Deep Learning CNN Cats-Dogs Datasets.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Assignment15_RickyGunawan_bestCNN.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "NkoomSAAfXBx"
      },
      "source": [
        "zip_path = '/content/drive/My\\ Drive/BCML/datasets/DeepLearning/cats-dogs.zip'\n",
        "\n",
        "!cp {zip_path} /content/\n",
        "\n",
        "!cd /content/\n",
        "\n",
        "!unzip -q /content/cats-dogs.zip -d /content\n",
        "\n",
        "!rm /content/cats-dogs.zip"
      ],
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "f1h4UgpjJ7ab"
      },
      "source": [
        "def plot_loss(history):\n",
        "    plt.plot(history.history['loss'], label='loss')\n",
        "    plt.plot(history.history['val_loss'], label='val_loss')\n",
        "    plt.xlabel('Epoch')\n",
        "    plt.ylabel('Loss')\n",
        "    plt.legend()\n",
        "    plt.grid(True)\n",
        "def plot_accuracy(history):\n",
        "    plt.plot(history.history['accuracy'], label='accuracy')\n",
        "    plt.plot(history.history['val_accuracy'], label='val_accuracy')\n",
        "    plt.xlabel('Epoch')\n",
        "    plt.ylabel('accuracy')\n",
        "    plt.legend()\n",
        "    plt.grid(True)"
      ],
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IXyxQWL7v12W"
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "import os\n",
        "\n",
        "import tensorflow as tf\n",
        "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
        "from tensorflow.keras.utils import to_categorical\n",
        "from tensorflow.keras.models import Sequential\n",
        "from tensorflow.keras.layers import Conv2D\n",
        "from tensorflow.keras.layers import MaxPooling2D\n",
        "from tensorflow.keras.layers import Dense\n",
        "from tensorflow.keras.layers import Flatten\n",
        "from tensorflow.keras.layers import Dropout\n",
        "from tensorflow.keras.optimizers import SGD, Adam, RMSprop\n",
        "from tensorflow.keras.layers import BatchNormalization"
      ],
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8pU82B52kiSi"
      },
      "source": [
        "dataset_dir = '/content/'\n",
        "train_datagen = ImageDataGenerator(rescale=1.0/255.0,horizontal_flip=True,zoom_range=0.1,shear_range=0.1)\n",
        "test_datagen = ImageDataGenerator(rescale=1.0/255.0)"
      ],
      "execution_count": 5,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "i2PWokbG7S8F",
        "outputId": "003bd6a0-eb90-4ce0-f967-978fa1494c8b",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "train_augmented_iterator = train_datagen.flow_from_directory(os.path.join(dataset_dir, 'train'), class_mode='binary', batch_size=128, target_size=(200, 200))\n",
        "test_augmented_iterator = test_datagen.flow_from_directory(os.path.join(dataset_dir, 'test'), class_mode='binary', batch_size=128, target_size=(200, 200))"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Found 12749 images belonging to 2 classes.\n",
            "Found 2252 images belonging to 2 classes.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wxLDfa8iiUyJ",
        "outputId": "8a632b5c-2316-45bf-e0a8-591b8873338d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 765
        }
      },
      "source": [
        "model = Sequential()\n",
        "\n",
        "model.add(Conv2D(32, (3, 3), padding='same', input_shape=(200,200,3), activation='relu'))\n",
        "model.add(Conv2D(32, (3, 3), padding='same', activation='relu'))\n",
        "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
        "\n",
        "model.add(Conv2D(64, (3, 3), padding='same', activation='relu'))\n",
        "model.add(Conv2D(64, (3, 3), padding='same', activation='relu'))\n",
        "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
