Keras_CNN_CIFAR10_Exam.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Keras_CNN_CIFAR10_Exam.ipynb",
      "version": "0.3.2",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.6.8"
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/Tejas1415/Cifar10-image-classification-using-MLP-in-TF-backend./blob/master/Keras_CNN_CIFAR10_Exam.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pgu7oI-eqJ6r",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from __future__ import print_function\n",
        "import keras\n",
        "from keras.datasets import mnist, cifar10\n",
        "from keras.layers import Dense, Flatten\n",
        "from keras.layers import Conv2D, MaxPooling2D\n",
        "from keras.models import Sequential\n",
        "import matplotlib.pylab as plt\n",
        "import numpy as np \n",
        "\n",
        "# load the data already splits into train and test data sets\n",
        "(x_train, y_train), (x_test, y_test) = cifar10.load_data()\n",
        "\n",
        "# input image dimensions\n",
        "img_x, img_y, channels = 32, 32, 3\n",
        "\n",
        "keras.backend.set_image_data_format('channels_last')\n",
        "input_shape = (img_x, img_y, channels)\n",
        "# reshape the data into a 4D tensor - (sample_number, x_img_size, y_img_size, num_channels)\n",
        "x_train = x_train.reshape(x_train.shape[0], img_x, img_y, channels)\n",
        "x_test = x_test.reshape(x_test.shape[0], img_x, img_y, channels)\n",
        "\n",
        "cifar10_labels={\n",
        "    0: \"airplane\", \n",
        "    1: \"automobile\", \n",
        "    2: \"bird\",\n",
        "    3: \"cat\",\n",
        "    4: \"deer\",\n",
        "    5: \"dog\",\n",
        "    6: \"frog\",\n",
        "    7: \"horse\",\n",
        "    8: \"ship\",\n",
        "    9: \"truck\"\n",
        "}\n",
        "\n",
        "num_classes = len(cifar10_labels)\n",
        "\n",
        "# convert class vectors to binary class matrices for categorical_crossentropy loss\n",
        "y_train = keras.utils.to_categorical(y_train, num_classes)\n",
        "y_test = keras.utils.to_categorical(y_test, num_classes)\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HnaM54FkqJ6v",
        "colab_type": "code",
        "outputId": "da046c78-d5a9-4ea4-a6e6-668f7035d353",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "x_train.shape, y_train.shape"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((50000, 32, 32, 3), (50000, 10))"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 36
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DUTk6xDZqJ6y",
        "colab_type": "code",
        "outputId": "f6df317d-ca6d-428a-a2ab-fa206d88d57e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "x_test.shape, y_test.shape"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((10000, 32, 32, 3), (10000, 10))"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 37
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jr1CtEekqJ61",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8_xRWJh6SKhV",
        "colab_type": "text"
      },
      "source": [
        "# Converting the images (x_train and x_test) to float32\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oAuHJ-JuqJ64",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "new_x_train = np.array(x_train, dtype='f')\n",
        "new_x_test = np.array(x_test, dtype='f')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qA8dHQouzSLU",
        "colab_type": "code",
        "outputId": "5e9a8f5a-64ff-4448-84be-b6b6dda18e0f",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "new_x_train.dtype, new_x_test.dtype"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(dtype('float32'), dtype('float32'))"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8PJSfXXOSR3V",
        "colab_type": "text"
      },
      "source": [
        "# Scale the pixel values (x_train and x_test) from range 0..255 to 0..1\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4czubIKlqJ66",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "scaled_x_train = new_x_train / 255.0\n",
        "scaled_x_test = new_x_test / 255.0"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kTJdi6ErSYoa",
        "colab_type": "text"
      },
      "source": [
        "# Model1: Create the CNN architecture provided\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8vl1QdG4qJ68",
        "colab_type": "code",
        "outputId": "4deabc11-aaf9-4d85-cebe-7f08749c08d7",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 88
        }
      },
      "source": [
        "#Layer 1\n",
        "model1 = Sequential()\n",
        "model1.add(Conv2D(32, kernel_size=(3, 3), activation='relu', input_shape=(32, 32, 3)))\n",
        "\n",
        "#Flatten\n",
        "model1.add(Flatten())\n",
        "\n",
        "#Softmax classifier\n",
        "model1.add(Dense(10, activation='softmax'))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow/python/framework/op_def_library.py:263: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Colocations handled automatically by placer.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RjDUqHCySgkA",
        "colab_type": "text"
      },
      "source": [
        "## Display the layers and the size of the outputs. Should match the architecture provided for Model 1.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vi8Ko836qJ6-",
        "colab_type": "code",
        "outputId": "f2cf9c05-c2cf-4212-a74f-8fa071c6a1dd",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 238
        }
      },
      "source": [
        "model1.summary()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_1 (Conv2D)            (None, 30, 30, 32)        896       \n",
            "_________________________________________________________________\n",
            "flatten_1 (Flatten)          (None, 28800)             0         \n",
            "_________________________________________________________________\n",
            "dense_1 (Dense)              (None, 10)                288010    \n",
            "=================================================================\n",
            "Total params: 288,906\n",
            "Trainable params: 288,906\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6Pg7yNPrTHCV",
        "colab_type": "text"
      },
      "source": [
        "## Train your model1. You should achieve less than 50% accuracy on the test set."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SrUCujpUqJ7A",
        "colab_type": "code",
        "outputId": "141ef75e-4963-47fb-a1f1-a13d9069737a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 428
        }
      },
      "source": [
        "from keras.callbacks import EarlyStopping\n",
        "from keras.optimizers import Adam\n",
        "\n",
        "model1.compile(loss='categorical_crossentropy',\n",
        "                  optimizer=Adam(lr=0.01, decay=1e-6),\n",
        "                  metrics=['accuracy'])\n",
        "\n",
        "model1.fit(scaled_x_train, y_train,\n",
        "              batch_size=128,\n",
        "              shuffle=True,\n",
        "              epochs=15,\n",
        "              validation_data=(scaled_x_test, y_test),\n",
        "              callbacks=[EarlyStopping(min_delta=0.001, patience=3)])"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow/python/ops/math_ops.py:3066: to_int32 (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Use tf.cast instead.\n",
