ResNet-VGG-Inception-4Piece-Indiv-Concat-Model.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "fd1efdf1",
   "metadata": {},
   "source": [
    "### Importing Necessary Libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "id": "0938de48",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import pickle\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras.preprocessing.image import load_img, img_to_array\n",
    "\n",
    "from tensorflow.keras.layers import Input, Flatten, Dense, Dropout, GlobalAvgPool2D, BatchNormalization, Concatenate\n",
    "# from tensorflow.keras.applications import EfficientNetV2L\n",
    "from tensorflow.keras.regularizers import l2\n",
    "\n",
    "from tensorflow.keras.optimizers import Adam\n",
    "\n",
    "from sklearn.metrics import precision_recall_fscore_support as score"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c954b162",
   "metadata": {},
   "source": [
    "## Data"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "68302673",
   "metadata": {},
   "source": [
    "A detailed discussion and visualization of the data can be seen in [here](Data-Overview.ipynb)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "id": "75b06a70",
   "metadata": {},
   "outputs": [],
   "source": [
    "labels = pd.read_csv(\"./Data/train_labels.csv\")\n",
    "images = ['./Data/train/train/'+str(name)+'.tif' for name in labels['id']]\n",
    "labels['images'] = images\n",
    "labels = labels[['id', 'images', 'label']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "fd2fe082",
   "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>id</th>\n",
       "      <th>images</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>./Data/train/train/0.tif</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>./Data/train/train/1.tif</td>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>./Data/train/train/2.tif</td>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>./Data/train/train/3.tif</td>\n",
       "      <td>14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>./Data/train/train/4.tif</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15995</th>\n",
       "      <td>15995</td>\n",
       "      <td>./Data/train/train/15995.tif</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15996</th>\n",
       "      <td>15996</td>\n",
       "      <td>./Data/train/train/15996.tif</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15997</th>\n",
       "      <td>15997</td>\n",
       "      <td>./Data/train/train/15997.tif</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15998</th>\n",
       "      <td>15998</td>\n",
       "      <td>./Data/train/train/15998.tif</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15999</th>\n",
       "      <td>15999</td>\n",
       "      <td>./Data/train/train/15999.tif</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>16000 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          id                        images  label\n",
       "0          0      ./Data/train/train/0.tif      1\n",
       "1          1      ./Data/train/train/1.tif     13\n",
       "2          2      ./Data/train/train/2.tif     13\n",
       "3          3      ./Data/train/train/3.tif     14\n",
       "4          4      ./Data/train/train/4.tif      6\n",
       "...      ...                           ...    ...\n",
       "15995  15995  ./Data/train/train/15995.tif      2\n",
       "15996  15996  ./Data/train/train/15996.tif     15\n",
       "15997  15997  ./Data/train/train/15997.tif      3\n",
       "15998  15998  ./Data/train/train/15998.tif      9\n",
       "15999  15999  ./Data/train/train/15999.tif      9\n",
       "\n",
       "[16000 rows x 3 columns]"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "id": "276e2059",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_labels = pd.read_csv('./Data/train_labels_1.csv')\n",
    "test_labels = pd.read_csv('./Data/test_labels_1.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "2aea9ea9",
   "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>id</th>\n",
       "      <th>images</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>./Data/train/train/0.tif</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>./Data/train/train/1.tif</td>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>./Data/train/train/2.tif</td>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>./Data/train/train/3.tif</td>\n",
       "      <td>14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>./Data/train/train/4.tif</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15195</th>\n",
       "      <td>15995</td>\n",
       "      <td>./Data/train/train/15995.tif</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15196</th>\n",
       "      <td>15996</td>\n",
       "      <td>./Data/train/train/15996.tif</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15197</th>\n",
       "      <td>15997</td>\n",
       "      <td>./Data/train/train/15997.tif</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15198</th>\n",
       "      <td>15998</td>\n",
       "      <td>./Data/train/train/15998.tif</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15199</th>\n",
       "      <td>15999</td>\n",
       "      <td>./Data/train/train/15999.tif</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>15200 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          id                        images  label\n",
       "0          0      ./Data/train/train/0.tif      1\n",
       "1          1      ./Data/train/train/1.tif     13\n",
       "2          2      ./Data/train/train/2.tif     13\n",
       "3          3      ./Data/train/train/3.tif     14\n",
       "4          4      ./Data/train/train/4.tif      6\n",
       "...      ...                           ...    ...\n",
       "15195  15995  ./Data/train/train/15995.tif      2\n",
       "15196  15996  ./Data/train/train/15996.tif     15\n",
       "15197  15997  ./Data/train/train/15997.tif      3\n",
       "15198  15998  ./Data/train/train/15998.tif      9\n",
       "15199  15999  ./Data/train/train/15999.tif      9\n",
       "\n",
       "[15200 rows x 3 columns]"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "id": "b9ea19c9",
   "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>id</th>\n",
       "      <th>images</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>18</td>\n",
       "      <td>./Data/train/train/18.tif</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>25</td>\n",
       "      <td>./Data/train/train/25.tif</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>33</td>\n",
       "      <td>./Data/train/train/33.tif</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>41</td>\n",
       "      <td>./Data/train/train/41.tif</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>65</td>\n",
       "      <td>./Data/train/train/65.tif</td>\n",
       "      <td>14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>795</th>\n",
       "      <td>15816</td>\n",
       "      <td>./Data/train/train/15816.tif</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>796</th>\n",
       "      <td>15824</td>\n",
       "      <td>./Data/train/train/15824.tif</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>797</th>\n",
       "      <td>15832</td>\n",
       "      <td>./Data/train/train/15832.tif</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>798</th>\n",
       "      <td>15863</td>\n",
       "      <td>./Data/train/train/15863.tif</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>799</th>\n",
       "      <td>15994</td>\n",
       "      <td>./Data/train/train/15994.tif</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>800 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        id                        images  label\n",
       "0       18     ./Data/train/train/18.tif     11\n",
       "1       25     ./Data/train/train/25.tif      8\n",
       "2       33     ./Data/train/train/33.tif      6\n",
       "3       41     ./Data/train/train/41.tif      1\n",
       "4       65     ./Data/train/train/65.tif     14\n",
       "..     ...                           ...    ...\n",
       "795  15816  ./Data/train/train/15816.tif      2\n",
       "796  15824  ./Data/train/train/15824.tif      2\n",
       "797  15832  ./Data/train/train/15832.tif     11\n",
       "798  15863  ./Data/train/train/15863.tif     15\n",
       "799  15994  ./Data/train/train/15994.tif      6\n",
       "\n",
       "[800 rows x 3 columns]"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "id": "29190fd4",
   "metadata": {},
   "outputs": [],
   "source": [
    "class_labels = list(set(labels['label']))\n",
    "class_names = [\n",
    "    'letter', 'form', 'email', 'handwritten', 'advertisement', 'scientific report', 'scientific publication',\n",
    "    'specification', 'file folder', 'news article', 'budget', 'invoice', 'presentation', 'questionnaire', 'resume',\n",
    "    'memo'\n",
    "]\n",
    "label_names = pd.DataFrame({\n",
    "    'labels': class_labels,\n",
    "    'names': class_names\n",
    "})"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c2e251c1",
   "metadata": {},
   "source": [
    "## Data Preparation"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "85b2ccd1",
   "metadata": {},
   "source": [
    "We have already generated the Extracted Feature Vectors from the ResNet Model now we train using that, so we load those feature vectors."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "id": "37a36e2b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(16000, 5, 2048)"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fine_tune_data_resnet = np.load('./Data/ResNet-4P-train-precompute/whole_train_data_precomp.npy')\n",
    "fine_tune_data_resnet.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "id": "4cc9fe34",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(16000, 5, 512)"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fine_tune_data_vgg = np.load('./Data/VGG16-4P-train-precompute/whole_train_data_precomp.npy')\n",
    "fine_tune_data_vgg.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "id": "40e99544",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(16000, 5, 1536)"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fine_tune_data_inception_resnet = np.load('./Data/Inception-ResNet-4P-train-precompute/whole_train_data_precomp.npy')\n",
    "fine_tune_data_inception_resnet.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "6e4e8b18",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(16000, 1536)"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# fine_tune_data = np.concatenate([fine_tune_data_resnet, fine_tune_data_vgg, fine_tune_data_inception_resnet], axis=-1)\n",
    "# fine_tune_data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "id": "568ffcb1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(16000,)"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fine_tune_labels = labels['label'].to_numpy()\n",
    "fine_tune_labels.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "id": "937e1e86",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(15200, 5, 2048)"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fine_tune_data_resnet_train = fine_tune_data_resnet[list(train_labels['id'])]\n",
    "fine_tune_data_resnet_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "id": "79d0a4f9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(15200, 5, 512)"
      ]
     },
     "execution_count": 83,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fine_tune_data_vgg_train = fine_tune_data_vgg[list(train_labels['id'])]\n",
    "fine_tune_data_vgg_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "id": "91479f50",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(15200, 5, 1536)"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fine_tune_data_inception_resnet_train = fine_tune_data_inception_resnet[list(train_labels['id'])]\n",
    "fine_tune_data_inception_resnet_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "af08c830",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(15200,)"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fine_tune_labels_train = train_labels['label'].to_numpy()\n",
    "fine_tune_labels_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "id": "c00f0b32",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(800, 5, 2048)"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fine_tune_data_resnet_test = fine_tune_data_resnet[list(test_labels['id'])]\n",
    "fine_tune_data_resnet_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "id": "213fc255",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(800, 5, 512)"
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fine_tune_data_vgg_test = fine_tune_data_vgg[list(test_labels['id'])]\n",
    "fine_tune_data_vgg_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "id": "32995292",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(800, 5, 1536)"
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fine_tune_data_inception_resnet_test = fine_tune_data_inception_resnet[list(test_labels['id'])]\n",
    "fine_tune_data_inception_resnet_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "id": "16f4bb93",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(800,)"
      ]
     },
     "execution_count": 88,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fine_tune_labels_test = test_labels['label'].to_numpy()\n",
    "fine_tune_labels_test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dff1f776",
   "metadata": {},
   "source": [
    "## Fine-Tuning Models"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7380947f",
   "metadata": {},
   "source": [
    "The fine-tuning part of our model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "id": "96a2bdfe",
   "metadata": {},
   "outputs": [],
   "source": [
    "def ftmodel():\n",
    "    inputs1 = Input(shape=(5, 2048))\n",
    "    inputs2 = Input(shape=(5, 512))\n",
    "    inputs3 = Input(shape=(5, 1536))\n",
    "\n",
    "    top_dropout_rate = 0.5\n",
    "    \n",
    "    x = Flatten()(inputs1)\n",
    "    x = BatchNormalization()(x)\n",
    "    x = Dropout(top_dropout_rate)(x)\n",
    "    x = Dense(512, kernel_initializer=\"he_normal\", kernel_regularizer=l2(5e-5), activation='relu')(x)\n",
    "    x = BatchNormalization()(x)\n",
    "    \n",
    "    y = Flatten()(inputs2)\n",
    "    y = BatchNormalization()(y)\n",
    "    y = Dropout(top_dropout_rate)(y)\n",
    "    y = Dense(512, kernel_initializer=\"he_normal\", kernel_regularizer=l2(5e-5), activation='relu')(y)\n",
    "    \n",
    "    z = Flatten()(inputs3)\n",
    "    z = BatchNormalization()(z)\n",
    "    z = Dropout(top_dropout_rate)(z)\n",
    "    z = Dense(512, kernel_initializer=\"he_normal\", kernel_regularizer=l2(5e-5), activation='relu')(z)\n",
    "    \n",
    "    a = Concatenate()([x, y, z])\n",
    "    \n",
    "    # x = Dropout(top_dropout_rate, name='top_dropout_1')(x)\n",
    "    # x = Dense(1024, name='top_dense_1', kernel_initializer=\"he_normal\", kernel_regularizer=l2(5e-5), activation='relu')(x)\n",
    "    # x = BatchNormalization()(x)\n",
    "    a = Dropout(top_dropout_rate)(a)\n",
    "    a = Dense(512, kernel_initializer=\"he_normal\", kernel_regularizer=l2(5e-5), activation='relu')(a)\n",
    "    a = BatchNormalization()(a)\n",
    "    \n",
    "    a = Dropout(top_dropout_rate)(a)\n",
    "    # x = Dense(256, name='top_dense_3', kernel_initializer=\"he_normal\", kernel_regularizer=l2(5e-5), activation='relu')(x)\n",
    "    # x = BatchNormalization()(x)\n",
    "    # x = Dropout(top_dropout_rate, name='top_dropout_4')(x)\n",
    "    outputs = Dense(16, activation='softmax', name='pred', kernel_regularizer=l2(5e-5))(a)\n",
    "\n",
    "    model1 = tf.keras.Model([inputs1, inputs2, inputs3], outputs, name='RestNet-VGG-Inception-ResNet-FineTune')\n",
    "    return model1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "id": "c6574006",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"RestNet-VGG-Inception-ResNet-FineTune\"\n",
      "__________________________________________________________________________________________________\n",
      " Layer (type)                   Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      " input_11 (InputLayer)          [(None, 5, 2048)]    0           []                               \n",
      "                                                                                                  \n",
      " flatten_10 (Flatten)           (None, 10240)        0           ['input_11[0][0]']               \n",
      "                                                                                                  \n",
      " input_12 (InputLayer)          [(None, 5, 512)]     0           []                               \n",
      "                                                                                                  \n",
      " input_13 (InputLayer)          [(None, 5, 1536)]    0           []                               \n",
      "                                                                                                  \n",
      " batch_normalization_18 (BatchN  (None, 10240)       40960       ['flatten_10[0][0]']             \n",
      " ormalization)                                                                                    \n",
      "                                                                                                  \n",
      " flatten_11 (Flatten)           (None, 2560)         0           ['input_12[0][0]']               \n",
      "                                                                                                  \n",
      " flatten_12 (Flatten)           (None, 7680)         0           ['input_13[0][0]']               \n",
      "                                                                                                  \n",
      " dropout_10 (Dropout)           (None, 10240)        0           ['batch_normalization_18[0][0]'] \n",
      "                                                                                                  \n",
      " batch_normalization_20 (BatchN  (None, 2560)        10240       ['flatten_11[0][0]']             \n",
      " ormalization)                                                                                    \n",
      "                                                                                                  \n",
      " batch_normalization_21 (BatchN  (None, 7680)        30720       ['flatten_12[0][0]']             \n",
      " ormalization)                                                                                    \n",
      "                                                                                                  \n",
      " dense_8 (Dense)                (None, 512)          5243392     ['dropout_10[0][0]']             \n",
      "                                                                                                  \n",
      " dropout_11 (Dropout)           (None, 2560)         0           ['batch_normalization_20[0][0]'] \n",
      "                                                                                                  \n",
      " dropout_12 (Dropout)           (None, 7680)         0           ['batch_normalization_21[0][0]'] \n",
      "                                                                                                  \n",
      " batch_normalization_19 (BatchN  (None, 512)         2048        ['dense_8[0][0]']                \n",
      " ormalization)                                                                                    \n",
      "                                                                                                  \n",
      " dense_9 (Dense)                (None, 512)          1311232     ['dropout_11[0][0]']             \n",
      "                                                                                                  \n",
      " dense_10 (Dense)               (None, 512)          3932672     ['dropout_12[0][0]']             \n",
      "                                                                                                  \n",
      " concatenate_2 (Concatenate)    (None, 1536)         0           ['batch_normalization_19[0][0]', \n",
      "                                                                  'dense_9[0][0]',                \n",
      "                                                                  'dense_10[0][0]']               \n",
      "                                                                                                  \n",
      " dropout_13 (Dropout)           (None, 1536)         0           ['concatenate_2[0][0]']          \n",
      "                                                                                                  \n",
      " dense_11 (Dense)               (None, 512)          786944      ['dropout_13[0][0]']             \n",
      "                                                                                                  \n",
      " batch_normalization_22 (BatchN  (None, 512)         2048        ['dense_11[0][0]']               \n",
      " ormalization)                                                                                    \n",
      "                                                                                                  \n",
      " dropout_14 (Dropout)           (None, 512)          0           ['batch_normalization_22[0][0]'] \n",
      "                                                                                                  \n",
      " pred (Dense)                   (None, 16)           8208        ['dropout_14[0][0]']             \n",
      "                                                                                                  \n",
      "==================================================================================================\n",
      "Total params: 11,368,464\n",
      "Trainable params: 11,325,456\n",
      "Non-trainable params: 43,008\n",
      "__________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model1 = ftmodel()\n",
    "model1.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "id": "04f47ff8",
   "metadata": {},
   "outputs": [],
   "source": [
    "# tf.keras.utils.plot_model(model1, show_shapes=True, to_file='./Images/ResNet-VGG-Inception-ResNet-4Piece-PCA-Model.png')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e56505e2",
   "metadata": {},
   "source": [
    "### Model Training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "id": "96524578",
   "metadata": {},
   "outputs": [],
   "source": [
    "model1.compile(loss = 'sparse_categorical_crossentropy', metrics = ['accuracy'], optimizer = Adam(learning_rate = 0.0003))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "id": "20df8897",
   "metadata": {},
   "outputs": [],
   "source": [
    "EPOCHS = 1000\n",
    "checkpoint_filepaths = ['./Models/ResNet-VGG-Inception-4Piece-Indiv-Concat/auto-ResNet-VGG-Inception-ResNet-FT-model-weight']\n",
    "model_checkpoint_callback = [tf.keras.callbacks.ModelCheckpoint(\n",
    "    filepath=checkpoint_filepath,\n",
    "    save_weights_only=True,\n",
    "    monitor = \"loss\",\n",
    "    mode='min',\n",
    "    save_best_only=True) for checkpoint_filepath in checkpoint_filepaths]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "id": "03354278",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-10-09 17:56:49.492498: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:113] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "18/19 [===========================>..] - ETA: 0s - loss: 2.6925 - accuracy: 0.3557"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-10-09 17:56:50.765168: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:113] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "19/19 [==============================] - 2s 70ms/step - loss: 2.6650 - accuracy: 0.3615 - val_loss: 2.1840 - val_accuracy: 0.3725\n",
      "Epoch 2/1000\n",
      "19/19 [==============================] - 1s 57ms/step - loss: 1.9439 - accuracy: 0.5116 - val_loss: 1.8921 - val_accuracy: 0.5025\n",
      "Epoch 3/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 1.7399 - accuracy: 0.5618 - val_loss: 1.8229 - val_accuracy: 0.5375\n",
      "Epoch 4/1000\n",
      "19/19 [==============================] - 1s 49ms/step - loss: 1.5852 - accuracy: 0.6024 - val_loss: 1.7414 - val_accuracy: 0.5737\n",
      "Epoch 5/1000\n",
      "19/19 [==============================] - 1s 50ms/step - loss: 1.4732 - accuracy: 0.6289 - val_loss: 1.6793 - val_accuracy: 0.5938\n",
      "Epoch 6/1000\n",
      "19/19 [==============================] - 1s 47ms/step - loss: 1.3657 - accuracy: 0.6555 - val_loss: 1.6296 - val_accuracy: 0.5950\n",
      "Epoch 7/1000\n",
      "19/19 [==============================] - 1s 46ms/step - loss: 1.2854 - accuracy: 0.6810 - val_loss: 1.5629 - val_accuracy: 0.6250\n",
      "Epoch 8/1000\n",
