Range_Doppler_LSTM_v2_datalarla_dropout01.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Range_Doppler_LSTM_v2_datalarla_dropout01.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "machine_shape": "hm",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/mertege/FMCW-Data-Classification-/blob/main/Range_Doppler_LSTM_v2_datalarla_dropout01.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kN4ixvH39HW6"
      },
      "source": [
        "# Dropout 0.1 secildi\n",
        "# 5 defa run edildi ve değerler aşağıda paylaşıldı.\n",
        "# Mean test accuracy is 0.974, mean test f1 score is 0.970, max test accuracy is 0.994, max test f1 score is 0.993, min test accuracy is 0.959, min test f1 score is 0.951, std of test accuracy is 0.012, std of test f1 score is 0.014\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lzJwN6GfkN8Z"
      },
      "source": [
        "from keras.models import Sequential\n",
        "from tensorflow.keras import layers\n",
        "from keras.models import Model\n",
        "from keras.layers import Activation, Dropout, Flatten, Dense, BatchNormalization, Normalization, Input, Conv2D, MaxPooling2D, Concatenate, GRU, LSTM, GRU, TimeDistributed, Bidirectional\n",
        "import tensorflow as tf\n",
        "import scipy.io\n",
        "import numpy as np\n",
        "import cv2\n",
        "import matplotlib.pyplot as plt\n",
        "import random\n",
        "from numpy.random import seed\n",
        "from sklearn.model_selection import KFold, StratifiedKFold\n",
        "import time\n",
        "from sklearn.metrics import precision_recall_fscore_support\n",
        "from keras.callbacks import EarlyStopping\n",
        "from numpy import array\n",
        "from numpy import argmax\n",
        "from sklearn.preprocessing import LabelEncoder\n",
        "from sklearn.preprocessing import OneHotEncoder\n",
        "from keras import backend as K \n",
        "import gc\n",
        "import os"
      ],
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8FERVnWrkrwE",
        "outputId": "66f675c1-801a-4385-ef04-1c2225f132b9"
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "o19JHieQ81cS"
      },
      "source": [
        "# Get Range-Doppler data from\n",
        "range_doppler_fast_resized = scipy.io.loadmat('/content/drive/MyDrive/data/range_doppler_fast_resizedv2.mat')\n",
        "range_doppler_fast_resized = range_doppler_fast_resized['range_doppler_fast_resized']\n",
        "range_doppler_fast_resized = np.transpose(range_doppler_fast_resized, (2, 0, 1))\n",
        "# range_doppler_fast_resized = np.delete(range_doppler_fast_resized,(49), axis=0) # 50th row is deleted since there is no 50th row in spectrogram fast data.\n",
        "range_doppler_fast_label = scipy.io.loadmat('/content/drive/MyDrive/data/range_doppler_fast_label.mat')\n",
        "range_doppler_fast_label = range_doppler_fast_label['range_doppler_fast_label']  \n",
        "\n",
        "range_doppler_slow_resized = scipy.io.loadmat('/content/drive/MyDrive/data/range_doppler_slow_resizedv2.mat')\n",
        "range_doppler_slow_resized = range_doppler_slow_resized['range_doppler_slow_resized']\n",
        "range_doppler_slow_resized = np.transpose(range_doppler_slow_resized, (2, 0, 1))\n",
        "range_doppler_slow_label = scipy.io.loadmat('/content/drive/MyDrive/data/range_doppler_slow_label.mat')\n",
        "range_doppler_slow_label = range_doppler_slow_label['range_doppler_slow_label']  \n",
        "\n",
        "range_doppler_slow_pocket_resized = scipy.io.loadmat('/content/drive/MyDrive/data/range_doppler_slow_pocket_resizedv2.mat')\n",
        "range_doppler_slow_pocket_resized = range_doppler_slow_pocket_resized['range_doppler_slow_pocket_resized']\n",
        "range_doppler_slow_pocket_resized = np.transpose(range_doppler_slow_pocket_resized, (2, 0, 1))\n",
        "range_doppler_pocket_label = scipy.io.loadmat('/content/drive/MyDrive/data/range_doppler_pocket_label.mat')\n",
        "range_doppler_pocket_label = range_doppler_pocket_label['range_doppler_pocket_label']  \n",
        "# Get Range-Doppler data from\n",
        "spectrogram_fast_resized = scipy.io.loadmat('/content/drive/MyDrive/data/spectrogram_fast_resized.mat')\n",
        "spectrogram_fast_resized = spectrogram_fast_resized['spectrogram_fast_resized']\n",
        "spectrogram_fast_resized = np.transpose(spectrogram_fast_resized, (2, 0, 1))\n",
        "spectrogram_fast_label = scipy.io.loadmat('/content/drive/MyDrive/data/spectrogram_fast_label.mat')\n",
        "spectrogram_fast_label = spectrogram_fast_label['spectrogram_fast_label']  \n",
        "\n",
        "spectrogram_slow_resized = scipy.io.loadmat('/content/drive/MyDrive/data/spectrogram_slow_resized.mat')\n",
        "spectrogram_slow_resized = spectrogram_slow_resized['spectrogram_slow_resized']\n",
        "spectrogram_slow_resized = np.transpose(spectrogram_slow_resized, (2, 0, 1))\n",
        "spectrogram_slow_label = scipy.io.loadmat('/content/drive/MyDrive/data/spectrogram_slow_label.mat')\n",
        "spectrogram_slow_label = spectrogram_slow_label['spectrogram_slow_label']  \n",
        "\n",
        "spectrogram_slow_pocket_resized = scipy.io.loadmat('/content/drive/MyDrive/data/spectrogram_slow_pocket_resized.mat')\n",
        "spectrogram_slow_pocket_resized = spectrogram_slow_pocket_resized['spectrogram_slow_pocket_resized']\n",
        "spectrogram_slow_pocket_resized = np.transpose(spectrogram_slow_pocket_resized, (2, 0, 1))\n",
        "spectrogram_slow_pocket_label = scipy.io.loadmat('/content/drive/MyDrive/data/spectrogram_slow_pocket_label.mat')\n",
        "spectrogram_slow_pocket_label = spectrogram_slow_pocket_label['spectrogram_slow_pocket_label']  "
      ],
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wFp_PErnMLN2"
      },
      "source": [
        "# Concat range-doppler data\n",
        "range_doppler_concat = np.concatenate((range_doppler_fast_resized,range_doppler_slow_resized),axis=0)\n",
        "range_doppler_concat = np.concatenate((range_doppler_concat,range_doppler_slow_pocket_resized),axis=0)\n",
        "range_doppler_concat = range_doppler_concat[:,:,:,np.newaxis] \n",
        "range_doppler_concat_label = np.zeros((range_doppler_concat.shape[0],1))\n",
        "range_doppler_concat_label[:range_doppler_fast_resized.shape[0],:] = 1\n",
        "# Shuffle concat range doppler\n",
        "shuffle_indx = random.sample(range(0, range_doppler_concat.shape[0]), range_doppler_concat.shape[0]) # split validation data\n",
        "range_doppler_concat_shuffle = range_doppler_concat[shuffle_indx,:,:,:]\n",
        "range_doppler_concat_label_shuffle = range_doppler_concat_label[shuffle_indx,:]\n",
        "# Concat range-doppler data\n",
        "spectrogram_concat = np.concatenate((spectrogram_fast_resized,spectrogram_slow_resized),axis=0)\n",
        "spectrogram_concat = np.concatenate((spectrogram_concat,spectrogram_slow_pocket_resized),axis=0)\n",
        "spectrogram_concat = spectrogram_concat[:,:,:,np.newaxis] \n",
        "spectrogram_concat_label = np.zeros((spectrogram_concat.shape[0],1))\n",
        "spectrogram_concat_label[:spectrogram_fast_resized.shape[0],:] = 1\n",
        "# Shuffle concat range doppler\n",
        "# shuffle_indx = random.sample(range(0, spectrogram_concat.shape[0]), spectrogram_concat.shape[0]) # split validation data\n",
        "spectrogram_concat_shuffle = spectrogram_concat[shuffle_indx,:,:,:]\n",
        "spectrogram_concat_label_shuffle = spectrogram_concat_label[shuffle_indx,:]"
      ],
      "execution_count": 5,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_1aBNkjoLCvI"
      },
      "source": [
        "# ---------------- Augmente and shuffle (train and test) data data ----------------\n",
        "data_augmentation = tf.keras.Sequential([\n",
        "  layers.RandomFlip(\"horizontal_and_vertical\"),\n",
        "  layers.RandomRotation(0.1),\n",
        "])\n",
        "\n",
        "\n",
        "def mixup_augmentation(images,range_doppler_training_data, labels, repeat_of_mixup, alpha=0.2):\n",
        "    batch_size = images.shape[0]\n",
        "    concat_images = np.zeros((batch_size*(repeat_of_mixup+1),images.shape[1],images.shape[2],images.shape[3]))\n",
        "    concat_images_range_doppler = np.zeros((batch_size*(repeat_of_mixup+1),range_doppler_training_data.shape[1],\\\n",
        "                                            range_doppler_training_data.shape[2],range_doppler_training_data.shape[3]))\n",
        "    concat_label = np.zeros((batch_size*(repeat_of_mixup+1),labels.shape[1]))\n",
        "    for ii in range(repeat_of_mixup):\n",
        "      # shuffle train dataset\n",
        "      shuffle_indx_1 = random.sample(range(0, images.shape[0]), images.shape[0]) # split validation data\n",
        "      images_shuffled_1 = images[shuffle_indx_1,:,:,:]\n",
        "      range_doppler_training_data_shuffled_1 = range_doppler_training_data[shuffle_indx_1,:,:,:]\n",
        "      labels_shuffled_1 = labels[shuffle_indx_1,:]\n",
        "\n",
        "      shuffle_indx_2 = random.sample(range(0, images.shape[0]), images.shape[0]) # split validation data\n",
        "      images_shuffled_2 = images[shuffle_indx_2,:,:,:]\n",
        "      range_doppler_training_data_shuffled_2 = range_doppler_training_data[shuffle_indx_2,:,:,:]\n",
        "      labels_shuffled_2 = labels[shuffle_indx_2,:]\n",
        "\n",
        "      # Sample lambda and reshape it to do the mixup\n",
        "      gaussian_mean = 0.2\n",
        "      gaussian_std = 0.02\n",
        "      ll = np.random.normal(gaussian_mean, gaussian_std, (batch_size,1,1,1))\n",
        "      x_l = np.reshape(ll, (batch_size,1,1,1))\n",
        "      y_l = np.reshape(ll, (batch_size,1))\n",
        "      \n",
        "      # Perform mixup on both images and labels by combining a pair of images/labels\n",
        "      images_mixup = images_shuffled_1 * x_l + images_shuffled_2 * (1 - x_l)\n",
        "      images_mixup_range_doppler = range_doppler_training_data_shuffled_1 * x_l + range_doppler_training_data_shuffled_2 * (1 - x_l)\n",
        "      labels_mixup = labels_shuffled_1 * y_l + labels_shuffled_2 * (1 - y_l)\n",
        "      concat_images[ii*batch_size:(ii+1)*batch_size,:,:,:] = images_mixup\n",
        "      concat_images_range_doppler[ii*batch_size:(ii+1)*batch_size,:,:,:] = images_mixup_range_doppler\n",
        "      concat_label[ii*batch_size:(ii+1)*batch_size,:] = labels_mixup\n",
        "\n",
        "    concat_images[repeat_of_mixup*batch_size:,:,:,:] = images\n",
        "    concat_images_range_doppler[repeat_of_mixup*batch_size:,:,:,:] = range_doppler_training_data\n",
        "    concat_label[repeat_of_mixup*batch_size:,:] = labels\n",
        "    return (concat_images,concat_images_range_doppler, concat_label)\n",
        "def split_and_augmentation_of_training(spectrogram_concat_shuffle_train,range_doppler_concat_shuffle_train,range_doppler_concat_label_shuffle_train,\\\n",
        "                                       repeat_of_mixup, augmentation_enable):\n",
        "  # ---------------- Parameters ----------------\n",
        "  repeat_of_augmentation_for_fast = 1\n",
        "  repeat_of_augmentation_for_slow = np.floor(repeat_of_augmentation_for_fast/2)\n",
        "  repeat_of_augmentation_for_slow = int(repeat_of_augmentation_for_slow)\n",
        "  # size_of_validation = 30\n",
        "  alpha = 0.2\n",
        "  dummy_label = np.zeros((spectrogram_concat_shuffle_train.shape[0],1))\n",
        "  for randomlist_for_train_indx, randomlist_for_validation_indx in kfold.split(spectrogram_concat_shuffle_train,dummy_label):   \n",
        "    randomlist_for_validation_indx\n",
        "  # Split validation\n",
        "  # randomlist_for_validation_indx = random.sample(range(0, range_doppler_concat_shuffle_train.shape[0]), size_of_validation) # split validation data\n",
        "  # randomlist_for_train_indx = np.delete(range(0, range_doppler_concat_shuffle_train.shape[0]), randomlist_for_validation_indx) # split training data\n",
        "  # get validation data\n",
        "  spectrogram_validation_data = spectrogram_concat_shuffle_train[randomlist_for_validation_indx,:,:,:]\n",
        "  spectrogram_validation_labels = range_doppler_concat_label_shuffle_train[randomlist_for_validation_indx,:]\n",
        "  range_doppler_validation_data = range_doppler_concat_shuffle_train[randomlist_for_validation_indx,:,:,:]\n",
        "  # get training data\n",
        "  spectrogram_training_data = spectrogram_concat_shuffle_train[randomlist_for_train_indx,:,:,:]\n",
        "  spectrogram_training_labels = spectrogram_concat_label_shuffle_train[randomlist_for_train_indx,:]\n",
        "  range_doppler_training_data = range_doppler_concat_shuffle_train[randomlist_for_train_indx,:,:,:]\n",
        "\n",
        "  # Rotate Augmentation\n",
        "  # get slow and fast indexes of training data\n",
        "  slow_indexes = np.where(spectrogram_training_labels == 0)[0]\n",
        "  fast_indexes = np.delete(range(0, spectrogram_training_labels.shape[0]), slow_indexes)  \n",
        "\n",
        "  slow_spectrograms_train = spectrogram_training_data[slow_indexes,:,:,:]\n",
        "  size_of_samples_slow = slow_spectrograms_train.shape[0]\n",
        "\n",
        "  fast_spectrograms_train = spectrogram_training_data[fast_indexes,:,:,:]  \n",
        "  size_of_samples_fast = fast_spectrograms_train.shape[0]\n",
        "\n",
        "  slow_range_train = range_doppler_training_data[slow_indexes,:,:,:]\n",
        "  fast_range_train = range_doppler_training_data[fast_indexes,:,:,:]  \n",
        "\n",
        "  if augmentation_enable == True: \n",
        "    # ---------------- Augmente Train Data for Fast ----------------\n",
        "    augmented_image_fast = np.zeros((size_of_samples_fast*repeat_of_augmentation_for_fast,fast_spectrograms_train.shape[1],fast_spectrograms_train.shape[2],1))\n",
        "    spectrograms_fast_label = np.ones((size_of_samples_fast*(repeat_of_augmentation_for_fast+1),1))\n",
        "    # augmented_image_fast = np.flip(fast_spectrograms_train,axis=2)\n",
        "    for jj in range(repeat_of_augmentation_for_fast):\n",
        "      for ii in range(size_of_samples_fast):\n",
        "        augmented_image_fast[size_of_samples_fast*jj+ii,:,:,:] = data_augmentation(fast_spectrograms_train[ii,:,:,:])\n",
        "    augmented_image_fast = np.concatenate((augmented_image_fast,fast_spectrograms_train),axis=0)   \n",
        "\n",
        "    augmented_image_fast_range = np.zeros((size_of_samples_fast*repeat_of_augmentation_for_fast,fast_range_train.shape[1],fast_range_train.shape[2],1))\n",
        "    # augmented_image_fast_range  = np.flip(fast_range_train,axis=2)\n",
        "    for jj in range(repeat_of_augmentation_for_fast):\n",
        "      for ii in range(size_of_samples_fast):\n",
        "        augmented_image_fast_range[size_of_samples_fast*jj+ii,:,:,:]  =data_augmentation(fast_range_train[ii,:,:,:])\n",
        "    augmented_image_fast_range = np.concatenate((augmented_image_fast_range,fast_range_train),axis=0)   \n",
        "\n",
        "    # ---------------- Augmente Train Data for Slow ----------------\n",
        "    augmented_image_slow = slow_spectrograms_train\n",
        "    augmented_image_slow_range = slow_range_train\n",
        "\n",
        "  else:\n",
        "    augmented_image_fast = fast_spectrograms_train\n",
        "    augmented_image_slow = slow_spectrograms_train\n",
        "    augmented_image_fast_range = fast_range_train\n",
        "    augmented_image_slow_range = slow_range_train\n",
        "    spectrograms_fast_label = np.ones((size_of_samples_fast,1))\n",
        "    \n",
        "  spectrograms_slow_label = np.zeros((size_of_samples_slow,1))\n",
        "\n",
        "  spectrogram_training_data = np.concatenate((augmented_image_fast,augmented_image_slow),axis=0)\n",
        "  range_doppler_training_data = np.concatenate((augmented_image_fast_range,augmented_image_slow_range),axis=0)\n",
        "  spectrogram_training_labels = np.concatenate((spectrograms_fast_label,spectrograms_slow_label),axis=0)\n",
        "\n",
        "  (spectrogram_augmented_image,range_doppler_augmented_image,spectrograms_label)=\\\n",
        "   mixup_augmentation(spectrogram_training_data,range_doppler_training_data, spectrogram_training_labels, repeat_of_mixup, alpha=0.2)\n",
        "\n",
        "  return (spectrogram_augmented_image,range_doppler_augmented_image,spectrograms_label,\\\n",
        "     spectrogram_validation_data,range_doppler_validation_data, spectrogram_validation_labels)\n",
        "def normalize_inputs(range_doppler_concat_shuffle, normalize_inputs_enable):\n",
        "  # ---------------- Normalize Inputs ----------------\n",
        "  if normalize_inputs_enable == True:\n",
        "    layer = Normalization(axis=None)\n",
        "    layer.adapt(range_doppler_concat_shuffle)\n",
        "    range_doppler_concat_shuffle = layer(range_doppler_concat_shuffle)\n",
        "  else:\n",
        "    range_doppler_concat_shuffle = range_doppler_concat_shuffle\n",
        "  return(range_doppler_concat_shuffle)"
      ],
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8CUvxggth7ZQ"
      },
      "source": [
        "normalize_inputs_enable = 1\n",
        "range_doppler_concat_shuffle = normalize_inputs(range_doppler_concat_shuffle, normalize_inputs_enable)\n",
        "spectrogram_concat_shuffle = normalize_inputs(spectrogram_concat_shuffle, normalize_inputs_enable)\n",
        "range_doppler_concat_shuffle = np.float32(range_doppler_concat_shuffle)\n",
        "spectrogram_concat_shuffle = np.float32(spectrogram_concat_shuffle)\n"
      ],
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "n_features = range_doppler_concat_shuffle.shape[1]\n",
        "n_steps = range_doppler_concat_shuffle.shape[2]\n",
        "range_doppler_concat_shuffle = np.transpose(range_doppler_concat_shuffle, axes = (0,2,1,3)) \n",
        "spectrogram_concat_shuffle = np.transpose(spectrogram_concat_shuffle, axes = (0,2,1,3)) "
      ],
      "metadata": {
        "id": "38jkSYXM4aey"
      },
      "execution_count": 8,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rNpkBBIadpPo",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "2d432b7f-59b5-4543-d885-fc797567427a"
      },
      "source": [
        "t = time.time()\n",
        "# ---------- Parameters ----------------\n",
        "augmentation_enable = True\n",
        "normalize_inputs_enable = True\n",
        "num_folds = 5\n",
        "kfold = StratifiedKFold(n_splits=num_folds, shuffle=True, random_state = None) # random_state = 1 ile split run'dan run'a sabit.\n",
        "test_accuracy_per_run = []\n",
        "f1_score_per_run = []\n",
        "epoch_number = 100\n",
        "batch_size = 32\n",
        "dense_size = 64\n",
        "dropout_prob_dense = 0.5\n",
        "repeat_of_mixup = 5\n",
        "number_of_repeat = 5\n",
        "lstm_dropout_rate = 0.1\n",
        "unit_number_of_lstm = 16\n",
        "dense_unit_of_lstm_function = 256\n",
        "decoder_dense = 256\n",
        "\n",
        "\n",
        "for repeat_run_number in range(number_of_repeat):\n",
        "  test_accuracy_per_fold = []\n",
        "  f1_score_per_fold = []\n",
        "  if repeat_run_number > 0:\n",
        "    del range_doppler_concat_shuffle_test\n",
        "    del spectrogram_concat_shuffle_test\n",
        "    del range_doppler_augmented_image\n",
        "    del range_doppler_concat_shuffle_train\n",
        "    del spectrogram_concat_shuffle_train\n",
        "    del spectrogram_augmented_image\n",
        "   \n",
        "  for randomlist_for_train_indx, randomlist_for_test_indx in kfold.split(range_doppler_concat_shuffle,range_doppler_concat_label_shuffle):   \n",
        "    gc.collect()\n",
        "    K.clear_session()\n",
        "    \n",
        "    # test data\n",
        "    range_doppler_concat_shuffle_test = range_doppler_concat_shuffle[randomlist_for_test_indx,:,:,:]\n",
        "    spectrogram_concat_shuffle_test = spectrogram_concat_shuffle[randomlist_for_test_indx,:,:,:]\n",
        "    range_doppler_concat_label_shuffle_test = range_doppler_concat_label_shuffle[randomlist_for_test_indx,:]\n",
        "    #train data\n",
        "    range_doppler_concat_shuffle_train = range_doppler_concat_shuffle[randomlist_for_train_indx,:,:,:]\n",
        "    spectrogram_concat_shuffle_train = spectrogram_concat_shuffle[randomlist_for_train_indx,:,:,:]\n",
        "    spectrogram_concat_label_shuffle_train = spectrogram_concat_label_shuffle[randomlist_for_train_indx,:]\n",
        "      # ---------------- MixUp Augmentation ----------------\n",
        "    (spectrogram_augmented_image,range_doppler_augmented_image,spectrogram_concat_label_shuffle_concat,\\\n",
        "     validation_spectrogram,validation_range_doppler, spectrogram_validation_labels)  =\\\n",
        "      split_and_augmentation_of_training(spectrogram_concat_shuffle_train,range_doppler_concat_shuffle_train,\\\n",
        "                                         spectrogram_concat_label_shuffle_train,\\\n",
        "                                         repeat_of_mixup, augmentation_enable)\n",
        "    \n",
        "    # ---------------- Neural Network Architecture ----------------\n",
        "\n",
        "\n",
        "\n",
        "    def lstm_encoder_network_1(input_shape):\n",
        "        input = Input(shape=input_shape)\n",
        "        x = Bidirectional(LSTM(unit_number_of_lstm, return_sequences=True, dropout = 0.5))(input)\n",
        "        x = Flatten()(x)\n",
        "        x = BatchNormalization()(x)\n",
        "        x = Dropout(0.5)(x)\n",
        "        x = Dense(dense_unit_of_lstm_function)(x)\n",
        "        x = BatchNormalization()(x)\n",
        "        x = Activation('LeakyReLU')(x)\n",
        "        x = Dropout(dropout_prob_dense)(x)\n",
        "        return Model(input, x)\n",
        "\n",
        "\n",
        "    def decoder_for_concat(input_shape):\n",
        "      input = Input(shape=input_shape)\n",
        "      x = Dense(decoder_dense)(input)\n",
        "      # x = BatchNormalization()(x)\n",
        "      x = Activation('LeakyReLU')(x)\n",
        "      x = Dropout(0.3)(x)\n",
        "      x = Dense(dense_size)(x)\n",
        "      # x = BatchNormalization()(x)\n",
        "      x = Activation('LeakyReLU')(x)\n",
        "      x = Dropout(dropout_prob_dense)(x)\n",
        "      x = Dense(1, activation=\"sigmoid\")(x)\n",
        "      return Model(input, x)\n",
        "\n",
        "    # input_shape = range_doppler_concat_shuffle.shape[1:]\n",
        "    # base_network_range_doppler = encoder_for_range_doppler(input_shape)\n",
        "    # range_doppler_input  = Input(shape=input_shape)\n",
        "    # processed_range_doppler  = base_network_range_doppler(range_doppler_input)\n",
        "\n",
        "    input_shape = range_doppler_concat_shuffle.shape[1:3]\n",
        "    base_network_lstm = lstm_encoder_network_1(input_shape)\n",
        "    range_doppler_input  = Input(shape=input_shape)\n",
        "    processed_lstm  = base_network_lstm(range_doppler_input)\n",
        "\n",
        "    base_decoder_network = decoder_for_concat((processed_lstm.shape[1]))\n",
        "    out = base_decoder_network(processed_lstm)\n",
        "\n",
        "    model = Model(inputs=[range_doppler_input], outputs=[out]) \n",
        "    if repeat_run_number == 0:\n",
        "      print(base_network_lstm.summary())\n",
        "      print(model.summary())\n",
        "    # ---------------- Compile and Fit ----------------\n",
        "    model.compile(loss='binary_crossentropy',\n",
        "                  optimizer='adam',\n",
        "                  metrics=['accuracy'])\n",
        "    \n",
        "    earlyStopping = EarlyStopping(monitor='val_loss', patience=15, verbose=0,restore_best_weights=True, mode='min')\n",
        "    history = model.fit((range_doppler_augmented_image),(spectrogram_concat_label_shuffle_concat),\n",
        "                    epochs=epoch_number,\n",
        "                    batch_size=batch_size,\n",
        "                    shuffle = True,\n",
        "                    callbacks=[earlyStopping],\n",
        "                    validation_data = ((validation_range_doppler) , (spectrogram_validation_labels)))\n",
        "    test_loss, test_accuracy  = model.evaluate([range_doppler_concat_shuffle_test],\\\n",
        "                                               [range_doppler_concat_label_shuffle_test],\n",
        "                  batch_size=batch_size)\n",
        "    gc.collect()\n",
        "    # ---------------- Get Test Results ----------------\n",
        "    y_test_predicted = model.predict((range_doppler_concat_shuffle_test), batch_size=batch_size)\n",
        "    # ----- Binarize y_test_predicted values -----\n",
        "    y_test_predicted_binary = np.zeros(y_test_predicted.size)\n",
        "    for ii in range(y_test_predicted.size):\n",
        "      if y_test_predicted[ii] < 0.5:\n",
        "        y_test_predicted_binary[ii] = 0\n",