            "Train on 50000 samples, validate on 10000 samples\n",
            "Epoch 1/15\n",
            "50000/50000 [==============================] - 5s 109us/step - loss: 1.9537 - acc: 0.3284 - val_loss: 1.8347 - val_acc: 0.3343\n",
            "Epoch 2/15\n",
            "50000/50000 [==============================] - 3s 53us/step - loss: 1.6605 - acc: 0.4122 - val_loss: 1.8757 - val_acc: 0.3683\n",
            "Epoch 3/15\n",
            "50000/50000 [==============================] - 3s 53us/step - loss: 1.5759 - acc: 0.4415 - val_loss: 1.6798 - val_acc: 0.3944\n",
            "Epoch 4/15\n",
            "50000/50000 [==============================] - 3s 53us/step - loss: 1.5238 - acc: 0.4607 - val_loss: 1.7393 - val_acc: 0.4164\n",
            "Epoch 5/15\n",
            "50000/50000 [==============================] - 3s 53us/step - loss: 1.4883 - acc: 0.4762 - val_loss: 1.5483 - val_acc: 0.4588\n",
            "Epoch 6/15\n",
            "50000/50000 [==============================] - 3s 53us/step - loss: 1.4415 - acc: 0.4928 - val_loss: 1.5284 - val_acc: 0.4589\n",
            "Epoch 7/15\n",
            "50000/50000 [==============================] - 3s 53us/step - loss: 1.4017 - acc: 0.5085 - val_loss: 1.5287 - val_acc: 0.4635\n",
            "Epoch 8/15\n",
            "50000/50000 [==============================] - 3s 53us/step - loss: 1.3767 - acc: 0.5179 - val_loss: 1.5849 - val_acc: 0.4465\n",
            "Epoch 9/15\n",
            "50000/50000 [==============================] - 3s 54us/step - loss: 1.3552 - acc: 0.5273 - val_loss: 1.5933 - val_acc: 0.4592\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7fa6257fa2b0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Nq22zOMHRI7x",
        "colab_type": "text"
      },
      "source": [
        "## Plot your losses and/or accuracy"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WqonxHI-qJ7C",
        "colab_type": "code",
        "outputId": "8f785cf1-f089-44f7-fcfb-b15edd2804e0",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 573
        }
      },
      "source": [
        "def plot(model):\n",
        "    # Plot training & validation accuracy values\n",
        "    plt.plot(model.history.history['acc'])\n",
        "    plt.plot(model.history.history['val_acc'])\n",
        "    plt.title('Model accuracy')\n",
        "    plt.ylabel('Accuracy')\n",
        "    plt.xlabel('Epoch')\n",
        "    plt.legend(['Train', 'Test'], loc='upper left')\n",
        "    plt.show()\n",
        "\n",
        "    # Plot training & validation loss values\n",
        "    plt.plot(model.history.history['loss'])\n",
        "    plt.plot(model.history.history['val_loss'])\n",
        "    plt.title('Model loss')\n",
        "    plt.ylabel('Loss')\n",
        "    plt.xlabel('Epoch')\n",
        "    plt.legend(['Train', 'Test'], loc='upper left')\n",
        "    plt.show()\n",
        "plot(model1)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZIAAAEWCAYAAABMoxE0AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEWCAYAAACJ0YulAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSiml2mgoKKWUavP/Ad6f8TVzm06CAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Qzs77pLRTXvG",
        "colab_type": "text"
      },
      "source": [
        "# Model 2: Modify the model by adding at least one more layer consisting of convolutions and MaxPooling. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-EGIcvCOqJ7F",
        "colab_type": "code",
        "outputId": "ac57c544-95f2-470d-b576-1fa16e1272c6",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 340
        }
      },
      "source": [
        "#Creating the model with one more conv layer, maxpooling \n",
        "model2 = Sequential()\n",
        "\n",
        "model2.add(Conv2D(32, kernel_size=(3, 3), activation='relu', input_shape=(32, 32, 3)))\n",
        "model2.add(Conv2D(64, kernel_size=(3, 3), activation='relu'))\n",
        "model2.add(MaxPooling2D(pool_size=(2, 2)))\n",
        "model2.add(Dropout(0.25))\n",
        "\n",
        "#Flatten\n",
        "model2.add(Flatten())\n",
        "\n",
        "model2.add(Dense(10, activation='softmax'))\n",
        "\n",
        "model2.summary()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_8 (Conv2D)            (None, 30, 30, 32)        896       \n",
            "_________________________________________________________________\n",
            "conv2d_9 (Conv2D)            (None, 28, 28, 64)        18496     \n",
            "_________________________________________________________________\n",
            "max_pooling2d_4 (MaxPooling2 (None, 14, 14, 64)        0         \n",
            "_________________________________________________________________\n",
            "dropout_7 (Dropout)          (None, 14, 14, 64)        0         \n",
            "_________________________________________________________________\n",
            "flatten_8 (Flatten)          (None, 12544)             0         \n",
            "_________________________________________________________________\n",
            "dense_14 (Dense)             (None, 10)                125450    \n",
            "=================================================================\n",
            "Total params: 144,842\n",
            "Trainable params: 144,842\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Yjs71ln8WVB2",
        "colab_type": "text"
      },
      "source": [
        "## Model2: Training"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "icgrzydGTjBy",
        "colab_type": "code",
        "outputId": "a7b1877c-9ad2-481d-9b3c-eccaa0840887",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1751
        }
      },
      "source": [
        "# Compile the model\n",
        "model2.compile(loss='categorical_crossentropy',\n",
        "                  optimizer=Adam(lr=0.0001, decay=1e-6),\n",
        "                  metrics=['accuracy'])\n",
        "# training\n",
        "model2.fit(scaled_x_train, y_train,\n",
        "              batch_size=128,\n",
        "              shuffle=True,\n",
        "              epochs=50,\n",
        "              validation_data=(scaled_x_test, y_test),\n",
        "              callbacks=[EarlyStopping(min_delta=0.001, patience=3)])"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Train on 50000 samples, validate on 10000 samples\n",
            "Epoch 1/50\n",
            "50000/50000 [==============================] - 5s 108us/step - loss: 1.8711 - acc: 0.3406 - val_loss: 1.6285 - val_acc: 0.4380\n",
            "Epoch 2/50\n",
            "50000/50000 [==============================] - 5s 96us/step - loss: 1.5658 - acc: 0.4514 - val_loss: 1.4945 - val_acc: 0.4743\n",
            "Epoch 3/50\n",
            "50000/50000 [==============================] - 5s 96us/step - loss: 1.4536 - acc: 0.4914 - val_loss: 1.4068 - val_acc: 0.5092\n",
            "Epoch 4/50\n",
            "50000/50000 [==============================] - 5s 95us/step - loss: 1.3724 - acc: 0.5207 - val_loss: 1.3433 - val_acc: 0.5306\n",
            "Epoch 5/50\n",
            "50000/50000 [==============================] - 5s 95us/step - loss: 1.3118 - acc: 0.5443 - val_loss: 1.2864 - val_acc: 0.5566\n",
            "Epoch 6/50\n",
            "50000/50000 [==============================] - 5s 95us/step - loss: 1.2618 - acc: 0.5623 - val_loss: 1.2470 - val_acc: 0.5696\n",
            "Epoch 7/50\n",
            "50000/50000 [==============================] - 5s 96us/step - loss: 1.2224 - acc: 0.5782 - val_loss: 1.2177 - val_acc: 0.5786\n",
            "Epoch 8/50\n",
            "50000/50000 [==============================] - 5s 96us/step - loss: 1.1908 - acc: 0.5881 - val_loss: 1.2014 - val_acc: 0.5828\n",
            "Epoch 9/50\n",
            "50000/50000 [==============================] - 5s 95us/step - loss: 1.1624 - acc: 0.5992 - val_loss: 1.1727 - val_acc: 0.5964\n",
            "Epoch 10/50\n",
            "50000/50000 [==============================] - 5s 96us/step - loss: 1.1364 - acc: 0.6095 - val_loss: 1.1495 - val_acc: 0.6041\n",
            "Epoch 11/50\n",
            "50000/50000 [==============================] - 5s 95us/step - loss: 1.1146 - acc: 0.6177 - val_loss: 1.1283 - val_acc: 0.6093\n",
            "Epoch 12/50\n",
            "50000/50000 [==============================] - 5s 95us/step - loss: 1.0940 - acc: 0.6246 - val_loss: 1.1089 - val_acc: 0.6158\n",
            "Epoch 13/50\n",
            "50000/50000 [==============================] - 5s 95us/step - loss: 1.0759 - acc: 0.6321 - val_loss: 1.1175 - val_acc: 0.6169\n",
            "Epoch 14/50\n",
            "50000/50000 [==============================] - 5s 95us/step - loss: 1.0583 - acc: 0.6359 - val_loss: 1.0905 - val_acc: 0.6243\n",
            "Epoch 15/50\n",
            "50000/50000 [==============================] - 5s 95us/step - loss: 1.0418 - acc: 0.6432 - val_loss: 1.0834 - val_acc: 0.6303\n",
            "Epoch 16/50\n",
            "50000/50000 [==============================] - 5s 95us/step - loss: 1.0282 - acc: 0.6467 - val_loss: 1.0633 - val_acc: 0.6341\n",
            "Epoch 17/50\n",
            "50000/50000 [==============================] - 5s 95us/step - loss: 1.0125 - acc: 0.6542 - val_loss: 1.0556 - val_acc: 0.6370\n",
            "Epoch 18/50\n",
            "50000/50000 [==============================] - 5s 95us/step - loss: 1.0001 - acc: 0.6584 - val_loss: 1.0502 - val_acc: 0.6374\n",
            "Epoch 19/50\n",
            "50000/50000 [==============================] - 5s 95us/step - loss: 0.9873 - acc: 0.6632 - val_loss: 1.0428 - val_acc: 0.6433\n",
            "Epoch 20/50\n",
            "50000/50000 [==============================] - 5s 95us/step - loss: 0.9762 - acc: 0.6669 - val_loss: 1.0261 - val_acc: 0.6448\n",
            "Epoch 21/50\n",
            "50000/50000 [==============================] - 5s 95us/step - loss: 0.9643 - acc: 0.6706 - val_loss: 1.0231 - val_acc: 0.6445\n",
            "Epoch 22/50\n",
            "50000/50000 [==============================] - 5s 95us/step - loss: 0.9534 - acc: 0.6742 - val_loss: 1.0188 - val_acc: 0.6477\n",
            "Epoch 23/50\n",
            "50000/50000 [==============================] - 5s 95us/step - loss: 0.9427 - acc: 0.6765 - val_loss: 1.0139 - val_acc: 0.6524\n",
            "Epoch 24/50\n",
            "50000/50000 [==============================] - 5s 95us/step - loss: 0.9342 - acc: 0.6809 - val_loss: 1.0137 - val_acc: 0.6522\n",
            "Epoch 25/50\n",
            "50000/50000 [==============================] - 5s 97us/step - loss: 0.9281 - acc: 0.6832 - val_loss: 1.0104 - val_acc: 0.6531\n",
            "Epoch 26/50\n",
            "50000/50000 [==============================] - 5s 96us/step - loss: 0.9214 - acc: 0.6855 - val_loss: 0.9956 - val_acc: 0.6556\n",
            "Epoch 27/50\n",
            "50000/50000 [==============================] - 5s 96us/step - loss: 0.9122 - acc: 0.6879 - val_loss: 0.9969 - val_acc: 0.6546\n",
            "Epoch 28/50\n",
            "50000/50000 [==============================] - 5s 95us/step - loss: 0.9033 - acc: 0.6912 - val_loss: 0.9939 - val_acc: 0.6558\n",
            "Epoch 29/50\n",
            "50000/50000 [==============================] - 5s 95us/step - loss: 0.8963 - acc: 0.6963 - val_loss: 0.9821 - val_acc: 0.6611\n",
            "Epoch 30/50\n",
            "50000/50000 [==============================] - 5s 95us/step - loss: 0.8892 - acc: 0.6961 - val_loss: 0.9716 - val_acc: 0.6662\n",
            "Epoch 31/50\n",
            "50000/50000 [==============================] - 5s 96us/step - loss: 0.8820 - acc: 0.6978 - val_loss: 0.9764 - val_acc: 0.6629\n",
            "Epoch 32/50\n",
            "50000/50000 [==============================] - 5s 95us/step - loss: 0.8755 - acc: 0.7009 - val_loss: 0.9664 - val_acc: 0.6637\n",
            "Epoch 33/50\n",
            "50000/50000 [==============================] - 5s 95us/step - loss: 0.8722 - acc: 0.7008 - val_loss: 0.9755 - val_acc: 0.6655\n",
            "Epoch 34/50\n",
            "50000/50000 [==============================] - 5s 95us/step - loss: 0.8647 - acc: 0.7041 - val_loss: 0.9588 - val_acc: 0.6688\n",
            "Epoch 35/50\n",
            "50000/50000 [==============================] - 5s 95us/step - loss: 0.8580 - acc: 0.7072 - val_loss: 0.9724 - val_acc: 0.6669\n",
            "Epoch 36/50\n",
            "50000/50000 [==============================] - 5s 95us/step - loss: 0.8504 - acc: 0.7093 - val_loss: 0.9601 - val_acc: 0.6692\n",
            "Epoch 37/50\n",
            "50000/50000 [==============================] - 5s 95us/step - loss: 0.8452 - acc: 0.7116 - val_loss: 0.9570 - val_acc: 0.6696\n",
            "Epoch 38/50\n",
            "50000/50000 [==============================] - 5s 94us/step - loss: 0.8426 - acc: 0.7128 - val_loss: 0.9581 - val_acc: 0.6699\n",
            "Epoch 39/50\n",
            "50000/50000 [==============================] - 5s 94us/step - loss: 0.8373 - acc: 0.7121 - val_loss: 0.9535 - val_acc: 0.6719\n",
            "Epoch 40/50\n",
            "50000/50000 [==============================] - 5s 95us/step - loss: 0.8306 - acc: 0.7165 - val_loss: 0.9474 - val_acc: 0.6741\n",
            "Epoch 41/50\n",
            "50000/50000 [==============================] - 5s 95us/step - loss: 0.8237 - acc: 0.7180 - val_loss: 0.9609 - val_acc: 0.6691\n",
            "Epoch 42/50\n",
            "50000/50000 [==============================] - 5s 95us/step - loss: 0.8219 - acc: 0.7180 - val_loss: 0.9411 - val_acc: 0.6745\n",
            "Epoch 43/50\n",
            "50000/50000 [==============================] - 5s 94us/step - loss: 0.8166 - acc: 0.7198 - val_loss: 0.9476 - val_acc: 0.6743\n",
            "Epoch 44/50\n",
            "50000/50000 [==============================] - 5s 95us/step - loss: 0.8106 - acc: 0.7240 - val_loss: 0.9404 - val_acc: 0.6752\n",
            "Epoch 45/50\n",
            "50000/50000 [==============================] - 5s 94us/step - loss: 0.8085 - acc: 0.7239 - val_loss: 0.9363 - val_acc: 0.6777\n",
            "Epoch 46/50\n",
            "50000/50000 [==============================] - 5s 94us/step - loss: 0.7996 - acc: 0.7271 - val_loss: 0.9434 - val_acc: 0.6744\n",
            "Epoch 47/50\n",
            "50000/50000 [==============================] - 5s 95us/step - loss: 0.7975 - acc: 0.7283 - val_loss: 0.9319 - val_acc: 0.6769\n",
            "Epoch 48/50\n",
            "50000/50000 [==============================] - 5s 94us/step - loss: 0.7938 - acc: 0.7291 - val_loss: 0.9340 - val_acc: 0.6753\n",
            "Epoch 49/50\n",
            "50000/50000 [==============================] - 5s 95us/step - loss: 0.7907 - acc: 0.7285 - val_loss: 0.9345 - val_acc: 0.6762\n",
            "Epoch 50/50\n",
