\n",
            "Epoch 359/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6667 - accuracy: 0.7483 - val_loss: 0.5537 - val_accuracy: 0.8243\n",
            "Epoch 360/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6559 - accuracy: 0.7348 - val_loss: 0.5600 - val_accuracy: 0.8378\n",
            "Epoch 361/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6424 - accuracy: 0.7500 - val_loss: 0.5663 - val_accuracy: 0.8041\n",
            "Epoch 362/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6367 - accuracy: 0.7399 - val_loss: 0.5545 - val_accuracy: 0.8108\n",
            "Epoch 363/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6389 - accuracy: 0.7416 - val_loss: 0.5455 - val_accuracy: 0.8243\n",
            "Epoch 364/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6611 - accuracy: 0.7449 - val_loss: 0.5311 - val_accuracy: 0.8446\n",
            "Epoch 365/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6382 - accuracy: 0.7601 - val_loss: 0.5531 - val_accuracy: 0.8378\n",
            "Epoch 366/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6453 - accuracy: 0.7601 - val_loss: 0.5434 - val_accuracy: 0.8514\n",
            "Epoch 367/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6517 - accuracy: 0.7517 - val_loss: 0.5531 - val_accuracy: 0.8378\n",
            "Epoch 368/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6621 - accuracy: 0.7280 - val_loss: 0.5518 - val_accuracy: 0.8311\n",
            "Epoch 369/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6459 - accuracy: 0.7399 - val_loss: 0.5369 - val_accuracy: 0.8581\n",
            "Epoch 370/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6454 - accuracy: 0.7382 - val_loss: 0.5663 - val_accuracy: 0.8243\n",
            "Epoch 371/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6090 - accuracy: 0.7601 - val_loss: 0.5331 - val_accuracy: 0.8243\n",
            "Epoch 372/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6372 - accuracy: 0.7500 - val_loss: 0.5281 - val_accuracy: 0.8041\n",
            "Epoch 373/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6290 - accuracy: 0.7500 - val_loss: 0.5250 - val_accuracy: 0.8176\n",
            "Epoch 374/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6448 - accuracy: 0.7331 - val_loss: 0.5443 - val_accuracy: 0.8378\n",
            "Epoch 375/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6342 - accuracy: 0.7399 - val_loss: 0.5513 - val_accuracy: 0.8108\n",
            "Epoch 376/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6438 - accuracy: 0.7483 - val_loss: 0.5371 - val_accuracy: 0.8311\n",
            "Epoch 377/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6038 - accuracy: 0.7348 - val_loss: 0.5476 - val_accuracy: 0.8311\n",
            "Epoch 378/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6105 - accuracy: 0.7635 - val_loss: 0.5588 - val_accuracy: 0.7973\n",
            "Epoch 379/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6188 - accuracy: 0.7652 - val_loss: 0.5487 - val_accuracy: 0.8108\n",
            "Epoch 380/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6664 - accuracy: 0.7517 - val_loss: 0.5885 - val_accuracy: 0.8243\n",
            "Epoch 381/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6547 - accuracy: 0.7297 - val_loss: 0.5547 - val_accuracy: 0.8311\n",
            "Epoch 382/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6655 - accuracy: 0.7145 - val_loss: 0.5694 - val_accuracy: 0.7973\n",
            "Epoch 383/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6410 - accuracy: 0.7365 - val_loss: 0.5421 - val_accuracy: 0.8176\n",
            "Epoch 384/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6383 - accuracy: 0.7500 - val_loss: 0.5288 - val_accuracy: 0.8581\n",
            "Epoch 385/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6549 - accuracy: 0.7230 - val_loss: 0.5583 - val_accuracy: 0.8378\n",
            "Epoch 386/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6729 - accuracy: 0.7365 - val_loss: 0.5607 - val_accuracy: 0.8243\n",
            "Epoch 387/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6467 - accuracy: 0.7348 - val_loss: 0.5393 - val_accuracy: 0.8378\n",
            "Epoch 388/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6644 - accuracy: 0.7517 - val_loss: 0.5540 - val_accuracy: 0.8446\n",
            "Epoch 389/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6540 - accuracy: 0.7348 - val_loss: 0.5488 - val_accuracy: 0.8514\n",
            "Epoch 390/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6312 - accuracy: 0.7534 - val_loss: 0.5391 - val_accuracy: 0.8176\n",
            "Epoch 391/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6384 - accuracy: 0.7449 - val_loss: 0.5669 - val_accuracy: 0.8243\n",
            "Epoch 392/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6244 - accuracy: 0.7753 - val_loss: 0.5391 - val_accuracy: 0.8514\n",
            "Epoch 393/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6615 - accuracy: 0.7382 - val_loss: 0.5443 - val_accuracy: 0.8378\n",
            "Epoch 394/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6428 - accuracy: 0.7584 - val_loss: 0.5426 - val_accuracy: 0.8378\n",
            "Epoch 395/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6465 - accuracy: 0.7534 - val_loss: 0.5427 - val_accuracy: 0.8311\n",
            "Epoch 396/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6292 - accuracy: 0.7584 - val_loss: 0.5264 - val_accuracy: 0.8514\n",
            "Epoch 397/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6254 - accuracy: 0.7753 - val_loss: 0.5512 - val_accuracy: 0.8446\n",
            "Epoch 398/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6319 - accuracy: 0.7635 - val_loss: 0.5383 - val_accuracy: 0.8311\n",
            "Epoch 399/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6425 - accuracy: 0.7483 - val_loss: 0.5636 - val_accuracy: 0.8176\n",
            "Epoch 400/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6747 - accuracy: 0.7703 - val_loss: 0.5679 - val_accuracy: 0.8311\n",
            "Epoch 401/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6299 - accuracy: 0.7635 - val_loss: 0.5200 - val_accuracy: 0.8581\n",
            "Epoch 402/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6148 - accuracy: 0.7686 - val_loss: 0.5469 - val_accuracy: 0.8514\n",
            "Epoch 403/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6728 - accuracy: 0.7449 - val_loss: 0.5363 - val_accuracy: 0.8649\n",
            "Epoch 404/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6018 - accuracy: 0.7770 - val_loss: 0.5486 - val_accuracy: 0.8108\n",
            "Epoch 405/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5960 - accuracy: 0.7905 - val_loss: 0.5355 - val_accuracy: 0.8446\n",
            "Epoch 406/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6956 - accuracy: 0.7095 - val_loss: 0.5673 - val_accuracy: 0.8446\n",
            "Epoch 407/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6621 - accuracy: 0.7264 - val_loss: 0.5548 - val_accuracy: 0.8378\n",
            "Epoch 408/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6312 - accuracy: 0.7584 - val_loss: 0.5362 - val_accuracy: 0.8378\n",
            "Epoch 409/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6054 - accuracy: 0.7669 - val_loss: 0.5416 - val_accuracy: 0.8378\n",
            "Epoch 410/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6596 - accuracy: 0.7399 - val_loss: 0.5468 - val_accuracy: 0.8378\n",
            "Epoch 411/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6243 - accuracy: 0.7804 - val_loss: 0.5422 - val_accuracy: 0.8311\n",
            "Epoch 412/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6159 - accuracy: 0.7686 - val_loss: 0.5515 - val_accuracy: 0.8378\n",
            "Epoch 413/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6304 - accuracy: 0.7432 - val_loss: 0.5242 - val_accuracy: 0.8649\n",
            "Epoch 414/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6750 - accuracy: 0.7348 - val_loss: 0.5544 - val_accuracy: 0.8784\n",
            "Epoch 415/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6216 - accuracy: 0.7753 - val_loss: 0.5198 - val_accuracy: 0.8446\n",
            "Epoch 416/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5996 - accuracy: 0.7770 - val_loss: 0.5492 - val_accuracy: 0.8514\n",
            "Epoch 417/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6618 - accuracy: 0.7196 - val_loss: 0.5744 - val_accuracy: 0.8378\n",
            "Epoch 418/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6548 - accuracy: 0.7601 - val_loss: 0.5502 - val_accuracy: 0.8176\n",
            "Epoch 419/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6643 - accuracy: 0.7399 - val_loss: 0.5643 - val_accuracy: 0.8176\n",
            "Epoch 420/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6205 - accuracy: 0.7601 - val_loss: 0.5362 - val_accuracy: 0.8378\n",
            "Epoch 421/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6765 - accuracy: 0.7280 - val_loss: 0.5983 - val_accuracy: 0.8446\n",
            "Epoch 422/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6518 - accuracy: 0.7449 - val_loss: 0.5645 - val_accuracy: 0.8514\n",
            "Epoch 423/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6652 - accuracy: 0.7584 - val_loss: 0.5650 - val_accuracy: 0.8378\n",
            "Epoch 424/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6270 - accuracy: 0.7703 - val_loss: 0.5450 - val_accuracy: 0.8581\n",
            "Epoch 425/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6796 - accuracy: 0.7280 - val_loss: 0.5636 - val_accuracy: 0.8446\n",
            "Epoch 426/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6280 - accuracy: 0.7382 - val_loss: 0.5388 - val_accuracy: 0.8108\n",
            "Epoch 427/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6200 - accuracy: 0.7720 - val_loss: 0.5208 - val_accuracy: 0.8716\n",
            "Epoch 428/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6378 - accuracy: 0.7618 - val_loss: 0.5614 - val_accuracy: 0.8514\n",
            "Epoch 429/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6767 - accuracy: 0.7264 - val_loss: 0.5681 - val_accuracy: 0.8446\n",
            "Epoch 430/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6544 - accuracy: 0.7635 - val_loss: 0.5499 - val_accuracy: 0.8243\n",
            "Epoch 431/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6512 - accuracy: 0.7720 - val_loss: 0.5498 - val_accuracy: 0.8446\n",
            "Epoch 432/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6219 - accuracy: 0.7635 - val_loss: 0.5399 - val_accuracy: 0.8108\n",
            "Epoch 433/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5963 - accuracy: 0.7736 - val_loss: 0.5353 - val_accuracy: 0.8311\n",
            "Epoch 434/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6271 - accuracy: 0.7466 - val_loss: 0.5325 - val_accuracy: 0.8176\n",
            "Epoch 435/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6902 - accuracy: 0.7264 - val_loss: 0.5630 - val_accuracy: 0.8243\n",
            "Epoch 436/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6452 - accuracy: 0.7449 - val_loss: 0.5463 - val_accuracy: 0.8446\n",
            "Epoch 437/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6220 - accuracy: 0.7534 - val_loss: 0.5469 - val_accuracy: 0.8311\n",
            "Epoch 438/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6433 - accuracy: 0.7399 - val_loss: 0.5398 - val_accuracy: 0.8581\n",
            "Epoch 439/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6642 - accuracy: 0.7416 - val_loss: 0.5482 - val_accuracy: 0.8716\n",
            "Epoch 440/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6421 - accuracy: 0.7568 - val_loss: 0.5320 - val_accuracy: 0.8378\n",
            "Epoch 441/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6343 - accuracy: 0.7601 - val_loss: 0.5268 - val_accuracy: 0.8108\n",
            "Epoch 442/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6909 - accuracy: 0.7432 - val_loss: 0.5557 - val_accuracy: 0.8108\n",
            "Epoch 443/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6450 - accuracy: 0.7399 - val_loss: 0.5391 - val_accuracy: 0.8378\n",
            "Epoch 444/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6110 - accuracy: 0.7466 - val_loss: 0.5515 - val_accuracy: 0.8311\n",
            "Epoch 445/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6411 - accuracy: 0.7432 - val_loss: 0.5637 - val_accuracy: 0.8311\n",
            "Epoch 446/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6428 - accuracy: 0.7483 - val_loss: 0.5788 - val_accuracy: 0.8378\n",
            "Epoch 447/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6392 - accuracy: 0.7551 - val_loss: 0.5842 - val_accuracy: 0.8311\n",
            "Epoch 448/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6936 - accuracy: 0.7230 - val_loss: 0.5780 - val_accuracy: 0.8041\n",
            "Epoch 449/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6653 - accuracy: 0.7331 - val_loss: 0.5439 - val_accuracy: 0.8311\n",
            "Epoch 450/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6440 - accuracy: 0.7314 - val_loss: 0.5285 - val_accuracy: 0.8243\n",
            "Epoch 451/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6648 - accuracy: 0.7280 - val_loss: 0.5656 - val_accuracy: 0.8378\n",
            "Epoch 452/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6540 - accuracy: 0.7517 - val_loss: 0.5427 - val_accuracy: 0.8446\n",
            "Epoch 453/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6435 - accuracy: 0.7247 - val_loss: 0.5364 - val_accuracy: 0.8446\n",
            "Epoch 454/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6684 - accuracy: 0.7584 - val_loss: 0.5621 - val_accuracy: 0.8243\n",
            "Epoch 455/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6712 - accuracy: 0.7314 - val_loss: 0.5731 - val_accuracy: 0.7973\n",
            "Epoch 456/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6437 - accuracy: 0.7382 - val_loss: 0.5631 - val_accuracy: 0.7703\n",
            "Epoch 457/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6270 - accuracy: 0.7753 - val_loss: 0.5352 - val_accuracy: 0.8041\n",
            "Epoch 458/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6336 - accuracy: 0.7551 - val_loss: 0.5479 - val_accuracy: 0.8378\n",
            "Epoch 459/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6593 - accuracy: 0.7517 - val_loss: 0.5409 - val_accuracy: 0.8446\n",
            "Epoch 460/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6712 - accuracy: 0.7466 - val_loss: 0.5387 - val_accuracy: 0.8649\n",
            "Epoch 461/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6552 - accuracy: 0.7584 - val_loss: 0.5381 - val_accuracy: 0.8446\n",
            "Epoch 462/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6553 - accuracy: 0.7551 - val_loss: 0.5560 - val_accuracy: 0.8446\n",
            "Epoch 463/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6705 - accuracy: 0.7264 - val_loss: 0.5432 - val_accuracy: 0.8378\n",
            "Epoch 464/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5875 - accuracy: 0.7872 - val_loss: 0.5204 - val_accuracy: 0.8378\n",
            "Epoch 465/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6759 - accuracy: 0.7635 - val_loss: 0.5715 - val_accuracy: 0.8108\n",
            "Epoch 466/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6525 - accuracy: 0.7618 - val_loss: 0.5654 - val_accuracy: 0.8243\n",
            "Epoch 467/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6527 - accuracy: 0.7534 - val_loss: 0.5431 - val_accuracy: 0.8311\n",
            "Epoch 468/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6236 - accuracy: 0.7652 - val_loss: 0.5711 - val_accuracy: 0.7973\n",
            "Epoch 469/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6651 - accuracy: 0.7331 - val_loss: 0.5583 - val_accuracy: 0.8311\n",
            "Epoch 470/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6734 - accuracy: 0.7551 - val_loss: 0.5612 - val_accuracy: 0.8176\n",
            "Epoch 471/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6481 - accuracy: 0.7534 - val_loss: 0.5579 - val_accuracy: 0.8311\n",
            "Epoch 472/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6361 - accuracy: 0.7500 - val_loss: 0.5476 - val_accuracy: 0.8378\n",
            "Epoch 473/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6399 - accuracy: 0.7399 - val_loss: 0.5599 - val_accuracy: 0.8378\n",
            "Epoch 474/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6788 - accuracy: 0.7331 - val_loss: 0.5632 - val_accuracy: 0.8108\n",
            "Epoch 475/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6309 - accuracy: 0.7652 - val_loss: 0.5455 - val_accuracy: 0.8176\n",
            "Epoch 476/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6175 - accuracy: 0.7618 - val_loss: 0.5301 - val_accuracy: 0.8581\n",
            "Epoch 477/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6604 - accuracy: 0.7348 - val_loss: 0.5198 - val_accuracy: 0.8716\n",
            "Epoch 478/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6425 - accuracy: 0.7449 - val_loss: 0.5585 - val_accuracy: 0.8446\n",
            "Epoch 479/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6761 - accuracy: 0.7264 - val_loss: 0.5493 - val_accuracy: 0.8514\n",
            "Epoch 480/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6583 - accuracy: 0.7568 - val_loss: 0.5316 - val_accuracy: 0.8378\n",
            "Epoch 481/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6221 - accuracy: 0.7551 - val_loss: 0.5683 - val_accuracy: 0.8176\n",
            "Epoch 482/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6409 - accuracy: 0.7584 - val_loss: 0.5767 - val_accuracy: 0.8108\n",
            "Epoch 483/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6926 - accuracy: 0.7230 - val_loss: 0.5809 - val_accuracy: 0.8243\n",
            "Epoch 484/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6511 - accuracy: 0.7348 - val_loss: 0.5490 - val_accuracy: 0.8446\n",
            "Epoch 485/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6166 - accuracy: 0.7753 - val_loss: 0.5503 - val_accuracy: 0.8108\n",
            "Epoch 486/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6337 - accuracy: 0.7568 - val_loss: 0.5641 - val_accuracy: 0.8243\n",
            "Epoch 487/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7182 - accuracy: 0.7095 - val_loss: 0.5885 - val_accuracy: 0.8176\n",
            "Epoch 488/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6547 - accuracy: 0.7568 - val_loss: 0.5347 - val_accuracy: 0.8378\n",
            "Epoch 489/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6373 - accuracy: 0.7703 - val_loss: 0.5460 - val_accuracy: 0.8243\n",
            "Epoch 490/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5911 - accuracy: 0.7855 - val_loss: 0.5066 - val_accuracy: 0.8378\n",
            "Epoch 491/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6247 - accuracy: 0.7500 - val_loss: 0.5300 - val_accuracy: 0.8243\n",
            "Epoch 492/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6158 - accuracy: 0.7753 - val_loss: 0.5003 - val_accuracy: 0.8311\n",
            "Epoch 493/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6339 - accuracy: 0.7618 - val_loss: 0.5232 - val_accuracy: 0.8446\n",
            "Epoch 494/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6048 - accuracy: 0.7753 - val_loss: 0.5359 - val_accuracy: 0.8311\n",
            "Epoch 495/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6445 - accuracy: 0.7432 - val_loss: 0.5417 - val_accuracy: 0.8311\n",
            "Epoch 496/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6584 - accuracy: 0.7314 - val_loss: 0.5673 - val_accuracy: 0.8514\n",
            "Epoch 497/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6602 - accuracy: 0.7348 - val_loss: 0.5656 - val_accuracy: 0.8378\n",
            "Epoch 498/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6289 - accuracy: 0.7618 - val_loss: 0.5315 - val_accuracy: 0.8311\n",
            "Epoch 499/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6614 - accuracy: 0.7382 - val_loss: 0.5443 - val_accuracy: 0.8514\n",
            "Epoch 500/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6212 - accuracy: 0.7618 - val_loss: 0.5652 - val_accuracy: 0.8311\n",
            "5/5 [==============================] - 0s 4ms/step - loss: 0.5652 - accuracy: 0.8311\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:62: DeprecationWarning: scipy.interp is deprecated and will be removed in SciPy 2.0.0, use numpy.interp instead\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/500\n",
