vgg16.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "vgg16.ipynb",
      "version": "0.3.2",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "TPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/Juhi-Purswani/Offline_Signature_Verification/blob/master/vgg16.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "G26qlx234Bhw",
        "colab_type": "code",
        "outputId": "ca68c0ef-a4b6-4e1b-cfa7-61c5b1a80e26",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 125
        }
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Go to this URL in a browser: https://accounts.google.com/o/oauth2/auth?client_id=947318989803-6bn6qk8qdgf4n4g3pfee6491hc0brc4i.apps.googleusercontent.com&redirect_uri=urn%3Aietf%3Awg%3Aoauth%3A2.0%3Aoob&scope=email%20https%3A%2F%2Fwww.googleapis.com%2Fauth%2Fdocs.test%20https%3A%2F%2Fwww.googleapis.com%2Fauth%2Fdrive%20https%3A%2F%2Fwww.googleapis.com%2Fauth%2Fdrive.photos.readonly%20https%3A%2F%2Fwww.googleapis.com%2Fauth%2Fpeopleapi.readonly&response_type=code\n",
            "\n",
            "Enter your authorization code:\n",
            "··········\n",
            "Mounted at /content/drive\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qQSnV__e6htc",
        "colab_type": "code",
        "outputId": "8f5f1858-46d2-48da-fc97-a2195101e92d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        " !ls"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "drive  sample_data\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "aTaF0Q9t6jp5",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import os\n",
        "os.chdir('./drive/My Drive/SignForge')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KL1wcrmK7YLo",
        "colab_type": "code",
        "outputId": "4678dd6c-7d42-4e62-c5f4-9bd2fccebd9b",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "import cv2                 # working with, mainly resizing, images\n",
        "import numpy as np         # dealing with arrays\n",
        "import os                  # dealing with directories\n",
        "from random import shuffle # mixing up or currently ordered data that might lead our network astray in training.\n",
        "from tqdm import tqdm \n",
        "import numpy as np\n",
        "from keras.utils import to_categorical\n",
        "\n",
        "from keras.models import Sequential, Model, load_model\n",
        "from keras import applications\n",
        "from keras import optimizers\n",
        "from keras.layers import Dropout, Flatten, Dense\n",
        "\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "TRAIN_DIR = 'data/train_set'\n",
        "TEST_DIR = 'data/test_set'\n",
        "IMG_SIZE = 224\n",
        "img_rows, img_cols, img_channel = 224, 224, 3\n"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Using TensorFlow backend.\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IZUjW2f475un",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import cv2                 # working with, mainly resizing, images\n",
        "import numpy as np         # dealing with arrays\n",
        "import os                  # dealing with directories\n",
        "from random import shuffle # mixing up or currently ordered data that might lead our network astray in training.\n",
        "from tqdm import tqdm      \n",
        "\n",
        "TRAIN_DIR = 'data/train_set'\n",
        "TEST_DIR = 'data/test_set'\n",
        "IMG_SIZE = 224\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZhmN1wsr8JCo",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def person_img(img):\n",
        "    label = img.split('.')\n",
        "    k = label[0]\n",
        "    return [k[-3:]]\n",
        "def create_features(DIR):\n",
        "    features = []\n",
        "    labels = []\n",
        "    for img in tqdm(os.listdir(DIR)):\n",
        "        label = label_img(img)\n",
        "        person = person_img(img)\n",
        "       \n",
        "        path = os.path.join(DIR,img)\n",
        "        img = cv2.imread(path)\n",
        "        img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)\n",
        "        img = cv2.resize(img, (IMG_SIZE,IMG_SIZE))\n",
        "        features.append([img])\n",
        "        labels.append(np.array(label))\n",
        "    return features\n",
        "def label_img(img):\n",
        "    image_label = img.split('.')\n",
        "    name = image_label[0]\n",
        "    # checking if the signature is forged or genuine\n",
        "    #                            \n",
        "    \n",
        "    if name[4:7] == name[-3:]:\n",
        "        return 1\n",
        "  \n",
        "    elif name[4:7] != name[-3:]:\n",
        "        return 0"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ww5DW7UY8Lhn",
        "colab_type": "code",
        "outputId": "554c653b-2b82-4389-b303-749766b9e0c4",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 105
        }
      },
      "source": [
        "train_features  = np.array(create_features(TRAIN_DIR))\n",
        "test_features  = np.array(create_features(TEST_DIR))\n",
        "print (test_features.shape)\n",
        "train_features_reshaped = train_features.reshape(-1,224,224,3)\n",
        "test_features_reshaped = test_features.reshape(-1,224,224,3)\n",
        "print (test_features_reshaped.shape)\n",
        "print (train_features_reshaped.shape)\n",
        "\n",
        "train_features_reshaped = train_features_reshaped.astype('float32')\n",
