CatsDogsClassifier.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Cats vs. dogs classifier\n",
    "## Folder structure preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "nxzJ7fHo0Ap0"
   },
   "outputs": [],
   "source": [
    "!unzip kaggle_data.zip\n",
    "!rm kaggle_data.zip\n",
    "!mkdir kaggle_data\n",
    "!mv test.zip kaggle_data\n",
    "!mv train.zip kaggle_data\n",
    "!mv sample_submission.csv kaggle_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "hXvcS-C51H16",
    "outputId": "25891307-db92-42a1-f485-9d15e67c4dde"
   },
   "outputs": [],
   "source": [
    "%cd kaggle_data\n",
    "\n",
    "!unzip test.zip\n",
    "!unzip train.zip\n",
    "\n",
    "%cd .."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "CygFeNGFS7YE",
    "outputId": "9120d22d-9476-4089-b515-26289640267b"
   },
   "outputs": [],
   "source": [
    "!mkdir kaggle_data/input\n",
    "!mkdir kaggle_data/input/dogs\n",
    "!mkdir kaggle_data/input/cats\n",
    "!mv kaggle_data/test kaggle_data/input/dogs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "id": "q-AvvCkMUp7a"
   },
   "outputs": [],
   "source": [
    "import os\n",
    "import shutil\n",
    "import random\n",
    "\n",
    "def distribute_train_validation_split(validation_size=0.2):\n",
    "\n",
    "    all_images = os.listdir('./kaggle_data/train/')\n",
    "    random.shuffle(all_images)\n",
    "\n",
    "    all_dogs = list(filter(lambda image: 'dog' in image, all_images))\n",
    "    all_cats = list(filter(lambda image: 'cat' in image, all_images))\n",
    "\n",
    "    index_to_split = int(len(all_dogs) - len(all_dogs) * validation_size)\n",
    "    training_dogs = all_dogs[:index_to_split]\n",
    "    validation_dogs = all_dogs[index_to_split:]\n",
    "    training_cats = all_cats[:index_to_split]\n",
    "    validation_cats = all_cats[index_to_split:]\n",
    "\n",
    "    shutil.rmtree('./processed_data', ignore_errors=True)\n",
    "    os.makedirs('./processed_data/train/dogs/', exist_ok=True)\n",
    "    os.makedirs('./processed_data/train/cats/', exist_ok=True)\n",
    "    os.makedirs('./processed_data/validation/dogs/', exist_ok=True)\n",
    "    os.makedirs('./processed_data/validation/cats/', exist_ok=True)\n",
    "\n",
    "    copy_images_to_dir(training_dogs, './processed_data/train/dogs')\n",
    "    copy_images_to_dir(validation_dogs, './processed_data/validation/dogs')\n",
    "    copy_images_to_dir(training_cats, './processed_data/train/cats')\n",
    "    copy_images_to_dir(validation_cats, './processed_data/validation/cats')\n",
    "\n",
    "def copy_images_to_dir(images_to_copy, destination):\n",
    "    for image in images_to_copy:\n",
    "        shutil.copyfile(f'./kaggle_data/train/{image}', f'{destination}/{image}')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "id": "imaaJ18DUp7b"
   },
   "outputs": [],
   "source": [
    "distribute_train_validation_split(0.20)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Setting up environment"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%pip install scipy\n",
    "%pip install pandas"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "OMmYr5IxUp7c",
    "outputId": "dd9cf709-9869-47bc-ac59-f99e1f358d91"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[PhysicalDevice(name='/physical_device:GPU:0', device_type='GPU')]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
    "\n",