            "50000/50000 [==============================] - 5s 94us/step - loss: 0.7866 - acc: 0.7337 - val_loss: 0.9307 - val_acc: 0.6786\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7fa4e7c20710>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 45
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hftHGMTpTj1K",
        "colab_type": "text"
      },
      "source": [
        "## Model2: Plot losses and/or accuracy"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-wGZm_lbTjFk",
        "colab_type": "code",
        "outputId": "290cf3a8-73a5-4ab0-b690-32df63de68be",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 573
        }
      },
      "source": [
        "plot(model2)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYsAAAEWCAYAAACXGLsWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nWbIwxhjTI0sWxhhjevT/AVzEosghppI1AAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEWCAYAAACJ0YulAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nhYIxxphuFgrGGGO6WSgYY4zp9v8B1Ew+6qWlVMwAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hJMmaLw-WhyM",
        "colab_type": "text"
      },
      "source": [
        "# Model3: Find the smallest model you can that achieves 99% accuracy or better. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": false,
        "id": "-27rj-PrqJ7H",
        "colab_type": "code",
        "outputId": "bf530b8f-f45d-4bee-af3c-e6c63b521c1c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 578
        }
      },
      "source": [
        "#Creating the model with one more conv layer, maxpooling \n",
        "model3 = Sequential()\n",
        "\n",
        "model3.add(Conv2D(32, kernel_size=(3, 3), activation='relu', input_shape=(32, 32, 3)))\n",
        "model3.add(Conv2D(64, kernel_size=(3, 3), activation='relu'))\n",
        "model3.add(MaxPooling2D(pool_size=(2, 2)))\n",
        "model3.add(Dropout(0.25))\n",
        "\n",
        "model3.add(Conv2D(128, kernel_size=(3, 3), activation='relu'))\n",
        "model3.add(MaxPooling2D(pool_size=(2, 2)))\n",
        "model3.add(Conv2D(128, kernel_size=(3, 3), activation='relu'))\n",
        "model3.add(MaxPooling2D(pool_size=(2, 2)))\n",
        "model3.add(Dropout(0.25))\n",
        "\n",
        "model3.add(Flatten())\n",
        "model3.add(Dense(1024, activation='relu'))\n",
        "model3.add(Dropout(0.5))\n",
        "model3.add(Dense(10, activation='softmax'))\n",
        "\n",
        "\n",
        "# #Flatten\n",
        "# model3.add(Flatten())\n",
        "\n",
        "# model3.add(Dense(10, activation='softmax'))\n",
        "\n",
        "model3.summary()\n"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_10 (Conv2D)           (None, 30, 30, 32)        896       \n",
            "_________________________________________________________________\n",
            "conv2d_11 (Conv2D)           (None, 28, 28, 64)        18496     \n",
            "_________________________________________________________________\n",
            "max_pooling2d_5 (MaxPooling2 (None, 14, 14, 64)        0         \n",
            "_________________________________________________________________\n",
            "dropout_8 (Dropout)          (None, 14, 14, 64)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_12 (Conv2D)           (None, 12, 12, 128)       73856     \n",
            "_________________________________________________________________\n",
            "max_pooling2d_6 (MaxPooling2 (None, 6, 6, 128)         0         \n",
            "_________________________________________________________________\n",
            "conv2d_13 (Conv2D)           (None, 4, 4, 128)         147584    \n",
            "_________________________________________________________________\n",
            "max_pooling2d_7 (MaxPooling2 (None, 2, 2, 128)         0         \n",
            "_________________________________________________________________\n",
            "dropout_9 (Dropout)          (None, 2, 2, 128)         0         \n",
            "_________________________________________________________________\n",
            "flatten_9 (Flatten)          (None, 512)               0         \n",
            "_________________________________________________________________\n",
            "dense_15 (Dense)             (None, 1024)              525312    \n",
            "_________________________________________________________________\n",
            "dropout_10 (Dropout)         (None, 1024)              0         \n",
            "_________________________________________________________________\n",
            "dense_16 (Dense)             (None, 10)                10250     \n",
            "=================================================================\n",
            "Total params: 776,394\n",
            "Trainable params: 776,394\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0SuYvRoLXPto",
        "colab_type": "text"
      },
      "source": [
        "## Model3: training"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6jivFtMMXQGs",
        "colab_type": "code",
        "outputId": "3bc0b381-4645-4332-de4f-2aeec3ee4889",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1717
        }
      },
      "source": [
        "# Compile the model\n",
        "model3.compile(loss='categorical_crossentropy',\n",
        "              optimizer=Adam(lr=0.0001, decay=1e-6),\n",
        "              metrics=['accuracy'])\n",
        "\n",
        "# training\n",
        "model3.fit(scaled_x_train, y_train,\n",
        "              batch_size=128,\n",
        "              shuffle=True,\n",
        "              epochs=150,\n",
        "              validation_data=(scaled_x_test, y_test),\n",
        "              callbacks=[EarlyStopping(min_delta=0.001, patience=3)])"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Train on 50000 samples, validate on 10000 samples\n",
            "Epoch 1/150\n",
            "50000/50000 [==============================] - 7s 144us/step - loss: 2.0023 - acc: 0.2486 - val_loss: 1.7467 - val_acc: 0.3638\n",
            "Epoch 2/150\n",
            "50000/50000 [==============================] - 6s 128us/step - loss: 1.6703 - acc: 0.3774 - val_loss: 1.5339 - val_acc: 0.4444\n",
            "Epoch 3/150\n",
            "50000/50000 [==============================] - 6s 128us/step - loss: 1.5249 - acc: 0.4375 - val_loss: 1.4263 - val_acc: 0.4899\n",
            "Epoch 4/150\n",
            "50000/50000 [==============================] - 6s 128us/step - loss: 1.4357 - acc: 0.4745 - val_loss: 1.3414 - val_acc: 0.5242\n",
            "Epoch 5/150\n",
            "50000/50000 [==============================] - 6s 128us/step - loss: 1.3694 - acc: 0.5024 - val_loss: 1.2749 - val_acc: 0.5422\n",
            "Epoch 6/150\n",
            "50000/50000 [==============================] - 6s 128us/step - loss: 1.3089 - acc: 0.5280 - val_loss: 1.2326 - val_acc: 0.5703\n",
            "Epoch 7/150\n",
            "50000/50000 [==============================] - 6s 127us/step - loss: 1.2567 - acc: 0.5474 - val_loss: 1.1672 - val_acc: 0.5895\n",
            "Epoch 8/150\n",
            "50000/50000 [==============================] - 6s 128us/step - loss: 1.2090 - acc: 0.5663 - val_loss: 1.1259 - val_acc: 0.6028\n",
            "Epoch 9/150\n",
            "50000/50000 [==============================] - 6s 128us/step - loss: 1.1698 - acc: 0.5858 - val_loss: 1.0945 - val_acc: 0.6126\n",
            "Epoch 10/150\n",
