driver-distraction.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "6590bbdd",
   "metadata": {
    "_cell_guid": "b1076dfc-b9ad-4769-8c92-a6c4dae69d19",
    "_uuid": "8f2839f25d086af736a60e9eeb907d3b93b6e0e5",
    "execution": {
     "iopub.execute_input": "2023-01-28T18:25:46.238841Z",
     "iopub.status.busy": "2023-01-28T18:25:46.237967Z",
     "iopub.status.idle": "2023-01-28T18:25:51.592429Z",
     "shell.execute_reply": "2023-01-28T18:25:51.591290Z"
    },
    "papermill": {
     "duration": 5.363296,
     "end_time": "2023-01-28T18:25:51.595406",
     "exception": false,
     "start_time": "2023-01-28T18:25:46.232110",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "import numpy as np # linear algebra\n",
    "import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\n",
    "import matplotlib.pyplot as plt\n",
    "import os\n",
    "from PIL import Image\n",
    "from IPython.display import display \n",
    "import random\n",
    "import tensorflow as tf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "5c42c782",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-01-28T18:25:51.604214Z",
     "iopub.status.busy": "2023-01-28T18:25:51.603125Z",
     "iopub.status.idle": "2023-01-28T18:25:51.659513Z",
     "shell.execute_reply": "2023-01-28T18:25:51.658475Z"
    },
    "papermill": {
     "duration": 0.063243,
     "end_time": "2023-01-28T18:25:51.662024",
     "exception": false,
     "start_time": "2023-01-28T18:25:51.598781",
     "status": "completed"
    },
    "tags": []
   },
   "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>subject</th>\n",
       "      <th>classname</th>\n",
       "      <th>img</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>p002</td>\n",
       "      <td>c0</td>\n",
       "      <td>img_44733.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>p002</td>\n",
       "      <td>c0</td>\n",
       "      <td>img_72999.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>p002</td>\n",
       "      <td>c0</td>\n",
       "      <td>img_25094.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>p002</td>\n",
       "      <td>c0</td>\n",
       "      <td>img_69092.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>p002</td>\n",
       "      <td>c0</td>\n",
       "      <td>img_92629.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22419</th>\n",
       "      <td>p081</td>\n",
       "      <td>c9</td>\n",
       "      <td>img_56936.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22420</th>\n",
       "      <td>p081</td>\n",
       "      <td>c9</td>\n",
       "      <td>img_46218.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22421</th>\n",
       "      <td>p081</td>\n",
       "      <td>c9</td>\n",
       "      <td>img_25946.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22422</th>\n",
       "      <td>p081</td>\n",
       "      <td>c9</td>\n",
       "      <td>img_67850.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22423</th>\n",
       "      <td>p081</td>\n",
       "      <td>c9</td>\n",
       "      <td>img_9684.jpg</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>22424 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      subject classname            img\n",
       "0        p002        c0  img_44733.jpg\n",
       "1        p002        c0  img_72999.jpg\n",
       "2        p002        c0  img_25094.jpg\n",
       "3        p002        c0  img_69092.jpg\n",
       "4        p002        c0  img_92629.jpg\n",
       "...       ...       ...            ...\n",
       "22419    p081        c9  img_56936.jpg\n",