    "tf.config.list_physical_devices('GPU')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "physical_devices = tf.config.experimental.list_physical_devices('GPU')\n",
    "if len(physical_devices) > 0:\n",
    "    tf.config.experimental.set_memory_growth(physical_devices[0], True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "DIMENSIONS = 200"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "jFdt2-mSUp7c",
    "outputId": "b8dad32e-7a97-4e5c-e8f3-1d793ee381a2"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 20000 images belonging to 2 classes.\n",
      "Found 5000 images belonging to 2 classes.\n"
     ]
    }
   ],
   "source": [
    "train_imagedatagenerator = ImageDataGenerator(\n",
    "                            rescale=1. / 255,\n",
    "                            rotation_range=40,\n",
    "                            width_shift_range=0.2,\n",
    "                            height_shift_range=0.2,\n",
    "                            shear_range=0.2,\n",
    "                            zoom_range=0.2,\n",
    "                            horizontal_flip=True,\n",
    "                            fill_mode='nearest')\n",
    "validation_imagedatagenerator = ImageDataGenerator(rescale=1/255.0)\n",
    "\n",
    "train_iterator = train_imagedatagenerator.flow_from_directory(\n",
    "    './processed_data/train',\n",
    "    target_size=(DIMENSIONS, DIMENSIONS),\n",
    "    batch_size=200,\n",
    "    class_mode='binary')\n",
    "\n",
    "validation_iterator = validation_imagedatagenerator.flow_from_directory(\n",
    "    './processed_data/validation',\n",
    "    target_size=(DIMENSIONS, DIMENSIONS),\n",
    "    batch_size=50,\n",
    "    class_mode='binary')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "id": "60FiOfEzUp7c"
   },
   "outputs": [],
   "source": [
    "from tensorflow import keras\n",
    "\n",
    "model = keras.Sequential([\n",
    "    keras.layers.Conv2D(16, (3, 3), activation='relu', input_shape=(DIMENSIONS, DIMENSIONS, 3)),\n",
    "    keras.layers.MaxPool2D((2, 2)),\n",
    "    keras.layers.Conv2D(32, (3, 3), activation='relu'),\n",
    "    keras.layers.MaxPool2D((2, 2)),\n",
    "    keras.layers.Conv2D(64, (3, 3), activation='relu'),\n",
    "    keras.layers.MaxPool2D((2, 2)),\n",
    "    keras.layers.Flatten(),\n",
    "    keras.layers.Dense(1024, activation='relu'),\n",
    "    keras.layers.Dense(512, activation='relu'),\n",
    "    keras.layers.Dense(1, activation='sigmoid')\n",
    "])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "KBtVyfiiUp7c",
    "outputId": "1718f694-3e43-4ed1-f839-2a22d90d543d"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " conv2d (Conv2D)             (None, 198, 198, 16)      448       \n",
      "                                                                 \n",
      " max_pooling2d (MaxPooling2D  (None, 99, 99, 16)       0         \n",
      " )                                                               \n",
      "                                                                 \n",
      " conv2d_1 (Conv2D)           (None, 97, 97, 32)        4640      \n",
      "                                                                 \n",
      " max_pooling2d_1 (MaxPooling  (None, 48, 48, 32)       0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " conv2d_2 (Conv2D)           (None, 46, 46, 64)        18496     \n",
      "                                                                 \n",