            "50000/50000 [==============================] - 6s 128us/step - loss: 1.1288 - acc: 0.5993 - val_loss: 1.0643 - val_acc: 0.6251\n",
            "Epoch 11/150\n",
            "50000/50000 [==============================] - 6s 127us/step - loss: 1.0969 - acc: 0.6112 - val_loss: 1.0507 - val_acc: 0.6292\n",
            "Epoch 12/150\n",
            "50000/50000 [==============================] - 6s 128us/step - loss: 1.0716 - acc: 0.6197 - val_loss: 1.0355 - val_acc: 0.6386\n",
            "Epoch 13/150\n",
            "50000/50000 [==============================] - 6s 127us/step - loss: 1.0375 - acc: 0.6351 - val_loss: 0.9891 - val_acc: 0.6537\n",
            "Epoch 14/150\n",
            "50000/50000 [==============================] - 6s 127us/step - loss: 1.0098 - acc: 0.6422 - val_loss: 0.9728 - val_acc: 0.6538\n",
            "Epoch 15/150\n",
            "50000/50000 [==============================] - 6s 128us/step - loss: 0.9881 - acc: 0.6519 - val_loss: 0.9451 - val_acc: 0.6660\n",
            "Epoch 16/150\n",
            "50000/50000 [==============================] - 6s 127us/step - loss: 0.9628 - acc: 0.6604 - val_loss: 0.9135 - val_acc: 0.6812\n",
            "Epoch 17/150\n",
            "50000/50000 [==============================] - 6s 128us/step - loss: 0.9419 - acc: 0.6698 - val_loss: 0.8841 - val_acc: 0.6858\n",
            "Epoch 18/150\n",
            "50000/50000 [==============================] - 6s 128us/step - loss: 0.9204 - acc: 0.6757 - val_loss: 0.8852 - val_acc: 0.6859\n",
            "Epoch 19/150\n",
            "50000/50000 [==============================] - 6s 128us/step - loss: 0.9005 - acc: 0.6847 - val_loss: 0.8486 - val_acc: 0.7022\n",
            "Epoch 20/150\n",
            "50000/50000 [==============================] - 6s 128us/step - loss: 0.8836 - acc: 0.6896 - val_loss: 0.8330 - val_acc: 0.7046\n",
            "Epoch 21/150\n",
            "50000/50000 [==============================] - 6s 129us/step - loss: 0.8619 - acc: 0.6961 - val_loss: 0.8272 - val_acc: 0.7098\n",
            "Epoch 22/150\n",
            "50000/50000 [==============================] - 6s 128us/step - loss: 0.8450 - acc: 0.7045 - val_loss: 0.8077 - val_acc: 0.7131\n",
            "Epoch 23/150\n",
            "50000/50000 [==============================] - 6s 129us/step - loss: 0.8299 - acc: 0.7083 - val_loss: 0.8040 - val_acc: 0.7179\n",
            "Epoch 24/150\n",
            "50000/50000 [==============================] - 6s 127us/step - loss: 0.8185 - acc: 0.7150 - val_loss: 0.7932 - val_acc: 0.7194\n",
            "Epoch 25/150\n",
            "50000/50000 [==============================] - 6s 127us/step - loss: 0.8014 - acc: 0.7190 - val_loss: 0.7725 - val_acc: 0.7301\n",
            "Epoch 26/150\n",
            "50000/50000 [==============================] - 6s 127us/step - loss: 0.7903 - acc: 0.7228 - val_loss: 0.7741 - val_acc: 0.7284\n",
            "Epoch 27/150\n",
            "50000/50000 [==============================] - 6s 127us/step - loss: 0.7770 - acc: 0.7296 - val_loss: 0.7738 - val_acc: 0.7270\n",
            "Epoch 28/150\n",
            "50000/50000 [==============================] - 6s 127us/step - loss: 0.7582 - acc: 0.7346 - val_loss: 0.7445 - val_acc: 0.7435\n",
            "Epoch 29/150\n",
            "50000/50000 [==============================] - 6s 127us/step - loss: 0.7482 - acc: 0.7384 - val_loss: 0.7361 - val_acc: 0.7426\n",
            "Epoch 30/150\n",
            "50000/50000 [==============================] - 6s 128us/step - loss: 0.7352 - acc: 0.7423 - val_loss: 0.7230 - val_acc: 0.7450\n",
            "Epoch 31/150\n",
            "50000/50000 [==============================] - 6s 128us/step - loss: 0.7234 - acc: 0.7475 - val_loss: 0.7180 - val_acc: 0.7497\n",
            "Epoch 32/150\n",
            "50000/50000 [==============================] - 6s 127us/step - loss: 0.7141 - acc: 0.7494 - val_loss: 0.7100 - val_acc: 0.7529\n",
            "Epoch 33/150\n",
            "50000/50000 [==============================] - 6s 128us/step - loss: 0.6990 - acc: 0.7552 - val_loss: 0.7072 - val_acc: 0.7524\n",
            "Epoch 34/150\n",
            "50000/50000 [==============================] - 6s 128us/step - loss: 0.6913 - acc: 0.7581 - val_loss: 0.6931 - val_acc: 0.7577\n",
            "Epoch 35/150\n",
            "50000/50000 [==============================] - 6s 127us/step - loss: 0.6803 - acc: 0.7622 - val_loss: 0.6797 - val_acc: 0.7604\n",
            "Epoch 36/150\n",
            "50000/50000 [==============================] - 6s 127us/step - loss: 0.6690 - acc: 0.7651 - val_loss: 0.6773 - val_acc: 0.7643\n",
            "Epoch 37/150\n",
            "50000/50000 [==============================] - 6s 127us/step - loss: 0.6643 - acc: 0.7674 - val_loss: 0.6888 - val_acc: 0.7578\n",
            "Epoch 38/150\n",
            "50000/50000 [==============================] - 6s 127us/step - loss: 0.6540 - acc: 0.7698 - val_loss: 0.6697 - val_acc: 0.7648\n",
            "Epoch 39/150\n",
            "50000/50000 [==============================] - 6s 128us/step - loss: 0.6366 - acc: 0.7771 - val_loss: 0.6876 - val_acc: 0.7589\n",
            "Epoch 40/150\n",
            "50000/50000 [==============================] - 6s 128us/step - loss: 0.6336 - acc: 0.7767 - val_loss: 0.6556 - val_acc: 0.7705\n",
            "Epoch 41/150\n",
            "50000/50000 [==============================] - 6s 128us/step - loss: 0.6213 - acc: 0.7825 - val_loss: 0.6606 - val_acc: 0.7703\n",
            "Epoch 42/150\n",
            "50000/50000 [==============================] - 6s 127us/step - loss: 0.6140 - acc: 0.7850 - val_loss: 0.6485 - val_acc: 0.7756\n",
            "Epoch 43/150\n",
            "50000/50000 [==============================] - 6s 127us/step - loss: 0.6060 - acc: 0.7866 - val_loss: 0.6439 - val_acc: 0.7737\n",
            "Epoch 44/150\n",
            "50000/50000 [==============================] - 6s 128us/step - loss: 0.5980 - acc: 0.7908 - val_loss: 0.6409 - val_acc: 0.7752\n",
            "Epoch 45/150\n",
            "50000/50000 [==============================] - 6s 128us/step - loss: 0.5900 - acc: 0.7939 - val_loss: 0.6396 - val_acc: 0.7793\n",
            "Epoch 46/150\n",
            "50000/50000 [==============================] - 6s 129us/step - loss: 0.5791 - acc: 0.7967 - val_loss: 0.6289 - val_acc: 0.7816\n",
            "Epoch 47/150\n",
            "50000/50000 [==============================] - 6s 128us/step - loss: 0.5749 - acc: 0.8015 - val_loss: 0.6308 - val_acc: 0.7819\n",
            "Epoch 48/150\n",
            "50000/50000 [==============================] - 6s 130us/step - loss: 0.5675 - acc: 0.8008 - val_loss: 0.6333 - val_acc: 0.7784\n",
            "Epoch 49/150\n",
            "50000/50000 [==============================] - 6s 128us/step - loss: 0.5596 - acc: 0.8037 - val_loss: 0.6323 - val_acc: 0.7835\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7fa4e7588860>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 48
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "PdyXGciMXso-",
        "colab_type": "text"
      },
      "source": [
        "## Model3: Plot losses and/or accuracy"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "g6C05fJmXs2W",
        "colab_type": "code",
        "outputId": "707082c0-05b9-4569-f8b1-be07cc5cd450",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 573
        }
      },
      "source": [