        "test_features_reshaped = test_features_reshaped.astype('float32')\n",
        "train_features_reshaped = train_features_reshaped / 255.\n",
        "test_features_reshaped = test_features_reshaped / 255.\n",
        "#print (train_features_reshaped)"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "100%|██████████| 179/179 [01:00<00:00,  3.03it/s]\n",
            "100%|██████████| 121/121 [00:39<00:00,  3.49it/s]\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "(121, 1, 224, 224, 3)\n",
            "(121, 224, 224, 3)\n",
            "(179, 224, 224, 3)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "A6fX1xZt8WE3",
        "colab_type": "code",
        "outputId": "5bd2e7cb-28ac-4d73-e7ae-42fa5a6b300e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "train_features_reshaped.shape"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(179, 224, 224, 3)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "17l3Xvr79c04",
        "colab_type": "code",
        "outputId": "dbbec004-cd08-4d60-dead-84f0274197f2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 87
        }
      },
      "source": [
        "def create_labels(DIR):\n",
        "    labels = []\n",
        "    for img in tqdm(os.listdir(DIR)):\n",
        "        label = label_img(img)\n",
        "        labels.append(np.array(label))\n",
        "    return labels\n",
        "train_labels = np.array(create_labels(TRAIN_DIR))\n",
        "test_labels = np.array(create_labels(TEST_DIR))\n",
        "print (test_labels.shape)\n",
        "train_labels_one_hot = to_categorical(train_labels)\n",
        "#print (train_labels_one_hot)\n",
        "test_labels_one_hot = to_categorical(test_labels)\n",
        "print (test_labels_one_hot.shape)"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "100%|██████████| 179/179 [00:00<00:00, 263246.99it/s]\n",
            "100%|██████████| 121/121 [00:00<00:00, 223081.66it/s]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "(121,)\n",
            "(121, 2)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qC5jcqeL9ymY",
        "colab_type": "code",
        "outputId": "c77c5dd3-259b-4b35-d3ba-55948c630da6",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "base_model = applications.VGG16(weights='imagenet', include_top=False, input_shape=(img_rows, img_cols, img_channel))\n",
        "base_model.summary()\n",
        "\n",
        "add_model = Sequential()\n",
        "add_model.add(Flatten(input_shape=base_model.output_shape[1:]))\n",
        "add_model.add(Dense(256, activation='relu'))\n",
        "add_model.add(Dense(2, activation='sigmoid'))\n",
        "\n",
        "model = Model(inputs=base_model.input, outputs=add_model(base_model.output))\n",
        "model.compile(loss='binary_crossentropy', optimizer=optimizers.SGD(lr=1e-4, momentum=0.9),\n",
        "              metrics=['accuracy'])\n",
        "\n",
        "model.summary()"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING: Logging before flag parsing goes to stderr.\n",
            "W0713 14:59:58.362632 140133506992000 deprecation_wrapper.py:119] From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:74: The name tf.get_default_graph is deprecated. Please use tf.compat.v1.get_default_graph instead.\n",
            "\n",
            "W0713 14:59:58.431775 140133506992000 deprecation_wrapper.py:119] From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:517: The name tf.placeholder is deprecated. Please use tf.compat.v1.placeholder instead.\n",
            "\n",
            "W0713 14:59:58.446002 140133506992000 deprecation_wrapper.py:119] From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:4138: The name tf.random_uniform is deprecated. Please use tf.random.uniform instead.\n",
            "\n",
            "W0713 14:59:58.510613 140133506992000 deprecation_wrapper.py:119] From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:3976: The name tf.nn.max_pool is deprecated. Please use tf.nn.max_pool2d instead.\n",
            "\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Downloading data from https://github.com/fchollet/deep-learning-models/releases/download/v0.1/vgg16_weights_tf_dim_ordering_tf_kernels_notop.h5\n",
            "58892288/58889256 [==============================] - 1s 0us/step\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "W0713 15:00:00.073915 140133506992000 deprecation_wrapper.py:119] From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:174: The name tf.get_default_session is deprecated. Please use tf.compat.v1.get_default_session instead.\n",
            "\n",
            "W0713 15:00:00.075343 140133506992000 deprecation_wrapper.py:119] From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:181: The name tf.ConfigProto is deprecated. Please use tf.compat.v1.ConfigProto instead.\n",
            "\n",
            "W0713 15:00:00.686165 140133506992000 deprecation_wrapper.py:119] From /usr/local/lib/python3.6/dist-packages/keras/optimizers.py:790: The name tf.train.Optimizer is deprecated. Please use tf.compat.v1.train.Optimizer instead.\n",
            "\n",
            "W0713 15:00:00.697702 140133506992000 deprecation.py:323] From /usr/local/lib/python3.6/dist-packages/tensorflow/python/ops/nn_impl.py:180: add_dispatch_support.<locals>.wrapper (from tensorflow.python.ops.array_ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Use tf.where in 2.0, which has the same broadcast rule as np.where\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "input_1 (InputLayer)         (None, 224, 224, 3)       0         \n",