      "19/19 [==============================] - 1s 47ms/step - loss: 1.2286 - accuracy: 0.6955 - val_loss: 1.4825 - val_accuracy: 0.6513\n",
      "Epoch 9/1000\n",
      "19/19 [==============================] - 1s 57ms/step - loss: 1.1560 - accuracy: 0.7121 - val_loss: 1.4341 - val_accuracy: 0.6625\n",
      "Epoch 10/1000\n",
      "19/19 [==============================] - 1s 47ms/step - loss: 1.0847 - accuracy: 0.7304 - val_loss: 1.3578 - val_accuracy: 0.6775\n",
      "Epoch 11/1000\n",
      "19/19 [==============================] - 1s 45ms/step - loss: 1.0467 - accuracy: 0.7412 - val_loss: 1.2951 - val_accuracy: 0.7000\n",
      "Epoch 12/1000\n",
      "19/19 [==============================] - 1s 47ms/step - loss: 1.0023 - accuracy: 0.7576 - val_loss: 1.2323 - val_accuracy: 0.7200\n",
      "Epoch 13/1000\n",
      "19/19 [==============================] - 1s 46ms/step - loss: 0.9376 - accuracy: 0.7714 - val_loss: 1.2057 - val_accuracy: 0.7188\n",
      "Epoch 14/1000\n",
      "19/19 [==============================] - 1s 53ms/step - loss: 0.8981 - accuracy: 0.7855 - val_loss: 1.1577 - val_accuracy: 0.7150\n",
      "Epoch 15/1000\n",
      "19/19 [==============================] - 1s 47ms/step - loss: 0.8719 - accuracy: 0.7921 - val_loss: 1.1262 - val_accuracy: 0.7225\n",
      "Epoch 16/1000\n",
      "19/19 [==============================] - 1s 46ms/step - loss: 0.8275 - accuracy: 0.8057 - val_loss: 1.0935 - val_accuracy: 0.7250\n",
      "Epoch 17/1000\n",
      "19/19 [==============================] - 1s 47ms/step - loss: 0.7881 - accuracy: 0.8167 - val_loss: 1.0597 - val_accuracy: 0.7350\n",
      "Epoch 18/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.7568 - accuracy: 0.8248 - val_loss: 1.0353 - val_accuracy: 0.7512\n",
      "Epoch 19/1000\n",
      "19/19 [==============================] - 1s 46ms/step - loss: 0.7230 - accuracy: 0.8386 - val_loss: 1.0207 - val_accuracy: 0.7462\n",
      "Epoch 20/1000\n",
      "19/19 [==============================] - 1s 46ms/step - loss: 0.6968 - accuracy: 0.8432 - val_loss: 1.0129 - val_accuracy: 0.7462\n",
      "Epoch 21/1000\n",
      "19/19 [==============================] - 1s 46ms/step - loss: 0.6747 - accuracy: 0.8469 - val_loss: 0.9830 - val_accuracy: 0.7562\n",
      "Epoch 22/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.6528 - accuracy: 0.8571 - val_loss: 0.9925 - val_accuracy: 0.7537\n",
      "Epoch 23/1000\n",
      "19/19 [==============================] - 1s 47ms/step - loss: 0.6250 - accuracy: 0.8652 - val_loss: 0.9791 - val_accuracy: 0.7625\n",
      "Epoch 24/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.6086 - accuracy: 0.8722 - val_loss: 0.9757 - val_accuracy: 0.7662\n",
      "Epoch 25/1000\n",
      "19/19 [==============================] - 1s 47ms/step - loss: 0.5793 - accuracy: 0.8793 - val_loss: 0.9879 - val_accuracy: 0.7650\n",
      "Epoch 26/1000\n",
      "19/19 [==============================] - 1s 49ms/step - loss: 0.5723 - accuracy: 0.8787 - val_loss: 0.9682 - val_accuracy: 0.7700\n",
      "Epoch 27/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.5546 - accuracy: 0.8849 - val_loss: 0.9837 - val_accuracy: 0.7675\n",
      "Epoch 28/1000\n",
      "19/19 [==============================] - 1s 47ms/step - loss: 0.5414 - accuracy: 0.8891 - val_loss: 0.9892 - val_accuracy: 0.7725\n",
      "Epoch 29/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.5264 - accuracy: 0.8951 - val_loss: 0.9994 - val_accuracy: 0.7788\n",
      "Epoch 30/1000\n",
      "19/19 [==============================] - 1s 47ms/step - loss: 0.5000 - accuracy: 0.9047 - val_loss: 0.9943 - val_accuracy: 0.7788\n",
      "Epoch 31/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.4951 - accuracy: 0.9063 - val_loss: 1.0129 - val_accuracy: 0.7788\n",
      "Epoch 32/1000\n",
      "19/19 [==============================] - 1s 46ms/step - loss: 0.4800 - accuracy: 0.9114 - val_loss: 1.0070 - val_accuracy: 0.7675\n",
      "Epoch 33/1000\n",
      "19/19 [==============================] - 1s 51ms/step - loss: 0.4668 - accuracy: 0.9149 - val_loss: 1.0190 - val_accuracy: 0.7812\n",
      "Epoch 34/1000\n",
      "19/19 [==============================] - 1s 49ms/step - loss: 0.4606 - accuracy: 0.9181 - val_loss: 1.0206 - val_accuracy: 0.7788\n",
      "Epoch 35/1000\n",
      "19/19 [==============================] - 1s 46ms/step - loss: 0.4405 - accuracy: 0.9216 - val_loss: 1.0294 - val_accuracy: 0.7850\n",
      "Epoch 36/1000\n",
      "19/19 [==============================] - 1s 51ms/step - loss: 0.4346 - accuracy: 0.9259 - val_loss: 1.0477 - val_accuracy: 0.7825\n",
      "Epoch 37/1000\n",
      "19/19 [==============================] - 1s 45ms/step - loss: 0.4267 - accuracy: 0.9272 - val_loss: 1.0401 - val_accuracy: 0.7763\n",
      "Epoch 38/1000\n",
      "19/19 [==============================] - 1s 50ms/step - loss: 0.4212 - accuracy: 0.9288 - val_loss: 1.0551 - val_accuracy: 0.7700\n",
      "Epoch 39/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.4176 - accuracy: 0.9303 - val_loss: 1.0664 - val_accuracy: 0.7775\n",
      "Epoch 40/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.4080 - accuracy: 0.9358 - val_loss: 1.0748 - val_accuracy: 0.7738\n",
      "Epoch 41/1000\n",
      "19/19 [==============================] - 1s 49ms/step - loss: 0.3937 - accuracy: 0.9372 - val_loss: 1.0854 - val_accuracy: 0.7788\n",
      "Epoch 42/1000\n",
      "19/19 [==============================] - 1s 46ms/step - loss: 0.3928 - accuracy: 0.9386 - val_loss: 1.0722 - val_accuracy: 0.7775\n",
      "Epoch 43/1000\n",
      "19/19 [==============================] - 1s 51ms/step - loss: 0.3859 - accuracy: 0.9413 - val_loss: 1.0643 - val_accuracy: 0.7825\n",
      "Epoch 44/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.3757 - accuracy: 0.9458 - val_loss: 1.0779 - val_accuracy: 0.7850\n",
      "Epoch 45/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.3668 - accuracy: 0.9493 - val_loss: 1.0839 - val_accuracy: 0.7812\n",
      "Epoch 46/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.3693 - accuracy: 0.9468 - val_loss: 1.1051 - val_accuracy: 0.7788\n",
      "Epoch 47/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.3579 - accuracy: 0.9505 - val_loss: 1.1046 - val_accuracy: 0.7850\n",
      "Epoch 48/1000\n",
      "19/19 [==============================] - 1s 47ms/step - loss: 0.3510 - accuracy: 0.9520 - val_loss: 1.1148 - val_accuracy: 0.7850\n",
      "Epoch 49/1000\n",
      "19/19 [==============================] - 1s 51ms/step - loss: 0.3493 - accuracy: 0.9552 - val_loss: 1.1103 - val_accuracy: 0.7850\n",
      "Epoch 50/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.3502 - accuracy: 0.9520 - val_loss: 1.1093 - val_accuracy: 0.7825\n",
      "Epoch 51/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.3458 - accuracy: 0.9559 - val_loss: 1.1109 - val_accuracy: 0.7862\n",
      "Epoch 52/1000\n",
      "19/19 [==============================] - 1s 50ms/step - loss: 0.3337 - accuracy: 0.9593 - val_loss: 1.1294 - val_accuracy: 0.7812\n",
      "Epoch 53/1000\n",
      "19/19 [==============================] - 1s 31ms/step - loss: 0.3369 - accuracy: 0.9570 - val_loss: 1.1363 - val_accuracy: 0.7800\n",
      "Epoch 54/1000\n",
      "19/19 [==============================] - 1s 49ms/step - loss: 0.3290 - accuracy: 0.9580 - val_loss: 1.1277 - val_accuracy: 0.7875\n",
      "Epoch 55/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.3310 - accuracy: 0.9601 - val_loss: 1.1408 - val_accuracy: 0.7788\n",
      "Epoch 56/1000\n",
      "19/19 [==============================] - 1s 47ms/step - loss: 0.3209 - accuracy: 0.9626 - val_loss: 1.1286 - val_accuracy: 0.7763\n",
      "Epoch 57/1000\n",
      "19/19 [==============================] - 1s 49ms/step - loss: 0.3189 - accuracy: 0.9630 - val_loss: 1.1518 - val_accuracy: 0.7788\n",
      "Epoch 58/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "19/19 [==============================] - 1s 47ms/step - loss: 0.3137 - accuracy: 0.9653 - val_loss: 1.1528 - val_accuracy: 0.7788\n",
      "Epoch 59/1000\n",
      "19/19 [==============================] - 1s 51ms/step - loss: 0.3117 - accuracy: 0.9666 - val_loss: 1.1521 - val_accuracy: 0.7788\n",
      "Epoch 60/1000\n",
      "19/19 [==============================] - 1s 33ms/step - loss: 0.3136 - accuracy: 0.9629 - val_loss: 1.1438 - val_accuracy: 0.7900\n",
      "Epoch 61/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.3061 - accuracy: 0.9682 - val_loss: 1.1414 - val_accuracy: 0.7912\n",
      "Epoch 62/1000\n",
      "19/19 [==============================] - 1s 50ms/step - loss: 0.3051 - accuracy: 0.9674 - val_loss: 1.1653 - val_accuracy: 0.7900\n",
      "Epoch 63/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.3039 - accuracy: 0.9671 - val_loss: 1.1552 - val_accuracy: 0.7825\n",
      "Epoch 64/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.3038 - accuracy: 0.9685 - val_loss: 1.1495 - val_accuracy: 0.7900\n",
      "Epoch 65/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.2980 - accuracy: 0.9695 - val_loss: 1.1630 - val_accuracy: 0.7912\n",
      "Epoch 66/1000\n",
      "19/19 [==============================] - 1s 55ms/step - loss: 0.2958 - accuracy: 0.9707 - val_loss: 1.1452 - val_accuracy: 0.7925\n",
      "Epoch 67/1000\n",
      "19/19 [==============================] - 1s 26ms/step - loss: 0.2958 - accuracy: 0.9703 - val_loss: 1.1674 - val_accuracy: 0.7950\n",
      "Epoch 68/1000\n",
      "19/19 [==============================] - 1s 46ms/step - loss: 0.2928 - accuracy: 0.9722 - val_loss: 1.1519 - val_accuracy: 0.7875\n",
      "Epoch 69/1000\n",
      "19/19 [==============================] - 1s 52ms/step - loss: 0.2910 - accuracy: 0.9706 - val_loss: 1.1668 - val_accuracy: 0.7862\n",
      "Epoch 70/1000\n",
      "19/19 [==============================] - 1s 49ms/step - loss: 0.2860 - accuracy: 0.9753 - val_loss: 1.1511 - val_accuracy: 0.7950\n",
      "Epoch 71/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2928 - accuracy: 0.9719 - val_loss: 1.1843 - val_accuracy: 0.7987\n",
      "Epoch 72/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.2882 - accuracy: 0.9724 - val_loss: 1.1745 - val_accuracy: 0.7950\n",
      "Epoch 73/1000\n",
      "19/19 [==============================] - 1s 47ms/step - loss: 0.2830 - accuracy: 0.9755 - val_loss: 1.2071 - val_accuracy: 0.7950\n",
      "Epoch 74/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.2846 - accuracy: 0.9747 - val_loss: 1.1869 - val_accuracy: 0.8012\n",
      "Epoch 75/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.2787 - accuracy: 0.9763 - val_loss: 1.1722 - val_accuracy: 0.7937\n",
      "Epoch 76/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.2816 - accuracy: 0.9750 - val_loss: 1.1939 - val_accuracy: 0.7925\n",
      "Epoch 77/1000\n",
      "19/19 [==============================] - 1s 47ms/step - loss: 0.2781 - accuracy: 0.9774 - val_loss: 1.1831 - val_accuracy: 0.7987\n",
      "Epoch 78/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2791 - accuracy: 0.9759 - val_loss: 1.1983 - val_accuracy: 0.7950\n",
      "Epoch 79/1000\n",
      "19/19 [==============================] - 1s 51ms/step - loss: 0.2776 - accuracy: 0.9764 - val_loss: 1.2127 - val_accuracy: 0.8050\n",
      "Epoch 80/1000\n",
      "19/19 [==============================] - 1s 50ms/step - loss: 0.2741 - accuracy: 0.9769 - val_loss: 1.2137 - val_accuracy: 0.7937\n",
      "Epoch 81/1000\n",
      "19/19 [==============================] - 1s 54ms/step - loss: 0.2672 - accuracy: 0.9782 - val_loss: 1.1952 - val_accuracy: 0.7937\n",
      "Epoch 82/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2700 - accuracy: 0.9786 - val_loss: 1.1899 - val_accuracy: 0.7937\n",
      "Epoch 83/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.2724 - accuracy: 0.9774 - val_loss: 1.1772 - val_accuracy: 0.7800\n",
      "Epoch 84/1000\n",
      "19/19 [==============================] - 1s 50ms/step - loss: 0.2666 - accuracy: 0.9784 - val_loss: 1.1928 - val_accuracy: 0.7987\n",
      "Epoch 85/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2680 - accuracy: 0.9781 - val_loss: 1.1752 - val_accuracy: 0.7937\n",
      "Epoch 86/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.2675 - accuracy: 0.9797 - val_loss: 1.2071 - val_accuracy: 0.7837\n",
      "Epoch 87/1000\n",
      "19/19 [==============================] - 1s 47ms/step - loss: 0.2623 - accuracy: 0.9811 - val_loss: 1.1933 - val_accuracy: 0.8012\n",
      "Epoch 88/1000\n",
      "19/19 [==============================] - 0s 24ms/step - loss: 0.2633 - accuracy: 0.9803 - val_loss: 1.2304 - val_accuracy: 0.7912\n",
      "Epoch 89/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.2644 - accuracy: 0.9803 - val_loss: 1.2218 - val_accuracy: 0.8000\n",
      "Epoch 90/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2643 - accuracy: 0.9805 - val_loss: 1.2196 - val_accuracy: 0.8012\n",
      "Epoch 91/1000\n",
      "19/19 [==============================] - 1s 49ms/step - loss: 0.2621 - accuracy: 0.9813 - val_loss: 1.2266 - val_accuracy: 0.7862\n",
      "Epoch 92/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.2624 - accuracy: 0.9800 - val_loss: 1.2425 - val_accuracy: 0.7987\n",
      "Epoch 93/1000\n",
      "19/19 [==============================] - 1s 46ms/step - loss: 0.2608 - accuracy: 0.9818 - val_loss: 1.2297 - val_accuracy: 0.7925\n",
      "Epoch 94/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.2548 - accuracy: 0.9823 - val_loss: 1.2334 - val_accuracy: 0.7937\n",
      "Epoch 95/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.2568 - accuracy: 0.9811 - val_loss: 1.2565 - val_accuracy: 0.7875\n",
      "Epoch 96/1000\n",
      "19/19 [==============================] - 1s 51ms/step - loss: 0.2521 - accuracy: 0.9831 - val_loss: 1.2424 - val_accuracy: 0.8000\n",
      "Epoch 97/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2562 - accuracy: 0.9814 - val_loss: 1.2509 - val_accuracy: 0.7975\n",
      "Epoch 98/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.2524 - accuracy: 0.9822 - val_loss: 1.2671 - val_accuracy: 0.7912\n",
      "Epoch 99/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2542 - accuracy: 0.9828 - val_loss: 1.2960 - val_accuracy: 0.7912\n",
      "Epoch 100/1000\n",
      "19/19 [==============================] - 1s 25ms/step - loss: 0.2554 - accuracy: 0.9818 - val_loss: 1.2918 - val_accuracy: 0.8012\n",
      "Epoch 101/1000\n",
      "19/19 [==============================] - 1s 46ms/step - loss: 0.2520 - accuracy: 0.9825 - val_loss: 1.2964 - val_accuracy: 0.7950\n",
      "Epoch 102/1000\n",
      "19/19 [==============================] - 1s 47ms/step - loss: 0.2493 - accuracy: 0.9843 - val_loss: 1.2615 - val_accuracy: 0.7900\n",
      "Epoch 103/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2504 - accuracy: 0.9834 - val_loss: 1.2371 - val_accuracy: 0.8112\n",
      "Epoch 104/1000\n",
      "19/19 [==============================] - 1s 49ms/step - loss: 0.2440 - accuracy: 0.9851 - val_loss: 1.2621 - val_accuracy: 0.8025\n",
      "Epoch 105/1000\n",
      "19/19 [==============================] - 1s 31ms/step - loss: 0.2446 - accuracy: 0.9857 - val_loss: 1.2808 - val_accuracy: 0.7950\n",
      "Epoch 106/1000\n",
      "19/19 [==============================] - 1s 49ms/step - loss: 0.2437 - accuracy: 0.9860 - val_loss: 1.2463 - val_accuracy: 0.8050\n",
      "Epoch 107/1000\n",
      "19/19 [==============================] - 1s 31ms/step - loss: 0.2466 - accuracy: 0.9851 - val_loss: 1.2656 - val_accuracy: 0.7950\n",
      "Epoch 108/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.2469 - accuracy: 0.9843 - val_loss: 1.2319 - val_accuracy: 0.7912\n",
      "Epoch 109/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.2506 - accuracy: 0.9829 - val_loss: 1.2519 - val_accuracy: 0.7900\n",
      "Epoch 110/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.2445 - accuracy: 0.9849 - val_loss: 1.2432 - val_accuracy: 0.7925\n",
      "Epoch 111/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.2432 - accuracy: 0.9851 - val_loss: 1.2607 - val_accuracy: 0.7900\n",
      "Epoch 112/1000\n",
      "19/19 [==============================] - 1s 46ms/step - loss: 0.2393 - accuracy: 0.9862 - val_loss: 1.2573 - val_accuracy: 0.7900\n",
      "Epoch 113/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2416 - accuracy: 0.9857 - val_loss: 1.2276 - val_accuracy: 0.7925\n",
      "Epoch 114/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.2395 - accuracy: 0.9855 - val_loss: 1.2396 - val_accuracy: 0.8000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 115/1000\n",
      "19/19 [==============================] - 1s 31ms/step - loss: 0.2404 - accuracy: 0.9860 - val_loss: 1.2619 - val_accuracy: 0.7925\n",
      "Epoch 116/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.2441 - accuracy: 0.9849 - val_loss: 1.2455 - val_accuracy: 0.7937\n",
      "Epoch 117/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.2396 - accuracy: 0.9858 - val_loss: 1.2249 - val_accuracy: 0.7875\n",
      "Epoch 118/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.2398 - accuracy: 0.9859 - val_loss: 1.2453 - val_accuracy: 0.7950\n",
      "Epoch 119/1000\n",
      "19/19 [==============================] - 1s 49ms/step - loss: 0.2378 - accuracy: 0.9859 - val_loss: 1.2408 - val_accuracy: 0.7875\n",
      "Epoch 120/1000\n",
      "19/19 [==============================] - 1s 46ms/step - loss: 0.2346 - accuracy: 0.9859 - val_loss: 1.2620 - val_accuracy: 0.7962\n",
      "Epoch 121/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.2364 - accuracy: 0.9861 - val_loss: 1.2546 - val_accuracy: 0.7962\n",
      "Epoch 122/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2363 - accuracy: 0.9868 - val_loss: 1.2674 - val_accuracy: 0.7925\n",
      "Epoch 123/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2395 - accuracy: 0.9847 - val_loss: 1.2837 - val_accuracy: 0.7975\n",
      "Epoch 124/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2410 - accuracy: 0.9855 - val_loss: 1.2812 - val_accuracy: 0.7975\n",
      "Epoch 125/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.2394 - accuracy: 0.9852 - val_loss: 1.2620 - val_accuracy: 0.8012\n",
      "Epoch 126/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2347 - accuracy: 0.9865 - val_loss: 1.2438 - val_accuracy: 0.7975\n",
      "Epoch 127/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2391 - accuracy: 0.9848 - val_loss: 1.3026 - val_accuracy: 0.7862\n",
      "Epoch 128/1000\n",
      "19/19 [==============================] - 1s 47ms/step - loss: 0.2336 - accuracy: 0.9872 - val_loss: 1.2915 - val_accuracy: 0.7962\n",
      "Epoch 129/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2360 - accuracy: 0.9872 - val_loss: 1.2706 - val_accuracy: 0.7950\n",
      "Epoch 130/1000\n",
      "19/19 [==============================] - 1s 46ms/step - loss: 0.2330 - accuracy: 0.9865 - val_loss: 1.2626 - val_accuracy: 0.7950\n",
      "Epoch 131/1000\n",
      "19/19 [==============================] - 1s 51ms/step - loss: 0.2299 - accuracy: 0.9872 - val_loss: 1.2762 - val_accuracy: 0.7825\n",
      "Epoch 132/1000\n",
      "19/19 [==============================] - 1s 53ms/step - loss: 0.2296 - accuracy: 0.9878 - val_loss: 1.2579 - val_accuracy: 0.7912\n",
      "Epoch 133/1000\n",
      "19/19 [==============================] - 1s 46ms/step - loss: 0.2274 - accuracy: 0.9880 - val_loss: 1.2646 - val_accuracy: 0.7925\n",
      "Epoch 134/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2296 - accuracy: 0.9872 - val_loss: 1.3125 - val_accuracy: 0.7875\n",
      "Epoch 135/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.2296 - accuracy: 0.9871 - val_loss: 1.2858 - val_accuracy: 0.7975\n",
      "Epoch 136/1000\n",
      "19/19 [==============================] - 1s 50ms/step - loss: 0.2236 - accuracy: 0.9887 - val_loss: 1.2863 - val_accuracy: 0.8000\n",
      "Epoch 137/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.2276 - accuracy: 0.9870 - val_loss: 1.2774 - val_accuracy: 0.7950\n",
      "Epoch 138/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.2296 - accuracy: 0.9870 - val_loss: 1.2667 - val_accuracy: 0.7912\n",
      "Epoch 139/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.2338 - accuracy: 0.9856 - val_loss: 1.2528 - val_accuracy: 0.7987\n",
      "Epoch 140/1000\n",
      "19/19 [==============================] - 1s 26ms/step - loss: 0.2269 - accuracy: 0.9880 - val_loss: 1.2994 - val_accuracy: 0.8000\n",
      "Epoch 141/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.2238 - accuracy: 0.9884 - val_loss: 1.2663 - val_accuracy: 0.8025\n",
      "Epoch 142/1000\n",
      "19/19 [==============================] - 1s 46ms/step - loss: 0.2218 - accuracy: 0.9888 - val_loss: 1.2324 - val_accuracy: 0.8075\n",
      "Epoch 143/1000\n",
      "19/19 [==============================] - 1s 50ms/step - loss: 0.2205 - accuracy: 0.9900 - val_loss: 1.2430 - val_accuracy: 0.7975\n",
      "Epoch 144/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.2239 - accuracy: 0.9879 - val_loss: 1.2462 - val_accuracy: 0.7950\n",
      "Epoch 145/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2324 - accuracy: 0.9854 - val_loss: 1.2830 - val_accuracy: 0.7987\n",
      "Epoch 146/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2287 - accuracy: 0.9873 - val_loss: 1.3356 - val_accuracy: 0.8012\n",
      "Epoch 147/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2280 - accuracy: 0.9874 - val_loss: 1.2955 - val_accuracy: 0.8025\n",
      "Epoch 148/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.2233 - accuracy: 0.9877 - val_loss: 1.2983 - val_accuracy: 0.7975\n",
      "Epoch 149/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2246 - accuracy: 0.9879 - val_loss: 1.3610 - val_accuracy: 0.7912\n",
      "Epoch 150/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.2224 - accuracy: 0.9882 - val_loss: 1.3192 - val_accuracy: 0.7975\n",
      "Epoch 151/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2245 - accuracy: 0.9874 - val_loss: 1.2851 - val_accuracy: 0.7925\n",
      "Epoch 152/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.2217 - accuracy: 0.9884 - val_loss: 1.2872 - val_accuracy: 0.8000\n",
      "Epoch 153/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.2242 - accuracy: 0.9870 - val_loss: 1.3228 - val_accuracy: 0.8037\n",
      "Epoch 154/1000\n",
      "19/19 [==============================] - 1s 46ms/step - loss: 0.2191 - accuracy: 0.9888 - val_loss: 1.3241 - val_accuracy: 0.7987\n",
      "Epoch 155/1000\n",
      "19/19 [==============================] - 1s 26ms/step - loss: 0.2195 - accuracy: 0.9893 - val_loss: 1.3579 - val_accuracy: 0.7987\n",
      "Epoch 156/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2231 - accuracy: 0.9884 - val_loss: 1.3542 - val_accuracy: 0.7950\n",
      "Epoch 157/1000\n",
      "19/19 [==============================] - 1s 49ms/step - loss: 0.2159 - accuracy: 0.9904 - val_loss: 1.3581 - val_accuracy: 0.7900\n",
      "Epoch 158/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.2240 - accuracy: 0.9877 - val_loss: 1.3656 - val_accuracy: 0.7925\n",
      "Epoch 159/1000\n",
      "19/19 [==============================] - 1s 46ms/step - loss: 0.2128 - accuracy: 0.9917 - val_loss: 1.3400 - val_accuracy: 0.7887\n",
      "Epoch 160/1000\n",
      "19/19 [==============================] - 1s 26ms/step - loss: 0.2213 - accuracy: 0.9876 - val_loss: 1.3325 - val_accuracy: 0.7900\n",
      "Epoch 161/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2199 - accuracy: 0.9875 - val_loss: 1.3272 - val_accuracy: 0.7987\n",
      "Epoch 162/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.2242 - accuracy: 0.9861 - val_loss: 1.3146 - val_accuracy: 0.8000\n",
      "Epoch 163/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.2219 - accuracy: 0.9874 - val_loss: 1.3726 - val_accuracy: 0.7937\n",
      "Epoch 164/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.2205 - accuracy: 0.9870 - val_loss: 1.3055 - val_accuracy: 0.8037\n",
      "Epoch 165/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2223 - accuracy: 0.9882 - val_loss: 1.3274 - val_accuracy: 0.8012\n",
      "Epoch 166/1000\n",
      "19/19 [==============================] - 1s 32ms/step - loss: 0.2221 - accuracy: 0.9869 - val_loss: 1.3091 - val_accuracy: 0.8012\n",
      "Epoch 167/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2178 - accuracy: 0.9882 - val_loss: 1.3321 - val_accuracy: 0.7975\n",
      "Epoch 168/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.2191 - accuracy: 0.9886 - val_loss: 1.3563 - val_accuracy: 0.7962\n",
      "Epoch 169/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2201 - accuracy: 0.9861 - val_loss: 1.3486 - val_accuracy: 0.7912\n",
      "Epoch 170/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.2179 - accuracy: 0.9884 - val_loss: 1.3020 - val_accuracy: 0.8012\n",
      "Epoch 171/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2186 - accuracy: 0.9882 - val_loss: 1.2957 - val_accuracy: 0.8025\n",