            "74/74 [==============================] - 1s 10ms/step - loss: 1.8336 - accuracy: 0.4814 - val_loss: 1.3490 - val_accuracy: 0.6757\n",
            "Epoch 2/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 1.3807 - accuracy: 0.5389 - val_loss: 1.2811 - val_accuracy: 0.6824\n",
            "Epoch 3/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 1.2860 - accuracy: 0.5405 - val_loss: 1.2189 - val_accuracy: 0.6824\n",
            "Epoch 4/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 1.1949 - accuracy: 0.4983 - val_loss: 1.1333 - val_accuracy: 0.6554\n",
            "Epoch 5/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 1.1045 - accuracy: 0.5304 - val_loss: 1.0562 - val_accuracy: 0.6824\n",
            "Epoch 6/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 1.0320 - accuracy: 0.5321 - val_loss: 0.9886 - val_accuracy: 0.5743\n",
            "Epoch 7/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.9731 - accuracy: 0.5794 - val_loss: 0.9364 - val_accuracy: 0.6892\n",
            "Epoch 8/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.9460 - accuracy: 0.5557 - val_loss: 0.9072 - val_accuracy: 0.7703\n",
            "Epoch 9/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8965 - accuracy: 0.5845 - val_loss: 0.8522 - val_accuracy: 0.7095\n",
            "Epoch 10/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8725 - accuracy: 0.5929 - val_loss: 0.8189 - val_accuracy: 0.7095\n",
            "Epoch 11/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8596 - accuracy: 0.5811 - val_loss: 0.8167 - val_accuracy: 0.6892\n",
            "Epoch 12/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8283 - accuracy: 0.6301 - val_loss: 0.8081 - val_accuracy: 0.6622\n",
            "Epoch 13/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8588 - accuracy: 0.5895 - val_loss: 0.7710 - val_accuracy: 0.7838\n",
            "Epoch 14/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.8246 - accuracy: 0.6368 - val_loss: 0.7670 - val_accuracy: 0.7297\n",
            "Epoch 15/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8196 - accuracy: 0.6537 - val_loss: 0.7658 - val_accuracy: 0.7838\n",
            "Epoch 16/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.8720 - accuracy: 0.6030 - val_loss: 0.7977 - val_accuracy: 0.7635\n",
            "Epoch 17/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8427 - accuracy: 0.6149 - val_loss: 0.7663 - val_accuracy: 0.7905\n",
            "Epoch 18/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.8631 - accuracy: 0.6351 - val_loss: 0.7789 - val_accuracy: 0.7365\n",
            "Epoch 19/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8415 - accuracy: 0.6267 - val_loss: 0.7841 - val_accuracy: 0.7297\n",
            "Epoch 20/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.8136 - accuracy: 0.6318 - val_loss: 0.7906 - val_accuracy: 0.6554\n",
            "Epoch 21/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8010 - accuracy: 0.6267 - val_loss: 0.7443 - val_accuracy: 0.7770\n",
            "Epoch 22/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7902 - accuracy: 0.6537 - val_loss: 0.7045 - val_accuracy: 0.8041\n",
            "Epoch 23/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7975 - accuracy: 0.6841 - val_loss: 0.7024 - val_accuracy: 0.7838\n",
            "Epoch 24/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.8232 - accuracy: 0.6520 - val_loss: 0.7471 - val_accuracy: 0.7838\n",
            "Epoch 25/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.8138 - accuracy: 0.6723 - val_loss: 0.7407 - val_accuracy: 0.8041\n",
            "Epoch 26/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7865 - accuracy: 0.6807 - val_loss: 0.7019 - val_accuracy: 0.7973\n",
            "Epoch 27/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7911 - accuracy: 0.6723 - val_loss: 0.7063 - val_accuracy: 0.7905\n",
            "Epoch 28/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8033 - accuracy: 0.6757 - val_loss: 0.6933 - val_accuracy: 0.8108\n",
            "Epoch 29/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7868 - accuracy: 0.6757 - val_loss: 0.6840 - val_accuracy: 0.7703\n",
            "Epoch 30/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7710 - accuracy: 0.6740 - val_loss: 0.6871 - val_accuracy: 0.7973\n",
            "Epoch 31/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7826 - accuracy: 0.6740 - val_loss: 0.6646 - val_accuracy: 0.8108\n",
            "Epoch 32/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8013 - accuracy: 0.6757 - val_loss: 0.6955 - val_accuracy: 0.8243\n",
            "Epoch 33/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7873 - accuracy: 0.6909 - val_loss: 0.6908 - val_accuracy: 0.8041\n",
            "Epoch 34/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7896 - accuracy: 0.6774 - val_loss: 0.6697 - val_accuracy: 0.8311\n",
            "Epoch 35/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7708 - accuracy: 0.6824 - val_loss: 0.6662 - val_accuracy: 0.7973\n",
            "Epoch 36/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7890 - accuracy: 0.6622 - val_loss: 0.6391 - val_accuracy: 0.8446\n",
            "Epoch 37/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7422 - accuracy: 0.7179 - val_loss: 0.6269 - val_accuracy: 0.8514\n",
            "Epoch 38/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7390 - accuracy: 0.6993 - val_loss: 0.5992 - val_accuracy: 0.8243\n",
            "Epoch 39/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7793 - accuracy: 0.6807 - val_loss: 0.6588 - val_accuracy: 0.8378\n",
            "Epoch 40/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7748 - accuracy: 0.6639 - val_loss: 0.6758 - val_accuracy: 0.8446\n",
            "Epoch 41/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7737 - accuracy: 0.6740 - val_loss: 0.6751 - val_accuracy: 0.8514\n",
            "Epoch 42/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7425 - accuracy: 0.6824 - val_loss: 0.6358 - val_accuracy: 0.8041\n",
            "Epoch 43/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7821 - accuracy: 0.6807 - val_loss: 0.6635 - val_accuracy: 0.7905\n",
            "Epoch 44/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7580 - accuracy: 0.6926 - val_loss: 0.6517 - val_accuracy: 0.8243\n",
            "Epoch 45/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7411 - accuracy: 0.7078 - val_loss: 0.6307 - val_accuracy: 0.8108\n",
            "Epoch 46/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7486 - accuracy: 0.7044 - val_loss: 0.6401 - val_accuracy: 0.8176\n",
            "Epoch 47/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7637 - accuracy: 0.6926 - val_loss: 0.6550 - val_accuracy: 0.8176\n",
            "Epoch 48/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7754 - accuracy: 0.6706 - val_loss: 0.6521 - val_accuracy: 0.8041\n",
            "Epoch 49/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7569 - accuracy: 0.6807 - val_loss: 0.6655 - val_accuracy: 0.8108\n",
            "Epoch 50/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7540 - accuracy: 0.6689 - val_loss: 0.6293 - val_accuracy: 0.7905\n",
            "Epoch 51/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7391 - accuracy: 0.6774 - val_loss: 0.6339 - val_accuracy: 0.7905\n",
            "Epoch 52/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7076 - accuracy: 0.7247 - val_loss: 0.6192 - val_accuracy: 0.8108\n",
            "Epoch 53/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7643 - accuracy: 0.6639 - val_loss: 0.6396 - val_accuracy: 0.8378\n",
            "Epoch 54/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7586 - accuracy: 0.6841 - val_loss: 0.6234 - val_accuracy: 0.8108\n",
            "Epoch 55/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7160 - accuracy: 0.7078 - val_loss: 0.6054 - val_accuracy: 0.8176\n",
            "Epoch 56/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7476 - accuracy: 0.6943 - val_loss: 0.6384 - val_accuracy: 0.8108\n",
            "Epoch 57/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7623 - accuracy: 0.6723 - val_loss: 0.6305 - val_accuracy: 0.7905\n",
            "Epoch 58/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7537 - accuracy: 0.6774 - val_loss: 0.6296 - val_accuracy: 0.8108\n",
            "Epoch 59/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7674 - accuracy: 0.6655 - val_loss: 0.6522 - val_accuracy: 0.7703\n",
            "Epoch 60/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7569 - accuracy: 0.6689 - val_loss: 0.6484 - val_accuracy: 0.7770\n",
            "Epoch 61/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7610 - accuracy: 0.7027 - val_loss: 0.6658 - val_accuracy: 0.7432\n",
            "Epoch 62/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7824 - accuracy: 0.6875 - val_loss: 0.6353 - val_accuracy: 0.8108\n",
            "Epoch 63/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7331 - accuracy: 0.7095 - val_loss: 0.6092 - val_accuracy: 0.8108\n",
            "Epoch 64/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7664 - accuracy: 0.6723 - val_loss: 0.6196 - val_accuracy: 0.8311\n",
            "Epoch 65/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7249 - accuracy: 0.7196 - val_loss: 0.6207 - val_accuracy: 0.7973\n",
            "Epoch 66/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7298 - accuracy: 0.7145 - val_loss: 0.6311 - val_accuracy: 0.7838\n",
            "Epoch 67/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7373 - accuracy: 0.7078 - val_loss: 0.6302 - val_accuracy: 0.8311\n",
            "Epoch 68/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7678 - accuracy: 0.6858 - val_loss: 0.6572 - val_accuracy: 0.8243\n",
            "Epoch 69/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7537 - accuracy: 0.7010 - val_loss: 0.6158 - val_accuracy: 0.8378\n",
            "Epoch 70/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7438 - accuracy: 0.6909 - val_loss: 0.6326 - val_accuracy: 0.8108\n",
            "Epoch 71/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7097 - accuracy: 0.7162 - val_loss: 0.6137 - val_accuracy: 0.8311\n",
            "Epoch 72/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7282 - accuracy: 0.7111 - val_loss: 0.6167 - val_accuracy: 0.7973\n",
            "Epoch 73/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7506 - accuracy: 0.6723 - val_loss: 0.6109 - val_accuracy: 0.8378\n",
            "Epoch 74/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7639 - accuracy: 0.6791 - val_loss: 0.6504 - val_accuracy: 0.8378\n",
            "Epoch 75/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7376 - accuracy: 0.7078 - val_loss: 0.6407 - val_accuracy: 0.8243\n",
            "Epoch 76/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7046 - accuracy: 0.7483 - val_loss: 0.6138 - val_accuracy: 0.8311\n",
            "Epoch 77/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7284 - accuracy: 0.6926 - val_loss: 0.6048 - val_accuracy: 0.8108\n",
            "Epoch 78/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7322 - accuracy: 0.7010 - val_loss: 0.5971 - val_accuracy: 0.8176\n",
            "Epoch 79/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7348 - accuracy: 0.6892 - val_loss: 0.6111 - val_accuracy: 0.8176\n",
            "Epoch 80/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7125 - accuracy: 0.7213 - val_loss: 0.5827 - val_accuracy: 0.8378\n",
            "Epoch 81/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7237 - accuracy: 0.7010 - val_loss: 0.6123 - val_accuracy: 0.8311\n",
            "Epoch 82/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7134 - accuracy: 0.7196 - val_loss: 0.5736 - val_accuracy: 0.8311\n",
            "Epoch 83/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7432 - accuracy: 0.7061 - val_loss: 0.6336 - val_accuracy: 0.8243\n",
            "Epoch 84/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7296 - accuracy: 0.7095 - val_loss: 0.6412 - val_accuracy: 0.8041\n",
            "Epoch 85/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7289 - accuracy: 0.7027 - val_loss: 0.6093 - val_accuracy: 0.8176\n",
            "Epoch 86/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7152 - accuracy: 0.7365 - val_loss: 0.6049 - val_accuracy: 0.8108\n",
            "Epoch 87/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7357 - accuracy: 0.6909 - val_loss: 0.6173 - val_accuracy: 0.8581\n",
            "Epoch 88/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7018 - accuracy: 0.7179 - val_loss: 0.6075 - val_accuracy: 0.8041\n",
            "Epoch 89/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7173 - accuracy: 0.7145 - val_loss: 0.6124 - val_accuracy: 0.8041\n",
            "Epoch 90/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7059 - accuracy: 0.7179 - val_loss: 0.5837 - val_accuracy: 0.8108\n",
            "Epoch 91/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6882 - accuracy: 0.7348 - val_loss: 0.5799 - val_accuracy: 0.8176\n",
            "Epoch 92/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6988 - accuracy: 0.7297 - val_loss: 0.5946 - val_accuracy: 0.8108\n",
            "Epoch 93/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7010 - accuracy: 0.6993 - val_loss: 0.5863 - val_accuracy: 0.8311\n",
            "Epoch 94/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7120 - accuracy: 0.7078 - val_loss: 0.6001 - val_accuracy: 0.8311\n",
            "Epoch 95/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7297 - accuracy: 0.7095 - val_loss: 0.5843 - val_accuracy: 0.8378\n",
            "Epoch 96/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7471 - accuracy: 0.7027 - val_loss: 0.6045 - val_accuracy: 0.8041\n",
            "Epoch 97/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6941 - accuracy: 0.7365 - val_loss: 0.5745 - val_accuracy: 0.8176\n",
            "Epoch 98/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6741 - accuracy: 0.7280 - val_loss: 0.5784 - val_accuracy: 0.8378\n",
            "Epoch 99/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7376 - accuracy: 0.7247 - val_loss: 0.5973 - val_accuracy: 0.8311\n",
            "Epoch 100/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6876 - accuracy: 0.7247 - val_loss: 0.5848 - val_accuracy: 0.8311\n",
            "Epoch 101/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7157 - accuracy: 0.7145 - val_loss: 0.6053 - val_accuracy: 0.7973\n",
            "Epoch 102/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7210 - accuracy: 0.7264 - val_loss: 0.5907 - val_accuracy: 0.8176\n",
            "Epoch 103/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7295 - accuracy: 0.7128 - val_loss: 0.6156 - val_accuracy: 0.8041\n",
            "Epoch 104/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6842 - accuracy: 0.7297 - val_loss: 0.5854 - val_accuracy: 0.8041\n",
            "Epoch 105/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6841 - accuracy: 0.7247 - val_loss: 0.5785 - val_accuracy: 0.8041\n",
            "Epoch 106/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7244 - accuracy: 0.7297 - val_loss: 0.6096 - val_accuracy: 0.8108\n",
            "Epoch 107/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6565 - accuracy: 0.7551 - val_loss: 0.5576 - val_accuracy: 0.8176\n",
            "Epoch 108/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6758 - accuracy: 0.7365 - val_loss: 0.5848 - val_accuracy: 0.8311\n",
            "Epoch 109/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7252 - accuracy: 0.6959 - val_loss: 0.5963 - val_accuracy: 0.8378\n",
            "Epoch 110/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7232 - accuracy: 0.7179 - val_loss: 0.6030 - val_accuracy: 0.8311\n",
            "Epoch 111/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7085 - accuracy: 0.7027 - val_loss: 0.5959 - val_accuracy: 0.8311\n",
            "Epoch 112/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6660 - accuracy: 0.7348 - val_loss: 0.5806 - val_accuracy: 0.8176\n",
            "Epoch 113/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6780 - accuracy: 0.7230 - val_loss: 0.5677 - val_accuracy: 0.8446\n",
            "Epoch 114/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7175 - accuracy: 0.7111 - val_loss: 0.6015 - val_accuracy: 0.8176\n",
            "Epoch 115/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7059 - accuracy: 0.7196 - val_loss: 0.6046 - val_accuracy: 0.7703\n",
            "Epoch 116/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6834 - accuracy: 0.7365 - val_loss: 0.5635 - val_accuracy: 0.8041\n",
            "Epoch 117/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6913 - accuracy: 0.7145 - val_loss: 0.5532 - val_accuracy: 0.8176\n",
            "Epoch 118/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6687 - accuracy: 0.7365 - val_loss: 0.5854 - val_accuracy: 0.8176\n",
            "Epoch 119/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6991 - accuracy: 0.7162 - val_loss: 0.5783 - val_accuracy: 0.8311\n",
            "Epoch 120/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7479 - accuracy: 0.6841 - val_loss: 0.5959 - val_accuracy: 0.8108\n",
            "Epoch 121/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7032 - accuracy: 0.7483 - val_loss: 0.5725 - val_accuracy: 0.8311\n",
            "Epoch 122/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6801 - accuracy: 0.7399 - val_loss: 0.6094 - val_accuracy: 0.8243\n",
            "Epoch 123/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6898 - accuracy: 0.7331 - val_loss: 0.5787 - val_accuracy: 0.8108\n",
            "Epoch 124/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6882 - accuracy: 0.7517 - val_loss: 0.5557 - val_accuracy: 0.8311\n",
            "Epoch 125/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6917 - accuracy: 0.7416 - val_loss: 0.5668 - val_accuracy: 0.8311\n",
            "Epoch 126/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6586 - accuracy: 0.7416 - val_loss: 0.5415 - val_accuracy: 0.8378\n",
            "Epoch 127/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6842 - accuracy: 0.7348 - val_loss: 0.5758 - val_accuracy: 0.8243\n",
            "Epoch 128/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6561 - accuracy: 0.7669 - val_loss: 0.5907 - val_accuracy: 0.8176\n",
            "Epoch 129/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7094 - accuracy: 0.6959 - val_loss: 0.6017 - val_accuracy: 0.8446\n",
            "Epoch 130/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6931 - accuracy: 0.7584 - val_loss: 0.6002 - val_accuracy: 0.8311\n",
            "Epoch 131/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6811 - accuracy: 0.7213 - val_loss: 0.5562 - val_accuracy: 0.8311\n",
            "Epoch 132/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7190 - accuracy: 0.7128 - val_loss: 0.5786 - val_accuracy: 0.8446\n",
            "Epoch 133/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6939 - accuracy: 0.7264 - val_loss: 0.5910 - val_accuracy: 0.8176\n",
            "Epoch 134/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6839 - accuracy: 0.7078 - val_loss: 0.5736 - val_accuracy: 0.8243\n",
            "Epoch 135/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6879 - accuracy: 0.7314 - val_loss: 0.5767 - val_accuracy: 0.8446\n",
            "Epoch 136/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6575 - accuracy: 0.7348 - val_loss: 0.5535 - val_accuracy: 0.8311\n",
            "Epoch 137/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6682 - accuracy: 0.7297 - val_loss: 0.5467 - val_accuracy: 0.7973\n",
            "Epoch 138/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6877 - accuracy: 0.7213 - val_loss: 0.5680 - val_accuracy: 0.8581\n",
            "Epoch 139/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6489 - accuracy: 0.7399 - val_loss: 0.5288 - val_accuracy: 0.8243\n",
            "Epoch 140/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7029 - accuracy: 0.7027 - val_loss: 0.5587 - val_accuracy: 0.8378\n",
            "Epoch 141/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6729 - accuracy: 0.7297 - val_loss: 0.5649 - val_accuracy: 0.8446\n",
            "Epoch 142/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6720 - accuracy: 0.7635 - val_loss: 0.5624 - val_accuracy: 0.8311\n",
            "Epoch 143/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6893 - accuracy: 0.7264 - val_loss: 0.5530 - val_accuracy: 0.8446\n",
            "Epoch 144/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6679 - accuracy: 0.7331 - val_loss: 0.5573 - val_accuracy: 0.8311\n",