        "plot(model2)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYsAAAEWCAYAAACXGLsWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nWbIwxhjTI0sWxhhjevT/AVzEosghppI1AAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEWCAYAAACJ0YulAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nhYIxxphuFgrGGGO6WSgYY4zp9v8B1Ew+6qWlVMwAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "aFmv3zMCY9qH",
        "colab_type": "text"
      },
      "source": [
        "# Make a prediction on one of the test images. Sample code is below. Modify as needed.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NQaWGqyWqJ7J",
        "colab_type": "code",
        "outputId": "8d274dfa-9053-4aca-ab27-5fbf8f6b566e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 354
        }
      },
      "source": [
        "\n",
        "input_image_index = 0\n",
        "print(x_test[input_image_index,:,:,:].shape)\n",
        "\n",
        "plt.imshow(x_test[input_image_index,:,:,:])\n",
        "plt.show()\n",
        "\n",
        "print(\"y=\", y_test[input_image_index])\n",
        "print()\n",
        "\n",
        "predictions = model3.predict_classes(x_test[input_image_index:])[0:10]\n",
        "# print(predictions)\n",
        "print()\n",
        "\n",
        "for i, prediction in enumerate(predictions):\n",
        "    if i==input_image_index:\n",
        "      print(\"img belogs to the following class index:\",prediction) \n"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(32, 32, 3)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "y= [0. 0. 0. 1. 0. 0. 0. 0. 0. 0.]\n",
            "\n",
            "\n",
            "img belogs to the following class index: 3\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "m8s37TwmQt9e",
        "colab_type": "text"
      },
      "source": [
        "# TODO: Transfer Learning (VGG16)\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Sxg-p8wVqJ7L",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Choose any image classification model you want (ResNEt, MobileNet, VGG, etc) \n",
        "# from ny framework (TensorFlow, Keras, etc.). Find wights for it. Retrain it on the Cifar10 dataset. \n",
        "# You may have to chanege the finaly kayer to output 10 classes.\n",
        "\n",
        "# Using VGG16\n",
        "from keras.applications import VGG16\n",
        "conv_base = VGG16(weights='imagenet',\n",
        "include_top=False,\n",
        "input_shape=(32, 32, 3))\n",
        "\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rDhvZI1z3GiA",
        "colab_type": "code",
        "outputId": "03b4d42a-096b-4804-a206-68ea48befa73",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 782
        }
      },
      "source": [
        "conv_base.summary()\n"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "input_2 (InputLayer)         (None, 32, 32, 3)         0         \n",
            "_________________________________________________________________\n",
            "block1_conv1 (Conv2D)        (None, 32, 32, 64)        1792      \n",
            "_________________________________________________________________\n",
            "block1_conv2 (Conv2D)        (None, 32, 32, 64)        36928     \n",
            "_________________________________________________________________\n",
            "block1_pool (MaxPooling2D)   (None, 16, 16, 64)        0         \n",
            "_________________________________________________________________\n",
            "block2_conv1 (Conv2D)        (None, 16, 16, 128)       73856     \n",
            "_________________________________________________________________\n",
            "block2_conv2 (Conv2D)        (None, 16, 16, 128)       147584    \n",
            "_________________________________________________________________\n",
            "block2_pool (MaxPooling2D)   (None, 8, 8, 128)         0         \n",
            "_________________________________________________________________\n",
            "block3_conv1 (Conv2D)        (None, 8, 8, 256)         295168    \n",
            "_________________________________________________________________\n",
            "block3_conv2 (Conv2D)        (None, 8, 8, 256)         590080    \n",
            "_________________________________________________________________\n",
            "block3_conv3 (Conv2D)        (None, 8, 8, 256)         590080    \n",
            "_________________________________________________________________\n",
            "block3_pool (MaxPooling2D)   (None, 4, 4, 256)         0         \n",
            "_________________________________________________________________\n",
            "block4_conv1 (Conv2D)        (None, 4, 4, 512)         1180160   \n",
            "_________________________________________________________________\n",
            "block4_conv2 (Conv2D)        (None, 4, 4, 512)         2359808   \n",
            "_________________________________________________________________\n",
            "block4_conv3 (Conv2D)        (None, 4, 4, 512)         2359808   \n",
            "_________________________________________________________________\n",
            "block4_pool (MaxPooling2D)   (None, 2, 2, 512)         0         \n",
            "_________________________________________________________________\n",
            "block5_conv1 (Conv2D)        (None, 2, 2, 512)         2359808   \n",
            "_________________________________________________________________\n",
            "block5_conv2 (Conv2D)        (None, 2, 2, 512)         2359808   \n",
            "_________________________________________________________________\n",
            "block5_conv3 (Conv2D)        (None, 2, 2, 512)         2359808   \n",
            "_________________________________________________________________\n",
            "block5_pool (MaxPooling2D)   (None, 1, 1, 512)         0         \n",
            "=================================================================\n",
            "Total params: 14,714,688\n",
            "Trainable params: 14,714,688\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zVdIJZ49EbJU",
        "colab_type": "code",
        "outputId": "167cbb26-839a-45fb-8a11-5fde79301c06",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 918
        }
      },
      "source": [
        "from keras.layers import Dropout\n",
        "from keras.engine import Model\n",
        "\n",
        "last = conv_base.get_layer('block5_pool').output\n",
        "# Add classification layers on top of it\n",
        "x = Flatten()(last)\n",
        "max\n",
        "x = Dropout(0.5)(x)\n",
        "pred = Dense(10, activation='sigmoid')(x)\n",
        "\n",
        "tl_model = Model(conv_base.input, pred)\n",
        "\n",
        "for layer in conv_base.layers:\n",
        "    layer.trainable = False\n",
        "\n",
        "tl_model.compile(loss='binary_crossentropy',\n",
        "              optimizer=SGD(lr=1e-3, momentum=0.9),\n",
        "              metrics=['accuracy'])\n",
        "\n",
        "tl_model.summary()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "input_2 (InputLayer)         (None, 32, 32, 3)         0         \n",
            "_________________________________________________________________\n",
            "block1_conv1 (Conv2D)        (None, 32, 32, 64)        1792      \n",
            "_________________________________________________________________\n",