      "Epoch 172/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.2144 - accuracy: 0.9891 - val_loss: 1.2862 - val_accuracy: 0.7975\n",
      "Epoch 173/1000\n",
      "19/19 [==============================] - 1s 25ms/step - loss: 0.2198 - accuracy: 0.9881 - val_loss: 1.3067 - val_accuracy: 0.8025\n",
      "Epoch 174/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2188 - accuracy: 0.9874 - val_loss: 1.3967 - val_accuracy: 0.7837\n",
      "Epoch 175/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.2137 - accuracy: 0.9884 - val_loss: 1.3141 - val_accuracy: 0.8000\n",
      "Epoch 176/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.2158 - accuracy: 0.9888 - val_loss: 1.3059 - val_accuracy: 0.7925\n",
      "Epoch 177/1000\n",
      "19/19 [==============================] - 1s 33ms/step - loss: 0.2179 - accuracy: 0.9875 - val_loss: 1.3317 - val_accuracy: 0.7887\n",
      "Epoch 178/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2154 - accuracy: 0.9886 - val_loss: 1.3324 - val_accuracy: 0.7912\n",
      "Epoch 179/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.2106 - accuracy: 0.9903 - val_loss: 1.3127 - val_accuracy: 0.8000\n",
      "Epoch 180/1000\n",
      "19/19 [==============================] - 1s 26ms/step - loss: 0.2180 - accuracy: 0.9884 - val_loss: 1.3184 - val_accuracy: 0.7925\n",
      "Epoch 181/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.2133 - accuracy: 0.9879 - val_loss: 1.2941 - val_accuracy: 0.8012\n",
      "Epoch 182/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.2144 - accuracy: 0.9889 - val_loss: 1.3226 - val_accuracy: 0.8062\n",
      "Epoch 183/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2131 - accuracy: 0.9889 - val_loss: 1.3583 - val_accuracy: 0.8062\n",
      "Epoch 184/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.2121 - accuracy: 0.9904 - val_loss: 1.3377 - val_accuracy: 0.8012\n",
      "Epoch 185/1000\n",
      "19/19 [==============================] - 1s 32ms/step - loss: 0.2156 - accuracy: 0.9886 - val_loss: 1.3098 - val_accuracy: 0.7900\n",
      "Epoch 186/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.2159 - accuracy: 0.9882 - val_loss: 1.3074 - val_accuracy: 0.8087\n",
      "Epoch 187/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2140 - accuracy: 0.9886 - val_loss: 1.3439 - val_accuracy: 0.8012\n",
      "Epoch 188/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.2121 - accuracy: 0.9891 - val_loss: 1.3299 - val_accuracy: 0.7950\n",
      "Epoch 189/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.2123 - accuracy: 0.9886 - val_loss: 1.3481 - val_accuracy: 0.8000\n",
      "Epoch 190/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.2123 - accuracy: 0.9884 - val_loss: 1.3312 - val_accuracy: 0.8050\n",
      "Epoch 191/1000\n",
      "19/19 [==============================] - 1s 52ms/step - loss: 0.2056 - accuracy: 0.9912 - val_loss: 1.3205 - val_accuracy: 0.8050\n",
      "Epoch 192/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2097 - accuracy: 0.9895 - val_loss: 1.3191 - val_accuracy: 0.7912\n",
      "Epoch 193/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2083 - accuracy: 0.9895 - val_loss: 1.3060 - val_accuracy: 0.7962\n",
      "Epoch 194/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2074 - accuracy: 0.9893 - val_loss: 1.2938 - val_accuracy: 0.8025\n",
      "Epoch 195/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.2072 - accuracy: 0.9893 - val_loss: 1.3212 - val_accuracy: 0.8050\n",
      "Epoch 196/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2083 - accuracy: 0.9900 - val_loss: 1.3098 - val_accuracy: 0.7975\n",
      "Epoch 197/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2155 - accuracy: 0.9874 - val_loss: 1.3555 - val_accuracy: 0.7950\n",
      "Epoch 198/1000\n",
      "19/19 [==============================] - 1s 32ms/step - loss: 0.2080 - accuracy: 0.9894 - val_loss: 1.3311 - val_accuracy: 0.7937\n",
      "Epoch 199/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2058 - accuracy: 0.9903 - val_loss: 1.3674 - val_accuracy: 0.8050\n",
      "Epoch 200/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2103 - accuracy: 0.9882 - val_loss: 1.3348 - val_accuracy: 0.7987\n",
      "Epoch 201/1000\n",
      "19/19 [==============================] - 1s 47ms/step - loss: 0.2035 - accuracy: 0.9905 - val_loss: 1.3486 - val_accuracy: 0.8037\n",
      "Epoch 202/1000\n",
      "19/19 [==============================] - 1s 25ms/step - loss: 0.2050 - accuracy: 0.9896 - val_loss: 1.3859 - val_accuracy: 0.7937\n",
      "Epoch 203/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.2059 - accuracy: 0.9897 - val_loss: 1.3282 - val_accuracy: 0.8050\n",
      "Epoch 204/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2077 - accuracy: 0.9902 - val_loss: 1.3426 - val_accuracy: 0.7950\n",
      "Epoch 205/1000\n",
      "19/19 [==============================] - 1s 47ms/step - loss: 0.2017 - accuracy: 0.9905 - val_loss: 1.3846 - val_accuracy: 0.8012\n",
      "Epoch 206/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.2075 - accuracy: 0.9892 - val_loss: 1.3656 - val_accuracy: 0.7925\n",
      "Epoch 207/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2048 - accuracy: 0.9901 - val_loss: 1.3903 - val_accuracy: 0.7925\n",
      "Epoch 208/1000\n",
      "19/19 [==============================] - 1s 50ms/step - loss: 0.1978 - accuracy: 0.9920 - val_loss: 1.3858 - val_accuracy: 0.7962\n",
      "Epoch 209/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2055 - accuracy: 0.9889 - val_loss: 1.3532 - val_accuracy: 0.8025\n",
      "Epoch 210/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2038 - accuracy: 0.9905 - val_loss: 1.3997 - val_accuracy: 0.7912\n",
      "Epoch 211/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.2009 - accuracy: 0.9903 - val_loss: 1.3628 - val_accuracy: 0.7975\n",
      "Epoch 212/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.2035 - accuracy: 0.9907 - val_loss: 1.3691 - val_accuracy: 0.8075\n",
      "Epoch 213/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1993 - accuracy: 0.9916 - val_loss: 1.3331 - val_accuracy: 0.7925\n",
      "Epoch 214/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1999 - accuracy: 0.9907 - val_loss: 1.3371 - val_accuracy: 0.8025\n",
      "Epoch 215/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2025 - accuracy: 0.9902 - val_loss: 1.3629 - val_accuracy: 0.7962\n",
      "Epoch 216/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2007 - accuracy: 0.9903 - val_loss: 1.3917 - val_accuracy: 0.7912\n",
      "Epoch 217/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2016 - accuracy: 0.9901 - val_loss: 1.3684 - val_accuracy: 0.7788\n",
      "Epoch 218/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.2004 - accuracy: 0.9896 - val_loss: 1.3729 - val_accuracy: 0.7950\n",
      "Epoch 219/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.2002 - accuracy: 0.9907 - val_loss: 1.4036 - val_accuracy: 0.8000\n",
      "Epoch 220/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.2032 - accuracy: 0.9891 - val_loss: 1.3590 - val_accuracy: 0.8012\n",
      "Epoch 221/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.2022 - accuracy: 0.9896 - val_loss: 1.3373 - val_accuracy: 0.8025\n",
      "Epoch 222/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.2000 - accuracy: 0.9901 - val_loss: 1.3332 - val_accuracy: 0.8000\n",
      "Epoch 223/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.2026 - accuracy: 0.9900 - val_loss: 1.3798 - val_accuracy: 0.8000\n",
      "Epoch 224/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.2002 - accuracy: 0.9907 - val_loss: 1.3906 - val_accuracy: 0.7900\n",
      "Epoch 225/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1995 - accuracy: 0.9900 - val_loss: 1.3448 - val_accuracy: 0.7962\n",
      "Epoch 226/1000\n",
      "19/19 [==============================] - 1s 47ms/step - loss: 0.1964 - accuracy: 0.9909 - val_loss: 1.3585 - val_accuracy: 0.7962\n",
      "Epoch 227/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "19/19 [==============================] - 1s 31ms/step - loss: 0.1970 - accuracy: 0.9911 - val_loss: 1.4003 - val_accuracy: 0.7912\n",
      "Epoch 228/1000\n",
      "19/19 [==============================] - 1s 52ms/step - loss: 0.1940 - accuracy: 0.9923 - val_loss: 1.3259 - val_accuracy: 0.7937\n",
      "Epoch 229/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1966 - accuracy: 0.9920 - val_loss: 1.3600 - val_accuracy: 0.7912\n",
      "Epoch 230/1000\n",
      "19/19 [==============================] - 1s 31ms/step - loss: 0.1947 - accuracy: 0.9918 - val_loss: 1.3246 - val_accuracy: 0.7975\n",
      "Epoch 231/1000\n",
      "19/19 [==============================] - 1s 25ms/step - loss: 0.1969 - accuracy: 0.9909 - val_loss: 1.2996 - val_accuracy: 0.8000\n",
      "Epoch 232/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1978 - accuracy: 0.9901 - val_loss: 1.3246 - val_accuracy: 0.8025\n",
      "Epoch 233/1000\n",
      "19/19 [==============================] - 1s 47ms/step - loss: 0.1937 - accuracy: 0.9912 - val_loss: 1.3698 - val_accuracy: 0.8025\n",
      "Epoch 234/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1983 - accuracy: 0.9901 - val_loss: 1.3766 - val_accuracy: 0.8025\n",
      "Epoch 235/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1979 - accuracy: 0.9893 - val_loss: 1.3648 - val_accuracy: 0.8012\n",
      "Epoch 236/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1971 - accuracy: 0.9901 - val_loss: 1.4247 - val_accuracy: 0.7975\n",
      "Epoch 237/1000\n",
      "19/19 [==============================] - 1s 33ms/step - loss: 0.1945 - accuracy: 0.9908 - val_loss: 1.3670 - val_accuracy: 0.8012\n",
      "Epoch 238/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.1912 - accuracy: 0.9922 - val_loss: 1.3641 - val_accuracy: 0.7975\n",
      "Epoch 239/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1961 - accuracy: 0.9908 - val_loss: 1.3499 - val_accuracy: 0.8062\n",
      "Epoch 240/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1984 - accuracy: 0.9895 - val_loss: 1.3808 - val_accuracy: 0.7987\n",
      "Epoch 241/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1959 - accuracy: 0.9900 - val_loss: 1.3706 - val_accuracy: 0.7925\n",
      "Epoch 242/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1923 - accuracy: 0.9912 - val_loss: 1.3669 - val_accuracy: 0.7962\n",
      "Epoch 243/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1941 - accuracy: 0.9911 - val_loss: 1.3482 - val_accuracy: 0.8037\n",
      "Epoch 244/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1990 - accuracy: 0.9895 - val_loss: 1.4039 - val_accuracy: 0.7962\n",
      "Epoch 245/1000\n",
      "19/19 [==============================] - 1s 25ms/step - loss: 0.1932 - accuracy: 0.9914 - val_loss: 1.3819 - val_accuracy: 0.7887\n",
      "Epoch 246/1000\n",
      "19/19 [==============================] - 1s 50ms/step - loss: 0.1892 - accuracy: 0.9920 - val_loss: 1.3598 - val_accuracy: 0.8062\n",
      "Epoch 247/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1901 - accuracy: 0.9920 - val_loss: 1.3454 - val_accuracy: 0.7962\n",
      "Epoch 248/1000\n",
      "19/19 [==============================] - 1s 47ms/step - loss: 0.1890 - accuracy: 0.9926 - val_loss: 1.3492 - val_accuracy: 0.7937\n",
      "Epoch 249/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.1884 - accuracy: 0.9914 - val_loss: 1.3185 - val_accuracy: 0.8037\n",
      "Epoch 250/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1918 - accuracy: 0.9905 - val_loss: 1.3426 - val_accuracy: 0.8037\n",
      "Epoch 251/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1920 - accuracy: 0.9903 - val_loss: 1.3177 - val_accuracy: 0.8037\n",
      "Epoch 252/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1916 - accuracy: 0.9911 - val_loss: 1.3534 - val_accuracy: 0.8000\n",
      "Epoch 253/1000\n",
      "19/19 [==============================] - 1s 25ms/step - loss: 0.1917 - accuracy: 0.9907 - val_loss: 1.3296 - val_accuracy: 0.8012\n",
      "Epoch 254/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1923 - accuracy: 0.9901 - val_loss: 1.3673 - val_accuracy: 0.8000\n",
      "Epoch 255/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1886 - accuracy: 0.9920 - val_loss: 1.4128 - val_accuracy: 0.7925\n",
      "Epoch 256/1000\n",
      "19/19 [==============================] - 1s 51ms/step - loss: 0.1878 - accuracy: 0.9923 - val_loss: 1.3883 - val_accuracy: 0.7950\n",
      "Epoch 257/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1907 - accuracy: 0.9920 - val_loss: 1.4183 - val_accuracy: 0.7962\n",
      "Epoch 258/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1905 - accuracy: 0.9916 - val_loss: 1.4047 - val_accuracy: 0.8000\n",
      "Epoch 259/1000\n",
      "19/19 [==============================] - 1s 50ms/step - loss: 0.1861 - accuracy: 0.9922 - val_loss: 1.3626 - val_accuracy: 0.8050\n",
      "Epoch 260/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1872 - accuracy: 0.9924 - val_loss: 1.3383 - val_accuracy: 0.8050\n",
      "Epoch 261/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1893 - accuracy: 0.9914 - val_loss: 1.4176 - val_accuracy: 0.8012\n",
      "Epoch 262/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1870 - accuracy: 0.9918 - val_loss: 1.4075 - val_accuracy: 0.7925\n",
      "Epoch 263/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1864 - accuracy: 0.9928 - val_loss: 1.4234 - val_accuracy: 0.8000\n",
      "Epoch 264/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1887 - accuracy: 0.9909 - val_loss: 1.4466 - val_accuracy: 0.7987\n",
      "Epoch 265/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1870 - accuracy: 0.9921 - val_loss: 1.4114 - val_accuracy: 0.8012\n",
      "Epoch 266/1000\n",
      "19/19 [==============================] - 1s 53ms/step - loss: 0.1859 - accuracy: 0.9909 - val_loss: 1.3871 - val_accuracy: 0.8050\n",
      "Epoch 267/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1869 - accuracy: 0.9911 - val_loss: 1.4128 - val_accuracy: 0.8025\n",
      "Epoch 268/1000\n",
      "19/19 [==============================] - 1s 46ms/step - loss: 0.1858 - accuracy: 0.9915 - val_loss: 1.3757 - val_accuracy: 0.8000\n",
      "Epoch 269/1000\n",
      "19/19 [==============================] - 1s 46ms/step - loss: 0.1827 - accuracy: 0.9917 - val_loss: 1.3874 - val_accuracy: 0.8012\n",
      "Epoch 270/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1836 - accuracy: 0.9929 - val_loss: 1.3959 - val_accuracy: 0.8012\n",
      "Epoch 271/1000\n",
      "19/19 [==============================] - 1s 26ms/step - loss: 0.1905 - accuracy: 0.9901 - val_loss: 1.3818 - val_accuracy: 0.8050\n",
      "Epoch 272/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1874 - accuracy: 0.9909 - val_loss: 1.3547 - val_accuracy: 0.8037\n",
      "Epoch 273/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1861 - accuracy: 0.9920 - val_loss: 1.3521 - val_accuracy: 0.8012\n",
      "Epoch 274/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1870 - accuracy: 0.9907 - val_loss: 1.3589 - val_accuracy: 0.8062\n",
      "Epoch 275/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1864 - accuracy: 0.9918 - val_loss: 1.3914 - val_accuracy: 0.8000\n",
      "Epoch 276/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1872 - accuracy: 0.9905 - val_loss: 1.3493 - val_accuracy: 0.7962\n",
      "Epoch 277/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1897 - accuracy: 0.9904 - val_loss: 1.3935 - val_accuracy: 0.7925\n",
      "Epoch 278/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1887 - accuracy: 0.9901 - val_loss: 1.3609 - val_accuracy: 0.7875\n",
      "Epoch 279/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1851 - accuracy: 0.9911 - val_loss: 1.3529 - val_accuracy: 0.7900\n",
      "Epoch 280/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1893 - accuracy: 0.9901 - val_loss: 1.3955 - val_accuracy: 0.7962\n",
      "Epoch 281/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1857 - accuracy: 0.9915 - val_loss: 1.4217 - val_accuracy: 0.7875\n",
      "Epoch 282/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1866 - accuracy: 0.9905 - val_loss: 1.3918 - val_accuracy: 0.7962\n",
      "Epoch 283/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "19/19 [==============================] - 1s 49ms/step - loss: 0.1824 - accuracy: 0.9921 - val_loss: 1.4396 - val_accuracy: 0.8025\n",
      "Epoch 284/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1848 - accuracy: 0.9920 - val_loss: 1.4059 - val_accuracy: 0.7962\n",
      "Epoch 285/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1834 - accuracy: 0.9924 - val_loss: 1.3571 - val_accuracy: 0.8062\n",
      "Epoch 286/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1836 - accuracy: 0.9914 - val_loss: 1.3566 - val_accuracy: 0.8012\n",
      "Epoch 287/1000\n",
      "19/19 [==============================] - 1s 51ms/step - loss: 0.1811 - accuracy: 0.9925 - val_loss: 1.3814 - val_accuracy: 0.8112\n",
      "Epoch 288/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1918 - accuracy: 0.9887 - val_loss: 1.3536 - val_accuracy: 0.8012\n",
      "Epoch 289/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1820 - accuracy: 0.9924 - val_loss: 1.3608 - val_accuracy: 0.7975\n",
      "Epoch 290/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1838 - accuracy: 0.9912 - val_loss: 1.3223 - val_accuracy: 0.8075\n",
      "Epoch 291/1000\n",
      "19/19 [==============================] - 1s 33ms/step - loss: 0.1811 - accuracy: 0.9925 - val_loss: 1.3575 - val_accuracy: 0.8000\n",
      "Epoch 292/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1833 - accuracy: 0.9920 - val_loss: 1.4050 - val_accuracy: 0.7862\n",
      "Epoch 293/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1850 - accuracy: 0.9908 - val_loss: 1.3853 - val_accuracy: 0.7937\n",
      "Epoch 294/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1838 - accuracy: 0.9911 - val_loss: 1.3685 - val_accuracy: 0.8025\n",
      "Epoch 295/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1822 - accuracy: 0.9907 - val_loss: 1.4334 - val_accuracy: 0.7937\n",
      "Epoch 296/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1831 - accuracy: 0.9916 - val_loss: 1.4038 - val_accuracy: 0.8037\n",
      "Epoch 297/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1828 - accuracy: 0.9923 - val_loss: 1.3616 - val_accuracy: 0.8087\n",
      "Epoch 298/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1828 - accuracy: 0.9912 - val_loss: 1.3950 - val_accuracy: 0.7912\n",
      "Epoch 299/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1845 - accuracy: 0.9908 - val_loss: 1.3250 - val_accuracy: 0.8025\n",
      "Epoch 300/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1851 - accuracy: 0.9905 - val_loss: 1.3931 - val_accuracy: 0.8000\n",
      "Epoch 301/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1860 - accuracy: 0.9904 - val_loss: 1.4049 - val_accuracy: 0.7912\n",
      "Epoch 302/1000\n",
      "19/19 [==============================] - 1s 52ms/step - loss: 0.1798 - accuracy: 0.9925 - val_loss: 1.3853 - val_accuracy: 0.7875\n",
      "Epoch 303/1000\n",
      "19/19 [==============================] - 1s 47ms/step - loss: 0.1785 - accuracy: 0.9930 - val_loss: 1.3649 - val_accuracy: 0.8025\n",
      "Epoch 304/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1815 - accuracy: 0.9918 - val_loss: 1.3540 - val_accuracy: 0.7987\n",
      "Epoch 305/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1847 - accuracy: 0.9910 - val_loss: 1.3755 - val_accuracy: 0.7900\n",
      "Epoch 306/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1805 - accuracy: 0.9920 - val_loss: 1.4306 - val_accuracy: 0.7987\n",
      "Epoch 307/1000\n",
      "19/19 [==============================] - 1s 25ms/step - loss: 0.1797 - accuracy: 0.9912 - val_loss: 1.3618 - val_accuracy: 0.8025\n",
      "Epoch 308/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1792 - accuracy: 0.9925 - val_loss: 1.4295 - val_accuracy: 0.7937\n",
      "Epoch 309/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1807 - accuracy: 0.9909 - val_loss: 1.3912 - val_accuracy: 0.7975\n",
      "Epoch 310/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1822 - accuracy: 0.9914 - val_loss: 1.3404 - val_accuracy: 0.8050\n",
      "Epoch 311/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1792 - accuracy: 0.9914 - val_loss: 1.3340 - val_accuracy: 0.7975\n",
      "Epoch 312/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.1770 - accuracy: 0.9922 - val_loss: 1.3931 - val_accuracy: 0.7962\n",
      "Epoch 313/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1798 - accuracy: 0.9920 - val_loss: 1.4341 - val_accuracy: 0.7937\n",
      "Epoch 314/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1786 - accuracy: 0.9924 - val_loss: 1.4138 - val_accuracy: 0.7950\n",
      "Epoch 315/1000\n",
      "19/19 [==============================] - 0s 24ms/step - loss: 0.1787 - accuracy: 0.9921 - val_loss: 1.4093 - val_accuracy: 0.7862\n",
      "Epoch 316/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1798 - accuracy: 0.9925 - val_loss: 1.3914 - val_accuracy: 0.7925\n",
      "Epoch 317/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1793 - accuracy: 0.9918 - val_loss: 1.3623 - val_accuracy: 0.7900\n",
      "Epoch 318/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1795 - accuracy: 0.9913 - val_loss: 1.3998 - val_accuracy: 0.7987\n",
      "Epoch 319/1000\n",
      "19/19 [==============================] - 1s 46ms/step - loss: 0.1765 - accuracy: 0.9932 - val_loss: 1.4002 - val_accuracy: 0.7962\n",
      "Epoch 320/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1777 - accuracy: 0.9919 - val_loss: 1.3890 - val_accuracy: 0.8012\n",
      "Epoch 321/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1775 - accuracy: 0.9926 - val_loss: 1.3597 - val_accuracy: 0.7925\n",
      "Epoch 322/1000\n",
      "19/19 [==============================] - 0s 24ms/step - loss: 0.1812 - accuracy: 0.9908 - val_loss: 1.4029 - val_accuracy: 0.7925\n",
      "Epoch 323/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1828 - accuracy: 0.9905 - val_loss: 1.3469 - val_accuracy: 0.8037\n",
      "Epoch 324/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1798 - accuracy: 0.9916 - val_loss: 1.3641 - val_accuracy: 0.8075\n",
      "Epoch 325/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1800 - accuracy: 0.9911 - val_loss: 1.3530 - val_accuracy: 0.8000\n",
      "Epoch 326/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1801 - accuracy: 0.9910 - val_loss: 1.3359 - val_accuracy: 0.8025\n",
      "Epoch 327/1000\n",
      "19/19 [==============================] - 0s 24ms/step - loss: 0.1798 - accuracy: 0.9916 - val_loss: 1.3861 - val_accuracy: 0.7925\n",
      "Epoch 328/1000\n",
      "19/19 [==============================] - 0s 24ms/step - loss: 0.1768 - accuracy: 0.9921 - val_loss: 1.3649 - val_accuracy: 0.7987\n",
      "Epoch 329/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1766 - accuracy: 0.9916 - val_loss: 1.4015 - val_accuracy: 0.7887\n",
      "Epoch 330/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1785 - accuracy: 0.9916 - val_loss: 1.4010 - val_accuracy: 0.8100\n",
      "Epoch 331/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1816 - accuracy: 0.9913 - val_loss: 1.4123 - val_accuracy: 0.7975\n",
      "Epoch 332/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1768 - accuracy: 0.9916 - val_loss: 1.4071 - val_accuracy: 0.8025\n",
      "Epoch 333/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.1763 - accuracy: 0.9924 - val_loss: 1.3487 - val_accuracy: 0.8025\n",
      "Epoch 334/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1767 - accuracy: 0.9921 - val_loss: 1.3691 - val_accuracy: 0.8012\n",
      "Epoch 335/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1831 - accuracy: 0.9908 - val_loss: 1.3874 - val_accuracy: 0.7900\n",
      "Epoch 336/1000\n",
      "19/19 [==============================] - 0s 24ms/step - loss: 0.1788 - accuracy: 0.9912 - val_loss: 1.3717 - val_accuracy: 0.7950\n",
      "Epoch 337/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1790 - accuracy: 0.9905 - val_loss: 1.3443 - val_accuracy: 0.7887\n",
      "Epoch 338/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1799 - accuracy: 0.9910 - val_loss: 1.3563 - val_accuracy: 0.7950\n",
      "Epoch 339/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "19/19 [==============================] - 1s 48ms/step - loss: 0.1731 - accuracy: 0.9933 - val_loss: 1.3027 - val_accuracy: 0.7900\n",
      "Epoch 340/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1789 - accuracy: 0.9913 - val_loss: 1.3242 - val_accuracy: 0.8012\n",
      "Epoch 341/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1821 - accuracy: 0.9903 - val_loss: 1.3767 - val_accuracy: 0.7862\n",
      "Epoch 342/1000\n",
      "19/19 [==============================] - 0s 24ms/step - loss: 0.1767 - accuracy: 0.9914 - val_loss: 1.3620 - val_accuracy: 0.7950\n",