       "22420    p081        c9  img_46218.jpg\n",
       "22421    p081        c9  img_25946.jpg\n",
       "22422    p081        c9  img_67850.jpg\n",
       "22423    p081        c9   img_9684.jpg\n",
       "\n",
       "[22424 rows x 3 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "img_list = pd.read_csv('../input/state-farm-distracted-driver-detection/driver_imgs_list.csv')\n",
    "img_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "0568b98e",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-01-28T18:25:51.671445Z",
     "iopub.status.busy": "2023-01-28T18:25:51.669854Z",
     "iopub.status.idle": "2023-01-28T18:25:51.675569Z",
     "shell.execute_reply": "2023-01-28T18:25:51.674691Z"
    },
    "papermill": {
     "duration": 0.012335,
     "end_time": "2023-01-28T18:25:51.677678",
     "exception": false,
     "start_time": "2023-01-28T18:25:51.665343",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "img_directory = '../input/state-farm-distracted-driver-detection/imgs/train'\n",
    "img_directory_test = '../input/state-farm-distracted-driver-detection/imgs/test'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "548eec1c",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-01-28T18:25:51.686593Z",
     "iopub.status.busy": "2023-01-28T18:25:51.685110Z",
     "iopub.status.idle": "2023-01-28T18:26:12.480403Z",
     "shell.execute_reply": "2023-01-28T18:26:12.478832Z"
    },
    "papermill": {
     "duration": 20.80219,
     "end_time": "2023-01-28T18:26:12.483198",
     "exception": false,
     "start_time": "2023-01-28T18:25:51.681008",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 22424 files belonging to 10 classes.\n",
      "Using 17940 files for training.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-01-28 18:26:03.359553: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2023-01-28 18:26:03.360564: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2023-01-28 18:26:03.554284: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2023-01-28 18:26:03.555717: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2023-01-28 18:26:03.557022: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2023-01-28 18:26:03.558277: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2023-01-28 18:26:03.566871: I tensorflow/core/platform/cpu_feature_guard.cc:142] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 AVX512F FMA\n",
      "To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.\n",
      "2023-01-28 18:26:03.851661: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2023-01-28 18:26:03.852551: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2023-01-28 18:26:03.853325: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2023-01-28 18:26:03.854032: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2023-01-28 18:26:03.854709: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2023-01-28 18:26:03.855441: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2023-01-28 18:26:06.727957: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2023-01-28 18:26:06.728898: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2023-01-28 18:26:06.729589: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2023-01-28 18:26:06.730279: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2023-01-28 18:26:06.730956: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2023-01-28 18:26:06.731602: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1510] Created device /job:localhost/replica:0/task:0/device:GPU:0 with 13789 MB memory:  -> device: 0, name: Tesla T4, pci bus id: 0000:00:04.0, compute capability: 7.5\n",
