neural_network.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "import plotly\n",
    "from plotly.express import scatter"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Pregnancies</th>\n",
       "      <th>Glucose</th>\n",
       "      <th>BloodPressure</th>\n",
       "      <th>SkinThickness</th>\n",
       "      <th>Insulin</th>\n",
       "      <th>BMI</th>\n",
       "      <th>DiabetesPedigreeFunction</th>\n",
       "      <th>Age</th>\n",
       "      <th>Outcome</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>6</td>\n",
       "      <td>148</td>\n",
       "      <td>72</td>\n",
       "      <td>35</td>\n",
       "      <td>0</td>\n",
       "      <td>33.6</td>\n",
       "      <td>0.627</td>\n",
       "      <td>50</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>85</td>\n",
       "      <td>66</td>\n",
       "      <td>29</td>\n",
       "      <td>0</td>\n",
       "      <td>26.6</td>\n",
       "      <td>0.351</td>\n",
       "      <td>31</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>8</td>\n",
       "      <td>183</td>\n",
       "      <td>64</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>23.3</td>\n",
       "      <td>0.672</td>\n",
       "      <td>32</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>89</td>\n",
       "      <td>66</td>\n",
       "      <td>23</td>\n",
       "      <td>94</td>\n",
       "      <td>28.1</td>\n",
       "      <td>0.167</td>\n",
       "      <td>21</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>137</td>\n",
       "      <td>40</td>\n",
       "      <td>35</td>\n",
       "      <td>168</td>\n",
       "      <td>43.1</td>\n",
       "      <td>2.288</td>\n",
       "      <td>33</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Pregnancies  Glucose  BloodPressure  SkinThickness  Insulin   BMI  \\\n",
       "0            6      148             72             35        0  33.6   \n",
       "1            1       85             66             29        0  26.6   \n",
       "2            8      183             64              0        0  23.3   \n",
       "3            1       89             66             23       94  28.1   \n",
       "4            0      137             40             35      168  43.1   \n",
       "\n",
       "   DiabetesPedigreeFunction  Age  Outcome  \n",
       "0                     0.627   50        1  \n",
       "1                     0.351   31        0  \n",
       "2                     0.672   32        1  \n",
       "3                     0.167   21        0  \n",
       "4                     2.288   33        1  "
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = pd.read_csv('./datasets/diabetes.csv')\n",
    "\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = data.iloc[:, 1:-1]\n",
    "y = data.iloc[:, -1]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Split Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "scaler = MinMaxScaler()\n",
    "X = scaler.fit_transform(X)\n",
    "y = pd.get_dummies(y)\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=123)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Build Neural Network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/50\n",
      "12/12 [==============================] - 4s 74ms/step - loss: 0.6892 - categorical_accuracy: 0.5174 - recall_71: 0.5590 - val_loss: 0.6675 - val_categorical_accuracy: 0.6302 - val_recall_71: 0.0411\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.6536 - categorical_accuracy: 0.6632 - recall_71: 0.0051 - val_loss: 0.6658 - val_categorical_accuracy: 0.6198 - val_recall_71: 0.0000e+00\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6462 - categorical_accuracy: 0.6615 - recall_71: 0.0000e+00 - val_loss: 0.6655 - val_categorical_accuracy: 0.6198 - val_recall_71: 0.0000e+00\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6392 - categorical_accuracy: 0.6615 - recall_71: 0.0000e+00 - val_loss: 0.6564 - val_categorical_accuracy: 0.6198 - val_recall_71: 0.0000e+00\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.6317 - categorical_accuracy: 0.6615 - recall_71: 0.0000e+00 - val_loss: 0.6469 - val_categorical_accuracy: 0.6198 - val_recall_71: 0.0000e+00\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 28ms/step - loss: 0.6270 - categorical_accuracy: 0.6632 - recall_71: 0.0103 - val_loss: 0.6389 - val_categorical_accuracy: 0.6354 - val_recall_71: 0.0411\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.6206 - categorical_accuracy: 0.6597 - recall_71: 0.0000e+00 - val_loss: 0.6367 - val_categorical_accuracy: 0.6198 - val_recall_71: 0.0000e+00\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6136 - categorical_accuracy: 0.6649 - recall_71: 0.0154 - val_loss: 0.6278 - val_categorical_accuracy: 0.6406 - val_recall_71: 0.0548\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.6065 - categorical_accuracy: 0.6701 - recall_71: 0.0462 - val_loss: 0.6165 - val_categorical_accuracy: 0.6615 - val_recall_71: 0.1233\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5970 - categorical_accuracy: 0.6823 - recall_71: 0.0923 - val_loss: 0.6069 - val_categorical_accuracy: 0.6823 - val_recall_71: 0.1781\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5878 - categorical_accuracy: 0.7031 - recall_71: 0.1897 - val_loss: 0.5965 - val_categorical_accuracy: 0.6979 - val_recall_71: 0.2329\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5786 - categorical_accuracy: 0.7049 - recall_71: 0.1949 - val_loss: 0.5911 - val_categorical_accuracy: 0.7031 - val_recall_71: 0.2329\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5703 - categorical_accuracy: 0.7101 - recall_71: 0.2000 - val_loss: 0.5801 - val_categorical_accuracy: 0.6979 - val_recall_71: 0.2740\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5644 - categorical_accuracy: 0.7066 - recall_71: 0.3385 - val_loss: 0.5705 - val_categorical_accuracy: 0.7031 - val_recall_71: 0.3425\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5545 - categorical_accuracy: 0.7118 - recall_71: 0.2821 - val_loss: 0.5671 - val_categorical_accuracy: 0.7135 - val_recall_71: 0.3014\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5487 - categorical_accuracy: 0.7240 - recall_71: 0.2769 - val_loss: 0.5583 - val_categorical_accuracy: 0.7188 - val_recall_71: 0.3699\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.5425 - categorical_accuracy: 0.7188 - recall_71: 0.4154 - val_loss: 0.5515 - val_categorical_accuracy: 0.7396 - val_recall_71: 0.4795\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5333 - categorical_accuracy: 0.7205 - recall_71: 0.3590 - val_loss: 0.5491 - val_categorical_accuracy: 0.7448 - val_recall_71: 0.4247\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5287 - categorical_accuracy: 0.7309 - recall_71: 0.3179 - val_loss: 0.5431 - val_categorical_accuracy: 0.7500 - val_recall_71: 0.4658\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.5213 - categorical_accuracy: 0.7292 - recall_71: 0.3744 - val_loss: 0.5349 - val_categorical_accuracy: 0.7552 - val_recall_71: 0.5205\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.5206 - categorical_accuracy: 0.7465 - recall_71: 0.5128 - val_loss: 0.5329 - val_categorical_accuracy: 0.7708 - val_recall_71: 0.5205\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5124 - categorical_accuracy: 0.7448 - recall_71: 0.3846 - val_loss: 0.5277 - val_categorical_accuracy: 0.7708 - val_recall_71: 0.5205\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5077 - categorical_accuracy: 0.7465 - recall_71: 0.5231 - val_loss: 0.5224 - val_categorical_accuracy: 0.7760 - val_recall_71: 0.5616\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.5017 - categorical_accuracy: 0.7587 - recall_71: 0.4410 - val_loss: 0.5243 - val_categorical_accuracy: 0.7448 - val_recall_71: 0.4521\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.4959 - categorical_accuracy: 0.7691 - recall_71: 0.4615 - val_loss: 0.5164 - val_categorical_accuracy: 0.7812 - val_recall_71: 0.5753\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 25ms/step - loss: 0.4923 - categorical_accuracy: 0.7535 - recall_71: 0.5179 - val_loss: 0.5151 - val_categorical_accuracy: 0.7708 - val_recall_71: 0.5479\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.4918 - categorical_accuracy: 0.7674 - recall_71: 0.4718 - val_loss: 0.5115 - val_categorical_accuracy: 0.7708 - val_recall_71: 0.5616\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4853 - categorical_accuracy: 0.7674 - recall_71: 0.5641 - val_loss: 0.5121 - val_categorical_accuracy: 0.7552 - val_recall_71: 0.4932\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4848 - categorical_accuracy: 0.7569 - recall_71: 0.4154 - val_loss: 0.5070 - val_categorical_accuracy: 0.7760 - val_recall_71: 0.5753\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.4790 - categorical_accuracy: 0.7535 - recall_71: 0.5795 - val_loss: 0.5049 - val_categorical_accuracy: 0.7812 - val_recall_71: 0.6027\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4750 - categorical_accuracy: 0.7691 - recall_71: 0.4769 - val_loss: 0.5079 - val_categorical_accuracy: 0.7604 - val_recall_71: 0.5205\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4690 - categorical_accuracy: 0.7726 - recall_71: 0.5333 - val_loss: 0.5000 - val_categorical_accuracy: 0.8021 - val_recall_71: 0.6575\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4690 - categorical_accuracy: 0.7795 - recall_71: 0.5487 - val_loss: 0.5018 - val_categorical_accuracy: 0.8021 - val_recall_71: 0.6438\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4662 - categorical_accuracy: 0.7708 - recall_71: 0.5333 - val_loss: 0.4991 - val_categorical_accuracy: 0.7969 - val_recall_71: 0.6438\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.4625 - categorical_accuracy: 0.7708 - recall_71: 0.5179 - val_loss: 0.4993 - val_categorical_accuracy: 0.7969 - val_recall_71: 0.6164\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4603 - categorical_accuracy: 0.7726 - recall_71: 0.5231 - val_loss: 0.4965 - val_categorical_accuracy: 0.8073 - val_recall_71: 0.6575\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4611 - categorical_accuracy: 0.7778 - recall_71: 0.5385 - val_loss: 0.5009 - val_categorical_accuracy: 0.7917 - val_recall_71: 0.6027\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4573 - categorical_accuracy: 0.7708 - recall_71: 0.5846 - val_loss: 0.4964 - val_categorical_accuracy: 0.7969 - val_recall_71: 0.6575\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4568 - categorical_accuracy: 0.7743 - recall_71: 0.5282 - val_loss: 0.4988 - val_categorical_accuracy: 0.7917 - val_recall_71: 0.6027\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4541 - categorical_accuracy: 0.7795 - recall_71: 0.5949 - val_loss: 0.4978 - val_categorical_accuracy: 0.7969 - val_recall_71: 0.6164\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4533 - categorical_accuracy: 0.7674 - recall_71: 0.5385 - val_loss: 0.4993 - val_categorical_accuracy: 0.7969 - val_recall_71: 0.6301\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4510 - categorical_accuracy: 0.7795 - recall_71: 0.5333 - val_loss: 0.4990 - val_categorical_accuracy: 0.7969 - val_recall_71: 0.6164\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 25ms/step - loss: 0.4490 - categorical_accuracy: 0.7795 - recall_71: 0.5744 - val_loss: 0.4979 - val_categorical_accuracy: 0.7969 - val_recall_71: 0.6164\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4498 - categorical_accuracy: 0.7743 - recall_71: 0.6000 - val_loss: 0.5033 - val_categorical_accuracy: 0.7865 - val_recall_71: 0.5890\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.4481 - categorical_accuracy: 0.7778 - recall_71: 0.5333 - val_loss: 0.4987 - val_categorical_accuracy: 0.7969 - val_recall_71: 0.6164\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4487 - categorical_accuracy: 0.7865 - recall_71: 0.5436 - val_loss: 0.4973 - val_categorical_accuracy: 0.8021 - val_recall_71: 0.6712\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4493 - categorical_accuracy: 0.7847 - recall_71: 0.6615 - val_loss: 0.5044 - val_categorical_accuracy: 0.7865 - val_recall_71: 0.5890\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.4515 - categorical_accuracy: 0.7691 - recall_71: 0.4769 - val_loss: 0.4993 - val_categorical_accuracy: 0.7969 - val_recall_71: 0.6575\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4448 - categorical_accuracy: 0.7812 - recall_71: 0.6205 - val_loss: 0.5008 - val_categorical_accuracy: 0.7969 - val_recall_71: 0.6301\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4466 - categorical_accuracy: 0.7847 - recall_71: 0.5538 - val_loss: 0.4995 - val_categorical_accuracy: 0.7969 - val_recall_71: 0.6438\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(<keras.engine.sequential.Sequential at 0x26e02a1b210>,\n",
       " 0.7864583134651184,\n",
       " 0.6615384817123413,\n",
       " 0.8072916865348816,\n",
       " 0.6712328791618347)"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def NurNet(norons_in_layer, epochs, batch_size):\n",
    "    Layers_num = len(norons_in_layer)\n",
    "    # str = 'tf.keras.models.Sequential([tf.keras.layers.Dense(norons_in_layers[0], activation=\"relu\", input_shape=(7, ))])'\n",
    "    str=\"tf.keras.models.Sequential([tf.keras.layers.Dense(norons_in_layer[0], activation='relu', input_shape=(7,)),\"\n",
    "\n",
    "    for i in range(1, Layers_num):\n",
    "        # str = str+f\"tf.keras.layers.Dense({norons_in_layer[i]}, activation='relu')\"\n",
    "        # str = str+f\"tf.keras.layers.Dense(2, actovation='softmax')\"\n",
    "        str=str+f\"tf.keras.layers.Dense({norons_in_layer[i]}, activation='relu'),\"\n",
    "        str=str+f\"tf.keras.layers.Dense(2, activation='softmax')])\"\n",
    "\n",
    "        model = eval(str)\n",
    "        # model.compile(optimizer=tf.keras.optimizers.Adam(0.001), loss=tf.keras.losses.CategorycalCrossentropy(), metrics=[tf.keras.metrics.CategoricalCrossentropy(), tf.keras.metrics.Recall(class_id=1)])\n",
    "        model.compile(optimizer=tf.keras.optimizers.Adam(0.001),\n",
    "                  loss=tf.keras.losses.CategoricalCrossentropy(),\n",
    "                  metrics=[tf.keras.metrics.CategoricalAccuracy(),\n",
    "                           tf.keras.metrics.Recall(class_id=1)]\n",
    "                 )\n",
    "        history = model.fit(X_train, y_train, epochs=epochs, validation_data=(X_test, y_test), batch_size=batch_size)\n",
    "        keys = list(history.history.keys())\n",
    "\n",
    "        accuracy_train = np.max(history.history['categorical_accuracy'])\n",
    "        accuracy_test = np.max(history.history[keys[2]])\n",
    "        recall_train = np.max(history.history[keys[4]])\n",
    "        recall_test = np.max(history.history[keys[5]])\n",
    "\n",
    "        return (model, accuracy_train, accuracy_test, recall_train, recall_test)\n",
    "    \n",
    "\n",
    "NurNet([25, 34, 45, 36, 18, 12, 7], 50, 50)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "results = pd.DataFrame(columns=['Layers_num', 'norons_in_layer', 'model', 'accuracy_train', 'accuracy_test', 'recall_train', 'recall_test'])\n",
    "architecture_num = 0"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Train Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/50\n",
      "12/12 [==============================] - 4s 61ms/step - loss: 0.6990 - categorical_accuracy: 0.4618 - recall_72: 0.5949 - val_loss: 0.6778 - val_categorical_accuracy: 0.6354 - val_recall_72: 0.0411\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.6590 - categorical_accuracy: 0.6632 - recall_72: 0.0051 - val_loss: 0.6680 - val_categorical_accuracy: 0.6198 - val_recall_72: 0.0000e+00\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6490 - categorical_accuracy: 0.6615 - recall_72: 0.0000e+00 - val_loss: 0.6633 - val_categorical_accuracy: 0.6198 - val_recall_72: 0.0000e+00\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.6377 - categorical_accuracy: 0.6615 - recall_72: 0.0000e+00 - val_loss: 0.6514 - val_categorical_accuracy: 0.6250 - val_recall_72: 0.0137\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 21ms/step - loss: 0.6292 - categorical_accuracy: 0.6632 - recall_72: 0.0051 - val_loss: 0.6431 - val_categorical_accuracy: 0.6250 - val_recall_72: 0.0137\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 22ms/step - loss: 0.6205 - categorical_accuracy: 0.6632 - recall_72: 0.0154 - val_loss: 0.6326 - val_categorical_accuracy: 0.6458 - val_recall_72: 0.0822\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 20ms/step - loss: 0.6114 - categorical_accuracy: 0.6719 - recall_72: 0.0564 - val_loss: 0.6231 - val_categorical_accuracy: 0.6562 - val_recall_72: 0.1096\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.6012 - categorical_accuracy: 0.6753 - recall_72: 0.0923 - val_loss: 0.6137 - val_categorical_accuracy: 0.6771 - val_recall_72: 0.1644\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.5929 - categorical_accuracy: 0.6788 - recall_72: 0.1179 - val_loss: 0.6020 - val_categorical_accuracy: 0.6771 - val_recall_72: 0.2055\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.5818 - categorical_accuracy: 0.6927 - recall_72: 0.1846 - val_loss: 0.5920 - val_categorical_accuracy: 0.6927 - val_recall_72: 0.2466\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.5717 - categorical_accuracy: 0.7014 - recall_72: 0.2256 - val_loss: 0.5797 - val_categorical_accuracy: 0.7031 - val_recall_72: 0.2877\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 20ms/step - loss: 0.5590 - categorical_accuracy: 0.7049 - recall_72: 0.2205 - val_loss: 0.5659 - val_categorical_accuracy: 0.7135 - val_recall_72: 0.3151\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5490 - categorical_accuracy: 0.7135 - recall_72: 0.3641 - val_loss: 0.5544 - val_categorical_accuracy: 0.7240 - val_recall_72: 0.3562\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.5367 - categorical_accuracy: 0.7135 - recall_72: 0.2667 - val_loss: 0.5484 - val_categorical_accuracy: 0.7240 - val_recall_72: 0.3425\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.5287 - categorical_accuracy: 0.7292 - recall_72: 0.4051 - val_loss: 0.5360 - val_categorical_accuracy: 0.7396 - val_recall_72: 0.4110\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.5166 - categorical_accuracy: 0.7344 - recall_72: 0.3641 - val_loss: 0.5280 - val_categorical_accuracy: 0.7500 - val_recall_72: 0.4247\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5090 - categorical_accuracy: 0.7535 - recall_72: 0.4513 - val_loss: 0.5172 - val_categorical_accuracy: 0.7708 - val_recall_72: 0.5616\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.5018 - categorical_accuracy: 0.7483 - recall_72: 0.4410 - val_loss: 0.5112 - val_categorical_accuracy: 0.7708 - val_recall_72: 0.5342\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 36ms/step - loss: 0.4939 - categorical_accuracy: 0.7639 - recall_72: 0.4923 - val_loss: 0.5074 - val_categorical_accuracy: 0.7604 - val_recall_72: 0.4795\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 31ms/step - loss: 0.4896 - categorical_accuracy: 0.7674 - recall_72: 0.4718 - val_loss: 0.5003 - val_categorical_accuracy: 0.7760 - val_recall_72: 0.5616\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 41ms/step - loss: 0.4823 - categorical_accuracy: 0.7708 - recall_72: 0.5026 - val_loss: 0.5002 - val_categorical_accuracy: 0.7552 - val_recall_72: 0.4932\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 23ms/step - loss: 0.4774 - categorical_accuracy: 0.7708 - recall_72: 0.5128 - val_loss: 0.4928 - val_categorical_accuracy: 0.7917 - val_recall_72: 0.5890\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4731 - categorical_accuracy: 0.7726 - recall_72: 0.5026 - val_loss: 0.4886 - val_categorical_accuracy: 0.7917 - val_recall_72: 0.6164\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4699 - categorical_accuracy: 0.7674 - recall_72: 0.5385 - val_loss: 0.4895 - val_categorical_accuracy: 0.7812 - val_recall_72: 0.5616\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.4664 - categorical_accuracy: 0.7691 - recall_72: 0.4718 - val_loss: 0.4822 - val_categorical_accuracy: 0.8073 - val_recall_72: 0.6575\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.4690 - categorical_accuracy: 0.7760 - recall_72: 0.5795 - val_loss: 0.4864 - val_categorical_accuracy: 0.7708 - val_recall_72: 0.5479\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.4590 - categorical_accuracy: 0.7743 - recall_72: 0.5487 - val_loss: 0.4844 - val_categorical_accuracy: 0.7865 - val_recall_72: 0.5890\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4614 - categorical_accuracy: 0.7691 - recall_72: 0.5385 - val_loss: 0.4903 - val_categorical_accuracy: 0.7760 - val_recall_72: 0.5342\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4548 - categorical_accuracy: 0.7726 - recall_72: 0.5026 - val_loss: 0.4786 - val_categorical_accuracy: 0.7969 - val_recall_72: 0.6301\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.4574 - categorical_accuracy: 0.7726 - recall_72: 0.6154 - val_loss: 0.4904 - val_categorical_accuracy: 0.7656 - val_recall_72: 0.5205\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.4695 - categorical_accuracy: 0.7622 - recall_72: 0.4718 - val_loss: 0.4822 - val_categorical_accuracy: 0.7917 - val_recall_72: 0.6986\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4519 - categorical_accuracy: 0.7830 - recall_72: 0.5641 - val_loss: 0.4959 - val_categorical_accuracy: 0.7656 - val_recall_72: 0.5068\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4485 - categorical_accuracy: 0.7830 - recall_72: 0.5641 - val_loss: 0.4798 - val_categorical_accuracy: 0.7969 - val_recall_72: 0.6438\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4460 - categorical_accuracy: 0.7882 - recall_72: 0.5897 - val_loss: 0.4848 - val_categorical_accuracy: 0.7917 - val_recall_72: 0.5890\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4472 - categorical_accuracy: 0.7726 - recall_72: 0.4821 - val_loss: 0.4813 - val_categorical_accuracy: 0.8073 - val_recall_72: 0.6438\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.4454 - categorical_accuracy: 0.7882 - recall_72: 0.6154 - val_loss: 0.4828 - val_categorical_accuracy: 0.8021 - val_recall_72: 0.6301\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 26ms/step - loss: 0.4428 - categorical_accuracy: 0.7917 - recall_72: 0.5846 - val_loss: 0.4822 - val_categorical_accuracy: 0.8073 - val_recall_72: 0.6438\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4401 - categorical_accuracy: 0.7917 - recall_72: 0.6000 - val_loss: 0.4858 - val_categorical_accuracy: 0.7917 - val_recall_72: 0.5890\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4412 - categorical_accuracy: 0.7760 - recall_72: 0.5333 - val_loss: 0.4831 - val_categorical_accuracy: 0.8021 - val_recall_72: 0.6301\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4394 - categorical_accuracy: 0.7934 - recall_72: 0.6154 - val_loss: 0.4938 - val_categorical_accuracy: 0.7812 - val_recall_72: 0.5616\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4394 - categorical_accuracy: 0.7812 - recall_72: 0.5231 - val_loss: 0.4838 - val_categorical_accuracy: 0.7865 - val_recall_72: 0.6712\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4377 - categorical_accuracy: 0.7951 - recall_72: 0.6308 - val_loss: 0.4899 - val_categorical_accuracy: 0.7865 - val_recall_72: 0.5753\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4364 - categorical_accuracy: 0.7847 - recall_72: 0.5538 - val_loss: 0.4846 - val_categorical_accuracy: 0.7969 - val_recall_72: 0.6438\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4371 - categorical_accuracy: 0.7882 - recall_72: 0.5692 - val_loss: 0.4852 - val_categorical_accuracy: 0.7969 - val_recall_72: 0.6575\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4397 - categorical_accuracy: 0.7865 - recall_72: 0.6410 - val_loss: 0.4927 - val_categorical_accuracy: 0.7917 - val_recall_72: 0.5890\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4341 - categorical_accuracy: 0.7847 - recall_72: 0.5949 - val_loss: 0.4898 - val_categorical_accuracy: 0.7969 - val_recall_72: 0.6164\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4333 - categorical_accuracy: 0.7795 - recall_72: 0.5385 - val_loss: 0.4868 - val_categorical_accuracy: 0.7917 - val_recall_72: 0.6575\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4362 - categorical_accuracy: 0.7899 - recall_72: 0.6718 - val_loss: 0.5069 - val_categorical_accuracy: 0.7760 - val_recall_72: 0.5479\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 20ms/step - loss: 0.4426 - categorical_accuracy: 0.7760 - recall_72: 0.4872 - val_loss: 0.4879 - val_categorical_accuracy: 0.7812 - val_recall_72: 0.6575\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4396 - categorical_accuracy: 0.7760 - recall_72: 0.6410 - val_loss: 0.5002 - val_categorical_accuracy: 0.7812 - val_recall_72: 0.5616\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 4s 66ms/step - loss: 0.6810 - categorical_accuracy: 0.5729 - recall_73: 0.4513 - val_loss: 0.6639 - val_categorical_accuracy: 0.6198 - val_recall_73: 0.0274\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6487 - categorical_accuracy: 0.6615 - recall_73: 0.0000e+00 - val_loss: 0.6582 - val_categorical_accuracy: 0.6198 - val_recall_73: 0.0000e+00\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6374 - categorical_accuracy: 0.6615 - recall_73: 0.0000e+00 - val_loss: 0.6552 - val_categorical_accuracy: 0.6198 - val_recall_73: 0.0000e+00\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.6308 - categorical_accuracy: 0.6615 - recall_73: 0.0000e+00 - val_loss: 0.6460 - val_categorical_accuracy: 0.6198 - val_recall_73: 0.0000e+00\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.6233 - categorical_accuracy: 0.6615 - recall_73: 0.0000e+00 - val_loss: 0.6348 - val_categorical_accuracy: 0.6198 - val_recall_73: 0.0000e+00\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.6154 - categorical_accuracy: 0.6615 - recall_73: 0.0000e+00 - val_loss: 0.6266 - val_categorical_accuracy: 0.6198 - val_recall_73: 0.0000e+00\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.6053 - categorical_accuracy: 0.6562 - recall_73: 0.0000e+00 - val_loss: 0.6151 - val_categorical_accuracy: 0.6250 - val_recall_73: 0.0411\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5962 - categorical_accuracy: 0.6684 - recall_73: 0.0410 - val_loss: 0.6083 - val_categorical_accuracy: 0.6354 - val_recall_73: 0.0685\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.5866 - categorical_accuracy: 0.6771 - recall_73: 0.0923 - val_loss: 0.5970 - val_categorical_accuracy: 0.7031 - val_recall_73: 0.2466\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5782 - categorical_accuracy: 0.6858 - recall_73: 0.1436 - val_loss: 0.5887 - val_categorical_accuracy: 0.6875 - val_recall_73: 0.2466\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5702 - categorical_accuracy: 0.6979 - recall_73: 0.3128 - val_loss: 0.5796 - val_categorical_accuracy: 0.7292 - val_recall_73: 0.3836\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5569 - categorical_accuracy: 0.7049 - recall_73: 0.2872 - val_loss: 0.5750 - val_categorical_accuracy: 0.7188 - val_recall_73: 0.3288\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5478 - categorical_accuracy: 0.7153 - recall_73: 0.2667 - val_loss: 0.5677 - val_categorical_accuracy: 0.7292 - val_recall_73: 0.3699\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5378 - categorical_accuracy: 0.7344 - recall_73: 0.3795 - val_loss: 0.5570 - val_categorical_accuracy: 0.7031 - val_recall_73: 0.3973\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5274 - categorical_accuracy: 0.7396 - recall_73: 0.4000 - val_loss: 0.5516 - val_categorical_accuracy: 0.7188 - val_recall_73: 0.4110\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5184 - categorical_accuracy: 0.7483 - recall_73: 0.3590 - val_loss: 0.5485 - val_categorical_accuracy: 0.7292 - val_recall_73: 0.3973\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 29ms/step - loss: 0.5094 - categorical_accuracy: 0.7691 - recall_73: 0.5179 - val_loss: 0.5388 - val_categorical_accuracy: 0.7604 - val_recall_73: 0.5342\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5084 - categorical_accuracy: 0.7431 - recall_73: 0.3436 - val_loss: 0.5374 - val_categorical_accuracy: 0.7552 - val_recall_73: 0.4932\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4926 - categorical_accuracy: 0.7778 - recall_73: 0.5641 - val_loss: 0.5337 - val_categorical_accuracy: 0.7552 - val_recall_73: 0.5068\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4849 - categorical_accuracy: 0.7726 - recall_73: 0.4667 - val_loss: 0.5326 - val_categorical_accuracy: 0.7604 - val_recall_73: 0.4932\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4791 - categorical_accuracy: 0.7847 - recall_73: 0.5692 - val_loss: 0.5297 - val_categorical_accuracy: 0.7552 - val_recall_73: 0.5068\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4778 - categorical_accuracy: 0.7760 - recall_73: 0.4564 - val_loss: 0.5249 - val_categorical_accuracy: 0.7760 - val_recall_73: 0.6027\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.4685 - categorical_accuracy: 0.7847 - recall_73: 0.6154 - val_loss: 0.5308 - val_categorical_accuracy: 0.7448 - val_recall_73: 0.4795\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4646 - categorical_accuracy: 0.7882 - recall_73: 0.4872 - val_loss: 0.5236 - val_categorical_accuracy: 0.7760 - val_recall_73: 0.6301\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4612 - categorical_accuracy: 0.7899 - recall_73: 0.6615 - val_loss: 0.5268 - val_categorical_accuracy: 0.7656 - val_recall_73: 0.5479\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4603 - categorical_accuracy: 0.7899 - recall_73: 0.5282 - val_loss: 0.5243 - val_categorical_accuracy: 0.7812 - val_recall_73: 0.6164\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4505 - categorical_accuracy: 0.7917 - recall_73: 0.5692 - val_loss: 0.5292 - val_categorical_accuracy: 0.7604 - val_recall_73: 0.5479\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4524 - categorical_accuracy: 0.7882 - recall_73: 0.6205 - val_loss: 0.5332 - val_categorical_accuracy: 0.7604 - val_recall_73: 0.5342\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4567 - categorical_accuracy: 0.7847 - recall_73: 0.4615 - val_loss: 0.5255 - val_categorical_accuracy: 0.7812 - val_recall_73: 0.6438\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4572 - categorical_accuracy: 0.7969 - recall_73: 0.7487 - val_loss: 0.5335 - val_categorical_accuracy: 0.7656 - val_recall_73: 0.5616\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4577 - categorical_accuracy: 0.7778 - recall_73: 0.4615 - val_loss: 0.5283 - val_categorical_accuracy: 0.7760 - val_recall_73: 0.6301\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4432 - categorical_accuracy: 0.7934 - recall_73: 0.6718 - val_loss: 0.5332 - val_categorical_accuracy: 0.7656 - val_recall_73: 0.5753\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 26ms/step - loss: 0.4411 - categorical_accuracy: 0.7951 - recall_73: 0.5590 - val_loss: 0.5317 - val_categorical_accuracy: 0.7708 - val_recall_73: 0.6027\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4376 - categorical_accuracy: 0.7899 - recall_73: 0.5949 - val_loss: 0.5352 - val_categorical_accuracy: 0.7708 - val_recall_73: 0.5753\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4349 - categorical_accuracy: 0.8003 - recall_73: 0.5897 - val_loss: 0.5332 - val_categorical_accuracy: 0.7760 - val_recall_73: 0.6164\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4336 - categorical_accuracy: 0.7917 - recall_73: 0.6103 - val_loss: 0.5346 - val_categorical_accuracy: 0.7760 - val_recall_73: 0.6164\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4348 - categorical_accuracy: 0.7899 - recall_73: 0.5795 - val_loss: 0.5355 - val_categorical_accuracy: 0.7760 - val_recall_73: 0.6164\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4321 - categorical_accuracy: 0.7951 - recall_73: 0.6154 - val_loss: 0.5354 - val_categorical_accuracy: 0.7760 - val_recall_73: 0.6164\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4322 - categorical_accuracy: 0.7969 - recall_73: 0.6000 - val_loss: 0.5398 - val_categorical_accuracy: 0.7708 - val_recall_73: 0.6027\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4297 - categorical_accuracy: 0.7917 - recall_73: 0.6000 - val_loss: 0.5372 - val_categorical_accuracy: 0.7760 - val_recall_73: 0.6164\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4290 - categorical_accuracy: 0.7899 - recall_73: 0.6205 - val_loss: 0.5432 - val_categorical_accuracy: 0.7604 - val_recall_73: 0.5753\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4288 - categorical_accuracy: 0.7986 - recall_73: 0.5744 - val_loss: 0.5390 - val_categorical_accuracy: 0.7708 - val_recall_73: 0.6164\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 20ms/step - loss: 0.4287 - categorical_accuracy: 0.7917 - recall_73: 0.6154 - val_loss: 0.5452 - val_categorical_accuracy: 0.7604 - val_recall_73: 0.5753\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4258 - categorical_accuracy: 0.8021 - recall_73: 0.6051 - val_loss: 0.5430 - val_categorical_accuracy: 0.7708 - val_recall_73: 0.6164\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4301 - categorical_accuracy: 0.7899 - recall_73: 0.6564 - val_loss: 0.5514 - val_categorical_accuracy: 0.7604 - val_recall_73: 0.5342\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4283 - categorical_accuracy: 0.8056 - recall_73: 0.5795 - val_loss: 0.5414 - val_categorical_accuracy: 0.7656 - val_recall_73: 0.6301\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 25ms/step - loss: 0.4268 - categorical_accuracy: 0.7969 - recall_73: 0.6359 - val_loss: 0.5511 - val_categorical_accuracy: 0.7656 - val_recall_73: 0.5753\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.4252 - categorical_accuracy: 0.7969 - recall_73: 0.6205 - val_loss: 0.5463 - val_categorical_accuracy: 0.7708 - val_recall_73: 0.6164\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4241 - categorical_accuracy: 0.7934 - recall_73: 0.6462 - val_loss: 0.5504 - val_categorical_accuracy: 0.7604 - val_recall_73: 0.5616\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4241 - categorical_accuracy: 0.7986 - recall_73: 0.5692 - val_loss: 0.5451 - val_categorical_accuracy: 0.7708 - val_recall_73: 0.6164\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 3s 66ms/step - loss: 0.6649 - categorical_accuracy: 0.6545 - recall_74: 0.0769 - val_loss: 0.6693 - val_categorical_accuracy: 0.6198 - val_recall_74: 0.0000e+00\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6483 - categorical_accuracy: 0.6615 - recall_74: 0.0000e+00 - val_loss: 0.6604 - val_categorical_accuracy: 0.6198 - val_recall_74: 0.0000e+00\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.6373 - categorical_accuracy: 0.6649 - recall_74: 0.0154 - val_loss: 0.6430 - val_categorical_accuracy: 0.6302 - val_recall_74: 0.0274\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.6283 - categorical_accuracy: 0.6632 - recall_74: 0.0154 - val_loss: 0.6358 - val_categorical_accuracy: 0.6302 - val_recall_74: 0.0274\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.6209 - categorical_accuracy: 0.6649 - recall_74: 0.0103 - val_loss: 0.6287 - val_categorical_accuracy: 0.6302 - val_recall_74: 0.0274\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.6105 - categorical_accuracy: 0.6667 - recall_74: 0.0410 - val_loss: 0.6136 - val_categorical_accuracy: 0.6615 - val_recall_74: 0.1096\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.6029 - categorical_accuracy: 0.6858 - recall_74: 0.1333 - val_loss: 0.6030 - val_categorical_accuracy: 0.6771 - val_recall_74: 0.1644\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 20ms/step - loss: 0.5920 - categorical_accuracy: 0.6823 - recall_74: 0.1179 - val_loss: 0.5956 - val_categorical_accuracy: 0.6771 - val_recall_74: 0.1644\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5853 - categorical_accuracy: 0.6823 - recall_74: 0.1128 - val_loss: 0.5850 - val_categorical_accuracy: 0.6979 - val_recall_74: 0.2192\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5744 - categorical_accuracy: 0.7049 - recall_74: 0.2103 - val_loss: 0.5717 - val_categorical_accuracy: 0.7292 - val_recall_74: 0.3562\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.5694 - categorical_accuracy: 0.7031 - recall_74: 0.3231 - val_loss: 0.5651 - val_categorical_accuracy: 0.7292 - val_recall_74: 0.3425\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5588 - categorical_accuracy: 0.7135 - recall_74: 0.3128 - val_loss: 0.5557 - val_categorical_accuracy: 0.7708 - val_recall_74: 0.4658\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 29ms/step - loss: 0.5507 - categorical_accuracy: 0.7205 - recall_74: 0.2974 - val_loss: 0.5489 - val_categorical_accuracy: 0.7760 - val_recall_74: 0.4658\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 23ms/step - loss: 0.5418 - categorical_accuracy: 0.7309 - recall_74: 0.3795 - val_loss: 0.5373 - val_categorical_accuracy: 0.7760 - val_recall_74: 0.5205\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5332 - categorical_accuracy: 0.7378 - recall_74: 0.3744 - val_loss: 0.5318 - val_categorical_accuracy: 0.7760 - val_recall_74: 0.4932\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.5252 - categorical_accuracy: 0.7552 - recall_74: 0.4462 - val_loss: 0.5262 - val_categorical_accuracy: 0.7708 - val_recall_74: 0.4932\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5172 - categorical_accuracy: 0.7552 - recall_74: 0.3949 - val_loss: 0.5178 - val_categorical_accuracy: 0.7656 - val_recall_74: 0.5068\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.5101 - categorical_accuracy: 0.7500 - recall_74: 0.4051 - val_loss: 0.5119 - val_categorical_accuracy: 0.7760 - val_recall_74: 0.5479\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.5042 - categorical_accuracy: 0.7691 - recall_74: 0.4872 - val_loss: 0.5065 - val_categorical_accuracy: 0.7812 - val_recall_74: 0.5616\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.4973 - categorical_accuracy: 0.7691 - recall_74: 0.5179 - val_loss: 0.5034 - val_categorical_accuracy: 0.7760 - val_recall_74: 0.5479\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.4906 - categorical_accuracy: 0.7760 - recall_74: 0.5026 - val_loss: 0.4999 - val_categorical_accuracy: 0.7812 - val_recall_74: 0.5616\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4882 - categorical_accuracy: 0.7726 - recall_74: 0.5231 - val_loss: 0.4950 - val_categorical_accuracy: 0.8021 - val_recall_74: 0.6301\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.4822 - categorical_accuracy: 0.7726 - recall_74: 0.5128 - val_loss: 0.4947 - val_categorical_accuracy: 0.8073 - val_recall_74: 0.6164\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4764 - categorical_accuracy: 0.7760 - recall_74: 0.5128 - val_loss: 0.4900 - val_categorical_accuracy: 0.8021 - val_recall_74: 0.6301\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4738 - categorical_accuracy: 0.7691 - recall_74: 0.4923 - val_loss: 0.4896 - val_categorical_accuracy: 0.7865 - val_recall_74: 0.5753\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4745 - categorical_accuracy: 0.7674 - recall_74: 0.5846 - val_loss: 0.4918 - val_categorical_accuracy: 0.7865 - val_recall_74: 0.5753\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 28ms/step - loss: 0.4681 - categorical_accuracy: 0.7778 - recall_74: 0.4872 - val_loss: 0.4880 - val_categorical_accuracy: 0.7917 - val_recall_74: 0.5890\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 24ms/step - loss: 0.4607 - categorical_accuracy: 0.7812 - recall_74: 0.5333 - val_loss: 0.4853 - val_categorical_accuracy: 0.7969 - val_recall_74: 0.6164\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4602 - categorical_accuracy: 0.7691 - recall_74: 0.5333 - val_loss: 0.4918 - val_categorical_accuracy: 0.7760 - val_recall_74: 0.5479\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4598 - categorical_accuracy: 0.7726 - recall_74: 0.5538 - val_loss: 0.4864 - val_categorical_accuracy: 0.7917 - val_recall_74: 0.5890\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4536 - categorical_accuracy: 0.7812 - recall_74: 0.5128 - val_loss: 0.4836 - val_categorical_accuracy: 0.7969 - val_recall_74: 0.6027\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4535 - categorical_accuracy: 0.7795 - recall_74: 0.5538 - val_loss: 0.4869 - val_categorical_accuracy: 0.7865 - val_recall_74: 0.5753\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4497 - categorical_accuracy: 0.7795 - recall_74: 0.6051 - val_loss: 0.4850 - val_categorical_accuracy: 0.7865 - val_recall_74: 0.5890\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4472 - categorical_accuracy: 0.7778 - recall_74: 0.5077 - val_loss: 0.4855 - val_categorical_accuracy: 0.7812 - val_recall_74: 0.5890\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4453 - categorical_accuracy: 0.7865 - recall_74: 0.5846 - val_loss: 0.4865 - val_categorical_accuracy: 0.7760 - val_recall_74: 0.5753\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4456 - categorical_accuracy: 0.7865 - recall_74: 0.5128 - val_loss: 0.4836 - val_categorical_accuracy: 0.7865 - val_recall_74: 0.6027\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4424 - categorical_accuracy: 0.7865 - recall_74: 0.5846 - val_loss: 0.4869 - val_categorical_accuracy: 0.7760 - val_recall_74: 0.5753\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.4399 - categorical_accuracy: 0.7865 - recall_74: 0.5846 - val_loss: 0.4897 - val_categorical_accuracy: 0.7760 - val_recall_74: 0.5753\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4414 - categorical_accuracy: 0.7882 - recall_74: 0.5077 - val_loss: 0.4862 - val_categorical_accuracy: 0.7812 - val_recall_74: 0.6027\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4464 - categorical_accuracy: 0.7795 - recall_74: 0.6513 - val_loss: 0.4984 - val_categorical_accuracy: 0.7812 - val_recall_74: 0.5616\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4403 - categorical_accuracy: 0.7830 - recall_74: 0.5026 - val_loss: 0.4881 - val_categorical_accuracy: 0.7760 - val_recall_74: 0.6575\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.4437 - categorical_accuracy: 0.7865 - recall_74: 0.6410 - val_loss: 0.4931 - val_categorical_accuracy: 0.7760 - val_recall_74: 0.5753\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 28ms/step - loss: 0.4356 - categorical_accuracy: 0.7917 - recall_74: 0.5795 - val_loss: 0.4889 - val_categorical_accuracy: 0.7760 - val_recall_74: 0.5890\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.4358 - categorical_accuracy: 0.7899 - recall_74: 0.5692 - val_loss: 0.4947 - val_categorical_accuracy: 0.7760 - val_recall_74: 0.5753\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4323 - categorical_accuracy: 0.7951 - recall_74: 0.5795 - val_loss: 0.4896 - val_categorical_accuracy: 0.7708 - val_recall_74: 0.6438\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4373 - categorical_accuracy: 0.7899 - recall_74: 0.5897 - val_loss: 0.4984 - val_categorical_accuracy: 0.7760 - val_recall_74: 0.5753\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4370 - categorical_accuracy: 0.7917 - recall_74: 0.6410 - val_loss: 0.4929 - val_categorical_accuracy: 0.7812 - val_recall_74: 0.5890\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4363 - categorical_accuracy: 0.7865 - recall_74: 0.5333 - val_loss: 0.4909 - val_categorical_accuracy: 0.7604 - val_recall_74: 0.5890\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4362 - categorical_accuracy: 0.8038 - recall_74: 0.6615 - val_loss: 0.4963 - val_categorical_accuracy: 0.7708 - val_recall_74: 0.5753\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4343 - categorical_accuracy: 0.7917 - recall_74: 0.6462 - val_loss: 0.4970 - val_categorical_accuracy: 0.7708 - val_recall_74: 0.5753\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 4s 100ms/step - loss: 0.6861 - categorical_accuracy: 0.6042 - recall_75: 0.0974 - val_loss: 0.6820 - val_categorical_accuracy: 0.6198 - val_recall_75: 0.0000e+00\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.6643 - categorical_accuracy: 0.6615 - recall_75: 0.0000e+00 - val_loss: 0.6748 - val_categorical_accuracy: 0.6198 - val_recall_75: 0.0000e+00\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.6517 - categorical_accuracy: 0.6615 - recall_75: 0.0000e+00 - val_loss: 0.6702 - val_categorical_accuracy: 0.6198 - val_recall_75: 0.0000e+00\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.6431 - categorical_accuracy: 0.6615 - recall_75: 0.0000e+00 - val_loss: 0.6648 - val_categorical_accuracy: 0.6198 - val_recall_75: 0.0000e+00\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 37ms/step - loss: 0.6337 - categorical_accuracy: 0.6615 - recall_75: 0.0000e+00 - val_loss: 0.6545 - val_categorical_accuracy: 0.6198 - val_recall_75: 0.0000e+00\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.6246 - categorical_accuracy: 0.6615 - recall_75: 0.0000e+00 - val_loss: 0.6460 - val_categorical_accuracy: 0.6198 - val_recall_75: 0.0000e+00\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.6148 - categorical_accuracy: 0.6615 - recall_75: 0.0000e+00 - val_loss: 0.6347 - val_categorical_accuracy: 0.6198 - val_recall_75: 0.0000e+00\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.6030 - categorical_accuracy: 0.6632 - recall_75: 0.0051 - val_loss: 0.6193 - val_categorical_accuracy: 0.6302 - val_recall_75: 0.0274\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 21ms/step - loss: 0.5879 - categorical_accuracy: 0.6719 - recall_75: 0.0359 - val_loss: 0.6082 - val_categorical_accuracy: 0.6354 - val_recall_75: 0.0548\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.5756 - categorical_accuracy: 0.6944 - recall_75: 0.1179 - val_loss: 0.5935 - val_categorical_accuracy: 0.6719 - val_recall_75: 0.1644\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 24ms/step - loss: 0.5644 - categorical_accuracy: 0.7049 - recall_75: 0.1846 - val_loss: 0.5843 - val_categorical_accuracy: 0.7031 - val_recall_75: 0.2603\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.5499 - categorical_accuracy: 0.7240 - recall_75: 0.3538 - val_loss: 0.5721 - val_categorical_accuracy: 0.7552 - val_recall_75: 0.4521\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 43ms/step - loss: 0.5366 - categorical_accuracy: 0.7326 - recall_75: 0.3641 - val_loss: 0.5628 - val_categorical_accuracy: 0.7448 - val_recall_75: 0.4384\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 26ms/step - loss: 0.5253 - categorical_accuracy: 0.7309 - recall_75: 0.3282 - val_loss: 0.5562 - val_categorical_accuracy: 0.7500 - val_recall_75: 0.4521\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.5141 - categorical_accuracy: 0.7604 - recall_75: 0.5333 - val_loss: 0.5468 - val_categorical_accuracy: 0.7500 - val_recall_75: 0.5342\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.5035 - categorical_accuracy: 0.7622 - recall_75: 0.4615 - val_loss: 0.5468 - val_categorical_accuracy: 0.7500 - val_recall_75: 0.4932\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 22ms/step - loss: 0.4949 - categorical_accuracy: 0.7708 - recall_75: 0.5026 - val_loss: 0.5389 - val_categorical_accuracy: 0.7552 - val_recall_75: 0.5479\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4860 - categorical_accuracy: 0.7760 - recall_75: 0.5179 - val_loss: 0.5367 - val_categorical_accuracy: 0.7604 - val_recall_75: 0.5616\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4820 - categorical_accuracy: 0.7743 - recall_75: 0.6000 - val_loss: 0.5394 - val_categorical_accuracy: 0.7448 - val_recall_75: 0.4795\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.4753 - categorical_accuracy: 0.7691 - recall_75: 0.4667 - val_loss: 0.5302 - val_categorical_accuracy: 0.7552 - val_recall_75: 0.5890\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4698 - categorical_accuracy: 0.7778 - recall_75: 0.5897 - val_loss: 0.5352 - val_categorical_accuracy: 0.7448 - val_recall_75: 0.5342\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4665 - categorical_accuracy: 0.7812 - recall_75: 0.5949 - val_loss: 0.5339 - val_categorical_accuracy: 0.7396 - val_recall_75: 0.5479\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 21ms/step - loss: 0.4595 - categorical_accuracy: 0.7847 - recall_75: 0.5333 - val_loss: 0.5315 - val_categorical_accuracy: 0.7448 - val_recall_75: 0.5753\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4600 - categorical_accuracy: 0.7778 - recall_75: 0.5128 - val_loss: 0.5300 - val_categorical_accuracy: 0.7604 - val_recall_75: 0.6438\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4645 - categorical_accuracy: 0.7986 - recall_75: 0.6974 - val_loss: 0.5410 - val_categorical_accuracy: 0.7500 - val_recall_75: 0.4932\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.4541 - categorical_accuracy: 0.7812 - recall_75: 0.5487 - val_loss: 0.5310 - val_categorical_accuracy: 0.7656 - val_recall_75: 0.6164\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4529 - categorical_accuracy: 0.7830 - recall_75: 0.5333 - val_loss: 0.5305 - val_categorical_accuracy: 0.7708 - val_recall_75: 0.6301\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4511 - categorical_accuracy: 0.7847 - recall_75: 0.6615 - val_loss: 0.5379 - val_categorical_accuracy: 0.7656 - val_recall_75: 0.5479\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.4425 - categorical_accuracy: 0.7847 - recall_75: 0.5590 - val_loss: 0.5328 - val_categorical_accuracy: 0.7656 - val_recall_75: 0.6164\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4421 - categorical_accuracy: 0.8108 - recall_75: 0.6872 - val_loss: 0.5454 - val_categorical_accuracy: 0.7344 - val_recall_75: 0.4795\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4423 - categorical_accuracy: 0.7882 - recall_75: 0.5744 - val_loss: 0.5346 - val_categorical_accuracy: 0.7708 - val_recall_75: 0.6027\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4399 - categorical_accuracy: 0.7951 - recall_75: 0.5897 - val_loss: 0.5332 - val_categorical_accuracy: 0.7604 - val_recall_75: 0.6164\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4348 - categorical_accuracy: 0.7969 - recall_75: 0.6513 - val_loss: 0.5358 - val_categorical_accuracy: 0.7708 - val_recall_75: 0.6164\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4339 - categorical_accuracy: 0.7899 - recall_75: 0.5949 - val_loss: 0.5352 - val_categorical_accuracy: 0.7708 - val_recall_75: 0.5890\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4342 - categorical_accuracy: 0.7986 - recall_75: 0.6205 - val_loss: 0.5386 - val_categorical_accuracy: 0.7812 - val_recall_75: 0.6027\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4324 - categorical_accuracy: 0.7986 - recall_75: 0.6410 - val_loss: 0.5378 - val_categorical_accuracy: 0.7708 - val_recall_75: 0.6027\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4299 - categorical_accuracy: 0.8003 - recall_75: 0.6103 - val_loss: 0.5350 - val_categorical_accuracy: 0.7656 - val_recall_75: 0.6164\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4312 - categorical_accuracy: 0.8090 - recall_75: 0.6769 - val_loss: 0.5432 - val_categorical_accuracy: 0.7760 - val_recall_75: 0.5890\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4277 - categorical_accuracy: 0.8038 - recall_75: 0.6359 - val_loss: 0.5381 - val_categorical_accuracy: 0.7708 - val_recall_75: 0.6027\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 20ms/step - loss: 0.4254 - categorical_accuracy: 0.7969 - recall_75: 0.6205 - val_loss: 0.5405 - val_categorical_accuracy: 0.7760 - val_recall_75: 0.6027\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4272 - categorical_accuracy: 0.8038 - recall_75: 0.6359 - val_loss: 0.5390 - val_categorical_accuracy: 0.7760 - val_recall_75: 0.6301\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4258 - categorical_accuracy: 0.8090 - recall_75: 0.6872 - val_loss: 0.5391 - val_categorical_accuracy: 0.7812 - val_recall_75: 0.6301\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4271 - categorical_accuracy: 0.7847 - recall_75: 0.5590 - val_loss: 0.5373 - val_categorical_accuracy: 0.7656 - val_recall_75: 0.6301\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4271 - categorical_accuracy: 0.8038 - recall_75: 0.6667 - val_loss: 0.5449 - val_categorical_accuracy: 0.7708 - val_recall_75: 0.5890\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.4256 - categorical_accuracy: 0.8021 - recall_75: 0.6718 - val_loss: 0.5449 - val_categorical_accuracy: 0.7812 - val_recall_75: 0.6164\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4230 - categorical_accuracy: 0.8038 - recall_75: 0.6205 - val_loss: 0.5425 - val_categorical_accuracy: 0.7760 - val_recall_75: 0.6164\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4191 - categorical_accuracy: 0.8003 - recall_75: 0.6308 - val_loss: 0.5423 - val_categorical_accuracy: 0.7812 - val_recall_75: 0.6301\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4198 - categorical_accuracy: 0.8021 - recall_75: 0.6410 - val_loss: 0.5432 - val_categorical_accuracy: 0.7812 - val_recall_75: 0.6301\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4184 - categorical_accuracy: 0.8038 - recall_75: 0.6410 - val_loss: 0.5493 - val_categorical_accuracy: 0.7708 - val_recall_75: 0.5890\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4212 - categorical_accuracy: 0.8090 - recall_75: 0.6564 - val_loss: 0.5427 - val_categorical_accuracy: 0.7812 - val_recall_75: 0.6301\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 3s 66ms/step - loss: 0.6616 - categorical_accuracy: 0.6615 - recall_76: 0.0000e+00 - val_loss: 0.6696 - val_categorical_accuracy: 0.6198 - val_recall_76: 0.0000e+00\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.6460 - categorical_accuracy: 0.6615 - recall_76: 0.0000e+00 - val_loss: 0.6663 - val_categorical_accuracy: 0.6198 - val_recall_76: 0.0000e+00\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.6397 - categorical_accuracy: 0.6615 - recall_76: 0.0000e+00 - val_loss: 0.6573 - val_categorical_accuracy: 0.6198 - val_recall_76: 0.0000e+00\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.6322 - categorical_accuracy: 0.6615 - recall_76: 0.0000e+00 - val_loss: 0.6456 - val_categorical_accuracy: 0.6198 - val_recall_76: 0.0000e+00\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6251 - categorical_accuracy: 0.6632 - recall_76: 0.0103 - val_loss: 0.6349 - val_categorical_accuracy: 0.6198 - val_recall_76: 0.0137\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6162 - categorical_accuracy: 0.6632 - recall_76: 0.0256 - val_loss: 0.6271 - val_categorical_accuracy: 0.6250 - val_recall_76: 0.0274\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6074 - categorical_accuracy: 0.6649 - recall_76: 0.0564 - val_loss: 0.6153 - val_categorical_accuracy: 0.6406 - val_recall_76: 0.0959\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5974 - categorical_accuracy: 0.6719 - recall_76: 0.1077 - val_loss: 0.6001 - val_categorical_accuracy: 0.6979 - val_recall_76: 0.3014\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.5893 - categorical_accuracy: 0.6823 - recall_76: 0.2974 - val_loss: 0.5898 - val_categorical_accuracy: 0.7031 - val_recall_76: 0.3562\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5784 - categorical_accuracy: 0.6875 - recall_76: 0.2564 - val_loss: 0.5862 - val_categorical_accuracy: 0.6979 - val_recall_76: 0.3014\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5753 - categorical_accuracy: 0.6962 - recall_76: 0.2103 - val_loss: 0.5748 - val_categorical_accuracy: 0.7083 - val_recall_76: 0.3836\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.5661 - categorical_accuracy: 0.7101 - recall_76: 0.4205 - val_loss: 0.5656 - val_categorical_accuracy: 0.7240 - val_recall_76: 0.4658\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5595 - categorical_accuracy: 0.7014 - recall_76: 0.3333 - val_loss: 0.5611 - val_categorical_accuracy: 0.7135 - val_recall_76: 0.3973\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.5512 - categorical_accuracy: 0.7118 - recall_76: 0.3949 - val_loss: 0.5530 - val_categorical_accuracy: 0.7083 - val_recall_76: 0.4110\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.5461 - categorical_accuracy: 0.7101 - recall_76: 0.3795 - val_loss: 0.5474 - val_categorical_accuracy: 0.7188 - val_recall_76: 0.4384\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5398 - categorical_accuracy: 0.7153 - recall_76: 0.4205 - val_loss: 0.5449 - val_categorical_accuracy: 0.7135 - val_recall_76: 0.4110\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.5372 - categorical_accuracy: 0.7066 - recall_76: 0.4154 - val_loss: 0.5381 - val_categorical_accuracy: 0.7344 - val_recall_76: 0.4658\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5284 - categorical_accuracy: 0.7170 - recall_76: 0.4000 - val_loss: 0.5313 - val_categorical_accuracy: 0.7448 - val_recall_76: 0.5068\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5255 - categorical_accuracy: 0.7344 - recall_76: 0.4513 - val_loss: 0.5323 - val_categorical_accuracy: 0.7448 - val_recall_76: 0.4658\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5194 - categorical_accuracy: 0.7274 - recall_76: 0.4513 - val_loss: 0.5199 - val_categorical_accuracy: 0.7604 - val_recall_76: 0.5479\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5128 - categorical_accuracy: 0.7465 - recall_76: 0.4513 - val_loss: 0.5264 - val_categorical_accuracy: 0.7656 - val_recall_76: 0.4932\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 24ms/step - loss: 0.5077 - categorical_accuracy: 0.7517 - recall_76: 0.4872 - val_loss: 0.5114 - val_categorical_accuracy: 0.7708 - val_recall_76: 0.6027\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.5037 - categorical_accuracy: 0.7535 - recall_76: 0.5077 - val_loss: 0.5115 - val_categorical_accuracy: 0.7656 - val_recall_76: 0.5616\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5015 - categorical_accuracy: 0.7431 - recall_76: 0.5077 - val_loss: 0.5100 - val_categorical_accuracy: 0.7604 - val_recall_76: 0.5205\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.4957 - categorical_accuracy: 0.7569 - recall_76: 0.5026 - val_loss: 0.5015 - val_categorical_accuracy: 0.7708 - val_recall_76: 0.6027\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4913 - categorical_accuracy: 0.7569 - recall_76: 0.5436 - val_loss: 0.5040 - val_categorical_accuracy: 0.7604 - val_recall_76: 0.5479\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4878 - categorical_accuracy: 0.7483 - recall_76: 0.4974 - val_loss: 0.5000 - val_categorical_accuracy: 0.7708 - val_recall_76: 0.5753\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4840 - categorical_accuracy: 0.7569 - recall_76: 0.5333 - val_loss: 0.4947 - val_categorical_accuracy: 0.7760 - val_recall_76: 0.6027\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.4812 - categorical_accuracy: 0.7569 - recall_76: 0.5128 - val_loss: 0.4944 - val_categorical_accuracy: 0.7708 - val_recall_76: 0.5753\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4818 - categorical_accuracy: 0.7622 - recall_76: 0.5846 - val_loss: 0.4933 - val_categorical_accuracy: 0.7708 - val_recall_76: 0.5753\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4807 - categorical_accuracy: 0.7656 - recall_76: 0.4769 - val_loss: 0.4889 - val_categorical_accuracy: 0.7812 - val_recall_76: 0.6164\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.4759 - categorical_accuracy: 0.7708 - recall_76: 0.6308 - val_loss: 0.4925 - val_categorical_accuracy: 0.7760 - val_recall_76: 0.5890\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4713 - categorical_accuracy: 0.7517 - recall_76: 0.4821 - val_loss: 0.4876 - val_categorical_accuracy: 0.7812 - val_recall_76: 0.6164\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.4664 - categorical_accuracy: 0.7656 - recall_76: 0.5333 - val_loss: 0.4858 - val_categorical_accuracy: 0.7812 - val_recall_76: 0.6164\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4635 - categorical_accuracy: 0.7674 - recall_76: 0.5487 - val_loss: 0.4856 - val_categorical_accuracy: 0.7812 - val_recall_76: 0.6164\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.4610 - categorical_accuracy: 0.7726 - recall_76: 0.5590 - val_loss: 0.4885 - val_categorical_accuracy: 0.7865 - val_recall_76: 0.6027\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.4617 - categorical_accuracy: 0.7726 - recall_76: 0.5487 - val_loss: 0.4859 - val_categorical_accuracy: 0.7812 - val_recall_76: 0.6027\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4592 - categorical_accuracy: 0.7708 - recall_76: 0.5385 - val_loss: 0.4809 - val_categorical_accuracy: 0.7917 - val_recall_76: 0.6575\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4553 - categorical_accuracy: 0.7743 - recall_76: 0.6154 - val_loss: 0.4853 - val_categorical_accuracy: 0.7865 - val_recall_76: 0.6164\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4539 - categorical_accuracy: 0.7778 - recall_76: 0.5436 - val_loss: 0.4806 - val_categorical_accuracy: 0.7917 - val_recall_76: 0.6575\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4543 - categorical_accuracy: 0.7830 - recall_76: 0.5897 - val_loss: 0.4840 - val_categorical_accuracy: 0.7969 - val_recall_76: 0.6301\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4539 - categorical_accuracy: 0.7830 - recall_76: 0.6154 - val_loss: 0.4853 - val_categorical_accuracy: 0.7969 - val_recall_76: 0.6301\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4485 - categorical_accuracy: 0.7847 - recall_76: 0.5897 - val_loss: 0.4831 - val_categorical_accuracy: 0.7917 - val_recall_76: 0.6301\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4463 - categorical_accuracy: 0.7760 - recall_76: 0.5538 - val_loss: 0.4821 - val_categorical_accuracy: 0.7865 - val_recall_76: 0.6164\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4472 - categorical_accuracy: 0.7917 - recall_76: 0.6615 - val_loss: 0.4899 - val_categorical_accuracy: 0.7917 - val_recall_76: 0.5890\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4587 - categorical_accuracy: 0.7691 - recall_76: 0.4872 - val_loss: 0.4820 - val_categorical_accuracy: 0.7969 - val_recall_76: 0.7123\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.4543 - categorical_accuracy: 0.7951 - recall_76: 0.7026 - val_loss: 0.4940 - val_categorical_accuracy: 0.7917 - val_recall_76: 0.5890\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4440 - categorical_accuracy: 0.7847 - recall_76: 0.5744 - val_loss: 0.4821 - val_categorical_accuracy: 0.7917 - val_recall_76: 0.6438\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.4444 - categorical_accuracy: 0.7847 - recall_76: 0.5795 - val_loss: 0.4886 - val_categorical_accuracy: 0.8021 - val_recall_76: 0.6164\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4389 - categorical_accuracy: 0.7865 - recall_76: 0.6154 - val_loss: 0.4835 - val_categorical_accuracy: 0.7917 - val_recall_76: 0.6438\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 3s 59ms/step - loss: 0.6710 - categorical_accuracy: 0.6649 - recall_77: 0.0256 - val_loss: 0.6768 - val_categorical_accuracy: 0.6198 - val_recall_77: 0.0000e+00\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6532 - categorical_accuracy: 0.6615 - recall_77: 0.0000e+00 - val_loss: 0.6696 - val_categorical_accuracy: 0.6198 - val_recall_77: 0.0000e+00\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6442 - categorical_accuracy: 0.6615 - recall_77: 0.0000e+00 - val_loss: 0.6602 - val_categorical_accuracy: 0.6198 - val_recall_77: 0.0000e+00\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.6371 - categorical_accuracy: 0.6615 - recall_77: 0.0000e+00 - val_loss: 0.6505 - val_categorical_accuracy: 0.6250 - val_recall_77: 0.0137\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6303 - categorical_accuracy: 0.6597 - recall_77: 0.0000e+00 - val_loss: 0.6393 - val_categorical_accuracy: 0.6302 - val_recall_77: 0.0274\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6214 - categorical_accuracy: 0.6649 - recall_77: 0.0154 - val_loss: 0.6322 - val_categorical_accuracy: 0.6302 - val_recall_77: 0.0274\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6131 - categorical_accuracy: 0.6684 - recall_77: 0.0256 - val_loss: 0.6218 - val_categorical_accuracy: 0.6354 - val_recall_77: 0.0411\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6052 - categorical_accuracy: 0.6840 - recall_77: 0.1077 - val_loss: 0.6059 - val_categorical_accuracy: 0.6719 - val_recall_77: 0.1644\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.5930 - categorical_accuracy: 0.6892 - recall_77: 0.1231 - val_loss: 0.6016 - val_categorical_accuracy: 0.6719 - val_recall_77: 0.1507\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5856 - categorical_accuracy: 0.7014 - recall_77: 0.1846 - val_loss: 0.5846 - val_categorical_accuracy: 0.6979 - val_recall_77: 0.2603\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5752 - categorical_accuracy: 0.7118 - recall_77: 0.2718 - val_loss: 0.5783 - val_categorical_accuracy: 0.7083 - val_recall_77: 0.2603\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.5671 - categorical_accuracy: 0.7083 - recall_77: 0.1949 - val_loss: 0.5739 - val_categorical_accuracy: 0.7083 - val_recall_77: 0.2603\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5594 - categorical_accuracy: 0.7170 - recall_77: 0.2718 - val_loss: 0.5596 - val_categorical_accuracy: 0.7135 - val_recall_77: 0.3288\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.5519 - categorical_accuracy: 0.7257 - recall_77: 0.2923 - val_loss: 0.5537 - val_categorical_accuracy: 0.7135 - val_recall_77: 0.3288\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5443 - categorical_accuracy: 0.7309 - recall_77: 0.3590 - val_loss: 0.5452 - val_categorical_accuracy: 0.7292 - val_recall_77: 0.3973\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5358 - categorical_accuracy: 0.7413 - recall_77: 0.3692 - val_loss: 0.5386 - val_categorical_accuracy: 0.7344 - val_recall_77: 0.4247\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.5294 - categorical_accuracy: 0.7309 - recall_77: 0.3795 - val_loss: 0.5320 - val_categorical_accuracy: 0.7396 - val_recall_77: 0.4384\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5226 - categorical_accuracy: 0.7361 - recall_77: 0.3897 - val_loss: 0.5309 - val_categorical_accuracy: 0.7500 - val_recall_77: 0.4384\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.5194 - categorical_accuracy: 0.7431 - recall_77: 0.3744 - val_loss: 0.5209 - val_categorical_accuracy: 0.7396 - val_recall_77: 0.4795\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5115 - categorical_accuracy: 0.7500 - recall_77: 0.4256 - val_loss: 0.5153 - val_categorical_accuracy: 0.7552 - val_recall_77: 0.5205\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5034 - categorical_accuracy: 0.7500 - recall_77: 0.4359 - val_loss: 0.5121 - val_categorical_accuracy: 0.7604 - val_recall_77: 0.5205\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4978 - categorical_accuracy: 0.7656 - recall_77: 0.4923 - val_loss: 0.5097 - val_categorical_accuracy: 0.7552 - val_recall_77: 0.5068\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.4931 - categorical_accuracy: 0.7622 - recall_77: 0.4615 - val_loss: 0.5053 - val_categorical_accuracy: 0.7604 - val_recall_77: 0.5342\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4889 - categorical_accuracy: 0.7656 - recall_77: 0.4974 - val_loss: 0.5026 - val_categorical_accuracy: 0.7656 - val_recall_77: 0.5616\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4840 - categorical_accuracy: 0.7691 - recall_77: 0.4821 - val_loss: 0.5032 - val_categorical_accuracy: 0.7552 - val_recall_77: 0.4932\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4796 - categorical_accuracy: 0.7795 - recall_77: 0.5077 - val_loss: 0.4994 - val_categorical_accuracy: 0.7708 - val_recall_77: 0.5342\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4761 - categorical_accuracy: 0.7691 - recall_77: 0.4872 - val_loss: 0.4942 - val_categorical_accuracy: 0.7917 - val_recall_77: 0.6438\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.4726 - categorical_accuracy: 0.7708 - recall_77: 0.5436 - val_loss: 0.4986 - val_categorical_accuracy: 0.7656 - val_recall_77: 0.5068\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4701 - categorical_accuracy: 0.7743 - recall_77: 0.5128 - val_loss: 0.4954 - val_categorical_accuracy: 0.7760 - val_recall_77: 0.5479\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4665 - categorical_accuracy: 0.7743 - recall_77: 0.5231 - val_loss: 0.4933 - val_categorical_accuracy: 0.7656 - val_recall_77: 0.5342\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4632 - categorical_accuracy: 0.7795 - recall_77: 0.5128 - val_loss: 0.4968 - val_categorical_accuracy: 0.7708 - val_recall_77: 0.5205\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4600 - categorical_accuracy: 0.7865 - recall_77: 0.5282 - val_loss: 0.4899 - val_categorical_accuracy: 0.7917 - val_recall_77: 0.6438\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4612 - categorical_accuracy: 0.7760 - recall_77: 0.5641 - val_loss: 0.4986 - val_categorical_accuracy: 0.7708 - val_recall_77: 0.5205\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.4572 - categorical_accuracy: 0.7795 - recall_77: 0.5179 - val_loss: 0.4905 - val_categorical_accuracy: 0.7917 - val_recall_77: 0.6164\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4582 - categorical_accuracy: 0.7812 - recall_77: 0.6103 - val_loss: 0.4979 - val_categorical_accuracy: 0.7656 - val_recall_77: 0.5205\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4574 - categorical_accuracy: 0.7830 - recall_77: 0.4974 - val_loss: 0.4904 - val_categorical_accuracy: 0.7760 - val_recall_77: 0.6301\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4554 - categorical_accuracy: 0.7917 - recall_77: 0.6205 - val_loss: 0.5009 - val_categorical_accuracy: 0.7656 - val_recall_77: 0.5205\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4518 - categorical_accuracy: 0.7830 - recall_77: 0.4974 - val_loss: 0.4935 - val_categorical_accuracy: 0.7812 - val_recall_77: 0.6164\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.4579 - categorical_accuracy: 0.7882 - recall_77: 0.6615 - val_loss: 0.5044 - val_categorical_accuracy: 0.7656 - val_recall_77: 0.5205\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4543 - categorical_accuracy: 0.7812 - recall_77: 0.5128 - val_loss: 0.4943 - val_categorical_accuracy: 0.7760 - val_recall_77: 0.6301\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.4491 - categorical_accuracy: 0.7847 - recall_77: 0.5744 - val_loss: 0.4975 - val_categorical_accuracy: 0.7760 - val_recall_77: 0.6027\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4451 - categorical_accuracy: 0.7812 - recall_77: 0.5487 - val_loss: 0.4980 - val_categorical_accuracy: 0.7760 - val_recall_77: 0.5890\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4450 - categorical_accuracy: 0.7847 - recall_77: 0.5692 - val_loss: 0.4957 - val_categorical_accuracy: 0.7812 - val_recall_77: 0.6301\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 24ms/step - loss: 0.4448 - categorical_accuracy: 0.7865 - recall_77: 0.5641 - val_loss: 0.4969 - val_categorical_accuracy: 0.7760 - val_recall_77: 0.6164\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.4423 - categorical_accuracy: 0.7899 - recall_77: 0.5744 - val_loss: 0.4988 - val_categorical_accuracy: 0.7760 - val_recall_77: 0.6164\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4407 - categorical_accuracy: 0.7865 - recall_77: 0.5744 - val_loss: 0.5011 - val_categorical_accuracy: 0.7708 - val_recall_77: 0.5753\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.4412 - categorical_accuracy: 0.7951 - recall_77: 0.6000 - val_loss: 0.5006 - val_categorical_accuracy: 0.7760 - val_recall_77: 0.6027\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.4408 - categorical_accuracy: 0.7830 - recall_77: 0.5436 - val_loss: 0.5003 - val_categorical_accuracy: 0.7865 - val_recall_77: 0.6301\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4445 - categorical_accuracy: 0.7778 - recall_77: 0.5487 - val_loss: 0.5021 - val_categorical_accuracy: 0.7865 - val_recall_77: 0.6164\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4459 - categorical_accuracy: 0.7882 - recall_77: 0.6308 - val_loss: 0.5121 - val_categorical_accuracy: 0.7656 - val_recall_77: 0.5479\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 3s 79ms/step - loss: 0.6688 - categorical_accuracy: 0.6615 - recall_78: 0.0000e+00 - val_loss: 0.6759 - val_categorical_accuracy: 0.6198 - val_recall_78: 0.0000e+00\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.6554 - categorical_accuracy: 0.6615 - recall_78: 0.0000e+00 - val_loss: 0.6736 - val_categorical_accuracy: 0.6198 - val_recall_78: 0.0000e+00\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6487 - categorical_accuracy: 0.6615 - recall_78: 0.0000e+00 - val_loss: 0.6707 - val_categorical_accuracy: 0.6198 - val_recall_78: 0.0000e+00\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6443 - categorical_accuracy: 0.6615 - recall_78: 0.0000e+00 - val_loss: 0.6666 - val_categorical_accuracy: 0.6198 - val_recall_78: 0.0000e+00\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6401 - categorical_accuracy: 0.6615 - recall_78: 0.0000e+00 - val_loss: 0.6627 - val_categorical_accuracy: 0.6198 - val_recall_78: 0.0000e+00\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6357 - categorical_accuracy: 0.6615 - recall_78: 0.0000e+00 - val_loss: 0.6564 - val_categorical_accuracy: 0.6198 - val_recall_78: 0.0000e+00\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.6282 - categorical_accuracy: 0.6615 - recall_78: 0.0000e+00 - val_loss: 0.6509 - val_categorical_accuracy: 0.6198 - val_recall_78: 0.0000e+00\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.6191 - categorical_accuracy: 0.6615 - recall_78: 0.0000e+00 - val_loss: 0.6395 - val_categorical_accuracy: 0.6198 - val_recall_78: 0.0000e+00\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.6098 - categorical_accuracy: 0.6615 - recall_78: 0.0000e+00 - val_loss: 0.6313 - val_categorical_accuracy: 0.6198 - val_recall_78: 0.0000e+00\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.6024 - categorical_accuracy: 0.6615 - recall_78: 0.0000e+00 - val_loss: 0.6244 - val_categorical_accuracy: 0.6198 - val_recall_78: 0.0000e+00\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5961 - categorical_accuracy: 0.6615 - recall_78: 0.0000e+00 - val_loss: 0.6206 - val_categorical_accuracy: 0.6198 - val_recall_78: 0.0000e+00\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 22ms/step - loss: 0.5861 - categorical_accuracy: 0.6615 - recall_78: 0.0000e+00 - val_loss: 0.6122 - val_categorical_accuracy: 0.6198 - val_recall_78: 0.0000e+00\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.5794 - categorical_accuracy: 0.6615 - recall_78: 0.0000e+00 - val_loss: 0.6087 - val_categorical_accuracy: 0.6250 - val_recall_78: 0.0137\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.5721 - categorical_accuracy: 0.6632 - recall_78: 0.0051 - val_loss: 0.6022 - val_categorical_accuracy: 0.6250 - val_recall_78: 0.0137\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5630 - categorical_accuracy: 0.6788 - recall_78: 0.0615 - val_loss: 0.5958 - val_categorical_accuracy: 0.6458 - val_recall_78: 0.0685\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5545 - categorical_accuracy: 0.6927 - recall_78: 0.1077 - val_loss: 0.5888 - val_categorical_accuracy: 0.6615 - val_recall_78: 0.1644\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5459 - categorical_accuracy: 0.7066 - recall_78: 0.1641 - val_loss: 0.5832 - val_categorical_accuracy: 0.6719 - val_recall_78: 0.2329\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5377 - categorical_accuracy: 0.7205 - recall_78: 0.2359 - val_loss: 0.5775 - val_categorical_accuracy: 0.6719 - val_recall_78: 0.2329\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5301 - categorical_accuracy: 0.7431 - recall_78: 0.3179 - val_loss: 0.5725 - val_categorical_accuracy: 0.6719 - val_recall_78: 0.2603\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5233 - categorical_accuracy: 0.7378 - recall_78: 0.2821 - val_loss: 0.5659 - val_categorical_accuracy: 0.6771 - val_recall_78: 0.3014\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5163 - categorical_accuracy: 0.7656 - recall_78: 0.4000 - val_loss: 0.5619 - val_categorical_accuracy: 0.6823 - val_recall_78: 0.3288\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.5092 - categorical_accuracy: 0.7639 - recall_78: 0.3949 - val_loss: 0.5571 - val_categorical_accuracy: 0.6927 - val_recall_78: 0.3425\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5042 - categorical_accuracy: 0.7726 - recall_78: 0.4615 - val_loss: 0.5513 - val_categorical_accuracy: 0.7135 - val_recall_78: 0.4110\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.4980 - categorical_accuracy: 0.7743 - recall_78: 0.4513 - val_loss: 0.5478 - val_categorical_accuracy: 0.7240 - val_recall_78: 0.4384\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4942 - categorical_accuracy: 0.7691 - recall_78: 0.4359 - val_loss: 0.5450 - val_categorical_accuracy: 0.7292 - val_recall_78: 0.4521\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4886 - categorical_accuracy: 0.7760 - recall_78: 0.4923 - val_loss: 0.5453 - val_categorical_accuracy: 0.7292 - val_recall_78: 0.4521\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4888 - categorical_accuracy: 0.7743 - recall_78: 0.5333 - val_loss: 0.5422 - val_categorical_accuracy: 0.7292 - val_recall_78: 0.4658\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.4817 - categorical_accuracy: 0.7726 - recall_78: 0.4308 - val_loss: 0.5392 - val_categorical_accuracy: 0.7344 - val_recall_78: 0.4932\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4770 - categorical_accuracy: 0.7743 - recall_78: 0.5436 - val_loss: 0.5339 - val_categorical_accuracy: 0.7552 - val_recall_78: 0.5616\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4731 - categorical_accuracy: 0.7691 - recall_78: 0.4974 - val_loss: 0.5370 - val_categorical_accuracy: 0.7396 - val_recall_78: 0.5205\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4693 - categorical_accuracy: 0.7812 - recall_78: 0.5077 - val_loss: 0.5364 - val_categorical_accuracy: 0.7396 - val_recall_78: 0.5205\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4668 - categorical_accuracy: 0.7778 - recall_78: 0.5590 - val_loss: 0.5329 - val_categorical_accuracy: 0.7396 - val_recall_78: 0.5342\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4707 - categorical_accuracy: 0.7708 - recall_78: 0.4769 - val_loss: 0.5316 - val_categorical_accuracy: 0.7500 - val_recall_78: 0.5753\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.4634 - categorical_accuracy: 0.7795 - recall_78: 0.6205 - val_loss: 0.5349 - val_categorical_accuracy: 0.7396 - val_recall_78: 0.5205\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.4597 - categorical_accuracy: 0.7812 - recall_78: 0.5128 - val_loss: 0.5410 - val_categorical_accuracy: 0.7604 - val_recall_78: 0.5205\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.4580 - categorical_accuracy: 0.7865 - recall_78: 0.5128 - val_loss: 0.5308 - val_categorical_accuracy: 0.7448 - val_recall_78: 0.5753\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.4589 - categorical_accuracy: 0.7760 - recall_78: 0.5949 - val_loss: 0.5365 - val_categorical_accuracy: 0.7448 - val_recall_78: 0.5342\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 22ms/step - loss: 0.4537 - categorical_accuracy: 0.7847 - recall_78: 0.5282 - val_loss: 0.5309 - val_categorical_accuracy: 0.7552 - val_recall_78: 0.6027\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4546 - categorical_accuracy: 0.7812 - recall_78: 0.6513 - val_loss: 0.5340 - val_categorical_accuracy: 0.7396 - val_recall_78: 0.5479\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.4518 - categorical_accuracy: 0.7847 - recall_78: 0.5282 - val_loss: 0.5348 - val_categorical_accuracy: 0.7500 - val_recall_78: 0.5616\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4506 - categorical_accuracy: 0.7882 - recall_78: 0.6256 - val_loss: 0.5344 - val_categorical_accuracy: 0.7500 - val_recall_78: 0.5753\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4474 - categorical_accuracy: 0.7865 - recall_78: 0.5333 - val_loss: 0.5368 - val_categorical_accuracy: 0.7552 - val_recall_78: 0.5616\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4462 - categorical_accuracy: 0.7778 - recall_78: 0.5846 - val_loss: 0.5349 - val_categorical_accuracy: 0.7396 - val_recall_78: 0.5753\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4447 - categorical_accuracy: 0.7847 - recall_78: 0.6051 - val_loss: 0.5377 - val_categorical_accuracy: 0.7552 - val_recall_78: 0.5616\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4457 - categorical_accuracy: 0.7847 - recall_78: 0.5590 - val_loss: 0.5346 - val_categorical_accuracy: 0.7552 - val_recall_78: 0.6164\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4436 - categorical_accuracy: 0.7830 - recall_78: 0.5795 - val_loss: 0.5376 - val_categorical_accuracy: 0.7604 - val_recall_78: 0.5753\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4406 - categorical_accuracy: 0.7847 - recall_78: 0.5641 - val_loss: 0.5354 - val_categorical_accuracy: 0.7500 - val_recall_78: 0.6027\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 21ms/step - loss: 0.4430 - categorical_accuracy: 0.7795 - recall_78: 0.5897 - val_loss: 0.5376 - val_categorical_accuracy: 0.7656 - val_recall_78: 0.5890\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4386 - categorical_accuracy: 0.7830 - recall_78: 0.5692 - val_loss: 0.5341 - val_categorical_accuracy: 0.7604 - val_recall_78: 0.6301\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.4411 - categorical_accuracy: 0.7865 - recall_78: 0.6615 - val_loss: 0.5412 - val_categorical_accuracy: 0.7656 - val_recall_78: 0.5753\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 3s 59ms/step - loss: 0.7023 - categorical_accuracy: 0.3906 - recall_79: 0.5949 - val_loss: 0.6965 - val_categorical_accuracy: 0.5729 - val_recall_79: 0.0548\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6842 - categorical_accuracy: 0.6545 - recall_79: 0.0154 - val_loss: 0.6859 - val_categorical_accuracy: 0.6198 - val_recall_79: 0.0000e+00\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6698 - categorical_accuracy: 0.6615 - recall_79: 0.0000e+00 - val_loss: 0.6794 - val_categorical_accuracy: 0.6198 - val_recall_79: 0.0000e+00\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6595 - categorical_accuracy: 0.6615 - recall_79: 0.0000e+00 - val_loss: 0.6751 - val_categorical_accuracy: 0.6198 - val_recall_79: 0.0000e+00\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.6540 - categorical_accuracy: 0.6615 - recall_79: 0.0000e+00 - val_loss: 0.6718 - val_categorical_accuracy: 0.6198 - val_recall_79: 0.0000e+00\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6489 - categorical_accuracy: 0.6615 - recall_79: 0.0000e+00 - val_loss: 0.6666 - val_categorical_accuracy: 0.6198 - val_recall_79: 0.0000e+00\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6440 - categorical_accuracy: 0.6615 - recall_79: 0.0000e+00 - val_loss: 0.6622 - val_categorical_accuracy: 0.6198 - val_recall_79: 0.0000e+00\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6394 - categorical_accuracy: 0.6615 - recall_79: 0.0000e+00 - val_loss: 0.6582 - val_categorical_accuracy: 0.6198 - val_recall_79: 0.0000e+00\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.6329 - categorical_accuracy: 0.6615 - recall_79: 0.0000e+00 - val_loss: 0.6499 - val_categorical_accuracy: 0.6198 - val_recall_79: 0.0000e+00\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 25ms/step - loss: 0.6267 - categorical_accuracy: 0.6615 - recall_79: 0.0000e+00 - val_loss: 0.6397 - val_categorical_accuracy: 0.6198 - val_recall_79: 0.0000e+00\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.6167 - categorical_accuracy: 0.6632 - recall_79: 0.0051 - val_loss: 0.6295 - val_categorical_accuracy: 0.6250 - val_recall_79: 0.0137\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6051 - categorical_accuracy: 0.6684 - recall_79: 0.0205 - val_loss: 0.6168 - val_categorical_accuracy: 0.6302 - val_recall_79: 0.0411\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.5950 - categorical_accuracy: 0.6823 - recall_79: 0.0923 - val_loss: 0.6033 - val_categorical_accuracy: 0.6562 - val_recall_79: 0.1370\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5840 - categorical_accuracy: 0.6927 - recall_79: 0.1282 - val_loss: 0.5972 - val_categorical_accuracy: 0.6562 - val_recall_79: 0.1370\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5773 - categorical_accuracy: 0.6910 - recall_79: 0.1282 - val_loss: 0.5838 - val_categorical_accuracy: 0.6875 - val_recall_79: 0.2466\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.5654 - categorical_accuracy: 0.7257 - recall_79: 0.2821 - val_loss: 0.5734 - val_categorical_accuracy: 0.7083 - val_recall_79: 0.3562\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5581 - categorical_accuracy: 0.7188 - recall_79: 0.3128 - val_loss: 0.5661 - val_categorical_accuracy: 0.7135 - val_recall_79: 0.3699\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5489 - categorical_accuracy: 0.7135 - recall_79: 0.2923 - val_loss: 0.5612 - val_categorical_accuracy: 0.7188 - val_recall_79: 0.3699\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5417 - categorical_accuracy: 0.7101 - recall_79: 0.2974 - val_loss: 0.5501 - val_categorical_accuracy: 0.7292 - val_recall_79: 0.4110\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.5317 - categorical_accuracy: 0.7257 - recall_79: 0.3795 - val_loss: 0.5452 - val_categorical_accuracy: 0.7240 - val_recall_79: 0.3836\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5257 - categorical_accuracy: 0.7292 - recall_79: 0.3897 - val_loss: 0.5381 - val_categorical_accuracy: 0.7396 - val_recall_79: 0.4521\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5161 - categorical_accuracy: 0.7396 - recall_79: 0.4051 - val_loss: 0.5371 - val_categorical_accuracy: 0.7240 - val_recall_79: 0.3836\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 27ms/step - loss: 0.5133 - categorical_accuracy: 0.7378 - recall_79: 0.3795 - val_loss: 0.5214 - val_categorical_accuracy: 0.7917 - val_recall_79: 0.6301\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5030 - categorical_accuracy: 0.7639 - recall_79: 0.5487 - val_loss: 0.5243 - val_categorical_accuracy: 0.7344 - val_recall_79: 0.4247\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5051 - categorical_accuracy: 0.7483 - recall_79: 0.3538 - val_loss: 0.5135 - val_categorical_accuracy: 0.7708 - val_recall_79: 0.5616\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4925 - categorical_accuracy: 0.7760 - recall_79: 0.5641 - val_loss: 0.5127 - val_categorical_accuracy: 0.7656 - val_recall_79: 0.5479\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.4862 - categorical_accuracy: 0.7708 - recall_79: 0.4974 - val_loss: 0.5072 - val_categorical_accuracy: 0.7760 - val_recall_79: 0.5753\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.4830 - categorical_accuracy: 0.7708 - recall_79: 0.5385 - val_loss: 0.5040 - val_categorical_accuracy: 0.7760 - val_recall_79: 0.6027\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4769 - categorical_accuracy: 0.7812 - recall_79: 0.5692 - val_loss: 0.5064 - val_categorical_accuracy: 0.7708 - val_recall_79: 0.5205\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4723 - categorical_accuracy: 0.7795 - recall_79: 0.5436 - val_loss: 0.5008 - val_categorical_accuracy: 0.7812 - val_recall_79: 0.6027\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4722 - categorical_accuracy: 0.7760 - recall_79: 0.5692 - val_loss: 0.5046 - val_categorical_accuracy: 0.7865 - val_recall_79: 0.5753\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4687 - categorical_accuracy: 0.7656 - recall_79: 0.5385 - val_loss: 0.4972 - val_categorical_accuracy: 0.7865 - val_recall_79: 0.6438\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4633 - categorical_accuracy: 0.7830 - recall_79: 0.5692 - val_loss: 0.5012 - val_categorical_accuracy: 0.7917 - val_recall_79: 0.6027\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 21ms/step - loss: 0.4608 - categorical_accuracy: 0.7899 - recall_79: 0.5897 - val_loss: 0.4960 - val_categorical_accuracy: 0.8021 - val_recall_79: 0.6438\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4588 - categorical_accuracy: 0.7830 - recall_79: 0.5026 - val_loss: 0.4984 - val_categorical_accuracy: 0.7865 - val_recall_79: 0.6027\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.4540 - categorical_accuracy: 0.7830 - recall_79: 0.5897 - val_loss: 0.4964 - val_categorical_accuracy: 0.7969 - val_recall_79: 0.6301\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4509 - categorical_accuracy: 0.7760 - recall_79: 0.5333 - val_loss: 0.5008 - val_categorical_accuracy: 0.7865 - val_recall_79: 0.5890\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4495 - categorical_accuracy: 0.7865 - recall_79: 0.5436 - val_loss: 0.4966 - val_categorical_accuracy: 0.7917 - val_recall_79: 0.6438\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4494 - categorical_accuracy: 0.7847 - recall_79: 0.6513 - val_loss: 0.4990 - val_categorical_accuracy: 0.7917 - val_recall_79: 0.6164\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.4483 - categorical_accuracy: 0.7917 - recall_79: 0.5385 - val_loss: 0.4981 - val_categorical_accuracy: 0.7969 - val_recall_79: 0.6438\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4448 - categorical_accuracy: 0.7812 - recall_79: 0.5744 - val_loss: 0.5005 - val_categorical_accuracy: 0.7812 - val_recall_79: 0.6027\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4422 - categorical_accuracy: 0.7899 - recall_79: 0.5949 - val_loss: 0.4996 - val_categorical_accuracy: 0.7865 - val_recall_79: 0.6164\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4416 - categorical_accuracy: 0.7865 - recall_79: 0.5590 - val_loss: 0.5021 - val_categorical_accuracy: 0.7812 - val_recall_79: 0.6027\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.4410 - categorical_accuracy: 0.7795 - recall_79: 0.5641 - val_loss: 0.4987 - val_categorical_accuracy: 0.8073 - val_recall_79: 0.6712\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4371 - categorical_accuracy: 0.7951 - recall_79: 0.6154 - val_loss: 0.5000 - val_categorical_accuracy: 0.7865 - val_recall_79: 0.6164\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4352 - categorical_accuracy: 0.7882 - recall_79: 0.5744 - val_loss: 0.5016 - val_categorical_accuracy: 0.7969 - val_recall_79: 0.6438\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4356 - categorical_accuracy: 0.7899 - recall_79: 0.5949 - val_loss: 0.5045 - val_categorical_accuracy: 0.7812 - val_recall_79: 0.6027\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4331 - categorical_accuracy: 0.7882 - recall_79: 0.5744 - val_loss: 0.5000 - val_categorical_accuracy: 0.8125 - val_recall_79: 0.6849\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4325 - categorical_accuracy: 0.7969 - recall_79: 0.6154 - val_loss: 0.5057 - val_categorical_accuracy: 0.7865 - val_recall_79: 0.6164\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4322 - categorical_accuracy: 0.7934 - recall_79: 0.6103 - val_loss: 0.5058 - val_categorical_accuracy: 0.7812 - val_recall_79: 0.6027\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 3s 59ms/step - loss: 0.7111 - categorical_accuracy: 0.3385 - recall_80: 1.0000 - val_loss: 0.6921 - val_categorical_accuracy: 0.4010 - val_recall_80: 1.0000\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6915 - categorical_accuracy: 0.5538 - recall_80: 0.7026 - val_loss: 0.6816 - val_categorical_accuracy: 0.6771 - val_recall_80: 0.2877\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.6788 - categorical_accuracy: 0.6632 - recall_80: 0.0872 - val_loss: 0.6759 - val_categorical_accuracy: 0.6198 - val_recall_80: 0.0000e+00\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.6693 - categorical_accuracy: 0.6615 - recall_80: 0.0051 - val_loss: 0.6712 - val_categorical_accuracy: 0.6198 - val_recall_80: 0.0000e+00\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 23ms/step - loss: 0.6616 - categorical_accuracy: 0.6615 - recall_80: 0.0000e+00 - val_loss: 0.6667 - val_categorical_accuracy: 0.6198 - val_recall_80: 0.0000e+00\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6546 - categorical_accuracy: 0.6615 - recall_80: 0.0000e+00 - val_loss: 0.6629 - val_categorical_accuracy: 0.6198 - val_recall_80: 0.0000e+00\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6486 - categorical_accuracy: 0.6615 - recall_80: 0.0000e+00 - val_loss: 0.6592 - val_categorical_accuracy: 0.6198 - val_recall_80: 0.0000e+00\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6437 - categorical_accuracy: 0.6615 - recall_80: 0.0000e+00 - val_loss: 0.6548 - val_categorical_accuracy: 0.6198 - val_recall_80: 0.0000e+00\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.6375 - categorical_accuracy: 0.6615 - recall_80: 0.0000e+00 - val_loss: 0.6491 - val_categorical_accuracy: 0.6198 - val_recall_80: 0.0000e+00\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6321 - categorical_accuracy: 0.6615 - recall_80: 0.0000e+00 - val_loss: 0.6446 - val_categorical_accuracy: 0.6198 - val_recall_80: 0.0000e+00\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6271 - categorical_accuracy: 0.6615 - recall_80: 0.0000e+00 - val_loss: 0.6380 - val_categorical_accuracy: 0.6198 - val_recall_80: 0.0000e+00\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6203 - categorical_accuracy: 0.6615 - recall_80: 0.0000e+00 - val_loss: 0.6292 - val_categorical_accuracy: 0.6198 - val_recall_80: 0.0000e+00\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.6135 - categorical_accuracy: 0.6615 - recall_80: 0.0000e+00 - val_loss: 0.6206 - val_categorical_accuracy: 0.6198 - val_recall_80: 0.0000e+00\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.6071 - categorical_accuracy: 0.6597 - recall_80: 0.0000e+00 - val_loss: 0.6099 - val_categorical_accuracy: 0.6250 - val_recall_80: 0.0137\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.5994 - categorical_accuracy: 0.6615 - recall_80: 0.0308 - val_loss: 0.6029 - val_categorical_accuracy: 0.6458 - val_recall_80: 0.0685\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.5919 - categorical_accuracy: 0.6736 - recall_80: 0.0821 - val_loss: 0.5942 - val_categorical_accuracy: 0.6771 - val_recall_80: 0.1644\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5841 - categorical_accuracy: 0.6788 - recall_80: 0.1385 - val_loss: 0.5865 - val_categorical_accuracy: 0.6979 - val_recall_80: 0.2329\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.5768 - categorical_accuracy: 0.6858 - recall_80: 0.1692 - val_loss: 0.5796 - val_categorical_accuracy: 0.6927 - val_recall_80: 0.2466\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5703 - categorical_accuracy: 0.6892 - recall_80: 0.2000 - val_loss: 0.5721 - val_categorical_accuracy: 0.7292 - val_recall_80: 0.3836\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5633 - categorical_accuracy: 0.7031 - recall_80: 0.3282 - val_loss: 0.5648 - val_categorical_accuracy: 0.7448 - val_recall_80: 0.4932\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5582 - categorical_accuracy: 0.7257 - recall_80: 0.4103 - val_loss: 0.5593 - val_categorical_accuracy: 0.7396 - val_recall_80: 0.4521\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.5549 - categorical_accuracy: 0.6997 - recall_80: 0.2359 - val_loss: 0.5577 - val_categorical_accuracy: 0.7188 - val_recall_80: 0.3425\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.5456 - categorical_accuracy: 0.7188 - recall_80: 0.3538 - val_loss: 0.5488 - val_categorical_accuracy: 0.7396 - val_recall_80: 0.5342\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.5410 - categorical_accuracy: 0.7413 - recall_80: 0.4821 - val_loss: 0.5427 - val_categorical_accuracy: 0.7344 - val_recall_80: 0.5479\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5355 - categorical_accuracy: 0.7396 - recall_80: 0.4359 - val_loss: 0.5408 - val_categorical_accuracy: 0.7552 - val_recall_80: 0.4932\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.5313 - categorical_accuracy: 0.7309 - recall_80: 0.3744 - val_loss: 0.5368 - val_categorical_accuracy: 0.7552 - val_recall_80: 0.4932\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 22ms/step - loss: 0.5253 - categorical_accuracy: 0.7378 - recall_80: 0.4308 - val_loss: 0.5302 - val_categorical_accuracy: 0.7500 - val_recall_80: 0.5753\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.5227 - categorical_accuracy: 0.7535 - recall_80: 0.5026 - val_loss: 0.5279 - val_categorical_accuracy: 0.7708 - val_recall_80: 0.5753\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5178 - categorical_accuracy: 0.7361 - recall_80: 0.4256 - val_loss: 0.5247 - val_categorical_accuracy: 0.7656 - val_recall_80: 0.5616\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5138 - categorical_accuracy: 0.7465 - recall_80: 0.4615 - val_loss: 0.5209 - val_categorical_accuracy: 0.7656 - val_recall_80: 0.5890\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5105 - categorical_accuracy: 0.7483 - recall_80: 0.4615 - val_loss: 0.5188 - val_categorical_accuracy: 0.7656 - val_recall_80: 0.5753\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5066 - categorical_accuracy: 0.7552 - recall_80: 0.4821 - val_loss: 0.5144 - val_categorical_accuracy: 0.7656 - val_recall_80: 0.6027\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5039 - categorical_accuracy: 0.7622 - recall_80: 0.5282 - val_loss: 0.5121 - val_categorical_accuracy: 0.7708 - val_recall_80: 0.6164\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4998 - categorical_accuracy: 0.7656 - recall_80: 0.4923 - val_loss: 0.5124 - val_categorical_accuracy: 0.7760 - val_recall_80: 0.5753\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4973 - categorical_accuracy: 0.7674 - recall_80: 0.4615 - val_loss: 0.5084 - val_categorical_accuracy: 0.7812 - val_recall_80: 0.6027\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.4935 - categorical_accuracy: 0.7622 - recall_80: 0.4974 - val_loss: 0.5055 - val_categorical_accuracy: 0.7865 - val_recall_80: 0.6301\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4916 - categorical_accuracy: 0.7674 - recall_80: 0.5385 - val_loss: 0.5041 - val_categorical_accuracy: 0.7865 - val_recall_80: 0.6164\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.4878 - categorical_accuracy: 0.7656 - recall_80: 0.5077 - val_loss: 0.5026 - val_categorical_accuracy: 0.7865 - val_recall_80: 0.6164\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.4863 - categorical_accuracy: 0.7812 - recall_80: 0.4974 - val_loss: 0.5009 - val_categorical_accuracy: 0.7812 - val_recall_80: 0.6164\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4881 - categorical_accuracy: 0.7604 - recall_80: 0.6051 - val_loss: 0.4982 - val_categorical_accuracy: 0.7969 - val_recall_80: 0.6575\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.4808 - categorical_accuracy: 0.7812 - recall_80: 0.4974 - val_loss: 0.5017 - val_categorical_accuracy: 0.7760 - val_recall_80: 0.5753\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4802 - categorical_accuracy: 0.7830 - recall_80: 0.4974 - val_loss: 0.4956 - val_categorical_accuracy: 0.7969 - val_recall_80: 0.6438\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4770 - categorical_accuracy: 0.7726 - recall_80: 0.5692 - val_loss: 0.4946 - val_categorical_accuracy: 0.7865 - val_recall_80: 0.6164\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4752 - categorical_accuracy: 0.7812 - recall_80: 0.5026 - val_loss: 0.4957 - val_categorical_accuracy: 0.7865 - val_recall_80: 0.6027\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.4718 - categorical_accuracy: 0.7778 - recall_80: 0.5231 - val_loss: 0.4923 - val_categorical_accuracy: 0.7969 - val_recall_80: 0.6438\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4709 - categorical_accuracy: 0.7726 - recall_80: 0.5487 - val_loss: 0.4918 - val_categorical_accuracy: 0.7917 - val_recall_80: 0.6301\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4691 - categorical_accuracy: 0.7778 - recall_80: 0.5231 - val_loss: 0.4918 - val_categorical_accuracy: 0.7917 - val_recall_80: 0.6301\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4674 - categorical_accuracy: 0.7708 - recall_80: 0.5590 - val_loss: 0.4899 - val_categorical_accuracy: 0.7917 - val_recall_80: 0.6438\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 22ms/step - loss: 0.4658 - categorical_accuracy: 0.7778 - recall_80: 0.5282 - val_loss: 0.4936 - val_categorical_accuracy: 0.7812 - val_recall_80: 0.5890\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4638 - categorical_accuracy: 0.7917 - recall_80: 0.5282 - val_loss: 0.4890 - val_categorical_accuracy: 0.7865 - val_recall_80: 0.6575\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 3s 58ms/step - loss: 0.6747 - categorical_accuracy: 0.6597 - recall_81: 0.0000e+00 - val_loss: 0.6894 - val_categorical_accuracy: 0.6198 - val_recall_81: 0.0000e+00\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6643 - categorical_accuracy: 0.6615 - recall_81: 0.0000e+00 - val_loss: 0.6840 - val_categorical_accuracy: 0.6198 - val_recall_81: 0.0000e+00\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6534 - categorical_accuracy: 0.6615 - recall_81: 0.0000e+00 - val_loss: 0.6673 - val_categorical_accuracy: 0.6198 - val_recall_81: 0.0000e+00\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6472 - categorical_accuracy: 0.6615 - recall_81: 0.0000e+00 - val_loss: 0.6585 - val_categorical_accuracy: 0.6198 - val_recall_81: 0.0000e+00\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.6390 - categorical_accuracy: 0.6615 - recall_81: 0.0000e+00 - val_loss: 0.6533 - val_categorical_accuracy: 0.6198 - val_recall_81: 0.0000e+00\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6313 - categorical_accuracy: 0.6615 - recall_81: 0.0000e+00 - val_loss: 0.6392 - val_categorical_accuracy: 0.6250 - val_recall_81: 0.0137\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6224 - categorical_accuracy: 0.6615 - recall_81: 0.0000e+00 - val_loss: 0.6312 - val_categorical_accuracy: 0.6250 - val_recall_81: 0.0137\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 20ms/step - loss: 0.6132 - categorical_accuracy: 0.6615 - recall_81: 0.0205 - val_loss: 0.6172 - val_categorical_accuracy: 0.6354 - val_recall_81: 0.0548\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.6041 - categorical_accuracy: 0.6615 - recall_81: 0.0410 - val_loss: 0.6066 - val_categorical_accuracy: 0.6562 - val_recall_81: 0.1096\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.5947 - categorical_accuracy: 0.6597 - recall_81: 0.0410 - val_loss: 0.5978 - val_categorical_accuracy: 0.6615 - val_recall_81: 0.1370\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.5852 - categorical_accuracy: 0.6997 - recall_81: 0.2410 - val_loss: 0.5835 - val_categorical_accuracy: 0.6927 - val_recall_81: 0.2466\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.5742 - categorical_accuracy: 0.6944 - recall_81: 0.1897 - val_loss: 0.5753 - val_categorical_accuracy: 0.7031 - val_recall_81: 0.2603\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.5644 - categorical_accuracy: 0.7066 - recall_81: 0.2513 - val_loss: 0.5600 - val_categorical_accuracy: 0.7344 - val_recall_81: 0.3699\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.5566 - categorical_accuracy: 0.7118 - recall_81: 0.2718 - val_loss: 0.5500 - val_categorical_accuracy: 0.7396 - val_recall_81: 0.3836\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5468 - categorical_accuracy: 0.7292 - recall_81: 0.3846 - val_loss: 0.5411 - val_categorical_accuracy: 0.7396 - val_recall_81: 0.3973\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5364 - categorical_accuracy: 0.7344 - recall_81: 0.3795 - val_loss: 0.5318 - val_categorical_accuracy: 0.7812 - val_recall_81: 0.5479\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.5289 - categorical_accuracy: 0.7396 - recall_81: 0.4205 - val_loss: 0.5314 - val_categorical_accuracy: 0.7448 - val_recall_81: 0.4247\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 23ms/step - loss: 0.5230 - categorical_accuracy: 0.7465 - recall_81: 0.5026 - val_loss: 0.5220 - val_categorical_accuracy: 0.7656 - val_recall_81: 0.4932\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5169 - categorical_accuracy: 0.7326 - recall_81: 0.3333 - val_loss: 0.5167 - val_categorical_accuracy: 0.7708 - val_recall_81: 0.5205\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5020 - categorical_accuracy: 0.7674 - recall_81: 0.4821 - val_loss: 0.5103 - val_categorical_accuracy: 0.8021 - val_recall_81: 0.6438\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.4974 - categorical_accuracy: 0.7778 - recall_81: 0.5487 - val_loss: 0.5109 - val_categorical_accuracy: 0.7760 - val_recall_81: 0.5205\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 21ms/step - loss: 0.4894 - categorical_accuracy: 0.7726 - recall_81: 0.4974 - val_loss: 0.5044 - val_categorical_accuracy: 0.7865 - val_recall_81: 0.5616\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4852 - categorical_accuracy: 0.7795 - recall_81: 0.4923 - val_loss: 0.5011 - val_categorical_accuracy: 0.7865 - val_recall_81: 0.5890\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4785 - categorical_accuracy: 0.7812 - recall_81: 0.5077 - val_loss: 0.5037 - val_categorical_accuracy: 0.7812 - val_recall_81: 0.5479\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.4732 - categorical_accuracy: 0.7760 - recall_81: 0.5077 - val_loss: 0.4974 - val_categorical_accuracy: 0.7917 - val_recall_81: 0.5890\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 22ms/step - loss: 0.4683 - categorical_accuracy: 0.7726 - recall_81: 0.5282 - val_loss: 0.4979 - val_categorical_accuracy: 0.7917 - val_recall_81: 0.6027\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4638 - categorical_accuracy: 0.7795 - recall_81: 0.5077 - val_loss: 0.4921 - val_categorical_accuracy: 0.8021 - val_recall_81: 0.6575\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.4614 - categorical_accuracy: 0.7830 - recall_81: 0.5385 - val_loss: 0.4952 - val_categorical_accuracy: 0.7865 - val_recall_81: 0.5890\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4568 - categorical_accuracy: 0.7812 - recall_81: 0.5590 - val_loss: 0.4955 - val_categorical_accuracy: 0.7917 - val_recall_81: 0.6027\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4569 - categorical_accuracy: 0.7812 - recall_81: 0.4769 - val_loss: 0.4907 - val_categorical_accuracy: 0.7917 - val_recall_81: 0.6575\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4651 - categorical_accuracy: 0.7951 - recall_81: 0.7179 - val_loss: 0.5092 - val_categorical_accuracy: 0.7760 - val_recall_81: 0.5342\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.4590 - categorical_accuracy: 0.7830 - recall_81: 0.4718 - val_loss: 0.4932 - val_categorical_accuracy: 0.7865 - val_recall_81: 0.6986\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4515 - categorical_accuracy: 0.7969 - recall_81: 0.5897 - val_loss: 0.4988 - val_categorical_accuracy: 0.7812 - val_recall_81: 0.5753\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4462 - categorical_accuracy: 0.7969 - recall_81: 0.6000 - val_loss: 0.4970 - val_categorical_accuracy: 0.7812 - val_recall_81: 0.5890\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4464 - categorical_accuracy: 0.7865 - recall_81: 0.5333 - val_loss: 0.4932 - val_categorical_accuracy: 0.7969 - val_recall_81: 0.6438\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.4507 - categorical_accuracy: 0.8021 - recall_81: 0.7231 - val_loss: 0.5050 - val_categorical_accuracy: 0.7865 - val_recall_81: 0.5753\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.4416 - categorical_accuracy: 0.7865 - recall_81: 0.5179 - val_loss: 0.4995 - val_categorical_accuracy: 0.7812 - val_recall_81: 0.5890\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.4386 - categorical_accuracy: 0.7865 - recall_81: 0.6051 - val_loss: 0.4998 - val_categorical_accuracy: 0.7760 - val_recall_81: 0.5890\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4398 - categorical_accuracy: 0.7899 - recall_81: 0.5231 - val_loss: 0.4996 - val_categorical_accuracy: 0.7812 - val_recall_81: 0.6027\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4380 - categorical_accuracy: 0.8003 - recall_81: 0.6359 - val_loss: 0.5024 - val_categorical_accuracy: 0.7708 - val_recall_81: 0.5753\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 23ms/step - loss: 0.4334 - categorical_accuracy: 0.7934 - recall_81: 0.6051 - val_loss: 0.5017 - val_categorical_accuracy: 0.7812 - val_recall_81: 0.6027\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4335 - categorical_accuracy: 0.7917 - recall_81: 0.5795 - val_loss: 0.5010 - val_categorical_accuracy: 0.7865 - val_recall_81: 0.6301\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4332 - categorical_accuracy: 0.7969 - recall_81: 0.6154 - val_loss: 0.5030 - val_categorical_accuracy: 0.7760 - val_recall_81: 0.6301\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4338 - categorical_accuracy: 0.7899 - recall_81: 0.5897 - val_loss: 0.5055 - val_categorical_accuracy: 0.7865 - val_recall_81: 0.6164\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.4318 - categorical_accuracy: 0.7917 - recall_81: 0.6051 - val_loss: 0.5042 - val_categorical_accuracy: 0.7865 - val_recall_81: 0.6438\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4348 - categorical_accuracy: 0.8073 - recall_81: 0.6667 - val_loss: 0.5138 - val_categorical_accuracy: 0.7760 - val_recall_81: 0.5753\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4311 - categorical_accuracy: 0.7882 - recall_81: 0.5385 - val_loss: 0.5073 - val_categorical_accuracy: 0.7760 - val_recall_81: 0.6712\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.4324 - categorical_accuracy: 0.8038 - recall_81: 0.6974 - val_loss: 0.5189 - val_categorical_accuracy: 0.7760 - val_recall_81: 0.5753\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.4361 - categorical_accuracy: 0.7899 - recall_81: 0.5436 - val_loss: 0.5083 - val_categorical_accuracy: 0.7708 - val_recall_81: 0.6575\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.4291 - categorical_accuracy: 0.7951 - recall_81: 0.6308 - val_loss: 0.5112 - val_categorical_accuracy: 0.7812 - val_recall_81: 0.6027\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 3s 69ms/step - loss: 0.6872 - categorical_accuracy: 0.5955 - recall_82: 0.0513 - val_loss: 0.6820 - val_categorical_accuracy: 0.6198 - val_recall_82: 0.0000e+00\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.6671 - categorical_accuracy: 0.6615 - recall_82: 0.0000e+00 - val_loss: 0.6733 - val_categorical_accuracy: 0.6198 - val_recall_82: 0.0000e+00\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6544 - categorical_accuracy: 0.6615 - recall_82: 0.0000e+00 - val_loss: 0.6680 - val_categorical_accuracy: 0.6198 - val_recall_82: 0.0000e+00\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.6468 - categorical_accuracy: 0.6615 - recall_82: 0.0000e+00 - val_loss: 0.6650 - val_categorical_accuracy: 0.6198 - val_recall_82: 0.0000e+00\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6420 - categorical_accuracy: 0.6615 - recall_82: 0.0000e+00 - val_loss: 0.6611 - val_categorical_accuracy: 0.6198 - val_recall_82: 0.0000e+00\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.6371 - categorical_accuracy: 0.6615 - recall_82: 0.0000e+00 - val_loss: 0.6554 - val_categorical_accuracy: 0.6198 - val_recall_82: 0.0000e+00\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6319 - categorical_accuracy: 0.6615 - recall_82: 0.0000e+00 - val_loss: 0.6495 - val_categorical_accuracy: 0.6198 - val_recall_82: 0.0000e+00\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.6271 - categorical_accuracy: 0.6615 - recall_82: 0.0000e+00 - val_loss: 0.6426 - val_categorical_accuracy: 0.6198 - val_recall_82: 0.0000e+00\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6215 - categorical_accuracy: 0.6615 - recall_82: 0.0000e+00 - val_loss: 0.6366 - val_categorical_accuracy: 0.6198 - val_recall_82: 0.0000e+00\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6157 - categorical_accuracy: 0.6615 - recall_82: 0.0000e+00 - val_loss: 0.6311 - val_categorical_accuracy: 0.6198 - val_recall_82: 0.0000e+00\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6096 - categorical_accuracy: 0.6597 - recall_82: 0.0000e+00 - val_loss: 0.6221 - val_categorical_accuracy: 0.6250 - val_recall_82: 0.0137\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.6033 - categorical_accuracy: 0.6632 - recall_82: 0.0103 - val_loss: 0.6150 - val_categorical_accuracy: 0.6250 - val_recall_82: 0.0137\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5940 - categorical_accuracy: 0.6684 - recall_82: 0.0513 - val_loss: 0.6026 - val_categorical_accuracy: 0.6510 - val_recall_82: 0.1096\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.5853 - categorical_accuracy: 0.6771 - recall_82: 0.1077 - val_loss: 0.5953 - val_categorical_accuracy: 0.6823 - val_recall_82: 0.2055\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5773 - categorical_accuracy: 0.6858 - recall_82: 0.1487 - val_loss: 0.5861 - val_categorical_accuracy: 0.7031 - val_recall_82: 0.2877\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 21ms/step - loss: 0.5693 - categorical_accuracy: 0.6910 - recall_82: 0.2564 - val_loss: 0.5783 - val_categorical_accuracy: 0.7135 - val_recall_82: 0.3288\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5652 - categorical_accuracy: 0.7014 - recall_82: 0.3590 - val_loss: 0.5706 - val_categorical_accuracy: 0.7135 - val_recall_82: 0.3973\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5558 - categorical_accuracy: 0.6962 - recall_82: 0.2410 - val_loss: 0.5714 - val_categorical_accuracy: 0.7031 - val_recall_82: 0.3014\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5481 - categorical_accuracy: 0.6962 - recall_82: 0.2667 - val_loss: 0.5598 - val_categorical_accuracy: 0.7292 - val_recall_82: 0.4658\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.5448 - categorical_accuracy: 0.7361 - recall_82: 0.4872 - val_loss: 0.5553 - val_categorical_accuracy: 0.7396 - val_recall_82: 0.4795\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.5357 - categorical_accuracy: 0.7240 - recall_82: 0.3641 - val_loss: 0.5541 - val_categorical_accuracy: 0.7344 - val_recall_82: 0.4247\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.5298 - categorical_accuracy: 0.7274 - recall_82: 0.3590 - val_loss: 0.5458 - val_categorical_accuracy: 0.7500 - val_recall_82: 0.5068\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5243 - categorical_accuracy: 0.7378 - recall_82: 0.4564 - val_loss: 0.5406 - val_categorical_accuracy: 0.7656 - val_recall_82: 0.5342\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 25ms/step - loss: 0.5191 - categorical_accuracy: 0.7344 - recall_82: 0.4718 - val_loss: 0.5372 - val_categorical_accuracy: 0.7604 - val_recall_82: 0.5205\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5138 - categorical_accuracy: 0.7500 - recall_82: 0.4359 - val_loss: 0.5331 - val_categorical_accuracy: 0.7604 - val_recall_82: 0.5342\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5130 - categorical_accuracy: 0.7413 - recall_82: 0.5282 - val_loss: 0.5257 - val_categorical_accuracy: 0.7812 - val_recall_82: 0.5890\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5058 - categorical_accuracy: 0.7569 - recall_82: 0.4308 - val_loss: 0.5277 - val_categorical_accuracy: 0.7552 - val_recall_82: 0.5068\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.5011 - categorical_accuracy: 0.7535 - recall_82: 0.5179 - val_loss: 0.5184 - val_categorical_accuracy: 0.7865 - val_recall_82: 0.6164\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4960 - categorical_accuracy: 0.7674 - recall_82: 0.4462 - val_loss: 0.5200 - val_categorical_accuracy: 0.7604 - val_recall_82: 0.5068\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4925 - categorical_accuracy: 0.7587 - recall_82: 0.5538 - val_loss: 0.5126 - val_categorical_accuracy: 0.7969 - val_recall_82: 0.6301\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.4883 - categorical_accuracy: 0.7674 - recall_82: 0.4615 - val_loss: 0.5121 - val_categorical_accuracy: 0.7917 - val_recall_82: 0.6027\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.4858 - categorical_accuracy: 0.7726 - recall_82: 0.5897 - val_loss: 0.5086 - val_categorical_accuracy: 0.8021 - val_recall_82: 0.6301\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 20ms/step - loss: 0.4843 - categorical_accuracy: 0.7674 - recall_82: 0.4667 - val_loss: 0.5074 - val_categorical_accuracy: 0.7969 - val_recall_82: 0.6164\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.4783 - categorical_accuracy: 0.7708 - recall_82: 0.4769 - val_loss: 0.5053 - val_categorical_accuracy: 0.8021 - val_recall_82: 0.6301\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4793 - categorical_accuracy: 0.7622 - recall_82: 0.6205 - val_loss: 0.5029 - val_categorical_accuracy: 0.7917 - val_recall_82: 0.6301\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4706 - categorical_accuracy: 0.7882 - recall_82: 0.4974 - val_loss: 0.5085 - val_categorical_accuracy: 0.7865 - val_recall_82: 0.5616\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.4666 - categorical_accuracy: 0.7899 - recall_82: 0.5128 - val_loss: 0.5002 - val_categorical_accuracy: 0.7969 - val_recall_82: 0.6575\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4670 - categorical_accuracy: 0.7726 - recall_82: 0.6051 - val_loss: 0.5005 - val_categorical_accuracy: 0.8021 - val_recall_82: 0.6438\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.4623 - categorical_accuracy: 0.7743 - recall_82: 0.5282 - val_loss: 0.5011 - val_categorical_accuracy: 0.7969 - val_recall_82: 0.6164\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4600 - categorical_accuracy: 0.7743 - recall_82: 0.5179 - val_loss: 0.5005 - val_categorical_accuracy: 0.7865 - val_recall_82: 0.5890\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4586 - categorical_accuracy: 0.7743 - recall_82: 0.5077 - val_loss: 0.4996 - val_categorical_accuracy: 0.7865 - val_recall_82: 0.6027\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 22ms/step - loss: 0.4571 - categorical_accuracy: 0.7830 - recall_82: 0.5692 - val_loss: 0.4993 - val_categorical_accuracy: 0.7865 - val_recall_82: 0.6027\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4542 - categorical_accuracy: 0.7778 - recall_82: 0.5282 - val_loss: 0.5021 - val_categorical_accuracy: 0.7865 - val_recall_82: 0.5753\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4542 - categorical_accuracy: 0.7778 - recall_82: 0.5846 - val_loss: 0.4993 - val_categorical_accuracy: 0.8021 - val_recall_82: 0.6301\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.4534 - categorical_accuracy: 0.7812 - recall_82: 0.4974 - val_loss: 0.5019 - val_categorical_accuracy: 0.7865 - val_recall_82: 0.5890\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4492 - categorical_accuracy: 0.7726 - recall_82: 0.5590 - val_loss: 0.4982 - val_categorical_accuracy: 0.7865 - val_recall_82: 0.6575\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.4480 - categorical_accuracy: 0.7778 - recall_82: 0.5590 - val_loss: 0.5037 - val_categorical_accuracy: 0.7865 - val_recall_82: 0.5890\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 8ms/step - loss: 0.4515 - categorical_accuracy: 0.7847 - recall_82: 0.5436 - val_loss: 0.4987 - val_categorical_accuracy: 0.7917 - val_recall_82: 0.6575\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.4479 - categorical_accuracy: 0.7865 - recall_82: 0.5282 - val_loss: 0.5049 - val_categorical_accuracy: 0.7760 - val_recall_82: 0.5616\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 21ms/step - loss: 0.4476 - categorical_accuracy: 0.7795 - recall_82: 0.6000 - val_loss: 0.5000 - val_categorical_accuracy: 0.7812 - val_recall_82: 0.6575\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 3s 66ms/step - loss: 0.7673 - categorical_accuracy: 0.3438 - recall_83: 1.0000 - val_loss: 0.7133 - val_categorical_accuracy: 0.3750 - val_recall_83: 0.9863\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.6982 - categorical_accuracy: 0.4740 - recall_83: 0.7795 - val_loss: 0.6792 - val_categorical_accuracy: 0.6198 - val_recall_83: 0.0685\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6640 - categorical_accuracy: 0.6580 - recall_83: 0.0103 - val_loss: 0.6681 - val_categorical_accuracy: 0.6198 - val_recall_83: 0.0000e+00\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 21ms/step - loss: 0.6504 - categorical_accuracy: 0.6615 - recall_83: 0.0000e+00 - val_loss: 0.6651 - val_categorical_accuracy: 0.6198 - val_recall_83: 0.0000e+00\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6447 - categorical_accuracy: 0.6615 - recall_83: 0.0000e+00 - val_loss: 0.6623 - val_categorical_accuracy: 0.6198 - val_recall_83: 0.0000e+00\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.6407 - categorical_accuracy: 0.6615 - recall_83: 0.0000e+00 - val_loss: 0.6570 - val_categorical_accuracy: 0.6198 - val_recall_83: 0.0000e+00\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 9ms/step - loss: 0.6359 - categorical_accuracy: 0.6615 - recall_83: 0.0000e+00 - val_loss: 0.6528 - val_categorical_accuracy: 0.6198 - val_recall_83: 0.0000e+00\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6304 - categorical_accuracy: 0.6615 - recall_83: 0.0000e+00 - val_loss: 0.6472 - val_categorical_accuracy: 0.6198 - val_recall_83: 0.0000e+00\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.6245 - categorical_accuracy: 0.6615 - recall_83: 0.0000e+00 - val_loss: 0.6410 - val_categorical_accuracy: 0.6198 - val_recall_83: 0.0000e+00\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 31ms/step - loss: 0.6186 - categorical_accuracy: 0.6615 - recall_83: 0.0000e+00 - val_loss: 0.6341 - val_categorical_accuracy: 0.6198 - val_recall_83: 0.0000e+00\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.6130 - categorical_accuracy: 0.6615 - recall_83: 0.0000e+00 - val_loss: 0.6266 - val_categorical_accuracy: 0.6250 - val_recall_83: 0.0137\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 29ms/step - loss: 0.6062 - categorical_accuracy: 0.6597 - recall_83: 0.0051 - val_loss: 0.6205 - val_categorical_accuracy: 0.6354 - val_recall_83: 0.0411\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 27ms/step - loss: 0.5999 - categorical_accuracy: 0.6632 - recall_83: 0.0205 - val_loss: 0.6134 - val_categorical_accuracy: 0.6406 - val_recall_83: 0.0548\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 24ms/step - loss: 0.5939 - categorical_accuracy: 0.6753 - recall_83: 0.0923 - val_loss: 0.6050 - val_categorical_accuracy: 0.6719 - val_recall_83: 0.1644\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5863 - categorical_accuracy: 0.6806 - recall_83: 0.1128 - val_loss: 0.5999 - val_categorical_accuracy: 0.6875 - val_recall_83: 0.2055\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 22ms/step - loss: 0.5816 - categorical_accuracy: 0.6979 - recall_83: 0.1897 - val_loss: 0.5907 - val_categorical_accuracy: 0.6979 - val_recall_83: 0.3014\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5749 - categorical_accuracy: 0.6892 - recall_83: 0.1744 - val_loss: 0.5894 - val_categorical_accuracy: 0.6875 - val_recall_83: 0.2192\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 37ms/step - loss: 0.5700 - categorical_accuracy: 0.6892 - recall_83: 0.2051 - val_loss: 0.5785 - val_categorical_accuracy: 0.7188 - val_recall_83: 0.3562\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.5619 - categorical_accuracy: 0.6962 - recall_83: 0.2154 - val_loss: 0.5760 - val_categorical_accuracy: 0.7188 - val_recall_83: 0.3425\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5552 - categorical_accuracy: 0.6997 - recall_83: 0.2410 - val_loss: 0.5683 - val_categorical_accuracy: 0.7344 - val_recall_83: 0.3973\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5487 - categorical_accuracy: 0.7135 - recall_83: 0.3026 - val_loss: 0.5613 - val_categorical_accuracy: 0.7500 - val_recall_83: 0.4658\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.5433 - categorical_accuracy: 0.7135 - recall_83: 0.3590 - val_loss: 0.5572 - val_categorical_accuracy: 0.7552 - val_recall_83: 0.4795\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5372 - categorical_accuracy: 0.7205 - recall_83: 0.3795 - val_loss: 0.5494 - val_categorical_accuracy: 0.7656 - val_recall_83: 0.5068\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.5320 - categorical_accuracy: 0.7240 - recall_83: 0.3641 - val_loss: 0.5456 - val_categorical_accuracy: 0.7604 - val_recall_83: 0.4932\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5271 - categorical_accuracy: 0.7292 - recall_83: 0.4667 - val_loss: 0.5404 - val_categorical_accuracy: 0.7500 - val_recall_83: 0.4932\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5254 - categorical_accuracy: 0.7135 - recall_83: 0.3282 - val_loss: 0.5386 - val_categorical_accuracy: 0.7500 - val_recall_83: 0.4932\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5157 - categorical_accuracy: 0.7326 - recall_83: 0.4462 - val_loss: 0.5360 - val_categorical_accuracy: 0.7552 - val_recall_83: 0.5205\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.5101 - categorical_accuracy: 0.7326 - recall_83: 0.4564 - val_loss: 0.5346 - val_categorical_accuracy: 0.7552 - val_recall_83: 0.5205\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5063 - categorical_accuracy: 0.7344 - recall_83: 0.4821 - val_loss: 0.5322 - val_categorical_accuracy: 0.7552 - val_recall_83: 0.5479\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5048 - categorical_accuracy: 0.7378 - recall_83: 0.4154 - val_loss: 0.5298 - val_categorical_accuracy: 0.7552 - val_recall_83: 0.5479\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 27ms/step - loss: 0.5016 - categorical_accuracy: 0.7431 - recall_83: 0.4974 - val_loss: 0.5325 - val_categorical_accuracy: 0.7552 - val_recall_83: 0.5068\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4964 - categorical_accuracy: 0.7413 - recall_83: 0.4205 - val_loss: 0.5246 - val_categorical_accuracy: 0.7396 - val_recall_83: 0.5616\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.4966 - categorical_accuracy: 0.7535 - recall_83: 0.5385 - val_loss: 0.5263 - val_categorical_accuracy: 0.7604 - val_recall_83: 0.5342\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4907 - categorical_accuracy: 0.7552 - recall_83: 0.4872 - val_loss: 0.5200 - val_categorical_accuracy: 0.7552 - val_recall_83: 0.5616\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 1s 71ms/step - loss: 0.4882 - categorical_accuracy: 0.7535 - recall_83: 0.5231 - val_loss: 0.5193 - val_categorical_accuracy: 0.7500 - val_recall_83: 0.5616\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.4858 - categorical_accuracy: 0.7552 - recall_83: 0.5128 - val_loss: 0.5205 - val_categorical_accuracy: 0.7604 - val_recall_83: 0.5479\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 21ms/step - loss: 0.4830 - categorical_accuracy: 0.7517 - recall_83: 0.4923 - val_loss: 0.5140 - val_categorical_accuracy: 0.7656 - val_recall_83: 0.6301\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.4810 - categorical_accuracy: 0.7587 - recall_83: 0.5385 - val_loss: 0.5165 - val_categorical_accuracy: 0.7656 - val_recall_83: 0.5616\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4782 - categorical_accuracy: 0.7587 - recall_83: 0.5077 - val_loss: 0.5133 - val_categorical_accuracy: 0.7760 - val_recall_83: 0.6164\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4769 - categorical_accuracy: 0.7604 - recall_83: 0.5590 - val_loss: 0.5122 - val_categorical_accuracy: 0.7760 - val_recall_83: 0.6164\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 30ms/step - loss: 0.4723 - categorical_accuracy: 0.7604 - recall_83: 0.5128 - val_loss: 0.5167 - val_categorical_accuracy: 0.7865 - val_recall_83: 0.5753\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4723 - categorical_accuracy: 0.7639 - recall_83: 0.5333 - val_loss: 0.5107 - val_categorical_accuracy: 0.7760 - val_recall_83: 0.6164\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4696 - categorical_accuracy: 0.7656 - recall_83: 0.5538 - val_loss: 0.5102 - val_categorical_accuracy: 0.7760 - val_recall_83: 0.6164\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 25ms/step - loss: 0.4691 - categorical_accuracy: 0.7622 - recall_83: 0.5538 - val_loss: 0.5101 - val_categorical_accuracy: 0.7812 - val_recall_83: 0.6164\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4661 - categorical_accuracy: 0.7674 - recall_83: 0.5282 - val_loss: 0.5077 - val_categorical_accuracy: 0.7865 - val_recall_83: 0.6301\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4648 - categorical_accuracy: 0.7726 - recall_83: 0.5897 - val_loss: 0.5070 - val_categorical_accuracy: 0.7865 - val_recall_83: 0.6301\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4627 - categorical_accuracy: 0.7674 - recall_83: 0.5538 - val_loss: 0.5074 - val_categorical_accuracy: 0.7812 - val_recall_83: 0.6164\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4624 - categorical_accuracy: 0.7656 - recall_83: 0.5333 - val_loss: 0.5062 - val_categorical_accuracy: 0.7812 - val_recall_83: 0.6301\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 31ms/step - loss: 0.4608 - categorical_accuracy: 0.7795 - recall_83: 0.6051 - val_loss: 0.5041 - val_categorical_accuracy: 0.8073 - val_recall_83: 0.6986\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 20ms/step - loss: 0.4599 - categorical_accuracy: 0.7743 - recall_83: 0.5231 - val_loss: 0.5083 - val_categorical_accuracy: 0.7865 - val_recall_83: 0.6164\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 4s 84ms/step - loss: 0.6769 - categorical_accuracy: 0.6615 - recall_84: 0.0000e+00 - val_loss: 0.6845 - val_categorical_accuracy: 0.6198 - val_recall_84: 0.0000e+00\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.6628 - categorical_accuracy: 0.6615 - recall_84: 0.0000e+00 - val_loss: 0.6796 - val_categorical_accuracy: 0.6198 - val_recall_84: 0.0000e+00\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 24ms/step - loss: 0.6556 - categorical_accuracy: 0.6615 - recall_84: 0.0000e+00 - val_loss: 0.6732 - val_categorical_accuracy: 0.6198 - val_recall_84: 0.0000e+00\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 23ms/step - loss: 0.6492 - categorical_accuracy: 0.6615 - recall_84: 0.0000e+00 - val_loss: 0.6660 - val_categorical_accuracy: 0.6198 - val_recall_84: 0.0000e+00\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 28ms/step - loss: 0.6430 - categorical_accuracy: 0.6615 - recall_84: 0.0000e+00 - val_loss: 0.6579 - val_categorical_accuracy: 0.6198 - val_recall_84: 0.0000e+00\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 20ms/step - loss: 0.6362 - categorical_accuracy: 0.6615 - recall_84: 0.0000e+00 - val_loss: 0.6518 - val_categorical_accuracy: 0.6198 - val_recall_84: 0.0000e+00\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.6266 - categorical_accuracy: 0.6615 - recall_84: 0.0000e+00 - val_loss: 0.6388 - val_categorical_accuracy: 0.6198 - val_recall_84: 0.0000e+00\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.6179 - categorical_accuracy: 0.6615 - recall_84: 0.0051 - val_loss: 0.6281 - val_categorical_accuracy: 0.6302 - val_recall_84: 0.0274\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.6081 - categorical_accuracy: 0.6667 - recall_84: 0.0256 - val_loss: 0.6212 - val_categorical_accuracy: 0.6354 - val_recall_84: 0.0411\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 24ms/step - loss: 0.5996 - categorical_accuracy: 0.6684 - recall_84: 0.0615 - val_loss: 0.6068 - val_categorical_accuracy: 0.6719 - val_recall_84: 0.1507\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.5895 - categorical_accuracy: 0.6875 - recall_84: 0.1179 - val_loss: 0.5972 - val_categorical_accuracy: 0.6823 - val_recall_84: 0.1781\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5813 - categorical_accuracy: 0.7083 - recall_84: 0.2205 - val_loss: 0.5860 - val_categorical_accuracy: 0.7031 - val_recall_84: 0.2740\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5706 - categorical_accuracy: 0.7118 - recall_84: 0.2103 - val_loss: 0.5798 - val_categorical_accuracy: 0.6979 - val_recall_84: 0.2603\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5616 - categorical_accuracy: 0.7205 - recall_84: 0.2564 - val_loss: 0.5688 - val_categorical_accuracy: 0.7083 - val_recall_84: 0.3014\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 0.5557 - categorical_accuracy: 0.7135 - recall_84: 0.2564 - val_loss: 0.5603 - val_categorical_accuracy: 0.7396 - val_recall_84: 0.3973\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.5487 - categorical_accuracy: 0.7118 - recall_84: 0.2718 - val_loss: 0.5552 - val_categorical_accuracy: 0.7344 - val_recall_84: 0.3699\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5380 - categorical_accuracy: 0.7274 - recall_84: 0.3538 - val_loss: 0.5459 - val_categorical_accuracy: 0.7344 - val_recall_84: 0.4110\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5328 - categorical_accuracy: 0.7257 - recall_84: 0.4000 - val_loss: 0.5394 - val_categorical_accuracy: 0.7344 - val_recall_84: 0.4384\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5253 - categorical_accuracy: 0.7326 - recall_84: 0.3744 - val_loss: 0.5397 - val_categorical_accuracy: 0.7396 - val_recall_84: 0.3973\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 28ms/step - loss: 0.5204 - categorical_accuracy: 0.7326 - recall_84: 0.3744 - val_loss: 0.5302 - val_categorical_accuracy: 0.7396 - val_recall_84: 0.4521\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5150 - categorical_accuracy: 0.7292 - recall_84: 0.3692 - val_loss: 0.5281 - val_categorical_accuracy: 0.7344 - val_recall_84: 0.4384\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5128 - categorical_accuracy: 0.7517 - recall_84: 0.4872 - val_loss: 0.5244 - val_categorical_accuracy: 0.7448 - val_recall_84: 0.4521\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5053 - categorical_accuracy: 0.7448 - recall_84: 0.4256 - val_loss: 0.5185 - val_categorical_accuracy: 0.7500 - val_recall_84: 0.4932\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4990 - categorical_accuracy: 0.7483 - recall_84: 0.4308 - val_loss: 0.5192 - val_categorical_accuracy: 0.7500 - val_recall_84: 0.4521\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4968 - categorical_accuracy: 0.7431 - recall_84: 0.4615 - val_loss: 0.5144 - val_categorical_accuracy: 0.7656 - val_recall_84: 0.5068\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4906 - categorical_accuracy: 0.7552 - recall_84: 0.4359 - val_loss: 0.5120 - val_categorical_accuracy: 0.7448 - val_recall_84: 0.4932\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4867 - categorical_accuracy: 0.7500 - recall_84: 0.4821 - val_loss: 0.5081 - val_categorical_accuracy: 0.7708 - val_recall_84: 0.5616\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4822 - categorical_accuracy: 0.7569 - recall_84: 0.4974 - val_loss: 0.5109 - val_categorical_accuracy: 0.7500 - val_recall_84: 0.4795\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 1s 42ms/step - loss: 0.4788 - categorical_accuracy: 0.7587 - recall_84: 0.4718 - val_loss: 0.5069 - val_categorical_accuracy: 0.7656 - val_recall_84: 0.5205\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4747 - categorical_accuracy: 0.7691 - recall_84: 0.5026 - val_loss: 0.5048 - val_categorical_accuracy: 0.7812 - val_recall_84: 0.5616\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4709 - categorical_accuracy: 0.7726 - recall_84: 0.5179 - val_loss: 0.5034 - val_categorical_accuracy: 0.7865 - val_recall_84: 0.5753\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4683 - categorical_accuracy: 0.7830 - recall_84: 0.5692 - val_loss: 0.5074 - val_categorical_accuracy: 0.7708 - val_recall_84: 0.5205\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4677 - categorical_accuracy: 0.7726 - recall_84: 0.4923 - val_loss: 0.4999 - val_categorical_accuracy: 0.8021 - val_recall_84: 0.6301\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4620 - categorical_accuracy: 0.7778 - recall_84: 0.5436 - val_loss: 0.5076 - val_categorical_accuracy: 0.7708 - val_recall_84: 0.5205\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4596 - categorical_accuracy: 0.7656 - recall_84: 0.4718 - val_loss: 0.4994 - val_categorical_accuracy: 0.7969 - val_recall_84: 0.6301\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4600 - categorical_accuracy: 0.7743 - recall_84: 0.6051 - val_loss: 0.5067 - val_categorical_accuracy: 0.7812 - val_recall_84: 0.5479\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4556 - categorical_accuracy: 0.7865 - recall_84: 0.5179 - val_loss: 0.4986 - val_categorical_accuracy: 0.7917 - val_recall_84: 0.6301\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4539 - categorical_accuracy: 0.7865 - recall_84: 0.5692 - val_loss: 0.5019 - val_categorical_accuracy: 0.7969 - val_recall_84: 0.6027\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 28ms/step - loss: 0.4528 - categorical_accuracy: 0.7812 - recall_84: 0.6051 - val_loss: 0.5037 - val_categorical_accuracy: 0.7969 - val_recall_84: 0.6027\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4478 - categorical_accuracy: 0.7882 - recall_84: 0.5744 - val_loss: 0.5001 - val_categorical_accuracy: 0.7969 - val_recall_84: 0.6301\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4456 - categorical_accuracy: 0.7865 - recall_84: 0.5385 - val_loss: 0.5010 - val_categorical_accuracy: 0.7969 - val_recall_84: 0.6301\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4446 - categorical_accuracy: 0.7812 - recall_84: 0.5949 - val_loss: 0.5072 - val_categorical_accuracy: 0.7865 - val_recall_84: 0.5753\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 21ms/step - loss: 0.4480 - categorical_accuracy: 0.7795 - recall_84: 0.5487 - val_loss: 0.5006 - val_categorical_accuracy: 0.7865 - val_recall_84: 0.6575\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.4412 - categorical_accuracy: 0.7951 - recall_84: 0.5795 - val_loss: 0.5065 - val_categorical_accuracy: 0.7969 - val_recall_84: 0.6301\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4395 - categorical_accuracy: 0.7847 - recall_84: 0.5641 - val_loss: 0.5042 - val_categorical_accuracy: 0.7969 - val_recall_84: 0.6438\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4451 - categorical_accuracy: 0.7882 - recall_84: 0.6000 - val_loss: 0.5187 - val_categorical_accuracy: 0.7760 - val_recall_84: 0.5479\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4409 - categorical_accuracy: 0.7847 - recall_84: 0.5692 - val_loss: 0.5053 - val_categorical_accuracy: 0.7917 - val_recall_84: 0.6438\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 21ms/step - loss: 0.4379 - categorical_accuracy: 0.7865 - recall_84: 0.5538 - val_loss: 0.5086 - val_categorical_accuracy: 0.7917 - val_recall_84: 0.6301\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 44ms/step - loss: 0.4347 - categorical_accuracy: 0.7865 - recall_84: 0.6051 - val_loss: 0.5080 - val_categorical_accuracy: 0.7969 - val_recall_84: 0.6438\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.4345 - categorical_accuracy: 0.7812 - recall_84: 0.5641 - val_loss: 0.5123 - val_categorical_accuracy: 0.7865 - val_recall_84: 0.6164\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 5s 128ms/step - loss: 0.6640 - categorical_accuracy: 0.6615 - recall_85: 0.0000e+00 - val_loss: 0.6844 - val_categorical_accuracy: 0.6198 - val_recall_85: 0.0000e+00\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.6586 - categorical_accuracy: 0.6615 - recall_85: 0.0000e+00 - val_loss: 0.6789 - val_categorical_accuracy: 0.6198 - val_recall_85: 0.0000e+00\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 20ms/step - loss: 0.6543 - categorical_accuracy: 0.6615 - recall_85: 0.0000e+00 - val_loss: 0.6742 - val_categorical_accuracy: 0.6198 - val_recall_85: 0.0000e+00\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 1s 48ms/step - loss: 0.6489 - categorical_accuracy: 0.6615 - recall_85: 0.0000e+00 - val_loss: 0.6698 - val_categorical_accuracy: 0.6198 - val_recall_85: 0.0000e+00\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.6437 - categorical_accuracy: 0.6615 - recall_85: 0.0000e+00 - val_loss: 0.6633 - val_categorical_accuracy: 0.6198 - val_recall_85: 0.0000e+00\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6380 - categorical_accuracy: 0.6615 - recall_85: 0.0000e+00 - val_loss: 0.6544 - val_categorical_accuracy: 0.6198 - val_recall_85: 0.0000e+00\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6309 - categorical_accuracy: 0.6615 - recall_85: 0.0000e+00 - val_loss: 0.6476 - val_categorical_accuracy: 0.6198 - val_recall_85: 0.0000e+00\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.6245 - categorical_accuracy: 0.6615 - recall_85: 0.0000e+00 - val_loss: 0.6431 - val_categorical_accuracy: 0.6198 - val_recall_85: 0.0000e+00\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6181 - categorical_accuracy: 0.6615 - recall_85: 0.0000e+00 - val_loss: 0.6302 - val_categorical_accuracy: 0.6198 - val_recall_85: 0.0000e+00\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 23ms/step - loss: 0.6094 - categorical_accuracy: 0.6632 - recall_85: 0.0154 - val_loss: 0.6234 - val_categorical_accuracy: 0.6198 - val_recall_85: 0.0000e+00\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 27ms/step - loss: 0.6020 - categorical_accuracy: 0.6649 - recall_85: 0.0154 - val_loss: 0.6169 - val_categorical_accuracy: 0.6302 - val_recall_85: 0.0274\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 35ms/step - loss: 0.5941 - categorical_accuracy: 0.6788 - recall_85: 0.0667 - val_loss: 0.6052 - val_categorical_accuracy: 0.6667 - val_recall_85: 0.1370\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 24ms/step - loss: 0.5846 - categorical_accuracy: 0.6788 - recall_85: 0.0718 - val_loss: 0.6007 - val_categorical_accuracy: 0.6510 - val_recall_85: 0.0959\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.5757 - categorical_accuracy: 0.6910 - recall_85: 0.1538 - val_loss: 0.5869 - val_categorical_accuracy: 0.7135 - val_recall_85: 0.2877\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5663 - categorical_accuracy: 0.6979 - recall_85: 0.2000 - val_loss: 0.5824 - val_categorical_accuracy: 0.6771 - val_recall_85: 0.1918\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.5565 - categorical_accuracy: 0.6979 - recall_85: 0.1949 - val_loss: 0.5709 - val_categorical_accuracy: 0.7188 - val_recall_85: 0.3014\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5474 - categorical_accuracy: 0.7274 - recall_85: 0.3179 - val_loss: 0.5609 - val_categorical_accuracy: 0.7240 - val_recall_85: 0.3699\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5383 - categorical_accuracy: 0.7274 - recall_85: 0.3026 - val_loss: 0.5537 - val_categorical_accuracy: 0.7292 - val_recall_85: 0.3425\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5298 - categorical_accuracy: 0.7465 - recall_85: 0.4051 - val_loss: 0.5453 - val_categorical_accuracy: 0.7292 - val_recall_85: 0.3973\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.5212 - categorical_accuracy: 0.7378 - recall_85: 0.3333 - val_loss: 0.5360 - val_categorical_accuracy: 0.7344 - val_recall_85: 0.4247\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 34ms/step - loss: 0.5131 - categorical_accuracy: 0.7587 - recall_85: 0.4410 - val_loss: 0.5276 - val_categorical_accuracy: 0.7708 - val_recall_85: 0.5479\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 26ms/step - loss: 0.5060 - categorical_accuracy: 0.7604 - recall_85: 0.4667 - val_loss: 0.5221 - val_categorical_accuracy: 0.7656 - val_recall_85: 0.5479\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 27ms/step - loss: 0.4987 - categorical_accuracy: 0.7743 - recall_85: 0.4821 - val_loss: 0.5157 - val_categorical_accuracy: 0.7656 - val_recall_85: 0.5479\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 1s 50ms/step - loss: 0.4920 - categorical_accuracy: 0.7812 - recall_85: 0.5333 - val_loss: 0.5123 - val_categorical_accuracy: 0.7708 - val_recall_85: 0.5616\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 37ms/step - loss: 0.4847 - categorical_accuracy: 0.7847 - recall_85: 0.5026 - val_loss: 0.5084 - val_categorical_accuracy: 0.7656 - val_recall_85: 0.5342\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 27ms/step - loss: 0.4838 - categorical_accuracy: 0.7726 - recall_85: 0.4769 - val_loss: 0.5032 - val_categorical_accuracy: 0.7812 - val_recall_85: 0.6301\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.4832 - categorical_accuracy: 0.7812 - recall_85: 0.5846 - val_loss: 0.5142 - val_categorical_accuracy: 0.7552 - val_recall_85: 0.4795\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 32ms/step - loss: 0.4728 - categorical_accuracy: 0.7778 - recall_85: 0.4769 - val_loss: 0.4973 - val_categorical_accuracy: 0.7812 - val_recall_85: 0.6164\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4736 - categorical_accuracy: 0.7847 - recall_85: 0.5795 - val_loss: 0.5006 - val_categorical_accuracy: 0.7604 - val_recall_85: 0.5068\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4639 - categorical_accuracy: 0.7917 - recall_85: 0.5333 - val_loss: 0.4930 - val_categorical_accuracy: 0.7812 - val_recall_85: 0.6301\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4656 - categorical_accuracy: 0.7760 - recall_85: 0.6513 - val_loss: 0.4975 - val_categorical_accuracy: 0.7656 - val_recall_85: 0.5205\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4739 - categorical_accuracy: 0.7743 - recall_85: 0.4256 - val_loss: 0.4939 - val_categorical_accuracy: 0.7760 - val_recall_85: 0.5890\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.4637 - categorical_accuracy: 0.7760 - recall_85: 0.6308 - val_loss: 0.4950 - val_categorical_accuracy: 0.7812 - val_recall_85: 0.5753\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 33ms/step - loss: 0.4583 - categorical_accuracy: 0.7969 - recall_85: 0.5282 - val_loss: 0.4906 - val_categorical_accuracy: 0.7656 - val_recall_85: 0.5890\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4552 - categorical_accuracy: 0.7917 - recall_85: 0.6051 - val_loss: 0.4942 - val_categorical_accuracy: 0.7812 - val_recall_85: 0.5753\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4501 - categorical_accuracy: 0.7934 - recall_85: 0.5744 - val_loss: 0.4895 - val_categorical_accuracy: 0.7708 - val_recall_85: 0.6027\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 26ms/step - loss: 0.4494 - categorical_accuracy: 0.7951 - recall_85: 0.6103 - val_loss: 0.4940 - val_categorical_accuracy: 0.7708 - val_recall_85: 0.5616\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.4473 - categorical_accuracy: 0.7934 - recall_85: 0.5692 - val_loss: 0.4900 - val_categorical_accuracy: 0.7656 - val_recall_85: 0.5890\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 25ms/step - loss: 0.4455 - categorical_accuracy: 0.7865 - recall_85: 0.6103 - val_loss: 0.4903 - val_categorical_accuracy: 0.7708 - val_recall_85: 0.5753\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 21ms/step - loss: 0.4452 - categorical_accuracy: 0.7934 - recall_85: 0.5795 - val_loss: 0.4885 - val_categorical_accuracy: 0.7917 - val_recall_85: 0.6712\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4426 - categorical_accuracy: 0.7882 - recall_85: 0.6000 - val_loss: 0.4927 - val_categorical_accuracy: 0.7760 - val_recall_85: 0.5753\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 1s 54ms/step - loss: 0.4416 - categorical_accuracy: 0.7899 - recall_85: 0.5692 - val_loss: 0.4901 - val_categorical_accuracy: 0.7708 - val_recall_85: 0.6027\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 35ms/step - loss: 0.4444 - categorical_accuracy: 0.8003 - recall_85: 0.6974 - val_loss: 0.4934 - val_categorical_accuracy: 0.7708 - val_recall_85: 0.5753\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 26ms/step - loss: 0.4467 - categorical_accuracy: 0.7882 - recall_85: 0.5385 - val_loss: 0.4914 - val_categorical_accuracy: 0.7812 - val_recall_85: 0.6301\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 23ms/step - loss: 0.4433 - categorical_accuracy: 0.7865 - recall_85: 0.6308 - val_loss: 0.4920 - val_categorical_accuracy: 0.7604 - val_recall_85: 0.5753\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 23ms/step - loss: 0.4374 - categorical_accuracy: 0.7951 - recall_85: 0.6154 - val_loss: 0.4912 - val_categorical_accuracy: 0.7604 - val_recall_85: 0.5890\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 39ms/step - loss: 0.4383 - categorical_accuracy: 0.7865 - recall_85: 0.5641 - val_loss: 0.4920 - val_categorical_accuracy: 0.7604 - val_recall_85: 0.5753\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 26ms/step - loss: 0.4379 - categorical_accuracy: 0.7986 - recall_85: 0.6718 - val_loss: 0.4900 - val_categorical_accuracy: 0.7865 - val_recall_85: 0.6575\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4545 - categorical_accuracy: 0.7847 - recall_85: 0.5077 - val_loss: 0.4938 - val_categorical_accuracy: 0.7812 - val_recall_85: 0.6164\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.4366 - categorical_accuracy: 0.7969 - recall_85: 0.6821 - val_loss: 0.4907 - val_categorical_accuracy: 0.7812 - val_recall_85: 0.6575\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 5s 113ms/step - loss: 0.6915 - categorical_accuracy: 0.5434 - recall_86: 0.6308 - val_loss: 0.6816 - val_categorical_accuracy: 0.6510 - val_recall_86: 0.1781\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6762 - categorical_accuracy: 0.6736 - recall_86: 0.0821 - val_loss: 0.6736 - val_categorical_accuracy: 0.6406 - val_recall_86: 0.0548\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.6654 - categorical_accuracy: 0.6649 - recall_86: 0.0205 - val_loss: 0.6667 - val_categorical_accuracy: 0.6250 - val_recall_86: 0.0137\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6567 - categorical_accuracy: 0.6615 - recall_86: 0.0000e+00 - val_loss: 0.6611 - val_categorical_accuracy: 0.6198 - val_recall_86: 0.0000e+00\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6481 - categorical_accuracy: 0.6615 - recall_86: 0.0000e+00 - val_loss: 0.6537 - val_categorical_accuracy: 0.6198 - val_recall_86: 0.0000e+00\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.6393 - categorical_accuracy: 0.6615 - recall_86: 0.0000e+00 - val_loss: 0.6467 - val_categorical_accuracy: 0.6198 - val_recall_86: 0.0000e+00\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6326 - categorical_accuracy: 0.6615 - recall_86: 0.0000e+00 - val_loss: 0.6401 - val_categorical_accuracy: 0.6198 - val_recall_86: 0.0000e+00\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 38ms/step - loss: 0.6248 - categorical_accuracy: 0.6615 - recall_86: 0.0000e+00 - val_loss: 0.6310 - val_categorical_accuracy: 0.6302 - val_recall_86: 0.0274\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.6179 - categorical_accuracy: 0.6649 - recall_86: 0.0205 - val_loss: 0.6223 - val_categorical_accuracy: 0.6406 - val_recall_86: 0.0548\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6116 - categorical_accuracy: 0.6771 - recall_86: 0.0667 - val_loss: 0.6132 - val_categorical_accuracy: 0.6615 - val_recall_86: 0.1096\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.6047 - categorical_accuracy: 0.6771 - recall_86: 0.0821 - val_loss: 0.6059 - val_categorical_accuracy: 0.6615 - val_recall_86: 0.1096\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5980 - categorical_accuracy: 0.6858 - recall_86: 0.1179 - val_loss: 0.5988 - val_categorical_accuracy: 0.6719 - val_recall_86: 0.1507\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5923 - categorical_accuracy: 0.6910 - recall_86: 0.1385 - val_loss: 0.5909 - val_categorical_accuracy: 0.7031 - val_recall_86: 0.2329\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5853 - categorical_accuracy: 0.6962 - recall_86: 0.1692 - val_loss: 0.5845 - val_categorical_accuracy: 0.7031 - val_recall_86: 0.2466\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 21ms/step - loss: 0.5791 - categorical_accuracy: 0.7014 - recall_86: 0.1846 - val_loss: 0.5773 - val_categorical_accuracy: 0.7031 - val_recall_86: 0.2466\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5731 - categorical_accuracy: 0.7101 - recall_86: 0.2615 - val_loss: 0.5692 - val_categorical_accuracy: 0.7240 - val_recall_86: 0.3288\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.5687 - categorical_accuracy: 0.7188 - recall_86: 0.3385 - val_loss: 0.5618 - val_categorical_accuracy: 0.7344 - val_recall_86: 0.3836\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5617 - categorical_accuracy: 0.7170 - recall_86: 0.2923 - val_loss: 0.5607 - val_categorical_accuracy: 0.7240 - val_recall_86: 0.3288\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.5576 - categorical_accuracy: 0.7188 - recall_86: 0.2769 - val_loss: 0.5520 - val_categorical_accuracy: 0.7396 - val_recall_86: 0.3973\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5537 - categorical_accuracy: 0.7205 - recall_86: 0.4051 - val_loss: 0.5459 - val_categorical_accuracy: 0.7552 - val_recall_86: 0.4658\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5462 - categorical_accuracy: 0.7205 - recall_86: 0.3487 - val_loss: 0.5464 - val_categorical_accuracy: 0.7344 - val_recall_86: 0.3699\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5433 - categorical_accuracy: 0.7309 - recall_86: 0.3487 - val_loss: 0.5396 - val_categorical_accuracy: 0.7500 - val_recall_86: 0.4384\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5384 - categorical_accuracy: 0.7240 - recall_86: 0.3795 - val_loss: 0.5347 - val_categorical_accuracy: 0.7760 - val_recall_86: 0.5068\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 22ms/step - loss: 0.5343 - categorical_accuracy: 0.7222 - recall_86: 0.3487 - val_loss: 0.5343 - val_categorical_accuracy: 0.7500 - val_recall_86: 0.4384\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5306 - categorical_accuracy: 0.7257 - recall_86: 0.3744 - val_loss: 0.5289 - val_categorical_accuracy: 0.7708 - val_recall_86: 0.4932\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5263 - categorical_accuracy: 0.7222 - recall_86: 0.3846 - val_loss: 0.5242 - val_categorical_accuracy: 0.7812 - val_recall_86: 0.5616\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5229 - categorical_accuracy: 0.7344 - recall_86: 0.4154 - val_loss: 0.5236 - val_categorical_accuracy: 0.7656 - val_recall_86: 0.4932\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5190 - categorical_accuracy: 0.7431 - recall_86: 0.4359 - val_loss: 0.5194 - val_categorical_accuracy: 0.7865 - val_recall_86: 0.5479\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5162 - categorical_accuracy: 0.7431 - recall_86: 0.4000 - val_loss: 0.5176 - val_categorical_accuracy: 0.7917 - val_recall_86: 0.5479\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5127 - categorical_accuracy: 0.7535 - recall_86: 0.4564 - val_loss: 0.5135 - val_categorical_accuracy: 0.7760 - val_recall_86: 0.5616\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5094 - categorical_accuracy: 0.7604 - recall_86: 0.4564 - val_loss: 0.5129 - val_categorical_accuracy: 0.7812 - val_recall_86: 0.5479\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 23ms/step - loss: 0.5062 - categorical_accuracy: 0.7587 - recall_86: 0.4872 - val_loss: 0.5092 - val_categorical_accuracy: 0.7812 - val_recall_86: 0.5616\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 24ms/step - loss: 0.5027 - categorical_accuracy: 0.7656 - recall_86: 0.4872 - val_loss: 0.5076 - val_categorical_accuracy: 0.7865 - val_recall_86: 0.5616\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5003 - categorical_accuracy: 0.7726 - recall_86: 0.4872 - val_loss: 0.5062 - val_categorical_accuracy: 0.7812 - val_recall_86: 0.5479\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4959 - categorical_accuracy: 0.7743 - recall_86: 0.4974 - val_loss: 0.5038 - val_categorical_accuracy: 0.7812 - val_recall_86: 0.5616\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4931 - categorical_accuracy: 0.7708 - recall_86: 0.5077 - val_loss: 0.5021 - val_categorical_accuracy: 0.7812 - val_recall_86: 0.5616\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4921 - categorical_accuracy: 0.7708 - recall_86: 0.5385 - val_loss: 0.4987 - val_categorical_accuracy: 0.7917 - val_recall_86: 0.5890\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4876 - categorical_accuracy: 0.7795 - recall_86: 0.5077 - val_loss: 0.5009 - val_categorical_accuracy: 0.7760 - val_recall_86: 0.5479\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4853 - categorical_accuracy: 0.7795 - recall_86: 0.4974 - val_loss: 0.4959 - val_categorical_accuracy: 0.8021 - val_recall_86: 0.6164\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.4834 - categorical_accuracy: 0.7743 - recall_86: 0.5436 - val_loss: 0.4958 - val_categorical_accuracy: 0.8021 - val_recall_86: 0.6164\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4805 - categorical_accuracy: 0.7726 - recall_86: 0.5333 - val_loss: 0.4943 - val_categorical_accuracy: 0.7969 - val_recall_86: 0.6027\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4780 - categorical_accuracy: 0.7778 - recall_86: 0.5179 - val_loss: 0.4932 - val_categorical_accuracy: 0.7969 - val_recall_86: 0.6027\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4766 - categorical_accuracy: 0.7812 - recall_86: 0.5179 - val_loss: 0.4918 - val_categorical_accuracy: 0.7969 - val_recall_86: 0.6164\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 24ms/step - loss: 0.4736 - categorical_accuracy: 0.7760 - recall_86: 0.5436 - val_loss: 0.4913 - val_categorical_accuracy: 0.7969 - val_recall_86: 0.6164\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4713 - categorical_accuracy: 0.7795 - recall_86: 0.5231 - val_loss: 0.4917 - val_categorical_accuracy: 0.7812 - val_recall_86: 0.5753\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4714 - categorical_accuracy: 0.7760 - recall_86: 0.5590 - val_loss: 0.4886 - val_categorical_accuracy: 0.7969 - val_recall_86: 0.6301\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4711 - categorical_accuracy: 0.7830 - recall_86: 0.4923 - val_loss: 0.4913 - val_categorical_accuracy: 0.7812 - val_recall_86: 0.5753\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4720 - categorical_accuracy: 0.7760 - recall_86: 0.5846 - val_loss: 0.4869 - val_categorical_accuracy: 0.8021 - val_recall_86: 0.6849\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4640 - categorical_accuracy: 0.7847 - recall_86: 0.5385 - val_loss: 0.4925 - val_categorical_accuracy: 0.7812 - val_recall_86: 0.5616\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4637 - categorical_accuracy: 0.7812 - recall_86: 0.5077 - val_loss: 0.4868 - val_categorical_accuracy: 0.7917 - val_recall_86: 0.6027\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 5s 117ms/step - loss: 0.6661 - categorical_accuracy: 0.6615 - recall_87: 0.0000e+00 - val_loss: 0.6701 - val_categorical_accuracy: 0.6198 - val_recall_87: 0.0000e+00\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6550 - categorical_accuracy: 0.6615 - recall_87: 0.0000e+00 - val_loss: 0.6684 - val_categorical_accuracy: 0.6198 - val_recall_87: 0.0000e+00\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6499 - categorical_accuracy: 0.6615 - recall_87: 0.0000e+00 - val_loss: 0.6689 - val_categorical_accuracy: 0.6198 - val_recall_87: 0.0000e+00\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6475 - categorical_accuracy: 0.6615 - recall_87: 0.0000e+00 - val_loss: 0.6690 - val_categorical_accuracy: 0.6198 - val_recall_87: 0.0000e+00\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6455 - categorical_accuracy: 0.6615 - recall_87: 0.0000e+00 - val_loss: 0.6668 - val_categorical_accuracy: 0.6198 - val_recall_87: 0.0000e+00\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6433 - categorical_accuracy: 0.6615 - recall_87: 0.0000e+00 - val_loss: 0.6661 - val_categorical_accuracy: 0.6198 - val_recall_87: 0.0000e+00\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6417 - categorical_accuracy: 0.6615 - recall_87: 0.0000e+00 - val_loss: 0.6636 - val_categorical_accuracy: 0.6198 - val_recall_87: 0.0000e+00\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 20ms/step - loss: 0.6400 - categorical_accuracy: 0.6615 - recall_87: 0.0000e+00 - val_loss: 0.6624 - val_categorical_accuracy: 0.6198 - val_recall_87: 0.0000e+00\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6377 - categorical_accuracy: 0.6615 - recall_87: 0.0000e+00 - val_loss: 0.6595 - val_categorical_accuracy: 0.6198 - val_recall_87: 0.0000e+00\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.6348 - categorical_accuracy: 0.6615 - recall_87: 0.0000e+00 - val_loss: 0.6558 - val_categorical_accuracy: 0.6198 - val_recall_87: 0.0000e+00\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.6311 - categorical_accuracy: 0.6615 - recall_87: 0.0000e+00 - val_loss: 0.6508 - val_categorical_accuracy: 0.6198 - val_recall_87: 0.0000e+00\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6276 - categorical_accuracy: 0.6615 - recall_87: 0.0000e+00 - val_loss: 0.6473 - val_categorical_accuracy: 0.6198 - val_recall_87: 0.0000e+00\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.6236 - categorical_accuracy: 0.6615 - recall_87: 0.0000e+00 - val_loss: 0.6421 - val_categorical_accuracy: 0.6198 - val_recall_87: 0.0000e+00\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6200 - categorical_accuracy: 0.6615 - recall_87: 0.0000e+00 - val_loss: 0.6367 - val_categorical_accuracy: 0.6198 - val_recall_87: 0.0000e+00\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 23ms/step - loss: 0.6153 - categorical_accuracy: 0.6615 - recall_87: 0.0000e+00 - val_loss: 0.6333 - val_categorical_accuracy: 0.6198 - val_recall_87: 0.0000e+00\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.6098 - categorical_accuracy: 0.6615 - recall_87: 0.0000e+00 - val_loss: 0.6260 - val_categorical_accuracy: 0.6198 - val_recall_87: 0.0000e+00\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6046 - categorical_accuracy: 0.6615 - recall_87: 0.0000e+00 - val_loss: 0.6196 - val_categorical_accuracy: 0.6198 - val_recall_87: 0.0000e+00\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.6001 - categorical_accuracy: 0.6580 - recall_87: 0.0000e+00 - val_loss: 0.6122 - val_categorical_accuracy: 0.6250 - val_recall_87: 0.0137\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5931 - categorical_accuracy: 0.6649 - recall_87: 0.0205 - val_loss: 0.6059 - val_categorical_accuracy: 0.6354 - val_recall_87: 0.0411\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5867 - categorical_accuracy: 0.6753 - recall_87: 0.0513 - val_loss: 0.5994 - val_categorical_accuracy: 0.6562 - val_recall_87: 0.0959\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5809 - categorical_accuracy: 0.6875 - recall_87: 0.1077 - val_loss: 0.5907 - val_categorical_accuracy: 0.6615 - val_recall_87: 0.1507\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5738 - categorical_accuracy: 0.6944 - recall_87: 0.1333 - val_loss: 0.5849 - val_categorical_accuracy: 0.6562 - val_recall_87: 0.1370\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5673 - categorical_accuracy: 0.6997 - recall_87: 0.1590 - val_loss: 0.5760 - val_categorical_accuracy: 0.7083 - val_recall_87: 0.2740\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5597 - categorical_accuracy: 0.7205 - recall_87: 0.2308 - val_loss: 0.5694 - val_categorical_accuracy: 0.7240 - val_recall_87: 0.3151\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5532 - categorical_accuracy: 0.7361 - recall_87: 0.3231 - val_loss: 0.5626 - val_categorical_accuracy: 0.7448 - val_recall_87: 0.3836\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5440 - categorical_accuracy: 0.7344 - recall_87: 0.2872 - val_loss: 0.5594 - val_categorical_accuracy: 0.7188 - val_recall_87: 0.3151\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5352 - categorical_accuracy: 0.7309 - recall_87: 0.2872 - val_loss: 0.5498 - val_categorical_accuracy: 0.7292 - val_recall_87: 0.3699\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5279 - categorical_accuracy: 0.7448 - recall_87: 0.4051 - val_loss: 0.5403 - val_categorical_accuracy: 0.7552 - val_recall_87: 0.4384\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5188 - categorical_accuracy: 0.7344 - recall_87: 0.3282 - val_loss: 0.5372 - val_categorical_accuracy: 0.7500 - val_recall_87: 0.4247\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5115 - categorical_accuracy: 0.7448 - recall_87: 0.4103 - val_loss: 0.5284 - val_categorical_accuracy: 0.7604 - val_recall_87: 0.4658\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5033 - categorical_accuracy: 0.7396 - recall_87: 0.4154 - val_loss: 0.5264 - val_categorical_accuracy: 0.7656 - val_recall_87: 0.4658\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4968 - categorical_accuracy: 0.7500 - recall_87: 0.4564 - val_loss: 0.5181 - val_categorical_accuracy: 0.7708 - val_recall_87: 0.5068\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 25ms/step - loss: 0.4904 - categorical_accuracy: 0.7656 - recall_87: 0.5231 - val_loss: 0.5182 - val_categorical_accuracy: 0.7708 - val_recall_87: 0.4932\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4856 - categorical_accuracy: 0.7622 - recall_87: 0.4615 - val_loss: 0.5091 - val_categorical_accuracy: 0.7760 - val_recall_87: 0.5479\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4831 - categorical_accuracy: 0.7726 - recall_87: 0.5949 - val_loss: 0.5077 - val_categorical_accuracy: 0.7865 - val_recall_87: 0.5753\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4748 - categorical_accuracy: 0.7726 - recall_87: 0.5179 - val_loss: 0.5067 - val_categorical_accuracy: 0.7865 - val_recall_87: 0.5479\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4698 - categorical_accuracy: 0.7778 - recall_87: 0.5590 - val_loss: 0.5046 - val_categorical_accuracy: 0.7865 - val_recall_87: 0.5616\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4647 - categorical_accuracy: 0.7812 - recall_87: 0.5590 - val_loss: 0.5022 - val_categorical_accuracy: 0.8021 - val_recall_87: 0.6301\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4608 - categorical_accuracy: 0.7743 - recall_87: 0.5282 - val_loss: 0.5010 - val_categorical_accuracy: 0.7969 - val_recall_87: 0.6301\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 22ms/step - loss: 0.4577 - categorical_accuracy: 0.7795 - recall_87: 0.5949 - val_loss: 0.4987 - val_categorical_accuracy: 0.7969 - val_recall_87: 0.6712\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4568 - categorical_accuracy: 0.7812 - recall_87: 0.5744 - val_loss: 0.4981 - val_categorical_accuracy: 0.7969 - val_recall_87: 0.6712\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4513 - categorical_accuracy: 0.7847 - recall_87: 0.5897 - val_loss: 0.5006 - val_categorical_accuracy: 0.8125 - val_recall_87: 0.6712\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4519 - categorical_accuracy: 0.7760 - recall_87: 0.5590 - val_loss: 0.4980 - val_categorical_accuracy: 0.7969 - val_recall_87: 0.6849\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.4477 - categorical_accuracy: 0.7812 - recall_87: 0.5949 - val_loss: 0.5042 - val_categorical_accuracy: 0.8073 - val_recall_87: 0.6575\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4460 - categorical_accuracy: 0.7934 - recall_87: 0.6359 - val_loss: 0.4995 - val_categorical_accuracy: 0.7865 - val_recall_87: 0.6712\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4455 - categorical_accuracy: 0.7865 - recall_87: 0.5744 - val_loss: 0.5032 - val_categorical_accuracy: 0.8021 - val_recall_87: 0.6575\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4473 - categorical_accuracy: 0.7934 - recall_87: 0.6615 - val_loss: 0.5023 - val_categorical_accuracy: 0.7917 - val_recall_87: 0.6575\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 24ms/step - loss: 0.4425 - categorical_accuracy: 0.7951 - recall_87: 0.6462 - val_loss: 0.5025 - val_categorical_accuracy: 0.7969 - val_recall_87: 0.6712\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4434 - categorical_accuracy: 0.7951 - recall_87: 0.5795 - val_loss: 0.5018 - val_categorical_accuracy: 0.8021 - val_recall_87: 0.6849\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4387 - categorical_accuracy: 0.7934 - recall_87: 0.6564 - val_loss: 0.5027 - val_categorical_accuracy: 0.7917 - val_recall_87: 0.6575\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 5s 107ms/step - loss: 0.6799 - categorical_accuracy: 0.6580 - recall_88: 0.0000e+00 - val_loss: 0.6850 - val_categorical_accuracy: 0.6198 - val_recall_88: 0.0000e+00\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6693 - categorical_accuracy: 0.6615 - recall_88: 0.0000e+00 - val_loss: 0.6807 - val_categorical_accuracy: 0.6198 - val_recall_88: 0.0000e+00\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6614 - categorical_accuracy: 0.6615 - recall_88: 0.0000e+00 - val_loss: 0.6773 - val_categorical_accuracy: 0.6198 - val_recall_88: 0.0000e+00\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6558 - categorical_accuracy: 0.6615 - recall_88: 0.0000e+00 - val_loss: 0.6737 - val_categorical_accuracy: 0.6198 - val_recall_88: 0.0000e+00\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.6476 - categorical_accuracy: 0.6615 - recall_88: 0.0000e+00 - val_loss: 0.6645 - val_categorical_accuracy: 0.6198 - val_recall_88: 0.0000e+00\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.6395 - categorical_accuracy: 0.6615 - recall_88: 0.0000e+00 - val_loss: 0.6558 - val_categorical_accuracy: 0.6198 - val_recall_88: 0.0000e+00\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 25ms/step - loss: 0.6313 - categorical_accuracy: 0.6615 - recall_88: 0.0000e+00 - val_loss: 0.6462 - val_categorical_accuracy: 0.6198 - val_recall_88: 0.0000e+00\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6218 - categorical_accuracy: 0.6649 - recall_88: 0.0103 - val_loss: 0.6364 - val_categorical_accuracy: 0.6250 - val_recall_88: 0.0137\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6134 - categorical_accuracy: 0.6684 - recall_88: 0.0308 - val_loss: 0.6272 - val_categorical_accuracy: 0.6458 - val_recall_88: 0.0685\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6048 - categorical_accuracy: 0.6771 - recall_88: 0.0718 - val_loss: 0.6165 - val_categorical_accuracy: 0.6562 - val_recall_88: 0.0959\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5961 - categorical_accuracy: 0.6840 - recall_88: 0.1179 - val_loss: 0.6033 - val_categorical_accuracy: 0.6771 - val_recall_88: 0.1644\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5859 - categorical_accuracy: 0.6910 - recall_88: 0.1538 - val_loss: 0.5943 - val_categorical_accuracy: 0.6719 - val_recall_88: 0.1644\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5769 - categorical_accuracy: 0.6962 - recall_88: 0.1949 - val_loss: 0.5847 - val_categorical_accuracy: 0.7031 - val_recall_88: 0.2466\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 20ms/step - loss: 0.5665 - categorical_accuracy: 0.7083 - recall_88: 0.2359 - val_loss: 0.5737 - val_categorical_accuracy: 0.7031 - val_recall_88: 0.3014\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5596 - categorical_accuracy: 0.7188 - recall_88: 0.3026 - val_loss: 0.5667 - val_categorical_accuracy: 0.7188 - val_recall_88: 0.3288\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5491 - categorical_accuracy: 0.7309 - recall_88: 0.3487 - val_loss: 0.5576 - val_categorical_accuracy: 0.7448 - val_recall_88: 0.4247\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.5391 - categorical_accuracy: 0.7309 - recall_88: 0.3333 - val_loss: 0.5527 - val_categorical_accuracy: 0.7448 - val_recall_88: 0.4110\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5296 - categorical_accuracy: 0.7257 - recall_88: 0.3282 - val_loss: 0.5482 - val_categorical_accuracy: 0.7500 - val_recall_88: 0.4384\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5213 - categorical_accuracy: 0.7361 - recall_88: 0.3692 - val_loss: 0.5393 - val_categorical_accuracy: 0.7604 - val_recall_88: 0.4658\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5134 - categorical_accuracy: 0.7535 - recall_88: 0.4410 - val_loss: 0.5370 - val_categorical_accuracy: 0.7656 - val_recall_88: 0.4795\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 24ms/step - loss: 0.5056 - categorical_accuracy: 0.7639 - recall_88: 0.5077 - val_loss: 0.5325 - val_categorical_accuracy: 0.7604 - val_recall_88: 0.4795\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4997 - categorical_accuracy: 0.7500 - recall_88: 0.3949 - val_loss: 0.5297 - val_categorical_accuracy: 0.7604 - val_recall_88: 0.4795\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4927 - categorical_accuracy: 0.7639 - recall_88: 0.5077 - val_loss: 0.5263 - val_categorical_accuracy: 0.7656 - val_recall_88: 0.5068\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4864 - categorical_accuracy: 0.7674 - recall_88: 0.5282 - val_loss: 0.5228 - val_categorical_accuracy: 0.7760 - val_recall_88: 0.5342\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 21ms/step - loss: 0.4819 - categorical_accuracy: 0.7622 - recall_88: 0.4974 - val_loss: 0.5220 - val_categorical_accuracy: 0.7656 - val_recall_88: 0.5205\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4769 - categorical_accuracy: 0.7708 - recall_88: 0.5179 - val_loss: 0.5199 - val_categorical_accuracy: 0.7656 - val_recall_88: 0.5342\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 22ms/step - loss: 0.4745 - categorical_accuracy: 0.7656 - recall_88: 0.5128 - val_loss: 0.5192 - val_categorical_accuracy: 0.7865 - val_recall_88: 0.6027\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4691 - categorical_accuracy: 0.7708 - recall_88: 0.5641 - val_loss: 0.5217 - val_categorical_accuracy: 0.7656 - val_recall_88: 0.5342\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4660 - categorical_accuracy: 0.7743 - recall_88: 0.5282 - val_loss: 0.5210 - val_categorical_accuracy: 0.7760 - val_recall_88: 0.5753\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4664 - categorical_accuracy: 0.7639 - recall_88: 0.5795 - val_loss: 0.5226 - val_categorical_accuracy: 0.7552 - val_recall_88: 0.5068\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4646 - categorical_accuracy: 0.7812 - recall_88: 0.4974 - val_loss: 0.5172 - val_categorical_accuracy: 0.7865 - val_recall_88: 0.6164\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4576 - categorical_accuracy: 0.7691 - recall_88: 0.5692 - val_loss: 0.5185 - val_categorical_accuracy: 0.7865 - val_recall_88: 0.6027\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4553 - categorical_accuracy: 0.7743 - recall_88: 0.5692 - val_loss: 0.5246 - val_categorical_accuracy: 0.7656 - val_recall_88: 0.5479\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.4551 - categorical_accuracy: 0.7795 - recall_88: 0.5128 - val_loss: 0.5225 - val_categorical_accuracy: 0.7865 - val_recall_88: 0.6027\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4525 - categorical_accuracy: 0.7778 - recall_88: 0.5897 - val_loss: 0.5224 - val_categorical_accuracy: 0.7917 - val_recall_88: 0.6164\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4497 - categorical_accuracy: 0.7865 - recall_88: 0.5795 - val_loss: 0.5234 - val_categorical_accuracy: 0.7917 - val_recall_88: 0.6164\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4514 - categorical_accuracy: 0.7847 - recall_88: 0.5744 - val_loss: 0.5297 - val_categorical_accuracy: 0.7656 - val_recall_88: 0.5479\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4471 - categorical_accuracy: 0.7812 - recall_88: 0.5846 - val_loss: 0.5246 - val_categorical_accuracy: 0.7917 - val_recall_88: 0.6301\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 25ms/step - loss: 0.4460 - categorical_accuracy: 0.7830 - recall_88: 0.5744 - val_loss: 0.5246 - val_categorical_accuracy: 0.7917 - val_recall_88: 0.6301\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4470 - categorical_accuracy: 0.7795 - recall_88: 0.5744 - val_loss: 0.5307 - val_categorical_accuracy: 0.7812 - val_recall_88: 0.6027\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4478 - categorical_accuracy: 0.7795 - recall_88: 0.6205 - val_loss: 0.5286 - val_categorical_accuracy: 0.7865 - val_recall_88: 0.6164\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4451 - categorical_accuracy: 0.7778 - recall_88: 0.5333 - val_loss: 0.5361 - val_categorical_accuracy: 0.7812 - val_recall_88: 0.5890\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4468 - categorical_accuracy: 0.7812 - recall_88: 0.6256 - val_loss: 0.5321 - val_categorical_accuracy: 0.7865 - val_recall_88: 0.6164\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4421 - categorical_accuracy: 0.7830 - recall_88: 0.5436 - val_loss: 0.5341 - val_categorical_accuracy: 0.7865 - val_recall_88: 0.6164\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4412 - categorical_accuracy: 0.7882 - recall_88: 0.6205 - val_loss: 0.5314 - val_categorical_accuracy: 0.7760 - val_recall_88: 0.6164\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4408 - categorical_accuracy: 0.7795 - recall_88: 0.5538 - val_loss: 0.5344 - val_categorical_accuracy: 0.7865 - val_recall_88: 0.6164\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 31ms/step - loss: 0.4408 - categorical_accuracy: 0.7865 - recall_88: 0.6154 - val_loss: 0.5377 - val_categorical_accuracy: 0.7865 - val_recall_88: 0.6164\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4397 - categorical_accuracy: 0.7882 - recall_88: 0.6154 - val_loss: 0.5380 - val_categorical_accuracy: 0.7865 - val_recall_88: 0.6301\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4381 - categorical_accuracy: 0.7795 - recall_88: 0.5692 - val_loss: 0.5350 - val_categorical_accuracy: 0.7812 - val_recall_88: 0.6301\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4377 - categorical_accuracy: 0.7899 - recall_88: 0.6359 - val_loss: 0.5382 - val_categorical_accuracy: 0.7865 - val_recall_88: 0.6301\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 5s 98ms/step - loss: 0.6654 - categorical_accuracy: 0.6615 - recall_89: 0.0000e+00 - val_loss: 0.6872 - val_categorical_accuracy: 0.6198 - val_recall_89: 0.0000e+00\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 22ms/step - loss: 0.6611 - categorical_accuracy: 0.6615 - recall_89: 0.0000e+00 - val_loss: 0.6850 - val_categorical_accuracy: 0.6198 - val_recall_89: 0.0000e+00\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.6581 - categorical_accuracy: 0.6615 - recall_89: 0.0000e+00 - val_loss: 0.6821 - val_categorical_accuracy: 0.6198 - val_recall_89: 0.0000e+00\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6559 - categorical_accuracy: 0.6615 - recall_89: 0.0000e+00 - val_loss: 0.6804 - val_categorical_accuracy: 0.6198 - val_recall_89: 0.0000e+00\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6545 - categorical_accuracy: 0.6615 - recall_89: 0.0000e+00 - val_loss: 0.6800 - val_categorical_accuracy: 0.6198 - val_recall_89: 0.0000e+00\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.6525 - categorical_accuracy: 0.6615 - recall_89: 0.0000e+00 - val_loss: 0.6775 - val_categorical_accuracy: 0.6198 - val_recall_89: 0.0000e+00\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.6507 - categorical_accuracy: 0.6615 - recall_89: 0.0000e+00 - val_loss: 0.6752 - val_categorical_accuracy: 0.6198 - val_recall_89: 0.0000e+00\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 22ms/step - loss: 0.6491 - categorical_accuracy: 0.6615 - recall_89: 0.0000e+00 - val_loss: 0.6736 - val_categorical_accuracy: 0.6198 - val_recall_89: 0.0000e+00\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.6475 - categorical_accuracy: 0.6615 - recall_89: 0.0000e+00 - val_loss: 0.6721 - val_categorical_accuracy: 0.6198 - val_recall_89: 0.0000e+00\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6452 - categorical_accuracy: 0.6615 - recall_89: 0.0000e+00 - val_loss: 0.6692 - val_categorical_accuracy: 0.6198 - val_recall_89: 0.0000e+00\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.6431 - categorical_accuracy: 0.6615 - recall_89: 0.0000e+00 - val_loss: 0.6666 - val_categorical_accuracy: 0.6198 - val_recall_89: 0.0000e+00\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6409 - categorical_accuracy: 0.6615 - recall_89: 0.0000e+00 - val_loss: 0.6639 - val_categorical_accuracy: 0.6198 - val_recall_89: 0.0000e+00\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.6381 - categorical_accuracy: 0.6615 - recall_89: 0.0000e+00 - val_loss: 0.6610 - val_categorical_accuracy: 0.6198 - val_recall_89: 0.0000e+00\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 21ms/step - loss: 0.6358 - categorical_accuracy: 0.6615 - recall_89: 0.0000e+00 - val_loss: 0.6577 - val_categorical_accuracy: 0.6198 - val_recall_89: 0.0000e+00\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.6322 - categorical_accuracy: 0.6615 - recall_89: 0.0000e+00 - val_loss: 0.6533 - val_categorical_accuracy: 0.6198 - val_recall_89: 0.0000e+00\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6285 - categorical_accuracy: 0.6597 - recall_89: 0.0000e+00 - val_loss: 0.6491 - val_categorical_accuracy: 0.6250 - val_recall_89: 0.0137\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6242 - categorical_accuracy: 0.6597 - recall_89: 0.0000e+00 - val_loss: 0.6444 - val_categorical_accuracy: 0.6302 - val_recall_89: 0.0274\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.6194 - categorical_accuracy: 0.6597 - recall_89: 0.0000e+00 - val_loss: 0.6384 - val_categorical_accuracy: 0.6302 - val_recall_89: 0.0274\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6134 - categorical_accuracy: 0.6597 - recall_89: 0.0000e+00 - val_loss: 0.6325 - val_categorical_accuracy: 0.6302 - val_recall_89: 0.0274\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6053 - categorical_accuracy: 0.6615 - recall_89: 0.0051 - val_loss: 0.6233 - val_categorical_accuracy: 0.6406 - val_recall_89: 0.0548\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5977 - categorical_accuracy: 0.6667 - recall_89: 0.0256 - val_loss: 0.6151 - val_categorical_accuracy: 0.6458 - val_recall_89: 0.0959\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5901 - categorical_accuracy: 0.6771 - recall_89: 0.0667 - val_loss: 0.6069 - val_categorical_accuracy: 0.6562 - val_recall_89: 0.1233\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5839 - categorical_accuracy: 0.6927 - recall_89: 0.1179 - val_loss: 0.5997 - val_categorical_accuracy: 0.6562 - val_recall_89: 0.1644\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5758 - categorical_accuracy: 0.6979 - recall_89: 0.1385 - val_loss: 0.5916 - val_categorical_accuracy: 0.6771 - val_recall_89: 0.2192\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5690 - categorical_accuracy: 0.7049 - recall_89: 0.1897 - val_loss: 0.5849 - val_categorical_accuracy: 0.6823 - val_recall_89: 0.2329\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5620 - categorical_accuracy: 0.7135 - recall_89: 0.2256 - val_loss: 0.5783 - val_categorical_accuracy: 0.6875 - val_recall_89: 0.2466\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5561 - categorical_accuracy: 0.7135 - recall_89: 0.2308 - val_loss: 0.5722 - val_categorical_accuracy: 0.7031 - val_recall_89: 0.3014\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 25ms/step - loss: 0.5491 - categorical_accuracy: 0.7135 - recall_89: 0.2769 - val_loss: 0.5657 - val_categorical_accuracy: 0.7240 - val_recall_89: 0.3699\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5438 - categorical_accuracy: 0.7135 - recall_89: 0.2667 - val_loss: 0.5624 - val_categorical_accuracy: 0.7083 - val_recall_89: 0.3151\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5347 - categorical_accuracy: 0.7292 - recall_89: 0.3385 - val_loss: 0.5525 - val_categorical_accuracy: 0.7448 - val_recall_89: 0.4658\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5349 - categorical_accuracy: 0.7274 - recall_89: 0.4564 - val_loss: 0.5469 - val_categorical_accuracy: 0.7448 - val_recall_89: 0.4658\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5287 - categorical_accuracy: 0.7344 - recall_89: 0.3128 - val_loss: 0.5516 - val_categorical_accuracy: 0.7344 - val_recall_89: 0.3699\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5197 - categorical_accuracy: 0.7413 - recall_89: 0.3897 - val_loss: 0.5395 - val_categorical_accuracy: 0.7448 - val_recall_89: 0.4932\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5148 - categorical_accuracy: 0.7431 - recall_89: 0.4205 - val_loss: 0.5383 - val_categorical_accuracy: 0.7552 - val_recall_89: 0.4932\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 21ms/step - loss: 0.5106 - categorical_accuracy: 0.7448 - recall_89: 0.4513 - val_loss: 0.5325 - val_categorical_accuracy: 0.7552 - val_recall_89: 0.5068\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5103 - categorical_accuracy: 0.7535 - recall_89: 0.3949 - val_loss: 0.5329 - val_categorical_accuracy: 0.7500 - val_recall_89: 0.4795\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5050 - categorical_accuracy: 0.7517 - recall_89: 0.4974 - val_loss: 0.5269 - val_categorical_accuracy: 0.7708 - val_recall_89: 0.5479\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4974 - categorical_accuracy: 0.7639 - recall_89: 0.4923 - val_loss: 0.5265 - val_categorical_accuracy: 0.7448 - val_recall_89: 0.4658\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4953 - categorical_accuracy: 0.7604 - recall_89: 0.4821 - val_loss: 0.5218 - val_categorical_accuracy: 0.7708 - val_recall_89: 0.5616\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4898 - categorical_accuracy: 0.7656 - recall_89: 0.5128 - val_loss: 0.5214 - val_categorical_accuracy: 0.7604 - val_recall_89: 0.5068\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4868 - categorical_accuracy: 0.7726 - recall_89: 0.4923 - val_loss: 0.5200 - val_categorical_accuracy: 0.7604 - val_recall_89: 0.5068\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4837 - categorical_accuracy: 0.7726 - recall_89: 0.4974 - val_loss: 0.5178 - val_categorical_accuracy: 0.7760 - val_recall_89: 0.5479\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4804 - categorical_accuracy: 0.7726 - recall_89: 0.5333 - val_loss: 0.5137 - val_categorical_accuracy: 0.7760 - val_recall_89: 0.6027\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4791 - categorical_accuracy: 0.7708 - recall_89: 0.5179 - val_loss: 0.5138 - val_categorical_accuracy: 0.7708 - val_recall_89: 0.5616\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4751 - categorical_accuracy: 0.7726 - recall_89: 0.5333 - val_loss: 0.5121 - val_categorical_accuracy: 0.7708 - val_recall_89: 0.5753\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4726 - categorical_accuracy: 0.7778 - recall_89: 0.5436 - val_loss: 0.5108 - val_categorical_accuracy: 0.7708 - val_recall_89: 0.5753\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 20ms/step - loss: 0.4704 - categorical_accuracy: 0.7743 - recall_89: 0.5436 - val_loss: 0.5107 - val_categorical_accuracy: 0.7656 - val_recall_89: 0.5479\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4677 - categorical_accuracy: 0.7795 - recall_89: 0.5282 - val_loss: 0.5089 - val_categorical_accuracy: 0.7760 - val_recall_89: 0.5890\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4663 - categorical_accuracy: 0.7691 - recall_89: 0.5590 - val_loss: 0.5087 - val_categorical_accuracy: 0.7708 - val_recall_89: 0.5753\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4641 - categorical_accuracy: 0.7812 - recall_89: 0.5333 - val_loss: 0.5094 - val_categorical_accuracy: 0.7760 - val_recall_89: 0.5890\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 4s 103ms/step - loss: 0.6771 - categorical_accuracy: 0.6510 - recall_90: 0.0000e+00 - val_loss: 0.6845 - val_categorical_accuracy: 0.6198 - val_recall_90: 0.0000e+00\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 26ms/step - loss: 0.6603 - categorical_accuracy: 0.6615 - recall_90: 0.0000e+00 - val_loss: 0.6788 - val_categorical_accuracy: 0.6198 - val_recall_90: 0.0000e+00\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.6544 - categorical_accuracy: 0.6615 - recall_90: 0.0000e+00 - val_loss: 0.6746 - val_categorical_accuracy: 0.6198 - val_recall_90: 0.0000e+00\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.6477 - categorical_accuracy: 0.6615 - recall_90: 0.0000e+00 - val_loss: 0.6657 - val_categorical_accuracy: 0.6198 - val_recall_90: 0.0000e+00\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6417 - categorical_accuracy: 0.6615 - recall_90: 0.0000e+00 - val_loss: 0.6573 - val_categorical_accuracy: 0.6198 - val_recall_90: 0.0000e+00\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6355 - categorical_accuracy: 0.6615 - recall_90: 0.0000e+00 - val_loss: 0.6487 - val_categorical_accuracy: 0.6198 - val_recall_90: 0.0000e+00\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6279 - categorical_accuracy: 0.6615 - recall_90: 0.0000e+00 - val_loss: 0.6381 - val_categorical_accuracy: 0.6198 - val_recall_90: 0.0000e+00\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 25ms/step - loss: 0.6208 - categorical_accuracy: 0.6615 - recall_90: 0.0000e+00 - val_loss: 0.6333 - val_categorical_accuracy: 0.6250 - val_recall_90: 0.0137\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.6118 - categorical_accuracy: 0.6649 - recall_90: 0.0154 - val_loss: 0.6218 - val_categorical_accuracy: 0.6406 - val_recall_90: 0.0548\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.6055 - categorical_accuracy: 0.6701 - recall_90: 0.0308 - val_loss: 0.6144 - val_categorical_accuracy: 0.6406 - val_recall_90: 0.0548\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5954 - categorical_accuracy: 0.6806 - recall_90: 0.0769 - val_loss: 0.6022 - val_categorical_accuracy: 0.6771 - val_recall_90: 0.1781\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5890 - categorical_accuracy: 0.6788 - recall_90: 0.0923 - val_loss: 0.5954 - val_categorical_accuracy: 0.6771 - val_recall_90: 0.1781\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5803 - categorical_accuracy: 0.6962 - recall_90: 0.2154 - val_loss: 0.5824 - val_categorical_accuracy: 0.7135 - val_recall_90: 0.3014\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5718 - categorical_accuracy: 0.6979 - recall_90: 0.2410 - val_loss: 0.5743 - val_categorical_accuracy: 0.7188 - val_recall_90: 0.3014\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 