      "Epoch 343/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1754 - accuracy: 0.9925 - val_loss: 1.3541 - val_accuracy: 0.8087\n",
      "Epoch 344/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1777 - accuracy: 0.9916 - val_loss: 1.3636 - val_accuracy: 0.8012\n",
      "Epoch 345/1000\n",
      "19/19 [==============================] - 1s 24ms/step - loss: 0.1766 - accuracy: 0.9922 - val_loss: 1.3397 - val_accuracy: 0.8000\n",
      "Epoch 346/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1753 - accuracy: 0.9922 - val_loss: 1.3614 - val_accuracy: 0.7837\n",
      "Epoch 347/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1760 - accuracy: 0.9922 - val_loss: 1.3550 - val_accuracy: 0.8012\n",
      "Epoch 348/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1762 - accuracy: 0.9921 - val_loss: 1.3202 - val_accuracy: 0.7962\n",
      "Epoch 349/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1744 - accuracy: 0.9926 - val_loss: 1.3555 - val_accuracy: 0.8025\n",
      "Epoch 350/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1740 - accuracy: 0.9920 - val_loss: 1.3704 - val_accuracy: 0.7937\n",
      "Epoch 351/1000\n",
      "19/19 [==============================] - 1s 47ms/step - loss: 0.1711 - accuracy: 0.9928 - val_loss: 1.3453 - val_accuracy: 0.7937\n",
      "Epoch 352/1000\n",
      "19/19 [==============================] - 1s 25ms/step - loss: 0.1723 - accuracy: 0.9932 - val_loss: 1.3426 - val_accuracy: 0.7950\n",
      "Epoch 353/1000\n",
      "19/19 [==============================] - 1s 53ms/step - loss: 0.1709 - accuracy: 0.9930 - val_loss: 1.3535 - val_accuracy: 0.8012\n",
      "Epoch 354/1000\n",
      "19/19 [==============================] - 1s 26ms/step - loss: 0.1720 - accuracy: 0.9925 - val_loss: 1.3729 - val_accuracy: 0.7987\n",
      "Epoch 355/1000\n",
      "19/19 [==============================] - 1s 47ms/step - loss: 0.1688 - accuracy: 0.9930 - val_loss: 1.3885 - val_accuracy: 0.7975\n",
      "Epoch 356/1000\n",
      "19/19 [==============================] - 1s 33ms/step - loss: 0.1733 - accuracy: 0.9924 - val_loss: 1.3432 - val_accuracy: 0.8050\n",
      "Epoch 357/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1691 - accuracy: 0.9935 - val_loss: 1.3688 - val_accuracy: 0.8075\n",
      "Epoch 358/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1708 - accuracy: 0.9924 - val_loss: 1.3573 - val_accuracy: 0.8050\n",
      "Epoch 359/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1697 - accuracy: 0.9934 - val_loss: 1.3504 - val_accuracy: 0.8037\n",
      "Epoch 360/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1693 - accuracy: 0.9926 - val_loss: 1.3517 - val_accuracy: 0.8012\n",
      "Epoch 361/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.1661 - accuracy: 0.9947 - val_loss: 1.3452 - val_accuracy: 0.8037\n",
      "Epoch 362/1000\n",
      "19/19 [==============================] - 1s 26ms/step - loss: 0.1707 - accuracy: 0.9927 - val_loss: 1.3685 - val_accuracy: 0.7975\n",
      "Epoch 363/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1707 - accuracy: 0.9920 - val_loss: 1.3713 - val_accuracy: 0.7925\n",
      "Epoch 364/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1690 - accuracy: 0.9931 - val_loss: 1.3357 - val_accuracy: 0.7912\n",
      "Epoch 365/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1738 - accuracy: 0.9919 - val_loss: 1.3831 - val_accuracy: 0.7937\n",
      "Epoch 366/1000\n",
      "19/19 [==============================] - 0s 24ms/step - loss: 0.1731 - accuracy: 0.9921 - val_loss: 1.3824 - val_accuracy: 0.7950\n",
      "Epoch 367/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1688 - accuracy: 0.9926 - val_loss: 1.3646 - val_accuracy: 0.7975\n",
      "Epoch 368/1000\n",
      "19/19 [==============================] - 1s 25ms/step - loss: 0.1695 - accuracy: 0.9924 - val_loss: 1.3787 - val_accuracy: 0.7975\n",
      "Epoch 369/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1714 - accuracy: 0.9920 - val_loss: 1.3817 - val_accuracy: 0.8000\n",
      "Epoch 370/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1672 - accuracy: 0.9932 - val_loss: 1.3544 - val_accuracy: 0.7987\n",
      "Epoch 371/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1708 - accuracy: 0.9928 - val_loss: 1.3347 - val_accuracy: 0.7987\n",
      "Epoch 372/1000\n",
      "19/19 [==============================] - 0s 24ms/step - loss: 0.1675 - accuracy: 0.9938 - val_loss: 1.3246 - val_accuracy: 0.7975\n",
      "Epoch 373/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1674 - accuracy: 0.9927 - val_loss: 1.3761 - val_accuracy: 0.7975\n",
      "Epoch 374/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1690 - accuracy: 0.9922 - val_loss: 1.4379 - val_accuracy: 0.7925\n",
      "Epoch 375/1000\n",
      "19/19 [==============================] - 1s 49ms/step - loss: 0.1650 - accuracy: 0.9940 - val_loss: 1.4087 - val_accuracy: 0.8025\n",
      "Epoch 376/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1718 - accuracy: 0.9917 - val_loss: 1.3972 - val_accuracy: 0.7900\n",
      "Epoch 377/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1679 - accuracy: 0.9921 - val_loss: 1.4416 - val_accuracy: 0.7862\n",
      "Epoch 378/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1699 - accuracy: 0.9926 - val_loss: 1.4170 - val_accuracy: 0.7925\n",
      "Epoch 379/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1701 - accuracy: 0.9931 - val_loss: 1.4510 - val_accuracy: 0.7837\n",
      "Epoch 380/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1651 - accuracy: 0.9946 - val_loss: 1.4093 - val_accuracy: 0.7875\n",
      "Epoch 381/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1676 - accuracy: 0.9928 - val_loss: 1.4066 - val_accuracy: 0.7912\n",
      "Epoch 382/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1737 - accuracy: 0.9928 - val_loss: 1.4226 - val_accuracy: 0.7962\n",
      "Epoch 383/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1712 - accuracy: 0.9917 - val_loss: 1.3953 - val_accuracy: 0.7925\n",
      "Epoch 384/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1693 - accuracy: 0.9920 - val_loss: 1.4193 - val_accuracy: 0.7912\n",
      "Epoch 385/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1705 - accuracy: 0.9921 - val_loss: 1.4140 - val_accuracy: 0.7950\n",
      "Epoch 386/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1687 - accuracy: 0.9920 - val_loss: 1.4210 - val_accuracy: 0.7875\n",
      "Epoch 387/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1684 - accuracy: 0.9924 - val_loss: 1.4086 - val_accuracy: 0.7800\n",
      "Epoch 388/1000\n",
      "19/19 [==============================] - 1s 49ms/step - loss: 0.1647 - accuracy: 0.9937 - val_loss: 1.3799 - val_accuracy: 0.7937\n",
      "Epoch 389/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1676 - accuracy: 0.9930 - val_loss: 1.3676 - val_accuracy: 0.7837\n",
      "Epoch 390/1000\n",
      "19/19 [==============================] - 1s 47ms/step - loss: 0.1633 - accuracy: 0.9933 - val_loss: 1.3751 - val_accuracy: 0.7850\n",
      "Epoch 391/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1663 - accuracy: 0.9924 - val_loss: 1.3732 - val_accuracy: 0.7987\n",
      "Epoch 392/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1660 - accuracy: 0.9930 - val_loss: 1.3660 - val_accuracy: 0.7837\n",
      "Epoch 393/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1637 - accuracy: 0.9943 - val_loss: 1.4061 - val_accuracy: 0.8000\n",
      "Epoch 394/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1648 - accuracy: 0.9939 - val_loss: 1.4033 - val_accuracy: 0.7937\n",
      "Epoch 395/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1672 - accuracy: 0.9934 - val_loss: 1.4022 - val_accuracy: 0.7837\n",
      "Epoch 396/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1653 - accuracy: 0.9930 - val_loss: 1.4017 - val_accuracy: 0.7862\n",
      "Epoch 397/1000\n",
      "19/19 [==============================] - 1s 46ms/step - loss: 0.1623 - accuracy: 0.9941 - val_loss: 1.3832 - val_accuracy: 0.7825\n",
      "Epoch 398/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1667 - accuracy: 0.9918 - val_loss: 1.3714 - val_accuracy: 0.7887\n",
      "Epoch 399/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1671 - accuracy: 0.9930 - val_loss: 1.4522 - val_accuracy: 0.7825\n",
      "Epoch 400/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1654 - accuracy: 0.9926 - val_loss: 1.4097 - val_accuracy: 0.7962\n",
      "Epoch 401/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1656 - accuracy: 0.9929 - val_loss: 1.4267 - val_accuracy: 0.7975\n",
      "Epoch 402/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1685 - accuracy: 0.9930 - val_loss: 1.4017 - val_accuracy: 0.7962\n",
      "Epoch 403/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1669 - accuracy: 0.9918 - val_loss: 1.4282 - val_accuracy: 0.7962\n",
      "Epoch 404/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1645 - accuracy: 0.9923 - val_loss: 1.3895 - val_accuracy: 0.7875\n",
      "Epoch 405/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1646 - accuracy: 0.9929 - val_loss: 1.3698 - val_accuracy: 0.7937\n",
      "Epoch 406/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1642 - accuracy: 0.9928 - val_loss: 1.3921 - val_accuracy: 0.8000\n",
      "Epoch 407/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1637 - accuracy: 0.9928 - val_loss: 1.4085 - val_accuracy: 0.7875\n",
      "Epoch 408/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1664 - accuracy: 0.9929 - val_loss: 1.4131 - val_accuracy: 0.7962\n",
      "Epoch 409/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1658 - accuracy: 0.9922 - val_loss: 1.3677 - val_accuracy: 0.8025\n",
      "Epoch 410/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1644 - accuracy: 0.9930 - val_loss: 1.3719 - val_accuracy: 0.8000\n",
      "Epoch 411/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1701 - accuracy: 0.9908 - val_loss: 1.3947 - val_accuracy: 0.7925\n",
      "Epoch 412/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1661 - accuracy: 0.9916 - val_loss: 1.4495 - val_accuracy: 0.7900\n",
      "Epoch 413/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1664 - accuracy: 0.9914 - val_loss: 1.3934 - val_accuracy: 0.7987\n",
      "Epoch 414/1000\n",
      "19/19 [==============================] - 1s 25ms/step - loss: 0.1673 - accuracy: 0.9926 - val_loss: 1.4164 - val_accuracy: 0.7900\n",
      "Epoch 415/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1687 - accuracy: 0.9926 - val_loss: 1.4309 - val_accuracy: 0.7950\n",
      "Epoch 416/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1661 - accuracy: 0.9911 - val_loss: 1.4081 - val_accuracy: 0.7825\n",
      "Epoch 417/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1658 - accuracy: 0.9927 - val_loss: 1.3894 - val_accuracy: 0.7950\n",
      "Epoch 418/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1632 - accuracy: 0.9932 - val_loss: 1.3661 - val_accuracy: 0.7937\n",
      "Epoch 419/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1656 - accuracy: 0.9925 - val_loss: 1.4160 - val_accuracy: 0.7900\n",
      "Epoch 420/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1654 - accuracy: 0.9918 - val_loss: 1.4243 - val_accuracy: 0.7937\n",
      "Epoch 421/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1685 - accuracy: 0.9920 - val_loss: 1.3929 - val_accuracy: 0.7975\n",
      "Epoch 422/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1674 - accuracy: 0.9920 - val_loss: 1.3686 - val_accuracy: 0.7975\n",
      "Epoch 423/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1661 - accuracy: 0.9922 - val_loss: 1.3853 - val_accuracy: 0.7925\n",
      "Epoch 424/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1662 - accuracy: 0.9925 - val_loss: 1.4006 - val_accuracy: 0.7900\n",
      "Epoch 425/1000\n",
      "19/19 [==============================] - 1s 50ms/step - loss: 0.1621 - accuracy: 0.9926 - val_loss: 1.3786 - val_accuracy: 0.7962\n",
      "Epoch 426/1000\n",
      "19/19 [==============================] - 1s 32ms/step - loss: 0.1648 - accuracy: 0.9926 - val_loss: 1.3935 - val_accuracy: 0.7887\n",
      "Epoch 427/1000\n",
      "19/19 [==============================] - 1s 26ms/step - loss: 0.1637 - accuracy: 0.9928 - val_loss: 1.4086 - val_accuracy: 0.7887\n",
      "Epoch 428/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1672 - accuracy: 0.9922 - val_loss: 1.3906 - val_accuracy: 0.7912\n",
      "Epoch 429/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1659 - accuracy: 0.9921 - val_loss: 1.4197 - val_accuracy: 0.7788\n",
      "Epoch 430/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1714 - accuracy: 0.9907 - val_loss: 1.3804 - val_accuracy: 0.7875\n",
      "Epoch 431/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1657 - accuracy: 0.9914 - val_loss: 1.3676 - val_accuracy: 0.7937\n",
      "Epoch 432/1000\n",
      "19/19 [==============================] - 1s 49ms/step - loss: 0.1617 - accuracy: 0.9943 - val_loss: 1.3882 - val_accuracy: 0.7875\n",
      "Epoch 433/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1631 - accuracy: 0.9933 - val_loss: 1.3750 - val_accuracy: 0.7962\n",
      "Epoch 434/1000\n",
      "19/19 [==============================] - 1s 25ms/step - loss: 0.1620 - accuracy: 0.9942 - val_loss: 1.3754 - val_accuracy: 0.7925\n",
      "Epoch 435/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1620 - accuracy: 0.9938 - val_loss: 1.3544 - val_accuracy: 0.7912\n",
      "Epoch 436/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.1605 - accuracy: 0.9932 - val_loss: 1.4237 - val_accuracy: 0.7925\n",
      "Epoch 437/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1621 - accuracy: 0.9938 - val_loss: 1.3593 - val_accuracy: 0.7975\n",
      "Epoch 438/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.1600 - accuracy: 0.9935 - val_loss: 1.3426 - val_accuracy: 0.8025\n",
      "Epoch 439/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1605 - accuracy: 0.9944 - val_loss: 1.3426 - val_accuracy: 0.8000\n",
      "Epoch 440/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1634 - accuracy: 0.9923 - val_loss: 1.3376 - val_accuracy: 0.7987\n",
      "Epoch 441/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1623 - accuracy: 0.9929 - val_loss: 1.3521 - val_accuracy: 0.8000\n",
      "Epoch 442/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1624 - accuracy: 0.9924 - val_loss: 1.3027 - val_accuracy: 0.7975\n",
      "Epoch 443/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1619 - accuracy: 0.9936 - val_loss: 1.3538 - val_accuracy: 0.8012\n",
      "Epoch 444/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.1597 - accuracy: 0.9935 - val_loss: 1.3516 - val_accuracy: 0.7950\n",
      "Epoch 445/1000\n",
      "19/19 [==============================] - 1s 26ms/step - loss: 0.1605 - accuracy: 0.9929 - val_loss: 1.3888 - val_accuracy: 0.7925\n",
      "Epoch 446/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1600 - accuracy: 0.9932 - val_loss: 1.3254 - val_accuracy: 0.7987\n",
      "Epoch 447/1000\n",
      "19/19 [==============================] - 1s 46ms/step - loss: 0.1591 - accuracy: 0.9937 - val_loss: 1.3128 - val_accuracy: 0.7975\n",
      "Epoch 448/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1610 - accuracy: 0.9934 - val_loss: 1.3569 - val_accuracy: 0.8050\n",
      "Epoch 449/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1602 - accuracy: 0.9928 - val_loss: 1.3510 - val_accuracy: 0.7962\n",
      "Epoch 450/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1592 - accuracy: 0.9932 - val_loss: 1.3442 - val_accuracy: 0.8037\n",
      "Epoch 451/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "19/19 [==============================] - 1s 47ms/step - loss: 0.1570 - accuracy: 0.9940 - val_loss: 1.3403 - val_accuracy: 0.8050\n",
      "Epoch 452/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1613 - accuracy: 0.9920 - val_loss: 1.3263 - val_accuracy: 0.8012\n",
      "Epoch 453/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1578 - accuracy: 0.9939 - val_loss: 1.3153 - val_accuracy: 0.7900\n",
      "Epoch 454/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1583 - accuracy: 0.9940 - val_loss: 1.3807 - val_accuracy: 0.7987\n",
      "Epoch 455/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1605 - accuracy: 0.9932 - val_loss: 1.3743 - val_accuracy: 0.8037\n",
      "Epoch 456/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1573 - accuracy: 0.9941 - val_loss: 1.4084 - val_accuracy: 0.7912\n",
      "Epoch 457/1000\n",
      "19/19 [==============================] - 1s 25ms/step - loss: 0.1621 - accuracy: 0.9925 - val_loss: 1.3734 - val_accuracy: 0.8050\n",
      "Epoch 458/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1597 - accuracy: 0.9929 - val_loss: 1.3596 - val_accuracy: 0.7987\n",
      "Epoch 459/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1643 - accuracy: 0.9919 - val_loss: 1.3145 - val_accuracy: 0.8037\n",
      "Epoch 460/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1655 - accuracy: 0.9913 - val_loss: 1.3811 - val_accuracy: 0.7850\n",
      "Epoch 461/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1613 - accuracy: 0.9914 - val_loss: 1.3739 - val_accuracy: 0.7987\n",
      "Epoch 462/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1637 - accuracy: 0.9915 - val_loss: 1.3828 - val_accuracy: 0.7987\n",
      "Epoch 463/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1612 - accuracy: 0.9917 - val_loss: 1.3856 - val_accuracy: 0.8000\n",
      "Epoch 464/1000\n",
      "19/19 [==============================] - 1s 25ms/step - loss: 0.1590 - accuracy: 0.9930 - val_loss: 1.4015 - val_accuracy: 0.7950\n",
      "Epoch 465/1000\n",
      "19/19 [==============================] - 1s 32ms/step - loss: 0.1598 - accuracy: 0.9936 - val_loss: 1.3913 - val_accuracy: 0.8025\n",
      "Epoch 466/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1612 - accuracy: 0.9924 - val_loss: 1.4098 - val_accuracy: 0.7987\n",
      "Epoch 467/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1603 - accuracy: 0.9928 - val_loss: 1.4040 - val_accuracy: 0.7975\n",
      "Epoch 468/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1586 - accuracy: 0.9934 - val_loss: 1.4136 - val_accuracy: 0.8037\n",
      "Epoch 469/1000\n",
      "19/19 [==============================] - 1s 37ms/step - loss: 0.1608 - accuracy: 0.9926 - val_loss: 1.4247 - val_accuracy: 0.8050\n",
      "Epoch 470/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1623 - accuracy: 0.9926 - val_loss: 1.3987 - val_accuracy: 0.8100\n",
      "Epoch 471/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1615 - accuracy: 0.9922 - val_loss: 1.3867 - val_accuracy: 0.8025\n",
      "Epoch 472/1000\n",
      "19/19 [==============================] - 1s 33ms/step - loss: 0.1603 - accuracy: 0.9929 - val_loss: 1.4254 - val_accuracy: 0.8000\n",
      "Epoch 473/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1596 - accuracy: 0.9930 - val_loss: 1.4282 - val_accuracy: 0.8000\n",
      "Epoch 474/1000\n",
      "19/19 [==============================] - 1s 31ms/step - loss: 0.1615 - accuracy: 0.9928 - val_loss: 1.4310 - val_accuracy: 0.8012\n",
      "Epoch 475/1000\n",
      "19/19 [==============================] - 1s 32ms/step - loss: 0.1584 - accuracy: 0.9928 - val_loss: 1.4276 - val_accuracy: 0.7987\n",
      "Epoch 476/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1610 - accuracy: 0.9920 - val_loss: 1.4353 - val_accuracy: 0.8000\n",
      "Epoch 477/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1618 - accuracy: 0.9926 - val_loss: 1.4357 - val_accuracy: 0.7900\n",
      "Epoch 478/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1592 - accuracy: 0.9938 - val_loss: 1.3951 - val_accuracy: 0.7937\n",
      "Epoch 479/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1599 - accuracy: 0.9938 - val_loss: 1.3785 - val_accuracy: 0.7925\n",
      "Epoch 480/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1609 - accuracy: 0.9924 - val_loss: 1.3691 - val_accuracy: 0.8012\n",
      "Epoch 481/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1608 - accuracy: 0.9923 - val_loss: 1.3928 - val_accuracy: 0.8025\n",
      "Epoch 482/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1623 - accuracy: 0.9924 - val_loss: 1.4157 - val_accuracy: 0.7912\n",
      "Epoch 483/1000\n",
      "19/19 [==============================] - 1s 31ms/step - loss: 0.1581 - accuracy: 0.9937 - val_loss: 1.3818 - val_accuracy: 0.7962\n",
      "Epoch 484/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1602 - accuracy: 0.9914 - val_loss: 1.4107 - val_accuracy: 0.7975\n",
      "Epoch 485/1000\n",
      "19/19 [==============================] - 1s 35ms/step - loss: 0.1629 - accuracy: 0.9924 - val_loss: 1.3692 - val_accuracy: 0.7925\n",
      "Epoch 486/1000\n",
      "19/19 [==============================] - 1s 36ms/step - loss: 0.1600 - accuracy: 0.9925 - val_loss: 1.3851 - val_accuracy: 0.7875\n",
      "Epoch 487/1000\n",
      "19/19 [==============================] - 1s 31ms/step - loss: 0.1624 - accuracy: 0.9919 - val_loss: 1.4044 - val_accuracy: 0.7862\n",
      "Epoch 488/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1575 - accuracy: 0.9940 - val_loss: 1.3558 - val_accuracy: 0.7925\n",
      "Epoch 489/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1586 - accuracy: 0.9921 - val_loss: 1.3806 - val_accuracy: 0.8025\n",
      "Epoch 490/1000\n",
      "19/19 [==============================] - 1s 34ms/step - loss: 0.1601 - accuracy: 0.9926 - val_loss: 1.4265 - val_accuracy: 0.7937\n",
      "Epoch 491/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1573 - accuracy: 0.9932 - val_loss: 1.3894 - val_accuracy: 0.7900\n",
      "Epoch 492/1000\n",
      "19/19 [==============================] - 1s 31ms/step - loss: 0.1597 - accuracy: 0.9930 - val_loss: 1.4129 - val_accuracy: 0.7887\n",
      "Epoch 493/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1603 - accuracy: 0.9921 - val_loss: 1.3998 - val_accuracy: 0.7862\n",
      "Epoch 494/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1571 - accuracy: 0.9943 - val_loss: 1.3280 - val_accuracy: 0.7825\n",
      "Epoch 495/1000\n",
      "19/19 [==============================] - 1s 33ms/step - loss: 0.1593 - accuracy: 0.9933 - val_loss: 1.3648 - val_accuracy: 0.7950\n",
      "Epoch 496/1000\n",
      "19/19 [==============================] - 1s 50ms/step - loss: 0.1569 - accuracy: 0.9930 - val_loss: 1.3463 - val_accuracy: 0.7950\n",
      "Epoch 497/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1593 - accuracy: 0.9934 - val_loss: 1.3115 - val_accuracy: 0.8000\n",
      "Epoch 498/1000\n",
      "19/19 [==============================] - 1s 54ms/step - loss: 0.1543 - accuracy: 0.9944 - val_loss: 1.3392 - val_accuracy: 0.8012\n",
      "Epoch 499/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1547 - accuracy: 0.9938 - val_loss: 1.3397 - val_accuracy: 0.7987\n",
      "Epoch 500/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1571 - accuracy: 0.9932 - val_loss: 1.3329 - val_accuracy: 0.7950\n",
      "Epoch 501/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1565 - accuracy: 0.9936 - val_loss: 1.3304 - val_accuracy: 0.7937\n",
      "Epoch 502/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1551 - accuracy: 0.9935 - val_loss: 1.3680 - val_accuracy: 0.7937\n",
      "Epoch 503/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1572 - accuracy: 0.9938 - val_loss: 1.3392 - val_accuracy: 0.7975\n",
      "Epoch 504/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1606 - accuracy: 0.9921 - val_loss: 1.3216 - val_accuracy: 0.8025\n",
      "Epoch 505/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1600 - accuracy: 0.9920 - val_loss: 1.3881 - val_accuracy: 0.8000\n",
      "Epoch 506/1000\n",
      "19/19 [==============================] - 1s 49ms/step - loss: 0.1542 - accuracy: 0.9935 - val_loss: 1.4302 - val_accuracy: 0.7900\n",
      "Epoch 507/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1587 - accuracy: 0.9928 - val_loss: 1.3848 - val_accuracy: 0.7987\n",
      "Epoch 508/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1551 - accuracy: 0.9938 - val_loss: 1.3905 - val_accuracy: 0.8000\n",
      "Epoch 509/1000\n",
      "19/19 [==============================] - 1s 32ms/step - loss: 0.1570 - accuracy: 0.9932 - val_loss: 1.3949 - val_accuracy: 0.8037\n",
      "Epoch 510/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1574 - accuracy: 0.9938 - val_loss: 1.3991 - val_accuracy: 0.7912\n",
      "Epoch 511/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1551 - accuracy: 0.9928 - val_loss: 1.3728 - val_accuracy: 0.7962\n",
      "Epoch 512/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1583 - accuracy: 0.9926 - val_loss: 1.3536 - val_accuracy: 0.8075\n",
      "Epoch 513/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1566 - accuracy: 0.9924 - val_loss: 1.3769 - val_accuracy: 0.8062\n",
      "Epoch 514/1000\n",