            "block1_conv2 (Conv2D)        (None, 32, 32, 64)        36928     \n",
            "_________________________________________________________________\n",
            "block1_pool (MaxPooling2D)   (None, 16, 16, 64)        0         \n",
            "_________________________________________________________________\n",
            "block2_conv1 (Conv2D)        (None, 16, 16, 128)       73856     \n",
            "_________________________________________________________________\n",
            "block2_conv2 (Conv2D)        (None, 16, 16, 128)       147584    \n",
            "_________________________________________________________________\n",
            "block2_pool (MaxPooling2D)   (None, 8, 8, 128)         0         \n",
            "_________________________________________________________________\n",
            "block3_conv1 (Conv2D)        (None, 8, 8, 256)         295168    \n",
            "_________________________________________________________________\n",
            "block3_conv2 (Conv2D)        (None, 8, 8, 256)         590080    \n",
            "_________________________________________________________________\n",
            "block3_conv3 (Conv2D)        (None, 8, 8, 256)         590080    \n",
            "_________________________________________________________________\n",
            "block3_pool (MaxPooling2D)   (None, 4, 4, 256)         0         \n",
            "_________________________________________________________________\n",
            "block4_conv1 (Conv2D)        (None, 4, 4, 512)         1180160   \n",
            "_________________________________________________________________\n",
            "block4_conv2 (Conv2D)        (None, 4, 4, 512)         2359808   \n",
            "_________________________________________________________________\n",
            "block4_conv3 (Conv2D)        (None, 4, 4, 512)         2359808   \n",
            "_________________________________________________________________\n",
            "block4_pool (MaxPooling2D)   (None, 2, 2, 512)         0         \n",
            "_________________________________________________________________\n",
            "block5_conv1 (Conv2D)        (None, 2, 2, 512)         2359808   \n",
            "_________________________________________________________________\n",
            "block5_conv2 (Conv2D)        (None, 2, 2, 512)         2359808   \n",
            "_________________________________________________________________\n",
            "block5_conv3 (Conv2D)        (None, 2, 2, 512)         2359808   \n",
            "_________________________________________________________________\n",
            "block5_pool (MaxPooling2D)   (None, 1, 1, 512)         0         \n",
            "_________________________________________________________________\n",
            "flatten_7 (Flatten)          (None, 512)               0         \n",
            "_________________________________________________________________\n",
            "dense_9 (Dense)              (None, 512)               262656    \n",
            "_________________________________________________________________\n",
            "dropout_3 (Dropout)          (None, 512)               0         \n",
            "_________________________________________________________________\n",
            "dense_10 (Dense)             (None, 10)                5130      \n",
            "=================================================================\n",
            "Total params: 14,982,474\n",
            "Trainable params: 267,786\n",
            "Non-trainable params: 14,714,688\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_F5Fsm8HE6fo",
        "colab_type": "code",
        "outputId": "b153f217-fe13-4a11-bafc-5d1815ae706c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 2023
        }
      },
      "source": [
        "tl_model.fit(scaled_x_train, y_train,\n",
        "              batch_size=128,\n",
        "              shuffle=True,\n",
        "              epochs=128,\n",
        "              validation_data=(scaled_x_test, y_test),\n",
        "              callbacks=[EarlyStopping(min_delta=0.001, patience=3)])"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Train on 50000 samples, validate on 10000 samples\n",
            "Epoch 1/128\n",
            "50000/50000 [==============================] - 11s 217us/step - loss: 0.3544 - acc: 0.8888 - val_loss: 0.3136 - val_acc: 0.9000\n",
            "Epoch 2/128\n",
            "50000/50000 [==============================] - 9s 184us/step - loss: 0.3241 - acc: 0.8999 - val_loss: 0.3041 - val_acc: 0.9000\n",
            "Epoch 3/128\n",
            "50000/50000 [==============================] - 9s 185us/step - loss: 0.3140 - acc: 0.8999 - val_loss: 0.2968 - val_acc: 0.9000\n",
            "Epoch 4/128\n",
            "50000/50000 [==============================] - 9s 187us/step - loss: 0.3058 - acc: 0.9000 - val_loss: 0.2908 - val_acc: 0.9000\n",
            "Epoch 5/128\n",
            "50000/50000 [==============================] - 9s 188us/step - loss: 0.2995 - acc: 0.9001 - val_loss: 0.2857 - val_acc: 0.9000\n",
            "Epoch 6/128\n",
            "50000/50000 [==============================] - 9s 187us/step - loss: 0.2945 - acc: 0.9001 - val_loss: 0.2813 - val_acc: 0.9000\n",
            "Epoch 7/128\n",
            "50000/50000 [==============================] - 9s 186us/step - loss: 0.2901 - acc: 0.9002 - val_loss: 0.2775 - val_acc: 0.9001\n",
            "Epoch 8/128\n",
            "50000/50000 [==============================] - 9s 186us/step - loss: 0.2866 - acc: 0.9005 - val_loss: 0.2741 - val_acc: 0.9002\n",
            "Epoch 9/128\n",
            "50000/50000 [==============================] - 9s 185us/step - loss: 0.2830 - acc: 0.9007 - val_loss: 0.2711 - val_acc: 0.9004\n",
            "Epoch 10/128\n",
            "50000/50000 [==============================] - 9s 185us/step - loss: 0.2800 - acc: 0.9008 - val_loss: 0.2684 - val_acc: 0.9006\n",
            "Epoch 11/128\n",
            "50000/50000 [==============================] - 9s 185us/step - loss: 0.2771 - acc: 0.9012 - val_loss: 0.2659 - val_acc: 0.9009\n",
            "Epoch 12/128\n",
            "50000/50000 [==============================] - 9s 186us/step - loss: 0.2749 - acc: 0.9014 - val_loss: 0.2637 - val_acc: 0.9015\n",
            "Epoch 13/128\n",
            "50000/50000 [==============================] - 9s 187us/step - loss: 0.2723 - acc: 0.9017 - val_loss: 0.2617 - val_acc: 0.9018\n",
            "Epoch 14/128\n",
            "50000/50000 [==============================] - 9s 186us/step - loss: 0.2703 - acc: 0.9018 - val_loss: 0.2599 - val_acc: 0.9020\n",
            "Epoch 15/128\n",
            "50000/50000 [==============================] - 9s 186us/step - loss: 0.2685 - acc: 0.9022 - val_loss: 0.2581 - val_acc: 0.9023\n",
            "Epoch 16/128\n",
            "50000/50000 [==============================] - 9s 187us/step - loss: 0.2667 - acc: 0.9026 - val_loss: 0.2565 - val_acc: 0.9026\n",
            "Epoch 17/128\n",
            "50000/50000 [==============================] - 9s 186us/step - loss: 0.2651 - acc: 0.9028 - val_loss: 0.2551 - val_acc: 0.9029\n",
            "Epoch 18/128\n",
            "50000/50000 [==============================] - 9s 187us/step - loss: 0.2638 - acc: 0.9029 - val_loss: 0.2537 - val_acc: 0.9032\n",
            "Epoch 19/128\n",