            "Epoch 145/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7178 - accuracy: 0.7044 - val_loss: 0.5883 - val_accuracy: 0.8243\n",
            "Epoch 146/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6793 - accuracy: 0.7348 - val_loss: 0.5595 - val_accuracy: 0.8243\n",
            "Epoch 147/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6765 - accuracy: 0.7365 - val_loss: 0.5461 - val_accuracy: 0.8446\n",
            "Epoch 148/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6798 - accuracy: 0.7264 - val_loss: 0.5463 - val_accuracy: 0.8311\n",
            "Epoch 149/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6529 - accuracy: 0.7416 - val_loss: 0.5447 - val_accuracy: 0.8446\n",
            "Epoch 150/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6806 - accuracy: 0.7280 - val_loss: 0.5666 - val_accuracy: 0.8176\n",
            "Epoch 151/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6580 - accuracy: 0.7534 - val_loss: 0.5619 - val_accuracy: 0.8378\n",
            "Epoch 152/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6681 - accuracy: 0.7432 - val_loss: 0.5594 - val_accuracy: 0.8378\n",
            "Epoch 153/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7087 - accuracy: 0.7128 - val_loss: 0.5641 - val_accuracy: 0.8311\n",
            "Epoch 154/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6505 - accuracy: 0.7365 - val_loss: 0.5341 - val_accuracy: 0.8176\n",
            "Epoch 155/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6832 - accuracy: 0.7280 - val_loss: 0.5698 - val_accuracy: 0.8243\n",
            "Epoch 156/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6501 - accuracy: 0.7551 - val_loss: 0.5287 - val_accuracy: 0.8311\n",
            "Epoch 157/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6560 - accuracy: 0.7449 - val_loss: 0.5326 - val_accuracy: 0.8378\n",
            "Epoch 158/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6583 - accuracy: 0.7551 - val_loss: 0.5283 - val_accuracy: 0.8649\n",
            "Epoch 159/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6593 - accuracy: 0.7382 - val_loss: 0.5365 - val_accuracy: 0.8581\n",
            "Epoch 160/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6083 - accuracy: 0.7534 - val_loss: 0.5674 - val_accuracy: 0.7905\n",
            "Epoch 161/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6287 - accuracy: 0.7534 - val_loss: 0.5497 - val_accuracy: 0.7973\n",
            "Epoch 162/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7218 - accuracy: 0.6993 - val_loss: 0.6048 - val_accuracy: 0.8514\n",
            "Epoch 163/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6732 - accuracy: 0.7382 - val_loss: 0.5539 - val_accuracy: 0.8378\n",
            "Epoch 164/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6256 - accuracy: 0.7416 - val_loss: 0.5463 - val_accuracy: 0.8176\n",
            "Epoch 165/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6935 - accuracy: 0.7044 - val_loss: 0.5712 - val_accuracy: 0.8041\n",
            "Epoch 166/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6733 - accuracy: 0.7111 - val_loss: 0.5389 - val_accuracy: 0.8581\n",
            "Epoch 167/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6688 - accuracy: 0.7213 - val_loss: 0.5295 - val_accuracy: 0.8446\n",
            "Epoch 168/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6845 - accuracy: 0.7247 - val_loss: 0.5633 - val_accuracy: 0.8108\n",
            "Epoch 169/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6788 - accuracy: 0.7247 - val_loss: 0.5548 - val_accuracy: 0.8446\n",
            "Epoch 170/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6245 - accuracy: 0.7703 - val_loss: 0.5347 - val_accuracy: 0.8243\n",
            "Epoch 171/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6537 - accuracy: 0.7162 - val_loss: 0.5178 - val_accuracy: 0.8446\n",
            "Epoch 172/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6590 - accuracy: 0.7297 - val_loss: 0.5332 - val_accuracy: 0.8378\n",
            "Epoch 173/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6591 - accuracy: 0.7416 - val_loss: 0.5398 - val_accuracy: 0.8243\n",
            "Epoch 174/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6464 - accuracy: 0.7297 - val_loss: 0.5392 - val_accuracy: 0.8378\n",
            "Epoch 175/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6274 - accuracy: 0.7348 - val_loss: 0.5369 - val_accuracy: 0.8176\n",
            "Epoch 176/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6626 - accuracy: 0.7365 - val_loss: 0.5449 - val_accuracy: 0.8446\n",
            "Epoch 177/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6367 - accuracy: 0.7280 - val_loss: 0.5186 - val_accuracy: 0.8446\n",
            "Epoch 178/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6600 - accuracy: 0.7247 - val_loss: 0.5298 - val_accuracy: 0.8243\n",
            "Epoch 179/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6582 - accuracy: 0.7314 - val_loss: 0.5062 - val_accuracy: 0.8446\n",
            "Epoch 180/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6169 - accuracy: 0.7635 - val_loss: 0.5012 - val_accuracy: 0.8514\n",
            "Epoch 181/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6457 - accuracy: 0.7314 - val_loss: 0.4861 - val_accuracy: 0.8446\n",
            "Epoch 182/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6100 - accuracy: 0.7466 - val_loss: 0.4837 - val_accuracy: 0.8581\n",
            "Epoch 183/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6416 - accuracy: 0.7568 - val_loss: 0.5377 - val_accuracy: 0.8041\n",
            "Epoch 184/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6321 - accuracy: 0.7534 - val_loss: 0.4975 - val_accuracy: 0.8446\n",
            "Epoch 185/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7060 - accuracy: 0.7179 - val_loss: 0.5418 - val_accuracy: 0.8514\n",
            "Epoch 186/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6366 - accuracy: 0.7179 - val_loss: 0.5077 - val_accuracy: 0.8378\n",
            "Epoch 187/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6559 - accuracy: 0.7230 - val_loss: 0.5421 - val_accuracy: 0.8041\n",
            "Epoch 188/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6475 - accuracy: 0.7213 - val_loss: 0.5121 - val_accuracy: 0.8311\n",
            "Epoch 189/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6027 - accuracy: 0.7703 - val_loss: 0.5068 - val_accuracy: 0.8378\n",
            "Epoch 190/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6890 - accuracy: 0.7314 - val_loss: 0.5255 - val_accuracy: 0.8514\n",
            "Epoch 191/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6263 - accuracy: 0.7297 - val_loss: 0.5002 - val_accuracy: 0.8514\n",
            "Epoch 192/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6504 - accuracy: 0.7348 - val_loss: 0.5313 - val_accuracy: 0.8446\n",
            "Epoch 193/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6724 - accuracy: 0.7466 - val_loss: 0.5385 - val_accuracy: 0.8514\n",
            "Epoch 194/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6595 - accuracy: 0.7432 - val_loss: 0.5594 - val_accuracy: 0.7973\n",
            "Epoch 195/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6129 - accuracy: 0.7618 - val_loss: 0.5298 - val_accuracy: 0.7973\n",
            "Epoch 196/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6519 - accuracy: 0.7500 - val_loss: 0.5272 - val_accuracy: 0.8446\n",
            "Epoch 197/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6445 - accuracy: 0.7534 - val_loss: 0.5313 - val_accuracy: 0.8176\n",
            "Epoch 198/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6402 - accuracy: 0.7449 - val_loss: 0.5275 - val_accuracy: 0.8243\n",
            "Epoch 199/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6422 - accuracy: 0.7297 - val_loss: 0.5113 - val_accuracy: 0.8243\n",
            "Epoch 200/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5985 - accuracy: 0.7382 - val_loss: 0.4880 - val_accuracy: 0.8378\n",
            "Epoch 201/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6045 - accuracy: 0.7601 - val_loss: 0.4959 - val_accuracy: 0.8243\n",
            "Epoch 202/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6643 - accuracy: 0.7314 - val_loss: 0.5450 - val_accuracy: 0.7973\n",
            "Epoch 203/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6697 - accuracy: 0.7128 - val_loss: 0.5782 - val_accuracy: 0.8243\n",
            "Epoch 204/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6457 - accuracy: 0.7230 - val_loss: 0.5191 - val_accuracy: 0.8311\n",
            "Epoch 205/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6602 - accuracy: 0.7145 - val_loss: 0.5367 - val_accuracy: 0.8311\n",
            "Epoch 206/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6576 - accuracy: 0.7230 - val_loss: 0.5299 - val_accuracy: 0.8446\n",
            "Epoch 207/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6372 - accuracy: 0.7703 - val_loss: 0.4817 - val_accuracy: 0.8378\n",
            "Epoch 208/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6492 - accuracy: 0.7331 - val_loss: 0.5100 - val_accuracy: 0.8243\n",
            "Epoch 209/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6224 - accuracy: 0.7432 - val_loss: 0.4874 - val_accuracy: 0.8311\n",
            "Epoch 210/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6323 - accuracy: 0.7500 - val_loss: 0.5073 - val_accuracy: 0.8378\n",
            "Epoch 211/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6595 - accuracy: 0.6993 - val_loss: 0.5138 - val_accuracy: 0.8176\n",
            "Epoch 212/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6159 - accuracy: 0.7534 - val_loss: 0.4911 - val_accuracy: 0.8446\n",
            "Epoch 213/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6479 - accuracy: 0.7128 - val_loss: 0.5185 - val_accuracy: 0.8108\n",
            "Epoch 214/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6537 - accuracy: 0.7179 - val_loss: 0.5141 - val_accuracy: 0.8446\n",
            "Epoch 215/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6294 - accuracy: 0.7568 - val_loss: 0.5054 - val_accuracy: 0.8446\n",
            "Epoch 216/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6361 - accuracy: 0.7382 - val_loss: 0.5239 - val_accuracy: 0.8311\n",
            "Epoch 217/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6502 - accuracy: 0.7331 - val_loss: 0.5072 - val_accuracy: 0.8311\n",
            "Epoch 218/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6095 - accuracy: 0.7551 - val_loss: 0.5058 - val_accuracy: 0.8176\n",
            "Epoch 219/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6291 - accuracy: 0.7568 - val_loss: 0.5241 - val_accuracy: 0.8243\n",
            "Epoch 220/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6447 - accuracy: 0.7416 - val_loss: 0.5127 - val_accuracy: 0.8311\n",
            "Epoch 221/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6285 - accuracy: 0.7584 - val_loss: 0.5180 - val_accuracy: 0.8311\n",
            "Epoch 222/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6334 - accuracy: 0.7331 - val_loss: 0.5104 - val_accuracy: 0.8243\n",
            "Epoch 223/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6280 - accuracy: 0.7432 - val_loss: 0.4999 - val_accuracy: 0.8446\n",
            "Epoch 224/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6374 - accuracy: 0.7297 - val_loss: 0.4991 - val_accuracy: 0.8108\n",
            "Epoch 225/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6209 - accuracy: 0.7382 - val_loss: 0.4935 - val_accuracy: 0.8311\n",
            "Epoch 226/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6347 - accuracy: 0.7382 - val_loss: 0.4938 - val_accuracy: 0.8378\n",
            "Epoch 227/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6354 - accuracy: 0.7432 - val_loss: 0.5132 - val_accuracy: 0.8243\n",
            "Epoch 228/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6140 - accuracy: 0.7618 - val_loss: 0.4962 - val_accuracy: 0.8243\n",
            "Epoch 229/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6207 - accuracy: 0.7449 - val_loss: 0.4951 - val_accuracy: 0.8446\n",
            "Epoch 230/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6161 - accuracy: 0.7669 - val_loss: 0.5133 - val_accuracy: 0.8176\n",
            "Epoch 231/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5974 - accuracy: 0.7618 - val_loss: 0.5168 - val_accuracy: 0.8311\n",
            "Epoch 232/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6215 - accuracy: 0.7500 - val_loss: 0.5062 - val_accuracy: 0.8108\n",
            "Epoch 233/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6357 - accuracy: 0.7247 - val_loss: 0.5039 - val_accuracy: 0.8176\n",
            "Epoch 234/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6165 - accuracy: 0.7534 - val_loss: 0.5055 - val_accuracy: 0.8243\n",
            "Epoch 235/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5712 - accuracy: 0.7872 - val_loss: 0.4990 - val_accuracy: 0.8311\n",
            "Epoch 236/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6022 - accuracy: 0.7635 - val_loss: 0.4972 - val_accuracy: 0.8041\n",
            "Epoch 237/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6287 - accuracy: 0.7449 - val_loss: 0.5106 - val_accuracy: 0.8108\n",
            "Epoch 238/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6267 - accuracy: 0.7348 - val_loss: 0.5256 - val_accuracy: 0.8108\n",
            "Epoch 239/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5999 - accuracy: 0.7432 - val_loss: 0.5107 - val_accuracy: 0.8176\n",
            "Epoch 240/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6099 - accuracy: 0.7517 - val_loss: 0.4948 - val_accuracy: 0.8176\n",
            "Epoch 241/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6576 - accuracy: 0.7399 - val_loss: 0.5225 - val_accuracy: 0.8243\n",
            "Epoch 242/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6288 - accuracy: 0.7280 - val_loss: 0.5046 - val_accuracy: 0.8176\n",
            "Epoch 243/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6079 - accuracy: 0.7483 - val_loss: 0.4962 - val_accuracy: 0.8243\n",
            "Epoch 244/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6181 - accuracy: 0.7568 - val_loss: 0.4919 - val_accuracy: 0.8446\n",
            "Epoch 245/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6242 - accuracy: 0.7432 - val_loss: 0.4902 - val_accuracy: 0.8176\n",
            "Epoch 246/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6244 - accuracy: 0.7720 - val_loss: 0.5144 - val_accuracy: 0.8446\n",
            "Epoch 247/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5898 - accuracy: 0.7787 - val_loss: 0.5146 - val_accuracy: 0.8041\n",
            "Epoch 248/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6210 - accuracy: 0.7686 - val_loss: 0.5199 - val_accuracy: 0.8243\n",
            "Epoch 249/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6062 - accuracy: 0.7449 - val_loss: 0.5091 - val_accuracy: 0.8108\n",
            "Epoch 250/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6201 - accuracy: 0.7753 - val_loss: 0.5264 - val_accuracy: 0.8176\n",
            "Epoch 251/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6582 - accuracy: 0.7331 - val_loss: 0.5306 - val_accuracy: 0.7973\n",
            "Epoch 252/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6406 - accuracy: 0.7432 - val_loss: 0.5200 - val_accuracy: 0.8243\n",
            "Epoch 253/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6466 - accuracy: 0.7652 - val_loss: 0.5284 - val_accuracy: 0.8176\n",
            "Epoch 254/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6849 - accuracy: 0.7179 - val_loss: 0.5425 - val_accuracy: 0.8243\n",
            "Epoch 255/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6284 - accuracy: 0.7652 - val_loss: 0.5418 - val_accuracy: 0.8108\n",
            "Epoch 256/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6381 - accuracy: 0.7584 - val_loss: 0.5157 - val_accuracy: 0.8311\n",
            "Epoch 257/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6088 - accuracy: 0.7652 - val_loss: 0.4905 - val_accuracy: 0.8243\n",
            "Epoch 258/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6296 - accuracy: 0.7382 - val_loss: 0.5026 - val_accuracy: 0.8311\n",
            "Epoch 259/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5778 - accuracy: 0.7703 - val_loss: 0.4888 - val_accuracy: 0.8378\n",
            "Epoch 260/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6251 - accuracy: 0.7483 - val_loss: 0.5150 - val_accuracy: 0.8176\n",
            "Epoch 261/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6348 - accuracy: 0.7500 - val_loss: 0.5257 - val_accuracy: 0.8378\n",
            "Epoch 262/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6338 - accuracy: 0.7669 - val_loss: 0.5140 - val_accuracy: 0.8243\n",
            "Epoch 263/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6052 - accuracy: 0.7416 - val_loss: 0.4913 - val_accuracy: 0.8446\n",
            "Epoch 264/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6139 - accuracy: 0.7517 - val_loss: 0.5045 - val_accuracy: 0.8311\n",
            "Epoch 265/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6602 - accuracy: 0.7365 - val_loss: 0.5428 - val_accuracy: 0.8176\n",
            "Epoch 266/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6349 - accuracy: 0.7449 - val_loss: 0.5136 - val_accuracy: 0.8446\n",
            "Epoch 267/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6565 - accuracy: 0.7297 - val_loss: 0.4993 - val_accuracy: 0.8446\n",
            "Epoch 268/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6313 - accuracy: 0.7399 - val_loss: 0.5076 - val_accuracy: 0.8446\n",
            "Epoch 269/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6510 - accuracy: 0.7213 - val_loss: 0.5160 - val_accuracy: 0.8378\n",
            "Epoch 270/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6171 - accuracy: 0.7348 - val_loss: 0.5095 - val_accuracy: 0.8378\n",
            "Epoch 271/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6247 - accuracy: 0.7399 - val_loss: 0.4823 - val_accuracy: 0.8378\n",
            "Epoch 272/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6357 - accuracy: 0.7382 - val_loss: 0.5122 - val_accuracy: 0.8311\n",
            "Epoch 273/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6431 - accuracy: 0.7230 - val_loss: 0.5156 - val_accuracy: 0.8514\n",
            "Epoch 274/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6057 - accuracy: 0.7652 - val_loss: 0.5039 - val_accuracy: 0.8311\n",
            "Epoch 275/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6095 - accuracy: 0.7500 - val_loss: 0.4970 - val_accuracy: 0.8378\n",
            "Epoch 276/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6044 - accuracy: 0.7669 - val_loss: 0.4911 - val_accuracy: 0.8243\n",
            "Epoch 277/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6292 - accuracy: 0.7652 - val_loss: 0.5185 - val_accuracy: 0.8243\n",
            "Epoch 278/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6457 - accuracy: 0.7382 - val_loss: 0.5254 - val_accuracy: 0.8378\n",
            "Epoch 279/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6119 - accuracy: 0.7280 - val_loss: 0.5317 - val_accuracy: 0.8108\n",
            "Epoch 280/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6077 - accuracy: 0.7635 - val_loss: 0.5114 - val_accuracy: 0.8041\n",
            "Epoch 281/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6281 - accuracy: 0.7652 - val_loss: 0.5052 - val_accuracy: 0.8176\n",
            "Epoch 282/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5983 - accuracy: 0.7584 - val_loss: 0.4993 - val_accuracy: 0.8108\n",
            "Epoch 283/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6144 - accuracy: 0.7601 - val_loss: 0.4855 - val_accuracy: 0.8378\n",
            "Epoch 284/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5677 - accuracy: 0.7770 - val_loss: 0.4913 - val_accuracy: 0.8378\n",
            "Epoch 285/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6837 - accuracy: 0.7061 - val_loss: 0.5146 - val_accuracy: 0.8176\n",
            "Epoch 286/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6305 - accuracy: 0.7466 - val_loss: 0.4970 - val_accuracy: 0.8378\n",
            "Epoch 287/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5984 - accuracy: 0.7618 - val_loss: 0.4929 - val_accuracy: 0.8176\n",
            "Epoch 288/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5986 - accuracy: 0.7635 - val_loss: 0.4989 - val_accuracy: 0.8378\n",