      " max_pooling2d_2 (MaxPooling  (None, 23, 23, 64)       0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " flatten (Flatten)           (None, 33856)             0         \n",
      "                                                                 \n",
      " dense (Dense)               (None, 1024)              34669568  \n",
      "                                                                 \n",
      " dense_1 (Dense)             (None, 512)               524800    \n",
      "                                                                 \n",
      " dense_2 (Dense)             (None, 1)                 513       \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 35,218,465\n",
      "Trainable params: 35,218,465\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "Kg6NBLjZUp7c",
    "outputId": "7825f082-c226-43a5-a6d4-9bcc12ecc237"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "100/100 [==============================] - 171s 2s/step - loss: 0.8158 - accuracy: 0.5721 - val_loss: 0.6468 - val_accuracy: 0.6188\n",
      "Epoch 2/100\n",
      "100/100 [==============================] - 149s 1s/step - loss: 0.6187 - accuracy: 0.6583 - val_loss: 0.5715 - val_accuracy: 0.6982\n",
      "Epoch 3/100\n",
      "100/100 [==============================] - 174s 2s/step - loss: 0.5818 - accuracy: 0.6918 - val_loss: 0.5387 - val_accuracy: 0.7206\n",
      "Epoch 4/100\n",
      "100/100 [==============================] - 151s 2s/step - loss: 0.5623 - accuracy: 0.7070 - val_loss: 0.5245 - val_accuracy: 0.7300\n",
      "Epoch 5/100\n",
      "100/100 [==============================] - 146s 1s/step - loss: 0.5407 - accuracy: 0.7243 - val_loss: 0.5057 - val_accuracy: 0.7454\n",
      "Epoch 6/100\n",
      "100/100 [==============================] - 146s 1s/step - loss: 0.5384 - accuracy: 0.7266 - val_loss: 0.4984 - val_accuracy: 0.7472\n",
      "Epoch 7/100\n",
      "100/100 [==============================] - 146s 1s/step - loss: 0.5173 - accuracy: 0.7442 - val_loss: 0.4610 - val_accuracy: 0.7756\n",
      "Epoch 8/100\n",
      "100/100 [==============================] - 168s 2s/step - loss: 0.5116 - accuracy: 0.7430 - val_loss: 0.4564 - val_accuracy: 0.7738\n",
      "Epoch 9/100\n",
      "100/100 [==============================] - 158s 2s/step - loss: 0.5031 - accuracy: 0.7505 - val_loss: 0.4456 - val_accuracy: 0.7820\n",
      "Epoch 10/100\n",
      "100/100 [==============================] - 146s 1s/step - loss: 0.4888 - accuracy: 0.7605 - val_loss: 0.4468 - val_accuracy: 0.7856\n",
      "Epoch 11/100\n",
      "100/100 [==============================] - 146s 1s/step - loss: 0.4869 - accuracy: 0.7635 - val_loss: 0.4376 - val_accuracy: 0.7904\n",
      "Epoch 12/100\n",
      "100/100 [==============================] - 1490s 15s/step - loss: 0.4626 - accuracy: 0.7800 - val_loss: 0.4268 - val_accuracy: 0.7972\n",
      "Epoch 13/100\n",
      "100/100 [==============================] - 146s 1s/step - loss: 0.4650 - accuracy: 0.7789 - val_loss: 0.4182 - val_accuracy: 0.8046\n",
      "Epoch 14/100\n",
      "100/100 [==============================] - 170s 2s/step - loss: 0.4558 - accuracy: 0.7832 - val_loss: 0.4111 - val_accuracy: 0.8040\n",
      "Epoch 15/100\n",
      "100/100 [==============================] - 148s 1s/step - loss: 0.4426 - accuracy: 0.7905 - val_loss: 0.4037 - val_accuracy: 0.8078\n",
      "Epoch 16/100\n",