        "      else:\n",
        "        y_test_predicted_binary[ii] = 1\n",
        "    \n",
        "    test_precision, test_recall, test_f1_score, support = precision_recall_fscore_support(range_doppler_concat_label_shuffle_test, y_test_predicted_binary, average='macro')\n",
        "\n",
        "    test_accuracy_per_fold.append(test_accuracy)\n",
        "    f1_score_per_fold.append(test_f1_score)\n",
        "    del model\n",
        "  test_accuracy_per_run.append(sum(test_accuracy_per_fold)/num_folds)\n",
        "  f1_score_per_run.append(sum(f1_score_per_fold)/num_folds)\n",
        "  print(test_accuracy_per_run)\n",
        "  print(f1_score_per_run)\n",
        "print(f'Mean test accuracy is {\"{:.2f}\".format(sum(test_accuracy_per_run)/number_of_repeat)}, mean test f1 score is {\"{:.2f}\".format(sum(f1_score_per_run)/number_of_repeat)}, \\\n",
        "max test accuracy is {\"{:.2f}\".format(max(test_accuracy_per_run))}, max test f1 score is {\"{:.2f}\".format(max(f1_score_per_run))}, \\\n",
        "min test accuracy is {\"{:.2f}\".format(min(test_accuracy_per_run))}, min test f1 score is {\"{:.2f}\".format(min(f1_score_per_run))}, \\\n",
        "std of test accuracy is {\"{:.2f}\".format(np.std(test_accuracy_per_run, axis=0))}, std of test f1 score is {\"{:.2f}\".format(np.std(f1_score_per_run, axis=0))}')\n",
        "elapsed = time.time() - t\n",
        "print(f'Time elapsed through all process: {\"{:.2f}\".format(elapsed)}, sec')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"model\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " input_1 (InputLayer)        [(None, 5120, 52)]        0         \n",
            "                                                                 \n",
            " bidirectional (Bidirectiona  (None, 5120, 32)         8832      \n",
            " l)                                                              \n",
            "                                                                 \n",
            " flatten (Flatten)           (None, 163840)            0         \n",
            "                                                                 \n",
            " batch_normalization (BatchN  (None, 163840)           655360    \n",
            " ormalization)                                                   \n",
            "                                                                 \n",
            " dropout (Dropout)           (None, 163840)            0         \n",
            "                                                                 \n",
            " dense (Dense)               (None, 256)               41943296  \n",
            "                                                                 \n",
            " batch_normalization_1 (Batc  (None, 256)              1024      \n",
            " hNormalization)                                                 \n",
            "                                                                 \n",
            " activation (Activation)     (None, 256)               0         \n",
            "                                                                 \n",
            " dropout_1 (Dropout)         (None, 256)               0         \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 42,608,512\n",
            "Trainable params: 42,280,320\n",
            "Non-trainable params: 328,192\n",
            "_________________________________________________________________\n",
            "None\n",
            "Model: \"model_2\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " input_2 (InputLayer)        [(None, 5120, 52)]        0         \n",
            "                                                                 \n",
            " model (Functional)          (None, 256)               42608512  \n",
            "                                                                 \n",
            " model_1 (Functional)        (None, 1)                 82305     \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 42,690,817\n",
            "Trainable params: 42,362,625\n",
            "Non-trainable params: 328,192\n",
            "_________________________________________________________________\n",
            "None\n",
            "Epoch 1/100\n",
            "28/28 [==============================] - 19s 410ms/step - loss: 0.4983 - accuracy: 0.5022 - val_loss: 0.1127 - val_accuracy: 1.0000\n",
            "Epoch 2/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.3874 - accuracy: 0.5403 - val_loss: 0.1352 - val_accuracy: 0.9259\n",
            "Epoch 3/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.3523 - accuracy: 0.5570 - val_loss: 0.1486 - val_accuracy: 0.9630\n",
            "Epoch 4/100\n",
            "28/28 [==============================] - 11s 375ms/step - loss: 0.3524 - accuracy: 0.5570 - val_loss: 0.3293 - val_accuracy: 0.9259\n",
            "Epoch 5/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.3393 - accuracy: 0.5604 - val_loss: 0.3139 - val_accuracy: 0.9259\n",
            "Epoch 6/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.3198 - accuracy: 0.5649 - val_loss: 0.1980 - val_accuracy: 0.9259\n",
            "Epoch 7/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.3131 - accuracy: 0.5694 - val_loss: 0.2040 - val_accuracy: 0.8519\n",
            "Epoch 8/100\n",
            "28/28 [==============================] - 11s 378ms/step - loss: 0.3045 - accuracy: 0.5694 - val_loss: 0.1493 - val_accuracy: 1.0000\n",
            "Epoch 9/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.3117 - accuracy: 0.5694 - val_loss: 0.2544 - val_accuracy: 0.9259\n",
            "Epoch 10/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.3124 - accuracy: 0.5682 - val_loss: 0.1478 - val_accuracy: 1.0000\n",
            "Epoch 11/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.3018 - accuracy: 0.5660 - val_loss: 0.2471 - val_accuracy: 0.9259\n",
            "Epoch 12/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2835 - accuracy: 0.5694 - val_loss: 0.0777 - val_accuracy: 1.0000\n",
            "Epoch 13/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2881 - accuracy: 0.5705 - val_loss: 0.1566 - val_accuracy: 1.0000\n",
            "Epoch 14/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.3009 - accuracy: 0.5671 - val_loss: 0.0763 - val_accuracy: 1.0000\n",
            "Epoch 15/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2891 - accuracy: 0.5705 - val_loss: 0.0718 - val_accuracy: 1.0000\n",
            "Epoch 16/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2883 - accuracy: 0.5705 - val_loss: 0.0825 - val_accuracy: 1.0000\n",
            "Epoch 17/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2760 - accuracy: 0.5705 - val_loss: 0.0849 - val_accuracy: 1.0000\n",
            "Epoch 18/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2896 - accuracy: 0.5682 - val_loss: 0.1645 - val_accuracy: 0.9630\n",
            "Epoch 19/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2770 - accuracy: 0.5705 - val_loss: 0.1625 - val_accuracy: 0.9259\n",
            "Epoch 20/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2728 - accuracy: 0.5705 - val_loss: 0.1214 - val_accuracy: 0.9259\n",
            "Epoch 21/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2778 - accuracy: 0.5705 - val_loss: 0.0515 - val_accuracy: 1.0000\n",
            "Epoch 22/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2813 - accuracy: 0.5705 - val_loss: 0.0487 - val_accuracy: 1.0000\n",
            "Epoch 23/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2725 - accuracy: 0.5705 - val_loss: 0.1160 - val_accuracy: 0.9630\n",
            "Epoch 24/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2697 - accuracy: 0.5705 - val_loss: 0.0766 - val_accuracy: 1.0000\n",
            "Epoch 25/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2646 - accuracy: 0.5705 - val_loss: 0.0452 - val_accuracy: 1.0000\n",
            "Epoch 26/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2684 - accuracy: 0.5705 - val_loss: 0.0894 - val_accuracy: 1.0000\n",
            "Epoch 27/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2634 - accuracy: 0.5705 - val_loss: 0.0644 - val_accuracy: 1.0000\n",
            "Epoch 28/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2692 - accuracy: 0.5705 - val_loss: 0.2096 - val_accuracy: 0.8889\n",
            "Epoch 29/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2637 - accuracy: 0.5705 - val_loss: 0.0526 - val_accuracy: 1.0000\n",
            "Epoch 30/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2774 - accuracy: 0.5694 - val_loss: 0.0457 - val_accuracy: 1.0000\n",
            "Epoch 31/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2645 - accuracy: 0.5705 - val_loss: 0.0510 - val_accuracy: 1.0000\n",
            "Epoch 32/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2665 - accuracy: 0.5705 - val_loss: 0.0860 - val_accuracy: 1.0000\n",
            "Epoch 33/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2589 - accuracy: 0.5705 - val_loss: 0.1035 - val_accuracy: 0.9630\n",
            "Epoch 34/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2712 - accuracy: 0.5705 - val_loss: 0.1360 - val_accuracy: 0.9259\n",
            "Epoch 35/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2620 - accuracy: 0.5705 - val_loss: 0.0338 - val_accuracy: 1.0000\n",
            "Epoch 36/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2626 - accuracy: 0.5705 - val_loss: 0.0584 - val_accuracy: 1.0000\n",
            "Epoch 37/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2673 - accuracy: 0.5705 - val_loss: 0.1384 - val_accuracy: 0.9259\n",
            "Epoch 38/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2621 - accuracy: 0.5705 - val_loss: 0.0875 - val_accuracy: 1.0000\n",
            "Epoch 39/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2674 - accuracy: 0.5705 - val_loss: 0.0496 - val_accuracy: 1.0000\n",
            "Epoch 40/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2583 - accuracy: 0.5705 - val_loss: 0.1227 - val_accuracy: 1.0000\n",
            "Epoch 41/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2604 - accuracy: 0.5705 - val_loss: 0.1914 - val_accuracy: 0.8889\n",
            "Epoch 42/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2563 - accuracy: 0.5705 - val_loss: 0.1213 - val_accuracy: 0.9259\n",
            "Epoch 43/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2532 - accuracy: 0.5705 - val_loss: 0.1568 - val_accuracy: 0.9630\n",
            "Epoch 44/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2561 - accuracy: 0.5705 - val_loss: 0.3123 - val_accuracy: 0.8889\n",
            "Epoch 45/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2575 - accuracy: 0.5705 - val_loss: 0.0832 - val_accuracy: 1.0000\n",
            "Epoch 46/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2577 - accuracy: 0.5705 - val_loss: 0.1335 - val_accuracy: 0.9630\n",
            "Epoch 47/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2599 - accuracy: 0.5705 - val_loss: 0.2065 - val_accuracy: 0.8889\n",
            "Epoch 48/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2601 - accuracy: 0.5705 - val_loss: 0.2600 - val_accuracy: 0.8519\n",
            "Epoch 49/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2619 - accuracy: 0.5694 - val_loss: 0.0498 - val_accuracy: 1.0000\n",
            "Epoch 50/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2560 - accuracy: 0.5705 - val_loss: 0.0936 - val_accuracy: 1.0000\n",
            "2/2 [==============================] - 1s 101ms/step - loss: 0.0755 - accuracy: 1.0000\n",
            "Model: \"model\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " input_1 (InputLayer)        [(None, 5120, 52)]        0         \n",
            "                                                                 \n",
            " bidirectional (Bidirectiona  (None, 5120, 32)         8832      \n",
            " l)                                                              \n",
            "                                                                 \n",
            " flatten (Flatten)           (None, 163840)            0         \n",
            "                                                                 \n",
            " batch_normalization (BatchN  (None, 163840)           655360    \n",
            " ormalization)                                                   \n",
            "                                                                 \n",
            " dropout (Dropout)           (None, 163840)            0         \n",
            "                                                                 \n",
            " dense (Dense)               (None, 256)               41943296  \n",
            "                                                                 \n",
            " batch_normalization_1 (Batc  (None, 256)              1024      \n",
            " hNormalization)                                                 \n",
            "                                                                 \n",
            " activation (Activation)     (None, 256)               0         \n",
            "                                                                 \n",
            " dropout_1 (Dropout)         (None, 256)               0         \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 42,608,512\n",
            "Trainable params: 42,280,320\n",
            "Non-trainable params: 328,192\n",
            "_________________________________________________________________\n",
            "None\n",
            "Model: \"model_2\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " input_2 (InputLayer)        [(None, 5120, 52)]        0         \n",
            "                                                                 \n",
            " model (Functional)          (None, 256)               42608512  \n",
            "                                                                 \n",
            " model_1 (Functional)        (None, 1)                 82305     \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 42,690,817\n",
            "Trainable params: 42,362,625\n",
            "Non-trainable params: 328,192\n",
            "_________________________________________________________________\n",
            "None\n",
            "Epoch 1/100\n",
            "28/28 [==============================] - 15s 405ms/step - loss: 0.5703 - accuracy: 0.5218 - val_loss: 0.3070 - val_accuracy: 0.8148\n",
            "Epoch 2/100\n",
            "28/28 [==============================] - 10s 368ms/step - loss: 0.3809 - accuracy: 0.5713 - val_loss: 0.2160 - val_accuracy: 0.9259\n",
            "Epoch 3/100\n",
            "28/28 [==============================] - 10s 364ms/step - loss: 0.3780 - accuracy: 0.5816 - val_loss: 0.2376 - val_accuracy: 0.9259\n",
            "Epoch 4/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.3650 - accuracy: 0.5816 - val_loss: 0.2053 - val_accuracy: 0.9630\n",
            "Epoch 5/100\n",
            "28/28 [==============================] - 10s 369ms/step - loss: 0.3410 - accuracy: 0.5908 - val_loss: 0.2179 - val_accuracy: 0.9259\n",
            "Epoch 6/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.3150 - accuracy: 0.5908 - val_loss: 0.1934 - val_accuracy: 0.9259\n",
            "Epoch 7/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.3324 - accuracy: 0.5874 - val_loss: 0.1666 - val_accuracy: 0.9630\n",
            "Epoch 8/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.3041 - accuracy: 0.5966 - val_loss: 0.1477 - val_accuracy: 0.9630\n",
            "Epoch 9/100\n",
            "28/28 [==============================] - 10s 368ms/step - loss: 0.3136 - accuracy: 0.5966 - val_loss: 0.1897 - val_accuracy: 0.9259\n",
            "Epoch 10/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2978 - accuracy: 0.5989 - val_loss: 0.1327 - val_accuracy: 0.9630\n",
            "Epoch 11/100\n",
            "28/28 [==============================] - 10s 368ms/step - loss: 0.2872 - accuracy: 0.6011 - val_loss: 0.1536 - val_accuracy: 0.9630\n",
            "Epoch 12/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2895 - accuracy: 0.5977 - val_loss: 0.1306 - val_accuracy: 0.9630\n",
            "Epoch 13/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.3003 - accuracy: 0.5977 - val_loss: 0.1110 - val_accuracy: 1.0000\n",
            "Epoch 14/100\n",
            "28/28 [==============================] - 10s 368ms/step - loss: 0.2848 - accuracy: 0.6011 - val_loss: 0.1251 - val_accuracy: 0.9259\n",
            "Epoch 15/100\n",
            "28/28 [==============================] - 10s 367ms/step - loss: 0.2984 - accuracy: 0.6000 - val_loss: 0.1214 - val_accuracy: 0.9630\n",
            "Epoch 16/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2967 - accuracy: 0.5954 - val_loss: 0.1025 - val_accuracy: 0.9630\n",
            "Epoch 17/100\n",
            "28/28 [==============================] - 10s 368ms/step - loss: 0.2768 - accuracy: 0.6011 - val_loss: 0.1423 - val_accuracy: 0.9630\n",
            "Epoch 18/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2755 - accuracy: 0.5989 - val_loss: 0.1024 - val_accuracy: 0.9630\n",
            "Epoch 19/100\n",
            "28/28 [==============================] - 10s 368ms/step - loss: 0.2726 - accuracy: 0.6011 - val_loss: 0.1351 - val_accuracy: 0.9630\n",
            "Epoch 20/100\n",
            "28/28 [==============================] - 10s 368ms/step - loss: 0.2724 - accuracy: 0.6023 - val_loss: 0.1031 - val_accuracy: 0.9630\n",
            "Epoch 21/100\n",
            "28/28 [==============================] - 10s 369ms/step - loss: 0.2624 - accuracy: 0.6023 - val_loss: 0.1406 - val_accuracy: 0.9259\n",
            "Epoch 22/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2716 - accuracy: 0.6011 - val_loss: 0.0793 - val_accuracy: 0.9630\n",
            "Epoch 23/100\n",
            "28/28 [==============================] - 10s 367ms/step - loss: 0.2649 - accuracy: 0.6023 - val_loss: 0.1255 - val_accuracy: 0.9630\n",
            "Epoch 24/100\n",
            "28/28 [==============================] - 10s 368ms/step - loss: 0.2574 - accuracy: 0.6023 - val_loss: 0.0920 - val_accuracy: 0.9630\n",
            "Epoch 25/100\n",
            "28/28 [==============================] - 10s 368ms/step - loss: 0.2632 - accuracy: 0.6000 - val_loss: 0.0959 - val_accuracy: 0.9630\n",
            "Epoch 26/100\n",
            "28/28 [==============================] - 10s 368ms/step - loss: 0.2657 - accuracy: 0.6023 - val_loss: 0.1077 - val_accuracy: 0.9630\n",
            "Epoch 27/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2623 - accuracy: 0.6011 - val_loss: 0.1153 - val_accuracy: 0.9630\n",
            "Epoch 28/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2694 - accuracy: 0.6000 - val_loss: 0.1276 - val_accuracy: 0.9259\n",
            "Epoch 29/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2612 - accuracy: 0.6023 - val_loss: 0.0822 - val_accuracy: 1.0000\n",
            "Epoch 30/100\n",
            "28/28 [==============================] - 10s 367ms/step - loss: 0.2586 - accuracy: 0.6023 - val_loss: 0.1167 - val_accuracy: 0.9630\n",
            "Epoch 31/100\n",
            "28/28 [==============================] - 10s 369ms/step - loss: 0.2554 - accuracy: 0.6023 - val_loss: 0.1507 - val_accuracy: 0.9630\n",
            "Epoch 32/100\n",
            "28/28 [==============================] - 10s 369ms/step - loss: 0.2496 - accuracy: 0.6023 - val_loss: 0.1111 - val_accuracy: 0.9630\n",
            "Epoch 33/100\n",
            "28/28 [==============================] - 10s 368ms/step - loss: 0.2536 - accuracy: 0.6011 - val_loss: 0.1351 - val_accuracy: 0.9259\n",
            "Epoch 34/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2761 - accuracy: 0.5989 - val_loss: 0.0970 - val_accuracy: 0.9630\n",
            "Epoch 35/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2669 - accuracy: 0.6023 - val_loss: 0.0443 - val_accuracy: 1.0000\n",
            "Epoch 36/100\n",
            "28/28 [==============================] - 10s 368ms/step - loss: 0.2613 - accuracy: 0.6011 - val_loss: 0.0734 - val_accuracy: 0.9630\n",
            "Epoch 37/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2559 - accuracy: 0.6023 - val_loss: 0.1022 - val_accuracy: 0.9630\n",
            "Epoch 38/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2539 - accuracy: 0.6023 - val_loss: 0.0924 - val_accuracy: 0.9630\n",
            "Epoch 39/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2522 - accuracy: 0.6023 - val_loss: 0.0953 - val_accuracy: 0.9630\n",
            "Epoch 40/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2586 - accuracy: 0.6023 - val_loss: 0.0959 - val_accuracy: 1.0000\n",
            "Epoch 41/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2629 - accuracy: 0.6023 - val_loss: 0.0906 - val_accuracy: 1.0000\n",
            "Epoch 42/100\n",
            "28/28 [==============================] - 10s 368ms/step - loss: 0.2572 - accuracy: 0.6023 - val_loss: 0.0908 - val_accuracy: 1.0000\n",
            "Epoch 43/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2689 - accuracy: 0.6000 - val_loss: 0.1144 - val_accuracy: 0.9259\n",
            "Epoch 44/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2667 - accuracy: 0.6023 - val_loss: 0.1019 - val_accuracy: 0.9630\n",
            "Epoch 45/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2626 - accuracy: 0.6011 - val_loss: 0.0664 - val_accuracy: 0.9630\n",
            "Epoch 46/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2471 - accuracy: 0.6023 - val_loss: 0.1291 - val_accuracy: 0.9630\n",
            "Epoch 47/100\n",
            "28/28 [==============================] - 10s 368ms/step - loss: 0.2563 - accuracy: 0.6011 - val_loss: 0.0882 - val_accuracy: 1.0000\n",
            "Epoch 48/100\n",
            "28/28 [==============================] - 10s 368ms/step - loss: 0.2513 - accuracy: 0.6023 - val_loss: 0.1298 - val_accuracy: 0.9630\n",
            "Epoch 49/100\n",
            "28/28 [==============================] - 10s 369ms/step - loss: 0.2473 - accuracy: 0.6023 - val_loss: 0.0941 - val_accuracy: 0.9630\n",
            "Epoch 50/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2442 - accuracy: 0.6023 - val_loss: 0.0909 - val_accuracy: 0.9630\n",
            "2/2 [==============================] - 1s 91ms/step - loss: 0.1072 - accuracy: 0.9706\n",
            "Model: \"model\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " input_1 (InputLayer)        [(None, 5120, 52)]        0         \n",
            "                                                                 \n",
            " bidirectional (Bidirectiona  (None, 5120, 32)         8832      \n",
            " l)                                                              \n",
            "                                                                 \n",
            " flatten (Flatten)           (None, 163840)            0         \n",
            "                                                                 \n",
            " batch_normalization (BatchN  (None, 163840)           655360    \n",
            " ormalization)                                                   \n",
            "                                                                 \n",
            " dropout (Dropout)           (None, 163840)            0         \n",
            "                                                                 \n",
            " dense (Dense)               (None, 256)               41943296  \n",
            "                                                                 \n",
            " batch_normalization_1 (Batc  (None, 256)              1024      \n",
            " hNormalization)                                                 \n",
            "                                                                 \n",
            " activation (Activation)     (None, 256)               0         \n",
            "                                                                 \n",
            " dropout_1 (Dropout)         (None, 256)               0         \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 42,608,512\n",
            "Trainable params: 42,280,320\n",
            "Non-trainable params: 328,192\n",
            "_________________________________________________________________\n",
            "None\n",
            "Model: \"model_2\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " input_2 (InputLayer)        [(None, 5120, 52)]        0         \n",
            "                                                                 \n",
            " model (Functional)          (None, 256)               42608512  \n",
            "                                                                 \n",
            " model_1 (Functional)        (None, 1)                 82305     \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 42,690,817\n",
            "Trainable params: 42,362,625\n",
            "Non-trainable params: 328,192\n",
            "_________________________________________________________________\n",
            "None\n",