      "19/19 [==============================] - 1s 49ms/step - loss: 0.1528 - accuracy: 0.9941 - val_loss: 1.3658 - val_accuracy: 0.8050\n",
      "Epoch 515/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1602 - accuracy: 0.9920 - val_loss: 1.3406 - val_accuracy: 0.8075\n",
      "Epoch 516/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1597 - accuracy: 0.9929 - val_loss: 1.3218 - val_accuracy: 0.8087\n",
      "Epoch 517/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1584 - accuracy: 0.9933 - val_loss: 1.3601 - val_accuracy: 0.8112\n",
      "Epoch 518/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1579 - accuracy: 0.9920 - val_loss: 1.3493 - val_accuracy: 0.8037\n",
      "Epoch 519/1000\n",
      "19/19 [==============================] - 1s 32ms/step - loss: 0.1553 - accuracy: 0.9939 - val_loss: 1.4120 - val_accuracy: 0.8000\n",
      "Epoch 520/1000\n",
      "19/19 [==============================] - 1s 26ms/step - loss: 0.1581 - accuracy: 0.9925 - val_loss: 1.3614 - val_accuracy: 0.8087\n",
      "Epoch 521/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1560 - accuracy: 0.9936 - val_loss: 1.3836 - val_accuracy: 0.8025\n",
      "Epoch 522/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1533 - accuracy: 0.9941 - val_loss: 1.3442 - val_accuracy: 0.8125\n",
      "Epoch 523/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1556 - accuracy: 0.9934 - val_loss: 1.3517 - val_accuracy: 0.8075\n",
      "Epoch 524/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1550 - accuracy: 0.9931 - val_loss: 1.3746 - val_accuracy: 0.8025\n",
      "Epoch 525/1000\n",
      "19/19 [==============================] - 1s 32ms/step - loss: 0.1543 - accuracy: 0.9937 - val_loss: 1.3167 - val_accuracy: 0.8112\n",
      "Epoch 526/1000\n",
      "19/19 [==============================] - 1s 32ms/step - loss: 0.1552 - accuracy: 0.9938 - val_loss: 1.3592 - val_accuracy: 0.8025\n",
      "Epoch 527/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1546 - accuracy: 0.9937 - val_loss: 1.3329 - val_accuracy: 0.8075\n",
      "Epoch 528/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1557 - accuracy: 0.9930 - val_loss: 1.3343 - val_accuracy: 0.8037\n",
      "Epoch 529/1000\n",
      "19/19 [==============================] - 1s 32ms/step - loss: 0.1551 - accuracy: 0.9935 - val_loss: 1.3501 - val_accuracy: 0.8025\n",
      "Epoch 530/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1555 - accuracy: 0.9929 - val_loss: 1.3587 - val_accuracy: 0.8012\n",
      "Epoch 531/1000\n",
      "19/19 [==============================] - 1s 34ms/step - loss: 0.1549 - accuracy: 0.9934 - val_loss: 1.3544 - val_accuracy: 0.8000\n",
      "Epoch 532/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1541 - accuracy: 0.9936 - val_loss: 1.3443 - val_accuracy: 0.8000\n",
      "Epoch 533/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1564 - accuracy: 0.9936 - val_loss: 1.3449 - val_accuracy: 0.8012\n",
      "Epoch 534/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1550 - accuracy: 0.9932 - val_loss: 1.3529 - val_accuracy: 0.8087\n",
      "Epoch 535/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.1516 - accuracy: 0.9948 - val_loss: 1.3324 - val_accuracy: 0.8012\n",
      "Epoch 536/1000\n",
      "19/19 [==============================] - 1s 26ms/step - loss: 0.1571 - accuracy: 0.9943 - val_loss: 1.3409 - val_accuracy: 0.7987\n",
      "Epoch 537/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.1510 - accuracy: 0.9944 - val_loss: 1.3398 - val_accuracy: 0.7975\n",
      "Epoch 538/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1516 - accuracy: 0.9932 - val_loss: 1.3387 - val_accuracy: 0.7925\n",
      "Epoch 539/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1522 - accuracy: 0.9940 - val_loss: 1.3338 - val_accuracy: 0.7987\n",
      "Epoch 540/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1532 - accuracy: 0.9939 - val_loss: 1.3074 - val_accuracy: 0.8025\n",
      "Epoch 541/1000\n",
      "19/19 [==============================] - 1s 46ms/step - loss: 0.1493 - accuracy: 0.9945 - val_loss: 1.4129 - val_accuracy: 0.7912\n",
      "Epoch 542/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1546 - accuracy: 0.9924 - val_loss: 1.3895 - val_accuracy: 0.8025\n",
      "Epoch 543/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1531 - accuracy: 0.9934 - val_loss: 1.4021 - val_accuracy: 0.8050\n",
      "Epoch 544/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1522 - accuracy: 0.9937 - val_loss: 1.3782 - val_accuracy: 0.7925\n",
      "Epoch 545/1000\n",
      "19/19 [==============================] - 1s 25ms/step - loss: 0.1538 - accuracy: 0.9927 - val_loss: 1.4284 - val_accuracy: 0.7875\n",
      "Epoch 546/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1563 - accuracy: 0.9924 - val_loss: 1.4131 - val_accuracy: 0.7950\n",
      "Epoch 547/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1541 - accuracy: 0.9933 - val_loss: 1.3740 - val_accuracy: 0.7950\n",
      "Epoch 548/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1523 - accuracy: 0.9936 - val_loss: 1.3445 - val_accuracy: 0.7987\n",
      "Epoch 549/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1529 - accuracy: 0.9932 - val_loss: 1.3436 - val_accuracy: 0.8037\n",
      "Epoch 550/1000\n",
      "19/19 [==============================] - 1s 32ms/step - loss: 0.1548 - accuracy: 0.9926 - val_loss: 1.3257 - val_accuracy: 0.8037\n",
      "Epoch 551/1000\n",
      "19/19 [==============================] - 1s 31ms/step - loss: 0.1535 - accuracy: 0.9930 - val_loss: 1.3939 - val_accuracy: 0.7987\n",
      "Epoch 552/1000\n",
      "19/19 [==============================] - 1s 35ms/step - loss: 0.1517 - accuracy: 0.9936 - val_loss: 1.3338 - val_accuracy: 0.8050\n",
      "Epoch 553/1000\n",
      "19/19 [==============================] - 1s 31ms/step - loss: 0.1516 - accuracy: 0.9938 - val_loss: 1.3296 - val_accuracy: 0.7987\n",
      "Epoch 554/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1534 - accuracy: 0.9929 - val_loss: 1.3206 - val_accuracy: 0.7987\n",
      "Epoch 555/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1538 - accuracy: 0.9928 - val_loss: 1.2996 - val_accuracy: 0.8087\n",
      "Epoch 556/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1522 - accuracy: 0.9939 - val_loss: 1.3240 - val_accuracy: 0.8025\n",
      "Epoch 557/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1534 - accuracy: 0.9937 - val_loss: 1.3631 - val_accuracy: 0.8050\n",
      "Epoch 558/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1518 - accuracy: 0.9932 - val_loss: 1.3698 - val_accuracy: 0.7925\n",
      "Epoch 559/1000\n",
      "19/19 [==============================] - 1s 26ms/step - loss: 0.1529 - accuracy: 0.9925 - val_loss: 1.3763 - val_accuracy: 0.7962\n",
      "Epoch 560/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1537 - accuracy: 0.9933 - val_loss: 1.4328 - val_accuracy: 0.7975\n",
      "Epoch 561/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1522 - accuracy: 0.9936 - val_loss: 1.4143 - val_accuracy: 0.7937\n",
      "Epoch 562/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1530 - accuracy: 0.9933 - val_loss: 1.4273 - val_accuracy: 0.7900\n",
      "Epoch 563/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1526 - accuracy: 0.9934 - val_loss: 1.3595 - val_accuracy: 0.8037\n",
      "Epoch 564/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1517 - accuracy: 0.9930 - val_loss: 1.3813 - val_accuracy: 0.8000\n",
      "Epoch 565/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1550 - accuracy: 0.9930 - val_loss: 1.3685 - val_accuracy: 0.7950\n",
      "Epoch 566/1000\n",
      "19/19 [==============================] - 1s 32ms/step - loss: 0.1513 - accuracy: 0.9936 - val_loss: 1.3706 - val_accuracy: 0.7962\n",
      "Epoch 567/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1520 - accuracy: 0.9932 - val_loss: 1.3781 - val_accuracy: 0.8025\n",
      "Epoch 568/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1529 - accuracy: 0.9926 - val_loss: 1.4063 - val_accuracy: 0.7900\n",
      "Epoch 569/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1543 - accuracy: 0.9927 - val_loss: 1.3791 - val_accuracy: 0.7950\n",
      "Epoch 570/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1556 - accuracy: 0.9921 - val_loss: 1.3522 - val_accuracy: 0.7987\n",
      "Epoch 571/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1563 - accuracy: 0.9930 - val_loss: 1.3519 - val_accuracy: 0.8025\n",
      "Epoch 572/1000\n",
      "19/19 [==============================] - 1s 32ms/step - loss: 0.1550 - accuracy: 0.9931 - val_loss: 1.3262 - val_accuracy: 0.7950\n",
      "Epoch 573/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1522 - accuracy: 0.9944 - val_loss: 1.2925 - val_accuracy: 0.8025\n",
      "Epoch 574/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1523 - accuracy: 0.9932 - val_loss: 1.3324 - val_accuracy: 0.8012\n",
      "Epoch 575/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1532 - accuracy: 0.9930 - val_loss: 1.3683 - val_accuracy: 0.8025\n",
      "Epoch 576/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1532 - accuracy: 0.9931 - val_loss: 1.3466 - val_accuracy: 0.8000\n",
      "Epoch 577/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1533 - accuracy: 0.9933 - val_loss: 1.3255 - val_accuracy: 0.8012\n",
      "Epoch 578/1000\n",
      "19/19 [==============================] - 1s 26ms/step - loss: 0.1504 - accuracy: 0.9934 - val_loss: 1.4062 - val_accuracy: 0.7887\n",
      "Epoch 579/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1503 - accuracy: 0.9940 - val_loss: 1.3701 - val_accuracy: 0.7987\n",
      "Epoch 580/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1547 - accuracy: 0.9926 - val_loss: 1.3670 - val_accuracy: 0.8000\n",
      "Epoch 581/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1503 - accuracy: 0.9938 - val_loss: 1.3733 - val_accuracy: 0.8037\n",
      "Epoch 582/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1528 - accuracy: 0.9929 - val_loss: 1.3376 - val_accuracy: 0.7987\n",
      "Epoch 583/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1526 - accuracy: 0.9925 - val_loss: 1.3333 - val_accuracy: 0.8037\n",
      "Epoch 584/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1524 - accuracy: 0.9934 - val_loss: 1.3620 - val_accuracy: 0.7962\n",
      "Epoch 585/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1526 - accuracy: 0.9929 - val_loss: 1.3652 - val_accuracy: 0.8025\n",
      "Epoch 586/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1504 - accuracy: 0.9932 - val_loss: 1.4057 - val_accuracy: 0.7925\n",
      "Epoch 587/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1544 - accuracy: 0.9925 - val_loss: 1.3305 - val_accuracy: 0.8037\n",
      "Epoch 588/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1519 - accuracy: 0.9934 - val_loss: 1.4110 - val_accuracy: 0.8050\n",
      "Epoch 589/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1534 - accuracy: 0.9928 - val_loss: 1.3866 - val_accuracy: 0.8012\n",
      "Epoch 590/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1521 - accuracy: 0.9922 - val_loss: 1.3336 - val_accuracy: 0.8100\n",
      "Epoch 591/1000\n",
      "19/19 [==============================] - 1s 31ms/step - loss: 0.1548 - accuracy: 0.9926 - val_loss: 1.3535 - val_accuracy: 0.8050\n",
      "Epoch 592/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1500 - accuracy: 0.9930 - val_loss: 1.3301 - val_accuracy: 0.8062\n",
      "Epoch 593/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1514 - accuracy: 0.9941 - val_loss: 1.3266 - val_accuracy: 0.8037\n",
      "Epoch 594/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1526 - accuracy: 0.9922 - val_loss: 1.3615 - val_accuracy: 0.8025\n",
      "Epoch 595/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1506 - accuracy: 0.9942 - val_loss: 1.3298 - val_accuracy: 0.7975\n",
      "Epoch 596/1000\n",
      "19/19 [==============================] - 1s 26ms/step - loss: 0.1498 - accuracy: 0.9944 - val_loss: 1.3334 - val_accuracy: 0.7900\n",
      "Epoch 597/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1534 - accuracy: 0.9930 - val_loss: 1.3790 - val_accuracy: 0.7837\n",
      "Epoch 598/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1503 - accuracy: 0.9941 - val_loss: 1.3393 - val_accuracy: 0.7950\n",
      "Epoch 599/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1511 - accuracy: 0.9936 - val_loss: 1.3203 - val_accuracy: 0.8012\n",
      "Epoch 600/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1524 - accuracy: 0.9934 - val_loss: 1.3560 - val_accuracy: 0.8037\n",
      "Epoch 601/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1494 - accuracy: 0.9947 - val_loss: 1.3842 - val_accuracy: 0.7875\n",
      "Epoch 602/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1511 - accuracy: 0.9934 - val_loss: 1.3560 - val_accuracy: 0.7875\n",
      "Epoch 603/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1528 - accuracy: 0.9931 - val_loss: 1.3730 - val_accuracy: 0.7975\n",
      "Epoch 604/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1535 - accuracy: 0.9931 - val_loss: 1.3353 - val_accuracy: 0.7950\n",
      "Epoch 605/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1530 - accuracy: 0.9927 - val_loss: 1.3304 - val_accuracy: 0.7975\n",
      "Epoch 606/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1521 - accuracy: 0.9926 - val_loss: 1.3660 - val_accuracy: 0.8012\n",
      "Epoch 607/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1540 - accuracy: 0.9925 - val_loss: 1.3521 - val_accuracy: 0.7987\n",
      "Epoch 608/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1504 - accuracy: 0.9932 - val_loss: 1.3284 - val_accuracy: 0.8087\n",
      "Epoch 609/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1498 - accuracy: 0.9941 - val_loss: 1.3130 - val_accuracy: 0.8137\n",
      "Epoch 610/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1495 - accuracy: 0.9945 - val_loss: 1.2587 - val_accuracy: 0.8137\n",
      "Epoch 611/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1498 - accuracy: 0.9942 - val_loss: 1.3276 - val_accuracy: 0.8050\n",
      "Epoch 612/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.1482 - accuracy: 0.9941 - val_loss: 1.3424 - val_accuracy: 0.8062\n",
      "Epoch 613/1000\n",
      "19/19 [==============================] - 1s 26ms/step - loss: 0.1494 - accuracy: 0.9937 - val_loss: 1.3600 - val_accuracy: 0.8037\n",
      "Epoch 614/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1483 - accuracy: 0.9940 - val_loss: 1.3639 - val_accuracy: 0.8087\n",
      "Epoch 615/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1522 - accuracy: 0.9923 - val_loss: 1.3367 - val_accuracy: 0.8100\n",
      "Epoch 616/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1535 - accuracy: 0.9931 - val_loss: 1.3657 - val_accuracy: 0.8062\n",
      "Epoch 617/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1482 - accuracy: 0.9938 - val_loss: 1.3403 - val_accuracy: 0.8025\n",
      "Epoch 618/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1503 - accuracy: 0.9932 - val_loss: 1.3151 - val_accuracy: 0.8025\n",
      "Epoch 619/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1512 - accuracy: 0.9928 - val_loss: 1.2814 - val_accuracy: 0.8087\n",
      "Epoch 620/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1532 - accuracy: 0.9930 - val_loss: 1.2563 - val_accuracy: 0.8100\n",
      "Epoch 621/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1500 - accuracy: 0.9938 - val_loss: 1.2587 - val_accuracy: 0.8087\n",
      "Epoch 622/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1504 - accuracy: 0.9934 - val_loss: 1.3210 - val_accuracy: 0.7937\n",
      "Epoch 623/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1524 - accuracy: 0.9935 - val_loss: 1.3037 - val_accuracy: 0.8000\n",
      "Epoch 624/1000\n",
      "19/19 [==============================] - 1s 50ms/step - loss: 0.1470 - accuracy: 0.9946 - val_loss: 1.3234 - val_accuracy: 0.8037\n",
      "Epoch 625/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1477 - accuracy: 0.9942 - val_loss: 1.3097 - val_accuracy: 0.8000\n",
      "Epoch 626/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1513 - accuracy: 0.9933 - val_loss: 1.3367 - val_accuracy: 0.7962\n",
      "Epoch 627/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1491 - accuracy: 0.9937 - val_loss: 1.3267 - val_accuracy: 0.8012\n",
      "Epoch 628/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1500 - accuracy: 0.9939 - val_loss: 1.3080 - val_accuracy: 0.7987\n",
      "Epoch 629/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1527 - accuracy: 0.9932 - val_loss: 1.2819 - val_accuracy: 0.8062\n",
      "Epoch 630/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1475 - accuracy: 0.9938 - val_loss: 1.3368 - val_accuracy: 0.8062\n",
      "Epoch 631/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1514 - accuracy: 0.9920 - val_loss: 1.3394 - val_accuracy: 0.7887\n",
      "Epoch 632/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1498 - accuracy: 0.9938 - val_loss: 1.3059 - val_accuracy: 0.7987\n",
      "Epoch 633/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1514 - accuracy: 0.9925 - val_loss: 1.3229 - val_accuracy: 0.8025\n",
      "Epoch 634/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1520 - accuracy: 0.9938 - val_loss: 1.3398 - val_accuracy: 0.8025\n",
      "Epoch 635/1000\n",
      "19/19 [==============================] - 1s 26ms/step - loss: 0.1486 - accuracy: 0.9936 - val_loss: 1.3349 - val_accuracy: 0.7937\n",
      "Epoch 636/1000\n",
      "19/19 [==============================] - 1s 26ms/step - loss: 0.1499 - accuracy: 0.9924 - val_loss: 1.2924 - val_accuracy: 0.7975\n",
      "Epoch 637/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1488 - accuracy: 0.9940 - val_loss: 1.2965 - val_accuracy: 0.7987\n",
      "Epoch 638/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1478 - accuracy: 0.9939 - val_loss: 1.2851 - val_accuracy: 0.7987\n",
      "Epoch 639/1000\n",
      "19/19 [==============================] - 1s 33ms/step - loss: 0.1517 - accuracy: 0.9933 - val_loss: 1.2958 - val_accuracy: 0.7975\n",
      "Epoch 640/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1472 - accuracy: 0.9950 - val_loss: 1.3438 - val_accuracy: 0.7950\n",
      "Epoch 641/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1470 - accuracy: 0.9936 - val_loss: 1.3295 - val_accuracy: 0.7875\n",
      "Epoch 642/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1488 - accuracy: 0.9939 - val_loss: 1.2928 - val_accuracy: 0.8000\n",
      "Epoch 643/1000\n",
      "19/19 [==============================] - 1s 31ms/step - loss: 0.1485 - accuracy: 0.9940 - val_loss: 1.3207 - val_accuracy: 0.7962\n",
      "Epoch 644/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1502 - accuracy: 0.9937 - val_loss: 1.3496 - val_accuracy: 0.7962\n",
      "Epoch 645/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1473 - accuracy: 0.9936 - val_loss: 1.3006 - val_accuracy: 0.7987\n",
      "Epoch 646/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.1464 - accuracy: 0.9934 - val_loss: 1.2642 - val_accuracy: 0.8075\n",
      "Epoch 647/1000\n",
      "19/19 [==============================] - 1s 26ms/step - loss: 0.1493 - accuracy: 0.9936 - val_loss: 1.3161 - val_accuracy: 0.7987\n",
      "Epoch 648/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1467 - accuracy: 0.9944 - val_loss: 1.2670 - val_accuracy: 0.8037\n",
      "Epoch 649/1000\n",
      "19/19 [==============================] - 0s 24ms/step - loss: 0.1505 - accuracy: 0.9928 - val_loss: 1.3134 - val_accuracy: 0.8012\n",
      "Epoch 650/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1491 - accuracy: 0.9934 - val_loss: 1.3073 - val_accuracy: 0.7975\n",
      "Epoch 651/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1497 - accuracy: 0.9935 - val_loss: 1.3240 - val_accuracy: 0.7975\n",
      "Epoch 652/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1473 - accuracy: 0.9938 - val_loss: 1.3248 - val_accuracy: 0.7925\n",
      "Epoch 653/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1478 - accuracy: 0.9932 - val_loss: 1.3148 - val_accuracy: 0.7975\n",
      "Epoch 654/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1476 - accuracy: 0.9937 - val_loss: 1.3056 - val_accuracy: 0.8062\n",
      "Epoch 655/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1480 - accuracy: 0.9934 - val_loss: 1.3227 - val_accuracy: 0.8100\n",
      "Epoch 656/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1470 - accuracy: 0.9945 - val_loss: 1.3271 - val_accuracy: 0.8050\n",
      "Epoch 657/1000\n",
      "19/19 [==============================] - 1s 45ms/step - loss: 0.1462 - accuracy: 0.9938 - val_loss: 1.3282 - val_accuracy: 0.7925\n",
      "Epoch 658/1000\n",
      "19/19 [==============================] - 1s 44ms/step - loss: 0.1457 - accuracy: 0.9941 - val_loss: 1.3129 - val_accuracy: 0.7987\n",
      "Epoch 659/1000\n",
      "19/19 [==============================] - 1s 44ms/step - loss: 0.1449 - accuracy: 0.9947 - val_loss: 1.3711 - val_accuracy: 0.7937\n",
      "Epoch 660/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1478 - accuracy: 0.9938 - val_loss: 1.3641 - val_accuracy: 0.7987\n",
      "Epoch 661/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1473 - accuracy: 0.9943 - val_loss: 1.3882 - val_accuracy: 0.8050\n",
      "Epoch 662/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1450 - accuracy: 0.9947 - val_loss: 1.3491 - val_accuracy: 0.7950\n",
      "Epoch 663/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1474 - accuracy: 0.9931 - val_loss: 1.2992 - val_accuracy: 0.8000\n",
      "Epoch 664/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1463 - accuracy: 0.9941 - val_loss: 1.3471 - val_accuracy: 0.7900\n",
      "Epoch 665/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1486 - accuracy: 0.9932 - val_loss: 1.3008 - val_accuracy: 0.8012\n",
      "Epoch 666/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1473 - accuracy: 0.9937 - val_loss: 1.2953 - val_accuracy: 0.8000\n",
      "Epoch 667/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1526 - accuracy: 0.9932 - val_loss: 1.2939 - val_accuracy: 0.7912\n",
      "Epoch 668/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1453 - accuracy: 0.9943 - val_loss: 1.3280 - val_accuracy: 0.7950\n",
      "Epoch 669/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1449 - accuracy: 0.9945 - val_loss: 1.3396 - val_accuracy: 0.7950\n",
      "Epoch 670/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1474 - accuracy: 0.9932 - val_loss: 1.3569 - val_accuracy: 0.7962\n",
      "Epoch 671/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1455 - accuracy: 0.9941 - val_loss: 1.3367 - val_accuracy: 0.8012\n",
      "Epoch 672/1000\n",
      "19/19 [==============================] - 1s 26ms/step - loss: 0.1482 - accuracy: 0.9928 - val_loss: 1.3941 - val_accuracy: 0.7987\n",
      "Epoch 673/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1484 - accuracy: 0.9932 - val_loss: 1.3847 - val_accuracy: 0.7975\n",
      "Epoch 674/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1486 - accuracy: 0.9928 - val_loss: 1.3081 - val_accuracy: 0.7937\n",
      "Epoch 675/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1473 - accuracy: 0.9938 - val_loss: 1.3410 - val_accuracy: 0.8000\n",
      "Epoch 676/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1466 - accuracy: 0.9931 - val_loss: 1.3328 - val_accuracy: 0.8025\n",
      "Epoch 677/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1496 - accuracy: 0.9934 - val_loss: 1.3316 - val_accuracy: 0.8025\n",
      "Epoch 678/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1469 - accuracy: 0.9936 - val_loss: 1.3276 - val_accuracy: 0.8000\n",
      "Epoch 679/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1488 - accuracy: 0.9932 - val_loss: 1.2886 - val_accuracy: 0.8037\n",
      "Epoch 680/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1483 - accuracy: 0.9930 - val_loss: 1.2841 - val_accuracy: 0.7987\n",
      "Epoch 681/1000\n",
      "19/19 [==============================] - 1s 47ms/step - loss: 0.1445 - accuracy: 0.9944 - val_loss: 1.3221 - val_accuracy: 0.8012\n",
      "Epoch 682/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1457 - accuracy: 0.9936 - val_loss: 1.3307 - val_accuracy: 0.7975\n",
      "Epoch 683/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1461 - accuracy: 0.9936 - val_loss: 1.3097 - val_accuracy: 0.7987\n",
      "Epoch 684/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1457 - accuracy: 0.9939 - val_loss: 1.3515 - val_accuracy: 0.7987\n",
      "Epoch 685/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1481 - accuracy: 0.9934 - val_loss: 1.3732 - val_accuracy: 0.8000\n",
      "Epoch 686/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1498 - accuracy: 0.9936 - val_loss: 1.3849 - val_accuracy: 0.7975\n",
      "Epoch 687/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1483 - accuracy: 0.9934 - val_loss: 1.3944 - val_accuracy: 0.7962\n",