      "2023-01-28 18:26:06.734823: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2023-01-28 18:26:06.735482: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1510] Created device /job:localhost/replica:0/task:0/device:GPU:1 with 13789 MB memory:  -> device: 1, name: Tesla T4, pci bus id: 0000:00:05.0, compute capability: 7.5\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 22424 files belonging to 10 classes.\n",
      "Using 4484 files for validation.\n"
     ]
    }
   ],
   "source": [
    "# 4:3 ratio images at 640x480, we'll scale it down to make it easier for the network\n",
    "train_ds = tf.keras.utils.image_dataset_from_directory(img_directory,validation_split=0.2 ,subset=\"training\",seed=123, image_size =(256,192), batch_size=32)\n",
    "val_ds = tf.keras.utils.image_dataset_from_directory(img_directory,validation_split=0.2 ,subset=\"validation\",seed=123, image_size =(256,192), batch_size=32)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "8ab7d991",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-01-28T18:26:12.493222Z",
     "iopub.status.busy": "2023-01-28T18:26:12.492382Z",
     "iopub.status.idle": "2023-01-28T18:26:12.502820Z",
     "shell.execute_reply": "2023-01-28T18:26:12.501778Z"
    },
    "papermill": {
     "duration": 0.017722,
     "end_time": "2023-01-28T18:26:12.505215",
     "exception": false,
     "start_time": "2023-01-28T18:26:12.487493",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "def create_model(input_dims, output_dims):\n",
    "    with tf.name_scope(\"cnn\"):\n",
    "        model = tf.keras.Sequential()\n",
    "        model.add(tf.keras.layers.InputLayer(input_shape=input_dims))\n",
    "        model.add(tf.keras.layers.Rescaling(1./255))\n",
    "        model.add(tf.keras.layers.Conv2D(32,3, activation=\"relu\"))\n",
    "        model.add(tf.keras.layers.MaxPooling2D())\n",
    "        model.add(tf.keras.layers.Conv2D(64,3, activation=\"relu\"))\n",
    "        model.add(tf.keras.layers.MaxPooling2D())\n",
    "        model.add(tf.keras.layers.Conv2D(128,3,activation=\"relu\"))\n",
    "        model.add(tf.keras.layers.Flatten())\n",
    "        \n",
    "        #fully connected layer\n",
    "        model.add(tf.keras.layers.Dense(128, activation=\"relu\"))\n",
    "        model.add(tf.keras.layers.Dense(64, activation=\"relu\"))\n",
    "        model.add(tf.keras.layers.Dense(output_dims, activation=\"softmax\"))\n",
    "        \n",
    "        model.compile(optimizer='adam', loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=False), metrics=['accuracy'])\n",
    "        \n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "36b94176",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-01-28T18:26:12.513699Z",
     "iopub.status.busy": "2023-01-28T18:26:12.513364Z",
     "iopub.status.idle": "2023-01-28T18:26:12.787799Z",
     "shell.execute_reply": "2023-01-28T18:26:12.786804Z"
    },
    "papermill": {
     "duration": 0.282143,
     "end_time": "2023-01-28T18:26:12.791009",
     "exception": false,
     "start_time": "2023-01-28T18:26:12.508866",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "model = create_model([256,192,3], 10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "dbaa59a5",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-01-28T18:26:12.800618Z",
     "iopub.status.busy": "2023-01-28T18:26:12.800227Z",
     "iopub.status.idle": "2023-01-28T18:26:12.807476Z",
     "shell.execute_reply": "2023-01-28T18:26:12.806351Z"
    },