            "Epoch 1/100\n",
            "27/27 [==============================] - 15s 411ms/step - loss: 0.5684 - accuracy: 0.5128 - val_loss: 0.1798 - val_accuracy: 0.9259\n",
            "Epoch 2/100\n",
            "27/27 [==============================] - 10s 375ms/step - loss: 0.4034 - accuracy: 0.5618 - val_loss: 0.1574 - val_accuracy: 0.9259\n",
            "Epoch 3/100\n",
            "27/27 [==============================] - 10s 374ms/step - loss: 0.3573 - accuracy: 0.5804 - val_loss: 0.1676 - val_accuracy: 0.9630\n",
            "Epoch 4/100\n",
            "27/27 [==============================] - 10s 371ms/step - loss: 0.3444 - accuracy: 0.5851 - val_loss: 0.2075 - val_accuracy: 0.9630\n",
            "Epoch 5/100\n",
            "27/27 [==============================] - 10s 372ms/step - loss: 0.3304 - accuracy: 0.5874 - val_loss: 0.1710 - val_accuracy: 0.9630\n",
            "Epoch 6/100\n",
            "27/27 [==============================] - 10s 373ms/step - loss: 0.3286 - accuracy: 0.5909 - val_loss: 0.1629 - val_accuracy: 0.9630\n",
            "Epoch 7/100\n",
            "27/27 [==============================] - 10s 372ms/step - loss: 0.3182 - accuracy: 0.5897 - val_loss: 0.1631 - val_accuracy: 1.0000\n",
            "Epoch 8/100\n",
            "27/27 [==============================] - 10s 374ms/step - loss: 0.3117 - accuracy: 0.5862 - val_loss: 0.1254 - val_accuracy: 1.0000\n",
            "Epoch 9/100\n",
            "27/27 [==============================] - 10s 376ms/step - loss: 0.3075 - accuracy: 0.5932 - val_loss: 0.1144 - val_accuracy: 0.9630\n",
            "Epoch 10/100\n",
            "27/27 [==============================] - 10s 372ms/step - loss: 0.2986 - accuracy: 0.5967 - val_loss: 0.1563 - val_accuracy: 0.9630\n",
            "Epoch 11/100\n",
            "27/27 [==============================] - 10s 376ms/step - loss: 0.2967 - accuracy: 0.5956 - val_loss: 0.1084 - val_accuracy: 0.9630\n",
            "Epoch 12/100\n",
            "27/27 [==============================] - 10s 375ms/step - loss: 0.2869 - accuracy: 0.5932 - val_loss: 0.0885 - val_accuracy: 0.9630\n",
            "Epoch 13/100\n",
            "27/27 [==============================] - 10s 376ms/step - loss: 0.2925 - accuracy: 0.5956 - val_loss: 0.0656 - val_accuracy: 1.0000\n",
            "Epoch 14/100\n",
            "27/27 [==============================] - 10s 375ms/step - loss: 0.2857 - accuracy: 0.5967 - val_loss: 0.0542 - val_accuracy: 1.0000\n",
            "Epoch 15/100\n",
            "27/27 [==============================] - 10s 372ms/step - loss: 0.2853 - accuracy: 0.5956 - val_loss: 0.0799 - val_accuracy: 0.9630\n",
            "Epoch 16/100\n",
            "27/27 [==============================] - 10s 371ms/step - loss: 0.2775 - accuracy: 0.5967 - val_loss: 0.1068 - val_accuracy: 0.9630\n",
            "Epoch 17/100\n",
            "27/27 [==============================] - 10s 372ms/step - loss: 0.2836 - accuracy: 0.5956 - val_loss: 0.1401 - val_accuracy: 0.9630\n",
            "Epoch 18/100\n",
            "27/27 [==============================] - 10s 372ms/step - loss: 0.2906 - accuracy: 0.5944 - val_loss: 0.0968 - val_accuracy: 0.9630\n",
            "Epoch 19/100\n",
            "27/27 [==============================] - 10s 373ms/step - loss: 0.2778 - accuracy: 0.5967 - val_loss: 0.1688 - val_accuracy: 0.9630\n",
            "Epoch 20/100\n",
            "27/27 [==============================] - 10s 372ms/step - loss: 0.2723 - accuracy: 0.5944 - val_loss: 0.0815 - val_accuracy: 0.9630\n",
            "Epoch 21/100\n",
            "27/27 [==============================] - 10s 374ms/step - loss: 0.2699 - accuracy: 0.5967 - val_loss: 0.1692 - val_accuracy: 0.9259\n",
            "Epoch 22/100\n",
            "27/27 [==============================] - 10s 373ms/step - loss: 0.2663 - accuracy: 0.5967 - val_loss: 0.0822 - val_accuracy: 0.9630\n",
            "Epoch 23/100\n",
            "27/27 [==============================] - 10s 374ms/step - loss: 0.2727 - accuracy: 0.5967 - val_loss: 0.0938 - val_accuracy: 0.9630\n",
            "Epoch 24/100\n",
            "27/27 [==============================] - 10s 372ms/step - loss: 0.2768 - accuracy: 0.5967 - val_loss: 0.0718 - val_accuracy: 0.9630\n",
            "Epoch 25/100\n",
            "27/27 [==============================] - 10s 372ms/step - loss: 0.2700 - accuracy: 0.5967 - val_loss: 0.1161 - val_accuracy: 0.9630\n",
            "Epoch 26/100\n",
            "27/27 [==============================] - 10s 372ms/step - loss: 0.2672 - accuracy: 0.5956 - val_loss: 0.1192 - val_accuracy: 0.9630\n",
            "Epoch 27/100\n",
            "27/27 [==============================] - 10s 370ms/step - loss: 0.2634 - accuracy: 0.5967 - val_loss: 0.1364 - val_accuracy: 0.9630\n",
            "Epoch 28/100\n",
            "27/27 [==============================] - 10s 371ms/step - loss: 0.2621 - accuracy: 0.5967 - val_loss: 0.1104 - val_accuracy: 0.9630\n",
            "Epoch 29/100\n",
            "27/27 [==============================] - 10s 374ms/step - loss: 0.2694 - accuracy: 0.5967 - val_loss: 0.0979 - val_accuracy: 0.9630\n",
            "2/2 [==============================] - 1s 99ms/step - loss: 0.0696 - accuracy: 1.0000\n",
            "Model: \"model\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " input_1 (InputLayer)        [(None, 5120, 52)]        0         \n",
            "                                                                 \n",
            " bidirectional (Bidirectiona  (None, 5120, 32)         8832      \n",
            " l)                                                              \n",
            "                                                                 \n",
            " flatten (Flatten)           (None, 163840)            0         \n",
            "                                                                 \n",
            " batch_normalization (BatchN  (None, 163840)           655360    \n",
            " ormalization)                                                   \n",
            "                                                                 \n",
            " dropout (Dropout)           (None, 163840)            0         \n",
            "                                                                 \n",
            " dense (Dense)               (None, 256)               41943296  \n",
            "                                                                 \n",
            " batch_normalization_1 (Batc  (None, 256)              1024      \n",
            " hNormalization)                                                 \n",
            "                                                                 \n",
            " activation (Activation)     (None, 256)               0         \n",
            "                                                                 \n",
            " dropout_1 (Dropout)         (None, 256)               0         \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 42,608,512\n",
            "Trainable params: 42,280,320\n",
            "Non-trainable params: 328,192\n",
            "_________________________________________________________________\n",
            "None\n",
            "Model: \"model_2\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " input_2 (InputLayer)        [(None, 5120, 52)]        0         \n",
            "                                                                 \n",
            " model (Functional)          (None, 256)               42608512  \n",
            "                                                                 \n",
            " model_1 (Functional)        (None, 1)                 82305     \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 42,690,817\n",
            "Trainable params: 42,362,625\n",
            "Non-trainable params: 328,192\n",
            "_________________________________________________________________\n",
            "None\n",
            "Epoch 1/100\n",
            "28/28 [==============================] - 16s 416ms/step - loss: 0.4579 - accuracy: 0.5263 - val_loss: 0.2002 - val_accuracy: 0.8889\n",
            "Epoch 2/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.4135 - accuracy: 0.5502 - val_loss: 0.3618 - val_accuracy: 0.8519\n",
            "Epoch 3/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.3664 - accuracy: 0.5674 - val_loss: 0.2569 - val_accuracy: 0.8889\n",
            "Epoch 4/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.3404 - accuracy: 0.5788 - val_loss: 0.2622 - val_accuracy: 0.8889\n",
            "Epoch 5/100\n",
            "28/28 [==============================] - 10s 368ms/step - loss: 0.3345 - accuracy: 0.5811 - val_loss: 0.2413 - val_accuracy: 0.8889\n",
            "Epoch 6/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.3084 - accuracy: 0.5856 - val_loss: 0.2519 - val_accuracy: 0.8889\n",
            "Epoch 7/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2924 - accuracy: 0.5913 - val_loss: 0.2188 - val_accuracy: 0.9259\n",
            "Epoch 8/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.3142 - accuracy: 0.5845 - val_loss: 0.2008 - val_accuracy: 0.8889\n",
            "Epoch 9/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.3011 - accuracy: 0.5890 - val_loss: 0.2434 - val_accuracy: 0.8889\n",
            "Epoch 10/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2967 - accuracy: 0.5913 - val_loss: 0.2486 - val_accuracy: 0.8889\n",
            "Epoch 11/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2840 - accuracy: 0.5925 - val_loss: 0.2049 - val_accuracy: 0.8889\n",
            "Epoch 12/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2751 - accuracy: 0.5925 - val_loss: 0.1808 - val_accuracy: 0.8889\n",
            "Epoch 13/100\n",
            "28/28 [==============================] - 10s 369ms/step - loss: 0.2782 - accuracy: 0.5925 - val_loss: 0.2162 - val_accuracy: 0.8889\n",
            "Epoch 14/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2779 - accuracy: 0.5925 - val_loss: 0.3076 - val_accuracy: 0.8889\n",
            "Epoch 15/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2785 - accuracy: 0.5936 - val_loss: 0.1761 - val_accuracy: 0.9259\n",
            "Epoch 16/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2764 - accuracy: 0.5913 - val_loss: 0.1767 - val_accuracy: 0.9259\n",
            "Epoch 17/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2672 - accuracy: 0.5936 - val_loss: 0.3039 - val_accuracy: 0.8889\n",
            "Epoch 18/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2701 - accuracy: 0.5913 - val_loss: 0.2051 - val_accuracy: 0.8889\n",
            "Epoch 19/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2756 - accuracy: 0.5925 - val_loss: 0.1824 - val_accuracy: 0.8889\n",
            "Epoch 20/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2704 - accuracy: 0.5925 - val_loss: 0.2446 - val_accuracy: 0.8889\n",
            "Epoch 21/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2693 - accuracy: 0.5936 - val_loss: 0.1795 - val_accuracy: 0.9259\n",
            "Epoch 22/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2633 - accuracy: 0.5936 - val_loss: 0.1339 - val_accuracy: 0.9259\n",
            "Epoch 23/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2585 - accuracy: 0.5936 - val_loss: 0.1358 - val_accuracy: 0.9259\n",
            "Epoch 24/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2550 - accuracy: 0.5925 - val_loss: 0.1608 - val_accuracy: 0.9259\n",
            "Epoch 25/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2565 - accuracy: 0.5936 - val_loss: 0.2112 - val_accuracy: 0.9259\n",
            "Epoch 26/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2642 - accuracy: 0.5936 - val_loss: 0.1996 - val_accuracy: 0.8889\n",
            "Epoch 27/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2567 - accuracy: 0.5936 - val_loss: 0.2175 - val_accuracy: 0.8889\n",
            "Epoch 28/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2604 - accuracy: 0.5936 - val_loss: 0.2039 - val_accuracy: 0.9259\n",
            "Epoch 29/100\n",
            "28/28 [==============================] - 10s 369ms/step - loss: 0.2524 - accuracy: 0.5936 - val_loss: 0.2539 - val_accuracy: 0.8889\n",
            "Epoch 30/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2531 - accuracy: 0.5936 - val_loss: 0.2019 - val_accuracy: 0.8889\n",
            "Epoch 31/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2504 - accuracy: 0.5936 - val_loss: 0.2110 - val_accuracy: 0.8889\n",
            "Epoch 32/100\n",
            "28/28 [==============================] - 10s 368ms/step - loss: 0.2569 - accuracy: 0.5936 - val_loss: 0.2834 - val_accuracy: 0.8889\n",
            "Epoch 33/100\n",
            "28/28 [==============================] - 10s 365ms/step - loss: 0.2521 - accuracy: 0.5936 - val_loss: 0.2614 - val_accuracy: 0.8889\n",
            "Epoch 34/100\n",
            "28/28 [==============================] - 10s 365ms/step - loss: 0.2689 - accuracy: 0.5936 - val_loss: 0.1820 - val_accuracy: 0.8889\n",
            "Epoch 35/100\n",
            "28/28 [==============================] - 10s 362ms/step - loss: 0.2487 - accuracy: 0.5925 - val_loss: 0.1812 - val_accuracy: 0.9259\n",
            "Epoch 36/100\n",
            "28/28 [==============================] - 10s 366ms/step - loss: 0.2545 - accuracy: 0.5936 - val_loss: 0.2385 - val_accuracy: 0.8889\n",
            "Epoch 37/100\n",
            "28/28 [==============================] - 10s 367ms/step - loss: 0.2531 - accuracy: 0.5936 - val_loss: 0.1793 - val_accuracy: 0.9259\n",
            "2/2 [==============================] - 1s 102ms/step - loss: 0.0518 - accuracy: 1.0000\n",
            "Model: \"model\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " input_1 (InputLayer)        [(None, 5120, 52)]        0         \n",
            "                                                                 \n",
            " bidirectional (Bidirectiona  (None, 5120, 32)         8832      \n",
            " l)                                                              \n",
            "                                                                 \n",
            " flatten (Flatten)           (None, 163840)            0         \n",
            "                                                                 \n",
            " batch_normalization (BatchN  (None, 163840)           655360    \n",
            " ormalization)                                                   \n",
            "                                                                 \n",
            " dropout (Dropout)           (None, 163840)            0         \n",
            "                                                                 \n",
            " dense (Dense)               (None, 256)               41943296  \n",
            "                                                                 \n",
            " batch_normalization_1 (Batc  (None, 256)              1024      \n",
            " hNormalization)                                                 \n",
            "                                                                 \n",
            " activation (Activation)     (None, 256)               0         \n",
            "                                                                 \n",
            " dropout_1 (Dropout)         (None, 256)               0         \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 42,608,512\n",
            "Trainable params: 42,280,320\n",
            "Non-trainable params: 328,192\n",
            "_________________________________________________________________\n",
            "None\n",
            "Model: \"model_2\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " input_2 (InputLayer)        [(None, 5120, 52)]        0         \n",
            "                                                                 \n",
            " model (Functional)          (None, 256)               42608512  \n",
            "                                                                 \n",
            " model_1 (Functional)        (None, 1)                 82305     \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 42,690,817\n",
            "Trainable params: 42,362,625\n",
            "Non-trainable params: 328,192\n",
            "_________________________________________________________________\n",
            "None\n",
            "Epoch 1/100\n",
            "27/27 [==============================] - 15s 410ms/step - loss: 0.5119 - accuracy: 0.5357 - val_loss: 0.3040 - val_accuracy: 0.9630\n",
            "Epoch 2/100\n",
            "27/27 [==============================] - 10s 370ms/step - loss: 0.3684 - accuracy: 0.6048 - val_loss: 0.3435 - val_accuracy: 0.9259\n",
            "Epoch 3/100\n",
            "27/27 [==============================] - 10s 370ms/step - loss: 0.3504 - accuracy: 0.6095 - val_loss: 0.3933 - val_accuracy: 0.8519\n",
            "Epoch 4/100\n",
            "27/27 [==============================] - 10s 370ms/step - loss: 0.3186 - accuracy: 0.6107 - val_loss: 0.3341 - val_accuracy: 0.9259\n",
            "Epoch 5/100\n",
            "27/27 [==============================] - 10s 366ms/step - loss: 0.2975 - accuracy: 0.6202 - val_loss: 0.4117 - val_accuracy: 0.8148\n",
            "Epoch 6/100\n",
            "27/27 [==============================] - 10s 366ms/step - loss: 0.2895 - accuracy: 0.6250 - val_loss: 0.3836 - val_accuracy: 0.8519\n",
            "Epoch 7/100\n",
            "27/27 [==============================] - 10s 372ms/step - loss: 0.2778 - accuracy: 0.6286 - val_loss: 0.2634 - val_accuracy: 0.9259\n",
            "Epoch 8/100\n",
            "27/27 [==============================] - 10s 366ms/step - loss: 0.2729 - accuracy: 0.6286 - val_loss: 0.2903 - val_accuracy: 0.8889\n",
            "Epoch 9/100\n",
            "27/27 [==============================] - 10s 366ms/step - loss: 0.2844 - accuracy: 0.6286 - val_loss: 0.3610 - val_accuracy: 0.8148\n",
            "Epoch 10/100\n",
            "27/27 [==============================] - 10s 365ms/step - loss: 0.2719 - accuracy: 0.6298 - val_loss: 0.2682 - val_accuracy: 0.8889\n",
            "Epoch 11/100\n",
            "27/27 [==============================] - 10s 368ms/step - loss: 0.2635 - accuracy: 0.6286 - val_loss: 0.2143 - val_accuracy: 0.9259\n",
            "Epoch 12/100\n",
            "27/27 [==============================] - 10s 368ms/step - loss: 0.2643 - accuracy: 0.6286 - val_loss: 0.3128 - val_accuracy: 0.8889\n",
            "Epoch 13/100\n",
            "27/27 [==============================] - 10s 371ms/step - loss: 0.2672 - accuracy: 0.6298 - val_loss: 0.2263 - val_accuracy: 0.9259\n",
            "Epoch 14/100\n",
            "27/27 [==============================] - 10s 369ms/step - loss: 0.2698 - accuracy: 0.6310 - val_loss: 0.2573 - val_accuracy: 0.9259\n",
            "Epoch 15/100\n",
            "27/27 [==============================] - 10s 374ms/step - loss: 0.2586 - accuracy: 0.6310 - val_loss: 0.2132 - val_accuracy: 0.9630\n",
            "Epoch 16/100\n",
            "27/27 [==============================] - 10s 370ms/step - loss: 0.2660 - accuracy: 0.6286 - val_loss: 0.2961 - val_accuracy: 0.8889\n",
            "Epoch 17/100\n",
            "27/27 [==============================] - 10s 371ms/step - loss: 0.2567 - accuracy: 0.6298 - val_loss: 0.2442 - val_accuracy: 0.9630\n",
            "Epoch 18/100\n",
            "27/27 [==============================] - 10s 371ms/step - loss: 0.2466 - accuracy: 0.6310 - val_loss: 0.2749 - val_accuracy: 0.9259\n",
            "Epoch 19/100\n",
            "27/27 [==============================] - 10s 371ms/step - loss: 0.2469 - accuracy: 0.6310 - val_loss: 0.2733 - val_accuracy: 0.9259\n",
            "Epoch 20/100\n",
            "27/27 [==============================] - 10s 373ms/step - loss: 0.2420 - accuracy: 0.6310 - val_loss: 0.2020 - val_accuracy: 0.9259\n",
            "Epoch 21/100\n",
            "27/27 [==============================] - 10s 370ms/step - loss: 0.2566 - accuracy: 0.6286 - val_loss: 0.2336 - val_accuracy: 0.9259\n",
            "Epoch 22/100\n",
            "27/27 [==============================] - 10s 368ms/step - loss: 0.2582 - accuracy: 0.6286 - val_loss: 0.5164 - val_accuracy: 0.8519\n",
            "Epoch 23/100\n",
            "27/27 [==============================] - 10s 369ms/step - loss: 0.2504 - accuracy: 0.6310 - val_loss: 0.2188 - val_accuracy: 0.9630\n",
            "Epoch 24/100\n",
            "27/27 [==============================] - 10s 371ms/step - loss: 0.2376 - accuracy: 0.6310 - val_loss: 0.2338 - val_accuracy: 0.9259\n",
            "Epoch 25/100\n",
            "27/27 [==============================] - 10s 373ms/step - loss: 0.2404 - accuracy: 0.6310 - val_loss: 0.1883 - val_accuracy: 0.9259\n",
            "Epoch 26/100\n",
            "27/27 [==============================] - 10s 372ms/step - loss: 0.2455 - accuracy: 0.6298 - val_loss: 0.1815 - val_accuracy: 0.9259\n",
            "Epoch 27/100\n",
            "27/27 [==============================] - 10s 373ms/step - loss: 0.2326 - accuracy: 0.6310 - val_loss: 0.1616 - val_accuracy: 0.9630\n",
            "Epoch 28/100\n",
            "27/27 [==============================] - 10s 370ms/step - loss: 0.2419 - accuracy: 0.6310 - val_loss: 0.1751 - val_accuracy: 0.9630\n",
            "Epoch 29/100\n",
            "27/27 [==============================] - 10s 371ms/step - loss: 0.2347 - accuracy: 0.6310 - val_loss: 0.2197 - val_accuracy: 0.9630\n",
            "Epoch 30/100\n",
            "27/27 [==============================] - 10s 371ms/step - loss: 0.2390 - accuracy: 0.6310 - val_loss: 0.1758 - val_accuracy: 0.9259\n",
            "Epoch 31/100\n",
            "27/27 [==============================] - 10s 372ms/step - loss: 0.2362 - accuracy: 0.6310 - val_loss: 0.1549 - val_accuracy: 0.9630\n",
            "Epoch 32/100\n",
            "27/27 [==============================] - 10s 370ms/step - loss: 0.2346 - accuracy: 0.6310 - val_loss: 0.2008 - val_accuracy: 0.9630\n",
            "Epoch 33/100\n",
            "27/27 [==============================] - 10s 371ms/step - loss: 0.2369 - accuracy: 0.6298 - val_loss: 0.1884 - val_accuracy: 0.9630\n",
            "Epoch 34/100\n",
            "27/27 [==============================] - 10s 372ms/step - loss: 0.2348 - accuracy: 0.6310 - val_loss: 0.1508 - val_accuracy: 0.9630\n",
            "Epoch 35/100\n",
            "27/27 [==============================] - 10s 370ms/step - loss: 0.2376 - accuracy: 0.6310 - val_loss: 0.2084 - val_accuracy: 0.9630\n",
            "Epoch 36/100\n",
            "27/27 [==============================] - 10s 368ms/step - loss: 0.2352 - accuracy: 0.6310 - val_loss: 0.1737 - val_accuracy: 0.9630\n",
            "Epoch 37/100\n",
            "27/27 [==============================] - 10s 368ms/step - loss: 0.2360 - accuracy: 0.6310 - val_loss: 0.1981 - val_accuracy: 0.9630\n",
            "Epoch 38/100\n",
            "27/27 [==============================] - 10s 371ms/step - loss: 0.2352 - accuracy: 0.6298 - val_loss: 0.2043 - val_accuracy: 0.9630\n",
            "Epoch 39/100\n",
            "27/27 [==============================] - 10s 369ms/step - loss: 0.2288 - accuracy: 0.6310 - val_loss: 0.2013 - val_accuracy: 0.9630\n",
            "Epoch 40/100\n",
            "27/27 [==============================] - 10s 369ms/step - loss: 0.2328 - accuracy: 0.6310 - val_loss: 0.2238 - val_accuracy: 0.9259\n",
            "Epoch 41/100\n",
            "27/27 [==============================] - 10s 370ms/step - loss: 0.2366 - accuracy: 0.6310 - val_loss: 0.1934 - val_accuracy: 0.9630\n",
            "Epoch 42/100\n",
            "27/27 [==============================] - 10s 368ms/step - loss: 0.2306 - accuracy: 0.6298 - val_loss: 0.2082 - val_accuracy: 0.9630\n",
            "Epoch 43/100\n",
            "27/27 [==============================] - 10s 370ms/step - loss: 0.2307 - accuracy: 0.6310 - val_loss: 0.2509 - val_accuracy: 0.9630\n",
            "Epoch 44/100\n",
            "27/27 [==============================] - 10s 370ms/step - loss: 0.2292 - accuracy: 0.6310 - val_loss: 0.1892 - val_accuracy: 0.9630\n",
            "Epoch 45/100\n",
            "27/27 [==============================] - 10s 370ms/step - loss: 0.2260 - accuracy: 0.6310 - val_loss: 0.2746 - val_accuracy: 0.8889\n",
            "Epoch 46/100\n",
            "27/27 [==============================] - 10s 369ms/step - loss: 0.2318 - accuracy: 0.6298 - val_loss: 0.2793 - val_accuracy: 0.9259\n",
            "Epoch 47/100\n",
            "27/27 [==============================] - 10s 369ms/step - loss: 0.2282 - accuracy: 0.6310 - val_loss: 0.2352 - val_accuracy: 0.9259\n",
            "Epoch 48/100\n",
            "27/27 [==============================] - 10s 369ms/step - loss: 0.2299 - accuracy: 0.6310 - val_loss: 0.1986 - val_accuracy: 0.9630\n",
            "Epoch 49/100\n",
            "27/27 [==============================] - 10s 372ms/step - loss: 0.2333 - accuracy: 0.6310 - val_loss: 0.2020 - val_accuracy: 0.9259\n",
            "2/2 [==============================] - 1s 99ms/step - loss: 0.1432 - accuracy: 0.9118\n",