            "Epoch 289/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5948 - accuracy: 0.7686 - val_loss: 0.5029 - val_accuracy: 0.8378\n",
            "Epoch 290/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6371 - accuracy: 0.7297 - val_loss: 0.5380 - val_accuracy: 0.8311\n",
            "Epoch 291/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6458 - accuracy: 0.7331 - val_loss: 0.5483 - val_accuracy: 0.8108\n",
            "Epoch 292/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5830 - accuracy: 0.7635 - val_loss: 0.5043 - val_accuracy: 0.8446\n",
            "Epoch 293/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6492 - accuracy: 0.7534 - val_loss: 0.5236 - val_accuracy: 0.8378\n",
            "Epoch 294/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6679 - accuracy: 0.7399 - val_loss: 0.5177 - val_accuracy: 0.8243\n",
            "Epoch 295/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5979 - accuracy: 0.7720 - val_loss: 0.4900 - val_accuracy: 0.8514\n",
            "Epoch 296/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6124 - accuracy: 0.7382 - val_loss: 0.5149 - val_accuracy: 0.8446\n",
            "Epoch 297/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5879 - accuracy: 0.7466 - val_loss: 0.5064 - val_accuracy: 0.8378\n",
            "Epoch 298/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5886 - accuracy: 0.7635 - val_loss: 0.4860 - val_accuracy: 0.8446\n",
            "Epoch 299/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6459 - accuracy: 0.7348 - val_loss: 0.4979 - val_accuracy: 0.8446\n",
            "Epoch 300/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6172 - accuracy: 0.7686 - val_loss: 0.4846 - val_accuracy: 0.8378\n",
            "Epoch 301/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6188 - accuracy: 0.7568 - val_loss: 0.5014 - val_accuracy: 0.8514\n",
            "Epoch 302/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5820 - accuracy: 0.7500 - val_loss: 0.4981 - val_accuracy: 0.8446\n",
            "Epoch 303/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5484 - accuracy: 0.7635 - val_loss: 0.4799 - val_accuracy: 0.8378\n",
            "Epoch 304/500\n",
            "74/74 [==============================] - 1s 8ms/step - loss: 0.6159 - accuracy: 0.7466 - val_loss: 0.5073 - val_accuracy: 0.8514\n",
            "Epoch 305/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6329 - accuracy: 0.7483 - val_loss: 0.4973 - val_accuracy: 0.8446\n",
            "Epoch 306/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6229 - accuracy: 0.7432 - val_loss: 0.5054 - val_accuracy: 0.8311\n",
            "Epoch 307/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5483 - accuracy: 0.7703 - val_loss: 0.4850 - val_accuracy: 0.8446\n",
            "Epoch 308/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5904 - accuracy: 0.7652 - val_loss: 0.4958 - val_accuracy: 0.8446\n",
            "Epoch 309/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5937 - accuracy: 0.7618 - val_loss: 0.5044 - val_accuracy: 0.8243\n",
            "Epoch 310/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6138 - accuracy: 0.7449 - val_loss: 0.5166 - val_accuracy: 0.8243\n",
            "Epoch 311/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6371 - accuracy: 0.7297 - val_loss: 0.4998 - val_accuracy: 0.8378\n",
            "Epoch 312/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5756 - accuracy: 0.7939 - val_loss: 0.4833 - val_accuracy: 0.8243\n",
            "Epoch 313/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5911 - accuracy: 0.7652 - val_loss: 0.4906 - val_accuracy: 0.8311\n",
            "Epoch 314/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6075 - accuracy: 0.7314 - val_loss: 0.5053 - val_accuracy: 0.8176\n",
            "Epoch 315/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5751 - accuracy: 0.7635 - val_loss: 0.4835 - val_accuracy: 0.8514\n",
            "Epoch 316/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6432 - accuracy: 0.7500 - val_loss: 0.5194 - val_accuracy: 0.8243\n",
            "Epoch 317/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5968 - accuracy: 0.7635 - val_loss: 0.4806 - val_accuracy: 0.8446\n",
            "Epoch 318/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6117 - accuracy: 0.7264 - val_loss: 0.4970 - val_accuracy: 0.8378\n",
            "Epoch 319/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5749 - accuracy: 0.7635 - val_loss: 0.4717 - val_accuracy: 0.8514\n",
            "Epoch 320/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6160 - accuracy: 0.7584 - val_loss: 0.5095 - val_accuracy: 0.8378\n",
            "Epoch 321/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6651 - accuracy: 0.7331 - val_loss: 0.5513 - val_accuracy: 0.8176\n",
            "Epoch 322/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6429 - accuracy: 0.7534 - val_loss: 0.5331 - val_accuracy: 0.8243\n",
            "Epoch 323/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6650 - accuracy: 0.7314 - val_loss: 0.5278 - val_accuracy: 0.8378\n",
            "Epoch 324/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6125 - accuracy: 0.7517 - val_loss: 0.4996 - val_accuracy: 0.8176\n",
            "Epoch 325/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6133 - accuracy: 0.7517 - val_loss: 0.4821 - val_accuracy: 0.8378\n",
            "Epoch 326/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6082 - accuracy: 0.7500 - val_loss: 0.4929 - val_accuracy: 0.8311\n",
            "Epoch 327/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6047 - accuracy: 0.7736 - val_loss: 0.4957 - val_accuracy: 0.8378\n",
            "Epoch 328/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6179 - accuracy: 0.7534 - val_loss: 0.4902 - val_accuracy: 0.8041\n",
            "Epoch 329/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5705 - accuracy: 0.7703 - val_loss: 0.4853 - val_accuracy: 0.8311\n",
            "Epoch 330/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5588 - accuracy: 0.7905 - val_loss: 0.4676 - val_accuracy: 0.8243\n",
            "Epoch 331/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5775 - accuracy: 0.7872 - val_loss: 0.4952 - val_accuracy: 0.8378\n",
            "Epoch 332/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6250 - accuracy: 0.7568 - val_loss: 0.4802 - val_accuracy: 0.8514\n",
            "Epoch 333/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6009 - accuracy: 0.7568 - val_loss: 0.4878 - val_accuracy: 0.8243\n",
            "Epoch 334/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5843 - accuracy: 0.7652 - val_loss: 0.4993 - val_accuracy: 0.8378\n",
            "Epoch 335/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6330 - accuracy: 0.7804 - val_loss: 0.4972 - val_accuracy: 0.8378\n",
            "Epoch 336/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5885 - accuracy: 0.7601 - val_loss: 0.4699 - val_accuracy: 0.8176\n",
            "Epoch 337/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6169 - accuracy: 0.7449 - val_loss: 0.4801 - val_accuracy: 0.8378\n",
            "Epoch 338/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6205 - accuracy: 0.7517 - val_loss: 0.4847 - val_accuracy: 0.8446\n",
            "Epoch 339/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6268 - accuracy: 0.7365 - val_loss: 0.5065 - val_accuracy: 0.8311\n",
            "Epoch 340/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6201 - accuracy: 0.7551 - val_loss: 0.4846 - val_accuracy: 0.8378\n",
            "Epoch 341/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5856 - accuracy: 0.7601 - val_loss: 0.4840 - val_accuracy: 0.8514\n",
            "Epoch 342/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5913 - accuracy: 0.7584 - val_loss: 0.5049 - val_accuracy: 0.8446\n",
            "Epoch 343/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5968 - accuracy: 0.7770 - val_loss: 0.4864 - val_accuracy: 0.8581\n",
            "Epoch 344/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6165 - accuracy: 0.7703 - val_loss: 0.4860 - val_accuracy: 0.8446\n",
            "Epoch 345/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6180 - accuracy: 0.7399 - val_loss: 0.4948 - val_accuracy: 0.8176\n",
            "Epoch 346/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6159 - accuracy: 0.7618 - val_loss: 0.4936 - val_accuracy: 0.8176\n",
            "Epoch 347/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5794 - accuracy: 0.7584 - val_loss: 0.4678 - val_accuracy: 0.8176\n",
            "Epoch 348/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6234 - accuracy: 0.7551 - val_loss: 0.4924 - val_accuracy: 0.8243\n",
            "Epoch 349/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6198 - accuracy: 0.7517 - val_loss: 0.4830 - val_accuracy: 0.8514\n",
            "Epoch 350/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6192 - accuracy: 0.7517 - val_loss: 0.4755 - val_accuracy: 0.8311\n",
            "Epoch 351/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5894 - accuracy: 0.7703 - val_loss: 0.4877 - val_accuracy: 0.8243\n",
            "Epoch 352/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6223 - accuracy: 0.7297 - val_loss: 0.5099 - val_accuracy: 0.8311\n",
            "Epoch 353/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5976 - accuracy: 0.7517 - val_loss: 0.4849 - val_accuracy: 0.8108\n",
            "Epoch 354/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5967 - accuracy: 0.7787 - val_loss: 0.4890 - val_accuracy: 0.8446\n",
            "Epoch 355/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5669 - accuracy: 0.7551 - val_loss: 0.4632 - val_accuracy: 0.8378\n",
            "Epoch 356/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6359 - accuracy: 0.7872 - val_loss: 0.4853 - val_accuracy: 0.8378\n",
            "Epoch 357/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5806 - accuracy: 0.7551 - val_loss: 0.4867 - val_accuracy: 0.8243\n",
            "Epoch 358/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5991 - accuracy: 0.7770 - val_loss: 0.4900 - val_accuracy: 0.8311\n",
            "Epoch 359/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6233 - accuracy: 0.7432 - val_loss: 0.4708 - val_accuracy: 0.8514\n",
            "Epoch 360/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5790 - accuracy: 0.7635 - val_loss: 0.4617 - val_accuracy: 0.8311\n",
            "Epoch 361/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6114 - accuracy: 0.7432 - val_loss: 0.4957 - val_accuracy: 0.8243\n",
            "Epoch 362/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6006 - accuracy: 0.7669 - val_loss: 0.4876 - val_accuracy: 0.8041\n",
            "Epoch 363/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6269 - accuracy: 0.7500 - val_loss: 0.5001 - val_accuracy: 0.8243\n",
            "Epoch 364/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5966 - accuracy: 0.7635 - val_loss: 0.4785 - val_accuracy: 0.8378\n",
            "Epoch 365/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5743 - accuracy: 0.7922 - val_loss: 0.4552 - val_accuracy: 0.8311\n",
            "Epoch 366/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6144 - accuracy: 0.7618 - val_loss: 0.4750 - val_accuracy: 0.8311\n",
            "Epoch 367/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5807 - accuracy: 0.7787 - val_loss: 0.4635 - val_accuracy: 0.8514\n",
            "Epoch 368/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5869 - accuracy: 0.7872 - val_loss: 0.4674 - val_accuracy: 0.8514\n",
            "Epoch 369/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5544 - accuracy: 0.7753 - val_loss: 0.4597 - val_accuracy: 0.8243\n",
            "Epoch 370/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5756 - accuracy: 0.7787 - val_loss: 0.4708 - val_accuracy: 0.8649\n",
            "Epoch 371/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5764 - accuracy: 0.7838 - val_loss: 0.4769 - val_accuracy: 0.8378\n",
            "Epoch 372/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6320 - accuracy: 0.7736 - val_loss: 0.4981 - val_accuracy: 0.8446\n",
            "Epoch 373/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5741 - accuracy: 0.7804 - val_loss: 0.4752 - val_accuracy: 0.8311\n",
            "Epoch 374/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5827 - accuracy: 0.7720 - val_loss: 0.4663 - val_accuracy: 0.8514\n",
            "Epoch 375/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5620 - accuracy: 0.7855 - val_loss: 0.4568 - val_accuracy: 0.8446\n",
            "Epoch 376/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5849 - accuracy: 0.7703 - val_loss: 0.4791 - val_accuracy: 0.8378\n",
            "Epoch 377/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5905 - accuracy: 0.7669 - val_loss: 0.4848 - val_accuracy: 0.8243\n",
            "Epoch 378/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6159 - accuracy: 0.7601 - val_loss: 0.4660 - val_accuracy: 0.8243\n",
            "Epoch 379/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5732 - accuracy: 0.7872 - val_loss: 0.4677 - val_accuracy: 0.8446\n",
            "Epoch 380/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5646 - accuracy: 0.7804 - val_loss: 0.4752 - val_accuracy: 0.8378\n",
            "Epoch 381/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6126 - accuracy: 0.7889 - val_loss: 0.4656 - val_accuracy: 0.8446\n",
            "Epoch 382/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5415 - accuracy: 0.7703 - val_loss: 0.4632 - val_accuracy: 0.8378\n",
            "Epoch 383/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6182 - accuracy: 0.7601 - val_loss: 0.4646 - val_accuracy: 0.8649\n",
            "Epoch 384/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6189 - accuracy: 0.7601 - val_loss: 0.4617 - val_accuracy: 0.8649\n",
            "Epoch 385/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6059 - accuracy: 0.7686 - val_loss: 0.4924 - val_accuracy: 0.8649\n",
            "Epoch 386/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5837 - accuracy: 0.7939 - val_loss: 0.4830 - val_accuracy: 0.8514\n",
            "Epoch 387/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5402 - accuracy: 0.7753 - val_loss: 0.4453 - val_accuracy: 0.8581\n",
            "Epoch 388/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6172 - accuracy: 0.7584 - val_loss: 0.4721 - val_accuracy: 0.8446\n",
            "Epoch 389/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5784 - accuracy: 0.7838 - val_loss: 0.4670 - val_accuracy: 0.8446\n",
            "Epoch 390/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6029 - accuracy: 0.7517 - val_loss: 0.4806 - val_accuracy: 0.8446\n",
            "Epoch 391/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5789 - accuracy: 0.7483 - val_loss: 0.4782 - val_accuracy: 0.8446\n",
            "Epoch 392/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5616 - accuracy: 0.7720 - val_loss: 0.4474 - val_accuracy: 0.8378\n",
            "Epoch 393/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5847 - accuracy: 0.7669 - val_loss: 0.4560 - val_accuracy: 0.8446\n",
            "Epoch 394/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5967 - accuracy: 0.7652 - val_loss: 0.4825 - val_accuracy: 0.8311\n",
            "Epoch 395/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6038 - accuracy: 0.7618 - val_loss: 0.4971 - val_accuracy: 0.8378\n",
            "Epoch 396/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6018 - accuracy: 0.7618 - val_loss: 0.4620 - val_accuracy: 0.8514\n",
            "Epoch 397/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5994 - accuracy: 0.7753 - val_loss: 0.4700 - val_accuracy: 0.8378\n",
            "Epoch 398/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6071 - accuracy: 0.7365 - val_loss: 0.4526 - val_accuracy: 0.8378\n",
            "Epoch 399/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6249 - accuracy: 0.7416 - val_loss: 0.4678 - val_accuracy: 0.8581\n",
            "Epoch 400/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5657 - accuracy: 0.7652 - val_loss: 0.4291 - val_accuracy: 0.8581\n",
            "Epoch 401/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5913 - accuracy: 0.7669 - val_loss: 0.4905 - val_accuracy: 0.8446\n",
            "Epoch 402/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6140 - accuracy: 0.7584 - val_loss: 0.4912 - val_accuracy: 0.8446\n",
            "Epoch 403/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5756 - accuracy: 0.7635 - val_loss: 0.4662 - val_accuracy: 0.8446\n",
            "Epoch 404/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6153 - accuracy: 0.7382 - val_loss: 0.4775 - val_accuracy: 0.8243\n",
            "Epoch 405/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5982 - accuracy: 0.7584 - val_loss: 0.4688 - val_accuracy: 0.8311\n",
            "Epoch 406/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5955 - accuracy: 0.7652 - val_loss: 0.4784 - val_accuracy: 0.8514\n",
            "Epoch 407/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6032 - accuracy: 0.7314 - val_loss: 0.4803 - val_accuracy: 0.8649\n",
            "Epoch 408/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5181 - accuracy: 0.7855 - val_loss: 0.4393 - val_accuracy: 0.8243\n",
            "Epoch 409/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6118 - accuracy: 0.7365 - val_loss: 0.4672 - val_accuracy: 0.8446\n",
            "Epoch 410/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6212 - accuracy: 0.7872 - val_loss: 0.4772 - val_accuracy: 0.8311\n",
            "Epoch 411/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5766 - accuracy: 0.7584 - val_loss: 0.4565 - val_accuracy: 0.8514\n",
            "Epoch 412/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5772 - accuracy: 0.7990 - val_loss: 0.4583 - val_accuracy: 0.8311\n",
            "Epoch 413/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5839 - accuracy: 0.7601 - val_loss: 0.4951 - val_accuracy: 0.8378\n",
            "Epoch 414/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6039 - accuracy: 0.7584 - val_loss: 0.4977 - val_accuracy: 0.8243\n",
            "Epoch 415/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5830 - accuracy: 0.7500 - val_loss: 0.4753 - val_accuracy: 0.8311\n",
            "Epoch 416/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5512 - accuracy: 0.7601 - val_loss: 0.4577 - val_accuracy: 0.8649\n",
            "Epoch 417/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6097 - accuracy: 0.7703 - val_loss: 0.4714 - val_accuracy: 0.8716\n",
            "Epoch 418/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6357 - accuracy: 0.7432 - val_loss: 0.4968 - val_accuracy: 0.8514\n",
            "Epoch 419/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6086 - accuracy: 0.7686 - val_loss: 0.4848 - val_accuracy: 0.8243\n",
            "Epoch 420/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5976 - accuracy: 0.7736 - val_loss: 0.4789 - val_accuracy: 0.8716\n",
            "Epoch 421/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6290 - accuracy: 0.7551 - val_loss: 0.4593 - val_accuracy: 0.8649\n",
            "Epoch 422/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5621 - accuracy: 0.7652 - val_loss: 0.4440 - val_accuracy: 0.8649\n",
            "Epoch 423/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5907 - accuracy: 0.7770 - val_loss: 0.4589 - val_accuracy: 0.8514\n",
            "Epoch 424/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5973 - accuracy: 0.7584 - val_loss: 0.4591 - val_accuracy: 0.8581\n",
            "Epoch 425/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5766 - accuracy: 0.7517 - val_loss: 0.4689 - val_accuracy: 0.8514\n",
            "Epoch 426/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5864 - accuracy: 0.7736 - val_loss: 0.4662 - val_accuracy: 0.8446\n",
            "Epoch 427/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6173 - accuracy: 0.7399 - val_loss: 0.4800 - val_accuracy: 0.8581\n",
            "Epoch 428/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6123 - accuracy: 0.7601 - val_loss: 0.4810 - val_accuracy: 0.8378\n",
            "Epoch 429/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5757 - accuracy: 0.7703 - val_loss: 0.4548 - val_accuracy: 0.8311\n",
            "Epoch 430/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6004 - accuracy: 0.7534 - val_loss: 0.4952 - val_accuracy: 0.8243\n",
            "Epoch 431/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5829 - accuracy: 0.7703 - val_loss: 0.4719 - val_accuracy: 0.8514\n",
            "Epoch 432/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5823 - accuracy: 0.7483 - val_loss: 0.4482 - val_accuracy: 0.8311\n",