      "100/100 [==============================] - 146s 1s/step - loss: 0.4333 - accuracy: 0.7983 - val_loss: 0.3927 - val_accuracy: 0.8204\n",
      "Epoch 17/100\n",
      "100/100 [==============================] - 146s 1s/step - loss: 0.4338 - accuracy: 0.7997 - val_loss: 0.4712 - val_accuracy: 0.7694\n",
      "Epoch 18/100\n",
      "100/100 [==============================] - 146s 1s/step - loss: 0.4165 - accuracy: 0.8094 - val_loss: 0.4010 - val_accuracy: 0.8126\n",
      "Epoch 19/100\n",
      "100/100 [==============================] - 146s 1s/step - loss: 0.4201 - accuracy: 0.8044 - val_loss: 0.3948 - val_accuracy: 0.8120\n",
      "Epoch 20/100\n",
      "100/100 [==============================] - 931s 9s/step - loss: 0.4070 - accuracy: 0.8130 - val_loss: 0.3685 - val_accuracy: 0.8296\n",
      "Epoch 21/100\n",
      "100/100 [==============================] - 155s 2s/step - loss: 0.3976 - accuracy: 0.8185 - val_loss: 0.3892 - val_accuracy: 0.8290\n",
      "Epoch 22/100\n",
      "100/100 [==============================] - 147s 1s/step - loss: 0.4019 - accuracy: 0.8159 - val_loss: 0.3603 - val_accuracy: 0.8366\n",
      "Epoch 23/100\n",
      "100/100 [==============================] - 147s 1s/step - loss: 0.3819 - accuracy: 0.8296 - val_loss: 0.3673 - val_accuracy: 0.8302\n",
      "Epoch 24/100\n",
      "100/100 [==============================] - 145s 1s/step - loss: 0.3774 - accuracy: 0.8313 - val_loss: 0.3394 - val_accuracy: 0.8484\n",
      "Epoch 25/100\n",
      "100/100 [==============================] - 148s 1s/step - loss: 0.3674 - accuracy: 0.8360 - val_loss: 0.3253 - val_accuracy: 0.8600\n",
      "Epoch 26/100\n",
      "100/100 [==============================] - 146s 1s/step - loss: 0.3706 - accuracy: 0.8355 - val_loss: 0.3170 - val_accuracy: 0.8642\n",
      "Epoch 27/100\n",
      "100/100 [==============================] - 146s 1s/step - loss: 0.3569 - accuracy: 0.8400 - val_loss: 0.3226 - val_accuracy: 0.8546\n",
      "Epoch 28/100\n",
      "100/100 [==============================] - 146s 1s/step - loss: 0.3449 - accuracy: 0.8486 - val_loss: 0.3380 - val_accuracy: 0.8518\n",
      "Epoch 29/100\n",
      "100/100 [==============================] - 148s 1s/step - loss: 0.3497 - accuracy: 0.8426 - val_loss: 0.3215 - val_accuracy: 0.8596\n",
      "Epoch 30/100\n",
      "100/100 [==============================] - 146s 1s/step - loss: 0.3379 - accuracy: 0.8516 - val_loss: 0.3096 - val_accuracy: 0.8620\n",
      "Epoch 31/100\n",
      "100/100 [==============================] - 146s 1s/step - loss: 0.3387 - accuracy: 0.8519 - val_loss: 0.3026 - val_accuracy: 0.8690\n",
      "Epoch 32/100\n",
      "100/100 [==============================] - 146s 1s/step - loss: 0.3407 - accuracy: 0.8471 - val_loss: 0.3305 - val_accuracy: 0.8484\n",
      "Epoch 33/100\n",
      "100/100 [==============================] - 146s 1s/step - loss: 0.3257 - accuracy: 0.8558 - val_loss: 0.2965 - val_accuracy: 0.8756\n",
      "Epoch 34/100\n",
      "100/100 [==============================] - 146s 1s/step - loss: 0.3190 - accuracy: 0.8623 - val_loss: 0.2888 - val_accuracy: 0.8780\n",
      "Epoch 35/100\n",
      "100/100 [==============================] - 146s 1s/step - loss: 0.3133 - accuracy: 0.8630 - val_loss: 0.2749 - val_accuracy: 0.8808\n",
      "Epoch 36/100\n",
      "100/100 [==============================] - 146s 1s/step - loss: 0.3077 - accuracy: 0.8642 - val_loss: 0.3073 - val_accuracy: 0.8652\n",