            "WARNING:tensorflow:5 out of the last 9 calls to <function Model.make_predict_function.<locals>.predict_function at 0x7f33080aab00> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has experimental_relax_shapes=True option that relaxes argument shapes that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for  more details.\n",
            "[0.9764705777168274]\n",
            "[0.973400290736091]\n",
            "Epoch 1/100\n",
            "28/28 [==============================] - 15s 413ms/step - loss: 0.5360 - accuracy: 0.4909 - val_loss: 0.5005 - val_accuracy: 0.7037\n",
            "Epoch 2/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.4088 - accuracy: 0.5548 - val_loss: 0.3533 - val_accuracy: 0.8889\n",
            "Epoch 3/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.3618 - accuracy: 0.5742 - val_loss: 0.3415 - val_accuracy: 0.8148\n",
            "Epoch 4/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.3439 - accuracy: 0.5731 - val_loss: 0.3249 - val_accuracy: 0.8148\n",
            "Epoch 5/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.3144 - accuracy: 0.5833 - val_loss: 0.2745 - val_accuracy: 0.8889\n",
            "Epoch 6/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.3275 - accuracy: 0.5811 - val_loss: 0.3112 - val_accuracy: 0.9259\n",
            "Epoch 7/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.3106 - accuracy: 0.5856 - val_loss: 0.2474 - val_accuracy: 0.8889\n",
            "Epoch 8/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.3029 - accuracy: 0.5845 - val_loss: 0.3058 - val_accuracy: 0.8519\n",
            "Epoch 9/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.3047 - accuracy: 0.5845 - val_loss: 0.3079 - val_accuracy: 0.8889\n",
            "Epoch 10/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2851 - accuracy: 0.5879 - val_loss: 0.2204 - val_accuracy: 0.9259\n",
            "Epoch 11/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2911 - accuracy: 0.5868 - val_loss: 0.2428 - val_accuracy: 0.8889\n",
            "Epoch 12/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2887 - accuracy: 0.5890 - val_loss: 0.2229 - val_accuracy: 0.8889\n",
            "Epoch 13/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2841 - accuracy: 0.5868 - val_loss: 0.1702 - val_accuracy: 0.9630\n",
            "Epoch 14/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2820 - accuracy: 0.5868 - val_loss: 0.3041 - val_accuracy: 0.8519\n",
            "Epoch 15/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2864 - accuracy: 0.5890 - val_loss: 0.2319 - val_accuracy: 0.8889\n",
            "Epoch 16/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2815 - accuracy: 0.5879 - val_loss: 0.2471 - val_accuracy: 0.8148\n",
            "Epoch 17/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2802 - accuracy: 0.5879 - val_loss: 0.3210 - val_accuracy: 0.9259\n",
            "Epoch 18/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2683 - accuracy: 0.5879 - val_loss: 0.1968 - val_accuracy: 0.9259\n",
            "Epoch 19/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2703 - accuracy: 0.5879 - val_loss: 0.1620 - val_accuracy: 0.9630\n",
            "Epoch 20/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2664 - accuracy: 0.5890 - val_loss: 0.2052 - val_accuracy: 0.9259\n",
            "Epoch 21/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2624 - accuracy: 0.5890 - val_loss: 0.2618 - val_accuracy: 0.9259\n",
            "Epoch 22/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2713 - accuracy: 0.5890 - val_loss: 0.2486 - val_accuracy: 0.9259\n",
            "Epoch 23/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2636 - accuracy: 0.5890 - val_loss: 0.1707 - val_accuracy: 0.9259\n",
            "Epoch 24/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2640 - accuracy: 0.5879 - val_loss: 0.2491 - val_accuracy: 0.8519\n",
            "Epoch 25/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2630 - accuracy: 0.5890 - val_loss: 0.1789 - val_accuracy: 0.9259\n",
            "Epoch 26/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2636 - accuracy: 0.5890 - val_loss: 0.2728 - val_accuracy: 0.8889\n",
            "Epoch 27/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2662 - accuracy: 0.5890 - val_loss: 0.2060 - val_accuracy: 0.9259\n",
            "Epoch 28/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2642 - accuracy: 0.5890 - val_loss: 0.2330 - val_accuracy: 0.9259\n",
            "Epoch 29/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2578 - accuracy: 0.5890 - val_loss: 0.2486 - val_accuracy: 0.9259\n",
            "Epoch 30/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2638 - accuracy: 0.5879 - val_loss: 0.2484 - val_accuracy: 0.8889\n",
            "Epoch 31/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2593 - accuracy: 0.5890 - val_loss: 0.2175 - val_accuracy: 0.9259\n",
            "Epoch 32/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2572 - accuracy: 0.5890 - val_loss: 0.3012 - val_accuracy: 0.8519\n",
            "Epoch 33/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2630 - accuracy: 0.5890 - val_loss: 0.2688 - val_accuracy: 0.9259\n",
            "Epoch 34/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2534 - accuracy: 0.5890 - val_loss: 0.3396 - val_accuracy: 0.8889\n",
            "2/2 [==============================] - 1s 94ms/step - loss: 0.0647 - accuracy: 1.0000\n",
            "WARNING:tensorflow:6 out of the last 11 calls to <function Model.make_predict_function.<locals>.predict_function at 0x7f32f0265cb0> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has experimental_relax_shapes=True option that relaxes argument shapes that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for  more details.\n",
            "Epoch 1/100\n",
            "27/27 [==============================] - 15s 416ms/step - loss: 0.5346 - accuracy: 0.5127 - val_loss: 0.8023 - val_accuracy: 0.6667\n",
            "Epoch 2/100\n",
            "27/27 [==============================] - 10s 377ms/step - loss: 0.4088 - accuracy: 0.5509 - val_loss: 0.4208 - val_accuracy: 0.7407\n",
            "Epoch 3/100\n",
            "27/27 [==============================] - 10s 377ms/step - loss: 0.3558 - accuracy: 0.5729 - val_loss: 0.3013 - val_accuracy: 0.8148\n",
            "Epoch 4/100\n",
            "27/27 [==============================] - 10s 378ms/step - loss: 0.3403 - accuracy: 0.5752 - val_loss: 0.1933 - val_accuracy: 0.9630\n",
            "Epoch 5/100\n",
            "27/27 [==============================] - 10s 374ms/step - loss: 0.3357 - accuracy: 0.5799 - val_loss: 0.2632 - val_accuracy: 0.8519\n",
            "Epoch 6/100\n",
            "27/27 [==============================] - 10s 377ms/step - loss: 0.3080 - accuracy: 0.5868 - val_loss: 0.1778 - val_accuracy: 1.0000\n",
            "Epoch 7/100\n",
            "27/27 [==============================] - 10s 378ms/step - loss: 0.3213 - accuracy: 0.5845 - val_loss: 0.1548 - val_accuracy: 0.9259\n",
            "Epoch 8/100\n",
            "27/27 [==============================] - 10s 375ms/step - loss: 0.2962 - accuracy: 0.5868 - val_loss: 0.1908 - val_accuracy: 0.8519\n",
            "Epoch 9/100\n",
            "27/27 [==============================] - 10s 375ms/step - loss: 0.3040 - accuracy: 0.5856 - val_loss: 0.1679 - val_accuracy: 0.8889\n",
            "Epoch 10/100\n",
            "27/27 [==============================] - 10s 377ms/step - loss: 0.2943 - accuracy: 0.5880 - val_loss: 0.1195 - val_accuracy: 1.0000\n",
            "Epoch 11/100\n",
            "27/27 [==============================] - 10s 375ms/step - loss: 0.2885 - accuracy: 0.5891 - val_loss: 0.1276 - val_accuracy: 1.0000\n",
            "Epoch 12/100\n",
            "27/27 [==============================] - 10s 380ms/step - loss: 0.2854 - accuracy: 0.5891 - val_loss: 0.0667 - val_accuracy: 1.0000\n",
            "Epoch 13/100\n",
            "27/27 [==============================] - 10s 377ms/step - loss: 0.2906 - accuracy: 0.5845 - val_loss: 0.1263 - val_accuracy: 1.0000\n",
            "Epoch 14/100\n",
            "27/27 [==============================] - 10s 375ms/step - loss: 0.2853 - accuracy: 0.5868 - val_loss: 0.1177 - val_accuracy: 0.9630\n",
            "Epoch 15/100\n",
            "27/27 [==============================] - 10s 377ms/step - loss: 0.2820 - accuracy: 0.5880 - val_loss: 0.1312 - val_accuracy: 0.9630\n",
            "Epoch 16/100\n",
            "27/27 [==============================] - 10s 375ms/step - loss: 0.2769 - accuracy: 0.5891 - val_loss: 0.1086 - val_accuracy: 1.0000\n",
            "Epoch 17/100\n",
            "27/27 [==============================] - 10s 378ms/step - loss: 0.2841 - accuracy: 0.5891 - val_loss: 0.0546 - val_accuracy: 1.0000\n",
            "Epoch 18/100\n",
            "27/27 [==============================] - 10s 375ms/step - loss: 0.2752 - accuracy: 0.5891 - val_loss: 0.1235 - val_accuracy: 0.9630\n",
            "Epoch 19/100\n",
            "27/27 [==============================] - 10s 376ms/step - loss: 0.2713 - accuracy: 0.5891 - val_loss: 0.1238 - val_accuracy: 0.9630\n",
            "Epoch 20/100\n",
            "27/27 [==============================] - 10s 375ms/step - loss: 0.2676 - accuracy: 0.5903 - val_loss: 0.1169 - val_accuracy: 0.9259\n",
            "Epoch 21/100\n",
            "27/27 [==============================] - 10s 378ms/step - loss: 0.2663 - accuracy: 0.5903 - val_loss: 0.0515 - val_accuracy: 1.0000\n",
            "Epoch 22/100\n",
            "27/27 [==============================] - 10s 377ms/step - loss: 0.2756 - accuracy: 0.5891 - val_loss: 0.0849 - val_accuracy: 1.0000\n",
            "Epoch 23/100\n",
            "27/27 [==============================] - 10s 375ms/step - loss: 0.2661 - accuracy: 0.5903 - val_loss: 0.1025 - val_accuracy: 1.0000\n",
            "Epoch 24/100\n",
            "27/27 [==============================] - 10s 373ms/step - loss: 0.2704 - accuracy: 0.5903 - val_loss: 0.0655 - val_accuracy: 1.0000\n",
            "Epoch 25/100\n",
            "27/27 [==============================] - 10s 376ms/step - loss: 0.2595 - accuracy: 0.5903 - val_loss: 0.0652 - val_accuracy: 1.0000\n",
            "Epoch 26/100\n",
            "27/27 [==============================] - 10s 374ms/step - loss: 0.2670 - accuracy: 0.5903 - val_loss: 0.0949 - val_accuracy: 0.9630\n",
            "Epoch 27/100\n",
            "27/27 [==============================] - 10s 371ms/step - loss: 0.2618 - accuracy: 0.5903 - val_loss: 0.1191 - val_accuracy: 0.9259\n",
            "Epoch 28/100\n",
            "27/27 [==============================] - 10s 374ms/step - loss: 0.2599 - accuracy: 0.5903 - val_loss: 0.0675 - val_accuracy: 1.0000\n",
            "Epoch 29/100\n",
            "27/27 [==============================] - 10s 374ms/step - loss: 0.2573 - accuracy: 0.5903 - val_loss: 0.1181 - val_accuracy: 0.9630\n",
            "Epoch 30/100\n",
            "27/27 [==============================] - 10s 377ms/step - loss: 0.2557 - accuracy: 0.5903 - val_loss: 0.0432 - val_accuracy: 1.0000\n",
            "Epoch 31/100\n",
            "27/27 [==============================] - 10s 375ms/step - loss: 0.2635 - accuracy: 0.5891 - val_loss: 0.0714 - val_accuracy: 1.0000\n",
            "Epoch 32/100\n",
            "27/27 [==============================] - 10s 373ms/step - loss: 0.2616 - accuracy: 0.5903 - val_loss: 0.0783 - val_accuracy: 0.9630\n",
            "Epoch 33/100\n",
            "27/27 [==============================] - 10s 374ms/step - loss: 0.2568 - accuracy: 0.5903 - val_loss: 0.0839 - val_accuracy: 1.0000\n",
            "Epoch 34/100\n",
            "27/27 [==============================] - 10s 374ms/step - loss: 0.2554 - accuracy: 0.5903 - val_loss: 0.0687 - val_accuracy: 1.0000\n",
            "Epoch 35/100\n",
            "27/27 [==============================] - 10s 375ms/step - loss: 0.2611 - accuracy: 0.5903 - val_loss: 0.0918 - val_accuracy: 0.9630\n",
            "Epoch 36/100\n",
            "27/27 [==============================] - 10s 375ms/step - loss: 0.2540 - accuracy: 0.5891 - val_loss: 0.1382 - val_accuracy: 0.9259\n",
            "Epoch 37/100\n",
            "27/27 [==============================] - 10s 372ms/step - loss: 0.2503 - accuracy: 0.5903 - val_loss: 0.0965 - val_accuracy: 0.9259\n",
            "Epoch 38/100\n",
            "27/27 [==============================] - 10s 374ms/step - loss: 0.2494 - accuracy: 0.5903 - val_loss: 0.0883 - val_accuracy: 0.9630\n",
            "Epoch 39/100\n",
            "27/27 [==============================] - 10s 375ms/step - loss: 0.2516 - accuracy: 0.5903 - val_loss: 0.1072 - val_accuracy: 0.9630\n",
            "Epoch 40/100\n",
            "27/27 [==============================] - 10s 374ms/step - loss: 0.2588 - accuracy: 0.5903 - val_loss: 0.1039 - val_accuracy: 0.9630\n",
            "Epoch 41/100\n",
            "27/27 [==============================] - 10s 376ms/step - loss: 0.2531 - accuracy: 0.5903 - val_loss: 0.0846 - val_accuracy: 0.9630\n",
            "Epoch 42/100\n",
            "27/27 [==============================] - 10s 376ms/step - loss: 0.2531 - accuracy: 0.5903 - val_loss: 0.0716 - val_accuracy: 1.0000\n",
            "Epoch 43/100\n",
            "27/27 [==============================] - 10s 377ms/step - loss: 0.2514 - accuracy: 0.5903 - val_loss: 0.0740 - val_accuracy: 1.0000\n",
            "Epoch 44/100\n",
            "27/27 [==============================] - 10s 374ms/step - loss: 0.2539 - accuracy: 0.5903 - val_loss: 0.0811 - val_accuracy: 1.0000\n",
            "Epoch 45/100\n",
            "27/27 [==============================] - 10s 375ms/step - loss: 0.2507 - accuracy: 0.5903 - val_loss: 0.1216 - val_accuracy: 0.9259\n",
            "2/2 [==============================] - 1s 99ms/step - loss: 0.1319 - accuracy: 1.0000\n",
            "Epoch 1/100\n",
            "28/28 [==============================] - 15s 408ms/step - loss: 0.6031 - accuracy: 0.4600 - val_loss: 0.4139 - val_accuracy: 0.7037\n",
            "Epoch 2/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.4268 - accuracy: 0.5320 - val_loss: 0.2262 - val_accuracy: 0.9259\n",
            "Epoch 3/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.3780 - accuracy: 0.5514 - val_loss: 0.2038 - val_accuracy: 0.9630\n",
            "Epoch 4/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.3729 - accuracy: 0.5537 - val_loss: 0.2852 - val_accuracy: 0.8519\n",
            "Epoch 5/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.3384 - accuracy: 0.5639 - val_loss: 0.1811 - val_accuracy: 1.0000\n",
            "Epoch 6/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.3344 - accuracy: 0.5708 - val_loss: 0.1721 - val_accuracy: 1.0000\n",
            "Epoch 7/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.3191 - accuracy: 0.5753 - val_loss: 0.1640 - val_accuracy: 1.0000\n",
            "Epoch 8/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.3196 - accuracy: 0.5731 - val_loss: 0.1461 - val_accuracy: 1.0000\n",
            "Epoch 9/100\n",
            "28/28 [==============================] - 10s 368ms/step - loss: 0.3148 - accuracy: 0.5742 - val_loss: 0.1647 - val_accuracy: 0.9259\n",
            "Epoch 10/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.3036 - accuracy: 0.5742 - val_loss: 0.1239 - val_accuracy: 0.9630\n",
            "Epoch 11/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2997 - accuracy: 0.5731 - val_loss: 0.1406 - val_accuracy: 0.9630\n",
            "Epoch 12/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2950 - accuracy: 0.5753 - val_loss: 0.1756 - val_accuracy: 0.9259\n",
            "Epoch 13/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2936 - accuracy: 0.5731 - val_loss: 0.1110 - val_accuracy: 0.9630\n",
            "Epoch 14/100\n",
            "28/28 [==============================] - 10s 369ms/step - loss: 0.2949 - accuracy: 0.5776 - val_loss: 0.1489 - val_accuracy: 0.9259\n",
            "Epoch 15/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2967 - accuracy: 0.5742 - val_loss: 0.1526 - val_accuracy: 0.9630\n",
            "Epoch 16/100\n",
            "28/28 [==============================] - 10s 367ms/step - loss: 0.2910 - accuracy: 0.5753 - val_loss: 0.3396 - val_accuracy: 0.8519\n",
            "Epoch 17/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2836 - accuracy: 0.5776 - val_loss: 0.1072 - val_accuracy: 0.9630\n",
            "Epoch 18/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2873 - accuracy: 0.5753 - val_loss: 0.1395 - val_accuracy: 0.9259\n",
            "Epoch 19/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2791 - accuracy: 0.5776 - val_loss: 0.2320 - val_accuracy: 0.9259\n",
            "Epoch 20/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2764 - accuracy: 0.5765 - val_loss: 0.1196 - val_accuracy: 0.9630\n",
            "Epoch 21/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2825 - accuracy: 0.5753 - val_loss: 0.1543 - val_accuracy: 0.9630\n",
            "Epoch 22/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2789 - accuracy: 0.5776 - val_loss: 0.0986 - val_accuracy: 1.0000\n",
            "Epoch 23/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2800 - accuracy: 0.5753 - val_loss: 0.1938 - val_accuracy: 0.9259\n",
            "Epoch 24/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2811 - accuracy: 0.5776 - val_loss: 0.0994 - val_accuracy: 1.0000\n",
            "Epoch 25/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2760 - accuracy: 0.5765 - val_loss: 0.0785 - val_accuracy: 1.0000\n",
            "Epoch 26/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2816 - accuracy: 0.5765 - val_loss: 0.1030 - val_accuracy: 0.9630\n",
            "Epoch 27/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2731 - accuracy: 0.5776 - val_loss: 0.0772 - val_accuracy: 0.9630\n",
            "Epoch 28/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2657 - accuracy: 0.5765 - val_loss: 0.1025 - val_accuracy: 0.9630\n",
            "Epoch 29/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2630 - accuracy: 0.5776 - val_loss: 0.1176 - val_accuracy: 0.9630\n",
            "Epoch 30/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2688 - accuracy: 0.5776 - val_loss: 0.0873 - val_accuracy: 0.9630\n",
            "Epoch 31/100\n",
            "28/28 [==============================] - 10s 369ms/step - loss: 0.2646 - accuracy: 0.5765 - val_loss: 0.1022 - val_accuracy: 0.9259\n",
            "Epoch 32/100\n",
            "28/28 [==============================] - 10s 369ms/step - loss: 0.2654 - accuracy: 0.5776 - val_loss: 0.0836 - val_accuracy: 1.0000\n",
            "Epoch 33/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2618 - accuracy: 0.5776 - val_loss: 0.1094 - val_accuracy: 0.9630\n",
            "Epoch 34/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2621 - accuracy: 0.5776 - val_loss: 0.1092 - val_accuracy: 0.9259\n",
            "Epoch 35/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2694 - accuracy: 0.5776 - val_loss: 0.3266 - val_accuracy: 0.8889\n",
            "Epoch 36/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2703 - accuracy: 0.5776 - val_loss: 0.1044 - val_accuracy: 0.9630\n",
            "Epoch 37/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2619 - accuracy: 0.5776 - val_loss: 0.0861 - val_accuracy: 0.9630\n",
            "Epoch 38/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2646 - accuracy: 0.5776 - val_loss: 0.1116 - val_accuracy: 0.9630\n",
            "Epoch 39/100\n",
            "28/28 [==============================] - 10s 368ms/step - loss: 0.2674 - accuracy: 0.5776 - val_loss: 0.1516 - val_accuracy: 0.9630\n",
            "Epoch 40/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2594 - accuracy: 0.5776 - val_loss: 0.1477 - val_accuracy: 0.9630\n",
            "Epoch 41/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2565 - accuracy: 0.5776 - val_loss: 0.1076 - val_accuracy: 0.9630\n",
            "Epoch 42/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2745 - accuracy: 0.5776 - val_loss: 0.1688 - val_accuracy: 0.9630\n",
            "2/2 [==============================] - 1s 95ms/step - loss: 0.0812 - accuracy: 1.0000\n",
            "Epoch 1/100\n",
            "28/28 [==============================] - 16s 414ms/step - loss: 0.5109 - accuracy: 0.5000 - val_loss: 0.0969 - val_accuracy: 1.0000\n",
            "Epoch 2/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.3996 - accuracy: 0.5578 - val_loss: 0.1882 - val_accuracy: 0.9259\n",
            "Epoch 3/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.3610 - accuracy: 0.5714 - val_loss: 0.1390 - val_accuracy: 1.0000\n",
            "Epoch 4/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.3542 - accuracy: 0.5737 - val_loss: 0.1288 - val_accuracy: 0.9630\n",
            "Epoch 5/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.3422 - accuracy: 0.5771 - val_loss: 0.1711 - val_accuracy: 0.9259\n",
            "Epoch 6/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.3214 - accuracy: 0.5805 - val_loss: 0.1663 - val_accuracy: 0.9630\n",
            "Epoch 7/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.3173 - accuracy: 0.5828 - val_loss: 0.1197 - val_accuracy: 1.0000\n",
            "Epoch 8/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.3053 - accuracy: 0.5828 - val_loss: 0.1323 - val_accuracy: 0.9630\n",
            "Epoch 9/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.3022 - accuracy: 0.5828 - val_loss: 0.1530 - val_accuracy: 1.0000\n",
            "Epoch 10/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2949 - accuracy: 0.5873 - val_loss: 0.2800 - val_accuracy: 0.8519\n",
            "Epoch 11/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2932 - accuracy: 0.5862 - val_loss: 0.1301 - val_accuracy: 0.9259\n",
            "Epoch 12/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2896 - accuracy: 0.5850 - val_loss: 0.1656 - val_accuracy: 0.9259\n",
            "Epoch 13/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2839 - accuracy: 0.5850 - val_loss: 0.1792 - val_accuracy: 0.9259\n",
            "Epoch 14/100\n",
            "28/28 [==============================] - 11s 375ms/step - loss: 0.2788 - accuracy: 0.5862 - val_loss: 0.1018 - val_accuracy: 0.9630\n",
            "Epoch 15/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2728 - accuracy: 0.5850 - val_loss: 0.0714 - val_accuracy: 1.0000\n",
            "Epoch 16/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2789 - accuracy: 0.5873 - val_loss: 0.0983 - val_accuracy: 1.0000\n",
            "Epoch 17/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2684 - accuracy: 0.5862 - val_loss: 0.1244 - val_accuracy: 0.9630\n",
            "Epoch 18/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2734 - accuracy: 0.5850 - val_loss: 0.1084 - val_accuracy: 0.9630\n",
            "Epoch 19/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2773 - accuracy: 0.5839 - val_loss: 0.1023 - val_accuracy: 0.9630\n",
            "Epoch 20/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2642 - accuracy: 0.5873 - val_loss: 0.0995 - val_accuracy: 0.9630\n",
            "Epoch 21/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2671 - accuracy: 0.5873 - val_loss: 0.1842 - val_accuracy: 0.8519\n",
            "Epoch 22/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2712 - accuracy: 0.5862 - val_loss: 0.0751 - val_accuracy: 1.0000\n",
            "Epoch 23/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2638 - accuracy: 0.5873 - val_loss: 0.1515 - val_accuracy: 0.8889\n",
            "Epoch 24/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2811 - accuracy: 0.5862 - val_loss: 0.0712 - val_accuracy: 1.0000\n",
            "Epoch 25/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2626 - accuracy: 0.5862 - val_loss: 0.0574 - val_accuracy: 1.0000\n",
            "Epoch 26/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2664 - accuracy: 0.5873 - val_loss: 0.1220 - val_accuracy: 0.9259\n",
            "Epoch 27/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2662 - accuracy: 0.5862 - val_loss: 0.0441 - val_accuracy: 1.0000\n",
            "Epoch 28/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2549 - accuracy: 0.5873 - val_loss: 0.0737 - val_accuracy: 1.0000\n",
            "Epoch 29/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2603 - accuracy: 0.5873 - val_loss: 0.0994 - val_accuracy: 1.0000\n",
            "Epoch 30/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2591 - accuracy: 0.5873 - val_loss: 0.0652 - val_accuracy: 1.0000\n",
            "Epoch 31/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2711 - accuracy: 0.5862 - val_loss: 0.0485 - val_accuracy: 1.0000\n",
            "Epoch 32/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2575 - accuracy: 0.5873 - val_loss: 0.0986 - val_accuracy: 0.9630\n",
            "Epoch 33/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2538 - accuracy: 0.5873 - val_loss: 0.1447 - val_accuracy: 0.9259\n",
            "Epoch 34/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2561 - accuracy: 0.5873 - val_loss: 0.1522 - val_accuracy: 0.8889\n",