            "Epoch 433/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5847 - accuracy: 0.7584 - val_loss: 0.4610 - val_accuracy: 0.8514\n",
            "Epoch 434/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5943 - accuracy: 0.7652 - val_loss: 0.4480 - val_accuracy: 0.8581\n",
            "Epoch 435/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5537 - accuracy: 0.7770 - val_loss: 0.4464 - val_accuracy: 0.8581\n",
            "Epoch 436/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5813 - accuracy: 0.7652 - val_loss: 0.4563 - val_accuracy: 0.8446\n",
            "Epoch 437/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5866 - accuracy: 0.7939 - val_loss: 0.4767 - val_accuracy: 0.8176\n",
            "Epoch 438/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5859 - accuracy: 0.7584 - val_loss: 0.4585 - val_accuracy: 0.8716\n",
            "Epoch 439/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5995 - accuracy: 0.7736 - val_loss: 0.4768 - val_accuracy: 0.8581\n",
            "Epoch 440/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5581 - accuracy: 0.7753 - val_loss: 0.4665 - val_accuracy: 0.8716\n",
            "Epoch 441/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5570 - accuracy: 0.7956 - val_loss: 0.4674 - val_accuracy: 0.8446\n",
            "Epoch 442/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6128 - accuracy: 0.7517 - val_loss: 0.4692 - val_accuracy: 0.8378\n",
            "Epoch 443/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5910 - accuracy: 0.7584 - val_loss: 0.4603 - val_accuracy: 0.8716\n",
            "Epoch 444/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5655 - accuracy: 0.7855 - val_loss: 0.4594 - val_accuracy: 0.8378\n",
            "Epoch 445/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6105 - accuracy: 0.7534 - val_loss: 0.4853 - val_accuracy: 0.8446\n",
            "Epoch 446/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5706 - accuracy: 0.7753 - val_loss: 0.4734 - val_accuracy: 0.8176\n",
            "Epoch 447/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6478 - accuracy: 0.7382 - val_loss: 0.5013 - val_accuracy: 0.8446\n",
            "Epoch 448/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6228 - accuracy: 0.7534 - val_loss: 0.4850 - val_accuracy: 0.8446\n",
            "Epoch 449/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6042 - accuracy: 0.7551 - val_loss: 0.4867 - val_accuracy: 0.8514\n",
            "Epoch 450/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6323 - accuracy: 0.7618 - val_loss: 0.5162 - val_accuracy: 0.8378\n",
            "Epoch 451/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6042 - accuracy: 0.7736 - val_loss: 0.4947 - val_accuracy: 0.8649\n",
            "Epoch 452/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6224 - accuracy: 0.7432 - val_loss: 0.4953 - val_accuracy: 0.8243\n",
            "Epoch 453/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5713 - accuracy: 0.7956 - val_loss: 0.4839 - val_accuracy: 0.8176\n",
            "Epoch 454/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6277 - accuracy: 0.7399 - val_loss: 0.4735 - val_accuracy: 0.8311\n",
            "Epoch 455/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5777 - accuracy: 0.7686 - val_loss: 0.4755 - val_accuracy: 0.8514\n",
            "Epoch 456/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5818 - accuracy: 0.7686 - val_loss: 0.4867 - val_accuracy: 0.8176\n",
            "Epoch 457/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5985 - accuracy: 0.7703 - val_loss: 0.4740 - val_accuracy: 0.8243\n",
            "Epoch 458/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6049 - accuracy: 0.7872 - val_loss: 0.4852 - val_accuracy: 0.8378\n",
            "Epoch 459/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6006 - accuracy: 0.7534 - val_loss: 0.4755 - val_accuracy: 0.8446\n",
            "Epoch 460/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5807 - accuracy: 0.7652 - val_loss: 0.4726 - val_accuracy: 0.8784\n",
            "Epoch 461/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6416 - accuracy: 0.7365 - val_loss: 0.4974 - val_accuracy: 0.8378\n",
            "Epoch 462/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6125 - accuracy: 0.7770 - val_loss: 0.4708 - val_accuracy: 0.8243\n",
            "Epoch 463/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5675 - accuracy: 0.7956 - val_loss: 0.4689 - val_accuracy: 0.8311\n",
            "Epoch 464/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6321 - accuracy: 0.7432 - val_loss: 0.4765 - val_accuracy: 0.8514\n",
            "Epoch 465/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6090 - accuracy: 0.7838 - val_loss: 0.5037 - val_accuracy: 0.8243\n",
            "Epoch 466/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6156 - accuracy: 0.7736 - val_loss: 0.4741 - val_accuracy: 0.8446\n",
            "Epoch 467/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5844 - accuracy: 0.7770 - val_loss: 0.4726 - val_accuracy: 0.8311\n",
            "Epoch 468/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5903 - accuracy: 0.7770 - val_loss: 0.4777 - val_accuracy: 0.8311\n",
            "Epoch 469/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5863 - accuracy: 0.7872 - val_loss: 0.4609 - val_accuracy: 0.8243\n",
            "Epoch 470/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6415 - accuracy: 0.7618 - val_loss: 0.4496 - val_accuracy: 0.8514\n",
            "Epoch 471/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5446 - accuracy: 0.8091 - val_loss: 0.4496 - val_accuracy: 0.8514\n",
            "Epoch 472/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5887 - accuracy: 0.7787 - val_loss: 0.4886 - val_accuracy: 0.8581\n",
            "Epoch 473/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6328 - accuracy: 0.7703 - val_loss: 0.5163 - val_accuracy: 0.8514\n",
            "Epoch 474/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6342 - accuracy: 0.7584 - val_loss: 0.4976 - val_accuracy: 0.8514\n",
            "Epoch 475/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6390 - accuracy: 0.7618 - val_loss: 0.5071 - val_accuracy: 0.8446\n",
            "Epoch 476/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5526 - accuracy: 0.7872 - val_loss: 0.4533 - val_accuracy: 0.8446\n",
            "Epoch 477/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6156 - accuracy: 0.7669 - val_loss: 0.4715 - val_accuracy: 0.8446\n",
            "Epoch 478/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5955 - accuracy: 0.7568 - val_loss: 0.4706 - val_accuracy: 0.8311\n",
            "Epoch 479/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5827 - accuracy: 0.7720 - val_loss: 0.4715 - val_accuracy: 0.8514\n",
            "Epoch 480/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5872 - accuracy: 0.7872 - val_loss: 0.4748 - val_accuracy: 0.8243\n",
            "Epoch 481/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6663 - accuracy: 0.7517 - val_loss: 0.5153 - val_accuracy: 0.8514\n",
            "Epoch 482/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5964 - accuracy: 0.7652 - val_loss: 0.4896 - val_accuracy: 0.8581\n",
            "Epoch 483/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5639 - accuracy: 0.7872 - val_loss: 0.4510 - val_accuracy: 0.8378\n",
            "Epoch 484/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6044 - accuracy: 0.7753 - val_loss: 0.4627 - val_accuracy: 0.8716\n",
            "Epoch 485/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5682 - accuracy: 0.7686 - val_loss: 0.4665 - val_accuracy: 0.8446\n",
            "Epoch 486/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5579 - accuracy: 0.7956 - val_loss: 0.4580 - val_accuracy: 0.8311\n",
            "Epoch 487/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5684 - accuracy: 0.7804 - val_loss: 0.4677 - val_accuracy: 0.8176\n",
            "Epoch 488/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6145 - accuracy: 0.7483 - val_loss: 0.4867 - val_accuracy: 0.8378\n",
            "Epoch 489/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6342 - accuracy: 0.7517 - val_loss: 0.4941 - val_accuracy: 0.8446\n",
            "Epoch 490/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6030 - accuracy: 0.7601 - val_loss: 0.4824 - val_accuracy: 0.8311\n",
            "Epoch 491/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5981 - accuracy: 0.7736 - val_loss: 0.4915 - val_accuracy: 0.8243\n",
            "Epoch 492/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5857 - accuracy: 0.7770 - val_loss: 0.4657 - val_accuracy: 0.8243\n",
            "Epoch 493/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5892 - accuracy: 0.7584 - val_loss: 0.4840 - val_accuracy: 0.8311\n",
            "Epoch 494/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5696 - accuracy: 0.7584 - val_loss: 0.4730 - val_accuracy: 0.8514\n",
            "Epoch 495/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5854 - accuracy: 0.7703 - val_loss: 0.4848 - val_accuracy: 0.8311\n",
            "Epoch 496/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6293 - accuracy: 0.7601 - val_loss: 0.5031 - val_accuracy: 0.8311\n",
            "Epoch 497/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5412 - accuracy: 0.7889 - val_loss: 0.4743 - val_accuracy: 0.8378\n",
            "Epoch 498/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6135 - accuracy: 0.7652 - val_loss: 0.4770 - val_accuracy: 0.8581\n",
            "Epoch 499/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6036 - accuracy: 0.7348 - val_loss: 0.4665 - val_accuracy: 0.8514\n",
            "Epoch 500/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5700 - accuracy: 0.7956 - val_loss: 0.4586 - val_accuracy: 0.8378\n",
            "5/5 [==============================] - 0s 3ms/step - loss: 0.4586 - accuracy: 0.8378\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:62: DeprecationWarning: scipy.interp is deprecated and will be removed in SciPy 2.0.0, use numpy.interp instead\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/500\n",
            "74/74 [==============================] - 1s 10ms/step - loss: 1.6511 - accuracy: 0.5000 - val_loss: 1.3414 - val_accuracy: 0.5203\n",
            "Epoch 2/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 1.3171 - accuracy: 0.5439 - val_loss: 1.2428 - val_accuracy: 0.5203\n",
            "Epoch 3/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 1.1793 - accuracy: 0.5642 - val_loss: 1.1327 - val_accuracy: 0.5068\n",
            "Epoch 4/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 1.0933 - accuracy: 0.5405 - val_loss: 1.0409 - val_accuracy: 0.5203\n",
            "Epoch 5/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.9950 - accuracy: 0.5794 - val_loss: 0.9600 - val_accuracy: 0.5676\n",
            "Epoch 6/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.9302 - accuracy: 0.6047 - val_loss: 0.9221 - val_accuracy: 0.5878\n",
            "Epoch 7/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.9297 - accuracy: 0.5642 - val_loss: 0.8838 - val_accuracy: 0.5946\n",
            "Epoch 8/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8538 - accuracy: 0.6233 - val_loss: 0.8565 - val_accuracy: 0.5608\n",
            "Epoch 9/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8665 - accuracy: 0.6064 - val_loss: 0.8378 - val_accuracy: 0.6689\n",
            "Epoch 10/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.8657 - accuracy: 0.6166 - val_loss: 0.7903 - val_accuracy: 0.7027\n",
            "Epoch 11/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8497 - accuracy: 0.6503 - val_loss: 0.8030 - val_accuracy: 0.6959\n",
            "Epoch 12/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8347 - accuracy: 0.6672 - val_loss: 0.7761 - val_accuracy: 0.7905\n",
            "Epoch 13/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8240 - accuracy: 0.6622 - val_loss: 0.7388 - val_accuracy: 0.7635\n",
            "Epoch 14/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8299 - accuracy: 0.6267 - val_loss: 0.7820 - val_accuracy: 0.6959\n",
            "Epoch 15/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.8181 - accuracy: 0.6588 - val_loss: 0.7364 - val_accuracy: 0.7770\n",
            "Epoch 16/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8219 - accuracy: 0.6689 - val_loss: 0.7812 - val_accuracy: 0.7095\n",
            "Epoch 17/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.8170 - accuracy: 0.6757 - val_loss: 0.7713 - val_accuracy: 0.7297\n",
            "Epoch 18/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8073 - accuracy: 0.6706 - val_loss: 0.7467 - val_accuracy: 0.7635\n",
            "Epoch 19/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.8481 - accuracy: 0.6588 - val_loss: 0.7704 - val_accuracy: 0.7500\n",
            "Epoch 20/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.8614 - accuracy: 0.6774 - val_loss: 0.7845 - val_accuracy: 0.7568\n",
            "Epoch 21/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8462 - accuracy: 0.6723 - val_loss: 0.7482 - val_accuracy: 0.7568\n",
            "Epoch 22/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.8210 - accuracy: 0.6740 - val_loss: 0.7410 - val_accuracy: 0.7973\n",
            "Epoch 23/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.8332 - accuracy: 0.6706 - val_loss: 0.7475 - val_accuracy: 0.7703\n",
            "Epoch 24/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7938 - accuracy: 0.6943 - val_loss: 0.7134 - val_accuracy: 0.7770\n",
            "Epoch 25/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.8056 - accuracy: 0.6943 - val_loss: 0.7126 - val_accuracy: 0.8108\n",
            "Epoch 26/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7806 - accuracy: 0.7145 - val_loss: 0.6978 - val_accuracy: 0.7905\n",
            "Epoch 27/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7988 - accuracy: 0.6926 - val_loss: 0.6987 - val_accuracy: 0.7905\n",
            "Epoch 28/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7783 - accuracy: 0.7044 - val_loss: 0.6798 - val_accuracy: 0.7838\n",
            "Epoch 29/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7695 - accuracy: 0.7179 - val_loss: 0.6820 - val_accuracy: 0.7568\n",
            "Epoch 30/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7630 - accuracy: 0.6892 - val_loss: 0.6971 - val_accuracy: 0.7568\n",
            "Epoch 31/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7775 - accuracy: 0.6875 - val_loss: 0.6990 - val_accuracy: 0.8041\n",
            "Epoch 32/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7836 - accuracy: 0.6892 - val_loss: 0.6982 - val_accuracy: 0.7770\n",
            "Epoch 33/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7398 - accuracy: 0.7179 - val_loss: 0.6579 - val_accuracy: 0.7703\n",
            "Epoch 34/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7960 - accuracy: 0.6841 - val_loss: 0.6930 - val_accuracy: 0.7635\n",
            "Epoch 35/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.8096 - accuracy: 0.6706 - val_loss: 0.6691 - val_accuracy: 0.7905\n",
            "Epoch 36/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7493 - accuracy: 0.7365 - val_loss: 0.6592 - val_accuracy: 0.7973\n",
            "Epoch 37/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7631 - accuracy: 0.7297 - val_loss: 0.6519 - val_accuracy: 0.8176\n",
            "Epoch 38/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7922 - accuracy: 0.6892 - val_loss: 0.6850 - val_accuracy: 0.7838\n",
            "Epoch 39/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7613 - accuracy: 0.7145 - val_loss: 0.6669 - val_accuracy: 0.8041\n",
            "Epoch 40/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7672 - accuracy: 0.6943 - val_loss: 0.6647 - val_accuracy: 0.7905\n",
            "Epoch 41/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7682 - accuracy: 0.6875 - val_loss: 0.6566 - val_accuracy: 0.7770\n",
            "Epoch 42/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7503 - accuracy: 0.6706 - val_loss: 0.6573 - val_accuracy: 0.7703\n",
            "Epoch 43/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7515 - accuracy: 0.7027 - val_loss: 0.6585 - val_accuracy: 0.8108\n",
            "Epoch 44/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.8000 - accuracy: 0.6791 - val_loss: 0.6938 - val_accuracy: 0.7973\n",
            "Epoch 45/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7971 - accuracy: 0.6824 - val_loss: 0.6791 - val_accuracy: 0.8041\n",
            "Epoch 46/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7916 - accuracy: 0.6791 - val_loss: 0.6832 - val_accuracy: 0.7838\n",
            "Epoch 47/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7533 - accuracy: 0.7027 - val_loss: 0.6508 - val_accuracy: 0.8243\n",
            "Epoch 48/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7508 - accuracy: 0.7196 - val_loss: 0.6343 - val_accuracy: 0.8176\n",
            "Epoch 49/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7821 - accuracy: 0.7128 - val_loss: 0.6490 - val_accuracy: 0.8108\n",
            "Epoch 50/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7604 - accuracy: 0.7213 - val_loss: 0.6968 - val_accuracy: 0.7635\n",
            "Epoch 51/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7876 - accuracy: 0.6824 - val_loss: 0.6724 - val_accuracy: 0.8041\n",
            "Epoch 52/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7462 - accuracy: 0.6943 - val_loss: 0.6634 - val_accuracy: 0.8041\n",
            "Epoch 53/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7653 - accuracy: 0.6740 - val_loss: 0.6279 - val_accuracy: 0.7973\n",
            "Epoch 54/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7584 - accuracy: 0.6976 - val_loss: 0.6253 - val_accuracy: 0.8176\n",
            "Epoch 55/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7651 - accuracy: 0.6926 - val_loss: 0.6420 - val_accuracy: 0.8041\n",
            "Epoch 56/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7576 - accuracy: 0.6774 - val_loss: 0.6575 - val_accuracy: 0.7973\n",
            "Epoch 57/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7895 - accuracy: 0.6841 - val_loss: 0.6470 - val_accuracy: 0.7973\n",
            "Epoch 58/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7497 - accuracy: 0.7162 - val_loss: 0.6151 - val_accuracy: 0.8041\n",
            "Epoch 59/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7568 - accuracy: 0.6959 - val_loss: 0.6190 - val_accuracy: 0.8581\n",
            "Epoch 60/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7853 - accuracy: 0.6976 - val_loss: 0.6503 - val_accuracy: 0.8041\n",
            "Epoch 61/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7468 - accuracy: 0.7078 - val_loss: 0.6290 - val_accuracy: 0.7905\n",
            "Epoch 62/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7768 - accuracy: 0.6875 - val_loss: 0.6654 - val_accuracy: 0.7905\n",
            "Epoch 63/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7329 - accuracy: 0.6943 - val_loss: 0.6282 - val_accuracy: 0.8243\n",
            "Epoch 64/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7178 - accuracy: 0.7145 - val_loss: 0.6218 - val_accuracy: 0.8176\n",
            "Epoch 65/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7466 - accuracy: 0.6993 - val_loss: 0.6361 - val_accuracy: 0.8176\n",
            "Epoch 66/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7438 - accuracy: 0.6909 - val_loss: 0.6238 - val_accuracy: 0.7770\n",
            "Epoch 67/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7363 - accuracy: 0.7044 - val_loss: 0.6042 - val_accuracy: 0.7973\n",
            "Epoch 68/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7261 - accuracy: 0.7179 - val_loss: 0.6361 - val_accuracy: 0.7838\n",
            "Epoch 69/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7361 - accuracy: 0.6993 - val_loss: 0.5929 - val_accuracy: 0.8243\n",
            "Epoch 70/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7357 - accuracy: 0.7382 - val_loss: 0.6206 - val_accuracy: 0.8041\n",
            "Epoch 71/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7377 - accuracy: 0.7162 - val_loss: 0.6301 - val_accuracy: 0.7703\n",
            "Epoch 72/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7574 - accuracy: 0.6875 - val_loss: 0.6333 - val_accuracy: 0.8243\n",
            "Epoch 73/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7385 - accuracy: 0.7230 - val_loss: 0.5877 - val_accuracy: 0.8176\n",
            "Epoch 74/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7466 - accuracy: 0.7382 - val_loss: 0.6289 - val_accuracy: 0.8243\n",