        "\n",
        "model.add(Conv2D(128, (3, 3), padding='same', activation='relu'))\n",
        "model.add(Conv2D(128, (3, 3), padding='same', activation='relu'))\n",
        "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
        "\n",
        "model.add(Conv2D(256, (3, 3), padding='same', activation='relu'))\n",
        "model.add(Conv2D(256, (3, 3), padding='same', activation='relu'))\n",
        "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
        "\n",
        "model.add(Flatten())\n",
        "model.add(Dense(256, activation='relu'))\n",
        "model.add(Dropout(0.5))\n",
        "\n",
        "model.add(Dense(256, activation='relu'))\n",
        "model.add(Dropout(0.5))\n",
        "\n",
        "model.add(Dense(1,activation='sigmoid'))\n",
        "\n",
        "opt=RMSprop(lr=0.0001)\n",
        "\n",
        "model.compile(loss='binary_crossentropy',optimizer=opt,metrics=['accuracy'])\n",
        "\n",
        "model.summary()"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d (Conv2D)              (None, 200, 200, 32)      896       \n",
            "_________________________________________________________________\n",
            "conv2d_1 (Conv2D)            (None, 200, 200, 32)      9248      \n",
            "_________________________________________________________________\n",
            "max_pooling2d (MaxPooling2D) (None, 100, 100, 32)      0         \n",
            "_________________________________________________________________\n",
            "conv2d_2 (Conv2D)            (None, 100, 100, 64)      18496     \n",
            "_________________________________________________________________\n",
            "conv2d_3 (Conv2D)            (None, 100, 100, 64)      36928     \n",
            "_________________________________________________________________\n",
            "max_pooling2d_1 (MaxPooling2 (None, 50, 50, 64)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_4 (Conv2D)            (None, 50, 50, 128)       73856     \n",
            "_________________________________________________________________\n",
            "conv2d_5 (Conv2D)            (None, 50, 50, 128)       147584    \n",
            "_________________________________________________________________\n",
            "max_pooling2d_2 (MaxPooling2 (None, 25, 25, 128)       0         \n",
            "_________________________________________________________________\n",
            "conv2d_6 (Conv2D)            (None, 25, 25, 256)       295168    \n",
            "_________________________________________________________________\n",
            "conv2d_7 (Conv2D)            (None, 25, 25, 256)       590080    \n",
            "_________________________________________________________________\n",
            "max_pooling2d_3 (MaxPooling2 (None, 12, 12, 256)       0         \n",
            "_________________________________________________________________\n",
            "flatten (Flatten)            (None, 36864)             0         \n",
            "_________________________________________________________________\n",
            "dense (Dense)                (None, 256)               9437440   \n",
            "_________________________________________________________________\n",
            "dropout (Dropout)            (None, 256)               0         \n",
            "_________________________________________________________________\n",
            "dense_1 (Dense)              (None, 256)               65792     \n",
            "_________________________________________________________________\n",
            "dropout_1 (Dropout)          (None, 256)               0         \n",
            "_________________________________________________________________\n",