            "_________________________________________________________________\n",
            "block1_conv1 (Conv2D)        (None, 224, 224, 64)      1792      \n",
            "_________________________________________________________________\n",
            "block1_conv2 (Conv2D)        (None, 224, 224, 64)      36928     \n",
            "_________________________________________________________________\n",
            "block1_pool (MaxPooling2D)   (None, 112, 112, 64)      0         \n",
            "_________________________________________________________________\n",
            "block2_conv1 (Conv2D)        (None, 112, 112, 128)     73856     \n",
            "_________________________________________________________________\n",
            "block2_conv2 (Conv2D)        (None, 112, 112, 128)     147584    \n",
            "_________________________________________________________________\n",
            "block2_pool (MaxPooling2D)   (None, 56, 56, 128)       0         \n",
            "_________________________________________________________________\n",
            "block3_conv1 (Conv2D)        (None, 56, 56, 256)       295168    \n",
            "_________________________________________________________________\n",
            "block3_conv2 (Conv2D)        (None, 56, 56, 256)       590080    \n",
            "_________________________________________________________________\n",
            "block3_conv3 (Conv2D)        (None, 56, 56, 256)       590080    \n",
            "_________________________________________________________________\n",
            "block3_pool (MaxPooling2D)   (None, 28, 28, 256)       0         \n",
            "_________________________________________________________________\n",
            "block4_conv1 (Conv2D)        (None, 28, 28, 512)       1180160   \n",
            "_________________________________________________________________\n",
            "block4_conv2 (Conv2D)        (None, 28, 28, 512)       2359808   \n",
            "_________________________________________________________________\n",
            "block4_conv3 (Conv2D)        (None, 28, 28, 512)       2359808   \n",
            "_________________________________________________________________\n",
            "block4_pool (MaxPooling2D)   (None, 14, 14, 512)       0         \n",
            "_________________________________________________________________\n",
            "block5_conv1 (Conv2D)        (None, 14, 14, 512)       2359808   \n",
            "_________________________________________________________________\n",
            "block5_conv2 (Conv2D)        (None, 14, 14, 512)       2359808   \n",
            "_________________________________________________________________\n",
            "block5_conv3 (Conv2D)        (None, 14, 14, 512)       2359808   \n",
            "_________________________________________________________________\n",
            "block5_pool (MaxPooling2D)   (None, 7, 7, 512)         0         \n",
            "=================================================================\n",
            "Total params: 14,714,688\n",
            "Trainable params: 14,714,688\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "input_1 (InputLayer)         (None, 224, 224, 3)       0         \n",
            "_________________________________________________________________\n",
            "block1_conv1 (Conv2D)        (None, 224, 224, 64)      1792      \n",
            "_________________________________________________________________\n",
            "block1_conv2 (Conv2D)        (None, 224, 224, 64)      36928     \n",
            "_________________________________________________________________\n",
            "block1_pool (MaxPooling2D)   (None, 112, 112, 64)      0         \n",
            "_________________________________________________________________\n",
            "block2_conv1 (Conv2D)        (None, 112, 112, 128)     73856     \n",
            "_________________________________________________________________\n",
            "block2_conv2 (Conv2D)        (None, 112, 112, 128)     147584    \n",
            "_________________________________________________________________\n",
            "block2_pool (MaxPooling2D)   (None, 56, 56, 128)       0         \n",
            "_________________________________________________________________\n",
            "block3_conv1 (Conv2D)        (None, 56, 56, 256)       295168    \n",
            "_________________________________________________________________\n",
            "block3_conv2 (Conv2D)        (None, 56, 56, 256)       590080    \n",
            "_________________________________________________________________\n",
            "block3_conv3 (Conv2D)        (None, 56, 56, 256)       590080    \n",
            "_________________________________________________________________\n",
            "block3_pool (MaxPooling2D)   (None, 28, 28, 256)       0         \n",
            "_________________________________________________________________\n",
            "block4_conv1 (Conv2D)        (None, 28, 28, 512)       1180160   \n",
            "_________________________________________________________________\n",
            "block4_conv2 (Conv2D)        (None, 28, 28, 512)       2359808   \n",
            "_________________________________________________________________\n",
            "block4_conv3 (Conv2D)        (None, 28, 28, 512)       2359808   \n",