            "Epoch 35/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2560 - accuracy: 0.5873 - val_loss: 0.0755 - val_accuracy: 0.9630\n",
            "Epoch 36/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2525 - accuracy: 0.5873 - val_loss: 0.1461 - val_accuracy: 0.9259\n",
            "Epoch 37/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2600 - accuracy: 0.5873 - val_loss: 0.2157 - val_accuracy: 0.8519\n",
            "Epoch 38/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2552 - accuracy: 0.5873 - val_loss: 0.1660 - val_accuracy: 0.9259\n",
            "Epoch 39/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2570 - accuracy: 0.5873 - val_loss: 0.1081 - val_accuracy: 0.9630\n",
            "Epoch 40/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2509 - accuracy: 0.5873 - val_loss: 0.1518 - val_accuracy: 0.9259\n",
            "Epoch 41/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2534 - accuracy: 0.5873 - val_loss: 0.0817 - val_accuracy: 1.0000\n",
            "Epoch 42/100\n",
            "28/28 [==============================] - 10s 376ms/step - loss: 0.2588 - accuracy: 0.5873 - val_loss: 0.1108 - val_accuracy: 0.9630\n",
            "2/2 [==============================] - 1s 97ms/step - loss: 0.0737 - accuracy: 1.0000\n",
            "Epoch 1/100\n",
            "28/28 [==============================] - 16s 413ms/step - loss: 0.5326 - accuracy: 0.4943 - val_loss: 0.2178 - val_accuracy: 0.9630\n",
            "Epoch 2/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.3971 - accuracy: 0.5571 - val_loss: 0.2436 - val_accuracy: 0.8889\n",
            "Epoch 3/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.3544 - accuracy: 0.5662 - val_loss: 0.1679 - val_accuracy: 0.9630\n",
            "Epoch 4/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.3444 - accuracy: 0.5696 - val_loss: 0.1430 - val_accuracy: 0.9630\n",
            "Epoch 5/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.3233 - accuracy: 0.5731 - val_loss: 0.1537 - val_accuracy: 0.9630\n",
            "Epoch 6/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.3167 - accuracy: 0.5731 - val_loss: 0.1636 - val_accuracy: 0.8889\n",
            "Epoch 7/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.3231 - accuracy: 0.5776 - val_loss: 0.2058 - val_accuracy: 0.8889\n",
            "Epoch 8/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.3077 - accuracy: 0.5776 - val_loss: 0.1627 - val_accuracy: 0.9259\n",
            "Epoch 9/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2946 - accuracy: 0.5811 - val_loss: 0.1305 - val_accuracy: 0.9630\n",
            "Epoch 10/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.3029 - accuracy: 0.5788 - val_loss: 0.1306 - val_accuracy: 0.9630\n",
            "Epoch 11/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2885 - accuracy: 0.5799 - val_loss: 0.1714 - val_accuracy: 0.9259\n",
            "Epoch 12/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2893 - accuracy: 0.5811 - val_loss: 0.1368 - val_accuracy: 0.9259\n",
            "Epoch 13/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2932 - accuracy: 0.5822 - val_loss: 0.1375 - val_accuracy: 0.9630\n",
            "Epoch 14/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2770 - accuracy: 0.5799 - val_loss: 0.1194 - val_accuracy: 0.9630\n",
            "Epoch 15/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2731 - accuracy: 0.5811 - val_loss: 0.0971 - val_accuracy: 0.9630\n",
            "Epoch 16/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2821 - accuracy: 0.5822 - val_loss: 0.1171 - val_accuracy: 0.9630\n",
            "Epoch 17/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2759 - accuracy: 0.5822 - val_loss: 0.1124 - val_accuracy: 0.9630\n",
            "Epoch 18/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2746 - accuracy: 0.5822 - val_loss: 0.1512 - val_accuracy: 0.9259\n",
            "Epoch 19/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2778 - accuracy: 0.5811 - val_loss: 0.1880 - val_accuracy: 0.8889\n",
            "Epoch 20/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2734 - accuracy: 0.5822 - val_loss: 0.1339 - val_accuracy: 0.9630\n",
            "Epoch 21/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2696 - accuracy: 0.5822 - val_loss: 0.1625 - val_accuracy: 0.9630\n",
            "Epoch 22/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2683 - accuracy: 0.5811 - val_loss: 0.1795 - val_accuracy: 0.9259\n",
            "Epoch 23/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2630 - accuracy: 0.5822 - val_loss: 0.1568 - val_accuracy: 0.9259\n",
            "Epoch 24/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2638 - accuracy: 0.5811 - val_loss: 0.1665 - val_accuracy: 0.8889\n",
            "Epoch 25/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2747 - accuracy: 0.5822 - val_loss: 0.1308 - val_accuracy: 0.9630\n",
            "Epoch 26/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2677 - accuracy: 0.5822 - val_loss: 0.1381 - val_accuracy: 0.9630\n",
            "Epoch 27/100\n",
            "28/28 [==============================] - 10s 369ms/step - loss: 0.2619 - accuracy: 0.5822 - val_loss: 0.1426 - val_accuracy: 0.9630\n",
            "Epoch 28/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2581 - accuracy: 0.5822 - val_loss: 0.1300 - val_accuracy: 0.9630\n",
            "Epoch 29/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2620 - accuracy: 0.5822 - val_loss: 0.1586 - val_accuracy: 0.8889\n",
            "Epoch 30/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2602 - accuracy: 0.5822 - val_loss: 0.1023 - val_accuracy: 0.9630\n",
            "2/2 [==============================] - 1s 121ms/step - loss: 0.1621 - accuracy: 0.9706\n",
            "[0.9764705777168274, 0.9941176414489746]\n",
            "[0.973400290736091, 0.9934299516908212]\n",
            "Epoch 1/100\n",
            "28/28 [==============================] - 16s 413ms/step - loss: 0.5204 - accuracy: 0.4875 - val_loss: 0.2745 - val_accuracy: 0.8519\n",
            "Epoch 2/100\n",
            "28/28 [==============================] - 11s 375ms/step - loss: 0.4049 - accuracy: 0.5499 - val_loss: 0.2440 - val_accuracy: 0.9630\n",
            "Epoch 3/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.3753 - accuracy: 0.5556 - val_loss: 0.2638 - val_accuracy: 0.9630\n",
            "Epoch 4/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.3398 - accuracy: 0.5680 - val_loss: 0.1922 - val_accuracy: 0.9630\n",
            "Epoch 5/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.3234 - accuracy: 0.5760 - val_loss: 0.2410 - val_accuracy: 0.9630\n",
            "Epoch 6/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.3232 - accuracy: 0.5771 - val_loss: 0.2108 - val_accuracy: 0.9630\n",
            "Epoch 7/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.3064 - accuracy: 0.5805 - val_loss: 0.1961 - val_accuracy: 0.9259\n",
            "Epoch 8/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.3017 - accuracy: 0.5816 - val_loss: 0.2096 - val_accuracy: 0.9259\n",
            "Epoch 9/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2951 - accuracy: 0.5828 - val_loss: 0.2209 - val_accuracy: 0.9259\n",
            "Epoch 10/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2857 - accuracy: 0.5828 - val_loss: 0.2816 - val_accuracy: 0.8519\n",
            "Epoch 11/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2937 - accuracy: 0.5828 - val_loss: 0.1672 - val_accuracy: 0.9259\n",
            "Epoch 12/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2883 - accuracy: 0.5805 - val_loss: 0.2003 - val_accuracy: 0.9259\n",
            "Epoch 13/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2764 - accuracy: 0.5850 - val_loss: 0.1276 - val_accuracy: 0.9630\n",
            "Epoch 14/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2878 - accuracy: 0.5805 - val_loss: 0.1253 - val_accuracy: 0.9630\n",
            "Epoch 15/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2828 - accuracy: 0.5839 - val_loss: 0.2739 - val_accuracy: 0.9259\n",
            "Epoch 16/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2685 - accuracy: 0.5850 - val_loss: 0.1622 - val_accuracy: 0.9630\n",
            "Epoch 17/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2763 - accuracy: 0.5839 - val_loss: 0.2299 - val_accuracy: 0.9630\n",
            "Epoch 18/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2727 - accuracy: 0.5839 - val_loss: 0.1520 - val_accuracy: 0.9259\n",
            "Epoch 19/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2672 - accuracy: 0.5850 - val_loss: 0.2009 - val_accuracy: 0.9259\n",
            "Epoch 20/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2701 - accuracy: 0.5850 - val_loss: 0.1700 - val_accuracy: 0.9630\n",
            "Epoch 21/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2617 - accuracy: 0.5850 - val_loss: 0.1723 - val_accuracy: 0.9259\n",
            "Epoch 22/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2629 - accuracy: 0.5839 - val_loss: 0.1607 - val_accuracy: 0.9630\n",
            "Epoch 23/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2618 - accuracy: 0.5850 - val_loss: 0.1389 - val_accuracy: 0.9630\n",
            "Epoch 24/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2593 - accuracy: 0.5850 - val_loss: 0.1210 - val_accuracy: 0.9630\n",
            "Epoch 25/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2669 - accuracy: 0.5850 - val_loss: 0.1313 - val_accuracy: 0.9630\n",
            "Epoch 26/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2659 - accuracy: 0.5850 - val_loss: 0.1682 - val_accuracy: 0.9259\n",
            "Epoch 27/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2680 - accuracy: 0.5850 - val_loss: 0.4076 - val_accuracy: 0.8889\n",
            "Epoch 28/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2700 - accuracy: 0.5850 - val_loss: 0.2120 - val_accuracy: 0.9630\n",
            "Epoch 29/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2644 - accuracy: 0.5850 - val_loss: 0.1930 - val_accuracy: 0.9259\n",
            "Epoch 30/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2603 - accuracy: 0.5850 - val_loss: 0.2032 - val_accuracy: 0.9259\n",
            "Epoch 31/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2568 - accuracy: 0.5850 - val_loss: 0.1944 - val_accuracy: 0.9259\n",
            "Epoch 32/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2582 - accuracy: 0.5850 - val_loss: 0.1687 - val_accuracy: 0.9259\n",
            "Epoch 33/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2583 - accuracy: 0.5850 - val_loss: 0.1398 - val_accuracy: 0.9630\n",
            "Epoch 34/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2541 - accuracy: 0.5850 - val_loss: 0.1874 - val_accuracy: 0.9259\n",
            "Epoch 35/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2573 - accuracy: 0.5850 - val_loss: 0.1642 - val_accuracy: 0.9259\n",
            "Epoch 36/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2585 - accuracy: 0.5850 - val_loss: 0.1623 - val_accuracy: 0.9630\n",
            "Epoch 37/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2524 - accuracy: 0.5850 - val_loss: 0.1399 - val_accuracy: 0.9630\n",
            "Epoch 38/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2490 - accuracy: 0.5850 - val_loss: 0.1427 - val_accuracy: 0.9630\n",
            "Epoch 39/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2511 - accuracy: 0.5850 - val_loss: 0.1423 - val_accuracy: 0.9259\n",
            "2/2 [==============================] - 1s 95ms/step - loss: 0.1451 - accuracy: 0.9706\n",
            "Epoch 1/100\n",
            "28/28 [==============================] - 15s 411ms/step - loss: 0.4914 - accuracy: 0.5204 - val_loss: 0.6511 - val_accuracy: 0.8148\n",
            "Epoch 2/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.3860 - accuracy: 0.5714 - val_loss: 0.3981 - val_accuracy: 0.8148\n",
            "Epoch 3/100\n",
            "28/28 [==============================] - 11s 375ms/step - loss: 0.3473 - accuracy: 0.5794 - val_loss: 0.2774 - val_accuracy: 0.8519\n",
            "Epoch 4/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.3437 - accuracy: 0.5884 - val_loss: 0.2760 - val_accuracy: 0.8519\n",
            "Epoch 5/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.3106 - accuracy: 0.5975 - val_loss: 0.2725 - val_accuracy: 0.7778\n",
            "Epoch 6/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.3005 - accuracy: 0.6054 - val_loss: 0.2306 - val_accuracy: 0.8519\n",
            "Epoch 7/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.3002 - accuracy: 0.5975 - val_loss: 0.2384 - val_accuracy: 0.8889\n",
            "Epoch 8/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2910 - accuracy: 0.6054 - val_loss: 0.1946 - val_accuracy: 0.8889\n",
            "Epoch 9/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2897 - accuracy: 0.6043 - val_loss: 0.2889 - val_accuracy: 0.8519\n",
            "Epoch 10/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2820 - accuracy: 0.6020 - val_loss: 0.2096 - val_accuracy: 0.8519\n",
            "Epoch 11/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2849 - accuracy: 0.6032 - val_loss: 0.1822 - val_accuracy: 0.8889\n",
            "Epoch 12/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2777 - accuracy: 0.6043 - val_loss: 0.2081 - val_accuracy: 0.8889\n",
            "Epoch 13/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2732 - accuracy: 0.6032 - val_loss: 0.1811 - val_accuracy: 0.8889\n",
            "Epoch 14/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2727 - accuracy: 0.6054 - val_loss: 0.1437 - val_accuracy: 0.9630\n",
            "Epoch 15/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2639 - accuracy: 0.6043 - val_loss: 0.1790 - val_accuracy: 0.9259\n",
            "Epoch 16/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2674 - accuracy: 0.6043 - val_loss: 0.1307 - val_accuracy: 0.9630\n",
            "Epoch 17/100\n",
            "28/28 [==============================] - 11s 375ms/step - loss: 0.2585 - accuracy: 0.6054 - val_loss: 0.1284 - val_accuracy: 0.9630\n",
            "Epoch 18/100\n",
            "28/28 [==============================] - 11s 375ms/step - loss: 0.2622 - accuracy: 0.6043 - val_loss: 0.2007 - val_accuracy: 0.8519\n",
            "Epoch 19/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2571 - accuracy: 0.6043 - val_loss: 0.1838 - val_accuracy: 0.8889\n",
            "Epoch 20/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2665 - accuracy: 0.6054 - val_loss: 0.1353 - val_accuracy: 0.9630\n",
            "Epoch 21/100\n",
            "28/28 [==============================] - 11s 375ms/step - loss: 0.2528 - accuracy: 0.6054 - val_loss: 0.1883 - val_accuracy: 0.9259\n",
            "Epoch 22/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2605 - accuracy: 0.6054 - val_loss: 0.1599 - val_accuracy: 0.9259\n",
            "Epoch 23/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2596 - accuracy: 0.6054 - val_loss: 0.1919 - val_accuracy: 0.8889\n",
            "Epoch 24/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2554 - accuracy: 0.6054 - val_loss: 0.1796 - val_accuracy: 0.9259\n",
            "Epoch 25/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2550 - accuracy: 0.6054 - val_loss: 0.1380 - val_accuracy: 0.9630\n",
            "Epoch 26/100\n",
            "28/28 [==============================] - 11s 375ms/step - loss: 0.2530 - accuracy: 0.6054 - val_loss: 0.1302 - val_accuracy: 0.9630\n",
            "Epoch 27/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2512 - accuracy: 0.6043 - val_loss: 0.1381 - val_accuracy: 1.0000\n",
            "Epoch 28/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2513 - accuracy: 0.6054 - val_loss: 0.1010 - val_accuracy: 0.9630\n",
            "Epoch 29/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2508 - accuracy: 0.6054 - val_loss: 0.1565 - val_accuracy: 0.9630\n",
            "Epoch 30/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2399 - accuracy: 0.6054 - val_loss: 0.1279 - val_accuracy: 0.9630\n",
            "Epoch 31/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2474 - accuracy: 0.6054 - val_loss: 0.1490 - val_accuracy: 0.9630\n",
            "Epoch 32/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2448 - accuracy: 0.6054 - val_loss: 0.1482 - val_accuracy: 0.9630\n",
            "Epoch 33/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2507 - accuracy: 0.6054 - val_loss: 0.1848 - val_accuracy: 0.9630\n",
            "Epoch 34/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2397 - accuracy: 0.6054 - val_loss: 0.1283 - val_accuracy: 0.9630\n",
            "Epoch 35/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2429 - accuracy: 0.6054 - val_loss: 0.1956 - val_accuracy: 0.9630\n",
            "Epoch 36/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2368 - accuracy: 0.6054 - val_loss: 0.1576 - val_accuracy: 0.9630\n",
            "Epoch 37/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2418 - accuracy: 0.6054 - val_loss: 0.1660 - val_accuracy: 0.9630\n",
            "Epoch 38/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2453 - accuracy: 0.6054 - val_loss: 0.2253 - val_accuracy: 0.8889\n",
            "Epoch 39/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2430 - accuracy: 0.6054 - val_loss: 0.1366 - val_accuracy: 0.9630\n",
            "Epoch 40/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2436 - accuracy: 0.6054 - val_loss: 0.1514 - val_accuracy: 0.9630\n",
            "Epoch 41/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2412 - accuracy: 0.6054 - val_loss: 0.1996 - val_accuracy: 0.8889\n",
            "Epoch 42/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2411 - accuracy: 0.6054 - val_loss: 0.1568 - val_accuracy: 0.9630\n",
            "Epoch 43/100\n",
            "28/28 [==============================] - 11s 378ms/step - loss: 0.2415 - accuracy: 0.6054 - val_loss: 0.1521 - val_accuracy: 0.9259\n",
            "2/2 [==============================] - 1s 115ms/step - loss: 0.0596 - accuracy: 1.0000\n",
            "Epoch 1/100\n",
            "28/28 [==============================] - 16s 410ms/step - loss: 0.4742 - accuracy: 0.5377 - val_loss: 0.3258 - val_accuracy: 0.8148\n",
            "Epoch 2/100\n",
            "28/28 [==============================] - 10s 368ms/step - loss: 0.3800 - accuracy: 0.5765 - val_loss: 0.2613 - val_accuracy: 0.8519\n",
            "Epoch 3/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.3591 - accuracy: 0.5868 - val_loss: 0.2004 - val_accuracy: 1.0000\n",
            "Epoch 4/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.3382 - accuracy: 0.5822 - val_loss: 0.1738 - val_accuracy: 1.0000\n",
            "Epoch 5/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.3281 - accuracy: 0.5936 - val_loss: 0.2278 - val_accuracy: 0.8519\n",
            "Epoch 6/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.3220 - accuracy: 0.5890 - val_loss: 0.1471 - val_accuracy: 1.0000\n",
            "Epoch 7/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.3095 - accuracy: 0.5993 - val_loss: 0.1564 - val_accuracy: 0.9259\n",
            "Epoch 8/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.3062 - accuracy: 0.5913 - val_loss: 0.1398 - val_accuracy: 0.9630\n",
            "Epoch 9/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2835 - accuracy: 0.5993 - val_loss: 0.1414 - val_accuracy: 0.9630\n",
            "Epoch 10/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2835 - accuracy: 0.5982 - val_loss: 0.1103 - val_accuracy: 0.9630\n",
            "Epoch 11/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2824 - accuracy: 0.5982 - val_loss: 0.0951 - val_accuracy: 1.0000\n",
            "Epoch 12/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2897 - accuracy: 0.6005 - val_loss: 0.1042 - val_accuracy: 0.9630\n",
            "Epoch 13/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2766 - accuracy: 0.5982 - val_loss: 0.1466 - val_accuracy: 0.9259\n",
            "Epoch 14/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2739 - accuracy: 0.5982 - val_loss: 0.1045 - val_accuracy: 1.0000\n",
            "Epoch 15/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2788 - accuracy: 0.5982 - val_loss: 0.1789 - val_accuracy: 0.8889\n",
            "Epoch 16/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2722 - accuracy: 0.5982 - val_loss: 0.0806 - val_accuracy: 1.0000\n",
            "Epoch 17/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2712 - accuracy: 0.5993 - val_loss: 0.1076 - val_accuracy: 1.0000\n",
            "Epoch 18/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2668 - accuracy: 0.5993 - val_loss: 0.1234 - val_accuracy: 1.0000\n",
            "Epoch 19/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2717 - accuracy: 0.6005 - val_loss: 0.0682 - val_accuracy: 1.0000\n",
            "Epoch 20/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2683 - accuracy: 0.6005 - val_loss: 0.0537 - val_accuracy: 1.0000\n",
            "Epoch 21/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2641 - accuracy: 0.6005 - val_loss: 0.0922 - val_accuracy: 1.0000\n",
            "Epoch 22/100\n",
            "28/28 [==============================] - 10s 368ms/step - loss: 0.2578 - accuracy: 0.6005 - val_loss: 0.0648 - val_accuracy: 1.0000\n",
            "Epoch 23/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2625 - accuracy: 0.6005 - val_loss: 0.0788 - val_accuracy: 1.0000\n",
            "Epoch 24/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2483 - accuracy: 0.6005 - val_loss: 0.0603 - val_accuracy: 1.0000\n",
            "Epoch 25/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2513 - accuracy: 0.6005 - val_loss: 0.0885 - val_accuracy: 1.0000\n",
            "Epoch 26/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2547 - accuracy: 0.6005 - val_loss: 0.0828 - val_accuracy: 1.0000\n",
            "Epoch 27/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2531 - accuracy: 0.6005 - val_loss: 0.0871 - val_accuracy: 1.0000\n",
            "Epoch 28/100\n",
            "28/28 [==============================] - 10s 368ms/step - loss: 0.2529 - accuracy: 0.6005 - val_loss: 0.0978 - val_accuracy: 1.0000\n",
            "Epoch 29/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2506 - accuracy: 0.6005 - val_loss: 0.1575 - val_accuracy: 0.9630\n",
            "Epoch 30/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2569 - accuracy: 0.6005 - val_loss: 0.0617 - val_accuracy: 1.0000\n",
            "Epoch 31/100\n",
            "28/28 [==============================] - 10s 369ms/step - loss: 0.2534 - accuracy: 0.5993 - val_loss: 0.0883 - val_accuracy: 1.0000\n",
            "Epoch 32/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2477 - accuracy: 0.6005 - val_loss: 0.1202 - val_accuracy: 1.0000\n",
            "Epoch 33/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2394 - accuracy: 0.6005 - val_loss: 0.0826 - val_accuracy: 1.0000\n",
            "Epoch 34/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2567 - accuracy: 0.6005 - val_loss: 0.1276 - val_accuracy: 0.9630\n",
            "Epoch 35/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2398 - accuracy: 0.6005 - val_loss: 0.0783 - val_accuracy: 1.0000\n",
            "2/2 [==============================] - 1s 93ms/step - loss: 0.1348 - accuracy: 0.9706\n",
            "Epoch 1/100\n",
            "28/28 [==============================] - 16s 412ms/step - loss: 0.5392 - accuracy: 0.4863 - val_loss: 0.8374 - val_accuracy: 0.7778\n",
            "Epoch 2/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.4235 - accuracy: 0.5320 - val_loss: 0.4117 - val_accuracy: 0.8148\n",
            "Epoch 3/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.3935 - accuracy: 0.5525 - val_loss: 0.3638 - val_accuracy: 0.8148\n",
            "Epoch 4/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.3613 - accuracy: 0.5605 - val_loss: 0.3172 - val_accuracy: 0.8148\n",
            "Epoch 5/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.3404 - accuracy: 0.5628 - val_loss: 0.2975 - val_accuracy: 0.8519\n",
            "Epoch 6/100\n",
            "28/28 [==============================] - 11s 378ms/step - loss: 0.3432 - accuracy: 0.5651 - val_loss: 0.2864 - val_accuracy: 0.8148\n",
            "Epoch 7/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.3117 - accuracy: 0.5719 - val_loss: 0.2195 - val_accuracy: 0.9630\n",
            "Epoch 8/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.3266 - accuracy: 0.5662 - val_loss: 0.3607 - val_accuracy: 0.8889\n",
            "Epoch 9/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.3225 - accuracy: 0.5674 - val_loss: 0.1714 - val_accuracy: 1.0000\n",
            "Epoch 10/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.3104 - accuracy: 0.5696 - val_loss: 0.1638 - val_accuracy: 0.9630\n",
            "Epoch 11/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.3021 - accuracy: 0.5719 - val_loss: 0.1667 - val_accuracy: 0.9630\n",