            "dense_2 (Dense)              (None, 1)                 257       \n",
            "=================================================================\n",
            "Total params: 10,675,745\n",
            "Trainable params: 10,675,745\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "aEbwT2zTsBwO",
        "outputId": "b68cafd7-7df2-4651-a557-ec07c76c4d9d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "history=model.fit(train_augmented_iterator, steps_per_epoch=len(train_augmented_iterator), validation_data=test_augmented_iterator, validation_steps=len(test_augmented_iterator), epochs=75)"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/75\n",
            "  2/100 [..............................] - ETA: 40s - loss: 0.6966 - accuracy: 0.5156WARNING:tensorflow:Callbacks method `on_train_batch_end` is slow compared to the batch time (batch time: 0.2490s vs `on_train_batch_end` time: 0.5601s). Check your callbacks.\n",
            "100/100 [==============================] - 185s 2s/step - loss: 0.6873 - accuracy: 0.5360 - val_loss: 0.6598 - val_accuracy: 0.5764\n",
            "Epoch 2/75\n",
            "100/100 [==============================] - 181s 2s/step - loss: 0.6341 - accuracy: 0.6349 - val_loss: 0.5689 - val_accuracy: 0.6998\n",
            "Epoch 3/75\n",
            "100/100 [==============================] - 182s 2s/step - loss: 0.5871 - accuracy: 0.6911 - val_loss: 0.5514 - val_accuracy: 0.7225\n",
            "Epoch 4/75\n",
            "100/100 [==============================] - 182s 2s/step - loss: 0.5634 - accuracy: 0.7129 - val_loss: 0.5291 - val_accuracy: 0.7180\n",
            "Epoch 5/75\n",
            "100/100 [==============================] - 182s 2s/step - loss: 0.5391 - accuracy: 0.7328 - val_loss: 0.5123 - val_accuracy: 0.7385\n",
            "Epoch 6/75\n",
            "100/100 [==============================] - 183s 2s/step - loss: 0.5234 - accuracy: 0.7385 - val_loss: 0.5012 - val_accuracy: 0.7611\n",
            "Epoch 7/75\n",
            "100/100 [==============================] - 181s 2s/step - loss: 0.5097 - accuracy: 0.7558 - val_loss: 0.4702 - val_accuracy: 0.7784\n",
            "Epoch 8/75\n",
            "100/100 [==============================] - 180s 2s/step - loss: 0.4846 - accuracy: 0.7670 - val_loss: 0.4530 - val_accuracy: 0.7842\n",
            "Epoch 9/75\n",
            "100/100 [==============================] - 181s 2s/step - loss: 0.4675 - accuracy: 0.7803 - val_loss: 0.4942 - val_accuracy: 0.7722\n",
            "Epoch 10/75\n",
            "100/100 [==============================] - 181s 2s/step - loss: 0.4500 - accuracy: 0.7904 - val_loss: 0.4075 - val_accuracy: 0.8108\n",
            "Epoch 11/75\n",
            "100/100 [==============================] - 179s 2s/step - loss: 0.4335 - accuracy: 0.8034 - val_loss: 0.4350 - val_accuracy: 0.7953\n",
            "Epoch 12/75\n",
            "100/100 [==============================] - 182s 2s/step - loss: 0.4137 - accuracy: 0.8176 - val_loss: 0.4016 - val_accuracy: 0.8166\n",
            "Epoch 13/75\n",
            "100/100 [==============================] - 181s 2s/step - loss: 0.4017 - accuracy: 0.8183 - val_loss: 0.3836 - val_accuracy: 0.8197\n",
            "Epoch 14/75\n",
            "100/100 [==============================] - 181s 2s/step - loss: 0.3800 - accuracy: 0.8320 - val_loss: 0.4072 - val_accuracy: 0.8139\n",
            "Epoch 15/75\n",
            "100/100 [==============================] - 179s 2s/step - loss: 0.3714 - accuracy: 0.8348 - val_loss: 0.3484 - val_accuracy: 0.8459\n",
            "Epoch 16/75\n",
            "100/100 [==============================] - 179s 2s/step - loss: 0.3595 - accuracy: 0.8419 - val_loss: 0.3756 - val_accuracy: 0.8335\n",
            "Epoch 17/75\n",