            "Epoch 75/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7675 - accuracy: 0.7078 - val_loss: 0.6288 - val_accuracy: 0.8243\n",
            "Epoch 76/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7451 - accuracy: 0.7027 - val_loss: 0.6221 - val_accuracy: 0.8243\n",
            "Epoch 77/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7043 - accuracy: 0.7213 - val_loss: 0.5746 - val_accuracy: 0.8108\n",
            "Epoch 78/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7322 - accuracy: 0.7179 - val_loss: 0.6080 - val_accuracy: 0.8243\n",
            "Epoch 79/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7291 - accuracy: 0.7297 - val_loss: 0.6102 - val_accuracy: 0.8176\n",
            "Epoch 80/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7448 - accuracy: 0.7078 - val_loss: 0.6131 - val_accuracy: 0.8243\n",
            "Epoch 81/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7100 - accuracy: 0.7348 - val_loss: 0.5949 - val_accuracy: 0.8176\n",
            "Epoch 82/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7198 - accuracy: 0.7314 - val_loss: 0.6009 - val_accuracy: 0.8108\n",
            "Epoch 83/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7158 - accuracy: 0.7230 - val_loss: 0.6617 - val_accuracy: 0.7635\n",
            "Epoch 84/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7504 - accuracy: 0.7095 - val_loss: 0.6429 - val_accuracy: 0.8243\n",
            "Epoch 85/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7360 - accuracy: 0.7145 - val_loss: 0.6259 - val_accuracy: 0.8378\n",
            "Epoch 86/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7629 - accuracy: 0.7044 - val_loss: 0.6337 - val_accuracy: 0.8311\n",
            "Epoch 87/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7537 - accuracy: 0.7264 - val_loss: 0.6198 - val_accuracy: 0.8243\n",
            "Epoch 88/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7067 - accuracy: 0.7314 - val_loss: 0.5797 - val_accuracy: 0.8446\n",
            "Epoch 89/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7454 - accuracy: 0.7213 - val_loss: 0.6125 - val_accuracy: 0.8243\n",
            "Epoch 90/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7079 - accuracy: 0.7382 - val_loss: 0.5891 - val_accuracy: 0.8446\n",
            "Epoch 91/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6859 - accuracy: 0.7483 - val_loss: 0.5740 - val_accuracy: 0.8446\n",
            "Epoch 92/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7108 - accuracy: 0.7145 - val_loss: 0.5970 - val_accuracy: 0.7905\n",
            "Epoch 93/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7161 - accuracy: 0.7601 - val_loss: 0.5738 - val_accuracy: 0.8378\n",
            "Epoch 94/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7134 - accuracy: 0.7331 - val_loss: 0.5862 - val_accuracy: 0.8243\n",
            "Epoch 95/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7245 - accuracy: 0.7466 - val_loss: 0.6273 - val_accuracy: 0.8041\n",
            "Epoch 96/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7242 - accuracy: 0.7280 - val_loss: 0.5915 - val_accuracy: 0.8108\n",
            "Epoch 97/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7325 - accuracy: 0.7230 - val_loss: 0.6096 - val_accuracy: 0.8041\n",
            "Epoch 98/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7059 - accuracy: 0.7348 - val_loss: 0.5751 - val_accuracy: 0.8311\n",
            "Epoch 99/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7194 - accuracy: 0.7348 - val_loss: 0.5891 - val_accuracy: 0.8378\n",
            "Epoch 100/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6863 - accuracy: 0.7348 - val_loss: 0.5627 - val_accuracy: 0.8378\n",
            "Epoch 101/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7192 - accuracy: 0.7095 - val_loss: 0.5817 - val_accuracy: 0.8311\n",
            "Epoch 102/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6991 - accuracy: 0.7264 - val_loss: 0.5735 - val_accuracy: 0.8311\n",
            "Epoch 103/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7180 - accuracy: 0.7196 - val_loss: 0.5812 - val_accuracy: 0.8514\n",
            "Epoch 104/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7196 - accuracy: 0.7314 - val_loss: 0.5839 - val_accuracy: 0.8243\n",
            "Epoch 105/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7550 - accuracy: 0.6875 - val_loss: 0.5993 - val_accuracy: 0.8514\n",
            "Epoch 106/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7281 - accuracy: 0.7247 - val_loss: 0.5782 - val_accuracy: 0.8378\n",
            "Epoch 107/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7240 - accuracy: 0.6841 - val_loss: 0.5678 - val_accuracy: 0.8108\n",
            "Epoch 108/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7021 - accuracy: 0.7382 - val_loss: 0.5871 - val_accuracy: 0.7973\n",
            "Epoch 109/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7038 - accuracy: 0.7111 - val_loss: 0.5690 - val_accuracy: 0.8378\n",
            "Epoch 110/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7137 - accuracy: 0.7247 - val_loss: 0.5775 - val_accuracy: 0.8243\n",
            "Epoch 111/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6830 - accuracy: 0.7500 - val_loss: 0.5585 - val_accuracy: 0.8514\n",
            "Epoch 112/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7165 - accuracy: 0.7196 - val_loss: 0.5570 - val_accuracy: 0.8446\n",
            "Epoch 113/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6968 - accuracy: 0.7213 - val_loss: 0.5491 - val_accuracy: 0.8649\n",
            "Epoch 114/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7239 - accuracy: 0.7162 - val_loss: 0.5776 - val_accuracy: 0.8581\n",
            "Epoch 115/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7166 - accuracy: 0.7247 - val_loss: 0.5894 - val_accuracy: 0.8446\n",
            "Epoch 116/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6904 - accuracy: 0.7145 - val_loss: 0.5589 - val_accuracy: 0.8581\n",
            "Epoch 117/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6835 - accuracy: 0.7348 - val_loss: 0.5308 - val_accuracy: 0.8581\n",
            "Epoch 118/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6892 - accuracy: 0.7280 - val_loss: 0.5631 - val_accuracy: 0.8378\n",
            "Epoch 119/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6649 - accuracy: 0.7618 - val_loss: 0.5687 - val_accuracy: 0.8514\n",
            "Epoch 120/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7126 - accuracy: 0.7128 - val_loss: 0.5740 - val_accuracy: 0.8311\n",
            "Epoch 121/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6936 - accuracy: 0.7128 - val_loss: 0.5558 - val_accuracy: 0.8446\n",
            "Epoch 122/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7077 - accuracy: 0.7213 - val_loss: 0.5661 - val_accuracy: 0.8378\n",
            "Epoch 123/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6510 - accuracy: 0.7483 - val_loss: 0.5206 - val_accuracy: 0.8514\n",
            "Epoch 124/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6549 - accuracy: 0.7331 - val_loss: 0.5362 - val_accuracy: 0.8243\n",
            "Epoch 125/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7119 - accuracy: 0.7264 - val_loss: 0.5594 - val_accuracy: 0.8649\n",
            "Epoch 126/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7400 - accuracy: 0.6959 - val_loss: 0.5820 - val_accuracy: 0.8514\n",
            "Epoch 127/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6592 - accuracy: 0.7618 - val_loss: 0.5454 - val_accuracy: 0.8243\n",
            "Epoch 128/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7236 - accuracy: 0.7111 - val_loss: 0.5920 - val_accuracy: 0.7838\n",
            "Epoch 129/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7309 - accuracy: 0.7010 - val_loss: 0.5923 - val_accuracy: 0.8311\n",
            "Epoch 130/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6916 - accuracy: 0.7399 - val_loss: 0.5589 - val_accuracy: 0.8311\n",
            "Epoch 131/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7137 - accuracy: 0.7044 - val_loss: 0.5594 - val_accuracy: 0.8581\n",
            "Epoch 132/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6820 - accuracy: 0.7551 - val_loss: 0.5543 - val_accuracy: 0.8108\n",
            "Epoch 133/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6832 - accuracy: 0.7297 - val_loss: 0.5671 - val_accuracy: 0.8378\n",
            "Epoch 134/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6942 - accuracy: 0.7416 - val_loss: 0.5488 - val_accuracy: 0.8311\n",
            "Epoch 135/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6793 - accuracy: 0.7314 - val_loss: 0.5303 - val_accuracy: 0.8716\n",
            "Epoch 136/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6942 - accuracy: 0.7264 - val_loss: 0.5675 - val_accuracy: 0.8649\n",
            "Epoch 137/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7160 - accuracy: 0.7044 - val_loss: 0.5711 - val_accuracy: 0.8649\n",
            "Epoch 138/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6876 - accuracy: 0.7230 - val_loss: 0.5523 - val_accuracy: 0.8649\n",
            "Epoch 139/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6730 - accuracy: 0.7416 - val_loss: 0.5458 - val_accuracy: 0.8581\n",
            "Epoch 140/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6971 - accuracy: 0.7280 - val_loss: 0.5750 - val_accuracy: 0.8311\n",
            "Epoch 141/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7074 - accuracy: 0.7027 - val_loss: 0.5581 - val_accuracy: 0.8581\n",
            "Epoch 142/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6612 - accuracy: 0.7584 - val_loss: 0.5195 - val_accuracy: 0.8716\n",
            "Epoch 143/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7002 - accuracy: 0.7196 - val_loss: 0.5765 - val_accuracy: 0.8649\n",
            "Epoch 144/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7010 - accuracy: 0.7128 - val_loss: 0.5658 - val_accuracy: 0.8311\n",
            "Epoch 145/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6671 - accuracy: 0.7483 - val_loss: 0.5239 - val_accuracy: 0.8784\n",
            "Epoch 146/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6861 - accuracy: 0.7416 - val_loss: 0.5581 - val_accuracy: 0.8716\n",
            "Epoch 147/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6845 - accuracy: 0.7297 - val_loss: 0.5471 - val_accuracy: 0.8243\n",
            "Epoch 148/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7078 - accuracy: 0.7010 - val_loss: 0.5549 - val_accuracy: 0.8784\n",
            "Epoch 149/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7046 - accuracy: 0.7111 - val_loss: 0.5587 - val_accuracy: 0.8514\n",
            "Epoch 150/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6808 - accuracy: 0.7382 - val_loss: 0.5345 - val_accuracy: 0.8716\n",
            "Epoch 151/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6869 - accuracy: 0.7635 - val_loss: 0.5995 - val_accuracy: 0.8108\n",
            "Epoch 152/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.7052 - accuracy: 0.7061 - val_loss: 0.5669 - val_accuracy: 0.8649\n",
            "Epoch 153/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6525 - accuracy: 0.7483 - val_loss: 0.5355 - val_accuracy: 0.8311\n",
            "Epoch 154/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6989 - accuracy: 0.7348 - val_loss: 0.5179 - val_accuracy: 0.8514\n",
            "Epoch 155/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6744 - accuracy: 0.7399 - val_loss: 0.5268 - val_accuracy: 0.8649\n",
            "Epoch 156/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6674 - accuracy: 0.7584 - val_loss: 0.5490 - val_accuracy: 0.8649\n",
            "Epoch 157/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6699 - accuracy: 0.7179 - val_loss: 0.5170 - val_accuracy: 0.8716\n",
            "Epoch 158/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6502 - accuracy: 0.7736 - val_loss: 0.5238 - val_accuracy: 0.8581\n",
            "Epoch 159/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7068 - accuracy: 0.7348 - val_loss: 0.5530 - val_accuracy: 0.8649\n",
            "Epoch 160/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6753 - accuracy: 0.7297 - val_loss: 0.5381 - val_accuracy: 0.8446\n",
            "Epoch 161/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7036 - accuracy: 0.7128 - val_loss: 0.5313 - val_accuracy: 0.8784\n",
            "Epoch 162/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6805 - accuracy: 0.7196 - val_loss: 0.5357 - val_accuracy: 0.8649\n",
            "Epoch 163/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6611 - accuracy: 0.7534 - val_loss: 0.5208 - val_accuracy: 0.8649\n",
            "Epoch 164/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6699 - accuracy: 0.7432 - val_loss: 0.5153 - val_accuracy: 0.8784\n",
            "Epoch 165/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6494 - accuracy: 0.7466 - val_loss: 0.4993 - val_accuracy: 0.8514\n",
            "Epoch 166/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6856 - accuracy: 0.7145 - val_loss: 0.5588 - val_accuracy: 0.8514\n",
            "Epoch 167/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6634 - accuracy: 0.7280 - val_loss: 0.5367 - val_accuracy: 0.8581\n",
            "Epoch 168/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6932 - accuracy: 0.7264 - val_loss: 0.5391 - val_accuracy: 0.8243\n",
            "Epoch 169/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6516 - accuracy: 0.7348 - val_loss: 0.5078 - val_accuracy: 0.8446\n",
            "Epoch 170/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6751 - accuracy: 0.7500 - val_loss: 0.5388 - val_accuracy: 0.8446\n",
            "Epoch 171/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6588 - accuracy: 0.7703 - val_loss: 0.5149 - val_accuracy: 0.8311\n",
            "Epoch 172/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6297 - accuracy: 0.7686 - val_loss: 0.5164 - val_accuracy: 0.8378\n",
            "Epoch 173/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6844 - accuracy: 0.7365 - val_loss: 0.4967 - val_accuracy: 0.8784\n",
            "Epoch 174/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6604 - accuracy: 0.7416 - val_loss: 0.4982 - val_accuracy: 0.8851\n",
            "Epoch 175/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6494 - accuracy: 0.7416 - val_loss: 0.5087 - val_accuracy: 0.8716\n",
            "Epoch 176/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6987 - accuracy: 0.7365 - val_loss: 0.5511 - val_accuracy: 0.8514\n",
            "Epoch 177/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6337 - accuracy: 0.7736 - val_loss: 0.5186 - val_accuracy: 0.8784\n",
            "Epoch 178/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6719 - accuracy: 0.7095 - val_loss: 0.5236 - val_accuracy: 0.8649\n",
            "Epoch 179/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6565 - accuracy: 0.7483 - val_loss: 0.5279 - val_accuracy: 0.8581\n",
            "Epoch 180/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6925 - accuracy: 0.7365 - val_loss: 0.5276 - val_accuracy: 0.8851\n",
            "Epoch 181/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6688 - accuracy: 0.7416 - val_loss: 0.5245 - val_accuracy: 0.8919\n",
            "Epoch 182/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7066 - accuracy: 0.7111 - val_loss: 0.5548 - val_accuracy: 0.8784\n",
            "Epoch 183/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6591 - accuracy: 0.7382 - val_loss: 0.5205 - val_accuracy: 0.8581\n",
            "Epoch 184/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6578 - accuracy: 0.7500 - val_loss: 0.5139 - val_accuracy: 0.8378\n",
            "Epoch 185/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6660 - accuracy: 0.7162 - val_loss: 0.4965 - val_accuracy: 0.8716\n",
            "Epoch 186/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6531 - accuracy: 0.7213 - val_loss: 0.5144 - val_accuracy: 0.8378\n",
            "Epoch 187/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6731 - accuracy: 0.7399 - val_loss: 0.5232 - val_accuracy: 0.8311\n",
            "Epoch 188/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6403 - accuracy: 0.7432 - val_loss: 0.5118 - val_accuracy: 0.8514\n",
            "Epoch 189/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6548 - accuracy: 0.7551 - val_loss: 0.5097 - val_accuracy: 0.8446\n",
            "Epoch 190/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6755 - accuracy: 0.7449 - val_loss: 0.5170 - val_accuracy: 0.8581\n",
            "Epoch 191/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6553 - accuracy: 0.7432 - val_loss: 0.5050 - val_accuracy: 0.8311\n",
            "Epoch 192/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6382 - accuracy: 0.7601 - val_loss: 0.5089 - val_accuracy: 0.8514\n",
            "Epoch 193/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6352 - accuracy: 0.7669 - val_loss: 0.4921 - val_accuracy: 0.8378\n",
            "Epoch 194/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6591 - accuracy: 0.7432 - val_loss: 0.5245 - val_accuracy: 0.8514\n",
            "Epoch 195/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6535 - accuracy: 0.7162 - val_loss: 0.5174 - val_accuracy: 0.8446\n",
            "Epoch 196/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.7020 - accuracy: 0.7280 - val_loss: 0.5461 - val_accuracy: 0.8581\n",
            "Epoch 197/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6725 - accuracy: 0.7517 - val_loss: 0.5221 - val_accuracy: 0.8514\n",
            "Epoch 198/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6390 - accuracy: 0.7348 - val_loss: 0.5068 - val_accuracy: 0.8649\n",
            "Epoch 199/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6596 - accuracy: 0.7399 - val_loss: 0.5002 - val_accuracy: 0.8514\n",
            "Epoch 200/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6145 - accuracy: 0.7787 - val_loss: 0.4880 - val_accuracy: 0.8716\n",
            "Epoch 201/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6296 - accuracy: 0.7584 - val_loss: 0.5053 - val_accuracy: 0.8514\n",
            "Epoch 202/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.7100 - accuracy: 0.7247 - val_loss: 0.5564 - val_accuracy: 0.8446\n",
            "Epoch 203/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6633 - accuracy: 0.7382 - val_loss: 0.5403 - val_accuracy: 0.8311\n",
            "Epoch 204/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6756 - accuracy: 0.7247 - val_loss: 0.5438 - val_accuracy: 0.8514\n",
            "Epoch 205/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6575 - accuracy: 0.7432 - val_loss: 0.5259 - val_accuracy: 0.8649\n",
            "Epoch 206/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6339 - accuracy: 0.7399 - val_loss: 0.5030 - val_accuracy: 0.8378\n",
            "Epoch 207/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6383 - accuracy: 0.7297 - val_loss: 0.4970 - val_accuracy: 0.8784\n",
            "Epoch 208/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6685 - accuracy: 0.7365 - val_loss: 0.5328 - val_accuracy: 0.8716\n",
            "Epoch 209/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6906 - accuracy: 0.7264 - val_loss: 0.5350 - val_accuracy: 0.8243\n",
            "Epoch 210/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6352 - accuracy: 0.7601 - val_loss: 0.5014 - val_accuracy: 0.8514\n",
            "Epoch 211/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6683 - accuracy: 0.7382 - val_loss: 0.5172 - val_accuracy: 0.8716\n",
            "Epoch 212/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6530 - accuracy: 0.7416 - val_loss: 0.5100 - val_accuracy: 0.8649\n",
            "Epoch 213/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6426 - accuracy: 0.7213 - val_loss: 0.5099 - val_accuracy: 0.8784\n",
            "Epoch 214/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6490 - accuracy: 0.7449 - val_loss: 0.4914 - val_accuracy: 0.8514\n",
            "Epoch 215/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6301 - accuracy: 0.7534 - val_loss: 0.4919 - val_accuracy: 0.8446\n",
            "Epoch 216/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6285 - accuracy: 0.7432 - val_loss: 0.4875 - val_accuracy: 0.8716\n",
            "Epoch 217/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6509 - accuracy: 0.7483 - val_loss: 0.4918 - val_accuracy: 0.8784\n",
            "Epoch 218/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6588 - accuracy: 0.7264 - val_loss: 0.5140 - val_accuracy: 0.8243\n",