22ms/step - loss: 0.5631 - categorical_accuracy: 0.7049 - recall_90: 0.2821 - val_loss: 0.5678 - val_categorical_accuracy: 0.7135 - val_recall_90: 0.2877\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5559 - categorical_accuracy: 0.7066 - recall_90: 0.2462 - val_loss: 0.5614 - val_categorical_accuracy: 0.7135 - val_recall_90: 0.2877\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5495 - categorical_accuracy: 0.7170 - recall_90: 0.3436 - val_loss: 0.5527 - val_categorical_accuracy: 0.7344 - val_recall_90: 0.3699\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.5411 - categorical_accuracy: 0.7222 - recall_90: 0.3128 - val_loss: 0.5449 - val_categorical_accuracy: 0.7500 - val_recall_90: 0.4247\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5334 - categorical_accuracy: 0.7344 - recall_90: 0.3795 - val_loss: 0.5431 - val_categorical_accuracy: 0.7396 - val_recall_90: 0.3836\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 24ms/step - loss: 0.5254 - categorical_accuracy: 0.7396 - recall_90: 0.3692 - val_loss: 0.5314 - val_categorical_accuracy: 0.7500 - val_recall_90: 0.4658\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5209 - categorical_accuracy: 0.7552 - recall_90: 0.4974 - val_loss: 0.5246 - val_categorical_accuracy: 0.7604 - val_recall_90: 0.4932\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5173 - categorical_accuracy: 0.7378 - recall_90: 0.3333 - val_loss: 0.5235 - val_categorical_accuracy: 0.7500 - val_recall_90: 0.4521\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5075 - categorical_accuracy: 0.7552 - recall_90: 0.5026 - val_loss: 0.5171 - val_categorical_accuracy: 0.7708 - val_recall_90: 0.5205\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5000 - categorical_accuracy: 0.7552 - recall_90: 0.4359 - val_loss: 0.5172 - val_categorical_accuracy: 0.7708 - val_recall_90: 0.5068\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.4958 - categorical_accuracy: 0.7656 - recall_90: 0.4821 - val_loss: 0.5130 - val_categorical_accuracy: 0.7708 - val_recall_90: 0.5068\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4892 - categorical_accuracy: 0.7674 - recall_90: 0.4410 - val_loss: 0.5062 - val_categorical_accuracy: 0.7865 - val_recall_90: 0.5479\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4841 - categorical_accuracy: 0.7708 - recall_90: 0.5077 - val_loss: 0.5022 - val_categorical_accuracy: 0.7969 - val_recall_90: 0.6164\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4788 - categorical_accuracy: 0.7882 - recall_90: 0.6000 - val_loss: 0.5048 - val_categorical_accuracy: 0.7656 - val_recall_90: 0.5068\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4752 - categorical_accuracy: 0.7760 - recall_90: 0.4359 - val_loss: 0.4976 - val_categorical_accuracy: 0.7917 - val_recall_90: 0.6164\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4742 - categorical_accuracy: 0.7795 - recall_90: 0.6103 - val_loss: 0.5062 - val_categorical_accuracy: 0.7552 - val_recall_90: 0.4795\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4645 - categorical_accuracy: 0.7812 - recall_90: 0.4923 - val_loss: 0.4958 - val_categorical_accuracy: 0.7917 - val_recall_90: 0.6164\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4630 - categorical_accuracy: 0.7778 - recall_90: 0.6051 - val_loss: 0.5091 - val_categorical_accuracy: 0.7604 - val_recall_90: 0.4932\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4653 - categorical_accuracy: 0.7760 - recall_90: 0.4410 - val_loss: 0.4956 - val_categorical_accuracy: 0.8125 - val_recall_90: 0.6986\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4621 - categorical_accuracy: 0.7847 - recall_90: 0.6359 - val_loss: 0.5003 - val_categorical_accuracy: 0.7760 - val_recall_90: 0.5616\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 23ms/step - loss: 0.4553 - categorical_accuracy: 0.7865 - recall_90: 0.5077 - val_loss: 0.4957 - val_categorical_accuracy: 0.7865 - val_recall_90: 0.6027\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4506 - categorical_accuracy: 0.7865 - recall_90: 0.5282 - val_loss: 0.4989 - val_categorical_accuracy: 0.7708 - val_recall_90: 0.5616\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4510 - categorical_accuracy: 0.7847 - recall_90: 0.5692 - val_loss: 0.4989 - val_categorical_accuracy: 0.7760 - val_recall_90: 0.5753\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4449 - categorical_accuracy: 0.7934 - recall_90: 0.5692 - val_loss: 0.4996 - val_categorical_accuracy: 0.7708 - val_recall_90: 0.5753\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 23ms/step - loss: 0.4435 - categorical_accuracy: 0.7917 - recall_90: 0.5641 - val_loss: 0.5036 - val_categorical_accuracy: 0.7708 - val_recall_90: 0.5616\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4428 - categorical_accuracy: 0.7951 - recall_90: 0.5949 - val_loss: 0.5012 - val_categorical_accuracy: 0.7760 - val_recall_90: 0.5890\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.4418 - categorical_accuracy: 0.7934 - recall_90: 0.5487 - val_loss: 0.5004 - val_categorical_accuracy: 0.7812 - val_recall_90: 0.6027\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4413 - categorical_accuracy: 0.7830 - recall_90: 0.6051 - val_loss: 0.5038 - val_categorical_accuracy: 0.7760 - val_recall_90: 0.5890\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4407 - categorical_accuracy: 0.7882 - recall_90: 0.5333 - val_loss: 0.5037 - val_categorical_accuracy: 0.7760 - val_recall_90: 0.5890\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 27ms/step - loss: 0.4369 - categorical_accuracy: 0.7865 - recall_90: 0.5590 - val_loss: 0.5054 - val_categorical_accuracy: 0.7760 - val_recall_90: 0.5890\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4374 - categorical_accuracy: 0.7986 - recall_90: 0.6256 - val_loss: 0.5068 - val_categorical_accuracy: 0.7812 - val_recall_90: 0.5890\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4379 - categorical_accuracy: 0.7986 - recall_90: 0.5385 - val_loss: 0.5031 - val_categorical_accuracy: 0.7865 - val_recall_90: 0.6164\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4383 - categorical_accuracy: 0.7917 - recall_90: 0.6513 - val_loss: 0.5149 - val_categorical_accuracy: 0.7656 - val_recall_90: 0.5479\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4384 - categorical_accuracy: 0.7899 - recall_90: 0.5179 - val_loss: 0.5072 - val_categorical_accuracy: 0.7865 - val_recall_90: 0.6301\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4373 - categorical_accuracy: 0.8003 - recall_90: 0.6000 - val_loss: 0.5072 - val_categorical_accuracy: 0.7812 - val_recall_90: 0.6027\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4339 - categorical_accuracy: 0.7934 - recall_90: 0.6308 - val_loss: 0.5167 - val_categorical_accuracy: 0.7656 - val_recall_90: 0.5479\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 5s 109ms/step - loss: 0.6713 - categorical_accuracy: 0.6615 - recall_91: 0.0000e+00 - val_loss: 0.6755 - val_categorical_accuracy: 0.6198 - val_recall_91: 0.0000e+00\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.6604 - categorical_accuracy: 0.6615 - recall_91: 0.0000e+00 - val_loss: 0.6712 - val_categorical_accuracy: 0.6198 - val_recall_91: 0.0000e+00\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 20ms/step - loss: 0.6530 - categorical_accuracy: 0.6615 - recall_91: 0.0000e+00 - val_loss: 0.6690 - val_categorical_accuracy: 0.6198 - val_recall_91: 0.0000e+00\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 22ms/step - loss: 0.6475 - categorical_accuracy: 0.6615 - recall_91: 0.0000e+00 - val_loss: 0.6658 - val_categorical_accuracy: 0.6198 - val_recall_91: 0.0000e+00\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6433 - categorical_accuracy: 0.6615 - recall_91: 0.0000e+00 - val_loss: 0.6625 - val_categorical_accuracy: 0.6198 - val_recall_91: 0.0000e+00\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.6392 - categorical_accuracy: 0.6615 - recall_91: 0.0000e+00 - val_loss: 0.6593 - val_categorical_accuracy: 0.6198 - val_recall_91: 0.0000e+00\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6363 - categorical_accuracy: 0.6615 - recall_91: 0.0000e+00 - val_loss: 0.6567 - val_categorical_accuracy: 0.6198 - val_recall_91: 0.0000e+00\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.6330 - categorical_accuracy: 0.6615 - recall_91: 0.0000e+00 - val_loss: 0.6529 - val_categorical_accuracy: 0.6198 - val_recall_91: 0.0000e+00\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6306 - categorical_accuracy: 0.6615 - recall_91: 0.0000e+00 - val_loss: 0.6505 - val_categorical_accuracy: 0.6198 - val_recall_91: 0.0000e+00\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 31ms/step - loss: 0.6275 - categorical_accuracy: 0.6615 - recall_91: 0.0000e+00 - val_loss: 0.6475 - val_categorical_accuracy: 0.6198 - val_recall_91: 0.0000e+00\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6244 - categorical_accuracy: 0.6615 - recall_91: 0.0000e+00 - val_loss: 0.6441 - val_categorical_accuracy: 0.6198 - val_recall_91: 0.0000e+00\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6212 - categorical_accuracy: 0.6615 - recall_91: 0.0000e+00 - val_loss: 0.6424 - val_categorical_accuracy: 0.6198 - val_recall_91: 0.0000e+00\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.6176 - categorical_accuracy: 0.6615 - recall_91: 0.0000e+00 - val_loss: 0.6378 - val_categorical_accuracy: 0.6198 - val_recall_91: 0.0000e+00\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6147 - categorical_accuracy: 0.6615 - recall_91: 0.0000e+00 - val_loss: 0.6337 - val_categorical_accuracy: 0.6198 - val_recall_91: 0.0000e+00\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6104 - categorical_accuracy: 0.6615 - recall_91: 0.0000e+00 - val_loss: 0.6313 - val_categorical_accuracy: 0.6198 - val_recall_91: 0.0000e+00\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.6063 - categorical_accuracy: 0.6615 - recall_91: 0.0000e+00 - val_loss: 0.6269 - val_categorical_accuracy: 0.6198 - val_recall_91: 0.0000e+00\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.6027 - categorical_accuracy: 0.6615 - recall_91: 0.0000e+00 - val_loss: 0.6233 - val_categorical_accuracy: 0.6250 - val_recall_91: 0.0137\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 26ms/step - loss: 0.5983 - categorical_accuracy: 0.6667 - recall_91: 0.0154 - val_loss: 0.6178 - val_categorical_accuracy: 0.6354 - val_recall_91: 0.0411\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5940 - categorical_accuracy: 0.6667 - recall_91: 0.0256 - val_loss: 0.6144 - val_categorical_accuracy: 0.6354 - val_recall_91: 0.0411\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5892 - categorical_accuracy: 0.6667 - recall_91: 0.0256 - val_loss: 0.6111 - val_categorical_accuracy: 0.6406 - val_recall_91: 0.0548\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5856 - categorical_accuracy: 0.6632 - recall_91: 0.0205 - val_loss: 0.6076 - val_categorical_accuracy: 0.6562 - val_recall_91: 0.0959\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5810 - categorical_accuracy: 0.6823 - recall_91: 0.0769 - val_loss: 0.6012 - val_categorical_accuracy: 0.6667 - val_recall_91: 0.1370\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5759 - categorical_accuracy: 0.6944 - recall_91: 0.1436 - val_loss: 0.5965 - val_categorical_accuracy: 0.6771 - val_recall_91: 0.1918\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5715 - categorical_accuracy: 0.7066 - recall_91: 0.1744 - val_loss: 0.5946 - val_categorical_accuracy: 0.6719 - val_recall_91: 0.1918\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5671 - categorical_accuracy: 0.7101 - recall_91: 0.1795 - val_loss: 0.5906 - val_categorical_accuracy: 0.6823 - val_recall_91: 0.2192\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5624 - categorical_accuracy: 0.7135 - recall_91: 0.1949 - val_loss: 0.5881 - val_categorical_accuracy: 0.6823 - val_recall_91: 0.2192\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.5579 - categorical_accuracy: 0.7153 - recall_91: 0.2000 - val_loss: 0.5847 - val_categorical_accuracy: 0.6823 - val_recall_91: 0.2329\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 26ms/step - loss: 0.5531 - categorical_accuracy: 0.7292 - recall_91: 0.2769 - val_loss: 0.5784 - val_categorical_accuracy: 0.6875 - val_recall_91: 0.2877\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5486 - categorical_accuracy: 0.7274 - recall_91: 0.2872 - val_loss: 0.5762 - val_categorical_accuracy: 0.6875 - val_recall_91: 0.2603\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5441 - categorical_accuracy: 0.7326 - recall_91: 0.2872 - val_loss: 0.5728 - val_categorical_accuracy: 0.6979 - val_recall_91: 0.2877\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.5405 - categorical_accuracy: 0.7309 - recall_91: 0.3179 - val_loss: 0.5676 - val_categorical_accuracy: 0.7031 - val_recall_91: 0.3425\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5364 - categorical_accuracy: 0.7378 - recall_91: 0.3487 - val_loss: 0.5652 - val_categorical_accuracy: 0.7031 - val_recall_91: 0.3425\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5318 - categorical_accuracy: 0.7431 - recall_91: 0.3385 - val_loss: 0.5643 - val_categorical_accuracy: 0.7031 - val_recall_91: 0.3288\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5287 - categorical_accuracy: 0.7500 - recall_91: 0.3692 - val_loss: 0.5605 - val_categorical_accuracy: 0.7135 - val_recall_91: 0.3699\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5242 - categorical_accuracy: 0.7552 - recall_91: 0.3744 - val_loss: 0.5593 - val_categorical_accuracy: 0.7083 - val_recall_91: 0.3562\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5217 - categorical_accuracy: 0.7639 - recall_91: 0.4154 - val_loss: 0.5549 - val_categorical_accuracy: 0.7083 - val_recall_91: 0.4110\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 35ms/step - loss: 0.5173 - categorical_accuracy: 0.7674 - recall_91: 0.4256 - val_loss: 0.5545 - val_categorical_accuracy: 0.7083 - val_recall_91: 0.3699\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.5139 - categorical_accuracy: 0.7656 - recall_91: 0.4000 - val_loss: 0.5519 - val_categorical_accuracy: 0.7240 - val_recall_91: 0.4247\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5116 - categorical_accuracy: 0.7674 - recall_91: 0.4462 - val_loss: 0.5478 - val_categorical_accuracy: 0.7240 - val_recall_91: 0.4658\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5079 - categorical_accuracy: 0.7743 - recall_91: 0.4667 - val_loss: 0.5486 - val_categorical_accuracy: 0.7240 - val_recall_91: 0.4247\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5068 - categorical_accuracy: 0.7674 - recall_91: 0.4051 - val_loss: 0.5486 - val_categorical_accuracy: 0.7135 - val_recall_91: 0.3973\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5026 - categorical_accuracy: 0.7760 - recall_91: 0.4564 - val_loss: 0.5460 - val_categorical_accuracy: 0.7292 - val_recall_91: 0.4384\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.5016 - categorical_accuracy: 0.7674 - recall_91: 0.4205 - val_loss: 0.5461 - val_categorical_accuracy: 0.7188 - val_recall_91: 0.4110\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4969 - categorical_accuracy: 0.7743 - recall_91: 0.4872 - val_loss: 0.5414 - val_categorical_accuracy: 0.7188 - val_recall_91: 0.4658\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4947 - categorical_accuracy: 0.7812 - recall_91: 0.5026 - val_loss: 0.5420 - val_categorical_accuracy: 0.7292 - val_recall_91: 0.4521\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 29ms/step - loss: 0.4922 - categorical_accuracy: 0.7726 - recall_91: 0.4615 - val_loss: 0.5427 - val_categorical_accuracy: 0.7292 - val_recall_91: 0.4384\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4897 - categorical_accuracy: 0.7760 - recall_91: 0.4667 - val_loss: 0.5397 - val_categorical_accuracy: 0.7292 - val_recall_91: 0.4521\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4879 - categorical_accuracy: 0.7795 - recall_91: 0.5077 - val_loss: 0.5391 - val_categorical_accuracy: 0.7240 - val_recall_91: 0.4384\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4847 - categorical_accuracy: 0.7795 - recall_91: 0.4513 - val_loss: 0.5397 - val_categorical_accuracy: 0.7292 - val_recall_91: 0.4384\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4831 - categorical_accuracy: 0.7865 - recall_91: 0.4974 - val_loss: 0.5370 - val_categorical_accuracy: 0.7292 - val_recall_91: 0.4658\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 5s 115ms/step - loss: 0.6909 - categorical_accuracy: 0.5660 - recall_92: 0.6872 - val_loss: 0.6779 - val_categorical_accuracy: 0.6250 - val_recall_92: 0.2329\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.6680 - categorical_accuracy: 0.6597 - recall_92: 0.0564 - val_loss: 0.6673 - val_categorical_accuracy: 0.6198 - val_recall_92: 0.0000e+00\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6525 - categorical_accuracy: 0.6615 - recall_92: 0.0000e+00 - val_loss: 0.6603 - val_categorical_accuracy: 0.6198 - val_recall_92: 0.0000e+00\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6434 - categorical_accuracy: 0.6615 - recall_92: 0.0000e+00 - val_loss: 0.6553 - val_categorical_accuracy: 0.6198 - val_recall_92: 0.0000e+00\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 33ms/step - loss: 0.6339 - categorical_accuracy: 0.6615 - recall_92: 0.0000e+00 - val_loss: 0.6477 - val_categorical_accuracy: 0.6198 - val_recall_92: 0.0000e+00\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.6252 - categorical_accuracy: 0.6615 - recall_92: 0.0000e+00 - val_loss: 0.6374 - val_categorical_accuracy: 0.6198 - val_recall_92: 0.0000e+00\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.6175 - categorical_accuracy: 0.6615 - recall_92: 0.0000e+00 - val_loss: 0.6292 - val_categorical_accuracy: 0.6198 - val_recall_92: 0.0000e+00\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6096 - categorical_accuracy: 0.6684 - recall_92: 0.0205 - val_loss: 0.6182 - val_categorical_accuracy: 0.6354 - val_recall_92: 0.0411\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.6010 - categorical_accuracy: 0.6701 - recall_92: 0.0410 - val_loss: 0.6107 - val_categorical_accuracy: 0.6510 - val_recall_92: 0.0822\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5931 - categorical_accuracy: 0.6806 - recall_92: 0.0923 - val_loss: 0.6024 - val_categorical_accuracy: 0.6615 - val_recall_92: 0.1233\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5851 - categorical_accuracy: 0.6875 - recall_92: 0.1128 - val_loss: 0.5949 - val_categorical_accuracy: 0.6615 - val_recall_92: 0.1370\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 28ms/step - loss: 0.5770 - categorical_accuracy: 0.7014 - recall_92: 0.1795 - val_loss: 0.5841 - val_categorical_accuracy: 0.6823 - val_recall_92: 0.2329\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5693 - categorical_accuracy: 0.7083 - recall_92: 0.2564 - val_loss: 0.5777 - val_categorical_accuracy: 0.6927 - val_recall_92: 0.2603\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5629 - categorical_accuracy: 0.7101 - recall_92: 0.2308 - val_loss: 0.5717 - val_categorical_accuracy: 0.7188 - val_recall_92: 0.3288\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5557 - categorical_accuracy: 0.7240 - recall_92: 0.2718 - val_loss: 0.5648 - val_categorical_accuracy: 0.7188 - val_recall_92: 0.3425\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5487 - categorical_accuracy: 0.7413 - recall_92: 0.3846 - val_loss: 0.5570 - val_categorical_accuracy: 0.7604 - val_recall_92: 0.4658\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5444 - categorical_accuracy: 0.7396 - recall_92: 0.3795 - val_loss: 0.5547 - val_categorical_accuracy: 0.7188 - val_recall_92: 0.3562\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5394 - categorical_accuracy: 0.7500 - recall_92: 0.4308 - val_loss: 0.5485 - val_categorical_accuracy: 0.7552 - val_recall_92: 0.4795\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5321 - categorical_accuracy: 0.7517 - recall_92: 0.3897 - val_loss: 0.5486 - val_categorical_accuracy: 0.7292 - val_recall_92: 0.3973\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5259 - categorical_accuracy: 0.7448 - recall_92: 0.4051 - val_loss: 0.5414 - val_categorical_accuracy: 0.7448 - val_recall_92: 0.4795\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.5206 - categorical_accuracy: 0.7552 - recall_92: 0.4308 - val_loss: 0.5385 - val_categorical_accuracy: 0.7396 - val_recall_92: 0.4521\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5147 - categorical_accuracy: 0.7587 - recall_92: 0.4410 - val_loss: 0.5361 - val_categorical_accuracy: 0.7448 - val_recall_92: 0.4658\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5089 - categorical_accuracy: 0.7604 - recall_92: 0.4513 - val_loss: 0.5375 - val_categorical_accuracy: 0.7344 - val_recall_92: 0.4384\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.5015 - categorical_accuracy: 0.7656 - recall_92: 0.4513 - val_loss: 0.5355 - val_categorical_accuracy: 0.7448 - val_recall_92: 0.4521\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 26ms/step - loss: 0.5014 - categorical_accuracy: 0.7604 - recall_92: 0.4103 - val_loss: 0.5337 - val_categorical_accuracy: 0.7448 - val_recall_92: 0.4521\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4948 - categorical_accuracy: 0.7778 - recall_92: 0.5538 - val_loss: 0.5293 - val_categorical_accuracy: 0.7708 - val_recall_92: 0.5753\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4879 - categorical_accuracy: 0.7639 - recall_92: 0.4769 - val_loss: 0.5354 - val_categorical_accuracy: 0.7396 - val_recall_92: 0.4384\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4855 - categorical_accuracy: 0.7674 - recall_92: 0.4462 - val_loss: 0.5318 - val_categorical_accuracy: 0.7448 - val_recall_92: 0.4521\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4797 - categorical_accuracy: 0.7847 - recall_92: 0.5385 - val_loss: 0.5262 - val_categorical_accuracy: 0.7500 - val_recall_92: 0.5479\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4765 - categorical_accuracy: 0.7743 - recall_92: 0.5026 - val_loss: 0.5307 - val_categorical_accuracy: 0.7552 - val_recall_92: 0.4795\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4738 - categorical_accuracy: 0.7812 - recall_92: 0.4821 - val_loss: 0.5263 - val_categorical_accuracy: 0.7500 - val_recall_92: 0.5205\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4701 - categorical_accuracy: 0.7812 - recall_92: 0.5282 - val_loss: 0.5247 - val_categorical_accuracy: 0.7448 - val_recall_92: 0.5342\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 38ms/step - loss: 0.4679 - categorical_accuracy: 0.7969 - recall_92: 0.6000 - val_loss: 0.5266 - val_categorical_accuracy: 0.7552 - val_recall_92: 0.5342\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4675 - categorical_accuracy: 0.7865 - recall_92: 0.4872 - val_loss: 0.5264 - val_categorical_accuracy: 0.7552 - val_recall_92: 0.5342\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4660 - categorical_accuracy: 0.7899 - recall_92: 0.6410 - val_loss: 0.5227 - val_categorical_accuracy: 0.7552 - val_recall_92: 0.5890\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4601 - categorical_accuracy: 0.7812 - recall_92: 0.5128 - val_loss: 0.5308 - val_categorical_accuracy: 0.7656 - val_recall_92: 0.5342\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4572 - categorical_accuracy: 0.7847 - recall_92: 0.5282 - val_loss: 0.5219 - val_categorical_accuracy: 0.7500 - val_recall_92: 0.5890\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4558 - categorical_accuracy: 0.7917 - recall_92: 0.6051 - val_loss: 0.5271 - val_categorical_accuracy: 0.7604 - val_recall_92: 0.5479\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4527 - categorical_accuracy: 0.7934 - recall_92: 0.5333 - val_loss: 0.5258 - val_categorical_accuracy: 0.7708 - val_recall_92: 0.5753\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4503 - categorical_accuracy: 0.7986 - recall_92: 0.5846 - val_loss: 0.5234 - val_categorical_accuracy: 0.7708 - val_recall_92: 0.5890\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4483 - categorical_accuracy: 0.7969 - recall_92: 0.5897 - val_loss: 0.5221 - val_categorical_accuracy: 0.7656 - val_recall_92: 0.5753\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4469 - categorical_accuracy: 0.7969 - recall_92: 0.5692 - val_loss: 0.5228 - val_categorical_accuracy: 0.7760 - val_recall_92: 0.5890\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4448 - categorical_accuracy: 0.7969 - recall_92: 0.6000 - val_loss: 0.5228 - val_categorical_accuracy: 0.7708 - val_recall_92: 0.5890\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.4444 - categorical_accuracy: 0.7951 - recall_92: 0.5590 - val_loss: 0.5250 - val_categorical_accuracy: 0.7760 - val_recall_92: 0.6027\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4472 - categorical_accuracy: 0.7899 - recall_92: 0.6103 - val_loss: 0.5266 - val_categorical_accuracy: 0.7708 - val_recall_92: 0.5616\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4405 - categorical_accuracy: 0.7934 - recall_92: 0.5846 - val_loss: 0.5256 - val_categorical_accuracy: 0.7708 - val_recall_92: 0.6301\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4389 - categorical_accuracy: 0.7969 - recall_92: 0.6103 - val_loss: 0.5279 - val_categorical_accuracy: 0.7656 - val_recall_92: 0.5616\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 26ms/step - loss: 0.4409 - categorical_accuracy: 0.7951 - recall_92: 0.5692 - val_loss: 0.5242 - val_categorical_accuracy: 0.7760 - val_recall_92: 0.6164\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4383 - categorical_accuracy: 0.7951 - recall_92: 0.6205 - val_loss: 0.5299 - val_categorical_accuracy: 0.7760 - val_recall_92: 0.5753\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4353 - categorical_accuracy: 0.7986 - recall_92: 0.5897 - val_loss: 0.5292 - val_categorical_accuracy: 0.7760 - val_recall_92: 0.6027\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 5s 108ms/step - loss: 0.7070 - categorical_accuracy: 0.3524 - recall_93: 0.8923 - val_loss: 0.6883 - val_categorical_accuracy: 0.6198 - val_recall_93: 0.5068\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6871 - categorical_accuracy: 0.6267 - recall_93: 0.1538 - val_loss: 0.6784 - val_categorical_accuracy: 0.6406 - val_recall_93: 0.0685\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6734 - categorical_accuracy: 0.6615 - recall_93: 0.0359 - val_loss: 0.6702 - val_categorical_accuracy: 0.6302 - val_recall_93: 0.0274\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6628 - categorical_accuracy: 0.6580 - recall_93: 0.0000e+00 - val_loss: 0.6637 - val_categorical_accuracy: 0.6198 - val_recall_93: 0.0000e+00\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6540 - categorical_accuracy: 0.6580 - recall_93: 0.0000e+00 - val_loss: 0.6589 - val_categorical_accuracy: 0.6198 - val_recall_93: 0.0000e+00\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 27ms/step - loss: 0.6469 - categorical_accuracy: 0.6597 - recall_93: 0.0000e+00 - val_loss: 0.6554 - val_categorical_accuracy: 0.6198 - val_recall_93: 0.0000e+00\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.6407 - categorical_accuracy: 0.6597 - recall_93: 0.0000e+00 - val_loss: 0.6518 - val_categorical_accuracy: 0.6198 - val_recall_93: 0.0000e+00\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6353 - categorical_accuracy: 0.6597 - recall_93: 0.0000e+00 - val_loss: 0.6486 - val_categorical_accuracy: 0.6198 - val_recall_93: 0.0000e+00\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6316 - categorical_accuracy: 0.6597 - recall_93: 0.0000e+00 - val_loss: 0.6463 - val_categorical_accuracy: 0.6198 - val_recall_93: 0.0000e+00\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6269 - categorical_accuracy: 0.6597 - recall_93: 0.0000e+00 - val_loss: 0.6404 - val_categorical_accuracy: 0.6198 - val_recall_93: 0.0000e+00\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6239 - categorical_accuracy: 0.6580 - recall_93: 0.0000e+00 - val_loss: 0.6346 - val_categorical_accuracy: 0.6198 - val_recall_93: 0.0000e+00\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 27ms/step - loss: 0.6187 - categorical_accuracy: 0.6562 - recall_93: 0.0000e+00 - val_loss: 0.6301 - val_categorical_accuracy: 0.6302 - val_recall_93: 0.0274\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6142 - categorical_accuracy: 0.6580 - recall_93: 0.0051 - val_loss: 0.6275 - val_categorical_accuracy: 0.6250 - val_recall_93: 0.0137\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6103 - categorical_accuracy: 0.6580 - recall_93: 0.0051 - val_loss: 0.6229 - val_categorical_accuracy: 0.6354 - val_recall_93: 0.0411\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6051 - categorical_accuracy: 0.6649 - recall_93: 0.0359 - val_loss: 0.6152 - val_categorical_accuracy: 0.6510 - val_recall_93: 0.0822\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.6007 - categorical_accuracy: 0.6806 - recall_93: 0.0974 - val_loss: 0.6095 - val_categorical_accuracy: 0.6562 - val_recall_93: 0.1233\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5958 - categorical_accuracy: 0.6806 - recall_93: 0.1128 - val_loss: 0.6050 - val_categorical_accuracy: 0.6615 - val_recall_93: 0.1370\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5911 - categorical_accuracy: 0.6840 - recall_93: 0.1282 - val_loss: 0.5985 - val_categorical_accuracy: 0.6979 - val_recall_93: 0.2329\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5857 - categorical_accuracy: 0.6910 - recall_93: 0.1641 - val_loss: 0.5935 - val_categorical_accuracy: 0.6979 - val_recall_93: 0.2329\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5810 - categorical_accuracy: 0.6927 - recall_93: 0.1744 - val_loss: 0.5892 - val_categorical_accuracy: 0.6875 - val_recall_93: 0.2329\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5763 - categorical_accuracy: 0.6962 - recall_93: 0.1846 - val_loss: 0.5833 - val_categorical_accuracy: 0.6927 - val_recall_93: 0.2603\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5711 - categorical_accuracy: 0.6997 - recall_93: 0.2205 - val_loss: 0.5775 - val_categorical_accuracy: 0.7083 - val_recall_93: 0.3151\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5664 - categorical_accuracy: 0.6944 - recall_93: 0.2154 - val_loss: 0.5743 - val_categorical_accuracy: 0.6979 - val_recall_93: 0.2877\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5627 - categorical_accuracy: 0.6997 - recall_93: 0.2462 - val_loss: 0.5696 - val_categorical_accuracy: 0.7135 - val_recall_93: 0.3288\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 32ms/step - loss: 0.5574 - categorical_accuracy: 0.7066 - recall_93: 0.2513 - val_loss: 0.5647 - val_categorical_accuracy: 0.7240 - val_recall_93: 0.3562\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5547 - categorical_accuracy: 0.7118 - recall_93: 0.3231 - val_loss: 0.5595 - val_categorical_accuracy: 0.7240 - val_recall_93: 0.3836\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5492 - categorical_accuracy: 0.7153 - recall_93: 0.2974 - val_loss: 0.5587 - val_categorical_accuracy: 0.7240 - val_recall_93: 0.3562\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5445 - categorical_accuracy: 0.7135 - recall_93: 0.3026 - val_loss: 0.5521 - val_categorical_accuracy: 0.7292 - val_recall_93: 0.3973\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5416 - categorical_accuracy: 0.7309 - recall_93: 0.3897 - val_loss: 0.5478 - val_categorical_accuracy: 0.7292 - val_recall_93: 0.3973\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5369 - categorical_accuracy: 0.7205 - recall_93: 0.3231 - val_loss: 0.5454 - val_categorical_accuracy: 0.7344 - val_recall_93: 0.3973\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5315 - categorical_accuracy: 0.7292 - recall_93: 0.3641 - val_loss: 0.5392 - val_categorical_accuracy: 0.7344 - val_recall_93: 0.4521\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 24ms/step - loss: 0.5284 - categorical_accuracy: 0.7396 - recall_93: 0.4000 - val_loss: 0.5367 - val_categorical_accuracy: 0.7292 - val_recall_93: 0.4247\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5237 - categorical_accuracy: 0.7326 - recall_93: 0.3744 - val_loss: 0.5323 - val_categorical_accuracy: 0.7344 - val_recall_93: 0.4521\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5213 - categorical_accuracy: 0.7431 - recall_93: 0.4256 - val_loss: 0.5299 - val_categorical_accuracy: 0.7292 - val_recall_93: 0.4384\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5168 - categorical_accuracy: 0.7483 - recall_93: 0.4256 - val_loss: 0.5265 - val_categorical_accuracy: 0.7292 - val_recall_93: 0.4384\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 25ms/step - loss: 0.5140 - categorical_accuracy: 0.7448 - recall_93: 0.4256 - val_loss: 0.5225 - val_categorical_accuracy: 0.7448 - val_recall_93: 0.4932\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5099 - categorical_accuracy: 0.7465 - recall_93: 0.4359 - val_loss: 0.5206 - val_categorical_accuracy: 0.7448 - val_recall_93: 0.4795\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5067 - categorical_accuracy: 0.7465 - recall_93: 0.4308 - val_loss: 0.5193 - val_categorical_accuracy: 0.7396 - val_recall_93: 0.4658\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5038 - categorical_accuracy: 0.7552 - recall_93: 0.4359 - val_loss: 0.5149 - val_categorical_accuracy: 0.7604 - val_recall_93: 0.5342\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.5002 - categorical_accuracy: 0.7517 - recall_93: 0.4564 - val_loss: 0.5130 - val_categorical_accuracy: 0.7604 - val_recall_93: 0.5342\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4968 - categorical_accuracy: 0.7483 - recall_93: 0.4410 - val_loss: 0.5135 - val_categorical_accuracy: 0.7500 - val_recall_93: 0.5068\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4950 - categorical_accuracy: 0.7569 - recall_93: 0.4359 - val_loss: 0.5111 - val_categorical_accuracy: 0.7448 - val_recall_93: 0.4932\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 25ms/step - loss: 0.4932 - categorical_accuracy: 0.7535 - recall_93: 0.4821 - val_loss: 0.5047 - val_categorical_accuracy: 0.7812 - val_recall_93: 0.6164\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4904 - categorical_accuracy: 0.7535 - recall_93: 0.4564 - val_loss: 0.5093 - val_categorical_accuracy: 0.7448 - val_recall_93: 0.4932\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4858 - categorical_accuracy: 0.7674 - recall_93: 0.4718 - val_loss: 0.5032 - val_categorical_accuracy: 0.7656 - val_recall_93: 0.5616\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4862 - categorical_accuracy: 0.7535 - recall_93: 0.5487 - val_loss: 0.5008 - val_categorical_accuracy: 0.7656 - val_recall_93: 0.5753\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4831 - categorical_accuracy: 0.7622 - recall_93: 0.4564 - val_loss: 0.5039 - val_categorical_accuracy: 0.7552 - val_recall_93: 0.5205\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4783 - categorical_accuracy: 0.7743 - recall_93: 0.5282 - val_loss: 0.4979 - val_categorical_accuracy: 0.7865 - val_recall_93: 0.6301\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 28ms/step - loss: 0.4778 - categorical_accuracy: 0.7622 - recall_93: 0.5282 - val_loss: 0.4984 - val_categorical_accuracy: 0.7708 - val_recall_93: 0.5753\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4764 - categorical_accuracy: 0.7743 - recall_93: 0.4974 - val_loss: 0.4980 - val_categorical_accuracy: 0.7708 - val_recall_93: 0.5753\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 5s 106ms/step - loss: 0.7150 - categorical_accuracy: 0.4236 - recall_94: 0.7641 - val_loss: 0.6765 - val_categorical_accuracy: 0.6146 - val_recall_94: 0.0137\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 26ms/step - loss: 0.6589 - categorical_accuracy: 0.6615 - recall_94: 0.0051 - val_loss: 0.6705 - val_categorical_accuracy: 0.6198 - val_recall_94: 0.0000e+00\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6466 - categorical_accuracy: 0.6615 - recall_94: 0.0000e+00 - val_loss: 0.6690 - val_categorical_accuracy: 0.6198 - val_recall_94: 0.0000e+00\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.6414 - categorical_accuracy: 0.6615 - recall_94: 0.0000e+00 - val_loss: 0.6602 - val_categorical_accuracy: 0.6198 - val_recall_94: 0.0000e+00\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.6361 - categorical_accuracy: 0.6615 - recall_94: 0.0000e+00 - val_loss: 0.6539 - val_categorical_accuracy: 0.6198 - val_recall_94: 0.0000e+00\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.6300 - categorical_accuracy: 0.6649 - recall_94: 0.0103 - val_loss: 0.6436 - val_categorical_accuracy: 0.6302 - val_recall_94: 0.0274\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 27ms/step - loss: 0.6224 - categorical_accuracy: 0.6649 - recall_94: 0.0103 - val_loss: 0.6399 - val_categorical_accuracy: 0.6302 - val_recall_94: 0.0274\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6155 - categorical_accuracy: 0.6667 - recall_94: 0.0154 - val_loss: 0.6307 - val_categorical_accuracy: 0.6302 - val_recall_94: 0.0411\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.6097 - categorical_accuracy: 0.6684 - recall_94: 0.0769 - val_loss: 0.6186 - val_categorical_accuracy: 0.6562 - val_recall_94: 0.1233\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5993 - categorical_accuracy: 0.6771 - recall_94: 0.0974 - val_loss: 0.6113 - val_categorical_accuracy: 0.6667 - val_recall_94: 0.1507\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5962 - categorical_accuracy: 0.6771 - recall_94: 0.0821 - val_loss: 0.6071 - val_categorical_accuracy: 0.6562 - val_recall_94: 0.1233\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 31ms/step - loss: 0.5846 - categorical_accuracy: 0.6927 - recall_94: 0.2051 - val_loss: 0.5924 - val_categorical_accuracy: 0.7188 - val_recall_94: 0.3562\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5777 - categorical_accuracy: 0.6962 - recall_94: 0.2359 - val_loss: 0.5880 - val_categorical_accuracy: 0.6823 - val_recall_94: 0.2192\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5683 - categorical_accuracy: 0.6979 - recall_94: 0.2103 - val_loss: 0.5794 - val_categorical_accuracy: 0.7031 - val_recall_94: 0.2877\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5609 - categorical_accuracy: 0.6979 - recall_94: 0.2872 - val_loss: 0.5705 - val_categorical_accuracy: 0.7396 - val_recall_94: 0.3836\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5539 - categorical_accuracy: 0.6997 - recall_94: 0.2513 - val_loss: 0.5642 - val_categorical_accuracy: 0.7396 - val_recall_94: 0.3836\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5447 - categorical_accuracy: 0.7101 - recall_94: 0.3333 - val_loss: 0.5550 - val_categorical_accuracy: 0.7396 - val_recall_94: 0.3973\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5405 - categorical_accuracy: 0.7257 - recall_94: 0.3949 - val_loss: 0.5523 - val_categorical_accuracy: 0.7396 - val_recall_94: 0.3836\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5296 - categorical_accuracy: 0.7240 - recall_94: 0.3333 - val_loss: 0.5419 - val_categorical_accuracy: 0.7448 - val_recall_94: 0.4658\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5228 - categorical_accuracy: 0.7413 - recall_94: 0.4308 - val_loss: 0.5388 - val_categorical_accuracy: 0.7448 - val_recall_94: 0.4247\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 23ms/step - loss: 0.5161 - categorical_accuracy: 0.7413 - recall_94: 0.4308 - val_loss: 0.5353 - val_categorical_accuracy: 0.7396 - val_recall_94: 0.4247\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5103 - categorical_accuracy: 0.7535 - recall_94: 0.4256 - val_loss: 0.5274 - val_categorical_accuracy: 0.7708 - val_recall_94: 0.5205\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.5027 - categorical_accuracy: 0.7552 - recall_94: 0.4308 - val_loss: 0.5272 - val_categorical_accuracy: 0.7552 - val_recall_94: 0.4521\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4984 - categorical_accuracy: 0.7500 - recall_94: 0.4256 - val_loss: 0.5166 - val_categorical_accuracy: 0.7708 - val_recall_94: 0.5753\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4916 - categorical_accuracy: 0.7604 - recall_94: 0.4564 - val_loss: 0.5167 - val_categorical_accuracy: 0.7812 - val_recall_94: 0.5479\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4902 - categorical_accuracy: 0.7795 - recall_94: 0.5641 - val_loss: 0.5101 - val_categorical_accuracy: 0.7812 - val_recall_94: 0.5616\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4885 - categorical_accuracy: 0.7587 - recall_94: 0.4051 - val_loss: 0.5065 - val_categorical_accuracy: 0.7865 - val_recall_94: 0.5890\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 22ms/step - loss: 0.4864 - categorical_accuracy: 0.7726 - recall_94: 0.6769 - val_loss: 0.5037 - val_categorical_accuracy: 0.7865 - val_recall_94: 0.5890\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 20ms/step - loss: 0.4810 - categorical_accuracy: 0.7569 - recall_94: 0.4256 - val_loss: 0.5037 - val_categorical_accuracy: 0.7969 - val_recall_94: 0.5890\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4763 - categorical_accuracy: 0.7847 - recall_94: 0.6410 - val_loss: 0.5020 - val_categorical_accuracy: 0.7969 - val_recall_94: 0.6027\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4709 - categorical_accuracy: 0.7726 - recall_94: 0.4564 - val_loss: 0.5037 - val_categorical_accuracy: 0.7969 - val_recall_94: 0.5753\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4725 - categorical_accuracy: 0.7830 - recall_94: 0.6410 - val_loss: 0.4985 - val_categorical_accuracy: 0.8125 - val_recall_94: 0.6438\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 25ms/step - loss: 0.4622 - categorical_accuracy: 0.7743 - recall_94: 0.4974 - val_loss: 0.5002 - val_categorical_accuracy: 0.7917 - val_recall_94: 0.5753\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4570 - categorical_accuracy: 0.7812 - recall_94: 0.5692 - val_loss: 0.4956 - val_categorical_accuracy: 0.7917 - val_recall_94: 0.6438\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 33ms/step - loss: 0.4561 - categorical_accuracy: 0.7726 - recall_94: 0.5487 - val_loss: 0.4955 - val_categorical_accuracy: 0.7969 - val_recall_94: 0.6438\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4542 - categorical_accuracy: 0.7743 - recall_94: 0.5128 - val_loss: 0.4947 - val_categorical_accuracy: 0.7917 - val_recall_94: 0.6438\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4509 - categorical_accuracy: 0.7882 - recall_94: 0.5897 - val_loss: 0.4987 - val_categorical_accuracy: 0.7865 - val_recall_94: 0.5890\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 40ms/step - loss: 0.4488 - categorical_accuracy: 0.7847 - recall_94: 0.5385 - val_loss: 0.4938 - val_categorical_accuracy: 0.7708 - val_recall_94: 0.6027\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.4480 - categorical_accuracy: 0.7917 - recall_94: 0.6103 - val_loss: 0.4950 - val_categorical_accuracy: 0.7917 - val_recall_94: 0.6027\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4456 - categorical_accuracy: 0.7899 - recall_94: 0.5897 - val_loss: 0.4947 - val_categorical_accuracy: 0.7812 - val_recall_94: 0.6027\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4448 - categorical_accuracy: 0.7882 - recall_94: 0.5590 - val_loss: 0.4968 - val_categorical_accuracy: 0.7812 - val_recall_94: 0.6027\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4410 - categorical_accuracy: 0.7882 - recall_94: 0.5692 - val_loss: 0.4934 - val_categorical_accuracy: 0.7969 - val_recall_94: 0.6849\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4428 - categorical_accuracy: 0.7899 - recall_94: 0.5846 - val_loss: 0.4959 - val_categorical_accuracy: 0.7760 - val_recall_94: 0.6027\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4440 - categorical_accuracy: 0.7778 - recall_94: 0.5436 - val_loss: 0.4944 - val_categorical_accuracy: 0.7865 - val_recall_94: 0.6301\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 29ms/step - loss: 0.4409 - categorical_accuracy: 0.7951 - recall_94: 0.6000 - val_loss: 0.4954 - val_categorical_accuracy: 0.7812 - val_recall_94: 0.6301\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4386 - categorical_accuracy: 0.7951 - recall_94: 0.6256 - val_loss: 0.4986 - val_categorical_accuracy: 0.7812 - val_recall_94: 0.5890\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4358 - categorical_accuracy: 0.7969 - recall_94: 0.6051 - val_loss: 0.4976 - val_categorical_accuracy: 0.7760 - val_recall_94: 0.6164\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4425 - categorical_accuracy: 0.7934 - recall_94: 0.6718 - val_loss: 0.5064 - val_categorical_accuracy: 0.7760 - val_recall_94: 0.5753\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.4394 - categorical_accuracy: 0.7795 - recall_94: 0.5026 - val_loss: 0.4981 - val_categorical_accuracy: 0.7760 - val_recall_94: 0.6164\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4352 - categorical_accuracy: 0.7986 - recall_94: 0.6615 - val_loss: 0.5021 - val_categorical_accuracy: 0.7812 - val_recall_94: 0.6164\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 5s 118ms/step - loss: 0.6690 - categorical_accuracy: 0.6615 - recall_95: 0.0000e+00 - val_loss: 0.6798 - val_categorical_accuracy: 0.6198 - val_recall_95: 0.0000e+00\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6589 - categorical_accuracy: 0.6615 - recall_95: 0.0000e+00 - val_loss: 0.6794 - val_categorical_accuracy: 0.6198 - val_recall_95: 0.0000e+00\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6546 - categorical_accuracy: 0.6615 - recall_95: 0.0000e+00 - val_loss: 0.6746 - val_categorical_accuracy: 0.6198 - val_recall_95: 0.0000e+00\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6501 - categorical_accuracy: 0.6615 - recall_95: 0.0000e+00 - val_loss: 0.6696 - val_categorical_accuracy: 0.6198 - val_recall_95: 0.0000e+00\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 39ms/step - loss: 0.6466 - categorical_accuracy: 0.6615 - recall_95: 0.0000e+00 - val_loss: 0.6631 - val_categorical_accuracy: 0.6198 - val_recall_95: 0.0000e+00\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.6418 - categorical_accuracy: 0.6615 - recall_95: 0.0000e+00 - val_loss: 0.6576 - val_categorical_accuracy: 0.6198 - val_recall_95: 0.0000e+00\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6370 - categorical_accuracy: 0.6615 - recall_95: 0.0000e+00 - val_loss: 0.6541 - val_categorical_accuracy: 0.6198 - val_recall_95: 0.0000e+00\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6318 - categorical_accuracy: 0.6615 - recall_95: 0.0000e+00 - val_loss: 0.6479 - val_categorical_accuracy: 0.6198 - val_recall_95: 0.0000e+00\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.6264 - categorical_accuracy: 0.6615 - recall_95: 0.0000e+00 - val_loss: 0.6398 - val_categorical_accuracy: 0.6198 - val_recall_95: 0.0000e+00\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 36ms/step - loss: 0.6215 - categorical_accuracy: 0.6615 - recall_95: 0.0000e+00 - val_loss: 0.6349 - val_categorical_accuracy: 0.6250 - val_recall_95: 0.0137\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.6135 - categorical_accuracy: 0.6615 - recall_95: 0.0000e+00 - val_loss: 0.6249 - val_categorical_accuracy: 0.6250 - val_recall_95: 0.0137\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.6069 - categorical_accuracy: 0.6719 - recall_95: 0.0308 - val_loss: 0.6155 - val_categorical_accuracy: 0.6250 - val_recall_95: 0.0137\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5998 - categorical_accuracy: 0.6719 - recall_95: 0.0410 - val_loss: 0.6081 - val_categorical_accuracy: 0.6406 - val_recall_95: 0.0548\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5925 - categorical_accuracy: 0.6736 - recall_95: 0.0462 - val_loss: 0.6025 - val_categorical_accuracy: 0.6458 - val_recall_95: 0.0685\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.5830 - categorical_accuracy: 0.6979 - recall_95: 0.1282 - val_loss: 0.5885 - val_categorical_accuracy: 0.6875 - val_recall_95: 0.2055\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 24ms/step - loss: 0.5734 - categorical_accuracy: 0.7066 - recall_95: 0.2051 - val_loss: 0.5806 - val_categorical_accuracy: 0.7135 - val_recall_95: 0.2740\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5639 - categorical_accuracy: 0.7049 - recall_95: 0.2205 - val_loss: 0.5716 - val_categorical_accuracy: 0.7135 - val_recall_95: 0.2877\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5545 - categorical_accuracy: 0.7153 - recall_95: 0.2718 - val_loss: 0.5581 - val_categorical_accuracy: 0.7135 - val_recall_95: 0.3425\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5439 - categorical_accuracy: 0.7292 - recall_95: 0.3231 - val_loss: 0.5492 - val_categorical_accuracy: 0.7240 - val_recall_95: 0.3562\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5336 - categorical_accuracy: 0.7240 - recall_95: 0.3385 - val_loss: 0.5394 - val_categorical_accuracy: 0.7448 - val_recall_95: 0.4247\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.5258 - categorical_accuracy: 0.7378 - recall_95: 0.3795 - val_loss: 0.5294 - val_categorical_accuracy: 0.7604 - val_recall_95: 0.4795\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5180 - categorical_accuracy: 0.7448 - recall_95: 0.4359 - val_loss: 0.5218 - val_categorical_accuracy: 0.7760 - val_recall_95: 0.5205\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 21ms/step - loss: 0.5109 - categorical_accuracy: 0.7587 - recall_95: 0.4769 - val_loss: 0.5154 - val_categorical_accuracy: 0.7760 - val_recall_95: 0.5205\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.5054 - categorical_accuracy: 0.7448 - recall_95: 0.4154 - val_loss: 0.5125 - val_categorical_accuracy: 0.7760 - val_recall_95: 0.5205\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4984 - categorical_accuracy: 0.7587 - recall_95: 0.5128 - val_loss: 0.5043 - val_categorical_accuracy: 0.7812 - val_recall_95: 0.5890\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4971 - categorical_accuracy: 0.7500 - recall_95: 0.4564 - val_loss: 0.5066 - val_categorical_accuracy: 0.7760 - val_recall_95: 0.5205\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4871 - categorical_accuracy: 0.7639 - recall_95: 0.5333 - val_loss: 0.4976 - val_categorical_accuracy: 0.7917 - val_recall_95: 0.6301\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4839 - categorical_accuracy: 0.7674 - recall_95: 0.5641 - val_loss: 0.5030 - val_categorical_accuracy: 0.7812 - val_recall_95: 0.5342\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 23ms/step - loss: 0.4805 - categorical_accuracy: 0.7674 - recall_95: 0.4769 - val_loss: 0.4951 - val_categorical_accuracy: 0.7969 - val_recall_95: 0.6027\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4764 - categorical_accuracy: 0.7708 - recall_95: 0.5538 - val_loss: 0.4917 - val_categorical_accuracy: 0.7812 - val_recall_95: 0.6164\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4744 - categorical_accuracy: 0.7639 - recall_95: 0.5333 - val_loss: 0.4985 - val_categorical_accuracy: 0.7760 - val_recall_95: 0.5342\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4725 - categorical_accuracy: 0.7691 - recall_95: 0.5436 - val_loss: 0.4900 - val_categorical_accuracy: 0.7708 - val_recall_95: 0.6027\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4713 - categorical_accuracy: 0.7726 - recall_95: 0.5128 - val_loss: 0.4941 - val_categorical_accuracy: 0.7865 - val_recall_95: 0.5753\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4647 - categorical_accuracy: 0.7760 - recall_95: 0.5795 - val_loss: 0.4907 - val_categorical_accuracy: 0.7656 - val_recall_95: 0.5753\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 23ms/step - loss: 0.4636 - categorical_accuracy: 0.7726 - recall_95: 0.5231 - val_loss: 0.4938 - val_categorical_accuracy: 0.7812 - val_recall_95: 0.5753\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 24ms/step - loss: 0.4627 - categorical_accuracy: 0.7760 - recall_95: 0.6103 - val_loss: 0.4902 - val_categorical_accuracy: 0.7812 - val_recall_95: 0.6027\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4576 - categorical_accuracy: 0.7708 - recall_95: 0.5231 - val_loss: 0.4971 - val_categorical_accuracy: 0.7760 - val_recall_95: 0.5616\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4574 - categorical_accuracy: 0.7760 - recall_95: 0.5641 - val_loss: 0.4886 - val_categorical_accuracy: 0.7969 - val_recall_95: 0.6712\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4563 - categorical_accuracy: 0.7795 - recall_95: 0.5692 - val_loss: 0.4958 - val_categorical_accuracy: 0.7760 - val_recall_95: 0.5753\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4529 - categorical_accuracy: 0.7726 - recall_95: 0.5590 - val_loss: 0.4907 - val_categorical_accuracy: 0.7969 - val_recall_95: 0.6438\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4499 - categorical_accuracy: 0.7760 - recall_95: 0.5795 - val_loss: 0.4948 - val_categorical_accuracy: 0.7708 - val_recall_95: 0.5753\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4488 - categorical_accuracy: 0.7726 - recall_95: 0.5487 - val_loss: 0.4931 - val_categorical_accuracy: 0.7917 - val_recall_95: 0.6301\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4527 - categorical_accuracy: 0.7708 - recall_95: 0.5077 - val_loss: 0.4937 - val_categorical_accuracy: 0.7917 - val_recall_95: 0.6301\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4528 - categorical_accuracy: 0.7812 - recall_95: 0.6769 - val_loss: 0.4939 - val_categorical_accuracy: 0.7812 - val_recall_95: 0.6301\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4509 - categorical_accuracy: 0.7778 - recall_95: 0.5179 - val_loss: 0.4981 - val_categorical_accuracy: 0.7865 - val_recall_95: 0.6027\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4437 - categorical_accuracy: 0.7795 - recall_95: 0.5949 - val_loss: 0.4928 - val_categorical_accuracy: 0.7812 - val_recall_95: 0.6301\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 22ms/step - loss: 0.4429 - categorical_accuracy: 0.7812 - recall_95: 0.6103 - val_loss: 0.5015 - val_categorical_accuracy: 0.7656 - val_recall_95: 0.5479\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4435 - categorical_accuracy: 0.7830 - recall_95: 0.5590 - val_loss: 0.4986 - val_categorical_accuracy: 0.7865 - val_recall_95: 0.6027\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4420 - categorical_accuracy: 0.7847 - recall_95: 0.5897 - val_loss: 0.4995 - val_categorical_accuracy: 0.7812 - val_recall_95: 0.5890\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4396 - categorical_accuracy: 0.7812 - recall_95: 0.5949 - val_loss: 0.4945 - val_categorical_accuracy: 0.7917 - val_recall_95: 0.6712\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 5s 104ms/step - loss: 0.6824 - categorical_accuracy: 0.6406 - recall_96: 0.1179 - val_loss: 0.6779 - val_categorical_accuracy: 0.6198 - val_recall_96: 0.0000e+00\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 22ms/step - loss: 0.6681 - categorical_accuracy: 0.6615 - recall_96: 0.0000e+00 - val_loss: 0.6690 - val_categorical_accuracy: 0.6198 - val_recall_96: 0.0000e+00\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6547 - categorical_accuracy: 0.6615 - recall_96: 0.0000e+00 - val_loss: 0.6631 - val_categorical_accuracy: 0.6198 - val_recall_96: 0.0000e+00\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6456 - categorical_accuracy: 0.6615 - recall_96: 0.0000e+00 - val_loss: 0.6600 - val_categorical_accuracy: 0.6198 - val_recall_96: 0.0000e+00\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6388 - categorical_accuracy: 0.6615 - recall_96: 0.0000e+00 - val_loss: 0.6561 - val_categorical_accuracy: 0.6198 - val_recall_96: 0.0000e+00\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6319 - categorical_accuracy: 0.6615 - recall_96: 0.0000e+00 - val_loss: 0.6472 - val_categorical_accuracy: 0.6198 - val_recall_96: 0.0000e+00\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 28ms/step - loss: 0.6206 - categorical_accuracy: 0.6615 - recall_96: 0.0000e+00 - val_loss: 0.6355 - val_categorical_accuracy: 0.6198 - val_recall_96: 0.0000e+00\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6118 - categorical_accuracy: 0.6615 - recall_96: 0.0000e+00 - val_loss: 0.6267 - val_categorical_accuracy: 0.6198 - val_recall_96: 0.0000e+00\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6027 - categorical_accuracy: 0.6632 - recall_96: 0.0051 - val_loss: 0.6158 - val_categorical_accuracy: 0.6354 - val_recall_96: 0.0411\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5930 - categorical_accuracy: 0.6701 - recall_96: 0.0410 - val_loss: 0.6070 - val_categorical_accuracy: 0.6406 - val_recall_96: 0.0685\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5858 - categorical_accuracy: 0.6910 - recall_96: 0.1231 - val_loss: 0.5976 - val_categorical_accuracy: 0.6875 - val_recall_96: 0.2055\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.5767 - categorical_accuracy: 0.6979 - recall_96: 0.1436 - val_loss: 0.5913 - val_categorical_accuracy: 0.6927 - val_recall_96: 0.2192\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5684 - categorical_accuracy: 0.7031 - recall_96: 0.1692 - val_loss: 0.5852 - val_categorical_accuracy: 0.7031 - val_recall_96: 0.2603\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5584 - categorical_accuracy: 0.7135 - recall_96: 0.2513 - val_loss: 0.5750 - val_categorical_accuracy: 0.7396 - val_recall_96: 0.4110\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5505 - categorical_accuracy: 0.7240 - recall_96: 0.3487 - val_loss: 0.5650 - val_categorical_accuracy: 0.7396 - val_recall_96: 0.4247\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5400 - categorical_accuracy: 0.7309 - recall_96: 0.4103 - val_loss: 0.5574 - val_categorical_accuracy: 0.7396 - val_recall_96: 0.4384\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5306 - categorical_accuracy: 0.7396 - recall_96: 0.4410 - val_loss: 0.5472 - val_categorical_accuracy: 0.7552 - val_recall_96: 0.4932\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 24ms/step - loss: 0.5214 - categorical_accuracy: 0.7378 - recall_96: 0.4000 - val_loss: 0.5411 - val_categorical_accuracy: 0.7500 - val_recall_96: 0.4932\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.5141 - categorical_accuracy: 0.7431 - recall_96: 0.5333 - val_loss: 0.5358 - val_categorical_accuracy: 0.7604 - val_recall_96: 0.5205\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5100 - categorical_accuracy: 0.7361 - recall_96: 0.3641 - val_loss: 0.5297 - val_categorical_accuracy: 0.7604 - val_recall_96: 0.5342\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5020 - categorical_accuracy: 0.7569 - recall_96: 0.6205 - val_loss: 0.5248 - val_categorical_accuracy: 0.7552 - val_recall_96: 0.5342\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4983 - categorical_accuracy: 0.7483 - recall_96: 0.4000 - val_loss: 0.5203 - val_categorical_accuracy: 0.7708 - val_recall_96: 0.5753\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4887 - categorical_accuracy: 0.7674 - recall_96: 0.5897 - val_loss: 0.5192 - val_categorical_accuracy: 0.7604 - val_recall_96: 0.5479\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 25ms/step - loss: 0.4822 - categorical_accuracy: 0.7639 - recall_96: 0.4872 - val_loss: 0.5168 - val_categorical_accuracy: 0.7604 - val_recall_96: 0.5753\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4782 - categorical_accuracy: 0.7639 - recall_96: 0.5282 - val_loss: 0.5158 - val_categorical_accuracy: 0.7656 - val_recall_96: 0.5479\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4735 - categorical_accuracy: 0.7656 - recall_96: 0.5436 - val_loss: 0.5142 - val_categorical_accuracy: 0.7708 - val_recall_96: 0.5479\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4671 - categorical_accuracy: 0.7778 - recall_96: 0.5744 - val_loss: 0.5130 - val_categorical_accuracy: 0.7760 - val_recall_96: 0.5753\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4646 - categorical_accuracy: 0.7708 - recall_96: 0.5128 - val_loss: 0.5110 - val_categorical_accuracy: 0.7760 - val_recall_96: 0.5890\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4613 - categorical_accuracy: 0.7812 - recall_96: 0.6154 - val_loss: 0.5108 - val_categorical_accuracy: 0.7865 - val_recall_96: 0.6027\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4570 - categorical_accuracy: 0.7778 - recall_96: 0.5333 - val_loss: 0.5091 - val_categorical_accuracy: 0.7865 - val_recall_96: 0.6301\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4571 - categorical_accuracy: 0.7708 - recall_96: 0.5538 - val_loss: 0.5175 - val_categorical_accuracy: 0.7708 - val_recall_96: 0.5342\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 22ms/step - loss: 0.4518 - categorical_accuracy: 0.7691 - recall_96: 0.5744 - val_loss: 0.5090 - val_categorical_accuracy: 0.7865 - val_recall_96: 0.6301\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4530 - categorical_accuracy: 0.7882 - recall_96: 0.5231 - val_loss: 0.5149 - val_categorical_accuracy: 0.7812 - val_recall_96: 0.5753\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4520 - categorical_accuracy: 0.7830 - recall_96: 0.6564 - val_loss: 0.5114 - val_categorical_accuracy: 0.7865 - val_recall_96: 0.6164\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4461 - categorical_accuracy: 0.7760 - recall_96: 0.5538 - val_loss: 0.5111 - val_categorical_accuracy: 0.7865 - val_recall_96: 0.6301\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4420 - categorical_accuracy: 0.7812 - recall_96: 0.5846 - val_loss: 0.5150 - val_categorical_accuracy: 0.7865 - val_recall_96: 0.6164\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4417 - categorical_accuracy: 0.7795 - recall_96: 0.5949 - val_loss: 0.5134 - val_categorical_accuracy: 0.7865 - val_recall_96: 0.6301\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.4398 - categorical_accuracy: 0.7847 - recall_96: 0.5692 - val_loss: 0.5143 - val_categorical_accuracy: 0.7865 - val_recall_96: 0.6164\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4390 - categorical_accuracy: 0.7882 - recall_96: 0.5641 - val_loss: 0.5159 - val_categorical_accuracy: 0.7865 - val_recall_96: 0.6164\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4382 - categorical_accuracy: 0.7847 - recall_96: 0.6154 - val_loss: 0.5161 - val_categorical_accuracy: 0.7865 - val_recall_96: 0.6164\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4366 - categorical_accuracy: 0.7847 - recall_96: 0.5692 - val_loss: 0.5209 - val_categorical_accuracy: 0.7760 - val_recall_96: 0.5616\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.4344 - categorical_accuracy: 0.7899 - recall_96: 0.5846 - val_loss: 0.5187 - val_categorical_accuracy: 0.7812 - val_recall_96: 0.6301\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 23ms/step - loss: 0.4369 - categorical_accuracy: 0.8003 - recall_96: 0.5692 - val_loss: 0.5178 - val_categorical_accuracy: 0.7812 - val_recall_96: 0.6301\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4399 - categorical_accuracy: 0.7882 - recall_96: 0.6872 - val_loss: 0.5235 - val_categorical_accuracy: 0.7760 - val_recall_96: 0.5753\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.4322 - categorical_accuracy: 0.7899 - recall_96: 0.5692 - val_loss: 0.5210 - val_categorical_accuracy: 0.7760 - val_recall_96: 0.6027\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4334 - categorical_accuracy: 0.7917 - recall_96: 0.5692 - val_loss: 0.5218 - val_categorical_accuracy: 0.7812 - val_recall_96: 0.6301\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4339 - categorical_accuracy: 0.7865 - recall_96: 0.6718 - val_loss: 0.5269 - val_categorical_accuracy: 0.7760 - val_recall_96: 0.5616\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.4330 - categorical_accuracy: 0.7899 - recall_96: 0.5385 - val_loss: 0.5218 - val_categorical_accuracy: 0.7865 - val_recall_96: 0.6438\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 22ms/step - loss: 0.4303 - categorical_accuracy: 0.7934 - recall_96: 0.6359 - val_loss: 0.5248 - val_categorical_accuracy: 0.7760 - val_recall_96: 0.5890\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 35ms/step - loss: 0.4287 - categorical_accuracy: 0.7934 - recall_96: 0.6308 - val_loss: 0.5279 - val_categorical_accuracy: 0.7708 - val_recall_96: 0.5753\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 4s 93ms/step - loss: 0.6909 - categorical_accuracy: 0.5330 - recall_97: 0.6308 - val_loss: 0.6833 - val_categorical_accuracy: 0.6562 - val_recall_97: 0.1096\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6788 - categorical_accuracy: 0.6562 - recall_97: 0.0154 - val_loss: 0.6766 - val_categorical_accuracy: 0.6198 - val_recall_97: 0.0000e+00\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.6691 - categorical_accuracy: 0.6597 - recall_97: 0.0000e+00 - val_loss: 0.6706 - val_categorical_accuracy: 0.6198 - val_recall_97: 0.0000e+00\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6602 - categorical_accuracy: 0.6615 - recall_97: 0.0000e+00 - val_loss: 0.6636 - val_categorical_accuracy: 0.6198 - val_recall_97: 0.0000e+00\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6502 - categorical_accuracy: 0.6615 - recall_97: 0.0000e+00 - val_loss: 0.6584 - val_categorical_accuracy: 0.6198 - val_recall_97: 0.0000e+00\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 23ms/step - loss: 0.6429 - categorical_accuracy: 0.6615 - recall_97: 0.0000e+00 - val_loss: 0.6536 - val_categorical_accuracy: 0.6198 - val_recall_97: 0.0000e+00\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6367 - categorical_accuracy: 0.6615 - recall_97: 0.0000e+00 - val_loss: 0.6482 - val_categorical_accuracy: 0.6198 - val_recall_97: 0.0000e+00\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.6307 - categorical_accuracy: 0.6615 - recall_97: 0.0000e+00 - val_loss: 0.6411 - val_categorical_accuracy: 0.6198 - val_recall_97: 0.0000e+00\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.6250 - categorical_accuracy: 0.6597 - recall_97: 0.0000e+00 - val_loss: 0.6335 - val_categorical_accuracy: 0.6198 - val_recall_97: 0.0000e+00\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 24ms/step - loss: 0.6188 - categorical_accuracy: 0.6597 - recall_97: 0.0000e+00 - val_loss: 0.6260 - val_categorical_accuracy: 0.6250 - val_recall_97: 0.0137\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6130 - categorical_accuracy: 0.6615 - recall_97: 0.0205 - val_loss: 0.6180 - val_categorical_accuracy: 0.6406 - val_recall_97: 0.0548\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6058 - categorical_accuracy: 0.6615 - recall_97: 0.0308 - val_loss: 0.6117 - val_categorical_accuracy: 0.6458 - val_recall_97: 0.0685\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5997 - categorical_accuracy: 0.6649 - recall_97: 0.0410 - val_loss: 0.6032 - val_categorical_accuracy: 0.6562 - val_recall_97: 0.1096\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5927 - categorical_accuracy: 0.6684 - recall_97: 0.0974 - val_loss: 0.5937 - val_categorical_accuracy: 0.6875 - val_recall_97: 0.2055\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.5853 - categorical_accuracy: 0.6840 - recall_97: 0.1590 - val_loss: 0.5857 - val_categorical_accuracy: 0.6927 - val_recall_97: 0.2466\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5789 - categorical_accuracy: 0.6962 - recall_97: 0.2154 - val_loss: 0.5776 - val_categorical_accuracy: 0.7031 - val_recall_97: 0.3014\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5721 - categorical_accuracy: 0.6979 - recall_97: 0.2154 - val_loss: 0.5724 - val_categorical_accuracy: 0.6979 - val_recall_97: 0.2877\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5654 - categorical_accuracy: 0.7066 - recall_97: 0.2462 - val_loss: 0.5649 - val_categorical_accuracy: 0.7135 - val_recall_97: 0.3425\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5606 - categorical_accuracy: 0.7083 - recall_97: 0.2513 - val_loss: 0.5598 - val_categorical_accuracy: 0.7135 - val_recall_97: 0.3425\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5554 - categorical_accuracy: 0.7118 - recall_97: 0.3282 - val_loss: 0.5512 - val_categorical_accuracy: 0.7344 - val_recall_97: 0.4247\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 31ms/step - loss: 0.5484 - categorical_accuracy: 0.7101 - recall_97: 0.3282 - val_loss: 0.5472 - val_categorical_accuracy: 0.7396 - val_recall_97: 0.4110\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5449 - categorical_accuracy: 0.7153 - recall_97: 0.2974 - val_loss: 0.5419 - val_categorical_accuracy: 0.7344 - val_recall_97: 0.4247\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5393 - categorical_accuracy: 0.7257 - recall_97: 0.4103 - val_loss: 0.5362 - val_categorical_accuracy: 0.7708 - val_recall_97: 0.5205\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5351 - categorical_accuracy: 0.7292 - recall_97: 0.4308 - val_loss: 0.5337 - val_categorical_accuracy: 0.7500 - val_recall_97: 0.4658\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5303 - categorical_accuracy: 0.7274 - recall_97: 0.3692 - val_loss: 0.5297 - val_categorical_accuracy: 0.7604 - val_recall_97: 0.4932\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5260 - categorical_accuracy: 0.7257 - recall_97: 0.3641 - val_loss: 0.5277 - val_categorical_accuracy: 0.7500 - val_recall_97: 0.4658\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5231 - categorical_accuracy: 0.7413 - recall_97: 0.4564 - val_loss: 0.5187 - val_categorical_accuracy: 0.7760 - val_recall_97: 0.5753\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5163 - categorical_accuracy: 0.7465 - recall_97: 0.4667 - val_loss: 0.5171 - val_categorical_accuracy: 0.7760 - val_recall_97: 0.5205\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5137 - categorical_accuracy: 0.7483 - recall_97: 0.4205 - val_loss: 0.5115 - val_categorical_accuracy: 0.7552 - val_recall_97: 0.5205\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5093 - categorical_accuracy: 0.7500 - recall_97: 0.5026 - val_loss: 0.5078 - val_categorical_accuracy: 0.7656 - val_recall_97: 0.5479\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 23ms/step - loss: 0.5095 - categorical_accuracy: 0.7448 - recall_97: 0.4103 - val_loss: 0.5060 - val_categorical_accuracy: 0.7708 - val_recall_97: 0.5205\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5039 - categorical_accuracy: 0.7448 - recall_97: 0.5179 - val_loss: 0.4995 - val_categorical_accuracy: 0.7812 - val_recall_97: 0.6027\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.4986 - categorical_accuracy: 0.7517 - recall_97: 0.5077 - val_loss: 0.4974 - val_categorical_accuracy: 0.7865 - val_recall_97: 0.5753\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 20ms/step - loss: 0.4995 - categorical_accuracy: 0.7622 - recall_97: 0.4462 - val_loss: 0.4984 - val_categorical_accuracy: 0.7708 - val_recall_97: 0.5342\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4900 - categorical_accuracy: 0.7656 - recall_97: 0.5333 - val_loss: 0.4902 - val_categorical_accuracy: 0.8021 - val_recall_97: 0.6575\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4909 - categorical_accuracy: 0.7639 - recall_97: 0.5436 - val_loss: 0.4908 - val_categorical_accuracy: 0.7812 - val_recall_97: 0.5890\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4850 - categorical_accuracy: 0.7622 - recall_97: 0.5231 - val_loss: 0.4878 - val_categorical_accuracy: 0.7865 - val_recall_97: 0.6027\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4837 - categorical_accuracy: 0.7656 - recall_97: 0.5333 - val_loss: 0.4869 - val_categorical_accuracy: 0.7812 - val_recall_97: 0.5890\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 34ms/step - loss: 0.4800 - categorical_accuracy: 0.7604 - recall_97: 0.5179 - val_loss: 0.4829 - val_categorical_accuracy: 0.7969 - val_recall_97: 0.6438\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4777 - categorical_accuracy: 0.7639 - recall_97: 0.5487 - val_loss: 0.4834 - val_categorical_accuracy: 0.7917 - val_recall_97: 0.6301\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4755 - categorical_accuracy: 0.7674 - recall_97: 0.5231 - val_loss: 0.4812 - val_categorical_accuracy: 0.7969 - val_recall_97: 0.6438\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4730 - categorical_accuracy: 0.7691 - recall_97: 0.5692 - val_loss: 0.4777 - val_categorical_accuracy: 0.8073 - val_recall_97: 0.6712\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4726 - categorical_accuracy: 0.7691 - recall_97: 0.5641 - val_loss: 0.4797 - val_categorical_accuracy: 0.7969 - val_recall_97: 0.6301\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.4693 - categorical_accuracy: 0.7743 - recall_97: 0.5282 - val_loss: 0.4773 - val_categorical_accuracy: 0.7917 - val_recall_97: 0.6301\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 22ms/step - loss: 0.4670 - categorical_accuracy: 0.7760 - recall_97: 0.5538 - val_loss: 0.4755 - val_categorical_accuracy: 0.7969 - val_recall_97: 0.6438\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4654 - categorical_accuracy: 0.7708 - recall_97: 0.5590 - val_loss: 0.4756 - val_categorical_accuracy: 0.8021 - val_recall_97: 0.6438\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4640 - categorical_accuracy: 0.7691 - recall_97: 0.5538 - val_loss: 0.4750 - val_categorical_accuracy: 0.7969 - val_recall_97: 0.6301\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4621 - categorical_accuracy: 0.7778 - recall_97: 0.5487 - val_loss: 0.4760 - val_categorical_accuracy: 0.7969 - val_recall_97: 0.6164\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4627 - categorical_accuracy: 0.7708 - recall_97: 0.5641 - val_loss: 0.4744 - val_categorical_accuracy: 0.7969 - val_recall_97: 0.6301\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.4603 - categorical_accuracy: 0.7708 - recall_97: 0.5231 - val_loss: 0.4749 - val_categorical_accuracy: 0.7969 - val_recall_97: 0.6164\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 4s 102ms/step - loss: 0.6901 - categorical_accuracy: 0.5312 - recall_98: 0.3282 - val_loss: 0.6781 - val_categorical_accuracy: 0.6198 - val_recall_98: 0.0000e+00\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 23ms/step - loss: 0.6569 - categorical_accuracy: 0.6615 - recall_98: 0.0000e+00 - val_loss: 0.6749 - val_categorical_accuracy: 0.6198 - val_recall_98: 0.0000e+00\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.6480 - categorical_accuracy: 0.6615 - recall_98: 0.0000e+00 - val_loss: 0.6673 - val_categorical_accuracy: 0.6198 - val_recall_98: 0.0000e+00\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6401 - categorical_accuracy: 0.6615 - recall_98: 0.0000e+00 - val_loss: 0.6530 - val_categorical_accuracy: 0.6198 - val_recall_98: 0.0000e+00\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.6287 - categorical_accuracy: 0.6615 - recall_98: 0.0000e+00 - val_loss: 0.6381 - val_categorical_accuracy: 0.6198 - val_recall_98: 0.0000e+00\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 25ms/step - loss: 0.6160 - categorical_accuracy: 0.6597 - recall_98: 0.0000e+00 - val_loss: 0.6210 - val_categorical_accuracy: 0.6302 - val_recall_98: 0.0274\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.6025 - categorical_accuracy: 0.6701 - recall_98: 0.0513 - val_loss: 0.6064 - val_categorical_accuracy: 0.6615 - val_recall_98: 0.1233\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5928 - categorical_accuracy: 0.6806 - recall_98: 0.0769 - val_loss: 0.5942 - val_categorical_accuracy: 0.6771 - val_recall_98: 0.1644\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.5807 - categorical_accuracy: 0.7135 - recall_98: 0.2667 - val_loss: 0.5787 - val_categorical_accuracy: 0.7135 - val_recall_98: 0.3014\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5704 - categorical_accuracy: 0.7188 - recall_98: 0.2359 - val_loss: 0.5692 - val_categorical_accuracy: 0.7135 - val_recall_98: 0.3014\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 22ms/step - loss: 0.5602 - categorical_accuracy: 0.7205 - recall_98: 0.3436 - val_loss: 0.5560 - val_categorical_accuracy: 0.7448 - val_recall_98: 0.4384\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.5501 - categorical_accuracy: 0.7274 - recall_98: 0.3128 - val_loss: 0.5499 - val_categorical_accuracy: 0.7396 - val_recall_98: 0.4247\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.5433 - categorical_accuracy: 0.7222 - recall_98: 0.4256 - val_loss: 0.5423 - val_categorical_accuracy: 0.7604 - val_recall_98: 0.4932\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5364 - categorical_accuracy: 0.7292 - recall_98: 0.3179 - val_loss: 0.5374 - val_categorical_accuracy: 0.7604 - val_recall_98: 0.4932\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5317 - categorical_accuracy: 0.7344 - recall_98: 0.5231 - val_loss: 0.5284 - val_categorical_accuracy: 0.7760 - val_recall_98: 0.5479\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5175 - categorical_accuracy: 0.7448 - recall_98: 0.4103 - val_loss: 0.5325 - val_categorical_accuracy: 0.7552 - val_recall_98: 0.4658\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5163 - categorical_accuracy: 0.7413 - recall_98: 0.3692 - val_loss: 0.5200 - val_categorical_accuracy: 0.7760 - val_recall_98: 0.5753\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5097 - categorical_accuracy: 0.7483 - recall_98: 0.5282 - val_loss: 0.5203 - val_categorical_accuracy: 0.7604 - val_recall_98: 0.5068\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 26ms/step - loss: 0.5015 - categorical_accuracy: 0.7535 - recall_98: 0.4462 - val_loss: 0.5145 - val_categorical_accuracy: 0.7760 - val_recall_98: 0.5616\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.4964 - categorical_accuracy: 0.7656 - recall_98: 0.5128 - val_loss: 0.5110 - val_categorical_accuracy: 0.7760 - val_recall_98: 0.5753\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4902 - categorical_accuracy: 0.7622 - recall_98: 0.5282 - val_loss: 0.5092 - val_categorical_accuracy: 0.7760 - val_recall_98: 0.5753\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4856 - categorical_accuracy: 0.7691 - recall_98: 0.5026 - val_loss: 0.5057 - val_categorical_accuracy: 0.7760 - val_recall_98: 0.5890\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4807 - categorical_accuracy: 0.7708 - recall_98: 0.5282 - val_loss: 0.5063 - val_categorical_accuracy: 0.7656 - val_recall_98: 0.5342\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4795 - categorical_accuracy: 0.7639 - recall_98: 0.5436 - val_loss: 0.5074 - val_categorical_accuracy: 0.7656 - val_recall_98: 0.5205\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4740 - categorical_accuracy: 0.7778 - recall_98: 0.4718 - val_loss: 0.5010 - val_categorical_accuracy: 0.8073 - val_recall_98: 0.6712\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4770 - categorical_accuracy: 0.7743 - recall_98: 0.6359 - val_loss: 0.5058 - val_categorical_accuracy: 0.7760 - val_recall_98: 0.5479\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4705 - categorical_accuracy: 0.7830 - recall_98: 0.5128 - val_loss: 0.5003 - val_categorical_accuracy: 0.7917 - val_recall_98: 0.6301\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 26ms/step - loss: 0.4636 - categorical_accuracy: 0.7830 - recall_98: 0.5179 - val_loss: 0.5002 - val_categorical_accuracy: 0.7969 - val_recall_98: 0.6301\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.4616 - categorical_accuracy: 0.7760 - recall_98: 0.6154 - val_loss: 0.4999 - val_categorical_accuracy: 0.7969 - val_recall_98: 0.6575\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4557 - categorical_accuracy: 0.7899 - recall_98: 0.5538 - val_loss: 0.5045 - val_categorical_accuracy: 0.7865 - val_recall_98: 0.5890\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4557 - categorical_accuracy: 0.7830 - recall_98: 0.5333 - val_loss: 0.4985 - val_categorical_accuracy: 0.7969 - val_recall_98: 0.6986\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4560 - categorical_accuracy: 0.7812 - recall_98: 0.5590 - val_loss: 0.4990 - val_categorical_accuracy: 0.7969 - val_recall_98: 0.6712\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4515 - categorical_accuracy: 0.7760 - recall_98: 0.6154 - val_loss: 0.5024 - val_categorical_accuracy: 0.7969 - val_recall_98: 0.6575\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4525 - categorical_accuracy: 0.7986 - recall_98: 0.5128 - val_loss: 0.5003 - val_categorical_accuracy: 0.7917 - val_recall_98: 0.6712\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4491 - categorical_accuracy: 0.7778 - recall_98: 0.6359 - val_loss: 0.5032 - val_categorical_accuracy: 0.7969 - val_recall_98: 0.6301\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 29ms/step - loss: 0.4450 - categorical_accuracy: 0.7882 - recall_98: 0.5436 - val_loss: 0.5028 - val_categorical_accuracy: 0.8021 - val_recall_98: 0.6575\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4441 - categorical_accuracy: 0.7795 - recall_98: 0.5641 - val_loss: 0.5025 - val_categorical_accuracy: 0.7969 - val_recall_98: 0.6712\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4426 - categorical_accuracy: 0.7865 - recall_98: 0.6256 - val_loss: 0.5042 - val_categorical_accuracy: 0.7865 - val_recall_98: 0.6438\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4445 - categorical_accuracy: 0.7865 - recall_98: 0.5436 - val_loss: 0.5046 - val_categorical_accuracy: 0.7760 - val_recall_98: 0.6712\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4406 - categorical_accuracy: 0.7865 - recall_98: 0.6308 - val_loss: 0.5161 - val_categorical_accuracy: 0.7865 - val_recall_98: 0.5753\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4461 - categorical_accuracy: 0.7812 - recall_98: 0.5487 - val_loss: 0.5056 - val_categorical_accuracy: 0.7812 - val_recall_98: 0.6849\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.4450 - categorical_accuracy: 0.7882 - recall_98: 0.5538 - val_loss: 0.5096 - val_categorical_accuracy: 0.7865 - val_recall_98: 0.6301\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4403 - categorical_accuracy: 0.7969 - recall_98: 0.6462 - val_loss: 0.5110 - val_categorical_accuracy: 0.7917 - val_recall_98: 0.6301\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4356 - categorical_accuracy: 0.7882 - recall_98: 0.6000 - val_loss: 0.5121 - val_categorical_accuracy: 0.7812 - val_recall_98: 0.6301\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 25ms/step - loss: 0.4367 - categorical_accuracy: 0.7934 - recall_98: 0.5590 - val_loss: 0.5095 - val_categorical_accuracy: 0.7812 - val_recall_98: 0.6575\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4371 - categorical_accuracy: 0.7882 - recall_98: 0.6103 - val_loss: 0.5116 - val_categorical_accuracy: 0.7812 - val_recall_98: 0.6301\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4332 - categorical_accuracy: 0.7951 - recall_98: 0.5641 - val_loss: 0.5162 - val_categorical_accuracy: 0.7917 - val_recall_98: 0.6164\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4315 - categorical_accuracy: 0.7917 - recall_98: 0.5897 - val_loss: 0.5116 - val_categorical_accuracy: 0.7865 - val_recall_98: 0.6986\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4345 - categorical_accuracy: 0.7899 - recall_98: 0.6103 - val_loss: 0.5146 - val_categorical_accuracy: 0.7760 - val_recall_98: 0.6438\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4303 - categorical_accuracy: 0.7951 - recall_98: 0.6154 - val_loss: 0.5162 - val_categorical_accuracy: 0.7760 - val_recall_98: 0.6301\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 5s 97ms/step - loss: 0.7240 - categorical_accuracy: 0.3403 - recall_99: 1.0000 - val_loss: 0.7021 - val_categorical_accuracy: 0.3854 - val_recall_99: 1.0000\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6997 - categorical_accuracy: 0.3889 - recall_99: 0.9692 - val_loss: 0.6866 - val_categorical_accuracy: 0.6875 - val_recall_99: 0.6712\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 23ms/step - loss: 0.6794 - categorical_accuracy: 0.6771 - recall_99: 0.1744 - val_loss: 0.6771 - val_categorical_accuracy: 0.6198 - val_recall_99: 0.0000e+00\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6670 - categorical_accuracy: 0.6615 - recall_99: 0.0000e+00 - val_loss: 0.6714 - val_categorical_accuracy: 0.6198 - val_recall_99: 0.0000e+00\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.6560 - categorical_accuracy: 0.6615 - recall_99: 0.0000e+00 - val_loss: 0.6685 - val_categorical_accuracy: 0.6198 - val_recall_99: 0.0000e+00\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6487 - categorical_accuracy: 0.6615 - recall_99: 0.0000e+00 - val_loss: 0.6668 - val_categorical_accuracy: 0.6198 - val_recall_99: 0.0000e+00\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6440 - categorical_accuracy: 0.6615 - recall_99: 0.0000e+00 - val_loss: 0.6665 - val_categorical_accuracy: 0.6198 - val_recall_99: 0.0000e+00\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 24ms/step - loss: 0.6399 - categorical_accuracy: 0.6615 - recall_99: 0.0000e+00 - val_loss: 0.6651 - val_categorical_accuracy: 0.6198 - val_recall_99: 0.0000e+00\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6371 - categorical_accuracy: 0.6615 - recall_99: 0.0000e+00 - val_loss: 0.6631 - val_categorical_accuracy: 0.6198 - val_recall_99: 0.0000e+00\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.6346 - categorical_accuracy: 0.6615 - recall_99: 0.0000e+00 - val_loss: 0.6611 - val_categorical_accuracy: 0.6198 - val_recall_99: 0.0000e+00\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6315 - categorical_accuracy: 0.6615 - recall_99: 0.0000e+00 - val_loss: 0.6581 - val_categorical_accuracy: 0.6198 - val_recall_99: 0.0000e+00\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6277 - categorical_accuracy: 0.6615 - recall_99: 0.0000e+00 - val_loss: 0.6542 - val_categorical_accuracy: 0.6198 - val_recall_99: 0.0000e+00\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.6222 - categorical_accuracy: 0.6615 - recall_99: 0.0000e+00 - val_loss: 0.6463 - val_categorical_accuracy: 0.6198 - val_recall_99: 0.0000e+00\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.6155 - categorical_accuracy: 0.6615 - recall_99: 0.0000e+00 - val_loss: 0.6400 - val_categorical_accuracy: 0.6198 - val_recall_99: 0.0000e+00\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 28ms/step - loss: 0.6103 - categorical_accuracy: 0.6615 - recall_99: 0.0000e+00 - val_loss: 0.6345 - val_categorical_accuracy: 0.6198 - val_recall_99: 0.0000e+00\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.6051 - categorical_accuracy: 0.6615 - recall_99: 0.0000e+00 - val_loss: 0.6292 - val_categorical_accuracy: 0.6198 - val_recall_99: 0.0000e+00\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 22ms/step - loss: 0.5999 - categorical_accuracy: 0.6615 - recall_99: 0.0000e+00 - val_loss: 0.6214 - val_categorical_accuracy: 0.6198 - val_recall_99: 0.0000e+00\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5940 - categorical_accuracy: 0.6615 - recall_99: 0.0000e+00 - val_loss: 0.6165 - val_categorical_accuracy: 0.6198 - val_recall_99: 0.0000e+00\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5888 - categorical_accuracy: 0.6632 - recall_99: 0.0051 - val_loss: 0.6103 - val_categorical_accuracy: 0.6198 - val_recall_99: 0.0000e+00\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5831 - categorical_accuracy: 0.6736 - recall_99: 0.0462 - val_loss: 0.6014 - val_categorical_accuracy: 0.6510 - val_recall_99: 0.0822\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.5776 - categorical_accuracy: 0.6944 - recall_99: 0.1231 - val_loss: 0.5951 - val_categorical_accuracy: 0.6719 - val_recall_99: 0.1507\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5708 - categorical_accuracy: 0.6875 - recall_99: 0.1026 - val_loss: 0.5919 - val_categorical_accuracy: 0.6615 - val_recall_99: 0.1233\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5656 - categorical_accuracy: 0.6944 - recall_99: 0.1487 - val_loss: 0.5812 - val_categorical_accuracy: 0.7031 - val_recall_99: 0.2466\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 22ms/step - loss: 0.5590 - categorical_accuracy: 0.7170 - recall_99: 0.2205 - val_loss: 0.5762 - val_categorical_accuracy: 0.7031 - val_recall_99: 0.2466\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 21ms/step - loss: 0.5532 - categorical_accuracy: 0.7240 - recall_99: 0.2410 - val_loss: 0.5699 - val_categorical_accuracy: 0.7083 - val_recall_99: 0.2740\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.5463 - categorical_accuracy: 0.7309 - recall_99: 0.2821 - val_loss: 0.5617 - val_categorical_accuracy: 0.7240 - val_recall_99: 0.3288\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5414 - categorical_accuracy: 0.7257 - recall_99: 0.2769 - val_loss: 0.5562 - val_categorical_accuracy: 0.7188 - val_recall_99: 0.3288\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 24ms/step - loss: 0.5335 - categorical_accuracy: 0.7378 - recall_99: 0.3487 - val_loss: 0.5457 - val_categorical_accuracy: 0.7448 - val_recall_99: 0.4110\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5282 - categorical_accuracy: 0.7396 - recall_99: 0.3641 - val_loss: 0.5409 - val_categorical_accuracy: 0.7448 - val_recall_99: 0.4110\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5221 - categorical_accuracy: 0.7431 - recall_99: 0.3744 - val_loss: 0.5337 - val_categorical_accuracy: 0.7656 - val_recall_99: 0.4658\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.5158 - categorical_accuracy: 0.7517 - recall_99: 0.4308 - val_loss: 0.5264 - val_categorical_accuracy: 0.7760 - val_recall_99: 0.5205\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5114 - categorical_accuracy: 0.7604 - recall_99: 0.4615 - val_loss: 0.5230 - val_categorical_accuracy: 0.7708 - val_recall_99: 0.4932\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.5091 - categorical_accuracy: 0.7413 - recall_99: 0.3795 - val_loss: 0.5194 - val_categorical_accuracy: 0.7604 - val_recall_99: 0.4658\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5015 - categorical_accuracy: 0.7691 - recall_99: 0.4923 - val_loss: 0.5111 - val_categorical_accuracy: 0.7604 - val_recall_99: 0.5205\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 24ms/step - loss: 0.4972 - categorical_accuracy: 0.7708 - recall_99: 0.5026 - val_loss: 0.5107 - val_categorical_accuracy: 0.7656 - val_recall_99: 0.4932\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4933 - categorical_accuracy: 0.7708 - recall_99: 0.4718 - val_loss: 0.5106 - val_categorical_accuracy: 0.7708 - val_recall_99: 0.4932\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4908 - categorical_accuracy: 0.7674 - recall_99: 0.5128 - val_loss: 0.4994 - val_categorical_accuracy: 0.7865 - val_recall_99: 0.6027\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4845 - categorical_accuracy: 0.7726 - recall_99: 0.5077 - val_loss: 0.5046 - val_categorical_accuracy: 0.7708 - val_recall_99: 0.5068\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4818 - categorical_accuracy: 0.7708 - recall_99: 0.4974 - val_loss: 0.4960 - val_categorical_accuracy: 0.7812 - val_recall_99: 0.5890\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 27ms/step - loss: 0.4784 - categorical_accuracy: 0.7674 - recall_99: 0.5077 - val_loss: 0.4952 - val_categorical_accuracy: 0.7917 - val_recall_99: 0.5890\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4749 - categorical_accuracy: 0.7691 - recall_99: 0.5179 - val_loss: 0.4938 - val_categorical_accuracy: 0.7708 - val_recall_99: 0.5616\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.4728 - categorical_accuracy: 0.7726 - recall_99: 0.5385 - val_loss: 0.4946 - val_categorical_accuracy: 0.7760 - val_recall_99: 0.5616\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4690 - categorical_accuracy: 0.7760 - recall_99: 0.5282 - val_loss: 0.4959 - val_categorical_accuracy: 0.7656 - val_recall_99: 0.5479\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4687 - categorical_accuracy: 0.7795 - recall_99: 0.5077 - val_loss: 0.4952 - val_categorical_accuracy: 0.7708 - val_recall_99: 0.5616\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4707 - categorical_accuracy: 0.7691 - recall_99: 0.5795 - val_loss: 0.4891 - val_categorical_accuracy: 0.7812 - val_recall_99: 0.6164\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4625 - categorical_accuracy: 0.7778 - recall_99: 0.5282 - val_loss: 0.5004 - val_categorical_accuracy: 0.7604 - val_recall_99: 0.5342\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4618 - categorical_accuracy: 0.7795 - recall_99: 0.5282 - val_loss: 0.4906 - val_categorical_accuracy: 0.7760 - val_recall_99: 0.6027\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4591 - categorical_accuracy: 0.7847 - recall_99: 0.5897 - val_loss: 0.4896 - val_categorical_accuracy: 0.7760 - val_recall_99: 0.6027\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 22ms/step - loss: 0.4584 - categorical_accuracy: 0.7830 - recall_99: 0.5538 - val_loss: 0.4937 - val_categorical_accuracy: 0.7812 - val_recall_99: 0.6027\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4561 - categorical_accuracy: 0.7812 - recall_99: 0.5641 - val_loss: 0.4917 - val_categorical_accuracy: 0.7760 - val_recall_99: 0.6164\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 5s 99ms/step - loss: 0.6785 - categorical_accuracy: 0.6528 - recall_100: 0.0462 - val_loss: 0.6765 - val_categorical_accuracy: 0.6198 - val_recall_100: 0.0000e+00\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 22ms/step - loss: 0.6585 - categorical_accuracy: 0.6615 - recall_100: 0.0000e+00 - val_loss: 0.6740 - val_categorical_accuracy: 0.6198 - val_recall_100: 0.0000e+00\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6506 - categorical_accuracy: 0.6615 - recall_100: 0.0000e+00 - val_loss: 0.6693 - val_categorical_accuracy: 0.6198 - val_recall_100: 0.0000e+00\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.6452 - categorical_accuracy: 0.6615 - recall_100: 0.0000e+00 - val_loss: 0.6622 - val_categorical_accuracy: 0.6198 - val_recall_100: 0.0000e+00\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 27ms/step - loss: 0.6394 - categorical_accuracy: 0.6615 - recall_100: 0.0000e+00 - val_loss: 0.6589 - val_categorical_accuracy: 0.6198 - val_recall_100: 0.0000e+00\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6343 - categorical_accuracy: 0.6615 - recall_100: 0.0000e+00 - val_loss: 0.6520 - val_categorical_accuracy: 0.6198 - val_recall_100: 0.0000e+00\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6275 - categorical_accuracy: 0.6615 - recall_100: 0.0000e+00 - val_loss: 0.6424 - val_categorical_accuracy: 0.6250 - val_recall_100: 0.0137\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6220 - categorical_accuracy: 0.6719 - recall_100: 0.0308 - val_loss: 0.6343 - val_categorical_accuracy: 0.6302 - val_recall_100: 0.0274\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6152 - categorical_accuracy: 0.6719 - recall_100: 0.0359 - val_loss: 0.6273 - val_categorical_accuracy: 0.6458 - val_recall_100: 0.0685\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 27ms/step - loss: 0.6096 - categorical_accuracy: 0.6719 - recall_100: 0.0359 - val_loss: 0.6202 - val_categorical_accuracy: 0.6458 - val_recall_100: 0.0822\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6013 - categorical_accuracy: 0.6736 - recall_100: 0.0667 - val_loss: 0.6142 - val_categorical_accuracy: 0.6510 - val_recall_100: 0.0959\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5951 - categorical_accuracy: 0.6719 - recall_100: 0.1026 - val_loss: 0.6076 - val_categorical_accuracy: 0.6771 - val_recall_100: 0.1781\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5890 - categorical_accuracy: 0.6736 - recall_100: 0.1077 - val_loss: 0.6012 - val_categorical_accuracy: 0.6771 - val_recall_100: 0.1918\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 21ms/step - loss: 0.5832 - categorical_accuracy: 0.6788 - recall_100: 0.2000 - val_loss: 0.5926 - val_categorical_accuracy: 0.6823 - val_recall_100: 0.2329\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.5759 - categorical_accuracy: 0.6910 - recall_100: 0.2308 - val_loss: 0.5894 - val_categorical_accuracy: 0.6823 - val_recall_100: 0.2329\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5701 - categorical_accuracy: 0.6892 - recall_100: 0.2615 - val_loss: 0.5803 - val_categorical_accuracy: 0.7031 - val_recall_100: 0.3425\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5632 - categorical_accuracy: 0.6944 - recall_100: 0.2872 - val_loss: 0.5775 - val_categorical_accuracy: 0.6875 - val_recall_100: 0.2877\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5588 - categorical_accuracy: 0.7083 - recall_100: 0.3385 - val_loss: 0.5697 - val_categorical_accuracy: 0.7135 - val_recall_100: 0.3836\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5495 - categorical_accuracy: 0.7049 - recall_100: 0.3231 - val_loss: 0.5685 - val_categorical_accuracy: 0.7031 - val_recall_100: 0.2877\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 24ms/step - loss: 0.5454 - categorical_accuracy: 0.7170 - recall_100: 0.3179 - val_loss: 0.5569 - val_categorical_accuracy: 0.7135 - val_recall_100: 0.3699\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.5371 - categorical_accuracy: 0.7188 - recall_100: 0.3436 - val_loss: 0.5501 - val_categorical_accuracy: 0.7292 - val_recall_100: 0.4247\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5306 - categorical_accuracy: 0.7274 - recall_100: 0.3744 - val_loss: 0.5439 - val_categorical_accuracy: 0.7292 - val_recall_100: 0.4247\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5220 - categorical_accuracy: 0.7431 - recall_100: 0.3897 - val_loss: 0.5374 - val_categorical_accuracy: 0.7344 - val_recall_100: 0.4521\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5176 - categorical_accuracy: 0.7448 - recall_100: 0.4000 - val_loss: 0.5293 - val_categorical_accuracy: 0.7604 - val_recall_100: 0.5479\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5120 - categorical_accuracy: 0.7431 - recall_100: 0.4000 - val_loss: 0.5264 - val_categorical_accuracy: 0.7448 - val_recall_100: 0.4658\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.5091 - categorical_accuracy: 0.7552 - recall_100: 0.5333 - val_loss: 0.5220 - val_categorical_accuracy: 0.7656 - val_recall_100: 0.5205\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4998 - categorical_accuracy: 0.7604 - recall_100: 0.4564 - val_loss: 0.5174 - val_categorical_accuracy: 0.7604 - val_recall_100: 0.5205\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 21ms/step - loss: 0.4961 - categorical_accuracy: 0.7622 - recall_100: 0.4256 - val_loss: 0.5121 - val_categorical_accuracy: 0.7656 - val_recall_100: 0.5479\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4915 - categorical_accuracy: 0.7691 - recall_100: 0.5333 - val_loss: 0.5112 - val_categorical_accuracy: 0.7656 - val_recall_100: 0.5068\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4899 - categorical_accuracy: 0.7552 - recall_100: 0.3846 - val_loss: 0.5064 - val_categorical_accuracy: 0.7760 - val_recall_100: 0.5479\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4835 - categorical_accuracy: 0.7674 - recall_100: 0.6000 - val_loss: 0.5017 - val_categorical_accuracy: 0.7969 - val_recall_100: 0.6712\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4800 - categorical_accuracy: 0.7812 - recall_100: 0.4974 - val_loss: 0.5026 - val_categorical_accuracy: 0.7865 - val_recall_100: 0.5753\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4743 - categorical_accuracy: 0.7847 - recall_100: 0.5487 - val_loss: 0.4986 - val_categorical_accuracy: 0.8125 - val_recall_100: 0.6575\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4715 - categorical_accuracy: 0.7865 - recall_100: 0.5179 - val_loss: 0.4968 - val_categorical_accuracy: 0.8125 - val_recall_100: 0.6575\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 22ms/step - loss: 0.4684 - categorical_accuracy: 0.7812 - recall_100: 0.5590 - val_loss: 0.4936 - val_categorical_accuracy: 0.8073 - val_recall_100: 0.6712\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4679 - categorical_accuracy: 0.7760 - recall_100: 0.5128 - val_loss: 0.4933 - val_categorical_accuracy: 0.8073 - val_recall_100: 0.6712\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4636 - categorical_accuracy: 0.7812 - recall_100: 0.5641 - val_loss: 0.4928 - val_categorical_accuracy: 0.8125 - val_recall_100: 0.6712\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.4628 - categorical_accuracy: 0.7778 - recall_100: 0.5538 - val_loss: 0.4961 - val_categorical_accuracy: 0.7917 - val_recall_100: 0.6027\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4613 - categorical_accuracy: 0.7760 - recall_100: 0.5385 - val_loss: 0.4895 - val_categorical_accuracy: 0.8021 - val_recall_100: 0.6849\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.4621 - categorical_accuracy: 0.7847 - recall_100: 0.5128 - val_loss: 0.4897 - val_categorical_accuracy: 0.8021 - val_recall_100: 0.6712\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.4588 - categorical_accuracy: 0.7812 - recall_100: 0.5744 - val_loss: 0.4929 - val_categorical_accuracy: 0.8073 - val_recall_100: 0.6575\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4535 - categorical_accuracy: 0.7795 - recall_100: 0.5590 - val_loss: 0.4891 - val_categorical_accuracy: 0.8073 - val_recall_100: 0.6849\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 26ms/step - loss: 0.4522 - categorical_accuracy: 0.7847 - recall_100: 0.5897 - val_loss: 0.4917 - val_categorical_accuracy: 0.7917 - val_recall_100: 0.6164\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.4543 - categorical_accuracy: 0.7812 - recall_100: 0.5590 - val_loss: 0.4900 - val_categorical_accuracy: 0.8021 - val_recall_100: 0.6438\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4542 - categorical_accuracy: 0.7795 - recall_100: 0.5231 - val_loss: 0.4886 - val_categorical_accuracy: 0.8073 - val_recall_100: 0.6986\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4531 - categorical_accuracy: 0.7865 - recall_100: 0.5538 - val_loss: 0.4895 - val_categorical_accuracy: 0.7969 - val_recall_100: 0.6438\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4485 - categorical_accuracy: 0.7812 - recall_100: 0.6103 - val_loss: 0.4901 - val_categorical_accuracy: 0.8021 - val_recall_100: 0.6438\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4475 - categorical_accuracy: 0.7899 - recall_100: 0.5692 - val_loss: 0.4895 - val_categorical_accuracy: 0.7917 - val_recall_100: 0.6438\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4503 - categorical_accuracy: 0.7865 - recall_100: 0.5282 - val_loss: 0.4889 - val_categorical_accuracy: 0.8021 - val_recall_100: 0.6849\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4475 - categorical_accuracy: 0.7882 - recall_100: 0.6667 - val_loss: 0.4907 - val_categorical_accuracy: 0.7969 - val_recall_100: 0.6575\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 5s 96ms/step - loss: 0.6994 - categorical_accuracy: 0.4635 - recall_101: 0.6308 - val_loss: 0.6833 - val_categorical_accuracy: 0.6302 - val_recall_101: 0.0274\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 25ms/step - loss: 0.6687 - categorical_accuracy: 0.6597 - recall_101: 0.0000e+00 - val_loss: 0.6736 - val_categorical_accuracy: 0.6198 - val_recall_101: 0.0000e+00\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6556 - categorical_accuracy: 0.6615 - recall_101: 0.0000e+00 - val_loss: 0.6715 - val_categorical_accuracy: 0.6198 - val_recall_101: 0.0000e+00\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6511 - categorical_accuracy: 0.6615 - recall_101: 0.0000e+00 - val_loss: 0.6712 - val_categorical_accuracy: 0.6198 - val_recall_101: 0.0000e+00\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.6479 - categorical_accuracy: 0.6615 - recall_101: 0.0000e+00 - val_loss: 0.6699 - val_categorical_accuracy: 0.6198 - val_recall_101: 0.0000e+00\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 25ms/step - loss: 0.6455 - categorical_accuracy: 0.6615 - recall_101: 0.0000e+00 - val_loss: 0.6682 - val_categorical_accuracy: 0.6198 - val_recall_101: 0.0000e+00\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6432 - categorical_accuracy: 0.6615 - recall_101: 0.0000e+00 - val_loss: 0.6670 - val_categorical_accuracy: 0.6198 - val_recall_101: 0.0000e+00\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.6411 - categorical_accuracy: 0.6615 - recall_101: 0.0000e+00 - val_loss: 0.6644 - val_categorical_accuracy: 0.6198 - val_recall_101: 0.0000e+00\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6386 - categorical_accuracy: 0.6615 - recall_101: 0.0000e+00 - val_loss: 0.6622 - val_categorical_accuracy: 0.6198 - val_recall_101: 0.0000e+00\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 23ms/step - loss: 0.6363 - categorical_accuracy: 0.6615 - recall_101: 0.0000e+00 - val_loss: 0.6603 - val_categorical_accuracy: 0.6198 - val_recall_101: 0.0000e+00\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6338 - categorical_accuracy: 0.6615 - recall_101: 0.0000e+00 - val_loss: 0.6572 - val_categorical_accuracy: 0.6198 - val_recall_101: 0.0000e+00\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6312 - categorical_accuracy: 0.6615 - recall_101: 0.0000e+00 - val_loss: 0.6540 - val_categorical_accuracy: 0.6198 - val_recall_101: 0.0000e+00\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.6288 - categorical_accuracy: 0.6615 - recall_101: 0.0000e+00 - val_loss: 0.6510 - val_categorical_accuracy: 0.6198 - val_recall_101: 0.0000e+00\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6255 - categorical_accuracy: 0.6615 - recall_101: 0.0000e+00 - val_loss: 0.6471 - val_categorical_accuracy: 0.6198 - val_recall_101: 0.0000e+00\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6223 - categorical_accuracy: 0.6615 - recall_101: 0.0000e+00 - val_loss: 0.6440 - val_categorical_accuracy: 0.6198 - val_recall_101: 0.0000e+00\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.6192 - categorical_accuracy: 0.6615 - recall_101: 0.0000e+00 - val_loss: 0.6406 - val_categorical_accuracy: 0.6198 - val_recall_101: 0.0000e+00\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6152 - categorical_accuracy: 0.6615 - recall_101: 0.0000e+00 - val_loss: 0.6368 - val_categorical_accuracy: 0.6198 - val_recall_101: 0.0000e+00\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.6115 - categorical_accuracy: 0.6615 - recall_101: 0.0000e+00 - val_loss: 0.6314 - val_categorical_accuracy: 0.6198 - val_recall_101: 0.0000e+00\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 35ms/step - loss: 0.6073 - categorical_accuracy: 0.6615 - recall_101: 0.0000e+00 - val_loss: 0.6281 - val_categorical_accuracy: 0.6198 - val_recall_101: 0.0000e+00\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.6040 - categorical_accuracy: 0.6667 - recall_101: 0.0205 - val_loss: 0.6218 - val_categorical_accuracy: 0.6406 - val_recall_101: 0.0548\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5985 - categorical_accuracy: 0.6684 - recall_101: 0.0256 - val_loss: 0.6188 - val_categorical_accuracy: 0.6406 - val_recall_101: 0.0548\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5942 - categorical_accuracy: 0.6701 - recall_101: 0.0308 - val_loss: 0.6150 - val_categorical_accuracy: 0.6354 - val_recall_101: 0.0548\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5891 - categorical_accuracy: 0.6719 - recall_101: 0.0410 - val_loss: 0.6096 - val_categorical_accuracy: 0.6458 - val_recall_101: 0.0822\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5843 - categorical_accuracy: 0.6823 - recall_101: 0.0718 - val_loss: 0.6039 - val_categorical_accuracy: 0.6667 - val_recall_101: 0.1370\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5795 - categorical_accuracy: 0.7031 - recall_101: 0.1436 - val_loss: 0.5969 - val_categorical_accuracy: 0.6875 - val_recall_101: 0.2055\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5725 - categorical_accuracy: 0.7049 - recall_101: 0.1487 - val_loss: 0.5923 - val_categorical_accuracy: 0.6875 - val_recall_101: 0.2055\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.5667 - categorical_accuracy: 0.7031 - recall_101: 0.1487 - val_loss: 0.5860 - val_categorical_accuracy: 0.6875 - val_recall_101: 0.2192\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5598 - categorical_accuracy: 0.7118 - recall_101: 0.2000 - val_loss: 0.5787 - val_categorical_accuracy: 0.6979 - val_recall_101: 0.2740\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 27ms/step - loss: 0.5523 - categorical_accuracy: 0.7205 - recall_101: 0.2154 - val_loss: 0.5746 - val_categorical_accuracy: 0.7083 - val_recall_101: 0.2877\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5440 - categorical_accuracy: 0.7431 - recall_101: 0.2821 - val_loss: 0.5631 - val_categorical_accuracy: 0.7188 - val_recall_101: 0.3288\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5355 - categorical_accuracy: 0.7465 - recall_101: 0.3128 - val_loss: 0.5528 - val_categorical_accuracy: 0.7083 - val_recall_101: 0.3425\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5282 - categorical_accuracy: 0.7656 - recall_101: 0.4410 - val_loss: 0.5441 - val_categorical_accuracy: 0.7656 - val_recall_101: 0.5068\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5188 - categorical_accuracy: 0.7569 - recall_101: 0.3744 - val_loss: 0.5435 - val_categorical_accuracy: 0.7135 - val_recall_101: 0.3288\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5122 - categorical_accuracy: 0.7622 - recall_101: 0.3744 - val_loss: 0.5344 - val_categorical_accuracy: 0.7500 - val_recall_101: 0.4658\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.5067 - categorical_accuracy: 0.7656 - recall_101: 0.4205 - val_loss: 0.5282 - val_categorical_accuracy: 0.7656 - val_recall_101: 0.5068\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5030 - categorical_accuracy: 0.7812 - recall_101: 0.5333 - val_loss: 0.5265 - val_categorical_accuracy: 0.7604 - val_recall_101: 0.4932\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4940 - categorical_accuracy: 0.7812 - recall_101: 0.4513 - val_loss: 0.5235 - val_categorical_accuracy: 0.7656 - val_recall_101: 0.4932\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 29ms/step - loss: 0.4896 - categorical_accuracy: 0.7795 - recall_101: 0.4974 - val_loss: 0.5164 - val_categorical_accuracy: 0.7552 - val_recall_101: 0.5205\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4847 - categorical_accuracy: 0.7917 - recall_101: 0.5128 - val_loss: 0.5172 - val_categorical_accuracy: 0.7708 - val_recall_101: 0.5205\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4810 - categorical_accuracy: 0.7899 - recall_101: 0.5333 - val_loss: 0.5125 - val_categorical_accuracy: 0.7656 - val_recall_101: 0.5479\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4795 - categorical_accuracy: 0.7882 - recall_101: 0.4718 - val_loss: 0.5106 - val_categorical_accuracy: 0.7656 - val_recall_101: 0.5342\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4730 - categorical_accuracy: 0.7934 - recall_101: 0.5590 - val_loss: 0.5088 - val_categorical_accuracy: 0.7604 - val_recall_101: 0.5342\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4699 - categorical_accuracy: 0.7934 - recall_101: 0.5179 - val_loss: 0.5080 - val_categorical_accuracy: 0.7656 - val_recall_101: 0.5342\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4671 - categorical_accuracy: 0.7934 - recall_101: 0.5744 - val_loss: 0.5044 - val_categorical_accuracy: 0.7604 - val_recall_101: 0.5616\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 25ms/step - loss: 0.4668 - categorical_accuracy: 0.7882 - recall_101: 0.5077 - val_loss: 0.5051 - val_categorical_accuracy: 0.7656 - val_recall_101: 0.5342\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4607 - categorical_accuracy: 0.7882 - recall_101: 0.5692 - val_loss: 0.5035 - val_categorical_accuracy: 0.7760 - val_recall_101: 0.5753\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4587 - categorical_accuracy: 0.7986 - recall_101: 0.5590 - val_loss: 0.5052 - val_categorical_accuracy: 0.7865 - val_recall_101: 0.5753\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4555 - categorical_accuracy: 0.7969 - recall_101: 0.5692 - val_loss: 0.5021 - val_categorical_accuracy: 0.7708 - val_recall_101: 0.5753\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 26ms/step - loss: 0.4555 - categorical_accuracy: 0.7882 - recall_101: 0.5487 - val_loss: 0.5010 - val_categorical_accuracy: 0.7708 - val_recall_101: 0.5753\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4522 - categorical_accuracy: 0.7917 - recall_101: 0.5795 - val_loss: 0.5027 - val_categorical_accuracy: 0.7760 - val_recall_101: 0.5753\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 5s 138ms/step - loss: 0.6768 - categorical_accuracy: 0.6562 - recall_102: 0.0923 - val_loss: 0.6755 - val_categorical_accuracy: 0.6250 - val_recall_102: 0.0274\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6615 - categorical_accuracy: 0.6615 - recall_102: 0.0000e+00 - val_loss: 0.6696 - val_categorical_accuracy: 0.6198 - val_recall_102: 0.0000e+00\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.6515 - categorical_accuracy: 0.6615 - recall_102: 0.0000e+00 - val_loss: 0.6648 - val_categorical_accuracy: 0.6198 - val_recall_102: 0.0000e+00\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6424 - categorical_accuracy: 0.6615 - recall_102: 0.0000e+00 - val_loss: 0.6604 - val_categorical_accuracy: 0.6198 - val_recall_102: 0.0000e+00\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 22ms/step - loss: 0.6360 - categorical_accuracy: 0.6615 - recall_102: 0.0000e+00 - val_loss: 0.6539 - val_categorical_accuracy: 0.6198 - val_recall_102: 0.0000e+00\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 22ms/step - loss: 0.6291 - categorical_accuracy: 0.6597 - recall_102: 0.0000e+00 - val_loss: 0.6455 - val_categorical_accuracy: 0.6198 - val_recall_102: 0.0000e+00\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6216 - categorical_accuracy: 0.6615 - recall_102: 0.0051 - val_loss: 0.6344 - val_categorical_accuracy: 0.6406 - val_recall_102: 0.0548\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 23ms/step - loss: 0.6141 - categorical_accuracy: 0.6667 - recall_102: 0.0256 - val_loss: 0.6263 - val_categorical_accuracy: 0.6406 - val_recall_102: 0.0685\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.6067 - categorical_accuracy: 0.6684 - recall_102: 0.0308 - val_loss: 0.6206 - val_categorical_accuracy: 0.6406 - val_recall_102: 0.0685\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5989 - categorical_accuracy: 0.6753 - recall_102: 0.0718 - val_loss: 0.6111 - val_categorical_accuracy: 0.6406 - val_recall_102: 0.0959\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5913 - categorical_accuracy: 0.6823 - recall_102: 0.0923 - val_loss: 0.6062 - val_categorical_accuracy: 0.6458 - val_recall_102: 0.1096\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5853 - categorical_accuracy: 0.6944 - recall_102: 0.1692 - val_loss: 0.5966 - val_categorical_accuracy: 0.6562 - val_recall_102: 0.1918\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5764 - categorical_accuracy: 0.6979 - recall_102: 0.2051 - val_loss: 0.5910 - val_categorical_accuracy: 0.6615 - val_recall_102: 0.2055\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5700 - categorical_accuracy: 0.7014 - recall_102: 0.2205 - val_loss: 0.5835 - val_categorical_accuracy: 0.7083 - val_recall_102: 0.3288\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5635 - categorical_accuracy: 0.7049 - recall_102: 0.2667 - val_loss: 0.5760 - val_categorical_accuracy: 0.7188 - val_recall_102: 0.3699\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5566 - categorical_accuracy: 0.7083 - recall_102: 0.2821 - val_loss: 0.5717 - val_categorical_accuracy: 0.7188 - val_recall_102: 0.3699\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 26ms/step - loss: 0.5507 - categorical_accuracy: 0.7014 - recall_102: 0.2513 - val_loss: 0.5685 - val_categorical_accuracy: 0.7135 - val_recall_102: 0.3425\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5446 - categorical_accuracy: 0.7118 - recall_102: 0.3128 - val_loss: 0.5607 - val_categorical_accuracy: 0.7500 - val_recall_102: 0.4658\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5385 - categorical_accuracy: 0.7153 - recall_102: 0.3333 - val_loss: 0.5575 - val_categorical_accuracy: 0.7344 - val_recall_102: 0.4247\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5329 - categorical_accuracy: 0.7170 - recall_102: 0.3949 - val_loss: 0.5516 - val_categorical_accuracy: 0.7292 - val_recall_102: 0.4795\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 21ms/step - loss: 0.5276 - categorical_accuracy: 0.7222 - recall_102: 0.3744 - val_loss: 0.5510 - val_categorical_accuracy: 0.7448 - val_recall_102: 0.4384\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5221 - categorical_accuracy: 0.7274 - recall_102: 0.4000 - val_loss: 0.5438 - val_categorical_accuracy: 0.7292 - val_recall_102: 0.5068\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.5194 - categorical_accuracy: 0.7361 - recall_102: 0.5179 - val_loss: 0.5396 - val_categorical_accuracy: 0.7344 - val_recall_102: 0.5342\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.5137 - categorical_accuracy: 0.7344 - recall_102: 0.4205 - val_loss: 0.5416 - val_categorical_accuracy: 0.7552 - val_recall_102: 0.4932\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5088 - categorical_accuracy: 0.7413 - recall_102: 0.4667 - val_loss: 0.5322 - val_categorical_accuracy: 0.7396 - val_recall_102: 0.6027\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.5046 - categorical_accuracy: 0.7500 - recall_102: 0.5282 - val_loss: 0.5314 - val_categorical_accuracy: 0.7396 - val_recall_102: 0.5479\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 24ms/step - loss: 0.5007 - categorical_accuracy: 0.7500 - recall_102: 0.4821 - val_loss: 0.5276 - val_categorical_accuracy: 0.7500 - val_recall_102: 0.5753\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 20ms/step - loss: 0.5010 - categorical_accuracy: 0.7431 - recall_102: 0.4615 - val_loss: 0.5244 - val_categorical_accuracy: 0.7604 - val_recall_102: 0.5890\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4987 - categorical_accuracy: 0.7587 - recall_102: 0.6256 - val_loss: 0.5195 - val_categorical_accuracy: 0.7552 - val_recall_102: 0.6301\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4917 - categorical_accuracy: 0.7535 - recall_102: 0.5077 - val_loss: 0.5206 - val_categorical_accuracy: 0.7812 - val_recall_102: 0.5890\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 21ms/step - loss: 0.4877 - categorical_accuracy: 0.7569 - recall_102: 0.5231 - val_loss: 0.5147 - val_categorical_accuracy: 0.7760 - val_recall_102: 0.6027\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4851 - categorical_accuracy: 0.7656 - recall_102: 0.5744 - val_loss: 0.5107 - val_categorical_accuracy: 0.7760 - val_recall_102: 0.6164\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4832 - categorical_accuracy: 0.7691 - recall_102: 0.5385 - val_loss: 0.5117 - val_categorical_accuracy: 0.7812 - val_recall_102: 0.5890\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 24ms/step - loss: 0.4804 - categorical_accuracy: 0.7552 - recall_102: 0.5846 - val_loss: 0.5066 - val_categorical_accuracy: 0.7812 - val_recall_102: 0.6438\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4753 - categorical_accuracy: 0.7656 - recall_102: 0.5538 - val_loss: 0.5077 - val_categorical_accuracy: 0.7865 - val_recall_102: 0.5890\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4762 - categorical_accuracy: 0.7656 - recall_102: 0.4769 - val_loss: 0.5016 - val_categorical_accuracy: 0.8021 - val_recall_102: 0.6712\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4777 - categorical_accuracy: 0.7587 - recall_102: 0.6359 - val_loss: 0.5002 - val_categorical_accuracy: 0.8021 - val_recall_102: 0.6575\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4689 - categorical_accuracy: 0.7674 - recall_102: 0.5590 - val_loss: 0.5024 - val_categorical_accuracy: 0.7917 - val_recall_102: 0.6027\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4683 - categorical_accuracy: 0.7760 - recall_102: 0.5026 - val_loss: 0.4998 - val_categorical_accuracy: 0.8021 - val_recall_102: 0.6301\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4646 - categorical_accuracy: 0.7726 - recall_102: 0.5333 - val_loss: 0.4954 - val_categorical_accuracy: 0.7917 - val_recall_102: 0.6575\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4667 - categorical_accuracy: 0.7622 - recall_102: 0.6051 - val_loss: 0.4957 - val_categorical_accuracy: 0.8021 - val_recall_102: 0.6575\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4613 - categorical_accuracy: 0.7760 - recall_102: 0.5744 - val_loss: 0.4939 - val_categorical_accuracy: 0.7969 - val_recall_102: 0.6438\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4623 - categorical_accuracy: 0.7726 - recall_102: 0.5179 - val_loss: 0.4955 - val_categorical_accuracy: 0.7969 - val_recall_102: 0.6164\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 20ms/step - loss: 0.4579 - categorical_accuracy: 0.7795 - recall_102: 0.5897 - val_loss: 0.4919 - val_categorical_accuracy: 0.7917 - val_recall_102: 0.6575\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.4577 - categorical_accuracy: 0.7726 - recall_102: 0.5333 - val_loss: 0.4934 - val_categorical_accuracy: 0.8073 - val_recall_102: 0.6438\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4538 - categorical_accuracy: 0.7830 - recall_102: 0.5692 - val_loss: 0.4907 - val_categorical_accuracy: 0.7917 - val_recall_102: 0.6164\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4529 - categorical_accuracy: 0.7795 - recall_102: 0.5538 - val_loss: 0.4903 - val_categorical_accuracy: 0.8021 - val_recall_102: 0.6301\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4516 - categorical_accuracy: 0.7830 - recall_102: 0.5846 - val_loss: 0.4905 - val_categorical_accuracy: 0.8073 - val_recall_102: 0.6438\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.4522 - categorical_accuracy: 0.7760 - recall_102: 0.5436 - val_loss: 0.4884 - val_categorical_accuracy: 0.7917 - val_recall_102: 0.6712\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4495 - categorical_accuracy: 0.7865 - recall_102: 0.6205 - val_loss: 0.4895 - val_categorical_accuracy: 0.7969 - val_recall_102: 0.6027\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 5s 103ms/step - loss: 0.6690 - categorical_accuracy: 0.6615 - recall_103: 0.0000e+00 - val_loss: 0.6733 - val_categorical_accuracy: 0.6198 - val_recall_103: 0.0000e+00\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6570 - categorical_accuracy: 0.6615 - recall_103: 0.0000e+00 - val_loss: 0.6695 - val_categorical_accuracy: 0.6198 - val_recall_103: 0.0000e+00\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6506 - categorical_accuracy: 0.6615 - recall_103: 0.0000e+00 - val_loss: 0.6671 - val_categorical_accuracy: 0.6198 - val_recall_103: 0.0000e+00\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 26ms/step - loss: 0.6463 - categorical_accuracy: 0.6615 - recall_103: 0.0000e+00 - val_loss: 0.6642 - val_categorical_accuracy: 0.6198 - val_recall_103: 0.0000e+00\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6411 - categorical_accuracy: 0.6615 - recall_103: 0.0000e+00 - val_loss: 0.6583 - val_categorical_accuracy: 0.6198 - val_recall_103: 0.0000e+00\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.6357 - categorical_accuracy: 0.6615 - recall_103: 0.0000e+00 - val_loss: 0.6526 - val_categorical_accuracy: 0.6198 - val_recall_103: 0.0000e+00\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.6299 - categorical_accuracy: 0.6615 - recall_103: 0.0000e+00 - val_loss: 0.6469 - val_categorical_accuracy: 0.6198 - val_recall_103: 0.0000e+00\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 24ms/step - loss: 0.6239 - categorical_accuracy: 0.6615 - recall_103: 0.0000e+00 - val_loss: 0.6403 - val_categorical_accuracy: 0.6198 - val_recall_103: 0.0000e+00\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.6168 - categorical_accuracy: 0.6615 - recall_103: 0.0000e+00 - val_loss: 0.6314 - val_categorical_accuracy: 0.6250 - val_recall_103: 0.0137\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.6088 - categorical_accuracy: 0.6632 - recall_103: 0.0051 - val_loss: 0.6238 - val_categorical_accuracy: 0.6250 - val_recall_103: 0.0137\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.6007 - categorical_accuracy: 0.6632 - recall_103: 0.0103 - val_loss: 0.6150 - val_categorical_accuracy: 0.6302 - val_recall_103: 0.0274\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 29ms/step - loss: 0.5920 - categorical_accuracy: 0.6684 - recall_103: 0.0359 - val_loss: 0.6059 - val_categorical_accuracy: 0.6510 - val_recall_103: 0.0822\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.5834 - categorical_accuracy: 0.6736 - recall_103: 0.0564 - val_loss: 0.5960 - val_categorical_accuracy: 0.6615 - val_recall_103: 0.1233\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5739 - categorical_accuracy: 0.6962 - recall_103: 0.1692 - val_loss: 0.5846 - val_categorical_accuracy: 0.7031 - val_recall_103: 0.2603\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5624 - categorical_accuracy: 0.7031 - recall_103: 0.1744 - val_loss: 0.5794 - val_categorical_accuracy: 0.6927 - val_recall_103: 0.2055\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5538 - categorical_accuracy: 0.7101 - recall_103: 0.1949 - val_loss: 0.5694 - val_categorical_accuracy: 0.6927 - val_recall_103: 0.2466\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5444 - categorical_accuracy: 0.7205 - recall_103: 0.2821 - val_loss: 0.5594 - val_categorical_accuracy: 0.7240 - val_recall_103: 0.3562\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 22ms/step - loss: 0.5361 - categorical_accuracy: 0.7205 - recall_103: 0.2718 - val_loss: 0.5556 - val_categorical_accuracy: 0.7240 - val_recall_103: 0.3288\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5295 - categorical_accuracy: 0.7222 - recall_103: 0.3128 - val_loss: 0.5460 - val_categorical_accuracy: 0.7448 - val_recall_103: 0.4384\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.5257 - categorical_accuracy: 0.7309 - recall_103: 0.4256 - val_loss: 0.5389 - val_categorical_accuracy: 0.7500 - val_recall_103: 0.4658\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5192 - categorical_accuracy: 0.7309 - recall_103: 0.2974 - val_loss: 0.5415 - val_categorical_accuracy: 0.7344 - val_recall_103: 0.3562\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 25ms/step - loss: 0.5117 - categorical_accuracy: 0.7517 - recall_103: 0.4462 - val_loss: 0.5282 - val_categorical_accuracy: 0.7760 - val_recall_103: 0.5479\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5054 - categorical_accuracy: 0.7465 - recall_103: 0.4051 - val_loss: 0.5276 - val_categorical_accuracy: 0.7552 - val_recall_103: 0.4658\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.4976 - categorical_accuracy: 0.7569 - recall_103: 0.4462 - val_loss: 0.5207 - val_categorical_accuracy: 0.7760 - val_recall_103: 0.5479\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4914 - categorical_accuracy: 0.7656 - recall_103: 0.4769 - val_loss: 0.5172 - val_categorical_accuracy: 0.7812 - val_recall_103: 0.5616\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4886 - categorical_accuracy: 0.7622 - recall_103: 0.4103 - val_loss: 0.5168 - val_categorical_accuracy: 0.7500 - val_recall_103: 0.4795\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4801 - categorical_accuracy: 0.7760 - recall_103: 0.5128 - val_loss: 0.5085 - val_categorical_accuracy: 0.7917 - val_recall_103: 0.6164\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4774 - categorical_accuracy: 0.7795 - recall_103: 0.5487 - val_loss: 0.5100 - val_categorical_accuracy: 0.7708 - val_recall_103: 0.5342\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.4745 - categorical_accuracy: 0.7778 - recall_103: 0.4821 - val_loss: 0.5039 - val_categorical_accuracy: 0.7969 - val_recall_103: 0.6301\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 23ms/step - loss: 0.4700 - categorical_accuracy: 0.7865 - recall_103: 0.6154 - val_loss: 0.5027 - val_categorical_accuracy: 0.7969 - val_recall_103: 0.6164\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4673 - categorical_accuracy: 0.7726 - recall_103: 0.4718 - val_loss: 0.5021 - val_categorical_accuracy: 0.7865 - val_recall_103: 0.5753\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.4626 - categorical_accuracy: 0.7812 - recall_103: 0.5590 - val_loss: 0.4986 - val_categorical_accuracy: 0.7969 - val_recall_103: 0.6438\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 11ms/step - loss: 0.4590 - categorical_accuracy: 0.7847 - recall_103: 0.5590 - val_loss: 0.5034 - val_categorical_accuracy: 0.7812 - val_recall_103: 0.5479\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4631 - categorical_accuracy: 0.7778 - recall_103: 0.5744 - val_loss: 0.4994 - val_categorical_accuracy: 0.7917 - val_recall_103: 0.6027\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.4559 - categorical_accuracy: 0.7847 - recall_103: 0.5282 - val_loss: 0.4970 - val_categorical_accuracy: 0.7865 - val_recall_103: 0.6027\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4535 - categorical_accuracy: 0.7795 - recall_103: 0.5641 - val_loss: 0.4960 - val_categorical_accuracy: 0.7917 - val_recall_103: 0.6301\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4494 - categorical_accuracy: 0.7882 - recall_103: 0.5897 - val_loss: 0.4976 - val_categorical_accuracy: 0.7917 - val_recall_103: 0.6027\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4485 - categorical_accuracy: 0.7847 - recall_103: 0.5487 - val_loss: 0.4956 - val_categorical_accuracy: 0.7917 - val_recall_103: 0.6164\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4464 - categorical_accuracy: 0.7899 - recall_103: 0.5949 - val_loss: 0.4968 - val_categorical_accuracy: 0.7917 - val_recall_103: 0.6164\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4484 - categorical_accuracy: 0.7830 - recall_103: 0.6000 - val_loss: 0.4959 - val_categorical_accuracy: 0.7969 - val_recall_103: 0.6438\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 24ms/step - loss: 0.4464 - categorical_accuracy: 0.7812 - recall_103: 0.5231 - val_loss: 0.4996 - val_categorical_accuracy: 0.7917 - val_recall_103: 0.6027\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4450 - categorical_accuracy: 0.7830 - recall_103: 0.5949 - val_loss: 0.4986 - val_categorical_accuracy: 0.7969 - val_recall_103: 0.6164\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4402 - categorical_accuracy: 0.7917 - recall_103: 0.5795 - val_loss: 0.4965 - val_categorical_accuracy: 0.7865 - val_recall_103: 0.6438\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4380 - categorical_accuracy: 0.7917 - recall_103: 0.6154 - val_loss: 0.4979 - val_categorical_accuracy: 0.7865 - val_recall_103: 0.6438\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 12ms/step - loss: 0.4386 - categorical_accuracy: 0.7917 - recall_103: 0.5795 - val_loss: 0.4999 - val_categorical_accuracy: 0.8021 - val_recall_103: 0.6301\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4368 - categorical_accuracy: 0.7917 - recall_103: 0.6051 - val_loss: 0.4973 - val_categorical_accuracy: 0.7812 - val_recall_103: 0.6575\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.4360 - categorical_accuracy: 0.7934 - recall_103: 0.6359 - val_loss: 0.5020 - val_categorical_accuracy: 0.7865 - val_recall_103: 0.6164\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4349 - categorical_accuracy: 0.7951 - recall_103: 0.5897 - val_loss: 0.5007 - val_categorical_accuracy: 0.7865 - val_recall_103: 0.6301\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4338 - categorical_accuracy: 0.8003 - recall_103: 0.6000 - val_loss: 0.4986 - val_categorical_accuracy: 0.7865 - val_recall_103: 0.6575\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.4337 - categorical_accuracy: 0.8038 - recall_103: 0.6718 - val_loss: 0.5026 - val_categorical_accuracy: 0.7812 - val_recall_103: 0.6164\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 5s 137ms/step - loss: 0.7224 - categorical_accuracy: 0.3663 - recall_104: 1.0000 - val_loss: 0.6909 - val_categorical_accuracy: 0.5312 - val_recall_104: 0.7671\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.6765 - categorical_accuracy: 0.6493 - recall_104: 0.2769 - val_loss: 0.6700 - val_categorical_accuracy: 0.6198 - val_recall_104: 0.0137\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.6545 - categorical_accuracy: 0.6597 - recall_104: 0.0000e+00 - val_loss: 0.6644 - val_categorical_accuracy: 0.6198 - val_recall_104: 0.0000e+00\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 27ms/step - loss: 0.6411 - categorical_accuracy: 0.6597 - recall_104: 0.0000e+00 - val_loss: 0.6535 - val_categorical_accuracy: 0.6198 - val_recall_104: 0.0000e+00\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.6302 - categorical_accuracy: 0.6615 - recall_104: 0.0051 - val_loss: 0.6436 - val_categorical_accuracy: 0.6250 - val_recall_104: 0.0137\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.6211 - categorical_accuracy: 0.6632 - recall_104: 0.0205 - val_loss: 0.6302 - val_categorical_accuracy: 0.6302 - val_recall_104: 0.0411\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 41ms/step - loss: 0.6118 - categorical_accuracy: 0.6701 - recall_104: 0.0615 - val_loss: 0.6209 - val_categorical_accuracy: 0.6458 - val_recall_104: 0.1096\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.6029 - categorical_accuracy: 0.6823 - recall_104: 0.1179 - val_loss: 0.6102 - val_categorical_accuracy: 0.6562 - val_recall_104: 0.1370\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 13ms/step - loss: 0.5947 - categorical_accuracy: 0.6806 - recall_104: 0.1436 - val_loss: 0.6012 - val_categorical_accuracy: 0.6771 - val_recall_104: 0.2055\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 30ms/step - loss: 0.5911 - categorical_accuracy: 0.6806 - recall_104: 0.1385 - val_loss: 0.5952 - val_categorical_accuracy: 0.6771 - val_recall_104: 0.2055\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.5798 - categorical_accuracy: 0.6910 - recall_104: 0.2205 - val_loss: 0.5823 - val_categorical_accuracy: 0.7135 - val_recall_104: 0.3699\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.5724 - categorical_accuracy: 0.6892 - recall_104: 0.2410 - val_loss: 0.5783 - val_categorical_accuracy: 0.7031 - val_recall_104: 0.3151\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.5676 - categorical_accuracy: 0.6875 - recall_104: 0.2410 - val_loss: 0.5675 - val_categorical_accuracy: 0.7188 - val_recall_104: 0.3836\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.5604 - categorical_accuracy: 0.6927 - recall_104: 0.2564 - val_loss: 0.5620 - val_categorical_accuracy: 0.7188 - val_recall_104: 0.3836\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.5527 - categorical_accuracy: 0.6997 - recall_104: 0.3179 - val_loss: 0.5527 - val_categorical_accuracy: 0.7240 - val_recall_104: 0.4247\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 37ms/step - loss: 0.5502 - categorical_accuracy: 0.7118 - recall_104: 0.4462 - val_loss: 0.5470 - val_categorical_accuracy: 0.7292 - val_recall_104: 0.4384\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.5420 - categorical_accuracy: 0.7014 - recall_104: 0.3026 - val_loss: 0.5466 - val_categorical_accuracy: 0.7344 - val_recall_104: 0.4110\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 20ms/step - loss: 0.5382 - categorical_accuracy: 0.7101 - recall_104: 0.3897 - val_loss: 0.5368 - val_categorical_accuracy: 0.7292 - val_recall_104: 0.4384\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.5325 - categorical_accuracy: 0.7066 - recall_104: 0.3385 - val_loss: 0.5316 - val_categorical_accuracy: 0.7344 - val_recall_104: 0.4521\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.5263 - categorical_accuracy: 0.7309 - recall_104: 0.4513 - val_loss: 0.5271 - val_categorical_accuracy: 0.7448 - val_recall_104: 0.4795\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 32ms/step - loss: 0.5203 - categorical_accuracy: 0.7240 - recall_104: 0.4205 - val_loss: 0.5237 - val_categorical_accuracy: 0.7500 - val_recall_104: 0.4658\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 32ms/step - loss: 0.5158 - categorical_accuracy: 0.7292 - recall_104: 0.4359 - val_loss: 0.5162 - val_categorical_accuracy: 0.7656 - val_recall_104: 0.5205\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.5114 - categorical_accuracy: 0.7274 - recall_104: 0.4667 - val_loss: 0.5137 - val_categorical_accuracy: 0.7708 - val_recall_104: 0.5205\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.5090 - categorical_accuracy: 0.7309 - recall_104: 0.4718 - val_loss: 0.5102 - val_categorical_accuracy: 0.7656 - val_recall_104: 0.5205\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.5024 - categorical_accuracy: 0.7292 - recall_104: 0.4615 - val_loss: 0.5071 - val_categorical_accuracy: 0.7656 - val_recall_104: 0.5205\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.4985 - categorical_accuracy: 0.7448 - recall_104: 0.4923 - val_loss: 0.5023 - val_categorical_accuracy: 0.7604 - val_recall_104: 0.5342\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 21ms/step - loss: 0.4964 - categorical_accuracy: 0.7413 - recall_104: 0.4513 - val_loss: 0.5004 - val_categorical_accuracy: 0.7708 - val_recall_104: 0.5616\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.4911 - categorical_accuracy: 0.7535 - recall_104: 0.5333 - val_loss: 0.4961 - val_categorical_accuracy: 0.7917 - val_recall_104: 0.6164\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 40ms/step - loss: 0.4880 - categorical_accuracy: 0.7500 - recall_104: 0.4872 - val_loss: 0.5003 - val_categorical_accuracy: 0.7708 - val_recall_104: 0.5342\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.4841 - categorical_accuracy: 0.7483 - recall_104: 0.5128 - val_loss: 0.4915 - val_categorical_accuracy: 0.7969 - val_recall_104: 0.6164\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.4801 - categorical_accuracy: 0.7604 - recall_104: 0.5333 - val_loss: 0.4925 - val_categorical_accuracy: 0.7969 - val_recall_104: 0.6027\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.4772 - categorical_accuracy: 0.7483 - recall_104: 0.4821 - val_loss: 0.4888 - val_categorical_accuracy: 0.8021 - val_recall_104: 0.6164\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.4789 - categorical_accuracy: 0.7552 - recall_104: 0.6000 - val_loss: 0.4890 - val_categorical_accuracy: 0.7969 - val_recall_104: 0.6027\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.4717 - categorical_accuracy: 0.7535 - recall_104: 0.4923 - val_loss: 0.4870 - val_categorical_accuracy: 0.7917 - val_recall_104: 0.6027\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.4737 - categorical_accuracy: 0.7552 - recall_104: 0.5744 - val_loss: 0.4854 - val_categorical_accuracy: 0.8021 - val_recall_104: 0.6301\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 35ms/step - loss: 0.4769 - categorical_accuracy: 0.7517 - recall_104: 0.4462 - val_loss: 0.4821 - val_categorical_accuracy: 0.8021 - val_recall_104: 0.6575\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 20ms/step - loss: 0.4672 - categorical_accuracy: 0.7691 - recall_104: 0.6308 - val_loss: 0.4828 - val_categorical_accuracy: 0.8021 - val_recall_104: 0.6575\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.4645 - categorical_accuracy: 0.7639 - recall_104: 0.5077 - val_loss: 0.4822 - val_categorical_accuracy: 0.7969 - val_recall_104: 0.6438\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.4600 - categorical_accuracy: 0.7656 - recall_104: 0.5846 - val_loss: 0.4823 - val_categorical_accuracy: 0.8073 - val_recall_104: 0.6438\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 21ms/step - loss: 0.4580 - categorical_accuracy: 0.7622 - recall_104: 0.5641 - val_loss: 0.4818 - val_categorical_accuracy: 0.8073 - val_recall_104: 0.6438\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 21ms/step - loss: 0.4574 - categorical_accuracy: 0.7604 - recall_104: 0.5487 - val_loss: 0.4813 - val_categorical_accuracy: 0.8073 - val_recall_104: 0.6438\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 1s 71ms/step - loss: 0.4538 - categorical_accuracy: 0.7691 - recall_104: 0.5641 - val_loss: 0.4780 - val_categorical_accuracy: 0.8073 - val_recall_104: 0.6575\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.4523 - categorical_accuracy: 0.7778 - recall_104: 0.5641 - val_loss: 0.4826 - val_categorical_accuracy: 0.8073 - val_recall_104: 0.6438\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.4531 - categorical_accuracy: 0.7691 - recall_104: 0.5538 - val_loss: 0.4783 - val_categorical_accuracy: 0.8073 - val_recall_104: 0.6575\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.4525 - categorical_accuracy: 0.7812 - recall_104: 0.5538 - val_loss: 0.4827 - val_categorical_accuracy: 0.8021 - val_recall_104: 0.6301\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.4516 - categorical_accuracy: 0.7795 - recall_104: 0.6256 - val_loss: 0.4823 - val_categorical_accuracy: 0.8073 - val_recall_104: 0.6438\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.4475 - categorical_accuracy: 0.7795 - recall_104: 0.5487 - val_loss: 0.4802 - val_categorical_accuracy: 0.8125 - val_recall_104: 0.6575\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.4454 - categorical_accuracy: 0.7847 - recall_104: 0.6103 - val_loss: 0.4800 - val_categorical_accuracy: 0.8177 - val_recall_104: 0.6712\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 31ms/step - loss: 0.4472 - categorical_accuracy: 0.7778 - recall_104: 0.5641 - val_loss: 0.4781 - val_categorical_accuracy: 0.8125 - val_recall_104: 0.6849\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.4429 - categorical_accuracy: 0.7830 - recall_104: 0.6051 - val_loss: 0.4823 - val_categorical_accuracy: 0.8125 - val_recall_104: 0.6575\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 6s 126ms/step - loss: 0.6663 - categorical_accuracy: 0.6615 - recall_105: 0.0000e+00 - val_loss: 0.6677 - val_categorical_accuracy: 0.6198 - val_recall_105: 0.0000e+00\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.6493 - categorical_accuracy: 0.6615 - recall_105: 0.0000e+00 - val_loss: 0.6663 - val_categorical_accuracy: 0.6198 - val_recall_105: 0.0000e+00\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 28ms/step - loss: 0.6431 - categorical_accuracy: 0.6615 - recall_105: 0.0000e+00 - val_loss: 0.6642 - val_categorical_accuracy: 0.6198 - val_recall_105: 0.0000e+00\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 20ms/step - loss: 0.6368 - categorical_accuracy: 0.6615 - recall_105: 0.0000e+00 - val_loss: 0.6537 - val_categorical_accuracy: 0.6198 - val_recall_105: 0.0000e+00\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.6287 - categorical_accuracy: 