            "_________________________________________________________________\n",
            "block4_pool (MaxPooling2D)   (None, 14, 14, 512)       0         \n",
            "_________________________________________________________________\n",
            "block5_conv1 (Conv2D)        (None, 14, 14, 512)       2359808   \n",
            "_________________________________________________________________\n",
            "block5_conv2 (Conv2D)        (None, 14, 14, 512)       2359808   \n",
            "_________________________________________________________________\n",
            "block5_conv3 (Conv2D)        (None, 14, 14, 512)       2359808   \n",
            "_________________________________________________________________\n",
            "block5_pool (MaxPooling2D)   (None, 7, 7, 512)         0         \n",
            "_________________________________________________________________\n",
            "sequential_1 (Sequential)    (None, 2)                 6423298   \n",
            "=================================================================\n",
            "Total params: 21,137,986\n",
            "Trainable params: 21,137,986\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7ooCK9v3-RX4",
        "colab_type": "code",
        "outputId": "867ccf19-06c6-45ad-a06f-7060ec06b099",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "batch_size = 32\n",
        "epochs = 50\n",
        "\n",
        "history = model.fit(train_features_reshaped, train_labels_one_hot, batch_size=batch_size,epochs=epochs,verbose=1,validation_data=(test_features_reshaped, test_labels_one_hot))\n"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Train on 179 samples, validate on 121 samples\n",
            "Epoch 1/50\n",
            "179/179 [==============================] - 415s 2s/step - loss: 0.7668 - acc: 0.4860 - val_loss: 0.6976 - val_acc: 0.4917\n",
            "Epoch 2/50\n",
            "179/179 [==============================] - 410s 2s/step - loss: 0.7320 - acc: 0.4972 - val_loss: 0.7441 - val_acc: 0.4959\n",
            "Epoch 3/50\n",
            "179/179 [==============================] - 414s 2s/step - loss: 0.7009 - acc: 0.5503 - val_loss: 0.6880 - val_acc: 0.5579\n",
            "Epoch 4/50\n",
            "179/179 [==============================] - 413s 2s/step - loss: 0.6817 - acc: 0.5279 - val_loss: 0.6804 - val_acc: 0.5785\n",
            "Epoch 5/50\n",
            "179/179 [==============================] - 413s 2s/step - loss: 0.6621 - acc: 0.6117 - val_loss: 0.6707 - val_acc: 0.5620\n",
            "Epoch 6/50\n",
            "179/179 [==============================] - 413s 2s/step - loss: 0.6451 - acc: 0.6620 - val_loss: 0.6584 - val_acc: 0.6364\n",
            "Epoch 7/50\n",
            "179/179 [==============================] - 415s 2s/step - loss: 0.6293 - acc: 0.7067 - val_loss: 0.6498 - val_acc: 0.6488\n",
            "Epoch 8/50\n",
            "179/179 [==============================] - 413s 2s/step - loss: 0.6147 - acc: 0.7430 - val_loss: 0.6419 - val_acc: 0.6736\n",
            "Epoch 9/50\n",
            "179/179 [==============================] - 414s 2s/step - loss: 0.6025 - acc: 0.7598 - val_loss: 0.6343 - val_acc: 0.6777\n",
            "Epoch 10/50\n",
            "179/179 [==============================] - 414s 2s/step - loss: 0.5912 - acc: 0.7765 - val_loss: 0.6269 - val_acc: 0.6777\n",
            "Epoch 11/50\n",
            "179/179 [==============================] - 414s 2s/step - loss: 0.5755 - acc: 0.8101 - val_loss: 0.6189 - val_acc: 0.7025\n",
            "Epoch 12/50\n",
            "179/179 [==============================] - 413s 2s/step - loss: 0.5636 - acc: 0.8184 - val_loss: 0.6109 - val_acc: 0.7066\n",
            "Epoch 13/50\n",
            "179/179 [==============================] - 413s 2s/step - loss: 0.5613 - acc: 0.7765 - val_loss: 0.6091 - val_acc: 0.6901\n",
            "Epoch 14/50\n",
            "179/179 [==============================] - 413s 2s/step - loss: 0.5407 - acc: 0.8101 - val_loss: 0.5958 - val_acc: 0.7107\n",
            "Epoch 15/50\n",
            "179/179 [==============================] - 412s 2s/step - loss: 0.5220 - acc: 0.8408 - val_loss: 0.5863 - val_acc: 0.7190\n",
            "Epoch 16/50\n",
            "179/179 [==============================] - 414s 2s/step - loss: 0.5045 - acc: 0.8743 - val_loss: 0.5746 - val_acc: 0.7314\n",
            "Epoch 17/50\n",
            "179/179 [==============================] - 412s 2s/step - loss: 0.4867 - acc: 0.8911 - val_loss: 0.5652 - val_acc: 0.7438\n",
            "Epoch 18/50\n",
            "179/179 [==============================] - 410s 2s/step - loss: 0.4684 - acc: 0.9078 - val_loss: 0.5532 - val_acc: 0.7355\n",
            "Epoch 19/50\n",
            "179/179 [==============================] - 410s 2s/step - loss: 0.4523 - acc: 0.9134 - val_loss: 0.5407 - val_acc: 0.7521\n",
            "Epoch 20/50\n",
            "179/179 [==============================] - 414s 2s/step - loss: 0.4331 - acc: 0.9218 - val_loss: 0.5296 - val_acc: 0.7562\n",
            "Epoch 21/50\n",
            "179/179 [==============================] - 414s 2s/step - loss: 0.4133 - acc: 0.9218 - val_loss: 0.5213 - val_acc: 0.7769\n",
            "Epoch 22/50\n",
            "179/179 [==============================] - 412s 2s/step - loss: 0.3929 - acc: 0.9358 - val_loss: 0.5036 - val_acc: 0.7769\n",
            "Epoch 23/50\n",