            "50000/50000 [==============================] - 9s 185us/step - loss: 0.2625 - acc: 0.9031 - val_loss: 0.2524 - val_acc: 0.9035\n",
            "Epoch 20/128\n",
            "50000/50000 [==============================] - 9s 186us/step - loss: 0.2607 - acc: 0.9035 - val_loss: 0.2512 - val_acc: 0.9041\n",
            "Epoch 21/128\n",
            "50000/50000 [==============================] - 9s 186us/step - loss: 0.2594 - acc: 0.9038 - val_loss: 0.2501 - val_acc: 0.9044\n",
            "Epoch 22/128\n",
            "50000/50000 [==============================] - 9s 187us/step - loss: 0.2579 - acc: 0.9041 - val_loss: 0.2490 - val_acc: 0.9047\n",
            "Epoch 23/128\n",
            "50000/50000 [==============================] - 9s 185us/step - loss: 0.2569 - acc: 0.9043 - val_loss: 0.2480 - val_acc: 0.9047\n",
            "Epoch 24/128\n",
            "50000/50000 [==============================] - 9s 185us/step - loss: 0.2562 - acc: 0.9045 - val_loss: 0.2470 - val_acc: 0.9051\n",
            "Epoch 25/128\n",
            "50000/50000 [==============================] - 9s 186us/step - loss: 0.2552 - acc: 0.9047 - val_loss: 0.2461 - val_acc: 0.9055\n",
            "Epoch 26/128\n",
            "50000/50000 [==============================] - 9s 186us/step - loss: 0.2540 - acc: 0.9052 - val_loss: 0.2452 - val_acc: 0.9057\n",
            "Epoch 27/128\n",
            "50000/50000 [==============================] - 9s 186us/step - loss: 0.2529 - acc: 0.9052 - val_loss: 0.2444 - val_acc: 0.9059\n",
            "Epoch 28/128\n",
            "50000/50000 [==============================] - 9s 186us/step - loss: 0.2524 - acc: 0.9055 - val_loss: 0.2436 - val_acc: 0.9063\n",
            "Epoch 29/128\n",
            "50000/50000 [==============================] - 9s 185us/step - loss: 0.2516 - acc: 0.9060 - val_loss: 0.2429 - val_acc: 0.9065\n",
            "Epoch 30/128\n",
            "50000/50000 [==============================] - 9s 187us/step - loss: 0.2504 - acc: 0.9058 - val_loss: 0.2421 - val_acc: 0.9068\n",
            "Epoch 31/128\n",
            "50000/50000 [==============================] - 9s 186us/step - loss: 0.2498 - acc: 0.9059 - val_loss: 0.2414 - val_acc: 0.9072\n",
            "Epoch 32/128\n",
            "50000/50000 [==============================] - 9s 187us/step - loss: 0.2488 - acc: 0.9064 - val_loss: 0.2408 - val_acc: 0.9070\n",
            "Epoch 33/128\n",
            "50000/50000 [==============================] - 9s 186us/step - loss: 0.2487 - acc: 0.9063 - val_loss: 0.2402 - val_acc: 0.9073\n",
            "Epoch 34/128\n",
            "50000/50000 [==============================] - 9s 187us/step - loss: 0.2477 - acc: 0.9067 - val_loss: 0.2395 - val_acc: 0.9075\n",
            "Epoch 35/128\n",
            "50000/50000 [==============================] - 9s 186us/step - loss: 0.2468 - acc: 0.9069 - val_loss: 0.2389 - val_acc: 0.9077\n",
            "Epoch 36/128\n",
            "50000/50000 [==============================] - 9s 186us/step - loss: 0.2460 - acc: 0.9071 - val_loss: 0.2383 - val_acc: 0.9080\n",
            "Epoch 37/128\n",
            "50000/50000 [==============================] - 9s 185us/step - loss: 0.2455 - acc: 0.9072 - val_loss: 0.2378 - val_acc: 0.9082\n",
            "Epoch 38/128\n",
            "50000/50000 [==============================] - 9s 185us/step - loss: 0.2448 - acc: 0.9074 - val_loss: 0.2372 - val_acc: 0.9084\n",
            "Epoch 39/128\n",
            "50000/50000 [==============================] - 9s 187us/step - loss: 0.2448 - acc: 0.9075 - val_loss: 0.2367 - val_acc: 0.9086\n",
            "Epoch 40/128\n",
            "50000/50000 [==============================] - 9s 186us/step - loss: 0.2438 - acc: 0.9078 - val_loss: 0.2362 - val_acc: 0.9086\n",
            "Epoch 41/128\n",
            "50000/50000 [==============================] - 9s 186us/step - loss: 0.2432 - acc: 0.9079 - val_loss: 0.2357 - val_acc: 0.9089\n",
            "Epoch 42/128\n",
            "50000/50000 [==============================] - 9s 185us/step - loss: 0.2425 - acc: 0.9078 - val_loss: 0.2352 - val_acc: 0.9091\n",
            "Epoch 43/128\n",
            "50000/50000 [==============================] - 9s 185us/step - loss: 0.2421 - acc: 0.9084 - val_loss: 0.2347 - val_acc: 0.9094\n",
            "Epoch 44/128\n",
            "50000/50000 [==============================] - 9s 185us/step - loss: 0.2413 - acc: 0.9083 - val_loss: 0.2343 - val_acc: 0.9094\n",
            "Epoch 45/128\n",
            "50000/50000 [==============================] - 9s 186us/step - loss: 0.2410 - acc: 0.9084 - val_loss: 0.2338 - val_acc: 0.9097\n",
            "Epoch 46/128\n",
            "50000/50000 [==============================] - 9s 185us/step - loss: 0.2409 - acc: 0.9085 - val_loss: 0.2334 - val_acc: 0.9098\n",
            "Epoch 47/128\n",
            "50000/50000 [==============================] - 9s 186us/step - loss: 0.2402 - acc: 0.9089 - val_loss: 0.2330 - val_acc: 0.9097\n",
            "Epoch 48/128\n",
            "50000/50000 [==============================] - 9s 185us/step - loss: 0.2398 - acc: 0.9089 - val_loss: 0.2326 - val_acc: 0.9099\n",
            "Epoch 49/128\n",
            "50000/50000 [==============================] - 9s 186us/step - loss: 0.2392 - acc: 0.9093 - val_loss: 0.2322 - val_acc: 0.9100\n",
            "Epoch 50/128\n",
            "50000/50000 [==============================] - 9s 185us/step - loss: 0.2389 - acc: 0.9090 - val_loss: 0.2318 - val_acc: 0.9101\n",
            "Epoch 51/128\n",
            "50000/50000 [==============================] - 9s 185us/step - loss: 0.2385 - acc: 0.9090 - val_loss: 0.2314 - val_acc: 0.9102\n",
            "Epoch 52/128\n",
            "50000/50000 [==============================] - 9s 184us/step - loss: 0.2380 - acc: 0.9094 - val_loss: 0.2311 - val_acc: 0.9103\n",
            "Epoch 53/128\n",
            "50000/50000 [==============================] - 9s 186us/step - loss: 0.2376 - acc: 0.9096 - val_loss: 0.2307 - val_acc: 0.9105\n",
            "Epoch 54/128\n",
            "50000/50000 [==============================] - 9s 185us/step - loss: 0.2370 - acc: 0.9098 - val_loss: 0.2303 - val_acc: 0.9109\n",
            "Epoch 55/128\n",
            "50000/50000 [==============================] - 9s 186us/step - loss: 0.2369 - acc: 0.9097 - val_loss: 0.2300 - val_acc: 0.9109\n",
            "Epoch 56/128\n",
            "50000/50000 [==============================] - 9s 186us/step - loss: 0.2366 - acc: 0.9099 - val_loss: 0.2297 - val_acc: 0.9110\n",
            "Epoch 57/128\n",
            "50000/50000 [==============================] - 9s 184us/step - loss: 0.2360 - acc: 0.9101 - val_loss: 0.2293 - val_acc: 0.9110\n",
            "Epoch 58/128\n",
            "50000/50000 [==============================] - 9s 186us/step - loss: 0.2355 - acc: 0.9103 - val_loss: 0.2290 - val_acc: 0.9112\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7fa61828a780>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 28
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "U_RAixYPFpDo",
        "colab_type": "code",
        "outputId": "a037a51b-fee4-4c3f-ce90-0a3f50bcf6be",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        }
      },
      "source": [
        "plot(tl_model)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEWCAYAAACJ0YulAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "loC5wU4rHIBv",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}