            "Epoch 219/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6382 - accuracy: 0.7601 - val_loss: 0.5089 - val_accuracy: 0.8446\n",
            "Epoch 220/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6566 - accuracy: 0.7500 - val_loss: 0.5226 - val_accuracy: 0.8716\n",
            "Epoch 221/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6807 - accuracy: 0.7348 - val_loss: 0.5248 - val_accuracy: 0.8716\n",
            "Epoch 222/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6279 - accuracy: 0.7432 - val_loss: 0.4822 - val_accuracy: 0.8716\n",
            "Epoch 223/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6332 - accuracy: 0.7517 - val_loss: 0.4988 - val_accuracy: 0.8514\n",
            "Epoch 224/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6907 - accuracy: 0.6959 - val_loss: 0.5237 - val_accuracy: 0.8784\n",
            "Epoch 225/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6416 - accuracy: 0.7416 - val_loss: 0.4917 - val_accuracy: 0.8784\n",
            "Epoch 226/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6902 - accuracy: 0.7331 - val_loss: 0.5236 - val_accuracy: 0.8514\n",
            "Epoch 227/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6432 - accuracy: 0.7568 - val_loss: 0.4977 - val_accuracy: 0.8514\n",
            "Epoch 228/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6669 - accuracy: 0.7111 - val_loss: 0.4933 - val_accuracy: 0.8446\n",
            "Epoch 229/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5942 - accuracy: 0.7753 - val_loss: 0.4679 - val_accuracy: 0.8311\n",
            "Epoch 230/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6396 - accuracy: 0.7534 - val_loss: 0.5042 - val_accuracy: 0.8243\n",
            "Epoch 231/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6784 - accuracy: 0.7517 - val_loss: 0.5358 - val_accuracy: 0.8649\n",
            "Epoch 232/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6619 - accuracy: 0.7568 - val_loss: 0.5276 - val_accuracy: 0.8716\n",
            "Epoch 233/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6807 - accuracy: 0.7382 - val_loss: 0.5045 - val_accuracy: 0.8649\n",
            "Epoch 234/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6683 - accuracy: 0.7500 - val_loss: 0.5070 - val_accuracy: 0.8784\n",
            "Epoch 235/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6502 - accuracy: 0.7466 - val_loss: 0.4997 - val_accuracy: 0.8649\n",
            "Epoch 236/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6758 - accuracy: 0.7500 - val_loss: 0.5110 - val_accuracy: 0.8581\n",
            "Epoch 237/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6607 - accuracy: 0.7618 - val_loss: 0.5053 - val_accuracy: 0.8649\n",
            "Epoch 238/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6424 - accuracy: 0.7534 - val_loss: 0.5036 - val_accuracy: 0.8176\n",
            "Epoch 239/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6951 - accuracy: 0.6976 - val_loss: 0.5401 - val_accuracy: 0.8311\n",
            "Epoch 240/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6214 - accuracy: 0.7280 - val_loss: 0.5041 - val_accuracy: 0.8108\n",
            "Epoch 241/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6667 - accuracy: 0.7128 - val_loss: 0.5408 - val_accuracy: 0.8514\n",
            "Epoch 242/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6634 - accuracy: 0.7247 - val_loss: 0.5271 - val_accuracy: 0.8514\n",
            "Epoch 243/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6226 - accuracy: 0.7534 - val_loss: 0.5068 - val_accuracy: 0.8514\n",
            "Epoch 244/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6216 - accuracy: 0.7399 - val_loss: 0.4893 - val_accuracy: 0.8378\n",
            "Epoch 245/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6661 - accuracy: 0.7449 - val_loss: 0.5454 - val_accuracy: 0.8581\n",
            "Epoch 246/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6531 - accuracy: 0.7348 - val_loss: 0.5117 - val_accuracy: 0.8514\n",
            "Epoch 247/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6468 - accuracy: 0.7500 - val_loss: 0.5142 - val_accuracy: 0.8311\n",
            "Epoch 248/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6342 - accuracy: 0.7331 - val_loss: 0.5105 - val_accuracy: 0.8243\n",
            "Epoch 249/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6600 - accuracy: 0.7247 - val_loss: 0.5199 - val_accuracy: 0.8041\n",
            "Epoch 250/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6245 - accuracy: 0.7551 - val_loss: 0.5026 - val_accuracy: 0.8378\n",
            "Epoch 251/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6397 - accuracy: 0.7500 - val_loss: 0.4871 - val_accuracy: 0.8649\n",
            "Epoch 252/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6681 - accuracy: 0.7331 - val_loss: 0.5167 - val_accuracy: 0.8378\n",
            "Epoch 253/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6435 - accuracy: 0.7466 - val_loss: 0.4978 - val_accuracy: 0.8581\n",
            "Epoch 254/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6281 - accuracy: 0.7551 - val_loss: 0.5054 - val_accuracy: 0.8311\n",
            "Epoch 255/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6488 - accuracy: 0.7280 - val_loss: 0.4891 - val_accuracy: 0.8581\n",
            "Epoch 256/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6388 - accuracy: 0.7652 - val_loss: 0.5035 - val_accuracy: 0.8446\n",
            "Epoch 257/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6672 - accuracy: 0.7618 - val_loss: 0.5127 - val_accuracy: 0.8378\n",
            "Epoch 258/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6479 - accuracy: 0.7517 - val_loss: 0.5204 - val_accuracy: 0.8446\n",
            "Epoch 259/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6512 - accuracy: 0.7551 - val_loss: 0.5149 - val_accuracy: 0.8514\n",
            "Epoch 260/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6535 - accuracy: 0.7449 - val_loss: 0.5222 - val_accuracy: 0.8514\n",
            "Epoch 261/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6530 - accuracy: 0.7365 - val_loss: 0.5038 - val_accuracy: 0.8581\n",
            "Epoch 262/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6572 - accuracy: 0.7432 - val_loss: 0.5034 - val_accuracy: 0.8378\n",
            "Epoch 263/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6789 - accuracy: 0.7466 - val_loss: 0.5230 - val_accuracy: 0.8581\n",
            "Epoch 264/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6705 - accuracy: 0.7280 - val_loss: 0.5303 - val_accuracy: 0.8581\n",
            "Epoch 265/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6266 - accuracy: 0.7399 - val_loss: 0.5092 - val_accuracy: 0.8514\n",
            "Epoch 266/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6702 - accuracy: 0.7264 - val_loss: 0.5065 - val_accuracy: 0.8514\n",
            "Epoch 267/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6731 - accuracy: 0.7162 - val_loss: 0.5318 - val_accuracy: 0.8514\n",
            "Epoch 268/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6331 - accuracy: 0.7466 - val_loss: 0.5068 - val_accuracy: 0.8514\n",
            "Epoch 269/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6283 - accuracy: 0.7500 - val_loss: 0.4925 - val_accuracy: 0.8581\n",
            "Epoch 270/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6353 - accuracy: 0.7280 - val_loss: 0.4734 - val_accuracy: 0.8514\n",
            "Epoch 271/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6379 - accuracy: 0.7736 - val_loss: 0.4793 - val_accuracy: 0.8514\n",
            "Epoch 272/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6403 - accuracy: 0.7736 - val_loss: 0.5157 - val_accuracy: 0.8446\n",
            "Epoch 273/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6847 - accuracy: 0.7365 - val_loss: 0.5622 - val_accuracy: 0.7973\n",
            "Epoch 274/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6487 - accuracy: 0.7449 - val_loss: 0.5243 - val_accuracy: 0.8311\n",
            "Epoch 275/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6787 - accuracy: 0.7162 - val_loss: 0.5035 - val_accuracy: 0.8378\n",
            "Epoch 276/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6520 - accuracy: 0.7264 - val_loss: 0.5224 - val_accuracy: 0.8581\n",
            "Epoch 277/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6343 - accuracy: 0.7213 - val_loss: 0.4973 - val_accuracy: 0.8311\n",
            "Epoch 278/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6402 - accuracy: 0.7382 - val_loss: 0.5179 - val_accuracy: 0.8108\n",
            "Epoch 279/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6670 - accuracy: 0.7331 - val_loss: 0.5354 - val_accuracy: 0.8108\n",
            "Epoch 280/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6653 - accuracy: 0.7264 - val_loss: 0.5192 - val_accuracy: 0.8311\n",
            "Epoch 281/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6513 - accuracy: 0.7517 - val_loss: 0.4993 - val_accuracy: 0.8108\n",
            "Epoch 282/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6378 - accuracy: 0.7162 - val_loss: 0.5035 - val_accuracy: 0.8311\n",
            "Epoch 283/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6397 - accuracy: 0.7264 - val_loss: 0.4891 - val_accuracy: 0.8514\n",
            "Epoch 284/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6121 - accuracy: 0.7686 - val_loss: 0.4969 - val_accuracy: 0.8243\n",
            "Epoch 285/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6329 - accuracy: 0.7466 - val_loss: 0.5022 - val_accuracy: 0.8311\n",
            "Epoch 286/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6050 - accuracy: 0.7635 - val_loss: 0.4925 - val_accuracy: 0.8514\n",
            "Epoch 287/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6796 - accuracy: 0.7331 - val_loss: 0.5204 - val_accuracy: 0.8311\n",
            "Epoch 288/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6099 - accuracy: 0.7669 - val_loss: 0.4826 - val_accuracy: 0.8311\n",
            "Epoch 289/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6624 - accuracy: 0.7365 - val_loss: 0.4983 - val_accuracy: 0.8311\n",
            "Epoch 290/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6602 - accuracy: 0.7280 - val_loss: 0.4993 - val_accuracy: 0.8446\n",
            "Epoch 291/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6265 - accuracy: 0.7416 - val_loss: 0.4917 - val_accuracy: 0.8378\n",
            "Epoch 292/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6369 - accuracy: 0.7483 - val_loss: 0.4894 - val_accuracy: 0.8311\n",
            "Epoch 293/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6345 - accuracy: 0.7618 - val_loss: 0.4799 - val_accuracy: 0.8243\n",
            "Epoch 294/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6382 - accuracy: 0.7348 - val_loss: 0.4924 - val_accuracy: 0.8311\n",
            "Epoch 295/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6298 - accuracy: 0.7348 - val_loss: 0.4821 - val_accuracy: 0.8514\n",
            "Epoch 296/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6455 - accuracy: 0.7348 - val_loss: 0.4994 - val_accuracy: 0.8311\n",
            "Epoch 297/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6483 - accuracy: 0.7382 - val_loss: 0.5060 - val_accuracy: 0.8243\n",
            "Epoch 298/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6429 - accuracy: 0.7551 - val_loss: 0.4940 - val_accuracy: 0.8514\n",
            "Epoch 299/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6323 - accuracy: 0.7365 - val_loss: 0.4740 - val_accuracy: 0.8514\n",
            "Epoch 300/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6533 - accuracy: 0.7280 - val_loss: 0.5072 - val_accuracy: 0.8446\n",
            "Epoch 301/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5888 - accuracy: 0.7635 - val_loss: 0.4788 - val_accuracy: 0.8311\n",
            "Epoch 302/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6140 - accuracy: 0.7568 - val_loss: 0.4817 - val_accuracy: 0.8311\n",
            "Epoch 303/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6169 - accuracy: 0.7753 - val_loss: 0.4829 - val_accuracy: 0.8378\n",
            "Epoch 304/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6514 - accuracy: 0.7331 - val_loss: 0.4945 - val_accuracy: 0.8514\n",
            "Epoch 305/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6174 - accuracy: 0.7331 - val_loss: 0.4880 - val_accuracy: 0.8649\n",
            "Epoch 306/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6194 - accuracy: 0.7416 - val_loss: 0.4848 - val_accuracy: 0.8446\n",
            "Epoch 307/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6473 - accuracy: 0.7517 - val_loss: 0.4873 - val_accuracy: 0.8446\n",
            "Epoch 308/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6280 - accuracy: 0.7568 - val_loss: 0.4851 - val_accuracy: 0.8649\n",
            "Epoch 309/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6151 - accuracy: 0.7669 - val_loss: 0.4797 - val_accuracy: 0.8446\n",
            "Epoch 310/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6201 - accuracy: 0.7534 - val_loss: 0.4803 - val_accuracy: 0.8311\n",
            "Epoch 311/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6450 - accuracy: 0.7196 - val_loss: 0.5120 - val_accuracy: 0.8446\n",
            "Epoch 312/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6598 - accuracy: 0.7314 - val_loss: 0.5131 - val_accuracy: 0.8649\n",
            "Epoch 313/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6528 - accuracy: 0.7466 - val_loss: 0.5067 - val_accuracy: 0.8514\n",
            "Epoch 314/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6280 - accuracy: 0.7399 - val_loss: 0.4831 - val_accuracy: 0.8649\n",
            "Epoch 315/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6258 - accuracy: 0.7500 - val_loss: 0.4785 - val_accuracy: 0.8446\n",
            "Epoch 316/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5751 - accuracy: 0.7703 - val_loss: 0.4537 - val_accuracy: 0.8243\n",
            "Epoch 317/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6417 - accuracy: 0.7297 - val_loss: 0.4960 - val_accuracy: 0.8243\n",
            "Epoch 318/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6453 - accuracy: 0.7500 - val_loss: 0.4949 - val_accuracy: 0.8514\n",
            "Epoch 319/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6484 - accuracy: 0.7466 - val_loss: 0.5008 - val_accuracy: 0.8581\n",
            "Epoch 320/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6494 - accuracy: 0.7297 - val_loss: 0.5025 - val_accuracy: 0.8581\n",
            "Epoch 321/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6150 - accuracy: 0.7551 - val_loss: 0.4691 - val_accuracy: 0.8514\n",
            "Epoch 322/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6513 - accuracy: 0.7500 - val_loss: 0.4969 - val_accuracy: 0.8514\n",
            "Epoch 323/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6145 - accuracy: 0.7500 - val_loss: 0.4642 - val_accuracy: 0.8581\n",
            "Epoch 324/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5883 - accuracy: 0.7551 - val_loss: 0.4565 - val_accuracy: 0.8649\n",
            "Epoch 325/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6590 - accuracy: 0.7230 - val_loss: 0.5046 - val_accuracy: 0.8041\n",
            "Epoch 326/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6518 - accuracy: 0.7348 - val_loss: 0.4881 - val_accuracy: 0.8581\n",
            "Epoch 327/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6613 - accuracy: 0.7382 - val_loss: 0.4990 - val_accuracy: 0.8514\n",
            "Epoch 328/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6470 - accuracy: 0.7111 - val_loss: 0.4789 - val_accuracy: 0.8378\n",
            "Epoch 329/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6255 - accuracy: 0.7382 - val_loss: 0.4591 - val_accuracy: 0.8716\n",
            "Epoch 330/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5870 - accuracy: 0.7551 - val_loss: 0.4725 - val_accuracy: 0.8581\n",
            "Epoch 331/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6185 - accuracy: 0.7517 - val_loss: 0.4512 - val_accuracy: 0.8649\n",
            "Epoch 332/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6554 - accuracy: 0.7348 - val_loss: 0.5019 - val_accuracy: 0.8649\n",
            "Epoch 333/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6540 - accuracy: 0.7331 - val_loss: 0.4991 - val_accuracy: 0.8446\n",
            "Epoch 334/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6795 - accuracy: 0.7196 - val_loss: 0.5254 - val_accuracy: 0.8581\n",
            "Epoch 335/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6017 - accuracy: 0.7483 - val_loss: 0.4601 - val_accuracy: 0.8649\n",
            "Epoch 336/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6076 - accuracy: 0.7584 - val_loss: 0.4691 - val_accuracy: 0.8581\n",
            "Epoch 337/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6490 - accuracy: 0.7162 - val_loss: 0.5178 - val_accuracy: 0.8514\n",
            "Epoch 338/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6199 - accuracy: 0.7365 - val_loss: 0.4867 - val_accuracy: 0.8514\n",
            "Epoch 339/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6471 - accuracy: 0.7466 - val_loss: 0.5030 - val_accuracy: 0.8446\n",
            "Epoch 340/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6101 - accuracy: 0.7517 - val_loss: 0.4509 - val_accuracy: 0.8581\n",
            "Epoch 341/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5556 - accuracy: 0.7821 - val_loss: 0.4468 - val_accuracy: 0.8446\n",
            "Epoch 342/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6107 - accuracy: 0.7720 - val_loss: 0.4699 - val_accuracy: 0.8446\n",
            "Epoch 343/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6409 - accuracy: 0.7466 - val_loss: 0.5142 - val_accuracy: 0.8446\n",
            "Epoch 344/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6408 - accuracy: 0.7297 - val_loss: 0.4828 - val_accuracy: 0.8716\n",
            "Epoch 345/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6149 - accuracy: 0.7382 - val_loss: 0.4590 - val_accuracy: 0.8716\n",
            "Epoch 346/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6159 - accuracy: 0.7500 - val_loss: 0.4687 - val_accuracy: 0.8446\n",
            "Epoch 347/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6064 - accuracy: 0.7568 - val_loss: 0.4839 - val_accuracy: 0.8649\n",
            "Epoch 348/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6876 - accuracy: 0.7111 - val_loss: 0.4807 - val_accuracy: 0.8649\n",
            "Epoch 349/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5922 - accuracy: 0.7534 - val_loss: 0.4612 - val_accuracy: 0.8514\n",
            "Epoch 350/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5824 - accuracy: 0.7584 - val_loss: 0.4597 - val_accuracy: 0.8514\n",
            "Epoch 351/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6597 - accuracy: 0.6909 - val_loss: 0.4987 - val_accuracy: 0.8378\n",
            "Epoch 352/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6218 - accuracy: 0.7601 - val_loss: 0.4566 - val_accuracy: 0.8649\n",
            "Epoch 353/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6427 - accuracy: 0.7348 - val_loss: 0.4916 - val_accuracy: 0.8581\n",
            "Epoch 354/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6138 - accuracy: 0.7483 - val_loss: 0.4936 - val_accuracy: 0.8378\n",
            "Epoch 355/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6355 - accuracy: 0.7382 - val_loss: 0.4886 - val_accuracy: 0.8446\n",
            "Epoch 356/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6124 - accuracy: 0.7534 - val_loss: 0.4811 - val_accuracy: 0.8716\n",
            "Epoch 357/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6305 - accuracy: 0.7348 - val_loss: 0.4865 - val_accuracy: 0.8649\n",
            "Epoch 358/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6253 - accuracy: 0.7568 - val_loss: 0.4875 - val_accuracy: 0.8716\n",
            "Epoch 359/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6345 - accuracy: 0.7314 - val_loss: 0.4638 - val_accuracy: 0.8716\n",
            "Epoch 360/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6146 - accuracy: 0.7466 - val_loss: 0.4706 - val_accuracy: 0.8716\n",
            "Epoch 361/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6503 - accuracy: 0.7517 - val_loss: 0.4992 - val_accuracy: 0.8378\n",
            "Epoch 362/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6487 - accuracy: 0.7145 - val_loss: 0.5058 - val_accuracy: 0.8446\n",