            "179/179 [==============================] - 413s 2s/step - loss: 0.3758 - acc: 0.9441 - val_loss: 0.4932 - val_acc: 0.7851\n",
            "Epoch 24/50\n",
            "179/179 [==============================] - 415s 2s/step - loss: 0.3440 - acc: 0.9469 - val_loss: 0.4759 - val_acc: 0.7893\n",
            "Epoch 25/50\n",
            "179/179 [==============================] - 414s 2s/step - loss: 0.3278 - acc: 0.9469 - val_loss: 0.4603 - val_acc: 0.8058\n",
            "Epoch 26/50\n",
            "179/179 [==============================] - 414s 2s/step - loss: 0.3030 - acc: 0.9665 - val_loss: 0.4445 - val_acc: 0.8099\n",
            "Epoch 27/50\n",
            "179/179 [==============================] - 413s 2s/step - loss: 0.2819 - acc: 0.9525 - val_loss: 0.4298 - val_acc: 0.8182\n",
            "Epoch 28/50\n",
            "179/179 [==============================] - 413s 2s/step - loss: 0.2640 - acc: 0.9665 - val_loss: 0.4212 - val_acc: 0.8140\n",
            "Epoch 29/50\n",
            "179/179 [==============================] - 413s 2s/step - loss: 0.2434 - acc: 0.9553 - val_loss: 0.4078 - val_acc: 0.8264\n",
            "Epoch 30/50\n",
            "179/179 [==============================] - 413s 2s/step - loss: 0.2126 - acc: 0.9749 - val_loss: 0.4061 - val_acc: 0.8182\n",
            "Epoch 31/50\n",
            "179/179 [==============================] - 411s 2s/step - loss: 0.1917 - acc: 0.9777 - val_loss: 0.3777 - val_acc: 0.8347\n",
            "Epoch 32/50\n",
            "179/179 [==============================] - 413s 2s/step - loss: 0.1725 - acc: 0.9860 - val_loss: 0.3675 - val_acc: 0.8388\n",
            "Epoch 33/50\n",
            "179/179 [==============================] - 414s 2s/step - loss: 0.1508 - acc: 0.9888 - val_loss: 0.3548 - val_acc: 0.8595\n",
            "Epoch 34/50\n",
            "179/179 [==============================] - 414s 2s/step - loss: 0.1310 - acc: 0.9944 - val_loss: 0.3431 - val_acc: 0.8678\n",
            "Epoch 35/50\n",
            "179/179 [==============================] - 413s 2s/step - loss: 0.1160 - acc: 0.9972 - val_loss: 0.3357 - val_acc: 0.8595\n",
            "Epoch 36/50\n",
            "179/179 [==============================] - 413s 2s/step - loss: 0.1015 - acc: 0.9972 - val_loss: 0.3232 - val_acc: 0.8884\n",
            "Epoch 37/50\n",
            "179/179 [==============================] - 415s 2s/step - loss: 0.0896 - acc: 0.9972 - val_loss: 0.3182 - val_acc: 0.8802\n",
            "Epoch 38/50\n",
            "179/179 [==============================] - 414s 2s/step - loss: 0.0775 - acc: 1.0000 - val_loss: 0.3112 - val_acc: 0.8884\n",
            "Epoch 39/50\n",
            "179/179 [==============================] - 414s 2s/step - loss: 0.0681 - acc: 1.0000 - val_loss: 0.3086 - val_acc: 0.8843\n",
            "Epoch 40/50\n",
            "179/179 [==============================] - 416s 2s/step - loss: 0.0605 - acc: 1.0000 - val_loss: 0.3058 - val_acc: 0.8802\n",
            "Epoch 41/50\n",
            "179/179 [==============================] - 416s 2s/step - loss: 0.0536 - acc: 1.0000 - val_loss: 0.2985 - val_acc: 0.8926\n",
            "Epoch 42/50\n",
            "179/179 [==============================] - 414s 2s/step - loss: 0.0457 - acc: 1.0000 - val_loss: 0.3015 - val_acc: 0.8843\n",
            "Epoch 43/50\n",
            "179/179 [==============================] - 414s 2s/step - loss: 0.0403 - acc: 1.0000 - val_loss: 0.2977 - val_acc: 0.8884\n",
            "Epoch 44/50\n",
            "179/179 [==============================] - 415s 2s/step - loss: 0.0375 - acc: 1.0000 - val_loss: 0.3019 - val_acc: 0.8843\n",
            "Epoch 45/50\n",
            "179/179 [==============================] - 415s 2s/step - loss: 0.0335 - acc: 1.0000 - val_loss: 0.3004 - val_acc: 0.8884\n",
            "Epoch 46/50\n",
            "179/179 [==============================] - 415s 2s/step - loss: 0.0294 - acc: 1.0000 - val_loss: 0.2902 - val_acc: 0.8926\n",
            "Epoch 47/50\n",
            "179/179 [==============================] - 414s 2s/step - loss: 0.0262 - acc: 1.0000 - val_loss: 0.2965 - val_acc: 0.8926\n",
            "Epoch 48/50\n",
            "179/179 [==============================] - 416s 2s/step - loss: 0.0244 - acc: 1.0000 - val_loss: 0.2956 - val_acc: 0.8926\n",
            "Epoch 49/50\n",
            "179/179 [==============================] - 415s 2s/step - loss: 0.0212 - acc: 1.0000 - val_loss: 0.2896 - val_acc: 0.8967\n",
            "Epoch 50/50\n",
            "179/179 [==============================] - 416s 2s/step - loss: 0.0207 - acc: 1.0000 - val_loss: 0.2928 - val_acc: 0.8926\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jGsh_g2e-jJ3",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 545
        },
        "outputId": "aac77bab-5780-4a68-e555-7c70160c851c"
      },
      "source": [
        "acc = history.history['acc']\n",
        "val_acc = history.history['val_acc']\n",
        "loss = history.history['loss']\n",
        "val_loss = history.history['val_loss']\n",
        " \n",
        "epochs = range(len(acc))\n",
        " \n",
        "plt.plot(epochs, acc, 'b', label='Training acc')\n",
        "plt.plot(epochs, val_acc, 'r', label='Validation acc')\n",
        "plt.title('Training and validation accuracy')\n",
        "plt.legend()\n",
        " \n",
        "plt.figure()\n",
        " \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",
        " \n",