            "100/100 [==============================] - 179s 2s/step - loss: 0.3534 - accuracy: 0.8444 - val_loss: 0.3680 - val_accuracy: 0.8268\n",
            "Epoch 18/75\n",
            "100/100 [==============================] - 181s 2s/step - loss: 0.3382 - accuracy: 0.8538 - val_loss: 0.3855 - val_accuracy: 0.8215\n",
            "Epoch 19/75\n",
            "100/100 [==============================] - 180s 2s/step - loss: 0.3257 - accuracy: 0.8598 - val_loss: 0.3641 - val_accuracy: 0.8321\n",
            "Epoch 20/75\n",
            "100/100 [==============================] - 180s 2s/step - loss: 0.3203 - accuracy: 0.8601 - val_loss: 0.3209 - val_accuracy: 0.8588\n",
            "Epoch 21/75\n",
            "100/100 [==============================] - 180s 2s/step - loss: 0.3039 - accuracy: 0.8702 - val_loss: 0.3193 - val_accuracy: 0.8579\n",
            "Epoch 22/75\n",
            "100/100 [==============================] - 179s 2s/step - loss: 0.2908 - accuracy: 0.8807 - val_loss: 0.3122 - val_accuracy: 0.8686\n",
            "Epoch 23/75\n",
            "100/100 [==============================] - 179s 2s/step - loss: 0.2832 - accuracy: 0.8800 - val_loss: 0.3231 - val_accuracy: 0.8641\n",
            "Epoch 24/75\n",
            "100/100 [==============================] - 180s 2s/step - loss: 0.2765 - accuracy: 0.8840 - val_loss: 0.2928 - val_accuracy: 0.8761\n",
            "Epoch 25/75\n",
            "100/100 [==============================] - 179s 2s/step - loss: 0.2678 - accuracy: 0.8860 - val_loss: 0.3428 - val_accuracy: 0.8477\n",
            "Epoch 26/75\n",
            "100/100 [==============================] - 178s 2s/step - loss: 0.2553 - accuracy: 0.8936 - val_loss: 0.3394 - val_accuracy: 0.8606\n",
            "Epoch 27/75\n",
            "100/100 [==============================] - 180s 2s/step - loss: 0.2450 - accuracy: 0.8961 - val_loss: 0.2834 - val_accuracy: 0.8739\n",
            "Epoch 28/75\n",
            "100/100 [==============================] - 180s 2s/step - loss: 0.2360 - accuracy: 0.9042 - val_loss: 0.3324 - val_accuracy: 0.8690\n",
            "Epoch 29/75\n",
            "100/100 [==============================] - 180s 2s/step - loss: 0.2321 - accuracy: 0.9047 - val_loss: 0.3239 - val_accuracy: 0.8659\n",
            "Epoch 30/75\n",
            "100/100 [==============================] - 179s 2s/step - loss: 0.2222 - accuracy: 0.9121 - val_loss: 0.2893 - val_accuracy: 0.8774\n",
            "Epoch 31/75\n",
            "100/100 [==============================] - 179s 2s/step - loss: 0.2132 - accuracy: 0.9147 - val_loss: 0.2750 - val_accuracy: 0.8899\n",
            "Epoch 32/75\n",
            "100/100 [==============================] - 178s 2s/step - loss: 0.2043 - accuracy: 0.9172 - val_loss: 0.2875 - val_accuracy: 0.8806\n",
            "Epoch 33/75\n",
            "100/100 [==============================] - 181s 2s/step - loss: 0.1971 - accuracy: 0.9221 - val_loss: 0.2505 - val_accuracy: 0.9036\n",
            "Epoch 34/75\n",
            "100/100 [==============================] - 181s 2s/step - loss: 0.1910 - accuracy: 0.9225 - val_loss: 0.3177 - val_accuracy: 0.8872\n",
            "Epoch 35/75\n",
            "100/100 [==============================] - 182s 2s/step - loss: 0.1897 - accuracy: 0.9242 - val_loss: 0.2473 - val_accuracy: 0.9045\n",
            "Epoch 36/75\n",
            "100/100 [==============================] - 182s 2s/step - loss: 0.1785 - accuracy: 0.9296 - val_loss: 0.2507 - val_accuracy: 0.9103\n",
            "Epoch 37/75\n",
            "100/100 [==============================] - 182s 2s/step - loss: 0.1719 - accuracy: 0.9302 - val_loss: 0.2403 - val_accuracy: 0.9067\n",
            "Epoch 38/75\n",
            "100/100 [==============================] - 182s 2s/step - loss: 0.1692 - accuracy: 0.9345 - val_loss: 0.2919 - val_accuracy: 0.9063\n",