            "Epoch 12/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2979 - accuracy: 0.5696 - val_loss: 0.2001 - val_accuracy: 0.9630\n",
            "Epoch 13/100\n",
            "28/28 [==============================] - 11s 378ms/step - loss: 0.2973 - accuracy: 0.5719 - val_loss: 0.1489 - val_accuracy: 0.9259\n",
            "Epoch 14/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2894 - accuracy: 0.5731 - val_loss: 0.2162 - val_accuracy: 0.9630\n",
            "Epoch 15/100\n",
            "28/28 [==============================] - 11s 375ms/step - loss: 0.2903 - accuracy: 0.5731 - val_loss: 0.1602 - val_accuracy: 0.9630\n",
            "Epoch 16/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.3016 - accuracy: 0.5708 - val_loss: 0.1557 - val_accuracy: 0.9630\n",
            "Epoch 17/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2854 - accuracy: 0.5719 - val_loss: 0.1238 - val_accuracy: 0.9630\n",
            "Epoch 18/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2808 - accuracy: 0.5731 - val_loss: 0.1542 - val_accuracy: 0.9259\n",
            "Epoch 19/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2826 - accuracy: 0.5731 - val_loss: 0.3370 - val_accuracy: 0.8889\n",
            "Epoch 20/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2822 - accuracy: 0.5731 - val_loss: 0.2238 - val_accuracy: 0.9259\n",
            "Epoch 21/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2791 - accuracy: 0.5731 - val_loss: 0.2536 - val_accuracy: 0.9630\n",
            "Epoch 22/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2772 - accuracy: 0.5731 - val_loss: 0.1825 - val_accuracy: 0.9630\n",
            "Epoch 23/100\n",
            "28/28 [==============================] - 10s 369ms/step - loss: 0.2719 - accuracy: 0.5731 - val_loss: 0.1239 - val_accuracy: 0.9630\n",
            "Epoch 24/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2761 - accuracy: 0.5719 - val_loss: 0.1110 - val_accuracy: 0.9259\n",
            "Epoch 25/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2786 - accuracy: 0.5719 - val_loss: 0.1671 - val_accuracy: 0.9630\n",
            "Epoch 26/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2777 - accuracy: 0.5731 - val_loss: 0.1403 - val_accuracy: 0.9259\n",
            "Epoch 27/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2710 - accuracy: 0.5719 - val_loss: 0.1495 - val_accuracy: 0.9630\n",
            "Epoch 28/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2725 - accuracy: 0.5731 - val_loss: 0.1672 - val_accuracy: 0.9630\n",
            "Epoch 29/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2774 - accuracy: 0.5731 - val_loss: 0.2213 - val_accuracy: 0.9259\n",
            "Epoch 30/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2741 - accuracy: 0.5731 - val_loss: 0.1989 - val_accuracy: 0.8889\n",
            "Epoch 31/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2639 - accuracy: 0.5719 - val_loss: 0.1534 - val_accuracy: 0.9259\n",
            "Epoch 32/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2661 - accuracy: 0.5731 - val_loss: 0.1323 - val_accuracy: 0.9630\n",
            "Epoch 33/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2653 - accuracy: 0.5731 - val_loss: 0.1587 - val_accuracy: 0.9630\n",
            "Epoch 34/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2641 - accuracy: 0.5731 - val_loss: 0.2699 - val_accuracy: 0.9259\n",
            "Epoch 35/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2752 - accuracy: 0.5731 - val_loss: 0.2115 - val_accuracy: 0.9259\n",
            "Epoch 36/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2686 - accuracy: 0.5731 - val_loss: 0.1219 - val_accuracy: 0.9259\n",
            "Epoch 37/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2641 - accuracy: 0.5731 - val_loss: 0.1756 - val_accuracy: 0.9259\n",
            "Epoch 38/100\n",
            "28/28 [==============================] - 11s 375ms/step - loss: 0.2568 - accuracy: 0.5731 - val_loss: 0.2771 - val_accuracy: 0.9259\n",
            "Epoch 39/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2656 - accuracy: 0.5731 - val_loss: 0.1923 - val_accuracy: 0.8889\n",
            "2/2 [==============================] - 1s 92ms/step - loss: 0.1226 - accuracy: 0.9412\n",
            "Epoch 1/100\n",
            "28/28 [==============================] - 16s 415ms/step - loss: 0.5094 - accuracy: 0.5000 - val_loss: 0.2379 - val_accuracy: 0.9259\n",
            "Epoch 2/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.3773 - accuracy: 0.5526 - val_loss: 0.2620 - val_accuracy: 0.9259\n",
            "Epoch 3/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.3638 - accuracy: 0.5559 - val_loss: 0.2052 - val_accuracy: 0.9259\n",
            "Epoch 4/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.3410 - accuracy: 0.5682 - val_loss: 0.2111 - val_accuracy: 0.9259\n",
            "Epoch 5/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.3377 - accuracy: 0.5626 - val_loss: 0.2161 - val_accuracy: 0.9259\n",
            "Epoch 6/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.3276 - accuracy: 0.5660 - val_loss: 0.2900 - val_accuracy: 0.8519\n",
            "Epoch 7/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.3267 - accuracy: 0.5649 - val_loss: 0.2329 - val_accuracy: 0.9259\n",
            "Epoch 8/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.3076 - accuracy: 0.5682 - val_loss: 0.2409 - val_accuracy: 0.9259\n",
            "Epoch 9/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.3128 - accuracy: 0.5705 - val_loss: 0.2499 - val_accuracy: 0.8889\n",
            "Epoch 10/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2987 - accuracy: 0.5716 - val_loss: 0.2320 - val_accuracy: 0.9259\n",
            "Epoch 11/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2898 - accuracy: 0.5716 - val_loss: 0.1800 - val_accuracy: 0.9630\n",
            "Epoch 12/100\n",
            "28/28 [==============================] - 11s 378ms/step - loss: 0.2954 - accuracy: 0.5716 - val_loss: 0.1688 - val_accuracy: 0.9630\n",
            "Epoch 13/100\n",
            "28/28 [==============================] - 10s 376ms/step - loss: 0.2985 - accuracy: 0.5716 - val_loss: 0.2350 - val_accuracy: 0.9259\n",
            "Epoch 14/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2888 - accuracy: 0.5727 - val_loss: 0.1852 - val_accuracy: 0.9259\n",
            "Epoch 15/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2820 - accuracy: 0.5716 - val_loss: 0.2251 - val_accuracy: 0.9259\n",
            "Epoch 16/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2902 - accuracy: 0.5716 - val_loss: 0.2287 - val_accuracy: 0.9259\n",
            "Epoch 17/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2831 - accuracy: 0.5727 - val_loss: 0.2021 - val_accuracy: 0.9259\n",
            "Epoch 18/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2891 - accuracy: 0.5727 - val_loss: 0.2426 - val_accuracy: 0.9259\n",
            "Epoch 19/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2763 - accuracy: 0.5727 - val_loss: 0.1981 - val_accuracy: 0.9259\n",
            "Epoch 20/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2850 - accuracy: 0.5727 - val_loss: 0.1865 - val_accuracy: 0.9630\n",
            "Epoch 21/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2771 - accuracy: 0.5727 - val_loss: 0.1677 - val_accuracy: 0.9259\n",
            "Epoch 22/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2699 - accuracy: 0.5727 - val_loss: 0.1368 - val_accuracy: 0.9630\n",
            "Epoch 23/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2817 - accuracy: 0.5727 - val_loss: 0.1764 - val_accuracy: 0.9259\n",
            "Epoch 24/100\n",
            "28/28 [==============================] - 11s 379ms/step - loss: 0.2733 - accuracy: 0.5727 - val_loss: 0.1339 - val_accuracy: 0.9630\n",
            "Epoch 25/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2728 - accuracy: 0.5727 - val_loss: 0.1814 - val_accuracy: 0.9259\n",
            "Epoch 26/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2714 - accuracy: 0.5727 - val_loss: 0.1864 - val_accuracy: 0.8889\n",
            "Epoch 27/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2721 - accuracy: 0.5727 - val_loss: 0.1790 - val_accuracy: 0.9259\n",
            "Epoch 28/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2731 - accuracy: 0.5727 - val_loss: 0.1934 - val_accuracy: 0.9630\n",
            "Epoch 29/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2703 - accuracy: 0.5727 - val_loss: 0.2821 - val_accuracy: 0.8519\n",
            "Epoch 30/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2780 - accuracy: 0.5727 - val_loss: 0.2265 - val_accuracy: 0.8889\n",
            "Epoch 31/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2689 - accuracy: 0.5727 - val_loss: 0.2103 - val_accuracy: 0.8519\n",
            "Epoch 32/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2674 - accuracy: 0.5727 - val_loss: 0.2084 - val_accuracy: 0.9630\n",
            "Epoch 33/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2646 - accuracy: 0.5727 - val_loss: 0.1931 - val_accuracy: 0.9259\n",
            "Epoch 34/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2654 - accuracy: 0.5727 - val_loss: 0.2031 - val_accuracy: 0.8889\n",
            "Epoch 35/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2686 - accuracy: 0.5727 - val_loss: 0.1985 - val_accuracy: 0.8889\n",
            "Epoch 36/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2601 - accuracy: 0.5727 - val_loss: 0.1902 - val_accuracy: 0.9630\n",
            "Epoch 37/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2575 - accuracy: 0.5727 - val_loss: 0.1756 - val_accuracy: 0.9259\n",
            "Epoch 38/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2604 - accuracy: 0.5727 - val_loss: 0.1986 - val_accuracy: 0.9259\n",
            "Epoch 39/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2596 - accuracy: 0.5727 - val_loss: 0.2003 - val_accuracy: 0.9630\n",
            "2/2 [==============================] - 1s 97ms/step - loss: 0.1332 - accuracy: 0.9706\n",
            "[0.9764705777168274, 0.9941176414489746, 0.970588219165802]\n",
            "[0.973400290736091, 0.9934299516908212, 0.9670915511664816]\n",
            "Epoch 1/100\n",
            "28/28 [==============================] - 16s 411ms/step - loss: 0.5170 - accuracy: 0.5080 - val_loss: 0.1742 - val_accuracy: 0.9630\n",
            "Epoch 2/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.4101 - accuracy: 0.5537 - val_loss: 0.2392 - val_accuracy: 0.9259\n",
            "Epoch 3/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.3848 - accuracy: 0.5685 - val_loss: 0.2056 - val_accuracy: 0.9630\n",
            "Epoch 4/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.3533 - accuracy: 0.5788 - val_loss: 0.2710 - val_accuracy: 0.8519\n",
            "Epoch 5/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.3388 - accuracy: 0.5868 - val_loss: 0.2125 - val_accuracy: 0.9259\n",
            "Epoch 6/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.3253 - accuracy: 0.5890 - val_loss: 0.1981 - val_accuracy: 0.9630\n",
            "Epoch 7/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.3154 - accuracy: 0.5913 - val_loss: 0.1708 - val_accuracy: 1.0000\n",
            "Epoch 8/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.3106 - accuracy: 0.5913 - val_loss: 0.1632 - val_accuracy: 1.0000\n",
            "Epoch 9/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2931 - accuracy: 0.5936 - val_loss: 0.1113 - val_accuracy: 1.0000\n",
            "Epoch 10/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.3032 - accuracy: 0.5925 - val_loss: 0.1650 - val_accuracy: 0.9630\n",
            "Epoch 11/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2939 - accuracy: 0.5890 - val_loss: 0.1210 - val_accuracy: 1.0000\n",
            "Epoch 12/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2861 - accuracy: 0.5959 - val_loss: 0.1292 - val_accuracy: 1.0000\n",
            "Epoch 13/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2862 - accuracy: 0.5959 - val_loss: 0.0844 - val_accuracy: 0.9630\n",
            "Epoch 14/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2857 - accuracy: 0.5925 - val_loss: 0.1117 - val_accuracy: 0.9630\n",
            "Epoch 15/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2863 - accuracy: 0.5947 - val_loss: 0.0860 - val_accuracy: 0.9630\n",
            "Epoch 16/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2854 - accuracy: 0.5947 - val_loss: 0.1740 - val_accuracy: 0.9259\n",
            "Epoch 17/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2772 - accuracy: 0.5947 - val_loss: 0.1842 - val_accuracy: 0.9259\n",
            "Epoch 18/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2767 - accuracy: 0.5959 - val_loss: 0.2952 - val_accuracy: 0.9259\n",
            "Epoch 19/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2779 - accuracy: 0.5936 - val_loss: 0.1693 - val_accuracy: 0.9259\n",
            "Epoch 20/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2768 - accuracy: 0.5959 - val_loss: 0.1065 - val_accuracy: 0.9630\n",
            "Epoch 21/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2658 - accuracy: 0.5959 - val_loss: 0.1576 - val_accuracy: 0.9630\n",
            "Epoch 22/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2689 - accuracy: 0.5959 - val_loss: 0.0954 - val_accuracy: 0.9630\n",
            "Epoch 23/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2613 - accuracy: 0.5959 - val_loss: 0.0952 - val_accuracy: 0.9630\n",
            "Epoch 24/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2620 - accuracy: 0.5959 - val_loss: 0.1617 - val_accuracy: 0.9259\n",
            "Epoch 25/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2660 - accuracy: 0.5959 - val_loss: 0.2263 - val_accuracy: 0.9259\n",
            "Epoch 26/100\n",
            "28/28 [==============================] - 11s 375ms/step - loss: 0.2688 - accuracy: 0.5947 - val_loss: 0.0966 - val_accuracy: 0.9630\n",
            "Epoch 27/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2657 - accuracy: 0.5959 - val_loss: 0.1404 - val_accuracy: 0.9630\n",
            "Epoch 28/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2614 - accuracy: 0.5959 - val_loss: 0.1852 - val_accuracy: 0.9259\n",
            "2/2 [==============================] - 1s 106ms/step - loss: 0.1713 - accuracy: 0.9412\n",
            "Epoch 1/100\n",
            "27/27 [==============================] - 15s 414ms/step - loss: 0.4751 - accuracy: 0.5417 - val_loss: 0.3422 - val_accuracy: 0.8148\n",
            "Epoch 2/100\n",
            "27/27 [==============================] - 10s 375ms/step - loss: 0.3598 - accuracy: 0.5926 - val_loss: 0.4960 - val_accuracy: 0.7407\n",
            "Epoch 3/100\n",
            "27/27 [==============================] - 10s 379ms/step - loss: 0.3374 - accuracy: 0.5903 - val_loss: 0.2534 - val_accuracy: 0.9259\n",
            "Epoch 4/100\n",
            "27/27 [==============================] - 10s 376ms/step - loss: 0.3225 - accuracy: 0.6007 - val_loss: 0.3292 - val_accuracy: 0.8148\n",
            "Epoch 5/100\n",
            "27/27 [==============================] - 10s 378ms/step - loss: 0.3066 - accuracy: 0.6053 - val_loss: 0.2823 - val_accuracy: 0.8519\n",
            "Epoch 6/100\n",
            "27/27 [==============================] - 10s 375ms/step - loss: 0.2947 - accuracy: 0.6019 - val_loss: 0.2892 - val_accuracy: 0.8148\n",
            "Epoch 7/100\n",
            "27/27 [==============================] - 10s 377ms/step - loss: 0.2992 - accuracy: 0.6065 - val_loss: 0.2582 - val_accuracy: 0.9259\n",
            "Epoch 8/100\n",
            "27/27 [==============================] - 10s 376ms/step - loss: 0.2942 - accuracy: 0.6053 - val_loss: 0.2872 - val_accuracy: 0.8148\n",
            "Epoch 9/100\n",
            "27/27 [==============================] - 10s 375ms/step - loss: 0.2869 - accuracy: 0.6065 - val_loss: 0.2859 - val_accuracy: 0.8519\n",
            "Epoch 10/100\n",
            "27/27 [==============================] - 10s 375ms/step - loss: 0.2957 - accuracy: 0.6076 - val_loss: 0.2641 - val_accuracy: 0.8519\n",
            "Epoch 11/100\n",
            "27/27 [==============================] - 10s 378ms/step - loss: 0.2793 - accuracy: 0.6088 - val_loss: 0.2089 - val_accuracy: 0.9259\n",
            "Epoch 12/100\n",
            "27/27 [==============================] - 10s 375ms/step - loss: 0.2873 - accuracy: 0.6076 - val_loss: 0.2568 - val_accuracy: 0.8889\n",
            "Epoch 13/100\n",
            "27/27 [==============================] - 10s 378ms/step - loss: 0.2678 - accuracy: 0.6100 - val_loss: 0.1401 - val_accuracy: 0.9630\n",
            "Epoch 14/100\n",
            "27/27 [==============================] - 10s 374ms/step - loss: 0.2712 - accuracy: 0.6111 - val_loss: 0.2158 - val_accuracy: 0.8889\n",
            "Epoch 15/100\n",
            "27/27 [==============================] - 10s 375ms/step - loss: 0.2665 - accuracy: 0.6100 - val_loss: 0.2327 - val_accuracy: 0.8519\n",
            "Epoch 16/100\n",
            "27/27 [==============================] - 10s 374ms/step - loss: 0.2705 - accuracy: 0.6100 - val_loss: 0.2853 - val_accuracy: 0.7778\n",
            "Epoch 17/100\n",
            "27/27 [==============================] - 10s 377ms/step - loss: 0.2676 - accuracy: 0.6100 - val_loss: 0.1929 - val_accuracy: 0.9259\n",
            "Epoch 18/100\n",
            "27/27 [==============================] - 10s 376ms/step - loss: 0.2713 - accuracy: 0.6111 - val_loss: 0.1556 - val_accuracy: 1.0000\n",
            "Epoch 19/100\n",
            "27/27 [==============================] - 10s 373ms/step - loss: 0.2671 - accuracy: 0.6100 - val_loss: 0.1909 - val_accuracy: 0.8889\n",
            "Epoch 20/100\n",
            "27/27 [==============================] - 10s 376ms/step - loss: 0.2540 - accuracy: 0.6111 - val_loss: 0.1949 - val_accuracy: 0.9630\n",
            "Epoch 21/100\n",
            "27/27 [==============================] - 10s 376ms/step - loss: 0.2524 - accuracy: 0.6111 - val_loss: 0.1725 - val_accuracy: 0.9630\n",
            "Epoch 22/100\n",
            "27/27 [==============================] - 10s 372ms/step - loss: 0.2527 - accuracy: 0.6111 - val_loss: 0.1555 - val_accuracy: 0.9630\n",
            "Epoch 23/100\n",
            "27/27 [==============================] - 10s 376ms/step - loss: 0.2511 - accuracy: 0.6111 - val_loss: 0.2465 - val_accuracy: 0.8889\n",
            "Epoch 24/100\n",
            "27/27 [==============================] - 10s 373ms/step - loss: 0.2549 - accuracy: 0.6111 - val_loss: 0.2314 - val_accuracy: 0.9630\n",
            "Epoch 25/100\n",
            "27/27 [==============================] - 10s 371ms/step - loss: 0.2585 - accuracy: 0.6111 - val_loss: 0.2099 - val_accuracy: 0.8519\n",
            "Epoch 26/100\n",
            "27/27 [==============================] - 10s 371ms/step - loss: 0.2484 - accuracy: 0.6111 - val_loss: 0.2151 - val_accuracy: 0.8889\n",
            "Epoch 27/100\n",
            "27/27 [==============================] - 10s 372ms/step - loss: 0.2620 - accuracy: 0.6111 - val_loss: 0.1741 - val_accuracy: 0.9259\n",
            "Epoch 28/100\n",
            "27/27 [==============================] - 10s 373ms/step - loss: 0.2568 - accuracy: 0.6100 - val_loss: 0.2627 - val_accuracy: 0.9259\n",
            "2/2 [==============================] - 1s 113ms/step - loss: 0.1038 - accuracy: 1.0000\n",
            "Epoch 1/100\n",
            "28/28 [==============================] - 16s 412ms/step - loss: 0.5083 - accuracy: 0.4897 - val_loss: 0.4343 - val_accuracy: 0.8148\n",
            "Epoch 2/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.4300 - accuracy: 0.5253 - val_loss: 0.4842 - val_accuracy: 0.7407\n",
            "Epoch 3/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.4063 - accuracy: 0.5230 - val_loss: 0.3082 - val_accuracy: 0.8519\n",
            "Epoch 4/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.3708 - accuracy: 0.5402 - val_loss: 0.2732 - val_accuracy: 0.9259\n",
            "Epoch 5/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.3479 - accuracy: 0.5448 - val_loss: 0.2770 - val_accuracy: 0.9259\n",
            "Epoch 6/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.3502 - accuracy: 0.5425 - val_loss: 0.2848 - val_accuracy: 0.8889\n",
            "Epoch 7/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.3374 - accuracy: 0.5517 - val_loss: 0.2237 - val_accuracy: 0.9259\n",
            "Epoch 8/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.3244 - accuracy: 0.5540 - val_loss: 0.2527 - val_accuracy: 0.9259\n",
            "Epoch 9/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.3289 - accuracy: 0.5529 - val_loss: 0.2401 - val_accuracy: 0.9259\n",
            "Epoch 10/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.3147 - accuracy: 0.5529 - val_loss: 0.2195 - val_accuracy: 0.8889\n",
            "Epoch 11/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.3151 - accuracy: 0.5517 - val_loss: 0.2094 - val_accuracy: 0.9259\n",
            "Epoch 12/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.3088 - accuracy: 0.5563 - val_loss: 0.2519 - val_accuracy: 0.9630\n",
            "Epoch 13/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.3134 - accuracy: 0.5540 - val_loss: 0.2076 - val_accuracy: 0.9630\n",
            "Epoch 14/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.3127 - accuracy: 0.5540 - val_loss: 0.1930 - val_accuracy: 0.9630\n",
            "Epoch 15/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2952 - accuracy: 0.5552 - val_loss: 0.3046 - val_accuracy: 0.8519\n",
            "Epoch 16/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.3024 - accuracy: 0.5563 - val_loss: 0.3513 - val_accuracy: 0.7778\n",
            "Epoch 17/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.3036 - accuracy: 0.5540 - val_loss: 0.2276 - val_accuracy: 0.9259\n",
            "Epoch 18/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2968 - accuracy: 0.5552 - val_loss: 0.1693 - val_accuracy: 1.0000\n",
            "Epoch 19/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2952 - accuracy: 0.5552 - val_loss: 0.2206 - val_accuracy: 0.9630\n",
            "Epoch 20/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2878 - accuracy: 0.5552 - val_loss: 0.1204 - val_accuracy: 0.9259\n",
            "Epoch 21/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2912 - accuracy: 0.5563 - val_loss: 0.1495 - val_accuracy: 0.9259\n",
            "Epoch 22/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.3010 - accuracy: 0.5563 - val_loss: 0.1628 - val_accuracy: 0.9259\n",
            "Epoch 23/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2888 - accuracy: 0.5540 - val_loss: 0.2818 - val_accuracy: 0.8519\n",
            "Epoch 24/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2827 - accuracy: 0.5563 - val_loss: 0.2297 - val_accuracy: 0.8889\n",
            "Epoch 25/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2894 - accuracy: 0.5563 - val_loss: 0.1445 - val_accuracy: 0.9259\n",
            "Epoch 26/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2846 - accuracy: 0.5552 - val_loss: 0.1299 - val_accuracy: 0.9630\n",
            "Epoch 27/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2859 - accuracy: 0.5563 - val_loss: 0.1730 - val_accuracy: 0.9259\n",
            "Epoch 28/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2758 - accuracy: 0.5563 - val_loss: 0.0991 - val_accuracy: 1.0000\n",
            "Epoch 29/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2775 - accuracy: 0.5563 - val_loss: 0.1045 - val_accuracy: 1.0000\n",
            "Epoch 30/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2762 - accuracy: 0.5563 - val_loss: 0.1271 - val_accuracy: 0.9630\n",
            "Epoch 31/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2757 - accuracy: 0.5563 - val_loss: 0.3272 - val_accuracy: 0.8889\n",
            "Epoch 32/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2777 - accuracy: 0.5563 - val_loss: 0.1652 - val_accuracy: 0.9630\n",
            "Epoch 33/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2913 - accuracy: 0.5563 - val_loss: 0.1515 - val_accuracy: 1.0000\n",
            "Epoch 34/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2933 - accuracy: 0.5552 - val_loss: 0.1499 - val_accuracy: 0.9630\n",
            "Epoch 35/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2793 - accuracy: 0.5563 - val_loss: 0.1287 - val_accuracy: 1.0000\n",
            "Epoch 36/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2776 - accuracy: 0.5563 - val_loss: 0.1386 - val_accuracy: 1.0000\n",
            "Epoch 37/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2818 - accuracy: 0.5563 - val_loss: 0.1492 - val_accuracy: 1.0000\n",
            "Epoch 38/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2787 - accuracy: 0.5552 - val_loss: 0.1303 - val_accuracy: 1.0000\n",
            "Epoch 39/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2738 - accuracy: 0.5563 - val_loss: 0.1597 - val_accuracy: 0.9630\n",
            "Epoch 40/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2719 - accuracy: 0.5563 - val_loss: 0.1051 - val_accuracy: 1.0000\n",