      "Epoch 688/1000\n",
      "19/19 [==============================] - 1s 31ms/step - loss: 0.1514 - accuracy: 0.9931 - val_loss: 1.3599 - val_accuracy: 0.8112\n",
      "Epoch 689/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1497 - accuracy: 0.9928 - val_loss: 1.3213 - val_accuracy: 0.8037\n",
      "Epoch 690/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1452 - accuracy: 0.9945 - val_loss: 1.3655 - val_accuracy: 0.7987\n",
      "Epoch 691/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1504 - accuracy: 0.9922 - val_loss: 1.3307 - val_accuracy: 0.8037\n",
      "Epoch 692/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1463 - accuracy: 0.9934 - val_loss: 1.3314 - val_accuracy: 0.7987\n",
      "Epoch 693/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1460 - accuracy: 0.9937 - val_loss: 1.3759 - val_accuracy: 0.8000\n",
      "Epoch 694/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1456 - accuracy: 0.9945 - val_loss: 1.3222 - val_accuracy: 0.7950\n",
      "Epoch 695/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1445 - accuracy: 0.9953 - val_loss: 1.3386 - val_accuracy: 0.7975\n",
      "Epoch 696/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1461 - accuracy: 0.9947 - val_loss: 1.3441 - val_accuracy: 0.7912\n",
      "Epoch 697/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1453 - accuracy: 0.9943 - val_loss: 1.3381 - val_accuracy: 0.7950\n",
      "Epoch 698/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1467 - accuracy: 0.9941 - val_loss: 1.3498 - val_accuracy: 0.7912\n",
      "Epoch 699/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1451 - accuracy: 0.9945 - val_loss: 1.3823 - val_accuracy: 0.7750\n",
      "Epoch 700/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1475 - accuracy: 0.9937 - val_loss: 1.3171 - val_accuracy: 0.7837\n",
      "Epoch 701/1000\n",
      "19/19 [==============================] - 1s 49ms/step - loss: 0.1437 - accuracy: 0.9949 - val_loss: 1.3382 - val_accuracy: 0.7925\n",
      "Epoch 702/1000\n",
      "19/19 [==============================] - 0s 24ms/step - loss: 0.1460 - accuracy: 0.9945 - val_loss: 1.3285 - val_accuracy: 0.7900\n",
      "Epoch 703/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1462 - accuracy: 0.9938 - val_loss: 1.4064 - val_accuracy: 0.7800\n",
      "Epoch 704/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1458 - accuracy: 0.9934 - val_loss: 1.3532 - val_accuracy: 0.7900\n",
      "Epoch 705/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1446 - accuracy: 0.9934 - val_loss: 1.3904 - val_accuracy: 0.7887\n",
      "Epoch 706/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1476 - accuracy: 0.9933 - val_loss: 1.3934 - val_accuracy: 0.7900\n",
      "Epoch 707/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1459 - accuracy: 0.9938 - val_loss: 1.3135 - val_accuracy: 0.7975\n",
      "Epoch 708/1000\n",
      "19/19 [==============================] - 1s 47ms/step - loss: 0.1437 - accuracy: 0.9943 - val_loss: 1.2992 - val_accuracy: 0.7962\n",
      "Epoch 709/1000\n",
      "19/19 [==============================] - 0s 24ms/step - loss: 0.1461 - accuracy: 0.9946 - val_loss: 1.2889 - val_accuracy: 0.7975\n",
      "Epoch 710/1000\n",
      "19/19 [==============================] - 1s 49ms/step - loss: 0.1428 - accuracy: 0.9944 - val_loss: 1.3061 - val_accuracy: 0.7962\n",
      "Epoch 711/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1437 - accuracy: 0.9945 - val_loss: 1.3085 - val_accuracy: 0.7937\n",
      "Epoch 712/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1471 - accuracy: 0.9924 - val_loss: 1.2790 - val_accuracy: 0.7950\n",
      "Epoch 713/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.1424 - accuracy: 0.9955 - val_loss: 1.3304 - val_accuracy: 0.7887\n",
      "Epoch 714/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1445 - accuracy: 0.9945 - val_loss: 1.3112 - val_accuracy: 0.7900\n",
      "Epoch 715/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1492 - accuracy: 0.9925 - val_loss: 1.2905 - val_accuracy: 0.7975\n",
      "Epoch 716/1000\n",
      "19/19 [==============================] - 1s 52ms/step - loss: 0.1423 - accuracy: 0.9947 - val_loss: 1.2828 - val_accuracy: 0.7962\n",
      "Epoch 717/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1445 - accuracy: 0.9938 - val_loss: 1.2542 - val_accuracy: 0.8075\n",
      "Epoch 718/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1441 - accuracy: 0.9941 - val_loss: 1.3242 - val_accuracy: 0.7962\n",
      "Epoch 719/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1467 - accuracy: 0.9939 - val_loss: 1.3068 - val_accuracy: 0.8000\n",
      "Epoch 720/1000\n",
      "19/19 [==============================] - 1s 46ms/step - loss: 0.1409 - accuracy: 0.9953 - val_loss: 1.3022 - val_accuracy: 0.7950\n",
      "Epoch 721/1000\n",
      "19/19 [==============================] - 1s 31ms/step - loss: 0.1438 - accuracy: 0.9938 - val_loss: 1.3209 - val_accuracy: 0.7875\n",
      "Epoch 722/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1430 - accuracy: 0.9943 - val_loss: 1.3334 - val_accuracy: 0.7962\n",
      "Epoch 723/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1488 - accuracy: 0.9928 - val_loss: 1.3085 - val_accuracy: 0.7900\n",
      "Epoch 724/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1466 - accuracy: 0.9927 - val_loss: 1.3201 - val_accuracy: 0.7950\n",
      "Epoch 725/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1440 - accuracy: 0.9940 - val_loss: 1.2757 - val_accuracy: 0.7962\n",
      "Epoch 726/1000\n",
      "19/19 [==============================] - 1s 25ms/step - loss: 0.1448 - accuracy: 0.9932 - val_loss: 1.2915 - val_accuracy: 0.7925\n",
      "Epoch 727/1000\n",
      "19/19 [==============================] - 1s 25ms/step - loss: 0.1424 - accuracy: 0.9943 - val_loss: 1.3267 - val_accuracy: 0.7887\n",
      "Epoch 728/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1465 - accuracy: 0.9929 - val_loss: 1.3299 - val_accuracy: 0.7912\n",
      "Epoch 729/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1438 - accuracy: 0.9938 - val_loss: 1.3372 - val_accuracy: 0.7937\n",
      "Epoch 730/1000\n",
      "19/19 [==============================] - 1s 25ms/step - loss: 0.1440 - accuracy: 0.9940 - val_loss: 1.3149 - val_accuracy: 0.7925\n",
      "Epoch 731/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1439 - accuracy: 0.9947 - val_loss: 1.3244 - val_accuracy: 0.7950\n",
      "Epoch 732/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1454 - accuracy: 0.9936 - val_loss: 1.3159 - val_accuracy: 0.7975\n",
      "Epoch 733/1000\n",
      "19/19 [==============================] - 1s 34ms/step - loss: 0.1428 - accuracy: 0.9941 - val_loss: 1.3083 - val_accuracy: 0.7912\n",
      "Epoch 734/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1450 - accuracy: 0.9934 - val_loss: 1.3420 - val_accuracy: 0.7912\n",
      "Epoch 735/1000\n",
      "19/19 [==============================] - 0s 24ms/step - loss: 0.1423 - accuracy: 0.9946 - val_loss: 1.3307 - val_accuracy: 0.7900\n",
      "Epoch 736/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1442 - accuracy: 0.9938 - val_loss: 1.3800 - val_accuracy: 0.7887\n",
      "Epoch 737/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1420 - accuracy: 0.9948 - val_loss: 1.3599 - val_accuracy: 0.7987\n",
      "Epoch 738/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1437 - accuracy: 0.9943 - val_loss: 1.3651 - val_accuracy: 0.8037\n",
      "Epoch 739/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1501 - accuracy: 0.9929 - val_loss: 1.4069 - val_accuracy: 0.7987\n",
      "Epoch 740/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1414 - accuracy: 0.9946 - val_loss: 1.3438 - val_accuracy: 0.8037\n",
      "Epoch 741/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1419 - accuracy: 0.9945 - val_loss: 1.2605 - val_accuracy: 0.8012\n",
      "Epoch 742/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1453 - accuracy: 0.9934 - val_loss: 1.3021 - val_accuracy: 0.8000\n",
      "Epoch 743/1000\n",
      "19/19 [==============================] - 1s 46ms/step - loss: 0.1406 - accuracy: 0.9947 - val_loss: 1.3281 - val_accuracy: 0.8000\n",
      "Epoch 744/1000\n",
      "19/19 [==============================] - 1s 26ms/step - loss: 0.1439 - accuracy: 0.9940 - val_loss: 1.3496 - val_accuracy: 0.7950\n",
      "Epoch 745/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1462 - accuracy: 0.9929 - val_loss: 1.3237 - val_accuracy: 0.8000\n",
      "Epoch 746/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1422 - accuracy: 0.9945 - val_loss: 1.3034 - val_accuracy: 0.7912\n",
      "Epoch 747/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1443 - accuracy: 0.9930 - val_loss: 1.3785 - val_accuracy: 0.7912\n",
      "Epoch 748/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1430 - accuracy: 0.9943 - val_loss: 1.3448 - val_accuracy: 0.8012\n",
      "Epoch 749/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1437 - accuracy: 0.9934 - val_loss: 1.3605 - val_accuracy: 0.7975\n",
      "Epoch 750/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1463 - accuracy: 0.9930 - val_loss: 1.3528 - val_accuracy: 0.7912\n",
      "Epoch 751/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1446 - accuracy: 0.9938 - val_loss: 1.3427 - val_accuracy: 0.7962\n",
      "Epoch 752/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1447 - accuracy: 0.9938 - val_loss: 1.3125 - val_accuracy: 0.7887\n",
      "Epoch 753/1000\n",
      "19/19 [==============================] - 1s 25ms/step - loss: 0.1441 - accuracy: 0.9941 - val_loss: 1.3444 - val_accuracy: 0.7950\n",
      "Epoch 754/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1422 - accuracy: 0.9943 - val_loss: 1.3246 - val_accuracy: 0.7962\n",
      "Epoch 755/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1475 - accuracy: 0.9936 - val_loss: 1.4023 - val_accuracy: 0.7950\n",
      "Epoch 756/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1469 - accuracy: 0.9928 - val_loss: 1.3900 - val_accuracy: 0.7962\n",
      "Epoch 757/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1426 - accuracy: 0.9940 - val_loss: 1.3105 - val_accuracy: 0.7975\n",
      "Epoch 758/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1428 - accuracy: 0.9946 - val_loss: 1.2892 - val_accuracy: 0.7887\n",
      "Epoch 759/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1451 - accuracy: 0.9936 - val_loss: 1.3475 - val_accuracy: 0.7900\n",
      "Epoch 760/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1445 - accuracy: 0.9936 - val_loss: 1.3657 - val_accuracy: 0.7950\n",
      "Epoch 761/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1424 - accuracy: 0.9945 - val_loss: 1.3331 - val_accuracy: 0.8012\n",
      "Epoch 762/1000\n",
      "19/19 [==============================] - 1s 32ms/step - loss: 0.1472 - accuracy: 0.9926 - val_loss: 1.3392 - val_accuracy: 0.7962\n",
      "Epoch 763/1000\n",
      "19/19 [==============================] - 0s 24ms/step - loss: 0.1425 - accuracy: 0.9939 - val_loss: 1.2798 - val_accuracy: 0.7975\n",
      "Epoch 764/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1456 - accuracy: 0.9936 - val_loss: 1.3431 - val_accuracy: 0.7987\n",
      "Epoch 765/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1438 - accuracy: 0.9932 - val_loss: 1.3578 - val_accuracy: 0.8025\n",
      "Epoch 766/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1434 - accuracy: 0.9939 - val_loss: 1.3419 - val_accuracy: 0.8000\n",
      "Epoch 767/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1427 - accuracy: 0.9942 - val_loss: 1.3368 - val_accuracy: 0.7912\n",
      "Epoch 768/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1452 - accuracy: 0.9926 - val_loss: 1.3052 - val_accuracy: 0.7975\n",
      "Epoch 769/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1459 - accuracy: 0.9938 - val_loss: 1.3722 - val_accuracy: 0.7937\n",
      "Epoch 770/1000\n",
      "19/19 [==============================] - 1s 32ms/step - loss: 0.1446 - accuracy: 0.9934 - val_loss: 1.2896 - val_accuracy: 0.7962\n",
      "Epoch 771/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1453 - accuracy: 0.9933 - val_loss: 1.3409 - val_accuracy: 0.8012\n",
      "Epoch 772/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1444 - accuracy: 0.9932 - val_loss: 1.3062 - val_accuracy: 0.8075\n",
      "Epoch 773/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1447 - accuracy: 0.9948 - val_loss: 1.2944 - val_accuracy: 0.8037\n",
      "Epoch 774/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1448 - accuracy: 0.9934 - val_loss: 1.3320 - val_accuracy: 0.8050\n",
      "Epoch 775/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1436 - accuracy: 0.9933 - val_loss: 1.3147 - val_accuracy: 0.7975\n",
      "Epoch 776/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1454 - accuracy: 0.9949 - val_loss: 1.2867 - val_accuracy: 0.8100\n",
      "Epoch 777/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1436 - accuracy: 0.9938 - val_loss: 1.2764 - val_accuracy: 0.8100\n",
      "Epoch 778/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1426 - accuracy: 0.9942 - val_loss: 1.2696 - val_accuracy: 0.8100\n",
      "Epoch 779/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1450 - accuracy: 0.9936 - val_loss: 1.2613 - val_accuracy: 0.8050\n",
      "Epoch 780/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1435 - accuracy: 0.9937 - val_loss: 1.2756 - val_accuracy: 0.7975\n",
      "Epoch 781/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1449 - accuracy: 0.9938 - val_loss: 1.2548 - val_accuracy: 0.7912\n",
      "Epoch 782/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1444 - accuracy: 0.9934 - val_loss: 1.3127 - val_accuracy: 0.7937\n",
      "Epoch 783/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1446 - accuracy: 0.9937 - val_loss: 1.3109 - val_accuracy: 0.7937\n",
      "Epoch 784/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1482 - accuracy: 0.9938 - val_loss: 1.3906 - val_accuracy: 0.7763\n",
      "Epoch 785/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1450 - accuracy: 0.9937 - val_loss: 1.3748 - val_accuracy: 0.7875\n",
      "Epoch 786/1000\n",
      "19/19 [==============================] - 1s 24ms/step - loss: 0.1437 - accuracy: 0.9940 - val_loss: 1.3011 - val_accuracy: 0.7862\n",
      "Epoch 787/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1453 - accuracy: 0.9936 - val_loss: 1.3070 - val_accuracy: 0.8000\n",
      "Epoch 788/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1414 - accuracy: 0.9950 - val_loss: 1.3410 - val_accuracy: 0.7925\n",
      "Epoch 789/1000\n",
      "19/19 [==============================] - 1s 47ms/step - loss: 0.1403 - accuracy: 0.9950 - val_loss: 1.3171 - val_accuracy: 0.8037\n",
      "Epoch 790/1000\n",
      "19/19 [==============================] - 1s 25ms/step - loss: 0.1406 - accuracy: 0.9947 - val_loss: 1.3077 - val_accuracy: 0.8025\n",
      "Epoch 791/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1420 - accuracy: 0.9939 - val_loss: 1.2813 - val_accuracy: 0.7950\n",
      "Epoch 792/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1413 - accuracy: 0.9945 - val_loss: 1.2939 - val_accuracy: 0.8025\n",
      "Epoch 793/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1434 - accuracy: 0.9935 - val_loss: 1.2907 - val_accuracy: 0.8000\n",
      "Epoch 794/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1438 - accuracy: 0.9935 - val_loss: 1.3622 - val_accuracy: 0.7862\n",
      "Epoch 795/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1447 - accuracy: 0.9939 - val_loss: 1.2791 - val_accuracy: 0.8025\n",
      "Epoch 796/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1465 - accuracy: 0.9932 - val_loss: 1.2830 - val_accuracy: 0.7975\n",
      "Epoch 797/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1414 - accuracy: 0.9941 - val_loss: 1.3314 - val_accuracy: 0.7962\n",
      "Epoch 798/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1422 - accuracy: 0.9941 - val_loss: 1.2742 - val_accuracy: 0.8025\n",
      "Epoch 799/1000\n",
      "19/19 [==============================] - 1s 32ms/step - loss: 0.1422 - accuracy: 0.9946 - val_loss: 1.3011 - val_accuracy: 0.7987\n",
      "Epoch 800/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1439 - accuracy: 0.9935 - val_loss: 1.3162 - val_accuracy: 0.7937\n",
      "Epoch 801/1000\n",
      "19/19 [==============================] - 1s 48ms/step - loss: 0.1392 - accuracy: 0.9949 - val_loss: 1.3178 - val_accuracy: 0.8012\n",
      "Epoch 802/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1404 - accuracy: 0.9943 - val_loss: 1.3073 - val_accuracy: 0.8025\n",
      "Epoch 803/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1415 - accuracy: 0.9947 - val_loss: 1.2726 - val_accuracy: 0.8087\n",
      "Epoch 804/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1409 - accuracy: 0.9943 - val_loss: 1.3002 - val_accuracy: 0.8037\n",
      "Epoch 805/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1408 - accuracy: 0.9945 - val_loss: 1.2778 - val_accuracy: 0.8025\n",
      "Epoch 806/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1429 - accuracy: 0.9937 - val_loss: 1.2944 - val_accuracy: 0.8000\n",
      "Epoch 807/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1442 - accuracy: 0.9936 - val_loss: 1.2957 - val_accuracy: 0.8062\n",
      "Epoch 808/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1451 - accuracy: 0.9939 - val_loss: 1.2838 - val_accuracy: 0.8000\n",
      "Epoch 809/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1439 - accuracy: 0.9930 - val_loss: 1.3122 - val_accuracy: 0.7975\n",
      "Epoch 810/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1432 - accuracy: 0.9934 - val_loss: 1.2825 - val_accuracy: 0.7937\n",
      "Epoch 811/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1459 - accuracy: 0.9930 - val_loss: 1.2965 - val_accuracy: 0.7950\n",
      "Epoch 812/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1446 - accuracy: 0.9937 - val_loss: 1.3002 - val_accuracy: 0.8037\n",
      "Epoch 813/1000\n",
      "19/19 [==============================] - 1s 25ms/step - loss: 0.1431 - accuracy: 0.9938 - val_loss: 1.2925 - val_accuracy: 0.7850\n",
      "Epoch 814/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1419 - accuracy: 0.9944 - val_loss: 1.3198 - val_accuracy: 0.8000\n",
      "Epoch 815/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1435 - accuracy: 0.9938 - val_loss: 1.2911 - val_accuracy: 0.8050\n",
      "Epoch 816/1000\n",
      "19/19 [==============================] - 1s 25ms/step - loss: 0.1426 - accuracy: 0.9948 - val_loss: 1.2983 - val_accuracy: 0.7975\n",
      "Epoch 817/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1413 - accuracy: 0.9932 - val_loss: 1.3779 - val_accuracy: 0.7912\n",
      "Epoch 818/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1399 - accuracy: 0.9944 - val_loss: 1.3569 - val_accuracy: 0.7912\n",
      "Epoch 819/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1392 - accuracy: 0.9958 - val_loss: 1.3433 - val_accuracy: 0.7875\n",
      "Epoch 820/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1406 - accuracy: 0.9944 - val_loss: 1.3866 - val_accuracy: 0.7925\n",
      "Epoch 821/1000\n",
      "19/19 [==============================] - 1s 46ms/step - loss: 0.1371 - accuracy: 0.9955 - val_loss: 1.3658 - val_accuracy: 0.7900\n",
      "Epoch 822/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1382 - accuracy: 0.9951 - val_loss: 1.3241 - val_accuracy: 0.7887\n",
      "Epoch 823/1000\n",
      "19/19 [==============================] - 1s 25ms/step - loss: 0.1414 - accuracy: 0.9947 - val_loss: 1.3634 - val_accuracy: 0.7937\n",
      "Epoch 824/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1410 - accuracy: 0.9944 - val_loss: 1.4383 - val_accuracy: 0.7887\n",
      "Epoch 825/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1404 - accuracy: 0.9939 - val_loss: 1.3769 - val_accuracy: 0.7937\n",
      "Epoch 826/1000\n",
      "19/19 [==============================] - 1s 25ms/step - loss: 0.1429 - accuracy: 0.9941 - val_loss: 1.3546 - val_accuracy: 0.7825\n",
      "Epoch 827/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1426 - accuracy: 0.9932 - val_loss: 1.3567 - val_accuracy: 0.7900\n",
      "Epoch 828/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1432 - accuracy: 0.9932 - val_loss: 1.3485 - val_accuracy: 0.7975\n",
      "Epoch 829/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1409 - accuracy: 0.9938 - val_loss: 1.3735 - val_accuracy: 0.7950\n",
      "Epoch 830/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1430 - accuracy: 0.9934 - val_loss: 1.3512 - val_accuracy: 0.7937\n",
      "Epoch 831/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1402 - accuracy: 0.9942 - val_loss: 1.3852 - val_accuracy: 0.8037\n",
      "Epoch 832/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1438 - accuracy: 0.9945 - val_loss: 1.3924 - val_accuracy: 0.7962\n",
      "Epoch 833/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1437 - accuracy: 0.9938 - val_loss: 1.3184 - val_accuracy: 0.7987\n",
      "Epoch 834/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1405 - accuracy: 0.9945 - val_loss: 1.3070 - val_accuracy: 0.7987\n",
      "Epoch 835/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1415 - accuracy: 0.9943 - val_loss: 1.3655 - val_accuracy: 0.7962\n",
      "Epoch 836/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1416 - accuracy: 0.9939 - val_loss: 1.3820 - val_accuracy: 0.7975\n",
      "Epoch 837/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1401 - accuracy: 0.9942 - val_loss: 1.3557 - val_accuracy: 0.7912\n",
      "Epoch 838/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1448 - accuracy: 0.9931 - val_loss: 1.3734 - val_accuracy: 0.7837\n",
      "Epoch 839/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1414 - accuracy: 0.9936 - val_loss: 1.3742 - val_accuracy: 0.7925\n",
      "Epoch 840/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1435 - accuracy: 0.9934 - val_loss: 1.3800 - val_accuracy: 0.7912\n",
      "Epoch 841/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1412 - accuracy: 0.9943 - val_loss: 1.3609 - val_accuracy: 0.7900\n",
      "Epoch 842/1000\n",
      "19/19 [==============================] - 1s 25ms/step - loss: 0.1445 - accuracy: 0.9927 - val_loss: 1.4143 - val_accuracy: 0.7850\n",
      "Epoch 843/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1396 - accuracy: 0.9947 - val_loss: 1.3989 - val_accuracy: 0.7937\n",
      "Epoch 844/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1388 - accuracy: 0.9950 - val_loss: 1.3903 - val_accuracy: 0.7812\n",
      "Epoch 845/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1418 - accuracy: 0.9936 - val_loss: 1.3428 - val_accuracy: 0.7962\n",
      "Epoch 846/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1404 - accuracy: 0.9938 - val_loss: 1.3122 - val_accuracy: 0.7950\n",
      "Epoch 847/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1388 - accuracy: 0.9949 - val_loss: 1.3223 - val_accuracy: 0.7950\n",
      "Epoch 848/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1405 - accuracy: 0.9939 - val_loss: 1.3473 - val_accuracy: 0.7900\n",
      "Epoch 849/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1378 - accuracy: 0.9953 - val_loss: 1.3341 - val_accuracy: 0.7912\n",
      "Epoch 850/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1387 - accuracy: 0.9949 - val_loss: 1.3128 - val_accuracy: 0.7975\n",
      "Epoch 851/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1415 - accuracy: 0.9938 - val_loss: 1.3393 - val_accuracy: 0.7950\n",
      "Epoch 852/1000\n",
      "19/19 [==============================] - 1s 31ms/step - loss: 0.1425 - accuracy: 0.9941 - val_loss: 1.3292 - val_accuracy: 0.8075\n",
      "Epoch 853/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1402 - accuracy: 0.9938 - val_loss: 1.3074 - val_accuracy: 0.8025\n",
      "Epoch 854/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1394 - accuracy: 0.9937 - val_loss: 1.2836 - val_accuracy: 0.8000\n",
      "Epoch 855/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1418 - accuracy: 0.9939 - val_loss: 1.2622 - val_accuracy: 0.7987\n",
      "Epoch 856/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1381 - accuracy: 0.9950 - val_loss: 1.3084 - val_accuracy: 0.7925\n",
      "Epoch 857/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1394 - accuracy: 0.9942 - val_loss: 1.2937 - val_accuracy: 0.7950\n",
      "Epoch 858/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1403 - accuracy: 0.9944 - val_loss: 1.2760 - val_accuracy: 0.8012\n",
      "Epoch 859/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1399 - accuracy: 0.9943 - val_loss: 1.2848 - val_accuracy: 0.8000\n",
      "Epoch 860/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1387 - accuracy: 0.9945 - val_loss: 1.2799 - val_accuracy: 0.7975\n",
      "Epoch 861/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1388 - accuracy: 0.9949 - val_loss: 1.2782 - val_accuracy: 0.7900\n",