0.6615 - recall_105: 0.0000e+00 - val_loss: 0.6440 - val_categorical_accuracy: 0.6198 - val_recall_105: 0.0000e+00\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 32ms/step - loss: 0.6213 - categorical_accuracy: 0.6615 - recall_105: 0.0000e+00 - val_loss: 0.6335 - val_categorical_accuracy: 0.6198 - val_recall_105: 0.0000e+00\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.6118 - categorical_accuracy: 0.6615 - recall_105: 0.0000e+00 - val_loss: 0.6229 - val_categorical_accuracy: 0.6250 - val_recall_105: 0.0137\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 14ms/step - loss: 0.6024 - categorical_accuracy: 0.6649 - recall_105: 0.0154 - val_loss: 0.6143 - val_categorical_accuracy: 0.6250 - val_recall_105: 0.0137\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 27ms/step - loss: 0.5936 - categorical_accuracy: 0.6684 - recall_105: 0.0513 - val_loss: 0.6028 - val_categorical_accuracy: 0.6562 - val_recall_105: 0.1096\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 20ms/step - loss: 0.5832 - categorical_accuracy: 0.6892 - recall_105: 0.1179 - val_loss: 0.5936 - val_categorical_accuracy: 0.6667 - val_recall_105: 0.1507\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.5748 - categorical_accuracy: 0.6962 - recall_105: 0.1487 - val_loss: 0.5830 - val_categorical_accuracy: 0.6823 - val_recall_105: 0.2466\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.5645 - categorical_accuracy: 0.7135 - recall_105: 0.2308 - val_loss: 0.5757 - val_categorical_accuracy: 0.6875 - val_recall_105: 0.2466\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 35ms/step - loss: 0.5549 - categorical_accuracy: 0.7118 - recall_105: 0.2564 - val_loss: 0.5640 - val_categorical_accuracy: 0.7344 - val_recall_105: 0.3836\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.5454 - categorical_accuracy: 0.7170 - recall_105: 0.3231 - val_loss: 0.5531 - val_categorical_accuracy: 0.7396 - val_recall_105: 0.4521\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.5388 - categorical_accuracy: 0.7170 - recall_105: 0.3641 - val_loss: 0.5461 - val_categorical_accuracy: 0.7292 - val_recall_105: 0.4384\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 21ms/step - loss: 0.5309 - categorical_accuracy: 0.7153 - recall_105: 0.3795 - val_loss: 0.5380 - val_categorical_accuracy: 0.7500 - val_recall_105: 0.5068\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 25ms/step - loss: 0.5313 - categorical_accuracy: 0.7274 - recall_105: 0.3590 - val_loss: 0.5341 - val_categorical_accuracy: 0.7500 - val_recall_105: 0.5068\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 24ms/step - loss: 0.5246 - categorical_accuracy: 0.7396 - recall_105: 0.5692 - val_loss: 0.5278 - val_categorical_accuracy: 0.7708 - val_recall_105: 0.5616\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 32ms/step - loss: 0.5107 - categorical_accuracy: 0.7396 - recall_105: 0.4308 - val_loss: 0.5299 - val_categorical_accuracy: 0.7396 - val_recall_105: 0.4658\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.5063 - categorical_accuracy: 0.7326 - recall_105: 0.4256 - val_loss: 0.5203 - val_categorical_accuracy: 0.7708 - val_recall_105: 0.5753\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.5004 - categorical_accuracy: 0.7431 - recall_105: 0.4769 - val_loss: 0.5174 - val_categorical_accuracy: 0.7656 - val_recall_105: 0.5616\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.4980 - categorical_accuracy: 0.7569 - recall_105: 0.5128 - val_loss: 0.5158 - val_categorical_accuracy: 0.7604 - val_recall_105: 0.5342\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.4937 - categorical_accuracy: 0.7552 - recall_105: 0.4256 - val_loss: 0.5094 - val_categorical_accuracy: 0.7812 - val_recall_105: 0.6301\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 33ms/step - loss: 0.4885 - categorical_accuracy: 0.7569 - recall_105: 0.5795 - val_loss: 0.5083 - val_categorical_accuracy: 0.7812 - val_recall_105: 0.6027\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 20ms/step - loss: 0.4835 - categorical_accuracy: 0.7656 - recall_105: 0.4769 - val_loss: 0.5088 - val_categorical_accuracy: 0.7812 - val_recall_105: 0.5616\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.4823 - categorical_accuracy: 0.7656 - recall_105: 0.5487 - val_loss: 0.5042 - val_categorical_accuracy: 0.7969 - val_recall_105: 0.6301\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.4806 - categorical_accuracy: 0.7726 - recall_105: 0.4564 - val_loss: 0.5012 - val_categorical_accuracy: 0.8073 - val_recall_105: 0.6575\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4744 - categorical_accuracy: 0.7622 - recall_105: 0.5949 - val_loss: 0.4996 - val_categorical_accuracy: 0.8073 - val_recall_105: 0.6438\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.4716 - categorical_accuracy: 0.7778 - recall_105: 0.5077 - val_loss: 0.4986 - val_categorical_accuracy: 0.7969 - val_recall_105: 0.6164\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 38ms/step - loss: 0.4686 - categorical_accuracy: 0.7795 - recall_105: 0.6051 - val_loss: 0.4971 - val_categorical_accuracy: 0.7969 - val_recall_105: 0.6164\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.4681 - categorical_accuracy: 0.7812 - recall_105: 0.4821 - val_loss: 0.4976 - val_categorical_accuracy: 0.7917 - val_recall_105: 0.6027\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 34ms/step - loss: 0.4649 - categorical_accuracy: 0.7778 - recall_105: 0.6308 - val_loss: 0.4949 - val_categorical_accuracy: 0.7917 - val_recall_105: 0.6027\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.4583 - categorical_accuracy: 0.7743 - recall_105: 0.5026 - val_loss: 0.4971 - val_categorical_accuracy: 0.7812 - val_recall_105: 0.5753\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.4541 - categorical_accuracy: 0.7795 - recall_105: 0.5487 - val_loss: 0.4915 - val_categorical_accuracy: 0.7865 - val_recall_105: 0.6301\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.4539 - categorical_accuracy: 0.7795 - recall_105: 0.5538 - val_loss: 0.4912 - val_categorical_accuracy: 0.7917 - val_recall_105: 0.6164\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 32ms/step - loss: 0.4507 - categorical_accuracy: 0.7760 - recall_105: 0.6051 - val_loss: 0.4927 - val_categorical_accuracy: 0.7917 - val_recall_105: 0.6027\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 21ms/step - loss: 0.4495 - categorical_accuracy: 0.7812 - recall_105: 0.5436 - val_loss: 0.4916 - val_categorical_accuracy: 0.7865 - val_recall_105: 0.6027\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.4481 - categorical_accuracy: 0.7865 - recall_105: 0.5590 - val_loss: 0.4916 - val_categorical_accuracy: 0.7812 - val_recall_105: 0.6027\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.4449 - categorical_accuracy: 0.7882 - recall_105: 0.6051 - val_loss: 0.4917 - val_categorical_accuracy: 0.7812 - val_recall_105: 0.6164\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.4430 - categorical_accuracy: 0.7847 - recall_105: 0.5641 - val_loss: 0.4928 - val_categorical_accuracy: 0.7917 - val_recall_105: 0.6027\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4436 - categorical_accuracy: 0.7847 - recall_105: 0.6051 - val_loss: 0.4940 - val_categorical_accuracy: 0.7969 - val_recall_105: 0.6027\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 35ms/step - loss: 0.4414 - categorical_accuracy: 0.7882 - recall_105: 0.5590 - val_loss: 0.4921 - val_categorical_accuracy: 0.7865 - val_recall_105: 0.6301\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.4404 - categorical_accuracy: 0.7865 - recall_105: 0.6205 - val_loss: 0.4964 - val_categorical_accuracy: 0.7969 - val_recall_105: 0.6027\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4383 - categorical_accuracy: 0.7830 - recall_105: 0.5487 - val_loss: 0.4937 - val_categorical_accuracy: 0.7760 - val_recall_105: 0.6164\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.4416 - categorical_accuracy: 0.7795 - recall_105: 0.6154 - val_loss: 0.4993 - val_categorical_accuracy: 0.7917 - val_recall_105: 0.6027\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 33ms/step - loss: 0.4368 - categorical_accuracy: 0.7830 - recall_105: 0.6000 - val_loss: 0.4954 - val_categorical_accuracy: 0.7917 - val_recall_105: 0.6301\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4342 - categorical_accuracy: 0.7951 - recall_105: 0.6205 - val_loss: 0.4999 - val_categorical_accuracy: 0.7969 - val_recall_105: 0.6027\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.4346 - categorical_accuracy: 0.7847 - recall_105: 0.5795 - val_loss: 0.4973 - val_categorical_accuracy: 0.7812 - val_recall_105: 0.6164\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.4321 - categorical_accuracy: 0.7934 - recall_105: 0.6000 - val_loss: 0.4987 - val_categorical_accuracy: 0.7812 - val_recall_105: 0.6164\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 21ms/step - loss: 0.4330 - categorical_accuracy: 0.7934 - recall_105: 0.6359 - val_loss: 0.5015 - val_categorical_accuracy: 0.7865 - val_recall_105: 0.6027\n",
      "Epoch 1/50\n",
      "12/12 [==============================] - 7s 140ms/step - loss: 0.6680 - categorical_accuracy: 0.6597 - recall_106: 0.0000e+00 - val_loss: 0.6759 - val_categorical_accuracy: 0.6198 - val_recall_106: 0.0000e+00\n",
      "Epoch 2/50\n",
      "12/12 [==============================] - 0s 34ms/step - loss: 0.6556 - categorical_accuracy: 0.6615 - recall_106: 0.0000e+00 - val_loss: 0.6710 - val_categorical_accuracy: 0.6198 - val_recall_106: 0.0000e+00\n",
      "Epoch 3/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.6489 - categorical_accuracy: 0.6615 - recall_106: 0.0000e+00 - val_loss: 0.6646 - val_categorical_accuracy: 0.6198 - val_recall_106: 0.0000e+00\n",
      "Epoch 4/50\n",
      "12/12 [==============================] - 0s 38ms/step - loss: 0.6427 - categorical_accuracy: 0.6615 - recall_106: 0.0000e+00 - val_loss: 0.6567 - val_categorical_accuracy: 0.6198 - val_recall_106: 0.0000e+00\n",
      "Epoch 5/50\n",
      "12/12 [==============================] - 0s 26ms/step - loss: 0.6356 - categorical_accuracy: 0.6615 - recall_106: 0.0000e+00 - val_loss: 0.6440 - val_categorical_accuracy: 0.6250 - val_recall_106: 0.0137\n",
      "Epoch 6/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.6290 - categorical_accuracy: 0.6719 - recall_106: 0.0410 - val_loss: 0.6342 - val_categorical_accuracy: 0.6406 - val_recall_106: 0.0548\n",
      "Epoch 7/50\n",
      "12/12 [==============================] - 0s 31ms/step - loss: 0.6228 - categorical_accuracy: 0.6701 - recall_106: 0.0308 - val_loss: 0.6280 - val_categorical_accuracy: 0.6406 - val_recall_106: 0.0548\n",
      "Epoch 8/50\n",
      "12/12 [==============================] - 0s 36ms/step - loss: 0.6137 - categorical_accuracy: 0.6701 - recall_106: 0.0564 - val_loss: 0.6150 - val_categorical_accuracy: 0.6615 - val_recall_106: 0.1233\n",
      "Epoch 9/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.6055 - categorical_accuracy: 0.6701 - recall_106: 0.0718 - val_loss: 0.6062 - val_categorical_accuracy: 0.6615 - val_recall_106: 0.1233\n",
      "Epoch 10/50\n",
      "12/12 [==============================] - 0s 24ms/step - loss: 0.5978 - categorical_accuracy: 0.6753 - recall_106: 0.0974 - val_loss: 0.5960 - val_categorical_accuracy: 0.6771 - val_recall_106: 0.1918\n",
      "Epoch 11/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.5894 - categorical_accuracy: 0.6771 - recall_106: 0.1282 - val_loss: 0.5867 - val_categorical_accuracy: 0.6979 - val_recall_106: 0.2466\n",
      "Epoch 12/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.5809 - categorical_accuracy: 0.7101 - recall_106: 0.2513 - val_loss: 0.5762 - val_categorical_accuracy: 0.7188 - val_recall_106: 0.3151\n",
      "Epoch 13/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.5736 - categorical_accuracy: 0.7066 - recall_106: 0.2256 - val_loss: 0.5724 - val_categorical_accuracy: 0.7031 - val_recall_106: 0.2603\n",
      "Epoch 14/50\n",
      "12/12 [==============================] - 0s 33ms/step - loss: 0.5635 - categorical_accuracy: 0.7170 - recall_106: 0.2923 - val_loss: 0.5601 - val_categorical_accuracy: 0.7500 - val_recall_106: 0.4384\n",
      "Epoch 15/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.5569 - categorical_accuracy: 0.7257 - recall_106: 0.3590 - val_loss: 0.5550 - val_categorical_accuracy: 0.7552 - val_recall_106: 0.4384\n",
      "Epoch 16/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.5502 - categorical_accuracy: 0.7222 - recall_106: 0.3026 - val_loss: 0.5492 - val_categorical_accuracy: 0.7500 - val_recall_106: 0.4247\n",
      "Epoch 17/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.5466 - categorical_accuracy: 0.7361 - recall_106: 0.4256 - val_loss: 0.5398 - val_categorical_accuracy: 0.7500 - val_recall_106: 0.4795\n",
      "Epoch 18/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.5365 - categorical_accuracy: 0.7292 - recall_106: 0.3385 - val_loss: 0.5386 - val_categorical_accuracy: 0.7448 - val_recall_106: 0.4521\n",
      "Epoch 19/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.5308 - categorical_accuracy: 0.7309 - recall_106: 0.3692 - val_loss: 0.5330 - val_categorical_accuracy: 0.7500 - val_recall_106: 0.4795\n",
      "Epoch 20/50\n",
      "12/12 [==============================] - 0s 30ms/step - loss: 0.5250 - categorical_accuracy: 0.7326 - recall_106: 0.3846 - val_loss: 0.5269 - val_categorical_accuracy: 0.7552 - val_recall_106: 0.5068\n",
      "Epoch 21/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.5218 - categorical_accuracy: 0.7448 - recall_106: 0.4974 - val_loss: 0.5228 - val_categorical_accuracy: 0.7552 - val_recall_106: 0.5205\n",
      "Epoch 22/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.5164 - categorical_accuracy: 0.7431 - recall_106: 0.4154 - val_loss: 0.5235 - val_categorical_accuracy: 0.7604 - val_recall_106: 0.4932\n",
      "Epoch 23/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.5129 - categorical_accuracy: 0.7413 - recall_106: 0.4205 - val_loss: 0.5169 - val_categorical_accuracy: 0.7656 - val_recall_106: 0.5342\n",
      "Epoch 24/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.5087 - categorical_accuracy: 0.7448 - recall_106: 0.4513 - val_loss: 0.5152 - val_categorical_accuracy: 0.7656 - val_recall_106: 0.5342\n",
      "Epoch 25/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.5041 - categorical_accuracy: 0.7465 - recall_106: 0.4667 - val_loss: 0.5144 - val_categorical_accuracy: 0.7604 - val_recall_106: 0.5205\n",
      "Epoch 26/50\n",
      "12/12 [==============================] - 0s 27ms/step - loss: 0.5013 - categorical_accuracy: 0.7448 - recall_106: 0.4154 - val_loss: 0.5133 - val_categorical_accuracy: 0.7552 - val_recall_106: 0.5068\n",
      "Epoch 27/50\n",
      "12/12 [==============================] - 0s 31ms/step - loss: 0.5007 - categorical_accuracy: 0.7535 - recall_106: 0.5385 - val_loss: 0.5049 - val_categorical_accuracy: 0.7865 - val_recall_106: 0.6301\n",
      "Epoch 28/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.4923 - categorical_accuracy: 0.7552 - recall_106: 0.4410 - val_loss: 0.5113 - val_categorical_accuracy: 0.7500 - val_recall_106: 0.4658\n",
      "Epoch 29/50\n",
      "12/12 [==============================] - 0s 16ms/step - loss: 0.4907 - categorical_accuracy: 0.7604 - recall_106: 0.5077 - val_loss: 0.5015 - val_categorical_accuracy: 0.7865 - val_recall_106: 0.6301\n",
      "Epoch 30/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4939 - categorical_accuracy: 0.7361 - recall_106: 0.4103 - val_loss: 0.5039 - val_categorical_accuracy: 0.7500 - val_recall_106: 0.5068\n",
      "Epoch 31/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.4860 - categorical_accuracy: 0.7639 - recall_106: 0.5538 - val_loss: 0.4988 - val_categorical_accuracy: 0.7812 - val_recall_106: 0.5890\n",
      "Epoch 32/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.4808 - categorical_accuracy: 0.7587 - recall_106: 0.5077 - val_loss: 0.5015 - val_categorical_accuracy: 0.7500 - val_recall_106: 0.5068\n",
      "Epoch 33/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4789 - categorical_accuracy: 0.7656 - recall_106: 0.4872 - val_loss: 0.4965 - val_categorical_accuracy: 0.7760 - val_recall_106: 0.5753\n",
      "Epoch 34/50\n",
      "12/12 [==============================] - 0s 30ms/step - loss: 0.4752 - categorical_accuracy: 0.7604 - recall_106: 0.5026 - val_loss: 0.4958 - val_categorical_accuracy: 0.7708 - val_recall_106: 0.5616\n",
      "Epoch 35/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.4761 - categorical_accuracy: 0.7639 - recall_106: 0.5487 - val_loss: 0.4980 - val_categorical_accuracy: 0.7656 - val_recall_106: 0.5205\n",
      "Epoch 36/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.4720 - categorical_accuracy: 0.7656 - recall_106: 0.4974 - val_loss: 0.4929 - val_categorical_accuracy: 0.7865 - val_recall_106: 0.6164\n",
      "Epoch 37/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.4691 - categorical_accuracy: 0.7604 - recall_106: 0.5077 - val_loss: 0.4930 - val_categorical_accuracy: 0.7865 - val_recall_106: 0.6164\n",
      "Epoch 38/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.4677 - categorical_accuracy: 0.7587 - recall_106: 0.5179 - val_loss: 0.4934 - val_categorical_accuracy: 0.7812 - val_recall_106: 0.6027\n",
      "Epoch 39/50\n",
      "12/12 [==============================] - 0s 21ms/step - loss: 0.4662 - categorical_accuracy: 0.7639 - recall_106: 0.5179 - val_loss: 0.4934 - val_categorical_accuracy: 0.7969 - val_recall_106: 0.6164\n",
      "Epoch 40/50\n",
      "12/12 [==============================] - 1s 45ms/step - loss: 0.4656 - categorical_accuracy: 0.7743 - recall_106: 0.6000 - val_loss: 0.4928 - val_categorical_accuracy: 0.7969 - val_recall_106: 0.6164\n",
      "Epoch 41/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.4609 - categorical_accuracy: 0.7622 - recall_106: 0.5179 - val_loss: 0.4949 - val_categorical_accuracy: 0.7760 - val_recall_106: 0.5616\n",
      "Epoch 42/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.4591 - categorical_accuracy: 0.7691 - recall_106: 0.5282 - val_loss: 0.4904 - val_categorical_accuracy: 0.7969 - val_recall_106: 0.6438\n",
      "Epoch 43/50\n",
      "12/12 [==============================] - 0s 15ms/step - loss: 0.4589 - categorical_accuracy: 0.7656 - recall_106: 0.5487 - val_loss: 0.4912 - val_categorical_accuracy: 0.8021 - val_recall_106: 0.6301\n",
      "Epoch 44/50\n",
      "12/12 [==============================] - 0s 21ms/step - loss: 0.4570 - categorical_accuracy: 0.7691 - recall_106: 0.5692 - val_loss: 0.4916 - val_categorical_accuracy: 0.8021 - val_recall_106: 0.6301\n",
      "Epoch 45/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.4545 - categorical_accuracy: 0.7708 - recall_106: 0.5333 - val_loss: 0.4947 - val_categorical_accuracy: 0.7865 - val_recall_106: 0.5890\n",
      "Epoch 46/50\n",
      "12/12 [==============================] - 0s 33ms/step - loss: 0.4557 - categorical_accuracy: 0.7674 - recall_106: 0.5641 - val_loss: 0.4942 - val_categorical_accuracy: 0.7917 - val_recall_106: 0.6027\n",
      "Epoch 47/50\n",
      "12/12 [==============================] - 0s 17ms/step - loss: 0.4524 - categorical_accuracy: 0.7726 - recall_106: 0.5436 - val_loss: 0.4933 - val_categorical_accuracy: 0.7969 - val_recall_106: 0.6164\n",
      "Epoch 48/50\n",
      "12/12 [==============================] - 0s 20ms/step - loss: 0.4537 - categorical_accuracy: 0.7708 - recall_106: 0.6051 - val_loss: 0.4941 - val_categorical_accuracy: 0.7969 - val_recall_106: 0.6164\n",
      "Epoch 49/50\n",
      "12/12 [==============================] - 0s 19ms/step - loss: 0.4509 - categorical_accuracy: 0.7708 - recall_106: 0.5282 - val_loss: 0.4963 - val_categorical_accuracy: 0.7865 - val_recall_106: 0.5890\n",
      "Epoch 50/50\n",
      "12/12 [==============================] - 0s 18ms/step - loss: 0.4501 - categorical_accuracy: 0.7726 - recall_106: 0.5949 - val_loss: 0.4940 - val_categorical_accuracy: 0.7969 - val_recall_106: 0.6301\n"
     ]
    }
   ],
   "source": [
    "batch_size = 50\n",
    "epochs = 50\n",
    "\n",
    "for layers_num in range(3, 10):\n",
    "    for architecture in range(5):\n",
    "        norons_in_layers = np.random.randint(5, 50, size=layers_num)\n",
    "        model, bst_acc_train, bst_acc_test, bst_rec_train, bst_rec_test = NurNet(norons_in_layers, epochs, batch_size)\n",
    "        results.loc[architecture_num]= [layers_num, str(norons_in_layers), model, bst_rec_train, bst_rec_test, bst_rec_train, bst_rec_test]\n",
    "        architecture_num+=1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Layers_num</th>\n",
       "      <th>norons_in_layer</th>\n",
       "      <th>model</th>\n",
       "      <th>accuracy_train</th>\n",
       "      <th>accuracy_test</th>\n",
       "      <th>recall_train</th>\n",
       "      <th>recall_test</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>3</td>\n",
       "      <td>[37 43 39]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.807292</td>\n",
       "      <td>0.698630</td>\n",
       "      <td>0.807292</td>\n",
       "      <td>0.698630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>3</td>\n",
       "      <td>[36 46 19]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.781250</td>\n",
       "      <td>0.643836</td>\n",
       "      <td>0.781250</td>\n",
       "      <td>0.643836</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>[36 42 41]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.807292</td>\n",
       "      <td>0.657534</td>\n",
       "      <td>0.807292</td>\n",
       "      <td>0.657534</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>[48 49 13]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.781250</td>\n",
       "      <td>0.643836</td>\n",
       "      <td>0.781250</td>\n",
       "      <td>0.643836</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3</td>\n",
       "      <td>[42 39 13]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.802083</td>\n",
       "      <td>0.712329</td>\n",
       "      <td>0.802083</td>\n",
       "      <td>0.712329</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>4</td>\n",
       "      <td>[40 17 24 30]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.791667</td>\n",
       "      <td>0.643836</td>\n",
       "      <td>0.791667</td>\n",
       "      <td>0.643836</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>4</td>\n",
       "      <td>[37 15 44 40]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.765625</td>\n",
       "      <td>0.630137</td>\n",
       "      <td>0.765625</td>\n",
       "      <td>0.630137</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>4</td>\n",
       "      <td>[28 35 14 26]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.812500</td>\n",
       "      <td>0.684932</td>\n",
       "      <td>0.812500</td>\n",
       "      <td>0.684932</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>4</td>\n",
       "      <td>[12 18 15  8]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.796875</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.796875</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>4</td>\n",
       "      <td>[44 40 19 18]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.802083</td>\n",
       "      <td>0.698630</td>\n",
       "      <td>0.802083</td>\n",
       "      <td>0.698630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>5</td>\n",
       "      <td>[25 25 24 48 49]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.802083</td>\n",
       "      <td>0.657534</td>\n",
       "      <td>0.802083</td>\n",
       "      <td>0.657534</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>5</td>\n",
       "      <td>[10 43 38 32 34]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.807292</td>\n",
       "      <td>0.986301</td>\n",
       "      <td>0.807292</td>\n",
       "      <td>0.986301</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>5</td>\n",
       "      <td>[42 23 11 47  5]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.802083</td>\n",
       "      <td>0.657534</td>\n",
       "      <td>0.802083</td>\n",
       "      <td>0.657534</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>5</td>\n",
       "      <td>[24 49  6 41 44]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.791667</td>\n",
       "      <td>0.671233</td>\n",
       "      <td>0.791667</td>\n",
       "      <td>0.671233</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>5</td>\n",
       "      <td>[23 17 26 26 37]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.802083</td>\n",
       "      <td>0.684932</td>\n",
       "      <td>0.802083</td>\n",
       "      <td>0.684932</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>6</td>\n",
       "      <td>[13 43 18 41 48 35]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.812500</td>\n",
       "      <td>0.684932</td>\n",
       "      <td>0.812500</td>\n",
       "      <td>0.684932</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>6</td>\n",
       "      <td>[19 26 25 24 27  5]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.791667</td>\n",
       "      <td>0.630137</td>\n",
       "      <td>0.791667</td>\n",
       "      <td>0.630137</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>6</td>\n",
       "      <td>[24 12 12 41 46 24]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.776042</td>\n",
       "      <td>0.602740</td>\n",
       "      <td>0.776042</td>\n",
       "      <td>0.602740</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>6</td>\n",
       "      <td>[49 15 42  7 16  7]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.812500</td>\n",
       "      <td>0.698630</td>\n",
       "      <td>0.812500</td>\n",
       "      <td>0.698630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>6</td>\n",
       "      <td>[ 8 20 21 23 12 22]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.729167</td>\n",
       "      <td>0.465753</td>\n",
       "      <td>0.729167</td>\n",
       "      <td>0.465753</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>7</td>\n",
       "      <td>[17 33 14 34 23 44 42]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.776042</td>\n",
       "      <td>0.630137</td>\n",
       "      <td>0.776042</td>\n",
       "      <td>0.630137</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>7</td>\n",
       "      <td>[28  8 37 29  7  8 26]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.786458</td>\n",
       "      <td>0.630137</td>\n",
       "      <td>0.786458</td>\n",
       "      <td>0.630137</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>7</td>\n",
       "      <td>[47 38 35 19 15  5 34]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.812500</td>\n",
       "      <td>0.684932</td>\n",
       "      <td>0.812500</td>\n",
       "      <td>0.684932</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>7</td>\n",
       "      <td>[16 47 14 44 31 23 27]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.796875</td>\n",
       "      <td>0.671233</td>\n",
       "      <td>0.796875</td>\n",
       "      <td>0.671233</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>7</td>\n",
       "      <td>[47 21 12 40 23 39 17]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.786458</td>\n",
       "      <td>0.643836</td>\n",
       "      <td>0.786458</td>\n",
       "      <td>0.643836</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>8</td>\n",
       "      <td>[22 14  5 36 36 19 39 37]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.807292</td>\n",
       "      <td>0.671233</td>\n",
       "      <td>0.807292</td>\n",
       "      <td>0.671233</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>8</td>\n",
       "      <td>[45 46  9 27 31  8 30 40]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.807292</td>\n",
       "      <td>0.698630</td>\n",
       "      <td>0.807292</td>\n",
       "      <td>0.698630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>8</td>\n",
       "      <td>[ 9 36 30  8 18 40 17 36]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.791667</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.791667</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>8</td>\n",
       "      <td>[24 29 42 45 13 36 26 12]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.812500</td>\n",
       "      <td>0.698630</td>\n",
       "      <td>0.812500</td>\n",
       "      <td>0.698630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>8</td>\n",
       "      <td>[31 11 40 34  7 21 39 39]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.786458</td>\n",
       "      <td>0.575342</td>\n",
       "      <td>0.786458</td>\n",
       "      <td>0.575342</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>9</td>\n",
       "      <td>[48 10 32 40 34 47 11 35 20]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.807292</td>\n",
       "      <td>0.671233</td>\n",
       "      <td>0.807292</td>\n",
       "      <td>0.671233</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>9</td>\n",
       "      <td>[47 11  7 33 23 35  7 32 47]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.802083</td>\n",
       "      <td>0.657534</td>\n",
       "      <td>0.802083</td>\n",
       "      <td>0.657534</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>9</td>\n",
       "      <td>[37 30 46  6  5 10 13 11 26]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.817708</td>\n",
       "      <td>0.767123</td>\n",
       "      <td>0.817708</td>\n",
       "      <td>0.767123</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>9</td>\n",
       "      <td>[47 21 35 36 40 49 36 48 28]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.807292</td>\n",
       "      <td>0.657534</td>\n",
       "      <td>0.807292</td>\n",
       "      <td>0.657534</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>9</td>\n",
       "      <td>[20 33 42 37 31 14  8 25 36]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.802083</td>\n",
       "      <td>0.643836</td>\n",
       "      <td>0.802083</td>\n",
       "      <td>0.643836</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    Layers_num               norons_in_layer  \\\n",
       "0            3                    [37 43 39]   \n",
       "1            3                    [36 46 19]   \n",
       "2            3                    [36 42 41]   \n",
       "3            3                    [48 49 13]   \n",
       "4            3                    [42 39 13]   \n",
       "5            4                 [40 17 24 30]   \n",
       "6            4                 [37 15 44 40]   \n",
       "7            4                 [28 35 14 26]   \n",
       "8            4                 [12 18 15  8]   \n",
       "9            4                 [44 40 19 18]   \n",
       "10           5              [25 25 24 48 49]   \n",
       "11           5              [10 43 38 32 34]   \n",
       "12           5              [42 23 11 47  5]   \n",
       "13           5              [24 49  6 41 44]   \n",
       "14           5              [23 17 26 26 37]   \n",
       "15           6           [13 43 18 41 48 35]   \n",
       "16           6           [19 26 25 24 27  5]   \n",
       "17           6           [24 12 12 41 46 24]   \n",
       "18           6           [49 15 42  7 16  7]   \n",
       "19           6           [ 8 20 21 23 12 22]   \n",
       "20           7        [17 33 14 34 23 44 42]   \n",
       "21           7        [28  8 37 29  7  8 26]   \n",
       "22           7        [47 38 35 19 15  5 34]   \n",
       "23           7        [16 47 14 44 31 23 27]   \n",
       "24           7        [47 21 12 40 23 39 17]   \n",
       "25           8     [22 14  5 36 36 19 39 37]   \n",
       "26           8     [45 46  9 27 31  8 30 40]   \n",
       "27           8     [ 9 36 30  8 18 40 17 36]   \n",
       "28           8     [24 29 42 45 13 36 26 12]   \n",
       "29           8     [31 11 40 34  7 21 39 39]   \n",
       "30           9  [48 10 32 40 34 47 11 35 20]   \n",
       "31           9  [47 11  7 33 23 35  7 32 47]   \n",
       "32           9  [37 30 46  6  5 10 13 11 26]   \n",
       "33           9  [47 21 35 36 40 49 36 48 28]   \n",
       "34           9  [20 33 42 37 31 14  8 25 36]   \n",
       "\n",
       "                                                model  accuracy_train  \\\n",
       "0   <keras.engine.sequential.Sequential object at ...        0.807292   \n",
       "1   <keras.engine.sequential.Sequential object at ...        0.781250   \n",
       "2   <keras.engine.sequential.Sequential object at ...        0.807292   \n",
       "3   <keras.engine.sequential.Sequential object at ...        0.781250   \n",
       "4   <keras.engine.sequential.Sequential object at ...        0.802083   \n",
       "5   <keras.engine.sequential.Sequential object at ...        0.791667   \n",
       "6   <keras.engine.sequential.Sequential object at ...        0.765625   \n",
       "7   <keras.engine.sequential.Sequential object at ...        0.812500   \n",
       "8   <keras.engine.sequential.Sequential object at ...        0.796875   \n",
       "9   <keras.engine.sequential.Sequential object at ...        0.802083   \n",
       "10  <keras.engine.sequential.Sequential object at ...        0.802083   \n",
       "11  <keras.engine.sequential.Sequential object at ...        0.807292   \n",
       "12  <keras.engine.sequential.Sequential object at ...        0.802083   \n",
       "13  <keras.engine.sequential.Sequential object at ...        0.791667   \n",
       "14  <keras.engine.sequential.Sequential object at ...        0.802083   \n",
       "15  <keras.engine.sequential.Sequential object at ...        0.812500   \n",
       "16  <keras.engine.sequential.Sequential object at ...        0.791667   \n",
       "17  <keras.engine.sequential.Sequential object at ...        0.776042   \n",
       "18  <keras.engine.sequential.Sequential object at ...        0.812500   \n",
       "19  <keras.engine.sequential.Sequential object at ...        0.729167   \n",
       "20  <keras.engine.sequential.Sequential object at ...        0.776042   \n",
       "21  <keras.engine.sequential.Sequential object at ...        0.786458   \n",
       "22  <keras.engine.sequential.Sequential object at ...        0.812500   \n",
       "23  <keras.engine.sequential.Sequential object at ...        0.796875   \n",
       "24  <keras.engine.sequential.Sequential object at ...        0.786458   \n",
       "25  <keras.engine.sequential.Sequential object at ...        0.807292   \n",
       "26  <keras.engine.sequential.Sequential object at ...        0.807292   \n",
       "27  <keras.engine.sequential.Sequential object at ...        0.791667   \n",
       "28  <keras.engine.sequential.Sequential object at ...        0.812500   \n",
       "29  <keras.engine.sequential.Sequential object at ...        0.786458   \n",
       "30  <keras.engine.sequential.Sequential object at ...        0.807292   \n",
       "31  <keras.engine.sequential.Sequential object at ...        0.802083   \n",
       "32  <keras.engine.sequential.Sequential object at ...        0.817708   \n",
       "33  <keras.engine.sequential.Sequential object at ...        0.807292   \n",
       "34  <keras.engine.sequential.Sequential object at ...        0.802083   \n",
       "\n",
       "    accuracy_test  recall_train  recall_test  \n",
       "0        0.698630      0.807292     0.698630  \n",
       "1        0.643836      0.781250     0.643836  \n",
       "2        0.657534      0.807292     0.657534  \n",
       "3        0.643836      0.781250     0.643836  \n",
       "4        0.712329      0.802083     0.712329  \n",
       "5        0.643836      0.791667     0.643836  \n",
       "6        0.630137      0.765625     0.630137  \n",
       "7        0.684932      0.812500     0.684932  \n",
       "8        1.000000      0.796875     1.000000  \n",
       "9        0.698630      0.802083     0.698630  \n",
       "10       0.657534      0.802083     0.657534  \n",
       "11       0.986301      0.807292     0.986301  \n",
       "12       0.657534      0.802083     0.657534  \n",
       "13       0.671233      0.791667     0.671233  \n",
       "14       0.684932      0.802083     0.684932  \n",
       "15       0.684932      0.812500     0.684932  \n",
       "16       0.630137      0.791667     0.630137  \n",
       "17       0.602740      0.776042     0.602740  \n",
       "18       0.698630      0.812500     0.698630  \n",
       "19       0.465753      0.729167     0.465753  \n",
       "20       0.630137      0.776042     0.630137  \n",
       "21       0.630137      0.786458     0.630137  \n",
       "22       0.684932      0.812500     0.684932  \n",
       "23       0.671233      0.796875     0.671233  \n",
       "24       0.643836      0.786458     0.643836  \n",
       "25       0.671233      0.807292     0.671233  \n",
       "26       0.698630      0.807292     0.698630  \n",
       "27       1.000000      0.791667     1.000000  \n",
       "28       0.698630      0.812500     0.698630  \n",
       "29       0.575342      0.786458     0.575342  \n",
       "30       0.671233      0.807292     0.671233  \n",
       "31       0.657534      0.802083     0.657534  \n",
       "32       0.767123      0.817708     0.767123  \n",
       "33       0.657534      0.807292     0.657534  \n",
       "34       0.643836      0.802083     0.643836  "
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Best architecture with mos accuracy"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### best train_accuarcy = 0.78 and test_accuracy = 0.93"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Layers_num</th>\n",
       "      <th>norons_in_layer</th>\n",
       "      <th>model</th>\n",
       "      <th>accuracy_train</th>\n",
       "      <th>accuracy_test</th>\n",
       "      <th>recall_train</th>\n",
       "      <th>recall_test</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>4</td>\n",
       "      <td>[12 18 15  8]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.796875</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.796875</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>8</td>\n",
       "      <td>[ 9 36 30  8 18 40 17 36]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.791667</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.791667</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>5</td>\n",
       "      <td>[10 43 38 32 34]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.807292</td>\n",
       "      <td>0.986301</td>\n",
       "      <td>0.807292</td>\n",
       "      <td>0.986301</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>9</td>\n",
       "      <td>[37 30 46  6  5 10 13 11 26]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.817708</td>\n",
       "      <td>0.767123</td>\n",
       "      <td>0.817708</td>\n",
       "      <td>0.767123</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3</td>\n",
       "      <td>[42 39 13]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.802083</td>\n",
       "      <td>0.712329</td>\n",
       "      <td>0.802083</td>\n",
       "      <td>0.712329</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>3</td>\n",
       "      <td>[37 43 39]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.807292</td>\n",
       "      <td>0.698630</td>\n",
       "      <td>0.807292</td>\n",
       "      <td>0.698630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>8</td>\n",
       "      <td>[45 46  9 27 31  8 30 40]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.807292</td>\n",
       "      <td>0.698630</td>\n",
       "      <td>0.807292</td>\n",
       "      <td>0.698630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>4</td>\n",
       "      <td>[44 40 19 18]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.802083</td>\n",
       "      <td>0.698630</td>\n",
       "      <td>0.802083</td>\n",
       "      <td>0.698630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>6</td>\n",
       "      <td>[49 15 42  7 16  7]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.812500</td>\n",
       "      <td>0.698630</td>\n",
       "      <td>0.812500</td>\n",
       "      <td>0.698630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>8</td>\n",
       "      <td>[24 29 42 45 13 36 26 12]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.812500</td>\n",
       "      <td>0.698630</td>\n",
       "      <td>0.812500</td>\n",
       "      <td>0.698630</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    Layers_num               norons_in_layer  \\\n",
       "8            4                 [12 18 15  8]   \n",
       "27           8     [ 9 36 30  8 18 40 17 36]   \n",
       "11           5              [10 43 38 32 34]   \n",
       "32           9  [37 30 46  6  5 10 13 11 26]   \n",
       "4            3                    [42 39 13]   \n",
       "0            3                    [37 43 39]   \n",
       "26           8     [45 46  9 27 31  8 30 40]   \n",
       "9            4                 [44 40 19 18]   \n",
       "18           6           [49 15 42  7 16  7]   \n",
       "28           8     [24 29 42 45 13 36 26 12]   \n",
       "\n",
       "                                                model  accuracy_train  \\\n",
       "8   <keras.engine.sequential.Sequential object at ...        0.796875   \n",
       "27  <keras.engine.sequential.Sequential object at ...        0.791667   \n",
       "11  <keras.engine.sequential.Sequential object at ...        0.807292   \n",
       "32  <keras.engine.sequential.Sequential object at ...        0.817708   \n",
       "4   <keras.engine.sequential.Sequential object at ...        0.802083   \n",
       "0   <keras.engine.sequential.Sequential object at ...        0.807292   \n",
       "26  <keras.engine.sequential.Sequential object at ...        0.807292   \n",
       "9   <keras.engine.sequential.Sequential object at ...        0.802083   \n",
       "18  <keras.engine.sequential.Sequential object at ...        0.812500   \n",
       "28  <keras.engine.sequential.Sequential object at ...        0.812500   \n",
       "\n",
       "    accuracy_test  recall_train  recall_test  \n",
       "8        1.000000      0.796875     1.000000  \n",
       "27       1.000000      0.791667     1.000000  \n",
       "11       0.986301      0.807292     0.986301  \n",
       "32       0.767123      0.817708     0.767123  \n",
       "4        0.712329      0.802083     0.712329  \n",
       "0        0.698630      0.807292     0.698630  \n",
       "26       0.698630      0.807292     0.698630  \n",
       "9        0.698630      0.802083     0.698630  \n",
       "18       0.698630      0.812500     0.698630  \n",
       "28       0.698630      0.812500     0.698630  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "You must install pydot (`pip install pydot`) and install graphviz (see instructions at https://graphviz.gitlab.io/download/) for plot_model to work.\n"
     ]
    }
   ],
   "source": [
    "from tensorflow.keras.utils import plot_model\n",
    "\n",
    "results_sorted_acc = results.sort_values(by='accuracy_test', ascending=False)\n",
    "\n",
    "display(results_sorted_acc.head(10))\n",
    "best_architecture = results.iloc[0].model\n",
    "\n",
    "plot_model(best_architecture, show_layer_activations=True, show_shapes=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "6/6 [==============================] - 0s 4ms/step\n"
     ]
    }
   ],
   "source": [
    "y_pred = np.argmax(best_architecture.predict(X_test), axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Best architecture with highest Recall"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Layers_num</th>\n",
       "      <th>norons_in_layer</th>\n",
       "      <th>model</th>\n",
       "      <th>accuracy_train</th>\n",
       "      <th>accuracy_test</th>\n",
       "      <th>recall_train</th>\n",
       "      <th>recall_test</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>4</td>\n",
       "      <td>[12 18 15  8]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.796875</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.796875</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>8</td>\n",
       "      <td>[ 9 36 30  8 18 40 17 36]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.791667</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.791667</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>5</td>\n",
       "      <td>[10 43 38 32 34]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.807292</td>\n",
       "      <td>0.986301</td>\n",
       "      <td>0.807292</td>\n",
       "      <td>0.986301</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>9</td>\n",
       "      <td>[37 30 46  6  5 10 13 11 26]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.817708</td>\n",
       "      <td>0.767123</td>\n",
       "      <td>0.817708</td>\n",
       "      <td>0.767123</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3</td>\n",
       "      <td>[42 39 13]</td>\n",
       "      <td>&lt;keras.engine.sequential.Sequential object at ...</td>\n",
       "      <td>0.802083</td>\n",
       "      <td>0.712329</td>\n",
       "      <td>0.802083</td>\n",
       "      <td>0.712329</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    Layers_num               norons_in_layer  \\\n",
       "8            4                 [12 18 15  8]   \n",
       "27           8     [ 9 36 30  8 18 40 17 36]   \n",
       "11           5              [10 43 38 32 34]   \n",
       "32           9  [37 30 46  6  5 10 13 11 26]   \n",
       "4            3                    [42 39 13]   \n",
       "\n",
       "                                                model  accuracy_train  \\\n",
       "8   <keras.engine.sequential.Sequential object at ...        0.796875   \n",
       "27  <keras.engine.sequential.Sequential object at ...        0.791667   \n",
       "11  <keras.engine.sequential.Sequential object at ...        0.807292   \n",
       "32  <keras.engine.sequential.Sequential object at ...        0.817708   \n",
       "4   <keras.engine.sequential.Sequential object at ...        0.802083   \n",
       "\n",
       "    accuracy_test  recall_train  recall_test  \n",
       "8        1.000000      0.796875     1.000000  \n",
       "27       1.000000      0.791667     1.000000  \n",
       "11       0.986301      0.807292     0.986301  \n",
       "32       0.767123      0.817708     0.767123  \n",
       "4        0.712329      0.802083     0.712329  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "You must install pydot (`pip install pydot`) and install graphviz (see instructions at https://graphviz.gitlab.io/download/) for plot_model to work.\n"
     ]
    }
   ],
   "source": [
    "results_sorted_rec = results.sort_values(by='recall_test' ,ascending= False)\n",
    "display(results_sorted_rec.head(5))\n",
    "best_architecture = results.iloc[0].model\n",
    "plot_model(best_architecture, show_layer_activations=True, show_shapes=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "6/6 [==============================] - 0s 4ms/step\n"
     ]
    }
   ],
   "source": [
    "y_pred = np.argmax(best_architecture.predict(X_test), axis=1) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}