    "papermill": {
     "duration": 0.015881,
     "end_time": "2023-01-28T18:26:12.811437",
     "exception": false,
     "start_time": "2023-01-28T18:26:12.795556",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "rescaling (Rescaling)        (None, 256, 192, 3)       0         \n",
      "_________________________________________________________________\n",
      "conv2d (Conv2D)              (None, 254, 190, 32)      896       \n",
      "_________________________________________________________________\n",
      "max_pooling2d (MaxPooling2D) (None, 127, 95, 32)       0         \n",
      "_________________________________________________________________\n",
      "conv2d_1 (Conv2D)            (None, 125, 93, 64)       18496     \n",
      "_________________________________________________________________\n",
      "max_pooling2d_1 (MaxPooling2 (None, 62, 46, 64)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_2 (Conv2D)            (None, 60, 44, 128)       73856     \n",
      "_________________________________________________________________\n",
      "flatten (Flatten)            (None, 337920)            0         \n",
      "_________________________________________________________________\n",
      "dense (Dense)                (None, 128)               43253888  \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 64)                8256      \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 10)                650       \n",
      "=================================================================\n",
      "Total params: 43,356,042\n",
      "Trainable params: 43,356,042\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "f580143a",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-01-28T18:26:12.820077Z",
     "iopub.status.busy": "2023-01-28T18:26:12.819788Z",
     "iopub.status.idle": "2023-01-28T18:54:19.188920Z",
     "shell.execute_reply": "2023-01-28T18:54:19.187863Z"
    },
    "papermill": {
     "duration": 1686.375859,
     "end_time": "2023-01-28T18:54:19.191062",
     "exception": false,
     "start_time": "2023-01-28T18:26:12.815203",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/25\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-01-28 18:26:13.250481: I tensorflow/compiler/mlir/mlir_graph_optimization_pass.cc:185] None of the MLIR Optimization Passes are enabled (registered 2)\n",
      "2023-01-28 18:26:15.599400: I tensorflow/stream_executor/cuda/cuda_dnn.cc:369] Loaded cuDNN version 8005\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "561/561 [==============================] - 94s 149ms/step - loss: 0.6724 - accuracy: 0.7827 - val_loss: 0.0514 - val_accuracy: 0.9880\n",
      "Epoch 2/25\n",
      "561/561 [==============================] - 56s 99ms/step - loss: 0.0335 - accuracy: 0.9906 - val_loss: 0.0500 - val_accuracy: 0.9860\n",
      "Epoch 3/25\n",
      "561/561 [==============================] - 54s 96ms/step - loss: 0.0320 - accuracy: 0.9909 - val_loss: 0.0442 - val_accuracy: 0.9882\n",
      "Epoch 4/25\n",
      "561/561 [==============================] - 53s 94ms/step - loss: 0.0166 - accuracy: 0.9947 - val_loss: 0.0558 - val_accuracy: 0.9877\n",
      "Epoch 5/25\n",
      "561/561 [==============================] - 57s 100ms/step - loss: 0.0088 - accuracy: 0.9974 - val_loss: 0.0246 - val_accuracy: 0.9949\n",
      "Epoch 6/25\n",
      "561/561 [==============================] - 57s 100ms/step - loss: 0.0187 - accuracy: 0.9944 - val_loss: 0.0490 - val_accuracy: 0.9873\n",
      "Epoch 7/25\n",