      "Epoch 862/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1403 - accuracy: 0.9944 - val_loss: 1.3031 - val_accuracy: 0.8012\n",
      "Epoch 863/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1399 - accuracy: 0.9939 - val_loss: 1.3274 - val_accuracy: 0.8025\n",
      "Epoch 864/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1382 - accuracy: 0.9949 - val_loss: 1.3101 - val_accuracy: 0.8000\n",
      "Epoch 865/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1424 - accuracy: 0.9927 - val_loss: 1.3668 - val_accuracy: 0.7962\n",
      "Epoch 866/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1447 - accuracy: 0.9934 - val_loss: 1.3302 - val_accuracy: 0.7925\n",
      "Epoch 867/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1439 - accuracy: 0.9931 - val_loss: 1.3907 - val_accuracy: 0.7900\n",
      "Epoch 868/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1438 - accuracy: 0.9932 - val_loss: 1.3612 - val_accuracy: 0.7975\n",
      "Epoch 869/1000\n",
      "19/19 [==============================] - 1s 32ms/step - loss: 0.1412 - accuracy: 0.9939 - val_loss: 1.3304 - val_accuracy: 0.7950\n",
      "Epoch 870/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1382 - accuracy: 0.9947 - val_loss: 1.3426 - val_accuracy: 0.7937\n",
      "Epoch 871/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1420 - accuracy: 0.9932 - val_loss: 1.3230 - val_accuracy: 0.7962\n",
      "Epoch 872/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1426 - accuracy: 0.9935 - val_loss: 1.3229 - val_accuracy: 0.8012\n",
      "Epoch 873/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1441 - accuracy: 0.9936 - val_loss: 1.3267 - val_accuracy: 0.7812\n",
      "Epoch 874/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1427 - accuracy: 0.9932 - val_loss: 1.3448 - val_accuracy: 0.8012\n",
      "Epoch 875/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1411 - accuracy: 0.9936 - val_loss: 1.3282 - val_accuracy: 0.8000\n",
      "Epoch 876/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1419 - accuracy: 0.9941 - val_loss: 1.3674 - val_accuracy: 0.7987\n",
      "Epoch 877/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1393 - accuracy: 0.9949 - val_loss: 1.3969 - val_accuracy: 0.7937\n",
      "Epoch 878/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1441 - accuracy: 0.9930 - val_loss: 1.3862 - val_accuracy: 0.7975\n",
      "Epoch 879/1000\n",
      "19/19 [==============================] - 1s 25ms/step - loss: 0.1387 - accuracy: 0.9947 - val_loss: 1.3811 - val_accuracy: 0.7887\n",
      "Epoch 880/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1415 - accuracy: 0.9934 - val_loss: 1.3735 - val_accuracy: 0.7937\n",
      "Epoch 881/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1411 - accuracy: 0.9939 - val_loss: 1.3464 - val_accuracy: 0.7987\n",
      "Epoch 882/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1400 - accuracy: 0.9946 - val_loss: 1.3539 - val_accuracy: 0.7962\n",
      "Epoch 883/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1412 - accuracy: 0.9940 - val_loss: 1.3805 - val_accuracy: 0.7950\n",
      "Epoch 884/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1403 - accuracy: 0.9941 - val_loss: 1.3772 - val_accuracy: 0.7887\n",
      "Epoch 885/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1386 - accuracy: 0.9938 - val_loss: 1.3533 - val_accuracy: 0.7912\n",
      "Epoch 886/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1388 - accuracy: 0.9953 - val_loss: 1.3640 - val_accuracy: 0.7862\n",
      "Epoch 887/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1391 - accuracy: 0.9943 - val_loss: 1.3754 - val_accuracy: 0.7912\n",
      "Epoch 888/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1381 - accuracy: 0.9941 - val_loss: 1.3434 - val_accuracy: 0.7950\n",
      "Epoch 889/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1393 - accuracy: 0.9949 - val_loss: 1.4436 - val_accuracy: 0.7887\n",
      "Epoch 890/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1403 - accuracy: 0.9936 - val_loss: 1.4109 - val_accuracy: 0.7900\n",
      "Epoch 891/1000\n",
      "19/19 [==============================] - 1s 47ms/step - loss: 0.1370 - accuracy: 0.9948 - val_loss: 1.3998 - val_accuracy: 0.7862\n",
      "Epoch 892/1000\n",
      "19/19 [==============================] - 1s 50ms/step - loss: 0.1359 - accuracy: 0.9950 - val_loss: 1.3942 - val_accuracy: 0.7875\n",
      "Epoch 893/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1391 - accuracy: 0.9941 - val_loss: 1.3376 - val_accuracy: 0.7900\n",
      "Epoch 894/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1395 - accuracy: 0.9938 - val_loss: 1.3447 - val_accuracy: 0.7912\n",
      "Epoch 895/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1385 - accuracy: 0.9947 - val_loss: 1.3505 - val_accuracy: 0.7887\n",
      "Epoch 896/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1370 - accuracy: 0.9951 - val_loss: 1.3530 - val_accuracy: 0.7975\n",
      "Epoch 897/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1377 - accuracy: 0.9941 - val_loss: 1.3501 - val_accuracy: 0.7987\n",
      "Epoch 898/1000\n",
      "19/19 [==============================] - 1s 47ms/step - loss: 0.1358 - accuracy: 0.9942 - val_loss: 1.3537 - val_accuracy: 0.7900\n",
      "Epoch 899/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1381 - accuracy: 0.9945 - val_loss: 1.3316 - val_accuracy: 0.8012\n",
      "Epoch 900/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1389 - accuracy: 0.9946 - val_loss: 1.3286 - val_accuracy: 0.7962\n",
      "Epoch 901/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1390 - accuracy: 0.9945 - val_loss: 1.3595 - val_accuracy: 0.7912\n",
      "Epoch 902/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1377 - accuracy: 0.9945 - val_loss: 1.4145 - val_accuracy: 0.7950\n",
      "Epoch 903/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1379 - accuracy: 0.9940 - val_loss: 1.3403 - val_accuracy: 0.7912\n",
      "Epoch 904/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1366 - accuracy: 0.9947 - val_loss: 1.3455 - val_accuracy: 0.7900\n",
      "Epoch 905/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1386 - accuracy: 0.9946 - val_loss: 1.3698 - val_accuracy: 0.7887\n",
      "Epoch 906/1000\n",
      "19/19 [==============================] - 1s 31ms/step - loss: 0.1402 - accuracy: 0.9945 - val_loss: 1.3439 - val_accuracy: 0.7900\n",
      "Epoch 907/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1371 - accuracy: 0.9949 - val_loss: 1.3548 - val_accuracy: 0.7962\n",
      "Epoch 908/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1365 - accuracy: 0.9948 - val_loss: 1.3364 - val_accuracy: 0.8000\n",
      "Epoch 909/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1367 - accuracy: 0.9938 - val_loss: 1.3204 - val_accuracy: 0.7975\n",
      "Epoch 910/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1417 - accuracy: 0.9932 - val_loss: 1.3516 - val_accuracy: 0.7862\n",
      "Epoch 911/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1405 - accuracy: 0.9935 - val_loss: 1.3511 - val_accuracy: 0.7975\n",
      "Epoch 912/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1392 - accuracy: 0.9945 - val_loss: 1.3306 - val_accuracy: 0.7950\n",
      "Epoch 913/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1413 - accuracy: 0.9938 - val_loss: 1.2960 - val_accuracy: 0.8012\n",
      "Epoch 914/1000\n",
      "19/19 [==============================] - 1s 31ms/step - loss: 0.1379 - accuracy: 0.9939 - val_loss: 1.2804 - val_accuracy: 0.8050\n",
      "Epoch 915/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1394 - accuracy: 0.9931 - val_loss: 1.2850 - val_accuracy: 0.8087\n",
      "Epoch 916/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1383 - accuracy: 0.9940 - val_loss: 1.3185 - val_accuracy: 0.8012\n",
      "Epoch 917/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1410 - accuracy: 0.9929 - val_loss: 1.3262 - val_accuracy: 0.7937\n",
      "Epoch 918/1000\n",
      "19/19 [==============================] - 1s 54ms/step - loss: 0.1355 - accuracy: 0.9950 - val_loss: 1.3618 - val_accuracy: 0.7962\n",
      "Epoch 919/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1399 - accuracy: 0.9941 - val_loss: 1.3299 - val_accuracy: 0.8000\n",
      "Epoch 920/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1368 - accuracy: 0.9948 - val_loss: 1.3243 - val_accuracy: 0.7950\n",
      "Epoch 921/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1394 - accuracy: 0.9947 - val_loss: 1.2991 - val_accuracy: 0.8025\n",
      "Epoch 922/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1375 - accuracy: 0.9945 - val_loss: 1.3045 - val_accuracy: 0.8062\n",
      "Epoch 923/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1388 - accuracy: 0.9940 - val_loss: 1.3639 - val_accuracy: 0.7900\n",
      "Epoch 924/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1376 - accuracy: 0.9936 - val_loss: 1.3718 - val_accuracy: 0.7975\n",
      "Epoch 925/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1358 - accuracy: 0.9943 - val_loss: 1.3590 - val_accuracy: 0.7887\n",
      "Epoch 926/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1372 - accuracy: 0.9953 - val_loss: 1.3638 - val_accuracy: 0.7950\n",
      "Epoch 927/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1390 - accuracy: 0.9941 - val_loss: 1.3359 - val_accuracy: 0.7900\n",
      "Epoch 928/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1371 - accuracy: 0.9949 - val_loss: 1.3442 - val_accuracy: 0.7925\n",
      "Epoch 929/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1376 - accuracy: 0.9949 - val_loss: 1.3167 - val_accuracy: 0.7925\n",
      "Epoch 930/1000\n",
      "19/19 [==============================] - 1s 46ms/step - loss: 0.1345 - accuracy: 0.9951 - val_loss: 1.3030 - val_accuracy: 0.7975\n",
      "Epoch 931/1000\n",
      "19/19 [==============================] - 1s 47ms/step - loss: 0.1337 - accuracy: 0.9957 - val_loss: 1.3370 - val_accuracy: 0.8000\n",
      "Epoch 932/1000\n",
      "19/19 [==============================] - 1s 26ms/step - loss: 0.1386 - accuracy: 0.9943 - val_loss: 1.3449 - val_accuracy: 0.7962\n",
      "Epoch 933/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1388 - accuracy: 0.9944 - val_loss: 1.3610 - val_accuracy: 0.7900\n",
      "Epoch 934/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1365 - accuracy: 0.9945 - val_loss: 1.3554 - val_accuracy: 0.7975\n",
      "Epoch 935/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1401 - accuracy: 0.9936 - val_loss: 1.3157 - val_accuracy: 0.7987\n",
      "Epoch 936/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1383 - accuracy: 0.9941 - val_loss: 1.3471 - val_accuracy: 0.7937\n",
      "Epoch 937/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1395 - accuracy: 0.9936 - val_loss: 1.3323 - val_accuracy: 0.7962\n",
      "Epoch 938/1000\n",
      "19/19 [==============================] - 1s 31ms/step - loss: 0.1385 - accuracy: 0.9931 - val_loss: 1.3485 - val_accuracy: 0.7862\n",
      "Epoch 939/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1403 - accuracy: 0.9939 - val_loss: 1.3527 - val_accuracy: 0.7975\n",
      "Epoch 940/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1355 - accuracy: 0.9950 - val_loss: 1.3531 - val_accuracy: 0.7962\n",
      "Epoch 941/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1384 - accuracy: 0.9939 - val_loss: 1.3463 - val_accuracy: 0.7950\n",
      "Epoch 942/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1385 - accuracy: 0.9939 - val_loss: 1.3622 - val_accuracy: 0.8000\n",
      "Epoch 943/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1387 - accuracy: 0.9940 - val_loss: 1.3861 - val_accuracy: 0.7900\n",
      "Epoch 944/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1402 - accuracy: 0.9938 - val_loss: 1.3852 - val_accuracy: 0.7912\n",
      "Epoch 945/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1390 - accuracy: 0.9942 - val_loss: 1.3744 - val_accuracy: 0.7912\n",
      "Epoch 946/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1375 - accuracy: 0.9939 - val_loss: 1.4171 - val_accuracy: 0.7862\n",
      "Epoch 947/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1400 - accuracy: 0.9929 - val_loss: 1.3725 - val_accuracy: 0.7887\n",
      "Epoch 948/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1412 - accuracy: 0.9931 - val_loss: 1.3096 - val_accuracy: 0.8000\n",
      "Epoch 949/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1404 - accuracy: 0.9940 - val_loss: 1.3720 - val_accuracy: 0.7962\n",
      "Epoch 950/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1398 - accuracy: 0.9949 - val_loss: 1.3776 - val_accuracy: 0.7962\n",
      "Epoch 951/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1393 - accuracy: 0.9936 - val_loss: 1.3677 - val_accuracy: 0.7875\n",
      "Epoch 952/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1376 - accuracy: 0.9941 - val_loss: 1.3440 - val_accuracy: 0.7937\n",
      "Epoch 953/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1410 - accuracy: 0.9933 - val_loss: 1.3103 - val_accuracy: 0.7925\n",
      "Epoch 954/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1437 - accuracy: 0.9941 - val_loss: 1.3090 - val_accuracy: 0.7937\n",
      "Epoch 955/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1383 - accuracy: 0.9942 - val_loss: 1.3361 - val_accuracy: 0.8025\n",
      "Epoch 956/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1380 - accuracy: 0.9947 - val_loss: 1.3159 - val_accuracy: 0.8025\n",
      "Epoch 957/1000\n",
      "19/19 [==============================] - 1s 33ms/step - loss: 0.1364 - accuracy: 0.9941 - val_loss: 1.3068 - val_accuracy: 0.7962\n",
      "Epoch 958/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1408 - accuracy: 0.9938 - val_loss: 1.3115 - val_accuracy: 0.7912\n",
      "Epoch 959/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1372 - accuracy: 0.9945 - val_loss: 1.3246 - val_accuracy: 0.7950\n",
      "Epoch 960/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1372 - accuracy: 0.9949 - val_loss: 1.3388 - val_accuracy: 0.7975\n",
      "Epoch 961/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1380 - accuracy: 0.9939 - val_loss: 1.3616 - val_accuracy: 0.7912\n",
      "Epoch 962/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1392 - accuracy: 0.9942 - val_loss: 1.3386 - val_accuracy: 0.7962\n",
      "Epoch 963/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1366 - accuracy: 0.9942 - val_loss: 1.3744 - val_accuracy: 0.7962\n",
      "Epoch 964/1000\n",
      "19/19 [==============================] - 1s 30ms/step - loss: 0.1388 - accuracy: 0.9941 - val_loss: 1.3706 - val_accuracy: 0.7987\n",
      "Epoch 965/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1346 - accuracy: 0.9949 - val_loss: 1.3844 - val_accuracy: 0.7925\n",
      "Epoch 966/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1399 - accuracy: 0.9936 - val_loss: 1.3716 - val_accuracy: 0.7875\n",
      "Epoch 967/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1416 - accuracy: 0.9933 - val_loss: 1.3265 - val_accuracy: 0.7962\n",
      "Epoch 968/1000\n",
      "19/19 [==============================] - 1s 25ms/step - loss: 0.1384 - accuracy: 0.9936 - val_loss: 1.3391 - val_accuracy: 0.7937\n",
      "Epoch 969/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1405 - accuracy: 0.9936 - val_loss: 1.3188 - val_accuracy: 0.7925\n",
      "Epoch 970/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1398 - accuracy: 0.9937 - val_loss: 1.3180 - val_accuracy: 0.7975\n",
      "Epoch 971/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1421 - accuracy: 0.9940 - val_loss: 1.3159 - val_accuracy: 0.7925\n",
      "Epoch 972/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1374 - accuracy: 0.9941 - val_loss: 1.2843 - val_accuracy: 0.7900\n",
      "Epoch 973/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1406 - accuracy: 0.9930 - val_loss: 1.3188 - val_accuracy: 0.7962\n",
      "Epoch 974/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1402 - accuracy: 0.9938 - val_loss: 1.3804 - val_accuracy: 0.7937\n",
      "Epoch 975/1000\n",
      "19/19 [==============================] - 0s 24ms/step - loss: 0.1410 - accuracy: 0.9936 - val_loss: 1.3486 - val_accuracy: 0.7912\n",
      "Epoch 976/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1401 - accuracy: 0.9934 - val_loss: 1.3641 - val_accuracy: 0.7900\n",
      "Epoch 977/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1402 - accuracy: 0.9939 - val_loss: 1.3848 - val_accuracy: 0.7875\n",
      "Epoch 978/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1416 - accuracy: 0.9929 - val_loss: 1.3573 - val_accuracy: 0.7900\n",
      "Epoch 979/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1401 - accuracy: 0.9939 - val_loss: 1.3681 - val_accuracy: 0.7850\n",
      "Epoch 980/1000\n",
      "19/19 [==============================] - 1s 33ms/step - loss: 0.1398 - accuracy: 0.9947 - val_loss: 1.3634 - val_accuracy: 0.7950\n",
      "Epoch 981/1000\n",
      "19/19 [==============================] - 1s 26ms/step - loss: 0.1354 - accuracy: 0.9956 - val_loss: 1.3698 - val_accuracy: 0.8000\n",
      "Epoch 982/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1368 - accuracy: 0.9942 - val_loss: 1.3181 - val_accuracy: 0.8025\n",
      "Epoch 983/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1368 - accuracy: 0.9944 - val_loss: 1.3045 - val_accuracy: 0.8037\n",
      "Epoch 984/1000\n",
      "19/19 [==============================] - 1s 28ms/step - loss: 0.1352 - accuracy: 0.9951 - val_loss: 1.2986 - val_accuracy: 0.7975\n",
      "Epoch 985/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1365 - accuracy: 0.9945 - val_loss: 1.3053 - val_accuracy: 0.8025\n",
      "Epoch 986/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1358 - accuracy: 0.9940 - val_loss: 1.3206 - val_accuracy: 0.8050\n",
      "Epoch 987/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1369 - accuracy: 0.9951 - val_loss: 1.3197 - val_accuracy: 0.8037\n",
      "Epoch 988/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1356 - accuracy: 0.9948 - val_loss: 1.3505 - val_accuracy: 0.8000\n",
      "Epoch 989/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1363 - accuracy: 0.9943 - val_loss: 1.3810 - val_accuracy: 0.7962\n",
      "Epoch 990/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1369 - accuracy: 0.9945 - val_loss: 1.3069 - val_accuracy: 0.8012\n",
      "Epoch 991/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1344 - accuracy: 0.9951 - val_loss: 1.3168 - val_accuracy: 0.8025\n",
      "Epoch 992/1000\n",
      "19/19 [==============================] - 1s 27ms/step - loss: 0.1337 - accuracy: 0.9951 - val_loss: 1.3218 - val_accuracy: 0.8025\n",
      "Epoch 993/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1344 - accuracy: 0.9957 - val_loss: 1.3451 - val_accuracy: 0.7987\n",
      "Epoch 994/1000\n",
      "19/19 [==============================] - 0s 26ms/step - loss: 0.1349 - accuracy: 0.9946 - val_loss: 1.3742 - val_accuracy: 0.7900\n",
      "Epoch 995/1000\n",
      "19/19 [==============================] - 1s 29ms/step - loss: 0.1368 - accuracy: 0.9943 - val_loss: 1.3723 - val_accuracy: 0.7962\n",
      "Epoch 996/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1361 - accuracy: 0.9939 - val_loss: 1.3467 - val_accuracy: 0.8000\n",
      "Epoch 997/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1339 - accuracy: 0.9951 - val_loss: 1.3616 - val_accuracy: 0.7950\n",
      "Epoch 998/1000\n",
      "19/19 [==============================] - 1s 47ms/step - loss: 0.1330 - accuracy: 0.9948 - val_loss: 1.3549 - val_accuracy: 0.7925\n",
      "Epoch 999/1000\n",
      "19/19 [==============================] - 1s 31ms/step - loss: 0.1351 - accuracy: 0.9947 - val_loss: 1.3498 - val_accuracy: 0.7962\n",
      "Epoch 1000/1000\n",
      "19/19 [==============================] - 0s 25ms/step - loss: 0.1347 - accuracy: 0.9953 - val_loss: 1.3221 - val_accuracy: 0.7975\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<tensorflow.python.training.tracking.util.CheckpointLoadStatus at 0x2cef4e670>"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "history = model1.fit([fine_tune_data_resnet_train, fine_tune_data_vgg_train, fine_tune_data_inception_resnet_train], fine_tune_labels_train, epochs=EPOCHS, batch_size=800, validation_data=([fine_tune_data_resnet_test, fine_tune_data_vgg_test, fine_tune_data_inception_resnet_test], fine_tune_labels_test), validation_batch_size=800, callbacks=[model_checkpoint_callback])\n",
    "model1.save_weights(\"./Models/ResNet-VGG-Inception-4Piece-Indiv-Concat/resnet-vgg-inception-resnet-4p-model-weights\")\n",
    "model1.load_weights('./Models/ResNet-VGG-Inception-4Piece-Indiv-Concat/resnet-vgg-inception-resnet-4p-model-weights')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "id": "b1f883d7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<tensorflow.python.training.tracking.util.CheckpointLoadStatus at 0x36de26a30>"
      ]
     },
     "execution_count": 95,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model1.load_weights('./Models/ResNet-VGG-Inception-4Piece-Indiv-Concat/auto-ResNet-VGG-Inception-ResNet-FT-model-weight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "id": "9116a6a6",
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_train_loss(history, model_name):  \n",
    "    plt.figure(figsize=(8,6))\n",
    "    plt.plot(history.history['loss'])\n",
    "    plt.plot(history.history['accuracy'])\n",
    "    plt.plot(history.history['val_loss'])\n",
    "    plt.plot(history.history['val_accuracy'])\n",
    "    plt.title(model_name+' loss')\n",
    "    plt.ylabel('loss')\n",
    "    plt.xlabel('epochs')\n",
    "    plt.legend(['train loss', 'train accuracy', 'val loss', 'val accuracy'], loc='upper left')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "id": "c55b5870",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 800x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_train_loss(history, 'Fine-Tuned ResNet-VGG-Inception-ResNet-4Piece Model')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "03390fa6",
   "metadata": {},
   "source": [
    "### Model Evaluation"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "48701107",
   "metadata": {},
   "source": [
    "#### On Train Data"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b0594069",
   "metadata": {},
   "source": [
    "Now, after the model is being trained, we will look at the performance of the model on the training set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "id": "cf7ded38",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 26/475 [>.............................] - ETA: 1s"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-10-09 18:07:15.867983: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:113] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "475/475 [==============================] - 2s 4ms/step\n"
     ]
    }
   ],
   "source": [
    "train_preds = model1.predict([fine_tune_data_resnet_train, fine_tune_data_vgg_train, fine_tune_data_inception_resnet_train])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "id": "f2a28269",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_pred_labels = np.argmax(train_preds, axis=-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "id": "768775b2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 1, 13, 13, ...,  3,  9,  9])"
      ]
     },
     "execution_count": 100,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_pred_labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "id": "eff52a3d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "15197"
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.sum(train_pred_labels == train_labels['label'])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3fd403fa",
   "metadata": {},
   "source": [
    "Looking at the misclassified images to get a better idea about what features the model is unable to capture."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "id": "8f594a33",
   "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>id</th>\n",
       "      <th>images</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3597</th>\n",
       "      <td>3781</td>\n",
       "      <td>./Data/train/train/3781.tif</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12255</th>\n",
       "      <td>12896</td>\n",
       "      <td>./Data/train/train/12896.tif</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14569</th>\n",
       "      <td>15333</td>\n",
       "      <td>./Data/train/train/15333.tif</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          id                        images  label\n",
       "3597    3781   ./Data/train/train/3781.tif      0\n",
       "12255  12896  ./Data/train/train/12896.tif      9\n",
       "14569  15333  ./Data/train/train/15333.tif      7"
      ]
     },