        "plt.show()"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAEICAYAAACktLTqAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nLgZ5cHfOuRjkwd0552KQB3fnnItB/w/Z1W3tUKmflQAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAEICAYAAACktLTqAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "W4IXOIz0GwSb",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 351
        },
        "outputId": "9ac10bb8-4b27-44c9-c402-51f0536affe4"
      },
      "source": [
        "base_model1 = applications.VGG16(weights='imagenet', include_top=False, input_shape=(img_rows, img_cols, img_channel))\n",
        "for layer in base_model1.layers[:-12]:\n",
        "    layer.trainable = False\n",
        "    \n",
        "for layer in base_model1.layers:\n",
        "    print(layer, layer.trainable)"
      ],
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "<keras.engine.input_layer.InputLayer object at 0x7f154c171d68> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f154c171a58> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f154c1df550> False\n",
            "<keras.layers.pooling.MaxPooling2D object at 0x7f154c178518> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f154c178588> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f154c1a1da0> False\n",
            "<keras.layers.pooling.MaxPooling2D object at 0x7f154c135710> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f154c135860> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f154c165cf8> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f154c0fcf28> True\n",
            "<keras.layers.pooling.MaxPooling2D object at 0x7f154c110a90> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f154c1290f0> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f154c0c3ac8> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f154c0d4ac8> True\n",
            "<keras.layers.pooling.MaxPooling2D object at 0x7f154c0ecf60> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f154c0866d8> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f154c02f390> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f154c02fc18> True\n",
            "<keras.layers.pooling.MaxPooling2D object at 0x7f154c062be0> True\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Avc2BTUMI6Zy",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 843
        },
        "outputId": "dbd2354a-b2b8-46ae-c393-0b7ff8d6bb5f"
      },
      "source": [
        "add_model = Sequential()\n",
        "add_model.add(Flatten(input_shape=base_model1.output_shape[1:]))\n",
        "add_model.add(Dense(256, activation='relu'))\n",
        "add_model.add(Dense(2, activation='sigmoid'))\n",
        "\n",
        "model2 = Model(inputs=base_model1.input, outputs=add_model(base_model1.output))\n",
        "model2.compile(loss='binary_crossentropy', optimizer=optimizers.SGD(lr=1e-4, momentum=0.9),\n",
        "              metrics=['accuracy'])\n",
        "\n",
        "model2.summary()"
      ],
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "input_5 (InputLayer)         (None, 224, 224, 3)       0         \n",
            "_________________________________________________________________\n",
            "block1_conv1 (Conv2D)        (None, 224, 224, 64)      1792      \n",
            "_________________________________________________________________\n",
            "block1_conv2 (Conv2D)        (None, 224, 224, 64)      36928     \n",
            "_________________________________________________________________\n",
            "block1_pool (MaxPooling2D)   (None, 112, 112, 64)      0         \n",
            "_________________________________________________________________\n",
            "block2_conv1 (Conv2D)        (None, 112, 112, 128)     73856     \n",
            "_________________________________________________________________\n",
            "block2_conv2 (Conv2D)        (None, 112, 112, 128)     147584    \n",
            "_________________________________________________________________\n",
            "block2_pool (MaxPooling2D)   (None, 56, 56, 128)       0         \n",
            "_________________________________________________________________\n",
            "block3_conv1 (Conv2D)        (None, 56, 56, 256)       295168    \n",
            "_________________________________________________________________\n",
            "block3_conv2 (Conv2D)        (None, 56, 56, 256)       590080    \n",
            "_________________________________________________________________\n",
            "block3_conv3 (Conv2D)        (None, 56, 56, 256)       590080    \n",
            "_________________________________________________________________\n",
            "block3_pool (MaxPooling2D)   (None, 28, 28, 256)       0         \n",
            "_________________________________________________________________\n",
            "block4_conv1 (Conv2D)        (None, 28, 28, 512)       1180160   \n",
            "_________________________________________________________________\n",
            "block4_conv2 (Conv2D)        (None, 28, 28, 512)       2359808   \n",
            "_________________________________________________________________\n",
            "block4_conv3 (Conv2D)        (None, 28, 28, 512)       2359808   \n",
            "_________________________________________________________________\n",
            "block4_pool (MaxPooling2D)   (None, 14, 14, 512)       0         \n",
            "_________________________________________________________________\n",