            "Epoch 41/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2789 - accuracy: 0.5563 - val_loss: 0.0754 - val_accuracy: 1.0000\n",
            "Epoch 42/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2803 - accuracy: 0.5563 - val_loss: 0.1324 - val_accuracy: 1.0000\n",
            "Epoch 43/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2841 - accuracy: 0.5563 - val_loss: 0.1181 - val_accuracy: 0.9630\n",
            "Epoch 44/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2705 - accuracy: 0.5563 - val_loss: 0.1456 - val_accuracy: 0.9259\n",
            "Epoch 45/100\n",
            "28/28 [==============================] - 11s 375ms/step - loss: 0.2682 - accuracy: 0.5563 - val_loss: 0.1112 - val_accuracy: 1.0000\n",
            "Epoch 46/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2670 - accuracy: 0.5563 - val_loss: 0.1176 - val_accuracy: 0.9630\n",
            "Epoch 47/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2672 - accuracy: 0.5563 - val_loss: 0.0799 - val_accuracy: 1.0000\n",
            "Epoch 48/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2673 - accuracy: 0.5563 - val_loss: 0.1062 - val_accuracy: 1.0000\n",
            "Epoch 49/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2640 - accuracy: 0.5563 - val_loss: 0.1073 - val_accuracy: 1.0000\n",
            "Epoch 50/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2669 - accuracy: 0.5563 - val_loss: 0.1287 - val_accuracy: 1.0000\n",
            "Epoch 51/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2686 - accuracy: 0.5563 - val_loss: 0.1171 - val_accuracy: 1.0000\n",
            "Epoch 52/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2673 - accuracy: 0.5563 - val_loss: 0.1760 - val_accuracy: 1.0000\n",
            "Epoch 53/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2705 - accuracy: 0.5563 - val_loss: 0.1452 - val_accuracy: 1.0000\n",
            "Epoch 54/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2935 - accuracy: 0.5540 - val_loss: 0.3018 - val_accuracy: 0.8889\n",
            "Epoch 55/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2823 - accuracy: 0.5552 - val_loss: 0.2346 - val_accuracy: 0.8889\n",
            "Epoch 56/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2721 - accuracy: 0.5563 - val_loss: 0.2624 - val_accuracy: 0.8889\n",
            "2/2 [==============================] - 1s 99ms/step - loss: 0.0618 - accuracy: 1.0000\n",
            "Epoch 1/100\n",
            "28/28 [==============================] - 16s 411ms/step - loss: 0.5394 - accuracy: 0.4932 - val_loss: 0.1839 - val_accuracy: 0.9259\n",
            "Epoch 2/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.4278 - accuracy: 0.5263 - val_loss: 0.1363 - val_accuracy: 1.0000\n",
            "Epoch 3/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.3832 - accuracy: 0.5502 - val_loss: 0.1699 - val_accuracy: 0.9630\n",
            "Epoch 4/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.3595 - accuracy: 0.5559 - val_loss: 0.2058 - val_accuracy: 0.9630\n",
            "Epoch 5/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.3543 - accuracy: 0.5651 - val_loss: 0.2352 - val_accuracy: 0.9259\n",
            "Epoch 6/100\n",
            "28/28 [==============================] - 11s 375ms/step - loss: 0.3495 - accuracy: 0.5537 - val_loss: 0.1662 - val_accuracy: 1.0000\n",
            "Epoch 7/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.3264 - accuracy: 0.5616 - val_loss: 0.1997 - val_accuracy: 0.9259\n",
            "Epoch 8/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.3341 - accuracy: 0.5616 - val_loss: 0.1187 - val_accuracy: 0.9630\n",
            "Epoch 9/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.3241 - accuracy: 0.5651 - val_loss: 0.1621 - val_accuracy: 0.9630\n",
            "Epoch 10/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.3119 - accuracy: 0.5674 - val_loss: 0.3073 - val_accuracy: 0.8148\n",
            "Epoch 11/100\n",
            "28/28 [==============================] - 11s 375ms/step - loss: 0.3062 - accuracy: 0.5674 - val_loss: 0.1511 - val_accuracy: 1.0000\n",
            "Epoch 12/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.3036 - accuracy: 0.5674 - val_loss: 0.2769 - val_accuracy: 0.8519\n",
            "Epoch 13/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2950 - accuracy: 0.5696 - val_loss: 0.1094 - val_accuracy: 1.0000\n",
            "Epoch 14/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2965 - accuracy: 0.5708 - val_loss: 0.0842 - val_accuracy: 1.0000\n",
            "Epoch 15/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2928 - accuracy: 0.5685 - val_loss: 0.0871 - val_accuracy: 1.0000\n",
            "Epoch 16/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2939 - accuracy: 0.5696 - val_loss: 0.2236 - val_accuracy: 0.9259\n",
            "Epoch 17/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2853 - accuracy: 0.5685 - val_loss: 0.1149 - val_accuracy: 0.9630\n",
            "Epoch 18/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2888 - accuracy: 0.5696 - val_loss: 0.1901 - val_accuracy: 0.9259\n",
            "Epoch 19/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2870 - accuracy: 0.5696 - val_loss: 0.1710 - val_accuracy: 0.9259\n",
            "Epoch 20/100\n",
            "28/28 [==============================] - 11s 378ms/step - loss: 0.2809 - accuracy: 0.5708 - val_loss: 0.0703 - val_accuracy: 1.0000\n",
            "Epoch 21/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2871 - accuracy: 0.5708 - val_loss: 0.1064 - val_accuracy: 0.9259\n",
            "Epoch 22/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2790 - accuracy: 0.5708 - val_loss: 0.1552 - val_accuracy: 0.8889\n",
            "Epoch 23/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2776 - accuracy: 0.5708 - val_loss: 0.2104 - val_accuracy: 0.9259\n",
            "Epoch 24/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2834 - accuracy: 0.5696 - val_loss: 0.0899 - val_accuracy: 0.9630\n",
            "Epoch 25/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2887 - accuracy: 0.5708 - val_loss: 0.0892 - val_accuracy: 0.9630\n",
            "Epoch 26/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2752 - accuracy: 0.5708 - val_loss: 0.1547 - val_accuracy: 0.9259\n",
            "Epoch 27/100\n",
            "28/28 [==============================] - 11s 375ms/step - loss: 0.2768 - accuracy: 0.5708 - val_loss: 0.1086 - val_accuracy: 0.9630\n",
            "Epoch 28/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2732 - accuracy: 0.5708 - val_loss: 0.1261 - val_accuracy: 0.9630\n",
            "Epoch 29/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2714 - accuracy: 0.5708 - val_loss: 0.0853 - val_accuracy: 1.0000\n",
            "Epoch 30/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2721 - accuracy: 0.5685 - val_loss: 0.0520 - val_accuracy: 1.0000\n",
            "Epoch 31/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2775 - accuracy: 0.5708 - val_loss: 0.1309 - val_accuracy: 0.9630\n",
            "Epoch 32/100\n",
            "28/28 [==============================] - 11s 375ms/step - loss: 0.2740 - accuracy: 0.5708 - val_loss: 0.0855 - val_accuracy: 0.9630\n",
            "Epoch 33/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2700 - accuracy: 0.5708 - val_loss: 0.0537 - val_accuracy: 1.0000\n",
            "Epoch 34/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2723 - accuracy: 0.5708 - val_loss: 0.0633 - val_accuracy: 1.0000\n",
            "Epoch 35/100\n",
            "28/28 [==============================] - 11s 375ms/step - loss: 0.2713 - accuracy: 0.5708 - val_loss: 0.0909 - val_accuracy: 1.0000\n",
            "Epoch 36/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2704 - accuracy: 0.5708 - val_loss: 0.0720 - val_accuracy: 1.0000\n",
            "Epoch 37/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2702 - accuracy: 0.5708 - val_loss: 0.0720 - val_accuracy: 1.0000\n",
            "Epoch 38/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2753 - accuracy: 0.5708 - val_loss: 0.1134 - val_accuracy: 0.9630\n",
            "Epoch 39/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2713 - accuracy: 0.5696 - val_loss: 0.1612 - val_accuracy: 0.9630\n",
            "Epoch 40/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2657 - accuracy: 0.5708 - val_loss: 0.1873 - val_accuracy: 0.9630\n",
            "Epoch 41/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2654 - accuracy: 0.5708 - val_loss: 0.0666 - val_accuracy: 1.0000\n",
            "Epoch 42/100\n",
            "28/28 [==============================] - 10s 376ms/step - loss: 0.2664 - accuracy: 0.5708 - val_loss: 0.0797 - val_accuracy: 1.0000\n",
            "Epoch 43/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2771 - accuracy: 0.5708 - val_loss: 0.0746 - val_accuracy: 1.0000\n",
            "Epoch 44/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2688 - accuracy: 0.5708 - val_loss: 0.1185 - val_accuracy: 0.9630\n",
            "Epoch 45/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2575 - accuracy: 0.5708 - val_loss: 0.1569 - val_accuracy: 0.9259\n",
            "2/2 [==============================] - 1s 113ms/step - loss: 0.0473 - accuracy: 1.0000\n",
            "Epoch 1/100\n",
            "27/27 [==============================] - 15s 417ms/step - loss: 0.4561 - accuracy: 0.5451 - val_loss: 0.2607 - val_accuracy: 0.8889\n",
            "Epoch 2/100\n",
            "27/27 [==============================] - 10s 380ms/step - loss: 0.3730 - accuracy: 0.5775 - val_loss: 0.1710 - val_accuracy: 0.9630\n",
            "Epoch 3/100\n",
            "27/27 [==============================] - 10s 378ms/step - loss: 0.3477 - accuracy: 0.5752 - val_loss: 0.1485 - val_accuracy: 0.9630\n",
            "Epoch 4/100\n",
            "27/27 [==============================] - 10s 380ms/step - loss: 0.3374 - accuracy: 0.5903 - val_loss: 0.1435 - val_accuracy: 0.9630\n",
            "Epoch 5/100\n",
            "27/27 [==============================] - 10s 380ms/step - loss: 0.3049 - accuracy: 0.5880 - val_loss: 0.1256 - val_accuracy: 1.0000\n",
            "Epoch 6/100\n",
            "27/27 [==============================] - 10s 376ms/step - loss: 0.3088 - accuracy: 0.5868 - val_loss: 0.1612 - val_accuracy: 1.0000\n",
            "Epoch 7/100\n",
            "27/27 [==============================] - 10s 378ms/step - loss: 0.3005 - accuracy: 0.5903 - val_loss: 0.1038 - val_accuracy: 1.0000\n",
            "Epoch 8/100\n",
            "27/27 [==============================] - 10s 375ms/step - loss: 0.2914 - accuracy: 0.5938 - val_loss: 0.1299 - val_accuracy: 1.0000\n",
            "Epoch 9/100\n",
            "27/27 [==============================] - 10s 382ms/step - loss: 0.2894 - accuracy: 0.5938 - val_loss: 0.0973 - val_accuracy: 0.9630\n",
            "Epoch 10/100\n",
            "27/27 [==============================] - 10s 377ms/step - loss: 0.2950 - accuracy: 0.5903 - val_loss: 0.0974 - val_accuracy: 1.0000\n",
            "Epoch 11/100\n",
            "27/27 [==============================] - 10s 380ms/step - loss: 0.2889 - accuracy: 0.5926 - val_loss: 0.0827 - val_accuracy: 1.0000\n",
            "Epoch 12/100\n",
            "27/27 [==============================] - 10s 380ms/step - loss: 0.2743 - accuracy: 0.5949 - val_loss: 0.0766 - val_accuracy: 1.0000\n",
            "Epoch 13/100\n",
            "27/27 [==============================] - 10s 376ms/step - loss: 0.2816 - accuracy: 0.5926 - val_loss: 0.1524 - val_accuracy: 0.9630\n",
            "Epoch 14/100\n",
            "27/27 [==============================] - 10s 377ms/step - loss: 0.2710 - accuracy: 0.5949 - val_loss: 0.1168 - val_accuracy: 0.9630\n",
            "Epoch 15/100\n",
            "27/27 [==============================] - 10s 380ms/step - loss: 0.2695 - accuracy: 0.5938 - val_loss: 0.0624 - val_accuracy: 1.0000\n",
            "Epoch 16/100\n",
            "27/27 [==============================] - 10s 378ms/step - loss: 0.2651 - accuracy: 0.5949 - val_loss: 0.0750 - val_accuracy: 1.0000\n",
            "Epoch 17/100\n",
            "27/27 [==============================] - 10s 379ms/step - loss: 0.2668 - accuracy: 0.5949 - val_loss: 0.0688 - val_accuracy: 1.0000\n",
            "Epoch 18/100\n",
            "27/27 [==============================] - 10s 377ms/step - loss: 0.2772 - accuracy: 0.5938 - val_loss: 0.1083 - val_accuracy: 1.0000\n",
            "Epoch 19/100\n",
            "27/27 [==============================] - 10s 377ms/step - loss: 0.2602 - accuracy: 0.5949 - val_loss: 0.0908 - val_accuracy: 1.0000\n",
            "Epoch 20/100\n",
            "27/27 [==============================] - 10s 377ms/step - loss: 0.2593 - accuracy: 0.5949 - val_loss: 0.0843 - val_accuracy: 1.0000\n",
            "Epoch 21/100\n",
            "27/27 [==============================] - 10s 376ms/step - loss: 0.2641 - accuracy: 0.5949 - val_loss: 0.1409 - val_accuracy: 0.9630\n",
            "Epoch 22/100\n",
            "27/27 [==============================] - 10s 377ms/step - loss: 0.2582 - accuracy: 0.5949 - val_loss: 0.1146 - val_accuracy: 0.9630\n",
            "Epoch 23/100\n",
            "27/27 [==============================] - 10s 377ms/step - loss: 0.2578 - accuracy: 0.5949 - val_loss: 0.1374 - val_accuracy: 0.9259\n",
            "Epoch 24/100\n",
            "27/27 [==============================] - 10s 376ms/step - loss: 0.2532 - accuracy: 0.5949 - val_loss: 0.1867 - val_accuracy: 0.8889\n",
            "Epoch 25/100\n",
            "27/27 [==============================] - 10s 377ms/step - loss: 0.2541 - accuracy: 0.5949 - val_loss: 0.0831 - val_accuracy: 1.0000\n",
            "Epoch 26/100\n",
            "27/27 [==============================] - 10s 378ms/step - loss: 0.2608 - accuracy: 0.5949 - val_loss: 0.1029 - val_accuracy: 1.0000\n",
            "Epoch 27/100\n",
            "27/27 [==============================] - 10s 379ms/step - loss: 0.2536 - accuracy: 0.5949 - val_loss: 0.0961 - val_accuracy: 1.0000\n",
            "Epoch 28/100\n",
            "27/27 [==============================] - 10s 379ms/step - loss: 0.2578 - accuracy: 0.5949 - val_loss: 0.0813 - val_accuracy: 1.0000\n",
            "Epoch 29/100\n",
            "27/27 [==============================] - 10s 379ms/step - loss: 0.2533 - accuracy: 0.5949 - val_loss: 0.1109 - val_accuracy: 1.0000\n",
            "Epoch 30/100\n",
            "27/27 [==============================] - 10s 379ms/step - loss: 0.2465 - accuracy: 0.5949 - val_loss: 0.1341 - val_accuracy: 0.9630\n",
            "2/2 [==============================] - 1s 99ms/step - loss: 0.2344 - accuracy: 0.9118\n",
            "[0.9764705777168274, 0.9941176414489746, 0.970588219165802, 0.970588231086731]\n",
            "[0.973400290736091, 0.9934299516908212, 0.9670915511664816, 0.9651646403242149]\n",
            "Epoch 1/100\n",
            "28/28 [==============================] - 16s 418ms/step - loss: 0.5410 - accuracy: 0.4820 - val_loss: 0.3488 - val_accuracy: 0.8519\n",
            "Epoch 2/100\n",
            "28/28 [==============================] - 11s 378ms/step - loss: 0.3907 - accuracy: 0.5428 - val_loss: 0.3901 - val_accuracy: 0.9259\n",
            "Epoch 3/100\n",
            "28/28 [==============================] - 11s 380ms/step - loss: 0.3836 - accuracy: 0.5518 - val_loss: 0.3288 - val_accuracy: 0.8519\n",
            "Epoch 4/100\n",
            "28/28 [==============================] - 11s 378ms/step - loss: 0.3377 - accuracy: 0.5631 - val_loss: 0.2065 - val_accuracy: 0.9630\n",
            "Epoch 5/100\n",
            "28/28 [==============================] - 11s 378ms/step - loss: 0.3250 - accuracy: 0.5732 - val_loss: 0.1543 - val_accuracy: 0.9630\n",
            "Epoch 6/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.3408 - accuracy: 0.5687 - val_loss: 0.1746 - val_accuracy: 1.0000\n",
            "Epoch 7/100\n",
            "28/28 [==============================] - 11s 378ms/step - loss: 0.3192 - accuracy: 0.5755 - val_loss: 0.1433 - val_accuracy: 0.9259\n",
            "Epoch 8/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.3074 - accuracy: 0.5709 - val_loss: 0.1747 - val_accuracy: 1.0000\n",
            "Epoch 9/100\n",
            "28/28 [==============================] - 11s 379ms/step - loss: 0.3072 - accuracy: 0.5721 - val_loss: 0.1208 - val_accuracy: 0.9259\n",
            "Epoch 10/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2966 - accuracy: 0.5743 - val_loss: 0.1070 - val_accuracy: 0.9630\n",
            "Epoch 11/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2938 - accuracy: 0.5743 - val_loss: 0.1442 - val_accuracy: 1.0000\n",
            "Epoch 12/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2840 - accuracy: 0.5743 - val_loss: 0.1367 - val_accuracy: 0.9259\n",
            "Epoch 13/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2883 - accuracy: 0.5743 - val_loss: 0.1452 - val_accuracy: 0.9630\n",
            "Epoch 14/100\n",
            "28/28 [==============================] - 11s 380ms/step - loss: 0.2985 - accuracy: 0.5743 - val_loss: 0.1001 - val_accuracy: 0.9630\n",
            "Epoch 15/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2810 - accuracy: 0.5755 - val_loss: 0.1101 - val_accuracy: 0.9630\n",
            "Epoch 16/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2842 - accuracy: 0.5766 - val_loss: 0.1086 - val_accuracy: 0.9630\n",
            "Epoch 17/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2807 - accuracy: 0.5755 - val_loss: 0.1094 - val_accuracy: 0.9630\n",
            "Epoch 18/100\n",
            "28/28 [==============================] - 11s 379ms/step - loss: 0.2802 - accuracy: 0.5766 - val_loss: 0.0993 - val_accuracy: 1.0000\n",
            "Epoch 19/100\n",
            "28/28 [==============================] - 11s 379ms/step - loss: 0.2710 - accuracy: 0.5755 - val_loss: 0.0895 - val_accuracy: 0.9630\n",
            "Epoch 20/100\n",
            "28/28 [==============================] - 11s 378ms/step - loss: 0.2816 - accuracy: 0.5766 - val_loss: 0.1048 - val_accuracy: 0.9630\n",
            "Epoch 21/100\n",
            "28/28 [==============================] - 11s 379ms/step - loss: 0.2781 - accuracy: 0.5766 - val_loss: 0.0880 - val_accuracy: 1.0000\n",
            "Epoch 22/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2685 - accuracy: 0.5766 - val_loss: 0.0923 - val_accuracy: 0.9630\n",
            "Epoch 23/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2745 - accuracy: 0.5755 - val_loss: 0.0911 - val_accuracy: 0.9630\n",
            "Epoch 24/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2725 - accuracy: 0.5766 - val_loss: 0.1438 - val_accuracy: 0.9630\n",
            "Epoch 25/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2642 - accuracy: 0.5766 - val_loss: 0.1218 - val_accuracy: 0.9630\n",
            "Epoch 26/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2636 - accuracy: 0.5766 - val_loss: 0.1464 - val_accuracy: 0.9259\n",
            "Epoch 27/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2652 - accuracy: 0.5766 - val_loss: 0.1312 - val_accuracy: 1.0000\n",
            "Epoch 28/100\n",
            "28/28 [==============================] - 11s 379ms/step - loss: 0.2687 - accuracy: 0.5766 - val_loss: 0.0790 - val_accuracy: 1.0000\n",
            "Epoch 29/100\n",
            "28/28 [==============================] - 11s 378ms/step - loss: 0.2675 - accuracy: 0.5755 - val_loss: 0.1147 - val_accuracy: 0.9630\n",
            "Epoch 30/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2670 - accuracy: 0.5766 - val_loss: 0.1102 - val_accuracy: 0.9630\n",
            "Epoch 31/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2576 - accuracy: 0.5766 - val_loss: 0.1351 - val_accuracy: 0.9630\n",
            "Epoch 32/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2550 - accuracy: 0.5766 - val_loss: 0.1016 - val_accuracy: 1.0000\n",
            "Epoch 33/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2559 - accuracy: 0.5766 - val_loss: 0.1060 - val_accuracy: 0.9630\n",
            "Epoch 34/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2638 - accuracy: 0.5766 - val_loss: 0.0908 - val_accuracy: 0.9630\n",
            "Epoch 35/100\n",
            "28/28 [==============================] - 11s 378ms/step - loss: 0.2613 - accuracy: 0.5766 - val_loss: 0.0986 - val_accuracy: 1.0000\n",
            "Epoch 36/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2617 - accuracy: 0.5766 - val_loss: 0.2099 - val_accuracy: 0.8889\n",
            "Epoch 37/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2622 - accuracy: 0.5766 - val_loss: 0.1022 - val_accuracy: 0.9259\n",
            "Epoch 38/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2564 - accuracy: 0.5766 - val_loss: 0.1017 - val_accuracy: 0.9630\n",
            "Epoch 39/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2502 - accuracy: 0.5766 - val_loss: 0.1101 - val_accuracy: 0.9630\n",
            "Epoch 40/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2557 - accuracy: 0.5766 - val_loss: 0.1709 - val_accuracy: 1.0000\n",
            "Epoch 41/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2664 - accuracy: 0.5766 - val_loss: 0.1322 - val_accuracy: 0.9259\n",
            "Epoch 42/100\n",
            "28/28 [==============================] - 11s 379ms/step - loss: 0.2562 - accuracy: 0.5766 - val_loss: 0.0675 - val_accuracy: 1.0000\n",
            "Epoch 43/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2589 - accuracy: 0.5766 - val_loss: 0.1034 - val_accuracy: 1.0000\n",
            "Epoch 44/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2569 - accuracy: 0.5766 - val_loss: 0.0965 - val_accuracy: 1.0000\n",
            "Epoch 45/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2513 - accuracy: 0.5755 - val_loss: 0.0801 - val_accuracy: 0.9630\n",
            "Epoch 46/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2556 - accuracy: 0.5766 - val_loss: 0.1041 - val_accuracy: 0.9630\n",
            "Epoch 47/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2495 - accuracy: 0.5766 - val_loss: 0.1004 - val_accuracy: 0.9630\n",
            "Epoch 48/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2498 - accuracy: 0.5766 - val_loss: 0.1201 - val_accuracy: 0.9630\n",
            "Epoch 49/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2583 - accuracy: 0.5766 - val_loss: 0.1060 - val_accuracy: 0.9630\n",
            "Epoch 50/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2502 - accuracy: 0.5766 - val_loss: 0.1391 - val_accuracy: 0.9630\n",
            "Epoch 51/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2565 - accuracy: 0.5766 - val_loss: 0.1050 - val_accuracy: 0.9259\n",
            "Epoch 52/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2586 - accuracy: 0.5766 - val_loss: 0.1274 - val_accuracy: 1.0000\n",
            "Epoch 53/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2509 - accuracy: 0.5766 - val_loss: 0.1133 - val_accuracy: 1.0000\n",
            "Epoch 54/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2464 - accuracy: 0.5766 - val_loss: 0.0790 - val_accuracy: 1.0000\n",
            "Epoch 55/100\n",
            "28/28 [==============================] - 11s 378ms/step - loss: 0.2560 - accuracy: 0.5766 - val_loss: 0.2399 - val_accuracy: 0.9259\n",
            "Epoch 56/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2462 - accuracy: 0.5766 - val_loss: 0.1142 - val_accuracy: 0.9630\n",
            "Epoch 57/100\n",
            "28/28 [==============================] - 11s 379ms/step - loss: 0.2521 - accuracy: 0.5766 - val_loss: 0.1337 - val_accuracy: 0.9630\n",
            "2/2 [==============================] - 1s 99ms/step - loss: 0.0638 - accuracy: 1.0000\n",
            "Epoch 1/100\n",
            "27/27 [==============================] - 15s 416ms/step - loss: 0.5743 - accuracy: 0.4745 - val_loss: 0.1798 - val_accuracy: 0.9259\n",
            "Epoch 2/100\n",
            "27/27 [==============================] - 10s 378ms/step - loss: 0.4296 - accuracy: 0.5301 - val_loss: 0.2103 - val_accuracy: 0.9259\n",
            "Epoch 3/100\n",
            "27/27 [==============================] - 10s 377ms/step - loss: 0.3710 - accuracy: 0.5475 - val_loss: 0.2136 - val_accuracy: 0.9259\n",
            "Epoch 4/100\n",
            "27/27 [==============================] - 10s 375ms/step - loss: 0.3596 - accuracy: 0.5463 - val_loss: 0.2032 - val_accuracy: 1.0000\n",
            "Epoch 5/100\n",
            "27/27 [==============================] - 10s 378ms/step - loss: 0.3455 - accuracy: 0.5532 - val_loss: 0.1588 - val_accuracy: 1.0000\n",
            "Epoch 6/100\n",
            "27/27 [==============================] - 10s 376ms/step - loss: 0.3426 - accuracy: 0.5544 - val_loss: 0.1866 - val_accuracy: 1.0000\n",
            "Epoch 7/100\n",
            "27/27 [==============================] - 10s 375ms/step - loss: 0.3364 - accuracy: 0.5579 - val_loss: 0.1729 - val_accuracy: 1.0000\n",
            "Epoch 8/100\n",
            "27/27 [==============================] - 10s 372ms/step - loss: 0.3303 - accuracy: 0.5590 - val_loss: 0.1894 - val_accuracy: 1.0000\n",