     "execution_count": 102,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mis_train_data = train_labels[train_pred_labels != train_labels['label']]\n",
    "mis_train_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "2ea8f949",
   "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>precision</th>\n",
       "      <th>recall</th>\n",
       "      <th>fscore</th>\n",
       "      <th>support</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.998978</td>\n",
       "      <td>0.999488</td>\n",
       "      <td>978</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>925</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>925</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>947</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.998926</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.999463</td>\n",
       "      <td>930</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>962</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>998</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.998905</td>\n",
       "      <td>0.999452</td>\n",
       "      <td>913</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1005</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.998951</td>\n",
       "      <td>0.999475</td>\n",
       "      <td>953</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>916</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>955</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0.997957</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.998978</td>\n",
       "      <td>977</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>943</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>928</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>945</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    precision    recall    fscore  support\n",
       "0    1.000000  0.998978  0.999488      978\n",
       "1    1.000000  1.000000  1.000000      925\n",
       "2    1.000000  1.000000  1.000000      925\n",
       "3    1.000000  1.000000  1.000000      947\n",
       "4    0.998926  1.000000  0.999463      930\n",
       "5    1.000000  1.000000  1.000000      962\n",
       "6    1.000000  1.000000  1.000000      998\n",
       "7    1.000000  0.998905  0.999452      913\n",
       "8    1.000000  1.000000  1.000000     1005\n",
       "9    1.000000  0.998951  0.999475      953\n",
       "10   1.000000  1.000000  1.000000      916\n",
       "11   1.000000  1.000000  1.000000      955\n",
       "12   0.997957  1.000000  0.998978      977\n",
       "13   1.000000  1.000000  1.000000      943\n",
       "14   1.000000  1.000000  1.000000      928\n",
       "15   1.000000  1.000000  1.000000      945"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "precision, recall, fscore, support = score(train_labels['label'], train_pred_labels)\n",
    "pd.DataFrame({\n",
    "    'precision': precision,\n",
    "    'recall': recall,\n",
    "    'fscore': fscore,\n",
    "    'support': support\n",
    "})"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "659ae313",
   "metadata": {},
   "source": [
    "#### On Train-Test Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "id": "0d1e8f85",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "25/25 [==============================] - 0s 5ms/step\n"
     ]
    }
   ],
   "source": [
    "test_preds = model1.predict([fine_tune_data_resnet_test, fine_tune_data_vgg_test, fine_tune_data_inception_resnet_test])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "id": "49233509",
   "metadata": {},
   "outputs": [],
   "source": [
    "test_pred_labels = np.argmax(test_preds, axis=-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "id": "8f5cba39",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([11,  8,  6,  7, 10,  6, 13, 12, 14,  1,  6, 15,  7, 14,  1, 11,  0,\n",
       "        9,  6, 12,  8, 14,  3, 12,  6,  0,  3,  8,  3,  3, 11,  1,  9,  2,\n",
       "        1,  8,  4,  7,  1,  9,  6, 12, 14,  2,  5, 14,  3,  0, 13,  4,  0,\n",
       "       10, 12, 15, 11, 14, 15,  6,  9,  2, 15, 13, 11, 11,  6,  8,  8,  0,\n",
       "        6,  3,  4,  4, 12,  0,  8,  6,  7,  0, 12, 11, 11, 15, 12, 12,  9,\n",
       "        8,  1, 13, 13,  0, 10, 11,  4, 10,  1,  7, 11,  8, 13, 10,  3, 10,\n",
       "        5,  9,  7, 14,  8,  2,  1,  5,  7,  5,  9,  4, 12, 12, 12,  1, 11,\n",
       "        6, 13,  1,  0, 12,  1, 10, 12,  4, 15, 11,  3, 11,  6,  1,  6,  9,\n",
       "       14, 10,  7, 14, 10,  8, 11, 12,  8, 10,  8,  9,  2, 13, 15, 14,  4,\n",
       "       14,  4,  9,  4, 14,  7, 11,  8,  4,  4, 10, 13,  8, 15,  3,  8,  8,\n",
       "       13,  6,  3, 11,  7, 12, 10, 14, 14,  8,  6, 11, 13, 11,  6, 13,  4,\n",
       "       14, 11,  7, 15,  0,  3,  0, 10, 13, 12, 10, 11, 11, 11,  1,  9,  1,\n",
       "        9,  5, 13, 14, 12,  1,  4,  8, 11,  6, 13,  5,  4, 12,  8,  0,  9,\n",
       "        1, 15,  0,  7,  5,  2,  2,  6,  0,  7,  0,  3, 12,  3,  1,  2, 12,\n",
       "        4,  7, 12, 15, 12,  8, 13, 10,  0,  0, 11,  0, 11, 13, 15, 15,  8,\n",
       "        1, 10, 13,  5, 11,  9, 14,  6, 13, 12,  0,  1, 15, 11, 11, 14,  3,\n",
       "        0, 11,  4,  3,  1,  4,  6,  2,  7,  8, 10, 10,  8,  5,  7,  8, 13,\n",
       "        9,  8, 15,  0,  2,  3,  4, 10,  0,  8, 13,  0,  7, 15, 10, 10,  5,\n",
       "        3, 13,  3, 10,  9,  2,  5,  3, 15,  6, 15,  4, 14,  7, 13,  2, 15,\n",
       "        9, 15, 15,  5,  0, 15,  3, 11, 13, 15, 15,  1, 10,  4, 11, 15,  4,\n",
       "       13,  0,  3,  3,  2,  8,  9,  0, 11,  3,  3,  2, 12, 14,  4,  0,  3,\n",
       "        2, 12,  1,  3, 11,  5,  1,  3,  7, 15, 15,  3,  2, 12, 15,  3,  6,\n",
       "       15, 14, 14,  0,  5,  8, 14,  6, 12, 13,  6,  9,  8,  7, 10,  1,  2,\n",
       "       11, 14,  2, 14, 15,  5,  0,  1,  2, 12,  1,  0, 14,  6,  9,  0,  2,\n",
       "        7,  3, 14,  8, 13, 11,  4, 13, 13,  1,  4,  7, 13, 14,  5, 11,  8,\n",
       "        0,  1, 11,  0, 10,  3,  7,  2,  5,  6, 15,  2, 11, 15,  3, 11,  8,\n",
       "       12,  0,  3,  5,  5,  9,  6, 14,  7,  8,  3,  2, 13, 12,  1,  0, 11,\n",
       "        9,  8, 10, 14, 12,  0,  0,  1,  9,  9, 15, 12,  7, 14,  9, 15,  7,\n",
       "        5, 10,  4,  8,  6,  3,  7,  8,  6, 10,  1, 10,  7, 15, 11, 13, 10,\n",
       "        0,  3,  1, 10, 13,  7,  8,  5,  9,  1,  6,  7, 15, 14,  8,  7,  6,\n",
       "       12,  7,  0, 12, 11,  1, 10,  7,  9, 10, 14,  3,  1, 14,  6, 14,  7,\n",
       "        1, 14,  6, 14,  9, 13,  5, 14,  1,  5,  2,  3,  7,  2,  0,  8,  8,\n",
       "        0,  8,  6,  5, 14,  3,  6, 11,  8, 14,  6,  9,  1,  9, 11,  2,  0,\n",
       "        3, 10, 11, 10, 15,  2,  7,  5, 14,  9,  8,  1,  9, 13, 11, 10,  5,\n",
       "        9, 11,  6, 10, 12,  8,  5, 14, 14,  9,  0, 15,  8,  5,  8,  6,  5,\n",
       "       12,  5,  0, 15, 14,  5,  3,  2, 12,  1, 15, 11, 10,  4, 13,  1,  0,\n",
       "        1, 11, 14,  4,  4,  9, 11,  0, 14, 13,  4,  8,  5,  9, 13,  4, 15,\n",
       "       11,  0,  1,  5, 13,  7,  6, 12,  8,  7,  5, 15,  5,  9,  4,  7, 14,\n",
       "        1, 13,  6,  6,  3,  7,  7,  5,  5, 12,  6,  8,  0,  4,  7,  7,  0,\n",
       "       12, 14, 12,  0,  3, 11,  2,  3,  5,  4, 13,  1, 12,  0, 14, 15,  8,\n",
       "       11, 12, 10,  7,  2, 13, 12,  8, 11,  6,  9,  2, 10, 13,  7,  6,  3,\n",
       "       13,  3,  2, 14,  0,  8, 12, 12,  8, 13, 13, 14,  4, 10,  2,  5, 14,\n",
       "       12, 13, 13,  4, 12, 15, 10,  0,  8, 11,  1,  8,  4,  9,  2,  8,  0,\n",
       "        1,  2, 14, 15,  0,  3,  9, 11,  0,  8,  9,  7,  4, 11, 13, 11,  3,\n",
       "        7,  7, 13,  8, 12,  2, 12, 15,  7, 10, 14, 12, 12,  6, 14,  0,  7,\n",
       "        5,  3, 12, 13, 14,  7,  2,  8,  5, 13,  2,  9,  8, 14, 12, 14, 12,\n",
       "        3,  5,  8, 15,  4, 11, 14,  0,  9,  9,  5, 13, 10,  2,  2, 11, 15,\n",
       "        6])"
      ]
     },
     "execution_count": 105,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_pred_labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "id": "ee078822",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "634"
      ]
     },
     "execution_count": 106,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.sum(test_pred_labels == test_labels['label'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "c3db9b5c",
   "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>id</th>\n",
       "      <th>images</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>41</td>\n",
       "      <td>./Data/train/train/41.tif</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>65</td>\n",
       "      <td>./Data/train/train/65.tif</td>\n",
       "      <td>14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>67</td>\n",
       "      <td>./Data/train/train/67.tif</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>100</td>\n",
       "      <td>./Data/train/train/100.tif</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>372</td>\n",
       "      <td>./Data/train/train/372.tif</td>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>769</th>\n",
       "      <td>15453</td>\n",
       "      <td>./Data/train/train/15453.tif</td>\n",
       "      <td>14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>776</th>\n",
       "      <td>15528</td>\n",
       "      <td>./Data/train/train/15528.tif</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>779</th>\n",
       "      <td>15568</td>\n",
       "      <td>./Data/train/train/15568.tif</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>785</th>\n",
       "      <td>15621</td>\n",
       "      <td>./Data/train/train/15621.tif</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>788</th>\n",
       "      <td>15687</td>\n",
       "      <td>./Data/train/train/15687.tif</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>149 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        id                        images  label\n",
       "3       41     ./Data/train/train/41.tif      1\n",
       "4       65     ./Data/train/train/65.tif     14\n",
       "5       67     ./Data/train/train/67.tif      9\n",
       "7      100    ./Data/train/train/100.tif     10\n",
       "19     372    ./Data/train/train/372.tif     13\n",
       "..     ...                           ...    ...\n",
       "769  15453  ./Data/train/train/15453.tif     14\n",
       "776  15528  ./Data/train/train/15528.tif      5\n",
       "779  15568  ./Data/train/train/15568.tif      9\n",
       "785  15621  ./Data/train/train/15621.tif     15\n",
       "788  15687  ./Data/train/train/15687.tif     12\n",
       "\n",
       "[149 rows x 3 columns]"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mis_train_data = test_labels[test_pred_labels != test_labels['label']]\n",
    "mis_train_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "b538b721",
   "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>precision</th>\n",
       "      <th>recall</th>\n",
       "      <th>fscore</th>\n",
       "      <th>support</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.816327</td>\n",
       "      <td>0.740741</td>\n",
       "      <td>0.776699</td>\n",
       "      <td>54</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.638298</td>\n",
       "      <td>0.681818</td>\n",
       "      <td>0.659341</td>\n",
       "      <td>44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.974359</td>\n",
       "      <td>0.883721</td>\n",
       "      <td>0.926829</td>\n",
       "      <td>43</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.843137</td>\n",
       "      <td>0.934783</td>\n",
       "      <td>0.886598</td>\n",
       "      <td>46</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.853659</td>\n",
       "      <td>0.972222</td>\n",
       "      <td>0.909091</td>\n",
       "      <td>36</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.756757</td>\n",
       "      <td>0.560000</td>\n",
       "      <td>0.643678</td>\n",
       "      <td>50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.836735</td>\n",
       "      <td>0.891304</td>\n",
       "      <td>0.863158</td>\n",
       "      <td>46</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.846154</td>\n",
       "      <td>0.846154</td>\n",
       "      <td>0.846154</td>\n",
       "      <td>52</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.828125</td>\n",
       "      <td>0.981481</td>\n",
       "      <td>0.898305</td>\n",
       "      <td>54</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.844444</td>\n",
       "      <td>0.808511</td>\n",
       "      <td>0.826087</td>\n",
       "      <td>47</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>0.807692</td>\n",
       "      <td>0.724138</td>\n",
       "      <td>0.763636</td>\n",
       "      <td>58</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>0.779661</td>\n",
       "      <td>0.821429</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>56</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0.653846</td>\n",
       "      <td>0.790698</td>\n",
       "      <td>0.715789</td>\n",
       "      <td>43</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>0.868852</td>\n",
       "      <td>0.828125</td>\n",
       "      <td>0.848000</td>\n",
       "      <td>64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>0.896552</td>\n",
       "      <td>0.812500</td>\n",
       "      <td>0.852459</td>\n",
       "      <td>64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>0.772727</td>\n",
       "      <td>0.790698</td>\n",
       "      <td>0.781609</td>\n",
       "      <td>43</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    precision    recall    fscore  support\n",
       "0    0.816327  0.740741  0.776699       54\n",
       "1    0.638298  0.681818  0.659341       44\n",
       "2    0.974359  0.883721  0.926829       43\n",
       "3    0.843137  0.934783  0.886598       46\n",
       "4    0.853659  0.972222  0.909091       36\n",
       "5    0.756757  0.560000  0.643678       50\n",
       "6    0.836735  0.891304  0.863158       46\n",
       "7    0.846154  0.846154  0.846154       52\n",
       "8    0.828125  0.981481  0.898305       54\n",
       "9    0.844444  0.808511  0.826087       47\n",
       "10   0.807692  0.724138  0.763636       58\n",
       "11   0.779661  0.821429  0.800000       56\n",
       "12   0.653846  0.790698  0.715789       43\n",
       "13   0.868852  0.828125  0.848000       64\n",
       "14   0.896552  0.812500  0.852459       64\n",
       "15   0.772727  0.790698  0.781609       43"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "precision, recall, fscore, support = score(test_labels['label'], test_pred_labels)\n",
    "pd.DataFrame({\n",
    "    'precision': precision,\n",
    "    'recall': recall,\n",
    "    'fscore': fscore,\n",
    "    'support': support\n",
    "})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "ed4db76c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8123396122618025"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean(fscore)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "067c5096",
   "metadata": {},
   "source": [
    "#### On Validation Data"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3e665591",
   "metadata": {},
   "source": [
    "Generating Predictions for the Validation Set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "4b085c0c",
   "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>images</th>\n",
       "      <th>id</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>850</th>\n",
       "      <td>./Data/validation/validation/17801.tif</td>\n",
       "      <td>17801</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>898</th>\n",
       "      <td>./Data/validation/validation/17802.tif</td>\n",
       "      <td>17802</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>852</th>\n",
       "      <td>./Data/validation/validation/17803.tif</td>\n",
       "      <td>17803</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>798</th>\n",
       "      <td>./Data/validation/validation/17804.tif</td>\n",
       "      <td>17804</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>750</th>\n",
       "      <td>./Data/validation/validation/17805.tif</td>\n",
       "      <td>17805</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>246</th>\n",
       "      <td>./Data/validation/validation/18696.tif</td>\n",
       "      <td>18696</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>211</th>\n",
       "      <td>./Data/validation/validation/18697.tif</td>\n",
       "      <td>18697</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>760</th>\n",
       "      <td>./Data/validation/validation/18698.tif</td>\n",
       "      <td>18698</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>787</th>\n",
       "      <td>./Data/validation/validation/18699.tif</td>\n",
       "      <td>18699</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>891</th>\n",
       "      <td>./Data/validation/validation/18700.tif</td>\n",
       "      <td>18700</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>900 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                     images     id  label\n",
       "850  ./Data/validation/validation/17801.tif  17801     -1\n",
       "898  ./Data/validation/validation/17802.tif  17802     -1\n",
       "852  ./Data/validation/validation/17803.tif  17803     -1\n",
       "798  ./Data/validation/validation/17804.tif  17804     -1\n",
       "750  ./Data/validation/validation/17805.tif  17805     -1\n",
       "..                                      ...    ...    ...\n",
       "246  ./Data/validation/validation/18696.tif  18696     -1\n",
       "211  ./Data/validation/validation/18697.tif  18697     -1\n",
       "760  ./Data/validation/validation/18698.tif  18698     -1\n",
       "787  ./Data/validation/validation/18699.tif  18699     -1\n",
       "891  ./Data/validation/validation/18700.tif  18700     -1\n",
       "\n",
       "[900 rows x 3 columns]"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "validation_data = pd.DataFrame({'images':['./Data/validation/validation/'+name for name in os.listdir('./Data/validation/validation/')]})\n",
    "validation_data['id'] = [name.split('.')[0] for name in os.listdir('./Data/validation/validation/')]\n",
    "validation_data['label'] = -1 # Simply added to prevent re-writing code\n",
    "validation_data.sort_values(by=['id'], inplace=True)\n",
    "validation_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 270,
   "id": "53635253",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(900, 5, 2048)"
      ]
     },
     "execution_count": 270,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fine_tune_data_validation_resnet = np.load('./Data/ResNet-4P-validation-precompute/whole_validation_data_precomp.npy')\n",
    "fine_tune_data_validation_resnet.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 271,
   "id": "8e78b6f1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(900, 5, 512)"
      ]
     },
     "execution_count": 271,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fine_tune_data_validation_vgg = np.load('./Data/VGG16-4P-validation-precompute/whole_validation_data_precomp.npy')\n",
    "fine_tune_data_validation_vgg.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 272,
   "id": "88e39e91",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(900, 5, 1536)"
      ]
     },
     "execution_count": 272,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fine_tune_data_validation_inception_resnet = np.load('./Data/Inception-ResNet-4P-validation-precompute/whole_validation_data_precomp.npy')\n",
    "fine_tune_data_validation_inception_resnet.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 273,
   "id": "13b10856",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(900, 5, 1300)"
      ]
     },
     "execution_count": 273,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fine_tune_data_validation_red_resnet = np.zeros((fine_tune_data_validation_resnet.shape[0], 5, N_COMPS_1))\n",
    "for i in range(5):\n",
    "    fine_tune_data_validation_red_resnet[:,i,:] = pcas_resnet[i].transform(fine_tune_data_validation_resnet[:,i,:])\n",
    "fine_tune_data_validation_resnet = fine_tune_data_validation_red_resnet\n",
    "del fine_tune_data_validation_red_resnet\n",
    "fine_tune_data_validation_resnet.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 274,
   "id": "b0a31a0b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(900, 5, 300)"
      ]
     },
     "execution_count": 274,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fine_tune_data_validation_red_vgg = np.zeros((fine_tune_data_validation_vgg.shape[0], 5, N_COMPS_2))\n",
    "for i in range(5):\n",
    "    fine_tune_data_validation_red_vgg[:,i,:] = pcas_vgg[i].transform(fine_tune_data_validation_vgg[:,i,:])\n",
    "fine_tune_data_validation_vgg = fine_tune_data_validation_red_vgg\n",
    "del fine_tune_data_validation_red_vgg\n",
    "fine_tune_data_validation_vgg.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 275,
   "id": "5e12894e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(900, 5, 900)"
      ]
     },
     "execution_count": 275,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fine_tune_data_validation_red_inception_resnet = np.zeros((fine_tune_data_validation_inception_resnet.shape[0], 5, N_COMPS_3))\n",
    "for i in range(5):\n",
    "    fine_tune_data_validation_red_inception_resnet[:,i,:] = pcas_inception_resnet[i].transform(fine_tune_data_validation_inception_resnet[:,i,:])\n",
    "fine_tune_data_validation_inception_resnet = fine_tune_data_validation_red_inception_resnet\n",
    "del fine_tune_data_validation_red_inception_resnet\n",
    "fine_tune_data_validation_inception_resnet.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 276,
   "id": "586ac8df",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(900, 5, 2500)"
      ]
     },
     "execution_count": 276,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fine_tune_data_validation = np.concatenate([fine_tune_data_validation_resnet, fine_tune_data_validation_vgg, fine_tune_data_validation_inception_resnet], axis=-1)\n",
    "fine_tune_data_validation.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 277,
   "id": "db7d1256",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "29/29 [==============================] - 0s 3ms/step\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-10-09 15:57:30.272902: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:113] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    }
   ],
   "source": [
    "valid_preds = model1.predict(fine_tune_data_validation)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 278,
   "id": "18b17d2e",
   "metadata": {},
   "outputs": [],
   "source": [
    "valid_pred_labels = np.argmax(valid_preds, axis=-1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2bece594",
   "metadata": {},
   "source": [
    "### Submission"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "304c47de",
   "metadata": {},
   "source": [
    "For Kaggle Submission"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 279,
   "id": "e7ca9d19",
   "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>id</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>17801</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>17802</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>17803</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>17804</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>17805</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>895</th>\n",
       "      <td>18696</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>896</th>\n",
       "      <td>18697</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>897</th>\n",
       "      <td>18698</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>898</th>\n",
       "      <td>18699</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>899</th>\n",
       "      <td>18700</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>900 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        id  label\n",
       "0    17801      4\n",
       "1    17802      6\n",
       "2    17803      4\n",
       "3    17804      8\n",
       "4    17805      3\n",
       "..     ...    ...\n",
       "895  18696      7\n",
       "896  18697     12\n",
       "897  18698      8\n",
       "898  18699     10\n",
       "899  18700      4\n",
       "\n",
       "[900 rows x 2 columns]"
      ]
     },
     "execution_count": 279,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "submission_data = pd.DataFrame({'id':list(range(17801, 18701))})\n",
    "submission_data['label'] = valid_pred_labels\n",
    "submission_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 280,
   "id": "6eea7d9c",
   "metadata": {},
   "outputs": [],
   "source": [
    "submission_data.to_csv('./Submissions/submission-resnet-vgg-inception-4p.csv', index=False)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.8.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}