      "561/561 [==============================] - 56s 98ms/step - loss: 0.0092 - accuracy: 0.9976 - val_loss: 0.0546 - val_accuracy: 0.9886\n",
      "Epoch 8/25\n",
      "561/561 [==============================] - 56s 100ms/step - loss: 0.0141 - accuracy: 0.9963 - val_loss: 0.0410 - val_accuracy: 0.9900\n",
      "Epoch 9/25\n",
      "561/561 [==============================] - 57s 100ms/step - loss: 0.0124 - accuracy: 0.9967 - val_loss: 0.0548 - val_accuracy: 0.9895\n",
      "Epoch 10/25\n",
      "561/561 [==============================] - 57s 100ms/step - loss: 0.0077 - accuracy: 0.9979 - val_loss: 0.0435 - val_accuracy: 0.9904\n",
      "Epoch 11/25\n",
      "561/561 [==============================] - 56s 99ms/step - loss: 0.0064 - accuracy: 0.9983 - val_loss: 0.0335 - val_accuracy: 0.9924\n",
      "Epoch 12/25\n",
      "561/561 [==============================] - 54s 96ms/step - loss: 0.0042 - accuracy: 0.9990 - val_loss: 0.0473 - val_accuracy: 0.9906\n",
      "Epoch 13/25\n",
      "561/561 [==============================] - 54s 96ms/step - loss: 0.0131 - accuracy: 0.9956 - val_loss: 0.0499 - val_accuracy: 0.9891\n",
      "Epoch 14/25\n",
      "561/561 [==============================] - 54s 96ms/step - loss: 0.0053 - accuracy: 0.9980 - val_loss: 0.0469 - val_accuracy: 0.9902\n",
      "Epoch 15/25\n",
      "561/561 [==============================] - 54s 96ms/step - loss: 0.0035 - accuracy: 0.9991 - val_loss: 0.0330 - val_accuracy: 0.9958\n",
      "Epoch 16/25\n",
      "561/561 [==============================] - 57s 101ms/step - loss: 6.6821e-06 - accuracy: 1.0000 - val_loss: 0.0337 - val_accuracy: 0.9958\n",
      "Epoch 17/25\n",
      "561/561 [==============================] - 54s 95ms/step - loss: 2.8853e-06 - accuracy: 1.0000 - val_loss: 0.0340 - val_accuracy: 0.9958\n",
      "Epoch 18/25\n",
      "561/561 [==============================] - 54s 95ms/step - loss: 1.9916e-06 - accuracy: 1.0000 - val_loss: 0.0343 - val_accuracy: 0.9958\n",
      "Epoch 19/25\n",
      "561/561 [==============================] - 54s 95ms/step - loss: 1.4266e-06 - accuracy: 1.0000 - val_loss: 0.0346 - val_accuracy: 0.9958\n",
      "Epoch 20/25\n",
      "561/561 [==============================] - 53s 93ms/step - loss: 1.0404e-06 - accuracy: 1.0000 - val_loss: 0.0350 - val_accuracy: 0.9958\n",
      "Epoch 21/25\n",
      "561/561 [==============================] - 54s 96ms/step - loss: 7.6486e-07 - accuracy: 1.0000 - val_loss: 0.0354 - val_accuracy: 0.9958\n",
      "Epoch 22/25\n",
      "561/561 [==============================] - 54s 96ms/step - loss: 5.6509e-07 - accuracy: 1.0000 - val_loss: 0.0358 - val_accuracy: 0.9958\n",
      "Epoch 23/25\n",
      "561/561 [==============================] - 53s 94ms/step - loss: 4.1890e-07 - accuracy: 1.0000 - val_loss: 0.0362 - val_accuracy: 0.9958\n",
      "Epoch 24/25\n",
      "561/561 [==============================] - 54s 94ms/step - loss: 3.1112e-07 - accuracy: 1.0000 - val_loss: 0.0366 - val_accuracy: 0.9958\n",
      "Epoch 25/25\n",
      "561/561 [==============================] - 55s 96ms/step - loss: 2.3078e-07 - accuracy: 1.0000 - val_loss: 0.0369 - val_accuracy: 0.9958\n"
     ]
    }
   ],
   "source": [
    "EPOCHS = 25\n",
    "\n",
    "history = model.fit(train_ds, epochs=EPOCHS, validation_data=val_ds)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "ae72f4e1",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-01-28T18:54:20.774794Z",
     "iopub.status.busy": "2023-01-28T18:54:20.774422Z",
     "iopub.status.idle": "2023-01-28T18:54:20.779876Z",
     "shell.execute_reply": "2023-01-28T18:54:20.778832Z"