            "block5_conv1 (Conv2D)        (None, 14, 14, 512)       2359808   \n",
            "_________________________________________________________________\n",
            "block5_conv2 (Conv2D)        (None, 14, 14, 512)       2359808   \n",
            "_________________________________________________________________\n",
            "block5_conv3 (Conv2D)        (None, 14, 14, 512)       2359808   \n",
            "_________________________________________________________________\n",
            "block5_pool (MaxPooling2D)   (None, 7, 7, 512)         0         \n",
            "_________________________________________________________________\n",
            "sequential_2 (Sequential)    (None, 2)                 6423298   \n",
            "=================================================================\n",
            "Total params: 21,137,986\n",
            "Trainable params: 20,877,826\n",
            "Non-trainable params: 260,160\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KculvliYI7YF",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "ef45a8b4-c7f2-4d4d-8a7d-b4d12f40e658"
      },
      "source": [
        "batch_size = 32\n",
        "epochs = 50\n",
        "\n",
        "history = model2.fit(train_features_reshaped, train_labels_one_hot, batch_size=batch_size,epochs=epochs,verbose=1,validation_data=(test_features_reshaped, test_labels_one_hot))\n"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Train on 179 samples, validate on 121 samples\n",
            "Epoch 1/50\n",
            "179/179 [==============================] - 304s 2s/step - loss: 0.7103 - acc: 0.4860 - val_loss: 0.7008 - val_acc: 0.5000\n",
            "Epoch 2/50\n",
            "179/179 [==============================] - 300s 2s/step - loss: 0.6996 - acc: 0.4832 - val_loss: 0.6939 - val_acc: 0.5000\n",
            "Epoch 3/50\n",
            "179/179 [==============================] - 302s 2s/step - loss: 0.6778 - acc: 0.5447 - val_loss: 0.6868 - val_acc: 0.5083\n",
            "Epoch 4/50\n",
            "179/179 [==============================] - 302s 2s/step - loss: 0.6683 - acc: 0.5950 - val_loss: 0.6771 - val_acc: 0.5207\n",
            "Epoch 5/50\n",
            "179/179 [==============================] - 301s 2s/step - loss: 0.6495 - acc: 0.6201 - val_loss: 0.6651 - val_acc: 0.6033\n",
            "Epoch 6/50\n",
            "179/179 [==============================] - 301s 2s/step - loss: 0.6370 - acc: 0.6620 - val_loss: 0.6558 - val_acc: 0.5785\n",
            "Epoch 7/50\n",
            "179/179 [==============================] - 302s 2s/step - loss: 0.6215 - acc: 0.6983 - val_loss: 0.6480 - val_acc: 0.6364\n",
            "Epoch 8/50\n",
            "179/179 [==============================] - 302s 2s/step - loss: 0.6102 - acc: 0.7346 - val_loss: 0.6396 - val_acc: 0.6446\n",
            "Epoch 9/50\n",
            "179/179 [==============================] - 302s 2s/step - loss: 0.5979 - acc: 0.7402 - val_loss: 0.6356 - val_acc: 0.6364\n",
            "Epoch 10/50\n",
            "179/179 [==============================] - 303s 2s/step - loss: 0.5834 - acc: 0.7654 - val_loss: 0.6249 - val_acc: 0.6570\n",
            "Epoch 11/50\n",
            "179/179 [==============================] - 303s 2s/step - loss: 0.5688 - acc: 0.7821 - val_loss: 0.6164 - val_acc: 0.6612\n",
            "Epoch 12/50\n",
            "179/179 [==============================] - 302s 2s/step - loss: 0.5533 - acc: 0.8240 - val_loss: 0.6092 - val_acc: 0.6777\n",
            "Epoch 13/50\n",
            "179/179 [==============================] - 302s 2s/step - loss: 0.5419 - acc: 0.8380 - val_loss: 0.5997 - val_acc: 0.6942\n",
            "Epoch 14/50\n",
            "179/179 [==============================] - 301s 2s/step - loss: 0.5235 - acc: 0.8492 - val_loss: 0.5902 - val_acc: 0.6942\n",
            "Epoch 15/50\n",
            "179/179 [==============================] - 303s 2s/step - loss: 0.5093 - acc: 0.8436 - val_loss: 0.5801 - val_acc: 0.7190\n",
            "Epoch 16/50\n",
            "179/179 [==============================] - 303s 2s/step - loss: 0.4967 - acc: 0.8799 - val_loss: 0.5726 - val_acc: 0.7231\n",
            "Epoch 17/50\n",
            "179/179 [==============================] - 303s 2s/step - loss: 0.4763 - acc: 0.9078 - val_loss: 0.5589 - val_acc: 0.7273\n",
            "Epoch 18/50\n",
            "179/179 [==============================] - 303s 2s/step - loss: 0.4621 - acc: 0.8855 - val_loss: 0.5478 - val_acc: 0.7355\n",
            "Epoch 19/50\n",
            "179/179 [==============================] - 302s 2s/step - loss: 0.4426 - acc: 0.8994 - val_loss: 0.5365 - val_acc: 0.7438\n",
            "Epoch 20/50\n",
            "179/179 [==============================] - 303s 2s/step - loss: 0.4269 - acc: 0.9274 - val_loss: 0.5278 - val_acc: 0.7479\n",
            "Epoch 21/50\n",
            "179/179 [==============================] - 303s 2s/step - loss: 0.4063 - acc: 0.9302 - val_loss: 0.5127 - val_acc: 0.7645\n",
            "Epoch 22/50\n",
            "179/179 [==============================] - 304s 2s/step - loss: 0.3915 - acc: 0.9162 - val_loss: 0.4997 - val_acc: 0.7645\n",
            "Epoch 23/50\n",
            "179/179 [==============================] - 304s 2s/step - loss: 0.3730 - acc: 0.9441 - val_loss: 0.4912 - val_acc: 0.8017\n",
            "Epoch 24/50\n",
            "179/179 [==============================] - 304s 2s/step - loss: 0.3575 - acc: 0.9413 - val_loss: 0.4751 - val_acc: 0.7686\n",