      "Epoch 37/100\n",
      "100/100 [==============================] - 303s 3s/step - loss: 0.3053 - accuracy: 0.8687 - val_loss: 0.2694 - val_accuracy: 0.8860\n",
      "Epoch 38/100\n",
      "100/100 [==============================] - 146s 1s/step - loss: 0.2970 - accuracy: 0.8740 - val_loss: 0.2738 - val_accuracy: 0.8796\n",
      "Epoch 39/100\n",
      "100/100 [==============================] - 145s 1s/step - loss: 0.3003 - accuracy: 0.8684 - val_loss: 0.2715 - val_accuracy: 0.8876\n",
      "Epoch 40/100\n",
      "100/100 [==============================] - 146s 1s/step - loss: 0.3013 - accuracy: 0.8687 - val_loss: 0.2619 - val_accuracy: 0.8884\n",
      "Epoch 41/100\n",
      "100/100 [==============================] - 146s 1s/step - loss: 0.2935 - accuracy: 0.8722 - val_loss: 0.2821 - val_accuracy: 0.8856\n",
      "Epoch 42/100\n",
      "100/100 [==============================] - 146s 1s/step - loss: 0.2980 - accuracy: 0.8724 - val_loss: 0.2606 - val_accuracy: 0.8900\n",
      "Epoch 43/100\n",
      "100/100 [==============================] - 146s 1s/step - loss: 0.2848 - accuracy: 0.8785 - val_loss: 0.2630 - val_accuracy: 0.8936\n",
      "Epoch 44/100\n",
      "100/100 [==============================] - 146s 1s/step - loss: 0.2891 - accuracy: 0.8742 - val_loss: 0.2932 - val_accuracy: 0.8738\n",
      "Epoch 45/100\n",
      "100/100 [==============================] - 145s 1s/step - loss: 0.2868 - accuracy: 0.8749 - val_loss: 0.2612 - val_accuracy: 0.8894\n",
      "Epoch 46/100\n",
      "100/100 [==============================] - 336s 3s/step - loss: 0.2808 - accuracy: 0.8781 - val_loss: 0.2609 - val_accuracy: 0.8846\n",
      "Epoch 47/100\n",
      "100/100 [==============================] - 148s 1s/step - loss: 0.2758 - accuracy: 0.8809 - val_loss: 0.2682 - val_accuracy: 0.8912\n",
      "Epoch 48/100\n",
      "100/100 [==============================] - 146s 1s/step - loss: 0.2832 - accuracy: 0.8785 - val_loss: 0.2811 - val_accuracy: 0.8758\n",
      "Epoch 49/100\n",
      "100/100 [==============================] - 146s 1s/step - loss: 0.2767 - accuracy: 0.8794 - val_loss: 0.2462 - val_accuracy: 0.8958\n",
      "Epoch 50/100\n",
      "100/100 [==============================] - 146s 1s/step - loss: 0.2725 - accuracy: 0.8833 - val_loss: 0.2488 - val_accuracy: 0.8954\n",
      "Epoch 51/100\n",
      "100/100 [==============================] - 146s 1s/step - loss: 0.2680 - accuracy: 0.8843 - val_loss: 0.2303 - val_accuracy: 0.9040\n",
      "Epoch 52/100\n",
      "100/100 [==============================] - 151s 2s/step - loss: 0.2731 - accuracy: 0.8812 - val_loss: 0.2791 - val_accuracy: 0.8770\n",
      "Epoch 53/100\n",
      "100/100 [==============================] - 156s 2s/step - loss: 0.2636 - accuracy: 0.8856 - val_loss: 0.2277 - val_accuracy: 0.9024\n",
      "Epoch 54/100\n",
      "100/100 [==============================] - 152s 2s/step - loss: 0.2583 - accuracy: 0.8900 - val_loss: 0.2249 - val_accuracy: 0.9056\n",
      "Epoch 55/100\n",
      "100/100 [==============================] - 155s 2s/step - loss: 0.2635 - accuracy: 0.8859 - val_loss: 0.2376 - val_accuracy: 0.8998\n",
      "Epoch 56/100\n",
      "100/100 [==============================] - 155s 2s/step - loss: 0.2567 - accuracy: 0.8902 - val_loss: 0.2397 - val_accuracy: 0.8962\n",
      "Epoch 57/100\n",