    },
    "papermill": {
     "duration": 0.79387,
     "end_time": "2023-01-28T18:54:20.782048",
     "exception": false,
     "start_time": "2023-01-28T18:54:19.988178",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9891661119461059\n"
     ]
    }
   ],
   "source": [
    "print(sum(history.history['accuracy']) / EPOCHS)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "87b78cb6",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-01-28T18:54:22.512221Z",
     "iopub.status.busy": "2023-01-28T18:54:22.511867Z",
     "iopub.status.idle": "2023-01-28T18:54:22.747388Z",
     "shell.execute_reply": "2023-01-28T18:54:22.746532Z"
    },
    "papermill": {
     "duration": 1.243128,
     "end_time": "2023-01-28T18:54:22.749553",
     "exception": false,
     "start_time": "2023-01-28T18:54:21.506425",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7f1ea018aa50>]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "plt.plot(history.history['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "2617afc9",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-01-28T18:54:24.283730Z",
     "iopub.status.busy": "2023-01-28T18:54:24.283372Z",
     "iopub.status.idle": "2023-01-28T18:54:24.490233Z",
     "shell.execute_reply": "2023-01-28T18:54:24.489401Z"
    },
    "papermill": {
     "duration": 1.003913,
     "end_time": "2023-01-28T18:54:24.492097",
     "exception": false,
     "start_time": "2023-01-28T18:54:23.488184",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7f1ea0023150>]"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAD7CAYAAACRxdTpAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/NK7nSAAAACXBIWXMAAAsTAAALEwEAmpwYAAAY6UlEQVR4nO3df2zc933f8eeL94M8SiRPsmnrJy0nkys7iWzHnNNtRZetySavqJSgbSoDAxLsh1asWrNlGOpsgxt4GNZlW7r94Q1VuwDpsFT1stXVUK1q0abrOjSF6NixI8k/FM22SEsWLYmURJE88vjeH3ekjhR/HKWjTvf9vh7AQff93od376/PeN3nPvf9fj6KCMzMLFnaml2AmZk1nsPdzCyBHO5mZgnkcDczSyCHu5lZAjnczcwSqK5wl7RH0huSTkt6ZpHHf1nSK9Xbm5JGGl6pmZnVTSud5y4pA7wJfBoYBI4DT0fEySXa/wPg8Yj4Ww2u1czM6pSto82TwOmIOAMg6TCwD1g03IGngV9c6Unvvffe2LFjR51lmpkZwEsvvfRBRPSu1K6ecN8KnK3ZHgQ+sVhDSQ8ADwJ/uNKT7tixg4GBgTpe3szMZkl6p552jf5BdT/wrYgoL/agpAOSBiQNDA8PN/ilzcxsVj3hPgRsr9neVt23mP3Abyz1RBFxKCL6I6K/t3fFbxVmZnaL6gn348BOSQ9KylMJ8CMLG0naBWwA/rSxJZqZ2WqtGO4RMQ0cBI4Bp4AXIuKEpOck7a1puh84HJ5m0sys6er5QZWIOAocXbDv2QXbX2lcWWZmdjt8haqZWQI53M3MEqjlwv3425f46u++zsyMh/bNzJbScuH+vbMj/Mc/+gFXJ6ebXYqZ2V2r5cK9p5AD4Mr4VJMrMTO7e7VsuI9cd7ibmS2lZcN91D13M7MltVy4FzvzgMPdzGw5LRfuc8My46UmV2Jmdvdq2XB3z93MbGktF+4duTby2TaHu5nZMlou3CXRU8j5VEgzs2W0XLhDZWjGp0KamS2tZcPdwzJmZktryXAvOtzNzJbVkuHuYRkzs+W1ZLh3+wdVM7NltWS4FztzXJ2cpuxpf83MFtWS4e6ZIc3MltfS4T7icDczW1RLh7vPmDEzW1xd4S5pj6Q3JJ2W9MwSbT4n6aSkE5K+2dgy5yt2OtzNzJaTXamBpAzwPPBpYBA4LulIRJysabMT+DLwlyLisqT71qpgqF2wwzNDmpktpp6e+5PA6Yg4ExEl4DCwb0Gbvws8HxGXASLiQmPLnK/bP6iamS2rnnDfCpyt2R6s7qv1EPCQpP8r6TuS9jSqwMV4zN3MbHkrDsus4nl2Ap8EtgF/LOljETFS20jSAeAAQF9f3y2/WHs2QyGXcbibmS2hnp77ELC9ZntbdV+tQeBIRExFxP8D3qQS9vNExKGI6I+I/t7e3lutGfAUBGZmy6kn3I8DOyU9KCkP7AeOLGjzIpVeO5LupTJMc6ZxZd7MM0OamS1txXCPiGngIHAMOAW8EBEnJD0naW+12THgoqSTwLeBfxIRF9eqaICeToe7mdlS6hpzj4ijwNEF+56tuR/Al6q3O6KnkOPspet36uXMzFpKS16hCh6WMTNbTsuGuxfsMDNbWsuGe08hx/VSmdL0TLNLMTO767RuuHt+GTOzJbVuuPsqVTOzJTnczcwSKAHh7pkhzcwWSkC4u+duZrZQy4Z7sTMPwKjnlzEzu0nLhnt3R+Xi2tHx6SZXYmZ292nZcM9m2ljfnmXEY+5mZjdp2XAHT0FgZraUlg93L7VnZnazlg93L9hhZnazlg93D8uYmd2spcO96AU7zMwW1dLh7p67mdniWjrcuws5JqdnmJgqN7sUM7O7SkuHu6cgMDNbXEuHe9FzupuZLaqucJe0R9Ibkk5LemaRx78gaVjSK9Xb32l8qTeb7bn7dEgzs/myKzWQlAGeBz4NDALHJR2JiJMLmv5mRBxcgxqX5GEZM7PF1dNzfxI4HRFnIqIEHAb2rW1Z9XG4m5ktrp5w3wqcrdkerO5b6CclvSrpW5K2