            "Epoch 39/75\n",
            "100/100 [==============================] - 179s 2s/step - loss: 0.1626 - accuracy: 0.9374 - val_loss: 0.2383 - val_accuracy: 0.9072\n",
            "Epoch 40/75\n",
            "100/100 [==============================] - 179s 2s/step - loss: 0.1563 - accuracy: 0.9366 - val_loss: 0.2305 - val_accuracy: 0.9143\n",
            "Epoch 41/75\n",
            "100/100 [==============================] - 179s 2s/step - loss: 0.1442 - accuracy: 0.9442 - val_loss: 0.3033 - val_accuracy: 0.8832\n",
            "Epoch 42/75\n",
            "100/100 [==============================] - 179s 2s/step - loss: 0.1414 - accuracy: 0.9441 - val_loss: 0.2416 - val_accuracy: 0.9192\n",
            "Epoch 43/75\n",
            "100/100 [==============================] - 178s 2s/step - loss: 0.1364 - accuracy: 0.9479 - val_loss: 0.2757 - val_accuracy: 0.9107\n",
            "Epoch 44/75\n",
            "100/100 [==============================] - 178s 2s/step - loss: 0.1304 - accuracy: 0.9494 - val_loss: 0.2274 - val_accuracy: 0.9156\n",
            "Epoch 45/75\n",
            "100/100 [==============================] - 178s 2s/step - loss: 0.1276 - accuracy: 0.9520 - val_loss: 0.2214 - val_accuracy: 0.9156\n",
            "Epoch 46/75\n",
            "100/100 [==============================] - 179s 2s/step - loss: 0.1215 - accuracy: 0.9551 - val_loss: 0.2106 - val_accuracy: 0.9223\n",
            "Epoch 47/75\n",
            "100/100 [==============================] - 180s 2s/step - loss: 0.1180 - accuracy: 0.9562 - val_loss: 0.2473 - val_accuracy: 0.9085\n",
            "Epoch 48/75\n",
            "100/100 [==============================] - 180s 2s/step - loss: 0.1185 - accuracy: 0.9562 - val_loss: 0.2451 - val_accuracy: 0.9192\n",
            "Epoch 49/75\n",
            "100/100 [==============================] - 182s 2s/step - loss: 0.1079 - accuracy: 0.9589 - val_loss: 0.2559 - val_accuracy: 0.9298\n",
            "Epoch 50/75\n",
            "100/100 [==============================] - 183s 2s/step - loss: 0.1076 - accuracy: 0.9601 - val_loss: 0.2073 - val_accuracy: 0.9307\n",
            "Epoch 51/75\n",
            "100/100 [==============================] - 181s 2s/step - loss: 0.0993 - accuracy: 0.9617 - val_loss: 0.2628 - val_accuracy: 0.9223\n",
            "Epoch 52/75\n",
            "100/100 [==============================] - 181s 2s/step - loss: 0.1004 - accuracy: 0.9624 - val_loss: 0.2962 - val_accuracy: 0.9214\n",
            "Epoch 53/75\n",
            "100/100 [==============================] - 180s 2s/step - loss: 0.0966 - accuracy: 0.9634 - val_loss: 0.2531 - val_accuracy: 0.9210\n",
            "Epoch 54/75\n",
            "100/100 [==============================] - 179s 2s/step - loss: 0.0901 - accuracy: 0.9631 - val_loss: 0.2974 - val_accuracy: 0.9214\n",
            "Epoch 55/75\n",
            "100/100 [==============================] - 181s 2s/step - loss: 0.0961 - accuracy: 0.9658 - val_loss: 0.2888 - val_accuracy: 0.9081\n",
            "Epoch 56/75\n",
            "100/100 [==============================] - 183s 2s/step - loss: 0.0873 - accuracy: 0.9672 - val_loss: 0.2597 - val_accuracy: 0.9218\n",
            "Epoch 57/75\n",
            "100/100 [==============================] - 182s 2s/step - loss: 0.0839 - accuracy: 0.9689 - val_loss: 0.2369 - val_accuracy: 0.9161\n",
            "Epoch 58/75\n",
            "100/100 [==============================] - 183s 2s/step - loss: 0.0842 - accuracy: 0.9702 - val_loss: 0.2419 - val_accuracy: 0.9290\n",
            "Epoch 59/75\n",
            "100/100 [==============================] - 182s 2s/step - loss: 0.0863 - accuracy: 0.9692 - val_loss: 0.2523 - val_accuracy: 0.8908\n",
            "Epoch 60/75\n",