      "100/100 [==============================] - 153s 2s/step - loss: 0.2607 - accuracy: 0.8885 - val_loss: 0.2465 - val_accuracy: 0.8922\n",
      "Epoch 58/100\n",
      "100/100 [==============================] - 154s 2s/step - loss: 0.2526 - accuracy: 0.8900 - val_loss: 0.2252 - val_accuracy: 0.9006\n",
      "Epoch 59/100\n",
      "100/100 [==============================] - 156s 2s/step - loss: 0.2590 - accuracy: 0.8872 - val_loss: 0.2349 - val_accuracy: 0.9026\n",
      "Epoch 60/100\n",
      "100/100 [==============================] - 153s 2s/step - loss: 0.2495 - accuracy: 0.8943 - val_loss: 0.2395 - val_accuracy: 0.8934\n",
      "Epoch 61/100\n",
      "100/100 [==============================] - 156s 2s/step - loss: 0.2450 - accuracy: 0.8963 - val_loss: 0.2870 - val_accuracy: 0.8718\n",
      "Epoch 62/100\n",
      "100/100 [==============================] - 154s 2s/step - loss: 0.2493 - accuracy: 0.8918 - val_loss: 0.3280 - val_accuracy: 0.8614\n",
      "Epoch 63/100\n",
      "100/100 [==============================] - 153s 2s/step - loss: 0.2591 - accuracy: 0.8896 - val_loss: 0.2241 - val_accuracy: 0.9052\n",
      "Epoch 64/100\n",
      "100/100 [==============================] - 154s 2s/step - loss: 0.2425 - accuracy: 0.8956 - val_loss: 0.3026 - val_accuracy: 0.8724\n",
      "Epoch 65/100\n",
      "100/100 [==============================] - 150s 1s/step - loss: 0.2390 - accuracy: 0.8963 - val_loss: 0.2190 - val_accuracy: 0.9038\n",
      "Epoch 66/100\n",
      "100/100 [==============================] - 148s 1s/step - loss: 0.2414 - accuracy: 0.8991 - val_loss: 0.2326 - val_accuracy: 0.9000\n",
      "Epoch 67/100\n",
      "100/100 [==============================] - 150s 1s/step - loss: 0.2348 - accuracy: 0.9008 - val_loss: 0.2170 - val_accuracy: 0.9076\n",
      "Epoch 68/100\n",
      "100/100 [==============================] - 149s 1s/step - loss: 0.2406 - accuracy: 0.8985 - val_loss: 0.2283 - val_accuracy: 0.8980\n",
      "Epoch 69/100\n",
      "100/100 [==============================] - 149s 1s/step - loss: 0.2354 - accuracy: 0.9015 - val_loss: 0.2084 - val_accuracy: 0.9112\n",
      "Epoch 70/100\n",
      "100/100 [==============================] - 149s 1s/step - loss: 0.2430 - accuracy: 0.8952 - val_loss: 0.2876 - val_accuracy: 0.8762\n",
      "Epoch 71/100\n",
      "100/100 [==============================] - 149s 1s/step - loss: 0.2383 - accuracy: 0.9000 - val_loss: 0.2178 - val_accuracy: 0.9022\n",
      "Epoch 72/100\n",
      "100/100 [==============================] - 149s 1s/step - loss: 0.2284 - accuracy: 0.9031 - val_loss: 0.2273 - val_accuracy: 0.9030\n",
      "Epoch 73/100\n",
      "100/100 [==============================] - 148s 1s/step - loss: 0.2298 - accuracy: 0.9018 - val_loss: 0.2495 - val_accuracy: 0.8888\n",
      "Epoch 74/100\n",
      "100/100 [==============================] - 155s 2s/step - loss: 0.2344 - accuracy: 0.9004 - val_loss: 0.2403 - val_accuracy: 0.8974\n",
      "Epoch 75/100\n",
      "100/100 [==============================] - 162s 2s/step - loss: 0.2203 - accuracy: 0.9068 - val_loss: 0.2173 - val_accuracy: 0.9060\n",
      "Epoch 76/100\n",
      "100/100 [==============================] - 153s 2s/step - loss: 0.2255 - accuracy: 0.9058 - val_loss: 0.2333 - val_accuracy: 0.8996\n",
      "Epoch 77/100\n",
      "100/100 [==============================] - 149s 1s/step - loss: 0.2229 - accuracy: 0.9066 - val_loss: 0.2207 - val_accuracy: 0.9064\n",