            "Epoch 25/50\n",
            "179/179 [==============================] - 305s 2s/step - loss: 0.3374 - acc: 0.9581 - val_loss: 0.4674 - val_acc: 0.8182\n",
            "Epoch 26/50\n",
            "179/179 [==============================] - 304s 2s/step - loss: 0.3164 - acc: 0.9693 - val_loss: 0.4526 - val_acc: 0.7810\n",
            "Epoch 27/50\n",
            "179/179 [==============================] - 304s 2s/step - loss: 0.2959 - acc: 0.9721 - val_loss: 0.4403 - val_acc: 0.8140\n",
            "Epoch 28/50\n",
            "179/179 [==============================] - 304s 2s/step - loss: 0.2813 - acc: 0.9693 - val_loss: 0.4296 - val_acc: 0.8264\n",
            "Epoch 29/50\n",
            "179/179 [==============================] - 304s 2s/step - loss: 0.2588 - acc: 0.9777 - val_loss: 0.4213 - val_acc: 0.7810\n",
            "Epoch 30/50\n",
            "179/179 [==============================] - 303s 2s/step - loss: 0.2406 - acc: 0.9804 - val_loss: 0.4094 - val_acc: 0.8388\n",
            "Epoch 31/50\n",
            "179/179 [==============================] - 304s 2s/step - loss: 0.2271 - acc: 0.9804 - val_loss: 0.3973 - val_acc: 0.8430\n",
            "Epoch 32/50\n",
            "179/179 [==============================] - 303s 2s/step - loss: 0.2107 - acc: 0.9860 - val_loss: 0.3913 - val_acc: 0.8058\n",
            "Epoch 33/50\n",
            "179/179 [==============================] - 304s 2s/step - loss: 0.1930 - acc: 0.9860 - val_loss: 0.3831 - val_acc: 0.8306\n",
            "Epoch 34/50\n",
            "179/179 [==============================] - 303s 2s/step - loss: 0.1769 - acc: 0.9832 - val_loss: 0.3710 - val_acc: 0.8388\n",
            "Epoch 35/50\n",
            "179/179 [==============================] - 303s 2s/step - loss: 0.1624 - acc: 0.9916 - val_loss: 0.3666 - val_acc: 0.8388\n",
            "Epoch 36/50\n",
            "179/179 [==============================] - 303s 2s/step - loss: 0.1483 - acc: 0.9944 - val_loss: 0.3605 - val_acc: 0.8430\n",
            "Epoch 37/50\n",
            "179/179 [==============================] - 303s 2s/step - loss: 0.1360 - acc: 0.9944 - val_loss: 0.3490 - val_acc: 0.8554\n",
            "Epoch 38/50\n",
            "179/179 [==============================] - 303s 2s/step - loss: 0.1231 - acc: 0.9944 - val_loss: 0.3422 - val_acc: 0.8636\n",
            "Epoch 39/50\n",
            "179/179 [==============================] - 303s 2s/step - loss: 0.1133 - acc: 0.9972 - val_loss: 0.3405 - val_acc: 0.8636\n",
            "Epoch 40/50\n",
            "179/179 [==============================] - 304s 2s/step - loss: 0.1007 - acc: 0.9972 - val_loss: 0.3345 - val_acc: 0.8554\n",
            "Epoch 41/50\n",
            "179/179 [==============================] - 304s 2s/step - loss: 0.0933 - acc: 1.0000 - val_loss: 0.3300 - val_acc: 0.8636\n",
            "Epoch 42/50\n",
            "179/179 [==============================] - 304s 2s/step - loss: 0.0842 - acc: 1.0000 - val_loss: 0.3259 - val_acc: 0.8678\n",
            "Epoch 43/50\n",
            "179/179 [==============================] - 304s 2s/step - loss: 0.0760 - acc: 1.0000 - val_loss: 0.3262 - val_acc: 0.8678\n",
            "Epoch 44/50\n",
            "179/179 [==============================] - 304s 2s/step - loss: 0.0687 - acc: 1.0000 - val_loss: 0.3216 - val_acc: 0.8595\n",
            "Epoch 45/50\n",
            "179/179 [==============================] - 304s 2s/step - loss: 0.0629 - acc: 1.0000 - val_loss: 0.3194 - val_acc: 0.8719\n",
            "Epoch 46/50\n",
            "179/179 [==============================] - 304s 2s/step - loss: 0.0579 - acc: 1.0000 - val_loss: 0.3165 - val_acc: 0.8636\n",
            "Epoch 47/50\n",
            "179/179 [==============================] - 304s 2s/step - loss: 0.0518 - acc: 1.0000 - val_loss: 0.3155 - val_acc: 0.8719\n",
            "Epoch 48/50\n",
            "179/179 [==============================] - 304s 2s/step - loss: 0.0483 - acc: 1.0000 - val_loss: 0.3161 - val_acc: 0.8760\n",
            "Epoch 49/50\n",
            "179/179 [==============================] - 304s 2s/step - loss: 0.0442 - acc: 1.0000 - val_loss: 0.3136 - val_acc: 0.8760\n",
            "Epoch 50/50\n",
            "179/179 [==============================] - 304s 2s/step - loss: 0.0404 - acc: 1.0000 - val_loss: 0.3133 - val_acc: 0.8802\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yEzpO-MnJZJ-",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 545
        },
        "outputId": "81bc05d8-9462-4e52-dc77-9548065356c7"
      },
      "source": [
        "acc = history.history['acc']\n",
        "val_acc = history.history['val_acc']\n",
        "loss = history.history['loss']\n",
        "val_loss = history.history['val_loss']\n",
        " \n",
        "epochs = range(len(acc))\n",
        " \n",
        "plt.plot(epochs, acc, 'b', label='Training acc')\n",
        "plt.plot(epochs, val_acc, 'r', label='Validation acc')\n",
        "plt.title('Training and validation accuracy')\n",
        "plt.legend()\n",
        " \n",
        "plt.figure()\n",
        " \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",
        " \n",
        "plt.show()"
      ],
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAEICAYAAACktLTqAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAEICAYAAACktLTqAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SxuxrPsTH-VE",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}