            "100/100 [==============================] - 182s 2s/step - loss: 0.0872 - accuracy: 0.9682 - val_loss: 0.2962 - val_accuracy: 0.9187\n",
            "Epoch 61/75\n",
            "100/100 [==============================] - 183s 2s/step - loss: 0.0744 - accuracy: 0.9723 - val_loss: 0.2424 - val_accuracy: 0.9187\n",
            "Epoch 62/75\n",
            "100/100 [==============================] - 183s 2s/step - loss: 0.0751 - accuracy: 0.9742 - val_loss: 0.3186 - val_accuracy: 0.9316\n",
            "Epoch 63/75\n",
            "100/100 [==============================] - 183s 2s/step - loss: 0.0780 - accuracy: 0.9733 - val_loss: 0.3223 - val_accuracy: 0.9312\n",
            "Epoch 64/75\n",
            "100/100 [==============================] - 182s 2s/step - loss: 0.0782 - accuracy: 0.9726 - val_loss: 0.3150 - val_accuracy: 0.9161\n",
            "Epoch 65/75\n",
            "100/100 [==============================] - 182s 2s/step - loss: 0.0762 - accuracy: 0.9737 - val_loss: 0.3989 - val_accuracy: 0.9134\n",
            "Epoch 66/75\n",
            "100/100 [==============================] - 182s 2s/step - loss: 0.0693 - accuracy: 0.9755 - val_loss: 0.2647 - val_accuracy: 0.9298\n",
            "Epoch 67/75\n",
            "100/100 [==============================] - 182s 2s/step - loss: 0.0707 - accuracy: 0.9761 - val_loss: 0.2251 - val_accuracy: 0.9267\n",
            "Epoch 68/75\n",
            "100/100 [==============================] - 182s 2s/step - loss: 0.0646 - accuracy: 0.9778 - val_loss: 0.2267 - val_accuracy: 0.9263\n",
            "Epoch 69/75\n",
            "100/100 [==============================] - 182s 2s/step - loss: 0.0644 - accuracy: 0.9768 - val_loss: 0.3220 - val_accuracy: 0.9263\n",
            "Epoch 70/75\n",
            "100/100 [==============================] - 182s 2s/step - loss: 0.0805 - accuracy: 0.9748 - val_loss: 0.3376 - val_accuracy: 0.9241\n",
            "Epoch 71/75\n",
            "100/100 [==============================] - 182s 2s/step - loss: 0.0637 - accuracy: 0.9776 - val_loss: 0.3762 - val_accuracy: 0.9245\n",
            "Epoch 72/75\n",
            "100/100 [==============================] - 182s 2s/step - loss: 0.0637 - accuracy: 0.9782 - val_loss: 0.3225 - val_accuracy: 0.9281\n",
            "Epoch 73/75\n",
            "100/100 [==============================] - 181s 2s/step - loss: 0.0587 - accuracy: 0.9784 - val_loss: 0.3374 - val_accuracy: 0.9214\n",
            "Epoch 74/75\n",
            "100/100 [==============================] - 180s 2s/step - loss: 0.0598 - accuracy: 0.9794 - val_loss: 0.2716 - val_accuracy: 0.9218\n",
            "Epoch 75/75\n",
            "100/100 [==============================] - 181s 2s/step - loss: 0.0680 - accuracy: 0.9778 - val_loss: 0.2900 - val_accuracy: 0.9356\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZVDHpoev96Zr",
        "outputId": "977fb12a-02a9-4e9b-9ae6-a67b47cd4540",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 279
        }
      },
      "source": [
        "plot_loss(history)"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UsInz2-j9_eA",
        "outputId": "12ea7b33-2635-44e0-a878-5620087189e4",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 283
        }
      },
      "source": [
        "plot_accuracy(history)"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "V1Q5he-H-j2H"
      },
      "source": [
        "# **Kesimpulan**"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "iVMSb8yd-paS"
      },
      "source": [
        "Epoch 50/75\n",
        "\n",
        "100/100 [==============================] - 183s 2s/step - loss: 0.1076 - accuracy: 0.9601 - val_loss: 0.2073 - val_accuracy: 0.9307\n",
        "\n",
        "dari hasil terbaik yang saya dapat dengan melihat val_loss terkecil berada di angka 0.20 dengan akurasi di 93%"
      ]
    }
  ]
}