      "Epoch 78/100\n",
      "100/100 [==============================] - 151s 2s/step - loss: 0.2222 - accuracy: 0.9059 - val_loss: 0.2109 - val_accuracy: 0.9124\n",
      "Epoch 79/100\n",
      "100/100 [==============================] - 149s 1s/step - loss: 0.2252 - accuracy: 0.9055 - val_loss: 0.2063 - val_accuracy: 0.9120\n",
      "Epoch 80/100\n",
      "100/100 [==============================] - 149s 1s/step - loss: 0.2107 - accuracy: 0.9104 - val_loss: 0.1948 - val_accuracy: 0.9190\n",
      "Epoch 81/100\n",
      "100/100 [==============================] - 148s 1s/step - loss: 0.2183 - accuracy: 0.9079 - val_loss: 0.2155 - val_accuracy: 0.9066\n",
      "Epoch 82/100\n",
      "100/100 [==============================] - 148s 1s/step - loss: 0.2201 - accuracy: 0.9079 - val_loss: 0.2182 - val_accuracy: 0.9066\n",
      "Epoch 83/100\n",
      "100/100 [==============================] - 149s 1s/step - loss: 0.2127 - accuracy: 0.9106 - val_loss: 0.2015 - val_accuracy: 0.9140\n",
      "Epoch 84/100\n",
      "100/100 [==============================] - 150s 1s/step - loss: 0.2143 - accuracy: 0.9097 - val_loss: 0.1951 - val_accuracy: 0.9210\n",
      "Epoch 85/100\n",
      "100/100 [==============================] - 149s 1s/step - loss: 0.2089 - accuracy: 0.9114 - val_loss: 0.2159 - val_accuracy: 0.9020\n",
      "Epoch 86/100\n",
      "100/100 [==============================] - 149s 1s/step - loss: 0.2086 - accuracy: 0.9136 - val_loss: 0.2324 - val_accuracy: 0.8984\n",
      "Epoch 87/100\n",
      "100/100 [==============================] - 153s 2s/step - loss: 0.2136 - accuracy: 0.9097 - val_loss: 0.2352 - val_accuracy: 0.8972\n",
      "Epoch 88/100\n",
      "100/100 [==============================] - 152s 2s/step - loss: 0.2131 - accuracy: 0.9107 - val_loss: 0.2119 - val_accuracy: 0.9124\n",
      "Epoch 89/100\n",
      "100/100 [==============================] - 150s 1s/step - loss: 0.2099 - accuracy: 0.9118 - val_loss: 0.2268 - val_accuracy: 0.9016\n",
      "Epoch 90/100\n",
      "100/100 [==============================] - 149s 1s/step - loss: 0.2063 - accuracy: 0.9144 - val_loss: 0.1966 - val_accuracy: 0.9162\n",
      "Epoch 91/100\n",
      "100/100 [==============================] - 149s 1s/step - loss: 0.2090 - accuracy: 0.9118 - val_loss: 0.2008 - val_accuracy: 0.9154\n",
      "Epoch 92/100\n",
      "100/100 [==============================] - 149s 1s/step - loss: 0.2021 - accuracy: 0.9155 - val_loss: 0.1916 - val_accuracy: 0.9186\n",
      "Epoch 93/100\n",
      "100/100 [==============================] - 149s 1s/step - loss: 0.2046 - accuracy: 0.9119 - val_loss: 0.1895 - val_accuracy: 0.9192\n",
      "Epoch 94/100\n",
      "100/100 [==============================] - 149s 1s/step - loss: 0.2075 - accuracy: 0.9127 - val_loss: 0.1990 - val_accuracy: 0.9138\n",
      "Epoch 95/100\n",
      "100/100 [==============================] - 149s 1s/step - loss: 0.2116 - accuracy: 0.9101 - val_loss: 0.1986 - val_accuracy: 0.9152\n",
      "Epoch 96/100\n",
      "100/100 [==============================] - 149s 1s/step - loss: 0.2047 - accuracy: 0.9154 - val_loss: 0.1920 - val_accuracy: 0.9190\n",
      "Epoch 97/100\n",
      "100/100 [==============================] - 149s 1s/step - loss: 0.1943 - accuracy: 0.9198 - val_loss: 0.2430 - val_accuracy: 0.9026\n",
      "Epoch 98/100\n",