            "Epoch 9/100\n",
            "27/27 [==============================] - 10s 374ms/step - loss: 0.3257 - accuracy: 0.5602 - val_loss: 0.2097 - val_accuracy: 0.9259\n",
            "Epoch 10/100\n",
            "27/27 [==============================] - 10s 372ms/step - loss: 0.3113 - accuracy: 0.5613 - val_loss: 0.1956 - val_accuracy: 0.9630\n",
            "Epoch 11/100\n",
            "27/27 [==============================] - 10s 372ms/step - loss: 0.3109 - accuracy: 0.5648 - val_loss: 0.2241 - val_accuracy: 0.8889\n",
            "Epoch 12/100\n",
            "27/27 [==============================] - 10s 379ms/step - loss: 0.3130 - accuracy: 0.5637 - val_loss: 0.1477 - val_accuracy: 1.0000\n",
            "Epoch 13/100\n",
            "27/27 [==============================] - 10s 382ms/step - loss: 0.3072 - accuracy: 0.5637 - val_loss: 0.1082 - val_accuracy: 1.0000\n",
            "Epoch 14/100\n",
            "27/27 [==============================] - 10s 376ms/step - loss: 0.2931 - accuracy: 0.5648 - val_loss: 0.1233 - val_accuracy: 1.0000\n",
            "Epoch 15/100\n",
            "27/27 [==============================] - 10s 378ms/step - loss: 0.2906 - accuracy: 0.5648 - val_loss: 0.1645 - val_accuracy: 0.9630\n",
            "Epoch 16/100\n",
            "27/27 [==============================] - 10s 379ms/step - loss: 0.2972 - accuracy: 0.5625 - val_loss: 0.3544 - val_accuracy: 0.8519\n",
            "Epoch 17/100\n",
            "27/27 [==============================] - 10s 378ms/step - loss: 0.2831 - accuracy: 0.5648 - val_loss: 0.1726 - val_accuracy: 0.9630\n",
            "Epoch 18/100\n",
            "27/27 [==============================] - 10s 379ms/step - loss: 0.2933 - accuracy: 0.5637 - val_loss: 0.2312 - val_accuracy: 0.8519\n",
            "Epoch 19/100\n",
            "27/27 [==============================] - 10s 380ms/step - loss: 0.2954 - accuracy: 0.5637 - val_loss: 0.1028 - val_accuracy: 1.0000\n",
            "Epoch 20/100\n",
            "27/27 [==============================] - 10s 379ms/step - loss: 0.2876 - accuracy: 0.5648 - val_loss: 0.1180 - val_accuracy: 1.0000\n",
            "Epoch 21/100\n",
            "27/27 [==============================] - 10s 377ms/step - loss: 0.2928 - accuracy: 0.5648 - val_loss: 0.1090 - val_accuracy: 1.0000\n",
            "Epoch 22/100\n",
            "27/27 [==============================] - 10s 380ms/step - loss: 0.2865 - accuracy: 0.5648 - val_loss: 0.1286 - val_accuracy: 1.0000\n",
            "Epoch 23/100\n",
            "27/27 [==============================] - 10s 378ms/step - loss: 0.2883 - accuracy: 0.5637 - val_loss: 0.1720 - val_accuracy: 0.9630\n",
            "Epoch 24/100\n",
            "27/27 [==============================] - 10s 378ms/step - loss: 0.2791 - accuracy: 0.5648 - val_loss: 0.1189 - val_accuracy: 1.0000\n",
            "Epoch 25/100\n",
            "27/27 [==============================] - 10s 378ms/step - loss: 0.2737 - accuracy: 0.5648 - val_loss: 0.1136 - val_accuracy: 1.0000\n",
            "Epoch 26/100\n",
            "27/27 [==============================] - 10s 380ms/step - loss: 0.2879 - accuracy: 0.5637 - val_loss: 0.1214 - val_accuracy: 1.0000\n",
            "Epoch 27/100\n",
            "27/27 [==============================] - 10s 379ms/step - loss: 0.2741 - accuracy: 0.5648 - val_loss: 0.1424 - val_accuracy: 1.0000\n",
            "Epoch 28/100\n",
            "27/27 [==============================] - 10s 377ms/step - loss: 0.2726 - accuracy: 0.5648 - val_loss: 0.1083 - val_accuracy: 1.0000\n",
            "Epoch 29/100\n",
            "27/27 [==============================] - 10s 379ms/step - loss: 0.2773 - accuracy: 0.5648 - val_loss: 0.1548 - val_accuracy: 0.9259\n",
            "Epoch 30/100\n",
            "27/27 [==============================] - 10s 380ms/step - loss: 0.2660 - accuracy: 0.5648 - val_loss: 0.1559 - val_accuracy: 0.9259\n",
            "Epoch 31/100\n",
            "27/27 [==============================] - 10s 379ms/step - loss: 0.2753 - accuracy: 0.5648 - val_loss: 0.1131 - val_accuracy: 1.0000\n",
            "Epoch 32/100\n",
            "27/27 [==============================] - 10s 378ms/step - loss: 0.2680 - accuracy: 0.5648 - val_loss: 0.1055 - val_accuracy: 1.0000\n",
            "Epoch 33/100\n",
            "27/27 [==============================] - 10s 375ms/step - loss: 0.2724 - accuracy: 0.5648 - val_loss: 0.1862 - val_accuracy: 0.9259\n",
            "Epoch 34/100\n",
            "27/27 [==============================] - 10s 378ms/step - loss: 0.2706 - accuracy: 0.5648 - val_loss: 0.0759 - val_accuracy: 1.0000\n",
            "Epoch 35/100\n",
            "27/27 [==============================] - 10s 379ms/step - loss: 0.2740 - accuracy: 0.5648 - val_loss: 0.1035 - val_accuracy: 1.0000\n",
            "Epoch 36/100\n",
            "27/27 [==============================] - 10s 377ms/step - loss: 0.2806 - accuracy: 0.5637 - val_loss: 0.1436 - val_accuracy: 1.0000\n",
            "Epoch 37/100\n",
            "27/27 [==============================] - 10s 378ms/step - loss: 0.2688 - accuracy: 0.5648 - val_loss: 0.1533 - val_accuracy: 1.0000\n",
            "Epoch 38/100\n",
            "27/27 [==============================] - 10s 378ms/step - loss: 0.2652 - accuracy: 0.5648 - val_loss: 0.1815 - val_accuracy: 0.9259\n",
            "Epoch 39/100\n",
            "27/27 [==============================] - 10s 379ms/step - loss: 0.2571 - accuracy: 0.5648 - val_loss: 0.0978 - val_accuracy: 1.0000\n",
            "Epoch 40/100\n",
            "27/27 [==============================] - 10s 377ms/step - loss: 0.2682 - accuracy: 0.5648 - val_loss: 0.1585 - val_accuracy: 1.0000\n",
            "Epoch 41/100\n",
            "27/27 [==============================] - 10s 377ms/step - loss: 0.2656 - accuracy: 0.5648 - val_loss: 0.1266 - val_accuracy: 1.0000\n",
            "Epoch 42/100\n",
            "27/27 [==============================] - 10s 379ms/step - loss: 0.2753 - accuracy: 0.5648 - val_loss: 0.1011 - val_accuracy: 1.0000\n",
            "Epoch 43/100\n",
            "27/27 [==============================] - 10s 377ms/step - loss: 0.2660 - accuracy: 0.5648 - val_loss: 0.1413 - val_accuracy: 0.9630\n",
            "Epoch 44/100\n",
            "27/27 [==============================] - 10s 377ms/step - loss: 0.2639 - accuracy: 0.5648 - val_loss: 0.1614 - val_accuracy: 0.9259\n",
            "Epoch 45/100\n",
            "27/27 [==============================] - 10s 378ms/step - loss: 0.2599 - accuracy: 0.5648 - val_loss: 0.1338 - val_accuracy: 0.9630\n",
            "Epoch 46/100\n",
            "27/27 [==============================] - 10s 379ms/step - loss: 0.2610 - accuracy: 0.5637 - val_loss: 0.1704 - val_accuracy: 0.9630\n",
            "Epoch 47/100\n",
            "27/27 [==============================] - 10s 380ms/step - loss: 0.2630 - accuracy: 0.5648 - val_loss: 0.0879 - val_accuracy: 1.0000\n",
            "Epoch 48/100\n",
            "27/27 [==============================] - 10s 380ms/step - loss: 0.2651 - accuracy: 0.5648 - val_loss: 0.1959 - val_accuracy: 0.9630\n",
            "Epoch 49/100\n",
            "27/27 [==============================] - 10s 380ms/step - loss: 0.2588 - accuracy: 0.5648 - val_loss: 0.1276 - val_accuracy: 1.0000\n",
            "2/2 [==============================] - 1s 100ms/step - loss: 0.1615 - accuracy: 0.9118\n",
            "Epoch 1/100\n",
            "28/28 [==============================] - 16s 416ms/step - loss: 0.5332 - accuracy: 0.4764 - val_loss: 0.1547 - val_accuracy: 0.9630\n",
            "Epoch 2/100\n",
            "28/28 [==============================] - 11s 382ms/step - loss: 0.4128 - accuracy: 0.5507 - val_loss: 0.1546 - val_accuracy: 0.9259\n",
            "Epoch 3/100\n",
            "28/28 [==============================] - 11s 379ms/step - loss: 0.3801 - accuracy: 0.5619 - val_loss: 0.1260 - val_accuracy: 1.0000\n",
            "Epoch 4/100\n",
            "28/28 [==============================] - 11s 378ms/step - loss: 0.3454 - accuracy: 0.5687 - val_loss: 0.1294 - val_accuracy: 0.9630\n",
            "Epoch 5/100\n",
            "28/28 [==============================] - 11s 379ms/step - loss: 0.3414 - accuracy: 0.5698 - val_loss: 0.1049 - val_accuracy: 1.0000\n",
            "Epoch 6/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.3341 - accuracy: 0.5709 - val_loss: 0.1411 - val_accuracy: 1.0000\n",
            "Epoch 7/100\n",
            "28/28 [==============================] - 11s 380ms/step - loss: 0.3142 - accuracy: 0.5766 - val_loss: 0.1101 - val_accuracy: 1.0000\n",
            "Epoch 8/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.3160 - accuracy: 0.5743 - val_loss: 0.1458 - val_accuracy: 1.0000\n",
            "Epoch 9/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.3177 - accuracy: 0.5721 - val_loss: 0.1568 - val_accuracy: 1.0000\n",
            "Epoch 10/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.3095 - accuracy: 0.5755 - val_loss: 0.2258 - val_accuracy: 0.9259\n",
            "Epoch 11/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2935 - accuracy: 0.5777 - val_loss: 0.1203 - val_accuracy: 1.0000\n",
            "Epoch 12/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2900 - accuracy: 0.5788 - val_loss: 0.0810 - val_accuracy: 1.0000\n",
            "Epoch 13/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2894 - accuracy: 0.5766 - val_loss: 0.1060 - val_accuracy: 1.0000\n",
            "Epoch 14/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.3082 - accuracy: 0.5732 - val_loss: 0.1718 - val_accuracy: 1.0000\n",
            "Epoch 15/100\n",
            "28/28 [==============================] - 11s 380ms/step - loss: 0.2959 - accuracy: 0.5777 - val_loss: 0.0595 - val_accuracy: 1.0000\n",
            "Epoch 16/100\n",
            "28/28 [==============================] - 11s 378ms/step - loss: 0.2780 - accuracy: 0.5788 - val_loss: 0.0475 - val_accuracy: 1.0000\n",
            "Epoch 17/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2817 - accuracy: 0.5788 - val_loss: 0.0543 - val_accuracy: 1.0000\n",
            "Epoch 18/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2768 - accuracy: 0.5788 - val_loss: 0.1034 - val_accuracy: 1.0000\n",
            "Epoch 19/100\n",
            "28/28 [==============================] - 11s 378ms/step - loss: 0.2790 - accuracy: 0.5788 - val_loss: 0.1342 - val_accuracy: 0.9259\n",
            "Epoch 20/100\n",
            "28/28 [==============================] - 11s 379ms/step - loss: 0.2737 - accuracy: 0.5788 - val_loss: 0.0755 - val_accuracy: 1.0000\n",
            "Epoch 21/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2786 - accuracy: 0.5788 - val_loss: 0.0710 - val_accuracy: 1.0000\n",
            "Epoch 22/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2780 - accuracy: 0.5788 - val_loss: 0.0685 - val_accuracy: 1.0000\n",
            "Epoch 23/100\n",
            "28/28 [==============================] - 11s 378ms/step - loss: 0.2753 - accuracy: 0.5788 - val_loss: 0.0439 - val_accuracy: 1.0000\n",
            "Epoch 24/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2662 - accuracy: 0.5788 - val_loss: 0.0447 - val_accuracy: 1.0000\n",
            "Epoch 25/100\n",
            "28/28 [==============================] - 11s 380ms/step - loss: 0.2674 - accuracy: 0.5788 - val_loss: 0.0678 - val_accuracy: 1.0000\n",
            "Epoch 26/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2623 - accuracy: 0.5788 - val_loss: 0.0490 - val_accuracy: 1.0000\n",
            "Epoch 27/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2638 - accuracy: 0.5788 - val_loss: 0.0545 - val_accuracy: 1.0000\n",
            "Epoch 28/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2634 - accuracy: 0.5788 - val_loss: 0.0918 - val_accuracy: 1.0000\n",
            "Epoch 29/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2622 - accuracy: 0.5788 - val_loss: 0.0749 - val_accuracy: 1.0000\n",
            "Epoch 30/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2705 - accuracy: 0.5777 - val_loss: 0.0504 - val_accuracy: 1.0000\n",
            "Epoch 31/100\n",
            "28/28 [==============================] - 11s 379ms/step - loss: 0.2654 - accuracy: 0.5788 - val_loss: 0.0418 - val_accuracy: 1.0000\n",
            "Epoch 32/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2627 - accuracy: 0.5788 - val_loss: 0.0485 - val_accuracy: 1.0000\n",
            "Epoch 33/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2575 - accuracy: 0.5788 - val_loss: 0.0577 - val_accuracy: 1.0000\n",
            "Epoch 34/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2642 - accuracy: 0.5788 - val_loss: 0.0588 - val_accuracy: 1.0000\n",
            "Epoch 35/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2595 - accuracy: 0.5788 - val_loss: 0.0842 - val_accuracy: 1.0000\n",
            "Epoch 36/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2609 - accuracy: 0.5777 - val_loss: 0.0753 - val_accuracy: 1.0000\n",
            "Epoch 37/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2580 - accuracy: 0.5788 - val_loss: 0.0436 - val_accuracy: 1.0000\n",
            "Epoch 38/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.2625 - accuracy: 0.5788 - val_loss: 0.0479 - val_accuracy: 1.0000\n",
            "Epoch 39/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2534 - accuracy: 0.5788 - val_loss: 0.0570 - val_accuracy: 1.0000\n",
            "Epoch 40/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2606 - accuracy: 0.5788 - val_loss: 0.0864 - val_accuracy: 1.0000\n",
            "Epoch 41/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2590 - accuracy: 0.5788 - val_loss: 0.1186 - val_accuracy: 0.9630\n",
            "Epoch 42/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2628 - accuracy: 0.5788 - val_loss: 0.0624 - val_accuracy: 1.0000\n",
            "Epoch 43/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2612 - accuracy: 0.5788 - val_loss: 0.0583 - val_accuracy: 1.0000\n",
            "Epoch 44/100\n",
            "28/28 [==============================] - 11s 375ms/step - loss: 0.2492 - accuracy: 0.5788 - val_loss: 0.0555 - val_accuracy: 1.0000\n",
            "Epoch 45/100\n",
            "28/28 [==============================] - 11s 375ms/step - loss: 0.2553 - accuracy: 0.5788 - val_loss: 0.0818 - val_accuracy: 1.0000\n",
            "Epoch 46/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2603 - accuracy: 0.5788 - val_loss: 0.0351 - val_accuracy: 1.0000\n",
            "Epoch 47/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2644 - accuracy: 0.5788 - val_loss: 0.0399 - val_accuracy: 1.0000\n",
            "Epoch 48/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2515 - accuracy: 0.5788 - val_loss: 0.0745 - val_accuracy: 1.0000\n",
            "Epoch 49/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2535 - accuracy: 0.5788 - val_loss: 0.1160 - val_accuracy: 0.9630\n",
            "Epoch 50/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2588 - accuracy: 0.5788 - val_loss: 0.0997 - val_accuracy: 0.9630\n",
            "Epoch 51/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2648 - accuracy: 0.5788 - val_loss: 0.1405 - val_accuracy: 0.9630\n",
            "Epoch 52/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2567 - accuracy: 0.5788 - val_loss: 0.1160 - val_accuracy: 0.9630\n",
            "Epoch 53/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2549 - accuracy: 0.5788 - val_loss: 0.0656 - val_accuracy: 1.0000\n",
            "Epoch 54/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2497 - accuracy: 0.5788 - val_loss: 0.0672 - val_accuracy: 1.0000\n",
            "Epoch 55/100\n",
            "28/28 [==============================] - 10s 375ms/step - loss: 0.2518 - accuracy: 0.5788 - val_loss: 0.0795 - val_accuracy: 1.0000\n",
            "Epoch 56/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2519 - accuracy: 0.5788 - val_loss: 0.0413 - val_accuracy: 1.0000\n",
            "Epoch 57/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2601 - accuracy: 0.5788 - val_loss: 0.1459 - val_accuracy: 0.9259\n",
            "Epoch 58/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2550 - accuracy: 0.5788 - val_loss: 0.1057 - val_accuracy: 0.9630\n",
            "Epoch 59/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2492 - accuracy: 0.5788 - val_loss: 0.0857 - val_accuracy: 1.0000\n",
            "Epoch 60/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2488 - accuracy: 0.5788 - val_loss: 0.1110 - val_accuracy: 1.0000\n",
            "Epoch 61/100\n",
            "28/28 [==============================] - 11s 378ms/step - loss: 0.2480 - accuracy: 0.5788 - val_loss: 0.0678 - val_accuracy: 1.0000\n",
            "2/2 [==============================] - 1s 111ms/step - loss: 0.1020 - accuracy: 0.9706\n",
            "Epoch 1/100\n",
            "28/28 [==============================] - 16s 413ms/step - loss: 0.5496 - accuracy: 0.5034 - val_loss: 0.3534 - val_accuracy: 0.8519\n",
            "Epoch 2/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.3739 - accuracy: 0.5736 - val_loss: 0.3141 - val_accuracy: 0.8519\n",
            "Epoch 3/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.3533 - accuracy: 0.5816 - val_loss: 0.3365 - val_accuracy: 0.8519\n",
            "Epoch 4/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.3320 - accuracy: 0.5816 - val_loss: 0.2808 - val_accuracy: 0.9259\n",
            "Epoch 5/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.3368 - accuracy: 0.5851 - val_loss: 0.3733 - val_accuracy: 0.9259\n",
            "Epoch 6/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.3097 - accuracy: 0.5920 - val_loss: 0.2980 - val_accuracy: 0.8519\n",
            "Epoch 7/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.3058 - accuracy: 0.5931 - val_loss: 0.3236 - val_accuracy: 0.8519\n",
            "Epoch 8/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.3050 - accuracy: 0.5931 - val_loss: 0.2375 - val_accuracy: 0.9259\n",
            "Epoch 9/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.3007 - accuracy: 0.5931 - val_loss: 0.5448 - val_accuracy: 0.6296\n",
            "Epoch 10/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2944 - accuracy: 0.5954 - val_loss: 0.2515 - val_accuracy: 0.9259\n",
            "Epoch 11/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2849 - accuracy: 0.5943 - val_loss: 0.3119 - val_accuracy: 0.8519\n",
            "Epoch 12/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2915 - accuracy: 0.5966 - val_loss: 0.2628 - val_accuracy: 0.8148\n",
            "Epoch 13/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2758 - accuracy: 0.5954 - val_loss: 0.2853 - val_accuracy: 0.8519\n",
            "Epoch 14/100\n",
            "28/28 [==============================] - 11s 380ms/step - loss: 0.2854 - accuracy: 0.5954 - val_loss: 0.2246 - val_accuracy: 0.8889\n",
            "Epoch 15/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2802 - accuracy: 0.5966 - val_loss: 0.2510 - val_accuracy: 0.8519\n",
            "Epoch 16/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2796 - accuracy: 0.5954 - val_loss: 0.2858 - val_accuracy: 0.8519\n",
            "Epoch 17/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2689 - accuracy: 0.5966 - val_loss: 0.3577 - val_accuracy: 0.8519\n",
            "Epoch 18/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2792 - accuracy: 0.5966 - val_loss: 0.2437 - val_accuracy: 0.8889\n",
            "Epoch 19/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2809 - accuracy: 0.5977 - val_loss: 0.2373 - val_accuracy: 0.8889\n",
            "Epoch 20/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2817 - accuracy: 0.5977 - val_loss: 0.2535 - val_accuracy: 0.8519\n",
            "Epoch 21/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2699 - accuracy: 0.5977 - val_loss: 0.2572 - val_accuracy: 0.8889\n",
            "Epoch 22/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.2694 - accuracy: 0.5977 - val_loss: 0.2644 - val_accuracy: 0.8889\n",
            "Epoch 23/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2685 - accuracy: 0.5966 - val_loss: 0.2669 - val_accuracy: 0.8889\n",
            "Epoch 24/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2715 - accuracy: 0.5966 - val_loss: 0.2754 - val_accuracy: 0.9259\n",
            "Epoch 25/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2676 - accuracy: 0.5977 - val_loss: 0.2930 - val_accuracy: 0.8889\n",
            "Epoch 26/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.2670 - accuracy: 0.5977 - val_loss: 0.3948 - val_accuracy: 0.8148\n",
            "Epoch 27/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2971 - accuracy: 0.5931 - val_loss: 0.2293 - val_accuracy: 0.9630\n",
            "Epoch 28/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.2750 - accuracy: 0.5954 - val_loss: 0.2423 - val_accuracy: 0.9630\n",
            "Epoch 29/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2616 - accuracy: 0.5977 - val_loss: 0.3736 - val_accuracy: 0.7778\n",
            "2/2 [==============================] - 1s 106ms/step - loss: 0.1269 - accuracy: 0.9412\n",
            "Epoch 1/100\n",
            "28/28 [==============================] - 16s 412ms/step - loss: 0.4887 - accuracy: 0.5207 - val_loss: 0.5347 - val_accuracy: 0.8519\n",
            "Epoch 2/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.3821 - accuracy: 0.5552 - val_loss: 0.4043 - val_accuracy: 0.8889\n",
            "Epoch 3/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.3740 - accuracy: 0.5586 - val_loss: 0.3122 - val_accuracy: 0.8519\n",
            "Epoch 4/100\n",
            "28/28 [==============================] - 10s 370ms/step - loss: 0.3465 - accuracy: 0.5644 - val_loss: 0.3226 - val_accuracy: 0.8889\n",
            "Epoch 5/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.3276 - accuracy: 0.5713 - val_loss: 0.3491 - val_accuracy: 0.8889\n",
            "Epoch 6/100\n",
            "28/28 [==============================] - 11s 377ms/step - loss: 0.3258 - accuracy: 0.5667 - val_loss: 0.2427 - val_accuracy: 0.9259\n",
            "Epoch 7/100\n",
            "28/28 [==============================] - 10s 372ms/step - loss: 0.3158 - accuracy: 0.5701 - val_loss: 0.2698 - val_accuracy: 0.8889\n",
            "Epoch 8/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.3105 - accuracy: 0.5759 - val_loss: 0.2061 - val_accuracy: 0.9259\n",
            "Epoch 9/100\n",
            "28/28 [==============================] - 10s 371ms/step - loss: 0.3035 - accuracy: 0.5747 - val_loss: 0.2308 - val_accuracy: 0.8889\n",
            "Epoch 10/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.3009 - accuracy: 0.5701 - val_loss: 0.2066 - val_accuracy: 0.9259\n",
            "Epoch 11/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2975 - accuracy: 0.5759 - val_loss: 0.1958 - val_accuracy: 0.9259\n",
            "Epoch 12/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.2904 - accuracy: 0.5747 - val_loss: 0.2519 - val_accuracy: 0.8889\n",
            "Epoch 13/100\n",
            "28/28 [==============================] - 11s 376ms/step - loss: 0.2938 - accuracy: 0.5759 - val_loss: 0.1170 - val_accuracy: 1.0000\n",
            "Epoch 14/100\n",
            "28/28 [==============================] - 10s 374ms/step - loss: 0.3007 - accuracy: 0.5724 - val_loss: 0.1474 - val_accuracy: 1.0000\n",
            "Epoch 15/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.3002 - accuracy: 0.5759 - val_loss: 0.2002 - val_accuracy: 0.9259\n",
            "Epoch 16/100\n",
            "28/28 [==============================] - 10s 373ms/step - loss: 0.2847 - accuracy: 0.5770 - val_loss: 0.1611 - val_accuracy: 0.9259\n",
            "Epoch 17/100\n",
            " 2/28 [=>............................] - ETA: 9s - loss: 0.3263 - accuracy: 0.5000 "
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0N7v4PDP4Dv2",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "2795bfbb-29d9-48a9-de1f-7925be898d2f"
      },
      "source": [
        "print(f'Mean test accuracy is {\"{:.3f}\".format(sum(test_accuracy_per_run)/number_of_repeat)}, mean test f1 score is {\"{:.3f}\".format(sum(f1_score_per_run)/number_of_repeat)}, \\\n",
        "max test accuracy is {\"{:.3f}\".format(max(test_accuracy_per_run))}, max test f1 score is {\"{:.3f}\".format(max(f1_score_per_run))}, \\\n",
        "min test accuracy is {\"{:.3f}\".format(min(test_accuracy_per_run))}, min test f1 score is {\"{:.3f}\".format(min(f1_score_per_run))}, \\\n",
        "std of test accuracy is {\"{:.3f}\".format(np.std(test_accuracy_per_run, axis=0))}, std of test f1 score is {\"{:.3f}\".format(np.std(f1_score_per_run, axis=0))}')\n",
        "# print(f'Time elapsed through all process: {\"{:.3f}\".format(elapsed)}, sec') "
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mean test accuracy is 0.974, mean test f1 score is 0.970, max test accuracy is 0.994, max test f1 score is 0.993, min test accuracy is 0.959, min test f1 score is 0.951, std of test accuracy is 0.012, std of test f1 score is 0.014\n"
          ]
        }
      ]
    }
  ]
}