N6S6FRQL1Wl/He5mZvM06gfV/wnsiIjdwO8D31iskaQDkgYkDQwPD9/2i3Z1ZJFg9LpnhjQzq1VPuA8BtT3xbdV9cyLiYkRMVjd/DXhisSeKiEMR0R8R/b29vbdS7zxtbaKrPeueu5nZAvWE+3Fgp6QHJeWB/cCR2gaSNtds7gVONa7E5RU78w53M7MFVjxbJiKmJR0EjgEZ4OsRcULSc8BARBwBfl7SXmAauAR8YQ1rnqenkGPE4W5mNs+K4Q4QEUeBowv2PVtz/8vAlxtbWn08v4yZ2c1a+gpVcLibmS2m9cO906sxmZkt1PrhXl2NKSKaXYqZ2V0jEeE+PRNcL3naXzOzWS0f7kVPQWBmdpOWD3fPDGlmdrPEhLt77mZmN7R8uHc73M3MbtLy4T67GpNPhzQzu6Hlw31uzH3cM0Oamc1q+XBf354l0yYPy5iZ1Wj5cJfkKQjMzBZo+XCHG1epmplZRSLCvds9dzOzeRIR7j0FTx5mZlYrEeFedM/dzGyeRIS7V2MyM5svMeF+ZXyKmRlP+2tmBgkJ92JnjpmAa6XpZpdiZnZXSES4z80v49MhzcyAOsNd0h5Jb0g6LemZZdr9pKSQ1N+4ElfmmSHNzOZbMdwlZYDngaeAR4CnJT2ySLsu4IvAnzW6yJU43M3M5qun5/4kcDoizkRECTgM7Fuk3b8A/jUw0cD66jI7M6TD3cysop5w3wqcrdkerO6bI+njwPaI+J0G1lY3r8ZkZjbfbf+gKqkN+Brwj+toe0DSgKSB4eHh233pOR6WMTObr55wHwK212xvq+6b1QV8FPgjSW8DPwwcWexH1Yg4FBH9EdHf29t761UvUMhlyGfaHO5mZlX1hPtxYKekByXlgf3AkdkHI2I0Iu6NiB0RsQP4DrA3IgbWpOJFSKpOHuYFO8zMoI5wj4hp4CBwDDgFvBARJyQ9J2nvWhdYr55C1j13M7OqbD2NIuIocHTBvmeXaPvJ2y9r9bxgh5nZDYm4QhWg2Jl3uJuZVSUm3L0ak5nZDYkKd/fczcwqEhXuVyemKXvaXzOzZIU74OX2zMxIYLh7aMbMzOFuZpZIiQl3zwxpZnZDYsJ9bmZIh7uZWfLC3T13M7MEhXu3z5YxM5uTmHDvyGXoyLUxct0zQ5qZJSbcwVepmpnNcribmSVQosK9WPDMkGZmkLBw7/bMkGZmQMLCvaeQ89kyZmYkMNw9LGNmlrBwL3bmGCuVmSrPNLsUM7OmSlS4+ypVM7MKh7uZWQLVFe6S9kh6Q9JpSc8s8vjPSnpN0iuS/kTSI40vdWU9nhnSzAyoI9wlZYDngaeAR4CnFwnvb0bExyLiMeCrwNcaXWg95nruPh3SzFKunp77k8DpiDgTESXgMLCvtkFEXKnZXAc0ZSFTD8uYmVVk62izFThbsz0IfGJhI0k/B3wJyAN/tSHVrZLD3cysomE/qEbE8xHxYeAXgH++WBtJByQNSBoYHh5u1EvPmVuww8MyZpZy9YT7ELC9Zntbdd9SDgOfWeyBiDgUEf0R0d/b21t3kfXKZdpYl8+4525mqVdPuB8Hdkp6UFIe2A8cqW0gaWfN5o8DbzWuxNXxVapmZnWMuUfEtKSDwDEgA3w9Ik5Ieg4YiIgjwEFJnwKmgMvA59ey6OX0dHpmSDOzen5QJSKOAkcX7Hu25v4XG1zXLespZBkd92pMZpZuibpCFTwsY2YGDnczs0RKXLgXO/M+FdLMUi9x4d5TyDE5PcPEVLnZpZiZNU3iwr27eiGTV2QyszRLXLgXPQWBmVnywn1uCgKHu5mlWGLD3dP+mlmaJTfc3XM3sxRLXLgXOz0sY2aWuHDv6nDP3cwsceGeaRNdHVmfCmlmqZa4cIfK0Ix77maWZokM955CjpHrnhnSzNIrseHunruZpZnD3cwsgRIa7l6NyczSLaHhXum5R0SzSzEza4rEhvtUORj3tL9mllKJDPe5q1Q9v4yZpVRd4S5pj6Q3JJ2W9Mwij39J0klJr0r6A0kPNL7U+nl+GTNLuxXDXVIGeB54CngEeFrSIwuavQz0R8Ru4FvAVxtd6Go43M0s7erpuT8JnI6IMxFRAg4D+2obRMS3I+J6dfM7wLbGlrk6DnczS7t6wn0rcLZme7C6byl/G/hft1PU7fKc7maWdtlGPpmkvwn0A395iccPAAcA+vr6GvnS8/R0uuduZulWT899CNhes72tum8eSZ8C/hmwNyImF3uiiDgUEf0R0d/b23sr9dZlfT5LmxzuZpZe9YT7cWCnpAcl5YH9wJHaBpIeB36FSrBfaHyZq9PWpsrkYeOePMzM0mnFcI+IaeAgcAw4BbwQESckPSdpb7XZvwHWA/9N0iuSjizxdHdM5SrV6WaXYWbWFHWNuUfEUeDogn3P1tz/VIPrum2ePMzM0iyRV6gCdDvczSzFEhvuxc48o16ww8xSKrHh3lPIuuduZqmV4HCvDMvMzHjaXzNLn8SGe7GQZybgWslnzJhZ+iQ23D0FgZmlWWLDvduTh5lZiiU23Gd77lcc7maWQokN97nVmBzuZpZCiQ13z+luZmnmcDczS6DEhntnPkO2TV4k28xSKbHhLolip+eXMbN0Smy4Q+V0SJ8tY2ZplOhw97S/ZpZWiQ73oldjMrOUSnS4u+duZmmV/HD32TJmlkKJD/crE9OUPe2vmaVMssO9Mw/A1Qn33s0sXZId7r5K1cxSqq5wl7RH0huSTkt6ZpHHf1TSdyVNS/qpxpd5axzuZpZWK4a7pAzwPPAU8AjwtKRHFjR7F/gC8M1GF3g75maG9I+qZpYy2TraPAmcjogzAJIOA/uAk7MNIuLt6mMza1DjLXPP3czSqp5hma3A2Zrtweq+u57D3czS6o7+oCrpgKQBSQPDw8Nr/noOdzNLq3rCfQjYXrO9rbpv1SLiUET0R0R/b2/vrTzFqnTkMrRn2xzuZpY69YT7cWCnpAcl5YH9wJG1LatxfJWqmaXRiuEeEdPAQeAYcAp4ISJOSHpO0l4ASX9e0iDw08CvSDqxlkWvhueXMbM0qudsGSLiKHB0wb5na+4fpzJcc9cpdnpmSDNLn0RfoQqzPffpZpdhZnZHJT7cvRqTmaVR4sO9p5Bj5LqHZcwsXRIf7sVCnrFSmanyXXXxrJnZmkp8uPcUKr8Ze2jGzNIk+eHe6atUzSx9Eh/uxUJlwY4Rh7uZpUjiw73b88uYWQolPtxnJw/zmLuZpUlqwt0LdphZmqQm3D0sY2Zpkvhwz2fb6MxnHO5mliqJD3fwzJBmlj51zQrZ6noKOf7krQ/4R7/5Ch25NtqzGQr5DB3ZDB25trn77bk2CrkMHbkM69ozfGRLDx25TLPLNzNbtVSE+088uoUXXx5i4J1LjJdmmJwqMzFdZqocy/5dV3uWPR/dxGce38oPf+geMm26QxWbmd0eRSwfcGulv78/BgYGmvLas6bLM0xMzzAxVa65VbYvjZX4vZPv87vfP8+1yWnu725n76Nb2PfYVj6ypRvJQW9md56klyKif8V2aQ73ekxMlfmDUxf4rZeH+N9vXmCqHPy5+9bz2ce3svfRLWzf2NnsEs0sRRzua+DyWInfee0cv/3KEMffvgxA/wMb+MzjW/nxj21mw7r8mr7+pbESrw2Nsqm7g533rafNw0RmqeNwX2NnL13nyPfe48WXh3jrwjVyGfEXP3wvj27rYdfmbh7e3E3fxs5bHqePCH4wPMZL71zipXcuM/DOZc4Mj8093tWe5dHtRT7eV+Txvg083lek2Lm2Hy5m1nwO9zskIjh57govvjzEt98Y5szwNWaq/0kLuQwPberi4U1d7NrUxcObu9m1qXtupspaE1NlXh0c5aV3Ls8F+uXqVbXFzhxP9G3giR0beGxbkfNXJvjuu5d5+d0RXj9/lXL1BT907zoe6yvy8WrY/9D9XWQzN5/tOlWe4fzoBOdGJ3hvZJz3Rsd5b2SccyMTvFfdNz5VppDLVG75yhlEhZozizrymZse712fZ1NPgc09HWzu6WDjunzL/DZx8dokrw6O8urgKG9fHKNvYycPb+5i16bKh7S/JdndoqHhLmkP8B+ADPBrEfFLCx5vB34deAK4CPxMRLy93HMmJdwXmpgq89b71zh1/gqvn7vK6+evcOrclbmgBtjS08Guzd3s2tTFVHmGgXcu8/2h0bmzdz7Uu44n+jbQv2MDTzywkQ/3rlsyJK+Xpnl1cHQu7F9+9zIfXKusPNWZz7B7Ww8P3d/FxWslhkbGOTc6zoWrkyx823sKOTb3dLClWGBLsYN17Vkmp2YYL5UZn6rcJqbK87fnPXbzYij5bBubujvY1NPBlp6OueCvbBe4r7udjlyG9mwb7dm2O/ZBcGViiteqQf7q4AivDo4yNDIOgASbujt4/8rE3Id0Zz7DQ/d38fDmGx/QP7Spa+7qZ7M7qWHhLikDvAl8GhgEjgNPR8TJmjZ/H9gdET8raT/w2Yj4meWeN6nhvpiI4MLVSU6du8Lr56/y+rkrnDp3lR8MX6OtTeze2sMTOzbQ/8BGnnhgAxtvY+w+Ihi8PD4v7M98MEbv+na2FKu96mKBrcUONvdUgnxzT4F17bd3Vmx5Jrh4bZJz1W8E50fHa+5P8N7oOO9fmVj29NN8NeTbszcCP59to73mA6Ajl2F9e5bOfOXfddXb+vbM3P11+Szr2iuPF/IZ3r14ne9Vg/y1wVHOfHBjeKtvYycf29bDo9t62L2tyEe2dNPVkWO8VOatC1d5/dxVTp67Uv2QvjrvYritxQK7NnWxa3MXD9yzjnX5LIV8G4Vc5XU78ze+2cx+y/E3ALtdjQz3vwB8JSL+enX7ywAR8a9q2hyrtvlTSVngPNAbyzx5msJ9KZPTZYTIZ1NxoTAzM8HFsRLnqsF/4eokk1NlJqdnqrcyk1MzlMozTE5Vt6dnKE3fuD9eKjNWmmZsssy1yWlK0/Uvn3h/dzu7txXZvbWH3dsr/67mR/CI4P0rlQ/p2m9mPxgemxsaW0l7dTqMQi5DdyHHPevzbOjMc8+6PBvXtbNxfeX+hs4896zPs7F639dY2Kx6w72e7tpW4GzN9iDwiaXaRMS0pFHgHuCD+spNp/Zsuq5+bWsTvV3t9Ha1s3tbY55zqjzD2OQ01yanuV6qBP5Y9XZtssz10jSbujt4dHuR+7s7buu1JLGpOqz0V3bdN7d/YqrMhSuTjE9VXm+8ZvjqeqkynHW9VDOkVapsj46XuDRWYujyKBfHSlydmF7idSvDZkkL+eQcyer9/I/t5Cce3bKmr3FHr1CVdAA4ANDX13cnX9oSKpdpo9iZb+qZQh25DH333P71DqXpGUaul7g4Vgn92dvFsRKXx0pcvl666beSVhUk5EBu0Z34vaaecB8Cttdsb6vuW6zNYHVYpofKD6vzRMQh4BBUhmVupWCzpMpn27ivu4P7bvMbhhnUNyvkcWCnpAcl5YH9wJEFbY4An6/e/yngD5cbbzczs7W1Ys+9OoZ+EDhG5VTIr0fECUnPAQMRcQT4z8B/kXQauETlA8DMzJqkrjH3iDgKHF2w79ma+xPATze2NDMzu1XpOAfPzCxlHO5mZgnkcDczSyCHu5lZAjnczcwSqGlT/koaBt65xT+/l3RPbZDm40/zsUO6j9/HXvFARPSu9AdNC/fbIWmgnolzkirNx5/mY4d0H7+PfXXH7mEZM7MEcribmSVQq4b7oWYX0GRpPv40Hzuk+/h97KvQkmPuZma2vFbtuZuZ2TJaLtwl7ZH0hqTTkp5pdj13kqS3Jb0m6RVJiV+jUNLXJV2Q9P2afRsl/b6kt6r/bmhmjWtliWP/iqSh6vv/iqS/0cwa14qk7ZK+LemkpBOSvljdn5b3fqnjX9X731LDMvUs1p1kkt4G+iMiFef6SvpR4Brw6xHx0eq+rwKXIuKXqh/uGyLiF5pZ51pY4ti/AlyLiH/bzNrWmqTNwOaI+K6kLuAl4DPAF0jHe7/U8X+OVbz/rdZzfxI4HRFnIqIEHAb2NbkmWyMR8cdU1geotQ/4RvX+N6j8T584Sxx7KkTEuYj4bvX+VeAUlXWa0/LeL3X8q9Jq4b7YYt2rPugWFsDvSXqpuh5tGt0fEeeq988D9zezmCY4KOnV6rBNIoclaknaATwO/BkpfO8XHD+s4v1vtXBPux+JiI8DTwE/V/3qnlrVpRxbZ1zx9v0n4MPAY8A54N81tZo1Jmk98N+BfxgRV2ofS8N7v8jxr+r9b7Vwr2ex7sSKiKHqvxeA36IyTJU271fHJGfHJi80uZ47JiLej4hyRMwAv0qC339JOSrB9l8j4n9Ud6fmvV/s+Ff7/rdauNezWHciSVpX/XEFSeuAvwZ8f/m/SqTaxdg/D/x2E2u5o2aDreqzJPT9lyQq6zKfioiv1TyUivd+qeNf7fvfUmfLAFRP//n33Fis+182t6I7Q9KHqPTWobL27TeTfuySfgP4JJUZ8d4HfhF4EXgB6KMyq+jnIiJxPzwuceyfpPKVPIC3gb9XMwadGJJ+BPg/wGvATHX3P6Uy7pyG936p43+aVbz/LRfuZma2slYbljEzszo43M3MEsjhbmaWQA53M7MEcribmSWQw93MLIEc7mZmCeRwNzNLoP8PGIWc5uj3MAwAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history.history['loss'])"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.12"
  },
  "papermill": {
   "default_parameters": {},
   "duration": 1731.765155,
   "end_time": "2023-01-28T18:54:29.156889",
   "environment_variables": {},
   "exception": null,
   "input_path": "__notebook__.ipynb",
   "output_path": "__notebook__.ipynb",
   "parameters": {},
   "start_time": "2023-01-28T18:25:37.391734",
   "version": "2.3.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}