            "Epoch 363/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6110 - accuracy: 0.7466 - val_loss: 0.4817 - val_accuracy: 0.8581\n",
            "Epoch 364/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6246 - accuracy: 0.7601 - val_loss: 0.4918 - val_accuracy: 0.8311\n",
            "Epoch 365/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6205 - accuracy: 0.7416 - val_loss: 0.4687 - val_accuracy: 0.8514\n",
            "Epoch 366/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6056 - accuracy: 0.7314 - val_loss: 0.4681 - val_accuracy: 0.8581\n",
            "Epoch 367/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6019 - accuracy: 0.7686 - val_loss: 0.4504 - val_accuracy: 0.8784\n",
            "Epoch 368/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6227 - accuracy: 0.7517 - val_loss: 0.4629 - val_accuracy: 0.8176\n",
            "Epoch 369/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6162 - accuracy: 0.7652 - val_loss: 0.4702 - val_accuracy: 0.8446\n",
            "Epoch 370/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6011 - accuracy: 0.7466 - val_loss: 0.4601 - val_accuracy: 0.8378\n",
            "Epoch 371/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6493 - accuracy: 0.7230 - val_loss: 0.5114 - val_accuracy: 0.8311\n",
            "Epoch 372/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6418 - accuracy: 0.7145 - val_loss: 0.4959 - val_accuracy: 0.8378\n",
            "Epoch 373/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6314 - accuracy: 0.7449 - val_loss: 0.4835 - val_accuracy: 0.8311\n",
            "Epoch 374/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5951 - accuracy: 0.7534 - val_loss: 0.4724 - val_accuracy: 0.8378\n",
            "Epoch 375/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6365 - accuracy: 0.7314 - val_loss: 0.4927 - val_accuracy: 0.8378\n",
            "Epoch 376/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6013 - accuracy: 0.7568 - val_loss: 0.4579 - val_accuracy: 0.8446\n",
            "Epoch 377/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6367 - accuracy: 0.7382 - val_loss: 0.4737 - val_accuracy: 0.8446\n",
            "Epoch 378/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6786 - accuracy: 0.7162 - val_loss: 0.5120 - val_accuracy: 0.8311\n",
            "Epoch 379/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6050 - accuracy: 0.7449 - val_loss: 0.4761 - val_accuracy: 0.8378\n",
            "Epoch 380/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6540 - accuracy: 0.7314 - val_loss: 0.4850 - val_accuracy: 0.8649\n",
            "Epoch 381/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5799 - accuracy: 0.7736 - val_loss: 0.4471 - val_accuracy: 0.8514\n",
            "Epoch 382/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6478 - accuracy: 0.7348 - val_loss: 0.4705 - val_accuracy: 0.8311\n",
            "Epoch 383/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6079 - accuracy: 0.7618 - val_loss: 0.4581 - val_accuracy: 0.8378\n",
            "Epoch 384/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6450 - accuracy: 0.7568 - val_loss: 0.4833 - val_accuracy: 0.8514\n",
            "Epoch 385/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6053 - accuracy: 0.7787 - val_loss: 0.4832 - val_accuracy: 0.8649\n",
            "Epoch 386/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6127 - accuracy: 0.7635 - val_loss: 0.4686 - val_accuracy: 0.8446\n",
            "Epoch 387/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6443 - accuracy: 0.7669 - val_loss: 0.4855 - val_accuracy: 0.8378\n",
            "Epoch 388/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6493 - accuracy: 0.7213 - val_loss: 0.4740 - val_accuracy: 0.8514\n",
            "Epoch 389/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6214 - accuracy: 0.7601 - val_loss: 0.4649 - val_accuracy: 0.8581\n",
            "Epoch 390/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6155 - accuracy: 0.7652 - val_loss: 0.4532 - val_accuracy: 0.8514\n",
            "Epoch 391/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6100 - accuracy: 0.7584 - val_loss: 0.4564 - val_accuracy: 0.8446\n",
            "Epoch 392/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6157 - accuracy: 0.7432 - val_loss: 0.4645 - val_accuracy: 0.8649\n",
            "Epoch 393/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6050 - accuracy: 0.7703 - val_loss: 0.4699 - val_accuracy: 0.8514\n",
            "Epoch 394/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5937 - accuracy: 0.7889 - val_loss: 0.4567 - val_accuracy: 0.8514\n",
            "Epoch 395/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6266 - accuracy: 0.7466 - val_loss: 0.4669 - val_accuracy: 0.8446\n",
            "Epoch 396/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5702 - accuracy: 0.7804 - val_loss: 0.4556 - val_accuracy: 0.8514\n",
            "Epoch 397/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6088 - accuracy: 0.7584 - val_loss: 0.4676 - val_accuracy: 0.8581\n",
            "Epoch 398/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6216 - accuracy: 0.7618 - val_loss: 0.4624 - val_accuracy: 0.8514\n",
            "Epoch 399/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6221 - accuracy: 0.7449 - val_loss: 0.4758 - val_accuracy: 0.8514\n",
            "Epoch 400/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6174 - accuracy: 0.7399 - val_loss: 0.4977 - val_accuracy: 0.8446\n",
            "Epoch 401/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6193 - accuracy: 0.7483 - val_loss: 0.4712 - val_accuracy: 0.8446\n",
            "Epoch 402/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6011 - accuracy: 0.7669 - val_loss: 0.4633 - val_accuracy: 0.8514\n",
            "Epoch 403/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6397 - accuracy: 0.7213 - val_loss: 0.4809 - val_accuracy: 0.8581\n",
            "Epoch 404/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5959 - accuracy: 0.7838 - val_loss: 0.4879 - val_accuracy: 0.8446\n",
            "Epoch 405/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6114 - accuracy: 0.7348 - val_loss: 0.4619 - val_accuracy: 0.8446\n",
            "Epoch 406/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6314 - accuracy: 0.7601 - val_loss: 0.4758 - val_accuracy: 0.8378\n",
            "Epoch 407/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6174 - accuracy: 0.7517 - val_loss: 0.4568 - val_accuracy: 0.8243\n",
            "Epoch 408/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6264 - accuracy: 0.7399 - val_loss: 0.4692 - val_accuracy: 0.8176\n",
            "Epoch 409/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5872 - accuracy: 0.7534 - val_loss: 0.4523 - val_accuracy: 0.8378\n",
            "Epoch 410/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6134 - accuracy: 0.7432 - val_loss: 0.4768 - val_accuracy: 0.8176\n",
            "Epoch 411/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5870 - accuracy: 0.7753 - val_loss: 0.4630 - val_accuracy: 0.8378\n",
            "Epoch 412/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5867 - accuracy: 0.7652 - val_loss: 0.4952 - val_accuracy: 0.8514\n",
            "Epoch 413/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6113 - accuracy: 0.7399 - val_loss: 0.4740 - val_accuracy: 0.8378\n",
            "Epoch 414/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6001 - accuracy: 0.7534 - val_loss: 0.4574 - val_accuracy: 0.8378\n",
            "Epoch 415/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6380 - accuracy: 0.7449 - val_loss: 0.4698 - val_accuracy: 0.8514\n",
            "Epoch 416/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6331 - accuracy: 0.7314 - val_loss: 0.4818 - val_accuracy: 0.8378\n",
            "Epoch 417/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6313 - accuracy: 0.7111 - val_loss: 0.4892 - val_accuracy: 0.8378\n",
            "Epoch 418/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6258 - accuracy: 0.7247 - val_loss: 0.4782 - val_accuracy: 0.8108\n",
            "Epoch 419/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5927 - accuracy: 0.7652 - val_loss: 0.4580 - val_accuracy: 0.8311\n",
            "Epoch 420/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6124 - accuracy: 0.7314 - val_loss: 0.4658 - val_accuracy: 0.8581\n",
            "Epoch 421/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6354 - accuracy: 0.7365 - val_loss: 0.4820 - val_accuracy: 0.8514\n",
            "Epoch 422/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6024 - accuracy: 0.7382 - val_loss: 0.4417 - val_accuracy: 0.8514\n",
            "Epoch 423/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5926 - accuracy: 0.7551 - val_loss: 0.4413 - val_accuracy: 0.8581\n",
            "Epoch 424/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6094 - accuracy: 0.7500 - val_loss: 0.4855 - val_accuracy: 0.8243\n",
            "Epoch 425/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6578 - accuracy: 0.7128 - val_loss: 0.5088 - val_accuracy: 0.8446\n",
            "Epoch 426/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6317 - accuracy: 0.7095 - val_loss: 0.4958 - val_accuracy: 0.8378\n",
            "Epoch 427/500\n",
            "74/74 [==============================] - 1s 8ms/step - loss: 0.6133 - accuracy: 0.7432 - val_loss: 0.4555 - val_accuracy: 0.8378\n",
            "Epoch 428/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5736 - accuracy: 0.7770 - val_loss: 0.4425 - val_accuracy: 0.8581\n",
            "Epoch 429/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5948 - accuracy: 0.7669 - val_loss: 0.4619 - val_accuracy: 0.8446\n",
            "Epoch 430/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6141 - accuracy: 0.7466 - val_loss: 0.4608 - val_accuracy: 0.8446\n",
            "Epoch 431/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5510 - accuracy: 0.7889 - val_loss: 0.4309 - val_accuracy: 0.8649\n",
            "Epoch 432/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5931 - accuracy: 0.7686 - val_loss: 0.4607 - val_accuracy: 0.8378\n",
            "Epoch 433/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6352 - accuracy: 0.7111 - val_loss: 0.4886 - val_accuracy: 0.8378\n",
            "Epoch 434/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6284 - accuracy: 0.7297 - val_loss: 0.4980 - val_accuracy: 0.8108\n",
            "Epoch 435/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6153 - accuracy: 0.7584 - val_loss: 0.4911 - val_accuracy: 0.8243\n",
            "Epoch 436/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6310 - accuracy: 0.7618 - val_loss: 0.4879 - val_accuracy: 0.8378\n",
            "Epoch 437/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5907 - accuracy: 0.7517 - val_loss: 0.4734 - val_accuracy: 0.8378\n",
            "Epoch 438/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6060 - accuracy: 0.7703 - val_loss: 0.4657 - val_accuracy: 0.8378\n",
            "Epoch 439/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5983 - accuracy: 0.7703 - val_loss: 0.4561 - val_accuracy: 0.8311\n",
            "Epoch 440/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5857 - accuracy: 0.7466 - val_loss: 0.4354 - val_accuracy: 0.8311\n",
            "Epoch 441/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6052 - accuracy: 0.7568 - val_loss: 0.4429 - val_accuracy: 0.8243\n",
            "Epoch 442/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6297 - accuracy: 0.7551 - val_loss: 0.4774 - val_accuracy: 0.8378\n",
            "Epoch 443/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5985 - accuracy: 0.7652 - val_loss: 0.4622 - val_accuracy: 0.8378\n",
            "Epoch 444/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6015 - accuracy: 0.7483 - val_loss: 0.4575 - val_accuracy: 0.8446\n",
            "Epoch 445/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6260 - accuracy: 0.7399 - val_loss: 0.4582 - val_accuracy: 0.8649\n",
            "Epoch 446/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6369 - accuracy: 0.7517 - val_loss: 0.4773 - val_accuracy: 0.8378\n",
            "Epoch 447/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6146 - accuracy: 0.7314 - val_loss: 0.4600 - val_accuracy: 0.8378\n",
            "Epoch 448/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6171 - accuracy: 0.7686 - val_loss: 0.4697 - val_accuracy: 0.8446\n",
            "Epoch 449/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5932 - accuracy: 0.7568 - val_loss: 0.4547 - val_accuracy: 0.8581\n",
            "Epoch 450/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6134 - accuracy: 0.7230 - val_loss: 0.4601 - val_accuracy: 0.8243\n",
            "Epoch 451/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6147 - accuracy: 0.7196 - val_loss: 0.4764 - val_accuracy: 0.8311\n",
            "Epoch 452/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6213 - accuracy: 0.7432 - val_loss: 0.4967 - val_accuracy: 0.8378\n",
            "Epoch 453/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6567 - accuracy: 0.7162 - val_loss: 0.4986 - val_accuracy: 0.8378\n",
            "Epoch 454/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6044 - accuracy: 0.7500 - val_loss: 0.4786 - val_accuracy: 0.8243\n",
            "Epoch 455/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5998 - accuracy: 0.7517 - val_loss: 0.4379 - val_accuracy: 0.8446\n",
            "Epoch 456/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6203 - accuracy: 0.7365 - val_loss: 0.4754 - val_accuracy: 0.8514\n",
            "Epoch 457/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6204 - accuracy: 0.7432 - val_loss: 0.4716 - val_accuracy: 0.8514\n",
            "Epoch 458/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6146 - accuracy: 0.7179 - val_loss: 0.4450 - val_accuracy: 0.8311\n",
            "Epoch 459/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5708 - accuracy: 0.7601 - val_loss: 0.4248 - val_accuracy: 0.8446\n",
            "Epoch 460/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6115 - accuracy: 0.7348 - val_loss: 0.4459 - val_accuracy: 0.8176\n",
            "Epoch 461/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5837 - accuracy: 0.7703 - val_loss: 0.4447 - val_accuracy: 0.8243\n",
            "Epoch 462/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6168 - accuracy: 0.7382 - val_loss: 0.4398 - val_accuracy: 0.8649\n",
            "Epoch 463/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6137 - accuracy: 0.7314 - val_loss: 0.4500 - val_accuracy: 0.8784\n",
            "Epoch 464/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5982 - accuracy: 0.7568 - val_loss: 0.4525 - val_accuracy: 0.8649\n",
            "Epoch 465/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6243 - accuracy: 0.7264 - val_loss: 0.4707 - val_accuracy: 0.8311\n",
            "Epoch 466/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5803 - accuracy: 0.7770 - val_loss: 0.4592 - val_accuracy: 0.8446\n",
            "Epoch 467/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6172 - accuracy: 0.7399 - val_loss: 0.4823 - val_accuracy: 0.8581\n",
            "Epoch 468/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6075 - accuracy: 0.7534 - val_loss: 0.4566 - val_accuracy: 0.8514\n",
            "Epoch 469/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6566 - accuracy: 0.7213 - val_loss: 0.4992 - val_accuracy: 0.8784\n",
            "Epoch 470/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6346 - accuracy: 0.7399 - val_loss: 0.4947 - val_accuracy: 0.8649\n",
            "Epoch 471/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5734 - accuracy: 0.7736 - val_loss: 0.4479 - val_accuracy: 0.8649\n",
            "Epoch 472/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6653 - accuracy: 0.7179 - val_loss: 0.4947 - val_accuracy: 0.8311\n",
            "Epoch 473/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6288 - accuracy: 0.7618 - val_loss: 0.4745 - val_accuracy: 0.8378\n",
            "Epoch 474/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6104 - accuracy: 0.7466 - val_loss: 0.4630 - val_accuracy: 0.8581\n",
            "Epoch 475/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5929 - accuracy: 0.7770 - val_loss: 0.4584 - val_accuracy: 0.8581\n",
            "Epoch 476/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6080 - accuracy: 0.7449 - val_loss: 0.4493 - val_accuracy: 0.8446\n",
            "Epoch 477/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6366 - accuracy: 0.7264 - val_loss: 0.4949 - val_accuracy: 0.8514\n",
            "Epoch 478/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6093 - accuracy: 0.7449 - val_loss: 0.4607 - val_accuracy: 0.8581\n",
            "Epoch 479/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6112 - accuracy: 0.7416 - val_loss: 0.4633 - val_accuracy: 0.8378\n",
            "Epoch 480/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5984 - accuracy: 0.7500 - val_loss: 0.4583 - val_accuracy: 0.8446\n",
            "Epoch 481/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6055 - accuracy: 0.7855 - val_loss: 0.4575 - val_accuracy: 0.8378\n",
            "Epoch 482/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6009 - accuracy: 0.7466 - val_loss: 0.4654 - val_accuracy: 0.8378\n",
            "Epoch 483/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5755 - accuracy: 0.7534 - val_loss: 0.4559 - val_accuracy: 0.8378\n",
            "Epoch 484/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5863 - accuracy: 0.7449 - val_loss: 0.4578 - val_accuracy: 0.8311\n",
            "Epoch 485/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6171 - accuracy: 0.7466 - val_loss: 0.4560 - val_accuracy: 0.8243\n",
            "Epoch 486/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5792 - accuracy: 0.7720 - val_loss: 0.4117 - val_accuracy: 0.8378\n",
            "Epoch 487/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6324 - accuracy: 0.7483 - val_loss: 0.4848 - val_accuracy: 0.8581\n",
            "Epoch 488/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6070 - accuracy: 0.7568 - val_loss: 0.4418 - val_accuracy: 0.8649\n",
            "Epoch 489/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.6065 - accuracy: 0.7416 - val_loss: 0.4548 - val_accuracy: 0.8514\n",
            "Epoch 490/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5699 - accuracy: 0.7686 - val_loss: 0.4570 - val_accuracy: 0.8378\n",
            "Epoch 491/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6090 - accuracy: 0.7449 - val_loss: 0.4646 - val_accuracy: 0.8581\n",
            "Epoch 492/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6212 - accuracy: 0.7348 - val_loss: 0.4856 - val_accuracy: 0.8446\n",
            "Epoch 493/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5671 - accuracy: 0.7753 - val_loss: 0.4557 - val_accuracy: 0.8378\n",
            "Epoch 494/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5977 - accuracy: 0.7534 - val_loss: 0.4414 - val_accuracy: 0.8446\n",
            "Epoch 495/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.6067 - accuracy: 0.7500 - val_loss: 0.4687 - val_accuracy: 0.8851\n",
            "Epoch 496/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.6502 - accuracy: 0.7230 - val_loss: 0.4799 - val_accuracy: 0.8514\n",
            "Epoch 497/500\n",
            "74/74 [==============================] - 1s 7ms/step - loss: 0.5959 - accuracy: 0.7652 - val_loss: 0.4391 - val_accuracy: 0.8514\n",
            "Epoch 498/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5971 - accuracy: 0.7635 - val_loss: 0.4588 - val_accuracy: 0.8716\n",
            "Epoch 499/500\n",
            "74/74 [==============================] - 0s 6ms/step - loss: 0.5710 - accuracy: 0.7804 - val_loss: 0.4403 - val_accuracy: 0.8581\n",
            "Epoch 500/500\n",
            "74/74 [==============================] - 0s 7ms/step - loss: 0.5960 - accuracy: 0.7635 - val_loss: 0.4671 - val_accuracy: 0.8311\n",
            "5/5 [==============================] - 0s 3ms/step - loss: 0.4671 - accuracy: 0.8311\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:62: DeprecationWarning: scipy.interp is deprecated and will be removed in SciPy 2.0.0, use numpy.interp instead\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 864x864 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "[0.8108108043670654, 0.837837815284729, 0.8310810923576355, 0.837837815284729, 0.8310810923576355]\n",
            "Accuracy: 0.83\n",
            "Sensitivity: 0.8691\n",
            "Specificity: 0.7875\n",
            "MCC: 0.6580\n",
            "Precision: 0.8076\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "p5YD2gb0_Ktb",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}