      "100/100 [==============================] - 148s 1s/step - loss: 0.2043 - accuracy: 0.9156 - val_loss: 0.2117 - val_accuracy: 0.9120\n",
      "Epoch 99/100\n",
      "100/100 [==============================] - 150s 1s/step - loss: 0.1962 - accuracy: 0.9176 - val_loss: 0.2069 - val_accuracy: 0.9124\n",
      "Epoch 100/100\n",
      "100/100 [==============================] - 151s 2s/step - loss: 0.2037 - accuracy: 0.9137 - val_loss: 0.1993 - val_accuracy: 0.9176\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(train_iterator,\n",
    "                    validation_data=validation_iterator,\n",
    "                    steps_per_epoch=100,\n",
    "                    epochs=100,\n",
    "                    validation_steps=100)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Plotting results and saving it"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "id": "4XbxqT3jUp7d"
   },
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "def plot_result(history):\n",
    "    acc = history.history['accuracy']\n",
    "    val_acc = history.history['val_accuracy']\n",
    "    epochs = range(len(acc))\n",
    "\n",
    "    plt.plot(epochs, acc, 'b', label='Training accuracy')\n",
    "    plt.plot(epochs, val_acc, 'r', label='Validation accuracy')\n",
    "    plt.title('Training and validation accuracy')\n",
    "    plt.legend()\n",
    "    plt.figure()\n",
    "\n",
    "    loss = history.history['loss']\n",
    "    val_loss = history.history['val_loss']\n",
    "    plt.plot(epochs, loss, 'b', label='Training Loss')\n",
    "    plt.plot(epochs, val_loss, 'r', label='Validation Loss')\n",
    "    plt.title('Training and validation loss')\n",
    "    plt.legend()\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "id": "EqFZnrOyUp7d"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_result(history)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "id": "aWvTuFKsYSJX"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 12500 images belonging to 2 classes.\n",
      "12500/12500 [==============================] - 50s 4ms/step\n"
     ]
    }
   ],
   "source": [
    "test_generator = ImageDataGenerator(rescale=1. / 255)\n",
    "\n",
    "test_iterator = test_generator.flow_from_directory(\n",
    "    './kaggle_data/input/',\n",
    "    target_size=(DIMENSIONS, DIMENSIONS),\n",
    "    shuffle=False,\n",
    "    class_mode='binary',\n",
    "    batch_size=1)\n",
    "\n",
    "ids = []\n",
    "for filename in test_iterator.filenames:\n",
    "    ids.append(int(filename.split('/')[1].split('.')[0]))\n",
    "\n",
    "predict_result = model.predict(test_iterator, steps=len(test_iterator.filenames))\n",
    "predictions = []\n",
    "for index, prediction in enumerate(predict_result):\n",
    "    predictions.append([ids[index], prediction[0]])\n",
    "predictions.sort()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "id": "yRcbjmKeTzS7"
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "\n",
    "\n",
    "df = pd.DataFrame(data=predictions, index=range(1, 12501), columns=['id', 'label'])\n",
    "df = df.set_index(['id'])\n",
    "df.to_csv('submission.csv')"
   ]
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "collapsed_sections": [],
   "provenance": []
  },
  "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": "3.8.10"
  },
  "vscode": {
   "interpreter": {
    "hash": "5180b977f8d8d7dc150cd187960f6c154734bcfc9201f46d12b0f9d48f671324"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}